Methods in Molecular Biology™

Series Editor
John M. Walker
School of Life Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK

For other titles published in this series, go to

www.springer.com/series/7651
The Plant Cell Wall

Methods and Protocols

Edited by

Zoë A. Popper
Botany and Plant Science, School of Natural Sciences,
National University of Ireland, Galway,
Galway, Ireland
Editor
Zoë A. Popper, Ph.D.
Botany and Plant Science
School of Natural Sciences
National University of Ireland, Galway
Galway,
Ireland
[email protected]

ISSN 1064-3745
ISBN 978-1-61779-007-2
DOI 10.1007/978-1-61779-008-9
Springer New York Dordrecht Heidelberg London
© Springer Science+Business Media, LLC 2011
All rights reserved. This work may not be translated or copied in whole or in part without the written permission of
the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013,
USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of
information storage and retrieval, electronic adaptation, computer software, or by similar or �dissimilar methodology
now known or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified
as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of going to press, �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 part of Springer Science+Business Media (www.springer.com)

Preface

When they got home to dinner they met the Hemulen on the steps. He was beaming with happiness.
“Well?” said Moomintroll. “What is it?” “Nature study!” shouted the Hemulen. “I shall botanize …”
Tove Jansson

Plants are essential to life on earth, and, while some readers of this book may not be
entirely familiar with the cell wall per se, they will have come across it in many forms.
Cellulose, a major plant cell wall polysaccharide, is also the most abundantly occurring
natural biopolymer, with many other plant cell wall components being among the next
most abundant. Some ways in which people may be familiar with the cell wall and/or cell
wall components are as; textiles (many, such as cotton, are cellulose); paper; timber; pec-
tin, which is the gelling agent used in jams and other foods; dietary fiber; and cell wall
characteristics and metabolism control, for example, fruit ripening and texture. We are
therefore dependent on plant cell walls for health, food, and clothing, and a major current
area of research is their use as biofuels.
Plant cells are surrounded by a cell wall which is fundamental to their function and
survival. The cell wall and its constituent polysaccharides and proteins control nearly all
plant-based biological and biophysical processes including expansive plant cell growth,
plant development, cell shape and size, cell–cell communication, and interactions with,
and defence against pathogens. Understanding the cell wall is therefore not only funda-
mental to the plant sciences but it is also pertinent to aspects of human and animal nutri-
tion and health as well as plant–microbe and plant–animal interactions. Furthermore,
advanced cell wall analysis is the key to developing novel or improving current uses of the
wall and plants.
Comprehensive analysis of the plant cell wall demands a multidisciplinary approach
and employs a multitude of tools and techniques. This volume describes some of the
methods which are currently applied to investigate the many aspects of the plant cell wall
including its structure, biochemical composition, and metabolism, to name but a few.
Each chapter is written by leading experts in cell wall research and is written with the aim
that the protocol(s) can be carried out by someone without previous experience in that
particular method or specifically in cell wall research. The techniques included in this
volume­range from plant tissue culture techniques, which can be applied to investigating
cell wall structure and metabolism, to methods directed towards structural analysis and
occurrence of carbohydrates, to the development and use of microscopy-based tools
and techniques, to those which measure the physical properties of the wall, to methods
based on the application of molecular genetic approaches. Many of the methods have been
recently developed or are becoming more widely used with the development of advanced
instrumentation and technology, and several are high throughput and/or in situ tech-
niques which facilitate powerful new insights into cell wall biochemistry and metabolism.
While this volume aims to describe a wide-range of cell wall-directed protocols that
can be used to investigate the cell wall, there are other resources which the reader is also
likely to find extremely useful. The Growing Plant Cell Wall: Chemical and Metabolic

v
vi Preface

Analysis [1] provides detailed and user-friendly descriptions of cell wall-directed methods,
the majority of which are not contained in this volume, and which are widely used and
fundamental to research in the field. Furthermore, it also provides a comprehensive
and accessible introduction to the cell wall. The reader may also find helpful the video-
based explanations of protocols which are available from companies such as Megazyme in
addition to recent JoVE publications [2–4].
I would like to wish the reader every success with their plant-based conjectures,
hypotheses, and experiments (Fig. 1). Finally, I would like to thank all members of the cell
wall community and colleagues at NUI Galway who have supported and enabled this
project.
Zoë A. Popper

Fig. 1. Simon Popper. Copyright: The artist (Courtesy: Rachmaninoff’s, London and the artist).

References
1. Fry SC (2000) The growing plant cell wall: 3. Foster CE, Martin TM, Pauly M (2010)
chemical and metabolic analysis. Reprint Comprehensive compositional analysis of plant
Edition, Blackburn, Caldwell, NJ. [ISBN cell walls (lignocellulosic biomass). Part II: car-
1-930665-08-3] bohydrates. JoVE. 37. http://www.jove.com/
2. Foster CE, Martin TM, Pauly M (2010) index/details.stp?id=1837, doi: 10.3791/1837
Comprehensive analysis of plant cell walls 4. Durachko DM, Cosgrove DJ (2009)
(lignocellulosic biomass). Part I: lignin. JoVE. Measuring plant cell wall extension (creep)
37. http://www.jove.com/index/details. induced by acidic pH and alpha-expansin.
stp?id=1745, doi: 10.3791/1745 JoVE 25. http://www.jove.com/index/
details.stp?id=1263, doi: 10.3791/1263
Acknowledgment
This publication was grant-aided by the Publications Fund of National University
of Ireland, Galway.

vii
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

1 Plant Tissue Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Anna Kärkönen, Arja Santanen, Kuninori Iwamoto, and Hiroo Fukuda
2 Computerized Molecular Modeling of Carbohydrates . . . . . . . . . . . . . . . . . . . . . 21
Alfred D. French and Glenn P. Johnson
3 Oligosaccharide Mass Profiling (OLIMP) of Cell Wall Polysaccharides
by MALDI-TOF/MS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Markus Günl, Florian Kraemer, and Markus Pauly
4 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building
Blocks and Their Metabolic Precursors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Stephen C. Fry
5 Carbohydrate Gel Electrophoresis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Florence Goubet, Paul Dupree, and Katja Salomon Johansen
6 Capillary Electrophoresis with Detection by Laser-Induced Fluorescence . . . . . . . 93
Andrew Mort and Xiangmei Wu
7 Monoclonal Antibodies, Carbohydrate-Binding Modules,
and the Detection of Polysaccharides in Plant Cell Walls . . . . . . . . . . . . . . . . . . . 103
Cécile Hervé , Susan E. Marcus , and J. Paul Knox
8 Screening and Characterization of Plant Cell Walls
Using Carbohydrate Microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Iben Sørensen and William G.T. Willats
9 Electron Tomography and Immunogold Labelling as Tools
to Analyse De Novo Assembly of Plant Cell Walls . . . . . . . . . . . . . . . . . . . . . . . . 123
Marisa S. Otegui
10 Analysing Cellulose Biosynthesis with Confocal Microscopy . . . . . . . . . . . . . . . . . 141
Meera Nair and Seth DeBolt
11 Visual Mapping of Cell Wall Biosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Yumiko Sakuragi, Morten H.H. Nørholm, and Henrik V. Scheller
12 Atomic Force Microscopy of Plant Cell Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Andrew R. Kirby
13 Using Solid-State 13C NMR Spectroscopy to Study the Molecular
Organisation of Primary Plant Cell Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Tracey J. Bootten, Philip J. Harris, Laurence D. Melton,
and Roger H. Newman
14 Formation of Cellulose-Based Composites with Hemicelluloses
and Pectins Using Gluconacetobacter Fermentation . . . . . . . . . . . . . . . . . . . . . . . 197
Deirdre Mikkelsen and Michael J. Gidley

ix
x Contents

15 Structural Proteins of the Primary Cell Wall: Extraction,

Purification, and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Derek T.A. Lamport, Li Tan, and Marcia J. Kieliszewski
16 New Insights into the Control of Cell Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Claudia Blaukopf, Matthäus Z. Krol, and Georg J. Seifert
17 Extraction and Detection of Arabinogalactan Proteins . . . . . . . . . . . . . . . . . . . . . 245
Zoë A. Popper
18 Characterization of the Plant Cell Wall Proteome
Using High-Throughput Screens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Sang-Jik Lee and Jocelyn K.C. Rose
19 Knocking Out the Wall: Protocols for Gene Targeting
in Physcomitrella patens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Alison W. Roberts, Christos S. Dimos, Michael J. Budziszek, Jr.,
Chessa A. Goss, and Virginia Lai
20 Measuring In Vitro Extensibility of Growing Plant Cell Walls . . . . . . . . . . . . . . . 291
Daniel J. Cosgrove
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Contributors

Clauda Blaukopf • Department of Applied Genetics and Cell Biology,

University of Natural Resources and Applied Life Sciences, Vienna, Austria
Tracey J. Bootten • School of Biological Sciences, The University of Auckland,
Auckland, New Zealand; Industrial Research Limited, Lower Hutt, New Zealand
Michael J. Budziszek, Jr. • Department of Biological Sciences, Center for
Biotechnology and Life Sciences, University of Rhode Island, Kingston, RI, USA
Daniel J. Cosgrove • Department of Biology, Penn State University,
University Park, PA, USA
Seth Debolt • Department of Horticulture, University of Kentucky,
Lexington, KY, USA
Christos S. Dimos • Department of Biological Sciences, Center for Biotechnology
and Life Sciences, University of Rhode Island, Kingston, RI, USA
Paul Dupree • Department of Biochemistry, University of Cambridge, Cambridge, UK
Alfred D. French • Southern Regional Research Center,
U. S. Department of Agriculture, New Orleans, LA, USA
Stephen C. Fry • The Edinburgh Cell Wall Group, Institute of Molecular Plant
Sciences, School of Biological Sciences, University of Edinburgh, Edinburgh, UK
Hiroo Fukuda • Department of Biological Sciences, Graduate School of Science,
University of Tokyo, Tokyo, Japan
Michael J. Gidley • ARC Centre of Excellence in Plant Cell Walls, Centre for
Nutrition and Food Sciences, The University of Queensland, Brisbane, Queensland,
Australia
Chessa A. Goss • Department of Biological Sciences, Center for Biotechnology
and Life Sciences, University of Rhode Island, Kingston, RI, USA
Florence Goubet • Bayer BioScience NV, Ghent, Belgium
Markus Günl • Energy Biosciences Institute, University of California, Berkeley,
CA, USA
Philip J. Harris • School of Biological Sciences, The University of Auckland,
Auckland, New Zealand
Cécile Hervé • Station Biologique de Roscoff, UMR7139 Marine Plants
and Biomolecules, Roscoff, France
Kuninori Iwamoto • Department of Biological Sciences, Graduate School of Science,
University of Tokyo, Tokyo, Japan
Katja Salomon Johansen • Novozymes A/S, Bagsvaerd, Denmark
Glenn P. Johnson • Southern Regional Research Center,
U. S. Department of Agriculture, New Orleans, LA, USA
Anna Kärkönen • Department of Agricultural Sciences, University of Helsinki,
Helsinki, Finland; Department of Agricultural Sciences, MTT Agrifood Research
Finland, University of Helsinki, Helsinki, Finland

xi
xii Contributors

Marcia J. Kieliszewski • Department of Chemistry and Biochemistry,

Biochemistry Research Facility, Ohio University, Athens, OH, USA
Andrew R. Kirby • Institute of Food Research, Colney, Norwich, Norfolk, UK
J. Paul Knox • Centre for Plant Sciences, Faculty of Biological Sciences,
University of Leeds, Leeds, UK
Florian Kraemer • Energy Biosciences Institute, University of California,
Berkeley, CA, USA
Matthäus Z. Krol • Department of Applied Genetics and Cell Biology,
University of Natural Resources and Applied Life Sciences, Vienna, Austria;
University of Applied Sciences, Wiener Neustadt, Austria
Virginia Lai • Department of Biological Sciences, Center for Biotechnology
and Life Sciences, University of Rhode Island, Kingston, RI, USA
Derek T.A. Lamport • School of Life Sciences, University of Sussex, Falmer,
Brighton, UK
Sang-Jik Lee • Department of Plant Biology, Cornell University, Ithaca, NY, USA
Susan E. Marcus • Centre for Plant Sciences, Faculty of Biological Sciences,
University of Leeds, Leeds, UK
Laurence D. Melton • Food Science, Chemistry Department,
The University of Auckland, Auckland, New Zealand
Deirdre Mikkelsen • Centre of Excellence in Plant Cell Walls, Centre for Nutrition
and Food Sciences, The University of Queensland, Brisbane, Queensland, Australia
Andrew Mort • Department of Biochemistry and Molecular Biology,
Oklahoma State University, Stillwater, OK, USA
Meera Nair • Department of Horticulture, University of Kentucky, Lexington,
KY, USA
Roger H. Newman • Industrial Research Limited, Lower Hutt, New Zealand; Scion,
Rotorua, New Zealand
Morten H.H. Nørholm • The Department of Plant Biology and Biotechnology,
University of Copenhagen, Copenhagen, Denmark
Marisa S. Otegui • Department of Botany, University of Wisconsin, Madison,
WI, USA
Markus Pauly • Energy Biosciences Institute, University of California, Berkeley, CA, USA
Zoë A. Popper • Botany and Plant Science, School of Natural Sciences, National
University of Ireland, Galway, Galway, Ireland
Alison W. Roberts • Department of Biological Sciences, Center for Biotechnology
and Life Sciences, University of Rhode Island, Kingston, RI, USA
Jocelyn K.C. Rose • Department of Plant Biology, Cornell University, Ithaca,
NY, USA
Yumiko Sakuragi • The Department of Plant Biology and Biotechnology,
University of Copenhagen, Copenhagen, Denmark
Arja Santanen • Department of Agricultural Sciences, University of Helsinki,
Helsinki, Finland
Henrik V. Scheller • Joint Bioenergy Institution, Feedstocks Division,
Lawrence Berkeley National Laboratory, Emeryville, CA, USA
Georg J. Seifert • Department of Applied Genetics and Cell Biology,
University of Natural Resources and Applied Life Sciences, Vienna, Austria
 Contributors xiii

Iben Sørensen • Department of Biology, University of Copenhagen,

Copenhagen, Denmark
Li Tan • Complex Carbohydrate Research Center, University of Georgia, Athens, GA, USA
William G.T. Willats • Department of Biology, University of Copenhagen,
Copenhagen, Denmark
Xiangmei Wu • Department of Biochemistry and Molecular Biology, Oklahoma State
University, Stillwater, OK, USA
dlfkldfjk
 Chapter 1

Plant Tissue Cultures

Anna Kärkönen, Arja Santanen, Kuninori Iwamoto, and Hiroo Fukuda

Abstract
Plant tissue cultures are an efficient system to study cell wall biosynthesis in living cells in vivo. Tissue
cultures also provide cells and culture medium where enzymes and cell wall polymers can easily be sepa-
rated for further studies. Tissue cultures with tracheary element differentiation or extracellular lignin
formation have provided useful information related to several aspects of xylem and lignin formation. In
this chapter, methods for nutrient medium preparation, callus culture initiation, and its maintenance, as
well as those for protoplast isolation and viability observation, are described. As a case study, we describe
the establishment of a xylogenic culture of Zinnia elegans mesophyll cells.

Key words: Callus culture, Initiation, Maintenance, Nutrient medium, Protoplast, Tracheary element

1. Introduction

Plant cells and organs can be cultivated in vitro in aseptic condi-

tions (1). Plant tissue cultures are an efficient system to study cell
wall biosynthesis in living cells in vivo. Tissue cultures also pro-
vide cells and culture medium where enzymes and cell wall poly-
mers can easily be separated for further studies. In vitro cultures
allow investigations to be conducted in controlled conditions
independent of seasons. Factors related to cell wall formation can
be studied, for example, by adding the compound of interest into
the culture medium, and after incubation, the cells and the
medium are collected for further analysis. The culture medium
can be considered as a continuum of the plant cell wall as it con-
tains the proteins and the cell wall polymers that are sloughed off
from the cell wall. In callus culture, cells grow mainly as a mass of
undifferentiated cells, but there exist also some differentiated cells
making the callus an inhomogeneous mixture of cells. With cer-
tain growth regulators, organogenesis (shoots or roots) or somatic

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_1, © Springer Science+Business Media, LLC 2011

1
2 Kärkönen et al.

embryo formation may be induced. In the latter case, callus is

embryogenic. Tissue cultures with tracheary element (TE) dif-
ferentiation (2) and cell wall or extracellular lignin formation
(3–6) have provided useful information related to several aspects
of xylem and lignin formation. One of the most famous xylogenic
cultures is that of Zinnia elegans (2). Zinnia system is useful for
studying the sequence of events during xylem differentiation
largely because the differentiation is highly frequent and synchro-
nous, and all processes can be followed in single cells. Systematic
gene expression analysis and molecular markers have revealed that
many processes are common between in vitro and in situ TE for-
mation (7, 8). Studies using this system have clarified numbers of
physiological, biochemical, cell biological, and molecular biological
events underlying TE differentiation (9–11).
This chapter describes the procedures for surface sterilisation,
callus culture initiation, and maintenance, as well as for protoplast
preparation and viability observations. Finally, we describe the
basic method for the establishment of a xylogenic culture of iso-
lated Zinnia mesophyll cells.

2. Materials

2.1. Nutrient Medium Nutrient medium is a source of nutrients that plant normally
obtains from the soil. The medium also contains a carbon source
(often 1–4% w/v sucrose) and growth regulators the plant needs
for cell division and growth in vitro. Gelling agent can be added
to make the medium solid (1). As various species (even geno-
types) have different nutritional requirements for optimum
growth, a wide variety of nutrient media have been developed for
in vitro cultured plants. In order to select a medium for the spe-
cies of interest, it is useful to make a literature search. In Table 1
we show some widely used media that can be the choices to start
with (see Note 1). Also, the explants with successful callus initia-
tion are listed, since the developmental stage of the plant has a
great effect on the success of culture initiation.
Table 2 shows the nutrient salt composition of the various
media. Various macro- and microelement mixtures can be pur-
chased commercially or the stock solutions can be prepared from
the nutrient salts. A mixture of macroelements can be prepared as
10 times concentrated (10×) stock solution, whereas those of
microelements, vitamins, and growth regulators can be made as
100–1,000× stock solutions (see Notes 2–5). After combining all
components of the medium except the gelling agent, adjust the
pH, adjust to the final volume, add the gelling agent (e.g. agar),
and autoclave the medium at 121°C for 20 min. Let the medium
cool to ca. 60°C. In a laminar air-flow cabinet, filter-sterilise the
 Plant Tissue Cultures 3

Table 1
Types of explants, some widely used nutrient media, and growth regulator
concentrations used for successful callus culture initiation

Plant group Explant Medium Growth regulators

Monocotyledonous Immature and mature MS (14), N6 (24) Auxin (2,4-D)

plants embryos 1.0–18 mM
Leaf, or root segments of
aseptically germinated
seedlings
Dicotyledonous Young leaf, roots MS, WPM (25) Cytokinin
plants Stem segments (BA, 2iP, kinetin, zeatin)
Leaf, root, and stem segments 0.1–40 mM + auxin
of aseptically germinated (NAA, 2,4-D)
seedlings 0.5–10 mM
Gymnosperms Cambial/xylem strips MS Cytokinin (BA, kinetin)
Shoot tips Mod. Brown and 2–5 mM + auxin
Zygotic embryos Lawrence (3) (2, 4-D, NAA)
Mod. N6 (12) 9–16 mM 2.4-D alone:
11 mM (3)

heat-labile compounds (if any, see Note 6) and pour the medium
into Petri dishes (ca. 25 mL medium/Petri dish with a diameter
of 9 cm) (see Note 7).

2.2. Surface 1. 70% (v/v) ethanol.

Sterilisation 2. Diluted Na-hypochlorite: NaClO, 1–2% (v/v) active chlorine,
supplemented with a couple of drops of Tween 20.
3. Sterile distilled water.
4. 96% (v/v) ethanol for flaming.
5. Forceps.
6. Scalpels.
7. Sterile Petri dishes.
8. Parafilm®.

2.3. Maintenance 1. Fresh nutrient medium with a gelling agent or, alternatively,
without the gelling agent.
2. 96% (v/v) ethanol for flaming.
3. Forceps.
4. Parafilm®.
5. Sterile 5 mL pipette tips with cut tips.
6. Sterile measuring cylinders (e.g. 25 mL in volume).
7. Orbital shaker (in the case of liquid cultures).
8. Temperature- and light-adjusted growth chamber.
4 Kärkönen et al.

Table 2
Nutrient media constituents for plant tissue culture basal media (26)

Mod. Brown and

B5 (15) Lawrence (3) MS (14) N6 (24) WPM (25)
Macronutrients mg/L mM mg/L mM mg/L mM mg/L mM mg/L mM
NH4NO3 1,650 20.6 1,650 20.6 400 5
(NH4)2SO4 134 1.0 463 3.5
Ca(NO3)2 ∙ 4H2O 556 2.4
KNO3 2,528 25 1,900 18.8 1,900 18.8 2,830 28
MgSO4 · 7H2O 246 1.0 1,900 7.7 370 1.5 185 0.75 370 1.5
KH2PO4 340 2.5 170 1.25 400 2.94 170 1.25
NaH2PO4 · H2O 150 1.1
CaCl2 · 2H2O 150 1.0 22 0.15 440 3.0 166 1.1 96 0.65
K2SO4 990 5.7
Micronutrients mM mM mM mM mM
H3BO3 3.0 49 30.9 500 6.2 100 1.6 26 6.2 100
KI 0.75 4.5 4.15 25 0.83 5.0 0.8 4.8
MnSO4 · 4H2O 13.2 59.2 31.2 140 22.3 100 4.4 19.7 22.3 100
ZnSO4 · 7H2O 2.0 7.0 43.1 150 8.6 30 1.5 5.2 8.6 30
CuSO4 · 5H2O 0.025 0.1 1.0 4.0a 0.025 0.1 0.25 1.0
Na2MoO4 · 2H2O 0.25 1.0 1.2 5.0 0.25 1.0 0.25 1.0
CoCl2 · 6H2O 0.025 0.1 0.13 0.55 0.025 0.1
FeSO4 · 7H2O 27.8 100 27.8 100 27.8 100 27.8 100 27.8 100
Na2EDTA · 2H2O 37.2 100 37.2 100 37.2 100 37.2 100 37.2 100
Organic constituents
Myo-inositol 100 560 20 111 100 560 100 560 100 560
Nicotinic acid 1.0 8.1 0.5 4.1 0.5 4.1 0.5 4.1 0.5 4.1
Pyridoxine-HCl 1.0 4.9 0.1 0.49 0.5 2.4 0.5 2.4 0.5 2.4
Thiamine-HCl 10 30 0.1 0.3 0.1 0.3 1.0 3.0 1.0 3.0
Glycine 2.0 26.6 2.0 26.6 2.0 26.6
g/L mM g/L mM g/L mM g/L mM g/L mM
Sucrose 20 58.4 30 87.6 30 87.6 20 58.4 20 58.4
pH 5.7 5.5 5.7 5.7 5.7
a
L.B., Davin, personal communication

2.4. Protoplasts 1. Preplasmolysis solution, enzyme solution, and nutrient

medium for protoplast cultivation according to Table 3.
2. Syringes.
3. Syringe filters (0.2 mm pore size).
4. Forceps.
5. Scalpels.
6. 96% (v/v) ethanol for flaming.
 Plant Tissue Cultures 5

Table 3
Solutions for protoplast preparation and cultivation

Preplasmolysis solution
B5/MS macroelements
B5/MS Microelements
sucrose 60 mM
Mannitol/sorbitol 0.3–0.5 M
pH 5.7 (see Note 33)
Sterilise in autoclave.
Enzyme solution (make fresh each time)
0.5% (w/v) Cellulase and 0.2% (w/v) Macerase or 0.1–4% (w/v) Cellulase, 0.05–2% (w/v)
Pectolyase/Macerase and 0.1–2% (w/v) Hemicellulase in the preplasmolysis solution
Mix gently for 15–30 min to dissolve, filter-sterilise through syringe filters (0.2 mm pore size)
Nutrient medium for protoplast cultivation
B5/MS macroelements
B5/MS microelements
NaFe-EDTA 100 mM
B5/MS vitamins
Sucrose 60 mM
Mannitol/sorbitol 0.3–0.5 M
Myo-inositol 100 mg/L
Plant growth regulators
Auxin (2,4-D/NAA/IAA) 1–10 mM
Cytokinin (BA/2iP/kinetin/zeatin) 0.5–2.5 mM
pH 5.7 (see Note 33)
Sterilise in autoclave
See Table 2 for B5/MS medium constituents

7. Sterile nylon or steel sieves (70–100 mm pore size), screw-cap

centrifuge tubes.
8. 20% (w/v) sucrose solution (autoclaved).
9. Fuchs-Rosenthal modified haemocytometer.
10. Microscopic slides.
11. Cover glasses.
12. Agars with low melting point (m.p.) specifically designed for
protoplast culturing (e.g. A8678 Agar washed, m.p. 25–27°C;
A7921 Agar purified, m.p. 30–35°C, Sigma).
13. Sterile pipette tips.
14. Petri dishes.
15. Parafilm®.
Viability stains:
16. 5–10 mg/mL fluorescein diacetate (FDA) in acetone (stock
solution). This is then diluted immediately prior to use by add-
ing 20 mL of the stock solution to 1 mL of 0.65 M mannitol.
6 Kärkönen et al.

17. 0.025–0.25% (w/v) Evans blue (EVB) in 0.65 M mannitol.

18. 0.025–0.25% (w/v) Methanol blue in 0.65 M mannitol.
19. 0.1% (w/v) Phenosafranine in 0.65 M mannitol.
20. 0.01–0.1% (w/v) Tinopal CBS-X (disodium 4,4΄-bis[2-
sulfostyryl)biphenyl) in 0.65 M mannitol.

2.5. Zinnia Cultures 1. 0.25% Na-hypochlorite.

2.5.1. Germination 2. Mesh strainer.
of Zinnia Seeds 3. Vermiculite.
4. Plastic trays.
5. Growth chamber.
6. Liquid fertiliser: e.g. HYPONeX; N:P:K = 6:10:5 (HYPONeX
Japan, Osaka). Dilute 1:100 before use.

2.5.2. Isolation and Culture 1. Table 4 shows the composition of the nutrient medium (see
of Mesophyll Cells Note 8). Frequency of TE differentiation is optimal when
the nutrient medium is supplemented with 0.89 mM
6-benzyladenine (BA) and 0.54 mM 1-naphtalene acetic acid
(NAA). Medium without BA and/or NAA can be used for
control cultures in which TE differentiation does not occur.
2. 0.1% Na-hypochlorite with 0.001% (w/v) Triton X-100.
3. Sterile distilled water.
4. Sterile labware: Waring-type blender, stainless-steel cups,
nylon mesh (50–80 mm pore size), screw-cap centrifuge tubes,
pipette tips, culture tubes (30 mm internal diameter (i.d.)
× 200 mm, 18 mm i.d. × 180 mm or 12 mm i.d. × 105 mm)
capped with aluminium foil.
5. Revolving drum.
6. Growth chamber.

2.5.3. Observations 1. Glutaraldehyde.

of Zinnia Cells 2. 0.2 mg/mL 4΄,6-diamidino-2-phenylindole (DAPI), 1 mM
SYTO16 in DMSO (Molecular Probes).
3. Microscopic slides.
4. Cover glasses.
5. Haemocytometer.

3. Methods

3.1. Surface The idea of surface sterilisation is to selectively kill micro-organisms

Sterilisation on the plant material without killing the plant tissue. For culture
 Plant Tissue Cultures 7

Table 4
Medium for xylogenic culture of Zinnia mesophyll cells

Constituents Concentration (mg/L) Molarity

Macroelements mM
KNO3 2,020 20
MgSO4 · 7H2O 247 1
CaCl2 · 2H2O 147 1
KH2PO4 68 0.5
NH4Cl 54 1
Microelements I mM
MnSO4 · 4H2O 25 110
H3BO3 10 60
ZnSO4 7H2O 10 35
Na2MoO4 · 2H2O 0.25 1
CuSO4 · 5H2O 0.025 0.1
Microelements II mM
Na2EDTA · 2H2O 37 100
FeSO4 · 7H2O 28 100
Organic growth factors I mM
Myo-inositol 100 560
Nicotinic acid 5 41
Glycine 2 27
Pyridoxine-HCl 0.5 2.4
Thiamine-HCl 0.5 1.5
Biotin 0.05 0.2
Organic growth factors II mM
Folic acid 0.5 1.1
Growth regulators mM
NAA 0.1 0.54
BA 0.2 0.89
g/L mM
Sucrose 10 29.2
d-Mannitol 36.4 200
pH 5.5

initiation, it is important to select a healthy plant tissue as an

explant. If necessary, wash the plant organ with tap water and cut
it to ca. 1 cm pieces. Seeds are surface-sterilised intact. Carry out
the procedures aseptically in the laminar air-flow cabinet.
1. Pretreat the explants for 30–60 sec in 70% (v/v) ethanol.
2. Transfer the pieces into a diluted 1–2% (v/v) Na-hypochlorite
solution supplemented with couple of drops of Tween 20.
Incubate for 5–30 min (see Note 9) with occasional shaking.
8 Kärkönen et al.

3. Rinse the explants carefully with sterile distilled water

(three times with at least 1 min incubation in each rinse to
wash all surface sterilants away). Use alcohol-flamed forceps
to transfer the pieces from one solution into the other (see
Note 10).
4. As cut surfaces of the plant material are injured by contact
with the surface sterilising agent, cut the surfaces fresh with a
sterile scalpel by using a half of a sterile Petri dish as a cutting
board.
5. Aseptically dissect the tissue of interest (e.g. embryo, cambial
strips) out of the seed/plant organ and place it onto the sur-
face of the initiation medium. Use a stereomicroscope in the
laminar air-flow cabinet if needed. Seal the dish by using a
strip of Parafilm®.

3.2. Growth Conditions The temperature and light requirements depend on the plant species.
If no information exists in literature in relation to in vitro growth
conditions of the species (or related species) of your interest, it
might be useful to choose the light and temperature conditions in
which the plant grows in vivo. Usually, a constant temperature
(e.g. +25°C) is used; alternatively, the temperature is reduced for
night (+25°C during day, +20°C during night). The quality of
light is obtained by selection of lamps. Some species, like Norway
spruce, prefer fluorescent warm white lamps (5, 12).
The intensity of light and its rhythm are very important. It is
basically by trial and error you can estimate these values unless
some information is available in literature. In general, light inten-
sities of 20–200 mmol/m2/s are used. Some in vitro cultures,
however, are cultivated in the dark.

3.3. Maintenance After 2–6 weeks in culture, callus growth becomes visible at the
edges of the explant (Fig. 1). You have to subculture callus in
order to supplement the cells with fresh nutrients and growth
regulators. Subculture cells in 1- to 4-week intervals depending
on the growth of callus. You might need to modify the nutrient
medium at this stage in relation to nutrient and growth regulator
concentrations and types (see Notes 1 and 11).
1. Subculture callus by transferring the freshest cells (usually at
the edges of callus) onto the fresh medium with flamed for-
ceps (see Note 10). The size of the inoculum should be kept
in a constant size (ca. 0.9 × 0.9 × 0.5 cm; see Note 12). Do
not transfer inoculums which are too small in size because it
takes longer for cells to start dividing when they are subcul-
tured. Alternatively, if you subculture inoculums that are too
big, the cells divide very fast and enter the stationary phase
early (they also fill the growth container). This means more
frequent subculturing.
 Plant Tissue Cultures 9

Fig. 1. (a) Callus culture of Norway spruce (Picea abies). (b) Cell suspension culture of
Norway spruce composed of single cells and small cell aggregates.

2. You can also transfer callus into liquid culture (Fig. 1b).
For this, make the nutrient medium without the gelling agent,
aliquot it in 25 mL aliquots in 100 mL flasks (see Note 13),
close the flask with a double layer of aluminium foil, and auto-
clave. Inoculate the most friable callus cells into the liquid
medium (ca. 0.5 g of cells into 25 mL medium). It depends
on the type of callus whether you will get a fine cell suspension
with single cells and small cell aggregates or whether the callus
grows in big clumps with no cell detachment.
3. For aeration, keep the cultures on an orbital shaker (100 rpm)
in the same growth conditions as cultures on solid medium.
4. Subculture at regular intervals into fresh medium (see above)
by letting cells to settle down to the bottom of the flask.
Decant some culture medium off. Transfer ca. 5 mL cells into
20 mL of fresh medium, for example, with the help of a 5-mL
cut, autoclaved pipette tip or with a measuring cylinder.

3.4. Protoplasts Protoplasts are plant cells that have their cell wall removed by
digestion with plant cell wall-degrading enzymes pectinases,
hemicellulases, and cellulases (Table 5). Protoplasts can be iso-
lated enzymatically in two different ways. In a two-step method,
the cells are first separated to cell suspension with pectinases that
digest the pectinous middle lamella between cells. Then the
remaining cell walls are digested with cellulases and hemicellu-
lases. The one-step method uses a mixture of pectinases and cel-
lulases simultaneously for cell wall digestion (13).
Protoplasts can be produced from intact plant parts such as
root tips and leaves, or from suspension-cultured and callus cells.
10 Kärkönen et al.

Table 5
Some commercially available cell wall-digesting enzymes utilised
in protoplast isolation

Enzyme Source Supplier

Pectin-digesting enzymes
Macerozyme R-10 Rhizopus sp. Yakult Honsha, Japan
Macerase Rhizopus sp. Calbiochem
Pectinase Aspergillus niger Sigma
Pectolyase Aspergillus japonicus Sigma
Hemicellulose-digesting enzymes
Hemicellulase A. niger Sigma
Viscozyme Aspergillus sp. Novozymes Corp.
Cellulose-digesting enzymes
Onozuka R-10 Trichoderma viride Yakult Honsha, Japan
Cellulysin T. viride Calbiochem
Driselase Basidiomycetes sp. Sigma

Having no cell wall, protoplasts are very sensitive to osmotic stress

and must be handled in an isotonic/slightly hypertonic solution
to prevent rupture. Nutrient medium requirements of protoplasts
are quite similar to those of cultured plant cells. Extra calcium is
supplemented to stabilise plasma membranes, and optimisation of
commonly used MS (14) and B5 (15) medium for different
species is often essential.
Protoplasts can be used in plant breeding either through pro-
toplast fusion of related species or in transformation. After cell
wall development, regenerable cells can be induced to plant for-
mation. Protoplasts are also an excellent model to study cell wall
synthesis or transport through cell membranes. Cell wall develops
in protoplasts normally during 24–36 h incubation after which
cells are capable of division. Protoplasts lose their characteristic
spherical shape once the wall formation is complete (Fig. 2).

3.4.1. Protoplast Isolation Prepare the preplasmolysis and enzyme solutions according to
Table 3. Then continue as described below.
Leaves:
1. Surface sterilise young, fully expanded leaves as described in
Subheading 3.1.
2. Cut the leaf into narrow sections with a sharp scalpel in a
Petri dish that contains a small volume (10 mL) of the pre-
plasmolysis solution. Peeling of abaxial epidermis accelerates
cell wall digestion by the enzymes as they enter the intracel-
lular spaces more easily (see Note 14).
 Plant Tissue Cultures 11

Fig. 2. (a) Protoplasts made of Nicotiana tabacum leaves. (b) After a couple of days in
culture, the cell wall has regenerated and the cell has divided. (Photograph courtesy of
Enni Väisänen, University of Helsinki).

Suspension-cultured cells:
3. Centrifuge actively growing cell suspension culture (10 mL) in
the early logarithmic or exponential stage of growth for
5–10 min at 50–100 × g to separate the cells from the culture
medium.
4. Decant medium after centrifugation and transfer the cells into
a Petri dish with the preplasmolysis solution (see above).
Callus culture:
5. Transfer actively growing callus cells (from the edges of callus
pieces) into a Petri dish containing the preplasmolysis solution.
6. Incubate plant material in the preplasmolysis solution. After
30 min, replace the preplasmolysis solution with the enzyme
solution. Incubate for 0.5–20 h (see Note 15) in the dark at
room temperature.
7. After incubation, shake the Petri dish gently to see that the
tissue is digested; if not, incubate for 1–2 more hours.
8. To remove cell debris, pipette protoplasts through a nylon or
a steel sieve (70–100 mm pore size) into a sterile screw-cap
centrifuge tube. Centrifuge for 5–10 min at 50–100 × g.
9. Resuspend the protoplast pellet in the preplasmolysis solu-
tion. Alternatively, fractionate protoplasts from the cell debris
by pipetting the protoplast suspension on top of a 20% (w/v)
sucrose solution.
12 Kärkönen et al.

10. Centrifuge for 5–10 min at 50–100 × g. Cell debris sediments to

the bottom of the tube and protoplasts float at the interface of
the sugar layer and the enzyme solution. Transfer protoplast on
top of a fresh sucrose solution by pipetting, repeat washing for
three times (see Note 16). Resuspend protoplasts in the nutri-
ent medium at an appropriate density (see Notes 17 and 18).

3.4.2. Protoplast Viability Protoplast viability can be detected with different dyes which indi-
Tests cate viable or non-viable cells. Appropriate osmoticum has to be
added to the staining solution to avoid protoplast bursting. EVB is
excluded from living cells and only dead cells are stained blue.
Methanol blue (MB) enters both living and dead cells, but in living
cells the dye is reduced to a colourless compound. Phenosafranine
(PS) enters to dead protoplasts staining them red.
Fluorescent dyes: FDA accumulates inside protoplasts. In via-
ble cells FDA is cleaved to fluorescent fluorescein by an esterase.
Tinopal CBS-X is capable of permeating only dead cells (16, 17).
1. Select the dye you will use in your viability staining. Prepare
it as described in Subheading 2.4.
2. On a microscopic slide, mix equal volumes of staining solution
and the protoplast suspension and overlay with a cover glass.
3. Observe EVB, MB, or PS in a light microscope and count the
number of dead protoplasts per all protoplasts in some fields
(see Note 19).
4. Observe FDA in a fluorescent microscope with the excitation
and emission of 440–490 nm and 510 nm, respectively
(FITC, fluorescein isothiocyanate filter combination). Living
protoplasts have bright fluorescence (see Note 20).
5. Use excitation and emission of 334–385 nm and 420 nm,
respectively, for Tinopal CBS-X. Viable protoplasts have blue
fluorescence (see Note 21).

3.4.3. Culturing Protoplasts are usually cultured on semi-solid agar-containing

of Protoplasts medium or in liquid medium. The salts of MS (14) or B5 (15)
medium supplemented with an extra osmoticum, sorbitol, mannitol,
sucrose, or glucose are usually suitable (Table 3) (see Note 17).
1. Mix double density protoplast suspension with molten agar
(see Note 22), at a double concentration as required in the
final culture. Make sure that the agar is not too warm since
this kills your protoplasts (agar should be just above its melt-
ing point, ca. ³30°C).
2. Pipette quickly as small droplets (100–200 mL) or plate evenly
onto Petri dish.
3. Seal plates with Parafilm® and incubate in diffuse light
(5–10 mmol/m2/s) at room temperature.
 Plant Tissue Cultures 13

3.5. Special Case: Fukuda and Komamine established an in vitro experimental system
Zinnia Cultures in which single mesophyll cells of Z. elegans redifferentiate directly
into TEs independently of cell division (Fig. 3) (2). During TE
formation, cell wall structures undergo dynamic changes, such
as localised thickenings and lignification of secondary cell walls,
partial degradation of primary cell walls, and perforation at the
longitudinal end(s). In Zinnia xylogenic culture, concurrently
with the secondary cell wall formation in developing TEs, active
cell wall degradation takes place; pectin is one of the most actively
degraded substances (18). Thus, taking advantage of in vitro
xylogenic culture system, mechanisms concerning the structural
changes of cell walls can also be studied.

3.5.1. Germination The first true leaves of 14-day-old seedlings of Z. elegans are used
of Zinnia seeds for the in vitro xylogenic culture. Mesophyll cells should be pre-
pared from healthy leaves carefully grown under optimal
conditions.
1. Surface sterilise seeds of Z. elegans cv. Canary bird or Envy in
0.25% Na-hypochlorite solution for 10 min with occasional
shaking.
2. Wash the seeds with running water for 10 min in a mesh
strainer (see Note 23).
3. Sow seeds in moistened vermiculite (0.1 g of seeds/100 cm2)
in plastic trays (see Note 24).

Fig. 3. A mesophyll cell and a TE formed in in vitro Zinnia xylogenic culture. (a) A single
mesophyll cell just after isolation. (b) A TE with a thickened secondary cell wall.
14 Kärkönen et al.

4. Grow seedlings at 25° C for 14 days under a cycle of 14 h

light (approx. 100 mmol/m2/s, white light from fluorescent
lamps) and 10 h dark. Humidity in the growth chamber
should be kept under 45%. Water when surface of vermiculite
is dry (see Note 25). Feed 100-fold diluted liquid fertiliser
(HYPONeX; N:P:K = 6:10:5; HYPONeX Japan, Osaka) once
at the fourth day after sowing.

3.5.2. Isolation and Culture Because of weak attachment between mesophyll cells of Z.
of Mesophyll Cells elegans, single mesophyll cells can be isolated by mechanical
maceration using a Waring-type blender. Epidermal and vascular
cells are removed by filtration of the cell homogenate through
a nylon mesh because of their strong adhesion to each other.
Steps of isolation and culture of mesophyll cells are described
below.
1. Harvest first true leaves (80–120 leaves) that are 3–4 cm in
length (see Note 26).
2. Surface sterilise leaves for 10 min in 0.1% Na-hypochlorite
solution supplemented with 0.001% (w/v) Triton X-100 with
occasional stirring (see Note 27).
3. Rinse the leaves with autoclaved water three times (see Notes 23
and 28).
4. Transfer the leaves into a 100-mL stainless-steel cup contain-
ing 60 mL of nutrient medium.
5. Macerate the leaves at 10,000 rpm for 40 s using a Waring-
type blender (Fig. 4a, b, see Note 29).
6. Filter the homogenate through a nylon mesh (Fig. 4c, pore
size 50–80 mm) by pipetting using a large-bore pipette. Wash
the homogenate that remains on the nylon mesh with 40 mL
of additional nutrient medium (see Note 30).
7. Centrifuge the filtrate at 200 × g for 1 min.
8. Remove and discard the supernatant with a pipette or by
decantation. Suspend the pelleted cells in 80 mL of nutrient
medium by gentle shaking.
9. Centrifuge again at 200 × g for 1 min.
10. Resuspend the pelleted cells in nutrient medium at a cell den-
sity of ca. 8 × 104 cells/mL (see Note 18).
11. Distribute the cell suspension into culture tubes (20 mL for a
tube of 30 mm i.d. × 200 mm, 3 mL for a tube of 18 mm
i.d. × 180 mm, and 1 mL for a tube of 12 mm i.d. × 105 mm)
capped with aluminium foil.
12. Incubate cultures in darkness at 25–27°C on a revolving
drum at 10 rpm at an angle of elevation of 8° (Fig. 4d, see
Note 31).
 Plant Tissue Cultures 15

Fig. 4. Experimental apparatus used for the culture of Zinnia mesophyll cells.
(a) A Waring-type blender (b) Two sets of the stainless-steel cup and a blade used with
the blender (c) A nylon mesh is attached to a cylinder and set on a glass beaker for use
(d) A revolving drum, which is placed in a temperature-controlled incubator or room.

3.5.3. Determination At 72 h of culture, 30–50% of cells synchronously differentiate

of Frequencies of TE into TEs. These can be easily identified by characteristic patterns
Differentiation and Cell of secondary cell walls, which are observed even under a light
Division microscope (see Note 32). Therefore, the number of TEs formed
can be counted using haemocytometer without any pre-treatment.
The frequency of TE formation is determined as the number of
TEs per number of living cells plus TEs. The frequency of cell
division can be estimated from the number of septa, since initially
all mesophyll cells are single.

3.5.4. Observation TEs are distinguishable from other cells by their peculiar cell wall
of Zinnia Cells thickenings seen under a light microscope as described above.
TEs can also be detected by staining of lignified secondary cell
walls with phloroglucinol-HCl (19) or fluorochrome-conjugated
wheat germ agglutinin (20). Isolated cells of Z. elegans are suit-
able for observation under a fluorescence microscope and a con-
focal laser scanning microscope as well.
Upon maturation of TEs, intracellular components including
nuclei are lysed autonomously. This stage of differentiation can
be monitored by staining of nuclei with a DNA-specific fluoro-
chrome, DAPI.
1. Fix the cells by adding glutaraldehyde to a final concentration
of 2% (v/v).
16 Kärkönen et al.

2. Add 1/100 volume of 0.2 mg/mL DAPI and incubate briefly

in dark. Observe the nuclei under ultraviolet light using a
fluorescence microscope.
3. To visualise the nuclei in living TEs, add 1/1,000 volume of
1 mM SYTO16 and incubate for 10 min. Detect using a fluo-
rescent microscope. The dye is excited at 488-nm and the
fluorescence is detected at 515–545 nm (21).

4. Notes

1. Sometimes it is useful to include an undefined mixture of

organic substances (e.g. casein hydrolysate (0.1–1 g/L),
coconut milk (3–10%, v/v)) to the nutrient medium for culture
initiation. During the maintenance growth this is gradually
omitted, if possible, as the exact composition of the mixture
is not known and varies according to the lot.
2. Cytokinins (e.g. 6-benzyladenine (BA), 6-(g,g-dimethylal-
lylamino)purine (2iP), kinetin, zeatin) are usually dissolved in
a few drops of alkali (1 M NaOH), then filled with water to
the final volume. Auxins (e.g. 2,4-dichlorophenoxyacetic acid
(2,4-D), 3-indoleacetic acid (IAA), 1-naphthalene acetic acid
(NAA)) are dissolved in a few drops of absolute ethanol.
Boiling water (warmed in a water-bath) is poured over to let
the ethanol evaporate. After the solution has cooled, adjust
the final volume. Store at +4°C.
3. Store stock solutions of macroelements and microelements at
+4°C. Stock solutions of many macroelements (10×) can be
autoclaved to increase their storage time.
4. Iron can be supplied as NaFe(III)EDTA chelate. Make a sep-
arate stock solution (100×) out of this. Store at +4°C.
5. Mixtures of organic compounds, like vitamins, are prepared
as 1,000× stock solutions. Aliquot the stock solution (e.g.
1 mL aliquots) and store in –20°C freezer.
6. Heat-labile compounds (e.g. certain growth regulators, some
amino acids) are added to the autoclaved medium (when
cooled to ca. 60°C) by filter-sterilisation through syringe
filters (0.2 mm pore size).
7. Depending on the nutrient medium, you may be able to store
the ready-made dishes for some time. After plating the agar-
containing medium onto Petri dishes, let the medium solidify.
Pack the dishes into clean plastic bags (the ones that con-
tained the empty dishes) in a laminar air-flow cabinet, close
the bag with tape. Store at room temperature, in the dark,
agar-side down.
 Plant Tissue Cultures 17

8. A mixture of macroelements can be stored as a 50× stock

solution. Microelements I, microelements II, organic growth
factors I, and organic growth factors II can be stored sepa-
rately as 400× stock solutions. Microelements II should be
autoclaved to form a chelate. Folic acid should be dissolved
by adding a small amount of NaOH.
9. For leaf material, 5–10 min in Na-hypochlorite may be
enough. For stem fragments and seeds, 20–30 min will be
necessary.
10. After flaming, let the forceps cool down before touching the
explants. Cooling can be done, for example, by dipping into
sterile water or by pressing into the agar.
11. Sometimes the growth of callus ceases on nutrient medium
where the callus has previously grown well. It might help if
you transfer the cells onto a medium where either cytokinin
or auxin is depleted. If the growth continues, the cells have
started to produce the growth hormone by themselves. This
phenomenon is called habituation (22, 23).
12. At the beginning, especially if there is only little callus
growth, it is good to transfer the whole explant with the
new growth onto the fresh nutrient medium. Only when
you have enough callus, separate it from the explant for
subculturing.
13. You can use different volumes of liquid cultures, but make
sure that the medium to flask ratio is similar to 25 mL culture
per 100 mL flask. This is to make sure that enough air space
exists in the culture flask for gas exchange.
14. Epidermis can be easily removed from surface-sterilised leaves
with the help of forceps and a scalpel. It is also possible to
make protoplasts separately from the epidermal cell layer and
mesophyll cells.
15. Duration of incubation has to be determined for each plant
material.
16. Alternatively, protoplasts can be fractionated with Ficoll (10%
in 0.6 M mannitol) or with Percoll (20% with 0.25 M man-
nitol and 0.1 M CaCl2).
17. Some plant species prefer glucose (100–200 mM) to sucrose.
18. Protoplast/cell density can be determined by Fuchs-Rosenthal
modified haemocytometer. For many species, protoplast den-
sity of 1×103–1×105 is suitable.
19. Notice that most dyes are quite toxic and prolonged incuba-
tion of protoplasts in staining solution may kill cells.
20. With FDA, you have to count the number of all protoplasts
in the bright field since only the viable ones are detectable in
the dark field.
18 Kärkönen et al.

21. Tinopal CBS-X-stained protoplasts can be counted by

simultaneous illumination with UV and visible light.
22. Specifically designed agars for protoplast culture remain liquid
down to their melting point, which is ca. ³30°C. Plating into
agar facilitates further observations of protoplasts as they
become stationary. Agar concentration should be low enough
(0.6%, w/v) to give a soft medium. Agars with low melting
point are recommended for temperature-sensitive protoplasts
not successful in liquid culture.
23. Na-hypochlorite should be thoroughly removed after surface
sterilisation of seeds and leaves.
24. The size of the plastic trays we routinely use is 40 × 30 × 6.5 cm,
and the depth of vermiculite is about 4 cm. Seeds should be
covered with a little vermiculite after sowing. Seeds and liquid
fertiliser should be equally distributed in a plastic tray so that
seedlings grow uniformly.
25. Since too much watering of seedlings often causes serious dis-
eases, water only when the surface of soil is dry. During water-
ing, leaves should be kept dry. Use of leaves with splashes often
leads to bacterial contamination in the subsequent cell culture.
26. During harvesting, healthy leaves without withered area and
rough surface should be cut off with a pair of scissors steri-
lised with 70% (v/v) ethanol. Leaves harvested should be
kept in water in a plastic container until the next step.
27. Leaves are damaged when they are soaked in Na-hypochlorite
solution for longer than 10 min.
28. All steps between rinse of sterilised leaves and distribution of cell
suspension to tubes must be done under aseptic conditions.
29. The optimal condition for maceration of leaves depends on
the plant materials and the type of the blender used. Speed
and time of maceration should be adjusted to keep the number
of dead cells low and the number of living cells collected high.
Single mesophyll cells can also be isolated using mortar and
pestle with gentle maceration in the nutrient medium.
30. Application of filtrated cell suspension onto remaining homo-
genates on the nylon mesh increases the number of cells col-
lected (optional step).
31. Alternatively, 25 mL of cell suspension can be incubated in a
100-mL flask using a rotary shaker at 40 rpm.
32. Zinnia cells fixed with 0.25% (v/v) glutaraldehyde can be
stored at 4°C for at least a few months without significant
visual change.
33. pH can be adjusted to 5.7 with KOH. Alternatively, pH can
be buffered to 5.7 with 0.5% (w/v) MES (2-[N-Morpholino]
ethanesulfonic acid)-KOH.
 Plant Tissue Cultures 19

Acknowledgements

We thank University of Helsinki and the Academy of Finland for

the financial support of the work (A.K., A.S.). We also thank
Teresa Pehkonen (University of Helsinki) for useful comments
and Enni Väisänen (University of Helsinki) for allowing us to
include the protoplast figures.

References
1. Pierik, R.L.M. (1997) In vitro culture of higher 11. Ito, Y., Nakanomyo, I., Motose, H., Iwamoto,
plants. 4th ed. Kluwer, Dordrecht. 348 p. K., Sawa, S., Dohmae, N., and Fukuda, H.
2. Fukuda, H., and Komamine, A. (1980) (2006) Dodeca-CLE peptides as suppressors
Establishment of an experimental system for of plant stem cell. Science 313, 842–845.
the tracheary element differentiation from 12. Simola, L.K., and Santanen, A. (1990)
single cells isolated from the mesophyll of Improvement of nutrient medium for growth
Zinnia elegans. Plant Physiol 65, 57–60. and embryogenesis of megagametophyte and
3. Eberhardt, T. L., Bernards, M. A., He, L., Davin, embryo callus lines of Picea abies Physiol. Plant
L. B., Wooten, J. B., and Lewis, N. G. (1993) 80, 27–35.
Lignification in cell suspension cultures of Pinus 13. Bajaj, Y.P.S. (1996) Plant protoplasts and
taeda. In situ characterization of a gymnosperm genetic engineering VII. Springer, Berlin. 317
lignin. J Biol Chem 268, 21088–21096. p. ISBN 3-540-60876-1.
4. Brunow, G., Ämmälahti, E., Niemi, T., Sipilä, 14. Murashige, T., and Skoog, F. (1962) A revised
J., Simola, L. K., and Kilpeläinen, I. (1998) medium for rapid growth and bio assays
Labelling of a lignin from suspension cultures with tobacco tissue cultures. Physiol Plant 15,
of Picea abies. Phytochemistry 47, 1495–1500. 473–497.
5. Kärkönen, A., Koutaniemi, S., Mustonen, M., 15. Gamborg, O. L., Miller, R. A., and Ojima, K.
Syrjänen, K., Brunow, G., Kilpeläinen, I., Teeri, (1968) Nutrient requirements of suspension
T.H., and Simola, L. K. (2002) Lignification cultures of soybean root cells. Exp Cell Res
related enzymes in Picea abies suspension cul- 50, 151–158.
tures. Physiol Plant 114, 343–353. 16. Widholm, J. M. (1972) The use of fluorescein
6. Kärkönen, A., and Koutaniemi, S. (2010) diacetate and phenosafranine for determining
Lignin biosynthesis studies in plant tissue cul- viability of cultured plant cells. Stain Technol
tures. J Integr Plant Biol 52, 176-185. 47, 189–194.
7. Fukuda, H. (1997) Tracheary element differ- 17. Huang, C. N., Cornejo, M. J., Bush, D. S.,
entiation. Plant Cell 9, 1147–1156. and Jones, R. L. (1986) Estimating viability
8. Demura, T., Tashiro, G., Horiguchi, G., of plant protoplasts using double and single
Kishimoto, N., Kubo, M., Matsuoka, N., staining. Protoplasma 135, 80–87.
Minami, A., Nagata-Hiwatashi, M., Nakamura, 18. Ohdaira, Y., Kakegawa, K., Amino, S.,
K., Okamura, Y., Sassa, N., Suzuki, S., Yazaki, Sugiyama, M., and Fukuda, H. (2002) Activity
J., Kikuchi, S., and Fukuda, H. (2002) of cell-wall degradation associated with differ-
Visualization by comprehensive microarray entiation of isolated mesophyll cells of Zinnia
analysis of gene expression programs during elegans into tracheary elements. Planta 215,
transdifferentition of mesophyll cells into 177–184.
xylem cells. Proc Natl Acad Sci USA 99, 19. Siegel, S. M. (1953) On the biosynthesis of
15794–15799. lignin. Physiol Plant 6, 134–139.
9. Fukuda, H. (2004) Signals that govern plant 20. Hogetsu, T. (1990) Detection of hemicellulo-
vascular cell differentiation. Nat Rev Mol Cell ses specific to the cell wall of tracheary elements
Biol 5, 379–391. and phloem cells by fluorescein-conjugated
10. Motose, H., Sugiyama, M., and Fukuda, H. lectins. Protoplasma 156, 67–73.
(2004) A proteoglycan mediates inductive 21. Obara, K., Kuriyama, H., and Fukuda, H.
interaction during plant vascular develop- (2001) Direct evidence of active and rapid
ment. Nature 429, 873–878. nuclear degradation triggered by vacuole
20 Kärkönen et al.

rupture during programmed cell death in culture of rice through comparative experiments
Zinnia. Plant Physiol 125, 615–626. on the nitrogen sources Sci Sin 18, 659–668.
22. Christou, P. (1988) Habituation in in vitro 25. Lloyd, G. and McCown, B. (1980)
soybean cultures. Plant Physiol 87, 809–812. Commercially-feasible micropropagation of
23. Piscke, M. S., Huttlin, E.L., Hegeman, A. D., mountain laurel, Kalmia latifolia, by use of
and Sussman, M. R. (2006) A transcriptome- shoot-tip culture. Comb Proc Internat Plant
based characterization of habituation in plant Prop Soc 30, 421–427.
tissue culture Plant Physiol 140, 1255–1278. 26. Owen, H. R., Miller, A. R. (1992) An exami-
24. Chu, C. C., Wang, C. C., Sun, C. S., Hsu, C., nation and correction of plant tissue culture
Yin, K. C., Chu, C. Y., and Bi, B. Y. (1975) basal medium formulations. Plant Cell Tiss
Establishment of an efficient medium for anther Org Cult 28, 147–150.
 Chapter 2

Computerized Molecular Modeling of Carbohydrates

Alfred D. French and Glenn P. Johnson

Abstract
Computerized molecular modeling continues to increase in capability and applicability to carbohydrates.
This chapter covers nomenclature and conformational aspects of carbohydrates, perhaps of greater use to
carbohydrate-inexperienced computational chemists. Its comments on various methods and studies
might be of more use to computation-inexperienced carbohydrate chemists. New work on intrinsic vari-
ability of glucose, an overall theme, is described.

Key words: Carbohydrate, Disaccharide, Conformation, Puckering, Modeling, Quantum mechanics,

Molecular mechanics

1. Introduction

Various computer modeling software systems provide a graphical

user interface with a drawing function that lets the user start with-
out any information other than a vision in their head and a knowl-
edge of the pattern of atom connectivity. Such software “knows”
about atomic diameters, usual bond lengths, angles, etc., so it will
assist the user in creating a structure that may highly resemble the
actual molecule as represented by balls and/or sticks. Different
display options for the molecules provide a sense of artistic accom-
plishment and better convey information and may even enhance
the appeal of a particular research program. Still, the validity of
such pictures of typical carbohydrate molecules is questionable
because there are many alternative structures (conformations or
shapes) that can be formed by mostly changing the torsion angles.
Figure 1 shows the three staggered conformations of n-butane and

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_2, © Springer Science+Business Media, LLC 2011

21
22 French and Johnson

Fig. 1. Staggered conformations of butane, with their C1–C2–C3–C4 torsion angles indicated. Newman projections are
also shown. The angles refer to the angle of the C1–C2 bond relative to the C3–C4 bond, when viewed down the C2–C3
bond. The C1–C2–C3–C4 torsion angle is 0° when all four atoms are in a plane and C1 and C4 are cis. The vertical rods
indicate the axes about which the C4 groups rotate. All three conformations (−gauche, trans, and +gauche) correspond
to minima in the energy, but the 180° conformer has the lowest energy, i.e., it is the global minimum. The gauche forms
are also called syn, and the trans form is also called anti.

their torsion angles. Deciding which conformer best represents the

“real” population of molecules requires a measure of the relative
free energy of each alternative. That may seem simple, but for a
carbohydrate there are actually two complicated issues. One is
how to calculate the energy, and the other is to decide which
alternative structures to consider and which to ignore. Both issues
are typically approached with assumptions and approximations
that are being reduced as computer power and software sophisti-
cation increase.
Regarding alternative structures, the pyranosyl ring of aldo-
hexose sugars is often assumed to be in a chair form, even though
there are alternatives. In fact, idopyranose appears to have signifi-
cant populations for both chairs and a skew form of the ring in
solution (1). Even glucose has conformational ambiguity, as some
methylated or acylated cyclodextrin molecules have glucose resi-
dues that take alternative chair or skew forms in their experimen-
tally determined crystal structures (2, 3).
Energy can be calculated in many ways, starting historically
with a simple scan for short distances between atoms that are not
bonded to each other. If such contacts exist, that kind of analysis
would assign a potential energy of infinity, reflecting complete
 Computerized Molecular Modeling of Carbohydrates 23

improbability. One mainstream approach is Molecular Mechanics

(MM). The energy terms of these empirical force-field models
arise from our useful “cartoons” of molecular structure, with
potential functions for bond stretching, angle bending, and
charge–charge interactions. Quantum mechanics (QM) calcula-
tions consider distributions and interactions among individual
electrons. Thus, QM is also called electronic structure theory.
QM calculations are so expensive in terms of computer time and
memory that trisaccharides (4) are about the largest molecules to
consider. Also, explicit treatments of neighbor molecules such as
solvent or adjacent molecules in a crystal are rarely included in
QM calculations, nor are most of the alternative structures, all
because of the time required.
Sometimes errors in calculated energy cancel each other.
When comparing two different molecular shapes, the same errors
may affect both forms. Even if the absolute energy values are not
correct, the relative energies could still be useful. Also, there is a
useful but ultimately unreliable compensation for the lack of
explicit treatment of electron correlation in the Hartree–Fock
(HF) QM method. That error can often be countered by errors
from using a small basis set such as 6-31G*. Combining the HF
method and the 6-31G* basis set gives HF/6-31G*, a “magic”
level of theory (5).
To some extent, the present chapter follows up on proceed-
ings from a symposium on carbohydrates in 1989 (ACS Symposium
Series 430). Besides many articles in chemical and carbohydrate
journals, special issues have been at least partly dedicated to
­carbohydrate modeling (Molecular Simulation (vol. 4, issue 4);
THEOCHEM (vol. 395–397), Carbohydrate Research (vol. 340,
issue 5), ACS Symposium Series 930). Some of the issues treated
herein are covered in more detail in a recent review (6).

2. Structural
Descriptors
of Carbohydrates
Because most carbohydrates have numerous asymmetric centers, they
2.1. Nomenclature have kept their traditional nomenclature (7). b-d-­Glucopyranose
should be easier to remember than (2R, 3R, 4S, 5S, 6R) 2,3,4,5
tetrahydroxy, 6 methoxy oxacyclohexane. Another point is that
the first carbon atom in the parent acyclic sugar is number 1
instead of the heteroatom in the sugar ring as would be the case
if standard organic chemistry nomenclature were used. Suppose a
disaccharide is composed of two d-glucopyranose residues, linked
at the 1¢ and 4 positions. The particulars of the linkage between
the glucose residues define the compound, i.e., whether it is maltose
or cellobiose. The configuration at C1, the reducing end ­anomeric
center, defines whether it is the a- or b-anomer. Thus, b-maltose
24 French and Johnson

and a-cellobiose both exist, despite the opposite configurations

at the anomeric centers of their linkages. Rules for carbohydrates
composed of more than two monosaccharides are covered by
publications cited in (7).

2.2. Drawings To aid recognizability of ring compounds, the anomeric center

(the carbon bound to both the ring oxygen and a hydroxyl in the
native sugar) is drawn on the far right, and the bond between the
ring oxygen and the nonanomeric carbon is parallel to the plane of
the paper and indicated as being at the back of the ring. The ring
oxygen of five-membered (furanosyl) rings is shown at the back.

2.3. Ring Puckering Carbohydrate rings are puckered, not flat. An extensive analysis
of both furanose and pyranose ring shapes is presented in a study
of the conformationally and configurationally ambiguous sugar,
psicose (8). Furanose rings can be envelopes, with four coplanar
atoms and one atom out-of-plane. For example, a ring with its
oxygen atom out of plane is a characteristic form, denoted OE if
the oxygen is above the ring and EO if below. Otherwise, they
are twists, with three coplanar atoms and one atom above the
plane and one below (see Fig. 2 for 3T2 and E2 examples based

Fig. 2. Sample furanosyl rings with C1, O5, and C4 all coplanar. The 3T2 drawing has C3 above the plane, and C2 below,
where the E2 drawing has only C2 out of plane. Both of the b-d-glucopyranosyl rings (hydrogen atoms not shown) are
exactly the same structures, correctly described as OS2 rings. The ring on the right has been rotated about a line between
C1 and C4. Convention dictates that the ring is described as OS2 instead of 3S5. There are two planes that contain four of
the atoms. One contains C1, C3, C4, and C5, and the other contains C1, C2, C4, and O5.
 Computerized Molecular Modeling of Carbohydrates 25

on tetrahydrofuran). Each individual envelope (E) or twist (T)

is a “characteristic form.” Ten E forms of furanose rings exist, as well
as ten T forms, such as 4T3, a favored form for b-d-fructofuranose.
Facile transitions known as pseudorotation are permitted between
the adjacent alternating E and T forms, such as 4E, 4T3, and E3.
The characteristic forms do not necessarily correspond to energy
minima but are merely markers in ring-shape hyperspace for
understanding the shape of a particular ring.
Characteristic pyranosyl forms include two chairs, six boats,
and six skews (twist-boats, see Fig. 2). Also, there are 12 enve-
lopes and 12 half-chairs, and Boyen described 12 additional
(screw-boat) characteristic forms (9). For ordinary sugars, these
latter forms correspond to intermediates or transition states during
conversions and are not stable. There are only two unique chairs,
with convention dictating use of the lowest-numbered ring atoms.
For example, the same glucopyranosyl chair could be described as
4
C1, 2C5, or OC3 but only 4C1 is used. The other form is 1C4.
Similarly, each skew form could be described in two ways (Fig. 2),
but convention dictates the use of the lower-numbered atoms.
Because experimental and computational sugar rings usually
do not correspond exactly to a characteristic conformation, it is
useful to specify the shape quantitatively. That is the job of pucker-
ing parameters. The 15 x, y, z coordinates of the five atoms of a
furanosyl ring are reduced to two puckering parameters, and three
for the positions of the six atoms of a pyranosyl ring. Just as varia-
tions in bond lengths and angles are ignored in ordinary conforma-
tional analysis based on torsion angles, puckering parameters are
intended as overall descriptors, not a completely detailed descrip-
tion. Any computed puckering from an experimental or computed
structure can be translated to a phrase such as “a conformation that
is nearest in puckering space to OS2” (a skew form, with the ring
oxygen above the plane and C2 below the plane that includes C1,
C3, C4, and C5). Even when the shape is obvious, such as the fre-
quently found 4C1 (chair) form, assessment of puckering parame-
ters allows the degree of distortion from the normal shape to be
described to learn the effects of being packed in a crystal or com-
plexed with an enzyme. In another example, the degree of pucker-
ing for glucose could be compared with that of tetrahydropyran to
learn the effects of substituents on the ring shape.
Cremer–Pople puckering parameters (10) work well in most
circumstances but in cases where bond lengths are quite different,
the translation from the computed parameters to a characteristic
conformer may not meet expectations based on a visual assess-
ment. Besides the Cremer–Pople puckering parameters, which
apply to all sizes of rings, there are the Altoona–Sundaralingam
puckering parameters for furanosyl rings (11). Ring puckering
conformations also have descriptors that are derived from endo-
cyclic torsion angles (12–15) or flap angles (16, 17).
26 French and Johnson

Other measures of the ring shape are useful, especially when

they are interrelated with the conformation of polysaccharides.
For example, the distance between O1 and O4 of a-d-glucopyra-
nose varies between 3.88 and 4.84 Å (see below) in crystals of
mono- and oligosaccharides, despite a basic 4C1 shape. If those
residues are used to make models of the amylose polysaccharide,
the resulting helix shape will vary widely. Recently, we observed a
twist of the glucose ring, i.e., the O1–C1…C4–O4 virtual torsion
angle ranges over about 25° (18). Again, this parameter governs
the location of the adjacent residue in model polymers.

2.4. Exocyclic Group The orientation of primary alcohol groups is often an interesting
Orientation variable. Most often the three staggered rotamers are described
as gg (gauche–gauche), gt (gauche–trans), and tg (trans–gauche).
These different conformers are shown below in Fig. 2. The first
of the two letters corresponds, in d-glucose, to O5–C5–C6–O6
torsion angles (w) of −60° (−g), 60° (+g), and 180° (t) (see
Fig. 1). The second letter corresponds to C4–C5–C6–O6 torsion
angles of 60°, 180°, and −60°. Other authors prefer −g, +g, and t
for the O5–C5–C6–O6 angle. Because C5 is sp3 hybridized, the
use of two letters could be considered redundant, but it avoids
needing to remember the sign of the torsion angle. If considering
l-glucopyranose, the signs of gauche forms must be reversed for
the single-letter notation, but with the two-letter notation the
mirror image of d-glucose with O6 gt is l-glucose with O6 gt.
The tg conformation has been described as “forbidden” because
it was not observed in early crystal structure studies of molecules
having the gluco configuration at C4 (O4 equatorial), and some
NMR-based analyses of sugars have yielded a sum of the gt and
gg conformations slightly greater than 100%, implying negative
amounts of tg. More recent experimental work has found exam-
ples, especially in native cellulose I. The galacto configuration,
with O4 axial, has a low population of gg conformations (6).
Secondary alcohol (hydroxyl) orientations are important to the
calculated energy. They are often described in terms of clockwise
and counter-clockwise (reverse clockwise) systems of intra-residue
hydrogen bonds that provide maximum stabilization for isolated
molecules. Despite the significant lowering of the energy due to
having all of the secondary hydroxyl groups appearing to partici-
pate in a continuous donor–acceptor–donor–acceptor network, the
resulting long H…O distances and small O–H…O angles would
yield weak attractions. According to Bader’s Atoms-In-Molecules
(AIM) theory (19) (“electron density gradient vector field analysis”),
these energy-lowering orientations do not result in the bond paths
and bond critical points (20–22) needed for true hydrogen bonds
(23). The primary alcohol can form a hydrogen bond with O4, and
two axial hydroxyl groups with 1,3 spacing (such as in 1C4 glucose)
can form better hydrogen bonds (21).
 Computerized Molecular Modeling of Carbohydrates 27

Fig. 3. Energy distribution for stationary shapes of glucose at the B3LYP/6-31+G** level. Twenty-five of the energies
above 4.3 kcal/mol correspond to “saddle-point” or “transition state” structures.

All staggered rotamers for glucose can be studied by fairly

good QM. Putting each of the six rotatable groups into all three
orientations, there are 36 (=729) combinations. The Jaguar pro-
gram (24) was used with B3LYP/6-31+G** energy minimiza-
tion on each of the conformers as isolated (gas phase) molecules.
Most of the 729 combinations were unstable and one or more of
their exocyclic groups rotated to a different staggered orienta-
tion. Still, there were 150 unique stationary structures, and their
energy range (13.4 kcal/mol) is considerable. Figure 3 shows
their nearly continuous distribution of energies. Figure 4 shows
the six lowest-energy and the six highest-energy forms. The six
lowest-energy structures included the gt, gg, and tg conforma-
tions of O6 and both reverse-clockwise and clockwise secondary
hydroxyl groups. This would seem to support the notion of coop-
erative rings of hydrogen bonds, for which there is some experi-
mental support (25), despite the above AIM studies. Four of the
six highest-energy structures have the hydroxyl hydrogen on O1
located underneath the pyranose ring, a sterically disadvantageous
orientation, and the rings of OH interactions are absent.
The exocyclic group orientations have a substantial effect on
the molecular geometry, including the O1…O4 distance. Figure 5
shows answers to three different questions about this distance.
The top curve shows the probability distribution predicted by
pi = e −∆E / RT for a glucose residue stretched and compressed with
MM3 (26) calculations. The underlying energy curve obeys
Hooke’s Law almost perfectly. The middle graph presents the
28 French and Johnson

Fig. 4. Six lowest energy and six highest-energy stationary forms of a-d-glucopyranose.
Of these structures, only the lower left is in a transition state according to frequency
calculations.

O1…O4 distances in 2,582 experimental examples of a-d-glucose

and its derivatives having 4C1 rings, obtained from a scan of the
Cambridge Structural Database (27). The lowest graph shows the
O1…O4 distances within the 150 stable structures discussed
above. The top and bottom theoretical analyses indicate, respec-
tively, the elasticity of glucose and the intrinsic variability based on
the exocyclic group orientations. The experimental data combine
both these factors, with the elasticity corresponding to deforma-
tions from random crystal packing and strain from being part of
larger molecules, including macrocycles. The mean MM3 distance
 Computerized Molecular Modeling of Carbohydrates 29

Fig. 5. Upper : Probabilities calculated by MM3 for stretched and compressed a-d-­
glucopyranose for different O1–O4 distances. Center : Frequencies of experimental
O1–O4 distances in 2,582 a-d-glucopyranose rings from a scan of the Cambridge
Structural Database. The mean value is 4.356 Å. Lower : Frequencies of intrinsic O1–O4
distances in 150 stationary B3LYP/6-31+G** structures of a-d-glucopyranose.

is longer than the mean experimental value, as is the shortest

B3LYP/6-31+G** result.
Other methods for calculating the energy of the 729 con-
formers will give different numbers of final structures as well as
energy values. All of these 150 structures have the 4C1 shape;
many more stationary points would be found for other ring
shapes. Some of those alternatives will give energies lower than
the higher-energy 4C1 shapes.

2.5. Anomeric Centers In aqueous solution, a single enantiomer such as d-glucose is five
compounds: acyclic, and a- and b-pyranoses and furanoses. The
populations of the furanose and acyclic forms are minimal for glu-
cose, but must be considered for sugars such as the ketohexose,
psicose is a ketohexose (8). Opening and re-closing the ring allows
the configuration of glucose to change at C1, and the resulting
forms, a- and b-glucopyranose, are “anomers.” Experimental data
for reducing sugars are affected by this interconversion, which is
known as mutarotation. It can be avoided by substituting the
hydroxyl hydrogen on the glycosidic oxygen with a methyl group.

2.6. Di-, Oligo-, Atom numbers in the nonreducing residue of disaccharides are
and Polysaccharides primed, while longer molecules have increasing values of Roman
numerals for residues further from the reducing end. Linkages
30 French and Johnson

between monomeric units of larger molecules consist of either

two or three bonds, typically with the oxygen atom attached to
the anomeric carbon (the glycosidic oxygen) leaving during
synthesis. Thus, in the formation of cellobiose, O4 remains.
Disaccharide conformations are specified by the values of the
torsion angles for the glycosidic (C1¢–On) (f) and aglycon
(On–Cn) (y) bonds (n = 4 for cellobiose). Three-bond linkages
involve a primary alcohol group. Its conformation is described
with letters, sometimes upper case, e.g., GG, GT, and TG, or by
the w torsion angle. The central bond is specified by y, and f is
for rotation about the glycosidic bond. Polysaccharides have
disaccharide linkages so the same descriptors apply. For poly-
saccharides composed of regular repeating units, helix nomen-
clature applies. Helices are described by the number of units
per turn (n), and the rise, or advance, (h) along the helix axis.
Along with C1¢, O4, and C4, the four-atom definition of the
torsion angle f in maltose or cellobiose could involve any of the
three atoms H1¢, O5¢, or C2¢. Many workers have used H1¢,
especially if they have a background in NMR and are thinking of
using nuclear Overhauser effects to solve the structure. Others,
mindful of the difficulties in accurately locating hydrogen atoms
by X-ray crystallography, have opted for O5¢. No examples of C2¢
come to mind. The y torsion angle has been defined by all three
possible atoms. For cellobiose, they would be H4, C3 or C5. The
above reasons for favoring a hydrogen atom or a heavier atom
also apply, but the above-cited nomenclature (7) uses the lower-
numbered carbon atom. Thus, the standard definition of y for
cellobiose is C1¢–O4–C4–C3, along with the standard f of O5¢–
C1¢–O4–C4. Greek letter descriptors are italicized.
Standardization of the ends of the 360° ranges of f and y on
plots of energies, experimental points and molecular dynamics
(MD) trajectories would allow quick visual comparisons of differ-
ent plots. We argue for plots that are equivalent to fH and yH (e.g.,
H1¢–C1¢–O4–C4 and C1¢–O4–C4–H4) values from −180° to
+180° for two reasons. Firstly, experimentally determined points
for many reducing disaccharides are mostly in minima that have
fH and yH near 0° and therefore will fall in the center of the map.
If the ranges were 0–360°, the major populations could be sepa-
rated into as many as four visual groups despite close structural
similarity. Secondly, on such plots for maltose and cellobiose the
diagonal line from the lower right to the upper left corresponds
to helices with no chirality, separating right- and left-handed
forms. For maltose, the diagonal line corresponds to polymers
with h = 0. Long molecules having maltose-type linkages and f,y
values on the diagonal line would self-intersect, but molecules
with six, seven or eight glucose residues could form cyclodextrins.
Cellulose polymers with f,y values on the diagonal line are heli-
ces with n = 2. By adding or subtracting 120° as appropriate, the
 Computerized Molecular Modeling of Carbohydrates 31

f y

Fig. 6. HF/6-31G* energies from the depicted methylated maltose analog. Observed conformations in experimental
­crystal structures of maltose and related structures are shown as points. Structures within the 1 kcal/mol contour do not
possess the O2–O3′ hydrogen bond that is found for all other observed experimental structures. Axes correspond to fH
and yH (e.g., H1′–C1′–O4–C4 and C1′–O4–C4–H4) values from −180° to +180°. Note that a copy of this map can be
placed on each edge to test for periodicity.

ranges of f and y defined according to the nonhydrogen atoms

can cover the structures and retain the central location of most
experimental structures and the desirable diagonal line (see Fig. 6,
above).

3. Special
Problems
of Carbohydrates
Because of their many hydroxyl groups and substantial flexibility,
carbohydrates exaggerate many of the problems experienced in
modeling other molecules. Also, while some other molecules such
as dimethoxymethane (DMM) have sequences of atoms that cor-
respond to anomeric centers, such centers are relentlessly preva-
lent in carbohydrates. The following gives an overview of ways to
cope with these problems.

3.1. Sampling With all of the rotatable exocyclic substituents, it is necessary to

ascertain that the modeling results are based on a sufficient explo-
ration of the possible different states. How can one be certain
that the final results depend only on the energy calculations of a
32 French and Johnson

particular modeling method and not on a failure to consider the

best possible arrangements of these exocyclic groups? As seen
above, there are 729 combinations of staggered exocyclic orienta-
tions for glucose, and cellobiose has 59,049. Again, most will not
be stable, but many will and their stabilities are dependent on f
and y. This leaves a large conformation space to sample. One
approach is to use simulated annealing (28).
Two examples come from our recent work. In one (29), we
developed f,y energy maps to provide energies of distortion for
conformations of substrates in complexes with hydrolyzing
enzymes. The goal was to learn whether f,y distortion might be
part of the catalytic function. To sufficiently sample conformation
space, we use different “starting structures,” each with a particu-
lar combination of orientations of the exocyclic groups. The
energy is computed for each of these starting structures at each
f,y location. The f and y values were fixed at grid points in 20°
intervals for a total of (18 × 18=) 324 unique f,y points. Initial
studies, using Monte Carlo methods (30) and the OPLS-2005
(31) force field in MacroModel (24) identified 1,863 stable start-
ing structures for cellobiose when both rings were in the normal
chair forms, 3,485 when the reducing ring had the 2SO conforma-
tion, and 2,871 when the reducing ring had the 3S1 shape.
Thiocellobiose (where sulfur replaces the interresidue oxygen)
yielded 2,277 different structures to test at each f,y point.
In another project (32), methyl cellobioside, -tetraoside, and-
hexaoside were investigated with Replica Exchange Molecular
Dynamics studies in explicit TIP3P water. (Ref. (33) has a tutorial
study on replica exchange MD of disaccharides in vacuum).
Depending on the size of the carbohydrate, 714–3,741 molecules
of water were included in the AMBER (34) calculations, using
the GLYCAM-04 force field (35). To assure complete sampling
for the hexamer, replicas had temperatures of 297–557° in 42
increments, with each simulation lasting more than 16 ns.
To make Ramachandran surfaces, the f and y torsion angles
of a disaccharide conformation are adjusted in increments of
perhaps 10° or 20° and the energy calculated. Over the past
20 years, the energy has often been minimized for each f,y con-
formation, holding f and y constant but allowing all other
parameters to find their nearest local minimum. Such studies are
called “relaxed-residue” analyses. Their primary rationale is that
they avoid collisions that would occur if the rings were kept rigid.
(On one rigid-residue map, the crystallographic conformation
of the sucrose moiety in raffinose corresponded to an unreason-
able 100 kcal/mol (36)). Modeling software often provides for
these calculations to step through f,y space, using a tool called
“dihedral driver” or “scan.” Because the monosaccharide units
of the disaccharide are flexible, there is the strong likelihood that
an inelastic deformation of the molecule will occur during the scan.
 Computerized Molecular Modeling of Carbohydrates 33

For example, a hydroxyl group may simply rotate to a new

­staggered form, or the ring might lose its chair form. If that
deformed structure is used to start the subsequent minimiza-
tion, the energy at −180° will be different from the one at +180°.
This is a sampling issue because the points after the deformation
occurs will not be tested with the intended starting structure.
Our procedure for avoiding this problem is to use the same
starting geometry at each f,y point, with only rigid rotations to
the point in question.
Our approach is not without problems. Not having energy-
minimized structures at each preceding point can lead to an inter-
penetration of the two monosaccharide residues in regions of
high energy. The interpenetration causes much higher calculated
energies or failure of minimization altogether but can be partially
avoided by increasing the glycosidic bond angle to 150° for each
starting structure. A similar sampling problem is a concern during
MD simulations because simulations may be too short. At least
with MD, the deformed structures could recover if the simulation
runs long enough.

3.2. Hydrogen Bonding Hydrogen bonding is especially important for most carbohydrates
because of the high density of hydroxyl groups. If the stabiliza-
tion from a typical moderate to weak hydrogen bond is 5.0 kcal/
mol (37), model structures that have them would completely
dominate those with otherwise similar structures. Further, the
various acceptor oxygen atoms do not appear to accept hydrogen
bonds with equal eagerness. The oxygen atom of the glycosidic
linkage is a poor acceptor, and donations to the ring oxygen are
less stabilizing than to hydroxyl groups. Further, there are coop-
erative effects. These effects are found in continuous donor–
acceptor–donor–acceptor chains. Such sequences are more stable
than an equal number of discontinuous hydrogen bonds and their
H…O distances are shortened (38). In modeling hydrogen
bonds, some workers find that no further consideration is needed
after the charges are assigned to the atoms in their empirical
models. Other modelers have devised elaborate schemes to
­provide calculated energies whenever hydrogen bonding is
­present. With QM, most workers are finding that reliable calcula-
tion of hydrogen-bonding geometries and energies requires fairly
sophisticated techniques, such as post-Hartree–Fock theory or
Density Functional Theory (DFT) and correction for Basis Set
Superposition Error. Bader’s AIM theory (see above) has been
employed extensively in hydrogen bond studies (39).
Paradoxically, we have found that f,y conformations in car-
bohydrate crystal structures can be predicted by isolated models
with an elevated dielectric constant (e.g., 4.0–8.0) rather than the
prescribed value of 1.0 for CHARMM and AMBER calculations
or 1.5 for MM3 or MM4. Elevated dielectric constants reduce the
34 French and Johnson

strengths of interactions between charged atoms, including those

in hydrogen bonds. This work-around appears to provide a poten-
tial of mean force similar to what might otherwise have been
obtained by MD with explicit solvent. This approach is not appro-
priate for investigation of specific molecule–molecule interactions
such as might be found in modeling an entire crystal, for example.
It seems to work only for modeling condensed-phase systems
when the rest of the condensed phase is not explicitly present.
Less clear is the impact of C–H…O hydrogen bonds.
Geometric criteria identified 14 such bonds in the crystal struc-
ture of dicyclohexyl cellobioside (40). These fairly weak interac-
tions have typically been emphasized less when developing
empirical force fields.

3.3. Anomeric Effects The anomeric effect was the unexpected finding that the experi-
mental a:b ratio for compounds such as d-glucopyranose favored
the a-anomeric form more than would have been expected for an
axial substituent on a cyclohexane ring (6). The exoanomeric
effect was named for the preference of the substituent in methyl
glucopyranoside to take an orientation gauche to the ring oxygen.
This increased stabilization affects φO5 directly, favoring −60° and
+60° angles but not the trans, 180° angle. An external anomeric
torsional effect has also been proposed that affects y (41). The
term “general anomeric effect” covers the gauche preference for
any R–X–C–Y atom sequence in organic chemistry where X
denotes O, N, or S, and Y denotes any atom having lone pairs of
electrons. Thus, DMM, which has a C–O–C–O–C sequence, pre-
fers the gg conformation, while n-pentane prefers the all-trans
form. The analogy between small molecules and the C5–O5–C1–
O1–CMe sequence in methyl glucosides was noted many years ago
(42, 43).
Besides conformational preferences, anomeric effects cause
differences in bond lengths and angles. In the very accurate mul-
tipole refinement of crystalline sucrose at 20 K (44), the bonds
from the anomeric carbons to the ring oxygens are 1.4192 and
1.4146 Å for the pyranosyl and furanosyl rings, respectively,
whereas the distances between the ring oxygens and the other
carbon atoms (C5 and C5¢) are 1.4477 and 1.4543 Å, with stan-
dard deviations of 0.0005.
Often overlooked is the effect of an anomeric center, regard-
less of conformational details. It results in extra stability of the
compound, as discussed by Tvaroška and Bleha (45). A “bond
and group enthalpy increment scheme” can be used to calculate
heats of formation (46), either by QM or MM. The increment in
MM3 for O–C–O is 6.62 kcal, much larger than the corrections
for QM (0.505 kcal/mol for HF/6-31G* and −0.351 kcal/mol
for B3LYP/6-31G*). MM3 steric energies do not consider
­anomeric centers, so a large adjustment is needed. Fairly simple
 Computerized Molecular Modeling of Carbohydrates 35

QM does calculate most of this enthalpy, but some correction is

still needed.
Earlier, we used this method to calculate heats of formation
for analogs of some disaccharides (47). The analogs were based on
the native sugars but all exocyclic groups were replaced by hydro-
gen. We also joined two tetrahydropyran molecules with an ether
oxygen at the 4 positions (organic nomenclature), making iso-
meric pseudodisaccharide analogs with di-axial, axial-­equatorial,
and di-equatorial linkages. Analogs of a,a-, a,b-, and b,b-trehalose
had enthalpies of −149.9, −147.4, and −147.5 kcal/mol,
­respectively. The analogs of nigerose, laminarabiose, maltose,
­cellobiose, and galabiose gave −142.6, −141.7, −141.3, −140.6,
and −140.4 kcal/mol, and the di-axial, axial-equatorial, and di-
equatorial pseudosugar analogs had values of −135.1, −135.4,
and −135.8 kcal/mol. These were the B3LYP/6-31G* values,
but quite similar results were computed by MM3 and HF/6-
31G*. The trehalose analogs had two anomeric centers and
roughly 12 kcal of stabilization, and the analogs of the other
disaccharides had one center and 6 kcal, relative to the pseudo-
sugar analogs with no centers.
Anomeric effects are likely to have several different causes and
are affected by different factors. Magnitudes vary in different
solvents, suggesting that there is an electrostatic component, and
considerable effort has been directed to analyses of the changes in
electronic structure (6). One group suggested that gauche con-
formers for anomeric sequences are stabilized by C–H…O hydro-
gen bonds and carried out Natural Bond Order calculations to
confirm that result (48).

4. Quantum
Mechanics
Approaches
Empirical force fields (MM) summarize our knowledge of
­molecular structure and energy relationships for application to
other molecules. On the other hand, electronic structure theory
calculations (QM) are closer to experiment, with the possibility of
increasing the resolution by using more computer time and mem-
ory. There are two other important concepts: balance between
method and basis set, and DFT. Perdew et al. (49) has introduced
a “Jacob’s Ladder” to rate the various DFT methods, and Csonka
et al. (50) evaluated various DFT methods for applicability to
carbohydrates. The importance of choosing a QM method and
basis set cannot be exaggerated. Consider the calculations of
energy for four conformers of b-d-glucopyranose (5). Some levels
of theory favored the seldom-observed chair (1C4) by as much as
17 kcal/mol while others favored the dominant 4C1 form by that
same amount.
36 French and Johnson

The analogs described above in the studies of the anomeric

effect are much less demanding of a particular level of QM theory.
We have made relaxed f,y surfaces for the cellobiose analog at
HF/6-31G*, B3LYP/6-31G*, MP2/6-31+G**, and MP2/
­6-311+G** levels of theory, as well as B3LYP/6-311++G** calcu-
lations based on the B3LYP/6-31G* geometries. All of these
maps are very similar, with the B3LYP maps being slightly flatter
than the HF (51) and unpublished MP2 maps. Although it makes
little difference for these analogs, the diffuse function (indicated
by the +) is needed to avoid substantial overestimation of hydro-
gen bond energies for the native saccharides with the B3LYP and
MP2 methods and Pople basis sets.
Another lesson from those analogs was that QM maps are
fairly predictive of conformations of the native disaccharides in
crystals. The addition of methyl groups at C5 to complete the
carbon backbone improved predictions for cellobiose by reducing
the size of the 1 kcal/mol region (51). Figure 6 shows a map for
the methylated maltose analog with the crystalline maltose link-
ages (except cyclodextrins). (This energy map was shown earlier
with conformations from cyclodextrins (18).) Structures inside
the 1.0 kcal/mol contour are in accord with the exoanomeric
effect and do not have intramolecular hydrogen bonds in the
crystals. Their hydroxyl groups are either acylated and cannot
form hydrogen bonds or they form intermolecular linkages. All
other observed maltose-type structures have intramolecular,
interresidue hydrogen bonds between O2¢ and O3 that presum-
ably compensate for the higher energy of the analog “backbone.”
The diagonal line corresponding to the helical parameter value of
h = 0 is also shown. To its left are conformations that lead to left-
handed helices, with right-handed ones on the right. The second-
ary minimum at f = 80°, y = −280° is populated by some linkages
in larger cycloamyloses, and is also compatible with the exoano-
meric effect.
In at least one instance, the global minimum structure for the
fully hydroxylated native disaccharide is not in the region of the
crystal structures. That global minimum, for cellobiose, is stabi-
lized by an exceptional hydrogen bonding network (52). In an
important validation of that computational finding with QM, that
structure, and the global minimum for the closely related lactose
molecule have been observed experimentally in the gas phase at
very low temperature (53).
As might be expected because of the time required, there
are only a few molecular dynamics simulations based on QM
energies. The Car–Parinello method has been applied in a study
of the distortion energies of glucopyranose (54). The metady-
namics approach was applied to force the ring into different
puckering arrangements in a short time (44.4 ps). Our energy-
minimization approach to mapping the energy of cellobiose
 Computerized Molecular Modeling of Carbohydrates 37

against the linkage torsion angles was also computationally

expensive, with 181 different starting geometries, and only a
quarter of the entire f,y space was covered (55).

5. Empirical Force
Fields and Some
Applications
Empirical, or MM, force fields can be used with either energy
minimization methods or MD. A minute’s worth of MM time
can be equivalent to a month or more of QM time. Therefore,
many more issues can be evaluated with MM, making a more
complete study possible. Even if it is desired to ultimately study
the problem with QM, it can be worthwhile to study it first
with MM.
Development of force fields that are “carbohydrate aware”
continues. Besides the generally applicable MM3 and MM4 (56)
programs, the GLYCAM-06 (57) force field primarily for AMBER
is also intended for carbohydrates as well as proteins and lipids.
Refinements of GLYCAM continue with the addition of explicit
lone pairs on oxygen and nitrogen (58). The CHARMM (59)
program also has had carbohydrate force fields available, e.g., the
Ha et al. (60) parameterization and the CSFF parameters (61).
New parameterizations for CHARMM that include carbohydrates
are being released (62, 63). Some of the GROMOS parameter
sets also consider carbohydrates (64). Force fields that incorpo-
rate these parameterizations were recently compared, as well as
some older force fields (65). There are still substantial variations
among the various systems. One factor is that some force fields
have been parameterized so that aqueous solution data can be
reproduced with MD using explicit water molecules, especially
TIP3P (66).
DeMarco and Woods (67) reviewed several state-of-the-art
simulations and studies pertaining to carbohydrate–protein inter-
actions, both conformational and energetic. That paper then
advocates a new carbohydrate nomenclature that is more useful
for glycomics and integration with the Protein Data Bank proto-
cols and data formats.
Tvaroška has used a hybrid method for studying enzymatic
action on carbohydrates, in which the substrate is systematically
deformed in the protein. In that work, the active site atoms and
ligand are represented by QM and the rest of the protein and
water are represented by MM (68). Schramm’s group has a similar
approach (69).
In our nonintegral hybrid method, QM maps for the analog
furnish the conformational energy for the backbone of the disac-
charide, while the hydrogen bonding and other steric consider-
ations result from MM. The MM backbone map is subtracted
38 French and Johnson

from the MM disaccharide map and the QM map is added. This

approach is useful when the torsional energies of a particular link-
age are not well parameterized. Originally, the method was devel-
oped for studying sucrose (70), which has two adjacent anomeric
centers. More recently, similar hybrid studies of acarviosinde and
thiocellobiose compensated for incomplete parameters involving
the nitrogen and sulfur atoms, respectively (29).

6. Conclusions

It is not possible to completely understand the plant cell wall

without understanding the structures of the various carbohydrates
that compose it. Cellulose is, of course, the main component of
many cell walls, and other cell wall polysaccharides are closely
related to it in regard to their backbone structures. In this chapter,
we have of course cited some of our papers on cellulose, and we
have presented brief studies of the monomer and dimer of starch,
a-d-glucose, and maltose, as examples of different but chemically
similar materials. Without attempting to be intimidating, the exam-
ples are intended to illustrate the magnitude of the problems in
modeling carbohydrates as well as some ways to surmount the
problems. A checklist to be considered in a carbohydrate model-
ing study could be developed from the section headings above,
along with the specific issues that instigated the research in the
first place.
As early as the 1920s, ball-and-stick molecular modeling was
applied to help understand the structure of cellulose, and Jones
applied computerized modeling to cellulose in the late 1950s (see
Zugenmaier’s review (71)). The use of modeling to augment the
limited experimental data available from cellulose is well estab-
lished, but modeling has matured enough that it was able to make
fairly bold predictions for the lowest energy conformation of cel-
lobiose (52) that were subsequently confirmed by exotic experi-
ments (53).
Cell walls are of course much more complicated than the
small components that have been modeled so far. As computers
become even faster, larger assemblies of molecules can be studied,
along with physical and chemical processes. For example, current
studies include models of cellulose crystallites in water (72) and
dynamic conversions of such crystallites (73) and models com-
posed of up to a hundred thousand atoms are being studied.
Thus, the future of carbohydrate modeling will include the study
of previously undetermined structures, repeat studies of familiar
structures but with ever-improving methods, and ever-larger and
more complete representations of complex structures such as the
plant cell wall.
 Computerized Molecular Modeling of Carbohydrates 39

Numerous resources are available online, and at least two

should be mentioned. The Centre de Recherches sur les
Macromolécules Végétales (CERMAV) in Grenoble, France
maintains a site (http://www.cermav.cnrs.fr/glyco3d/index.php)
that has a library of disaccharide maps as well as a general tutorial
on carbohydrate modeling. Interactive views of various mono-
and oligosaccharides are also available. The GLYCAM (http://
www.glycam.com) site at the Complex Carbohydrate Research
Center at the University of Georgia in Athens provides several
interactive tools for the carbohydrate modeler, including a builder
for different structures.

References
1. Kurihara, Y. and Ueda, K. (2006) An investi- 10. Cremer, D. and Pople, J. A. (1975) A general
gation of the pyranose ring interconversion definition of ring puckering coordinates.
path of a-l-idose calculated using density J Am Chem Soc 97, 1354–1358.
functional theory. Carbohydr Res 341, 11. Altona, C. and Sundaralingam, M. (1972)
2565–2574. Conformational analysis of the sugar ring in
2. Steiner, T. and Saenger, W. (1998) Closure of nucleosides and nucleotides. A new descrip-
the cavity in permethylated cyclodextrins tion using the concept of pseudorotation.
through glucose inversion, flipping, and kink- J Am Chem Soc 94, 8205–8212.
ing. Angew Chem Int Ed 37, 3404–3407. 12. Haasnoot, C. A. G. (1992) The conformation
3. Añibarro, M., Gessler, K., Usón, I., Sheldrick, of six-membered rings described by puckering
G. M., Harata, K., Hirayama, K., Abe, Y. and coordinates derived from endocyclic torsion
Saenger, W. (2001) Effect of peracylation of angles. J Am Chem Soc 114, 882–887.
b-cyclodextrin on the molecular structure and 13. Zotov, A. Y., Palyulin, V. A. and Zefirov, N. S.
on the formation of inclusion complexes: an (1997) RICON – the computer program for
X-ray study. J Am Chem Soc 123, the quantitative investigations of cyclic organic
11854–11862. molecule conformations. J Chem Inf Comput
4. Gould, I. R., Bettley, H. A.-A. and Bryce, R. A. Sci 37, 766–773.
(2007) Correlated ab initio quantum chemical 14. Geremia, S., Vicentini, L. and Calligaris, M.
calculations of di- and trisaccharide conforma- (1998) Stereochemistry of ruthenium bis-
tions. J Comput Chem 28, 1965–1973. chelate disulfoxide complexes. A molecular
5. Barrows, S. E., Dulles, F. J., Cramer, C. J., mechanics investigation. Inorg Chem 37,
French, A. D. and Truhlar, D. G. (1995) Relative 4094–4103.
stability of alternative chair forms and hydroxy- 15. Bérces, A., Whitfield, D. M. and Nukada, T.
methyl conformations of b-glucopyranose. (2001) Quantitative description of six-­
Carbohydr Res 276, 219–251. membered ring conformations following the
6. Grindley, T. B. (2008) Structure and confor- IUPAC conformational nomenclature.
mation of carbohydrates. In: Fraser-Reid, B. Tetrahedron 57, 477–491.
O., Tatsuta, K. and Thiem, J. eds., 16. Joshi, N. V. and Rao, V. S. R. (1979) Flexibility
Glycosciences. Springer, Berlin, pp. 3–55. of the pyranose ring in a- and b-d-glucoses.
7. McNaught, A. D. (1996) Nomenclature of Biopolymers 18, 2993–3004.
carbohydrates (IUPAC Recommendations 17. Hill, D. and Reilly, P. J. (2007) Puckering coor-
1996). Pure Appl Chem 68, 1919–2008. dinates of monocyclic rings by triangular decom-
http://www.chem.qmul.ac.uk/iupac/ position. J Chem Inf Model 47, 1031–1035.
2carb/00n01.html#00. 18. French, A. D. and Johnson, G. P. (2007)
8. French, A. D. and Dowd, M. K. (1994) Analysis Linkage and pyranosyl ring twisting in cyclo-
of the ring-form tautomers of psicose with dextrins. Carbohydr Res 342, 1223–1237.
MM3 (92). J Comput Chem 15, 561–570. 19. Bader, R. F. W. (1990) Atoms in Molecules –
9. Boeyens, J.C.A. (1978) The conformation of six- A Quantum Theory. Oxford University Press,
membered rings. J Cryst Mol Struct 8, 317–320. Oxford.
40 French and Johnson

20. Csonka, G. I., Kolossváry, I., Császár, P., Éliás, K. length and temperature dependence. J Amer
and Csizmadia, I. G. (1997) The conforma- Chem Soc 131, 14786–14794.
tional space of selected aldo-pyrano-hexoses 33. Campen, R. K., Verde, A. V. and Kubicki, J. D.
J Mol Struct: THEOCHEM 395–396, 29–40. (2007) Influence of glycosidic linkage neigh-
21. Klein, R. A. (2002) Electron density topo- bors on disaccharide conformation in vacuum.
logical analysis of hydrogen bonding in glu- J Phys Chem B 111, 13775–13785.
copyranose and hydrated glucopyranose. J Am 34. Weiner, S. J., Kollman, P. A., Case, D. A.,
Chem Soc 124, 13931–19937. Singh, U. C., Ghio, C., Alagona, G., Profeta, S.,
22. Klein, R. A. (2006) Lack of intramolecular Jr. and Weiner, P. (1984) A new force field for
hydrogen bonding in glucopyranose: vicinal molecular mechanical simulation of nucleic acids
hydroxyl groups exhibit negative cooperativity. and proteins. J Am Chem Soc 106, 765–784.
Chem Phys Lett 433, 165–169. 35. Woods, R. J., Dwek, R. A., Edge, C. J. and
23. Koch, U. and Popelier, P. (1995) Fraser-Reid, B. (1995) Molecular mechanical
Characterization of C–H–O hydrogen bonds and molecular dynamic simulations of glyco-
on the basis of the charge density. J Phys Chem proteins and oligosaccharides. 1. GLYCAM_93
99, 9747–9754. parameter development. J Phys Chem 99,
24. Schrodinger, Portland, Oregon. http://www. 3832–3846.
schrodinger.com. 36. Ferretti, V., Bertolasi, V. and Gilli, G. (1984)
25. Çarçabal, P., Jockusch, R. A., Hunig, I., Structure of 6-kestose monohydrate, C18H31O16.
Snoek, L. C., Kroemer, R. T., Davis, B. G., H2O. Acta Crystallogr C 40, 531–535.
Gamblin, D. P., Compagnon, I., Oomens, J. 37. Jeffrey, G. A. (1997) Introduction to
and Simons, J. P. (2005) Hydrogen bonding Hydrogen Bonding. Oxford University Press,
and cooperativity in isolated and hydrated New York, p. 12.
sugars: mannose, galactose, glucose, and lac- 38. Parthasarathi, R., Elango, M., Subramanian, V.
tose. J Am Chem Soc 127, 11414–11425. and Sathyamurthy, N. (2009) Structure and
26. Allinger, N. L., Yuh, Y. H. and Lii, J.-H. stability of water chains (H2O)n, n = 5–20.
(1989) Molecular mechanics. The MM3 force J Phys Chem A 113, 3744–3749.
field for hydrocarbons. 1. J Am Chem Soc 39. Grabowski, S. J. (2006) Hydrogen Bonding –
111, 8551–8567. New Insights. Springer, Dordrecht, The
27. Allen, F. H. (2002) The Cambridge Structural Netherlands, 519pp.
Database: a quarter of a million crystal struc- 40. Yoneda, Y., Mereiter, K., Jaeger, C., Brecker, L.,
tures and rising. Acta Crystallogr B: Struct Sci Kosma, P., Rosenau, T. and French, A. (2008)
58, 380–388. van der Waals versus hydrogen-bonding forces
28. Naidoo, K. J. and Brady, J. W. (1997) The in a crystalline analog of cellotetraose:
application of simulated annealing to the con- ­cyclohexyl 4¢-o-cyclohexyl b-d-cellobioside
formational analysis of disaccharides. Chem cyclohexane solvate. J Am Chem Soc 130,
Phys 224, 263–273. 16678–16690.
29. Johnson, G. P., Petersen, L., French, A. D. 41. Lii, J.-H., Chen, K.-H., Johnson, G. P.,
and Reilly, P. J. (2009) Twisting of glycosidic French, A. D. and Allinger, N. L. (2005) The
bonds by hydrolases. Carbohydr Res 344, external-anomeric torsional effect. Carbohydr
2157–2166. Res 340, 853–862.
30. Mohamadi, F., Richards, N. G. J., Guida, W. C., 42. Jeffrey, G. A., Pople, J. A. and Radom, L.
Liskamp, R., Lipton, M., Caufield, C., Chang, G., (1972) The application of ab initio molecular
Hendrikson, T. and Still, W. C. (1990) orbital theory to the anomeric effect. A com-
Macromodel – an integrated software system parison of theoretical predictions and experi-
for modeling organic and bioorganic mole- mental data on conformations and bond
cules using molecular mechanics. J Comput lengths in some pyranoses and methyl pyrano-
Chem 11, 440–467. sides. Carbohydr Res 25, 117–131.
31. Kaminski, G. A., Friesner, R. A., Tirado-Rives, J. 43. Jeffrey, G. A., Pople, J. A. and Radom, L.
and Jorgensen, W. J. (2001) Evaluation and (1974) The application of ab initio molecular
reparametrization of the OPLS-AA force field orbital theory to structural moieties of carbo-
for proteins via comparison with accurate hydrates. Carbohydr Res 38, 81–95.
quantum chemical calculations on peptides. 44. Jaradat, D. M. M., Mebs, S., Chęcińska, L.
J Phys Chem B 105, 6474–6487. and Luger, P. (2007) Experimental charge
32. Shen, T., Langan, P., French, A. D., Johnson, density of sucrose at 20 K: bond topological,
G. P. and Gnanakaran, S. (2009) ­Conformational atomic, and intermolecular quantitative prop-
flexibility of soluble cellulose oligomers: chain erties. Carbohydr Res 342, 1480–1489.
 Computerized Molecular Modeling of Carbohydrates 41

45. Tvaroška, I. and Bleha, T. (1979) Lone pair Foley, B. L. and Woods, R. J. (2008)
interactions in dimethoxymethane and ano- GLYCAM06: a generalizable biomolecular
meric effect. Can J Chem 57, 424–435. force field. Carbohydrates. J Comput Chem
46. Allinger, N. L., Schmitz, L. R., Motoc, I., 29, 622–655.
Bender, C. and Labanowski, J. K. (1992) 58. Tschampel, S. M., Kennerty, M. R. and
Heats of formation of organic molecules. 2. Woods, R. J. (2007) TIP5P-consistent
The basis for calculations using either ab initio treatment of electrostatics for biomolecular
or molecular mechanics methods. Alcohols simulations. J Chem Theory Comput 3,
and ethers. J Am Chem Soc 114, 2880–2883. 1721–1733.
47. French, A. D., Kelterer, A.-M., Johnson, G. P. 59. Brooks, B. R., Bruccoleri, R. E., Olafson, B.
and Dowd, M. K. (2000) B3LYP/6-31G*, D., States, D. J., Swaminathan, S. and Karplus,
RHF/6-31G* and MM3 heats of formation M. (1983) CHARMM: a program for macro-
of disaccharide analogs. J Mol Struct 556, molecular energy, minimization, and dynam-
303–313. ics calculations. J Comput Chem 4, 187–217.
48. Takahashia, O., Yamasakia, K., Kohnob, Y., 60. Ha, S. N., Giammona, A., Field, M. and
Uedab, K., Suezawac, H. and Nishio, M. Brady, J. W. (1988) A revised potential-energy
(2009) The origin of the generalized ano- surface for molecular mechanics studies of
meric effect: possibility of CH/n and CH/p carbohydrates. Carbohydr Res 180, 207–221.
hydrogen bonds. Carbohydr Res 344, 61. Kuttel, M., Brady, J. W. and Naidoo, K. J.
1225–1229. (2002) Carbohydrate solution simulations:
49. Perdew, J. P., Ruzsinszky, A., Constantin, L. producing a force field with experimentally
A., Sun, J. and Csonka, G. I. (2009) Some consistent primary alcohol rotational frequen-
fundamental issues in ground-state density cies and populations. J Comput Chem 23,
functional theory: a guide for the perplexed. 1236–1243.
J Chem Theory Comput 5, 902–908. 62. Guvench, O., Greene, S. N., Kamath, G.,
50. Csonka, G. I., French, A. D., Johnson, G. P. Brady, J. W., Venable, R. M., Pastor, R. W.
and Stortz, C. A. (2009) Evaluation of den- and Mackerell, Jr., A. D. (2008) Additive
sity functionals and basis sets for carbohy- empirical force field for hexopyranose mono-
drates. J Chem Theory Comput 5, 679–692. saccharides. J Comput Chem 29, 2543–2564.
51. French, A. D. and Johnson, G. P. (2004) 63. Hatcher, E. R., Guvench, O. and
Advanced conformational energy surfaces for MacKerell, Jr., A. D. (2009) CHARMM addi-
cellobiose. Cellulose 11, 449–462. tive all-atom force field for acyclic polyalco-
52. Strati, G. L., Willett, J. L. and Momany, F. A. hols, acyclic carbohydrates, and inositol.
(2002) Ab initio computational study of J Chem Theory Comput 5, 1315–1327.
b-cellobiose conformers using B3LYP/6- 64. Oostenbrink, C., Soares, T. A., van der Vegt,
311++G**. Carbohydr Res 337, 1851–1859. N. F. A. and van Gusteren, W. F. (2005)
53. Cocinero, E. J., Gamblin, D. P., Davis, B. G. Validation of the 53A6 GROMOS force field.
and Simons, J. P. (2009) The building blocks Eur Biophys J 34, 273–284.
of cellulose: the intrinsic conformational 65. Stortz, C. A., Johnson, G. P., French, A. D.
structures of cellobiose, its epimer, lactose, and Csonka, G. I. (2009) Comparison of dif-
and their singly hydrated complexes. J Am ferent force fields for the study of disaccha-
Chem Soc 131, 11117–11123. rides. Carbohydr Res 344, 2217–2228.
54. Biarnés, X., Ardèvol, A., Planas, A., Rovira, 66. Jorgensen, W. L., Chandrasekhar, J.,
C., Laio, A. and Parrinello, M. (2007) The Madura, J. D., Impey, R. W. and Klein, M. L.
conformational free energy landscape of b-d- (1983) Comparison of simple potential func-
glucopyranose. Implications for substrate pre- tions for simulating liquid water. J Chem Phys
activation in b-glucoside hydrolases. J Am 79, 926–935.
Chem Soc 129, 10686–10693. 67. DeMarco, M. L. and Woods, R. J. (2008)
55. French, A. D. and Johnson, G. P. (2006) Structural glycobiology: a game of snakes and
Quantum mechanics studies of cellobiose ladders. Glycobiology 18, 426–440.
conformations. Can J Chem 84, 603–612. 68. Krupička, M. and Tvaroška, I. (2009) Hybrid
56. Lii, J.-H., Chen, K.-H. and Allinger, N. L. quantum mechanical/molecular mechanical
(2003) Alcohols, ethers, carbohydrates, and investigation of the b-1,4-galactosyltrans-
related compounds. IV. Carbohydrates. ferase-I mechanism. J Phys Chem B 113, (32),
J Comput Chem 24, 1504–1513. 11314–11319.
57. Kirschner, K. N., Yongye, A. B., Tschampel, 69. Zhang, Y., Luo, M. and Schramm, V. L. (2009)
S. M., González-Outeriño, J., Daniels, C. R., Transition states of Plasmodium falciparum
42 French and Johnson

and human orotate phosphoribosyltransferases. 72. Matthews, J. F., Skopec, C. E., Mason, P. E.,
J Am Chem Soc 131, 4685–4694. Zuccato, P., Torget, R. W., Sugiyama, J.,
70. French, A. D., Kelterer, A.-M., Cramer, C. J., Himmel, M. E. and Brady, J. W. (2006).
Johnson, G. P. and Dowd, M. K. (2000) A Computer simulation studies of microcrystal-
QM/MM analysis of the conformations of line cellulose Ib. Carbohydr Res 341,
crystalline sucrose moieties. Carbohydr Res 138–152.
326, 305–322. 73. Yui, T. and Hayashi, S. (2009) Structural sta-
71. Zugenmaier, P. (2008) Crystalline Cellulose bility of the solvated cellulose IIII crystal
and Derivatives. Characterization and models: a molecular dynamics study. Cellulose
Structrures. Springer, Berlin, pp. 8 and 38. 16, 151–165.
 Chapter 3

Oligosaccharide Mass Profiling (OLIMP) of Cell Wall

Polysaccharides by MALDI-TOF/MS
Markus Günl, Florian Kraemer, and Markus Pauly

Abstract
In today’s field of plant cell wall research, insights into the structure of wall components are obtained
using many different techniques, ranging from spectroscopic and microscopic to chemical and biochemi-
cal. In this chapter, we describe one method: oligosaccharide mass profiling (OLIMP). Using OLIMP,
we can harness the selective power of a specific wall hydrolase together with the speed and sensitivity of
mass spectrometry to provide highly reproducible structural and compositional information about the
wall molecule of interest.

Key words: Mass spectrometry, Matrix polysaccharides, Oligosaccharides, Glycosylhydrolases,

Xyloglucan, Xylan, Pectins

1. Introduction

Oligosaccharide mass profiling (OLIMP) is a useful, rapid, and

sensitive tool to reveal structural properties of cell wall polysac-
charides (1, 2). The sample preparation and analysis of wall poly-
saccharides by OLIMP is simple and quick, as shown in Fig. 1. It
comprises three steps: (1) the preparation of wall material from
plant tissues, (2) enzymatic release of specific oligosaccharides
from the wall materials, followed by (3) mass spectrometry on the
solubilised oligosaccharides. Based on the observed ions and the
known specificity of the enzymes used, specific structures can be
assigned to the ions. Hence, OLIMP is capable of giving valuable
insights into the diversity and substitution patterns of wall poly-
saccharides. The analysis of oligosaccharide composition with
matrix-assisted laser desorption/ionisation time-of-flight mass
spectrometry (MALDI-TOF/MS) is furthermore a method that
does not involve strong acid or base treatments, which allows a

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_3, © Springer Science+Business Media, LLC 2011

43
44 Günl, Kraemer, and Pauly

Etiolated Roots, All plant

Leaves
seedlings flower buds material

Alternative General protocol Laser dissection of Preparation of Golgi

protocol for AIR for preparation Arabidopsis tissues and vesicles and
preparation (3.2) of AIR (3.1) AIR preparation (3.4) extraction of AIR (3.3)

Enzymatic digest of cell

wall polysaccharides (3.5)

MALDI-TOF/MS
analysis (3.6)

Fig. 1. Flow chart for OLIMP procedure using MALDI-TOF/MS. The first step encompasses preparation of AIR from plant
material. AIR can be extracted from a variety of plant tissues, and the preparation method can be modified according to
the wall material to be analysed (see Subheadings 1–4). Oligosaccharides can be released from AIR with enzymes spe-
cific for various polysaccharides (see Subheading 5) and subsequently analysed by MALDI-TOF/MS (see Subheading 6).

comprehensive analysis of the polysaccharide structure including

O-acetylation and methylesterification substitution levels. OLIMP
is very sensitive, requiring walls of only 5,000 cells for a complete
analysis (3). In addition, a rapid analysis of the samples allows
high-throughput experiments as shown by Lerouxel et al. (1) and
Mouille et al. (4). However, structural isomers cannot be distin-
guished with OLIMP unless further structural information is
obtained by mass fragmentation methods such as post-source
decay (PSD). This procedure will generate several smaller frag-
ments of a single oligosaccharide, which are further analysed by
mass spectrometry (5, 6). This information can also be helpful for
the analysis of polysaccharides for which limited structural infor-
mation is available. Another limitation of OLIMP is its non-
quantitative nature. Although the relative abundance of the
oligosaccharide ions can be delineated from the spectra, the abso-
lute amount present in the sample cannot, because the enzyme
might not solubilise the polymer in its entirety from the wall (7),
or because of ion suppression of the mass spectrometer at high
concentrations or salt contaminants.
OLIMP can be used on a wide range of wall materials; it has
been successfully used for the analysis of various Arabidopsis tis-
sues such as leaves, roots, and flower buds. Because only very
small samples are required for OLIMP analysis, it is possible to
analyse specific cell types, cell compartments, or single etiolated
seedlings (3).
OLIMP can be used to analyse a variety of wall polymers,
limited only by the type of enzymes available (Fig. 2). A wide variety
of commercially available hydrolases are suitable for use in OLIMP.
 Oligosaccharide Mass Profiling (OLIMP) of Cell Wall Polysaccharides 45

XXXG
a xyloglucan

XXLG XXFG

XXFG
XXLG XLFG
XLFG

900 1100 1300 1500 1700

xylan
Arbitrary abundance

X 4A 1M
b X4A1G

X AM
X5A2G
X 5A 3G 5 2

X6A 2M
X6A3M
X5A1G

X6A2G
X5A3M

X6A3G

X7A4M
X4A2M

X6A4M
X4A2G

X6A4G

X7A 3M
X5A1M

X7A2G
X7A3G

X7A5M
X7A4G

X8A5M
X8A4M
X5A4M

X8A4G
X7A5G

X8A6M
600 850 1100 1350 pectin 1600
(GalA)4Me1Ace1

(GalA)5Me4Ace1
c
(GalA)4Me1
(GalA)4

700 800 900 1000

m/z

Fig. 2. OLIMP spectra of solubilised oligosaccharides (see Note 10). (a) Xyloglucan OLIMP spectrum from an Arabidopsis
hypocotyl using a xyloglucan-specific endoglucanase. Nomenclature of oligosaccharide structures is according to Fry
et al. (26), whereby the single letter indicates the reducing end sugar of the side-chain: e.g. G non-substituted glucosyl-
residue; X xylosyl; L galactosyl; F fucosyl-residue. Underlined structures indicate galactosyl 0-acetylation. (b) Xylan
OLIMP spectrum derived from rice leaves using a xylanase M6 (Megazyme). Nomenclature for putative structures from
Lee et al. (14). Xn number of pentoses, G glucuronic acid, and M methylated glucuronic acid. (c) Pectin OLIMP spectrum
from Arabidopsis leaf walls using endoPG and pectin methylesterase. (GalA)n number of galacturonic acid; Men number
of methylester; Acn number of o-acetyl substituents.

For example, for xyloglucan OLIMP, endo-b-(1–4)-glucanases

are commonly used (8–12); these enzymes can be specific for
xyloglucan (10–12). Similarly, for the analysis of xylan and pectin,
endo-b-(1–4)-xylanases (13, 14) and endo-polygalacturonases
(endoPG) (3, 15) are utilised, respectively. For the analysis of the
pectic polysaccharides, a pectin methylesterase (PME) is added to
the digestion to increase the susceptibility of pectin to hydrolysis
by endoPG (3, 15). A large number of other enzymes have been
recently cloned and are available to the public (16) with a poten-
tial to expand the wall components that can be assessed by
OLIMP. For instance, digestion of cell wall material with lichenase
leads to the release of mixed-linkage glucan-derived oligosaccha-
rides that can be analysed with MALDI-TOF/MS (17, 18).
Furthermore, with slightly altered experimental set-ups, glycosyl-
transferase and glycoside hydrolase activities can be assayed (19–
23). For this purpose, purified enzymes or plant extracts are
incubated with suitable substrates, and the product released is
analysed by mass spectrometry.
46 Günl, Kraemer, and Pauly

2. Materials

2.1. Preparation of 1. 2 mL Microcentrifuge tubes (Eppendorf, Hamburg, Germany)

Alcohol-Insoluble (see Note 1).
Residue (AIR) from 2. 70% (v/v) Aqueous ethanol.
Arabidopsis Plant
3. 1:1 (v/v) Chloroform:methanol.
Materials
4. Acetone, pure.

2.2. Alternative 1. 2 mL microcentrifuge tubes (Eppendorf, Hamburg, Germany)

Protocol for (see Note 1).
Preparation of AIR from 2. Methanol, pure.
Dark-Grown Seedlings
3. 1:1 (v/v) Chloroform:methanol.

2.3. Preparation of 1. Sucrose buffers: 0.1 M KH2PO4 buffer (pH 6.65), 10 mM

Golgi Vesicles and MgCl2, 8, 16, 33, 36, and 38% (w/v) sucrose.
Extraction of AIR 2. STM buffer: 0.25 M sucrose, 1 mM MgCl2, 10 mM Tris–HCl
(pH 8.0).
3. Ethanol, pure.
4. 80% (v/v) Aqueous ethanol.
5. 1:1 (v/v) Chloroform:methanol.

2.4. Laser Dissection 1. Ethanol dilution series for dehydration of leaf tissue: 20, 30,
and AIR Extraction 40, 50, 60, 70, 80, 90% (v/v) aqueous ethanol, and pure
from Arabidopsis Leaf ethanol.
Tissue 2. Pure xylene and 1:1 ethanol/xylene (v/v) solution.
3. Paraffin solutions for embedding: 30, 50, and 80% paraffin in
xylene (v/v) and pure paraffin.

2.5. Enzymatic Digest 1. Enzymes: Xyloglucan-specific endo-glucanase (XEG) (EC

of Cell Wall 3.2.1.151, (24)), PME (Novozymes, Bagsvaerd, Denmark)
Polysaccharides endo-polygalacturonase M2 (endoPG), and xylanase M6
(Megazyme, Bray, Ireland).
2. 1 M ammonium formate stock solution (pH 4.5).
3. 4 M sodium hydroxide.
4. 1 M hydrochloric acid.
5. Float-A-Lyzer G2 dialysis devices (MWCO 8–10 kDa;
Spectrum Labs, Greensboro, NC, USA).
6. 1 M sodium acetate stock solution (pH 6.0).

2.6. MALDI-TOF MS 1. 2,5-dihydroxybenzoic acid 10 mg/mL in water (see Note 2).

Analysis 2. Bio-Rad MSZ-501 (D) cation exchange resin beads (see Note 3).
 Oligosaccharide Mass Profiling (OLIMP) of Cell Wall Polysaccharides 47

3. Methods

3.1. General Procedure 1. 0.1–10 mg plant material is placed in a 2 mL microcentrifuge

for Preparation of tube and snap frozen in liquid nitrogen (see Note 4).
Alcohol-Insoluble 2. The plant tissue is ground using a Retsch ball mixer mill
Residue (AIR) from MM301 for 1 min at 25 Hz with a single steel ball (3 mm).
Arabidopsis Plant
3. After the tissue has been ground, the steel ball is removed
Material from the tube via a magnet.
4. 1 mL 70% Ethanol is added to the ground material; the
microcentrifuge tube is vortexed briefly and centrifuged for
10 min at 20,000 × g in a tabletop centrifuge.
5. The supernatant is removed and the pellet is washed with
1 mL chloroform:methanol and vortexed briefly; the plant
tissue is spun down for 10 min at 20,000 × g in a tabletop
centrifuge.
6. The supernatant is discarded and the pellet dried under a
stream of air. The dried tissue can be directly used for enzyme
digestion (see Subheading 3.5; Note 5).

3.2. Alternative 1. Approximately 10 seedlings (3–7 days old) are placed in a

Protocol for 2 mL microcentrifuge tube and 1 mL of methanol is added
Preparation of AIR (see Note 4).
from Dark-Grown 2. The tissue is homogenised using a Retsch ball mixer mill
Arabidopsis Seedlings MM301 for 1 min at 25 Hz with a 5-mm stainless steel ball.
3. After the tissue has been homogenised, the metal ball is
removed from the tube using a magnet.
4. The macerated material is centrifuged for 15 min at 20,000 × g
and the supernatant is discarded.
5. The pellet is resuspended in 1 mL chloroform:methanol and
centrifuged for 15 min at 20,000 × g.
6. The supernatant is discarded and the pellet air dried.
Alternatively, the material can be dried under vacuum.
7. The dried tissue is ready for enzyme digestion (see
Subheading 3.5).

3.3. Preparation of The method described here is based on Muñoz et al. (25).
Golgi Vesicles and Ultracentrifugation was performed using a Beckman Coulter SW
Extraction of AIR 32 Ti rotor and 38.5 mL thin-wall rotor tubes. To prevent degra-
dation, all preparation steps are performed in the cold room or on
ice.
1. 5–10 mg of fresh plant material is transferred into a cold Petri
dish and the material is finely chopped (1–2 mm sections)
with a razor blade.
48 Günl, Kraemer, and Pauly

2. The chopped material is added to a cold mortar and 8 mL of

cold 16% sucrose buffer is added.
3. The material is homogenised for 3 min with a pestle by rota-
tion, applying only light pressure.
4. The suspension is filtered through a nylon mesh (diameter
30 mm) into a 50 mL Falcon tube, and the filtrate is centri-
fuged for 10 min at 2,000 × g and 4°C.
5. The ultracentrifugation tube is prepared by adding 8 mL of
cold 38% sucrose buffer.
6. Add the supernatant of step 4 on top of the sucrose layer, try-
ing not to disturb the sucrose solution. This can be accom-
plished by transferring the supernatant with a Pasteur pipette
onto the side of the centrifuge tube walls. The sample is cen-
trifuged for 100 min at 100,000 × g. The microsomal fraction
will form as a milky layer on top of the 38% sucrose cushion
with the microsome free cell extract layer above.
7. The microsome free cell extract (top layer) is removed with a
Pasteur pipette.
8. 8 mL of cold 36% sucrose buffer is carefully added on top of
the microsomal layer. Be careful not to disturb the layer.
Then, 8 mL of cold 33% sucrose buffer is added on top of this
layer, with care not to disturb the layer.
9. The tube is filled up to 3.5 mL with cold 8% sucrose buffer
and centrifuged for 90 min at 100,000 × g. During this cen-
trifugation step, the microsomal fraction (on top of the 36%
sucrose layer) will separate and float on top, resulting in an
enriched ER and Golgi fraction.
10. The fraction above the 33% cushion is the Golgi-enriched
fraction. This fraction is transferred with a Pasteur pipette
into a new ultracentrifuge tube and the fraction is diluted 1:2
with water.
11. The diluted fraction is centrifuged for 60 min at
100,000 × g.
12. The supernatant is discarded, and the pellet is washed twice
with STM buffer to remove inorganic phosphate.
13. The pellet is resuspended in 1 mL STM buffer and stored at
−80°C until needed.
14. To prepare AIR for OLIMP using MALDI-TOF/MS analy-
sis, a portion of the solution containing 100 mg protein
equivalent is used. The protein content can be determined
(e.g. by Bradford assay).
15. Pure ethanol is added to the sample to reach a final concen-
tration of 80% (v/v), and then the sample is filled up with
80% ethanol to a final volume of 1 mL.
 Oligosaccharide Mass Profiling (OLIMP) of Cell Wall Polysaccharides 49

16. The sample is centrifuged for 10 min at 20,000 × g and the

supernatant is discarded.
17. 1 mL of chloroform:methanol is added to the pellet and the
pellet is vortexed gently.
18. The sample is centrifuged for 10 min at 20,000 × g, the super-
natant is discarded and the sample is air dried for 30 min.
19. The sample is now ready for enzyme digestion (see Sub­
heading 3.5).

3.4. Laser Dissection 1. A leaf from a 5-week-old plant is harvested, directly trans-
of Arabidopsis Tissues ferred to a 2 mL microcentrifuge tube containing 1 mL 20%
and Their AIR ethanol, and incubated at room temperature.
Preparation 2. After 6 h, the 20% ethanol is removed and the procedure is
repeated with increasing ethanol concentrations in 10% incre-
mentals, each 6 h long, until incubation with pure ethanol is
achieved. Incubation with pure ethanol is carried out twice.
The pure ethanol is removed and ethanol:xylene (1:1, v/v) is
added. The ethanol:xylene incubation is followed by three
incubations with pure xylene. All incubations are carried out
for at least 6 h with 1 mL of solvent.
3. After removal of pure xylene, the leaf tissue is incubated in
1 mL 30% paraffin xylene solution for 8 h at 42°C, followed by
sequential incubation in 50 and 80% paraffin for 8 h at 52 and
58°C, respectively. Then, the tissue is transferred to pure paraf-
fin and incubated for 7 h at 64°C. This step is repeated once.
4. The tissue is then placed in aluminium tray (3 mL) containing
melted paraffin. After the paraffin solidifies at room tempera-
ture, the leaf is excised and mounted on a cutting block.
5. Sections of 20–40 mm are made and placed on glass micro-
scope slides. The paraffin is removed by adding xylene drop-
wise onto the tissue, and soaking up the solubilised paraffin
with a paper tissue from the side.
6. Tissues of interest are dissected using a laser dissector, accord-
ing to the guidelines of the manufacturer. To analyse oligo-
saccharide composition, the equivalent of approximately
5,000 cells must be collected.
7. The collected cells (fragments) are transferred to a 0.5 mL
microcentrifuge tube, washed with 200 mL xylene, and spun
down at 20,000 × g in a tabletop centrifuge. After decanting
the supernatant, 200 mL methanol:chloroform is added and
the sample is briefly vortexed and centrifuged at 20,000 × g in
a tabletop centrifuge.
8. Discard the supernatant and air dry the pellet. The dried cell
fragments can be directly used for enzyme digestion (see
Subheading 3.5; Note 5).
50 Günl, Kraemer, and Pauly

3.5. Enzymatic Digest 1. Depending on the polysaccharide to be investigated, the

of Cell Wall following digests can be carried out.
Polysaccharides (a) Xyloglucan-specific endoglucanase (XEG) digest for analy-
sis of the cross-linking glycan xyloglucan: 50 mL of 50 mM
ammonium formate (made with 2.5 mL ammonium for-
mate stock solution) containing 0.2 U XEG (1 U of XEG
releases 1 mmol xyloglucan oligosaccharides per min) are
added to previously prepared AIR from plant material
(see Subheadings 3.1–3.4) and incubated overnight
(16 h) at 37°C and shaking at 120 rpm.
(b) Pectin digest with endo-polygalacturonase M2 (endoPG):
50 mL of 100 mM ammonium formate (prepared with
5 mL ammonium formate stock solution) containing
0.15 U endoPG (1 U of endoPG releases 1 mmol polyga-
lacturonic acid oligosaccharides per min) and 0.08 U PME
(1 U of PME releases 1 mmol methanol per min) are added
to previously prepared AIR from plant material and incu-
bated overnight (16 h) at 37°C, shaking at 120 rpm.
(c) Xylan digest: 200 mL of 4 M sodium hydroxide is added
to the previously prepared AIR and incubated for 1 h at
37°C under shaking (500 rpm). The samples are neutra-
lised by adding 800 mL of 1 M hydrochloric acid and
spun at 20,000 × g for 10 min in a tabletop centrifuge.
The supernatant is transferred into a Float-A-Lyzer G2
dialysis device (MWCO 8–10 kDa), and dialysis is per-
formed against ultrapure water with at least one change
of water. After dialysis, the sample is transferred into a
microcentrifuge tube and dried down in a speed vac. The
dried material is digested overnight (16 h) in 200 mL
50 mM sodium acetate (prepared with 10 mL sodium
acetate stock solution) containing 8 U xylanase M6 (1 U
of xylanase releases 1 mmol arabinoxylan oligosaccha-
rides) per min at 40°C and 120 rpm.
2. The digest is spun down for 10 min at 20,000 × g in a table-
top centrifuge, and the supernatant containing the released
oligosaccharide fragments is transferred to a new microcentri-
fuge tube (see Note 6).

3.6. MALDI-TOF/MS The analysis can be performed on a Kratos AXIMA CFR MALDI-
Analysis TOF/MS instrument using a stainless steel MALDI target type
DE1580TA (Kratos).
1. Prepare the MALDI-TOF/MS sample target by adding a
layer of matrix. Use 2 mL of 2,5-dihydroxybenzoic acid per
sample, spot and dry the matrix under vacuum (see Note 7).
2. Transfer 10 mL digest into a fresh 1.5 mL or 0.5 mL micro-
centrifuge tube.
 Oligosaccharide Mass Profiling (OLIMP) of Cell Wall Polysaccharides 51

3. Add approximately 5–10 cation exchange beads to each sample

and incubate at room temperature for 15 min. (see Notes 3
and 8).
4. Transfer the desalted remaining liquid into a new microcen-
trifuge tube.
5. Spot as many samples as possible within a 3 min time frame
(2 mL of the desalted oligosaccharide sample solutions) onto
the target plate mesas containing the dried matrix spot (step 1).
This time limit ensures that the first spot is not fully air-dried.
Wait another 2 min and dry the target under vacuum. Repeat
these steps until all your samples are spotted (see Note 9).
6. The mono- and oligosaccharides are detected by the MS as
their sodium [M+Na]+ and to a lesser degree potassium
[M+K]+ adducts. Therefore, the molecular mass detected will
be 23 or 39 m/z larger than that of your analyte.

4. Notes

1. Testing different microcentrifuge brands and sizes showed

that the Eppendorf 2 mL tubes are best suited to withstand
milling with the ball mixer at very low temperatures. Of
course, once in a while plastic pieces break or crack.
2. You can prepare a stock solution of the matrix chemical and
store aliquots at −20°C for as long as 6 months. Thaw the
aliquot 30 min before use and vortex vigorously several times.
Make sure that the matrix is completely dissolved. Other sol-
vents than water can be used, such as acetone, acetonitrile,
methanol, or chloroform. However, in our experience water
is the most effective solvent to achieve good spectra of
oligosaccharides.
3. The MSZ501 (D) resin comes as a mixture of anion and cat-
ion resin beads, and the two forms can be distinguished visu-
ally. We use only the cation exchange resin beads. The anion
exchange beads are lighter than the cation beads and have a
blue indicator dye irreversibly bound, whereas the cation
exchange beads are of a golden brown colour. To prepare the
cation exchange resin you have to separate the brown and
blue beads. For this purpose, transfer several spoonfuls of
Bio-Rad MSZ-501 (D) resin beads into a 50 mL Falcon tube
and fill with water. Swirl the tube a few times and decant the
lighter, blue cation beads into a second 50 mL Falcon tube.
Repeat until you have reached separation. The beads can be
stored at room temperature for 6 months in water. However,
before use, wash the beads several times with fresh water.
52 Günl, Kraemer, and Pauly

4. In our experience, a wide range of plant material and tissues

can be used, but there are some limitations. Tissues that work
very well are 4-day-old dark-grown seedlings and flower buds,
and leaves of 2–3-week-old plants. In many cases, weighing
of material may not be necessary, and we have found that the
following amounts work well: 1–20 four-day-old dark-grown
seedlings, flower buds of 1 inflorescence, 1 leaf of a 2–3-week-
old plant.
5. To dry samples more quickly, an additional washing step with
1 mL acetone is recommended.
6. For samples that contain very small amounts of material, it is
advisable to dry the supernatant containing the solubilised
oligosaccharides using a vacuum centrifuge followed by
re-dissolving the pellet in 10 mL water to increase the oligo-
saccharide concentration.
7. Prepare about 20 sample spots at a time with the matrix.
Preparation of too many sample spots at a time can lead to
longer drying times under vacuum, which can lead to poor
quality of the matrix crystals. Prepare more sample spots than
you have samples to provide some leeway in case some sam-
ples mix and you have to re-spot.
8. To add the ion exchange beads, use a small spoon-shaped
spatula to hold a small amount of beads. With a second
tapered spatula, move a small quantity of beads over the edge
of the spoon. The beads will stick to the tapered spatula. Tap
the tapered spatula on the rim of the microcentrifuge tube to
transfer the beads into the tube.
9. As a quality control and to simplify troubleshooting, it is
advisable to include a sample that is known to produce good
spectra on each target that you prepare. If possible, use a pre-
vious sample derived from plant material rather than a pure
standard.
10. Ionised oligosaccharides will be analysed as their sodium or
potassium adducts; therefore, you should not expect to find
masses that exactly match the predicted oligosaccharide
masses, but with an additional +23 or +39 m/z.

Acknowledgements

The authors would like to thank Kirk Schorr (Novozymes) for

providing the enzymes XEG and PME. Karen Bird, DOE-Plant
Research Lab, Michigan State University, is thanked for editing
the text.
 Oligosaccharide Mass Profiling (OLIMP) of Cell Wall Polysaccharides 53

References

1. Lerouxel, O., Choo, T. S., Seveno, M., (2008) Disrupting two Arabidopsis thaliana
Usadel, B., Faye, L., Lerouge, P., and Pauly, xylosyltransferase genes results in plants defi-
M. (2002) Rapid structural phenotyping of cient in xyloglucan, a major primary cell wall
plant cell wall mutants by enzymatic oligosac- component. Plant Cell 20, 1519–1537.
charide fingerprinting. Plant Physiology 130, 11. Hilz, H., de Jong, L. E., Kabel, M. A., Schols,
1754–1763. H. A., and Voragen, A. G. (2006) A compari-
2. Günl, M., Gille, S., and Pauly, M. (2010) son of liquid chromatography, capillary elec-
OLIigo Mass Profiling (OLIMP) of extracel- trophoresis, and mass spectrometry methods
lular polysaccharides. Journal of Visualized to determine xyloglucan structures in black
Experiments. http://jove.com/index/details. currants. Journal of Chromatography A 1133,
stp?id=2046:doi: 10.3791/2046. 275–286.
3. Obel, N., Erben, V., Schwarz, T., Kühnel, S., 12. Pauly, M., Eberhard, S., Albersheim, P.,
Fodor, A., and Pauly, M. (2009) Microanalysis Darvill, A., and York, W. S. (2001) Effects of
of plant cell wall polysaccharides. Molecular the mur1 mutation on xyloglucans produced
Plant 2, 922–923. by suspension-cultured Arabidopsis thaliana
4. Mouille, G., Witucka-Wall, H., Bruyant, cells. Planta 214, 67–74.
M. P., Loudet, O., Pelletier, S., Rihouey, 13. Brown, D. M., Goubet, F., Wong, V. W.,
C., Lerouxel, O., Lerouge, P., Höfte, Goodacre, R., Stephens, E., Dupree, P., and
H., and Pauly, M. (2006) Quantitative trait Turner, S. R. (2007) Comparison of five xylan
loci analysis of primary cell wall composi- synthesis mutants reveals new insight into the
tion in Arabidopsis. Plant Physiology 141, mechanisms of xylan synthesis. Plant Journal
1035–1044. 52, 1154–1168.
5. Ray, B., Loutelier-Bourhis, C., Lange, C., 14. Lee, C., Zhong, R., Richardson, E A.,
Condamine, E., Driouich, A., and Lerouge, P. Himmelsbach, D.S., McPhail, B.T., and Ye,
(2004) Structural investigation of hemicellu- Z. -H. (2007) The PARVUS gene is expressed
losic polysaccharides from Argania spinosa: in cells undergoing secondary wall thickening
characterisation of a novel xyloglucan motif. and is essential for glucuronoxylan biosynthe-
Carbohydrate Research 339, 201–208. sis. Plant and Cell Physiology 48, 1659–1672.
6. Yamagaki, T., Mitsuishi, Y., and Nakanishi, H. 15. Egelund, J., Obel, N., Ulvskov, P., Geshi, N.,
(1997) Structural analyses of xyloglucan hep- Pauly, M., Bacic, A., and Petersen, B. L. (2007)
tasaccharide by the post-source decay frag- Molecular characterization of two Arabidopsis
ment method using matrix-assisted laser thaliana glycosyltransferase mutants, rra1 and
desorption/ionization time-of-flight mass rra2, which have a reduced residual arabinose
spectrometry. Bioscience Biotechnology and content in a polymer tightly associated with
Biochemistry 61, 1411–1414. the cellulosic wall residue. Plant Molecular
7. Pauly, M., Albersheim, P., Darvill, A., and Biology 64, 439–451.
York, W. S. (1999) Molecular domains of the 16. Bauer, S., Vasu, P., Persson, S., Mort, A. J., and
cellulose/xyloglucan network in the cell walls Somerville, C. R. (2006) Development and
of higher plants. The Plant Journal 20, application of a suite of polysaccharide-degrading
629–639. enzymes for analyzing plant cell walls. Proceedings
8. Aboughe-Angone, S., Nguerna-Ona, E., of National Academy of Sciences of the United
Ghosh, P., Lerouge, P., Ishii, T., Rayb, B., and States of America 103, 11417–11422.
Driouich, A. (2008) Cell wall carbohydrates 17. Sørensen, I., Pettolino, F. A., Wilson, S. M.,
from fruit pulp of Argania spinosa: structural Doblin, M. S., Johansen, B., Bacic, A., and
analysis of pectin and xyloglucan polysaccha- Willats, W. G. T. (2008) Mixed-linkage (1 ® 3),
rides. Carbohydrate Research 343, 67–72. (1 ® 4)-beta-D-glucan is not unique to the
9. Vanzin, G. F., Madson, M., Carpita, N. C., Poales and is an abundant component of
Raikhel, N. V., Keegstra, K., and Reiter, W. Equisetum arvense cell walls. Plant Journal
D. (2002) The mur2 mutant of Arabidopsis 54, 510–521.
thaliana lacks fucosylated xyloglucan because 18. Doblin, M. S., Pettolino, F. A., Wilson, S. M.,
of a lesion in fucosyltransferase AtFUT1. Campbell, R., Burton, R. A., Fincher, G. B.,
Proceedings of National Academy of Sciences of Newbigin, E., and Bacic, A. (2009) A barley
the United States of America 99, 3340–3345. cellulose synthase-like CSLH gene mediates
10. Cavalier, D. M., Lerouxel, O., Neumetzler, (1,3; 1,4)-beta-D-glucan synthesis in trans-
L., Yamauchi, K., Reinecke, A., Freshour, G., genic Arabidopsis. Proceedings of National
Zabotina, O. A., Hahn, M. G., Burgert, I., Academy of Sciences of the United States of
Pauly, M., Raikhel, N. V., and Keegstra, K. America 106, 5996–6001.
54 Günl, Kraemer, and Pauly

19. Cavalier, D. M., and Keegstra, K. (2006) Two functional assessment of polysaccharide-active
xyloglucan xylosyltransferases catalyze the enzymes using MALDI-TOF mass spectrom-
addition of multiple xylosyl residues to cello- etry as exemplified on plant cell wall polysac-
hexaose. Journal of Biological Chemistry 281, charides. Analytical Biochemistry 373, 9–17.
34197–34207. 24. Pauly, M., Andersen, L. N., Kauppinen, S.,
20. Leonard, R., Pabst, M., Bondili, J. S., Kofod, L. V., York, W. S., Albersheim, P., and
Chambat, G., Veit, C., Strasser, R., and Darvill, A. (1999) A xyloglucan-specific endo-
Altmann, F. (2008) Identification of an beta-1,4-glucanase from Aspergillus aculeatus:
Arabidopsis gene encoding a GH95 alpha1, expression cloning in yeast, purification and
2-fucosidase active on xyloglucan oligo- characterization of the recombinant enzyme.
and polysaccharides. Phytochemistry 69, Glycobiology 9, 93–100.
1983–1988. 25. Muñoz, P., Norambuena, L., and Orellana, A.
21. Lee, C. H., O’Neill, M. A., Tsumuraya, Y., (1996) Evidence for a UDP-glucose trans-
Darvill, A. G., and Ye, Z. -H. (2007) The porter in golgi apparatus-derived vesicles from
irregular xylem9 mutant is deficient in xylan pea and its possible role in polysaccharide bio-
xylosyltransferase activity. Plant and Cell synthesis. Plant Physiology 112, 1585–1594.
Physiology 48, 1624–1634. 26. Fry, S. C., York, W. S., Albersheim, P., Darvill,
22. Iglesias, N., Abelenda, J. A., Rodino, M., A., Hayashi, T., Joseleau, J. P., Kato, Y., Pãrez
Sampedro, J., Revilla, G., and Zarra, I. (2006) Lorences, E., Maclachlan, G. A., McNeil, M.,
Apoplastic glycosidases active against xyloglu- Mort, A. J., Grant Reid, J. S., Seitz, H. U.,
can oligosaccharides of Arabidopsis thaliana. Selvendran, R. R., Voragen, A.G. J., and
Plant and Cell Physiology 47, 55–63. White, A. R. (1993) An unambiguous nomen-
23. Leboeuf, E., Immerzeel, P., Gibon, Y., Steup, clature for xyloglucan-derived oligosaccha-
M., and Pauly, M. (2008) High throughput rides. Physiologia Plantarum 89, 1–3.
 Chapter 4

High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall

Building Blocks and Their Metabolic Precursors
Stephen C. Fry

Abstract
HVPE is an excellent and often overlooked method for obtaining objective and meaningful information
about cell-wall “building blocks” and their metabolic precursors. It provides not only a means of analysis
of known compounds but also an insight into the charge and/or mass of any unfamiliar compounds that
may be encountered. It can be used preparatively or analytically. It can achieve either “class separations”
(e.g. delivering all hexose monophosphates into a single pool) or the resolution of different compounds
within a given class (e.g. ADP-Glc from UDP-Glc; or GlcA from GalA).
All information from HVPE about charge and mass can be obtained on minute traces of analytes,
especially those that have been radiolabelled, e.g. by in-vivo feeding of a 3H- or 14C-labelled precursor.
HVPE does not usually damage the substance under investigation (unless staining is used), so samples of
interest can be eluted intact from the paper ready for further analysis. Although HVPE is a technique that
has been available for several decades, recently it has tended to be sidelined, possibly because the appara-
tus is not widely available. Interested scientists are invited to contact the author about the possibility of
accessing the Edinburgh apparatus.

Key words: Charge:mass ratio, Electrophoresis, Hydroxyproline oligoarabinosides, Ionisation,

Monosaccharides, Nucleotide-sugars, Oligosaccharides, Radiolabelling, Sugar-phosphates, Uronic
acids

1. Introduction

High-voltage paper electrophoresis (HVPE) is a seriously under-

valued method for the analysis of small (<2,500 Da), hydrophilic,
charged compounds such as uronic acids, amino sugars, sugar
phosphates, sugar sulphates, sugar nucleotides, ascorbate metab-
olites and Krebs-cycle intermediates. In addition, it can be useful
for normally uncharged compounds, such as neutral sugars, that
can be given a charge by complexing with ions such as borate or

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_4, © Springer Science+Business Media, LLC 2011

55
56 Fry

molybdate. HVPE is a very rapid separation method, typical

run-times being 30–60 min with 10–20 samples typically being
run simultaneously.
The electrophoretic mobilities of compounds on HVPE are
not simply random values that need to be determined empirically.
On the contrary, a compound’s electrophoretic mobility is gov-
erned by a definable property, related to its charge:mass ratio, in
a highly predictable way, discussed below (1).
Ion-exchange chromatography is in some ways comparable
to HVPE and is an alternative method for analysis of charged
compounds. For example, an anion-exchange matrix will bind
anions, which can subsequently be released from the chromatog-
raphy column by a gradient of increasing ionic strength or changing
pH. This provides some evidence that the compound thus
obtained had a negative charge. However, there is the potential
pitfall that a compound might adsorb to an anion-exchange resin
by some means other than electrostatic bonding, such as hydro-
phobic interaction. In contrast, a compound can migrate towards
the anode on HVPE (relative to a neutral marker) only if it pos-
sesses a negative charge.
HVPE can be used analytically (e.g. 20 small samples on thin
paper) or preparatively (a single large sample “streak-loaded” on
a sheet of thick paper). The separated compounds can be detected
either by staining (which is usually destructive) or by fluorescence,
autoradiography (for 14C, 35S or 32P) or fluorography (for 3H),
after which the sample can usually be recovered for further work.
Alternatively, the compounds can be eluted from a preparative
paper ready for further analysis, e.g. by MS, NMR or bioassay.
Most of the recommended electrophoresis buffers are volatile, so
the eluted compounds simply need to be dried, not specially de-
salted, prior to further analysis. Another advantage of HVPE is
that – like paper chromatography – it provides a very convenient
means of archiving samples that have been separated but not yet
analysed: what you are storing is a dried paper with zones of
potentially interesting compounds on it; this is much more con-
venient than column chromatography separations in which liquid
fractions need to be stored.

2. Materials

2.1. HVPE Apparatus The basic principle of HVPE, a form of zone electrophoresis, is
that the sample is dried on a sheet of paper, which is then wetted
with an aqueous buffer and subjected to a voltage gradient. The
compounds in the sample migrate as zones towards the anode or
cathode according to their net charge. The higher the voltage
applied, the faster the compounds migrate; fast migration minimises
High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 57

+
lid
× • × • × • steel cool -
ing coils

glass rod
platinum cathode buffer
cathode
glass trough

paper (length
57 cm) wetted
with buffer

glass tank

toluene or
white spirit
(coolant)
platinum
foil anode buffer
(anode)

Fig. 1. Apparatus for high-voltage paper electrophoresis using a liquid coolant such as
white spirit or toluene. This type of apparatus is available at the University of Edinburgh
but not widely elsewhere; interested scientists are invited to contact the author about
the possibility of accessing the Edinburgh apparatus. The tank is shown in cross-section.
Filled circle = glass rod, plus = anode, minus = cathode, U8U8U8 = cooling coil
(suspended from the lid). Cold tap-water flows through the coil to cool the white spirit or
toluene. The anode and cathode are platinum wires encased in glass tubes and con-
nected at the tips to platinum foil.

diffusion of the spots. However, an excessively high voltage would

overheat the wet paper and possibly degrade the compounds of
interest. A cooling system must therefore also be used.
1. HVPE equipment: apparatus set up as shown in Fig. 1 (see
Note 1).
2. 42 × 57-cm Sheets of filter paper (see Note 2): two types of
paper are commonly used: Whatman No. 1 and the thicker
Whatman 3CHR (formerly called 3MM) (see Notes 3 and 4).
3. Coolant for wet paper: white spirit, toluene, or 20:1 (v/v)
toluene:pyridine. Cooling of the wet paper is best provided
by a large volume of a water-immiscible liquid (see Notes 5
and 6).
4. Glass trough: capable of containing ~350 mL.
58 Fry

5. Glass bar for holding the wet paper in the glass trough.
6. Aqueous running buffer: ~350 mL in the glass trough at the
cathode end and 1–2 L of identical running buffer at the
bottom of the tank and into which the platinum anode dips
(see Subheading 2.3).
7. Platinum cathode.
8. Platinum anode.
9. Steel cooling coils in the top 1–2 cm of the coolant to keep its
temperature below about 30°C.
10. Lid.

2.2. HVPE Apparatus: If the use of water-immiscible coolant liquids is not feasible, e.g.
Flat-Bed System because hydrophobic compounds are of interest, HVPE can also be
performed on a flat-bed system. The paper is laid on a polythene-
insulated metal plate (containing cooling coils) with the ends of the
paper dipping into troughs containing buffer and electrodes. An
insulated and padded lid is tightly clamped on top of the paper to
maximise uniform contact with the cooling plate (see Note 7).
1. 30 × 57-cm sheets of filter paper.
2. Polythene-insulated metal plate (containing cooling coils).
3. Insulated and padded lid to maximise uniform contact with
the cooling plate.
4. Troughs containing buffer (see Subheading 2.3).

2.3. Recommended Whenever possible, volatile buffers are used so that after the run
Buffers and Coolants the electrophoretograms can be freed of buffer salts simply by
drying. The three volatile buffers routinely used in our laboratory
are listed in Table 1. The buffer concentrations recommended are
a compromise between the need to provide adequate buffering
capacity (the higher the better) and the need to avoid excessive
heating during the run (a concentrated buffer has a higher conduc-
tivity and thus draws a higher current, giving excessive heating).
Approximate pKa values of the compounds employed in the buffers
are H2SO4, 2.0 (second ionisation); formic acid, 3.7; acetic acid,
4.7; pyridine, 5.3; borate, 9.2; molybdate, 6.0. White spirit (painters’
turpentine substitute) is used as the coolant for HVPE at pH 2.0
or 3.5. A toluene/pyridine mixture is used for cooling the pH 6.5
buffer because pure toluene or white spirit would extract a high
proportion of the pyridine (which is largely in its unionised form
at pH 6.5) from the wet paper into the coolant, thus decreasing
the pH of the buffer during the run.
Sodium borate and sodium molybdate are not volatile, so
compounds that have been purified by HVPE in these buffers and
eluted from the paper with water require de-salting, e.g. by
­re-electrophoresis in a volatile buffer such as at pH 2.0.
 Table 1
Recommended HVPE buffer/coolant/marker combinations a

Suitable coloured negative Suitable coloured

Buffer b Composition Coolant markers positive markers
Volatile buffers
pH 2.0 Formic acid/acetic acid/H2O White spirit Orange G [picrate would be Ne-2,4-dinitrophenyl-
(1:4:45 by vol.) lost into the white spirit] lysine, methyl green,
methyl violet
pH 3.5 Acetic acid/pyridine/H2O White spirit Orange G, picrate Methyl green
(10:1:189 by vol.)c
pH 6.5d Acetic acid/pyridine/H2O Toluene/pyridine (20:1) Orange G, picrate Methyl green
(10:1:189 by vol.)
Sugar-complexing buffers
Borate, pH 9.4 1.9% w/v Borax White spirit Orange G, picrate [Not required]
(Na2B4O7 · 10H2O), pH
adjusted with NaOH
Molybdate, pH 3–5 2.0% w/v Na2MoO4 · 2H2O, pH White spirit Orange G, bromophenol [Not required]
adjusted with formic acid or blue, picrate
H2SO4
a
In most buffer systems, a suitable neutral marker is glucose, which is revealed by staining. A better alternative is a non-ionic fluorescent compound, such as feruloyl-arabinose
(20), which can be used at trace concentrations as an internal marker and visualised by its autofluorescence without staining. On HVPE in borate or molybdate buffers, glucose
is unsuitable as a neutral marker since it may bind oxyanions; non-binding alternatives include 2,3,4,6-tetra-O-methylglucose
b
Except for pH 3.5, the buffer specified is used both at the electrodes and for wetting the paper
c
The composition of the electrode buffer is given; it should be diluted with an equal volume of water for wetting the paper
d
For quantitatively accurate conclusions about charge:mass ratio to be drawn, the effective pH of the buffer in the paper during electrophoresis must be known. In the case of the
pH 6.5 buffer, it sometimes happens that insufficient pyridine is added to the coolant (toluene), so that although the buffer used for wetting the paper had been accurately
adjusted to pH 6.5 the pH of the buffer within the paper gradually falls during the run because some of the pyridine partitions from the paper into the toluene. A good test
compound with which to show whether the pH in the paper during the run was correct is histidine, whose imido group has a pKa of 6.0. At pH 6.5, histidine has a net charge
High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks

of +0.24, increasing strongly at slightly lower pHs. On the other hand, lysine has a constant net charge of approximately +1.00 at all pH values between 4.5 and 7.5. At pH 6.5,
His should move towards the cathode at about 0.23× the rate of Lys. If His runs almost as fast as Lys, more pyridine should be added to the toluene
59
60 Fry

2.4. Buffers Used Since non-volatile salts (including buffers) can interfere in
in Sample Preparation ­electrophoresis, these should be avoided during sample prepara-
Prior to tion. Any buffer used, e.g. to control the pH of an enzyme, should
Electrophoresis preferably be volatile. For the pH range 2–6, this can be similar to
one of the mixtures given in Table 1 – if necessary diluted so that
pyridine, which at high concentrations may inhibit enzymes, does
not exceed ~2% v/v. For pHs in the 6–8 range, 1% v/v methyl­
pyridines (adjusted to the desired pH with acetic acid) form use-
ful volatile buffers e.g. 2,6-dimethylpyridine (lutidine; pKa » 6.7)
or 2,4,6-trimethylpyridine (collidine; pKa » 7.4).

2.5. Wetting the Paper 1. Tissue paper.

2. Pipette.
3. Two glass rods.
4. Neutral marker: examples include glucose, [14C]glucose, or
5-O-feruloyl arabinose.
5. Mobile marker: typically orange G (Table 1).
6. Glass plate: this should be slightly larger than the paper to be
loaded.
7. Wash-bottle.
8. Buffer (see Subheading 2.4).

2.6. Detection 1. Aniline hydrogen phthalate stain: 16 g phthalic acid in

of Analytes 490 mL acetone, 490 mL diethyl ether and 20 mL dH2O is
used to prepare a stock solution. 0.5 mL of aniline is added to
100 mL of stock solution immediately before use (2).
2. AgNO3 stain (2) solution A: 0.8 g AgNO3 is dissolved in
1.6 mL water, then diluted into in 104 mL acetone; if neces-
sary, a little extra water added dropwise to redissolve the
AgNO3.
3. AgNO3 stain solution B: 100 mL ethanol + 1.25 mL 10 M
NaOH.
4. AgNO3 stain solution C: 10% w/v sodium thiosulphate.
5. NH3 solution.
6. Folin–Ciocalteu reagent.
7. Ninhydrin stain containing isatin: 270 mg ninhydrin, 130 mg
isatin, 2 mL triethylamine in 100 mL acetone (2).
8. −80°C freezer.
9. Fluor: 7% (w/v) PPO in ether.
10. Autoradiography film.
11. Film for fluorography: autoradiography film pre-flashed with
a photographic flash gun at such a distance that the back-
ground of the film will be slightly fogged when developed.
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 61

12. pH indicator: 0.04% (w/v) bromophenol blue in 10 mM

NaOH.
13. Solution for removing pyridine: acetic acid/toluene 1:20
(v/v).
14. Solution for removing acetic acid: diethyl ether/methanol
3:1 (v/v).

3. Methods

3.1. Layout 1. If it is known that all compounds of interest in the samples are
of Electrophoretogram neutral or negatively charged, the samples are loaded 12 cm
from the cathode end of the paper. Conversely, if all compounds
of interest are neutral or positively charged, they are loaded
9 cm from the anode end. If anions and cations are both of
interest, the samples are loaded near the middle of the paper.
2. For analytical HVPE, the solutions are typically loaded at up
to 12 and 30 mL per sample on Whatman No. 1 and 3CHR
papers respectively. If necessary, multiple 12- or 30-mL por-
tions are applied, with drying between each application.
3. A small amount of a visible mobile marker (e.g. 5 mg orange
G; Table 1) is either pre-mixed with the sample (i.e. used as
an internal marker) or loaded as a series of spots alternating
with the samples (i.e. as an external marker).
4. A neutral marker: preferably internal (e.g. glucose, [14C]glu-
cose, or 5-O-feruloyl-arabinose; revealed by staining, auto­
radiography or autofluorescence, respectively), should also be
included so that the extent of electro-endo-osmosis (see
Subheading 3.3 below) can be monitored.

3.2. Wetting 1. The paper, with samples loaded, is laid on a large glass plate
and Running and wetted with electrophoresis buffer (Fig. 2a). The majority
the Electrophoretogram of the paper area can be wetted quite quickly, e.g. by use of a
wash-bottle.
2. Any excess buffer is very lightly blotted from the electropho-
retogram with dry tissue paper; care must be taken not to
crease the wet paper during this process.
3. Wetting in the vicinity of the samples should be done last, and
very carefully, with a pipette so that the samples do not dif-
fuse excessively. This can be achieved if the area of the paper
that includes the origin line is suspended between two glass
rods; the wetting of this part of the paper is left until last.
With practice, the wetting of this part of the paper can be
made to focus the sample spots into narrow bands, instead of
discs, along the origin line (Fig. 2b).
62 Fry

buffer application electrophoresis dried samples

paper

glass rods glass plate

samples focused by
uniformly wet paper buffer application

Fig. 2. Method for wetting the electrophoretogram with running buffer.

4. With the apparatus shown in Fig. 1, and with the samples

loaded near the end of the paper, we typically perform HVPE
at 4.5 kV for ~30 min (see Note 8). With a full-size sheet of
Whatman No. 1 paper (57 × 42 cm), appropriately wetted,
4.5 kV delivers a current of about 150 mA (pH 2.0 buffer),
80 mA (pH 3.5) or 130 mA (pH 6.5) (see Note 9).
5. It should be noted that the pH of the buffer has an effect on
the charge of the analytes (see Note 10, Tables 2 and 3).

3.3. Reporting The mobility (m) of a substance in HVPE is often quoted relative
the Electrophoretic to an easily detectable mobile marker, e.g. orange G, and a neutral
Mobility of an Analyte one, e.g. glucose. The neutral marker is important since neutral
compounds often move slightly away from the origin by electro-
endo-osmosis. If orange G is used, the electrophoretic mobility is
quoted as mOG, where:

(distance moved by compound) − (distance moved by glucose)

mOG = .
(distance moved by orange G) − (distance moved by glucose)

A compound that remains at or near the origin while the neu-

tral marker moves towards the anode should therefore usually be
recorded as having electrophoretic mobility towards the cathode,
as exemplified by AMP on HVPE at pH 2.0 (Fig. 3a). This is
based on the assumption that a compound remaining at the ori-
gin has not been insolubilised there (as would happen for example
with cellohexaose, which hydrogen-bonds strongly to the cellulose
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 63

Table 2
Typical pKa values of some functional groups involved in plant cell-wall and
apoplast biochemistry

Examples of compounds in the

Functional group Qa pKa value(s) stated pKa rangeb

–COOH – 1.3 Oxalic acid (1st ionisation)

1.8–2.5 a-Carboxy group of amino acids,
oxaloacetic (1st), pyruvic,
diketogulonic
3.0–5.2 Most typical carboxylic acids,
e.g. acetic, gluconic, glucuronic,
glucaric (both), tartaric (both),
malic (both), ferulic, side-chain of
glutamic and aspartic, oxalic
(2nd), citric (1st & 2nd)
5.0–6.5 Tri- and a few di-carboxylic acids
e.g. citric (3rd), succinic (2nd)
–SO3H – 1.3 Cysteic acid
F–OHc – 8.5–10.5 Phenolic –OH of tyrosine, isodity-
rosine, dityrosine (2nd), feruloyl
esters, ferulate, coumarate
6.7 Phenolic –OH of dityrosine (1st)
Amino + 10–11 Most typical amino groups, e.g. of
methylamine, lysine (side-chain),
putrescine (2nd); also imino group
of proline
8.5–10 a-Amino group of amino acids;
amino group of putrescine (1st);
also imino group of
hydroxyproline
7–8 Amino group of amino-sugars
Phosphate ester – 1–2 Phosphate group of Glc-6-P,
dihydroxyacetone phosphate etc.
(1st)
2− 6–6.5 Phosphate group of Glc-6-P,
dihydroxyacetone phosphate etc.
(2nd)
Imidazole + 6.0 Imidazole ring of histidine
a
Q = sign of charge when ionised
b
1st, 2nd, both etc., refer to compounds possessing more than one of the functional group mentioned
c
F = benzene ring

of paper) — an assumption that can easily be tested by rinsing in

water if there is any doubt. mOG values estimated from Fig. 3a,
with fructose as the neutral marker, are: ATP, 0.57; ADP, 0.36;
AMP, −0.05; and (by definition) fructose, 0.00; orange G, 1.00.
64 Fry

Table 3
Behaviour of the principal ionisable groups present in cell-wall components and
precursors (amino, carboxy and phosphate) on HVPE at pH 2.0 and 6.5

Little or no net
Observed net charge Positive at pH 2.0 charge at pH 2.0 Negative at pH 2.0

Positive at pH 6.5 −NH2 group(s) n/aa n/aa

present. Fewer or
no acidic groups.
Lysine, putrescine
Little or no net charge −NH2 group(s) No −COOH, −NH2, n/aa
at pH 6.5 present. Equal phosphate or
−COOH groups. sulphate groups.
Serine, isodityrosine Glucose, glucitol
Negative at pH 6.5 −NH2 group(s) Weak acid group(s) Strong acid group(s)
present. More present present e.g.
−COOH groups. e.g. −COOH. phosphate, sulphate,
Aspartate, Galacturonate, or a low-pKa
glutamate malate, ascorbate, −COOH. Few or no
tartrate, citrate −NH2 groups.
Glucose 6-phosphate,
oxalate
The sample is run by HVPE at pH 2.0 and 6.5, and the direction of migration (if any) at these two pHs leads to the
conclusions entered in the table. Specific examples of compounds in each category are given in italics
a
Would be an implausible result

3.4. Calibrating As mentioned above, the mobility of a compound during HVPE

with Electrophoretic is decided by its charge and mass. More precisely, mobility is pro-
Mobilities portional to the Q:Mr2/3 ratio, where Q is the net charge and Mr
of “Knowns” to the power of 2/3 is an indication of the molecule’s relative
surface area (1). It is useful to prepare a calibration curve of the
relationship between mOG and Q:Mr2/3 ratio for a few “knowns”
run as markers. The Mr of an authentic marker is usually given by
the suppliers (but subtract any contribution due to a counter-ion
such as the Cl in glucosamine hydrochloride, or any water of
crystallisation).
If the pKa of the “known” has been published (for example,
see http://research.chem.psu.edu/brpgroup/pKa_compilation.
pdf), then its Q can be calculated. The pKa is the pH at which
50% of the molecules are ionised (at the group under consider-
ation) and 50% are not. For example, glyceric acid has a pKa of
3.5; therefore, if a dilute solution of this compound is present
in pH 3.5 electrophoresis buffer, 50% of the molecules at
any moment are glycerate anions (CH2OH–CHOH–COO−,
­represented as A−), while the other 50% are un-ionised
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 65

a pH 2.0 b pH 6.5

anode →
anode →

P2
[orange
G]

C5P2 (−)
C5P2 C6 P1
P C6P2
P2
C6P2
UDP
UDPG UDP
C3P1 ATP C3P1
NADPH [orange G]
NADPH UDPG
ADP C5P1
ATP
C5P1 C6P1
C6P1 ADPG C7P1
C7P1 NADP+
P C6(−)P1
ADPG
ADP
AMP
NADP+

fructose (C6) fructose (C6)

AMP origin origin

ATP, ADP, AMP

GlcO-6-P
UDP-Glc

Glc-1-P

Glc-6-P
ATP, ADP, AMP

ADP-Glc

Rib-5-P

Sed-7-P
DHAP
Ru-5-P
NADP+, NADPH

Fru-6-P
UDP

Fructose + Pi
GlcO-6-P
UDP-Glc

Glc-1-P

Glc-6-P
ADP-Glc

Rib-5-P

Sed-7-P
DHAP
Ru-5-P

Fru-6-P

Fructose + Pi
UDP

Ru-1,5-P2
Fru-1,6-P2
Ru-1,5-P2
Fru-1,6-P2

NADP+, NADPH

PPi
PPi

← cathode
← cathode

Fig. 3. HVPE at pH 2.0 and 6.5 used for separating classes of phosphorylated metabolites involved in cell-wall ­biosynthesis.
Here, we are aiming not to resolve all possible phosphorylated metabolites, but to place them into classes of related
compounds (sharing approximately the same charge:mass ratio) prior to more detailed analysis. For example, the hexose
monophosphate pool can later be eluted from the electrophoretogram and hydrolysed to determine whether the sugar
present is Glc, Gal, Man, Fru etc. Specific compounds run are listed along the origin of each electrophoretogram; classes
of compounds are listed along the right-hand edge: for example, “C5P2” indicates a compound (ribulose 1,5-bisphos-
phate) with five carbon atoms and two phosphate groups. “C6(−)P1” is gluconate 6-phosphate: note that its carboxy group
is strongly ionised at pH 6.5 but not 2.0, causing it to migrate very differently on the two electrophoretograms. Nucleotides
are listed along the left edge. Orange G, a coloured marker, was added into each sample before electrophoresis. Some of
the standard solutions used contained Pi as a contaminant. Each compound was loaded at 50 mg per spot except dihy-
droxyacetone phosphate (DHAP; 25 mg), and Pi and PPi (15 mg each). Electrophoresis was conducted on Whatman no. 1
paper at 4.5 kV for 30 min (a, pH 2.0) or 35 min (b, pH 6.5), and the spots were stained with molybdate reagent (2).
Fructose (C6) is not phosphorylated and is therefore not immediately revealed by the molybdate reagent; its position
(dotted outline) gradually becomes visible when the electrophoretogram is stored. Neutral compounds such as fructose
(C6) migrate slightly towards the anode owing to electro-endo-osmosis. At pH 2.0, AMP has a small net positive charge
and therefore moves slightly towards the cathode (relative to the neutral marker, fructose).

­(CH2OH–CHOH–COOH, represented as HA). In other

words, at pH 3.5, the reaction
H + + A − ↔ HA
is precisely balanced with half the molecules on the left and half
on the right. If a buffer of pH 2.0 had been used (i.e. with a
32-fold higher concentration of H+ than at pH 3.5), then the
equilibrium would have been pushed towards the right and there
would have been much more HA and much less A−. Conversely,
if a buffer of pH 6.5 had been used, then the vast majority of the
66 Fry

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5
1.00 1.00
0.95 0.95
0.90 0.90
0.85 0.85
0.80 0.80
0.75 0.75
E xte n t o f n e ga tive ch a rge

0.70 0.70
0.65 0.65
0.60 0.60
0.55 0.55
0.50 0.50
0.45 0.45
0.40 0.40
0.35 0.35
pH pH pH
0.30 2.0 3.5 6.5 0.30
0.25 0.25
0.20 0.20
0.15 0.15
0.10 0.10
0.05 0.05
0.00 0.00
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5

p Ka

Fig. 4. Relationship between pH and pKa. The degree of ionisation of any anionic group with a pKa between 0.0 and 8.5
at each of the three recommended pH values for routine HVPE. For example, the graph shows that a carboxy group with
pKa = 6.0 has a charge of −0.76 in pH 6.5 buffer. The graph can also be used for cationic groups, but an amino group
with pKa = 6.0 has a charge of 1.00–0.76 = +0.24 in pH 6.5 buffer.

molecules would have been in the form A−. The ratio of [A−] to
[HA] at any given pH is given by the equation
log{[A − ] / [HA]} = pH − pK a .
Thus, with glucuronic acid (pKa = 3.2) in pH 3.5 buffer,
log {[A−]/[HA]} = 0.3, so [A−]/[HA] » 2.0, and therefore 67% is
present as the glucuronate anion and 33% as uncharged glucuronic
acid. This can conveniently be described as GlcA having a net
charge Q = −0.67 at pH 3.5. This and similar values can be read
off in Fig. 4.
With weak bases such as glucosamine (Glc–NH2), similar rules
apply, but raising the pH decreases the % ionisation, e.g.
Glc −NH3+ ↔ H + + Glc −NH 2 .

The equation for weak bases is:

log{[B] / [BH + ]} = pH − pK a ,
where, in this example, B is Glc–NH2 and BH+ is Glc–NH3+.
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 67

In compounds with two or more ionisable groups, each group

must be treated separately, and the net charge for the whole ­molecule
is then calculated by addition of the individual charges. For example,
with fumaric acid [which has two carboxy (–COOH) groups,
with pKa values of 3.0 and 4.4 respectively] in an electrophoresis
buffer of pH 3.5, it is calculated that the “first” and “second”
carboxy groups contribute partial charges of −0.76 and −0.11
respectively, and thus the compound has net charge Q = −0.87 at
this pH. As a second example, leucine has a carboxy group with
pKa = 2.3 and an amino (–NH2) group with pKa = 9.7; thus in a
buffer at pH 2.0 these two ionisable groups contribute partial
charges of −0.33 and very nearly +1.00 respectively, giving leucine
a net charge Q = +0.67.

3.5. Interpreting Armed with a graph plotting mOG against Q:Mr2/3 for several
Electrophoretic “knowns”, we can interpret the mOG values of unknown com-
Mobilities pounds, run under the same conditions, in terms of their Q:Mr2/3
of “Unknowns” ratios. If either of the parameters, Q or Mr, can be assumed, then
we can estimate the other. For example, if the charge is known to
be −1 at the pH of the electrophoresis buffer, then the ­compound’s
Mr can be estimated. Likewise, if the Mr is known (e.g. because we
know it is a hexuronic acid), then we can estimate the pKa, which
may identify which specific uronic acid it is.
On electrophoresis at pH 2.0, amino groups are fully posi-
tively charged. Therefore, a compound with a single amino group,
and no groups that possess an appreciable negative charge at that
pH, can be estimated for size by HVPE if we have a calibration
curve. An example of a calibration curve is given in Fig. 5.
The compounds tested here were reductively aminated sugars
[oligosaccharidyl-1-amino-1-deoxyalditols (OADs), prepared from
glucose and various authentic oligosaccharides of DP 2–9], which
can be assumed to have Q = +1.00 at pH 2.0. The reference com-
pounds were glucose and glucosamine, so the y-axis on this occa-
sion reports mGlcN rather than mOG. The graph exhibits a good
approximation to a straight line, so this calibration curve can be
used to estimate the size of other OADs, prepared from unknown
oligosaccharides.
At pH 6.5, most carboxy groups are almost fully negatively
ionised, so each carboxy group can be assumed to have Q  » −1.
Again, then, if the approximate Mr is known, a suitable calibration
curve allows us to “count” the carboxy groups. Thus, for example,
the C4 compound tartaric acid (with two COOH groups) is
readily distinguished from threonic acid (also a C4 compound but
with one COOH group). Likewise, if the number of carboxy
groups is known, then the Mr can be roughly estimated. These
approaches have proved useful in characterising novel apoplastic
metabolites of ascorbate by HVPE at pH 6.5 (3).
68 Fry

1.0

0.8

0.6
mGlcN

0.4

0.2

0.0
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035
Q / Mr2/3

Fig. 5. A calibration curve plotting electrophoretic mobility against the Q:Mr2/3 ratio for
several “knowns” (reductively aminated sugars). 1-Amino-1-deoxyalditols were obtained
by reaction of glucose or an oligosaccharide in the presence of NaCNBH3 + NH4HCO3. The
x-axis shows the Q:Mr2/3 ratio calculated on the basis that Q = +1.00; the y-axis plots
the observed electrophoretic mobility (as mGlcN) at pH 2.0. The top data-point is for glu-
cosamine itself (mGlcN = 1.00, by definition). Other data-points refer to (in order of
decreasing mGlcN) reductively aminated glucose, maltose, cellobiose, maltotriose, malto-
tetraose, isomaltotetraose, maltopentaose, maltohexaose, XXXG, maltoheptaose, XXLG
and XLLG (21).

3.6. Detection Staining can be used after HVPE to detect many cell-wall
of Analytes: Staining c­omponents with reasonable sensitivity, though usually destruc-
tively. Several methods are described in detail by Fry (2).

3.6.1. Staining for Sugars Sugars are readily detected by staining with aniline hydrogen-
phthalate or AgNO3. The former is quicker, distinguishes different
classes of sugar by colour differences, and is compatible with the
subsequent detection of radioactivity, but fails to detect non-
reducing sugars and is less sensitive than AgNO3. AgNO3 also
detects non-reducing sugars e.g. trehalose and sucrose as well as
“sugar-like” compounds such as glycerol, galactonate, threonate,
tartrate, ascorbate and dehydroascorbate. AgNO3 can detect
down to about 0.1 mg of arabinose; aniline hydrogen-phthalate
down to about 0.4 mg. Borate may interfere with AgNO3 staining
(see Note 11).
1. Aniline hydrogen-phthalate
(a) Dip the paper through the aniline hydrogen-phthalate
stain (see Subheading 2.6.1).
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 69

(b) Allow the paper to dry.

(c) Heat the paper in an oven at 105°C for 5 min.
2. AgNO3 staining
(a) Work in subdued light throughout. Dip the electropho-
retogram through solution A (see Subheading 2.6.2).
(b) Dry for ~15 min.
(c) Dip the electrophoretogram with a smooth continuous
motion through solution B (see Subheading 2.6.3).
(d) Dry ~15 min.
(e) Repeat steps 3 and 4 until the spots can be seen clearly.
(f ) Dip the paper through solution C (see Subheading 2.6.4)
and then wash in tap water for 1–2 h.
(g) Dry the paper.

3.6.2. Staining Phosphates, including sugar-phosphates and NDP-sugars, are

for Phosphates detected by molybdate reagent (2) (Fig. 3), although this is rela-
tively insensitive (down to 2 mg of Glc-6-P) and is mainly used for
localisation of markers run alongside (or mixed with) radioactive
samples of interest.

3.6.3. Staining Phenolics can often be visualised by their autofluorescence, some-

for Phenolics times intensified by exposure of the paper to NH3 vapour, or less
sensitively by spraying with commercially available Folin–Ciocalteu
phenol reagent followed by exposure of the paper to NH3 vapour
until the yellow background is decolourised.

3.6.4. Staining for Amino Amino acids can be detected very sensitively with ninhydrin,
Acids which is rendered more specific for hydroxyproline by inclusion
of isatin (2).

3.6.5. Staining Acidic Acidic compounds such as oxalic or citric acids, that are not phos-
Compounds phates or carbohydrate-related, and therefore cannot be stained
with molybdate or AgNO3, can be detected if the paper is sprayed
with an aqueous solution of a pH indicator (0.04% bromophenol
blue in 10 mM NaOH). Acidic analytes show up as yellow spots
on a blue background. Before use of a pH indicator, the electro-
phoretogram should be freed of any traces of pyridine (by dip-
ping through acetic acid/toluene 1:20 and re-drying) and then
thoroughly freed of acetic acid; this can be promoted by repeated
dipping of the paper through diethyl ether/methanol (3:1) and
re-drying. After the paper has been sprayed with the pH indicator,
the contrast between non-volatile acidic analytes and the back-
ground colour can be adjusted if the paper is lightly exposed to
NH3 vapour. The spots may be transient and should be recorded
as soon as they appear.
70 Fry

3.7. Detection Radioactivity is often detected on paper electrophoretograms by

of Radioactivity autoradiography (for 14C, 32P, 33P and 35S) or fluorography (3H).
3.7.1. Autoradiography 1. Expose the electrophoretogram to film.
2. After incubating in the dark develop the film in a darkroom.

3.7.2. Fluorography 1. The paper is dipped through a fluor (7% w/v PPO in ether)
and dried.
2. The film is pre-flashed (see Subheading 2.6.8).
3. The electrophoretogram is exposed to the film at a low
­temperature e.g. −80°C (see Note 12).

3.8. Detection Substances separated on the electrophoretogram can be eluted

of Analytes: Bioassay for further analysis (see Subheading 3.9) – including bio-assays
(for oligosaccharins), in which case the paper should be carefully
freed of all traces of the liquids used during electrophoresis (ace-
tic acid, pyridine, trace contaminants from the white spirit, etc.).
Anionic analytes on the paper (e.g. oligogalacturonides) will be
present as their pyridinium salts; thus, after electrophoresis, the
dried paper is dipped in toluene/acetic acid (20:1 v/v) and re-
dried to remove the pyridine, then dipped in diethyl ether/meth-
anol (3:1 v/v) and re-dried to remove the acetic acid. If necessary,
the paper can be very thoroughly washed in pure toluene (e.g. by
descending paper chromatography with the toluene as the solvent)
to remove contaminants picked up from the coolant (white spirit
or practical-grade toluene).

3.9. Elution Method Substances separated on the electrophoretogram can be eluted,

usually in water, for further analysis. Elution is achieved in a mini-
mal volume of eluent by the syringe-barrel method (4).
1. The relevant zone of the electrophoretogram is cut out with
scissors and tightly packed into a plastic syringe barrel, which
is then suspended in a plastic centrifuge tube (Fig. 6).
2. The paper is wetted with just enough water to moisten it.
3. The assembly is bench-centrifuged (e.g. 4,000 × g), causing
the eluate to collect in the bottom of the centrifuge tube.
4. Moistening and centrifugation are repeated, typically 4–5 times,
until the analyte of interest has been eluted. Simultaneously
eluting an orange G marker spot can indicate when elution is
complete.

3.10. Specific Monomeric uronic acids, aldonic acids, aldaric acids, Krebs-cycle
Examples intermediates and many ascorbate metabolites possess 1–3 ionis-
able carboxy groups. At pH 6.5, most carboxy groups are almost
3.10.1. Monomeric Sugar
fully ionised: any carboxy group with a pKa less than 5.2 will be
Acids and Related
>95% fully ionised (Fig. 4). Thus, the advantage of HVPE at
Metabolites
pH 6.5 is that it can be used to “count” the compound’s carboxy
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 71

Eluent

2.5-ml
2.5 plastic
syringe barrel

15-ml plastic
centrifuge tube

Paper from
HVPE

Eluate

Fig. 6. Method for the elution of analytes from electrophoresis paper.

groups if its approximate Mr is known, or to estimate its Mr if the

number of carboxy groups is known. If neither is known (as was
the case with novel metabolites of ascorbate e.g. 4-O-oxalyl-l-
threonate and a corresponding cyclic oxalyl ester; (3)), useful
information can still be obtained, for example that the compound
is either a (C6 2−)1 compound such as oxalyl-threonate or a
(C3 1−) compound such as glyceric acid.
At pH 6.5 there is little resolution between different mem-
bers of a given class e.g. hexuronic acids, which all have the same
Mr and a single carboxy group and are thus grouped together into
a tight zone on the electrophoretogram. On the other hand,
pH 3.5 is reasonably close to the pKa of many carboxy groups.
Therefore, small differences in pKa, such as exist between iso-
meric uronic acids (e.g. GalA, GlcA, ManA and GulA), enable
their separation from each other (Fig. 7) (5).

3.10.2. Acidic HVPE at pH 6.5, at which uronic acid carboxy groups are almost
Oligosaccharides fully ionised, is valuable for determining whether an acidic oligo-
saccharide of known size (degree of polymerisation, as estimated
for example by gel-permeation or paper chromatography) con-
sists only of acidic residues (e.g. galacturonobiose) or of a mixture
of neutral and acidic residues (e.g. a-d-glucuronosyl-(1→3)-l-
galactose (6)). As above, the charge:mass ratio is obtained.

i.e. a compound with 6 carbon atoms and two negative charges.

1
72 Fry

Fig. 7. HVPE at pH 3.5 used for separating specific sugar-acids, differing subtly in their pKa values. The specific com-
pounds are listed along the left edge; classes of compounds are listed on the right : for example, (“C6, 1−)” indicates a
compound (e.g. galacturonic acid) with six carbon atoms and one negatively ionised group. Orange G, a coloured marker,
was loaded between each sample. Each compound was loaded at 10 mg per spot except h and i, which were crude
preparations of unknown concentration; h is a mixture of ManA and GulA. Electrophoresis was conducted on Whatman
3CHR paper at 4.5 kV for 50 min, and the spots were stained with AgNO3. Several of the sugar-acids are accompanied by
neutral lactones, co-migrating with glucose. The suffixes -A, -O and -R indicate -uronic acids, -onic acids and -aric acids:
for example, GalA galacturonic acid; GalO galactonic acid; GalR galactaric acid. The left and right halves are from a single,
wide electrophoretogram, illustrating the good reproducibility of mOG values; the compounds shown acted as external
markers for a radioactive experimental sample (central band “S”; not shown in full).

3.10.3. Sugar Phosphates Since phosphate and sulphate are strongly acidic groups, they
and Sulphates remain appreciably ionised during HVPE at pH 2.0 (Fig. 3a),
unlike most of the carboxylic acid groups present in ­carbohydrates.
Thus, it is possible to obtain an excellent class-separation of sugar
phosphates from non-phosphorylated sugars at this pH (7). Here,
High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 73

we are aiming not to resolve all possible phosphorylated metabolites

from each other, but to place them into classes of related com-
pounds (sharing approximately the same charge:mass ratio) prior
to more detailed analysis. For example, the hexose monophos-
phate pool can later be eluted from the electrophoretogram and
hydrolysed; this will reveal whether the sugar present is Glc, Gal,
Man, Fru etc. At pH 2.0, all the sugar phosphates are tightly clus-
tered, though, as expected, the tendency is for some resolution
to occur on the basis of size: in order of decreasing mobility
towards the anode, we have triose-phosphates, tetrose-phosphates,
pentose-phosphates, hexose-phosphates and heptose-phosphates
(Fig. 3a). Gluconic acid 6-phosphate co-migrates with glucose
6-phosphate because the carboxy group of the former is not
appreciably ionised at pH 2.0. Sugar bisphosphates, such as ribu-
lose 1,5-bisphosphate and fructose 1,6-bisphosphate, migrate
considerably faster than the corresponding monophosphates
because both phosphate groups ionise.
At pH 6.5 (Fig. 3b), each phosphate group carries a stronger
negative charge than at pH 2.0 and there is a more reliable
­correlation between size and mobility (mobility of triose- >
­tetrose- > pentose- > hexose- > heptose-phosphates). In this case,
however, gluconic acid 6-phosphate runs much faster than glu-
cose 6-phosphate because its carboxy group is almost fully ionised
at pH 6.5.
Often the isolation of a given class of compounds by HVPE at
pH 6.5, e.g. the hexose monophosphates, provides the metabolic
information required for the project in hand, since Fru-6-P, Glc-
6-P, Glc-1-P and Gal-1-P are all readily inter-converted in vivo.
However, if necessary, a further class separation is easily achieved
by graded hydrolysis: the aldose 1-phosphates (Glc-1-P, Gal-1-P,
Man-1-P, etc.) are completely hydrolysed to the free monosac-
charides under mildly acidic conditions (0.1 M trifluoroacetic acid
at 100°C for 25 min), whereas aldose 6-phosphates (e.g. Glc-6-P,
Man-6-P), ketose 6-phosphates (e.g. Fru-6-P) and ketose 1-phos-
phates (e.g. Fru-1-P, although this probably does not occur in
plants) are not. Thus, the sugars from the former class are obtained
as free monosaccharides after mild acid hydrolysis; thereafter the
monosaccharides can be obtained from the remaining unhydroly-
sed hexose monophosphates belonging to the latter class by diges-
tion with a commercial phosphatase preparation (7).
The major hexose bisphosphate (Fru-1,6-P2) is not appreciably
hydrolysed by mild acid – unlike the UDP-sugars, which release
the monosaccharide almost quantitatively. Mild acid removes one
phosphate group (the one attached to the anomeric carbon)
from Glc-1,6-P2, Man-1,6-P2 and Fru-2,6-P2, leaving a hexose
6-phosphate.
Similar advice applies to phosphorylated sugars other than
hexoses but, in the case of pentoses for example, for “6” read “5”.
74 Fry

3.10.4. Nucleotides, As with sugar monophosphates versus bisphosphates at pH 2.0,

Including NDP-Sugars there is excellent resolution within a given series of nucleotides: in
and CoA-Thioesters order of decreasing mobility towards the anode, we have ATP,
ADP and AMP (Fig. 3a). Sugar-nucleotides tend to migrate fairly
close to the corresponding nucleoside diphosphates (e.g. UDP-
Glc near UDP, and ADP-Glc near ADP). Some of the nucleoside
moieties (adenosine, uridine etc.) differ in charge at pH 2.0
depending on the presence of amino groups. Adenosine has such
a significant positive charge at pH 2.0 that AMP (despite its nega-
tively charged phosphate group) has a small net positive charge
and therefore moves slightly towards the cathode (relative to the
neutral marker). Thus, ADP-hexoses are very well resolved from
UDP-hexoses (Fig. 3a). GDP-hexoses migrate only slightly faster
than ADP-hexoses at pH 2.0 (data not shown).
At pH 6.5, UDP-glucose resolves well from UDP-glucuronate
(8) because the GlcA residue carries a full negative charge.
However, at pH 2.0, UDP-glucose runs only slightly slower than
UDP-glucuronic acid because the GlcA residue is scarcely ionised
at such a low pH. Thus, at pH 2.0, it is possible to obtain very
useful class separations of UDP-sugars (including UDP-Glc and
UDP-GlcA) in one zone and ADP-sugars + GDP-sugars in a sec-
ond zone. These two zones can be eluted from the paper for
further analysis, e.g. by mild acid hydrolysis (0.1 M trifluoroacetic
acid at 100°C for 25 min) followed by paper chromatography or
TLC (or HVPE at pH 3.5 for the uronic acids) to resolve the
diverse monosaccharides – for example from the UDP-sugar pool,
containing UDP-Glc, UDP-d-Gal, UDP-Ara, UDP-Xyl, UDP-
Rha, UDP-GlcA and UDP-GalA. Note that the ribose moiety of
NDP-sugars is not released by mild acid hydrolysis.
At both pH values, 2.0 and 6.5, NADPH migrates faster
towards the anode than NADP+ because the former lacks the extra
positive charge (note that both NADPH and so-called “NADP+”
both possess a net negative charge; the “+” is only relative) (Fig. 3).
At pH 3.5 and 6.5, in contrast to pH 2.0, NDP-sugars consis-
tently migrate slower than the corresponding NDPs.
HVPE at pH 3.5 is useful for class-separating ADP-hexoses
(mOG » 0.58) from GDP-hexoses (mOG » 0.75). These two classes
can subsequently be further analysed, e.g. by acid hydrolysis of
the GDP-sugar zone to yield the monosaccharides (typically
d-mannose, l-fucose and l-galactose).
Thioesters of coenzyme-A (e.g. acetyl-CoA, feruloyl-CoA
etc.) are conveniently analysed by HVPE at pH 2.0 (9), in which
buffer they all possess a substantial net negative charge.
Commercially available dodecanoyl-CoA is a useful marker for
feruloyl-CoA, which has about the same charge:mass ratio. [14C]
Cinnamate metabolites with an appreciable net negative charge at
pH 2.0 are very likely to be CoA conjugates since nothing else
would give them such a charge. They also migrate very rapidly at
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 75

pH 3.5 (mOG values: CoA, 1.0; feruloyl-CoA, 0.61). They can be

further characterised by mild alkaline hydrolysis, yielding the
­former acyl residue in free form (ferulate, acetate, etc.).

3.10.5. Amino-Sugars The amino groups of ManN, GalN and GlcN have pKa values of
about 7.3, 7.7 and 7.8, respectively. These compounds thus carry
almost a full positive charge at all pH values commonly used for
HVPE and so are not resolved except slightly at pH 6.5.
Nevertheless, HVPE at pH 2.0, 3.5 or 6.5 permits the quick and
easy class separation of sugar amines from acidic and neutral sugars,
including N-acetyl amino-sugars e.g. GlcNAc. Similarly, HVPE
at pH 2.0 allows excellent class separation of amino-sugar deriva-
tives with crystal violet (CV) as a cationic marker and orange G
(OG) as an anionic marker (10):
mCV = 1.26: GlcN, GalN
mCV = 0.20: GlcN-1-P, GlcN-6-P, GalN-1-P
mCV = mOG = 0: GlcNAc, GalNAc
mOG = 0.57: GlcNAc-1-P, GlcNAc-6-P, GalNAc-1-P
mOG = 0.76: UDP-GlcNAc, UDP-GalNAc
Amino-sugars and their derivatives can be stained with aniline
hydrogen-phthalate (unless carrying a phosphate group at posi-
tion 1) or ninhydrin (unless acetylated on the N).

3.10.6. Amino Acids a-Amino acids (i.e. compounds with an amino group and a ­carboxy
and Polyamines group attached to the same carbon, the simplest being glycine)
have at least one amino group (contributing at least one full posi-
tive charge at pH 2.0, 3.5 and 6.5), and an a-carboxy group which
bears a full or partial negative charge depending on the pH. If the
amino group is free (not involved in a peptide bond), the
a-carboxy group has a very low pKa (in the range 1.8–2.5) and is
thus 61–24% fully anionic at pH 2.0, unlike most other carboxy
groups which are almost neutral at that pH. The side-chain car-
boxy groups of Asp and Glu have much higher pKa values and are
only ~1% charged at pH 2.0 but highly charged at pH 3.5 and 6.5.
Thus, on HVPE at pH 2.0, all the common a-amino acids have a
net positive charge; Arg, Lys and His have a particularly large one
owing to the presence of a second positively ionising group.
Cysteic acid (an oxidation product of Cys) is the only com-
monly encountered a-amino acid with a net negative charge at
pH 2.0: this is due to its negatively ionising –SO3H group
(pKa = 1.3). Other than cysteic acid, hydroxyproline stands out on
HVPE at pH 2.0 as the slowest-migrating major “amino acid”
(strictly an imino acid) because its carboxy group is unusually acidic
(pKa = 1.8). Hydroxyproline mono-, di-, tri- and tetra-arabinosides,
obtained from some cell-wall glycoproteins such as extensins,
migrate progressively slower still (11).
HVPE at pH 2.0 followed by staining with Folin and
Ciocalteu’s phenol reagent is useful for the detection of tyrosine
76 Fry

and its ­oxidative coupling products (isodityrosine, pulcherosine,

di-isodityrosine), since these are among the very few cationic
phenols. Others that do exist include tyramine and N-feruloyl-
putrescine.
4-Aminobutyrate (GABA) is a significant stress metabolite in
plants, sometimes found in the apoplast and thus relevant to wall
metabolism. Its carboxy group is not attached to the same C atom
as its amino group and thus has a more typical pKa value (4.0, as
opposed to 2.3 for 2-aminobutyrate). Therefore, at pH 2.0, GABA
has a net charge of about +1.0 and migrates towards the cathode
about 90% as rapidly as lysine, which has two amino groups.
Polyamines possess hydrocarbon chains, usually with two
­primary amino groups; in addition there may be one or more
­secondary amino groups. Major examples in plants are putrescine
(a diamine), spermidine (a triamine), and spermine and its isomer
thermospermine (tetraamines) (12). Their presence in solution in
the apoplast and covalently bonded to wall polymers has been
reported. Polyamines migrate particularly rapidly towards the
cathode during HVPE, not only at pH 2.0 but also at pH 6.5 since
the amino groups are essentially fully ionised, there are no nega-
tively ionising groups, and they have low Mr. HVPE thus enables
a very convenient class separation, resolving polyamines from all
other common phytochemicals. All the common polyamines have
mlysine » 1.6 at pH 2.0, and stain with ninhydrin. Resolution of the
various polyamines from each other by HVPE requires a higher
pH buffer, closer to the pKa of the amino groups, e.g. 0.2 M
ammonium carbonate (pH 8.7); however, better separation of
spermine, spermidine and putrescine is obtained by paper chroma-
tography in butan-1-ol/acetic acid/pyridine/water (4:1:1:2).
It has been suggested that polyamines such as putrescine may
be linked via amide (isopeptide) bonds to pectic GalA residues
(13). Model compounds with which to search for the natural
occurrence of such linkages were synthesised chemically [e.g.
N-d-galacturonoyl-putrescinamide (GalA–Put) and N,N´-di-d-
galacturonoyl-putrescinamide (GalA–Put–GalA)]. In addition,
promising diagnostic “fragments” were isolated by Driselase
digestion of artificially putrescine-conjugated homogalacturonan
[yielding products such as Put–GalA3 and GalA3–Put–GalA3].
These various novel glycoconjugates were characterised largely by
HVPE on the basis of the following rules:
●● Carboxy groups: almost fully ionised negatively at pH 6.5,
partially at pH 3.5, almost un-ionised at pH 2.0.
●● Amino groups: fully ionised positively in all three buffers.
●● Amide (–CONH–) groups: un-ionised in all three buffers.
Similar techniques were used in the preparation and charac-
terisation of Ne-d-galacturonoyl-l-lysine and related conjugates,
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 77

which are useful model compounds with which to test for the
possible occurrence of pectin–extensin amide linkages (14).
Glutathione (GSH, a tripeptide) and its disulphide-bridged
oxidation-product (GSSG), both of which have been reported in
the apoplast, can also be resolved quickly and cleanly from many
co-occurring compounds by HVPE. Pre-labelling with [35S]sul-
phate is particularly helpful as plants possess few extracellular
­sulphur compounds. GSH and GSSG have a net positive charge
at pH 2.0 and net negative charge at pH 6.5. Selected mOG values
at pH 6.5 are: GSH, 0.74; GSSG, 0.88; cysteic acid, 1.29; inor-
ganic sulphate, 2.8; cysteine, 0.00; cystine, 0.00; methionine,
0.00 (15).

3.10.7. Electrophoresis Neutral sugars and alditols do not migrate on HVPE in ordinary
of Neutral Sugars buffers e.g. at pH 2.0, 3.5 and 6.5. However, such sugars can
reversibly be given a negative charge by complexation with oxy­
anions such as borate, molybdate, tungstate, stannate or alumi-
nate (2, 16). Complexation is a very rapid and simple procedure:
the sugar sample is loaded in the normal way (Fig. 2), and the
paper is wetted with an aqueous solution of the oxyanion, which
is used as the electrophoresis buffer (Table 1).
Borate binds to suitably orientated pairs of −OH groups,
which occur in most sugars and alditols, conferring a negative
charge. Binding is optimal at high pH (e.g. 9.4) and progressively
weaker at lower pHs. The borate–sugar dissociation constant (and
thus the average number of borate ions bound per unit mass of
sugar at equilibrium) determines the electrophoretic mobility.
The method is useful for distinguishing oligosaccharides that
­differ in bond position ((1→2), (1→3), (1→4) etc.) (2, 17, 18).
There may also be good separation of pairs of oligosaccharides
that differ only in anomeric configuration, e.g. maltose versus cel-
lobiose. Wall-derived oligosaccharides can often be freed of con-
taminating malto-oligosaccharides on the basis of the slow
migration of the latter in borate buffer (19).
Molybdate reversibly binds to alditols and related compounds
possessing certain patterns of −OH groups; binding is optimal at
pH 2 and is progressively weaker at higher pHs. HVPE in molyb-
date is particularly effective at resolving reducing oligosaccharides
(e.g. the xyloglucan-derived nonasaccharide, XXFG), most of
which cannot bind molybdate, from the corresponding reduced
oligosaccharides (e.g. XXFGol), which can bind it specifically at
the glucitol moiety. Thus, XXFGol is mobile in molybdate buffer,
whereas XXFG is not. The mOG value of a reduced oligosaccharide
is strongly influenced by the position(s) at which the rest of the
oligosaccharide is attached to the alditol moiety, and can there-
fore give valuable information on sugar–sugar linkages in novel
radiolabelled oligosaccharides (2, 20).
78 Fry

4. Notes

1. The applied voltage is typically 4.5 kV.

2. The effective path-length is a little less than 57 cm, the two
ends of the paper being submerged in buffer.
3. HVPE can conveniently handle up to twenty samples per
sheet if the samples are spot-loaded on No. 1 or 3CHR paper
at, respectively, <0.3 or <1.0 mmol of total ions (including
any non-volatile salts) per spot. Non-ionic compounds, e.g.
glucose or glycerol, can be present in higher amounts as these
do not interfere in the ionisation and electrophoresis
processes.
4. For preparative work, a single sample containing up to about
25 mmol total ions (equivalent to ~5 mg of galacturonic acid)
can be streak-loaded per sheet of 3CHR.
5. If the use of water-immiscible coolant liquids is not feasible,
e.g. because hydrophobic compounds are of interest, HVPE
can also be performed on a flat-bed system.
6. One potential disadvantage of this method is that non-polar
compounds e.g. ferulic acid might partition into the coolant
and be lost off the paper. However, the great majority of
metabolites of interest in cell wall research are hydrophilic
enough to be retained in the aqueous buffer in the paper;
even isoleucine (the most “hydrophobic” of the 20 common
amino acids) and feruloylated sugars are retained.
7. In this laboratory’s experience, this method provides less
­uniform cooling and therefore greater irregularity in the
migration of replicate samples across the width of the paper.
8. The high voltage and short run-time minimise diffusion of
the analytes on the wet paper.
9. This entails a considerable heating effect (e.g. at pH 2.0,
power = 4,500 V × 0.15 A = 675 W). The thicker 3CHR paper
draws an even greater current, and the voltage must be
­correspondingly reduced (and the run time increased) so that
the power remains at <750 W.
10. The charges borne by compounds during electrophoresis
depend on the pKa values of the ionisable groups and the pH
of the buffer. Approximate pKa values are listed in Table 2.
Deductions that can be drawn from pH-dependent shifts in
electrophoretic mobility, are summarised in Table 3.
11. Borate may interfere with AgNO3 staining; to avoid this on
electrophoretograms run in borate buffer, the alkaline solu-
tion normally used for AgNO3 staining should be replaced by
 High-Voltage Paper Electrophoresis (HVPE) of Cell-Wall Building Blocks 79

80% (v/v) ethanol containing 2% (w/v) NaOH and 4% (w/v)

pentaerythritol.
12. Flashing and cooling are not beneficial for autoradiography.

Acknowledgements

I thank numerous past and present colleagues and students for

electrophoretic data and experience presented in this chapter, and
the UK BBSRC for financial support.

References

1. Offord, R. E. (1966). Electrophoretic mobilities 10. Piro, G., Perotto, S., Bonfante-Fasolo, P., and
of peptides on paper and their use in the Dalessandro, G. (1988) Metabolism of d-(U-
determination of amide groups. Nature 211, 14
C)glucosamine in seedlings of Calluna vul-
591–593. garis (L) Hull. Journal of Plant Physiology
2. Fry, S. C. (2000). The Growing Plant Cell 132, 695–701.
Wall: Chemical and Metabolic Analysis, Reprint 11. Lamport, D. T. A. (1967) Hydroxyproline-O-
Edition. The Blackburn Press, Caldwell, NJ, glycosidic linkage of plant cell plant cell wall gly-
pp. xviii + 333 [ISBN 1-930665-08-3]. coprotein extensin. Nature 216, 1322–1324.
3. Green, M. A., and Fry, S. C. (2005) Vitamin 12. Takahashi, T., Kakehi, J. -I. (2010) Polyamines:
C degradation in plant cells via enzymatic ubiquitous polycations with unique roles in
hydrolysis of 4-O-oxalyl-l-threonate. Nature growth and stress responses. Annals of Botany
433, 83–88. 105, 1–6
4. Eshdat, Y., and Mirelman, D. (1972). An 13. Lenucci M., Piro, G., Miller, J. G.,
improved method for the recovery of com- Dalessandro, G., and Fry, S. C. (2005) Do
pounds from paper chromatograms. Journal polyamines contribute to plant cell wall assem-
of Chromatography 65, 458–459. bly by forming amide bonds with pectins?
5. Wright, K., and Northcote, D. H. (1975) An Phytochemistry 66, 2581–2594.
acidic oligosaccharide from maize slime. 14. Perrone P., Hewage, C., Sadler, I. H., and
Phytochemistry 14, 1793–1798. Fry, S. C. (1998). Na- and Ne-d-galacturonoyl-
6. Popper, Z. A., Sadler, I. H., and Fry, S. C. l-lysine amides: properties and possible
(2003) a-d-Glucuronosyl-(1→3)-l-galactose, occurrence in plant cell walls. Phytochemistry
an unusual disaccharide from polysaccharides 49, 1879–90.
of the hornwort Anthoceros caucasicus. 15. Kärkönen A., Warinowski, T., Teeri, T. H.,
Phytochemistry 64, 325–335. Simola, L. K., and Fry, S. C. (2009) On the
7. Sharples S. C., and Fry, S. C. (2007) Radio- mechanism of apoplastic H2O2 production
isotope ratios discriminate between compet- during lignin formation and elicitation in cul-
ing pathways of cell wall polysaccharide and tured spruce cells; peroxidases after elicitation.
RNA biosynthesis in living plant cells. Plant Planta 230, 553–567.
Journal 52, 252–262. 16. Weigel, H. (1963). Paper electrophoresis of
8. Kärkönen, A., and Fry, S. C. (2006) Novel carbohydrates. Advances in Carbohydrate
characteristics of UDP-glucose dehydroge- Chemistry 18, 61–96.
nase activities in maize: non-involvement of 17. Dumville J.C., and Fry, S. C. (2003)
alcohol dehydrogenases in cell wall polysac- Gentiobiose: a novel oligosaccharin in ripen-
charide biosynthesis. Planta 223, 858–870. ing tomato fruit. Planta 216, 484–495.
9. Fry, S. C., Willis S. C., and Paterson, A. E. J. 18. Narasimham, S., Harpaz, N., Longmore, G.,
(2000). Intraprotoplasmic and wall-localised Carver, J. P., Grey, A. A., and Schachter, H.
formation of arabinoxylan-bound diferulates (1980) Control of glycoprotein synthesis:
and larger ferulate coupling-products in maize the purification by preparative paper
cell-suspension cultures. Planta 211, ­electrophoresis in borate of glycopeptides
679–692. containing high mannose and complex
80 Fry

­ ligosaccharide chains linked to asparagine.

o 20. Wende, G., and Fry, S. C. (1996).
Journal of Biological Chemistry 255, 2-O-b- d -Xylopyranosyl-(5-O-feruloyl)- l -
4876–4884. arabinose, a widespread component of grass
19. O’Looney, N., and Fry, S. C. (2005) cell walls. Phytochemistry 44, 1019–1030.
Oxaziclomefone, a new herbicide, inhibits 21. Miller J. G., Farkaš, V., Sharples, S. C., and
wall expansion in maize cell-cultures without Fry, S. C. (2007) O-Oligosaccharidyl-1-amino-
affecting polysaccharide biosynthesis, xyloglu- 1-deoxyalditols as intermediates for fluores-
can transglycosylation, peroxidase action or cent labelling of oligosaccharides. Carbohydrate
apoplastic ascorbate oxidation. Annals of Research 342, 44–54.
Botany 96, 1097–1107.
 Chapter 5

Carbohydrate Gel Electrophoresis

Florence Goubet, Paul Dupree, and Katja Salomon Johansen

Abstract
Polysaccharide analysis using carbohydrate gel electrophoresis (PACE) relies on derivatization of the
reducing ends of sugars with a fluorophore, followed by electrophoresis under optimized conditions in
polyacrylamide gels. PACE is a sensitive and simple tool for studying polysaccharide structure or quantity
and also has applications in the investigation of enzyme specificity.

Key words: Oligosaccharide, Monosaccharide, Polysaccharide, Pectin, Hemicellulose, Enzyme

1. Introduction

The PACE method is both simple and robust. It involves two

main steps:
1. Conjugation of a fluorophore onto the reducing end of a
sugar using different types of fluorophores depending on the
sugars under study. In the case of a highly negatively charged
sugar, the fluorophore used is uncharged (e.g. 2-aminoacri-
done; AMAC) whereas in the case of partially charged or
neutral sugars, the fluorophore used is charged (e.g. 8-amin-
onaphthalene-1,3,6-trisulfonic acid; ANTS).
2. Electrophoresis at high voltage in thin polyacrylamide gels.
PACE method can provide information on both polysaccha-
ride structure and substrate specificity of carbohydrate active
enzymes.
The main advantages of the method are:
●● The equipment required is not expensive and is easy
to use.

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_5, © Springer Science+Business Media, LLC 2011

81
82 Goubet, Dupree, and Salomon Johansen

●● The sample to be analysed does not need any type of

clean-up prior to the derivatization. Each gel will be used
once so the purity of the samples is not an issue for pre-
serving the equipment.
●● There is a linear relationship between molar amounts of
any sugar and the intensity of the signal. This provides data
for long oligomers that are superior to the data obtained
with the HPLC coupled with amperometric detection.
●● PACE can be used as a separation step prior to analysis
using mass spectrometry (MS).

2. Materials

2.1. Sample 1. 0.1 M ANTS in acetic acid/water (3/17, v/v).

Preparation 2. 0.2 or 1 M NaCNBH3 in DMSO.
3. 50 mM AMAC in acetic acid/DMSO (1.5/18.5, v/v).
4. 0.5 M NaCNBH3 in water.
5. 0.1 M NaCNBH3 in water.
6. 6 M urea in water.
7. Acetic acid/water/DMSO 3/17/20 v/v/v.
8. Monosaccharide or oligosaccharide standards: 1 mM sugars
(e.g. of a range of monosaccharides and oligosaccharides)
5 mL is added to a 0.1–2-mL tube (depending on the reaction
sample) and dried before derivatization.
9. Polysaccharides: 0.5 mg/mL of a range of polysaccharides in
buffer.
10. Enzymes (see Note 1).
11. Incubator.
12. Centrifugal vacuum evaporator.
  Buffers: for enzyme reaction
13. 0.1 M ammonium acetate adjusted to pH 4.5–6 with glacial
acetic acid.
14. 10 mM Tris–HCl, pH 7–9.
15. 0.1 mM Tris–HCl, 1 mM CaCl2, pH 8.

2.2. Electrophoresis Buffers:

1. 0.1 M Tris adjusted to pH 8.2 with boric acid.
2. 0.1 M Tris–HCl pH 8.2.
3. 0.15 M Tris adjusted to pH 8.5 with 0.15 M glycine.
 Carbohydrate Gel Electrophoresis 83

Gels:
4. 29:1 (w/v) Polyacrylamide/acrylamide: N,N-9-methylene­
bisacrylamide from Bio-Rad (Hertfordshire, UK).
5. Gel for analysis of neutral oligosaccharides: 20% (w/v) poly-
acrylamide gel containing 0.5% (w/v) N,N-9-methylene­bisa­
crylamide, 0.1 M Tris–borate pH 8.2. Stacking gel for analysis
of neutral oligosaccharides: 8% (w/v) polyacrylamide, 0.2%
N,N-9-methylenebisacrylamide, 0.1 M Tris–borate pH 8.2.
6. Gel for analysis of acidic oligosaccharides: 25% (w/v)
polyacrylamide gel containing 0.8% (w/v) N,N-9-methylene­
bisacrylamide, 0.1 M Tris–borate pH 8.2. Stacking gel for
analysis of acidic oligosaccharides: 10% (w/v) polyacrylam-
ide, 0.4% N,N-9-methylenebisacrylamide, 0.1 M Tris–borate
pH 8.2.
7. Hoefer SE 660 vertical slab gel electrophoresis apparatus
(Amersham, Buckinghamshire, UK) with 24-cm plates, 0.75-mm
spacer, and well of width 0.25 cm (other equipments can also
be used).
8. Micro-syringes.
9. Standard glass or low-fluorescent Pyrex plates.

2.3. Gel Imaging 1. MasterImager CCD camera system (Amersham).

(One of the Following 2. G:BOX Chemi HR16 (Syngene, Cambridge, UK).
Systems Can Be Used)
3. Standard UV transilluminator.

2.4. Gel Analysis 1. GeneTools software (Syngene, Cambridge UK).

2.5. ext raction from 1. Nanosep system.

a gel 2. MilliQ water.
3. 1% acetic acid.
4. Dialysis tubing.

3. Methods

3.1. Sample Polysaccharides (0.5 mg/mL; 100 mL) are suspended with
Preparation: enzymes (see Note 1) in a suitable buffer at a total volume of
Hydrolysis of Pure 250 mL. The suspension is incubated at room temperature (see
Polysaccharides Using Note 2) for 1 min to overnight (hydrolysis could be longer but
Enzyme Preparations special care must be taken to prevent contamination with fungi or
bacteria). Different buffers can be used with a preference of ammo-
nium acetate, which is volatile and thus leaves no salts behind.
Buffers containing amino group (e.g. Tris buffer) can increase the
background since they can react with the fluorophore.
84 Goubet, Dupree, and Salomon Johansen

For pH 4.5–6, the buffer used is 0.1 M ammonium acetate

adjusted with glacial acetic acid; for pH 7–9, the buffer is 10 mM
Tris–HCl. In order to study lyase activity, 0.1 mM Tris–HCl buffer
pH 8 with 1 mM CaCl 2 is added in the reaction. Controls
without substrates or enzymes are performed under the same
conditions to identify any unspecific compounds present in the
enzymes preparations, polysaccharides or/and labelling reagents.
The reactions are stopped by boiling for 30 min (see Note 3).
The samples are dried using a centrifugal vacuum evaporator.

3.2. Sample To study the polysaccharide architecture of plant cell walls, highly
Preparation: purified and well-defined enzymes with known activity are used
Hydrolysis of Plant to cleave the polysaccharides into smaller fragments. The result-
Cell Wall Materials ing oligomers can only arise from polysaccharides containing the
particular type of bonds which the enzyme can recognize and
cleave. The amount of released products can be quantified or the
pattern of bands can be used as an indicator for the presence or
absence of a particular saccharide in the cell wall sample.
The plant cell wall contains a mixture of polysaccharides in
different ratios (see Note 4). To study them, different protocols
could be used. For the study of polygalacturonan, which is very
abundant in the cell wall, only a small amount of cell wall material
is needed (50 mg; (1)) whereas for the analysis of mannan, which
is a minor component of the cell wall, considerably more cell wall
material is required (0.5 mg; (2)).
The specific compounds may in some cases be inaccessible to
the enzyme. For example, most polygalacturonases are not able
to cleave the glycosidic bonds of highly esterified pectin. In this
case, the accessibility of the enzyme can be improved by the
removal of the methyl groups by pre-treatment either with pectin
methyl esterase or by incubation of the cell wall material in an
alkaline solution (1). Another example is xylan, which can be
esterified and/or closely bound to the cell wall. To eliminate the
esterification and partially solubilize it a highly concentrated
NaOH solution can be used (3).
The samples are dried using a centrifugal vacuum evaporator.

3.3. Analysis 1. Derivatization is carried out in the tubes containing dried

of Neutral polysaccharides, oligosaccharides, or monosaccharides. ANTS
Oligosaccharides are prepared in acetic acid/water (3/17, v/v) at a final con-
Derivatized with ANTS centration of 0.1 M (made freshly or stored at −20°C for at
least 6 months). NaCNBH3 (0.1 M, made freshly and used
immediately, toxic) is solubilized in DMSO (in a fume hood)
for ANTS derivatization.
2. To each dry sample, 5 mL of ANTS solution and 5 mL of the
appropriate NaCNBH3 solution are added. The volume can
be slightly adjusted if large quantities of cell wall material is
used and a very low amount of oligosaccharides produced
 Carbohydrate Gel Electrophoresis 85

(e.g. to detect mannan in Arabidopsis, 0.5 mg of cell wall

material is needed and the oligosaccharide production is very
low (2)). In this case, more solvent can be added with a simi-
lar ratio (acetic acid/water/DMSO; 3/17/20 v/v/v) to
keep the compounds in suspension.
3. The reagents are briefly mixed (using a vortex), centrifuged
and incubated at 60°C overnight. We previously used 1 M of
NaCNBH3 and incubated at 37°C; this condition is optimal
for the study of most of the oligosaccharides. However, there
is a poor recovery of oligosaccharides containing glucosamine
using these conditions. To increase their labelling, a decrease
of NaCNBH3 concentration, to 0.2 M, and incubation at
60°C is required. Using these conditions, all types of oligo-
saccharides are derivatized similarly.
4. The solution is dried in a centrifugal vacuum evaporator for
2 h at 40°C (avoid using high temperatures that could increase
the background). The derivatized sugars are resuspended in
100 mL of 6 M urea and stored before use at −20°C, and are
stable for at least 6 months (the background signal can
increase afterwards).
5. Samples (0.5–4 mL depending of the sugar concentration) are
loaded to the gel using micro-syringes (see Note 4). In all
cases, an Hoefer SE 660 vertical slab gel electrophoresis appa-
ratus (Amersham, Buckinghamshire, UK) is used with 24-cm
plates, 0.75-mm spacer, and well of width 0.25 cm. Standard
glass or low-fluorescence Pyrex plates is used. Electrophoresis
is performed at 10°C in all cases to avoid any heating.
The 20% (w/v) polyacrylamide gel contained 0.5% (w/v)
N,N-9-methylenebisacrylamide with a stacking gel (2 cm)
of 8% (w/v) polyacrylamide and 0.2% (w/v) N,N-9-
methylenebisacrylamide (see Note 5); both gels are made in
0.1 M Tris–borate pH 8.2 (see Note 6). The gels are cast and
cooled at least 1 day before they are to be used in order to
allow complete polymerization of the acrylamide. The gel is
then stored overnight at 4°C so that it will be cold and ready
to use. Smaller electrophoresis equipment can also be used
but some oligosaccharides may be less well separated (4).
6. The electrophoresis buffer system (cooled at 10°C) is 0.1 M
Tris adjusted to pH 8.2 with boric acid (Tris–borate; see Note 6).
The samples are electrophoresed first at 200 V for 20 min and
then at 1,000 V for 90 min. The buffer can be used several
times (see Note 7).

3.4. Analysis of Acidic 1. AMAC is prepared in acetic acid/DMSO (1.5/18.5, v/v) at

Oligosaccharides 50 mM final concentration (made freshly to reduce the back-
Derivatized with AMAC ground). NaCNBH3 (0.5 M; made freshly and used immedi-
ately; toxic) is solubilized in water. To each dry sample, 5 mL
86 Goubet, Dupree, and Salomon Johansen

of AMAC solution and 5 mL of the appropriate NaCNBH3

solution were added. The reagents are mixed, centrifuged,
and incubated at 37°C overnight. The solution is dried in a
centrifugal vacuum evaporator for 2 h at 40°C. The deriva-
tized sugars are resuspended in 100 mL of 6 M urea.
2. To have good derivatization and better gel visualization,
when the polysaccharide studied is contained in the cell wall
at a low level, more derivatization solvent (to recover the
material produced) and a lower concentration of urea are
used. The samples can be stored at least 1 month at −20°C.
Long-term storage will create background in the samples.
3. Samples (0.5–4 mL depending on the sugar concentration)
are loaded to the gel using micro-syringes (see Note 4). In all
cases, an Hoefer SE 660 vertical slab gel electrophoresis appa-
ratus (Amersham, Buckinghamshire, UK) is used with 24-cm
plates, 0.75-mm spacer, and well of width 0.25 cm. Standard
glass or low-fluorescence Pyrex plates is used. The 25%
(w/v) polyacrylamide gel contained 0.8% (w/v) N,N-9-
methylenebisacrylamide with a stacking gel (2 cm) of
10% (w/v) polyacrylamide and 0.4% (w/v) N,N-9-
methylenebisacrylamide (see Note 5); both gels are made in
0.1 M Tris adjusted to pH 8.2 with HCl (Tris–HCl). The gels
are cast and cooled at least 1 day before they are used in order
to allow complete polymerization of the acrylamide. The gel
is then stored overnight at 4°C ready to be used. The acryl-
amide percent can be adjusted to study all oligosaccharides as
described in Goubet et al. (5) (see Note 8).
4. The electrophoresis buffer system (cooled at 10°C) is 0.1 M
Tris–HCl pH 8.2 as the anode reservoir buffer and 0.15 M
Tris adjusted to pH 8.5 with 0.15 M glycine as the cathode
reservoir buffer (see Note 6). The samples are electrophore-
sed first at 200 V for 20 min and then at 1,000 V for 2 h. The
buffer can be used several times (see Note 7).

3.5. Gel Imaging Different systems can be used to image the gels:
(a) Gels can be scanned using a MasterImager CCD camera system
(Amersham) with an excitation filter at 400 nm and a detec-
tion filter at 530 nm. Exposure time is optimized to increase
sensitivity without saturating the intense bands. An image of
the gel (resolution 100 mm) can be obtained and exported
into a 16-bit file to be quantified.
(b) A G:BOX Chemi HR16 (Syngene, Cambridge, UK) can also
be used. The stacking gel is removed and a small amount of
water is added onto each gel prior to imaging to flatten them
out and reduce the wrinkling. The gels are then transferred to
a G:BOX Chemi HR16 for imaging. Since the emission peak
 Carbohydrate Gel Electrophoresis 87

for ANTS is 356 nm and for AMAC is 420 nm the gels are
imaged using short wavelength UV, with UV and short pass
emission filters and without neutral fielding (http://www.
syngene.com/PACE_Intl_Labmate_Article.pdf).
(c) The gel could also be visualized using a standard UV transil-
luminator (wavelength 360 nm), but this has been found to
be less sensitive than the use of the MasterImager, particularly
in the case of ANTS derivatization (6, 7).

3.6. Oligosaccharide All oligosaccharides appear as one clear band (Fig. 1) except for
Separation glucosamine and oligo-glucosamines (oligo-chitosans). For these
last compounds, each sugar is represented by two bands (data not
shown). Compounds that can not be separated are cellobiose and
glucose as shown in Fig. 1. However, using different polyacryl-
amide conditions, those two compounds can be separated (8).

Fig. 1. Different oligosaccharides were derivatized with ANTS and separated in polyacrylamide
gel. There is oligosaccharide separation by the composition, size and also by the glycosidic
bond. Xyloglucan oligosaccharides (XXXG (DP 7), XXLG/XLXG (DP8) and XLLG (DP 9)), where
G is an unsubstituted glucose residue, X is a xylose-substituted glucose residue, and L is a
galactosylxylose-substituted glucose residue (14). The number close to each band is the DP
for each of the oligosaccharides – colour coded by type of oligosaccharide.
88 Goubet, Dupree, and Salomon Johansen

3.7. Gel Analysis Quantification is performed using GeneTools software (Syngene,

Cambridge UK), using rolling ball background detection.
Standards (single or multiple) are run in each gel to obtain a stan-
dard curve for quantification of sugars in the samples. Derivatized
sugars have a linear response between the concentration and sig-
nal level. One point to note is that for ANTS derivatization, the
signal can go through zero as there is low background; however
with AMAC derivatization, the signal can not go through zero
due to higher gel background. The consequence is that quantifi-
cation of ANTS derivatized oligosaccharides can be done using
only one standard; however, for AMAC, a minimum of two stan-
dards is required to determine the background level. For accuracy
in both cases, more standards will be needed. To obtain accurate
quantification, pure standards need to be used. The monosaccha-
rides are highly pure whereas not all commercial preparations of
oligosaccharides are, as shown in Goubet et al. (6). Man and
(Man)3 or GalU and (GalU)3 can be used as standards for quanti-
fication for either ANTS or AMAC derivatized samples (1). The
band intensity is independent of the sugar tested except for oli-
gochitosan. More than one standard gives a greater accuracy of
quantification.

3.8. Extraction 1. To determine the identity of oligosaccharides in specific

of Derivatized bands, a preparative gel with multiple adjacent lanes loaded
Oligosaccharides with 6 mL of derivatized oligosaccharides is prepared.
and MS Analysis 2. Bands can be excised while briefly viewing the gel under a UV
transilluminator (wavelength 360 nm), and suspended in
1 mL of milliQ water.
3. To extract the oligosaccharides, acrylamide gel slices are
partially crushed and subjected to three cycles of freeze/
thawing. In the case of purification of esterified oligosaccha-
rides, the pH needs to be slightly acid to reduce any demethy-
lesterification process (9), so acetic acid solution (1%) replaces
water to extract the compounds.
4. To eliminate any polyacrylamide fragments, the samples are
filtered using the Nanosep system (MWCO of 10,000 Da;
Pall, East Hills, NY) and then dialyzed against water using
dialysis tubing (MWCO of 500 Da – note that derivatized
monosaccharides will be lost).
5. The solution is dried and the pellet is suspended in
10–20 mL of water. Aliquots of the specific derivatized-
Me-OGA samples are loaded onto a gel to estimate their
quantity.
6. These samples are then analyzed by different MS configurations
as described by Goubet et al. (9).
 Carbohydrate Gel Electrophoresis 89

4. Results

PACE can resolve many carbohydrates from each other based on

their size and composition and function of the glycosidic bond
(Fig. 1). Already in monosaccharide form, some isomers can be
separated from each other (e.g. glucose (Glc) from galactose
(Gal)). The oligosaccharides can also be separated from each
other depending on their composition. For example, a dimer of
arabinose (Ara) is clearly separated from the dimer of xylose (Xyl);
and in this case both monosaccharides are pentose. Another
example is that dimer of Gal which is separated from the dimer of
mannose (Man) and in this case, the monosaccharide unit for
both dimers is a hexose.
The glycosidic bond can also play a role in their separation. In
the example given in Fig. 1, a b 1-3 oligoglucan can be separated
from a b 1-4 oligoglucan with the same degree of polymerization.
Similar observations are made for large oligosaccharides but to
achieve better separation, different compositions of polyacrylam-
ide gels need to be used. The same oligosaccharides with substi-
tutions in different places on the backbone can also clearly be
separated from each other as shown in Goubet et al. (9).
Similar observations were made studying charged oligosac-
charides as described in Goubet et al. (5) and shown in Fig. 2.
Carrageenans are polysaccharides of repeating disaccharide units
of 3-linked b-d-galactopyranose and 4-linked a-d-galactopyra-
nose. Three main structural groups are defined based on the
­presence or absence of a-d-anhydrogalactose in place of a-d-ga-
lactopyranose and on the position of sulphate groups as described
in Liners et al. (10). Oligosaccharides of iota- and kappa-carra-
geenans (kindly provided by F. Liners and P. Van Cutsem,
University of Namur, Belgium) were analysed by PACE. As previ-
ously shown, charged oligosaccharides have a size-dependent
“turning point” (5). As a consequence of this turning point, some
large oligosaccharides could migrate to similar positions as smaller
ones. In Fig. 2, two kappa-carrageenans with degree of polymer-
ization (DP) of 4 and 6 respectively co-migrated under the condi-
tions used. To separate them, different polyacrylamide conditions
can be used (5). The two types of carrageenan are separated from
each other so that PACE can be very useful to study carragenan
structure and also some slight changes of structures as already
been shown for the methylation of polygalacturonan (9).
Recently, separation of saturated oligosaccharides versus
unsaturated ones has also been shown possible using PACE tech-
nology (11). This technique could be used for rapid screening of
the mode of action of hydrolase and lyase activities.
90 Goubet, Dupree, and Salomon Johansen

Fig. 2. Analysis of kappa- and iota-oligocarrageenans by PACE – derivatization was

carried out using the AMAC fluorophore. The oligosaccharides separated by the size and
the composition. The number close to each band is the DP of each oligosaccharide.
“Mix” is a partial hydrolysis of the corresponding polysaccharides. Please note that the
standards contain additional compounds than the ones indicated. For example, the
kappa-carrageenan DP4 contains also two minor compounds that one has been identi-
fied as kappa-carrageenan DP3. The other one is unknown and could be a DP2. The
number close to each band is the DP for each oligosaccharide – colour coded by type of
oligosaccharide. The position of the kappa-penta-carrageenan (DP5) has been indicated
as a possible but it has not confirmed.

5. Conclusion

PACE is a versatile method for the separation and detection of

any kind of carbohydrate with a reducing end. The power to
separate each carbohydrate coupled with its high sensitivity
makes PACE an attractive alternative to HPLC, TLC and MS
based analysis.
Using this method, polysaccharide structure and quantity and
enzyme characteristics can be obtained using simple equipment.
 Carbohydrate Gel Electrophoresis 91

6. Notes

1. If the enzymes are contained in a mixture, pure polysaccharides

are required to study the enzyme characteristics as described
in Phalip et al. (12). Many polysaccharides and oligosac-
charides are available from Megazyme (http://www.mega-
zyme.com/) and Sigma Aldrich (http://www.sigmaaldrich.
com/sigma-aldrich/home.html) to study the enzyme
specificity.
2. Higher temperatures may be appropriate depending on the
properties of enzymes used (e.g. their thermostability (13)).
3. To avoid any product modification (e.g. degradation by some
enzymes) during derivatization, the reaction is stopped. This
will induce protein precipitation which is not an issue for
PACE analysis except during loading. To avoid any problems
with the micro-syringe, suspend the compounds fully before
loading.
4. Cell wall mass is accurately measured by using a cell wall sus-
pension (i.e. 0.5 mg/mL; homogenized using a glass potter)
and an aliquot is taken for the analysis. Before taking an
aliquot, the suspension is well mixed since the cell wall easily
sediments.
5. For acrylamide polymerization, TEMED and APS solutions
are used. TEMED solution can be bought ready to use and
can be stored at room temperature for at least 1 year. A solu-
tion of APS can be stored for couple of months at 4°C, but if
polymerization starts to take a longer time, a fresh solution
should be made.
6. A 10× stock solution of the following buffers can be made
and diluted when needed. The Tris–glycine stock solution
can be stored at 4°C for at least 6 months. The Tris–HCl and
Tris–borate solutions can be stored at room temperature for
at least 6 months. Borate can precipitate if stored for too long
or if the storage temperature is too cold.
7. The running solutions can be used several times. Borate salt
can precipitate which leads to deteriorating electrophoresis
materials. A sign of this is that the electrophoresis takes lon-
ger to run, the separation becomes less efficient and the back-
ground increases. Prepare a fresh solution if this situation
occurs.
8. To study charged oligosaccharides, 1–31% (w/v) polyacryl-
amide gel contained 0.5–1.1% (w/v) N,N-9-methylene­
bisacrylamide can be used (higher acrylamide percent can
also be used) (5).
92 Goubet, Dupree, and Salomon Johansen

References

1. Barton, C. J., Tailford, L., Welchman, H., 8. Jackson, P. (1994) The analysis of fluorophore-
Zhang, Z., Gilbert, H. J., Dupree, P., and labeled glycans by high-resolution polyacryl-
Goubet, F. (2006) Enzymatic fingerprinting amide gel electrophoresis. Anal Biochem 216,
of Arabidopsis pectic polysaccharides using 243–252.
PACE-polysaccharide analysis by carbohy- 9. Goubet, F., Ström, A., Quéméner, B., Stephens, E.,
drate gel electrophoresis. Planta 224, Williams, M. A. K., and Dupree, P. (2006)
163–174. Resolution of the structural isomers of partially
2. Handford, M. G., Baldwin, T. C., Goubet, F., methylesterified oligogalacturonides by poly-
Prime, T. A., Miles, J., Yu, X., and Dupree, P. saccharide analysis using carbohydrate gel elec-
(2003) Localisation and characterisation of trophoresis. Glycobiology 16, 29–35.
mannan cell wall polysaccharides in Arabidopsis 10. Liners, F., Helbert, W., and Van Cutsem, P.
thaliana. Planta 218, 27–36. (2005) Production and characterization of a
3. Brown, D. M., Goubet, F., Wong, V. W., phage-display recombinant antibody against car-
Goodacre, R., Stephens, E., Dupree, P., and rageenans: evidence for the recognition of a sec-
Turner, S. R. (2007) Comparison of five xylan ondary structure of carrageenan chains present
synthesis mutants reveals new insight into the in red algae tissues. Glycobiology 15, 849–860.
mechanisms of xylan synthesis. Plant J 52, 11. Phalip, V., Goubet, F., Carapito, R., and
1154–1168. Jeltsch, J. -M. (2009) Plant cell wall degrada-
4. Karousou, E., Porta, G., De Luca, G., and tion with a powerful fusarium graminearum
Passi, A. (2004) Analysis of fluorophore- enzymatic arsenal. J Microbiol Biotechnol 19,
labelled hyaluronan and chondroitin sulfate 573–581.
disaccharides in biological samples. J Pharmac 12. Phalip, V., Delalande, F., Carapito, C.,
Bioch Anal 34, 791–795. Goubet, F., Hatsch, D., Leize-Wagner, E.,
5. Goubet, F., Morriswood, B., and Dupree, P. Dupree, P., Dorsselaer, A. V., and Jeltsch, J. M.
(2003) Analysis of methylated and unmethy- (2005) Diversity of the exoproteome of
lated polygalacturonic acid structure by Fusarium graminearum grown on plant cell
PACE: polysaccharide analysis using carbohy- wall. Curr Genet 48, 366–379.
drate gel electrophoresis. Anal Biochem 321, 13. Palackal, N., Brennan, Y., Callen, N. C., Dupree,
174–182. P., Frey, G., Goubet, F., Hazlewood, G. P.,
6. Goubet, F., Jackson, P., Deery, M., and Healey, S., Kang, Y. E., Kretz, K. A., Lee, E.,
Dupree, P. (2002) Polysaccharide analysis Tan, X., Tomlinson, G. L., Verruto, J., Wong,
using carbohydrate gel electrophoresis V. W. K., Mathur, E. J., Short, J. M.,
(PACE): a method to study plant cell wall Robertson, D. E., and Steer, B. A. (2004) An
polysaccharides and polysaccharide hydro- evolutionary route to xylanase process fitness.
lases. Anal Biochem 300, 53–68. Protein Sci 13, 494–503.
7. Fliegmann, J., Mithöfer, A., Wanner, G., and 14. Fanutti, C., Gidley, M. J., and Reid, G. J. S.
Ebel, J. (2004) An ancient enzyme domain (1996) Substrate subsite recognition of the
hidden in the putative b-glucan elicitor recep- xyloglucan endo-transglycosylase or xyloglucan-
tor of soybean may play an active part in the specific endo-(1→4)-b-d-glucanase from the
perception of pathogen-associated molecular cotyledons of germinated nasturtium
patterns during broad host resistance. J Biol (Tropaeolum majus L.) seeds. Planta 200,
Chem 279, 1132–1140. 221–228.
 Chapter 6

Capillary Electrophoresis with Detection

by Laser-Induced Fluorescence
Andrew Mort and Xiangmei Wu

Abstract
The importance of capillary zone electrophoresis (CZE) has been increasing in use for: structural analysis
of plant cell walls, characterization of enzymes that degrade polysaccharides, and profiling of oligosac-
charides to characterize cell wall mutants. CZE with laser-induced fluorescence detection provides high
separation efficiencies, high speed analysis, with extremely small sample requirements. Here, we describe
the instrumentation we use and the methods for attaching fluorescent labels to oligosaccharides so that
they can be detected.

Key words: Capillary zone electrophoresis (CZE), Laser-induced fluorescence detection,

Oligo­saccharide, Separation, Profiling

1. Introduction

Plant cell walls contain polysaccharides with complex structures

such as xyloglucans and pectins which vary somewhat from species
to species, from cell type to cell type, and even at different locations
within the wall around a single cell (1). With the fairly recent avail-
ability of pure enzymes (2, 3), one can hydrolyze these polymers
into oligomers which can then be separated and characterized.
The oligomers can be labelled at their reducing ends by
reacting them with a fluorescent amine at the aldehyde group to
form an imine which can be selectively reduced with sodium
cyanoborohydride to a stable secondary amine. We use two dif-
ferent aromatic trisulfonated amines; 8-aminonaphthalene-1,3,
6-trisulfonate (ANTS), and 8-aminopyrene-1,3,6-trisulfonate
(APTS). ANTS is quite inexpensive and in our hands gives few
interfering peaks on the CZE trace. However, ANTS absorbs in
the UV range and so, requires a UV laser for excitation. APTS is

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_6, © Springer Science+Business Media, LLC 2011

93
94 Mort and Wu

about 100 times as expensive and gives more interfering signals just
from the reagent. It absorbs blue light and fluoresces green, so there
is little chance of interfering signals from plant compounds. The
sensitivity of detection for APTS-labelled oligomers excited with an
Argon-Ion laser at 448 nm is about 50 times greater than that of an
ANTS-labelled oligomer excited by a 325 nm Helium Cadmium
laser (4). The pyrene ring seems to make the APTS “sticky.” Thus,
the capillary needs cleaning frequently to avoid traily peaks.
Both ANTS and APTS have three negative charges on them
which ensure that all labelled oligomers have a charge of minus
three. We use uncoated capillaries for the electrophoresis with a
fairly high ionic strength buffer at a pH of 2.5. This low pH
ensures that silicic acid groups are protonated so there are no
fixed charges on the walls of the capillary and hence no electro-
osmotic flow. A pH of 2.5 also causes almost complete protona-
tion of uronic acids, thus both neutral and acidic oligosaccharides
should only be charged because of the three sulphonic acid groups
on the fluorescent amine label. Since all oligomers will have the
same net charge, the driving force of the electric field should be
the same for all oligomers. The frictional force on the oligomers
will be the product of the velocity of the oligomer times its fric-
tional coefficient. According to Stokes law, the frictional coeffi-
cient f = 6phr, where h is the viscosity of the medium and r is the
Stokes radius of the oligomer. Thus, the larger the hydrodynamic
radius of the oligomer the slower it will move in the electric field.
Electrophoresis of the labelled oligomers therefore allows one to
distinguish them according to their size.
So, using these methods one can profile the products of an
enzyme digest of cell wall polymers to follow the progress of the
digestion, or can compare the profile of products between wild
type and potential cell wall mutants.
If one labels a pure oligosaccharide, it can then be used to
characterize the mode of action of an enzyme, or to detect the
presence of enzymes that act on the oligomer, even in the pres-
ence of a very complex medium. For example, we labelled an
oligomer from a partial acid hydrolysate of rhamnogalacturonan
and used this to detect and characterize the mode of action of
rhamnogalacturonan lyase in intact cotton plants (5).

2. Materials

2.1. Preparation 1. 1 mg oligosaccharide in 20 mL dH2O: purified oligomers such

of Aminopyrene as xylohexaose (Megazyme International Ireland Ltd., Bray
Trisulfonate- (APTS) Business Park, Bray, Co. Wicklow, Ireland) should be used
Labelled Oligomers (see Notes 1 and 2).
2. 0.1 M APTS (Molecular Probes or Sigma) in 25% acetic acid
(see Note 3).
 Capillary Electrophoresis with Detection by Laser-Induced Fluorescence 95

3. 1 M sodium cyanoborohydride in dimethylsulfoxide (see Note 4).

4. 50 mM ammonium acetate buffer, pH 5.2: HPLC-grade
acetonitrile 3:1 v/v, briefly degassed with a water aspirator.
5. Chromatography materials: HW-40S gel filtration material
column (Toyopearl, Supelco) packed in a 100 × 10 mm stain-
less steel column.
6. Heating block or other reliable heat source set at 80°C with
holes suitable for 500 mL microfuge tubes.
7. Vortex.
8. Microfuge.
9. Speed vac concentrator.
10. Spectrophotometer.

2.2. Labelling with 1. Polysaccharide solution: 10 mg/mL

Aminonaphthalene 2. Internal standard: 1 mg/mL cellobiose or maltose solution.
Trisulfonate (ANTS)
3. Appropriate buffer.
4. 4 mUnits of the appropriate enzyme.
5. Heating block.
6. Pipette.
Labelling reagents:
7. 23 mM ANTS (Molecular Probes) in 3% w/v acetic acid.
8. 1 M solution of sodium cyanoborohydride in dimethylsulfoxide.
9. 500 mL microfuge tubes.
10. Vortex.
11. Microfuge.

2.3. Determination of 1. 1:100 dilution of stock APTS-labelled substrate prepared as

Enzyme Activity above.
In Vitro 2. 1 mU of an appropriate hydrolytic enzyme.
3. Spectrophotometer.
4. Heating block.
5. Standard mixture of oligomers prepared either by specific
enzyme digestion or TFA hydrolysis of an appropriate poly-
saccharide e.g. for determination of xylanase activity the oli-
gomers should be generated from xylan.

2.4. In Situ Detection 1. 6 pmol/mL APTS-labelled substrate, prepared as above.

of Enzyme Activity 2. Gas-tight syringe: 10 mL (1701 RNFS Hamilton Co., Reno,
Using APTS NV, USA) fitted with a needle made of a 15-cm section of
0.17-mm o.d. fused silica capillary (Alltech Associates, Inc.,
Deerfield, IL, USA).
3. Plant of interest.
96 Mort and Wu

4. Temperature-controlled plant growth chamber.

5. 125 mL Erlenmeyer flask.
6. Extracting solvent: 25 mM sodium acetate buffer, pH 5.2,
ice-cold.
7. Water aspirator.
8. Paper towels.
9. Tissue.
10. 2-mL Reacti-Vial (Supelco, Inc., Bellefonte, PA, USA).
11. Centrifuge.

2.5. Capillary Capillary columns:

Electrophoresis 1. Column for APTS-labelled oligomer separation: A fused-silica
Columns and Buffers capillary (TSP050375, Polymicro Technologies, http://www.
for APTS and ANTS- polymicro.com) of internal diameter 50 mm and length 31 cm
Labelled Oligomer with a 3–4 mm window burned into the plastic coating 5.7 cm
Separation from the anode end. The window is formed by resting the
capillary on a glass window etching device (pEZ-Window,
J & W Scientific, Folsom, CA, USA) on a heating plate and
covering the capillary at that point with a drop of concentrated
sulphuric acid. The capillary is carefully inserted into the plastic
tube for circulation of the coolant and assembled into the car-
tridge as instructed by the manufacturer (see Note 5).
2. Column for ANTS-labelled oligomer separation: A fused-
silica capillary (TSP050375, Polymicro Technologies, http://
www.polymicro.com) of internal diameter 50 mm and length
50 cm with a 3–4 mm window is burned into the plastic coat-
ing at 30 cm from the anode end as described above.
3. Running buffer: 0.1 M sodium phosphate, pH 2.5. This buffer is
made by slowly adding 0.1 M phosphoric acid to 0.1 M sodium
monophosphate until the pH just reaches 2.5 (see Note 6).
4. Rinsing buffer: 1 M NaOH.

2.6. Capillary A custom built CZE system consisting of a high voltage power
Electrophoresis System supply, helium cadmium laser, optical system built on an inexpen-
for ANTS-Labelled sive microscope, an intensified CCD camera for light detection,
Oligosaccharides and a computer-controlled variable-light-attenuator. A complete
description of the system is available (6) (Fig. 1).

2.7. Capillary 1. Instrument: Capillary zone electrophoresis (CZE) Biofocus-2000

Electrophoresis (Bio-Rad laboratories, http://www.bio-rad.com) CZE appara-
System for APTS- tus with laser-induced fluorescence detection.
Labelled 2. Gas pressure: for injection 4.5 lb/in.2 of helium pressure for
Oligosaccharides 0.22 s and for rinsing 80 lb/in.2.
3. Excitation: 488 nm light from a 5 mW argon ion laser.
 Capillary Electrophoresis with Detection by Laser-Induced Fluorescence 97

Fig. 1. Pictures showing the major components of the custom built CZE apparatus. (a) The complete setup excluding the
computer and controller boxes for the attenuator and camera. Upper right, Helium Cadmium Laser. Upper centre, intensified
CCD camera mounted on top of an inexpensive microscope. Middle left, micro-switch to ensure that high voltage is disabled
if the door to the safety and light excluding enclosure is not in place. Centre, microscope. Lower centre left and right, plastic
holders for the 1.5 mL microfuge vial electrode chambers. Bottom, high voltage power supply. (b) Close-up view of the holder
and X–Y positioner for alignment of the capillary under the microscope objective. (c) Close-up of the capillary and the fibre
optic held in a stainless steel cannula attached to an eccentric brass nut for movement of the fibre up and down so that it can
be aimed directly at the centre of the capillary.

4. Fluorescence emission collection: emission collected through

a 520 nm narrow band pass filter.
5. Electrophoresis conditions: 15 kV/70–100 mA with the cathode
at the inlet; controlled temperature of 20°C.
6. Data processing: BioFocus 2000 System operating software
and BioFocus system integration software.

3. Methods

3.1. Preparation of 1. Dissolve 1 mg of oligosaccharide in 20 mL of water in a 500-mL

Aminopyrene microcentrifuge tube.
Trisulfonate- (APTS) 2. Add 2 mL of 0.1 M APTS in 25% acetic acid and 20 mL
Labelled Substrate of a 1 M solution of sodium cyanoborohydride in
dimethylsulfoxide.
3. Vortex-mix well and centrifuge briefly to bring down the liquid,
cap the vial tightly.
98 Mort and Wu

4. Heat for 60 min at 80°C.

5. After the mixture has cooled to room temperature dilute to
about 200 mL with water and pass through a Toyopearl
HW-40S (100 × 10 mm) gel filtration column eluted with
25% acetonitrile and 75% 50 mM ammonium acetate buffer
v/v, pH 5.2 at 1 mL/min (see Note 7). The labelled sub-
strate elutes at around 5–6 min and the salts and excess label-
ling reagents around 10 min. The fractions containing the
labelled substrate and free-label can be detected using a fluo-
rescence or visible detector. However, at this scale one can
observe with the naked eye where each fraction is.
6. Pool the labelled substrate fractions and evaporate them to
dryness in a speed vac concentrator.
7. Re-dissolve the labelled oligomer in 100 mL of water and
store frozen.
8. The concentration of APTS-labelled substrate can be deter-
mined based on the extinction coefficient of 17,100 M/cm at
456 nm.

3.2. Determination 1. Prepare an appropriate labelled substrate as described above.

of the Mode of Action 2. Estimate the amount of enzyme needed to degrade the sub-
of an Enzyme strate over a period of several hours.
3. Incubate an amount of substrate, which will give a total fluo-
rescence intensity of at least 10–50 RFU with the enzyme in
an appropriate buffer, taking aliquots after various lengths of
time and stopping the reaction by heating at 80°C for 10 min.
For xylohexaose, 1 mL of a 1:100 dilution of the stock solu-
tion of labelled substrate in a 25-mL incubation with around
one micro-unit of enzyme is about right.
4. Generate a standard mixture of labelled oligomers for
­comparison with the enzyme-produced products e.g. for
xylanase take a xylan and digest it for a short time with a
xylanase, or hydrolyze it for a short time in trifluoroacetic
acid and then label the products as described above (see
Note 8).

3.3. Detection of Since APTS absorbs blue light and fluoresces green light, the
Enzyme Activity CZE detection system is oblivious to all imaginable plant
in a Complex Medium ­components. Thus, if one adds an APTS-labelled substrate to a
Such as an Intact complex system and then analyses the result using CZE with
Plant Using APTS laser-induced fluorescence detection, only the undigested
­substrate and any degradation products which still contain the
label will be visible.
1. Prepare an appropriate labelled substrate and dilute it so that
it has a concentration of about 30 pmol/5 mL.
 Capillary Electrophoresis with Detection by Laser-Induced Fluorescence 99

2. Inject 5 mL of the solution into the intercellular spaces of an

intact cotton cotyledon using a gas-tight 10 mL-syringe fitted
with a needle made of a 15-cm section of 0.17-mm o.d. fused
silica capillary.
3. Return the plant to the growth chamber for the desired
incubation time.
4. Cut the cotyledon from the plant and place it in an Erlenmeyer
flask containing about 30 mL of ice-cold extracting solvent
(25 mM sodium acetate buffer, pH 5.2).
5. Apply vacuum for 2 min from a water aspirator.
6. Release the vacuum to cause infiltration of the cotyledon’s
intercellular spaces by the extracting solvent.
7. Repeat the evacuation and vacuum release 2–3 times.
8. Transfer the cotyledon to paper towels and blot with tissue.
9. Roll the cotyledon in a conical shape and put into a 2-mL
Reacti-Vial reaching only half way to the bottom to avoid
contact of the cotyledon with the intercellular wash fluid during
centrifugation.
10. Centrifuge at 1,500 × g for 15 min. About 0.3 mL of intercel-
lular wash fluid per cotyledon will be collected.
11. Analyse the products by CZE and identify them by compari-
son to standard labelled oligomers.

3.4. Separation 1. Load the samples, running buffer, rinsing buffer, NaOH, and
of APTS-Labelled water in the inlet carousel according to the “configuration”
Compounds on stored in the computer memory. Load a vial for waste and
Capillary Zone running buffer in the outlet carousel according to the
Electrophoresis (CZE) “configuration” to be used.
2. Select a “method.” For following degradation of xylohexaose
we use:
(a) Inlet buffer: 0.1 M phosphate, pH 2.5.
(b) Outlet buffer: 0.1 M phosphate, pH 2.5.
(c) Cartridge temperature 20°C.
(d) Inverse polarity.
(e) Voltage 15 kV.
(f) Current limit 100 mA.
(g) Run time 8 min.
(h) Pre-run steps: high pressure rinse with NaOH for 30 s
High pressure rinse with wash buffer for 60 s.
Inject 2 psi*s.
Water dip to prevent carryover between samples.
Run.
100 Mort and Wu

3.5. Profiling and Time The range of products produced by digestion of a polysaccharide
Course Analysis of substrate with enzymes can be followed over time by taking small
Oligosaccharides aliquots of the digestion mixture at suitable time points and
Produced by Enzyme derivatizing them with ANTS for subsequent analysis by CZE. To
Degradation make the time course relatively quantitative one can add a ­constant
of Polysaccharides amount of a commercially available disaccharide such as maltose
or cellobiose as an internal standard for comparison of peak
heights or areas.
3.5.1. Labelling with
1. Incubate 25 mL of a 10 mg/mL solution of the polysaccharide
Aminonaphthalene
in an appropriate buffer with 4 mUnits of enzyme at the
Trisulfonate (ANTS)
­temperature optimum for the enzyme.
2. Take duplicate 1 mL aliquots at 0 time, 15, 30 min, 1, 2 and 4 h.
3. Put the aliquots in 500 mL microfuge tubes along with 1 mL
internal standard, 1 mL cyanoborohydride solution, and
10 mL ANTS solution. Mix well and then centrifuge briefly to
collect the liquid in the tip of the tube.
4. Heat for 1 h at 80°C.

3.5.2. Separation of 1. Turn on power to the laser, the camera, the attenuator
ANTS-Labelled Compounds controller, and the serial to parallel interface box.
by Capillary Zone 2. Start the CZE control programme on the computer.
Electrophoresis (CZE)
3. Set the attenuation to 1 and view the image from the
camera.
4. Draw a rectangle around the area that should be used for
detection of the fluorescence.
5. Move the rectangle to the top right hand corner of the image
and click on the background button.
6. Move the rectangle back over the image of the lumen of the
capillary.
7. Designate a file name and path for data storage.
8. Rinse the capillary with running buffer, or with NaOH and
then running buffer to remove contaminants from the walls of
the capillary. We use a 250 mL gas tight syringe with a replace-
able blunt ended needle adapted to press fit onto the capillary
with a 1 cm piece of tubing fitting over the needle and a piece
of Teflon tubing with ID of 360 nm pushed inside it.
9. Inject the sample by dipping the inlet end of the capillary into
the microfuge tube and elevating it 15 cm above the outlet
end of the capillary.
10. After 6 s lower the inlet to the same level as the outlet and
quickly transfer the inlet end of the capillary to the cathode
well (a 1.5 mL microfuge tube with two holes poked in its lid
with a push pin, one for the inlet platinum wire electrode and
the other for the capillary).
 Capillary Electrophoresis with Detection by Laser-Induced Fluorescence 101

11. Turn on the high voltage to 18 kV with the negative ­electrode

at the inlet. The current should be about 60 mA.
12. Start the run on the computer.
13. After the desired run time, turn off the high voltage and click
stop on the computer. The data will automatically be saved.
Figure 2 shows an example of the time course of reaction of endo-
polygalacturonase with pectic acid.

Fig. 2. A time course of degradation of pectic acid with endopolygalacturonase obtained from a Pichia clone expressing
open reading frame AN 8327. (a) No enzyme, just substrate plus cellobiose as an internal standard. The peak at 2.5 min
is from the great excess of free ANTS which had not found sugar to attach to. The peak at 3.5 min is from the labelled
cellobiose. Note the very low levels of oligomers stretching out until around 11 min where the peaks all fuse together.
(b) After 3 min of incubation, a series of oligomers is observed with low fluorescent intensity. Remember only the end of
the oligomers are labelled so, large oligomers only give a low fluorescence yield per mg. (c) After 10 min shorter oligom-
ers predominate. After longer times of incubation (d, e) only the monomer, dimer, and trimer of GalA remain. Each GalA
oligomer gives rise to two peaks because of the tendency of the labelled GalA residue to form a lactone between the
carboxyl group and C-3. Thus, part of the oligomer is in the lactone form and part as the free acid. This pattern of degra-
dation reflects a random attack of the polymer by the enzyme.
102 Mort and Wu

4. Notes

1. Megazyme has the widest selection of purified oligomers

from plant polysaccharides. Many potential substrate oligo-
saccharides are not commercially available, so must be gener-
ated and purified by the individual investigator.
2. An example of purification of rhamnogalacturonan oligomers
for investigation of rhamnogalacturonase and rhamnogalac-
turonan lyase is given in (5).
3. We add 200 mL of 25% acetic acid to the 10 mg vial, we
receive the reagent from the manufacturer to avoid the inevi-
table losses that would occur during transfer of the powder.
4. Use caution when preparing and handling this solution. The
cyanoborohydride breaks down slowly to produce HCN and
dimethylsulfoxide can carry dissolved substances through
skin.
5. We have found the manufacturer’s cartridges to be flimsy.
They break easily when you tighten the seal at the inlet end to
prevent arcing between the inlet electrode and the coolant
compartment. A replacement cartridge incorporating the
optical bench section of the original cartridge can be made by
a competent machine shop and is much sturdier.
6. Too much phosphoric acid makes the current through the
capillary at the suggested operating voltage too high.
7. The acetonitrile is necessary to keep the APTS from adsorbing
to the gel filtration material.
8. Sub mg amounts of oligosaccharide mixtures can be deriva-
tized in smaller volumes e.g. 5 mL sample in water, 0.5 mL
APTS solution and 5 mL cyanoborohydride solution.

References
1. Willats, W. G. T., Knox, J. P., and Mikkelsen, 1,4,6-trisulfonate-derivatized sugars by
J. D. (2006) Pectin: new insights into an old capillary electrophoresis with laser-induced
polymer are starting to gel. Trends Food Sci fluorescence detection. Anal Chem 67,
Technol 17, 97–104. 2239–2245.
2. Bauer, S., Vasu, P., Persson, S., Mort, A. J., and 5. Naran, R., Pierce, M. L., and Mort, A. J.
Somerville, C. R. (2006) Development and (2007) Detection and identification of rham-
application of a suite of polysaccharide-degrad- nogalacturonan lyase activity in intercellular
ing enzymes for analyzing plant cell walls. Proc spaces of expanding cotton cotyledons. Plant
Nat Acad Sci USA 103, 11417–11422. J 50, 95–107.
3. De Vries, R. P., and Visser, J. (2001) 6. Merz, J. M., and Mort, A. J. (1998) A com-
Aspergillus enzymes involved in degradation puter-controlled variable light attenuator for
of plant cell wall polysaccharides. Microbiol protection and autoranging of a laser-
Mol Biol Rev 65, 497–522. induced fluorescence detector for capillary
4. Evangelista, R. A., Liu, M., and Chen, F. zone electrophoresis. Electrophoresis 19,
(1995) Characterization of 9-aminopyrene- 2239–2242.
 Chapter 7

Monoclonal Antibodies, Carbohydrate-Binding Modules,

and the Detection of Polysaccharides in Plant Cell Walls
Cécile Hervé , Susan E. Marcus , and J. Paul Knox

Abstract
Plant cell walls are diverse composites of complex polysaccharides. Molecular probes such as monoclonal
antibodies (MABs) and carbohydrate-binding modules (CBMs) are important tools to detect and dissect
cell wall structures in plant materials. We provide an account of methods that can be used to detect cell
wall polysaccharide structures (epitopes) in plant materials and also describe treatments that can provide
information on the masking of sets of polysaccharides that may prevent detection. These masking
­phenomena may indicate potential interactions between sets of cell wall polysaccharides, and methods to
uncover them are an important aspect of cell wall immunocytochemistry.

Key words: Carbohydrate-binding module, Cell wall immunocytochemistry, Immunofluorescence

microscopy, Pectate lyase, Plant cell walls, Monoclonal antibody

1. Introduction

Plant cell walls are diverse composites of structurally complex

polysaccharides, phenolic polymers, proteins, ions and water. The
macromolecular polysaccharide and phenolic components gener-
ate the bulk of cell walls and are major contributors to cell wall
properties and functions. Cell wall polysaccharides are highly
diverse and vary both in relation to cell wall ultrastructures, cell
status and taxonomy (1). The majority of cell wall polysaccharides,
including xyloglucan and xylan hemicelluloses and the pectic poly-
saccharides, contain a range of structural variants that appear to be
integral to polymer properties and cell wall functions. To fully
understand how the diverse cell wall structures function in cell
processes and organ mechanics and respond to environmental
impacts, it is important to assess the presence of not only specific

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_7, © Springer Science+Business Media, LLC 2011

103
104 Hervé, Marcus, and Knox

polymers but also specific configurations of polymers in relation to

individual cell wall architectures, cell types, and cell status.
One of the best ways to detect and assess the presence of
polysaccharides in plant materials is by the use of tagged proteins
with specific recognition capacities. Currently, most proteins used
for cell wall polysaccharide recognition are rodent monoclonal
antibodies (MABs) and carbohydrate-binding modules (CBMs).
In the case of complex carbohydrate polymers, purification of
immunogens in sufficient amounts for antibody production is
often a limiting step. In nature, CBMs of microbial and plant
polysaccharide hydrolases are used for carbohydrate recognition.
These protein domains encompass a large collection of sequences
and a wide range of ligand specificities (2). When produced as
separate recombinant his-tagged modules, they can be readily
adapted to antibody-style procedures including immunocy-
tochemistry approaches (3). Large sets of these MAB and CBM
probes directed to cell wall polysaccharides are now available.
This chapter will focus on general factors relating to the use of
these probes to detect polysaccharides in conjunction with cell
wall imaging using immunocytochemistry with an emphasis on
immunofluorescence procedures.
Cell wall immunocytochemistry has entered an exciting and
intriguing phase with the recent discovery that polysaccharide
epitopes present in cell walls may not be directly detectable due to
the presence of other polymers – perhaps indicating intimate
associations. To date, this has been demonstrated for pectic
homogalacturonan (HG) polysaccharides obscuring or masking
xyloglucan and xylan polysaccharides in primary cell walls (4, 5).
Combining the use of molecular probes with specific enzymatic
treatments provides a more nuanced understanding of the occur-
rence of polysaccharides in cell walls. Treatments of plant materi-
als to explore these phenomena and to uncover masked epitopes
of hemicelluloses are described below.
All MAB and CBM probes can also be used to detect polysac-
charides (when isolated from plant cell walls) using microtitre
plates, nitrocellulose, and microarray substrates. The general
principles of MAB/CBM detection strategies, staged incubations,
importance of washing steps, etc. in these cases have been dis-
cussed elsewhere (6, 7).

2. Materials

2.1. Molecular Probes 1. A large range of rodent MABs that recognise plant cell wall
polysaccharides and proteoglycans is now available as detailed
at the online sites of Biosupplies (www.biosupplies.com.au),
Carbosource Services (www.carbosource.net) and PlantProbes
(www.plantprobes.net). Biosupplies and Carbosource MABs
 Monoclonal Antibodies, Carbohydrate-Binding Modules, and the Detection 105

are derived using mouse hybridoma systems, and thus the

probes require anti-mouse secondary reagents whereas those
at PlantProbes are mostly rat and require anti-rat secondary
reagents (see Note 1).
2. CBMs, derived from cell wall hydrolase enzymes, are gener-
ally used as recombinant proteins with polyhistidine tags to
allow detection with secondary reagents; however, they can
also be used directly as fusion proteins with fluorescent pro-
teins such as GFP, which allows direct imaging using fluores-
cence microscopy as shown in Fig. 1 (see Note 2). CBMs are
not yet as widely available as MABs, but some are available
commercially (see online sites above).

2.2. Preparation, 1. Four percent solution of formaldehyde in PEM buffer:

Fixation, 50 mM Pipes, 5 mM EGTA, 5 mM MgSO4; pH adjusted to
and Sectioning 7.0 with KOH. Make a 12 or 16% (w/v) stock solution of
of Plant Materials paraformaldehyde in water by heating up to 70oC and adding
1 M NaOH dropwise until the cloudy solution turns clear.
Cool to RT. A good alternative is 16% formaldehyde solution
(Agar Scientific, Stansted, UK). Aliquots can be stored at
−20°C for up to 6 months (see Note 3).
2. Ethanol to prepare aqueous solutions (30–100%) for dehy-
dration procedures.
3. Wax for embedding. Steedman’s wax, an ethanol-soluble low
melting point polyester wax (35–37oC), is prepared from a mix-
ture of polyethylene glycol 400 distearate and 1-hexadecanol
(cetyl alcohol) (Sigma-Aldrich, Gillingham, UK) (see Note 4).
4. LR White resin (hard grade, containing 0.5% of the catalyst
benzoin methyl ether, Agar Scientific) can be used for both
light and electron microscopy.

Fig. 1. Micrographs showing comparison of indirect and direct fluorescence labelling

procedures for the detection of xylan by the CBM probe CBM2b-1-2. Equivalent trans-
verse sections of tobacco stem showing (a) bright field showing all cells, (b) section
showing indirect labelling of secondary cell walls using his-tagged CBM2b-1-2 and
anti-his secondary reagents and (c) section showing direct immunolabelling of second-
ary cell walls with a CBM2b-1-2*GFP construct. Both methods are effective for the
detection of xylan in secondary cell walls. cp cortical parenchyma; pf phloem fibres;
c cambium; x xylem. Bar = 10 mm.
106 Hervé, Marcus, and Knox

5. Disposable base moulds (15 × 15 × 5 mm, Electron Microscopy

Sciences, Hatfield, USA) and embedding cassettes (Simport,
Beloeil, Canada) for wax embedding.
6. Gelatine capsules (Agar Scientific) for resin embedding.
7. Polysine-coated microscope slides (VWR, Lutterworth, UK).
8. Vectabond (Vector Laboratories, Peterborough, UK) coated
multitest eight-well glass slides (MP Biomedicals, Solon, USA).
9. Nickel grids (Agar Scientific) for electron microscopy.

2.3. Immuno- 1. Super PAP hydrophobic pen (Agar Scientific) for marking
Microscopies buffer incubation regions on glass slides.
2. Phosphate-buffered saline (PBS): Prepare 10× stock with
1.37 M NaCl, 27 mM KCl, 100 mM Na2HPO4, 18 mM
KH2PO4 (pH 7.4) and autoclave before storage at room tem-
perature. Prepare working solution by dilution of one part
with nine parts water. Alternatively, use prepared 10× PBS
(Severn Biotech, Kidderminster, UK).
3. Blocking/antibody dilution buffers. PBS with 3% (w/v) milk
protein (e.g. Marvel Milk) (PBS/MP) or 3% (w/v) bovine
serum albumin (Sigma-Aldrich) in PBS (PBS/BSA).
4. Secondary antibodies: anti-rat immunoglobulin (whole mol-
ecule) reagents coupled to FITC and gold; mouse anti-­
polyhistidine; anti-mouse immunoglobulin coupled to FITC
(Sigma-Aldrich), anti-polyhistidine coupled to Alexafluor 488
(Serotec, Kidlington, UK); anti-rat coupled to AlexaFluor
488 (Invitrogen).
5. Anti-fade reagents. Citifluor glycerol/PBS AF1 (Agar Scientific).
6. Microscope slide cover slips (Scientific Laboratory Supplies,
Nottingham, UK)
7. 0.25% (w/v) Calcofluor White (fluorescent brightener,
Sigma-Aldrich).

2.4. Enzymatic 1. High pH solution for pectin de-esterification. 0.1 M sodium

Pre-Treatments carbonate (pH 11.4).
(Pectic HG Removal) 2. Pectate lyase (from Cellvibrio japonicus) (Megazyme, Bray,
Ireland) (see Note 5).
3. CAPS buffer: 50 mM CAPS, 2 mM CaCl2, pH 10.

3. Methods

Here we focus on immunofluorescence microscopy as this is a

sensitive method that can provide an important overview of the
occurrence of cell wall epitopes across an organ as well as significant
 Monoclonal Antibodies, Carbohydrate-Binding Modules, and the Detection 107

detail in relation to individual cells. Electron microscopy is useful

to locate cell wall polysaccharides in specific cell wall domains and
other cell compartments such as the Golgi apparatus.

3.1. Plant Material 1. Small samples such as Arabidopsis seeds or seedlings can be
Preparation, Excision, plunged directly into formaldehyde fixative. Maintain in fix
and Fixation for at least 2 h and for no more than overnight. Transfer to
Procedures PEM buffer or PBS and store at 4oC until use.
2. Some relatively stiff materials such as stems are amenable to
direct sectioning by hand. Hand-cut sections can be prepared
with a razor blade and can be cut from a fresh stem directly
into fixative solution or into water if the material is prefixed.
3. For wax- and resin-embedding procedures, pieces of material
(generally no thicker than 5 mm) are excised from plant organs
and placed in fixative solution. Placing material under vacuum
(to expel air) can help with infiltration of the fixative.

3.2. Wax-Embedding The wax we use is known as Steedman’s wax (8), and is a low
Protocol melting point polyester wax with good sectioning properties. It is
soluble in ethanol and therefore removed prior to immunolabel-
ling resulting in good maintenance of antigenicity.
1. Wash fixed material in PEM buffer or PBS buffer, 2 times
10 min.
2. Dehydrate by incubation in an ascending ethanol series (30,
50, 70, 90, and 100%) with 30 min incubation for each
change at 4oC.
3. Move samples to 37oC for next steps.
4. Incubate in molten wax and ethanol (1:1, overnight) and
then 100% wax (2 times 1 h).
5. Keep wax molten using a 37oC oven.
6. Fill base mould with molten wax and place sample in the wax.
Take care to orientate the sample for optimal sectioning. Fill
up with molten wax, and when almost set, apply embedding
cassette.
7. Leave at RT overnight to solidify. Can be used 12–24 h after
embedding or can be stored in a cool, dry place indefinitely.

3.3. Sectioning 1. These instructions are for the use of a HM 325 rotary micro-
Wax-Embedded tome (Microm, Bicester, UK), but they will be readily adapted
Material to other systems.
2. Sections are cut to a thickness of ~10–12 mm to produce rib-
bons, which are transferred to paper. Sections are selected
and placed on polysine slides over a drop of water to promote
spreading.
3. Slides are allowed to dry in air overnight.
108 Hervé, Marcus, and Knox

4. To de-wax and re-hydrate sections, incubate slides with 100%

ethanol 3 times 10 min, 90% ethanol/water 10 min, 50%
ethanol/water 10 min, water 10 min, water 90 min.
5. Slides are then air-dried and can be stored at RT indefinitely.

3.4. Embedding 1. Wash in buffer minus fixative for 3 times 10 min (or overnight
Protocol for LR at 4oC).
White Resin 2. Dehydrate using an ascending ethanol series (30, 50, 70, 90,
and 100%) with 30 min each change.
3. Infiltrate with resin at 4oC by increasing from 10% resin in
ethanol 1 h, 20% 1 h, 30% 1 h, 50% 1 h, 70% 1 h, 90% 1 h,
100% resin overnight, then 8 h, then overnight.
4. Transfer to gelatine capsules and ensure appropriate orienta-
tion of plant material. Fill to the top with resin and seal to
exclude air.
5. Allow polymerisation of resin either at 37oC for 5 days or by
action of UV light at −20oC.

3.5. Sectioning 1. These instructions relate to the use of Reichert-Jung Ultracut

of Resin-Embedded Ultramicrotome.
Materials 2. Prepare glass knives.
3. For light microscopy, cut sections to a thickness of 1–2 mm
onto water.
4. Transfer sections to a drop of water on Vectabond-coated
slides and allow them to dry on to the slide in air.
5. For electron microscopy, cut ultrathin sections to a thickness
of ~80 nm when they are silvery gold in colour.
6. Collect sections on nickel grids.

3.6. Immunolabelling This procedure is for the indirect immunofluorescence labelling

of Plant Cell Walls of sections of plant material (see Note 6). Always ensure that
Using Monoclonal there is a no-primary-antibody-control to assess the extent of cell
Antibodies wall autofluorescence present in the material. Here we focus on
immunofluorescence procedures, but there are very effective
alternatives such as immunogold with silver enhancement for
light microscopy (9).
1. Use the hydrophobic pen to isolate regions around sections
that will contain incubation solutions.
2. Block non-specific binding sites by incubation with PBS/MP
for at least 30 min.
3. Incubate with PBS for 5 min.
4. Incubate with primary monoclonal antibody diluted in PBS/
MP for at least 1 h at RT or overnight at 4°C. A five to ten-
fold dilution of a hybridoma cell culture supernatant is a good
starting point for the primary antibody (see Note 7).
 Monoclonal Antibodies, Carbohydrate-Binding Modules, and the Detection 109

5. Wash with three changes of PBS with at least 5 min for each
change.
6. Incubate with a secondary antibody diluted in the region of
100-fold in PBS/MP for at least 1 h at RT. Anti-rat-IgG (whole
molecule) linked to FITC is widely used. Another good photo-
stable probe is anti-rat-IgG linked to AlexaFluor 488.
7. Wash with three changes of PBS with at least 5 min for each
change.
8. Incubate with a tenfold dilution of the Calcofluor White stock
solution for 5 min. (see Note 8).
9. Wash with three changes of PBS.
10. Mount samples using a small drop of anti-fade reagent, cover
with coverslip and examine. We use Citifluor AF1 glycerol/
PBS-based anti-fade.
11. Examine with a microscope fitted with epifluorescence optics.

3.7. Immunolabelling 1. Isolate sections on slides as appropriate and block non-specific

of Plant Cell Walls binding sites (as previously explained in Subheading 3.6).
Using Recombinant 2. Incubate with the CBM diluted in PBS/MP for at least 1 h at
CBMs RT. The most effective working concentration should be
determined by trial studies for each CBM, but most CBMs
can be used effectively in the range of 5–20 mg/mL.
3. Ensure that there is a no-CBM-control to assess cell wall
autofluorescence in the section.
4. Wash with three changes of PBS.
5. In the case of a CBM fused with a fluorescent protein, pro-
ceed directly to step 9. In the case of a CBM with a polyhisti-
dine tag, incubate with anti-polyhistidine antibody diluted in
the range 1,000-fold in PBS/MP for at least 1 h at RT.
6. Wash with three changes of PBS.
7. Incubate with the secondary antibody (e.g. anti-mouse cou-
pled with FITC at 100-fold dilution) in MP/PBS for at
least 1 h.
8. Wash with three changes of PBS.
9. Incubate with Calcofluor White if required as described in
Subheading 3.6.8.
10. Mount slides using anti-fade reagent and examine.

3.8. Immunogold 1. Block to prevent non-specific binding by floating the EM

Labelling for the grid section side down on a droplet (at least 20 mL) of PBS/
Electron Microscope BSA on Parafilm® for 30 min.
2. Transfer grid to a droplet of primary antibody diluted in
PBS/BSA. Monoclonal antibody cell culture supernatants
should be diluted between 5- and 200-fold.
110 Hervé, Marcus, and Knox

3. Wash grids by incubation in a minimum of three changes

of PBS.
4. Transfer grids to secondary antibody diluted 1 in 20 with PBS/
BSA. We routinely use anti-rat IgG coupled to 10 nm gold.
5. Wash as in step 3 and then extensively in distilled water.
6. Allow the grid to dry and then examine in an electron
microscope.

3.9. Section To date, the demonstrated cases of cell wall polysaccharide epitope
Pre-Treatments Prior masking are of hemicelluloses by pectic HG. Pectic HG is vari-
to Immunolabelling ously methyl-esterified, and to effect its most efficient removal by
pectate lyase or polygalacturonase enzymes, a pre-treatment of
the section with a high pH solution is required as shown in Fig. 2.
These pre-treatments can be applied to all sectioned materials
including wax- and resin-embedded materials (see Note 5).
1. Incubate section with a solution of 0.1 M sodium carbonate
(pH 11.4) for 2 h.
2. Wash 2 times 10 min with PBS.
3. Incubate with pectate lyase (10 mg/mL) in CAPS buffer for 2 h.

Fig. 2. Micrographs showing impacts of section pre-treatments on the binding of LM19

pectic HG and LM15 xyloglucan MABs to pith parenchyma cell walls in transverse sec-
tions of tobacco stem. Immunofluorescence labelling procedures were identical in all
cases and representative micrographs are shown with no pre-treatment, with a high pH
pre-treatment (sodium carbonate) that would remove methyl esters from HG and with a
high pH treatment followed by application of a pectic HG degrading enzyme (sodium
carbonate/pectate lyase). The LM15 xyloglucan epitope is abundantly detected at the
corners of intercellular spaces after the pectate lyase treatment. Arrows indicate the cell
walls at the corners of intercellular space. Bar = 100 mm.
 Monoclonal Antibodies, Carbohydrate-Binding Modules, and the Detection 111

4. Wash with three changes PBS.

5. Sections are now ready for immunolabelling as detailed in
Subheadings 3.6, 3.7 or 3.8.

4. Conclusion

It is an exciting time for plant cell wall immunochemistry. The

combination of extensive sets of molecular probes with method-
ologies for the enzyme deconstructions and specific removal of
cell wall polysaccharides will provide real insights into the range
of cell wall structures found in plant cells and organs.

5. Notes

1. The range of available MABs and CBMs is increasing rapidly.

Care must, therefore, be taken in probe selection when
embarking upon an immunochemical survey of cell walls –
especially if an overview is required and there is no focus on a
particular subset of cell wall polymers. A good place to start
would be with probes directed to pectic HG and also the
major hemicellulose that is known for that system/taxon.
2. In the case of a GFP tag, care must be taken to assess the
binding ability of the fused CBM. Indeed, depending on
the recombinant target, the folding of this bulky tag may
impair the recognition ability of the appended CBM by
covering its binding site. In this case, the use of another tag
is required.
3. Fixatives are needed to stop all cell reactions, and materials
are most commonly fixed using aldehyde fixatives. For light
microscopy, 4% (w/v) formaldehyde is widely used. For
electron microscopy, 2.5% (w/v) glutaraldehyde is used – this
is a good fixative but can result in sample autofluorescence
and thus is not generally used for light microscopy. However,
extensive glutaraldehyde-induced autofluorescence can be
effectively quenched by the resin-embedding procedure and
so glutaraldehyde-fixed resin-embedded material can be
used for both light and electron microscopies. The fixative
buffer preferred by some electron microscopists is 0.1 M
sodium cacodylate buffer, pH 7.0. Aldehyde fixatives do not
directly crosslink polysaccharides and some may remain sol-
uble. This can be assessed by other procedures such as tissue
printing (6). Specific fixative procedures to cross-link polysac-
charides into materials have not been explored extensively.
112 Hervé, Marcus, and Knox

4. Melt 900 g of polyethylene glycol 400 distearate and 100 g

1-hexadecanol in a large beaker in an incubator at 65oC.
When melted, stir wax very thoroughly using a stirring bar.
Pour the wax into a tray lined with aluminium foil (or
50 mL plastic conical tubes) and leave at RT to harden.
Prepared wax can be stored at RT indefinitely. For embed-
ding procedures, melt an appropriate amount at 37oC and
if using a water bath ensure that container is closed to keep
out moisture.
5. The recent discovery that pectic HG can mask or block the
detection of hemicellulose polysaccharides requires methods
for pectic HG removal from sections by the use of pectic
HG-degrading enzymes. Pectate lyase or polygalacturonase
can do this effectively. Both of these enzymes act on de-ester-
ified pectic HG and therefore a high pH pre-treatment to
remove pectic HG methylesters may optimise subsequent
enzyme action and HG removal. Application of pectin-
degrading enzymes to material not fixed to a glass slide is
likely to result in separation of cells and may cause degrada-
tion of samples. Section pre-treatments can also be extended
for the enzymatic removal of other cell wall polysaccharides
and the enzymes, buffers, and conditions required will need
to be determined accordingly.
6. Indirect procedures of immunofluorescence labelling of cell
wall are widely used as these are easy and can accommodate
the use of several antibodies in the same protocol and also
readily allow assessments of non-specific binding and back-
ground autofluorescence. The principles of staged incuba-
tions in the immunolabelling procedures are the same for
intact materials and hand-cut sections, and these materials
can be incubated in tubes or plates. Direct immunolabelling
procedures, requiring just one step, are rapid and also highly
effective as shown in Fig. 1.
7. The recommended dilution of an antibody is the highest dilu-
tion that results in a strong specific signal. It is often impor-
tant to assess a few dilutions to decide on a good working
dilution. Manufacturers of secondary reagents provide good
guidance. For primary MABs, a five to tenfold dilution of cell
culture supernatants is often used; however, in some cases, up
to a 200-fold dilution can be highly effective.
8. Calcofluor White is used as a counter stain as it binds widely
to b-glycans, including cellulose, and fluoresces under UV
excitation and therefore can indicate all cell walls in sections
and is useful for orientation and identification of immunola-
belling in relation to organ and tissue anatomy. An alternative
is to use a bright field image for this purpose.
 Monoclonal Antibodies, Carbohydrate-Binding Modules, and the Detection 113

Acknowledgements

We acknowledge the funding from the UK Biotechnology &

Biological Sciences Research Council.

References

1. Knox, J. P. (2008) Revealing the structural ­secondary plant cell walls. Plant J 58,
and functional diversity of plant cell walls. 413–422.
Curr Opin Plant Biol 11, 308–313. 6. Willats, W. G. T., Steele-King, C. G., Marcus, S.
2. Boraston, A. B., Bolam, D. N., Gilbert, H. J., E., and Knox, J. P. (2002) Antibody techniques.
and Davies, G. J. (2004) Carbohydrate- In: Molecular Plant Biology – Volume Two: A
binding modules: fine-tuning polysaccharide Practical Approach (Gilmartin P. M., Bowler C.
recognition. Biochem J 382, 769–781. (eds)), pp 199–219, Oxford University Press,
3. McCartney, L., Gilbert, H. J., Bolam, D. N., Oxford, UK.
Boraston, A. B., and Knox, J. P. (2004) 7. Moller, I., Sørensen, I., Bernal, A. J., Blaukopf,
Glycoside hydrolase carbohydrate-binding C., Lee, K., Øbro, J., Pettolino, F., Roberts,
modules as molecular probes for the analysis A., Mikkelsen, J. D., Knox, J. P., Bacic, A.,
of plant cell wall polymers. Anal Biochem and Willats, W. G. T. (2007) High-throughput
326, 49–54. mapping of cell wall polymers within and
4. Marcus, S. E., Verhertbruggen, Y., Hervé, C., between plants using novel microarrays. Plant
Ordaz-Ortiz, J. J., Farkas, V., Pedersen, H. J 50, 1118–1128.
L., Willats, W. G. T., and Knox, J. P. (2008) 8. Steedman, H. F. (1957) A new ribboning
Pectic homogalacturonan masks abundant embedding medium for histology. Nature
sets of xyloglucan epitopes in plant cell walls. 179, 1345.
BMC Plant Biol 8, 60. 9. Meloche, C. G., Knox, J. P., and Vaughn, K.
5. Hervé, C., Rogowski, A., Gilbert, H. J., C. (2007) A cortical band of gelatinous fibers
and Knox, J. P. (2009) Enzymatic treat- causes the coiling of redvine tendrils: a model
ments reveal differential capacities for xylan based upon cytochemical and immunocy-
recognition and degradation in primary and tochemical studies. Planta 225, 485–498.
 Chapter 8

Screening and Characterization of Plant Cell Walls

Using Carbohydrate Microarrays
Iben Sørensen and William G.T. Willats

Abstract
Plant cells are surrounded by cell walls built largely from complex carbohydrates. The primary walls of
growing plant cells consist of interdependent networks of three polysaccharide classes: cellulose, cross-
linking glycans (also known as hemicelluloses), and pectins. Cellulose microfibrils are tethered together
by cross-linking glycans, and this assembly forms the major load-bearing component of primary walls,
which is infiltrated with pectic polymers. In the secondary walls of woody tissues, pectins are much
reduced and walls are reinforced with the phenolic polymer lignin. Plant cell walls are essential for plant
life and also have numerous industrial applications, ranging from wood to nutraceuticals. Enhancing our
knowledge of cell wall biology and the effective use of cell wall materials is dependent to a large extent
on being able to analyse their fine structures. We have developed a suite of techniques based on microar-
rays probed with monoclonal antibodies with specificity for cell wall components, and here we present
practical protocols for this type of analysis.

Key words: Carbohydrate microarrays, Plant cell walls, Polysaccharides, Antibodies, CoMPP

1. Introduction

The “comprehensive microarray polymer profiling” or CoMPP

technique combines the specificity of monoclonal antibodies
(mAbs) with the high-throughput capacity of microarray technol-
ogy (1). Using CoMPP, profiles of the polysaccharide composi-
tions of large (hundreds) sets of cell wall samples can be rapidly
established (within a few days). CoMPP involves the extraction of
cell wall components using a series of solvents. The extractions are
printed as microarrays and then probed with a series of mAbs or
other probes. The signals from the arrays provide semi-quantitative
data about the relative abundance of polysaccharides across
the sample set (Fig. 1). The technique can be used to analyse a

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_8, © Springer Science+Business Media, LLC 2011

115
116 Sørensen and Willats

a
Extraction
Mixermill

b
Spotting Microarray robot

c
Probing Monoclonal antibodies

d Microarray software
Analysis

Fig. 1. Cell wall polymer extraction is performed using a 96-well format homogenizer,
such as a TissueLyser II (a). Using a microarray robot, the samples are printed as
microarrays (b). The microarrays are probed using monoclonal antibodies (c) and the
arrays are analysed using microarray analysis software such as ImaGene 6.0 (d).

wide range of cell wall materials including different species,

­developmental stages, mutants or plants exposed to different
growth conditions, or processing steps. CoMPP is complementary
to established biochemical techniques, such as monosaccharide
composition or linkage analysis because antibody binding provides
information about the occurrence of larger glycan structures
(epitopes) rather than individual sugars that cannot always be
assigned with confidence to polysaccharides. CoMPP is highly ver-
satile and can be modified to suit the particular needs of the exper-
iment. Starting materials can be fresh plant tissues, or purified or
semi-purified cell walls. The solvents and regimes used to extract
polysaccharides can be varied, and so can the set of probes used for
analysis. Since only a few milligrammes of starting material is
required, CoMPP can be used to fine-map polysaccharide occur-
rence within single plant organs. CoMPP arrays can be printed
onto a variety of surfaces including nitrocellulose membrane or
nitrocellulose-covered slides, and the arrays can be printed using
either pin-based or piezoelectric robots. The protocols below are
a detailed summary of previously published studies (1, 2).

2. Materials

2.1. Extraction Buffers 1. 50 mM diamino-cyclo-hexane-tetra-acetic acid (CDTA),

pH 7.5.
2. 4 M sodium hydroxide (NaOH) with 0.1% v/v Sodium
­borohydride (NaBH4).
 Screening and Characterization of Plant Cell Walls Using Carbohydrate Microarrays 117

3. Cadoxen (31% (v/v) 1,2-diaminoethane with 0.78 M

­cadmium oxide (CdO)). Prepare cadoxen by stirring 310 mL
1,2-diaminoethane with 720 mL H2O and 100 g CdO at
20°C for 3 h and 4°C for 18 h. Centrifuge and use super­
natant. Store at 4°C (3).

2.2. Other Buffers 1. Phosphate buffered saline (PBS): (140 mM NaCl, 2.7 mM
and Solutions KCl, 10 mM Na2HPO4, 1.7 mM KH2PO4, pH 7.5).
2. Milk powder/PBS (MP/PBS): 5% milk powder (w/v) in
PBS. Mix well.
3. Primary mAbs and secondary alkaline phosphatase-conjugated
mAbs are diluted in MP/PBS according to supplier’s
specifications.
4. Alkaline phosphatase (AP) buffer: 100 mM sodium chloride
(NaCl), 5 mM magnesium chloride (MgCl2), 100 mM dieth-
anolamine, pH 9.5.
5. 5-Bromo-4-chloro-3¢-indolyphosphate (BCIP) stock: 20 mg/
mL BCIP in de-ionised water (dH2O). Store at −20°C.
6. Nitro-blue tetrazolium chloride (NBT) stock: 50 mg/mL
NBT in methanol. Store at −20°C.
7. AP developing solution: 10 mL AP-buffer with 66 mL NBT
stock and 82.3 mL BCIP stock.

2.3. Materials 1. Nitrocellulose membrane, 0.45 mm pore size (Whatman,

Schleicher & Schuell, Dassel, Germany).
2. Eight-strip collection microtubes (Qiagen MM 200, Qiagen
Nordic, West Sussex).
3. Metal ball bearings (Qiagen, Qiagen Nordic, West Sussex, UK).

2.4. Equipment 1. TissueLyser II (Qiagen MM 200, Qiagen Nordic, West

Sussex, UK).
2. Microarray robot (Microgrid II, Genomic Solutions, Ann
Arbor, MI, USA) or (Sprint, ArrayJet, Roslin, Scotland, UK).
These are examples that we use and other arraying hardware
may perform equally well.
3. Rocking table.

3. Methods

3.1. Plant Cell Wall 1. Plant material is collected (e.g. appropriate cell type, tissue, and
Material organs samples), including biological replicates (see Note 1).
2. Samples are homogenised in liquid nitrogen using a mortar
and pestle or using a mechanical homogeniser such as the
Qiagen TissueLyser II, with appropriate inserts (see Note 2).
118 Sørensen and Willats

3. Alcohol insoluble residue (AIR) is made by adding five

­volumes of 70% ethanol to the powdered plant material and
leaving the sample on a rocking table for 1 h.
4. The samples are spun down at 2,500 × g for 10 min and the
supernatant discarded.
5. Five volumes of 70% ethanol is added and the samples are
shaken for 1 h.
6. Steps 4 and 5 are repeated until the supernatant is clear.
7. A final wash for 5 min in acetone is performed, and the sam-
ples are left to air dry, leaving the AIR extract.

3.2. Extraction of Cell 1. 10 mg of each AIR sample is weighed out and placed in eight-
Wall Fractions strip collection tubes in 96-tube boxes.
2. A metal ball bearing is placed in each tube and the tubes are
capped.
3. Before the extractions, the tubes are submerged in liquid
nitrogen and the samples are homogenised once more with
the TissueLyser II.
4. 300 mL CDTA is added to each tube, and the samples are
shaken on the highest speed (30 Hz) on the TissueLyser II
for 2 min before a 2-h extraction at 6–10 Hz (see Note 3).
5. The boxes are placed in a centrifuge and spun down at
2,500 × g for 10 min and the supernatants are collected.
6. 300 mL NaOH is added and the procedure is repeated with
another 2 h extraction.
7. After the NaOH extraction, the supernatants are collected,
300 mL Cadoxen is added, and the samples are shaken again
for 2 h or more.
8. The supernatants are collected.
9. Other solvents or extraction procedures may of course be
used, as needed.

3.3. Preparation 1. Samples are transferred to microtiter plates and dilutions are
of Microtiter Plates made in PBS, typically a 0, 5 and 25× serial dilution series.
and Printing of Arrays Importantly, ink (India ink or other appropriate marker) is
used as a marker for the outline of the arrays and is also added
to the appropriate wells in the microtiter plate (see Note 4).
2. The microarrayer is loaded with nitrocellulose membrane,
pins, and microtiter plates.
3. Parameters such as humidity, collection and dwell time, num-
ber replicates, and layout are set.
4. The samples are printed as microarrays (see Note 5).
 Screening and Characterization of Plant Cell Walls Using Carbohydrate Microarrays 119

3.4. Probing of Arrays The volumes below are optimised for arrays up to 5.7 × 5.7 mm;
however, the size of the arrays will vary according to the number
of samples printed, and smaller or larger probing containers and
volumes of mAb solutions might be used.
1. The membrane is cut into individual arrays.
2. Arrays are blocked individually, for example in weighing
boats, in 5 mL MP/PBS solution for 1 h. Remember to
include one array as a control (no primary mAb).
3. MP/PBS is discarded.
4. An appropriate dilution of 5 mL mAb in MP/PBS is added,
and arrays are left at room temperature for 2 h on a rocking
table or overnight at 4°C. The dilutions may have to be tested
empirically. The control microarray is incubated in MP/PBS
containing no primary antibody.
5. The mAb solution is discarded and arrays are washed 3 times
for 5 min in 10 mL PBS on a rocking table.
6. An appropriate dilution of AP-conjugated secondary mAb in
MP/PBS (5 mL) is added and arrays are left at room tem-
perature for 2 h on a rocking table, or overnight at 4°C.
7. The mAb solution is discarded, and the arrays are washed
3 times for 5 min in 10 mL PBS and 1 time in 10 mL dH2O
on a rocking table.
8. Arrays are developed using AP developing solution and rinsed
in H2O before placed on filter paper for drying.

3.5. Quantification 1. The arrays are scanned on a standard commercial desktop

and Analysis of Arrays flatbed scanner at highest resolution (1,200 dpi) and images
saved as 16-bit tiff files, in negative contrast image format
(i.e. light-coloured spots on a dark membrane background).
2. The tiff files are uploaded to microarray analysis software (e.g.
ImaGene 6.0) and each spot signal quantified, after background
corrections are made and the spot areas are appropriately
defined, using the parameters of the analytical software.
3. The data is saved as a .txt file and imported into statistical
analytical software for further analysis.
4. An appropriate lower limit cut-off value can be introduced in
order to avoid false-positive values (see Note 6).

3.6. Online Tools 1. A range of tools is available to further extract information

for Analysis from the data and to make more detailed statistical compari-
sons. For example, heatmaps can be made by uploading the
data to the online heatmapper tool (www.bar.utoronto.ca/
ntools/cgi-bin/ntools_heatmapper.cgi) in order to present
120 Sørensen and Willats

a b A ntibodies

CDTA

S am p les
S am p les
N aO H
C adox en

S am p les
c
A ntibodies
S am ples

Fig. 2. Example of an array after probing, converted into a negative 16-bit tiff file, is
shown in (a). The data can be converted into a heatmap format (b), or used for a cluster
analysis (c).

the data in a more visual way. An example is shown in Fig. 2b

where the highest value in the data set has been set to 100
and the others adjusted accordingly.
2. Cluster analysis can be performed by uploading the data to
the online clustering tool (http://www.bioinf.ebc.ee/EP/
EP/EPCLUST/) to reveal the relationship between specific
samples. An example is given in Fig. 2c.

4. Notes

1. All plants have different developmentally related cell wall

compositions, and when sampling plants outside the lab envi-
ronment, it is very important to note at least the dates of col-
lection and to preferably collect over a season to ensure the
inclusion of as many differentially expressed polymers as
possible.
2. Many plants have a very tough cell wall, which can be difficult
to homogenise even when frozen in liquid nitrogen and
ground with a pestle and mortar. There are several commer-
cially available grinding tools for the mixer mill system, for
 Screening and Characterization of Plant Cell Walls Using Carbohydrate Microarrays 121

example the Retsch stainless steel inserts (Retsch, Haan,

Germany), that can be of help.
3. Some plants contain pectic polymers at a level that will make
the supernatant in the CDTA extraction very viscous, and
depending on the plant material, the AIR to buffer ratio
might have to be adjusted.
4. As mentioned, we usually make a 0, 5, and 25× dilution series
of our extracts for printing. Depending on the extract, it may,
however, sometimes be necessary to start with a 5× dilution
to avoid spot saturation.
5. Cadoxen is a very potent cellulose solubiliser and as such will
degrade the nitrocellulose over time. Arrays including extracts
of this solvent can therefore not be stored as long.
6. The quantification method mentioned above is based on
AP-conjugated secondary antibodies but can be adjusted for
the use of horseradish peroxide (HRP) or fluorescent quanti-
fication systems. The latter is best suited for slide printing,
since it requires a slide scanning set-up.

References
1. Moller, I., Sørensen, I., Bernal, A. J., Blaukopf, Willats, W. G. T. (2008) Mixed linkage
C., Lee, K., Øbro, J., Roberts, A., Mikkelsen, (1→3),(1→4)-b-d-glucan is not unique to the
J. D., Knox, J. P., Bacic, A. and Willats, W. G. Poales and is an abundant component of
T. (2007) High-throughput mapping of cell- Equisetum arvense cell walls. Plant J 54,
wall polymers between and within plants using 510–521.
novel microarrays. Plant J 50, 1118–1128. 3. Fry, S. C. (1988) The Growing Plant Cell Wall:
2. Sørensen, I., Pettolino, F. A., Wilson, S. M., Chemical and Metabolic Analysis. The
Doblin, S. M., Johansen, B., Bacic, A. and Blackburn Press, Caldwell.
 Chapter 9

Electron Tomography and Immunogold Labelling as Tools

to Analyse De Novo Assembly of Plant Cell Walls
Marisa S. Otegui

Abstract
High-resolution imaging of the membranous intermediates and cytoskeletal arrays involved in the assembly
of a new cell wall during plant cytokinesis requires state-of-the-art electron microscopy techniques. The
combination of cryofixation/freeze-substitution methods with electron tomography (ET) has revealed
amazing structural details of this unique cellular process. This chapter deals with the main steps associ-
ated with these imaging techniques: selection of samples suitable for studying plant cytokinesis, sample
preparation by high-pressure freezing/freeze substitution, and ET of plastic sections. In addition, immu-
nogold approaches for the identification of proteins and polysaccharides during cell wall assembly are
discussed.

Key words: Electron tomography, Cell walls, Cytokinesis, Cryofixation, Immunolabelling

1. Introduction

The formation of a new cell wall during cytokinesis is a highly

regulated process that requires coordinated interactions between
the cytoskeleton and membrane trafficking pathways. The plant
cytokinetic machinery includes three distinct structures: the
phragmoplast, the cell plate, and the cell plate assembly matrix
(CPAM) (1–3).
The phragmoplast consists of two sets of anti-parallel micro-
tubules (MTs) and actin filaments with their plus/barbed ends
facing the division plane. The Golgi apparatus plays a central role
during cytokinesis, producing millions of vesicles that provide
the building materials for the future cell wall (1). These vesicles
are transported along the phragmoplast MTs and fuse with each
other at the division plane, giving rise to the cell plate. By the
coordinate disassembly and reassembly of phragmoplast MTs,

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_9, © Springer Science+Business Media, LLC 2011

123
124 Otegui

more vesicles are added to the growing edges of the expanding

cell plate. Finally, the cell plate fuses with the parental plasma
membrane and a new cell wall forms between the two daughter cells,
completing cytokinesis. In addition, a filamentous matrix called
CPAM has been found to enclose the cell plate growing edges,
fusing vesicles, and most of the MT plus ends at the phragmoplast
midline. Although the composition of this matrix is not known, it
has been postulated to play an important role in both stabilisation
of phragmoplast MT plus ends and membrane fusion (3).
The combination of cryofixation/freeze-substitution meth-
ods with electron tomography (ET) has revealed amazing details
of this complex process. By combining superb cellular preserva-
tion and three-dimensional (3D), high-resolution (4–7 nm) imag-
ing, it has been possible to analyse the architecture of the
membranous intermediates that arise during cell plate formation
in different plant cell types, the changes in phragmoplast organisa-
tion, and even the distribution of individual macromolecules, such
as kinesin-like molecules and dynamin rings (1–4). In addition,
ET has provided novel information on quantitative changes in cell
plate surface area and volume that have been essential to estimate
membrane dynamics during cytokinesis (1, 3). No other method
has achieved comparable 3D resolution to help us understand plant
cytokinesis at the molecular level in the cellular context (5).
To obtain reliable 3D electron tomographic data, it is very
important to work with very well-preserved biological sample.
Therefore, this chapter will not only deal with the process of cal-
culating and modelling electron tomograms but also with plant
sample preparation by the best preservation method available,
cryofixation and freeze-substitution. Since it is also very impor-
tant to be able to correlate structure with composition, protocols
for immunodetection of proteins and polysaccharides are also
included.

2. Materials

2.1. Plant and Cell 1. Sterile hood.

Culture Materials 2. 10% bleach.
3. 70% ethanol.
4. Sterile water.
5. Sterile glass Pasteur pipettes or sterile 1 mL pipette tips.
6. Arabidopsis seeds.
7. 0.8% agar plates containing ½ strength, i.e. 2.2 g of powder
per litre of Murashige and Skoog (MS) basal medium (Sigma-
Aldrich, St. Louis, MO).
 Electron Tomography and Immunogold Labelling as Tools to Analyse 125

8. Arabidopsis siliques containing developing seeds (between 5

and 10 days after pollination).
9. Tobacco Bright Yellow-2 (BY-2) cultured cells.
10. BY-2 culture medium: 0.43% (w/v) MS basal medium, 3 mM
thiamine-HCl (B1), 0.5 mM myo-inositol, 85 mM sucrose,
1 mM 2,4-dichlorophenoxyacetic acid (2,4-D), 1.3 mM
KH2PO4, pH 5.5–5.8; autoclave and store at room
temperature.
11. Aphidicolin (inhibitor of eukaryotic nuclear DNA replica-
tion) (Sigma-Aldrich).
12. Propyzamide (MT assembly inhibitor) (Sigma-Aldrich).

2.2. High-Pressure 1. High-pressure freezer (Leica EM HPM100 or Bal-tec/ABRA

Freezing Fluid AG HPM 010).
2. Freezing brass planchettes (“hats”) type B (Ted Pella,
Redding, CA) if a Bal-tec/ABRA Fluid AG HPM 010 is used.
3. Cryoprotectant: 0.1 M sucrose or 1-hexadecene.
4. 2.0 mL Cryovials (Nalge Nunc, Thermo Fisher Scientific,
Rochester, NY).
5. Forceps.
6. Tank of liquid nitrogen.

2.3. Freeze-Substitution 1. Cryovials containing 1.5 mL of 2% OsO4 in anhydrous acetone

and Resin Embedding (Electron Microscope Sciences, Hatfield, PA) (prepare in hood
using gloves and store in liquid nitrogen).
2. Cryovials containing 1.5 mL 0.2% glutaraldehyde plus 0.2%
uranyl acetate in anhydrous acetone (Electron Microscope
Sciences) (prepare in hood using gloves and store in liquid
nitrogen).
3. Freeze-substitution/low-temperature resin embedding sys-
tem (AFS, Leica, Bannockburn, IL).
4. Aluminium block with holes for fitting cryovials.
5. Eponate 12 kit (Ted Pella). Embed 812 kit from Electron
Microscope Sciences can also be used.
6. Eponate 12 resin mix without accelerator: 29 g of Eponate
12 resin, 16 g of DDSA, 14.3 g of NMA. It can be kept at
4°C for several days.
7. Eponate 12 resin prepared at the following concentrations:
10% Eponate 12 resin mix, 25% Eponate 12 resin mix, 50%
Eponate 12 resin mix, 75% Eponate 12 resin mix.
8. 100% Eponate 12 resin mix with accelerator: 2.5–3% BMPA
to freshly prepared Eponate 12 resin mix. It can be kept at
4°C for several days.
126 Otegui

9. Lowicryl HM20 resin kit (Polysciences, Warrington, PA, or

Electron Microscope Sciences).
10. Flat embedding moulds (Ted Pella).
11. Coverwell Imaging chambers (2.8 mm deep; 20 mm diame-
ter; Electron Microscope Sciences).
12. Dry ice and Styrofoam box.
13. Plastic mounting cylinders (Ted Pella).

2.4. Preparation 1. Copper/rhodium slot grids coated with 0.7–1% (w/v) Formvar
of Sections in ethylene dichloride (Electron Microscopy Sciences).
for Electron 2. Ultramicrotome.
Tomography
3. 2% uranyl acetate on 70% methanol.
4. Reynold’s lead citrate: 2.6% lead nitrate and 3.5% sodium cit-
rate, pH 12.
5. 10 or 15 nm colloidal gold particles (Electron Microscopy
Sciences) (store at 4°C).
6. Carbon coater.

2.5. Image Acquisition 1. Intermediate voltage (300 kV) electron microscope (for
and Calculation example FEI Tecnai G2 30 TWIN) equipped with high-tilt
of Dual-Axis Electron rod for tomographic image acquisition.
Tomograms 2. Software: SerialEM (6–8) for image acquisition and IMOD
(9) package for tomogram reconstruction (can be down-
loaded from http://bio3d.colorado.edu/docs/software.
html) (see Note 1).

2.6. Image Segmentation 1. IMOD package (9) (can be downloaded from http://bio3d.
(Modelling) colorado.edu/docs/software.html) (see Note 1).

2.7. Immunolabelling 1. Nickel or gold single slot grids coated with 0.25–0.5% (w/v)
Formvar in ethylene dichloride.
2. Phosphate-buffered saline (PBS): Prepare 1 L of 10× stock
solution containing 1.76 g of NaH2PO4, 11.49 g of
Na2HPO4, 85 g sodium chloride, pH 6.8 (store at room
temperature).
3. PBS-T-0.1%: Dilute 1 mL of 10× PBS with 9 mL of water
and add 10 mL of Tween-20.
4. PBS-T-0.5%: Dilute 100 mL of 10× PBS with 900 mL of
distilled water and add 0.5 mL of Tween-20.
5. Blocking buffer: 5% (w/v) non-fat milk in PBS-T-0.1%.
6. Primary antibody in blocking buffer.
7. Secondary antibody conjugated to gold particles (5, 10, or
15 nm in diameter) diluted (1:10) in blocking buffer.
 Electron Tomography and Immunogold Labelling as Tools to Analyse 127

8. Cryosubsitution medium: 0.2% glutaraldehyde plus 0.2% uranyl

acetate in anhydrous acetone.
9. HM20 resin mix: 2.98 g Crosslinker D, 17.02 g Monomer E,
0.1 g Initiator C. Mix the three ingredients in a brown-
coloured glass bottle (HM20 is sensitive to light) and keep it
at −20°C.
10. Three HM20 resin: 30, 60, and 100% in anhydrous acetone.

3. Methods

3.1. Plant Material On average, the apical root meristem region of an 8-day-old
3.1.1. Arabidopsis
Arabidopsis seedling consists of ~52 cells, each of which divides
Seedlings
every 18 h (10). Therefore, the chance of finding at least a few
cells undergoing cytokinesis in a given root tip is relatively high.
To obtain root tips from 8 to 10-day-old seedlings, it is best to
germinate seeds on 0.8% agar plates containing ½ strength MS
basal medium.
1. Place Arabidopsis seeds in plastic tube and add 10% bleach for
5 min, mixing occasionally.
2. Remove bleach using sterile glass Pasteur pipette or sterile
1 mL pipette tips (open the pipette tip box inside the hood)
and rinse 3 times with sterile water.
3. Add 70% ethanol, mix, and discard after 5 min.
4. Rinse 3 times with sterile water and place seeds on 0.8% agar
plates supplemented with ½ strength MS.

3.1.2. Developing Seeds Developing seeds are also a very good source of dividing cells. In
Arabidopsis, the analysis of developing seeds allows for the simul-
taneous study of somatic cytokinesis in embryo cells and an
unconventional cytokinesis that occurs during endosperm cellu-
larisation (1, 11, 12). The endosperm in Arabidopsis starts to cel-
lularise at the micropylar region, when the embryo has reached
the late globular stage (approximately 5–6 days after pollination).
The high rate of cell divisions continues in the embryo until the
torpedo/early bent cotyledon stage (approximately 10–12 days
after pollination).

3.1.3. Synchronisation The tobacco BY-2 cell line developed by Nagata and co-workers
of BY2 Cells (13, 14) responds well to synchronisation protocols and can provide
mitotic indexes of ~39–77%.
1. Grow BY-2 cells in medium containing 3–5 mg/mL aphidicolin
for 24 h.
2. Wash out the aphidicolin-containing medium and allow cells
to grow in fresh medium for 3 h.
128 Otegui

3. Add 6 mM propyzamide to the medium for 6 h (15).

4. After 90–180 min of washing out the propyzamide, most
dividing cells undergo cytokinesis (16).

3.2. High-Pressure 1. 1 mm segments of root tips, whole developing seeds, or

Freezing excised developing embryos are loaded in a type B freezing
planchette containing cryoprotectant (0.1 M sucrose or
1-hexadecene).
2. Another freezing planchette is placed on top to close the
chamber. It is important to completely fill the chamber with
cryoprotectant, not leaving air bubbles that could collapse
during high-pressure freezing.
3. If working with cultured cells grown in a medium with
sucrose, a soft pellet of cultured cells can likewise be loaded
directly into the freezing planchette.
4. Place freezing planchettes in the sample holder and freeze
them under high pressure in a HPM 010 unit.
5. Under liquid nitrogen, split open the two freezing planch-
ettes with the tips of a pair of forceps pre-cooled in liquid
nitrogen. The freezing planchettes containing the samples
can either be stored in liquid nitrogen (see Note 2) or placed
directly in cryosubstitution medium.

3.3. Freeze-Substitution The freeze-substitution medium and resin should be chosen

and Resin Embedding according to the type of analysis one wants to perform. To achieve
good preservation and staining of MTs and membranes, freeze-
substitution in 2% OsO4 in acetone followed by Eponate 12
embedding is recommended. However, OsO4 and epoxy-based
resins such as Eponate 12 are not suitable for most immunolabel-
ling approaches. Cryosubstitution in acetone without fixatives or
low concentrations of glutaraldehyde (0.2%) and uranyl acetate
(0.2%) (17) followed by embedding in methacrylate-based low
temperature UV curing resins, such as Lowicryl HM20, are pre-
ferred for immunogold labelling applications.
It is important to keep in mind that many antibodies raised
against cell wall polysaccharides, such as anti-callose (Biosupplies
Australia, Victoria, Australia), anti-xyloglucan (18), and CCRC-M1
(19), and carbohydrate epitopes on arabinogalactan proteins, such
as the JIM13 (20) and LM2 (21) antibodies, work well on osmi-
cated samples embedded in Eponate 12.

3.3.1. Structural Analysis 1. Place freezing planchettes with frozen samples in cryovials
Using Dry Ice and containing 1.5 mL of 2% OsO4 in acetone. Be sure to keep
Styrofoam Box cryovials in liquid nitrogen during planchette transferring
and to pre-cool the tip of the forceps in liquid nitrogen before
touching the freezing planchettes.
 Electron Tomography and Immunogold Labelling as Tools to Analyse 129

2. Place the cryovials in an aluminium block pre-cooled in dry

ice (at −80°C) inside a Styrofoam box and leave it there for
5 days. Refill the box with dry ice if necessary.
3. Transfer aluminium block with cryovials to a freezer at −20°C
for 24 h (see Note 3).
4. Transfer aluminium block with cryovials to fridge at 4°C for
at least 3 h (see Note 3).
5. Transfer aluminium block with cryovials to the fume hood
and leave it at room temperature for at least 1 h.
6. Discard substitution medium and rinse samples with fresh
anhydrous acetone at least 5 times (every 5 min).
7. Remove freezing planchettes (freezing planchettes can be reused
if they are cleaned by sonication in methanol or acetone).
8. Rinse samples one more time with fresh acetone.
9. Prepare Eponate 12 resin mix without accelerator.
10. Add increasing concentration of Eponate 12 resin mix (with-
out accelerator) in anhydrous acetone and keep the samples
at least 4 h in each of the following resin concentrations:
10% Eponate 12 resin mix, 25% Eponate 12 resin mix, 50%
Eponate 12 resin mix, 75% Eponate 12 resin mix.
11. Add 100% Eponate 12 resin mix (without accelerator) for at
least 8 h.
12. Prepare 100% Eponate 12 resin mix with accelerator (add
2.5–3% BMPA to freshly prepared Eponate 12 resin mix) (see
Note 4). It can be kept at 4°C for several days.
13. Add 100% Eponate 12 resin mix with accelerator for at least
12 h (repeat this step twice).
14. Place samples in flat embedding moulds and polymerise at
60°C for 24 h (see Note 5).
15. Select samples in resin blocks using a dissecting microscope
and mount pieces of resin containing the samples on plastic
mounting cylinders.

3.3.2. Immunolabelling Cryosubstitution for immunolabelling can also be performed

Using the Leica AFS using “custom” freeze-substitution devices as explained in
Subheading 3.3.1. However, if a low temperature UV curing
resin is used, it is advisable to use an automatic freeze-substitution
device such as the Leica AFS. This device allows for a very precise
control of the temperature during freeze-substitution and resin
embedding. In addition, it includes a UV lamp that can be directly
attached to the sample chamber for resin polymerisation.
1. Under liquid nitrogen, place freezing planchettes containing
samples in cryovials containing frozen cryosubstitution
medium.
130 Otegui

Table 1
Suggested steps/programmes to use during freeze-substitution and resin
embedding using a Leica AFS

Step/programme Temperature 1 Temperature 2 (°C) Ramp (°C/h) Duration (h)

0 −90°C −90 – 120

1 −90°C −60 5 54
2 −60°C −50 5 24
3 −50°C/h 18 5 24

2. Set the programme/s in the Leica AFS according to Table 1.

3. Transfer cryovials to the Leica AFS and leave them at −90°C
for 5 days.
4. After the AFS chamber has reached −60°C, discard cryosub-
stitution medium and rinse samples and freezing planchettes
with pre-cooled anhydrous acetone at least 3 times.
5. Remove freezing planchettes with pre-cooled forceps. Samples
should be free in the acetone by now, completely detached
from the freezing planchettes.
6. Prepare HM20 resin mix.
7. Start the infiltration with HM20 by adding increasingly
higher concentrations of resin in anhydrous acetone to the
cryovials. Three steps can be used: 30, 60, and 100% HM20
resin (at least 3 h for each concentration). All the resin solu-
tions should be pre-cooled to −60°C before being added to
the cryovials containing the samples.
8. Add fresh, pre-cooled 100% HM20 resin mix 4–5 times over
the next 24 h.
9. With a pre-cooled glass Pasteur pipette, remove samples from
cryovials and place them in Coverwell imaging chambers with
a glass coverslip on top. Be sure to completely fill the cham-
ber with resin; no air bubbles should be seen after placing the
glass coverslip.
10. Attach the UV lamp and adjust the temperature to −50°C.
11. Polymerise at −50°C for 24 h followed by a slow warming up
to 18°C over the following 24 h (see Table 1). After 48 h
under UV light, the HM20 resin should be completely
polymerised.
12. Remove resin blocks with samples from the imaging chambers.
13. Select samples in resin blocks using a dissecting microscope
and mount pieces of resin containing the samples on plastic
mounting cylinders.
 Electron Tomography and Immunogold Labelling as Tools to Analyse 131

3.3. Preparation High-pressure frozen/freeze-substituted samples embedded in

of Sections Eponate 12 can be sectioned in an ultramicrotome (60–70 nm
for Electron thick section) and analysed in a regular transmission electron
Tomography microscope to evaluate preservation quality and to identify cells
undergoing cytokinesis. Once the right specimens have been
selected, semi-thick section for ET can be prepared.
1. Collect 250–300 nm thick sections on single slot specimen
grids coated with Formvar.
2. Stain sections with a 2% uranyl acetate in 70% methanol for
10 min (see Note 6) followed by Reynold’s lead citrate for
5 min.
3. Apply 10 mL of 10- or 15-nm colloidal gold solution to each
side of the sections for 5 min. Remove the excess solution by
touching the grid with filter paper. The gold particles are used as
fiducials, aiding in the fine alignment of the tilt images during
the tomographic reconstruction.
4. Carbon-coat both sides of the grids with the sections to mini-
mise charging and drifting during imaging under the electron
beam.

3.4. Image Acquisition The resolution of a tomographic reconstruction depends on dif-

and Calculation ferent factors. Given that the quality of the tilt series (image focus
of Dual-Axis Electron and alignment) is optimal, the main factors affecting resolution are
Tomograms (a) the magnification at which the images are collected, (b) the
angular interval, (c) the angular range, and (d) section thickness
(22, 23). Before collecting the images for calculating tomographic
reconstructions, the variables involved in these four factors have
to be carefully considered.
(a) Magnification: Very often, large areas of the cells have to be
imaged to analyse forming cell walls and associated phragmo-
plasts. However, imaging at low magnification is not recom-
mended because of the resulting loss in resolution. If a
charge-couple device (CCD) camera is used to collect the
images, the magnification should be high enough that each
pixel in the image is equivalent to ~1 nm in the specimen. If the
structure of interest is so large that cannot be imaged in a
single frame, montaged images can be used to image large
areas without compromising resolution (5).
(b) The angular interval: 1 or 1.5° angular intervals are recom­
mended.
(c) Angular range: Contrary to medical computed tomography
in which images of the patient can be collected over a full
360°, the angular range allowed by the conventional tilting
specimen holders used for ET is much more restricted, resulting
in a wedge of missing information between the maximal tilt
angle collected and 90° (6, 24). This results in distorted
132 Otegui

tomograms with anisotropic resolution (22). To improve the

isotropy in resolution, it is recommended to collect images
from two orthogonal axes and combine the two resulting
tomograms into a dual-axis tomogram (6).
(d) Section thickness: If an intermediate-voltage (200–300 kV)
electron microscope is used, sections thicker than 300 nm
will likely result in poor resolution images, particularly at high
tilt angles. If larger volumes are required for imaging cell
plates and phragmoplasts, serial tomograms can be obtained
from serial sections (2, 3). A ribbon of several serial sections
can be placed on the same grid and a dividing cell can then be
located and imaged on each relevant section in the ribbon.
However, it is important to note that this approach suffers
from a 15–25 nm gap of missing information between serial
tomograms (25). For highly complex cellular structures, this
gap in information may cause difficulties in the alignment of
serial tomograms.
1. Place specimens in a high-tilt sample holder of an interme-
diate (300 kV) electron microscope and collect images at
1° angular intervals and over an angular range of ±60–70°
using the free software SerialEM (see Note 1) (Fig. 1).
2. After collecting the first stack of images, rotate the grid
90° and collect images at 1° angular intervals along the
second axis.
3. Process the resulting images using the eTomo programme
in the IMOD package (see Note 1) (Fig. 1).

3.5. Modelling Cellular structures contained in electron tomographic reconstruc-

and Quantitative tions can be manually segmented (modelling) using the 3dmod
Analysis programme of the IMOD package. Modelling is the creation of
of Tomographic graphic objects that accurately represent the 3D positions of
Models features of interest in a tomogram. Image segmentation is the
most time consuming part of the process, and it is somehow a
subjective task. The 3dmod programme allows the operator to
draw on the image data, placing points, chosen shapes (circles,
etc.), or sets of point (lines or curves that match the structure of
interest) as “overlays” on the image data. Each such representa-
tion is called a contour, and these are generally drawn on a single
tomographic slice extracted from the tomogram (Figs. 2 and 3).
Contours can be either closed (if they represent something like
the membrane that surrounds a vesicle or cisterna) or open (if
they represent a fibre, like a MT or actin filament).
Three different types of image displays are generally used
during modelling: the Zap window that displays tomographic
slices cut parallel to the surface of the physical section that was
reconstructed, the Slicer window that allows the operator to display
a slice at an arbitrary angle through the volume, and the Model
 Electron Tomography and Immunogold Labelling as Tools to Analyse 133

Fig. 1. Image acquisition and ET reconstruction. (a, c) Semi-thick (300 nm) sections imaged in an FEI Tecnai TF30 at 0°
tilt. The small black dots are 15 nm gold particles used as fiducials for image alignment. Images were taken every 1°
interval, from +60 to −60°, along two orthogonal axes. The two resulting single axis tomograms were combined in a
single dual-axis tomogram. (b, d) Single tomogramic slice show in great detail membrane profiles, microtubules (MTs),
ribosomes hardly distinguishable in the original semi-thick sections (a) and (c), respectively. CW cell wall; G Golgi; LB lipid
body; M mitochondrion; MVB multivesicular body; TGN trans-Golgi Network. Scale bars = 500 nm.

View window that shows contours and meshed objects as they are
created (Figs. 2 and 3). A general guide that provides a compre-
hensive description of the 3dmod modelling programme can be
found at http://bio3d.colorado.edu/IMOD.
Quantitative information obtained from tomograms (see
below) is based on the tomographic models that have been
meshed, a process that generates 3D graphic objects with defined
surfaces. However, meshes are derived from contours, so it is
extremely important to draw the contours accurately.

3.5.1. Microtubules In the simplest case, MTs and actin filaments can be modelled as
and Actin Filaments hollow tubes of a given diameter.
1. Create a new object under the “Edit” menu of 3dmod and
chose the “open contours” option (Fig. 2a).
2. Move along the stack of tomographic slices using the Zap
window, identify one end of the MT and place the first point,
134 Otegui

Fig. 2. Segmentation of phragmoplast MTs (a) and vesicles (b) during cell wall assembly using 3dmod. MTs are seg-
mented as open contours (times) in the Zap window and rendered as hollow cylinders 25 nm in diameter in the Model
View window. Vesicles are rendered as spheres of variable radii. Scale bars = 100 nm.

as close to the centre of the MT as possible. Place one or more

points at every fifth or sixth tomographic slice along the MT
length until reaching the other end of the MT.
3. Alternatively, use the “slicer” window to model MTs (24).
Adjust the X, Y, Z sliders to find a tomographic slice that
contains as long a segment as possible of the MT axis and
place points at the beginning and end of the segment. In either
case, each MT/filament should be considered a new contour
within one object.
4. To obtain a 3D representation of the MT/filament, use the
command imodmesh with the options -t, for tube, -d (the pixel
diameter of the tube, 25 nm for a MT, 5–9 nm for an actin
filament), and -E (this will “cap” the end of the tube).
 Electron Tomography and Immunogold Labelling as Tools to Analyse 135

Fig. 3. Segmentation of cell plate domains using the “Closed” object type option in 3dmod. Cell plate membranes are
traced as overlay contours in the Zap window and rendered as a meshed 3D object in the Model View window. Scale
bars = 100 nm.

Complete instructions for the use of the imodmesh command

can be found at http://bio3d.colorado.edu/imod/doc/man/
imodmesh.html (see Note 7). Be sure to save the model before
running imodmesh.

3.5.2. Vesicles To simplify the modelling of vesicles, it can be assumed that their
shape approximately correspond to a sphere. In this case, vesicles
can be modelled as “scattered points”, where a sphere is computed
from a single point placed in the centre of the vesicle, and its size
can be adjusted to match the diameter of the vesicle (Fig. 2b).
All similar vesicles (for example, all clathrin-coated vesicles) can be
contained in a single object.
1. Create a new object under the “Edit” menu and select the
“scattered” option.
2. Move along the stack of tomographic slices using the Zap
window and identify the centre of the vesicle.
3. Place a point in the centre of the vesicle and define the radius
by adjusting the “Sphere radius for points” option.
4. Spheres do not required to be meshed and can be directly
displayed as 3D objects in the “Model” window (Fig. 2b).
136 Otegui

3.5.3. Cell Plates Cell plates, endoplasmic reticulum, and any other membrane-bound
and Membrane-Bound organelle under analysis should be modelled as separate closed
Organelles objects. Non-connected portions of the cell plate or different
domains of the endoplasmic reticulum can be modelled as differ-
ent objects as well. When modelling membranous organelles,
it is recommended to place contours in the middle of the lipid
bilayer (1, 25).
1. Create a new object under the “Edit” menu and select the
“closed” option (Fig. 3).
2. Move along the stack of tomographic slices using the Zap
window and draw in each slice the outline of the cellular
structure being modelled (Fig. 3).
3. Save model and run the imodmesh command. imodmesh offers
a number of options for capping off objects, connecting con-
tours in non-adjacent tomographic slices, etc. (see Note 7).
For a complete reference of imodmesh options go to http://
bio3d.colorado.edu/imod/doc/man/imodmesh.html.

3.5.4. Quantitative Analysis One of the powerful advantages of ET is the possibility of obtain-
ing quantitative information from tomographic models. One can
analyse spatial relationships between vesicles and MTs, membrane
surface area changes during cell plate assembly, frequency and
distribution of MT plus ends in the phragmoplast (1, 3, 4). The
measurements are generally expressed in terms of the pixel size,
which is described in the model header as a number in nanome-
tres. The extent to which the section thinned in the electron beam
(“thinning factor”) is also described in the model header and is
applied to the calculations. Therefore, it is very important to enter
accurate values in the model header window (under the “Edit”
menu) before performing quantitative calculations.

3.5.5. Quantifying 1. Extract randomly located boxes of a defined size along the
the Volume and Surface cell plate.
Area of Cell Plates 2. Mesh the portions of cell plates enclosed in these boxes and
run the imodinfo command to calculate total surface and
volume of different modelled structures, using the option –s.
A complete description of all options available for the imodinfo
command can be found at http://bio3d.colorado.edu/imod/
doc/man/imodinfo.html (see Note 8).

3.5.6. Quantifying 1. Ensure that Subheading 3.2.1 all the individual items (for
the Density of Ribosomes, example, individual ribosomes) are points in the same con-
Vesicles, or Any Other tour of the same object.
Structures Modelled 2. Extract defined size boxes from the model and run the imodinfo
as Spheres command to obtain the number of points contained in the
box volume.
 Electron Tomography and Immunogold Labelling as Tools to Analyse 137

3.5.7. Spatial Relationships: To measure the distances between the objects in 3D and to
MicroTubule Kissing ­compute an average density of neighbouring items as a function
Analysis of distance between objects, one can use the microtubule kissing
(MTK) programme of the IMOD package (9, 26). This pro-
gramme considers structures of three kinds: open contours, like
MTs; scattered point objects, like vesicles; and meshed, closed
contour objects like cell plate domains. For MTs, each contour is
considered as a separate line. It is possible to calculate the spatial
relationship between any object and a whole MT, or to break the
MT into multiple segments and measure the closest distances
between a chosen object and each of the MT segments. Distances
can be measured from the central axis or surface of one object to
the central axis or surface of another. A complete explanation of
MTK and its commands can be found in http://bio3d.colorado.
edu/imod/doc/man/mtk.html.

3.6. Immunolabelling One of the main challenges in this type of imaging analysis is to
be able to identify the biochemical identity/composition of the
structures reconstructed in electron tomograms. One approach
to identify macromolecules in tomographic reconstructions is to
perform immunolabelling experiments. This can be done as a
correlative approach (the immunolabelling is performed in differ-
ent sections from the ones that are used for tomographic recon-
struction) or a direct approach (when the same sections are used
for both immunolocalisation and ET) (27). As an example, cor-
relative immunolabelling experiments with specific antibodies
suggested that the electron-dense rings constricting cell plate
membranes in endosperm cell plates are dynamin ADL1A/DRP1A
polymers (1) (Fig. 4).

Fig. 4. Identification of protein complexes in electron tomographic reconstructions of

Arabidopsis endosperm cell plates by correlative immunogold labelling. (a, b) Membrane-
associated rings in endosperm cell plates. (a) Tomographic slice and (a¢) corresponding
tomographic model of a cell plate tubule. Note the presence of an electron dense collar
(arrowheads) associated with a constricted tubule. The identity of the rings as dynamin
ADL1A/DRP1A polymers was determined by immunolabelling with specific antibodies
(b). Scales = 50 nm. Reproduced from Otegui et al. (1) (Copyright © 2001 American
Society of Plant Biologists).
138 Otegui

1. Float nickel or gold grids containing section on a drop of 5%

milk in PBT-T-0.1% for 20 min.
2. Transfer specimen grid to a drop of primary antibody diluted
in 5% milk in PBT-T-0.1% for 1 h.
3. Rinse the grid with a continuous stream of PBS-T-0.5% for
1 min.
4. Remove the excess liquid by touching the grid with filter
paper.
5. Place specimen grid in secondary antibody conjugated to
gold particles (5, 10, 15 nm in diameter) diluted 1:10 in 5%
milk in PBT-T-0.1% for 1 h.
6. Rinse the grid with a continuous stream of PBS-T-0.5% for
1 min, followed by rinsing with distilled water.
7. Post-stained grid with 2% uranyl acetate in 70% methanol and
Reynold’s lead citrate.

4. Notes

1. SerialEM is a free software package for image acquisition

(http://bio3d.colorado.edu/SerialEM) that is compatible
with FEI and JEOL electron microscopes. IMOD is a free
software package that runs in both PC and Macintosh sys-
tems and was developed primarily by David Mastronarde,
Rick Gaudette, Sue Held, and Jim Kremer at the Boulder
Laboratory for 3D Electron Microscopy of Cells. It contains
around 140 programmes for image processing, tomogram
calculation, image segmentation, display, and quantitative
analysis.
2. High-pressure frozen material can be stored in liquid nitro-
gen for months or even years without suffering changes in
cellular preservation.
3. OsO4 in acetone is highly volatile. Even when the acetonic
OsO4 solution is kept in closed cryovials, osmication of
objects around cryosubstitution vials can easily happen. If a
freezer/fridge is used during cryosubstitution with OsO4, it
is advisable to have a freezer/fridge fully dedicated to this use
to avoid contamination of other reagents and lab ware.
4. BDMA is recommended as the accelerator for the Eponate
12 resin mix instead of DMP-30 because BDMA has lower
viscosity and diffuses more rapidly into tissues.
5. Unpolymerised resin waste, dirty globes, and other resin-
contaminated items should be placed in the oven for 24 h for
polymerization before discarding them.
 Electron Tomography and Immunogold Labelling as Tools to Analyse 139

6. The use of methanolic solution of uranyl acetate is

­recommended because it increases the contrast of membrane
and cytoskeletal elements more than aqueous solutions do.
7. imodmesh generates triangles that connect neighbouring
points within a contour and nearby points on adjacent con-
tours. The resulting triangles represent a surface in space that
is an excellent approximation to all the contour information,
so they are used for all subsequent quantifications of area,
volume, and distance. They can also be used to generate a
shaded surface that provides a good visual representation of
the modelled object.
8. imodinfo provides information about IMOD models, such as
lists of objects, contours and point data, lengths and centroids
of contours, and surface areas and volumes of objects or
surfaces.

Acknowledgements

This work was supported by the National Research Initiative of

the USDA Cooperative State Research, Education and Extension
Service, grant number 2008-35304-18672 to M.S.O.

References

1. Otegui, M. S., Mastronarde, D. N., Kang, B. 6. Mastronarde, D. N. (1997) Dual-axis tomog-

H., Bednarek, S. Y., and Staehelin, L. A. raphy: an approach with alignment methods
(2001) Three-dimensional analysis of syncy- that preserve resolution. J Struct Biol 120,
tial-type cell plates during endosperm cellular- 343–352.
ization visualised by high resolution electron 7. Mastronarde, D. N. (2005) Automated elec-
tomography. Plant Cell 13, 2033–2051. tron microscope tomography using robust
2. Otegui, M. S., and Staehelin, L. A. (2004) prediction of specimen movements. J Struct
Electron tomographic analysis of post-meiotic Biol 152, 36–51.
cytokinesis during pollen development in 8. Mastronarde, D. N. (2008) Correction for
Arabidopsis thaliana. Planta 218, 501–515. non-perpendicularity of beam and tilt axis in
3. Segui-Simarro, J. M., Austin, J. R., II, White, tomographic reconstructions with the IMOD
E. A., and Staehelin, L. A. (2004) Electron package. J Microsc 230, 212–217.
tomographic analysis of somatic cell plate for- 9. Kremer, J. R., Mastronarde, D. N., and
mation in meristematic cells of Arabidopsis McIntosh, J. R. (1996) Computer visualiza-
preserved by high-pressure freezing. Plant tion of three-dimensional image data using
Cell 16, 836–856. IMOD. J Struct Biol 116, 71–76.
4. Austin, J. R., II, Segui-Simarro, J. M., and 10. Beemster, G. T. S., and Baskin, T. I. (1998)
Staehelin, L. A. (2005) Quantitative analysis Analysis of cell division and elongation under-
of changes in spatial distribution and plus-end lying the developmental acceleration of root
geometry of microtubules involved in plant- growth in Arabidopsis thaliana. Plant Physiol
cell cytokinesis. J Cell Sci 118, 3895–3903. 116, 1515–1526.
5. Otegui, M. S., and Austin, J. R., II (2007) 11. Otegui, M. S., and Staehelin, L. A. (2000)
Visualization of membrane-cytoskeletal inter- Cytokinesis in flowering plants: more than
actions during plant cytokinesis. Methods Cell one way to divide a cell. Curr Opin Plant Biol
Biol 79, 221–240. 3, 493–502.
140 Otegui

12. Otegui, M. S., and Staehelin, L. A. (2000) 20. Knox, J. P., Linstead, P. J., Peart, J., Cooper,
Syncytial-type cell plates: a novel kind of cell C., and Roberts, K. (1991) Developmentally
plate involved in endosperm cellularization of regulated epitopes of cell surface arabinoga-
Arabidopsis. Plant Cell 12, 933–947. lactan proteins and their relation to root tissue
13. Nagata, T., Nemoto, Y., and Hasezawa, S. pattern formation. Plant J 1, 317–326.
(1992) Tobacco BY-2 cell line as the “HeLa” 21. Smallwood, M., Yates, E. A., Willats, W. G. T.,
cell in the cell biology of higher plants. Int Martin, H., and Knox, J. P. (1996)
Rev Cytol 132, 1–30. Immunochemical comparison of membrane
14. Nagata, T., and Kumagai, F. (1999) Plant cell associated and secreted arabinogalactan-pro-
biology through the window of the highly teins in rice and carrot. Planta 198, 452–459.
synchronised tobacco BY-2 cell line. Methods 22. McEwen, B. F., and Frank, J. (2001) Electron
Cell Sci 21, 123–127. tomographic and other approaches for imag-
15. Kakimoto, T., and Shibaoka, H. (1988) ing molecular machines. Curr Op Neurobiol
Cytoskeletal ultrastructure of phragmoplast- 11, 594–600.
nuclei complexes isolated from cultured 23. Marsh, B. J. (2005) Lessons from tomo-
tobacco cells. Protoplasma S2, 95–103. graphic studies of the mammalian Golgi.
16. Samuels, L. A., Giddings, T. H., and Staehelin, Biochim Biophys Acta 1744, 273–292.
L. A. (1995) Cytokinesis in tobacco BY-2 and 24. O’Toole, E. T., Giddings, J. T. H., Dutcher,
root tip cells: a new model of cell plate forma- S. K., and McIntosh, J. R. (2007) Under­
tion in higher plants. J Cell Biol 130, standing microtubule organizing centers by
1345–1357. comparing mutant and wild type structures
17. Giddings, T. H. (2003) Freeze-substitution with electron tomography. Methods Cell Biol
protocols for improved visualization of mem- 79, 125–143.
branes in high-pressure frozen samples. J 25. Ladinsky, M. S., Mastronarde, D. N.,
Microsc 212, 53–61. McIntosh, J. R., Howell, K. E., and Staehelin,
18. Moore, P. J., Darvill, A. G., Albersheim, P., L. A. (1999) Golgi structure in three dimen-
and Staehelin, L. A. (1986) Immunogold sions: functional insights from the normal rat
localization of xyloglucan and rhamnogalac- kidney cell. J Cell Biol 144, 1135–1149.
turonan I in the cell walls of suspension-­ 26. Marsh, B. J., Mastronarde, D. N., Buttle, K.
cultured sycamore cells. Plant Physiol 82, F., Howell, K. E., and McIntosh, J. R. (2001)
787–794. Organellar relationship in the Golgi region of
19. Puhlmann, J., Bucheli, E., Swain, M. J., the pancreatic beta cell line, HIT-T15, visual-
Dunning, N., Albersheim, P., Darvill, A. G., ised by high-resolution electron tomography.
and Hahn, M.G. (1994) Generation of mono- Proc Natl Acad Sci U S A 98, 2399–2406.
clonal antibodies against plant cell-wall poly- 27. Donohoe, B. S., Kang, B. -H., and Staehelin,
saccharides. I. Characterization of a monoclonal L. A. (2007) Identification and characteriza-
antibody to a terminal alpha-(1®2)-linked tion of COPIa- and COPIb-type vesicle classes
fucosyl-containing epitope. Plant Physiol 104, associated with plant and algal Golgi. Proc
699–710. Natl Acad Sci U S A 104, 163–168.
 Chapter 10

Analysing Cellulose Biosynthesis with Confocal

Microscopy
Meera Nair and Seth DeBolt

Abstract
Plant cells are delimited by a rigid cell wall that resists internal turgor pressure, but extends with a remarkable
degree of control that allows the cell to grow and acquire specific shapes. Live cell fluorescence microscopy
systems have allowed an amazing view into the complex and dynamic lives of individual proteins during cell
morphogenesis. The current chapter will focus on methodology for live cell imaging of cellulose synthase
(CESA) in Arabidopsis, which will also provide a launching pad to explore ones specific protein of interest.

Key words: Cellulose synthesis, Live cell imaging, Confocal microscopy, CESA, AFP, YFP

1. Introduction

1.1. Background Once a challenging technique with limited accessibility, confocal

microscopy has rapidly evolved into a robust technique providing
proteome level localization data with quantitative precision (1–5).
The capacity to visualise your target protein relies on a fused
autofluorescent protein (AFP), which excites when laser is acti-
vated. Classic examples are green fluorescent protein (GFP)
derived from the jellyfish Aqueora victoriae being fused to your
plant protein of interest. Since the discovery of GFP 16 years ago
(1), numerous forms of AFP have been derived including blue
fluorescent protein (BFP), cyan fluorescent protein (CFP), yellow
fluorescent protein (YFP), and the mFruits collection (6–8).
Monomeric version of photoswitchable or DRfONPA (9) and
photoactivatable GFP exist (10). Designing an AFP expression
fusion can therefore be a daunting task as there are many consider-
ations including; which AFP, expression vectors, the choice of
where to fuse your AFP, what promoter to use, and whether to

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_10, © Springer Science+Business Media, LLC 2011

141
142 Nair and DeBolt

use stable or transient expression of your chimeric protein fusion.

Furthermore, fluorescence recovery after photobleaching (FRAP;
(11)), fluorescence resonance energy transfer (FRET; (12)) and
bimolecular fluorescence complementation (BiFC; (13)) could be
considered during experimental design depending on your applica-
tion. This chapter aims to provide researchers with a generalised
introduction to the materials and methods needed for performing
a live cell imaging analysis of proteins in the plant cell wall.
Live cell imaging has been developed to explore the location
of proteins that are fused to AFPs in living tissue by fluorescence
microscopy. These tools have been developed alongside sophisti-
cated advances in microscopy, specifically laser assisted confocal
microscopy that relies on precise spectral wavelengths produced
by different lasers. The morphology and motility of the protein-
AFP fusion can then be used to place a protein in a particular part
of the cell with the main limitation being the resolution of the
focal plane of the confocal microscope. Cell wall biosynthetic
proteins have been localised to the plasma membrane, cell wall,
the secretory system and the endomembrane system. The central
focus of this chapter is on the progression for cloning the gene,
sequence checking the gene, the myriad of factors involved in
selecting a vector for C or N terminal AFP fusion, selection of a
promoter to drive the fusion protein and the binary vector system
to introduce the AFP tagged gene into your plant either tran-
siently or stably. Where possible, the numerous pitfalls that must
be negotiated during protein localization experiments are high-
lighted. Once introduced into the transgenic plant, the AFP-
fusion protein can be imaged and a brief overview of imaging
systems and approaches to quantification has been provided.

1.2. Vector Selection The choice of promoter used to drive the target gene fused to an
AFP in the transgenic plant must be made. The two main choices
1.2.1. Promoter
are the use of a native promoter (defined as 1–2 kb of upstream
sequence from the target genes start codon), vs. a Cauliflower
mosaic virus (CaMV) 35S or double 35S constitutive promoter. If
one is interested in studying a single protein and there is a require-
ment for the AFP fusion to function as closely as possible to the
native protein with respect to localization, tissue-specificity, tim-
ing, and level of expression, then it may be necessary to express
the fusion in transgenic plants under control of the native pro-
moter. Moreover, if ones protein of interest can be knocked out in
a model organism, for instance in the model plant Arabidopsis,
one can functionally complement the knockout allele with the
AFP-fusion protein as a means to check that mutant phenotypes
are restored to that of wild-type plants. This experiment infers
functionality of the chimeric protein. At the other extreme, if there
is a need to localise thousands of fusions in a relatively short period
of time, then transient expression may be the most cost effective
as would a high throughput approach with a constitutive promoter
 Analysing Cellulose Biosynthesis with Confocal Microscopy 143

such as 35S using a Gateway® cloning (Invitrogen, Carlsbad, CA)

system, using vectors such as pSITE (14) or pMDC (15).
Some consider that expression under the control of native
promoters will be always superior to that of employing constitu-
tive promoters. However, simply using the native promoter, or
more often 1–2 kb of “upstream” sequence, ignores the fact that
promoter/gene duplication may affect expression levels, as will
genomic context since the AFP fusion is unlikely to be expressed
from the same genetic locus as the native gene (13). It also appears
to be a common belief that expression from 35S, or even double
35S, promoters necessarily results in accumulation of fusions pro-
teins of the levels higher than native proteins. However, all such
results are highly dependent upon the protein under investiga-
tion. Fusion of an AFP to a protein may stabilise it. In systems
such as Arabidopsis, where it is straightforward to obtain T-DNA
insertion alleles for your gene of interest (16), single gene com-
plementation by your AFP-fusion protein provides some confi-
dence of correct functionality of your fusion protein in planta.

1.2.2. Where to Fuse Your Where to fuse your AFP? Should one use an amino (N) or a car-
AFP: Amino (N), Carboxy (C), boxy (C) terminal fusion? While N or C terminal fusions are the
or Internal most common, and easiest to construct, some proteins may not
tolerate AFPs fused to a terminal end and it may be necessary to
insert the AFP into an internal site. Moreover, if your target protein
contains an N-terminal signal peptide it may result in mislocalisa-
tion if expressed as fusions to the C-termini of AFPs (17). To over-
come this, computational methods have been developed to predict
the effect of an AFP on a particular fusion (17). Most binary expres-
sion vector systems, sometimes referred to as destination vectors,
have been developed to suite N- or C-terminal fusions (18–20). To
head off any potential problems when testing an AFP fusion, both
C and N terminal fusions can be made for preliminary studies using
a transient expression system prior to stable transformation or
detailed quantitative imaging analysis. When considering the vector
system you aim to use a question is, whether the researcher will use
a Gateway® cloning (Invitrogen, Carlsbad, CA) or non-gateway
restriction enzyme based cloning such as pCAMBIA vectors
(CAMBIA Corporation, Canberra, Australia). Gateway® cloning
technology can be particularly useful for both high throughput
studies, as described above, and single protein studies due to its
robust and accurate cloning and most of the current vector systems
employ this technology. Both of these vector systems have C and N
terminal AFP variants or versions with no AFP to allow the user to
insert their AFP internally. For Gateway® cloning, the pMDC and
pSITE vector systems are good example and for restriction enzyme
based, the use of pCAMBIA is ideal for plant-based expression.

1.2.3. Choice of AFP The demonstration that the GFP (1) isolated from the jellyfish
Aqueora victoriae, could be linked to proteins of interest in order
144 Nair and DeBolt

to allow in vivo examination of protein localization and dynamics

in real-time has transformed cell biology in a manner similar to
the effect of the polymerase chain reaction on molecular biology.
Numerous AFP spectral variants are available in colours such as
red fluorescent protein (RFP) (21), YFP (8) and now banana,
orange, cherry, tomato, and plum in the mFruits collection (6).
Recently, a novel monomeric red fluorescent protein, TagRFP,
has been described (22). This protein is brighter and more resis-
tant to photobleaching than mRFP and can be used in combina-
tion with GFP in FRET experiments. To this end, despite GFP
being a popular AFP tag for creating fusion proteins, currently
the most common FRET pair is CFP/YFP. For the case study
described herein, YFP was chosen to fuse to cellulose synthase
(CESA) because this allowed for CFP to be fused to TUBULIN.
Spectral properties of YFP and CFP allow their excitation and
thus visualisation in different channels allowing the simultaneous
view of two fluorophors in a single plant cell (23).

1.3. Transformation There is one main question for the choice of transformation for a
of Your AFP Fusion live cell imaging experiment: are you going to use a stable transfor-
into a Plant mation or a transient one? This decision is not a trivial one. In high
throughput circumstances, creating stable transformations is time
consuming and can be restrictive depending on laboratory resources.
Therefore, for proteome-scale projects transient assays are often
preferred. Transient assay systems utilise agroinfiltration of your
AFP fusion into Nicotiana benthamiana leaves. This simple tech-
nique employs injecting an Agrobacterium (expressing your AFP
fusion) solution directly into the underside of a leaf blade, inoculat-
ing for 2–3 days and then imaging the living leaf tissue. Alternatively,
stable transformation in Arabidopsis thaliana utilises Agrobacterium
mediated floral dipping methods (24) to introduce your AFP fusion
into the genome as a T-DNA insertion. The advantage of stable
transformation is the capacity to localise your target protein in
numerous tissues and developmental stages. The presence of the
AFP fusion protein in the plant can then be checked by western
blot using antibodies against your chosen AFP (readily available
from common suppliers such as Sigma-Aldrich, St Louis MO).

2. Materials

2.1. Polymerase Chain 1. Proof reading enzyme for amplification, Platinum®PfxDNA

Reaction (PCR) Polymerase, stored at −20°C (Invitrogen, Carlsbad, CA).
to Amplify Gene 2. 5¢ and 3¢ Primers specific for promoter and gene for CESA.
of Interest 100 mM stock is diluted to 10 mM sub stock to be used as a
with Promoter working solution. These are stored at −20°C (see Note 1).
and Subcloning 3. 10× Pfx Buffer 25 mM MgSO4 ,100 mM dNTPs (Invitrogen
into an Entry Vector Carlsbad, CA), stored at −20°C (see Note 2).
 Analysing Cellulose Biosynthesis with Confocal Microscopy 145

4. Template as genomic DNA (gDNA) (500 ng/mL−1), stored

at −20°C (see Note 3).
5. pENTR® dTOPO® vector system (Invitrogen, Carlsbad, CA),
stored at −20°C.

2.2. Cloning of Your 1. pSITE2NA destination vector (see Note 4), stored at
Gene of Interest −20°C.
into a Compatible 2. Gateway® cloning kits (Invitrogen, Carlsbad, CA) stored at
Destination Vector −20°C.
to make AFP Fusions
3. Agrobacterium competent cells, stored at −80°C (Invitrogen,
and Introducing Carlsbad, CA).
this Into the Plant
by Agrobacterium
Mediated
Transformation

2.3. SDS- 1. Running buffer (5×): 125 mM Tris, 960 mM glycine, 0.5%
Polyacrylamide Gel (w/v) SDS. Store at room temperature.
Electrophoresis 2. 2× SDS sample loading buffer: 100 mM Tris–Cl pH 6.8,
(SDS-PAGE) 4% v/v SDS, 0.2% bromophenol blue, 20% w/v glycerol,
stored at room temperature.
3. Pre-stained molecular weight markers: Kaleidoscope markers
(Bio-Rad, Hercules, CA).
4. 0.5 M Tris-Cl buffer pH 8.
5. 100 mM Phenylmethanesulfonyl fluoride (PMSF; Sigma
Aldrich, St. Louis)) in isopropanol, stored at −20°C.

2.4. Western Blotting 1. Setup buffer: 25 mM Tris (do not adjust pH), 190 mM gly-
for YFP cine, 20% (v/v) methanol, stored at room temperature.
2. Transfer buffer: Setup buffer with the added inclusion of
0.05% (w/v) SDS. Store in the transfer apparatus at room
temperature (see Note 5).
3. Supported nitrocellulose membrane from Millipore, Bedford,
MA, and 3MM Chromatography paper from Whatman,
Maidstone, UK (see Note 6).
4. Tris-buffered saline with Tween (TBS-T): Prepare 10× stock
solution containing 1.37 M NaCl; 27 mM KCl; 250 mM
Tris–HCl, pH 7.4; 1% Tween-20.
5. Tris-buffered saline with Tween (TBS-T): Prepare 1× stock
for use by diluting 100 mL of 10× stock in 900 mL water.
6. Blocking buffer: 5% (w/v) non-fat dry milk in TBS-T.
7. Primary antibody dilution buffer: TBS-T supplemented with
2% (w/v) fraction bovine serum albumen (BSA).
8. Anti-YFP monoclonal antibody (available from Sigma Aldrich,
St. Louis) stored at −20°C (see Note 7).
146 Nair and DeBolt

9. Secondary antibody: Anti-mouse IgG conjugated to horse

radish peroxidase (available from Sigma Aldrich, St Louis)
stored at −20°C (see Note 7).
10. Chemiluminescence detection kit (SuperSignal West Pico
Chemiluminescent Substrate (Pierce Biotechnology,
Rockford, IL) stored at 4°C.
11. Detection using Bio-Rad ChemiDoc XRS+® (BioRad,
Hercules, CA).

2.5. Imaging Analysis 1. Cover slips 48 × 60, 24 × 60 (Menzel, Braunschweig Germany).

2. Fine point tweezers #5 Dumont (Electron Microscopy
Sciences, Hatfield, PA).
3. Dow Corning High Vacuum Grease (Specialty Fluids Co,
Valencia, CA).
4. Sterile water.

3. Methods

Herein, we describe a Gateway® cloning for tagging CESA to YFP

and the use of confocal microscopy to visualise it in plants. Gateway-
compatible binary vectors have greatly improved the cloning effi-
ciency of AFP tagging projects (15, 19). Briefly, Gateway® cloning
uses the lambda phage site-specific recombination system in order
to transfer DNA fragments between plasmids containing compat-
ible recombination sites (26). What makes this strategy so attrac-
tive is that once DNA clones of interest are captured into an entry
vector (pENTR/pDONR), they can be mobilised into a plethora
of destination vectors that permit expression in bacteria, insects,
yeasts, or plants but also allow for introducing the target gene into
an AFP vector with RFP, YFP, CFP, or GFP fluorophors.
Numerous factors influence imaging, for instance, the cell
type being imaged, whether your experiments is simply to deter-
mine the location of your target protein, or whether you wish to
track the dynamic behaviour of the your target protein by time
lapse imaging. The best results have been obtained by localising
protein dynamics in upper regions of the hypocotyls of etiolated
seedlings grown in a vertical position on Murashige-Skoog (MS)
agar plates for 2.5–4 days at room temperature (22°C), mounted
between cover slips in water and then imaged (23, 27).

3.1. Polymerase Chain 1. Independent amplification of the CESA and promoter or

Reaction (PCR) your gene of interest achieved using a PCR reaction com-
to Amplify Gene posed of 1.5 pmol of each primer, 0.3 mM dNTPs, 1× Pfx
of Interest Buffer, 1 mM MgSO4 (see Note 8) and 1.25 units
and Subcloning Platinum®PfxDNA Polymerase (Invitrogen, Carlsbad, CA),
into an Entry Vector 1 mL of gDNA, made to 20 mL with deionised H2O (dH2O).
 Analysing Cellulose Biosynthesis with Confocal Microscopy 147

The following thermocycler reaction times and temperature

can be used, a 3-min heating at 95°C was followed by 32
cycles of 95°C (30 s) denaturation, 55°C (30 s) annealing
temp (see Note 9), 1 min per kilobase product size extension
times, and a final extension of 7 min.
2. Gateway® cloning of the PCR product into an entry vector is
then achieved based on manufacturer’s instructions
(Invitrogen, Carlsbad, CA).

3.2. Cloning of Your 1. The CESA gene in entry vector can then be cloned in to a
Gene of Interest destination vector such as pSITE2N using the manufacturer’s
with Promoter manual for Gateway® cloning (Invitrogen, Carlsbad, CA).
into a Compatible The double CAMV 35S promoter can be excised using com-
Destination Vector patible restriction enzymes and the amplified promoter region
to make AFP Fusions ligated into the vector by directional cloning. This creates an
and Introducing additional step in the cloning and will depend on whether
this into the Plant you wish to examine localization of your target gene with a
by Agrobacterium constitutive or native promoter.
Mediated 2. At this stage a sequence verified version of your target gene
Transformation (see Note 10), for instance pSITE2NproCESA::CESA (see
Note 11), can then be transformed into electro or chemically
competent Agrobacterium tumefaciens cells according to
manufacturer’s manual (Invitrogen, Carlsbad, CA).
3. Introducing the transgene into the plant should follow pub-
lished protocols (24) (see Note 12).
4. Selection of stable transformants in A. thaliana will be
achieved by growing the T1 progeny on sterile 0.5 strength
MS media supplemented with 50 mg/mL−1 Kanamycin
(Kan50). Plants able to grow on the Kan50 media are then
transferred to soil and grown.

3.3. SDS Page For simplicity, this method assumes the use of Criterion cell and
and Western Blotting precast gels from BioRad® (Hercules, CA) as well as the BioRad
for Detection of YFP Trans-Blot® Electrophoretic Transfer Cell®.
in Plant Tissue 1. Total protein from individual transformants is extracted by
simply grinding 2–3 leaves using liquid nitrogen in mortar
and pestle and 300 mL 0.5 M Tris-Cl buffer with 0.1 mM
final concentration PMSF. 1 volume of 2× SDS loading buf-
fer is added to 1 volume of total protein and boiled for
5 min. It is then carefully loaded into the well of the precast
gel along with molecular weight markers and the gel is run
for 1 h at 80 mA in the Criterion cell system (BioRad®
Hercules, CA).
2. After protein separation by SDS-PAGE the samples are trans-
ferred to a nitrocellulose membrane according to the manu-
facturer’s instructions (assuming the use of a BioRad
Trans-Blot® Electrophoretic Transfer Cell®) (see Note 6).
148 Nair and DeBolt

3. The coloured molecular weight markers should be clearly

­visible on the membrane. Nitrocellulose membrane is care-
fully transferred in a solution of blocking buffer for 1 h at
room temperature on a shaker.
4. The nitrocellulose membrane is rinsed and then immersed in
a solution of 1:10,000 dilution of anti-YFP antibody in
TBST/2% BSA for 1 h on a rocking platform (see Note 13).
5. Wash three times for 5 min each with 50 mL TBS-T and then
immerse in 1:30,000-fold dilution of the secondary antibody
as above. Wash again 3 times for 10 min each with TBS-T.
Then incubate in equal volumes of Pico-West (total of 4 mL)
(Pierce Biotechnology, Rockford, IL) solution for 1 min, seal
in a plastic sleeve and examine chemiluminescent signal in the
BioRad® GelDoc. Select transgenic plants that show the pres-
ence of YFP for confocal imaging (see Note 14).

3.4. Preparation 1. Once stable transformants expressing YFP::CESA fusion are

of Plants for Imaging identified and transferred to soil, the plants are grown to
obtain seeds from them.
2. The seeds once ready are harvested in a newspaper piece and
dried in a dry corner of the lab for 3–4 days. They are then
sterilised using 30% bleach and 5% SDS for surface sterilisa-
tion. The seeds are washed thoroughly to remove sterilisation
solution using sterile water. It is then resuspended in 0.15%
Agar and stored in dark at 4°C for vernalisation for 3 days
(see Note 15).
3. A thaliana seedlings are grown in dark for 2.5 days on 0.5
strength MS-agar plates at 22°C.

3.5. Imaging Analysis 1. Single seedlings are gently removed from 0.5 strength
MS-agar plates and mounted in an aqueous solution between
a 48 × 60 and a 24 × 60 cover slip.
2. The silicon vacuum grease is carefully applied to the perime-
ter of the 24 × 60 cover slips to avoid any water loss or evapo-
ration over the duration of imaging and essentially avoids
compression of the epidermal cells of the tissue being imaged
(see Note 16).
3. For imaging YFP::CESA, a purpose built spinning disc confo-
cal microscope using Leica X 63 N.A. = 1.4 oil immersion
objective and Roper Cascade 512b EMCCD camera can be
used (see Note 17).
4. The confocal plane is focused then on the plasma membrane
focal plane with an exposure of 600 ms for YFP::CESA
(Fig. 1a). A method to improve the signal to background
noise ratio is to average multiple frames using the frame aver-
aging feature in the imaging software being used.
 Analysing Cellulose Biosynthesis with Confocal Microscopy 149

Fig. 1. Live-cell imaging of YFP::CESA (a) Plasma membrane focal plane (b) X and Y drift
during time lapse image acquisition can reduce image quality and make it difficult to
track protein dynamics (c) Z drift during time lapse image acquisition can reduce image
quality and distort protein dynamics analyses.

5. YFP is excited at 488 nm and data is collected through a

525.50 nm band pass filter (Chroma Technologies,
Brattleboro, VT) (23). The image can be acquired using
image acquisition software and analysed (see Note 18).
6. To obtain multiple frames of a single cell to provide a time-
lapse series (i.e. a movie), seek the time-lapse feature in the
image acquisition software being used (see Note 18) (select
the number of frames to be acquired and the amount of time
between each frame) (see Note 19).

4. Notes

1. When the primers arrive dilute them to 100 mM concentra-

tion using TE (10 nM Tris-Cl, 1 mM EDTA) buffer and
make a 10-mM working solution for yourself.
2. Each of the dNTPs is available as a 100 mM stock which can
be combined as a 10-mM sub stock for your use.
3. Here, the cloning strategy calls for the amplification of cellu-
lose synthesising enzyme or CESA with the promoter, thus
gDNA needs to be used to include introns as well as exons
and regulatory regions. gDNA preparation can be done using
the established protocols (25).
4. pMDC and pSITE vectors can be requested from the authors
(14, 15).
150 Nair and DeBolt

5. The transfer buffer should be cooled below room temperature

before you begin to do the transfer. This can be done by
­cooling the buffer at 4°C prior to doing western blots.
6. Nitrocellulose membranes should not be touched and should
not be handled without using gloves. It would be best to use
tweezers to handle the membrane at all times.
7. Primary and secondary antibodies should be aliquoted into
200 mL aliquots and stored at −20°C. When required, a vial
should be retrieved and used and the remaining can then be
stored at 4°C for multiple uses. This way the main stock will
not be contaminated if there is a chance of doing so.
8. MgSO4 levels can be optimised for your specific PCR reaction
through a little research.
9. The annealing temperatures have to be optimised for a spe-
cific pair of primers; usually a thumb rule of 1° than the melt-
ing temperature for the primers is used.
10. It is usually better to sequence your target gene for missense
mutations at the DNA level that confer changes in the amino
acid sequence of the target protein (Sequencing reactions
such as BigDye® (Applied Biosystems, Carlsbad, CA) and
nested primer design; for sequencing are not covered in this
chapter.
11. Amplification for C-terminal fusions (for purpose of this
chapter, this refers to the AFP fusion being fused to the
C-terminal end of the target gene) remove the stop codon
from the target gene in the 3΄ primer. For N-terminal fusions,
the AFP is referred to as being fused to the N-terminus of the
target gene.
12. You can also do Agrobacterium infiltration into Nicotiana
benthamiana (14) for transient expression.
13. An overnight addition of primary antibody can be done in
4°C cooler with a rocking bottom or shaker.
14. If no signal is obtained for YFP in transgenic plants in stable
transformants (A. thaliana), check for YFP in all plant sam-
ples by PCR in gDNA; if positive, select additional lines for
protein analysis by western blotting in the next generation or
if negative, retransform and proceed from there.
15. Seed sterilisation should be done under the hood to minimise
the chance of contamination. It is always worthwhile to use
only less than half of your seed stock so that if a mistake is
made in subsequent steps, there are always some transformant
seeds available to begin again.
16. For transient expression, a similar strategy is employed,
whereby a portion of the infiltrated N. benthamiana is physi-
cally removed from the leaf and mounted as described above.
 Analysing Cellulose Biosynthesis with Confocal Microscopy 151

17. Numerous confocal microscope systems are available and are

equally attractive for live-cell imaging experiments. The lead-
ing companies and examples of their confocal systems as of
January 2010 are Leica (Leica SP5 AOBS Point Scanning
Spectral Confocal Microscope), Olympus (Olympus Fluo View
FV1000MPE, a multiphoton laser scanning microscope),
Yokagawa (Yokogawa CSU-10 spinning disc confocal micro-
scope) Zeiss 780NLO Laser-Scanning Confocal Microscope.
18. Softwares for imaging analysis: Metamorph – Molecular
Devices (Sunnyvale, CA), which can be used for microscope
automation and image acquisition. LAS Image Analysis
optional software module developed by Leica Microsystems
(Bannockburn, IL) provides sequence control assisting
acquiring, detecting, and measuring multiple image features.
In addition to Leica, all major companies have similar systems
(see local representative). ImageJ is a Freeware software pro-
gramme that has numerous features developed by a world-
wide community of cell biologists to streamline quantitative
analysis of confocal imaging results. In addition to the numer-
ous image analysis tools built into ImageJ, there are Plugins
for tracking position over time by Kymograph, reslicing of
stacks, 3D reconstruction, time-stamping, conversion of
BioRad stacks to Quicktime format as well as most format
conversion, 3D rendering and quantitation (ImageJ, National
Institute of Health, Bethedsa, MD).
19. The main pitfall for time-lapse image collection is drift (x, y or
z drift). X and Y plane drift can be overcome fairly easily by
allowing the drift to occur and then use the Stackreg algo-
rithm ImageJ plugin (http://rsbweb.nih.gov/ij/plugins)
after acquisition to correct for the drift (Fig. 1a compared to
Fig. 1b). Z drift on the other hand cannot be corrected for
and therefore requires manual adjustment during acquisition
(Fig. 1c).

References

1. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. ­ igh-throughput fluorescent tagging of
H
W., and Prasher, D. C. (1994) Green fluores- ­full-length Arabidopsis gene products in
cent protein as a marker for gene expression. planta. Plant Physiol 135, 25–38.
Science 263, 802–805. 4. Heazlewood, J. L., Verboom, R. E., Tonti-
2. Cutler, S. R., Ehrhardt, D. W., Griffitts, J. S., Filippini, J., Small, I., and Millar, A. H. (2007)
and Somerville, C. (2000) Random SUBA: The Arabidopsis subcellular database.
GFP::cDNA fusions enable visualization of Nucleic Acids Res 35, D213–D218.
subcellular structures in cells of Arabidopsis at 5. Moore, I. and Murphy, A. (2009) Validating
a high frequency. Proc Natl Acad Sci USA 97, the location of fluorescent protein fusions in
3718–3723. the endomembrane system. Plant Cell 21,
3. Tian, G. W., Mohanty, A., Chary, S. N., Li, S., 1632–1636.
Paap, B., Drakakaki, G., Kopec, C. D., Li, J., 6. Shaner, N. C., Campbell, R. E., Steinbach, P. A.,
Ehrhardt, D., Jackson, D., et al (2004) Giepmans, B. N., Palmer, A. E., and ­Tsien, R. Y.
152 Nair and DeBolt

(2004) Improved monomeric red, orange and expression of autofluorescent protein fusions
yellow fluorescent proteins derived from in plants: probing Nicotiana benthamiana-
Discosoma sp. red fluorescent protein. Nat virus interactions. Mol Plant Microbe Interact
Biotechnol 22, 1567–1572. 20, 740–750.
7. Shaner, N. C., Steinbach, P. A., and Tsien, R. 19. Earley, K. W., Haag, J. R., Pontes, O., Opper,
Y. (2005) A guide to choosing fluorescent K., Juehne, T., Song, K., and Pikaard, C. S.
proteins. Nat Methods 2, 905–909. (2006) Gateway-compatible vectors for plant
8. Zhang, J., Campbell, R. E., Ting, A. Y., and functional genomics and proteomics. Plant J
Tsien, R. Y. (2002) Creating new fluorescent 45, 616–629.
probes for cell biology. Nat Rev Mol Cell Biol 20. Tzfira, T., Tian, G. W., Lacroix, B., Vyas, S.,
3, 906–918. Li, J., Leitner-Dagan, Y., Krichevsky, A.,
9. Habuchi, S., Ando, R., Dedecker, P., Taylor, T., Vainstein, A., and Citovsky, V.
Verheijen, W., Mizuno, H., Miyawaki, A., and (2005) pSAT vectors: a modular series of plas-
Hofkens, J. (2005) Reversible single-molecule mids for autofluorescent protein tagging and
photoswitching in the GFP-like fluorescent expression of multiple genes in plants. Plant
protein Dronpa. Proc Natl Acad Sci USA Mol Biol 57, 503–516.
102, 9511–9516. 21. Matz, M. V., Fradkov, A. F., Labas, Y.A.,
10. Patterson, G. H., and Lippincott-Schwartz, J. Savitsky, A. P., Zaraisky, A. G., Markelov,
(2004) Selective photolabeling of proteins M.L., and Lukyanov, S.A. (1999) Fluorescent
using photoactivatable GFP. Methods 32, proteins from nonbioluminescent Anthozoa
445–450. species. Nat Biotechnol 17, 969–973.
11. Bates, I. R., Wiseman, P. W., and Hanrahan, J. 22. Merzlyak, E. M., Goedhart, J., Shcherbo, D.,
W. (2006) Investigating membrane protein Bulina, M. E., Shcheglov, A. S., Fradkov, A. F.,
dynamics in living cells. Biochem Cell Biol 84, Gaintzeva, A., Lukyanov, K. A., Lukyanov, S.,
825–831. Gadella, T. W., et al (2007) Bright monomeric
12. Hink, M. A., Bisselin, T., and Visser, A. J. red fluorescent protein with an extended fluo-
(2002) Imaging protein-protein interactions rescence lifetime. Nat Methods 4, 555–557.
in living cells. Plant Mol Biol 50, 871–883. 23. Paredez, A. R., Somerville, C. R., and
13. Bhat, R. A., Lahaye, T., and Panstruga, R. Ehrhardt, D. W. (2006) Visualization of cel-
(2006) The visible touch: in planta visualization lulose synthase demonstrates functional asso-
of protein-protein interactions by fluoro- ciation with microtubules Science 312,
phore-based methods. Plant Methods 2, 12. 1491–1495.
14. Goodin, M. M., Chakrabarty, R., Banerjee, 24. Clough, S. J. and Bent, A. (1998) Floral dip:
R., Yelton, S., and DeBolt, S. (2007) New a simplified method for Agrobacterium-
gateways to discovery. Plant Physiol 145, mediated transformation of Arabidopsis thali-
1100–1109. ana. Plant J. 16, 735–743.
15. Curtis, M. D. and Grossniklaus, U. (2003) 25. Lukowitz, W., Gillmor, C. S., and Scheible,
A gateway cloning vector set for high- W. R. (2000) Positional cloning in Arabidopsis
throughput functional analysis of genes in Why it feels good to have a genome initiative
planta. Plant Physiol 133, 462–469. working for you. Plant Physiol 123,
16. Alonso, J. M., Stepanova, A. N., Leisse, T .J., 795–805.
Kim, C. J., Chen, H., Shinn, P., Stevenson, D. 26. Walhout, A. J., Temple, G. F., Brasch, M. A.,
K., Zimmerman, J., Barajas, P., Cheuk, R., Hartley, J. L., Lorson, M. A., van den Heuvel,
et al. (2003) Genome-wide insertional muta- S., and Vidal, M. (2000) GATEWAY recom-
genesis of Arabidopsis thaliana. Science 301, binational cloning: application to the cloning
653–657. of large numbers of open reading frames or
17. Simpson, J. C., Neubrand, V. E., Wiemann, ORFeomes. Methods Enzymol 328, 575–592.
S., and Pepperkok, R. (2001) Illuminating 27. Gutierrez, R., Lindeboom, J. J., Paredez, A.
the human genome. Histochem Cell Biol 115, R., Emons, A. M. C., and Ehrhardt, D. W.
23–29. (2009) Arabidopsis cortical microtubules
18. Chakrabarty, R., Banerjee, R., Chung, S. M., position cellulose synthase delivery to the
Farman, M., Citovsky, V., Hogenhout, S. A., plasma membrane and interact with cellulose
Tzfira, T., and Goodin, M. (2007) pSITE synthase trafficking compartments. Nat Cell
vectors for stable integration or transient Biol 11, 797–806.
 Chapter 11

Visual Mapping of Cell Wall Biosynthesis

Yumiko Sakuragi, Morten H.H. Nørholm, and Henrik V. Scheller

Abstract
Biosynthesis of pectin and hemicelluloses occurs in the Golgi apparatus and is thought to involve spatial
regulations and complex formation of biosynthetic enzymes and proteins. We have demonstrated that a
combination of heterologous expression of recombinant proteins tagged with fluorescent proteins and
live cell imaging with confocal laser scanning microscopy (CLSM) allows efficient visualization of biosyn-
thetic enzymes and proteins in subcellular compartments. We have also successfully utilized bimolecular
fluorescence complementation (BiFC) for in situ visualization of protein–protein interactions of pectin
biosynthetic enzymes and for the determination of their membrane topology in the Golgi apparatus.

Key words: Bioimaging, Cell wall biosynthesis, Uracil-excision cloning, Bimolecular fluorescence
complementation, Protein–protein interaction, Subcellular localization, Glycosyltransferase, Pectin

1. Introduction

1.1. Background Plant cells are surrounded by a strong wall composed primarily of
polysaccharides. The wall is essential for giving strength and shape
to plant cells, thereby allowing plants to stand upright, and in
addition serves as a barrier against invading pathogens. Cell wall
polysaccharides are traditionally grouped into cellulose, hemicel-
luloses, and pectin. Although these polysaccharides can have a
very complex structure, the basic structural features have been
reasonably well described. In contrast, the biosynthetic apparatus
responsible for synthesising the wall is poorly understood and its
identification and characterization is one of the major challenges
in plant biochemistry.
Pectin and hemicellulose biosynthesis has been shown to
occur in the lumen of the Golgi apparatus, although it cannot be
excluded that some early steps in the biosynthesis take place in the
endoplasmic reticulum (ER). Hence, most, if not all of the proteins,

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_11, © Springer Science+Business Media, LLC 2011

153
154 Sakuragi, Nørholm, and Scheller

involved in these biosynthetic processes must be localized in the

Golgi lumen either as soluble proteins or as embedded in the sur-
rounding membranes. Proteins involved include glycosyltrans-
ferases that catalyze transfer of activated sugar molecules to accep-
tor molecules, nucleotide sugar inter-converting enzymes that
supply the substrates to these glycosyltransferases, and nucleotide
sugar transporters and modifying enzymes (e.g. acetyltransferases
and methyltransferases). Glycosyl hydrolases may also take part in
the biosynthetic process, and they may function in the Golgi or
after secretion of precursor polysaccharides to the apoplast. The
majority of glycosyltransferases are predicted Type II membrane
proteins with the catalytic domains in the Golgi lumen. Important
exceptions to this are the glycosyltransferases belonging to the
Cellulose Synthase Like group of proteins. These are multi-span-
ning membrane proteins, and it is not yet clear on which side of the
Golgi membrane the catalytic domain is found. For recent reviews
on biosynthesis of hemicelluloses and pectin see refs. (1–4).
In the protein glycosylation pathways in animals, and in cel-
lulose and callose synthesis in plants, protein–protein interactions
act as an important organising principle with regard to biosyn-
thetic coordination, subcellular localisation, and direct regulation
of enzymatic activity (5–7). A school of thought exists that pro-
teins involved in biosynthesis of a particular pectin or hemicellu-
lose polymer form protein complexes or loose metabolons inside
the Golgi for optimal and coordinated actions of the enzymes.
However, until recently, direct evidence for such complexes had
been lacking in pectin and hemicellulose biosyntheses, largely due
to our limited knowledge of polysaccharide biosynthesis in
general. For example, overexpression of ARAD1, a putative
arabinosyltransferase, and XGD1, a putative xylosyltransferase, in
Arabidopsis did not lead to more arabinan or xylogalacturonan,
respectively (8, 9). This could suggest that these proteins func-
tion in biosynthetic complexes and that another component was
limiting in the over-expression studies. However, several other
explanations are possible, e.g. substrate limitation or feedback
regulation. Another study showed that mutation of one isoform
of UDP-glucose-4-epimerase resulted in a lack of galactose in ara-
binogalactan proteins and in a specific xyloglucan fraction,
whereas pectic galactan was unaffected (10). This result suggests
that specific epimerase isoforms are associated with a subset of
glycosyltransferases. UDP-arabinopyranose mutase, also known
as reversibly glycosylated protein, has been shown to form het-
erodimeric complexes (11, 12), although it is not yet clear how
these proteins are linked to polysaccharide biosynthesis. The most
compelling evidence for the role of protein complexes in biosyn-
thesis of pectin has been provided by Mohnen and colleagues.
GAUT1, a galacturonosyltransferase that is responsible for
homogalacturonan biosynthesis, has been shown to interact with
 Visual Mapping of Cell Wall Biosynthesis 155

GAUT7, a homolog of GAUT1 with no demonstrated enzymatic

activity (3, 13). This interaction was found to be essential for the
accumulation of GAUT1 in the Golgi apparatus (Atmodjo,
Sakuragi, Scheller, Mohnen, unpublished).
Rapid progress in determining genome sequences of plants
has led to the discovery of a large repertoire (ca. 1,500) of puta-
tive proteins involved in cell wall biosynthesis on the basis of their
sequence similarity to known proteins and their predicted subcel-
lular localizations. Experimental validation of the predicted sub-
cellular localisations, membrane topology, and protein complex
formation of these proteins remain to be carried out. Rapid meth-
ods for expression and visualisation of cell wall biosynthetic pro-
teins in planta are described here.

1.2. General Principles The protein of interest is recombinantly fused to fluorescent pro-
of the Methods teins (green fluorescent protein (GFP) and its variants) and tran-
siently expressed in the tobacco plant Nicotiana benthamiana.
Highly efficient recombinant gene technologies used in our labo-
ratories are Gateway® and uracil-excision cloning methods. A
detailed protocol of Gateway® cloning is made available by the
manufacturer (Invitrogen) and therefore is not discussed here.
The uracil-excision cloning method is a method for simple, yet
versatile ligase-independent cloning (14, 15). It requires only few
reagents and enzymes, and these can be purchased from commercial
vendors. Briefly, our compatible vectors contain cloning sites
­consisting of PacI and Nt. BbvCI restriction endonuclease recog-
nition sequences and are digested by these enzymes to generate
unique 3´-octanucleotide overhangs allowing directional inser-
tion of the genes of interests. A gene to be inserted is amplified by
the polymerase chain reaction (PCR) with synthetic oligonucle-
otides, which, in addition to the sequence specific to the target
DNA, contain uracil-containing tails. Upon treatment with
USER™ enzymes (New England Biolab), which is a mixture of
uracil DNA glycosylase and DNA glycosylase-lyase Endo VIII,
the uracils are excised, which results in 3´-overhangs complemen-
tary to those in the digested vector. Annealing between the insert
and the digested vector is stable during chemical transformation
of Escherichia coli and becomes covalently linked in vivo. There are
various approaches by which constructions of fluorescently tagged
proteins of interests can be performed. The simplest way is to
insert the gene coding for the protein of interest into an existing
vector that is uracil-excision-cloning compatible and contains a
gene encoding a fluorescent protein tag. A variety of such vectors
have been reported (14) and we have expanded this repertoire
with, e.g., C-terminal eGFP fusion vectors (Table 1, see Note 1).
When suitable vectors are not available, a three-fragment cloning
approach can be employed between the expression vector, the tag
and the gene-of-interest to generate a seamless translational fusion
156 Sakuragi, Nørholm, and Scheller

Table 1
Plasmid vectors and strains of Agrobacterium tumefaciens PGV3850 C58C1

Straina Plasmid Vector backbone GOIb Fluorescent tag Reference

N.A.c pCAMBI- pCAMBIA3300 (14)

A330035Su
N.A.c USER-eGFP pCAMBIA330035Su eGFP This work
YS28d pGFP-HDEL pTXS.P3C2 eGFP (18)
YS60 pSTtmd-YFP pCAMBIA330035Su STtmd Venus:1–238 Unpublished
YS63 pSTtmd-Yn pCAMBIA330035Su STtmd Venus:1–155 Unpublished
YS64 pSTtmd-Yc pCAMBIA330035Su STtmd Venus:156–238 Unpublished
YS66 pUXS2-YFP pCAMBIA330035Su UXS2 Venus:1–238 Unpublished
YS70 pUXS2-Yn pCAMBIA330035Su UXS2 Venus:1–155 Unpublished
YS71 pUXS2-Yc pCAMBIA330035Su UXS2 Venus:156–238 Unpublished
YS14 pARAD1-YFP pCAMBIA330035Su ARAD1 Venus:1–238 Unpublished
YS18 pARAD1-Yn pCAMBIA330035Su ARAD1 Venus:1–155 Unpublished
YS19 pARAD1-Yc pCAMBIA330035Su ARAD1 Venus:156–238 unpublished
P19 d
P19 pBin61 P19 (18)
a
Agrobacterial strain catalogue numbers in our laboratory
b
Gene of interest
c
Not available, only available as purified plasmids
d
Available but not distributable, the readers are advised to contact the original authors in respective references

protein construct in one step (15). For generating a series of

fusion constructs for a given protein, we have found that a
sequential two-step, two-fragment cloning approach is somewhat
more practical than those described above. A detailed procedure
is found in ref. (14). Briefly, the gene-of-interest is first inserted
into the uracil-excision cloning site of an expression vector while
regenerating the uracil-excision cloning site at the appropriate
end. In the second cloning step, genes coding for fluorescent
proteins (YFP, GFP, CFP, Yn, Yc, see below) are inserted into the
regenerated cloning site. This approach minimizes the number of
DNA sequence reactions to be carried out for verification of the
correct insert.
The translational fusion constructs generated above are intro-
duced into Agrobacterium tumefaciens for transient expression in
tobacco plants. Most of the fluorescent fusion proteins we have
tested so far resulted in detectable signals upon live cell imaging
as described below.
Live cell imaging directly on leaf tissues is carried out by con-
focal laser scanning microscopy (CLSM). CLSM allows optical
sectioning of complex biological specimen in a non-invasive manner,
 Visual Mapping of Cell Wall Biosynthesis 157

thus making it possible to determine subcellular localization and

dynamics of fluorescently tagged proteins in living cells. In com-
bination with an acousto-optical beam splitter that allows flexible
tuning of the detection wavelengths, simultaneous detection of
GFP- and YFP-tagged proteins is rendered straightforward (see
Note 2). In situ detection of protein complex formation is carried
out by using bimolecular fluorescent complementation (BiFC)
(16, 17). YFP is split in two non-fluorescent parts, and the
N-terminal (e.g. YFP amino acid residues 1–155, denoted as Yn)
and C-terminal (e.g. residues 156–238, denoted as Yc) halves are
fused to the proteins of interest and are brought together to
reconstitute the fluorescent properties of YFP, should the selected
proteins interact. We found that expression levels of the proteins
need to be carefully adjusted and appropriate negative controls
must be included in the analysis in order to minimize false posi-
tive unspecific interactions. We have also found that STtmd
(N-terminal 52 amino acid residues of rat sialyltransferase con-
taining the transmembrane domain) is a promiscuous interacter
that complements fluorescence in the BiFC assay with all Golgi
proteins we have tested so far. By exploiting this promiscuity, we
have established a BiFC-based system for determining the mem-
brane topology of Golgi localising proteins.

2. Materials

2.1. Growth 1. Cylindrical plastic pots (ca. 4 cm diameter, ca. 4 cm height)

of Nicotiana and a tray.
benthamiana 2. Potting soil available in gardening shops. (Optional)
Vermiculite can be added to potting soil in 1:2 ratio.
3. Greenhouse or growth facility with the following day and
night setting: daytime temperature 24°C and 12 h light; night
temperature 18°C and 12 h dark. These settings can be modi-
fied, for example, we have used plants grown at 28°C day
temperature successfully.
4. Bactimos (Garta, Odense, Denmark), a larvicide based on
Bacillus thuringiensis israelensis for minimising fungus gnat
infestation. In the US, a similar product named “Gnatrol
WDG” is manufactured by Valent (Walnut Creek, CA) and
available from various suppliers.
5. Plant nutrients available in gardening shops.

2.2. Plasmids, 1. For plasmids used in this study, see Table 1.

Bacterial Strains, 2. GenElute™ Plasmid Miniprep Kit (Sigma-Aldrich) or most of
Media, and Incubators commercial miniprep kits.
3. Chemical competent cells of E. coli DH10B or DH5a.
158 Sakuragi, Nørholm, and Scheller

4. Electrocompetent cells of Agrobacterium tumefaciens

PGV3850 C58C1.
5. Lurie-Bertani (LB) medium (1 L): 10 g Bacto Tryptone, 5 g
Yeast Extract, 5 g NaCl, and 15 g Bacto Agar for solid media.
Autoclaved.
6. 80% (v/v) Glycerol, autoclaved.
7. Double distilled water (ddH2O).

2.3. Uracil-Excision 1. Enzymes and buffers: PfuTurbo® Cx polymerase and 10×

Cloning, E. coli, and reaction buffer (Stratagene) (see Note 3); USER™ enzyme
A. tumefaciens mix, PacI, Nt. BbvCI, 10× NEB buffer 1, 10 mg/mL bovine
Transformation serum albumin (BSA) (New England Biolabs); TAE buffer
(40 mM Tris base, 40 mM acetic acid, 1 mM EDTA at
pH 8.0); a standard loading buffer.
2. 10 mM dNTP containing 2.5 mM each of dATP, dTTP,
dGTP, dCTP.
3. Synthetic oligonucleotides engineered with uracil-containing
tails at 5´ends and template DNAs (e.g. cDNA clones or
genomic DNA) containing genes of interest.
4. Antibiotics: in E. coli, kanamycin 50 mg/mL, ampicillin
100 mg/mL; in A. tumefaciens, kanamycin 50 mg/mL, ampi-
cillin 100 mg/mL, rifampicin 100 mg/mL.
5. A heat block or water bath set at 42°C and a shaker incubator
set at 37°C.
6. An electroporator, electroporation cuvettes (0.1 or 0.2 cm).

2.4. Transfection 1. Infiltration buffer: 10 mM 2-(N-morpholino)ethanesulfonic

of N. benthamiana acid (MES), pH. 5.6, 10 mM MgCl2, 100 mM acetosyringone.
2. 1-mL syringes.

2.5. Confocal Laser 1. Glass slides and double-sided adhesive tape.

Scanning Microscope 2. Leica TCS SP2 confocal scanning laser microscope equipped
with acousto-optical beam splitter, Argon laser (excitation
wavelengths at 458, 477, 488, and 514 nm) and He-Ne laser
(excitation wavelengths at 543 and 633 nm) and a 40× water
immersion objective (Leica Microsystems).

3. Methods

3.1. Growth 1. Plastic pots placed in a tray are filled with potting soil and are
of N. benthamiana showered with Bactimos until the soil is visibly soaked.
2. Two seeds of N. benthamiana are planted per pot by using a
toothpick.
 Visual Mapping of Cell Wall Biosynthesis 159

3. The tray is placed in the greenhouse. (Optional) The tray is

covered for a few days with a clear cover or a plastic wrap with
perforations for maintaining moisture. After germination, the
cover/wrap is removed.
4. The plants are watered two or three times per week. At least
once a week plant nutrient is included in the water. Plants
that are 4–5-week-old were used for transfection.

3.2. Construction 1. Preparation of uracil-excision compatible vectors: up to 2 mg

of Recombinant of column purified uracil-excision compatible vectors are
Proteins fused digested overnight at 37°C with 20 units of PacI in the
to YFP, GFP, Yn, or Yc presence of 1× NEB buffer 1 and 100 mg/mL BSA in a final
by Uracil-Excision volume of 20 mL. Five units each of PacI and the nicking
Cloning Strategy enzyme Nt. BbvCI are added to the mixture, and the mixture
is incubated for 5–6 h at 37°C. Forty microlitres of ddH2O is
added to the mixture, and the mixture is incubated at 80°C
for 20 min. The digested vectors can be stored at −20°C until
used (see Note 4).
2. Preparation of PCR products: reactions containing 1.25 units
PfuTurbo® Cx DNA polymerase, 1× reaction buffer, 0.5 mM
each of forward and reverse primers, 0.2 mM dNTPs, and
50 ng template DNAs are prepared. The following PCR con-
dition is routinely used: initial denaturing at 94°C for 2 min;
followed by 30 cycles of denaturing at 94°C for 0.5 min,
annealing at 50°C for 0.5 min, and extension at 72°C for
1 min/kb; final extension at 72°C for 10 min. Three
microlitres of the product is subjected to agarose-gel electro-
phoresis in the TAE buffer followed by visualization under
UV transilluminator light. The product that consists primar-
ily of the amplicon of the target and very little primer dimers
or satellite bands can be directly used without further purifi-
cation for annealing below. Otherwise, gel purification of the
target amplicon is highly recommended.
3. USER reaction and annealing: a typical reaction consists of
5 mL PCR product, 2 mL digested vector, and 1 mL USER™
enzyme mix. The reaction mix is incubated for 30 min at
37°C for uracil-excision, which is followed by 30 min incuba-
tion at 25°C for annealing the cohesive ends. The volume of
the PCR product and the digested vector can be varied (see
Note 5). The following negative controls should be included:
a reaction in which PCR product is replaced by ddH2O and a
reaction in which the digested vector is replaced with
ddH2O.
4. Chemical transformation: 40 mL E. coli competent cells are
thawed in ice and mixed with 4 mL of the uracil-excision
reaction. The mix is kept on ice for 30 min, thereafter trans-
ferred to 42°C for 2 min and then placed on ice for 1 min.
160 Sakuragi, Nørholm, and Scheller

One hundred microlitres of LB medium is added, and the mixture

is shaken at 37°C for 45 min. The entire mixture is plated over
an LB plate containing appropriate antibiotics. Colonies should
be visible after 12–14 h at 37°C (see Note 6).
5. Typically, three colonies per transformation are individually
transferred to 3 mL of LB medium containing appropriate
antibiotics and grown overnight at 37°C. The plasmids are
isolated from the culture and the insert verified by sequenc-
ing. The plasmids can be stored at −20°C. The plasmids are
introduced into Agrobacteria as described below.

3.3. Transformation 1. Ca. 200 ng of the isolated plasmid in no more than 1 mL

of A. tumefaciens water is added to 40 mL of A. tumefaciens electrocompetent
cells. The mixture is transferred to a 0.1-cm-cuvette, which is
subjected to electroporation with the following setup: resis-
tance, 400 W; voltage, 2.5 kV; capacitance, 25 mF.
2. Five hundred microlitres of LB medium is added to the
cuvette, and the mixture is incubated at 28°C for 1–2 h. Ten
and hundred microlitres of the mixture are spread on LB
plates containing rifampicin 100 mg/mL, ampicillin 50 mg/
mL, and appropriate antibiotics for selection of the plasmids.
The plates are incubated at 28°C for 2 days.
3. Colonies are transferred to LB medium containing the appro-
priate antibiotics and grown overnight at 28°C. Freezer stocks
are prepared by mixing with 80% glycerol to achieve a final
glycerol concentration of 20% (v/v).

3.4. Transfection 1. Three microlitres of overnight cultures of Agrobacterial

of N. benthamiana strains is centrifuged and the cell pellets are resuspended in
3 mL of the infiltration buffer.
2. For co-expression of recombinant proteins, the Agrobacterial
strains carrying the respective constructs are combined such
that the cell densities of individual strains adjusted to desired
values between 0.02 and 0.6 at optical density at 600 nm
(OD600nm). The combined resuspensions are allowed to stand
at room temperature for 1 h before infiltration.
3. By using a 1-mL syringe, the resuspension is infiltrated into
the leaf tissue. With the syringe tip gently pressed against the
underside (abaxial side) of the leaf, the resuspension is injected
into the leaf tissue by slowly pressing down the piston. An
area of at least 2 cm diameter should be infiltrated. The area
of successful infiltration is apparent by its water-soaked
appearance. For each construct, at least one leaf from
three independent plants is infiltrated. Up to four indepen-
dent infiltrations are made routinely in each leaf without
overlapping.
 Visual Mapping of Cell Wall Biosynthesis 161

4. The infiltrated plants are placed in greenhouse for subsequent

microscopical observation between 2 days post infiltration
(dpi) and 6 dpi.

3.5. Standard 1. For each expressed protein, the three independently infil-
Procedure for CLSM trated areas of the leaves are excised (ca. 1.5 cm × 1.5 cm)
with scissors and are fixed with upper side down on the micro-
scope glass slide by using double adhesive tape.
2. A drop of water is placed between each specimen and the 40×
water immersion objective.
3. The typical set up used for imaging fluorescent signals by
using Leica TCS SP2 CLSM is summarized is Table 2. Gain
of each photomultiplier (PMT) is typically between 700 and
850.
4. Images of at least three randomly selected positions per leaf
specimen are recorded, and three leaf specimens are analysed
for each expressed protein. The signal morphology and intensity
across different cells should be largely identical. If the signal
intensity is too weak or no signal was detected, the expression
may be optimized either by increasing the Agrobacterial
­inoculum in the infiltration mixture and/or by co-infiltrating
with Agrobacterial strain harbouring P19 for suppressing
post-transcriptional gene silencing mechanism (18). If the
signal is saturating and is observed in non-specific subcellular
localizations, the expression may be optimized by reducing
the cell density of the Agrobacterial inoculum or by making

Table 2
Parameters used for confocal scanning laser microscopy (Leica TCS SP2)

Excitation Emission Sequential Scanning

Fluorophores (nm) (nm) scanning speed (Hz) Line average
Single expression
PMT1 (GFP) 488 495–535 800, 1,000 8, 16
PMT1 (YFP) 514 535–560 800, 1,000 8, 16
Co-expression
PMT1 (Yn/Yc) 514 535–560 800, 1,000 8, 16
PMT1(GFP), PMT2 495, 514 495–510, Between 800, 1,000 8, 16
(YFP) 545–560 lines
Chloroplast detection
PMT3 488 or 514 650–707 800, 1,000 8, 16
PMT photomultiplier
162 Sakuragi, Nørholm, and Scheller

observations at later time point (e.g. 4 dpi and onwards).

For obtaining optimal expression conditions for a fusion
protein, we routinely start with OD600nm of 0.05, and optimize
the conditions by shifting the OD600nm values to 0.02, 0.05,
0.1, 0.2, 0.4, 0.6 with and without P19, which is routinely
included at the OD600nm of 0.1, and making observations at 2,
4, and 6pi.

3.6. Co-Expression 1. Optimization of expression conditions should be carried out

and Visualization of for individual proteins as described above. For ARAD1-YFP,
GFP- and YFP-Tagged STtmd-GFP, and GFP-HDEL, OD600nm of 0.05 without P19,
Proteins: ARAD1-YFP, at 4–6 dpi is optimal. STtmd is the N-terminal 52 amino acid
STtmd-GFP, and residues of the rat sialyltransferase that contains a transmem-
HDEL-GFP brane domain and is widely used as a Golgi marker (19).
GFP-HDEL is a widely used maker for ER, where the GFP is
retained in the ER by the tetrapepide ER retention signal
(HDEL) instead of being secreted to the apoplast (20).
2. Because of the overlap of the spectral properties of GFP and
YFP, it is critical that gain setup in each detection channel is
adjusted in order to minimize signal crosstalk. To this end,
negative controls that are transfected with individual con-
struct are prepared. The negative controls are scanned in the
sequential line scanning mode with two parameter sets:
(1) 488 nm excitation with 495–513 nm emission; (2)
514 nm excitation with 545–560 nm emission (see Table 2).
The gain of the photomultiplier detecting GFP (GFP chan-
nel) is first adjusted by using the YFP-expressing negative
control (i.e. ARAD1-YFP) such that either no or minimal sig-
nal crosstalk of YFP signal occurs in the GFP channel. The
gain of the YFP channel is then adjusted by using the GFP-
expressing negative control (i.e. STmd-GFP, GFP-HDEL)
such that either no or minimal signal crosstalk of GFP signal
occurs in the YFP channel.
3. With the gain setup chosen above, leaf specimen co-express-
ing both GFP- and YFP-tagged proteins are scanned (Fig. 1).
Examples of simultaneous imagines of GFP- and YFP-tagged
proteins are found in refs. (9) and (21).

3.7. In Situ Protein– 1. Optimization of parameters of transfection and expression

Protein Interaction should be carried out for individual proteins by using the full-
Detection by BiFC: length YFP fusion as described above. For ARAD1-YFP and
ARAD1 and UXS2 UXS2-YFP (see below), OD600nm of 0.05 without P19 at
4–6 dpi observation is optimal.
2. The detection of BiFC signals is carried out in the same way
as that of YFP (Table 2).
3. Negative controls. It has been reported that the split halves of
YFP can drive the interaction of proteins themselves and
 Visual Mapping of Cell Wall Biosynthesis 163

Fig. 1. Subcellular localization of ARAD1-YFP, STtmd-GFP, and GFP-HDEL imaged by confocal laser scanning microscopy.
(a–c) Co-expression of ARAD1-YFP and STtmd-GFP. (a) ARAD1-YFP detected in the YFP channel (magenta). (b) STtmd-
GFP detected in the GFP channel (green). (c) Merge of (a) and (b). The perfect overlap of the YFP and GFP signals is evi-
dent, which indicates that ARAD1-YFP localizes in Golgi. (d) Co-expression of ARAD1-YFP and GFP-HDEL. The images
acquired in the YFP and GFP channels are shown in green and red, respectively. Distinct patterns of the GFP and YFP
signals due to distinct subcelluar localizations of the two proteins are evident. Little signal cross-talk between the GFP
and YFP detection channels is observed.

thereby giving rise to false positive results (22). Therefore,

a set of negative controls that target the same cellular com-
partment and has an expression level similar to the test pro-
tein must be included in order to ascertain the relevance of
the observed interaction. Here we test the homodimerization
of ARAD1 and UXS2, an UDP-glucuronic acid decarboxy-
lase (23). It has been reported that these Golgi proteins form
homodimers ((23), Søgaard, Scheller, Sakuragi, unpublished).
They are not functionally related, thus they serve as a nega-
tive control. We transfected N. benthamiana with
(1) ARAD1-Yn, (2) ARAD1-Yc, (3) UXS2-Yn, (4) UXS2-Yc,
(5) ARAD1-Yn and UXS-Yc, (6) ARAD1-Yc and UXS-Yn,
(7) ARAD1-Yn and ARAD1-Yc, (8) UXS2-Yn and UXS2-Yc.
Experiments (1) to (4) show no signal, thus serving as techni-
cal controls (data not shown). Experiments (5) and (6) show
no signal, thus serving as negative interaction controls
(Fig. 2c). Only the experiments (7) and (8) give rise to fluo-
rescent signals with typical Golgi morphology (Fig. 2b, d).
164 Sakuragi, Nørholm, and Scheller

a b

YN YC YFP*

A B A B

c d

Fig. 2. BiFC analysis of ARAD1 and UXS2 homodimerization. (a) An illustration of the principle of BiFC, where the protein
A and B are fused to Yn and Yc and are brought together to complement fluorescence upon interaction between the
proteins. (b) Co-expression of ARAD1-Yn and ARAD1-Yc. (c) Co-expression of ARAD1-Yn and UXS2-Yc. (d) Co-expression
of UXS2-Yn and UXS-Yc. YFP fluorescence is shown in green. The red autofluorescence from chloroplasts is shown in
blue to make it distinguishable from the YFP fluorescence even for persons with impaired colour vision.

Combined results thus demonstrate that fluorescence

complementation between ARAD1-Yn and ARAD1-Yc is
mediated by homodimerization of ARAD1 protein in situ.
4. Positive controls. The following positive controls may be
included. The purpose of the positive controls is to ascertain
that the Yn and Yc fusion proteins are expressed and cor-
rectly folded to be compatible for BiFC. In split-ubiquitin
system, the wild-type Nub spontaneously assembles with
Cub due to a high affinity and is used for such control (24).
However, there has been no report of an equivalent positive
control for BiFC in Golgi. We observed that STtmd is a pro-
miscuous interacter and, fused to Yn and Yc, complements
fluorescence with all the Golgi-localising proteins tested so
far (Sakuragi, Scheller, unpublished). As a result, STtmd can
 Visual Mapping of Cell Wall Biosynthesis 165

serve as a positive control to show that Golgi-localized BiFC

constructs are functional. The basis for the observed promis-
cuity of STtmd-Yn/Yc is not clear. It is predicted that this
truncated protein consists of a short cytosolic N-terminus of
9 amino acid residues, a transmembrane domain of 17 amino
acid residues, and a Golgi-lumen-exposed C-terminus. The
C-terminal portion is predicted to be disordered by the three
web-based programmes, DisEMBL, DISOPRED, and
GlobPlots. It is known that intrinsically disordered proteins
are often involved in protein–protein interactions and that
disordered regions are thought to be structurally extended
and flexible, which enhances an initial, relatively non-specific,
association between the proteins (25). Therefore, it is plau-
sible that the non-specific interaction observed for STtmd-Yn
and Yc with the other BiFC proteins is mediated by the dis-
order in the C-terminus.

3.8. In Situ Protein Membrane topology of cell wall biosynthetic enzymes localized in
Membrane Topology the Golgi can be analysed by BiFC using STtmd-Yn/Yc. We have
Analysis (PROMTO) of verified by protease K susceptibility assay that the C-terminus of
Cell Wall Biosynthetic STtmd is indeed localized in the lumen while the N-terminus is in
Enzymes in Golgi the cytosol and that successful BiFC occurs between proteins that
expose the Yn and Yc domain to the same side of the membrane
(i.e. STtmd-Yn and STtmd-Yc, and Yn-STtmd and Yc-STtmd)
(Søgaard, Scheller, Sakuragi, unpublished). We have carried out
PROMTO analysis of ARAD1-Yn/Yc, UXS-Yn/Yc, GAUT7-Yn/
Yc, GUT1-Yn/Yc and observed fluorescence complementation
with STtmd-Yc/Yn. These results indicate that C-termini of these
CW biosynthetic proteins are oriented to the lumen side of the
Golgi membrane, which is in agreement with the bioinformatic
prediction of their type II membrane topology.

4. Notes

1. We constructed a vector for expression in plants with a

C-terminal eGFP-tag by amplifying the eGFP coding sequence
with the two oligonucleotides used for generating the plas-
mid pPS48uYFP (13). The resulting eGFP PCR fragment
was cloned by uracil-excision into pCAMBIA 2300u (13),
regenerating the uracil-excision compatible cloning site.
2. A series of fluorescent proteins with distinct spectral proper-
ties has been reported. In our laboratories, we have so far not
been able to detect fluorescent signals from tdTomato, mOr-
ange, and mCherry expressed transiently in tobacco.
3. PfuTurbo® Cx DNA polymerase is the only commercially
available proofreading DNA polymerase that can read through
166 Sakuragi, Nørholm, and Scheller

uracils in DNA templates (Stratagene). Recently, a new highly

efficient DNA polymerase was developed specifically for use
in uracil-excision cloning (26).
4. The complete digestion of the vector is crucial for efficient
cloning. It is recommended that a small aliquot (ca. 1 mL) of
the digest is analysed by gel electrophoresis before heat inac-
tivation of the enzymes. The digested vectors can be purified
by PCR purification kit (Qiagen), although we have not
noticed significant improvements in cloning efficiency with
purification as compared to heat inactivation.
5. Successful uracil-excision cloning appears to be dependent, to
some extent, on the sufficient amount of vector. If a spectro-
photometer for the assessment of DNA concentration is not
available, DNA concentration can be roughly estimated on
TAE agarose gel electrophoresis. In this case, 1 mL of digested
DNA should result in a band with an appropriate size with
intensity readily visible under UV light.
6. The efficiency of chemically competent cells need not be high,
and home-made chemically competent cells with an efficiency
of 107 CFU/mg DNA (supercoiled pUC19) should result in
sufficient numbers of colonies ranging up to 100–200 trans-
formants. Typically, we obtain between 10 and 200 transfor-
mants per reaction. For negative controls, the number of
transformants, or background, is kept less than 10. In case
this number is higher than 10, we carry out another round of
digestion in order to reduce background.

Acknowledgements

This work was supported by the U.S. Department of Energy,

Office of Science, Office of Biological and Environmental
Research, through contract DE-AC02-05CH11231 between
Lawrence Berkeley National Laboratory and the U.S. Department
of Energy; by the Danish Villum Kann Rasmussen Foundation
through the VKR centre Pro-active Plants; and by the Danish
Agency for Science, Technology and Innovation.

References

1. Lerouxel, O., Cavalier D. M., Liepman, A. 3. Mohnen, D. (2008) Pectin structure and bio-
H., and Keegstra K. (2006) Biosynthesis of synthesis. Curr Opin Plant Biol 11, 266–277
plant cell wall polysaccharides – a complex 4. Scheller, H. V. and Ulvskov, P. (2010)
process. Curr Op Plant Biol 9, 621–630 Hemicelluloses. Ann Rev Plant Biol, 61,
2. Scheller, H. V., Jensen, J. K., Sorensen, S. O., 263–289
Harholt, J., and Geshi, N. (2007) Biosynthesis 5. Doblin, M. S., Kurek, I., Jacob-Wilk, D., and
of pectin. Physiologia Plantarum 129, 283–295 Delmer, D. P. (2002) Cellulose biosynthesis
 Visual Mapping of Cell Wall Biosynthesis 167

in plants: from genes to rosettes. Plant Cell taneous fusion and cloning of multiple PCR
Physiol 43, 1407–1420 products. Nucleic Acids Res 35, e55
6. Hong, Z. L., Zhang, Z. M., Olson, J. M., 16. Hu, C. D., Chinenov, Y., and Kerppola, T. K.
Verma, D. P. S. (2001) A novel UDP-glucose (2002) Visualization of interactions among
transferase is part of the callose synthase com- bZIP and Rel family proteins in living cells
plex and interacts with phragmoplastin at the using biomoleuclar fluorescence complemen-
forming cell plate. Plant Cell 13, 769–779 tation. Mol Cell 9, 789–798
7. Young, W. W. (2004) Organization of Golgi 17. Kerppola, T. K. (2006) Visualization of molec-
glycosyltransferases in membranes: complex- ular interactions by fluorescence complemen-
ity via complexes. J Membr Biol 198, 1–13 tation. Nat Rev Mol Cell Biol 7, 449–456
8. Harholt, J., Jensen, J. K., Sørensen, S. O., 18. Voinnet, O., Rivas, S., Mestre, P., and
Orfila, C., Pauly, M., and Scheller, H. V. Baulcombe, D. (2003). An enhanced tran-
(2006) ARABINAN DEFICIENT 1 is a novel sient expression system in plans based on sup-
glycosyltransferase involved in biosynthesis of pression of gene silencing by the p19 protein
pectic arabinan in Arabidopsis. Plant Physiol of tomato bushy stunt virus. Plant J 33,
140, 49–58 949–956
9. Jensen, J. K., Sørensen, S. O., Harholt, J., 19. Boevink, P., Oparka, K., Santa Cruz, S.,
Geshi, N., Sakuragi, Y., Møller, I., Zandleven, Martin, B., Betteridge, A., and Hawes, C.
J., Bernal, A. J., Jensen, N. B., Sørensen, C., (1998) Stacks on tracks: the plant Golgi appa-
Pauly, M., Beldman, G., Willats, W. G. T., and ratus traffics on an actin/ER network. Plant J
Scheller, H. V. (2008) Identification of a 15, 441–447
xylogalacturonan xylosyltransferase involved 20. Boevink, P., Santa Cruz, S., Hawes, C., Harris,
in pectin biosynthesis in Arabidopsis. Plant N., and Oparka, K. J. (1996) Virus-mediated
Cell 20, 1289–1302 delivery of the green fluorescent protein to
10. Nguema-Ona, E., Andeme-Onzighi, C., the endoplasmic reticulum of plant cells. Plant
Aboughe-Angone, S., Bardor, M., Ishii, T., J 10, 935–941
Lerouge, P., and Driouich, A. (2006) The 21. Voinonen, J. P., Sakuragi, Y., Stael, S.,
reb1-1 mutation of Arabidopsis. Effect on the Tikkanen, M., Allahverdiyeva, Y., Paakkarinen,
structure and localization of galactose- V., Aro, E., Suorsa, M., Scheller, H. V., Vener,
containing cell wall polysaccharides. Plant A. V., and Aro, E. M. (2008) Light regulation
Physiol 140, 1406–1417 of CaS, a novel phosphoprotein in the thyla-
11. Langeveld, S. M. J., Vennik, M., Kottenhagen, koid membrane of Arabidopsis thaliana. FEBS
M., van Wijk, R., Buijk, A., Kijne, J. W., and J 275, 1767–1777
de Pater, S. (2002) Glucosylation activity and 22. Lalonde, S., Ehrhardt, D. W., Loque, D.,
complex formation of two classes of reversibly Chen, J., Rhee, and S. Y., Frommer, W. B.
glycosylated polypeptides. Plant Physiol 129, (2008) Molecular an cellular approaches for
278–289 the detection of protein-protein interactions:
12. de Pino, V., Borán, M., Norambuena, L., latest techniques and current limitations.
González, M., Reyes, F., Orellana, A., and Plant J 53, 610–635
Moreno, S. (2007) Complex formation regu- 23. Gu, X., and Bar-Peled, M. (2004) The bio-
lates the glycosylation of the reversibly glyco- synthesis of UDP-galacturonic acid in plants.
sylated polypeptide. Planta 226, 335–345 Functional cloning and characterization of
13. Sterling, J. D., Atmodjo, M. A., Inwood, S. Arabidopsis UDP-d-glucuronic acid 4-epimerase.
E., Kumar Kolli, V. S., Quigley, H. F., Hahn, Plant Physiol 136, 4256–4264
M. G., and Mohnen, D. (2006) Functional 24. Stagliar, I., Korostensky, C., Johnsson, N.,
identification of an Arabidopsis pectin biosyn- and te Heesen, S. (1998) A genetic system
thetic homogalacturonan galacturonosyltrans- based on split-ubiquitin for the analysis of
ferase. Proc Natl Acad Sci U S A 103, interactions between membrane proteins
5236–5241 in vivo. Proc Natl Acad Sci U S A 95,
14. Nour-Eldin, H. H., Hansen, B. G., Nørholm, 5187–5192
M. H., Jensen, J. K., and Halkier, B. A. (2006) 25. Tompa, P. (2005) The interplay between
Advancing uracil-excision based cloning structure and function in intrinsically unstruc-
towards an ideal technique for cloning PCR tured proteins. FEBS Lett 579, 3346–3356
fragments. Nucleic Acids Res 34, e122 26. Norholm, M. H. H. (2010) A mutant Pfu
15. Geu-Flores, F., Nour-Eldin, H. H., Nielsen, DNA polymerase designed for advanced uracil-
M. T., and Halkier, B. A. (2007) USER excision DNA engineering. BMC Biotechnol
fusion: a rapid and efficient method for simul- 10, 21
 Chapter 12

Atomic Force Microscopy of Plant Cell Walls

Andrew R. Kirby

Abstract
Atomic force microscopy (AFM) can be used to obtain high-resolution images on a wide variety of materials.
Unfortunately, plant cell wall material is typically too rough to be imaged as native tissue by AFM.
Small tissue fragments can be produced through careful ball milling. These fragments can subsequently
be imaged at high resolution in near native conditions showing the overall architecture and the arrange-
ment of the individual cellulose fibrils. An overview of items that can cause practical difficulties is given,
as is a description of common image artifacts.

Key words: Atomic force microscopy, Tissue, Artifact

1. Introduction

Since its inception in the 1980s, atomic force microscopy (AFM)

has matured into an immensely powerful technique for studying
surfaces at high resolution. It is particularly suited to tackle prob-
lems in biology since the sample preparation is usually kept to an
absolute minimum, without the need for fixation, dehydration or
metal coating. The technique employs a sharp microfabricated
stylus that “feels” the sample surface in a way that is analogous to
a blind person reading Braille. In order to minimize the forces
applied to the sample, and therefore reduce any potential dam-
age, the stylus (more usually called the tip) is mounted on the end
of an extremely flexible cantilever (Fig. 1).
The AFM tip can “feel” the sample so delicately that it is
­routinely possible to image individual molecules on flat surfaces
without displacing or shearing them. The tiny vertical movements
of the tip are detected using an optical lever; a laser beam is
reflected off the end of the cantilever and onto a photodiode

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_12, © Springer Science+Business Media, LLC 2011

169
170 Kirby

Fig. 1. A typical AFM cantilever with a pyramidal stylus on the end. This particular cantile-
ver has a “V” shape, but rectangular beam varieties are also available. Typical cantilever
lengths span 40–200 mm. SEM micrograph courtesy of Paul Gunning.

Fig. 2. On rough materials, there is a real danger that the underside of the cantilever can
foul on the sample, particularly when scanning over larger areas. This can lead to a
variety of artifacts in the image.

detector. Unfortunately, when the sample is relatively rough, i.e.,

its vertical topography varies in the order of a few microns (which
is in the same size range as the dimensions of the tip), then the tip
will track the sample surface poorly and the legs of the cantilever
may foul on the sample (Fig. 2).
This results in an inaccurate image of the sample being
recorded, with many of the features being artifacts. In this case,
more consideration of the sample preparation procedure and the
adjustment of the instrument parameters are essential. For a more
rigorous introduction to the technique see Morris et al. (1).

2. Materials

2.1. Microscopy 1. The principal requirement is an AFM with a reasonably large

Instrumentation range scanner. This is the piezoelectric device that raster scans
the tip relative to the sample. In order to cope with the large
variations in sample topography of the plant cell wall material,
 Atomic Force Microscopy of Plant Cell Walls 171

a scanner with ~10 mm in vertical displacement is desirable.

An AFM mounted on an optical microscope provides a quick
and convenient way of finding regions of interest before scan-
ning at high resolution with the AFM. This is important
because even large AFM scanners typically have only 100 mm
range in X and Y. Suitable Instruments that the author has
experience with include MFP-3D (Asylum Research, Santa
Barbara, CA), Bioscope Catalyst (Veeco Instruments, Santa
Barbara, CA), and the Nanowizard III (JPK Instruments AG,
Berlin, Germany), although no doubt others are available.
2. AFM cantilevers. NP general purpose 100–200 mm in length
(Veeco Instruments, Santa Barbara, CA).
3. The sample itself needs to be supported by a rigid substrate;
this can be a cleaved mica (Agar Scientific Ltd., Stansted, UK)
or simply a conventional clean glass microscope slide.
Chemicals:
4. 1.5% SDS and Na2S2O5 containing 5 mL octanol/L.
5. 0.5% SDS and Na2S2O5 containing 2.5 mL octanol/L.
6. I2KI solution: 3 g/L iodine crystals, 15 g/L potassium iodide
dissolved in water.

2.2. Sample Preparation 1. Ball mill for reducing the size of the plant tissue fragments
(Capco Test Equipments Ltd., Wickham Market, Suffolk,
UK).
2. Waring Blender (Christison Particle Technologies Ltd.,
Gateshead, UK).
3. Ystral Homogeniser (Ystral GmbH, Dottingen, Germany).
4. Nylon mesh (BioDesign Inc., NY).

3. Methods

3.1. Preparation Although this method is particular to potatoes, with minor modi-
of Cell Wall Material fications it can be successfully used to extract cell wall material
from Potatoes from many plants.
1. Slice potatoes transversely (approximately 5 mm thick).
2. Cut out parts within the vascular ring from the slices and
from the central third of the potato.
3. For interim storage freeze the tissue in liquid nitrogen.
4. Using a Waring blender blend 500 g batches of the potato
tissue for 3 min in 1 L of 1.5% SDS and Na2S2O5 solution + 5 mL
octanol.
172 Kirby

5. Filter out any unblended tissue with a 2 mm mesh.

6. Using an Ystral Homogeniser break down the filtrate at
16,000 rpm for 1 min.
7. Filter on a 200-mm nylon mesh.
8. Wash the retained material with water and then resuspend in
0.5% SDS and Na2S2O5 solution + 2.5 mL octanol.
9. Add the solution to a 2.5-L ball mill pot until only the very
tops of the balls are showing. A few drops of octanol can
reduce excessive foaming. Ball mill for 1–2 h at 60 rpm.
10. Pour out the liquor and wash each ball with water.
11. Remove the starch by filtering on a 100 mm nylon mesh and
wash the material with 10 L of water.
12. Resuspend in water, homogenize again at 1,600 rpm for
1 min.
13. Filter and wash with 10 L of water on 100 mm nylon mesh.
14. Check for the absence of starch by staining a small amount
of tissue on a microscope slide with I2KI solution. Any remain-
ing granules should appear blue/black in color. If there are
still a significant amount of granules remaining, repeat steps
12 and 13.
15. Resuspend the cell wall material in water then freeze until
needed.

3.2. Microscopy 1. Prepare the AFM for use by inserting a new undamaged
tip, then align the laser and photodiode in the normal
manner.
2. The instrument will be operated in contact mode, and should
be set up as for imaging in air.
3. If mica is selected as the preferred substrate, it will first need
to be cleaved by inserting a sharp object, such as a pointed set
of forceps, at one edge and peeling the layers apart to expose
a clean flat surface. Alternatively, a glass slide that has been
cleaned with water and ethanol can be used.
4. Apply a drop (~100 mL) of a dilute solution of tissue
­fragments onto the chosen substrate with a pipette. In
order to prevent the pipette tip from blocking, it may be
necessary to enlarge the aperture by cutting the end back
obliquely.
5. Using a low power optical microscope as a visual aid, distrib-
ute the tissue fragments across the substrate using a toothpick
or similar. This is to prevent the sheets of cell wall material
aggregating and stacking up, which would form a very rough
surface.
 Atomic Force Microscopy of Plant Cell Walls 173

Fig. 3. An instrument that is combined with an optical microscope is a great aid to locat-
ing appropriately sized tissue fragments and positioning the tip. On the left is the canti-
lever as seen from above. This cantilever is 200 mm in length. Notice that the actual tip
is not visible at this magnification.

6. Blot away any excess water using filter paper, but leave the
sample still moist (see Notes 1 and 2).
7. Select an area of interest using the optical microscope.
This will be a fragment of tissue ~100 mm in diameter
(Fig. 3), and set the AFM tip to approach this region (see
Note 3).
8. When the tip is in contact with the surface of the sample,
imaging can begin. This should be at a slow rate, for
example 0.5 Hz or less, i.e., one raster scan line every 2 s
or less. The scan size should initially be modest, for exam-
ple £3 mm. Check that the sample is not being pushed
around by the tip (see Notes 4–6). Typical detailed images
of a variety of plant tissues are shown (Fig. 4). Additionally,
the reader is directed to other examples in the literature
(2–6).
9. In addition to recording the topography image, it is a good
idea to also record the error signal image as this can provide
more fine details on rougher materials. It is worth stressing
that the topography image appears to degrade as the error
signal image improves and vice versa.
10. During the experiment, try increasing the instrument gain to
improve the tracking of the tip over tall features on the sample
(see Note 7).
11. If it is still difficult to obtain undistorted images, try experi-
menting with longer cantilevers (see Note 8).
174 Kirby

Fig. 4. Three examples of what you should see when things go correctly! Top : AFM error signal image of apple cell wall,
size 1 × 1 mm. Middle: AFM error signal image of potato cell wall. Size 2 × 2 mm. Below : AFM error signal image of nettle
cell wall. Size 2.5 × 2.5 mm. Nettle image courtesy of Patrick Gunning.

4. Notes

1. Over time, the sample will dry out and become more difficult to
image. The principle reason for keeping the cell wall tissue moist
is that the cellulose microfibrils stick upward like little “hairs”
when completely dry. The AFM cannot properly track this kind
of surface as the microfibrils are simply pushed aside by the tip
during the scanning process. The surface tension of the thin liq-
uid layer helps pull the microfibrils down so that they lay flat and
immobile on the substrate (5). Additional water may be required
after an hour or so, depending on the ambient conditions.
2. Do not forget to blot away any excess water (see Subheading 3.2,
part 6), otherwise the liquid layer will spill over onto the back of
 Atomic Force Microscopy of Plant Cell Walls 175

the cantilever and it will be impossible to find the laser spot due
to refraction of the laser beam. Worse still, the tissue fragments
may detach from the substrate and float above it.
3. If it is not possible to find any tissue fragments that are flat
enough to image, then the ball milling time could be empiri-
cally increased to reduce the fragment size. However, if the
tissue is ball milled for too long, then there is a risk of exces-
sive disruption and resulting images that are not truly repre-
sentative of the bulk sample.
4. Start by scanning small areas (see Subheading 3.2, part 8).
Larger scan areas contain greater extremes in topography and
are therefore more difficult, or impossible, for the AFM to
track properly, unless the instrument parameters are perfect.
5. You will most likely find that there is a practical limit to the
maximum scan size you can perform before the images appear
distorted or the tip mistracks (known as a “tip jump”) due to
excessive sample roughness (Figs. 5 and 6). Attempting to
perform even larger scan sizes will only result in tip damage
or contamination.

Fig. 5. Image artifacts can and will occur, particularly with rough samples. Top : an AFM
image where the sample has interacted with the sidewall of the tip causing the image
to appear “blurred.” Note the tell-tale “cliff-edge” feature in the center of the image.
If this type of artifact appears, it would be sensible to try reducing the scan speed.
Below : schematic illustrating this type of tip-sample interaction.
176 Kirby

Fig. 6. Streaks or tip jumps in the image. This is caused by the tip momentarily sticking
to the sample surface or sudden mistracking due to movement of the sample.

Fig. 7. The same features appear multiple times in the image (top). This is most com-
monly called a “multiple tip” artifact. This arises due to tip damage or contamination and
is most frequently observed at small scan sizes. Below: schematic illustrating this type
of tip-sample interaction.

6. Additionally, there is another artifact that can occur because

of tip wear, damage, or contamination. The apex of the tip
is no longer just a single sharp point but multiple points.
This manifests itself as repeating structures or side-by-side
features in the image (Fig. 7). When this occurs the only
option is to replace the tip.
 Atomic Force Microscopy of Plant Cell Walls 177

Fig. 8. Schematic diagram showing that shorter cantilevers move through large angles
(top). This can pose a problem with rough samples because the laser beam could miss
the photodiode altogether. In this case, longer cantilevers are preferred as the angular
sensitivity is reduced (bottom).

7. The speed of response of the AFM to large features can

be improved by increasing the instrument gain (see
Subheading 3.2, part 10). However, if the gain value is set
too high, it can lead to the appearance of vertical lines in the
image due to uncontrolled tip oscillations.
8. The cantilever length plays an important role in the vertical
sensitivity of the instrument. Shorter cantilevers can detect
smaller variations in the height of the sample because they
deflect the laser beam through wider angles than their longer
counterparts (Fig. 8). However, for rough samples, longer
cantilevers are preferred since it is vital that the reflected laser
beam actually remains on the face of the photodiode without
spilling over the side.

Acknowledgements

The research described in this article was funded through the

BBSRC core support grant to the Institute of Food Research.
178 Kirby

References

1. Morris, V. J., Kirby, A. R., and Gunning, A. P. 4. Kirby, A. R., Ng, A., Waldron, K. W., and
(2010) Atomic Force Microscopy for Morris, V. J. (2006) AFM investigations of
Biologists. Imperial College Press, London. cellulose fibers in bintje potato (Solanum
2. Marga, F., Grandbois, M., Cosgrove, D. J., tuberosum L.) cell wall fragments. Food
and Baskin, T. I. (2005) Cell wall extension Biophysics 1, 163–167.
results in the coordinate separation of parallel 5. Kirby, A. R., Gunning, A. P., Waldron, K. W.,
microfibrils: evidence from scanning electron Morris, V. J., and Ng, A. (1996) Visualization
microscopy and atomic force microscopy. of plant cell walls by atomic force microscopy.
Plant Journal 43, 181–190. Biophysical Journal 70, 1138–1143.
3. Ding, S. -Y., and Himmel, M. E. (2006) The 6. Thimm, J. C., Burritt, D. J., Ducker, W. A.,
maize primary cell wall microfibril: a new and Melton, L. D. (2000) Celery (Apium gra-
model derived from direct visualization. veolens L.) parenchyma cell walls examined by
Journal of Agricultural and Food Chemistry atomic force microscopy: effect of dehydration
54, 597–606. on cellulose microfibrils. Planta 212, 25–32.
 Chapter 13

Using Solid-State 13C NMR Spectroscopy to Study

the Molecular Organisation of Primary Plant Cell Walls
Tracey J. Bootten, Philip J. Harris, Laurence D. Melton,
and Roger H. Newman

Abstract
Studies of the mobilities of polysaccharides or parts of polysaccharides in a cell-wall preparation may give
clues about the molecular interactions among the polysaccharides in the cell wall and the relative loca-
tions of polysaccharides within the cell wall. A number of solid-state 13C NMR techniques have been
developed that can be used to investigate different types of polysaccharide mobilities: rigid, semi-rigid,
mobile, and highly mobile. In this chapter, techniques are described for obtaining spectra from primary
cell-wall preparations using CP/MAS, proton-rotating frame, proton spin-spin, spin-echo relaxation
spectra, and single-pulse excitation. We also describe how proton spin relaxation editing can be used to
obtain subspectra for cell-wall polysaccharides of different mobilities.

Key words: Primary cell walls, Polysaccharide mobility, Solid-state 13C NMR, Proton-spin relaxation
editing, Single-pulse excitation NMR, TEM, X-ray diffraction

1. Introduction

Primary walls surround growing plant cells and are composed

of rigid cellulose microfibrils embedded in a gel-like matrix
of ­non-cellulosic polysaccharides with a range of different structures
(1, 2). The major non-cellulosic, matrix polysaccharides of the
­primary cell walls of eudicotyledons and non-commelinid mono-
cotyledons are xyloglucans (XG), and pectic polysaccharides (2).
The pectic polysaccharides consist mainly of homogalacturonan
(HG) and rhamnogalacturonan I (RG-I), with smaller propor-
tions of the substituted galacturonans rhamnogalacturonan II
(RG-II) and xylogalacturonan (XGA). Arabinans, galactans, and
arabinogalactans are attached as side chains to RG-I (3). The major

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_13, © Springer Science+Business Media, LLC 2011

179
180 Bootten et al.

­ olysaccharides of the primary cell walls of many commelinid

p
­monocotyledons are glucuronoarabinoxylans (GAXs) with smaller
proportions of pectic polysaccharides and XGs (2, 4). Variable pro-
portions of (1→3),(1→4)-b-glucans also occur in the primary walls
of grasses and cereals (Poaceae) and related families (5).
Solid-state 13C nuclear magnetic resonance (NMR) spectros-
copy is one of several techniques used to study the molecular
organisation of such walls. In contrast to TEM, this technique
can be used on cell walls that have never been dried, thereby
minimising possible chemical and physical modifications.
Moreover, in contrast to X-ray diffraction, this technique does
not require multiple planes of ordered molecules. Solid-state
13
C NMR is sensitive to ordering over relatively short dimensions,
and measurements of chemical shifts can indicate differences
in the conformations of backbone chains or side-chains (6).
Measurements of spin relaxation time constants can indicate
­differences in the short-range environment, specifically the nature
and dynamics of neighbouring molecules. In studies of the cel-
lulose in walls, solid-state 13C NMR can be used to distinguish
between molecules on the surfaces of crystallites and those in the
interior, since their molecular conformations differ (7–9). This
information can be used to estimate the lateral dimensions of cel-
lulose crystallites. In this application, solid-state 13C NMR is par-
ticularly informative when the crystallites are of such a small
cross-section that there are as many chains exposed on surfaces as
contained in the interior, as occurs in primary-wall cellulose (6, 10).
TEM and X-ray diffraction are more informative if the cellulose
crystallites are of a relatively large cross-section (8).
Solid-state 13C NMR spectroscopy can also be used to distin-
guish cell-wall polymers that have different mobilities because of
their locations and interactions with other molecules within the
cell wall (11). A summary of how this type of mobility-resolved
NMR may help to identify polysaccharide domains and interac-
tions within the cell wall is shown in Fig. 1. For example, the
(1→4)-b-d-glucan backbones of XG molecules, or parts of mol-
ecules, adsorbed on to the surface of cellulose microfibrils in
muro are predicted to adopt a rigid, flattened conformation,
rather than the twisted backbone conformation of free XG (12).
These two different conformations can be detected by solid-state
13
C NMR spectroscopy (13, 14) (Fig. 1). Therefore, not only is
solid-state 13C NMR spectroscopy useful for determining the
mobilities of different components in plant cell walls, but it also
effectively makes in situ investigations (15).
Two main types of mobility-resolved solid-state 13C NMR
experiments are commonly used to investigate primary cell walls,
cross-polarisation (CP) NMR and single-pulse excitation (SPE)
NMR. Both are generally used in combination with MAS, hence
they are referred to as CP/MAS NMR and SPE/MAS NMR (16).
 Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 181

RIGID DOMAIN MOBILE DOMAIN VERY

MOBILE
Polysaccharides that respond to CP/MAS NMR DOMAIN
Polysaccharides that separate into Polysaccharides that separate into Polysaccharides
PSRE subspectrum A with long PSRE subspectrum A with short that do not
T1r(H) and short T2(H) T1r(H) and long T2(H) respond to
CP/MAS NMR
Inherently rigid polysaccharides Mobile polysaccharides
e.g. cellulose Very mobile
Also, more mobile polysaccharide Semi-rigid polysaccharide structures polysaccharides
structures affected by proton-spin affected by proton-spin diffusion from e.g. arabinans,
diffusion from rigid polysaccharides eg. the mobile polysaccharides galactans,
Glc and Xyl of XG adsorbed on to arabinogalactans
cellulose

Polysaccharides that respond to SPE/MAS (recovery delay 1 ms) NMR

Polysaccharides that respond to SE-SPE/MAS NMR

Polysaccharides Polysaccharides that separate into subspectrum B Very long T2(C)

that separate into of SE-PSRE
subspectrum A of
§ Short T2(C) IntermediateT2(C) Short T2(C) Very mobile
SE-PSRE
Mobile Semi-rigid Mobile polysaccharides
Long T2(C) polysaccharide polysaccharide polysaccharides e.g. arabinans,
Rigid structures structures galactans and
e.g. pectic
polysaccharides e.g. XG-Xyl of e.g. XG-Glc in XG arabinogalactans
homogalacturonan
e.g. cellulose XG adsorbed to cross-links between
cellulose cellulose microfibrils

Fig. 1. Solid-state 13C NMR, proton- and 13C- spin relaxation techniques described in this chapter and used to identify
polysaccharides of different mobilities in cell-wall preparations. §These mobile polysaccharides separate into the same
T1r(H) and T2(H) subspectra as cellulose due to proton spin diffusion from adjacent cellulose but separate into the mobile
T2(C) subspectra as these structures are inherently mobile.

CP/MAS NMR suppresses signals from relatively mobile molecules,

and SPE/MAS NMR suppresses signals from relatively rigid
­molecules (17).
Cross-polarisation NMR combines both the magnetic spins
of protons and 13C nuclei. By transferring the magnetisation from
protons to 13C, the range of dispersion of the chemical shifts is
increased compared to the range of chemical shift values for
­proton NMR, thus increasing the resolution of the signals from
solid-state samples (18). CP/MAS NMR also allows the indirect
measurement of proton NMR relaxation by variations in the
strengths of selected 13C signals (19). The proton relaxation
­process is complicated by spin diffusion where neighbouring
nuclei exchange spin information (20). Proton spin diffusion may
occur over ­distances of nanometers and over time scales of milli-
seconds (21). Therefore, proton spin relaxation time constants
182 Bootten et al.

1 td tp t sl tc ta
H

13
tc
C

FID output

Fig. 2. Schematic representation of a proton rotating-frame spin relaxation pulse

sequence. tp: preparation pulse; t sl: spin-locking pulse; t c: cross-polarisation contact
time; ta: data acquisition time; t d: recovery delay. In each case, t = time. Adapted from
Newman and Hemmingson (28), de Gruyter.

are mean values for all protons within a finite volume, irrespective
of the 13C-signal used to monitor the proton relaxation.
Two main approaches using CP/MAS NMR are used to
investigate polysaccharide mobility in primary cell walls. Both
approaches exploit the fact that the relaxation of spins of mag-
netic nuclei for polysaccharides in cell walls can be sensitive to the
molecular conformations of those polysaccharides. First, proton-
spin relaxation editing (PSRE) can be used to separate NMR sub-
spectra containing signals from polysaccharides in rigid and
mobile domains of the cell wall (7). To do this, a proton spin
relaxation event is introduced prior to the CP contact time.
A typical PSRE pulse sequence is illustrated in Fig. 2. Linear com-
binations, generated from spectra obtained under the normal
CP/MAS and the PSRE conditions, can then be used to edit the
CP/MAS spectra (17). Second, a number of PSRE experiments
may be carried out each with different recovery delays (22). From
these spectra, full relaxation curves for component signals may be
constructed. The shape of the relaxation curve and the time con-
stants associated with the proton relaxations can be used to
­distinguish between polysaccharides in rigid and mobile domains
in the cell wall, remembering that the proton spin relaxation time
constants of a particular polysaccharide will be similar to those of
the polysaccharides in its immediate vicinity.
Segmental motion within a polysaccharide can be investigated
by measuring relaxation of the 13C-nuclei by using a spin-echo
(SE) NMR pulse sequence. The rarity of the 13C-isotope (1.1% of
carbon) means the diffusion of 13C spin information is relatively
slow (23). Therefore, unlike proton spin relaxation, 13C relax-
ation is sensitive to segmental motion of the polysaccharide at the
site of the 13C nucleus and is much less sensitive to the motion of
more distant segments of the same polysaccharide chain or other
polysaccharide chains.
 Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 183

There are three main spin-relaxation processes that have been

used to investigate primary cell walls: spin-lattice relaxation with
time constants T1(H) and T1(C), rotating-frame relaxation with time
constants T1r(H) and T1r(C), and spin-spin relaxation with
time constants T2(H) and T2(C) (6, 18). Newman et al. (6) found
that the most useful relaxation processes for investigating proton
relaxations of polysaccharides in primary cell walls were the rotating-
frame and spin-spin experiments.
SPE/MAS NMR is an excellent complementary technique to
CP/MAS NMR (22) for investigating mobile polysaccharides in
cell-wall preparations. The CP/MAS NMR spectra of the cell-
wall preparations are generally dominated by signals assigned to
cellulose, and the broader underlying signals are assigned to non-
cellulosic polysaccharides, along with, in primary cell-wall prepa-
rations containing predominantly pectic polysaccharides, signals
at 173, 54, and 22 ppm assigned to carboxylic acid, methoxyl,
and acetyl carbons, respectively. As cellulose generally accounts
for less than half of the primary cell wall, a portion of the non-
cellulosic material must be too mobile to respond to the cross-
polarisation pulse sequence. Although SPE/MAS can potentially
detect all components in the cell-wall preparations (16), the
highly mobile cell-wall components, which are less responsive to
CP/MAS NMR (24, 25), are particularly responsive to SPE/
MAS. By using a short recovery delay in the SPE/MAS pulse
sequence (1 s), signals that have long T1(C) relaxations, such as
the rigid cellulose molecules, are suppressed (16, 25) (Fig. 1).
Studies of polysaccharide mobility in an isolated cell-wall
preparation should use not only the relaxation properties of the
proton or 13C of that moiety, as indicated in Table 1, but also
take into account alterations in the chemical shift for that poly-
saccharide or monosaccharide or part of the polysaccharide.
Generally, polysaccharides investigated using solution NMR will
have lower chemical shifts than in solid-state NMR (Table 1).
These different chemical shifts may indicate differences in
­molecular conformations, e.g. extended chains rather than coiled
chains (14). Chemical shifts can therefore provide insights into
the reasons for reduced mobility within the cell wall, particularly
when combined with the relaxation studies described above.

2. Materials

1. Plant material of interest.

2. Cell wall isolation buffer: 20 mM HEPES-KOH buffer,
pH 6.7 containing 10 mM dithiothreitol (DTT) (see Note 1).
3. DTT-free buffer: 20 mM HEPES-KOH, pH 6.7.
184 Bootten et al.

Table 1
Cited solid-state 13C NMR assignments for the main primary
cell-wall polysaccharides of eudicotyledons and non-
­commelinid monocotyledons

Assignment Chemical shift ppm Reference

C-1 Glc Cellulose Ia (interior) 105.6 (35)

Cellulose Ib (interior) 106.5, 104.4 (35)
Cellulose II 107.6 (36)
C-4 Glc Cellulose Ia (interior) 90.3 and 89.4 (37, 38)
Cellulose Ib (interior) 89.4 and 88.5 (37, 38)
Cellulose I (surface) 84.8 and 83.9 (39)
Cellulose II 89.4 and 88.2 (40)
C-6 Glc Cellulose Ia (interior) 65.2 (36)
Cellulose Ib (interior) 66.3 and 65.2 (36)
Cellulose I (surface) 63.1 and 62.1 (36)
Cellulose II 63.5 and 62.5 (38)
C-1 XG-Xyl Cyclamen seed XG 99.8 (41)
Tamarind XG 99.8 (42)
Arabidopsis cell walls 100.0 (43)
(chemically extracted
to remove pectic
polysaccharides)
Tamarind/BC 100.3 (44)
composite
Rubus suspension cells 100.4a (45)
Tamarind XG 99.1–100a (46)
Cyclamen seed XG 100.2a (41)
C-4 XG-Xyl Tamarind XG 70.1–70.4a (46)
Cyclamen seed XG 70.8a (41)
C-1 XG-Glc Tamarind XG 103.6 (42)
Rubus suspension cells 103.9a (45)
Tamarind XG 103.1–103.7a (46)
Cyclamen seed XG 103.5a (41)
C-4 Glc Cyclamen seed XG 82–85 (41)
Tamarind XG 70.3–80.5a (46)
Cyclamen seed XG 79.5–80.8a (41)
C-1 5-Ara Arabinans 108.4a (47)
5-Ara Arabinans 108.5a (48)
4-Gal Galactans 105.2a (47)
4-GalA HG 101.1 (29)
4-GalA HG 99.8a (48)
(continued)
 Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 185

Table 1
(continued)

Assignment Chemical shift ppm Reference

C-4 4-Ara Arabinans 83.2a (47)

4-GalA HG 79–80 (29)
4-Gal Galactans 78.5a (47)
4-Ara Arabinans 77.7a (47)
t-Gal Galactans 61.8a (47)
-COOCH3 HG 52.8–53.7 (29)
BC bacterial cellulose
a
Solution NMR spectroscopy; all other values were obtained with CP/MAS of solid
material

4. Ponceau 2R solution: Make up an aqueous solution of 0.2%

w/v Ponceau 2 R (CI 16150) and add 2 drops/100 mL of
18 M H2SO4. Store solution at 4°C.
5. Polydimethylsilane was from Hüls America Incorporated,
Cincinnati, OH, USA (see Note 2).
6. Polytrichlorofluoroethylene grease was from Halocarbon
Products Corporation, River Edge, NJ, USA (see Note 3).
7. Polytron blender (Model PT10-35, Kinematica, Luzern,
Switzerland).
8. 15 mL Tenbroeck ground glass homogenizer (Kontes Glass
Company, Vineland, NJ, USA).
9. Centrifuge.
10. Nylon mesh with pore size 11 mm.
11. 80% v/v ethanol.
12. Glass-fibre filter (GF/C, Whatman Scientific, Maidstone,
Kent, UK).
13. Freeze-drier.
14. 7 mm-diameter cylindrical silicon nitride rotor with Vespel
end caps.
15. Inova-200 NMR spectrometer (Varian, Palo Alto, CA, USA).

3. Methods

3.1. Preparation A cell-wall preparation should be free of cytoplasmic contents as

of Cell Walls NMR signals from proteins and lipid can interfere with the NMR
signals from polysaccharides. To achieve this, cell walls are
186 Bootten et al.

isolated by mechanically breaking the cells open and washing out

the cells contents with cold aqueous buffer (26, 27). All proce-
dures are carried out at 4°C.
1. Plant tissue (approximately 20 g wet weight) is homogenised
in 100 mL of cell wall isolation buffer, using a Polytron
blender on full power (3 times for 20 s).
2. The tissue can be further homogenised using a 15-mL-
Tenbroeck ground-glass homogeniser. Breakage of the cells
is monitored using bright-field microscopy after staining with
Ponceau 2R solution, which stains protein red (26).
3. The homogenate is centrifuged (250 × g, 10 min), and the
pellet washed 3 times by centrifugation with buffer (with
DTT omitted), followed by 6 times with water.
4. The pellet is resuspended in water, washed onto nylon mesh
(pore size 11 mm), and the residue on the mesh washed with
water (500 mL).
5. The water content of the cell-wall preparation should be
reduced but not completely eliminated for solid-state NMR
(see Note 4). Two methods can be used to do this:
(a) The preparation is washed 3 times by centrifugation with
80% (v/v) ethanol and the final suspension kept at 4°C in
80% (v/v) ethanol until NMR spectroscopy can be done
(see Note 5). Preparations in 80% (v/v) aqueous ethanol
are prepared for NMR by filtering portions of the suspen-
sions onto a glass-fibre filter (GF/C, Whatman Scientific,
Maidstone, Kent, UK) and part-drying in air to a water-
content of approximately 40% (w/w). This weight is esti-
mated from the dry weight of an aliquot of preparation
that is removed and freeze-dried.
(b) An aliquot of the preparation is removed and freeze-dried
to estimate the water. The remaining preparation is fro-
zen and dried, carefully, under vacuum to a water-content
of approximately 40% (w/w) based on the dry weight of
the sub-­sample. If this method is used, NMR spectros-
copy must be done immediately to avoid possible sample
degradation. The exact moisture content of the prepara-
tion can be determined by oven drying to constant
weight (105°C) following the completion of the NMR
experiments.

3.2. Solid-State 1. The never-dried cell-wall preparation is packed in a 7-mm-

13
C-NMR Spectroscopy diameter cylindrical silicon nitride rotor, and retained with
Vespel end caps. An internal standard of polydimethylsilane
can be added to the centre of the sample during the packing
of the rotor. Polydimethylsilane contributes a 13C signal at
 Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 187

−1.96 ppm (see Note 2). As the cell-wall preparations

are partly hydrated, polytrichlorofluoroethylene grease (see
Note 3) is used to ensure a water-tight seal between the cyl-
inder and the end caps. The grease should be spread on the
internal surface of the rotor, not on the cap, to avoid expel-
ling grease when the cap is inserted. It is important that there
should be no grease on the external surfaces of the rotor or
caps, otherwise the NMR stator will become contaminated
and sample spinning will be impeded.
2. The rotor is spun at 4 kHz in a magic-angle spinning probe
for 13C NMR spectroscopy at 50.3 MHz. In the worked
example, the probe was supplied by Doty Scientific (Columbia,
SC, USA) and the Inova-200 spectrometer was supplied by
Varian (Palo Alto, CA, USA). Slower spinning can cause
interference from spinning-sideband signals, while faster spin-
ning can cause physical degradation of the cell-walls because
of the centrifugal forces generated.

3.3. CP/MAS In cross-polarisation (CP) NMR experiments, each 90° proton

Spectroscopy preparation pulse is followed by a 1-ms CP contact time, 51 ms
of data acquisition, and a recovery delay of 1.0 s before the
sequence was repeated (see Note 6). The correct duration of
the 90° proton preparation pulse can be measured by using trial
­values and selecting the value that provides the best signal
strength. A typical value is 6 ms, as used to illustrate this chapter.
Both proton and 13C transmitters are left on for the duration of
the contact time. It is also important that the power levels are
adjusted for a Hartman-Hahn match, that is, the measurement
of a 90° pulse should give the same value for both nuclei.
Measurement of a 90° pulse for 13C is discussed below, under
the Subheading 3.5.2. The proton transmitter output is
increased during data acquisition, to provide adequate power
for spin decoupling, i.e., a target power level corresponding to
a precession frequency >40 kHz and typically between 53 and
59 kHz as in the experiments used to illustrate this chapter.
Lower power levels will cause noticeable broadening of the
NMR signals.

3.4. Proton Spin Proton rotating-frame spin relaxation with time constant T1r(H)
Relaxation and proton spin-spin relaxation with time constant T2(H) are
Experiments characterised by inserting relaxation intervals of duration t1 or t2,
respectively, between the proton preparation pulse and the CP
contact time (Fig. 2). The values for t1 and t2 are chosen to be
within the ranges of values for T1r(H) and T2(H), respectively, to
optimise the signal-to-noise ratios in separate subspectra (28).
Protons are spin-locked during t1, but the proton transmitter is
switched off during t2.
188 Bootten et al.

PSRE NMR subspectra are generated by combining spectra

labelled S, S¢ and S″, where S is obtained by the normal
­cross-polarisation pulse sequence, S¢ with t1 = 4 ms and S″ with
t2 = 15 ms. The experimental spectra are obtained in the order
CP/MAS (S), proton-rotating frame experiment (S¢) and proton
spin-spin experiment (S″).
The number of transients of the pulse sequences required to
obtain adequate signal to noise ratios from averaged spectra
needed for PSRE editing varies, but 40,000 up to 100,000 tran-
sients is usual. This corresponds to 30–80 h of data accumulation
time for a full PSRE experiment.

3.5. Separating The principles behind PSRE NMR have been described in
the Proton-Spin Newman (17) and are summarised in the Introduction. In the
Relaxation Edited simplest case, a spectrum S is the sum of subspectra A and B from
(PSRE) Subspectra two distinct types of domains.
A partly-relaxed spectrum S¢ is then:
S′ = f a A + f b B, (1)

where fa and fb are signal suppression factors.

The subspectra can then be separated by computing:
A = kS + k ′ S′, (2a)

B = (1 − k)S − k ′ S′, (2b)

where
k = f b / ( f b − f a ), (3a)

k ′ = −1 / ( f b − f a ). (3b)

In the context of a cell-wall preparation, subspectra A and B usually

contain signals from the cellulose and non-cellulosic polysaccha-
rides, respectively.
Two NMR signals, characteristic of the two mobility domains,
are selected for proton rotating-frame relaxation experiments. The
signal at 89 ppm (assigned to C-4 of cellulose crystallite-interior)
is selected as representative of cellulose crystallites, appearing at a
chemical shift for which there is little overlap with signals from
other polysaccharides. A signal at 69 ppm (assigned to C-2, C-3,
and C-5 of GalA residues in pectic homogalacturonans) (29, 30)
is selected as representative of mobile polysaccharides for those
primary walls containing high proportions of pectic polysaccha-
rides (1, 2). (NB. See ref. (25) for appropriate editing signals for
GAX-rich primary walls).
1. Because the T1r(H) relaxation time constants for the 69 ppm
signal were not greatly different, it was not possible to achieve
Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 189

Table 2
Suppression factors, linear combinations and the
corresponding relaxation time constants used to generate
PSRE subspectra A and B for mung bean cell walls

Suppression factors
T1r(H) ms fa 0.69
fb 0.27
T2(H) ms fa 0.30
fb 0.60
Linear combinations
T1r(H) ms Subspectrum A −0.64S + 2.38S¢
Subspectrum B 1.64S + 2.38S¢
T2(H) ms Subspectrum A 2.00S–3.33S″
Subspectrum B −1.00S + 3.33S″
Relaxation time constants
T1r(H) ms Subspectrum A 10.8
Subspectrum B 3.1
T2(H) ms Subspectrum A 9.7
Subspectrum B 14.8

total elimination of signals from non-cellulosic polysaccha-

rides without also suppressing signals from cellulose, so linear
combinations of S and S¢ are generated to enhance signal sup-
pression (Table 2).
2. Initial estimates of fa and fb are calculated from signal heights
(Fig. 3).
3. The values for fa and fb are used to determine the linear com-
binations used to separate the subspectra from the CP/MAS
data, using (3a) and (3b) then (2a) and (2b).
4. Values for fa and fb are then adjusted until the signal at 89
ppm is eliminated from subspectrum B and signal at 69 ppm,
assigned to pectic polysaccharides, are suppressed in subspec-
trum A without allowing any signals to become inverted. For
example, for the spectra from mung bean (Vigna radiata)
cell walls shown in Fig. 3, the final suppression factors from
the proton rotating frame experiment were:
fa = 0.69 and fb = 0.27; therefore, k = −0.64 and k¢ = −2.38
I n the worked example, the linear combinations used to sepa-
rate subspectra A and B from the CP/MAS NMR data were:
Subspectrum A = −0.64S + 2.38S¢ and Subspectrum
B = 1.64S + 2.38S¢
190 Bootten et al.

C-2,3,5
Example: T1r (H) fa = (S´d (89 ppm) /Sd (89 ppm))

fb = (S´d (69 ppm) /Sd (69 ppm))

where d = length (mm)

C-1

C-6
C-4

S´

S´´

190 170 150 130 110 90 70 50 30 10

Chemical shift (ppm)

Fig. 3. Calculating suppression factors from the CP/MAS and PSRE spectra obtained of
a mung bean cell-wall preparation. S obtained by the normal CP/MAS pulse sequence;
S ¢ with 4 ms of proton rotating-frame spin relaxation; S″ with 15 ms of proton spin-spin
relaxation. Carbon numbers refer to the Glc residues of cellulose.

The separation is successful when signals assigned to cel-

lulose and mobile non-cellulosic polysaccharides appeared in
subspectra A and B, respectively. The resulting separated sub-
spectra are shown in Fig. 4.
5. The final values of fa and fb can then be used to calculate
improved values of the proton spin relaxation time constants.
For example, if the spin relaxation process is exponential
then:

f a = exp[−t 1 (ms) / T1 r (H)]

Taking natural logarithms of both sides:

ln f a = −t 1 (ms) / T1r (H)

This can be rearranged to:

T1r (H) = −t 1 (ms) / ln f a

If fa is 0.69 and the t1 value for the T1r(H) experiment is 4 ms,

then the estimated T1r(H) value for subspectrum A in the
proton rotating-frame experiments is 10.8 ms. The ­relaxation
Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 191

CP/MAS spectrum

Subspectrum A

Subspectrum B

190 170 150 130 110 90 70 50 30 10

Chemical shift (ppm)

Fig. 4. Normal CP/MAS spectrum and separated PSRE subspectra of a cell-wall preparation
of the hypocotyls from mung bean seedlings. Subspectra are obtained by exploiting the
differences in proton rotating-frame spin relaxation. Subspectra A and B display signals
assigned primarily to cellulose and the non-cellulosic matrix, respectively.

time constant for subspectrum B using fb can be calculated

using the same equations.
6. The editing process is repeated to separate the T2(H) sub-
spectra. This spin relaxation process suppresses cellulose sig-
nals more than the other signals, but does not eliminate them
entirely. Linear combinations can again be generated to
enhance the amount of suppression. Like the T1r(H) relax-
ation, the separation is successful in that signals assigned to
cellulose and non-cellulosic polysaccharides appear in A and
B, respectively.
7. As for T1r(H) relaxation, the final values of fa and fb can be
used to calculate improved values of the spin-spin relaxation
time constants for the two subspectra. However, the relax-
ation curves for rigid solids, such as crystalline cellulose and
more mobile polysaccharides, such as pectic homogalactur-
onans, are described by different functions (6):
f a = exp[−{t 2 (ms) / T2 (H)}2 / 2],

f b = exp[−t 2 (ms) / T2 (H)]

Table 2 shows the adjusted suppression factors, linear combi-

nations and relaxation time constants from the PSRE experi-
ments obtained for a preparation of mung bean cell walls.
192 Bootten et al.

3.5.1. Spin Echo NMR As discussed above, unlike T1r(H) and T2(H), T2(C) is sensitive to
with PSRE (SE-PSRE) segmental motion of the polysaccharide at the site of the 13C
nucleus and is insensitive to the motion of more distant polysac-
charides. Therefore, in SE-PSRE NMR experiments, the T2(C)
values for a polysaccharide will indicate the mobility of that par-
ticular polysaccharide, whereas the T1r(H) and T2(H) relaxation
values will reflect the averaged molecular motion of the
­surrounding polysaccharides. For example, if a relatively rigid XG
chain extends through a domain containing relatively flexible
­pectic polysaccharides, values of T1r(H) and T2(H) measured from
XG signals will reflect molecular motion in the pectic polysaccha-
ride environment. In this example, the values of T2(C) measured
from xyloglucan signals will reflect the rigidity of the XG chain and
not the mobility of the pectic polysaccharide environment (11).
1. T2(C) relaxation is characterised by a spin-echo sequence in
which a delay of duration t2 is inserted between the CP con-
tact time and commencement of data acquisition (6). A 180°
refocusing pulse is applied halfway through t2.
2. Multiple values of t2 can be chosen so that 0.5t2 is always a
multiple of the rotor rotation period, and protons are decou-
pled with an attenuated power output corresponding to a
precession frequency of 43 kHz throughout t2.
3. Proton spin relaxation edited (PSRE) NMR subspectra
are generated for each of the T2(C) spectra using the same S
and S¢ values used for editing the proton spin-relaxation
spectra.
4. Relaxation time constants for T2(C) can be calculated as
described in Subheading 3.5. As indicated above, these relax-
ation values will reflect the mobility of that particular polysac-
charide in the cell-wall preparation.

3.5.2. Single Pulse As discussed in the Introduction, SPE/MAS is a useful technique

Excitation/Magic Angle for investigating the mobility of the very mobile polysaccharides
Spinning (SPE/MAS) NMR in the cell wall, such as the pectic polysaccharide side chains
on RG-I.
1. Single pulse excitation/magic angle spinning (SPE/MAS)
spectra are obtained with a pulse sequence in which each 90°
13
C excitation pulse is followed by 51 ms of data acquisition
time and a 1 s recovery delay (25). The correct duration of
the 90° 13C pulse can be measured by using trial values and
selecting the value that provides the best signal strength.
A typical value is 6 ms, as used to illustrate this chapter.
The 1.0 s delay is used to maximise the response from mobile
components and minimise the response from rigid
­components of the cell walls (25). The proton decoupler
 Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 193

Gal-1

Ara-1 Gal-4

Ara-4

120 110 100 90 80 70 60 50

Chemical shift (ppm)

Fig. 5. SPE/MAS NMR spectrum of cell-wall preparation from hypocotyls of mung

bean seedlings. Assignments refer to the carbon numbers of Ara or Gal in arabinans or
galactans (see Table 1).

transmitter power is increased during data acquisition and

should correspond to a precession frequency >40 kHz and
preferably between 53 and 59 kHz.
2. As with all NMR experiments, the signal to noise ratio will
improve with increased number of experimental transients.
However, care must be taken to avoid sample degradation
during long experiments (see Note 7). The SPE/MAS spec-
trum for a cell-wall preparation of mung bean hypocotyls,
shown in Fig. 5, was acquired by the averaging of 23,931
experiment transients. This equated to approximately 6.7 h of
data accumulation time.

3.5.3. Spin Echo with SPE/ The SE-SPE/MAS pulse sequence is similar to the SE-PSRE
MAS NMR (SE-SPE/MAS) sequence except it is run with SPE/MAS and not PSRE, and
spectra are not separated into mobility domains. The SE-SPE/
MAS experiments will provide mobility information about a par-
ticular monosaccharide component within the polysaccharide,
provided the signal is relatively free from other overlapping
signals.
Each 90° 13C excitation pulse is followed by a t2 delay then
51 ms of data acquisition time and a 1 s recovery delay. A 180°
refocusing pulse is applied halfway through t2, and values of t2 are
chosen so that 0.5t2 is always a multiple of the rotor rotation
period, and protons are decoupled with an attenuated power out-
put corresponding to a precession frequency >40 kHz through-
out t2. The durations of the 90° and 180° 13C pulses are typically
6 and 12 ms respectively.
194 Bootten et al.

4. Notes

1. The reducing agent DTT is added to the isolation buffer to

prevent the oxidation of phenols to quinones (26, 31).
Quinones can polymerise to form red/brown products and
may also form covalent bonds with proteins resulting in insol-
uble precipitates (26). Alternatively, buffer containing 10 mM
2-mercaptoethanol may replace DTT.
2. The negative chemical shift (−1.96 ppm) of polydimethylsi-
lane is outside the range normally seen for cell-wall polysac-
charides, and therefore does not interfere with signals from
cell-wall material. An alternative standard could be polydim-
ethylsiloxane, showing a 13C signal at 1.50 ppm, which is
closer to the range associated with cell wall material (32).
Polydimethylsiloxane is readily available from Sigma-Aldrich,
whereas no supplier of small amounts of polydimethylsilane
could currently be found by the authors.
3. The polytrichlorofluoroethylene grease from Halocarbon is
thickened with silica and provides a water tight seal for the Vespel
end caps. Neither the grease nor the thickener contributes signal
strength to the cross-polarisation 13C NMR spectra.
4. Moisture is essential for distinguishing polysaccharides in
different molecular environments in a cell-wall preparation,
for example, cellulose molecules at the surface of the crystal-
lite (7). However, over-drying or drying and re-hydrating of
the preparation can result in irreversible changes in the
molecular order of the polysaccharides (33, 34). Excessively
high moisture contents dilute down the amount of carbon
that can be packed into the rotor and therefore diminish the
NMR ­signal. The ideal water content is between 30 and 50%
w/w in many cases.
5. Exposure of cell-wall preparations to ethanol may, in principle
physically alter polysaccharides (33), as well as denature and
precipitate proteins. Although we have not seen evidence in
the NMR spectra of such changes, we recommend caution.
6. Preliminary T1(H) experiments indicated that a 1.0 s delay
was adequate for the recovery of proton magnetisation in
relatively mobile segments of polysaccharides, as was also
shown in experiments on other cell walls (6, 7).
7. It is desirable to check for sample degradation using SPE
experiments. This can be achieved by breaking the experi-
ment into several periods of data accumulation and comparing
the spectra to test for changes. If the spectra are all similar,
they may be added together to improve the signal-to-noise
ratio.
 Using Solid-State 13 C NMR Spectroscopy to Study the Molecular Organisation 195

References

1. Harris, P. J. (2005) Diversity in plant cell 13. Horii, F., Hirai, A., and Kitamaru, R. (1984)
walls. In: Plant diversity and evolution: geno- Cross-polarization/magic angle spinning 13C-
typic and phenotypic variation in higher plants. NMR study. Molecular chain conformations
Henry, R. J., ed. CAB International: of native and regenerated cellulose. In:
Wallingford, pp 201–227. Polymers for fibers and elastomers. Arthur, J.C.
2. Harris, P. J., and Stone, B. A. (2008) Chemistry Jr, Diefendorf, R.J., Yen, T.F., Needles, H.L.,
and molecular organization of plant cell walls. Schaefgen, J.R., Jaffe, M., and Logothetis,
In: Biomass recalcitrance: deconstructing the A.L. eds. American Chemical Society, 260,
plant cell wall for bioenergy. Himmel, M. E., pp 27–42.
ed. Blackwell Publishing, Oxford. pp 61–93. 14. Jarvis, M. C. (1994) Relationship of chemical
3. Mohnen, D. (2008) Pectin structure and bio- shift to glycosidic conformation in the solid
synthesis. Curr Opinion Plant Biol 11, state 13C NMR spectra of (1→4)-linked glu-
266–277. cose polymers and oligomers: anomeric and
4. Hsieh, Y. S. Y., and Harris, P. J. (2009) related effects. Carbohydr Res 259, 311–318.
Xyloglucans of monocotyledons have diverse 15. Jarvis, M. C., and Apperley, D. C. (1990)
structures. Mol Plant 2, 943–965. Direct observation of cell wall structure in
5. Trethewey, J. A. K., Campbell, L. M., and ­living plant tissues by solid-state 13C NMR
Harris, P. J. (2005) (1→3),(1→4)-b-d-­glucans spectroscopy. Plant Physiol 92, 61–65.
in the cell walls of the Poales (sensu lato): an 16. Tang, H., Belton, P. S., Ng, A., and Ryden, P.
immunogold labelling study using a monoclo- (1999) 13C MAS NMR studies of the effects
nal antibody. Am J Bot 92, 1660–1674. of hydration on the cell walls of potatoes and
6. Newman, R. H., Davies, L. M., and Harris, P. J. Chinese water chestnuts. J Agric Food Chem
(1996) Solid-state 13C nuclear magnetic reso- 47, 510–517.
nance characterisation of cellulose in the cell 17. Newman, R. H. (1999) Editing the information
walls of Arabidopsis thaliana leaves. Plant in solid-state carbon-13 NMR spectra of food.
Physiol 111, 475–485. In: Advances in magnetic resonance in food
7. Newman, R. H., Ha, M.-A., and Melton, L. ­science. Belton, P. S, Hills, B. P, and Webb, G.
D. (1994) Solid-state 13C NMR investigation A., eds. The Royal Society of Chemistry:
of molecular ordering in the cellulose of apple Cambridge, pp 144–157.
cell walls. J Agric Food Chem 42, 1402–1406. 18. Newman, R. H. (1992) Solid-state carbon-13
8. Newman, R. H. (1999) Estimation of the lat- NMR spectroscopy of multiphase biomateri-
eral dimensions of cellulose crystallites using als. In: Viscoelasticity of biomaterials. Glasser,
13
C NMR signal strengths. Solid State Nucl W. G., and Hatakeyama, H., eds. American
Magnet Reson 15, 21–29. Chemical Society: Washington, pp 311–319.
9. Newman, R. H., and Davidson, T. C. (2004) 19. Tekely, P., and Vignon, M. R. (1987) Proton
Molecular conformations at the cellulose- T1 and T2 relaxation times of wood compo-
water interface. Cellulose 11, 23–32. nents using 13C CP/MAS NMR. J Polym Sci
Part C Polym Lett 25, 257–261.
10. Bootten, T. J., Harris, P. J., Melton, L. D.,
and Newman, R. H. (2008) WAXS and 13C- 20. Hediger, S., Emsley, L., and Fischer, M.
NMR study of Gluconoacetobacter xylinus (1999) Solid-state NMR characterization of
­cellulose in composites with tamarind xylog- hydration on polymer mobility in onion cell-
lucan. Carbohyr Res 343, 221–229. wall material. Carbohydr Res 322, 102–112.
11. Bootten, T. J., Harris, P. J., Melton, L. D., 21. Zumbulyadis, N. (1983) Selective carbon
and Newman, R. H. (2004) Solid-state 13C- excitation and the detection of spatial hetero-
NMR spectroscopy shows that the xyloglu- geneity in cross-polarization magic-angle-
cans in the primary cell walls of mung bean spinning NMR. J Magn Reson 53, 486–494.
(Vigna radiata L.) occur in different 22. Tang, H., and Hills, B. P. (2003) Use of 13C
domains: a new model for xyloglucan-­ MAS NMR to study domain structure and
cellulose interactions in the cell wall. J Exp dynamics of polysaccharides in the native starch
Bot 55, 571–583. granules. Biomacromolecules 4, 1269–1276.
12. Levy, S., Maclachlan, G., and Staehelin, L. A. 23. VanderHart, D. L. (1987) Natural-abundance
(1997) Xyloglucan sidechains modulate
13
C-13C spin exchange in rigid crystalline
­binding to cellulose during in vitro binding ­solids. J Magn Reson 72, 13–47.
assays as predicted by conformational dynam- 24. Foster, T. J, Ablett, S., McCann, M. C.,
ics simulations. Plant J 11, 373–386. and Gidley, M. J. (1996) Mobility-resolved
196 Bootten et al.

13
C-NMR spectroscopy of primary plant 37. Atalla, R. H., and Vanderhart, D. L. (1984)
cell walls. Biopolymers 39, 51–66. Native cellulose: a composite of two distinct
25. Smith, B. G., Harris, P. J., Melton, L. D, and crystalline forms. Science 223, 283–285.
Newman, R. H. (1998) The range of mobility 38. Newman, R. H., and Hemmingson, J. A.
of the non-cellulosic polysaccharides is similar in (1995) Carbon-13 NMR distinction between
primary cell walls with different polysaccharide categories of molecular order and disorder in
compositions. Physiol Plant 103, 233–246. cellulose. Cellulose 2, 95–110.
26. Harris, P. J. (1983) Cell walls. In: Isolation of 39. Newman, R. H. (1998) Evidence for assign-
membranes and organelles from plant cell walls. ment of 13C NMR signals to cellulose crystal-
Hall, J. L. and Moore, A. L., eds. Academic: lite surfaces in wood, pulp and isolated
London, pp 25–53. celluloses. Holzforschung 52, 157–159.
27. Melton, L. D., and Smith, B. G. (2005). 40. Hirai, A., Horii, F., and Kitamaru, R. (1990)
Isolation of plant cell walls and fractionation Carbon-13 spin-lattice relaxation behaviour
of cell wall polysaccharides. In: Handbook of of the crystalline and non-crystalline compo-
food analytical chemistry: water, proteins, nents of native and regenerated celluloses.
enzymes, lipids and carbohydrates. Wrolstad, R. Cellulose Chem Technol 24, 703–711.
E., ed. Wiley: Hoboken, pp 697–719. 41. Braccini, I., Hervé du Penhoat, C., Michon, V.,
28. Newman, R. H., and Hemmingson, J. A. Goldberg. R., Clochard, M., Jarvis, M. C.,
(1990) Determination of the degree of cellu- Huang, Z.-H., and Gage, D.A. (1995) Structural
lose crystalinity in wood by carbon-13 nuclear analysis of cyclamen seed ­xyloglucan oligosac-
magnetic resonance spectroscopy. charides using cellulase digestion and spectro-
Holzforschung 44, 351–355. scopic methods. Carbohydr Res 276, 167–181.
29. Sinitsya, A., Čopíková, J., and Pavlíková, H. 42. Gidley, M. J., Lillford, P. J., Rowlands, D. W.,
(1998) 13C CP/MAS NMR spectroscopy in Lang, P., Dentini, M., Crescenzi, V., Edwards,
the analysis of pectins. J Carbohydr Chem 17, M., Fanutti, C., and Reid, J. S. G. (1991)
279–292. Structure and solution properties of tamarind-
30. Jarvis, M. C., and Apperley, D. C. (1995) seed polysaccharide. Carbohydr Res 214,
Chain conformation in concentrated pectic 299–314.
gels: evidence from 13C NMR. Carbohydr Res 43. Davies, L. M., Harris, P. J., and Newman, R.
275, 131–145. H. (2002) Molecular ordering of cellulose
31. Bootten, T. J., Harris, P. J., Melton, L. D., after extraction of polysaccharides from pri-
and Newman, R. H. (2009) A Solid-state 13C- mary cell walls of Arabidopsis thaliana: a
NMR study of a composite of tobacco xylog- solid-state CP/MAS 13C NMR study.
lucan and Gluconacetobacter xylinus cellulose: Carbohydr Res 337, 587–593.
molecular interactions between the compo- 44. Whitney, S. E. C., Brigham, J. E., Darke, A. H.,
nent polysaccharides. Biomacromolecules 10, Reid, J. S. G., and Gidley, M. J. (1995) In
2961–2967. vitro assembly of cellulose/xyloglucan net-
32. Jelinski, L. W., and Melchior, M. T. (1996) works: ultrastructural and molecular aspects.
High-resolution NMR of solids. In: NMR Plant J 8, 491–504.
spectroscopy techniques. Practical Spectroscopy 45. Joseleau, J. P., Cartier, N., Chambat, G., Faik, A.,
Series, 2. Bruch, M. D., ed. Marcel Dekker: and Ruel, K. (1992) Structural features and
New York, pp 417–486. biological activity of xyloglucans from suspen-
33. Thimm, J. C., Burritt D. J., Ducker, W. A., and sion-cultured plant cells. Biochemie 74, 81–88.
Melton, L. D. (2000). Celery (Apium graveolens L.) 46. York, W. S., Harvey, L. K., Guillen, R.,
parenchyma cell walls examined by atomic force Albersheim, P., and Darvill, A. G. (1993)
microscopy. Planta 212, 25–32. Structural analysis of tamarind seed xyloglu-
34. Newman, R. H. (2004) Carbon-13 NMR evi- can oligosacharides using b-galactosidase
dence for cocrystallization of cellulose as a digestion and spectroscopic methods.
mechanism for hornification of bleached kraft Carbohydr Res 248, 285–301.
pulp. Cellulose 11, 45–52. 47. Ryden, P., Colquhoun, I. J., and Selvendran,
35. Newman, R. H. (1997) Crystalline forms of R. R. (1989) Investigation of structural
cellulose in the silver tree fern. Cyathea ­features of the pectic polysaccharides of onion
Dealbata Cellulose 4, 269–278. by 13C-N.M.R. spectroscopy. Carbohydr Res
36. Newman, R. H., and Redgwell, R. J. (2002) 185, 233–237.
Cell wall changes in ripening kiwifruit: 13C 48. Saulnier, L., Brillouet, J. -M., and Joseleau, J. -P.
solid state NMR characterisation of relatively (1988) Structural studies of pectic substances
rigid cell wall polymers. Carbohydr Polym 49, from the pulp of grape berries. Carbohydr Res
121–129. 182, 63–78.
 Chapter 14

Formation of Cellulose-Based Composites

with Hemicelluloses and Pectins Using
Gluconacetobacter Fermentation
Deirdre Mikkelsen and Michael J. Gidley

Abstract
Gluconacetobacter xylinus synthesises cellulose in an analogous fashion to plants. Through fermentation
of Ga. xylinus in media containing cell wall polysaccharides from the hemicellulose and/or pectin fami-
lies, composites with cellulose can be produced. These serve as general models for the assembly, struc-
ture, and properties of plant cell walls. By studying structure/property relationships of cellulose
composites, the effects of defined hemicellulose and/or pectin polysaccharide structures can be investi-
gated. The macroscopic nature of the composites also allows composite mechanical properties to be
characterised.
The method for producing cellulose-based composites involves reviving and then culturing Ga.
xylinus in the presence of desired hemicelluloses and/or pectins. Different conditions are required for
construction of hemicellulose- and pectin-containing composites. Fermentation results in a floating mat
or pellicle of cellulose-based composite that can be recovered, washed, and then studied under hydrated
conditions without any need for intermediate drying.

Key words: Plant cell wall, Cellulose, Composites, Gluconacetobacter xylinus, Pectin, Hemicellulose,
Arabinoxylan, b-Glucan, Xyloglucan

1. Introduction

The cell walls of plants are typically complex in terms of their

measured average composition, with variation being exhibited
not only between different plant types, but also between local tis-
sue types and even within a single cell wall. While some informa-
tion on the relationships between composition and properties of
cell walls can be deduced through studies of e.g. plant mutants
lacking defined compositional features, the isolation of plant cell
wall material for the study of structure/property relationships has

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_14, © Springer Science+Business Media, LLC 2011

197
198 Mikkelsen and Gidley

Fig. 1. Schematic illustration of cellulose biosynthesis by Gluconacetobacter xylinus (adapted from Brown (1)).

major limitations due to microheterogeneity within the plant and

the need for harsh extraction conditions.
Gluconacetobacter xylinus (previously known as Acetobacter
xylinus or A. xylinum) has been used as a model for cellulose bio-
synthesis because it has the same general features of cellulose
deposition as plants (Fig. 1). A transmembrane assembly of cata-
lytic (cellulose synthase) and structural proteins produces strands
of (1→4)-b-d-glucan that associate, first into small sub-fibrils,
then into microfibrils which subsequently aggregate into a char-
acteristic flat, twisting ribbon (Fig. 1). Although plant cellulose
is usually more cylindrical in its final aggregated state, all other
stages of cellulose synthesis are shared with Ga. xylinus. Cellulose
secreted by bacteria into an aqueous fermentation medium is used
as a model for plant cellulose secreted extracellularly into a nascent
cell wall. In plants, cellulose is deposited into an environment
containing a complex mixture of other cell wall polymers. In the
bacterial model system, the extracellular environment is ­controlled
and can be used to investigate the potential of each type of cell
wall polymer to form composite structures with cellulose and to
investigate the molecular, microscopic, and macroscopic ­properties
of the resulting materials. Validation of the system as a model
comes from (a) microscopic observation of similar architectures
in cellulose/xyloglucan composites (2) as in de-pectinated
cell walls of e.g. onion (3), (b) responsiveness of hemicellulose/­
cellulose composites to application of expansin proteins (4) and
xyloglucan endo-transglycosylase (5) similar to that found or
 Formation of Cellulose-Based Composites with Hemicelluloses and Pectins 199

expected in plant tissues, and (c) conversion of highly crystalline

Ga. xylinus cellulose into a less crystalline form (characteristic of
plant cell walls) in the presence of hemicelluloses (2, 6). While
there are limitations to this model system, it is experimentally
straightforward and provides insights into the effects of polymer
components of defined chemistry that are not possible from cell
walls isolated from plants. Furthermore, as cellulosic materials can
be produced in multi-centimetre pieces, mechanical measurements
can be made that are highly informative in defining the principles
of cell wall materials properties (7–9).
In the absence of polymers in the fermentation medium, a
cellulose “mat” or pellicle is produced that is highly hydrated but
mechanically tough. In outline, the formation of cellulose com-
posites with hemicellulose or pectin is accomplished by ferment-
ing Ga. xylinus in liquid fermentation containing the hemicellulose
and/or pectin of interest. The cellulose produced by bacteria
(Fig. 1) thus comes into contact with the added polymer(s) as
soon as it is secreted from the bacteria. For hemicelluloses that
can associate (bind) molecularly to cellulose, the fact that the
crystallinity of cellulose is affected greatly (2, 6) is interpreted to
mean that added polymers can access cellulose microfibrils prior
to their aggregation into the final ribbon assembly (Fig. 1). The
negative charge on pectins means that there is no direct binding
between the two backbones (although arabinan side chains may
bind (10)), and composite formation requires the presence of a
pre-formed pectin network (11). Thus, the fermentation medium
contains a weak gel which can be varied by choice of pectin type
(particularly degree of methyl esterification) and calcium level.
Following fermentation, composites are recovered from the
medium, and washed to remove bacteria and any polymers that
are held non-specifically. The isolated solid composite material
can then be analysed by chemical, spectroscopic, microscopic, or
mechanical methods.

2. Materials

1. Laminar flow cabinet and appropriate facilities for observing

good microbiological practice (e.g. autoclave).
2. Bacterial strain: Ga. xylinus (formerly A. xylinus and A. xyli-
num) strain ATCC 53524 frozen (−80°C) stock (see Note 1).
Growth media:
3. Hestrin and Schramm (HS) medium (12) containing (per
litre) 5 g peptone, 5 g yeast extract, 3.38 g Na2HPO4 · H2O,
1.15 g citric acid, and 20 g glucose (see Note 2). HS agar
medium (containing 15 g/L agar) is used for maintenance
200 Mikkelsen and Gidley

and long-term storage of the bacterial strain. HS liquid medium

is used for composite preparation. Preferably, the medium is
made fresh as required. Alternatively, sterile medium stored at
4°C is stable for up to 1 week.
4. Modified HS broth medium containing hemicelluloses: The
medium is prepared as a concentrated (2×) solution by add-
ing 5 g peptone, 5 g yeast extract, 3.38 g Na2HPO4 · H2O,
and 1.15 g citric acid to 400 mL deionised water. The pH is
adjusted to 5.0 with 10 M HCl, and the volume of the solu-
tion is adjusted to 460 mL with sterile deionised water. The
medium is sterilised by autoclaving at 121°C for 15 min.
After cooling the sterile medium to 55°C, 40 mL of 0.22 mm
filtered sterilised glucose solution (50% w/v), and 500 mL of
the 1% (w/v) polysaccharide solution are added aseptically.
5. Modified HS broth medium containing pectins: The medium
is prepared as a concentrated (2×) solution by adding 5 g
peptone, 5 g yeast extract, 3.38 g Na2HPO4·H2O, and 1.15 g
citric acid to 400 mL deionised water. The pH is adjusted to
5.0 with 10 M HCl, and the volume of the solution is adjusted
to 460 mL with sterile deionised water. The medium is steri-
lised by autoclaving at 121°C for 15 min. After cooling the
sterile medium to 55°C, 40 mL of 0.22 mm filtered sterilised
glucose solution (50% w/v), and 500 mL of the 1% (w/v)
pectin solution is added aseptically.
6. Long-term storage of Ga. xylinus: Microbank™ preservation
cryovials (Pro-Lab Diagnostics, ON, Canada).
7. Hemicellulose solutions (see Note 3): Prepare a 1% (w/v)
solution by accurately weighing out 5 g of powdered sub-
strate with a sterilised spatula into a sterile 1 L dry Pyrex bea-
ker in a laminar flow cabinet. Add a sterile magnetic stirrer
bar, followed by 400 mL of hot (90°C) sterile deionised water
(see Note 4). Immediately place the beaker containing the
mixture on a magnetic hotplate-stirrer and heat at a setting of
100°C with vigorous stirring. Loosely cover the beaker with
aluminium foil, stirring and boiling the contents until the
hemicellulose dissolves completely (see Note 5). Adjust the
volume of the solution to 500 mL with sterile deionised
water. Make solution fresh as required (see Note 6).
8. Pectin solutions (see Note 7): Prepare a 1% (w/v) solution by
accurately weighing out in a laminar flow cabinet 7 g of pow-
dered pectin with a sterilised spatula into a sterile 1 L Schott
bottle containing 489.5 mL of sterile deionised water. Add a
sterile magnetic stirrer bar, place the sealed bottle containing
the mixture on a magnetic stirrer and stir vigorously over-
night (see Note 8). Once the pectin is completely in solution,
the volume of the solution can be adjusted to 500 mL with
sterile deionised water. Make solution fresh as required.
 Formation of Cellulose-Based Composites with Hemicelluloses and Pectins 201

Fig. 2. Examples of sterile containers used for composite preparation. These include
(from left to right) sterile 150 × 20 mm and 92 × 16 mm Petri dishes, as well as sterile
screw lid containers with 70 mL capacity and 40 mm diameter.

9. 0.125 M calcium chloride (CaCl2) solution, sterilised by

autoclaving at 121°C for 15 min.
10. 12.5 mM CaCl2 solution, sterilised by autoclaving at 121°C
for 15 min.
11. Sterile culture containers (see Note 9; Fig. 2).
12. 0.02% (w/v) sodium azide solution made with deionised
water (see Note 10).
13. 0.02% (w/v) sodium azide solution made with sterile
12.5 mM CaCl2 solution (see Note 10).
14. Deionised water, sterilised by autoclaving at 121°C for
15 min.
15. Temperature-controlled growth facilities.
16. Desiccator.

3. Methods

3.1. Revival and 1. Frozen stocks: Under aseptic conditions, open the cryovial
Maintenance of Ga. and using a sterile needle, remove one bead to inoculate each
xylinus HS agar plate (inoculate two plates). Use the bead to directly
streak onto the solid medium to get isolated pure colonies
(13). Incubate at 30°C for 3–4 days.
2. Maintenance of culture: Two plates of the culture are inoculated
as one plate is reserved for starting a new working culture when
required (see Note 11), and the other plate is used for routine
transfers as required. The plates are wrapped with Parafilm® and
stored at 4°C. The cultures remain viable for 1 month.

3.2. Long-Term 1. Under aseptic conditions open the screw cap of the
Storage of Ga. xylinus Microbank™ cryovial.
(see Note 12) 2. Inoculate the cryopreservative fluid with a loopful of young
colony growth (72 h) picked from a pure culture.
202 Mikkelsen and Gidley

3. Close vial tightly and invert four to five times to emulsify

organism. A vortex mixer must not be used.
4. The excess cryopreservative must be well aspirated leaving
the inoculated beads free of liquid as much as possible.
5. Inoculated cryovials are closed finger tight and stored at
−80°C (shelf-life 3–4 years).

3.3. Composite Bacterial cellulose–hemicellulose (0.5% w/v) composites are

Preparation with ­constructed as follows:
Hemicelluloses
1. For primary inoculum preparation: Inoculate 20 mL modi-
fied HS broth medium containing hemicelluloses with a few
colonies of bacteria from the HS agar plate used to maintain
the strain. Incubate under static conditions at 30°C for 72 h
(see Note 13).
2. For scale-up preparation: Inoculate 18 mL modified HS
broth medium with 2 mL primary inoculum and incubate
without agitation at 30°C for 48 h (arabinoxylan or b-glucan
composite) or 72 h (for xyloglucan composite).
3. Harvesting of hemicellulose composites: After incubation,
the composite pellicle is removed with forceps and washed at
room temperature by gentle agitation (50 rpm), in a sterile
3 L glass beaker containing excess ice-cold sterile deionised
water (Fig. 3) (see Note 14).
4. Short-term preservation: Composites are stored in the hydra­
ted state in 0.02% (w/v) sodium azide solution at 4°C (see
Note 15).

3.4. Composite The level of pectin incorporation within the cellulose network is
Preparation dependent on the interaction between pectin and Ca2+ ions pres-
with Pectins ent in the medium. A preformed gel of the desired strength,

Fig. 3. Purification of hemicellulose composites by washing with gentle agitation (50 rpm)
at room temperature, with excess ice-cold sterile deionised water (a). Having removed
the excess medium and bacterial cells, the purified pellicle is white (b).
 Formation of Cellulose-Based Composites with Hemicelluloses and Pectins 203

which will allow for sufficient gelling as well as cellulose microfibril

penetration, must be achieved in order to form a composite. The
following detailed method describes the preparation of bacterial
cellulose–pectin composites using pectin of degree of methyl
esterification (DM) of 30%. Previous research (11) has deter-
mined that the highest pectin incorporation in cellulose compos-
ites occurs when using pectin of ~DM 30. When attempting to
construct bacterial cellulose–pectin (0.5%) composites with other
DM values, different CaCl2 concentrations may be added to the
modified HS medium as appropriate.
1. For primary inoculum preparation: Inoculate 18 mL modi-
fied HS broth medium containing pectins with two to three
colonies of bacteria from the HS agar plate used to maintain
the strain. Place the inoculated container on a platform shaker
and under vigorous shaking (~250 rpm), add 2 mL of
0.125 M CaCl2 solution. Shake the inoculated container for a
further 5 min (see Note 16; Fig. 4). Incubate under static
conditions at 30°C for 72 h.
2. For scale-up preparation: Inoculate 16 mL modified HS
broth medium with 2 mL primary inoculum. As above, place
the inoculated container on a platform shaker and under vig-
orous shaking (~250 rpm), add 2 mL of 0.125 M CaCl2 solu-
tion. Shake the inoculated container for a further 5 min and
incubate under static conditions at 30°C for 72 h.
3. Harvesting of pectin composites: After incubation, remove
the composite pellicle by carefully pouring out the pellicle

Fig. 4. Preparation of bacterial cellulose–pectin composites. The inoculated container is placed on a platform shaker and
CaCl2 solution (to give a final concentration of typically 12.5 mM) is added under vigorous shaking (a), followed by a
further 5 min of shaking (b), to allow the HS-pectin medium to gel in a uniform manner. A uniform pre-formed gel results
in a relatively homogeneous composite.
204 Mikkelsen and Gidley

into a gloved hand, while letting the excess medium drain

through the fingers (like separating the yolk from the white of
an egg) (see Note 17). Place the pectin composite in a sterile
glass beaker containing excess ice-cold sterile CaCl2 solution
of the same concentration as the incubation medium (e.g.
12.5 mM), and wash at room temperature by gentle agitation
(50 rpm) (see Note 18).
4. Short-term preservation: Pectin composites are stored in the
hydrated state in 0.02% (w/v) sodium azide solution containing
the appropriate concentration of CaCl2, at 4°C (see Note 19).

4. Notes

1. Ga. xylinus ATCC 53524 is the strain of choice in our labora-

tory, as it produces cellulose that is chemically pure and highly
crystalline (14). In particular, this strain does not produce the
water-soluble polysaccharide acetan, a heteropolymer con-
taining glucose, mannose, glucuronic acid and rhamnose in a
molar ratio of 4:1:1:1 that is characteristic of other
Gluconacetobacter strains (15, 16).
2. The medium detailed in this chapter is typically used in our
laboratory. Alternatively, glucose can be substituted with
other carbon sources such as mannitol or glycerol (14), as
well as deuterated carbon sources including d-glucose and
d-glycerol (unpublished work).

3. The hemicelluloses routinely used in our laboratory include

arabinoxylan (wheat, medium viscosity ~25 cSt), b-glucan
(barley, medium viscosity ~28 cSt), and xyloglucan (tamarind
seed; high viscosity ~6.5 dL/g) (Megazyme International
Ireland Ltd., County Wicklow, Ireland). They are stored at
room temperature in a desiccator.
4. The hemicellulose solution cannot be sterilised by autoclav-
ing or micro-filtration as this may depolymerise or remove
some of the polymer respectively. Thus, in order to prevent
contamination, it is imperative to use sterilised equipment
and diluent when making up the solution. Although it is
common practice to add the polymer slowly to solvent, when
working in a laminar flow cabinet, the air flow makes it diffi-
cult to handle fine powders.
5. Wetting the plant cell wall polysaccharide with 40 mL of
95% (v/v) ethanol is recommended by the manufacturer.
However, this is not appropriate as Ga. xylinus is able to
utilise ethanol as a carbon source (19). Subsequently, this
results in a significant (P <0.05) increase in bacterial cell
Formation of Cellulose-Based Composites with Hemicelluloses and Pectins 205

Table 1
The effect of ethanol on the growth of Ga. xylinus ATCC 53524

Increase in cell growth Cellulose yield

HS medium (log10CFU mL−1)a Final pHb at 48 h (g)

With EtOH 4.58 3.23 0.0353

Without EtOH 2.85 5.01 0.0356
a
Values are the difference between cell growth at t = 0 h and t = 48 h. All values are
presented as mean colony counts
b
Initial pH = 5.0. Results are presented as means of triplicates

numbers, but not in the rate of cellulose production

(Table 1). By omitting substrate wetting with alcohol, the
approximate solubilisation times for arabinoxylan, b-glucan,
and xyloglucan are 30–60 min.
6. Arabinoxylan solutions typically have a slight off-white opal-
escent appearance, while xyloglucan and b-glucan solutions
are sometimes very slightly turbid. This may be due to the
presence of trace amounts of protein (Megazyme product
datasheet).
7. The pectins routinely used in our laboratory include com-
mercially produced citrus extracted pectins of degree of
methyl esterification (DM) 30–35 and 60–65. All pectins are
stored at −20°C as the DM is stable for up to a year. However,
whether the pectin is stored at room temperature or at −20°C,
it is recommended that when preparing the pectin solution,
the DM is routinely determined; this involves using titrimetry
(17, 18). Briefly, aliquots (1–5 mL) of a 1% pectin solution
are titrated to between pH 7 and 8 with 0.02 M NaOH and
the titre is recorded. Thereafter, 1 mL 0.5 M HCl solution is
added and the solution is titrated again to pH 7–8. Blank
values, obtained by substituting pectin with water, are sub-
tracted from the titre of pectin. The DM can be calculated
directly from the titres using the following equation:
DM − 100 × V s / V t
where Vs is the hydrolysed (or saponification) titre in millil-
itres, and Vt is the total titre (sum of initial titre plus hydroly-
sed titre) in millilitres (17, 18).
8. We choose to stir the pectin solution overnight as this ensures
complete dissolution. Due to this overnight process, and the
fact that the pectin solution cannot be sterilised by autoclav-
ing or means of filtration, contamination is minimised/pre-
vented by using sterilised equipment and diluent when making
up the solution. Pectins are heat labile, so no heating is used
during the dissolution process.
206 Mikkelsen and Gidley

9. The size of culture container depends on the desired size of

the composite pellicle. Sterile yellow lid specimen containers
(70 mL capacity; 40 mm diameter), standard size (92 × 16 mm)
sterile Petri dishes, and large size (150 × 20 mm) sterile Petri
dishes are commonly used in our laboratory. If yellow lid
specimen containers are used, care must be taken to ensure
that the lid is screwed on only very loosely. As this organism
is an obligate aerobe (16), this allows for adequate aeration of
the culture medium and subsequently pellicle formation is
not negatively impacted.
10. Great care must be taken when preparing 0.02% (w/v)
sodium azide solution. Appropriate personal protective equip-
ment (specified in the Material Safety Data Sheet) must be
worn when weighing out the powder. Sterilised deionised
water may be used for the preparation of the solution to be
used for the hemicellulose composites. For pectin compos-
ites, sterile CaCl2 solution of the appropriate concentration
(e.g. 12.5 mM for DM 30 pectin composites) should be used
when preparing the 0.02% (w/v) sodium azide solution.
Once sodium azide is added, under no circumstances should
these solutions be sterilised by autoclaving.
11. The culture from the revival plate is only sub-cultured once,
and cultures over 1 month old are discarded. This approach is
adopted to not only avoid spontaneous mutation of the bac-
terium, but also to consistently use healthy, viable cultures.
12. Microbank™ is a cryopreservation system commercially avail-
able from Pro-Lab Diagnostics (Ontario, Canada). Each vial
contains a cryopreservative solution and approximately 25
porous beads which serve as carriers to support microorgan-
isms. Using this long-term storage solution, we have success-
fully stored our stock cultures for over 4 years and still have
100% success with revival as well as no contamination.
13. Primary inoculum volumes depend on the desired scale-up vol-
ume. Always prepare sufficient primary inoculum to ensure that
exactly 10% (v/v) primary inoculum is used during scale-up.
14. Washing of composite pellicle after harvesting is carried out
to remove excess medium, polysaccharides non-specifically
trapped within the cellulose mat, and bacterial cells. This pro-
cess is carried out until the pellicle changes from off-white to
white colour. This is typically achieved by carrying out at least
six washes (three times 30 min, followed by three times
10 min washes).
15. It is our experience that hemicellulose/cellulose composites
are stable for up to approximately 2 months when stored at
4°C. After this period, degradation is observed with the edges
of the otherwise opaque pellicle becoming translucent – this
 Formation of Cellulose-Based Composites with Hemicelluloses and Pectins 207

observation is prominent in pellicles that have been stored at

4°C for 6 months.
16. The inoculated container is placed on a platform shaker and
CaCl2 solution is added under vigorous shaking, followed by a
further 5 min of shaking, as this allows the HS-pectin medium
to subsequently gel under quiescent conditions in a uniform
manner. Uniformity of the preformed gel is important for the
formation of a relatively homogeneous composite.
17. Avoid using forceps when harvesting the bacterial cellulose–
pectin composites, as the pellicles can be fragile and tear eas-
ily. Moreover, due to their highly extensible or “stretchy”
nature, the use of forceps can impact on their shape. This is
important to note when handing the samples for tensile
stress/strain measurements, and also applies to other “weak”
composites such as those produced after limited incubation
times (e.g. 24 h).
18. Note that the purification of pectin composites by washing
with CaCl2 solution causes the pellicle to shrink slightly (see
Fig. 5).
19. In order to ensure good storage conditions for pectin compos-
ites, it is critical to use 0.02% (w/v) sodium azide solution con-
taining the appropriate concentration of CaCl2. If sodium azide
solution made up with deionised water is used, composites are
stable for up to approximately 2 months when stored at 4°C.
After this period, pellicle degradation is observed, with the
sodium azide solution becoming viscous and the pellicle loosing
some of its characteristic “lumpy” texture (due to precipitation
of pectin from the composite) – this observation is prominent
in pellicles that have been stored at 4°C for 6 months.

Fig. 5. Bacterial cellulose–pectin composites before (a) and after (b) washing with CaCl2 solution (typically 12.5 mM),
demonstrating typical shrinkage from 4.2 cm in width to 3.8 cm after washing the pellicle.
208 Mikkelsen and Gidley

References
1. Brown, R. M. (1989) Bacterial cellulose. In: 10. Zykwinska, A. W., Ralet, M. C. J., Garnier, C.
Cellulose: structural and functional aspects, D., and Thibault, J. F. J. (2005) Evidence for
(Phillips, G. O., Kennedy, J. F., and Williams, in vitro binding of pectin side chains to cellu-
P. A., eds.), Ellis Horwood Ltd, New York, lose. Plant Physiol 139, 397–407.
pp. 145–151. 11. Chanliaud, E., and Gidley, M. J. (1999) In
2. Whitney, S. E. C., Brigham, J. E., Darke, A., vitro synthesis and properties of
Reid, J. S. G., and Gidley, M. J. (1995) In pectin/Acetobacter xylinus cellulose compos-
vitro assembly of cellulose/xyloglucan net- ites. Plant J 20, 25–35.
works: ultrastructural and molecular aspects. 12. Hestrin, S., and Schramm, M. (1954)
Plant J 8, 491–504. Synthesis of cellulose by Acetobacter xylinum.
3. McCann, M. C., Wells, B., and Roberts, K. 2. Preparation of freeze-dried cells capable of
(1990) Direct visualization of cross-links in polymerizing glucose to cellulose. Biochem
the primary plant-cell wall. J Cell Sci 96, J 58, 345–352.
323–334. 13. Willey, J. M., Sherwood, L. M., and Woolverton,
4. Whitney, S. E. C., Gidley, M. J., and McQueen- C. J. (2008) Prescott, Harley, and Klein’s micro-
Mason, S. J. (2000) Probing expansin action biology, McGraw-Hill, New York, p. 115.
using cellulose/hemicellulose composites. 14. Mikkelsen, D., Flanagan, B. M., Dykes, G. A.,
Plant J 22, 327–334. and Gidley, M. J. (2009) Influence of different
5. Chanliaud, E., DeSilva, J., Strongitharm, B., carbon sources on bacterial cellulose produc-
Jeronimidis, G., and Gidley, M. J. (2004) tion by Gluconacetobacter xylinus strain ATCC
Mechanical effects of plant cell wall enzymes 53524. J Appl Microbiol 107, 576–583.
on cellulose/xyloglucan composites. Plant 15. Couso, R. O., Ielpi, L., and Dankert, M. A.
J 38, 27–37. (1987) A xanthan-gum-like polysaccharide
6. Whitney, S. E. C., Brigham, J. E., Darke, A., from Acetobacter xylinum. J Gen Microbiol
Reid, J. S. G., and Gidley, M. J. (1998) 133, 2123–2135.
Structural aspects of the interaction of man- 16. Kersters, K., Lisdiyanti, P., Komagata, K., and
nan-based polysaccharides with bacterial cel- Swings, J. (2006) The family Acetobaceraceae:
lulose. Carbohydr Res 307, 299–309. the genera Acetobacter, Acidomonas, Asaia,
7. Chanliaud, E., Burrows, K. M., Jeronimidis, Gluconacetobacter, Gluconobacter and
G., and Gidley, M. J. (2002) Mechanical Kozakia. In: The prokaryotes: an evolving elec-
properties of primary plant cell wall analogues. tronic resource for the microbiological commu-
Planta 215, 989–996. nity, (Dwokin, M., ed.), Springer, New York,
8. McKenna, B. A., Mikkelsen, D., Wehr, J. B., pp. 163–200.
Gidley, M. J., and Menzies, N. W. (2009) 17. Marga, F., Morvan, C., and Morvan, H.
Mechanical and structural properties of native (1995) Pectins in normal and vitreous apple
and alkali-treated bacterial cellulose produced microplants cultured in liquid-medium. Plant
by Gluconacetobacter xylinus strain ATCC Physiol Biochem 33, 81–86.
53524. Cellulose 16, 1047–1055. 18. Wehr, J. B., Menzies, N. W., and Blamey, F. P.
9. Whitney, S. E. C., Gothard, M. G. E., C. (2004) Alkali hydroxide-induced gelation
Mitchell, J. T., and Gidley, M. J. (1999) Roles of pectin. Food Hydrocolloid 18, 375–378.
of cellulose and xyloglucan in determining the 19. Ross, P., Mayer, R., and Benziman, M. (1991)
mechanical properties of plant cell walls. Plant Cellulose biosynthesis and function in bacteria.
Physiol 121, 657–663. Microbiol Rev 55, 35–58.
 Chapter 15

Structural Proteins of the Primary Cell Wall: Extraction,

Purification, and Analysis
Derek T.A. Lamport, Li Tan, and Marcia J. Kieliszewski

Abstract
Structural proteins of the primary cell wall present unusual but interesting problems for structural biologists
in particular and plant biologists in general. As structure is the key to function; then the biochemical
isolation of these glycoproteins for further study is paramount. Here, we detail the “classical” method for
isolating soluble extensin monomers by elution of monomeric precursors to network extensin from tissue
cultures. We also outline an additional approach involving genetic engineering that can potentially yield
the complete genomic range of extensins and other hydroxyproline-rich glycoprotein (HRGPs) currently
underutilized for biotechnology.

Key words: Extensin, Primary cell wall, HRGPs, AGPs, Cultured cells

1. Introduction

The year 2010 marks the fiftieth birthday of the cell wall protein
extensin (1), a hydroxyproline-rich glycoprotein (HRGP), argu-
ably the planet’s most abundant. Belatedly recognized as the third
network of the primary cell wall (2) after cellulose (Anselme Payen
in 1838) and pectin (Henri Bracconot in 1825), extensins play an
essential role in cytokinesis as self-assembling amphiphiles that
template new cross wall deposition presumably as extensin ­pectate.
The discovery of cell wall protein by a graduate student working
with D.H. Northcote in the Department of Biochemistry at
Cambridge, was fortuitously in a lab adjacent to Fred Sanger, 1958
Nobel laureate and ­architect-in-chief of modern sequence analysis.
The inspirational Sanger school, the world centre of protein chem-
istry at that time, generated methods and advice that “topped up”
a pre-eminent undergraduate course of practical biochemistry
with classical methods still relevant – a judicious mix of enzymic

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_15, © Springer Science+Business Media, LLC 2011

209
210 Lamport, Tan, and Kieliszewski

degradation and ­chemical methodology ranging from the Sanger

reagent to partial and complete acid/base hydrolysis – precursor
to HF-solvolysis for the deglycosylation of glycoproteins (3).
There was just one problem – the Hyp-rich wall-bound
(crosslinked) protein resisted solubilization as an intact protein
(4); this stymied initial attempts to deploy classical methods of
protein characterization. For example, heat denaturation melts
the triple helix of Hyp-rich collagen to yield gelatin but did not
solubilize the cell wall protein. Miniscule amounts of soluble pro-
tein precursors ionically bound to the wall of sycamore or carrot
cells were insufficient for thorough analysis until an improved
experimental system – tomato cell suspension cultures – yielded
sufficient material. Subsequent chromatographic purification,
peptide mapping, and sequence analysis of extensin monomers
demonstrated the exceptional peptide periodicity of extensins (5)
and their diagnostic motifs, Ser-Hyp4 (6) and intramolecular isodi-
tyrosine (Idt) (7) amply confirmed by later genomic analysis.
Some members of the extensin superfamily, typically arabi-
nogalactan proteins (AGPs) although transiently GPI-anchored are
not covalently crosslinked and therefore are readily solubilized by
mechanical cell disruption (8). However, AGPs resist purification
because extensive glycosylation by acidic arabinogalactan polysac-
charides swamps properties of the protein. HF-solvolysis at 4°C
surmounts this problem by selective glycosidic bond cleavage
with peptide bonds remaining intact; deglycosylation of an AGP
enables purification of its polypeptide backbone. Thus, salt-elution
of extensin from intact cells and HF-deglycosylation of both
extensins and AGPs are effective methods for isolating the poly-
peptides when present in appreciable amounts.
A more recent general method involves genetic engineering
and yields the intact glycoproteins: overexpressing the gene of
choice attached to a green fluorescent protein tag enables visual
selection of transformed cells with generally high yields of
the native glycoprotein. Crucially, the hydrophobic GFP adduct
facilitates clean chromatographic separation of hydrophilic AGPs
compared with other methods that can yield ambiguous or
misleading results (9, 10).

2. Materials

2.1. Elution of Extensin 1. Cell culture of plant of interest (see Notes 1 and 2).
Monomers from Intact 2. Ice.
Cells
3. Vacuum (see Note 3).
4. Coarse-sintered funnel.
5. 50 mM CaCl2.
 Structural Proteins of the Primary Cell Wall: Extraction, Purification, and Analysis 211

6. Freeze-drier.
7. 5% Trichloroacetic acid (TCA).
8. Ice.
9. Bench centrifuge.
10. Centrifuge tubes.
11. Distilled water.
12. 30 mM Sodium phosphate buffer, pH 7.6.
13. Biorex 70 cation exchange column: 90× 1.5 cm.
14. Monitor suitable for detecting wavelength of column eluent.
15. Column reservoir buffer: 30 mM Sodium phosphate buffer
pH 6.1, 1 M NaCl.
16. Dialysis tubing.
17. Dialysis clips.

2.2. Isolation 1. Cell cultures (see Notes 1 and 2).

of Periplasmic and 2. Ice.
Extracellular AGPs
3. Coarse-sintered funnel.
4. Vaccum apparatus.
5. 1% NaCl.
6. Sonicator.
7. TCA.
8. Bench centrifuge.
9. Dialysis tubing.
10. Dialysis clips.
11. 2% NaCl.
12. b-d-Glucosyl Yariv reagent.
13. Sodium dithionite.
14. Screw-cap vial.
15. Centrifuge tubes.
16. Block heater or oven at 45°C.

2.3. Transformation 1. Agrobacterium competent tobacco BY-2 cells.

of Tobacco BY-2 Cells 2. pBI plasmid containing constructed gene cassettes.
3. Eppendorf tubes.
4. Liquid nitrogen.
5. 37°C incubator.
6. LB medium: 10 g tryptone, 5 g yeast extract, 5 g NaCl,
1 mL/L 1 N NaOH.
7. Incubated shaker at 28°C.
212 Lamport, Tan, and Kieliszewski

8. LB plates containing 30 mg/mL kanamycin.

9. Petri dishes.
10. Centrifuge.
11. LB culture medium containing 100 mg/L kanamycin,
400 mg/L timentin and solidified with 0.8–1.5% w/v agar.
12. LB culture medium containing 100 mg/mL kanamycin.

2.4. Isolation of Fusion 1. Transgenic cultures prepared as described in Subheading 3.5:

Proteins 12–18 day-old.
2. Rotoevaporator.
3. Dialysis tubing.
4. Filter paper.
5. Funnel.
6. Solid NaCl.
7. Centrifuge.
8. Pre-treated phenyl sepharose hydrophobic interaction chro-
matography column: Pretreat the column by washing with
distilled water and equilibrate with 2 M NaCl.
9. 2 M NaCl.
10. Fluorescence detector (excitation 488 nm, emission
520 nm).
11. Hamilton PrP-1 reverse-phase chromatography system.
12. 80% v/v Acetonitrile in 0.1% v/v TFA.
13. Double distilled water.

3. Methods

3.1. Elution of Extensin The primary cell wall is largely pectin that is highly methyl-
Monomers from Intact esterified and therefore weak acidic macromolecule binds to weak
Cells (11) basic extensin monomers as extensin pectate. Thus, extensins can
be eluted from acidic pectin because it behaves as a cation exchange
resin. Like other highly glycosylated glycoproteins extensin is sol-
uble in TCA. Low yields may result from crosslinkage by extensin
peroxidase in the presence of trace amounts of hydrogen peroxide
derived from reactive oxygen species (ROS) like superoxide
e.g. ROS → hydrogen peroxide → crosslinked extensin.
1. Grow appropriate species in shake culture (12) (see Notes 1
and 2).
2. Harvest after 4–8 days in rapid growth phase.
3. Cool cells on ice.
 Structural Proteins of the Primary Cell Wall: Extraction, Purification, and Analysis 213

4. Suction filter 50–500 g fresh weight cells gently on a

coarse-sinter funnel.
5. Wash briefly with cold water.
6. Elute cells with 2× volume of cold 50 mM CaCl2.
7. Freeze-dry eluate.
8. Redissolve eluate in ice-cold 5% TCA and stand for 1 h
(see Note 4).
9. Centrifuge at 9,000 rpm for 60 min and discard precipitate.
10. Dialyze supernate against distilled water at 4°C.
11. Freeze-dry.
12. Re-dissolve in 30 mM sodium phosphate buffer, pH 7.6.
13. Load onto a 90× 1.5 cm BioRex 70 cation exchange
column.
14. Elute from the column at a flow rate of 60 mL/h with a gradient
produced by having 300 mL of 30 mM sodium phosphate
buffer, pH 7.6 in the mixing chamber and 300 mL of 30 mM
sodium phosphate buffer, pH 6.1 containing 1 M NaCl in
the reservoir.
15. Monitor the eluent at 200 nm and collect appropriate fractions.
16. Dialyze, freeze-dry and store at −20°C.
Although network extensin per se cannot be isolated, peptide
fragments can be obtained, analyzed, and identified by chymotryp-
tic/tryptic degradation of isolated cell walls after partial removal
of polysaccharides by enzymic (4) or chemical degradation (13)
or removal of all polysaccharides by HF-solvolysis (14).

3.2. Isolation of In growing cells, phospholipase C continuously releases GPI-

Periplasmic and anchored AGPs that cover the outer surface of the plasma mem-
Extracellular AGPs brane. Following addition to the wall by apposition, these soluble
from Cultured Cells (8) AGPs trapped by the limiting porosity of the pectic matrix do not
diffuse freely but migrate through the expanding wall by “plug
flow” extrusion into the culture medium as soluble AGPs. Intact
cells typically contain a total of ~600 mg AGP/g fw of which
roughly half is soluble periplasmic AGP that is released by cell
disruption. AGPs in muro are trapped in the pectic matrix and
released by appropriate disruption of this matrix e.g. by boiling or
pectinase treatment. The method for isolation of periplasmic
AGPs is given below. AGPs from the culture medium filtrate can
be prepared in the same way except that the amount of extracel-
lular AGP in the medium of a friable culture is generally similar to
the soluble periplasmic AGP content of its intact cells.
1. Grow appropriate species in shake culture.
2. Harvest after 4–8 days in rapid growth phase.
214 Lamport, Tan, and Kieliszewski

3. Cool cells on ice.

4. Suction filter 1–100 g fresh weight cells on a coarse-sinter
funnel.
5. Wash briefly with 1% NaCl.
6. Disrupt cells by sonication in an ice bath.
7. Add TCA to final concentration of 5%.
8. Centrifuge at 10,000 rpm.
9. Discard the TCA precipitate.
10. Dialyze supernate exhaustively.
11. Freeze-dry.
12. Dissolve in a small volume of 2% aqueous NaCl.
13. Add excess (see Note 5) b-d-glycosyl Yariv reagent.
14. After 1 h at RT collect AGP-Yariv precipitate by low speed
centrifugation.
15. Add excess sodium dithionite in a small screw cap vial.
16. Fill tube to minimize oxidation of dithionite to yield elemental
sulphur.
17. Heat at 45°C till colour fades to disrupt the AGP-Yariv
complex.
18. Dialyze and freeze-dry the purified AGPs.

3.3. Design of Genes The repetitive peptide periodicity of extensins allows the use of a
for Expression and head-to-tail method (15, 16) or a modified polymerization
Detection of HRGP- method (17), to polymerize single repeats of oligonucleotides
GFP Chimeras encoding an extensin repeat into large repetitive genes. The head-
to-tail method involves polymerization of three pairs of partially
overlapping oligonucleotides. However, it is difficult to control the
degree of polymerization (Fig. 1). The longest gene synthesized
was 600 bp encoding (ThrPro)100 (16). The modified polymeriza-
tion method (17) can be used to synthesize genes of specific sizes;
e.g. the synthetic rsh gene encoding 13 repeats of a 28-amino-acid
peptide (Tan and Kieliszewski, unpublished data) (Fig. 2).
Four great advantages of making chimeras of extensin (or
AGPs) with the enhanced green fluorescent protein (GFP) are;
(1) GFP fluorescence enables facile identification of transformed
BY-2 cells, (2) GFP fluorescence indicates the highest yielding
cell lines, (3) GFP hydrophobicity facilitates chromatographic
separation of hydrophilic HRGPs, and (4) GFP fluorescence sim-
plifies the monitoring of column eluates.
1. Design gene using appropriate software, such as Primer
Premier (Premier Biosoft International).
2. Choose codons favoured by plants (18). Clone the synthetic
gene into the plasmid pUC18-SStob-EGFP vector (15) (Fig. 3)
Structural Proteins of the Primary Cell Wall: Extraction, Purification, and Analysis 215

Fig. 1. Flow chart of the head-to-tail polymerization method. In brief, the annealed
5¢-linker oligonucleotide set is mixed with an excess of the annealed internal repeats
oligonucleotide set which encode part of the target gene (e.g. 5 to 1 molar ratio of inter-
nal repeats to 5¢-linker oligonucleotide sets). The pool formed of 5¢-linker-internal
repeats of different internal repeats is then capped with 3¢-linker. The different sized
gene pool is then ligated by T4-DNA ligase. The restriction sites on both 5¢- and 3¢-end
linkers allow further cloning of the synthetic genes into pUC18-SStob-EGFP vector.

which contains the following restriction sites: BamHI–

(SStob)–XmaI–NcoI–(GFP)–BsrGI–SacI.
SStob represents an extensin signal sequence gene, while GFP
is a gene encoding the enhanced green fluorescent protein.
3. Ligate the synthetic gene:
a. I nto the pUC18-SStob-EGFP vector as a XmaI-NcoI frag-
ment to form pUC18-SStob-target gene plasmid (15, 17)
or
b. Behind the EGFP gene as a BsrGI-SacI fragment to form
pUC18-SStob-EGFP-target gene plasmid (19).

4. Subclone the resulting gene cassette, SStob-target gene-EGFP

or SStob-EGF target gene into the plant transformation ­vector,
pBI121, as a BamHI-SacI fragment replacing the b-glucuronidase
reporter gene. The gene cassette is then under regulation of a
CaMV S35 promoter for overexpression and contains the
neomycin phosphotransferase (NPT) II gene that confers
resistance to kanamycin.
216 Lamport, Tan, and Kieliszewski

c
Plasmid with one repeat of target gene

Digestion 1 Digestion 2
a a

b b

c c
DNA fragment 1 DNA fragment 2

Ligation

c
Plasmid with two repeats of target gene

Fig. 2. Flow chart of a controlled polymerization method. A plasmid is digested with two
sets of restriction enzymes, e.g. a/c and b/c as showed. This method requires that
restriction enzymes a and b should yield the same sticky ends. After digestion, the two
fragments each with a single repeat of the target gene are purified via agarose gel
electrophoresis, and ligated with T4-DNA ligase to yield two repeats of the target gene
with reconstituted restriction sites between them. These procedures can be repeated to
build desired target genes of defined sizes.

EGFP

sstob
pUC18-SS tob-EGFP

Amp r

Fig. 3. A plasmid map of pUC18-SStob-EGFP. SStob represents an extensin signal sequence

gene, while GFP encodes the enhanced green fluorescent protein. The SStob gene is
flanked by a BamHI restriction site and a XmaI restriction site, with XmaI site located at
its 3¢-end. Downstream of the XmaI site, an NcoI restriction site (at the 5¢-end) and a
BsrGI restriction site (at the 3¢-end) sandwich the GFP gene. A SacI restriction site behind
the BsrGI site, together with a BamHI site, allow further cloning of the gene cassette into
the pBI121 vector.
 Structural Proteins of the Primary Cell Wall: Extraction, Purification, and Analysis 217

3.4. Transformation Transform Agrobacterium tumefaciens with the pBI plasmid

of Tobacco BY-2 Cells constructs via the freeze-thaw method (20) as follows:
and Selection of Cell
1. Add ~1 mg of pBI plasmid containing the constructed gene
Lines
cassettes to 100 mL Agrobacterium competent cells, prepared
by a method similar to E. coli (20).
2. Freeze Agrobacterium cells in a sterile Eppendorf tube in
liquid nitrogen.
3. Transfer tube of frozen Agrobacterium cells to 37°C for
5 min.
4. Add 1 mL LB medium.
5. Incubate on shaker at 28°C for 3–4 h.
6. Spread Agrobacterium on LB plates containing 30 mg/mL
kanamycin.
7. Incubate plates at 28°C for 2–3 days.
8. The colonies contain transformed Agrobacterium which can
be used to transform tobacco BY-2 cells.
9. Transfer 5 mL tobacco BY-2 cells (4-days old) to 10 cm
Petri dish.
10. Add 200 mL from transformed Agrobacterium cultured
overnight on LB.
11. Co-culture for 2–3 days at 28°C in the dark.
12. Wash the infected cells with liquid culture medium three
times by low speed centrifugation (~500 × g).
13. Spread washed cells on solid culture medium containing
kanamycin and timentin.
14. Incubate for 4–5 weeks.
15. Identify transformed cells.
16. Subculture BY-2 cells into liquid culture medium containing
100 mg/mL kanamycin.
17. Shake cultures at 92 rpm at room temperature for protein
expression.

3.5. Isolation of Fusion 1. Filter culture medium from 12- to 18-day transgenic cultures.
Proteins from the 2. Concentrate filtrate via rotoevaporation at 28°C to 10% of
Culture Medium initial volume.
3. Dialyze against ddH2O for 2 days.
4. Add solid NaCl to a final concentration of 2 M.
5. Spin at 10,000 × g for 30 min.
6. Remove and discard any pellet.
7. Load clarified supernatant onto a phenyl sepharose hydro-
phobic interaction chromatography column, previously
washed with dH2O and equilibrated with 2 M NaCl.
218 Lamport, Tan, and Kieliszewski

8. Elute the column with a linear NaCl gradient decreasing from

2 to 0 M.
9. Monitor column via fluorometry (Excitation 488 nm; emission
520 nm).
10. Fluorescent fractions contain EGFP fusion proteins of ~85%
homogeneity.
11. Further purify by reversed phase chromatography on
Hamilton PRP-1 with a gradient of 0–70% of 80% acetonitrile
in 0.1% TFA in 100 min. The EGFP fusion glycoproteins
elute at ~35% acetonitrile in 0.1% TFA (21).

4. Notes

1. Expect a wide variation in yields between cultures. Yield is

typically in the range of 5–500 mg/g fresh weight of cells.
2. Ideally cultures should be friable. Friability maximizes release
of AGPs from an expanding wall. Cultures that grow as tight
micro-calli release much less as they expose a much smaller
proportion of the cell surface to the growth medium.
3. Vacuum can be provided either by using a vacuum pump at a
low setting or by suction provided using a system attached to
a tap with running water. The latter is easier to control.
4. Minimal time in TCA avoids degradation of acid-labile
arabinofuranosides.
5. Greater than 1 mg b-d-glycosyl Yariv reagent per mg of AGP.

References

1. Lamport, D. T. A., and Northcote, D. H. 6. Lamport, D. T. A., Katona, L., and Roerig, S.
(1960) Hydroxyproline in primary cell walls (1973) Galactosyl serine in extensin. Biochem J
of higher plants. Nature 188, 665–666. 133, 125–131.
2. Kerr, T., and Bailey, I. W. (1934) The cam- 7. Epstein, L., and Lamport, D. T. A. (1984) An
bium and its derivative tissues. X Structure, intracellular linkage involving isodityrosine in
optical properties and chemical composition extensin. Phytochemistry 23, 1241–1246.
of the so-called middle lamella. J Arnold Arbor 8. Lamport, D. T. A., Kieliszewski, M. J., and
15, 327–349. Showalter, A. M. (2006) Salt-stress upregu-
3. Mort, A. J., and Lamport, D. T. A. (1977) lates periplasmic arabinogalactans-proteins:
Anhydrous hydrogen fluoride deglycosy- using salt-stress to analyse AGP function. New
lates glycoproteins. Anal Biochem 82, Phytol 169, 479–492.
289–309. 9. Baldwin, T. C., McCann, M. C., and Roberts,
4. Lamport, D. T. A. (1965) The protein com- K. (1993) A novel hydroxyproline-deficient
ponent of primary cell walls. Adv Bot Res 2, arabinogalactan protein secreted by suspension-
151–218. cultured cells of Daucus carota. Plant Physiol
5. Smith, J. J., Muldoon, E. P., Willard, J. J., and 103, 115–123.
Lamport, D. T. A. (1986) Tomato extensin 10. Baldwin, T. C., Domingo, C., Schindler, T.,
precursors P1 and P2 are highly periodic Seetharaman, G., Stacey, N., and Roberts, K.
structures. Phytochemistry 25, 1021–1030. (2001) DcAGP1, a secreted arabinogalactans
 Structural Proteins of the Primary Cell Wall: Extraction, Purification, and Analysis 219

protein, is related to a family of basic arabinogalactans addition to arabinogalactan-

­proline-rich proteins. Plant Mol Biol 45, proteins. Plant Physiol 132, 1362–1369.
421–435. 17. Held, M. A., Tan, L., Kamyab, A., Hare, M.,
11. Smith, J. J., Muldoon, E. P., and Lamport, Shpak, E., and Kieliszewski, M. J. (2004)
D. T. A. (1984) Isolation of extensin pre- Di-isodityrosine is the intermolecular cross-
cursors by direct elution of intact tomato link of isodityrosine-rich extensin analogs
cell suspension cultures. Phytochemistry 23, cross-linked in vitro. J Biol Chem 279,
1233–1239. 55474–55482.
12. Lamport, D. T. A. (1964) Cell suspension 18. Li, S., and Showalter, A. M. (1996) Cloning
cultures of higher plants, isolation and growth and developmental/stress regulated expression
energetics. Exp Cell Res 33, 195–206. of a gene encoding a tomato arabinogalactan
13. Lamport, D. T. A. (1969) The isolation and protein. Plant Mol Biol 32, 641–652.
partial characterization of hydroxyproline-rich 19. Zhao, Z. D., Tan, L., Showalter, A. M.,
glycopeptides obtained by enzymic degrada- Lamport, D. T. A., and Kieliszewski, M. J.
tion of primary cell walls. Biochemistry 8, (2002) Tomato LeAGP-1 arabinogalactan-
1155–1163. protein purified from transgenic tobacco cor-
14. Frueauf, J. B., Dolata, M., Leykam, J. F., roborates the Hyp contiguity hypothesis.
Lloyd, E. A., Gonzales, M., VandenBosch, K., Plant J 31, 431–444.
and Kieliszewski, M. J. (2000) Peptides iso- 20. An, G., Ebert, P. R., Mitra, A., and Ha, S. B.
lated from cell walls of Medicago trunculata (1988) Binary vectors. Plant Molecular
nodules and uninfected root. Phytochemistry Biology Manual. Gelvin, S.B. and Schilperoort,
55, 429–438. R.A. (eds). Dordrecht, Netherlands: Martinus
15. Shpak, E., Leykam, J. F., and Kieliszewski, M. J. Nijhoff, p. 1–19.
(1999) Synthetic genes for glycoprotein 21. Xu, J., Tan, L., Goodrum, K. J., and
design and the elucidation of hydroxyproline- Kieliszewski, M. J. (2007) High-yields and
O-glycosylation codes. Proc Natl Acad Sci U S A extended serum half-life of human interferon
96, 14736–14741. alpha 2b expressed in tobacco cells as arabi-
16. Tan, L., Leykam, J. F., and Kieliszewski, M. J. nogalactan-protein fusions. Biotechnol Bioeng
(2003) Glycosylation motifs that direct 97, 997–1008.
 Chapter 16

New Insights into the Control of Cell Growth

Claudia Blaukopf, Matthäus Z. Krol, and Georg J. Seifert

Abstract
Undoubtedly, the function of the plant cell wall in the control of cell growth far exceeds its mechanical
role. The plant’s monitoring of cell wall function and integrity comprises a central checkpoint to inte-
grate cues for survival and division, expansion and differentiation, as well as fluctuations in the biotic and
abiotic environment (Somerville et al., Science 306:2206–2211, 2004). With their biochemical nature
yet unknown, the identification of molecular constituents of cell wall performance, and integrity control
initially depends on a combination of genetic and physiological approaches.

Key words: Cell wall integrity control, Forward genetics, Map based cloning, Programmed cell
death, Yariv, Arabinogalactan proteins

1. Introduction

In the recent past, there have been significant advances in the

analysis of structural components of the plant cell wall and like-
wise, the cell wall biosynthetic machinery is being elucidated at an
ever increasing pace, largely by using methods presented in this
volume. However, conceptually, the molecular components of
the hypothetical cell wall performance and integrity control sys-
tem are not even laid out yet. Some serendipitous insights into
potential ingredients have been derived from genetic screens for
constitutive activation of stress response or sugar signaling. The
finding that elevated expression of several stress-induced genes is
triggered by defects in cellulose biosynthesis, identified jasmonate
(JA) and ethylene signaling as potential pathways involved in cell
wall integrity control. Intriguingly, the combination of JA and eth-
ylene insensitivity partially suppressed the reduction of hypocotyl
elongation seen in the cesA3/cev1 cellulose synthase mutants (2).

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_16, © Springer Science+Business Media, LLC 2011

221
222 Blaukopf, Krol, and Seifert

Another example for the fortuitous identification of a component

of cell wall integrity control was the demonstration that the cell
wall matrix compositional mutants mur4, mur1 and mur3 (3) all
display PRL1 dependent glucose hypersensitivity (4). As the path-
ways that signal cell wall stress, sugar response and growth regula-
tor stimuli are thought to interact with many other regulatory
pathways; such pioneering findings encourage a more systematic
search for components of cell wall performance and integrity con-
trol. As it can be assumed that there exist receptor-like sensors of
cell wall integrity, a systematic reverse genetic search for receptor
mutants seems an attractive possibility. One recent example was
the identification of the pair of the FEI1 and FEI2 loci that in
combination are required for normal growth at elevated levels of
salt and sugar (5). Further genetic analysis indicated that FEI1
and FEI2 might act in the same pathway that is responsible for
salt oversensitivity in the sos5 mutant that expresses a defective
variant of the fasciclin-like arabinogalactan protein (AGP) 4
(FLA4) in Arabidopsis (6). Although a valuable contribution to
our knowledge of cell wall signaling, this study also highlights an
important limitation of reverse genetics, namely that genetic
redundancy has to be overcome by cumbersome creation of mul-
tiple mutant combinations. A central strategy to identify players
in cell wall performance and integrity control is forward genetics.
On the one hand, there will be new screens for genetic modifiers
of cell wall mutants. On the other hand, several chemicals induce
cell wall defects, and mutants displaying altered sensitivity towards
such drugs can lead to the identification of positive and negative
regulators of cell wall integrity sensing. A rationale behind sensi-
tized screens is the hypothesis that the physiological and morpho-
logical responses to a mutation in cell wall structure such as
inhibited growth, disease resistance, or cell death are the conse-
quence of a hyperactivated cascade of sensory and signal trans-
duction processes that normally probe the function of the wild
type cell wall during growth and stress. So far, there are only few
examples for systematic cell wall mutant modifier screens, the
most notable being the identification of the THESEUS1 (THE1)
locus encoding an active receptor-like kinase that is responsible
for the dwarfed phenotype of cellulose biosynthetic mutants (7)
and was originally identified in a screen for genetic suppressors of
the CesA6 mutant procuste. Although its biochemical mode of
action and potential ligands are unknown, the THE1 receptor-
like kinase comprises the first obvious candidate of a molecular
component of cell wall integrity sensing. Chemically sensitized
screens have previously been performed to isolate mutants insen-
sitive to cellulose synthase inhibitors (8–10) and cytoskeletal
drugs (11, 12). Although the identified genes were either directly
related to cell wall biosynthesis and cytoskeletal assembly or were
postulated to be involved in drug transport, there is a strong
 New Insights into the Control of Cell Growth 223

requirement for more extensive screens to identify components of

cell wall performance and integrity control. The fact that a muta-
tion in the FLA4 locus is responsible for salt oversensitive root
growth (6) highlights the potential role of AGPs in cell wall integ-
rity sensing. Because AGPs are often tethered to glycosylphos-
phoinositol lipid anchors at the plasma membrane and can bind
glycans, AGPs might act as coreceptors presenting cell wall struc-
ture to transmembrane receptor like kinases or membrane chan-
nel type receptors (13). Despite the genetic role of AGPs in cell
survival, cell division, and cell elongation, a direct implication of
plasma membrane AGPs in cell wall integrity control is still specu-
lative. Genetic approaches are needed to establish a link between
the phenotypes of AGP mutants and a cellular sensing and signal
transduction machinery. As a long-term goal, hypothetical recep-
tor complexes have to be reconstituted in recombinant systems
that allow for all the required posttranslational modifications that
are expected to be crucial in the function of AGPs.
In summary, very little is known about the molecular players
in cell wall performance and integrity control, and novel insights
in the control of cell growth can be obtained from well designed
modifier screens as well as chemically sensitized screens. As exam-
ples for screens involving defective AGPs, we describe how we
generated chemically mutagenized populations of the conditional
AGP glycosylation and protein structural mutants uge4-3 (14, 15)
and sos5 (6), respectively, how we screened for enhancer and sup-
pressor mutants including additional selection filters, and how we
are proceeding towards map-based cloning of the mutant loci.
We also describe how to cheaply synthesize gram quantities of the
AGP binding drug b-glucosyl Yariv that interferes with cell growth
(16) and survival (17, 18) and apply it in a genetically amenable
assay.

2. Materials

2.1. EMS Mutagenesis 1. Bulked mutant seeds (uge4-3, sos5).

2. 0.1% KCl.
3. Prepare in chemical hood on day of use: Ethyl methanesul-
fonate (EMS) buffer: 100 mM (290 mL per 20 mL) EMS,
0.1 M Na-phosphate buffer (pH 5), 5% dimethylsulfoxide
(DMSO).
4. Quenching buffer: 0.1 M Na-thiosulfate (0.791 g/50 mL).
5. Solid Na-thiosulfate.
6. Cryosanitized soil (peat:humus:sand 5:5:1; frozen at −80° for
48 h; this treatment in combination with keeping Perspex lids
224 Blaukopf, Krol, and Seifert

on plant trays during early growth helps to control insect and

mite infestations). Soil is thawed over night and distributed
into 6 × 6 cm pots before use.
7. Germination trays with Perspex hoods. “Seed propagator”
bought at local house ware supply store.

2.2. Primary and 1. Sterilization solution: Typically a 1:9 solution of hypochlorite

Secondary Screens containing household bleach. Alternative 3% sodium
of uge4-3 Enhancers hypochlorite and 0.2% sodium dodecyl sulfate (SDS) in
and sos5 Suppressors water.
2. Standard germination medium (SGM): Murashige and
Skoog salts including vitamins and MES-buffer (Duchefa,
M0255.0001), 1% sucrose, 0.8% Phytagel (Sigma-Aldrich,
P8169). Dissolve Murashige and Skoog salts and sucrose and
adjust the pH to 5.8, add Phytagel and autoclave, then bring
to 55°C before pouring into 90 mm Petri dishes.
3. Galactose-selection medium: SGM including 5 mM d-galactose
(prepare 1 M stock solution, sterile filtered, add to autoclaved
SGM prior to plating).
4. SGM including 4% sucrose.
5. High salt medium: SGM including 100 mM NaCl.
6. Seed dispensing pipette. Ultipette Variable Volume 1–10 mL
(Barky Instr. Intl., UK) and plastic tips, Ultipette capillary
tips, 0.5–10 mL (Barky Instr. Intl., UK).
7. Stereomicroscope (e.g., Olympus SZX7 magnification 0.8–
5.6 × 10). When plates are directly placed on the diffuse bright
field illumination screen, contrast of roots is too low for effi-
cient observation and should be increased by lifting dishes
3–5 cm above the diffuse light source on an workshop made
glass table. In other microscopes (Zeiss Stemi 2000-C or Leitz
LM10) the built-in illumination produces sufficient contrast.

2.3. Establishing 1. Suitable parental plants (acceptor and donor plants).

a Genetic Model 2. Sharp forceps.
and a Mapping
3. Labeling tape.
Population
4. Nail scissors (optional).
5. Magnifying goggles or Macromicroscope (optional).
6. High salt medium: SGM (see Subheading 2.2) including
100 mM NaCl.

2.4. Genomic DNA 1. Mortar.

Extraction 2. 1.5 mL Eppendorf tubes.
2.4.1. Protocol 1: Modified 3. 100–1,000 mL pipettes and tips.
from (19)
 New Insights into the Control of Cell Growth 225

4. DNA extraction buffer: 200 mM Tris–HCl (pH 7.5) (Duchefa

Biochemie, NL-Haarlem), 250 mM NaCl, 25 mM EDTA,
0.5% SDS.
5. Isopropanol.
6. Double distilled water.

2.4.2. Protocol 2: Rapid 1. 200 mL PCR tubes.

DNA Extraction from 2. 20–200 mL pipettes and tips.
Leaves
3. QuickExtract™ Plant DNA extraction solution (EPICENTRE
biotechnologies, Madison, WI, USA).
4. Thermocycler.

2.5. Genotyping: 1. Mastermix for a 10 mL reaction using Biotherm DNA poly-

Confirmation merase: 6.25 mL Double distilled water, 1 mL 2 mM
of the Mutant Di-nucleotide-tri-phosphates (dNTP) (Fermentas GmbH,
Status by CAPS Leon-Rot, Germany), 1 mL 10× Reaction buffer Biotherm
Markers containing 15 mM MgCl2 (Fermenta GmbH, Leon-Rot,
Germany), 0.5 mL 5 mM Forward primer, 0.5 mL 5 mM
Reverse primer, 0.05 mL DNA polymerase (5 units/mL;
Biotherm, Gene Craft, Cologne, Germany).
Add 0.7 mL plant DNA per reaction.
2. Mastermix for a 10 mL reaction using KAPA 2G Robust:
6.36 mL Double distilled water, 0.2 mL 10 mM dNTPs (Kapa
Biosystems), 2 mL 5× Buffer B (Kapa Biosystems, USA) ,0.3 mL
5 mM Forward primer, 0.3 mL 5 mM Reverse primer, 0.02 mL
Kapa 2G Robust polymerase (5 units/mL; Kapa Biosystems,
USA).
Add 0.7 mL Plant DNA per reaction.
3. AvaII mastermix for uge4-3 genotyping: 0.25 mL AvaII
restriction endonuclease (5 units/mL; Fermentas GmbH,
Leon-Rot, Germany), 2 mL 10× Buffer Red (Fermentas
GmbH, Leon-Rot, Germany), 7.75 mL Double distilled
water.
4. BfuI mastermix for sos5 genotyping: 0.25 mmL BfuI restric-
tion endonuclease (5 units/mL; Fermentas GmbH, Leon-
Rot, Germany), 2 mL 10× Buffer BfuI (Fermentas GmbH,
Leon-Rot, Germany), 7.75 mL Double distilled water.
5. PCR strips or plates.
6. Loading dye/buffer: (60% glycerol, 50 mM Tris–HCl
(pH 8.0), 25 mg/mL bromophenol blue).
7. Agarose gel (1.4% for CAPS markers): 1× TAE: 40 mM Tris-
acetate, 1 mM EDTA, LE Agarose 1 g/100 mL 1× TAE
(Biozym Scientific GmbH, Oldenburg, Germany).
226 Blaukopf, Krol, and Seifert

2.6. Rough Mapping 1. Materials (see Subheading 2.5).

2. Molecular markers to distinguish ecotypes Columbia and
Landsberg at various locations on each chromosome. We gen-
erally start with three different molecular markers per chro-
mosome. Additional markers might be required depending on
initial data. Sequences of all nga markers are published (20).
Other markers are described in http://www.arabidopsis.org.
Chromosome 1: nga63, F19G10 (forward: AGTTGGTC
CTCGAGCTCTCC, reverse: AAGAACTTAATTTCT
CTCACCCG), nga111.
Chromosome 2: RGA (RsaI for restriction digestion (see
Subheading 2.5) forward: TTCGATTCAGTTCGGTT
TAG, reverse: GTTTAAGCAAGCGAGTATGC), ciw3
(forward: GAAACTCAATGAAATCCACTT, reverse:
TGAACTTGTTGTGAGCTTTGA), nga168.
Chromosome 3: nga172, nga162, nga6.
Chromosome 4: nga8, ciw6 (forward: CTCGTAGTGCACT
TTCATCA, reverse: CACATGGTTAGGGAAACAATA),
nga1107.
Chromosome 5: nga106, nga76, ciw10 (forward: CCACA
TTTTCCTTCTTTCATA, reverse: CAACATTTAGCAA
ATCAACTT).

2.7. Fine Mapping 1. Materials (see Subheading 2.6).

2. Additional molecular markers in the region determined by
rough mapping. Markers can be found either on the
Arabidopsis homepage (http://www.arabidopsis.org; marker
search tool), or can be designed using the Monsanto polymor-
phism collection (also available at http://www.arabidopsis.
org; registration required to access the data).

2.8. Sequencing 1. DNA of the plants of interest.

2. Primers flanking the region of interest (5 mM for PCR,
1.5 mM for sequencing). Use a primer length of 20–25 base
pairs (bp) to ensure binding specificity, the GC content
should be ³40% and the first 1–2 bp at the 3¢ end should be
a G or a C.
3. PCR and Gel electrophoreses (see Subheading 2.5).
4. PCR purification or Gel extraction kit (Qiagen, Hilden,
Germany).
5. Sequence viewer software such as Chromas (Freeware,
Technelysium Pty Ltd, Tewantin QLD, Australia).
 New Insights into the Control of Cell Growth 227

2.9. Initial 1. 100 mM phosphate buffer, pH 7.

Characterization 2. Paraformaldehyde-glutaraldehyde-fixative, final concentrations:
of uge4-3 Enhancers
100 mM Phosphate buffer pH 7, 2% paraformaldehyde, 2.5%
by Immunohisto-
glutaraldehyde. Preparation of 10% paraformaldehyde stock
chemistry
solution from powder: add 2 g of PFA powder to 20 mL
2.9.1. Fixation dH2O and heat up to 60–65°C. Add a few drops of 1 N
and Embedding NaOH until the solution turns clear and leave to cool at room
temperature.
3. Ethanol in the following concentrations: 30, 50, 70, 80, 90,
and 96%.
4. LR White Embedding Kit for heat polymerization or acceler-
ated polymerization at room temperature.
5. Molding trays with lid (Plano, IL, USA) or gelatine capsules
with holder (Plano). Alternatively, 0.5 or 0.2 mL Eppendorf
tubes (depending on the specimen size) can be used
(Eppendorf, Hamburg, Germany).
6. 60°C cabinet for polymerization.
7. Razor blades for trimming (Plano).
8. Low melting point agarose (Sigma).
9. Two-component glue (e.g., UHU plus endfest 300).

2.9.2. Sectioning 1. (Ultra) Microtome (e.g., Leica Ultracut R, Leica, Wetzlar

Germany).
2. Diamond knife (Diatome, Biel, Switzerland, Diatome histo).
3. Multiwell (5–6 mm well-diameter) microscope slides, adhe-
sive (Tekdon inc., Myakka City, FL, Slide-ID: 12-54).
Alternatively, adhesive slides without wells can be used, and
wells can be produced with a grease pencil (e.g., Pap Pen,
Plano, IL, USA).
4. Distilled Water and Pasteur pipette for wetting microscope
slides.
5. Eyelash glued to matchstick to transfer sections to microscope
slide.
6. Heating block for drying slides.

2.9.3. Antibody Labeling 1. 1× PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4,

2 mM KH2PO4.
2. 3% Bovine serum albumin (BSA) in 1× PBS.
3. Staining jar (Plano, Plano, IL, USA).
4. Primary antibodies. Frequently used monoclonal antibodies
reacting with cell wall carbohydrate e.g., LM2 (21), LM5
(22), CCRC-M1, CCRC-M7 (23), were the generous gift of
228 Blaukopf, Krol, and Seifert

Paul Knox and Michael Hahn and can be obtained at

http://www.plantprobes.net/ and http://www.ccrc.uga.
edu/~carbosource/CSS_home.html.
5. Secondary antibody (goat antimouse IgG + IgM Immunogold
conjugated, British Bio Cell Int. Cardiff, UK). This second-
ary antimouse antibody can be used for both mouse and rat
primary antibodies.
6. Silver enhancement kit (BBI international, Cardiff, UK).
7. Glycerol (Roth, Karlsruhe, Germany).
8. Cover glass.

2.9.4. Microscopic 1. Fluorescence microscope (e.g., Zeiss Axiovert 200M)

Examination equipped with mercury lamp (OSRAM, Augsburg, Germany,
product no. HBO 103W/2), suitable filter set and camera
(e.g., AxioCam MRc5). Lenses: Zeiss Plan-Apochromat 63×
oil; Zeiss Plan-Neofluar 40× oil; Zeiss LC Achroplan 20×.
2. Confocal microscope.
Argon Laser (488 nm) for reflection contrast of immunogold
labeling, UV laser (405 nm) for Calcofluor. Reflection:
lem = 470–500; Calcofluor: lem = 400–450 nm; Image size
2,048 × 2,048 pixel. Lens: 63× oil.

2.10. Synthesis 1. p-Nitrophenyl-b-d-glucopyranoside (Sigma, Saint Louis, MO,

of b-Glucosyl Yariv USA).
Reagent (24) 2. Round bottom flask with septum and hollow needle.
3. Anhydrous methanol.
4. Charcoal with 10% Pd (Aldrich, Saint Louis, MO, USA).
5. Ammonium formate.
6. Celite packed column (250 mL capacity).
7. Vacuum pump with suction filter.
8. 5% H2SO4.
9. 5% NaOH.
10. Double distilled H2O.
11. Magnetic stirrer.
12. Temperature-controlled water bath.
13. Rotary Evaporator.
14. Sodium nitrite (Aldrich, Saint Louis, MO, USA) solution
(1.0288 g in 56 mL HQ H2O).
15. Phloroglucinol (Sigma, Saint Louis, MO, USA) solution
(0.5952 g in 192 mL HQ H2O).
16. Ethanol abs.
17. HPLC (and NMR) to test product purity.
 New Insights into the Control of Cell Growth 229

2.11. Assay for 1. 6-well-plates (Greiner, Frickenhausen, Germany).

b-Glucosyl Yariv 2. SGM (0.4% Phytagel) (see Subheading 2.3).
Sensitivity
3. Liquid SGM (SGM without Phytagel).
4. b-glucosyl Yariv (see Subheading 2.10).
5. Propidium iodide (PI) (Molecular Probes, Eugene, OR,
USA). CAUTION: Toxic! Use gloves!
6. Additional compounds to be tested in the assay (in dH2O or
DMSO).
7. Microscope (e.g., Zeiss Axiovert 200M with Axio Vision 4.6
software) equipped with a mercury lamp and a suitable fluo-
rescent filter set (Zeiss filter set 15 Ex 546/12 for propidium
iodide lex = 536 nm, lem = 617 nm).
8. Software for data analysis (ImageJ/NIH image, public
domain software; http://rsb.info.nih.gov/nih-image/).

3. Methods

3.1. EMS Mutagenesis 1. Imbibe 100–200 mg of homozygous uge4-3 and sos5 seeds
overnight in 20 mL 0.1% KCl at room temperature.
2. In chemical hood: After discarding the supernatant, incubate
seeds in 20 mL EMS solution for 3 h at room temperature on
a rotary shaker (200 rpm).
3. In chemical hood: Dispose EMS solution supernatant by pour-
ing into beaker containing 10 g solid Na-thiosulfate. Rinse
seeds twice in quenching buffer. After 30 min quenching,
EMS waste can be disposed off.
4. Rinse seeds four times in dH2O (nonsterile).
5. Suspend seeds in 30–50 mL dH2O and distribute 1 mL
­aliquots to 6 × 6 cm pots filled with cryosanitized soil.
Intermittent stirring of seed suspension ensures that numbers
of M0 individuals are similar in each pool.
6. Germinate seedlings under Perspex hood in greenhouse using
mixed natural and 16 h artificial light.
7. When plants start flowering, remove hoods and separate plant
pots sufficiently to avoid cross-contamination.
8. Upon senescence, wrap plant clusters in glassine paper bags
and keep watering until fully senescent.
9. Collect M1 seed pools for screening.

3.2. Primary Screen 1. Approximately 300–400 seeds from each M1 seeds pool are
of uge4-3 Enhancers suspended in sterilization solution in 1.5 mL plastic tubes,
and sos5 Suppressors mixed initially by vortexing and intermittently by inversion
230 Blaukopf, Krol, and Seifert

for 10–20 min. In separate tubes wild type (Col-0), mutant

background (uge4-3, sos5) and full loss of function uge4
mutants (uge4-4) are prepared as controls. Depending on the
experience of the worker 10–20 M1 populations can be pro-
cessed on 1 day (2–6,000 individual seeds).
2. Sterilization solution is decanted and seeds are rinsed three
times with sterile dH2O.
3. Seeds are placed individually onto SGM plates (for uge4-3) or
SGM 4% sucrose (for sos5) in 2–3 mm distance in lines 20 mm
apart.
4. To facilitate simultaneous germination, plates are cold-treated
at 4°C in the dark for 48 h.
5. For germination plates are incubated at 25°C in continuous
light in an almost vertical slightly inclined position in a way
that lines of seeds are horizontal.
6. Morphological appearance of root tips is visually screened in
a dissecting microscope at approximately five to tenfold mag-
nification and putative uge4-3 enhancers and sos5 suppressors
are marked on the bottom of the Petri dish. While uge4-3
looks like wild type under the screening conditions, putative
enhancer mutants of uge4-3 appear similar to full loss of func-
tion uge4-4, displaying root epidermal bulging, collapsed epi-
dermal cells, and wrinkled roots (14). Under the described
conditions, sos5 mutants can be distinguished from wild type
by their fatter root. Putative suppressors of sos5 display a more
wild-type-like root morphology. The typical sos5 phenotype is
only seen upon transfer to high salt medium (6). Germination
on high salt medium is not recommended as it reduces ger-
mination efficiency of both mutant and wild-type seedlings.
7. To perform an additional confirmation of the nature of the
putative modifier mutations, putative uge4-3 enhancer seed-
lings are transferred to plates containing Gal-selection medium
and putative sos5 suppressor seedlings are transferred to high
salt medium. Because the effect of the uge4 mutation on cell
walls is suppressed by 5 mM galactose (15), a uge4-3 enhancer
uge4-3 double mutant appears more like wild type on Gal-
selection medium. The phenotype of sos5 is clearly distinct
from wild type upon transfer to high salt medium. A sos5 sup-
pressor, sos5 double mutant looks more like wild type under
these conditions.
8. After the appearance of the putative modifier mutants is con-
firmed for 48 h on second selection plates, seedlings are trans-
ferred to soil to harvest M2 seeds.

3.3. Secondary Screen 1. To confirm heritability of the modifier trait, M2 families of

putative double mutants are plated and germinated on SGM
(see Subheading 3.2).
 New Insights into the Control of Cell Growth 231

2. Five days after germination, equal numbers of putative uge4-3

enhancers are transferred to Gal-selection medium and SGM.
Phenotype of a genuine enhancer should be clearly distinct
on the two media. Putative sos5 suppressors are transferred to
high salt medium and SGM. Phenotypes of genuine suppres-
sors should be similar on either medium.

3.4. Establishing A crucial part of mutant gene localization and mapping approaches
a Genetic Model and following the initial mutagenesis of Arabidopsis thaliana
and a Mapping plants is the crossing of mutant lines. The back-crosses are per-
Population formed between modifier mutants and their genetic background
i.e., uge4-3 and sos5 single mutants, respectively. Because neither
very weak uge4 nor sos5 mutant alleles exist in a non-Columbia
background, we generate outcrosses for mapping with Landsberg
erecta (Ler) plants. To generate mapping populations, the mutant
status of individual F2 plants at the background loci is to be deter-
mined by genotyping as described above.

3.4.1. Pollination 1. Choose a suitable acceptor plant (see Note 1) and remove all
interfering parts, like pollinated siliques, open flowers, or very
young buds (see Fig. 1) with sharp forceps or with scissors.

Fig. 1. Preparation for crossing. (a) Open donor flower with carpel (arrow 1) and stamens
containing mature pollen (arrow 2). (b–d) Preparation of the acceptor. (b) Removal
of very young buds. (c) Loosening of a not yet opened acceptor bud to reveal the carpel.
(d) Carpel after removal of petals and stamens ready for fertilization by a donor flower.
232 Blaukopf, Krol, and Seifert

Do not damage the stem while removing the siliques, flowers,

or buds.
2. Normally three to five closed big buds are used for the
crossing.
3. Puncture the sepal of the closed buds with sharp forceps and
carefully open the bud to make the anthers accessible (see
Note 2).
4. Remove all anthers to prevent self-fertilization (see Note 2).
5. Excise donor flowers shortly after opening of petals. Squeeze
the flower near the base with the forceps and brush it across
the exposed acceptor carpels. Pollen should be visible on the
stigma.
6. Label the acceptor inflorescence with the donor genotype
using adhesive tape. Check the plants after 4–5 days. If the
cross was successful, the siliques should be elongated.
7. Before opening of siliques, wrap the pollinated inflorescence
into a glassine paper bag (6 × 3 cm) to collect the seeds (see
Note 3).

3.4.2. Genetic Analysis 1. Sterilize and plate the seeds on SGM and observe seedlings.
of F1 and F2 Offspring 2. In case uge4-3 is genetically enhanced by a recessive second
3.4.2.1. F1 Backcrosses site mutation, F1 appear like wild type. In case of a dominant
enhancer locus, depending on its zygosity in the maternal
genome, 50 or 100% of F1 offspring show the enhanced
phenotype.
3. Suppressed sos5 mutant F1 seedlings are transferred to high
salt medium. In case of a recessive suppressor mutation in the
maternal genome, all seedlings display sos5 phenotype (6) (see
Note 1). In case the maternal genome contained a dominant
suppressor mutation, F1 offspring displays salt sensitivity sim-
ilar to wild type.
4. Allow F1 to self-pollinate to generate a large number of F2
offspring.

3.4.2.2. F2 Backcrosses 1. Sterilize, plate, and observe several hundred F2 seeds.

2. Count the phenotypic and wild type-looking plants.
3. Because the background is homozygous for uge4-3, recessive
and dominant uge4-3 enhancer mutations lead to 25 and 75%
phenotypic plants, respectively. Recessive and dominant sos5
suppressors segregate 75 and 25% phenotypic seedlings,
respectively after high salt selection. Deviations from these
values indicate the presence of more than one modifier muta-
tion in the population.
 New Insights into the Control of Cell Growth 233

4. When the segregation ratio indicates simple interactions,

additional backcrosses are essential before the mutants are
further characterized.

3.4.2.3. F2 Outcrosses 1. Sterilize, plate, and observe several hundred F2 seeds.

2. In the presence of an unlinked, recessive, or dominant uge4-3
enhancer, 6.25 or 18.75% respectively, of the offspring are
expected to be phenotypic. In the outcross of the sos5 sup-
pressor mutants, 18.75 or 6.25% of F1 show the normal sos5
phenotype (on high salt medium) in the presence of an
unlinked, recessive, or dominant suppressor, respectively.
3. Confirm the expected mutant status via genotyping of pheno-
typic uge4-3 plants. Genotype the phenotypically wild type
plants at the SOS5 locus to identify suppressed homozygous sos5
plants. These plants are expected to contain the sos5 suppressor
mutation and are subsequently used for rough mapping.

3.5. Genomic DNA The extraction of genomic DNA from a large number of indi-
Extraction vidual plants is essential for map-based cloning approaches and is
inherently time consuming. The presented two alternative extrac-
tion protocols are both “quick and dirty” but individually offer
advantages in higher DNA yield and lower cost or increased
speed, respectively. We recommend protocol 3.5.1 (see
Subheading 3.5.1) for rough mapping and protocol 3.5.2 (see
Subheading 3.5.2) for collection of recombinant chromosomes
during fine mapping.

3.5.1. Genomic DNA 1. Leaves or inflorescences are excised and macerated in

Extraction for Rough Eppendorf tubes containing 400 mL of DNA extraction
Mapping (Modified from buffer.
(19)) 2. Debris is pelleted by centrifugation.
3. 300 mL of the extract is transferred into a fresh Eppendorf
tube and 300 mL of isopropanol is added to precipitate the
DNA.
4. Centrifuge at 16,110 × g for 5 min and discard the solution to
dry the DNA pellet (see Note 4).
5. After all isopropanol is vaporized, the DNA pellet is resus-
pended with 50 mL sterile deionised water (see Note 5).

3.5.2. QuickExtract™ Plant 1. Pipette 35 mL of the QuickExtract™ buffer into 200 mL PCR
DNA Extraction for Fine tubes.
Mapping 2. Transfer excised leaf fragments (ca. 2 × 3 mm) into the
buffer.
3. Incubate the leaves at 37°C for 12 min in a thermocycler.
4. Inactivate the enzymes in the buffer at 65°C for 15 min.
234 Blaukopf, Krol, and Seifert

3.6. Genotyping: 1. Primers: UGE4Seq2F, AGATCATGGACAAAATCTCACC

Confirmation GTCCACCAG; EPIM5R, ATTACTTCCAAAAGTGTAAT
of the Mutant Status AGTCGCAATC.
by (CAPS) Markers 2. In 200 mL PCR tubes, add 0.7 mL DNA to 9.3 mL of the
3.6.1. uge4-3 Genotyping mastermix for each reaction. Include wild type and mutant
controls.
3. Amplify using the following PCR protocol.
94°C 2 min 1 cycle
94°C 10 s
56°C 10 s
72°C 35 s to step 2, 35 cycles
15°C for ever for ever

4. Add 10 mL of AvaII mastermix and incubate at 37°C for 12 h

and at 65°C for 20 min.
5. Add 3 mL of loading buffer to each reaction.
6. Load 10 mL of each reaction to 2% agarose gel and run elec-
trophoresis in 1× TAE buffer (80 V for a 100 mL gel or 100 V
for a 250 mL gel).
7. Image the gel in a UV transilluminator after 35 min. Compare
the DNA samples with your wild type control and your
mutant control to see if they have a uge4-3 mutation, are
heterozygote, or wild type plants.
8. The PCR reaction produces a 708 bp product, which in the
wild type, is cut by AvaII at one position to yield 425 bp and
283 bp fragments. The uge4-3 allele is not cut.

3.6.2. sos5 Genotyping 1. Primers: FLASeq3F, ATCACAGTCTTCGTCCCCACCG

ACTC; FLASeq3R, TAAAAATAAATAAAGCCCTAATC
GAAGG.
2. In 200 mL PCR tubes, add 0.7 mL DNA to 9.3 mL of the
mastermix for each reaction. Include wild type and mutant
controls. Amplify using the following PCR protocol.
92°C 2 min 1 cycle
92°C 10 s
68°C 10 s
72°C 45 s to step 2, 27 cycles−0.5°C per cycle
92°C 10 s
54°C 10 s
72°C 50 s to step 5, 30 cycles
15°C for ever
 New Insights into the Control of Cell Growth 235

3. Add 10 mL of BfuI mastermix and incubate at 37°C for 12 h

and 65°C for 20 min.
4. Analyze products on a 2% agarose gel.
5. The PCR reaction generates a 603 bp product, which in the
wild type, is cut by AvaII at one position to yield 324 bp and
279 bp fragments. The sos5-1 allele is not cut.

3.7. Rough Mapping Rough mapping positions a mutant locus to an approximate

genetic locus on a chromosomal arm. Here, we provide a proto-
col for a simple mapping procedure as applied to recessive modi-
fiers of uge4-3 and sos5. For general literature on gene mapping in
Arabidopsis see (20, 25).
1. F2 plants are selected as described in Subheading 3.4.2 (see
Note 1).
2. Identify homozygous double mutants for DNA isolation. (In
case of uge4-3 enhancers, confirm homozygous uge4-3 mutant
status and galactose rescuable uge4 phenotype. In case of sos5
suppressors, confirm homozygous sos5 mutant status and
absence of high salt induced root swelling.)
3. Make PCR mastermix and use the different primers of the
molecular markers to cover all five chromosomal arms (see
Subheadings 2.5 and 2.6).
4. Pipette 0.7 mL of DNA into 9.3 mL of the mastermix for each
molecular marker.
5. Modify the PCR protocol (see Subheading 3.5.1) by
­decreasing the annealing temperature to 50°C for nga63,
F19G10, nga111, RGA, ciw3, nga172, nga162, nga6, nga8,
ciw6, nga1107, nga106, and ciw10 and to 53° for nga168
and nga76.
6. The CAPS marker RGA needs an additional restriction endo-
nuclease digestion step (see Subheading 2.5) using RsaI as
the restriction enzyme.
7. Analyze PCR products on 4% agarose gel.
8. Determine the frequency of chromosomes from Columbia vs.
Landsberg: A single band counts for two chromosomes for
the respective genotype, a doublette counts one for each gen-
otype. Note that the frequency of Landsberg chromosomes
will be much less than 50% at two genetic loci. One locus is
the original mutation and the other locus is the modifier
mutation.
9. To confirm the recessive nature of the mutation, approxi-
mately 100 F3 seeds are germinated and inspected at selective
conditions to confirm homozygosity of the modifier allele.
236 Blaukopf, Krol, and Seifert

3.8. Fine Mapping After the approximate location has been roughly mapped, the
exact location can be determined by searching for wild type DNA
on either side of the rough location. Fine mapping usually requires
the generation of new markers at high density close to the mutant
locus. Markers can be self-designed by using the Monsanto poly-
morphism collection (also available at http://www.arabidopsis.
org; registration required to access the data), which comprises of
polymorphisms, either single nucleotide (SNP), or single sequence
length polymorphisms (SSLP) between the two ecotypes
Columbia (col-0) and Landsberg (ler).
1. Sterilize and plate seeds of the mapping population (see
Subheading 3.2).
2. Select phenotypic plants.
3. Extract DNA of phenotypic plants (see Subheading 3.5).
4. In case of a secondary mutation (enhancer or suppressor of a
primary mutant), genotype the primary mutant locus to make
sure the plants used for mapping are homozygous for the
mutation at this locus or heterozygous. This information is
crucial, as ratio of phenotypic to nonphenotypic plants in the
F3 generation determines the status of the modifier locus.
5. Identify recombinant plants by testing each DNA with two
markers, one on each site of the rough location of the muta-
tion. Plants in which the genotypes at these two markers dif-
fer from each other are considered recombinants and can be
used to further delimit the locus.
6. Transfer recombinant plants to soil and collect F3 seeds.
7. Plate approximately 100 F3 seeds of each population and
assess phenotype to determine mutant status and zygosity at
the modifier locus.
8. The presence of Columbia (mutant) and Landsberg (wild
type) chromosomes has to be equivalent to the genotype of
modifier locus as determined in the F3 population. If the
genotype at a particular marker is different from the genotype
of the modifier background, then the corresponding genomic
position can be ruled out as candidate region. Gradually,
identifying more genetic markers and recombinant chromo-
somes closer to the mutant locus allows delimitation of the
position of the modifier locus until sequencing of candidate
genes becomes economically feasible.

3.9. Sequencing 1. Using mutant DNA as template, amplify overlapping frag-

ments of open reading frame under investigation, and purify
DNA using PCR purification kit or gel extraction kit depend-
ing on the specificity of the reaction, according to manufac-
turer’s instructions.
 New Insights into the Control of Cell Growth 237

2. Supply isolated DNA and primers to commercial sequencing

service.
3. Compare resulting sequences with published wild type
sequence.

3.10. Initial 1. Up to eight samples (depending on the material) are arranged

Characterization side by side and embedded in a block of 2.5% low melting
of uge4-3 Enhancers agarose (LMA) prior to fixation. Place a Petri dish on a heat-
by Immunohisto­ ing table set to 30°C and add LMA (~1 mL per block). Place
chemistry the samples parallel to each other into a drop of 2.5% LMA
using forceps, until they are covered with LMA. Place them
3.10.1. Fixation closely to each other but avoid contact between them. Let the
and Embedding LMA drop harden at room temperature and cut out a block
containing the sample with a razor blade (see Note 6).
2. Transfer the agarose embedded samples into freshly prepared
paraformaldehyde-glutaraldehyde-fixative (1 mL in 1.5 mL
tubes).
3. Apply vacuum for 15 min to enhance fixation efficiency, fol-
lowed by incubation at 4°C for 3 h.
4. Rinse three times 20 min in 0.1 M phosphate buffer, pH 7.
5. Dehydrate the samples in a graded ethanol series in the
­following concentrations: 30, 50, 70, 80, 90, and 96%, each
step for 20 min at 4°C.
6. Remove all ethanol and incubate samples two times with
1 mL LR-White at 4°C, each time for 3 h.
7. Each sample is transferred to individual gelatine capsules or
molding trays containing LR-white (two-thirds of final
­volume). In case of using a molding tray, it is crucial that it
can be closed with a lid because oxygen inhibits LR polymer-
ization. Arrange the samples with forceps in a way that the
desired orientation is orthogonal to the length axis of the
gelatine capsule, fill container with LR White and close the lid
(see Note 7).
8. Polymerize at 60°C for 24 h or at room temperature using
the accelerator comprised in the embedding kit. Allow blocks
to remain at room temperature for at least 12 h after polym-
erization and before sectioning (see Note 8).

3.10.2. Sectioning 1. Remove the (gelatine) capsule and trim the area containing
the samples with a razor blade by cutting away the resin that
surrounds sample material. The resulting block containing
the sample should be as small as possible. Also make sure to
trim the cut surface in the desired angle (in case of transverse
sections, orientate the cut surface perpendicular to the length
axis of the sample).
238 Blaukopf, Krol, and Seifert

2. For immunohistochemistry, sections of 0.5–2 mm are cut

either with a glass knife or with a diamond knife on an
ultramicrotome.
3. Transfer sections to multiwell microscope slide (add drop of
water onto wells to allow sections to spread). If a diamond
knife with water bath is used, the sections float on the surface
and can be easily transferred with an eyelash. Depending on
the size of the section, 3–6 sections can be put on the same
well. Evaporate water using a 60°C hot heating block plate
(see Note 9).

3.10.3. Antibody Labeling 1. Block slides with 3% BSA in PBS for 1.5 h at room tempera-
ture by adding 40 mL aliquots onto slide wells.
2. Remove blocking solution carefully using a micropipette
and add 40 mL primary antibody (diluted in blocking ­solution)
to each well and incubate for 2 h at room temperature (see
Note 10). In case of the negative control, use blocking
­solution instead.
3. Wash three times with PBS in the same way as described for
other buffers for 1–5 min per washing step.
4. Incubate with secondary antibody (diluted in blocking solu-
tion) for 2 h at room temperature (see Note 10) (in case of
CBMs: antipoly Histidine, in case of JIMs and LMs gold-
conjugated antimouse).
5. Wash three times with PBS.
6. In case of CBMs: Incubate with gold-conjugated antimouse
diluted in blocking solution for 2 h at room temperature.
7. Washing (in staining jar): Three times with PBS, one time
with PBS containing 0.5% glutaraldehyde, three times with
PBS, and one time with dH2O.
8. Silver enhancement: Mix the required amount (1 drop = 38 mL)
of the enhancer and the initiator solution 1:1 and apply 20 mL
of the mixture to each well. Incubate for 2 min.
9. Rinse with tap water for at least 10 min.
10. Incubate 5 min with Calcofluor White (0.1% w/v).
11. Briefly wash two times with dH2O in staining jar.
12. Air dry slides, mount coverslip with glycerol (see Note 11).

3.11. Synthesis of 1. Place 4 g p-nitrophenyl-b-d-glucoside in a flame dried round

b-Glucosyl Yariv bottom flask, add 400 mL anhydrous methanol and stir until
Reagent (24) the compound is completely dissolved (sonication can be
used to facilitate dissolving). It is important to maintain anhy-
drous conditions because water deactivates the catalyst.
2. Add 0.8 g charcoal with 10% Pd as catalyst and stir until
evenly suspended.
 New Insights into the Control of Cell Growth 239

3. Add 3.84 g ammonium formate as H-donor for the ­reduction

of NO2 to NH2 and stir at room temperature until dissolved.
Close the flask with a septum and pierce it with a hollow
needle to allow H2 to effuse.
CAUTION: Use fumehood to prevent the formation of
oxyhydrogen!
4. Stir 20 min at 50°C to allow the conversion of NO2 to NH2.
Let the solution cool down completely.
5. Filter solution, under vacuum, through a methanol prewashed
Celite packed column (to remove 10% Pd/C) and wash Celite
column with ~300 mL methanol to enhance the yield.
6. Evaporate methanol at 40°C, 100 mbar (The substance starts
foaming when methanol is removed) and dissolve in 80 mL
HQ H2O.
7. Removal of ammonium formate: Add ~6 mL 5% H2SO4 until
pH 3 is reached (use pH test strips) to displace formic acid by
the stronger acid H2SO4, and distil off ~40 mL at 29°C,
11 mbar. Fill up to 80 mL with HQ water. In order to displace
ammonium by the stronger base NaOH, add ~12 mL 5%
NaOH until pH 9.2 is reached (use pH test strips) and distil off
~40 mL at 29°C, 11 mbar. Fill up to 80 mL with HQ H2O.
8. Cool the flask to 0–5°C in an ice-bath and monitor the tem-
perature of the contents. It is crucial that the temperature
stays between 0 and 5°C, to enable the subsequent reaction.
9. Prepare sodium nitrite (1.029 g in 56 ml H2O) and phlorog-
lucinol solution (0.5952 g in 192 mL H2O).
10. Add sodium nitrite solution drop-wise, keeping the tempera-
ture between 0 and 5°C.
11. Add phloroglucinol drop-wise, keeping the temperature
between 0 and 5°C.
12. Keep flask on ice (0–5°C) for 30 min.
13. Place flask onto bench allowing the contents to regain room
temperature.
14. Adjust to pH 9 adding 5% NaOH.
15. Leave reaction at room temperature for 1.5 h and adjust pH
to 9 using 10% NaOH in case it drops below 9 (pH stabilizes
after approximately 1.5 h).
16. Slowly add an equal volume (500 mL) absolute ethanol and
keep at 4°C over night.
17. Recover the precipitate (=b-glucosyl Yariv) by filtration through
a Buechner funnel (pore size 4) and wash two times with abso-
lute ethanol (centrifuge each time before ethanol is decanted).
18. Recover the precipitate by filtration and dry over night using
high-vacuum.
240 Blaukopf, Krol, and Seifert

19. Test purity of the product using HPLC (and NMR).

Expected results:
HPLC: Retention time for p-nitrophenyl-b-d-glucopyranoside is
13.5 min, retention time for b-d-glucosyl-Yariv is 12.0 min.
NMR: p-Nitrophenyl-b-d-glucoside (DMSO) d8.1 (d, 2H,
J = 12.0 Hz), d7.1 (d, 2H, J = 9.0 Hz); b-d-glucosyl-Yariv
(DMSO) d7.5 (d, 6H, J = 12.0 Hz), d7.1 (d, 6H,
J = 12.0 Hz).

3.12. A Semiquantitative This assay can be used to semiquantitatively determine the sensi-
Assay for b-Glucosyl tivity of seedlings to b-Yariv, a chemical that specifically binds
Yariv Sensitivity AGPs and hence interferes with their function. This treatment
of Arabidopsis Roots induces programmed cell death (PCD) in roots. Cell death is
measured using propidium iodide, a nuclear dye that accumulates
in damaged cells. Subsequently the effect of small molecular com-
pounds, or the sensitivity of mutants can be investigated.
1. Sterilize seeds and plate them on 6-well plates, each well con-
taining 1.5 mL SGM (0.4% Phytagel) in a horizontal row in
the upper third of each well, and keep plates at 4°C in the
dark for 2 days. Use three replicas (1 well = 1 replica) per
treatment (see Note 12).
2. Transfer the plates to light and place them horizontally, to
allow germination and root grow into the medium, along the
bottom of the plate, for 3–4 days. Place the plates vertically
for another 1–2 days to allow the roots to grow in parallel and
unidirectionally.
3. Apply 1.5 mL of liquid SGM containing b-glucosyl Yariv (to
50 mM of DMSO stock solution) and, for visualization of cell
death, PI (3 mL of 1 mg/mL) to each well in the sterile bench
and seal the plate with Parafilm® (see Note 13).
CAUTION: PI is toxic and Yariv dyes are irritant. Always
wear gloves!
4. After addition of liquid SGM, incubate plates horizontally.
5. Monitor the increase of PI staining directly on the plates
without removing the lid using an inverse fluorescence micro-
scope (e.g., Zeiss Axiovert 200 M with Axio Vision 4.6 soft-
ware) equipped with a mercury lamp and a suitable fluorescent
filter set (Zeiss filter set 15 Ex 546/12 for propidium iodide
lex = 536 nm, lem = 617 nm) and record images (e.g., Axio
Cam MRc5) of each root tip every 60 min (starting ~3 h after
the beginning of Yariv exposure ;which is the duration for the
first signs of cell death and hence propidium iodide staining
to become visible under the described conditions) for a period
of 24 h using a 5× lens (Zeiss A-Plan 5×/0.12). For each time
point record around ten root tips. To keep exposure within
 New Insights into the Control of Cell Growth 241

the limits of measurement, use PI stained positive control

roots (e.g., uge4-4) that exhibit a strong cell death pheno-
type. To set black level, adjust an empty area to zero.
6. In case of partial or complete suppression of cell death,
­contrast of PI images is low. Therefore, we also record bright
field images of each root of the same position and magnifica-
tion as used for PI images to facilitate the selection of the area
used for PI measurement.
7. Image analysis using ImageJ
This analysis is based on the determination of gray level histo-
grams in a defined area throughout a time course, resulting in
an approximate quantification of the increase of cell death.
Perform image adjustments to maximize contrast across the
entire data set (black level subtraction, contrast spread).
In ImageJ, select the root elongation area using the rectangu-
lar selection tool (see white box in Fig. 2a) starting at the
region just above the lateral root cap extending to the root
hair differentiation zone. Measure the distance from the root
tip to the onset of the selection rectangle in the first root and

Fig. 2. Analysis of Yariv assay. (a) Aligned (representative) PI and bright field images of wild
type root tips of each time point of Yariv exposure with selected area for histogram analy-
sis. (b) Curve of Histogram analysis representing the increase of PI staining over time.
242 Blaukopf, Krol, and Seifert

use this distance as guide for positioning the selection

­rectangle in all other roots. Use the corresponding bright
field images for orientation (Fig. 2a). Perform a histogram
analysis and save the values in tabular form. Given correct
image adjustments, the top 30% of the recorded gray values
represent very bright dead cells and cells undergoing cell
death (Fig. 2b).

4. Notes

1. Under our growth conditions in the green house (14 h ­artificial

light mixed with day light and shading of direct ­sunlight),
40–50 day old plants are at an optimal stage for crosses.
Pollination efficiency of greenhouse grown plants shows
­seasonal and diurnal variation. Sunny morning periods from
September to May have been found to yield the most consis-
tent results. Pests such as thrips and aphids greatly reduce suc-
cess rate hence high phytosanitary standards are essential.
2. Even though the anthers are visible to the naked eye, the use
of a magnifying device, such as a dissecting microscope or
magnifying goggles, is recommended. We prefer to use
­goggles in the greenhouse, because plants are continuously
exposed to greenhouse conditions and are not heat and light-
stressed by microscope illumination.
3. Due to seed dormancy, F1 seeds germinate most efficiently at
least 4 weeks after pollination.
4. Optionally, wash pellet with 70% ethanol for better removal
of isopropanol.
5. With plant DNA extracted with the QuickExtract™, amplifi-
cation using KAPA 2G Robust from Kapa Biosystems, we
consistently achieved good results.
6. The diameter of the gelatine capsule/tube used for polymer-
ization limits the block size. In case of a gelatine capsule size
0, the block should not be broader than 7 mm.
7. If orienting the sample in the desired way is not possible, the
solid resin fragment containing the sample can be cut off and
trimmed and glued to the polymerized LR White block in the
desired orientation using a two-component glue.
8. In case of heat polymerization, stable temperature conditions
are crucial for proper polymerization (±2°C). In case of
polymerization at room temperature using the accelerator,
polymerization takes place within 10–20 min. Use 1 drop of
accelerator per 10 mL LR-White. Addition of the accelerator
causes an exothermic reaction, therefore cooling of the
­molding tray or gelatine capsules on ice is recommended.
 New Insights into the Control of Cell Growth 243

9. The negative control is crucial for antibody labeling, ­therefore

an additional well with sections should be prepared.
10. All incubation steps at room temperature can alternatively be
done at 4°C over night.
11. In order to prevent the coverslip from slipping off the slide, it
is useful to fix it with a piece of sticky tape on both ends.
12. Use Phytagel as gelling agent! Agar is not suitable at such a
low concentration.
13. As negative control add the according amount of DMSO.
Alternatively, inactive Yariv reagent (e.g., b-Gal Yariv) can be
used.

Acknowledgements

We gratefully acknowledge the provision of immunoprobes from

Paul Knox and Michael Hahn and sos5 seeds from Jian-Zhang
Zhu. We thank Markus Blaukopf for assisting with the prepara-
tion of b-Yariv reagent. This work was supported by the Austrian
Science foundation FWF grant P19788.

References
1. Somerville, C., Bauer, S., Brininstool, G., encodes a putative cell surface adhesion pro-
Facette, M., Hamann, T., Milne, J., Osborne, E., tein and is required for normal cell expansion.
Paredez, A., Persson, S., Raab, T., Vorwerk, S., Plant Cell 15, 19–32.
and Youngs, H. (2004) Toward a systems 7. Hématy, K., Sado, P. E., Van Tuinen, A.,
approach to understanding plant cell walls. Rochange, S., Desnos, T., Balzergue, S.,
Science 306, 2206–2211. Pelletier, S., Renou, J. P., and Höfte, H.
2. Ellis, C., Karafyllidis, I., Wasternack, C., and (2007) A receptor-like kinase mediates the
Turner, J. G. (2002) The Arabidopsis mutant response of Arabidopsis cells to the inhibition
cev1 links cell wall signaling to jasmonate and of cellulose synthesis. Curr Biol 17, 922–931.
ethylene responses. Plant Cell 14, 8. Desprez, T., Vernhettes, S., Fagard, M.,
1557–1566. Refregier, G., Desnos, T., Aletti, E., Py, N.,
3. Reiter, W. D., Chapple, C., and Somerville, Pelletier, S., and Höfte, H. (2002) Resistance
C. R. (1997) Mutants of Arabidopsis thaliana against herbicide isoxaben and cellulose defi-
with altered cell wall polysaccharide composi- ciency caused by distinct mutations in same
tion. Plant J 12, 335–345. cellulose synthase isoform CESA6. Plant
4. Li, Y., Smith, C., Corke, F., Zheng, L., Merali, Z., Physiol 128, 482–490.
Ryden, P., Derbyshire, P., Waldron, K., and 9. Scheible, W. R., Eshed, R., Richmond, T.,
Bevan, M. W. (2007) Signaling from an altered Delmer, D., and Somerville, C. (2001)
cell wall to the nucleus mediates sugar-­responsive Modifications of cellulose synthase confer
growth and development in Arabidopsis thali- resistance to isoxaben and thiazolidinone
ana. Plant Cell 19, 2500–2515. ­herbicides in Arabidopsis Ixr1 mutants. Proc
5. Xu, S. L., Rahman, A., Baskin, T. I., and Natl Acad Sci U S A 98, 10079–10084.
Kieber, J. J. (2008) Two leucine-rich repeat 10. Scheible, W. R., Fry, B., Kochevenko, A.,
receptor kinases mediate signaling, linking cell Schindelasch, D., Zimmerli, L., Somerville, S.,
wall biosynthesis and ACC synthase in Loria, R., and Somerville, C. R. (2003) An
Arabidopsis. Plant Cell 20, 3065–3079. Arabidopsis mutant resistant to thaxtomin A,
6. Shi, H., Kim, Y., Guo, Y., Stevenson, B., and a cellulose synthesis inhibitor from Streptomyces
Zhu, J. K. (2003) The Arabidopsis SOS5 locus species. Plant Cell 15, 1781–1794.
244 Blaukopf, Krol, and Seifert

11. Gu, Y., Deng, Z., Paredez, A. R., DeBolt, S., death in Arabidopsis cell cultures and
Wang, Z. Y., and Somerville, C. (2008) ­implicates arabinogalactan protein involve-
Prefoldin 6 is required for normal microtu- ment. Plant J 19, 321–331.
bule dynamics and organization in Arabidopsis. 19. Edwards, K., Johnstone, C., and Thompson, C.
Proc Natl Acad Sci USA 105, 18064–18069. (1991) A simple and rapid method for the
12. Paredez, A. R., Persson, S., Ehrhardt, D. W., preparation of plant genomic DNA for PCR
and Somerville, C. R. (2008) Genetic ­evidence analysis. Nucleic Acids Res 19, 1349.
that cellulose synthase activity influences 20. Bell, C. J., and Ecker, J. R. (1994) Assignment
microtubule cortical array organization. Plant of 30 microsatellite loci to the linkage map of
Physiol 147, 1723–1734. Arabidopsis. Genomics 19, 137–144.
13. Seifert, G. J., and Roberts, K. (2007) The 21. Smallwood, M., Yates, E. A., Willats, W. G.,
biology of arabinogalactan-proteins. Annu Martin, H., and Knox, J. P. (1996)
Rev Plant Biol 58, 137–161. Immunochemical comparison of membrane
14. Baskin, T. I., Betzner, A. S., Hoggart, R., associated and secreted arabinogalactan-­
Cork, A., and Williamson, R. E. (1992) Root proteins in rice and carrot. Planta 198,
morphology mutants in Arabidopsis thaliana. 452–459.
Aust J Plant Physiol 19, 427–437. 22. McCartney, L., Steele-King, C. G., Jordan,
15. Seifert, G. J., Barber, C., Wells, B., Dolan, E., and Knox, J. P. (2003) Cell wall pectic
L., and Roberts, K. (2002) Galactose biosyn- (1→4)-b-d-galactan marks the acceleration of
thesis in Arabidopsis: genetic evidence for cell elongation in the Arabidopsis seedling
substrate channeling from UDP-d-galactose root meristem. Plant J 33, 447–454.
into cell wall polymers. Curr Biol 12, 23. Freshour, G., Clay, R. P., Fuller, M. S.,
1840–1845. Albersheim, P., Darvill, A. G., and Hahn, M.
16. Willats, W. G., and Knox, J. P. (1996) A role G. (1996) Developmental and tissue-specific
for arabinogalactan-proteins in plant cell structural alterations of the cell-wall polysac-
expansion: evidence from studies on the inter- charides of Arabidopsis thaliana roots. Plant
action of b-glucosyl Yariv reagent with seed- Physiol 110, 1413–1429.
lings of Arabidopsis thaliana. Plant J 9, 24. Basile, D. V., and Ganjian, I. (2004) Beta-d-
919–925. glucosyl and alpha-d-galactosyl Yariv reagents:
17. Guan, Y., and Nothnagel, E. A. (2004) syntheses from p-nitrophenyl-d-glycosides by
Binding of arabinogalactan proteins by Yariv transfer reduction using ammonium formate.
phenylglycoside triggers wound-like responses J Agric Food Chem 52, 7453–7456.
in Arabidopsis cell cultures. Plant Physiol 135, 25. Lister, C., and Dean, C. (1993) Recombinant
1346–1366. inbred lines for mapping Rflp and phenotypic
18. Gao, M., and Showalter, A. M. (1999) Yariv markers in Arabidopsis thaliana. Plant J 4,
reagent treatment induces programmed cell 745–750.
 Chapter 17

Extraction and Detection of Arabinogalactan Proteins

Zoë A. Popper

Abstract
Arabinogalactan proteins are a diverse group of plant cell wall-associated proteoglycans. While structural
and molecular genetic analyses have contributed to the emerging improved understanding of the wide-
range of biological processes in which AGPs are implicated; the ability to detect, localise, and quantify
them is fundamentally important. This chapter describes two commonly used methods, histological
staining and radial gel diffusion, both of which utilise the ability of Yariv reagent to bind to AGPs.

Key words: Yariv reagent, Arabinogalactan proteins, Radial gel diffusion, Semi-quantitative
detection, Arabinogalactan protein detection

1. Introduction

One of the defining features of plants is that their cells are sur-
rounded by a cell wall consisting of structural polysaccharides and
associated proteins. One class of proteins, the arabinogalactan
proteins (AGPs), has been strongly implicated in terrestrial plant
developmental processes, but their roles are as yet poorly defined
(1–7). A further complexity in understanding AGP function is
that microarray analysis suggests that it is likely that different
AGPs have different functions (8–11).
AGPs consist of two distinct moieties, the carbohydrate and
the protein domain. The carbohydrate component typically
accounts for 90–98% of an AGP by weight and is rich in arabinose
and galactose residues. The protein moiety, accounting for ~10%
of an AGP by weight is hydroxyproline-rich (7). However, there
is a wide range of variability in the structure and composition of
AGPs such that they are frequently classified as being classical or
non-classical AGPs (9–15). Classical AGPs contain a hydrophobic
transmembrane domain, which in mature AGPs is replaced by a

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_17, © Springer Science+Business Media, LLC 2011

245
246 Popper

glycosylphosphatidylinositol (GPI) lipid anchor (10, 14, 16–18);

this domain appears to be lacking from the non-classical AGPs
(10). Non-classical AGPs also tend to be less heavily glycosylated
(19). Differences in AGP composition have also been discovered,
which can be related to terrestrial plant taxonomy, for example,
bryophyte AGPs contain the sugar residue 3-O-methyl-l-
rhamnose, which appears to be absent from flowering plant
AGPs (20).
AGPs are widespread in land plants (21) and more recently
have been found in the cell walls of fresh-water green algae,
Micrasterias (22) and Chara corallina (23), and green seaweeds
including Codium fragile (24). It is not yet possible to determine
whether and which types of AGPs are present in specific plants
based on the analysis of genetic sequence data (5). There are
several methods, both in situ and ex situ, which can be used to
detect, quantify, and localise AGPs; each method has been essen-
tial to our understanding of AGP structure, function, and locali-
sation. However, all of the methods have some advantages and
disadvantages, caused by the structural and compositional diver-
sity present within both the protein and carbohydrate compo-
nents of AGPs. These are discussed in the introduction to this
chapter and because several of the methods are detailed within
other chapters of this volume, the protocols for radial-gel diffu-
sion and in situ staining with Yariv reagents only are included in
this chapter.
Immunocytochemistry (described by Hervé et al., Chapter 7)
is one of the most informative methods currently available for the
detection and localisation of AGPs within plant tissues.
Furthermore, there are several monoclonal antibodies including
CCRC-M7 (25), JIM4 (26–28), JIM13–16 (1, 28), LM2 (28,
29), LM14 (30), and MAC207 (21, 28) which are known to
recognise specific epitopes present in AGPs and which are com-
mercially available (Biosupplies, Australia; CarboSource, http://
www.ccrc.uga.edu/~carbosource/CSS_mabs7-07.html;
PlantProbes, http://www.plantprobes.net/). This epitope speci-
ficity means that it is likely that a specific monoclonal antibody
may recognise a specific AGP or a group of AGPs containing the
same epitope. However, it is this specificity which currently poses
some limitations because while there are several monoclonal
antibodies which are known to detect and bind to epitopes
present in AGPs (25–32), it is likely that at present not all AGPs
can currently be detected in this way. This problem will eventually
be circumvented by the generation and characterisation of a
greater array of AGP-specific monoclonal antibodies making
monoclonal antibody-based methods even more powerful for the
investigation of AGP function.
However, it can also be useful to apply more broad-spectrum
methods of detection and several such methods for detecting the
presence of AGPs have been developed. These methods nearly all
 Extraction and Detection of Arabinogalactan Proteins 247

employ a group of red-brown synthetic phenylglycoside dyes

known as Yariv reagents (33). Although the exact mechanism of
interaction between Yariv reagents and AGPs has not yet been
fully explained (5), they are widely used in detecting and purifying
AGPs (34). b-d-Glucosyl and b-d-galactosyl Yariv reagents bind
to and precipitate AGPs whereas a-d-galactosyl and a-d-man-
nosyl Yariv regents do not and are often used as controls (5, 33).
Yariv reagents can be used in both in situ and ex situ methods of
AGP detection and quantification. They have also been a key fac-
tor enabling determination of the function of AGPs (2, 33). A
method for in situ Yariv staining of plant tissues is detailed in
Subheading 2 and is widely used (34–38). However, it has the
disadvantage that staining may be obscured if the plant tissues are
deeply pigmented or similarly coloured (red–brown) to Yariv
reagent. Colorimetric methods have been used to determine the
concentration of AGPs present in plant tissues (39). This method
has the advantage that it is quantitative. However, it is not known
whether some AGPs are capable of binding greater (or lesser)
concentrations of AGPs than others, and it is suggested that non-
classical AGPs may display variable binding to b-d-glucosyl Yariv
reagent (40). A further problem associated with in situ staining and
the in situ colorimetric method is that Yariv reagents also bind to
cellulose (41) giving a slightly coloured background making very
low concentrations of AGPs less easy to determine and localise.
Therefore, while it has the disadvantage that information regarding
tissue-specific localisation is lost, it can be informative in some cases
to extract, concentrate, and detect AGPs by Yariv staining ex situ.
The radial gel diffusion assay developed by Van Holst and Clarke
(42) and described below enables detection of low concentrations
of AGPs in a cellulose-free environment. The method is semi-
quantitative in comparison with a dilution series of gum arabic, and
tissue specificity can be partially accommodated by carefully select-
ing the tissues prior to extraction and has been used to determine
which fractions of plant extracts contain AGPs (43, 44).

2. Materials

1. 10 g plant tissue of interest.

2. Extraction buffer: 50 mM Tris–HCl, pH 8, 10 mM EDTA,
0.1% v/v b-mercaptoethanol, 1% w/v Triton X-100 (see
Note 1).
3. Ethanol.
4. 50 mM Tris–HCl, pH 8.
5. Bench centrifuge.
6. 1% w/v NaCl.
248 Popper

7. 0.15 M NaCl.
8. Freeze drier.
9. Agarose gel containing b-d-glucosyl Yariv reagent: 1% w/v
agar, 0.02% w/v b-glucosyl Yariv reagent (commercially avail-
able from Biosupplies Australia, Victoria, Australia or they
can be synthesised (33, 45–47) as described by Blaukopf
et al., Chapter 16), 0.15 M NaCl, 0.02% w/v sodium azide
(see Notes 2 and 3).
10. Agarose gel containing a-d-galactosyl Yariv reagent: 1% w/v
agar, 0.02% w/v a-galactosyl Yariv reagent (Biosupplies
Australia, Victoria, Australia) (see Note 4), 0.15 M NaCl,
0.02% w/v sodium azide.
11. 2 mg/mL b-d-glucosyl Yariv reagent in 0.15 M NaCl.
12. 2 mg/mL a-d-galactosyl Yariv reagent in 0.15 M NaCl.
13. Autoclave.
14. Petri dishes.
15. 4 mg/mL gum arabic.
16. Parafilm®.
17. Aluminium foil.
18. Scanner or digital camera.
19. Automatic pipettes.
20. 50 and 50 mL screw-capped centrifuge tubes.
21. Magnetic stirrer bar and stirring plate.
22. Glass Pasteur pipette or core borer.
23. Mortar and pestle.
24. Liquid nitrogen.
25. Freezer: −20°C.
26. Microscope slides.
27. Cover slips.
28. Microscope with digital camera attachment.
29. White ceramic tile.
30. Backed razor blade.
31. Rocking platform.
32. 1.5 mL Eppendorf tubes.

3. Methods

3.1. In Situ Staining 1. Cut sections of the plant tissue of interest. We find that care-
of AGPs Using Yariv ful hand-sectioning with a backed razor-blade on a white
Reagents ceramic tile works well (see Note 5).
 Extraction and Detection of Arabinogalactan Proteins 249

2. Incubate the plant materials in an Eppendorf tube containing

1 mL of 2 mg/mL b-d-glucosyl Yariv reagent in 0.15 M
NaCl for 1 h, at room temperature, on a rocking
platform.
3. As a control, incubate a duplicate set of plant materials e.g.,
sections from the same plant tissues in 1 mL of 2 mg/mL
a-d-galactosyl Yariv reagent in 0.15 M NaCl under the same
conditions.
4. After 1 h, carefully remove the Yariv reagent taking care not
to damage the plant materials (see Note 6).
5. Add 1 mL of 0.15 M NaCl to each of the Eppendorf tubes
and incubate at room temperature for 5 min.
6. Steps 4 and 5 should be repeated several times until the
0.15 M NaCl is no longer appreciably red.
7. Transfer the sections to microscope slides, examine and record
using a light microscope with a digital camera attached to it
(35–38).

3.2. Extraction 1. Grind 10 g of plant material to a fine power in liquid nitrogen

of Arabinogalactan in a mortar and pestle (see Note 7).
Proteins 2. Add 10 mL of extraction buffer (50 mM Tris–HCl, pH 8,
10 mM EDTA, 0.1% v/v b-mercaptoethanol, 1% w/v Triton
X-100) to the plant material and incubate at 4°C for at
least 3 h.
3. Centrifuge for 10 min at 4,000 × g.
4. Carefully remove the supernatant using a Pasteur pipette and
precipitate polysaccharides and glycoproteins with 5 volumes
of ethanol at 4°C for at least 16 h.
5. Centrifuge for 2 min at 2,000 × g.
6. Carefully remove the supernatant taking care not to disturb
the pellet (see Note 8).
7. Resuspend the pellet in 5 mL of 50 mM Tris–HCl, pH 8.
8. Centrifuge for 10 min at 4,000 × g.
9. Collect the supernatant into a polypropylene tube.
10. Resuspend the remaining pellet in 5 mL 50 mM Tris–HCl,
pH 8.
11. Centrifuge for 10 min at 4,000 × g.
12. Carefully remove the supernatant and pool it with that col-
lected in step 9.
13. Freeze and freeze dry the supernatant.
14. Dissolve the dried supernatant in 500 mL 1% w/v NaCl (see
Notes 9–11) (48).
250 Popper

3.3. Detection Extracts using methods including the one described above may
of Arabinogalactan be used to detect the presence of AGPs in plant materials. It is
Proteins by Radial semi-quantitative if a serial dilution of a known source of AGPs
Gel-Diffusion e.g., gum arabic is included in the assay for comparison. A degree
of tissue specificity is enabled by carefully selecting the plant
materials from which the extracts are made. Additional samples
that can be investigated for the presence and concentration of
AGPs include the culture medium of plant tissue cultures. Some
AGPs are secreted into the culture medium (15, 49), which can
be filtered to remove any cell debris, and then freeze-dried prior
to use in the assay (42).
1. Use the end of a glass Pasteur pipette or a core borer to cut
out wells in the agarose gel containing b-d-glucosyl Yariv
reagent.
2. Into one well load 20–50 mL of 1% w/v NaCl (see Note 12).
3. Into another well load a known amount of gum arabic. If you
are investigating whether AGPs are present or not, then
20–40 mL of a 4 mg/mL solution of gum arabic is suitable.
However, if you are trying to quantify the concentration of
AGPs in the extract or for a given quantity of plant tissue, then
it is advisable to include several wells on the plate, which are
loaded with a dilution series of gum arabic starting with
2–4 mg/mL. The minimum concentration of gum arabic that
will give a positive result using this method is 0.25 mg/mL.
4. Load your extract prepared as described in Subheading 3.1
into remaining wells. It can be useful to load your extracts
into two different wells at two different loadings i.e., in one
well load twice the volume that you load in another well (see
Note 13).
5. Seal plates with Parafilm® to prevent them drying out.
6. Store the plates at room temperature, in darkness, for at least
48 h (see Notes 14 and 15).
7. After 48 h, the results can be recorded either by scanning the
plate or by taking a photograph using a digital camera (see
Note 16). An example is shown in Fig. 1.

4. Notes

1. The extraction buffer generally requires at least 2 h stirring to

sufficiently dissolve the Triton X-100.
2. Caution, sodium azide is highly toxic. Gloves should be worn.
3. Inclusion of sodium azide means that sterile conditions are
not required. The AGP extracts contain high sugar content
 Extraction and Detection of Arabinogalactan Proteins 251

Fig. 1. Radial gel diffusion assay. (a, b) 20- and 40-mL loadings of 0.15 M NaCl.
(c, d) 20- and 40-mL loadings of a solution of 4 mg/mL gum arabic in 0.15 M NaCl.

and could quickly become contaminated with microbes.

However, heat sterilisation could be damaging to the AGPs,
and filter sterilisation is not practical owing to the viscosity of
many of the extracts.
4. a-d-Mannosyl Yariv reagent (Biosupplies, Australia Pty Ltd.)
can be used in place of a-d-galactosyl Yariv reagent.
5. Microalgae and other plant materials can also be surface-
labelled using the same procedure but in this case are not
sectioned prior to incubation.
6. We often find that some very thin hand sections can be dam-
aged if bathing solutions are removed too rapidly. One way to
minimise this is by using a glass Pasteur pipette which has been
heated and stretched so that the opening is narrower than nor-
mal. This also helps prevent uptake into the glass Pasteur
pipette which may occur if the sections or plant materials that
are being stained are either very thin and/or very small.
7. In order to protect the mortar and pestle from possible dam-
age due to sudden temperature changes, it is best to cover the
mortar and pestle in cling film/Saran wrap® and place it in a
−20°C freezer for at least 2 h prior to use.
8. One of the easiest ways to do this is by using a Pasteur pipette
which has been stretched over a flame to reduce the width of
the tip.
9. It is useful to weigh the freeze-dried sample so that the mg/mL
of dissolve extract used in radial gel diffusion can be calculated.
10. Some of the extracts may have a high sugar content and fail
to freeze at −20°C, or once frozen they thaw rapidly. In this
252 Popper

case, the extracts may be frozen at −80°C or alternatively can be

dialysed against distilled water, at 4°C, to remove any sugars
with a low degree of polymerisation (DP). To prevent micro-
bial growth in the extract, it is recommended that dialysis is
against 0.05% w/v chlorbutol (1,1,1-trichloro-2-methyl-2-
propanol), which is volatile so will be removed on freeze
drying (49).
11. If at this concentration the solution is extremely viscous or if
there is a large amount (greater than 0.2 g) of freeze-dried
extract, it is advisable to dissolve the sample in a larger
volume of 1% w/v NaCl.
12. If your samples are not dissolved in 1% w/v NaCl, then load
the buffer that they were dissolved in the control wells instead
of the 1% w/v NaCl.
13. It is essential that each well on the plate is clearly labelled. We
find this easiest to do by printing a grid of 1.2 × 1.2 cm squares
onto an acetate sheet, which can be stuck to the back of the Petri
dish. Each well is then punched into the centre of a square.
14. Darkness can be achieved by wrapping the plates in alumin-
ium foil.
15. The plates can be stored for several weeks, even in the light,
at room temperature provided that they are sealed with
Parafilm® and do not dry out.
16. It takes ~48 h for the AGPs to diffuse into the gel and the
Yariv reagent to bind to and precipitate them.

Acknowledgements

Quentin Coster, a visiting MSc student from the Université

Catholique de Louvain, Belgium, and Tanya Slattery, an under-
graduate project student in Botany and Plant Science at NUI
Galway, are thanked for their technical assistance.

References
1. Knox, J. P., Linstead, P. J., Peart, J., Cooper, 3. Basile, D. V., Kushner, B. K., and Basile, M.
C., and Roberts, K. (1991) Developmentally- R. (1989) A new method for separating and
regulated epitopes of cell surface arabinoga- comparing arabinogalactan proteins in the
lactans-proteins and their relation to root Hepatica. Bryologist 90, 401–404.
tissue pattern formation. Plant J 1, 317–326. 4. Basile, D. V. (1980) A possible mode of
2. Lee, K. J. D., Sakata, Y., Mau, S.-L., Pettolino, action for morphoregulatory hydroxyproline-
F., Bacic, A., Quatrano, R.S., Knight, C. D., proteins. Bull Torrey Bot Club 107,
and Knox, J. P. (2005) Arabinogalactan pro- 325–338.
teins are required for apical cell extension in 5. Seifert, G. J., and Roberts, K. (2007) The
the moss Physcomitrella patens. Plant Cell 17, biology of arabinogalactan proteins. Annu
3051–3065. Rev Plant Biol 58, 137–161.
 Extraction and Detection of Arabinogalactan Proteins 253

6. Seifert, G. J., Barber, C., Wells, B., and 18. Youl, J. J., Bacic, A., and Oxley, D. (1998)
Roberts, K. (2004) Growth regulators and Arabinogalactan-proteins from Nicotiana
the control of nucleotide sugar flux. Plant alata and Pyrus communis containglyco-
Cell 16, 723–730. sylphosphatidylinositol membrane anchors.
7. Fincher, G. G., Stone, B. A., and Clarke, A. E. Proc Natl Acad Sci USA 95, 7921–7926.
(1983) Arabinogalactan-proteins: structure, 19. Showalter, A. M. (2001) Arabinogalactan
biosynthesis and function. Annu Rev Plant proteins: Structure, expression and function.
Physiol 34, 47–70. Cell Mol Life Sci 58, 1399–1417.
8. Schultz, C. J., Rumsewicz, M. P., Johnson, K. 20. Fu, H., Yadav, M. P., and Nothnagel, E. A.
L., Jones, B. J., Gaspar, Y. M., and Bacic, A. (2007) Physcomitrella patens arabinogalactan
(2002) Using genomic resources to guide proteins contain abundant terminal
research directions. The arabinogalactan pro- 3-O-methyl-l-rhamnosyl residues not found
tein gene family as a test case. Plant Physiol in Angiosperms. Planta 226, 1511–1524.
129, 1448–1463 21. Pennell, R. I., Knox, J. P., Scofield, G. N.,
9. Gaspar, Y., Johnson, K. L., McKenna, J. A., Selvendran, R. R., and Roberts, K. (1989) A
Bacic, A., Schultz, C. J. (2001) The complex family of abundant plasma membrane-associated
structures of arabinogalactan-proteins and the glycoproteins related to the arabinogalactan
journey towards understanding function. proteins is unique to flowering plants.
Plant Mol Biol 47, 161–176. J Cell Biol 108, 1967–1977.
10. Majewska-Sawka, A., and Nothnagel E. A. 22. Eder, M., Tenhaken, R., Driouich, A., and
(2000) The multiple roles of arabinogalactan Lütz-Mendel, U. (2008) Occurrence and
proteins in plant development. Plant Physiol characterization of arabinogalactan-like pro-
122, 3–9. teins and hemicelluloses in Micrasterias
(Streptophyta). J Phycol 44, 1221–1234.
11. Kieliszewski, M. J., and Lamport, D. T. A.
(1994) Extensin: repetitive motifs, functional 23. Domozych, D. S., Sørensen, I., and Willats,
sites, post-translational codes, and phylogeny. W. G. T. (2009) The distribution of cell wall
Plant J 5, 157–172. polymers during antheridium development
and spermatogenesis in the Charophycean
12. Nothnagel, E. A., Bacic, A., and Clarke, A. E. green alga, Chara corallina. Ann Bot 104,
(2000) Cell and Developmental Biology 1045–1056.
of Arabinogalactan-proteins. New York:
Kluwer. 24. Estevez, J. M., Fernández, P. V., Kasulin, L.,
Dupree, P., and Ciancia, M. (2009) Chemical
13. Nothnagel, E. A. (1997) Proteoglycans and and in situ characterization of macromolecular
related components in plant cells. Int Rev components of the cell walls from the green
Cytol 174, 195–291. seaweed Codium fragile. Glycobiology 19,
14. Du, H., Simpson, R. J., Clarke, A. E., and 212–228.
Bacic, A. (1996) Molecular characterization if 25. Steffan, W., Kovác, P., Albersheim, P., Darvill,
a stigma-specific gene encoding an arabinoga- A. G., and Hahn, M. G. (1995)
lactan protein (AGP) from Nicotiana alata. Characterization of a monoclonal antibody
Plant J 9, 313–323. that recognizes an arabinosylated (1→6)-b-d-
15. Mau, S. L., Chen, C. G., Pu, Z, Y, Moritz, R. galactan epitope in plant complex carbohy-
L., Simpson, R. J. and Bacic, A. (1995) drates. Carbohydr Res 275, 295–307.
Molecular cloning of cDNAs encoding the 26. Knox, J. P., Day, S., and Roberts, K. (1989) A
protein backbones for arabinogalactan pro- set of cell wall surface glycoproteins forms a
teins from the filtrate of suspension-cultured marker of cell position, but not cell type, in
cells of Pyrus communis and Nicotiana alata. the root apical meristem of Daucus carota L.
Plant J 8, 269–281. Development 106, 47–56.
16. Oxley, D., and Bacic, A. (1999) Structure 27. Stacey, N. J., Roberts, K., and Knox, J. P.
of the glycosylphosphatidylinositol mem- (1990) Patterns of expression of JIM4 arabi-
brane anchor of an arabinogalactan protein nogalactan protein epitope in cell cultures and
from Pyrus communis suspension-cultured during somatic embryogenesis in Daucus
cells. Proc Natl Acad Sci USA 6, carota L. Planta 180, 285–292.
14246–14251. 28. Yates, E. A., Valdor, J. F., Haslam, S. M.,
17. Svetek, J., Yadav, M. P., and Nothnagel, E. A. Morris, H. R., Dell, A., Mackie, W.,
(1999) Presence of a glycosylphosphati- and Knox, J. P. (1996) Characterization of
dylinositol lipid anchor on rose arabino­ carbohydrate structural features reco­gnised
galactan proteins. J Biol Chem 274, by anti-arabinogalactan-protein monoclonal
14724–14733. antibodies. Glycobiology 6, 131–139.
254 Popper

29. Smallwood, M., Yates, E. A., Willats, W. G. T., 39. Lamport, D. T. A., Kieliszewski, M. J.,
Martin, H., and Knox, J. P. (1996) Immuno­ and Showalter, A. M. (2005) Salt stress
chemical comparison of ­membrane-associated upregulates periplasmic arabinogalactan
and secreted arabinogalactan-proteins in rice proteins: using salt stress to analyse AGP func-
and carrot. Planta 198, 452–459. tion. New Phytol 169, 479–492.
30. Moller, I., Marcus, S. E., Haeger, A., 40. Sommer-Knudsen, J., Clarke, A. E., and
Verhertbruggen, Y., Verhoef, R., Schols, H., Bacic, A. (1997) Proline- and hydroxyproline-
Mikkelsen., J. D., Knox, J. P., and Willats, W. rich gene products in the sexual tissues of
G. T. (2008) High-throughput screening of flowers. Sex Plant Reprod 10, 253–260.
monoclonal antibodies against plant cell wall 41. Triplett, B. A., and Timpa, J. D. (1997)
glycans by hierarchical clustering of their car- b-Glucosyl and a-galactosyl Yariv reagents
bohydrate microarray binding profiles. bind to cellulose and other glucans. J Agric
Glycoconj J 25, 37–48. Food Chem 45, 4650–4654.
31. Norman, P. M., Wingate, V. P. M., Fitter, M. 42. Van Holst G. -J., and Clarke, A. E. (1985)
S., and Lamb, C. J. (1986) Monoclonal anti- Quantification of arabinogalactans-protein in
bodies to plant plasma-membrane antigens. plant extracts by single radial gel diffusion.
Planta 167, 452–459. Anal Biochem 148, 446–450.
32. Norman, P. M., Kjellbom, P., Bradley, D. J., 43. Osman, M., Menzies, A. R., Albo Martin,
Hahn, M. G., and Lamb, C. J. (1996) B., Williams, P. A., Phillips, G. O., and
Immunoaffinity purification and biochemical Baldwin, T.C. (1995) Characterization of
characterization of plasma-membrane arabi- gum arabic fractions obtained by anion-
nogalactans-rich glycoproteins of Nicotiana exchange chromatography. Phytochemistry
glutinosa. Planta 181, 365–373. 38, 409–417.
33. Yariv, M., Rapport, M. M., and Graf, L. 44. Parveen, S., Gupta, A. D., and Prasad, R.
(1962) The interaction of glycosides and (2006) Arabinogalactan protein from Arachis
saccharides with antibody to corresponding hypogaea: role as a carrier in drug formula-
phenyl glycosides. Biochem J 85, 383–388. tions. Int J Pharm 333, 78–86.
34. Willats, W. G. T., and Knox, J. P. (1996) A role 45. Yariv, J. H., Lis, E., Katchalshi, E. (1967)
for arabinogalactan-proteins in plant cell expan- Precipitation of arabic acid and some seed
sion: evidence from studies on the interaction polysaccharides by glycosyl phenylazo dyes.
of b-d-glucosyl Yariv reagent with seedlings of Biochem J 195, 1c–2c.
Arabidopsis thaliana. Plant J 9, 919–925. 46. Basile, D. V., and Ganjian, I. (2004) b-d-
35. Anderson, R. L. , Clarke, A. E. , Jermyn, M. Glucosyl and a-d-galactosyl Yariv reagents:
A., Knox, R .B. , and Stone, B. A. (1977) A Syntheses from p-nitrophenyl-d-glycosides by
carbohydrate binding arabinogalactan-protein transfer reduction using ammonium formate.
from liquid suspension cultures of Lolium J Agr Food Chem 52, 7453–7456.
multiflorum endosperm. Aust J Plant Physiol. 47. Blaukopf, C., Krol, M. Z., and Seifert, G. J.
4, 143–158. (2011) New insights into the control of cell
36. Clarke, A. E., Anderson, A. L., and Stone, B. growth. Zoë A. Popper (Ed.), The Plant Cell
A. (1979) Form and function of arabinogalac- Wall: Methods and Protocols. Humana Press,
tans and arabinogalactan proteins. New Jersey, USA, pp. 221–244.
Phytochemistry 18, 521–540. 48. Schultz, C. J., Johnson, K. L., Currie, G., and
37. Gleeson, P. A., and Clarke, A. E. (1979) Bacic, A. (2000) The classical arabinogalactan
Structural studies on the major component of protein gene family of Arabidopsis. Plant Cell
Gladiolus style mucilage, an arabinogalactan 12, 1751–1768.
protein. Biochem J 181, 607–621. 49. Fry, S. C. (2000). The Growing Plant Cell
38. Jermyn, M. A., and Yeow, M. (1975) A class Wall: Chemical and Metabolic Analysis.
of lectins present in the tissues of seed plants. Reprint Edition, Blackburn, Caldwell, NJ.
Aust J Plant Physiol 2, 501–531. [ISBN 1-930665-08-3].
 Chapter 18

Characterization of the Plant Cell Wall Proteome

Using High-Throughput Screens
Sang-Jik Lee and Jocelyn K.C. Rose

Abstract
Plant cell wall proteins play essential roles in many important biological processes, and yet there is still
not a comprehensive catalogue of the cell wall proteome, or “secretome”. Here, we describe three pro-
cedures, including a yeast secretion trap (YST), Agrobacterium-mediated transient expression using a
necrosis-inducing protein (NIP) and protein localization assay using a fluorescent protein, to identify and
confirm the localization of cell wall proteins. The approaches are orthogonal and collectively provide a
powerful suite of approaches to complement more commonly used strategies to isolate plant cell wall-
associated proteins.

Key words: Plant cell wall protein, Secreted protein, Secretome, Yeast secretion trap, Necrosis-inducing
protein, Agroinfiltration, Confocal fluorescence, Extracellular matrix

1. Introduction

A large number of plant proteins are secreted to the plant cell

wall, or apoplastic environment – where they contribute to rein-
forcement or restructuring of wall architecture, protection from
pathogen attack and abiotic stresses, signalling and metabolism of
apoplastic compounds (1, 2). However, there is still only a super-
ficial understanding of the plant cell wall proteome, both in terms
of the complement of secreted proteins and the dynamics of their
expression. There is therefore considerable interest in better
defining the composition of the plant cell wall proteome, or
“secretome”, either by non-targeted profiling of secreted protein
populations or by examining the localization, accumulation, and
structure of specific cell wall proteins.
Many proteomics projects target subcellular proteomes by
extracting highly pure organelle extracts and then fractionating

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_18, © Springer Science+Business Media, LLC 2011

255
256 Lee and Rose

and sequencing the constituent proteins, typically using advanced

mass spectrometry methods. However, in this regard the cell wall
proteome presents a number of conceptual and practical chal-
lenges (1). For example, unlike subcellular compartment such as
the chloroplast or nucleus, the apoplast is a continuum, which
limits the ability to isolate a more comprehensive and uncontami-
nated cell wall protein extract. In addition, many cell wall pro-
teins can be covalently linked into the cell wall matrix and so
highly resistant to extraction. Therefore, while there are many
studies that have examined the protein content of apoplastic fluids,
or cell wall associated extracts, such an approach inevitably
involves both protein losses and contamination by intracellular
proteins (3).
An alternative approach is to use computational prediction to
search for canonical N-terminal signal peptides that target proteins
to the classical secretory pathway (1). However, it is important to
bear in mind that this only identified proteins that are targeted to
the canonical secretory pathway, and the encoded proteins may be
retained within that pathway or targeted to intracellular organelles
or compartments. In other words, the presence of an ER-targeting
N-terminal signal peptide does not indicate that the associated
mature polypeptide is ultimately secreted to the cell wall.
Additionally, there is growing evidence that a subset of cell wall/
apoplastic proteins are secreted via non-classical secretion path-
ways, as has been reported in animals (4), and yeast (5).
A third option is to use so-called functional screens, where
genes encoding secreted proteins are identified based on the ulti-
mate localization of the encoded protein, utilizing protein markers
whose activities confirm targeting to the cell surface.
In this chapter, we describe the use of two such approaches
(the YST and NIP screens) to search cDNA libraries for secreted
proteins, as well as subsequent confirmation of their localization
using fluorescent protein markers (Fig. 1). Such functional
approaches can be valuable complements to more traditional strat-
egies involving the fractionation and sequencing of cell wall
protein extracts and will provide important information for plant
cell wall targeting and localisation of secreted proteins in vivo
(Fig. 1).

2. Materials

2.1. Yeast Secretion 1. Fresh or frozen tissues of interest.

Trap (YST) 2. cDNA Synthesis Kit (Stratagene) (see Note 1).
3. Oligotex mRNA Mini Kit (Qiagen) (see Note 1).
4. QIAquick Gel Extraction Kit (Qiagen) (see Note 1).
Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 257

Plant material

Protein extract RNA extract

Fractionation
cDNA library
( gel/LC )

Mass
spectrometry YST screen
analysis

Database NIP screen

search

Localization confirmation
(e.g. GFP/immunolocalization)

Fig. 1. Flow chart summarizing the steps for identification of plant cell wall proteins. LC
liquid chromatography.

5. QIAquick PCR Purification Kit (Qiagen) (see Note 1).

6. QIAprep Spin Miniprep Kit (Qiagen) (see Note 1).
7. Cloning vector(s): pYST-0, -1, and -2 three vector set (6).
These vectors contain alcohol dehydrogenase (ADH1) pro-
moter, invertase (SUC2) gene lacking its signal peptide, and
ADH1 terminator.
8. One Shot® TOP10 Electrocomp™ E. coli cells (Invitrogen):
Electrocompetent cells can be prepared manually (7, 8) or
obtained from commercial sources.
9. Random hexamer linker primer: 5¢-GAGAGAGAGAGAGA
GAACCGCGCGGCCGCCNNNNNN-3¢ (Not I restriction
enzyme site, underlined).
10. Yeast (Saccharomyces cerevisiae) strain(s): DBYa2445 (Mata,
suc2D9, lys2-801, ura3-52, ade2-101). The host yeast strain
carries a mutation which ensures that the endogenous SUC2
gene is not expressed (6).
11. Dimethyl sulfoxide (DMSO).
12. Glass beads 425–600 mm, acid-washed (Sigma).
13. Luria–Bertani (LB) medium (broth and agar plate): 10 g/L
of tryptone, 5 g/L yeast extract, 10 g/L NaCl, 20 g/L agar
258 Lee and Rose

(for plates only). Adjust pH to 7.0 with NaOH solution. Add

deionized H2O to a final volume of 1 L. Autoclave.
14. LB-ampicillin medium (150 × 15 mm plates): 10 g/L tryp-
tone, 5 g/L yeast extract, 10 g/L NaCl, 20 g/L agar,
100 mg/L ampicillin. Adjust pH to 7.0 with NaOH solution.
Add deionized H2O to a final volume of 1 L. Autoclave.
15. Microcentrifuge tubes and conical tubes.
16. MicroPulser Electroporator (Bio-Rad).
17. Petri dishes.
18. 3 M sodium acetate.
19. 70% (v/v) ethanol.
20. Restriction enzymes (Promega).
21. Sucrose selection medium: 20 g/L Difco peptone, 10 g/L
yeast extract, 20 g/L agar (for plates only). Adjust the pH to
6.5 with HCl solution if necessary. Add deionised H2O to a
final volume of 950 mL. Autoclave. Cool to 55°C. Add
sucrose (filter-sterilized) to 2% (v/v) by adding 50 mL of a
40% stock solution.
22. T4 DNA ligase (Promega).
23. TE-LiAc solution: Prepare fresh immediately prior to yeast trans-
formation by combining 0.2 mL of 10× TE with 0.2 mL of 10×
LiAc. Bring the total volume to 2 mL using sterile H2O.
24. TE-LiAc-PEG solution: Prepare fresh immediately prior to
yeast transformation by adding 1 mL of 10× TE, 1 mL of 10×
LiAc and 8 mL of 50% (w/v) PEG 3350 to make the final
volume of 10 mL.
25. Yeast lysis buffer: 2% (v/v) Triton X-100, 1% (w/v) SDS,
100 mM NaCl, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA.
26. YEASTMAKER™ Yeast Transformation System 2 (BD
Biosciences) contains components (10× LiAc, 10× TE, 50%
PEG, and Herring testes carrier DNA, denatured (10 mg/
mL)) for yeast transformation.
27. YPD medium: Dissolve 20 g/L peptone and 10 g/L yeast
extract. Adjust the pH to 6.5 with HCl solution if necessary.
Add deionized H2O to a final volume of 950 mL. Autoclave
and cool to 55°C. Add glucose (filter-sterilized with 0.2-mm
membrane filter) to 2% (v/v) by adding 50 mL of a 40% stock
solution.
28. Phenol/chloroform/isoamyl alcohol 25:24:1 (v/v/v).

2.2. In Planta Transient 1. Nicotiana benthamiana.

Expression of Secreted
2. Agrobacterium tumefaciens strains such as C58C1 (9) and
Protein Genes (NIP
GV3101 (10).
screen)
 Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 259

3. 20 mM CaCl2
4. Buffer for storing Agrobacterium competent cells: 20 mM
CaCl2 containing 20% glycerol. Sterilize and reduce tempera-
ture to ice-cold before use.
5. Liquid nitrogen for rapid freezing.
6. Microcentrifuge tubes.
7. Conical tubes.
8. Infiltration medium: 10 mM MES pH 5.6, 10 mM MgCl2
and 200 mM acetosyringone (11).
9. Agrobacterium culture medium: Luria–Bertani (LB) (see
Subheading 2.1).
10. Cloning vectors: pART-NIPF (see Note 2), pART-NIPM (see
Note 3), pART-XTH2SP:NIPM (see Note 4), and pBIN61S-
CymRSV p19 (see Note 5).
11. 1-mL syringe without needle.

2.3. Confocal 1. Onion (Allium cepa L.)

Fluorescence Imaging 2. PDS-1000/He biolistic transformation system (Bio-Rad).
for Plant Cell Wall
3. Confocal laser-scanning microscope equipped with helium/
Localization
neon lasers.
4. Cloning vector(s): pART-GFP (see Note 6) and pART-
PGSP:tdTOMATO (see Note 7).
5. M-10 tungsten particles.
6. Forceps.
7. 50% Glycerol.
8. Isopropanol for soaking rupture disk.
9. Microcentrifuge tubes.
10. 70% ethanol.
11. 100% ethanol.
12. 0.8 M Sucrose.
13. Petri dishes, 100 × 15 mm.
14. Whatman 3MM chromatography paper.
15. 2.5 M CaCl2.
16. 0.1 M Spermidine.

3. Methods

3.1. Yeast Secretion This screen is based on the principal of generating a library of
Trap Screen cDNAs fused to the N-terminus of invertase, and then transforming
the library into an invertase-deficient yeast mutant (6). The
260 Lee and Rose

transformed yeast mutant collection is then spread on plates of

sucrose, and any yeast colony that grows must have been trans-
formed with a gene encoding a secreted protein, the fusion of
which to invertase rescued the phenotype. This represents a rapid
means to identify large number of genes encoding secreted
proteins.

3.1.1.Construction 1. Isolate polyadenylated mRNA from total RNA using an

of the YST cDNA Library Oligotex mRNA Mini Kit, according to the manufacturer’s
instructions.
2. Synthesize the first- and second-strand cDNAs from the
mRNA (5 mg) using a random hexamer linker primer con-
taining a NotI restriction site for a directional cloning using a
cDNA Synthesis Kit (Stratagene) according to the manufac-
turer’s instructions.
3. Ligate the EcoRI adapters (provided with the Stratagene
cDNA Synthesis Kit) to the blunted cDNAs and ­phosphorylate
the EcoRI ends according to the manufacturer’s instructions.
4. Purify cDNA using a QIAquick PCR Purification Kit (Qiagen)
(see Note 8).
5. Add 5 mL of 10× NotI buffer and 5 mL of NotI (10 U/mL) to
digest 40 mL of the cDNAs. Incubate the reaction for 2 h at
37°C.
6. Run a 1% agarose gel to fractionate the cDNA fragments and
excise a gel slice corresponding to the region from approxi-
mately 0.3–1.0 kb, as judged using DNA ladder molecular
weight markers.
7. Purify the excised cDNAs with a QIAquick Gel Extraction
Kit (Qiagen).
8. Ligate the cDNAs into the EcoRI and NotI sites of the YST
vector(s) by adding the eluted cDNA (~20 ng), 1 mL of 10×
ligase buffer, 1 mL of T4 DNA ligase and the pYST vectors
digested with EcoRI and NotI (~200 ng, comprising equal
quantities of the pYST-0, -1 and -2 vectors), to a final vol-
ume of 10 mL. Incubate the reaction overnight at 16°C (see
Note 9).
9. Transform 2 mL of the ligation mix into 50 mL of a TOP10
electrocompetent E. coli strain (Invitrogen) by electropora-
tion using a MicroPulser (Bio-Rad) according to the manufac-
turer’s instructions. Immediately add 1 mL of SOC medium
(provided with TOP10 electrocompetent cells) to the tube to
recover the transformed cells. Incubate for 1 h at 37°C.
Plate 1 mL of the transformation mixture onto an agar plate
supplemented with ampicillin (100 mg/L). Incubate the plate
for 16–18 h at 37°C. Count the colonies (see Note 10).
 Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 261

10. Transform the remaining 8 mL ligation mixture into E. coli in

more electroporation reactions (see Note 11).
11. Plate 250 mL of the transformed E. coli cells onto each
LB-ampicillin agar plate (150 × 15 mm) at a density of roughly
5 × 104 CFU/plate and incubate the plates for 16–18 h at
37°C.
12. Add 20 mL of LB liquid medium per plate to harvest the
transformants from each plate and incubate at 37°C for 3 h
with gentle shaking.
13. Pool the overlaying LB cultures in a sterile 250-mL flask (see
Note 12).
14. Isolate the library plasmids from 30–50 mL of the bacterial
LB cultures according to the manufacturer’s instructions as
described for the Perfectprep Plasmid Midi Kit.
15. Quantify the amount of the library plasmid DNA using an
UV spectrophotometer and store at −20°C.

3.1.2. Yeast Transformation 1. Inoculate 1 mL of YPD liquid medium in a 1.5-mL micro-

centrifuge tube with one or two 2–3 m yeast colonies and
vortex the culture vigorously to disperse any clumps.
2. Transfer the cells to 50 mL of YPD in a 250-mL flask.
3. Incubate the culture at 30°C for 16–18 h with shaking at
250 rpm.
4. Check the OD at 600 nm using 1 mL of the culture every
30 min until OD at 600 nm reads 1.2.
5. Transfer the 50-mL culture into 300 mL of YPD in a 1-L
flask to produce 0.2–0.3 OD at 600 nm.
6. Incubate the culture for 3 h at 30°C, shaking at 250 rpm
until OD (600 nm) reaches 0.4–0.6.
7. Centrifuge the cells at 4,500 × g for 5 min at room
temperature.
8. Discard the supernatant and resuspend the cells by vortexing
in 50 mL of sterile H2O.
9. Centrifuge the cells at 4,500 × g for 5 min at room
temperature.
10. Discard the supernatant and resuspend cells in 1.5 mL of
freshly prepared TE-LiAc solution.
11. In a 50-mL tube, add 50 mg of cDNA library plasmids and
2 mg of carrier DNA (herring testes DNA or salmon sperm
DNA) and mix well.
12. In 1.5-mL tubes, add 100 ng of control plasmids (positive: a
secreted protein gene in-frame with suc2; and negative: a suc2
fusion construct lacking a signal peptide) and 0.1 mg of carrier
262 Lee and Rose

DNA (herring testes DNA or salmon sperm DNA) and mix

well (see Note 13).
13. Add 1 mL of yeast competent cells to the tube containing the
library plasmids (or 0.1 mL to control tubes) and mix well by
vortexing.
14. Add 6 mL of TE-LiAc-PEG solution to the tube containing
the library plasmids (or 0.6 mL to control tubes) and mix
well by vortexing.
15. Incubate at 30°C for 30 min with shaking at 200 rpm.
16. Add 700 mL of DMSO to the tube containing the library
plasmids (or 70 mL to control tubes) and mix well gently (see
Note 14).
17. Heat-shock the samples by incubating for 15 min in a 42°C
water bath and swirl gently every 5 min.
18. Place the tubes on ice for 5 min.
19. Centrifuge the cells at 4,500 × g for 5 min (or 15 s for con-
trols) at room temperature.
20. Remove the supernatant and resuspend cells in 5 mL (or
0.5 mL for controls) of 1× TE buffer.
21. Spread 200 mL of the transformation mixture onto a
150 × 15 mm plate containing the sucrose selection medium
(or 100 mL onto a 100 × 15 mm plate for controls).
22. Incubate plates at 30°C for 6–10 days (Fig. 2, see Note 15).

3.1.3. Identification of 1. Streak all the growing yeast transformants on the sucrose
cDNA Clone Encoding a selection medium and incubate at 30°C for 2 days.
Putative Signal Peptide 2. Inoculate yeast cells from the streaked plate with a sterile
spatula into 2 mL of sucrose selection medium and incubate
at 30°C for 16–18 h.

Fig. 2. Yeast secretion trap (YST) system to identify secreted proteins. The
invertase-deficient yeast mutant can be rescued when the cDNA encodes a secreted
protein fused with the invertase (SP:suc2, as shown at left), but the mutant yeast cells
are unable to grow on a sucrose selection medium because the invertase is not secreted
(suc2, as shown at right).
 Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 263

3. Transfer the culture into a 2.0-mL microcentrifuge tube and

centrifuge the cells at 14,000 × g for 2 min and discard the
supernatant.
4. Resuspend the pellet in 0.2 mL of yeast lysis buffer with 0.2 g
of glass beads by vortexing vigorously for 2 min.
5. Add 0.2 mL of phenol/chloroform/isoamyl alcohol (25:24:1
(v/v/v)) and vortex vigorously for 2 min (see Note 16).
6. Centrifuge at 14,000 × g for 5 min and transfer the upper
aqueous layer into a new tube.
7. Precipitate the DNA by adding 20 mL of 3 M sodium acetate
and 500 mL of absolute ethanol and mix well by vortexing.
Precipitate the DNA overnight at −20°C.
8. Centrifuge at 14,000 × g for 10 min to pellet the DNA.
9. Wash the DNA pellet by adding 0.5 mL of 70% (v/v) ethanol
and spin at 14,000 × g for 2 min at room temperature.
10. Resuspend the DNA pellet in 35 mL of sterile H2O or TE
buffer (see Note 17).
11. Transform 1 mL of DNA into electrocompetent E. coli cells
by electroporation and select transformants on LB plates sup-
plemented with the ampicillin.
12. Identify colonies that contain the YST vector by preparing
miniprep DNA from colonies on the LB-ampicillin agar plates
with a QIAprep Spin Miniprep Kit (Qiagen) and by restric-
tion enzyme digestion.
13. Perform DNA sequencing using a specific primer (5¢-TCCT
CGTCATTGTTCTCGTTCC-3¢) derived from YST vector
to confirm in-frame fusion between the cDNA and the suc2
invertase gene.
14. Signal peptides of the translated amino acid sequences can be
predicted computationally using publicly available signal pep-
tide prediction programs such as SignalP version 3.0 (www.
cbs.dtu.dk/services/SignalP)(12).

3.2. In Planta Transient This screen is somewhat similar to the YST screen described
Expression Screen of above, except that a “necrosis inducing protein” (NIP) is used as
Secreted Protein a marker, rather than invertase. NIP acts as a toxin and is known
Genes (NIP Screen) to induce rapid necrosis in compatible plant species when local-
ized in the apoplast, but not in the cytosol (13). Briefly, the NIP
screen works by fusing a cDNA population to the N-terminus of
a NIP gene. The NIP-fusion library is then transformed into
A. tumefaciens and a library spread among 96-well plates. Media
containing the individual A. tumefaciens transformants is then
flooded into the apoplast of N. benthamiana leaves. The presence
of a secreted NIP fusion protein will then be evident by a rapid
cell death/necrosis phenotype.
264 Lee and Rose

3.2.1. Preparation of 1. Streak out Agrobacterium strain from glycerol stock stored at
Agrobacterium Competent −80°C onto LB agar plate supplemented with the appropriate
Cells antibiotic and incubate the preculture for 3 days at 28°C.
2. Inoculate a single colony into 2 mL of LB liquid medium and
incubate the culture overnight at 28°C under constant agita-
tion at 250 rpm.
3. Transfer 1 mL of culture to inoculate 50 mL of LB liquid
medium in a 250-mL flask.
4. Incubate the culture at 28°C under constant agitation at
250 rpm to an OD (600 nm) of 0.5–1.0.
5. Chill the culture on ice for 5 min and harvest cells in pre-
chilled 50-mL tube by centrifugation for 10 min at 4,500 × g
at 4°C.
6. Resuspend the cells in 1 volume of ice-cold, sterile 20 mM
CaCl2.
7. Repeat Step 5.
8. Resuspend the cells in 1 mL of ice-cold, sterile 20 mM
CaCl2/20% glycerol.
9. Freeze 50 mL aliquots in 1.5-mL microfuge tubes in liquid
nitrogen and store at −80°C until use.

3.2.2. Agrobacterium 1. Thaw Agrobacterium competent cells on ice.

Transformation 2. Add plasmid DNA (0.5–1.0 mg, 5 mL of standard E. coli mini-
prep DNA) in a 1.5-mL microfuge tube containing 50 mL of
Agrobacterium competent cells.
3. Mix them together and place the tube on ice for 10 min.
4. Freeze the cells in liquid nitrogen for 5 min.
5. Place the tube on the benchtop at room temperature for
10 min.
6. Add 1 mL of LB to the tube.
7. Incubate for 2–4 h at 28°C with gentle agitation.
8. Plate the transformed cells (100–200 mL) on an LB agar plate
containing the appropriate antibiotic.
9. Incubate the cells for 3 days at 28°C.
10. Restreak the colonies on a new LB agar plate and incubate for
2 days at 28°C.
11. Grow 2 mL of culture from a single colony overnight at 28°C
under constant agitation at 250 rpm and confirm the ­presence
of the introduced plasmid DNA by PCR and/or sequence
analysis using miniprep DNA.
12. Store glycerol stocks at −80°C.
 Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 265

3.2.3. Agroinfiltration 1. Streak out Agrobacterium strain from glycerol stock stored at
−80°C onto LB agar plate supplemented with the appropriate
antibiotic and incubate the plate for 3 days at 28°C.
2. Inoculate a single colony into 2 mL of LB liquid medium and
incubate the culture overnight at 28°C under constant agita-
tion at 250 rpm.
3. Harvest cells by centrifugation at 4,500 × g for 15 min at
room temperature.
4. Resuspend the cell pellets in 1 volume of infiltration medium
(10 mM MES pH 5.6, 10 mM MgCl2 and 200 mM acetosy-
ringone) and incubate for 1 h at room temperature.
5. Dilute the culture for infiltration to an OD (600 nm) of 0.3
before infiltration.
6. Infiltrate Agrobacterium suspension into the abaxial side of
the leaves of 4–6-week-old N. benthamiana plants using a
1-mL syringe without needle.
7. Incubate infiltrated plants in the growth chamber in constant
light at 24°C for 3–5 days.
8. Photographs can be taken at 5 days post infiltration (Fig. 3,
see Note 18).

3.3. Confocal Once a candidate cell wall protein has been identified, it may be
Fluorescence Imaging advisable to confirm its extracellular localization using an orthog-
for Plant Cell Wall onal approach, such as examining the localization of a fluorescent
Localization protein marker fused to the candidate protein. A rapid approach
using transient expression of the fusion protein in onion epidermal
cells is described below, although stable transformation is another

Fig. 3. Transient expression of the necrosis-inducing protein (NIP) in agroinfiltrated Nicotiana

benthamiana. Leaves of N. benthamiana plants were infiltrated with A. tumefaciens
­carrying pART-NIPF (1), pART-NIPM (2), pART-XTH2SP:NIPM (3), and no plasmid DNA (4). The
P19 protein can be co-expressed for high-level transient expression of a protein of interest
in whole plants.
266 Lee and Rose

option that may, in some cases, be more robust. An example of

this approach is previously described (14).

3.3.1. Preparation 1. Weigh out tungsten particles (30 mg) and transfer to a
of Tungsten Particles microcentrifuge.
2. Add 1 mL of 70% ethanol.
3. Vortex for 10–15 min.
4. Centrifuge the particles at 14,000 × g for 10 s and remove
supernatant.
5. Repeat 70% ethanol wash twice. Pellet the particles by brief
centrifugation between washes.
6. Add 1 mL of sterile H2O, vortex for 1 min, let settle for
1 min, and then spin briefly (~2 s).
7. Repeat sterile H2O wash steps three times.
8. Remove supernatant. Add 500 mL of 50% glycerol. Vortex
briefly.
9. Freeze at −20°C until use.

3.3.2. Preparation of Onion 1. Remove the outer papery layers of a yellow or white onion
Tissues purchased at a local grocery store. Cut off the dry tips.
2. Split the onion lengthwise into quarters.
3. Discard two outermost layers of the leaves.
4. Cut the third outermost layer of leaves into slices of ~1.5 cm2
and place on a piece of Whatman paper moisturized with
5 mL of sterile H2O on the centre of a Petri dish. Bombardment
will occur on the epidermal peel on the inner side of an onion
piece.
5. Close the lid to keep from drying out.

3.3.3. Coating Microcarriers 1. Place plasmid DNA (0.5–1.0 mg, 5 mL of standard E. coli
miniprep DNA) in a 1.5-mL microcentrifuge tube. Total
­volume should be 26 mL (see Note 19).
2. Add 10 mL of tungsten particles and mix well (see Note 20).
3. Add 30 mL of 2.5 M CaCl2 and mix well.
4. Add 12.5 mL of 0.1 M spermidine and mix well.
5. Incubate for 20 min on ice with occasional vortexing.
6. Spin briefly (~10 s) and remove supernatant.
7. Add 300 mL of 70% ethanol and vortex.
8. Spin briefly (~10 s) and remove supernatant.
9. Add 300 mL of 100% ethanol and vortex.
10. Spin briefly (~10 s) and remove supernatant.
11. Resuspend the pellet in 30 mL of 100% ethanol and vortex
(see Note 21).
 Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 267

12. Place 15 mL onto a macrocarrier by spreading out the particles

uniformly. Allow the ethanol to evaporate.

3.3.4. Bombardment 1. Start vacuum pump.

2. Open helium tank.
3. Screw in adjustment handle until the pressure of the helium
tank is set at 1,300 psi.
4. Switch on the gene gun (left switch of three red buttons).
5. Load a rupture disk (1,100 psi-rated, soaked in isopropanol)
in the rupture disk retaining cap at the top of chamber (see
Note 22).
6. Place a stopping screen (convex side up) in the bottom of
macrocarrier assembly.
7. Invert a macrocarrier holder with DNA-coated macrocarrier
(tungsten particles should be facing down) and place over the
screen.
8. Screw the macrocarrier cover lid back on.
9. Slide the macrocarrier launch assembly into the second shelf
from the top.
10. Place the Petri dish with onion sample on the target plate
shelf and slide into the fourth shelf from the top.
11. Close the chamber door.
12. Pull vacuum (upper position on three-way middle switch of
three red buttons) until the vacuum reaches 27.5 in. Hg.
Hold the vacuum by quickly pressing the three-way switch all
the way down.
13. Press the fire switch (right switch of three red buttons) and
hold it down. Pressure will rise and the rupture disk will burst
when pressure reaches 1,100 psi.
14. Release vacuum.
15. Open the chamber door and remove sample.
16. Seal Petri dish with Parafilm®. Incubate in the dark for
16–24 h at 24°C.
17. Shut down the particle bombardment system.
(a) Close the helium tank valve.
(b) Close the chamber.
(c) Pull vacuum.
(d) Press the fire switch until the pressure on helium regula-
tor gauge reads zero.
(e) Release vacuum.
(f) Loosen pin until it turns freely.
(g) Turn off gene gun and vacuum pump.
268 Lee and Rose

Fig. 4. Transient expression of a secreted tdTOMATO in onion epidermal cells. GFP (as shown on left) was co-expressed
with the PGSP:tdTOMATO (as shown in middle) in order to differentiate between cytoplasm and cell wall. A merged image
(right) and the scale bar representing 50 mm are shown.

3.3.5. Confocal Imaging 1. Peel off the thin epidermis layer carefully with forceps.
2. Mount the epidermal cells in 0.8 M sucrose solution under a
coverslip for plasmolysis (see Note 23).
3. Incubate for 10–20 min at room temperature.
4. Observe fluorescent signals using a confocal laser-microscope
(Fig. 4, see Note 24).

4. Notes

1. Other similar commercial kits would be adequate and could

be used as an alternative.
2. To construct pART-NIPF, the full-length Phytophthora sojae
NIP coding sequence (NCBI accession number AAM48170)
(13) was amplified from P. sojae race 1 genomic DNA with
primers NIP1K (5¢-gcgcgcGGTACCCCATGAACCTCCG
CCCTGCA-3¢) and NIP3H (5¢-gcgcgcAAGCTTTTAAG
CGTAGTAGGCGTTGCC-3¢). The plasmid pKANNIBAL1-
NIP was constructed by ligating the KpnI/HindIII-digested
PCR product into KpnI/HindIII-digested pKANNIBAL
(15). The pKANNIBAL1-NIP construct was subcloned as
NotI fragments into pART27 (16) to generate pART-NIPF.
3. To construct pART-NIPM, the coding sequence of the mature
form of P. sojae NIP was PCR-amplified from the full-length
P. sojae NIP clone with primers NIP2K (5¢-gcgcgcGGTACC
CCATGAGCGTTATCAACCACGAC-3¢) and NIP3H. The
plasmid pKANNIBAL2-NIP was constructed by ligating the
KpnI/HindIII-digested PCR product into KpnI/HindIII-
digested pKANNIBAL. The pKANNIBAL2-NIP construct
was subcloned as NotI fragments into pART27. A new SgfI
restriction site was introduced into the cloning site upstream
of the NIP coding sequence to generate pART-NIPM.
Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 269

4. To construct pART-XTH2SP:NIPM, the sequence encoding

the signal peptide of tomato SlXTH2 (17), which is one of
the xyloglucan endotransglycosylase/hydrolase genes (NCBI
accession number AF176776), was amplified from the
­full-length SlXTH2 cDNA with primers XTH2E (5¢-gcgcgc
GAATTCATGATCAAAACATCAAGT-3¢) and XTH2S
(5¢-gcgcgcGCGATCGCCCAAAAGCCACCACTAC-
GAA-3¢). The EcoRI/SgfI-digested PCR product was cloned
into EcoRI/SgfI-digested pART-NIPM to generate pART-
XTH2SP:NIPM.
5. The pBIN61S-CymRSV P19 vector was kindly provided by
Dr. J. Burgyán (Hungary). Tombusvirus P19 has been shown
to suppress posttranscriptional gene silencing (18–20). The
P19 protein may be useful for high level of transient expres-
sion of proteins in planta (21).
6. To construct pART-GFP vector, the intron in pHANNIBAL
(15) was excised by KpnI/HindIII digestion and replaced with
the KpnI/HindIII-digested smGFP open reading frame from
pSMGFP (22). The NotI fragment (containing the CaMV 35S
promoter, GFP coding sequence and OCS terminator) of the
pHANNIBAL-GFP vector was cloned into the NotI site of the
binary vector pART27 to produce pART-GFP.
7. To construct pART-PGSP:tdTOMATO, the GFP open
­reading frame of pART-GFP was firstly replaced with the
KpnI/XbaI-digested tdTOMATO (23) amplified using the
primers TOM5 (5¢-gcgcgcGGTACCATGGTTTCCAAG
GGTGAG-3¢) and TOM3 (5¢-gcgcgcTCTAGATTACTTG-
TACAACTCGTC-3¢) using the plasmid DNA (LAT52_
tdTOMATO_NOS in pLAT1 vector; kindly provided by
Dr. B.A. McClure) as a template to generate pART-tdTO-
MATO. Secondly, the sequence encoding the signal peptide
of tomato PG-2a (NCBI accession number P05117) was
amplified from the full-length PG-2a cDNA with primers
PG1 (5¢-gcgcgcCTCGAGATGGTTATCCAAAGGAAT-3¢)
and PG2 (5¢-gcgcgcGGTACCGCTTCTACAAGTTGA
AAT-3¢). Then, the XhoI/KpnI-digested PCR products of
tomato PG-2a signal peptide-encoding region were cloned
into XhoI/KpnI-digested pART-tdTOMATO to produce
pART-PGSP:tdTOMATO.
8. We follow the protocol provided with the QIAquick PCR
Purification Kit by adding 5 volumes of buffer PB (provided
in the kit) to 1 volume of the reaction. Ensure that 40 mL of
elution buffer are added to the column to elute the cDNAs
for the subsequent reactions.
9. Shorter (<0.3 kb) or longer (>1.0 kb) cDNA fragments are
likely to contain stop codons, which could disrupt the in-
frame gene fusion with the invertase gene.
270 Lee and Rose

10. T4 DNA ligase buffer contains ATP, which degrades rapidly.

Avoid multiple freeze-thaw cycles by making smaller aliquots
of the buffer and storing them at −80°C.
11. We recommend commercially prepared electrocompetent
cells as these typically provide a high transformation efficiency
of >1 × 109 colony forming units (CFU)/mg. Approximately
200 colonies should be obtained from 1 mL of the transfor-
mation mixture.
12. The majority (>70% recombinants) of clones in the cDNA
library should contain cDNA inserts. Approximately 50
clones from the E. coli library should be selected at random to
check the percentage of recombinant clones, and the average
insert size of the cDNAs using the plasmid isolated according
to the manufacturer’s instructions as described for the
QIAprep Spin Miniprep Kit.
13. Yeast competent cells should be used immediately after
preparation.
14. Do not vortex as the yeast cells are more fragile.
15. Yeast transformants expressing the positive control (SP:suc2)
should grow on the sucrose selection medium, while yeast
expressing the same protein without the predicted SP should
not grow. The first colonies should appear after approximately
4–5 days although the slow-growing transformants may con-
tinue to appear until 10 days.
16. This should be handled under a fume hood. Wear safety
glasses, a lab coat, and gloves.
17. The DNA pellet should be briefly dried before adding sterile
H2O or TE buffer to remove residual ethanol.
18. The smaller size of the fusion protein can help maintain the
activity of NIP protein, as the fusion protein may abolish the
NIP activity.
19. This prepares enough particles for two shots.
20. Make sure that the tungsten particles are completely
resuspended.
21. The DNA-coated tungsten particle suspension can be briefly
sonicated.
22. Make sure that the rupture disk retaining cap should be firmly
tightened.
23. Onion epidermal cells can be bathed in a range of sucrose
solutions (e.g., 0.4–1.0 M).
24. It is important to note that onion epidermal cells should be
co-bombarded with plasmids expressing GFP (or GFP-ER)
and tdTOMATO for a dual-colour fluorescence imaging.
 Characterization of the Plant Cell Wall Proteome Using High-Throughput Screens 271

Acknowledgements

Research in this area was supported by grants from the NSF Plant
Genome Program (DBI-0606595), the New York State Office of
Science, Technology and Academic Research (NYSTAR) and
Cornell Center for a Sustainable Future (CCSF).

References
1. Lee, S.-J., Saravanan, R. S., Damasceno, C. 11. Bendahmane, A., Querci, M., Kanyuka, K.,
M. B., Yamane, H., Kim, B. -D. and Rose, J. and Baulcombe, D. C. (2000) Agrobacterium
K. C. (2004) Digging deeper into the plant transient expression system as a tool for the
cell wall proteome. Plant Physiol Biochem 42, isolation of disease resistance genes: applica-
979–988. tion to the Rx2 locus in potato. Plant J 21,
2. Jamet, E., Albenne, C., Boudart, G. Irshad, 73–81.
M., Canut H., and Pont-Lezica., R. (2008) 12. Bendtsen, J. D., Nielsen, H., von Heijne, G.,
Recent advances in plant cell wall proteomics. and Brunak, S. (2004) Improved prediction
Proteomics 8, 893–908. of signal peptides: SignalP 3.0. J Mol Biol 340,
3. Rose, J. K. C., Bashir, S., Giovannoni, J. J., 783–795.
Jahn, M. M., and Saravanan, R. S. (2004) 13. Qutob, D., Kamoun, S., and Gijzen, M.
Tackling the plant proteome: practical (2002) Expression of a Phytophthora sojae
approaches, hurdles, and experimental tools. necrosis inducing protein occurs during tran-
Plant J 39, 715–733. sition from biotrophy to necrotrophy. Plant J
4. Nickel, W. (2005) Unconventional secretory 32, 361–373.
routes: direct protein export across the plasma 14. Yamane, H., Lee, S. -J., Kim, B. -D., Tao, R.,
membrane of mammalian cells. Traffic 6, and Rose, J. K. C. (2005) A coupled yeast sig-
607–614. nal sequence trap and transient plant expres-
5. Nombela, C., Gil, C., and Chaffin, W. L. sion strategy to identify and confirm genes
(2006) Non-conventional protein secretion in encoding extracellular proteins from peach
yeast. Trends Microbiol 14, 15–21. pistils. J Exp Bot 56, 2229–2238.
6. Lee, S. -J., Kim, B. -D., and Rose, J. K. C. 15. Wesley, V. S., Helliwell, C., Smith, N. A.,
(2006) Identification of eukaryotic secreted and Wang, M. B., Rouse, D., Liu, Q., Gooding,
cell surface proteins using the yeast secretion P.S., Singh, S.R., Abbott, D., Stoutjesdijk, A.,
trap (YST) screen. Nat Protoc 1, 2439–2447. Robinson, S.P., Gleave, A. P., Green, A. G.,
7. Ausubel, F. M., Brent, R., Kingston, R. E., and Waterhouse, P. M. (2001) Construct
Moore, D. D., Seidman, J. G., Smith, J. A., design for efficient, effective and high-
and Struhl, K. (1987) (eds.) Current Protocols throughput gene silencing in plants. Plant J
in Molecular Biology, Wiley, New York. 27, 581–590.
8. Miller, E. M. and Nickoloff, J. A. (1995) 16. Gleave, A. P. (1992) A versatile binary vector
Escherichia coli electrotransformation, in system with a T-DNA organizational struc-
Methods in Molecular Biology 47, Nickoloff, J. ture conducive to efficient integration of
A., ed., Humana, Totowa, NJ, 105. cloned DNA into the plant genome. Plant
9. Koncz, C. and Schell, J. (1986) The promoter Mol Biol 20, 1203–1207.
of TL-DNA gene 5 controls the tissue-specific 17. Catalá, C., Rose, J. K. C., York, W. S.,
expression of chimaeric genes carried by a Albersheim, P., Darvill, A. G., and Bennett, A.
novel type of Agrobacterium binary vector. B. (2001) Characterization of a tomato xylo-
Mol Gen Genet 204, 383–396. glucan endotransglycosylase gene that is
10. Koncz, C., Kreuzaler, F., Kalman, Z., and down-regulated by auxin in etiolated hypoco-
Schell, J. (1984) A simple method to transfer, tyls. Plant Physiol 127, 1180–1192.
integrate and study expression of foreign 18. Qiu, W., Park, J. -W., and Scholthof, H .B.
genes, such as chicken ovalbumin and a-actin (2002) Tombusvirus P19-mediated suppres-
in plant tumors. EMBO J 3, 1029–1037. sion of virus-induced gene silencing is
272 Lee and Rose

c­ ontrolled by genetic and dosage features that 21. Voinnet, O., Rivas, S., Mestre, P., and
influence pathogenicity. Mol Plant Microbe Baulcombe, D. C. (2003) An enhanced tran-
Interact 15, 269–280. sient expression system in plants based on sup-
19. Qu, F. and Morris, T. J. (2002) Efficient pression of gene silencing by the p19 protein of
infection of Nicotiana benthamiana by tomato bushy stunt virus. Plant J 33, 949–956.
Tomato bushy stunt virus is facilitated by the 22. Davis, S. J. and Vierstra, R. D. (1998) Soluble,
coat protein and maintained by p19 through highly fluorescent variants of green fluores-
suppression of gene silencing. Mol Plant- cent protein (GFP) for use in higher plants.
Microbe Interact 15, 193–202. Plant Mol Biol 36, 521–528.
20. Silhavy, D., Molnár, A., Lucioli, A., Szittya, 23. Shaner, N. C., Campbell, R. E., Steinbach, P.
G., Hornyik, C., Tavazza, M and Burgyán, J. A., Giepmans, B. N., Palmer, A. E., and Tsien,
(2002) A viral protein suppresses RNA silencing R. Y. (2004) Improved monomeric red,
and binds silencing-generated, 21- to orange and yellow fluorescent proteins derived
25-nucleotide double-stranded RNAs. EMBO from Discosoma sp. red fluorescent protein.
J 21, 3070–3080. Nat Biotechnol 22, 1567–1572.
 Chapter 19

Knocking Out the Wall: Protocols for Gene Targeting

in Physcomitrella patens
Alison W. Roberts, Christos S. Dimos, Michael J. Budziszek Jr.,
Chessa A. Goss, and Virginia Lai

Abstract
The moss Physcomitrella patens has become established as a model for investigating plant gene function
due to the feasibility of gene targeting. The chemical composition of the P. patens cell wall is similar to
that of vascular plants and phylogenetic analyses of glycosyltransferase sequences from the P. patens
genome have identified genes that putatively encode cell wall biosynthetic enzymes, providing a basis for
investigating the evolution of cell wall polysaccharides and the enzymes that synthesize them. The pro-
tocols described in this chapter provide methods for targeted gene knockout in P. patens, from construct-
ing vectors and maintaining cultures to transforming protoplasts and analyzing the genotypes and
phenotypes of the resulting transformed lines.

Key words: Gene targeting, Immunocytochemistry, Physcomitrella patens, Gateway® cloning

1. Introduction

The moss Physcomitrella patens has become established as a model

for investigating plant gene function due to its uniquely high
rate of homologous recombination, which enables targeted gene
modification (1). Wild type alleles can be knocked out by target-
ing with vectors containing homologous sequences interrupted
by a selection cassette (2). Targeted mutagenesis can be accom-
plished by replacing genes with vectors carrying insertions, dele-
tions, or point mutations (3). Functional transgenes can be
inserted using expression vectors targeted to intergenic regions
(4). Translational fusions can also be created at the native locus
by gene targeting (5). A further advantage of P. patens is that
the dominant phase in its life-cycle is haploid, which enables

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_19, © Springer Science+Business Media, LLC 2011

273
274 Roberts et al.

detection of mutant phenotypes in primary transformants and

eliminates the need for backcrossing (6). Genomic resources for
P. patens include EST sequences and their corresponding full-
length cDNA clones (http://moss.nibb.ac.jp/), microarrays (7, 8),
and the annotated genome sequence http://www.cosmoss.org/
(9). Methods for gene silencing by RNAi (10) and miRNA (11)
and for producing temperature sensitive mutants (12) have also
been developed.
Several aspects of P. patens morphology and development
are advantageous for investigating cell wall biosynthesis and
evolution. Development of P. patens from haploid spores, pro-
toplasts, or fragmented tissue begins as photosynthetic chlo-
ronemal filaments that extend by apical division and tip growth
(13). Growth of chloronemal filaments can be maintained for
several weeks on medium containing ammonium tartrate. Thus,
in contrast to vascular plants, it is possible to produce tissue
consisting of a single cell type (chloronemal filaments) from
P. patens. This greatly simplifies the task of assigning changes in
cell wall composition to alterations in the expression of specific
genes and the activity of specific enzymes. In the absence of
ammonium tartrate, the apical cells of the chloronemal filaments
increase their growth rate to begin producing caulonemal fila-
ments (6). Caulonemal filaments produce buds that develop
into the familiar leafy portion of the moss plant, the gameto-
phore, which enlarges by diffuse growth. Zygotes derived from
fusion of gametes, produced at the gametophore apex, develop
into diploid sporophytes consisting of a short stalk and a
sporangium (13).
Like those of vascular plants, P. patens cell walls contain cel-
lulose, xyloglucan, mannan, pectins, and arabinogalactan pro-
teins (14–18). Like other bryophytes, they lack lignin (19, 20).
Phylogenetic analyses of glycosyltransferase sequences from the
P. patens genome have identified orthologs of genes that encode
cell wall biosynthetic enzymes in Arabidopsis and rice http://wiki.
genomics.purdue.edu/index.php/Cell_wall_composition_and_
glycosyltransferases_involved_in_cell_wall_formation (21) and
provide a basis for investigating the evolution of cell wall poly-
saccharides and the enzymes that synthesize them. Various
aspects of P. patens biology have been reviewed (1, 3, 6, 22–27)
and protocols for culture and transformation have been described
previously in print (28, 29) and online (http://moss.nibb.ac.
jp/, http://www.plant-biotech.net/; http://biology4.wustl.
edu/moss/methods.html). The following protocols provide a
guide for targeted gene knock out in P. patens, from construct-
ing a vector and maintaining cultures to transforming protoplasts
and analyzing the genotypes and phenotypes of the resulting
transformed lines.
 Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 275

2. Materials

2.1. Vector Construction 1. Sequence of interest as a query for searching the P. patens
genome.
2. Multisite Gateway® Pro 3.0 three-fragment or Multisite
Gateway® Pro Plus flexible cloning kit (Invitrogen, Carlsbad,
CA, USA).
3. Reagents for PCR.
4. Destination vector (available upon request from the authors).
5. Commercial plasmid DNA midi- or maxi-prep kit.
2.2. Culture of P. patens
1. Incubator set to 25°C with constant illumination at
50–80 mmol/m2/s.
2. Paint shaker (e.g., Tornado II, Blair Equipment Company,
Flint, MI, USA), optional.
3. BCDAT or BCD medium (see Table 1, Note 1).
4. Unvented Petri plates, 95 mm (Greiner Bio-One, Frickenhausen,
Germany, see Note 2).

Table 1
Media for culture and transformation of P. patens (29). Dry ingredients are added
to purified water, q.s. to 1 L, and sterilized by autoclaving. Calcium chloride
solution is added after autoclaving to prevent precipitation

BCD BCDAT PRMB PRMT PRML

(per L) (per L) (per L) (per L) (per L)
MgSO4 heptahydrate 0.25 g 0.25 g 0.25 g 0.25 g 0.25 g
KH2PO4 0.25 g 0.25 g 0.25 g 0.25 g 0.25 g
KNO3 1.0 g 1.0 g 1.0 g 1.0 g 1.0 g
FeSO4 septahydrate 12.5 mg 12.5 mg 12.5 mg 12.5 mg 12.5 mg
Diammonium tartrate – 0.92 g 0.92 g 0.92 g 0.92 g
Trace element solutiona 1 mL 1 mL 1 mL 1 mL 1 mL
Mannitol – – 60 g 80 g 80 g
Agar 7g 7g 8g 5g –
Calcium chloride solution 1 mL 1 mL 10 mL 10 mL 10 mL
55.5 g/L CaCl2 (sterile)
Add after autoclaving
55 mg/L cupric sulfate pentahydrate; 55 mg/L zinc sufate heptahydrate; 614 mg/L boric acid; 389 mg/L man-
a

ganous chloride tetrahydrate; 55 mg/L cobalt chloride hexahydrate; 28 mg/L potassium iodide; 25 mg/L sodium
molybdate dehydrate (32)
276 Roberts et al.

5. Sterile cellophane disks (Type 325P, AA Packaging Ltd.,

Lancashire, UK) interleaved with circles of copy paper and
autoclaved in glass Petri plates (see Note 3).
6. Inoculum of P. patens (Gransden, available from David Cove,
Washington University, St. Louis, MO, USA).
7. Sterile 15 mL disposable centrifuge tubes (polystyrene or
polypropylene).
8. Sterile 10 mL serological pipettes.
9. Sterile purified water.
10. Sterile stainless steel beads (3.2 mm, Biospec Products,
Bartlesville, OK, USA).

2.3. Transformation 1. Materials for subculture of P. patens (see above).

and Selection 2. Pipettors (20, 200, and 1,000 mL capacity) and sterile tips.
3. Heated water bath with thermometer.
4. Orbital shaker.
5. Hemocytometer.
6. 50 mg vector DNA: linearized, 1 mg/mL in sterile purified
water or 5 mM Tris, pH 7.5.
7. 8.5% (w/v) d-mannitol (sterilized by autoclaving).
8. 3 M solution: the following are added to 40 mL purified
water: 4.55 g d-mannitol, 750 mL of 1 M MgCl2, 5 mL of 1%
MES-KOH, pH 5.6; q.s. to 50 mL; filter sterilized; stored at
4°C for up to 6 months.
9. 2% Driselase solution: 1 g Driselase (Sigma, St. Louis, MO,
USA) is dissolved in 50 mL 8.5% d-mannitol; stirred gently
30 min 21–25°C; chilled 30 min 4°C; stirred 5 min 21–25°C;
centrifuged 2,500 × g 10 min; filter sterilized; stored as 3 mL
aliquots in sterile 15 mL centrifuge tubes at –20°C).
10. PEG solution: To prepare part 1: 9 mL of 8.5% d-mannitol is
combined with 1 mL of 1 M Ca(NO3)2 and 100 mL of 1 M
Tris-HCl, pH 8.0, and filter sterilized; to prepare part 2: 4 g
PEG 8000 (Sigma) is melted in a sterile 50 mL disposable
centrifuge tube by microwaving; Part 1 is added to Part 2 and
vortexed until completely mixed; kept at 21–25°C for 2 h
before use; stored in 1 mL aliquots at –20°C).
11. PRMB and PRMT media (see Table 1).
12. Sterile nylon filters (Cell Strainer, BD Falcon, Franklin Lakes,
NJ, USA).
13. Sterile 15 and 50 mL disposable conical centrifuge tubes and
15 mL round-bottom culture tubes.
14. Antibiotic stock: e.g., 15 mg/mL hygromycin in purified
water.
 Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 277

2.4. Genotype Analysis 1. Vector primers (Vector F = TGACAGATAGCTGGGCAATG,

by PCR Vector R = TCCGAGGGCAAAGAAATAGA) and flanking
primers (gene specific, see below).
2. Kontes Pellet Pestle® Micro Grinder (Kimble/Kontes,
Vineland, NJ, USA).
3. DNA extraction buffer for PCR: 0.2 M Tris-HCl, pH 9.0,
0.4 M LiCl, 25 mM EDTA, 1% SDS.
4. Isopropanol.
5. TE buffer: 10 mM Tris-HCl, pH 8, 1 mM EDTA.

2.5. Genotype Analysis 1. Transformed P. patens line cultured on BCDAT plate for
by Southern Blotting 7 days (see Subheading 3.2).
2. DNA extraction buffer for Southern blotting: the following
are combined in the order listed in a 15 mL disposable centri-
fuge tube: Solution 1: 350 mM sorbitol, 100 mM Tris–HCl,
pH 7.5, 5 mM EDTA, sterilized by autoclaving (1 mL);
3.8 mg sodium bisulfite; Solution 2: 200 mM Tris-HCl,
pH 7.5, 50 mM EDTA, 2% (w/v) CTAB, 2 M sodium chlo-
ride, sterilized by autoclaving (1 mL); 0.4 mL of 5% (w/v)
N-lauroylsarcosine; 5 mL of 10 mg/mL RNaseA; prepared
just before use.
3. Chloroform:octanol 24:1.
4. Cryocup Grinder (Research Products International,
Mt. Prospect, IL, USA).
5. Liquid nitrogen.
6. DNA precipitation buffer: 80 mL ethanol, 20 mL 1 M sodium
acetate, pH 7.0.
7. 70% ethanol.
8. Supplies and reagents for Southern blotting.

2.6. Immuno­ 1. 1 mL pipette tips, tips cut off with a razor blade at 1 cm,
fluorescent Cell Wall autoclaved.
Labeling 2. Primary antibodies or carbohydrate binding modules (CBM)
(see Plant Probes www.plantprobes.net; Biosupplies Australia
www.biosupplies.com.au).
3. Secondary antibodies (ALEXA Fluor® 488 labeled goat anti-
rat or antimouse as appropriate for primary antibody,
Invitrogen).
4. FlexiPERM® 12 well chambers (ISC BioExpress, Kaysville,
UT, USA).
5. Poly-L-lysine coated slides (Fisher Scientific).
6. Sterile nylon filters (see Subheading 2.3, step 12).
7. 15 and 50 mL disposable centrifuge tubes.
278 Roberts et al.

8. 2× fixative stock: 100 mM PIPES, pH 6.8, 5 mM magnesium

sulfate, 10 mM EGTA.
9. Fixative: 500 mL 2× fixative stock, 270 mL purified water,
230 mL 16% formaldehyde (methanol-free, Polysciences Inc.,
Warrington, PA, USA), prepared just before use.
10. Phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 mM
KCl, 10 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.4.
11. Blocking solution: 5% w/v nonfat dry milk in PBS, prepared
just before use.
12. Coverslips (24 × 60 mm).
13. Antifade reagent (SlowFade® Gold, Invitrogen).

2.7. Morphological
Analysis
(see Subheading 2.2)

3. Methods

3.1. Preparation of Replacement vectors for deleting target genes include homolo-
Knockout Vectors gous sequences upstream and downstream of the target gene
coding sequence, separated by a selection cassette (Fig. 1). The
selection cassette replaces the target gene when the vector is inte-
grated into the genome by homologous recombination. Vectors
are constructed using Gateway Multisite® Cloning (Invitrogen).
The following describes methods for cloning homologous
sequences upstream and downstream of the target gene into the
appropriate pDONR vectors for cloning into pBHSNRG (Fig. 1a,
see Note 4).
1. P. patens homologs of genes of interest are identified using the
BLAST function available at the Cosmoss homepage (http://
www.cosmoss.org/, see Note 5). Choose “GENOME”, then
“BLAST” from the menu. Paste your peptide or nucleic acid
sequence into the box, choose the appropriate molecule (i.e.,
peptide or nucleic acid) and click “Submit” at the bottom of
the page. When your search is complete, click the “Click
here …” link, then the “my query” link. The results of a search
are reported as a list of gene models producing significant
alignments. Clicking on a gene model will direct you to the
alignment, where you can click on the “Genome Browser”
icon to examine the gene model. To download the genomic
and flanking sequences, zoom out to show at least 2 kb of
flanking sequence on both ends of the coding sequence. From
the “Report and Analysis” dropdown box, choose “Download
sequence file” and click on “Configure”. At the top of the
Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 279

Fig. 1. (a) Multisite Gateway® destination vector pBHSNRG. Amplified 5¢ and 3¢ homolo-
gous regions cloned into pDONR 221 P1-P4 and pDONR 221 P3–P2, respectively, are
inserted in the destination vector when the att R1/ att R4 and att R3/ att R2 sites of
pBHSNRG recombine with the entry clone att  L1/ att L4 and att L3/ att L2 sites to make
the knockout vector. (b) Knockout vector, wild type target genomic locus and trans-
formed target genomic locus. The vector is linearized using BsrGI or restriction enzymes
chosen to cut near the ends of the homologous region (vertical arrows). The positions of
primers used to test for targeted integration (Flank F/Vector R and Vector F/Flank R),
deletion of the target gene (Target F/Target R), and absence of concatenated vector
sequences (Vector F/Vector R) are shown.

“Configure” page, click the radio button for “Save to Disk”

and then click the “Go” ­button at the bottom of the page to
save the text file.
2. Open the sequence text file in any text editor. After the start
and stop codons are identified, primers are designed to amplify
about 1 kb of homologous sequence upstream and 1 kb of
homologous sequence downstream of the gene of interest.
Sequences for cloning into pDONR 221 P1-P4 (Element 1)
are added to the upstream forward and reverse primers, and
sequences for cloning into pDONR 221 P3-P2 (Element 3)
are added to the downstream forward and reverse primers
(see Gateway Multisite® instruction manual).
280 Roberts et al.

3. Primers are used to amplify genomic DNA isolated from wild

type P. patens (see Subheading 3.5) and the amplified frag-
ments are cloned into their respective pDONR vectors using
the BP Clonase II recombination reaction as described in the
Gateway Multisite® instruction manual to construct Element
1 and Element 3.
4. Element 1 and Element 3 are cloned into BHSNRG using
the LR Clonase II Plus recombination reaction as described
in the Gateway Multisite® instruction manual. Clones are verified
by sequencing with the M13 Reverse primer supplied in the
Gateway Multisite® kit and Vector R (see Subheading 2.4,
step 1, Fig. 1b).
5. Plasmid DNA is prepared using a commercial midi- or maxi-
prep kit.
6. The vector is linearized using BsrGI or restriction enzymes
chosen to cut near the ends of the homologous regions if
BsrGI truncates the vector (Fig. 1b).

3.2. Subculture 1. Plates are prepared by melting BCDAT medium and pouring
of P. patens about 25 mL into each unvented Petri plate. After the medium
Chloronemal solidifies, plates are overlain with sterile cellophane disks using
Filaments sterile forceps. Cellophane is allowed to relax and flatten out
for 10 min, and then straightened, if necessary, with sterile
forceps.
2. For each line to be subcultured, 2–4 stainless steel beads and
sterile water (2 mL for each plate to be inoculated from a
single line) are added to a sterile 15 mL centrifuge tube.
3. Sterile forceps are used to scrape the filaments from the starter
culture into a mound and transfer them to the centrifuge
tube. Up to ¼ of a confluent plate (about 50 mg of tissue) is
used for each new plate to be inoculated. Alternatively, a
pinch of tissue from a colony maintained on BCDAT medium
may be used (see Subheading 3.3, step 18).
4. Tubes are capped tightly, clamped in a paint shaker, and
shaken for several min until large clumps are broken up and
small clumps of a few dozen cells remain (see Note 6). When
older starter cultures are used, longer shaking is required.
5. Using a serological pipette, 2 mL of suspension are trans-
ferred to the surface of each plate and spread evenly.
6. Plates are incubated at 25°C with constant illumination at
50–80 mmol/m2/s (see Note 7). They are not sealed with
Parafilm® or stacked.
7. After 1 week, tissue is subcultured or plates are sealed with
Parafilm® and transferred to an incubator set to 10°C with a
2 h photoperiod at 20 mmol/m2/s for storage up to 1 year.
 Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 281

Clones can also be conserved for several years by suspending

about 50 mg of tissue in 1 mL of sterile distilled water in a
sterile microcentrifuge tube and storing at 4°C (see Note 8).

3.3. Transformation 1. 5–7 days before transformation, chloronemal tissue from the
and Selection line to be transformed is subcultured from a fresh plate (see
Note 9) on cellophane-overlain BCDAT plates and incubated
as described in Subheading 3.2.
2. Protoplast regeneration plates (3 per transformation) are pre-
pared with PRMB medium and overlain with sterile cello-
phane (Subheading 3.2, step 1).
3. Before beginning protoplast preparation, all materials are
made ready: waterbath is equilibrated to 45°C for heat shock;
a 500 mL beaker containing 300 mL of water is equilibrated
to room temperature (21–25°C); PRMT medium is melted
and equilibrated to 45°C in the water bath; linearized vector
DNA is ethanol precipitated and dissolved in sterile water at
1 mg/mL; one aliquot of Driselase solution per line to be
transformed and one aliquot of PEG solution per 1–3 vectors
are thawed and completely redissolved.
4. Protoplast preparation is begun by pipetting 9 mL of 8.5%
mannitol into a Petri plate and adding chloronemal filaments
scraped from the plate using sterile forceps, followed by 3 mL
of 2% Driselase solution. Petri plates are incubated for 60 min
with shaking at 60 rpm on an orbital shaker at 21–25°C.
5. Using a serological pipette, protoplast suspension is gently
drawn from the Petri plate and passed through a nylon filter
placed on top of a 50 mL disposal centrifuge tube.
6. Filtrate is transferred to a 15 mL disposable conical centrifuge
tube and centrifuged at speed 2–3 in a clinical centrifuge (see
Note 10) for 7 min; supernatant is discarded.
7. Protoplasts are resuspended in 10 mL 8.5% mannitol pipetted
directly onto the pellet from a serological pipette. It is impor-
tant to resuspend protoplasts gently and to avoid aspirating
protoplasts into the pipette. The suspension is centrifuged for
7 min, speed 2–3; supernatant is discarded. This step is
repeated.
8. During step 7, 15–30 mL of each vector is pipetted into a
labeled 15 mL disposable round-bottomed centrifuge tube.
9. Protoplasts are resuspended in 10 mL of 8.5% mannitol. After
10 mL of suspension are removed and loaded into a hemocy-
tometer, the suspension is centrifuged for 7 min, speed 2–3.
10. Intact protoplasts (see Note 11) are counted and the density
is estimated; use the instructions supplied with your hemocy-
tometer; 2–4 × 105 protoplasts/mL is typical.
282 Roberts et al.

11. Supernatant is discarded and protoplasts are resuspended in

3 M solution at 2 × 106 protoplasts/mL; 2–4 mL of suspen-
sion is typical.
12. Protoplast suspension (0.3 mL) and PEG solution (0.3 mL)
are added to each tube containing vector DNA. The suspen-
sion is mixed gently but thoroughly and incubated at 21–25°C
for 10 min.
13. Protoplasts are heat-shocked for 3 min at 45°C and trans-
ferred immediately to a beaker of 21–25°C water for 10 min.
Important – the water bath heater is turned off just before
submerging the tubes containing the protoplasts to prevent
overheating should the heater turn on during incubation.
14. Protoplasts are resuspended in 5 mL PRMT held at 45°C and
1.6 mL of suspension is spread on each of three cellophane-
overlain PRMB plates. This should result in an inoculation
rate of 2 × 105 protoplasts/plate (i.e., 2 × 106 protoplasts/
mL × 0.3 mL/transformation ÷ 3 plates/transformation).
15. Plates are incubated for 5 days at 25°C with constant illumi-
nation at 50–80 mmol/m2/s.
16. The regeneration rate can be estimated on day 5 as follows:
using a mm ruler, determine the field area of your dissecting
microscope on high power. Using the same magnification,
count the number of regenerated protoplasts in 3–5 fields
and calculate the average. The area of a plate is about
6,400 mm2. Estimate the number of regenerated protoplasts
per plate by multiplying the average number per field by
6,400 mm2 and dividing by the field area in mm2. Expect
about 30% regeneration (see Note 12).
17. Selection is initiated after 5 days. Selection plates are prepared
by melting BCDAT medium, cooling slightly, adding antibi-
otic (e.g., 15 mg/mL hygromycin), and pouring about 25 mL
each into unvented Petri plates. After medium solidifies, cel-
lophane disks are lifted from PRMB plates with sterile forceps
and transferred to BCDAT/antibiotic plates, taking care to
avoid trapping air bubbles under the cellophane. Plates are
incubated for 7 days at 25°C with constant illumination at
50–80 mmol/m2/s.
18. Clones surviving after 7 days on selection consist of stable
and unstable transformants. To select for stable transformants,
cellophane disks are transferred to BCDAT plates for 7 days,
then back to BCDAT/antibiotic plates for 7 days. Typically,
clones that grow vigorously during the second round of selec-
tion are stable. When large enough, they are split and arrayed
on duplicate BCDAT plates without cellophane and incu-
bated for 7 days at 25°C with constant illumination at
50–80 mmol/m2/s, and then stored at 10°C with a 2 h
 Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 283

­ hotoperiod at 20 mmol/m2/s. A small amount of tissue

p
(about 25 mm3) is collected for genotype analysis.

3.4. Genotype Analysis 1. Genomic DNA for PCR is prepared by homogenizing tissue
by PCR (about 25 mm3) in a 1.5 mL microcentrifuge tube using a
Pellet Pestle® Micro Grinder and immediately adding 500 mL
of DNA extraction buffer. Debris is pelleted in a microcentri-
fuge at high speed for 5 min and 350 mL of supernatant is
transferred to a 1.5 mL microcentrifuge tube containing
350 mL of isopropanol. DNA is pelleted in a microcentrifuge
at high speed for 15 min. The supernatant is poured off and
the tube containing the pellet is dried upside down on a paper
towel. The pellet is dissolved in 400 mL of TE buffer by shak-
ing for 30 min at 21–25°C.
2. PCR primers oriented outward from the selection cassette
(Vector R/Vector F; see Subheading 2.4, step 1) are paired
with primers designed to amplify inward from the genomic
regions flanking the homologous sequences contained within
the knockout vector (Flank F/Flank R) to test for 5¢ and 3¢
integration of the vector. Forward and reverse primers
designed to amplify the target sequence (Target F/Target R)
are used to test for deletion of the target gene (Fig. 1b). Flank
and target sequence primers are gene-specific.
3. DNA extracted from each stable line (2–4 mL) is amplified
with primers Flank F/Vector R (Fig. 1b) in a 25 mL PCR
reaction to test for integration of the 5¢ end of the vector (see
Note 13).
4. For stable lines testing positive for 5¢ integration, genomic
DNA (2–4 mL) is amplified with primers Vector F/Flank R
(Fig. 1b) in a 25 mL PCR reaction to test for integration of
the 3¢ end of the vector.
5. For stable lines testing positive for both 5¢ and 3¢ integration,
genomic DNA (2–4 mL) is amplified with primers Target F/
Target R (Fig. 1b) in a 25 mL PCR reaction to test for dele-
tion of the target sequence.
6. For stable lines testing positive for 5¢ and 3¢ integration, and
negative for the target sequence, genomic DNA (2–4 mL) is
amplified with primers Vector F/Vector R (Fig. 1b) in a
25 mL PCR reaction to test for insertion of concatenated
vector.

3.5. Genotype Analysis Southern blots are performed to test for nonhomologous inte-
by Southern Blotting gration of the vector. The probe is synthesized using the selection
cassette (amplified or restriction fragment) as a template. Genomic
DNA (about 2–3 mg each) is digested with 2–4 restriction
enzymes, each chosen to excise a 3–10 kb fragment containing
284 Roberts et al.

the selection cassette and run on a 0.7% agarose gel at 1 V/cm for
18–24 h. DNA is transferred, hybridized, and developed using
standard methods. Hybridization only to the expected fragments
is evidence for integration of a single copy of the vector at the
target locus. The following procedure yields about 20 mg of
genomic DNA (enough for 6–8 digests) and requires only a
microcentrifuge:
1. Seven-day old chloronemal tissue is scraped from a plate and
squeezed firmly between layers of filter paper to remove excess
liquid.
2. Up to 240 mg of squeeze-dried tissue is ground in a liquid
nitrogen-cooled Cryocup grinder. The resulting powder is
transferred to a 15 mL disposable centrifuge tube containing
2.4 mL extraction buffer for Southern blotting.
3. Homogenate is incubated at 65°C for 20 min with occasional
mixing by inversion.
4. Homogenate is transferred to three microcentrifuge tubes
(0.75 mL each). After adding 0.75 mL of chloroform:octanol,
each tube is inverted 6 times to mix.
5. Phases are separated by microcentrifuging at high speed,
5 min. Upper aqueous phase is transferred into clean micro-
centrifuge tubes, 0.9 mL DNA precipitation buffer is added
and each tube is inverted 6 times and chilled at 4°C for
15 min.
6. DNA is pelleted in a microcentrifuge at high speed, 15 min.
7. Pellets are washed with 1.5 mL of 70% ethanol and micro-
centrifuged at high speed, 5 min.
8. Supernatant is removed and remaining ethanol is evaporated
at 21–25°C, 10 min. Pellets are dissolved in 100 mL TE buf-
fer and combined.

3.6. Immuno­ 1. Protoplasts are prepared as described in Subheading 3.3,

fluorescent Cell Wall steps 1–7 and plated in PRML (1 mL per plate) on cello-
Labeling phane-overlain PRMB at a density of about 15,000 per plate.
The cellophanes are transferred to BCDAT plates after 2 days
(see Note 14) and cultured on BCDAT for 3–5 days.
2. To collect colonies, the plate is flooded with 3–5 mL of sterile
H2O and a cut-off 1 mL pipette tip is used to pipette the sus-
pension through a nylon filter placed over a 50 mL disposable
centrifuge tube. This removes dead protoplasts, which create
undesirable background labeling. The filter is placed upside
down in a clean Petri plate and colonies are washed into the
plate with 3–5 mL of purified water before pipetting them
into a 15 mL tube.
3. Colonies are centrifuged for 7 min at speed 3 in a clinical
centrifuge (speed is adjusted as necessary to collect tissue
 Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 285

without damaging it). Water is removed and colonies are

­suspended in 1 mL of fixative for 20 min at 21–25°C or over-
night at 4°C.
4. Colonies are centrifuged for 7 min at speed 3 in a clinical
centrifuge. Supernatant is removed and colonies are sus-
pended in 3 mL of PBS. This step is repeated two more
times.
5. During the washes, the flexiPERM cell chamber is pressed
onto a poly-L-lysine coated slide. The slide is incubated on a
slide warmer at 80°C for 20 min to ensure adhesion, and then
cooled to 21–25°C.
6. Colonies are centrifuged for 7 min at speed 3 in a clinical
centrifuge. Supernatant is removed and colonies are sus-
pended in 3 mL of purified water. This step is repeated two
more times, resuspending in 1 mL of purified water after the
final wash. All salts must be removed as they interfere with
adhesion of the tissue to the poly-l-lysine coated slides.
7. A cut-off 1 mL pipette tip is used to pipette about 50 mL of
suspension into each well (including an extra well for a nega-
tive control). Tissue sinks to the bottoms of the wells and
should cover them completely. More suspension may be added,
if necessary. After 20 min, water is carefully removed and the
slide is allowed to sit for 20 min to maximize adhesion.
8. Following the schedule below, solutions are pipetted into
each well, and then removed using care not to dislodge the
tissue. A multichannel pipettor may be used, if available.
(a) 200 mL of blocking solution, 10 min.
(b) 50 mL of primary antibody or CBM (see Note 15) diluted
to the recommended working concentration with block-
ing solution OR; 50 mL of blocking solution for negative
control wells, 1.5 h at 21–25°C or overnight at 4°C.
(c) 200 mL of PBS, 5 min. Repeat for a total of three
washes.
(d) 50 mL of secondary antibody diluted 1:100 in blocking
solution, 1.5 h at 21–25°C in the dark.
(e) 200 mL of PBS, 5 min. Repeat for a total of three
washes.
(f) After removal of the final PBS wash, the flexiPERM® is
carefully removed, and slides are air dried, 5–10 min. A
few drops of antifade reagent and a coverslip are added
and the slide is left in a dark dry place overnight.

3.7. Morphological Cell wall defects are often manifested as alterations in the
Analysis ­morphology of specific cells or tissues. The following methods
can be used to test for alterations in various stages of P. patens
growth and development.
286 Roberts et al.

3.7.1. Tip Growth Assay Protoplasts are prepared as described in Subheading 3.3, steps 1–7
and plated on cellophane-overlain PRMB at a density of about
5,000 per plate. The cellophanes are transferred to BCDAT plates
after 4 days (see Note 14) and cultured on BCDAT for 3 days.
Chlorophyll autofluorescence from the resulting colonies is pho-
tographed at 63 × using a fluorescence dissecting microscope with
blue excitation (488 nm). An ImageJ macro for analyzing area,
circularity, and solidity is available from Luis Vidali, Worcester
Polytechnic Institute, Worcester, MA, USA (30).

3.7.2. Gametophore Protoplasts are prepared as described in Subheading 3.3, steps

Development Assay 1–7 and plated in PRML on cellophane-overlain PRMB at a den-
sity of about 1,000 per plate. The cellophanes are transferred to
BCD plates after 2–4 days (see Note 14) and cultured for 10 days
at 25°C with constant illumination at 50–80 mmol/m2/s.
Gametophore buds appear 6 days after transfer to BCD and
develop into leafy gametophores over the next 4 days.

3.7.3. Caulonemal Petri plates containing BCDAT with 1.2% agar and 350 mM
Gravitropism Assay sucrose are prepared. About 7 small clumps of fresh chloronemal
tissue are plated along the diameter of each plate. After incuba-
tion at 25°C with constant illumination at 50–80 mmol/m2/s,
the plates are positioned vertically and incubated in the dark at
25°C for 14 days. Pigmented caulonemal filaments are negatively
gravitropic and may exceed 2 cm in length (31).

3.7.4. Rhizoid Chloronemal tissue is plated on BCD medium supplemented

Development Assay with 0.1–1.0 mM naphthylacetic acid and cultured at 25°C with
constant illumination at 50–80 mmol/m2/s for 14 days.
Gametophores develop numerous rhizoids (5).

4. Notes

1. Some media formulas for P. patens use Ca(NO3)2 as a nitrate

source (32). However, calcium phosphate precipitates form
during autoclaving and this is prevented by using KNO3 as a
nitrate source and adding CaCl2 solution after autoclaving.
2. The use of unvented Petri plates reduces evaporation and
contamination. Alternatively, vented Petri plates can be sealed
with Micropore® tape (3 M Corporation, St. Paul, MN,
USA).
3. We have also used roll cellophane (Research Products
International) cut in 8.5 × 11 in. sheets, interleaved with copy
paper, cut into 9 cm circles, and autoclaved in glass Petri
Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 287

plates. However, growth inhibition has been noted with some

lots of this product.
4. Multisite Gateway® is a rapid method for producing knock-
out vectors. A destination vector (pBHSNRG) was con-
structed by inserting R1–R4 and R3–R2 Gateway® cassettes
into the multiple cloning sites of pBHSNR (gift of D. Schaefer,
University of Lausanne, constructed by inserting the hpt gene
driven by a double 35 S promoter (Sac1-NotI fragment from
pCAMBIA) in reverse orientation between the two loxP sites
of pBilox) (33). The Gateway® Reading Frame Cassette A
(Invitrogen) was modified by PCR (converting R2–R4 or
R1–R3) and cloned into pGEM-T Easy (Promega Corp.,
Madison, WI, USA). The R1–R4 and R3–R2 cassettes were
excised from pGEM-T Easy with SphI/SpeI and AvrII/NsiI,
respectively, and ligated into the SphI/XbaI and SpeI/NsiI
sites, respectively, of pBHSNR. pBHSNRG is available upon
request. Vector construction methods using standard restric-
tion digestion and ligation are available in print (29) and
online (http://moss.nibb.ac.jp/). Detailed treatments of the
effects of various vector parameters on recombination effi-
ciency are also available (34, 35).
5. The P. patens genome was sequenced by the U.S. Department
of Energy Joint Genome Institute Community Sequencing
Program http://www.jgi.doe.gov/. The annotated sequence
is hosted at Cosmoss http://www.cosmoss.org/.
6. As an alternative to shaking, protonema can be blended for
subculture using a Polytron® type homogenizer (36). We
have noted a reduced contamination frequency using the
shaking method. When subculturing fresh cultures, vortexing
can be substituted for shaking. However, older cultures are
often not sufficiently broken up by vortexing.
7. P. patens may be grown with continuous light or a long-day
photoperiod (typically 18 h). The cell division cycle becomes
synchronized to the long-day photoperiod (28).
8. Other methods for long-term storage, including cryopreser-
vation have been reported (28).
9. Chloronemal tissue used to generate protoplasts must be in
excellent condition. Dead filaments cause clumping of proto-
plasts and low transformation efficiency. Tissue should be
subcultured every 5–7 days until no dead, brown filaments
are visible with a dissecting microscope.
10. This protocol uses more gentle centrifugation (about 25 × g)
compared to published protocols (28, 29) and the resulting
pellet is easier to resuspend. The supernatant from the proto-
plast washes should be clear. A green tinge results from free
chloroplasts, a sign of protoplast lysis.
288 Roberts et al.

11. Intact protoplasts are spherical and appear turgid with their
chloroplasts pressed against the plasma membrane. Protoplasts
with chloroplasts aggregated in the centre will not regenerate
and should not be counted. Intact protoplasts should sub-
stantially outnumber damaged protoplasts.
12. The rate of protoplast regeneration after 5 days is an indicator
of successful transformation. At this stage protoplasts should
have divided several times and there should be more than
50,000 per plate. A good transformation yields hundreds of
unstable transformants, after the first round of selection, and
dozens of stable transformants, after the second round of
selection.
13. The following should be considered when interpreting the
PCR genotyping results. The percentage of stable transfor-
mants in which the vector is integrated by homologous
recombination at both the 5¢ and 3¢ ends typically ranges
from 25 to 100%. In some cases, several tandem and/or
inverted copies of the vector may be integrated at the target
locus. Protoplast fusion can occur during transformation,
producing diploid clones carrying both the vector ­integrated
by homologous recombination and the wild type gene. Refer
to detailed treatments of integration mechanisms and result-
ing genotypes (34, 35) for more information.
14. Developing protoplasts can be transferred from PRMB to
antibiotic-free BCDAT after 2 days. However, the protoplasts
require 4 days to develop antibiotic resistance.
15. The same protocol with modifications is used to label
with CBM. CBM is substituted for primary antibody in
Subheading 3.6, step 8(b), and an antipolyhistidine incuba-
tion and three washes are added before the secondary anti-
body incubation.

Acknowledgements

We thank the members of the moss community for their collegiality

and helpful suggestions. Magdalena Bezanilla and the members
of her group provided invaluable advice and assistance. We thank
Didier Schaefer for the gift of pBHSNR. This project was sup-
ported by the National Research Initiative Competitive Grant no.
2007-35318-18389 from the USDA National Institute of Food
and Agriculture.
 Knocking Out the Wall: Protocols for Gene Targeting in Physcomitrella patens 289

References
1. Schaefer, D. G. (2002) A new moss genetics: 11. Khraiwesh, B., Ossowski, S., Weigel, D.,
targeted mutagenesis in Physcomitrella patens. Reski, R., and Frank, W. (2008) Specific gene
Annu Rev Plant Biol 53, 477–501. silencing by artificial MicroRNAs in
2. Strepp, R., Scholz, S., Kruse, S., Speth, V., Physcomitrella patens: an alternative to tar-
and Reski, R. (1998) Plant nuclear gene geted gene knockouts. Plant Physiol 148,
knockout reveals a role in plastid division for 684–693.
the homolog of the bacterial cell division pro- 12. Vidali, L., Augustine, R. C., Fay, S. N., Franco,
tein FtsZ, an ancestral tubulin. Proc Natl P., Pattavina, K. A., and Bezanilla M. (2009)
Acad Sci USA 95, 4368–4373. Rapid screening for temperature-sensitive
3. Schaefer, D. and Zrÿd, J. -P. (2004) Principles alleles in plants. Plant Physiol 151, 506–514.
of targeted mutagenesis in the moss 13. Schumaker, K. S., and Dietrich, M. A. (1998)
Physcomitrella patens. In: Wood, A.J., Oliver, Hormone-induced signaling during moss
M.J., and Cove, D.J., eds. New Frontiers in development. Annu Rev Plant Physiol Plant
Bryology: Physiology, Molecular Biology and Mol Biol 49, 501–523.
Functional Genomics. Dordrecht: Kluwer; 14. Lee, K. J. D., Sakata, Y., Mau, S. -L., Pettolino,
37–49. F., Bacic, A., Quatrano, R. S., Knight, C. D.,
4. Perroud, P. -F. and Quatrano, R .S. (2006) Knox, J. P. (2005) Arabinogalactan proteins
The role of ARPC4 in tip growth and align- are required for apical cell extension in the
ment of the polar axis in filaments of moss Physcomitrella patens. Plant Cell 17,
Physcomitrella patens. Cell Motil Cytoskeleton 3051–3065.
63, 162–171. 15. Liepman, A. H., Nairn, C. J., Willats, W. G.
5. Sakakibara, K., Nishiyama, T., Sumikawa, N., T., Sørensen, I., Roberts, A. W., and Keegstra,
Kofuji, R., Murata, T. and Hasebe, M. (2003) K. (2007) Functional genomic analysis sup-
Involvement of auxin and a homeodomain- ports conservation of function among Cellulose
leucine zipper I gene in rhizoid development synthase like A gene family members and sug-
of the moss Physcomitrella patens. Development gests diverse roles of mannans in plants. Plant
130, 4835–4846. Physiol 143, 1881–1893.
6. Cove, D. (2005) The moss Physcomitrella pat- 16. Fu, H., Yadav, M. P., and Nothnagel, E. A.
ens. Annu Rev Genet 39, 339–58. (2007) Physcomitrella patens arabinogala­
7. Richardt, S., Timmerhaus, G., Lang, D., ctan proteins contain abundant terminal
Qudeimat, E., Corrêa, L.G.G., Reski, R., 3-O-methyl-l-rhamnosyl residues not found
Rensing, S. A., and Frank, W. (2010) in angiosperms. Planta 226, 1511–1524.
Microarray analysis of the moss Physcomitrella 17. Moller, I., Sørensen, I., Bernal, A. J., Blaukopf,
patens reveals evolutionarily conserved tran- C., Lee, K., Øbro, J., Pettolino, F., Roberts,
scriptional regulation of salt stress and abscisic A., Mikkelsen, J. D., Knox, J. P., Bacic, A.,
acid signalling. Plant Mol Biol 72, 27–45. and Willats,W. G. T. (2007) High-throughput
8. Cuming, A. C., Cho, S. H., Kamisugi, Y., mapping of cell-wall polymers within and
Graham, H., and Quatrano, R. S. (2007) between plants using novel microarrays. Plant
Microarray analysis of transcriptional responses J 50, 1118–1128.
to abscisic acid and osmotic, salt, and drought 18. Peña, M. J., Darvill, A. G., Eberhard, S., York,
stress in the moss, Physcomitrella patens. New W. S., and O’Neill, M. A. (2008) Moss and
Phytol 176, 275–287. liverwort xyloglucans contain galacturonic
9. Rensing, S. A., Lang, D., Zimmer, A. D., acid and are structurally distinct from the
et al. (2008) The Physcomitrella genome xyloglucans synthesized by hornworts and
reveals evolutionary insights into the conquest vascular plants. Glycobiology 18, 891–904.
of land by plants. Science 319, 64–69. 19. Graham, L. E., Cook, M. E., and Busse, J. S.
10. Bezanilla, M., Perroud, P. -F., Pan, A., Klueh, (2000) The origin of plants: Body plan
P., and Quatrano, R. S. (2005) An RNAi sys- changes contributing to a major evolutionary
tem in Physcomitrella patens with an internal radiation. Proc Natl Acad Sci USA 97,
marker for silencing allows for rapid identifi- 4535–4540.
cation of loss of function phenotypes. Plant 20. Niklas, K. J. (2004) The cell walls that bind
Biol 7, 251–257. the tree of life. BioScience 54, 831–841.
290 Roberts et al.

21. Roberts, A. W., and Bushoven, J. T. (2007) 30. Vidali, L., Augustine, R. C., Kleinman, K. P.,
The cellulose synthase (CESA) gene super- and Bezanilla, M. (2007) Profilin is essential
family of the moss Physcomitrella patens. Plant for tip growth in the moss Physcomitrella pat-
Mol Biol 63, 207–219. ens Plant Cell 19, 3705–3722.
22. Schaefer, D. G., and Zrÿd, J. -P. (2001) The 31. Cove, D. J., and Quatrano, R. S. (2004) The
moss Physcomitrella patens, now and then. use of mosses for the study of cell polarity. In:
Plant Physiol 127, 1430–1438. Wood, A. J., Oliver, M. J., and Cove, D. J.,
23. Lang, D., Zimmer, A. D., Rensing, S. A., and eds. New Frontiers in Bryology: Physiology,
Reski, R. (2008) Exploring plant biodiversity: Molecular Biology and Functional Genomics.
the Physcomitrella genome and beyond. Trends Dordrecht: Kluwer, 189–203.
Plant Sci 13, 542–9. 32. Ashton, N. W., Grimsley, N. H., and Cove, D.
24. Schaefer, D. G. (2001) Gene targeting in J. (1979) Analysis of gametophytic develop-
Physcomitrella patens. Curr Opin Plant Biol 4, ment in the moss, Physcomitrella patens, using
143–150. auxin and cytokinin resistant mutants Planta
25. Decker, E. L., Frank, W., Sarnighausen, E., 144, 427–435.
and Reski, R. (2006) Moss systems biology en 33. Thelander, M., Nilsson, A., Olsson, T.,
route: phytohormones in Physcomitrella devel- Johansson, M., Girod, P. -A., Schaefer D. G.,
opment. Plant Biol 8, 397–405. Zrÿd J. -P., and Ronne, H. (2007) The moss
26. Cove, D., Benzanilla, M., Harries, P., and genes PpSKI1 and PpSKI2 encode nuclear
Quatrano, R. (2006) Mosses as model systems SnRK1 interacting proteins with homolo­
for the study of metabolism and development. gues in vascular plants. Plant Mol Biol 64,
Annu Rev Plant Biol 57, 497–520. 559–573.
27. Frank, W., Decker, E. L., and Reski, R. (2005) 34. Kamisugi, Y., Cuming, A. C., and Cove, D. J.
Molecular tools to study Physcomitrella pat- (2005) Parameters determining the efficiency
ens. Plant Biol 7, 220–227. of gene targeting in the moss Physcomitrella
28. Cove, D. J., Perroud, P. F., Charron, A. J., patens. Nucleic Acids Res 33, e173.
McDaniel, S. F., Khandelwal, A., and Quatrano, 35. Kamisugi, Y., Schlink, K., Rensing, S. A.,
R. S. (2009) The moss Physcomitrella patens. A Schween, G., von Stackelberg, M., Cuming,
novel model system for plant development and A. C., Reski, R., and Cove, D. J. (2006) The
genomic studies. In: Behringer, R.R., Johnson, mechanism of gene targeting in Physcomitrella
A.D., Krumla\uf, R.E., eds. Emerging Model patens: homologous recombination, concate-
Organisms: A Laboratory Manual: Cold Spring nation and multiple integration. Nucleic Acids
Harbor Laboratory Press 69–104. Res 34, 6205–6214.
29. Knight, C. D., Cove, D. J., Cuming, A. C., and 36. Nishiyama, T., Hiwatashi, Y., Sakakibara, I.,
Quatrano, R. S. (2002) Moss gene technology. Kato, M., and Hasebe, M. (2000) Tagged
In: Gilmartin, P.M., C. B, eds. Molecular Plant mutagenesis and gene-trap in the moss,
Biology — A Practical Approach. Oxford, New Physcomitrella patens by shuttle mutagenesis.
York: Oxford Press; 285–301. DNA Res 7, 9–17.
 Chapter 20

Measuring In Vitro Extensibility of Growing Plant Cell Walls

Daniel J. Cosgrove

Abstract
This article summarizes the theory and practical aspects of measuring cell wall properties by four different
extensometer techniques and how the results of these methods relate to the concept and ideal measure-
ment of cell wall extensibility in the context of cell growth. These in vitro techniques are particularly
useful for studies of the molecular basis of cell wall extension. Measurements of breaking strength, elastic
compliance, and plastic compliance may be informative about changes in cell wall structure, whereas
measurements of wall stress relaxation and creep are sensitive to both changes in wall structure and wall-
loosening processes, such as those mediated by expansins and some lytic enzymes. A combination of
methods is needed to obtain a broader view of cell wall behavior and properties connected with the con-
cept of cell wall extensibility.

Key words: Extensibility, Cell wall, Stress relaxation, Creep, Expansin

1. Introduction

Acting like a firm corset, the cell wall gives plant cells their specific
shape and size by restraining the expansive tendency of the proto-
plast. In growing cells, this mechanical restraint is rather more
subtle than a corset, as the cell wall not only resists turgor pres-
sure but at the same time it also stretches slowly and irreversibly
and often anisotropically (directionally), a controlled process in
which the load-bearing network of wall matrix polymers and cel-
lulose microfibrils yields to the turgor-generated tensile forces in
the cell wall. As used here, the term “wall extensibility” refers to
this ability to extend irreversibly, but a close reading of the litera-
ture reveals this to be a fuzzy concept, with various technical defi-
nitions (when defined at all) and a variety of methods for estimating
its value. These concepts and methods were critically reviewed in
(1). The current article presents an update focused on the utility
and concepts underlying various measurements of wall properties

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9_20, © Springer Science+Business Media, LLC 2011

291
292 Cosgrove

in vitro, specifically by use of extensometers in which a cell wall

specimen is clamped and extended in various modes to measure
cell wall behavior that one might relate to cell wall extensibility in
the context of cell growth. Measurements at the single cell level
are also possible, but technically more challenging and difficult to
interpret, and are detailed elsewhere (2, 3).
As a hydrated composite of complex polysaccharides, the
growing plant cell wall has elastic, viscoelastic, and plastic prop-
erties, which have been measured in the past half century by a
variety of static and dynamic loading methods. These mechani-
cal techniques, in which a physical force is put on the isolated
wall and the resulting deformation is measured, assess something
a bit different from what we mean by wall extensibility in the
context of cell growth. In biophysical terms, wall extensibility is
usually defined as the local slope of the curve relating growth
rate to cell turgor pressure (1, 4). This discrepancy between
concept and technical measurement stems from the fact that cell
wall expansive growth is not a simple matter of inert polymer
mechanics; additionally, it depends on dynamic wall-loosening
processes that modify the matrix-cellulose network in a time-
dependent way, enabling the polymeric network to yield to wall
stresses. This results in cell wall creep (slow, time-dependent,
irreversible extension), in which the cellulose microfibrils shift
or separate from one another. We might call this “chemorheologi-
cal extensibility” (5) to emphasize that the wall’s flow (rheologi-
cal) behavior depends not only on cell wall structure and tensile
stress but also on dynamic processes of polymer creep catalyzed
by cell wall-loosening agents, such as expansins and lytic enzymes
whose activity may be rapidly modulated by changes in wall
pH, redox potential, hydration, ion concentrations, the supply
of cell wall materials, and other ephemeral conditions within the
cell wall space. The biophysical and cellular bases of this process
were recently reviewed (4). This turn of phrase, chemorheologi-
cal extensibility, may more clearly denote the dynamic nature of
wall extensibility, but it still leaves us with the problem of how
to measure this subtle property. This issue is relevant to a variety
of studies where walls become more, or less, extensible during
cell development and in response to hormones, light, gravity,
dehydration, and a variety of other environmental and biological
stresses that influence plant growth. Measuring wall extensibility
and understanding the molecular nature of its modulation are
key aspects of understanding how plants control their cell growth
by modifying their primary cell walls.
In vivo methods for measuring wall extensibility, summarized
in (1), have the advantage of fitting nicely into the biophysical
theory of plant growth in quantitatively meaningful ways, and
they are sensitive to the cell’s dynamic modification of the cell wall
environment. Methods for measuring wall stress relaxation in vivo,
 Measuring In Vitro Extensibility of Growing Plant Cell Walls 293

such as the pressure-block method (6) or the ­pressure-probe

method (7, 8), have some advantages over other in vivo methods,
but they require specialized equipment and moreover they have
limited utility for investigating cell wall loosening by exogenous
enzymes or for other studies where the molecular nature of wall
extension is investigated. For such studies, isolated cell walls –
i.e., where the living cells of the tissue have been removed or
otherwise made inactive – have particular merit because the results
are not compromised by complicated responses of the living cells
during the measurement and because a wider variety of treat-
ments and methods may be applied to the isolated cell walls with-
out concern about complex responses (even death!) of the living
cells. These in vitro methods are particularly well suited to studies
of the biochemical underpinnings of cell wall enlargement and are
the subject of this article.
One notable caveat with such in vitro studies, however, is that
a negative result (no measurable change in isolated wall proper-
ties) does not necessarily mean that in vivo wall extensibility does
not change during a particular growth response. For instant, cells
may rapidly modulate cell wall pH and thereby affect expansin
activity and increase wall extensibility, but such action would not
be detected in the methods outlined below. Thus, the results of
in vivo and in vitro methods may disagree, likely pointing to an
ephemeral, rather than structural, change in the cell wall.
Several methods for assessing extensibility of isolated cell
walls by use of extensometers have been devised, four of which
are summarized below. Note that they neglect, by their very
nature, the dynamic aspects of cellular modulation of cell wall
yielding, but there is still a chemorheological aspect because
expansins and other wall modifying enzymes may remain active in
the isolated walls (assuming no protein-denaturing treatments).

2. Materials

1. Plant material of interest: stems, hypocotyls, or plant organs

of similar geometry. Slices of tissues may also be used. The
plant material must be pretreated to disrupt the protoplasts
(see Notes 1 and 2).
2. Extensometer: can be commercial or custom-made (see
Note 3).
3. Carborundum powder, prewashed.
4. Microscope slides or other suitable glass plates.
5. Weight.
6. Liquid nitrogen or −80°C freezer.
294 Cosgrove

7. Adhesive tape or cyanoacrylate adhesive (“Superglue”).

8. Buffer: 20 mM sodium acetate, pH 4.5.

3. Methods

3.1. Preparation To disrupt the protoplasts and remove the bulk water and cellular
of Plant Tissues to fluids within the tissue, a simple procedure is as follows (see Note 4):
Remove Protoplasts
1. Freeze the sample in liquid nitrogen or in a −80°C freezer.
2. Thaw at room temperature or 4°C and cut to size suitable for
the extensometer clamps.
3. Permeabilize the cuticle, e.g., by abrasion with a slurry of
well-washed carborundum powder (see Note 5).
4. Treat the plant tissues to inactivate wall-bound enzyme activ-
ity, e.g., by boiling (see Note 6).
5. Press the sample between two glass plates (i.e., microscope
slides) for ~5 min under a weight to flatten the tissue and
express the cell sap.

3.2. Measuring Cell Four methods are detailed below for measuring (1) breaking
Wall Mechanics strength, (2) elastic and plastic compliances, (3) stress relaxation,
and (4) cell wall creep. These techniques measure different aspects
of the cell wall mechanics, with the creep method coming perhaps
closest to the measurement desired for in-vitro wall extensibility.
All of the methods are essentially described as follows:
1. Clamp the prepared cell walls in an extensometer.
2. Apply a tensile force to the walls.
3. Either extend the walls at a defined rate or measure their
extension.
A visual demonstration of these procedures may be found in (9).

3.2.1. Breaking Strength The principle here is simple, if not simple minded: measure the
force need to break the cell wall. There are various geometrical
means for causing breakage or mechanical failure of plant organs,
with pulling (axial tension) or buckling (compression or bending)
being the most straightforward for engineering analysis. For mea-
surements of axial breaking strength, a suitable sample is clamped
in an extensometer that can extend the distance between the
clamps while measuring the force needed for the extension. It is
important to clamp the sample in such a way that the clamp does
not cause tearing of the sample. Sometimes adhesive tape or
cyanoacrylate adhesive (“Superglue”) is used to help fix the walls
to the clamping device (10, 11). As the sample is extended, the
 Measuring In Vitro Extensibility of Growing Plant Cell Walls 295

Fig. 1. Schematic diagrams of force-extension curves for cell walls measured in an extensometer for (a) tensile strength
(breakage force) and (b) elastic and plastic moduli. For assessment of tensile strength the wall is extended until it fails,
and the maximum force, the total extension, and the area under the curve may be useful metrics of changes in cell wall
structure. For assessment of cell wall moduli, the wall is extended, then returned to its original size and extended a
second time. The slope of the second extension, near the end of the extension, is taken as the elastic modulus; its recip-
rocal is the elastic compliance. The slope of the first extension is the total modulus and the total compliance is its
­reciprocal. The plastic compliance is the difference between the total compliance and the elastic compliance.

force increases and then levels off as the sample begins to fail and
then drops quickly as the breakage is completed (Fig. 1). Typically,
the maximum force is taken as the breaking strength, but one can
also measure the percentage extension before failure and the area
under the curve, which is the energy input for breakage (12). The
breaking force of plant wall samples is often found to be a func-
tion of the extension rate, so this parameter needs to be the same
for valid comparisons. Another important detail: be sure that the
sample stays wet during the measurement, as wall dehydration
greatly increases cell wall strength.
The breaking strength of hypocotyls or inflorescence stems
has been used in recent times to characterize Arabidopsis mutants
with defects in cell wall composition or in wall assembly (11, 13–
15), thereby drawing inferences about the structural role of a par-
ticular wall polymer (Fig. 1a). What such measurements mean for
wall extensibility in the context of growth is more difficult to say,
as there is at best only an indirect connection between the two
concepts. Organ breakage occurs at the weakest point in the sam-
ple, which may be the middle lamella, that is, the adhesive layer of
matrix polysaccharides and structural proteins that cements adja-
cent cell walls together. Changes in organ anatomy could also
affect breaking strength. In contrast, wall extensibility depends
on rearrangements within the matrix-cellulose network. Thus,
while breaking strength may be informative about aspects of cell
wall structure or the glue that holds cells together, it is not a
296 Cosgrove

r­ eliable metric of wall extensibility. As a case in point, Arabidopsis

hypocotyls from a xyloglucan-deficient mutant are weaker than
wild type when assessed in mechanical tests of stiffness and break-
ing strength (14), yet they are stronger (less extensible) in assays
of cell wall creep (Y.B. Park and D.J. Cosgrove, unpublished
data). Thus, tensile breaking strength may be informative about
structural changes in cell walls or tissue architecture, but is in
general a poor measure of wall extensibility.

3.2.2. Elastic and Plastic The principle of this method is similar to that described above for
Compliances: Axial breaking strength, except that the sample is not extended to the
Extension breaking point, but is extended a small amount in two cycles.
In the first cycle, the sample is extended until a predetermined
force is reached (well before the breakage point), then returned
to the original length before a second extension is made. This
second force-extension curve differs from the first one (Fig. 1b),
but subsequent extensions cycles are reversible, at least to a first
approximation, and so extension #2 is taken as an elastic exten-
sion. The difference between the first curve (total extension) and
the second curve (elastic extension) gives the plastic, or irrevers-
ible, extension. The slope of the lines are estimated near the end
of each extension cycle, to give Dforce/Dextension; this ratio is
known as a modulus. The higher the modulus, the stiffer the
material.
The modulus depends on many characteristics of the cell wall,
but we do not have a comprehensive theory of this yet. Among
the most important wall characteristics are the number and bun-
dling of cellulose microfibrils in wall cross sectional area; the ori-
entation of the microfibrils relative to the direction of extension;
and the density, hydration, and cross linking of the matrix and its
connection to the cellulose microfibrils. A recent theoretical study
has attempted to explore the elastic modulus of cell walls by use
of finite element analysis to calculate the elastic behavior of a vir-
tual cellulose-hemicellulose network constructed to mimic aspects
of real cell walls (16), while a very different type of model, based
on the thermodynamics of hydrogen-bonded networks, was used
to predict the plastic behavior of similar idealized cell walls (17).
These and other theoretical models make significant steps toward
gaining molecular-scale insights into cell wall mechanics, but they
are still very simplified and limited compared to real cell walls.
The raw units for this modulus might be N/mm or g-force/
mm. If the Dforce is divided by the cross-sectional area of the
sample and the Dextension is calculated as fractional change in
length, then the units can be readily converted to standard units
of MPa (that is, stress divided by strain, or force per unit area
divided by the fractional increase in length). Estimates of cross-
sectional area, however, can be problematic (i.e., you cannot
count cell lumens; the sample gets thinner as it extends, etc.),
 Measuring In Vitro Extensibility of Growing Plant Cell Walls 297

introducing errors into the absolute values of the moduli. Hence,

such values reported in the literature should be examined with a
critical eye. As long as the cross sections are similar for all samples,
then a comparison of values is valid for reaching conclusions
about relative changes among groups. On the other hand, if dif-
ferent groups have different cross-sectional areas, then the inter-
pretation of the values can be more challenging, as differences in
organ anatomy can influence the results. Similarly, if wall thick-
nesses differ among comparison groups, then the interpretation
of the values needs careful analysis. For instance, at equal force a
thinner sample will have a larger stress than a thicker sample. The
mechanical properties of cell walls are strongly nonlinear, so com-
parisons should be made at similar values of wall stress. This may
mean one needs to apply a larger tensile force to a thicker tissue
sample in order to equalize stress in the two measurements.
Because wall cross-sectional area in a sample is difficult to
measure, a practical substitute is to use the cell wall mass per unit
length of the sample. For many samples, this value can be esti-
mated by cutting a sample to 1 cm length, freezing and thawing
to disrupt the cells, then washing and pressing the sample to
remove cell contents. What remains is mostly cell walls, which can
be weighed after drying.
The reciprocal of the force/extension value is known as a
compliance, which corresponds to a type of extensibility, with
units of strain/stress (or Dextension/Dforce for practical mea-
surements). The elastic or plastic compliances that one measures
with this technique are sometimes called extensibilities, but keep
in mind that these are purely mechanical extensibilities that
depend primarily on cell wall structure and thickness and do not
include the chemorheological aspects of cell wall extensibility.
As an example, a-expansin does not affect elastic or plastic
­compliances of cucumber hypocotyl cell walls (18). These
­compliances are good reporters for changes in cell wall structure.
For instance, treatment of cucumber cell walls with a family-12
endoglucanase caused large increases in both the elastic and ­plastic
compliances of the cell walls (18).

3.2.3. Elastic and Plastic A microscopic variant of the stress-strain method has been used
Compliances: for evaluating the mechanical properties of cell walls in single cells
Microindentation or parts of cells, such as at different parts of the apical dome of
pollen tubes (19). The method does not separate out elastic and
plastic components, as above, and the stiffness values obtained are
useful for relative comparisons, but not for obtaining absolute
values of wall modulus or compliance. The principle of the method
is simple: a small probe is used to deform a local region of the cell
wall and to measure the force on the wall. The resistance to such
deformation is a complex function of cell wall stiffness, cell
­geometry, and turgor pressure (20). Wall stiffness depends on the
298 Cosgrove

thickness of the wall and its modulus, which in turn depends on

the arrangement of its structural elements (cellulose microfibril
density, orientation, and interconnection by matrix polymers).
The major advantage of the method is that it may be used at the
single cell level and even to probe different parts of the cell (19),
but this comes at the cost of some uncertainty about the physical
interpretation of the measured values.

3.2.4. Stress Relaxation A crucial biophysical difference between growing and ­nongrowing
cell walls is that the former undergoes continuous stress relax-
ation (the physical face of cell wall loosening), which lowers cell
turgor pressure and creates the water potential gradient necessary
for sustained water uptake by the growing cell. Water uptake
physically enlarges the cell and counter balances wall stress relax-
ation so that turgor pressure stabilizes. This theory of wall relax-
ation was first enunciated qualitatively by Ray et al. (21), elaborated
in specific quantitative terms by Cosgrove (22), and demonstrated
experimentally in a series of studies in which water uptake into the
growing cells was prevented, thereby allowing stress relaxation to
proceed unabated by water uptake, resulting in a decay in turgor
pressure to the yield threshold (7, 8, 22).
These techniques for measuring in vivo wall stress relaxation
have a counterpart for isolated cell walls (23–25), in which the
isolated wall is clamped in an extensometer as above, rapidly
extended until a predetermined force is attained, and then held to
constant dimension while the holding force is monitored. The
practical time scale for these measurements is from about 50 ms
to 500 s. Longer times are possible if cell wall dehydration can be
prevented. Shorter times are limited by the mechanics of the
extensometer: it takes ~50 ms to extend the wall sample and to
allow time for the induced vibrations (mechanical “ringing”) in
the sample to dampen out.
During the extension process, the cellulose-matrix network is
elastically stretched; some of the wall polymers subsequently relax
to lower energy states, resulting in wall stress relaxation. As a
result, the holding force decays with time. The resulting decay in
force, or stress relaxation, may be converted into a form known as
a stress relaxation spectrum; a mathematically simpler operation,
which approximates the relaxation spectrum, is to convert relax-
ation to log time scale and to plot the rate as a function of log
time (Fig. 2), i.e., −Dforce/Dlog time vs. log time (26). Much of
this relaxation is the result of the viscoelastic nature of the cell
wall material, that is, a passive physical response that depends on
cell wall structure. This is the case for both growing and non-
growing walls, although the difference in wall structure in the
two cases may result in different stress relaxation behavior.
Additionally, any wall-loosening processes that are still active
in isolated cell walls may result in additional stress relaxation not
 Measuring In Vitro Extensibility of Growing Plant Cell Walls 299

Fig. 2. Diagram illustrating stress relaxation spectra of heat-inactivated cucumber hypo-

cotyl cell walls treated with buffer alone (−aExpansin) or with buffer containing
a-expansins (+aExpansin) (18, 29).

found in the nongrowing cell wall. Expansin-induced stress

­relaxation is readily demonstrated in these assays (27); in acidic
buffer, the cell walls exhibit faster stress relaxation than in neutral
buffer. This difference is eliminated by brief heat treatment, show-
ing that the difference is not simply due to a pH-dependence of
pectin rigidity in the cell wall. It is restored by addition of expansins,
showing them to be the major mediators of wall stress relaxation
in such isolated cell walls. On the other hand, the family-12 endo-
glucanase that increased elastic and plastic compliances (men-
tioned above) had no effect on the stress relaxation spectrum at
times >1 s; its action on wall plasticity could be seen as increased
stress relaxation at times <0.2 s (18). Thus, stress relaxation assays
are sensitive to changes in cell wall structure, expansin activity, and
potentially lytic enzymes that promote cell wall extension.

3.2.5. Creep The fourth extensometer method measures cell wall creep, which
is the time-dependent, irreversible extension of wall samples held
at constant force. Because wall creep is a slow process, the mea-
surement period is typically in the range of 30–150 min and the
wall samples are clamped in a buffer-filled cuvette to prevent
dehydration. Wall samples are typically clamped at a constant
force in a neutral buffer, and after 10–15 min the buffer is switched
to an acid one, initiating rapid extension which gradually slows
over 30–60 min to a constant or near constant rate (Fig. 3). A
variant of this method is to clamp the walls in acidic buffer at low
force which is insufficient to cause creep, then raise the force to a
value high enough to cause wall creep.
Of the four in vitro methods described in this article, the
creep method mimics the in-vivo wall extension process to the
closest degree, and it readily distinguishes between growing and
300 Cosgrove

Fig. 3. Diagram illustrating the creep behavior of cucumber cell walls when clamped in
a constant force extensometer at neutral pH and then switched to acidic buffer (26).

nongrowing cell walls for many plant tissues. The difference

between growing and nongrowing walls, when measured with
the creep method, is much larger than the other three extenso­
meter techniques described above. Moreover, creep assays are
more sensitive to the activity of both expansins and family-12
endoglucanases, making the creep assay more encompassing for
detecting wall-loosening activities with different mechanisms of
action and temporal signature.
There are two critical parameters for cell wall creep measure-
ments. First, one must establish a suitable force, which should be
large enough to cause cell wall creep, but not so large that break-
age becomes a problem. This is a matter of trial and error. In our
hands, single Arabidopsis hypocotyls (~0.1 mm diameter) give
good creep curves with 0.5 g-force; cucumber hypocotyls (~2 mm
diameter) creep well with 20 g-force. These axial forces are smaller
than the calculated axial forces that are generated internally by
cell turgor pressure in the living tissues.
The second parameter is the buffer: its pH, ionic strength,
and chemical nature influence cell wall creep. Sodium acetate, at
20 mM and pH 4.5, works well for most tissues in our hands.
At this pH endogenous expansins are active whereas pectin-
methyl esterase, whose activity inhibits cell wall creep, is inactive.
If enzymatic activities in the cell walls have not been inactivated
(for instance with a brief heat treatment), then the activity of
esterases and other enzymes attached to the cell wall may cause
changes in the cell wall itself and also lead to pH drift. Buffer
exchanges or higher buffer concentrations are potential solutions
to this drift, but high buffer concentrations reduce the creep rate.
We have tried a number of different buffers at the same pH and
concentration and we found >2× differences in creep rates,
­evidently due to the anion effects on cell wall creep. Some buffer
 Measuring In Vitro Extensibility of Growing Plant Cell Walls 301

anions, such as citrate, act as divalent chelators that can remove

pectin-bound calcium from the cell wall and thereby affect wall
physical properties.
Takahashi et al. (28) constructed a “programmable creep
meter” in which the axial force starts at low values and gradually
increases. In this way they were able to characterize creep rate as
a function of applied force and to make estimates of the yield
threshold for wall creep.

3.3. Conclusions For reasons discussed in the text, it is unlikely that cell wall exten-
sibility can be fully measured by in-vitro methods, as wall exten-
sibility is based not only on wall mechanics but also depends on
wall-loosening processes that are sensitive to ephemeral condi-
tions in the cell wall space. Nevertheless, the methods outlined
here can provide positive evidence for changes in wall structure
(indicated by elastic and plastic compliances) and changes in wall-
bound wall-loosening activities such as expansins and lytic
enzymes. These methods are particularly useful for investigations
of the molecular basis of cell wall extension and its dependence
on cell wall structure.

4. Notes

1. For these methods to be interpreted in terms of cell wall prop-

erties, it is important that the protoplast be disrupted so that
turgor pressure and mechanically driven water flows within
the tissue do not complicate the measurement. Turgor pres-
sure generates wall tension, making the cell wall stiffer, much
as a pressurized tyre is stiffer than a flaccid one. As a result, the
tensile properties of living tissues depend on both turgor pres-
sure and cell wall structure. If protoplasts are intact, deforma-
tions can be partially limited by water flows from the protoplast,
making for a more complicated interpretation of the results.
For these reasons, these assays are simpler to interpret when
the cell protoplasts are disrupted. On the other hand, there
may be circumstances where these complications are not
deemed important, for example see Abasolo et al. (10).
2. Sample variability typically requires 8–15 replicates to obtain
adequate confidence limits of the measurements. The source
of this variability is unknown but likely arises from differences
in the plant materials. Therefore, care in growing, harvesting,
selecting, and preparing the plant materials are important for
limiting variability.
3. Trace levels of metal ions such as copper, iron, and aluminum
are potent inhibitors of cell wall creep and stress relaxation.
Therefore, metal clamps and other metallic parts of the
302 Cosgrove

e­ xtensometer should be coated with epoxy, plastic, latex, or

other material to avoid interactions with the cell walls. This is
most important in methods entailing long measurement peri-
ods, e.g., the creep method and the stress relaxation method.
4. In some studies, researchers have boiled the cell walls in
methanol, then rehydrated in buffer prior to measurement.
This procedure disrupts the protoplast and inactivates most
enzymes bound to the cell wall sample. However, the proce-
dure does not inactivate expansins (26, 29) and also precipi-
tates polysaccharides, potentially causing irreversible changes
in wall mechanics. In most cases, this type of pretreatment is
best avoided.
5. This step (cuticle abrasion) is only needed when buffers or
other materials must be diffused into the wall samples or
when the walls must be thoroughly washed. It is commonly
omitted when measuring breaking strength and elastic and
plastic compliances.
6. If wall-bound enzymes are not inactivated by heat or other
treatment, care must be taken to limit enzymatic wall modifi-
cation, most importantly pectin demethylation and polysac-
charide hydrolysis, as such processes may modify the wall
behaviors measured in these assays.

Acknowledgements

This material is based upon work supported as part of The Center

for Lignocellulose Structure and Formation, an Energy Frontier
Research Center funded by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences under Award
Number DE-SC0001090. The research on cell wall creep was
supported by Award number DE-FG02-84ER13179 from the
Department of Energy Office of Basic Energy Sciences.

References

1. Cosgrove, D. J. (1993) Wall extensibility: its 5. Ray, P. M. (1987) Principles of Plant Cell
nature, measurement, and relationship to Growth. In Physiology of Cell Expansion
plant cell growth. New Phytol 124, 1–23. ­during Plant Growth (Symposium in Plant
2. Geitmann, A. (2006) Experimental approaches Physiology, Penn State Univ), Cosgrove DJ,
used to quantify physical parameters at cellu- Knievel DJ, eds. (American Society of Plant
lar and subcellular levels. Am J Bot 93, Physiologists: Rockville), pp. 1–17.
1380–1390. 6. Cosgrove, D. J. (1995) Measurements of wall
3. Burgert, I. (2006) Exploring the microme- stress relaxation in growing plant cells. Methods
chanical design of plant cell walls. Am J Bot Cell Biol 49, 229–241.
93, 1391–1401. 7. Cosgrove, D. J., Van Volkenburgh, E., and
4. Geitmann, A., and Ortega, J. K. E. (2009) Cleland, R. E. (1984) Stress relaxation of cell
Mechanics and modeling of plant cell growth. walls and the yield threshold for growth: dem-
Trends Plant Sci 14, 467–478. onstration and measurement by micro-pressure
 Measuring In Vitro Extensibility of Growing Plant Cell Walls 303

probe and psychrometer techniques. Planta 17. Veytsman, B. A., and Cosgrove, D. J. (1998)
162, 46–52. A model of cell wall expansion based on ther-
8. Cosgrove, D. J. (1987) Wall relaxation in modynamics of polymer networks. Biophys J
growing stems: comparison of four species 75, 2240–2250.
and assessment of measurement techniques. 18. Yuan, S., Wu, Y., and Cosgrove, D. J. (2001)
Planta 171, 266–278. A fungal endoglucanase with plant cell wall
9. Durachko, D. M., and Cosgrove, D. J. (2009) extension activity. Plant Physiol 127,
Measuring plant cell wall extension (creep) 324–333.
induced by acidic pH and by alpha-expansin. 19. Zerzour, R., Kroeger, J., and Geitmann, A.
J Vis Exp (25), 1263. (2009) Polar growth in pollen tubes is associ-
10. Abasolo, W., Eder, M., Yamauchi, K., Obel, ated with spatially confined dynamic changes
N., Reinecke, A., Neumetzler, L., Dunlop, J. in cell mechanical properties. Dev Biol 334,
W. C., Mouille, G., Pauly, M., Höfte, H. et al. 437–446.
(2009) Pectin may hinder the unfolding of 20. Bolduc, J. F., Lewis, L. J., Aubin, C. E., and
xyloglucan chains during cell deformation: Geitmann, A. (2006) Finite-element analysis
Implications of the mechanical performance of geometrical factors in micro-indentation of
of Arabidopsis hypocotyls with pectin altera- pollen tubes. Biomech Modeling Mechanobiol
tions. Mol Plant 2, 990–999. 5, 227–236.
11. Reiter, W. D., Chapple, C. C., and Somerville, 21. Ray, P. M., Green, P. B., and Cleland, R. E.
C. R. (1993) Altered growth and cell walls in (1972) Role of turgor in plant cell growth.
a fucose-deficient mutant of Arabidopsis. Nature 239, 163–164.
Science 261, 1032–1035. 22. Cosgrove, D. J. (1985) Cell wall yield proper-
12. Wainwright, S. A., Biggs, W. D., Currey, J. ties of growing tissues. Evaluation by in vivo
D., and Gosline, J. M. (1976). Mechanical stress relaxation. Plant Physiol 78, 347–356.
Design in Organisms. (Edward Arnold: 23. Yamamoto, R., Shinozaki, K., and Masuda, Y.
London), p. 423. (1970) Stress-relaxation properties of plant
13. Ryden, P., Sugimoto-Shirasu, K., Smith, A. cell walls with special reference to auxin
C., Findlay, K., Reiter, W. D., and McCann, action. Plant Cell Physiol 11, 947–956.
M. C. (2003) Tensile properties of Arabidopsis 24. Yamamoto, R., Kawamura, H., and Masuda,
cell walls depend on both a xyloglucan cross- Y. (1974) Stress relaxation properties of the
linked microfibrillar network and rhamnoga- cell wall of growing intact plants. Plant Cell
lacturonan II-borate complexes. Plant Physiol Physiol 15, 1073–1082.
132, 1033–1040. 25. Fujihara, S., Yamamoto, R., and Masuda, Y.
14. Cavalier, D. M., Lerouxel, O., Neumetzler, L., (1978) Viscolelastic properties of plant cell
Yamauchi, K., Reinecke, A., Freshour, G., walls II. Effect of pre-extension rate of stress
Zabotina, O. A., Hahn, M. G., Burgert, I., relaxation. Biorheology 15, 77–85.
Pauly, M. et al. (2008) Disrupting two 26. Cosgrove, D. J. (1989) Characterization of
Arabidopsis thaliana xylosyltransferase genes long-term extension of isolated cell walls from
results in plants deficient in xyloglucan, a growing cucumber hypocotyls. Planta 177,
major primary cell wall component. Plant Cell 121–130.
20, 1519–1537.
27. McQueen-Mason, S. J., and Cosgrove, D. J.
15. Zhong, R., Peña, M. J., Zhou, G. K., Nairn, (1995) Expansin mode of action on cell walls.
C. J., Wood-Jones, A., Richardson, E. A., Analysis of wall hydrolysis, stress relaxation,
Morrison, W. H., III, Darvill, A. G., York, W. and binding. Plant Physiol 107, 87–100.
S., and Ye, Z. -H. (2005) Arabidopsis fragile
fiber8, which encodes a putative glucuronyl- 28. Takahashi, K., Hirata, S., Kido, N., and Katou,
transferase, is essential for normal secondary K. (2006) Wall-yielding properties of cell
wall synthesis. Plant Cell 17, 3390–3408. walls from elongating cucumber hypocotyls in
relation to the action of expansin. Plant Cell
16. Kha, H., Tuble, S. C., Kalyanasundaram, S., Physiol 47, 1520–1529.
and Williamson, R. E. (2010) WallGen, soft-
ware to construct layered cellulose-hemicellu- 29. McQueen-Mason, S., Durachko, D. M., and
lose networks and predict their small Cosgrove, D. J. (1992) Two endogenous
deformation mechanics. Plant Physiol 152, proteins that induce cell wall expansion in
774–786. plants. Plant Cell 4, 1425–1433.
 Index

A Autofluorescent protein (AFP)

blue fluorescent protein (BFP)................................. 141
AFP. See Autofluorescent protein cyan fluorescent protein (CFP)..........141, 144, 146, 156
AGP. See Arabinogalactan proteins green fluorescent protein (GFP).......................141, 155,
Agrobacterium transformation.................................144, 145, 210, 214–216
147, 150, 158, 160, 211, 217, 258, 259, 264, 265 transformation of AFP fusion into a plant...... 142–145,
Agroinfiltration...................................................... 144, 265 147, 150
Alcohol-insoluble residue........................................... 46, 47 yellow fluorescent protein (YFP)............................. 141,
Antibodies 144–150, 156, 159–165
anti-mouse..................105, 106, 109, 146, 228, 238, 277 Autoradiography........................................ 56, 60, 61, 70, 79
anti-rat....................................... 105, 106, 109, 110, 277
biosupplies.........................................104, 128, 246, 277 B
carbosource................................................104, 228, 246
b-glucan..................................................180, 202, 204, 205
epitopes.............................. 104, 106, 110, 116, 128, 246
Bimolecular fluorescence complementation
PlantProbes.......................................104, 105, 246, 277
(BiFC).........................................142, 157, 162–165
primary antibodies....................................108, 109, 119,
Bioimaging..................................................................... 156
126, 138, 145, 150, 227, 228, 238, 277, 285, 288
Biosynthesis
secondary antibodies
cellulose..............141–151, 153, 154, 198, 221, 222, 274
coupled to FITC......................................... 106, 109
cell wall............. 1, 65, 142, 153–166, 198, 221, 222, 274
coupled to gold...................................106, 109–110,
Buffers.....................................18, 46, 48, 56, 58–62, 64–67,
126, 138, 228, 238
74, 76–78, 82–86, 91, 94–96, 98–100, 105–108,
Antibody labelling
110–112, 116–117, 121, 126, 144–150, 158–160,
embedding................................................................ 227
183, 186, 194, 211, 213, 223–225, 227, 229, 233,
fixing................................................................. 227, 243
234, 237, 238, 247, 249, 250, 252, 258–260, 262,
masking............................................................ 110, 112
263, 269, 270, 277, 278, 283, 284, 294, 299, 300, 302
pre-treatment.................................................... 110, 112
sectioning...................................................227, 238, 243 C
Arabidopsis thaliana
ecotypes Callus culture.........................................................1–3, 9, 11
Columbia.....................................226, 231, 235, 236 Capillary zone electrophoresis (CZE)................ 93, 96–101
Landsberg....................................226, 231, 235, 236 Carbohydrate
inflorescence............................................................. 295 anomeric centres...................... 23–24, 29, 31, 34, 35, 38
pollination........................................................ 231–232 ball-and-stick modelling............................................. 38
Arabinans.......................................................154, 179, 181, carbohydrate gel electrophoresis............81–91, 145, 159
184, 185, 193, 199 computerised modelling....................................... 21–39
Arabinogalactan proteins (AGPs) conformation..................... 21, 22, 24–27, 30–34, 36–38
detection drawings............................................................... 21, 24
ex situ............................................................246, 247 hydrogen bonding......................................33–34, 36, 37
in situ............................................................246–249 microarrays........................................104, 115–121, 245
extraction....................210, 211, 213, 214, 218, 245–252 nomenclature.......................................23–24, 30, 35, 37
Arabinogalactans............................................128, 154, 179, ring-puckering...................................................... 24–26
181, 210, 222, 245–252, 274 Carbohydrate-binding modules (CBMs).......103–112, 238,
Arabinoxylan.............................................50, 202, 204, 205 277, 285, 288

Zoë A. Popper (ed.), The Plant Cell Wall: Methods and Protocols, Methods in Molecular Biology, vol. 715,
DOI 10.1007/978-1-61779-008-9, © Springer Science+Business Media, LLC 2011

305
 The Plant Cell Wall: Methods and Protocols
306 Index
  

cDNA libraries........................................256, 260–261, 270 Conjugation of a fluorophore onto the reducing

Cell cultures.............................................. 18, 108, 109, 112, end of a sugar
124–125, 210, 211 fluorophore
Cell growth..............................................205, 221–243, 292 2-aminoacridone (AMAC).............81, 82, 85–88, 90
Cell plate................................................................124, 132, 8-aminonaphthalene-1,3,6-trisulfonic
135–137 acid (ANTS).................................81, 82, 84–85,
Cell plate assembly matrix...................................... 123, 136 87, 88, 93–96, 100–101
Cellulose negatively-charged sugars........................................... 81
biosynthesis, cellulose synthase (CESA).................. 144, partially-charged or uncharged sugars........................ 81
146–149, 154, 198, 221, 222 Crosses
cadoxen..................................................................... 121 F1 backcrosses.......................................................... 232
cellulose-based composites F2 backcrosses.................................................. 232–233
hemicellulose-containing............................ 197–207 F2 outcrosses............................................................ 233
pectin-containing........................................ 197–207 Cryofixation................................................................... 124
stability................................................200, 205–207 Culture medium...............................................1, 9, 11, 125,
staining with Calcofluor White................................ 112 206, 212, 213, 217–218, 250, 259
Cell wall Cytokinesis.....................................................123, 124, 127,
biosynthetic machinery............................................. 221 128, 131, 209
building-blocks..................................................... 55–79 CZE. See Capillary zone electrophoresis
creep.......................................... 292, 294, 296, 299–301
extensibility....................................................... 291–302 D
hydrolysis........................................................45, 73–75,
Disaccharide...........................23, 29, 30, 32, 35–39, 89, 100
84, 210, 302
DNA extraction, buffer................... 225, 233–234, 277, 283
integrity control................................................ 221–223
mechanical role.........................................117, 186, 199,
E
210, 291, 292, 294, 296–298, 301
molecular components...................................... 221, 222 Electrophoresis, buffer
mutants high-voltage paper electrophoresis
cesA3/cev1............................................................ 221 (HVPE)......................................... 56, 61, 64, 67, 77
mur1 .................................................................. 222 polysaccharide analysis using carbohydrate gel
mur3 .................................................................. 222 electrophoresis (PACE)........................................ 91
mur4 .................................................................. 222 Electrophoretic mobilities.........................56, 62–68, 77, 78
oligosaccharides..............................................43–52, 67, Embedding
71–72, 77, 84–86, 91, 94 LR-White resin.........................................105, 108, 237
polysaccharides, enzymatic digestion.............47, 49, 263 wax-embedding................................................ 106–108
primary cell wall.........................................13, 104, 179, EMS mutagenesis............................................223–224, 229
180, 182–185, 209–218, 292 Enzymes
proteins.............................................................209, 210, activity...............................................95–96, 98–99, 294
255–257, 265 cellulases............................................................. 5, 9, 10
proteome........................................................... 255–271 driselase...................................................................... 10
secondary cell wall.........................................13, 15, 105 endo-b-(1-4)-glucanases............................................. 45
secretome.................................................................. 255 endo-b-(1-4)-xylanases............................................... 45
stress relaxation.................................................292, 294, endo-polygalacturonase....................................45, 46, 50
298–299, 301, 302 glycosyl hydrolases.................................................... 154
structural components.............................................. 221 glycosyltransferase.......................................45, 154, 274
Charge:mass ratio................................ 56, 59, 65, 71, 73, 74 pectate lyase...............................................106, 110, 112
Cloning.......................................... 142–147, 149, 155, 156, pectin methyl esterase (PME)............. 45, 46, 50, 52, 84
158–160, 165, 166, 215, 216, 223, 233, 257, 259, use in combination with molecular
260, 268, 275, 278, 279, 287 probes..................................................104, 105, 111
Columns................................................. 56, 95, 96, 98, 159, use in OLIMP...........................................43–45, 47, 49
211–214, 217, 218, 228, 239, 269 use in protoplast isolation............ 2, 4, 5, 9, 10, 274, 294
Composites, stability.......................................200, 205–207 xyloglucan-specific endoglucanase
Comprehensive microarray polymer profiling (XEG).................................................45, 46, 50, 52
(CoMPP), quantification and analysis Ethylene................................................................. 126, 221
of arrays.............................................................. 119 Expansin.................................. 198, 292, 293, 297, 299–302
 The Plant Cell Wall: Methods and Protocols
Index   
307

Extensins.....................................................75, 77, 209–216 modelling and quantitative analysis of

Extensometer...................................292–295, 298–300, 302 tomographic models................................... 132–137
Extracellular matrix........................................................ 213 Immunocytochemistry, masking phenomenon............... 104
Immunohistochemistry...................................227, 237–238
F Immunolabelling
Fixing.............................. 15, 18, 32, 94, 105, 107, 108, 111, immunogold labelling................................109, 123–139
112, 128, 161, 169, 227, 237, 243, 278, 285, 294 using monoclonal antibodies.....................105, 107–112
Fluorography.........................................................56, 60, 70 using recombinant CBMs........................................ 109
Forward genetics............................................................ 222 section pre-treatments...................................... 110–112
Freeze-substitution..................................124–126, 128–131 Immunolocalisation........................................................ 137
Ion-exchange chromatography
G anion-exchange........................................................... 56
cation-exchange, preparation of resin............46, 51, 212
Galactans................................. 154, 179, 181, 184, 185, 193
Ionisation................................................... 43, 58, 63, 66, 78
Gateway cloning.............................................143, 145–147,
155, 275, 278–280, 287 J
GAX. See Glucuronoarabinoxylans
Gene targeting........................................................ 273–288 Jasmonate ( JA)............................................................... 221
Genetic screens............................................................... 221
K
Genotyping
sos5 . ..................................................225, 231, 233–235 Knock-outs......................................................142, 273–288
uge4-3 .............................................................. 231–235
Gluconacetobacter xylinus L
long-term storage............................................. 200–202 Labelled oligomers..........................................94–96, 98, 99
maintenance..............................................199, 201, 202 Laser-induced fluorescence detection....................... 93–102
revival....................................................................... 201 Live cell imaging
Glucuronoarabinoxylans (GAXs)........................... 180, 188 imaging analysis................................................ 148–149
Glycosyl hydrolases......................................................... 154 preparation of plants for imaging............................. 148
Golgi apparatus.......................................107, 123, 153, 155
Golgi vesicles, preparation.......................................... 46, 47 M
H Map based cloning................................................. 223, 233
Mass spectrometry, MALDI-TOF/MS..................... 43–54
High-pressure freezing............................125, 128, 131, 138 Matrix polysaccharides........................................... 179, 235
High-voltage paper electrophoresis (HVPE) Medium
apparatus, flat-bed system..................................... 58, 78 nutrient medium and salts
buffers................................ 56, 58–62, 64–67, 74, 76, 77 Agrobacterium, luria-Bertani (LB)
sugar-complexing.................................................. 59 medium................................................. 257, 259
volatile............................. 56, 58–60, 69, 78, 83, 252 Gluconacetobacter xylinus, Hestrin and Schramm
calibration........................................................64, 67, 68 (HS) medium................................................ 199
coolants......................................................57–59, 70, 78 plant
electro-endo-osmosis............................................ 61, 65 B5, 4, 5, 10, 12
electrophoretic mobilities....................56, 62–68, 77, 78 BCDAT medium (for Physcomitrella
electrophoretogram patens)................................................... 280, 282
elution of substances for further analysis........ 70, 71 BCD medium (for Physcomitrella
removal of contaminants................................. 65, 70 patens).....................................................275, 286
markers...............................56, 59–65, 69, 70, 72, 74, 75 Brown and Lawrence (modified)........................ 3, 4
Hydrolysis of plant cell wall materials.............................. 84 BY-2 culture medium......................................... 125
Hydroxyproline oligoarabinosides.................................... 75 inclusion of growth regulators...................... 3, 8, 16
Hydroxyproline-rich glycoprotein inclusion of undefined mixtures of organic
(HRGP)..............................................209, 214–216 substances....................................................... 16
Murashige and Skoog (MS)............................... 146
I
N6....................................................................... 3, 4
Image analysis standard germination medium (SGM)............... 224
gel imaging................................................................. 83 WPM................................................................. 3, 5
imageJ................................................151, 229, 241, 286 yeast, YPD medium............................................ 254
 The Plant Cell Wall: Methods and Protocols
308 Index
  

Medium (Continued) O
selection medium
galactose selection medium................................. 224 Oligosaccharide mass profiling (OLIMP).................. 43–52
sucrose selection medium....................258, 262, 270 Oligosaccharides....................................... 26, 39, 43–52, 67,
Metabolic precursors.................................................. 55–79 69, 77, 82–91, 94, 96, 97, 100, 102
Microscopy P
atomic force
Pectins
artifact..................................................170, 175, 176
homogalacturonan..............................................76, 104,
cantilever fouling................................................ 170
154, 179, 181
sample prepartion............................................... 169
rhamnogalacturonan I................................................ 94
confocal
rhamnogalacturonan II............................................. 179
fluorescence................................................ 259, 260
xylogalacturonan............................................... 154, 179
laser scanning microscopy
Phragmoplast..........................................123, 124, 134, 136
(CLSM).........................................156, 161, 163
Physcomitrella patens
electron tomography
caulonemal gravitropism assay.................................. 286
image acquisition........................................ 131–132
gametophore development assay............................... 286
image segmentation............................................ 132
genotype analysis.............................................. 277, 283
immunolabelling......................................... 137–138
protoplast
preparation of sections........................................ 126
preparation...................................................2, 5, 281
immunofluorescence................................................. 106
regeneration................................................ 281, 288
TEM........................................................................ 180
rhizoid development assay.................................2, 5, 281
Modelling.......................................................124, 126–127,
tip growth assay........................................................ 286
132–133, 135, 136
transformation.......................................................... 288
Molecular mechanics........................................................ 23
Picea abies callus culture...................................................... 9
Molecular probes
pKa of functional groups....................................... 58–61, 72
biosupplies................................................................ 104
Plant tissue cultures
carbohydrate-binding modules
callus...................................................... 1–3, 8, 9, 11, 17
(CBMs)...................................................... 103–113
carbon source........................................................ 2, 204
carbosource............................................................... 104
explant.............................................................2, 3, 8, 13
in combination with specific enzyme
growth conditions................................................. 8, 116
treatments................................................... 227, 246
growth regulators
monoclonal antibodies...................................... 103–113
auxins.................................................................. 3, 5
plantprobes.......................................................104, 105,
cytokinins................................................3, 5, 16, 18
246, 277
initiation..........................................................2, 7, 9, 16
Monoclonal antibodies........................................... 103–113
macroelements.................................................2, 5, 7, 17
Monosaccharides....................................... 73, 74, 84, 88, 89
maintenance........................................2–4, 16, 199, 201
microelements..................................................... 2, 5, 17
N
Nicotiana tabacum protoplast
Necrosis-inducing protein...................................... 263, 265 culture................................................................... 11
Nicotiana benthamiana..................................... 144, 150, 155, nutrient medium
157, 258, 265 B5............................................................4, 5, 10, 12
Nicotiana tabacum protoplast culture................................. 11 Brown and Lawrence (modified)........................ 3, 4
NIP screen.......................................256, 258–259, 263–265 BY-2 culture medium......................................... 125
NMR MS...................................................................... 3, 5
13
C-nuclei................................................................. 183 N6....................................................................... 3, 5
cross-polarisation NMR...................................180, 182, WPM................................................................. 3, 5
187, 189 organogenesis............................................................... 1
preparation of cell walls for NMR............................ 187 Picea abies callus culture................................................ 9
proton relaxation............................................... 181–183 protoplast
single-pulse excitation NMR.................................... 180 isolation................................................................ 10
solid-state 13C NMR........................................ 179–194 viability................................................................. 12
spin-echo (SE) NMR pulse sequence............... 182, 192 somatic embryo formation........................................ 1–2
Nucleotide sugars..............................................74, 154, 555 surface sterilisation....................................................... 2
Nutrient medium...........................................2–8, 10, 14–18 suspension-cultured cells............................................ 11
 The Plant Cell Wall: Methods and Protocols
Index   
309

tobacco bright yellow-2 (BY-2) Sequencing

cultured-cells, synchronisation.................... 127–128 Staining
tracheary element......................................................... 2 ex situ
xylogenic culture..................................................... 2, 13 acidic compounds, pH indicator..................... 61, 69
Zinnia elegans mesophyll cells................................. 2, 13 hydroxyproline...................................................... 69
Polymerase chain reaction (PCR)....................144–146, 153 isatin..................................................................... 69
Polysaccharide arabinogalactan proteins, Yariv........................... 214
epitopes..................................................................... 110 phenolics
methylesterification.................................................... 44 autofluorescence.................................................... 69
O-acetylation........................................................ 44, 45 folin-ciocalteu phenol reagent.............................. 75
oligosaccharide mass profiling of cell wall NH3 vapour.......................................................... 69
polysaccharides............................................... 43–54 molybdate reagent................................................ 69
polysaccharide analysis using carbohydrate gel starch, potassium iodide (I2KI) solution...... 171, 275
electrophoresis (PACE) aniline hydrogen phthalate........................60, 68, 75
conjugation of a fluorophore onto interference by borate............................... 68, 78–79
the reducing end of a sugar light-sensitivity..................................................... 68
negatively-charged sugars..................................... 77 in situ
partially-charged or uncharged sugars.................. 78 arabinogalactan proteins, Yariv........................... 214
electrophoresis buffer..........................56, 61, 64, 67, cellulose, Calcofluor White.........106, 109, 112, 238
77, 85, 86 lignin, phloroglucinol-HCl................................... 15
gel imaging..................................................... 83, 86 nuclei.................................................................... 15
sample preparation................................................ 82 protein, ponceau 2R solution...................... 185, 186
Primer..................................................... 144, 146, 149, 150, protoplasts to determine viability...................... 2, 12
159, 214, 225, 226, 234, 235, 237, 257, 260, 263, Stress response................................................................ 221
268, 269, 277, 279, 280, 283 Sub-cellular localisation.......................................... 152, 153
Profiling...........................................................100, 115, 255 Sugar-phosphates...................................................... 55, 69,
Programmed cell death................................................... 240 72–73
Protein-protein interaction..............................154, 162–165 Sugar signalling.............................................................. 221
Proton-spin relaxation editing (PRSE).......................... 182 Suspension-cultured cells............................................. 9, 11
Protoplasts
culture......................................................................... 18 T
generation..................................................281, 282, 288
TEM.............................................................................. 180
viability test.......................................................... 11, 12
Tobacco bright yellow-2 (BY-2) cultured-cells,
PRSE. See Proton-spin relaxation editing
synchronisation........................................... 127–128
Q Transfection.....................................................158–160, 162
Transformation.................................................10, 143–145,
Quantum mechanics......................................................... 23 147, 155, 158–160, 211–212, 215, 217, 258–262,
264–265, 270, 275–276, 281, 282, 287, 288
R
Radiolabelling U
14
C......................................................................... 56, 69
Uracil-excision cloning.................... 155, 156, 158–160, 166
detection
Uronic acids................................................................ 67, 71
autoradiography...................................56, 59, 60, 69
fluorography, pre-flashing film....................... 60, 70 V
3
H......................................................................... 56, 70
precursors.............................................................. 55–88 Vector construction................................................. 275, 287
in vivo..........................................................................73 Vector selection...................................................... 142–144
Resin embedding.....................................106, 107, 111, 130
W
S
Western blotting..............................................145, 147, 150
Screening.......................................................... 89, 114–120
Sectioning X
resin-embedded materials..........................104, 110, 111 X-ray diffraction............................................................. 180
wax-embedded materials.................................. 107–108 Xylan......................................... 45, 50, 84, 95, 98, 103–105
 The Plant Cell Wall: Methods and Protocols
310 Index
  

Xyloglucan staining..................................................................... 247

characterisation..................................................... 45–46 synthesis
digestion..................................................................... 45 HPLC........................................................ 228, 240
NMR.......................................................... 228, 240
Y Yeast secretion trap (YST)...................................... 256–257
Yariv Yeast transformation............................................... 258, 265
assay for sensitivity................................................... 229
radial gel diffusion assay................................... 247, 251
Z
reagents............................................. 211, 214, 218, 228,
238–240, 243, 246–252 Zinnia elegans.......................................................................2