(Methods in Molecular Biology 1257) Willem F. Wolkers, Harriëtte Oldenhof (Eds.) - Cryopreservation and Freeze-Drying Protocols-Springer-Verlag New York (2015)

Methods in
Molecular Biology 1257
Willem F. Wolkers
Harritte Oldenhof Editors
Cryopreservation
and Freeze-Drying
Protocols
Third Edition
METHODS IN M O L E C U L A R B I O LO G Y
Series Editor
John M. Walker
School of Life Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK
For further volumes:

http://www.springer.com/series/7651
Cryopreservation
and Freeze-Drying Protocols
Third Edition
Edited by
Willem F. Wolkers
Institute of Multiphase Processes, Leibniz Universitt Hannover, Hannover, Germany
Harritte Oldenhof
Unit for Reproductive Medicine, Clinic for Horses, University of Veterinary
Medicine Hannover, Hannover, Germany
Editors
Willem F. Wolkers Harritte Oldenhof
Institute of Multiphase Processes Unit for Reproductive Medicine, Clinic
Leibniz Universitt Hannover for Horses
Hannover, Germany University of Veterinary
Medicine Hannover
Hannover, Germany
ISSN 1064-3745 ISSN 1940-6029 (electronic)

ISBN 978-1-4939-2192-8 ISBN 978-1-4939-2193-5 (eBook)
DOI 10.1007/978-1-4939-2193-5
Springer New York Heidelberg Dordrecht London
Library of Congress Control Number: 2014955222
Springer Science+Business Media New York 2015

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this
legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for
the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher's location, in its current version, and permission for use must always be obtained from Springer. Permissions
for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution
under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication, neither
the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be
made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
Printed on acid-free paper
Humana Press is a brand of Springer

Springer is part of Springer Science+Business Media (www.springer.com)
Dedication
In memory of Stanley P. Leibo

Preface
Cryopreservation and freeze-drying are widely used for long-term storage of biological
materials. Methods to safely store specimens in a stable state for extended periods have
widespread applications in medicine and agriculture. Using cryopreservation for cell and
tissue banking allows samples to be available at the time of need. Ice-free cryopreservation,
referred to as vitrification, is receiving increased attention in biobanking. Whereas cryopre-
served samples are typically stored in liquid nitrogen, dried specimens can be stored at
room temperature, which has clear advantages for banking and transport. Freeze-drying
involves a freezing and a drying step and hence requires protectants that protect during
both freezing and drying.
A variety of biological materials can be cryopreserved using relatively easy and stan-
dard protocols. There are cases, however, which require custom-designed cryopreser-
vation strategies including the use of nonconventional cryoprotective agents, optimized
cooling and warming rates, selection and cleanup processing, and modification of cel-
lular properties. Drying of cells or biomolecular assemblies is generally more damaging
compared to cryopreservation, and requires addition of lyoprotectants. Freeze-drying
is widely used to stabilize biomolecules and macromolecular assemblies and has been
implicated as a method to preserve mammalian cells in a dry state. Cryopreservation
and dry preservation are highly interdisciplinary fields of research requiring insights
from biologists, chemists, physicists, as well as engineers to find rationally designed
preservation solutions for individual cases.
In this edition of the book Cryopreservation and Freeze-Drying Protocols we not
only aimed to provide a variety of standard protocols that can be used to cryopreserve or
freeze-dry different types of specimens, but also wanted to highlight methods that can be
used to obtain insights in cellular and macromolecular changes in response to freezing or
drying that can be used to rationally design preservation protocols. The book is divided
into four parts. Part I handles fundamental principles of cryopreservation, vitrification,
freeze-drying, and the use of mathematical modeling to design preservation protocols. In
Part II, microscopic, spectroscopic, as well as calorimetric methods are presented to study
cell and molecular behavior during freezing and drying, as well as thermodynamic proper-
ties of preservation solutions. In Part III, cryopreservation and vitrification approaches are
presented for a wide variety of samples including sperm, oocytes, blastocysts, mammalian
and plant cell lines, stem cells, blood cells, and tissues. In addition, various preparative pro-
cessing methods such as cleanup procedures and membrane modification strategies are
presented. In Part IV, freeze-drying methods are described for proteins, bacteria, sperm,
and extracellular tissue matrices.
vii
viii Preface
The book aims to serve as a practical guideline that can be used without the need of
other reference sources. In addition to protocols that rely on the use of specialized equip-
ment, practical and cheaper alternatives are also described. Our intended readers are
researchers and technical assistants in academia and industry with a background in life
sciences or engineering who want to investigate freezing and drying processes or set up
methods to safely store biological material while maintaining its function upon
reconstitution.
Hannover, Germany Willem F. Wolkers

Hannover, Germany Harritte Oldenhof
Contents
Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
PART I FUNDAMENTAL ASPECTS AND FREEZING TECHNOLOGY

1 Principles of Cryopreservation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
David E. Pegg
2 Principles of Cryopreservation by Vitrification. . . . . . . . . . . . . . . . . . . . . . . . . 21
Gregory M. Fahy and Brian Wowk
3 Modeling and Optimization of Cryopreservation . . . . . . . . . . . . . . . . . . . . . . 83
James D. Benson
4 The Principles of Freeze-Drying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Gerald D.J. Adams, Isobel Cook, and Kevin R. Ward
PART II METHODS TO STUDY FREEZING AND DRYING PROCESSES

5 Use of In Situ Fourier Transform Infrared Spectroscopy
to Study Freezing and Drying of Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Willem F. Wolkers and Harritte Oldenhof
6 Calorimetric Analysis of Cryopreservation
and Freeze-Drying Formulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Wendell Q. Sun
7 Measurement of Intracellular Ice Formation Kinetics
by High-Speed Video Cryomicroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Jens O.M. Karlsson
8 Laser Scanning Microscopy in Cryobiology . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Frank Stracke, Asger Kreiner-Mller, and Heiko Zimmermann
9 Low-Temperature Electron Microscopy: Techniques and Protocols . . . . . . . . . 243
Roland A. Fleck
PART III CRYOPRESERVATION PROTOCOLS

10 Cryopreservation of Semen from Domestic Livestock . . . . . . . . . . . . . . . . . . . 277
Harald Sieme and Harritte Oldenhof
11 Cryopreservation of Mammalian Oocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Victoria Keros and Barry J. Fuller
12 Vitrification: A Simple and Successful Method for Cryostorage
of Human Blastocysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Juergen Liebermann
ix
x Contents
13 Efficient Cryopreservation of Human Pluripotent

Stem Cells by Surface-Based Vitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Julia C. Neubauer, Axel F. Beier, Niels Geijsen,
and Heiko Zimmermann
14 Cryopreservation of Greenshell Mussel (Perna canaliculus) Sperm . . . . . . . . 329
Serean L. Adams, John F. Smith, Jolene Taylor,
Lindsay T. McGowan, and H. Robin Tervit
15 Membrane Modification Strategies for Cryopreservation . . . . . . . . . . . . . . . . . 337
Phillip H. Purdy and James K. Graham
16 Sperm Cleanup and Centrifugation Processing for Cryopreservation . . . . . . . . 343
17 Cryopreservation of Red Blood Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Johan W. Lagerberg
18 Cord Blood Clinical Processing, Cryopreservation, and Storage . . . . . . . . . . . 369
Heidi Elmoazzen and Jelena L. Holovati
19 Directional Freezing for Large Volume Cryopreservation . . . . . . . . . . . . . . . . 381
Joseph Saragusty
20 Vitrification of Heart Valve Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Kelvin G.M. Brockbank, Zhenzhen Chen, Elizabeth D. Greene,
and Lia H. Campbell
21 Cryopreservation of Plant Cell Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
Heinz Martin Schumacher, Martina Westphal,
and Elke Heine-Dobbernack
22 Writing Standard Operating Procedures (SOPs)
for Cryostorage Protocols: Using Shoot Meristem
Cryopreservation as an Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Keith Harding and Erica E. Benson
PART IV FREEZE-DRYING PROTOCOLS

23 Freeze-Drying of Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Baolin Liu and Xinli Zhou
24 Freeze-Drying of Lactic Acid Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
Fernanda Fonseca, Stphanie Cenard, and Stphanie Passot
25 Freeze-Drying of Mammalian Sperm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
Levent Keskintepe and Ali Eroglu
26 Freeze-Drying of Decellularized Heart Valve Tissues. . . . . . . . . . . . . . . . . . . . 499
Willem F. Wolkers and Andres Hilfiker
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
Contributors
GARY D. J. ADAMS Deceased

SEREAN L. ADAMS Cawthron Institute, Nelson, New Zealand
AXEL F. BEIER Fraunhofer Institute for Biomedical Technology, St. Ingbert, Germany;
Hubrecht Institute, Utrecht, The Netherlands
JAMES D. BENSON Department of Mathematical Sciences, Northern Illinois University,
DeKalb, IL, USA
ERICA E. BENSON Damar Research Scientists, Damar, Fife, Scotland, UK
KELVIN G.M. BROCKBANK Cell and Tissue Systems, Inc., North Charleston, SC, USA;
Institute for Bioengineering and Bioscience, Georgia Institute of Technology, Atlanta,
GA, USA; Department of Regenerative Medicine and Cell Biology, Medical University
of South Carolina, Charleston, SC, USA
LIA H. CAMPBELL Cell and Tissue Systems, Inc., North Charleston, SC, USA
STPHANIE CENARD Institut National de la Recherche Agronomique, Gnie et
Microbiologie des Procds Alimentaires, Thiverval-Grignon, France; AgroParisTech,
Gnie et Microbiologie des Procds Alimentaires, Thiverval-Grignon, France
ZHENZHEN CHEN Cell and Tissue Systems, Inc., North Charleston, SC, USA
ISOBEL COOK Biopharma Technology Ltd, Winchester, UK
HEIDI ELMOAZZEN Canadian Blood Services National Public Cord Blood Bank,
Ottawa, ON, Canada
ALI EROGLU Department of Medicine, Institute of Molecular Medicine and Genetics,
Medical College of Georgia, Georgia Regents University, Augusta, GA, USA;
Department of Obstetrics and Gynecology, Medical College of Georgia,
Georgia Regents University, Augusta, GA, USA
GREGORY M. FAHY 21st Century Medicine, Inc., Fontana, CA, USA
ROLAND A. FLECK Centre for Ultrastructural Imaging, Kings College London,
London, UK
FERNANDA FONSECA Institut National de la Recherche Agronomique,
Gnie et Microbiologie des Procds Alimentaires, Thiverval-Grignon, France;
AgroParisTech, Gnie et Microbiologie des Procds Alimentaires, Thiverval-Grignon,
France
BARRY J. FULLER University College London Medical School, London, UK
NIELS GEIJSEN Hubrecht Institute, Utrecht, The Netherlands
JAMES K. GRAHAM Department of Biomedical Sciences, Colorado State University,
Fort Collins, CO, USA
ELIZABETH D. GREENE Cell and Tissue Systems, Inc., North Charleston, SC, USA
KEITH HARDING Damar Research Scientists, Damar, Fife, Scotland, UK
ELKE HEINE-DOBBERNACK Plant Cell Culture Department, Leibniz-Institut,
German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany
xi
xii Contributors
ANDRES HILFIKER Leibniz Research Laboratories for Biotechnology and Artificial Organs,
Hannover Medical School, Hannover, Germany
JELENA L. HOLOVATI Department of Laboratory Medicine and Pathology, University
of Alberta, Edmonton, AB, Canada; Comprehensive Tissue Centre, Alberta Health
Services, Edmonton, AB, Canada
JENS O.M. KARLSSON Department of Mechanical Engineering, Villanova University,
Villanova, PA, USA
VICTORIA KEROS Department of Medicine, Centre for Andrology and Sexual Medicine,
Karolinska Institutet, Stockholm, Sweden; Reproductive Medicine, Karolinska University
Hospital Huddinge, Stockholm, Sweden
LEVENT KESKINTEPE School of Medicine, Sher Institute for Reproductive Medicine
and University of Nevada, Las Vegas, NV, USA
ASGER KREINER-MLLER Fraunhofer Institute for Biomedical Technology, St. Ingbert,
Germany
JOHAN W. LAGERBERG Department of Blood Cell Research, Sanquin Research,
Amsterdam, The Netherlands
JUERGEN LIEBERMANN Fertility Centers of Illinois, Chicago, IL, USA
BAOLIN LIU School of Medical Instrument and Food Engineering,
Institute of Biothermal Science, Shanghai, China
LINDSAY T. MCGOWAN AgResearch Ltd, Hamilton, New Zealand
JULIA C. NEUBAUER Fraunhofer Institute for Biomedical Technology, St. Ingbert,
Germany
HARRITTE OLDENHOF Unit for Reproductive Medicine, Clinic for Horses,
University of Veterinary Medicine Hannover, Hannover, Germany
STPHANIE PASSOT Institut National de la Recherche Agronomique, Gnie et
Microbiologie des Procds Alimentaires, Thiverval-Grignon, France; AgroParisTech,
Gnie et Microbiologie des Procds Alimentaires, Thiverval-Grignon, France
DAVID E. PEGG Department of Biology, University of York, York, UK
PHILLIP H. PURDY United States Department of Agriculture, Agricultural Research
Service, National Animal Germplasm Program, National Center for Genetic Resources
Preservation, Fort Collins, CO, USA
JOSEPH SARAGUSTY Department of Reproduction Management, Leibniz Institute
for Zoo and Wildlife Research, Berlin, Germany
HEINZ MARTIN SCHUMACHER Plant Cell Culture Department, Leibniz-Institut,
German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany
HARALD SIEME Unit for Reproductive Medicine, Clinic for Horses, University
of Veterinary Medicine Hannover, Hannover, Germany
JOHN F. SMITH Cawthron Institute, Nelson, New Zealand
FRANK STRACKE Fraunhofer Institute for Biomedical Technology, St. Ingbert, Germany
WENDELL Q. SUN Department of Electronic Science and Technology, School of
Information Science and Technology, University of Science and Technology of China,
Hefei, China; School of Medical Instruments and Food EngineeringUniversity of
Shanghai for Science and Technology, Shanghai, China
JOLENE TAYLOR Cawthron Institute, Nelson, New Zealand
H. ROBIN TERVIT Cawthron Institute, Nelson, New Zealand
KEVIN R. WARD Biopharma Technology Ltd, Winchester, UK
Contents xiii
MARTINA WESTPHAL Plant Cell Culture Department, Leibniz-Institut, German

Collection of Microorganisms and Cell Cultures, Braunschweig, Germany
WILLEM F. WOLKERS Institute of Multiphase Processes, Leibniz Universitt Hannover,
Hannover, Germany
BRIAN WOWK 21st Century Medicine, Inc., Fontana, CA, USA
XINLI ZHOU School of Medical Instrument and Food Engineering,
Institute of Biothermal Science, Shanghai, China
HEIKO ZIMMERMANN Fraunhofer Institute for Biomedical Technology, St. Ingbert,
Germany; Chair of Molecular and Cellular Biotechnology/NanotechnologySaarland
University, Saarbrcken, Germany
Part I
Fundamental Aspects and Freezing Technology

Chapter 1
Principles ofCryopreservation
DavidE.Pegg
Abstract
Cryopreservation is the use of very low temperatures to preserve structurally intact living cells and tissues.
Unprotected freezing is normally lethal and this chapter seeks to analyze some of the mechanisms involved
and to show how cooling can be used to produce stable conditions that preserve life. The biological effects
of cooling are dominated by the freezing of water, which results in the concentration of the solutes that are
dissolved in the remaining liquid phase. Rival theories of freezing injury have envisaged either that ice crystals
pierce or tease apart the cells, destroying them by direct mechanical action, or that damage is from secondary
effects via changes in the composition of the liquid phase. Cryoprotectants, simply by increasing the total
concentration of all solutes in the system, reduce the amount of ice formed at any given temperature; but to
be biologically acceptable they must be able to penetrate into the cells and have low toxicity. Many com-
pounds have such properties, including glycerol, dimethyl sulfoxide, ethanediol, and propanediol.
In fact, both damaging mechanisms are important, their relative contributions depending on cell type,
cooling rate, and warming rate. A consensus has developed that intracellular freezing is dangerous, whereas
extracellular ice is harmless. If the water permeability of the cell membrane is known it is possible to predict
the effect of cooling rate on cell survival and the optimum rate will be a trade-off between the risk of intra-
cellular freezing and effects of the concentrated solutes. However, extracellular ice is not always innocuous:
densely packed cells are more likely to be damaged by mechanical stresses within the channels where they
are sequestered and with complex multicellular systems it is imperative not only to secure cell survival but
also to avoid damage to the extracellular structure. Ice can be avoided by vitrificationthe production of
a glassy state that is defined by the viscosity reaching a sufficiently high value (~1013 poises) to behave like
a solid, but without any crystallization. Toxicity is the major problem in the use of vitrification methods.
Whether freezing is permitted (conventional cryopreservation) or prevented (vitrification), the cryo-
protectant has to gain access to all parts of the system. However, there are numerous barriers to the free
diffusion of solutes (membranes), and these can result in transient, and sometimes equilibrium, changes in
compartment volumes and these can be damaging. Hence, the processes of diffusion and osmosis have
important effects during the introduction of cryoprotectants, the removal of cryoprotectants, the freezing
process, and during thawing. These phenomena are amenable to experiment and analysis, and this has
made it possible to develop effective methods for the preservation of a very wide range of cells and some
tissues; these methods have found widespread applications in biology and medicine.
Key words Cryopreservation, Cryoprotectants, Intracellular freezing, Solution effects, Supercooling,

Vitrification
Willem F. Wolkers and Harritte Oldenhof (eds.), Cryopreservation and Freeze-Drying Protocols, Methods in Molecular Biology,
vol. 1257, DOI10.1007/978-1-4939-2193-5_1, Springer Science+Business Media New York 2015
3
4 DavidE.Pegg
1 Freezing Injury andCryoprotection
1.1 The Discovery Writers of science fiction have been greatly attracted by the con-
oftheCryoprotective cept of suspended animation, whereby the biochemistry of life
Effect ofGlycerol could be reversibly suspended for long periods of time and then
restored. Although such phenomena do occur in nature, though
rarely, it is unfortunately a fact that freezing is normally lethal. In
order to understand the effects of very low temperatures, we have
to recognize that many structures and processes are temperature
dependent and, consequently, cooling has extraordinarily complex
effects that produce conditions that are far removed from normal
physiology. When we cool below 0C the biological effects are
dominated by the freezing of water, which typically constitutes at
least 80% of the tissue mass. Freezing is the conversion of liquid
water to crystalline ice, which results in the concentration of dis-
solved solutes in the remaining liquid phase and the precipitation
of any solutes that exceed their solubility limit. It was not until
1948 that a general method was discovered that permitted the
freezing of many types of animal cells with subsequent restoration
of structure and function. In 1949, Polge, Smith, and Parkes pub-
lished their landmark paper [1] in which they showed that the
inclusion of 1020% of glycerol enabled the spermatozoa of the
cock to survive prolonged freezing at 80C.The theories of
freezing injury then extant envisaged ice crystals piercing or teas-
ing apart the cells and intracellular structures, destroying them by
direct mechanical action.
Glycerol, simply by increasing the total solute concentration,
would reduce the amount of ice formed in the same way that anti-
freeze (ethanediol) reduces the amount of ice forming in the cool-
ing system of an automobile engine. But it was also recognized
very early on that one effect of freezing an aqueous solution was to
increase the concentration of solutes in the dwindling volume of
the remaining solution, and that this could be a fundamental cause
of injury. In a series of classical papers published in the 1950s,
James Lovelock [2, 3] provided strong evidence that salt concen-
tration, rather than ice, is the cause of freezing injury to cells, and
that glycerol protects against this damage only to the extent that it
modulates the rise in salt concentration during freezing. It follows
that the effectiveness of glycerol, or of any similar cryoprotectant,
depends on a number of properties: (1) the compound must be
highly soluble in water and remain so at low temperatures in order
to produce a profound depression of the freezing temperature; (2)
it must be able to penetrate into the cells; and (3) it must have a
low toxicity so that it can be used in the high concentrations that
are required to produce these effects. Many compounds have these
properties. Those in common use include glycerol, dimethyl sulf-
oxide, ethanediol, and propanediol.
Principles of Cryopreservation 5
1.2 The Effect That degree of understanding provided a starting point for the
ofRate ofChange development of practical freeze-preservation techniques for a range
ofTemperature of cells, but it soon became clear that reality was considerably more
complex. First cooling rate, and then warming rate, were found to
be important determinants of survival and Lovelocks theories did
not account for such kinetic effects. In 1963, Mazur discovered
that the rate of change of temperature was important because it
controlled the transport of water across the cell membrane, and
hence, indirectly, the probability of intracellular freezing [4]. In
general, intracellular freezing is lethal. Mazur argued that the rate
of cooling controls the rate at which water is converted to ice;
hence it controls the rate at which the concentration of the solu-
tion surrounding the cells changes; therefore, by controlling the
osmolality of the surrounding fluid, the rate of change of tempera-
ture also influences the rate at which water is transported out of the
cells during cooling and into the cells during warming. Providing
water can leave the cells rapidly to maintain thermodynamic equi-
librium across the cell membrane, the cytoplasm will not cool
below its freezing point (supercool), and all the ice will be external
to the cells. On the other hand, if the cooling rate is too rapid for
the membrane of the cell in question to transport sufficient water
out of the cell, then the protoplasm will become supercooled, and
the greater the extent of supercooling, the more likely is the cell to
freeze internally (Fig.1). The combination of these two factors,
Fig. 1 Schematic representation of cells being cooled rapidly, and freezing internally or sufficiently slowly to
lose water and avoid intracellular ice
6 DavidE.Pegg
Fig. 2 The effect of cooling rate on the survival following freezing of four types of cell. (Based on ref. 5)
solution effects and intracellular freezing, causes each cell to show

maximal survival at a characteristic cooling rate; as the cooling rate
increases from very low rates so does survival because the deleteri-
ous effects of exposure to high salt concentrations are reduced, but
eventually survival drops off because intracellular freezing super-
venes. Each cell has an optimum cooling rate (Fig.2), although
absolute survival is usually extremely low unless a cryoprotectant is
present to reduce the damage at low cooling rates. Cryoprotectants,
like glycerol, have the effect of reducing the solution effects, result-
ing in a lower optimal cooling rate and an increase in the maximum
survival obtained (Fig.3). We will now examine these mechanisms
of cryoinjury and cryoprotection in a little more detail.
1.3 Solution Effects Lovelock [2] had actually shown that the extent of hemolysis that
occurred when a saline suspension of erythrocytes was cooled to
and thawed from a given subzero temperature was similar to that
suffered by a cell suspension that was exposed to the concentration
of sodium chloride produced by freezing to that temperature and
then returning to isotonic saline (Fig.4). Lovelock also demon-
strated [3] that when glycerol was present, hemolysis started at the
(lower) temperature at which the same critical concentration of salt
was produced (Fig.5). Correlation does not prove causation, but
in this case, if the solution changes were not causative of freezing
injury, then the correspondence would be a remarkable coinci-
dence indeed. It was these studies that led to the consensus that
extracellular ice is harmless to cells and that freezing injury is
caused by indirect effects of the formation of ice.
Fig. 3 The effect of cooling rate on the cryopreservation of mouse hemopoietic stem cells cooled in the pres-
ence of the indicated molar concentrations of glycerol. (Reprinted with permission from ref. 6)
Fig. 4 Human erythrocytes were frozen to the indicated temperatures and then thawed (triangles) compared
with exposure to equivalent salt concentrations and then being returned to isotonic conditions (circles).
(Reprinted with permission from ref. 7)
However, the salt in the suspending medium is not the only

solute to be concentratedthe cryoprotectant is concentrated to
the same degree. Pegg and Diaper [9] showed that red blood cells
8 DavidE.Pegg
Fig. 5 The increase in mole fraction of NaCl in solutions that have the indicated molality
of glycerol and are isotonic with respect to NaCl. Five percent hemolysis was observed
at a mole fraction of NaCI=0.0160.021. (Reprinted with permission from ref. 8)
actually suffer more damage when exposed to a given salt concen-

tration in the presence of glycerol than in its absence, and this
effect is dependent on the concentration of glycerol (Fig.6).
When red blood cells were frozen and thawed in the presence
of a range of concentrations of glycerol, they demonstrated that
the correspondence between the effects of salt exposure and of
freezing was retained. This observation is important for two rea-
sons: it shows that cryoprotectants are not innocuousoverall
they are protective but at a price; second, the observation adds
powerful support to the solution-effect theory.
1.4 Intracellular If the water permeability of the cell membrane is known, and the
Freezing temperature coefficient of water permeability can be estimated,
then it is possible to predict the effect of cooling rate on cell sur-
vival (Fig.7). The calculated degree of supercooling for different
rates of cooling shows that intracellular freezing is unlikely at
1C/min but is highly probable at a cooling rate of 10C/min.
For hepatocytes a cooling rate around 1C/min will essentially
eliminate the risk of intracellular freezing and faster cooling will be
preferred only if solution effects are a problem. The optimum rate
will be a trade-off between those two factors. Of course, other cells
have different water permeabilities and it has been shown by direct
experiment that the cooling rate that produces intracellular freez-
ing on a cryomicroscope corresponds with the cooling rate that
produces significant intracellular supercooling ([11]; Fig.8).
Fig. 6 Hemolysis observed when human erythrocytes were frozen to temperatures that produce the indicated
concentrations of NaCl, compared with exposure to equivalent salt concentrations followed by return to iso-
tonic conditions. The R values are the weight ratio of glycerol to NaCl in each solution. (Reprinted with permis-
sion from ref. 7)
Fig. 7 The calculated effect of cooling rate on the volume of hepatocytes and the extent of supercooling of the
cell contents. (a) Relative volume (V/Vo) of cells cooled at the indicated rates (C/min). The line labeled 0 is the
equilibrium line. (b) From the same calculations as in (a), the calculated degree of supercooling of the cell
contents at the indicated cooling rates. At 10C/min the cells are supercooled by 10C and therefore likely to
freeze internally. (Reproduced with permission from ref. 10)
In fact, very small amounts of intracellular ice are compatible

with recovery, and this is one reason why the warming rate has a
profound effect. The behavior of very small intracellular ice crystals
differs between slow and rapid warming: slow warming allows the
10 DavidE.Pegg
Fig. 8 The survival of three types of cells plotted against cooling rate and corre-
lated with the observed occurrence of intracellular freezing. (Reproduced with
permission from ref. 11)
crystals to recrystallize, to coalesce, and to grow. This has been

demonstrated to damage the cells in which it occurs; however, dur-
ing rapid warming there is insufficient time for this to happen and
the ice simply melts. Because the cooling rate influences the forma-
tion of intracellular ice, while warming rate controls what happens
to that ice subsequently, and because cells differ in their water per-
meability and probably also in their susceptibility to intracellular
ice, then it follows that cells will differ in their cooling and warm-
ing requirements and cooling rate will interact with warming rate.
1.5 The Cell Most studies of freezing injury have been carried out with rela-
PackingEffect tively dilute cell suspensions, whereas the cells are quite densely
packed in some systems that have to be preserved, for example red
blood cells for transfusion, and particularly in tissues and organs.
Experiment has shown that the proportion of red blood cells sus-
pended in 2.5M glycerol solution that are hemolyzed during freez-
ing and thawing is strongly dependent on hematocrit (the
percentage of cells by volume) when the hematocrit exceeds 50%.
The increase in hemolysis as the hematocrit is increased is amelio-
rated by increasing the glycerol concentration. At 2M glycerol
concentration, hemolysis is inversely dependent on warming rate
when the cooling rate is less than 1,000C/min and is directly
dependent on cooling rates at higher cooling rates [12].
These observations cannot be accounted for by the classical
mechanisms of cryoinjurysolution effects and intracellular freez-
ing. The most likely explanation is that densely packed cells are
more likely to be damaged by mechanical stresses when the chan-
nels within which they are sequestered change shape. This is a
result of recrystallization of the ice that forms their boundaries.
2 Cyroprotection
2.1 General Cryoprotection usually involves treatment of the cells or tissues

Considerations with cryoprotectant solutes, often in high concentration, and this
produces a driving force for the movement of water by osmosis and
of solutes by diffusion. Freezing involves changes in the concentra-
tion and composition of aqueous solutions and this also produces
driving forces for the movement of water and solutes. Biological
systems contain numerous barriers to the free diffusion of solutes
(membranes), and these can result in transient, and sometimes
equilibrium, changes in compartment volumes; if excessive, these
changes can be damaging. Hence, the processes of diffusion and
osmosis are very important for cryopreservation. Fortunately, the
quantitative description of mass transfer processes is well devel-
oped [13, 14]. The driving force for flow is pressure. Thus, the
flow of water, Jv, through a membrane is given by
J v = kP (1)
where k is a constant that is characteristic of the membrane-and-

water combination and P is the pressure difference across the filter.
J is given the subscript v to signify volumetric flux. When the driv-
ing force for the flow of water through a membrane is osmotic
pressure rather than hydrostatic pressure, flow can be described by
the same equation if osmotic pressure, , is substituted for hydro-
static pressure, thus,
J v = kp (2)
The constant k has the same value in the two equations, providing
only that the membrane and the solvent, water in this case, remain
the same. Thus, both hydrostatic and osmotic pressure differences
can be incorporated into a single equation. The constant k is then
known as the hydraulic conductivity Lp (the units are cm/satm).
J v = L p (P + p ) (3)

Under the conditions prevailing in cryobiology the hydrostatic
pressure term will normally be zero and can be calculated from
concentration by multiplication by the product of the universal gas
constant R, and the absolute temperature T. If the area of the
membrane is A, and internal and external osmolalities are denoted
by Ci and Ce, then we obtain
J v = L p ART (Ci - Ce ) (4)

The solute flux Js is described by
J s = ws ART (Se - Si ) (5)

12 DavidE.Pegg
This equation states that flux across unit area of membrane is

proportional to the solute permeability s, and the difference in con-
centration of the solute across the membrane. The more familiar
solute permeability Ps (units are cm/s) is equal to sRT. The con-
stant RT is 23,235atm/cm3mol. The convention for the direction
of flux is that outsideinside is positive. A somewhat more complex
formalism was elaborated by Kedem and Katchalsky [13] in 1958
and their equations are often used in cryobiology where they are
usually referred to as the KK equations. Kedem and Katchalsky
assumed that the solvent and solute used a common channel through
the membrane and they therefore added a solvent/solute interaction
term, , known as the reflection coefficient. This led to modification
of the equations for Jv and Js, as shown
J v = L p ART (Ci - Ce ) + s (Cip - Cep ) (6)

J s = ws ART (Se - Si ) + J v (1 - s ) c s (7)

In the equation for Jv the solutes are partitioned between impermeant
solutes (C) and permeating solutes (Cp), the latter interacting with
water in the common flow channel. The equation for Js has an addi-
tional term that represents solvent drag on the permeant solute, which
is present in the membrane at concentration cs. Clearly, the KK is
more complex and curve fitting routines can lead to uncertain results
because of the lack of independence of the parameter, . Kleinhans
[14] has discussed these problems in detail and moreover he has
argued that the KK formalism is often invalid because of the presence
of separate channels for water and solute. In practice, the simpler for-
malism is adequate for the current needs of cryobiologists. The two
equations are solved simultaneously by numerical methods and pro-
grams to carry out these calculations can be run on an ordinary PC.
Several methods are available for the determination of perme-
ability parameters in cryobiology. If the solute under study can be
radiolabeled, the time-course of isotope uptake is easily measured
but the calculation of concentration requires the additional
measurement of water content at each time-point. Permeating sol-
utes can be extracted after known times of exposure and high per-
formance liquid chromatography methods are often suitable for
their assay. The Karl Fischer method, using a back titration scheme,
is a convenient method for water [15]. If the compound under
study has a distinctive nuclear magnetic resonance spectrum,
nuclear magnetic resonance can be used to determine the time-
course of both solute and water content simultaneously, so this
technique yields concentration directly. A commonly used indirect
method for isolated cells is to record the time-course of cell volume
following exposure to a known concentration of the compound by
Coulter counter or light-scattering methods; the equations
described previously are then used to model the experimental data
and derive estimates of Lp and Ps. We will now consider in more

detail some situations in which these permeability parameters are
relevant to cryopreservation.
2.2 Introduction The exposure of cells to a high concentration of cryoprotectant

ofCryoprotectants causes osmotic dehydration. If the cryoprotective compound per-
meates, the cells then increase in volume, water entering along with
the cryoprotectant until the cells reach their final volume. The extent
of shrinkage and the rate of change in cell volume are determined by
the permeability parameters. The final equilibrium volume depends
on the concentration of impermeant solutes in the solution and is
the same as the normal volume only if the concentration of imper-
meant solutes is isotonic in molar (per liter) terms. This is because of
the fact that the cryoprotectant occupies space within the cells and
the volume of water must therefore be lower than the physiological
water content if the total volume is to be normal [16]. The rate of
change of volume, and particularly the equilibrium volume, are both
important and must be optimized in cryopreservation procedures.
2.3 Removal When a permeating cryoprotectant is removed by exposing the

ofCryoprotectants cells to a lower concentration of the compound, the osmotic uptake
of water causes the cells to swell above their initial volume. They
then shrink as the cryoprotectant moves out, accompanied by suf-
ficient water to maintain osmotic equilibrium; they return to physi-
ological volume only if nonpermeating solute has neither been lost
nor gained during the process. Because cells are generally more
sensitive to swelling than to shrinkage, removal of cryoprotectants
tends to be more hazardous than their addition. Again, both the
rate of change of volume and the final volume must be considered
when designing protocols for the recovery of cryopreserved cells.
2.4 Freezing Freezing causes the solution surrounding the cells to concentrate,
andThawing and as a consequence the cells shrink at a rate that depends upon
the rate of formation of ice, the cells Lp and its temperature coef-
ficient, and temperature itself. This phenomenon is an extremely
important determinant of intracellular freezing. The final extent of
shrinkage depends on the cryoprotectant concentration.
2.5 Exposure Cells immersed in a solution of nonpermeating solute reach an

toNonpermeating equilibrium volume that is an inverse function of the osmolality of
Solutes the solution; ideally,
1
V rel = (8)
M rel

where Vrel is the volume of intracellular water relative to the physi-
ological water content and Mrel is the external osmolality relative to
its physiological value.
14 DavidE.Pegg
This relationship requires that a plot of Vrel against Mrel, which

is known as a Boyle vant Hoff plot, is a straight line of slope=1,
which must pass through the origin because water content is zero
at infinite osmolality [17]. In reality there is always an intercept on
the y-axisthe so-called nonosmotic water volume or Vinf. This
probably represents a physically distinct portion of the cell water
that is so structured that it does not participate in solution phenom-
ena. Alternatively, it could reflect the nonideal behavior of the intra-
cellular solutes such that osmolality increases with concentration
more than in linear proportion. The value of Vinf, determined by a
Boyle vant Hoff plot, is needed to interpret volume/time data for
the cells in question and to calculate intracellular concentrations of
permeating solutes. Experimentally, the collection of such data can
usefully be combined with determining the upper and lower vol-
ume limits that the cells will tolerate without damage.
3 Preservation ofCells andTissues
3.1 Preservation The basic cryobiological knowledge reviewed here has made it
ofCells possible to develop effective methods for the preservation of a very
wide range of cells, and these have found widespread applications
in biology and medicine. Examples include the long-term preser-
vation of spermatozoa of many species, including cattle, laboratory
animals, and man, very early embryos and ova, red and white blood
cells, hemopoietic stem cells, tissue culture cells, and so on. For
each type of cell there is a set of conditions that is optimal for pres-
ervation, determined by the interaction of the particular properties
of the cell in question with the cryobiological factors that have
been discussed. If the characteristics of the cell are known, it is
usually possible to predict with reasonable precision the conditions
that will provide effective cryopreservation.
3.2 Preservation The situation becomes much more difficult when we move from
ofMulticellular single cells to complex multicellular systems. Cell survival is still
Systems required, of course, but tissues and organs contain a heterogeneous
collection of cells, which may have quite different optimum require-
ments for preservation, unlike the situation in cell preservation
where one is usually dealing with a single type of cell. Yet it is neces-
sary to find a method that will secure adequate survival of all the
cells that are important for the function of that tissue. Fortunately,
the use of high concentrations of cryoprotectant results in a flatten-
ing of the bell-shaped survival curve and a broadening of its peak:
with sufficiently high concentrations of cryoprotectant it is possible
to secure overlapping survival curves for many different cells.
Another problem is that it is not sufficient to obtain high levels of
survival for the various types of cell that are present in tissues and
organs; it is also imperative to avoid damage to important extracel-
lular structures and to retain normal interconnections between the

cells and their attachments to basement membranes [18]. Ice that
forms outside the cells when a cell suspension is frozen is outside
the system that it is desired to preserve, and it can damage the cells
only by indirect means (solution effects) or by exerting a shear or
compressive force on them externally. The situation is quite differ-
ent for organized tissues; here, extracellular ice is still within the
system that is to be preserved and can disrupt the structure of the
tissue directly. The first evidence of such an effect was provided by
Taylor and Pegg [19] when they showed that smooth muscle, fro-
zen to 21C by cooling at 2C/min in the presence of 2.56M
dimethyl sulfoxide, was functionally damaged, whereas exposure to
the solution conditions produced by freezing that solution to that
temperature, at the same temperature, was innocuous. Structural
studies using freeze substitution showed that ice formed within the
muscle bundles [20]. If cooling was slowed to 0.3C/min, freez-
ing produced less damage and ice was shown to form only between
the muscle bundles. This showed that extracellular ice damaged this
tissue, but the extent of such damage was dependent on the site at
which the ice formed. Damaging effects of extracellular ice have
also been demonstrated in kidneys and livers, where it has been
shown to cause rupture of the capillaries.
Rubinsky and Pegg [10] have proposed a mechanism for this
effect; ice forms within the vessel lumens, drawing in water from
the surrounding tissue until the volume of intraluminal ice exceeds
the elastic capacity of vessel and rupture ensues. In organs and tis-
sues that require an intact vasculature for function, vascular rup-
ture is lethal, even if many cells survive, and this mechanism
provides the major barrier to effective cryopreservation of such
systems. The avoidance of freezing, or at least limitation of the
amount of ice to very small quantities in the least susceptible loca-
tions, seems to be the only way to avoid this problem.
Attempts to cryopreserve complex multicellular systems simply
by adapting techniques from single-cell systems have generally
been unrewarding. In the medical field, the situation may be more
favorable with tissues that can be transplanted without revascular-
ization; it all depends on the precise requirements for surgical
acceptability. For example, the primary requirement for heart valve
grafts is that the collagen structure is intact, and it is unclear
whether the survival of donor fibroblasts has any useful effect.
Similarly, human skin can be cryopreserved by methods similar to
those used for cell suspensions and will then retain significant num-
bers of viable keratinocytes, although it is questionable whether
these influence the clinical results when skin grafts are used as a
temporary covering on seriously burned patients. For other tissues,
such as small elastic arteries, satisfactory methods have only been
developed relatively recently [21]. For corneas and for whole vas-
cularized organs there are no effective methods.
16 DavidE.Pegg
Fig. 9 Supplemented phase diagram for glycerol/water. The intersection of the

melting curve and the glass transition curve at Tg indicates the lowest concen-
tration of glycerol that, in theory, will vitrify. In practice, the lower temperatures
on the melting curve are unlikely to be reached owing to the high viscosity pre-
venting the crystallization of ice. (Reproduced with permission from ref. 23)
3.3 Vitrification Much of the very early work in cryobiology, notably by Luyet [22],
Methods had been based on the assumption that freezing damaged cells
directly and, consequently, that effective preservation would require
a technique that completely prevented the crystallization of ice.
Luyet devoted a great deal of effort to the search for conditions that
would produce a vitreous or glassy state with biological systems and
that living cells could survive. Vitrification is defined by the viscosity
of the solution reaching a sufficiently high value (~1013 poises) to
behave like a solid but without crystallization. In conventional
cryopreservation, the concentration of solute in the remaining liq-
uid increases during progressive freezing, and a temperature (Tg) is
eventually reached with many systems where the residual liquid vit-
rifies in the presence of ice (Fig.9). Cells can survive this situation;
they do so in conventional cryopreservation, but they will not toler-
ate exposure to the necessary concentration for vitrification without
freezing (~80g% [w/w]) at temperatures above 0C.Some other
solutes will vitrify at lower concentrations, for example butane-2,3-
diol at around 35% (w/w), but unfortunately this compound is
more toxic than glycerol. Luyet knew that it was possible to vitrify
Fig. 10 Diagram constructed from data by Luyet showing the time and tempera-
ture dependence of nucleation and ice crystal growth in a thin film of a 50%
(w/v) solution of polyvinylpyrrolidone. The arrows indicate cooling trajectories
that avoid nucleation (300C/s), nucleate without crystal growth (80C/s) and
produce ice crystals (20C/s). (Reprinted with permission from ref. 7)
solutions that are less concentrated than this if sufficiently rapid

cooling a risk that freezing will occur during warming; the reason is
that ice crystals form by the accretion of water molecules onto a
nucleus. Both the formation of the nuclei and the subsequent
growth of ice crystals are temperature dependent. Nucleation is
unlikely just below the equilibrium freezing point (hence the phe-
nomenon of supercooling), but it becomes more probable as the
temperature falls, reaches a maximum rate, and then decreases as
the movement of water is limited by viscosity. However, the growth
of ice crystals is maximal just below the freezing point and is pro-
gressively slowed, and eventually arrested, by cooling. The interac-
tion of these two processes creates three possibilities for a cooled
sample (Fig.10); if it cools rapidly it may escape both nucleation
and freezing; if it cools sufficiently slowly it will nucleate and then
freeze; and at an intermediate cooling rate it will nucleate but not
freeze. Upon warming, however, there are only two possibilities; if
heated sufficiently rapidly it will escape both nucleation and freez-
ing during warming; the alternative is that the trajectory passes
through both the nucleation and the ice crystal growth zones and,
therefore, it will nucleate (if it is not already nucleated) and the ice
crystals will then grow before eventually melting. Therefore, unless
a sufficient concentration of cryoprotectant has been used to ensure
that no ice can form under any circumstances, there is a risk that
freezing will occur during warming. The problem is that bulky tis-
sues and organs cannot be cooled much more rapidly than a few
18 DavidE.Pegg
degrees per minute in practice. For small samples it is more feasible

to cool rapidly, as was demonstrated by the successful vitrification of
Drosophila melanogaster embryos [24]. These were complex organ-
isms comprising some 50,000 cells with advanced differentiation
into organ systems, and they cannot be preserved by conventional
freezing methods. The successful method required careful permea-
bilization of the waxy vitelline membrane to allow penetration of
the cryoprotectant, exposure to 8.5M ethanediol, cooling at
100,000C/min, storage at approx 200C, and warming at
100,000C/min. The extremely high rate of warming was far more
critical than the rate of cooling, which is consistent with the crucial
importance of maintaining the vitreous state.
The demonstration that ice forming in tissues produces so
much damage has created renewed interest in the possibility of
using vitrification with very high concentrations of appropriate
cryoprotectants to avoid the formation of ice completely. Current
research aims to identify materials that will inhibit the formation of
ice crystals during warming [25, 26], and one interesting possibility
is the antifreeze proteins that some polar fish and overwintering
insects have evolved to avoid freezing in nature. One effect of such
compounds is to reduce the warming rate required to prevent ice
crystallization to more manageable rates. This approach is being
used in conjunction with electromagnetic heating [27, 28] to
achieve more rapid and more uniform heating. However, despite
progress in the design of vitrification cocktails with reduced toxicity,
the major problem remains cryoprotectant toxicity. One approach
to this problem is to increase the concentration of cryoprotectant
progressively during cooling so that the tissue concentration fol-
lows the liquidus curve: ice does not form but the cells do not
experience any greater concentration of cryoprotectant than occurs
during freezing. This has recently proved to be practical and very
effective for the cryopreservation of articular cartilage, an otherwise
recalcitrant tissue [29]. The same method may potentially be effec-
tive for other resistant tissues and perhaps even for organs.
References
1. Polge C, Smith AU, Parkes AS (1949) Revival 5. Pegg DE (1972) Cryobiology. In: Proceedings
of spermatozoa after vitrification and dehydra- of the fourth international cryogenic engineer-
tion at low temperatures. Nature 164:666 ing conference, Eindhoven. IPC Science and
2. Lovelock JE (1953) The haemolysis of human Technology Press, Guilford, UK, pp4754
red blood cells by freezing and thawing. 6. Pegg DE (1972) Cryobiologya review. In:
Biochim Biophys Acta 10:414426 Timmerhaus KD (ed) Advances in cryogenic
3. Lovelock JE (1953) The mechanism of the engineering. Plenum Publishing Corporation,
protective action of glycerol against haemolysis NewYork, NY, pp116136
by freezing and thawing. Biochim Biophys 7. Pegg DE (1987) Mechanisms of freezing dam-
Acta 11:2836 age. In: Bowler K, Fuller BJ (eds) Temperature
4. Mazur P (1963) Kinetics of water loss from cells and animal cells. Symposia XXXXI of the soci-
at subzero temperatures and the likelihood of ety for experimental biology. The Company of
intracellular freezing. J Gen Physiol 47:347369 Biologists Ltd., Cambridge, UK, pp363378
8. Pegg DE (1970) Organ storagea review. In: tissue stored at 21C or 60C.Cryobiology
Maxwell Anderson J (ed) The biology and sur- 20:3640
gery of tissue transplantation. Blackwell 20. Hunt CJ, Taylor MJ, Pegg DE (1982) Freeze-
Scientific, Oxford, UK, pp195216 substitution and isothermal freeze-fixation
9. Pegg DE, Diaper MP (1988) On the mecha- studies to elucidate the pattern of ice formation
nism of injury to slowly frozen erythrocytes. in smooth muscle at 252K (21C). J Microsc
Biophys J 54:471488 125:177186
10. Rubinsky B, Pegg DE (1988) A mathematical 21. Pegg DE, Wusteman MC, Boylan S (1997)
model for the freezing process in biological tis- Fractures in cryopreserved elastic arteries.
sue. Proc R Soc Lond B Biol Sci 234:343358 Cryobiology 34:183192
11. Leibo SP (1977) Fundamental cryobiology of 22. Luyet BJ, Gehenio PM (1940) Life and death at
mouse ova and embryos. In: The freezing of low temperatures. Biodynamica, Normandy, MO
mammalian embryos. Ciba Foundation sympo- 23. Pegg DE, Diaper MP (1990) Freezing versus
sium 52 (new series). Elsevier, Amsterdam, vitrification: basic principles. In: Smit Sibinga
The Netherlands, pp6992 CT, Das PC, Meryman HT (eds)
12. Pegg DE, Diaper MP, leB Skaer H, Hunt CJ Cryopreservation and low temperature biology
(1984) The effect of cooling rate and warming in blood transfusion. Kluwer Academic
rate on the packing effect in human erythro- Publishers, Dordrecht, Netherlands, pp5569
cytes frozen and thawed in the presence of 2M 24. Mazur P, Cole KW, Hall JW, Schreuders PD,
glycerol. Cryobiology 21:491502 Mahowald AP (1992) Cryobiological preserva-
13. Kedem O, Katchalsky A (1958) Thermodynamic tion of Drosophila embryos. Science
analysis of the permeability of biological mem- 258:19321935
branes to non-electrolytes. Biochim Biophys 25. Sutton RL (1991) Critical cooling rates to
Acta 27:229246 avoid ice crystallization in solutions of cryopro-
14. Kleinhans FW (1989) Membrane permeability tective agents. J Chem Soc Faraday Trans
modelling: Kedem-Katchalsky vs a two- 87:101105
parameter formalism. Cryobiology 37:271289 26. Sutton RL, Pegg DE (1993) Devitrification in
15. Huang Q, Pegg DE, Kearney JN (2004) butane-2,3-diol solutions containing anti-
Banking of non-viable skin allografts using freeze peptide. CryoLetters 14:1320
high concentrations of glycerol or propylene 27. Evans S, Rachman MJ, Pegg DE (1992)
glycol. Cell Tissue Bank 5:321 Design of a UHF applicator for rewarming of
16. Pegg DE (1984) Red cell volume in glycerol/ cryopreserved biomaterials. IEEE Trans
sodium chloride/water mixtures. Cryobiology Biomed Eng 39:217225
21:234239
28. Robinson MP, Wusteman MC, Wang L-H,
17. Pegg DE, Hunt CJ, Fong LP (1987) Osmotic Pegg DE (2002) Electromagnetic rewarming
properties of the rabbit corneal endothelium of cryopreserved tissues: effect of choice of
and their relevance to cryopreservation. Cell cryoprotectant and sample shape on unifor-
Biophys 10:169191 mity of heating. Phys Med Biol 47:
18. Pegg DE (1987) Ice crystals in tissues and 23112325
organs. In: Pegg DE, Karow AM Jr (eds) The 29. Pegg DE, Wang L-H, Vaughan D (2006)
biophysics of organ preservation. Plenum Cryopreservation of Articular Cartilage 3. The
Press, NewYork, pp117140 liquidus tracking method. Cryobiology
19. Taylor MJ, Pegg DE (1983) The effect of ice 52:360368
formation on the function of smooth muscle
Chapter 2
Principles ofCryopreservation by Vitrification

GregoryM.Fahy andBrianWowk
Abstract
Vitrification is an alternative approach to cryopreservation that enables hydrated living cells to be cooled
to cryogenic temperatures in the absence of ice. Vitrification simplifies and frequently improves cryo-
preservation because it eliminates mechanical injury from ice, eliminates the need to find optimal cooling
and warming rates, eliminates the importance of differing optimal cooling and warming rates for cells in
mixed cell type populations, eliminates the need to find a frequently imperfect compromise between solu-
tion effects injury and intracellular ice formation, and enables cooling to be rapid enough to outrun
chilling injury, but it complicates the osmotic effects of adding and removing cryoprotective agents and
introduces a greater risk of cryoprotectant toxicity during the addition and removal of cryoprotectants.
Fortunately, a large number of remedies for the latter problem have been discovered over the past 30+
years, and the former problem can in most cases be eliminated or adequately controlled by careful atten-
tion to technique. Vitrification is therefore beginning to realize its potential for enabling the superior and
convenient cryopreservation of most types of biological systems (including molecules, cells, tissues, organs,
and even some whole organisms), and vitrification is even beginning to be recognized as a successful strat-
egy of nature for surviving harsh environmental conditions. However, many investigators who employ
vitrification or what they incorrectly imagine to be vitrification have only a rudimentary understanding of
the basic principles of this relatively new and emerging approach to cryopreservation, and this often limits
the practical results that can be achieved. A better understanding may therefore help to improve present
results while pointing the way to new strategies that may be yet more successful in the future. To assist this
understanding, this chapter describes the basic principles of vitrification and indicates the broad potential
biological relevance of vitrification.
Key words Vitrification, Freezing, Intracellular ice formation, Devitrification, Recrystallization,

Chilling injury, Cryoprotective agents, Cryoprotectant toxicity, Osmotic limits, Protein denaturation,
Biobanking, Glass transition, Glassy state, Optimal cooling rate, Organ preservation
1 Introduction andGeneral Orientation
1.1 Overview Vitrification is the solidification of a liquid into a noncrystalline or

amorphous (meaning, literally, without structure) solid known as
a glass [1, 2]. The industrial significance of vitrification is well
understood and long-standing. Beyond the manufacturing of famil-
iar glassy items such as porcelain and windows, for example, obsid-
ian, which is a vitrified form of lava [3], was used to make artifacts
21
22 GregoryM.Fahy andBrianWowk
such as arrowheads as long ago as the Stone Age [3], and it is anticipated
that, in the future, vitrification may be used to trap radioactive waste
to prevent it from escaping into the biosphere [4, 5]. On the other
hand, the potential biological significance of vitrification has been
reasonably well appreciated for less than 80years.
The possibility of vitrifying water was postulated as long ago as
1860 [6]. In 1980, the successful vitrification of a 0.1M CuCl2
solution as well as of pure water by ultrarapid cooling was reported
[7], and in 1981, the vitrification of quenched 1m droplets of
pure water was claimed on the basis of an absence of visible ice
crystals in electron microscopic images [8]. But the successful use
of vitrification to preserve biological viability or molecular stability
in the vitreous state, which is the focus of this chapter, was most
unambiguously achieved even earlier, in 1968, when human eryth-
rocytes were vitrified in a rapidly cooled aqueous solution of 8.6M
glycerol and remained intact when rewarmed [9].
Vitrification is usually induced by cooling, which in nonfreez-
ing aqueous solutions eventually elevates viscosity to ~1013P, at
which point the liquid is considered to have reverted to the glassy
or vitreous state [2, 10]. But vitrification or something very close
to it can also be achieved in nature or in the laboratory by drying,
and some organisms [11, 12] and many proteins [13] can be pre-
served successfully in this way.
The ability of vitrification to preserve molecules, cells, tissues,
whole organs, and even some whole organisms has many current
and future agricultural, medical, scientific, and ecological ramifica-
tions. The application of vitrification to cryopreservation has been
growing exponentially since the early 1980s ([1416] and Fig.1)
and may eventually enable the preservation even of systems as
complex and massive as whole human organs for transplantation
[14, 17, 18]. Given the broad potential biological relevance of vit-
rification, which is illustrated in detail in the remaining contribu-
tions to the present volume, an understanding of the basic principles
of vitrification is becoming increasingly important.
1.2 Basic The glass transition temperature, or TG, is the temperature at

Terminology which vitrification, the transition from a liquid-like state into the
glassy state, takes place on cooling. TG is usually defined on the
basis of a change in heat capacity detected by, for example, differ-
ential scanning calorimetry (DSC). Additional discussion of the
nature of the glass transition is given below. TG can be measured
during either cooling or warming, but there is no generally
accepted word that describes the reverse of vitrification, i.e., the
onset of liquid-like behavior as the temperature is raised from
below TG to above TG. The terms vitromelting or vitrofusion
were once suggested to describe this transition [19, 20], but they
have not been adopted.
Principles of Vitrification 23
Fig. 1 Mentions of the word vitrification in PubMed over time. The data are cen-
sored prior to 1986 to avoid extraneous references, but are not censored thereafter
TG is a theoretical temperature reached when freezing is able to

concentrate the unfrozen liquid portion of a solution until its melt-
ing point becomes equal to its glass transition temperature [21, 22].
Such extreme freeze concentration is rarely, if ever, achieved in real
life in the case of aqueous solutions of the low-molecular-weight
glass-forming solutes typically used for cryopreservation by vitrifi-
cation [23] (Fig.2), given that ice cannot continue to grow at a
measurable rate as viscosities are approached that characterize the
glass transition (>1010P), and most literature depictions of TG as
determined during warming probably actually represent the TIM,
the temperature of incipient melting, which is the temperature
of onset of melting well above TG (at the extreme limit of the TM
curve in Fig.2) [2429]. No further comment is made about the
somewhat controversial concept of TG in this review.
Freezing is the reorganization of water molecules into ice
crystals [30]. Although freezing is often used to mean arresting
motion or change, the use of this sense of the word in the context
of vitrification, in which the object is to avoid ice crystallization, is
misleading and inappropriate.
Thawing is the melting of ice. Reference to the thawing
of vitreous systems is common but is inaccurate and is to be
Fig. 2 The behavior of a frozen glycerol solution, leading to freeze concentration of the solution until the
residual unfrozen portion of the solution becomes incapable of further freezing. Continuing cooling then leads
to vitrification of the concentrated unfrozen solution. The thin line, denoted by Tm, is the solution melting tem-
perature. The thick shaded line denoted by TG is the glass transition temperature. The thick arrowed line shows
the concentration of glycerol in the remaining unfrozen solution during slow (~1C/min) cooling. It tracks the
melting temperature until increasing viscosity prevents sufficient ice growth to attain equilibrium. Below the
TG line, the sample consists of a mixture ice and glass
avoided. Rewarming, or the warming of a previously cryopre-

served system, whether frozen or vitrified, is a more accurate term
than thawing in the context of vitrified systems.
Devitrification is not the reverse of vitrification. Instead, it is
the formation of ice during warming after previous vitrification
[3133] and is explained in detail below.
Recrystallization is the transfer of water molecules from small
ice crystals to larger crystals (for an early discussion, see [34]; for
particularly illustrative photos, see [35]). This can happen under iso-
thermal conditions, during which no net change in the total quan-
tity of ice may occur, or during warming, in which case the quantity
of ice may change even as recrystallization proceeds. Recrystallization
tends to be damaging because it results in the conversion of a large
number of relatively innocuous small ice crystals into a smaller num-
ber of larger and more damaging ice crystals. As noted below, recrys-
tallization appears to be more important than devitrification per se
in governing the fate of previously vitrified systems.
The ability of ice to form during either cooling or warming
depends on how much time is available for ice nucleation and
growth. Therefore, at sufficiently high cooling and warming rates,
it is possible to outrun the kinetics of ice formation. The critical

cooling rate is the cooling rate above which appreciable ice
formation is not observed [36], and the critical warming rate
is the warming rate that completely or sufficiently suppresses ice
formation during warming [36].
The critical cooling [37] and warming [38] rates for a given
system depend very strongly on the total solute content of the system
as well as on the chemical nature of the solute. The solutes used for
vitrification are generally the same as or similar to those used to pro-
tect against freezing injury and are generally referred to as cryopro-
tective agents (CPAs) or cryoprotectants [3941]. CPAs that are
of sufficiently low molecular mass to pass reasonably quickly across
cell membranes are referred to as penetrating or permeating
CPAs (pCPAs), while those of higher molecular mass are referred to
as non-penetrating CPAs (npCPAs). Glycerol, which has a molec-
ular mass of 92.1Da, is near the limit for defining the difference
between pCPAs and npCPAs, although less hydrophilic CPAs may
penetrate despite somewhat higher total masses. pCPAs include
dimethyl sulfoxide, ethylene glycol, glycerol, and propylene glycol,
but many others have been identified [40]. To cross cell membranes,
pCPAs must not possess a net charge. npCPAs include polyvinylpyr-
rolidone, polyethylene glycol, sucrose, trehalose, and many others.
Although it has been argued that when cryoprotectants are
used to enable vitrification they should be called vitrificants
rather than cryoprotectants [42], cryoprotectants are defined to
be agents that reduce or prevent freezing injury, so the term cryo-
protectants remains proper in the context of vitrification because
in fact these agents continue to prevent freezing injury, even if they
do so by preventing ice formation altogether. However, the term
vitrificants, while not widely used, is also correctly descriptive of
agents that facilitate vitrification.
A relatively new type of cryoprotectant is the ice blocker,
which is a molecule that is capable of undergoing specific inter-
actions with ice or ice nucleating agents so as to reduce or pre-
vent ice nucleation, ice growth, or both [16, 4347]. Antifreeze
proteins are proteins that can adsorb to the surface of ice crys-
tals and prevent them from growing even when the temperature
is lowered below the thermodynamic melting point of the ice
[41, 48]. Although antifreeze proteins, or AFPs, were the first
natural examples of ice blockers, ice blockers are usually thought
of as being lower in mass and either synthetic or non-protein-
aceous natural products.
A vitrification solution [49] is a solution of cryoprotectants suf-
ficiently concentrated to enable extracellular and intracellular vitrifica-
tion of the system at hand under the intended cooling conditions.
A carrier solution is the physiological support medium in which
CPAs are dissolved to enable cells to be exposed to CPAs without
injury beyond the injury associated with the CPAs themselves.
Chilling injury is injury caused by cooling per se. Although

chilling injury is most conspicuous in the absence of ice, there has
been some speculation and some strong evidence that it can occur
also during freezing in specific cases [5054]. Thermal shock or
cold shock is injury caused by rapid cooling but not by slow
cooling, whereas chilling injury is observed during slow cooling
and may even be outrun by very rapid cooling if the system is
not subject to injury from thermal shock.
Anhydrobiosis [55, 56] is the survival of life in a desiccated
state. It is relevant to vitrification in the sense that sufficiently con-
centrated cytoplasm undergoes a glass transition that contributes
to the survival of organisms, cells, or seeds that are adapted to or
prepared for preservation by drying [57, 58]. Although it is of
considerable potential applied and ecological significance [5760],
this chapter focuses primarily on low-temperature vitrification
rather than on high-temperature anhydrobiosis or aestivation.
1.3 Vitrification Vitrification is important for protecting cells and tissues against
andMolecular freezing damage, but it is not as important for preserving the basic
Stability at Low molecular inventory of cells and tissues. Most molecular constitu-
Temperatures ents of cells are reasonably stable under low-temperature condi-
tions in situ even without special precautions, although there are
exceptions. Generally speaking, neither freezing and thawing nor
cooling per se causes the formation or breakage of covalent chemi-
cal bonds. The reversible formation of S-S cross-links in frozen
thiogels [61], one particular protein (but not others) extracted
from freeze-killed cabbage [62], and one of five SH groups in
F-actin [63] has been reported, but no change in S-S or S-H con-
tent was found in lethally frozen sea urchin eggs [64], and an
increase of S-H content in frozen-thawed bull spermatozoon
membranes was observed [65].
Temperature reduction inhibits most chemical reactions (e.g.,
alkaline phosphatase catalysis is about 95% slower at 25C than
at 0C [66]), and although reactants can be concentrated greatly
by the freezing process, chemical reactions are generally not quickly
driven forward as a consequence, although enzymatic reaction
rates may briefly increase at high subzero temperatures in frozen
model systems [66]. Exothermic phase changes such as the crystal-
lization of water or the formation of the liquid crystalline gel state
[67, 68] or HEXII state [69] of membrane lipids are favored over
limited temperature ranges, but these phase changes do not destroy
but only rearrange the participating molecules and in most cases
are reversible. Protein cold denaturation, discussed in more detail
below, may or may not be spontaneously reversible, but usually
does not involve covalent modification of the protein.
In some cases, the cryoprotectants used for vitrification may
inhibit and in some cases may promote covalent or non-covalent
changes in biomolecules, but their main purpose is to prevent
physical changes, particularly involving cell distortion, that precede

and thus are generally far more important than chemical reactions,
phase transitions, or protein denaturation for the survival of living
cells during cryopreservation.
1.4 Cryopreservation A historical introduction to the field of cryopreservation by vitrifi-

by Vitrification: cation will help to convey some of the key overall concepts in
AConceptual History approximately the order in which they were originally developed.
Additional reviews of the history of biological vitrification are avail-
able elsewhere [15, 16, 42, 7072].
Cryopreservation by vitrification was apparently first intro-
duced conceptually although without much clarity or influence by
Stiles [73] in 1930. Apparently independently, the idea was rein-
troduced actively, clearly, and influentially by Luyet [31] in 1937.
Luyet was inspired in part by Tammanns finding that 38% of
tested organic compounds could be vitrified by rapid cooling [74]
and in part by indications that aqueous gelatin gels could be vitri-
fied [75, 76]. The original concept was that if the water in living
systems could be cooled rapidly enough, there would be insuffi-
cient time for crystals to form before reaching the glass transition
temperature of water, and the living system could therefore be
trapped in the vitreous state [31].
This approach in principle limited vitrification to samples that
could be cooled and warmed very rapidly, but it was nevertheless
pursued energetically by Luyet and his associates for 21years [70].
In 1958, evidence emerged indicating that Luyets primary indica-
tion of vitrification, optical transparency of thin films or thin living
systems after cooling followed by opacification or continued trans-
parency on warming, had been misleading and had deceived Luyet
into believing he had attained complete vitrification when in fact
what he had attained was partial vitrification involving predomi-
nantly the formation of ice crystals known as spherulites that were
too small or too thin to scatter visible light enough to be visible to
the naked eye [77, 78].
This setback apparently chastened Luyet, who never again
advocated or claimed cryopreservation by vitrification. Instead, he
turned his focus onto ice and studied its formation and morphol-
ogy under an extensive array of conditions [70]. Eventually, he and
his colleagues learned that the presence of very high concentra-
tions of cryoprotectants, including both small molecules [27, 28,
79, 80] and larger polymers [26], could in fact enable vitrification
even at low cooling rates [70, 81]. This is the key observation that
has enabled most modern methods of vitrification, but ironically,
Luyet himself never proposed using high concentrations of cryo-
protectants to cryopreserve living systems by vitrification.
The ability of cryoprotectants to enable vitrification only began
to be elucidated and disseminated by Luyet and his colleagues as
of the end of 1966 [80] and continued until 1970 [2628, 79].
At least in part for that reason, in 1965, a historical opportunity

was missed. In that year, Farrant [82] reported that he could main-
tain guinea pig uteri in the liquid state at 79C (the temperature
of subliming dry ice) using 55%v/v dimethyl sulfoxide (Me2SO)
as the cryoprotectant and recover these whole organs with excel-
lent contractile function after rewarming and removal of the
Me2SO.As later pointed out by Fahy etal. [83], 55%v/v Me2SO
is sufficient for vitrification, so Farrant could have actually achieved
successful vitrification of whole organs in 1965 merely by cooling
them another 55C or so. However, in 1965, it was still assumed
that cooling below 79C would lead to ice formation [82].
As noted above, the first definitive report of the successful vit-
rification of a living cell was published in 1968 by Rapatz and
Luyet, who showed that erythrocytes cooled at high rates in the
presence of 8.6M glycerol remained intact (did not hemolyze)
under conditions in which freeze-fracture electron microscopy
demonstrated the absence of discernable intracellular and extracel-
lular ice crystals [9]. Before this, it is possible that some of Luyets
successes in recovering life after very rapid cooling and warming in
the presence of cryoprotectants (which were used to achieve dehy-
dration prior to cooling so as to reduce the volume of water that
required vitrification) [8486] might have included some level of
useful vitrification, but this is difficult to infer from available knowl-
edge. Ironically, the vitrification of red cells by Rapatz and Luyet,
which might have been regarded as the culmination of Luyets life
work, was noted little, if at all, outside of Luyets laboratory for
many years and is still almost never cited.
In 1970 [87] and 1972 [88], Rapatz reported successfully
cooling frog hearts to 79C and rewarming them with good
recovery using 11M ethylene glycol (EG) and variations on
Farrants [82] pioneering method. He presumably could have vit-
rified and successfully recovered these hearts in a viable condition
and, unlike Farrant, must have understood that this was possible,
but he reported no attempts to do so. This is likely to be because,
as he reported from the podium during his 1970 presentation [87]
(but did not mention in his published abstract [87]), when hearts
loaded with EG were transferred into liquid nitrogen, they shat-
tered, as he put it (a general problem that is discussed in detail
below). He later reported that 10M EG was the minimum con-
centration allowing recovery of frog hearts from 79C, but that
rat hearts could not tolerate more than 5M EG and therefore
could not be successfully preserved [89]. Nevertheless, establish-
ing that frog hearts, at least, can theoretically be vitrified and
recovered remains one of the most outstanding achievements in
biological vitrification.
From 1970 [90] to 1972 [91], Elford similarly worked out a
method for preserving strips of intestinal smooth muscle in a
supercooled state at 79C using variations of Farrants method.
In the 1972 work, although it was not reported formally, it has
been noted anecdotally (D.E.Pegg, personal communication) that

some muscle strips cooled in liquid nitrogen (the ones that had not
experienced the same kind of shattering or fracturing observed
by Rapatz) recovered after warming and would therefore have
been the first successfully and definitively vitrified organized tis-
sues. However, once again, there was no suggestion that vitrifica-
tion as opposed to deep supercooling might be used as a method
of cryopreservation.
The first intimation that high concentrations of cryoprotec-
tants might in theory be used to enable cryopreservation by vitrifi-
cation came in 1978, when Pierre Boutron, a physicist who turned
to cryobiology partly on the basis of the phase diagram work of
Luyet and colleagues after previously having studied the structure
of amorphous solid water [92], published a landmark paper that
contemplated vitrification in a new way. This paper was the first to
thoroughly describe the kinetics of ice formation in vitrifiable
aqueous cryoprotectant solutions (containing glycerol, Me2SO,
and mixtures of the two) in view of the theoretical possibility of
completely vitrifying cell suspensions using mixtures of cryopro-
tectants to facilitate vitrification and reduce toxicity [36]. It also
pioneered the combined use of X-ray diffraction and differential
scanning calorimetry (DSC) to investigate ice formation and glass
transitions in aqueous solutions, introduced the concepts of the
critical cooling rate and the critical warming rate, and introduced
mathematical models of the kinetics of ice formation relevant to
vitrifiable solutions [36]. However, Boutrons aim was to find a
very stable amorphous state of the whole solution even for diluted
solutions (emphasis added), whereas the accumulated work of
Luyet and his colleagues had clearly established that amorphous
solutions can only be stable against ice formation when they are
concentrated, not when they are dilute. Indeed, Boutrons own
paper showed that, for example, the critical warming rate of a 45%
w/w solution of glycerol is about 31011C/min, although the
critical warming rate for 45% w/w Me2SO was found to be only
17,000C/min [36].
Undeterred, Boutron and Kaufmann went on to study, in
19781979, the stability of the amorphous state in aqueous solu-
tions of ethanol [93], glycerol plus ethylene glycol [94], glycerol
plus ethanol [95], and, most significantly, propylene glycol (PG, or
1,2-propanediol) [96]. Remarkably, 35% w/w PG could be
vitrified when cooled at only 320C/min, and 40% PG vitrified
when cooled at ~40C/min. The critical warming rate for 45%
w/w PG was a remarkable 260C/min, but the critical warming
rate for 40% w/w PG was extrapolated to be 76,000C/min.
In later years, Boutron and colleagues continued to seek solu-
tions that could be vitrified at relatively low concentrations on the
theory that such solutions would be less toxic than solutions that
required higher concentrations to vitrify, regardless of the practical
difficulties of escaping from ice formation during the warming of
these solutions from the vitreous state. Perhaps because such meth-
ods would not be applicable to systems much larger than single
cells and because single cells can generally be preserved well by
freezing using much lower and hence less toxic concentrations of
cryoprotectants, Boutrons method did not seem to inspire much
attention at the time.
In 19811984, Fahy [83, 97102] proposed a different
approach to vitrification that, in principle, is both definitive and
universally applicable to almost all biological systems. Inspired by
his desire to overcome mechanical injury from ice in whole organs
[15, 49, 72, 83, 97], his method relies on the fact that at suffi-
ciently high concentrations, both the critical cooling rates and the
critical warming rates needed to suppress ice formation become
low enough to enable, in principle, even the vitrification of objects
as large as human organs. Unlike Boutrons approach of attempt-
ing to sidestep toxicity by using lower concentrations, Fahy elected
to attack the problem of high-concentration toxicity head on [83,
103106] so as to avoid the need for high cooling and warming
rates. Also, unlike the approaches of Farrant, Elford, and Rapatz,
which required introducing cryoprotectant at temperatures as low
as 55C [87], Fahys approach sought to enable the use of cryo-
protectants at much higher temperatures that were more compat-
ible with organ perfusion and the very temperature-dependent rate
of passage of pCPAs across cell membranes.
This approach was shown to work when applied to mouse
embryos by Rall and Fahy in 1985 [49]. This demonstration,
which showed that vitrification was at long last a feasible general
method for cryopreservation, inspired much subsequent research
on a variety of living systems using a variety of cryoprotectant solu-
tions and methods (Fig.1). The emphasis since 1985 has been
largely on refining the basic but complex parameters of cryoprotec-
tant selection; the concentration, temperature, timing, and osmotic
effects of each step of cryoprotectant introduction and removal;
methods and equipment for cooling and warming; and methods
for avoiding fracturing. These methods have been applied to a
wide variety of living systems (see, e.g., the listings in [15]), includ-
ing plant systems [107, 108]. The number of papers devoted to
these topics since 1985 is beyond the scope of this introductory
review, but their contents are reflected by the scope of the other
contributions to this volume.
1.5 Advantages The overall purpose of vitrification is to avoid freezing. To under-

andDisadvantages stand the advantages and disadvantages of vitrification, it is there-
ofVitrification fore necessary to understand something about freezing injury.
Conventional cryopreservation by freezing involves, by
definition, the formation and dissolution of ice during cooling
and warming, respectively. Ice is an almost completely pure sub-
stance, so its formation subtracts solvent water from a freezing

solution, leaving the dissolved solutes in a reduced volume of sol-
vent. The effects of ice formation are in part due to this concen-
trating action, which increases both the osmotic concentration
of the cellular environment and the individual concentrations
of dissolved solutes such as electrolytes, buffers, etc. [109, 110].
If cooling proceeds sufficiently slowly, ice formation begins extra-
cellularly [111, 112], there is time for cells to lose water down the
transmembrane osmotic gradient established by the extracellular
ice, and the cells will consequently shrink. If shrinkage proceeds
too far, osmotic injury may result [113116]. If cooling proceeds
more rapidly, the rate of water subtraction from the cell fails to
keep up with the rate of water subtraction from the extracellular
environment, leaving the cell interior significantly more dilute
than the extracellular solution [117]. This means that the thermo-
dynamic freezing point of the cell fails to fall as rapidly as the
prevailing temperature, i.e., that the temperature of the cytosol
begins to fall farther and farther below its nominal freezing point.
This defines a state of supercooling (cooling below the freezing
point without ice formation), and as supercooling increases, the
risk of ice formation within the cytosol increases.
In summary, cells cooled too slowly are liable to injury related
to shrinkage and changes in solution composition (solution
effects injury), whereas cells cooled too quickly are liable to injury
related to intracellular ice formation (IIF) [117121]. Between
these two ends of the spectrum, there is an optimum cooling rate
that minimizes both sources of injury [117121].
These considerations are relevant to vitrification for three rea-
sons. First, the existence of an optimum cooling rate is problem-
atic. The optimum cooling rate can only be determined
experimentally for every cell type of interest, which is inconvenient,
particularly for multicellular tissues, which may contain not only
multiple cell types but also cells in different relationships to each
other and to the extracellular environment, all of which affect the
optimal cooling rate [118]. Moreover, given the existence of a dif-
ferent optimal cooling rate for different cells, finding a compro-
mise rate that gives high recoveries of all cells may be difficult and
has been proposed as a limiting factor for cryopreserving complex
systems [118]. Furthermore, even for a given type of cell, the opti-
mum cooling rate very often fails to yield 100% cell survival [118,
122]. And finally, the use of cryoprotective agents to increase sur-
vival at the optimal cooling rate also changes the optimal cooling
rate itself [122], again in a way that will be cell type dependent.
For these reasons, vitrification is advantageous in part because it
transcends the need to find an optimal cooling rate, to compromise
the survival of one cell type to ensure survival of another cell type,
and to accept cell survival rates that are unlikely to approach 100%.
The second point of relevance to vitrification is that, as noted,

pCPAs must generally be used to obtain high survivals after freez-
ing and thawing, since they mitigate solutions effects injury.
Although relatively low concentrations of pCPAs are needed to
prevent solution effects injury in many cells, the concentrating
effect of freezing on dissolved solutes pertains just as much to
pCPAs as it does to other solutes, the result being that pCPA con-
centrations may be driven high enough in the frozen state to
induce toxic effects of their own [123126]. Interestingly, the
concentrations generated by freezing actually exceed the concen-
trations required for the vitrification of even large living systems
[83, 101, 127], so the advantage of using lower concentrations for
freezing is not necessarily as large as it at first appears.
The third reason the constraint of an optimum cooling rate is
relevant to vitrification is that some important living systems such
as oocytes are subject to chilling injury (see below), and attempts
to cool more rapidly than the kinetics of chilling injury proceed are
precluded if the result is death secondary to IIF (intracellular ice
formation). Vitrification eliminates that obstacle by eliminating IIF
at high cooling rates and has often been pursued for that reason
[52, 54, 128, 129].
Beyond changes in solution composition and IIF, freezing can
result in injury in at least two additional ways, both of them being
mechanical in nature. First, the physical displacement of structures
in organized tissues by the simple growth of extracellular ice can
cause considerable damage to both the vascular bed and to paren-
chymal structures [98, 130139]. In fact, it was the observation
that dog kidneys frozen to 30C and stored for a week using 3M
glycerol could perfuse normally and respond well to pressors
invitro but urinated whole blood and stopped perfusing within
1020min of being transplanted (Fahy, Goldman, and Meryman,
unpublished results) that inspired the proposal to investigate vitri-
fication as a more promising approach to organ cryopreservation.
Fortunately, Taylor and his colleagues have provided extensive
microscopic evidence using freeze substitution methods that vitri-
fiable solutions successfully prevent tissue distortion by ice [16,
140, 141]. Second, even single cells can be injured by intracellular
[118] and extracellular [134, 135] recrystallization. Both forms of
mechanical injury secondary to ice formation argue for vitrification
as a potentially less damaging preservation method, particularly for
complex organized tissues and organs.
Vitrification does have significant disadvantages as well, how-
ever [142, 143]. First, the need to tolerate very high concentra-
tions of CPA requires relatively sophisticated methods of adding
and removing these agents and careful selection of the right CPA
blend for the living system at hand. Second, it is not always clear
what CPA concentration and exposure time is needed to ensure

vitrification and maintenance of an ice-free state upon warming in
specific cases, and investigation of these points may be needed.
Third, rapid cooling to below the glass transition temperature
and/or rapid warming below TG may induce fracturing of the
glass in which the biological system is embedded [144, 145] (see
also Subheading2.7), which may cleave cells or tissues, causing
irreversible injury [146] and additional ice nucleation [147].
Fourth, to avoid the third problem, in some cases, storage at
intermediate temperatures [~130 to 160C, i.e., below TG
but above the temperature of boiling liquid nitrogen (196C)]
may be needed to ensure long-term preservation without crack
formation. Finally, although rapid warming is generally beneficial
for frozen systems, it is even more important for vitrified ones
due to the need to avoid injury from devitrification and subse-
quent recrystallization.
A potential disadvantage of certain techniques of vitrification
comes from the use of container-free cooling methods to acceler-
ate the cooling and warming rate and thereby enable the use of
minimal concentrations of cryoprotectant. The lack of a container
may result in contamination of the sample being preserved [142].
The need for such methods is questionable, however, and it seems
likely that closed-system vitrification will eventually remove the risk
of contamination.
1.6 Vitrification It is reassuring that nature has often drawn the same conclusion as
inNature the cryobiologist in pursuing ice-free cryopreservation in prefer-
ence to freezing.
A number of insects, for example, survive the winter by
freeze avoidance [57, 58, 148, 149], achieved by suppressing
the presence of ice nucleating substances, synthesizing high con-
centrations of cryoprotectants such as glycerol, and producing
antifreeze proteins that bind to ice and prevent it from grow-
ing (see Subheading2.6 below for more discussion of antifreeze
proteins and ice blockers). In one case, that of the larval Alaskan
red flat bark beetle (Cucujus clavipes puniceus) [149], more than
half of the individuals tested supercooled to below 60 to 70C
and none showed exotherms indicative of freezing when cooled
to 150C in a DSC.All showed large whole body glass transi-
tions between 58 and 76C (DSC curve inflection points;
mean TG, 71C). Two large larvae had a second small TG at
96 or 98C.When unselected larvae were cooled to
71.51.5C or to 100C, the survival rates were about 50%
and 7%, respectively, although the latter rate was probably
reduced by mechanical damage sustained by the methods used.
In any case, at least some larvae appear able to survive cooling to

below even the lowest observed TGs. Further, the coldest tem-
peratures recorded in nature (79.8C in Alaska [150], 89.2C
at the Russian Vostok Station in Antarctica [151], and most
recently 94.7C in eastern Antarctica [150]) are all below this
insects main TG and well below the highest measured TG of
58C.This suggests that some organisms have actually sur-
vived low temperatures in a vitreous state under natural condi-
tions using endogenous cryoprotectants similar in both molecular
weight and concentration to those being used for artificial vitri-
fication in cryobiological laboratories to those being used for
artificial vitrification in cryobiological laboratories (in C. c. puni-
ceus, up to 6.5M glycerol was measured in the cited study, and
up to 10M glycerol has been reported elsewhere [152]).
C. c. puniceus, in addition to elaborating protective substances,
concentrates them to vitrifiable levels in part by water loss to the
environment. Species whose moisture content varies with the ambi-
ent humidity are said to be poikilohydric, and this water loss can be
sufficient to induce cytoplasmic vitrification [108, 151, 153, 154]
and may even make vitrification under natural conditions more com-
mon than survival by freezing tolerance [151, 155]. As one example,
soil nematodes dried to below 0.3g of water per g dry weight sur-
vived cooling in liquid nitrogen and showed no evidence of freezing
[156]. However, vitrification by dehydration enables survival at high
temperatures as well as at low temperatures [11, 12].
The lowest common terrestrial temperatures are typically
between about 30 and 60C [149, 151], but the cells of
many species have been shown to have glass transition tempera-
tures above 50C [57, 153, 157]. Twigs of Populus balsam-
ifera were shown to undergo a glass transition at about 45C
and are known to be able to survive immersion in liquid nitro-
gen [157]. Highly frost-hardy plants in general, according to
Sakai, survive in conjunction with their ability to form intracel-
lular glasses [108].
1.7 Vitrification When ice forms in the presence of cryoprotectants, their concen-
During Freezing trations are elevated by loss of water from the solution into the ice
ofLiving Cells phase, until eventually they preclude further ice formation during
continued cooling, resulting ultimately in vitrification of the resid-
ual unfrozen solution [79, 83] (see Fig.2; also Subheading2.1).
Similarly, freeze concentration of the extracellular solution and
concomitant osmotic reduction of cell volume (see also
Subheading 1.5 above) result in vitrification of cytoplasm when
cooling is slow enough to preclude IIF [83, 158161]. Therefore,
most cells survive cryopreservation as a result of vitrification, even

if the medium surrounding them is not completely vitrified.
2 The Physical Principles ofVitrification
2.1 Vitrification Figure2 illustrates the process described in Subheading1.7, show-

Depends ontheSolute ing the example of the slow freezing of a 10% glycerol solution in
Concentration water on a glycerol-water phase diagram. The solution may initially
ofAqueous Solutions supercool before the first ice crystal forms, but thereafter the con-
centration of the remaining unfrozen solution follows the melting
temperature line (Tm) with continued cooling. Increasing solution
viscosity during cooling eventually inhibits ice growth, causing a
small departure from thermodynamic equilibrium [23]. Final cool-
ing then continues with little change in concentration until the
glass transition temperature (TG) is reached [83]. Below TG, the vis-
cosity of unfrozen solution between ice crystals becomes so high
(>1013P) that the solution behaves as a solid. It does so while retain-
ing the random molecular arrangement of a liquid. A solid with
the same unstructured molecular arrangement as a liquid is called a
glass [1].
With a sufficiently high solute concentration and/or cooling
rate, it is possible to cool all the way to the glass transition tempera-
ture without significant ice formation [83, 101]. This is the basis of
cryopreservation by vitrification. During cryopreservation by vitrifi-
cation, the entire sample volume remains substantially free of ice
during cooling. As shown in Fig.3, this can be achieved by using
either low solute concentrations and fast cooling rates or higher con-
centrations and slower cooling rates. Heat transfer limitations neces-
sitate the use of high solute concentration and slow cooling rates
when vitrifying large volumes, such as tissues and organs.
As hinted at in Fig.3, there is a way to predict the concentration
of CPA that will enable vitrification at moderate cooling rates
(around 10C/min). The curve labeled Th designates the homo-
geneous nucleation temperature, which is further described in
Subheading2.3. Th sets the limit beyond which the solution cannot
be supercooled without ice nucleation. Careful analysis of the thresh-
old concentration required for vitrification (CV or CNV, the con-
centration needed for vitrification) based on visual inspection of
~8ml volumes of CPA solutions cooled at about 10C/min showed
that, for glycerol-water, ethylene glycol-water, dimethyl sulfoxide-
water, and propylene glycol-water systems (and the latter at 1, 1,000,
and 1,500atm of applied pressure), CV coincided with the concen-
tration required to make Th equal to TG in every case [83].
Fig. 3 Vitrification at three different concentrations of glycerol in water. Unlike freezing, with vitrification the
solution concentration remains constant during cooling because cooling is too rapid for ice to form or grow
appreciably. Unstable vitrification requires cooling at thousands of degrees per minute, or more, due to high
ice nucleation and growth rates associated with homogeneous nucleation. Metastable vitrification is typically
possible at cooling rates on the order of 10C/min. Stable vitrification (equilibrium vitrification) is possible
at arbitrarily low cooling rates
2.2 The Physical Vitrification occurs when thermal energy becomes insufficient for
Nature andBasis molecules to overcome potential energy barriers that must be
ofVitrification overcome for translational rearrangements within a liquid. Below
the glass transition temperature, molecules lose the ability to wander
among other molecules over the timescale of measurements being
made. They instead vibrate in place. As a consequence, the measured
values of thermodynamic response functions such as heat capacity,
thermal expansion coefficient, and compressibility fall from those of
a liquid to those of a solid. The glass transition is typically detected
calorimetrically by an observed drop in heat capacity during cooling.
In contrast, thermodynamic state variables such as volume, energy,
and entropy do not change at the glass transition. Only their slope as
a function of temperature undergoes change [2].
Although the glass transition is a material phase change (liquid
to solid), it is not a thermodynamic phase change from one equi-
librium state to another. The glass transition is a kinetic phenom-
enon in which viscosity delays intermolecular rearrangements that
are thermodynamically favored. In essence, vitrification locks in
a nonequilibrium thermodynamic state. As cooling rates are varied
by orders of magnitude, measured glass transition temperatures
can vary by several degrees Celsius [162], with slower cooling rates
resulting in lower measured glass transition temperatures. This is
due to the kinetic nature of the glass transition. Slower cooling

rates provide more time for intermolecular rearrangements that
release heat, contract volume, and otherwise approach equilibrium
before rising viscosity stops liquid-like behavior.
The measured decrease in heat capacity that occurs during
passage below the glass transition temperature provides a more
abstract interpretation of the glass transition. The heat capacity of
a liquid above the glass transition temperature is greater than that
of a crystal at the same temperature. Entropy varies as heat flow
divided by temperature. Therefore, during cooling, the entropy of
a liquid decreases faster than the entropy of a crystal of the same
composition. This leads to a projected temperature called the
Kauzmann temperature (TK) below which the liquid is extrapo-
lated to have a lower entropy than the crystal [163]. Since a disor-
dered liquid state is supposed to have higher entropy than an
ordered crystal, cooling a liquid to TK would create a paradox. The
decrease in heat capacity at the glass transition prevents this ther-
modynamic paradox. Although kinetic in nature, the eventual
occurrence of a glass transition can be viewed as a thermodynamic
necessity for crystallizable liquids with a Kauzmann temperature
greater than absolute zero.
As mentioned above, thermodynamic nonequilibrium is
intrinsic to glasses at the time of their formation. Lack of equilib-
rium pertains to the glassy phase in which cryopreserved cells are
suspended irrespective of whether cryopreservation is by freezing
or vitrification.
Four types of nonequilibrium are noteworthy in the context of
cryopreservation. First, there is vapor pressure nonequilibrium
between ice and the unfrozen sample volume. This is important for
cryopreservation by vitrification because the aim is for the sample to
remain substantially free of ice during cooling and storage despite
strong thermodynamic driving forces (vapor pressure nonequilib-
rium) favoring ice growth. Unequal vapor pressures between ice
crystals of different sizes also play a role in ice recrystallization [34,
164], which can damage cells during rewarming of frozen samples
or vitrified samples that form ice during rewarming [134, 165,
166]. Second, there is chemical nonequilibrium, in the sense of
chemical instability and change. At relatively high temperatures,
chemical or protein conformational changes in cells are able to pro-
ceed beyond the controls of normal metabolism, but these changes
can be slowed and ultimately arrested as temperatures decline and
increasing viscosity combined with insufficient activation energies
prevents chemical reactions. Chemical nonequilibrium may relate
to the rate of viability loss during cooling. Third, there is chemical
potential nonequilibrium. Nonequilibrium of cryoprotectant con-
centration between different regions of a vitrified sample [17] or
across the cell membrane [83, 167] can make the difference between
satisfactory and unsatisfactory p reservation. Fourth, there is
mechanical nonequilibrium. Mechanical nonequilibrium is relevant

to mechanical stress and strain during cryopreservation, which is an
especially important consideration for large samples such as tissues
and organs, and is discussed in more detail in Subheading2.7.
2.3 Ice Nucleation Ice formation begins with a process called nucleation [168]. There
are two kinds of nucleation, homogeneous and heterogeneous.
During nucleation, water molecules begin organizing into the
structure of ice on a nanometer scale. The resulting nascent ice
crystals, or ice nuclei, tend to be unstable. In accordance with the
Gibbs-Thomson equation [169], ice crystals of small size have low
melting temperatures (high vapor pressure) caused by sharp curva-
ture of the crystal surface. Consequently, newly formed ice nuclei
tend to melt at any temperatures warmer than about 38C in
pure water. This defines the homogeneous nucleation temperature
(Th), the lowest temperature to which small [30, 170] samples can
be cooled under normal conditions without ice formation and the
highest temperature at which small samples are likely to form ice
when preexisting ice crystals or contaminants that mimic ice crys-
tals (heterogeneous nucleators, discussed below) are absent [30,
70, 170173]. As shown in Fig.3, Th decreases with increasing
solute concentration [79].
Ice can form at temperatures above Th in the presence of het-
erogeneous nucleators [30, 168, 174]. Heterogeneous nucleators
are particles or surfaces that mimic the structure of ice on a molecu-
lar scale or otherwise induce water to assume a more icelike con-
figuration with a larger radius of curvature. Ice crystals with a larger
radius of curvature upon their initial formation have a lower surface
energy, allowing them to avoid melting at warmer temperatures
[174]. The most potent heterogeneous nucleators can cause ice to
form at temperatures only 1C below Tm [175, 176]. Heterogeneous
nucleators are ubiquitous environmental contaminants [168, 174].
They are usually responsible for initial ice nucleation events in vol-
umes larger than ~10m diameter water droplets prepared as aero-
sols or emulsions to study homogeneous nucleation [177].
The phase diagram of a cryoprotectant solution can be divided
into heterogeneous and homogeneous nucleation zones (Fig.3).
These zones give rise to at least three distinct types of vitrification.
Vitrification using solute concentrations insufficient to prevent
passage through the homogeneous nucleation zone will nucleate
at an innumerably large number of points in the solution [164]
(e.g., 2,500 nuclei/m3 in one case [178]). Since in the homo-
geneous nucleation zone water is self-nucleating, passage through
this zone should be considered unstable vitrification. (In glass sci-
ence, solutions that homogeneously nucleate are said to be doubly
unstable due to instability with respect to both crystal nucleation
and growth [179].) Unstable vitrification can be survived if warm-
ing is sufficiently rapid, as discussed below.
Vitrification using a solute concentration high enough to avoid

homogeneous nucleation but low enough that ice formation is still
thermodynamically favored (T<Tm) can be defined as metastable
vitrification [83, 180, 181]. It is metastable because such a s olution
can supercool all the way to the glass transition without ice nucleation
necessarily occurring, but ice is still thermodynamically favored.
Ice nucleation events will essentially be confined to discrete loca-
tions where heterogeneous nucleators are present.
A third form of vitrification is stable vitrification, or equilib-
rium vitrification [16], which uses solute concentrations so high
that ice cannot exist in the solution [2, 83]. A practical approxima-
tion to the edge of the stable vitrification zone can be defined as
the concentration that is sufficient to make devitrification vanish
during slow warming [18, 83], but complete stability requires the
higher concentrations that are typical of those that preclude ice
growth during previous slow freezing (the unfreezable concen-
tration, or CU [15, 72]).
Ice nucleation occurs by local reorientation of water molecules.
The nucleation rate is therefore driven primarily by the free energy
difference (thermodynamic driving force) between ice and liquid
water in solution. Rates of nucleation rise as temperature decreases,
reaching a maximum near the glass transition temperature [2,
182]. Only below the glass transition temperature do astronomical
viscosity and loss of rotational freedom begin to slow ice nucle-
ation [2, 182, 183].
2.4 Kinetic Aspects Boutron was the first to characterize the quantitative relationships
ofIce Formation between the cooling rate of CPA-water solutions of different con-
inVitrification centrations and their ability to escape from ice formation on cool-
Solutions During ing, as evidenced by the lack of exotherms recorded using
Cooling differential scanning calorimetry [94]. His modeling eventually led
[37] to an equation
(( ) )
- ln (1 - x 1/3 ) + 0.5 ln (1 + x 1/3 + x 2 /3 ) + 3 arctg 3 x 1/3 / ( 2 + x 1/3 ) = k 4 / v
that accurately predicts the amount of ice formed during the cool-
ing of cryoprotectant solutions at different rates given certain
starting information, such as the amount of ice that forms at very
low cooling rates, in which ice formation is maximum. In this
equation, k4 is a constant and x is q/qmax, where q is the calori-
metrically determined amount of ice observed to form at cooling
rate v and qmax is the similarly determined maximum amount of ice
that can form at very low cooling rates. When k4 equals v, x=0.036,
so k4 is equivalent to the cooling rate that reduces ice formation to
3.6% of the maximum amount that can form [184].
The critical cooling rate can be defined as the cooling rate suf-
ficient to reduce x to whatever any particular living system can
tolerate. In the limit of low x, which is the limit of interest for vit-
rification, the above equation becomes simplified. If x=106, then
v = 100k 4 / 3.
If x is set to a value reflecting 0.2% w/w ice, then the equation

becomes [184]
(1 / 3 )
v = k4 / 3 ( 0.2 / qmax ) ,

where qmax is the maximum mass percent of the solution calori-

metrically observed to freeze while cooling very slowly (arbitrarily
using the 334J/g heat of fusion of pure water at 0C to convert
the heat of the exotherm into the mass of ice formed) and the 0.2
factor represents a mass percent of solution crystallized of ~0.2%
(which should be low enough even for the survival of a vitrified
kidney [17]). The convention of choosing 0.2% w/w ice as a stan-
dard for defining the critical cooling rate is based in part on the fact
that this is the minimum amount of ice that has been considered to
be quantifiable by DSC [184].
Figure4 shows the effect of k4 when the critical cooling rate is
defined as the rate that reduces x to 0.1, 0.2, 0.5, or 1% of the
maximum amount of ice that can form in the system. Using
Fig. 4 Universal relationship between Boutrons k4 values and the critical cooling rate for vitrification based on
the choice of x (0.11% of the maximum possible amount of ice)
x=0.2/qmax, Baudot etal. have calculated critical cooling rates for

many cryoprotectant solutions and compared them to the critical
warming rates for the same solutions [185].
2.5 Devitrification Once ice has nucleated into stable nanoscale [186] nascent ice
andRecrystallization crystals, it tends to grow. However, unlike nucleation, which
depends on local molecular motion, ice growth requires diffusion
to supply water molecules to a growing ice front and dissipate
solutes not incorporated into the crystal. This dependence on dif-
fusion makes the rate of ice growth strongly, and inversely, depen-
dent upon solution viscosity. Consequently, ice in vitrification
solutions grows most rapidly at temperatures not far below Tm [2].
This is the opposite of the behavior of ice nucleation, which has a
maximum rate at temperatures near TG. This separation between
the temperature optima for ice nucleation and growth has impor-
tant implications for determining the cooling and warming rates
necessary for avoiding appreciable ice formation.
During cooling, ice growth will be limited by the small num-
ber of nuclei (typically those arising from heterogeneous nucle-
ation) that form at higher temperatures where the growth rate is
significant. Nucleation continues and accelerates during cooling
until TG is passed, but at those temperatures, the nuclei are too
cold to grow. Upon warming, nucleation resumes when the tem-
perature range near TG is again traversed. This means not only
that many more nuclei are present during warming than are
Fig. 5 Critical cooling and critical warming rates to avoid ice formation during cooling or warming various
concentrations of ethylene glycol in water. Rates increase exponentially as concentration decreases, with the
calculated critical warming rate becoming practically unattainable at low concentrations. Data from Baudot
and Odagescu [188]
present during cooling [2, 182, 187] but also that during warming,
vastly more nuclei will be present at temperatures that favor rapid
ice growth. This means that total ice development is much more
rapid during warming, and hence warming rates required to avoid
significant devitrification are found to be far higher than cooling
rates initially required to achieve vitrification (e.g., Fig.5, [188],
and [38]). For glycerol, propylene glycol, and ethylene glycol, as
the critical cooling rate increases from 10C/min to 100C/
min, the critical warming rate increases from ~102103C/min
to ~1052107C/min [15, 72] (a 5-log increase for ethylene
glycol and glycerol and perhaps a 3-log increase for propylene
glycol). Fahy found that for vitrification solutions of propylene
glycol and Me2SO, which vitrify at about 10C/min, the critical
warming rate is 1,000C/min (ignoring the effect of the carrier
on the latter) [189, 190].
The nucleation [191] and growth [192] rates of ice in vitrifica-
tion solutions are also dependent on the carrier solution. Carrier
solutions based on NaCl, which include many typical culture
media, have minimal effects, but carrier solutions based on sugars,
which displace more water and have a stronger effect on solution
viscosity, are comparable in effect to CPAs, gram for gram, in
increasing vitrification tendency and reducing the critical warming
rate [72, 191].
Another feature of rewarming after forming large numbers of
small nuclei on cooling has to do with the visual appearance of the
solution. Converting a relatively dilute CPA solution into a heavily
nucleated glass does not necessarily change the visual appearance
of the solution because the crystals can be too small to scatter light,
making it possible for the solution to remain transparent, albeit
there may be a blue coloration as a telltale sign of the presence of
ice nuclei that may be overlooked [164]. During rewarming, such
dilute solutions tend to opacify at a temperature higher than the
temperature shown by differential scanning calorimetry (DSC) to
be the temperature of devitrification (Fig.6) because opacification
requires the small crystals formed during devitrification to recrys-
tallize until the prevailing crystal size is larger than the wavelength
of light [164]. At higher CPA concentrations, the onset of devitri-
fication is synonymous with the onset of opacification because, at a
lower nucleation density, each nucleus must grow to a larger size
to evolve a detectable amount of heat. At rapid warming rates,
opacification may not be observed if insufficient recrystallization
takes place even if devitrification is extensive. Therefore, using
visual appearance as evidence of vitrification and the lack of devit-
rification is valid only for solutions that can be vitrified at modest
cooling rates (near 10C/min) and is not reliable for significantly
more dilute solutions [164].
Fig. 6 Lack of equality of solution opacification and devitrification when solution

concentrations are below the concentrations needed for metastable vitrification.
TOPAQUE is the temperature of maximum opacity during warming at 80C/min. TD
is the temperature of the maximum of the devitrification peak during warming at
80C/min, and TG is the temperature of onset of the glass transition of the
quenched solution as measured during rewarming at the same rate. All solutes
are dissolved in water. For propylene glycol (PG), ethylene glycol (EG), and glyc-
erol (G, squares), CV at a cooling rate of ~10C/min is ~4346% w/w, 5255%
w/w, and 5760% w/w, respectively. TOPAQUE and TD are therefore projected to
converge ~58% w/w below CV for PG and ~03% w/w below CV for EG.Drawn
from the data of [164]
On the opposite end of the spectrum is the use, for whole

organs, of a vitrification solution (M22) that has a critical cooling
rate of less than 0.1C/min and a critical warming rate of around
0.4C/min [14, 193]. For kidneys vitrified with M22, warming at
~10C/min has resulted in appreciable ice formation due to
incomplete distribution of the cryoprotectant into all parts of the
organ [17, 18]. In this case, the critical warming rates are low
enough to be addressed by radiofrequency or microwave frequency
electromagnetic warming [194]. Such warming has been achieved
by either electric coupling to ions and polar molecules in cryopro-
tected tissues [195198] or magnetic coupling to exogenous
magnetic nanoparticles [199].
2.6 Antinucleation In recent years, there has been growing interest in agents able to
andSpecific Ice inhibit ice nucleation or ice growth by specific molecular recogni-
Growth Inhibition tion of ice or ice nucleators, sometimes called ice blockers.
Antifreeze proteins and antifreeze glycoproteins are natural exam-
ples, but they are difficult to obtain in useful quantities and at
affordable cost for tissue cryopreservation applications.
As alternatives, the synthetic polymers polyvinyl alcohol (PVA)
and polyglycerol (PGL) have been recognized as ice growth and
ice nucleation inhibitors, respectively, for cryopreservation applica-
tions [43, 44], with PGL specifically effective against heteroge-
neous nucleators of bacterial origin. The utility of ice blocking
compounds is that they can have large effects on ice formation
even while present in small quantities (Fig.7), even quantities as
small as one part per million [43]. They can also inhibit ice recrys-
tallization [200], making them useful if devitrification cant be
avoided during warming or if cryopreservation by conventional
freezing is used [201]. They are being productively used in an
increasing number of vitrification applications [202208].
A flavonol glycoside antinucleator also significantly improved
survival of vitrified shoot apices at a concentration of only 0.05%
[209], and a number of other small molecules may have future
applications as practical ice blockers for vitrification [16, 4547],
but currently they have not been tested for this purpose. A new
synthetic molecule, polylysine with about 65mol% of its side chains
replaced with carboxyl groups, has shown recrystallization
inhibition and good cryopreservation by freezing [210, 211] as
well as successful applications to vitrification [212215].
Fig. 7 Semivitrified 500g samples of ethylene glycol (EG) solutions cooled to 128C.Small quantities of poly-
vinyl alcohol (PVA) and polyglycerol (PGL) ice blockers dramatically reduce the amount of ice formed during cool-
ing. The PVA and PGL used were, respectively, X-1000 and Z-1000 ice blockers from 21st Century Medicine, Inc.
2.7 Thermally Like most matter, cryoprotectant solutions contract with cooling,
Induced Volume possessing a linear thermal expansion coefficient of ~90ppm/C
Changes, Strain, [216]. Below the glass transition temperature, the thermal expan-
andFracture sion coefficient is observed to fall to ~40ppm/C [216] due to
Formation kinetic inhibition of molecular movements required to continue
liquid behavior. This causes mechanical stress in samples during
vitrification. Colder outer parts of a sample will solidify (take on a
lower thermal expansion coefficient) before warmer inner parts of
a sample do. The resulting tendency of the interior to contract
more than the exterior after passage through the glass transition
and subsequent approach to thermal equilibrium creates stress.
Cryoprotectant glasses have a fracture strain of ~0.3% and fracture
stress of ~3MPa [217], which is only one tenth that of silica glass
(due to hydrogen bonding in the former versus covalent bonding
in the latter). This makes samples prone to fracturing during vitri-
fication [144, 145, 147]. Fracturing is undesirable for cell samples
because fracture planes nucleate ice [147] and unacceptable for
tissue or organs.
Stress during cooling to temperatures far below TG has been
found to be proportional to the cooling rate and the square of the
linear sample size [218]. It is best managed by slowing the cooling
rate as TG is approached [145] and ensuring sample temperature
uniformity as best as reasonably possible during passage below TG.
Ensuring that samples dont adhere to container surfaces dur-
ing cooling is especially important for fracture avoidance [71]
because containers typically respond differently to cooling than
cryoprotectant samples. Hydrophobic polymers that permit sam-
ple retraction away from container walls during cooling are to be
preferred over hydrophilic materials such as borosilicate glass.
3 The Biological Principles ofVitrification
3.1 Are As reviewed above, high-temperature vitrification can be achieved

Cryoprotectants in some cells and organisms simply by drying, and some organisms
Necessary can also dry into a glassy state during the winter at low terrestrial
forVitrification? temperatures. But low-temperature vitrification has also been
attempted in the absence of added cryoprotectants for some cells
that are not desiccation tolerant. It has been known for many years
that very small biological systems and even pure water can be
cooled rapidly enough to avoid ice crystals that are visible in the
electron microscope [8], but that is a different problem than vitri-
fying cells and recovering them in a viable state upon rewarming.
The problem of vitrifying cells without pCPAs is a topic of some
current interest and discussion, and a detailed examination of this
question is appropriate for illustrating additional principles of cryo-
preservation by vitrification.
The vitrification tendency of cryoprotectant solutions has been

studied extensively as a function of cryoprotectant concentration,
but living cells contain solutes in addition to any cryoprotectants
they may take up, and these natural solutes definitely augment the
vitrification tendency of cytoplasm in the presence of added pCPAs
[9, 83, 143, 219]. The quantitative contribution of intracellular
solutes to intracellular vitrification in different cells or in different
organelles has, however, not been well studied. Cytoplasm typically
contains 15% w/w protein and just 80% water, the endoplasmic
reticulum contains only 65% water, nuclei contain only 61% water,
and mitochondria contain only 59% water [189]. Exactly how
these low water contents influence vitrification tendency deserves
to be considered more carefully. However, since most cells and
their organelles are in osmotic equilibrium at ~300 milliosmolal
(mOsm), they do not have an extraordinary water activity and
readily experience IIF and behave very much as though their water
is present as an ideal dilute solution [117119, 220].
Ice formation is also affected by the presence of solid surfaces,
whose organization of local water structure tends to inhibit ice for-
mation. According to Merymans discussion [221] of the work of
Hori (in 1956), the spontaneous freezing temperature of water
between glass plates separated by 10m is 30C, and a plate sepa-
ration on the order of 0.14m results in no ice formation even at
100C and a negligible water vapor pressure. Water structuring
by other types of surface has been investigated at some length [222,
223]. The water in intracellular compartments may be influenced
by this same general type of effect near the plasma membrane,
organelle membranes, and the cytoskeleton, further favoring vitrifi-
cation, and the effect might be particularly significant for sperm
considering their highly elongated, thin shape and packed DNA.
The water content of sperm, as estimated from their osmoti-
cally inactive volume (the b value in the Boyle-vant Hoff equation
described below), is just 2355% according to Isachenko etal.
[224] or 60% according to Morris [225]. This has led to specula-
tion that sperm might be vitrifiable internally in the absence of
pCPAs even at relatively low cooling rates (~150250C/min
[224]), but direct evidence for this speculation is lacking. Others
have cooled sperm at high rates without pCPAs and obtained sur-
vival, but again have not verified that the survival was due to intra-
cellular vitrification and the prevention of devitrification [226229],
and another report notes ice formation in the sperm tail, midpiece,
and neck after rapid freezing [230]. The possibility of intracellular
vitrification given cooling rates of up to 7.2105C/min [228]
seems plausible but has to be compared to the fact that the esti-
mated cooling rate required to vitrify 0.3M glycerol in an isotonic
salt solution, which would have a lower activity of water than iso-
tonic cells including sperm cells, can be estimated from curves
originating from the work of Toner to be on the order of

21062107C/min or more [15]. In addition, viability was
found to be more or less independent of cooling rate, which is
consistent with the survival of IIF.
A general problem with attempting to prove that cells can survive
vitrification without cryoprotectants is that if cooling is sufficiently
rapid, IIF leads to ice crystals that are too small to kill the cell o
utright,
and if warming is sufficiently rapid to preclude recrystallization, these
crystals will continue to be innocuous during rewarming, so IIF can
be mistaken for vitrification if the main criterion for vitrification is
survival [15, 189] (see further discussion and examples in the next
section). Although the water content of sperm is limited, the extent to
which that water is immobilized by contact with intracellular solutes
and structures may not be reflected by the water content per se.
Morris [231] checked human sperm for intracellular ice using
freeze-fracture/freeze-etch electron microscopy and freeze sub-
stitution after freezing at various rates after previously equilibrat-
ing them with ice at 7C with and without glycerol pretreatment.
He observed no clear evidence for intracellular ice, even after
rewarming to 40C.The results, though, do not speak directly
to the question of whether sperm can be vitrified without pCPAs,
because cells pre-equilibrated at 7C were also equilibrated,
prior to rapid cooling, with either about 20% glycerol, both extra-
cellularly and intracellularly, or with an osmoticum about 13 times
more concentrated (~3.8Osm) than native sperm contents, and
both exposures would dramatically increase resistance to IIF com-
pared to sperm quenched under isotonic conditions [225].
Moreover, sperm frozen to 7C and below without glycerol
were not viable. In this regard, sperm are no different than mus-
cle: slow freezing of muscle results in vitrification of the unfrozen
residual liquid once the water content has been reduced to about
20% by mass [29], but muscle cells rendered vitrifiable by such
extreme dehydration would not be viable.
Morris has estimated intracellular Th values for isotonic, pCPA-
free sperm as a function of intracellular protein concentration,
finding Th~43C for 30% protein and Th~58C for 60% pro-
tein [225]. However, a Th of 58C would be the equivalent of
~30% w/w glycerol, which would require a Tm of ~7C, vs. the
0.5C characteristic of isotonic cells, and Morris estimates the
intracellular protein concentration in real cells at around 20%. He
measured the TG of a previously freeze-concentrated solution of 10%
glycerol+40% bovine serum albumin at about 29C, but did not
(and could not) measure the TG of the unfrozen solution. For
comparison, the TG of isotonic muscle tissue is well below 100C
[29] and probably close to 130C.
Another footnote to this discussion is that reference to vitrify-
ing sperm in the absence of CPAs would normally imply vitrifying
not just the cells but their environment. Proponents of this method,
however, are not normally aiming to vitrify the extracellular milieu,
so their use of the term vitrify is not synonymous with the use of
this term for other forms of biological vitrification. If cooling were
done sufficiently rapidly to vitrify the extracellular solution, it
would presumably also be rapid enough to vitrify cells within the
solution as well, but devitrification would still be expected.
The titles of many papers cited in this section assert the achieve-
ment of vitrification in the absence of any proof. It is one thing to
postulate vitrification but something else to claim it. Luyet made
the error of claiming vitrification only to be disappointed when his
suppositions were disproven. As a matter of scientific rigor, unsub-
stantiated claims should in general not be made.
3.2 Vitrification As noted above, doubly unstable solutions are nucleated homoge-
into Doubly Unstable neously and therefore require very high warming rates to prevent
Glasses andOne- devitrification, if devitrification can be prevented at all. Warming
Way Vitrification vitrified samples as fast as possible, such as by immersion in a warm
bath at temperatures as high as +50C [232] for dilute vitrification
solutions of low toxicity, may be hazardous yet still insufficient to
prevent devitrification.
However, it has been apparent for some time [189] that strict
avoidance of devitrification is not necessary [15, 72], and this has
been underscored in recent times by, for example, the successful
vitrification of cells using solute concentrations as low as 2M
(15%w/v) propylene glycol plus 0.5M (17%w/v) trehalose
[233]. This solution has a critical cooling rate on the order of
300,000C/min, the limit of the equipment used (unstable vitri-
fication). If we assume that the achievable warming rate is within
an order of magnitude of this rate, then considering that, for exam-
ple, 40% w/w ethylene glycol, which can be expected to be much
more resistant to devitrification, has a critical warming rate of
1010C/min (Fig.5), it seems clear that the cells in these experi-
ments survived despite extensive devitrification.
Figure8 documents the effect of warming rate on the survival of
cells that were rapidly cooled under conditions that led to IIF when
vitrification solutions were not used and vitrification was not the objec-
tive. As can be seen, despite extensive IIF, cells were able to survive in
high proportions as long as they were warmed at rates in the vicinity of
1,000C/min, which presumably rescued these cells from the recrys-
tallization of intracellular and perhaps extracellular [134] ice.
Similarly, survival of devitrification is presumably explained by
limitation of the sizes of ice crystals. Homogeneous nucleation will
result in very large numbers of very small ice crystals, but below a
critical size on the order of 100 [234] to 300 [235] nanometers,
intracellular ice crystals have been found to be survivable.
Recrystallization of these small ice crystals is rapid and can kill the
cell [117, 118] when crystal sizes exceed ~400nm [235], so pre-
Fig. 8 Rescue of a wide variety of intracellularly frozen cells by rapid warming. For original references,
see [189]. Drawn from tabular data in [189]
sumably the critical warming rate necessary to survive devitrifica-

tion is equal to the warming rate necessary to avoid fatal
recrystallization, which is why it can be orders of magnitude lower
than the warming rate necessary to avoid devitrification per se [15,
72, 165, 181]. An interesting benchmark for future research to
define is the critical warming rate necessary to survive devitrifica-
tion, which to date has not been systematically evaluated or defined.
Methods that preserve cells by forming partially crystallized
glasses [72, 236] that can further devitrify upon rewarming deserve
to be distinguished from methods that achieve and maintain a pre-
dominantly amorphous state. Quench-limited freezing may be a
suitable term for such methods if ice formation on cooling is par-
ticularly extensive; doubly unstable vitrification [72] or, as sug-
gested here, unstable vitrification (Fig.3) would be descriptive
when ice formation on cooling is more subtle. If, in such cases,
significant devitrification cannot be avoided, the term one-way
vitrification has been suggested to describe the fact that the living
system or its environment does not remain amorphous during
warming [72].
Describing rapid freezing or unstable vitrification methods
simply as vitrification obscures the true physical nature of the
process being employed.
3.3 Carrier Solutions All cells normally survive in an environment to which they have
andCryoprotectants been adapted. In the mammalian body, this environment has some
common characteristics, including, typically, a total osmolality a
little less than 300mOsm (which depresses the freezing point to
about 0.55C), a high sodium concentration (~145mM), a low

potassium concentration (~4.5mM), a high chloride content, a
pH in the vicinity of 7.4, and a variety of other electrolytes, pro-
teins, signaling molecules, etc. To survive exposure to low-
temperature conditions, this environment must in some form be
maintained within viable limits, and that is the function of the
carrier solution, which is the physiological support medium in
which cryoprotectants are dissolved. Cells suspended in pCPAs
alone would not be able to maintain their volume, pH, ionic con-
tent, and membrane integrity, so there must be a basic solution
that is maintained around the cells in both the presence and absence
of cryoprotectants. The carrier solution in a sense carries the cryo-
protectants to and from living cells while allowing them to avoid
injury that is unrelated to the cryoprotectants per se. Carrier solu-
tion compositions vary widely, but all are designed with this basic
supportive role in mind.
Cryoprotectants must generally be added to carrier solutions
to enable vitrification. To be an effective pCPA, the agent must
be miscible or soluble in water to high concentrations, be of low
toxicity, be able to remain in solution even at very low tempera-
tures, and, as noted above, be able to cross the cell membrane
[39, 125]. npCPAs have similar requirements but are normally
used at much lower concentrations and by definition do not
enter cells. More details on the correct use of pCPAs and npCPAs
are provided below.
The toxic effects of cryoprotectants have been shown to
depend on the choice of the carrier solution [141, 237239].
Often, the reason for this dependence is unclear, but carriers that
do not support cell viability under hypothermic conditions may
add extraneous hypothermic injury to any injury that may be due
to CPA exposure per se.
As noted above (Subheading2.5), the carrier solution also
plays an important role in limiting ice nucleation and ice growth
rates during both cooling and warming. An influence of this effect
on survival rates, though, remains to be demonstrated.
A major function of the carrier solution in addition to main-
taining viability is to control cell volume. The principles of prepar-
ing carrier solutions to accomplish this in the presence and absence
of cryoprotectants are described in the next section.
3.4 Osmosis, Osmosis is the movement of water from a region of high water
Osmotic Limits, concentration or vapor pressure to an area of low water
andOsmotic Protocols concentration or vapor pressure. The importance of osmosis is par-
ticularly high for vitrification protocols because of the compara-
tively enormous concentrations of cryoprotectants required in
comparison to freezing protocols, which necessarily reduce water
concentration greatly, and even to close to the limits of compatibility
with life. Cavalier use of both pCPAs and npCPAs without adequate
avoidance of osmotic shifts is a frequent preventable cause of injury
associated with the use of vitrifiable concentrations of cryoprotec-
tants. More detailed descriptions of osmotic effects during the
introduction and removal of cryoprotectants are available in many
publications (e.g., [116, 121, 240, 241]), but here we will simply
describe the basic phenomena of relevance.
The rate of water movement across the cell membrane depends
on the transmembrane difference in water vapor pressure, or osmo-
lality, without regard to the nature of the solutes whose presence
generates the transmembrane vapor pressure gradient. For pCPAs,
the effect of the CPA on the transmembrane osmotic gradient is
transient. Initially, since water moves more rapidly than the CPA,
extracellular pCPA raises extracellular osmolality more than intra-
cellular osmolality, and the cell loses water in response. Later, as
the pCPA enters the cell down its own transmembrane concentra-
tion gradient, water diffuses back into the cell to maintain osmotic
equilibrium. This sequence of events is often referred to as the
shrink-swell cycle.
The end result of this process then depends on the carrier
solution [121]. When the pCPA concentration is (nominally)
the same, per unit liquid volume, on both sides of the mem-
brane, the cell will have returned to its original volume provided
the osmotic effect of the carrier solution, which was equal to the
osmotic effect of intracellular molecules before CPA addition, is
also the same, per unit solution volume, as it was prior to pCPA
addition. Although cells contain proteins and many other com-
plex molecules whose osmotic coefficient might be expected to
differ from the osmotic coefficient of the carrier solution, in
practice, this difference is small, and the cell can reasonably be
modeled, to a first approximation, as a dilute salt solution having
the osmolality of plasma (~285mOsm) [117, 121, 220]. If the
osmotic coefficient of the carrier is about the same as the osmotic
coefficient of cytosolic solutes, then for the cell to return to its
original volume, the carrier must have the same concentration
per unit volume of extracellular solution as it did prior to the
addition of the cryoprotectant [121].
In other words, the correct way to compose a cryoprotectant
solution, if the goal is no volume change after pCPA equilibra-
tion, is to add all the carrier solutes needed for a given volume,
add all the CPAs needed for the same given volume, and then
bring the solution to the final desired volume by the addition of
water. This method essentially replaces water, volume for volume,
with pCPA, such that the molar concentration of carrier solution
solutes is unchanged.
This is important to note for three reasons. The first reason is
that making vitrification solutions on a % w/w basis, which may be
meaningful for physical reasons, is not meaningful for biological

purposes, because carrier solution and other impermeant solutes
included on this basis are not readily maintained at a molar concen-
tration that can be evaluated with respect to the effect of the solu-
tion on the volumes of cells that may be placed into it. A procedure
for preparing solutions on a % w/w basis while still maintaining
isotonicity has been described [196] and should be employed, with
suitable modifications if a defined tonicity other than isotonic is
desired. The procedure to maintain isotonicity in the presence of
X% w/w pCPA is: (a) prepare a concentrate of the carrier solution,
such as dissolving the nonaqueous components of 1L of carrier in
water so as to obtain a final total volume of 200ml (this would be
a 5 carrier concentrate); (b) weigh the carrier solution concen-
trate; (c) add a weight of pCPA equal to the weight of the carrier
times (X/(100X)) (thus creating a solution that has the correct
% w/w concentration of pCPA); and (d) q.s. to 1L with X% w/w
pCPA in water (thus creating an isotonic solution with the correct
final % w/w).
The second reason to understand the principles of isotonicity
is that it is common for cell freezing labs to add a pure cryoprotec-
tant to a cell culture medium prior to freezing, which dilutes the
culture medium. Since cell freezing often employs, for example,
10%v/v Me2SO, the error is not very significant in that case. But
when composing a much more concentrated vitrification solution,
diluting the carrier by, for example, 50% would require intracel-
lular solutes to be diluted to the same extent to maintain osmotic
equilibrium at the end of pCPA equilibration, which means a dou-
bling of cell liquid space [121]. Not only might this be harmful for
the cell, but it will also dilute the intracellular solutes that also
contribute to vitrification tendency of the cytoplasm [83, 167],
making the cell at greater risk of IIF during both cooling and sub-
sequent warming. In fact, it seems desirable to vitrify cells when
they are still in the shrinkage phase of the shrink-swell cycle rather
than after complete pCPA equilibrium, in part to reduce exposure
time to the pCPA and in part because it is at that time that the
cytoplasm is the most concentrated [49, 83, 167], but this strategy
will not be completely efficient if the carrier solution has been
diluted during the shrinkage phase. See the next section for a spe-
cial case in which deliberate hypotonicity during vitrification solu-
tion loading can be used to actually expedite CPA equilibration.
The third reason to understand how to adjust tonicity in the
presence of CPAs is that chilling injury and its avoidance or mini-
mization depend on the effective tonicity of the vitrification solu-
tion [18]. More details about this topic are given below.
When the pCPA concentration is diluted, extracellular water
at first rushes into the cell down its concentration gradient, caus-
ing the cell to swell. Over time, as the pCPA in the cell diffuses to
the extracellular space, taking water with it to maintain equilib-

rium, the cell volume begins to return back toward normal.
Eventually, if the extracellular carrier solution is isotonic
(unchanged in molarity from its molarity in the absence of pCPA),
the cell will in principle return to its original volume. This swell-
shrink sequence is the reverse of the shrink-swell behavior
observed during pCPA administration.
Although cells can be damaged by either excessive shrinkage or
excessive swelling, cells can generally withstand about a fourfold
increase in extracellular osmolality but only around a twofold
decrease (the osmotic limits of the cell) [241]. This means that
a given fold change in extracellular concentration is more hazard-
ous during the removal of pCPA than during its introduction. To
offset cell swelling during pCPA removal, a non-penetrating agent
such as mannitol, sucrose, or trehalose can be included in the car-
rier solution to increase the effective osmolality of the extracellular
medium and thus offset, to some extent, the decrease in effective
osmolality caused by dilution of the pCPA, thus limiting the extent
of swelling [241]. Used in this way, extracellular agents are some-
times referred to as osmotic buffers.
At equilibrium, the relationship between the volume of intra-
cellular water and the net extracellular osmolality, , is given by the
Boyle-vant Hoff equation [242], which, under isothermal condi-
tions, can be written as
Vc=b+Voo/,
where Vc is the cell volume, Vo is the volume of intracellular water
under normal (isotonic) conditions, b is the osmotically inactive
volume of the cell, and o is the isotonic extracellular osmolality.
This equation says that the volume of a cell is a linear function of
the reciprocal of the effective extracellular osmolality and that
when the extracellular osmolality approaches infinity, the volume
of the cell approaches its non-osmotic (usually its dry) volume, b.
Since the volume of intracellular water, Vw, is given by
Vw=Vcb, we can rewrite the equation as
V w / Vo = P o / P ,
which shows that the volume of intracellular water changes in
inverse proportion to the extracellular osmolality. For example, if
the extracellular osmolality decreases by a factor of 2, the cell water
content will double, and if the extracellular osmolality increases by
a factor of 2, the cell water content will be halved. This rough rule
of thumb is handy for estimating acceptable changes in concentra-
tion, assuming b is relatively small.
When pCPAs are added in steps, each step contributes to the
value of b, because as successive steps of concentration are added, pre-
viously added intracellular pCPA cannot leave the cell, since intra
cellular concentrations can never exceed extracellular concentrations.
As b increases to a maximum following each step of pCPA addition,

the redefined volume of intracellular water under isotonic conditions
(in the presence of the pCPA) correspondingly decreases (Vo=Vcb).
When the next step of pCPA addition takes place, there is less intra-
cellular water to extract, so the osmotic consequence of a given
step change in pCPA concentration is reduced even if the fold
change in water content is the same. This enables the step size to
be safely increased as pCPA a dministration proceeds. In addition,
since, from the above expression, Vw responds to the fold change
in effective extracellular osmolality and not to the absolute change,
the concentration of pCPA can be incremented exponentially
rather than linearly. This procedure, by speeding pCPA addition
and thus reducing total exposure time to the cryoprotectant, can
actually result in less toxicity [105].
During washout, intracellular water content is again responsive
to the fold change in extracellular osmolality, so washout can pro-
ceed exponentially as well, but the b value decreases rather than
increasing with time as intracellular pCPA is subtracted, which is one
reason the hazard of excessive cell swelling increases as the concen-
tration approaches zero. For this reason, including an osmotic buffer
throughout the pCPA washout process, including the step when the
pCPA concentration reaches zero, is generally recommended.
A non-penetrating solute, unlike a pCPA, cannot be added
without lowering cell volume relative to what it would otherwise
have been. If the pCPA that accompanies an npCPA happens to
reach transmembrane concentration equilibrium (which it will
generally not), the npCPA will have just as much of an osmotic
effect on the cell in the presence of the pCPA as it would have had
in the absence of the pCPA, even if the npCPA contributes a very
small fraction of the total extracellular osmolality, since the effects
of intracellular and extracellular pCPA cancel each other out. Also,
under the influence of the npCPA, pCPAs behave like water: any
decrease in cell volume caused by the npCPA will concentrate the
intracellular pCPA, causing efflux to the more dilute extracellular
pCPA pool and thus a reduction in that part of the b value that is
contributed by the pCPA.
With the above concepts in mind, it is possible to explore
unusual osmotic protocols that are intended to enable more rapid
cryoprotectant addition and removal without exceeding but while
taking advantage of the osmotic limits of the cell. For example, in
a scheme by Meryman [241], a cell is exposed to a concentration
of pCPA that brings the cell transiently to its hypertonic osmotic
limit (about four times isotonic) but contains a carrier solution that
is only half as concentrated as an isotonic solution. After
equilibration, the cell swells to twice its isotonic volume (neglect-
ing b), which then allows the osmotic concentration of the medium
to be raised by a factor of 8in the next step rather than 4: an
increase of twofold would merely bring the cell back to its isotonic
volume, so an increase of fourfold in addition to that is needed to
once again bring the cell to its hypertonic osmotic limit (again,
neglecting the b value of the cell, which would allow the fold
change to be even greater). Upon washout, the cell is placed into
a solution containing a four times isotonic carrier plus whatever
pCPA is needed to avoid the hypotonic transient osmotic limit of
the cell during swelling, after which the cell comes to its hyper-
tonic osmotic limit prior to the next dilution step. The next dilu-
tion step can then reduce total extracellular osmolality by a factor
of 8. In principle, the steps involved in this scheme could be even
greater, even beyond accounting for the b value, because in prac-
tice significant permeation of the pCPA will usually take place dur-
ing the shrinkage or swelling phases of the process, thus limiting
volume extremes.
Although the above guidelines are helpful for determining
the final equilibrium state after concentration changes and
boundary conditions on volume excursions during transient
shrinking or swelling, the design of a cryoprotectant addition
and removal protocol also generally requires some knowledge of
the permeability of the system to the cryoprotectants employed.
The timing, magnitude, and temperature of concentration steps
can be considerably improved by using computer modeling of
the shrink-swell and swell-shrink processes under a variety of vir-
tual conditions (e.g., [243246]), and this approach is recom-
mended whenever possible. When this is not possible, our
experience has been that a protocol in which pCPA concentra-
tion is doubled on every addition step and halved on every dilu-
tion step is effective in avoiding osmotic injury during the
preparation of rabbit renal cortical slices [190].
Clever computer-modeled schemes have been proposed to
exploit known permeation rates and the osmotic limits of the cell
so as to maximally accelerate CPA addition and removal [247,
248], though most have not been tried. One particularly interest-
ing and innovative current scheme that also factors in toxicity min-
imization is described in the next section.
The problem of introducing and removing cryoprotectants is
more difficult when the pCPA and npCPA must be administered
by perfusion. In a kidney, for example, exponential addition of
pCPA is counterproductive because lags in distribution of the
CPA to the medulla result in little benefit of the exponential addi-
tion rate, whereas accelerating exposure of the renal cortex to the
CPA and then holding the cortex at the highest concentrations
long enough for the medulla to catch up can be lethal [17,
194]. As a general rule, the osmotic protocol for each system
should be t ailored to the specific needs and characteristics of that
system.
3.5 Procedures The central problem of vitrification has always been inducing living
forAvoiding cells to tolerate enormous concentrations of CPAs and low con-
Cryoprotectant centrations of water. In 1984, Fahy etal. suggested seven
Toxicity approaches to controlling cryoprotectant toxicity in vitrification
procedures, i.e., avoid osmotic injury; employ cryoprotectant mix-
tures so their mutual dilution minimizes specific sources of toxic-
ity; use one or more npCPAs to allow reduction in the intracellular
pCPA concentration (see next section for more details); maintain
temperature as low as possible; select an appropriate carrier solu-
tion; keep exposure time to the CPA to a minimum; and, when
possible, employ cryoprotectant toxicity neutralization (CTN; see
Subheading3.7 below).
Implicit in this list at the time was also the very first step in
designing a vitrification solution, and that is to determine exactly
how much CPA is needed (CV). Extensive lists of solutions that
are, within 1%w/v total concentration, exactly sufficient in con-
centration to avoid visible ice crystals (i.e., that are at their CV) on
cooling at about 10C/min have been published (e.g., [70, 83,
127]), and means of interpolating between known solutions to
estimate the CVs of arbitrary mixtures of CPAs have also been
described (e.g., [143, 190]). However, some investigators deter-
mine CV only to the nearest 5%, potentially exposing their cells to
as much as 4%w/v or v/v (or w/w) more CPAs than needed,
which may have strong [249] and unnecessary effects on the total
toxicity observed. CV is a function of solution composition, cool-
ing rate, and applied pressure [70, 97, 99, 127, 191, 250], so it
must frequently be redetermined for new circumstances.
These methods have been used successfully, but since 1984,
five additional approaches have been added: use ice blockers to
reduce the overall quantity of pCPA required [43, 44], use methox-
ylated CPAs in moderation [251], employ creative addition and
washout methods to minimize the overall cost function of toxic-
ity for the solution in question when the system is a cell suspension
or simple tissue [252, 253], preferentially employ weak glass form-
ers [203], and use a combination of constant pressure and constant
flow methods when CPA must be introduced by perfusion so as to
speed equilibration of poorly circulated tissues [194]. Organ per-
fusion techniques are still being demonstrated, so further elabora-
tion will be provided here only on the use of ice blockers,
methoxylated CPAs, cost function optimization, and preferential
use of weak glass formers. The latter is further discussed from a
theoretical point of view in Subheading3.8 below.
As noted above, ice blockers interact with, and inhibit, extra-
cellular ice, but they do not have access to the intracellular space.
This inaccessibility is immaterial, however, because cells do not
appear to contain heterogeneous nucleating agents of any
significant effectiveness. Evidence for this conclusion comes from a
variety of observations (see, e.g., Table1 and accompanying discus-

sion in [34]), but the most direct have been experiments in which
single cells were contained in the type of emulsions used to detect
homogeneous nucleation and were found to have minimal nucleat-
ing activity in this system [171, 254]. The fact remains, however,
that relying on the use of excessive extracellular ice growth inhibition,
whether by adding ice blockers or by adding viscosity-enhancing
agents, runs the theoretical risk of making the extracellular medium
more resistant to ice nucleation than the cytoplasm and thus poten-
tially deceptive in predicting the fate of cells cooled in it.
Replacing hydroxyl groups on OH-bearing pCPAs with
methoxy groups has two favorable effects: glass-forming tendency
is greatly improved and membrane permeability is greatly increased
[251]. There are, however, two counterbalancing negative effects:
increasing the glass-forming tendency of pCPAs generally increases
toxicity in the context of vitrification (see Subheading3.8), and
increased membrane permeability is due to increased hydrophobic-
ity, which also tends to make pCPAs more toxic [255]. However, with
at least one methoxylated pCPA, methoxy-glycerol (3-methoxy-
1,2-propanediol), a net positive effect can be obtained [18].
Methoxy groups are more effective for vitrification not entirely
because of their superior ability to hydrogen bond with water but
also because interactions between the methoxy groups are impos-
sible, since these agents present only the lone pair electrons of their
oxygen atoms to the solution for bonding, which do not hydrogen
bond to one another, and therefore these oxygens are more free to
interact with water and therefore are not wasted. Hydroxyl
groups, on the other hand, can hydrogen bond to each other, and
these self-associations reduce interaction with water, thus requiring
more pCPA to induce vitrification. In 3-methoxy-1,2-propanediol,
two hydroxyls are present to maintain reasonable hydrophilicity,
helping to offset the hydrophobic effect of the methoxy group.
This modality for toxicity control appears effective [18] but has
been little investigated to date.
Regarding cost function minimization [252], the toxicity of a
cryoprotectant is a function of its concentration in the cell and its
time of exposure. Both aspects of toxicity can be minimized by
adding cryoprotectant while the cell is maintained at its hypotonic
volume limit: this increases the gradient for pCPA uptake while
keeping the intracellular concentration as low as possible. The
intracellular concentration is then abruptly raised to the target
value by osmotically shrinking the cell, which is a rapid process
since water crosses the cell membrane very quickly. To remove the
pCPA, the cell is again immediately brought to its hypotonic vol-
ume limit, which immediately dilutes the intracellular pCPA, and
the cell is maintained at this hypotonic limit while pCPA concen-
trations are lowered, after which the cell is restored to its desired
volume, which transiently increases the lingering intracellular

pCPA concentration, thus speeding its exit from the cell, while not
raising its concentration to a toxic level. This approach has yielded
positive results with oocytes, which were able to be loaded with
Me2SO in a remarkably short time yet recover very well [253].
Applicability to other systems remains to be explored.
Both empirically and theoretically, pCPAs that interact most
weakly with water are generally the ones that are the least damaging
when included in a vitrification solution (see also Subheading3.8).
The principle of favoring weak glass formers for the formulation of
vitrification solutions (VSs) has enabled the development of less
toxic VSs [18, 203], but there are limits on its application. For
example, the weak glass former, ethylene glycol, may have specific
toxic effects at higher concentrations [203] or may be required in
such high concentrations as to begin to deplete water to the point
where water becomes inadequately available despite weak interac-
tions with the pCPA.Glycerol, which is another weak glass former,
is frequently too impermeable, too viscous, or too toxic for other
reasons to be used in concentrations that might be more ideal in
theory. Despite such limitations, however, favoring weak glass
formers has, in combination with the other modalities discussed
above, including CTN, the use of 3-methoxy-1,2-propanediol, ice
blockers, and the use of another npCPA, enabled the creation of
M22, a 9.4M solution whose critical cooling and warming rates are
extremely low (~0.1C/min and 0.4C/min, respectively) 14, yet
which can be perfused through a kidney with only transient dys-
function after transplantation [18].
Despite these advances, there is still both an academic and an
applied need to better understand CPA toxicity so as to enable still
more effective countermeasures to it. This topic is discussed in
Subheading3.8 below.
3.6 Extracellular As noted in Subheading3.1, living cells contain significant amounts

Agents inVitrification of protein, metabolic intermediates, nucleic acids, and other sol-
utes. These solutes might, when water content is significantly
reduced by the use of pCPAs, contribute more to glass-forming
tendency of the cytosol than the carrier solution contributes to the
stability of the extracellular solution. In this case, the use of suffi-
cient pCPA to ensure vitrification of the extracellular solution would
actually require more pCPA than needed to vitrify the intracellular
solutions and would therefore be more toxic than necessary. To
correct this problem, it was suggested in 1981 [99] and reported in
1982 [100] that extracellular pCPA concentrations could be
reduced by using npCPAs as counterparts to intracellular solutes,
thus maintaining the vitrification tendency of the vitrification solu-
tion with less pCPA and still allowing cells in contact with this
solution to vitrify. This was later documented photographically,

and many examples of the use of npCPAs were presented [83].
npCPAs, in addition to facilitating vitrification, also tend to
increase the viscosity and the effective tonicity of the solution. The
effect on viscosity is beneficial for vitrification but may become
counterproductive when pCPA diffusion rates need to be maxi-
mized or when pCPAs have to be introduced by perfusion [17].
The effect on tonicity may be limiting if the living system to be
preserved is sensitive to chilling injury and chilling injury is a func-
tion of the tonicity of the medium [18] (see Subheading3.9 below).
When neither of these drawbacks intervenes, in principle, the
amount of npCPA that can be used is limited only by the upper
osmotic limit of the cell in question, the extent to which pCPA
levels can be reduced without compromising the vitrification ten-
dency of the VS, and the extent to which the npCPA may induce
specific toxic effects [206, 256] (see also below). Although npCPAs
do not directly perturb cytosolic proteins, interactions with the
external leaflet of the plasma membrane and with integral mem-
brane proteins that communicate with the cytoplasm are still pos-
sible. In principle, the amount of npCPA that can be used might be
increased by reducing carrier solution solutes in favor of more
npCPAs should cell shrinkage be the limiting factor.
By increasing npCPA concentration, cell volume will be
reduced and intracellular solutes will become more concentrated,
thus further stabilizing the cell against ice formation. In retrospect,
in fact, cell shrinkage may be a primary mechanism of action of
npCPAs in vitrification, since, as noted above, the effect of the car-
rier solution on ice nucleation extracellularly may not be greatly
different than the effect of the intracellular solutes intracellularly.
In addition, because the cell is shrunken, the absolute amount of
pCPA inside the cell will also be reduced, and this means that, after
vitrification, the cell can be diluted more abruptly and more exten-
sively before the hypotonic osmotic limit of the cell will be reached
(see Subheading3.4).
A study by Shaw etal. [257] brought out many important fac-
ets to the use of npCPA that should be kept in mind. Excessive
amounts of high-molecular-mass npCPAs tend to raise Tm and
increase the total solute concentration needed to vitrify (both
potentially adverse effects) while having minimal effects on TG.
However, TD (the temperature of devitrification) tended to be
raised, which is beneficial.
Kuleshova etal. studied sugars as npCPAs for vitrification of
oocytes and embryos [258]. Glucose, fructose, and sorbitol were
equal in effectiveness to ethylene glycol, mole for mole, in support-
ing vitrification and raised TG. Trehalose, sucrose, and raffinose,
being larger molecules, required an increase in total solute concen-
tration, when measured on a weight percent basis, to replace EG
mole for mole, thus depleting the solution of extra water, but also
raised TG, raffinose being particularly active in the latter respect.
3.7 Toxicity There are cases in which the addition of a nontoxic concentration
Neutralization of one cryoprotectant to a toxic concentration of another can neu-
tralize the toxicity of the latter [259, 260], while the then-
acceptable presence of the latter can then allow the former to be
maintained at a nontoxic concentration and still allow the solution
to vitrify [18, 190, 203, 206, 249]. The same principle has also
allowed improved recovery after freezing and thawing [104].
Cryoprotectant toxicity neutralization (CTN) is thus far restricted
to the neutralization of amide toxicity by Me2SO, and even for
amides, the effect is not universal [260]. CTN is strong for for-
mamide and urea, weak for acetamide and N-methylformamide, and
nonexistent for dimethylformamide and N-methylacetamide [260].
The mechanisms that underlie these effects are presently
unknown. The idea that a compatible solute effect might be
involved analogous to the protection of proteins against urea by
solutes in nature that have a protein-stabilizing effect sufficient to
offset the protein-denaturing effect of urea [190, 261264] was
not experimentally supported [249]. Pursuant to the original sug-
gestion [265] that CTN may involve physical interaction between
amides and Me2SO, it was determined that the magnitude of the
exothermic heats of mixing when Me2SO is mixed with formamide,
ethylene glycol, and propylene glycol has the same rank order as
the viability of kidney slices exposed to mixtures of Me2SO with
these same solutes (formamide>ethylene glycol>propylene gly-
col), and the heat of mixing between Me2SO and
N-methylformamide, whose toxicity is only marginally neutralized
by Me2SO [260], was minimal [190]. On the other hand, when
similar experiments were done in the presence of water, the affinity
between Me2SO and water is so strong that the interaction between
Me2SO and formamide in aqueous solution is thermochemically
repulsive (endothermic) [249]. Still, there remains a correlation
between the effectiveness of CTN for a given amide and its strength
of interaction with Me2SO in aqueous solutions: as the interaction
becomes more thermochemically repulsive, CTN becomes weaker
and disappears (cf. [249] and [260]).
Despite the current lack of information about the molecular
basis of CTN, a specific molecular example of CTN is available that
may be instructive in future studies. The inactivation of membrane
Na+,K+-ATPase by urea is blocked by Me2SO [266].
Although the combination of amides and Me2SO was origi-
nally proposed to neutralize the toxic effects of Me2SO [249, 265],
this effect has not been observed subsequently [249, 267].
Although there is no known toxicity neutralizer that can protect
against Me2SO toxicity at this time, acetylcholinesterase inhibition
by Me2SO can be physiologically blocked by atropine [268], some

oxidative effects of Me2SO can be reversed with reducing agents
[259, 260], and there is one still-unconfirmed report that glucose
can prevent irreversible binding of Me2SO to proteins and also
reduce its toxicity [237].
So far, few systems have been evaluated for their ability to ben-
efit from CTN.CTN is known to apply to rabbit renal cortical,
liver, and brain slices and to murine osteoblasts [206] and seems
very likely to pertain as well to rat liver [207], renal cortical and
medullary [269], and brain [205] slices, and amide+Me2SO mix-
tures have also been successfully applied to other systems ([16,
140], and unpublished results). The only tested systems that so far
do not seem to benefit from CTN are human oocytes and early
stage embryos (S.F.Mullen, unpublished results).
3.8 Specific Ironically, the central problem of vitrification, cryoprotectant tox-

andNonspecific icity, remains, in 2014, practically unexplored mechanistically. This
Biochemical Effects attests to the decades-long lack of interest in cryobiological prob-
ofCryoprotectants lems that has prevailed on the part of biochemists, with very rare
exceptions (most notably, [265]). Criteria proposed 24years ago
now [249] for showing that a given biochemical modification is an
important cause of observed cryoprotectant toxicity under practi-
cal cryobiological conditions have rarely, if ever, been met.
Fortunately, however, this picture is finally beginning to change.
Examples of cryoprotectant metabolism invivo or invitro at
elevated temperatures, such as the phosphorylation of glycerol
leading to ATP depletion [270, 271] or the transformation of eth-
ylene glycol into toxic by-products [272, 273], have no demon-
strated relevance to events at the lower temperatures at which
pCPAs are usually administered or in tissues or cells whose meta-
bolic activities are different than those studied. Me2SO, which has
a largely undeserved reputation as being particularly toxic and has
numerous pharmacological effects at body temperature [274] and
can chemically react with tissue sulfhydryl groups [275], under-
standably has no demonstrated pharmacological effects at 0C, at
which temperature its reaction with sulfhydryl groups may be too
slow to be meaningful for cellular viability [206, 259]. It has been
shown to react with steroids and triterpenoids [276], but never to
do so under biologically meaningful conditions. As summarized
elsewhere [249] (references given therein), pCPAs have been
shown to elevate membrane phase transition temperatures; rear-
range the cytoskeleton, including most significantly the meiotic
spindle; cause membrane blistering; fuse cell membranes; change
gene expression; alter RNA polymerase; weaken DNA-nucleosome
binding; destabilize nucleic acid duplexes; impair ribosome assem-
bly; and induce many other adverse changes, but for the most part,
the relevance of these observed changes, if any, to most cells being
prepared for cryopreservation is currently unknown. There seems
to be little or no generalized effect of vitrification solutions on pas-

sive membrane permeability to sodium and potassium under
practical conditions [190]. Some pCPAs can induce differentiation
of leukemia cells [277] and can change their chromatin and DNA
conformation [278], but these are not what would normally be
considered toxic effects. In summary, it seems that the literature on
the biochemical effects of cryoprotectants and the literature on the
toxic effects of CPAs in cryobiological applications are mostly dis-
connected from one another.
A proposed protein-altering mechanism of Me2SO toxicity
involving specific interaction between Me2SO and protein surface
lysine residues [265] has not been supported by subsequent inves-
tigations [249, 259] and is not favored by the general observation
that small molecules, including Me2SO [279], tend to be preferen-
tially excluded from the hydration layer surrounding protein sur-
faces, thus stabilizing them against denaturation [263, 280282]
and even enhancing renaturation after previous denaturation [283,
284]. Mixtures of pCPAs in vitrification solutions intended for use
at high hydrostatic pressures did show increased protein destabili-
zation tendencies, but these effects were correlated inversely with
toxicity [249]. In addition, the toxic effects of individual com-
monly used pCPAs do not appear to be accounted for by their
general protein denaturation tendency [249, 260, 279, 285, 286]
or by their ability to increase the permeability of membranes when
used below 15C [267], and Arakawa etal. have argued that dis-
ruption of the hydration layer surrounding proteins and mem-
branes by pCPAs may account for pCPA toxicity at high
temperatures but not at low temperatures [286].
Nevertheless, protein denaturation has not been ruled out in
ambient pressure vitrification solutions composed of mixtures of
pCPAs, which is the most common type of VS in use today. Even
though each individual pCPA in the VS mixture may be at a nomi-
nally non-denaturing concentration, the sum total of all of the
pCPAs might still have a cumulative denaturing effect, especially
when cold destabilization of proteins is also factored in, and some
evidence indicates that pCPAs do not protect effectively against
cold denaturation [287]. There is now, in fact, some evidence for
the possibility that high-concentration VSs might induce protein
denaturation under low-temperature exposure conditions. This
evidence has been derived in two independent ways.
The first line of evidence is indirect and arises from the finding
that the toxicity of a large number of VSs can be correlated strongly
with a compositional variable, qv*, which defines the number of
moles of water associated, on average, with each water-bonding
polar group on the pCPAs of a VS when that solution is exactly at
its CV [203]. The evidence indicates that pCPAs that interact
strongly with water [288], such as 1,2-propanediol, are more toxic

precisely because they interact strongly with water, presumably
competing with biological molecules for access to water, whereas
more weakly bonding solutes, such as ethylene glycol, are required
in higher concentrations to vitrify water because of their weaker
interaction with water, yet leave the water that remains more free
to hydrate biomolecules despite a lower overall concentration of
water in the solution [203]. It is unclear whether this mechanism
of nonspecific toxicity arises as a result of protein denaturation or
some other water-dependent effect(s), but it is compatible with a
role for protein denaturation in the manifestation of toxicity.
The second line of evidence is more direct and is based on a
recent microarray analysis of alterations in transcription following
exposure of rat liver slices to two candidate (8.8 and 8.9M) VSs
[207]. It was published as part of an analysis of chilling injury (see
next section), and so full results were not provided, but the results
that were presented are illuminating.
Although neither VS reduced slice ATP, 1985 transcripts were
changed, of which 92 increased at least 1.5-fold and 49 decreased
at least 1.5-fold. The VSs increased transcripts for 11 heat shock
genes, and one particular Hsp70 family member, Hspa1b, was ele-
vated 12.6-fold, although transcription of Dnajc12 (an Hsp40
homolog) was slightly decreased. Also consistent with the loss of
some proteins to denaturation, eight genes associated with ribo-
some biogenesis were induced as well, especially the 5S rRNA
gene, whose transcripts increased 4.5-fold. VS exposure decreased
transcription for genes in the p38 signaling pathway about 25%
and depressed TGF-1 and TGF-3 transcripts, which normally
lead to stimulation of the p38 pathway. In contrast, the VSs greatly
increased transcription related to the ERK and especially to the
JNK pathway, a third finding consistent with the possibility of pro-
tein denaturation. Though not clearly related to denaturation, it
was also of note that expression of Hmox-1 [heme oxygenase
(decycling) 1], which functions primarily as a major defense against
oxidative stress and injury [289], was decreased by 1.9-fold.
It is important to point out that, even if protein denatur-
ation is involved in the mechanism of VS toxicity, the proteins
undergoing denaturation may be a small subset of the total
[190]. Cryoprotectants decrease the solubility of tyrosine, leu-
cine, alanine, cystine, and glycine [190, 279], which should have
a generally inhibitory effect on denaturation, although 20%
Me2SO increases the solubility of tryptophan by about 40%
[279]. Different proteins respond differently to cryoprotectants
[255, 290], suggesting that the most susceptible proteins may
be selectively involved in toxic responses. More globally, and
dramatically, closely related tissues as a whole also react differ-
ently to cryoprotectants despite the fact that the same basic
housekeeping enzymes are essentially common to all cells. For

example, guinea pig uteri [291] and intestinal smooth muscle
[91] can tolerate the levels of Me2SO required for metastable
cooling to dry ice temperature without freezing, whereas rabbit
renal cortex cannot [292]. Similarly, frog hearts [87, 88] but not
rat hearts [89] can tolerate the 1011M level of ethylene glycol
required for the same purpose.
At least two additional biochemical studies on the nature of
cryoprotectant toxicity are currently underway, and a third is being
planned. In addition, efforts are being made to develop better
methods for measuring protein denaturation on a global basis to
enable direct examination of the role of protein denaturation in
both cryoprotectant toxicity and chilling injury. The potential con-
tributions of molecular biology to cryobiology could be revolu-
tionary and have been long awaited.
3.9 Chilling Injury Chilling injury is observed both in nature, at temperatures above
andIts Modification or 0C [293295], and in the laboratory, at temperatures well below
Avoidance zero [18, 52, 53, 207]. It has been linked to phase changes in
membranes [67, 68, 296299] and associated defects in mem-
brane permeability [298, 300, 301] and can be blocked in some
cases by directly modifying cell membrane composition [68, 302,
303], by using antifreeze proteins [300, 304], or by using genetic
engineering [305, 306] to block these phase transitions and inhibit
membrane leakage. Chilling injury may also result from protein
denaturation based on protection by prior heat shock [307], the
production of heat shock proteins in response to chilling [307],
and the production of cold shock proteins in response to chilling
as well [308, 309]. Also suggestive is the observation that the
unfolded protein response (ER stress response) is induced by 18h
of cold storage of human corneal endothelial cells and inhibited
using the specific blocking agent, salubrinal [310]. However, chill-
ing injury is generally observed to be an immediate response rather
than a delayed response to temperature reduction [18, 52, 53].
In none of the above mechanistic and interventional studies
was the system of interest saturated with a multi-molar concentra-
tion of cryoprotectant before the induction of chilling injury.
Further, chilling injury associated with vitrification is typically
observed primarily or exclusively at temperatures far below those
of the lipid phase transition temperatures and denaturation phe-
nomena noted above [18, 52, 53, 128, 207].
Nevertheless, one DNA microarray study of chilling injury in
the context of vitrification has recently been done, and it verified
changes suggestive of the ER stress/unfolded protein response
and altered lipid metabolism [207]. Precision-cut rat liver slices
loaded with either of two vitrification solutions showed no drop in
ATP content compared to controls when held at 0C, but a
2030% drop after cooling without freezing to 15C for 10min
(as detected after incubation at 37C following CPA washout).

Principal component analysis indicated clear separation between
the effects of CPA administration and the effects of chilling in the
presence of CPA.Comparing CPA treatment to CPA treatment
plus chilling, 1,108 transcripts changed in abundance with chilling,
but of these, only 31 increased more than 1.5-fold and only 6
decreased more than 1.5-fold, so the changes observed were in
general mild, in keeping with the mild change in ATP content, and
likely were indicative of the first changes induced by chilling. The
primary observations were an increase in heat shock protein and
heat shock factor transcripts, an increase in ribosomal RNA tran-
scripts (which would favor more protein synthesis to replace dena-
tured proteins), a lack of activation of apoptotic pathways
(suggesting ER stress did not reach levels sufficient to induce
apoptosis), activation of DNA damage sensing genes, activation of
two of the three MAP kinase stress pathways (involving increased
JNK and ERK signaling without increased p38 signaling), changes
that tend to reduce cholesterol synthesis and remove it from the
cell membrane and transfer it to the endoplasmic reticulum, and,
rather paradoxically, changes that tend to reduce synthesis of poly-
unsaturated fatty acids and increase synthesis of saturated fatty
acids as well as reduce fatty acid oxidation and metabolism. A num-
ber of other changes in expression were seen whose significance is
less easy to interpret. In summary, the effects of chilling and the
effects of VS exposure were remarkably similar, as though chilling
injury were an extension of VS toxicity.
Pig oocytes [311] and embryos [312] are particularly sensitive
to chilling injury in part as a consequence of containing globules of
cytoplasmic fat that cause damage to the cell membrane on cooling.
This problem has been reduced by various lipid removal or segrega-
tion techniques [311, 313], although birth rates have tended to be
low despite adding additional interventions [311, 314, 315].
Oocytes are in general quite susceptible to cooling injury, and
much of this susceptibility is related to disassembly of the meiotic
spindle and subsequent abnormal or incomplete reassembly [316,
317]. Cryoprotectants can stabilize [318] but can also damage
[319] the spindle. Nevertheless, with proper methodology, oocytes
can be preserved without significant spindle damage [319].
The vitrification of oocytes is motivated in no small part by the
utility of vitrification for outrunning chilling injury [54]. The
successful cryopreservation of Drosophila embryos was also enabled
by the ability of vitrification to allow rapid chilling injury to be
outrun in this species as well [52, 53, 128].
The demonstrated methods mentioned above for altering chill-
ing injury (consisting of modifying plasma membrane composition
or using antifreeze proteins to prevent membrane leakage) are
generally inconvenient or impractical for many applications and may
not be pertinent to chilling injury below 0C.McGrath [320], in
non-cryoprotected systems, and Fahy etal. [18, 321], in cryoprotected

systems, showed that chilling injury can be reduced or prevented by
an increase in medium tonicity. In the latter case, the optimum
tonicity for avoiding chilling injury during vitrification was found to
be between about 1.3 and 1.5 times isotonic, whereas for porcine
embryos a tonicity of ~2.8 times isotonic was effective [320].
Chilling injury is not universally observed in systems prepared
for vitrification [206]. It is seen in rabbit but not rat renal cortical
slices, is seen in rabbit and rat liver slices but not in monkey liver
slices, and may be absent in rat or rabbit hippocampal slices.
Comparing susceptible and non-susceptible tissues of the same
type might offer another way of understanding chilling injury and
seeking new mitigation strategies for it.
3.10 Storage Little is known about the safety of various durations of storage in
intheVitreous or the vitreous state at temperatures in the vicinity of TG, but this is an
Near-Vitreous State important topic for several reasons. First, the risk of fracture for-
mation increases as vitreous samples are cooled to the temperature
of liquid nitrogen (see Subheading2.7 above) and fractures may
damage organs, tissues, oocytes, embryos, and other systems as
well as creating sites of ice nucleation [147] that may indirectly
damage vitrified cells during warming. Second, liquid nitrogen
immersion has a number of practical, safety, and potential contami-
nation issues that could be avoided by storing in the vapor phase if
this were known to be safe. Therefore, one would like to know
how far below TG a sample must be cooled to protect it for long-
term storage and to verify that this temperature is still warm enough
to minimize the risk of fracturing.
Song etal. [322] reported that vitrified rabbit jugular veins
(TG~123C) stored at 130C for 4weeks or for 4months or
stored below 160C in liquid nitrogen vapor for either 4weeks or
for 4months all recovered as well as veins stored for only 24h and
approached the functionality of fresh controls. Heart valves and carti-
lage yielded similar results [322]. Our laboratory has stored rabbit
hippocampal slices under isothermal conditions in the vicinity of
145C (TG~124C) for months as well and has seen no deteriora-
tion [206]. In Song etal.s experience, there was no visual develop-
ment of ice during storage, and freeze substitution showed no ice
development after 5months of storage in liquid nitrogen vapor [322].
Fahy and Rall [15] and Mullen and Fahy [72] reported calcu-
lated safe storage times based on the assumption that the rate of
biological deterioration is governed by the mobility of molecules
near TG and therefore by the viscosity of the cryoprotectant solu-
tion. For 40%w/v DAP10+6% PEG 6000in an RPS-2 carrier
solution (40% DAP10=10%w/v 1,2-propanediol plus 30%w/v
of a mixture of Me2SO and acetamide in a 1:1mol ratio; TG
assumed to be about 122C), the amount of diffusion equivalent
to 1min at 20C was found to require 1week at 88C, 1month
at 93C, 1year at 100C, 10years at 106C, and 10,000years

at 116C [15]. For the intensively studied and used M22
vitrification solution (TG~123 C [18]), diffusion requiring 10s
at 0C was found to require 1week at ~93C, 1month at
~97C, 1year at ~103C, 10years at ~109C, and
10,000years at ~119C [72].
If not for the likelihood of nucleation above TG [72, 182],
these times suggest that storage near but actually above TG would
provide sufficiently long storage times for all practical purposes.
According to one scheme [99], intensive nucleation above TG
might not be a problem for the survival of a living system because
homogeneous nucleation will generate ice crystals that are too
small to be harmful. If the liquids in the sample are allowed to
nucleate completely, the lack of heterogeneity in crystal size would
preclude recrystallization, and warming would then simply melt
the ice nuclei, perhaps without significant grain growth. The work-
ability of this scheme has never been experimentally tested and
might lead to interesting results when tested in the future.
If storage is to be below TG to minimize nucleation, how far
below TG is cold enough for this purpose? In the M22 vitrification
solution, detectable nucleation can be extrapolated to be extin-
guished at about 136 to 137C or about 1314C below TG
[2]. Mehl [182], comparing the devitrification tendency of the
VS41A vitrification solution after 6months of storage at 1353C
(a mean of 12C below TG) to that of unstored samples, found
that the warming rate required to observe zero ice on warming
increased from 50C/min without storage to only 100150C/
min after storage, which supports the idea that extensive nucle-
ation during holding near TG may not lead to insurmountable
problems on warming. Mehl also pointed out that the number of
nuclei may not matter if the aim is to suppress their growth on
warming and they are all the same size at the beginning of warm-
ing, which will tend to be the case since ice crystals nucleated near
or below TG are not likely to grow until warming begins.
Although it may be academic given the above long projected
storage times near TG, relaxation times below TG take on the form of
Arrhenius kinetics, rising exponentially rather than super-exponentially
as temperature continues to fall [2]. Essentially, the extension of loga-
rithmic viscosity plots above TG to temperatures below TG is described
approximately by the tangent to the curve at TG.
Empirically, Rowe found no difference in stability of rapidly
frozen red cells between 165C and 196C over 16years
[323]. Valeri and Pivacek found no difference in the recovery of
frozen peripheral blood mononuclear cells stored at 135, 150,
and 196C for 22.4years [324]. And red cells frozen in
4045%w/v glycerol were stored successfully for 21years even at
80C [325]. Although indirect, these observations are consistent
with the possibility of storing vitrified systems not far below TG.
From time to time, investigators who correlate the glass t ransition

with the stability of dried systems have noted that deterioration of
those systems can take place even very far below the nominal glass
transition temperature(s) of those systems [326329]. The reason for
this is unknown, but it must be remembered that the physical state of
dried systems is far different from that of hydrated vitrified systems.
The former may be more susceptible to mechanical or chemical
sources of injury, the nature and meaning of the glassy state may be
different, and deterioration may be measured at much higher abso-
lute temperatures than for the case of hydrated vitrified systems.
According to Sun, however, deterioration of dried liposomes can
sometimes be stopped by using more sugar prior to drying, perhaps
because the extra sugar provides space between the liposomes so that
they dont fuse [326]. This may be more analogous to a hydrated
system, in which water, the smallest major biological molecule, fills in
more gaps and bonds the system together more strongly.
3.11 Proteins at Low We conclude by noting a little-referenced area of molecular biol-

Temperatures ogy that has a direct bearing on the principles of biological vitrifi-
cation and that may one day provide additional insights that will
help to guide the cryobiologist to more successful vitrification
methods. This field of research, called cryoenzymology, has pro-
vided a significant body of literature documenting the ability of
mixed cosolvent systems comprising water and molecules that
depress its freezing point to maintain protein secondary, tertiary,
and quaternary structure sufficiently well in some cases to enable
normal enzymatic catalysis to proceed, albeit at greatly reduced
reaction rates, at temperatures as low as 70C [330, 331].
Because vitrification preserves cells and molecules in the absence of
ice, cryoenzymology offers many direct windows on phenomena
that may affect the success of vitrification, including changing
dielectric constants and pKa values with decreasing temperature
and increasing cryoprotectant concentrations [331], which may
bear on such phenomena as cold denaturation and chilling injury
during vitrification. On the other hand, one of the observations
made is that although protein cold denaturation is a real phenom-
enon [332334], it can be prevented in some cases by rapid cool-
ing to temperatures too low to favor it kinetically [331, 332], and
cryoprotectants such as sucrose can prevent protein denaturation
during freezing [332], while glycerol, for example, has been shown
to prevent enzymes such as glucose-6-phosphate dehydrogenase,
carbamyl phosphate synthetase, and pyruvate decarboxylase from
dissociating into subunits due to cooling [331].
So far, such observations have not been applied to events tak-
ing place in living cells. Perhaps, in the future, an alliance between
cryoenzymologists and cryobiologists could lead to interesting and
potentially utilitarian results. By the same token, there may be les-
sons to be learned from those who freeze proteins either in purified
form or in situ [66, 335].
4 Summary andConclusions
Successful biological vitrification is the result of reconciling the

physics of aqueous solutions with the biology of surviving expo-
sure to very water-poor mixed solvent systems. The toxicity of a
cryoprotectant solution in the context of vitrification has meaning
only in the context of the physical properties of the solution, and
the physical properties of the solution, in turn, have meaning only
in the context of their compatibility with life and cellular health.
From the mutual embrace of these two very different scientific
realms has come an extremely broad, beneficial, and growing tech-
nology for preserving life at low temperatures. For much of the
modern history of this now-united field, the major scientific
emphasis has been on intensive study of the physics of vitrification
and devitrification, with the biological side of the equation being
confined more or less to the modeling of pCPA permeation rates
in relatively simple living systems and to hypothesis-driven process
optimization without the benefit of a mechanistic understanding
of cryoprotectant toxicity, but that mechanistic understanding is
now beginning to emerge.
The use of vitrification to preserve living systems and the con-
stituents thereof is currently growing at an exponential rate, and
the occurrence of vitrification in the natural world is becoming
increasingly appreciated. Successful applications have been reported
for a wide variety of mammalian cells and tissues and even for an
intact mammalian kidney but extend as well to the plant kingdom
and to the preservation of insects and other invertebrates of scien-
tific, medical, and commercial interest.
The success of applied vitrification depends on the choice and
final concentrations of cryoprotectants employed, the avoidance of
osmotic damage during their introduction and removal, the selec-
tion of appropriate temperatures and durations of each phase of
treatment with cryoprotectants, the avoidance of chilling injury, the
avoidance of fracturing (especially but not exclusively in larger sys-
tems), and, increasingly, the use of new types of chemical agent that
can interact specifically and directly with ice to inhibit nucleation
and growth of ice crystals. Fortunately, much is currently known
about all of these factors. Mathematical modeling is also allowing
increasingly sophisticated and creative protocols for speeding the
introduction and washout of pCPAs while minimizing toxicity
throughout, but equipment for continuous rather than stepwise
introduction and removal of cryoprotectants, which is needed for
the most efficient protocols, is not yet widely available.
The most important barrier to vitrification since the inception
of the concept of using high concentrations of cryoprotectants to
achieve it remains the same: the intrinsic toxicity of the vitrification
solution. Spectacular improvements in the control of vitrification
solution toxicity have already been achieved, but more progress is

still needed. Fortunately, for the first time, the powerful and com-
prehensive omics tools, as well as the more focused and selec-
tive tools, of modern molecular biology and enzymology are finally
beginning to be applied to the problem of gaining a comprehen-
sive understanding of the still-mysterious and pivotal problems of
cryoprotectant toxicity and chilling injury.
There is no guarantee that deeper understanding will lead to
fundamental new applied breakthroughs, but the insights gained
to date are already providing useful clues to potentially more effec-
tive interventions. All in all, the future of cryopreservation by vit-
rification looks bright and full of exciting new possibilities.
Acknowledgments
We would like to thank Igor Katkov for providing helpful references

to the lack of IIF in cryoprotected and non-cryoprotected sperm.
This research was supported by 21st Century Medicine, Inc.
References
1. Goldstein M, Simha R (1976) The glass tran- suspensions containing various concentra-
sition and the nature of the glassy state, vol tions of glycerol. Biodynamica 10:193210
279, Ann N Y Acad Sci. NewYork Academy 10. Angell CA, Sichina W (1976) Thermodynamics
of Sciences, NewYork, NY of the glass transition: empirical aspects. Ann
2. Wowk B (2010) Thermodynamic aspects of N Y Acad Sci 279:5367
vitrification. Cryobiology 60:1122 11. Sakurai M, Furuki T, Akao K, Tanaka D,
3. Shackley MS (2013) Archeological obsidian Nakahara Y, Kikawada T, Watanabe M, Okuda
studies: method and theory, vol 3, Adv. T (2008) Vitrification is essential for anhydro-
Archeol. Museum Sci. Plenum press, biosis in an African chironomid, Polypedilum
NewYork, NY, pp1243 vanderplanki. Proc Natl Acad Sci U S A
4. Ojovan MJ, Lee WE (2013) An introduction 105:50935098
to nuclear waste immobilization, 2nd edn, 12. Hengherr S, Worland MR, Reuner A,
Elsevier insights. Elsevier, Amsterdam Brummer F, Schill RO (2009) High-
5. Hanford Vitrification Plant. Available from: temperature tolerance in anhydrobiotic tardi-
http://www.bechtel.com/hanford_waste_ grades is limited by glass transition. Physiol
treatment.html Biochem Zool 82:749755
6. Brayley EW (1860) Notes on the apparent uni- 13. Povey JF, Perez-Moral N, Noel TR, Parker R,
versality of a principle analogous to regelation, Howard MJ, Smales CM (2009) Investigating
on the physical nature of glass, and on the prob- variables and mechanisms that influence pro-
able existence of water in a state corresponding tein integrity in low water content amorphous
to that of glass. Proc Roy Soc 10:450460 carbohydrate matrices. Biotechnol Prog
7. Bruggeller P, Mayer E (1980) Complete vitri- 25:12171227
fication in pure liquid water and dilute aque- 14. Fahy GM, Wowk B, Wu J (2006)
ous solutions. Nature 288:569571 Cryopreservation of complex systems: the
8. Dubochet J, McDowall AW (1981) missing link in the regenerative medicine sup-
Vitrification of pure water for electron micros- ply chain. Rejuvenation Res 9:279291
copy. J Microsc 124:RP3RP4 15. Fahy GM, Rall WF (2007) Vitrification: an
9. Rapatz G, Luyet B (1968) Electron micro- overview. In: Liebermann J, Tucker MJ (eds)
scope study of erythrocytes in rapidly cooled Vitrification in assisted reproduction: a users
manual and troubleshooting guide. Informa 29. Rasmussen D (1969) A note about phase
Healthcare, London, pp120 diagrams of frozen tissues. Biodynamica
16. Taylor MJ, Song YC, Brockbank KG (2004) 10:333339
Vitrification in tissue preservation: new devel- 30. Knight CA (1967) The freezing of super-
opments. In: Fuller BJ, Lane N, Benson EE cooled liquids. D.Van Nostrand, Princeton,
(eds) Life in the frozen state. CRC Press, NJ
Boca Raton, FL, pp603641 31. Luyet B (1937) The vitrification of organic
17. Fahy GM, Wowk B, Pagotan R, Chang A, colloids and of protoplasm. Biodynamica
Phan J, Thomson B, Phan L (2009) Physical 1:114
and biological aspects of renal vitrification. 32. MacFarlane DR, Forsyth M, Barton CA
Organogenesis 5:167175 (1991) Vitrification and devitrification in
18. Fahy GM, Wowk B, Wu J, Phan J, Rasch C, cryopreservation. Adv Low Temp Biol
Chang A, Zendejas E (2004) Cryopreservation 1:221277
of organs by vitrification: perspectives and 33. Mehl PM (1996) Crystallization and vitrifica-
recent advances. Cryobiology 48:157178 tion in aqueous glass-forming solutions. Adv
19. Luyet BJ, Gehenio PM (1940) Life and death Low Temp Biol 3:185255
at low temperatures. Biodynamica, Normandy, 34. Mazur P (1966) Physical and chemical basis
MO of injury in single-celled micro-organisms
20. Luyet BJ, Gehenio PM (1939) The physical subjected to freezing and thawing. In:
states of protoplasm at low temperatures. Meryman HT (ed) Cryobiology. Academic,
Biodynamica 2:1128 NewYork, NY, pp214315
21. Franks F, Asquith MH, Hammond CC, Skaer 35. Luyet BJ, Sager D, Gehenio PM (1967) The
HB, Echlin P (1977) Polymer cryoprotec- phenomenon of premelting recrystalliza-
tants in the preservation of biological ultra- tion. Biodynamica 10:123132
structure. I.Low temperature states of 36. Boutron P, Kaufmann A (1978) Stability of
aqueous solutions of hydrophilic polymers. J the amorphous state in the system water-
Microsc 110:223228 glycerol-dimethylsulfoxide. Cryobiology
22. Levine H, Slade L (1988) Thermomechanical 15:93108
properties of small-carbohydrate-water glasses 37. Boutron P (1986) Comparison with the the-
and rubbers. J Chem Soc Faraday Trans ory of the kinetics and extent of ice crystalli-
84:26192633 zation and of the glass-forming tendency in
23. Morris GJ, Goodrich M, Acton E, Fonseca F aqueous cryoprotective solutions.
(2006) The high viscosity encountered dur- Cryobiology 23:88102
ing freezing in glycerol solutions: effects on 38. Boutron P, Mehl P (1990) Theoretical pre-
cryopreservation. Cryobiology 52:323334 diction of devitrification tendency: determi-
24. Rasmussen DH, Luyet B (1972) Thermal nation of critical warming rates without using
analysis of partially dehydrated heart tissue. finite expansions. Cryobiology 27:359377
Biodynamica 11:149155 39. Meryman HT (1971) Cryoprotective agents.
25. Luyet B, Rasmussen DH (1973) On some Cryobiology 8:173183
inconspicuous changes occurring in aqueous 40. Karow AM Jr (1969) Cryoprotectants a
systems subjected to below zero C tempera- new class of drugs. J Pharm Pharmacol
tures. Biodynamica 11:209215 21:209223
26. Luyet B, Rasmussen D (1967) Study by dif- 41. Fuller BJ (2004) Cryoprotectants: the essen-
ferential thermal analysis of the temperatures of tial antifreezes to protect life in the frozen
instability in rapidly cooled solutions of polyvi- state. Cryo Lett 25:375388
nylpyrrolidone. Biodynamica 10:137147 42. Katkov II, Isachenko V, Isachenko E (2007)
27. Luyet B, Rasmussen D (1968) Study by dif- Vitrification in small quenched volumes with
ferential thermal analysis of the temperatures a minimal amount of, or without vitrificants:
of instability of rapidly cooled solutions of basic biophysics and thermodynamics. In:
glycerol, ethylene glycol, sucrose, and glu- Tucker MJ, Liebermann J (eds) Vitrification
cose. Biodynamica 10:167191 in assisted reproduction. Informa UK,
28. Rasmussen D, Luyet B (1969) Complementary London, pp2132
study of some non-equilibrium phase transi- 43. Wowk B, Leitl E, Rasch CM, Mesbah-Karimi
tions in frozen solutions of glycerol, ethylene N, Harris SB, Fahy GM (2000) Vitrification
glycol, glucose, and sucrose. Biodynamica enhancement by synthetic ice blocking agents.
10:319331 Cryobiology 40:228236
44. Wowk B, Fahy GM (2002) Inhibition of bac- 59. Crowe JH, Crowe LM, Tablin F, Wolkers W,
terial ice nucleation by polyglycerol polymers. Oliver AE, Tsvetkova NM (2004) Stabilization
Cryobiology 44:1423 of cells during freeze-drying: the trehalose
45. Kuwabara C, Wang D, Endoh K, Fukushi Y, myth. In: Fuller BJ, Lane N, Benson EE (eds)
Arakawa K, Fujikawa S (2013) Analysis of Life in the frozen state. CRC Press, Boca
supercooling activity of tannin-related poly- Raton, FL, pp581601
phenols. Cryobiology 67:4049 60. Acker JP, Chen T, Fowler A, Toner M (2004)
46. Kuwabara C, Wang D, Kasuga J, Fukushi Y, Engineering desiccation tolerance in mamma-
Arakawa K, Koyama T, Inada T, Fujikawa S lian cells: tools and techniques. In: Fuller BJ,
(2012) Freezing activities of flavonoids in Lane N, Benson EE (eds) Life in the frozen
solutions containing different ice nucleators. state. CRC Press, Boca Raton, FL,
Cryobiology 64:279285 pp563580
47. Walters KR, Serianni AS, Sformo T, Barnes 61. Levitt J (1965) Thiogel a model system for
BM, Duman JG (2009) A nonprotein ther- demonstrating intermolecular disulfide bond
mal hysteresis-producing xylomannan anti- formation on freezing. Cryobiology
freeze in the freeze-tolerant Alaskan beetle 1:312316
Upis ceramboides. Proc Natl Acad Sci U S A 62. Morton WM (1969) Effects of freezing and
106:2021020215 hardening on the sulfhydryl groups of protein
48. Graether SP (2011) Biochemistry and func- fractions from cabbage leaves. Plant Physiol
tion of antifreeze proteins. In: Uversky VN 44:168172
(ed) Molecular anatomy and physiology of 63. Ishiwata S (1976) Freezing of actin. Reversible
proteins. Nova, NewYork, NY oxidation of a sulfhydryl group and structural
49. Rall WF, Fahy GM (1985) Ice-free cryo- change. J Biochem 80:595609
preservation of mouse embryos at 196C by 64. Takahashi T, Asahina E (1977) Protein-
vitrification. Nature 313:573575 bound SH groups in frozen-thawed egg cells
50. McGrath JJ (1987) Cold shock: thermoelastic of the sea urchin. Cryobiology 14:367372
stress in chilled biological membranes. In: 65. Chatterjee S, de Lamirande E, Gagnon C
Diller KR (ed) Network thermodynamics, heat (2001) Cryopreservation alters membrane
and mass transfer in biotechnology. United sulfhydryl status of bull spermatozoa: protec-
Engineering Center, NewYork, NY, pp5766 tion by oxidized glutathione. Mol Reprod
51. McGrath JJ, Morris GJ (1985) Cold shock Dev 60:498506
injury is a significant factor in freezing injury: 66. Tappel AL (1966) Effects of low tempera-
a position for. Cryobiology 22:628 tures and freezing on enzymes and enzyme
52. Steponkus PL, Myers SP, Lynch DV, Gardner systems. In: Meryman HT (ed) Cryobiology.
L, Bronshteyn V, Leibo SP, Rall WF, Pitt RE, Academic, NewYork, NY, pp163177
Lin TT, MacIntyre RJ (1990) 67. Crowe JH, Hoekstra FA, Crowe LM,
Cryopreservation of Drosophila melanogaster Anchordoguy TJ, Drobnis E (1989) Lipid
embryos. Nature 345:170172 phase transitions measured in intact cells with
53. Mazur P, Schneider U, Mahowald AP (1992) Fourier transform infrared spectroscopy.
Characteristics and kinetics of subzero chill- Cryobiology 26:7884
ing injury in Drosophila embryos. 68. Zeron Y, Tomczak M, Crowe J, Arav A
Cryobiology 29:3968 (2002) The effect of liposomes on thermo-
54. Martino A, Songsasen N, Leibo SP (1996) tropic membrane phase transitions of bovine
Development into blastocysts of bovine spermatozoa and oocytes: implications for
oocytes cryopreserved by ultra-rapid cooling. reducing chilling sensitivity. Cryobiology
Biol Reprod 54:10591069 45:143152
55. Crowe JH, Clegg JS (eds) (1973) 69. Webb MS, Hui SW, Steponkus PL (1993)
Anhydrobiosis. Dowden, Hutchinson & Dehydration-induced lamellar-to-hexagonal-
Ross, Stroudsburg, PA, p477 II phase transitions in DOPE/DOPC mix-
56. Crowe JH, Clegg JS (1979) Dry biological tures. Biochim Biophys Acta 1145:93104
systems. Academic, NewYork, NY 70. Fahy GM (1988) Vitrification. In: McGrath
57. Leather SR, Walters KFA, Bale JS (1993) The JJ, Diller KR (eds) Low temperature biotech-
ecology of insect overwintering. Cambridge nology: emerging applications and engineer-
University Press, Cambridge ing contributions. ASME, NewYork, NY,
pp113146
58. Lee RE Jr, Denlinger DL (1991) Insects at
low temperature. Chapman and Hall, 71. Fahy GM (1990) Vitrification as an approach
NewYork, NY to organ cryopreservation: past, present, and
future. In: Smit Sibinga CT, Das PC, 88. Rapatz G (1972) Recovery of activity of frog
Meryman HT (eds) Cryopreservation and hearts after exposure to 78C.Cryobiology
low temperature biology in blood transfusion. 9:322
Kluwer, Boston, MA, pp255268 89. Rapatz G, Keener R (1974) Effect of concen-
72. Mullen SF, Fahy GM (2011) Fundamental tration of ethylene glycol on the recovery of
aspects of vitrification as a method of repro- frog hearts after freezing to low temperatures.
ductive cell, tissue, and organ cryopreserva- Cryobiology 11:571572
tion. In: Donnez J, Kim S (eds) Principles & 90. Elford BC (1970) Functional recovery of
practice of fertility preservation. Cambridge smooth muscle after exposure to dimethyl
University Press, Cambridge, pp145163 sulfoxide and low temperatures. Cryobiology
73. Stiles W (1930) On the cause of cold death of 7:148153
plants. Protoplasma 9:459468 91. Elford BC, Walter CA (1972) Effects of elec-
74. Tammann G (1898) Ueber die Abhangigkeit trolyte composition and pH on the structure
der Zahl der Kerne, welche sich in verschie- and function of smooth muscle cooled to
denen unterkuhlten Flussigkeiten bilden, vom 79C in unfrozen media. Cryobiology
der temperatur. Z Phys Chem 25:441479 9:82100
75. Moran T (1926) The freezing of gelatin gel. 92. Boutron P, Alben R (1975) Structural model
Proc Roy Soc A 112:3046 for amorphous solid water. J Chem Phys
76. Hardy WB (1926) A microscopic study of the 62:48484853
freezing of gel. Proc Roy Soc A 112:4761 93. Boutron P, Kaufmann A (1978) Metastable
77. Luyet B, Rapatz G (1958) Patterns of ice for- states in the system water-ethanol. Existence
mation in some aqueous solutions. of a second hydrate, curious properties of
Biodynamica 8:168 both hydrates. J Chem Phys 68:50325041
78. Meryman HT (1958) X-ray analysis of rapidly 94. Boutron P, Kaufmann A (1979) Stability of
frozen gelatin gels. Biodynamica 8:6972 the amorphous state in the system water-
79. Rasmussen D, Luyet B (1970) Contribution glycerol-ethylene glycol. Cryobiology
to the establishment of the temperature- 16:8389
concentration curves of homogeneous nucle- 95. Boutron P, Kaufmann A (1979) Maximum in
ation in solutions of some cryoprotective the stability of the amorphous state in the sys-
agents. Biodynamica 11:3344 tem water-glycerol-ethanol. Cryobiology
80. Luyet B, Kroener C (1966) The temperature 16:372389
of the glass transition in aqueous solutions 96. Boutron P (1979) Stability of the amorphous
of glycerol and ethylene glycol. Biodynamica state in the system water-1,2-propanediol.
10:3340 Cryobiology 16:557568
81. Luyet B (1969) On the amount of water 97. Fahy GM (1981) Prospects for vitrification of
remaining amorphous in frozen aqueous solu- whole organs. Cryobiology 18:617
tions. Biodynamica 10:277291 98. Fahy GM (1981) Analysis of solution
82. Farrant J (1965) Mechanism of cell damage effects injury: cooling rate dependence of
during freezing and thawing and its preven- the functional and morphological sequelae of
tion. Nature 205:12841287 freezing in rabbit renal cortex protected with
83. Fahy GM, MacFarlane DR, Angell CA, dimethyl sulfoxide. Cryobiology 18:550570
Meryman HT (1984) Vitrification as an 99. MacFarlane DR, Angell CA, Fahy GM (1981)
approach to cryopreservation. Cryobiology Homogeneous nucleation and glass forma-
21:407426 tion in cryoprotective systems at high pres-
84. Gonzales F, Luyet B (1950) Resumption of sures. Cryo Lett 2:353358
heart beat in chick embryo frozen in liquid 100. Fahy GM, MacFarlane DR, Angell CA (1982)
nitrogen. Biodynamica 7:15 Recent progress toward vitrification of kid-
85. Luyet BJ, Gonzales F (1953) Growth of neys. Cryobiology 19:668669
nerve tissue after freezing in liquid nitrogen. 101. Fahy GM, Hirsh A (1982) Prospects for

Biodynamica 7:171174 organ preservation by vitrification. In: Pegg
86. Luyet B, Hodapp A (1938) Revival of frogs DE, Jacobsen IA, Halasz NA (eds) Organ
spermatozoa vitrified in liquid air. Proc Meet preservation, basic and applied aspects. MTP
Soc Exp Biol 39:433434 Press, Lancaster, pp399404
87. Rapatz G (1970) Resumption of activity in 102. Fahy GM, MacFarlane DR, Angell CA,

frog hearts after exposure to very low tem- Meryman HT (1983) Vitrification as an
peratures. Cryobiology 6:588 approach to cryopreservation. Cryobiology
20:699
103. Fahy GM (1982) Prevention of toxicity from 119. Mazur P (1984) Freezing of living cells:

high concentrations of cryoprotective agents. mechanisms and implications. Am J Physiol
In: Pegg DE, Jacobsen IA, Halasz NA (eds) 247:C125C142 (Cell Physiology 16)
Organ preservation, basic and applied aspects. 120. Mazur P (1988) Stopping biological time.

MTP Press, Lancaster, pp367369 The freezing of living cells. Ann N Y Acad Sci
104. Fahy GM (1983) Cryoprotectant toxicity
541:514531
neutralizers reduce freezing damage. Cryo 121. Mazur P (2004) Principles of cryobiology. In:
Lett 4:309314 Fuller BJ, Lane N, Benson EE (eds) Life in
105. Fahy GM (1984) Cryoprotectant toxicity:
the frozen state. CRC Press, Boca Raton, FL,
biochemical or osmotic? Cryo Lett 5:7990 pp365
122. Mazur P, Leibo SP, Farrant J, Chu EHY,

reduction: specific or nonspecific? Cryo Lett Hanna MG, Smith LH (1970) Interactions of
5:287294 cooling rate, warming rate and protective
107. Sakai A, Kobayashi S, Oiyama I (1990)
additive on the survival of frozen mammalian
Cryopreservation of nucellar cells of navel cells. In: Wolstenholme GEW, OConnor M
orange (Citrus sinensis Osb. va. brasiliensis (eds) The frozen cell. J&A Churchill, London,
Tanaka) by vitrification. Plant Cell Rep pp6985
9:3033 123.
Fahy GM, Karow AM Jr (1977)
108. Sakai A (2004) Plant cryopreservation. In:
Ultrastructure-function correlative studies for
Fuller BJ, Lane N, Benson EE (eds) Life in cardiac cryopreservation. V.Absence of a cor-
the frozen state. CRC Press, Boca Raton, FL, relation between electrolyte toxicity and cryo-
pp329345 injury in the slowly frozen, cryoprotected rat
109. Lovelock JE (1953) The mechanism of the heart. Cryobiology 14:418427
protective action of glycerol against haemoly- 124. Fahy GM (1986) The relevance of cryopro-
sis by freezing and thawing. Biochim Biophys tectant toxicity to cryobiology. Cryobiology
Acta 11:2836 23:113
110. Kylin H (1917) Uber die Kalteresistenze der 125. Lovelock JE (1954) The protective action of
Meeresalgen. Ber Desch Bot Ges neutral solutes against haemolysis by freezing
35:370384 and thawing. Biochem J 56:265270
111. Muller-Thurgau H (1886) Uber das Gefrieren 126. Meryman HT, Williams RJ, Douglas MSJ

and Erfrieren der Pflanzen. II Theil. Landw (1977) Freezing injury from solution
Jahrb (Berlin) 15:453610 effects and its prevention by natural or artifi-
112. Chambers R, Hale HP (1932) The formation cial cryoprotection. Cryobiology
of ice in protoplasm. Proc Roy Soc Ser B 14:287302
110:336352 127. Fahy GM (1986) Vitrification: a new approach
113. Meryman HT (1970) The exceeding of a
to organ cryopreservation. Prog Clin Biol Res
minimum tolerable cell volume in hypertonic 224:305335
suspension as a cause of freezing injury. In: 128. Mazur P, Cole KW, Hall JW, Schreuders PD,
Wolstenholme GEW, OConnor M (eds) The Mahowald AP (1992) Cryobiological preser-
frozen cell. J & A Churchill, London, vation of Drosophila embryos. Science
pp5164 258:18961897
114. Meryman HT (1968) Modified model for the 129. Berthelot F, Martinat-Botte F, Perreau C

mechanism of freezing injury in erythrocytes. (2001) Birth of piglets after OPS vitrification
Nature 218:333336 and transfer of compacted morula stage
115. Meryman HT (1971) Osmotic stress as a
embryos with intact zona pellucida. Reprod
mechanism of freezing injury. Cryobiology Nutr Dev 41:267272
8:489500 130. Pollack GA, Pegg DE, Hardie IR (1986) An
116. Meryman HT (1974) Freezing injury and its isolated perfused rat mesentery model for
prevention in living cells. Annu Rev Biophys direct observation of the vasculature during
Bioeng 3:341363 cryopreservation. Cryobiology 23:500511
117. Mazur P (1963) Kinetics of water loss from 131. Taylor MJ, Pegg DE (1983) The effect of ice
cells at subzero temperatures and the likeli- formation on the function of smooth muscle
hood of intracellular freezing. J Gen Physiol tissue stored at 21 or 60C.Cryobiology
47:347369 20:3640
118. Mazur P (1970) Cryobiology: the freezing of 132. Pegg DE, Diaper MP (1982) The mechanism
biological systems. Science 168:939949 of cryoinjury in glycerol-treated rabbit kid-
neys. In: Pegg DE, Jacobsen IA, Halasz NA
(eds) Organ preservation, basic and applied tions and disappearance of the cracks during
aspects. MTP Press, Ltd., Lancaster, rewarming. Biodynamica 10:4752
pp389393 145. Fahy GM, Saur J, Williams RJ (1990) Physical
133. Hunt CJ (1984) Studies on cellular structure problems with the vitrification of large bio-
and ice location in frozen organs and tissues: logical systems. Cryobiology 27:492510
the use of freeze-substitution and related 146. Rall WF, Meyer TK (1989) Zona fracture

techniques. Cryobiology 21:385402 damage and its avoidance during the cryo-
134. Pegg DE (1987) Ice crystals in tissues and preservation of mammalian embryos.
organs. In: Pegg DE, Karow AM Jr (eds) The Theriogenology 31:683692
biophysics of organ cryopreservation. Plenum, 147. Williams RJ, Carnahan DL (1990) Fracture
NewYork, NY, pp117140 faces and other interfaces as ice nucleation
135. Pegg DE (2010) The relevance of ice crystal sites. Cryobiology 27:479482
formation for the cryopreservation of tissues 148. Nickell PK, Sass S, Verleve D, Blumenthal

and organs. Cryobiology 60:S36S44 EM, Duman JG (2013) Antifreeze proteins in
136. Taylor MJ (1984) Sub-zero preservation and the primary urine of larvae of the beetle
the prospect of long-term storage of multicel- Dendroides canadensis. J Exp Biol
lular tissues and organs. In: Calne RY (ed) 216:16951703
Transplantation immunology: clinical and 149. Sformo T, Walters K, Jeannet K, Wowk B,
experimental. Oxford University Press, Fahy G, Barnes BM, Duman JG (2010) Deep
Oxford, pp360390 supercooling, vitrification and limited survival
137. Hunt CJ, Taylor MJ, Pegg DE (1982) Freeze- to 100C in the Alaskan beetle Cucujus cla-
substitution and isothermal freeze fixation vipes puniceus (Coleoptera: Cucujidae). J Exp
studies to elucidate the pattern of ice forma- Biol 213:502509
tion on smooth muscle at 252K (21C). J 150. Cosman B (2013) This is the new coldest

Microsc 125:177186 temperature ever recorded on earth
138. Jacobsen IA, Pegg DE, Starklint H, Chemnitz 151. Elster J, Benson EE (2004) Life in the polar
J, Hunt C, Barfort P, Diaper MP (1984) terrestrial environment with a focus on algae
Effect of cooling and warming rate on glyc- and cyanobacteria. In: Fuller BJ, Lane N,
erolized rabbit kidneys. Cryobiology Benson EE (eds) Life in the frozen state.
21:637653 CRC Press, Boca Raton, FL, pp111150
139. Pegg DE, Jacobsen IA, Armitage WJ, Taylor 152. Bennett VA, Sformo T, Walters K, Toien O,
MJ (1979) Mechanisms of cryoinjury in Jeannet K, Hochstrasser R, Pan Q, Serianni
organs. In: Pegg DE, Jacobsen IA (eds) AS, Barnes BM, Duman JG (2005)
Organ preservation II.Churchill Livingstone, Comparative overwintering physiology of
NewYork, NY, pp132144 Alaska and Indiana populations of the beetle
140.
Song YC, Khirabadi BS, Lightfoot F, Cucujus clavipes (Fabricius): roles of anti-
Brockbank KG, Taylor MJ (2000) Vitreous freeze proteins, polyols, dehydration and dia-
cryopreservation maintains the function of pause. J Exp Biol 208(Pt 23):44674477
vascular grafts. Nat Biotechnol 18:296299 153. Sun WQ, Leopold AC (1997) Cytoplasmic
141. Brockbank KGM, Taylor MJ (2007) Tissue vitrification and survival of anhydrobiotic
preservation. In: Baust JG, Baust JM (eds) organisms. Comp Biochem Physiol
Advances in biopreservation. Taylor & Francis 117A:327333
Group, Boca Raton, FL, pp157196 154. Burke MJ (1986) The glassy state and survival
142. Vajta G, Kuwayama M, Vanderzwalmen P
of anhydrous biological systems. In: Leopold
(2007) Disadvantages and benefits of vitrifica- AC (ed) Membranes, metabolism, and dry
tion. In: Tucker MJ, Liebermann J (eds) organisms. Cornell University Press, Ithaca,
Vitrification in assisted reproduction. A users NY, pp358364
manual and trouble-shooting guide. Informa 155. Holmstrup M, Bayley M, Ramlov H (2002)
UK, London, pp3344 Supercool or dehydrate? An experimental
143. Fahy GM, Takahashi T, Meryman HT (1986) analysis of overwintering strategies in small
Practical aspects of ice-free cryopreservation. permeable arctic invertebrates. Proc Natl
In: Smit Sibinga CT, Das PC, Greenwalt TJ Acad Sci 99:57165720
(eds) Future developments in blood banking. 156. Crowe JH, Jackson S, Crowe LM (1983)

Martinus Nijhoff Publishing, Boston, MA, Nonfreezable water in anhydrobiotic nema-
pp111122 todes. Mol Physiol 3:99105
144. Kroener C, Luyet B (1966) Formation of
157. Hirsh AG, Williams RJ, Meryman HT (1985)
cracks during the vitrification of glycerol solu- A novel method of natural cryoprotection:
intracellular glass formation in deeply frozen 173.

Pruppacher HR, Klett JD (1996)
populus. Plant Physiol 79:4156 Homogeneous nucleation. In: Pruppacher
158. Rall WF, Reid DS, Farrant J (1980) Innocuous HR, Klett JD (eds) Microphysics of clouds
biological freezing during warming. Nature and precipitation. Springer, Dordrecht,
286:511514 pp191215, Second revised and expanded
159. Rall WF (1981) The role of intracellular ice in edition with an introduction to cloud chemis-
the slow warming injury of mouse embryos. try and cloud electricity
In: Zeilmaker GH (ed) Frozen storage of lab- 174.
Pruppacher HR, Klett JD (1996)
oratory animals. Gustav Fischer Verlag, Heterogeneous nucleation. In: Pruppacher
NewYork, NY, pp3344 HR, Klett JD (eds) Microphysics of clouds
160.
Lehn-Jensen H, Rall WF (1983) and precipitation. Springer, Dordrecht,
Cryomicroscopic observations of cattle pp287360, Second revised and expanded
embryos during freezing and thawing. edition with an introduction to cloud chemis-
Theriogenology 19:263277 try and cloud electricity
161. Rall WF, Reid DS, Polge C (1984) Analysis of 175. Maki LR, Galyan EL, Chang-Chien MM,

slow-warming injury of mouse embryos by Caldwell DR (1974) Ice nucleation induced
cryomicroscopical and physicochemical meth- by pseudomonas syringae. Appl Environ
ods. Cryobiology 21:106121 Microbiol 28:456459
162. Ediger MD, Angell CA, Nagel SR (1996)
176.
Gavish M, Popovitz-Biro R, Lahav M,
Supercooled liquids and glasses. J Phys Chem Leiserowitz L (1990) Ice nucleation by alco-
100:1320013212 hols arranged in monolayers at the surface of
water drops. Science 250:973975
163. Kauzmann W (1948) The nature of the glassy
state and the behavior of liquids at low tem- 177. Earle ME, Kuhn T, Khalizov AF, Sloan JJ

peratures. Chem Rev 43:219256 (2010) Volume nucleation rates for homoge-
neous freezing in supercooled water micro-
164.
Forsyth M, MacFarlane DR (1986) droplets: results from a combined experimental
Recrystallization revisited. Cryo Lett and modelling approach. Atm Chem Phys
7:367378 10:79457961
165. Mazur P, Seki S (2011) Survival of mouse 178. Dupuy J, Jal JF, Ferradou C, Chieux P, Wright
oocytes after being cooled in a vitrification AF, Calemczuk R, Angell CA (1982)
solution to 196C at 95 to 70,000C/min Controlled nucleation and quasi-ordered
and warmed at 610 to 118,000C/min: a growth of ice crystals from low temperature
new paradigm for cryopreservation by vitrifi- electrolyte solutions. Nature 296:135140
cation. Cryobiology 62:17
179. Angell CA, Sare EJ, Donnella J, MacFarlane
166. Boutron P, Arnaud F (1984) Comparison of DR (1981) Homogeneous nucleation and
the cryoprotection of red blood cells by glass transition temperatures in solutions of
1,2-propanediol and glycerol. Cryobiology lithium salts in water-D2 and water. Doubly
21:348358 unstable glass regions. J Phys Chem
167. Rall WF (1987) Factors affecting the survival 85:14611464
of mouse embryos cryopreserved by vitrifica- 180. Fahy GM (1998) Fundamentals of vitrifica-
tion. Cryobiology 24:387402 tion: tissues and organs. Cryobiology
168. Dorsey NE (1948) The freezing of super-
37:380381
cooled water. Trans Am Phil Soc 181. Wowk B (2013) Metastable vitrification of

38:247328 cryoprotectant solutions. Cryobiology
169. Jackson CL, McKenna GB (1990) The melt- 67:409
ing behavior of organic materials confined in 182. Mehl P (1993) Nucleation and crystal growth
porous solids. J Chem Phys 93:90029011 in a vitrification solution tested for organ
170. Angell CA (1982) Supercooled water. In:
cryopreservation by vitrification. Cryobiology
Franks F (ed) Water, a comprehensive treatise. 30:509518
Plenum, NewYork, NY, pp181 183. Wowk B, Fahy GM (2007) Ice nucleation and
171. Rasmussen DH, MacCaulay MN, MacKenzie growth in concentrated vitrification solutions.
AP (1975) Supercooling and nucleation of ice Cryobiology 55:330
in single cells. Cryobiology 12:328339 184. Boutron P (1993) Glass-forming tendency

172. Franks F (1982) The properties of aqueous and stability of the amorphous state in solu-
solutions at subzero temperatures. In: Franks tions of a 2,3-butanediol containing mainly
F (ed) Water, a comprehensive treatise. the levo and dextro isomers in water, buffer,
Plenum, NewYork, NY, pp215338 and Euro-Collins. Cryobiology 30:8697
185. Baudot A, Alger L, Boutron P (2000) Glass- 198. Wowk B (2013) Adaptation of a commercial
forming tendency in the system water- diathermy machine for radiofrequency warm-
dimethyl sulfoxide. Cryobiology 40:151158 ing of vitrified organs. Cryobiology 67:404
186. Vigier G, Vassoille R (1987) Ice nucleation 199. Etheridge ML, Xu Y, Choi J, Bischof JC

and crystallization in water-glycerol mixtures. (2013) Radiofrequency heating of magnetic
Cryobiology 24:345354 nanoparticle cryoprotectant solutions for
187. Luyet B (1967) On the possible biological improved cryopreservation protocols.
significance of some physical changes encoun- Cryobiology 67:398399
tered in the cooling and the rewarming of 200. Inada T, Lu S (2003) Inhibition of recrystal-
aqueous solutions. In: Asahina E (ed) Cellular lization of ice grains by adsorption of
injury and resistance in freezing organisms. poly(vinyl alcohol) onto ice surfaces. Cryst
Hokkaido University, Sapporo, pp120 Growth Des 3:747752
188. Baudot A, Odagescu V (2004) Thermal prop- 201. Deller RC, Vatish M, Mitchell DA, Gibson
erties of ethylene glycol and aqueous solu- MI (2014) Synthetic polymers enable non-
tions. Cryobiology 48:283294 vitreous cellular cryopreservation by reducing
189. Fahy GM (1987) Biological effects of vitrifi- ice crystal growth during thawing. Nat
cation and devitrification. In: Pegg DE, Commun 5:3244
Karow AM Jr (eds) The biophysics of organ 202. Tan X, Song E, Liu X, Liu G, Cheng H, Wan
cryopreservation. Plenum Press, NewYork, F (2012) Successful vitrification of mouse
NY, pp265293 ovaries using less-concentrated cryoprotec-
190. Fahy GM, Levy DI, Ali SE (1987) Some
tants with Supercool X-1000 supplementa-
emerging principles underlying the physical tion. In Vitro Cell Dev Biol Anim 48:6974
properties, biological actions, and utility of 203. Fahy GM, Wowk B, Wu J, Paynter S (2004)
vitrification solutions. Cryobiology Improved vitrification solutions based on pre-
24:196213 dictability of vitrification solution toxicity.
191. Fahy GM (1995) The role of nucleation in Cryobiology 48:2235
cryopreservation. In: Lee RE, Warren GJ, 204. Ting AY, Yeoman RR, Lawson MS, Zelinski
Gusta LV (eds) Biological ice nucleation and MB (2012) Synthetic polymers improve vitri-
its applications. APS Press, St. Paul, MO, fication outcomes of macaque ovarian tissue
pp315336 as assessed by histological integrity and the
192. Baudot A, Peyridieu JF, Boutron P, Mazuer J, invitro development of secondary follicles.
Odin J (1996) Effects of saccharides on the glass- Cryobiology 65:111
forming tendency and stability of solutions of 205. Pichugin Y, Fahy GM, Morin R (2006)

2,3-butanediol, 1,2-propanediol, or 1,3-butane- Cryopreservation of rat hippocampal slices by
diol in water, phosphate-buffered saline, Euro- vitrification. Cryobiology 52:228240
Collins solution, or Saint Thomas cardioplegic 206. Fahy GM, Guan N, De Graaf IAM, Tan Y,
solution. Cryobiology 33:363375 Griffin L, Groothuis GMM (2013)
193. Wowk B, Fahy GM (2005) Toward large
Cryopreservation of precision-cut tissue
organ vitrification: extremely low critical cool- slices. Xenobiotica 43:113132
ing and warming rates of M22 vitrification 207. Guan N, Blomsma SA, Fahy GM, Groothuis
solution. Cryobiology 51:362 GM, de Graaf IAM (2013) Analysis of gene
194. Fahy G (2013) Consequences and control of expression changes to elucidate the mecha-
ice formation in the renal inner medulla. nism of chilling injury in precision-cut liver
Cryobiology 67:409410 slices. Toxicol In Vitro 27:890899
195. Wusteman M, Robinson M, Pegg D (2004) 208. Ting AY, Yeoman RR, Campos JR, Lawson
Vitrification of large tissues with dielectric MS, Mullen SF, Fahy GM, Zelinski MB
warming: biological problems and some (2013) Morphological and functional preser-
approaches to their solution. Cryobiology vation of pre-antral follicles after vitrification
48:284294 of macaque ovarian tissue in a closed system.
196. Ruggera PS, Fahy GM (1990) Rapid and uni- Hum Reprod 28:12671279
form electromagnetic heating of aqueous 209. Kami D, Kasuga J, Arakawa K, Fujikawa S
cryoprotectant solutions from cryogenic tem- (2008) Improved cryopreservation by diluted
peratures. Cryobiology 27:465478 vitrification solution with supercooling-
197. Burdette EC, Wiggins S, Brown R, Karow
facilitating flavonol glycoside. Cryobiology
AM Jr (1980) Microwave thawing of frozen 57:242245
kidneys: a theoretically based experimentally- 210.
Matsumura K, Hyon SH (2009)
effective design. Cryobiology 17:393402 Polyampholytes as low toxic efficient cryopro-
tective agents with antifreeze protein proper- 224. Isachenko E, Isachenko V, Katkov II, Sanchez
ties. Biomaterials 30:48424849 R, van der Ven H, Nawroth F (2007)
211. Matsumura K, Bae JY, Hyon SH (2010)
Cryoprotectant-free vitrification of spermato-
Polyampholytes as cryoprotective agents for zoa. In: Tucker MJ, Liebermann J (eds)
mammalian cell cryopreservation. Cell Vitrification in assisted reproduction. A users
Transplant 19:691699 manual and trouble-shooting guide. Informa
212. Matsumura K, Bae JY, Kim HH, Hyon SH Healthcare, London, pp87105
(2011) Effective vitrification of human 225. Morris GJ, Acton E, Murray BJ, Fonseca F
induced pluripotent stem cells using carboxyl- (2012) Freezing injury: the special case of the
ated -poly-l-lysine. Cryobiology 63:7683 sperm cell. Cryobiology 64:7180
213. Maehara M, Sato M, Watanabe M, Matsunari 226. Nawroth F, Isachenko V, Dessole S, Rahimi
H, Kokubo M, Kanai T, Sato M, Matsumura G, Farina M, Vargiu N, Mallmann P, Dattena
K, Hyon SH, Yokoyama M, Mochida J, M, Capobianco G, Peters D etal (2002)
Nagashima H (2013) Development of a novel Vitrification of human spermatozoa without
vitrification method for chondrocyte sheets. cryoprotectants. Cryo Lett 23:93102
BMC Biotechnol 13:58 227. Isachenko E, Isachenko V, Katkov II, Rahimi
214. Shibao Y, Fujiwara K, Kawasaki Y, Matsumura G, Schondorf T, Mallmann P, Dessole S,
K, Hyon SH, Kashiwazaki N (2014) The Nawroth F (2004) DNA integrity and motil-
effect of a novel cryoprotective agent, carbox- ity of human spermatozoa after standard slow
ylated epsilon-poly-l-lysine, on the develop- freezing versus cryoprotectant-free vitrifica-
mental ability of re-vitrified mouse embryos at tion. Hum Reprod 19:932939
the pronuclear stage. Cryobiology 68:200 228. Isachenko V, Isachenko E, Katkov II, Montag
215. Watanabe H, Kohaya N, Kamoshita M, M, Dessole S, Nawroth F, Van der Ven H
Fujiwara K, Matsumura K, Hyon SH, Ito J, (2004) Cryoprotectant-free cryopreservation
Kashiwazaki N (2013) Efficient production of of human spermatozoa by vitrification and
live offspring from mouse oocytes vitrified with freezing in vapor: effect on motility, DNA
a novel cryoprotective agent, carboxylated integrity, and fertilization ability. Biol Reprod
epsilon-L-lysine. PLoS One 8:e83613 71:11671173
216. Jimenez-Rios JL, Rabin Y (2006) Thermal 229. Merino O, Sanchez R, Risopatron J, Isachenko
expansion of blood vessels in low cryogenic tem- E, Katkov II, Figueroa E, Valdebenito I,
peratures, part II: vitrification with VS55, DP6, Mallmann P, Isachenko V (2012)
and 7.05M DMSO.Cryobiology 52:284294 Cryoprotectant-free vitrification of fish
217. Rabin Y, Steif PS, Hess JL, Jimenez-Rios JL, (Oncorhynchus mykiss) spermatozoa: first
Palastro MC (2006) Fracture formation in report. Andrologia 44(Suppl 1):390395
vitrified thin films of cryoprotectants. 230. Sherman JK, Kiu KC (1982) Ultrastructure
Cryobiology 53:7595 before freezing, while frozen, and after thaw-
218. Steif PS, Palastro MC, Rabin Y (2007) The ing in assessing cryoinjury of mouse epididy-
effect of temperature gradients on stress mal spermatozoa. Cryobiology 19:503510
development during cryopreservation via vit- 231. Morris GJ (2006) Rapidly cooled human

rification. Cell Preserv Tech 5:104115 sperm: no evidence of intracellular ice forma-
219. Nei T (1976) Freezing injury to erythrocytes. tion. Hum Reprod 21:20752083
I.Freezing patterns and post-thaw hemolysis. 232. Isachenko V, Soler C, Isachenko E, Perez-

Cryobiology 13:278286 Sanchez F, Grishchenko V (1998) Vitrification
220. Levin R, Cravalho EG, Huggins CE (1977) of immature porcine oocytes: effects of lipid
Effect of solution non-ideality on erythrocyte droplets, temperature, cytoskeleton, and
volume regulation. Biochim Biophys Acta addition and removal of cryoprotectant.
465:179190 Cryobiology 36:250253
221. Meryman HT (1966) Review of biological
233. He X, Park EYH, Fowler A, Yarmush ML,
freezing. In: Meryman HT (ed) Cryobiology. Toner M (2008) Vitrification by ultra-fast
Academic, NewYork, NY, pp1114 cooling at a low concentration of cryoprotec-
tants in a quartz micro-capillary: a study using
222. Henniker JC, McBain JW (1948) The depth murine embryonic stem cells. Cryobiology
of the surface zone of a liquid. Technical 56:223232
Report no. 5, N60ri-154-T.O.II.Stanford
Research Institute, Stanford, CA 234. Asahina E, Shimada K, Hisada Y (1970) A
stable state of frozen protoplasm with invisi-
223. Vogler EA (1998) Structure and reactivity of ble intracellular ice crystals obtained by rapid
water at biomaterial surfaces. Adv Colloid cooling. Exp Cell Res 59:349358
Interface Sci 74:69117
235. Takahashi T, Hirsh A, Erbe E, Williams RJ permeable cryoprotectants: isolated cells.

(1988) Mechanism of cryoprotection by Cryobiology 18:3248
extracellular polymeric solutes. Biophys J 248. Levin R (1982) A generalized method for the
54:509518 minimization of cellular osmotic stresses and
236. MacFarlane DR (1987) Physical aspects of
strains during the introduction and removal
vitrification in aqueous solutions. Cryobiology of permeable cryoprotectants. J Biomec Eng
24:181195 104:8186
237. Clark P, Fahy GM, Karow AM Jr (1984)
249. Fahy GM, Lilley TH, Linsdell H, St. John
Factors influencing renal cryopreservation. Douglas M, Meryman HT (1990)
II.Toxic effects of three cryoprotectants in Cryoprotectant toxicity and cryoprotectant
combination with three vehicle solutions in toxicity reduction: in search of molecular
non-frozen rabbit cortical slices. Cryobiology mechanisms. Cryobiology 27:247268
21:274284 250. Fahy GM (2005) Vitrification as an approach
238. Karow AM Jr, McDonald M, Dendle T, Rao to cryopreservation: general perspectives.
R (1986) Functional preservation of the Cryobiology 51:348349
mammalian kidney. VII.Autologous trans- 251. Wowk B, Darwin M, Harris SB, Russell SR,
plantation of dog kidneys after treatment with Rasch CM (1999) Effects of solute methoxyl-
dimethylsulfoxide (2.8 and 4.2 M). ation on glass-forming ability and stability of
Transplantation 41:669674 vitrification solutions. Cryobiology
239. Pegg DE, Jacobsen IA, Diaper MP, Foreman 39:215227
J (1986) Optimization of a vehicle solution 252. Benson JD, Kearsley AJ, Higgins AZ (2012)
for the introduction and removal of glycerol Mathematical optimization of procedures for
with rabbit kidneys. Cryobiology 23:5363 cryoprotectant equilibration using a toxicity
240.
Meryman HT, Douglas MSJ (1982) cost function. Cryobiology 64:144151
Isotonicity in the presence of penetrating 253. Karlsson JO, Szurek EA, Higgins AZ, Lee SR,
cryoprotectants. Cryobiology 19:565569 Eroglu A (2013) Optimization of cryoprotec-
241. Meryman HT (2007) Cryopreservation of
tant loading into murine and human oocytes.
living cells: principles and practice. Transfusion Cryobiology 68:1828
47:935945 254. Franks F, Mathias SF, Galfre P, Webster SD,
242. Nobel PS (1969) The Boyle-Vant Hoff rela- Brown D (1983) Ice nucleation and freezing
tion. J Theor Biol 23:375379 in undercooled cells. Cryobiology
243. Vian AM, Higgins AZ (2014) Membrane
20:298309
permeability of the human granulocyte to 255. Fink AL (1986) Effects of cryoprotectants on
water, dimethyl sulfoxide, glycerol, propylene enzyme structure. Cryobiology 23:2837
glycol and ethylene glycol. Cryobiology 256. Blow AM, Botham GM, Fisher D, Goodall
68:3542 AH, Tilcock CP, Lucy JA (1978) Water and
244. Mullen SF, Li M, Li Y, Chen ZJ, Critser JK calcium ions in cell fusion induced by
(2008) Human oocyte vitrification: the per- poly(ethylene glycol). FEBS Lett
meability of metaphase II oocytes to water 94:305310
and ethylene glycol and the appliance toward 257. Shaw JM, Kuleshova LL, MacFarlane DR,

vitrification. Fertil Steril 89:18121825 Trounson AO (1997) Vitrification properties
245. Jin B, Higashiyama R, Nakata Y, Yonezawa J, of solutions of ethylene glycol in saline con-
Xu S, Miyake M, Takahashi S, Kikuchi K, taining PVP, ficoll, or dextran. Cryobiology
Yazawa K, Mizobuchi S, Kitayama M, 35:219229
Koshimoto C, Matsukawa K, Kasai M, 258. Kuleshova LL, MacFarlane DR, Trounson

Edashige K (2013) Rapid movement of water AO, Shaw JM (1999) Sugars exert a major
and cryoprotectants in pig expanded blasto- influence on the vitrification properties of
cysts via channel processes: its relevance to ethylene glycol-based solutions and have low
their higher tolerance to cryopreservation. toxicity to embryos and oocytes. Cryobiology
Biol Reprod 89:87 38:119130
246. Kleinhans FW (1998) Membrane permeabil- 259. Fahy GM, da Mouta C, Tsonev L, Khirabadi
ity modeling: Kedem-Katchalsky vs a two- BS, Mehl P, Meryman HT (1995) Cellular
parameter formalism. Cryobiology injury associated with organ cryopreservation:
37:271289 chemical toxicity and cooling injury. In:
247. Levin R, Miller TW (1981) An optimum
Lemasters JJ, Oliver C (eds) Cell biology of
method for the introduction or removal of trauma. CRC Press, Boca Raton, FL

273. Leth PM, Gregersen M (2005) Ethylene gly-
neutralization. Cryobiology 60:S45S53 col poisoning. Forensic Sci Int 155:179184
261. Yancey PG, Clark ME, Hand SC, Rowlus RD, 274. Jacob SW, Herschler R (1986) Pharmacology
Somero GN (1982) Living with water stress: of DMSO.Cryobiology 23:1427
evolution of osmolyte systems. Science 275. Snow JT, Finley JW, Friedman M (1975)

217:12141222 Oxidation of sulfhydryl groups to disulfides
262. Somero GN (1986) From dogfish to dogs: by sulfoxides. Biochem Biophys Res Commun
trimethylamines protect proteins from urea. 64:441
News Physiol Sci 1:912 276. Agarwal SP, Nwaiwu J (1984) Dimethyl sulf-
263. Wang A, Bolen DW (1997) A naturally occur- oxide as a spray reagent for the detection of
ring protective system in urea-rich cells: triterpenoids and some steroids on thin-layer
mechanisms of osmolyte protection of pro- plates. J Chromatogr 295:537542
teins against urea denaturation. Biochemistry 277. Friend C, Scher W, Preisler HD, Holland JG
36:91019108 (1973) Studies on erythroid differentiation of
264. Lin TY, Timasheff SN (1994) Why do some Friend virus-induced murine leukemia cells.
organisms use a urea-methylamine mixture as Bibl Haematol 39:916923
osmolyte? Thermodynamic compensation of 278. Reboulleau CP, Shapiro HS (1983) Chemical
urea and trimethylamine N-oxide interactions inducers of differentiation cause conforma-
with protein. Biochemistry 33:1269512701 tional changes in the chromatin and deoxyri-
265. Baxter S, Lathe G (1971) Biochemical effects bonucleic acid of murine erythroleukemia
on kidney of exposure to high concentrations cells. Biochemistry 22:45124517
of dimethyl sulphoxide. Biochem Pharmacol 279. Arakawa T, Kita Y, Timasheff SN (2007)
30:10791091 Protein precipitation and denaturation by
266. Mirsalikhova NM (1978) Stabilizatsiia Na+, dimethyl sulfoxide. Biophys Chem 131:6270
K+-adenozintrifosfatazy dimetilsulfoxsidom pri 280. Gekko K, Timasheff SN (1981) Mechanism
inaktivatsii mochevinoi. Biokhimiia 43:3439 of protein stabilization by glycerol: preferen-
267. Anchordoguy TJ, Carpenter JF, Crowe JH, tial hydration in glycerol-water mixtures.
Crowe LM (1992) Temperature-dependent Biochemistry 20:46674676
perturbation of phospholipid bilayers by 281. Arakawa T, Timasheff SN (1982) Stabilization
dimethylsulfoxide. Biochim Biophys Acta of protein structure by sugars. Biochemistry
1104:117122 21:65366544
268. Shlafer M, Matheny JL, Karow AM Jr (1976) 282. Timasheff SN (2002) Protein hydration, ther-
Cardiac chronotropic mechanisms of dimethyl modynamic binding, and preferential hydra-
sulfoxide: inhibition of acetylcholinesterase tion. Biochemistry 41:1347313482
and antagonism of negative chronotropy by 283. Kim SH, Yan YB, Zhou HM (2006) Role of osmo-
atropine. Arch Int Pharmacodyn Ther lytes as chemical chaperones during the refolding of
221:2131 aminoacylase. Biochem Cell Biol 84:3038
269. de Graaf IA, Draaisma AL, Schoeman O,
284. Hamagushi K (1964) Structure of murami-
Fahy GM, Groothuis GM, Koster HJ (2007) dase (lysozyme). VIII.Effect of dimethyl
Cryopreservation of rat precision-cut liver sulfoxide on the stability of muramidase.

and kidney slices by rapid freezing and vitrifi- J Biochem 56:441449
cation. Cryobiology 54:112
285. Fujita Y, Izumiguchi S, Noda Y (1982) Effect
270. Burch HB, Lowry OH, Meinhardt L, Max P, of dimethylsulfoxide and its homologues
Chyu K-J (1970) Effect of fructose, dihy- on the thermal denaturation of lysozyme as
droxyacetone, glycerol and sucrose on metab- measured by differential scanning calorimetry.
olites and related compounds in liver and J Pept Prot Res 19:2531
kidney. J Biol Chem 245:20922102
286. Arakawa T (1990) The basis for toxicity of
271. Jans AWH, Willem R (1988) 13C-NMR
certain cryoprotectants: a hypothesis. Cryo
study of glycerol metabolism in rabbit renal biology 27:401415
cells of proximal convoluted tubules. Eur J
Biochem 174:6773 287. Strambini GB, Balestreri E, Galli A, Gonnelli
M (2008) Effects of sugars and polyols on the
272. Booth ED, Dofferhoff O, Boogaard PJ,
stability of azurin in ice. J Phys Chem B
Watson WP (2004) Comparison of the 112:43724380
metabolism of ethylene glycol and glycolic
acid invitro by precision-cut tissue slices from 288. MacFarlane DR, Forsyth M (1990) Recent
female rat, rabbit, and human liver. insights on the role of cryoprotective agents
Xenobiotica 34:3148 in vitrification. Cryobiology 27:345358
289. Yoshida T, Maulik N, Ho YS, Alam J, Das DK 302.

He L, Bailey JL, Buhr MM (2001)
(2001) H(mox-1) constitutes an adaptive Incorporating lipids into boar sperm decreases
response to effect antioxidant cardioprotection: chilling sensitivity but not capacitation poten-
a study with transgenic mice heterozygous for tial. Biol Reprod 64:6979
targeted disruption of the Heme oxygenase-1 303. Horvath G, Seidel GE Jr (2006) Vitrification
gene. Circulation 103:16951701 of bovine oocytes after treatment with
290. Arakawa T, Bhat R, Timasheff SN (1990)
cholesterol-loaded methyl-beta-cyclodextrin.
Why preferential hydration does not always Theriogenology 66:10261033
stabilize the native structure of globular pro- 304. Tomczak MM, Hincha DK, Estrada SD,

teins. Biochemistry 29:19241931 Wolkers WF, Crowe LM, Feeney RE, Tablin
291. Farrant J (1965) Permeability of guinea-pig F, Crowe JH (2002) A mechanism for stabili-
smooth muscle to non-electrolytes. J Physiol zation of membranes at low temperatures by
178:113 an antifreeze protein. Biophys J 82:874881
292. Fahy GM (1980) Analysis of solution
305. Wada H, Gombos Z, Murata N (1990)

effects injury: rabbit renal cortex frozen in Enhancement of chilling tolerance of a cyano-
the presence of dimethyl sulfoxide. Cryo bacterium by genetic manipulation of fatty
biology 17:371388 acid desaturation. Nature 347:200203
293. Sonna LA, Fujita J, Gaffin SI, Lilly CM
306. Murata N, Ishizaki-Nishizawa O, Higashi S,
(2002) Effects of heat and cold stress on Hayashi H, Tasaka Y, Nishida I (1992)
mammalian gene expression. J Appl Physiol Genetically engineered alteration in the chill-
92:17251742 ing sensitivity of plants. Nature 356:710713
294. Al-Fageeh MB, Marchant RJ, Carden MJ, Smales 307. Burton V, Mitchell HK, Young P, Petersen
CM (2006) The cold shock response in cultured NS (1988) Heat shock protection against
mammalian cells: harnessing the response for the cold stress of Drosophila melanogaster. Mol
improvement of recombinant protein production. Cell Biol 8:35503552
Biotechnol Bioeng 93:829835 308.
VanBogelen RA, Neidhardt FC (1990)
295. Inouye M, Yamanaka K (eds) (2000) Cold Ribosomes as sensors of heat and cold shock
shock response and adaptation. Horizon in Escherichia coli. Proc Natl Acad Sci U S A
Scientific Press, Norfolk 87:55895593
296. Lyons JM, Raison JK (1970) Oxidation activ- 309. Schindelin H, Marahiel MA, Heinemann U
ity of mitochondria isolated from plant tissue (1993) Universal nucleic acid-binding domain
sensitive and resistant to chilling injury. Plant revealed by crystal structure of the B. subtilis
Physiol 45:386389 major cold-shock protein. Nature 364:
297. Lyons JM (1972) Phase transitions and
164168
control of cellular metabolism at low temper- 310. Corwin WL, Baust JM, Baust JG, Van Buskirk
atures. Cryobiology 9:341350 RG (2011) The unfolded protein response in
298. Ghetler Y, Yavin S, Shalgi R, Arav A (2005) human corneal endothelial cells following
The effect of chilling on membrane lipid hypothermic storage: implications of a novel
phase transition in human oocytes and stress pathway. Cryobiology 63:4655
zygotes. Hum Reprod 20:33853389 311. Zhou GB, Li N (2009) Cryopreservation of
299.
Drobnis EZ, Crowe LM, Berger T, porcine oocytes: recent advances. Mol Hum
Anchordoguy TJ, Overstreet JW, Crowe JH Reprod 15:279285
(1993) Cold shock damage is due to lipid 312. Zhang W, Yi K, Yan H, Zhou X (2012)

phase transitions in cell membranes: a demon- Advances on invitro production and cryo-
stration using sperm as a model. J Exp Zool preservation of porcine embryos. Anim
265:432437 Reprod Sci 132:115122
300. Hays LM, Feeney RE, Crowe LM, Crowe
313. Cameron RD, Beebe LF, Blackshaw AW

JH, Oliver AE (1996) Antifreeze glycopro- (2006) Cryopreservation and transfer of pig
teins inhibit leakage from liposomes during embryos. Soc Reprod Fertil Suppl 62:
thermotropic phase transitions. Proc Natl 277291
Acad Sci U S A 93:68356840 314. Nagashima H, Hiruma K, Saito H, Tomii R,
301. Hays LM, Crowe JH, Wolkers W, Rudenko S Ueno S, Nakayama N, Matsunari H, Kurome
(2001) Factors affecting leakage of trapped M (2007) Production of live piglets following
solutes from phospholipid vesicles during cryopreservation of embryos derived from
thermotropic phase transitions. Cryobiology invitro-matured oocytes. Biol Reprod 76:
42:88102 900905
315. Beeb LF, Cameron RD, Blackshaw AW,

surements in human peripheral blood mono-
Higgins A, Nottle MB (2002) Piglets born nuclear cells. Transfusion 36:303308
from centrifuged and vitrified early and peri- 325. Valeri CR, Pivacek LE, Gray AD, Cassidy GP,
hatching blastocysts. Theriogenology Leavy ME, Dennis RC, Melaragno AJ, Niehoff
57:21552165 J, Yeston N, Emerson CP, Altschule MD
316. Pickering SJ, Johnson MH (1987) The influ- (1989) The safety and therapeutic effectiveness
ence of cooling on the organization of the of human red cells stored at 80C for as long
meiotic spindle of the mouse oocyte. Hum as 21 years. Transfusion 29:429437
Reprod 2:207216 326. Sun WQ, Leopold AC, Crowe LM, Crowe
317. Wang WH, Meng L, Hackett RJ, Odenbourg JH (1996) Stability of dry liposomes in sugar
R, Keefe DL (2001) Limited recovery of mei- glasses. Biophys J 70:17691776
otic spindles in living human oocytes after 327. Walters C (2004) Temperature dependency of
cooling-rewarming observed using polarized molecular mobility in preserved seeds.
light microscopy. Hum Reprod Biophys J 86:12531258
16:23742378 328. Walters C (2004) Longevity of cryogenically
318. Yang D, Winslow KL, Nguyen K, Duffy D, stored seeds. Cryobiology 48:229244
Freeman M, Al-Shawaf T (2010) Comparison 329. Shamblin SL, Tang X, Chang LJ, Hancock
of selected cryoprotective agents to stabilize BC, Pikal MJ (1999) Characterization of the
meiotic spindles of human oocytes during time scales of molecular motion in pharmaco-
cooling. J Exp Clin Assist Reprod. 7: pii: 4 logically important glasses. J Phys Chem B
319. Mandelbaum J, Anastasiou O, Levy R, Guerin 103:41134121
J-F, de Larouziere V, Antoine JM (2004) 330. Fink AL (1979) Enzyme-catalyzed reactions
Effects of cryopreservation on the meiotic in unfrozen, noncellular systems at subzero
spindle of human oocytes. Eur J Obstet temperatures. In: Fennema O (ed) Proteins at
Gynecol Reprod Biol 113(Suppl 1):S17S23 low temperatures. American Chemical
320. McGrath JJ (1989) Protecting biological
Society, Washington, DC, pp3554
material against chilling injury. European 331. Douzou P (1977) Cryobiochemistry. An

Patent Office introduction. Academic, NewYork, NY
321. Fahy GM (2007) Hypertonic reduction of
332. Brandts JF, Fu J, Nordin JH (1970) The low
chilling injury, USPTO (ed). United States: temperature denaturation of chymotrypsino-
21st Century Medicine, Inc. gen in aqueous solution and in frozen aque-
322. Song YC, Li CY, Lightfoot FG, Boggs JM, ous solution. In: Wolstenholme GEW,
Greene ED, Taylor MJ, Chen ZZ, Brockbank OConnor M (eds) The frozen cell. J & A
KGM (2002) Stability of vitrified tissues in Churchill, London, pp189209
the vapor phase of liquid nitrogen. 333. Dias CL, Ala-Nissila T, Wong-ekkabut J,

Cryobiology 45:243 Vattulainen I, Grant M, Karttunen M (2010)
323. Rowe AW (1987) Sixteen year cryogenic stor- The hydrophobic effect and its role in cold
age of blood frozen by droplet freezing tech- denaturation. Cryobiology 60:9199
nique and a low glycerol-rapid freeze 334. Lopez CF, Darst RK, Rossky PJ (2008)

procedure. In: Proc XVII Intern Congress Mechanistic elements of protein cold dena-
Refrigeration. Paris: International Institute of turation. J Phys Chem B 112:59615967
Refrigeration, pp135140
335. Fennema O (1979) Proteins at low tempera-
324. Valeri CR, Pivacek LE (1996) Effects of the tures. In: Comstock MJ (ed) Advances in
temperature, the duration of frozen storage, chemistry, vol 180. American Chemical
and the freezing container on invitro mea- Society, Washington, DC
Chapter 3
Modeling andOptimization ofCryopreservation

JamesD. Benson
Abstract
Modeling plays a critical role in understanding the biophysical processes behind cryopreservation. It facili-
tates understanding of the biophysical and some of the biochemical mechanisms of damage during all
phases of cryopreservation including CPA equilibration, cooling, and warming. Modeling also provides a
tool for optimization of cryopreservation protocols and has yielded a number of successes in this regard.
While modern cryobiological modeling includes very detailed descriptions of the physical phenomena that
occur during freezing, including ice growth kinetics and spatial gradients that define heat and mass trans-
port models, here we reduce the complexity and approach only a small but classic subset of these problems.
Namely, here we describe the process of building and using a mathematical model of a cell in suspension
where spatial homogeneity is assumed for all quantities. We define the models that describe the critical cell
quantities used to describe optimal and suboptimal protocols and then give an overview of classical meth-
ods of how to determine optimal protocols using these models.
Key words Mass transport, Boyle vant Hoff, Chemical potential, Freezing point depression, Phase
diagram, Virial equation, Optimization
1 Introduction
Theoretical and practical inroads from mathematical modeling

began over 50 years ago in cryobiology. At the intersection of biol-
ogy, engineering, physics, chemistry, and mathematics, and like
other similar fields, the relationship between biology and biophys-
ics forged in cryobiology has been fruitful for both fields.
The first foundational cryobiological model was Mazurs bio-
physical model of intracellular state as a function of cooling rate
proposed in 1963 [1]. This model lent support to an experimen-
tally verifiable theory of cell death as a function of too-high cooling
rates and set the stage for modelers to use similar approaches to
optimize and understand cryobiological protocols. This was fol-
lowed by a biochemical argument for the damage from too-low
cooling rates [2]. The combination of these two creates a proto-
type of optimization of cooling rates in (quasi) equilibrium cryo-
preservation protocols. In fact, modern cooling rate optimization
83
84 James D. Benson
still combines a mass transport model (to estimate intracellular

water concentration), the phase diagram, and an ice kinetics model
(see, e.g., [3] or [4]).
A relatively more recent development in cryobiological model-
ing is the modeling of CPA equilibration processes. Addressed in
more detail below, note that the step-change exposure of cells to
high concentrations of permeating cryoprotectants causes the cells
to experience a rapid loss of water volume due to the temporary
large osmotic transmembrane gradient coupled with the differen-
tial permeability to water and CPA in the cell membrane. This
exosmosis can cause the cell to shrink below a critical volume,
called a lower osmotic tolerance limit, associated with irrevers-
ible cell damage, often times after swelling back to isosmotic vol-
ume, suggesting possible membrane fusion at low volumes as the
mechanism of damage [5]. Modeling has played a significant role
both in demonstrating that some practices are likely unsuccessful
due to these damaging effects [6] and in suggesting safe CPA
equilibration strategies that end with an equivalent final concentra-
tion but through step-wise or other gradual approaches [713].
Modeling has matured since the 1960s, both from a biophys-
ical model point of view and from a computational point of view.
Most of this chapter will address some of the changes to the for-
mer. To the latter, note that one can now easily solve nonlinear
differential equations with minimal forethought, numerical opti-
mization of protocols can be done using off the shelf packages,
and numerical visualization and graphics production is trivial.
These are distinct advantages to all modern cryobiologists, but
one of the additional benefits is that the computational modeling
of cryobiological processes is considerably more accessible to
non-mathematician scientists. In fact, one of the aims of this
chapter is to convince the non-mathematician scientist that mod-
eling provides a valuable tool for optimization of cryopreserva-
tion protocols.
While considerable modeling advances in cryopreservation
have been made, there is still much work at the forefront of the
field, including attempts to understand the relationship between
cooling rate, concentration, viscosity and the likelihood of crystal-
lization or recrystallization events. Moreover, there are new ques-
tions about the necessity of optimal cooling rates if ultrarapid
warming rates are available [14] (see Note 1), and there are open
questions about model selection, model temperature dependence,
appropriate solution theories, among many others.
The scope of modeling in modern cryobiology is very broad,
and includes heat and mass transport in tissues, organs, and hybrid
systems, as well as complicated ice dynamics and formation model-
ing in the presence and absence of cellular systems [1518]. It now
even includes informatics approaches to understanding cryobio-
logical outcomes [19]. This chapter, however, presents the now
Modeling andOptimization ofCryopreservation 85
classical modeling and optimization of single-cell cryopreservation

protocols. For a compete experimental approach that includes bio-
physical measurement of parameters, prediction of optimal cryo-
preservation protocols based on those parameters and the
approaches described in this chapter, and the experimental valida-
tion and discussion of the results of these models see the series by
Kashuba etal.[2022]. Here we also assume no spatially depen-
dent gradients in temperature or concentration. These systems
include the cryopreservation of most cultured cells [21, 22], gam-
etes [23], and even embryos and blastocysts [4], among others. In
this chapter our aims are as follows: to present the ideas needed to
construct and understand the standard single cell models and then
present several optimization schemas.
2 Model Selection
2.1 Cell Volume Modeling cryobiological protocols depends on a thorough knowl-

edge of the cellular state including mole fraction or concentration
of all of the intracellular components as a function of time, tem-
perature, and protocol. Typical experiments rarely yield complete
cellular state informationthe available measurement is usually
either cell volume or intracellular water volume or their proxies.
Moreover, cell volume limits (known as osmotic tolerance limits,
see Subheading3.1) are given in terms of total cell volume. In prac-
tice, then, cryobiological modeling takes advantage of the relation-
ships between the volumes of the components to describe the total
volume, and deduces the state of all intracellular constituents
(see Note 2).
The total volume V of a cell is given by
J K
V = W + v si Si + vni N i + Vb* , (1)
i =1 i =1
where Si and Ni indicate moles of J intracellular permeating and K
non-permeating solutes with their associated partial molar volumes
by v si and vni , respectively, and Vb* is the so-called osmotically
inactive volume of the cell, consisting of cell and organelle mem-
branes, protein complexes, their associated bound water and sol-
utes, and other non-transportable material.
Differentiating Eq.1 with respect to time, t, gives
J
dV dW dS K
dN i
= + v si i + vni . (2)
dt dt i =1 dt i =1 dt
Here, the usual assumption is made that, on the time scale of
interest for most cryobiological experiments and procedures,
dN i dS j
vni vs j for all i = 1, , K and at least one j = 1, , J .
dt dt
86 James D. Benson
Inother words, while we recognize that there is a vast body of

literature on ionic transport, the relative permeability of water

(often referred to in terms of Hydraulic Conductivity Lp) is an
order of magnitude greater than that of permeating cryoprotec-
tants such as Me2SO, 1,2-propanediol, etc. which in turn are at
least an order of magnitude greater than those of ionic compo-
nents such as salts [24]. Therefore, for prediction of the critical cell
volume, cellular water volume, and intracellular CPA concentra-
tion, it is considerably simpler and effective in the cryobiological
K
dN i
case to assume vni = 0 . This assumption is not well explored
i =1 dt
in the cryobiology literature.
K
dN i
Using the assumption vni = 0 , the total osmotically
i =1 dt
K
inactive volume is defined as Vb := vni N i + Vb* . Dividing through
i =1
by the isosmotic volume Viso, and defining v = V V iso , gives a
normalized volume equation:
J

v = W + v si Si V iso + vbfrac , (3)
i =1
where vbfrac := Vb V iso is the osmotically inactive fraction of the
isosmotic volume, a commonly reported value in the literature.
2.2 Boyle vant Hoff An intrinsic assumption of membrane mass transport modeling is
that cell water volume, and thus, total cell volume, behaves as an
ideal osmometer. In fact, this is a property observed in nearly all
cell types [2533] over a range of osmolalities from 1/3 i sosmolal
to 10 isosmolal. Classically [34], the Boyle vant Hoff relation-
ship is defined by pW = p 0W0 where and 0 are intracellular
osmolalities, W is the intracellular water volume, and subscript 0
defines a particular known state. Typically 0 is isosmolal and W0 is
the corresponding intracellular water volume. Therefore, the cell
water volume as a function of osmolality can be described by the
relationship W = p 0W0 p .
Prickett etal. [35] point out that the impermeability of some
solutes allows a variant of the Boyle vant Hoff relationship to be
derived from the relationship N=N0, where N is the moles of
intracellular non-permeating solute, by showing that this is equiva-
lent to m0W0 = mW where m is the molality of non-permeating
solute. In this case, we have a molal version of the Boyle vant
Hoff relationship. They show that using this relationship and
anonideal description of osmolality as a function of molality
(see Subheading2.3) a different and potentially more accurate esti-
mate for the osmotically inactive volume may be obtained, and the
accuracy of their model over the usual Boyle vant Hoff relation is
enhanced in the extremely concentrated solutions of interest to

cryobiology. Nevertheless, in either case, this is how the Boyle
vant Hoff relationship is used most oftento allow the replace-
ment of the mole fraction, concentration, or molality of intracel-
lular non-permeating solute in the membrane transport equations
with the inverse of intracellular osmolality or molality. Note that in
the dilute case when 0 >1, the behavior of the molal and osmolal
models is nearly identical as a function of molality.
Mathematically, the Boyle vant Hoff relationship implies that
the water volume is inversely proportional to the intracellular
osmolality (or molality). If a cell satisfies this relationship over a
range of osmolalities, it is said to behave as a linear osmometer.
However, this relationship is understood in the isothermal case. In
particular, the constant of proportionality is related to several tem-
perature dependent parameters that include the relative density of
intracellular water and possibly also the relative density of water as
a function of the concentration of non-permeating solutes.
This determination has an additional benefit. Using the total
volume equation(1), the total equilibrium volume of a cell in anis-
osmotic media containing only non-permeating solutes is
V = W + Vb . Replacing W with the expression from the Boyle vant
Hoff relationship gives
p0
V = W0 + V b . (4)
p
Thus plotting volumes against the inverse of extracellular
osmolality yields what is known as a Boyle vant Hoff plot
(see Fig.1). Note that in these plots it is assumed that the volume
has had sufficient time to equilibrate, but insufficient time for the
non-permeating solute assumption to become invalid. Figure1
shows a typical Boyle vant Hoff plot with a linear regression to
the cell volume axis. This is equivalent to taking the limit as goes
to infinity, and at this limit, the intracellular water volume should
be 0, leaving the osmotically inactive volume, Vb. Frequently
the axes are normalized as in Eq.3, which is the form shown in
Fig.1. To recover the volume, one may multiply by the normal-
izing value, Viso.
Finally, there has been some recent discussion about the cor-
rect experimental and statistical approach to this regression. In par-
ticular, Katkov [36, 37] makes several arguments, including that
the regression line should be forced to go through the point
defined by isosmotic volume at isosmolality. However, we show in
[38] that this argument is invalid on theoretical and statistical
grounds and has the potential to introduce errors.
2.3 Osmolality In most physiologic literature, the approximation of the osmolality

J +K
andChemical
Potential
by p m
i =1
i
and chemical potential i by mi = RT ln mi is
88 James D. Benson
Fig. 1 Boyle vant Hoff plot for mouse B6 embryonic stem cells. Data are from
Kashuba-Benson etal. [20]. The line represents a linear regression of all of the
data. Here we present the normalized volume as a function of the normalized
inverse osmolality. This plot demonstrates that these cells behave as linear
osmometers. By extrapolating to the y-axis we may determine the value of Vb.
Here Vb=0.402. Note that as the plot increases along the x-axis, the osmolality
is decreasing; in other words, values greater than one indicate hyposomolal con-
ditions and values less than one indicate hyperosmolal conditions. Finally, note
that in [20], hyposmotic values were excluded from the regression (see [20] for
details)
reasonable because most physiologic media are relatively dilute and

can be considered ideal solutions, where osmolality and chemi-
cal potential are linear functions of their constituent molalities. In
cryobiological settings, though, this approximation, while often
used in modeling literature, is most often invalid. To wit: the
standard cryopreservation protocol for many cultured cell types
requires equilibration of cells in 10% (v/v) Me2SO.This corre-
sponds to more than 1mol/kg solute which is far from what most
consider dilute. Further, the action of cooling at 1K/min under
the standard cryopreservation protocol causes extracellular ice
to nucleate and crystallize further concentrating the remaining
solution, making it even less dilute. Therefore, dilute approxi-
mations in subzero (and even suprazero transport models of CPA
equilibration) are not likely to produce very accurate predictions of
intracellular state.
While dilute approximations are unlikely to yield accurate pre-
dictions of the intracellular state during cryopreservation proto-
cols, there is still great conceptual utility in exploring relationships
and dynamics with a less precise model. For example, the optimal
CPA equilibration strategies developed by Benson etal. [8, 9, 39]
yield parameter independent extremal trajectories that are likely to

be optimal in the case of nondilute, nonideal modelsa statement
that remains to be mathematically proven and, importantly, experi-
mentally verified.
There have been two approaches to model membrane trans-
port at lower temperatures in these nondilute cases. The first is to
physically measure the osmolality of the particular solution of
interest as a function of temperature and concentration of its con-
stituents. In this case, a differential scanning calorimeter is typically
used to measure the melting temperature of a family of solutions as
a function of the relative concentration of their constituents. Using
Raoults Law (see, e.g.,[40]), or a more thermodynamically appli-
cable variant [41], the relationship between osmolality and freez-
ing point depression is used to determine solution osmolality. To
formulate a model for this relationship, measurements of the melt-
ing temperature of solutions containing concentrations of its con-
stituent species are made. With enough of these measurements, a
graph over the surface of concentration of solutes can be found
that describes the phase diagram. We note that the system dimen-
sion increases with each constituent and thus the number of mea-
surements increases exponentially with the number of constituents.
For example, if n measurements are required to accurately describe
a binary mixture, on the order n2 measurements are required to
accurately describe a ternary mixture, n3 for a quaternary mixture,
etc. This is one of the primary motivators for the synthetic or theo-
retically generated phase diagrams discussed below.
Typically of interest to cryobiologists is the case where the
extracellular media contains one CPA and a primary non-
permeating salt such as NaCl. In this case, the isopleth defined by
fixing the ratio R of, for example, salt and CPA is an important
quantity. This produces a concentration or mass fraction vs melt-
ing temperature curve in Fig.2. These isopleths are useful because
they provide a functional relationship between system water con-
tent and osmolality. Putting a number of these isopleths together a
two-solute (ternary) system can be generated (shown in Fig.3)
where the freezing point depression has a phenomenological model
-Tm = -(38.3 - 0.2145R)w - (81.19 - 0.2909R)w 2 , (5)
where Tm is the freezing point depression.

The second approach is to use a thermodynamic model for
chemical potential or osmolality as a function of the state variables.
In general, thermodynamic models of osmolality and chemical
potential are complicated and require the measurement of mixture
specific parameters [see42, for a brief review]. This measurement,
in essence, is very similar to measuring freezing point depression of
specific mixtures, and as such there was little historic interest in
their utilization in cryobiology. However, two models have been
90 James D. Benson
Fig. 2 Isopleths of the water-rich portion of the ternary system ethylene glycolsodium chloridewater in
terms of freezing point depression with R=5 and R=45. Data are from Benson etal. [106]. To use these, one
could express mass fraction in terms of molality of ethylene glycol and sodium chloride
proposed recently that are accurate enough for solutions of interest

in cryobiology yet require no mixture specific data. In particular,
both models base their predictions on data from models of osmo-
lality of binary solutions such as glycerol and water or sodium chlo-
ride and water. The osmolality of these binary solutions is well
modeled using a quadratic or cubic function in molality:
p (m) = Am + Bm 2 + Cm 3 , (6)
where A, B, and C are coefficients to be determined by fitting, for
example, freezing point depression data. While Kleinhans and
Mazur propose that model (6) is a phenomenological model based
on experimental data for a variety of binary solutions, Elliott etal.
propose that this is essentially an osmotic virial expansion of the
chemical potential in molality, and a similar formulation may be
found using mole fraction [42], an observation originating from
classical thermodynamics [see43, p.267]. Using model (6) for two
solutes, say sodium chloride and glycerol indicated by subscripts 1
and 2, respectively, gives two separate binary osmolality models
with different parameters:
p 1 (m1 ) = A1m1 + B1m12 + C1m13 ,
(7)
p 2 (m2 ) = A2m2 + B2m22 + C 2m23 .

Elliott etal.s thermodynamic derivation of Eqs.6 and 7 pre-
scribes that Ai=1 for all i, though the Kleinhans and Mazur model
does not have this restriction. Additionally, the necessity of the
Fig. 3 Phase diagram of the water-rich portion of the ternary system ethylene gly-
colsodium chloridewater in terms of freezing point depression. Data are from
Benson etal. [106]. Here the phase diagram is a contour plot of freezing point
depression = -Tm = -(38.3 - 0.2145R)w - (81.19 - 0.2909R)w 2
where variables w and R, are, classical to reports of phase diagrams in the cryobio-
logical literature, total solute mass fraction the ratio of ethylene glycol to sodium
chloride, respectively. This formulation is convenient as in slow cooling protocols,
the mass fraction is the only variable that changes as the crystallization of water
into ice increases w with decreasing temperatures (see Fig.2), thus for any initial
point in the plot, the mass fraction then becomes a function of temperature. Here,
solid lines are degrees C of freezing point depression contours, and isopleths are
indicated by the dashed vertical lines on the contour plot
cubic term is dependent on the solute. Finally in this case m1 is the

total molality of the dissociated salt. Elliott etal. account for this
latter quantity by finding the dissociation constant kdiss as part of
fitting binary solution data to model (6) (e.g., setting m1 = kdissmNaCl
and letting mNaCl be the non-dissociated molality of the salt).
Prickett etal.show that for most solutes of cryobiological interest,
the cubic term is negligible and likely superfluous [41], though the
cubic term was critical in modeling larger solutes such as
hemoglobin.
92 James D. Benson
To arrive at a model of osmolality as a function of molality of

both solutes, Kleinhans and Mazur propose a simple additive
model of osmolality where the relative osmolalities of binary mix-
tures as a function of molality are simply summed [44]:
p (m1 , m2 ) = p 1 (m1 ) + p 2 (m2 ),
(8)
= A1m1 + B1m12 + C1m13 + A2m2 + B2m22 + C 2m23 .
On the other hand, Elliott etal.[42] suggest that there are

interactions between solutes that are not sufficiently captured by
model (8) and propose the solute mixing terms with A1=A2 = 1
B1 + B2
p (m1 , m2 ) = p 1 (m1 ) + p 2 (m2 ) + m1m2
2
1 1
+(C12C 2 )3 m12m2 + (C1C 22 )3 m1m22 , (9)
2 3 B + B2 2 3
= m1 + B1m + C1m + m2 + B2m + C 2m + 1
1 1 m1m22 2
2
1 1
+(C12C 2 )3 m12m2 + (C1C 22 )3 m1m22 .

In fact, for an arbitrary number of solutes, Elliott etal. propose
using the arithmetic mean for the quadratic B terms and a geo-
metric mean for the cubic C terms (see Note 3). In particular,
with m = (m1 , m2 , , mn )T ,
n n Bi + B j n
p (m) = mi + mim j + (C C C ) i j k
1 3
mim j mk . (10)
2
i =1 i , j =1 i , j ,k =1

This formulation allows the construction of aqueous phase
diagrams for solutions containing an arbitrary number of solutes.
It also allows the comparison of the solution theory with experi-
mental measurement in Fig.4.
2.3.1 Application The relationship between osmolality and freezing point depression
ofOsmolalityModels (e.g. Raoults Law or its more thermodynamic appropriate
analogue [see, e.g.,41]) along with the fixed ratio R in the pre-
ceding work allows one to calculate extracellular molality or con-
centration of the constituents at a given temperature via either the
phenomenological models defined by fitting experimentally derived
phase diagrams or the synthesized osmolality models (8) and (9).
For example, using Eq.9 and assuming Ci=0 for i=1,2, first define
R = m1 m2 , and thus m1 = Rm2 . Then, at any given tempera-
ture (with unitsC) and using Raoults Law, replace m1 through-
out Eq.9 yielding
q = -1.86p (m1 , m2 ) = -1.86p (Rm2 , m2 ),
B + B2 (11)
= (R + 1)m2 + (B1R 2 + B2 )m22 + 1 Rm22 .
2
Fig. 4 Comparison of measured freezing point depression for the ternary mixture ethylene glycol, sodium
chloride, and water. The solid points are data from [106] measured using differential scanning calorimetry, the
solid line is the phenomenological model in Fig.3 fit to the data, the other lines represent models (8) (Additive
Model) and (9) (Quadratic Virial, where Ci=0, and Cubic Virial). This figure is modified and redrawn from [106].
For further examples and analysis of these comparisons, see [41]
Therefore at any given temperature Eq.9 and m1 = Rm2 yields

a quadratic function in m2 with solution
m1 = Rm2 ,
1 + 4B2q + R 2 (1 + 4B1q ) + 2R(1 + B1q + B2q ) - 1 - R (12)
m2 = .
B2 (2 + R) + B1R(1 + 2R)
2.3.2 Chemical Potential Osmolality models facilitate the prediction of the melting tempera-
ture and the likelihood that water will crystalize. It is useful to
derive the concentrations or molalities of the constituent solutes.
However, water and solute transport is driven by chemical poten-
tial gradients. In the case of water transport, note that
m w = m w0 - RT p , where w0 is the chemical potential of pure water
at standard temperature and pressure. Thus osmolality is sufficient
for modeling water transport. For solute transport, however, other
models must be used. The most common approximation for chem-
ical potential is that ms (ms ) RT ln ms , but one may arrive at a
more accurate form by starting with the same virial energy used
to derive model (9), and differentiating with respect to the moles
of solute [see45, for details]. Using the same formalism and mixing
rules defined for the osmotic virial Eq.9, the chemical potential of
the ith solute as a function of m = (m1 , , mn )T is
n
mi (m) = RT ln mi + yi* + (Bi + B j )m j , (13)
j =1
94 James D. Benson
where yi* is a function of temperature and pressure, and Bi are

defined above.
2.4 Membrane Membrane transport models vary widely in complication, but most
Transport Models reduce to the following premise: the rate of flux per unit area of
membrane is a function of the difference in chemical potentials
across the membrane. For passive transport, this premise comes
from the combination of the Reynolds transport theorem, the
argument that cell membranes are relatively thin with respect to
the operating diffusion lengths, and an appropriate choice of con-
stitutive diffusion flux laws (c.f. [46]). The particular proportional-
ity function (linear, quadratic, exponential, etc.) is related to the
underlying constitutive law chosen for the model, and most appli-
cations adopt Ficks law, which is linear, i.e. the mass flux
J x = a (mxe - mxi ), where a is some constant of proportionality. Note
that this holds for both water and permeating solutes, and in this
case, an n-solute and water system can be written as
dW
= Pw A (mwe - mwi ) = -L p ART (p e - p i ),
dt
dS1
= Pw A (mse1 - msi1 ), (14)
dt

dSn
= Pw A (msen - msin ),
dt
where A is cellular surface area, and Px and x are the permeability
coefficients and chemical potentials, respectively, for each species
x that may depend on the local quantities of the other species, Lp is
the hydraulic conductivity, and is the osmolality. The chemical
potential is then written as a function of either the mole fraction x,
concentration c, or molality m of each of the species being mod-
eled, e.g. mw = mw (mw , ms1 , ms 2 , , msn ) as in Subheading2.3. This
in conjunction with auxiliary equations defining water concentra-
tions at the membrane yields a closed system of equations. We note
that it is standard to assume that the cellular surface area is fixed,
even while total cell volume changes.
There are other potential models that purport to be free of the
shortcomings of the linear, Ficks law based, model (14). One such
model is proposed by Elmoazzen etal.[47], where
dW
= Pw A sinh(mwe - mwi ),
dt
dS1
= Pw A sinh(mse1 - msi1 ),
dt (15)

dSn
= Pw A sinh(msen - msin ).
dt
Note that as mx - mx 0 , sinh(mx - mx ) (mx - mx ) , and

e i e i e i
Benson [45] demonstrated that the local behavior of these systems

at rest points is identical.
An alternative approach is one based on the irreversible ther-
modynamics construct of Onsager by which the Kedem and
Katchalsky formalism is derived [48]. In this case, the model
assumes that fluxes are linearly proportional to forces. In short, this
model is very similar to Model (14), except that the interaction of
solutes is accounted for using the parameter . However, Kleinhans
published a thorough comparison and analysis of the Kedem
Katchalsky and a simplified form of Model (14) [49], where he
showed that the computational differences between the two mod-
els were slight under typical cryobiological conditions. He then
argued that the introduction of the third parameter introduces
more uncertainty than the precision it might contribute, as its
physical interpretation is unclear except in the most direct experi-
mental designs, echoing comments by Finkelstein [50].
2.4.1 Chemical Potential The 2p model (14), with one permeating and one non-perme-
Approximations ating solute, is the most widely used model describing water
transport during freezing of cells. This model, for example,
would be appropriate in the case of a cell placed in media con-
taining one permeating CPA and non-permeating solutes (e.g.,
Phosphate Buffered Saline). Let ms and mn be the molality (see
Note 4) of permeating and non-permeating solutes, respec-
tively. The system is then further simplified by assuming that
chemical potential differences increase linearly with molality,
that is m we - m wi = -RT p w -RT (mse + mne ) + RT (msi + mni ) and
mse - msi c se - c si , for concentrations cs.
In fact, in our application the chemical potential of the
permeating solute always appears as a difference across the mem-

brane. In this case note that a better approximation using lowest

order terms can be obtained by truncating the equation for
thechemical potential of the nth permeating species (13) to
msen - msin ln msen - ln msin = ln(msen msin ) . This is in effect the same
as assuming that the virial coefficients Bi0 for i=1 n, which
notably imply that the freezing point depression is linear in molal-
ity. Elliott etal. show that most solutes are quadratic or cubic in
molality, therefore the linear model is likely only valid for small
molalities (i.e., in the dilute region where ms2n 1). This being
said, however, if msen msin (e.g., the cell is near equilibrium), then
the expansion of lnx in terms of its power series in 1 x gives

(-1)n +1 e
ln(msen msin ) = (msn msin - 1)n . (16)
n =1 n
96 James D. Benson
This power series has unit radius of convergence corresponding

to the interval 0 < msen msin 2 . Now, taking only the first term
yields
ln msen - ln msin (msen msin - 1),
mse - msi (17)
= n i n.
msn

The error using this approximation after one term is bounded
1 e
by (msn msin - 1)2 , yielding an estimate of when the approxima-
2
tion will be valid.
This approximation, however, is less than ideal as it contains
msin in the numerator and is functionally undefined when msin = 0,
a very common cryobiological initial condition. For example, dur-
ing CPA equilibration protocols, msin (t ) = 0 at time t=0. Therefore,
a different approximation must be used. Let msavg n
= (msin + msen ) 2
and apply the approximation from Eq.17 to arrive at
1
ln msen - ln msin (msen msavg - 1) - (msin msavg - 1) = (msen - msin ). (18)
n n
msavg
n

As above, the error from truncating the power series after the
1
first term is bounded by
2
(
(msen msavg
n
- 1)2 + (msin msavg
n
- 1)2 . )
To illustrate the error from this approximation, Fig.5 shows
ln msen - ln msin and its approximation by Eq.18 in two forms. First
is the pointwise error at any given r, shown by the solid line. The
dashed line shows the error assuming that msavg n
= 5 which would
be relevant in the case where, for example, the cell initially has no
intracellular permeating solutes (msin (0) = 0 and the extracellular
molality of permeating solutes is 10mol/kg (msen 10). Finally,
note, however, that msn in fact contains additional terms (using the
virial expansion in Eq.13), therefore this approach approximates
the approximation!
Additionally, note that in this formulation, there is an implicit
concentration dependence in the solute permeability term where
Ps (mse , msi ) = Ps RT m avg unless msavg is fixed at 1 (see Fig.5 and
caption for discussion). In this case, system (14) becomes
dW
= Pw A (mwe - mwi ),
dt
= L p ART (p i - p e ),
(19)
-L p ART (mse + mne - msi - mni ),
dS
= Ps A (mse - msi ) P s A (mse - msi )
dt
Fig. 5 Percent error of the approximation of chemical potential differences given in Eq.18 as a function of the
ratio of intra- and extracellular molalities. Specifically, with r = msi mse , the solid line shows plot of
100 1-r 2(r - 1)
% error (r ) = ln(1 r ) - = 100 1 + in both panels. The
ln(1 r ) 1 2+r 2 (1 + r )(- ln r )
dashed lines show specific selections for ms avg . In particular large dashes indicate msavg = 5 , which one
would expect to be appropriate for the equilibration of a cell with 10mol/kg CPA, and small dashes indicate
msavg=1, which incidentally is the case where ln mse - ln msi mse - msi . In this case, errors are bounded
above by 60% when mes > msi , (middle plot). In the bottom plot, however note that that errors are consider-
ably worse for msavg=1 when msi > mse . Importantly, though, the error during equilibration in these cases will
decrease very rapidly due to the rapid efflux of water causing mes msi . Finally, this error is going to be
linearly proportional to the error in flux of permeating solute at any given concentration due to Eq.14
98 James D. Benson
where Lp is the hydraulic conductivity. Taking mavg=1, and noting

that msi = S (rwW ) and mni = N (rwW ) where w is the density
of water, and N the moles of intracellular non-permeating solute,
Eq.19 can be rewritten as
dW N +S
= -L p ART rw-1 rwmse + rwmne - ,
dt W
dS S
= Ps A rw-1 rwmse - , (20)
dt W
dN
= 0.
dt
This can be coupled with an initial condition for W(0)=W0
and S(0)=S0 and the Boyle vant Hoff equation to arrive at
N (0) = W0mni = N yielding the closed system of ODEs:
dW N +S
= -L p ART rw-1 rwmse + rwmne - ,
dt W
(21)
dS S
-1
= Ps A r rwmse -
w .
dt W
This system is known as the 2p model in cryobiological
literature.
Finally, to recover total cell volume, we use Eq.1 with the solu-
tion of system (21). It is notationally convenient to define the solu-
tion of (21) as the vector (see Note 5) X(t)=(W(t)S(t))T and then
define the vector G = (1v s )T so that the cell volume is
V (t ) = G X (t ) + Vb .
Finally, consider the jth permeating solute again. Suppose that
the mi for i j are small in the sense that they could be considered
dilute if in a binary solution and suppose that mii mie 1. Then
we may assume that for i j the (B j + Bi )mi terms of Eq.13 are
negligible compared to 2B j m j , and use the nearly full approxima-
tion of the chemical potential (13) as follows:
n n
m je - m ij = ln m ej + (B j + Bi )mie - ln m ij - (B j + Bi )mii ,
i =1 i =1
n
= ln m m + (B j + Bi )(m - mii ),
e
j
i
j
e
i
i =1
(22)
m ej - m ij e i
+ 2B j (m - m ), j j
m ave
j
1 e
= 2B j + ave i
(m j - m j ).
mj

This expression will retain improved accuracy over Eq.18.
2.4.2 Temperature There is a well-known temperature dependence of the hydraulic

Dependence conductivity and solute permeability, Lp and Pw. Because Lp and Ps
are derived from diffusion models, it is reasonable and standard to
assume that these parameters follow the Arrhenius model:
E
P (T ) = P0 exp - a (T -1 - T0-1 ) , (23)
R
where P=Lp or Ps and P0 indicates a value at temperature T0. In the
range of super-zero temperatures (e.g., 037C) utilization of this
model has been carried out in a very wide range of cell types [5,
5159]. There are some criticisms of this model, however, includ-
ing that there are other larger temperature dependent causes for
changing parameter values [6062], including membrane phase
transitions [63]. Another criticism comes from Katkov [37], who
claims that the temperature dependence of Lp is correct but that
the temperature dependence of Ps should be modeled using Ps= R T
where is a solute mobility term that follows the Arrhenius
model. In our view, this argument is based on adapting the Kedem
and Katchalsky formalism and derivation to the 2p model, when in
fact their derivations are fundamentally different. The Ps term is, in
fact, a lumped parameter that includes diffusivity, solute mobil-
ity, partition coefficients, and even concentration, each with its
own temperature dependence. Therefore, the precise model of
temperature dependence of Ps is difficult to predict. Unfortunately
there are only a very few studies that examine the Ps in a sufficient
temperature range to suggest that one model is superior to another
(the differential scanning calorimetry studies by Devireddy
etal.[6062] may be an exception to this, however, they use the
KedemKatchalsky model, and avoid some aspects of this ques-
tion). It is of interest to note that the model proposed by Katkov
follows the form of the modified Arrhenius model [64], which
takes the general form
n
T E
P (T ) = P0 exp - a (T -1 - T0-1 ) . (24)
T
0 R

2.4.3 Nondimensional It is nearly always advantageous to nondimensionalize mathemati-

Model cal models. This allows examination of the relative sizes of terms,
and shows the dependence of behaviors on lumped parameters.
For the 2p model (21), this was first proposed in the cryobio-
logical literature by Katkov [65], and subsequently extended by
Benson [39] used by Benson etal. [8, 9, 39], Lusianti etal. [66],
and Davidson etal. [7] among others. The nondimensionalization is
achieved as follows. First, let W=wwo, S = smo wo , rwme = memo ,
b = Ps (L p ARTm0 )-1 , and t = t *t := w0 (L p ARTmo rw )-1t where is
our new unitless time-variable, t is a characteristic time scale of the
100 James D. Benson
system, and subscript o is a value at a specific quantitytypically

an isosmotic value, e.g. wo = wiso , mo = miso . Then
dt dt = L p ART rw-1wo-1mo leaving
dw 1+ s
= mse + mse - ,
dt w (25)
ds s
= -b mse - .
dt w
In this nondimensional version of the 2p model, there are

two parameters, t and b. Note that two cells with identical b but
different t will trace out the same solutions in water and solute
vs.-time. This is very useful for comparing behavior between cells
as we note that the critical surface area to volume ratio appears only
in the t term. Therefore, cells with the same membrane permeabil-
ity characteristics but with different isosmotic volumes will yield
identical plots.
Temperature Dependence Note that the temperature dependence of the parameter b also
follows the Arrhenius model if Lp and Ps do. To wit, using Eq.23
for both Lp and Ps we get that b(T) also can be modeled with
L
Eq.23 with E ab = E a p - E aPs and b0 = Ps 0 (L p0 ARTm0 )-1 :
dw 1+ s
= m se(T ,t ) + m se(T ,t ) - ,
dt w
ds s
= -b (T ) m se(T ,t ) - (26)
dt w
dT
= g(t ),
dt
where g() is the cooling rate and the temperature dependence of
b(T) is defined by Eq.23 with P0 = b0 defined above.
Virial Expansion Models We may use the same nondimensional variables for the virial
expansion, as long as we change the values of the virial coefficients
appropriately. For example, with p (m1 , m2 ) = m1 + m2 + B1m12 + B2m22 ,
we use mim0 = mi and Bimo = Bi , for i=1,2 to get
p (m1 , m2 ) = p (m1m0 , m2m0 ) = mo (m1 + m2 + B1m12 + B2m22 ) and
everything else will scale as in Eq.25.
2.4.4 Reparametrization While system (25) is easily solved using standard numerical
forStiff Solutions integration techniques, there are many advantages to analytical
andAnalytic Solution solutions of differential equations. For example, in the case of sys-
tem (21), note that when w approaches 0as is the case during
some CPA equilibration protocols and also during slow cooling
system (25) becomes stiff (see, e.g., Chap.21 of [67]). Numerical
solvers for stiff ODEs give up speed and accuracy. Benson etal.[68]
showed that for constant mes and mne , a new time variable ,
rescaled by setting the differentials d =wd , allows the factoring
out of a 1w term from the right-hand side of both equations in
system (25) to arrive at
dw
= (m se + m ne )w - s - 1,
dq (27)
ds
= b (m se w - s ).
dq
This linear second order differential equation is easily solved
using standard techniques (see, e.g., [69]). To recover the original
unitless time , one must integrate the differential:
q
t = w(x ) dx . (28)
0
In the usual suprazero cryobiological case where ms e and
e
m 0 are constant, System(27) may be solved analytically as fol-
n
lows. First, note with the vector x=(w,s)T, System(27) is of the
form x = Ax + e1 where
m e + m e 1 (29)
A = s en ,
-bms b

and e1 = (1 , 0)T is the first unit basis vector. Then define a new vec-
tor y = x + A -1e1 (see Benson [45] for a proof that A is invertible
when mne 0 ). This yields
dy
= Ay , (30)
dq
and the solution can be written in terms of the matrix exponent
y=exp(A ) (see, e.g., Chap.7.8 of [69]) or as a function of the
fundamental matrix solution defined by the eigenvalues and eigen-
vectors of A (see, e.g., Chap.7 of [69]). Finally, x may be recovered
by subtracting A1e1.
There are two critically important applications of this solution
technique. First is that one may solve, analytically, for the time and
volume at which the cell reaches its maximal or minimal volume
during a CPA equilibration protocol. This allows one to calculate
whether a particular protocol for a particular cell type will cause a
cell to exceed its osmotic tolerance limits without having to numer-
ically solve a differential equation. The analytical solution for rela-
tive extrema under CPA equilibration protocols is provided in
dimensional variables (i.e., for system (21)) in Benson etal. [68].
This is a refinement over previous work by Katkov [70] and Zhang
102 James D. Benson
and Chen [71] who found a time-free form of the extremal volume
as a function of initial conditions.
Additionally, the analytic solution greatly facilitates numerical
optimization of CPA equilibration protocols. Optimization of
multistep protocols where step length and extracellular permeating
and non-permeating solute concentration at each step are control
variables requires an extremely efficient numerical solution of sys-
tem (25). This becomes challenging when optimal equilibration
protocols drive water volumes to zero as in the time-optimal con-
trols defined by Benson etal. [39], dwelling in the stiff region of
the phase-space. The exact solution allows optimization of an easily
differentiable function. For an example of this approach, see
Lusianti etal. [66] and Davidson etal. [7].
This solution technique works even in the cooling regime if
the extracellular concentrations are only temperature dependent
(not temperature and time dependent) turning system (26) into
dw
= (m se (T ) + m ne (T ))w - s - 1,
dq
ds
= -b (T ) (m se (T )w - s ) (31)
dq
dT
= g (q )w.
dq
e e
Under most conditions mn (T ) and mn (T ) will be nonlinear,
however, so even with constant g(), a closed-form analytic solu-
tion is unlikely to be found, though other techniques exist due to
the linear nature of the ODE.Nevertheless, the avoidance of divid-
ing by the small w term that is nearly always encountered during
equilibrium cooling protocols may make this worth the effort.
2.4.5 Effects To demonstrate the effects of this assumption Fig.6 shows the vol-
oftheSelection ume versus time plot for a hypothetical cell (modeled after a human
ofChemical Potential oocyte) exposed to 1.2mol/kg propylene glycol in 290mOsm
Approximation saline solution using model (25), and model (14) with the osmotic
virial expansion for osmolality (Eq.9) and the two more accurate
approximations of chemical potential differences (Eqs.13 and 22).
To illustrate the differences, the permeability parameter is fixed at
b=1.62 [7]. Note, however, that in practice, one would choose a
model and fit for the permeability appropriate for that model. This
was performed by Elmoazzen etal.[72] who found concentration
dependent differences in fit Ps as a function of model selection.
2.5 Ice Formation While the modeling of extracellular ice has a long and active his-
Models tory in and beyond the cryobiological literature, there are essen-
tially two models of intracellular ice formation used to predict the
likelihood that a particular modeled cryopreservation protocol will
cause the formation of potentially lethal intracellular ice. The first
Fig. 6 Plot of water volume as a function of time after exposure to 1.2 molal propylene glycol for three models
using three different approximations of chemical potential differences across the membrane with a fixed per-
meability coefficient b. The solid gray line corresponds to the system (25), the solid black line to system (14)
with the quadratic osmotic virial Eq.9 and solute chemical potential approximation (22), and the dashed black
line to system (14) with the quadratic osmotic virial Eq.9 and solute chemical potential approximation (13).
Osmotic virial coefficients are from [107], and the nondimensional permeability b=1.62 was used by Benson
etal. [8] and Davidson etal. [7] to model propylene glycol permeability in human oocytes
model was defined in Mazurs seminal paper [1]. Mazurs method

for prediction of intracellular ice is based on an experimentally
observed and modeling-justified statement that the likelihood of
intracellular ice increases dramatically when the cellular solution is
supercooled to two Kelvin below its melting temperature.
More modern approaches to formulate models to predict
intracellular ice formation have been proposed by Toner etal.[73]
and Karlsson etal.[74, 75]. In particular they use the hypothesis
that the growth rate of ice crystals is limited by the diffusivity of
water during cooling. Combined with temperature and viscosity
dependent stochastic models of ice nucleation that include the
likelihood of nucleation within the cell, on the cell membrane, and
outside of the cell, the model provides repeatable predictions of
the likelihood that a cell will undergo intracellular ice formation
given a particular cooling protocol. Because of the complexity this
model is not provided here.
3 Optimization
3.1 CPA Equilibration Cryoprotective agents are necessary for the successful
Protocols cryopreservation of cells in suspension. These CPAs work in con-
centrations usually exceeding 1mol/kg. Therefore, as discussed
104 James D. Benson
100
85% survival
80
% survival
60
40
20
1/8 1/4 1/2 1 2 4 8

Osmolality (x isosmolal)
Fig. 7 Plot of survival of a hypothetical cell as a function of extracellular non-

permeating osmolality with 85% survival line shown with its corresponding
extracellular osmolalities. In particular, here the lower and upper osmolalities
associated with 85% survival are 1/2 isosmolal and 2 isosmolal, respec-
tively (e.g., around 150 and 600mOsm, respectively). These osmolalities then
are used in conjunction with the Boyle vant Hoff plot (see Fig.1 or Eq.4) to
determine the volumes associated with the osmolalities corresponding to 85%
cell survival, these volume limits are known as osmotic tolerance limits
above, the abrupt exposure of cells to these high concentrations

may cause damage due to excessive water volume flux (see Fig.6).
The equilibration of cells with and from high concentrations of
cryoprotectant agents is an important part of nearly all cryopreser-
vation procedures. While the biophysics of the equilibration pro-
cesses is dependent on many parameters including cell type, CPA
type, temperature, among others, the protocol used to achieve
equilibration can have a dramatically damaging effect on cell viabil-
ity, even before any cooling has occurred. This damage is under-
stood to be dominated by both physical and biochemical effects
[8]. The former is linked to transport driven volume fluxes causing
cells to exceed volume limits, and the latter due to cytotoxicity,
whether acute or accumulated, of CPAs.
In the cryobiological literature, osmotic tolerance limits are
total cell volume limits within which the cell can shrink or swell
with minimal damage. These limits have been explored in a wide
variety of cell types [5, 20, 22, 25, 29, 7689]. These limits are
assumed to be hard limits, even though they are usually defined by
a fixed decrease in population viability; e.g., the relative volume
limits that allow, say, 85% of the population to survive (see Fig.7)
There may be confounding effects, such as temperature, concen-
tration, or duration away from isosmotic volumes, but these have
yet to be definitively explored.
Fig. 8 Prototypical volume response to CPA addition protocols. The dashed line
indicates the volume response in a single-step protocol that may cause the cell
to exceed a lower volume limit, shown in the gray dot-dash line. A two-step
protocol may be used to avoid excessive volume fluxes, though will expose cells
to high concentrations of solutes for longer times
Osmotic tolerance limits are critical to optimization of CPA

equilibration protocols. Figure8 demonstrates the classic problem.
Suppose one wishes to equilibrate cells with a given concentration
C of CPA.Depending on the parameters of model (21), the cell
volume response to the abrupt exposure to this concentration C
may drive exosmosis of water and its associated cell volume loss
beyond the lower osmotic tolerance limit. A two-step protocol
where cells are first equilibrated with the concentration C2 for a
length of time, and then with the concentration C, may cause the
cell volume to remain within the osmotic tolerance limit, and mini-
mal volume related damage is expected.
Mathematically, these osmotic tolerance limits can be written
as V low V (t ) V up where V (t) is the time dependent total volume
of the cell. However, because the osmotically inactive volume Vb
from Eq. 4 does not change, it may be subtracted, leaving
V low - Vb W (t ) + v s S (t ) = G X (t ) V up - Vb using the vector nota-
tion from above. In terms of the nondimensional variables, this
expression is equivalent to k* G x (t ) k * where x(t) is the non-
dimensional form of the state vector X(t), is the n ondimensional
form of the vector of relative partial volumes, and k and k are the
nondimensional forms of the lower and upper osmotic tolerance
limits, respectively.
Competing with the volume flux induced damage due to
exceeding the osmotic tolerance limits is the time, temperature, and
concentration dependence of the accumulated damage of exposure
to CPA solutions. The use of permeating CPAs has facilitated suc-
cessful cryopreservation because, in part, they mitigate multimolal
salt solutions that would be encountered in CPA free cooling pro-
tocols, but there is still an inherent toxicity due to exposure to CPAs
and this toxicity is concentration and time dependent [9092].
106 James D. Benson
The simplest approach to minimize this toxicity is to attempt

to determine a minimal-time equilibration protocol. However, if
accumulated cell damage is also concentration dependent then in
order to determine an optimal protocol, there must be a way to
quantify the cumulative effects of the cell concentration. These
effects are succinctly summarized with the definition of a toxicity
cost functional (see Note 6), first defined in terms of time only in
Benson [93] and then, more generally, in terms of a concentration
dependent power law in Benson etal.[8].
The most general form of a toxicity cost functional would be
tf
(32)
J (m e ) = f (m (t ), t ) dt ,
i
0
where mi and me are the vectors of intra- and extracellular m olalities,
respectively, and tf is the time at which the cell reaches a desired
intracellular state (e.g., ms i (t f ) = msdes = 10 mol/kg)note that
this tf requirement may be strict in the sense that exact control-
lability of the system is desired (e.g.,msi (t f ) = msdes ), or it may be
expressed in terms of a tolerance (e.g.| msi (t f ) - msdes | tol where
tol is an acceptable tolerance. The cost J represents the accumu-
lated damage to the cell as a function of equilibration protocol.
While there may be a very complicated functional relationship of
instantaneous damage, Benson etal.[8] use existing studies of
time and concentration dependent toxicities to propose the model
tf
(m (t ))
a
J (m ) = e i
s dt , (33)
0
i
where m is the intracellular molality of the permeating solute and
s
is a constant. This superseded the time-optimal model proposed
by Benson [93] and Karlsson [94] where =0, and includes the
toxicity cost functional defined by Benson etal. [8], who cited
existing studies to support =1.6.
While there was overlap between two studies to support this
model, there is much need for further exploration of appropriate
toxicity cost functionals. For example, one might expect a depen-
dence on non-permeating solute molality as well, that could be
included in the cost functional:
tf
(m (t )) + e (mni (t )) dt .
a b
e
J (m ) = i
s
(34)
0
Or one could hypothesize that damage is also a function of
integrated distance away from isosmolality. In this case the cost
functional might be
tf
(m (t ))
a
+ e (V iso - V (t )) dt .
2
e
J (m ) = i
s
(35)
0
Regardless of the choice of specific cost functional, the

c ombination of the osmotic tolerance limits with the toxicity cost
functional allow the definition of the state-constrained optimal
control problem:
Find the optimal time dependent choice of m e A to minimize
thecost functional J(me) subject to the mass transport Eq.14 or its
variants and state constraints defined by osmotic tolerance limits
V low - Vb G X (t ) V up - Vb , where A is the set of admissible
control functions [95].
In its most general case, the set A may be all measurable
functions (c.f. [96]) that drive the cellular state to reach a desired
value [95]note that this class of functions includes the usual
smooth, and piecewise constant functions one may imagine, but
also chattering functions that vary infinitely often in an infinitesi-
mal length of timea less-than-desirable function class for imple-
mentation in the real world. In more restrictive cases, one might
expect that A contains functions that are bounded, or that are
piecewise linear or constant. The theory of optimal control works
best with the most general, but bounded, A , but restrictions to
more physically relevant functions are possible.
This is the approach adopted by Benson etal. [8, 9, 39, 93],
Lusianti etal. [66], and Davidson etal. [7], where they use the
dilute reparametrized nondimensional 2p model with the cost
functional defined in Eq.33. In particular, noticing that Eq.33 is
equivalent to
tf a
S (t )
0 W (t ) dt , (36)
e
J (m ) =

and adopting the nondimensionalization and reparametrization
scheme from Subheading2.4.4, the cost functional becomes
qf a
s (t )
J (m e ) = 0 w(t ) w(t ) dt , (37)

where f is the final time in the new time variable, and the dynamics
are governed by Eq.27.
In particular, the minimal time protocol defined by =0 will
be associated with the cost,
qf
(38)
J (m e ) = w(t ) dt ,
0
and the toxicity cost functional defined by Benson etal. [8] will be
qf
(39)
J (m e ) = s (t ) w(t )-0.6 dt .
1.6
0
108 James D. Benson
Classical optimal control theory may be applied if the set of

admissible controls is allowed to be general [95] (i.e., the extracel-
lular concentrations as a function of time are not restricted to, say,
piecewise constant functions) and this approach was used by
Benson etal. [9, 39, 93], where the theory of geometrical optimal
control [97] was used to define intracellular-state-dependent con-
trol functions to achieve time-optimal control. This approach has
several distinct advantages. First, it prescribes a feedback control
for the cell, where if the state of the cell is known, one may pre-
scribe the optimal control at that instant. Second, it yields insight
into general schemes of optimization and optimal control in these
cases. For instance, Benson etal.show that the time-optimal CPA
equilibration protocol is that which causes the cell to remain at its
lower osmotic tolerance limit for as long as possible while increas-
ing or decreasing extracellular concentrations [9]. While not
exactly a theorem, it can be conjectured that this rule-of-thumb
can be extended to admissible sets with more restrictions, such as
piecewise linear or piecewise constant functions. This is natural due
to the cost functional containing only w(t), the normalized water
volume. Naturally, if w(t) is minimized throughout the protocol
through the control of extracellular solute concentrations, this
integral will also be minimized, regardless of the admissible func-
tion set. In fact, this was borne out in work with human red blood
cells by Lusianti etal. [66] where minimal deglycerolization time
approaches were achieved when cells remained at lower water
volumes.
Classical and geometric optimal control theory has been only
used in preliminary results [98] to analyze the cases where >0,
but these cases have been investigated numerically, first in Benson
etal. [8] and in Davidson etal. [7, 13]. From the geometrical per-
spective, the opposite holds true for the =1.6 case, namely, that
the toxicity optimal protocol is that which drives the cell to its
upper osmotic tolerance limit for as long as possible while increas-
ing or decreasing extracellular concentrations. Again, this is natural
due to the cost function containing w(t) in the denominatorif
w(t) is maximized throughout the protocol through the control of
extracellular solute concentrations, the cost functional will be
minimized.
A final note on this approach is that the optimal addition and
removal approaches for both time-optimal and toxicity-optimal
(=0 and =1.6, respectively) are such that for CPA addition pro-
tocols, mn e 0, and for CPA removal protocols mse 0. This maxi-
mizes the ds dt term throughout the protocol.
3.1.1 Classical CPA Classically, optimal CPA equilibration strategies have been multi-
Equilibration Optimization step (piecewise constant in the context of the above section) pro-
Approach tocols where the choice of step concentrations and durations have
typically been driven by minimizing the number of steps while
keeping cells within osmotic tolerance limits (see, e.g., Fig.8). As

an example, suppose that we wish to equilibrate cells to a desired
final intracellular CPA molality msdes . In order to determine the
optimal first step, Eq.14 or its variants are solved for the CPA con-
centration ms1e that causes the cell volume to meet but not exceed
its lower or upper osmotic tolerance limit. This may be accom-
plished numerically, by solving the system of differential equa-
tions(21) and using a computational package to optimize the
choice of mse that yields the appropriate minimal or maximal vol-
ume, or analytically using either the method discussed by Katkov
or Zhang and Chen [70, 71] or the method of Benson etal. [68].
An analytical solution will certainly be faster and more accurate,
but these advantages are of little consequence in terms of modern
computing power unless this step is part of a much larger optimiza-
tion problem (see, e.g., Lusianti etal. [66]).
To continue to implement the method for a CPA addition pro-
tocol, for example, if ms1e > ms des , assign ms1e = msdes and the opti-
mal protocol will have only a single step. If not, and ms1e has been
determined, the process may be repeated, assuming that the initial
condition is characterized by a cell equilibrated with CPA of con-
centration ms1e , and the concentration ms e2 will be determined so
that it causes the cell volume to meet but not exceed the lower
osmotic tolerance limit. As with the first step, if ms e2 > ms des , assign
ms e2 = ms des and the optimal protocol will have two steps. This pro-
cess may be repeated as needed until msne > ms des , in which case the
optimal protocol will have n-steps. For examples of this approach,
see Gilmore etal.[6], who examine single-step CPA addition and
removal protocols for human spermatozoa to show that a one-step
CPA addition protocol will not cause excessive shrinking if the
CPA ethylene glycol is used, but will do so if glycerol is the CPA,
Agca etal.[5] who examine the relative effects of multistep
addition and dilution protocols for mouse spermatozoa, or Mullen
etal.[12] who look at multistep CPA removal protocols for human
oocytes in this fashion.
3.2 Cooling Rate The most widely accepted theory of damage during slow cooling
protocols where ice is allowed to nucleate, sequester water, and
concentrate the remaining solutes is called the two-factor hypoth-
esis proposed by Mazur, Liebo, and Chu [99]. The hypothesis
was an attempt to explain why damage occurred during sufficiently
slow cooling protocols where the cytoplasm could concentrate
enough to avoid intracellular ice formation. In particular, injury
due to slow-cooling is the accumulation of the so-called deleteri-
ous solution effects. The premise is that even at subzero tem-
peratures, exposures to extremely high concentrations of solutes
causes irreparable damage to cells. This damage mechanism pro-
vides a first rule for cooling rate optimization:
Cool as quickly as possible to avoid solution effects injury.
110 James D. Benson
Fig. 9 Cooling rate as a function of survival defined by competing effects for a

generalized cell. The so-called solution effects occur during slow cooling regimes
and damage due to intracellular ice formation occurs during too-fast cooling
regimes. Summing survival produces the inverted U shaped survival curve. The
scale on the x-axis is cell and CPA dependent. Redrawn from Muldrew etal.[108]
Unfortunately, under too-fast cooling protocols, the intracel-

lular water may not exit fast enough and the intracellular freezing
point depression will not be enough to prevent intracellular ice
formation. Therefore, a second rule of cooling rate optimization is:
Cool slowly enough to avoid intracellular formation.
The combination of these rules creates a two-factor
echanism of damage and produces the theoretical inverted U
m
shaped survival curve shown in Fig.9 that is borne out in experi-
ment (see Fig.10).
Therefore, there is a critical need to understand the intracel-
lular state as a function of temperature, and to couple this intracel-
lular state with a model of intracellular ice formationeither
directly from the phase diagram, or via other models to be dis-
cussed below.
3.2.1 Mazur Model Recall that the Mazur model of ice formation during cooling is the
hypothesis that cells are likely to be ice-free if there is less than two
degrees of intracellular supercooling. Karlsson shows that this is
likely to be dramatically wrong for certain combinations of con-
centrations and temperatures [75]. Yet the Mazur model yields
success even in recent literature [21, 22]. Perhaps this is in part
due to underestimation of the intracellular melting temperature.
To wit, typically the melting temperature of the intracellular
milieux is modeled using the ternary phase diagram of NaCl-CPA-
Water or KCl-CPA-Water (see Subheading2.3). This approach
Mouse
embryos
60
RBC
50
Mouse Sperm
Survival (%)
40
Human Sperm
30
20
10
0
0.1 1 10 100 1000 104
Cooling Rate C min
Fig. 10 Cooling rate vs survival in a variety of species. Note that these cooling
rates are in the slow cooling regime, as the amount of CPA used in these
experiments was not high enough to suppress ice nucleation and/or significant
crystal growth. Redrawn and modified from [109]
overlooks large quantities of intracellular proteins and other

structures that contribute (nonideally) to melting point depres-
sion. Whether it is as precise as other options or not, the Mazur
model is straightforward to implement and will at least provide
order-of-magnitude or ballpark approximations of optimal cool-
ing rates. With these caveats in place, note that there are a variety
of approaches one can use to optimize cooling rates under this
assumption.
Constant Cooling Rate This first approach yields optimal constant cooling rates (defining
Approach temperature as a linear function of time). Typically this means that
cells will be cooled in a controlled rate freezer to a specific subzero
temperature and then removed and immediately plunged into liq-
uid nitrogen. The critical observation here is that the amount of
intracellular supercooling before reaching the plunge temperature
increases monotonically with increasing cooling rates and with
decreasing temperatures. In particular, due to the temperature
dependence of membrane water permeability (see Eq.23) if cells
are cooled at a fixed rate their ability to lose water to keep up
with extracellular ice-formation decreases with temperature.
The decision about what temperature or intracellular state to
achieve before plunging must be made before optimizing cooling
rate. This temperature or state is that which causes negligible crys-
tallization during plunging into liquid nitrogen. The rule of thumb
adopted by Liu etal. [4], Kashuba Benson etal. [20], and others
is that plunging when the intracellular CPA is at a critical
112 James D. Benson
Fig. 11 Intracellular supercooling as a function of (constant) cooling rate for three

different goal concentrations. Maximal (optimal) cooling rates are those that just
reach 2C supercooling at the desired intracellular plunge concentration. Any
larger cooling rate would cause more than 2C supercooling. Thus, the optimal
cooling rate is given by the intersection of the curves generated and the
Supercooling=2C line, and is indicated by the arrows for each goal concentra-
tion. Data and model are from [20]
concentration of 40% (w/w) should be sufficiently safe, a state

usually occurring between 40 and 80C.In theory this critical
concentration percentage should be dependent on CPA and the
temperature at which this occurs, but note that cooling rates
through the most dangerous temperatures with regard to ice crys-
tal growth with its attendant damage are improved in this fashion
as the Leidenfrost effect is minimized due to much lower pre-
plunge temperatures, and that the thermal conductivity of ice
which now makes up a large majority of the system volume, is
nearly four times that of water.
In this case, optimization of cooling rates is simply a matter of
simulating cooling at increasing rates until the maximal supercool-
ing exceeds 2C.This is shown in Fig.11 where the intracellular
supercooling at three possible critical concentration CPA percent-
ages (40%, 45%, and 50%) is given as a function of constant cool-
ing rate for mouse embryonic stem cells loaded with 1mol/kg
DMSO.This figure demonstrates the monotonicity of supercool-
ing as a function of goal concentration. To generate this figure,
model (21) with temperature dependent permeability given by
the Arrhenius law in Eq. 23 as well as the simple ODE:
dT dt = -b with initial condition T(0)=T0melt, the initial melt-
ing point of the solution, are coupled with a experimental phase
diagram model (5) or synthetic phase diagram model (9) to deter-
mine the extracellular osmolality and concentration as a function
of temperature (e.g., as in Eq.12). In practice this optimization

takes very little computational time, though one could choose
a rapidly converging numerical optimization scheme if this was
an issue.
Woelders Approach The second and very elegant approach developed by Woelders and
Chaviero [3] uses a Raoults law approximation of freezing point
depression as a function of osmolality: -q 1.86p . This is then
solved for the osmolality such that q = T + p where T is the intra-
cellular temperature and p is the degree of intracellular supercool-
ing allowed. In particular, the difference of osmolalities across the
membrane with the intracellular space fixed at p degrees of super-
cooling will be p1.86. Using this in Eq.14 and assuming dSdt=0
yields
dW
= -L p (T )ARTp 1.86. (40)
dt
Next, using the Boyle vant Hoff relationship, the intracellu-
larwater may be expressed directly in terms of osmolality:
W = N p , which is combined with p = -(q + p) 1.86 to get
W = -1.86N (q + p). Differentiating this expression with respect
to T noting q = T - 273.15 yields
dW 1.86N
= . (41)
dT (q + p)2
Finally, using the chain rule and Eqs.40 and 41,

dT dW dt
-CR opt = = ,
dt dW dT
-L p (T )ARTp 1.86
= , (42)
1.86N
.
(q + p)2
L p (T )ARTp (q + p)2
=- .
1.862 N
Note that this elegant approach may be generalized to systems
with permeating solutes (e.g., see Woelders and Chaviero for the
case with one permeating solute [3]). The challenge here is that
nonlinear cooling rates are difficult to achieve. Typically controlled
rate freezers achieve linear cooling rate protocols, however there
are some controlled rate freezers that are more flexible in this
regard [100]. Nevertheless, with enough piecewise temperature
versus time intervals over a number of time steps, a reasonable
approximation to any cooling profile may be made, and modeling
can be performed to ensure that these approximations do not cause
excessive intracellular supercooling (in other words, one may use
linear interpolation to approximate any thermal profile).
114 James D. Benson
PIF Models Note that the same two optimization approaches above apply for
the more modern probability of ice formation models. Namely,
one can decide on a maximal acceptable likelihood of intracellular
ice formation and solve for a nonlinear (in time) temperature pro-
file. Or, one may prescribe a linear cooling protocol, and then
observe that the likelihood of intracellular ice formation is still a
monotonically increasing function with cooling rate. See Karlsson
etal. [101] for a complete example of this approach with mouse
oocytes. Alternatively, an excellent application of a variant of this
approach is given by Liu etal. [4] in which an interrupted cooling
protocol was developed.
3.3 Warming Because any ice crystals formed during cooling will grow during
warming, it is generally accepted that one should maximize warm-
ing rates unless warming rates will cause fracturing and other
stresses due to differential thermal expansion in the sample [102].
In fact recent work has shown that samples cooled in a suboptimal
method may be rescued by sufficiently high warming rates [14].
Maximal warming rates for straws and most sample containers are
achieved in a circulating water bath, which provides nearly an order
of magnitude faster warming rates than warming in air [103].
4 Conclusions
This chapter demonstrates the many facets of mathematical model-

ing of single cell cryopreservation. Considerations here include the
appropriate choice of transmembrane flux model, chemical poten-
tial model, ice formation model, and others. The models consid-
ered here are only a subset of a larger system; heat and mass
transport does not exist only on the cellular level, and as such one
cannot in general ignore the effects of spatial gradients of heat and
concentration. In fact, there is a vast body of literature on the
effects of unstirred-layers and solute-polarization on membrane
mass transport (see, e.g., [104]), but these effects have largely been
ignored in the cryobiological community. This may be because
unstirred layers are often modeled in the literature as additional
permeable membranes in series, yielding a lumped permeability
parameter that includes the unstirred layer. The difficulty here is
that membrane permeability measurements and the cryopreserva-
tion of cells (in sample tubes) are often performed in very different
environments (e.g., turbulent versus still), that may generate very
different unstirred layer thicknesses. There is also differential heat
transport from the outside of a sample container compared to the
inside of a sample container, even in relatively slow, quasi-
equilibrium cooling protocols. This may generate differential sur-
vival in cells that are particularly sensitive to cooling rates. There
are challenges to modeling CPA equilibration protocols that
involve extremely high concentrations of cryoprotectants, as the

viscosity affects diffusivity, advection, and momentum equations.
While the foundations of these models are also applicable in tis-
sues, the heat and mass transport models must be adapted to
account for the spatial gradients and the inherent complex geom-
etry and structure of the tissue.
Mathematical modeling provides cryobiologists a powerful
tool to approach general cryobiological problems, facilitating the
development of cryopreservation strategies for cells and tissues
with scientific and clinical utility. Cryobiology is also an exciting
area for applied mathematicians as it provides a rich source of inter-
esting and challenging clinically, biologically, and financially rele-
vant problems that are based on classical physical models, yet
require a delicate balance of specificity and utility. The challenges
of cryobiological modeling not only encompass analytic, computa-
tional, and foundational modeling problems but also generate real-
world and relevant optimization and optimal control problems
that require novel analytical tools and careful mathematical
approaches to ensure that their utility is preserved.
5 Notes
1. The majority of mathematical modeling in the cryobiology

literature is concerned with cooling rates, leaving warming
rates relatively unexplored, perhaps because warming rates are
typically an order of magnitude or more greater than the asso-
ciated cooling rates.
2. Here we follow the approach and notation outlined in Benson
[38].
3. A table of these parameters is available in [105].
4. Dissociated if a salt
5. Where T indicates the transpose of the row vector
6. The use of the mathematical term functional is more precise
than function and appropriate here as it is a function that maps
to the real numbers. Often the term functional is used when
the argument is a function on which a global operation such as
integration is performed.
References
1. Mazur P (1963) Kinetics of water loss from 3. Woelders H, Chaveiro A (2004) Theoretical
cells at subzero temperatures and the likeli- prediction of optimal freezing programmes.
hood of intracellular freezing. J Gen Physiol Cryobiology 49:258271
47:347369 4. Liu J, Woods E, Agca Y, Critser E, Critser J
2. Mazur P, Leibo S, Chu E (1972) A two-factor (2000) Cryobiology of rat embryos II: a the-
hypothesis of freezing injury. Evidence from oretical model for the development of inter-
Chinese hamster tissue-culture cells. Exp Cell rupted slow freezing procedures. Biol Reprod
Res 71:345355 63:13031312
116 James D. Benson
5. Agca Y, Gilmore J, Byers M, Woods EJ, Liu J, and growth during both vitrification and slow
Critser JK (2002) Osmotic characteristics of freezing. J Appl Phys 105:114701
mouse spermatozoa in the presence of extend- 17. Chang A, Dantzig JA, Darr BT, Hubel A
ers and sugars. Biol Reprod 67:14931501 (2007) Modeling the interaction of biological
6. Gilmore J, Liu J, Gao D, Critser J (1997) cells with a solidifying interface. J Comput
Determination of optimal cryoprotectants Phys 226:18081829
and procedures for their addition and removal 18. Liu Z, Wan R, Muldrew K, Sawchuk S,
from human spermatozoa. Hum Reprod 12: Rewcastle J (2004) A level set variational for-
112118 mulation for coupled phase change/mass
7. Davidson AF, Benson JD, Higgins AZ (2014) transfer problems: application to freezing of
Mathematically optimized cryoprotectant equil- biological systems. Finite Elem Anal Des
ibration procedures for cryopreservation of 40:16411663
human oocytes. Theor Biol Med Model 11:13 19. Zeng C, He L, Peng W, Ding L, Tang K,
8. Benson JD, Kearsley AJ, Higgins AZ (2012) Fang D, Zhang Y (2014) Selection of optimal
Mathematical optimization of procedures for reference genes for quantitative RT-PCR
cryoprotectant equilibration using a toxicity studies of boar spermatozoa cryopreserva-
cost function. Cryobiology 64:144151 tion. Cryobiology 68:113121. http://
9. Benson JD, Chicone CC, Critser JK (2012) dx.doi.org/10.1016/j.cryobiol.2014.01.004
Analytical optimal controls for the state con- 20. Kashuba Benson CM, Benson JD, Critser JK
strained addition and removal of cryoprotec- (2008) An improved cryopreservation
tive agents. Bull Math Biol 74:15161530 method for a mouse embryonic stem cell line.
10. Levin RL (1982) A generalized method Cryobiology 56:120130
forthe minimization of cellular osmotic 21. Kashuba CM, Benson JD, Critser JK (2014)
stresses and strains during the introduction Rationally optimized cryopreservation of
and removal of permeable cryoprotectants. multiple mouse embryonic stem cell lines:
J Biomech Eng 104:8186 IImathematical prediction and experimen-
11. Songsasen N, Leibo SP (1997) tal validation of optimal cryopreservation
Cryopreservation of mouse spermatozoa. II. protocols. Cryobiology 68:176184. http://
Relationship between survival after cryo- dx.doi.org/10.1016/j.cryobiol.2013.12.003
preservation and osmotic tolerance of sper- 22. Kashuba CM, Benson JD, Critser JK (2014)
matozoa from three strains of mice. Rationally optimized cryopreservation of
Cryobiology 35:255269 multiple mouse embryonic stem cell lines:
12. Mullen SF, Li M, Li Y, Chen ZJ, Critser JK Icomparative fundamental cryobiology of
(2008) Human oocyte vitrification: the per- multiple mouse embryonic stem cell lines and
meability of metaphase II oocytes to water the implications for embryonic stem cell cryo-
and ethylene glycol and the appliance toward preservation protocols. Cryobiology 68:
vitrification. Fertil Steril 89:18121825 166175. http://dx.doi.org/10.1016/j.
13. Karlsson JO, Szurek EA, Higgins AZ, Lee cryobiol.2013.12.007
SR, Eroglu A (2014) Optimization of cryo- 23. Agca Y, Liu J, Critser E, Critser J (2000)
protectant loading into murine and human Fundamental cryobiology of rat immature
oocytes. Cryobiology 68:1828. http://dx. and mature oocytes: hydraulic conductivity in
doi.org/10.1016/j.cryobiol.2013.11.002 the presence of Me(2)SO, Me(2)SO permea-
14. Seki S, Jin B, Mazur P (2014) Extreme rapid bility, and their activation energies. J Exp
warming yields high functional survivals of Zool 286:523533
vitrified 8-cell mouse embryos even when sus- 24. Lacelle PL, Rothstein A (1966) The passive
pended in a half-strength vitrification solution permeability of the red blood cell to cations.
and cooled at moderate rates to 196 C. J Gen Physiol 50:171188
Cryobiology 68:7178 25. Agca Y, Liu J, Mullen S, Johnson-Ward J,
15. Karlsson JOM, Cravalho EG, Rinkes IHMB, Gould K, Chan A, Critser J (2005)
Tompkins RG, Yarmush ML, Toner M (1993) Chimpanzee (pan troglodytes) spermatozoa
Nucleation and growth of ice crystals inside osmotic tolerance and cryoprotectant perme-
cultured-hepatocytes during freezing in the ability characteristics. J Androl 26:470477
presence of dimethyl-sulfoxide. Biophys J 26. Newton H, Pegg DE, Barrass R, Gosden RG
65:25242536 (1999) Osmotically inactive volume, hydrau-
16. Yang G, Zhang A, Xu LX, He X (2009) lic conductivity, and permeability to dimethyl
Modeling the cell-type dependence of sulphoxide of human mature oocytes.
diffusion- limited intracellular ice nucleation J Reprod Fertil 117:2733
27. Gao D, Chang Q, Liu C, Farris K, Harvey K, 40. Moore WJ (1972) Physical chemistry, 4th
McGann L, English D, Jansen J, Critser J edn. Prentice-Hall, New Jersey
(1998) Fundamental cryobiology of human 41. Prickett RC, Elliott JAW, McGann LE (2011)
hematopoietic progenitor cells i: osmotic Application of the multisolute osmotic virial
characteristics and volume distribution. equation to solutions containing electrolytes.
Cryobiology 36:4048. doi:10.1006/cryo. J Phys Chem B 115:1453114543
1997.2060 42. Elliott JAW, Prickett R, Elmoazzen H, Porter
28. Woods EJ, Zieger MA, Lakey JR, Liu J, K, McGann L (2007) A multisolute osmotic
Critser JK (1997) Osmotic characteristics of virial equation for solutions of interest in biol-
isolated human and canine pancreatic islets. ogy. J Phys Chem B 111:17751785
Cryobiology 35:106113. doi:10.1006/ 43. Landau L, Lifshitz E (1980) Statistical phys-
cryo.1997.2029 ics. Course of theoretical physics, vol 5, 3rd
29. Willoughby CE, Mazur P, Peter AT, Critser edn. Pergamon Press, Oxford
JK (1996) Osmotic tolerance limits and prop- 44. Kleinhans F, Mazur P (2007) Comparison of
erties of murine spermatozoa. Biol Reprod actual vs. synthesized ternary phase diagrams
55:715727 for solutes of cryobiological interest.
30. Du J, Tao J, Kleinhans FW, Peter AT, Critser Cryobiology 54:212222
JK (1994) Determination of boar spermato- 45. Benson JD (2011) Stability analysis of several
zoa water volume and osmotic response. non-dilute multiple solute transport equa-
Theriogenology 42:11831191 tions. J Math Chem 49:859869
31. Du J, Tao J, Kleinhans FW, Mazur P, Critser 46. Bird RB, Stewart WE, Lightfoot EN (2002)
JK (1994) Water volume and osmotic behav- Transport phenomena, 2nd edn. Wiley,
iour of mouse spermatozoa determined by NewYork
electron paramagnetic resonance. J Reprod
Fertil 101:3742 47. Elliott J, Elmoazzen H, McGann L (2000) A
method whereby Onsager coefficients may be
32. Benson C, Liu C, Gao D, Critser E, Critser J evaluated. J Chem Phys 113:65736578
(1993) Determination of the osmotic charac-
teristics of hamster pancreatic islets and iso- 48. Kedem O, Katchalsky A (1958)
lated pancreatic islet cells. Cell Transplant 2: Thermodynamic analysis of the permeability
461465 of biological membranes to non-electrolytes.
Biochim Biophys Acta 27:229246
33. Mazur P, Schneider U (1986) Osmotic
responses of preimplantation mouse and 49. Kleinhans F (1998) Membrane permeability
bovine embryos and their cryobiological modeling: Kedem-Katchalsky vs a two-
implications. Cell Biophys 8:259285 parameter formalism. Cryobiology 37:
271289
34. Shapiro H (1948) The change in osmotically
inactive fraction produced by cell activation. 50. Finkelstein A (1987) Water movement
J Gen Physiol 32:3451 through lipid bilayers, pores, and plasma
membranes: theory and reality. Wiley,
35. Prickett RC, Elliott JAW, Hakda S, McGann NewYork
LE (2008) A non-ideal replacement for the
Boyle vant Hoff equation. Cryobiology 51. Ebertz S, McGann L (2002) Osmotic param-
57:130136 eters of cells from a bioengineered human
corneal equivalent and consequences for
36. Katkov II (2011) On proper linearization, cryopreservation. Cryobiology 45:109117
construction and analysis of the Boylevant
Hoff plots and correct calculation of the 52. Mazur P, Koshimoto C (2002) Is intracellular
osmotically inactive volume. Cryobiology ice formation the cause of death of mouse
62:232241 sperm frozen at high cooling rates? Biol
Reprod 66:14851490
37. Katkov II (2008) Challenge from the simple:
some caveats in linearization of the Boyle- 53. Fedorow C, McGann L, Korbutt G, Rayat G,
vant Hoff and Arrhenius plots. Cryobiology Rajotte R, Lakey J (2001) Osmotic and cryo-
57:142149 protectant permeation characteristics of islet
cells isolated from the newborn pig pancreas.
38. Benson JD (2012) Some comments on recent Cell Transplant 10:651659
discussion of the Boyle vant Hoff relation-
ship. Cryobiology 64:118120 54. Benson CT, Liu C, Gao D, Critser E, Benson
JD, Critser JK (1998) Hydraulic conductivity
39. Benson JD, Chicone CC, Critser JK (2011) A (Lp) and its activation energy (Ea), cryopro-
general model for the dynamics of cell volume, tectant agent permeability (Ps) and its Ea, and
global stability and optimal control. J Math reflection coefficients (sigma) for golden
Biol 63:339359
118 James D. Benson
hamster individual pancreatic islet cell mem- 66. Lusianti RE, Benson JD, Acker JP, Higgins
branes. Cryobiology 37:290299 AZ (2013) Rapid removal of glycerol from
55. Benson C, Liu C, Gao D, Critser E, Benson J, frozen-thawed red blood cells. Biotechnol
Critser J (1998) Hydraulic conductivity (Lp) Prog 69:609620
and its activation energy (Ea), cryoprotectant 67. Kreyszig E (2006) Advanced engineering
agent permeability (Ps) and its Ea, and reflec- mathematics, 9th edn. Wiley, NewYork
tion coefficients (sigma) for golden hamster 68. Benson JD, Chicone CC, Critser JK (2005)
individual pancreatic islet cell membranes. Exact solutions of a two parameter flux model
Cryobiology 37:290299. doi:10.1006/ and cryobiological applications. Cryobiology
cryo.1998.2124 50:308316
56. Liu J, Zieger M, Lakey J, Woods E, Critser J 69. Boyce W, DiPrima R (1992) Elementary dif-
(1997) The determination of membrane per- ferential equations and boundary value prob-
meability coefficients of canine pancreatic islet lems, 6th edn. Wiley, NewYork
cells and their application to islet cryopreser- 70. Katkov I (2002) The point of maximum cell
vation. Cryobiology 35:113 water volume excursion in case of presence of an
57. Benson CT, Liu C, Gao DY, Critser ES, impermeable solute. Cryobiology 44:193203
Benson JD, Critser JK (1996) Water permea- 71. Zhang S, Chen G (2002) Analytical solution
bility (Lp) and its activation energy (Ea), for the extremums of cell water volume and
cryoprotectant permeability (Ps) and its Ea, cell volume using a two-parameter model.
and reflection coefficients (sigma) for golden Cryobiology 44:204209
hamster individual pancreatic islet cell mem-
branes. Cryobiology 33:290--299 72. Elmoazzen HY, Elliott JAW, McGann LE
(2009) Osmotic transport across cell mem-
58. Liu C, Benson C, Gao D, Haag B, Mcgann L, branes in nondilute solutions: a new nondi-
Critser J (1995) Water permeability and its lute solute transport equation. Biophys J
activation-energy for individual hamster 96:25592571
pancreatic-islet cells. Cryobiology 32:493
502. doi:10.1006/cryo.1995.1049 73. Toner M, Cravalho EG, Karel M (1993)
Cellular-response of mouse oocytes to freez-
59. Rule GS, Law P, Kruuv J, Lepock JR (1980) ing stress- prediction of intracellular ice for-
Water permeability of mammalian cells as a mation. J Biomech Eng 115:169174
function of temperature in the presence of
dimethylsulfoxide: correlation with the state 74. Karlsson J, Cravalho E, Toner M (1994) A
of the membrane lipids. J Cell Physiol model of diffusion-limited ice growth inside
103:407416 biological cells during freezing. J Appl Phys
75:44424455
60. Devireddy R, Fahrig B, Godke R, Leibo S
(2004) Subzero water transport characteris- 75. Karlsson JOM (2010) Effects of solution
tics of boar spermatozoa confirm observed composition on the theoretical prediction of
optimal cooling rates. Mol Reprod Dev ice nucleation kinetics and thermodynamics.
67:446457 Cryobiology 60:4351
61. Devireddy R, Smith D, Bischof J (1999) Mass 76. Glazar AI, Mullen SF, Liu J, Benson JD,
transfer during freezing in rat prostate tumor Critser JK, Squires EL, Graham JK (2009)
tissue. (AIChE Journal) 45:639654 Osmotic tolerance limits and membrane per-
meability characteristics of stallion spermato-
62. Devireddy R, Raha D, Bischof J (1998) zoa treated with cholesterol. Cryobiology
Measurement of water transport during freez- 59:201206
ing in cell suspensions using a differential
scanning calorimeter. Cryobiology 77. Yoshimori T, Takamatsu H (2009) 3-D mea-
36:124155 surement of osmotic dehydration of isolated
and adhered PC-3 cells. Cryobiology 58:52
63. Drobnis E, Crowe L, Berger T, Anchordoguy 61. http://dx.doi.org/10.1016/j.
T, Overstreet J, Crowe J (1993) Cold shock cryobiol.2008.10.128
damage is due to lipid phase transitions in cell
membranes: a demonstration using sperm as a 78. Blanco JM, Long JA, Gee G, Donoghue AM,
model. J Exp Zool 265:432437 Wildt DE (2008) Osmotic tolerance of avian
spermatozoa: influence of time, temperature,
64. Mcnaught AD, Wilkinson A (1997) cryoprotectant and membrane ion pump
IUPAC.Compendium of chemical terminol- function on sperm viability. Cryobiology
ogy (the Gold Book), 2nd edn. Wiley 56:814
Blackwell, Oxford. ISBN 978-0865426849
79. Salinas-Flores L, Adams SL, Lim MH (2008)
65. Katkov I (2000) A two-parameter model of Determination of the membrane permeability
cell membrane permeability for multisolute characteristics of pacific oyster, crassostrea
systems. Cryobiology 40:6483
gigas, oocytes and development of optimized 90. Fahy G, Wowk B, Wu J, Paynter S (2004)
methods to add and remove ethylene glycol. Improved vitrification solutions based on the
Cryobiology 56:4352 predictability of vitrification solution toxicity.
80. Si W, Benson J, Men H, Critser J (2006) Cryobiology 48:2235
Osmotic tolerance limits and effects of cryo- 91. Elmoazzen HY, Poovadan A, Law GK, Elliott
protectants on the motility, plasma membrane JAW, McGann LE, Jomha NM (2007)
integrity and acrosomal integrity of rat sperm. Dimethyl sulfoxide toxicity kinetics in intact
Cryobiology 53:336348 articular cartilage. Cell Tissue Bank 8:125133
81. Agca Y, Mullen S, Liu J, Johnson-Ward J, 92. Wang L, Liu J, Zhou GB, Hou YP, Li JJ,
Gould K, Chan A, Critser J (2005) Osmotic Zhu SE (2011) Quantitative investigations on
tolerance and membrane permeability charac- the effects of exposure durations to the com-
teristics of rhesus monkey (macaca mulatta) bined cryoprotective agents on mouse oocyte
spermatozoa. Cryobiology 50:114 vitrification procedures. Biol Reprod 85:
82. Walters EM, Men H, Agca Y, Mullen SF, 884894
Critser ES, Critser JK (2005) Osmotic toler- 93. Benson JD (2009) Mathematical problems
ance of mouse spermatozoa from various from cryobiology. Ph.D. thesis, University of
genetic backgrounds: acrosome integrity, Missouri
membrane integrity, and maintenance of 94. Karlsson JO, Younis AI, Chan AW, Gould
motility. Cryobiology 50:193205 KG, Eroglu A (2009) Permeability of the rhe-
83. Hunt C, Armitage S, Pegg D (2003) sus monkey oocyte membrane to water and
Cryopreservation of umbilical cord blood: 2. common cryoprotectants. Mol Reprod Dev
Tolerance of CD34(+) cells to multimolar 76:321333
dimethyl sulphoxide and the effect of cooling 95. Lee E, Markus L (1968) Foundations of opti-
rate on recovery after freezing and thawing. mal control theory. The SIAM series in
Cryobiology 46:7687 applied mathematics. Wiley, NewYork
84. Guthrie H, Liu J, Critser J (2002) Osmotic 96. Royden HL (1988) Real analysis, 3rd edn.
tolerance limits and effects of cryoprotectants Prentice-Hall, New Jersey
on motility of bovine spermatozoa. Biol 97. Schttler H, Ledzewicz U (2012) Geometric
Reprod 67:18111816 optimal control: theory, methods and exam-
85. Koshimoto C, Mazur P (2002) The effect of ples. Springer, NewYork
the osmolality of sugar-containing media, the 98. Benson JD (2013) Cost functional depen-
type of sugar, and the mass and molar dence of optimal CPA equilibration trajecto-
concentration of sugar on the survival of
ries. Cryobiology 67:404
frozen- thawed mouse sperm. Cryobiology
45:8090 99. Mazur P (1977) The role of intracellular
freezing in the death of cells cooled at supra-
86. Liu J, Christian J, Critser J (2002) Canine optimal rates. Cryobiology 14:251272
RBC osmotic tolerance and membrane per-
meability. Cryobiology 44:258268 100. Morris G, Acton E, Avery S (1999) A novel
approach to sperm cryopreservation. Hum
87. Koshimoto C, Gamliel E, Mazur P (2000) Reprod 14:10131021
Effect of osmolality and oxygen tension on
the survival of mouse sperm frozen to various 101. Karlsson JO, Eroglu A, Toth TL, Cravalho
temperatures in various concentrations of EG, Toner M (1996) Fertilization and devel-
glycerol and raffinose. Cryobiology 41: opment of mouse oocytes cryopreserved
204231 using a theoretically optimized protocol.
Hum Reprod 11:12961305
88. Gao DY, Liu J, Liu C, McGann LE, Watson
PF, Kleinhans FW, Mazur P, Critser ES, 102. Wowk B (2010) Thermodynamic aspects of
Critser JK (1995) Prevention of osmotic vitrification. Cryobiology 60:1122
injury to human spermatozoa during addition 103. Seki S, Mazur P (2009) The dominance of
and removal of glycerol. Hum Reprod warming rate over cooling rate in the survival
10:11091122 of mouse oocytes subjected to a vitrification
89. Gao DY, Ashworth E, Watson PF, Kleinhans procedure. Cryobiology 59:7582
FW, Mazur P, Critser JK (1993) Hyperosmotic 104. Barry P, Diamond J (1984) Eflects of

tolerance of human spermatozoa: separate unstirred layers on membrane phenomena.
effects of glycerol, sodium chloride, and Physiol Rev 64:110
sucrose on spermolysis. Biol Reprod 49: 105. Prickett RC (2010) The application of the

112123 multisolute osmotic virial equation to
120 James D. Benson
cryobiology. Ph.D. thesis, University of 108. Muldrew K, Acker JP, Elliott JA, McGann LE
Alberta, Edmonton, AB (2004) The water to ice transition:
106. Benson JD, Bagchi A, Han X, Critser JK, Woods Implications for living cells. In: Fuller BJ,
EJ (2010) Melting point equations for the Lane N, Benson EE (eds) Life in the frozen
ternary system water/sodium chloride/ethylene state. CRC Press, London
glycol revisited. Cryobiology 61:352356 109. Mazur P, Miller R (1976) Permeability of the
107. Prickett RC, Elliott JAW, McGann LE (2010) human erythrocyte to glycerol in 1 and 2M
Application of the osmotic virial equation in solutions at 0 or 20 degrees C.Cryobiology
cryobiology. Cryobiology 60:3042 13:507522
Chapter 4
The Principles of Freeze-Drying

Gerald D.J. Adams, Isobel Cook, and Kevin R. Ward
Abstract
This chapter provides an up-to-date overview of freeze-drying (lyophilization) with particular relevance to
stabilizing live cells or viruses for industrial applications as vaccines or seed culture. The chapter discusses
the importance of formulation, cycle development, validation, and the need to satisfy pharmaceutical regu-
latory requirements necessary for the commercial exploitation of freeze-dried products.
Key words Freeze-drying, Lyophilization, Lyoprotectants, Secondary drying, Sublimation
1 Introduction
1.1 General Water is essential to life, providing a universal solvent supporting

Overview biochemical activities within cells, which enables metabolisms to
continue and sustains all living processes. Quite simply, in the
absence of water, life as we define it will cease, resulting in a state
of death or dormancy in live cells or inhibiting biochemical activity
in cellular extracts. Water also plays a major role in the degradation
of stored material, providing conditions that potentiate autolysis or
promote the growth of spoilage organisms [1].
In order to stabilize labile products, it is therefore necessary to
immobilize or reduce the water content of stored samples. Vaccines,
other biological materials, and microorganisms can be stabilized by
chilling or freezing. However, maintaining and transporting sam-
ples in the frozen state is costly, whereas freezer breakdown may
result in the complete loss of valuable product [2].
Alternatively, bioproducts can be dried in air using high process-
ing temperatures. Traditional drying typically results in marked
changes in the physical and chemical properties of the product by
high solute concentration or thermal inactivation and is more appro-
priate for dehydrating low-cost products such as foodstuffs. Freeze-
drying combines the benefits of both freezing and drying to provide
a dry, active, shelf-stable, and readily soluble product [3, 4].
vol. 1257, DOI 10.1007/978-1-4939-2193-5_4, Springer Science+Business Media New York 2015
121
122 Gerald D.J. Adams et al.
1.2 Defining Freeze-drying is also referred to as lyophilization. The term

Freeze-Drying lyophilization, which means to make solvent loving, is
less descriptive than the alternative definition freeze-drying.
Several alternative definitions have been used to describe
freeze-drying. Operationally we could define freeze-drying as
a controllable method of dehydrating labile products by vac-
uum desiccation.
Earlier accounts of freeze-drying suggested that ice was only
removed by sublimation and defined this step as primary drying.
The cycle was then described as being extended by secondary dry-
ing or desorption. Although these definitions are applicable to
ideal systems, they incompletely define the process for typical sys-
tems that form an amorphous matrix or glass when cooled [5].
Technically, freeze-drying may be defined as:
1. Cooling of the liquid sample, followed by the conversion of
freezable solution water into ice, crystallization of crystallizable
solutes and the formation of an amorphous matrix comprising
non-crystallizing solutes associated with unfrozen moisture
2. Sublimation of ice under vacuum
3. Evaporation of water from the amorphous matrix
4. Desorption of unfrozen water resident in the apparently dried cake
1.3 History The method can be traced back to prehistoric times and was used
by the Aztecs and Arctic peoples for preserving foodstuffs. Toward
the end of the 1880s, the process was used on a laboratory scale
and the basic principles understood at that time. Practically, the
method remained a laboratory technique until the 1930s when
there was the need to process heat-labile antibiotics and blood
products. At this time, refrigeration and vacuum technologies had
advanced sufficiently to enable production freeze-dryers to be
developed, and since then the process has been used industrially in
both the food and pharmaceutical industries [3, 6].
1.4 Advantages Freeze-drying has a number of advantages over alternative stabilizing

of Freeze-Drying methods. These may be summarized by the following criteria [6]:
1. The need to stabilize materials for storage or distribution.
2. There may be no suitable alternative to freeze-drying due to
material sensitivities.
3. There may be a legal requirement to freeze-dry the product to
satisfy regulatory demands.
4. Freezing will reduce thermal inactivation of the product and
immobilize solution components.
5. Concentration effects such as salting out of proteins, altera-
tions in the distribution of components within the drying and
dried product, and so on may be minimized by freeze-drying.
Principles of Freeze-Drying 123
6. The water content of the dried product can be reduced to low

levels, and in general samples are more shelf stable when dried
to low moisture contents, although overdrying may reduce
shelf stability in sensitive biomaterials.
7. Because the product is normally sealed under vacuum or an
inert gas, oxidative denaturation is reduced.
8. Loss of water equates to a loss of product weight, and this may
be important where transport costs are significant.
9. Sample solubility, shrinkage, unacceptable appearance, or loss
of activity may all be improved when freeze-drying is used
rather than an alternative technique.
10. The need to compete with competitors supplying similar products.
11. The requirement to launch a product on the market while less
costly drying techniques are being developed.
12. The production of intermediate bulk or requirement to
remove solvents such as ethanol.
13. The need to maximize investment in drying plants by freeze-
drying a minor product rather than invest in an alternative and
more costly drying process.
14. The need to separately dry two or more components that
would be incompatible if dispensed together as a solution
within a single container.
1.5 Types of Freeze- Freeze-dried products may be classified as:

Dried Products
1. Nonbiologicals, where the process is used to dehydrate or
concentrate reactive or heat-sensitive chemicals.
2. Nonliving bioproducts. These comprise the major area of
application and include enzymes, hormones, antibiotics, vita-
mins, blood products, antibiotics, inactivated or attenuated
vaccines, polymerase chain reaction (PCR) components,
nanoparticles, and so on. This subgroup includes pharmaceu-
ticals, which may be used diagnostically or therapeutically.
3. Bone and other body tissues for surgical or medical use; foods
where organoleptic properties are important; industrial
bioproducts.
4. Living organisms for vaccine or seed culture use, which must
grow and multiply to produce new progeny after drying and
reconstitution.
5. Miscellaneous, for example, flood-damaged books, museum
artifacts, and so on.
However, freeze-drying is less appropriate for:
1. Oily or nonaqueous solutions where the material has a low
melting temperature
2. Products that form impervious surface skins, thereby preventing

vapor migration from the drying sample during processing
3. Eukaryotic cells that are able to retain viability when frozen
only in the presence of additives, which may be incompatible
with the freeze-drying process
2 The Process of Freeze-Drying
2.1 Description For convenience, the freeze-drying process may be divided into a
of Process number of discrete steps that may be summarized as [7, 8]:
1. For the processing of cell or other bioproducts, a variety of
preparatory processing steps may be required, e.g., vaccine
preparation, extraction, purification, and formulation in a suit-
able medium for freeze-drying.
2. Sample freezing, which reduces thermal denaturation of prod-
uct, immobilizes solution components, and prevents foaming
when the vacuum is applied. Freezing also induces a desired ice
crystal structure within the sample, which facilitates drying.
3. Primary drying (sublimation) where conditions must be main-
tained in the drying chamber to sustain water migration from
the sample ice during drying. During primary drying, the sam-
ple temperature (strictly freeze-drying interface temperature)
must be maintained below the eutectic, glass transition, col-
lapse, or melt temperature as appropriate to minimize sample
damage during drying.
4. A secondary drying stage during which resident moisture
adsorbed to the apparently dry structure is removed by
desorption.
5. Sealing the dried sample in a vacuum or under an inert gas at
the end of the process, both of which exclude the entry of
reactive, destabilizing, atmospheric gases such as oxygen or
carbon dioxide into the dried sample and prevent the ingress
of damp air into the freeze-dried sample. Note that a freeze-
dried product will have a vastly expanded dry surface area and
is therefore particularly sensitive to air denaturation or mois-
ture uptake.
6. The samples are then removed from the freeze-dryer, stored,
and/or distributed for use prior to reconstitution for injec-
tion, application, or regrowth.
2.2 Processing Freeze-drying is a complex process during which drying may pro-
Principles ceed more or less rapidly within individual samples throughout the
process batch, such that parts of the product will be frozen, whereas
other areas are drying or will have dried depending on the nature of
the sample and stage in the cycle. The precise freezing and drying
behavior will be determined by the interrelationship between the
sample and shelf temperature, system pressure, extent of product
dryness, and variations in drying conditions throughout the cycle.
Often regarded as a gentle method of drying materials, freeze-
drying is in reality a potentially damaging process where the individual
process stages should be regarded as a series of interrelated stresses,
each of which can damage sensitive bioproducts. Damage sustained
during one step in the process may be exacerbated at succeeding
stages in the process chain, and even apparently trivial changes in the
process, such as a change in container, may be sufficient to transform
a successful process to one which is unacceptable [8].
Freeze-drying will not reverse damage incurred prior to for-
mulation, and care must be exercised when selecting an appropri-
ate cell type or technique used to culture or purify the cell or its
extracts prior to freeze-drying. The essence of the formulation
exercise should be to minimize freeze-drying damage, loss of via-
bility, or activity. To ensure minimal losses of activity, the sample
may require dilution in a medium containing protective additives,
specifically selected for the product or application. Although fre-
quently described as protectants, these additives may not be
effective at all stages of the process but may protect only during
particular steps in the drying cycle. At other stages, the additive
may fail to protect the active component and indeed may be
incompatible with the process. It is also important to appreciate
that individual stages in the process can result in damage, which
initially remains undetected, becoming evident only when the
dried sample is rehydrated. Particular attention must be applied to
the selection and blending of the additive mixes in the formula-
tion, and the importance of formulation will be discussed at
greater length later [1, 4, 911].
Freeze-dried products should:
1. Be minimally changed by the process
2. Be dry
3. Be active
4. Be shelf stable
5. Be clean and sterile (for pharmaceutical applications)
6. Be ethically acceptable
7. Be pharmaceutically elegant
8. Be readily soluble and simple to reconstitute
9. Have process that should be economically practicable
Products should be formulated to ensure batch product uni-
formity, whereas there may be particular requirements relating to
product use. In this context, vaccines freeze-dried for oral or aero-
sol delivery may require the inclusion of excipients that minimize

damage when the dried product is exposed to moist air [10].
A wide range of containers can be used to freeze-dry vaccines,
microorganisms, and others, including all glass ampoules, rubber-
stoppered vials, double-chambered vials, and prefilled syringes that
hold both dried vaccine and diluent, bifurcated needles.
Alternatively, vaccines can be dried in bulk in stainless steel or plas-
tic trays and the resultant powder tableted, capsulated, sachet filled,
or dispensed into aerosol devices for lung or nasal delivery.
2.3 Sample Freezing Regarded as the first step in the process, the formulated product
must be frozen before evacuating the chamber to induce sublima-
tion [12, 13]. Freezing will:
1. Immobilize the components in the solution and prevent foam-
ing as the vacuum is applied.
2. Reduce thermal inactivation of the dispensed product.
3. Induce a specific ice crystal structure within the frozen mass,
which will facilitate or inhibit vapor migration from the drying
cake. In short, the ice structure formed during freezing will
dictate subsequent freeze-drying behavior and the ultimate
morphology of the dried cake.
Ideally freezing should minimize solute concentration effects
and result in a sample where all the components are spatially
arranged as in the dispensed solution. However, it may not be pos-
sible to achieve this ideal when solutions or suspensions are frozen.
When addressing the freezing of aqueous solutions or suspensions,
there is the need to consider both the solvent (water in the case of
aqueous solutions) and solute(s) in the formulation.
Frequently, the terms cooling and freezing are erroneously
interchanged, and confusion in understanding the process may
occur and may be compounded by failing to distinguish between
shelf or product cooling and freezing. Cooling refers to the reduc-
tion of temperature of the freeze-dryer shelves, the fluid circulat-
ing through the shelves, the vial and tray mass, interior of the
freeze-dryer, and the dispensed solution or suspension. Cooling
does not assume a change in state from liquid to solid and strictly
should be used to describe reducing temperature during the initial
stage of freeze-drying. Freezing refers to the abrupt phase change
when water freezes as ice. Except for very complex biomolecules or
cold-sensitive cells, cooling in the absence of freezing (chilling) is
generally not damaging to biomaterials.
When solutions or suspensions are frozen, they may cool
appreciably below their measured freezing point prior to ice for-
mation, a phenomenon defined as supercooling (undercooling or
subcooling). The extent of supercooling depends on cooling rate,
sample composition and cleanliness, dispensed fill volume, con-
tainer type, method of sample cooling, and so on. Even when a

simple solution is repeatedly cooled or warmed, the onset and
extent of supercooling will vary from cycle to cycle. In the super-
cooled state, while the composition of the solution remains
unchanged, the cooled liquid is thermodynamically unstable and
sensitive to ice formation. As the solution is cooled to lower tem-
peratures, the probability of ice crystallization will correspondingly
increase. For optimized freeze-drying, the intention should be to
induce supercooling in the suspension to encourage uniform cool-
ing and freezing throughout the sample contents [1416].
Sample freezing may be defined as the abrupt conversion of
the suspension into a mixture of ice and solute concentrate.
Freezing is a two-step process during which water initially nucle-
ates, followed by the growth of the ice crystals that pervade the
solute phase resulting in a mixture of ice and solute concentrate.
Under typical processing conditions, ice nucleates heterogeneously
around microscopic particles within the suspension and is encour-
aged by reducing temperature and agitating the supercooled sus-
pension to increase the probability of contact between nucleating
foci and water clusters. Nucleation depends on the number and
physical nature of particulate impurities within the suspension or
solution. Ice is a particularly effective nucleation focus, and cryobi-
ologists may deliberately seed samples with ice to induce nucle-
ation. Other effective ice nucleators include glass shards and
specifically formulated nucleation promoters. Whereas nucleation
aids can be added to experimental systems, deliberate attempts to
add ice inducers to pharmaceutical materials would be at variance
with good pharmaceutical manufacturing practice [17].
In contrast to nucleation, ice growth (proliferation) is encour-
aged by raising the temperature, thereby decreasing the suspension
viscosity. Ice nucleation and proliferation are inhibited at tempera-
tures below the glass transition temperature (Tg), whereas above
the melting temperature (Tm), the suspension or solution will
melt. The consequences and measurements of these parameters are
important elements in the formulation exercise [13, 1720].
To facilitate the sublimation of water vapor from the drying
mass, the ice crystals should be large, wide, and contiguous,
extending from the product base toward its surface, thereby pro-
viding an optimized structure for vapor migration. Crystal struc-
tures commonly observed during freeze-drying when solutions are
frozen in trays or vials include dendritic structuring, where the ice
crystal branches continuously from the nucleating focus and the
spherulite form, and where sub-branching is discouraged because
the solution viscosity is high, or fast rates of cooling are used.
Cooling or freezing rates are defined as slow (suboptimal),
rapid (superoptimal), or optimal as assessed by criteria such as
post-freezing cell survival or biopolymer activity and are ambiguous
unless conditions are more precisely defined. Cooling rates may be

defined in terms of:
1. The rate at which the shelf temperature is cooled per unit time
2. The rate at which the solution cools per unit time
3. The depth of liquid within the vial (in mm) which cools per unit time
2.4 Controlled The use of controlled ice nucleation in freeze-drying cycles is

Nucleation receiving increased attention [21] to reduce the random occur-
rence of the ice nucleation temperature, which affects primary dry-
ing times and product porosity and uniformity. Any method that
allows for a control of the nucleation temperature results in several
degrees of freedom for the freeze-drying process. In principle,
there are three parameters which can directly be controlled: the
nucleation temperature, the isothermal hold time post-nucleation,
and the cooling rate of the shelf post-nucleation.
Methods to achieve controlled ice nucleation include:
1. High-pressure/depressurization technique [21]
2. The ice fog method [22]
3. Ultrasound ice crystallization [23]
4. Electrofreezing [24]
Controlled nucleation may assist in [25]:
1. Reducing the primary drying time/overall drying time
2. Reduction of damage during cooling and freezing
3. Control of product porosity
4. Improved scale-up ability
5. Improved cake appearance, morphology, and homogeneity
6. Reduced reconstitution time
2.5 Shelf- The shelf-cooling rate [26] is the simplest parameter to control,
Cooling Rate and programmed rates of cooling are standard options on research
and production freeze-dryers. Because shelf temperature and prod-
uct responses are not identical, defining shelf-cooling rate will not
fully define product behavior. Although we are concerned with the
cooling rate achievable within each vial, this parameter is less easy
to monitor compared with shelf cooling, and freeze-drying cycles
generally are controlled by programmed shelf cooling rather than
feedback control from the sample. Cooling rates of the product/
cell suspension will vary considerably from vial and throughout the
sample within the vial, and consequently, measuring the tempera-
ture of vial contents at a fixed position will give only an approxima-
tion of the sample temperature variation.
Observing the freezing pattern of a number of vials arranged
on a shelf will demonstrate that while the contents of some vials
will freeze slowly from the vial base, neighboring vials may remain
unfrozen and supercool appreciably before freezing instantly. This
random freezing pattern will reflect differences in ice structure
from vial to vial and translated into different drying geometries
from sample-to-sample vials. In summary, freezing patterns will be
related to:
1. The ice forming potential within each vial
2. The relative position of the vial on the shelf causing exposure
of individual vials to cold or hot spots
3. Edge effects where samples in vials on the periphery of each
shelf will be subjected to heat transmitted through the cham-
ber walls or door
4. The insertion of temperature probe into the sample, which
will induce ice crystallization
5. The evolution of latent heat as samples freeze, which will tend
to warm adjacent containers
6. Variations in container base geometry, which may impede
thermal contact between sample and shelf
The ice and solute crystal structure resulting from sample freeze
has a major impact on subsequent freeze-drying behavior, encour-
aging the sample to dry efficiently or with defects such as melt or
collapse depending on freezing rate used. The preferred ice struc-
ture comprising large contiguous ice crystals is induced by freezing
the sample at a slow rate of ca. 0.21.0 C/min. Slow cooling will
also induce the crystallization of solutes reluctant to crystallize
when faster rates of cooling are used. However, a slow rate of cool-
ing may exacerbate the development of a surface skin, which inhib-
its sublimation efficiency. Slow cooling can also inactivate a
bioproduct by prolonging sample exposure to the solute concen-
trate biomolecules. However, a fast rate of cooling can result in the
formation of numerous, small, randomly orientated ice crystals
embedded in an amorphous solute matrix, which may be difficult to
freeze-dry. Complicating the choice of freezing regimes is the fact
that the optimal cooling rate cannot be sustained where the sample
fill depth exceeds 10 mm. In short, defining cooling rates often
requires a compromise in sample requirements.
2.6 Ice Structure A period of consolidation (defined as the hold time) is necessary at
and Freeze the end of sample cooling to ensure that all the vial contents in the
Consolidation sample batch have frozen adequately, although excessive hold times
will increase the time of sample freeze and impact on the overall
cycle time. It is a fallacy to assume that the ice structure induced
remains unchanged during this consolidation period and an ice
structure comprising a large number of small ice crystals, induced
by rapid cooling, is thermodynamically less stable than an ice struc-
ture comprising fewer, larger crystals. The thermodynamic equilibrium

can be maintained by Ostwald ripening of ice from small to large
crystals, a process termed grain growth. Although ice structure
changes take place randomly from vial to vial, the hold period is a
major factor in ice recrystallization resulting in significant variation
in crystal structure and subsequent sublimation efficiency from
sample to sample the longer the hold period is employed.
As an alternative to increasing the length of the hold time to
encourage ice recrystallization, a more controlled and time-efficient
method of inducing recrystallization is to heat anneal the frozen
sample [27]. Essentially heat annealing is achieved by:
1. Cooling the product to freeze the solvent (usually water) and
any readily crystallizable solutes.
2. Raising the product temperature during the freezing stage to
recrystallize ice from a small to a large ice crystal matrix. (Note:
this warming phase may also crystallize solutes that are reluc-
tant to crystallize upon initial cooling.)
3. Cooling the product to terminal hold temperature prior to
chamber evacuation (although this may not be necessary if
annealing has caused the critical temperature of the formula-
tion to increase above the holding temperature).
Heat annealing (also defined as tempering) is particularly use-
ful to:
1. Convert an ice structure to a crystalline form, which improves
sublimation efficiency.
2. Crystallize solutes that are reluctant to crystallize during
cooling.
3. Provide a more uniform, dry structure throughout the prod-
uct batch.
4. Integrated with rapid cooling, minimize the development of a
surface skin on the sample, thereby facilitating sublimation.
5. Induce a more porous cake structure with improved drying
efficiency to achieve a lower dried sample moisture content,
with improved solubility.
Although heat annealing will increase the length of the freez-
ing stage of the cycle, overall freeze-drying cycle times may be
significantly reduced because of improvements in drying efficiency
resulting from heat annealing.
Care should be exercised when selecting temperatures and
hold times for heat annealing, particularly when defining the upper
temperature for sample warming. Subjecting a labile product, such
as a vaccine, to temperatures above the eutectic temperature will
expose the sample to hypertonic solution concentrates as the sam-
ple partially melts, which can damage sensitive biomolecules.
2.7 Freezing Solute Regardless of the precise freezing pattern, the formation of ice will
Behavior concentrate the remaining solution within the container. As the pro-
portion of ice increases within the mixture, solute concentration will
correspondingly increase. In the case of an aqueous 1 % (w/v) saline
solution, this concentration effect will be considerable, increasing to
approximately 30 % (w/v) just prior to freezing, and damage to bio-
molecules results as a consequence of solute concentration exposure
rather than direct damage by ice crystals. The behavior of the
solute(s) within the solute concentrate depends on the nature, con-
centration, cooling rate, and interactions between individual solutes
present in the medium and forms the basis for experimental review
during a formulation development exercise [2831].
Overall, four patterns of solute response are observed during
freeze-drying:
1. Solute crystallizes readily, regardless of cooling rate or freezing
conditions, to form a mixture of ice and solute crystals (this
behavior is termed eutectic freezing).
2. Solute crystallizes but only when the solution is subjected to a
slow rate of cooling.
3. Solute crystallizes only after the solution has been heat annealed.
4. Solute fails to crystallize regardless of cooling rate or regime
adopted, and solute remains associated with unfrozen water as
a metastable amorphous mass or glass.
For a crystallizing solute, the eutectic point is the lowest tem-
perature in a system in which a residual liquid phase and solid phase
are in equilibrium. Above the eutectic point, ice and solute con-
centrate persist, whereas below the eutectic point, a mixture of ice
and solute crystals is formed. Eutectic temperatures for aqueous
solutions containing crystallizing salts are characteristic for each
solute and are significantly below the freezing point of water (e.g.,
eutectic temperature for sodium chloride (21.4 C)). Exposing
cells or proteins for prolonged periods to a eutectic solution com-
prising hypertonic salt concentrations can cause damage by plas-
molysis or precipitation by salting out [32].
The eutectic zone is the range of temperatures encompassing
all the eutectic temperatures within the system. For a two-part
water/solute system, the eutectic temperature is a discrete, quanti-
fiable temperature in contrast to multisolute systems where a
eutectic zone may be observed that represents a range of tempera-
tures where the minimum eutectic temperature is lower than that
of any individual eutectic temperatures in the medium.
Typical freeze-dried vaccine formulations fail to crystallize
completely when cooled, and a proportion of the solutes in the
sample persist as an amorphous, noncrystalline glass. When exposed
to temperatures above their glass transition (Tg) or collapse tem-
perature (Tc), these samples may warm during sublimation causing

the amorphous mass to soften, so that the freeze-drying progresses
with sample collapse to form a sticky structureless residue within
the vial. Less severe collapse will result in the formation of a
shrunken, distorted, or split cake [19, 3337].
Samples of freeze-dried products freeze-dried using an unsuit-
able process cycle (sample A) compared with an identically formu-
lated sample (sample B) where processing conditions have been
optimized to provide an acceptable product. Both samples are for-
mulated in an identical medium and filled to a depth of 60 mm,
which greatly exceeds the recommended fill depth maximum for
freeze-drying of 10 mm. Although such excessive fill depths are
often required for commercial or marketing reasons, such fills greatly
impede sublimation resulting in prolonged cycles, collapsed prod-
uct, poor sample solubility, reduced product activity, and shelf stabil-
ity. In addition, such samples often exhibit high moisture content
and unacceptable pharmaceutical elegance and display cake fracture
and physical loss of sample from the vial as drying progresses (defined
as ablation and of particular concern when live vaccines or cytotoxic
drugs are freeze-dried). Sample A was freeze-dried using a conven-
tional freezing and drying cycle and resulted in a cycle time in excess
of 10 days with extensive vial breakage, in addition to the unaccept-
able features noted previously. Sample B was dried using a cycle
designed to ensure satisfactory freezing to induce an optimized fro-
zen structure conducive to rapid rates of sublimation. In addition,
the author was able to accelerate drying by adjusting shelf tempera-
ture and chamber pressure so that drying times were reduced to
3 days. Product quality was assured by maintaining sample tempera-
tures sufficiently below collapse (Tc) or glass transition (Tg) tem-
perature, defined during formulation and process development.
Collapsed cakes are not only cosmetically unacceptable but
may be poorly soluble, exhibit reduced activity, or compromised
shelf stability. Collapse may be exacerbated by the formation of a
surface skin, which impedes vapor migration from the drying
structure. To avoid sample collapse, it is necessary to maintain the
sublimation interface below Tg or Tc throughout primary drying
and to include excipients in the formulation, which reduce the
severity of collapse. It is therefore essential to characterize formula-
tions experimentally during the process development program.
Although collapse may cause operational difficulties during freeze-
drying, the induction and maintenance of the amorphous state
may be essential for protecting labile biomolecules during freezing,
drying, and storage [27, 38, 39].
2.8 Freezing Samples may be frozen in a variety of ways depending on opera-

in Practice tional requirements:
1. Samples may be frozen in a freezer or a cooling tunnel prior to
transfer to the freeze-dryer for desiccation. Advantages include
increased annual sample throughput because the freeze-dryer

is used only for drying. Disadvantages include the greater risk
of sample melt or contamination resulting from the need to
transfer samples from the freezer into the drier.
2. Samples may be frozen in the absence of cooling by evacuating
the container and relying on evaporative cooling to freeze the
sample. However, the need to prevent sample foaming when
the dryer is evacuated precludes the widespread use of the
method.
3. Pellet freezing. Strictly this is not a method of freezing but can
be useful when bulk products, including vaccines for subse-
quent powder filling, are processed. The suspension is sprayed
into a cryogenic liquid or onto a cold surface to form frozen
sample droplets, which are then placed into trays or flasks for
freeze-drying. Under these conditions, sublimation rates are
typically very high because the thickness of the dry layer is
restricted only by the pellet radius, and drying proceeds in a
virtually unimpeded manner from each pellet.
4. The most widely used technique is to freeze the samples
directly on the freeze-dryer shelf. Although this method has
the disadvantage that the drier is used for part of the cycle as a
freezer, freezing and drying samples within a single machine
eliminate the need to transfer samples from freezer to drier
and therefore improve sample control, as well as reducing
product vulnerability; this is generally the preferred method
by regulatory bodies.
5. Samples may be frozen using controlled nucleation which
assists in obtaining a more homogeneous ice crystal size,
which can positively impact the success of the freezing stage
and length of the freeze-drying cycle.
3 The Process of Water Removal
3.1 Sublimation Under atmospheric conditions, liquid water is converted into vapor
and Drying by warming, a process defined as evaporation. However the three
states of water ice, liquid, and vapor coexist at the triple point
and illustrate that at subatmospheric pressures, ice can convert
directly to vapor by sublimation. Ice sublimation from a frozen
sample results in an open, porous, dry structure where solutes are
spatially arranged as in the original solution or suspension. In con-
trast to evaporation, where components are concentrated as drying
progresses, sublimation under vacuum minimizes concentration
effects providing a dry product that is active and readily soluble.
Having frozen the solution, the next step is to dry the sample
by subliming ice directly into water vapor. In order to maintain
freeze-drying conditions, it is essential to lower the partial pressure
of water below the triple point (approx. 800 mBar at 0 C) to

ensure the direct conversion of ice into water vapor and prevent
sample melt. Vacuum will (1) reduce the air concentration above
the product and encourage sublimation and (2) ensure that air
leaking into the system is removed.
3.2 Sublimation Rate Decreasing the chamber pressure will increase the rate of sublima-
and Chamber Pressure tion by reducing the gas/vapor concentration above the sample to
Conditions provide minimal resistance to water molecules migrating from the
sample. Reducing system pressure too low will not increase subli-
mation rate, and indeed, contrary to expectations, at very low sys-
tem pressures, the sublimation rate will decrease. This apparent
paradox can be explained by assuming that two separate factors
influence sublimation efficiency:
1. System pressure reduction sufficient to thin the chamber
atmosphere and facilitate vapor migration from the sample.
2. A system pressure containing sufficient gas or vapor molecules
in the chamber to conduct heat energy from the shelf into the
sample. Essentially, under high vacuum conditions, a thermos
flask effect is induced in the chamber, which inhibits heat trans-
fer from the shelf. Under high-pressure (poor vacuum) condi-
tions, heat transfer from the shelf to the sample is gas/vapor
conduction in contrast to high vacuum conditions where heat
transfer by conduction is reduced and product heat is predomi-
nantly by radiation, which is a relatively inefficient mechanism.
3.3 Vapor To sustain freeze-drying, it is necessary to establish a pressure gra-

Differential Pressure dient from a sample (highest pressure), to condenser, finally to
and Drying Efficiency vacuum pump (lowest pressure) so that water migrates from the
sample as drying progresses. Although the temperature of the sam-
ple must be higher than that of the condenser to ensure a net
migration of water from the sample, the system driving force rep-
resents the difference in vapor pressure (VP) rather than the differ-
ence in temperature between sample and condenser and can be
calculated as the difference in VP between the two. For example,
sample at 20 C has a VP = 0.78 torr, and with the condenser at
40 C (equivalent to a VP of 0.097 torr), driving force will be
0.780.097 or 0.683 torr. Little improvement in driving force is
achieved by operating the condenser at 70 C. (VP = 0.002 torr,
providing a VP differential of 0.78 [sample] 0.002 [condenser]
of 0.778 torr.) The example illustrates that greater sublimation
efficiency is derived by increasing sample temperature rather than
reducing condenser temperature, and the selection of suitable
excipients that enable high processing temperatures to be used
during freeze-drying without compromising sample quality plays
an important role in process and cycle development.
3.4 Heat and Mass The essence of the freeze-drying process depends on maintaining a
Transfer critical balance between the conversion of ice into water vapor by
sublimation under vacuum and the removal of that vapor from the
frozen mass [40, 41]. To maintain sublimation, heat energy is
applied to the product to compensate for sublimation cooling.
However, the heat extracted from the drying sample as water vapor
must carefully balance the amount of energy added to the sample.
Unless this equilibrium can be maintained, the product tem-
perature will either decrease, thereby reducing drying efficiency, or
increase, which may compromise product quality by inducing melt
or collapse. This critical balance between sample warming to increase
drying rate and vapor extraction is defined by the heat and mass
transfer equation. In the early stages of sublimation, the equilibrium
is simple to maintain because the dry structure offers minimal resis-
tance to vapor flow. However, as drying progresses and the depth of
the dry layer increases, impedance to vapor flow will also increase
and the sample may warm sufficiently to melt or collapse unless the
process temperature is reduced. One consequence of reducing the
energy input will be to reduce drying rate and prolong cycle times,
but this may be unavoidable if sample quality is to be preserved.
3.5 Cooling The shelves fitted into the freeze-dryer to support sample contain-
and Warming ers may be alternatively cooled to initially freeze the sample or
the Product maintain shelf at a constant temperature throughout the drying
cycle or warmed to provide energy for drying. Basically two sys-
tems may be fitted:
1. An independent cooling coil is embedded in the shelf through
which cold refrigerant is supplied (this system is termed direct
expansion) and a heating element is bonded into or onto the
base of the shelf. Shelf control is maintained by alternately
operating either the heater or cooler. Direct expansion systems
are relatively inexpensive but fail to achieve temperature con-
trol much better than 5 C.
2. For industrial or development activities, where shelf control
to 1 C is necessary to meet good manufacturing practice
(GMP) requirements, a diatherm fluid, which is invariably sili-
cone fluid, is circulated through the shelves, and a separate
refrigerator/heat exchanger maintains the diatherm fluid at a
preset temperature.
The mechanism and the relative quantities of heat entering the
product will depend on:
1. The nature of the product, its fill depth, consistency, and so on
2. The dimensions and geometry of the sample container and
whether the container rests directly on a shelf or is supported
in a tray
3. The freeze-dryer design

4. Chamber vacuum conditions
Product temperature can be maintained by either raising or
reducing shelf temperature or by alternating system pressure which
has the effect of improving or reducing heat transfer efficiency.
Regardless of the precise system incorporated into the freeze-dryer,
shelf temperature conditions may be controlled manually or pro-
grammed using a PC or microprocessor control.
3.6 The Drying For clarity, it is usual to separate the drying cycle into primary dry-
Cycle: Primary Drying ing (the sublimation stage) and secondary drying or desorption.
The first step in the drying cycle is defined as primary drying and
represents the stage where ice, which constitutes between approxi-
mately 70 and 95 % of the sample moisture, is converted into water
vapor. Sublimation is a relatively efficient process although the precise
length of primary drying will vary depending on the sample formula-
tion, cake depth, and so on. During primary drying, the sample dries
at a discrete boundary (the sublimation interface), which recedes
through the sample from surface to base as drying progresses.
3.7 The Sublimation Variously described as the drying front, freeze-drying front, and so
Interface on, macroscopically the sublimation interface can be observed as a
discrete boundary that moves through the frozen sample to form an
increasingly deeper layer of dried sample above the frozen sample.
Heat is conducted from the shelf through the vial base and the fro-
zen sample layer to the sublimation front where ice is converted into
water vapor. Several consequences result from this progressive reces-
sion of the sublimation front through the dry layer, which include:
1. The maintenance of the frozen zone at a low temperature
because of sublimation cooling.
2. An increase in the resistance to vapor migration and a decrease
in sublimation rate as the dry layer increases in thickness.
3. Because the sublimation interface represents a zone represent-
ing maximum change of sample temperature and moisture
content, the interface represents the zone over which struc-
tural softening or collapse is likely to occur.
4. Water migrating from the sublimation front can reabsorb into
the dried material above the sublimation interface.
Because the sublimation interface is the region where freeze-
drying takes place, temperature monitoring of the interface is of
paramount importance for product monitoring. However, because
the sublimation front is constantly moving through the sample,
interface temperature cannot be effectively monitored using tradi-
tional temperature probes. Although the sublimation interface is
defined as a discrete boundary, this is true only for ideal eutectic
formulations, where ice crystals are large, open, and contiguous
with each other. For typical amorphous formulations, such as vac-

cines, the sublimation front is much broader and comprises indi-
vidual ice crystals imbedded in the amorphous phase. Under these
conditions, although ice sublimes within the isolated crystals, the
water vapor must diffuse through the amorphous phase (which is
itself progressively drying) until it can migrate freely from the dry-
ing sample matrix. Under these conditions, sublimation rates are
much lower than those anticipated from data derived using eutectic
model systems. Complicating a precise prediction of sublimation
rate is the fact that fractures in the dry cake between the ice crystals
can improve drying efficiency. All of these factors, including system
impedances caused by the development of a surface skin on the
sample, have to be considered during sample formulation and cycle
development programs. Notwithstanding these complications in
precisely defining primary drying, sublimation is nevertheless a rela-
tively efficient process, and conditions used for primary drying
include the use of shelf temperatures high enough to accelerate sub-
limation without comprising sample quality by inducing collapse or
melt, combined with high system pressures designed to optimize
heat conduction from shelf into product. Removing the product
when sublimation has been judged as complete will provide a vac-
cine which appears dry but which displays a high moisture content
that is invariably too high (e.g., 10 %) to provide long-term storage
stability, and the drying cycle is extended to remove additional
moisture by desorption during secondary drying.
3.8 Secondary In contrast to primary drying, which is a dynamic process associ-

Drying ated with high vapor flow rates, secondary drying is much less effi-
cient with secondary drying times representing approximately
2040 % of the total process time but only removing 510 % of the
total sample moisture. Under secondary drying conditions, the
sample approaches steady-state conditions where moisture is
desorbed or absorbed from or into the sample in response to rela-
tive humidity and shelf temperatures. Desorption is favored by
increasing shelf temperature, using high vacuum conditions in the
chamber, thereby reducing the system vapor pressure or relative
humidity. Conversely, when the shelf temperature is reduced and
the vapor pressure in the system increased by warming the con-
denser, dried samples will reabsorb moisture and exhibit an increase
in moisture content. Although sample collapse during secondary
drying is generally less likely than collapse during primary drying,
it is possible to induce collapse in the dried matrix by exposing the
sample to temperature above its glass transition temperature (Tg).
3.9 Stoppering A freeze-dried product is both hygroscopic and has an enormously

the Product exposed surface area. Consequently, exposing the dried product
to atmosphere will result in reabsorption of damp air into the
product. Both water and air are damaging to a dried sample, caus-
ing degradative changes resulting in poor stability, and it is there-
fore prudent to stopper samples within the freeze-dryer prior to

removal. It may also be necessary to dry stoppers prior to use or
ensure appropriate low-moisture stoppers are used to prevent
damage by moisture ingress from the stopper; the significance of
this will be dependent on the amount and properties of the prod-
uct. Stoppering under a full vacuum provides ideal conditions for
ensuring product stability because reactive atmospheric gases are
reduced to a minimum. However, injecting water into a sample in
a fully evacuated vial can induce foaming, which can be reduced
by backfilling vials with an inert gas, such as nitrogen, before stop-
pering. Maintaining a certain level of vacuum, e.g., 2/3 of atmo-
sphere, can help ensure a good stopper/vial seal is achieved and
prevent foaming during rehydration.
4 Reconstituting the Product
It is often supposed that because freeze-drying only removes water,

then all products will be fully active by rehydrating only with water.
This may not be the case, and freeze-dried products often exhibit
enhanced activity when reconstituted in an isotonic medium, such
as saline, rather than water. Establishing the correct rehydration
method can be as critical as developing a suitable and robust cycle;
this was demonstrated in a recent research project investigating the
freeze-drying and reconstitution success of red blood cells [42].
5 Freeze-Dryer Design
The need to operate the freeze-dryer under low-pressure conditions

to convert ice directly into water vapor (a process termed sublima-
tion) adds to the complexity and cost of dryers because the chamber
holding the sample must withstand the differential pressure from
vacuum to atmosphere. Although a suitable vacuum pump is essen-
tial for initially evacuating the chamber and eliminating air that may
leak into the dryer during operation, vacuum pumps are not capable
of continuously removing water vapor subliming from the sample,
and a refrigerated trap (termed the process condenser) must be
placed between the sample and the pump to condense the moisture
migrating from the drying sample. In reality, it is the condenser that
comprises the pumping force of the system. Process condensers
may be incorporated into the drying chamber (referred as an inter-
nal condenser) or located in a separate chamber between the sample
chamber and pump (external condenser). Each geometry has advan-
tages and disadvantages although either design may be used. Stainless
steel is typically used to fabricate research or production dryers
because this metal can be cleaned by a wide range of sanitizers
including steam. For GMP manufacture, the freeze-dryer is invari-
ably sterilized by pressurized steam, and this adds to the complexity

and expense of the dryer because it has to conform to the require-
ments to operate under subatmospheric and pressure conditions.
Modern freeze-dryers are also fitted with internal stoppering devices
for sealing vials at the end of the cycle, valves, and monitoring devices
for assessing drying efficiency and are typically computer or micro-
processor controlled so that cycles can be reproduced and evaluated
for regulatory purposes. When freeze-drying vaccines, it may be nec-
essary to incorporate protective devices and introduce processing
protocols that ensure both safe operation and prevent product cross-
contamination [43].
6 Sample Damage During Freeze-Drying
6.1 Damaging Damage to a product for freeze-drying may occur:

Factors
1. When the solution is cooled (described as cold or chill shock).
2. During freezing as the ice forms and the unfrozen solute phase
concentrates.
3. During drying, particularly when the sample collapses as dry-
ing progresses.
4. By protein polymerization when high shelf temperature is
used for secondary drying.
5. During drying and storage because of damage by reactive
gases, such as oxygen, and it is important to appreciate that
even in a vacuum, sufficient gas molecules will be present in
the sealed sample to cause inactivation.
6. During storage by free radical damage or Maillard reactions.
7. During reconstitution, particularly if the sample is poorly
soluble.
6.2 Chill Damage Reducing temperature in the absence of ice formation is generally
(Cold Shock) not damaging to biomolecules or live organisms, although sensi-
tive biopolymers may be damaged by cold shock [2].
6.3 Freezing Damage Reducing temperature in the presence of ice formation is the first
major stress imposed on a biomolecule. Direct damage by ice is not
generally damaging except when living cells are frozen as the for-
mation of intracellular or extracellular ice could rupture the cell.
Biomolecules are more likely to be damaged by an increase in sol-
ute concentration as ice forms. Freezing will result in:
1. Ice formation [44]
2. A rise in solute concentration (this effect can be appreciable,
and a 1 % solution of sodium chloride will increase to 30 % by
freeze concentration as ice forms) [45]
3. Changes in solution tonicity [46]

4. Concentration of all solutes, including cells and biomolecules
that are encouraged to aggregate [47, 48]
5. An increase in solute concentration that may result in salting
out of protein molecules [47]
6. Differential crystallization of individual buffer salts resulting
in marked changes in solution pH as the solution freezes [48]
7. Concentration of potentially toxic impurities above a toxic
threshold sufficient for the impurities to become toxic [48]
8. Disruption of sulfur bonds
9. Generation of anaerobic conditions as freezing progresses
7 Factors Effecting Dried Products
7.1 Storage Stability Freeze-dried vaccines should be formulated to minimize storage

decay and should tolerate storage at ambient temperatures for dis-
tribution purposes. However, it is a fallacy to suppose that a freeze-
dried product remains immune to damage during storage, and
factors which damage freeze-dried products include:
1. Temperature. Whereas a freeze-dried product is more shelf
stable than its solution counterpart, freeze-dried materials are
sensitive to thermal decay and will be influenced by storage
temperature [49].
2. Moisture content [5058].
3. Reactive gases [59].
4. Light.
5. Free radical damage [60].
6. Background nuclear radiation.
7. Specific chemical reactions including Maillard reactions [61].
The interrelationship between sample formulation, dried cake
moisture, storage conditions, and glass transition temperature (Tg)
is complex. In general terms, any physical distortion of the dry
cake during storage will often result in a much more rapid loss of
sample activity than predicted using the Arrhenius equation for
reviewing similar samples [20, 33, 36, 37, 6264].
7.2 Influence Attempts to freeze-dry cells in water or a simple salt solution typi-
of Suspending cally result in poor survival. A wide range of protective media has
Medium Composition been developed for preserving freeze-dried vaccines, including
on Survival of Live augmented growth media or sugar solutions. Carbohydrates are
Cells to Freeze-Drying widely used as freeze-drying protectants either individually or in
combination with other solutes. They should be chosen on the
basis of experimentally determining their freeze-drying character-

istics rather than on a pragmatic basis. Monosaccharides, such as
glucose, provide good bioprotection during freezing and freeze-
drying but exhibit low glass transition (Tg) or collapse tempera-
tures (Tc) and dry with collapse when orthodox freeze-drying
cycles are used. Disaccharides are effective freeze-drying protec-
tants, and because they display higher collapse than monosaccha-
rides, they freeze-dry successfully when conventional drying cycles
are used [64]. Reducing sugars may induce damaging Maillard
reactions, thereby compromising stability [65], and for this reason,
nonreducing disaccharides, such as sucrose or trehalose, are pre-
ferred to reducing sugars such as lactose [65]. The addition of salts
to formulations containing sugars will markedly depress Tg or Tc
[1, 3, 19, 27, 35, 63, 66]. Morgan et al. [67] provide a helpful
review article discussing microorganism preservation by various
drying technologies highlighting factors such as growth phase and
growth medium, as well as cell concentration and bacteria type.
Although presenting technical difficulties such as sample col-
lapse, during freeze-drying, the amorphous phase may be an essen-
tial prerequisite for stabilizing biomaterials, such as vaccines and
live cells, by providing an integration of the protective additive and
biomolecule, thereby minimizing damage level during freeze-
drying and drying.
References
1. Adams GDJ (1995) The preservation of inoc- tion, science et technique du froid, Refrigeration
ula. In: Brown MRW, Gilbert P (eds) Science and Technology, Commission C1,
Microbiological quality assurance: a guide Paris, France, pp 2134
towards relevance and reproducibility of inoc- 7. Oetjen G (1999) Industrial freeze-drying for
ula. CRC Press, London, UK, pp 89119 pharmaceutical applications. In: Rey L, May JC
2. Fanget B, Francon A (1996) A varicella vaccine (eds) Freeze-drying/lyophilization of pharma-
stable at 5 degrees C. Dev Biol Stand ceutical and biological products. Marcel
87:167171 Dekker, New York, pp 267335
3. Adams GDJ (1996) Lyophilization of vaccines. 8. Adams GDJ (1995) Freeze-dryingthe inte-
In: Robinson A, Farrar GH, Wiblin CN (eds) grated approach pharmaceutical manufacturing
Vaccine protocols. Humana Press, Totowa, NJ, international published by Sterling Publications
pp 167185 Limited, London, UK, pp 177180
4. Gheorghiu M, Lagranderie M, Balazuc AM 9. Carpenter JF, Izutsu K, Randolph TW
(1996) Stabilisation of BCG vaccines. Dev Biol (1999) Freezing- and drying induced per-
Stand 87:251261 turbations of protein structure and mecha-
5. Bindschaedler C (1999) Lyophilization pro- nisms of protein protection by stabilizing
cess validation. In: Rey L, May JC (eds) Freeze- additives. In: Rey L, May JC (eds) Freeze-
drying/lyophilization of pharmaceutical and drying/lyophilization of pharmaceutical
biological products. Marcel Dekker, New York, and biological products. Marcel Dekker,
pp 373408 New York, pp 123161
6. Mackenzie AP (1985) A current understand- 10. Cammack KA, Adams GDJ (1985) Formulation
ing of the freeze-drying of representative aque- and storage. In: Spiers RE, Griffiths JB (eds)
ous solutions. In: Refrigeration science and Animal cell biotechnology, vol 2. Academic,
technology: fundamentals and applications of London, UK, pp 251288
freeze-drying to biological materials, drugs and 11. Adebayo AA, Sim-Brandenburg JW, Emmel H,
foodstuffs. International institute of refrigera- Olaeye DO, Niedrig M (1998) Stability of
17D Yellow fever virus vaccine using different lized monoclonal antibody. Int J Pharm
stabilisers. Biologicals 26:309316 450:7078
12. Mackenzie AP (1977) The physico-chemical 26. Rowe TWG (1971) Machinery and methods in
basis for the freeze-drying process. In: freeze-drying. Cryobiology 8:153172
Developments in biological standards, vol 36. 27. Adams GDJ (1996) Technologically chal-
Karger, Basel, Switzerland, pp 5157 lengedfreeze drying damage prevention. Med
13. Willemer H (1999) Experimental freeze- Lab World 1996:4344
drying: procedures and equipment. In: Rey L, 28. Grout B, Morris J, McLellan M (1990)
May JC (eds) Freeze-drying/lyophilization of Cryopreservation and the maintenance of cell
pharmaceutical and biological products. Marcel lines. Trends Biotechnol 8:293297
Dekker, New York, pp 79121
29. Mazur P, Leibo SP, Chu CHY (1972) A two
14. Franks F (1990) Freeze-drying: from empiri- factor hypothesis of freezing injury. Evidence
cism to predictability. CryoLetters 11:93110 from Chinese hamster tissue cells. Exp Cell Res
15. Franks F (1989) Improved freeze-drying: an 71:345355
analysis of the basic scientific principles. Process 30. Franks F (1992) Freeze-drying: from empiri-
Biochem 24:38 cism to predictability. The significance of glass
16. Mackenzie AP (1966) Basic principles of transitions. In: Developments in biological
freeze-drying for pharmaceuticals. Bull standards. Karger, Basel, Switzerland, pp 919
Parenter Drug Assoc 26:101129 31. Franks F, Hatley RHM, Mathias SF (1991)
17. Cameron P (1997) Good pharmaceutical Materials science and the production of shelf-
freeze-drying practice. Interpharm Press, stable biologicals. Pharm Technol Int 3:2434
Englewood, CO 32. Franks F (1985) Biophysics and biochemistry
18. Jennings TA (1986) Validation of the lyophili- at low temperatures. Cambridge University
zation process. In: Carleton FJ, Agalloco JP Press, Cambridge, UK, p 37
(eds) Validation of aseptic pharmaceutical pro- 33. Bellows RJ, King CJ (1972) Freeze-drying
cesses. Marcel Dekker, New York, pp 595633 aqueous solutions: maximal allowable operat-
19. Pikal MJ (1999) Mechanisms of protein stabi- ing temperatures. Cryobiology 9:559561
lization during freeze-drying and storage: the
34. Levine H, Slade L (1988) Water as plasticizer:
relative importance of thermo stabilization and
physico-chemical aspects of low moisture poly-
glassy state relaxation dynamics. In: Rey L,
meric systems. Water Sci Rev 5:79185
May JC (eds) Freeze-drying/lyophilization of
pharmaceutical and biological products. Marcel 35. Pikal MJ (1991) Freeze-drying of proteins II:
Dekker, New York, pp 161198 formulation selection. Pharm Technol Int
3:4043
20. Franks F (1989) Improved freeze-drying: an
analysis of the basic scientific principles. Process 36. Adams GDJ, Irons LI (1992) Practical aspects
Biochem 24:23 of formulation: the avoidance of product col-
lapse. Pharm J 249:442443
21. Geidobler R, Winter G (2013) Controlled ice
nucleation in the field of freeze-drying: funda- 37. Adams GDJ, Irons LI (1993) Some implica-
mentals and technology review. Eur J Pharm tions of structural collapse during freeze drying
Biopharm 85:214222 using Erwinia caratovora L-asparaginase as a
22. Patel SM, Bhugra C, Pikal MJ (2009) Reduced model. J Chem Technol Biotechnol 58:7176
pressure ice fog technique for controlled ice 38. Rey LR (1999) Glimpses into the realm of
nucleation during freeze-drying. AAPS freeze-drying: classic issues and new ventures.
PharmSciTech 10:14061411 In: Rey L, May JC (eds) Freeze-drying/lyoph-
23. Passot S, Trla IC, Marin M, Galan M, Morris ilization of pharmaceutical and biological
GJ, Fonseca F (2009) Effect of controlled ice products. Marcel Dekker, New York, pp 130
nucleation on primary drying stage and protein 39. Livesey RG, Rowe TWG (1987) A discussion
recovery in vials cooled in a modified freeze- of the effect of chamber pressure on heat and
dryer. J Biomech Eng 131:074511 (5 pages) mass transfer in freeze-drying. J Parenter Sci
24. Petersen A, Rau G, Glasmacher B (2006) Technol 41:169171
Reduction of primary freeze-drying time by 40. Adams GDJ, Ramsay JR (1996) Optimizing
electric field induced ice nucleus formation. the lyophilization cycle and the consequences
Heat Mass Transfer 42:929938 of collapse on the pharmaceutical acceptability
25. Awotwe-Otoo D, Agarabi C, Read EK, Lute S, of Erwinia L-asparaginase. J Pharm Sci
Brorson KA, Khan MA, Shah RB (2013) 85:13011305
Impact of controlled ice nucleation on process 41. Adams GDJ (1994) Freeze-drying of biohaz-
performance and quality attributes of a lyophi- ardous products. In: Hambleton P, Melling J,
Salusbury TT (eds) Biosafety in industrial bio- Branch (HFA 305). Food and Drug
technology. Blackie Academic and Professional, Administration, Rockville, MD
London, UK, pp 178212 56. Phillips GO, Harrop R, Wedlock DJ, Srbova
42. Ward K, Cowen A, Peacock T (2012) Patent H, Celba V, Drevo M (1981) A study of the
WO/2012/098358 water binding in lyophilised viral vaccine sys-
43. de Rizzo E, Pereira CA, Fang FL, Takata CS, tems. Cryobiology 18:414419
Tenrio EC, Pral MM, Mendes IF, Gallina NM 57. Adams GDJ (1990) Residual moisture and the
(1990) Photo sensitivity and stability of freeze- freeze-dried product. In: Lyophilization technol-
dried and/or reconstituted measles vaccines. ogy handbook. The Center for Professional
Rev Saude Publica 24:5159 Advancement, East Brunswick, NJ, pp 581604
44. Meryman HI, Williams RJ, St J, Douglas M 58. Cook IA, Ward KR (2011) Headspace mois-
(1977) Freezing injury from solution effects ture mapping and the information that can be
and its prevention by natural or artificial cryo- gained about freeze-dried materials and pro-
protection. Cryobiology 14:287302 cesses. PDA J Pharm Sci Technol 65:457467
45. Grout BWW, Morris GJ (1987) The effects of 59. Greiff D, Rightsel WA (1969) Stabilities of
low temperatures on biological systems. freeze-dried suspensions of influenza virus
Edward Arnold, London, UK sealed in vacuum or under different gases. Appl
46. Farrant J (1980) General observations on cell Microbiol 17:830835
preservation. In: Ashwood-Smith M, Farrant J 60. Heckly RJ, Quay J (1983) Adventitious chem-
(eds) Low temperature preservation in medi- istry at reduced water activities: free radicals
cine and biology. Pitman Medical, Tonbridge and polyhydroxy compounds. Cryobiology
Wells, UK, pp 118 20:613624
47. Ashwood-Smith MT, Farrant J (1980) Low 61. Cox CS (1991) Roles of Maillard reactions in
temperature preservation in medicine and biol- disease. HMSO Publications, London, UK
ogy. Pitman Medical, Tunbridge Wells, UK 62. Cowdery S, Frey M, Orlowski S, Gray A (1977)
48. Arakawa T, Carpenter JF, Kita YA, Crowe JH Stability characteristics of freeze-dried human
(1990) The basis for toxicity of certain cryo- live virus vaccines. In: International sympo-
protectants: a hypothesis. Cryobiology sium on freeze-drying of biological products,
27:401415 vol 36. Developments in biological standards.
49. Taylor MJ (1981) The meaning of pH at low Karger, Basel, Switzerland, pp 297303
temperature. Cryobiology 2:231239 63. Nicholson AE (1977) Predicting stability of
50. Greiff D, Rightsel WA (1965) Stabilities of sus- lyophilized products. In: International sympo-
pensions of virus after vacuum sublimation and sium on freeze-drying of biological products,
storage. Cryobiology 3:435443 vol 36. Developments in biological standards.
51. Greiff D, Rightsel WA (1968) Stabilities of Karger, Basel, Switzerland, pp 6975
influenza virus dried to different contents of 64. Tsourouflis S, Funk JM, Karel M (1976) Loss
residual moisture by sublimation in vacuo. of structure in freeze-dried carbohydrate solu-
Appl Microbiol 16:835840 tions: the effect of temperature, moisture con-
52. Greiff D (1971) Protein structure and freeze- tent and composition. J Sci Food Agricult
drying: the effects of residual moisture and 27:509519
gases. Cryobiology 8:145152 65. Patapoff TW, Overcashier D, Hsu C, Nguyen
53. Bellissent-Funel M, Teixera Q (1999) TH, Borchardt RT (1996) Effects of reducing
Structural and dynamic properties of bulk and sugars on the chemical stability of human
confined water additives. In: Rey L, May JC relaxin in the lyophilized state. J Pharm Sci
(eds) Freeze-drying/lyophilization of pharma- 85:873877
ceutical and biological products. Marcel 66. Griffin CW, Cook FC, Mehaffrey MA (1981)
Dekker, New York, pp 5377 Predicting the stability of freeze-dried
54. Beale PT (1983) Water in biological systems. Fusobacterium mortiferum. Proficiency testing
Cryobiology 20:528531 samples by accelerated storage tests.
Cryobiology 18:420425
55. Center for Biologies Evaluation and Research
(1990) Guidelines for the determination of 67. Morgan CA, Herman N, White PA, Vesey G
residual moisture in dried biological products. (2006) Preservation of microorganisms by dry-
Docket No. 89D-0 140 Docket Management ing; a review. J Microbiol Methods 66:183193
Part II
Methods to Study Freezing and Drying Processes

Chapter 5
Use of In Situ Fourier Transform Infrared Spectroscopy

to Study Freezing and Drying of Cells
Willem F. Wolkers and Harritte Oldenhof
Abstract
An infrared spectrum gives information about characteristic molecular vibrations of specific groups in
molecules. Fourier transform infrared spectroscopy can be applied to study lipids and proteins in cells or
tissues. Spectra can be collected during cooling, heating, or dehydration of a sample using a temperature-
controlled sample holder or a sample holder for controlled dehydration. In the current chapter, acquisition
and analysis of infrared spectra during cooling, warming, or dehydration is described. Spectra analysis
involving assessment of specific band positions, areas, or ratios is described. Special emphasis is given on
studying membrane phase behavior and protein denaturation in cells or tissues. In addition, methods are
presented to determine the water-to-ice phase change during freezing, dehydration kinetics, and the glass
transition temperature of amorphous systems.
Key words Dehydration, Freeze-drying, Freezing, Fourier transform infrared spectroscopy (FTIR),
Membrane phase behavior, Protein denaturation
1 Introduction
Fourier transform infrared spectroscopy (FTIR) is emerging as a

powerful technique for characterization and structural analysis of
biological samples including cells and tissues. In situ FTIR studies
during freezing, heating, or dehydration provide information on
conformational and phase changes of endogenous biomolecules,
including membrane lipids and proteins [1, 2]. By following char-
acteristic peaks in the FTIR spectra of cells, thermotropic and lyo-
tropic membrane phase behavior and changes in protein secondary
structure can be studied. FTIR studies are minimally invasive and
do not require labeling. Moreover, samples can be measured in any
physical state including in hydrated, frozen, and dried state.
IR spectroscopy relies on characteristic vibrations of molecu-
lar groups within molecules. The vibrational frequency
of a molecular group vibration depends on the mass of the vibrat-
ing atoms and on intra- and intermolecular interactions.
147
148 Willem F. Wolkers and Harritte Oldenhof
Dehydration or changes in temperature affect the intra- and

intermolecular interactions of biomolecules, which may result in
conformational or phase changes. This in turn affects the amount
of energy needed for a particular molecular vibration to take
place, which is visible as a change in the position and shape of the
absorption band in the IR spectrum.
Disruption of the plasma membrane is one of the primary
causes of freezing and drying injury. FTIR is one of the few tech-
niques that can be used to study changes in membrane phase state
during freezing [3] or drying [4] in real time. Membrane phase
behavior can be determined using the symmetric CH2 stretching
band arising from lipid acyl chains [5]. The water-to-ice phase
transition can be captured using the libration and bending combi-
nation band of water. Changes in protein secondary structure asso-
ciated with heat-induced denaturation can be evaluated using the
protein amide bands [6]. The glass transition temperature of cryo-
preservation and freeze-drying solutions can be determined using
the OH stretching band [7]. All these transitions can be analyzed
simultaneously by analyzing different spectral regions.
FTIR thus provides a powerful tool to study lipids and pro-
teins in cells or tissues during the extreme conditions that samples
are exposed to during cryopreservation or freeze-drying.
2 Materials
2.1 Fourier 1. A Fourier transform infrared (FTIR) spectrometer (e.g., a

Transform Infrared model 100 FTIR; PerkinElmer, Norwalk, CT, USA), equipped
Spectrometer with a narrow band mercury cadmium telluride (MCT) liquid
nitrogen-cooled IR detector (see Note 1).
2. FTIR purge gas generator (Whatman, Clifton, NJ, USA) (see
Note 2).
3. Personal computer with spectra acquisition software for col-
lecting spectra as a function of time (e.g., Timebase from
PerkinElmer) (see Note 3).
4. Software for spectral analysis (e.g., Spectrum procedures
from PerkinElmer) (see Note 3).
2.2 Temperature- 1. Temperature-controlled demountable liquid transmission cell/

Controlled Sample sample holder with Luer lock fittings and a temperature con-
Holder troller for controlled heating of the device (e.g., from Harrick
for Transmission Scientific, Pleasantville, NY, USA). The sample holder should
Spectra Acquisition allow mounting of samples sandwiched between two IR trans-
parent windows (25 mm diameter), while the sample edge
should have an access point for a thermocouple (to accurately
measure the sample temperature) or for touching the sample
with a copper wire to induce ice nucleation (see Note 4).
In situ Infrared Spectroscopy 149
2. Pump system using liquid nitrogen as a coolant, e.g., a Linkam

stage (Linkam Scientific Instruments, Tadworth, Surrey, UK),
connected to a liquid nitrogen dewar (1 L minimum) and the
sample holder described above (see Note 5).
3. Data logging thermometer with type J thermocouple (Fluke,
Everett, WA, USA) and stopwatch.
4. Two CaF2 windows, 25 2 mm (e.g., from Korth Kristalle,
Kiel, Germany), and a Teflon spacer 25 0.25 mm (e.g., from
Harrick Scientific).
2.3 Attenuated Total 1. Attenuated total reflection (ATR) accessory with a diamond/
Reflection Accessory ZnSe crystal, 1 1 mm (e.g., a PerkinElmer Universal ATR
and Setup Sampling accessory).
for Controlling 2. Donut-shaped shallow dish with lid, which can be filled with
the Sample Humidity saturated salt solution and placed such that it surrounds the
sample holder.
3. Saturated salt solutions to create a defined relative humidity
(RH) [8]. The following solutions can be used: water (~94 %
RH) and saturated solutions of NaCl (~75 % RH), MgCl2
(~33 % RH), and LiCl (~13 % RH). Saturated salt solutions
are prepared by adding salt to water under continuous stirring
until the salt no longer dissolves and salt crystals are visible.
4. Thermometer and hygrometer (e.g., from Fluke), for measur-
ing the sample temperature and the relative humidity in the
sample area.
3 Methods
3.1 Setting 1. Turn on the FTIR purge gas generator (pressure of ~50 psi).
Up and Background Make sure all compartment areas of the FTIR system are
Spectrum Acquisition connected to the purge system, and dry air is supplied in the
sample area.
2. Cool the MCT detector with liquid nitrogen (LN2): slowly
add LN2 in small portions using a funnel. The detector is
properly cooled when the energy level of the background
spectrum reaches its maximum and remains stable. Make sure
to refill the detector in time with LN2, which typically should
be done after about 48 h.
3. In case the temperature-controlled transmission sample holder
is used, clean two CaF2 windows with 70 % ethanol and mount
them in the sample holder. Insert a Teflon spacer between the
two windows to avoid fringing (see Note 6). Place the sample
holder in the FTIR, close the lid around the sample holder
area, and wait for about 510 min until the sample area
is thoroughly flushed with dry air from the purge system
(see Note 7). In case the ATR accessory is used, clean the
diamond/ZnSe crystal with 70 % ethanol.
4. Open the program for acquisition of FTIR spectra. Set the
instrumental settings in the program: (i) give a background
name (yymmdd-bckg); (ii) add scan settings including type of
scan (background), spectral region (4,000900 cm1), and
number of scans [8]; (iii) set spectrometer resolution (4 cm1),
and deselect/select the automatic CO2/H2O correction (see
Note 8); (iv) select the type of accessory that is used (trans-
mission mode in case of the temperature-controlled sample
holder and reflection mode in case of the ATR accessory); and
(v) select the detector (MCT) and adjust the iris opening if
needed (see Note 9).
5. Acquire a background spectrum and save it. Make sure that
the sample area is thoroughly flushed with dry air and that the
contribution of water vapor and CO2 to the background spec-
trum does not change (see Note 10) (see Fig. 1).
6. Open the program for acquisition of FTIR spectra at timed
intervals. If needed, acquire a background spectrum within
this program using similar settings as indicated above (8 co-
added interferograms, 4,000900 cm1 wavenumber range,
4 cm1 resolution).
(CaF2 windows; background)
with purge
energy (a.u.)
water vapor
CO2-stretch.
without purge
4000 3500 3000 2500 2000 1500 1000

wavenumber (cm1)
Fig. 1 Background spectra as recorded in the 4,000900 cm1 region, using CaF2
windows. Spectra were recorded when using the purge system (upper trace),
and not using purge to flush with dry air (lower trace, gray line). Bands arising
from water vapor are indicated, as well as the CO2 stretching vibration
3.2 Acquisition 1. Concentrate cells from a suspension by centrifugation

of Transmission (1,200 g, 1 min) and transfer about 20 L hydrated pellet on
Spectra Using a CaF2 window. Add a second window so that the sample is
the Temperature- sandwiched between two windows. For dried specimens use a
Controlled Sample Teflon spacer between the windows to avoid sticking at high
Holder temperatures. Grease can be used around the windows to seal
the sample and prevent dehydration during heating.
2. Place the windows with the cell sample in the temperature-
controlled sample holder and insert the thermocouple such
that it touches the edge of the sample (via a Luer lock, grease
can be used to facilitate contact). Check connections for heat-
ing and cooling of the sample holder, close the lid around the
sample holder area, and wait for about 510 min until the area
is thoroughly flushed with dry air from the purge system.
3. Acquire a spectrum of the sample using settings as indicated
above (8 co-added interferograms, 4,000900 cm1 wave-
number range, 4 cm1 resolution, in absorbance) to verify the
quality of the spectrum. Prepare a new sample if needed (see
Note 11). Figure 2 shows typical spectra of hydrated, frozen,
and dried cell pellets.
4. Turn on both the cooling and heating devices. Set the begin
temperature using the controller unit (heater device) and wait
(fibroblasts)
H2O-libration, -bending
OH-stretching
CH-stretching
Amide-I
Amide-II
Amide-III
absorbance (a.u.)
dried
frozen
hydrated
4000 3500 3000 2500 2000 1500 1000

wavenumber (cm1)
Fig. 2 In situ infrared spectra of fibroblast cell pellets. Spectra were acquired at
20 C (hydrated, lower trace) and at 30 C (frozen, middle trace) from a
hydrated sample, as well as at 20 C for cells that were air-dried (upper trace).
The OH stretching and the H2O-libration and H2O-bending combination bands are
indicated, as well as the CH-stretching region mainly arising from membrane
lipids and protein amide-I, amide-II, and amide-III bands. Data adapted from [9]
until the sample holder reaches this temperature and remains

stable. Note that the thermometer with the thermocouple
touching the sample provides the actual sample temperature,
which generally differs (several degrees) from the value indi-
cated by the controller unit.
In case a Linkam stage is used for cooling with LN2 (see
above), turn on the manual pumping mode and use mode 1
for temperatures ranging from 0 to 20 C, mode 3 for tem-
peratures from 0 to 20 C, and mode 5 for temperatures
below 20 C down to 60 C.
5. Enter the temperature ramp settings using the controller unit.
Most sample holders for spectral data acquisition are not suited
for high cooling or warming rates. A ramp rate of 1 C min1
is suitable for most studies.
When a Harrick controller unit for heating is used (see
above), settings should be as follows: thermocouple type
(input/set, Ln1) H; control type (9LbL/set, CntL) nor; set-
ting units (9LbL/set, C_F) C; low temperature limit in C
(input/set, rL1) 200; high temperature limit in C (input/
set, rH1) 750; low power limit in % (9LbL/set, LoP) 0; high
power limit in % (9LbL/set, HiP) 35; autotune set point
(9LbL/set, AtSP) 90; ramp setting (9LbL/set, Rp), i.e., StPt
as mode for entering a set point temperature to be reached;
and ramp rate in C/min (9LbL/set, Rp), i.e., 1 for a cool-
ing/heating rate of 1 C min1.
6. In the program for acquisition of spectra at timed intervals,
enter the settings for spectra acquisition: (i) enter data collec-
tion mode (timed interval, e.g., 30 s) and duration of the run
(dependent on the temperature ramp and range), (ii) supply a
name for saving of spectra (folder, yymmdd-n; spectra, nNN#),
(iii) add scan settings (8 co-added interferograms, 4,000
900 cm1 wavenumber range), and (iv) select view during
spectra acquisition (stack plot of spectra versus time).
7. Start acquisition of spectra during the temperature scan:
(i) start the program for spectra acquisition at timed intervals,
(ii) start the temperature ramp by entering the end tempera-
ture (set point) on the controller unit, (iii) start the data log-
ger attached to the thermocouple monitoring the actual
sample temperature, and (iv) start a stopwatch. Make sure that
the heater and cooling devices are in correct modes, dependent
on the temperature scan that is performed (adjust during the run
when needed). Check spectra quality and sample temperature
during the course of the run. As an alternative for using the
data logger attached to the thermocouple, the sample tem-
perature can be noted manually as a function of time (every
210 min) during the run. When the end temperature is
reached, return the sample holder temperature to 20 C.
8. Construct a plot in which the sample temperature (y, in C) is

plotted versus the time point in the temperature run (x, in s);
and fit using a linear regression line (y = ax + b, where a repre-
sents the actual cooling/warming rate in C/s and b the offset
or start temperature in C) (see Note 12).
9. List the time points (x, in s) at which spectra were recorded
during the scan (each labeled with its own #), and calculate
their temperature (y, in C), using the linear regression equa-
tion determined above. In a spreadsheet program (e.g., Excel),
enter columns listing: (i) spectra numbers, (ii) time points of
collection, and (iii) calculated sample temperature during col-
lection. Results from spectral analysis can be added later.
When Timebase software (see above) is used, time points at
which individual spectra were acquired can be found in the
summarizing csv file.
10. Extract and save the individual spectra files (each labeled with
its own #) that are acquired during the temperature run, in a
separate folder (yymmdd-n, nNN; label with date and experi-
ment run number).
3.3 Acquisition 1. Concentrate cells from a suspension by centrifugation and

of Spectra During transfer a defined amount of hydrated pellet (e.g., 2 L) on
Drying of a Sample the diamond/ZeSn crystal. Alternatively, use a tissue slice.
Under Controlled 2. Close the donut-shaped dish with saturated salt solution
Humidity Conditions, around the sample area, to maintain the defined relative
Using the ATR Device humidity conditions.
3. In the program for acquisition of spectra at timed intervals,
enter the settings for spectra acquisition (timed interval of,
e.g., 30 s, duration of the run, and file name for the spectra)
and scan settings (8 co-added interferograms, 4,000900 cm1
wavenumber range).
4. Start the program for spectra acquisition at timed intervals.
Check spectra acquisition during the course of the run.
Figure 3 shows spectra of liposomes during drying.
5. Create a listing (in, e.g., Excel) with the time points (x, in s) at
which spectra were recorded (each labeled with its own #).
The results from the spectral analysis can be added later in the
same file.
6. Save the individual spectra that are collected during the course of
the run, in a separate folder (label with date and run number).
3.4 Spectral 1. Membrane phase behavior can be evaluated by analyzing the

Analysis: Membrane CH2 stretching region in the FTIR spectra (3,0002,800 cm1
Phase State and Phase wavenumber range). This region contains the asymmetric and
Behavior symmetric CH2 stretching vibration bands at ~2,925 cm1 and
~2,850 cm1, respectively, arising from the lipid acyl chains.
(liposomes)
asym. CH2-stretch.
OH-stretching
hydrated
sym. CH2-stretching
PO4-stretching
C=O-stretching
H2O-scissoring
absorbance (a.u.)
dried
4000 3500 3000 2500 2000 1500 1000

wavenumber (cm1)
Fig. 3 Infrared spectra of vesicles composed of egg phosphatidylcholine, during
drying at low relative humidity for different durations (2 L sample at 13 % RH
for up to 1 h; solid line, at 1 min; dark gray line, at 30 min; light gray line, at
50 min). The OH stretching and H2O scissoring bands arising from water, and the
CH2, C=O and PO4 stretching bands from the acyl chains, ester bonds, and head-
groups of phospholipids are indicated. Data adapted from [4]
Second derivative analysis can be used to resolve the band

positions more clearly. The average of the spectral position at
80 % of the peak height can be used to determine the peak
position of the band more easily. Membrane phase transitions
can be evaluated by plotting the position of the symmetric
CH2 stretching band as a function of the temperature or water
content of the sample (see Figs. 4 and 5).
2. Use the following procedure to calculate the band position of
the symmetric CH2 stretching vibration for each spectrum
(each labeled with its own #):
(a) Open spectrum and calculate the second derivative spec-
trum, using a 13-point smoothing factor.
(b) Select the spectral region between 2,865 and 2,835 cm1.
(c) Invert the spectrum, by multiplication by 1.
(d) Normalize this region, such that the peak around
2,850 cm1 is normalized to 1.
(e) Determine the peak position as the midpoint of the horizon-
tal line that intersects at 80 % of the peak height, and save it.
(f) Steps (a) through (e) need to be applied to all the spectra
that have been recorded during the run.
a b c
normalized second derrivative (r.u.)

(fibroblasts)
OH-stretching region CH2-stretching region
T (C):
absorbance (a.u.)
absorbance (a.u.)
-40
-30
-20
-10
0
10
20
2800 2600 2400 2200 2000 1800 3000 2950 2900 2850 2800 3000 2950 2900 2850 2800
wavenumber (cm1) wavenumber (cm1) wavenumber (cm1)
Fig. 4 Infrared spectra in the H2O bending (a) and CH2 stretching region (b) as acquired at various sample
temperatures of fibroblasts cooled from 20 to 40 C at 1 C min1. Dotted lines indicate the wavenumber
regions which were used for calculating the area of the H2O-libration and H2O-bending band (2,6901,960 cm1)
and position of the symmetric CH2 stretching vibration band (2,8652,835 cm1). Normalized second deriva-
tive spectra were calculated from the CH2 stretching region (C) to determine the band position (~2,850 cm1)
more easily, at 80 % of the peak height
2854 8
(fibroblasts)
7
2853 ice nucleation
(Tn) 6
AH2O (a.u.)
CH2 (cm1)
2852 5
4
2851 3
membrane phase transition 2

2850
(supra-zero) (sub-zero)
1
2849 0
40 30 20 10 0 0 10 20 30 40
temperature (C) temperature (C)
Fig. 5 Membrane phase behavior of fibroblasts at suprazero (left panel), as well as subzero (right panel) tem-
peratures. Spectra were acquired during cooling from 40 to 40 C at 1 C min1, while ice nucleation was
induced manually at 3 C. The band position of the symmetric CH2 stretching vibration band arising from
endogenous lipids was determined and plotted as a function of the sample temperature, to reveal changes in
membrane fluidity. The band area of the H2O-libration and H2O-bending combination band was determined to
assess when ice nucleation took place (open circles). The temperature-dependent decrease in membrane
conformational disorder is illustrated with lines. Discontinuities from this indicate when cellular membranes
undergo phase transitions, which are indicated. Figure adapted from [10]
3. In a spreadsheet program enter columns listing: (i) spectrum

number, (ii) time point of collection and/or sample tempera-
ture during collection, and (iii) symmetric CH2 stretching band
position (CH2). Create a plot in which CH2 is plotted as a

function of the time point of spectra acquisition or temperature
of the sample (see Note 13) (see Figs. 4 and 5). First derivative
analysis can be used to resolve phase transitions more clearly.
3.5 Spectral 1. The water-to-ice or ice-to-water phase transition can be stud-

Analysis: Water-to-Ice ied by following the change in the position and shape of the
Phase Transition H2O-libration and H2O-bending combination band around
2,200 cm1 in FTIR spectra acquired during cooling or heat-
ing of a sample (see Fig. 4a).
2. Use the following procedure to calculate the area of the H2O-
libration and H2O-bending combination band of each spec-
trum (each labeled with its own #):
(a) Open spectrum and select the spectral region between
2,690 and 1,965 cm1.
(b) Calculate the baseline-corrected area from 2,690 and
1,965 cm1 (the region may need to be adjusted depen-
dent on the sample), and save the result.
(c) Steps (a) and (b) need to be applied to all the spectra that
have been recorded during the run.
ture during collection, and (iii) area of the H2O-libration and
H2O-bending combination band (AH2O). Create a plot in
which AH2O is plotted as a function of the sample tempera-
ture (see Fig. 5b).
3.6 Spectral 1. Protein secondary structure can be evaluated by analyzing the

Analysis: Protein amide-I band in FTIR spectra (1,7001,600 cm1 wavenumber
Secondary Structure range). This region contains the C=O stretching vibration band
and Heat Denaturation around 1,655 cm1, arising from the protein backbone. Different
bands in the amide-I region represent different types of secondary
structure: -helical structures and turn/-sheet structures can be
found at ~1,655 cm1 and ~1,635 cm1, respectively. The H2O
band interferes with the amide-I band, which complicates analy-
sis. Difference spectra analysis can be used to resolve this issue.
Protein denaturation coincides with an abrupt change in the
amide-I band profile, which is reflected in the band area of sec-
ond derivative difference spectra (see Fig. 6).
2. Use the following procedure to calculate the area of the amide-
I band of each spectrum (each labeled with its own #):
(a) Open spectrum (recorded at a particular time point or
temperature) and subtract the first recorded spectrum
(reference spectrum, typically recorded at 0 or 20 C).
(b) Calculate the second derivative of the difference spectrum
obtained in step (a), using a 13-point smoothing factor.
a b c
inverted second derrivative (a.u.)

(fibroblasts)
80C
absorbance difference (a.u.)
amide-I region
area Amide-I (a.u.)

T (C): protein
0 denaturation
30 0C (Td)
40
50
60
80
1700 1680 1660 1640 1620 1600 1700 1680 1660 1640 1620 1600 0 10 20 30 40 50 60 70 80
1 1 temperature (C)
wavenumber (cm ) wavenumber (cm )
Fig. 6 Infrared spectra of fibroblasts were collected during heating from 0 to 90 C at 2 C min1. Panel a shows
difference spectra, in the amide-I region, of the spectrum recorded at 0 C and those at the indicated tempera-
ture. Difference spectra were calculated to subtract the interfering contributions of H2O in this region. To
resolve different components within the amide-I band, second derivatives of the difference spectra were
calculated (b). Denaturation coincides with a decrease in the band at 1,625 cm1 and a decrease in the band
at 1,655 cm1. The area of the second derivative band from 1,640 to 1,605 cm1 (indicated with dotted lines)
was determined and plotted as a function of the sample temperature, to reveal changes in protein secondary
structure (c). Figure adapted from [2]
(c) Select the spectral region between 1,700 and 1,600 cm1.
(d) Invert the spectrum, by multiplication by 1.
(e) Calculate the baseline-corrected area from 1,640 to
1,605 cm1 (the region may need to be adjusted depen-
dent on the sample), and save the result.
number, (ii) sample temperature during acquisition, and (iii)
calculated area of the 1,6401,605 cm1 spectral range in the
inverted second derivative difference spectrum. Create a plot
in which the area is plotted as a function of the temperature at
which the spectra were recorded (see Fig. 6).
3.7 Spectral 1. This procedure is suitable for a liposome model system. In this
Analysis: Drying case, the line-height ratio between the water scissoring band
Kinetics in Liposome (H2O, ~1,650 cm1) and the lipid ester band (CO,
Model System ~1,736 cm1) can be used as a measure for the water content
of the sample during drying.
2. Use the following procedure to calculate the line-height ratio
between the water scissoring band and the lipid ester band
(IH2O/ICO) of each spectrum (each labeled with its own #):
(a) Open spectrum, and select the spectral region from 1,750
to 1,700 cm1.
(b) Determine the intensity of the peak at 1,736 cm1 (ICO),

using a baseline from 1,750 to 1,700 cm1, and save the
result.
(c) Select the spectral region from 1,700 to 1,600 cm1.
(d) Determine the intensity of the peak at 1,650 cm1
(IH2O), using a baseline from 1,700 to 1,600 cm1, and
save the result.
(e) Determine IH2O/ICO from steps (b) and (d) and save
the result.
number, (ii) time point of collection, (iii) peak intensity of water
scissoring band (IH2O) and lipid ester band (CO), as well as
their ratio (IH2O/ICO). Create a plot in which the IH2O/
ICO is plotted as a function of the drying time. Drying is vis-
ible as a decrease in this ratio, and its value provides a relative
measure for the water content of a sample (see Fig. 7).
3.8 Spectral 1. This procedure is particularly suitable to study sugar glasses.

Analysis: Glass Glass transitions can be studied by following the band position
Transition of the OH stretching vibration band around 3,300 cm1, aris-
Temperature ing from sugar OH groups, as a function of the temperature.
of Amorphous 2. Use the following procedure to calculate the position of the OH
Systems stretching band of each spectrum (each labeled with its own #):
a b c
band height H2O/CO (a.u.)
(liposomes) CO, H2O (sugar glass)

absorbance difference (a.u.)
time (min):
2
OH (r.u.)
glass
transition
8 (Tg)
12
1800 1700 1600 1500 0 5 10 15 20 25 30 0 20 40 60 80 100 120

wavenumber (cm1) time (min) temperature (C)
Fig. 7 Infrared spectra of egg phosphatidyl choline were collected during drying (2 L sample at 94 % RH for
up to 30 min). Panel a shows the 1,8001,500 cm1 wavenumber range, which contains the lipid ester (CO,
~1,736 cm1) and water scissoring (H2O, ~1,650 cm1) bands. The band height ratio between these bands
(IH2O/ICO) was calculated to capture dehydration kinetics (b). Panel c shows glassy behavior for a sucrose
glass. The data points reflect the relative shift in the band position of the OH stretching band as a function of
temperature. Data adapted from [4, 7]
(a) Open spectrum, and select the spectral region between

3,600 and 3,000 cm1.
(b) Normalize this region, such that the peak around
3,300 cm1 is normalized to 1.
(c) Determine the peak position as the midpoint of the hori-
zontal line that intersects at 80 % of the peak height, and
save the result.
(d) Steps (a) through (c) need to be applied to all the spectra
ture during collection, and (iii) position of the OH band
(H). Create a plot in which H is plotted as a function of
the sample temperature. The glass transition temperature can
be determined as the intersection of linear regression lines in
the glassy and liquid state (see Fig. 7c).
4 Notes
1. A TGS (triglycine sulfate) detector can also be used. The

signal-to-noise ratio of a TGS detector is not as good com-
pared to that of an MCT (mercury cadmium telluride) detec-
tor, which increases the data acquisition time to obtain spectra
of sufficient quality for data analysis. Therefore, an MCT
detector is preferred.
2. An air compressor is needed to run the FTIR purge gas gen-
erator. If no air compressor system is available, the generator
can also be hooked up to a nitrogen gas cylinder.
3. Procedures and methods can easily be adapted using FTIR
systems from other suppliers (e.g., Thermo Scientific, Brker).
Furthermore, LabVIEW and MATLAB software can also be
used for data acquisition and spectra analysis.
4. Commercially available temperature-controlled sample hold-
ers like the one from Harrick can be replaced by custom-
designed devices, allowing more flexibility in sample handling
and temperature regimes.
5. A liquid nitrogen cryogenic pump micro-dosing system from
Norhof (Maarssen, Netherlands) can also be used for cooling
the sample holder.
6. Fringing is visible as a sinusoid shape interfering throughout
the spectra. If fringing is observed, mount the windows again
in the sample holder or use another pair of windows and
acquire a new background spectrum.
7. Dry air prevents condensation of the sample windows at tem-

peratures below 5 C. Contributions of water vapor from
ambient air to the spectrum strongly interfere, particularly
with the protein bands.
8. It is recommended to select an automatic CO2/H2O correc-
tion when ATR studies are done. For transmission studies the
automatic correction is not needed if both background and
sample are recorded after sufficient purging of the sample area.
9. The iris opening can be adjusted if needed.
10. Make sure that both background and sample spectra are
recorded after sufficient purging of the sample area (minimally
5 min). The disappearance of water vapor after purging is par-
ticularly visible in the 1,8001,500 cm1 region (see Fig. 1).
11. If the absorbance values exceed 2, this indicates the sample is
too thick, which may result in distortion of the spectral bands.
12. The actual ramp rate of the sample is generally different from the
set value of the ramp rate, due to discrepancies between the tem-
perature that is indicated on the controller unit and the actual
sample temperature which is measured by the thermometer.
13. Calibrate the system by measuring phase transition temperatures
of pure lipid systems with a known melting temperature (i.e.,
DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine, and
DOPC, dioleoyl-sn-glycero-3-phosphocholine). Lipid phase
transition temperatures should be measured during heating,
since literature values are typically reported from heating scans.
Acknowledgment
This work is supported by funding from the German Research

Foundation (DFG, Deutsche Forschungsgemeinschaft) for the
Cluster of Excellence From Regenerative Biology to Reconstructive
Therapy (REBIRTH).
References
1. Wolkers WF, Hoekstra FA (2003) In situ FTIR 4. Wolkers WF, Oldenhof H, Glasmacher B
assessment of desiccation-tolerant tissues. (2010) Effect of trehalose on dehydration
Spectroscopy 17:297313 kinetics of phospholipid vesicles, as measured
2. Wolkers WF, Oldenhof H (2010) In situ FTIR in real time using ATR infrared spectroscopy.
studies on mammalian cells. Spectroscopy Cryobiology 61:108114
24:525534 5. Mantsch HH, McElhaney RN (1991)
3. Wolkers WF, Balasubramanian SK, Ongstad Phospholipid phase transitions in model and
EL, Zec H, Bischof JC (2007) Effects of freez- biological membranes as studied by infrared
ing on membranes and proteins in LNCaP spectroscopy. Chem Phys Lipids 57:213226
prostate tumor cells. Biochim Biophys Acta 6. Goormaghtigh E, Cabiaux V, Ruysschaert JM
1768:728736 (1994) Determination of soluble and
membrane protein structure by Fourier 8. OBrien FEM (1948) The control of humidity by
transform infrared spectroscopy. I. Assignments saturated salt solutions. J Sci Instrum 25:7376
and model compounds. Subcell Biochem 9. Akhoondi M, Oldenhof H, Sieme H, Wolkers
23:329362 WF (2012) Freezing-induced removal of water
7. Wolkers WF, Oldenhof H, Alberda M, from phospholipid head groups in biomem-
Hoekstra FA (1998) A Fourier transform infra- branes. BSI 1:293302
red microspectroscopy study of sugar glasses: 10. Oldenhof H, Akhoondi M, Sieme H, Wolkers
application to anhydrobiotic higher plant cells. WF (2013) Use of Fourier transform infrared
Biochim Biophys Acta 1379:8396 spectroscopy to study membrane properties of
cells at subzero temperatures. BSI 2:8390
Chapter 6
Calorimetric Analysis of Cryopreservation

and Freeze-Drying Formulations
Wendell Q. Sun
Abstract
Differential scanning calorimetry (DSC) is a commonly used thermal analysis technique in cryopreserva-
tion and freeze-drying research. It has been used to investigate crystallization, eutectic formation, glass
transition, devitrification, recrystallization, melting, polymorphism, molecular relaxation, phase separa-
tion, water transport, thermochemistry, and kinetics of complex reactions (e.g., protein denaturation).
Such information can be used for the optimization of protective formulations and process protocols. This
chapter gives an introduction to beginners who are less familiar with this technique. It covers the instru-
ment and its basic principles, followed by a discussion of the methods as well as examples of specific
applications.
Key words Cryopreservation, Crystallization, Devitrification, Differential scanning calorimeter,

Freeze concentration, Freeze-drying, Glass transition, Phase separation, Vitrification
1 Introduction
1.1 General Differential scanning calorimetry (DSC) is probably the most com-
Introduction monly used thermal technique in cryopreservation and freeze-
drying research. It is a simple but very powerful tool. DSC has
been used for research in cryopreservation and freeze-drying, par-
ticularly to investigate crystallization, eutectic formation, glass
transition, devitrification, recrystallization, melting, polymor-
phism, molecular relaxation, phase separation, thermochemistry,
and kinetics of complex reactions (e.g., protein denaturation).
Such information can be used for the development of protective
formulations and to rationally design process protocols. This chap-
ter provides an overview for beginners who are less familiar with
thermal analysis techniques and wish to apply DSC in cryopreser-
vation or freeze-drying research. It starts with a brief introduction
to the instrument and its basic principles, followed by a discussion
of methods and examples of specific applications.
163
164 Wendell Q. Sun
Fig. 1 A differential scanning calorimeter with a cooling unit attached for cryopreservation and freeze-drying
research. Displayed is a DSC Q2000 (TA Instruments, Inc., New Castle, DE)
1.2 Basics Figure 1 shows a DSC system with a cooling unit attached for
of a Differential cryopreservation and freeze-drying research. The DSC system
Scanning Calorimeter: consists of several subsystems: cooling/heating elements, a tem-
Instrument perature controller, a gas flow rate controller, a signal amplification
and Principles module, a differential scanning detector, and a data collection sub-
system. The cooling/heating elements provide the temperature
conditions that are required for sample measurement.
Cryopreservation and freeze-drying research generally use low-
temperature DSC models, typically in the temperature range
between 170 C and 250 C. The temperature controller regu-
lates the cooling and heating rates, as specified by the investigator
for individual experiments. The signal amplification module ampli-
fies the very small thermal difference detected by the thermal cou-
ples and increases the sensitivity and accuracy for thermal
measurements. The differential scanning detector is the most criti-
cal core component that includes sample cells and thermal couples
for temperature measurement and signal transduction. The gas
flow regulator controls the gaseous environment in which the mea-
surement is made. The data collection subsystem (a computer)
automatically records and stores the measurement data and allows
data retrieval for subsequent analysis.
DSC measures the heat flow of a sample during the cooling
and/or heating scans. There are two cells in the detector, one refer-
ence cell and one sample cell. An empty crucible is placed into the
reference cell, whereas a crucible containing the sample is placed
into the sample cell. During a DSC experiment, both reference cell
and sample cell are cooled and/or heated according to a selected
temperature/time ramping program over a temperature range of
interest. The scanning detector records the difference in heat flow
Calorimetric Analysis 165
or power between the reference and the sample as a function of

temperature and/or time, which is called a thermogram. The ther-
mogram shows the release and absorption of heat energy by the
sample or the change of heat capacity at different temperatures.
The basic principle involved in the calorimetrical analysis of
cryopreservation solution and freeze-drying formulations is that,
on one hand, the thermal changes of water and their correspond-
ing quantities of energy are greatly affected by the presence of sol-
utes and solvents or other biomaterials and that, on the other hand,
the thermal behaviors of solutes and other materials in the system
are affected by the presence of water. For example, as water con-
tent decreases and/or the solute concentration increases, the onset
freezing and melting temperature of water decreases, while at the
same time, the glass transition temperature increases due to the
reduced plasticization effect by water. By analyzing the thermo-
gram of a cryopreservation solution, one gains insight into the
nature of the thermophysical and/or thermochemical events that
might occur during cooling and warming. Such information is of
great value for the development of cryopreservation solutions and
successful cooling/warming protocols. Similarly, by analyzing the
thermograms of a freeze-drying formulation and the freeze-dried
product that is produced by a given freeze-drying process, one
obtains the necessary information for reformulation and optimiza-
tion of process control parameters, as well as information about the
stability and shelf life of the product.
1.3 Basics When a substance is cooled or heated to a given temperature, it

of a Differential may melt, crystallize, solidify, change the state, adsorb, desorb,
Scanning Calorimeter: decompose, and react. The associated changes in the enthalpy yield
Thermogram various thermal effects. A plot of heat flow against measurement
temperature or time is called a thermogram. From the thermo-
gram, one can see the number of thermal events, their positions,
magnitudes, and shapes. The number of thermal events, such as
peaks and steplike changes, shows how many thermochemical and
thermophysical changes have occurred. Their positions show the
temperatures and/or time at which these changes start and end.
The magnitudes of thermal events are represented by the peak
heights and peak areas that relate to the energy involved in the
individual thermal events. The direction of thermal events indi-
cates the energy change to be endothermic (i.e., taking energy) or
exothermic (i.e., releasing energy).
Figure 2 illustrates the cooling and warming thermograms of a
hypothetical aqueous solution. It should be noted that the curves
in this section are drawn only to demonstrate probable thermal
events and not to represent a real cryopreservation solution or a
freeze-drying formulation. Upon cooling, the hypothetical solu-
tion undertakes three significant thermal events. First, this solution
supercools (also known as undercools) to below its equilibrium
166 Wendell Q. Sun
Fig. 2 Cooling and warming thermogram of a hypothetical aqueous solution. The cooling part of the thermo-
gram shows supercooling, water crystallization, salt precipitation or hydrate formation, and glass formation of
the freeze-concentrated amorphous domain, whereas the warming part of the thermogram shows the glass
transition of the amorphous domain, eutectic melt, ice recrystallization, glass transition of the maximally
freeze-concentrated amorphous domain, as well as ice melting. See the text for a detailed description
freezing temperature, and a part of water nucleates and crystallizes

to form ice at a lower temperature, resulting in a transient increase
in solution temperature during the cooling due to the release of
latent heat in liquid water (the freeze loop in Fig. 2) (exother-
mic). A cryopreservation or freeze-drying solution crystallizes
below its equilibrium freezing point only in the presence of a seed
nucleus. In the absence of a seed nucleus, the solution maintains its
liquid state down to its homogeneous nucleation temperature.
Second, the formation of ice causes the concentrations of all sol-
utes in the solution to increase. When the concentration of a solute
(e.g., a salt) reaches above its saturation point, the solute may pre-
cipitate out of the solution and/or solidify as the eutectic composi-
tion (exothermic). The eutectic event rarely occurs in
cryopreservation studies but is observed quite often in various
freeze-drying solutions. The third thermal event is the glass forma-
tion of the freeze-concentrated domain, exhibited as a steplike
transition or a gradual baseline shift in the thermogram due to the
change of heat capacity from one state to another state. This distin-
guishes it from freezing and melting transitions, which will pro-
duce peaks. The glass transition event is often masked by the large
water crystallization peak and is usually only visible after expanding
the y-axis. Upon warming, a series of related but often more com-
plex thermal events occur in the frozen system. As the temperature
increases, the solidified amorphous domain undergoes a glass
transition. At higher subzero temperatures, unfrozen water may
Fig. 3 The heating thermogram of a hypothetical low-moisture sample. As temperature increases, the sample
undergoes glass transition, crystallization, melting, exothermic reactions, and decomposition
crystallize, likely resulting in further concentration of the solutes in

the amorphous domain. Small ice crystals that are preformed dur-
ing cooling may recrystallize to form large ice crystals. The eutectic
composition will melt (endothermic), and the maximally freeze-
concentrated amorphous domain will undergo another glass tran-
sition. Finally ice melts (endothermic).
Figure 3 is a typical thermogram of a low-moisture sample
such as the freeze-dried protein with the presence of crystallizing
bulk agents upon heating from ambient temperature to 300 C. As
the temperature increases, the sample first undergoes a glass transi-
tion, which is followed by crystallization of bulk agents and then
melting of the crystals. As temperature increases further, several
exothermic reactions could occur, including oxidation and cross-
linking, before the sample decomposes. Crystallization and melt-
ing are not always observed for many low-moisture samples.
However, the other three thermal events (glass transition, exother-
mic reactions, and sample decomposition) are usually detectable.
Calorimetric analysis is an informative tool for analyzing sam-
ples for cryopreservation and freeze-drying. In addition to the
qualitative measurements as illustrated in Figs. 2 and 3, it is also
widely used to do quantitative measurements, such as water trans-
port across cell membrane during extracellular freezing, the kinet-
ics of crystallization and phase separation in viscous liquids, the
effect of annealing treatments in frozen systems, molecular relax-
ation dynamics of freeze-dried products, as well as kinetics of reac-
tions and decomposition, etc. However, calorimetric analysis has
its own limits. If the events do not involve thermal changes, DSC
168 Wendell Q. Sun
analysis is not applicable. There are also difficulties in correct iden-

tification of origin for some thermal events in heterogeneous, com-
plex biological samples.
1.4 Basics The operation of running calorimetric tests is simple. Modern

of Differential computer and thermal analysis software enable one with little back-
Scanning Calorimeter: ground of mathematical and physical sciences to perform data
Experimental analysis. However, new users may be disappointed when they ini-
Conditions tially obtain poorly reproducible results. The thermogram of the
same material may vary with crucibles used and among different
DSC models. Even using the same DSC unit, the result may differ
among operators. The temperature, shape, height, and area of
peaks can vary considerably. This problem arises mainly from the
complexity of heat transport phenomenon, which could be affected
by many factors. The solution to this problem is to understand
precisely what should be measured and then to select the right
experimental conditions for specific tests.
Preparation of Samples: DSC is applicable to solutions, vis-
cous liquids, and solids or powders. During the measurement, the
test sample is contained in a pair of crucibles or pans. Heat trans-
port properties of crucibles are affected by the material type,
design, and size, and a beginner may need to consult more expe-
rienced users for advice to choose suitable crucibles (see Note 1).
Solid and powder samples sometimes do not make good contact
with the bottom of the crucibles and may need to be pressed
before testing for better heat transport results. Atmosphere, mois-
ture, and pressure may influence thermophysical changes and
thermochemical reactions of certain samples, and therefore such
samples may need to be prepared with sealable crucibles in the
environment of inert gases and/or under the right pressure condi-
tion. Enough sample material should be added to allow accurate
measurement, but an oversized sample reduces the resolution and
often causes adjacent peaks that overlap. In most situations, a sam-
ple size of 1020 mg is suitable.
Temperature Ramping Program: The cooling and heating
rates affect both the position (temperature/time) of a thermal
event and its magnitude (height or area). In general, a faster rate
results in larger peak area but is more likely to deviate from the
equilibrium condition due to nonuniform heating. Similarly to the
oversize effect, rapid scanning may also lead to a lower resolution
and overlapping of adjacent peaks. A slow rate can separate peaks
better and measure more accurately but requires the instrument to
have a higher sensitivity. It should be noted that thermal events of
cryopreservation and freeze-drying solutions, as well as biological
materials, are not just thermodynamic events but also kinetic
events, i.e., time-dependent processes. Therefore, the tempera-
ture/time ramping program must be carefully designed to meet
the requirements for specific measurements.
2 Materials
2.1 Equipment 1. DSC Q2000 (TA Instruments, Inc., New Castle, DE) with a
and Materials TA refrigerant cooling system (RCS, 90 C).
2. A computer with the Advantage and TA Universal Analysis
software installed.
3. Hermetic aluminum crucibles (pans and covers, Tzero).
4. TA Instruments Q Series sample encapsulating press.
5. Indium (99.9999 %, Standard Reference Material 2232 certi-
fied by the National Institute of Standards and Technology).
6. 10 % maltodextrin (M180) solution (prepared in saline).
3 Methods
3.1 DSC Verification Sensors and the conversion of thermocouple voltage to heat flow
of the DSC units are periodically calibrated. It is a good practice to
verify the instruments accuracy of measurements. The SRM-2232
indium is used as the thermometric standard for verification.
1. Prepare a SRM-2232 sample (58 mg) in an aluminum cruci-
ble. Record the actual mass with precision to 0.0001 g. Run
the sample with the following 4-step temperature ramping
program: Ramp at 20 C/min from the room temperature to
180 C; ramp at 20 C/min to 130 C; isothermal for 5 min;
ramp at 2 C/min to 170 C.
2. Use the Universal Analysis software to determine the onset tem-
perature (Tm) of fusion (melting) and heat (enthalpy) of fusion
(H) of the indium sample according to the data of the second
slow-heating segment (room 130 C to 170 C at 2 C/min).
3. The certified Tm and H values of SRM-2232 indium by NIST
are 156.5985 0.00034 C and 28.51 0.19 J/g, respectively.
The measured values of Tm and H meet the accuracy require-
ment if Tm is 156.6 0.3 C and H is 28.5 0.5 J/g.
3.2 Operation This procedure below outlines the operation of the TA differential
Procedure scanning calorimeter (DSC-Q2000) to measure the Tg of 10 %
maltodextrin (M180) solution. The basic operation of the instru-
ment is the same for different types of measurements, except with
different temperature/time ramping programs.
1. Prepare the sample: Weigh and record the mass of two pairs of
empty crucibles. Load approximately 1015 L test solution in
the center of the pan. Weigh again to obtain the net sample
weight. Place the cover and seal the crucible with the press.
Make the other pair of crucibles as the reference.
170 Wendell Q. Sun
2. Load the reference crucible and the test sample into the DSC
cell: The DSC has an auto-lid assembly. Touch the LID OPEN
key on the DSC Control Menu to open the cell. Place the refer-
ence on the rear or left platform, and place the sample on the
front or right platform (see Fig. 1). Touch the LID CLOSED
key on the DSC Control Menu to close the DSC cell.
3. Check that the purge gas is connected and set to the desired
flow rate at 50 mL/min (see Note 2).
4. Check the refrigerant cooling system (RCS), and make sure it is
ready. If an RCS, be sure a base purge gas is used (see Note 3).
5. Open the Q Series Explorer program (the software that con-
trols the instrument). Select the SUMMARY PAGE, and
enter the following information: Mode, standard, and Test,
custom. Then enter sample name, net sample mass (weight),
crucible mass, and comments (information you wish to enter
about the sample or measurement). Enter a data file name and
specify the path to save the electronic data file.
6. Select the PROCEDURE PAGE and create a test method.
Name the new method in the procedure page. Click on
EDITOR (to the right of the Name field) to open the
method dialog box. In the method dialog box, a segment list
will appear on the right. To choose any segment function, dou-
ble-click on the function. Enter the method parameters as fol-
lows: Ramp at 5 C/min from the room temperature to 0 C;
ramp at 0.5 C/min to 40 C; isothermal for 10 min; ramp at
2 C/min to 20 C. Click the APPLY button at the bottom
of the page to save the method.
7. Select the NOTES PAGE. Record the operators initials and
other information.
8. Start the DSC measurement. Double-click on the red arrow
on the upper left side of the screen. Wait for test completion.
9. Upon completion, unload the test sample from the DSC cell.
3.3 Data Analysis 1. Open the TA Universal Analysis software. From the main
menu, open the saved electronic file. The first window will
bring out sample information. Make any correction needed if
the information displayed on the screen is not correct.
2. Click on the SIGNALS button to bring out the signal selec-
tion window. Select the desired y-axis signals to plot (heat
flow), using the drop-down lists. Select the desired x-axis sig-
nal to plot (temperature). Click the OK button when to plot
data.
3. Expand the warming segment of the DSC thermogram
between 30 C and 10 C, using the graph-plotting tools.
Use the GLASS TRANSITION on the ANALYZE menu
Fig. 4 Data analysis window in a TA Universal Analysis program displaying the onset, end, and midpoint of the
glass transition
to calculate the onset glass transition temperature. When you

select the above option, two markers will be displayed on the
graph (Fig. 4).
4. Double-click on the curve at a point about 25 C to position
the first marker (before the transition), and then double-click
to position the second marker between 15 C and 10 C
(after the transition). Right click to display the ANALYZE
pop-up menu, and select ACCEPT LIMITS (or press
Enter). Glass transition temperatures (onset, end, and mid-
transition) will be displayed on the graph (Fig. 4). The onset
Tg is 21.3 C as displayed in the inset.
5. To further analyze the region of interest on the curve, select
the region of interest, select INSERT (right click on the
mouse), and specify the x- and y-axis range (the inset in Fig. 4).
4 Examples of Specific DSC Applications in Cryopreservation and Freeze-Drying
4.1 Solutions Traditional cryopreservation methods use slow and controlled

for Traditional rates to freeze biomaterials of interest for long-term preservation at
Cryopreservation ultra-low temperatures, typically in 80 C freezer or liquid nitro-
(Slow Freezing) gen freezer (196 C). Under such conditions, ice formation does
occur, normally in the extracellular space. Various slow and con-
trolled rate methods are well established for various mammalian
172 Wendell Q. Sun
cells, including oocytes, embryos, sperms, stem cells, and tissues.

These methods consist of one or more programmed freezing steps
at or to certain low temperatures before plunging samples into
liquid nitrogen freezers or placed into 80 C freezers for long-
term cryopreservation. Lethal intracellular ice formation is pre-
vented when cooling is slow to allow sufficient cellular water to
flux out to the extracellular space. The key for successful cryo-
preservation of living materials is to find a balance by preventing
intracellular ice formation, minimizing extracellular ice formation,
avoiding the detrimental effects of excessive dehydration and high
solute concentration, and at the same time achieving the required
stability during long-term storage and rewarming.
Optimal cooling and rewarming rates during cryopreservation
depend on the cell type, size, and water permeability across the cell
membrane, as well as the use of cryoprotectants. DSC has been
used as a tool to measure water transport across the cell membrane
during extracellular freezing and to optimize cryopreservation
parameters [13].
Figure 5 shows an example where the cryopreservation solution
is suboptimal for slow freezing. In this example, human red blood
cells are cryopreserved by plunging directly into liquid nitrogen in
the presence of 12 % hydroxyethyl starch (average molecular weight
of 263 kDa, molar substitution of 0.45) [4]. Hydroxyethyl starch is
a non-permeable cryoprotectant and requires fast cooling and
warming rates to avoid intracellular ice formation and excessive cel-
lular dehydration. DSC analysis has revealed that rapidly frozen
sample is unstable at elevated subzero temperatures. First, after
Fig. 5 DSC warming thermograms of human red blood cells cryopreserved with 12 % (wt/vol) hydroxyethyl
starch by plunging into LN2. It shows the instability of the frozen sample upon warming above 130 C at a
rate of 3 C/min. The inset is a thermogram of a frozen HES/erythrocyte sample after annealing at 40 C for
4 days. Curves were redrawn according to Sun et al. [4]
glass transition at 121 C upon warming, devitrification is observed

at approximately 110 C (Td). The devitrification peak is very
small as compared to the ice melting peak that occurs above 10 C,
indicating that bulk ice formation happened during rapid cooling
when the sample is plunged into liquid nitrogen. Second, there is a
broad second-order transition between 90 C and 50 C. This
broad transition shows the presence of nonhomogenous amor-
phous domains. The blood cell suspension has been frozen rapidly,
resulting in the formation of multiple amorphous domains with
varying degrees of freeze concentration. During warming, glass
transitions of these domains overlapped one another, giving rise a
continuous baseline shift over a broad temperature range. Third,
between 45 C and 10 C just before ice melting, there are glass
transitions of a less concentrated amorphous domain starting at
38 C, recrystallization in freeze-concentrated domains around
20 C to 25 C, and glass transition of the maximally freeze-
concentrated domain at 17 C. The inset shows the thermogram
of a frozen sample after isothermal annealing at 40 C. Annealing
eliminates all thermal transitions below 40 C and also avoids
recrystallization between 25 C and 20 C. Some of these ther-
mal changes have important implications affecting the survival of
frozen cells. Devitrification (crystallization) is thought to be a major
event damaging rapidly frozen cells during slow warming and iso-
thermal storage at elevated subzero temperature. In the system
studied, devitrification around 110 C does not cause significant
damage to frozen blood cells. When frozen samples are transferred
from 196 C to 78 C for isothermal storage, time-dependent
hemolysis remains less than 4 % within the first 6 h, and maximum
hemolysis has been reported to be 1012 % after extended storage.
Following the broad transition above 80 C, the stability of the
frozen cells declined significantly. Within this temperature range,
no particular DSC-detectable thermal event has been linked to the
instability of frozen erythrocytes. Slow recrystallization (undetect-
able by DSC) would occur at this temperature range. The abrupt
increase in hemolysis occurs at temperatures above 20 C, which
is associated with the glass transition of freeze-concentrated amor-
phous domains and probably also to the melting of ice [4]. For
optimization purposes, the DSC study points to the need of increas-
ing the warming rate above 40 C.
In plant biotechnology, DSC has been used to investigate the
mechanism of actions for the effect of desiccation (or dehydra-
tion), pre-culture or osmotic preconditioning, and treatments with
cryoprotectant solutions on the survival of cells and tissues after
cryopreservation. Most studies used DSC to measure the reduc-
tion of ice formation and the increase of glass transition tempera-
ture after various treatments and used the data for the optimization
of cryopreservation protocols. One of the successful approaches is
the encapsulation-dehydration cryopreservation protocol, which
174 Wendell Q. Sun
involves the encapsulation of the osmotically preconditioned cells

and/or tissues in alginate and then desiccation to a tolerable lower
water content in order to increase the glass transition temperature
and to minimize ice damage during cooling. The freezing behav-
iors of treated specimens under different cooling programs can be
obtained from freezing thermograms. The amount of ice formed
in a sample can be quantified by integrating the enthalpy of the
freezing peak(s) and/or ice melting peak(s), and the glass transi-
tion temperature can be found in the cooling and warming ther-
mograms. An example is given in Fig. 6, where the researchers
characterized thermal profiles of silica gel desiccated encapsulated
shoot tips of Ribes nigrum upon cooling and warming [5]. Five
hours of desiccation reduced the water content of shoot tip sam-
ples from ~3.8 to ~0.4 g/g dry mass and increased the glass transi-
tion temperature to ca. 80 C, and only a small amount of ice
formation was observed from the warming curve. The protocol
achieved >60 % post-cryopreservation viability. Seven hours of des-
iccation reduced the water content further to ~0.3 g/g dry mass,
increased the glass transition temperature to ca. 55 C, eliminated
ice formation completely, and achieved ice-free cryopreservation
(see the next section).
Fig. 6 DSC thermograms of encapsulated meristems of Ribes nigrum cv. Consort after desiccation with silica
gel at 16 C for 5 h and 7 h. Meristems were pretreated by placing on 0.75 M sucrose for 7 days, encapsulated
in alginate, and dehydrated in 0.57 M liquid sucrose medium for 22 h. Samples were cooled at a rate of 10 C/
min from 25 C to 100 C and then at 5 C/min to 150 C. Sample was held isothermally for 5 min before
ramping up to 100 C at a rate of 10 C/min and at 5 C/min to 25 C. A small fraction of water (~0.1 g/g
sample) was still freezable in the 5 h-desiccated sample. Curves were redrawn according to Sherlock et al. [5]
4.2 Solutions DSC analysis can be used to optimize the cryoprotectant formula-
for Ice-Free tions for ice-free cryopreservation and to determine the thermal
Cryopreservation critical cooling and warming rate for specific vitrification protocols.
(Vitrification) Ice-free cryopreservation or vitrification was introduced into cryo-
preservation of mammalian cells by Rall and Fahy [6] and for plant
cells and tissues by Uragami et al. [7] and Sakai et al. [8].
Vitrification methods have the advantage over traditional slow-
freezing cryopreservation methods because they avoid damaging
ice crystal formation.
In general practice, there are two methods to achieve vitrifica-
tion: using very high cooling/warming rates and high cryoprotectant
concentration. Vitrification of water can be achieved by an extremely
rapid cooling rate (>106 C/s) even without the presence of cryo-
protectants. With the presence of high cryoprotectant concentra-
tion, the cooling rate can be decreased. However, the damage to
cells and tissues occurs at high concentration due to the toxicity of
certain cryoprotectants and the high osmolality of vitrification solu-
tions. In most cases, poor post-cryopreservation survival could be
attributed to osmotic injury, solution toxicity, and insufficient pen-
etration of vitrification solutions into the specimen.
PVS1 and PVS2 have been widely used for plant cell cryo-
preservation by vitrification for more than two decades. Few
attempts have been made to improve vitrification solutions or to
develop new vitrification solutions after PVS2 was invented. The
options are quite limited for cryopreservation of plant cells and tis-
sues by vitrification. In their study, Suzuki et al. [9] aimed to
develop an alternative vitrification solution with a potentially wide
applicability to the cryopreservation of plant-cultured materials
and genetic resources. Figure 7 shows the warming thermograms
of four cryoprotectant formulations (PVS1, PVS2, VSL, and
VSL+). DSC warming thermograms show that all four solutions
vitrified upon rapid cooling to liquid nitrogen temperatures; how-
ever, devitrification occurred after glass transition during warming.
PVS2 and VSL+ have much smaller devitrification exotherms due
to the increase in the concentration of glycerol, Me2SO, and/or
sucrose. The authors further studied the effect of cooling rates.
Cooling rates between 100 C/min and 10 C/min did not
affect the warming thermograms of vitrified solutions, demonstrat-
ing that the new solutions (VSL and VSL+) could be vitrified even
at moderate cooling rates. PVS2 contains glycerol (30 % w/v) and
its high viscosity leads to slower penetration into tissues. VSL is less
viscous than PVS2 because of lower glycerol concentration and
higher ethylene glycol concentration, permitting faster penetration
and/or faster dehydration of tissues. The newly developed vitrifica-
tion solutions (VSL and VSL+) were also tested for cryopreserva-
tion of gentian axillary buds. Excised stem segments (shoot apices)
were pre-cultured with sucrose to induce osmotic tolerance prior
to cryopreservation. VSL exhibited 78 % survival as determined by
176 Wendell Q. Sun
Fig. 7 DSC warming thermograms of vitrification solutions PVS1, PVS2, VSL, and VSL+. Samples were cooled
at 100 C/min from 25 C to 175 C and held there for 2.5 min before being warmed at 10 C/min from
175 C to 25 C. Tg glass transition temperature, Td devitrification temperature, Tm melting temperature.
PVS1 contains 22 % w/v glycerol, 15 % w/v ethylene glycol, 15 % w/v propylene glycol, 7 % w/v Me2SO, and
0.5 M sorbitol in MS medium (Uragami et al. [7]). PVS2 contains 30 % w/v glycerol, 15 % w/v ethylene glycol,
15 % w/v dimethyl sulfoxide, and 0.4 M sucrose in MS medium (Sakai et al. [8]). VSL contains 20 % w/v glyc-
erol, 30 % w/v ethylene glycol, 10 % w/v Me2SO, and 5 % w/v sucrose in 10 mM CaCl2. VSL+ contains
20 % w/v glycerol, 30 % w/v ethylene glycol, 10 % w/v Me2SO, and 15 % w/v sucrose in 10 mM CaCl2. Curves
were redrawn according to Suzuki et al. [9]
the regrowth capacity, which was comparable to PVS2 and PVS1.

VSL had a wider optimal incubation time (2045 min) than PVS2
and was more suitable for cryopreserving gentian buds.
The warming rate is thought to be far more critical than the
cooling rate for successful cryopreservation by vitrifications. The
crystallization of vitrified solution and the recrystallization of small
ice crystals formed during cooling are probably responsible for the
lethality of cells and tissues during slow warming. While the cool-
ing rates between 100 and 10 C/min are sufficiently fast to
achieve vitrification of PVS1, PVS2, VSL, and VSL+ [9], the warm-
ing rate at 10 C/min is not rapid enough to avoid the crystalli-
zation of vitrified solutions upon warming above
80 C. Calorimetric studies of the vitrified solutions using a series
of warming rates will permit one to determine the critical warming
rate for retention of the vitrified state.
4.3 Glass Transition Freeze-drying is increasingly used to preserve labile biological

of the Maximally materials including proteins, cells, and tissue specimens for various
Freeze- applications. The development of a successful freeze-drying
Concentrated System method can be a laborious and costly process. The formulation has
to be optimized to protect the biologic against stresses during
freeze and dehydration, as well as to give the best possible stability

during storage. The knowledge of the critical processing parame-
ters has to be obtained for the best retention of the desired biologi-
cal structures and functions.
The collapse temperature of the frozen samples during freeze-
drying is a critical process parameter and is traditionally determined
by DSC indirectly (see Note 4). Each freeze-drying formulation
has its own collapse temperature, depending on the type and con-
centration of excipient or protectant and biologic in the formula-
tion. The product temperature during freeze-drying has to be
maintained below the collapse temperature but as high as possible
for rapid drying. The glass transition temperature of the maximally
freeze-concentrated sample (Tg) is considered to be the upper
limit of safe product temperature during freeze-drying because the
collapse temperature is slightly higher than the Tg, typically
510 C above [10]. Tg has to be accurately measured in order to
select suitable conditions for its successful and safe processing in a
reasonable time frame.
The change in heat capacity associated with the glass transition
is small, often accompanied by an endothermic relaxation, and can
be mistaken for a melting transition. Another problem is that Tg is
dependent on the scanning rate of the measurement, and both the
cooling rate and warming rate will affect the Tg value. In order to
accurately measure the Tg of the maximally freeze-concentrated
sample for the development of a freeze-drying protocol, the cool-
ing and warming rates must be kept low, typically less than 12 C/
min. It is important to remember that Tg is not only affected by
the nature of the formulations but also by the presence of the
materials that are intended to be preserved. Therefore the Tg value
of the same protective formulation may differ when it is used for
freeze-drying of different biologics or of the same biologics at dif-
ferent concentrations. For example, when freeze-drying liposomal
formulations, changing the sucrose to lipid ratio has a large impact
on the Tg value (Fig. 8) [11].
4.4 Residual Storage stability of air-dried, freeze-dried, or vacuum-dried prod-

Moisture and Storage ucts can be investigated using DSC. The product in the amor-
Stability phous form has a Tg that can be easily determined using DSC. Other
structural information, such as melting transitions of lipids and
denaturation transitions of proteins, can also be measured. Changes
with time and storage conditions offer insight into the product
stability, i.e., how well the protectant formulation is able to main-
tain structural integrity of preserved materials.
Trehalose is among the most effective carbohydrates in pre-
serving protein structure during freeze-drying and subsequent
storage. Figure 9 shows an example where DSC was used to study
phase separation and crystallization of freeze-dried glucose/treha-
lose with preserved glucose-6-phosphate dehydrogenase during
178 Wendell Q. Sun
Fig. 8 DSC-determined Tg values of sucrose/liposome formulations with varying sucrose/lipid ratios. Curves
are redrawn according to Kett [11]
Fig. 9 DSC thermograms of freeze-dried glucose/trehalose samples with glucose-6-phosphate dehydroge-

nase after storage at 60 C. For samples stored for 12 h or longer, arrows indicate glass transitions of glucose-
rich and trehalose-rich amorphous domains. Note that the size of the sugar crystal melting peak increases
with storage time. A scan rate of 10 C/min was used (Sun and Davidson [12])
storage at an elevated temperature (60 C). Samples had a residual

moisture of 0.10 g/g dry mass and a Tg of ~13 C. DSC detected
phase separation and crystallization during storage and identified
three separate domains in stored samples (i.e., sugar crystals,
glucose-rich, and trehalose-rich amorphous domains). Phase sepa-

ration and crystallization were correlated to the loss of activity of
the preserved enzyme. This study demonstrated the impact of for-
mulations on protein stability [12].
5 Notes
1. For measurements between room temperature and subzero

temperature, 1520 L standard aluminum crimp sample cru-
cibles are a good choice. However, hermetic sample crucibles
are preferred for measurements above room temperatures to
prevent the loss of moisture and/or volatiles.
2. Purified nitrogen gas is usually used as the purge gas. Helium
may be used to improve resolution between thermal events
due to its better heat conductivity.
3. If liquid nitrogen cooling is used, check the liquid nitrogen
supply before use.
4. Freeze-drying microscopy permits one to observe the gross
structural changes associated with collapse and can be directly
used to determine the collapse temperature.
References
1. Devireddy RV, Raha D, Bischof JC (1998) 7. Uragami A, Sakai A, Nagai M, Takahashi T

Measurement of water transport during freez- (1989) Survival of cultured cells and somatic
ing in cell suspensions using a differential scan- embryos of Asparagus officinalis cryopreserved
ning calorimeter. Cryobiology 36:124155 by vitrification. Plant Cell Rep 8:418421
2. Devireddy RV, Swanlund DJ, Roberts KP, 8. Sakai A, Kobayashi S, Oiyama I (1990)
Bischof JC (1999) Subzero water permeability Cryopreservation of nucellar cells of navel
parameters of mouse spermatozoa in the pres- orange (Citrus sinensis Obs. var. brasiliensis
ence of extracellular ice and cryoprotective Tanaka) by vitrification. Plant Cell Rep
agents. Biol Reprod 61:764775 9:3033
3. Devireddy RV, Swanlund DJ, Roberts KP, 9. Suzuki M, Tandon P, Ishikawa M, Toyomasu T
Pryor JL, Bischof JC (2000) The effect of (2008) Development of a new vitrification
extracellular ice and cryoprotective agents on solution, VSL, and its application to the cryo-
the water permeability parameters of human preservation of gentian axillary buds. Plant Cell
sperm plasma membrane during freezing. Rep 2:123131
Hum Reprod 15:11251135 10. Sun WQ (1997) Temperature and viscosity for
4. Sun WQ, Wagner CT, Liversey SA, Connor J structural collapse and crystallization of amor-
(2003) Instability of frozen human erythro- phous carbohydrate solutions. CryoLetters
cytes at elevated temperatures. Cell Preserv 18:99106
Technol 1:255267 11. Kett V (2010) Development of freeze-dried
5. Sherlock G, Block W, Benson EE (2005) formulations using thermal analysis and micros-
Thermal analysis of the plant encapsulation- copy. Am Pharma Rev 13(6), September 1
dehydration cryopreservation protocol using sil- 12. Sun WQ, Davidson P (1998) Protein inactiva-
ica gel as the desiccant. CryoLetters 26:4554 tion in amorphous sucrose and trehalose matri-
6. Rall WF, Fahy GM (1985) Ice-free cryopreser- ces: effects of phase separation and
vation of mouse embryos at 196 degrees C by crystallization. Biochim Biophys Acta
vitrification. Nature 313:573575 1425:235244
Chapter 7
Measurement ofIntracellular Ice Formation Kinetics

byHigh-Speed Video Cryomicroscopy
JensO.M.Karlsson
Abstract
Quantitative information about the kinetics and cumulative probability of intracellular ice formation is
necessary to develop minimally damaging freezing procedures for the cryopreservation of cells and tissue.
Conventional cryomicroscopic assays, which rely on indirect evidence of intracellular freezing (e.g., opac-
ity changes in the cell cytoplasm), can yield significant errors in the estimated kinetics. In contrast, the
formation and growth of intracellular ice crystals can be accurately detected using temporally resolved
imaging methods (i.e., video recording at sub-millisecond resolution). Here, detailed methods for the
setup and operation of a high-speed video cryomicroscope system are described, including protocols for
imaging of intracellular ice crystallization events, and stochastic analysis of the ice formation kinetics in a
cell population. Recommendations are provided for temperature profile design, sample preparation, and
configuration of the video acquisition parameters. Throughout this chapter, the protocols incorporate best
practices that have been drawn from over a decade of experience with high-speed video cryomicroscopy in
our laboratory.
Key words Cryomicroscope, High-speed imaging, Ultra-slow motion, Intracellular ice formation,
Flashing, Nucleation, Kinetics, Cumulative probability, Cumulative hazard, Nelson-Aalen estimator
1 Introduction
The formation of ice within cells during freezing has long been
recognized as a major mode of cryoinjury [1]. In fact, a review of
experimental literature reveals a near one-to-one correspondence
between freezing processes that lead to irreversible cell damage and
those that result in intracellular ice formation [2], a correlation that
holds across a diverse range of cell types and freezing conditions: for
example, the critical cooling rate (the rate of cooling which, if
exceeded, causes the risk of intracellular ice injury to increase
above 50%) can differ by three orders of magnitude when compar-
ing the freezing response of different cell species [2], whereas the
characteristic temperature of intracellular ice formation can range
from above 5C to below 40C, depending on cell type [3].
181
182 JensO.M.Karlsson
Due to the importance of intracellular crystallization, researchers

interested in freezing injury have investigated this phenomenon by
direct microscopic observation of cells exposed to low tempera-
tures, starting with the pioneering work by Gppert [4]. Such
experimental studies have been facilitated by the development of
cryomicroscopes, optical microscopes that incorporate cooling
technology making possible manipulation of specimen tempera-
ture. Whereas the earliest cryomicroscope design was described
over a century ago [5], key advances in cryomicroscopy instrumen-
tation came in the 1970s, with the introduction of automatic tem-
perature regulation using feedback control [6], and the adoption of
then-nascent video recording technology for acquisition of micro-
graphs at sampling rates up to 60Hz [7]. The continuous record-
ing capabilities afforded by video cryomicroscopy are essential for
obtaining kinetic information about the stochastic processes that
govern the probability of intracellular freezing. Such data are valu-
able for estimating the risk of cryoinjury associated with various
freezing conditions, and can therefore be used to guide the devel-
opment of cryopreservation procedures. Furthermore, quantitative
measurements of intracellular ice formation kinetics are required to
calibrate mathematical models of this damage mechanism, by the
use of curve-fitting techniques to estimate model parameters [8
12]. In turn, such models make possible computer-aided optimiza-
tion of complex temperature profiles for minimally damaging
freezing and thawing processes [1315].
Experimental characterization of intracellular ice formation
kinetics is generally done under conditions of rapid cooling (e.g.,
>2K/s), which minimize the loss of cell water and typically lead to
intracellular freezing in ~100% of the cells observed. The benefit of
preventing freezing-induced cell dehydration is that the estimation
of thermodynamic and kinetic coefficients by model fitting will not
be confounded by changes in the intracellular solute concentrations
[8, 9, 14, 16]: if water transport is negligible, any concentration-
dependent model parameters can be treated as constants, while the
relationship between temperature and supercooling can be described
by a simple linear transformation. Unfortunately, conventional cry-
omicroscopic imaging technology cannot accurately detect the
appearance of ice within rapidly cooled cells. Historically, investiga-
tors have for the most part relied on observations of a characteristic
sudden increase in cytoplasmic opacity as indirect evidence of
intracellular ice formation in the supercooled cell. In transmitted-
light microscopy, this optical phenomenon manifests as a darkening
of the cell, which early researchers described as a flashing [1719]
or blacking out [20]. In larger cells (e.g., oocytes), the darkening
may appear to spread through the cytoplasm at speeds ranging
from ~0.1m/ms [21] to ~1m/ms [1822]. Because cells that
darken during rapid cooling typically remain dark for several seconds
or longer, it is possible to detect the opacity change using video or
High-Speed Video Cryomicroscopy 183
still micrography. However, the actual ice crystals that form inside
rapidly cooled cells cannot be detected by conventional video
imaging, whereas the familiar darkening events are due to second-
ary processes, which are not triggered until after ice has already
filled the cell [23]. Due to this decoupling of the cell darkening
reaction from the preceding intracellular ice formation event (and
the attendant time delay between the two phenomena), significant
errors in the estimation of intracellular ice formation kinetics can
result if darkening is used as a proxy indicator for the crystallization
of cell water [23].
The formation of ice crystals in rapidly cooled cells can be visu-
alized using high-speed video cryomicroscopy [23]. Ultra-slow
motion playback of high-speed video recordings reveals that when
ice crystals fill supercooled cells, there is no appreciable change in
cell opacity, whereas the advancing ice-liquid interface is readily dis-
cernible in the form of a solitary wavefront that travels through the
cell interior [23]. Because the crystal growth velocity is typically on
the order ~10m/ms, while somatic cell diameters tend to be on
the order ~10m, cryomicroscopy images must be sequentially
acquired at a temporal resolution (i.e., sampling interval) no longer
than 0.11ms to allow accurate detection of intracellular ice forma-
tion events. Thus, to advance our understanding of the mechanisms
of intracellular ice crystallization, and to measure the true kinetics
of this deleterious phase transformation process, video recordings
of cryomicroscopy experiments must use image acquisition rates in
the range 103104 frames per second (fps), or better. In contrast,
conventional analog video recording technology is constrained to
frame rates fixed at 29.97 fps (National Television System
Committee standard) or 25 fps (Phase Alternating Line standard),
yielding a temporal resolution (33 or 40ms, respectively) that is
insufficient to capture intracellular crystallization events. With digital
imaging technology, the situation is even worse: off-the-shelf solu-
tions that are commercially available at the time of this writing have
inferior temporal resolution and are not suitable for detection of
intracellular ice formation. For example, turnkey systems in which
digital image acquisition is integrated with cryomicroscope tem-
perature measurement and control include Linksys 32DV (Linkam
Scientific Instruments, Surrey, United Kingdom) and PAX-it
(Midwest Information Systems, Inc., Villa Park, IL); in the former,
the best temporal resolution for digital video acquisition is approxi-
mately 150ms [24], whereas in the latter, the shortest interval
between successive images is 10s [25].
Because imaging at sub-millisecond temporal resolution is
required to study intracellular ice formation processes, whereas
high-speed video cryomicroscopy systems are not yet commer-
cially available, this chapter presents a solution for integrating a
third-party high-speed digital video camera with an off-the-shelf
cryomicroscope stage. In addition, standard protocols for conducting
a cryomicroscopy experiment to acquire high-speed video recordings

of intracellular ice formation are described, including best-practice
techniques for analysis of kinetic data from such videos (see Note 1).
2 Materials
2.1 High-Speed 1. High-speed video camera (e.g., Phantom v7.1, Vision

Imaging System Research, Wayne, NJ, USA), including power supply and data
andAccessories transfer cable (see Note 2).
2. Camera control software (see Note 3).
3. Capture bundle cable (Vision Research) (see Note 4).
4. Pushbutton trigger switch, with BNC male terminated cable
(e.g., Part No. VRI-TRIG-PIC, Vision Research).
2.2 Microscope 1. Upright, transmitted-light microscope, with long-working-

andAccessories distance optics (e.g., DM2500 M, Leica Microsystems,
Wetzlar, Germany) (see Note 5).
2. Microscope coupling adapter for camera, F-mount interface
(e.g., Part No. D10HCF, SPOT Imaging Solutions, Sterling
Heights, Michigan, USA) (see Note 6).
3.
External light source (optional): SOLA Light Engine
(Lumencor, Beaverton, OR, USA), 3-mm-diameter liquid
light guide, collimating microscope adapter.
2.3 Temperature 1.
Temperature-controlled cryomicroscopy system (Linkam
Control System Scientific Instruments, Surrey, United Kingdom): biological
andAccessories cryostage (BCS196), system controller (T95-Linksys), liquid
nitrogen pump with 2-L dewar (LNP95), software (Linksys32)
(see Note 7).
2. Adapter clamps or plate for attaching the stage to the micro-
scope (e.g., Part No. 9670, Linkam Scientific Instruments)
(see Note 8).
3. Vacuum controller expansion board (VC95, Linkam Scientific
Instruments) (see Note 9).
4. FDCS196 stage cable (Linkam Scientific Instruments) (see
Note 9).
5.
Vacuum gauge connector, 9-pin (Linkam Scientific
Instruments) (see Notes 9 and 10).
2.4 Computer 1. Computer (see Note 11).

System andElectronic 2. Spreadsheet software (e.g., Excel, Microsoft Corporation,
Accessories Redmond, WA, USA) (see Note 12).
3. BNC male to terminal block adapter (e.g., Delock No. 65323,
Tragant, Berlin, Germany) (see Note 13).
4. BNC tee splitter, 3-way female adapter.
2.5 Tools 1. Handheld blow-dryer.

2. Bulb air blower (e.g., Rocket Blaster, Giottos Industrial,
Taipei, Taiwan).
3. Two pairs of fine-pointed forceps, angled (e.g., Fisherbrand
No. 08-875, Thermo Fisher Scientific, Waltham, Massachusetts,
USA).
4. Adjustable micropipette, 110L, with disposable tip.
5.
Vacuum tweezers (e.g., PEN-VAC Pro Series, Virtual
Industries, Colorado Springs, Colorado, USA) (see Note 14).
6. Stopwatch.
7. Flashlight.
2.6 Materials 1. Syringe, 5mL, Luer-lock.

andSupplies 2. Blunt-tipped dispensing needle, 16-G, Luer-lock.
3. Silicone grease (e.g., High Vacuum Grease, Dow Corning,
Midland, Michigan, USA).
4. Isopropyl alcohol.
5. Glass coverslips, circular, 16-mm diameter (e.g., Part No.
W16G, Linkam Scientific Instruments) (see Note 15).
6. Low-lint delicate task wipers (e.g., Kimtech Science Kimwipes
No. 34155, Kimberly-Clark, Roswell, Georgia, USA).
7. Liquid nitrogen.
8. Cell suspension, approximately 106 cells/mL (see Note 16).
3 Methods
3.1 Trigger To obtain accurate kinetic data, it is necessary to synchronize as

InterfaceCable closely as possible the independent timing signals (i.e., the clocks)
of the cryomicroscope system controller and the high-speed imag-
ing system. Cryomicroscope systems currently on the market do
not offer capabilities for communication with third-party high-
speed video cameras, which is a prerequisite for achieving such
synchronization (see Note 17). Thus, the solution described here
takes advantage of the availability of a dedicated signal channel
used for sampling pressure gauge measurements in the system con-
trollers that are designed for vacuum-capable thermal stages pro-
duced by Linkam Scientific Instruments (see Notes 18 and 19). To
synchronize the camera clock with the cryomicroscope system
clock, a trigger interface cable should be constructed as described
below (see Note 20):
1. Connect the black (signal ground) wire from the 9-pin vacuum
gauge connector to the outer shell (ground, or negative terminal)
of a BNC male connector, as well as the red (signal) wire from the
9-pin connector to the central pin (signal, or positive terminal) of

the BNC connector (see Notes 10, 21, and 22).
2. Using a three-way BNC female tee adapter, create a trigger
interface cable by joining the BNC male connectors of the
vacuum signal cable (from step 1 above), the cameras trigger
signal cable, as well as the trigger switch cable. When assem-
bled, the trigger interface cable will thus comprise three coaxial
cables: one cable leads to the 9-pin connector that will be con-
nected to the cryomicroscope system controller unit (see
Subheading3.2), another cable leads to a camera-specific con-
nector that will be joined to the camera system trigger input
socket (see step 2 in Subheading3.3), and the third cable is
terminated by the pushbutton switch in the manual trigger.
The wiring diagram in Fig.1 depicts the electrical interconnec-
tions in the trigger interface cable (see Note 22).
3.2 Instrumentation An overview of the complete high-speed video cryomicroscopy sys-

Setup tem is shown in Fig.2. This section provides instructions for setting
up the main components of the system, except for those necessary
to supply the liquid nitrogen coolant (setup of the cooling system is
outlined in Subheading3.3 below). The instrumentation setup
steps described in this section should typically be performed only
Manual trigger
pushbutton
switch
Camera capture
Vacuum
bundle cable
gauge
connector
6 B
1 P
BNC T
adapter Camera
external signal
connector
Fig. 1 Wiring diagram for the trigger interface cable. Pin 6 of the vacuum gauge
connector is connected to the cameras trigger signal input (pin B), and a com-
mon signal ground is provided by interconnecting pin 1 of the vacuum gauge
connector and pin P of the camera connector. When the pushbutton switch is
closed, the trigger signal is short-circuited to the signal ground, allowing both the
cryomicroscope control system and the camera control system to register the
resulting voltage drop
Cryomicroscope
System Controller
Trigger Interface Cable
Trigger Serial Communication

Switch Cable
Data Transfer Cable
Monitor Computer
Camera
Data Archive
Stage Cable
Insulated Capillary Stage
Instrument
Microscope Bus Cable
Nitrogen Withdrawal
Tubing (and Return)
Dewar Light Source Liquid Nitrogen Pump
Fig. 2 Main components and interconnections of the integrated high-speed video

cryomicroscopy system, including pathways for light (orange arrow), liquid and
gaseous nitrogen (blue arrows), analog electrical signals (red lines), and digital
information (green lines). See text for details (Subheadings3.2 and 3.3). For
interpretation of the references to color in this figure legend, the reader is
referred to the electronic version of the chapter
once; in contrast, the experimental setup protocol described in

Subheading 3.3 must be completed at the start of each day of
experimentation:
1. If using an external light source (see Note 23), remove the
transmitted-light lamp housing from the microscope stand,
attach the collimating adapter to the illuminator mount,
and then couple the light source output to the microscope

collimating adapter using a liquid light guide (see Note 24).
2. If not already preinstalled, install the Linkam Scientific vacuum
controller expansion card into Slot 1 on the motherboard
inside the Linkam Scientific system controller unit. Place the
system controller unit to the left of the microscope.
3. Remove the four transit screws (and corresponding washers)
from the bottom of the Linkam Scientific liquid nitrogen
pump (LNP) unit, then gently turn the unit upright, and stack
it on top of the system controller unit (see Note 25).
4. Install the following cables into the designated sockets on the
back panel of the system controller unit:
(a) Connect the FDCS196 stage cables 25-pin connector to
the socket labeled Stage (see Note 26).
(b) Connect the serial communication cable to the socket
labeled RS232.
(c) Connect the instrument bus cable to the purple socket
labeled Instrument Bus.
(d) Connect the 9-pin vacuum gauge connector of the trigger
interface cable (see Subheading3.1) to the 9-pin socket on
the vacuum controller expansion card in Slot 1.
(e) Attach the power cord to the three-prong connector.
5. Install the following cables into the designated sockets on the
back panel of the LNP unit:
(a) Connect the free end of the instrument bus cable to either
one of the two instrument bus sockets on the LNP unit.
(b) Attach the power cord to the three-prong connector.
6. Connect the free end of the serial communication cable to a
serial (COM) port on the computer, which should be p
owered
off before the connection is made.
7. Connect the system controller unit and the LNP unit to AC
power.
8. Insert the G16.3 sample carrier into the cryomicroscope
stage chamber via the access door in the right side of the stage
body; close the side door and tighten the thumbscrew to seal
the chamber.
9. Attach the microscope mounting hardware (adapter plate or
clamps) to the bottom of the cryomicroscope stage.
10. Mount the cryomicroscope stage to the microscope, using the
attached mounting hardware (see Note 27).
11. Mount the high-speed video camera to the camera port on the
microscope, using an F-mount adapter (see Notes 2830).
3.3 Experimental 1. Attach the LEMO connector (on the free end of the stage
Setup cable) to the matching socket on the left side of the cryomicro-
scope stage body.
2. Attach the trigger interface cable (see Subheading3.1) to the
cameras trigger input.
3. Connect the data transfer cable (e.g., Ethernet cable) between
the camera and the computer.
4. Power up the camera by connecting its power adapter.
5. Power up the computer (see Note 31).
6. Launch the Linksys software, and select Connect from the
File menu (see Note 32). When prompted by the software
to power on the connected Linkam equipment, first push the
LNP units power button and wait for its power indicator LED
to turn on, and then push the system controllers power but-
ton. When the power indicator LEDs of the two units are both
turned on, click OK at the software prompt (see Note 33).
7. Confirm that the stage temperature is displayed in the
Temperature Control toolbar and that a pressure value (which
should be in the range 110Pa) is displayed in the Pressure
Control toolbar; confirm that the pressure display reads No
Gauge when the camera trigger switch is closed.
8. Set the LNP control to manual mode, and set the flowrate to
zero (by pressing the downward arrow next to the Lnp box
in the Temperature Control Panel).
9. Select Temperature Profile from the View menu. Program
or open a temperature profile that specifies as its first step a tem-
perature value that is above room temperature (e.g., in the range
3040C), with an indefinite hold (e.g., set Time=99,999)
(see Note 34). Start execution of the temperature profile by
clicking the blue triangle in the Temperature Control toolbar,
and confirm that the displayed stage temperature increases to
the specified value.
10. Attach the LNP tubing to the BCS196 stage, as follows. First,
attach the nitrogen withdrawal tubing (double-walled silicone
tubing) to the coolant outlet pipe on the left side of the stage
body. Next, attach the purge tubing extension (which extends
from the T-junction in the nitrogen withdrawal tubings connec-
tor) to the stage chamber, by inserting the tubings metal fitting
into the front gas port of the BCS196 stage (see Note 35).
11. Fill the LNP dewar with liquid nitrogen, adding the liquid a little
bit at a time by pouring from a transfer vessel (see Note 36). Fill
to a level not past the beveled edge 3cm (1.25in.) below the
dewar rim, and wait until liquid nitrogen boiling subsides.
12. Cap the LNP dewar using the dewar lid, by slowly immersing
the siphon tubing into the liquid nitrogen and waiting to
latch the lid until after the liquid nitrogen boil-off has largely
subsided (see Note 37).
13. Place the filled LNP dewar to the left of the microscope, and
attach the dewars insulated capillary tubing to the coolant
inlet pipe on the left side of the BCS196 stage body. First,
thread the black siphon capillary tube into the coolant inlet
pipe, and then slide the white plastic connector over the inlet
pipe, using a twisting motion. At this point, the system con-
figuration should correspond to the schematic shown in Fig.2.
14. Using the Linksys32 software Temperature Control Panel, set
the LNP control to automatic mode.
15. Modify the first row of the temperature profile (see Note 38)
to set stage temperature to the desired sample loading tem-
perature (see Note 39), with an indefinite hold (e.g., set
Limit=37.0, Time=99,999).
16. Program the remaining temperature profile ramps required for
the freezing experiment (e.g., the profile shown in Table1 is
appropriate for mammalian cells frozen in an isotonic solution
without cryoprotectant additives), and save the programmed
profile (see Note 46).
in the Temperature Control Panel) (see Note 47).
18. Maximize clearance between the BCS196 lid and the micro-
scope objectives by using the coarse focus controls to lower the
stage (see Notes 48 and 49).
Table 1
Representative temperature profile for measurement of intracellular ice
formation kinetics
Ramp Rate Limit Time Delay

1 130 37.0a 99,999
2 60 0.0b 99,999
3 30 1.0 c
5 d

4 130 e
50.0 f
1 g

5 130 37.0a 99,999
a
Sample loading temperature (see Note 39)
b
Seeding temperature (see Note 40)
c
Equilibration hold temperature (see Note 41)
d
Equilibration hold time (see Note 42)
e
Cooling rate for freezing ramp (see Note 43)
f
End temperature of freezing ramp (see Note 44)
g
Post-freezing hold time (see Note 45)
19. Thread the LNP window tubing through the designated hole
in the window tubing clip, then attach the window tubing clip
to the stage lid, and adjust the position of the tubing outlet
(see Note 50).
20. After the appropriate length of window tubing has been pulled
through the window tubing clip, carefully detach the clip
from the cryomicroscope stage lid (without altering the length
of tubing that has been threaded through the hole in the
metal clip).
21. Unscrew and remove the BCS196 stage lid, and set it upside
down onto a clean, dry surface.
22. Use the BCS196 stage X- and Y-drives to center the sample
carrier on the silver block in the cryomicroscope chamber.
23. Visually inspect the silver block to ensure that it is clean and
dry. Dust particles can be removed using a bulb air blower.
24. Replace the cryomicroscope stage lid to minimize dust ingress
during sample preparation. It is not necessary to tighten the lid
completely.
3.4 Initial 1. Launch the camera control software and navigate to the video
Configuration ofHigh- acquisition settings interface (see Note 51).
Speed Imaging System 2. If a camera setup file has previously been saved (see step 11
below), then load the saved settings into the software (see
Note 52), and proceed to the sample preparation procedure
(Subheading3.5). Otherwise, complete the camera configu-
ration procedure described below.
3. Set the camera sensors active region, or region of interest
(ROI), by selecting the desired image dimensions (in pixels)
from the Resolution drop-down menu or by typing in a cus-
tom value and pressing Enter (see Note 53).
4. Set the camera exposure time to 100s (see Note 54).
5. Set the camera image acquisition rate (Sample rate in the
Phantom Camera Control software) to a value in the range
2,0008,000 fps (see Note 55).
6. Configure the video recording termination time by specifying
the number of frames to be acquired after the trigger event is
detected. Start by placing the trigger position at the beginning
of the recorded image range (i.e., make all frames post-trigger)
(see Note 56).
7. Determine the post-trigger recording time by dividing the
number of post-trigger frames by the cameras image acquisi-
tion rate.
8. Confirm that the available post-trigger recording duration is
approximately equal to the duration of the rapid-cooling ramp
in the temperature profile (Ramp 4in Table1), which is
computed by taking the difference between the ramps start

and end temperatures and then dividing this difference by the
cooling rate. If the post-trigger recording length is signifi-
cantly shorter than the duration of the cooling portion of the
ramp, one should decrease the image acquisition rate or the
ROI dimensions to increase the available recording duration
(see Note 57).
9. Confirm that the end of the video recording will occur before
the start of the warming ramp in the temperature profile (Ramp
5in Table1). To do so, compute the total duration of the pre-
ceding rapid-cooling ramp and isothermal hold (Ramp 4in
Table1), by taking the difference between the ramps start and
end temperatures, dividing this difference by the cooling rate,
then adding any hold time specified for Ramp 4, and, finally,
multiplying the resulting value by the conversion factor 60s/
min (to convert the ramp duration to seconds). The total dura-
tion thus calculated should be at least ~1s longer than the
post-trigger recording duration (from step 7 above). If neces-
sary, increase the image acquisition rate or the ROI dimensions
to decrease the total recording duration, or increase the hold
time associated with Ramp 4 (Table1) to extend the ramp
duration (see Note 58).
10. Iteratively repeat steps 79 above until the video recording
interval is properly synchronized with the rapid-cooling ramp.
11. Save the final camera configuration as a camera setup file
(see Note 59).
3.5 Sample 1. Create a grease dispenser by removing the plunger from a

Preparation 5-mL Luer-lock syringe, dispensing ~2-mL silicone grease into
the syringe barrel, and then reinserting the plunger into the
barrel (see Note 60). Attach a blunt-tipped Luer-lock dispens-
ing needle (see Notes 61 and 62).
2. Place a 16-mm-diameter circular coverslip onto a clean flat sur-
face, and stabilize it by applying a downward force using the
tips of a clean pair of angled forceps.
3. Push the syringe plunger until a small amount (<0.5mm) of
grease protrudes from the tip of the dispensing needle.
4. Apply a small (<0.5mm in diameter) dot of silicone grease
approximately 2mm away from the coverslip edge, by daub-
ing the surface of the coverslip with the dispensing needle tip
(see Note 63).
5. Continue to apply a total of approximately 20 such silicone
grease daubs to the coverslip, evenly spaced along a circle
~12mm in diameter (always ~2mm away from the coverslip
edge), as shown in Fig.3a. Apply pressure to the syringe plunger
to advance more silicone grease as necessary (see Note 64).
a b
GR GR
CS
CC CC
FT CC FT
WS CC
d e
FT FT FT FT
CC CC
FT FT FT FT
Fig. 3 Initial steps of the sample preparation procedure. (a) Top view, depicting
~20 small daubs of silicone grease (GR) that are applied along the perimeter of
a 16-mm-diameter circular coverslip (CC). (b) After smearing the grease dots
together to form a ring, a droplet of cell suspension (CS) is pipetted onto the
center of the coverslip. After gently placing a second circular coverslip on top of
the sample, the alignment of the two coverslips can be corrected by applying
lateral forces using two pairs of forceps (ce). (c) Side view (not to scale), depict-
ing the coverslip sandwich and forceps tips (FT) in contact with the work surface
(WS); the left pair of forceps is in contact with the bottom coverslip, whereas the
right pair of forceps is in contact with the top coverslips overhang, and both pairs
of forceps are moved slowly toward each other (arrows). (d) Top view, showing
the placement of the tips of the two pairs of forceps at the start of the coverslip
alignment procedure. (e) The alignment procedure is complete when the two
circular coverslips are concentric. The remainder of the sample preparation pro-
tocol is described in Subheading3.5
6. Use the tip of the dispensing needle to smear the grease dots
together, to form a continuous grease circle as shown in Fig.3b
(see Note 65).
7. Using a micropipette, dispense a 4-L drop of cell suspension
onto the coverslip surface in the center of the grease circle
(see Fig.3b).
8. Using forceps or vacuum tweezers, hold a second 16-mm cir-
cular coverslip 23mm directly above the first coverslip, so
that the two coverslips are parallel to each other and concentri-
cally aligned. Release the top coverslip and allow it to drop
onto the bottom coverslip.
9. If necessary, align the two coverslips to each other by applying
horizontal forces with two sets of angled forceps, as shown in
Fig.3ce (see Note 66).
10. Using the tips of a pair of forceps, gently push down on the top
coverslip to simultaneously spread out the liquid sample and
seal the grease ring. Stop when the liquid has spread over a
~1-cm2 area, just making contact with the inner edge of the
grease ring (see Notes 67 and 68).
11. Place the coverslip sandwich onto a Kimwipe tissue. Without
applying excessive pressure (which can squeeze liquid out of the
sample), dry the top and bottom surface of the sample. In addi-
tion, apply the Kimwipe tissue along the entire perimeter of the
coverslip sandwich preparation, to draw out (by capillary action)
any liquid that has escaped the grease barrier (see Note 69).
3.6 Sample Loading 1. Remove the cryomicroscope stage lid, and set it upside down
onto a clean, dry surface.
2. Use vacuum tweezers to transfer the sample into the cryomi-
croscope stage chamber, carefully dropping it (from a height
<5mm) into the G16.3 sample carrier ring on the silver block
of the BCS196 stage (see Note 70).
3. Without delay, start recording of temperature data by clicking
the red triangle in the Data Capture toolbar (see Note 71).
4. Replace the cryomicroscope stage lid, and screw it tight to seal
the chamber.
5. Purge the air from the cryomicroscope stage chamber using
dry nitrogen gas, by completing the procedure described in
steps 612 below (see Note 72).
6. Manually increase the liquid nitrogen flowrate to its maximum
value (by pressing the upward arrow next to the Lnp box in
the Temperature Control Panel, until Lnp=30) (see Note 73).
7. Insert the supplied gas valve fitting (barbed end facing out-
ward) into the gas port on the back of the BCS196 stage, until
the valve clicks open.
8. Direct gas flow through the purge tubing (which was previously
attached to the front gas port, in step 10 of Subheading3.3),
by blocking flow through the LNP window tubing (pinching
the tubing shut between the thumb and index finger of the
left hand) (see Note 74) while simultaneously blocking the
exhaust outlet in the plastic connector of the LNP nitrogen
withdrawal tubing (using the middle finger of the left hand).
Continue to block both gas exhaust paths until step 11 below
(see Note 75).
9. Using the right hand, reach behind the cryomicroscope stage
body and block the rear gas valve with a finger, to pressurize
the chamber gas; after 12s, lift the finger from the rear gas
port, and allow the gas to flow out for approximately 510s.
Repeat this process periodically for 12min, alternatively pres-
surizing and releasing the chamber gas (see Note 76).
10. While continuing to block the LNP tubing connector exhaust
outlet and the window tubing, remove the valve opening con-
nector from the rear gas port by pushing the valve sleeve
toward the stage body until the valve clicks shut.
11. Release the window tubing and stop blocking the LNP tubing
connector exhaust outlet.
12. Set the LNP control to automatic mode (see Note 77).
3.7 Freezing 1. Attach the window tubing clip to the stage lid, and adjust the
Experiment position of the tubing outlet (see Note 50).
2. Turn on the microscope illuminator (see Note 78).
3. If the cryomicroscope stage body was laterally repositioned
during the experimental setup procedure (see Note 48), then
the stage body should now be re-centered until light can be
observed to pass through the BCS196 silver block aperture
(see Note 49).
4. Configure the microscope for bright-field imaging, and move
a low-power (e.g., 5) objective into the microscopes optical
path (see Note 79).
5. Bring the specimen into focus (relative to the camera live
image, not the eyepieces), and then configure Khler illumina-
tion (again using the camera live image) (see Note 80).
6. If the BCS196 silver block aperture is not centered in the cam-
era image, make the necessary fine adjustments to the cryomi-
croscope stage body position (see Note 81).
7. Using the microscopes coarse focus controls, lower the
mechanical stage as far as possible without disturbing the con-
denser position.
8. Modify the first row of the temperature profile in the Linksys
software (see Table1), by changing the entry in the Time
column of Ramp 1 from a value of 99,999 to a value of 0, and

then press Enter (see Note 82). Confirm that the displayed
stage temperature decreases as the stage is cooled to the seed-
ing temperature (see Note 40).
9. When the seeding temperature has been reached, use the cryo-
microscope stage sample carrier X- and Y-drives to position the
sample edge over the BCS196 seeding cold spot (see Note 83).
10. Set the LNP control to manual mode, and manually increase
the liquid nitrogen flowrate to its maximum value (by pressing
the upward arrow next to the Lnp box in the Temperature
Control Panel, until Lnp=30). Simultaneously start a stop-
watch to time the seeding process.
11. Continuously monitor the cold-spot area through the BCS196
stage lid window. When ice formation is observed at the sam-
ple edge (in the sample region that was resting on the cold
spot), quickly perform the actions described in steps 1215
below (see Note 84).
12. Stop the stopwatch that was started in step 10 above (see
Note 85).
13. Switch the LNP control to automatic mode.
14. Turn the cryomicroscope stage X- and Y-drives counterclock-
wise to move the sample away from the cold spot, positioning
the sample carrier so that the entire sample rests on the main
(cylindrical) silver block.
15. Advance the execution of the temperature profile (see Table1),
by changing the entry in the Time column of Ramp 2
from a value of 99,999 to a value of 0, and then press Enter
(see Note 86).
16. The microscope and imaging system manipulations that are
described in steps 1726 below must be completed before the
end of the equilibration hold time (see Note 42) that is speci-
fied in Ramp 3 of the temperature profile (e.g., 5min in the
example in Table1). If there is insufficient time to complete
these actions, program a longer equilibration hold time in the
temperature profile.
17. Move the desired microscope objective into the optical path.
18. Use the microscope focus controls to bring the specimen into
focus in the live camera image.
19. Use the BCS196 sample carrier X- and Y-drives to move a region
of interest in the specimen into the field of view (see Note 87).
20. Optionally, acquire fluorescent images (e.g., to identify live
and dead cells in the field of view), and then switch the micro-
scope optics back to bright-field illumination.
21. Make any necessary adjustments to the fine focus and to the
Khler illumination (see Note 88).
22. Adjust the condenser iris to achieve an adequate balance
between image contrast, brightness, and optical resolution
(see Note 89).
23. Make any necessary adjustments to the camera exposure, image
size, and other settings in the camera control software.
24. Calibrate the camera black level (see Note 90).
25. Start capturing video into the cameras circular memory buf-
fer. The camera should now be waiting for the trigger signal.
26. Pick up the trigger switch, and be ready to push the trigger
button during the next step.
27. Simultaneously monitor the Linksys stage temperature display
and status indicators, as well as the camera live image. When
the Linksys software automatically advances the temperature
profile to the rapid-cooling ramp (Ramp 4in Table1), push
the trigger button (see Notes 91 and 92).
28. When the high-speed video recording is completed, click the
blue square button in the Linksys software Data Capture tool-
bar to stop recording temperature data. When prompted, spec-
ify a directory and name for the Linksys data (.iml) file, and
click Save (see Note 93).
29. Turn off the microscope illumination.
30. Save the acquired video onto the computer hard drive. If any
still images were captured during the experiment, save these as
well (see Note 94).
in the Temperature Control Panel) (see Note 47).
33. Detach the window tubing clip from the BCS196 stage lid,
and then unscrew and remove the lid. Set the lid upside down
onto a clean, dry surface.
34. Use the BCS196 stage X- and Y-drives to center the sample
carrier on the silver block in the cryomicroscope chamber. Use
vacuum tweezers to remove the previous sample from the sil-
ver block (see Note 95).
35. Visually inspect the silver block to ensure that it is clean and
dry. Dust particles can be removed using a bulb air blower.
36. Replace the cryomicroscope stage lid to minimize dust ingress.
It is not necessary to tighten the lid completely.
37. If additional experiments are planned, reload the previously

saved temperature profile (Table1) into the Linksys software
(see Note 96). When saving of the current video has completed
(see Note 94), repeat the above protocols starting with the
sample preparation procedure (Subheading3.5).
3.8 Experiment 1. Set the stage temperature to a value in the range 3040C, by
Shutdown modifying the active row of the temperature profile table in
Linksys (see Note 38); specify an indefinite hold (e.g., set
Time=99,999).
3. Detach the LNP purge tubing extension from the BCS196
front gas port by pushing the valve sleeve toward the stage
body until the valve clicks shut.
4. Disconnect the LNP liquid nitrogen withdrawal tubing from
the coolant exhaust pipe on the BCS196 stage (see Note 97).
5. Disconnect the dewars insulated capillary tubing from the
coolant inlet pipe on the BCS196 stage (see Notes 97 and 98).
6. Terminate control of the stage temperature, by clicking the
blue square button in the Temperature Control toolbar in the
Linksys interface to stop execution of the temperature profile.
7. Select Disconnect from the File menu; confirm that the
temperature display reads No Comm.
8. Exit the Linksys32 software, and then power down the liquid
nitrogen pump and the cryomicroscope system controller.
9. Unclasp the latches of the lid on the LNP dewar, and gently
remove the lid. If necessary, discard any leftover liquid nitro-
gen using appropriate disposal methods.
10. Detach the LNP window tubing from the BCS196 stage lid,
and remove the metal clip from the end of the tubing.
11. Disconnect the LEMO connector from the socket on the left
side of the BCS196 stage body, by pulling it straight out.
12. When all image and videos from the experiments have been
saved, close the camera control software and shut down the
computer.
13. Power down the camera by disconnecting its power adapter.
14. Disconnect the data transfer cable and the trigger interface
cable from the camera.
15. Optionally, detach the camera from the microscope; immedi-
ately cap the camera sensor and the microscope camera port to
prevent dust ingress.
16. Optionally, disconnect the BCS196 stage body from the

microscope stand, and remove the mounting hardware from
the stage. Return the BCS196 stage, as well as the valve
opening connector and the window tubing clip to their
storage case.
3.9 Data Analysis To analyze the kinetics of the mechanisms that cause intracellular
ice crystallization, the cumulative hazard of intracellular ice forma-
tion can be computed using the Nelson-Aalen estimator, a versatile
nonparametric method that can account for competition between
multiple mechanisms, as well as data censoring [12, 26].
Nonetheless, in the procedure described below, no attempt is made
to distinguish between different mechanisms of intracellular ice
formation (i.e., only the total rate of intracellular ice formation is
determined), and data censoring is treated only in a rudimentary
manner. It is assumed that one or more replicate cryomicroscopy
experiments were performed and that high-speed video recordings
were acquired starting from the beginning of the rapid-cooling
ramp (e.g., Ramp 4in Table1) and ending before the beginning
of the warming ramp (e.g., Ramp 5in Table1):
1. In every video, display the first post-trigger frame (see Note 99),
and assign each cell a unique identifying label. Reviewing the full
length of each video (starting from the first post-trigger frame),
if any of these cells becomes obscured at any point during the
experiment (e.g., by drifting out of the field of view or out of
focus), omit such cells from further analysis (see Note 100).
2. For each remaining cell, using slow-motion and frame-by-frame
playback, find the video image in which intracellular ice crystal-
lization is initiated (i.e., the first frame in which any evidence
of intracellular ice can be seen) (see Notes 101 and 102).
Record the timestamp of each such video frame (with time
measured relative to the time at which the camera detected the
falling edge of the trigger signal).
3. For each video, also go to the first post-trigger frame and to
the final video frame, and record the corresponding time-
stamps (with time measured relative to the time at which the
camera detected the falling edge of the trigger signal).
4. Using the Linksys software, convert each saved temperature
data file from its native format (.iml) to a text file with comma-
separated values (see Note 103), then open the resulting file
using any standard spreadsheet or word processing software,
or optionally, import the data into a computational software
platform such as MATLAB (see Note 12).
5. For each video, use the following interpolation technique to
estimate the cryomicroscope stage temperature associated with
the timestamps collected in steps 23 above. In the text file
containing the Linksys measurements, the data are arranged in

rows containing three values each; if imported into a spread-
sheet program, these will be allocated to three columns. The
first column represents the time (in seconds), relative to the
start of data logging (see step 3 in Subheading3.6); the second
column represents the stage temperature (in Celsius); the third
column contains data from the vacuum controller expansion
card (used here to monitor the trigger signal; see Note18):
(a) Start by identifying the row in which the vacuum card
value in the third column first drops to zero (or to a value
near zero): this is the ttrigger event, and the corresponding
time value (first column) will be designated ttrigger in what
follows.
(b) If necessary, convert the timestamp value for the video
frame of interest to units of seconds (see Note 104); the
corresponding value (in seconds) will be designated tframe.
(c) Compute the interpolated time value (designated tint) by
adding ttrigger and tframe:
t int = t trigger + t frame

(d) In the first column of the Linksys data file, find the pair of
time successive values (designated tk and tk+1) that span the
interpolated time value (i.e., such that tktinttk+1). Make
a note of the temperature datum (in the second column)
corresponding to time tk; this temperature will be denoted
Tk. Likewise, the quantity Tk+1 designates the temperature
value corresponding to the subsequent timepoint tk+1 in
the Linksys data file.
(e) Compute the interpolated temperature value (designated
Tint) using the following formula (see Note 105):
(Tk +1 - Tk ) (t int - t k )
Tint = Tk + .
t k +1 - t k

6. Among the interpolated temperature values corresponding to
each videos trigger event (see step 3 above), identify the low-
est temperature (which corresponds to the video with the lat-
est trigger time). Any isolated intracellular ice formation events
occurring at higher temperatures (i.e., earlier) in any of the
other videos should be excluded from analysis (see Note 106).
7. Count the total number of cells in each video, excluding those
cells that were disqualified in steps 1 and 6 above (see Note 107).
This is the initial size of the so-called risk group.
8. Among the interpolated temperature values corresponding
to video end frames (see step 3 above), identify the highest
temperature (which corresponds to the earliest end time).

Kinetics of intracellular ice formation will not be estimated
past this temperature (see Note 106).
9. Make a list of all interpolated intracellular ice formation tem-
peratures (from all videos) that fall between the temperatures
identified in steps 6 and 8 above (inclusive). Using a spread-
sheet program (or any computational software with similar
functionality), enter these values into a data column and sort
the temperatures in descending order (i.e., in order of increas-
ing time along the temperature profile), as shown in Table2.
10. Adjacent to the temperature column from the previous step,
create a new column to contain values of the risk group size.
Next to the first (highest) intracellular ice formation tempera-
ture, enter the initial size of the risk group (from step 7 above).
For each subsequent intracellular ice formation temperature,
decrement the risk group size by one as shown in Table2 (see
column labeled R) (see Note 109).
11. Adjacent to the risk group size column from the previous step,
create a new column to contain values of the incremental con-
tributions to the Nelson-Aalen estimator. For each row, the
value in this column should equal the inverse of the corre-
sponding risk group size (see column labeled A in Table2).
12. Adjacent to the previous column, create a new column to
contain values of the Nelson-Aalen estimator of cumulative
hazard. Fill the rows of the new column by computing a run-
ning sum of the values in the previous column, as shown in
Table2 (see column labeled A).
Table 2
Sample calculation of intracellular ice formation kinetics using the
Nelson-Aalen estimatora
Tb Rc Ad Ae Pf
7.250 5 0.2000 0.2000 0.1813
11.123 4 0.2500 0.4500 0.3624
11.124 3 0.3333 0.7833 0.5431
13.768 2 0.5000 1.2833 0.7229
a
In this hypothetical example, it has been assumed that four intracellular freezing events
were observed within the period of analysis, whereas the initial risk group comprised five
unfrozen cells (see Note 108)
b
Interpolated intracellular ice formation temperatures (in C) for all events not
disqualified
c
Risk group size (number of unfrozen cells just prior to the recorded intracellular ice
formation event)
d
Incremental contribution to the Nelson-Aalen estimator of cumulative hazard (A=1/R)
e
Estimated cumulative hazard of intracellular ice formation
f
Cumulative probability of intracellular ice formation (P=1eA)
13. If desired, the cumulative probability of intracellular ice formation

(P) at each temperature can now be computed from the esti-
mated cumulative hazard values (A), using the following
transformation:
P = 1 - e- A
An example of this calculation is shown in Table2 (see column

labeled P) (see Note 110).
14. It is often informative to plot the estimated cumulative hazard
(A) as a function of descending temperature. If the slope of the
resulting correlation (which can be estimated by regression
analysis or by finite differencing algorithms) is multiplied by
the cooling rate for the experiment, then the product will rep-
resent the hazard rate of the intracellular ice formation process
(i.e., the expected number of intracellular freezing events per
unit time, per unfrozen cell). This quantity can be interpreted
as a nucleation rate, for example.
4 Notes
1. This chapter is written with the assumption that the reader

already has prerequisite knowledge and skills pertaining to the
use of the individual instruments that are described herein,
including the configuration and operation of their optical
microscope and high-speed imaging camera, as well as a basic
familiarity with the cryomicroscope stage, temperature con-
trol system, and software. Moreover, this chapter does not
address all safety concerns that may be relevant to the use of
the protocols (including, but not limited to: liquid nitrogen
safety precautions, biosafety precautions, optical radiation
safety precautions, as well as general selection and use of per-
sonal protective equipment). It is the readers responsibility
to obtain the necessary safety training and to take all precau-
tions required to ensure personal safety and compliance with
applicable safety regulations.
2. There are a large number of high-speed video camera systems
available on the market, at various price points. Minimum
requirements for a camera system include continuous image
acquisition rates of at least 2,000 fps (preferably up to 8,000
fps) at a reduced sensor resolution no smaller than 128128pix-
els (preferably at full resolution). Furthermore, the imaging
system must incorporate a circular buffer architecture with at
least 2GB of memory and the ability to configure the timing of
the trigger event that controls termination of image acquisi-
tion; the camera must accept an external trigger signal for this
purpose. In addition, the frame rate, image resolution, and

exposure time should be user adjustable. The cameras light
sensitivity and signal-to-noise ratio should be sufficient to yield
publication-quality micrographs at exposure times (i.e., shutter
speeds) of 100s; this must be confirmed by imaging tests
using the actual cryomicroscope setup on which experiments
will be performed, including a representative cell sample.
3. High-speed imaging systems typically include a dedicated
software package for configuration and control of the camera.
Older Vision Research Phantom camera models, such as the
v7.1 model used here, require a pre-2010 implementation of
the Vision Research Phantom Camera Control application,
which is referred to as the legacy version of the software. In
contrast, more modern Phantom camera models use a post-
2010 version of the Vision Research Phantom Camera
Control program, which is generally referred to as the PCC
software. In this chapter, distinct instructions are provided for
both versions of the software.
4. The capture bundle cable is a breakout cable that provides
access to the cameras hardware trigger input (as well as other
signal connectors), typically via a BNC connector. The required
capture cable hardware may be different for different camera
models. In the Vision Research Phantom v7.1 camera used
here, the capture cable joins a 19-pin Amphenol military stan-
dard circular connector to five BNC male c onnectors (which
correspond to the cameras trigger, F-sync, pre-trigger,
IRIG-in, and analog video signals, respectively); this capture
bundle cable is supplied with the v7.1 camera system. In other
Vision Research camera systems, access to the trigger signal
may require a separately purchased breakout box (e.g., Part
No. VRI-BOB-2 for the Phantom v7.3 camera) or a different
style of capture cable (e.g., for the Vision Research Miro-series
cameras). In the most recent v-series cameras from Vision
Research, the trigger signal connector is available directly on
the back of the camera chassis via a BNC female connector; for
these cameras, a standard BNC male to BNC male coaxial
cable can be used in place of the capture bundle cable.
5. The microscope used here is a Leica DM2500M stand with a
5/0.12 NA, 14.0-mm working-distance objective (Part No.
11566077), a 20/0.40 NA, 6.9-mm working-distance objec-
tive (Part No. 11506242), as well as a universal condenser
(Part No. 11551062) with a 0.50-NA, 15-mm working-
distance top lens (Part No. 11501037); the microscope system
also includes a 100-W lamp housing (Part No. 11504080), a
photo tube (Part No. 11551511), and a mechanical stage with
XY-drive (Part No. 11888705). A wide variety of microscope
brands and models can be used. Some minimum requirements
include the availability of stage clamps from Linkam Scientific

Instruments that are compatible with the microscope, the avail-
ability of objectives with a free working distance of at least
4.5mm, and a condenser with a free working distance of at
least 12.7mm (although Linkam Scientific Instruments also
makes available condenser extension lenses for many common
condenser models). The microscope must also include a cam-
era port and light source capable of generating ~3,000 lumen
(e.g., a 100-W, 12-V tungsten-halogen lamp).
6. The adapter should be selected to match the microscope camera
port and the camera lens mount. Although C-mount hardware
is available for Vision Research Phantom cameras, the F-mount
interface is preferable because it causes less vignetting.
7. The system includes a complete set of cables, including power
cables, a serial communication cable, an instrument bus cable,
and a BCS196 stage cable. However, to establish communica-
tion with the vacuum board, the BCS196 stage cable will not be
used; instead, the BCS196 stage must be connected to the sys-
tem controller via a FDCS196 stage cable, purchased separately.
8. Stage clamps should be selected to match the microscope.
Linkam Scientific Instruments Part No. 9670 is designed to
fit a Leica DM2500M microscope with a mechanical XY-
stage (Part No. 11888705).
9. This item is available from Linkam Scientific Instruments as
part of a T95 external trigger kit that includes a VC95 vac-
uum controller expansion card, a 9-pin vacuum gauge con-
nector (with pre-wired leads), as well as an FDCS196 stage
cable [24]. All of these items are required to implement the
synchronization scheme described in this chapter.
10. The 9-pin vacuum gauge connector obtained from Linkam
Scientific Instruments must be used, because it contains an inte-
grated circuit (wired to pins 4, 5, and 9) that allows the cryomi-
croscope system controller to communicate with the connector.
The connector supplied by Linkam Scientific Instruments is also
pre-wired with a signal lead (red) that connects to pin 6, and a
signal ground lead (black) that connects to pin 1.
11. Current versions of the Vision Research PCC software and the
Linkam Scientific Linksys32 software require the use of
Microsoft Windows operating systems from Windows XP up
to and including Windows 7. Minimum hardware require-
ments include a Pentium-class microprocessor (1.7GHz or
higher), 4GB of memory, and at least 80GB of disk space.
The computer must have at least one free serial (COM) port,
as well as a 10-GB Ethernet adapter. It is recommended that
the computer be outfitted with a system for efficiently archiving
image data (e.g., a tape drive or a disk array system).
12. Alternatively, one may use any general-purpose numerical soft-

ware platform (e.g., MATLAB, The MathWorks, Natick, MA).
13. It is possible to use any BNC male connector, but a BNC to
terminal block adapter simplifies assembly of the trigger inter-
face cable, because no soldering or crimping is necessary.
14. The Linkam Scientific Instruments temperature-controlled
microscopy stages ship with a set of vacuum tweezers pow-
ered by a motorized vacuum pump. The vacuum tweezers
supplied by Linkam Scientific Instruments are adequate for
moving samples to and from the cryomicroscope stage as
called for in the protocols, whereas a self-contained pen-
style vacuum pickup tool has the advantage of not being teth-
ered to a pump via tubing.
15. Circular coverslips supplied by Linkam Scientific Instruments
are rated as 0.17mm thick (No. 1.5 thickness), whereas No. 1
coverslips are also acceptable (0.13-mm thickness). In select
countries, 16-mm-diameter circular glass coverslips can also be
sourced from various distributors of laboratory supplies (e.g.,
Part No. 631-0152, VWR International, Radnor, PA, USA).
16. If the cell concentration is lower than 105 cells/mL, cells will
be difficult to locate within the sample, and typically, the field
of view will contain at most a single cell. On the other hand,
if the cell concentration exceeds 107 cells/mL, then the cells
will become crowded within the field of view (with a mean
separation of ~50m or less), with a high propensity for
aggregation (which may affect the intracellular ice formation
kinetics). The sample temperature profile shown in Table1
assumes that the suspending medium is isotonic (~300mOs-
mol/kg); if the solution contains cryoprotectant additives,
the profile temperatures must be modified.
17. Linkam Scientific Instruments control units starting with
model T95 do incorporate a programmable high-impedance
input port, as well as two programmable open-drain output
ports, which are designed to offer communication capabili-
ties with external devices. However, these features are not
enabled in the most recent release of the Linksys32 software
as of the date of this writing (version 2.3.0). According to a
representative of Linkam Scientific Instruments, a new ver-
sion of the Linksys software, due to be released before the
end of 2014, will include capabilities to communicate via the
T95 input/output ports [24]. The synchronization tech-
niques described in this chapter can easily be adapted for use
with the system controllers programmable input port, after
the functionality of this communication channel has been
enabled in the upcoming software release.
18. System controllers supplied with vacuum-capable temperature-
controlled stages, including the FDCS196 freeze-drying sys-
tem and the THMS350V vacuum system, are outfitted with a
vacuum controller expansion board that is capable of reading

an analog 210-V signal produced by a Pirani vacuum gauge.
The system controller samples this signal every 50ms, along
with the stage temperature data, and converts the sensor volt-
age (V, in units of V) to a pressure value (p, in Pa) according
the exponential relationship p=10V4 [24]. In versions of the
Linksys control software available to date, the digitized tem-
perature and pressure measurements are downsampled to an
effective sampling rate of ~3Hz before being written to a
data file on the computer hard drive [24]. The approach
described here for synchronizing the Linksys clock and cam-
era clock is to generate a trigger signal that will be simultane-
ously detected by both systems. In particular, a TTL-level
signal, which is pulled down from a high state (~45V) to a
low state (~0V) by the closure of a trigger switch, is con-
nected to the input channel of a vacuum controller expansion
board installed in the T95. As a result, at the moment the
trigger switch is closed, the pressure values recorded in the
Linksys data file will exhibit a drop from a steady nonzero
value (in the range 110Pa) to a value of 0.0Pa. The time of
this apparent pressure drop provides a common reference
point that can be used to synchronize the cryomicroscope
system clock and the video camera clock during data analysis
(as described in Subheading3.9).
19. As a result of multiplexing of the cryomicroscope system con-
troller data channels [24], the actual arrival time of the falling
edge in the trigger voltage may be anywhere from 88ms
before to 12ms after the timestamp associated with the cor-
responding pressure drop recorded in the Linksys data file. In
addition, the real temperature sampling time may be any-
where from 38ms before to 12ms after the corresponding
Linksys timestamp. Timestamp discrepancies in the Linksys
temperature and pressure data affect the synchronization
algorithm in opposite directions, yielding a net synchroniza-
tion error in the range 50 to +100ms. Furthermore, the
resulting temperature estimation error is the product of the
net timing error and the rate of cooling. Accordingly, for a
cooling rate of 150C/min, the interpolated temperature Tint
(see step 5 in Subheading3.9) may be as much as 0.25C
lower than the actual temperature of the observed event or up
to 0.12C higher than the actual temperature.
20. The specific instructions provided here are for a Vision Research
Phantom v7.1 camera, in which the external trigger mechanism
is implemented using a dry contact that supplies an isolated
5-V source via a 470- current-limiting resistor. Thus, by wir-
ing a switch between this trigger input and the cameras iso-
lated signal ground, closure of the switch will pull the trigger
signal low; the falling edge of the resulting voltage drop is

detected by the camera electronics within 3s and timestamped.
The trigger input contact corresponds to pin B of the cameras
19-pin Amphenol military standard circular connector and can
be accessed using the capture bundle cable available from
Vision Research (in which the trigger signal and signal ground
are routed from the 19-pin connector to the central pin and
shell, respectively, of a BNC male connector at the end of a
breakout cable). If using other camera models with different
trigger mechanisms, it may be necessary to modify the design
of the trigger interface cable. If the instructions in the protocol
are modified, it is recommended that the manufacturer of both
the camera and the cryomicroscope controller unit be con-
sulted, to avoid damage to these instruments.
21. To extend the reach of this vacuum signal cable, one may
terminate the red and black wires using a BNC female con-
nector, instead of the male connector specified in the proto-
col. This makes it possible to use a standard (male-to-male)
BNC coaxial cable as a convenient extension.
22. If ones high-speed imaging system does not use BNC connec-
tors for the trigger signal, then the electrical connections
described here should be made using appropriate adapters and
connectors.
23. Incandescent 100-W, 12-V tungsten-halogen lamps, such as
those commonly fitted in the standard lamp housing that is
furnished with research-grade optical microscopes, emit ~3,000
lumen of luminous flux. When using a high-speed video cam-
era that has adequate light sensitivity, the brightness of such
halogen bulbs is sufficient for routine detection of intracellular
ice formation in suspended or attached cells, and an external
illuminator is not necessary. However, for special applications
in which higher light intensity is required, one may replace the
microscope lamp housing by an external light source (e.g., the
Lumencor SOLA, which can supply illumination that is approx-
imately twice as bright as that of a standard 100-W halogen
lamp). If the original microscope lamp is replaced by a brighter
light source, it is important to pay critical attention to optical
radiation safety, because exposure to high-intensity light may
cause serious personal injury (e.g., vision loss). The risks are
even higher if using a metal-halide lamp as an external illumina-
tor, because gas-discharge lamps emit high levels of ultraviolet
radiation. To mitigate such dangers, it is recommended that
the microscope eyepieces be permanently removed and the
ocular tubes capped (so that the operator can view a specimen
only via the live image from a camera).
24. Liquid light guides are fragile and if stressed may form bub-
bles in the liquid core, which can significantly reduce light
transmission. In particular, liquid light guides should always

be installed with sufficient slack and not be allowed to form
any bends with a radius of curvature less than ~10cm.
25. The transit screws are marked by small yellow labels on the bot-
tom of the LNP unit. These four screws are installed for ship-
ping, to secure the shock-mounted pumps. When the transit
screws have been removed, the LNP should be handled gently
and protected from large accelerations that can damage the
anti-vibration mounts. The screws and washers can be fastened
to the four holes provided for storage on the units back panel.
26. It is important to use the FDCS196 stage cable, and not the
original BCS196 stage cable. Only if using the FDCS196
cable will the Linksys software be able to register the trigger
signal that is input via the vacuum controller card.
27. Alternatively, if one does not want the cryomicroscope stage
to be permanently installed on the microscope, this step can be
performed at the beginning of the experimental setup proce-
dure (i.e., before completing any of the steps in Subheading3.3).
28. Alternatively, if one does not want the camera to be perma-
nently installed on the microscope, this step can be performed
at the beginning of the experimental setup procedure (i.e.,
before completing any of the steps in Subheading3.3).
29. If the camera body or the microscope camera port cannot be
configured to accept F-mount hardware, a C-mount adapter
can be used as an alternative. Nonetheless, because high-
speed imaging sensors typically have large dimensions, an
F-mount is preferable, because it provides a wider field of
view (and thus prevents vignetting).
30. Depending on the weight of the camera model used, the
mechanical strength of the microscope stand, and the configura-
tion of the camera port and coupling adapter, it may be necessary
to support the weight of the high-speed camera (to prevent
undue mechanical stress on the microscope components). For
example, a tripod or a scissor jack platform may be used to coun-
teract the forces and torques resulting from the camera weight.
31. Confirm that there is sufficient storage space on the hard drive to
hold the video files that will be generated. If there is insufficient
space on the computer hard drive, delete any videos from previ-
ous experiments, after first backing up the files to a data archive.
32. Before establishing a connection to the system controller the
first time, software settings should be configured as described
in the Linksys32 User Guide. In particular, go to Temperature
Controller under the Setup menu, select the
TempControl entry in the navigation tree, and then specify
the computer COM port number to which the controller
units serial communication cable was connected (in step 6 of
Subheading 3.2). In the same setup window, check the box

labeled Use profile, and then close the window. Next, right-
click in the Data Capture toolbar to bring up the Capture
Setup window, and set the option Control from capture
control start stop, and then close the setup window.
33. The sequence is important. The system controller must be
powered up after the LNP is turned on, or else the controller
will not recognize the pump. The OK button in the soft-
ware prompt must not be clicked before the hardware is pow-
ered up, or else the software will not be able to establish
communication with the controller unit.
34. The purpose of raising the stage temperature above room tem-
perature is to prevent condensation when the liquid nitrogen
is attached. For instructions on profile programming, refer to
the Linksys32 User Guide. After changing any value in the
temperature profile table, always use the Enter key, or alter-
natively, use the mouse to click in a different table cell; this
ensures that the system controller registers the updated value.
35. Push the fitting in firmly until the valve clicks open. If prop-
erly inserted, it should not be possible to pull the fitting out
of the gas port.
36. Inspect the dewar before adding liquid nitrogen: the dewar
interior must be free of particulates, condensation, frost, or
ice, as these can cause the siphon inlet filter to clog. If there is
frost or significant condensation on the interior surfaces of
the dewar, it may be necessary to warm up the dewar using a
handheld blow-dryer.
37. Ensure that the brass intake filter at the end of the dewar lid
siphon tubing is at or above room temperature before submerg-
ing it in the liquid nitrogen; if the metal is cold (e.g., if the
dewar was recently in use), then condensation may form, caus-
ing the filter to be blocked by ice when immersed into the liquid
nitrogen. Do not close the lid too quickly after lowering the
siphon into the liquid, as the boil-off may cause pressure buildup
that can result in ejection of liquid nitrogen through the capil-
lary tube. Whenever handling the lid, take care not to damage
the black siphon capillary tubing; for example, when setting the
lid down, always turn it upside down (interior side up).
38. It is possible to directly modify the hold temperature of the
currently active ramp (which is highlighted in blue in the tem-
perature profile table). Simply revise temperature value in the
Limit column, and then press Enter.
39. The sample loading temperature is the temperature at which
the sample will be held before the freezing process starts.
This temperature should be above room temperature (to
prevent condensation) and often corresponds to the normal
incubation temperature for the cell type under study (e.g.,

37C for mammalian cells).
40. The seeding temperature is the temperature at which e xtracellular
ice crystallization will be induced, at the start of the cell freezing
experiment. This temperature should be near (typically within
1C) the melting point of the solution. If the seeding tem-
perature is too high, the time of contact required between the
sample and the seeding cold spot to achieve ice formation may
become very long, resulting in (uncontrolled) cooling of large
parts of the sample. Furthermore, if the seeding temperature is
too high, the ice formed during the seeding process may melt
before the temperature profile can be advanced to the next ramp
(the equilibration hold). Conversely, if the seeding temperature
is too low, rapid growth of extracellular ice throughout the spec-
imen may cause damage to some of the cells. As a general guide-
line, the seeding temperature should be selected so that the
seeding process results inlocalized ice formation (not extending
more than a few millimeters beyond the seeding cold spot)
within 1020s; moreover, the size of the resulting ice patch
should not change by more than 50% in the time it takes for
the sample to reach the equilibration hold temperature (Ramp
3in Table1). When experimenting with different seeding tem-
peratures, it is helpful to directly observe the formation and sub-
sequent changes to the localized ice patch (which appears as a
white cloudy area), by using a small flashlight to illuminate the
sample edge through the BCS196 stage lid window.
41. The equilibration temperature is the temperature at which the
sample will be held after seeding, to allow the extracellular ice
to propagate throughout the sample, until it reaches equilib-
rium with the unfrozen solution. This temperature must be
below the melting point of the solution. The equilibration
temperature should be chosen according to the needs of the
experiment. If the equilibration temperature is low, then the
unfrozen liquid fraction will be lower (possibly causing cell
crowding or excessive deformation if too low), and the initial
water content within the cell will be reduced (which will
depress intracellular freezing temperatures). If the equilibra-
tion temperature is high (approaching the melting tempera-
ture), then cell dehydration will be minimal, but the volume
fraction of ice in the sample may be too low. If the amount of
extracellular ice in the sample is so low that many cells are
initially not in contact with ice, then the experimental results
may be more difficult to interpret (because the probability of
intracellular ice formation is known to be affected by contact
with extracellular ice crystals). Moreover, if rapid freezing
(Ramp 4in Table1) is initiated while the unfrozen liquid
fraction is large, then fine dendritic ice crystals will fill the
unfrozen spaces in the early stages of the freezing ramp, which

may cause the view of the cells to be obscured.
42. The equilibration hold time is the time at which the sample will
be held at the equilibration hold temperature before rapid freez-
ing (Ramp 4in Table1) is initiated. To facilitate interpretation
of experimental results, the hold time should be sufficiently long
to allow the ice field to advance through the area of the sample
to be observed during freezing, and to allow cells in this region
to reach osmotic equilibrium with the freeze-concentrated liq-
uid in the unfrozen fraction. The equilibration hold also needs
to be sufficiently long to allow time for selecting a field of view,
adjusting the microscope optics (e.g., fine focus, Khler illumi-
nation, and contrast enhancement), and any other pre-freezing
experimental procedures (such as acquisition of fluorescence
images). The length of the equilibration hold will also affect the
size and shape of the extracellular ice crystals (due to annealing)
and, if excessively long, may cause some degradation of cell
quality due to the action of solution effects mechanisms.
Because both of these factors (geometry of ice-cell interactions,
as well as exposure to the unfrozen solution) may affect the
subsequent intracellular ice formation kinetics, it is important to
keep the equilibration hold time consistent from experiment to
experiment. In the example in Table1, the hold time has been
set to 5min. Note that the Linksys32 software allows for the
unit of time to be configured as either minutes or seconds. The
time field in the Temperature Control Panel will be labeled
Hold mins if the unit of time is minutes, or Hold secs if the
time is measured in seconds; in the latter case, the temperature
profile Ramp 3in Table1 should be modified so that
Time=300in order to achieve a 5-min equilibration hold.
43. The rate of cooling (in C/min) for the main freezing ramp is
an important experimental parameter. For measurement of
intracellular ice formation kinetics, the cooling rate is typically
set to the fastest controlled rate achievable by the cryomicro-
scope system. This minimizes the amount of cell dehydration
during freezing, which significantly simplifies data interpreta-
tion, by allowing the analyses of intracellular ice formation and
water transport to be decoupled (because, for negligible levels
of water loss, the chemical composition of the intracellular liq-
uid can be assumed to be constant during freezing).
Alternatively, the cooling rate may be varied as part of the
experimental design (e.g., to estimate the critical cooling rate,
at which the cumulative incidence of intracellular ice formation
is 50%). In some cases (especially when working with thicker
samples, which can develop vertical temperature gradients dur-
ing rapid cooling), one may also choose to limit the cooling
rate in order to reduce temperature measurement errors.
44. The experiment end temperature should be selected so that

intracellular ice formation has been completed before the end
of the freezing ramp. For mammalian cells frozen rapidly (i.e.,
with negligible dehydration) in isotonic solution without
cryoprotectants, the lowest expected intracellular freezing
point is approximately 40C (corresponding to ice forma-
tion by homogeneous nucleation); the theoretical minimum
will be lower if the freezing point is depressed by the use of
cryoprotectants additives, etc. In practice, intracellular freez-
ing events may occur at stage temperatures lower than the
expected minimum nucleation temperature, as a result of
unanticipated cell dehydration or vertical temperature gradi-
ents. Thus, an additional margin of 510C past the lowest
expected intracellular ice formation temperature is advisable.
45. Data analysis will be greatly simplified if the duration of the
video recording is longer than the duration of the freezing ramp
(the cooling step in Ramp 4) but shorter than the total duration
of Ramp 4 (including the cooling step and the isothermal hold
period). Thus, the data analysis approach described in
Subheading3.9 assumes that the warming ramp (Ramp 5) does
not begin until after the video recording has ended. To ensure
that the video recording does not overlap the warming ramp, it
may be necessary to adjust the length of the isothermal hold at
the end of the cooling step, to extend the total duration of
Ramp 4. For example, in Table1, the duration of the rapid-
freezing ramp is 82.6s (including the specified 1-min hold at
50C). Thus, if recording 8-bit video at a resolution of
300300pixels and a frame rate of 2,000 fps, a camera memory
capacity of 4GB will yield 23.9s of recording time; for the tem-
perature profile specified in Table1, this recording duration is
sufficient to capture the cooling from 1 to 50C (22.6s),
whereas the video recording will have terminated well before
the start of the warming ramp. Conversely, without the 1-min
hold time at 50C, the last ~2,500 frames of the video record-
ing would depict the warming response from 50C to approx-
imately 47C (i.e., the initial part of Ramp 5).
46. After entering all values in the temperature profile table, save
the profile by selecting Save, then Temperature Profile,
under the File menu. Specify the desired location and name
of the temperature profile file, and then click the Save but-
ton. The protocols described in Subheadings3.6 and 3.7
assume that the saved temperature profile starts and ends with
an indefinite hold at the sample loading temperature (e.g.,
Ramp 1 and Ramp 5in Table1).
47. This will reduce the accumulation of frost on the coolant inlet pipe
inside the BCS196 chamber. Any water allowed to accumulate
within the chamber may make the purging procedure less effective.
48. Clearance can be further improved by rotating the objective

turret so that the shortest objective (or an empty objective
position) faces the front. If the BCS196 mounting allows for
translational (XY) movement of the cryomicroscope stage
body relative to the microscope, then the BCS196 should be
moved as far toward the front of the microscope as possible.
49. Note that for certain combinations of cryomicroscope mount-
ing hardware designs and microscope condenser optics, it is
possible for the cryomicroscopes mounting adapter to make
contact with the condenser top lens if the BCS196 stage body
is moved laterally (in the XY plane) while the condenser focus
position has been set for Khler illumination. Thus, to avoid
damage in such cases, one should make a habit of always low-
ering the condenser before making any changes to the posi-
tion of the cryomicroscope stage body.
50. The end of the window tubing should touch the edge of the
stage lid window. It may be necessary to position the tubing
closer to the window center (if frost forms in the light path
during freezing experiments) or further away from the win-
dow edge (if the tubing interferes with objectives that have
shorter working distance).
51. In the modern (post-2010) versions of Vision Research Phantom
Camera Control (PCC) software, click the Live panel tab, and
expand the Cine Settings section. In the legacy version of the
Vision Research Phantom Camera Control application, select
Setup and Recording under the Acquisition menu.
52. To load a previously saved camera setup file for Vision Research
Phantom cameras (.stp format), complete the following steps:
First, in the PCC software, expand the Advanced Settings
section of the Live panel, and click the Load Settings but-
ton; conversely, in the legacy camera control application, click
the Open button in the Setup and Recording window.
Second, if using the PCC software, set the file type selection to
All (*.*); in the legacy application, ensure the file type is set
to Setup files (*.STP,*STG). Third, select the desired setup
file (.stp extension) and click the Open button.
53. The ROI dimensions should be selected to accommodate the
size of the field under observation, including a margin to allow
for sample drift during cooling (e.g., due to thermal contraction
of the coolant pipes); for example, on a Linkam Scientific
Instruments BCS196 cryomicroscope stage, the average drift is
on the order of 1m/K.The larger the ROI, the more data can
be collected for each freezing run (because of the larger number
of cells in the field of view). The maximum ROI size is deter-
mined by the physical dimensions of the camera sensor. In addi-
tion, because the view of the specimen is restricted by the
1.3-mm-diameter circular aperture in the silver block of the

cryomicroscope stage, any part of the image that is obstructed
by silver block is not usable and should be cropped out (because
imaging of obscured regions uses up memory and bandwidth).
Therefore, at microscope magnifications that are sufficiently low
to allow the full area of the silver block aperture to fit within the
camera image, one should set the ROI to a square shape with
edges that do not exceed the aperture diameter. In this scenario,
the upper bound on the ROI size will be determined by the
shortest dimension of the camera sensor (conventionally the
vertical dimension). For modern high-speed cameras, the maxi-
mum vertical resolution that can be used at 2,000 fps (which is
the minimum recommended frame rate for imaging of intracel-
lular ice formation) is typically in the range 6001,600pixels,
corresponding to maximum ROI dimensions that range from
600600pixels to 1,6001,600pixels. Nonetheless, for many
high-speed video cryomicroscopy applications, a reduced-size
ROI is preferable, because keeping the ROI small allows for
longer recording times and faster frame rates. It should be noted
that if the microscope magnification is optimally matched to the
camera pixel pitch (so that a pixel pair spans the Rayleigh resolu-
tion limit), a typical somatic cell can be resolved in an ROI with
linear dimensions on the order of 20pixels. Adding to this a
margin to account for sample drift, an ROI size of 128128pix-
els may be suggested as a practical minimum. At this resolution,
every gigabyte of camera memory yields approximately 1530s
of recording time at a frame rate of 2,000 fps (the exact record-
ing time depends primarily on the sensor bit depth, i.e., the
number of distinct gray levels that can be resolved by each pixel);
in contrast, under the same conditions, a 600600-pixel ROI
will yield ~1s of recording time, and a 1,6001,600-pixel ROI
will only allow for ~0.1s of recording time.
54. As a general rule of thumb, to avoid motion blur in video
recordings of a feature that is moving at a velocity v, the camera
exposure should not exceed the image spatial resolution,
divided by v. The spatial resolution in this context is the larger
of the camera sensor resolution (i.e., pixel pitch divided by
magnification) and the Rayleigh limit (which is on the order
~1m for high-speed video cryomicroscopy, as a result of long-
working-distance lenses and the need to stop down the con-
denser diaphragm for improved image contrast). Thus, for a
high-speed camera sensor with pixel pitch of ~20m, to image
intracellular ice formation (in which the solidification front
typically advances at a velocity v10 m/ms), the maximum
exposure is 400 and 200s for magnifications of 5 and 10,
respectively; for magnifications of 20 and greater, the camera
exposure should not exceed 100s. If the image of the cells is
saturated (overexposed), either the illumination levels or the
exposure time can be reduced. Conversely, if the image is

underexposed, the best remedy is to increase illumination lev-
els, if possible; the exposure time should not be increased
beyond the upper bound specified here. If illumination levels
and exposure time have both been maximized, but the image is
still too dark, then one may use any available sensitivity-enhanc-
ing functionality specific to the camera model (e.g., preampli-
fier gain increase or output of the analog-to-digital converters
lower bits) or, as a last resort, apply digital image processing
tools such as brightness increase or gamma correction.
55. To unambiguously identify intracellular ice formation, at least
two images of the ice crystal interface must be captured as it
advances through the cell interior. Therefore, the velocity of
intracellular crystal growth, when divided by one-half the
smallest diameter of the projected cell image, yields a lower
bound for the image acquisition rate that is required to detect
intracellular ice formation. Thus, for a crystallization velocity
on the order 10m/ms, and a projected cell diameter of
~10m, the minimum frame rate is ~2,000 fps. In practice,
there is also an upper bound on the image acquisition rate.
For example, the frame rate cannot exceed the inverse of the
camera exposure time (set in step 4 of Subheading3.4),
because the exposure of one frame must be completed before
the next frame can be exposed. Similarly, high-speed cameras
typically have a limiting data throughput rate, so that the
maximum possible frame rate will be determined by the ROI
dimensions that were specified in step 3 of Subheading3.4.
Furthermore, increasing the frame rate will decrease the avail-
able video recording duration, which may in some cases result
in loss of data. On the other hand, to detect intracellular ice
formation, there is typically no advantage to having more
than 510 sequential video frames depicting the growth of
the intracellular ice. Therefore, for cells that have a projected
diameter on the order ~10m, the image acquisition rate
does not need to be faster than ~8,000 fps.
56. Modern high-speed imaging systems save the acquired images
in a circular (first-in-first-out) memory buffer and terminate the
recording after a preset number of frames have been acquired
following detection of the trigger signal. Thus, some portion of
the video that is retained in the camera memory will have been
acquired prior to the trigger event, whereas the post-trigger
frames all represent images that were acquired after the trigger
event. In the data analysis procedure described in Subheading3.9,
only the video images acquired after the trigger event are used.
57. If the recording length is significantly shorter than the dura-
tion of the cooling portion of the ramp, then the cryomicros-
copy videos will be prematurely truncated, which may result
in missing intracellular ice formation events. One may esti-

mate the temperature at which the video recording will ter-
minate by expressing the computed post-trigger recording
duration in minutes and multiplying by the cooling rate from
Ramp 4in Table1 (in C/min) and then subtracting this
product from the ramp start temperature (i.e., the Limit
value in Ramp 3 of Table1).
58. Alternatively, if the recording end time occurs too late, one
may reduce the number of post-trigger frames. In particular,
to establish an appropriate number of post-trigger frames,
subtract 1s from the calculated cooling ramp duration, and
then multiply the result by the image acquisition rate. Note
that any video frames that are acquired before the trigger will
not be used in the data analysis. If less than half of the total
number of frames available for recording will be acquired
after the trigger, then a more suitable approach to reducing
the recording duration is to partition the camera memory.
59. To save Vision Research Phantom camera setup file (.stp for-
mat), complete the following steps: First, in the PCC soft-
ware, expand the Advanced Settings section of the Live
panel, and click the Save Settings button; conversely, in the
legacy camera control application, click the Save button
in the Setup and Recording window. Second, if using the
PCC software, set the file type selection to All (*.*); in the
legacy application, set the file type to Setup files (*.STP).
Third, specify the desired location and name of the setup file,
including the .stp extension (the file will not be properly saved
if the .stp extension is not explicitly appended to the specified
file name), and click the Save button.
60. To clean off grease residue from the external surfaces of the
syringe, isopropyl alcohol works well.
61. Dispensing needles in the range 1620G are recommended.
With the larger bores, less pressure is needed to dispense
grease, whereas the smaller bores offer finer control of the
amount dispensed.
62. If the dispenser has not been used for some time, the dispens-
ing tip should be cleared by pushing some grease through the
syringe, before use in sample preparation.
63. The dispensing needle should be held perpendicular to the
glass surface. Hold the coverslip steady by applying a down-
ward force using the forceps.
64. The total amount of silicone grease transferred to the cover-
slip surface should be approximately 2mg.
65. Small gaps in the circle perimeter are acceptable. The sample
does not need to be hermetically sealed. The purpose of the
grease ring is fourfold: (a) to provide adhesion to hold the

coverslip sandwich construct together, (b) to reduce the risk
of leakage of the liquid sample onto the cryomicroscope silver
block, (c) to control the thickness of the liquid sample, and
(d) to reduce the rate of evaporation of water from the sample.
66. Do not push down on the coverslips before they have been
aligned. Hold the two angled tips of the first pair of forceps
against the work surface, approximately 1cm apart and perpen-
dicular to the surface; this pair of tips will serve as a brace for
the bottom coverslip. Gently nudge the bottom coverslip so
that its edge makes contact with both tips of the bracing for-
ceps, while the overhang of the misaligned top coverslip should
be oriented to be diametrically opposed to the bracing forceps.
In a similar manner, the second set of forceps can be held with
its angled tips perpendicular to the work surface and brought
into contact with the top coverslip overhang (i.e., on the side
opposite the first set of forceps), as shown in Fig.3c-d. By care-
fully moving the two sets of forceps toward each other, the two
coverslips will be aligned within the square-shaped arrange-
ment of the four tips (see Fig.3e).
67. For best results, push only on the part of the coverslip that is
directly above the grease ring, by running the tip of the forceps
along the perimeter of the grease ring while applying gentle
pressure. Observe the response of the grease and the aqueous
sample during this process. The grease should spread to a width
of 12mm, but there will typically be significant variations in
width along the perimeter of the grease circle (thus, it is impor-
tant to ensure that the silicone grease never spreads so far as to
reach the edge of the coverslip). Simultaneously, the liquid sam-
ple will spread out due to the capillary forces that arise when the
two coverslips are brought closer together. Although it is nor-
mal for the aqueous liquid to occasionally breach the grease
barrier, one should attempt to minimize such leaks by removing
all pressure when the liquid reaches the inside edge of the grease
ring. At no time should any liquid be squeezed out of the cov-
erslip sandwich (i.e., past the coverslip edge). If necessary, one
may use a slightly smaller volume of liquid in step 7 of
Subheading3.5 (e.g., ~3L), although this may result in air
gaps within the grease circle (which are also undesirable).
68. The techniques described here will yield a coverslip sandwich
in which the thickness of the liquid sample is approximately
30 m. This sample thickness is appropriate for cells with
diameters in the range 1030m. If necessary, the sample
thickness can be varied by altering the amount of grease or the
liquid sample volume. However, the sample thickness should
be kept as small as possible without deforming the cells.
69. It is important that the surfaces and edges of the sample be

completely dry before it is loaded into the cryomicroscope
stage. Residual liquid on the sample can cause serious prob-
lems, including contamination of the silver block surface
(which can affect temperature control), capillary adhesion of
the bottom coverslip to the silver block (which may cause the
sandwich construct to delaminate when attempting to move
the sample), and ice formation on the external surfaces of the
coverslips (which will cause the view of the sample to be
obscured or result in confounding optical interference that
can be misinterpreted during data analysis).
70. If the sample positioning is not perfectly accurate, then the
sample edge may come to rest on top of the sample carrier
ring, preventing thermal contact between the silver block and
the samples bottom coverslip. Thus, one should check
whether the sample is laying flat inside the carrier ring, by
nudging the sample horizontally (lightly touching the sample
using the edge of the vacuum tweezers suction cup works
well for this purpose). It is also advisable at this time to move
the carrier back and forth using the BCS196 stage X- and
Y-drives while observing the sample to confirm that it can be
positioned laterally without problems (in particular, check
that the carrier ring does not slide into the space between the
samples top and bottom coverslip).
71. For record-keeping and quality assurance purposes, it is useful
to have a continuous recording of the complete temperature
history experienced by the sample, from the time of loading
into the cryomicroscope stage until the end of the
experiment.
72. The purging process is also described in the BCS196 User
Guide, and a video tutorial can be viewed on the Linkam
Scientific Instruments website (http://www.linkam.co.uk/
video-manuals/) after obtaining login credentials (by submit-
ting a product registration for the cryomicroscope stage).
Purging is necessary to reduce the moisture content within
the cryomicroscope stage chamber, to prevent condensation
and frost from forming on the sample during freezing.
73. If the LNP flowrate is increased rapidly, the sample tempera-
ture may temporarily drop (by ~1C). If the specimen is sen-
sitive to temperature fluctuations of this magnitude, the LNP
flowrate should be incremented slowly, while monitoring the
stage temperature.
74. The metal window tubing clip will be positioned near the end
of the tubing. Try not to disturb the position of the clip rela-
tive to the tubing outlet.
75. There should be an audible hissing sound as the nitrogen gas

flows through the cryomicroscope stage chamber and exits
the rear gas port, flushing out the chamber air.
76. The purpose of this process is to promote gas mixing within the
stage chamber, to more efficiently purge out any moist air. One
may confirm that the chamber is adequately pressurized by
confirming that the stage lid window deflects slightly whenever
the gas valve is blocked (e.g., by observing distortion of light
that is reflected in the lid window). The absence of any notice-
able deflection may indicate a leak in the chamber lid and/or
windows, or insufficient nitrogen flow from the pump. It is
very important to wear protective eyewear during the pressur-
ization process, especially while visually observing deflections
in the stage lid window. It should also be noted that the sample
temperature may temporarily fluctuate (by approximately
1C) during the purging process, due to the perturbation of
coolant flow through the cryomicroscope stage. If the speci-
men is sensitive to temperature fluctuations of this magnitude,
the duration of blockage of the rear gas port should be reduced,
and the stage temperature should be monitored during the
pressurization process and subsequent release.
77. Because the LNP flowrate decreases rapidly when the pump is
switched back to automatic control, the sample temperature
may temporarily rise (by ~1C). If the specimen is sensitive to
temperature fluctuations of this magnitude, the LNP flowrate
should be manually decreased at a slow to moderate pace,
before switching to automatic control.
78. To minimize the risk of sample damage due to phototoxicity or
excessive heating by the light source, the microscope illumina-
tion should remain off (or set to minimal power output) while
light is not needed. It is also recommended that control experi-
ments be performed to test for possible influence of high-inten-
sity illumination on intracellular ice formation kinetics [23].
79. The objective magnification should be sufficiently low that the
edges of the BCS196 silver block aperture can be seen in the
camera image. If necessary, one may temporarily expand the
field of view by increasing the dimensions of the cameras
region of interest until the silver block aperture is visible. This
is necessary to make possible adjustment to the centering of the
silver block aperture in the microscopes optical axis, thus maxi-
mizing the effective numerical aperture of the condenser.
80. If the live image is too dark or too small to observe the sample
for focusing purposes, it may be necessary to apply digital
image processing options, such as zoom, gain, or gamma
adjustment, in the camera control software. Conversely, if the
camera image is saturated, temporarily reduce illumination lev-
els or decrease the camera exposure time to allow focusing.
81. If the BCS196 mounting allows for lateral (XY) movement of

the cryomicroscope stage body relative to the microscopes
optical axis, use the XY-control on the mechanical stage to
center the silver block aperture in the light path. Alternatively,
the BCS196 mounting hardware may include centering
screws that can be used to align the silver block.
82. Setting a value Time=0in an active ramp (i.e., the tempera-
ture profile row highlighted in blue in the Linksys software)
that is currently holding at the ramps limit temperature will
cause the isothermal hold to terminate, thus advancing the
temperature profile execution to the next ramp in the table.
83. The BCS196 cryomicroscope has a small rectangular silver
post attached to the liquid nitrogen inlet pipe, next to the
main (cylindrical) silver block; this post can be chilled by forc-
ing nitrogen coolant through the inlet pipe, thus producing a
localized cold spot that can be used to induce ice formation in
the sample. To induce (i.e., seed) ice formation, the edge of
the sample must first be brought to rest over the cold spot.
With the G16.3 sample carrier, the sample will be in the cor-
rect position for seeding when both the X- and Y-drives have
been turned clockwise as far as possible (however, one should
be careful not to turn these screw drives further when the end
of travel has been reached, because this may result in damage
to the stage). When the sample is in the correct position for
seeding, the aperture in the main silver block will be partially
blocked by the edge of the G16.3 carrier.
84. If seeding is successful, the extracellular ice nucleation mani-
fests as a sudden formation of a white cloudy region in the
area of the sample that is located over the cold spot (and typi-
cally extending a few millimeters beyond the cold spot). To
aid observation of the seeding process, a small flashlight can
be used to illuminate the cold-spot region through the
BCS196 stage lid window.
85. Wait until after the sample has reached the equilibration tem-
perature (after step 15 of Subheading3.7), and then record
the stopwatch time (which represents the time required to
induce ice formation in the sample). This time should nor-
mally be in the range 1020s; seeding times outside this
range may indicate that the seeding temperature was not
properly selected. When running multiple experiments with
replicate samples, outliers in the seeding time may signify a
problem with the sample (e.g., dehydration during sample
preparation) or with the cryomicroscope stage (e.g., impeded
heat transfer due to scratches or foreign material on the silver
block surface, or blockage of liquid nitrogen flow).
86. This will advance the execution of the temperature profile to
Ramp 3in Table1, thus bringing the sample to the post-
seeding equilibration temperature. Steps 14 and 15

(Subheading3.7) must be completed in quick succession, so
that the ice patch formed during seeding does not melt (which
may happen if the seeding temperature is above the solution
melting temperature) or grow too rapidly (which may happen
if the seeding temperature is below the equilibration tempera-
ture). As a general guideline, the size of the ice patch at the
sample edge should not change by more than 50% before
the sample has reached the equilibration hold temperature.
87. During cooling, the sample may drift laterally, due to thermal
contraction of the metals in the BCS196 stage. Thus, when
selecting a field of view, allow for a margin between the cells
of interest and the image edge.
88. All adjustments should be made with reference to the camera
live image (not the microscope eyepieces). For high-
magnification objectives, if the imaged field of view is too
small to contain the blade edges of the stopped-down field
diaphragm, one may temporarily expand the field of view by
increasing the dimensions of the cameras region of inter-
est. Only minor adjustments to the condenser focus and
field diaphragm should be necessary, because the micro-
scope was already pre-configured for Khler illumination in
step 5 of Subheading3.7.
89. The intracellular ice crystals (which have not yet formed) do
not have a high degree of contrast. Thus, the condenser dia-
phragm aperture typically needs to be significantly restricted
in order to detect intracellular ice formation. Some iteration
may be necessary to select the best aperture size, because the
contrast of the intracellular ice crystals can only be checked
during slow-motion playback of the high-speed video that is
recorded during freezing.
90. In the Vision Research Phantom cameras, this is done using
the Current Session Reference function. For optimal image
quality, the camera should be recalibrated before each video
recording, because the black levels are sensitive to the changes
in temperature as the camera warms up during continuous
operation. Black level calibration requires that light be pre-
vented from reaching the camera sensor; this can be achieved
by using the designated microscope control to temporarily
divert all light away from the camera port (and onto the
eyepieces).
91. Because it may not be possible to determine whether or not a cell
has frozen unless a high-speed video recording of the intracellu-
lar ice formation event has been captured, it is extremely impor-
tant to initiate the video acquisition before any intracellular
crystallization events occur. Thus, if possible, the video record-
ing should begin at the very start of the rapid-freezing ramp.

However, if necessary due to camera memory limitations, one
may delay the start of the high-speed video recording (i.e., the
trigger time) slightly, as long as the stage temperature remains
sufficiently high to preclude any intracellular ice formation
prior to the push of the trigger button. If any of the cells in
the field of view have already frozen prior to the beginning
of the acquired video, then the data analysis methods described
in Subheading3.9 will yield invalid results.
92. During rapid cooling, the sample may go out of focus due to
thermal contraction of the metal components to which the
silver block is mounted. With practice, one may correct such
problems by making small adjustments to the microscope fine
focus during the early stages of freezing. However, it is impor-
tant to avoid adjusting the microscope focus further when
temperatures have reached a regime in which intracellular ice
formation is likely, because dynamic changes in specimen
appearance in the recorded video will confound detection of
intracellular freezing events.
93. In the current version of the Linksys software, this is the only
available opportunity to save the recorded temperature data.
Thus, it is very important to save the .iml file immediately
when prompted. Do not click Cancel, because this will
result in permanent loss of the data.
94. At the end of the video acquisition, the image data exist only in
volatile memory. Thus, the video must be downloaded from
the camera memory and written to the computer hard drive.
The time required to save the video will depend on the amount
of data contained in the memory, the write-speed of the hard
drive, and the data throughput rate of the communication
channels that connect the two devices. It is a good idea to test
(and, if possible, optimize) the video download time in advance,
to determine the delay time between successive experiments.
95. If there is any air gap between the bottom surface of the sample
carrier ring and the top surface of the silver cooling block that
is greater than ~0.1mm, then the sample carrier may occasion-
ally trap the bottom coverslip of the sample. Thus, to prevent
forces from being transmitted to the sample carrier when lifting
the sample out, it is advisable to first check whether the sample
is stuck. This can be done by gently nudging the sample hori-
zontally, to check whether it slides freely before making contact
with the inner edge of the sample carrier ring (there should be
a space of approximately 0.3mm between the sample coverslip
edge and the sample carrier ring).
96. To load a previously saved profile file, complete the following
steps: First, stop the profile execution by clicking the blue square
button in the Temperature Control toolbar. Second, click on

any cell in the Delay column of the temperature profile table.
Third, go to the File menu and select Open and then
Temperature Profile; select the desired profile file (.pro exten-
sion) and click the Open button. Finally, restart profile execu-
tion as soon as possible, by clicking the blue triangle in the
Temperature Control toolbar. Ensure that the stage tempera-
ture is holding at the desired sample loading temperature.
97. To reduce wear, employ a twisting motion to ease the tubing
connectors off the metal pipes, instead of pulling them off.
98. Be careful not to twist or bend the black capillary tubing. Be
mindful of small amounts of liquid nitrogen that may be
ejected from the end of the capillary tube, due to pressure
buildup in the dewar.
99. In Vision Research Phantom cameras, the first post-trigger
frame always corresponds to frame #0.
100. It is important to avoid introducing bias: the criteria for

removing cells from further analysis should be independent of
the probability of ice formation in the cells. For example, even
if a cell does not disappear from the field of view until after its
intracellular freezing event, it cannot be included in the tally.
In addition, one should also omit cells that are not yet visible
at the video trigger time but that drift into the field of view
during the course of the freezing ramp. It is recommended
that attempts be made during the cryomicroscope experiment
to minimize the probability of data censoring (e.g., by select-
ing a field of view in which there is a sufficient margin between
the cells and the edges of the field to account for drift).
Furthermore, it should be noted that the Nelson-Aalen algo-
rithm does allow for a more flexible approach to data censor-
ing, but such variations of the data analysis technique are
outside the scope of this chapter.
101. If the temporal resolution of the acquired video is sufficiently
fine (e.g., sampling intervals on the order 0.5ms or better),
then the majority of intracellular ice formation events will
manifest as a solitary wavefront that travels through the cell
interior (typically over the course of a few milliseconds,
although the crystal growth velocity may vary by an order of
magnitude), whereas concomitant changes in cell opacity are
minimal. Example high-speed video recordings that depict
intracellular freezing are available in the online supplementary
materials of recently published articles [12, 23]. If the video
acquisition rate is less than 1,000 fps, then intracellular ice
formation will typically appear in the form of so-called
twitching events [27], which are more difficult to discern.
102. When searching for intracellular ice crystallization events in

high-speed video recordings, especially when the acquisition
frame rate is too slow (e.g., less than 1,000 fps), it is possible
to misidentify as intracellular ice formation certain unrelated
changes in cell appearance that are not caused by ice crystals
within the cell. For example, a small change in the vertical
position of the cell relative the camera focal plane may alter
the appearance of the cell image in a manner similar to the
changes caused by the twitching form of intracellular ice
formation. To avoid such misidentifications, one should keep
in mind that true intracellular ice formation events are typi-
cally stochastic (random) and relatively independent (espe-
cially in cell suspensions). Thus, if changes in cell appearance
occur at the same time as similar changes in appearance of
other cells in the field of view, it is unlikely that these events
represent intracellular freezing (which seldom occurs simulta-
neously in two cells)a change of focal plane is a more likely
explanation. Likewise, if changes in the appearance of a cell
are spatially and temporally correlated with movements in the
surrounding extracellular ice field, one should consider inter-
actions with the external ice as an alternative explanation for
the observed cellular changes. Finally, especially in the absence
of clear view of the advancing intracellular solidification front,
one may use the duration of the observed event as a criterion
to confirm whether or not the recorded phenomenon repre-
sent intracellular ice formation. In particular, for somatic cells
(with diameters in the range 550m) that are frozen rapidly
(with minimal cell dehydration), the duration of intracellular
crystallization processes typically falls in the range 0.110ms.
Thus, cellular events that occur over a much longer time scale
should be critically evaluated before a decision is made to clas-
sify the observation as an intracellular ice formation event.
103. To convert a Linksys data file, first go to the File menu and
select Open, then Data File; select the desired data file
(.iml extension) and click the Open button. A Data Chart
window showing the recorded temperature data will be dis-
played; click the Export icon in the toolbar of this window.
In the export window, select CSV Data (Text) (*.txt) under
the Save as type options. Next, specify a name and location
for the exported data file, and click the Save button.
104. For example, in the Vision Research Phantom Camera Control
application, the timestamps for each frame (Elapsed Time
from Trigger) are displayed in microseconds. Thus, multiply
by 106s/s to convert to seconds. Do not perform any
rounding on the resulting value, to prevent loss of precision.
105. Save the resulting value to a precision commensurate with the
video frame rate and temperature rate of change. For the analy-
sis of kinetics, it will be necessary to distinguish between interpo-
lated temperature values corresponding to events in successive

video frames. Thus, the minimum numerical precision should
equal the magnitude of the temperature interval Tk+1Tk,
divided by the number of video frames acquired in the time
interval tk+1tk. For example, if the cooling rate was 150C/
min and the image acquisition rate was 2,500 fps, then the
required precision in the interpolated temperature calculation
is at least three decimal places (0.001C).
106. This represents a simplified approach to analysis of censored
data. Alternative approaches exist that do not require data
points to be discarded, but such techniques are beyond the
scope of this chapter. Nonetheless, if consistency is main-
tained in the experimental procedures (i.e., if the start and
end temperatures of each video recording are consistent
among replicate trials), then the risk of having to discard
intracellular freezing events from analysis will be low.
107. It is important to note that one should not exclude from this
total any cells in which no intracellular ice formation was
detected (unless the cell was already omitted from analysis on
the basis of criteria given in step 1 in Subheading3.9).
108. One possible scenario that would yield the data shown in

Table2 is the following hypothetical case. In this example, it is
assumed that there were two replicate experiments with three
cells in each video, all of which underwent intracellular freez-
ing and were never obscured from view. In one video, the
recording start and end times corresponded to interpolated
temperatures of 1.000 and 21.000C, respectively; the cor-
responding intracellular ice formation events had interpolated
temperature values of 3.890, 11.123, and 13.768C.In
the other video recording in this example, the start and end
temperatures were 4.000 and 24.000C, whereas the intra-
cellular ice formation events occurred at 7.250, 11.124, and
22.446C.Thus, the kinetic analysis is performed for the
overlapping temperature range only (i.e., between 4.000 and
21.000C). As a result, the cell that was observed to freeze
at 3.890C in the first video is omitted from the risk group.
The cell that experienced intracellular ice formation at
22.446C in the second video is also not included as an
event in the table; however, this cell is counted as part of the
initial size of the risk group (because it is unfrozen at the start
of the kinetic analysis, when the temperature is 4.000C).
109. If all non-disqualified cells undergo intracellular crystalliza-
tion, then the final value in the risk group column will be
one. If some cells do not experience intracellular ice forma-
tion prior to the analysis end temperature, then the final
value in the risk group column will equal one plus the num-
ber of unfrozen cells.
110. A conventional method to present intracellular ice formation

data is to plot the cumulative probability of intracellular freez-
ing as a function of descending temperature. Another com-
mon metric used to quantify the experimental results is the
median intracellular ice formation temperature, which is
defined as the temperature at which P=0.5. Evaluation of the
median intracellular ice formation temperature does not
require computation of the cumulative probability, because it
is equivalent to the temperature at which the cumulative haz-
ard A=ln(2).
Acknowledgments
The author gratefully acknowledges Peter Grocutt of Linkam

Scientific Instruments for valuable discussions and technical sup-
port during the development of the high-speed video cryomicros-
copy system and during the writing of this chapter. This work was
supported by National Science Foundation grant CBET-1066619.
References
1. Mazur P, Leibo SP, Chu EHY (1972) A two- ice formation based on the extent of supercool-
factor hypothesis of freezing injury: evidence ing. Cryobiology 29:359373
from Chinese hamster tissue-culture cells. Exp 9. Toner M, Tompkins RG, Cravalho EG etal
Cell Res 71:345355 (1992) Transport phenomena during freezing of
2. Toner M (1993) Nucleation of ice crystals isolated hepatocytes. AIChE J 38:15121522
inside biological cells. In: Steponkus P (ed) 10. Irimia D, Karlsson JOM (2005) Kinetics of
Advances in low-temperature biology, vol 2. intracellular ice formation in one-dimensional
JAI Press, London, pp151 arrays of interacting biological cells. Biophys J
3. Karlsson JOM, Cravalho EG, Toner M (1993) 88:647660
Intracellular ice formation: causes and conse- 11. Karlsson JOM (2010) Effects of solution com-
quences. Cryo-Letters 14:323336 position on the theoretical prediction of ice
4. Gppert HR (1830) Ueber die Wrme- nucleation kinetics and thermodynamics.
Entwickelung in den Pflanzen, deren Gefrieren Cryobiology 60:4351
und die Schutzmittel gegen dasselbe. Joseph 12. Higgins AZ, Karlsson JOM (2013) Effects of
Max, Breslau intercellular junction protein expression on
5. Molisch H (1897) Untersuchungen ber das intracellular ice formation in mouse insulinoma
Erfrieren der Pflanzen. Gustav Fischer, Jena cells. Biophys J 105:20062015
6. Diller KR, Cravalho EG (1971) A cryomicro- 13. Toner M, Cravalho EG, Stachecki J etal (1993)
scope for the study of freezing and thawing Nonequilibrium freezing of one-cell mouse
processes in biological cells. Cryobiology 7: embryos. Membrane integrity and develop-
191199 mental potential. Biophys J 64:19081921
7. Brown MS, Reuter FW (1974) Freezing of
14. Karlsson JOM, Eroglu A, Toth TL etal
nonwoody plant tissues. III.Videotape microg- (1996) Fertilization and development of
raphy and the correlation between individual mouse oocytes cryopreserved using a theoreti-
cellular freezing events and temperature cally optimized protocol. Hum Reprod 11:
changes in the surrounding tissue. Cryobiology 12961305
11:185191 15. Karlsson JOM (2001) A theoretical model of
8. Pitt RE, Chandrasekaran M, Parks JE (1992) intracellular devitrification. Cryobiology 42:
Performance of a kinetic model for intracellular 154169
16. Toner M, Cravalho EG, Karel M (1990) 22. Guenther JF, Seki S, Kleinhans FW etal (2006)
Thermodynamics and kinetics of intracellular Corrigendum to Extra- and intra-cellular ice
ice formation during freezing of biological formation in Stage I and II Xenopus laevis
cells. J Appl Phys 67:15821592 oocytes. Cryobiology 53:294
17. Chambers R, Hale HP (1932) The formation 23. Stott SL, Karlsson JOM (2009) Visualization
of ice in protoplasm. Proc R Soc Lond B 110: of intracellular ice formation using high-
336352 speed video cryomicroscopy. Cryobiology 58:
18. Luyet BJ, Gibbs MC (1937) On the mecha- 8495
nism of congelation and of death in the rapid 24. Grocutt P (2014) Linkam Scientific Instruments,
freezing of epidermal plant cells. Biodynamica Ltd., Surrey, UK.Personal communication
1:118 25. Seki S, Mazur P (2010) The temperature and
19. Asahina E (1953) Analysis of the freezing pro- type of intracellular ice formation in preim-
cess of living organisms. X.Freezing process of plantation mouse embryos as a function of
egg cell of sea-urchin. Low Temp Sci 10:8192 the developmental stage. Biol Reprod 82:
20. Smith AU, Smiles J (1953) Microscopic studies 11981205
of mammalian tissues during cooling to and 26. Aalen OO, Borgan , Gjessing HK (2008)
rewarming from -79 C.J R Microsc Soc 73: Survival and event history analysis: a process
134139 point of view. Springer, NewYork
21. Toner M, Cravalho EG, Karel M etal (1991) 27. Scheiwe MW, Krber C (1987) Quantitative
Cryomicroscopic analysis of intracellular ice cryomicroscopic analysis of intracellular freez-
formation during freezing of mouse oocytes ing of granulocytes without cryoadditive.
without cryoadditives. Cryobiology 28:5571 Cryobiology 24:473483
Chapter 8
Laser Scanning Microscopy in Cryobiology

Frank Stracke, Asger Kreiner-Mller, and Heiko Zimmermann
Abstract
Laser scanning microscopy is emerging as a powerful imaging tool in cryobiology. The basic microscopy
system can be combined with various imaging modalities including Raman spectroscopy, fluorescence
correlation spectroscopy, fluorescence lifetime imaging, or multiphoton imaging. Multiphoton imaging
can be used to study intracellular ice formation at the subcellular level. A Raman imaging modality can be
used for chemical mapping of frozen samples. A Raman spectrum gives information about characteristic
molecular vibrations of specific groups in molecules. Raman images can be used to determine the localiza-
tion of intra- and extracellular constituents and the various forms of water in freeze-concentrated solutions.
Spectra can be collected during freezing and thawing of a sample using a temperature-controlled sample
holder. In this chapter, various advanced cryoimaging methods are described. Special emphasis is given on
the different imaging modalities that can be used to study the various aspects of cryopreservation.
Key words Cryoimaging, Laser scanning microscopy, Raman microspectroscopy, Fluorescence

imaging
1 Introduction
Little is known about processes taking place in cells during

cryopreservation processing. Established methods to study cellular
dynamics are based on their immersion in water, and cannot be
used to study frozen samples. Cryopreservation studies often rely
on experiments conducted before and after cryopreservation (inter
alia [16]). This yields fractions of viable cells, or cells with reduced
functionality as a result of the accumulated impacts during the
preservation procedure. It is not possible to derive the specific
moment causing cell death post-thawing.
Since the introduction of the first computer-controlled cryo-
microscopy chamber, cryomicroscopy has allowed investigation of
the entire cryopreservation procedure without disturbance of the
sample [7]. Cryomicroscopy, however, is restricted to wide-field
transmission imaging and the information content is rather low
and exclusively morphologic.
229
230 Frank Stracke et al.
Modeling and predicting cellular behavior und medium

properties by extrapolation of ambient temperature observations
to low temperature is not possible because several phase transitions
take place in the surrounding medium [8, 9], the cellular mem-
branes [10, 11], and biomacromolecules [1214]. Such transitions
imply substantial changes in the material properties. The homog-
enous medium turns into a complex multiphase system after freez-
ing. The composition and distribution of the different phases
depend on the initial medium composition and the cooling proto-
col. Furthermore, kinetic hindrance of thermodynamically driven
processes (supercooling, supersaturation, glass formation) is
becoming important at low temperatures, leading to metastable
domains in the sample. Hence, the preserved cell responds in a
completely different way to a completely different environment
compared to a cell under physiologic conditions.
A perfect examination technique would allow investigating
cellular state and functioning as well as the physical and chemical
characterization of the solidified surrounding. With the recent
adoption of laser scanning techniques into cryomicroscopy
[8, 1519] it is now possible to exert functional investigations dur-
ing all phases of the cryopreservation cycle. The principle of image
formation in laser scanning microscopy (LSM) is the same for
different LSM systems: A focused laser beam is scanned over a
sample (Fig. 1). At each position (x, y, z) a measurement of a par-
ticular interaction (I) of incident light and the sample is carried
out, which is then reconstructed to form an image I (x, y, z).
Various interactions determine the image contrast: reflected light,
fluorescence, Raman scattering [20], and some nonlinear effects
[21]. A vast number of fluorescent markers and indicators make
sample properties accessible on a microscopic level. LSM enables
cryobiologists to noninvasively explore entirely new aspects of cel-
lular cryopreservation processes. The physical basics of LSM in
cryobiology are the same as at room temperature. So the reader is
referred to the literature [2224]. Both fluorescence [1517] and
Raman scattering [8, 18, 19] are useful techniques in cryobiology.
Fluorescence can be used to visualize ice formation within and out-
side the cell [17], to highlight cellular compartments [17] and to
explore the intercellular cohesion in cryopreserved tissue models
[25]. Due to its superior sample penetration depth and low bleach-
ing the two-photon microscopy (TPM) is one of the best fluores-
cence imaging modalities (Fig. 2). In TPM, two photons of half
the required energy are absorbed simultaneously to successively
generate fluorescence (a nice introduction into nonlinear LSM can
be found in [21]). A simultaneous absorption of two photons is
very improbable and only takes place using high excitation intensi-
ties. This is achieved by means of a pulsed laser with a pulse length
on the order of 100 fs and a strongly confined focus due to high
numerical aperture optics. The effect only occurs in the minute
Laser Scanning Cryomicroscopy 231
Fig. 1 Excitation beam path and image formation principle of laser scanning
microscopy: The excitation laser is coupled in by a dichroic beam splitter and
tilted by means of electromechanic mirrors. Deliberate optics transforms the tilt
angle into the aimed x- and y-deflection at the focal plane. The collected emis-
sion light goes the reverse way and passes the dichroic beamsplitter towards the
detection (left out here)
focal volume and hence fluorescence is generated only there and

no confocal pinhole is needed to associate any fluorescence signal
to the focal volume. Confocal Raman microscopy (CRM) uses a
standard confocal concept, but with Raman scattering instead of
fluorescence. Emitted light by inelastic Raman scattering is fre-
quency-shifted compared to the incident light [20]. The frequency
shift equals the vibration frequencies of the scattering molecule.
Hence a Raman spectrum contains a fingerprint of the molecular
vibrations inside the focal volume enabling chemical imaging.
CRM can either be performed fast using only a single Raman band
in order to map the concentration of a single compound or more
slowly in the spectral imaging mode, where each pixel contains an
entire Raman spectrum for subsequent data analysis. CRM is nicely
suited to map the phase texture of frozen samples and to monitor
the intracellular chemistry during freezing and thawing.
Fig. 2 Detection beam path and spatial resolution principle of confocal (left each) and nonlinear microscopy
(right each): (a) Photons from the focal volume are detected properly in both cases. The focal spot is imaged
onto the confocal pinhole. (b) No photons from outside the focal volume are detected in both cases. In nonlin-
ear microscopy simply no photons are generated there. In confocal imaging all light generated outside the
focal volume is directed away from the pinhole aperture and rejected from detection. (c) Photons from the focal
volume which are scattered in the specimen are lost for confocal microscopy, since they left the imaging
course of beam. In nonlinear microscopy also scattered photons can be detected, since no confocal pinhole
rejects them. They originate from the focal volume anyway. This is one of the main reasons for the enhanced
penetration depth of nonlinear microscopy into turbid samples
The key to proper microscopic investigations of cryopreserva-

tion processes is to mimic the freezing conditions during biobank-
ing in the microscopy cryochamber as much as possible, particularly
with respect to thermal gradients, volume effects, surface effects
and sample handling. Cryo-LSM is not a routine technique. First
of all, the appropriate contrast method and image modality must
be chosen. The choice of laser wavelength, filter sets, detector, and
acquisition parameters have to be made according to the experi-
mental requirements. In this chapter we give a checklist for choos-
ing the right laser scanning technique and some other helpful tips
for practical cryo-LSM.
2 Materials
2.1 Equipment 1. A laser scanning microscope for cryomicroscopy should be

equipped with the appropriate light sources, filter sets, and
detectors for the respective experiment. All the main microscopy
companies offer at least one confocal fluorescence microscopy

system, and most of them also setups for nonlinear microscopy.
The TPM-micrographs used here have been recorded on a
Zeiss 510 META NLO (Carl Zeiss, Jena, Germany) using a
Chameleon ultra-tunable Titanium:Sapphire-laser (Coherent,
Santa Clara, USA) as excitation source. For confocal Raman
microscopy sample scanning is usually employed instead of
beam scanning since image acquisition is rather slow here. A
combination of a polychromator and a line camera are used as
spectral resolving detector. Raman-LSM is offered by compa-
nies like Witec GmbH (Witec GmbH, Ulm, Germany) or
Renishaw Ltd. (Renishaw Ltd., Wotton-Under-Edge, UK).
The confocal Raman micrographs shown here are recorded
with a home-built setup comprised of a Nikon Eclipse LV100
upright microscope (Nikon, Tokyo, Japan), a continuous wave
Coherent Compass 315 M 532 nm excitation laser (Coherent,
Santa Clara, USA), a Razor Edge adapted filter set (Semrock,
Rochester, USA) and a spectrograph assembly (Andor Shamrock
303i, Andor iDus 409, Andor Ltd., Belfast, UK). The col-
lected Raman emission is guided to the detector by means of a
50 m fiber. The front aperture of the fiber acts as a confocal
pinhole. Sample scanning is done by a piezo scanning stage
(Physik Instrumente GmbH & co, Karlsruhe, Germany).
2. The selection of computer-controlled microscopy chambers
for temperatures down to 180 C applicable for biology use
is limited. Most studies are carried out using chambers from
Linkam Ltd. (Linkam Ltd. Tadworth, UK). Here we used the
Linkam MDS600 model. The Linkam chambers are not com-
patible with inverted microscopes. This problem can be over-
come using an objective inverter (LSMtech, Wellsville, USA),
a device for inverting the illumination direction. It is mounted
into the objective thread and transfers the scanning movement
1:1 to an upright sample position next to the microscope. For
high-resolution scanning one will need to remove the outer
window of the chamber since the required objectives have
short working distances. Imaging can be performed using high
NA air objectives (e.g., 20, NA 0.65). This has two implica-
tions related to the ambient air reaching the inner window
now: (1) air humidity will form ice on the window and (2) a
temperature gradient over the sample. Since the temperature
sensor is located in the cryochambers cooling block (usually a
large piece of silver) the actual sample temperature will be a
few degrees higher than the sensor tells you. The temperature
deviation is relatively small here, but it can be difficult to keep
the air humidity away. When using an oil objective (e.g., 40,
NA 1.3) air humidity on the sample is not a problem. Ice will
form at the window and the objective, but not in the optical path.
The temperature difference between sample and sensor will be

significant, since the oil contact transfers some heat from the
microscope body to the sample. See Notes 110 on possible
implications and tips.
3 Methods
3.1 Image Modalities There is no standard procedure for cryo-LSM. Samples should be
to Study Cryobiology prepared closely matching the conditions used in the cryopreserva-
tion protocol. Translate the freezing procedure to an appropriate
cooling program of the cryo-microscopy stage. Use an optical
setup that gives the desired information with preferably low inter-
ference and background. In the following we will give a checklist
to determine the proper LSM modality for a specific cryobiology
imaging task.
1. A wide-field transmission microscope system can be used to
investigate morphology and freezing-induced morphological
changes only in a qualitative manner. This is the easiest and
fastest way to obtain images of your sample.
2. Quantitative morphological studies that are three dimension-
ally resolved should be done using a scanning fluorescence
microscopy (confocal or TPM). Use a vital cytoplasmic dye
(e.g., fluorescein diacetate) and a simple non-membrane-
permeable stain (fluorescent dextrans work fine) for the
medium. One can nicely study ice recrystallization and cell vol-
ume this way (Fig. 3).
3. Fast dynamics, i.e., processes on a time scale below minutes,
can be studied using Raman spectroscopy. Keep in mind, how-
ever, that Raman scattering is a weak phenomenon requiring
some integration time. Use Raman spot measurements instead
of full imaging or only image a single compound specified by a
single wavelength.
4. Studies on the distribution of constituents in a sample should
be done using Raman microscopy. This can be used to extract
quantitative concentrations using appropriate calibration
curves (Fig. 4).
Fig. 3 (continued) of an unstained L929 cell. The endogenous fluorescence of

(predominantly) NAD(P)H is employed for contrast in this example. The fluores-
cence light is analyzed and color-coded for its decay time (fluorescence lifetime
imaging FLIM, see [17] for experimental description). A slight increase of the
NAD(P)H fluorescence decay time with decreasing temperature is found (which
is not spectacular or surprising in any way, but exemplarily shows how fluores-
cence may be used to gain information beyond morphology)
Fig. 3 Examples of two-photon excited fluorescence microscopythe most

popular nonlinear imaging technique (All samples with 10 % Me 2SO):
( a ) Mesenchymal stem cells from the umbilical cord, frozen down to 80 C at
a cooling rate of 100 K/min. Such rapid cooling leads to the formation of fine-
grained intracellular ice, which can be nicely studied this way. Vital staining was
carried out using calcein (AM-ester) for the cytoplasm and Hoechst 33342 for the
nuclei prior to cryopreservation. (b) Moderate ice formation inside an L929
mouse fibroblast after cooling the sample down to 50 C at a cooling rate of
1 K/min. Displacement of mitochondria is visible here. Mitochondria were vitally
stained by rhodamine 123. (c) A temperature sequence (30 C, 50 C, 70 C)
Fig. 4 Confocal Raman microscopy of two L929 cells. (ac) CRM of ice, cellular
matter (from the integrated CH2 band), and hydrohalite (NaCl2H2O). Hydrohalite is
a crystalline compound that forms from saline at subzero temperatures [8]. Scale
bar is 10 m. Details on data processing can be found in [19]. (d) Transmission
image. Comparing the CRM to the transmission image reveals information that
cannot be gained through morphology studies, i.e. the spatial distribution of the
major compounds in the sample. (e) Exemplifying Raman spectra at three points
given in the CRM images. The band at 3,100 cm1 is used to image ice. The band
at 2,900 cm1 is used to image cellular matter. The band at 3,400 cm1 is used to
image hydrohalite. All bands are background corrected (linear link between the
integration borders) and integrated to yield the respective pixel value
5. If you want to know, if the water in your sample is crystalline

ice or glassy, then choose Raman microscopy. Ice and amor-
phous water have distinctive peaks in Raman spectra. If you
want to study physical or chemical properties like viscosity or
pH in the cytoplasm or the medium, scanning fluorescence
microscopy in combination with appropriate fluorescent indi-
cators is the correct choice.
6. By using scanning fluorescence microscopy it is also possible to
go for sophisticated fluorescence techniques like fluorescence
correlation spectroscopy (FCS), fluorescence lifetime imaging
(FLIM) [15, 17, 26] fluorescence recovery after photobleach-
ing (FRAP) and the like.
7. A very important distinction is between thin samples and vol-
ume samples. A thin sample for example is obtained when a
small droplet is sandwiched between two cover slips. All mea-
surement techniques can be used for thin samples, but not for
volume samples. When measurements are performed on larger
volumes the penetration depth of the measurement technique
has to be considered. Ordinary confocal techniques will have
interactions between the incident light and the sample in the
optical path through the sample and not only in the focus. This
leads to scattering effects, bleaching, and a strong unwanted
background signal. This is avoided in nonlinear measurement
techniques such as TPM, where the interaction is only taking
place within the focal volume. This leads to a high penetration
depth and is thus the most suitable microscopy technique for
volume samples.
4 Notes
1. When high spatial resolution is wanted, the application of oil

immersion objectives is inevitable since these enable a higher
numerical aperture compared to air objectives. The spatial res-
olution is dependent on the numerical aperturenot magnifi-
cation. We use the immersion oil from Zeiss and found it to
work well down to at least 80 C. Oil immersion has the
advantage to inherently keep the air humidity out of the opti-
cal path. At very low temperatures the heat flux via the objec-
tive may lead to the condensation of dew or even frost on the
objective back aperture. This can be avoided by flushing the
microscope stand with dry air or nitrogen is helpful. At ultralow
temperatures < 100 C the oil becomes very viscous, so moving
the objective against the sample is constricted. Due to the ther-
mal contact between sample and optics, the sample may have a
somewhat higher temperature than the cooling block of the
microscopy stage. A miniaturized thermo sensor within the
sample is a good idea to measure the exact sample tempera-

ture. Furthermore temperature changes within the sample
become visible, else hidden due to the large thermal inertia of
the cooling block. Another thing to consider with regard to
using oil objectives is that the oil might lift the sample away
from the sample holder due to capillary forces. This can be
avoided by placing a small water droplet between the cooling
block and sample. The sample will then be held stationary by
capillary forces or ice (at lower temperatures).
2. If you do not use oil immersion, then you need to keep the air
humidity away from your sample and the front optics. We
found a piece of foam insulation tube to work fine. It is sold at
any DIY market as heating pipe insulation. Choose a diameter
that fits your objective tightly and press the open end gently
against your sample window. Some piezo scanning stages are
very delicate and any force transferred from the sample holder
to the microscope objective might damage the scanning
stagecaution is advised in such circumstances. You may
furthermore flush the tubes inner with the cold and dry nitro-
gen flux from the microscopy chamber exhaust by means of a
syringe needle.
3. In order to quantify the volume fraction of the remaining
interdendritic fluid as the sample solidifies by accumulation of
fluorescence in the fluid, take care in the choice of stains. The
fluorescence increase is a very elegant measure for the fluid
volume decrease only if the dye that is used does not respond
to changes in sample properties. A couple of dyes have a
fluorescence efficiency dependent on temperature, viscosity,
polarity, oxygen content, pH and other solutes. For instance,
rhodamine B has a fluorescence efficiency strongly depending
on temperature, viscosity and polarity. Also be aware of self-
quenching, which is often observed with fluorescein. A suit-
able dye insensitive for most outer parameters over a reasonable
range of concentration is rhodamine 6G, quite uncommon in
biology but obtainable in high purity from laser dye vendors.
4. The best way to detect organic material (e.g., cellular material)
in a sample is to map the Raman CH2-stretching vibration at
2,800 cm1. If the medium contains significant amounts of
organic additives like dimethyl sulfoxide (Me2SO) or ethylene
glycol, however, the CH2-vibration no longer specifies a single
compound. Instead we suggest to use the amide-II band of
proteins at about 1,655 cm1 which is not interfered by most
medium additives (except proteins like albumin).
5. Fluorescence can be very broad banded and therefore unwanted
in Raman experiments, since this can lead to very high back-
ground or even drown the weak Raman signal. Unknown
narrow banded fluorescence might also complicate the assign-

ment of Raman bands in a given Raman spectrum. Most cell
media contain phenol-red which is weakly fluorescent, but
enough to drown the Raman signal. Prior to any imaging, the
biospecimen should therefore be transferred to a media phe-
nol-red-free media such as phosphate-buffered saline.
6. When investigating the collected Raman spectra consider
which Raman bands should be used for imaging. Keep in mind
that different compounds might have overlapping Raman
bands. This might lead to situations where two different com-
pounds cannot be distinguished with Raman microscopy.
7. Not all compounds exhibit Raman scattering; it is in fact only
compounds containing molecular bonds, and metals and single
atomic ions will as such not be visible in Raman spectra. If it is
possible we advise that before any Raman microscopy is per-
formed Raman spectra of each single compound should be
measured. This will greatly help in assignment of Raman bands
in post processing of the data. Keep in mind that different
phases of the same compound might give rise to different
Raman spectra and that temperature changes tend to slightly
shift the whole Raman spectra towards lower frequency.
8. When choosing the appropriate laser wavelength for Raman
microscopy account for several aspects: (1) The Raman scatter-
ing cross section decreases with the fourth power of the inci-
dent lights wavelength; that is, shorter is better. (2) In a
biospecimen there is often endogenous fluorescence. The exci-
tation of this autofluorescence is usually stronger at short
wavelengths; that is, longer is better. (3) In contrast to most
researchers cryobiologists are interested in water (in its differ-
ent forms). The near infrared diode laser line at 785 nm is get-
ting more and more popular for Raman techniques, in
particular because it will not evoke cellular autofluorescence.
The high frequency Raman shift of water (around 3,200 cm1)
will lead to a respective emission at about 1,050 nm. Silicon-
based detectors (like all CCD and most CMOS) are not sensi-
tive at wavelengths beyond 1,000 nm and will fail to show any
water signal. So the choice of wavelength depends on how
much autofluorescence is to be expected and how important
information on water is. You will likely end up with a laser
wavelength somewhere between 532 and 700 nm.
9. Keep in mind that the probing volume is not infinitesimal but
some hundred nm in x and y and m sized in z, depending on
the size of the objectives focal volume. If you have a structure
(e.g., a cell) and want to probe its chemical composition by
CRM, at least at the structure boundary you will have signal
contributions of the surrounding medium. This is particularly
the case for adherent cells. Apart from the nucleus adherent
cells are very flat. It is not possible to get a pure intracellular
Raman spectrum in such cases. This exemplifies the limitation
of the spectral resolution.
10. For safety purpose: Think about the amount of liquid nitrogen
as well as dry nitrogen for flushing and what volume fraction of
your lab it might fill as a gas. If only liquid nitrogen is used,
you may evaporate a (liquid) volume of about 0.25 % of your
lab volume in order to keep at least 18 % oxygen to breathe.
Lower oxygen concentrations start to be dangerous. Dangerous
situations may be prevented through the use of oxygen alarms.
References
1. Ehrhart F, Schulz JC, Katsen-Globa A et al 10. Bagatolli LA, Gratton E (2000) Two photon
(2009) A comparative study of freezing single fluorescence microscopy of coexisting lipid
cells and spheroids: towards a new model sys- domains in giant unilamellar vesicles of binary
tem for optimizing freezing protocols for phospholipid mixtures. Biophys J 78:290305
cryobanking of human tumours. Cryobiology 11. Glfke C, Akhoondi M, Oldenhof H et al
58:119127 (2012) Cryopreservation of platelets using
2. Holm F, Strom S, Inzunza J et al (2010) An trehalose: the role of membrane phase behav-
effective serum- and xeno-free chemically ior during freezing. Biotechnol Prog 28:
defined freezing procedure for human embry- 13471354
onic and induced pluripotent stem cells. Hum 12. Somero GN (1995) Proteins and temperature.
Reprod 25:12711279 Annu Rev Physiol 57:4368
3. Reubinoff BE, Pera MF, Vajta G et al (2001) 13. Smeller L (2002) Pressure-temperature phase
Effective cryopreservation of human embry- diagrams of biomolecules. Biochim Biophys
onic stem cells by the open pulled straw vitrifi- Acta 1595:1129
cation method. Hum Reprod 16:21872194 14. Hatley RMH, Franks F (1993) Cold destabili-
4. Donnelly ET, McClure N, Lewis SE (2001) sation of enzymes. Faraday Discuss
Cryopreservation of human semen and pre- 93:249257
pared sperm: effects on motility parameters 15. Stark M, Drr D, Ehlers A et al (2007)
and DNA integrity. Fertil Steril 76:892900 Multiphoton imaging and fluorescence life-
5. McGinnity DF, Soars MG, Urbanowicz RA time studies on unstained cells and tissues at
et al (2004) Evaluation of fresh and cryopre- cryogenic conditions. Proc SPIE 6628:
served hepatocytes as in vitro drug metabolism 662809662834
tools for the prediction of metabolic clearance. 16. Ishiguro H, Horimizu T (2008) Three-
Drug Metab Dispos 32:12471253 dimensional microscopic freezing and thawing
6. Schneider S, Klein HH (2011) Preserved insu- behavior of biological tissues revealed by real-
lin secretion capacity and graft function of time imaging using confocal laser scanning
cryostored encapsulated rat islets. Regul Pept microscopy. Int J Heat Mass Trans 51:
166:135138 56425649
7. Diller KR, Cravalho EG (1971) A cryo- 17. Drr D, Stark M, Ehrhart F et al (2009)
microscope for the study of freezing and thaw- Multiphoton microscopy for the in-situ obser-
ing processes in biological cells. Cryobiology vation of cellular processes and integrity in
7:191199 cryopreservation. Biotechnol J 4:12151221
8. Kreiner-Mller A, Stracke F, Zimmermann H 18. Dong J, Malsam J, Bischof JC (2010) Spatial
(2013) Confocal Raman microscopy as a non- distribution of the state of water in frozen
invasive tool to investigate the phase composi- mammalian cells. Biophys J 99:24532459
tion of frozen complex cryopreservation 19. Kreiner-Mller A, Stracke F, Zimmermann H
media. CryoLetters 34:248254 (2014) Hydrohalite spatial distribution in fro-
9. Cocks FH, Brower EH (1974) Phase diagram zen cell cultures measured using confocal
relationships in cryobiology. Cryobiology 11: Raman microscopy. Cryobiology 69(1):4147
340358
20. Raman CV, Krishnan KS (1928) A new type of 24. Dieing T, Hollricher O, Toporski J (eds)
secondary radiation. Nature 121:501502 (2010) Confocal Raman microscopy. Springer,
21. Zipfel WR, Williams RM, Webb WW (2003) Berlin, Heidelberg
Nonlinear magicmultiphoton microscopy 25. Beier AFJ, Schultz JR, Drr D et al (2011)
in the life sciences. Nat Biotechnol 21: Effective surface-based cryopreservation of
13691377 human embryonic stem cells by vitrification.
22. Beiser L (2003) Unified optical scanning tech- Cryobiology 63:175185
nology. Wiley, New Jersey 26. Knig K, Uchugonova A, Breunig HG (2014)
23. Masters BR (2006) Confocal and multiphoton High-resolution multiphoton cryomicroscopy.
excitation microscopy. SPIE Press, Bellingham Methods 66:230236
Chapter 9
Low-Temperature Electron Microscopy: Techniques

and Protocols
Roland A. Fleck
Abstract
Low-temperature electron microscopy endeavors to provide solidification of a biological specimen by cooling
with the aim of minimal displacement of its components through the use of low temperature as a physical
fixation strategy (Steinbrecht and Zierold, Cryotechniques in biological electron microscopy. Springer-
Verlag, Berlin, p 293, 1987). The intention is to maintain confidence that the tissue observed retains the
morphology and dimensions of the living material while also ensuring soluble cellular components are not
displaced. As applied to both scanning and transmission electron microscopy, cryo-electron microscopy is a
strategy whereby the application of low-temperature techniques are used to reduce or remove processing
artifacts which are commonly encountered in more conventional room temperature electron microscopy
techniques which rely heavily on chemical fixation and heavy metal staining. Often, cryo-electron micros-
copy allows direct observation of specimens, which have not been stained or chemically fixed.
Key words Electron microscopy, Cryo-electron microscopy, SEM, TEM
1 Introduction
1.1 Use of Physical Water is by far the most abundant cellular constituent and this
Fixation of Biological presents a fundamental challenge for electron microscopy (EM)
Material by Freezing where the electron beam is generated under vacuum at pressures
to Preserve and temperatures nominally incompatible with liquid water.
Ultrastructure Traditionally, this has been overcome by converting the live
hydrated tissue to an anhydrous stable state via a series of steps
including; chemical fixation, alcohol dehydration, and resin infil-
tration (for a range of room temperature protocols, refer to
Glauert and Lewis [1]). Although well accepted in biological
research, these steps introduce processing artifacts (perturbations
to structure). These include effects on tissue structure caused by
shrinkage or swelling of tissue [2], shrinkage of cellular organelles,
and extraction or redistribution of cellular constituents such as
lipids (e.g., for samples fixed with glutaraldehyde lipids may be
redistributed into whorled bodies which have been misinterpreted
243
244 Roland A. Fleck
as organelle), proteins, and DNA. Even the relatively simple

negative-staining technique [3], routinely employed for the visu-
alization of small particles (e.g., viruses), can perturb structure
and introduce apparent pleomorphic forms [4, 5]. Negative stain-
ing is influenced by the size and surface properties of the structure
as well as by the properties of supporting films which limit the
value of this approach for 3D reconstruction of structures [68].
These can be overcome by the adoption of specialist cryotech-
niques. [4, 5] In addition, most chemical fixatives react with pro-
teins and crosslink peptide chains as part of their action [1]. These
reactions can be highly deleterious to epitopes and therefore com-
promise immunohistochemical studies. The preservation of bio-
logical structure in a state as close to life as possible is best
achieved by fixation of all components of the sample at the same
time. From this perspective, the use of chemical fixatives is not
desirable due to the time taken to diffuse fixatives through the
sample and the specificity of the chemical cross-links formed [9].
An increasingly valuable approach to avoid the processing arti-
facts described above for both scanning (SEM) and transmission
(TEM) electron microscopy and able to fix cellular constituents
without introducing significant structural alterations is cryo-
fixation. Cryo-fixation has two distinct advantages over chemical
fixation. Firstly, it is rapid (measured in milliseconds), providing
confidence that the sample is preserved in a close-to-life state at
the point of initiating fixation, and secondly it ensures simultane-
ous immobilization of all macromolecular components [10]. As
many protein networks are labile and prone to disruption with
even a slight osmotic or temperature change, cryo-fixation mini-
mizes these deleterious effects. Thus, cryotechniques allow the
study of biological samples with improved ultrastructural preserva-
tion and can facilitate the study of dynamic processes. Rapid freez-
ing can produce superior preservation of epitopes and prevents
artifactual aggregations of proteins and tissue fixation, and pro-
cessing can be performed without (or with minimum) use of cross-
linking fixatives producing samples with higher antigenicity
[1113]. Thus, the adoption of rapid freezing can avoid the com-
mon perturbations to structure caused by chemical fixatives pre-
serving structures, hydrated, in a state close to life.
To achieve this level of preservation, successful cryo-fixation
relies on vitrification (the transformation of water from a liquid to
an amorphous state without inducing the nucleation of ice crys-
tals). Poor fixation will result in reticular ice crystal formation
which can disrupt ultrastructure and create structural artifacts. The
nucleation of ice crystals is temperature and pressure dependent,
and crystallization depends on cooling rate (dependent on the
thermal properties of water, the sample thickness, and the heat
extraction flow at the surface of the specimen) as freezing is a time-
dependent process.
Low Temperature Electron Microscopy 245
As with any complex technique, cryo-EM techniques have

often been regarded as highly specialist and technically challeng-
ing. For these reasons only a small number of advanced laborato-
ries have routinely adopted cryo-EM procedures. This is, however,
changing as the importance of close-to-life preservation with
minimal artifacts has increased with the advent of techniques such
as electron tomography, which allow visualization of three-
dimensional structures at the macromolecular level [9, 1416].
However, it would be remiss not to consider the differences
between cryopreservation for the maintenance of viability after
thawing and cryo-fixation for EM. For preservation, high levels of
post-thaw viability (>90 %) are the principal goals, and changes to
ultrastructure and/or cellular constituents are acceptable compro-
mises in the search for good preservation. Indeed, slow-cooling
protocols for cryopreservation rely to a degree on the concentra-
tion of intracellular components by cryodehydration with the
resulting intracellular space being more readily vitrified due to a
lower % volume of water. For EM, perfect retention of ultra-
structure and cellular constituents is the only goal.
There are a variety of methods in use for rapid freeze fixation
of tissues. The depth of vitrification that can be achieved by most
methods, however, is very limited. Methods that involve plunging
the sample into liquid cryogen (vitreous thin film/plunge freezing
and immersion freezing) are generally limited to a few microns of
vitrified sample. Penetration of vitrification can be improved by
methods such as cold metal block freezing (slam freezing) or
freezing by spraying the sample with liquid cryogen (jet freezing)
but are still likely to only reach depths of a few tens of micrometers
[9, 17]. Slam freezing or immersion freezing combined with freeze
fracture and deep etching has been used to produce high-quality
structural information (for examples see [1723]). However, the
depth of good freezing in these methods is severely limited.
Currently the most common cryo-fixation methods are plunge
freezing for small samples (e.g., viruses and proteins) and high-
pressure freezing for larger samples (e.g., cell suspensions and tis-
sues). The protocols discussed below will use one or other of these
physical fixation methods, however, could be adapted to fix mate-
rial using alternative cryo-fixation devices (Fig. 1).
1.2 High-Pressure The idea of freezing biological samples under pressure for electron
Freezing (HPF) microscopy was first introduced by Moor and Riehle in 1968 [24]
and high-pressure freezing (HPF) devices became commercially
available in 1985. All high-pressure freezing machines currently
available, despite having different designs, deliver synchronized
pressurization and cooling of the sample within 20 ms in a highly
reproducible manner [14, 25, 26]. Freezing larger volumes of tis-
sue requires high pressures to be exerted on the sample, in combi-
nation with cryogenic temperatures. High-pressure freezing (HPF)
246 Roland A. Fleck
Room Temperature Low Temperature Specimen Handling and Processing

Specimen Handling
Warm/ Live Sample Plunge Freezing (VTF)

Physically Fixed Vitrified Specimen Fixation
Propane Jet Freezing
Slam Freezing Freeze Fracture
Chemical Fixation High Pressure Freezing
Cryo-ultramicrotomy
CEMOVIS
Freeze-substitution
Processing
Ultramicrotomy
Replica
Conventional Room
temperature Sample
Preparation Techniques Tokuyasu/
Immunolabelling
Technique
Imaging by either SEM or TEM at conventional imaging conditions of Imaging by either SEM or TEM under Visualization
temperature and pressure cryogenic conditions
Fig. 1 Low-temperature preparation workflows. Areas to the left of the dashed line are at room temperature,
indicating that samples can be handled without the requirement for specialist cryo-operating procedures.
Samples to the right of the dashed line must be handled in a manor suitable for vitreous biological samples
with specific care taken to pre-cool tools before approaching or manipulating samples. Arrows denote the
direction of workflow and key stages required to process a sample from living/close to life state to final
image
can extend the depth of vitrification of the sample to 200 m [15,

27]. At sufficiently high pressures (210 MPa), the cooling rate
required for the vitrification of water is reduced from several
100,000 K/s to a few 1,000 K/s, making vitrification of relatively
thick samples practicable [15]. Good heat conduction through the
sample and between the sample and the carrier upon which it is
mounted is vital for effective freezing. Air-filled spaces between the
tissue and the sample holder used for freezing can reduce the freez-
ing efficiency by acting as insulators. Furthermore, air-filled spaces
are likely to collapse when high pressures are applied, deforming
the tissue. To solve these problems, sample holders are often loaded
with cryoprotectant fillers, acting as both a cryoprotectant to
improve freezing and protecting the sample against pressure-
induced shearing [28]. Fillers are often substances that cannot
penetrate the tissue and have low osmotic activity (e.g., solutions
of high-molecular-weight (HMW) dextran or the inert solvent
1-hexadecene (CH3(CH2)13CH=CH2)). Alternatively cryoprotec-
tants that penetrate the tissue, for example, sucrose, can be used.
The use of penetrating cryoprotectants can be undesirable, as their

penetration into the tissue can lead to structural changes. However,
in some cases this penetration seems to have no effect, probably
due to diffusion barriers preventing entry to the tissue during the
short period of exposure to the cryoprotectant [15, 29, 30].
1.3 Freeze After freezing, samples can be handled in a variety of ways and
Substitution examined either at cryogenic temperatures or at room temperature
after additional processing. Freeze substitution (FS) is a common
technique for the low-temperature dehydration of samples, before
embedding in resin and examination at room temperature. Water
in the samples is replaced by a solvent (usually acetone) at a tem-
perature around 90 C, a temperature lower than the lowest tem-
perature (70 C) at which secondary ice crystals have been shown
to form in biological samples [31]. Fixatives and heavy metal,
electron-dense stains such as osmium tetroxide and uranyl acetate
are usually added to the sample to stabilize the frozen structure
and improve image contrast. Because chemical fixation occurs
when the sample is already frozen, there is no liquid water present
and unwanted osmotic effects should not occur [32]. Once dehy-
dration and infiltration of fixatives have occurred, the sample is
slowly warmed to a suitable temperature for resin embedding.
Although HPF/FS protocols eliminate many of the artifacts
associated with conventional fixation and embedding, the process
of rapid freezing and substitution can itself cause artifactual changes
to the tissue, and protocols used for HPF/FS must be tailored
specifically to the tissue in question [33]. Freezing artifacts, such as
the growth of small ice crystals within the tissue due to insuffi-
ciently rapid freezing or subsequent warming, are a common prob-
lem. Ice crystals formed during freezing cause dehydration
(cryodehydration) in the surrounding cytoplasm, and solutes are
concentrated as water is recruited to the growing ice crystal. After
FS these effects are observed as characteristic holes in the sample
(from the ice crystal branches) and aggregates of solute in the sur-
rounding cytoplasm. In severe cases these changes in solute con-
centration can lead to the rupture of membranes [28, 34]. Where
growth of small ice crystals has occurred, a reticulated pattern is
also often visible in the sample [9].
2 Materials
Prepare all aqueous solutions using glass-distilled water. Prepare

and store all reagents as small aliquots (110 ml) and unless oth-
erwise stated store at 4 C for short periods (>10 days) or 20 C
for longer periods (<10 days). Indicative shelf lives for reagents
are provided; however, the use of freshly prepared solutions is
recommended.
248 Roland A. Fleck
2.1 Physical Fixation 1. Liquid nitrogen (LN).

by Low Temperature: 2. Ethane gas (CP grade) or propane gas (see Note 1).
Vitreous Thin Films
3. TEM grids.
(VTF)
2.2 High-Pressure 1. LN.

Freezing Using 2. Isopropanol (Sigma Chemical Company, St. Louis, MO, USA).
a Baltec HPM010/Abre
3. Specimen carrier planchets (Leica microsystems, Vienna,
HPM010: Operation
Austria; Wohlwend, GmbH, Switzerland).
of the HPM010
2.3 Sample 1. 2040 % HMW dextran (if required as a filler or cryoprotec-
Preparation for High- tant or for CEMOVIS to provide a vitreous sample with cor-
Pressure Freezing rect cutting constancy necessary for cryo-ultramicrotomy and
subsequent imaging of ultrastructure in the fully hydrated
state). 20 % high-molecular-weight dextran (w/v) (Sigma
Chemical Company, St. Louis, MO, USA) prepared by adding
2 g HMW dextran to 10 ml dH2O or cell media. Vortex until
dextran has dissolved and store (4 C) in aliquots. 40 % HMW
dextran (w/v) prepared by adding 4 g HMW dextran to 10 ml
dH2O or cell media. Vortex until dextran has dissolved.
Dissolving dextran at 40 % (w/v) is protracted and may be
helped by gentle heating in a water bath.
2. 1-hexadecene (CH3(CH2)13CH=CH2) (Fluka/Sigma
Chemical Company, St. Louis, MO, USA) (if required).
2.4 Reagents 1. 4 % (v/v) formaldehyde (FA) stock solution, dilute ampoule of

for Tokuyasu 16 % FA (Polysciences Inc, Warrington, PA, USA) 1:4 with
Technique PBS (2 ml 16 % FA and 6 ml PBS).
for Immunolabeling 2. 4 % FA 0.2 % (v/v) glutaraldehyde (GA), dilute ampoule of
of Biological Samples 8 % GA (Polysciences Inc, Warrington, PA, USA) 1:40 and
for TEM 16 % FA 1:4 with PBS (2 ml 16 % FA, 0.2 ml 8 % GA, and
5.8 ml PBS).
3. 1 % GA in phosphate-buffered saline (PBS), dilute ampoule of
8 % GA 1:8 with PBS (1 ml GA and 7 ml PBS).
4. 1 % (w/v) gelatin (Sigma Chemical Company, St. Louis, MO,
USA), mix 0.1 g gelatin with 10 ml PBS on a hot plate (60 C)
until solution is clear. Cool (37 C) and continue mixing.
Store at 4 C.
5. 2 % (w/v) gelatin, mix 0.2 g gelatin with 10 ml PBS on a hot
plate (60 C) until solution is clear. Cool (37 C) and continue
mixing. Store at 4 C.
6. 15 % (w/v) gelatin, mix 1.5 g gelatin with 10 ml PBS on a hot
plate (60 C) until solution is clear. Cool (37 C) and continue
mixing. Store at 4 C.
7. 2.3 M sucrose (Sigma Chemical Company, St. Louis, MO,

USA), dissolve 78.73 g sucrose in 70 ml PBS in a volumetric
flask, once dissolved top up the volume to 100 ml with
PBS. Store at 4 C in 10 ml aliquots.
8. 2 % (w/v) methyl cellulose (MC) (Sigma Chemical Company,
St. Louis, MO, USA), heat 196 ml dH2O to 90 C and add 4 g
MC while stirring. Cool the solution on ice, while stirring until
the solution reaches approximately 10 C. Add dH2O to a final
volume of 200 ml and stir O/N (4 C). Leave the solution to
mature (3 days, 4 C) without stirring, and then spin at
100,000 g for 1.5 h. Aliquot into 10 ml portions and store in
the dark at 4 C. Keeps for up to 90 days.
9. MC-section pickup solution (1.15 M sucrose, 1 % MC), mix
2.3 M sucrose and 2 % methyl cellulose 1:1, keep on ice at all
times.
10. 0.15 % (w/v) glycine, add 0.15 g glycine to 100 ml PBS and
mix until dissolved. Store at 4 C.
11. Blocking buffer, mix Bovine Serum Albumin (BSA) (1 ml),
0.1 ml Tween 20, and 1 ml normal animal (same animal source
as used to raise 1 antibodies) serum to 100 ml PBS or
TBS. Store at 4 C.
12. UA/MC solution, add 9 ml 2 % MC to 1 ml aqueous UA
(4 %) and mix gently. Centrifuge for 95 min at 35,280 g at
4 C. Store in the dark (4 C, <90 days).
2.5 Freeze 1. 20 % (w/v) uranyl acetate (UA) (TAAB, Aldermaston, Berks.,

Substitution UK) in methanol, dissolve 20 g UA in methanol (>99 %).
for Improved 2. 1 % (w/v) aqueous UA, dissolve 0.1 g UA to 10 ml dH2O.
Ultrastructural
3. Hawes rapid freeze substitution and immunolabeling-specific
Preservation and/or
reagents: 0.2 % (v/v) UA in acetone (from 20 % UA in metha-
Immunolabeling nol stock), 2 % (v/v) UA in acetone (from 20 % UA in metha-
nol stock).
4. High membrane contrast-specific reagents (for improved ultra-
structure not immunolabeling): 3 % GA, 1 % OsO4 (TAAB,
Aldermaston, Berks., UK), and 0.5 % UA (in methanol).
5. Standard osmium stock reagent (if required): 15 % OsO4.
6. Standard membrane contrast media mix: 0.5 % GA, 1 % OsO4,
and 0.2 % UA 15 % H2O.
7. HM20 (TAAB, Aldermaston, Berks., UK) or other suitable
freeze-substitution resin (for all resins follow manufactures
instructions): Combine cross-linker (2.98 g) and monomer
(17.02 g) by gentle mixing (avoiding air bubbles and agitation
of the solution) with a glass rod (35 min). Add initiator
(0.10 g), and continue gently mixing until completely dis-
solved. HM20 resin may be stored (20 C, 90 days).
250 Roland A. Fleck
2.6 Freeze-Fracture 1. Cleaning solution 1; 70 % sulfuric acid (stock solution). Slowly

Replica Cleaning add 7 ml concentrated sulfuric acid to 3 ml dH2O.
2. Cl Cleaning solution 2; 40 % chromic acid solution (w/v), dis-
solve 4 g of CrO3 (Sigma Chemical Company, St. Louis, MO,
USA) in 10 ml concentrated H2SO4.
3 Methods
3.1 Physical Fixation 1. Preparation of grids: Grids need to be glow discharged or

by Low Temperature: plasma cleaned immediately before plunge freezing (e.g., Leica
Vitreous Thin Films EM ACE200 Glow Discharge, Quorum Technologies Q150R
(VTF) Glow Discharge, Fischione 1020 Plasma Cleaner). This step
reduces hydrophobicity and allows even spreading of the sam-
ple across the grid (Fig. 2).
2. Plunge freezing: Cool the plunge-freezing apparatus with liq-
uid nitrogen (LN) and position the ethane container below the
forceps. The apparatus may be a commercial system (e.g.,
Gatan CP3, Leica microsystems CPC, or EM GP or FEI
Vitrobot) or a home-built manual apparatus (please refer to
Fig. 2 The effect of plasma cleaning on wetting and spreading of a sample (4 l) on a TEM Quantifoil grid
for VTF
Grid gripped at
edge with forceps
Liquid Ethane Sample (4l)

crucible pipetted onto
TEM-grid with
holey support film
LN-Dewar
Liquid Ethane level
LN level
Fig. 3 Generic diagram showing the essential parts of a VTF plunge-freezing

instrument. The crucible retaining the liquid ethane is located in a bath of LN. The
sample is first pipetted onto a TEM grid, held above the liquid ethane on a
guillotine-type mechanism. Once blotted to leave a continuous even film of
sample/liquid on the grid, the grid is propelled into the liquid ethane. On release,
the guillotine will propel the grid with sample into the liquid ethane to a depth
sufficient to completely immerse the grid in the ethane without contacting (and
damaging/bending the grid) the base of the crucible
the operating manual for specific guidance on the correct and

safe operation of the device) (Fig. 3). Once it has reached the
desired temperature (either according to manufactures instruc-
tions or if using a homemade guillotine, when the LN has
stopped boiling), make the liquid ethane.
3. Creating liquid ethane: Ensure the ethane crucible is dry and
free from LN, open the ethane cylinder, and carefully open the
regulator.
(a) First, flush the delivery tube to ensure only ethane is deliv-
ered to the chamber and not a mix of air-ethane. To do
this place the pipette tip/hollow metal needle at the end
of the ethane tube perpendicular to the vessel containing
LN; when the ethane is gently flowing from the tip, you
will see a stream of milky-colored gas disturbing the
vapor phase above the LN.
(b) Next place the tip of the ethane delivery tube fully into the
precooled ethane crucible and fill with freshly condensed
liquid ethane (Fig. 3). Initially, the tip should be located
against the base of the crucible intended to hold the eth-
ane (this is the coldest point and will allow the gaseous
ethane to liquefy). There will be a clearly identifiable
change in the sound of dispensing gas (to a gurgling) as
252 Roland A. Fleck
the pressurized gas starts to liquefy. Use this change in

sound to gauge when the gas starts to liquefy and slowly
fill the crucible with liquid ethane (see Note 2).
4. Preparing the sample: Various sample types can be plunge fro-
zen (e.g., virus particles, lipid vesicles, proteins, bacteria); here
are a few guidelines to getting the best results (see Note 3).
5. Plunging the sample: Grip (pickup) the grid at its edge without
the forceps touching the fragile grid bars or support film
(Fig. 3). Tighten the forceps and gently tap the edge of the
forceps to ensure that the grid does not move or drop out.
Load the forceps into the plunge-freezing apparatus or guillo-
tine and secure. If using a manual plunge-freezing apparatus or
commercial system which does not regulate ethane tempera-
ture, thaw the solid ethane by adding fresh liquid ethane to the
ethane container; the ideal consistency for the ethane is vis-
cous to a consistency comparable to glycerol.
6. Mix the sample gently and remove 4 l of sample using a
pipette. Holding the pipette, as near to horizontal as possible,
pipette the sample gently onto the grid.
7. Excess liquid is then removed from the grid by blotting with
filter paper (blotting may be performed from either a single
or both sides). With automated commercial systems, blotting
time can be programmed and adjusted to ensure consistent
conditions between samples. In addition, many commercial
systems maintain and blot the sample in a high-humidity
chamber to ensure near identical blotting conditions between
samples. As a guide, blotting should be for approximately 3 s.
However, blotting times will depend on sample concentra-
tion, composition, and relative humidity, and optimization of
conditions is necessary for new samples. Once the excess liq-
uid has been removed, immediately plunge the grid into the
liquid ethane. Again this step has been built into the auto-
mated process of many commercial systems.
8. After plunging, remove the forceps from the guillotine (taking
care to maintain the sample under cryogenic conditions) and
rapidly transfer the grid into LN (Fig. 5). There can be a
droplet of ethane on the grid; this can be removed by gently
touching the grid to the side of the ethane container or by
touching the grid against an LN-cooled tip of filter paper.
Place the grid into a grid storage box by moving it under the
LN or well within the nitrogen vapor phase if using a commer-
cially available plunge freezer, ready for transfer into the micro-
scope (Fig. 4) (see Note 4).
9. Storage: Grids can be stored in grid storage boxes, with screw-
top lids, for a few months under LN. If the grids are to be shipped
to another site, ship in a dry-phase LN dewar (see Note 5).
Fig. 4 Steps in the preparation and TEM visualization of vitreous samples. (a) TEM grid held in forceps, (b)
single-side blot of a sample (4 l) pipetted onto the TEM grid, (c) plunging the sample into the liquid cryogen
(ethane), (d) transferring the vitreous sample under cryogenic conditions into a cryo-TEM grid storage box, (e)
loading the TEM cryo-rod under cryogenic conditions in the cryo-loading station, (f) inserting the cryo-TEM rod
into the TEM with the cryo-shield in the closed position, protecting the sample from contamination, (g) virus-
like particles (VLP), viewed as by VTF cryo-TEM. Scale bar represents 30 nm
3.2 High-Pressure 1. Start up: Firstly refill the alcohol (isopropanol), open the pressure
Freezing Using valve, and depressurize the alcohol system. Once the pressure
a Baltec HPM010/Abre has dropped, fill alcohol to the desired level through the filling
HPM010: Operation tube located on top of the high-pressure freezer (HPF). Refit
of the HPM010 the pressure cap (screw) to the filling tube and pressurize the
system to 3 bar.
2. Turn on the main switch and the key switch.
3. Press the air heater button and leave the machine to dry for at
least 1 h. This is important to avoid ice forming and blocking
the HPF during operation.
4. Turn off the air heater and open the LN supply valve.
5. Start the HPF cooling by pressing start and nitrogen.
6. Wait (approximately 3045 min) until the nitrogen (green)
light goes off, indicating the machine is cold and the LN dewar
is full.
7. Press the green drive in button and wait until the drive in
light has gone off; press the button marked auto and wait
for the oil pressure to rise to 300 bar.
8. Perform a test shot with the test sample holder to confirm that
the machine is operating optimally by inserting the test sample
holder and the locking pin, and ensure the sensor from the test
sample holder is connected to HPF.
254 Roland A. Fleck
Recess to hold sample
Planchet: Specimen Carrier

Type A
Sample sandwiched
between 2 planchets Recess to hold sample
200m
100m
2mm
3mm 300 m
Planchet: Specimen
Carrier Type B
Fig. 5 Planchets and planchet sandwiches for HPF of biological samples. The lower type B carrier is showing
scoring on the flat surface to grip samples during subsequent freeze-fracture sample processing.
Specimen carrier planchets of type A and B are shown with indication of the relative depth of the machined
recess. By combining flat surfaces (type B) and either 100, 200 (type A) or 300 m (type B) recesses, sample
volume of between 100 and 600 m depth are possible. In addition, 6 mm diameter planchets and slot
planchets for tissue biopsy (type C) are available for the EM-HPM 100 (Leica Microsystems) and Baltec/Abre
HPM 010
9. Press the yellow jet button to perform a freezing shot.

10. The HPF will then preform a freezing cycle, immediately after
the freeze denoted by a loud hiss and bang; rapidly remove
the locking pin and then the test head. Confirm the HPF oper-
ating values (they should be approximately pressure mainte-
nance 0.55 s, cooling time between 6 and 10 ms, and
temperature maintenance of about 4.5 s).
11. Warm and dry the test sample holder and perform a second
test shot to confirm the values (see Note 6).
3.3 Sample 1. Cell suspensions: Gently pellet cells by centrifugation, filtra-

Preparation for High- tion or if growing on agar plates or in monolayers by scraping
Pressure Freezing (this can be done in the presence of a small volume of media
or phosphate-buffered saline (PBS) or other suitable simple
buffer).
2. Pipette ~1 l of the cell suspension into each well of the speci-
men carrier planchets and assemble as a thin sandwich (Fig. 5)
(see Note 7).
3. Cell suspensions and/or live organisms (e.g., yeast, protozoa,
algae, embryos) can be frozen as above may also be aspirated
by capillary action into short segments of cellulose capillary
tubes; these are then placed into the planchet and are sur-
rounded by a filler (1-hexadecene by preference).
4. Tissue samples, small sections of tissue are placed into the

planchet and surrounded with a filler (1-hexadecene by prefer-
ence); a second planchet is placed on top of the first to form a
sandwich.
5. Load the planchet into the sample holder and screw shut
tightly. Load the sample holder into the HPM010 and lock
into place with the locking pin. Freeze as for the test shot
(Subheading 3.2, step 10). Once the sample is frozen, remove
the locking pin and rapidly transfer the tip of the sample holder
into liquid nitrogen. Under LN, unscrew the sample holder tip
and push out the frozen planchet sandwich. The frozen sample
can now be either used immediately or can be transferred to a
storage dewar until required.
6. After freezing is complete, shut down the HPM010. To
turn off the HPM 010, push the stop button then turn off the
key switch. Turn off the main switch at least 4 h after turning
off the key switch (see Note 8).
3.4 Cryo-to-Room 1. Room temperature fixation: Gently pellet cells by centrifugation

Temperature and remove supernatant. Add a small volume of fixative (4 %
Techniques: Tokuyasu formaldehyde (FA) or 4 % FA + 0.2 % glutaraldehyde (GA))
Technique and resuspend pellet gently; leave at room temperature for 1 h.
for Immunolabeling If not using cells, immediately transfer to 4 C for storage
of Biological Samples overnight (O/N)/over the weekend.
for TEM 2. Gelatin embedding: Gently pellet cells, remove supernatant,
and wash twice in cooled PBS (if cells are very clumped and
stick to each other, re-pellet cells and add 1 % bovine gelatin in
PBS at 37 C gently resuspend). Pellet cells and remove super-
natant; add 15 % gelatin and mix with straightened paper clip.
Place the Eppendorf tube into ice to set the gelatin; once set,
cut the bottom, containing cells, off the Eppendorf tube and
extract the lump of gelatin. Cut the gelatin into 1 mm3 pieces
and infiltrate with sucrose.
Alternatively, pipette cells in gelatin between two microscope
slides on Parafilm placed ~1 mm apart; place a microscope slide
on top. Place at 4 C until set; cut the gelatin into 1 mm3 pieces
and infiltrate with sucrose.
3. Sucrose infiltration: Place the cubed gelatin blocks in an
Eppendorf with 2.3 M sucrose solution. Allow sucrose to infil-
trate O/N at 4 C on a daisy wheel or shaker (or at room
temperature for a minimum time 2 h).
4. Cryo-fixation: Place the gelatin cubes onto a specimen pin;
wick off the excess sucrose with filter paper, leaving a small
amount around the block. Plunge into LN. Mount the pin in
a precooled cryo-microtome cooled to 80 C (e.g., Leica
microsystems UC6 or 7 with FC6 or 7; RMC CR-X cryosec-
tioning system).
256 Roland A. Fleck
Rotate Through 75 Rotate Through 30
Trapezoid Trimming
Square Trimming
Rotate Through 90
Fig. 6 Trimming square or trapezoid volumes for ultramicrotomy. Vertical arrows indicate cutting points for the
trimming knife. Horizontal arrows denote the direction of rotation of the specimen in the microtome. For sec-
tioning in the cryo-ultramicrotome, all adjustments are by way of a tool which allows the cooled sample block
to be rotated and locked in position while at cryogenic temperatures. Chucks for ultramicrotomy are available
with internal clamping mechanisms designed to clamp and retain samples of different geometries which have
been prepared by either high-pressure freezing of planchets or tubes, for Tokuyasu, or by freeze substitution.
Adapted from NIBSC cryoworkshop course notes
5. Sectioning: First trim the frozen blocks to form either a trap-

ezoid or square pyramid, using a 20 or 45 cryotrim (Diatome,
Switzerland) diamond knife (Fig. 6). Use a speed of 100 mm/s
and a feed of approximately 300 nm with the temperature set
to 80 C.
Light microscope (LM) sections may be collected to confirm
that the sample is being sectioned and the block has been
trimmed sufficiently to allow the sample to be accessible for
ultramicrotomy. To do this, use a diamond knife (35 cryo-
immuno knife (Diatome, Switzerland)) and cut 300 nm thick
sections at 100 to 120 C to check by LM. Use an eyelash
to guide the sections down the knife surface. Turn off the anti-
static device and pick up the cut sections with a drop of
sucrose/methyl cellulose (MC) solution (kept at 4 C) on a
loop; just before the point of freezing (when only the edges of
the drop are frosty), touch the drop onto the sections and lift
them from the knife. Place the sections on a cover slip.
EM sections, if cells are present in the LM test sections, cut
100 nm thick sections with a speed between 2 and 20 mm/s.
Pick up on sucrose/MC drop. Place thawed drop on naked or

carbon formvar-coated grid. Sections may then be processed
for immunolabeling (refer to Sarraf [35] for information on
immunolabeling strategies) (see Note 9).
3.5 Freeze Samples (cells/tissues) must first be HPF cryo-fixed to allow freeze
Substitution substitution to occur. Freeze substitution is a technique whereby a
for Improved biological sample is dehydrated from a cryo-fixed state by exchange
Ultrastructural of ice (vitreous water) against an organic solvent [36] and was first
Preservation and/or established and published by Fernndez-Morn in 1957 [37].
Immunolabeling Since then the technique has undergone a number of refinements
and modifications and is now employed for both TEM and SEM
sample preparation (e.g., TEM [38] and SEM [39]). However,
protocols employed do vary and are often developed empirically
[40]. Here a generic approach to performing freeze substitution is
described with references to established and proven protocols for a
range of biological samples (Table 1). The most effective physical
cryo-fixation method is HPF [41]. The following represents the
key steps in performing freeze substitution:
1. For suspensions of cells, gently pellet cells by centrifugation,
filtration, or if growing on agar plates or in monolayers by
scraping (this can be done in the presence of a small volume of
media or PBS). Pipette ~1 l of the cell suspension into a well
of the high-pressure freezer specimen carrier (planchet) and
assemble as a thin sandwich. It is possible to mix the cell pellet
1:1 with 2040 % HMW dextran or 1-hexadecene which acts as
a cryoprotectant and filler, respectively, which will allow
medium or thick sandwiches to be produced as the cryoprotec-
tant will ensure vitrification of thicker samples. Cell suspensions
and/or live organisms (e.g., yeast, protozoa, algae, embryos)
can be frozen as above may also be aspirated by capillary action
into short segments of cellulose capillary tubes [42]; these are
then placed into the planchet and are surrounded by a filler
(1-hexadecene by preference). For small sections of tissue, tis-
sue biopsies (collected by needle biopsy [43]), or cells grown
on substrates to maintain orientation, the samples are placed
into the planchet and surrounded with a filler (1-hexadecene in
preference); a second planchet is placed on top of the first to
form a sandwich (Fig. 5).
2. For automated freeze substitution using the Leica FSP-AFS
system, fill the dewar with LN and precool the Leica FSP-AFS
to desired start temperature. Prepare the freeze substitution
reagents (acetone, substitution media, and resin) and place
10 ml of each into plastic reagent bottles; place the bottles at
20 C until needed. Once the Leica FSP-AFS has reached the
desired temperature, load the specimen container and add
4 ml substitution media into the bottom of the container.
258 Roland A. Fleck
Table 1
Key freeze-substitution references for guidance (adapted from Leica microsystems
EM AFS Recipe Book)
Hawes, P., Netherton, C.L., Mller, M., Wileman, T., and Monaghan, P. (2007) Rapid freeze-
substitution preserves membranes in high-pressure frozen tissue culture cells. J. Microsc. 226,
182189
Ap Gwynn, I., Wade, S., Kaeaeb, M.J., Owen, G.R., and Richards, R.G. (2000) Freeze-substitution
of rabbit tibial articular cartilage reveals that radial zone collagen fibres are tubules. J. Microsc. 197,
159172
Barlow, D.I. and Sleigh, M.A. (1979) Freeze substitution for preservation of ciliated surfaces for
scanning electron microscopy. J. Microsc. 115, 8195
Bourett, T.M., Czymmek, K.J., and Howard, R.J. (1999) Ultrastructure of chloroplast protuberances
in rice leaves preserved by high pressure freezing. Planta 208, 47479
Edelmann, L. (1991) Freeze-substitution and the preservation of diffusible ions. J. Microsc. 161,
217228
Gilbert, C.S. and Parmley, R.T. (1998) Morphology of human neutrophils: A comparison of
cryofixation, routine glutaraldehyde fixation, and the effects of dimethyl sulfoxide. Anat. Record
252, 254263
Graham, L.L. and Beveridge, T.J. (1990) Evaluation of freeze-substitution and conventional
embedding protocols for routine electron microscopic processing of eubacteria. J. Bacteriol. 172,
21412149
He, Y. and Wetzstein, H.Y. (1997) Improved structural preservation and immunolocalization of the
microtubule cytoskeleton in plant and animal cells by freeze substitution/fixation. Method Cell Sci.
19, 91100
Hoch, H.C. (1986) Freeze-substitution of fungi. In: Aldrich, H.C., Todd, W.J. Eds. Ultrastructure
techniques for microorganisms. Plenum Press, N.Y., 183212
Holland, D.J., Cunningham, A.L., and Boadle, R.A. (1998) The axonal transmission of Herpes
simplex virus to epidermal cells: A novel use of the freeze substitution technique applied to explant
cultures retained on cover slips. J. Microsc. 192, 6972
Monaghan, P., Perusinghe, N., and Mller, M. (1998) High-pressure freezing for
immunocytochemistry. J. Microsc. 192, 248258
Palsgard, E., Lindh, U., and Godfried, M.R. (1994) Comparative study of freeze-substitution
techniques for X-ray microanalysis of biological tissue. Microsc. Res. Techniq. 28, 254258
Parthasarathy, M.V. (1995) Freeze-substitution. In: Galbraith, D.W., Bohnert, H.J., Bourque,
D.P. Eds. Methods in cell biology. Vol. 49, 5769. N.Y., Academic Press
Studer, D., Michel, M., Wohlwend, M., Hunziker, E.B., and Buschmann, M.D. (1995) Vitrification
of articular cartilage by high-pressure freezing. J. Microsc. 179, 321332
Thirion, S., Troadec, J.D., Pagnotta, S., Andrews, S.B., Leapman, R.D., and Nicaise, G. (1997)
Calcium in secretory vesicles of neurohypophysial nerve endings: Quantitative comparison by X-ray
microanalysis of cryosectioned and freeze-substituted specimens. J. Microsc. 186, 2834
Tiedemann, J.H., Hohenberg, H., and Kollmann, R. (1998): High-pressure freezing of plant cells
cultured in cellulose microcapillaries. J. Microsc. 189, 163171
(continued)
Table 1
(continued)
Wang, Y., Chen Y., Lavin, C., Gretz M.R. (2000) Extracellular matrix assembly in diatoms
(Bacillariophyceae). IV. Ultrastructure of Achnanthes longipes and Cymbella cistula as revealed by
high-pressure freezing/freeze substitution and cryo-field emission scanning electron microscopy.
J. Appl. Phycol. 36, 36378
Wharton, D.A. (1991) Freeze-substitution techniques for preparing nematodes for scanning electron
microscopy. J. Microsc. 164, 187196
Wilson, M.T., Farmer, M.A., and Karwoski, C.J. (1998) Ultrastructure of the frog retina after
high-pressure freezing and freeze substitution. J. Microsc. 189, 219235
Yokota, S. and Okada, Y. (1997) Quantitative evaluation of preparation procedures affecting
immunogold staining in post embedding immunochemistry. Acta Histochem Cytochem 30, 497504
Load the reagent bottles and place 10 ml ethanol around the

outside of the containers (to improve contact between the
reagents/sample and the cooling chamber). Remove the lids
from all of the bottles. Check the fluid exchange needle, as per
the manufacturers instructions. Insert the needle/syringe
into the automatic liquid transfer device (AFS) by inserting
the end of the plunger into the clamps; this should give you a
volume of 5.5 ml of air in the syringe.
3. Transfer planchets individually into the chamber of the Leica
FSP-AFS. If the planchets are firmly attached to each other,
prize them open with a precooled pick. Place one half of the
planchet sandwich sample side up either on top of or in the
substitution media frozen at the bottom of the specimen
container.
4. Attach the automatic head to the freeze substitution dewar and
lock into place. Start the run. Check the LN every 12 days
(see Note 10).
3.6 Freeze Freeze substitution without automation of liquid exchange (e.g.,

Substitution Without Leica FSP or RMC FS-8500) or a fully manual home-built sys-
Automation of Liquid tem requiring basic laboratory tools: a platform shaker, LN, a
Exchange metal block with holes for cryotubes, and an insulated container
such as an ice bucket [44].
1. Prepare and fix the sample by HPF, prize the planchets apart
under LN, and if necessary remove the cellulose capillaries.
Place the sample into a substitution solution precooled to
90 to ~ 80 C. Excellent freeze substitution results can be
obtained in as little as 90 min for cells of small volume such as
bacteria and monolayers of tissue culture cells. For cells of
greater volume or for samples that have significant diffusion
barriers such as cuticles or thick cell walls, extended substitu-
tion periods of 3 h or more are required [44] (see Note 11).
260 Roland A. Fleck
3.7 Sectioning 1. Removing planchet: trim excess resin from around the planchet,
a Freeze- using a razor blade; make sure that there is no resin around the
Substituted Sample sides of the aluminum planchet. Plunge the tip of the resin
block into LN, just to cover the planchet. After a few seconds,
remove the resin block and lever the planchet off the resin block
with a razor blade, if it does not pop off during warming.
2. Insert the block into the sample chuck and orientate it so that
it is optimal for sectioning. Ideally you want either the largest
concentration or piece of tissue or cellulose capillary orientated
vertically (or horizontally) so that, when you start trimming
the block, you are trimming away the minimum amount of
material of interest, leaving you with a pyramid that contains a
large amount of material of interest.
3. Trim the blocks to form either trapezoid using a 20 or 45
trim diamond knife. Use a speed of 100 mm/s and a feed of
approximately 300 nm (Fig. 6). Use the antistatic device on
full power during trimming to keep the knife clear.
4. Sectioning: collect LM sections first to confirm that the sample
is being sectioned. Change the diamond knife to an ultra 35
(Diatome, Switzerland) diamond knife. Fill the boat with
dH2O so that the knife surface is fully covered but not so full
that the water is bulging over the top of the boat. Cut 300 nm
thick sections to inspect by LM. Place the LM sections onto a
glass cover slip or microscope slide, dry onto a slide using a hot
plate at 70 C, place a drop of toluidine blue stain onto the
cover slip, and lightly, but not fully, dry with hot plate wash
with H2O, and check for cells on a light microscope.
5. EM sections: if cells are present in the LM sections, cut 100 nm
thick sections with a speed between 2 and 20 mm/s. Stretch
the sections, to remove any folds and creases, by passing a
chloroform-wetted swab or filter paper over the sections, allow-
ing the vapor to smooth the sections. Pick up the sections by
placing a prefect loop (or homemade loop) over a few sec-
tions and transfer to a grid (formvar-/carbon-coated by prefer-
ence). Touch a piece of filter paper to the side of the loop to
remove excess water and then carefully remove the grid from
the loop; leave to air-dry before storing or viewing. Sections can
be stained and/or immunolabeled if required [45].
6. Staining of sections to improve contrasts: float sections on 1 %
(w/v) aqueous uranyl acetate for 25 min. Rinse the sections
quickly with distilled water. Check the staining in the TEM. If
more contrast is required, the sections can be stained using lead
citrate. Dilute the lead citrate solution 1:5 to 1:1,000 times
with 0.1 N NaOH; float grid on top of droplet of stain in a
basic atmosphere for 510 min (methacrylates) or 1530 min
(epoxy resins). Wash by plunging grid into 0.02 N NaOH fol-
lowed by dH2O and then air-dry.
3.8 Freeze Fracture Freeze fracture/freeze etch of biological tissues permits reliefs of
fractured surfaces to be prepared. These replicas, consisting of a
thin metal (e.g., platinum-carbon) cast, may subsequently be
examined in the TEM. Replicas are especially suited to studies of
membrane structure, notably the internal aspects of membrane
lipid bilayer, and provide a three-dimensional representation of
general tissue/cellular organization at the level of macromolecular
organization of the membranes. In comparison to thin-section
TEM studies where significant contrast exists between the speci-
men and the embedding media (particularly for traditionally pre-
pared room temperature sections stained with heavy metals),
freeze-fracture replicas have no such contrast between specimen
and surrounding fracture plane. Instead, the contrast comes from
variations in the thickness of the metal replica. By applying the
metal coating from a single side at an oblique angle (e.g., 45), a
3D impression of a surface is created, in much the same way as a
mountain is coated in snow, with slopes facing the wind receiving
more snow than sheltered valleys and protected slopes facing away
from the wind! Thus, for coating applied from a single direction
and angle, slopes facing the coating source receive a large deposit
of metal, and slopes shielded from the coating source will receive
no coating. In the TEM, the coated areas will be electron dense
and the uncoated areas will allow the electrons to pass through the
section, creating an area with no contrast. The result is a 2D image
giving the impression of a 3D form (Fig. 7).
1. For freeze fracture, samples must be first cryo-fixed.
Concentrate cells as described above. Place a small volume
(23 l) of pelleted material or tissue into the planchet, alter-
natively scratch the flat surfaces of a planchet with a razor
blade, place a small volume (<1 l) on the flat face of the plan-
chet and sandwich with another scratched planchet (Fig. 5).
Samples can also be prepared by plunge freezing small volumes
of material placed on gold carriers into a liquid cryogen (e.g.,
ethane, propane, Freon 22). For specialist investigations, e.g.,
of membrane injury due to low-temperature events, it is pos-
sible to build the experimental conditions into the preparation
steps for freeze fracture. For plant cold acclimation studies, leaf
sections (<1 mm3) are mounted in gold specimen supports
(Leica microsystems, Vienna) and placed in a cold chamber
with cold stage supplied with a continuously recirculating sup-
ply of refrigerant to allow controlled cooling of specimens to
predetermined nadir temperature at rate of 0.1 C/min.
Tissues are initially cooled to 2 C, then seeded with ice to
simulate extracellular freezing of the leaf (e.g., due to frost
forming) which would in turn promote freezing of extracellu-
lar spaces and cryodehydration of the leaf.
2. For freeze-fracture/freeze-etch TEM, a fracture instrument is
required. For the purposes of this protocol, the Baltec/Leica
262 Roland A. Fleck
Fig. 7 Freeze-fracture SEM and freeze-fracture TEM. (a) Freeze-fracture SEM image of Euplotes sp. showing
Golgi apparatus and nucleus with nucleopores; scale bar represents 1 m. (b) Freeze-fracture TEM image of
Arabidopsis leaf, showing nucleus with nucleopores; scale bar represents 200 nm. Arrow indicates direction
of shadowing
BAF060 is described. However, the protocol can be readily

adapted to other systems (e.g., Baltzers/Baltec BAF 300/400,
JEOL JFD III/V and Cressington CFE-60). To perform
ultrahigh-resolution freeze-fracture/freeze-etch SEM requires
a high-vacuum fracture system with electron beam guns
(ebeam), capable of double-axis rotary shadowing (DARS)
and equipped with a cryo-vacuum transfer system. Currently
the only systems capable would be Leica BAF060 or EM
ACE600 with Leica EM VCT100 (there is one JEOL JFD III,
in Japan, which has been modified and equipped with Leica
EM VCT100). Fracture/cryo-SEM can be performed with
simpler dedicated cryo-SEM stages connected directly to the
SEM itself (e.g., Gatan Alto 1000 or 2500, Quorum
Technologies PP3010T); however, these solutions perform at
a lower resolution.
3. First dry the BAF by activating the thawing cycle, which drives
dry air through the system (30 min). Secondly, prepare, load, and
condition the ebeam guns; cool the BAF to 160 C (Table 2).
4. Test the ebeam guns: adjust the elevation of the guns to the
desired angles, e.g., Gun1 45 and Gun2 90. Select Gun1 and
adjust voltage and emission for the evaporation source (Table 2).
Activate Gun2 and adjust voltage and emission for the evapora-
tion source (Table 2). Verify that the guns work correctly by
selecting each gun and running it; confirm the evaporation
rates with a quartz crystal thickness monitor (Baltec QSG100)
evaporation rates for Pt/C of at least 0.02 nm/s and for carbon
about 0.2 nm/s.
Table 2
Ebeam gun sources, conditioning parameters, and coating conditions for freeze fracture (adapted and translated from the Baltec BAF060 user
manual)
Material Pt-C C Ta-W Cr W Au

Degassing 1 kV/50 mA 1.5 kV/50 mA 1.5 kV/50 mA 0.6 kV/20 mA 1.5 kV/50 mA N/A
ca. 150 s ca. 150 s (for ca. 150 s ca. 60 s ca. 150 s
new C source
double times)
Melting 1.6 kV/60 mA N/A 1.9 kV/90 mA 1.95 kV/98 mA 0.45 kV/13 mA
ca. 10 s (of ca. 10 s (of ca. 10 s (of
evaporation) evaporation) evaporation)
Evaporation 1.6 kV/60 mA 1.9 kV/90 mA 1.9 kV/90 mA 1.2 kV/20 mA 1.95 kV/98 mA 0.34 kV/18 mA
Low Temperature Electron Microscopy
263
264 Roland A. Fleck
5. Specimen loading and transfer using the VCT100: attach the

VCT100 shuttle to the BAF. Fill the LN dewar on the VCT100
and cool the VCT100. Vent the VCT100 and connect it to the
sample-loading box and load a sample and transfer it to the
BAF (for a detailed workflow, please refer to the manufactur-
ers instructions).
6. Fracturing and coating using the knife: adjust stage tempera-
ture to desired fracture temperature (normally between 100
and 110 C) and wait until the temperature of the specimen
holder has reached the temperature of the specimen stage
(allow at least 20 min for the sample and sample block to equil-
ibrate at the desired temperature). Adjust the level of the knife
to the specimen surface (carrier). Fracture sample with the
knife (manually, semiautomatic, or automatically) by perform-
ing a single positive cut, splitting the sandwiched planchets.
Or for samples prepared as open blocks, skim the sample with
the knife to reveal a smooth vitreous area of interest. After
fracturing, cover specimen with shutter manually or in auto-
matic mode and sublime (etch) ice from the specimen or coat
immediately (Table 3).
7. For cryo-SEM, transfer specimen to the cryo-SEM (Fig. 7).
Table 3
Etching and sublimation rages for freeze-fracture freeze etch. Sublimation rates of ice when the
vacuum is 100 > than the saturation vapor pressure of specimen. Adapted from the Baltec BAF060
user manual
Temperature (C) Saturation pressure (mbar) Sublimation rate (g/cm2 s) Etch rate (1 s1)
80 5.36 104 7.16 106 77.0 nm
85 2.29 104 3.10 106 33.3 nm
5 6
90 9.35 10 1.28 10 13.7 nm
5 7
95 3.61 10 5.02 10 5.39 nm
100 1.32 105 1.87 107 2.00 nm
105 4.57 106 6.55 108 0.70 nm
6 8
110 1.48 10 2.15 10 0.23 nm
7 9
115 4.45 10 6.58 10 70.4 pm
120 1.24 107 1.86 109 19.9 pm
9 10
130 7.38 10 1.15 10 1.22 pm
10 12
140 2.88 10 4.64 10 49.5 fm
150 6.68 1012 1.12 1013 1.19 fm
160 8.02 1014 1.40 1015 0.01 fm
8. For TEM replica preparation, detach replica by dipping the

specimen carrier slowly into distilled water. Transfer to 70 %
sulfuric acid for 2030 min. Transfer replica to distilled water
for 5 min. Transfer replica to 40 % chromic acid for 30 min (see
Note 12). Transfer to distilled water and wash for 5 min each.
Collect replica with a clean copper grid. Depending on replica
dimensions, select 200 or 400 mesh. Collect replica with mat
side of formvar-coated copper TEM grid and draw off water
with a filter paper. Clean replica of fatty specimen additionally
in chloroform-methanol (1:1) (see Note 13).
3.9 CEMOVIS A close to native structure of bulk biological specimens can be

for Imaging imaged by cryo-electron microscopy of vitreous sections
of Ultrastructure (CEMOVIS). However, CEMOVIS has not routinely been
in the Fully Hydrated adopted due to the high technical threshold for success. The two
State critical steps consist of (1) producing a frozen section ribbon of a
few mm in length and (2) transferring the ribbon onto a TEM grid.
1. Prepare samples for HPF as described above. Mix the cell pellet
1:1 with 2040 % HMW dextran; pipette ~1 l of the cell sus-
pension into a well of the planchet and form one of the plan-
chet sandwiches (Fig. 6). Or for small sections of tissue (e.g.,
tissue biopsy [43]), place samples into the planchet and sur-
round with 20 % (w/v) HMW dextran; place a second planchet
on top of the first to form a sandwich (Fig. 5) (see Note 14).
2. HPF of tubes: Gently concentrate cells as described above.
Mix the cell pellet 1:1 with 40 % (w/v) HMW dextran. Place
a droplet of the suspension onto a clean piece of parafilm and
place the one end of the tube into the droplet (if using the
EMPACT system, assemble the tube into the cassette first).
Use a thin length of wire to draw the sample into the tube.
Turn the tube around and repeat at the opposite end, to
ensure even filling of the tube and to minimize air bubbles.
Once frozen cut the tube using the cutting tool provided with
the EMPACT.
3. Cryosectioning: Insert and lock the trimming and sectioning
knives into the knife holder (trimming knife to the left, cut-
ting knife to the right). Insert the pumping tube into the LN
dewar; then once it has cooled down, set the microtome to
cool to 140 C. Once the chamber has reached the correct
temperature, transfer the HPF tube/membrane carrier or
aluminum planchet under LN into the chamber. Place the
sample into the specimen holder and secure.
4. Trim at an initial feed 150 m speed maximum (100 mm/s);
cut to a depth of 30 m. Check the surface to see if it is smooth
and black to confirm a vitreous sample. For tubes, make a
square pyramid at a feed of 100 m to a depth of 30 m
266 Roland A. Fleck
(Fig. 6). For planchets and membrane carriers, open the plan-
chet sandwich in LN and place one half of the sandwich into
the sample holder. First cut the planchets to remove all of the
excess metal and to reveal a smooth black vitreous surface.
With the whole width of the trimming knife, cut a flat surface
(Feed 150 m speed maximum (100 mm/s)). Make a square
pyramid at a feed of 100 m to a depth of 30 m (Fig. 6).
5. CEMOVIS sectioning using the cutting knife section (cutting
speed is 310 mm/s; feed 50 nm): With the antistatic ionizer
on (change the intensity to ~50 % and adjust if needed), start-
ing at the left-hand edge of the knife, cut a small ribbon. Switch
off the antistatic ionizer.
During these steps, the first sections of the ribbon are wrapped
around an eyelash. With a ribbon sufficiently attached to the
eyelash, the operator guides the emerging ribbon of vitreous
sample. The more robust technical solution currently available
is with two micromanipulators [47]. One of the micromanipu-
lators guides an electrically conductive fiber to which the rib-
Spurr/ Epon/
60 LR White
Time
Resin Curing temperature polymerisation by UV light below 0C

10 Unicryl
20 LR Gold
30 K4M
GA Fixation
40
Temperature C
Graded series:
50 33% resin in solvent HM20
66% resin in solvent
60 100% resin K11M
100% resin
OsO4 Fixation e.g., 1% OsO4 in acetone
70
80 HM23
Devitrification
90
e.g., 0.2% UA in acetone
100
196 Open planchets remove Dehydration in solvent Fixation Embedding in

capillaries under LN e.g. acetone. May resin through
include fixation graded series
Fig. 8 Generic overview of a freeze-substitution workflow, showing critical steps in the process and indicative
processing temperatures. Adapted from: Humbel & Schwarz [46]. Horizontal lines indicate threshold tempera-
tures above which temperature-dependent events occur. The key stages in the freeze-substitution workflow
are described at the bottom of the figure with the relative curing temperatures of common resins on the right
bon sticks. The second micromanipulator positions the grid

beneath the newly formed section ribbon, and with the help of
an ionizer, the ribbon is attached to the grid. Although manip-
ulations (Diatome, manip) greatly facilitate sectioning, section-
ing artifacts (compression, crevasses, and knife marks) remain.
6. Once the ribbon is long enough stop the ionizer and hold the
ribbon above the grid, activate the static charge-based cou-
pling unit (Leica microsystems, EM Crion) and detach the
eyelash/manipulator from the ribbon. Transfer the grid to the
precooled grid box and either transfer under LN to the micro-
scope or to a storage dewar (Fig. 2).
7. After sectioning, remove specimen to LN storage. Remove
diamond knives and wash in cold water to remove ice, then
wash with distilled water and then clean the diamond with
cleaning stick wetted with 66 % (v/v) ethanol in dH2O. Dry
knives and place in boxes. Loosen all the screws on the speci-
men holder and warm up chamber. CEMOVIS sections can be
stored in grid storage boxes, with screw-top lids, for a few
months under LN (see Note 15).
3.10 Freeze-Drying Samples prepared by HPF and sectioned following the CEMOVIS
protocol can be freeze-dried. To avoid devitrification and shrink-
age, samples are normally freeze-dried at temperatures below
90 C for up to 24 h (Baltec/Leica microsystems BAF060, Leica
microsystems ACE600, EMITECH K775). This extended and
slow drying gives superior preservation of ultrastructure (see Note
16). Samples can be further analyzed in the TEM by a range of
techniques (e.g., energy-dispersive and wavelength-dispersive
X-ray and electron energy-loss spectroscopies) able to provide ele-
mental microanalytical information [48, 49].
3.11 Imaging For imaging, samples which have been cryo-fixed and processed to
a room temperature stable state (e.g., freeze substituted, Tokuyasu,
freeze-fracture replica) can be imaged directly in the SEM or TEM
under normal operating conditions. For those samples which are
vitreous, at and during viewing (e.g., VTF, CEMOVIS, freeze-
fracture SEM), imaging must be performed under cryogenically
controlled conditions (Fig. 1). This requires a LN-cooled speci-
men rod for the TEM (e.g., 626 or 914 Gatan, USA; 1900F/J
cryogenic holder Hummingbird, USA; 2550 cryotomography
holder Fischione, USA). The rod needs a cryo-shield to minimize
contamination during loading of the TEM. In addition, the TEM
requires an anti-contamination device to minimize contamination
of the specimen and decay of the vacuum in the column. Cryo-
SEM is best on a field emission type of SEM and requires a suitable
cryogenic stage with vacuum transfer capability. Systems may be
dedicated (permanently mounted on the SEM, e.g., 2500 or 1000
268 Roland A. Fleck
Alto Gatan, USA; PP3010T Quorum Technologies) or non-dedicated

where all fracturing and coating is performed within a specialist-
dedicated high-vacuum fracture/coating system (BAF060 or EM
ACE600 Leica Microsystems) and subsequently transferred at high
vacuum and cryogenically shielded to the SEM cryostage (VCT100
Leica Microsystems). In all cases, a robust vacuum system and anti-
contaminator is required.
Vitreous samples in both the SEM and TEM are sensitive to
beam-induced warming, which in turn will lead to devitrification
of the sample. Furthermore, beam-induced damage to macromo-
lecular structures is possible. To avoid these deleterious events,
imaging is performed under specific operating conditions. For
TEM, high accelerating voltage microscopes (200400 keV) are
common (balancing electron specimen penetration against losses
in electron detection) and imaging is performed under low total
accumulated electron dose conditions (<100 e/2) [50]. For
SEM, the microscope is nominally operated at low keV (<3 keV)
with total specimen electron dose being controlled by limiting the
total number of beam scans the specimen receives and through
the adoption of beam deceleration techniques (e.g., Gentle Beam
JEOL, Japan).
4 Notes
1. Pure propane melts at 190 C; however, many commercial

sources of propane may be readily supercooled and remain liq-
uid between 190 C and 196 C.
2. Freezing directly in LN does not produce well-vitrified mate-
rial due to the Leidenfrost effect. When a warm surface (e.g.,
an EM grid) is introduced to LN, the LN boils and evapo-
rates (forming bubbles of gas between the LN and the grid),
slowing the rate of freezing of the grid and sample. To avoid
this it is preferable to use an intermediate cryogen with a
large temperature difference between gaseous and solid states
(e.g., ethane, propane, isopentane). These cryogens are
cooled in LN to just above their freezing point to obtain the
coldest liquid possible. Ethane has a large temperature differ-
ence (94.7 C) between gas and solid forms. This permits
rapid freezing of the sample (in the order of 1315 kK/s for
ethane, 1012 kK/s for propane, and <9 kK/s for isopentane
as compared to 0.5 kK/s for LN and 2 kK/s for LN slush),
without an insulation effect from boiling at the liquid-sample
interface. Ethane, however, is highly flammable and can cause
irritation to the skin and respiratory tract as well as burns.
Eye protection is necessary as liquefied ethane can do serious
damage to the eyes. Ethane can also violently splash if it
comes into contact with LN.
3. Total specimen volume required for a study is fairly small

(sample concentrations should be approximately 0.11 mg/ml)
with an initial starting volume of at least 50 l. Proteins, objects
to be imaged, should generally be >100200 kDa (there may
be some exceptions, e.g., smaller sizes can be imaged for sam-
ples of DNA or RNA). In general, sample purity should be
very high (>95 %, perhaps higher depending on the nature of
the contaminant). Certain common buffer constituents have
an adverse effect on image contrast in the TEM. Specifically,
glycerol must be avoided if at all possible. Sucrose (commonly
use in differential centrifugation to isolate/purify small parti-
cles) also reduces contrast but is preferable to glycerol (sucrose
concentration above 3 % (w/v) will be challenging to image).
Buffers/solvents containing large molecules may produce a
strong noise background in the image, making analysis
difficult. The best situation is a fairly simple buffer (however,
fairly high salt concentrations (~0.5 M) can be tolerated).
Detergent solubilization is acceptable, but concentrations
must be kept below or detergent vesicles will obscure the spec-
imen. Do not plunge-freeze samples which have been concen-
trated/purified and contain Optiprep as the final vitreous
sample is electron dense in the TEM and the particles cannot
been visualized (to dilute out the Optiprep, the dilution
required is such that finding particles is unlikely). For consis-
tent and reproducible sample preparation, use of commercially
prepared carbon support files of constant thickness, with regu-
lar patterning of holes, is recommended (e.g., C-flat
(Protochips, USA) or Quantifol (Quantifoil Micro Tools,
GmbH)). For optimum image formation, the particles are
imaged in the vitreous film within the holes of the carbon sup-
port film, not on the carbon film itself.
4. Once frozen the sample is extremely vulnerable to warming
and devitrification. All steps must be carried out at cryogenic
temperatures and it is critical to ensure that all tools used to
manipulate samples are precooled in LN before approaching
the sample (Fig. 4). Thicker solutions (lipids, emulsions, and
gels) need longer blotting times. When working in high-
humidity environments or on a humid day, blotting time may
also need to be increased; the inverse is also true.
5. Avoid sample loss during shipping; it is important to confirm
the performance of the dry-shipping container. Dewars should
be tested (every 6 months and before use) to ensure they main-
tain cryogenic temperatures necessary to protect the sample
(<180 C) for the duration of anticipated transit time.
6. A domestic hair dryer is a useful addition to any cryo-EM labo-
ratory, allowing holders and tools to be safely warmed and
dried between freezing runs.
270 Roland A. Fleck
7. It is possible to mix the cell pellet 1:1 with 2040 % (w/v)

HMW dextran or 1-hexadecene, which act as a cryoprotectant
and filler, respectively; if this is the case, either the medium or
thick sandwiches can be created, as the cryoprotectant will
ensure vitrification of the thicker sample. Planchets are avail-
able with 100, 200 and 300 nm recesses and can be combined
to create samples of different total volume.
8. Although the HPM010 is the most common HPF found in EM
facilities, alternative instruments with significant differences in
their freezing procedures and operation exist. These include the
Wohlwend GmbH Compact 01 and Compact 02, Leica EM
HPM100, Leica EM PACT, and EM PACT2. For protocols
highlighting operation of the Leica EM HPM100 and Leica
EM PACT2 HPF, please refer to Kaech and Ziegler [51] and for
the Wohlwend Compact, Sherman and Huang [52].
9. A 2 % methyl cellulose solution is prepared in advance to the
following recipe: Heat 196 ml dH2O to 90 C and add 4 g
methyl cellulose while stirring. Cool the solution on ice, while
stirring until approximately 10 C. Add water to a volume of
200 ml and stir overnight at 4 C. Leave the solution to mature
at 4 C with no stirring for 3 days, and then spin at 100,000 g
for 1.5 h. Aliquot into 10 ml portions and store in the dark at
4 C (shelf life of up to 3 months).
It is useful to keep a plastic pipette in the cryo-chamber to
blow away the excess material that sticks to the knife. Plastic
pipettes can be stretched by pulling through pliers to reduce
their diameter and provide more controlled and focused jets of
cold gas. Use the antistatic device on full power during trim-
ming and adjust during sectioning if the ribbons fly away.
Loops can be made using fine soldering or platinum wire
connected to a wooden applicator. Alternatively, and for con-
venience, a Perfect Loop can be purchased. Perfect Loop
(Diatome, Switzerland) is designed to allow the consistent
pickup sections consistently, without causing them any dam-
age. The Perfect Loop has the same external diameter as a
TEM grid and internal diameter slightly larger than the
observed area in the TEM which helps attract the sections to
the loop by water surface tension and accurate placement of
the sections on the grid.
An antistatic device (Diatome, Switzerland) is essential to
reduce static changing during cutting and prevent sections
from flying away due to static attraction to surfaces and tools.
It is better to use non-copper grids for immunolabeling.
10. Usually it is best to allow 24 h resin infiltration at resin con-
centrations of 33 % (v/v) and 66 % (v/v) and overnight infiltra-
tion in 100 % resin with a final 24 h in fresh 100 % resin. The
temperature of embedding and polymerization depends on
the resin mixture used (e.g., K4M 30 C, HM20 30 C to

50 C, K11M 60 C, and HM23 as low as 70 C (Fig. 8)).
11. If you are using osmium in your substitution media, care needs
to be taken not to release any toxic osmium vapors, and all
other users of the freezer need to be informed of this risk. After
substitution at the low temperature, move the vials to 35 to
20 C (usually a lab freezer) and infiltrate with resin of choice.
For LR white and epoxy resins, move the vials to room tem-
perature before adding resin then polymerize at 50 C; for
methacrylates polymerize with UV light at 35 C or lower, or
according to manufacturers instructions.
With commercial systems, program the temperature ramps and
perform manual exchange of each substitution solution. With
a home-built system, regulate and monitor temperature
with thermocouples (Fig. 8).
12. High-valent chromium is carcinogenic and teratogenic and
causes severe environmental damage; the use of chromic acid is
only recommended for difficult to clean samples (e.g., plant
tissue).
13. For cryo-SEM, only a single thin metal (14 nm) coating is
required and should be applied using the DARS technique.
For TEM, PT/C is applied at 45 (24 nm) and C (20 nm)
at 90.
There are many protocols described in the literature for clean-
ing replica, including those intended to allow for subsequent
immunogold labeling [53]. It is recommend that careful review
of the replica-cleaning protocol is performed at the outset of a
project to minimize the health and safety risk from the chemi-
cals employed, considering the likely difficulty of successfully
cleaning replica.
14. For some tissue samples, especially brain, it is essential to infil-
trate the sample with cryoprotectant (usually 5 % (w/v) sucrose
and 40 % (w/v) HMW dextran) for 530 min prior to HPF.
15. Shaking or overstretching may break the ribbon particularly
when using an eyelash to collect and pull the ribbon. The rib-
bon can also wrap around itself or fly away due to static, thus
becoming unusable. Cutting under low-humidity conditions
will reduce contamination on the sections.
It is possible to perform a very rapid freeze substitution on
sections prepared by CEMOVIS [54]. In addition, CEMOVIS
can be combined with tomography, or for some cases struc-
tural information can be combined with X-ray data leading to
atomic resolution in situ.
16. For any biological freezing/freeze-drying protocol, intending
subsequent ultrastructural analysis, the recrystallization of ice is
a factor. At temperatures where this phenomena is considered
272 Roland A. Fleck
(130 C), the sublimation rate is in the order of 1 105 nm/s,

requiring extended freeze-drying times and a robust vacuum
system (e.g., turbo pumped with extensive cryo-shielding (e.g.,
Baltec BAF 300, 400, 060, Cressington CFE 60, JEOL JFD))
as the saturated vapor pressure of ice at 130 C is in the order
of 1 106 Pa. (7.6 109 Torr). From the experience of the
author, there are clear indications of ice crystal growth at tem-
peratures warmer than 130 C and optimal freeze-drying is
provided by loading specimens at close to 196 C into a pre-
cooled freeze dryer, equilibrating the sample at 160 C for
20 min, then warming (1 C/min) to 90 C holding isother-
mally for at least 4 h, before warming to room temperature
(0.5 C/min).
Acknowledgments
Kirsty MacLellan-Gibson who has helped organize and run the

NIBSC cryoworkshop, the team at CUI, Leanne Glover, Gema
Vizcay, Monika Balys, Fiona Winning, and Pippa Hawes (The
Pirbright Institute) for her helpful suggestions and comments.
References
1. Glauert AM, Lewis PR (1998) Biological speci- cellular ultrastructure. J Electron Micr Tech 3:
men preparation for transmission electron 177210
microscopy. Portland Press, London, p 326 10. Robards AW, Sleytr UB (1995) Low tempera-
2. Boyde A, Maconnachie E (1979) Volume ture methods in biological electron microscopy.
changes during preparation of mouse embry- In: Glauert AM (ed) Practical methods in elec-
onic tissue for scanning electron microscopy. tron microscopy, vol 10. Elsevier, Amsterdam,
Scanning 2:149163 pp 5146
3. Brenner S, Horne RW (1959) A negative stain- 11. Kellenberger E (1991) The potential of cryo-
ing method for high-resolution electron fixation and freeze substitution: observations
microscopy of viruses. Biochim Biophys Acta and theoretical considerations. J Microsc 161:
34:103110 183203
4. Nermut MV, Frank H (1971) Fine structure of 12. Hayat MA (2000) Chemical fixation. In: Hayat
Influenza A2 (SINGAPORE) as revealed by MA (ed) Principles and techniques of electron
negative staining, freeze-drying and freeze- microscopy: biological applications, 4th edn.
etching. J Gen Virol 10:3751 Cambridge University Press, Cambridge pp 480
5. Nermut MV (1991) Unorthodox methods of 13. Claeys M, Vanhecke D, Couvreur M, Tytgat T,
negative staining. Micron Microsc Acta 22: Coomans A, Borgonie G (2004) High-pressure
327339 freezing and freeze substitution of gravid
6. Baumeister W (1982) Towards higher resolu- Caenorhabditis elegans (Nematoda:
tion in biomolecular electron microscopy. Rhabditida) for transmission electron micros-
Ultramicroscopy 9:151158 copy. Nematology 6:319327
7. Frank J (1989) Image analysis of single macro- 14. Studer D, Graber W, Al-Amoudi A, Eggli P
molecules. Electron Microsc Rev 2:5374 (2001) A new approach for cryofixation by
8. Carrascosa JL, Abella G, Marco S, Carazo JM high-pressure freezing. J Microsc 203:285
(1990) Three-dimensional reconstruction of 15. Studer D, Humbel BM, Chiquet M (2008)
the sevenfold form of Bacillus subtilis Gro El Electron microscopy of high pressure frozen
chaperonin. J Struct Biol 104:28 samples: bridging the gap between cellular
9. Gilkey JC, Staehelin LA (1986) Advances in ultrastructure and atomic resolution.
ultra-rapid freezing for the preservation of Histochem Cell Biol 130:877889
16. Lucic V, Rigort A, Baumeister W (2013) high-pressure freezing: equipment and

Cryo-electron tomography: the challenge of specimen-loading methods. In: Kuo J (ed)
doing structural biology in situ. J Cell Biol Electron microscopy: methods and protocols,
202:407419 2nd edn. Humana Press, Totowa, NJ, USA,
17. Kachar B, Parakkal M, Kurc M, Zhao Y, pp 143173
Gillespie PG (2000) High-resolution structure 31. Steinbrecht RA (1985) Recrystallization and
of hair-cell tip links. Proc Natl Acad Sci U S A ice-crystal growth in a biological specimen, as
97:1333613341 shown by a simple freeze substitution method.
18. Branton D (1966) Fracture faces of frozen J Microsc 140:4146
membranes. Proc Natl Acad Sci U S A 55: 32. Studer D, Hennecke H, Mller M (1992)
10481055 High-pressure freezing of soybean nodules
19. Dodge JD (1968) An atlas of biological ultra- leads to an improved preservation of ultrastruc-
structure. Edward Arnold, London, p 80 ture. Planta 188:155163
20. Fleck RA (2001) A guide to freeze fractured/ 33. Koster AJ, Klumperman J (2003) Electron
freeze etched Arabidopsis thaliana. Quekett microscopy in cell biology: integrating struc-
J Microsc 39:163177 ture and function. Nat Cell Biol:Ss6Ss10
21. Forge A, Davies S, Zajic G (1991) Assessment 34. Dubochet J (2007) The physics of rapid cool-
of ultrastructure in isolated cochlear hair-cells ing and its implications for cryoimmobilization
using a procedure for rapid freezing before of cells. Methods Cell Biol 79:721
freeze-fracture and deep-etching. J Neurocytol 35. Sarraf CE (2000) Immunolabelling for elec-
20:471484 tron microscopy. In: George AJT, Urch CE
22. Hirokawa N, Tilney LG (1982) Interactions (eds) Methods in molecular medicine, vol 40,
between actin-filaments and between actin- Diagnostic and therapeutic antibodies.
filaments and membranes in quick-frozen and Springer, London, UK, pp 439452
deeply etched hair-cells of the chick ear. J Cell 36. Steinbrecht RA, Zierold K (1987)
Biol 95:249261 Cryotechniques in biological electron micros-
23. Stolinski C, Breathnach AS (1975) Freeze frac- copy. Springer, Berlin, p 293
ture replication of biological tissues. Academic, 37. Fernandez-Moran H (1957) Electron micros-
London, p 181 copy of nervous tissue. In: Richter D (ed)
24. Moor H, Riehle U (1968) Snap-freezing under Metabolism of the nervous system. Pergamon
high pressure: a new fixation technique for Press, Oxford, pp 134
freeze-etching. Proceedings of the fourth 38. van Harreveld A, Crowell J (1964) Electron
European Regional Conference of Electron microscopy after rapid freezing on a metal sur-
Microscopy 2:3334 face and substitution fixation. Anat Rec 149:
25. Mller M, Moor H (1984) Cryofixation of 381385
thick specimens by high pressure freezing. In: 39. Barlow DI, Sleigh MA (1979) Freeze substitu-
Revel JP, Barnard T, Haggis GH (eds) Science tion for preservation of ciliated surfaces for
of biological specimen preparation. SEM Inc, scanning electron microscopy. J Microsc 115:
AMF OHare, Chicago, pp 131138 8195
26. Riehle U (1968) ber die Vitrifizierung 40. Nicolas MT, Bassot JM (1993) Freeze substi-
verdnnter wssriger Lsungen. Federal tution after fast freezing fixation in preparation
Institute of Technology (ETH) for immunocytochemistry. Microsc Res Tech
27. Studer D, Michel M, Wohlwend M, Hunziker 24:474487
EB, Buschmann MD (1995) Vitrification of 41. Steinbrecht RA, Mller M (1987) Freeze-
articular cartilage by high-pressure freezing. J substitution and freeze-drying. In: Steinbrecht
Microsc 179:321332 RA, Zierold K (eds) Cryotechniques in bio-
28. Galway ME, Heckman JW, Hyde GJ, Fowke logical electron microscopy. Springer-Verlag,
LC (1995) Chapter 1. Advances in high- Berlin, Heidelberg, pp 149172
pressure and plunge-freeze fixation. In: 42. Hohenberg H, Mannweiler K, Mller M
Galbraith LDW, Bohnert HJ, Bourque DP (1994) High-pressure freezing of cell suspen-
(eds) Methods in cell biology, vol 49. Elsevier, sions in cellulose capillary tubes. J Microsc
Philadelphia, PA, USA, pp 319 175:3443
29. Dahl R, Staehelin LA (1989) High-pressure 43. Hohenberg H, Tobler M, Mller M (1996)
freezing for the preservation of biological High-pressure freezing of tissue obtained by
structure: theory and practice. J Electron Micr fine-needle biopsy. J Microsc 183:133139
Tech 13:165174 44. McDonald KL, Webb RI (2011) Freeze sub-
30. McDonald KLM, Morphew M, Verkade PM, stitution in 3 hours or less. J Microsc 243:
Muller-Reichert T (2007) Recent advances in 227233
274 Roland A. Fleck
45. Polak JM, Varndell LM (1985) Immunolabelling 50. Murphy GE, Jensen GJ (2007) Electron cryoto-
for electron microscopy. Elsevier, Oxford mography. Biotechniques 43(4):413420
46. Humbel BM, Schwarz H (1989) Freeze- 51. Kaech A, Ziegler U (2014) High-pressure
substitution for immunochemistry. In: Verkleij freezing: current state and future prospects.
AJ, Leunissen JLM (eds) Immuno-gold label- In: Kuo J (ed) Electron microscopy: methods
ing in cell biology. CRC Press, Boca Raton, and protocols. Springer, New York, USA,
pp 115134 pp 151171
47. Studer D, Klein A, Iacovache I, Gnaegi H, 52. Sherman DM, Huang CP (2006) High Pressure
Zuber B (2014) A new tool based on two Freezing (HPF) using the compact HPF 01
micromanipulators facilitates the handling of unit and evaluation of freeze-substitution meth-
ultrathin cryosection ribbons. J Struct Biol ods for cyanobacteria and plant cells. Microsc
185:125128 Microanal 12(Suppl 2):442443
48. Warely A (1993) Qualitative X-ray microanaly- 53. Fujimoto K (1995) Freeze-fracture replica
sis of thin sections in neology: appraisal and electron microscopy combined with SDS diges-
interpretation of results. In: Sigee DC, Morgan tion for cytochemical labeling of integral mem-
AJ, Summer AT, Warley A (eds) X-ray micro- brane proteins. J Cell Sci 108:34433449
analysis in biology, experimental techniques 54. Karreman MA, van Donselaar EG, Gerritsen
and applications. Cambridge University Press, HC, Verrips CT, Verkleij AJ (2011) VIS2FIX:
Cambridge, pp 4757 a high-speed fixation method for immuno-
49. Kapp N, Studer D, Gehr P, Geiser M (2007) electron microscopy. Traffic 12:806814
Electron energy-loss spectroscopy as a tool for 55. White DL, Andrews SB, Faller JW, Barrnett RJ
elemental analysis in biological specimens. In: (1976) Chemical nature of osmium tetroxide
Kuo J (ed) Electron microscopy: methods and fixation and staining of membranes by X-ray
protocols. Humana Press, Totowa, NJ, USA, photoelectron-spectroscopy. Biochim Biophys
pp 431447 Acta 436:577592
Part III
Cryopreservation Protocols
Chapter 10
Cryopreservation of Semen from Domestic Livestock

Abstract
In modern livestock breeding, cryopreserved semen is routinely used for artificial insemination. Sperm
cryopreservation secures future reproduction, and insemination doses can be easily shipped. Processing of
semen for cryopreservation can be done with minimal efforts and relatively low costs. In this chapter we
describe the entire cryopreservation process for stallion and bull sperm including dilution of sperm in
primary and freezing extender, cooling and packaging in straws, freezing in liquid nitrogen vapor, cryo-
genic storage, and thawing. Special emphasis is given on preparation of commonly used primary and freez-
ing extenders (skim milk extender INRA-82, TRIS-egg yolk extender TEY) used in a two-step
dilution approach. Furthermore the different cooling rates needed in different temperature regimes during
the freezing process are being described. Cryopreservation procedures are described in case of using both
specialized automated equipment and simple equipment.
Key words Stallion, Bull, Semen, Spermatozoa, Cryopreservation
1 Introduction
Cryopreserved sperm can be used for artificial insemination, irre-

spective of the availability and location of the donor and recipient.
Sperm cryopreservation secures future reproduction for patients
undergoing a treatment which could induce sterility [1].
Cryopreservation of sperm involves collection of semen, dilution
in so-called extenders (containing nutrients and protectants), pack-
aging (typically in straws), cooling and freezing, and subsequent
storage in liquid nitrogen. During frozen storage, metabolic pro-
cesses and damaging reactions are slowed down and longevity is
increased [2].
Despite the fact that sperm is reputed for its large inter- and
intraspecies variation in cryosurvival, cryopreserved sperm is
increasingly used in breeding programs. In modern livestock breed-
ing, use of cryopreserved semen for artificial insemination is favored
above use of other reproductive biotechnological approaches such
as use of cryopreserved embryos. This is related to the relatively low
277
278 Harald Sieme and Harritte Oldenhof
costs and efforts needed for cryopreservation of semen [3]. In the

cattle breeding industry, insemination almost entirely relies on
cryopreserved semen [35]. In the equine breeding industry,
cold-stored (5 C) diluted semen is predominantly used, but cryo-
preserved semen is increasingly being used [6, 7]. For porcine
species, cryopreservation is rarely used and diluted semen is stored
at 1620 C [8, 9].
For domestic species such as stallion, collection of semen for
cryopreservation typically takes place outside the breeding season.
After a period of sexual rest, extragonadal sperm reserves should be
depleted by repeated semen collections prior to collecting sperm
for cryopreservation. Regular semen collection intervals (e.g.,
every 48 h) should be employed in order to ensure sperm quality.
After collection, semen should be directly diluted with at least one
volume of pre-warmed (37 C) primary extender, which contains
nutrients and protectants (salts, sugars, as well as proteins and lip-
ids). After primary dilution, sperm can be subjected to (density)
centrifugation processing to remove part of the seminal plasma
and/or nonviable sperm, and to concentrate the sample. The next
step is to dilute with freezing extender, after which sperm samples
can be packaged in straws, cooled down to 5 C, frozen, and sub-
sequently stored. Freezing extenders contain cryoprotective agents
such as glycerol stabilizing the cells during freezing [5, 10, 11].
Sample size and geometry greatly affect the freezing and thaw-
ing rate. Optimal freezing rates for sperm resulting in maximum
survival after thawing typically range from 10 to 100 C min1.
Diluted semen can be cryopreserved in glass ampoules, aluminum
or plastic tubes, or vials. Semen can also be frozen as pellets.
Nowadays, however, semen is generally cryopreserved in 0.25- or
0.5-mL plastic straws, which have a large surface area to volume
ratio. Straws can be filled (semi-)automatically using a pressure
gradient, and sealed using, e.g., heat or sealing balls/powder.
Conventional equiaxial freezing of semen packaged in straws is
done using different cooling velocities in different temperature
regimes. Typically, sperm diluted in freezing extender is slowly
cooled from room temperature to 5 C at a rate of about 0.1
0.3 C min1, followed by freezing at a rate of 1060 C min1
down to temperatures as low as 80 or 120 C, after which sam-
ples are plunged into liquid nitrogen for storage. Freezing is done
by placing the straws in horizontal position on metal racks. The
racks are placed in a controlled rate freezer, allowing freezing at a
defined cooling rate, while the temperature is evenly distributed in
the chamber. Alternatively, the racks can be placed in the vapor
phase above liquid nitrogen in a closed polystyrene box, for a
defined duration and distance above the liquid nitrogen.
There are many protocols available for semen cryopreservation.
In this chapter we describe the basic components that sperm cryo-
preservation approaches have in common including (1) dilution of
Cryopreservation of Sperm 279
raw semen in a primary extender, (2) centrifugation for concentra-

tion of spermatozoa and removal of seminal plasma, (3) suspension
of sperm in a freezing extender, (4) packaging of sperm samples in
straws, (5) controlled freezing in nitrogen vapor and storage in liq-
uid nitrogen, and (6) thawing.
2 Materials
2.1 Materials 1. Phase contrast microscope (with heated stage), slides, and
for Processing of Raw cover glasses.
Semen, and Dilution 2. Hemocytometer (Neubauer counting chamber) for determin-
ing sperm concentration. Alternatively, a device for automated
counting can be used, e.g., a photometer (see Note 1) or
NucleoCounter (e.g., model SP-100 from ChemoMetec,
Allerd, DK-3450 Denmark).
3. Water bath set at 37 C, for pre-warming of diluents.
4. Extenders for diluting and storage of semen; primary extender
for initial dilution and refrigerated storage and/or freezing
extender for cryopreservation. Homemade extenders can be
used (e.g., INRA-82 skim milk extender for stallion; TRIS-egg
yolk extender for bull) as well as commercial extenders (e.g.,
EZ-Mixin/Freezin, INRA 96/Freeze, BotuSemen/Cryo for
stallion; one/two-step Triladyl or Biladyl for bull; from Animal
Reproduction Systems, Chino, CA, USA; IMV-Technologies,
LAigle, France; Botupharma, Botucatu/SP, Brasil; Minitb,
Tiefenbach, Germany).
5. Centrifuge (for 4001,000 g), and tubes for centrifugation of
sperm samples. Polypropylene conical tubes for 50 mL can be
used (e.g., from Corning Life Sciences, Lowell, MA, USA), or
round bottom centrifuge glasses for volumes up to
100 mL. Alternatively, 40-mL nipple-bottom centrifugation
tubes can be used (e.g., from Pesce Lab Sales, Kennett Square,
PA, USA). In addition, cushion-fluid (e.g., Optiprep from
Axis-Shield PoC, Oslo, Norway) can be used for high-speed
cushioned centrifugation.
6. Device for removing supernatant (vacuum pump, or syringe).
2.2 Materials 1. Cooling cabinet or refrigerator/cold room, for handling of

and Equipment materials (sperm samples, straws) at 5 C.
for Cooling, 2. Cryogenic container and goblets; for storage of straws.
Cryopreservation, Cryoprotective gloves, goggles, and tweezers; for handling of
and Thawing of Sperm liquid nitrogen and straws.
Samples 3. Water bath set at 37 C for thawing of straws.
4. Polypropylene or polyvinyl-chloride 0.25-mL (for bull) or 0.5-mL
(for stallion) straws (e.g., from IMV-Technologies, or Minitb).
5. Device for labeling of straws, or permanent marker.

6. Devices for (semi-)automated filling and sealing of straws.
Alternatively, for filling straws manually, a 1-mL syringe can be
used with a small piece of tubing attached on the opening to fit
in a straw. Sealing can also be done using heat, sealing powder,
or balls.
7. Controlled rate freezer (e.g., model Icecube from Minitb,
or the Minidigitcool from IMV-technologies), with racks
for straws. Alternatively, a Styrofoam box with liquid nitrogen
can be used with racks placed at a fixed position above the liq-
uid nitrogen level or by using a floating freeze rack (e.g.,
from Minitb) (see Note 2).
2.3 Materials 1. Phase contrast microscope equipped with a heated stage and
for Assessment 10 100 magnification with oil immersion, slides, and cover
of Parameters glasses.
Describing Sperm 2. Eosin-nigrosin staining solution (dissolve 5 g nigrosin, 0.6 g
Quality eosin, 3 g sodium citrate dehydrate in 100 mL water; adjust
the pH to 7.0, pass through filter paper; store at 4 C) for
evaluation of sperm viability/membrane integrity and mor-
phology. Alternatively, a buffered formal saline solution (dis-
solve 2.9 g trisodium citrate dihydrate in 100 mL water,
remove 4 mL solution, add 4 mL 37 % formalin solution;
adjust the pH to 7.0, store at 4 C) or Hancock solution [12]
can be used for assessing sperm morphology.
3. Optional: computer-assisted sperm analysis (CASA) system for
assessing sperm motility characteristics. Flow cytometer/fluo-
rescence microscope for assessing membrane integrity using
fluorescent dyes (see Note 3).
3 Methods
3.1 Preparation The procedures described below are examples of frequently used
of Extenders extenders for preserving stallion and bull sperm. We refer to reviews
for Diluting of others [4, 5, 11] and references therein for a more comprehen-
and Storage of Semen sive listing of details (see Note 4).
1. To prepare (clarified) egg yolk use fresh hen eggs (free from
diseases).
(a) Separate egg yolks from whites, and then carefully roll
them on filter paper to remove remnants. Cut the sur-
rounding membrane using a scalpel.
(b) Recover the yolks in a cylinder. This egg yolk solution
(~100 % solution) can be directly used for preparing TRIS-
egg yolk extender.
(c) For preparing clarified egg yolk: add an equal volume of

distilled water (gives a ~50 % egg yolk solution), mix well,
and centrifuge at 10,000 g for 20 min (using an
ultracentrifuge).
(d) Recover the clear water-soluble fraction: discard the lipid
material floating at the surface and the pellet. Filter
through 12 m filters. If needed, freeze as aliquots for
later use.
2. For stallion sperm, e.g., skim milk extender INRA-82 can be
used as a primary extender, while this can be supplemented
with clarified egg yolk and glycerol for use as freezing extender.
(a) Prepare buffered glucose saline solution by dissolving the
following chemicals in 500 mL: 25 g glucose, 1.5 g lactose
monohydrate, 1.5 g raffinose pentahydrate, 0.25 g sodium
citrate dihydrate, 0.41 g potassium citrate monohydrate,
4.76 g HEPES, and antibiotics (see Note 5). Mix well (use
a magnetic stirrer). The pH should be adjusted to 7.2
using KOH. The osmolality should be adjusted to
~330 mOsm kg1 by adding water or glucose if needed (see
Note 6). Filter-sterilize and store at 4 C. This solution
can be frozen as aliquots for later use.
(b) INRA-82 is prepared by mixing equal volumes (typically
50 or 500 mL each) of the above described buffered glu-
cose saline solution and commercial 0.3 % ultra-heat-
treated skim milk (see Note 7). This should give a pH of
6.87.0 and osmolality of ~300330 mOsm kg1. Prepare
fresh/prior to use.
(c) Freezing extender for the so-called two-step dilution is
prepared by supplementing INRA-82 with 4 % egg yolk
and 5 % glycerol (twofold the final concentrations). For
this, prepare modified buffered saline solution by adding
the above mentioned amounts of chemicals in 370 mL
(instead of 500 mL as described above). Then, prepare
1 L INRA-82-freezing extender by also adding: 500 mL
commercial 0.3 % ultra-heat-treated skim milk, 80 mL
clarified egg yolk (~50 % solution, prepared as described
above), and 50 mL glycerol (~100 % solution). Mix well,
and if needed freeze as aliquots for later use.
(d) In case of performing a one-step (direct) dilution with
freezing extender, prepare INRA-82 containing 2 % (v/v)
egg yolk and 2.5 % (v/v) glycerol.
3. For bull sperm, e.g., TRIS-egg yolk (TEY) extenders can be
used. Both TEY-1 and -2 contain egg yolk, whereas TEY-2
added for cryopreservation also includes glycerol. Final con-
centrations for cryopreservation are 20 % (v/v) egg yolk and
6.4 % (v/v) glycerol.
(a) Prepare TRIS stock solution by dissolving the following

chemicals in 1 L distilled water: 36.05 g TRIS, 20.24 g
citric acid monohydrate, 14.88 g d-fructose, and antibiot-
ics (see Note 5). The pH should be adjusted to 7.0. Filter-
sterilize and store at 4 C.
(b) Extenders TEY-1 and TEY-2, for the so-called two-step
dilution, are prepared as follows.
For 1 L TEY-1 add together: 672 mL TRIS stock solution,
128 mL water, and 200 mL egg yolk (~100 % solution).
For 1 L TEY-2 add together: 672 mL TRIS stock solution,
200 mL egg yolk (~100 % solution), and 128 mL glycerol
(~100 % solution).
(c) In case of performing a one-step dilution with freezing
extender, prepare TEY extender. Therefore, add together:
672 mL TRIS stock solution, 64 mL water, 200 mL egg
yolk (~100 % solution), and 64 mL glycerol (~100 % solu-
tion). TEY is also prepared by mixing equal volumes of
TEY-1 and TEY-2.
4. Prior to use, warm primary extender to 37 C in a water bath.
3.2 Semen Collection Before semen collection for cryopreservation (for commercial pur-
and Dilution poses), sires should have passed a complete breeding soundness
with Primary Extender examination and federal health tests. Different methods and
approaches can be employed for collecting semen from domestic
species including stallion and bull (see Note 8).
1. Directly after semen collection: determine the ejaculate (gel-
free) volume, appearance/consistency, and (approximate)
sperm concentration.
2. Dilute raw semen by slowly adding primary extender (e.g.,
INRA-82 for stallion, TEY-1 for bull) pre-warmed at
37 C. Add at least an equal volume of extender, or dilute at
~100 106 sperm mL1. This is referred to as diluted semen.
3. Evaluate sperm motility and morphology using a phase con-
trast microscope with heated stage. For later evaluation of
sperm morphology, sperm can be stained using eosin-nigrosin
solution and dried on a slide, or fixed in buffered formal saline
solution. When available, CASA can be used to assess sperm
motility characteristics. Furthermore, flow cytometric analysis
of sperm stained with specific fluorescent dyes can be used.
4. Let the diluted semen sample cool down to room temperature
(for further handling at room temperature).
3.3 Centrifugation Stallion sperm samples are generally subjected to centrifugation

Processing processing prior to cryopreservation, whereas bull sperm samples
for Removal are not. Centrifugation can be done using a higher speed when a
of Seminal Plasma cushion fluid is used to increase the recovery rate while preventing
and Concentration sperm packing in a dense pellet.
1. When performing ordinary centrifugation:

(a) Transfer diluted sperm sample in a tube (e.g., 50-mL
tube), and centrifuge for 10 min at 600 g at ambient
temperature.
(b) Carefully remove the supernatant; using a vacuum pump
or syringe, leaving a little volume (~5 mL in case of 50 mL
original volume) (see Note 9). Gently resuspend the pel-
let, using a shaker.
2. When performing high-speed cushioned centrifugation:
(a) Add diluted sperm in a centrifugation tube. Underlay a
small amount of cushion fluid in the bottom of the cen-
trifugation tube, such that it covers the full diameter of the
tube (i.e., ~4 mL in case of a 50-mL tube). Centrifuge for
20 min at 1,000 g at ambient temperature.
(b) Carefully remove the supernatant. Cushion fluid can be
aspirated from below the layer with sperm, but this is not
necessary when using nipple tubes (see Note 10). Transfer
the sperm to a new tube when needed.
3. Determine the sperm concentration in the resuspended pellet or
recovered layer, using a hemocytometer or NucleoCounter.
After this either a one- or two-step dilution can be followed
for adding freezing extender (see Note 11).
4. When performing a one-step dilution, dilute (a concentrated
sperm sample) by adding freezing extender containing the
desired final concentrations of protectants. Add slowly, drop-
wise on the side of the tube with the sperm sample while gently
turning the tube to ensure mixing. For stallion sperm aim for
100400 106 sperm mL1 in INRA-82 supplemented with
2 % egg yolk and 2.5 % glycerol. For bull sperm aim for 60 106
sperm mL1 in TEY containing 20 % egg yolk and 6.4 % glyc-
erol (see Note 12).
5. When performing a two-step dilution:
(a) First add primary extender (INRA-82 for stallion, or
TEY-1 for bull) to dilute to twofold the desired final sperm
concentration (e.g., 200800 or 120 106 sperm mL1 for
stallion and bull, respectively).
(b) Then add an equal volume of freezing extender containing
twofold the desired final concentrations of cryoprotective
agents (e.g., INRA-82 supplemented with 4 % egg yolk
and 5 % glycerol for stallion; TEY-2 containing 20 % egg
yolk and 12.8 % glycerol for bull). Add slowly, drop-wise
on the side of the tube while gently turning to ensure
mixing.
3.4 Slow Cooling 1. Slowly cool the sperm sample in freezing extender down to
and Freezing of Sperm 5 C, at ~0.1 C min1. Such a cooling rate is reached by placing
Packaged in Straws a tube with sperm sample of ambient temperature (e.g., 50 mL)
in a beaker with room temperature water (e.g., 250 mL) at
5 C during 24 h (see Note 13). For bull sperm, equilibration
periods at 5 C up to 18 h can be used.
2. Load sperm sample into 0.25-mL (bull) or 0.5-mL (stallion)
labeled straws, while maintaining samples at 5 C (in a cooling
cabinet/cold room) (see Note 14). Use an automatic filling
machine, or a syringe attached to the plugged site of the straw.
Make sure that the sealing powder on the sealed site is moist-
ened, to close the straw on that site.
3. Seal the straws using a sealing machine, sealing powder, or
balls. Place straws on the racks used for freezing.
4. (a) When using a controlled rate freezer: place racks with straws
in the freezing chamber of the device hold at 5 C, and cool
down to 120 C, at 1060 C min1 (see Note 15).
(b) Alternatively, straws can be frozen in the vapor phase
above liquid nitrogen in a Styrofoam box. In this case, a
cooling rate of ~45 C min1 is reached by placing straws
in horizontal position ~5 cm above the surface of liquid
nitrogen for 20 min.
5. Plunge the straws in liquid nitrogen, after which they can be
transferred into goblets, and stored in liquid nitrogen
containers.
3.5 Thawing 1. Thaw straws by incubating in a water bath: thaw 0.5-mL straws
of Cryopreserved for 30 s at 37 C, and 0.25-mL straws for 10 s at 40 C (see
Samples Note 16).
4 Notes
1. A photometer can be used to estimate the cell density for raw

semen, but not when sperm is diluted in extender containing
skim milk and/or egg yolk. Note that photometer measure-
ments (for optical density) do not discriminate between cells
and debris.
2. One can use a Styrofoam box, e.g., with dimensions of
40 20 20 cm. Mark in the box to what level it should be
filled with liquid nitrogen, such that a rack can be positioned/
fixed about 5 cm above this level (during 20 min). Use a rack
in which straws can be positioned in horizontal position in
separate slots.
3. A double staining using the fluorescent dyes propidium iodide
(PI) and SYBR (or Hoechst) is typically done to discriminate
between plasma membrane damaged and intact sperm, respec-

tively. Instead using SYBR, peanut agglutinin labeled with a
fluorophore like fluorescein (FITC-PNA) can be used to assess
acrosomal membrane integrity.
4. Many extender recipes for semen cryopreservation are avail-
able, as well as commercial extenders for which compositions
are not necessarily given. In all cases it is important to check
the pH and osmolality of the extender, and sperm osmotic tol-
erance limits for the species used. Also, federal regulations
should be taken into account concerning testing for diseases of
sires and use of antibiotics [13].
5. Effective and targeted antibiotics as recommended by the
World Organization for Animal Health (OIE) should be added
to frozen semen. Names and concentrations of the antibiotics
added should be stated in the certificate accompanying the
semen during exportation. We typically supplement INRA-82
with 50,000 U L1 penicillin and 50 mg L1 gentamicin. TEY
can be supplemented with a mixture of tylosin (50 mg L1),
gentamicin (250 mg L1), and lincomycin/spectinomycin
(150/300 mg L-1).
6. When the osmolality of a solution is too high: add water. When
the osmolality is too low: calculate the number of osmoles
missing (difference from, e.g., 300330). In case of INRA-82
the major component is glucose, thus add as glucose. Glucose
does not dissociate in solution. Calculate grams to be added
representing onefold the amount of moles to be added (miss-
ing osmoles).
7. Skim milk used in extenders should be heated to ~90 C for
10 min in order to inactivate the lactenin which is toxic to
equine sperm. Skim milk can be replaced with defined proteins
such as phosphocaseinates (e.g., INRA-96 from IMV-
Technologies, or EquiPro from Minitb).
8. See reviews of others [4, 14] and references therein for a more
comprehensive overview on semen collection for domestic
species, including stallion and bull.
9. Presence of large contents of seminal plasma in preserved
semen is described to negatively affect sperm quality. Seminal
plasma contents of ~5 %, however, have been reported to result
in higher percentages of motile sperm post-thaw [15].
10. Iodixanol (commercially available as Optiprep) has been
described to have cryoprotective properties; via increasing the
glass transition temperature of the freezing extender and affect-
ing ice crystal formation [16]. For cryopreservation, 12 % final
iodixanol concentrations can remain in sperm samples.
11. In case of a 1-step dilution, freezing extender containing the
desired final concentrations of cryoprotective agents is directly
added to a concentrated sperm sample. In case of a 2-step dilution,

a sperm sample is first diluted with a primary extender which
does not contain glycerol (and egg yolk); this is done to a con-
centration twofold the desired final sperm concentration.
Then, for cryopreservation, an equal volume of freezing
extender is added which contains twofold the desired final con-
centrations of cryoprotective agents.
12. For commercial purposes, take into account regulations on
minimum values for numbers and/or percentages of motile
and morphological sperm which need to be present in a dose
for use for artificial insemination. Generally the total number
of sperm per straw or insemination dose is calculated. For
bulls, one 0.25-mL straw containing ~20 106 sperm typically
results in ~10 106 progressively motile sperm post-thaw,
when taking into account a 50 % survival after cryopreserva-
tion. For stallions, 48 straws of 0.5 mL containing ~50
100 106 sperm typically result in ~250 106 progressively
motile sperm post-thaw.
13. Filling of straws can be done at ambient temperature, after
which cooling down to 5 C can be done using a controlled
rate freezer, prior to freezing to subzero temperatures.
14. Straws should be labeled with at least information concerning
the date of semen collection, donor, and manufacturer.
15. If desired one can use a protocol with a hold at a particular
subzero temperature at which ice nucleation can be induced
(mechanically).
16. More rapid thawing, as attained when using higher incubation
temperatures for a shorter period, has been described to result
in higher sperm motility post-thaw. This has been attributed to
reduced re-crystallization during thawing. Thawing of straws
for artificial insemination, however, is generally done at
37 C. This can be easily done using available equipment, and
represents the temperature in the female reproductive tract.
Acknowledgments
Employees of the National Stud of Lower Saxony (Celle, Germany)

and Masterrind GmbH (Verden, Germany) are acknowledged for
sharing protocols.
References
1. Johnson MD, Cooper AR, Jungheim ES, 2. Curry MR (2000) Cryopreservation of semen
Lanzendorf SE, Odem RR, Ratts VS (2013) from domestic livestock. Rev Reprod 5:4652
Sperm banking for fertility preservation: a 3. Rodriguez-Martinez RM (2013) Sperm bio-
20-year experience. Eur J Obstet Gynecol technologies in domestic species: state of the
Reprod Biol 170:177182 art. Anim Reprod 10:268276
4. Baracaldo MI, Barth AD, Bertrand W (2006) DD (eds) Equine reproduction, 2nd edn.
Steps for freezing bovine semen: from semen Blackwell Publishing Ltd., Chichester, UK,
collection to the liquid nitrogen tank. In: IVIS pp 29722982
Reviews in veterinary medicine, International 11. Sieme H (2011) Semen extenders for frozen
Veterinary Information ServiceIVIS (eds) semen. In: McKinnon AO, Squires EL, Vaala
Ithaca, NY. www.ivis.org WE, Varner DD (eds) Equine reproduction,
5. Vishvanath R, Shannon P (2000) Storage of 2nd edn. Blackwell Publishing Ltd., Chichester,
bovine semen in liquid and frozen state. Anim UK, pp 29642971
Reprod Sci 62:2353 12. Hancock JL (1957) The morphology of boar
6. Samper JC, Morris CA (1998) Current meth- spermatozoa. J Roy Microsc Soc 76:8497
ods for stallion semen cryopreservation: a sur- 13. Thibier M, Guerin B (2000) Hygienic aspects
vey. Theriogenology 49:895903 of storage and use of semen for artificial insemi-
7. Vidament M (2005) French field results (1985 nation. Anim Reprod Sci 62:233251
2005) on factors affecting fertility of frozen 14. Hurtgen JP (2000) Semen collection in stallions.
stallion semen. Anim Reprod Sci 89:115136 In: Samper JC (ed) Equine breeding manage-
8. Johnson LA, Weitze KF, Fiser P, Maxwell ment and artificial insemination. W.B. Saunders
WMC (2000) Storage of boar semen. Anim Company, Pennsylvania, pp 6169
Reprod Sci 62:143172 15. Moore AI, Squires EL, Graham JK (2005)
9. Didion BA, Braun GD, Duggan MV (2013) Effect of seminal plasma on the cryopreserva-
Field fertility of frozen boar semen: a retro- tion of equine spermatozoa. Theriogenology
spective report comprising over 2600 AI ser- 63:23722381
vices spanning a four year period. Anim Reprod 16. Saragusty J, Gacitua H, Rozenboim I, Arav A
Sci 137:189196 (2009) Protective effects of iodixanol during
10. Sieme H (2011) Freezing semen. In: bovine sperm cryopreservation. Theriogenology
McKinnon AO, Squires EL, Vaala WE, Varner 71:14251432
Chapter 11
Cryopreservation of Mammalian Oocytes

Victoria Keros and Barry J. Fuller
Abstract
Two methods for the laboratory-focused cryopreservation of mammalian oocytes are described, based on
work with murine oocytes. One method uses a relatively low concentration of the cryoprotectant propane-
diol plus sucrose and requires controlled rate cooling equipment to achieve a slow cooling rate. Such a
method has also produced live births from cryopreserved human oocytes. The second method described
employs a high concentration of the cryoprotectant dimethyl sulfoxide plus a low concentration of poly-
ethylene glycol. This is a vitrification method which involves ultrarapid cooling by plunging standard
straws into liquid nitrogen vapor, hence avoiding the need for specialized equipment, but requires techni-
cal ability to manipulate the oocytes quickly in the highly concentrated cryoprotectant solutions. Murine
oocytes vitrified using this technique has resulted in live births.
Key words Oocyte, Slow-cooling, Vitrification, Murine, Human
1 Introduction
Storage of unfertilized oocytes has numerous applications, and

oocyte cryopreservation has become increasingly important over
the past decade. In the treatment of human infertility these appli-
cations include storage of oocytes collected prior to receipt of can-
cer treatments or premature ovarian failure; banking of excess
oocytes produced as a result of in vitro fertilization treatments; and
donation of oocytes to others (storage giving the added advantage
of allowing time to screen donors for potential disease transmission
and for avoiding the necessity of synchronization of the ovarian
stimulation of the donor and recipients cycles). Oocyte cryo-
preservation could also serve as a means of delaying child bearing
to an age when natural fertility and/or oocyte quality has declined.
Applications in animal management include the preservation of
important genetic lineage; the ability to restock following out-
breaks of diseases such as foot and mouth; the preservation of
genetically modified strains, thereby reducing the cost of con-
tinuous breeding and avoiding problems of genetic drift; and the
289
290 Victoria Keros and Barry J. Fuller
preservation of endangered species. In certain circumstances,

oocyte preservation is considered preferable to embryo preserva-
tion in humans because ethical concerns are fewer for unfertilized
gametes. Cryopreservation of the oocytes is the method of choice
to preserve fertility in women who do not have a current partner
to fertilize their oocytes for subsequent cryopreservation of
embryos. In some countries, cryopreservation of human embryos
is banned or is strictly limited. In animal management, oocyte
cryopreservation gives greater flexibility in breeding programs than
embryo cryopreservation.
The first live births from cryopreserved ovulated mammalian
oocytes were reported in 1977 in mice [1]. Much of the funda-
mental cryobiology of mammalian oocytes was investigated in the
murine model. Since then pregnancies and live births resulting
from cryopreserved oocytes have been reported in a number of
species including rabbits [2], cows [3], horses [4], felines [5], and
humans [6]. However, in most species success rates tend to be
poorer than that for cryopreserved embryos.
Historically, oocytes have been considered to be very difficult
cells to cryopreserve for a number of reasons. The oocyte is a large
single cell with relatively low permeabilities to water and cryopro-
tective agents (CPA). This means that oocytes have a tendency to
retain water as ice starts to form in the suspending medium during
cooling and, if this leads to formation of intracellular ice as cooling
proceeds, damage to the cell results. Permeability of oocytes varies
between species, strains and maturational status of the oocyte. The
mature oocyte is also a short-lived cell that must undergo fertiliza-
tion in order for it to continue to survive and develop. For fertiliza-
tion to occur naturally, the oocyte must retain integrity of a number
of its unique structural features. These include the surrounding
zona pellucida, the cortical granules and the microtubular spindle
on which the chromosomes are organized. The zona pellucida is a
glycoprotein coat surrounding the oocyte. Changes to this layer,
triggered by the action of a single sperm binding to its receptor,
induce the cortical granules to release their contents. The enzymes
released from the cortical granules act to crosslink the glycopro-
teins of the zona pellucida, thus rendering it impenetrable to fur-
ther sperm [7]. Cryopreservation has been shown to result in the
premature release of the contents of cortical granules, thus creat-
ing a block to sperm penetration [8]. Cryoprotectants themselves
can cause a transient increase in intracellular calcium similar to the
initial increase caused by sperm entry and subsequent induction of
zona hardening [9]. Conversely cryopreservation can lead to phys-
ical damage to the zona pellucida resulting in multiple sperm entry.
Both of these problems can be overcome by application of the
intracytoplasmic sperm injection (ICSI) technique, whereby a sin-
gle sperm is injected into the oocyte to achieve fertilization. The
microtubular spindle is the structure upon which the condensed
Cryopreservation of Oocytes 291
chromosomes are aligned in mature oocytes and is responsible for

the movement of chromosomes during cell division. Damage to
this structure can lead to aneuploidy after fertilization. During
cooling the microtubular spindle has been shown to disassemble,
although there is growing evidence that the spindle is capable of
repair on warming and post-thaw culture [1012]. Whether such
repair has adverse consequences during development of the result-
ing embryo is still uncertain. One way of avoiding the potential for
damage to the microtubular spindle is to cryopreserve the imma-
ture oocyte before the spindle is formed, when the chromosomes
are contained within the germinal vesicle and the oocyte is termed
immature. However, success following cryopreservation of imma-
ture oocytes is less than that following cryopreservation of mature
oocytes. If protocols for hormonal stimulation of ovaries to induce
release of numerous mature oocytes are not available or not easily
applicable, for example in patients with polycystic ovary syndrome
or in cancer patients, when harvesting of ovarian tissue is the only
option for fertility preservation, then only immature oocytes, con-
tained within the ovaries, may be available for storage. The major
problem with cryopreserving immature oocytes is that they must
be matured in vitro in order to become fertilisable. This matura-
tion involves communication with the cumulus cells that surround
the oocyte [13]. The cumulus cells are much smaller than the
oocyte and are connected with it by numerous gap junctions.
Development of cryopreservation protocols that allow survival of
the oocyte, the cumulus cells and connections between the two is
difficult [14]. However it was shown that preservation of this cell-
to-cell connection can be achieved even within such a complex
ovarian tissue when the method of vitrification was applied [15].
Also, techniques for maturation in vitro require further refinement
in most species. In vitro maturation of oocytes prior to cryopreser-
vation may produce poor quality oocytes that are more prone to
damage during cryopreservation. As well as these general prob-
lems, the oocytes of some species have further characteristics that
make them still more difficult to freeze; for example, porcine
oocytes have a high lipid content and are sensitive to chilling [16].
The protocols used for the cryopreservation of unfertilized
oocytes have largely been adopted from embryo cryopreservation
techniques. As with embryo cryopreservation both slow controlled
rate cooling and vitrification techniques have been successfully
applied. Murine oocytes have been successfully cryopreserved
using a range of CPA such as glycerol or dimethyl sulfoxide, whilst
the technique most commonly applied to the controlled-rate cryo-
preservation of human oocytes is that of slow cooling in the pres-
ence of propanediol and sucrose. This technique, combined with
ICSI, first yielded a live birth from cryopreserved human oocytes in
1997 [17]. Since then, the number of human live births has been
increasing steadily, with live births escalating into hundreds [11, 18].
In the past 10 years, vitrification has also produced accumulating

numbers of human live births [19, 20], with more than 250 recorded
by 2009 [21]. Significant improvement of oocyte vitrification has
been achieved by the Cryotop method [22, 23]. Combination of
the (1) stepwise equilibration at the room temperature in increas-
ing concentrations of permeating CPA with addition of non-
permeating agents to the vitrification solution (minimizing the
toxic effect by using a mixture of cryoprotectants and equilibration
at room temperature); (2) placing an oocyte on the surface of the
fine propylene strip attached to a plastic handle in the minimum
(1 L) volume of vitrification solution (reaching a high cooling
rate to minimize the toxic effect of the CPAs and associated osmotic
injury, also preventing any accidental ice formation, when a lower
concentration of CPA is used, and avoiding any fracture of the
vitrified oocyte or surrounding solution); (3) direct plunging into
liquid nitrogen in an open system thus reaching a high cooling rate
(if required, to avoid contamination of the sample, liquid nitrogen
can be sterilized by filtration or exposure to UV); (4) rapid warm-
ing at 37 C with stepwise equilibration of oocytes in decreasing
concentration of sucrose to prevent the effects of de-vitrification
and osmotic shock resulted in good morphological survival of 95 %
of vitrified oocytes, high rates of fertilization (91 %), in vitro blas-
tocyst formation (50 %) and subsequent delivery of healthy babies.
The Cryotop method using different mixtures of CPAs is the most
commonly used worldwide method for oocyte vitrification.
As both controlled rate slow cooling and vitrification have
been more commonly used around the world, the debate contin-
ues about which (or, if either) technique, is superior for oocyte
cryopreservation. For example, in some studies, vitrification was
found to give better overall outcomes than slow cooling in human
oocytes [2426]. In other investigations, outcomes in human oocyte
cryopreservation were found to be broadly similar using either
technique [27]. A recent systematic review in human oocyte cryo-
preservation comparing either slow cooling methods or vitrifica-
tion techniques found some evidence that oocyte recoveries,
fertilization, and embryo development were higher in the vitrifica-
tion groups [28], although the number of studies which met suit-
able inclusion criteria was small, and most of these had small
subject numbers. The authors acknowledged the difficulties in
conducting such prospective studies and called for more large
scale trials. Due to different requirements in operator skills, pro-
cess times, equipment availability and species-specific factors in
oocyte biology, no single cryopreservation method is likely to be
suitable in all situations, for all species or for all developmental
stages of oocytes. Many ancillary factors such as oocyte quality
and selection, hormonal stimulation regime, and in vitro manipula-
tion conditions can impact on cryopreservation outcomes [29, 30].
In addition, further studies on the fundamental biophysics of
cryogenic exposure are needed to optimize both slow cooling and

vitrification protocols. As far as can be assessed from current infor-
mation, oocyte cryopreservation does not impact in a negative
way on birth abnormalities [21]. Live births can be achieved after
cryogenic storage for at least 5 years [31]. As noted by Boldt [30],
human oocyte cryopreservation has enjoyed a renaissance in
importance over the past decade, fuelled by increasing success in
post-thaw recoveries, and this is likely to feed through into appli-
cations in other species. Oocyte cryopreservation is moving away
from assignment as an experimental procedure, towards recogni-
tion of the true value of this powerful technique with far-reaching
potential in reproductive technologies.
Two methods of oocyte cryopreservation are described below,
with the necessary technical details to introduce these into a stan-
dard laboratory setting and which have been developed over sev-
eral years for murine oocytes. One method is for slow controlled
rate cooling and the other is a vitrification method. They can yield
reliable, good outcomes, but (as discussed above) there may still be
opportunities for further optimization in the various steps, should
investigators be interested to pursue those.
2 Materials
2.1 Slow Controlled 1. A controlled rate freezing machine (available from several
Rate Cooling sources such as Planer plc, Middlesex, UK or Asymptote,
Cambridge, UK) set to hold at 20 C, then cool at 2 C min1
to 7 C, allow for seeding (see Note 1), hold at 7 C for
10 min after seeding, then cool at 0.3 C min1 to 30 C,
then at50 to150 C and finally hold at150 C for 10 min.
2. Plastic straws (IMV, LAigle, France).
3. Plugs for straws (IMV, LAigle, France), sealing powder or
heat sealer (see Note 2).
4. Pulled glass pipettes.
5. Tissue culture dishes (Falcon, Becton and Dickinson Co., USA).
6. Dissecting microscope.
7. Hotplate set at 37 C.
8. Forceps cooled in liquid nitrogen (optionalsee Note 1).
9. Liquid nitrogen dewars, preferably at least two storage dewars
(see Note 3) and one for transporting samples to storage dewar
and cooling forceps, if used.
10. Scalpel.
11. Syringes and needle (optionalsee Notes 4 and 5).
12. Heated water bath set at 30 C.
13. Safety equipment, e.g., cryo-gloves, face shield, oxygen depletion

monitor.
14. Heated gassed incubator.
15. Oocyte culture medium, e.g., for human oocytes Fertilization
Medium (Cook IVF, Brisbane, Australia), for murine oocytes
Tyrodes medium (Invitrogen, Paisley, UK).
16. Dulbeccos phosphate-buffered solution (PBS; Invitrogen,
Paisley, UK) for human oocytes supplemented with 30 % pro-
tein supplement, e.g., plasma protein supplement (Baxter AG,
Vienna, Austria) or serum protein supplement (Pacific
Andrology, CGA/Diasint, Florence, Italy) for mouse oocytes
supplemented with 5 % heat-inactivated fetal bovine serum
(Invitrogen, Paisley, UK).
17. Freezing solutions1.5 M 1,2-propanediol (PrOH) and
1.5 M PrOH plus 0.2 M sucrose, both made up in PBS with
protein supplement.
18. Thawing and dilution solutions1.0 M PrOH plus 0.2 M
sucrose; 0.5 M PrOH plus 0.2 M sucrose; 0.2 M sucrose, all
made up in PBS with protein supplement.
19. Hyaluronidase (optionalsee Note 6).
2.2 Vitrification 1. Plastic straws (IMV, LAigle, France) (see Note 7).
2. Plugs for straws (IMV, LAigle, France) or sealing powder or
heat sealer (see Note 2).
3. Pulled glass pipettes.
4. Tissue culture dishes (Falcon, Becton and Dickinson Co.,
USA).
5. Dissecting microscope.
6. Thermocouple.
7. Device to hold straw horizontally above liquid nitrogen vapor
without covering area of straw containing oocytes (see Note 8).
8. Hotplate set at 37 C.
9. Liquid nitrogen dewars, preferably at least two storage dewars
(see Note 3), plus a container capable of holding liquid nitro-
gen and accommodating a straw held horizontally.
10. Scalpel.
11. 3 1 mL syringe and 2 needle.
12. Heated water bath set at 20 C.
13. Safety equipment, e.g., cryo-gloves, face shield, oxygen deple-
tion monitor.
14. Heated gassed incubator.
15. Oocyte culture medium, e.g., for human oocytes Fertilization

Medium (Cook IVF, Brisbane, Australia), for murine oocytes
Tyrodes medium (Invitrogen, Paisley, UK).
16. Dulbeccos phosphate-buffered solution (PBS; Invitrogen,
Paisley, UK) for human oocytes supplemented with 30 % pro-
tein supplement, e.g., plasma protein supplement (Baxter AG,
Vienna, Austria) or serum protein supplement (Pacific
Andrology, CGA/Diasint, Florence, Italy) for mouse oocytes
supplemented with 5 % heat-inactivated fetal bovine serum
(Invitrogen, Paisley, UK).
17. Vitrification solutions6 M dimethyl sulfoxide (Me2SO) plus
1 mg mL1 polyethylene glycol (PEG; MW 8000) added to 4
strength PBS medium without CaCl2 (see Note 9). The solu-
tion (for a volume of 10 mL) should be made up by adding
1 mL of distilled water to 2.5 mL of 4 PBS prior to adding
the Me2SO (4.69 mL) and PEG (1 mL of 10 mg mL1 aque-
ous stock). Then add protein supplement (e.g., 0.5 mL FBS)
and make up almost to volume with water prior to adding
CaCl2 (100 L of aqueous stock solution of 15.9 g L1). Finally,
add distilled water to make up to the final volume (10 mL in
this case). Keep at room temperature until required. This solu-
tion will be referred to as VSDP (see Note 10). Make up dilu-
tions of 25 % and 65 % VSDP using PBS containing protein
supplement as required.
18. Dilution solution consisting of 1 M sucrose made up in PBS
containing protein supplement.
19. Hyaluronidase (optionalsee Note 6).
3 Methods
3.1 Slow Controlled The slow controlled rate cooling method described uses a mixture
Rate Cooling of the permeating cryoprotectant PrOH and non-permeating
sucrose. Recent studies suggest that 0.3 M sucrose yields better
survival than 0.2 M sucrose in human oocytes [32, 33]. This is
thought to be largely due to greater dehydration of the cells prior
to freezing. Optimal exposure time to PrOH plus 0.2 M sucrose is
510 min, whereas optimal exposure time to PrOH plus 0.3 M
sucrose is yet to be determined. However, an exposure time of
2 min will give a level of dehydration equivalent to that achieved
with 5 min exposure to PrOH plus 0.2 M sucrose [19]. A further
potential modification to the method described below is the choice
of media in which to dilute the cryoprotectants. Recent studies
have shown a medium in which some of the sodium has been
replaced by choline to be preferable to PBS [3436].
1. Straws should be loaded with a column of 1.5 M PrOH plus

0.2 M sucrose and left at room temperature (20 2 C) until
required (see Note 4).
2. Place one 0.5 mL droplet each of PBS plus protein supple-
ment, 1.5 M PrOH and 1.5 M PrOH plus 0.2 M sucrose in a
tissue culture dish. Label each droplet.
3. No more than five oocytes (see Note 11) should be placed in
the droplet of PBS (cumulus cells may be removed, see Note 6)
and then placed in the droplet of 1.5 M PrOH for 10 min at
room temperature. The oocytes are then moved to the droplet
of 1.5 M PrOH plus 0.2 M sucrose for 5 min at room
temperature.
4. The oocytes are then loaded into the prepared straws within
the column of 1.5 M PrOH plus 0.2 M sucrose and the straw
sealed (see Note 2).
5. The straws should be placed within a freezing machine set at
20 C. The machine should be cooled to 7 C at 2 C min1.
Having reached this temperature ice nucleation should then be
initiated in the solution containing the oocytes (see Note 1).
The straws are then replaced in the cooling machine and the
cooling regime resumed with the temperature being held at
7 C for 10 min to allow dissipation of the heat of crystalliza-
tion. Once the cooling protocol has been completed and the
samples are being held at 150 C, wearing appropriate pro-
tective equipment, the straws are placed in a liquid nitrogen
storage dewar (see Note 12).
6. Prior to warming the straws, prepare dishes containing a
0.51 mL droplet each of 1.0 M PrOH plus 0.2 M sucrose,
0.5 M PrOH plus 0.2 M sucrose, 0.2 M sucrose and PBS with
protein supplement. Label each droplet.
7. Wearing suitable protective equipment, warm the straws by
holding them in air for 30 s, hold either end of the straw not
the area containing the oocytes. Then place the straws in the
water bath set at 30 C for 30 s or until the ice has just melted.
8. Wipe the straw dry. The plug, if used, should be removed from
the straw or the ends of the straw cut with a scalpel and the
contents then expelled (see Note 5) into the dish containing
1 M PrOH plus 0.2 M sucrose.
9. The oocytes should remain in this solution for 5 min at room
temperature before being moved into the droplet of 0.5 M
PrOH plus 0.2 M sucrose again for 5 min at room temperature.
10. The oocytes are moved into 0.2 M sucrose for 10 min at room
temperature.
11. The oocytes are then placed in the droplet of PBS for 20 min,
10 min at room temperature and 10 min at 37 C on a
hotplate.
12. The oocytes should then be cultured (23 h for human, 30 min
for mouse) within an incubator in a suitable culture medium to
allow recovery prior to attempted fertilization (see Note 13).
3.2 Vitrification Vitrification is advantageous over slow controlled rate cooling in

that expensive cooling machines are not required and the tech-
nique can be performed in the field. However, the need for high
concentrations of cryoprotectant means that the times stated for
cryoprotectant exposure must be strictly adhered to and manipula-
tion of oocytes in these highly viscous solutions is technically
demanding. Different CPA mixtures and minimum volume vitrifi-
cation have been proven to be effective for oocyte vitrification.
Toxicity of vitrification solution can be reduced by combination of
decreased concentration of permeating cryoprotectant, which can
be balanced by increasing concentrations of non-permeating agents
like polymers and sugars [23, 37]. Successful vitrification can result
in oocytes with very good morphological recovery post-rewarming
(see Fig. 1). The vitrification method described using VSDP has
resulted in high blastocyst formation [38] and live births [39] in
mice, although the technique is prone to variability [37]. The
cryoprotectant mixture and cryopreservation vessel could be mod-
ified but the basic techniques described are applicable to all vitrifi-
cation protocols.
Fig. 1 A human MII oocyte recovered from a vitrification protocol. The oocyte retains good refractive cytoplasm
with intact plasma membrane (solid black arrow), typical first polar body (dashed black arrow), and intact,
homogenous zona pellucida (white arrow). Image kindly provided by V Keros
1. Straws should be prepared by pushing the cotton plug

approximately one third of the way down the straw. Using a
needle and syringe, inject a column of sucrose solution, via
what was the open end, so that the plug is wetted. Take care
not to wet the sides of the straw with sucrose (see Note 14).
The column of sucrose should fill approximately one third of
the straw. Using a separate needle and syringe, inject a small
~0.5 cm column of 100 % VSDP leaving at least 1 cm between
the sucrose and the vitrification solution. Leave prepared straw
at room temperature until required.
2. Place one 50 L droplet each of 25, 65 and 100 % VSDP in a
tissue culture dish. Label each droplet.
3. Place liquid nitrogen, to a level equivalent to at least the length
of the straw, into the vessel capable of accommodating the
straw horizontally. Use the thermocouple to determine the
point above the liquid nitrogen at which the temperature is
140 C and temporarily affix the thermocouple at this level.
4. Pipette a maximum of five oocytes (see Note 11) into 25 %
VSDP droplet and leave at room temperature for 35 min.
5. Transferring as little of the solution as possible (see Note 15),
pipette the oocytes from the 25 % VSDP into 65 % VSDP
droplet.
6. As quickly as possible, move oocytes into droplet of 100 % VSDP.
7. Immediately draw up a small amount of 100 % VSDP into a
glass pipette and collect the oocytes. Transfer the oocytes into
the column of 100 % VSDP contained within the prepared
straw.
8. Seal the open end of the straw (see Note 2). Hold straw using
holder and, wearing suitable protective equipment, position
the straw horizontally at the position above the liquid nitrogen
at which the temperature is 140 C (see Note 16). Keep the
straw in this position for 3 min and then plunge the straw into
liquid nitrogen.
9. Transfer the straw to a liquid nitrogen storage vessel (see Note 12).
10. Prior to warming the straw, place 1 mL of 1 M sucrose solution
in a culture dish, half fill a 1 mL syringe with 1 M sucrose and
place 2 50 L droplets of sucrose solution and 2 50 L
droplets of PBS with protein supplement in a culture dish.
Label the droplets.
11. Wearing suitable protective equipment, remove the straw
from liquid nitrogen storage. Hold the straw in air for 10 s,
hold either end of the straw not the area containing the
oocytes, then plunge the straw into the water bath at 20 C
for 10 s (see Note 16).
12. Wipe the straw dry and cut through the straw using a scalpel in
the area containing the sucrose. Remove the plug from the
other end, if used, or cut with a scalpel. Attach the syringe
containing sucrose to one end of the straw and hold the other
end over the dish containing 1 mL 1 M sucrose solution. Flush
the contents of the straw and syringe into the dish and ensure
good mixing.
13. Immediately begin to look for the oocytes using the dissecting
microscope (see Note 17). As soon as the oocytes are identified
place them in one of the droplets of 1 M sucrose solution.
14. Immediately move the oocytes into the second droplet of 1 M
sucrose solution. When the oocytes have been in contact with
1 M sucrose for a total of 5 min, transfer the oocytes into a
droplet of PBS and leave for 10 min at room temperature.
Move the oocytes into the second droplet of PBS and leave
them for 10 min on the hotplate.
15. The oocytes should then be placed in oocyte culture medium,
within an incubator, (23 h for human, 30 min for mouse)
prior to addition of sperm (see Note 13).
4 Notes
1. Ice nucleation can be initiated automatically by some freezing

machines or can be initiated manually by removing the straws
from the machine and touching the solution with forceps or
cotton tipped applicators/cotton swabs that have been pre-
cooled in liquid nitrogen or by insertion of an ice crystal from
a pipette tip.
2. Straws can be sealed by heating of the ends (and is required for
clinical samples), although care must be taken not to heat the
solution containing the oocytes, or by inserting plugs, which
should be wet in order to give a good seal, or with sealing
powder.
3. Ideally, oocytes should be stored individually within straws and
the straws kept in at least two different locations so that should
an accident occur with one straw then all is not lost. Accurate
records should be kept of their location within the storage
vessel.
4. The straw can be filled using a small syringe and needle.
Pushing the cotton plug along the straw will reduce the capac-
ity of the straw. The plug should be wetted as this will help
prevent liquid nitrogen entry. If the plug is wetted by insertion
of a sizeable column of the primary thawing/dilution solution
(in this case a 4 cm column of 1 M sucrose solution), then
expulsion of this solution along with the oocytes following
thawing will aid dilution of the cryoprotectant. An air gap

(e.g., a 0.5 cm column) should be left between this section
(1 M sucrose solution) and the area to contain the oocytesa
1 cm column of VSDP cryoprotectant. Straws should be clearly
labelled, with an appropriate marker.
5. The straw contents can be expelled upon thawing by inserting
a 1 ml syringe containing at least 0.5 mL of primary thawing/
dilution solution and expelling the contents of both the straw
and syringe into a dish.
6. Studies have been performed with oocytes that have been
denuded of cumulus cells (by means of gentle pipetting and/
or treatment with hyaluronidase) prior to freezing or frozen
with the cumulus intact. No clear advantage either way is evi-
dent. Recently it has been shown that mouse oocytes with
remaining intact cumulus cells successfully survive vitrification
in calcium-free medium with ethylene glycol as CPA [40].
7. Faster cooling/warming rates can be achieved and the risk of
ice crystal formation reduced, by using straws that have been
pulled to reduce the wall thickness, or by use of such devices as
nylon loops or microscope grids. However, such devices are
even more susceptible to temperature change and some of
these systems are open, i.e., the sample is in direct contact
with liquid nitrogen (see Note 12).
8. A piece of plastic that can be inserted into a straw and is
approximately twice the length of a straw is ideal. The plastic is
inserted into the end of the straw and then bent through 90.
The straw can then be placed close to the nitrogen in the hori-
zontal position and the holder can be held by hand at a safe
distance from the nitrogen.
9. Addition of the high concentrations of cryoprotectant used in
vitrification solutions directly to the PBS medium will result in
reduction of the salt concentrations within that medium. Using
a concentrated PBS medium allows the cryoprotectant to be
added prior to the addition of water to give the required vol-
ume of single strength PBS medium.
10. The vitrification solution could be replaced with other combi-
nations of cryoprotectant, for example, ethylene glycol
together with sucrose, or ethylene glycol plus propanediol plus
sucrose often in combination with a macromolecule such as
Ficoll or trehalose, which has been used successfully for the
preservation of bovine and human oocytes, or in combination
with newly applied cryoprotective agent carboxylated -poly-L-
lysine resulted in production of live offspring in mouse [41].
Exposure to this cryoprotectant mixture can be performed at
37 C thereby avoiding any cooling-related damage.
11. In order to adhere to the timing of each step of the procedure

oocytes should be processed in small batches.
12. It should be noted that straws allow rapid heat transfer and
hence are susceptible to temperature change during handling,
for example on transfer from the cooling machine to liquid
nitrogen for storage and during handling for identification pur-
poses prior to thawing. Straws can be stored in the liquid or
vapor phase of liquid nitrogen. If stored within the liquid phase
care should be taken to prevent leakage of liquid nitrogen into
the straw as this will expand on warming and may cause the
straw to explode or crack. Also, the liquid nitrogen may be con-
taminated with viruses, which have been shown to survive such
temperatures. Straws can also be placed within a second vessel,
for example a second larger straw (straw in straw method), to
reduce liquid nitrogen entry. One end of the larger straw should
be pre-sealed and plunged into liquid nitrogen, then vitrifica-
tion straw containing oocytes placed inside of it and another
end of the covering straw accurately sealed avoiding warming of
the sample. However, debate continues about the impact on
cooling and warming rates which enclosure in a second straw
produce, and whether this has an impact on oocyte survival
[30]. If storing in the vapor phase, the level of liquid nitrogen
should be carefully monitored to ensure stability of storage
temperature. Alarms and automatic filling systems are available,
but it is recommended that the level of liquid nitrogen still be
checked manually at regular intervals.
13. There is evidence to suggest that the microtubular spindle is
capable of repair during a period of culture post thaw.
14. The presence of the sucrose solution on the sides of the straw
would dilute the concentration of cryoprotectant in the VSDP
column and, more importantly, may allow the propagation of
ice crystals from the sucrose solution into the area within which
the oocytes are contained.
15. Care must be taken not to dilute the cryoprotectant
concentration in these small droplets, the use of small droplets
aids location of the oocytes in these viscous solutions.
16. The straw is held at 140 C (below the glass transition tem-
perature range for aqueous CPS mixtures), rather than plung-
ing directly into liquid nitrogen in order to reduce the
occurrence of cracks in the vitrified glass which, on warming,
may form sites of ice nucleation. It is important to measure the
temperature as extreme fluctuations in temperature are present
across a few centimeters in liquid nitrogen vapor. Higher cooling/
warming rates are recommended when lower CPA concentra-
tions are used and oocyte is vitrified in the lowless than 1 L
volumeof the vitrification solution. Immediate warming in a

large volume of warming solution at 37 C gives better survival
if oocytes are vitrified in a closed system [42].
17. Oocytes are particularly difficult to locate in this solution and
operator practice is important. The oocytes must spend no
longer than 5 min in 1 M sucrose solution at this point and
this time can pass quickly. Other groups have developed post-
vitrification dilution steps using 0.5 M sucrose solutions [41]
which could be applied here, although we have not direct evi-
dence about their use with VSDP protocols.
Acknowledgement
We would like to gratefully acknowledge the significant role played

by Dr Sharon Paynter, who as a collaborator over many years
developed the practical aspects of the two methods described in
this chapter.
References
1. Whittingham DG (1977) Fertilization in vitro 8. Carroll J, Depypere H, Matthews CD (1990)
and development to term of unfertilized mouse Freeze-thaw induced changes of the zona
oocytes previously stored at 196. J Reprod pellucida explains decreased rates of fertilisation
Fertil 49:8994 in frozen-thawed mouse oocytes. J Reprod
2. Al-Hasani A, Kirsch J, Diedrich K, Blanke S, Fertil 90:547553
van der Ven H, Krebs D (1989) Successful 9. Larman MG, Sheehan CB, Gardner DK (2006)
embryo transfer of cryopreserved and in vitro Calcium-free vitrification reduces
fertilised rabbit oocytes. Hum Reprod 4:7779 cryoprotectant-induced zona pellucida harden-
3. Fuku E, Kojima T, Shoiya Y, Marcus GJ, ing and increases fertilization rates in mouse
Downey BR (1992) In vitro fertilization and oocytes. Reproduction 131:5361
development of frozen-thawed bovine oocytes. 10. Bianchi V, Coticchio G, Fava L, Flamigni C,
Cryobiology 29:485492 Borini A (2005) Meiotic spindle imaging in
4. MacLellan LJ, Carnevale EM, da Silva MAC, human oocytes frozen with a slow freezing pro-
Scoggin CF, Bruemmer JE, Squires EL cedure involving high sucrose concentration.
(2002) Pregnancies from vitrified equine Hum Reprod 20:10781083
oocytes collected from super-stimulated and 11. Stachecki JJ, Cohen J (2004) An overview of
non-stimulated mares. Theriogenology 58: oocyte cryopreservation. Reprod Biomed
911919 Online 9:152163
5. Pope CE, Gmez MC, Kagawa N, Kuwayama 12. Rienzi L, Martinez F, Ubaldi F, Minasi MG,
M, Leibo SP, Dresser BL (2012) In vivo sur- Iacobelli M, Tesarik J, Greco E (2004) Polscope
vival of domestic cat oocytes after vitrification, analysis of meiotic spindle changes in living meta-
intracytoplasmic sperm injection and embryo phase II human oocytes during the freezing and
transfer. Theriogenology 77:531538 thawing procedures. Hum Reprod 19:655659
6. Chen C (1986) Pregnancy after human oocyte 13. Brower PT, Schultz RM (1982) Intercellular
cryopreservation. Lancet 1:884886 communication between granulosa cells and
7. Bleil JD, Wassarman PM (1980) Structure and mouse oocytes: existence and possible nutri-
function of the zona pellucida: identification tional role during oocyte growth. Dev Biol
and characterization of the proteins of the 90:144153
mouse oocytes zona pellucida. Dev Biol 76: 14. Ruppert-Lingham CJ, Paynter SJ, Godfrey J,
185202 Fuller BJ, Shaw RW (2003) Developmental
potential of murine germinal vesicle stage 26. Edgar DH, Gook DA (2012) A critical appraisal
cumulus-oocyte complexes following exposure of cryopreservation (slow cooling versus vitrifi-
to dimethylsulphoxide or cryopreservation: cation) of human oocytes and embryos. Hum
loss of membrane integrity of cumulus cells Reprod Update 18:536554
after thawing. Hum Reprod 18:392398 27. Noyes N, Knopman J, Labella P, McCaffrey C,
15. Keros V, Xella S, Hultenby K, Pettersson K, Clark-Williams M, Grifo J (2010) Oocyte
Sheikhi M, Volpe A, Hreinsson J, Hovatta O cryopreservation outcomes including pre-
(2009) Vitrification versus controlled-rate cryopreservation and post-thaw meiotic spin-
freezing in cryopreservation of human ovarian dle evaluation following slow cooling and
tissue. Hum Reprod 24:16701683 vitrification of human oocytes. Fertil Steril 94:
16. Isachenko V, Soler C, Isachenko E, Perez- 20782082
Sanchez F, Grischenko V (1998) Vitrification 28. Cobo A, Diaz C (2012) Clinical application of
of immature porcine oocytes: effects of lipid oocyte vitrification: a systematic review and
droplets, temperature, cytoskeleton, and addi- meta-analysis of randomized controlled trials.
tion and removal of cryoprotectant. Fertil Steril 96:277285
Cryobiology 36:250253 29. Paynter SJ (2000) Current status of the cryo-
17. Porcu E, Fabbri R, Seracchioli R, Ciotti PM, preservation of human unfertilised oocytes.
Magrini O, Flamigni C (1997) Birth of a Hum Reprod Update 6:449456
healthy female after intracytoplasmic sperm 30. Boldt J (2011) Current results with slow freez-
injection of cryopreserved human oocytes. ing and vitrification of the human oocyte.
Fertil Steril 68:724726 Reprod Biomed Online 23:314322
18. Bianchi V, Lappi M, Bonu MA, Borini A 31. Kim TJ, Hong SW (2011) Successful live birth
(2012) Oocyte slow freezing using a 0.20.3 M from vitrified oocytes after 5 years of cryo-
sucrose concentration protocol: is it really the preservation. J Assist Reprod Genet 28:7376
time to trash the cryopreservation machine? 32. Fabbri R, Porcu E, Marsella T, Rocchetta G,
Fertil Steril 97:11011107 Venturoli S, Flamigni C (2001) Human oocyte
19. Kuleshova L, Gianaroli L, Magli C, Ferraretti cryopreservation: new perspectives regarding
A, Trounson A (1999) Birth following vitrifica- oocyte survival. Hum Reprod 16:411416
tion of a small number of human oocytes. Hum 33. Paynter SJ, Borini A, Bianchi V, De Santis L,
Reprod 14:30773079 Flamigni C, Coticchio G (2005) Volume
20. Yoon TK, Chung HM, Lim JM, Han SY, Ko changes of mature human oocytes on expo-
JJ, Cha KY (2000) Pregnancy and delivery of sure to cryoprotectant solutions used in slow
healthy infants developed from vitrified oocytes cooling procedures. Hum Reprod 20:
in a stimulated in vitro fertilization-embryo 11941199
transfer program. Fertil Steril 74:180181 34. Stachecki JJ, Cohen J, Willadsen SM (1998)
21. Noyes N, Porcu E, Borini A (2009) Over 900 Cryopreservation of unfertilized mouse
oocyte cryopreservation babies born with no oocytes: the effect of replacing sodium with
apparent increase in congenital anomalies. choline in the freezing medium. Cryobiology
Reprod Biomed Online 18:769776 37:346354
22. Kuwayama M, Vajta G, Kato O, Leibo SP 35. Quintans CJ, Donaldson MJ, Bertolino MV,
(2005) Highly efficient vitrification method for Pasqualini RS (2002) Birth of two babies using
cryopreservation of human oocytes. Reprod oocytes that were cryopreserved in a choline-
Biomed Online 11:300308 based freezing medium. Hum Reprod 17:
23. Kuwayama M (2007) Highly efficient vitrifica- 31493152
tion for cryopreservation of human oocytes and 36. Boldt J, Cline D, McLaughlin D (2003)
embryos: the Cryotop method. Theriogenology Human oocyte cryopreservation as an adjunct
67:7380 to IVF-embryo transfer cycles. Hum Reprod
24. Cao YX, Xing Q, Li L, Cong L, Zhang ZG, 18:12501255
Wei ZL, Zhou P (2009) Comparison of sur- 37. Fahy GM, Wowk B, Wu J, Paynter SJ (2004)
vival and embryonic development in human Improved vitrification solutions based on the
oocytes cryopreserved by slow-freezing and predictability of vitrification solution toxicity.
vitrification. Fertil Steril 92:13061311 Cryobiology 48:2235
25. Fadini R, Brambillasca F, Renzini MM, Merola 38. ONeil L, Paynter SJ, Fuller BJ, Shaw RW
M, Comi R, De Ponti E, Dal Canto MB (2009) (1997) Vitrification of mature mouse oocytes:
Human oocyte cryopreservation: comparison improved results following addition of polyeth-
between slow and ultrarapid methods. Reprod ylene glycol to a dimethyl sulfoxide solution.
Biomed Online 19:171180 Cryobiology 34:295301
39. ONeil L, Paynter SJ, Fuller BJ, Shaw RW 41. Watanabe H, Kohaya N, Kamoshita M,
(1999) Birth of live young from mouse oocytes Fujiwara K, Matsumura K, Hyon SH, Ito J,
vitrified in 6 M dimethyl sulfoxide supple- Kashiwazaki N (2013) Efficient production of
mented with 1 mg/ml polyethylene glycol. live offspring from mouse oocytes vitrified with
Cryobiology 39:284 a novel cryoprotective agent, carboxylated
40. Kohaya N, Fujiwara K, Ito J, Kashiwazaki N -poly-L-lysine. PLoS One 8:e83613
(2011) High developmental rates of mouse 42. De Munck N, Verheyen G, Van Landuyt L,
oocytes cryopreserved by an optimized vitrifi- Stoop D, Van de Velde H (2013) Survival and
cation protocol: the effects of cryoprotectants, post-warming in vitro competence of human
calcium and cumulus cells. J Reprod Dev 57: oocytes after high security closed system vitrifi-
675680 cation. J Assist Reprod Genet 30:361369
Chapter 12
Vitrification: A Simple and Successful Method

for Cryostorage of Human Blastocysts
Juergen Liebermann
Abstract
Cryopreservation is one of the keystones in clinical infertility treatment. Especially vitrification has become
a well-established and widely used routine procedure that allows important expansion of therapeutic strat-
egies when in vitro fertilization (IVF) is used to treat infertility. Vitrification of human blastocysts allows
us to maximize the potential for conception from any one in vitro fertilization cycle and prevents wastage
of embryos. This goes even further toward to best utilize a patients supernumerary oocytes after retrieval,
maximizing the use of embryos from a single stimulation cycle. The technology may even be used to elimi-
nate fresh embryo transfers for reasons of convenience, uterine receptivity, fertility preservation, preim-
plantation genetic diagnosis, or emergency management. In this chapter, the application of vitrification
technology for cryopreserving human blastocyst will be revealed through step-by-step protocols. The
results that are presented using the described protocols underscore the robustness of the vitrification tech-
nology for embryo cryopreservation.
Key words Human blastocyst, Cryopreservation, Embryo transfer, Cryostorage, Vitrification
1 Introduction
The impact of cryopreservation on the growth and improved effi-

ciency of assisted reproduction in humans has become increasingly
appreciated. With approximately one quarter of a million babies
born following cryopreservation, cryopreservation has been shown
to increase pregnancy rates while allowing for further selection of
embryos. Therefore, it is possible to achieve implantation and preg-
nancy rates with frozen-thawed embryos as high as those achieved
with fresh embryos. Blastocysts have been shown to increase preg-
nancy rates while allowing for further selection of embryos. Lower
numbers of embryos are being transferred resulting in less high-
order multiple pregnancies and increased implantations. In addi-
tion, decreased numbers of embryos are transferred increasing the
potential for more embryos to be placed in frozen storage and
305
306 Juergen Liebermann
expose the patient to fewer fresh cycles. The fundamental objectives

for successful cryostorage of cells in liquid nitrogen at 196 C can
be summarized as follows: (1) arresting the metabolism reversibly,
(2) maintaining structural and genetic integrity, (3) achieving
acceptable survival rates after thawing, (4) maintaining develop-
mental competence post thaw and (5) achieving a technique that is
reliable and repeatable.
Cryopreservation slows or totally prevents unwanted physical
and chemical changes. The major disadvantage of using cryostor-
age is that it can lead to crystallization of water and thereby can
create new and unwanted physical and chemical events that may
injure the cells that are being preserved. Vitrification, however,
avoids ice formation altogether during the cooling process by
establishing a glassy or vitreous state, wherein molecular transla-
tional motions are arrested without structural reorganization of
the liquid in which the reproductive cells are suspended. To achieve
this glass-like solidification of living cells for cryostorage, high
cooling rates in the range of 2,50030,000 C min1 or greater in
combination with high concentrations of cryoprotectants are used.
During vitrification water is transformed directly from the liquid
phase to a glassy vitrified state. The physical definition of vitrifica-
tion is the solidification of a solution at low temperature, not by ice
crystallization but by extreme elevation in viscosity during cooling
[1, 2]. A primary strategy for vitrifying cells and tissue is to increase
the speed of thermal conductivity, while decreasing the concentra-
tion of the vitrificants to reduce their potential toxicity. In general,
the rate of cooling/warming and the concentration of the cryo-
protectant required to achieve vitrification are inversely related. In
addition, recent publications have shown the dominance of warm-
ing rate over cooling rates in the survival of oocytes subjected to a
vitrification procedure [3, 4].
The earliest attempts using vitrification as an ice-free cryo-
preservation method for embryos were first reported in 1985 [5].
In 1993 successful vitrification of mouse embryos was demon-
strated [6]. Furthermore, bovine oocytes and cleavage stages were
vitrified and warmed successfully a few years later [7]. In 1999 and
2000 successful pregnancies and deliveries after vitrification and
warming of human oocytes were reported [8, 9]. Since that time,
and because it seems to be that both entities appear to be espe-
cially chill-sensitive cells in artificial reproduction technologies
(ART), oocytes and blastocysts seem to receive a potentially sig-
nificant boost in survival rates by avoiding ice crystallization using
vitrification [10]. In general, vitrification solutions are aqueous
cryoprotectant solutions that do not freeze when cooled at high
cooling rates to very low temperature. Vitrification is very simple,
requires no expensive programmable freezing equipment, and
relies especially on the placement of the embryo in a very small
Vitrification of Blastocysts 307
volume of vitrification medium that must be cooled at extreme

rates not obtainable in traditional enclosed cryostorage devices
such as straws and vials.
Although initially reported in 1985 as a successful cryopreser-
vation approach for mouse embryos [5], vitrification has taken a
backseat in human-assisted reproduction. One drawback consid-
ered by embryologists who are not familiar with the vitrification
technique, is the use of high concentrations of cryoprotectants,
which does potentially mean that vitrification solutions are more
toxic than their counterpart solutions used for conventional slow
freezing. This is necessitated by the practical limit for the rate of
cooling and the biological limit of tolerance of the cells for the
concentration of toxic cryoprotectants being used to achieve the
cryopreserved state. It is important to note that recently published
papers [1114] have shown that the use of relatively high concen-
tration of cryoprotectants such as 15 % (v/v) ethylene glycol (EG)
used in an equimolar mixture with dimethyl sulfoxide (Me2SO)
had no negative effect on the perinatal outcomes from blastocyst
transfers following vitrification when compared with those from
fresh blastocyst transfers.
Cryoprotective agents (CPAs) are essential for the cryopreser-
vation of cells. Basically two groups of cryoprotectants exist: (1)
permeating (glycerol, ethylene glycol, dimethyl sulfoxide) and (2)
non-permeating (saccharides, proteins, polymers) agents. The
essential component of a vitrification solution is the permeating
agent. These compounds are hydrophilic nonelectrolytes with a
strong dehydrating effect. Furthermore, these CPAs are able to
depress the freezing point of the solution. Because the permeat-
ing CPA is responsible for the toxicity (the key limiting factor in
cryobiology), different cryoprotectants have been tested for their
relative toxicity, and the results indicate that ethylene glycol is the
least toxic followed by glycerol. In general, the permeating CPA
should be chosen firstly by their permeating property and secondly
on the basis of their potential toxicity. Additionally, these highly
permeating cryoprotectants are also more likely to rapidly diffuse
out of the cells so that the cells quickly regain their original volume
upon warming, thus preventing osmotic injury. The second com-
ponent of a vitrification solution is disaccharides, for example,
sucrose, which does not penetrate the cell membrane, but helps to
draw out more water from cells by osmosis and therefore lessen the
exposure time of the cells to the toxic effects of the cryoprotec-
tants. The non-permeating sucrose also acts as an osmotic buffer to
reduce the osmotic shock that might otherwise result from the
dilution of the cryoprotectant after cryostorage. In addition, per-
meating agents are able to bind with intracellular water and there-
fore water is very slowly removed from the cell. Hence, the critical
intracellular salt concentration is reached at a lower temperature.
Removal of the cryoprotective agent during warming can present

a very real problem in terms of trying to reduce toxicity to the
cells. Firstly, because of the toxicity of the vitrification solutions,
quick dilution of them after warming is necessary; and secondly,
during dilution water permeates more rapidly in to the cell than
the cryoprotective additive diffuses out. As a consequence of the
excess water inflow, the cells are threatened by injury from osmotic
swelling. In this situation the non-permeating sucrose acts as an
osmotic buffer to reduce the osmotic shock. During warming
using a high extracellular concentration of sucrose (e.g., 1.0 M)
counterbalances the high concentration of the cryoprotectant
agents in the cell, as it reduces the difference in osmolarity between
the intra- and extracellular compartments. The high sucrose con-
centration cannot totally prevent the cell from swelling, but it can
reduce the speed and magnitude of swelling [1517].
2 Materials
2.1 General Supplies 1. HSV (High Security Vitrification Kit, Cryo Bio System).
2. Heat sealer (Cryo Bio System).
3. Polycarbonate micropipettes, 170 and 140 m end holes
(MidAtlantic Diagnostics).
4. Brady labeler (TS 2000).
5. Brady labels (PTL-19-427).
6. 90 15 mm Petri dish (Nunclon).
7. Center-well organ culture dish (Falcon).
8. Styrofoam container.
9. Visotubes 10 mm (IMV).
10. CryoCanes aluminum (Thermo Scientific).
2.2 Reagents 1. Serum Protein Supplement (SPS) (Sage).

2. Modified Human Tubal Fluid (HTF-HEPES) (Sage).
3. Vit KitFreeze (Irvine Scientific).
4. Vit KitThaw (Irvine Scientific).
2.3 Equipment 1. Dissecting stereomicroscope (Olympus SZX-12, Bausch Lomb

or Leica) with warming stage.
2. Laminar flow hood (Origio).
3. Inverted microscope (Olympus 1-71).
4. Infrared 1.48 m diode laser (Hamilton ThorneZILOS
laser, Hamilton Thorne Research, Beverly, MA).
3 Methods
3.1 Stepwise Vitrification of blastocysts needs to be undertaken utilizing a

Blastocyst Vitrification closed system (HSV: High Security Vitrification Kit; Cryo Bio
Procedure System, LAigle, France; FDA 510(k) clearance for cleavage stage
embryos in blastocysts) (see Notes 1 and 2) after a two-step load-
ing with cryoprotectant agents at 24 C (see Notes 3 and 4). If
assisted collapsing is done prior to vitrification, then the blastocyst
needs to be put on an inverted microscope equipped with a laser
system (ZILOS-tk, Hamilton Thorne) and the junction of two
trophectoderm cells in each blastocyst needs to be located, and one
shot (100 % power, 500 s pulse length) was applied. Then the
blastocysts can be moved back in the incubators for 510 min.
Briefly, blastocysts have to be placed in equilibration solution,
which is the base medium (M199 with 20 % Serum Supplement
Substitution (SSS) containing 7.5 % (v/v) ethylene glycol (EG)
and 7.5 % (v/v) dimethyl sulfoxide (Me2SO)). After 57 min, the
blastocysts need to be washed quickly in vitrification solution,
which is the base medium containing 15 % (v/v) Me2SO, 15 %
(v/v) EG, and 0.5 M sucrose, for 4560 s and transferred onto the
HSV using a micropipette. Immediately after the loading of not
more than two blastocysts in a 1 L drop on the HSV, the straws
can be heat sealed, then plunged in liquid nitrogen (LN2), and
secondarily stored inside 5 mL LN2-prefilled canes (Visotube
Rond, IMV, France). Each single step is described in detail below:
1. Aseptic techniques are required at all stages. For equilibration
and vitrification procedures, ensure the benchwarmer is at
room temperature (~25 C) (see Notes 3 and 4).
2. Take reagents from the refrigerator and allow them to warm to
room temperature.
3. Separate the blastocysts to freeze into a separate well. Bring
this dish to the inverted microscope, and with the embryo
positioned with the laser objective, use a single pulse to hit the
blastocysts between two trophectoderm cells to collapse the
embryo. Place the dish back into the incubator for 510 min.
4. Label a Petri dish with the patients name under the lid as fol-
lows: HTF-HEPES, ES, and VS. Prepare 2 50 L of HTF-
HEPES, 2 50 L of ES, and 4 50 L of VS (see Fig. 1).
5. Brady label should include the patients last name, first name,
accession #, MPI #, date, and number and type of embryos.
6. Before vitrification, use a Stripper tip with 200 m end hole for
loading the blastocysts on the top.
7. Fill Styrofoam container with LN2.
8. Each sample that is vitrified will be done in a separate hood and
verified by a second embryologist before proceeding. Vitrify
good expanded/hatching blastocysts on days 5, 6, and 7.
Fig. 1 Setup for the vitrification procedure on a plain 90 mm dish lid surface
9. Remove embryos from culture dishes using a Stripper tip into

the HTF-HEPES (drop 1), gently aspirating to remove any
traces of culture media.
10. Pipette from mHTF (drop 1) to the other drop of mHTF
(drop 2), and immediately merge it with the first drop of ES
(drop 3). Set the timer for 5 min.
11. When the time is up, transfer the embryos to the remaining
drop of ES (drop 4). Set the timer for 3 min. Place the embryos
on the top of the drop and let them settle to the bottom.
12. Then load the blastocysts in a VS-backloaded Stripper tip and
rinse through the four droplets of VS (drop 58); between
each droplet clean your tip.
13. Placement into the VS and loading of the cryotop should take
less than 1 min, so that the total incubation time in 15 % VS
is 30 s. After 30 s, gently transfer them to the tip of the HSV
by using a Stripper tip to load the blastocysts in as small vol-
ume (less than 0.5 L) as possible onto the edge of the stick
(see Note 5).
14. Visually confirm placement (see Note 6).
15. Before loading, apply the label to the open end of the empty
straw. Load the HSV stick into the empty straw, the side with
the embryos first. Use the blue handle to make sure the stick
is in as far as it goes. Then, using the heat sealer, seal the open
end of the stick and plunge the whole straw into the LN2.
Place the straw in a precooled aluminum cane for further stor-
age (see Notes 7 and 8).
16. Store at the cane in a nitrogen tank.

17. Make sure to record cane location on the freezing worksheet
and cryo inventory log.
18. Complete all paperwork and recheck that all vial locations are
logged into the Embryo Inventory.
3.2 Stepwise Important to be mentioned is that regardless of the day of cryo-

Blastocyst Warming preservation of the embryo (whether day 5, 6, or 7), at thawing,
Procedure blastocysts should be treated as if they had been frozen on the fifth
day of development. To remove the cryoprotectants, blastocysts
need to be warmed and diluted in a three-step process. With the
HSV submerged in LN2, the inner straw should be removed, and
then the carrier with the blastocysts can be removed from the LN2
and placed directly into a pre-warmed (37 C) organ culture dish
containing 1 mL of 1.0 M sucrose (see Note 9). Blastocysts can be
picked up directly from the HSV and placed in a fresh drop of
1.0 M sucrose at 24 C and immediately connected with a drop of
0.5 M sucrose. After 5 min blastocysts can be transferred to 0.5 M
sucrose solution and connected with drops of base medium for
additional 5 min. Even when switching the cells between different
concentrations of warming solutions, fill up the pipette with the
next lower concentration of warming solution before picking up
the cells for moving in the following concentration (see Note 10).
Then the blastocysts can be washed in the base medium for 3 min
and returned to the culture medium (SAGE Blastocyst Medium,
Trumbull, CT, USA) until transfer (Table 1). Each single step is
described in detail below:
1. Take reagents from the refrigerator and allow them to warm to
room temperature. All cryoprotectants are removed at 25 C.
2. Place a 200 L drop of TS on a Petri dish and place on a warm-
ing plate (see Note 9).
Table 1
Summary of vitrification and warming solutions (Irvine Vit Kit Freeze and Warm)
Composition
Ethylene glycol (%) Me2SO (%) Sucrose (M)

Equilibration solution (ES) 7.5 7.5 0
Vitrification solution (VS) 15 15 0.5
Warming solution (TS) 0 0 1.0
Diluent solution (DS) 0 0 0.5
Wash solution (WS) 0 0 0
Fig. 2 Setup for the warming procedure on a plain 90 mm dish lid surface
3. Label a Petri dish (Nunc) with the patients name under the lid
as follows: TS, DS, and WS. Prepare 1 50 L of TS, 4 50 L
of DS, and 6 50 L of WS (see Fig. 2).
4. Before warming, use a Stripper tip with 200 m end hole for
removing the blastocysts from the top.
5. Fill Styrofoam container with LN2.
6. Confirm location and identification with a second embryolo-
gist before warming any HSV kit. Warm one kit at a time.
7. Each sample that is warmed will be done in a separate hood
and verified by a second embryologist before proceeding.
8. With the HSV kit under LN2, open the kit by cutting the outer
straw. Use the blue handle to remove the inner stick.
9. Submerge HSV kit directly in the pre-warmed drop containing
TS, which should be as close as possible to the LN2 Styrofoam
container (see Note 11). As soon as the HSV kit contents liq-
uefy (within 1 min), try to locate the blastocysts before remov-
ing them with a Stripper tip. After locating all the blastocysts,
remove them from the tip and place them in the droplet of TS
(drop A) and connect immediately with the 1st droplet of DS
(drop B). Wait for shrinkage and re-expansion.
10. When they start to wrinkle, connect with the second droplet
(drop C) and finally with the third droplet of DS (drop D).
11. When they stop reacting and start to reshrink, transfer blas-
tocysts to 0.5 M sucrose (drop E) by placing at the top of
this drop so they float to the bottom. When the reaction is

complete, connect with the first of WS (drop F; wait for
about 90 % re-expansion).
12. After 100 % expansion, connect with droplet #2 (drop G),
then with droplet #3 (drop H) of WS. Turn on the bench-
warmer, and finally dilute through a series of three wash drops
of HS (I to K).
13. Place the blastocysts into a culture dish and put it in the incu-
bator for subsequent culture.
14. Record the survival and appearance of all blastocysts. Update log
with warm data, and notify the physician of result (see Note 12).
3.3 Results Between 2004 and 2013 the Fertility Centers of Illinois IVF
on Blastocyst Laboratory River North (Chicago) has vitrified 21,027 blasto-
Vitrification cysts from 5,389 patients (Table 2). After close of 10 years of vitri-
fying blastocysts using an open as well as closed system and more
than 4,100 FET with an average number of 1.8 embryos trans-
ferred, the perinatal outcome is as follows (babies delivered until
February 2013): 1,495 babies (766 girls and 729 boys) were born
(Table 3). No abnormalities were recorded.
The outcome with regard to the day of development and age
of the patient are summarized in Tables 4 and 5. In good prognosis
patient under 35 years old with transferring day 5 blastocysts, an
ongoing pregnancy and implantation of 47.1 and 41.0 % were
noted (Table 4). In contrast, transferring day 6 blastocysts in
patients younger than 35 of age, an ongoing pregnancy and
implantation of 34.0 and 29.2 % were recorded (Table 5).
Between 2007 and September 2012, the Fertility Centers of
Illinois IVF Laboratory River North (Chicago) performed 1,482
frozen transfers (FET) without collapsing prior vitrification with a
mean age of the patients of 35.4 4.9 years (group A) and 712
FET (group B) with a mean age of the patients of 35.5 5.0 years
where artificial collapse was performed prior vitrification steps
(Table 6). On average, 1.8 embryos were transferred in group A
and B. Survival in group A vs. group B was not significantly differ-
ent (98.6 % vs. 99.5 %). However, there was a significant improvement
Table 2
Retrospective data from 5,389 patients (average age 34.2 4.9) with blastocyst cryopreservation by
vitrification between 2004 and 2013
Day of development
Day 5 Day 6 Day 7 Total

Number of blastocysts vitrified 10,614 (50.5 %) 9,970 (47.4 %) 443 (2.1 %) 21,027 (100 %)
Table 3
Perinatal outcome of vitrified blastocysts after more than 4,100 transfers
between 2004 and 2013 (babies delivered until February 2013)
Day of development
Day 5 Day 6
Deliveries (total) 1,209 748 461
Babies born (total) 1,495 939 556
Female 766 488 278
Male 729 451 278
Singletons 931 (77.0) 561 (75.0) 370 (80.0)
twins 270 (22.0) 183 (24.5) 87 (19.0)
triplets 8 (1.0) 4 (0.5) 4 (1.0)
Values are numbers unless otherwise described; percentages are indicated between brackets
Table 4
Retrospective outcome data (20042013) at the Fertility Centers of Illinois,
Chicago, from vitrified day 5 blastocysts in regard to the patients age
Patients age (year)
<35 3537 3840 >40 Donor

Cycles 1,142 482 311 155 227
Transfers 1,141 482 309 155 227
Blastocysts survived 98.0 98.1 98.8 97.9 98.8
Blastocysts transferred (mean) 1.8 1.7 1.8 1.8 1.8
Positive pregnancy/VET (%) 60.9 58.9 56.0 51.6 59.5
Clinical pregnancy/VET (%) 53.8 50.2 45.0 40.6 52.0
Ongoing/delivered pregnancies (%) 47.1 41.1 34.6 32.0 42.7
Implantations 835 304 180 75 152
Implantation rate (%) 41.0 37.2 31.9 29.7 38.2
Values are numbers, unless otherwise described
in group B compared with group A for the following: (a) clinical

pregnancy rate (cPR), 57.5 % vs. 43.2 %; (b) ongoing pregnancy
(oPR), 49.6 % vs. 35.0 %; and (c) implantation rate (IR), 42.1 % vs.
32.0 % (Table 6).
When the vitrified-warmed blastocysts were divided into
day 5 and 6 groups, the following data was gathered (Table 7).
Table 5
Retrospective outcome data (20042013) at the Fertility Centers of Illinois, Chicago, from vitrified
day 6 blastocysts in regard to the patients age
Patients age (year)
<35 3537 3840 >40 Donor

Cycles 799 411 300 174 139
Transfers 792 406 299 172 139
Blastocysts survived (%) 97.3 98.4 98.4 96.6 100.0
Blastocysts transferred (mean) 1.8 1.7 1.8 1.8 1.8
Positive pregnancy/VET (%) 48.4 43.8 44.1 39.0 46.0
Clinical pregnancy/VET (%) 41.0 36.7 38.5 31.4 36.7
Ongoing/delivered pregnancies (%) 34.0 29.3 30.5 22.1 27.3
Implantations 422 187 146 62 59
Implantation rate (%) 29.2 26.2 27.2 20.2 24.1
Values are numbers, unless otherwise described
Table 6
A comparison of retrospective data from the cryopreservation program
(Fertility Centers of Illinois, Chicago) of vitrified blastocysts without AC
(group A) and with AC (group B) using a closed carrier system between
2007 and 2013
Technique
Group A (no AC) Group B (with AC)

Patients age (y) 35.3 5.0 35.5 5.0
Transfers 1,482 712
Blastocysts warmed 2,697 1,282
Blastocysts survived 2,660 (98.6) 1,276 (99.5)
Blastocysts transferred 2,618 1,257
Blastocysts transferred (mean) 1.8 1.8
Implantations 836 (32.0)a 529 (42.1)a
Positive pregnancy/VET 752 (50.7)b 484 (68.0)b
CLINICAL pregnancy/VET 640 (43.2)c 409 (57.5)c
Ongoing/delivered pregnancies 518 (35.0)d 353 (49.6)d
VET = vitrified embryo transfer. Values are numbers unless otherwise described; per-
centages are indicated between brackets. aP < 0.01; b,c,dP < 0.001
Table 7
A comparison of retrospective data from the cryopreservation program (Fertility Centers of Illinois,
Chicago) of vitrified day 5 and day 6 blastocysts without AC (group A) and with AC (group B) using a
closed carrier system between 2007 and 2013
Technique

Day 5 Day 6 Day 5 Day 6
Patients age (y) 35.1 5.1 35.5 4.9 35.1 4.9 36.2 4.9
Transfers 863 619 441 271
Blastocysts warmed 1,553 1,144 782 500
Blastocysts survived 1,538 (99.0) 1,122 (98.1) 778 (99.5) 498 (99.6)
Blastocysts transferred 1,511 1,107 773 484
Blastocysts transferred (mean) 1.8 1.8 1.8 1.8
Implantations 551 (36.5)a 285 (25.7)aa 358 (46.3)a 171 (35.3)aa
Positive pregnancy/VET 485 (56.2)b 267 (43.1)bb 314 (71.2)b 170 (62.7)b
Clinical pregnancy/VET 415 (48.1)c 225 (36.3)cc 271 (61.5)c 138 (50.9)cc
Ongoing/delivered pregnancies 341 (39.5)d 177 (28.9)dd 239 (54.2)d 114 (42.1)dd
VET = vitrified embryo transfer. Values are numbers unless otherwise described; percentages are indicated between
brackets. Day 5, aP < 0.01; b,c,dP < 0.001; day 6, aaP < 0.01; bb,cc,ddP < 0.001
In 863 FETs transferring day 5 blastocysts from group A (mean

age of 35.1 5.1), the IR, cPR, and oPR were 36.5 %, 48.1 %,
and 39.5 % compared to 46.3 %, 61.5 %, and 54.2 % of day 5
blastocysts from group B (mean age of 35.1 4.9). As shown in
Table 7, implantation, cPR, and oPR occurring in the day 5
blastocysts from group B were significantly higher than in the
day 5 blastocyst from group A (2; P < 0.001, respectively).
If we compare day 6 in group A (mean age of 35.5 4.9) with
day 6 outcome in group B (mean age of 36.2 4.9), the following
data in terms of implantation, cPR, and oPR was observed: 25.7 %,
36.3 %, 28.9 % vs. 35.3 %, 50.9 %, and 42.1 %, respectively
(Table 7). As shown in Table 7, implantation, cPR, and oPR occur-
ring in the day 6 blastocysts of group B were significantly higher
than transferring day 6 blastocysts from group A (2; P < 0.001 for
any comparison, respectively).
In Table 8 the results for patients under 35 are summarized.
Comparing day 5 from group A (n = 438) with day 5 from group
B (n = 228), we found the following for IR, cPR, and oPR: 40.6 %
vs. 51.4 %, 51.4 % vs. 66.7 %, and 44.7 % vs. 59.6 %. We observed
the same trend and tendency for day 6 blastocysts (Table 8).
Table 8
A comparison of retrospective data from the cryopreservation program (Fertility Centers of Illinois,
Chicago) of vitrified day 5 and 6 blastocysts without AC (group A) and with AC (group B) using a
closed carrier system in patients younger than 35 years old between 2007 and 2013
Technique
Less than 35 years old Less than 35 years old
Day 5 Day 6 Day 5 Day 6

Patients age (year) 31.2 2.4 31.5 0.6 31.1 2.3 31.2 2.7
Transfers 438 287 228 98
Blastocysts warmed 796 543 398 180
Blastocysts survived 782 (98.2) 530 (97.6) 395 (99.2) 179 (99.4)
Blastocysts transferred 768 521 393 179
Blastocysts transferred (mean) 1.8 1.8 1.7 1.8
Implantations 312 (40.6)a 155 (29.8)aa 202 (51.4)a 69 (38.5)aa
Positive pregnancy/VET 254 (58.0)b 140 (48.8)bb 173 (75.9)b 66 (67.3)bb
Clinical pregnancy/VET 225 (51.4)c 119 (41.5)cc 152 (66.7)c 54 (55.1)cc
Ongoing/delivered pregnancies 196 (44.7)d 100 (34.8)dd 136 (59.6)d 46 (46.9)dd
a.b,c,d
Values are numbers unless otherwise described; percentages are indicated between brackets. Day 5, P < 0.001; day
6, aa,bb,cc,ddP < 0.001; VET = vitrified embryo transfer
4 Notes
1. Special care must be given to the selection of the carriers. It is

necessary to use the types of carrier or vessel material with
rapid heat transfer that also support the process of uniform
heat exchange to achieve higher cooling rates.
2. In addition, although no report of contamination in human
IVF exists, the user should be encouraged to choose a closed
carrier system, which will work for blastocysts without any
problems.
3. To minimize the toxicity of the cryoprotectant a stepwise
exposure of cells to precooled concentrated solutions (around
room temperature of 24 C) is recommended.
4. Utilizing higher concentrations of cryoprotectant allows
shorter exposure times to the cryoprotectantbut be careful
the toxicity of the cryoprotectants might arise with their con-
centration. Because almost all cryoprotectants are toxic to
some extent, it is important to carefully monitor the duration

of exposure to the final cryoprotectant before plunging into
liquid nitrogen.
5. To facilitate vitrification by even higher cooling rates, it is also
necessary to minimize the volume of the vitrification solution
(VS) as much as practical (preferable less than 1 L). From this
point of view, it is very important to use a small pulled pipette.
Furthermore, by collecting the cells on one place and loading
not more as four cells at the same time in the pipette, it is pos-
sible to keep the volume small. However, if the load of media
is too large, it can still be reduced before plunging in LN2 by
drawing down the droplet to flatten the blastocysts slightly
while removing all surplus vitrification solution.
6. To make sure, that the cells are loaded on the carrier, perform
the loading process under a light microscope. Check the num-
ber of loaded cells.
7. After sealing the carrier submerge the carrier loaded with the
cells directly in liquid nitrogen by passing fast through the
vapor phase (nitrogen gas).
8. Store CryoCane in a prechilled PVC CryoSleeve sitting in the
goblet in the Dewar. It is essential to maintain exposure of the
HSV to LN2 at all times to eliminate risk of warming and
devitrification.
9. Before moving the carrier fast from the nitrogen in the warming
solution, pull a glass pipette or have a stripper tip ready. Fill
the pipette with a small amount of the first warming solution
(TS). Using the HSV as carrier, rinse the open edge of straw
after placing in the pre-warmed (37 C) 1.0 M sucrose, or
because the droplet is so small it warms immediately, it is
essential to pull/stir the blastocysts off the carrier surface
directly as soon as possible to avoid any toxic effect of the
VS. A stirring motion is recommended when plunging into
the warm TS to agitate the cell off the carrier surface without
the need to remove it actively.
10. Even when switching the cells between different concentra-
tions of warming solutions, fill up the pipette with the next
lower concentration of warming solution before picking up the
cells for moving in the following concentration.
11. In general, during vitrification and warming the LN2 Styrofoam
box needs to be as close as possible to the working area mini-
mizing lag in cooling and warming rates.
12. Be aware of the expiration dates of the vitrification and warming
media; once opened, shelf life is 6 (six) weeks (according to the
manufacture).
Acknowledgment
The author wants to thank the Fertility Centers of Illinois (FCI)

and the embryologists at the FCI IVF Laboratory River North
(Elissa Pelts, BS; Jill Matthews, BS; Sara Sanchez, BS; Rebecca
Brohammer, BS; Yuri Wagner, BS; and Ewelina Pawlowska, MS)
for their invaluable contributions and support in pushing vitrifica-
tion to become our standard protocol for cryopreservation of
human oocytes and blastocysts within our program since 2004.
References
1. Fahy GM, MacFarlane DR, Angell CA, in a stimulated in vitro fertilization-embryo
Meryman HT (1984) Vitrification as an transfer program. Fertil Steril 74:180181
approach to cryopreservation. Cryobiology 21: 10. Walker DL, Tummon IS, Hammitt DG, Session
407426 DR, Dumesic DA, Thornhill AR (2004)
2. Fahy GM (1986) Vitrification: a new approach Vitrification versus programmable rate freezing
to organ cryopreservation. In: Merryman HT of late stage murine embryos: a randomized
(ed) Transplantation: approaches to graft rejec- comparison prior to application in clinical
tion. Alan R Liss, New York, pp 305335 IVF. Reprod Biomed Online 8:558568
3. Seki S, Mazur P (2009) The dominance of 11. Takahashi K, Mukaida T, Goto T, Oka C
warming rate over cooling rate in the survival (2005) Perinatal outcome of blastocyst transfer
of mouse oocytes subjected to a vitrification with vitrification using cryoloop: a 4-year fol-
procedure. Cryobiology 59:7582 low-up study. Fertil Steril 84:8892
4. Mazur P, Seki S (2011) Survival of mouse 12. Liebermann J, Tucker MJ (2006) Comparison
oocytes after being cooled in a vitrification of vitrification versus conventional cryopreser-
solution to196 C at 95 to 70,000 C/min vation of day 5 and day 6 blastocysts during
and warmed at 610 to 118,000 C/min: A clinical application. Fertil Steril 86:2026
new paradigm for cryopreservation by vitrifica- 13. Liebermann J (2009) Vitrification of human
tion. Cryobiology 62:17 blastocysts: an update. Reprod Biomed Online
5. Rall WF, Fahy GM (1985) Ice-free cryopreser- 19(Suppl 4):105114
vation of mouse embryos at196 C by vitrifi- 14. Liebermann J (2011) More than six years of
cation. Nature 313:573575 blastocyst vitrificationwhat is the verdict? US
6. Ali J, Shelton JN (1993) Vitrification of preim- Obstet Gynecol 5:1417
plantation stages of mouse embryos. J Reprod 15. Liebermann J, Tucker MJ (2002) Effect of
Fertil 98:459465 carrier system on the yield of human oocytes
7. Vajta G, Holm P, Kuwayama M, Booth PJ, and embryos as assessed by survival and
Jacobsen H, Greve T, Callesen H (1998) Open developmental potential after vitrification.
pulled straws (OPS) vitrification: a new way to Reproduction 124:483489
reduce cryoinjuries of bovine ova and embryos. 16. Liebermann J, Nawroth F, Isachenko V,
Mol Reprod Dev 51:5358 Isachenko E, Rahimi G, Tucker MJ (2002)
8. Kuleshova L, Gianaroli L, Magli C, Ferraretti Potential importance of vitrification in repro-
A, Trounson A (1999) Birth following vitrifica- ductive medicine. Biol Reprod 67:16711680
tion of a small number of human oocytes: case 17. Liebermann J, Dietl J, Vanderzwalmen P,
report. Hum Reprod 14:30773079 Tucker MJ (2003) Recent developments in
9. Yoon TK, Chung HM, Lim JM, Han SY, Ko human oocyte, embryo and blastocyst vitrifica-
JJ, Cha KY (2000) Pregnancy and delivery of tion: where are we now? Reprod Biomed
healthy infants developed from vitrified oocytes Online 7:623633
Chapter 13
Efficient Cryopreservation of Human Pluripotent Stem

Cells by Surface-Based Vitrification
Julia C. Neubauer, Axel F. Beier, Niels Geijsen, and Heiko Zimmermann
Abstract
Efficient cryopreservation of human stem cells is crucial for guaranteeing a permanent supply of high-quality
cell material for drug discovery or regenerative medicine. Conventionally used protocols usually employ-
ing slow freezing rates, however, result in low recovery rates for human pluripotent stem cells due to their
complex colony structure. In this chapter, a surface-based vitrification protocol for pluripotent stem cells
is presented based on a procedure for human embryonic stem cells developed by Beier et al. (Cryobiology
63:175185, 2011). This simple and highly efficient cryopreservation method allows cryopreservation of
large numbers of ready-to-use adherent cells that maintain pluripotency.
Key words Cryopreservation, Pluripotent stem cells, Surface-based vitrification, Vitrification
1 Introduction
Human pluripotent stem cells have been of central interest in bio-

medical research since their first derivation from human blastocysts
by Thomson et al. in 1998 [1], regardless of origin, either derived
from blastocysts or produced by epigenetic reprogramming tech-
niques. Pluripotent stem cells possess the potential to differentiate
into cells of all three germ layers [2] and to renew themselves by
mitotic cell division [3]. Due to these unique properties, pluripo-
tent stem cells are of great interest for both basic and applied
research in biomedical sciences. Furthermore, pluripotent stem
cells have the potential to accelerate drug discovery and drug
development [4, 5]. However, prospective applications, e.g., high-
throughput screenings of new target drugs or new high-content-
screening technologies (using, for instance, surface-based hanging
drop cultivation [6]), require a readily available, reliable, and safe
supply of high-quality stem cells or stem cell-derived progeny cells.
Current standard cryopreservation procedures are based on
slowly freezing detached cells in suspension using cryomedia, typi-
cally containing 10 % Me2SO and varying concentrations of FCS
321
322 Julia C. Neubauer et al.
with cooling rates of approximately 1 C/min. Although a proce-

dure for successful cryopreservation of attached multipotent cells
based on this standard method has been introduced recently [7],
this technique is not applicable for human pluripotent cells due to
their tight colony formation and sensibility to cell to cell contact
disruption [8, 9]. A major cause of cell death during the slow-
freezing procedure is the formation of microscopic ice crystals,
which damage the cellular membrane structures. The added
Me2SO does not adequately prevent ice formation, and fragile
cells, such as human pluripotent stem cells, seem particularly sensi-
tive to cellular damage [10, 11].
In contrast to the slow-freezing procedure, vitrification entails
the rapid (flash)-freezing of cells or tissues, typically by immersion
in liquid nitrogen. The high freezing rate, combined with a very
high concentration of cryoprotective agents, instantly transforms
the cytoplasm into a glasslike structure, preventing the formation
of intracellular ice crystallization [12].
While the recovery rate of slow-frozen hESCs is often below
10 %, vitrification of hESC cultures consistently yields high recov-
ery rates [13]. Although viability of hESCs is quite high directly
after slow-rate cryopreservation, only few colonies are able to
adhere again. hESC colonies possess a compact colony structure
with a high number of tight and gap junctions together with differ-
ent cell adhesion proteins, like connexin 43 and claudin 6 [1416].
Disruption of cell contacts during cell detachment and colony dis-
sociation results in spontaneous differentiation and cell death [17].
Although the use of a selective Rho-associated kinase inhibitor
(ROCKi, Y27632 [18]) significantly improves the survival of cells
during passaging and dissociation [19, 20] and investigations on
cryomedia, cooling rate, and seeding temperature led to slight
improvements of cell viability and recovery [21, 22], mechanisms
of freezing-related cell loss are still not completely understood and
recovery rates remain low. Therefore, expansion of small, mainly
disintegrated colonies after thawing to an extent suitable for later
applications, e.g., for drug screenings, still severely limits research.
Furthermore, a complete prevention of ice crystallization using
vitrification could improve the cryopreservation efficiency of hESCs,
as intracellular ice formation leads to disruption of cell contacts and
lethal damage. Vitrification transforms the cell cytosol into a glass-
like structure [23], preventing the formation of intracellular ice.
Vitrification of hESCs in straws already results in very high sur-
vival and recovery rates with low post-thaw differentiation [24].
Unfortunately, the number of cells that can be vitrified is limited to
about eight hESC clumps per straw due to the required high heat
transfer and the complex procedure.
Surface-based vitrification can overcome these limitations [14],
which is reviewed in this chapter. Cells are undergoing less stress
prior to and during the cryopreservation process (e.g., through
Vitrification of Pluripotent Stem Cells 323
chemical dissociation) [25]. In addition, cell contacts and protein

interactions are maintained during the freezing process due to pre-
vention of ice crystal formation. The procedure is based on a modi-
fied Thermanox coverslip, which allows precise handling, exact
incubation times, and freezing at high cooling rates. Based on this
procedure, a sterile device avoiding direct contact of the cells with
liquid nitrogen has already been developed by Beier et al. [26].
2 Materials
2.1 Cultivation 1. H9 human embryonic stem cells (WiCell, Madison, WI, USA).
Components for H9 2. Treated cell culture dishes (60 mm diameter, Thermo Scientific
Human Embryonic Nunc, Schwerte, Germany).
Stem Cells
3. PBS (without Mg and Ca; Gibco/Life Technologies, Darmstadt,
on Inactivated MEF
Germany).
4. H9 culture medium: add 20 % syntactical serum replacer, 2 mM
L-glutamine, 0.1 mM -mercaptoethanol, 1 % nonessential
amino acids, 4 ng/mL human recombinant bFGF, 100 U/mL
penicillin, and 100 g/mL streptomycin (all Invitrogen/Life
Technologies, Darmstadt, Germany) to DMEM/F12 (Gibco/
Life Technologies, Darmstadt, Germany).
5. Mitotically inactivated MEF cells (Merck Millipore, Darmstadt,
Germany).
6. MEF culture medium: add 10 % heat-inactivated FCS (Gibco/
Life Technologies, Darmstadt, Germany) and 1 % nonessential
amino acids to DMEM (Gibco/Life Technologies, Darmstadt,
Germany).
7. Coating solution for hESCs and MEFs: dissolve 0.1 % gelatine
from porcine skin (Sigma-Aldrich, Munich, Germany) in water
and autoclave it.
2.2 Components 1. Vitrification solution 1 (VS1): add 10 % Me2SO (WAK-Chemie

for Surface-Based Medical, Steinbach, Germany) and 10 % ethylene glycol
Vitrification (Sigma-Aldrich, Munich, Germany) to H9 culture medium.
2. Vitrification solution 2 (VS2): add 300 mM sucrose (Sigma-
Aldrich, Munich, Germany), 20 % Me2SO, and 20 % ethylene
glycol to H9 culture medium.
3. Warming solution 1 (WS1): add 200 mM sucrose to H9 cul-
ture medium.
4. Warming solution 2 (WS2): add 100 mM sucrose to H9 cul-
ture medium.
5. Thermanox coverslips with a diameter of 13 mm (Thermo
Scientific Nunc, Schwerte, Germany).
6. 70 % ethanol.
7. A Dewar vessel with liquid nitrogen.
2.3 Components 1. Live/dead staining solution: add 50 L ethidium bromide

for Analysis (stock solution, 1 mg/mL in ddH2O; Molecular Probes/Life
of Survival Rate Technologies, Darmstadt, Germany) and 8 L fluorescein
of Adherent hESC diacetate (stock solution, 5 mg/mL in acetone; Molecular
Colonies Probes/Life Technologies, Darmstadt, Germany) to 5 mL
PBS (without Mg and Ca).
2. Scalpel.
3. 4- or 24-well plate (Thermo Scientific Nunc, Schwerte,
Germany).
4. Stereo fluorescence microscope (SMZ 1500, Nikon, Japan).
5. Software ImageJ (NIH, USA).
3 Methods
3.1 Preparation 1. To improve the handling, add a handlebar to the coverslip by

of Thermanox making two parallel, about 3 mm-long cuts using a sterile scis-
Coverslips sor and afterwards bend it up (Fig. 1).
and Vitrification 2. Add individual marks, e.g. lines, on the untreated side of the
Thermanox coverslips using a sterile scalpel for simplified
identification of identical colonies before and after
vitrification.
3. After modification, sterilize the substrates by washing with
70 % ethanol for 20 min and incubation under UV radiation
for 15 min.
4. Submerge a 60 mm culture dish in a Dewar vessel containing
liquid nitrogen (LN).
Fig. 1 Preparation of the Thermanox coverslips. (a) Coverslip with handlebar for better handling. (b) Spatial
arrangement of the coverslips for cell cultivation
3.2 Cell Seeding 1. For vitrification of hESCs, distribute up to ten Thermanox

coverslips in a 60 mm cell culture dish, add 5 mL 0.1 % gela-
tine solution, and incubate it at 37 C overnight.
2. Aspirate the 0.1 % gelatine solution, plate 15 104 mitotically
inactivated MEFs, and cultivate the cells in MEF culture
medium in an incubator with 95 % humidity, 37 C, and 5 %
CO2 for 24 h (see Note 1).
3. Aspirate the MEF culture medium, wash once with PBS, and
add the H9 colonies detached using your standard protocol in
H9 culture medium.
4. Cultivate the cells for 57 days with medium change every
other day in an incubator.
3.3 Vitrification 1. Take pictures of individual hESC colonies considering the

added marks with 7.5 magnification on a stereo fluorescence
microscope before vitrification.
2. Take the coverslips with cells at the handlebar with a long for-
ceps, incubate it in 5 mL VS1 for 1 min and afterwards in 5 mL
VS2 for 5 s (see Note 2).
3. Plunge the coverslip into the Dewar containing LN and place it
in the 60 mm culture dish submerged in the Dewar (see Note 3).
4. After 10 s transfer the 60 mm culture dish (still containing liq-
uid nitrogen) to the vapor phase of a nitrogen storage tank (see
Note 4).
3.4 Thawing 1. Prepare a Dewar vessel with liquid nitrogen.

2. Pull out the rack containing the 60 mm culture dish with the
Thermanox coverslips from the nitrogen storage tank.
3. Fill the 60 mm culture dish containing the Thermanox cov-
erslips immediately with liquid nitrogen (see Note 4).
4. Transfer the filled 60 mm culture dish to the Dewar vessel con-
taining liquid nitrogen (LN) and submerge it.
5. Transfer the coverslips from the liquid nitrogen directly into
the prepared warming solutions.
6. Take the coverslips at the handlebar with a long forceps, incu-
bate it in 5 mL WS1 for 1 min and afterwards in 5 mL WS2 for
5 min (see Note 5).
7. Transfer the coverslips to H9 culture medium and cultivate it
in an incubator (37 C, 5 % CO2, 95 % humidity).
8. Use the cells after 24 h recovery for the planned application or
cultivate them further on (see Note 6).
1. Transfer the Thermanox coverslips for analysis after thawing

3.5 Analysis of the to a 4- or 24-well plate, one coverslip per well.
Survival Rate of 2. Add 1 mL live/dead staining solution per well.
Adherent hESC Colonies
3. Incubate cells in the dark at room temperature for 5 min.
After Vitrification
Fig. 2 Analysis of the vital area of a hESC colony using ImageJ. (a) Transmitted light picture before cryopreser-
vation. The colony area was circled manually and the number of pixel was measured. (b) Stained hESC colony
after thawing. Cells were stained with FDA/EB. Only the vital areas of the colony (green) were circled and
number of pixel was measured
4. Aspirate the live/dead staining solution and add 1 mL PBS

(without Mg and Ca) per well.
5. Take pictures of the same hESC colonies as before vitrification
considering the added marks using the stereo fluorescence
microscope with 7.5 magnification, one picture using a FITC
filter and one using a Texas Red filter of the same position.
6. Merge both pictures using the merge feature of the ImageJ
software.
7. For comparison of colony size before and after vitrification,
open the picture of a hESC colony before vitrification in the
ImageJ software (see Fig. 2).
8. Circle the hESC colony manually and calculate the number of
pixels using the ImageJ software (Fig. 3).
9. Apply the same procedure to the fluorescence images taken of
the identical hESC colony after vitrification but only circling
the vital and green area of the colony.
10. Calculate the survival rate using the following formula: size of
vital colony area [pixel]/colony size before cryopreservation
[pixel] 100.
4 Notes
1. Overlapping of the Thermanox coverslips during cell seed-

ing, cell cultivation, and storage should be avoided to guaran-
tee homologous cell distribution and to prevent cell damage.
2. When removing the Thermanox coverslips from the vitrifica-
tion solution 2, make sure that only a very thin liquid layer
Fig. 3 Overview of the analysis method using ImageJ. (a) Microscopic image of hESCs prior to cryopreserva-
tion. (b) Same area of the substrate shown in (a) after thawing. Cells are stained with FDA/EB. Red arrows mark
individually labeled colonies. White arrows show specific and individual marks in the substrate. The number of
pixel per hESC colony was measured using the software ImageJ
remains on the cells. Otherwise, vitrification efficiency will be

influenced.
3. Try to keep the forceps as dry as possible before vitrification, as
otherwise the Thermanox coverslips become frozen to them.
4. Temperature of vitrified samples must not exceed 130 C to
prevent recrystallization or devitrification.
5. Before taking out the Thermanox coverslips with cells from
the transport vessel with liquid nitrogen for thawing, precool
the forceps by immersing them into liquid nitrogen.
6. Some of the vitrified Thermanox coverslips show damages
after vitrification due to high thermal stress on the material,
resulting in damaged hESC colony areas with low viability. But
the dead cells will detach within 24 h and the remaining vital
cells will repopulate these areas.
Acknowledgments
The work with human embryonic stem cells was permitted by the
Robert Koch Institute (18 and 44 permission) and carried out
according to German law.
References
1. Thomson JA, Itskovitz-Eldor J, Shapiro SS, embryonic stem cell lines. Stem Cells 19:
Waknitz MA, Swiergiel JJ, Marshall VS, Jones JM 193204
(1998) Embryonic stem cell lines derived from 3. Wright WE, Piatyszek MA, Rainey WE, Byrd
human blastocysts. Science 282:11451147 W, Shay JW (1996) Telomerase activity in
2. Odorico JS, Kaufman DS, Thomson JA (2001) human germline and embryonic tissues and
Multilineage differentiation from human cells. Dev Genet 18:173179
4. Lerou PH, Daley GQ (2005) Therapeutic 15. Richards M, Tan SP, Tan JH, Chan WK,
potential of embryonic stem cells. Blood Rev Bongso A (2004) The transcriptome profile of
19:321331 human embryonic stem cells as defined by
5. Strulovici Y, Leopold PL, OConnor TP, SAGE. Stem Cells 22:5164
Pergolizzi RG, Crystal RG (2007) Human 16. Sathananthan H, Pera M, Trounson A (2002)
embryonic stem cells and gene therapy. Mol The fine structure of human embryonic stem
Ther 15:850866 cells. Reprod Biomed Online 4:5661
6. Schulz JC, Stumpf PS, Katsen-Globa A, 17. Amit M, Carpenter MK, Inokuma MS, Chiu
Sachinidis A, Hescheler J, Zimmermann H CP, Harris CP, Waknitz MA, Itskovitz-Eldor J,
(2012) First steps towards the successful Thomson JA (2000) Clonally derived human
surface-based cultivation of human embryonic embryonic stem cell lines maintain pluripo-
stem cells in hanging drop systems. Eng Life tency and proliferative potential for prolonged
Sci 12:584587 periods of culture. Dev Biol 227:271278
7. Katsen-Globa A, Meiser I, Petrenko YA, Ivanov 18. Narumiya S, Ishizaki T, Uehata M (2000) Use
RV, Lozinsky VI, Zimmermann H, Petrenko and properties of ROCK-specific inhibitor
AY (2014) Towards ready-to-use 3-D scaffolds Y-27632. Methods Enzymol 325:273284
for regenerative medicine: adhesion-based 19. Watanabe K, Ueno M, Kamiya D, Nishiyama
cryopreservation of human mesenchymal stem A, Matsumura M, Wataya T et al (2007) A
cells attached and spread within alginate-gelatin ROCK inhibitor permits survival of dissociated
cryogel scaffolds. J Mater Sci Mater Med 25: human embryonic stem cells. Nat Biotechnol
857871 25:681686
8. Acker JP, Larese A, Yang H, Petrenko A, 20. Li XY, Meng GL, Krawetz R, Liu SY, Rancourt
McGann LE (1999) Intracellular ice forma- DE (2008) The ROCK inhibitor Y-27632
tion is affected by cell interactions. Cryobiology enhances the survival rate of human embryonic
38:363371 stem cells following cryopreservation. Stem
9. Heng BC, Ye CP, Liu H, Toh WS, Rufaihah AJ, Cells Dev 17:10791085
Yang Z et al (2006) Loss of viability during 21. Kim SJ, Park JH, Lee JE, Kim JM, Lee JB, Moon
freeze-thaw of intact and adherent human SY et al (2004) Effects of type IV collagen and
embryonic stem cells with conventional slow- laminin on the cryopreservation of human
cooling protocols is predominantly due to embryonic stem cells. Stem Cells 22:950961
apoptosis rather than cellular necrosis. J Biomed 22. Lee JY, Lee JE, Kim DK, Yoon TK, Chung
Sci 13:433445 HM, Lee DR (2010) High concentration of
10. Liu JC, Guan X, Ryan JA, Rivera AG, Mock C, synthetic serum, stepwise equilibration and
Agrawal V et al (2013) High mitochondrial slow cooling as an efficient technique for large-
priming sensitizes hESCs to DNA-damage- scale cryopreservation of human embryonic
induced apoptosis. Cell Stem Cell 13:483491 stem cells. Fertil Steril 93:976985
11. Dumitru R, Gama V, Fragan BM, Bower JJ, 23. Chen SU, Lien YR, Chao K, Lu HF, Ho HN,
Swahari V, Pevny LH, Deshmukh M (2012) Yang YS (2000) Cryopreservation of mature
Human embryonic stem cells have constitutively human oocytes by vitrification with ethylene
active Bax at the Golgi and are primed to glycol in straws. Fertil Steril 74:804808
undergo rapid apoptosis. Mol Cell 46:573583 24. Reubinoff BE, Pera MF, Vajta G, Trounson
12. Wowk B (2010) Thermodynamic aspects of vit- AO (2001) Effective cryopreservation of
rification. Cryobiology 60:1122 human embryonic stem cells by the open pulled
13. Richards M, Fong CY, Tan S, Chan WK, straw vitrification method. Hum Reprod 16:
Bongso A (2004) An efficient and safe xeno- 21872194
free cryopreservation method for the storage of 25. Hornung J, Muller T, Fuhr G (1996)
human embryonic stem cells. Stem Cells Cryopreservation of anchorage-dependent
22:779789 mammalian cells fixed to structured glass and
14. Beier AF, Schulz JC, Drr D, Katsen-Globa A, silicon substrates. Cryobiology 33:260270
Sachinidis A, Hescheler J, Zimmermann H 26. Beier AF, Schulz JC, Zimmermann H (2013)
(2011) Effective surface-based cryopreserva- Vitrification with a twist: towards a sterile,
tion of human embryonic stem cells by vitrifi- serum-free surface-based vitrification of hESCs.
cation. Cryobiology 63:175185 Cryobiology 66:816
Chapter 14
Cryopreservation of Greenshell Mussel

(Perna canaliculus) Sperm
Serean L. Adams, John F. Smith, Jolene Taylor, Lindsay T. McGowan,
and H. Robin Tervit
Abstract
Cryopreservation is a valuable technique for aquaculture as it enables a library or bank of genetically valuable
animals to be maintained in a cost-effective manner. Here, we describe a method to cryopreserve the sperm
of the Greenshell mussel (Perna canaliculus) and how to use the sperm post-thawing to maximize larval
production from thawed sperm in selective breeding.
Key words Sperm, Greenshell mussel, Cryopreservation, Fertility, Larvae
1 Introduction
Cryopreservation can be a powerful tool in selective breeding and

hatchery production of shellfish for aquaculture. In selective breed-
ing, it enables breeders to make parental crosses on demand: for
example, crosses out of season or between stock types with differ-
ent breeding seasons, between generations, and even beyond the
lifespan of the donors, giving breeders much more flexibility and
enabling genetic gains to be made faster. It also manages the bio-
logical and commercial risks that are associated with selective
breeding: for example, animal loss due to disease, breeding for
resistance to new or emerging diseases, or economic and market
changes [15]. In ecotoxicology, the gametes and larvae of marine
invertebrates are used to carry out direct toxicity assessments for
monitoring and assessing the impact of pollutants on the marine
environment [610]. The ability to use cryopreserved material
may enable these assessments to be carried out year-round with the
same species and genetic material.
There are numerous papers published describing methods for
cryopreserving marine invertebrate gametes and larvae (e.g., [15],
[1018]). However, shellfish cryopreservation is only recently being
329
330 Serean L. Adams et al.
incorporated into selective breeding for shellfish as protocols are

determined to be sufficiently reliable.
The Greenshell mussel (Perna canaliculus) is the main shell-
fish species farmed in New Zealand [19]. The industry currently
obtains spat (juveniles) to on-grow from two wild populations.
However, there is a move toward hatchery spat production to
ensure a reliable supply for on-growing and to capture the benefits
of selective breeding. The method for cryopreserving Greenshell
mussel sperm presented here is based on our previously published
method but with some modifications [15]. Recommendations for
how to use the sperm post-thawing for selective breeding purposes
are also given here.
2 Materials
Before beginning, familiarize yourself with the safe handling and

material safety data sheets for dimethyl sulfoxide, trehalose, and
liquid nitrogen [21].
2.1 Chemicals Prepare solutions using ultrapure water (Milli-Q water purified to
and Solutions attain a sensitivity of 18.2 M cm at 25 C) (see Note 1). Dimethyl
sulfoxide (Me2SO) should be reagent grade or higher (see Note 2).
Trehalose should be of food grade (e.g., TREHA, Hayashibara,
Japan) or higher.
1. Trehalose solution (1 M): Weigh 15.2 g of trehalose and add
30 mL of ultrapure water in a sterile bottle. Mix until dissolved
(see Note 3).
2. Cryoprotective agent (CPA) solution: In a sterile bottle, mix
12 mL of 1 M trehalose solution, 20 mL of ultrapure water,
and 8 mL of Me2SO (see Note 4). Cool solution in the refrig-
erator prior to use (46 C).
2.2 Cryopreservation 1. Storage Dewar containing liquid nitrogen.

Equipment 2. Transport Dewar containing liquid nitrogen.
3. Metal rack attached to polystyrene frame 3 cm high for cryo-
preserving straws on (see Fig. 1).
4. Polystyrene box and lid (25 cm 33 cm 21.5 cm).
5. Storage goblet.
6. Safety equipment: cryo-gloves and protective face shield.
7. Long metal forceps.
8. Thermocouple.
9. Water bath.
10. Dish support.
11. Micro-aspirator or tubing with filter to aspirate dilute sperm
into straws.
Cryopreservation of Mussel Sperm 331
Fig. 1 0.5 mL straws are laid out on a metal rack attached to 3 cm-high polysty-
rene frame that is floated over liquid nitrogen
2.3 Cryopreservation 1. 0.5 mL plastic cryo-straws (IMV, LAigle, France).

Consumables 2. Colored sealing powder.
3. Bubbler inserts and dish.
4. Test tubes (12 75 mm) with stoppers.
5. Tissues.
6. Pipettes and pipette tips.
2.4 Other Equipment 1. Controlled-temperature 1 m filtered seawater.

and Consumables 2. Spawning trays.
3. Containers for individual mussels.
4. Neubauer hemocytometer.
5. 12 well cell culture plates.
3 Methods
3.1 Broodstock Mature mussels are able to be spawned almost all year-round
both in the hatchery and in the wild, although a distinct spawn-
ing period exists for wild populations ([20] and Adams, pers.
obs.). In the hatchery, broodstock are maintained on a mixed diet
of pond algae at ambient temperature, although it is preferable to
limit the maximum and minimum temperatures that the broodstock

experience during the peaks of summer and winter to within that
experienced in the wild.
3.2 Gamete 1. Induce mussels to spawn using thermal cycling [22]. Place
Collection mature mussels into shallow spawning trays (approximately 30
and Handling per tray) containing 1 m filtered seawater that is approxi-
mately 4 C higher than the temperature broodstock are cur-
rently maintained in. Cover the tray for approximately 2 h
before reducing the water temperature to approximately 4 C
below ambient broodstock seawater temperature. After
30 min, increase the temperature to 4 C above ambient
broodstock temperature. Repeat water changes every 30 min
until animals have begun spawning. In general, animals should
spawn after two to four cycles.
2. Rinse spawning individuals with 1 m filtered seawater and
separate into individual containers.
3. Place spawning males anterior up in individual containers and
leave to spawn dry (i.e., without seawater). Collect sperm
every 30 min and store in a refrigerator at 46 C for up to
24 h until ready to cryopreserve.
4. Place spawning females into individual containers with
~500 mL of cold filtered seawater (see Note 5). Collect
oocytes every 30 min and store in a refrigerator at 46 C for
up to 6 h until ready to fertilize. Replace seawater in the con-
tainer after each collection.
5. Determine sperm concentration using a Neubauer
hemocytometer.
6. Examine a sample of oocytes from each collection kept at ambi-
ent seawater temperature for at least 20 min microscopically.
Discard collections that have already been fertilized (as indi-
cated by the presence of polar bodies and/or cleaved embryos).
7. Pool uncontaminated oocytes from the same female and deter-
mine oocyte density by counting 20 L aliquots from a known
dilution of concentrated oocytes.
3.3 Cryopreservation 1. Aliquot 12 mL of sperm into test tubes precooled on ice.

2. Dilute sperm 1:1 with CPA solution in four equal volume steps
2030 s apart (see Note 6). Stopper tubes and invert gently to
mix contents between each addition.
3. Aspirate diluted sperm into 0.5 mL plastic straws and place on
the bubbler insert to create an air gap in the end of each straw.
4. Seal the straws by tapping the end of each straw into a con-
tainer of colored sealing powder. Place the straws, sealing the
powder end down into a beaker containing a small volume of
Milli-Q water to wet the sealing powder.
5. Wipe each straw with a tissue and place on a stainless steel rack
attached to a polystyrene frame precooled over ice.
6. Fill a polystyrene box (25 cm 33 cm 21.5 cm) with approxi-
mately 3.5 L of liquid nitrogen.
7. Once the liquid nitrogen has cooled the polystyrene box, gen-
tly float the rack with the straws on the liquid nitrogen and
place the lid on the box immediately (see Note 7).
8. After 10 min (see Note 8), remove the straws from the rack
and plunge into liquid nitrogen using forceps. Pack straws into
goblets and transfer to a storage Dewar until required.
3.4 Thawing 1. Carefully remove straws from the storage Dewar using forceps
and plunge into an ambient seawater temperature water bath
(normally 1418 C).
3.5 Small-Scale The post-thaw fertility of sperm may be assessed in a small-scale

Post-thaw Fertilization assay to determine whether sperm from a given individual has fro-
Assay zen well without the need to thaw several straws from the same
male. Fertilization is measured over a range of sperm concentra-
tions to generate a sigmoidal fertilization curve for each individual
frozen. Cryopreserved sperm have a lower fertility than fresh sperm
and must therefore be evaluated over a higher concentration range.
1. Bovine serum albumin (BSA) solution: Dissolve 12.0 g of BSA
(AlbuMAX I lipid-rich BSA, Gibco Invitrogen Corporation,
New Zealand) in 600 mL filtered seawater to give a 2 % w/v
solution. Aliquot into 50 mL plastic flasks and store in a 20 C
freezer until required for use.
2. Ethylenediaminetetraacetic acid (EDTA) solution: Dissolve
14.88 g of EDTA in approximately 800 mL of ultrapure water
in a sterile bottle and adjust pH to 8.2. Add additional water
to bring the volume to 1 L to give a 40 mM solution.
3. Prepare fertilization seawater for small-scale assays by adding
50 mL BSA solution and 2.5 mL of EDTA solution to
947.5 mL of 1 m filtered seawater to give final concentrations
of 0.1 % BSA and 100 M EDTA.
4. Aliquot 2.75 mL of fertilization seawater into each well of a
12-well cell culture plate. Prepare 1 plate per individual.
5. Dilute oocytes to a concentration of 2,667/mL using fertiliza-
tion seawater.
6. Add 225 L of diluted oocyte suspension to each well of the
12-well plates.
7. Dilute thawed sperm to 109 sperm/mL using fertilization sea-
water in a glass test tube (see Note 9).
8. Prepare a dilution series of sperm in glass test tubes. Add
315 L of sperm at 109 sperm/mL to a test tube containing
685 L of fertilization seawater. Repeat this step to obtain

sperm concentrations of between 106.5 and 109 sperm/mL.
9. For each sperm concentration, fertilize 2 wells by adding 30 L
of the serially diluted sperm to each well so that six sperm
concentrations occupy all 12 wells in a plate (see Note 10).
10. An assay with fresh sperm may also be prepared to confirm the
fertility of sperm before it is frozen and/or to assess the quality
of the oocytes used in the post-thaw fertilization assay. Fresh
sperm should be diluted as outlined in steps 7 and 8 but
diluted further so that sperm concentrations between 104 and
108 sperm/mL are assayed to give final concentrations of
between 102 and 106 sperm/mL.
11. Two wells containing oocytes but no sperm may also be used
as negative controls.
12. Incubate 12-well cell culture plates at ambient seawater tem-
perature (1418 C) for approximately 3 h.
13. Fix oocytes/embryos by adding 150 L of 10 % formalin (in
1 m filtered seawater).
14. Examine under a light microscope and determine the percent-
age fertilized by counting approximately 100200 oocytes/
embryos. Oocytes that are fertilized will have cleaved and be at
early embryonic stages (216 cell stage).
15. Sperm that have frozen well will have a sigmoidal fertilization
curve over the sperm concentration range evaluated.
3.6 Fertilization 1. Fill a glass beaker with 800 mL of 1 m filtered seawater and
and Incubation 2 mL of 40 mM EDTA solution. Add 800,0004,000,000
for Production of oocytes to each beaker and 5,000 thawed sperm per oocyte
D-Larvae from (see Note 11).
Cryopreserved Sperm 2. Stir every few minutes for 1520 min (see Note 12).
3. Transfer the contents of the beakers to incubation tanks
containing 1 m filtered seawater and 4 M EDTA at ambient
seawater temperature. Incubate at a density of up to 130
embryos/mL.
4. Collect D-larvae 3660 h postfertilization by screening the
contents of the tank through a 45 m mesh sieve (see Note
13). The D-larvae are then transferred to larval rearing tanks
for rearing under normal hatchery conditions.
4 Notes
1. Solutions are generally prepared on the day of use but may be

stored for up to a week in the refrigerator at 46 C if required.
2. Dimethyl sulfoxide crystallizes at 16 C and can be melted by
warming in a warm water bath (37 C) if required.
3. A warm water bath at 37 C can be used to assist with dissolv-

ing the trehalose into solution.
4. The addition of Me2SO to a solution results in an exothermic
reaction. Allow time for the solution to cool prior to being
used.
5. The temperature of the water is usually the same as that used
during the cold cycle for spawning (i.e., 4 C below
ambient).
6. Single-step addition generally gives equivalent post-thaw
fertility.
7. The time is flexible in that no difference in fertility was found
between treatments exposed to the CPA solution from 5 to
30 min.
8. Times of between 10 and 20 min have shown no difference in
fertility.
9. While preparation of the sperm dose-response dilutions and
then fertilization are normally done immediately upon thaw-
ing, intervals of up to 30 min between thawing and fertiliza-
tion have not affected fertility.
10. The sperm addition is normally done in sequence starting at
the lowest sperm dose to minimize any chance of
contamination.
11. Ideally a ratio of 5,00010,000 sperm per oocyte should be
used to maximize fertilization success. However, if oocyte
numbers are not limited and the goal is to maximize the num-
ber of embryos obtained from a limited number of cryopre-
served sperm, then the operator can potentially increase the
number of embryos by fertilizing more eggs at a lower sperm
per oocyte ratio.
12. If desired, a sample of fertilized embryos can be taken (160
600 oocytes/embryos) after 15 min and incubated in a 12-well
culture plate with 3 mL of fertilization seawater for a further
3 h to determine fertilization success.
13. The time taken for embryos to reach the D-larval stage is
dependent upon the temperature of the seawater during
incubation.
Acknowledgment
This research was funded by the New Zealand Ministry of Business,

Innovation, and Employment (CAWX0802 and CAWX1315). We
thank SPATnz and Kono for supplying mussels and technical assis-
tance and also thank the staff of Cawthron Aquaculture Park (in par-
ticular Jonathan Morrish and Ellie Watts) for technical assistance.
References
1. McFadzen IRB (1995) Cryopreservation of 12. Adams SL, Hessian PA, Mladenov PV (2004)
sperm of the Pacific oyster Crassostrea gigas. Cryopreservation of sea urchin (Evechinus chlo-
Methods Mol Bio 38:145149 roticus) sperm. Cryo Letters 25:287299
2. Adams SL, Smith JF, Roberts RD, Jankec AR, 13. Dong Q, Eudeline B, Huang C, Allen SK Jr,
Kaspar HF, Tervit HR, Pugh PA, Webb SC, Tiersch TR (2005) Commercial-scale sperm
King NG (2004) Cryopreservation of sperm of cryopreservation of diploid and tetraploid
the Pacific oyster (Crassostrea gigas): develop- Pacific oysters, Crassostrea gigas. Cryobiology
ment of a practical method for commercial spat 50:116
production. Aquaculture 242:271282 14. Adams SL, Smith JF, Tervit HR, McGowan
3. Tervit HR, Adams SL, Roberts RD, McGowan LT, Roberts RD, Janke AR, King NG, Gale SL,
LT, Pugh PA, Smith JF, Janke AR (2005) Webb SC (2011) Cryopreservation of mollus-
Successful cryopreservation of Pacific oyster can sperm: Pacific oyster, Green-lipped mussel
(Crassostrea gigas) oocytes. Cryobiology 51: and Paua abalone. In: Tiersch TR, Green CC
142151 (eds) Cryopreservation in aquatic species, 2nd
4. Tiersch TR, Yang H, Jenkins JA, Dong Q edn. World Aquaculture Society, Baton Rouge,
(2007) Sperm cryopreservation in fish and LA, pp 562573
shellfish. Soc Reprod Fertil Suppl 65:493508 15. Smith JF, Adams SL, Gale SL, McGowan LT,
5. Adams SL, Smith JF, Roberts RD, Janke AR, Tervita HR, Roberts RD (2012)
King NG, Tervit HR, Webb SC (2008) Cryopreservation of Greenshell mussel (Perna
Application of sperm cryopreservation in selec- canaliculus) sperm. I. Establishment of freezing
tive breeding of the Pacific oyster, Crassostrea protocol. Aquaculture 334337:199204
gigas (Thunberg). Aquacult Res 39:14341442 16. Paredes E, Adams SL, Tervit HR, Smith JF,
6. Environment Agency UK (2007) The direct McGowan LT, Gale SL, Morrish JR, Watts E
toxicity assessment of aqueous environmental (2012) Cryopreservation of Greenshell mus-
samples using the oyster (Crassostrea gigas) sel (Perna canaliculus) trochophore larvae.
embryo-larval development test: methods for Cryobiology 65:256262
the examination of waters and associated materi- 17. Paredes E, Bellas J, Adams SL (2013)
als, Environment Agency UK, Bristol, UK. Comparative cryopreservation study on two
http://www.environment-agency.gov.uk/ species of bivalves: Pacific oyster (Crassostrea
static/documents/Resear ch/oyster- gigas) and blue mussel (Mytilus galloprovincia-
209jan30_1388168.pdf. Accessed 26 Feb 2014. lis). Cryobiology 67:274279
7. Environment Canada (2011) Biological test 18. Liu Y, Xu T, Robinson N, Qin J, Li X (2014)
method: fertilization assay using echinoids (Sea Cryopreservation of sperm in farmed Australian
Urchins and Sand Dollars) EPS 1/RM/27, greenlip abalone Haliotis laevigata.
2nd ed, Ottawa, Canada. http://www.ec.gc. Cryobiology 68:185193
ca/Publications/047B08AB-530E-49EA-8EC8- 19. Aquaculture New Zealand (2012) New
8AB8E8D75DA4/1_BiologicalTestMethod Zealand aquaculture: a sector overview with
FertilizationAsayUsingEchinoidsSeaUrchins key facts, statistics and trends. http://aquacul-
SandDollars2ndEdition.pdf. Accessed 26 Feb ture.org.nz/wp-content/uploads/2012/05/
2014. NZ-Aquaculture-Facts-2012.pdf. Accessed 13
8. ASTM (2012) E724: standard guide for con- Dec 2013.
ducting static acute toxicity tests starting 20. Alfaro AC, Jeffs AG, Hooker SH (2001)
with embryos of four species of saltwater Reproductive behaviour of the green-lipped
bivalve molluscs. ASTM International, West mussel, Perna canaliculus, in northern New
Conshohocken, PA Zealand. Bull Mar Sci 69:10951108
9. ASTM (2012) E1563: standard guide for con- 21. Martinez-Pastor F, Adams SL (2008)
ducting static acute toxicity tests with echi- Cryobiological material and handling proce-
noid embryos. ASTM International, West dures. In: Cabrita E, Robels V, Herraez P (eds)
Conshohocken, PA Methods in reproductive aquaculture: marine
10. Paredes E, Bellas J (2009) Cryopreservation of and freshwater species. CRC, Boca Raton, FL,
sea urchin embryos (Paracentrotus lividus) pp 295319
applied to marine ecotoxicological studies. 22. Adams SL, Tervit HR, McGowan LT, Smith
Cryobiology 59:344350 JF, Roberts RD, Salinas-Flores L, Gale SL,
11. Paniagua-Chavez CG, Tiersch TR (2001) Webb SC, Mullen SF, Critser JK (2009)
Laboratory studies of cryopreservation of Towards cryopreservation of Greenshell mussel
sperm and trochophore larvae of the eastern (Perna canaliculus) oocytes. Cryobiology
oyster. Cryobiology 43:211223 58:6974
Chapter 15
Membrane Modification Strategies for Cryopreservation

Phillip H. Purdy and James K. Graham
Abstract
Cell membranes can be modified using cyclodextrins loaded with lipids or unilamellar liposomes. Lipid
choice can greatly influence the organization of the targeted membrane and result in a cell that is more
capable of surviving cryopreservation due to altered membrane-phase transition properties or membrane
reorganization that may alter the normal physiologic processes of the treated cell. The protocols described
here explain the preparation of the cyclodextrins and liposomes, impact of the amount and type of lipids,
and general principles for treating cells using either of these technologies.
Key words Cyclodextrin, Liposome, Lipid, Cryopreservation
1 Introduction
The plasma membrane organization of a cell comprises phospho-

lipids, proteins, cholesterol, and other lipids. The constituents will
vary according to the cell type (e.g., spermatozoa, oocyte, somatic)
and the species from which it was obtained. This variation in mem-
brane components also results in varied responses to the cryo-
preservation process. As an example, sperm from bulls, roosters,
and stallions respond very differently when placed into an environ-
ment below 12 C without the presence of a penetrating cryopro-
tectant (e.g., glycerol) or a non-penetrating cryoprotectant (e.g.,
egg yolk). The rooster sperm will be essentially unaffected, but a
significant portion of bull sperm will die or exhibit some altered
physiology, and the vast majority of the stallion sperm will die. The
reason for these different responses is due to the differences in the
cholesterol and phospholipid content and accordingly their molar
ratios [1]. Cells with a cholesterol to phospholipid ratio approaching
1 (roosters) have a membrane that is capable of remaining fluid at
lower temperatures and consequently can respond to the stresses
of the cryopreservation process whereas cells with significantly
lower ratios (stallions) are not capable of responding to the stresses
and cell damage and death ensues [1]. Cells with intermediate
337
338 Phillip H. Purdy and James K. Graham
phospholipid-to-cholesterol ratios (bull) have a mixed response

and may survive the process but with altered membrane organiza-
tion and physiology [1].
The rationale for lipid-loaded cyclodextrins and liposomes is
based on these concepts. Specifically, if the cholesterol to phospho-
lipid ratio can be altered and consequently the membrane fluidity
at low temperatures increased, then cells should be better suited to
survive cryopreservation. Application of these techniques demon-
strates that greater proportions of sperm from boars [24], bucks
[5, 6], bulls [710], rams [1113], stallions [1417], sperm from
other species [1820] as well as oocytes [21, 22], and somatic cells
[23, 24] survive cryopreservation following treatment with either
method. However, there is tremendous deviation in the success
which is attributed to dose, type of phospholipid and chain length
[8, 25], cholesterol and its analogues [26], and type of cyclodex-
trins as well as species variation.
Treating cells using these technologies can also impact cellular
function. Introduction of lipids into membranes at these non-
physiologic levels can alter sperm capacitation and the acrosome
reaction [25, 27], and modulate intracellular calcium during these
processes [28]. In addition, because their membranes have been
modified cells can experience greater tolerance to osmotic stress
and can consequently, survive osmotic pressures that are higher or
lower that what untreated cells can withstand [4]. Exploration of
these technologies has enabled us to better understand the function
of the plasma membrane and how it responds to cryopreservation.
2 Materials
1. Cholesterol, methanol, cyclodextrins, and all chemicals are

reagent grade (Sigma Aldrich, St. Louis, MO, USA).
2. All media is prepared using distilled, deionized Nanopure water.
3. Glass petri dishes are 80 mm 15 mm.
4. Nitrogen gas.
5. Glass vials with caps to hold the dried loaded cyclodextrins.
6. Metal spatulas for removing the loaded cyclodextrins from the
petri dishes.
For preparation of liposomes:
7. Lipids, for preparation of liposomes, may be purchased from
Avanti Polar Lipids, Alabaster, Alabama, USA.
8. Liposome buffering medium: 277.5 mM glucose, 8.3 mM lac-
tose, 5.0 mM raffinose, 1.7 mM sodium citrate dehydrate,
2.5 mM potassium citrate, 29.8 mM HEPES, pH 7.27.4,
300 mOsm.
Membrane Modification Strategies 339
9. Probe sonicator: Branson Sonifier 450, Danbury, CT, USA;

Output control setting of 2 and Duty cycle % setting of
Constant.
10. Extruder: The Extruder from Lipex Biomembranes,
Vancouver, B.C. Canada equipped with a 0.2 m pore filter.
11. Centrifuge.
12. Ammonium ferrothiocyanate solution: 27.03 g of ferric chlo-
ride hexahydrate and 30.4 g of ammonium thiocyanate dis-
solved to 1 L in deionized distilled water.
13. Glass Pasteur pipettes.
3 Methods
3.1 Cholesterol- 1. Dissolve 200 mg of cholesterol in 1 mL of chloroform

Loaded Cyclodextrin (see Note 1).
(CLC) Preparation 2. Dissolve 1 g of cyclodextrin in 2 mL of methanol in a separate
beaker (see Note 2).
3. Add 0.45 mL of the cholesterol solution to the cyclodextrin
solution and stir. The solution should become clear. If it does
not become clear adding a few drops of methanol should make
the solution homogeneous (clear).
4. Transfer the solution to a clean glass petri dish and use a stream
of nitrogen gas to remove the chloroform and methanol.
5. Allow the resulting crystals to dry, without nitrogen gas, for an
additional 24 h.
6. Scrape the crystals from the petri dish and store them in a
sealed glass container at room temperature.
7. A working solution of the CLCs is prepared by adding 50 mg
of the CLC to 1 mL of buffered media, warming to 37 C,
followed by vortex mixing to clarify the solution.
3.2 Preparation 1. Preferred lipids (single or combinations) are dissolved in chlo-

of Liposomes roform in glass test tubes (see Note 3).
2. The tubes are flushed with nitrogen gas in a fume hood until
all of the chloroform is evaporated.
3. The sample is then diluted with 12 mL of liposome buffering
medium.
4. Samples are emulsified using a probe sonicator and sonicated
for 5 min or until the solution is clarified.
5. During sonication the tubes are continuously flushed with
nitrogen gas.
6. Samples are then centrifuged at 27,000 g for 15 min and the
supernatant is transferred to a clean tube.
3.3 Alternative 1. Unilamellar liposome vesicles are then prepared by forcing this
Approach solution through the extruder (see Note 4).
for Liposome 2. The resulting lipids are stored in aliquots at 80 C and can be
Preparation thawed in a 39 C water bath prior to use.
3.4 Determination 1. Prepare a solution of ammonium ferrothiocyanate (AF)

of the Concentration (see Note 5).
of Lipids 2. Prepare a standard solution by dissolving 10 mg of the lipids
in the Liposome used for preparation of the liposomes in chloroform to a
Preparation 100 mL final volume.
3. Serial dilutions of the lipid standard are then prepared by dilut-
ing known volumes of the lipid (0.11.0 mL) with 2 mL of the
AF solution.
4. The final volume of the lipids and AF solution is adjusted to
4 mL with chloroform.
5. The solution is then mixed with a vortex mixer for 1 min.
6. The lower phase of the solution is then removed with a Pasteur
pipette and placed in a spectrophotometer cuvette. If necessary
the solution can be clarified with a small amount of anhydrous
sodium sulfate.
7. The optical densities of the standards containing known quan-
tities of lipids are determined at 488 nm and the transmission
is plotted to form a standard curve.
8. Samples of liposomes are then prepared in the same manner as
the controls, the optical density is measure, and the lipid con-
tent determined from the standard curve.
3.5 Treating Cells 1. Aliquots of desired cell numbers may be diluted in buffered
with Lipid-Loaded solutions containing no lipids (milk, egg yolk) or cells can be
Cyclodextrins or treated in neat, concentrated form.
Liposomes 2. Samples can be treated at room temperature or an ambient
temperature specific to a cell type.
3. Incubation typically lasts for 15 min for mammalian species
but should be optimized by cell type.
4. The samples can then be diluted with cryopreservation diluent
containing lipids such as milk or egg yolk, and frozen using
practices customized to a cell type.
4 Notes
1. Preparation of CLCs is based on a protocol created by Klein

et al. [14] and modified by Purdy and Graham [7].
2. The choice of the type of cyclodextrin impacts cellular quality after
treatment. Each type (, , methyl-, etc.) of cyclodextrin has
varying size and as a result interacts with a membrane differently.
Membrane Modification Strategies 341
Consequently, it is prudent to test the lipid-cyclodextrin interac-

tion when attempting use of lipids that are previously untested.
3. Choice of lipid for use in either the cyclodextrin loading or the
liposomes is up to the user. However, when the cells are treated
the responses can be highly variable due to membrane reorga-
nization and modulation of physiologic responses due to the
type and size of the lipid.
4. We have prepared and utilized liposomes according to Hope
et al. [29] or in a modified method according to Wilhelm
et al. [15]; both work equally well.
5. The method for determination of the lipid content is based
on the manuscript by Stewart [30].
References
1. Parks JE, Lynch DV (1992) Lipid composition uration on the motility of bull spermatozoa after
and thermotropic phase behavior of boar, bull, cold shock and freezing. Cryobiology 24:4252
stallion and rooster sperm membranes. 9. Moraes EA, Graham JK, Torres CAA, Meyers
Cryobiology 29:255266 M, Spizziri B (2010) Delivering cholesterol or
2. He L, Bailey JL, Buhr MM (2001) cholestanol to bull sperm membranes improves
Incorporating lipids into boar sperm decreases cryosurvival. Anim Reprod Sci 118:148154
chilling sensitivity but not capacitation poten- 10. Rpke T, Oldenhof H, Leiding C, Sieme H,
tial. Biol Reprod 64:6979 Bollwein H, Wolkers WF (2011) Liposomes
3. Toms C, Blanch E, Fazeli A, Moc E (2013) for cryopreservation of bovine sperm.
Effect of a pre-freezing treatment with Theriogenology 76:14651472
cholesterol-loaded cyclodextrins on boar sperm 11. Holt WV, North RD (1988) The role of
longevity, capacitation dynamics, ability to membrane-active lipids in the protection of
adhere to porcine oviductal epithelial cells ram spermatozoa during cooling and storage.
in vitro and DNA fragmentation dynamics. Gamete Res 19:7789
Reprod Fert Dev 25:935946 12. Moc E, Purdy PH, Graham JK (2010)
4. Toms C, Blanch E, Cebrin B, Moc E (2013) Treating ram sperm with cholesterol-loaded
In vivo fertilising ability of frozen-thawed boar cyclodextrins improves cryosurvival. Anim
sperm treated with cholesterol-loaded cyclo- Reprod Sci 118:236247
dextrins prior to cryopreservation. Anim 13. Motamedi-Mojdehi R, Roostaei-Ali Mehr M,
Reprod Sci 140:7782 Rajabi-Toustani R (2014) Effect of different
5. Konyali C, Toms C, Blanch E, Gmez EA, levels of glycerol and cholesterol-loaded cyclo-
Graham JK, Moc E (2013) Optimizing condi- dextrin on cryosurvival of ram spermatozoa.
tions for treating goat semen with cholesterol- Reprod Dom Anim 49:6570
loaded cyclodextrins prior to freezing to improve 14. Klein U, Gimpl G, Fahrenholz F (1995)
cryosurvival. Cryobiology 67:124131 Alteration of the myometrial plasma membrane
6. Farshad A, Amidi F, Khor AK, Rashidi A cholesterol content with -cyclodextrin modu-
(2011) Effect of cholesterol-loaded- lates the binding affinity of the oxytocin recep-
cyclodextrin in presence and absence of egg tor. Biochemistry 34:1378413793
yolk during freezing step on quality of mark- 15. Wilhelm KM, Graham JK, Squires EL (1996)
hoz bucks spermatozoa. Asian-Aust J Anim Effects of phosphatidylserine and cholesterol
Sci 24:181189 liposomes on the viability, motility, and acroso-
7. Purdy PH, Graham JK (2004) Effect of mal integrity of stallion spermatozoa prior to
cholesterol-loaded cyclodextrin on the cryo- and after cryopreservation. Cryobiology
survival of bull sperm. Cryobiology 48:3645 33:320329
8. Graham JK, Foote RH (1987) Effect of several 16. Pillet E, Labbe C, Batellier F, Duchamp G,
lipids, fatty acyl chain length, and degree of unsat- Beaumal V, Anton M, Desherce S, Schmitt E,
Magistrini M (2012) Liposomes as an alterna- 23. Stadnick H, Stoll C, Wolkers WF, Acker JP,
tive to egg yolk in stallion freezing extender. Holovati JL (2011) The effect of liposome treat-
Theriogenology 77:268279 ment on the quality of hypothermically stored
17. Oldenhof H, Gojowsky M, Wang S, Henke S, Yu red blood cells. Biopreserv Biobank 9:335342
C, Rohn K, Wolkers WF, Sieme H (2013) 24. Stoll C, Holovati JL, Acker JP, Wolkers WF (2012)
Osmotic stress and membrane phase changes Synergistic effects of liposomes, trehalose, and
during freezing of stallion sperm: mode of action hydroxyethyl starch for cryopreservation of human
of cryoprotective agents. Biol Reprod 88:68 erythrocytes. Biotechnol Prog 28:364371
18. He Q, Lu G, Che K, Zhao E, Fang Q, Wang 25. Graham JK, Foote RH, Parrish JJ (1986)
H, Liu J, Huang C, Dong Q (2011) Sperm Effect of dilauroylphosphatidylcholine on the
cryopreservation of the endangered red spot- acrosome reaction and subsequent penetration
ted grouper, Epinephelus akaara, with a special of bull spermatozoa into zona-free hamster
emphasis on membrane lipids. Aquaculture eggs. Biol Reprod 35:413424
318:185190 26. Amorim EAM, Graham JK, Spizziri B, Meyers
19. Kiso WK, Asano A, Travis AJ, Schmitt DL, M, Torres CAA (2009) Effect of cholesterol or
Brown JL, Pukazhenthi BS (2012) cholesteryl conjugates on the cryosurvival of
Pretreatment of Asian elephant (Elephas maxi- bull sperm. Cryobiology 58:210214
mus) spermatozoa with cholesterol-loaded 27. Purdy PH, Graham JK (2004) Effect of adding
cyclodextrins and glycerol addition at 4 C cholesterol to bull sperm membranes on sperm
improves cryosurvival. Reprod Fert Dev capacitation, the acrosome reaction, and fertil-
24:11341142 ity. Biol Reprod 71:522527
20. Hussain SA, Lessard C, Anzar M (2013) A 28. Nolan JP, Graham JK, Hammerstedt RH
strategy for improvement of postthaw quality (1992) Artificial induction of exocytosis in bull
of bison sperm. Theriogenology 79:108115 sperm. Arch Biochem Biophys 292:311322
21. Zeron Y, Tomczak M, Crowe J, Arav A (2002) 29. Hope MJ, Bally MB, Webb G, Cullis PR
The effect of liposomes on thermotropic mem- (1985) Production of large unilamellar vesicles
brane phase transitions of bovine spermatozoa by a rapid extrusion procedure. Characterization
and oocytes: implications for reducing chilling of size distribution, trapped volume and ability
sensitivity. Cryobiology 45:143152 to maintain a membrane potential. Biochem
22. Horvath G, Seidel GE Jr (2006) Vitrification Biophys Acta 812:5565
of bovine oocytes after treatment with 30. Stewart JCM (1980) Colorimetric determina-
cholesterol-loaded methyl--cyclodextrin. tion of phospholipids with ammonium ferro-
Theriogenology 66:10261033 thiocyanate. Anal Biochem 104:1014
Chapter 16
Sperm Cleanup and Centrifugation Processing

for Cryopreservation
Abstract
Fertility rates with artificial insemination are highest with good-quality sperm samples. Therefore, nonviable
sperm, cellular debris, and seminal plasma are preferably removed from semen samples prior to use or for
preservation. Such compounds are sources where reactive oxygen species are generated during storage or
upon cryopreservation, impairing sperm function. In this chapter we describe methods to remove seminal
plasma and cellular debris from sperm samples, and for selecting morphologically normal motile sperm.
The methods that are described here include: ordinary centrifugation, sperm swim-up, glass wool and
Sephadex filtration/adherence, and single-layer as well as discontinuous two-layer iodixanol density gradi-
ent centrifugation.
Key words Semen, Spermatozoa, Sperm selection, Subpopulations, Cleanup
1 Introduction
In order to increase fertility rates, sperm processing methods can

be employed to remove damaging compounds present in an ejacu-
late and to select morphologically normal motile sperm. In addi-
tion to viable sperm, an ejaculate contains seminal plasma, immature
and dead sperm, cellular debris, as well as leukocytes. These com-
pounds have been reported to increase levels of reactive oxygen
species causing a decrease in viable and functional sperm during
storage and after cryopreservation and consequently decreased fer-
tility rates [1].
Different procedures can be used for removal of seminal
plasma and sperm selection, including migration (swim up) [2]
and filtration/adhesion (e.g., using glass wool or Sephadex) [35]
methods, as well as ordinary and density centrifugation [6]. In
addition, more advanced methods can be used such as magnetic
activated cell sorting (MACS) or electrostatic charge-based sperm
separation [7, 8].
343
Dilution and washing via ordinary centrifugation is a simple

and quick semen processing method for removal of seminal plasma
and to concentrate sperm in a small volume. Sperm is collected in
a pellet, the supernatant is discarded, and sperm are washed with
fresh diluent. This procedure collects particles in a pellet, and does
not discriminate between different cell types or sperm subpopula-
tions (i.e., with differences in membrane/chromatin intactness or
motility characteristics). Collection of sperm in a dense pellet in
close proximity with cellular debris can be damaging to sperm [9,
10]. Sperm recovery is dependent on the g-force that is applied
and the duration of centrifugation, as well as the type and geome-
try of the tube that is used and the sample volume [11, 12].
Migration methods rely on motility characteristics of sperm
within a sample. They are commonly used to enrich semen samples
with morphologically normal motile sperm and to remove debris.
The swim-up procedure involves centrifugation to collect sperm in
a sperm pellet, which is subsequently covered with medium allow-
ing motile sperm to migrate during an incubation period. Motile
sperm are collected from this layer (see Fig. 1). Even though only
motile sperm are selected by this method, the number of recovered
sperm is generally very low.
Filtration methods make use of sperm motility and their ten-
dency to adhere to matrices such as glass wool or beads. Motile sperm
can be separated from immotile cells, since they can move through
a b c
centrifuge diluted semen overlay collected sperm motile sperm migrate
collect sperm in a pellet with fresh medium recover sperm from upper layer
remove supernatant place tube in decanted position
Incubate at 37C
Fig. 1 Schematic presentation of sperm swim-up method, for recovering motile sperm. First, centrifugation is
used to collect and concentrate sperm in a pellet (a), after which the supernatant is removed and the sperm
layer is carefully overlayed with fresh medium (b). The tube is placed at an angle of 45 to increase the contact
area between the two layers, and incubated at 37 C. Motile sperm will migrate to the layer with fresh medium;
from which they can be recovered using a Pasteur pipette (c)
Sperm Clean-up and Processing 345
densely packed glass wool fibers, or Sephadex beads (see Fig. 2).
This approach can be used for whole ejaculates to enrich the sample
with motile sperm, and to effectively remove leukocytes. Small debris,
however, is not removed from the sample.
Density gradient centrifugation is used to select a sperm sub-
population from other particles by centrifugation through a den-
sity gradient solution. During centrifugation, sperm passes
through layers that have a lower density until they appear as a
a b c
in a syringe/column add Sephadex layer overlay with diluted semen
add glass wool layer wash column with diluent recover sperm
that migrate through the column
Fig. 2 Schematic presentation of preparing column with glass wool and Sephadex layers for sperm filtration/
migration. First glass wool is added in the bottom of a column/syringe (a), then an additional Sephadex layer
can be added on top of this (b), and the column can be washed using diluent. A stopper can be used to seal
the column, when needed. Diluted sperm should be carefully added, until the column is completely filtrated.
Particles will adhere to the column material, and motile sperm can migrate through the column material when
adding diluent. Sperm are recovered in a collection tube underneath the column (c)
band at a density coinciding with their own. With single-layer

density centrifugation, sperm are collected as a pellet underneath
a colloid of a lower density [13]. In the two-layer (discontinu-
ous) density centrifugation method, a bottom layer with a higher
density than sperm is used to collect sperm on top of a cushion
layer (see Fig. 3) [1416].
Separation approaches for sperm typically are based on differ-
ences in properties of subpopulations, including differences in
viability, motility, morphology, size, and density. In this chapter we
describe examples of sperm cleanup and selection approaches
including (1) dilution and washing via ordinary centrifugation, (2)
sperm migration/sedimentation methods (swim-up), (3) filtra-
tion/adherence methods (i.e., using glass wool and/or Sephadex),
and (4) density gradient centrifugation (i.e., using polyvinylpyr-
rolidone/PVP or silane-coated silica particles, or iodixanol).
a b c d
in a centrifuge tube underneath overlay with diluted semen recover sperm
add top layer add bottom layer centrifuge that passed through layers
with lower density as sperm
Fig. 3 Schematic presentation of iodixanol double-layer density gradient centrifugation. To prepare the discon-
tinuous double-layer gradient: first add the upper layer in a centrifuge tube (a), after which the lower (denser)
layer can be added underneath using a syringe with needle (b). Carefully add the diluted semen on top of the
upper layer, without disturbing the density gradient layers (c). During centrifugation, sperm will pass through
the layers which have a lower density. In case of using an upper and a lower layer which have a lower and
higher density as sperm, respectively, sperm will be collected at the interface of these layers (d). The diluent
in which the sperm was originally diluted will remain above the upper density gradient layer, while dense
particles will be collected in the pellet below the lower density gradient layer
2 Materials
2.1 Dilution 1. Extender for diluting and stabilizing semen. Extenders can be
and Centrifugation home-made, but commercially available extenders can also be
used. Water bath set at 37 C, for pre-warming of diluents.
2. Centrifuge (for 3001,000 g), and tubes for centrifugation of
samples. Plastic round bottom tubes for 10 mL, or conical
tubes for 50 mL can be used (e.g., from Corning Life Sciences,
Lowell, MA, USA), as well as round bottom centrifuge glasses
for volumes up to 100 mL. In addition, cushion-fluid (e.g.,
Optiprep from Axis-Shield PoC, Oslo, Norway) can be used.
3. Device for removing supernatant (vacuum pump, or syringe).
4. Phase contrast microscope (with heated stage), slides and cover
glasses. For determining the sperm concentration: a hemocy-
tometer (Neubauer counting chamber).
2.2 Sperm Swim-Up 1. Modified Tyrodes or the so-called TALP medium (96 mM
NaCl, 3.1 mM KCl, 0.4 mM MgSO4, 5 mM glucose, 15 mM
NaHCO3, 2 mM CaCl2, 0.3 mM KH2PO4, 20 mM HEPES,
21.7 mM sodium lactate, 1 mM sodium pyruvate, 3 mg BSA
mL1; pH at 37 C of 7.4 and osmolality of 300 mOsmol kg1.
Prepare fresh TALP medium. Dissolve in 100 mL water:
19.2 mL 0.5 M NaCl, 620 L 0.5 M KCl, 400 L 0.1 M MgSO4,
1,000 L 0.5 M glucose, 5 mL 0.3 M NaHCO3, 400 L 0.5 M
CaCl2, 300 L 0.1 M KH2PO4, 10 mL 0.2 M HEPES, (if desired
add: 500 L 20 mg mL1 gentamycin), 0.243 g sodium lactate,
0.011 g sodium pyruvate (or: 312 L of a 60 % syrup solution),
adjust the pH to 7.6 at 20 C. Check the osmolality, and adjust
to ~300 mOsm kg1 by adding NaCl if needed. Filter sterilize,
and add 3 g BSA prior to use. Let the solution equilibrate (for
several hours) in an incubator set at 37 C supplemented with
5 % CO2. Stock solutions should be filter-sterilized, and stored at
4 C. Sodium lactate can be prepared as aliquots in microtubes.
TALP medium preferably should be prepared fresh (can be
stored for up to 1 week at 4 C).
2.3 Glass Wool 1. Glass wool (e.g., coarse structure glass wool from Jrgens,
and Sephadex Hannover, Germany; glass wool microfiber code 112, from
Filtration Mannville Co, Denver, CO, USA; or fine-structure glass wool
like SpermFertil from MELLO Ltd, UK) (see Note 1).
2. Sephadex-G15 (e.g., from Sigma-Aldrich, Deisenhofen,
Germany) (see Note 2).
3. Plastic disposable syringes (e.g., for 10 or 50 mL) that can
serve as column, as well as holder for positioning the column
upright, and collection tubes for underneath.
2.4 Density Gradient 1. Percoll (e.g., Percoll of ~1.13 g mL1, from GE Healthcare
Centrifugation BioSciences AB, Uppsala, Sweden), or other commercially
available coated colloid optimized for single layer density cen-
trifugation of sperm from species of interest (e.g.,
PureSperm100 from Nidacon, Gothenburg, Sweden).
2. Iodixanol: a 60 % solution of ~1.32 g mL1 is commercially
available as OptiPrep (e.g., via Axis-Shield PoC, Oslo, Norway)
(see Note 3). Isotonic buffer for preparing solutions of differ-
ent densities: e.g., Hanks buffered salt solution (HBSS;
5.33 mM KCl, 0.441 mM KH2PO4, 4.17 mM NaHCO3,
137.92 mM NaCl, 0.338 mM NaH2PO4, 5.56 mM glucose;
pH of 7.4 and osmolality of 300 mOsmol kg1).
(a) Prepare a 16 % iodixanol solution of ~1.090 g mL1, for
use as upper layer: mix 13.3 mL 60 % iodixanol and
36.7 mL isotonic buffer.
(b) Prepare a 30 % iodixanol solution of ~1.165 g mL1, for
use as lower layer: mix 25 mL 60 % iodixanol and 25 mL
isotonic buffer.
3 Methods
3.1 Dilution 1. Dilute raw semen by slowly adding your favorite primary
and Washing via expender/diluent pre-warmed at 37 C. Add at least an equal
Ordinary volume of diluent, or dilute to ~50100 106 sperm mL1.
Centrifugation 2. Transfer diluted semen into a centrifuge tube (e.g., ~10
50 mL), and centrifuge for 10 min at 600 g, at ambient tem-
perature (see Notes 4 and 5).
3. Carefully take the tube from the centrifuge; without disturbing
the pellet. Remove the supernatant using a vacuum pump or
syringe, leaving a little volume (see Note 6). Gently resuspend
the pellet, using mild agitation.
4. Determine the sperm concentration for the resuspended pellet;
using a hemocytometer. Determine the recovered sperm number
by multiplying the cell concentration with the sample volume.
Calculate the recovered sperm fraction by dividing the recovered
sperm number by the original cell number (see Note 4).
3.2 Migration/ 1. Add an equal volume of TALP medium (e.g., 1 mL each) to

Sedimentation diluted semen.
Methods (Swim-Up) 2. Centrifuge for 5 min at 1,000 g, at ambient temperature.
Remove the supernatant, leaving 500 L in which the sperm
pellet can be resuspended (see Note 7).
3. Carefully add a layer of 500 L fresh TALP medium on top of
the sperm suspension (omit mixing with the sperm layer).
Incubate at 37 C for about 4560 min (in an incubator with
5 % CO2; to stimulate sperm capacitation when needed). In

order to increase the contact area between the two layers, the
tube can be positioned at an angle of 45 (see Fig. 1).
4. Transfer the upper layer, containing the migrated sperm, to a
clean tube (e.g., about 100 L), using a Pasteur pipette.
3.3 Filtration/ 1. Glass wool filtration (see Note 8):

Adherence Methods (a) Wash glass wool with the same diluent that has been used
to dilute the semen (to moisten and remove glass wool
fragments). Add the glass wool in a 5 mL syringe; press
tightly up to the mark indicating 1.5 mL.
(b) Place the syringe containing glass wool in a holder, and
position a collection tube below.
(c) Add 1 mL diluted semen in the syringe, on top of the glass
wool. When fully infiltrated in the glass wool, wash twice
with 1 mL diluent to recover sperm in the collection tube.
2. Glass wool-Sephadex filtration:
(a) Prepare 20 % Sephadex-G15 in advance: add 20 g
Sephadex-G15 in 100 mL of the diluent used for extend-
ing/diluting semen. Incubate for about 3 h at room tem-
perature while mixing occasionally (see Note 9). Mix well
before use.
(b) Wash glass wool with sperm extender, and add in a syringe.
Add about 1.5 g glass wool in a 50 mL syringe; press
tightly up to the mark indicating 6 mL.
(c) Place the syringe containing glass wool in a holder, and
overlay the glass wool layer with 20 mL 20 % Sephadex-G15
slurry. If necessary, the column can be washed using fresh
diluent, or the opening of the syringe can be sealed.
(d) Carefully add diluted semen on top of the Sephadex layer
(without disturbing the layer), until it has completely infil-
trated the column.
(e) Wash the column using 25 mL diluent, to recover sperm
in the collection tube (see Fig. 2). The initial volume
fraction coming from the column not containing sperm
can be discarded, to obtain a more concentrated sperm
sample.
3.4 Density Gradient 1. Single-layer density gradient centrifugation:

Centrifugation (a) Add density gradient solution in a centrifuge tube, and
carefully overlay with diluted semen sample. In a 10 mL
conical tube: add 1 mL diluted semen on top of 4 mL
80 % Percoll solution (omit mixing of layers). In case of
using a commercial single-layer density gradient solution,
follow the instructions supplied by the manufacturer.
(b) Centrifuge for 30 min at 400 g, at room temperature.

This results in: an upper layer containing the diluent in
which sperm was originally extended, the Percoll/density
gradient solution underneath, and a pellet of sperm which
were able to sediment through this layer.
(c) Carefully remove the upper layers, and recover the pellet
in a small volume.
2. Double-layer density gradient centrifugation:
(a) Prepare a discontinuous density gradient consisting of two
layers in a 50 mL conical plastic tube. First add 10 mL of
the top layer (16 % iodixanol solution of ~1.090 g mL1).
Using a syringe with needle, add 10 mL of the bottom
layer (30 % iodixanol solution of ~1.165 g mL1) under-
neath (see Note 10).
(b) Carefully add a layer of diluted semen (20 mL) on top of
the density gradient.
(c) Centrifugation at 1,000 g for 20 min (and let brake
slowly). Spermatozoa show up as a band (at the interface)
between the upper and bottom layer (see Fig. 3). Take
from the centrifuge without disturbing the layers.
(d) Carefully remove the upper layer, and transfer the layer
containing sperm into a clean tube. A syringe to which a
straw is attached can be used for this.
4 Notes
1. Different types of glass wool are available which can be tested

for their efficacy. Glass wool properties which should be taken
into account include: the type (i.e., borate glass, silicate glass,
or quartz glass), surface structure and charge, as well as thick-
ness of the fibers or the pore size of the filter.
2. Different types of Sephadex are available which can be tested
for their efficacy. Sephadex-G15 is typically used for sperm
cleanup methods.
3. Other agents such as Percoll or Ficoll can also be used to
prepare (discontinuous) density gradients. For clinical pur-
poses, sterile silane-coated silica colloids are available.
Compounds like sucrose can also be used to prepare density
gradients, but cannot be used here because this will expose
sperm to osmotic stress. The densities of solutions can be
checked using a refractometer, when needed.
4. The g-force applied for collecting sperm typically is in the
range of 350700 g, with a duration ranging from 10 to
15 min. Sperm quality after centrifugation is generally better
using a lower centrifugation force for a longer duration.

However, this coincides with greater losses of sperm. With
ordinary centrifugation, typically about 5080 % of the sperm
are recovered; when centrifuging 40 mL sperm solution
(~100 106 sperm mL1) in a 50 mL tube at 600 g for 10 min.
5. To minimize dense packing of sperm in the pellet during cen-
trifugation, tubes with a larger bottom surface can be used;
that is, round bottom tubes can be used instead of conical
tubes. Moreover, cushion centrifugation can be utilized. With
cushion centrifugation the sperm sample is layered on top of a
dense solution (e.g., a 60 % iodixanol solution) covering the
full diameter of the tube. With cushion centrifugation a higher
g-force can be used for longer duration (e.g., 1,000 g for
20 min), which yields a higher sperm recovery. After centrifu-
gation, sperm are present as a layer on top of the cushion fluid,
which has a higher density as the sperm.
6. High seminal plasma contents in preserved semen samples
have been reported to have deleterious effects, both for stor-
age at 5 C and for cryopreservation. However, low amounts
of seminal plasma in sperm freezing extenders (525 %) result
in increased motility post-thaw [17].
7. Since a centrifugation step is utilized, damage coinciding with
collection of sperm and debris in a dense pellet may occur. To
avoid using centrifugation, a migration-sedimentation proce-
dure can be used which uses a special tube with an inner cone.
Spermatozoa swim up directly from liquefied semen into the
medium layered on top, and subsequently sediment in the
inner cone.
8. Nonspecific adherence of cells to matrices can be explained by
their surface properties; for example they can exhibit sticky
molecules at their membrane surface, or can be charged.
Specific components can also be removed making use of beads
coated with antibodies against those components; e.g. leuko-
cytes can be removed from a sample using magnetic beads cov-
ered with the common leukocyte antigen CD45.
9. The Sephadex solution can be prepared in a diluent which can
be stored for longer periods at 4 C (e.g., 3 % sodium citrate
solution). The column should be washed extensively with the
diluent used for diluting semen before adding sperm sample.
10. Preparing discontinuous density gradients can be done both
either by carefully adding layers on top of each other, or by
adding more dense layers below each other, using a syringe
with needle. When recovering a layer of interest, one can
remove layers starting at the top (or bottom), or alternatively
position a syringe with needle within the layer of interest to
directly recover the sample.
The two-layer iodixanol density centrifugation procedure

described here is referred to as purification by sedimentation
and is described in detail by Stuhtmann et al. [16]. Smith et al.
[18] have described an alternative approach in which sperm is
layered below the density gradient layers. With this so-called
purification by flotation, good-quality sperm will swim up
against the applied g-force.
References
1. Henkel R (2012) Sperm preparation: state-of- subjected to centrifugation. Cell Biochem
the-artphysiological aspects and application of Biophys 31:231245
advanced sperm preparation methods. Asian J 12. Knop K, Hoffmann N, Rath D, Sieme H
Androl 14:260269 (2005) Effects of cushioned centrifugation
2. Mahadevan M, Baker G (1984) Assessment technique on sperm recovery and sperm quality
and preparation of semen for in vitro fertiliza- in stallions with good and poor semen freez-
tion. In: Wood C, Trounson AO (eds) Clinical ability. Anim Reprod Sci 89:294297
in vitro fertilization. Springer, Berlin, p 83 13. Morrell JM, Rodriguez-Martinez H,
3. Jeyendran RS, Perez-Palaez M, Crabo BG Johannisson A (2010) Single layer centrifuga-
(1986) Concentration of viable spermatozoa for tion of stallion spermatozoa consistently selects
artificial insemination. Fertil Steril 45:132134 the most robust spermatozoa from the rest of
4. Graham EF, Graham JK (1978) The effect of the ejaculate in a large sample size: data from 3
whole ejaculate filtration on the morphology breeding seasons. Equine Vet J 42:579585
and the fertility of bovine semen. J Dairy Sci 14. Morrell JM, Johannisson A, Dalin AM,
73:9197 Rodriguez-Martinez H (2009) Morphology
5. Sieme H, Martinsson G, Rauterberg H, Walter and chromatin integrity of stallion spermato-
K, Aurich C, Petzoldt R, Klug E (2003) zoa prepared by density gradient and single
Application of techniques for sperm selection layer centrifugation through silica colloids.
in fresh and frozen-thawed stallion semen. Reprod Domest Anim 44:512517
Reprod Dom Anim 38:134140 15. Mortimer D, Mortimer ST (2013) Density
6. Henkel RR, Schill WB (2003) Sperm prepara- gradient separation of sperm for artificial
tion for ART. Reprod Biol Endocrinol 1:108 insemination. In: Carrell DT, Aston KI (eds)
7. Sakkas D (2013) Novel technologies for select- Spermatogenesis: methods and protocols, vol
ing the best sperm for in vitro fertilization and 927, Methods in molecular biology. Humana
intracytoplasmic sperm injection. Fert Steril Press, New York, NY, pp 217226
99:10231029 16. Stuhtmann G, Oldenhof H, Peters P, Klewitz J,
8. Mehta A, Sigman M (2014) Identification and Martinsson G, Sieme H (2012) Iodixanol den-
preparation of sperm for ART. Urol Clin North sity centrifugation for selecting stallion sperm
Am 41:169180 for cold storage and cryopreservation. Anim
9. Aitken RJ, Clarkson JS (1988) Significance of Reprod Sci 133:184190
reactive oxygen species and antioxidants in 17. Sieme H (2011) Freezing semen. In: McKinnon
defining the efficacy of sperm preparation tech- AO, Squires EL, Vaala WE, Varner DD (eds)
niques. J Androl 9:367376 Equine reproduction, 2nd edn. Blackwell
10. Mortimer D (1994) Sperm recovery tech- Publishing Ltd., Chichester, pp 29722982
niques to maximize fertilizing capacity. Reprod 18. Smith TT, Byers M, Kaftani D, Whitford W
Fertil Dev 6:2531 (1997) The use of iodixanol as a density gradi-
11. Katkov I, Mazur P (1999) Factors affecting ent material for separating human sperm from
yield and survival of cells when suspensions are semen. Arch Androl 38:223230
Chapter 17
Cryopreservation of Red Blood Cells

Johan W. Lagerberg
Abstract
Cryopreservation of red blood cell concentrates (RBCs) is an important method for maintaining an inventory
of rare RBC units and managing special transfusion circumstances. The permeating additive glycerol is
used as a cryoprotectant to protect RBCs against freezing damage. The use of thawed RBCs was hampered
a 24-h outdating period due to potential bacterial contamination when a functionally open system was
used for addition and removal of the glycerol. With the introduction of a functionally closed system for the
glycerolization and deglycerolization of RBC units, extended post-thaw storage became possible. Here, we
describe the cryopreservation of red blood cells according to the high-glycerol method, using a function-
ally closed processing system.
Key words Cryopreservation, Glycerol, Red blood cells
1 Introduction
Cryopreservation is a valuable approach for managing an inventory

of rare red blood cell concentrate (RBC) units to provide compat-
ible blood for patients which are negative for blood group antigens
with a very high frequency in the population (so-called public
antigens). Frozen RBCs can also be used for military deployments,
which are characterized by logistical problems and unpredictable
needs of blood components [1, 2].
Glycerol is the most commonly used RBC cryoprotectant.
Freezing RBCs with glycerol as a cryoprotectant dates back to
1950 [3] after an accidental discovery the previous year [4]. In the
1960s, it was found that accelerating the rate of freezing could
significantly reduce the required concentration of glycerol to about
20 % [5, 6]. After several modifications [711], the two methods,
referred to as the high- (HGM) and low-glycerol method (LGM),
respectively, are still in use for freezing units of RBCs. With the
LGM, glycerol is added to a final concentration of approximately
20 % (w/v). The RBCs are frozen rapidly (i.e. >100 C/min) and
are stored at temperatures below 140 C, normally in or above
353
354 Johan W. Lagerberg
liquid nitrogen. On the other hand, RBCs can be frozen slowly

(i.e., ~13 C/min) by the high-glycerol method (HGM). With
this method, RBCs are frozen with a final concentration of approx-
imately 40 % glycerol (w/v) and stored at temperatures between
65 C and 80 C. Overall, RBC preservation can be extended to
at least 10 years if the correct storage temperature is guaranteed
[12, 13].
Despite their higher costs [14, 15], frozen RBCs have several
advantages that are mainly related to the necessary washing proce-
dure, eliminating cell debris, WBCs, cytokines, and free Hb [16,
17]. Besides cost, the actual use of frozen RBCs is hampered
mainly by processing time and a 24-h outdating period due to
potential bacterial contamination if thawing and washing is per-
formed in a non-closed system [12]. With the introduction of an
automated cell processing system (ACP215, Haemonetics
Braintree, MA) it became possible to both glycerolize and deglyc-
erolize RBCs in a functionally closed system, resulting in reduced
potential for bacterial contamination and allowing prolonged post-
thaw storage [18, 19]. Using the ACP215, all fluids needed for the
glycerolization and deglycerolization procedure are administered
through 0.22 m bacteria barrier filters, and all tubing connections
are achieved using a sterile connection device to maintain a closed
system. When the ACP215 is used for processing, post-thaw stor-
age time of HGM frozen RBCs can be extended to 27 days [20]
when stored in saline-adenine-glucose-mannitol (SAGM) solution
and to 14 days when stored in additive solution-3 (AS-3) [19].
In this chapter, the cryopreservation of red blood cells accord-
ing to the high-glycerol method using a functionally closed pro-
cessing system is described. The described process is used in our
institute for the preservation of rare RBC units. Although the
ACP215 has been developed for processing RBCs according to the
HGM, with some adjustment in the procedure, it can also be used
for deglycerolization of LGM units [21]. Since with the LGM
method, freezing was performed in aluminum containers that can-
not be sterilely connected, post-thaw storage of these units is lim-
ited to 24 h.
2 Materials
2.1 Equipment 1. ACP215 automatic cell processor (Haemonetics, Braintree, MA).

2. Centrifuge for blood bags (e.g., Hettich Roto Silenta RP).
3. Centrifuge for tubes (e.g., Hettich Rotanta).
4. Freezer 80 C.
5. HemoCue plasma/low Hb analyzer.
6. Kocher clamp.
7. Manual plasma extractor.
Cryopreservation of Red Blood Cells 355
8. Micro hematocrit centrifuge.

9. Non-contact thermometer.
10. Scale.
11. Sterile connection device (Terumo).
12. Tube sealer (e.g., Fresenius CompoSeal Mobilea).
13. Vacuum bag sealer.
14. Water bath.
2.2 Reagents 1. Glycerol 57 % solution (6.2 M, S.A.L.F. S.p.A., Cenate Sotto

(Bergamo) Italy), bottle 500 mL. Contains per 100 mL: 57 g
glycerol, 1.6 g sodium lactate, 30 mg monobasic sodium phos-
phate (monohydrate), 124 g dibasic sodium phosphate (water
free), pH 6.8.
2. 12 % sodium chloride solution (Bio-Deglyc, Bioluz, Saint-
Jean-de-Luz, France), bag 250 mL.
3. Washing solution, 0.9 % NaCl, 0.2 % dextrose (Bio-wash,
Bioluz), bag 2,000 mL.
4. Red cell additive solution SAG-M (Haemonetics, bag 300 mL).
5. Red blood cell concentrate, leukocyte depleted.
2.3 Disposables 1. Capillaries for hematocrit analysis.

2. Cardboard box (Cekumed, Ooltgensplaat, the Netherlands).
3. Deglycerolization set (LN235, Haemonetics).
4. Freezing bag: 1,000 mL PVC bag VSE 6002Z (MacoPharma,
Mouvaux, France).
5. Glycerolization set (LN225, Haemonetics).
6. Plastic overwrap bag.
7. Transfer bag 150 mL (e.g., P4159, Fresenius HemoCare,
Emmer-Compascuum, the Netherlands).
8. Transfer bags 500 mL (e.g., P4162, Fresenius HemoCare).
3 Methods
3.1 Glycerolization The starting materials in our process are leukocyte-reduced red
and Freezing of Red blood cell concentrates (RBC) in SAG-M. The RBC are prepared
Blood Cell according to the buffy coat method from 500 mL whole blood
Concentrates donations after overnight hold at ambient temperature [22].
Briefly, after a hard spin, the whole blood is separated in components
using an automated blood component separator (Compomat,
Fresenius HemoCare). RBC are diluted with 110 mL SAGM and
leukocyte reduced using the inline filter. The RBCs are glycero-
lized and frozen shortly after production, preferably within 7 days
after collection (see Note 1). The RBC should be stored at 26 C

in the period between production and freezing. Before glyceroliza-
tion, the supernatant (additive solution) of the RBC is removed.
3.1.1 Removal 1. Sterilely connect a 150 mL transfer bag to the RBC bag, open
of Additive Solution the weld and transfer about 5 mL cell suspension to the trans-
fer bag. This will be used for quality control of the starting
product.
2. Weigh and record the weight of the unit. To obtain the net weight,
subtract the weight of the empty bag from the gross weight.
3. Sterilely connect a new 150 mL transfer bag to the RBC bag.
This step will also disconnect the first 150 mL bag from the
RBC bag. Do not open the weld.
4. Place the RBC bag (and attached transfer bag) with the ports
upright in the bucket of the blood bag centrifuge and fill the
bucket with water-filled bags to position the blood bag. Weigh
the filled bucket and pair it with a bucket with the same weight
(2 g).
5. Spin the unit at 3,200 g in a 22 C centrifuge for 5 min. To
minimize resuspension of the red blood cell-additive solution
interface, the brake should be set low.
6. Gently remove the unit from the centrifuge and place it in a
plasma extractor. Place a Kocher clamp on the tube between
blood bag and transfer bag.
7. Open the weld. Open the Kocher clamp and express the super-
natant fluid to the transfer bag. Replace the Kocher clamp on
the tubing between the two bags.
8. Detach the waste bag (containing supernatant) from the RBC
bag and discard.
9. Weigh and record the weight of the supernatant-reduced unit.
To obtain the net weight, subtract the weight of the empty bag
from the gross weight.
3.1.2 Warming The uptake of glycerol by RBC is an active process [23]. For that
Procedure reason, at the time of glycerolization, the RBC and glycerol solu-
tion should be within a temperature range of 2530 C. The glyc-
erol solution can be stored at room temperature. The RBCs should
be removed from the refrigerator and stored at room temperature
for about 1 h prior to glycerolization.
1. Using a non-contact thermometer, check the temperature of
the RBC bag and the glycerol solution.
2. To warm the RBC bag or glycerol bottle place them in a plastic
overwrap. Immerse the wrapped RBC bag or glycerol bottle in
a 37 C water bath until the desired temperature is reached.
Take care that no water enters the plastic overwrap bag.
3.1.3 Glycerolization Based on the weight of the RBC, the ACP215 automatically adds
Procedure a volume of 57 % glycerol to the RBC to achieve a final glycerol
concentration of 40 % (w/v). The glycerolization procedure takes
about 10 min.
1. Power on the ACP215. The ACP215 automatically performs a
series of internal system checks.
2. Select the glycerolization protocol.
3. Open the glycerolization disposable set and inspect it for visi-
ble defects.
4. Install the disposable set according to the manufacturers
instructions [24].
5. Close the ratchet clamp on the spike connector tubing segment.
6. Sterilely connect the RBC bag to the glycerolization dispos-
able. Do not open the weld.
7. Load the blood pump tubing with the segment of tubing
between the two pump stops. Be sure that the blue pump stop
is toward the back of the machine. Secure the front and back
pump stops through their respective guides.
8. Load the pump outlet tubing into the blood pump air detector
(BLAD).
9. Connect the draw pressure monitor (DPM) filter to the pres-
sure sensor located on the front of the ACP215.
10. Remove the metal pull tab from the glycerol bottle, disinfect
the rubber stopper with an alcohol swab (70 %) and aseptically
insert the spike end of the glycerolization disposable into the
glycerol bottle. Assure that the air vent on the spike is open to
allow air to flow into the bottle while it empties. If necessary,
squeeze the drip chamber to prime the chamber.
11. Suspend the glycerol bottle on the raised IV pole of the ACP215.
12. Place the RBC bag on the shaker platform and secure with the
three shaker magnets. For optimum performance the shaker
should be placed at the same level as the machine. Unclamp
the ratchet clamp on the glycerol line.
13. Check to confirm that the set is installed properly and press
YES. After doing this, the machine guides the user through a
set of verification questions to ensure the proper loading of the
disposable set. Upon the question Sterile dock RBC bag open
the weld between the RBC bag and the glycerolization set.
14. The ACP215 prompts the operator to enter his/hers initials
(see Note 2).
15. Press Modify Program and fill in the weight of the RBC unit
using the YES/UP or NO/DOWN arrows.
16. Press START to begin the glycerolization process. The glyc-
erol solution will be added to the RBC at a computed flow rate.
The glycerol flow rate will progressively increase during the

process assuring that the osmolarity rate will not increase more
than 500 mOsm/kg/min.
17. When the computed volume of glycerol has been processed,
the pump stops. The shaker stops 30 s later. The message
Glycerolization complete will be displayed.
18. The printer will print out information related to the glycero-
lization procedure.
19. Heat seal the tubing between the RBC bag and the glycero-
lization set.
20. Detach the RBC bag from the rest of the glycerolization set.
21. Remove the disposable set from the machine and discard the
disposable set.
22. Sterilely dock a 1,000 mL freezing bag to the bag containing
the glycerolized red blood cells. Leave about 10 cm of tubing
between the bags.
23. Open the weld and transfer the cells to the freezing bag by grav-
ity. Weigh and record the net weight of the unit (see Note 3).
3.1.4 Removal Before freezing the RBC unit, the volume of the unit is reduced
of Supernatant Glycerol and the hematocrit of the suspension is increased by removing
supernatant glycerol [11] (see Note 4).
1. Sterilely dock a 500 mL transfer bag to the freezing bag. Do
not open the weld.
2. Roll the bottom of the bag until the RBCs are firmly packed at
the top of the freezing bag and fixate with elastic bands.
3. Spin the unit at 1,248 g in a 22 C centrifuge for 10 min. To
minimize red cell mixing, the brake should be set low.
4. Gently remove the unit from the centrifuge and place it in a
plasma expressor.
5. Squeeze weld and express all air and visible supernatant glyc-
erol to the transfer bag.
6. Place a Kocher clamp on the tubing between the two bags.
7. Remove the glycerolized RBC from the plasma expressor and
thoroughly mix the unit.
3.1.5 Freezing the Unit The glycerolized cells should be frozen slowly, with a cooling rate
of 13 C/min. This can be achieved by placing the glycerolized
units in a cardboard box at the bottom of a chest freezer.
1. Sterilely dock a 150 mL transfer bag to the freezing bag.
Squeeze weld and transfer about 10 mL of cell suspension to
the transfer bag. This will be used for quality control and filling
reference samples (cryovials).
2. Weigh the glycerolized unit and record the net weight of

the unit.
3. Label the unit and freezing box as required.
4. Fold the unit in half with the ports and tubing tucked under
the top of the bag. Place the unit inside a sealable bag.
5. Seal the plastic bag using an impulse sealer so that there is as
little trapped air as possible, preferably by using a vacuum sealer.
6. Place the bag plus the reference samples (cryovials containing
1.5 mL glycerolized RBC solution) inside a cardboard freez-
ing box.
7. Place the cardboard box in a 80 C freezer. Units should be
frozen in a location within the freezer where optimal freez-
ing can occur (preferably at the bottom of a chest freezer).
To avoid improper freezing, the units should not be stacked
on each other.
8. After 24 h, the units can be stacked and stored in chest or
upright 80 C freezers.
3.2 Thawing In the deglycerolization process, the ACP215 first dilutes the glyc-
and Deglycerolization erolized blood with 50 mL of 12 % NaCl while shaking the unit
of Red Blood Cell thoroughly on the shaking platform. After an equilibration delay,
Concentrates 340 mL of 0.9 % NaCl/0.2 % glucose solution is added while using
the shaking platform to ensure adequate mixing. The diluted cells
are transferred to the bowl where the supernatant glycerol and
saline is removed. The product is pumped back to the bag and
400 mL of 0.9 % NaCl/0.2 % glucose solution is added while using
the shaking platform to ensure adequate mixing. In both dilution
steps the dilution rate is calculated by the ACP to prevent an osmo-
larity decrease rate exceeding 500 mOsm/kg/min. After the sec-
ond dilution, the diluted cells are transferred to the bowl where the
washing procedure in the bowl, using five washing steps with 0.9 %
NaCl/0.2 % glucose solution, begins. Effluent exiting the bowl
will be collected in the waste bag. The line sensor monitors the free
hemoglobin level and cell spillage in the effluent. At the comple-
tion of the washing procedure, the cells are washed once with addi-
tive solution and finally resuspended in the additive solution. The
deglycerolized-resuspended RBCs transferred from the washing
bowl into the blood product bag of the disposable set. The entire
deglycerolization procedure takes about 6070 min.
3.2.1 Thawing the Unit 1. Remove the frozen unit from the freezer and from the card-
board box.
2. Place the unit with the overwrap bag in a water bath main-
tained at 40 C.
3. After 20 min, remove the overwrap and unfold the blood bag.
4. Monitor the temperature of the blood bag using a non-contact

thermometer. The temperature of the blood bag should be
between 25 and 30 C. Warm the unit if needed, taking care
not to wet the ports of the bag.
3.2.2 Deglycerolization 1. Power on the ACP215. The ACP215 automatically performs a

Procedure series of internal system checks.
2. Select the deglycerolization protocol.
3. Open the deglycerolization disposable set and inspect it for
visible defects.
4. Install the disposable set according to the manufacturers
instructions.
5. After loading the disposable set, the display posts the message
asking the operator if the Line Sensor Calibration Check is
needed. It is recommended to perform this check each day of
use. The line sensor is calibrated using the filters labeled 2 and
3 that are delivered with the ACP215. To calibrate, follow the
instructions on the display (see Note 5).
6. After the Line Sensor Calibration check has been completed or
bypassed, the Bowl Seating Test will be performed to ensure
that the bowl is properly loaded. Upon the message Is centri-
fuge cover locked? close the centrifuge lid and press YES.
The centrifuge will start running for 30 s. In the display the
following message appears Did bowl produce unusual noise.
Press NO if the bowl turns smoothly without loud noise. If
the bowl produced loud noise, press YES. In this case dis-
card the disposable and install a new disposable set.
7. Hang the washing solutions on the upper hook of the raised
IV pole. Using aseptic technique, connect the washing solu-
tions to the disposable:
(a) Spike the 12 % NaCl bag to the blue striped line.
(b) Spike the Saline/Glucose bag to the yellow striped line.
(c) Connect the Luer lock connector of the SAGM bag to the
orange striped line and crack the breakable seal.
8. Sterilely connect the red striped line to the tubing of the
thawed RBC bag. Leave the weld closed.
9. Place the thawed RBC unit on the shaker, using the magnetic
pins with the ports toward the front of the shaker.
10. Verify that all colored tubing segments are loaded properly in
the appropriate valves and press YES. After doing this, the
machine guides the user through a set of verification questions
to ensure the proper loading of the disposable set. Mistakes in
loading the disposable set can be corrected in this stage.
Improper installation of the disposable set may result in prod-
uct damage.
11. Upon the message Sterile Connect RBC bag to RED striped
tubing open the weld between RBC bag and disposable set.
Check the weld. If the weld is not correct reconnect the unit
and note that the unit is not sterilely processed and has a 24 h
post-wash storage limit.
12. The ACP215 prompts the operator to enter his/hers initials
(see Note 2).
13. Upon the message press modify to set parameters press
Modify and enter the weight of the RBC unit using the YES/
UP or NO/DOWN arrows. Press Modify. The ACP215 will
display the following message: Thawed unit hematocrit. Use
the YES/UP or NO/DOWN arrows to adjust the setting to
60 % in case glycerol was removed before freezing or to 30 % if
glycerol was not removed before freezing.
14. Press START to begin the deglycerolization process.
15. After about 15 min, the ACP215 will evaluate the integrity of
the connection between the RBC bag and the disposable by
holding and monitoring a negative pressure for a period of
time. The message Sterile connection check in progress will
be displayed. A steady pressure level indicates an intact weld,
the machine will continue operation as normal. If the test fails,
the intactness of the weld should be visually inspected. If the
weld is intact, press START to continue the procedure. If
the weld is incorrect and NO is pressed, a statement will be
printed on the print out of the procedure stating that the shelf
life of the blood product is limited to 24 h (see Note 6).
16. When the deglycerolization procedure is completed the
machine will beep and the following message will be displayed:
Is the color of waste supernatant acceptable? The color should
be nearly colorless. Press YES if so, or NO if the color is
not acceptable. Pressing the YES or NO key causes the
ACP215 to automatically resume operation. If NO is chosen,
it will be marked on the print out that the color was not
acceptable (see Note 7).
17. The printer will print out information related to the deglycero-
lization procedure.
18. Close all clamps on the lines of the disposable set to prevent
flowing of fluids.
19. Heat seal the tubing between the RBC bag and the deglycero-
lization set and between the waste bag and the deglyceroliza-
tion set. Label the RBC bag before disconnecting from the
ACP215.
20. Detach the RBC bag and waste bag from the rest of the
deglycerolization set.
21. Weigh the RBC bag and the waste bag and record the net weights.
22. Sterilely connect a sample bag to the RBC bag and transfer
about 5 mL of cell suspension for quality control.
23. Remove the disposable set from the machine and discard the
disposable set.
24. Store the deglycerolized RBC at 26 C until use.
3.3 Deglycerolization Although the ACP215 has been developed for processing RBCs
of LGM Frozen Units according to the HGM, it can also be used for deglycerolization of
LGM units [21]. The only adjustment needed in the deglycero-
lization procedure is the dilution of the initially added hypertonic
salt solution. While HGM frozen cells required a first dilution with
12 % NaCl, LGM cells are first diluted with 6 % NaCl. Because the
LGM procedure glycerolization and freezing is not performed in a
functionally closed system, post-thaw storage of these units is lim-
ited to 24 h.
1. Aseptically dilute the 12 % NaCl solution 1:1 with water to
obtain a 6 % NaCl solution.
2. Transfer the thawed LGM cells from the storage container to a
1,000 mL transfer bag.
3. Perform the deglycerolization procedure as described above,
using the 6 % NaCl solution instead of the 12 % NaCl solution.
4. Upon the message: Thawed unit hematocrit, use the YES/
UP or NO/DOWN arrows to adjust the setting to 30 % (LGM
units are frozen without volume reduction).
3.4 Quality Control The original RBC bag, the glycerolized RBC bag, the waste bag
and the deglycerolized RBC bag are weight and net weight is
recorded. The volume of the blood components is calculated from
the net weight divided by the specific gravity: 1.027 for plasma,
1.100 for RBCs, 1.100 for 40 % glycerol and 1.006 for
SAGM. Using the hematocrit of the solution, the specific gravity of
RBC in additive solution is calculated from these values. For the
deglycerolization waste solution a specific density of 1.04 is used.
The original RBC bag, the glycerolized RBC bag and the deglyc-
erolized RBC bag are sampled using sterilely connected sample
bags (see Note 8).
1. Analyze samples for hemoglobin concentration using
HemoCue Hb analyzer and hematocrit using a micro hemato-
crit centrifuge (see Note 9).
2. Calculate total hemoglobin (g) as: Volume (mL) Hb concen-
tration (g/dL)/100.
3. Calculate freeze recovery (%) as: total Hb glycerolized RBC/
total Hb original RBC 100.
Table 1
Representative example of a freeze/thaw process. Values marked with
grey are measured values; all other values are calculated
Starting material
Nett weight RBC in SAGM (g) 317
Specific gravity (g/mL) 1.064
Volume starting material (mL) 298
Hb concentration (g/dL) 18.9
Total Hb (g) starting material 56.3
Glycerolization
Nett weight RBC after SAGM removal (g) 197
Nett weight RBC after glycerolization (g) 586
Nett weight RBC after glycerol removal (g) 296
Hematocrit after glycerol removal (%) 68
Volume frozen suspension (mL) 269
Hb concentration (g/dL) 20.0
Total Hb frozen suspension (g) 53.0
Freeze recovery (%) 94
Deglycerolization procedure
Waste fluid
Nett weight bag waste fluid (g) 1,683
Volume waste fluid (mL) 1,618
Hb concentration waste fluid (g/dL) 0.3
Total Hb in waste fluid (g) 5.2
Deglycerolized product
Nett weight RBC in SAGM (g) 302
Hematocrit (%) 50
Specific gravity (g/mL) 1.052
Volume RBC in SAGM (mL) 287
Hb concentration RBC in SAGM (g/dL) 14.7
Total Hb deglycerolized product 42.2
Freeze/thaw/wash recovery 75 %
4. Calculate freeze/thaw/wash recovery as: Total Hb deglycerolized

RBC/total Hb original RBC 100.
A representative example of the calculations of a glycero-
lization/deglycerolization procedure is depicted in Table 1. In
the validation of the deglycerolization procedure the efficacy
of the washing process should be determined. This can be
done by measuring the osmolarity of the supernatant of the
deglycerolized product (see Note 10).
5. After deglycerolization, the RBCs should be stored at 26 C
Since small changes in the glycerolization/deglycerolization
procedure can influence the post-thaw stability of the cells, the
post-thaw outdating time should be determined by each insti-
tute individually (see Note 11).
4 Notes
1. The military blood bank (MBB) in the Netherlands holds a

stock of frozen (80 C) RBCs (blood group O) for military
deployments. Their starting material is leukoreduced whole
blood. Whole blood units are leukoreduced after overnight
hold at ambient temperature. The units are processed and fro-
zen within 24 h after collection. Before glycerolization, the
plasma of the whole blood units is removed and before freezing
supernatant glycerol is removed [1, 2].
2. We use this for identification of the ACP215 used (AA, BB).
3. With an RBC volume of about 170 mL after removal of
supernatant, the volume of the original RBC storage bag
(600 mL) is large enough to hold the glycerolized cells (final
volume about 470 mL). With larger RBCs, for instance made
from leukodepleted whole blood, the RBCs should be trans-
ferred to the 1 L freezing bag before removal of the supernatant.
After removal of supernatant, the hematocrit of the suspension
is very high making transfer to another bag very hard.
4. Recently it has been shown, that removal of supernatant glyc-
erol before freezing is not necessary [25]. This simplifies the
glycerolization procedure, and it may also reduce RBC lesions
by shortening exposure to centrifugal force. Because of the
larger volume of the glycerolized cell suspension, omitting
pre-freeze volume reduction could require larger cardboard
boxes, storage containers and therefore more storage capacity.
After thawing, the deglycerolization procedure takes slightly
longer, but is efficient in removing all glycerol. Thawed units
frozen according to the modified glycerolization procedure
meet the quality requirements of the European Council and
AABB standards, and there are indications that the post-thaw
stability is improved in comparison to units that were volume-
reduced before freezing [25, 26].
5. Filters have to be tested in the right order. If tested in the

incorrect order, turn the power of the machine off and back
on. The Line Sensor Calibration Check can now be repeated.
6. Because the used freezing bag has a greater volume (1,000 mL)
than programmed in the ACP215, the machine can errone-
ously conclude that the sterile connection is not intact. To pre-
vent false negative result, at the start of the test, a Kocher
clamp can be placed on the line between the weld and the RBC
bag. Do not forget to remove the clamp after the test is
finished.
7. In the rare event that the color of the supernatant is not accept-
able (too red), or when the print-out indicates that the hemo-
globin content of the final wash step was above 50 mg%, a
manual washing step could be performed:
(a) Sterilely connect a 500 mL transfer bag to the RBC bag.
(b) Centrifuge to unit at 3,200 g in a 22 C centrifuge for
5 min.
(c) Place the RBC unit in a plasma extractor and express the
supernatant fluid. Check if supernatant is now colorless, if
not, the unit should be discarded.
(d) Add an equal volume of SAGM to the RBC unit.
The shelf life of a washed unit is limited to 24 h.
8. To take a representative RBC sample, the RBC bag should be
thoroughly mixed. After connection of the sample bag, the
RBC should be mixed between the RBC bag and the sample
bag at least three times before the final sample is taken.
9. The freezing and thawing process results in increased permea-
bility of the RBC membrane. Because of this, hematology ana-
lyzers could give wrong values for mean corpuscular volume
and hematocrit. For a correct determination of the hematocrit,
the spun hematocrit should be determined.
10. The supernatant should have an osmolality comparable with
that of the additive solution used. In general, an osmolality
value below 400 mOsm/kg indicates a residual glycerol level
of less than 1 g% [19]. An alternative method to determine the
efficient removal of glycerol is the measurement of the glycerol
concentration in the supernatant by use of a plasma triglycer-
ides test [21].
11. A major quality parameter in determining the post-thaw stor-
age time is hemolysis. According to European guidelines [12]
this should be below 0.8 % at the end of the storage period
(according to the AABB guidelines this should be below 1 %
[13]). Another parameter determining the storage time could
be the ATP content of the cells. There are indications that
there is a correlation between total adenylate [27] or ATP
content [28] and in vivo survival. For this reason, Sanquin

Blood Supply included the requirement that at the end of the
storage period, the ATP content should be above 2.7 mol/g
Hb. With this concentration of ATP it can be predicted that
the in vivo survival will be at least 75 % [28]. Based on both
quality parameters, the out-dating period of thawed units
resuspended in SAGM was set on 48 h [20]. The additive
solution AS-3 has been shown to better maintain RBC integ-
rity after thawing [21, 29]. When resuspended in AS 3, RBCs
can be stored for up to 14 days [19].
Acknowledgement
The author wishes to thank Dr. Femke Noorman, Military Blood

Bank, the Netherlands, for sharing knowledge, providing informa-
tion, and reviewing this manuscript.
References
1. Lelkens CC, Koning JG, de Kort B, Floot IB, 10. Valeri CR (1975) Simplification of the methods
Noorman F (2006) Experiences with frozen for adding and removing glycerol during
blood products in the Netherlands military. freeze-preservation of human red blood cells
Transfus Apher Sci 34:289298 with the high or low glycerol methods: bio-
2. Holley A, Marks DC, Johnson L, Reade MC, chemical modification prior to freezing.
Badloe JF, Noorman F (2013) Frozen blood Transfusion 15:195218
products: clinically effective and potentially 11. Valeri CR, Valeri DA, Anastasi J, Vecchione JJ,
ideal for remote Australia. Anaesth Intensive Dennis RC, Emerson CP (1981) Freezing in
Care 41:1019 the primary polyvinylchloride plastic collection
3. Smith AU (1950) Prevention of haemolysis bag: a new system for preparing and freezing
during freezing and thawing of red blood cells. nonrejuvenated and rejuvenated red blood
Lancet 2:910911 cells. Transfusion 21:138149
4. Polge C, Smith AU, Parkes AS (1949) Revival 12. Council of Europe (2013) Guide to the prepa-
of spermatozoa after vitrification and dehydra- ration, use and quality assurance of blood com-
tion at low temperatures. Nature 164:666 ponents, 17th edn. Council of Europe
5. Krijnen HW, de Wit JJ, Kuivenhoven AC, Loos Publishing, Strasbourg Cedex, France
JA, Prins HK (1964) Glycerol treated human 13. AABB (2011) Standards for blood banks and
red cells frozen with liquid nitrogen. Vox Sang transfusion services, 27th edn. AABB,
9:559572 Bethesda, MD
6. Rowe AW, Eyster E, Kellner A (1968) Liquid 14. Scott KL, Lecak J, Acker JP (2005)
nitrogen preservation of red blood cells for Biopreservation of red blood cells: past, present,
transfusion; a low glycerol-rapid freeze proce- and future. Transfus Med Rev 19:127142
dure. Cryobiology 5:119128 15. Hess JR (2004) Red cell freezing and its impact
7. Huggins CE (1966) Frozen blood: principles on the supply chain. Transfus Med 14:18
of practical preservation. Monogr Surg Sci 16. Crowley JP, Wade PH, Wish C, Valeri CR
3:133173 (1977) The purification of red cells for transfu-
8. Almond DV, Valeri CR (1967) The in vivo sion by freeze-preservation and washing.
effects of deglycerolized agglomerated erythro- V. Red cell recovery and residual leukocytes
cytes transfused in multiple units to stable ane- after freeze-preservation with high concentra-
mic patients. Transfusion 7:95104 tions of glycerol and washing in various sys-
9. Meryman HT, Hornblower M (1972) A tems. Transfusion 17:17
method for freezing and washing red blood cells 17. Chaplin H Jr (1982) The proper use of previ-
using a high glycerol concentration. Transfusion ously frozen red blood cells for transfusion.
12:145156 Blood 59:11181120
18. Valeri CR, Ragno G, Pivacek L, ONeill EM 24. Haemonetics ACP215 Automated Blood
(2001) In vivo survival of apheresis RBCs, fro- Glycerolization & Deglycerolization System -
zen with 40-percent (wt/vol) glycerol, deglyc- Operators Manual.
erolized in the ACP 215, and stored at 4 25. List J, Horvath M, Leitner GC, Weigel G
degrees C in AS-3 for up to 21 days. Transfusion (2012) Cryopreservation of red blood cell
41:928932 units with a modified method of glyceroliza-
19. Valeri CR, Ragno G, Pivacek LE, Srey R, Hess tion and deglycerolization with the ACP 215
JR, Lippert LE, Mettille F, Fahie R, ONeill device complies with American and European
EM, Szymanski IO (2001) A multicenter study requirements. Immunohematology 28:6773
of in vitro and in vivo values in human RBCs 26. Lagerberg JW, Paeper VW, Hagen WK, de
frozen with 40-percent (wt/vol) glycerol and Korte D (2011) Omitting glycerol superna-
stored after deglycerolization for 15 days at 4 tant reduction before freezing increases the
degrees C in AS-3: assessment of RBC process- RBC stability after thawing. Vox Sang 101:
ing in the ACP 215. Transfusion 41:933939 164165
20. Lagerberg JW, Truijens-de Lange R, de Korte 27. Hogman CF, de Verdier CH, Ericson A,
D, Verhoeven AJ (2007) Altered processing of Hedlund K, Sandhagen B (1985) Studies on
thawed red cells to improve the in vitro quality the mechanism of human red cell loss of viabil-
during postthaw storage at 4 degrees C. ity during storage at +4 degrees C in vitro.
Transfusion 47:22422249 I. Cell shape and total adenylate concentration
21. Lelkens CC, Noorman F, Koning JG, as determinant factors for posttransfusion sur-
Truijens-de Lange R, Stekkinger PS, Bakker vival. Vox Sang 48:257268
JC, Lagerberg JW, Brand A, Verhoeven AJ 28. Heaton WA (1992) Evaluation of posttransfu-
(2003) Stability after thawing of RBCs frozen sion recovery and survival of transfused red
with the high- and low-glycerol method. cells. Transfus Med Rev 6:153169
Transfusion 43:157164 29. Bohonek M, Petras M, Turek I, Urbanov J,
22. Pietersz RN, de Korte D, Reesink HW, Dekker Hrdek T, Chmtal P, Staroprazsk V, Kostrov
WJ, van den Ende A, Loos JA (1989) Storage J, Horcickov D, Duchkov S, Svobodov J,
of whole blood for up to 24 hours at ambient Tejckov E (2010) Quality evaluation of frozen
temperature prior to component preparation. apheresis red blood cell storage with 21-day
Vox Sang 56:145150 postthaw storage in additive solution 3 and saline-
23. Carlsen A, Wieth JO (1976) Glycerol transport adenine-glucose-mannitol: biochemical and
in human red cells. Acta Physiol Scand chromium-51 recovery measures. Transfusion
97:501513 50:10071013
Chapter 18
Cord Blood Clinical Processing, Cryopreservation,

and Storage
Heidi Elmoazzen and Jelena L. Holovati
Abstract
Allogeneic umbilical cord blood (UCB) hematopoietic stem cell transplantation has become a crucial
advancement in the treatment for a variety of diseases including hematopoietic and non-hematopoietic
malignancies, BM failure syndromes, hemoglobinopathies, and metabolic and immunodeficiency disor-
ders. It has been well documented that the success of UCB engraftment is tied to UCB banking processes,
and now there are established guidelines for standardization of collection, banking, processing, and cryo-
preservation for unrelated UCB units with purpose of achieving consistent production of high quality
placental and UCB units for administration. In 2011, Canadas Ministry of Health has announced Canadas
first national, publicly funded umbilical cord blood bank, which aims to provide altruistic donations for
unrelated allogeneic hematopoietic stem cell transplant. In this chapter, we describe specific protocols for
clinical processing, cryopreservation, and storage of UCB used by the Canadian Blood Services National
Public Umbilical Cord Blood Bank.
Key words Umbilical cord blood (UCB), Hematopoietic stem cells (HSC), Transplant,
Cryopreservation, Processing, Cord blood bank
1 Introduction
The first successful umbilical cord blood (UCB) transplantation

was reported in a 5-year-old boy with Fanconis anemia in 1988
[1]. This patient had a healthy HLA-identical sister who was shown
by prenatal testing to be unaffected by that specific hematological
disorder. Therefore, her UCB was collected at birth, cryopreserved
and, after thawing, used for transplantation [1]. The patient
showed complete hematological reconstitution and donor chime-
rism, with no graft-versus-host disease [1].
This success story initiated three decades of major advances in
the field of allogeneic UCB hematopoietic stem cell transplanta-
tion that led to improved treatments and patients outcomes. Over
the last 25 years, more than 600,000 UCB units have been banked
for transplantation worldwide, and >30,000 UCB transplants have
369
370 Heidi Elmoazzen and Jelena L. Holovati
been performed worldwide for a variety of diseases including

hematopoietic and non-hematopoietic malignancies, BM failure
syndromes, hemoglobinopathies, and metabolic and immunodefi-
ciency disorders [26]. There are a number of advantages in using
unrelated allogeneic UCB for hematopoietic stem cell transplanta-
tion over more traditional approaches, using bone marrow stem
cells and peripheral blood stem cells, including: less stringent
requirements for HLA matching of donor and recipient; lower
incidence of Graft vs. Host Disease, the relative ease of procure-
ment, the absence of risk for mothers and the donor baby, the
reduced likelihood of transmitting infections, and the urgent access
to HLA-typed and virally tested UCB units in the frozen state,
available for immediate use [27]. In addition, compared to adult
cells, UBC hematopoietic stem cells produce larger hematopoietic
colonies in vitro, have longer telomeres, produce fewer cytokines
and engraft SCID-human mice in the absence of additional human
growth factors [711]. Recent studies have confirmed that patient
survival with umbilical cord blood transplants is comparable to
that with related or unrelated bone marrow transplants [1215].
However, challenges remain, the main one being is the low stem
cell yields in UBC units, resulting in higher rates of graft failure,
poorer immune cell reconstitution, as well as delayed time to
engraftment compared to unrelated adult hematopoietic stem cell
transplants [2, 3, 1517].
Regardless of challenges, the increasing success of reported
UCB transplants resulted in worldwide establishment of cryopre-
served cord blood banks for clinical use. In 2011, Canadas
Ministry of Health has announced Canadas first national, publicly
funded umbilical cord blood bank, which aims to provide altruistic
donations for unrelated allogeneic hematopoietic stem cell trans-
plant. A national public solution is especially important for
Canadian patients because of the countrys ethnically diverse pop-
ulation. Only 25 % of patients who require a stem cell transplant
will find a match within their own family, so majority of Canadian
patients must look outside of their families for a match. The best
chance of finding the stem cell match is within their own ethnic
group, which may be difficult with multiple ethnic, Aboriginal and
racial minorities that may be unique to Canada, which are generally
underrepresented in stem cell transplant registries. To improve the
chance of finding high-quality cord blood for Canadian patients,
the Canadian Blood Services (CBS) national public cord blood
became operational in September 2013 in Ottawa. Healthy preg-
nant women, 18 years of age or older, reaching 34 weeks or later
in their pregnancy, are able to donate UCB with their signed con-
sent in four major Canadian cities. UCB donation process also
includes a medical questionnaire and infectious disease testing of
maternal blood. At the delivery, the UCB can be collected in one
of two ways: in utero collection, in which a hospital physician or
Cryopreservation and Processing of Cord Blood 371
licensed midwife collects the cord blood after the baby is delivered,
but before the placenta is delivered; and ex utero collection, where
designated CBS personnel collect the cord blood after the baby
and placenta are delivered. Following the collection, the UCB is
sent to a CBS manufacturing facility for further processing. It has
been well documented that the success of UCB engraftment is tied
to UCB banking processes, and now there are established guide-
lines for standardization of collection, banking, processing, and
cryopreservation for unrelated UCB units with purpose of achiev-
ing consistent production of high quality placental and UCB units
for administration. In this chapter, we describe protocols for clini-
cal processing, cryopreservation, and storage of UCB used by the
Canadian Blood Services National Public Umbilical Cord Blood
Bank.
2 Materials
2.1 Cord Blood 1. Syringes (3, 5, 10, 30, 60 mL).

Processing and 2. 16 G 1 inch needle, alcohol swabs.
Cryopreservation
3. Balance.
Materials and
Equipment 4. Biological safety cabinet.
5. Cryovials.
6. Me2SO 55 % w/v.
7. Dextran-40 5 % w/v.
8. EDTA microtainers.
9. Hematology analyzer.
10. Heparin vacutainer.
11. Pentaspan.
12. Tube sealer.
13. Sampling Site Coupler.
14. Sepax 2, SepaxNet and single-use Sepax Separation Kit.
15. Sterile docking device.
16. Sterile gloves.
17. Timer.
18. Tube stripper.
19. Barcode scanner.
20. Becton Dickinson 10 cc glass syringe.
21. BioArchive Freezer including CRF module.
22. BioArchive canister.
23. Calibrated clock.
24. Canister opening tool.
25. Computer workstation.

26. Coolmix AS-210.
27. Controlled-rate freezer (CRF).
28. Me2SO 55 % w/v/Dextran-40 5 % w/v single use syringe.
29. MVE LN2 storage tank.
30. Overwrap bag.
31. Overwrap sealer.
32. Syringe pump.
33. Tube sealer.
34. Sealing jig.
3 Methods
3.1 Cord Blood UCB units must be processed within 48 h of collection

Processing Procedure (see Note 1).
1. Once UCB unit is received, initiate procedure by documenting
equipment and supplies. Obtain and inspect supplies for steril-
ity by checking for intact packaging and change in expected
color/turbidity. Record suppliers, lot numbers, and expiry and
equipment calibration dates, as applicable.
2. Switch on the Sepax 2 Instrument. The following settings are
important for cord blood processing:
(a) Red blood cell (RBC) extraction (so RBCs will be returned
to the input bag at the end of the procedure).
(b) Pentaspan % is set to 20 %.
(c) Final buffy coat (BC) volume is set to 22.0 mL.
3. Weigh cord blood unit and calculate UCB volume the volume
(see Note 2). Calculate and record the Pentaspan solution vol-
ume required corresponding to 20 % of the UCB input volume
(UCB + anticoagulant volume + 1 mL corresponding to the
dead volume between the connector and the filter).
4. Prepare Single-use Sepax Separation kit in the biological safety
cabinet. Ensure the kit sterility indicator is showing the
appropriate result. Ensure all kit components are present,
seals are intact and the unit is not discolored. Label the sam-
pling accessory components and plasma bag with the UCB
number. Remove the canister from the plastic bag and label
with the UCB number. Label the buffy coat bag with the
UCB number.
5. Connect the UCB bag to the Single-use Sepax Separation kit
with a sterile docking device (see Note 3). Do not remove the
cap of the spike.
6. In biological safety cabinet, remove the cap on Pentaspan and fill

syringe with calculated Pentaspan volume. Connect the syringe
filled with Pentaspan to the filter on the Single-use Sepax
Separation kit, open the Pentaspan clamp and inject the Pentaspan
solution slowly with one hand, while mixing the UCB bag thor-
oughly with the other (see Note 4). Seal off the Pentaspan filter
tube using tube sealer. Incubate the Pentaspan with the UCB
10 min at room temperature (1525 C).
7. Install the Single-use Sepax Separation kit into the Sepax
instrument separation chamber. Verify that the stopcocks are
aligned in the T-position. Open the stopcock holder by push-
ing down the two levers, install the stopcocks on the Sepax 2
rotary drive pins, and close the holder by pushing the levers
up (see Note 5). Insert the separation chamber tubing line
into the optical sensor area. Close the centrifuge cover and
connect the pressure sensor line to the pressure sensor port on
top of the Sepax.
8. Hang the plasma bag on the hook provided on the right of
Sepax. Hang the buffy coat bag on the hook provided on the
left of Sepax. Place the bubble chamber in its support and
hang the UCB on the hook provided. Adjust the height of
the UCB input bag above the Sepax processing unit to
ensure the flow of cord blood. Start the automated proce-
dure (see Note 6).
9. Remove bags and air filter at the end of the automated proce-
dure (see Note 7). Using tube stripper, strip the RBC line from
the blue stopcock to the input bag while holding the line in
one hand. Manually remove the air from the BC bag by squeez-
ing the 5 mL compartment and filling it with buffy coat (BC).
Remove all of the air from the 20 mL compartment, close the
clamps, and dismount the kit. Detach the RBC line and the
separation chamber. Retain RBC bag and plasma bag for
microbial culture testing and reference samples.
10. Obtain the BC post-processing sample by manually mixing the
BC bag, opening the clamp on the sampling chamber line,
positioning the BC bag upside down, and releasing pressure
on the sampling chamber to obtain approximately 2.0 mL in
the sampling chamber. Seal the clamp on the sampling cham-
ber line and place BC bag in a refrigerator (28 C) to cool for
minimum of 15 min.
11. Label post-processing UCB sample containers with UCB
identification number, including, cryovials for UCB plasma
storage as reference samples, cryovials for UCB buffy coat stor-
age as reference samples, 2 EDTA microtainers for flow
cytometry assay and one Heparin vacutainer for colony-
forming unit (CFU) assay.
12. In a biologic safety cabinet, inoculate post-processing UCB

samples, including:
(a) Luer lock a 10 mL syringe and acquire approximately
3.6 mL of plasma. Aliquot 1.8 mL of plasma into each of
the two labeled cryogenic vials and close caps.
(b) Connect a 3 mL Luer-lock syringe to sampling chamber
and withdraw post-processing UCB sample. Aliquot
approximately 0.25 mL of UCB into each of the two
labeled cryogenic vials and close caps.
(c) Aliquot 0.5 mL of UCB into each of 2 EDTA microtainers
for flow cytometry (total nucleated cell count, CD34
count and viability assay).
(d) Aliquot 0.5 mL of UCB into a heparin vacutainer for
colony-forming unit assay.
(e) Inoculate microbial culture bottles, according to the
manufacturers instructions.
13. Forward microbial culture bottles, EDTA and heparin post-
processing UCB samples for quality control testing. Store
plasma and BC post-processing UCB reference samples in
LN2 freezer. Determine post-processing total nucleated cell
count (TNC, see Note 8).
3.2 Cord Blood 1. Initiate procedure by documenting equipment and supplies.

Cryopreservation Obtain and inspect supplies for sterility by checking for
Procedure intact packaging and change in expected color/turbidity.
Record suppliers, lot numbers, and expiry and equipment
calibration dates, as applicable. Place all supplies to be used
into the biological safety cabinet as follows: Me2SO Syringe,
Me2SO extension line, 10 mL syringe, Sealing Jig, Overwrap
Bag, Canister.
2. Switch on the Coolmix AS-210 device with the main switch
on the rear panel of the instrument. Record the calibration
due date of Coolmix AS-210. Leave the Coolmix AS-210
cover closed at all times; open it only to insert or remove
the BC bag.
3. Obtain Me2SO 55 % (w/v)/Dextran-40 5 % (w/v) syringe
and visually check to ensure contents are clear and colorless.
Obtain Me2SO extension line and cord blood buffy coat bag
obtained from Sepax processing. In the biological safety cabi-
net attach the Me2SO 55 % w/v/Dextran-40 5 % w/v syringe
to the extension line by attaching clear cap on Me2SO syringe
to white cap on Me2SO extension line. Prime the Me2SO line
until the Me2SO solution has just passed through the filter
(see Note 3).
4. Load Me2SO syringe on the syringe pump (see Note 9). Ensure
the following settings are entered into the syringe pump and
initiate syringe pump start-up:
(a) Syringe diameter in mm (Becton Dickinson glass syringe
10 cc = 14.34 mm).
(b) Volume setting of 5.00 mL.
(c) Rate setting of 0.333 mL/min (5 mL/15 min).
(d) Units setting of mL min1.
5. Place the buffy coat bag in Coolmix AS-210 cooler platform
and close the separator arm (see Note 3). At the same time,
initiate the Coolmix AS-210 and the syringe pump operation.
Press start on cryopreservation timer (30 min). When the
alarm sounds at the end of 30 min, at the same time press stop
on the Coolmix AS-210 and syringe pump. Remove buffy coat
bag from Coolmix AS-210 and syringe from syringe pump.
6. Make UCB segments in the biological safety cabinet. Close blue
clips on Me2SO extension line and filter line and detach Me2SO
extension line from Me2SO filter by unscrewing it. Obtain a
10 mL syringe, take approximately 2 mL air into syringe then
attach syringe to air filter. Open white clip and push air from
syringe into buffy coat bag. Mix cells by pressing gently on small
compartment of buffy coat bag 510 times. Continue to express
the air from the 20 mL compartment until segment line fills
with cells. Close white clamp and place buffy coat bag in sealing
jig. With tube sealer, seal bottom of buffy coat bag, making
three segments. Label segments with UCB label.
7. Initiate the control rate freezing process by loading the UCB
into an overwrap bag (see Note 3). Align aliquot tubing seg-
ments along top of ports and in between large and small com-
partments. Fold filled freezing bag at channels between large
and small compartments. Grasping open end of overwrap bag,
shake overwrap in a downward motion to force freezing bag
down against bottom seam. Seal overwrap bag using the over-
wrap sealer.
8. Place Overwrapped Buffy Coat bag into the BioArchive can-
ister. Hold BioArchive canister so writing on canister and
temperature monitoring opening are facing up. Using Canister
Opening Tool, insert the BC bag into the canister, aligning the
bottom of the BC bag with the bottom of the canister and the
small compartment next to the front receptacle. Tuck left
upper corner of overwrap bag underneath buffy coat bag and
close the canister (see Note 10).
9. Insert the canister with the front receptacle up in the open slot
at the forward side of the doors of the controlled rate
freezer. Select the port to be used and remove the port plug.
Gently insert and seat the CRF fully into the port with the
exposed side of the canister facing the periscope shaft. Select
Cord Blood Freeze profile, with the following parameters:
(a) Pre-freeze rate 10 C/min.
(b) Freeze start temperature 3 C.
(c) Freeze power 100 %.
(d) Freeze exit temperature 12 C.
(e) Post-freeze rate 2 C/min.
(f) Target temperature 50 C.
10. Record total cryopreservation time for duration of cell exposure
prior to freezing and actual time on clock for start of cryo-
preservation (see Note 11). Obtain freeze graph upon
completion of control rate freeze and review for acceptability,
according to Fig. 1, and the following instructions:
Freezing Record
Sample: X60001200007483
Storage Date: 10/19/2012 10:37:51 AM
30
25
20
15
Temperature (Celsius)
10
5
0
5
10
15
20
25
30
35
40
45
50
55
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Time (Minutes)
Operator ID: ROUELLET Location: Rack: 1 Slot: 10

Ring: 7 Angle: 12.00
Profile Used: OMPCBB
Port Number: 2
Pre-Freeze: 10C/Min. End Freeze: 12C CRF Serial Number: 1781

Start Freeze: 3C Post Freeze: 2C/Min. CRF Version: 1.10
Fan Power: 100% End Temp: 50C
Fig. 1 Example of an UBC unit controlled-rate cryopreservation graph

(a) The time between 15 and 0 C should be read from the

graph. This time should be 1.5 min (for a programmed rate
of 10 C/min), with a range of 13 min (515 C/min).
(b) There should be a nonlinearity (a change of slope) between
5 and 15 C, but not at lower temperatures. This indi-
cates that the ice has been nucleated in the sample during
the freeze region.
11. The time between 20 and 40 C should be read from the graph.
This time should be 10 min (for 2 C/min cooling rate) with a
range between 6.8 min (3 C/min) and 20 min (1 C/min).
This is important for viability of stem cells in this sample.
12. Using Cord Blood Sample Inventory Record, determine, and
assign storage locations for each UCB reference sample (cord
blood plasma, cord blood nucleated cells, maternal plasma and
maternal nucleated cells) based upon currently available loca-
tions. Once frozen, UCB samples and units can be stored
indefinitely.
4 Notes
1. Please refer to relevant national and international standards

and guidelines for hematopoietic stem cell transplantation,
such as FACT International standards for cord blood collec-
tion, processing, testing, banking, selection, and release.
2. Volume determinations of cord blood unit (CBU) can be
obtained by weighing the cord blood and using the following
calculation:
Volume (mL) = Net weight of CBU (g)/CBU Specific
Gravity 1.06 (g/mL).
3. Please refer to appropriate manufacturer manuals, instruc-
tions, and guidelines.
4. Connect the CBU bag to the Single-use Sepax Separation kit
with a sterile docking device at a location of approximately
15 cm from the UCB bag and approximately 5 cm from the
spike on the Single-use Sepax Separation kit.
5. When installing the Single-use Sepax Separation kit into the
Sepax instrument separation chamber, do not close the
separation chamber pit covers until the kit has been positioned
in the separation chamber pit and the tubing has been placed
in the optical sensor. The tubing connector on top of the sepa-
ration chamber should not be rotated once the pit covers have
been closed to avoid any risk of leakage.
6. Avoid touching the Sepax and the Single-use Sepax Separation
kit during the automated procedure. Moving bags, tubes,
stopcocks and covers may cause errors requiring the process to
be restarted.
7. At the end of automated Sepax procedure, detach the RBC

line and the plasma bag by placing three seals 2 cm from the
stopcock manifold and cutting the middle seal. Detach the
separation chamber by placing three seals 2 cm from the Y
connection and cutting the middle seal.
8. Total nucleated cell count (TNC) in a CBU can be calculated
according to the following formula: Volume for post-
processing CBU sample will be 20 mL.
TNC = [(nucleated red blood cell count + white blood cell
count) (109/L) post-processing CBU volume (mL)]/1,000.
9. Total time allowed between start of Me2SO addition to start
of controlled-rate freeze is 30 min.
10. It is important to verify that the controlled-rate freezer is com-
pletely dry prior to placing UCB canister inside. Control-rate
freezer should sit in the drying box located on the control
system electronics box a minimum of 15 min before reuse. It
is also recommended that the CRF be at room temperature
prior to inserting canister.
11. Total cryopreservation time for duration of cell exposure prior
to freezing and actual time on clock for start of cryopreserva-
tion should not be greater than 30 min. Total time from
collection of UCB to cryopreservation completion should not
be greater than 48 h.
Disclaimer
This publication is provided without any representations or warran-

ties of any kind, whether express or implied, regarding the complete-
ness, accuracy, reliability, or suitability of the information contained
herein or its fitness for a particular purpose or use. None of Canadian
Blood Services, the authors or other persons involved in the creation
of this manuscript shall be responsible for the consequences of any
action taken on the basis of the information contained in this publi-
cation, or any errors or omissions therefrom. There are no conflicts
of interest from the authors of this publication.
References
1. Gluckman E, Broxmeyer HE, Auerbach AD, 2. Ballen KK, Gluckman E, Broxmeyer HE

Friedman HS, Douglas GW, Devergie A, (2013) Umbilical cord blood transplantation:
Esperou H, Thierry D, Socie G, Lehn P, Cooper the first 25 years and beyond. Blood
S, English D, Kurtzberg J, Bard J, Boyse EA 122:491498
(1989) Hematopoietic reconstitution in a 3. Gratwohl A, Baldomero H (2009) Trends of
patient with Fanconis anemia by means of hematopoietic stem cell transplantation in the
umbilical-cord blood from an HLA-identical third millennium. Curr Opin Hematol 16:
sibling. N Engl J Med 321:11741178 420426
4. Ljungman P, Bregni M, Brune M, Cornelissen 12. Chen YH, Xu LP, Liu DH, Chen H, Zhang XH,
J, de Witte T, Dini G, Einsele H, Gaspar HB, Han W, Wang FR, Wang JZ, Wang Y, Huang XJ,
Gratwohl A, Passweg J, Peters C, Rocha V, Liu KY (2013) Comparative outcomes between
Saccardi R, Schouten H, Sureda A, Tichelli A, cord blood transplantation and bone marrow or
Velardi A, Niederwieser D, European Group peripheral blood stem cell transplantation from
for Blood and Marrow Transplantation unrelated donors in patients with hematologic
(2010) Allogeneic and autologous transplan- malignancies: a single-institute analysis. Chin
tation for haematological diseases, solid Med J 126:24992503
tumours and immune disorders: current prac- 13. Locatelli F, Kabbara N, Ruggeri A,
tice in Europe 2009. Bone Marrow Transplant Ghavamzadeh A, Roberts I, Li CK, Bernaudin
45:219234 F, Vermylen C, Dalle JH, Stein J, Wynn R,
5. Wagner JE (2010) Umbilical cord transplan- Cordonnier C, Pinto F, Angelucci E, Soci G,
tation: state of the art 2010. Semin Hematol Gluckman E, Walters MC, Rocha V (2013)
47:12 Outcome of patients with hemoglobinopathies
6. Worth AJJ, Booth C, Veys P (2013) Stem cell given either cord blood or bone marrow trans-
transplantation for primary immune deficiency. plantation from an HLA-identical sibling.
Curr Opin Hematol 20:501508 Blood 122:10721078
7. Cohen Y, Nagler A (2004) Umbilical cord 14. Zhang H, Chen J, Que W (2012) A meta-
blood transplantation how, when and for analysis of unrelated donor umbilical cord
whom? Blood Rev 18:167179 blood transplantation versus unrelated donor
8. Emerson SG (1996) Ex vivo expansion of bone marrow transplantation in acute leuke-
hematopoietic precursors, progenitors, and mia patients. Biol Blood Marrow Transplant
stem cells: the next generation of cellular ther- 18:11641173
apeutics. Blood 87:30823088 15. Metheny L, Caimi P, de Lima M (2013) Cord
9. Koegler G, Radke TF, Lefort A, Sensken S, blood transplantation: can we make it better?
Fischer J, Sorg RV, Wernet P (2003) Cytokine Front Oncol 3(238):111
production and hematopoiesis supporting 16. Flores-Guzmn P, Fernndez-Snchez V,
activity of cord bloodderived unrestricted Mayani H (2013) Concise review: ex vivo
somatic stem cells. Exp Hematol 33: expansion of cord blood-derived hematopoi-
573583 etic stem and progenitor cells: basic principles,
10. Kimura M, Gazitt Y, Cao X, Zhao X, Lansdorp experimental approaches, and impact in regen-
PM, Aviv A (2010) Synchrony of telomere erative medicine. Stem Cells Trans Med
length among hematopoietic cells. Exp 2:830838
Hematol 38:854859 17. Wagner JE, Barker JN, DeFor TE, Baker KS,
11. Piacibello W, Sanavio F, Severino A, Dan A, Blazar BR, Eide C, Goldman A, Kersey J,
Gammaitoni L, Fagioli F, Perissinotto E, Krivit W, MacMillan ML, Orchard PJ, Peters
Cavalloni G, Kollet O, Lapidot T, Aglietta M C, Weisdorf DJ, Ramsay NK, Davies SM
(1999) Engraftment in nonobese diabetic (2002) Transplantation of unrelated donor
severe combined immunodeficient mice of umbilical cord blood in 102 patients with
human CD34(+) cord blood cells after ex vivo malignant and non-malignant diseases: influ-
expansion: evidence for the amplification and ence of CD34 cell dose and HLA disparity on
self-renewal of repopulating stem cells. Blood treatment-related mortality and survival.
93:37363749 Blood 100:16111618
Chapter 19
Directional Freezing for Large Volume Cryopreservation

Joseph Saragusty
Abstract
Cryopreservation is currently the method of choice when it comes to long-term preservation of viable
biological samples. The process, and consequently the volume of the sample, however, is limited by the
ability to achieve homogenous and efficient heat removal. When this cannot be properly managed, ice
crystals will grow uncontrollably resulting in extensive damage to the cryopreserved cells or tissues.
Directional freezing is a technique that can be used to precisely control heat dissipation and ice crystal
growth and morphology even when freezing large volumes. The technique has been used over the years to
cryopreserve spermatozoa, oocytes, embryos, tissue slices and whole organs from a wide variety of domes-
tic and wild species. In this chapter a protocol for directional freezing of spermatozoa is described and its
benefits and shortcomings are discussed.
Key words Cryoinjury, Directional freezing, Gamete cryopreservation, Semen
1 Introduction
When cryopreservation of viable cells is concerned, regardless of

whether they are in suspension or being part of an assembly of
cells, it is primarily the water in the system that needs to be dealt
with and the better it is handled the better the outcome. At the
most simplistic level, living cells can be described as lipid and pro-
tein membranes encompassing relatively large volume of water
with organelles and solutes in them. These membranes, while act-
ing as a barrier between the intracellular and extracellular environ-
ments, are largely permeable to water. Most of the water within the
cells is non-bound and osmotically active so it will move across the
cellular membrane in the direction of an osmotic gradient. Only
about 10 % of the water in cells is hydrogen-bound to proteins and
lipids and is therefore incapable of freezing or participating in the
osmotic activity [1]. During the process of cell suspension cryo-
preservation, the system is cooled to subzero temperatures. Because
ice formation is a stochastic process that is dependent on the tem-
perature and presence of ice nucleation sites, the probability of ice
381
382 Joseph Saragusty
formation is higher outside the cells. Once ice crystals start growing,
and as long as the temperature stays under their melting tempera-
ture, extracellular ice will continue to form, thus removing water
from the system. The osmolarity of the remaining solution will
increase, causing more water to move out of the cells. In the
absence of alternatives to substitute the lost water, the cells will
shrink and will eventually get damaged and die. Among their many
other functions, that include cell dehydration and stabilization of
cellular membranes, permeating cryoprotectants (CPs) enter the
cells to replace the lost water and maintain acceptable volume that
will permit cell survival.
Equiaxed solidification, the most commonly used freezing
technique in cryobiology, involves taking a sample and cooling it
gradually to a given temperature below the solutions equilibrium
freezing point. At this temperature, ice nucleation occurs primarily
in the extracellular compartment. From these seeding points, ice
dendrites grow rapidly towards the supercooled solution, releasing
the latent heat of fusion in the process. This heat is removed by
conduction through the frozen area down the thermal gradient,
resulting in an abrupt increase in its temperature towards the melt-
ing point. At this stage the ice front bordering the solution may go
through repeated cycles of freezing and thawing, resulting in
recrystallisation and possible damage to the cells. As cooling pro-
ceeds, ice crystals will grow from the colder periphery towards the
still liquid center in an uncontrolled way in terms of velocity and
morphology. The released heat from the process is conducted from
the unfrozen solution through the surrounding ice to the colder
exterior [2], resulting again in possible melting and refreezing at
the ice-liquid interface. This process and the morphology of the ice
crystals are cooling velocity-dependent.
In directional solidification or directional freezing, after the
initial seeding stage, samples are advanced at a constant velocity
through a linear temperature gradient (Fig. 1). This allows precise
control over the freezing process, thus facilitating accurate and
homogenous cooling of the samples. The constant sample propa-
gation also results in continuous seeding, thus controlling ice crys-
tal growth in a direction opposite to the direction in which the
sample is moved, down the temperature gradient. This directional
growth of ice crystals produces lamellae between which the cells in
the sample are trapped. Dissipation of heat from the sample is pri-
marily achieved in two ways. The highly heat-conductive metal
block snugly surrounding the sample efficiently removes heat away
from the sample during freezing. Because of the temperature gra-
dient and the directional movement of the sample, heat also con-
tinuously moves away from the ice front into the unfrozen fraction
of the sample that is still located at the higher temperature end of
the gradient. This controlled and directional heat dissipation pro-
tects the ice front from melting, thus avoiding damages to the cells.
Directional Freezing 383
Fig. 1 Schematic diagram of directional freezing. Samples are propagated from

the warm block to the cold block at a constant velocity (V). Gap (d) between the
blocks and the difference between the warm block temperature (TH) and the cold
block temperature (TC) result in temperature gradient (G). Heat is dissipated
(small arrows) from the sample towards the warm block and the unfrozen frac-
tion of the sample, and to the cold block that snugly surrounds the sample and is
made of highly heat-conductive metal (color figure online)
All these processes result in highly controlled and cell-friendly ice

crystal morphology that contributes to reducing mechanical
damages [3, 4].
To put the directional freezing process in mathematical terms,
it can be described by the following two equations:
G = (| TC | +TH ) / d (1)
DT / Dt = G V = B ( 2)
where G is the temperature gradient, TC is the cold base tempera-

ture in absolute numbers, TH is the hot base temperature, d is the
gap distance between the hot and cold bases in mm, T is the
change in temperature and t is the change in time, so that T/
t is cooling rate (B), and V is the velocity at which the sample
moves through the temperature gradient.
The efficient heat removal and controlled ice crystal propaga-
tion makes it possible to use this technology to freeze small and
large volumes alike (for recent reviews see [5, 6]). For example, we
presently freeze sperm suspension in glass tubes containing 8 mL
[79]. Similarly, large and complex cellular systems such as whole
ovaries [10, 11], livers [12], and hearts [13] can also be successfully
cryopreserved. The advantages of large volume freezing are further
apparent when one has to consider storage space and costs of a large
number of samples in liquid nitrogen over extended periods of time
[7, 14]. This is of course in addition to the costs of straws or vials
and the inconvenience involved in handling, identifying, and using
these, especially when large volumes are needed for insemination.
Yet, the directional freezing technique is also suitable for small

volume freezing and even vitrification in submicroliter drop size.
Actually, in its early days of use for cell cryopreservation, directional
freezing was employed to vitrify oocytes and embryos [1517].
A detailed protocol is described for directional freezing of
spermatozoa. While differences between species are vast, and each
species requires its optimal cryoprotective cocktail, as far as the
directional freezing protocol is concerned, we found that one basic
protocol works well for all species tested thus far. In this respect the
directional freezing technique contributes an important part to
the ultimate goal of achieving a universal cryopreservation proto-
col for spermatozoa of all species.
2 Materials
Use double distilled water and reagent grade or higher materials to

prepare all solutions used for handling the sperm (not described;
vary between species). Make sure to keep solutions and anything
else that might come in touch with the sperm (pipette tips, syringes,
long needles, glass tubes, etc.) at the same temperature as the sam-
ple at each specific stage of the process. Follow all waste (including
biological waste) disposal regulations. Diligently follow safety reg-
ulations relevant to work with liquid nitrogen as it may cause frost
burns as well as suffocation.
2.1 Semen Collection Anything that will come in contact with the semen directly or indi-
and Handling rectly should be warmed to or near the animals body temperature.
1. Microscope with heated stage (normally at 37 C), ideally a
fluorescence microscope (optional, if one or more of the evalu-
ation methods use fluorescence dyes).
2. Heated plate (normally at 37 C) to warm slides, coverslips,
and pipette tips.
3. Centrifuge with swinging buckets and deceleration-free option
(optional, only for species in which removal of seminal plasma
and/or transport extender are required). If centrifugation is
used, the following are also recommended:
(a) OptiPrep (60 % iodixanol).
(b) Syringes with needles long enough to go all the way to the
bottom of the centrifugation tube.
4. Computer-Assisted Sperm Analysis (CASA) system (e.g.,
AndroVision or SpermVision from Minitube, IVOS II or
CEROS II from Hamilton Thorne, Sperm Class Analyzer
from Microptic, etc.)ideal but only optional.
5. Water bath, heated to the desired temperature (usually 37 C).
6. Microscope slides with and without frosted end.
7. Coverslips, best the square ones, ~1820 mm in size.

8. 100 and 1,000 L pipettes and suitable tips.
9. Graduated containers, for semen collection and volume
determination.
10. Hemocytometer (or other cell counting device) for concentra-
tion estimation (optional, if CASA is not used to estimate
concentration).
11. Counter (handheld or benchtop) to count the cells (optional, if
CASA is not used to estimate concentration or if needed for mor-
phology, viability, acrosome integrity, or other assessments).
12. 1.5-mL Eppendorf tubes.
13. Stains and fixatives for morphology (e.g., Hancocks fixative),
acrosome integrity (e.g., FITC-PNA or FITC-PSA), viability
and membrane integrity (e.g., hypoosmotic swelling test, SYBR-
14/PI or eosin/nigrosin), mitochondrial membrane potential
(e.g., JC-1), DNA fragmentation (e.g., comet assay, acridine
orange staining), and/or any other evaluation as needed.
14. Suitable extenderdepending on the species, sperm sensitiv-
ity, and need for seminal plasma removal. Extenders can be
homemade or, in some cases commercially available. More
details in the following section.
15. Graduated disposable plastic tubes (15 mL, 50 mL) for semen
collection, dilution, chilling, and, if needed, centrifugation.
Tubes with conical end are best for centrifugation.
16. Suitable laboratory stands for 15 mL and/or 50 mL tubes
(species-dependent) as well as a stand for Eppendorf tubes.
17. Containers for disposable waste, including one for glass and
one for needles waste.
18. Waterproof marker.
19. Pencil (and pencil sharpener, just in case).
2.2 Sperm Chilling All parts that will come directly or indirectly in contact with the
and Freezing sperm need to be chilled alongside the sample so that everything
will be at the same temperature when used.
1. Refrigerator (4 C), if possiblea walk-in refrigerator is ideal.
2. Directional freezing machine (see Fig. 2)models such as MTG
516, MTG 550, MTG 1314 are all suitable (Core Dynamics,
Ness Ziona, Israel and Harmony CryoCare, Chester, UK).
Other models may also be available, including models suitable
for straw freezing. If models MTG 516 or MTG 550 are used
then HollowTubes-carrying cassettes (Fig. 2a) are needed
and should be chilled along with the sperm samples. Depending
on the model, other parts may also be needed including rods
Fig. 2 Directional freezing machine and parts. The directional freezing machine is seen in the main picture. The
following parts are seen in the inset: (A) Cassette to hold the tubes. (B) U-shaped holders for part C. (C) Rod to
push the sample through the holes in part A and the directional freezing machine. The right dark cover on the
freezing machine is where the cassettes with the HollowTubes are loaded, and the left is where the collection
chamber is located. Once the freezing process is completed, the frozen samples are collected from here and
transferred to liquid nitrogen for storage (color figure online)
(see Fig. 2c) to push the sample and the U-shaped metal parts
to hold these rods in place (see Fig. 2b).
3. Liquid nitrogen tank for the freezing machine. Models vary
depending on the freezing machine model used.
4. Liquid nitrogen storage tank to store the frozen samples.
5. Small, benchtop container with liquid nitrogen into which the
frozen samples are collected immediately after freezing and
before they are transferred into the storage tank.
6. Accessories for liquid nitrogen handling: gloves, protection
glasses, and long tweezers.
7. Small plastic or glass container that can hold the 15-mL tubes
with the extended semen and some water during the slow
cooling process.
8. HollowTubes of 2.5 mL and/or 8 mL, according to the vol-
ume to be frozen, and their dedicated silicon stoppers (IMT
International, Chester, UK). Some directional freezing machine
models also support straw freezing so when these are used
freezing in plastic straws of 0.25 or 0.5 mm is also possible.
9. 10 mL sterile glass pipette with suitable sucking/expulsion

pump or 10 mL syringe with long 1216 G needle, to fill up
the HollowTubes.
10. Stand for the filled HollowTubes.
11. Labels for marking the stoppers.
2.3 Sample Thawing 1. Protection accessories for handling liquid nitrogen (glasses,
gloves).
2. Long tweezers to remove the sample(s) from the liquid
nitrogen.
3. Water bath and metal thawing device (see Fig. 3a, b; Harmony
Cryocare, Chester, UK).
4. Glass -shaped extension (see Fig. 3d).
5. Thermometer.
6. Timer.
7. Dedicated stand for the tube (see Fig. 3c). Alternatively, a flat
piece of Styrofoam such as the lid of a Styrofoam box can also
be used.
Fig. 3 Parts used for thawing in a water bath. (A) Heater and water expulsion
device. (B) Metal stand into which the water is propelled by device A to allow
continuous and efficient heat transfer. (C) Stand for the HollowTube during
warming at room temperature before it is inserted into part B for thawing. (D)
-shaped glass tube that is inserted into the central hole in the stopper of the
HollowTube. Water flowing through the central tube of the HottlowTube and
this -shaped glass tube cause the HollowTube to rotate, thus achieving
homogeneous and more efficient heat transfer
8. Warm 15-mL disposable tubes to put the thawed sample in

after thawing.
9. Everything needed for post-thawing evaluation, including
microscope with heated stage, a CASA system, warm plate,
slides and coverslips, pipettes and pipette tips, viability, mor-
phology, acrosome, and/or mitochondrial membrane potential
stains (see Subheading 2.1 above).
3 Methods
Perform all procedures at room temperature till the extended

semen is placed into the refrigerator. Handle chilled semen in 4 C
environment, ideally in a walk-in refrigerator. Note that through-
out the collection and evaluation procedures the semen sample
should be kept warm (ideally, at or near its temperature at ejacula-
tion) and, if possible, protected from light.
3.1 Semen Collection 1. Warm the collection tubes, extenders, microscope stage, warm
and Handling plate, slides, coverslips, and pipette tips. Maintain collection
tubes insulated during the collection process, especially when
collection is performed in cold environments. When electroe-
jaculation is used, but also when other stimulation techniques
(e.g., rectal or penile massage) are used, it is best to replace
collection tubes between stimulations to protect already col-
lected samples from possible urine contamination. Mark frac-
tion numbers on the collection tubes with a waterproof marker.
During evaluation, similar-quality samples can be combined
while those of quality too poor to be considered for freezing
can be discarded or used for other purposes. For some species
(e.g., pigs) it is beneficial to put measured volume of washing
extender in the collection container or to dilute the sample
immediately after collection to protect the spermatozoa and
extend their in vitro viability. Maintain the sample at or near its
initial temperature while doing the preliminary evaluation and
deciding if it is suitable for freezing.
2. After collection, the sample is evaluated for total collected vol-
ume. If extender was added to the collection tube, its volume
should be deducted from the total volume. Color and smell
should also be noted. Ideally, for most species, the color should
be milky off-white and the sample should be free of urine smell.
If possible, especially when in doubt, measure the pH to make
sure there is no urine contamination. However, keep in mind
that in some species urine is acidic while in others it is basic.
3. Motility evaluation: Dilute an aliquot of the sample in
extender free of cryoprotective agents (e.g., glycerol, Me2SO)
to a final concentration of about 50 106 spermatozoa per mL.
When CASA is not used, place a 10-L drop on a pre-warmed

microscope slide, cover with pre-warmed coverslip, and
examine under phase contrast microscope at 200 magnifi-
cation (100 magnification and/or using dark field micros-
copy is also possible but less ideal). Estimate motility in
several locations across the center of the slide. While estimat-
ing total motility, also take note of motility characteristics
(forward motility, hyper activation, stationary motility, joint
motility of two or more cells together) and get a general
impression of sperm morphologyrate of tail-less heads,
rate of cytoplasmic droplets, rate of coiled tails or bent tails
or broken tails, large/small heads and so on. Keep in mind
that our vision is designed to be attracted to movement so
when evaluating sperm motility, our eyes naturally note the
motile cells more than the immotile ones. So, when evaluat-
ing subjective motility, pay special attention to the non-
moving cells. If CASA is used, follow the manufacturers
instructions and examine several fields randomly to get a
representative estimate. Some CASA machines rotate auto-
matically between fields while in others this should be done
manually. Make sure the gating settings are properly adjusted
for the species in question so that all or at least the vast
majority of the cells are identified and as few non-cellular
particles are included. In some systems the wrongly marked
particles can be manually removed and wrongly classified
spermatozoa be reclassified after acquisition to make the
reading more accurate.
4. If motility justifies processing the sample for freezing, concen-
tration needs to be estimated to determine how much extender
to add. Concentration can be estimated by the CASA system,
by using a hemocytometer or by a dedicated automatic cell
counter. For hemocytometer counting, we normally dilute a
known volume of the sample in a known volume of tap water.
The aim is to achieve a concentration of approximately
50100 106 spermatozoa per mL after dilution. With experi-
ence one can estimate the dilution factor during motility evalu-
ation. Alternatively, dilution can be done based on the expected
concentration known for the species. After dilution, the sample
can be further diluted if still too concentrated (not ideal) or a
new aliquot can be diluted at a lower dilution factor if the ini-
tial one resulted in a too diluted sample or at a higher dilution
factor if the sample was too concentrated.
5. The sample can be further processed for any other tests (fixed
in fixative, incubated with fluorescence or other dyes, etc.) as
required. One should, however, remember that the sample
should not let standing for too long while proceeding with all
these tests.
6. If centrifugation is required, this can be done while concentration

is estimated. Centrifugation is done in disposable test tubes
with conical end, volume of which depends on the species and
ejaculated volume. For most species we worked with, the
15-mL tubes are good while for elephants and pigs we some-
times use 50-mL tubes. The sample is put in one or more pre-
warmed tubes. Then, using a syringe and a long needle,
underlay the sample with 12 mL of pre-warmed OptiPrep,
which acts during centrifugation as a cushion to protect the
cells from overt centrifugation damage [18]. The use of
OptiPrep allows exploiting somewhat higher centrifugation
force for better yield. When using OptiPrep we often use
1,000 g with good results. At the end of centrifugation, the
spermatozoa are found at the interface between the OptiPrep
at the bottom and the seminal fluids and washing or transport
extender at the top. Remove the top fraction and then slowly
insert a long needle connected to a syringe through the sperm
layer to the bottom of the tube and aspirate the OptiPrep.
Small residues of OptiPrep are not to worry about and can
actually be beneficial to the sperm during freezing [4, 19].
7. Dilution of the sample in extender: Based on the initial con-
centration and after deducting estimated loss during centrifu-
gation, calculate the dilution factor of the pellet so as to achieve
the desired concentration for freezing. If no centrifugation
took place, dilution is done based on the original sample con-
centration. When possible (depending on the species and ini-
tial sample concentration) we normally dilute the sample to
100150 106 spermatozoa per mL. One should remember
that if higher dilution factor is used, it might be desirable to
adjust the composition of the freezing extender, primarily the
concentration of the cryoprotective agent(s), to prevent dam-
age to the cells. Generally speaking, two systems exist with
respect to how the cryoprotectants are presented to the cells.
For the more robust sperm, the sample is diluted prior to
chilling with extender containing the cryoprotective agent so
as to achieve the final desired concentration. For more sensi-
tive spermatozoa, a two-stage dilution system is sometimes
used. At the first stage the sample is diluted to double the final
desired spermatozoa concentration with extender containing
no cryoprotective agent or just minimal quantity of it (about
10 % of the final volume). Then, after chilling and shortly
before freezing, the sample is diluted at 1:1 ratio with isother-
mal extender containing the cryoprotective agent in proper
concentration so as to achieve the final desired cryoprotective
concentration. Either way, the addition of the extender is done
very slowly while periodically swivelling the tube to ensure
proper mixing. Regardless of how the extender is added, it
should always be at the same temperature as the sample.
3.2 Sperm Chilling 1. After diluting the sample with extender, the tubes containing
and Freezing the extended sperm are placed in a suitable plastic or glass con-
tainer with isothermal water, let stand at room temperature for
3060 min for equilibration and then placed in the refrigerator
to cool slowly down to 45 C, a process that may take 23 h,
depending on the refrigerator and the volume of sample and
water in the container. It is thus best not to use too large a
container. Cooling without a water bath is sometimes done for
chilling-resistant spermatozoa but even for these, slow cooling
may be beneficial. A thermometer can be placed in the water so
that the temperature of the sample can be checked without
disturbing the sample (see Note 1).
2. Preparation for freezing: Prepare the stoppers for the
HollowTubes by putting the labels on them. Once the sam-
ple has reached the desired temperature, using a sterile glass or
disposable plastic pipette or a syringe with long needle (18 G
or larger), the sample is transferred into the HollowTubes.
The HollowTubes are then cupped with the stoppers and the
samples are kept standing or lying in the refrigerator ready for
freezing. When filling the tubes, make sure to fill them so that
about 1 cm of space will be left between the sample and the
stopper (see Note 2). The stoppers should be inserted with
care while pushing them and turning them till they are prop-
erly in place. Excessive force may break the central tube of the
HollowTube, resulting in spillage of the sample (see Note 3).
Stoppers that were not put all the way in or that are not straight
might cause leakage or problems during the passage of the
tube through the freezing machine. Sometimes the residual air
inside the tube prevents the stopper from going in. This air
must be released to allow space for the stopper to go in. The
stoppers were designed with this in mind so they come with an
appendage that allows the air to escape while pushing the stop-
per into place. When using such stoppers, the appendage is
removed once the stopper is in place to permit proper sealing
(see Note 4).
3. Freezing (see Note 5): Depending on the model of freezing
machine used, the process would be slightly different.
Description here will be based on models MTG 516 and MTG
550. These models come with cassettes with five holes in them,
where up to five tubes can be put for freezing (see Fig. 2a).
Thus, five tubes at a time is the maximum freezing capacity of
these machines. These tubes can be either those of the 2.5 mL
or those of the 8 mL. To freeze both sizes at the same time,
extensions for the 2.5-mL tubes are needed so that they will be
pushed in parallel to the larger tubes. The cassettes should
be kept in the refrigerator so they are at the same temperature
as the samples. Other models have built-in positions where the
samples are placed and yet other models may also have the
possibility to freeze straws and have suitable grooves for plac-
ing them in before freezing. During chilling, many of the cells
stop moving and sink to the bottom of the tube, so, before
putting the tubes in the cassette, the tubes should be tilted
slowly back and forth several times to resuspend the cells before
freezing.
Generally speaking, for freezing, a cassette with the sam-
ples is put into position, the start button is pressed and the
engine starts pushing the rods that push the samples. Once the
samples reach the predefined initial seeding position, the
machine stops and the sample is seeded with ice crystals, which
will then grow along the tube during the freezing process.
After 60 s of seeding (time can be adjusted through the control
panel of the machine), the engine kicks into action again and
starts moving the sample from the cassette through the tem-
perature gradient and into the cold block and then through
the cold block and out from its other end into the collection
chamber. From here the samples are removed directly into liq-
uid nitrogen and transferred to the storage tank.
The seeding distance is the distance the tube, when placed
in the cassette with the stopper facing forward, is pushed from
the home position till the entire stopper plus about 23 mm of
the tube are pushed into the cold block of the freezing machine.
While the machine is chilling in preparation for freezing, or even
ahead of time, the freezing machine should be programmed,
using its control panel, for the proper seeding distance. In case
both small and large tubes are to be frozen in the same session
but not simultaneously with the extension mentioned above,
the seeding distance should be adjusted when shifting from one
kind of tube to the other. If desired, it is possible to adjust the
velocity at which the engine pushes the sample across the tem-
perature gradient, from the cassette (5 C) to the cold block
(usually 50 C), and thus determine the cooling rate. From our
experience, best results are achieved when this is set to a rate of
1.01.5 mm/s (in the range of 47.1470.71 C/min). Faster or
slower rates can be achieved by adjusting the velocity of sample
propagation in the control panel.
About 30 min before freezing actually starts (usually when
the sample has reached the desired temperature and before
putting it into the HollowTubes), the machine should be
turned on, the pressure release valve of the liquid nitrogen
tank closed and the heater in the tank should be turned on so
that pressure can be built in the tank. Once enough pressure is
built in the tank, liquid nitrogen start being injected into the
machine to cool the cold block and collection chamber down
to their preset temperatures (usually 50 C and 100 C,
respectively). The cooling process may take about half an hour
so plan accordingly. Once the cold block has reached its preset
temperature, place five (or less) same-sized tubes (or, if needed,
small tubes with extension alongside large tubes) in the cassette,
all with the stopper facing in the same direction (in the direc-
tion of the arrow mark on the cassette; see Fig. 2a), place the
cassette in its position in the freezing machine with the arrow
pointing at the cold block, and press the Start button. Make
sure to be next to the machine and ready to handle any prob-
lems during the freezing process (see Note 5). Once the tubes
went through the machine and into the collection chamber,
they can be removed from there into liquid nitrogen and trans-
ferred to the storage tank (see Note 6).
3.3 Sample Thawing 1. Have everything prepared for the HollowTubes thawing
process: water bath (with the dedicated thawing device con-
nected; see Fig. 3a, b) at the desired thawing temperature (nor-
mally 37 C but sometimes going up to 6070 C for faster
warming rates of shorter duration), stand for the sample (see
Fig. 3c or flat Styrofoam instead), timer set to 2 min and 30 s
(when 37 C is the thawing temperature) or two timersone
set to 90 s and the other to 2 min and 30 s, -shaped glass (see
Fig. 3d), gloves, protective glasses, long tweezers, paper tow-
els, warm 15-mL disposable tube.
2. Remove the sample from the liquid nitrogen and start the
timer. If two times are used, start both simultaneously. For the
first 1015 s or so, hold the sample slanted with the stopper
facing down to allow expulsion of any liquid nitrogen that may
have managed to get into the tube. Then turn the sample with
the stopper up, place it in the dedicated stand or on the
Styrofoam and start massaging the stopper with your fingers
while turning it and alternating hands as the fingers get cold
after a while if gloves are not used. Do not use excessive force
as this might break the frozen stopper. The aim is to get only
the top of the stopper to warm and become soft enough to
allow the insertion of the -shaped glass into the hole in its
middle so that everything is ready when 90 s have elapsed (60 s
are left). This massaging process is less efficient when one wears
gloves on the hands as less heat is transferred to the stopper. At
this point, once the stopper is soft, the -shaped glass is in
place, and 90 s have elapsed, put the tube into the dedicated
thawing device in the water bath where it will thaw for 1 min.
When the tube is put inside the thawing device, water will flaw
over the outside wall of the tube as well as through the center
of the tube and out through the arms of the -shaped glass.
Because of the shape of these -shaped glass arms, the force of
the water will cause the tube to rotate and thus make warming
more homogenous. When thawing 2.5-mL tubes, first put an
empty 2.5-mL tube in the thawing device with the opening
facing up, so it will fill with water and sink to the bottom. Then
when putting in the tube for thawing, it will stand on top of
the sunken tube and not sink deep into the thawing device.
Sometimes friction, or the level of water in the bath, result in
tubes that do not rotate or rotate too slowly. Proper rotation
can then be achieved manually by turning the sample with the
hand. Adjustments to the water level can only be done before
thawing the next sample (see Note 7).
3. Once the time is up, remove the sample from the water. At this
point it would be at about 20 C and completely thawed.
Detach the -shaped glass and thoroughly dry the tube and
stopper from any residual water. Turn the tube slowly and care-
fully from side to side a few times to mix the sample inside.
Then, very slowly and carefully remove the stopper. First
remove it half way and dry again any water that may have col-
lected at the rim of the tube, and then remove it completely.
Pour the tubes contents into a pre-warmed and marked 15-mL
tube and proceed with post-thawing sample evaluations.
4 Notes
1. Chilling the sample at the right cooling rate is very important.

Cooling too fast can often harm the cells in the sample. It can
be generally said that cooling too slow is better than cooling
too fast. Therefore, avoid devising ways to shorten the cooling
duration, for example by adding ice to the water around the
sample. Usually, the drop in temperature is faster during the
beginning of the chilling process (large temperature gradient)
and slows down as you progress to lower temperatures. Thus,
the last few degrees before reaching the 5 C target may take
relatively long time. Have patience. In some refrigerators there
might be relatively large differences between the top and bot-
tom shelves. As cold air tends to move down while warm air
goes up, samples are better placed on the bottom shelf to make
sure adequate cooling rate is achieved.
2. When filling the HollowTubes with the sample one should
take care to put the sample only inside the tube. Any drop of
sample or water on the surface of the tube will surely freeze
once the sample moves into the cold block of the freezing
machine. This external ice is one of the causes for the freezing
process to stop in mid operation (see Note 5). In case spillage
did occur during HollowTube filling, make sure to thor-
oughly wipe it dry.
3. The area at the bottom of the HollowTube, where the inter-
nal tube is connected to the outside tube is the weakest point
in these containers. When attempting to push the stopper into
place before freezing or remove it after thawing, the tube

might break in this area. One way to prevent losing the sample
completely when this happens is to work above a sterile con-
tainer so that if spillage does occur, at least the sample will spill
into this container so it can be collected from there and put
into new tube. If one works cautiously, however, tube break-
age is rare.
4. Sometimes pushing the stopper into the tube is actually hin-
dered by the special appendage that was added to help release
trapped air in the tube. In such cases one can remove this
appendage and try pushing the stopper into place without it.
Alternatively, the appendage can be held separately in the tube
near the opening and then be removed once the stopper has
been pushed into place. Yet another option is to use a thin wire
or needle instead of the appendage. Sometimes a stopper might
be a bit too soft and it becomes impossible to push it into the
tube. In such cases one can try to put it in the freezer for a few
minutes to get it to stiffen a bit (but then be careful not to
bring it in direct contact with the sample) or simply replace it
with another stopper. As such incidences are not very rare,
make sure to have a few spare stoppers and labels just in case.
5. Humidity in the room and liquid droplets on the HollowTubes
are the enemies of the directional freezing machine. Because
the HollowTubes so snugly fit into the holes in the cold
block of the directional freezing machine, any ice accumulat-
ing in there may cause problems. Problems start with difficul-
ties of the engine to push the tubes through. The engine is not
an extremely powerful one so when it meets resistance it starts
jumping in its place instead of pushing the sample forward. A
stronger engine would have crushed the tube if met with resis-
tance. The HollowTubes are pushed by rods that, in some
models, are held in place with inverted U-shaped metal parts,
one for each rod (Fig. 2b). These may start moving or even
popping out of place when the tubes encounter too much
resistance. If too much ice accumulates in any of the tunnels in
the cold block, the tube in that tunnel will get stuck and every-
thing sill stop moving as the engine pushes all samples simulta-
neously. To avoid this from happening make sure the machine
is dry before turning it on. Also make sure the HollowTubes
are dry at all times. During the freezing process occasionally
rotate the rods that push the samples. When a tube meets resis-
tance it will be difficult to rotate the associated rod. Turning
the rod before complete blockage happens may help a bit. If resis-
tance it too much and the engine starts jumping, sometimes
it helps to hold it in place for a few seconds. This may help it
overcome a temporary block and proceed with the freezing
process. If nothing helps and the machine gets stuck, removing
the rod associated with the stuck tube by removing the

U-shaped holder at its rear end will let the machine continue
with the other tubes so that at least those samples are not lost.
When freezing is over and the machine is warmed so that all
accumulated ice is thawed, the tube that got stuck inside can
be removed. Because the temperature of the cold block is only
50 C, that sample can probably be considered lost. In some
models (e.g., MTG 1314) it is impossible to remove a rod dur-
ing operation. In other models (now sold by Harmony
Cryocare) the top of the cold block can be opened even during
operation to remove interferences or the sample that got stuck
without excessive disruption to the freezing process and with-
out losing the sample in most cases.
6. Sometimes, especially when too much humidity exists in the
system, the Hollowtubes may stick to the rods that push
them through the cold block and into the collection chamber.
As the machine is designed to retract the rods back to the
home position once they have pushed the samples completely
into the collection chamber, it is best to open the lid of the
collection chamber a few seconds before this happens and
release any tube that got stuck to the pushing rod. If this is not
done, the rod may pull the sample back with it.
7. For best thawing results, rotation of the HollowTube inside
the metal part (see Fig. 3b) is desired. To achieve that the stop-
per must be softened between the fingers during the time the
tube is held at room temperature. About 1015 s before the
tube is to be transferred into the water (7580 s after removing
it from the liquid nitrogen) push the -shaped glass (see
Fig. 3d) into the central hole in the stopper. Make sure it is
properly attached and straight. The tube might not rotate if
the -shaped glass is not placed properly or if the level of water
in the bath is too high or too low. For the specific tube already
in the water it is too late to make any adjustments so rotation
can be achieved by doing it manually. Adjustments to the water
level can be made before the next tube is thawed. If rotation
does not work at all or not well enough, one can also remove
the -shaped glass from the tube and transfer the tube directly
into the water bath where it should be moved back and forth
along its long axis allowing proper heat exchange with the sur-
rounding water.
Acknowledgement
The author is grateful to Dr. Amir Arav for useful suggestions on

the manuscript, and to Drs. Roland Fray and Gabriela Galateanu
for help with the figures.
References
1. Mazur P (2004) Principles of cryobiology. In: 11. Maffei S, Pennarossa G, Brevini TAL, Arav A,
Fuller BJ, Lane N, Benson EE (eds) Life in the Gandolfi F (2013) Beneficial effect of direc-
frozen state. CRC Press, Boca Raton, FL, tional freezing on in vitro viability of cryopre-
pp 365 served sheep whole ovaries and ovarian cortical
2. Watson PF (2000) The causes of reduced fertil- slices. Hum Reprod 29:114124
ity with cryopreserved semen. Anim Reprod 12. Gavish Z, Ben-Haim M, Arav A (2008)
Sci 6061:481492 Cryopreservation of whole murine and porcine
3. Saragusty J, Gacitua H, Rozenboim I, Arav A livers. Rejuvenation Res 11:765772
(2009) Do physical forces contribute to cryo- 13. Elami A, Gavish Z, Korach A, Houminer E,
damage? Biotechnol Bioeng 104:719728 Schneider A, Schwalb H, Arav A (2008)
4. Saragusty J, Gacitua H, Rozenboim I, Arav A Successful restoration of function of frozen and
(2009) Protective effects of iodixanol during thawed isolated rat hearts. J Thorac Cardiovasc
bovine sperm cryopreservation. Theriogenology Surg 135:666672.e661
71:14251432 14. Arav A, Zeron Y, Shturman H, Gacitua H
5. Arav A (2014) Cryopreservation of oocytes (2002) Successful pregnancies in cows follow-
and embryos. Theriogenology 81:96102 ing double freezing of a large volume of semen.
6. Arav A, Saragusty J (2013) Directional freezing Reprod Nutr Dev 42:583586
of spermatozoa and embryos. Reprod Fertil 15. Rubinsky B, Arav A, DeVries AL (1991)
Dev 26:8390 Cryopreservation of oocytes using directional
7. Saragusty J, Gacitua H, Zeron Y, Rozenboim I, cooling and antifreeze glycoproteins. CryoLetters
Arav A (2009) Double freezing of bovine 12:93106
semen. Anim Reprod Sci 115:1017 16. Arav A (1992) Vitrification of oocytes and
8. Saragusty J, Hildebrandt TB, Behr B, embryos. In: Gandolfi F, Lauria A (eds) New
Knieriem A, Kruse J, Hermes R (2009) trends in embryo transfer. Portland Press,
Successful cryopreservation of Asian elephant Cambridge, pp 255264
(Elephas maximus) spermatozoa. Anim Reprod 17. Arav A, Rubinsky B, Fletcher G, Seren E
Sci 115:255266 (1993) Cryogenic protection of oocytes with
9. Hermes R, Saragusty J, Gritz F, Bartels P, antifreeze proteins. Mol Reprod Dev 36:
Potier R, Baker B, Streich WJ, Hildebrandt TB 488493
(2013) Freezing African elephant semen as a 18. Revell SG, Pettit MT, Ford TC (1997) Use of
new population management tool. PLoS One centrifugation over iodixanol to reduce damage
8:e57616 when processing stallion sperm for freezing.
10. Revel A, Elami A, Bor A, Yavin S, Natan Y, J Reprod Fertil 19:38 (abstract)
Arav A (2001) Intact sheep ovary cryopreser- 19. Kim S, Agca C, Agca Y (2013) Effects of iodix-
vation and transplantation. Fertil Steril 76:S42 anol during rat epididymal sperm cryopreser-
S43 (abstract) vation. Cryobiology 67:432 (abstract)
Chapter 20
Vitrification of Heart Valve Tissues

Kelvin G.M. Brockbank, Zhenzhen Chen,
Elizabeth D. Greene, and Lia H. Campbell
Abstract
Application of the original vitrification protocol used for pieces of heart valves to intact heart valves has
evolved over time. Ice-free cryopreservation by Protocol 1 using VS55 is limited to small samples where
relatively rapid cooling and warming rates are possible. VS55 cryopreservation typically provides extracel-
lular matrix preservation with approximately 80 % cell viability and tissue function compared with fresh
untreated tissues. In contrast, ice-free cryopreservation using VS83, Protocols 2 and 3, has several advantages
over conventional cryopreservation methods and VS55 preservation, including long-term preservation
capability at 80 C; better matrix preservation than freezing with retention of material properties; very
low cell viability, reducing the risks of an immune reaction in vivo; reduced risks of microbial contamina-
tion associated with use of liquid nitrogen; improved in vivo functions; no significant recipient allogeneic
immune response; simplified manufacturing process; increased operator safety because liquid nitrogen is
not used; and reduced manufacturing costs.
Key words Heart valves, Vitrification, Tissue banking
1 Introduction
1.1 Overview Three types of heart valves are employed for replacement of defective
of Heart Valve Tissue valves in patients: mechanical, xenogeneic tissue, and allogeneic
Banking human valves derived from donors post-mortem. Most patients
receive either xenogeneic tissue or mechanical valves; however, the
use of cryopreserved human valve allografts became established
during the 1970s and 1980s for certain patient subsets. Two tri-
leaflet valves, the aortic and pulmonary, with associated arterial
conduit and cardiac muscle band, are dissected from cadaver donor
hearts for patient use. Following cryopreservation, aortic and
pulmonary valve allografts have typically been used to reconstruct
the left or right ventricular outflow tract for repair of diverse con-
genital cardiac anomalies. Cryopreserved allografts have especially
benefited children with congenital heart disease since the use of
alternative mechanical and xenogeneic tissue valves has historically
Willem F. Wolkers and Harriette Oldenhof (eds.), Cryopreservation and Freeze-Drying Protocols, Methods in Molecular Biology,
399
400 Kelvin G.M. Brockbank et al.
been limited in this patient population. They are also used in

women of childbearing age and older patients with memory prob-
lems who may not be relied upon to keep up with the medications
required for mechanical valves.
In the last century, cryopreserved, control rate frozen Me2SO-
protected, human heart valves were used in approximately 20 % of
the tissue heart valve procedures performed annually [14]. They
are currently less utilized because of new xenogeneic valves, which
are impacting aortic valve utilization, and because of failures in
young patients [5], particularly infants [6, 7]. The pathophysiology
of allograft heart valve failure is not fully understood [810].
Immunology definitely played a significant role in failure in young
patients; this led to the development of decellularization methods
to minimize the recipient immune response upon implantation.
We, however, hypothesized that the rapid deterioration seen in
some allograft heart valve recipients might be due, at least in part,
to disruptive interstitial ice damage that occurred during cryo-
preservation leading subsequently to accelerated valve degeneration
upon implantation. About 75 % of the area of conventional Me2SO-
cryopreserved heart valve leaflets, the most important functional
component of a heart valve, was occupied by ice [11] (see Fig. 1).
It was hard to imagine that such extensive ice formation was not
damaging the tissue. Both light and electron microscopy demon-
strated that the ice was extracellular. Upon thawing and rehydra-
tion, the leaflets looked normal histologically because the ice
domains within the extracellular matrix closed up as the ice melted.
Development of ice-free cryopreservation by vitrification meth-
ods for heart valves was initially stimulated by our demonstration of
large ice crystal domains within heart valve tissues [11] (see Fig. 1).
Vitrification as we have practiced it for heart valve tissues is a non-
equilibrium approach (Protocol 1) in contrast to the equilibrium
approach developed by Pegg and his colleagues (reviewed in 12).
This method for the avoidance of ice was based on the early work of
Fahy and Rall [1315]. The basic principles of this approach have
been discussed and reviewed in great detail many times [12, 13,
1520]. In addition to heart valve tissues, we have used Protocol 1
for many other types of tissue including blood vessels [2124],
cartilage [2528], pancreatic islets [29], and engineered tissues
[3032]. The functional survival of these vitrified tissues was
approximately 80 % or higher, whereas the frozen counterparts
yielded less than 20 % survival (see the The emerging 80/20 rule
[12]). This marked contrast (~80 % vs. 2030 %) was consistent
irrespective of the nature of the tissue except in the case of cardio-
vascular tissues where equivalent post-cryopreservation cell viability
can be obtained by either freezing or vitrification methods. In the
case of tissues that were particularly susceptible to cryoprotectant
cytotoxicity, the final steps in addition and first steps in removal of
the cryoprotectants could be performed at subzero temperatures.
Vitrification of Heart Valve Tissues 401
Fig. 1 Montage of micrographs from a cryo-substituted porcine heart valve leaf-

let demonstrating the presence of ice (white spaces) throughout the tissue. This
tissue sample was slow frozen in culture medium containing 1 M Me2SO. Cryo-
substitution was performed with methanol at 90 C as previously reported [21]
Our first heart valve vitrification studies utilized pieces of

porcine aortic valves and intact rat aortic valves in small volumes
of a vitrification solution (VS) called VS55 to reflect its 55 %
(w/v) of cryoprotectant solutes that was originally designated
VS41A by its originators [14]. VS55 consists of an 8.4 M mixture
of 1,2-propanediol, formamide, and dimethyl sulfoxide in Euro-
Collins (EC) solution. EC is an extracellular-type preservation
solution that was developed in the 1970s by Collins and subsequently
Table 1
Viability post-VS55 heart valve leaflet preservation
Post-recovery time incubation (days) 0 1 2

a
Relative florescent units/mg dry weight (mean, N = 6) 2,047 2,318 2,743
Results in % of fresh untreated controls 80 90 107
a
Resazurin metabolic assay [45]
modified in Europe, where it was the gold standard for hypothermic

storage of human kidneys for several years. The protocol is
described in US Patent #6,740,484 [33, 34] and in Protocol 1
below. In the porcine studies, we demonstrated that in contrast
to conventionally frozen heart valves, vitrified leaflets had no vis-
ible ice by light microscopy. There was some discussion of whether
or not very small ice crystals could be observed by electron
microscopy. In vitro studies with animal tissues have repeatedly
demonstrated that Protocol 1 maintained ~80 % heart valve leaflet
cell viability immediately after rewarming, returning to control
values over a few days in tissue culture (example, Table 1). The
emphasis in our rat studies was assessment of transplant out-
comes. Two implant models were employed (1) a subcutaneous
implantation calcification model and (2) descending thoracic
aorta implants [35]. The calcification rate of frozen valves was
significantly greater (p < 0.01) than that of vitrified valves in both
syngeneic and allogeneic recipients, supporting prior observa-
tions that ice-free cryopreservation reduces allogeneic heart valve
calcification. However, cryopreservation by freezing and vitrifica-
tion resulted in only mild morphological changes in 2- and
4-week syngeneic explants, a slight decrease in leaflet cellularity,
and a more rapid onset of intimal hyperplasia than in fresh valve
explants. The allograft explant groups exhibited changes, fresh
allograft leaflets were relatively acellular, and frozen leaflets were
hypercellular due to acute inflammatory reactions compared with
the vitrified allografts (see Fig. 2). At that time, we decided that
these observations provided only weak support for the hypoth-
esis that the rapid deterioration seen in some allograft heart valve
recipients was due to disruptive interstitial ice damage that occurs
during cryopreservation by freezing [35]. In hindsight it is
likely that the decreased inflammation observed in ice-free vitri-
fied heart valve explants was actually a sign of the decreased
immune response subsequently observed in sheep.
1.2 Development When our rat studies were published in 2004 [35], our vitrified
of a Vitrification heart valve research had encountered a significant hurdle; a stron-
Protocol ger demonstration of superiority to conventional cryopreserva-
tion was needed to warrant further development. In 2006 the
Fig. 2 Representative heterotopic rat heart valve explant histology. The valves were explanted at 2 weeks (4
week explants were similar in appearance). The syngeneic valves all appear similar; however, both the fresh
allograft and frozen valve leaflets appear thickened relative to vitrified valves. Fresh allograft leaflets were
relatively acellular, and frozen leaflets were hypercellular due to acute inflammatory reactions. Leaflets indi-
cated by arrows, hematoxylin and eosin stain, 10 magnification
publication of a paper using advanced morphological methods

that demonstrated significant differences in extracellular matrices
of fresh and Me2SO-cryopreserved frozen heart valves [36, 37]
provided the tools and experience to compare vitrified and frozen
heart valves. Autofluorescence and particularly second harmonic
generated microscopy images revealed that conventional frozen
cryopreservation of cardiac valves, when compared with fresh or
vitrified tissues, led to the loss of normal ECM structures in valve
leaflets [38]. Similar results were found in associated cardiac mus-
cle and artery suggesting that structural deterioration of the
ECM is a common consequence of frozen cryopreservation.
These observations led to a renewed enthusiasm for further
development and optimization of the vitrification protocol for
heart valves. Our first protocol was excellent for small-volume
tissue samples where relatively rapid cooling and warming rates
that avoided ice formation could be achieved. However, upon
scale up to full-sized heart valves, 80100 mL volume, we
encountered two problems: (1) cracking at vapor-phase nitrogen
temperatures, particularly in dry nitrogen shippers, and (2) the VS55

solution demonstrated ice formation during rewarming [39].
The solution to both problems was to increase the concentrations
of the three cryoprotectants in VS55 from 55 to 83 % to make
VS83. Furthermore, modulated differential scanning calorimetry
studies indicated that this new formulation was potentially stable
at 80 C, free from ice at 80 C [39], which would make it
easier and cheaper to store and ship the tissue samples.
Three in vivo juvenile sheep implant studies were performed to
test the VS83 protocol (Protocol 2). The first two studies are
unpublished. Protocol 2 was similar to Protocol 1 except that
VS83 was employed and after cooling the tissues were stored at
80 C and shipped on dry ice to the implant site. There were dif-
ferences in the post-warming wash in each study. Another issue
that should be made clear is that in contrast to early work with
VS55, there was no effort to keep heart valve tissue cells alive,
viable, with VS83. The reason for this is that at this time there is no
convincing evidence in the literature suggesting that maintenance
of graft cell viability after cryopreservation is beneficial in alloge-
neic heart valve recipients. In the early years, 1980s and 1990s, of
allograft heart valve processing, cell viability was thought to be
very important for long-term maintenance of the heart valve post-
transplant in patients. Since then analysis of explanted heart valves
has supported the assumption that living cells are not necessary in
a functioning transplant. Explant analyses indicated that most of
the original cells either vanished, especially in the leaflet cusps, or
were replaced by donor fibroblasts [40, 41]. Only a few living cells
from the original donor could be detected after long-term trans-
plantation and normal graft leaflet cellularity was not observed in a
single explant [42]. Furthermore, many groups are now decellular-
izing heart valves in order to reduce their concerns regarding
immunogenicity and its consequences in patients [43, 44].
Therefore, we did not attempt to preserve cell viability during fur-
ther development of the VS83 method for heart valves. We had
previously demonstrated that the VS83 formulation cryoprotec-
tant components are cytotoxic for cultured porcine myofibroblasts
at high concentrations [45].
In the first study, we compared conventional frozen aortic
sheep heart valves with VS83 vitrified valves stored at 80 C or in
vapor-phase nitrogen over a period of 6 months in vivo. These
valves went through a multistep wash protocol similar to that
described in Protocol 1. Explant pathology studies indicated that
frozen heart valves had most deterioration at explant and 2/4
explants had changes that may have been due to ice damage (Fig. 3
and Table 2). Even though all heart valves exhibited changes, the
VS83 formulation combined with 80 C storage demonstrated
the least structural deterioration. Explant gross pathology and his-
tology revealed little difference, except that two cryopreserved by
Fig. 3 Sheep heart valve explant gross and histopathology. Heart valves were explanted at 6 months and pho-
tographed before further processing. (a) VS83 formulation stored at 80 C in a mechanical storage freezer,
(b) H&E-stained section of VS83 80 C explant after 6 months in vivo demonstrating cell-free leaflet with
recipient cell-derived pannus overgrowth on the artery (bottom and leaflet top), (c) and (d) cryopreserved by
freezing controls, (d) also demonstrating a hole in the base of leaflet, (e) and (f) VS83 formulation stored in
vapor-phase nitrogen
Table 2
Pathology overview
AI at 6
Group Description Implant pathologya monthsb
1 Conventional Mild diffuse leaflet fibrosis. 3/4 exhibited focal 12+, 12+,
freezing nodular vegetation (organized thrombi) and 2+, 23+
method 1/4 exhibited calcified foci. 1/4 exhibited a
basal dehiscence and 1/4 a paravalvular
aneurysm. 1/4 conduits mineralized
2 VS83 stored Implants all exhibited mild diffuse fibrosis; 2 of 4 2+,
at 80 C conduits were mineralized none,1+,
trace-1+
3 VS83 stored in Diffuse and focal fibrosis, 2 of 4 exhibited thrombi None, 23+,
vapor-phase and 1 of 4 exhibited focal nodular vegetation on 23+, 23+
nitrogen leaflets. 3 of 4 conduits were mineralized
a
Leaflets in all groups were devoid of cells except for recipient overgrowth that appears to be a chronic low-grade inflam-
matory response with a marked tendency for all leaflets to exhibit some thinning with thicker fibrous overgrowth at the
base of the cusps resulting in leaflet thickening and stiffening
b
Aortic Insufficiency(AI) was assessed by echocardiographic evaluation just prior to sacrifice
Fig 4 Calcium contents in sheep heart valve explants. Leaflet results are on the left and conduit results on the
right. VPN = vapor-phase nitrogen. Calcium-specific mineralization occurring in aortic allografts was quantified
using atomic absorption spectroscopy [59]. There were no significant differences
freezing controls had single holes present at explant that could

have been due to ice damage (example, Fig. 3d). There were no
significant differences in calcium content in explanted leaflet or
artery samples (Fig. 4) and least development of aortic insuffi-
ciency (AI) (Table 2).
These test results suggested that preservation methods based
upon modifications of VS83 combined with 80 C storage in a
mechanical freezer may provide an alternative for liquid nitrogen
storage of allogeneic heart valves, reducing the costs of both long-
term storage and distribution. Due to the small size of each experi-
mental group (n = 4), it was not possible to state that the benefits
of the VS83 protocol combined with 80 C storage were signifi-
cantly different from cryopreservation by freezing. However, this
test group demonstrated less structural deterioration and less aor-
tic insufficiency than traditionally cryopreserved by freezing
allograft heart valves after 6 months in vivo in sheep.
The second study was an unpublished Current Good
Laboratory Practice (cGLP) study in which decellularized arterial
tissue, sheep pulmonary artery, was vitrified (n = 5 sheep) and
compared with cryopreserved tissue (n = 2 sheep) by implanta-
tion as patches in the pulmonary artery of juvenile female sheep
(results on file pending future regulatory submissions). The
cryopreserved/decellularized/vitrified samples had an average
99.8 0.05 % reduction in extractable DNA. Two patches were
implanted in each sheep, and they were explanted after 140 days.
One explant from each sheep was used for calcium analysis (non-
GLP inductively coupled plasma optical emission spectroscopy)
and the other for histopathology. Histopathology included review
of H&E, -actin smooth muscle, TUNEL for apoptosis, Hsp47
(marker for the production of type I and III collagens) (Fig. 5),
and factor VIII endothelial cell marker-stained sections. Panel
reactive antibody analysis was performed using serum obtained
pre-implant and 10 and 20 weeks post implant. None of the
Fig. 5 Representative explant histology from 20-week decellularized VS83-preserved artery patch explants: (a)
hematoxylin and eosin stain, 100; (b) TUNEL stain to identify apoptotic cells, 200; (c) actin stain to identify
smooth muscle cells, 100; and (d) HSP47-stained cells to identify cells actively producing collagen, 100.
Factor VIII-positive endothelial cells were also observed (not shown)
animals demonstrated MHC Class I antibodies. One of the two

controls demonstrated a 54.65 % shift to the right over baseline
for MHC Class II. One of five experimental sheep experienced an
11.16 % shift over baseline for MHC Class II. There were no dif-
ferences in blood work, echocardiography, or necropsy results
between the experimental and control groups. The cryopreserved
patches decreased in cellularity and calcified. The decellularized
vitrified patches recellularized and resisted calcification, and the
infiltrating cells were producing -smooth muscle actin, factor
VIII, and Hsp47 after 20 weeks in vivo. An increase in TUNEL-
positive cells was observed in vitrified tissue compared with cryo-
preserved controls, 45 vs. 34 % positive. The cellularity was the
result of fibroblast and myofibroblast infiltration, and the amount
of inflammation seen was little to none. The results were similar
to published work on the decellularization method employed
using ovine patch grafts in the same animal model [4648]. These
results suggest that VS83-preserved tissues should be safe in
humans. However, the increased apoptosis noted in the TUNEL
staining led to in vitro studies to determine whether rewarmed,

washed vitrified tissues were cytotoxic (discussed later).
The third study was a research study in which the VS83 proto-
col was employed with simplified washing steps to process ovine
pulmonary heart valves (Protocol 2 [49, 50]). Vitrified and con-
ventionally frozen valves were then stored at 80 C and in vapor-
phase nitrogen, respectively, for approximately 1 year. Vitrified
pulmonary valves were shipped from Charleston, South Carolina,
to the implantation site in Berlin, Germany, on dry ice and stored
at 80 C until implantation. The control frozen pulmonary valves
were shipped in a nitrogen dry shipper and stored at 135 C in
Berlin. Aortic valve tissues were kept in Charleston to evaluate the
impact of preservation without implantation. The pulmonary
valves were implanted in the pulmonary position. We had concerns
that our earlier unpublished sheep studies, presented above, uti-
lized sheep that may have been highly inbred because crossbred
Whiteface heart donors and recipients were used. Therefore, two
different strains of sheep were chosen, crossbred Whiteface heart
valve donors from Minnesota and Heidschnucke, a Nordic short-
tailed breed, recipients from Germany to increase the possibility of
using a true allogeneic model.
Multiphoton microscopy revealed reduced, but not signifi-
cantly, damaged extracellular matrix before implantation in frozen
valves compared with ice-free tissues. Viability assessment revealed
significantly less metabolic activity in the ice-free valve leaflets and
artery samples compared with frozen tissues (p < 0.05). After 3
and 6 months, in vivo valve function was determined by two-
dimensional echo-Doppler, and at 7 months the valves were
explanted. Severe valvular stenosis with right heart failure was
observed in recipients of frozen valves; the echo data revealed
increased velocity and pressure gradients compared to ice-free
valve recipients (p = 0.0403, p = 0.0591). Histopathology showed
significantly thickened leaflets in the frozen valves (p < 0.05) and
infiltrating CD3+ T-cells (p < 0.05) compared with ice-free vitri-
fied valve leaflets. Multiphoton microscopy at explant revealed
reduced inducible autofluorescence and extracellular matrix dam-
age in the frozen explants and well-preserved structures in the
ice-free explant leaflets. We concluded that ice-free cryopreserva-
tion of heart valve transplants at 80 C avoids ice formation and
tissue-glass cracking and preserves extracellular matrix integrity
resulting in minimal inflammation and improved hemodynamics
in allogeneic juvenile sheep [4951].
Biomaterial properties of ice-free preserved heart valves have
also been assessed [39]. We observed occasional tissue cracking
and fracturing of the vitrified solution in experiments involving
storage of heart valves in VS55 below the solution glass transition
temperature in vapor-phase nitrogen. Substitution of VS83 for
VS55 corrected this for samples stored at 135 C. Biomechanical
testing was performed on fresh untreated control and VS83 cryo-

preserved leaflets and stored at 80 and 135 C. There were no
significant differences in Youngs modulus and ultimate stress
(strain) observed between the 80 and 135 C stored leaflets for
any of the mechanical test comparisons; similarly neither group
differed significantly when compared with the fresh untreated con-
trol leaflet samples (Fig. 6). Thakrar [52] reported that vitreous
cryopreservation, employing 40 % w/v 1,2-propanediol in DMEM
culture medium maintains the viscoelastic property of human vas-
cular grafts in contrast to conventional cryopreservation by freez-
ing with Me2SO. We also assessed the impact of VS83 preservation
upon ECM integrity at the two storage temperatures. Multiphoton
imaging of VS83-preserved heart valves stored at each temperature
demonstrated branched elastic fibers and wavy bundles of collagen
in almost all vitrified heart valve leaflets, aortic trunk, and cardiac
muscle specimens. There were no significant differences in SHG
signal intensities (Fig. 7) or in qualitative mean rating scores on
tissue quality determined by the operator during imaging. These
studies provided further support for the storage of VS83-preserved
Fig. 6 Summary of VS83 biomechanical testing. Results comparing VS83-preserved leaflets, stored at 80 or
135 C, and fresh untreated control leaflets are presented as (a) ultimate force, (b) ultimate stress, (c) ulti-
mate strain, and (d) Youngs modulus. Data presented as the mean 1 standard deviation, n = 56, for each
individual sample mean. There were no significant differences [39]
Fig. 7 Summary of second harmonic generated signal intensities. Collagenous

structures in heart valve tissues were observed by spectral fingerprinting after
VS83 preservation at 80 or 135 C. Data presented as the mean 1 standard
deviation, n = 4, of individual sample mean maximal intensity emission values
detected at peak emission wavelengths of 414425 nm. There were no signifi-
cant differences [39]
tissues at 80 C provided that cell viability was not required.

However, these tissues are technically not vitrified since we are
storing them above the glass transition temperature. Therefore, in
most of our more recent publications, we refer to tissues preserved
at 80 C as being ice-free cryopreserved.
Further process simplification has since been performed
(Protocol 3) with in vitro evaluation studies to assess the impact of
VS83 vitrification on cell viability and tissue hemocompatibility.
During the development of Protocol 3, we examined the impact of
varying the number of addition or removal steps in combination
with cooling and found that it made little if any difference for cell
survival. Room temperature incubation in VS83 without subse-
quent cooling and storage had similar effects on cell viability. The
vitrified artery and leaflet components were significantly less viable
than either fresh or Me2SO-cryopreserved frozen tissues. The mus-
cle component, while less viable, was not significantly different.
The changes in tissue viability were a function of cryoprotectant
exposure and cytotoxicity, not temperature reduction during storage.
Therefore, the simple one-step addition at room temperature was
selected for use in further experiments. Multiple changes of washing

solution were employed to minimize the residual cryoprotectant
concentrations after rewarming. Subsequent studies confirmed
the impact of VS83 on heart valve tissue cell viability using Protocol
3 [52] and demonstrated that heart valve tissues were hemocom-
patible using multiple assays [53, 54]. Protocol 3 will be tested
in vivo with either ice-cold or room temperature wash solutions in
future studies.
1.3 Current State In the evolution of our vitrification protocols, we started with
Protocol 1 based upon VS55 (reviewed in 12, 55). This protocol
resulted in excellent avoidance of ice formation with retention of
cell viability and matrix integrity provided that the sample size and
geometry permitted rapid cooling and warming. Ice formed during
slow rewarming of VS55 but not during rewarming of a more con-
centrated cryoprotectant solution, VS83 [39]. The use of Protocol
2 with VS83 permitted the use of large sample volumes and heart
valve tissue storage at 80 C at the expense of cell viability.
A variety of reasons for allograft heart valve failure were dis-
cussed in the past, and most investigators have emphasized immu-
nological issues [41, 56]. Standard quantitative and qualitative
cellular and matrix evaluations such as biochemical, immunohisto-
chemical screening, and routine histology did not help to solve
the controversial discussion whether remaining allogeneic cells or
potentially altered extracellular matrix contributed to the observed
degeneration. Preliminary data on patients treated with decellu-
larized allografts has recently demonstrated that decellularization
did not significantly improve outcome in terms of pressure gradi-
ents and structural deterioration compared to nondecellularized
allografts [57]. These early clinical results question the validity of
theories suggesting that an immune reaction to the remaining
donor cells in allogeneic heart valves is the sole cause of structural
deterioration. The impact of vitrification on the immunoreactivity
of tissue grafts was first clearly observed in the in vivo sheep model
study employing different sheep strains discussed above [49, 50].
Explant leaflets demonstrated well-preserved extracellular matrix
structures, with few CD3+ T-cells and significantly improved
hemodynamics. In marked contrast, leaflets in the frozen control
group demonstrated significant matrix damage and T-cell-mediated
inflammation. The observation that explanted vitrified leaflets
were nearly devoid of active immune cells suggested that a state
equivalent to decellularization was accomplished by the ice-free
cryopreservation step. Further support for this observation was
obtained in studies combining human cells with rewarmed,
washed tissues in vitro [58]. Concerns that the apparent benefits
of vitrification were due to cytotoxic cryoprotectant residuals were
dismissed by demonstration that cells cultured with these tissues
survived [58] (Figs. 8 and 9). Our current working hypothesis to
Fig. 8 Assessment of VS83-preserved cytotoxicity in vitro. Human umbilical vein-derived endothelial cells were
cultured with ice-free preserved vessels for 7 days in non-tissue culture grade tubes. The constructs were
tested on days 2, 3, 4, and 7 using the resazurin reduction assay. There was a gradual decrease over time in
culture for all groups, no significant differences except for comparison of the no cell group with the HUVEC
groups (p < 0.05). No cell = ice-free preserved vessels without added endothelial cells, Decell = decellularized
vessels plus cells. Rewarmed ice-free preserved vessels were subjected to 2, 4, and 8 wash steps. Note: The
no cell and 28 step groups were not decellularized. The gradually increasing viability in the no cell group is
due to proliferation of a small number of smooth muscle cells, which were used to engineer the vessel, surviv-
ing the ice-free preservation process
Fig. 9 Endothelial cell-vessel co-culture histology after 7 days. Human umbilical vein-derived endothelial cells
were cultured with ice-free preserved vessels for 7 days in non-tissue culture grade tubes. The tissues were
then fixed, embedded, sectioned, and stained for the CD31 endothelial cell marker and photographed at 40
magnification. (a) No cell = ice-free preserved vessel without added endothelial cells, (b) control decellularized
vessel plus endothelial cells, and ice-free preserved vessel with either (c) 2 wash steps, (d) 4 wash steps, or
(e) 8 wash steps. No qualitative differences were observed between the decellularized and IFC groups. Note:
CD31-positive endothelial cells at the upper edge of the tissue in be. The interstitial staining is not associated
with endothelial cells since it is also seen in a, the no cell added group
explain these changes in vitrified tissue immunoreactivity is that

this vitrification method is modifying or masking tissue signals
perceived by responder cells. These tissue signals are most likely
damage-associated molecular pattern molecules that can initiate
and perpetuate immune responses.
The potential advantages of VS83 heart valve storage at 80 C
employing Protocol 3 includes reduced infrastructural needs for
preservation, storage, and shipping in comparison with traditional
freezing methods while maintaining extracellular matrix integrity
and material properties. The loss of cell viability is probably a ben-
efit contributing to the reduction of immunogenicity observed
in vivo and in vitro. The typical methods employed for conven-
tional cryopreserved frozen heart valves require a control rate
freezer, storage in nitrogen-cooled tanks, and a continuous supply
of liquid nitrogen. Furthermore, conventionally cryopreserved
valves also require shipment in nitrogen dry shippers to the implan-
tation site where the valves need to be kept in nitrogen-cooled
freezers until implantation. Dry shippers are expensive, and the
cost of return shipment adds to the cost of the heart valve to the
end user. Liquid nitrogen is also a safety hazard for employees, and
precautions must be taken during operation of nitrogen-cooled
equipment. In contrast, no expensive equipment and liquid nitro-
gen are needed for the Protocol 3 VS83 storage method that has
evolved. There may also be a reduced risk of microbial contamina-
tion associated without use of liquid nitrogen. The only equipment
required is a mechanical storage freezer, and long-term preserva-
tion can be performed at 80 C. Shipping requires only an insu-
lated box with dry ice. The removal of liquid nitrogen from the
process also reduces the training required and safety hazards for
employees. In the future, we anticipate removal of the cooling
below 100 C step in the initial cooling procedure in order to
further reduce the complexity and cost of this method. In vivo
large animal studies of VS83 preservation combined with 80 C
storage employing pulmonary valves in juvenile sheep have already
demonstrated better function and explant pathology than tradi-
tionally frozen cryopreserved heart valves [49, 50]. VS83 preserva-
tion protocols combined with 80 C storage method should be
particularly beneficial in developing countries with limited financial
and logistic resources that have a high incidence of aortic heart
valve disease.
2 Materials
Prepare all solutions using deionized water (prepared with an acti-

vated carbon filter, a mixed bed working deionizer, and a mixed
bed polishing deionizer to attain a sensitivity of 10 M cm at room
temperature). The solutions are prepared with raw materials that
meet or exceed requirements of the American Chemical Society. If

no such specifications exist, chemicals of the highest purity avail-
able should be used. The solutions are made at room temperature
and stored at 4 C for up to 4 weeks. The solution formulations
described are for 1 L batches.
2.1 Euro-Collins 1. Put 0.5 L water in a container (graduated cylinder or glass

Solutions beaker), then add a stir bar followed by addition of each com-
ponent in the following order, 174.76 g Dextrose, 10.2 g
KH2PO4, 36.5 g K2HPO4, 5.6 g KCl, 4.2 g NaHCO3 make up
to 1 L with water to make a 5 stock solution. Stir continu-
ously during formulation; when the solution is clear of parti-
cles, it should be filtered using 0.2 m filters. We do not test
the osmolality or pH of 5 Euro-Collins. Dilute 1:5 with water
to make 1 Euro-Collins and check the pH (~7.4) and osmo-
lality (365 5). Store at 4 C.
2.2 Vitrification 1. VSs are formulated by placing a stir bar and then each compo-
Solutions nent, in the order that they appear in the list below (Table 3),
in a graduated container. Stir continuously; when finished and
it is clear of particles, it should be filtered. We use bottle top
0.2 m filters, and it takes a long time (hours) to filter due to
the high viscosity of the solutions. At the conclusion of prepa-
ration, check the final pH and make a 1:20 dilution for assess-
ment of osmolality. Store at 4 C.
2.3 Vitrification 1. Sequential vitrification solutions, both addition and removal

Solutions solutions, are used in certain protocols (see Note 1). The VS
chemical component concentration can be increased from 1/8,
Table 3
VS formulation, quantities
VS55 VS83
5 Euro-Collins (mL) 200 200
HEPES buffer (g) 2.39 2.39
Propylene glycola (g) 168.38 252.57
Formamidea (g) 139.56 209.34
a
Me2SO (g) 242.14 363.21
Water (mL) to 1 L to 1 L
pH (units) 7.98.1 7.98.1
Osmolalityb (mOsm/kg) 460 5 670 5
a
Liquids measured by weight
b
20 dilution = 50 L in 950 L of distilled water
, , and to 100 % of the final full-strength VS formulation

being employed. The diluted VS solutions are prepared by the
addition of appropriate volumes of 1 EC. The solutions
employed for removal are similar to those employed for addi-
tion except that 200400 mM mannitol is added to the formu-
lation as an osmotic buffer. For example, using seven steps, the
VS CPAs are decreased from full strength to 0 M plus mannitol
to 0 M, plain EC, and finally Dulbeccos Modified Eagle
Medium (DMEM).
3 Methods
3.1 Protocol 1: 1. Tissues should be gradually infiltrated with VS55 at 4 C

Suitable for Small (see Note 1). Precooled diluted VS solutions are employed in
Heart Valve Tissue six 15 min steps at 4 C of increasing CPA concentration as
Samples (3 mL Total previously described [12] (see Note 2).
Volume, Sample Plus 2. The top of the vitrification solution is then covered with
Solution) 2-methylbutane (isopentane; see Note 3).
with Maximum 3. The samples should be cooled rapidly to 100 C followed by
Retention of Cell slow cooling to 135 C (see Note 4).
Viability and Matrix
4. Finally the samples should be stored at 135 C in a mechani-
Integrity
cal storage freezer or at 135 C near the top in a vapor-
phase nitrogen freezer.
5. Rewarming of vitrified tissues is a two-stage process including
slow warming to 100 C and then rapid warming to ~4 C
(see Note 5). Wipe the container with 70 % ethanol to mini-
mize the risk of microbial contamination and use sterile pick-
ups to transfer the tissue to the first VS removal solution.
6. The vitrification solution is then removed in a stepwise manner
at 4 C in six or seven sequential, 15 min steps of increasingly
dilute VS solutions containing mannitol as previously described
[12] (see Note 1). Finally the tissues are placed in DMEM
culture medium.
3.2 Protocol 2: 1. Tissues are gradually infiltrated with VS83 in six 15 min steps
Suitable for Large of increasing concentration at 4 C.
Heart Valve Tissue 2. After the final step, the tissues are placed individually in sterile
Samples (Up polyester bags containing 7080 mL of the vitrification solu-
to 100 mL Total tion. Each bag is then evacuated of air by squeezing (see Note
Volume) 3) and heat-sealed.
with Preservation 3. The bags containing tissues need to be cooled rapidly to
of Matrix Integrity 100 C in a precooled 2-methylbutane bath and then trans-
and Minimal Residual ferred to storage (see Note 4).
Cell Viability
4. Store at 80 C in a mechanical storage freezer (see Note 6).
5. Rewarming needs to be performed by placing each bag in a

37 C water bath until the solution moves freely.
6. The bag is then transferred onto ice and the bag externally
cleaned with 70 % ethanol to reduce the risk of microbial con-
tamination. Cut the bag open with sterile scissors and use sterile
pickups to transfer the tissue to the first VS removal solution.
7. A multistep VS removal protocol was employed in our first
unpublished sheep study, similar to that used in Protocol 1,
above. Simpler, washing methods have been employed in a
subsequent sheep study [49, 50] and an unpublished study
with arterial patches (see Note 7).
3.3 Protocol 3: 1. Place the tissues in sterile polyethylene bags containing 80 mL

Simplified Versions of VS83 at room temperature (see Note 1).
of Protocol 2 2. Evacuate the air and heat-seal the bags (see Note 3).
in Development
3. Incubate the tissues with continuous shaking for at least 1 h at
for Heart Valves
room temperature prior to cooling.
4. The cooling process can then be achieved by placing the bags
for 10 min in a precooled bath of 2-methylbutane (< 100 C,
see Note 4).
5. The tissues can then be placed at 80 C for storage.
6. Rewarming of the tissue is then performed by submersion of
the tissue in its bag in a 37 C water bath.
7. The bag is then dried and externally cleaned with 70 % ethanol
to reduce the risk of microbial contamination. Cut the bag
open with sterile scissors and use sterile pickups to transfer the
tissue to the first wash step.
8. Washing to remove CPAs has been performed using five wash
steps of 5 min duration each with ice Euro-Collins solution or
with several washes with Lactated Ringers solution containing
5 % dextrose (see Note 8).
4 Notes
1. Stepwise VS addition and removal procedures are employed in

Protocol 1 in order to minimize damage to cells as they respond
to osmotic changes in their environment as a consequence of
being placed in VS solutions with high concentrations of CPAs
(reviewed in 12). This is not an issue in Protocols 2 and 3
because we are not trying to maximize cell viability. Loss of cell
viability is considered advantageous because it appears to
reduce immunogenicity in vivo [49, 50].
2. In the last step, the tissues are placed in vials; we usually use
glass scintillation vials (25 60 mm), containing 12 mL of
precooled full-strength VS solution. In circumstances where

unexpected cytotoxicity is observed, the last one or two steps
can be performed at subzero temperatures in a cold bath to
minimize cytotoxicity.
3. We use 0.71.0 mL of 2-methylbutane for the glass vials
described in Note 2. The 2-methylbutane has a freezing point
of 160 C and density of 0.62 g/cm3. The 2-methylbutane
layer prevents direct air contact minimizing the risk of ice
nucleation. Alternatively, the vial can be purged with a dry
gas, such as nitrogen, or a container without excess air space
can be used.
4. Rapid cooling to < 100 C is done by placing the vials in a
precooled (135 C) 2-methylbutane bath; slow cooling is
performed by placing the vials in air in the top of storage
freezer, either a 135 C mechanical freezer or near the top of
a vapor-phase nitrogen-cooled freezer. More rapid cooling
can be obtained, if needed, using liquid ethane at
~ 183 C. However, this strategy may result in the develop-
ment of thermal stresses within the sample because the edges
will cool rapidly and the inside more slowly. Such stresses may
lead to cracking during storage or upon rewarming.
5. A dummy sample prepared under the same conditions as the
preserved samples with a thermocouple is used to monitor
cooling and warming. Make sure that the thermocouple has
been calibrated using a NIST standard. The dummy sample is
placed in close proximity to the samples to be rewarmed in the
storage freezer. Plug the dummy sample thermocouple into a
reader. Place the samples and dummy sample at the top of the
storage freezer and hold just inside the lid. When the tempera-
ture reaches 100 C (~1 min), initiate the rapid warming
step. Place both vials in a room temperature 30 % Me2SO bath
and swirl the vials manually until the glass begins to melt (the
VS starts to move). Open the vial and remove the VS with a
transfer pipette and add the first dilution solution. Start timer
for 15 min.
6. The samples can be stored in vapor-phase nitrogen but VS
solution fractures may be observed. Although the visual impact
of the fracturing is disturbing, we have rarely seen tissue frac-
tures, and on the rare occasions when tissue cracking has
occurred, it has been during the use of nitrogen dry shippers
for transport. If samples are stored in vapor-phase nitrogen,
the two-step rewarming should be performed as described in
Protocol 1.
7. Ice-free artery patches were rewarmed by placing the bag con-
taining the tissue in a 37 C water bath for 1 min and then washed
by transferring the contents of the bag to an AlloFlow continuous
gradient washout chamber with 1 L of room temperature iso-

tonic saline over 20 min. Ice-free cryopreserved valves were
rinsed briefly in ice-cold EC solution containing 200 mM
mannitol. Then three sequential rinses were performed for
15 min each with continuous careful agitation in ice-cold EC
solution with 200 mM mannitol, followed by EC solution alone
and finally DMEM. The valves were stored briefly on ice in
DMEM culture medium until implantation or evaluation.
8. We are presently evaluating the use of Lactated Ringers solu-
tion containing 5 % dextrose (LRD5) for washing because it is
commonly available in surgery suites where ice-free cryopre-
served tissues may be used. Human vessels generated in the
laboratory by culturing human smooth muscle cells on a poly-
glycolic acid scaffold [31] were employed. The tissues were
subjected to a total of 2, 4, 8, or 15 min room temperature
wash steps in Lactated Ringers solution containing 5 % dex-
trose to remove residual cryoprotectants. In control decellular-
ized tissues, the cellular material was removed using enzymes
and detergents leaving an extracellular matrix tube as previ-
ously described [31], which was refrigerated in phosphate-
buffered saline until tested. Trypan blue excluding human
umbilical vein endothelial cells was added to each tissue sample
under physiological cell culture conditions in non-tissue cul-
ture tubes (to minimize cell migration from the tissues). There
were no statistically significant viability differences comparing
endothelial cell seeded decellularized vessels and ice-free pre-
served tissue-engineered blood vessels that were not decellu-
larized after 2, 4, or 8 washes in cell culture (Fig. 8). The
endothelial cells formed confluent monolayers on all surfaces
of the washed tissues, regardless of the number of post-
rewarming washes (Fig. 9), demonstrating that any residual
cryoprotectants present were not cytotoxic. In parallel xeno-
graft tissue studies, we have also demonstrated that viability of
human peripheral blood mononuclear cells is not impacted by
culture with rewarmed tissues washed 5 with EC.
Acknowledgments
This work was supported in part by a US Public Health Grant from

the National Institute of Biomedical Imaging and Bioengineering,
Grant # R43 EB014614, to KGMB. The content is solely the
responsibility of the authors and does not necessarily represent the
official views of the National Institute of Biomedical Imaging and
Bioengineering or the National Institutes of Health. Commercial
use of protocols disclosed in this work is subject to several issued
US Patents (6,194,137; 6,596,531; 6,740,484; 7,157,222;
8,440,390) and International Patents (available upon request).
References
1. Angell WW, DeLanerolle P, Shumway NE approach to cryopreservation. Cryobiology 21:
(1973) Valve replacement: present status of 407426
homograft valves. Prog Cardiovasc Dis 15: 14. Rall WF, Fahy GM (1985) Ice-free cryopreser-
589622 vation of mouse embryos at 196 C by vitrifi-
2. Stelzer P, Jones DJ, Elkins RC (1989) Aortic cation. Nature 313:573575
root replacement with pulmonary autograft. 15. Rall WF (1987) Factors affecting the survival
Circulation 80:209213 of mouse embryos cryopreserved by vitrifica-
3. Angell WW, Oury JH, Lamberti JJ, Koziol J tion. Cryobiology 24:387402
(1989) Durability of the viable aortic allograft. 16. Fahy GM (1988) Vitrification. In: McGrath JJ,
J Thorac Cardiovasc Surg 98:4856 Diller KR (eds) Low temperature biotechnol-
4. OBrien MF, McGiffin DC, Stafford EG, ogy: emerging applications and engineering
Gardner MA, Pohlner PF, McLachlan GJ, Gall contributions. The American Society of
K, Smith S, Murphy E (1991) Allograft aortic Mechanical Engineers, New York, pp 113146
valve replacement: long-term comparative clin- 17. Fahy GM (1989) Vitrification as an approach
ical analysis of the viable cryopreserved and to organ cryopreservation: past, present, and
antibiotic 4C stored valves. J Card Surg 6: future. In: Smit Sibinga CT, Das PC, Meryman
534543 HT (eds) Cryopreservation and low tempera-
5. O'Brien MF, Stafford EG, Gardner MAH, ture biology in blood transfusion. Kluwer
Pohlner PF, Tesar PJ, Cochrane AD, Mau TK, Academic Publishers, Dordrecht, pp 255268
Gall KL, Smith SE (1995) Allograft aortic valve 18. Armitage WJ, Rich SJ (1990) Vitrification of
replacement: long-term follow-up. Ann Thorac organized tissues. Cryobiology 27:483491
Surg 60:565570 19. Pegg DE, Diaper MP (1990) Freezing versus vit-
6. Clarke DR, Campbell DN, Hayward AR, rification: basic principles. In: Smit Sibinga CT,
Bishop DA (1993) Degeneration of aortic Das PC, Meryman HT (eds) Cryopreservation
valve allografts in young recipients. J Thorac and low temperature biology in blood transfu-
Cardiovasc Surg 105:934942 sion, vol 24. Kluwer Academic Publishers,
7. Yankah AC, Alexi-Meskhishvili V, Weng Y, Dordrecht, pp 5569
Schorn K, Lange RE, Hetzer R (1995) 20. MacFarlane DR, Forsyth M, Barton CA (1992)
Accelerated degeneration of allografts in the first Vitrification and devitrification in cryopreser-
two years of life. Ann Thorac Surg 60:7177 vation. In: Steponkus PL (ed) Advances in low
8. Wolfinbarger L Jr, Hopkins RA (1989) Biology temperature biology, vol 1. JAI Press,
of heart valve cryopreservation. In: Hopkins Greenwich, CT, pp 221278
PA (ed) Cardiac reconstructions with allograft 21. Song YC, Khirabadi BS, Lightfoot F, Brockbank
valves. Springer, New York, pp 2136 KGM, Taylor MJ (2000) Vitreous cryopreser-
9. Mitchell RN, Jonas RA, Schoen FJ (1995) vation maintains the function of vascular grafts.
Structure-function correlations in cryopre- Nat Biotechnol 18:296299
served allograft cardiac valves. Ann Thorac 22. Rabin Y, Taylor MJ, Walsh JR, Baicu S, Steif PS
Surg 60:81088113 (2005) Cryomacroscopy of vitrification, Part I:
10. Mitchell RN, Jonas RA, Schoen FJ (1998) a prototype and experimental observations on
Pathology of explanted cryopreserved allograft the cocktails VS55 and DP6. Cell Preserv
heart valves: comparison with aortic valves Technol 3:169183
from orthotopic heart transplants. J Thorac 23. Baicu S, Taylor MJ, Chen Z, Rabin Y (2006)
Cardiovasc Surg 115:118127 Vitrification of carotid artery segments: an
11. Brockbank KGM, Lightfoot FG, Song YC, integrated study of thermophysical events
Taylor MJ (2000) Interstitial ice formation in and functional recovery towards scale-up for
cryopreserved homografts: a possible cause of clinical applications. Cell Preserv Technol 4:
tissue deterioration and calcification in vivo. J 236244
Heart Valve Dis 9:200206 24. Brockbank KGM, Taylor MJ (2007) Tissue
12. Taylor MJ, Song YC, Brockbank KGM (2004) preservation. In: Baust JG (ed) Advances in
Vitrification in tissue preservation: new devel- biopreservation. CRC Press, Boca Raton,
opments. In: Benson E, Fuller B, Lane N (eds) pp 157196
Life in the frozen state. Taylor and Francis 25. Song YC, An YH, Kang QK, Li C, Boggs JM,
Books, London, pp 603641 Chen Z, Taylor MJ, Brockbank KGM (2004)
13. Fahy GM, MacFarlane DR, Angell CA, Vitreous preservation of articular cartilage
Meryman HT (1984) Vitrification as an grafts. J Invest Surg 17:6570
26. Song YC, Lightfoot FG, Chen Z, Taylor MJ, preservation of extracellular matrix in cardiac
Brockbank KGM (2004) Vitreous preserva- tissues: implications for long-term graft dura-
tion of rabbit articular cartilage. Cell Preserv bility. Ann Thorac Surg 83:16411650
Technol 2:6774 39. Brockbank KGM, Wright GJ, Yao H, Greene
27. Brockbank KGM, MacLellan WR, Xie J, ED, Chen ZZ, Schenke-Layland K (2011)
Hamm-Alvarez SF, Chen ZZ, Schenke-Layland Allogeneic heart valve preservation allogeneic
K (2008) Quantitative second harmonic gen- heart valve storage above the glass transition at
eration imaging of cartilage damage. Cell 80 C. Ann Thorac Surg 91:18291835
Tissue Bank 9:299308 40. Hazekamp MG, Koolbergen DR, Braun J,
28. Brockbank KGM, Chen ZZ, Song YC (2010) Sugihara H, Cornelisse CJ, Goffin YA,
Vitrification of porcine articular cartilage. Huysmans HA (1995) In situ hybridization: a
Cryobiology 60:217221 new technique to determine the origin of
29. Song YC, Chen ZZ, Mukherjee N, Lightfoot fibroblasts in cryopreserved aortic homograft
FG, Taylor MJ, Brockbank KGM, Sambanis valve explants. J Thorac Cardiovasc Surg 110:
A (2005) Vitrification of tissue engineered 248257
pancreatic substitute. Transplant Proc 37: 41. Koolbergen DR, Hazekamp MG, de Heer E,
253255 Bruggemans EF, Huysmans HA, Dion RA,
30. Elder E, Chen Z, Ensley A, Nerem R, Bruijn JA (2002) The pathology of fresh and
Brockbank K, Song Y (2005) Enhanced tissue cryopreserved homograft heart valves: an anal-
strength in cryopreserved, collagen-based ysis of forty explanted homograft valves. J
blood vessel constructs. Transplant Proc 37: Thorac Cardiovasc Surg 124:689697
46254629 42. Armiger LC (1998) Postimplantation leaflet
31. Dahl S, Chen Z, Solan A, Lightfoot F, Li C, cellularity of valve allografts: are donor cells
Brockbank KGM, Niklason L, Song YC (2006) beneficial or detrimental? Ann Thorac Surg
Tissue engineered blood vessels. Tissue Eng 66(suppl):S233S235
12:291300 43. Navarro FB, Costa FD, Mulinari LA, Pimentel
32. Farooque TM, Chen ZZ, Schwartz Z, Wick GK, Roderjan JG, Vieira ED, Noronha L,
TM, Boyan BD, Brockbank KGM (2009) Miyague NI (2010) Evaluation of the biologi-
Protocol development for vitrification of cal behavior of decellularized pulmonary
tissue-engineered cartilage. Bioprocessing homografts: an experimental sheep model. Rev
(Williamsbg Va) 8:2936 Bras Cir Cardiovasc 25:377387
33. Khirabadi BS, Song YC, Brockbank KGM 44. da Costa FD, Costa AC, Prestes R, Domanski
(2004) Method of cryopreservation of tissues AC, Balbi EM, Ferreira AD, Lopes SV (2010)
by vitrification. United States Patent #6,740, The early and midterm function of decellular-
484, 2004 ized aortic valve allografts. Ann Thorac Surg
34. Khirabadi BS, Song YC, Brockbank KGM 90:18541860
(2007) Method of cryopreservation of tissues 45. Campbell LH, Brockbank KGM (2010)
by vitrification. United States Patent #7, Cryopreservation of porcine aortic heart valve
157,222 leaflet-derived myofibroblasts. Biopreserv
35. Brockbank KGM, Song YC (2004) Biobank 8:211217
Morphological analyses of ice-free and frozen 46. Ketchedjian A, Jones AL, Krueger P, Robinson
cryopreserved heart valve explants. J Heart E, Crouch K, Wolfinbarger L Jr, Hopkins R
Valve Dis 13:297301 (2005) Recellularization of decellularized
36. Schenke-Layland K, Madershahian N, Riemann allograft scaffolds in ovine great vessel recon-
I, Starcher B, Halbhuber KJ, Knig K, Stock structions. Ann Thorac Surg 79:888896
UA (2006) Impact of cryopreservation on 47. Ketchedjian A, Krueger P, Lukoff H, Robinson
extracellular matrix structures of heart valve E, Jones A, Crouch K, Wolfinbarger L, Hopkins
leaflets. Ann Thorac Surg 81:918926 RA (2005) Ovine panel reactive antibody assay
37. Schenke-Layland K, Riemann I, Damour O, of HLA responsivity to allograft bioengineered
Stock UA, Knig K (2006) Two-photon vascular scaffolds. J Thorac Surg 129:155166
microscopes and in vivo multiphoton tomo- 48. Linthurst Jones A, Moore M (2009)
graphs powerful diagnostic tools for tissue MATRACELLTM Decellularized allograft bio-
engineering and drug delivery. Adv Drug Deliv implants critical applications for cardiovascu-
Rev 58:878896 lar surgery a bio-implants brief. LifeNet Health
38. Schenke-Layland K, Xie J, Haydarkhan-Hagvall Nosfolk, VA, USA
S, Hamm-Alvarez SF, Stock UA, Brockbank 49. Lisy M, Pennecke J, Brockbank KGM, Fritze
KGM, MacLellan WR (2007) Optimized O, Schleicher M, Schenke-Layland K, Kaulitz
R, Riemann I, Weber CN, Braun J, Mueller (2012) Preclinical evaluation of ice-free cryo-
KE, Fend F, Scheunert T, Gruber AD, Albes preserved arteries: structural integrity and
JM, Ziemer G, Stock UA (2010) The perfor- hemocompatibility. Cells Tissues Organs 196:
mance of ice-free cryopreserved heart valve 262270
allografts in an orthotopic pulmonary sheep 55. Brockbank KGM, Song YC, Greene ED,
model. Biomaterials 31:53065311 Taylor MJ (2007) Quantitative analyses of vit-
50. Brockbank KGM, Schenke-Layland K, Greene rified autologous venous arterial bypass graft
ED, Chen Z, Fritze O, Schleicher M, Kaulitz explants. Cell Preserv Technol 5:6876
R, Riemann I, Fend F, Albes JM, Stock UA, 56. Rajani B, Mee RB, Ratliff NB (1998) Evidence
Lisy M (2012) Ice-free cryopreservation of for rejection of homograft cardiac valves in
heart valve allografts: better extracellular matrix infants. J Thorac Cardiovasc Surg 115:
preservation in vivo and preclinical results. Cell 111117
Tissue Bank 13:663671 57. Bechtel JF, Stierle U, Sievers HH (2008) Fifty-
51. Brockbank KGM (2013) Methods for ice-free two months mean follow up of decellularized
preservation of tissues. US Patent #8,440,390 SynerGraft-treated pulmonary valve allografts.
52. Thakrar RR, Patel VP, Hamilton G, Fuller J Heart Valve Dis 17:98104
BJ, Seifalian AM (2006) Vitreous cryo- 58. Brockbank KGM, Campbell LH, Chen Z,
preservation maintains the viscoelastic prop- Greene ED, Stock UA, Seifert M (2013)
erty of human vascular grafts. FASEB J Minimization of allogeneic tissue immunoge-
20:874881 nicity by cryopreservation. Presented at the
53. Huber AJ, Brockbank KGM, Aberle T, 22nd annual congress of the European
Schleicher M, Chen Z, Greene ED, Lisy M, Association of Tissue Banks, November
Stock UA (2012) Development of a simplified (abstract)
ice-free cryopreservation method for heart 59. Valente M, Faggian G, Billingham ME, Talentie
valves. Biopreserv Biobank 10:479484 E, Calabrese F, Casula R, Shumway NE, Thiene
54. Huber AJT, Brockbank KGM, Riemann I, G (1995) The aortic valve after heart transplan-
Schleicher M, Fritze O, Wendel H, Stock UA tation. Ann Thorac Surg 60(suppl):S135S140
Chapter 21
Cryopreservation of Plant Cell Lines

Heinz Martin Schumacher, Martina Westphal,
and Elke Heine-Dobbernack
Abstract
Plant cell cultures may consist of dedifferentiated cells as well as of cells showing embryogenic poten-
tial. They can be used for very different purposes in research and biotechnology as well as for plant
propagation. For such cell cultures, cryopreservation is the only means for long-term preservation.
Most of the different cryopreservation approaches, which are generally used for plant tissues, have also
been applied to plant cell cultures; they include slow freezing, vitrification, and encapsulation/dehy-
dration approaches. The controlled-rate slow freezing approach which is described here, however,
remains to be the gold standard for cell cultures. In this chapter, a standard cryopreservation procedure
is presented for plant cell cultures.
Key words Suspension culture, Embryogenic, Dedifferentiated, Callus culture, Slow freezing,
Controlled-rate freezing, Immobilization, Alginate
1 Introduction
Plant cell cultures have a wide range of different applications

including secondary metabolite production [1], phytopharming
[2], the propagation of plants by embryogenic cell cultures [3],
and as model systems in fundamental research [4, 5]. For all of
these applications, a reliable long-term preservation method is
essential. So far, cryopreservation is the only possibility to realize
long-term preservation of plant cell cultures. A standard protocol
based on a controlled-rate freezing approach has been developed
in 1980 by Withers and King [6]. In recent years several funda-
mentally different techniques have been developed for plant tis-
sues, and many of them have also been applied to cell suspension
cultures such as vitrification [7, 8] and encapsulation/dehydration
[911]. Most dedifferentiated cell lines, however, are still pre-
served using controlled-rate slow freezing [12, 13]. It has to be
pointed out that the described method in this chapter is only an
example, which has been successfully applied for the Nicotiana
423
424 Heinz Martin Schumacher et al.
tabacum BY2 cell line at the German Collection of Microorganisms

and Cell Cultures (DSMZ). Many parameters such as pre-culture
osmotic conditions; pre-culture duration, type, and concentration
of cryoprotectant; and the freezing program have to be adjusted
for each specific cell line. In Subheading 4 of this chapter, param-
eters applied for different cell lines are listed (see Notes 112).
Embryogenic cultures are normally more clumpy than dedifferen-
tiated cell lines and require longer times for pre-culturing and
incubation with cryoprotectants. On the other hand, they often
show a higher level of drought resistance. Both embryogenic and
dedifferentiated cell cultures can be preserved by controlled-rate
freezing [12]. Droplet-vitrification [14] and encapsulation/dehy-
dration [15] have been applied more often for embryogenic cul-
tures than for dedifferentiated cell lines [14, 16]. Since these
methods are technically completely different from controlled-rate
freezing, they are not described here. Vitrification has been only
rarely applied to cell cultures, and we refer to Reinhoud et al. [17]
for a simple and efficient vitrification method. In this chapter, we
describe a typical procedure for cryopreservation of plant cell cul-
tures using a controlled-rate freezing approach.
2 Materials
1. Normal/standard liquid culture/growth medium (prepared as

usual).
2. Calcium-free growth medium containing 3 % sodium alginate
(complete dissolution of alginate occurs only after autoclaving)
(see Note 4).
3. Calcium-free growth medium.
4. Growth medium containing 100 mM calcium chloride.
5. Growth medium containing 0.9 M sorbitol as pre-culture
osmotic (see Note 5) (prepared as usual).
6. Petri dishes containing normal solid growth medium.
7. Dimethyl sulfoxide (Me2SO) (see Note 7).
8. The following materials have to be wrapped in aluminum foil
and sterilized: culture jars (normally Erlenmeyer flask for rou-
tine cultivation on a gyratory shaker), three suction filtration
flasks (250300 mL), one Buchner funnel (60 mm diameter)
with Nylon net (100 m pore size), two Buchner funnels
(60 mm diameter) without Nylon net, flammable metal spatula
or spoon.
9. The following materials have to be purchased sterilized: 50 mL
Falcon tubes, 2 mL cryovials.
Cryopreservation of Plant Cell Lines 425
10. Pipette tips with widened opening. To widen the opening of a

pipette tip: cut using scissors or a scalpel. A self-made position-
ing device helps to standardize this (see Note 9).
11. Technical equipment: gyratory shaker, programmable cooling
device (e.g., a Planer K10 Planer plc; or for cooling rates of
1 C/min a Mr. Frosty from Nalgene).
3 Methods
3.1 Immobilization 1. Use cultures in the logarithmic growth phase (see Note 1).
and Pre-culture Harvest cells by filtration through a Nylon filter in a Buchner
funnel. Let the medium gently drain off and do not apply vac-
uum (see Note 2).
2. Wash cells twice in the Buchner funnel with calcium-free
growth medium.
3. Fill a volume of 10 mL cells in a sterile 50 mL Falcon tube
using a sterile metal spoon or spatula. Fill up to 30 mL
(see Note 3) with growth medium containing 3 % alginate
(see Note 4).
4. Close the Falcon tube and mix gently to suspend the cells.
5. Transfer the suspension carefully and dropwise into 250 mL of
a solution of growth medium containing 100 mM calcium
chloride in a 500 mL beaker. This should be done drop by
drop using a pipette tip with a widened opening while stirring
gently.
6. For polymerization: stir gently for 20 min.
7. Let the alginate beads settle down, pour off the medium con-
taining calcium, and wash the beads twice for ten minutes with
standard growth medium.
8. After washing, separate the beads from the medium by filtering
through a sterile Buchner funnel (without a Nylon net) (see
Note 2).
9. Transfer the beads into a 300 mL Erlenmeyer flask.
10. Fill up to a total volume of 100 mL normal liquid growth
medium (see Note 3).
11. Cultivate under normal growth conditions for 3 days.
12. Let the beads settle down and pour off the medium.
13. Fill up the Erlenmeyer flask containing the beads to a total
volume of 100 mL with growth medium containing 0.9 M
sorbitol (see Note 5).
14. Cultivate cells under routine conditions for 2 more days.
3.2 Cryoprotection 1. Cool the Erlenmeyer flask containing cells immobilized in algi-
nate beads to 4 C (while shaking).
2. Add 5 mL 100 % Me2SO to the Erlenmeyer flask containing
100 mL cell/beads suspension, resulting in a final concentra-
tion of 5 % Me2SO in the suspension (see Note 6). Work under
sterile conditions (see Note 7).
3. Incubate for a total duration of 60 min at 4 C (see Note 8).
3.3 Freezing 1. Separate the beads and medium by filtering through a Buchner
funnel (without Nylon net), and drain off the medium (see
Note 2).
2. Transfer 5 beads into a 2 mL cryovial using a sterilized/flamed
forceps (see Note 9).
3. Place the cryovials into the cooling device and cool from 4
down to 40 C using a cooling rate of 0.25 C/min (see
Note 10).
4. Hold the vials at a temperature of 40 C for 15 min (see Note 10).
5. Transfer the vials immediately into liquid nitrogen.
6. Store samples in liquid nitrogen.
3.4 Recovery 1. For thawing, transfer vials from liquid nitrogen immediately
into a water bath of 40 C (see Note 11).
2. Keep the samples at 40 C until completely thawed (~3 min).
3. Pour the thawed beads onto a sterile Petri dish (60 mm diam-
eter) containing normal solid growth medium (see Note 12).
4. Press the beads gently onto the medium. Make sure that not
more than 1/3 of the beads is submersed in the medium.
5. After 4 h, transfer the beads to another Petri dish with fresh
solid growth medium.
6. Within 26 weeks cells grow out of the beads and can be used
to reinitiate a new suspension culture.
4 Notes
1. In the initial and logarithmic phase of a cell culture, the cellular

water content is generally low and increases at the end of the
growth curve. Furthermore, cells are most vigorous in the log-
arithmic growth phase and best suited for freezing.
2. The BY2 cell line grows as a very fine suspension. Therefore,
for harvesting it has to be poured through a filter. For many
more clumpy cell cultures, it may be possible to let cells settle
down (incline the Erlenmeyer flask), after which the medium
can be poured off and the culture flask can be filled with the
fresh medium containing pre-culture osmotic.
3. In many cases, the cells-to-medium ratio is important for the

cryopreservation procedure and needs to be experimentally
determined. In any case, it is important to use the same ratio in
repeated experiments. This is typically determined by compar-
ing/measuring the volume of settled/immobilized cells (scv,
settled cell volume) versus the total volume of the suspension.
4. In the presented method, controlled-rate freezing is combined
with immobilization of cells in alginate. Immobilization is not
necessary but seems to stabilize the cells. Furthermore, immo-
bilization/encapsulation in alginate beads may slow down dif-
fusion processes during pre-culturing with osmotic and
cryoprotective agents.
5. The best substances to be used as pre-culture osmotic, its opti-
mum concentration, as well as the duration of pre-culturing all
have to be determined experimentally. The most frequently
used osmotic substances are sorbitol, mannitol, and sucrose,
with concentrations ranging from 0.3 M up to 1.2 M and pre-
culture durations from 2 to 6 days. For embryogenic cultures,
the pre-culturing period may be even longer.
6. Me2SO can be used as the only cryoprotective agent at concen-
trations ranging from 2.5 % to even 10 %. For many cultures,
we obtained good results when using 5 % Me2SO. Originally,
Withers and King [6] used the following mixtures for cryopro-
tection: DGS (1 M Me2SO, 1 M glycerol, 2 M sucrose; in
normal growth medium) and DPG (1 M Me2SO, 1 M glyc-
erol, 2 M proline; in normal growth medium). These mixtures
are filter sterilized. However, they are so viscous that moderate
warming is needed to allow proper filter sterilization. For cryo-
preservation, half a volume of such a cryopreservation solution
is mixed with an equal volume of pre-cultured cell suspension.
It is not recommended to use the mixtures described above in
combination with immobilized cells.
7. We use Me2SO under sterile conditions in the clean bench, and
do not filter sterilize this. Toxic oxidation products may accu-
mulate with time, and therefore only small batches/bottles of
Me2SO should be used. When stored at colder temperatures,
the substance may solidify. Me2SO may dissolve certain filter
materials, which are used for filter sterilization.
8. The cryoprotection procedure is generally carried out at cold
temperatures, even though in some cases there is no clear proof
of this having a beneficial effect. Apart from reduced toxicity of
Me2SO, a reduced diffusion rate at low temperatures may have
a positive effect.
9. When transferring/pipetting the (not immobilized) suspen-
sion into the cryovials, occasionally shake the culture jar to
make sure that cells remain suspended. Since plant cell cultures
are always more or less clumpy, use a pipette tip with widened
opening. Furthermore, pipetting should be performed quickly
to avoid settling of cells, which may block the pipette tip dur-
ing emptying.
10. For each specific cell line, the cooling program has to be
adjusted to its needs. In the older literature, cooling rates of
approximately 1 C/min have been reported. More recently,
cooling rates of 0.50.25 C/min have been shown to yield
better results for most cell lines [12]. This is in agreement with
our experiences. Hold-temperatures range from 35 to
40 C, for durations of 520 min.
11. During the thawing procedure, water may enter into the vials
and cause microbial contaminations. We therefore recommend
the use of a sterilized water bath and/or at least sterilized
water.
12. In most cases, regrowth on solid medium is better than
regrowth in liquid medium.
References
1. Tabata H (2004) Paclitaxel production by tobacco cell cultures. In Vitro Cell Dev Biol
plant cell culture technology. Adv Biochem Plant 45:750757
Eng Biotechnol 87:123 9. Shibli RA, Haagenson DM, Cunningham SM,
2. Hellwig S, Drossard J, Twyman RM, Fischer R Berg WK, Volenec JJ (2001) Cryopreservation
(2004) Plant cell cultures for the production of alfalfa (Medicago sativa L.) by encapsula-
of recombinant proteins. Nat Biotechnol tiondehydration. Plant Cell Rep 20:445450
11:14151422 10. Bachiri Y, Bajon C, Sauvant A, Gazeau C,
3. Ziv M (2005) Simple bioreactors for mass Morisset C (2000) Effect of osmotic stress tol-
propagation of plants. In: Hvoslef-Eide AK, erance of air-drying and cryopreservation of
Preil W (eds) Liquid culture systems for in- Arabidopsis thaliana suspension cells.
vitro plant propagation. Springer, Berlin, Protoplasma 214:227243
pp 7993 11. Ogawa Y, Suzuki H, Sakurai N, Akoi K,
4. Toshiyuki N, Seichiro H, Inz D (2010) Shibata D (2008) Cryopreservation and meta-
Tobacco BY2 cells. In: Lrz H, Widholm J bolic profiling analysis of Arabidopsis thaliana
(eds) Biotechnology in agriculture and for- T87 suspension cultures cells. Cryo Letters
estry. Springer, Berlin 29:427436
5. Kim KW, Bamba T, Harada K, Fukusaki E, 12. Keller EJR, Senula A, Hfer M, Heine-
Kobayashi A (2006) Time course of metabolic Dobbernack E, Schumacher HM (2013)
profiling in Arabidopsis thaliana cell cultures Cryopreservation of plant cells. In: Flickinger
after salt stress treatment. J Exp Bot MC (ed) Encyclopedia of industrial biotech-
58:415424 nology: bioprocess, bioseparation and cell
6. Withers L, King P (1980) A simple freezing technology. John Wiley & Sons, (2013), New
unit and routine cryopreservation method Jersey 114
for plant cell cultures. Cryo Letters 13. Lynch PT, Benson EE, Jones J, Cocking EC,
1:213220 Power JB, Davey MR (1994) Rice cell cryo-
7. Hao YJ, You CX, Deng XX (2002) Effects of preservation: the influence of culture methods
cryopreservation on developmental compe- and the embryogenic potential of cell suspen-
tency, cytological and molecular stability of sion on the post-thaw recovery. Plant Sci
citrus callus. Cryo Letters 23:2735 98:185192
8. Van Eck JV, Keen P (2009) Continued expres- 14. Fki L, Bouaziz N, Sahnoun N, Swennen R,
sion of plant-made vaccines following long- Dir N, Panis B (2011) Palm cryobanking.
term cryopreservation of antigen-expressing Cryo Letters 32:451462
15. Gonzlez-Benito ME, Martin C, Vidal JR regrowth, embryogenic competence and DNA
(2009) Cryopreservation of embryogenic cell content. Cryo Letters 29:409418
suspensions of the Spanish grapevine cultivars 17. Reinhoud PJ, Uragami S, Sakai A, van Iren F
Albarino and Tempranillo. Vitis (1995) Vitrification of plant cell suspen-
48:131136 sions. In: Day JG, McClellan MR (eds)
16. Mikula A, Olas M, Sliwinska E, Rybczynski JJ Cryopreservation and freeze-drying proto-
(2008) Cryopreservation by encapsulation of cols. Humana Press, Totowa, NJ,
Gentiana ssp.: cell suspensions maintains p 113120
Chapter 22
Writing Standard Operating Procedures (SOPs)

for Cryostorage Protocols: Using Shoot Meristem
Cryopreservation as an Example
Keith Harding and Erica E. Benson
Abstract
Standard operating procedures are a systematic way of making sure that biopreservation processes, tasks,
protocols, and operations are correctly and consistently performed. They are the basic documents of
biorepository quality management systems and are used in quality assurance, control, and improvement.
Methodologies for constructing workflows and writing standard operating procedures and work instruc-
tions are described using a plant cryopreservation protocol as an example. This chapter is pertinent to
other biopreservation sectors because how methods are written, interpreted, and implemented can affect
the quality of storage outcomes.
Key words Cryopreservation, Encapsulation/dehydration, Plant germplasm, Quality management

systems, Standard operating procedures
1 Introduction
The storage of authentic, pure (free from contaminating organisms),

and stable biological resources that conform to quality standards is
pertinent to all biopreservation sectors [1]. Biorepository and
biobank managers and beneficiaries of storage technologies are
therefore placing greater emphasis on the quality management of
biological resources [211]. Standard operating procedures (SOPs)
are the essential documents of quality management systems (QMS),
and their production from cryopreservation protocols is important
for successful storage and research outcomes. SOPs (see Note 1)
are an integral part of cryopreservation quality assurance (QA) and
quality control (QC), and they are especially relevant for large-
scale cryobanks holding high numbers of diverse samples that are
stored for long periods of time.
431
432 Keith Harding and Erica E. Benson
1.1 Scope, Aims, A prototype SOP for the cryostorage of shoot meristems has been
and Structure produced previously by the authors [12]. The aim of this chapter
is to explain how to write SOPs (see Note 2) for cryostorage meth-
ods using a validated, encapsulation/dehydration plant cryopreser-
vation protocol as an example [1315]. This method is selected on
the basis that (a) encapsulation is used to cryopreserve cells, tis-
sues, organs, and organisms representing diverse taxa, including
microbial, protist, plant, animal, and human [5, 1622] and (b) it
demonstrates how SOPs may be written for multistep storage pro-
tocols. This chapter differs in style and structure from those of
previous editions [23, 24] because it specifically concerns the
methodology for writing cryopreservation SOPs.
1.2 Standard To orientate readers unfamiliar with quality documents, the fol-
Operating Procedures lowing sections explain why SOPs are important for the standard-
Explained ization of biopreservation methods (see Note 2).
1.3 What Is an SOP? An SOP is a step-by-step, precisely written instruction that describes
how to undertake a task, procedure, process, or protocol. It is the
basic QMS document (see Note 1) and is used in a specific context
to assure different personnel always complete the same protocols
or procedures consistently and safely. Where deviations from SOPs
do occur, they are dealt with in a formal process that involves cor-
rective actions and quality improvement. SOPs are distinguished
from other methodological documents because they (a) are usually
written by one person who is the member of staff responsible for
the procedure; (b) are formatted in the style of a quality document;
(c) are developed, written, reviewed, revised, and improved by a
formal process that is overseen by a named quality manager who is
the authorizing signatory for the final version; (d) are controlled
documents, once finalized they cannot be changed other than by a
formal review process; (e) are developed, written, and used by
trained personnel who demonstrate an acceptable level of compe-
tency; (f) conform to safety and regulatory obligations; and (g)
include procedures for nonconforming work, and deviations from
SOPs are reported and acted upon.
1.4 Why Are SOPs Standard operating procedures are a systematic way [2] of making
Important for sure that biopreservation processes, tasks, protocols, and opera-
Biopreservation? tions are correctly, consistently, and uniformly performed (see Note 2).
They reduce the chances of an error occurring and help to assure
through QA/QC procedures and checks (see Note 1) that storage
protocols adhere to quality, risk management, and regulatory
guidelines. Reasons for the implementation of SOPs in bioreposi-
tories are explained as follows:
(a) Biorepository QMS: SOPs deliver quality management (see Note 1).
They systematically detail routine work and document the way
activities are performed in biorepository QMS. SOPs maintain
Standard Operating Procedures for Cryostorage 433
consistent quality outcomes as they provide the administrative,

technical, and analytical procedures, process chain, and storage
methods for a biorepositorys entire operations, including
facilities and the calibration and use of equipment (see Note 2).
(b) Safety: SOPs are approved and authorized procedures that com-
ply with health and safety regulations and risk and disaster
management, and they are used as reference documents fol-
lowing incidents and accidents.
(c) Training: SOPs are used in training and continuing profes-
sional development.
(d) Consistency: SOPs minimize variation and enable the consistent
implementation of a procedure by reducing the chances of
miscommunication and misinterpretation. They assure that a
task is performed correctly, every time, by different people,
even when staff leave or retire or when there are transitory
changes in personnel.
(e) Deviations from norms: SOPs can be used as checklists to ensure
that work is followed correctly; they help to identify and rectify
any deviations from normal operations.
(f) Standards: SOPs may be used to create, classify, and satisfy
attainable quality standards; they provide a documentary
record of the procedures needed to assure standards.
(g) Quality assurance, control, and improvement: SOPs support all
aspects of quality management; their regular review drives up
standards and facilitates quality improvement.
(h) Auditing: SOPs are used by inspectors during quality, regula-
tory, and risk auditing.
Biopreservation SOPs are written by the individuals and teams
who actually undertake the work, and capturing their cumulative
expertise is critical for SOP development, management, and qual-
ity improvement [2].
1.5 Developing The development of SOPs is an evidence-based process that considers

an SOP the input of all personnel involved in assuring the success and safety
of a procedure. Feasibility studies determine the practicality of a
method as a candidate SOP, for example, with regard to personnel
health and safety, risk management, technical simplicity, cost/benefit
analyses, effectiveness, and efficiency on scale-up. Procedural SOPs
are different from other types as they rely on the accuracy of scien-
tific and technical detail which is verified by using original research
papers to avoid ambiguities or errors being generated through sec-
ondary citations. SOPs are internally and/or externally validated;
this process assures a procedure meets its desired targets and satisfies
quality standards [14, 15, 25]. In the case of plant germplasm cryo-
preservation, this may be the setting of standards for post-storage
acceptance/rejection criteria related to viability (e.g., 85 % for seeds),
recovery (e.g., 40 % full plant regeneration from cryopreserved

shoot meristems), genetic stability, and trueness to type [18, 26,
27]. Developing, validating, and writing SOPs is achieved through
teamwork [2] and the witnessing of technical procedures. Providing
each member of staff who uses the SOP with a draft version to test
helps ensure that the final version fulfills its purpose; internal and
external reviewers, technical advisers, and IT specialists facilitate
SOP writing and document control.
1.6 What Does The inclusion of different elements and the sequence of content
an SOP Include? will vary dependent upon the type and purpose (e.g., administra-
tive, safety, or procedural) of the SOP and the institutional policies
and the procedures used to create and manage quality documents.
Limiting superfluous information and checking for appropriate
levels of detail are considerations when drafting SOPs, for which
examples of various elements and their content are listed below.
SOP element Description of element and examples of
content.
Title Unique name conveying SOP purpose.
Type For example, administration, procedural,
regulatory, facilities management, document
control, risk management, IT, storage, and
material handling.
Number A unique number, code, or ID for easy refer-
ence and traceability.
Dates Of approval, issue (when SOP becomes
effective), latest version.
Version reference Tracking system or version number which
concurs with the issue date to make sure that
the most recently authorized version is used.
Supersedes Version (as a code) that the latest SOP
supersedes.
Affiliations Department, division, facility, laboratory,
and institution.
Author Name of SOP author.
Approved by Name of person authorized to approve the
SOP.
Personnel Names of all other individuals that are autho-
rized and responsible for writing and admin-
istering the SOP and competent (trained) in
its use.
Document control Named, authorized individuals and proce-
dures used to manage documents.
Purpose Description of protocol or method
function.
Scope Range, application, and scope of the SOP.
Related documents Lists of workflows and other SOPs that are
connected to the SOP.
Risks/safety Safety requirements, risk assessments, and

instructions that all personnel authorized to
use the SOP understand the risk assessment,
are able to implement its procedures, and
have read the health and safety manual.
Detailed description of health and safety
measures associated with SOP with cross-
reference to relevant risk and safety regula-
tions, incident and accident mitigation and
reporting, and disaster response plans.
Safety equipment Description of the personal protective equip-
ment (PPE) and safety equipment (e.g.,
cryogenic safety, low-level O2 sensors).
Equipment List of equipment required to perform SOPs,
links to information concerning model, pur-
chase date, serial number, inventory num-
ber, name of manufacturer, dates for
maintenance, and servicing/safety checks
(e.g., pressurized cryogenic equipment, liq-
uid nitrogen [LN] storage Dewars, pro-
grammable freezers).
Materials Lists of chemicals, reagents, cryoprotectants,
media, materials, and supplies needed to per-
form the procedure. The SOP may also
include a checklist or links to information
about suppliers, catalogue numbers, lot num-
bers, and expiration dates for materials used.
Cross-references Reference citations; related SOPs; work-
flows; work instructions (WI); regulatory,
safety, policy, and administrative documents;
operations; maintenance; quality assurance;
and control procedures.
Procedure Step-by-step instructions written in suffi-
cient detail to ensure that the SOP can be
repeated and consistently reproduced by
authorized personnel. Instructions may
include sequential steps to be followed, time
allowed for each step, critical information
(e.g., temperature at which steps are per-
formed, environmental control parameters,
asepsis, and containment regimes and best
practices).
Workflows Inserted diagrams that map the SOP.
Training Staff training records, persons delivering
training, type of training, methods used to
test competence, training evidence log, com-
petency assessments, and personal profes-
sional development records.
Quality assurance/
quality control Detailed descriptions of QA/QC measures
and checks connected to the SOP, including
cross-reference to internal/external QA/
QC measures (see Note 1).
Regulatory issues Information about pertinent regulatory
information, e.g., related to risks and mate-
rial transfer agreements.
Revision history Date revised, authors, document control-
lers, and summary of revisions.
Bibliography List of wider supporting references, guide-
lines, and best practices.
Appendix/annex Any other information in support of the
SOP: glossaries, abbreviations, keywords,
calculations, and notes.
An SOP is not a stand-alone document, and it can include
links to supporting documents: health and safety instructions, risk
assessments, QA/QC procedures, policies, regulations, best prac-
tices, guidelines, and standards. To avoid an SOP becoming com-
plex and difficult to follow, information is cross-referenced to
other SOPs or placed in annexes and appendices. SOPs are col-
lated in hardcopy quality management manuals (see Note 2) and
QMS-dedicated webpages, databases, or Laboratory Information
Management Systems (LIMS).
1.7 Different Types Creating SOPs for a biorepositorys entire QMS is a considerable
of SOPs task and requires the production of various types of document (see
Note 2). Barnes et al. [2] identified the main categories of SOP
applicable for biorepositories as follows: management, regulatory,
policies, administration procedures, personnel, recruitment and
training, safety, technical, facilities/equipment, operations, pro-
cesses, QA/QC, records/documentation control, biological mate-
rials acquisition, storage, handling, release, distribution, and
material transfer documentation. Although the type, content, and
style of individual SOPs vary, the basic elements are the same
(see Subheading 3.5.1).
1.8 Work Work instructions are often considered to be a suborder of an SOP

Instructions (WIs) as they provide extra detail in circumstances where the task or pro-
cedure of the main SOP is complex or involves many different
steps and interconnected processes. In which case, the SOP
becomes a higher level document that gives information about
responsible personnel and what will be done, whereas the WI com-
prises technical, step-by-step instructions. Deciding when to use
WIs is rationalized on the basis that (a) detailed instructions are
necessary to ensure that complicated or numerous sub-procedures
involving many parallel, preparative, or sequential steps are
performed correctly; (b) detailed instructions are necessary to

describe adjunct processes connected to the main SOP; and (c)
there is a need to refine, adapt, or optimize individual steps of a
basic biopreservation protocol, for example, as might be the case
for recalcitrant species/genotypes and diverse collections of genetic
resources [3, 6].
1.9 Systemization Flowcharts or flow diagrams are graphical tools that are used to
of Procedures, visualize procedures and decision-making (see Note 3); they depict
Flowcharts, complex and multistep tasks and are often considered as QC tools
and Workflows (see Note 1). When flowcharts are used in SOPs, they are called
workflows (WF) as they systemize the individual steps of a pro-
cedure in a logical sequence and transform written, step-by-step
instructions into a visual map. Workflows are used in biobank pro-
cess chains to enable protocol validation and enhance the quality of
storage and sample handling outcomes [10]. The sequence of
operations in a workflow is represented as instructions or options
depicted by shapes such as rectangles or diamonds linked by arrows
that are defined by flowchart conventions (see Subheading 3).
1.10 SOP Formats Quality managers choose the best way to format SOPs; their objec-
tive is to produce an easy-to-follow document that gives unam-
biguous instructions that will be understandable to all staff that are
trained in the use of the method. A biorepositorys SOP portfolio
will thus contain all the appropriate information for the safe and
effective completion of tasks without superfluous or confusing
content. This is not a straightforward process and several factors
may help to determine the most suitable SOP format to use:
(a) Purpose of SOP (administrative, regulatory, process, procedure,
risk, QA/QC)
(b) Number of steps, sub-steps, and subroutines
(c) Number and complexity of preparative/associated tasks on
which the SOP depends
(d) Type and number of decisions that need to be made within
the SOP
(e) Timelines and timed events of each of the individual steps
(f) Different types of manipulations, equipment, and facilities
(g) Mapping points and locations for different SOP stages
Routine procedures that are short, require few decisions, use
simple facilities, and limited equipment may be best written using
a simple step-by-step instructional format. Procedures, involving
more than ten steps, with few decisions may be better written in
sequential steps with clear subheadings. More complicated, longer
procedures that comprise different stages, subroutines, and supple-
mentary manipulations or preparative stages may be better broken
down into a series of shorter SOPs that cross-reference one another

and are supported by workflows and WIs appended to the main
SOP document. Calculations, results, and assessment tables may
also be included as separate SOPs or WIs. Images demonstrating a
manipulation (e.g., meristem excision) or visual performance indi-
cators such as morphogenetic responses (see Note 4) can be
appended to the main SOP or videoed and uploaded onto a QMS-
dedicated database, LIMS, or website. Producing generic SOP
templates, forms, and tables makes the process of developing, writ-
ing, and revising multiple SOPs more efficient, particularly across
biorepository consortia [2].
1.11 Writing an SOP: Standard operating procedures need to give unambiguous, clear,
Language, Style, and concise instructions and be understandable to all the personnel
and Level of Detail who are trained to use them; they are written in plain language
using short steps and sentences that are easy to understand and
check off. A definitive, active (instructive) voice and the use of
present tense verbs are preferred (see Note 5). The terms you
and should are not normally used and wordy, redundant, and
repetitive phrases are avoided. The information and instructions
provided in an SOP need to be clear, concise, and explicit so as to
remove any doubt or potential for misinterpretation. Procedural
SOPs are written using short, direct sentences so that the operators
can understand and perform the steps easily without the need for
further clarification. Deciding on the appropriate level of detail to
include in an SOP can be a difficult task which should not be
underestimated; precise language and clear instructions are neces-
sary, particularly where differences in comprehension between
workers may occur, such as in the use of multilingual SOPs.
Conversely, omitting critical details in essential steps can lead to
variation among workers and compromise performance indicators.
Several drafts are required to produce an SOP, and quality manag-
ers need to balance giving too much detail with providing suffi-
cient information to assure that everyone performs the procedure
safely and correctly, in the same way, every time.
1.12 Combining In the case of multicomponent procedures that involve intercon-

SOPs, Work nected and different preparative manipulations, it may be more
Instructions, appropriate to combine SOPs, WIs, and workflows; the main SOP
and Workflows then becomes the overarching quality document which is supple-
mented by one or more workflows. The weakness of using this
format is that the level of detail is limited and multiple illustrative
steps can be hard to follow, in which case different types of work-
flows are used to explain how, when, and where decisions are
required and indicate where supporting SOPs and WIs are used
(see Note 3).
1.13 Managing Replica copies of individual, authorized SOPs and their risk assess-
and Improving SOPs ments are placed in the area where members of staff routinely
undertake the work and procedures. SOPs, WIs, and workflows are
collated in a master file, usually the quality manual or handbook
(see Note 2), which comprise hard and electronic copies that are
stored in secure physical and virtual locations. Documents are
overseen by the quality manager using the institutional document
control policy which contains all the information about how SOPs,
WIs, and workflows are coded, controlled, backed up, archived,
filed, and managed, together with descriptions of procedures about
annotations, corrections, and how to report and manage devia-
tions from controlled documents. Periodic revisions improve SOPs
and WIs, they can involve external and internal reviewers or audi-
tors, and they are informed by new research findings, technological
advances, and changes in risk management.
1.14 SOP Training The provision of training in SOP writing, use, and management
complements staff development and encourages quality improve-
ment, guidance, and mentoring is necessary to avoid staff writing
suboptimal SOPs that could cause inconsistencies in work resulting
in poor storage outcomes. However, it is important to recognize
that even a well-written SOP is not a substitute for more compre-
hensive training that explains to personnel the significance of bio-
preservation standards, QA, and QC and why and how procedures
need to be correctly performed in a QMS.
2 Materials
2.1 Materials 1. Validated protocols on which the SOP and WIs will be based
Needed to (see Note 6).
Write an SOP 2. Health and safety/risk assessments, regulatory guidelines,
other supporting documents.
3. Information/IT management system (e.g., webpage, LIMS,
database).
4. Graphic tools and software for producing workflows (Fig. 1).
5. A document control SOP.
2.2 Equipment 1. Dewar (1 L).

for Cryopreservation 2. Nunc cryovials (2 mL).
Procedures Used
3. CryoCanes.
in SOP and WI
Templates 4. Pasteur pipettes.
5. Petri dishes (90 and 50 mm).
6. Sterile filter papers (90 and 50 mm).
7. Syringes fitted with hypodermic needles (10 mL).
8. Forceps.
9. Scissors.
10. Scalpels.
11. Storage cryotank.
12. Plastic Pastette (3 mL).
13. Mesh sieve (m).
14. Reciprocal shaker.
15. Binocular dissecting microscope.
Fig. 1 Examples of commonly used workflow shapes and conventions. Generic shapes are based on those
provided by commercial, flow diagram software packages and the basic flowchart shapes available in
Microsoft Office Word and PowerPoint (see Note 3)
2.3 Materials 1. Liquid nitrogen.

Needed 2. Liquid RIB medium.
for Cryopreservation
3. Liquid RIB culture (pregrowth) medium containing 0.75 M
Procedures Used
sucrose.
in SOP and WI
Templates 4. Calcium-free, liquid RIB culture medium containing 3 %
(w/v) sodium alginate (SIGMAR 2 % low viscosity, sodium salt
derived from sea kelp).
5. Liquid RIB culture medium containing 100 mM CaCl2.
6. Solid RIB recovery medium.
3 Methods
This section demonstrates how to (a) format workflows for a cryo-

preservation SOP, (b) write a procedural SOP based on an encap-
sulation/dehydration protocol, and (c) write a WI for a preparative
(meristem excision) cryopreservation procedure. It is usual for an
institution or biorepository network to produce sets of generic
templates for each type of document (SOPs, WIs, workflows);
these are then adapted for each individual procedure. The rationale
for using different combinations of SOPs and WIs is considered in
the following sections. The authors aim to describe basic approaches
which may be adapted by individual biorepositories to meet the
needs of local QMS, personnel, and resources.
3.1 Writing Technical It is assumed that technical, procedural SOPs will have been properly
SOPs: Research, validated by internal and/or external procedures [3, 11, 14, 15,
Validation, Quality 25]. Cryopreservation research identifies the critical point factors
Assurance, [4, 15] that could affect storage outcomes and quality standards.
and Control The exemplar SOP described below is based on an encapsulation/
dehydration protocol that has been optimized using differential
scanning calorimetry (DSC) to profile critical thermal events
(nucleation, melts, and glass transitions Tgs) during cryopreser-
vation [28, 29]. As glass stability is crucial for meristem survival,
this evidence-based development of a cryopreservation SOP con-
firmed the existence of the vitrified state during cooling and
rewarming [2830]. Research literature [20, 21, 31], technology
transfers, publications produced by collaborating biorepositories,
reviews [2, 10, 11, 16, 17, 19], and post-storage stability assess-
ments [18, 26, 27] can all contribute to SOP validation and
the development of biopreservation QA measures and QC tests
(see Note 1).
3.2 Designing A workflow (see Note 3) is an easy-to-follow, navigational tool

and Formatting that guides an operator through the tasks and processes that are
Workflows associated with SOPs and WIs. Higher-order workflows map the
for Cryopreservation full operations of an institutional QMS, and they make explicit the
SOPs connectivity between different biorepository operations, such as
management, administrative, technical, safety, and regulatory.
Workflows orientate operators as they progress along process
chains, they are used as task checklists to ensure that a procedure
is followed accurately and in correct sequence, and they can iden-
tify and help to correct deviations from documented procedures.
Workflows may thus be considered as QC tools (see Note 1). SOPs
are supported by various types of customized workflows, which
can also be used in training; they are sometimes considered as
quality management elements [10] because the methods they
describe can affect biopreservation outcomes (see Note 1).
Interoperability and consistency in the use of workflows are upheld
by using shape conventions (Fig. 1) for different processes and
procedures. There are a numerous software programs available for
constructing workflows, although the conventions can vary in
shape, style, and complexity. ISO 5807, for information process-
ing, describes documentation symbols and conventions for data,
program, and system flowcharts, program network charts, and
system resources charts.
A demonstration of flowchart shape conventions that may be
used in cryopreservation workflows is illustrated in Figs. 2 and 3
which chart the sequence and complexity of various cryopreserva-
tion procedures. For example, curved-base rectangles (Fig. 1) are
used to depict single and multiple documents or records. Noting
that documents represent published or non-published factual
information, whereas records are registered information or
recorded evidence that provide permanent, traceable points of ref-
erence, it is thus important to recognize that conventions for
records and documents convey a distinct meaning. Specifically
records are systematically archived and used in regulatory and
institutional inspections and audits, and they have an important
function in both quality and risk management. Another example is
data storage, which is depicted by the drum symbol (Fig. 1) and is
usually labeled in flowchart conventions as a magnetic disk. Flow
diagram style and format can vary as they are customized for dif-
ferent sectors (healthcare, computing, and manufacturing). As yet
there is no generally accepted workflow convention that is specifi-
cally attributed to biopreservation, biorepository, genebank, or
biobank processes.
3.3 Linear Workflows Figure 2 is a linear workflow that maps cryopreservation SOPs and
and Mapping Process WIs for each preparative, principal, and associated procedure that
Chains is involved in shoot meristem encapsulation, dehydration, cryo-
preservation, and recovery. The graphic guides the operator
Fig. 2 Example of a linear workflow for shoot meristem encapsulation/dehydration cryopreservation and asso-
ciated and preparative procedures. Abbreviations: LN = liquid nitrogen; RIB = Ribes medium; SOP = standard
operating procedure; WI = work instruction: numerical codes are denoted for individual SOPs (14), multiple
SOPs (n = x) representing different options for post-storage stability evaluations or WIs (16). The continual
process of quality assurance, control, and improvement is demonstrated as a double arrow connecting SOPs
and WIs (from SOP 1 to SOP n = x)
Fig. 3 Example of a systems workflow for shoot meristem cryopreservation. Abbreviations: LN = liquid nitro-
gen; RIB = Ribes medium; QA/QC = quality assurance/control, SOP = standard operating procedure; WI = work
instruction; numerical codes are denoted as individual SOPs (14) or WIs (16) or multiple SOPs (n = x) repre-
senting different options for post-storage stability evaluations. Duration of tasks and procedures (indicative
operational times may vary) is depicted as numbers in circles, arrow connectors indicate interconnected
workflows, and vertical lines and boxes identify the facilities and locations where work will be carried out
through a process chain that comprises multiple and interrelated

tasks, making it easier to progress through each step by identifying
points at which information is recorded or where decisions need to
be made.
The cryopreservation of in vitro plant germplasm involves sev-
eral different, interconnected processes; these may be combined in
one, several, or multiple SOPs and WIs. Figure 2 demonstrates
various options for delineating and creating cryopreservation
workflows, SOPs, and WIs. Indicative content of the different
types of SOPs and WIs that might contribute to a cryopreservation
workflow is provided in the examples below.
Cryopreservation SOP. The main steps of the encapsulation/dehy-
dration cryopreservation protocol starting with alginate encapsula-
tion and ending with in vitro meristem recovery are provided in
SOP 3 which connects to SOP 4 (Fig. 2). The method for micro-
propagating in vitro plants used as the source material for cryo-
preservation are placed in SOP 1, which also includes the
preparation of basic Ribes (RIB) medium [14].
Preparative WIs. The preparative stages of the cryopreservation
protocol (SOP 3) are described in shorter WIs (Fig. 2) that includes
the preparation of cryopreservation media and solutions. They are
WI I (excision of apical/nodal shoots), WI 2 (sucrose acclimation),
WI 3 (meristem excision/dissection), WI 4 (preparation of algi-
nate solution), WI 5 (preparation of calcium chloride solution),
and WI 6 (preparation of sucrose medium for osmotic
dehydration).
QA/QC SOP. SOP 2 (Fig. 2) provides an example of where a QA/
QC checkpoint (see Note 1) may be placed in an SOP. In this
case, it is checking bead moisture content (MC) with a target of
2530 % residual MC (fresh weight basis) after evaporative desic-
cation. The purpose of this QC procedure is to verify that the criti-
cal MC (equivalent to ~0.4 g H2Og1 dry weight measured using
DSC) is achieved in order to assure the stability of the vitrified state
during bead cooling and rewarming [5, 2830].
Cryobank risk management SOP. As connected to SOP 3 (Fig. 2),
this represents where an SOP will need to be supported by a cryo-
bank risk management protocol which describes the best practices
involved in the addition, storage, and removal of cryovials from the
cryotank to avoid inadvertent rewarming above critical, glass tran-
sition temperatures (Tgs). Risk management SOPs provide details
of LN replenishment, low-level LN surveillance/alarms, and
health, safety, and risk management procedures for the use of LN,
O2 monitoring, and cryogenic equipment. In addition, SOP 3 may
also cross-reference the QA measures and QC tests that need to be
in place (see Note 1) to assure cryobank safety.
Post-storage recovery SOP. SOP 4 (Fig. 2) provides an example of
the methods involved in monitoring and recording the post-
storage recovery of cryopreserved shoot meristems. This SOP will
cross-reference with best practices and standards that are used to
assure (QA) the sustained regrowth of surviving shoot meristems.
These can include visual indicators (images/descriptors) that are
used in QC checks to assess stress symptoms (bleaching, browning,
necrosis) and undesirable modes (callus, adventitious shoots) of
morphogenetic development that predispose plants recovered
from cryopreserved meristems to somaclonal variation [4, 27, 31].
Post-storage stability evaluations SOPs. SOP n = x (Fig. 2) indicates
various options, approaches, and methods that may be transformed
into SOPs that are used to assess the morphogenetic, genetic, and
epigenetic stability and trueness to type of plants recovered from
cryopreserved meristems [26, 27]. These may also be considered

as SOPs for QC tests (see Note 1).
The advantage of delineating the main cryopreservation pro-
cedures and their associated operations into individual SOPs and
the preparative steps into separate WIs (WI 1 to WI 6) is that
each stage may be adapted and refined while the main cryopreser-
vation method (SOP 3) remains the same. This flexible approach
(see Subheadings 3.7 and 3.8) is more practical when optimizing
cryopreservation protocols for genetically diverse and recalcitrant
genotypes. It can also facilitate the refinement of cryopreserva-
tion protocols for large-scale genebanks holding thousands of
accessions representing cultivated and wild species, landraces,
and cultivars.
3.4 Systems Systems workflows (Fig. 3) give extra information about time lines,
Workflows: Logistics, duration of critical tasks, and the locations where the work
Tasks, and Timelines described in SOPs and WIs will be undertaken (see Note 3). Linear
workflows (Fig. 2), although useful, do not always reveal interac-
tions across procedures, whereas systems flow diagrams have the
advantages of incorporating location, time, and other organiza-
tional attributes, such as points at which QA/QC checks are
required (Fig. 3). They make more explicit where personnel,
resources, and equipment are needed to accomplish different bio-
preservation tasks, and they indicate how long each step will take.
The systems approach is especially pertinent for large-scale biore-
positories that hold thousands of accessions and have dispersed
facilities, manage logistics, and operations involving large work-
forces, multiple process chains, and extended storage timelines.
Linear workflows (Fig. 2) may not reveal how processes and
personnel interact to accomplish tasks and achieve end goals.
Whereas, systems workflows (Fig. 3) identify where quality, health,
safety, and risk management needs to be in place, as well as the
resources, personnel, and facilities that will be required to under-
take the SOPs and WIs. Scheduling the logistics of different cryo-
preservation stages (Fig. 2) can be complex, particularly if they
include or are associated with several critical QA/QC checkpoints,
for example, SOP 1 (endophytic contamination assessments), WI 3
(competency in meristem excision), SOP 2 (critical MC test); SOP
3 (LN level, cryotank temperature, cryovial labeling, good cryo-
genic practices), SOP 4 (avoidance of adventitious shoots, callus,
somaclonal variation), and SOP n = x (trueness-to-type, genetic/
epigenetic stability assessments). Systems workflows thus have the
added advantage that they can be adapted for different facilities
and institutions and consider local as well as dispersed work envi-
ronments. Workflows are updated in line with QA/QC, SOP, and
WI revisions (Fig. 2) and changes in risk management, regulations,
facilities, and personnel.
3.5 Writing an SOP It is out with the scope of this chapter to describe all the SOPs and
for a Routine Storage WIs related to the cryopreservation of plant shoot meristems
Protocol described in the workflows (Figs. 2 and 3); therefore, an example
of SOP and WI is provided so that it may be used to help custom-
ize the specific requirements of individual biorepositories. The fol-
lowing is an exemplar of a technical procedure SOP which is based
on the encapsulation/dehydration protocol of Fabre and
Dereuddre et al. [13]. This has been adapted for the cryopreserva-
tion of sucrose-acclimated shoot meristems of Ribes nigrum by
Johnston et al. [31]. Thermal analysis using DSC was applied to
optimize cryoprotectant regimes [2830], and the protocol has
been externally validated [14, 15]. SOPs can also be based on pub-
lished best practices [4, 5, 9], adapted from generic SOPs provided
by an accredited body (see Note 6) or produced by a collaborating
biobank consortium [2].
3.5.1 Example Indicative content of a procedural, biopreservation SOP is

of a Cryopreservation SOP described with explanations of element descriptions and suggested
content (italicized) as follows:
Title Cryopreservation of R. nigrum using encapsula-
tion/dehydration.
Type Procedural SOP.
Number SOP CP 3.00 (identifier, CP = notation for a
cryopreservation SOP).
Dates 1.1.2014 (date written) 10.1.2014 (date
approved).
Version reference SOP CP 3.00; 11.1. 2014 (date issued/effec-
tive date).
Supersedes Not applicable (as a new SOP).
Affiliations Plant cryopreservation laboratory (place
where work is done).
Author Author (name of person responsible).
Approved by Quality manager (name of person responsible).
Personnel Trained person(s) (names of personnel
responsible).
Document control Document controller (name of person
responsible).
Purpose This SOP describes (a) the procedure for
cryopreserving encapsulated-dehydrated
shoot-tip meristems of R. nigrum excised
from sucrose-acclimated, nodal stem seg-
ments and (b) procedures for rewarming
and recovering cryopreserved encapsulated-
dehydrated shoot tips.
Scope Applicable to all cultivars of R. nigrum, with
the exception of cryopreservation recalci-
trant genotypes.
Related Documents CP SOPs 1, 2, 4, n = x; CP WIs 16 (see Figs.

2 and 3). Best practices for in vitro and asep-
tic techniques [4].
Risks/safety Confirm with lab manager/health and
safety officer an understanding of all risk
assessments for safe use of LN/cryogenic
equipment and associated cryogenic/non-
cryogenic steps and procedures (cross-refer-
ence to SOP risk assessments health and safety
manual).
Safety equipment Lab coat, cryogenic safety equipment gloves,
aprons, glasses, face shields, low-level oxy-
gen sensor and alarm, and LN safety signs.
Equipment 1 L Dewar, 2.0 mL Nunc cryovials,
CryoCanes, Pasteur pipettes, 90 and 50 mm
Petri dishes, 90 and 50 mm sterile filter
papers, forceps, scissors, scalpels, storage
cryotank, 3 mL plastic Pastette, sieve, and
reciprocal shaker.
Materials LN, liquid RIB culture (pregrowth) medium
containing 0.75 M sucrose, aliquoted
(20 mL) into 100 mL conical flasks; cal-
cium-free liquid RIB culture medium con-
taining 3 % (w/v) sodium alginate (SIGMA
2 % low viscosity, sodium salt derived from
sea kelp) dispensed as 20 mL aliquots into
50 mL bottles; liquid RIB culture medium
containing 100 mM CaCl2 dispensed as
30 mL aliquots into conical flasks or bea-
kers; and solid RIB recovery medium in
50 mm Petri dishes.
Cross-references Original encapsulation/dehydration
method by Fabre and Dereuddre [13];
Ribes spp. method validation by Reed et al.
[14, 15]; and in vitro conservation and
cryopreservation best practices by Benson
et al. [4].
Procedure Based on the method of Johnston et al. [31].
Use good aseptic technique throughout all
in vitro manipulations.
Calcium-alginate encapsulation
1. Transfer shoot-tip meristems to 3 % (w/v) alginate solution,
swirl the vial to immerse tissues, and take care not to form air
bubbles.
2. Pipette 2 mL of alginate solution with 58 meristems in a
3 mL plastic Pastette, hold the Pastette vertically over 0.1 M
CaCl2 solution, and dispense the alginate in droplets; uniform

spherical beads form on contact with CaCl2.
3. Polymerize beads for 15 min, and strain through a sieve to
remove the calcium solution.
Dehydration-desiccation
4. Using forceps, transfer beads with encapsulated shoot tips
to a 100 mL conical flask containing 25 mL 0.75 M sucrose
(final level) made up in liquid RIB medium.
5. Put the flasks on an orbital shaker, and mix for 1622 h
(overnight) in the culture room.
6. The next morning, pour the beads through a sieve and blot dry
on sterile filter papers.
7. Using forceps, transfer the beads to the base of a 90 mm sterile
Petri dish and evenly distribute across the dish, and make sure
the beads do not touch one another.
8. Place the Petri dish at the back of a horizontal laminar airflow
hood, desiccate for 4 h.
Quality control
9. Record the temperature and relative humidity (RH) of the flow
bench at the beginning and end of the 4 h desiccation period.
10. To ensure cryogenic glass stability, after evaporative desicca-
tion target residual MC is ~25 % (fresh weight basis); equiva-
lent to 0.350.4 g H2Og1 dry weight.
Cryostorage
11. Transfer the desiccated beads to a cryovial and place in a
cryotank.
Rewarming
12. Warm at ambient temperature in a laminar airflow hood for
2030 min.
13. Wipe cryovials with a sterilant solution (70 % v/v ethanol or
Hibitane).
14. Expel beads from the cryovial and transfer to liquid RIB
medium.
15. Rehydrate for 20 min to remove sucrose.
Recovery
16. Transfer encapsulated shoot-tip meristems to RIB recovery
medium, and monitor regrowth weekly for 68 weeks.
17. Continue weekly assessments, and record recovery as shoot
regrowth and plantlet regeneration; check for sustained sur-
vival and health in recovered shoots/plantlets.
Workflows Workflows 1 and 2 (see Figs. 2 and 3).

Training SOP training record (names of competent
personnel, dates of training, and competency
assessment).
Quality assurance/ Internal QA/QC measures/checks; exam-
quality control ples include in vitro culture, expiry dates of
media, cryoprotectant reagents, data logger
for growth room environmental control,
aseptic technique, and checks for presence
of endophytes before and after cryopreser-
vation; extended post-storage recovery to
confirm sustained acceptable recovery (e.g.,
40 % of cryopreserved shoot meristems produce
plants); and contingencies to assure LN
deliveries, fill-level checks for LN level, LN
depletion alarms, cryobank disaster man-
agement plans, good practice procedures
for adding/removing cryovials to and from
cryotanks, accuracy of annotation, label-
ing, tracking, internal validation, proficiency
testing, witnessing, and audits.
Regulatory issues For example, maintenance of cryogenic
equipment, material transfer agreements,
and phytosanitary regulations.
Revision history Next revision date, versions, author, and
change summary.
Bibliography Fabre and Dereuddre [13]; Dumet et al.
[29]; Reed [22]; Reed et al. [14, 15];
Benson et al. [4, 5, 28, 30]; Johnston et al.
[31]; and Harding et al. [27].
Appendix/annex Risk assessments connected to SOP. Lists
of culture/cryoprotection media, suppli-
ers, and catalogue numbers. Procedures for
detection/elimination of endophytic con-
taminants. Indicators images and descriptors
used to assess viability/regrowth. Glossary
of keywords.
3.6 Writing a Work This WI is a preparative procedure that supports the main cryo-
Instruction preservation SOP as described in Subheading 3.5 and illustrated in
for a Preparative Figs. 2 and 3 as SOP 3. Indicative content of a WI is provided in
Cryopreservation the following example for which explanations of element descrip-
Procedure tions are italicized:
Title Excision of R. nigrum shoot meristems from
sucrose-acclimated, nodal stem segments.
Type Preparative, procedural, work instruction.
Number WI CP 3.00 (Work instruction, CP = nota-

tion for cryopreservation).
Dates 1.1.2014 (date written) 10.1.2014 (date
approved).
Version reference WI CP 3.00; 11.1. 2014 (date issued/effec-
tive date).
Supersedes Not applicable (as a new WI).
Affiliations Plant cryopreservation laboratory (place
where work is done).
Author Author (name of person responsible).
Approved by Quality manager (name of person
responsible).
Personnel Trained person(s) (names of personnel
responsible).
Document control Document controller (name of person
responsible).
Purpose Excision of shoot-tip meristems from
sucrose-acclimated, nodal stem segments of
Ribes spp.; this is a preparative procedure for
SOP CP 3.00; 11.1. 2014 (encapsulation/
dehydration, cryopreservation).
Scope Applicable to all cultivars of R. nigrum.
Risks/safety Confirm with lab manager/health and
safety officer an understanding of the risk
assessment connected to this WI (reference
to health and safety manual).
Safety equipment Lab coat, safety glasses, and sharps disposal
bin.
Equipment Binocular dissecting microscope (magnifi-
cation 20); 2 10 mL syringes fitted with
hypodermic needles; 90 and 50 mm Petri
dishes, 90 and 50 mm sterile filter papers,
forceps, scissors, scalpels, Pasteur pipettes,
3 mL plastic Pastette, and m mesh sieve.
Materials Liquid RIB medium and liquid RIB pre-
growth medium (with 0.75 M sucrose).
Cross-references Cryopreservation SOPs 14; WIs 16 (see
Figs. 2 and 3).
Related documents Best practices for in vitro and aseptic tech-
niques of Benson et al. [4].
Procedure Based on the method of Benson and
Harding [12]. Use good aseptic technique
throughout all in vitro manipulations.
1. Place 2 layers of a 50 mm sterile filter paper on the base of a
90 mm sterile Petri dish.
2. Add a few drops of liquid RIB pregrowth medium to the
filters.
3. Transfer 23 sucrose-acclimated R. nigrum nodal cutting onto

the surface of the medium-moistened sterile filter papers and
transfer to the platform of a dissecting microscope (2040)
positioned in a sterile laminar airflow hood.
4. Fit two hypodermic needles onto two 10 ml syringes, and
using the hypodermic needles, remove the larger expanded
leaves from the shoot tips, retain several none or partly
expanded leaf primordia and a small amount of stem tissue,
and trim the shoot tip to a size of 23 mm.
5. Immediately after dissection transfer the shoot tips to filter
paper bridges placed in a 50 mm Petri dish soaked with liquid
RIB pregrowth medium.
6. Proceed to cryopreservation SOP CP 3.00 (see Subheading 3.5).
Workflows Workflows 1 and 2 (see Figs. 2 and 3).
Training WI training record (names of competent per-
sonnel, dates of training, and competency
assessment).
Quality assurance/ Internal QA/QC measures/checks.
quality control Examples include proficiency in exci-
sion of shoot meristems demonstrated as
sustained shoot/plant regrowth, good
aseptic technique, and absence of microbial
contamination.
Revision history Next revision date, versions, author, summary
of changes, and document control.
Bibliography Benson and Harding [30], Harding et al.
[27], and Johnston et al. [31].
Appendix/annex Risk assessments connected to WI.
Images/video demonstration of shoot-tip exci-
sion technique.
Glossary of keywords.
3.7 Work Work instructions can be used as training aids, for example, skill in
Instructions, Training, meristem excision is critical for successful cryostorage and profi-
and Flexible Quality ciency testing; this technique assures that personnel are well prac-
Improvement Options ticed before they undertake routine cryopreservation (see Note 1).
Work instructions provide flexibility in optimizing storage proto-
cols for genetically diverse collections, for which the refinement
and quality improvement of multiple steps (acclimation, prepara-
tive in vitro manipulations, meristem excision, osmotic pretreat-
ments), use of special additives (dimethylsulfoxide [Me2SO],
antioxidants, hormones, and plant growth regulators), and special
recovery regimes (light/dark phasing) are often necessary to
improve storage outcomes for individual genotypes [3, 27].
3.8 Work Standard operating procedures may need to be refined to ensure

Instructions, Storage that acceptable post-storage outcomes satisfy quality standards for
Recalcitrance, storage recalcitrant species and diverse genotypes that have differ-
and Special ing responses to preparative and cryogenic treatments. WIs have
Treatments the advantage that critical manipulations can be delineated into
individual steps which may be systematically optimized using
appropriate controls. This approach circumvents the need to revise
the entire cryopreservation SOP for more problematic genotypes.
Using WI 3 (meristem excision) as an exemplar, storage recalci-
trant genotypes can be prone to wounding and osmotic and oxida-
tive stress, and they easily succumb to dehydration injury during
and following dissection. The quality improvement of WI 3 can be
undertaken by including precautionary measures that support
recovery before, during, and following meristem excision.
Examples include (a) preventing exposure to inadvertent changes
in temperature or osmotica that will invalidate acclimation regimes;
(b) minimizing the deleterious effects of wounding, osmotic, des-
iccation, and oxidative stress by applying protective agents, osmot-
ica, and free radical scavengers (e.g., Me2SO); and (c) incorporating
pretreatments that allow meristems time to recover from dissection
injury before proceeding with cryopreservation [12].
4 Notes
1. A quality management system (QMS) is the overarching quality

framework, including quality assurance and quality control
(QA/QC) that is adopted by an institution to deliver and docu-
ment the responsibilities, procedures, processes, and tasks to
achieve maximum quality and reliability, effectiveness, risk man-
agement, and safety in the most cost-effective and efficient way.
(a) Quality assurance (QA) involves the management of activi-
ties that ensure the quality of the processes by which
desired outcomes are achieved or products are developed.
QA involves the prevention of defects. In the context of bio-
preservation, QA aims to assure the best possible storage
outcomes and prevent variable, undesirable, and deleteri-
ous changes (e.g., low viability, low and/or abnormal
recovery and regrowth, loss of functionality, totipotency
and morphogenetic capability, genetic instability) in stored
samples that could affect the fitness for purpose of speci-
mens, samples, germplasm, or regenerable plants or organ-
isms recovered from cryogenic storage.
(b) Quality control (QC) involves activities that focus on iden-
tifying problematic outcomes or defects in the end prod-
uct. QC involves the detection of defects. In the context of
biopreservation, QC aims to detect suboptimal processes,
procedures, or conditions that lead to or directly cause
undesirable, substandard storage outcomes, and QC

informs or puts in place actions or conditions to prevent
them. Information from QC assessments can be used to
find and eliminate the sources of problematic quality out-
comes associated with biopreservation and support the
successful realization of storage standards.
2. SOPs are the basic instrument of a QMS (see Note 1); they
comprise the entire documented operations of a biorepository.
Routine protocols are converted into individual SOPs which
are then collated into groups of related tasks, procedures, or
processes and collated in logical sequence in a Quality
Handbook in electronic or web-based formats. Examples of
the different types of SOPs used in a biorepository include
administration, staff recruitment/training, records and docu-
ment management, facilities management, equipment, mainte-
nance, operations, QA/QC, health and safety, risk and disaster
management, chemicals, materials tracking, inventories and
labeling, acquisition and distribution, in vitro/tissue culture,
storage, regulations, and material transfer agreements.
3. There are three basic types of flowcharts: (a) a roadmap,
which navigates the planned stages of a project or gives an over-
view of a complex procedure; (b) a process flowchart that graph-
ically presents series of stages (including timelines/timed
events), where the end point is a product, result, or goal; (c) an
algorithm which is a stepwise list of directions to resolve a prob-
lem or make the best decision to achieve a desired outcome.
4. When deciding on a format for a procedural SOP, it is helpful
to make a detailed list of each step of the protocol in the order
that they are done; this can be enabled by observing personnel
that are well experienced in performing the procedure(s) then
writing down everything that they do and checking this with
them to ensure that each step has been accurately recorded.
This list becomes the first draft of the procedure, as appropri-
ate images, workflows, and WIs may be incorporated. It is
likely that several draft revisions may be needed before an SOP
becomes an authorized quality document.
5. Examples of (a) an instructive writing style and (b) an indirect
writing style:
(a) Excise the shoot-tip meristem using two hypodermic
needles.
(b) The shoot-tip meristem should be excised using two
hypodermic needles. Style (a) is usually preferred when
writing an SOP as it gives a definitive instruction.
6. Published method, internally/externally validated method,
best practices, guidelines, and SOPs provided by a collaborat-
ing institute, accredited body, or society.
Disclaimer
Risk and quality management information is provided for general

use; readers have the ultimate responsibility to ensure quality and
safety issues are addressed according to institutional obligations.
Mention of trade names or commercial products is for the sole
purpose of providing scientific information and does not imply any
recommendation or endorsement by the authors.
Acknowledgements
The authors gratefully acknowledge Dr Jason Johnston and the

helpful advice and guidance of colleagues at the Integrated Biobank
of Luxembourg.
References
1. Stacey GN, Day JG (2007) Long-term ex situ 8. OECD (2009) Guidelines on human biobanks
conservation of biological resources and the role and genetic research databases. OECD, Paris,
of biological resource centers. In: Day JG, Stacey France
G (eds) Methods in molecular biology, vol 368, 9. ISBER (2012) Best practices for repositories
2nd edn, Cryopreservation and freeze drying 3rd edition. Biopreserv Biobank 10:76161
protocols. Humana Press, Totowa, NJ, pp 114 10. Perskvist N, Bjrklund C, Dillner J (2014) A
2. Barnes R, Albert M, Damaraju S et al (2013) complex intervention for workflow enhance-
Generating a comprehensive set of standard ment at the Swedish cervical cytology biobank.
operating procedures for a biorepository Biopreserv Biobank 12:6973
network-the CTRNet experience. Biopreserv 11. Smith D, Ryan M (2012) Implementing
Biobank 11:337396 best practice and validation of cryopreser-
3. Benson EE (2008) Cryopreservation of phyto- vation techniques for microorganisms.
diversity: a critical appraisal of theory and prac- ScientificWorldJournal 2012:805659. doi:
tice. Crit Rev Plant Sci 27:141219 10.1100/2012/805659
4. Benson EE, Harding K, Debouck D et al 12. Benson EE, Harding K (2012)
(2011) Refinement and standardization of Cryopreservation of shoots and meristems: an
storage procedures for clonal crops - global overview of contemporary methodologies. In:
public goods phase 2: part III. Multi-crop Loyola-Vargas VM, Ocho-Alejo N (eds) Plant
guidelines for developing in vitro conservation cell culture protocols, 3rd edn. Humana Press,
best practices for clonal crops. System-wide New York, pp 191226
genetic resources programme, Rome, Italy 13. Fabre J, Dereuddre J (1990) Encapsulation-
5. Benson EE, Betsou F, Fuller BJ et al (2013) dehydration a new approach to cryopreserva-
Translating cryobiology principles into tion of Solanum shoot-tips. Cryo Letters
trans-disciplinary storage guidelines for biore- 11:413426
positories and biobanks: a concept paper. Cryo 14. Reed BM, Dumet DJ, Denoma JM et al (2001)
Letters 34:277312 Validation of cryopreservation protocols for
6. Harding K, Benson EE, Nunes EC et al (2013) plant germplasm conservation: a pilot study
Can biospecimen science expedite the ex situ using Ribes L. Biodiv Conserv 10:939949
conservation of plants in megadiverse coun- 15. Reed BM, Kovalchuk I, Kushnarenko S et al
tries? A focus on the flora of Brazil. Crit Rev (2005) Evaluation of critical points in technol-
Plant Sci 34:277312 ogy transfer of cryopreservation protocols to
7. OECD (2007) Best practice guidelines for bio- international plant conservation laboratories.
logical resource centres. OECD, Paris, France Cryo Letters 25:341352
16. Gonzalez-Arnao MT, Engelmann F (2006) 24. Benson EE, Harding K, Johnston J (2007)
Cryopreservation of plant germplasm using Cryopreservation of shoot-tips and meri-
the encapsulation-dehydration technique: stems. In: Day JG, Stacey G (eds) Methods in
review and case study on sugarcane. Cryo molecular biology, vol 368, 2nd edn,
Letters 27:155168 Cryopreservation and freeze drying protocols.
17. Harding K, Friedl T, Timmermann H et al Humana Press, Totowa, NJ, pp 163184
(2008) Deployment of the encapsulation/ 25. Day JG, Lorenz M, Wilding TA et al (2007)
dehydration protocol to cryopreserve microal- The use of physical and virtual infrastructures
gae held at Sammlung Von Algenkulturen, for the validation of algal cryopreservation
Universitt Gttingen, Germany. Cryo Letters methods in international culture collections.
29:1520 Cryo Letters 28:359376
18. Harding K, Mller J, Day JG et al (2010) 26. Harding K (2004) Genetic integrity of cryo-
Encapsulation-dehydration and colligative preserved plant cells: a review. Cryo Letters
cryoprotective strategies and the use of ampli- 25:322
fied fragment length polymorphism (AFLP) 27. Harding K, Johnston JW, Benson EE (2009)
markers to verify the identity and genetic sta- Exploring the physiological basis of cryo-
bility of cryopreserved Euglena gracilis. Cryo preservation success and failure in clonally
Letters 31:460472 propagated in vitro crop plant germplasm. Agr
19. Jenderek MM, Holman GE, DeNoma J et al Food Sci 18:316
(2013) Medium- and long-term storage of the 28. Benson EE, Reed BM, Brennan R et al
Pycnanthemum (mountain mint) germplasm (1996) Use of thermal analysis in the evalu-
collection. Cryo Letters 34:490496 ation of cryopreservation protocols for Ribes
20. Malpique R, Osorio LM, Ferreira DS et al nigrum L. germplasm. Cryo Letters
(2010) Alginate encapsulation as a novel strat- 17:347362
egy for the cryopreservation of neurospheres. 29. Dumet DJ, Block W, Worland R et al (2000)
Tiss Eng Part C Methods 16:965977 Profiling cryopreservation protocols for Ribes
21. Massie I, Selden C, Morris J (2011) ciliatum using differential scanning calorime-
Cryopreservation of encapsulated liver spher- try. Cryo Letters 21:367378
oids using a cryogen-free cooler: high func- 30. Benson EE, Johnston J, Muthusamy J et al
tional recovery using a multi-step cooling (2005) Physical and engineering perspec-
profile. Cryo Letters 32:158165 tives of in vitro plant cryopreservation. In:
22. Reed BM (2008) Plant cryopreservation: a Dutta Gupta S, Ibaraki Y (eds) Plant tissue
practical guide. Springer, New York culture engineering. Springer, Netherlands,
23. Benson EE (1995) Cryopreservation of shoot- pp 441473
tips and meristems. In: Day JG, Stacey G (eds) 31. Johnston JW, Harding K, Benson EE (2007)
Methods in molecular biology, vol 38, 1st edn, Antioxidant status and genotypic tolerance of
Cryopreservation and freeze drying protocols. Ribes in vitro cultures to cryopreservation.
Humana Press, Totowa, NJ, pp 121132 Plant Sci 172:524534
Part IV
Freeze-Drying Protocols
Chapter 23
Freeze-Drying of Proteins
Baolin Liu and Xinli Zhou
Abstract
Freeze-drying has become one of the most important processes for the preservation of biological products.
This chapter provides protocols for freeze-drying of proteins and discusses the importance of formulation,
cycle development, and validation. Specific formulations for stabilization of proteins are presented as well
as advice on common problems with freeze-drying of proteins.
Key words Freeze-drying, Proteins, Formulation, Stabilizer, Lyophilization
1 Introduction
Proteins are becoming an important class of drugs in human health

care. However, their marginal stability always presents challenges
in their development for pharmaceutical and other industrial use.
Since many proteins do not have adequate long-term storage sta-
bility in the aqueous state, a solid-state dosage form is preferred.
Freeze-drying is most commonly used to prepare dehydrated
proteins, in the areas of protein purification, protein reagent prepa-
ration, and manufacturing of protein biomolecules for therapeutic
and diagnostic applications. Freeze-dried proteins have long-term
stability at ambient temperatures, which offers advantages for stor-
age as well as for shipping and distribution.
1.1 Theory The freeze-drying process primarily consists of three stages.

of Freeze-Drying The first stage is freezing, which involves freezing the product
and creating a solid matrix suitable for drying. The second
stage is primary drying, sometimes preceded by an additional
step called annealing or thermal cycling. Primary drying
involves the removal of ice through sublimation by reducing
the pressure of the product environment while maintaining the
product temperature at a low target level. The third stage in
the process is called secondary drying, in which bound water is
459
460 Baolin Liu and Xinli Zhou
removed until the residual moisture content reaches its

targeted level. Although freeze-drying is the process of choice
for improving the shelf life of proteins, the process itself and
the subsequent long-term storage generate several stresses
responsible for protein denaturation. Even in the solid state,
chemical and physical degradation such as deamidation, oxida-
tion, and aggregation can occur during the freeze-drying pro-
cess, as well as during distribution and use. Stability problems
occurring during processing or storage are usually addressed
through process and formulation optimization.
In order to design an optimum freeze-drying process, pro-
cess development scientists need to know the critical properties
of the formulation and how to apply this information to design
the process. The critical formulation properties include the col-
lapse temperature of the formulation, the stability of the pro-
tein, and the properties of the excipients used. Many approaches
have been explored to determine the thermal properties of for-
mulations for freeze-drying, including differential thermal anal-
ysis (DTA), differential scanning calorimetry (DSC) [1],
electrical resistance analysis (ERA) [2], and freeze-drying
microscopy (FDM) [3].
To develop a protein formulation that minimizes protein
unfolding during freezing and drying, it is crucial that the specific
conditions (e.g., pH and specific stabilizing ligands) for optimum
protein stability be established and that the appropriate nonspe-
cific stabilizing additives (i.e., those excipients that generally stabi-
lize any protein) be incorporated into the formulation. Other
physical factors such as the glass transition temperature and the
residual moisture content of the dried solid must also be opti-
mized to assure storage stability in the dry state [4]. Here we will
describe how to design formulations that protect proteins during
both freezing and drying and discuss the mechanisms by which
additives stabilize proteins.
We wish to emphasize that each protein has unique physico-
chemical characteristics, which often manifest themselves as specific
routes of chemical and physical degradation. Thus, there is a great
need to increase the fundamental understanding of the mechanisms
by which protein stabilizers act and to document, by case studies,
the applicability of the general rules to individual proteins.
1.2 Impact The freeze-drying process generates a variety of stresses to dena-

of Freeze-Drying ture proteins. These include [5, 6]:
on Proteins 1. Low-temperature stress: Studies on low-temperature denatur-
ation of model proteins indicated a temperature of maximum
stability exists, and both high and low temperature can desta-
bilize a protein. The possible mechanism may be the decreas-
ing solvophobic interaction in proteins can reach a point where
protein stability reaches zero, causing cold denaturation.
Freeze-Drying of Proteins 461
2. Concentration effect: Freezing a protein solution rapidly

increases the concentration of all solutes due to ice formation.
Thus, all physical properties related to concentration may
change, such as ionic strength and relative composition of sol-
utes due to selective crystallization. These changes may poten-
tially destabilize a protein. Generally, lowering the temperature
reduces the rate of chemical reactions. However, chemical
reactions may actually accelerate in a partially frozen aqueous
solution due to increased solute concentration.
3. Formation of ice-water interface: Freezing a protein solution
generates an ice-water interface. Proteins can be adsorbed to
the interface, loosening the native fold of proteins and result-
ing in surface-induced denaturation. Rapid (quench) cooling
generates a large ice-water interface while a smaller interface is
induced by slow cooling.
4. pH changes during freezing: Many proteins are stable only in a
narrow pH range. At extreme pHs, increased electrostatic
repulsion between charged groups in proteins tends to cause
protein unfolding or denaturation. Freezing a buffered protein
solution may selectively crystallize one buffering species, caus-
ing pH changes. Na2HPO4 crystallizes more readily than
NaH2PO4. Because of this, a sodium phosphate buffer at pH 7
has a molar [NaH2PO4]:[Na2HPO4] ratio of 0.72, but this
ratio increases to 57 at the ternary eutectic temperature during
freezing. This can lead to a significant pH drop during freez-
ing, which then denatures pH-sensitive proteins.
5. Phase separation during freezing: Freezing polymer solutions
may cause phase separation due to polymers altered solubili-
ties at low temperatures and created a large excess of interface,
denaturing the protein. Freezing-induced phase separation can
easily occur in a solution containing two incompatible poly-
mers such as dextran and Ficoll. Several strategies have been
proposed to mitigate or prevent phase separation-induced pro-
tein denaturation during freezing. These include use of alter-
native salts, adjustment of the relative composition of polymers
to avoid or to rapidly pass over a temperature region where the
system may result in liquidliquid phase separation, and chemi-
cal modification of the protein such as pegylation.
6. Dehydration stresses: A fully hydrated protein has a monolayer
of water covering the protein surface, which is termed the
hydration shell. Lyophilization removes part of the hydration
shell. Removal of the hydration shell may disrupt the native
state of a protein and cause denaturation. A hydrated protein,
when exposed to a water-poor environment during dehydra-
tion, tends to transfer protons to ionized carboxyl groups and
thus abolishes as many charges as possible in the protein. The
decreased charge density may facilitate proteinprotein hydro-
phobic interaction, causing protein aggregation.
1.3 Stabilizers To protect a protein from freezing (cryoprotection) and/or dehy-

for Cryo- and Lyo- dration (lyoprotection) damage, protein stabilizers may be used.
Protection of Protein
1. Sugars/polyols: Sugar and polyol have similar protective mech-
anisms because they have the same functional groups (hydrox-
yls). When they are added to the protein formulation, the
hydroxyl groups in the stabilizers can form hydrogen bonds to
the surface of protein molecule as water does, and substitute
for the hydrogen bonding interaction with water that is lost
during drying. The common sugars/polyols-type protective
agents are shown in Table 1. Generally, sucrose and trehalose
are the most widely used protectants to stabilize proteins.
Mannitol and sorbitol are used as bulking agents and glycerol
is used as cryoprotective agents.
2. Polymers: Polymers are used as protective agents because they
can increase the glass transition temperature of the solution.
Generally, polymer-type protective agents have the following
characteristics: (1) Crystallize firstly during freezing process;
(2) Have higher surface activity; (3) Produce steric hindrance
effect among protein molecules; (4) Increase viscosity of the
Table 1
Sugars and polyols used in freeze-drying formulations for proteins [7, 8]
Type Name Formula MW (g/mol) Tg (C) Tcol (C)

Mono-saccharide Glucose C6H12O6 180.16 43 41
Galactose 41
Mannose 41
Fructose 42
Ribose C5H10O5 150.13 47
Xylose 48
Oligo-saccharide Sucrose C12H22O11 342.30 32 31
Lactose 30.5
Maltose monohydrate C12H22O11H2O 360.32 30
Trehalose dehydrate C12H22O112H2O 378.34 29
Raffinose pentahydrate C18H32O165H2O 594.53 26
Poly-saccharide Mannitol C6H14O6 182.17 27
Glycerol C3H8O3 92.09 65
Sorbitol C6H14O6 182.17 44 54
Xylitol C5H12O5 152.15 47
Inositol C6H12O6 180.16
Table 2
Examples of polymers used for freeze-drying of proteins [7, 8]
Name Formula MW (g/mol) Tg (C) Tcol (C)

Polyethylene glycol (PEG) H[OCH2CH2]xOH 2400 102 13
Dextran [C6H10O5]x 1200 104 10
Hydroxyethyl starch (HES) 12 > 5
4
Ficoll 740 10 20
Gum arabic (acacia) 25 105
Gelatin 8
4
Polyvinylpyrrolidone (PVP) [CHN(CH2)4CO]x 336 10 24 to 27
Cellulose
-Cyclodextrin C48H80O40 1,135.00
Methocel 418 104 9
Maltodextrin 860
Sephadex G200 10
Bovine serum albumin (BSA) 67,000 11
solution; (5) Increase the glass transition temperature signifi-

cantly; (6) Restrain crystallization of excipients with small mol-
ecules (e.g., sucrose); (7) Keep the pH value of solutions. The
common polymer-type protective agents are shown in Table 2.
3. Surfactants: Surfactants are wetting agents composed of hydro-
philic and oleophilic groups that can reduce the surface tension
of a liquid and reduce the interfacial tension between two liquids.
It is soluble in oil because of the C-H chains and soluble in water
because of polar groups (-COOH, -OH). When these molecules
are located at an air-water or oil-water interface, the hydrophilic
groups face the water phase, while the oleophilic groups point at
gas and oil phase. In the freeze-drying of proteins, surfactants
can reduce denaturation during freezing and dehydration. The
surfactants also act as humidifying agents during the rehydration
process. Typical surfactants are shown in Table 3.
4. Amino acids: An amino acid is a molecule containing both
amine and carboxyl functional groups. It can inhibit pH
changes of the solution during low-temperature storage and
freeze-drying of proteins. Typical amino acids used in freeze-
drying are shown in Table 4.
5. Miscellaneous excipients: Bulking agents are substances that can
prevent the effective components of the formulation from
escaping along with the water vapor, and promote fixation of
Table 3
Typical surfactants used for freeze-drying of proteins [8]
Name
Tween 80
Triton X-100
Sucrose fatty acid monoester
3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS)
Hydroxypropyl--cyclodextrin (HP--CD)
Sodium dodecanesulfonate (SDS)
Brij35 Brij30
Lubrol-px
Pluronic F127
Table 4
Amino acids as protective agents for proteins [7, 8]
Name Formula MW (g/mol) Tg (C)

Proline (CH2)3NHCHCOOH 115.13
Glycine CH2NH2COOH 75.07 37
Glutamic acid (CH2)2NH2CH(COOH)2 147.13 17
Histidine NHCHNCHCCH2CH(NH2)COOH 155.16 32
Arginine HNC(NH2)NH(CH2)3CH(NH2)COOH 174.20
4-Hydroxyproline NHCH2CH(OH)CH2CHCOOH 131.13
L-Serine HOCH2CH(NH2)COOH 105.09
-Alanine CH2(NH2)CH2COOH 89.09 65
Lysine hydrochloride H2N(CH2)4CH(NH2)COOHHCl 182.65
Lysine H2N(CH2)4CH(NH2)COOH 146.19
Sarcosine CH3NHCH2COOH 89.09
-Aminobutyric acid H2NCH2CH2CH2COOH 103.12
the effective components in the material. Typical bulking com-

pounds are mannitol, lactose, and gelatin. The addition of
bulking agents (easy to crystallize) into the formulation of
freeze-drying products has additional functions such as (1) pro-
viding adequate mechanical support for the final freeze-dried
products, (2) improving the appearance of the freeze-dried
products, (3) increasing the solubility of solutes, and (4)
Table 5
Antioxidants used in freeze-drying formulations for proteins [7, 8]
Name Formula MW (g/mol)

Antisterility factor (vitamin E) C29H50O2 430.72
Ascorbic acid (vitamin C) C6H8O6 176.13
Lecithin C40H82NO9P 752.08
D()-Isoascorbic acid C6H8O6 176.13
L-Ascorbic sodium C6H7 NaO6 198.11
Sodium thiosulfate anhydrous Na2S2O3 158.11
3-Tert-butyl-4-hydroxyanisole C11H16O2 180.25
Butylated hydroxy toluene CH3C6H2(OH)[C(CH3)3]2 220.36
Propyl gallate (HO)3C6H2COOCH2CH3 212.20
Ethylene diamine tetraacetic acid disodium salt C10H14N2Na2O82H2O 372.24
dihydrate (Na2EDTA)
preventing the freeze-dried products from collapse or overflow.

Antioxidants, such as vitamin D, vitamin E, protein hydroly-
sate, and sodium hyposulfite, are used to prevent oxidation dur-
ing freeze-drying and storage (Table 5). Buffer and chelating
agents, such as phosphoric acid, amino acids, and EDTA, regu-
late the pH value of the material and sequester ions (Table 6).
2 Materials
2.1 Freeze-Drying 1. A freeze dryer (e.g., Virtis Advantage, Millrock, Tofflon, or

Systems similar unit).
2. Suitable containers for freeze-drying the product, e.g., round-
bottomed glass flasks, glass vials.
3. Freezing system, e.g., liquid nitrogen, a 80 C commercial
freezer.
2.2 Formulations 1. Selection of the optimal pH value: An optimum pH is needed

for Freeze-Drying to keep a protein stable and soluble in solution (see Note 1).
of Proteins 2. Selection of buffer agents: Many buffering agents covering a
wide pH range are available for selection in formulating solid
proteins. These agents include acetate, citrate, glycine, histi-
dine, phosphate, and Tris (see Note 2).
3. Selection of bulking agents: Mannitol and glycine are two typi-
cal bulking agents. Most amino acids are potential bulking
agents as they easily crystallize out (see Note 3).
Table 6
Typical buffer agents used in protein formulations [7, 8]
Name Formula MW (g/mol) Tg (C) Tcol (C)

Citric acid monohydrate C6H8O7H2O 210.14 54
Phosphoric acid H3PO4 98.00
Ethylenediamine tetraacetic acid (EDTA) (CH2)2N2(CH2COOH)4 292.24
Tartaric acid C4H6O6 150.09
4-(2-Hydroxyethyl)-1- C8H18N2O4S 238.31 63
piperazineethanesulfonic acid (HEPES)
Histidine NHCHNCHCCH2 155.16 33
CH(NH2)COOH
Potassium acetate CH3COOK 98.14 76
Potassium citrate HOC(COOK) 306.42 62
(CH2COOK)2
Potassium phosphate monobasic KH2PO4 136.09 55
Sodium acetate CH3COONa 82.03 64
Sodium bicarbonate Na2CO3 105.99 52
Sodium citrate HOC(COONa) 258.10 41
(CH2COONa)2
Sodium phosphate NaH2PO4 119.98 45
Tris base 55
Tris HCl 65
4. Selection of cryoprotectants and lyoprotectants: A stabilizer(s)

for a solid protein product should be at least partially amor-
phous and able to replace water, forming intimate hydrogen
bonds with the protein. See Subheading 1.3.
5. Overall consideration of formulation excipients (see Note 4):
Several freeze-drying formulations for a variety of proteins and
their protective efficacy are shown in Table 7.
3 Methods
3.1 Preparation For proteins, it is necessary to add some additives before freeze-
of the Materials drying in order to keep their activity and get good quality of final
products. It is called formulation of the mixture of proteins and
additives. The additives are used for stabilizing the formulation or
for therapeutic reasons. The additives are specified according to
the type of proteins (see Note 5).
Table 7
Examples of formulation for freeze-drying of proteins [5, 8]
Name Formulation Protective efficacy Experimental conditions

Catalase (ox liver) Activity retention: Freezing between 15 and
70 C for 10 min
8.4 g/mL catalase, 10 mM 80 %
phosphate buffer (pH 7.0)
+1 M glycerol 95 %
+0.5 mM sodium dodecyl sulfate 90 %
Elastase Activity retention: Storage at temperature of
40 C and relative
20 mg/mL elastase, 10 mM 33 %
humidity of 79 % for 2
sodium acetate buffer solution
weeks
(pH 5.0)
+10 % (w/v) sucrose 86 %
+10 % (w/v) dextran 40 82 %
-Galactosidase Activity retention: Freeze-drying
20 g/mL -galactosidase, 47 %
200 mM phosphate buffer
(pH 7.4), 100 mM inositol
+400 mM inositol 90 %
+400 mM inositol, 1 mg/mL 100 %
dextran
+400 mM inositol, 1 mg/mL 100 %
CMC-Na
Recombinant Activity retention: Freeze-drying
factor XIII
2 mg/mL recombinant factor 82 %
(rFXIII)
XIII, 0.1 mM EDTA, 10 mM
Tris buffer (pH 8)
+100 mM mannitol 90 %
+1 % (w/v) PEG 115 %
+3.5 % (w/v) dextran 109 %
+100 mM trehalose 117 %
+100 mM sucrose 111 %
+0.002 % (w/v) Tween 20 106 %
(continued)
Table 7
(continued)

H+-ATPase Activity retention: Freeze-drying
1 mg/mL H+-ATPase, 1 Mm 4%
EDTA-Tris buffer (pH 7.0)
+20 mg trehalose/1 mg protein 100 %
+20 mg maltose/1 mg protein 91 %
+20 mg sucrose/1 mg protein 84 %
+20 mg glucose/1 mg protein 72 %
+20 mg galactose/1 mg protein 37 %
rhIL-1ra Degeneration: Storage at temperature of
50 C
10 mg/mL rhIL-1ra, 2 % 1.8 %/week
aminoacetic acid, 10 mM
sodium citrate (pH 6.5)
+1 % sucrose (w/v) 0.15 %/week
+1 % maltose (w/v) 6.4 %/week
+1 % sorbitol (w/v) 2.3 %/week
+1 % trehalose (w/v) 0.43 %/week
IL-2 Activity retention: Storage at temperature of
45 C for 4 weeks
0.8 mg/mL freeze-dried IL-2, 47 %
MES buffer (pH 7)
+0.5 % sucrose (w/v) 86 %
+5 % mannitol (w/v) 28 %
LDH Activity retention: Freeze-drying
2 g/mL LDH, 50 mM sodium 9%
phosphate buffer (pH 7.4)
+400 mM mannitol 14 %
+200 mM sucrose 30 %
+10 mg/mL CHAPS 76 %
+10 mg/mL HP--CD 73 %
+10 mg/mL sodium cholate 60 %
+10 mg/mL PEG3000 30 %
+10 mg/mL PEG20000 38 %
(continued)
Table 7
(continued)

PFK Activity retention: Freeze-drying
2 mg/mL PFK, 20 mM 0%
potassium phosphate buffer
(pH 8.0)
+1 % PEG 20 %
+10 mM mannitol 5%
+10 mM lactose 5%
+10 mM trehalose 0%
+1 % PEG + 10 mM mannitol 30 %
+1 % PEG + 10 mM lactose 103 %
+1 % PEG + 10 mM trehalose 100 %
GDH Activity retention: Freeze-drying
50 mg/mL GDH, 20 mM 65 %
potassium phosphate buffer
(pH 7.5)
+500 mM KCl 32 %
+500 mM trehalose 86 %
+500 mM mannosylglycerate 95 %
hGH Rate constant: Storage at temperature of
50 C
Lyophilized formulation 2 mg/ 4.5 per day
mL hGH
+Mannitol at 31:1 molar ratio 0.004 per day
+Trehalose at 31:1 molar ratio 0.002 per day
+Cellobiose at 31:1 molar ratio 0.0003 per day
Insulin Degradation: Storage at temperature of
35 C for 7 days
Lyophilized formulation
1 mg/mL insulin (pH 4) 0.04 mg/mL
(18 % dimers)
+5 mg/mL trehalose (0.02 mg/mL 1 %
dimers)
3.2 Filling Container 1. The formulation must be filled into containers before
freezing.
2. Choose containers made of glassware such as round-bottomed
glass flasks or glass vials (see Note 6).
3. The method of dispensing the product depends on the num-
ber of containers to be filled and the accuracy and consistency
of the desired fill volume.
4. Containers should not be filled more than one-third of their
total volume. Sudden volume and thermal changes may result
in shattering of the container with catastrophic loss or cross-
contamination of valuable samples.
3.3 Steps The general steps for freeze-drying proteins are as follows:
of a Freeze-Drying
1. Switch on freeze dryer and vacuum pump
Protocol
2. Make formulation mixtures by adding additives to the protein
solution
3. Dispense mixtures into glassware
4. Freeze containers with the sample mixtures
5. Primary drying
6. Secondary drying
7. Conditioning-packing and storage
3.4 Freeze-Drying Generally, a freeze-drying cycle includes freezing (solidification),

Cycle Design primary drying (sublimation), and secondary drying (desorption)
as shown in Fig. 1. This figure also indicates the variations of the
material temperature and the moisture during freeze-drying pro-
cesses [8].
3.5 Freezing Samples should be properly frozen during the freezing step
(see Note 7).
1. Freezing can be done in one of the three ways as (1) by dip-
ping or immersing into liquid nitrogen, (2) in a specified
freezer, or (3) inside the freeze dryer. The freezing rate is
important (see Note 8).
2. The end freezing temperature can be derived from studies of
the critical glass transition by DSC or the collapse temperature
by freeze-drying microscopy experiments.
3. The sample temperature should be measured during the entire
freezing and drying process (see Note 9).
4. Make sure the length of the freezing is sufficient to ensure that
all samples are completely frozen (see Note 10).
Tmax2
40
Tw2
20
Temperature /C
0
Tmax1
20
Tw1
40 Tcs
60
Time
solidification desorption
Relative residual moisture
sublimation
100%
free water
RMF
bound water
0
Time
Fig. 1 Variations in sample temperature and moisture content during freeze-

drying processes
3.6 Primary Drying Water can be classified as free water that is freezable at low tem-
(Sublimation Drying) peratures and bound water that cannot be frozen. Primary drying
refers sublimation of frozen free water directly to vapor.
1. For proteins, the shelf temperature can be set between 40 C
and 80 C according to the specifications of the freeze dryer.
2. Set the temperature during sublimation lower than the maxi-
mum allowable temperature Tmax1 of proteins as shown in
Fig. 1 (see Note 11).
3. The vacuum inside the freeze dryer usually maintains at
10100 bar [9].
4. The primary drying time ranges from several hours to several
days dependent on the drying system and solvent being removed.
5. During primary drying, heat can be provided to the samples
through conduction or radiation (see Note 12).
3.7 Secondary After primary drying, when most of the water is removed, the sec-
Drying (Desorption ondary drying process can start (see Note 13).
Drying) 1. During the secondary drying process, the sample temperature,
Tw2, should be less than the maximum allowable temperature,
Tmax2 (see Fig. 1).
press down to seal
rubber bottle
stopper
vapor channel
ampoule or vial
materiel
Fig. 2 Vial for freeze-drying with a rubber stopper for stoppering the product
under vacuum at the end of the process
2. The maximum allowable temperature, Tmax2, is dependent on

the type of protein (see Note 14).
3. The vacuum pressure should be kept the same as during pri-
mary drying.
4. The secondary drying process ends when the remaining mois-
ture final (RMF) of the proteins is less than 5 % (see Note 15).
3.8 Conditioning- Sealing and packing should be conducted by breaking vacuum in a

Packing and Storage dry environment (see Note 16). The packing and sealing should be
carried out either in the vacuum chamber or in a chamber with an
inert gas (nitrogen or argon). For freeze-dried proteins in bottles
or vials, sealing may be carried out in the drying chamber, with
rubber stoppers as shown in Fig. 2. Bulk products in ampoules can
be taken out of the drying chamber through a vacuum channel,
delivered to a vacuum chamber (or a chamber with sufficient inert
gas), and sealed by a manipulator. Product parameters should be
assessed (see Note 17). The freeze-dried samples can be stored at a
specific temperature (see Note 18).
4 Notes
1. The solution pH must be optimal to minimize protein dena-

turation during lyophilization and to allow maximum long-
term stability of the lyophilized proteins. To meet all these
requirements at the different stages, the pH of the formula-
tion needs to be carefully chosen. Very often, the most stable
pH for proteins in solution does not offer the best stability in
solid state. This is because the inactivation mechanisms of pro-
teins are different in the two different states. In such cases, a
balanced pH must be used.
2. A buffering agent, such as histidine for freeze-dried FVIII SQ
that also stabilizes the protein, is preferable. For pH-sensitive
proteins, sodium phosphate should be avoided because the selective

crystallization of Na2HPO4 can cause a significant pH drop dur-
ing freezing, thus denaturing the protein. Instead, potassium
phosphate, citrate, histidine, or Tris can be used, which result in
minimal pH changes during freezing. In addition, decreasing the
buffer concentration can also mitigate the pH shift.
3. Amorphous excipients (e.g., sucrose, trehalose) in a protein
formulation may inhibit crystallization of the bulking agents,
thus affecting protein stability. Therefore, proper selection of
a suitable bulking agents and its relative amount is critical.
4. Generally, the weight content of solids in the protein formula-
tion is between 2 % and 10 %. When the weight fraction is less
than 2 %, hard freeze-dried products cannot be obtained. When
the weight fraction is more than 10 %, it is difficult to com-
pletely dry the products and rehydration is very difficult. The
physical/chemical function between the additives is mutual. It
is important to maintain a suitable amount of additives in the
formulation. It is preferable if one type of additive fulfils several
protective functions simultaneously. For example, several sug-
ars can be used as cryoprotectant as well as lyoprotectant. In
addition, many buffer solutions or salts cannot be used because
they may change the pH during the freezing process or decrease
the glass transition temperature of the formula.
5. Additives can be classified as lyoprotectant, emulsifier, bulking
agent, antioxidant, and buffer agent. Formulations of lyoprotec-
tants for specific proteins are listed in Tables 1, 2, 3, 4, 5, and 6.
6. Glassware should be washed without detergents and heated/
autoclaved to sterilize. Containers used should be capable of
withstanding quick freezing such as the immersion in liquid
nitrogen or freezing within a freezer at 70 C or lower with-
out cracking. The containers must be capable to withstand
exposure to deep vacuum conditions.
7. Not only the free water in the formulation should be frozen to
form crystalline ice completely, but also the other constitutes
of the formulation should be solidified, to form noncrystalline
solid (glassy solid). Rapid freezing rates are recommended,
especially if the target protein rapidly loses activity in the liquid
state or if the selected formulation results in marked pH shifts
during freezing and denaturation. However, rapid freezing
may result in small ice crystals and longer drying times. If
rapid freezing is necessary an annealing step may be needed to
correct the effects [10].
8. The cooling rate and the solidification process are also impor-
tant factors. In general, rapid cooling would lead to formation
of partial glassy state and prevent excessive dehydration of the
material during cooling. However, if the cooling rate is too
fast it may cause fracture. The variation in temperature during

the process is called thermal history. The thermal history
during the cooling process is important because it affects the
thermal properties of the formulation. It has been found that
annealing affects the glass transition temperature Tg.
9. Insert a thermocouple into the samples to monitor the freez-
ing and drying process. It should be noted, however, that the
thermocouple will affect the nucleation point.
10. The material should be completely solidified after cooling.
The solidified proteins contain both crystalline solid and non-
crystalline hard reticulation structure solid. The cooling and
solidification process is an extremely important process, which
is often underestimated. The end temperature of the sample
after the freezing step should be lower than the eutectic tem-
perature, Te, or the glass transition temperature, Tg (see Fig. 1).
11. As shown in Fig. 1, the highest permissible temperature, Tmax,
is the glass transition temperature Tg, or the eutectic tempera-
ture Te. If the sublimation temperature is too high, the mate-
rial will appear softened or collapsed.
12. Two basic conditions must be satisfied to assure the progress
of the sublimation process: the vapor must be moved away
unceasingly from the sublimation interface, and heat must be
provided for sublimation. If either of the two is not met, soft-
ening, thawing, bulging, or collapse may occur. In fact subli-
mation is a process where heat and mass transfer simultaneously
take place. Only when the heat that is transferred to the inter-
face of sublimation is equal to the heat needed for the vapor to
escape from the interface, sublimation carries on smoothly.
13. At the end of primary drying, there is still unfrozen water
adsorbed on the porous structure surface and on the polar
groups of the proteins. Secondary drying is conducted at a
higher temperature, so that the bound unfrozen water absorbs
the heat of desorption and becomes free water. Then the free
water absorbs the heat needed for evaporation and finally turns
into vapor escaping the material. Since the adsorption energy
is quite high, desorption of bound water requires a higher
temperature and sufficient heat. The temperature should not
be too high to avoid protein denaturation.
14. For protein pharmaceuticals, the maximum allowable temper-
ature should generally be lower than 40 C. For food prod-
ucts, such as meats, fruits, and vegetables, the maximum
permissible temperature may be 6070 C or higher.
15. At the end of the desorption process, the sample reaches its
remaining moisture final (RMF). The optimum residual water
content of dried proteins should be determined on a case-by-
case basis. If the RMF is too high or too low, this can be
detrimental to the protein. An RMF that is too high may result

in degradation during storage. On the other hand if the RMF is
too low, this can also harm the active material [11]. The samples
need to be kept at an optimum water content to maintain activ-
ity of the protein [12]. Moisture content measurements are
necessary to predict storage stability of the dried protein formu-
lation. Water contents can be determined using the colorimetric
Karl Fischer method or by thermogravimetric analysis [13].
16. Avoiding contact with oxygen or water vapor in the air
increases storage stability of freeze-dried products. When the
product is needed, rehydration in water is necessary in most
cases.
17. The recovery and functional activity of the freeze-dried pro-
teins should be assessed after the process. The appearance and
residual moisture content should also be assessed. Visually
inspect all vials in the batch and assess the appearance of the
freeze-dried samples. Heterogeneity should be documented
to see whether it correlates with a particular location in the
freeze dryer. Poor appearance may indicate suboptimal
drying.
18. Some freeze-dried proteins can be stored at room tempera-
ture, whereas other proteins need to be stored at 4 C or
18 C. Accelerated aging studies at elevated temperatures
and humidities can be used to estimate storage stability of the
freeze-dried protein. The trends in activity loss can be ana-
lyzed using Arrhenius or other kinetic models and used to pre-
dict stability at the storage temperature [14]. The recovered
biological activity should be measured against controls taken
before lyophilization, to determine the formulation that gives
the best activity after rehydration.
References
1. Passot S, Fonseca F, Alarcon-Lorca M (2005) VL (eds) Biotechnology and biopharmaceuti-

Physical characterisation of formulations for cal manufacturing, processing and preserva-
the development of two stable freeze-dried tion. Interpharm Press, Buffalo Grove, IL,
proteins during both dried and liquid storage. pp 199264
Eur J Pharm Biopharm 60:335348 5. Wang W (2000) Lyophilization and develop-
2. Ma X, Wang W, Bouffard R (2001) ment of solid protein pharmaceuticals. Int J
Characterization of murine monoclonal anti- Pharm 203:160
body to tumor necrosis factor (TNF-MAb) for- 6. Chang LL, Pikal MJ (2009) Mechanisms of
mulation for freeze-drying cycle development. protein stabilization in the solid state. J Pharm
Pharm Res 18:196202 Sci 98:28862908
3. Nail SL, Her LM, Proffitt CPB (1994) An 7. Hua TC (2006) New technology of freeze-
improved microscope stage for direct observa- drying. Science Press, Beijing
tion of freezing and freeze drying. Pharm Res 8. Hua TC, Liu BL, Zhang H (2010) Freeze-
11:10981100 drying of pharmaceutical and food products.
4. Carpenter JF, Chang BS (1996) Lyophilization Cambridge, Science Press, Beijing and CRC
of protein pharmaceuticals. In: Avis KE, Wu Press LLC, Boca Raton
9. Matejtschuk P (2007) Lyophilization of 12. Breen ED, Curley JG, Overcashier DE, Hsu
proteins. In: Day JG, Stacey GN (eds) CC, Shire SJ (2001) Effect of moisture on the
Cryopreservation and freeze-drying protocols, stability of a lyophilized humanised monoclo-
2nd edn. Humana Press, Totowa, NJ, pp 5972 nal antibody formulation. Pharm Res
10. Searles JA, Carpenter JF, Randolph TW (2001) 18:13451353
Annealing to optimize primary drying rate, 13. May JC, Wheeler RM, Etz N, Del Grosso A
reduce freeze-induced drying rate heterogene- (1992) Measurement of final container residual
ity and determine the Tg in pharmaceutical moisture in freeze dried biological products.
lyophilization. J Pharm Sci 90:872877 Dev Biol Stand 74:153164
11. Jiang S, Nail SL (1998) Effect of process con- 14. Kirkwood TBL (1984) Design and analysis of
ditions on recovery of protein activity after accelerated degradation tests for the stability of
freezing and freeze-drying. Eur J Pharm biological standards III. Principles of design.
Biopharm 45:245257 J Biol Stand 12:215224
Chapter 24
Freeze-Drying of Lactic Acid Bacteria

Fernanda Fonseca, Stphanie Cenard, and Stphanie Passot
Abstract
Lactic acid bacteria are of great importance for the food and biotechnology industry. They are widely used
as starters for manufacturing food (e.g., yogurt, cheese, fermented meats, and vegetables) and probiotic
products, as well as for green chemistry applications. Freeze-drying or lyophilization is a convenient
method for preservation of bacteria. By reducing water activity to values below 0.2, it allows long-term
storage and low-cost distribution at suprazero temperatures, while minimizing losses in viability and func-
tionality. Stabilization of bacteria via freeze-drying starts with the addition of a protectant solution to the
bacterial suspension. Freeze-drying includes three steps, namely, (1) freezing of the concentrated and
protected cell suspension, (2) primary drying to remove ice by sublimation, and (3) secondary drying to
remove unfrozen water by desorption. In this chapter we describe a method for freeze-drying of lactic acid
bacteria at a pilot scale, thus allowing control of the process parameters for maximal survival and function-
ality recovery.
Key words Lactic acid bacteria (LAB), Starter, Fermentation, Freeze-drying, Lyophilization,
Formulation, Lyoprotectors, Preservation
1 Introduction
Freeze-drying (or lyophilization) is an invaluable technique for

stabilizing products that are greatly sensitive to high temperatures
such as lactic acid bacteria (starters). The process involves freezing
of the aqueous solution containing bacteria, followed by primary
drying to sublimate ice and, finally, secondary drying to remove
unfrozen water [1].
The freezing step is the first and the shortest step of the lyophi-
lization process. It governs ice crystal size, distribution, and mor-
phology (Fig. 1a) [2, 3]. Sublimation requires low vapor pressure
(vacuum) and heating for changing ice into vapor. The heat input
must be controlled in order to keep the product temperature lower
than the collapse temperature of the matrix [4, 5]. A water vapor
pressure gradient, necessary for water removal from the sample, is
achieved by trapping water by condensation (cold trap). Following
sublimation, the remaining concentrated solute phase still contains
477
478 Fernanda Fonseca et al.
Fig. 1 Cryo-SEM of Lactobacillus delbrueckii ssp. bulgaricus CFL1, frozen in a protective matrix. Panel A shows
LAB cells in the concentrated matrix. Panel B shows more clearly the interface solid matrix-ice (matrix-air after
drying)
significant amounts of unfrozen water (2030 %) incompatible

with storage stability of freeze-dried products. Desorption of
unfrozen water is achieved by heating the product to moderate
suprazero temperatures (2030 C), thus allowing reducing water
content (<5 % dry weight) and water activity (<0.2) [6].
Even if low product temperatures are applied, varying degrees of
viability recovery have been reported for freeze-dried bacteria [7, 8].
Freeze-Drying of Bacteria 479
Freeze-dried powders contain a combination of dead, viable, and

sublethally injured cells in variable proportions which depend on
strain sensitivity to environmental stresses induced by the stabiliza-
tion process itself. Freezing induces ice crystal formation and con-
centration of solutes. Low cooling rates typically applied in
freeze-drying recipes cause cell packing into the freeze-concentrated
material and cell dehydration (Fig. 1). Dehydration may cause
irreversible changes in bacterial cells, such as changes in the physi-
cal state of membrane lipids and in the structure of sensitive
proteins that often result in severe loss in bacterial viability [911].
Moreover, oxidation appears to be an important mechanism of cel-
lular damage during drying and storage [12, 13].
Formulation is a paramount step for stabilization of lactic acid
bacteria (LAB) to limit freeze-drying and storage stresses. LAB are
currently freeze-dried in the presence of sugars and polymers [11,
12, 1418]. The amorphous matrix that is formed by the combina-
tion of sugars and polymers has a high viscosity allowing low
molecular mobility, thus limiting diffusion-controlled degradation
reactions. Moreover, protective molecules such as sugars protect
cells by replacing water-forming hydrogen bonds with biomole-
cules, otherwise lost during drying [911, 19]. Protective agents
exhibiting antioxidant activity are usually also added to the formu-
lation for scavenging of reactive oxygen species [14, 20, 21].
Furthermore, operating conditions such as the freezing kinetics,
product temperature during freeze-drying, final water content,
and storage conditions (temperature, relative humidity, and atmo-
sphere) have a major impact on the stability of freeze-dried LAB
[6, 22, 23]. Consequently, they have to be considered for defining
freeze-drying protocols allowing high survival and activity recov-
ery of dried LAB.
2 Materials
2.1 Lactic Acid 1. Strains: culture collections are stored either frozen at 80 C
Bacteria/Biomass or in liquid nitrogen or freeze-dried at 80 C.
Production 2. Pre-culture media for inocula preparation (see Note 1):
(a) De Man, Rogosa, and Sharpe (MRS) medium for lactoba-
cilli. Typical composition: 20 g/L glucose,10.0 g/L poly-
peptone, 10.0 g/L meat extract, 5.0 g/L yeast extract,
5.0 g/L sodium acetate, 2.0 g/L dipotassium phosphate,
2.0 g/L ammonium citrate, 1.0 g/L Tween-80, 0.2 g/L
magnesium sulfate, 0.05 g/L manganese sulfate. Prepare
by weighing each ingredient and suspending in demineral-
ized water. Autoclave the broth at 121 C for 15 min, and
store at 48 C for up to 15 days.
(b) M17 medium for lactococci and streptococci. Typical
composition: 19.0 g/L sodium glycerophosphate, 5.0 g/L
lactose monohydrate, 5.0 g/L papaic digest of soybean

meal, 5.0 g/L meat extract, 2.5 g/L tryptone, 2.5 g/L pep-
tic digest from meat, 2.5 g/L yeast extract, 0.5 g/L ascor-
bic acid, 0.25 g/L magnesium sulfate. Prepare by weighing
each ingredient and suspending in demineralized water.
Autoclave the broth for 20 min at 121 C, and store at
48 C for up to 15 days.
3. Culture medium. Mild whey-based medium (60 g/L mild
whey, 15 g/L yeast extract). Suspend 60 g of mild whey pow-
der (see Note 2) in 1 L of demineralized water and then heat
at 110 C for 20 min. Separate the supernatant from precipi-
tated proteins by centrifugation (17,700 g for 30 min at
4 C) and filter through filter paper. Supplement the superna-
tant with 15 g/L yeast extract. Autoclave the medium at
110 C for 20 min, and store at 48 C for up to 2 days.
4. Glass flasks (Erlenmeyer or others) for cell pre-culturing.
5. Fermentor (1.5, 5, or 10 L working volume) for homogeniza-
tion by agitation and control of the temperature and pH of the
culture.
6. Neutralizing solution (about 2 M NaOH). Dissolve 100 g
(70 mL) of NaOH 35 % (w/v) in 230 mL of demineralized
water.
7. Balance for measuring NaOH solution consumption during
fermentation (precision weight 0.01 g).
8. Peristaltic pump for adding neutralizing solution during
fermentation.
2.2 Formulation 1. Prepare sugar solution: suspend 200 g of sucrose and 9 g of

Medium sodium chloride in 500 mL demineralized water by stirring,
and autoclave the solution at 120 C for 20 min. Store the
protective solution at 4 C for up to 1 month.
2. Prepare antioxidant solution: suspend 5 g of sodium ascorbate
in 500 mL demineralized water and filter-sterilize by passing the
solution trough a 0.2 m sterile membrane filter. Protect the
solution from light (in a dark flask and/or cover with aluminum
foil), and use immediately or store at 48 C for up to 1 day.
3. Mix equal volumes of sugar solution and antioxidant solution,
to obtain protective medium with the following composition:
200 g/L sucrose, 9 g/L sodium chloride, and 5 g/L sodium
ascorbate (see Notes 3 and 4). Use immediately or store at
48 C for up to 1 day, protected from light (in a dark flask
and/or cover with aluminum foil).
2.3 Freeze-Drying 1. Freeze-dryer pilot plant requirements: shelf surface about

Equipment 0.5 m2, cold trap/condenser temperature lower than 55 C,
and Materials and filter installed on the air injection valves (used for pressure
regulation and vacuum breakage) for ensuring sterility. A system

making it possible to stopper vials in the freeze-dryer is a plus.
2. Thoroughly clean the freeze-dryer chamber and shelf with a
disinfectant spray (efficient against bacteria, fungi, and viruses)
and rinse with sterile water before utilization (see Note 5).
3. Temperature sensors (e.g., thermocouples) for monitoring
product temperature during the freeze-drying process. Clean
and calibrate them before utilization (see Note 6).
4. Sensor for detecting the end of sublimation (moisture sensor,
Pirani gauge) (see Note 7).
5. Four milliliter sterile glass vials for laboratory freeze-drying
studies (see Note 8).
6. Stainless steel trays for placing and transferring vials.
7. Sterile dried rubber stoppers adapted for open vials, thus allow-
ing water vapor to flow out from the vial during sublimation.
If necessary, dry stoppers in sterile container at 90 C for 24 h
and store them in a low-moisture-content environment (i.e.,
by using a desiccant) before utilization.
8. Screw caps adapted for vials used.
9. Computer for data acquisition and visualization of tempera-
ture and pressure parameters during freeze-drying.
10. Aluminum bags for storage of freeze-dried samples.
11. Equipment for sealing aluminum bags under vacuum.
3 Methods
Figure 2 depicts the flowchart of the whole LAB stabilization

process, including the stages of culture preparations, formulation,
and freeze-drying.
3.1 Culturing, 1. Sterilize the cleaned empty fermentor at 120 C for 20 min.
Harvesting, 2. Supplement the mild whey-based medium with yeast extract
Concentrating, and add (1.5, 5, or 10 L) in the fermentor (see Note 9).
and Formulating
3. Sterilize the fermentor containing the culture medium at
Bacterial Suspensions 110 C for 20 min and let it cool to the optimal growth
temperature.
4. Thaw pure inoculum, in a water bath set at the optimal growth
temperature. Alternatively, rehydrate the freeze-dried inocu-
lum with sterile demineralized water. Aseptically inoculate (see
Note 4) a flask of pre-culture medium (MRS or M17 depend-
ing on the strain) (see Note 10) at 1 % (v/v) inoculation ratio
(i.e., 10 mL/L).
Inoculum from culture Pre-culture and culture

collection (80C) media preparation
Pre-culture Pre-culture and culture

inoculum media sterilisation
Batch fermentation
(control of pH and temperature)
Cooling and harvesting Protective medium

(centrifugation at 4C) preparation and sterilisation
Formulation
(addition of protective medium at 4C)
Distribution into sterile containers (vials)
Transfer containers to freeze-drier (4C)
Freeze-drying (Lyophilisation)
1. Freezing to 50C
2. Primary drying (20C, 20 Pa)
3. Secondary drying (25C, minimal pressure)
Stoppering and capping
Conditioning (vacuum, aluminium bags)
Storage <+15C
Fig. 2 Flowchart showing the stages of the experimental approach applied for
lactic acid bacteria production and stabilization by freeze-drying
5. Incubate at the optimal growth temperature and regularly

measure the optical density for following the cell growth
(see Note 11).
6. Stop the pre-culture at the beginning of the stationary growth
phase, for obtaining bacteria for fermentor inoculation.
7. Aseptically inoculate (see Note 4) the fermentor containing the
mild whey-based medium with bacteria from pre-culture (1 %
(v/v) inoculation ratio). Grow bacteria under the optimal tem-
perature and pH conditions while stirring for homogenization.
Control the pH at the optimum value during fermentation, by
adding neutralizing solution (NaOH) (see Note 12).
8. Monitor the weight of the neutralizing (NaOH) solution
added, and use to calculate the neutralizing consumption rate
(g solution/min) throughout fermentation (see Note 13).
9. Stop the culture 1 h after reaching the maximum neutralizing

consumption rate (maximum of acidification rate), which cor-
responds to the beginning of the stationary phase, by reducing
the temperature to 4 C (see Note 14).
10. Aseptically transfer the cell suspension to sterile centrifugation
flasks (see Note 4).
11. Centrifuge (17,000 g, 4 C, 30 min) (see Note 15), and
remove the supernatant.
12. Mix cell pellets with protective medium at a ratio 1:2 (1 g of
cells and 2 g of protective medium). Suspend cells thoroughly
by using a sterile spatula and a vortex mixer till complete cell
resuspension (see Note 16).
3.2 Freeze-Drying 1. Aseptically aliquot the final protected bacterial suspensions

of Bacterial into labeled sterile vials (about 0.5 cm product fill depth)
Suspensions (see Notes 4 and 7).
2. Place the vials in a disinfected stainless steel tray and insert the
sterilized vial stoppers halfway into the vials, thus allowing sub-
sequent vapor evacuation during sublimation and desorption.
3. Place calibrated thermocouples in the bottom of some vials
(one per vial) and fix them with specific holders.
4. Distribute the vials with the thermocouples in different places
in the tray to take heterogeneity inside the batch into account
(see Note 17).
5. Load the trays containing the vials onto the shelf precooled
at 4 C (see Note 18).
6. Decrease shelf temperature to 50 C at 0.5 C/min and hold
this temperature until the sample temperature stabilizes (about
2 h) to ensure complete solidification of the product.
7. After freezing the samples, cool the condenser to 65 C,
switch on the vacuum pump, and decrease the chamber pres-
sure to 20 Pa. Then increase the shelf temperature to 20 C
at 0.25 C/min to initiate the ice sublimation phase.
8. Monitor the process variables during the freeze-drying process
(temperature, pressure) (Fig. 3) in order to detect any
dysfunction and also to identify the end of sublimation (see
Note 19). The end of the sublimation phase can be detected
by a significant increase of product temperature and a vapor
pressure decrease compared to values obtained during subli-
mation (Fig. 3) (see Note 20).
9. Three to six hours after the detection of the end of ice subli-
mation (after 2440 h of sublimation), increase the shelf tem-
perature to 25 C at 0.25 C/min to initiate the secondary
drying phase and reduce the chamber pressure to minimal value
(see Note 21).
40 1.2
20 1
0 Secondary drying (DII)

Temperature (C)
Pressure (mbar)
Desorption 0.8
20
0.6
40 Primary drying (DI)
End of sublimation
Sublimation
0.4
60
0.2
80 Freezing
100 0
0 5 10 15 20 25 30 35 40 45
Time (h)
Tfluide (C) Tfluide out (C) Tcold trap (C)

Th1 Th2 Th3
Th4 Pchamber, Pirani (mbar) Pchamber capacitance(mbar)
Fig. 3 Evolution of process variables with time during the freeze-drying process, including fluid shelf tempera-
ture, cold trap temperature, product temperature in four vials (Th14), and chamber pressure (and capacitance
manometer)
10. After 10 h of desorption, break the vacuum by injection of dry

air, stopper the vials rapidly (in the freeze-dryer if a stoppering
system available), and cap them.
11. Pack the vials under vacuum in aluminum bags and store at
temperatures lower than 15 C (generally 4 C) until their uti-
lization (see Note 22).
12. Aseptically add sterile nutrient broth to each vial for rehydra-
tion. The volume to be added must be equal to the initial vol-
ume of the cell suspension prior to freeze-drying. Vortex closed
vial for 5 min (see Notes 4 and 23).
4 Notes
1. MRS and M17 broths have become standard culture media for
lactobacilli and streptococci and lactococci, respectively. They
are available in dried form and made up by suspending in
demineralized water. Other pre-culture media can be used,
depending on the strain or the experimental requirements
(e.g., lactococci are currently cultivated in M17 broth, but
Elliker broth can also be used).
2. Mild whey powder (pH 6.5 and low salt content) is better
adapted for LAB fermentation than acid whey (pH 4.5, high
salt content), especially due to its lower ionic concentration.
3. Concentration of protective molecules can be modified according

to strain requirements. Increasing antioxidant concentration
up to 10 g/L may be necessary, particularly for strains that are
sensitive to oxidation reactions. Polymers may be added to
ensure structure preservation if there is a risk of collapse (e.g.,
if sublimation and/or storage temperature is higher than those
recommended here). In any case, the concentration of solutes
in the protective medium should not be higher than 250 g/L.
4. Use a laminar hood for preparing sterile media and for asepti-
cally transferring cell samples. If laminar hood is not available,
work close to a gas burner.
5. For freeze-drying of LAB starters, freeze-dryer sterilization is
not compulsory (like in pharmaceutical applications). However,
we advise a thorough disinfection of the chamber, shelves, and
trays between batches for avoiding cross contamination.
6. Fine thermocouples should be chosen (e.g., type T) to mini-
mize perturbation of the sublimation process. At least three to
four thermocouples should be used for product temperature
monitoring. They need to be disinfected (e.g., ethanol and
sterile water) and calibrated before utilization according to
manufacturers recommendations.
7. Different sensors make it possible to determine the end of sub-
limation: moisture sensors (aluminum oxide probe) and com-
parative pressure measurement (Pirani gauge versus capacitance
manometer) (illustrated in Fig. 3) [24].
8. Stainless steel trays are usually used for LAB suspensions in
industrial freeze-drying conditions. Other containers (plastic,
aluminum) may also be used provided the fill depth is kept
between 0.5 and 1 cm. Vials allow to study different experi-
mental conditions with low biomass requirements.
9. For specific purposes other media can be used for fermentation.
LAB usually grow on different kinds of sugar sources (lac-
tose, glucose, etc.), but need nitrogen sources (yeast extract,
amino acids, etc.), and may have specific growth factor
requirements (vitamins, salts, surfactant, etc.). Optimization
of the fermentation medium composition may be necessary
for specific applications.
10. The pre-culture inoculum concentration and volume have to
be adapted in order to have several duplication times of cells
and to provide enough cell concentration for the following
culture.
11. In order to avoid contamination, an independent culture flask
is used for optical density measurements.
12. Fermentation medium and growth conditions (temperature,
pH, O2 concentration, harvesting growth stage) for optimal
preservation of LAB functionality and viability following

freeze-drying vary among LAB strains. Generally, fermenta-
tion conditions leading to high biomass yield are used.
However, if high losses of viability and/or acidification activity
(or other functionality) are observed, then the optimization of
the fermentation conditions should be considered to minimize
damage during freeze-drying and storage.
13. Use, if available, a software allowing controlling and visualiz-
ing fermentation variables and neutralizing solution consump-
tion rate (g solution/min) throughout fermentation thus
facilitating identification of the harvesting time (beginning of
stationary phase, corresponding to about 1 h after the maximal
neutralizing consumption rate) [25, 26]. Another alternative is
aseptic sampling during the fermentation to identify the begin-
ning of the stationary phase by optical density measurements.
14. Circulate cold water through the fermentor jacket (e.g., water-
ice mixture) to cool the fermentor at the end of fermentation.
15. Centrifugation is the method currently employed for concen-
trating lactic acid starters. However, some cells are sensitive to
this process. If this is the case, centrifugation conditions have
to be optimized by modifying the centrifuge force and dura-
tion. As an alternative, membrane processes such as cross-flow
microfiltration can also be used if available [27].
16. Maintain at 4 C during all intermediate steps of the pro-
cesses to limit cell degradation reactions [28]. In an appro-
priate medium such as the protective one, LAB may also
restart fermentation with the concomitant physiological
state modification of cells.
17. Locate vials with thermocouples at different places in the freeze-
dryer. If only one shelf is used, place three in the center of the
tray and two in the borders of the batch to take heterogeneity
of the product temperature within the batch into account.
18. Samples already frozen (because of particular experimental
requirements or freeze-dryer availability) must be transferred
to the freeze-drier after precooling the shelf to 50 C to avoid
thawing of the samples. Other freezing conditions include
cooling in an air chamber at 80 C or immersion in liquid
nitrogen [29].
19. Use, if available, software allowing data acquisition and repre-
sentation of process variables during the process: shelf tempera-
ture, product temperature, cold trap temperature, and chamber
pressure. This facilitates identification of the end of each of the
process steps (freezing, primary drying, secondary drying) and
recording of any anomalous behavior during the cycle.
20. The Pirani vacuum gauge measures gas thermal conductivity
and responds to the gaseous composition of the chamber.
The capacitance manometer indicates the absolute vacuum

and is not affected by vapor pressure. When the two gauges
exhibit the same value, the sublimation is considered to be
completed (Fig. 3) [24].
21. Take a safety margin for secondary drying initiation in order to
avoid collapse of the samples. Thermocouples accelerate the
sublimation process and generally give a premature sublima-
tion end point. Moreover, since thermocouples are distributed
at different places of the freeze-dryer, the end of the sublima-
tion is frequently different from one thermocouple to another.
Safety margins can be considered: 46 h after the increase of
product temperature or 34 h after obtaining equal values for
Pirani gauge and capacitance manometer.
22. Storage at lower temperatures (20 C, 80 C) should be
chosen for long storage periods (>1 month). In the case of
oxygen-sensitive strains, dry inert gas such as nitrogen can be
used for controlling pressure and for breaking vacuum at the
end of the freeze-drying process and also for conditioning the
freeze-dried powders. If vacuum packaging is not feasible, seal
the aluminum bags containing samples after injection of nitro-
gen gas, thus limiting oxidation reactions [30].
23. For rehydration current practice is to rehydrate in sterilized
skim milk for a fast activity recovery. However, other rehydra-
tion media (demineralized water, saline solution, culture
broth) can be used with variable recovery according to the
strains [31]. The rehydration temperature usually corresponds
to the optimal growth temperature.
Acknowledgment
This work was supported by the National Institute of Agronomic

Research (Paris, France) and by the Paris Institute of Technology
for Life, Food and Environmental Sciences (AgroParisTech, Paris,
France).
References
1. Jennings T (2002) Lyophilization: introduction 4. Fonseca F, Passot S, Cunin O, Marin M
and basic principles. CRC, Washington, DC (2004) Collapse temperature of freeze-dried
2. Roy ML, Pikal MJ (1989) Process control in Lactobacillus bulgaricus suspensions and
freeze drying: determination of the end point protective media. Biotechnol Prog 20:
of sublimation drying by an electronic moisture 229238
sensor. J Parenter Sci Technol 43:6066 5. Fonseca F, Passot S, Lieben P, Marin M (2004)
3. Searles JA, Carpenter JF, Randolph TW (2001) Collapse temperature of bacterial suspensions:
The ice nucleation temperature determines the the effect of cell type and concentration. Cryo
primary drying rate of lyophilization for sam- Letters 25:425434
ples frozen on a temperature-controlled shelf. 6. Passot S, Cenard S, Douania I, Trla IC,
J Pharm Sci 90:860871 Fonseca F (2012) Critical water activity and
amorphous state for optimal preservation of 19. Santivarangkna C, Kulozik U, Foerst P (2006)
lyophilised lactic acid bacteria. Food Chem Effect of carbohydrates on the survival of
132:16991705 Lactobacillus helveticus during vacuum drying.
7. Font de Valdez G, Savoy de Giori G, Pesce de Lett Appl Microbiol 42:271276
Ruiz Holgado A, Oliver G (1983) Comparative 20. Castro HP, Teixeira PM, Kirby R (1996)
study of the efficiency of some additives in pro- Changes in the cell membrane of Lactobacillus
tecting lactic acid bacteria against freeze- bulgaricus during storage following freeze-
drying. Cryobiology 20:560566 drying. Biotechnol Lett 18:99104
8. Miyamoto-Shinohara Y, Sukenobe J, Imaizumi 21. Teixeira P, Castro H, Kirby R (1996) Evidence
T, Nakahara T (2006) Survival curves for of membrane lipid oxidation of spray-dried
microbial species stored by freeze-drying. Lactobacillus bulgaricus during storage. Lett
Cryobiology 52:2732 Appl Microbiol 22:3438
9. Crowe JH, Crowe LM, Carpenter JF, Rudolph 22. Schoug A, Olsson J, Carlfors J, Schnurer J,
AS, Wistrom CA, Spargo BJ, Anchordoguy TJ Hakansson S (2006) Freeze-drying of
(1988) Interactions of sugars with membranes. Lactobacillus coryniformis Si3: effects of
Biochim Biophys Acta 947:367384 sucrose concentration, cell density, and freez-
10. Crowe JH, Crowe LM, Carpenter JF (1993) ing rate on cell survival and thermophysical
Preserving dry biomaterials: the water replace- properties. Cryobiology 53:119127
ment hypothesis, part 1. Biopharmacology 6: 23. Zhang J, Du GC, Zhang YP, Liao XY, Wang
2833 M, Li Y, Chen J (2010) Glutathione protects
11. Leslie SB, Israeli E, Lighthart B, Crowe JH, Lactobacillus sanfranciscensis against freeze-
Crowe LM (1995) Trehalose and sucrose pro- thawing, freeze-drying, and cold treatment.
tect both membranes and proteins in intact Appl Environ Microbiol 76:29892996
bacteria during drying. Appl Environ Microbiol 24. Passot S, Trla IC, Marin M, Galan M, Morris
61:35923597 GJ, Fonseca F (2009) Effect of controlled ice
12. Carvalho AS, Silva J, Ho P, Teixeira P, Malcata nucleation on primary drying stage and protein
FX, Gibbs P (2004) Relevant factors for the recovery in vials cooled in a modified freeze-
preparation of freeze-dried lactic acid bacteria. dryer. J Biomech Eng 131:07451110745115
Int Dairy J 14:835847 25. Fonseca F, Bal C, Corrieu G (2000) Method
13. Kurtmann L, Carlsen CU, Skibsted LH, Risbo of quantifying the loss of acidification activity
J (2009) Water activity-temperature state dia- of lactic acid starters during freezing and fro-
grams of freeze-dried Lactobacillus acidophilus zen storage. J Dairy Res 67:8390
(La-5): influence of physical state on bacterial 26. Fonseca F, Marin M, Morris GJ (2006)
survival during storage. Biotechnol Prog 25: Stabilization of frozen Lactobacillus delbrueckii
265270 subsp. bulgaricus in glycerol suspensions: freez-
14. Castro HP, Teixeira PM, Kirby R (1995) ing kinetics and storage temperature effects.
Storage of lyophilized cultures of Lactobacillus Appl Environ Microbiol 72:64746482
bulgaricus under different relative humidities 27. Streit F, Corrieu G, Bal C (2010) Effect of
and atmospheres. Appl Microbiol Biotechnol centrifugation conditions on the cryotolerance
44:172176 of Lactobacillus bulgaricus CFL1. Food
15. Champagne CP, Gardner N, Brochu E, Bioprocess Technol 3:3642
Beaulieu Y (1991) The freeze-drying of lactic 28. Fonseca F, Bal C, Mihoub F, Marin M,
acid bacteria. A review. Can Inst Food Sci Corrieu G (2003) Improvement of cryopreser-
Technol 24:118128 vation of Lactobacillus delbrueckii subsp. bul-
16. Meng XC, Stanton C, Fitzgerald GF, Daly C, garicus CFL1 with additives displaying different
Ross RP (2008) Anhydrobiotics: the challenges protective effects. Int Dairy J 67:8390
of drying probiotic cultures. Food Chem 106: 29. Trla IC, Passot S, Marin M, Fonseca F
14061416 (2009) Model for heat and mass transfer in
17. Tymczyszyn EE, Sosa N, Gerbino E, Hugo A, freeze-drying pellets. J Biomech Eng 131:
Gomez-Zavaglia A, Schebor C (2012) Effect 07450110745017
of physical properties on the stability of 30. Chavez BE, Ledeboer AM (2007) Drying of
Lactobacillus bulgaricus in a freeze-dried probiotics: optimization of formulation and
galacto-oligosaccharides matrix. Int J Food process to enhance storage survival. Drying
Microbiol 155:217221 Technol 25:11931201
18. Santivarangkna C, Aschenbrenner M, Kulozik 31. Font de Valdez G, Savoy de Giori G, Pesce de
U, Foerst P (2011) Role of glassy state on sta- Ruiz Holgado A, Oliver G (1985) Rehydration
bilities of freeze-dried probiotics. J Food Sci conditions and viability of freeze-dried lactic
76:R152R156 acid bacteria. Cryobiology 22:574577
Chapter 25
Freeze-Drying of Mammalian Sperm

Levent Keskintepe and Ali Eroglu
Abstract
Long-term preservation of mammalian sperm at suprazero temperatures is desired to save storage and
space costs as well as to facilitate transport of preserved samples. This can be accomplished by the freeze-
drying of sperm samples. Although freeze-drying results in immotile and membrane-compromised sperm,
intracytoplasmic sperm injection (ICSI) can be used to introduce such an immotile sperm into an oocyte
and thus start the fertilization process. So far, it has been shown that improved freeze-drying protocols
preserve chromosomal integrity and oocyte-activating factor(s) at 4 C for several years and at ambient
temperature for approximately 1 month, which permits shipping freeze-dried samples at ambient tempera-
ture. This chapter concisely reviews freeze-drying of mammalian sperm first and then presents a simple
freeze-drying protocol.
Key words Sperm, Spermatozoa, Freezing, Freeze-drying, Cryopreservation, Long-term storage,

Intracytoplasmic sperm injection, ICSI, Cryoprotectant, Trehalose
1 Introduction
Although interest in sperm preservation goes back several centuries

[1], reliable cryopreservation of sperm was achieved for the first
time in 1949 when Polge et al. serendipitously discovered cryopro-
tective properties of glycerol while working with fowl sperm [2].
This discovery ultimately led to successful methods for low-tem-
perature storage of bull [3, 4] and human sperm [5]. Initially, dry
ice (78.5 C) was used to cryopreserve and store sperm, which
resulted in declined sperm motility with increased storage time. In
1963, this problem was overcome by storing sperm in liquid nitro-
gen (LN2, 196 C) [6]. These developments in sperm cryopreser-
vation greatly facilitated both treatment of human infertility and
artificial insemination and thus genetic improvement of livestock.
Other applications of sperm cryopreservation include (1) fertility
preservation for male cancer patients, (2) conservation of endan-
gered species and maintenance of genetic diversity, and (3) bank-
ing of genetically engineered animal models for human diseases.
489
490 Levent Keskintepe and Ali Eroglu
Meanwhile, cryopreservation of sperm from different species has

been established although success rates greatly vary among species
and are still not satisfactory in some cases [1, 7, 8].
While long-term storage of mammalian sperm in LN2 is cur-
rently the method of choice, the ability to preserve sperm at higher
temperatures (ideally at ambient temperature) is preferred for a
number of reasons. It would significantly reduce the storage costs
and spaces, facilitate sample transport while reducing associated
expenses, and avoid the viral contamination risk among samples [9,
10]. Furthermore, LN2 is not readily accessible in some countries.
Consequently, there has been a long-term interest in cell preserva-
tion methods that would allow storage of samples at suprazero
temperatures. Freeze-drying represents such a method and brings
samples to a dried state through a multistep process involving pri-
mary and secondary drying and two phase transitions. For the pri-
mary drying, samples are first converted from a liquid phase into a
solid phase (ice) by cooling to below their eutectic temperature (see
Fig. 1, path A to B). Essentially, this step results in a physical
separation of water as ice from the solids present in samples. Next,
a vacuum is applied to the sample environment to lower the partial
pressure of water below the triple point (see Fig. 1, path B to C)
and thus to remove the frozen water by directly transitioning it
t
al oin
m ng p al int
r
No ezi o rm g po
N ilin
fre bo
Liquid
1.013
B A
Pressure (Bar)
Solid
(ice) Triple point
Gas
0.006
(water vapor)
C E
D
0 0.01 100
Temperature (C)
Fig. 1 Hypothetical freeze-drying paths superimposed on water phase diagram

(not in scale). Freeze-drying involves first converting specimens from a liquid
phase to solid (ice) by freezing to below their eutectic temperature (path A to B),
then lowering pressure around specimens below the triple point by applying vac-
uum (path B to C), and finally converting the solid (ice) to gas (water vapor) by either
further lowering the pressure below the sublimation line (path C to D) or controlled
heating of specimens under vacuum (path C to E). The triple point refers to the
intersection on the phase diagram, where three different phases can coexist
Freeze-Drying of Sperm 491
from solid to gas (water vapor) without passing through the inter-
mediate liquid phase again. This latter phase transition is known as
sublimation and can be accomplished in theory by either further
lowering the pressure (see Fig. 1, path C to D) or supplying heat to
the system (see Fig. 1, path C to E). The latter path is more practi-
cal and commonly used in freeze-drying protocols to sublimate
ice. To achieve efficient sublimation, a concentration gradient of
water vapor has to be accomplished between the drying front/
chamber and condenser, wherein the temperature is considerably
lower than that in the drying front/chamber and thus water vapor
migrates to the condenser and freezes there. The primary drying
ends by removal of the frozen unbound water and results in around
810 % moisture content depending on sample composition. To
ensure sample stability, the remaining unfrozen bound water can
be removed by desorption in the secondary drying step, wherein
the samples are further heated under lowest attainable vacuum to
form water vapor from the bound water. The resulting samples can
potentially be stored at ambient temperature and be reconstituted
by adding water, whenever needed.
Early studies on freeze-drying of sperm reported encouraging
results including up to 50 % post-rehydration motility with fowl
sperm [2] and bull sperm [11], birth of 12 litters with rabbit sperm
[12], and pregnancies after artificial insemination with freeze-dried
bull sperm [13, 14]. However, subsequent studies failed to repro-
duce the initial success [1519] and led to loss of interest in freeze-
drying of sperm for a while. A breakthrough came in 1998 when
Wakayama and Yanagimachi demonstrated development of normal
mice from oocytes fertilized with freeze-dried sperm [20].
Interestingly, their freeze-drying approach was not extraordinary
and resulted in 0 % motile sperm after rehydration. In fact, all rehy-
drated sperm were dead based on a red/green dual fluorescence
viability assay. The use of intracytoplasmic sperm injection (ICSI)
to introduce freeze-dried immotile sperm into oocytes was the key
to their success. Clearly, their results showed that freeze-drying of
mouse sperm in a cell culture medium (i.e., DMEM and CZB) was
sufficient to preserve integrity of sperm DNA and oocyte-activating
factor(s) for up to 3 months at 4 C and for a few weeks at ambient
temperature. It should be noted that in an earlier study, similar
findings (i.e., development to term) were reported by Goto et al.
[21] after microinjection of bovine sperm that were dead as a result
of cryoprotectant-free freezing and thawing. Taken together, these
studies revealed that to support full-term development, mamma-
lian sperm does not need to be viable in the conventional sense as
long as its genetic integrity is intact. The breakthrough study of
Wakayama and Yanagimachi revived the interest in freeze-drying
of mammalian sperm, and several studies reporting successful
results with freeze-dried sperm from different species followed (see
Table 1 for details). The species include rat [22, 23], hamster [24],
rabbit [25], bovine [2628], porcine [29, 30], equine [31], canine
[32], rhesus monkey [33], and human [34, 35]. While these stud-
ies together proved the feasibility of freeze-drying of mammalian
sperm, efforts to improve the procedure are still ongoing. It has
been shown that stability of freeze-dried mouse sperm can be
improved by using a simple TRIS-buffered freeze-drying solution
containing a calcium chelator such as EGTA and EDTA [36, 37].
The chelation of calcium seems to be important for inhibition of
endonucleases that compromise DNA integrity [36, 38].
Furthermore, a slightly alkaline freeze-drying solution (pH 8)
proved to be better than near-neutral or acidic (pH 7.46.0) solu-
tions in terms of chromosomal integrity and developmental poten-
tial [39]. The outcome of freeze-drying also appears to be affected
by the maturation status of sperm. Unlike mature epididymal
sperm, immature testicular sperm lacks extensive disulfide bonds in
its chromatin, which seems to make its nucleus vulnerable to
stresses associated with freeze-drying. Nevertheless, immature
sperm can be made resistant to such stresses by oxidizing its free
thiols to disulfides using diamide supplementation [40]. As a result
of these improvements, it was possible to store freeze-dried rat
sperm at 4 C for up to 5 years without significant decline in its
ability to support full-term development [23]. In contrast, long-
term storage of freeze-dried sperm at ambient temperature still
remains unsatisfactory. Both modeling and experimental data sug-
gest that chromosomal integrity of freeze-dried sperm is compro-
mised when stored at ambient temperature for more than 1 month
[41, 42]. It is likely that sugar- and protein-packed sperm nucleus
is stabilized in a glass-like vitrified state upon removal of water by
freeze-drying. However, this may not be sufficient for its long-
term preservation at ambient temperature without stabilizing its
surroundings in a vitrified state as well. Indeed, improved DNA
integrity has been reported after adding trehalose, a good glass
former, to freeze-drying solution [30, 43]. Taken together, further
comprehensive research is needed to achieve long-term preserva-
tion of mammalian sperm at ambient temperature.
In the following, we present a freeze-drying protocol based on
utilization of a manifold-type freeze dryer. With some minor modi-
fications, it should be adaptable to freeze-drying of mammalian
sperm using a shelf freeze dryer.
2 Materials
In addition to standard laboratory equipment and supplies, the fol-

lowing items are required for freeze-drying:
1. Manifold-type freeze dryer such as Flexi-Dry (FTS Systems,
Inc. Stoner Ridge, NY 12484) and FreeZone 77530-00
(Labconco Corporation, Kansas City, MO, USA) (see Note 1).
Table 1
Summary of some significant results obtained using freeze-dried sperm from different mammalian
species
Storage Storage Blastocyst Live

Species Reference temperature (C) duration Fertilization (%) formation (%) offspring (%)
Mouse [20] FC ND ND ND
4 3 months 100 92 28
25 1 month 98 76 18
[36] FC 100 86 ND
4 34 weeks 84 72 ND
[45] FC ND ND 58
4 1.5 years ND ND 21
[41] FC 95 59 ND
4 3 months 94 21 ND
4 6 months 91 13 ND
[37] FC ND ND ND
4 1 year 96 ND 65
[46] FC ND ND ND
4 6 months 98 50 21
[47] FC 100 ND 39
4 3 years 95 ND 33
Rat [22] FC 99 27 35
4 26 weeks 94 30 14
[48] CC 1 year 79 ND 36
4 1 year 70 ND 7
25 1 year 75 ND 0
[49] FC 97 50 ND
4 1 year 56 ND 16
[23] FC ND ND ND
4 5 years ND ND 11
Rabbit [25] FC 97 30 ND
4 2 years 70 24 0.4
Canine [32] FC 62 ND ND
80 C 1 month 92 ND ND
Bovine [26] CC 82 34 ND
4 13 months 74 30 ND
[27] CC 54 38 ND
4 3 months 58 34 ND
[28] FC 67 21 ND
4 1 week 36 1 ND
Porcine [29] FC 83 17 ND
4 1 month 83 11 ND
25 1 month 82 0 ND
[30] FC 72 33 ND
4 Not given 61 12 ND
Equine [31] CC 81 28 ND
4 3.5 78 6 29
Monkey [33] FC 63 0 ND
23 2 months 73 0 ND
Since it is impossible to fit all published data in a single table, we tried to select some representative ones. FC fresh
control, CC cryopreserved control, ND not determined
2. Long-necked borosilicate glass ampules (e.g., W651506,

Wheaton, Millville, NJ, USA). These glass ampules can be
autoclaved or heat sterilized in advance (see Note 2).
3. Air/gas torch for flame sealing.
4. TRIS-buffered freeze-drying solution containing a calcium
chelator: The two most commonly used freeze-drying solu-
tions are (a) 10 mM TrisHCl containing 1 mM EDTA (pH 8)
and (b) 10 mM TrisHCl containing 50 mM NaCl and 50 mM
EGTA (pH 8). We suggest selecting one of these solutions and
preparing its concentrated 2 stock solution in ultrapure water.
Next, its pH can be adjusted to 8 and then diluted to its 1
solution by adding ultrapure water. The resulting final solution
is stable at ambient temperature or 4 C once sterilized using a
0.22-m filter.
3 Methods
3.1 Freeze-Drying 1. Complete species-specific steps of sperm collection. The fol-

lowing includes preparation of murine sperm for freeze-drying
(see Note 3 for other species).
2. Aseptically dissect two cauda epididymides and place them in
an empty sterile culture dish.
3. Gently puncture the cauda epididymides with a 30-G needle
under a dissecting microscope and transfer them to the bottom
of a 1.5-mL microcentrifuge tube containing 1 mL freeze-dry-
ing solution (see Note 4).
4. Incubate the microcentrifuge tube at 37 C for 10 min to dis-
perse the sperm.
5. Carefully aspirate the top 0.8 mL sperm suspension and dis-
tribute into sterile glass ampules as 100 L aliquots.
6. Turn on your freeze dryer and make it ready for the freeze-
drying process depending on specifications of your freeze dryer.
7. Partially immerse glass ampules containing 100 L of sperm
suspension into liquid nitrogen for 3060 s to freeze the sperm
samples.
8. Quickly connect each ampule to your manifold freeze dryer.
9. Freeze-dry the sperm samples for 4 h under vacuum pressure
at around 0.03 mBar (0.003 kPa).
10. Flame seal each ampule using a gas torch under vacuum (0.03
mBar) after freeze-drying.
11. Store the sealed ampules at 4 C in the dark to avoid any
potential light-induced damage (see Note 5).
3.2 Rehydration 1. Open an ampule, preferentially under a laminar flow hood,

and Preparation and rehydrate the sperm by adding 100 L of sterile ultra-
of Freeze-Dried Sperm pure water.
for ICSI 2. Aspirate 5 L sperm suspension and mix it with a small drop
(2030 L) of a HEPES-buffered medium (e.g., HEPES-
CZB) containing 12 % (w/v) polyvinylpyrrolidone (PVP; Mr
360,000; ICN Pharmaceuticals).
3. For ICSI, select only sperm displaying normal morphology
(see Note 6).
4 Notes
1. For research purposes, a shelf freeze dryer may serve better

because it would allow more precise controlling of different
parameters such as holding temperature, heating during primary
and secondary drying, and avoiding collapse temperature.
2. Pre-scored ampules with a barcode might be helpful for error-free
maintenance of sample inventory.
3. Motile sperm from different species can be separated by a
discontinuous Percoll density gradient [44] or similar gradi-
ent systems [26]. Alternatively, an aliquot of liquefied semen
can be transferred to the bottom of a 1.5-mL microcentri-
fuge tube containing 1 mL of freeze-drying solution and
then be incubated at 37 C for 10 min for dispersion. Next,
the upper part of the freeze-drying solution containing
motile sperm can carefully be aspirated and distributed into
freeze-drying glass ampules [34]. It might be helpful to
determine the concentration of the separated motile sperm
and adjust it to 0.5 105 sperm/mL before its distribution
into the glass ampules.
4. It is important to work quickly at this step in order to prevent
evaporative drying of cauda epididymides and sperm. Also,
only sterile solutions, tools, and supplies should be used at this
step and throughout the procedure.
5. Freeze-dried sperm samples can be shipped and stored at ambi-
ent temperature for up to few weeks without any significant
adverse effect.
6. Before ICSI, it is important to wash sperm in another drop of
the HEPES-buffered medium containing 12 % PVP to mini-
mize the introduction of the calcium chelator (i.e., EDTA or
EGTA) into the oocyte cytoplasm because it may interfere
with normal activation of the microinjected oocyte.
References
1. Walters EM, Benson JD, Woods EJ, Critser JK 18. Nei T, Nagase H (1961) Attempts to freeze-
(2009) The history of sperm cryopreservation. dry bull spermatozoa. Low Temp Sci Ser B
In: Pacey AA, Tomlinson MJ (eds) Sperm 19:107115
banking: theory and practice. Cambridge 19. Meryman HT, Kafig E (1963) Freeze-drying of
University Press, Cambridge, UK, pp 117 bovine spermatozoa. J Reprod Fertil 5:8794
2. Polge C, Smith AU, Parkes AS (1949) Revival 20. Wakayama T, Yanagimachi R (1998)
of spermatozoa after vitrification and dehydra- Development of normal mice from oocytes
tion at low temperatures. Nature 164:666 injected with freeze-dried spermatozoa. Nat
3. Polge C (1952) Fertilizing capacity of bull Biotechnol 16:639641
spermatozoa after freezing at 79 degrees C. 21. Goto K, Kinoshita A, Takuma Y, Ogawa K
Nature 169:626627 (1990) Fertilisation of bovine oocytes by the
4. Stewart DL (1951) Storage of bull spermato- injection of immobilised, killed spermatozoa.
zoa at low temperatures. Vet Rec 63:6566 Vet Rec 127:517520
5. Bunge RG, Sherman JK (1953) Fertilizing 22. Kaneko T, Kimura S, Nakagata N (2007)
capacity of frozen human spermatozoa. Nature Offspring derived from oocytes injected with
172:767768 rat sperm, frozen or freeze-dried without cryo-
6. Sherman JK (1990) Cryopreservation of protection. Theriogenology 68:10171021
human semen. In: Keel BA, Webster BW (eds) 23. Kaneko T, Serikawa T (2012) Successful long-
Handbook of the laboratory diagnosis and term preservation of rat sperm by freeze-
treatment of infertility. CRC, Boston, MA, drying. PLoS One 7:e35043
pp 229260 24. Muneto T, Horiuchi T (2011) Full-term devel-
7. Curry MR (2000) Cryopreservation of semen opment of hamster embryos produced by
from domestic livestock. Rev Reprod 5:4652 injecting freeze-dried spermatozoa into
8. Mazur P, Leibo SP, Seidel GE (2008) oocytes. J Mamm Ova Res 28:3239
Cryopreservation of the germplasm of animals 25. Liu JL, Kusakabe H, Chang CC, Suzuki H,
used in biological and medical research: impor- Schmidt DW, Julian M, Pfeffer R, Bormann
tance, impact, status, and future directions. CL, Tian XC, Yanagimachi R, Yang XZ (2004)
Biol Reprod 78:212 Freeze-dried sperm fertilization leads to full-
9. Tedder RS, Zuckerman MA, Goldstone AH, term development in rabbits. Biol Reprod
Hawkins AE, Fielding A, Briggs EM, Irwin D, 70:17761781
Blair S, Gorman AM, Patterson KG et al (1995) 26. Keskintepe L, Pacholczyk G, Machnicka A,
Hepatitis B transmission from contaminated Norris K, Curuk MA, Khan I, Brackett BG
cryopreservation tank. Lancet 346:137140 (2002) Bovine blastocyst development from
10. Bielanski A, Nadin-Davis S, Sapp T, Lutze- oocytes injected with freeze-dried spermato-
Wallace C (2000) Viral contamination of zoa. Biol Reprod 67:409415
embryos cryopreserved in liquid nitrogen. 27. Martins CF, Bao SN, Dode MN, Correa GA,
Cryobiology 40:110116 Rumpf R (2007) Effects of freeze-drying on
11. Meryman HT, Kafig E (1989) Survival of sper- cytology, ultrastructure, DNA fragmentation,
matozoa following drying. Nature 184:470471 and fertilizing ability of bovine sperm.
Theriogenology 67:13071315
12. Yushchenko NP (1957) Proof of the possibility
of preserving mammalian spermatozoa in a 28. Hara H, Abdalla H, Morita H, Kuwayama M,
dried state. Proc Lenin Acad Agr Sci 22:3740 Hirabayashi M, Hochi S (2011) Procedure for
bovine ICSI, not sperm freeze-drying, impairs
13. Meryman HT (1960) Drying of living
the function of the microtubule-organizing
mammalian cells. Ann N Y Acad Sci 85:
center. J Reprod Dev 57:428432
729734
29. Kwon IK, Park KE, Niwa K (2004) Activation,
14. Larson EV, Graham EF (1976) Freeze-drying
pronuclear formation, and development in vitro
of spermatozoa. Dev Biol Stand 36:343348
of pig oocytes following intracytoplasmic injec-
15. Sherman JK (1954) Freezing and freeze-drying tion of freeze-dried spermatozoa. Biol Reprod
of human spermatozoa. Fertil Steril 5:357371 71:14301436
16. Bialy G, Smith VR (1957) Freeze-drying of 30. Men NT, Kikuchi K, Nakai M, Fukuda A,
bovine spermatozoa. J Dairy Sci 40:739745 Tanihara F, Noguchi J, Kaneko H, Linh NV,
17. Saacke RG, Almquist JO (1961) Freeze-drying Nguyen BX, Nagai T, Tajima A (2013) Effect
of bovine spermatozoa. Nature 192:995996 of trehalose on DNA integrity of freeze-dried
boar sperm, fertilization, and embryo develop- solution on the chromosome integrity and
ment after intracytoplasmic sperm injection. developmental ability of mouse spermatozoa.
Theriogenology 80:10331044 Biol Reprod 68:136139
31. Choi YH, Varner DD, Love CC, Hartman DL, 40. Kaneko T, Whittingham DG, Overstreet JW,
Hinrichs K (2011) Production of live foals via Yanagimachi R (2003) Tolerance of the mouse
intracytoplasmic injection of lyophilized sperm sperm nuclei to freeze-drying depends on their
and sperm extract in the horse. Reproduction disulfide status. Biol Reprod 69:18591862
142:529538 41. Kawase Y, Araya H, Kamada N, Jishage K,
32. Watanabe H, Asano T, Abe Y, Fukui Y, Suzuki Suzuki H (2005) Possibility of long-term pres-
H (2009) Pronuclear formation of freeze-dried ervation of freeze-dried mouse spermatozoa.
canine spermatozoa microinjected into mouse Biol Reprod 72:568573
oocytes. J Assist Reprod Genet 26:531536 42. Kaneko T, Nakagata N (2005) Relation
33. Sanchez-Partida LG, Simerly CR, Ramalho- between storage temperature and fertilizing
Santos J (2008) Freeze-dried primate sperm ability of freeze-dried mouse spermatozoa.
retains early reproductive potential after intra- Comp Med 55:140144
cytoplasmic sperm injection. Fertil Steril 89: 43. Martins CF, Dode MN, Bao SN, Rumpf R
742745 (2007) The use of the acridine orange test and
34. Kusakabe H, Yanagimachi R, Kamiguchi Y the TUNEL assay to assess the integrity of
(2008) Mouse and human spermatozoa can be freeze-dried bovine spermatozoa DNA. Genet
freeze-dried without damaging their chromo- Mol Res 6:94104
somes. Hum Reprod 23:233239 44. Donnelly ET, McClure N, Lewis SE (2001)
35. Gianaroli L, Magli MC, Stanghellini I, Crippa Cryopreservation of human semen and pre-
A, Crivello AM, Pescatori ES, Ferraretti AP pared sperm: effects on motility parameters and
(2012) DNA integrity is maintained after DNA integrity. Fertil Steril 76:892900
freeze-drying of human spermatozoa. Fertil 45. Ward MA, Kaneko T, Kusakabe H, Biggers JD,
Steril 97:10671073 Whittingham DG, Yanagimachi R (2003)
36. Kusakabe H, Szczygiel MA, Whittingham DG, Long-term preservation of mouse spermatozoa
Yanagimachi R (2001) Maintenance of genetic after freeze-drying and freezing without cryo-
integrity in frozen and freeze-dried mouse protection. Biol Reprod 69:21002108
spermatozoa. Proc Natl Acad Sci U S A 98: 46. Kawase Y, Tachibe T, Jishage K, Suzuki H
1350113506 (2007) Transportation of freeze-dried mouse
37. Kaneko T, Nakagata N (2006) Improvement in spermatozoa under different preservation con-
the long-term stability of freeze-dried mouse ditions. J Reprod Dev 53:11691174
spermatozoa by adding of a chelating agent. 47. Kaneko T, Serikawa T (2012) Long-term pres-
Cryobiology 53:279282 ervation of freeze-dried mouse spermatozoa.
38. Nakai M, Kashiwazaki N, Takizawa A, Cryobiology 64:211214
Maedomari N, Ozawa M, Noguchi J, Kaneko 48. Hochi S, Watanabe K, Kato M, Hirabayashi M
H, Shino M, Kikuchi K (2007) Effects of che- (2008) Live rats resulting from injection of
lating agents during freeze-drying of boar sper- oocytes with spermatozoa freeze-dried and stored
matozoa on DNA fragmentation and on for one year. Mol Reprod Dev 75:890894
developmental ability in vitro and in vivo after 49. Kaneko T, Kimura S, Nakagata N (2009)
intracytoplasmic sperm head injection. Zygote Importance of primary culture conditions for
15:1524 the development of rat ICSI embryos and
39. Kaneko T, Whittingham DG, Yanagimachi R long-term preservation of freeze-dried sperm.
(2003) Effect of pH value of freeze-drying Cryobiology 58:293297
Chapter 26
Freeze-Drying of Decellularized Heart Valve Tissues

Willem F. Wolkers and Andres Hilfiker
Abstract
Decellularized xeno-antigen-depleted porcine pulmonary heart valves tissues may be used as matrix
implants for patients with malfunctioning heart valves. Decellularized tissues are biological scaffolds com-
posed of extracellular matrix components. Biological scaffolds closely resemble properties of native tissue,
but lack immunogenic factors of cellular components. Decellularized heart valve scaffolds need to be
stored to be readily available whenever needed. Scaffolds can be stored at reduced supra-zero tempera-
tures, cryopreserved or freeze-dried. The advantage of freeze-drying is that it allows long-term storage
at room temperature. This chapter outlines the entire process from decellularization to freeze-drying to
obtain dry decellularized porcine heart valve scaffolds.
Key words Freeze-drying, Sucrose, Pulmonary heart valve conduits, Matrix implants, Anhydrobiosis
1 Introduction
Valve-related diseases result in a reduction of the forward blood

flow limiting oxygenation of tissues. Heart valve matrix implants
can be used to replace failing heart valves. Heart valve implants can
be produced using decellularization approaches [13]. Decellu-
larization refers to the removal of cells and cellular remnants while
leaving biological scaffold material composed of extracellular
matrix components. The obtained extracellular matrix closely
resembles properties of native tissue but lacks immunogenic factors
of cellular components leading to immune responses in patients.
Components of the extracellular heart valve leaflet matrix are pre-
dominantly collagen, elastin, and proteoglycans, with collagen
being the most abundant protein (about 60 % on a dry weight
basis [4]). Storage of decellularized heart valve scaffolds in bio-
banking facilities allows samples to be readily available whenever
needed. Tissue scaffolds can be stored at reduced supra-zero tem-
peratures (i.e., in standard refrigerators). This can be done by storing
the tissue in a highly concentrated glycerol [5] or sucrose [6] solution.
499
500 Willem F. Wolkers and Andres Hilfiker
Alternatively, they can be cryopreserved by vitrification [7] or

freeze-dried [8]. Freeze-drying has the advantage of allowing stor-
age in a dry state, making banking and transport easier. Freeze-
drying comprises of both a freezing and a drying step. The drying
process is divided in a primary and a secondary drying step. During
primary drying, the freezable water fraction is removed from fro-
zen material under reduced pressure by sublimation. During sec-
ondary drying, the unfrozen water is removed through desorption,
resulting in a product with a water content typically less than 3 %
(w/w). This is a slower process, which requires higher tempera-
tures. Freeze-drying is damaging to most biological materials and
requires protective additives such as sucrose stabilizing structure
and function of biomolecules. Sucrose as well as other disaccha-
rides protects biomolecules by replacing water that is normally
hydrogen bonded [9] and by forming a stabilizing glassy state
[10]. Care should be taken that the product remains in a glassy
state during the entire freeze-drying process.
If the damaging effects of freeze-drying can be overcome,
freeze-drying is the preferred method for biobanking of decellular-
ized tissues. This chapter outlines the decellularization process to
remove cells from porcine heart valve tissues, the loading step with
sucrose, as well as the freeze-drying procedure. Furthermore, a
standard histological procedure is described to assess the tissue
structure after rehydration.
2 Materials
2.1 Decellularization 1. Braunol disinfection solution (7.5 % povidone-iodine).

and Freeze-Drying 2. Decellularization solution: 0.5 % (w/v) sodium deoxycholate
of Heart Valves and 0.5 % (w/v) sodium dodecyl sulfate (SDS) in filter-
sterilized water.
3. Phosphate-buffered saline solution (PBS): 137 mM NaCl,
27 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4.
4. PBS+: PBS supplemented with 1 % (w/v) penicillin/
streptomycin, 1 % (w/v) partricin, and 1 % (w/v) gentamicin.
5. Shaking incubator for continuously shaking during decellular-
ization and washing.
6. Lyoprotectant incubation and freeze-drying medium.
(a) PBS/sucrose: 5 % (w/v) sucrose in PBS. Prepare 100 mL,
filter-sterilize using 0.22-m filters, and store at 4 C until
use.
(b) PBS/sucrose/HES: 2.5 % (w/v) sucrose and 2.5 % (w/v)
hydroxyl ethylene starch in PBS. Prepare 100 mL, filter-
sterilize using 0.22-m filters, and store at 4 C until use.
Freeze-Drying of Heart Valve Tissues 501
7. Water bath or incubator for loading the scaffolds with protectants

at elevated temperatures (37 C).
8. Freeze-dryer (e.g., VirTis AdVantage Plus benchtop freeze-
dryer from SP Scientific, Warminster, PA, USA) (see Note 1).
2.2 Histological 1. Formalin (3.5 % v/v) solution for fixation of tissue specimens.
Evaluation of Tissue 2. Ethanol (70, 75, 80, 85, 90, 95, and 100 % v/v solutions in
Structure distilled water), isopropanol (100 %), and Roticlear (Carl Roth
GmbH, Karlsruhe, Germany) for dehydration of tissue speci-
mens, and liquid paraffin for embedding.
3. Leica TP 1020 automatic tissue processor (Leica, Wetzlar,
Germany) for dehydration of tissue samples with increasing
alcohol concentrations and final embedding in liquid paraffin.
4. Microtome (e.g., manual rotary Reichert-Jung 2040 autocut
microtome, Leica, Wetzlar, Germany) for preparing tissue
sections.
5. Corbit-Balsam mounting medium (Hecht, Hamburg, Germany).
6. Hematoxylin-eosin staining solution. Filter solution prior to use.
3 Methods
3.1 Preparation 1. Obtain porcine hearts from a local slaughterhouse and dissect
of Heart Valves heart valves from the hearts.
and Decellularization 2. Disinfect by incubation in Braunol solution for 5 min, and
then remove this solution by incubation in PBS+ for 20 min.
3. Place tissues into the decellularization solution for two times 12 h
at room temperature while continuously shaking the solution.
4. Remove the detergents by washing two times 12 h in distilled
water.
5. Transfer tissue into PBS+, and incubate for 5 days at room
temperature, while continuously shaking and changing the
solution every 12 h (see Note 2).
6. Store decellularized tissues in PBS+ at 4 C, until further
processing. Conduit mass is reduced by the decellularization
procedure. Decellularization reduces cellular components of
myocardial tissue but it does not completely remove all cellular
components, which is evident from the brownish color of the
muscular myoglobin (Fig. 1).
3.2 Loading 1. Incubate decellularized heart valve tissues in freeze-drying

Decellularized Heart medium/lyoprotectant solution (PBS/sucrose or PBS/
Valve Tissue sucrose/HES), at 37 C for 4 h (see Note 3).
with Lyoprotectants 2. Transfer the valve tissue to an open container (i.e., a Petri dish)
and Freeze-Drying and put on the freeze-dryer shelves with the shelve tempera-
ture set at 20 C.
Fig. 1 Porcine pulmonary valve conduits in native (a) and decellularized (b) state composed of pulmonary
artery parts (multiplication sign), myocardial rim (hash), and properly pulmonary heart valve comprising three
semilunar-shaped leaflets. Conduit mass and tissue stability are significantly reduced by the decellularization
procedure. The adventitia layer of pulmonary artery with the vasa vasorum is clearly visible in the native state
and is mostly removed after decellularization. Decellularization also reduces cellular components of myocar-
dial tissue but removal is not complete, which is evident from the weakened but not abolished brownish color
of muscular myoglobin
3. Start the freeze-drying protocol.

(a) Freeze the samples by cooling the shelves at a cooling rate
of 1 C min-1 until the shelve temperature reaches 40 C,
and keep at 40 C for 3 h.
(b) Increase shelf temperature to 30 C at a rate of 1 C min1.
(c) For primary drying, decrease the chamber pressure to
60 mTorr while maintaining the shelf temperature at
30 C, and dry for 16 h (see Note 4).
(d) For secondary drying, increase the shelve temperature
from 30 to 20 C at a rate of 0.1 C min1 while main-
taining the pressure at 60 mTorr. After reaching 20 C,
dry for another 69 h (see Note 5).
4. After freeze-drying, transfer the dried valves into a dry envi-
ronment or seal with a vacuum sealer, and store at room
temperature (see Note 6).
5. Rehydrate dried valves in distilled water for minimally 1 h
(see Note 7).
3.3 Histological 1. Rehydrate dried valves in water for at least 1 h at room tem-
Analysis perature prior histological inspection.
2. Fix hydrated valve tissue by incubating in formalin solution for
24 h at room temperature.
3. Transfer the tissue samples into distilled water (twice), 25 %

ethanol (twice), and 50 % ethanol (twice), respectively, keeping
the samples for 30 min in each solution.
4. Transfer tissue samples into the automatic tissue processor
(see Note 8) to dehydrate by exposing the sample to ethanol
solutions with increasing ethanol concentration. After dehy-
dration in ethanol, infiltrate in liquid paraffin. Immerse the
samples in the following solutions:
(a) 75 % ethanol, 30 min
(b) 80 % ethanol, 30 min
(c) 85 % ethanol, 30 min
(d) 90 % ethanol, 30 min
(e) 95 % ethanol, 45 min
(f) 100 % ethanol, 45 min
(g) Isopropanol, 45 min
(h) Isopropanol, 45 min
(i) Roticlear, 45 min
(j) Roticlear, 45 min
(k) Paraffin, 6 h
(l) Paraffin, 24 h
5. Transfer tissue samples infiltrated with paraffin into embedding
templates and pour more hot paraffin wax (60 C) over the
samples. Cool the paraffin blocks on a cold plate (at 10 C).
6. Make serial paraffin sections of 10 m thickness using a manual
rotary microtome. Transfer sections onto glass slides.
7. Dry and flatten specimens on a heated plate (40 C) for several
hours.
8. In order to remove paraffin from the sections, first incubate the
slides at 60 C for 30 min. To rehydrate the specimens, subject
the slides to the following incubations at room temperature:
(a) Xylene, 10 min
(b) Xylene, 10 min
(c) 100 % ethanol, 2 min
(d) 90 % ethanol, 2 min
(e) 80 % ethanol, 2 min
(f) 70 % ethanol, 2 min
(g) Distilled water, 0.5 min
9. Stain the slides with hematoxylin-eosin staining solution to
visualize nuclei and the general tissue architecture (see Note 9):
(a) Incubate slides in hematoxylin solution for 8 min, followed
by 10 min washing under running tap water.
(b) Incubate slides for 5 min in eosin solution.
Fig. 2 H&E staining of paraffin sections of porcine pulmonary heart valve tissue. Representative areas from
longitudinal sections of pulmonary artery ((a) native state, (b) decellularized state) and pulmonary heart valve
leaflet ((c) native state, (d) decellularized state) are shown. A remarkable reduction of dark purple-stained
nuclei (hematoxylin) between (a) and (b) can be seen. Endothelial coating and most parts of subendothelial
connective tissue of intima layer (asterisk) are removed by decellularization. Structural architecture of media
layer (multiplication sign) mainly composed of pinkish stained (eosin) collagen and elastin fibers is largely
preserved but significantly loosened up. Same holds true for leaflet sections (c) and (d). The characteristic
trilaminar leaflet structure (consisting of lamina arterialis (a), lamina ventricularis (v), lamina fibrosa (f), and
lamina spongiosa (s)) is preserved after decellularization
(c) Incubate the slides in 80 and 90 % ethanol, 2 min each.

(d) Incubate the slides two times in xylene, 5 min each.
(e) Mount the specimens in mounting medium.
10. Inspect specimens using light microscopy. Hematoxylin-eosin
(H&E) staining can be used to study the overall tissue archi-
tecture of native and decellularized tissues. A remarkable
reduction of dark purple-stained nuclei is visible in the decel-
lularized specimens (Fig. 2).
4 Notes
1. A freeze-dryer system with temperature-controlled shelves

and control of chamber pressure is preferred. Freeze-drying,
however, can also be done with a simpler benchtop manifold
freeze-drying system lacking temperature-controlled shelves or
control of the chamber pressure. In this case, freezing can be
done by placing the samples in a 80 C freezer. The frozen
sample can be transported on dry ice to the freeze-dryer to
avoid thawing of the sample.
2. This step is done to remove residual detergents in the tissue.
Residual detergents may inhibit regrowth of cells in vivo.
3. It takes less than 4 h to load porcine heart valve tissue with
sucrose [8]. It should be noted that HES or other polymers
diffuse significantly slower through the tissue. Therefore, if
formulations HES or other polymers are used, incubation
times should be increased (i.e., 824 h).
4. The end of the primary drying phase can be detected by a pres-
sure rise test. In this procedure, the valve between condenser
and drying chamber is closed. A pressure increase after closure
of valves indicates not all frozen water has been evaporated yet.
5. Secondary drying is done to remove unfrozen water from the
sample. The water content of the sample depends on the end
temperature of the shelves that is selected in the secondary dry-
ing cycle. If lower water contents are desired, the end tempera-
ture of the shelves can be increased to a higher end temperature
(i.e., 40 C).
6. Most freeze-drying systems have a stoppering device allowing
sealing the sample under vacuum. There are no suitable com-
mercial containers (vial + stopper) available for freeze-drying
of heart valves. Such containers, however, can be constructed,
which would allow to seal the samples under vacuum. Alter-
natively, a vacuum sealer can be used to seal the freeze-dried
samples. In any case, samples should be stored in a dry environ-
ment (i.e., using a desiccant). Samples can be stored at room
temperature, in a standard refrigerator, or in a 20 C freezer.
7. Water can be replaced several times with fresh water to remove
the protectants. Sucrose and HES, however, are nontoxic pro-
tectants that do not need to be completely removed prior to
transplantation.
8. Alternatively, samples can be manually transferred in the differ-
ent solutions.
9. Hematoxylin has a deep blue-purple color and stains proteins
nonspecifically. In a typical tissue, nuclei are stained blue,
whereas cytoplasm and extracellular matrix have varying degrees
of pink staining.
Acknowledgment
This work is supported by funding from the CORTISS Foundation

and the German Research Foundation (DFG, Deutsche Fors-
chungsgemeinschaft) for the Cluster of Excellence From
Regenerative Biology to Reconstructive Therapy (REBIRTH).
References
1. Lichtenberg A, Tudorache I, Cebotari S, Ramos T, Gruys E, Voorhout W (2002)
Ringes-Lichtenberg S, Sturz G, Hoeffler K, Processing of ovine cardiac valve allografts: 1.
Hurscheler C, Brandes G, Hilfiker A, Haverich Effects of preservation method on structure
A (2006) In vitro re-endothelialization of and mechanical properties. Cell Tissue Bank
detergent decellularized heart valves under 3:7989
simulated physiological dynamic conditions. 6. Drury FJ, Olsen EGJ, Ross DN (1982)
Biomaterials 27:42214229 Morphological assessment of sucrose preser-
2. Cebotari S, Lichtenberg A, Tudorache I, vation for porcine heart valves. Thorax 37:
Hilfiker A, Mertsching H, Leyh R, Breymann 466471
T, Kallenbach K, Maniuc L, Batrinac A, Repin 7. Brockbank KGM, Schenke-Layland K, Greene
O, Maliga O, Ciubotaru A, Haverich A (2006) ED, Chen Z, Fritze O, Schleicher M, Kaulitz
Clinical application of tissue engineered human R, Riemann I, Fend F, Albes JM, Stock UA,
heart valves using autologous progenitor cells. Lisy M (2012) Ice-free cryopreservation of
Circulation 114:I132I137 heart valve allografts: better extracellular matrix
3. Cebotari S, Tudorache I, Ciubotaru A, Boethig preservation in vivo and preclinical results. Cell
D, Sarikouch S, Goerler A, Lichtenberg A, Tissue Bank 13:663671
Cheptanaru E, Barnaciuc S, Cazacu A, Maliga 8. Wang S, Goecke T, Meixner C, Haverich A,
O, Repin O, Maniuc L, Breymann T, Haverich Hilfiker A, Wolkers WF (2012) Freeze-dried
A (2011) Use of fresh decellularized allografts heart valve scaffolds. Tissue Eng 18:517525
for pulmonary valve replacement may reduce the 9. Crowe JH, Crowe LM, Carpenter JF,
reoperation rate in children and young adults: Prestrelski S, Hoekstra FA, de Araujo P, Panek
early report. Circulation 124:S115S123 AD (1997) Anhydrobiosis: cellular adaptation
4. Kunzelman KS, Cochran RP, Murphree SS, to extreme dehydration. In: Dantzler WH (ed)
Ring WS, Verrier ED, Eberhart RC (1993) Comparative physiology, handbook of physiol-
Differential collagen distribution in the mitral ogy, vol II. Oxford University Press, Oxford,
valve and its influence on biomechanical behav- UK, pp 14451477
ior. J Heart Valve Dis 2:236244 10. Crowe JH, Carpenter JF, Crowe LM (1989)
5. Aidulis D, Pegg DE, Hunt CJ, Goffin YA, The role of vitrification in anhydrobiosis. Annu
Vanderkelen A, van Hoeck B, Santiago T, Rev Physiol 60:73103
INDEX
A Crystallization
recrystallization.....................................10, 24, 32, 33, 37,
Alginate encapsulation ..................................... 444, 448449 4044, 47, 48, 67, 84, 130, 163, 166, 173, 176,
234, 271, 286, 327
B
water-ice phase change measurements..........................26
Biobanking ....................................................... 232, 499, 500 Cyclodextrin. See Methyl--cyclodextrin
Blastocyst........................................................... 85, 292, 297,
305318, 321, 493 D
human blastocyst vitrification ..................... 305318, 321 Deglycerolization...............................108, 354, 355, 359364
Blood cells Differential scanning calorimetry (DSC) .................... 22, 29,
red blood cell cryopreservation ...................7, 8, 108, 172, 33, 39, 40, 42, 93, 99, 163, 164, 167179, 404,
353366, 372, 378 408, 441, 445, 447, 460, 470
umbilical cord blood cryopreservation ................369371 Directional freezing ..................................................381396
umbilical cord blood processing..........................369271
Boyle vant Hoff ..........................................14, 46, 53, 8688, E
98, 104, 113
Electron microscopy
C cryo-electron microscopy ............................................265
scanning electron microscopy
Cellular dehydration (SEM)......................244, 257259, 262, 267, 268
FTIR measurements ..................................................147 transmission electron microscopy
microscopic measurements .........................................211 (TEM) ....................................244, 248251, 253,
modeling .....................................................................182 255257, 260262, 265, 267271
Chemical potential ..........................37, 8798, 102, 103, 114 Embryo transfer .......................................................315317
Chilling injury ..................................................26, 32, 52, 59,
6366, 6870 F
Cryoinjury ...................................................... 6, 10, 181, 182
Fourier transform infrared spectroscopy (FTIR).........147160
Cryomicroscopy.............................8, 181226, 229, 230, 232
Freeze-drying
Cryopreservation
decellularized heart valve freeze-drying ..............499505
blood cell cryopreservation ............................ 7, 8, 10, 14,
DSC measurements ............................................ 163, 460
108, 172, 353366, 372, 378
FTIR measurements ..........................................147160
directional freezing .............................................381396
lactic acid bacteria freeze-drying ........................477487
modeling ...............................................................83115
principles ............................................ 121141, 163165
oocyte cryopreservation ...................... 289, 290, 292, 293
proteins freeze-drying................................ 122, 131, 139,
plant germplasm cryopreservation ...................... 433, 444
140, 147, 148, 151, 156157, 159, 163, 167, 176,
principles ..................................................... 318, 2170,
177, 179, 459475, 479, 480, 499, 505
163165, 230, 400
sperm freeze-drying .................................... 172, 489495
sperm cryopreservation ............................... 277, 278, 489
FTIR. See Fourier transform infrared spectroscopy (FTIR)
stem cell cryopreservation ...........................7, 14, 88, 112,
172, 235, 321327, 369, 370, 377 G
Cryoprotective agents
freezing point depression ..............................................89 Glassy state
mechanisms .........................................59, 62, 63, 84, 462 glass transition measurements......................... 16, 22, 474
toxicity ............................. 30, 50, 5458, 62, 63, 65, 69, 1 principles ...................................................... 318, 2170
04, 105, 297, 307, 308 Greenshell mussel .................................................329335
vol. 1257, DOI 10.1007/978-1-4939-2193-5, Springer Science+Business Media New York 2015
507
CRYOPRESERVATION AND FREEZE-DRYING PROTOCOLS
508 Index
H P
Heart valves Phase
decellularized heart valve freeze-drying ..............499505 diagram ......... 16, 29, 35, 38, 84, 89, 91, 92, 110, 112, 490
heart valve vitrification ...............................................401 separation ....................................163, 167, 177179, 461
Plant
I callus culture ...............................................................445
Ice nucleation .............................................24, 25, 33, 35, 36, plant cell line cryopreservation ...........................423428
3839, 41, 43, 44, 50, 57, 59, 66, 103, 111, plant germplasm cryopreservation ...................... 433, 444
127, 128, 148, 155, 220, 286, 296, 299, 301, 381, suspension culture ............................................... 423, 426
382, 417 Primary drying ................. 122, 124, 128, 132, 136, 137, 459,
ICSI. See Intracytoplasmic sperm injection (ICSI) 470472, 474, 477, 486, 490, 491, 500, 502, 505
Imaging Protein structure
cryo-imaging ..............................................................182 denaturation.................................................. 68, 177, 474
fluorescence imaging...................................................230 stability ...............................................................177179
high speed imaging .............................184, 185, 191192,
Q
202, 203, 207, 208, 215
spectroscopic imaging ................................. 233, 234, 237 Quality management .................431433, 436, 442, 453, 455
Intracellular ice formation
imaging ....................................................... 182, 183, 214 R
principles ................................................................2170 Raman microspectroscopy ................................................234
Intracytoplasmic sperm injection
(ICSI) ..................................... 290, 291, 491, 495 S
Secondary drying ............................................. 122, 124, 136,
L
137, 139, 459, 470472, 474, 477, 483, 486, 487,
Lactic acid bacteria (LAB) .......................................477487 490, 491, 495, 500, 502, 505
Laser scanning microscopy (LSM) ...........................229240 Semen. See Sperm
Liposome preparation.......................................................340 Solution effects
Lyophilization. See Freeze-drying freeze-concentration ........................................... 139, 173
Lyoprotectants principles ...................................................... 318, 2170
formulations................................................ 465466, 473 SOP. See Standard operating procedures (SOP)
principles ............................................ 466, 473, 500502 Sperm
bovine sperm cryopreservation....................................491
M clean-up ..............................................................343352
Mass transport .......................................84, 86, 107, 114, 115 density centrifugation ..................278, 343, 346, 348, 352
Matrix implants ................................................................499 equine sperm cryopreservation ...................................285
Membrane phase behavior........................ 147, 148, 153, 155 freeze-drying of sperm........................................489495
Methyl--cyclodextrin......................................................340 Greenshell mussel sperm cryopreservation..............330
large volume cryopreservation ............................381396
N membrane modification......................................337341
Spermatozoa. See Sperm
Nelson-Aalen estimator............................................ 199, 201 Standard operating procedures (SOP) ......................431455
Stem cells
O
hematopoietic stem cell
Oocyte cryopreservation .............................. 369, 370, 377
human oocyte cryopreservation .......................... 292, 293 human pluripotent stem cell vitrification............321327
murine oocyte cryopreservation .......................... 291, 293 Sublimation ..............................................122, 124, 126, 127,
Optimal cooling rate 129, 130, 132138, 264, 272, 459, 470, 471, 474,
modeling ....................................................... 84, 111, 112 477, 481, 483, 485, 487, 490, 491, 500
principles ..................................................... 118, 2170, Supercooling .................................................3, 5, 8, 9, 17, 29,
121141 31, 110113, 126, 127, 166, 182, 230
Osmosis Surface-based vitrification ........................................321327
osmotic tolerance limits ..........................84, 85, 101, 104,
105, 107109, 285 U
principles ...................................................... 318, 2170 Umbilical cord blood. See Blood cells
CRYOPRESERVATION AND FREEZE-DRYING PROTOCOLS
Index
509
V heart valve vitrification ...............................................401

human blastocyst vitrification ..................... 305318, 321
Virial equation ................................................93, 95, 96, 100, organ preservation .................................... 18, 30, 32, 311
102, 103 principles ........................................................ 2170, 400
Vitrification stem cell vitrification ..........................................321327
devitrification .................................24, 33, 3944, 4649,
59, 67, 69, 163, 173, 175, 176, 267269, 318, 327
DSC measurements ........................................ 22, 33, 445

(Methods in Molecular Biology 1257) Willem F. Wolkers, Harriëtte Oldenhof (Eds.) - Cryopreservation and Freeze-Drying Protocols-Springer-Verlag New York (2015)

Uploaded by

Copyright:

Available Formats

You might also like

(Methods in Molecular Biology 1257) Willem F. Wolkers, Harriëtte Oldenhof (Eds.) - Cryopreservation and Freeze-Drying Protocols-Springer-Verlag New York (2015)

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

(Methods in Molecular Biology 1257) Willem F. Wolkers, Harriëtte Oldenhof (Eds.) - Cryopreservation and Freeze-Drying Protocols-Springer-Verlag New York (2015)

Uploaded by

Copyright:

Available Formats

Methods in

Molecular Biology 1257

For further volumes:

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Library of Congress Control Number: 2014955222

Springer Science+Business Media New York 2015

Printed on acid-free paper

Humana Press is a brand of Springer

In memory of Stanley P. Leibo

Hannover, Germany Willem F. Wolkers

PART I FUNDAMENTAL ASPECTS AND FREEZING TECHNOLOGY

PART II METHODS TO STUDY FREEZING AND DRYING PROCESSES

PART III CRYOPRESERVATION PROTOCOLS

13 Efficient Cryopreservation of Human Pluripotent

PART IV FREEZE-DRYING PROTOCOLS

GARY D. J. ADAMS Deceased

MARTINA WESTPHAL Plant Cell Culture Department, Leibniz-Institut, German

Fundamental Aspects and Freezing Technology

Key words Cryopreservation, Cryoprotectants, Intracellular freezing, Solution effects, Supercooling,

1 Freezing Injury andCryoprotection

solution effects and intracellular freezing, causes each cell to show

However, the salt in the suspending medium is not the only

actually suffer more damage when exposed to a given salt concen-

In fact, very small amounts of intracellular ice are compatible

crystals to recrystallize, to coalesce, and to grow. This has been

2.1 General Cryoprotection usually involves treatment of the cells or tissues

where k is a constant that is characteristic of the membrane-and-

This equation states that flux across unit area of membrane is

J v = L p ART (Ci - Ce ) + s (Cip - Cep ) (6)

and derive estimates of Lp and Ps. We will now consider in more

2.2 Introduction The exposure of cells to a high concentration of cryoprotectant

2.3 Removal When a permeating cryoprotectant is removed by exposing the

2.5 Exposure Cells immersed in a solution of nonpermeating solute reach an

This relationship requires that a plot of Vrel against Mrel, which

3 Preservation ofCells andTissues

lular structures and to retain normal interconnections between the

Fig. 9 Supplemented phase diagram for glycerol/water. The intersection of the

solutions that are less concentrated than this if sufficiently rapid

degrees per minute in practice. For small samples it is more feasible

Principles ofCryopreservation by Vitrification

Key words Vitrification, Freezing, Intracellular ice formation, Devitrification, Recrystallization,

1 Introduction andGeneral Orientation

1.1 Overview Vitrification is the solidification of a liquid into a noncrystalline or

1.2 Basic The glass transition temperature, or TG, is the temperature at

TG is a theoretical temperature reached when freezing is able to

avoided. Rewarming, or the warming of a previously cryopre-

it is possible to outrun the kinetics of ice formation. The critical

Chilling injury is injury caused by cooling per se. Although

physical changes, particularly involving cell distortion, that precede

1.4 Cryopreservation A historical introduction to the field of cryopreservation by vitrifi-

At least in part for that reason, in 1965, a historical opportunity

been noted anecdotally (D.E.Pegg, personal communication) that

1.5 Advantages The overall purpose of vitrification is to avoid freezing. To under-

stance, so its formation subtracts solvent water from a freezing

The second point of relevance to vitrification is that, as noted,

what CPA concentration and exposure time is needed to ensure

In any case, at least some larvae appear able to survive cooling to

most cells survive cryopreservation as a result of vitrification, even