2017 Book PlantHormones

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/310484689
Hormone Proﬁling in Plant Tissues
Chapter in Methods in molecular biology (Clifton, N.J.) · November 2017

DOI: 10.1007/978-1-4939-6469-7_20
CITATION READS
1 282
2 authors:
Maren Müller Sergi Munné-Bosch

University of Barcelona University of Barcelona
68 PUBLICATIONS 755 CITATIONS 224 PUBLICATIONS 8,258 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Oxidative stress as a master regulator of plant trade-offs View project
12th International Congress on Plant Biotechnology and Agriculture BIOVEG 2019 in Cuba (from May 27 to 31) View project
All content following this page was uploaded by Maren Müller on 22 August 2018.
The user has requested enhancement of the downloaded file.

Methods in
Molecular Biology 1497
Jürgen Kleine-Vehn
Michael Sauer Editors
Plant
Hormones
Methods and Protocols
Third Edition
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK
For further volumes:

http://www.springer.com/series/7651
Plant Hormones
Methods and Protocols
Third Edition
Edited by
Jürgen Kleine-Vehn
Department for Applied Genetics and Cell Biology, University of Natural
Resources and Life Sciences Vienna (BOKU), Austria
Michael Sauer
Department of Plant Physiology, Institute of Biochemistry and Biology,
University of Potsdam, Germany
Editors
Jürgen Kleine-Vehn Michael Sauer
Department for Applied Genetics Department of Plant Physiology
and Cell Biology (DAGZ) Institute of Biochemistry and Biology
University of Natural Resources and Universität Potsdam
Life Sciences Vienna (BOKU) Potsdam, Germany
Austria
ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-6467-3 ISBN 978-1-4939-6469-7 (eBook)
DOI 10.1007/978-1-4939-6469-7
Library of Congress Control Number: 2016955955
© Springer Science+Business Media New York 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
Printed on acid-free paper
This Humana Press imprint is published by Springer Nature

The registered company is Springer Science+Business Media LLC
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface
The field of Plant Hormone Biology is currently advancing at a rapid pace. In little more
than a decade we have seen an impressive development that both broadened and deepened
our knowledge of how small molecules influence plant physiology and development.
Besides the classical hormones, we are now aware of novel compounds that exert hormonal
functions, such as strigolactones, karrikins, or signaling peptides, where molecular signaling
pathways still unfold. At the same time the understanding of other hormonal pathways has
become so detailed that more and more sophisticated questions can be asked. Interactions
between members of signaling networks even at the quantitative level, cross-talk between
different hormonal pathways, or detailed molecular analyses of activity are now all within
the reach of the experimenter.
In this book, we aim to present a representative cross section of modern experimental
approaches relevant to Plant Hormone Biology. They range from relatively simple physi-
ological assays, which can be performed in any laboratory with standard equipment, to
highly sophisticated methods, which require specialized instrumentation. Some of the
chapters describe novel, previously undescribed methods, while others are refined varia-
tions of existing protocols. The first four chapters are dedicated to physiological and devel-
opmental assays. In line with the increasing demand for high-throughput methods, three
chapters on automated phenotyping follow. We tried to cover the wide spectrum of
microscopy-based techniques with six chapters ranging from response quantification to
four-dimensional tissue reconstruction. Mechanistic insight into hormonal pathways can
be gained by interaction studies, and four chapters outline different experimental
approaches. Traditionally, the plant hormone field has developed numerous analytical
methods to measure hormone contents, and we feature four examples of recent develop-
ments. Finally, we conclude with two chapters which outline how the use of heterologous
systems can significantly advance the field.
We trust the reader finds this book useful in a twofold way: On the one hand, it can be
used as a cookbook, which quickly aids in the setup of a particular experiment directly rel-
evant to the researcher’s interest. On the other hand, we encourage the reader to browse
through the chapters and explore whether some of the methods may be adapted also for
their own research.
We wish our readers the best of luck for their experiments!
Vienna, Wien, Austria Jürgen Kleine-Vehn

Potsdam, Germany Michael Sauer
v
Contents
Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1 Real-Time Analysis of the Apical Hook Development . . . . . . . . . . . . . . . . . . . 1

Qiang Zhu, Petra Žádníková, Dajo Smet, Dominique Van Der Straeten,
and Eva Benková
2 Grafting with Arabidopsis thaliana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Charles W. Melnyk
3 Tips and Tricks for Exogenous Application of Synthetic
Post-translationally Modified Peptides to Plants. . . . . . . . . . . . . . . . . . . . . . . . 19
Nathan Czyzewicz, Elisabeth Stes, and Ive De Smet
4 Assaying Germination and Seedling Responses of Arabidopsis to Karrikins . . . . 29
Mark T. Waters, Gavin R. Flematti, and Steven M. Smith
5 Low-Cost Microprocessor-Controlled Rotating Stage
for Medium-Throughput Time-Lapse Plant Phenotyping . . . . . . . . . . . . . . . . 37
Francis Barbez, Jürgen Kleine-Vehn, and Elke Barbez
6 Genome-Wide Association Mapping of Root Traits in the Context
of Plant Hormone Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Daniela Ristova and Wolfgang Busch
7 High-Throughput Scoring of Seed Germination . . . . . . . . . . . . . . . . . . . . . . . 57
Wilco Ligterink and Henk W.M. Hilhorst
8 Histochemical Staining of β-Glucuronidase and Its Spatial Quantification . . . . 73
Chloé Béziat, Jürgen Kleine-Vehn, and Elena Feraru
9 Imaging TCSn::GFP, a Synthetic Cytokinin Reporter,
in Arabidopsis thaliana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Jingchun Liu and Bruno Müller
10 Highlighting Gibberellins Accumulation Sites in Arabidopsis thaliana
Root Using Fluorescently Labeled Gibberellins . . . . . . . . . . . . . . . . . . . . . . . . 91
Hilla Schayek, Eilon Shani, and Roy Weinstain
11 In Silico Methods for Cell Annotation, Quantification of Gene Expression,
and Cell Geometry at Single-Cell Resolution Using 3DCellAtlas . . . . . . . . . . . 99
Petra Stamm, Soeren Strauss, Thomas D. Montenegro-Johnson,
Richard Smith, and George W. Bassel
12 Analyzing Cell Wall Elasticity After Hormone Treatment:
An Example Using Tobacco BY-2 Cells and Auxin . . . . . . . . . . . . . . . . . . . . . 125
Siobhan A. Braybrook
13 FRET-FLIM for Visualizing and Quantifying Protein Interactions
in Live Plant Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Alejandra Freire Rios, Tatyana Radoeva, Bert De Rybel, Dolf Weijers,
and Jan Willem Borst
vii
viii Contents
14 In Vivo Identification of Plant Protein Complexes Using IP-MS/MS . . . . . . . 147

Jos R. Wendrich, Sjef Boeren, Barbara K. Möller, Dolf Weijers,
and Bert De Rybel
15 Assaying Auxin Receptor Activity Using SPR Assays with F-Box Proteins
and Aux/IAA Degrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Mussa Quareshy, Veselina Uzunova, Justyna M. Prusinska,
and Richard M. Napier
16 Studying Transcription Factor Binding to Specific Genomic Loci
by Chromatin Immunoprecipitation (ChIP) . . . . . . . . . . . . . . . . . . . . . . . . . . 193
S. Vinod Kumar and Doris Lucyshyn
17 Hormone Receptor Glycosylation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Ulrike Vavra, Christiane Veit, and Richard Strasser
18 Highly Sensitive Salicylic Acid Quantification in Milligram Amounts
of Plant Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Víctor Carrasco Loba and Stephan Pollmann
19 High-Resolution Cell-Type Specific Analysis of Cytokinins
in Sorted Root Cell Populations of Arabidopsis thaliana . . . . . . . . . . . . . . . . . 231
Ondřej Novák, Ioanna Antoniadi, and Karin Ljung
20 Hormone Profiling in Plant Tissues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Maren Müller and Sergi Munné-Bosch
21 Use of Xenopus laevis Oocytes to Study Auxin Transport . . . . . . . . . . . . . . . . . 259
Astrid Fastner, Birgit Absmanner, and Ulrich Z. Hammes
22 Characterizing Auxin Response Circuits in Saccharomyces cerevisiae
by Flow Cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Edith Pierre-Jerome, R. Clay Wright, and Jennifer L. Nemhauser
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Contributors
BIRGIT ABSMANNER • Cell Biology and Plant Biochemistry, University of Regensburg,

Regensburg, Germany
IOANNA ANTONIADI • Department of Forest Genetics and Plant Physiology, Umeå Plant
Science Centre, Swedish University of Agricultural Sciences, Umeå, Sweden
ELKE BARBEZ • Department of Applied Genetics and Cell Biology (DAGZ),
University of Natural Resources and Life Sciences (BOKU), Vienna, Austria;
Gregor Mendel Institute (GMI) of Molecular Plant Biology, Vienna, Austria
FRANCIS BARBEZ • Department of Applied Genetics and Cell Biology (DAGZ),
University of Natural Resources and Life Sciences (BOKU), Vienna, Austria
GEORGE W. BASSEL • School of Biosciences, University of Birmingham, Birmingham, UK
EVA BENKOVÁ • Department of Life Sciences, Institute of Science and Technology Austria,
Klosterneuburg, Austria
CHLOÉ BÉZIAT • Department of Applied Genetics and Cell Biology (DAGZ),
SJEF BOEREN • Laboratory of Biochemistry, Wageningen University, Wageningen,
The Netherlands
JAN WILLEM BORST • Laboratory of Biochemistry, Wageningen University, Wageningen,
The Netherlands; Microspectroscopy Center, Wageningen University, Wageningen,
The Netherlands
SIOBHAN A. BRAYBROOK • The Sainsbury Laboratory, University of Cambridge, Cambridge,
UK
WOLFGANG BUSCH • Gregor Mendel Institute (GMI), Austrian Academy of Sciences Vienna
Biocenter (VBC), Vienna, Austria
NATHAN CZYZEWICZ • Division of Plant and Crop Sciences, School of Biosciences,
University of Nottingham, Loughborough, UK
BERT DE RYBEL • Laboratory of Biochemistry, Wageningen University, Wageningen, The
Netherlands; Department of Plant Systems Biology, Flemish Institute of
Biotechnology,VIB, Ghent, Belgium; Department of Plant Biotechnology and
Bioinformatics, Ghent University, Ghent, Belgium
IVE DE SMET • Division of Plant and Crop Sciences, School of Biosciences, University of
Nottingham, Loughborough, UK; Department of Plant Biotechnology and
Bioinformatics, Ghent University, Ghent, Belgium; Department of Plant Systems Biology,
Flemish Institute of Biotechnology (VIB), Ghent, Belgium; Centre for Plant Integrative
Biology, University of Nottingham, Loughborough, UK
ASTRID FASTNER • Cell Biology and Plant Biochemistry, University of Regensburg,
Regensburg, Germany
ELENA FERARU • Department of Applied Genetics and Cell Biology (DAGZ),
GAVIN R. FLEMATTI • School of Chemistry and Biochemistry, University of Western
Australia, Crawley, WA, Australia
ULRICH Z. HAMMES • Cell Biology and Plant Biochemistry, University of Regensburg,
Regensburg, Germany
ix
x Contributors
HENK W.M. HILHORST • Wageningen Seed Lab, Laboratory of Plant Physiology,

Wageningen University, Wageningen, The Netherlands
JÜRGEN KLEINE-VEHN • Department of Applied Genetics and Cell Biology (DAGZ),
University of Natural Resources and Life Sciences Vienna (BOKU), Vienna, Austria
S. VINOD KUMAR • Cell and Developmental Biology, John Innes Centre, Norwich, UK
WILCO LIGTERINK • Wageningen Seed Lab, Laboratory of Plant Physiology,
Wageningen University, Wageningen, The Netherlands
JINGCHUN LIU • Zurich-Basel Plant Science Center, Department of Plant and Microbial
Biology, University of Zurich, Zollikerstrasse, Zurich, Switzerland
KARIN LJUNG • Department of Forest Genetics and Plant Physiology, Umeå Plant Science
Centre, Swedish University of Agricultural Sciences, Umeå, Sweden
VÍCTOR CARRASCO LOBA • Centro de Biotecnología y Genómica de Plantas (CBGP),
Campus de Montegancedo, Pozuelo de Alarcón, Madrid, Spain
DORIS LUCYSHYN • Department of Applied Genetics and Cell Biology, University of Natural
Resources and Life Sciences (BOKU), Vienna, Austria
CHARLES W. MELNYK • The Sainsbury Laboratory, University of Cambridge, Cambridge, UK
BARBARA K. MÖLLER • Laboratory of Biochemistry, Wageningen University, Wageningen,
The Netherlands; Department of Plant Systems Biology, VIB Ghent University,
Ghent, Belgium
THOMAS D. MONTENEGRO-JOHNSON • School of Mathematics, University of Birmingham,
Edgbaston, Birmingham, UK
BRUNO MÜLLER • Zurich-Basel Plant Science Center, Department of Plant and Microbial
Biology, University of Zurich, Zollikerstrasse, Zurich, Switzerland
MAREN MÜLLER • Departament de Biologia Vegetal, Facultat de Biologia,
University of Barcelona, Barcelona, Spain
SERGI MUNNÉ-BOSCH • Departament de Biologia Vegetal, Facultat de Biologia,
University of Barcelona, Barcelona, Spain
RICHARD M. NAPIER • School of Life Sciences, University of Warwick, Coventry, UK
JENNIFER L. NEMHAUSER • Department of Biology, University of Washington, Seattle, WA, USA
ONDŘ EJ NOVÁK • Laboratory of Growth Regulators, Centre of the Region Haná
for Biotechnological and Agricultural Research, Institute of Experimental Botany
AS CR & Faculty of Science, Palacký University, Olomouc, Czech Republic
EDITH PIERRE-JEROME • Department of Biology, University of Washington, Seattle, WA, USA
STEPHAN POLLMANN • Centro de Biotecnología y Genómica de Plantas (CBGP), Campus de
Montegancedo, Pozuelo de Alarcón, Madrid, Spain
JUSTYNA M. PRUSINSKA • School of Life Sciences, University of Warwick, Coventry, UK
MUSSA QUARESHY • School of Life Sciences, University of Warwick, Coventry, UK
TATYANA RADOEVA • Laboratory of Biochemistry, Wageningen University, Wageningen,
The Netherlands
ALEJANDRA FREIRE RIOS • Laboratory of Biochemistry, Wageningen University,
Wageningen, The Netherlands
DANIELA RISTOVA • Gregor Mendel Institute (GMI), Austrian Academy of Sciences,
Vienna, Austria
HILLA SCHAYEK • Department of Molecular Biology and Ecology of Plants, Life Sciences
Faculty, University of Tel-Aviv, Tel-Aviv, Israel
EILON SHANI • Department of Molecular Biology and Ecology of Plants, Life Sciences
Faculty, Tel-Aviv University, Tel-Aviv, Israel
Contributors xi
DAJO SMET • Laboratory of Functional Plant Biology, Department of Physiology,

Ghent University, Ghent, Belgium
STEVEN M. SMITH • School of Biological Sciences, University of Tasmania, Hobart, TAS,
Australia; State Key Laboratory of Plant Genomics & National Center for Plant Gene
Research (Beijing), Institute of Genetics and Developmental Biology, Beijing, China
RICHARD SMITH • Department of Comparative Development and Genetics, Max Planck
Institute for Plant Breeding Research, Cologne, Germany
PETRA STAMM • School of Biosciences, University of Birmingham, Birmingham, UK
ELISABETH STES • Department of Plant Systems Biology, Flemish Institute of Biotechnology
VIB, Ghent, Belgium; Department of Plant Biotechnology and Bioinformatics, Ghent
University, Ghent, Belgium; Medical Biotechnology Center, Flemish Institute of
Biotechnology VIB, Ghent, Belgium; Department of Biochemistry, Ghent University,
Ghent, Belgium
RICHARD STRASSER • Department of Applied Genetics and Cell Biology, University
of Natural Resources and Life Sciences, BOKU, Vienna, Austria
SOEREN STRAUSS • Department of Comparative Development and Genetics, Max Planck
Institute for Plant Breeding Research, Cologne, Germany
VESELINA UZUNOVA • School of Life Sciences, University of Warwick, Coventry, UK
DOMINIQUE VAN DER STRAETEN • Laboratory of Functional Plant Biology,
Department of Physiology, Ghent University, Ghent, Belgium
ULRIKE VAVRA • Department of Applied Genetics and Cell Biology, University of Natural
Resources and Life Sciences, BOKU, Vienna, Austria
CHRISTIANE VEIT • Department of Applied Genetics and Cell Biology, University of Natural
Resources and Life Sciences, BOKU, Vienna, Austria
MARK T. WATERS • Australian Research Council Centre of Excellence in Plant Energy
Biology, University of Western Australia, Crawley, WA, Australia; School of Chemistry
and Biochemistry, University of Western Australia, Crawley, WA, Australia
DOLF WEIJERS • Laboratory of Biochemistry, Wageningen University, Wageningen,
The Netherlands
ROY WEINSTAIN • Department of Molecular Biology and Ecology of Plants, Life Sciences
Faculty, Tel-Aviv University, Tel-Aviv, Israel
JOS R. WENDRICH • Laboratory of Biochemistry, Wageningen University, Wageningen, The
Netherlands
R. CLAY WRIGHT • Department of Biology, University of Washington, Seattle, WA, USA
PETRA ŽÁDNÍKOVÁ • Institut für Genetik, Heinrich-Heine-University Düsseldorf,
Düsseldorf, Germany
QIANG ZHU • Department of Life Sciences, Institute of Science and Technology Austria,
Klosterneuburg, Austria
Chapter 1
Real-Time Analysis of the Apical Hook Development

Qiang Zhu, Petra Žádníková, Dajo Smet,
Dominique Van Der Straeten, and Eva Benková
Abstract
Mechanisms for cell protection are essential for survival of multicellular organisms. In plants, the apical
hook, which is transiently formed in darkness when the germinating seedling penetrates towards the soil
surface, plays such protective role and shields the vitally important shoot apical meristem and cotyledons
from damage. The apical hook is formed by bending of the upper hypocotyl soon after germination, and
it is maintained in a closed stage while the hypocotyl continues to penetrate through the soil and rapidly
opens when exposed to light in proximity of the soil surface. To uncover the complex molecular network
orchestrating this spatiotemporally tightly coordinated process, monitoring of the apical hook develop-
ment in real time is indispensable. Here we describe an imaging platform that enables high-resolution
kinetic analysis of this dynamic developmental process.
Key words Differential growth, Apical hook development, Hormonal cross talk, Real-time imaging,
Phenotype analysis
1 Introduction
To compensate for their sessile lifestyle, plants developed unique

mechanisms that provide them with an unusual level of develop-
mental plasticity. Bending of plant organs in response to gravi- and
photostimulation is a typical example of such plant-specific adapta-
tion strategies. A particularly intriguing process which comprises
the bending and consecutive unbending/straightening of the
upper hypocotyl is the development of the apical hook. The apical
hook is formed by the folding of the hypocotyl during early seed-
ling development to shield the tender shoot apical meristem and
cotyledons from damage while penetrating the soil. It is main-
tained in a closed stage while seedlings grow through the soil, and
rapidly open when exposed to light. Hence, apical hook development
gradually progresses through three distinct phases (formation,
maintenance, and opening phase), each of them depending on a
specific coordination of tissue and cell growth dynamics [1–3].
The common mechanistic basis underlying bending of various
Jürgen Kleine-Vehn and Michael Sauer (eds.), Plant Hormones: Methods and Protocols, Methods in Molecular Biology, vol. 1497,
DOI 10.1007/978-1-4939-6469-7_1, © Springer Science+Business Media New York 2017
1
2 Qiang Zhu et al.
200
formation maintenance opening Col
180
160
140
angles [°]
120
100
80
60
40
20
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120
hours after germination
Fig. 1 Real-time monitoring of the apical hook development. Hook growth was continuously recorded from
germination on and the angle of curvature was measured. Typically, the hook undergoes three developmental
phases: (1) formation, which is the period from seed germination until the hook angle reaches 180°; (2) main-
tenance, during which the hook remains fully closed; and (3) opening, when it gradually opens to reach an
angle of 0° (adapted from [12])
organs, including the apical hook, is differential growth, which,

through unequal rates of cell elongation at two opposite organ
sides, ultimately results in organ curving [4–6]. To form the hook
curvature the elongation rate of cells on the outer side of the upper
hypocotyl needs to exceed that of those on the inner side. In con-
trast the growth rate of cells on the concave (inner side) must exceed
that of cells on the convex (outer side) of the hook to straighten the
hypocotyl during the opening phase [1–3, 7] (Fig. 1).
Plant hormones are indispensable endogenous regulators of api-
cal hook development. Among them auxin plays a fundamental role.
Defects in auxin metabolism, transport, and signaling dramatically
affect all phases of apical hook development [8–15]. Particularly the
tightly controlled asymmetric auxin distribution is linked with dif-
ferential cell growth—the driving force of apical hook development
[12, 13, 16, 17]. Accumulation of auxin defines the concave side of
the apical curvature during formation phase, whereas balancing of
the auxin levels between the concave and convex side results in the
opening of the apical hook. Besides auxin, a multitude of other hor-
mone signaling pathways including that of ethylene, brassino-
steroids, and gibberellins coordinate this developmental process.
Increased ethylene levels in ethylene overproducer mutants (eto1,
eto2, eto3) result in an enhancement of the apical hook curvature
[6, 18, 19] whereas ethylene-insensitive mutants, such as ethylene
resistant1 (etr1) and ethylene-insensitive2 (ein2), exhibit a hookless
phenotype [6]. Gibberellins and brassinosteroids contribute to hook
establishment and their interaction with auxin and ethylene has been
described [6, 14, 20–25].
Apical Hook Development 3
Despite recent progress in dissecting the regulatory pathways

and complex hormonal network that guides the development of
the apical hook, we are still far from a full understanding of this
process. To hasten the elucidation of the molecular components
and mechanisms that control the progress of the apical hook
through its three phases of development, a reliable high-throughput
monitoring of the whole process is one of the major technical pre-
requisites. Molecular pathways that define (1) the kinetics of the
bending and angle of the hypocotyl curvature during the forma-
tion phase, (2) the kinetics with which the hypocotyl straightens in
response to light stimuli, as well as (3) the transition from forma-
tion to maintenance and from maintenance to opening cannot be
assessed unless employing continuous monitoring of the whole
process over time. Here we describe an imaging platform that
enables accurate kinematic analyses of apical hook development in
darkness and its opening in response to a light stimulus.
2 Materials and Chemicals
1. Seeds of Arabidopsis thaliana and mutants of interest.

2. Square Petri dishes (120 × 120 × 17 mm).
3. Calcium hypochlorite [Ca(ClO)2].
4. Triton-X-100.
5. Murashige and Skoog (MS) medium.
6. Sucrose (for plant tissue culture).
7. Agar for plant tissue culture.
8. Potassium hydroxide (KOH).
9. Optional: Phytohormones or other biologically active molecules
of interest.
2.1 Equipment 1. Dark growth chamber to accommodate plate, camera, and IR

light, a cube with 400 mm side length should be adequate (Fig. 2).
2. Infrared light source with 880 nm or alternatively 850 nm
wavelength, such as used in conjunction with surveillance
cameras (examples are 880 nm IR LED; Velleman, Belgium;
alternatively IR LED illuminator 850 nm; ABUS Security
Center, Affing, Germany).
3. Spectrum-enhanced digital camera which allows imaging of
infrared light and that can be remote controlled by a computer
and appropriate software. Frequently, this type of camera is sold
as optimized for astro-photography. For guidance, we give the
following examples: EOS 600D Canon Rebel T3i, 400DH
with built-in clear wideband-multicoated filter, equipped with a
standard 18–55 mm f3.5–5.6 lens and standard accessories
(Canon), operated by the EOS utility software (for one camera)
4 Qiang Zhu et al.
Fig. 2 Setup of the infrared imaging platform. Petri dishes are placed into a dark
box. Typically two Petri dishes are aligned for monitoring by one camera. As dur-
ing cultivation water might condense at the lid, Petri dishes are positioned with
the lid side facing away from the camera. The manual focus and automatic sta-
bilizer are used for image acquisition. The infrared light (IR-LED) is fixed in the
dark box to obtain homogenous illumination. The CCD camera is placed within
the dark box and connected to a computer. The camera is operated through
DSLR Remote Pro Multi-Camera software with adjustable frequency of picture
acquisition and synchronized switching on of the infrared light
or DSLR Remote Pro Multi-Camera software (for one or

Multiple Cameras). Another example would be Hercules opti-
cal glass USB-type CCD camera without an infrared filter
(Guillemot, La Gacilly, France), steered by Active WebCam v.4
software (PY Software, Etobicoke, Canada).
The infrared light is fixed in the dark box as indicated at

Fig. 2 to obtain homogenous illumination. The camera is
placed within the dark box and connected to a computer. The
camera is operated through DSLR Remote Pro Multi-Camera
software alternatively EOS utility software with adjustable fre-
quency of picture acquisition and synchronized switching on
of the infrared light. Synchronization can be set up using
appropriate software (e.g., Q Light Controller).
4. To monitor light-triggered opening of the apical hook a
computer-controlled light switch can be installed. This can be
quite simply achieved utilizing components frequently used for
stage and theatrical lighting adhering to the common DMX
standard. With these components, also light pulses of desired
spectral quality (by using RGB light strips) can be achieved. A free
open-source software solution is the program Qlight control-
ler (http://sourceforge.net/projects/qlc/). We suggest using
an LED dimmer controller in combination with an LED Red/
Green/Blue strip which is frequently used also for home/
ambient lighting applications. As technical advance and product
cycles in the lighting field are fast paced, we do not recommend
a specific product but rather encourage the reader to research
the market for a practical solution.
3 Methods
1. Sterilized seeds are plated on a square Petri dish containing 45 ml

of half-strength Murashige and Skoog (MS) medium with 1 %
sucrose and 0.8 % agar (pH 5.7 adjusted with KOH) and sealed
with one layer of micropore tape (3M MICROPORE). Optional:
According to the experimental design MS medium might be sup-
plemented with plant hormones or other biologically active mole-
cules. For optimal resolution two rows of seeds (15 seeds each) are
sown per plate. After stratification for 2 days at 4 °C in darkness,
seeds are exposed to light for 6 h at 21 °C (see Note 1).
2. Petri dishes are placed in a dark box at 21 °C. Typically two
Petri dishes are aligned for monitoring by one camera. As dur-
ing cultivation water might condense on the lid, Petri dishes
are positioned with lid facing away from the camera. Manual
focus and automatic stabilizer are used for image acquisition.
Typical frequency of the image acquisition is every hour for a
period of 8–10 days, during which seedlings progress from
formation to full opening of the apical hook.
3. To monitor light-triggered opening of the apical hook, the
light pulse of desired quality is applied for a defined time period
to apical hooks in the maintenance phase and their opening is
subsequently recorded.
6 Qiang Zhu et al.
Fig. 3 Kinetic analysis of the apical hook development. The angle of curvature α is defined as 180° minus the
angle formed by the tangential of the apical part and the axis of the lower part of the hypocotyl. In case the
bending exceeds 180° leading to formation of an exaggerated hook, the angle of curvature is defined as 180°
plus α (adapted from [12, 13])
4. The kinetic analysis of the apical hook development is performed

by the measurement of the angle between the hypocotyl axis
and cotyledons using ImageJ (NIH; http://rsb.info.nih.gov/ij).
Typically, 15–20 seedlings are processed. By convention, the
angle of curvature α is defined as 180° minus the angle formed
by the tangential of the apical part and the axis of the lower
part of the hypocotyl (Fig. 3) [12, 13]. In the case the bending
exceeds 180°, which leads to the formation of an exaggerated
hook, the angle of curvature is defined as 180° plus α (Fig. 3)
[13]. In our previous studies [12, 13, 25], the consecutive
phases of apical hook development were defined as follows: the
hook formation phase is the period from germination until the
time point at which the angle of hook curvature reaches 95 % of
its maximum value, and the maintenance phase comprises the
plateau in which hook angles differ at most 5 % from the maxi-
mum angle of curvature. This is then succeeded by the hook
opening phase resulting in full straightening of the hypocotyl
(see Note 2).
4 Notes
1. During early germination, the seed coat attached to the germi-

nating seedling might prevent reliable observation of the hook
bending. To improve visualization of this early hook formation
phase the seed coat can be peeled off. For these purposes seeds
are imbibed in water for 6 h at 4 °C in darkness and mature
Fig. 4 Monitoring of early phases of apical hook development. To improve visualization of early hook formation
phase the seed coat is peeled off and the mature embryo is dissected. The apical hook formation is monitored
as described
embryos are cautiously dissected from seed coats using tweezers.

For convenient manipulation a stereomicroscope placed in the
sterile bench might be used (Fig. 4).
2. Using the real-time platform for examination of apical hook
development the heterogeneity in seed germination might be
tolerated. This is one of the serious drawbacks of classical
“steady-state” assays in which hook curvatures of seedlings
grown in darkness are measured for a defined time period.
Here, variability in germination might lead to considerable
inaccuracies in an assessment of the apical hook developmental
phase. In contrast, continuous monitoring of growing seed-
lings allows to compensate for germination irregularities by
setting the time point zero for each individual seedling. Time
point zero is considered as the moment when the radicle pro-
trudes through the seed coat.
Acknowledgements
We thank Herman Höfte, Todor Asenov, Robert Hauschield, and

Marcal Gallemi for help with the establishment of the real-time
imaging platform and technical support. This work was supported
by the Czech Science Foundation (GA13-39982S) to Eva Benková.
Dominique Van Der Straeten acknowledges the Research
Foundation Flanders for financial support (G.0656.13N). Dajo
Smet holds a PhD fellowship of the Research Foundation Flanders.
References
1. Silk WK, Erickson RO (1978) Kinematics of 4. Darwin C, Darwin F (1881) The power of
hypocotyl curvature. Am J Bot 65:310–319 movement in plants. D. Appleton and Co,
2. Gendreau E, Traas J, Desnos T, Grandjean O, New York
Caboche M, Höfte H (1997) Cellular basis of 5. Boysen-Jensen P (1911) La transmission de
hypocotyl growth in Arabidopsis thaliana. l’irritation phototropique dans l’avena. Bulletin
Plant Physiol 114:295–305 Academie Des Sciences et Lettres de
3. Raz V, Ecker JR (1999) Regulation of differen- Montpellier 3:1–24
tial growth in the apical hook of Arabidopsis. 6. Guzmán P, Ecker JR (1990) Exploiting the
Development 126:3661–3668 triple response of Arabidopsis to identify
8 Qiang Zhu et al.
ethylene-related mutants. Plant Cell 2: 16. Kuhn H, Galston AW (1992) Physiological

513–523 asymmetry in etiolated pea epicotyls: relation
7. Raz V, Koornneef M (2001) Cell division to patterns of auxin distribution and phototro-
activity during apical hook development. Plant pic behavior. Photochem Photobiol 55:
Physiol 125:219–226 313–318
8. Boerjan W, Cervera MT, Delarue M, Beeckman 17. Žádníková P, Smet D, Zhu Q, Van Der Straeten
T, Dewitte W, Bellini C, Caboche M, Van D, Benková E (2015) Strategy of seedlings to
Onckelen H, Van Montagu M, Inzé D (1995) overcome their sessile nature: auxin in the
Superroot, a recessive mutation in Arabidopsis, mobility control. Front Plant Sci 6:218
confers auxin overproduction. Plant Cell 18. Vogel JP, Schuerman P, Woeste K, Brandstatter I,
7:1405–1419 Kieber JJ (1998) Isolation and characterization
9. Lehman A, Black R, Ecker JR (1996) of Arabidopsis mutants defective in the induc-
HOOKLESS1, an ethylene response gene, is tion of ethylene biosynthesis by cytokinin.
required for differential cell elongation in the Genetics 149:417–427
Arabidopsis hypocotyl. Cell 85:183–194 19. Woeste KE, Ye C, Kieber JJ (1999) Two
10. Zhao Y, Christensen SK, Fankhauser C, Arabidopsis mutants that overproduce ethylene
Cashman JR, Cohen JD, Weigel D, Chory J are affected in the posttranscriptional regula-
(2001) A role for flavin monooxygenase-like tion of 1-aminocyclopropane-1-carboxylic acid
enzymes in auxin biosynthesis. Science 291: synthase. Plant Physiol 119:521–530
306–309 20. Achard P, Vriezen WH, Van Der Straeten D,
11. Dharmasiri N, Dharmasiri S, Weijers D, Harberd NP (2003) Ethylene regulates
Lechner E, Yamada M, Hobbie L, Ehrismann Arabidopsis development via the modulation
JS, Jürgens G, Estelle M (2005) Plant develop- of DELLA protein growth repressor function.
ment is regulated by a family of auxin receptor Plant Cell 15:2816–2825
F box proteins. Dev Cell 9:109–119 21. Vriezen WH, Achard P, Harberd NP, Van Der
12. Žádníková P, Petrášek J, Marhavý P, Raz V, Straeten D (2004) Ethylene-mediated enhance-
Vandenbussche F, Ding Z, Schwarzerová K, ment of apical hook formation in etiolated
Morita MT, Tasaka M, Hejátko J et al (2010) Arabidopsis thaliana seedlings is gibberellin
Role of PIN-mediated auxin efflux in apical dependent. Plant J 37:505–516
hook development of Arabidopsis thaliana. 22. De Grauwe L, Vandenbussche F, Tietz O,
Development 137:607–617 Palme K, Van Der Straeten D (2005) Auxin,
13. Vandenbussche F, Petrásek J, Zádníková P, ethylene and brassinosteroids: tripartite control
Hoyerová K, Pesek B, Raz V, Swarup R, Bennett of growth in the Arabidopsis hypocotyl. Plant
M, Zazímalová E, Benková E et al (2010) The Cell Physiol 46:827–836
auxin influx carriers AUX1 and LAX3 are 23. An F, Zhang X, Zhu Z, Ji Y, He W, Jiang Z, Li
involved in auxin-ethylene interactions during M, Guo H (2012) Coordinated regulation of
apical hook development in Arabidopsis thaliana apical hook development by gibberellins and
seedlings. Development 137:597–606 ethylene in etiolated Arabidopsis seedlings.
14. Gallego-Bartolomé J, Arana MV, Cell Res 22:915–927
Vandenbussche F, Zádníková P, Minguet EG, 24. Gendron JM, Haque A, Gendron N, Chang T,
Guardiola V, Van Der Straeten D, Benkova E, Asami T, Wang ZY (2008) Chemical genetic
Alabadí D, Blázquez MA (2011) Hierarchy of dissection of brassinosteroid-ethylene interac-
hormone action controlling apical hook devel- tion. Mol Plant 1:368–379
opment in Arabidopsis. Plant J 67:622–634 25. Smet D, Žádníková P, Vandenbussche F,
15. Mazzella MA, Casal JJ, Muschietti JP, Fox AR Benková E, Van Der Straeten D (2014)
(2014) Hormonal networks involved in apical Dynamic infrared imaging analysis of apical
hook development in darkness and their hook development in Arabidopsis: the case of
response to light. Front Plant Sci 5:52 brassinosteroids. New Phytol 202:1398–1411
Chapter 2
Grafting with Arabidopsis thaliana

Charles W. Melnyk
Abstract
Generating chimeric organisms is an invaluable way to study cell-to-cell movement and non-cell-
autonomous actions of molecules. Plant grafting is an ancient method of generating chimeric organisms
and recently has been used to study the movement of hormones, proteins, and RNAs. Here, I describe a
simple and efficient way to graft Arabidopsis thaliana at the seedling stage to generate plants with roots
and shoots of different genotypes. Using this protocol, success rates of over 80 % with up to 80 grafts
assembled per hour can be achieved.
Key words Arabidopsis thaliana, Micro-grafting, Chimeric plants, Mobile molecules, Graft-
transmissible signal
1 Introduction
People have cut and joined together different plant varieties for
thousands of years to generate chimeric organisms that have
increased stress resistance, increased yields, or improved plant size
[1]. This technique, termed grafting, has been used more recently to
study the non-cell-autonomous actions of molecules including
RNAs, proteins, and hormones [2–4]. Although requiring technical
know-how and skill, grafting is far easier and less time consuming
than other methods used to generate chimeric plants such as tissue-
specific expression of a transgene or through the generation of sec-
tors following transposition, mutagenesis, or recombination.
Nonetheless, grafting has been limited to whole organ or tissue chi-
meras, such as the grafting of a leaf, inflorescence, or root system.
Arabidopsis grafting was first described over 20 years ago [5],
and since then, grafting with Arabidopsis has proven a very useful
and informative technique. Many tissues of Arabidopsis are suitable
for grafting including cotyledons [6], inflorescence stems [7],
developing leaves with the shoot apical meristem [8], the young
shoot/young root [2], and the mature shoot/mature root [5].
Most commonly, Arabidopsis grafting is performed on young
9
10 Charles W. Melnyk
seedlings between the shoot and root since many grafts can be
assembled rapidly and it allows a greater time for molecules to
move as the plant matures. Due to the small size of Arabidopsis
seedlings, this is a technically challenging process requiring the use
of a stereomicroscope. The below protocol is adapted from previ-
ously published butt-grafting protocols [2, 9, 10] and does not
require the use of tubing. With steady hands and sufficient practise,
high success rates (over 80 % with wild-type Columbia accession)
can be obtained with rapid rates (over 80 grafts per hour) for two
segment grafts. More advanced techniques such as three segment
grafts (where up to three genotypes can be grafted together as a
single plant) or Y-grafts (where one shoot is grafted to a second
intact plant) are also possible, but with lower success rates and more
time required.
Such shoot-root Arabidopsis grafting has become routine prac-
tice. To date, it has been used to study the movement of small
RNAs [4, 10, 11], nutrients [12, 13], secondary metabolites [14],
and hormones including jasmonic acid [15], strigolactones [2],
gibberellic acid [16], and cytokinin precursors [17]. It has also
been informative to study signals associated with flowering time
[3], leaf development [18], and disease resistance [19] and to
study vascular regeneration [9, 20].
2 Materials
1. Sterile Arabidopsis seed.

2. Ultra Fine Micro Knives (manufactured by Fine Science Tools;
catalogue number 10315-12; see Note 1).
3. Fine forceps.
4. Whatman 3MM Chr cellulose chromatography paper,
46 × 57 cm.
5. Hybond N membrane, 20 cm × 3 m.
6. Aluminium foil.
7. Plates with ½ Murashige and Skoog (MS) medium and 0.8–1 %
agar.
8. 9 cm Round Petri dishes.
9. Dissecting microscope.
10. Laminar flow hood.
11. 20 °C Growth cabinet.
12. 70 % Ethanol for sterilization.
13. Sterile, autoclaved water.
14. Parafilm.
Grafting Arabidopsis thaliana 11
3 Methods
3.1 Two Segment 1. Sprinkle or pipette out sterilized Arabidopsis seed on suitable
Shoot-Root Graft media such as ½MS plates with 0.8–1 % agar, leaving several
millimeters between seeds. Plates without sucrose and without
antibiotics work best, but 1 % sucrose can be included if neces-
sary. Leave at 4 °C in the dark for 2–7 days.
2. Move plates from the cold to a growth space set at 20–22 °C
with 80–100 μmol/m2/s of light. Mount plates vertically to
ensure correct hypocotyl growth. The growth space can be set
for either short-day (8-h light) or long-day (16-h light) condi-
tions. Short day-grown plants are grafted 7 days after moving
out. Long day-grown plants are grafted 5 days after moving
out. I prefer short day-grown plants, as the hypocotyls are
longer and easier to graft.
3. Cut the Whatman paper to ~8.5 cm circles, and Hybond mem-
brane to 2.5 × 4 cm strips. For each grafting Petri dish, two
Whatman circles and one Hybond membrane are required.
Cut strips of Whatman paper (approximately 3 × 8 cm long) to
use for adding or removing water during grafting. Wrap the
cut Whatman paper and Hybond membrane in aluminum foil
and autoclave to sterilize.
4. Perform grafting in a laminar flow hood under a dissecting
microscope that has been wiped clean with 70 % ethanol.
Forceps and micro knife (see Note 1) are kept sterile with 70 %
ethanol, but allowed to dry prior to grafting (see Note 2).
Residual ethanol in the grafting plate will inhibit graft
formation.
5. Add sterile water to an empty petri dish, and add two sterile
Whatman circles and then one sterile Hybond strip. Sterile
water works best for grafting. Liquid ½ MS can also be used
with lower efficiency. Sucrose reduces grafting efficiency and
should be avoided. Remove the water-soaked Hybond and
Whatman with forceps, allow excess water to drip, and then
place these in a new petri dish with one Hybond strip on top
of two circles of Whatman paper (Fig. 1). During grafting, the
Whatman paper maintains the correct moisture, whereas the
Hybond ensures that the roots don’t anchor and become
attached to the Whatman paper.
6. At 5 or 7 days of growth (see Subheading 3.1, step 2), move
6–12 Arabidopsis seedlings from the ½MS plate and place
these in a row on the Hybond membrane (Fig. 2a). Select
healthy seedlings that have straight hypocotyls and cotyledons
that are at right angles to the hypocotyl. Cotyledons that are
bent over the hypocotyl make for difficult grafting. The first
true set of leaves should be barley visible (Fig. 2b). Be careful
Arabidopsis seedlings
1X Nylon membrane
2X Whatman paper
9cm Petri dish
Fig. 1 Arabidopsis grafting setup. Two Whatman circles and one Hybond mem-
brane strip are hydrated, excess water left to drip off, and placed in a Petri dish.
Arabidopsis seedlings are then moved on top of the Hybond membrane
Fig. 2 Two-segment Arabidopsis hypocotyl grafting. Arabidopsis seedlings are placed on Hybond membrane,
cut (dotted lines), switched with the desired genotype (arrows), and reassembled (arrows) (a–f). Triangles
denote the graft junction. Before sealing, make sure that water is barely visible around the hypocotyl (f) and
not excessive (g). Plates are then sealed with Parafilm (h) and mounted vertically for 7–10 days to recover
before transfer to media or soil
when moving seedlings not to damage the hypocotyl, and root

or shoot apical meristem. If two genotypes are to be grafted to
each other, two rows of seedlings can be made (top row
contains the shoot genotype; bottom row contains the root
genotype). Alternatively, genotypes can be alternated in the
same row such as odd numbers one genotype, even numbers
the other.
7. Cut off and discard one cotyledon (Fig. 2b, c), usually the one
that is smaller, damaged, or bent suboptimally. This allows the
shoot to lie flat on the Hybond membrane. Leave the petiole
attached, as later this is useful for picking up the shoot. Make
a transverse cut through the hypocotyl close to the shoot (90°
to the axis of elongation; a butt-end cut). The cut should be as
clean and straight as possible, avoiding crushing the tissue
(Fig. 2b, c). Cuts made in the middle or lower portions of the
hypocotyl lead to adventitious root formation and graft failure.
Ensure that some water is visible around the plants, as this
facilitates cutting.
8. Take the cut shoot from one plant and place it close to the cut
root from a different plant. This can be accomplished by care-
ful pushing or by picking up the shoot via the exposed petiole
(Fig. 2d). The root hypocotyl should lie flat against the
Hybond membrane and not be moved. If it is not flat, roll the
hypocotyl by carefully pushing the hypocotyl at the
root/hypocotyl junction with closed forceps. Be careful not to
grab or crush the hypocotyl, as damage to this or the roots will
strongly inhibit graft formation.
9. To assemble the graft, keep the forceps closed and gently push
on the shoot apical meristem region, cotyledon, or petiole
until the cut shoot contacts the cut root (Fig. 2e). Push careful
to align and reposition while avoiding grabbing or damaging
the tissue. No tubing is used and adhesion of the two cut sur-
faces is sufficient. The level of moisture here is critical. Some
excess water is helpful for cutting, but if the plate is too dry, the
plants will stick to the forceps and, in extreme cases, wilt. Too
much water will make aligning and adhesion extremely diffi-
cult. For aligning and joining the pieces, water should be visi-
ble but not excessively pooling on the plate (Fig. 2f, g). Sterile
strips of Whatman paper (3 × 8 cm) are useful for removing or
adding sterile water (see Note 3). Alternatively, plates can be
left open in the flow hood to dry for a couple minutes.
10. Repeat this procedure until all the plants on the plate have
been grafted (see Note 4). To increase efficiency, cut all plants
on the plate at once with the micro knife at low magnification.
Discard the cut cotyledons and move the cut shoots. Carefully
roll any non-flat root hypocotyls. Use a higher magnification
to gently push the cut shoots onto the cut roots. If cutting and
alignment become progressively difficult as the plate dries out,
add extra water to the plate to facilitate grafting.
11. After grafting, place the lid on the petri dish (Fig. 2h). Water
should just be barely visible around some hypocotyls at this
stage (Fig. 2f) but not excessive (Fig. 2g). Excessive water will
lead to adventitious root formation. Seal the plates with one to
two layers of Parafilm and move these into the growth space at
20–22 °C. Elevated temperatures of 27 °C for 5–7 days are
helpful for graft recovery [2], but not necessary. Mount the
plates vertically in 80–100 μmol/m2/s of light in either short-
or long-day conditions. Under these conditions, root growth
of the grafted plants begins 4–6 days after grafting.
12. 7–10 days after grafting, inspect the plates. Successful grafts
have no adventitious roots (roots formed above the graft junc-
tion), are well attached, and show signs of new root growth
usually in the form of lateral roots. Primary root growth stops
after grafting, and does not usually resume. At this point,
transfer the grafts to soil or to media. With experience, this
method allows success rates of over 80 % and grafting rates of
approximately 80 grafts per hour depending on the genotype.
13. Inspect the plants 1–2 weeks after transferring to soil or media
to insure that no adventitious roots have formed and the plants
are growing normally. Those that have adventitious roots or
are not growing should be discarded as the graft has failed. For
the first week, plants on soil should have a propagator lid to
increase humidity to ensure efficient recovery. Do not bury the
graft junction, as this will promote adventitious rooting and
make the junction harder to inspect.
14. After the experiment is finished, typically when plants are
mature, plants can be inspected for adventitious root forma-
tion, though it can be difficult to detect the graft junction in
very mature plants. The most reliable indicator is to graft with
a shoot or root expressing a visual reporter (i.e., GUS or GFP)
[2], or genotype the grafted material.
3.2 Three-Segment 1. Three-segment (or interstock) grafts are set up in a similar

Graft (Interstock Graft) manner as two-segment grafts (Subheading 3.1). These types
of grafts are informative if the middle segment can block graft
formation or a graft-transmissible signal [9] and are used in
non-Arabidopsis species to improve graft compatibility [1].
The main difference compared to two segment grafts is that
two cuts are made in the hypocotyl instead of one (Fig. 3a).
The first cut is made in a similar location as for two-segment
grafts but the second cut should be approximately 1 mm below
the first cut. Longer segments can be used and are easier to
move but reduce the grafting success rates.
2. After cutting, you should end up with segments from the desired
shoot genotype, middle genotype, and root genotype (Fig. 3b).
Three-segment grafting is facilitated by having multiple rows of
plants (up to three), or by alternating genotypes in a row and
discarding the tissues not required. Care should be taken not to
mix up tissues of the various genotypes, so it is recommended
not to graft too many plants on one plate.
Fig. 3 Three-segment, self-, and Y-grafts. Three-segment (interstock) grafts are made by making two cuts in
the hypocotyl (dotted lines) and moving the segments as desired (arrows) (a and b). Note that middle segment
is moved by pushing either end with forceps (b). The lower junction is attached first (c), followed by the upper
junction (d). Self-grafts (two segment) are also made by cutting the hypocotyl twice (e), but the middle seg-
ment is discarded (f) before joining (g). Y-grafts involve cutting one hypocotyl halfway through, and the other
in a V pattern (h). The V segment is then attached to the partially cut hypocotyl (i and j). Triangles denote the
graft junction
3. After cutting, do not move root segments except for a gentle

roll at the root/hypocotyl junction to get the hypocotyl lying
flat if necessary. Move the middle segment by closing the for-
ceps and pushing at one cut end (Fig. 3b). Pushing at the side
(the epidermis) should be minimized as this damages the tissue.
Push the middle segment (using either cut end) to the cut root
until the two segments join (Fig. 3c). Maintain the correct
orientation of the middle segment, as upside-down segments
will not graft. This forms the lower junction.
4. Move the cut shoot into place using the petiole or by gentle
pushing of the cotyledon or meristem. Then close the for-
ceps and gently push the cut shoot onto the middle segment
(Fig. 3d). Care should be taken not to dislodge the lower
junction.
5. The remaining grafts are assembled, paying careful attention to
water levels, before the plate is sealed as in Subheading 3.1.
Transfer grafts to media or soil after 10 days since grafts take
longer to heal. Expect a lower grafting efficiency with three-
segment grafts, as adventitious roots can form above the upper
or lower junction. Those plants with adventitious roots should
be discarded. It is best to practise with two-segment grafts
before attempting three segment.
3.3 Two-Segment 1. Two-segment self-grafts are essentially the same protocol as

Self-Graft two-segment grafts (Subheading 3.1), except that two cuts are
made in the hypocotyl and the middle segment (~0.5 mm) is
discarded (Fig. 3e–g). The 0.5 mm piece is discarded to ensure
that the same surfaces are not simply realigned, and that new
tissue is used for the graft.
2. Self-grafts are used when the shoot and root need to be the same
genotype but the seeds are from a segregating population or the
phenotype is variable such as expression from a transgene. Self-
grafting has been useful for studying gene expression changes
at the graft junction and for dissecting the genetic require-
ments of graft formation [9]. Self-grafts are also useful controls
(see Note 4).
3.4 Two-Shoot 1. Y-grafts involve adding an additional shoot to an intact plant.

Y-Graft This type of graft is useful to test if the molecule of interest
moves within the shoot. For instance, Y-grafting was used to
demonstrate that the protein FT was mobile from shoot to
shoot [3].
2. Prepare plants and the grafting setup identical to that in
Subheading 3.1. Different genotypes can be set up in different
rows, or beside one another. On one side of the recipient
(intact) plant, cut off one cotyledon and make a diagonal cut
in the hypocotyl near the shoot that does not cut completely
through the hypocotyl (Fig. 3h). The cut should be approxi-
mately halfway through. Push gently the shoot with forceps
after cutting to slightly widen the hypocotyl cut (Fig. 3i).
3. For the donor shoot, cut off one cotyledon and then cut the
hypocotyl in a V-fashion near the upper part of the hypocotyl
(Fig. 3h, i). Using the petiole, move the donor shoot close to
the recipient hypocotyl. Using closed forceps, push the donor
shoot into place. Ideally, the donor shoot’s cotyledon should
face the opposite direction of the recipient shoot’s cotyledon
to ensure that there is sufficient room (Fig. 3j).
4. These grafts are time consuming and to avoid the plate from
drying out, fewer grafts should be made per plate. Once finished,
the plates are sealed with Parafilm and mounted vertically as in
Subheading 3.1.
4 Notes
1. Using an Ultra Fine Micro Knife is important for success, but

if unavailable, sharp double-sided razor blades will work,
though with lower success rates. The Ultra Fine Micro Knives
are fragile and take care not to bend or damage the tips. Usually
after several hundred grafts the knives need to be replaced to
maintain clean cuts.
2. Periodically dip the forceps and micro knife in 70 % ethanol and

leave to dry. Use aseptic technique, avoid or minimize sources
of contamination, and sterilize the hood and all materials with
ethanol prior to grafting. Contamination of the grafted plants
reduces the grafting efficiency or leads to a complete loss of
grafted material.
3. The amount of water is critical to success or failure. You
should see some water on the Hybond membrane when cut-
ting and grafting as it facilitates both processes. If the plate
starts to dry out while grafting, add a bit more sterile water as
this makes grafting easier. After making all grafts, the water
needs to be reduced usually by waiting a minute or two for the
plate to partially dry out. You should see a shimmer of water
that is hardly noticeable under a couple of the hypocotyls
(Fig. 2f). Do not have not much water before sealing (Fig. 2g).
One sign of too much water is that you will have a lot of
adventitious roots formed above the graft junction several
days after grafting. For cutting there should be visible water,
for grafting an intermediate level of water, and prior to sealing
the plates, a low level of water that is barely visible around
some of the hypocotyls.
4. Controls should be included in all experiments. In addition to
the grafted samples between different genotypes, genotypes
grafted to themselves are normally included. This control is
necessary to ensure that the effect observed is due to move-
ment of molecules, rather than an effect from grafting itself.
Occasionally ungrafted controls are also useful. These have one
cotyledon removed but are not cut in the hypocotyl region.
Controls can be included in a separate plate, or placed on a
second strip of Hybond membrane above or below the Hybond
membrane with grafted samples.
Acknowledgement
I thank Elliot Meyerowitz and Raymond Wightman for critical

reading. This work was funded by a Clare College Junior Research
Fellowship and through Gatsby Charitable Trust grants
GAT3272/C and GAT3273-PR1.
References
1. Melnyk CW, Meyerowitz EM (2015) Plant 3. Corbesier L, Vincent C, Jang S, Fornara F, Fan
grafting. Curr Biol 25:R183–R188 Q, Searle I, Giakountis A, Farrona S, Gissot L,
2. Turnbull CG, Booker JP, Leyser HM (2002) Turnbull C et al (2007) FT protein movement
Micrografting techniques for testing long- contributes to long-distance signaling in floral
distance signalling in Arabidopsis. Plant induction of Arabidopsis. Science 316:
J 32:255–262 1030–1033
4. Molnar A, Melnyk CW, Bassett A, Hardcastle GM, Gojon A et al (2011) High nitrogen
TJ, Dunn R, Baulcombe DC (2010) Small insensitive 9 (HNI9)-mediated systemic repres-
silencing RNAs in plants are mobile and direct sion of root NO3− uptake is associated with
epigenetic modification in recipient cells. changes in histone methylation. Proc Natl
Science 328:872–875 Acad Sci U S A 108:13329–13334
5. Rhee SY, Somerville CR (1995) Flat-surface 14. Andersen TG, Nour-Eldin HH, Fuller VL,
grafting in Arabidopsis thaliana. Plant Mol Biol Olsen CE, Burow M, Halkier BA (2013)
Rep 13:118–123 Integration of biosynthesis and long-distance
6. Yoo SJ, Hong SM, Jung HS, Ahn JH (2013) transport establish organ-specific glucosinolate
The cotyledons produce sufficient FT protein profiles in vegetative Arabidopsis. Plant Cell
to induce flowering: evidence from cotyledon 25:3133–3145
micrografting in Arabidopsis. Plant Cell Physiol 15. Gasperini D, Chauvin A, Acosta IF, Kurenda
54:119–128 A, Stolz S, Chetelat A, Wolfender JL, Farmer
7. Nisar N, Verma S, Pogson BJ, Cazzonelli CI EE (2015) Axial and radial oxylipin transport.
(2012) Inflorescence stem grafting made easy Plant Physiol 169:2244–2254
in Arabidopsis. Plant Methods 8:50 16. Ragni L, Nieminen K, Pacheco-Villalobos D,
8. Huang NC, Yu TS (2015) A pin-fasten graft- Sibout R, Schwechheimer C, Hardtke CS
ing method provides a non-sterile and highly (2011) Mobile gibberellin directly stimulates
efficient method for grafting Arabidopsis at Arabidopsis hypocotyl xylem expansion. Plant
diverse developmental stages. Plant Methods Cell 23:1322–1336
11:38 17. Matsumoto-Kitano M, Kusumoto T, Tarkowski
9. Melnyk CW, Schuster C, Leyser O, P, Kinoshita-Tsujimura K, Vaclavikova K,
Meyerowitz EM (2015) A developmental Miyawaki K, Kakimoto T (2008) Cytokinins
framework for graft formation and vascular are central regulators of cambial activity. Proc
reconnection in Arabidopsis thaliana. Curr Natl Acad Sci U S A 105:20027–20031
Biol 25:1306–1318 18. Van Norman JM, Frederick RL, Sieburth LE
10. Brosnan CA, Mitter N, Christie M, Smith NA, (2004) BYPASS1 negatively regulates a root-
Waterhouse PM, Carroll BJ (2007) Nuclear derived signal that controls plant architecture.
gene silencing directs reception of long- Curr Biol 14:1739–1746
distance mRNA silencing in Arabidopsis. Proc 19. Xia Y, Suzuki H, Borevitz J, Blount J, Guo Z,
Natl Acad Sci U S A 104:14741–14746 Patel K, Dixon RA, Lamb C (2004) An extra-
11. Pant BD, Buhtz A, Kehr J, Scheible WR (2008) cellular aspartic protease functions in
MicroRNA399 is a long-distance signal for the Arabidopsis disease resistance signaling. EMBO
regulation of plant phosphate homeostasis. J 23:980–988
Plant J 53:731–738 20. Yin H, Yan B, Sun J, Jia P, Zhang Z, Yan X,
12. Green LS, Rogers EE (2004) FRD3 controls Chai J, Ren Z, Zheng G, Liu H (2012) Graft-
iron localization in Arabidopsis. Plant Physiol union development: a delicate process that
136:2523–2531 involves cell-cell communication between scion
13. Widiez T, El Kafafi S, Girin T, Berr A, Ruffel S, and stock for local auxin accumulation. J Exp
Krouk G, Vayssieres A, Shen WH, Coruzzi Bot 63:4219–4232
Chapter 3
Tips and Tricks for Exogenous Application of Synthetic

Post-translationally Modified Peptides to Plants
Nathan Czyzewicz, Elisabeth Stes, and Ive De Smet
Abstract
The first signaling peptide discovered and purified was insulin in 1921. However, it was not until 1991 that
the first peptide signal, systemin, was discovered in plants. Since the discovery of systemin, peptides have
emerged as a potent and diverse class of signaling molecules in plant systems. Peptides consist of small
amino acid sequences, which often act as ligands of receptor kinases. However, not all peptides are created
equal, and signaling peptides are grouped into several subgroups dependent on the type of post-transla-
tional processing they undergo. Here, we focus on the application of synthetic, post-translationally modi-
fied peptides (PTMPs) to plant systems, describing several methods appropriate for the use of peptides in
Arabidopsis thaliana and crop models.
Key words Post-translationally modified peptide, Synthetic peptide, Arabidopsis, Cereal crops,
In vitro growth
1 Introduction
In addition to phytohormones, such as auxin, cytokinin, gibberel-

lin, and abscisic acid (see other chapters in this book), peptides also
impact on various aspects of plant growth and development [1–3].
The terms “peptide,” “peptide signal,” “signaling peptide,” and
“peptide hormone” (not to be confused with “signal sequences”
which are small domains of larger proteins that dictate cellular
localization) are often used synonymously, although all of these are
relatively vague terms. In mammalian models, there are three main
classes of hormones, comprising peptides, steroids, and amino acid
derivatives. Each of these classes are further grouped by mecha-
nism of signaling, namely autocrine (self), paracrine (adjacent/
nearby tissue), or endocrine (systemic) signaling [4]. In this regard,
“classical” plant hormones bear similarity to steroids (as isoprenoid
derivatives) or amino acid derivatives, and act systemically. Using
this system of classification, most plant peptides can be classified
as “paracrine peptide hormones”, although there are exceptions
19
20 Nathan Czyzewicz et al.
(e.g., Glycine max RHIZOBIA-INDUCED CLE1 and 2) [5],

which act in an endocrine manner. Peptides consist mainly of rela-
tively short chains of amino acid monomers linked by amide bonds
(<100 amino acids), and are—so far—generally acting as ligands
for receptor kinases, initiating signaling cascades that steer the
development of the plant or a specific organ, in response to envi-
ronmental, developmental or physiological stimuli. Most (but not
all) peptides are proteolytic cleavage products of larger precursor
sequences [1–3], and are found in two main families, post-transla-
tionally modified peptides (PTMPs) and cysteine-rich peptides.
Cysteine-rich peptides, as the name suggests, have multiple cyste-
ine residues, which form disulfide bridges as the peptide folds, and
are critical for maintaining the three-dimensional shape of the pep-
tide [1, 2]. Here, we focus on one class, namely PTMPs. PTMPs
are usually shorter than cysteine-rich peptides, and are subjected to
an array of modifications prior to proteolytic cleavage, which are
able to alter the bioactivity of the peptide, presumably by altering
the shape or affinity of the peptide for its receptor [6]. Since
PTMPs tend to exist as short 12–15 amino acid sequences, it is
relatively easy to procure synthetic forms of these peptide signals
(even with some modifications) and examine their activity on plant
models.
Functional analysis of PTMPs requires multiple approaches,
including (but not restricted to) loss-of-function (if available) and
gain-of-function (overexpression lines or application of synthetic
peptides) studies, analysis of expression patterns, and biochemical
ligand-receptor studies. Whilst not an ideal approach, in part due
to the off-target effects, exogenous application of synthetic vari-
ants of PTMPs can still give an idea of the peptide function; and
can similarly facilitate the study of downstream effects and the
mode-of-action.
In this chapter, we describe several PTMP-associated experi-
mental setups and highlight important aspects and considerations
in the use of synthetic peptide application. Key points to the use of
synthetic peptides that are described are:
● Pure peptide—check for contamination
● Define a relevant biological assay
● Use correct solvent for dissolution/stock creation
● Determine optimum working concentration experimentally
(and peptide should be active at low concentrations to increase
biological relevance)
● Determine best route of application for required experiment
● Complement with genetic loss- and gain-of-function
approaches and expression analysis
Applying Synthetic Peptides 21
2 Materials
2.1 Reconstitution 1. Synthetic peptide (several companies provide these).

and Dilution of Peptide 2. Water (ideally MilliQ 18 MΩ cm at 25 °C) (see Note 1).
3. −20 °C Freezer.
2.2 Analysis 1. Seed stocks (wild-type, reporters, and mutants of interest)

of Peptide Effect (acquired from for example NASC (http://arabidopsis.info/)
or ABRC (https://abrc.osu.edu/)).
2. Microcentrifuge tubes.
3. 70 % ethanol.
4. Sterilization solution: 4 % hypochlorite (NaOCl), 0.05 %
Tween 20.
5. Growth medium: 2.15 g/l Murashige and Skoog salts, 0.1 g/l
myo-inositol, 0.5 g/l, 2-(N-Morpholino)ethanesulfonic acid
hydrate. Adjust pH to 5.7 using 1 M NaOH (see Note 2).
6. 200 ml conical flasks.
7. Square petri dishes, 120 × 120 × 17 mm.
8. Glass test tubes 20 mm diameter × 150 mm height and appro-
priate sterilisable caps.
9. 50 ml sterile centrifuge tubes.
10. Pipette filler and sterile 25 ml pipettes.
11. Micropore tape.
12. Laminar flow cabinet.
13. 10, 200, and 1000 μl micropipettes and sterile tips.
14. Forceps.
15. Growth chamber (21 °C, continuous light).
16. −20 °C freezer.
17. Rotary shaker (21 °C, continuous light).
3 Methods
3.1 Initial Reconsti- 1. Determine the pI of the peptide using the “Compute pI/Mw”
tution and Dilution tool via ExPASy Bioinformatics Resource Portal.
2. To dilute the peptide to a consistent concentration, accounting
for purity of the peptide (see Note 3), calculate the total vol-
ume of solvent required using the following equation (Eq. 1)
prior to dissolution.
æ Mol ö
Volume required ( L ) = çç ÷÷ ´
è concentration required ( mol / L ) ø
æ Purity ( % ) ö
ç ÷
è 100 ø ( 1)
3. Reconstitution should always first be attempted in a small vol-

ume of sterile, distilled water, (ideally MilliQ 18 MΩ cm at
25 °C), but other chemicals may be required to adjust pH in
order to fully dissolve the peptide (see Note 4).
3.2 Sterilization 1. Distribute seeds into microcentrifuge tubes.

of Arabidopsis Seeds 2. Add 1 ml of 70 % ethanol, invert to ensure full coverage of
seeds and incubate at ambient temperature for 30 s. A rotary
shaker can be used to keep large amounts of seeds suspended
for even coverage.
3. Replace the ethanol with 1 ml of sterilization solution, invert
to ensure full coverage of seeds and incubate at ambient tem-
perature for 15 min. A rotary shaker can be used to keep large
amounts of seeds suspended for even coverage.
4. In a laminar flow cabinet, rinse the seeds three times in 1 ml
sterile, distilled water to remove sterilization solution.
5. Vernalize the seeds by storage at 4 °C in the dark for two
nights.
3.3 Sterilization of 1. Distribute seeds into 50 ml centrifuge tubes.

Wheat/Tomato/Oilseed 2. Add 30 ml of 70 % ethanol, invert to ensure full coverage of
Rape/Brachypodium seeds and incubate at ambient temperature for 30 s. A rotary
Seeds shaker can be used to keep large amounts of seeds suspended
for even coverage.
3. Replace the ethanol with 30 ml of sterilization solution, invert
to ensure full coverage of seeds and incubate at ambient tem-
perature for 30 min. A rotary shaker can be used to keep large
amounts of seeds suspended for even coverage.
4. In a laminar flow cabinet, rinse the seeds three times in 30 ml
sterile, distilled water to remove sterilization solution.
5. Vernalize the seeds by storage at 4 °C in the dark for 1 week.
3.4 Inclusion 1. Melt solid growth medium in a covered 100 °C water bath or
of Synthetic Peptide microwave (or use directly from the autoclave).
in Solid Medium 2. Cool media to <50 °C.
for Arabidopsis
3. Working in a laminar flow cabinet, use a 50 ml centrifuge tube
to dilute peptide to required working concentration in 50 ml
cooled media. Invert five times to mix.
4. Pour media into square petri dishes and allow to set.
5. Add approximately 10–20 surface-sterilized seeds (see

Subheading 3.2) using a 10 μl micropipette and sterile tips.
6. Seal plates with Micropore tape.
7. Place in growth chamber until required for analysis.
3.5 Inclusion 1. Melt solid growth medium in a covered 100 °C water bath or
of Synthetic Peptide microwave (or use directly from the autoclave).
in Solid Medium 2. Cool media to <50 °C.
for Crop Plants
3. Working in a laminar flow cabinet, aliquot required amount of
peptide stock for dilution in 20 ml media into sterile glass test
tubes.
4. Use a pipette filler and sterile 25 ml pipettes to dilute peptide
to required concentration in 20 ml cooled media. Pipette up
and down to mix, and allow to set.
5. Add a single surface-sterilized seed (see Subheading 3.3) to the
surface of media in each test tube using sterilized forceps.
6. Replace the sterile cap on each test tube and seal with Micropore
tape.
7. Place in growth chamber until required for analysis.
3.6 Inclusion 1. After preparing 100 ml aliquots of ½ MS liquid medium (see

of Synthetic Peptide Note 5) and sterilizing in conical flasks to be used for the sub-
in Liquid Medium sequent experiment, cool to ambient temperature.
for Arabidopsis 2. Working in a laminar flow hood, remove and retain aluminum
foil seal.
3. Add surface-sterilized seeds (approximately 50 mg dry weight
seeds per 100 ml media) using a 1000 μl micropipette and
sterile tips.
4. Peptide can be added at this stage by diluting it to the required
concentration in the liquid medium (see Note 6).
5. Reseal culture vessel with aluminum foil (sterilized with flask).
6. Place on rotary shaker at 90–100 rpm in optimal growth con-
ditions until required for analysis.
3.7 Direct 1. Dilute peptide to appropriate working concentration—deter-

Application of Peptide mined by dose–response analysis (see Subheading 3.8).
to Tissue 2. Autoclave or melt growth medium, cool to <50 °C.
3. Working in a laminar flow cabinet, measure out 50 ml of
medium using a sterile 50 ml centrifuge tube, aliquot into
square petri dishes and allow to set.
4. Add approximately 10–20 surface-sterilized seeds (see
Subheading 3.2) using a 10 μl micropipette and sterile tips.
5. Seal plates with Micropore tape.
6. Place in growth chamber until required for treatment.

7. In a laminar flow cabinet, remove seal, and apply peptide
directly to the tissue of interest.
8. Reseal plate and replace in growth chamber until required for
analysis.
3.8 Dose–Response 1. Create stocks of peptide at a range of concentrations between

Analysis 10 mM and 1 nM by serial dilution.
2. These stocks can be included in medium (usually 1:1000) as
detailed in Subheadings 3.3 and 3.4, giving a final concentra-
tion range of 10 μM to 1 pM.
3. Arabidopsis or crop plants can be grown on this medium as
detailed in Subheadings 3.3 and 3.4, respectively.
4 Applications
4.1 Determination To determine the effect of a synthetic peptide on plants, it must be

of Relevant Biological applied to a tissue, organ or even the whole plant to induce a bio-
Assay for Measuring logical effect. Since peptides are able to affect a range of develop-
Peptide Bioactivity mental, physiological and biochemical processes [2], it is important
to establish a relevant biological assay for determining bioactivity
either prior to, or upon first application of the peptide. This can be
at the level of a growth, physiological or biochemical read-out or
assessing the response of relevant markers [7, 8]. Transcriptional
fusions can illustrate the type of tissue in which the peptide acts or
is secreted from, allowing for some idea of whether any observed
phenotype is relevant. Reporter lines can give an indication of how
peptides can affect the downstream response [7]. With respect to
biological assays, examples include (but are not limited to) root
growth, meristem size, medium acidification, membrane depolar-
ization, and measuring oxidative burst [8–10]. Additionally, when
studying a large peptide family (e.g., CLAVATA3/EMBRYO-
SURROUNDING REGION (CLE) peptides), it is important to
factor for functional redundancy within the family. This can be
achieved by screening multiple peptides using the same biological
assay, and determining any difference in peptide bioactivity.
However, the best indication of relevance is given by comparison
with expression pattern analysis and loss-of-function mutants.
4.2 Screening A continuing question surrounding studies conducted in

for Homologous Arabidopsis, is whether any discovery made in this model organism
Signaling Modules is transferrable to other species. Utilizing synthetic peptides, it is
Across Multiple possible to treat other model organisms with synthetic peptides in
Species the same way as Arabidopsis, with some minor adaptations for the
size and amount of growth of the species.
4.3 Amino Acid It is possible to obtain a range of synthetic peptides with modified
Substitutions amino acid sequences from most peptide suppliers. Substitution of
amino acids with inert residues can be useful in providing insight
into peptide conformation, and also can reveal information on
exposed side chains important for interaction of peptide and recep-
tor. Replacement of amino acids in this manner can also occasionally
identify peptides that inhibit the function of the native peptide via
competitive inhibition—known as antagonistic peptides [11–13].
Alanine scanning is a useful technique for determining critical
amino acids for peptide function, and can additionally provide
insight into peptide secondary structure. Since peptides can be
synthesized directly, it is relatively simple and far less time consum-
ing to apply synthetic peptides where every amino acid is sequen-
tially replaced by an alanine (or another inert amino acid) to probe
for bioactivity than it is to create and transform constructs into
Arabidopsis. Occasionally, alanine scanning (and subsequent physi-
ological or biochemical analysis) can also lead to identification of
mutant peptides that are able to produce the opposite phenotype
compared to the unmodified peptide (putative antagonistic pep-
tides). Antagonistic peptide technology should be applied with
caution however, since there is no strict single modification or sub-
stitution of amino acid residues that works for every peptide, as
different peptides will interact with their individual receptors via a
different combination of exposed side chains and overall 3D struc-
ture [11, 13].
4.4 Modification Proteins are known to undergo multiple forms of modification,

of Peptides and peptides are no exception, often undergoing multiple forms of
modification in vivo prior to proteolytic cleavage. There are several
common post-translational modifications that are able to affect
peptide bioactivity, for example tyrosine sulfation and/or hydroxy-
prolination (and subsequent l-arabinosylation), are known to alter
the bioactivity of the peptide by adjusting the shape and/or bind-
ing affinity of the peptide to its receptor [6]. Peptides can be syn-
thesized with a number of modifications, which can be useful if the
user desires to determine if amino acid modification affects struc-
ture and/or bioactivity.
5 Caveats
5.1 Nonspecific Perhaps the most obvious caveat to peptide application is that the
Effects applied peptide is made available to cell types and tissues that
would not normally be exposed, and may produce nonspecific
effects in the studied plant. Since the mature sequence of many
peptides is reasonably short, there is often a relatively large amount
of similarity in the mature peptide sequences within a family, which
may activate receptors in the incorrect tissue and mislead the
analyst with an unrelated phenotype. From this point of view, it is

critical that any data generated by peptide application is supported
by a corresponding expression analysis, and—if possible—compari-
son with a phenotypic or biochemical analysis of knockout and
(mild) overexpression lines to determine if observed effects to the
plant correspond accordingly. Comparison with activity of similar
peptides can also help to negate any concern over nonspecificity.
5.2 Contamination There are multiple routes by which contamination can affect exper-
iments utilizing peptide application, including (but not restricted
to) contamination of the peptide during synthesis, stock solution
generation, and preparation of medium. It is important to confirm
that any observed effects on a biological model are indeed the
result of applying the peptide of interest and not some form of
contamination [e.g., see ref. 14 on flg22]. HPLC chromatogram
data (provided with synthesized peptide) should indicate any con-
tamination, however this is not infallible if the contaminant is also
a similarly sized/charged peptide or present in extremely low (but
sufficient for a response) quantities. If there is concern as to the
specificity of a peptide, obtaining a peptide of high purity decreases
the risk of contamination-related effects, and checking the supplied
LC chromatograph for additional peaks can be used to determine
whether any potential contaminant is present in the peptide stock.
Working with ultrapure water, using analytical grade reagents, and
observing sterile technique while preparing solutions and media
will further minimize any user-introduced contamination.
6 Notes
1. Other chemicals may be required for optimizing pH of recon-

stitution solution, according to peptide provider’s instructions.
2. For solid medium add 10 g/l plant tissue culture agar prior to
autoclaving.
For liquid medium, autoclave individual 100 ml aliquots in
200 ml conical flasks topped with an aluminum foil cap.
3. Peptide purity is important for minimizing contamination.
When purchasing synthetic peptides, first consider for what
purpose they are needed.
● For initial screening of peptide function, aliquots of approx-
imately 70 % purity are relatively inexpensive and can be
used to determine whether any of the peptides are able to
induce a change to morphology/biochemical processes.
● For further experiments, it is important to use peptides of
higher purity (at least >85 %, ideally higher), to safeguard
against contamination either from a biological or chemical
source within the synthesized peptide aliquot. If there is

particular concern over the biological effect of the peptide,
aliquots of >99 % purity can be obtained to ensure that any
observed effect is specific to the peptide.
4. If the peptide does not fully dissolve in water, further chemi-
cals may be added to adjust pH, allowing dissolution of the
peptide.
● Basic peptides are best dissolved in acidic solutions (acetic
acid/TFA).
● Acidic peptides are best dissolved in basic solutions
(NH4OH).
● Uncharged peptides are best dissolved in organic solvents
(acetonitrile/methanol/isopropanol) according to the
peptide provider’s instructions.
● If any chemical other than water is used in the preparation of
peptide, it is prudent to additionally create a blank solution,
to determine any effect of these chemicals on the organism,
tissue, or reporter system downstream of treatment.
5. It is important not to overfill the culture flasks, or the medium
will not remain in motion on a rotary shaker and will become
stagnant. Use 100 ml media in a 200 ml conical flask for up to
50 mg (dry weight) seeds. This can be scaled up for larger
weights of seeds.
6. Synthetic peptide can alternatively be added at a later stage if
analysis focuses on more short-term response to peptide
treatment.
References
1. Murphy E, Smith S, De Smet I (2012) Small 6. Shinohara H, Matsubayashi Y (2013) Chemical

signaling peptides in Arabidopsis development: synthesis of Arabidopsis CLV3 glycopeptide
how cells communicate over a short distance. reveals the impact of hydroxyproline arabino-
Plant Cell 24(8):3198–3217 sylation on peptide conformation and activity.
2. Czyzewicz N, Yue K, Beeckman T et al (2013) Plant Cell Physiol 54(3):369–374
Message in a bottle: small signalling peptide 7. Czyzewicz N, Shi CL, Vu LD et al (2015)
outputs during growth and development. Modulation of Arabidopsis and monocot root
J Exp Bot 64(17):5281–5296 architecture by CLAVATA3/EMBRYO
3. Tavormina P, De Coninck B, Nikonorova N SURROUNDING REGION 26 peptide.
et al (2015) The plant peptidome: an expand- J Exp Bot 66(17):5229–5243
ing repertoire of structural features and bio- 8. Qi Z, Verma R, Gehring C et al (2010) Ca2+
logical functions. Plant Cell 27(8):2095–2118 signaling by plant Arabidopsis thaliana Pep
4. Petraglia F, Florio P, Nappi C et al (2013) peptides depends on AtPepR1, a receptor with
Peptide signaling in human placenta and mem- guanylyl cyclase activity, and cGMP-activated
branes: autocrine, paracrine, and endocrine Ca2+ channels. Proc Natl Acad Sci U S A
mechanisms. Endocr Rev 17(2):156–186 107(49):21193–21198
5. Reid D, Li D, Ferguson B et al (2013) Structure– 9. Pearce G, Strydom D, Johnson S et al (1991)
function analysis of the GmRIC1 signal peptide A polypeptide from tomato leaves induces
and CLE domain required for nodulation con- wound-inducible proteinase inhibitor proteins.
trol in soybean. J Exp Bot 64(6):1575–1585 Science 253(5022):895–897
10. Butenko M, Wildhagen M, Albert M et al 13. Xu T, Ren S, Song X et al (2015) CLE19

(2014) Tools and strategies to match peptide- expressed in the embryo regulates both cotyle-
ligand receptor pairs. Plant Cell 26(5):1838– don establishment and endosperm develop-
1847 ment in Arabidopsis. J Exp Bot 66(17):
11. Czyzewicz N, Wildhagen M, Cattaneo P et al 5217–5227
(2015) Antagonistic peptide technology for 14. Mueller K, Chinchilla D, Albert M et al (2012)
functional dissection of CLE peptides revisited. Contamination risks in work with synthetic
J Exp Bot 66(17):5367–5374 peptides: flg22 as an example of a pirate in
12. Song XF, Guo P, Ren SC et al (2013) commercial peptide preparations. Plant Cell
Antagonistic peptide technology for functional 24(8):3193–3197
dissection of CLV3/ESR genes in Arabidopsis.
Plant Physiol 161(3):1076–1085
Chapter 4
Assaying Germination and Seedling Responses

of Arabidopsis to Karrikins
Mark T. Waters, Gavin R. Flematti, and Steven M. Smith
Abstract
Karrikins are a small family of naturally occurring plant growth regulators present in the smoke and char
produced from burning plant material in wildfires. They can stimulate germination of dormant seed and
can influence seedling morphogenesis. Although Arabidopsis thaliana is not considered to be a smoke-
responsive species, karrikins will stimulate seed germination under the appropriate circumstances and will
cause repression of hypocotyl elongation in low light. This chapter describes how to conduct assays of the
activity of karrikins on Arabidopsis seeds and seedlings. The methods presented can potentially be modified
for use in a range of Arabidopsis genotypes or in other plant species.
Key words Karrikins, Arabidopsis thaliana, Seed germination, Dormancy, Hypocotyl elongation,
Photomorphogenesis
1 Introduction
Originally isolated from smoke from burning plant material, kar-

rikins (KAR) stimulate the germination of seed in a wide variety of
plant species. In Arabidopsis thaliana, freshly harvested seed exhib-
its primary dormancy, such that the seed will not germinate readily
under conditions that are otherwise permissive for germination.
The depth of primary dormancy depends on the ecotype in ques-
tion and the growth conditions experienced by the parental plant.
However, in many cases KAR at micromolar concentrations can
overcome this primary dormancy [1]. Besides germination, KAR
also influence seedling development and light sensitivity: KAR-
treated seedlings show increased inhibition of hypocotyl elonga-
tion, and greater cotyledon expansion [2]. Consistent with these
effects, genetic studies indicate that KAR act through a signaling
pathway that promotes germination and seedling responses to
light [3, 4]. Thus, KAR are potent exogenous plant growth regula-
tors that have revealed a distinct signaling mechanism of plant
development.
29
30 Mark T. Waters et al.
Here we describe two generic methods for measuring the

effects of KAR on Arabidopsis—though they could equally be
applied to other plant growth regulators, and to other species. A
key aspect of detecting an effect of KAR on seed germination is
obtaining suitably dormant seed: if the seeds are too dormant,
KAR will not be effective, and if the seeds have lost their primary
dormancy through prolonged storage (after-ripening) then no fur-
ther stimulation of seed germination will be detectable. As dor-
mancy is highly variable across genotypes, ecotypes, and growth
environments, there is no “one-size-fits-all” approach. However,
we believe that the method we describe here will assist researchers
by serving as a guide for optimizing experimental conditions.
Observing effects of KAR on hypocotyl elongation requires atten-
tion to the light conditions but is potentially more transferable
between different genotypes or plant species.
2 Materials
2.1 Seed 1. Round 6 cm diameter Petri dishes.

Germination 2. Growth cabinet, temperature controlled.
Equipment
3. Low-magnification stereomicroscope.
4. Ultralow-temperature freezer for seed storage.
5. Microfuge tubes or seed envelopes for storage.
2.2 Seed 1. Arabidopsis seed of desired genotype.

Germination Reagents 2. Bacto-agar or phytoagar.
and Materials
3. Plant growth regulator stock solutions (1000× or greater).
4. Silica gel desiccant.
2.3 Hypocotyl 1. Round 6 cm diameter Petri dishes.

Elongation Assay 2. Growth cabinet, temperature controlled with LED red light
Equipment source (see Note 1).
3. High-resolution digital camera on a camera stand (SLR with
macro lens is recommended).
4. PC running ImageJ software (http://imagej.nih.gov/ij/) or
similar (see Note 2).
2.4 Hypocotyl 1. 0.05 % (v/v) Triton-X100 in 70 % (v/v) ethanol.

Elongation Assay 2. 100 % (v/v) ethanol.
Reagents
3. Arabidopsis seed of desired genotype.
4. Half-strength (0.5×) Murashige and Skoog (MS) basal medium,
pH 5.9 adjusted with KOH, solidified with 0.7 % (w/v) agar.
5. Plant growth regulator stock solutions (1000× or greater, in
appropriate solvent).
Assaying Responses to Karrikins 31
3 Methods
3.1 Generating Seed 1. Grow Arabidopsis under consistent, controlled conditions;

for Germination artificial growth facilities that maintain light levels, tempera-
ture, and humidity are highly recommended. Recommended
conditions for enhancing primary seed dormancy and rapid
growth are long (16 h) days, 22 °C day/18 °C night, 60 %
relative humidity, and white light intensity 100–150 μmol/
m2/s. Recent data indicate that shifting the plants to a lower
temperature (16 °C) after transition to flowering dramatically
increases seed dormancy [5]; this technique may be beneficial
for generating seed of Arabidopsis accessions with otherwise
little to no primary dormancy (e.g., Columbia-0).
2. Grow sufficient maternal plants to yield at least three independent
seed batches for each genotype, with 3–4 plants contributing to a
single batch. The plants should be randomized with respect to
position within a tray; ideally, rotate the positions weekly to ensure
that all plants grow at approximately similar rates. If your geno-
types vary substantially in flowering time, then you may need to
determine this in advance, and stagger the growth such that
flowering occurs synchronously among all genotypes.
3. Allow the maternal plants to set seed and the siliques to mature.
Inevitably, early siliques will mature first, and thus the seed
collected from any given plant will be heterogeneous in age. In
our view, it is impractical to collect seed from individual siliques
as they mature, but for very subtle germination phenotypes, a
staged collection strategy might be necessary. In general, the
longer the seed is allowed to dry on the plant, the sooner the
primary dormancy is lost. Therefore, the time of harvesting is
a compromise between obtaining a suitable amount of mature
seed and maintaining a useable level of primary dormancy.
Finding the right time to harvest is critical and depends on the
particular growth conditions. It must be determined empiri-
cally, perhaps by harvesting seed from different plants at vary-
ing levels of maturity over a period of 4 weeks. Typically we
stop watering once flowering has ceased (the primary shoot
apex has terminated), but before all siliques have turned brown.
The youngest siliques will still be maturing at this stage, but
these can be removed before seed harvesting. We then harvest
seed about 1 week later.
4. Harvest the seed and pool from 3 to 4 plants to form one seed
batch. Collect the seed in small envelopes or in 1.5 ml
microfuge tubes. Envelopes are easier to label and will assist
subsequent drying, but are perhaps less convenient to handle
and store in the freezer.
5. Place the seed under silica gel desiccant at room temperature
for 48–72 h, to improve viability of seed under storage. A plas-
tic food container containing anhydrous silica gel is suitable for

this purpose (see Note 3).
6. Store the seed in an ultralow-temperature freezer (−70 to
−86 °C). This reduces the rate of loss of primary dormancy. If
the seeds are stored at higher temperatures, after-ripening will
gradually occur and the effects of KAR will be hard or impos-
sible to observe (see Note 4).
3.2 Conducting Before setting up a very large experiment, we suggest testing one
Germination Tests genotype—typically wild type—under different incubation condi-
tions. Consider the gelling agent (step 1), the incubation tempera-
ture, and the diurnal period (step 6). When you have something
that appears to work well, set up the full experiment.
1. Prepare 1 % (w/v) agar in milliQ water by autoclaving. We have
used bacto-agar with success in the past, but we have found that
certain batches of agar inhibit germination. Phytoagar is more
consistent; agarose could also be considered. If required, add
chemical compounds using 1000× stocks in acetone (e.g., 1 mM
KAR1), and mix well. When pouring the plates, 6 cm petri dishes
hold 7–8 ml of agar. Allow plates to dry out for about 30 min in
a laminar flow hood, to avoid condensation later. Use fresh
plates each time you set up an experiment. When labeling plates,
we recommend a “blind” approach.
2. Surface sterilize the seed. A convenient method when dealing
with many samples is by exposure to gaseous chlorine:
Dispense sufficient seed into 1.5 ml microfuge tubes, and
place in a desiccator jar that contains 100 ml 10 % (w/v)
sodium hypochlorite solution. In a fume cupboard, add 3 ml
concentrated HCl to the solution, and quickly close the jar.
Leave for about 3–4 h. For an alternative sterilization method,
see Subheading 3.3 below.
3. Sow seed, about 100 per plate. Punch a small hole in the top
of the tube with a pin or needle, invert the tube, and tap to
sprinkle the seed on the plates. Try to avoid clumps of seed.
4. Seal the plates with gas-permeable tape, often referred to as
“surgical tape” or “Leukopor”®.
5. On the underside of the plate mark out with a small black dot,
four groups of 25 seeds (100 in total) to make germination
counting easier. Because the seeds can move or clump together
while imbibing, it is best to mark the seeds a couple of hours
after sowing, but before germination starts.
6. Incubate the plates under white light, 100–150 μmol/m2/s, at
20–22 °C. If the seeds are not very dormant, you can slow down
germination by increasing the temperature to 28 °C (thermoin-
hibition). This often makes the effect of germination stimulants
easier to detect. We typically use constant light conditions, but
others report 12-h light:12-h dark with success [5].
100
mock
90
Cumulative germination, percent ± SE

1 µM KAR1
80 1 µM KAR2
70 nitrate
60
50
40
30
20
10
0
0 40 60 80 100 120
Time since sowing, h
Fig. 1 Germination response of primary dormant Arabidopsis seed (ecotype

Landsberg erecta) to karrikins. Seed were sown on water-agar supplemented
with 0.1 % (v/v) acetone (mock) or with 1 μM of two different karrikins, KAR1 and
KAR2. For “nitrate,” seeds were sown on 0.5× MS medium + 1 % (w/v) agar,
stratified in the dark for 72 h, and transferred to the same growth conditions as
the other treatments. Note that the germination response to karrikins is not as
strong as to stratification in the presence of nitrates. Germination was inspected
every 24 h. Error bars are mean ± SE of n = 3 seed batches
Count germinated seeds at least every 24 h using a stereomi-

croscope. Classify a seed as germinated when you can see unam-
biguous radicle emergence (e.g., 1 mm), and mark such seeds with
a different color. Be aware that some inks fade rapidly under fluo-
rescent light in incubators! Failure to inspect plates regularly will
make it hard to detect germination, as the first seeds to germinate
can quickly obscure the others. Length of time to germination
depends on many factors—ecotype, incubation temperature, and
seed age. With Landsberg erecta, germination normally starts after
48 h, and Col-0 within 24 h. Most experiments have run their course
within 6 days; cumulative germination expressed as a percentage
typically follows a sigmoidal curve (Fig. 1).
For statistical analysis, we treat each seed batch within a geno-
type/treatment as an independent replicate. Because the germina-
tion data are expressed as percentages, standard procedure is to use
the arcsine transformation before statistical analysis. The entire
experiment should be repeated—ideally with different batches of
seed—to ensure that a conclusion is robust.
3.3 Hypocotyl For this assay, seedlings are grown in the presence of KAR under
Elongation Assays red light (see Note 1). Seedlings grown under red light have much
longer hypocotyls than under white (or blue) light, making the
effect of KAR on seedling development easier to observe.
1. Surface sterilize seed using chlorine as above (Subheading 3.2,

step 2), ethanol (below), or your preferred method. Determine
how much seed you will require for all treatments, assuming
that you will sow approximately 40 seedlings per plate. We
dispense seed into 1.5 ml microfuge tubes, add 70 % (v/v)
ethanol containing 0.05 % (v/v) Triton-X100, and mix for
5 min by occasional inversion. Allow the seed to settle, remove
the solution by pipetting, and replace with approximately 1 ml
of 100 % (v/v) ethanol. Mix briefly, allow the seed to settle,
and aspirate as much of the ethanol as possible. Close the lid,
flick the tube to disperse the seed, and leave the tube open and
on its side in the laminar flow hood to allow residual ethanol
to evaporate. After about 10 min, the seed will be dry and
ready to sow.
2. In the meantime, prepare the MS-agar plates by adding KAR
at the desired concentration, as described for seed germination
assays above.
3. Sow seed, seal the plates with permeable tape, and stratify at
4 °C for 72 h in the dark.
4. Transfer the plates under white light (~100–150 μmol/m2/s)
for 3 h at 20–22 °C. This exposure to broad-spectrum light
will ensure homogeneous germination.
5. Transfer the plates into darkness for 21 h at 20–22 °C, placing
them in the same location as the red light source. This period
of darkness is not essential but it means that red light treat-
ment will begin at a consistent 24 h after removal of the seed
from cold.
6. Incubate the plates at 20–22 °C under continuous red light for
4 days.
7. Remove the plates from the incubator and photograph them
on a blue or black background to enhance contrast. If there
are too many seedlings to clearly see each one individually,
you can remove them manually and arrange them on another
agar plate prior to photography. (IMPORTANT: Also photo-
graph an object of known size (e.g., a coin, or a grid pattern).
If you do not have a camera stand that allows each photo-
graph to be uniformly composed, this object should be in
every photograph and the image dimensions calibrated for
each photograph.)
8. Import each image into ImageJ software, one at a time.
9. Calibrate the image dimensions (to convert pixels into mm)
as follows. Draw a line of known physical length across your
calibration image taken in step 7. Use the “Set Scale” command
under the “Analyze” menu to define the length in millimeter
of the calibration object.
5
0.5× MS
Hypocotyl length, mm, mean ± SE

1 µM KAR1
4 10 µM KAR1
1 µM KAR2
10 µM KAR2
1 µM rac-GR24
3
0
Ler
Fig. 2 Example of KAR-treated seedlings from the hypocotyl elongation assay. Approximately 20 wild-type
Landsberg erecta seedlings from each treatment group were arranged flat on an agar plate after 4 days of
growth under continuous red light. KAR1 and KAR2 are two different karrikins; rac-GR24 is a synthetic strigo-
lactone analogue that has similar effects on seedling growth as KAR (but note that rac-GR24 does not promote
cotyledon expansion). “Mock” corresponds to seedlings grown on 0.5× MS media supplemented with 0.1 %
(v/v) acetone as a treatment control. The grid pattern, with 13 mm spacing, serves as a size calibration for the
image. This image has been processed in Adobe Photoshop to enhance contrast for reproduction in grayscale.
The chart on the right depicts data from three such images derived using ImageJ, each representing an inde-
pendent experimental replicate
10. Measure hypocotyl lengths manually using the “segmented

lines” tool. Measure from the root-hypocotyl junction to the
apical meristem, following the line of the hypocotyl. Do this
for 15–20 seedlings, and copy the data into a spreadsheet.
The procedure described above, performed once, comprises
one experimental replicate. Because of sampling error, we perform
this procedure three times or more (on different dates) for a given
experiment, and use each occasion as a separate experimental rep-
licate for statistical analysis. Thus, individual seedlings can be
nested within a replicate, but each replicate should be considered
the independent sample. In practise, we simply take the mean
hypocotyl length for each of the three replicates, and compare the
mean of these replicates across genotypes and treatments (Fig. 2).
4 Notes
1. Suitable LEDs can be found at electronics supply stores and

online, often sold as self-adhesive strips with red, green, and
blue LEDs, complete with power supply and remote control
(e.g., www.eaglelight.com). Ideally these should be characterized
with a spectrometer to determine the wavelength and intensity

of each light source. The peak output of the red LED should
lie between 600 and 660 nm, with integrated intensity of
between 2 and 20 μmol/m2/s, but these parameters are not
critical for this assay.
2. ImageJ is a public domain, Java-based image processing pro-
gram developed at the National Institutes of Health (http://
imagej.nih.gov/ij/).
3. Seeds harvested for germination assays may be equilibrated in
an atmosphere of 15 % relative humidity if you have facilities to
do this, instead of incubation in a box containing anhydrous
silica gel, prior to cold storage.
4. While −70 to −86 °C is preferred for storage of seeds to
maintain dormancy from many months to years, −20 °C may
be adequate for several months.
References
1. Nelson DC, Riseborough J-A, Flematti GR, strigolactone signaling in Arabidopsis thali-
Stevens J, Ghisalberti EL, Dixon KW et al ana. Proc Natl Acad Sci U S A 108:
(2009) Karrikins discovered in smoke trigger 8897–8902
Arabidopsis seed germination by a mechanism 4. Waters MT, Nelson DC, Scaffidi A, Flematti
requiring gibberellic acid synthesis and light. GR, Sun YK, Dixon KW et al (2012)
Plant Physiol 149:863–873 Specialisation within the DWARF14 protein
2. Nelson DC, Flematti GR, Riseborough J-A, family confers distinct responses to karrikins
Ghisalberti EL, Dixon KW, Smith SM (2010) and strigolactones in Arabidopsis. Development
Karrikins enhance light responses during germi- 139:1285–1295
nation and seedling development in Arabidopsis 5. Chen M, MacGregor DR, Dave A, Florance H,
thaliana. Proc Natl Acad Sci U S A 107: Moore K, Paszkiewicz K et al (2014) Maternal
7095–7100 temperature history activates Flowering Locus
3. Nelson DC, Scaffidi A, Dun EA, Waters MT, T in fruits to control progeny dormancy accord-
Flematti GR, Dixon KW et al (2011) F-box ing to time of year. Proc Natl Acad Sci U S A
protein MAX2 has dual roles in karrikin and 111:18787–18792
Chapter 5
Low-Cost Microprocessor-Controlled Rotating Stage

for Medium-Throughput Time-Lapse Plant Phenotyping
Francis Barbez, Jürgen Kleine-Vehn, and Elke Barbez
Abstract
Here we provide the instructions to build a cost-friendly rotating stage, which enables time-lapse pheno-
typing of seedlings, grown vertically on in vitro plates, in a medium-throughput manner.
Key words Rotating stage, Plant phenotyping, Arduino
1 Introduction
Due to their sessile lifestyle, plants feature an enormous capacity

for dynamic growth regulation in response to developmental as
well as environmental signals. To improve our understanding of
these dynamic developmental processes, the use of time-lapse plant
phenotyping approaches is becoming increasingly important.
Nowadays, a high range of excellent time-lapse phenotyping plat-
forms have been developed and are custom-made as well as com-
mercially available [1–3]. Unfortunately, the development and
construction or the commercial purchase of these platforms are
expensive and may not be cost effective for certain projects. Here,
we present an in-house-built rotating stage which enables custom-
ized time-lapse plant phenotyping in a medium-throughput man-
ner. Using the open-source Arduino prototyping platform, we
developed and programmed a microcontroller that steers a rotat-
ing stage carrying up to eight in vitro-grown plates (Fig. 1). In
combination with an ordinary consumer-grade single-lens reflex
camera, the micro-controlled rotating stage significantly increases
the number of seedlings that can be monitored over time. Below
are comprehensive instructions, a complete electric schematic dia-
gram, and the adjustable Arduino script for building a micro-
controlled rotating platform for less than €200.
37
38 Francis Barbez et al.
Front view Top view

30 cm
in vitro plate
L profile
(long) iron strip
L profile metal plate (long)
(short)
L profile
(long)
iron strip
stepper
hand screw iron strip motor
30 cm
electro-
magnet metal
bracket
metal plate
(short)
hand screw
24 cm
wooden plank
adjustable electro-magnet
rubber foot
14 cm
L profile
(short)
Side view
wooden plank
Adjustable
rubber foot
14 cm
L profile L profile
(long) (short)
hand screw
iron strip
metal plate
metal stepper metal 13 cm (short)
bracket bracket hand
motor screw
electro- black paper
magnet
L profile
(long)
electro-
wooden plank adjustable L profile metal plate magnet
rubber foot (short) (long)
stepper metal
24 cm iron strip motor bracket
plug socket
Fig. 1 Schematic drawing of construction
2 Materials
2.1 The Rotating Table 1 depicts the materials needed to construct the rotating
Stage stage.
2.2 The Table 2 depicts the materials needed (see Note 1) to construct the
Microcontroller microcontroller.
2.3 Software The software used is free of charge.

1. Arduino to upload the provided script on the Arduino-Uno
microcontroller: download from https://www.arduino.cc/
2. Fiji for image analysis: download from http://fiji.sc/Fiji
2.4 Imaging 1. Digital, single-lens reflex (or mirrorless-type system) camera

including imaging software to control the camera from a com-
Rotating Stage for in vitro Plant Phenotyping 39
Table 1
Material needed for constructing the rotor
# Symbol Name Specifications

1× Stepper motor 6 wires (SanyoDENKI 103-547-52500)
1× Electromagnet 100 N 12 V/DC 3.8 W
# Name Specifications Comments

1× Wooden plank 14 cm × 24 cm The basis of the stage
4× Adjustable rubber feet To keep the stage in a stable position
4× L-shaped metal bracket To attach the motor and electromagnet on
the plank
4× Aluminum L profile 2.5 cm × 2.5 cm × 30 cm For the frame
(long) (2 mm thick)
8× Aluminum L profile 2.5 cm × 2.5 cm × 30 cm For the plate holders
(short) (2 mm thick)
1× Metal plate (long) 29 cm × 4 cm (2 mm thick) To attach the frame to the motor
2× Metal plate (short) 6 cm × 4 cm (2 mm thick) To attach the magnetic strips
8× Hand screws and nuts To tighten the in vitro plates on the frame
4× Magnetic strip To position the stage exactly above the
electromagnet after turning
Additional screws To construct the frame and attach the

and nuts different parts
4× Black paper 30 cm × 12 cm To provide a black background behind the

in vitro plates
puter. There are multiple options for equipping this setup with
a camera and the pace of technical advance in this field is fast;
therefore, we do not recommend any specific brand or make.
Just make sure that the camera/lens combination can bring
two in vitro culture plates side by side into focus (see Fig. 1)
and can be remote controlled by a computer.
2. Laptop to control the camera.
2.5 Tools No specialized tools are required to build the rotating stage.
Standard tools include Soldering iron, screwdrivers, metal
glue, drill, and saw.
Table 2
Electric components needed for constructing the control unit

1× Arduino microcontroller Arduino-UNO (US only)/
Genuino-UNO (outside of US)
1× Big Easy Driver Stepper motor Sparkfun, Big Easy Driver

driver board ROB-12859
2× Diode 1000 V/1 A
2× Optocoupler 4N35
2× Mosfet/hexfet IRF520
5× Led 5 mm, round, 20 mA, 2.5 V
1× Electrolytic capacitor 2.5 mm 100 μF

16 V/DC 20 % (Ø × l)
6.3 mm × 11 mm
SY 100 μF/16 V
6.3 × 11 mm
Resistor Carbon film-fixed resistors
(continued)
Table 2
(continued)
1× Relais 4 Pole relais 12 V/DC 7 A
1× Round rocker switch 250 V/AC

6A
1× Rocker switch 250 V/AC

10 A
# Name Specifications
1× Panel-mounted socket with flange 6 contacts, 4 A, 34 V
1× Plug with insulated handle 6 contacts, 4 A, 34 V
1× 6-Wire control cable 6 × 0.25 mm2
Electric wire 0.14–0.25 mm2
Perfboard, dot-PCB For example 75 mm × 100 mm,
epoxy, Cu 35 μm
1× USB cable
1× Adaptor/power supply 12 V/3 A
1× Adaptor/power supply 5 V/1 A
Soldering tin
3 Methods
3.1 The Stage Figure 1 gives a detailed representation of how the stage can be
built.
1. Drill two holes (with a diameter slightly bigger than the hand
screws) in one flank of every long L-profile. The holes are
located 8 cm from each end.
2. To construct the frame, screw the long L-profiles tightly

together at the corners so that the perforated flanks stand
upright and the horizontal flanks face the inside of the frame.
3. Screw the long metal plate in between two sides of the frame.
Drill a hole, exactly in the middle of the metal plate, to later on
attach the frame on the axle of the stepper motor.
4. Screw on the two short metal plates in the middle of the two
sides that were not connected with the long metal plate.
5. Attach the iron strips to the lower sides of the metal plates
using metal glue.
6. Glue the eight short L-profiles onto the horizontal flanks of
the frame to create a groove, which is slightly wider than the
thickness of the in vitro plates used. Attach a profile facing
every perforation.
7. Glue the eight nuts belonging to the hand screws over the
perforations on the outside of the frame. When the glue is
thoroughly dry, screw the hand screws loosely in the nuts. The
hand screws will enable the in vitro plates to be fixed tightly on
the frame (see Note 2).
8. Screw the adjustable rubber feet onto the bottom of the
wooden plank.
9. Use 2 × 2 metal brackets to attach the stepper motor and the
electromagnet to the plank as shown in Fig. 1 (see Note 3).
10. The presence of the electromagnet will keep the stage in the
right position after every 90° turn (see Note 4). The optimal
distance between the iron strip and the electromagnet
(3–6 mm) may have to be optimized by trial and error.
11. Install the socket in a leftover piece of L-profile and attach this
to the wooden plank. The six contacts of the socket will be
later connected to the four cables of the stepper-motor and the
two cables of the electromagnet according to the electric dia-
gram given in Fig. 2.
3.2 Microcontroller Figure 2 gives a detailed representation of how the microcontroller

can be built (see Note 5).
1. Build the microcontroller on a perfboard, also called dot-PCB
(or, depending on your electronics skills, a breadboard, strip-
board, or even a printed circuit board of your own design)
using a soldering iron according to the electric diagram in
Fig. 2 (see Notes 6 and 7).
2. We chose to build the microcontroller in the shell of an old
external hard drive (Fig. 2). It can, however, be built into any
handmade or commercially available heat-resistant electronic
housing.
S tepper m otor
power supply 5V via USB

+12V
IN4007
IN4007
OPTO
5V GR N D 4N35
1 6
relais
10
IRF520
220 electro-magnet
9 2 5
8
3 4
5
1 6
2k
IRF520
4 220
2 5
3
ARDUINO 3 4
1k
2
OPTO
1k
4N35 100 µF
100 k
100 k
1k
1k
power supply 12V
Big easy driver

LED
Relais
2 1
Socket
round-rocker
switch
B
GR N D
STEP GRND
DIR + Arduino UNO

Circuit board
Round rocker
switch
BIG EASY DRIVER
Fig. 2 Electric circuit for the control unit
3.3 Uploading Here we provide an Arduino script, which allows steering of the
the Arduino Script stage’s rotation every desired time interval (in our case 15 min)
over an angle of 90°. The script also contains a manual paragraph
enabling a single 90° turn of the stage on command, depending on
the configuration of the switches (explained in Subheading 3.4).
1. Download the open-source Arduino software IDE from
https://www.arduino.cc/.
2. Download the Arduino script to steer the rotating stage from
http://www.dagz.boku.ac.at/pgz/kleine-vehn/tools/
supplements-rotating-stage/.
3. Connect the Arduino-UNO to the computer using the USB
connection.
4. Open the Arduino script.
5. Make sure that the software is compatible with the Arduino-
Uno hardware. Tools > board > Arduino/Genuino UNO.
6. Adjust the time interval if necessary.
X = desired time between the start of two
90°-turns in milli-seconds – 3000
Y = desired time between the start of two
90°-turns in milli-seconds – 18000
For example, in case of the given program, we want the stage to
turn 90°, every 15 min.
X= 900 000 msec(15 min)-3000 = 897 000

Y= 900 000 msec (15 min) -18000 = 882 000
7. Click upload to upload the script on the Arduino-board (see
Note 8).
8. Disconnect the Arduino-UNO from the computer.
9. The Arduino-UNO is now able to steer the stage according to
the status of the switches (manual vs. automatic, Subheading 3.4).
3.4 How to Use The rotating stage can be operated in the absence of a computer by
the Stage using the rocker switch (on/off) and the round rocker switch to
change between the manual (a single turn of 90°) and the auto-
matic mode (the stage turns 90° every time interval as required).
1. Note that the main power (rocker switch) should be switched
off while connecting or disconnecting any cables.
2. Place the eight in vitro plates on the stage, the lid facing the
inside of the frame, and tighten them using the hand screws.
To improve the image quality, placing black paper behind the
plates is recommended (Fig. 1).
3. To use the stage: Switch on the main power (rocker switch).
The main power control LED will glow.
4. In the manual mode (I), the stage will turn 90° every time
the switch is activated. This allows for correct positioning of
the turning stage carrying the in vitro plates in front of the
camera.
5. In the automatic mode (II), the stage will turn 90° at the end
of every time interval as programmed (in our case, every
15 min) until the switch is turned off (0).
6. Set the camera to take a picture every time interval (in our
case, 15 min) using the remote control software of your choice.
Start the time lapse around 30 s after the start of the rotating
stage to avoid pictures being taken during a stage turn.
7. In our case, every 15 min, one side of the stage will be imaged
so that every plate will be imaged once an hour.
4 Notes
1. In Table 2, we provide the electric components which we used to

build the microcontroller. However, most of the electric compo-
nents can be purchased from any other company as long as they
meet the technical specifications provided in the online data
sheets which can be downloaded from http://www.dagz.boku.
ac.at/pgz/kleine-vehn/tools/supplements-rotating-stage/.
2. The in vitro plate holders can be replaced by a horizontal stage

enabling the imaging of plants in pots.
3. To ensure that the components are tightly attached to the
wooden plank, covering the plank with a perforated metal
plate before attaching the motor and the electromagnet is rec-
ommended. The perforation allows for the easy attachment of
the metal brackets.
4. The frame can be built out of any light and strong material,
but note that the material used must not be magnetizable by
the electromagnet.
5. Data sheets for the components used can be downloaded from
http://www.dagz.boku.ac.at/pgz/kleine-vehn/tools/
supplements-rotating-stage/.
6. The stepper motor used is wired with six cables. However, only
four are operational. The unused cables can be identified by
measuring the resistance between the different cable pairs. The
resistance between two used cables will be double then between
a used/unused cable pair. (In the case of the stepper motor
used, the unused cables are black and white.)
7. The active state of several modes (main switch, manual mode,
automatic mode, and relays) is indicated by glowing LED
lamps. To ease the interpretation of the different glowing
LEDs, we decide to use LED lamps in different colors.
8. If it is not possible to upload the Arduino script on the Arduino-
UNO board, the default export-PORT in the Arduino software
may be incorrect. Select another PORT under TOOLS > PORT
and try again.
Acknowledgements
We would like to thank the Arduino community for setting up this

wonderful and easily accessible prototyping platform. We are grate-
ful to David Whittaker for help with the manuscript. This work was
supported by the Vienna Science and Technology Fund (WWTF)
(Vienna Research Group), Austrian Science Fund (FWF)
(P26568-B16; P26591-B16), and the European Research Council
(ERC) (Starting Grant 639478-AuxinER).
References
1. Araus JL, Cairns JE (2014) Field high- phenotyping is ready for a close-up. Curr Opin
throughput phenotyping: the new crop breed- Plant Biol 24:93–99
ing frontier. Trends Plant Sci 19:52–61 3. Fiorani F, Schurr U (2013) Future scenarios
2. Fahlgren N, Gehan MA, Baxter I (2015) for plant phenotyping. Annu Rev Plant Biol
Lights, camera, action: high-throughput plant 64:267–291
Chapter 6
Genome-Wide Association Mapping of Root Traits

in the Context of Plant Hormone Research
Daniela Ristova and Wolfgang Busch
Abstract
Genome-wide association (GWA) mapping is a powerful method for the identification of alleles that
underlie quantitative traits. It enables one to understand how genetic variation translates into phenotypic
variation. In particular, plant hormone signaling pathways play a key role in shaping phenotypes. This
chapter presents a protocol for genome-wide association mapping of root traits of Arabidopsis thaliana in
the context of hormone research. We describe a specific protocol for acquiring primary and lateral root
trait data that is appropriate for GWA studies using FIJI (ImageJ), and subsequent GWA mapping using a
user-friendly Internet application.
Key words Natural variation, Phenotyping, GWAS, Hormones, Root development
1 Introduction
Understanding how genotypic variation translates into phenotypic

variation remains one of the fundamental challenges in biology. In
recent years, tremendous progress has been made in this regard
mainly through the application of genome-wide association studies
(GWAS) [3]. GWAS is a method in which variation of a phenotype
of interest is associated with sequence polymorphisms throughout
the genome in a large number of genetically distinct individuals. In
Arabidopsis thaliana, natural variation in conjunction with GWAS
can be used to map alleles responsible for different quantitative
traits (Fig. 1). This is enabled by the large collection of Arabidopsis
accessions (naturally occurring strains), which were collected from
a broad range of habitats around the world and genotyped at high
resolution [1]. Different accessions of Arabidopsis thaliana exhibit
high levels of phenotypic variation [2, 6], which can be used to
relate diverse phenotypes to the genetic variation. Thus, GWAS
have become a very powerful tool for the identification of genes
and their alleles that underlie different quantitative traits [3].
Moreover, recent progress in addressing the potential confounding
47
48 Daniela Ristova and Wolfgang Busch
Fig. 1 Schematic representation of a GWAS in Arabidopsis roots. Variation of root phenotypes in diverse acces-
sions is quantified. This phenotypic variation is associated with genetic variation at a single-nucleotide level.
In this example, accessions with a short root show a highly significant difference to accessions from long roots
at one position of the genome which results in a significant association above the threshold in the third chro-
mosome (Manhattan plot)
effect of population structure [3] has made it even more powerful

and practical [4, 5].
In this chapter, we present a protocol for performing GWAS
on root traits acquired from Arabidopsis thaliana accessions grown
on media supplemented with phytohormones. In principle the
same protocol is generally applicable for root phenotypes, as GWAS
can be performed on any trait of interest that can be described as a
numeric value. We describe how to culture Arabidopsis accessions
in sterile conditions, and acquire root phenotypes. Further we
explain how to quantify root traits using the image analysis soft-
ware FIJI or ImageJ. Finally, we describe how to perform GWAS
by submitting the quantified phenotype to a free online Web appli-
cation, the GWA-Portal [7].
GWAS of Root Traits in Plant Hormone Research 49
2 Materials
2.1 Materials 1. Arabidopsis thaliana seeds from multiple accessions, approxi-

Needed for Culturing mately 15 seeds per accession. For one experimental set ten
Plants seeds are used for plating. However, start with 15 seeds to have
few extra during the sterilization and handling.
2. Other materials needed: Tube rack for 1.5 ml microcentrifuge
tubes and 1.5 ml microcentrifuge tubes; 250 ml glass beaker;
household Bleach containing 5–8 % sodium hypochlorite; 37 %
hydrochloric acid; 15 ml polypropylene conical centrifuge tube;
polycarbonate lockable airtight box; sterile square Petri dishes,
12 × 12 × 1.7 cm; transfer pipette, tips (1000 μl), parafilm; sterile
pipettes (50 ml) and electronic pipette; non-woven ventilating
tape; custom-made support rack to hold plates vertically (Fig. 2a);
leveling table, for filling media into plates (Fig. 2c); 0.22 μm filter,
Fig. 2 Culturing plants and acquisition of root trait data. (a) Custom-made support racks for vertically posi-
tioned plates. (b) Lockable airtight box for gas sterilization of seeds. (c) Culture media plate production using
a leveling table to ensure identical growth conditions. (d) Seed distribution on culture plates using plating
model. (e) Seedling transfer to MS media supplemented with hormone
for sucrose sterilization; water bath, for media cooling; Murashige

and Skoog including MES plant growth medium, sucrose, plant
agar, pH meter, KOH solution (1 M).
3. Hormones: 3-Indoleacetic acid (IAA), kinetin (CK), and
(+)-abscisic acid (ABA) are all solved in dimethyl sulfoxide
(DMSO).
2.2 Media Preparing media (1 l): Prepare sterile sucrose (1 M) solution, and
Preparation filter-sterilize it in the laminar hood, using a 0.22 μm filter. Weigh
an appropriate amount of Murashige and Skoog including MES,
into 966.67 ml of sterile water into a beaker containing a magnetic
stirrer bar, and mix well. Adjust pH to 5.7 with 1 M KOH solu-
tion. Add the media into a glass bottle, and add 8 g (0.8 %) of plant
agar. Sterilize the media at 125 °C for 15 min. After sterilization,
cool the bottle/media to 60 °C in water bath, and move the bottle
to the laminar hood. Add 33.33 ml of 1 M sucrose solution mix
well and pour 57 ml of media in each plate. Leave the plates open
for 45 min, then close them, wrap the plates in sterile plastic bags,
and store upside-down at room temperature until plating. Media
should be prepared 1–3 days before the plating. For preparation of
hormone media, after addition of sucrose, add hormones diluted
in DMSO to the desired final concentration.
2.3 Image Data At least one conventional flatbed scanner capable of 1200 dpi image
Acquisition data resolution is required. We recommend using multiple scanners
for increasing throughput, and using a custom-made support frame
for the plates in order to keep the plate position constant. To oper-
ate multiple scanners, you will need a desktop UNIX computer and
the multiscan interface (download link and instructions: http://
www.gmi.oeaw.ac.at/ research-groups/wolfgang- busch/
resources/brat) [8].
2.4 Quantification Quantification and extraction of root traits are conducted using
of Root Traits the Fiji software (a distribution of ImageJ; http://fiji.sc/
Downloads). The exported trait values can be further processed in
a spreadsheet application (e.g., Microsoft Excel) or any statistical
software (e.g., R) to further calculate mean and median trait value
for each accession/treatment.
2.5 Genome-Wide The GWA-Portal is an online Internet application (http://gwas.

Association Mapping gmi.oeaw.ac.at/) that requires a browser supporting HTML5.
The input file for GWA-Portal is a text with comma-separated
columns of accession IDs (unique number assigned for each
Arabidopsis accession, for example for Col-0 is 6909) and trait val-
ues. Accessions IDs can be retrieved on the GWA-Portal Web
(http://gwas.gmi.oeaw.ac.at/#/taxonomy/1/passports?alleleAss
ayId=0).
3 Methods
3.1 Culturing Plants 1. Surface sterilization of seeds: For each Arabidopsis accession
place 15 dry seeds in an open 1.5 ml microcentrifuge tube in a
tube rack, and place the rack into an airtight lockable box
along with a beaker containing a magnetic stirrer bar and
100 ml of bleach. Place the lid partially on the box. Put 3.5 ml
of 37 % hydrochloric acid in a 15 ml conical tube, lift the lid on
the side, quickly decant the HCl into the beaker to generate
chlorine gas (see Note 1), and then quickly close and secure
the lid. Place the box on a magnetic stirrer (speed = 2.5 rpm)
for 1 h of sterilization time (Fig. 2b).
2. After 1 h, unclamp the box, remove the beaker from the box,
and slightly tilt the lid in order to vent the rest of the chlorine
gas for approximately 15–20 min.
3. Close the lid. Transfer the box with the sterilized seeds inside
to a laminar hood and leave the tube rack with open tubes to
vent for 30 min.
4. To each tube add 2–3 drops of sterile water, close the tubes,
vortex, and centrifuge briefly. Place the rack in the dark at 4 °C
for stratification for 3 days.
5. Plating seeds: Plating is performed under sterile conditions
(laminar hood) to avoid contamination. Use a regular pipette or
a sterile custom-made pipette for speeding the plating process
(Fig. 2d). To make the custom-made pipette cut one transfer
pipette just below the base, and secure a sterile tip (1000 μl) to
the transfer pipette with parafilm. Plate four accessions per plate
(use background design for easy plating (Fig. 2d)). First plate
ten seeds/accession on MS media without hormone (control
media), and on day 7 transfer the five most common-appearing
plants from each accession to MS media supplemented with
hormone or to the control media (Fig. 2e). Final hormone con-
centrations depend on the research project and are best estab-
lished during pilot experiments prior to the GWAS. In our
studies we used IAA (0.5 μM), CK, and ABA (1 μM).
6. After plating, carefully seal the edges of the plates with non-woven
ventilating tape impermeable to bacteria (Fig. 2a).
7. Place plates vertically into the custom-made rack, and place the
rack into growth chamber (Fig. 2a).
3.2 Image Data 1. Transfer the rack with petri plates to the image acquisition
Acquisition room (see Note 2).
2. Place each plate (without removing the tape) onto the hori-
zontally oriented scanners. They should fit adequately into the
support frames (Fig. 3a).
Fig. 3 Image acquisition and root trait quantification. (a) CCD scanner with a custom frame to allow for identical
plate positions during scanning. (b) Example of root length extraction in FIJI (ImageJ). The example plate has twenty
10-day-old Arabidopsis seedlings (from four accessions) in two rows, grown on 500 nM IAA for 3 days
3. Scan the plates with the following parameters: 1200 dpi 8-bit
RGB TIFF files, and name the plates. For optimal image
quality, scan in a darkroom with scanner lid open. After image
acquisition, return the rack with plates to the growth chamber
(see Note 3).
3.3 Trait 1. Open Fiji, and then open the image of one plate. Choose
Quantification “Segmented Line” to draw the length of the primary root
(Fig. 3b), and run the “Measure” function by selecting from
the “Analyze” pull-down menu bar on top of the window or
alternatively press cmd + M (Mac) (length shown in pixels).
Repeat for each length, then copy the values, and export them
into Excel-file format.
2. For performing GWAS use mean and/or medium of all indi-
viduals of one accession for each root trait.
3. Quantified traits (example): length of primary root on transfer
day (P1), length of primary root after transfer to hormone
(P2), total length of lateral roots (TLRL), and number of lat-
eral root (LR.No). After obtaining these values, calculate the
following traits: total length of primary root (P = P1 + P2),
average lateral root length (LRL = TLRL/LR.No), total root
length (TRL = TLRL + P), lateral root density (LRd = LR.
No/P), and length ratio (LRR = TLRL/P).
3.4 Genome-Wide To identify alleles associated with the phenotype (root trait) of
Association Mapping interest, perform genome-wide association mapping by submitting
the quantified phenotype to GWA-Portal. GWA-Portal allows one
to quickly run GWAS on the four available datasets (250k SNP,

1001 Fullsequence, Swedish genomes, or Imputed Fullsequence)
(see Note 4).
1. Open the GWA-Portal web page (http://gwas.gmi.oeaw.
ac.at) and create an account. Press “Take a tour” to familiarize
yourself with the web page (see Note 5) [7].
2. To perform GWAS on your traits, click on “Create” in the
“New GWAS analysis.” A new page will open where the user
needs to fill in information for six steps (left column). Step 1 is
“Study”; here the user can choose already uploaded phenotypes
or can create a new one by clicking “Create new study.” When
“Create new study” is chosen, a new, smaller window will open
to add information for the new experiment, including name,
originator, design, and comments. After each step Click “Next”
(in the lower right side) to move to the next one.
3. Step 2, “Phenotype.” Click “Upload phenotype” and upload
your trait of interest. Use the .csv file format, where the first
column is “Accession_ID” and the second column is the
trait values (see Note 6). After uploading your file click
“SAVE.” Select your newly created phenotype and then click
“Next.”
4. Step 3, “Genotype.” Here you can select which sequencing
data to use. Use the 250k SNP chip data set (first option) [5].
5. To transform the phenotype data distribution, go to step 4,
“Transformation.” Choose the desired type of transformation
or no transformation.
6. Step 5 is where the user chooses the type of analysis. We rec-
ommend using the “AMM” method (accelerated mixed
model) that efficiently addresses the confounding effects of
population structure. The other two methods (KW and LM)
might produce population structure confounded association
mappings [4–6]. Finally, go to step 6, “Summary,” where all of
the previous user choices are shown. Press “Finish” to proceed
to the next page where the analysis is performed.
7. A new page will open where the summary is shown on the left
side. In the middle under “GWAS Status” click on the
drop-down arrow “N/A,” and choose “Run analysis,” after
which the user can monitor the progress of the analysis indi-
cated in percentage (see Note 5). When finished, click on “Plot
(AMM)” to visualize the interactive Manhattan plots for each
chromosome (Fig. 1). A Manhattan plot is a type of scatter
plot where genomic coordinates of the five chromosomes are
displayed on the X-axis, while the negative logarithm of the
association P-value for each SNP is displayed on the Y-axis.
On the interactive Manhattan plot the user can zoom into the
region of interest by selecting the region above the association.
To save the plot click “Download plot”; a new window will

open where one can choose the “mac” (minor allele count)
threshold, chromosomes, and the format.
4 Notes
1. Chlorine is a strong oxidizing agent with a yellow-green color,

and it is a very toxic gas that irritates the respiratory system.
Therefore wear gloves and safety goggles, and work in a well-
ventilated chemical hood.
2. To prevent condensation of droplets on the plate lids, the
temperature in the image acquisition room should be the same
as in the growth room or slightly higher.
3. Images acquired on scanners appear in a mirrored orientation.
This must be accounted for that when quantifying the root
traits and assigning accession number (i.e., flip the image to
obtain the positions as visible from top view during plating).
4. “250k SNP” refers to sequencing data genotyped using a cus-
tom Affymetrix single-nucleotide polymorphism (SNP) chip
containing 250,000 SNPs [5], “1001 Fullsequence” dataset
contains full sequence data (~10 million SNPs) for 1135 acces-
sions, while the “Imputed Fullsequence” dataset has the same
amount of SNPs for 2029 accessions (unpublished). “Swedish
genomes” are 259 Swedish accessions (6 million SNPs) [10].
5. At times the GWA-Portal server is quite busy; thus the page
can take some time to load.
6. The phenotype file has to be in the .csv file format (which can
be created in excel) and contain at least two columns: the
first column should contain the GWAS IDs of accessions (a
number which has already been assigned by the GWA-
Portal, see web page), while the second column contains the
mean and/or median values for the trait of interest for the
respective accession. Additional traits can be added in addi-
tional columns. The phenotype file must contain headers
with unique trait names. Do not use space characters in trait
names. Missing data must be denoted as “NA.”.
Acknowledgement
We are grateful to Elke Barbez, Takehiko Ogura, and Santosh

Satbhai for comments and suggestion for this protocol, Ümit Seren
for information about the GWA-Portal, and Matthew Watson for
editing.
References
1. Horton MW, Hancock AM, Huang YS et al for genome-wide association studies in structured
(2012) Genome-wide patterns of genetic variation populations. Nat Genet 44:825–830
in worldwide Arabidopsis thaliana accessions from 6. Atwell S, Huang YS, Vilhjálmsson BJ et al
the RegMap panel. Nat Genet 44:212–216 (2010) Genome-wide association study of
2. Koornneef M, Alonso-Blanco C, Vreugdenhil 107 phenotypes in Arabidopsis thaliana
D (2004) Naturally occurring genetic variation inbred lines. Nature 465:627–631
in Arabidopsis thaliana. Annu Rev Plant Biol 7. Seren Ü et al (2012) GWAPP: a web applica-
55:141–172 tion for genome-wide association mapping in
3. Weigel D (2012) Natural variation in Arabidopsis. Plant Cell 24(12):4793–4805
Arabidopsis: from molecular genetics to eco- 8. Slovak R et al (2014) A scalable open-source
logical genomics. Plant Physiol 158:2–22 pipeline for large-scale root phenotyping of
4. Platt A, Horton M, Huang YS et al (2010) The Arabidopsis. Plant Cell 26(6):2390–2403
scale of population structure in Arabidopsis 9. Long Q et al (2013) Massive genomic variation
thaliana. PLoS Genet. doi:10.1371/journal. and strong selection in Arabidopsis thaliana
pgen.1000843 lines from Sweden. Nat Genet 45(8):|
5. Segura V, Vilhjálmsson BJ, Platt A et al (2012) 884–890
An efficient multi-locus mixed-model approach
Chapter 7
High-Throughput Scoring of Seed Germination

Wilco Ligterink and Henk W.M. Hilhorst
Abstract
High-throughput analysis of seed germination for phenotyping large genetic populations or mutant col-
lections is very labor intensive and would highly benefit from an automated setup. Although very often
used, the total germination percentage after a nominated period of time is not very informative as it lacks
information about start, rate, and uniformity of germination, which are highly indicative of such traits as
dormancy, stress tolerance, and seed longevity. The calculation of cumulative germination curves requires
information about germination percentage at various time points. We developed the GERMINATOR
package: a simple, highly cost-efficient, and flexible procedure for high-throughput automatic scoring and
evaluation of germination that can be implemented without the use of complex robotics. The
GERMINATOR package contains three modules: (I) design of experimental setup with various options to
replicate and randomize samples; (II) automatic scoring of germination based on the color contrast
between the protruding radicle and seed coat on a single image; and (III) curve fitting of cumulative ger-
mination data and the extraction, recap, and visualization of the various germination parameters.
GERMINATOR is a freely available package that allows the monitoring and analysis of several thousands
of germination tests, several times a day by a single person.
Key words Arabidopsis thaliana, Automatic scoring, Curve-fitting, Germination, High-throughput

analysis, Image analysis
1 Introduction
Fundamental and applied seed biology research relies heavily on

accurate quantification of seed germination. Nowadays, large-scale
experiments using large genetic populations or mutant collections
are popular tools to unravel molecular aspects of seed develop-
ment, germination, dormancy, and seed performance [1, 2].
Detailed analysis of these traits requires the complete cumulative
germination curve and extraction of parameters such as start, rate,
and uniformity of germination. Arabidopsis thaliana is one of the
most used model species for plant science and it has also become a
very useful model plant to study seed biology. However germina-
tion of the very small A. thaliana seeds (200–400 μm) is often
evaluated by using a binocular microscope, which makes this a very
57
58 Wilco Ligterink and Henk W.M. Hilhorst
Fig. 1 Photographs of different stages of Arabidopsis seed germination. From left to right: fully imbibed seed
(6 h after start of imbibition), seed coat rupture, and radicle protrusion
laborious task. This often hampers the collection of cumulative

germination data in large-scale experiments and therefore often
only the less informative percentage of germination after a nomi-
nated period of time is used to describe the germination perfor-
mance of a seed lot. With this in mind, the GERMINATOR was
developed, a cost-efficient and flexible procedure for high-
throughput automatic scoring of germination that can be imple-
mented without the use of complex robotics [3].
The uptake of water during seed imbibition is triphasic, and
consists of a rapid initial uptake (phase I), followed by a plateau
phase (phase II) and a further increase (phase III). During this last
phase, the embryo axis elongates and breaks through the seed coat,
which consists of dead tissue. The next step is the morphological
completion of germination, which is characterized by the protrusion
of the radicle through the endosperm (Fig. 1).
1.1 GERMINATOR Most approaches to the automatic evaluation of germination based

on image analysis extract information from a time series of images.
For example, change in seed size over time can be used to evaluate
the progress of germination very precisely. However, this analysis
requires fully aligned images, which are often obtained by using
fixed setups in specialized cabinets, flat-bed scanners, or camera
systems over Jacobsen tables [4–7]. This type of setup has impor-
tant consequences not only for the amount of samples that can be
measured and the flexibility to follow germination under various
environmental conditions, but also for the costs of the equipment.
Especially this flexibility and level of throughput are essential char-
acteristics of the GERMINATOR.
Evaluation of germination by using image analysis requires
good contrast between background, seed, and protruding radicle.
Therefore, a blue filter paper specifically designed for germination
High Throughput Germination Scoring 59
tests is used. By using two different color threshold levels on a single

image we can distinguish between background and seed coat, or
background, seed coat, and radicle. This allows germinated and
non-germinated seeds to be distinguished. To ensure the most
accurate scoring, the number of non-germinated seeds to monitor
germination is used. With this approach it is not necessary to
obtain fully aligned images between the various time points at
which germination is evaluated. The GERMINATOR package
consists of three separate modules, based on the components of
our procedure: (I) experiment design, (II) image analysis, and (III)
curve fitting. Module I uses Microsoft Excel combined with Visual
Basic scripts to create a database of experiments, and assists in
experiment design. Germination is documented by photographs
taken manually at various intervals during germination. Automated
scripts were developed for Adobe Photoshop and ImageJ to auto-
matically analyze the images and to minimize user input as much as
possible. With the help of ImageJ, a two-color threshold analysis is
performed on each individual image: first, a color threshold that
selects the seed coat only, and second, a color threshold that selects
both the seed coat and the protruding radicle. The specific settings
for the color thresholds can be adjusted for each seed type during
a system calibration. The output of the image analysis consists of
four tables with information about the total number of seeds and
the position and size of each individual seed that has been detected.
In module II, these tables are then coupled with the experiment
setup tables from module I, and cumulative germination curves are
calculated. The calculation uses position and seed size in the vari-
ous color threshold analyses to determine whether a seed has ger-
minated. Module III performs a curve-fit analysis of the cumulative
germination data. This module uses data generated by the image
analysis of module II, or cumulative germination data of other
sources (for example derived from manual counting). This module
executes a curve fit using the four-parameter Hill function, which
can accurately fit the typical sigmoid curve that describes the course
of germination [8]. As soon as the best-fitted curve has been deter-
mined, typical germination parameters are extracted from the
curve. Generally Gmax (maximum percentage of germination),
t50 (time to reach 50 % germination), tx (time to reach x % germi-
nation), uniformity (time interval between, e.g., 16 and 84 % of
viable seeds to germinate), and AUC (integration of area under the
germination curve until a certain point of time) are determined. As
described by El-Kassaby and co-workers [8] the AUC can also be
used to calculate a dormancy index (DI), by extracting the area
under the curve after dormancy release by cold stratification, by
the area under the curve without cold stratification. By the same
analogy the AUC can be used to measure the effect of any stress
treatment and calculate a stress index (SI). The GERMINATOR
script will summarize the results by calculating averages and
standard errors for repeated samples, performing a student-t test,

and provide a clearly formatted output including graphs for the
different germination parameters.
With the GERMINATOR package, a single person can easily
handle more than 1000 germination tests at a time and monitor
germination precisely over time. This greatly facilitates large-scale
screens of mutant collections or genetic populations, such as recom-
binant inbred line populations [9, 10]. The package was tested and
optimized for Arabidopsis seeds and it has been shown to also work
for other species including Brassica spp. and rice. The whole
GERMINATOR package, including a manual and video tutorials,
is available from http://www.wageningenseedlab.nl/germinator.
2 Materials
2.1 Germination 1. Germination trays (see Note 1).

Assays 2. Colored germination paper that fits in the germination trays
(see Note 2).
3. A plastic flexible mold/template with the positions for the dif-
ferent germination tests in the same size as the germination
paper (Fig. 2).
4. A germination cabinet (see Note 3).
2.2 For Image 1. A digital single-lens reflex camera with more than ten million
Acquisition pixels and a macro lens.
Fig. 2 Plastic mold/template to put in the germination trays to ensure that the different germination tests will
be on the correct positions
Fig. 3 Camera setup for the GERMINATOR
2. A power adaptor and USB connector cable for the camera.

3. A repro-stand.
4. Proper, preferably indirect, lightning. A Perspex screen is used
around the camera setup to soften the light (Fig. 3) in a room
without daylight in order to keep the light conditions as con-
stant as possible. Our Perspex screen is 60 cm high, 55 cm wide,
and 35 cm deep. As light source two fluorescent tubes (TL) of
30 cm long on both sides of the camera and one of 115 cm
shining towards the ceiling of 115 cm are used (see Note 4).
5. A Windows-based PC (XP or Windows 7).
6. Software to control the camera. Nikon Control Pro version 2.0
is used in our setup.
2.3 For Image 1. A Windows-based PC.

Analysis 2. Microsoft Excel (version 2003 or 2010 and English language
version).
3. Adobe Photoshop CS3 or newer (English language version).
4. The GERMINATOR package (can be downloaded from www.
wageningenseedlab.nl/germinator).
5. ImageJ version 1.40g with Montpellier RIO Imaging plug-ins
(included in the GERMINATOR package).
3 Methods
3.1 Installing Scripts 1. Unzip the whole “Germinator.zip” file to a specific directory
and Macros (keeping the folder structure of the file).
2. Adjust settings in Microsoft Excel to allow macros and install
the Solver add-in.
For Microsoft Excel 2003:

1. Tools > add in > SOLVER ADD IN.
2. Tools > Macro > visual basic editor > tools > references > SOLVER.
3. Tools > options > security > macro security > trusted publish-
ers > trust access to visual basic project.
For Microsoft Excel 2010:
1. Go to File > Options > Add-Ins. Choose Excel Add-ins and
click Go. Select Solver Add-in and click OK.
2. Go to File > Options > Trust Center > Trust Center Settings >
Macro Settings > Check “Enable all macros” and “Trust access
to the VBA project object model” and click OK.
3. Adjust settings for Adobe Photoshop.
(a) Select the window “Actions” with |Alt F9|. Open a menu
on the right top of this window and select “load actions.”
Browse to the GERMINATOR package, select folder
“Photoshop,” and select “germinator.atn” (see Note 5).
(b) In Photoshop, open one image of a typical seed germina-
tion test. In the “Actions” panel go to Germinator > crop
and save and double-click the first “crop.” Adjust the crop
position that appears on the image to encompass the first
germination test (see Note 6), and press enter and subse-
quently Ctrl + z to go back to the original image. Then do
the same steps for the following five “crops” that are shown
in the “Actions” panel and close Photoshop.
4. Install and adjust settings for Image J.
(a) Install Image J 1.40g from the GERMINATOR package
(see Note 7).
(b) Extract Zip file “mri-plugins-base.zip” to the ImageJ
1.40g directory.
(c) Extract Zip file “mri-all-plugins.zip” to the ImageJ 1.40g
directory.
(d) Copy the Germinator folder in the ImageJ folder of the
Germinator.zip file to the ImageJ\_applications directory.
(e) Open ImageJ program, go to Plugins > Montpellier RIO
Imaging > MRI VisualScripting. Go to Applications >
Applications > Germinator > germination score + white,
and click on the O. Make sure that the macro file points
to the correct location of the “particle analysis macro.
txt” file. Close the window, right click on the top white
frame (just above the blue part), and click save. Close
the script and do the same thing for germination
score − white.
3.2 Prepare 1. Open Germinator_menu 1.01.xls and set defaults: Provide a

for a Germination name for the overview and click Browse to select a default
Experiment directory in to store the germination results. Click Save twice.
2. Click Browse and select the germinator table file from the
GERMINATOR package (Germinator_table 1.05.xls) by
double-clicking and subsequently click save (see Note 8).
3. Click on the “add new experiment” button, fill in the different
parameters for the germination experiment that should be
performed, and finish by clicking Add (see Note 9).
4. Click button “Add tables,” select “New tables,” and click the
button “Make Tables” in the newly opened Excel file.
5. Fill in the parameters for the germination experiment like
details, the number of samples that should be analyzed, and
the number repetitions and treatments. The number of treat-
ment option is there for when a germination test should be
performed at, e.g., 22 and 25 °C or in water and in a certain
concentration of ABA.
6. By checking the “Optional sample file name” box, sample
names can be extracted from an Excel file (according to the
format as shown in the “sample file name example.xls” file in
the Excel directory of the GERMINATOR package). If bio-
logical replicates should have different names, these can be
placed in different columns in this file. Select “separate bio-
logical replicates” to have the biological replicates in separate
trays in the experiment setup.
7. After clicking start, the tables that show the experiment setup in
which for each sample the tray number and position in the tray
are indicated can be printed (see Note 10). The Excel file will be
saved in the default directory under the experiment’s name.
8. The printed table is brought to the lab to start the germination
experiment.
3.3 Start 1. Take a tray and insert two sheets of blue blotter paper (see
a Germination Note 11).
Experiment 2. Mark the sheets with a plate number (depending on the
amount of germination tests that will be performed) in the
upper-left corner with a pencil.
3. Add 50 ml of demi water (see Notes 12 and 13).
4. Place the plastic mold/template in the tray to ensure that the
different germination tests will be on the correct positions
(Fig. 2) (see Notes 14 and 15).
5. Add around 50 A. thaliana seeds in each well (Fig. 4, see
Note 16) according to the printed Germinator table and note
the time of the start of imbibition in the table.
Fig. 4 A picture of a typical germination experiment with A. thaliana seeds
6. Pile up to 17 filled trays, add 1 tray with filter paper and water
at the bottom of the pile and 2 at the top + a lid, and pack the
whole pile in a big plastic bag (see Note 17).
7. Place the pile in a germination cabinet and set the desired tem-
perature (see Note 18).
3.4 Optimize Image 1. Attach the camera in a repro-stand and adjust height to ~53 cm
Acquisition (see Note 19).
2. Switch camera to full manual control and connect with USB to
a PC.
3. Place the Perspex screen around the camera setup to soften the
light, place the lights, and try to avoid interference of other,
unstable, light sources (Fig. 3).
4. Place a mask that fits the germination trays underneath the
camera to fix the imaging position (Fig. 3) (see Note 20).
5. Put a germination tray as described in Subheading 3.1 under
the camera and focus.
6. Use Nikon camera Control Pro software to set resolution to
maximum and file format to jpg with normal compression.
7. Adjust camera settings (aperture, ISO, and shutter speed) with
Nikon camera Control Pro software and if necessary light condi-
tions to optimize the pictures with respect to contrast between
germination paper, seed coat, and radicle (see Note 21).
8. Save the optimal settings via Settings > Save Control Settings
(as *.ncc file) (see Note 22). Also the tone compensation
curves can be adjusted manually to fine-tune the pictures

via Camera > Edit Camera Curves. Optimal tone compensa-
tion curves can be saved by clicking Save… under the curve
(as *.ntc file).
3.5 Take Pictures 1. Open the Camera Control Pro program, select settings > set-
of the Germination tings > load control settings, and select the previously saved
Experiment optimal settings (*.ncc file). Go to image processing > tone
comp > edit > load and select previously saved optimal settings
for tone compensation curves (*.ntc file).
2. Go to tools > download options and select a folder to store the
pictures. Subsequently go to edit file name. In prefix, first
month, then the day followed by the year and time that the
pictures are taken are indicated in a mmddyyyy-hhmm format
followed by # (e.g., 10212015-0923#). Set start numbering at
1 and length of number at 3 digits. Finish by clicking OK.
3. Start taking pictures of all germination trays in the experiment
starting with tray number 1 and adjust time settings every
5–10 min as indicated under Subheading 3.5, step 2. An
experienced user can easily take pictures of 30 plates in 10 min
(see Note 23).
4. Repeat taking pictures in certain intervals. For Arabidopsis rou-
tinely pictures are taken two or three times a day (see Note 24).
3.6 Determining 1. Open a photo of germinating seeds (some already germinated,

the Optimal ImageJ others not) with ImageJ (see Fig. 4 for an example).
Settings to Score 2. Go to Process > Enhance Contrast and adjust contrast of the
Germination picture to 0.2 % saturated pixels and click OK (see Note 25).
3. Go to Plugins > Segmentation > Threshold Colour, select YUV
(at the bottom of the window) (see Note 26) and adjust color
settings of the three channels until settings are found that selects
all seeds and all radicles but no background (Fig. 5, Filter +).
Fig. 5 A photograph of a typical germination experiment and the two color thresholds used to count seed
germination. Filter +: an example of filtering everything but the background. Filter −: an example of filtering
only for the seed coat color. Threshold: example of the Filter—picture after binarization and inversion and
ready for particle analysis
Write these settings down. For the example image, these settings
are Y 50-255, U 0-130, and V 0-255.
4. Adjust color settings until the settings are found that select the
seed coat, but not the radicles and the background (Fig. 5, Filter
−). For the example image, these settings are Y 50-255, U 0-90,
and V 0-255. Write these settings down (see Note 27).
5. Select Threshold and close the Colour Threshold window.
6. Go to Process > Binary > Make Binary followed by Edit > Invert.
This results in a black-and-white image with black seeds on a
white background (Fig. 5, Threshold).
7. To count the black spots on such a photo go to Analyse > Set
Measurements and select Area, Circularity, Centroid, Perimeter
and click OK. Go to Analyse > Analyse Particles, select size
0–Infinity, Circularity 0.00–1.00, Show Outlines, Display
Results, Summarize, Exclude on Edges, and Include Holes,
and click OK.
8. In the Summary table that will appear the total number of
seeds that are counted can be found (Count) and in the Results
table parameters for the individual seeds are shown. The Area
is a measure for size of the individual seeds. This area will be
used in the Excel Germinator script.
9. To prevent counting of clustered seeds or other artifacts, such
as remnants of siliques, a limit needs to be set for the size of
particles that are still measured and considered as single seeds.
The lower size limit is typically set at approximately 0.5 × the
smallest seed that is counted and the upper limit at approxi-
mately 2 × the biggest seed that is counted (see Note 28).
10. Open the “particle analysis macro.txt” file (from
Subheading 3.1, step 4e). Adjust the line “size = 40–175” by
replacing the bold numbers with the limits obtained under
Subheading 3.6, step 9, and save the file.
11. Open the visual scripting tool of ImageJ by going to
Plugins > Montpellier RIO Imaging > MRI VisualScripting.
Go to Applications > Applications > Germinator > germination
score + white and click on the O. Fill in the values obtained
under Subheading 3.6, step 3, for the three different channels
in the fields min/max channel 1–3. Close the window, right
click on the top white frame (just above the blue part), and
click save.
12. Close the script and do the same thing for germination
score − white with the settings obtained under Subheading 3.6,
step 4.
3.7 Analyze 1. When germination is completed pictures are analyzed for auto-
a Germination matic scoring of germination in batch. The first step is to divide
Experiment the pictures into the separate germination tests. To do so, first
Adobe Bridge is opened, all pics to be analyzed are selected,

and if the pictures are upside down they are rotated first: go to
Edit > Rotate 180° (see Note 29). Subsequently go to
Tools > Photoshop > Batch and select the following settings:
Set: Germinator; Action: Crop and Save; Source: Bridge;
Destination: Folder. Click Source and select the folder in which
the cropped separate pictures should be saved. Click override
action “save as” and click OK (see Note 30).
2. Open ImageJ and go to Plugins > Montpellier RIO
Imaging > MRI VisualScripting. Go to Applications >
Applications > Germinator > germination score + white and
click the big blue button. Select file type as all images and select
all images to be analyzed. Click Open. When ImageJ is fin-
ished, there are two windows. One named Summary, which
should be saved as summary+.txt, and one named Results,
which should be saved as results+.txt (see Note 31). Perform
the same steps on the same pictures for Applications >
Applications > Germinator > germination score − white (see
Note 32) and save the two windows as summary-.txt and
results-.txt, respectively.
3. The first time the GERMINATOR is used, the optimal set-
tings for variances that are allowed for the position of seeds and
differences in seed area will have to be determined and thereby
the performance of the system will be optimized. To do so first
count the number of germinated seeds from some samples and
save them in an Excel file according to the format as shown in
the “manualcount-example file.xls” file in the Excel directory
of the GERMINATOR package.
4. Open the Excel table that was made under Subheading 3.2,
step 7, click Optimize parameters, and select the txt files that
were created under Subheading 3.7, step 2 (see Note 33).
Under Name manual results file, select the Excel file created
under Subheading 3.7, step 3. Select Area and Absolute
(see Note 34) and select the range in which to look for the
optimal settings for both the variance in Area and XY position.
Use first a rough screen (e.g., for our system with Arabidopsis
for Area from 10 to 100 in steps of 10 and for XY variance
from 0 to 2 in steps of 0.2) to get an idea about the optimal
range and do subsequent analyses to zoom in to the best set-
tings (see Notes 35 and 36).
5. To measure germination for the whole germination experi-
ment with the optimal settings, click Calculate; select the txt
files that were created under Subheading 3.7, step 2; fill in
the optimized parameters as obtained under Subheading 3.7,
step 4; and click Calculate. The results file with the cumula-
tive germination table of the experiment will be automatically
saved in the default directory.
3.8 Curve Fitting 1. A cumulative germination table in itself does not provide much
and Interpretation information. Therefore the GERMINATOR also has a curve-
of the Data fitting module to convert the cumulative germination data
into useful parameters that can be used to compare germina-
tion characteristics of seed batches. After the cumulative ger-
mination table has been calculated, there will be an option to
fit curves. Click Yes and select the Germinator_curve-fitting 1
29.xls file in the Excel directory of the GERMINATOR pack-
age and click Save.
2. In the OUTPUT sheet of the curve-fitting file, several param-
eters for the curve fitting can be adjusted: the minimal germi-
nation at which a germination curve will be fitted through the
data, the number of hours until which the area under the curve
(AUC) should be calculated (see Note 37), the critical p-value
that should be used for the statistical analysis (default on 0.05),
which additional parameter should be calculated apart from
t50 (e.g., t10 or t90), a lower limit for the r-square of the fit,
the uniformity parameter that should be calculated (e.g.,
u8416 or u7525), and if letters should be added above the bars
in the graphs to indicate statistical differences between the dif-
ferent measurements.
3. Click GO in the OUTPUT or INPUT sheet to fit all germina-
tion curves and calculate the different parameters per curve.
These will be shown in the OUTPUT sheet. By placing 1 in
column R behind a specific germination sample and clicking
Graph, the curves of the selected samples will be plotted.
4. For germination samples with the same names, averages will be
calculated with standard errors and statistical differences between
samples will be calculated and shown in the statistics sheet.
5. Also after the curves are fitted, some parameters as mentioned
under Subheading 3.8, step 2, can be adjusted and statistical
analysis will be repeated after clicking the “recalculate statistics”
button in the OUTPUT sheet.
4 Notes
1. Our system is optimized for stackable 15 × 21 cm trays (ref 109,

DBP plastics, Belgium. www.dbp.be), but any other germina-
tion tray can also be used. The advantage of the DBP trays is
that they are stackable and allow for convenient storage in
germination incubators in which each tray functions as the lid
of the tray below and as such avoids evaporation as much as
possible.
2. The germination paper should have a color (in wet conditions)
which is sufficiently contrasting to the color of the seed coat
and radicle. We use 14 × 19.5 cm blue blotter germination

paper from Anchor (Anchor paper company, St Paul, MN,
USA; www.seedpaper.com). It can be ordered in any desired
size if it should be used in different size germination trays.
3. Any germination cabinet that is used for normal germination
tests can be used. For working with Arabidopsis it is important
to have light and preferably from the side, so that all trays get
equal exposure to light.
4. As alternative, A3 paper sheets can be placed between the light
and the trays to get a similar effect of indirect lightning and
also the strength and position of the lights can be changed to
optimize the light conditions.
5. There are also specific Photoshop action files for use with 3, 20,
and 24 germination tests per tray. Other formats can be requested
from the author.
6. It is wise to select the cropping areas as broad as possible with
as little as possible chance to select seeds of the neighboring
germination tests.
7. It is important to use ImageJ version 1.40. Newer versions
unfortunately do not support Visual Scripting.
8. Select Germinator_table 2.01.xls if a different number of ger-
mination tests on one tray is used. Version 2.01 allows the
formation of germination tables with any number of germina-
tion tests per tray.
9. The content of the drop-down selectors can be changed or
items can be added to it in the “Experiment descriptors” sheet.
10. Make sure that for the date format in cell B10 “-” is used as
separator instead of “/”. This can be changed in the Region
and Language control panel of the used Windows system
under the Short date format.
11. It is important to use two sheets of germination paper to obtain
a big enough buffer of solution. This will make the system less
sensitive to influences on germination due to evaporation.
12. The amount of solution added to the trays is very important.
Too little will have influence on the germination, but too much
will hamper the image analysis: too much water will cause a
light-reflecting meniscus around the seed that will be mistak-
enly detected as a radicle by the system.
13. Instead of adding water, also salt, mannitol, PEG, ABA, or any
other compound can be added to the trays to analyze the effect
of these compounds on seed germination.
14. The mold is made manually by cutting holes in a plastic sheet.
In our system, a mold for 6 germination tests per tray is used,
but as mentioned before, also molds for, e.g., 2, 3, or 24 ger-
mination tests per tray can be made.
15. The mold is reused for a next tray, after sowing one tray.
16. The amount of seeds depends on the size of the used seeds and
of the holes in the used mold. Take care to have the seeds as
evenly distributed as possible since too many seeds too close
together will hamper reliable automatic scoring of germination.
17. The trays on top and bottom of the pile are there to prevent
unequal evaporation as much as possible.
18. For dormancy breaking purposes, the piles can also be incubated
at 4 °C for 3–4 days before incubation in the germination
cabinet.
19. The 53 cm is the optimal distance for the 15 × 21 cm trays, but
should obviously be adapted for trays with different sizes.
20. The mask does not need to fit the trays very accurately, since
our system is not sensitive to slight changes in position of the
trays at the different time points. Changes up till 5 mm will not
have an effect on the automatic scoring.
21. Low light with longer shutter speed works in general very well
to enhance contrast.
22. In our setup, ISO 400, F/18, and 1/3 s are used, but these
settings have to be defined depending on the specific light
conditions. It is advised to be very punctilious in optimizing
the camera settings. The better the pictures, the better the
automatic scoring will perform.
23. The first picture should be taken after the initial imbibition,
but before the first seeds begin to germinate, since taking a
picture before initial imbibition (imbibition phase I) is com-
pleted will have an effect on the size of the seeds and the count
of non-germinated seeds on the first picture will determine the
total amount of seeds that is counted in the germination test.
24. The frequency of taken pictures depends on the germination
speed. In general it is advised to have at least three pictures in
the steep part of the germination curve. After all germination
tests within an experiment have reached t50, the frequency can
also go down. In general, pictures are taken till the moment
that no further germination is expected.
25. In our setup 0.2 % saturated pixels is used to enhance con-
trast, but slightly other settings might be optimal for other
setups.
26. In our setup the best results are obtained with YUV channels,
but HSB, RGB, or CIE channels might work better in other
setups.
27. Make sure that there is no, or as little as possible, size difference
between both settings for non-germinated seeds, but that
there is an obvious size difference for germinated seeds.
28. The upper size limit will also filter out seeds with long radicles,
but this will not hamper the analysis, because seeds not present
in the + white filter will be automatically counted as germinated
seeds.
29. In our setup pictures are automatically taken upside down due
to the orientation of the camera on the repro-stand.
30. The system is designed for the Photoshop version in the
English language. The names of the files for the cropped
separate germination tests should end with “copy” followed
by a number (e.g., 10212015-0923#001 copy 4.jpg).
31. It is important to save the files with the .txt extension and not
with the default .xls extension due to compatibility issues with
the subsequent analysis by the Excel scripts.
32. Both the germination score + white and germination
score − white analysis can be performed simultaneously, which
saves PC calculation time. To do so, ImageJ has to be started
twice and the subsequent steps have to be followed in each
open instance of ImageJ for germination score + white and ger-
mination score − white, respectively. In this case it is important
to keep the two appearing Summary and Results windows
apart. One way to do so is to drag the two windows of the
germination score − white analysis to the left of the desktop
and the two windows of the germination score + white to the
right of the desktop.
33. When the suggested naming of the txt files is used, only the
first txt file (summary+.txt) has to be selected and Excel auto-
matically will fill in the other three files.
34. In our system with Arabidopsis the Area variance and absolute
variances are used. For other setups and seed types, other set-
tings might work better. Also the Perimeter can be used instead
of the Area of the seeds, or relative differences (%) as compared
to absolute differences.
35. The optimal settings for our system with Arabidopsis seeds are
48 for Area and 0.8 for xy variance.
36. The maximum number of combinations that can be tested at
one time is 255.
37. The AUC is determined until a user-defined time point. This
time point is preferably chosen shortly after most germination
tests have just reached maximum germination. When using an
earlier time point, the AUC will mainly correlate with the t50
and using a much later time point will result in AUCs that are
highly correlated with the maximum germination. If chosen at
the right position, AUC will be a combinational measurement
reflecting aspects of both the t50 and Gmax.
References
1. Penfield S, King J (2009) Towards a systems 6. Teixeira PCN, Coelho Neto JA, Rocha H, De
biology approach to understanding seed dor- Oliveira JM (2007) An instrumental set up for
mancy and germination. Proc R Soc B Biol Sci seed germination studies with temperature
276:3561–3569 control and automatic image recording. Braz
2. Ligterink W, Joosen RVL, Hilhorst HWM J Plant Physiol 19:99–108
(2012) Unravelling the complex trait of seed 7. Wagner M-H, Demilly D, Ducournau S, Dürr
quality: using natural variation through a com- C, Léchappé J (2011) Computer vision for
bination of physiology, genetics and -omics monitoring seed germination from dry state to
technologies. Seed Sci Res 22:S45–S52 young seedlings. Seed Sci 142:49–51
3. Joosen RVL, Kodde J, Willems LA, Ligterink W, 8. El-Kassaby YA, Moss I, Kolotelo D, Stoehr M
van der Plas LH, Hilhorst HW (2010) Germinator: (2008) Seed germination: mathematical repre-
a software package for high-throughput scoring sentation and parameters extraction. Forest Sci
and curve fitting of Arabidopsis seed germination. 54:220–227
Plant J 62:|148–159 9. Nguyen T-P, Keizer P, van Eeuwijk F, Smeekens
4. Dell’Aquila A (2009) Digital imaging infor- S, Bentsink L (2012) Natural variation for seed
mation technology applied to seed germina- longevity and seed dormancy are negatively
tion testing. A review. Agron Sustain Dev 29: correlated in Arabidopsis thaliana. Plant Physiol
213–221 160(4):2083–2092
5. Ducournau S, Feutry A, Plainchault P, Revollon 10. Joosen RVL, Arends D, Willems LAJ,
P, Vigouroux B, Wagner MH (2005) Using Ligterink W, Jansen RC, Hilhorst HWM
computer vision to monitor germination time (2012) Visualizing the genetic landscape of
course of sunflower (Helianthus annuus L.) Arabidopsis seed performance. Plant Physiol
seeds. Seed Sci Technol 33:329–340 158:570–589
Chapter 8
Histochemical Staining of β-Glucuronidase and Its Spatial

Quantification
Chloé Béziat, Jürgen Kleine-Vehn, and Elena Feraru
Abstract
Microscope images of plant specimens showing expression of GUS markers, besides being very beautiful,
provide useful information regarding various biological processes. However, the information extracted
from these images is often purely qualitative, and in many publications is not subjected to quantification.
Here, we describe a very simple quantification method for GUS histochemical staining that enables detec-
tion of subtle differences in gene expression at cellular, tissue, or organ level. The quantification method
described is based on the freely available image analysis software ImageJ that is widely used by the scientific
community. We exemplify the method by quantifying small and precise changes (at the cellular level) as
well as broad changes (at the organ level) in the expression of two previously published reporter lines, such as
the pPILS2::GUS and pPILS5::GUS. The method presented here represents an easy tool for converting visual
information from GUS histochemical staining images into quantifiable data and is of general importance for
plant biologists performing GUS activity-based evaluation of reporter genes.
Key words GUS staining, Reporter gene expression, PILS, Auxin, Quantification, ImageJ
1 Introduction
The β-glucuronidase (GUS)-based reporter system, proposed by

Jefferson and co-authors several decades ago for the analysis of
gene expression in transformed plants, remains a very powerful
and highly sensitive technique to study gene regulation and pro-
tein stability [1–4]. Around 5000 citations recorded in PubMed
show that the GUS reporter system is used in hundreds of labs,
making it one of the most widely used tools in molecular plant
biology [http://www.ncbi.nlm.nih.gov/pubmed/?term=beta-
glucuronidase+plant]. The reporter system contains the Escherichia
coli GUS gene as a fusion marker and provides quantitative
information or visualization of gene expression activity in vivo and
in situ [1–4]. The expression of GUS can be accurately quantified
using methods such as fluorometric assays on tissue extracts, while
its histochemical staining provides visual information on the
73
74 Chloé Béziat et al.
localization of gene activity in cells, tissues, or organs [2, 5].

The methods available for quantification of GUS activity are often
laborious and time consuming, whereas histochemical staining
protocols usually are less labor intensive. However, the latter are
normally employed only in a qualitative manner and, in many pub-
lications, not subjected to quantification.
Here, we describe a simple method for quantification of GUS
activity in histochemically stained specimens, using the open-
access image processing and analysis software ImageJ ([6]; http://
rsb.info.nih.gov/ij/). We exemplify this method by visualizing
and quantifying the expression of GUS activity in roots of trans-
genic pPILS2::GUS and pPILS5::GUS [7]. PILS (PIN-like) pro-
teins are a family of putative auxin carriers that localize to the
endoplasmic reticulum and regulate intracellular auxin homeosta-
sis [7, 8]. pPILS2 and pPILS5 activity depends on their putative
substrate [7].
The most common substrate for GUS histochemical staining is
5-bromo-4-chloro-3-indolyl-beta-d-glucuronide (X-Gluc) that, in
the final reaction, produces a blue precipitate which shows the
localization of transgene expression [2]. The method presented
here requires the conversion of this blue color intensity into gray
values quantifiable with ImageJ software, and encompasses two
main steps: (1) conversion of GUS histochemical staining images
from RGB (stands for Red Green Blue) color mode to HSB (stands
for Hue Saturation Brightness) mode, and (2) quantification of
GUS dye present in the sample by measuring the signal intensity in
the Saturation channel. Saturation is a value depicting the relative
bandwidth of the color information. Since the intensity of the blue
color (from GUS staining) correlates with the saturation value only
in largely transparent tissues, a thorough clearing of the tissue is
strictly required. Notably, a similar approach has been previously
proposed [9], propounding the suitability of measuring the satura-
tion for GUS quantification. The method allows precise quantifica-
tion of subtle intensity differences in GUS histochemically stained
samples that are otherwise used only for the visualization of gene
expression.
2 Materials
2.1 Plant Medium 1. Arabidopsis seedlings are grown in a plant cabinet under long-
and Seedling Growth day photoperiod for 5 days: pPILS2::GUS and pPILS5::GUS in
Arabidopsis thaliana Col-0 background have been previously
published [7].
2. Plant growth medium contains 0.5 g/l MES, 2.3 g/l Murashige
and Skoog salt, 10 g/l sucrose, and 8 g/l plant agar (pH 5.9
KOH).
Histochemical Staining of β-Glucuronidase and Its Spatial Quantification 75
3. In vitro growth is carried out in standard, sterile square Petri

dishes (120 × 120 × 17) with vents.
4. 70 % Ethanol is used for seed sterilization.
2.2 Auxin Treatment 1. 24-Well culture plates.

2. Featherweight entomological tweezers.
3. 1-Naphthaleneacetic acid (NAA) solved in dimethyl sulfoxide
(DMSO).
2.3 GUS Staining 1. 24-Well culture plates.

2. Cold acetone 90 % (optional) (see Note 1).
3. Na-phosphate buffer (0.1 M; pH 7) made of 39 ml (2.76 g to
100 ml H2O) NaH2PO4·H2O (0.1 M final concentration),
61 ml (5.35 g to 100 ml H2O) Na2HPO4·H2O (0.1 M final
concentration), and 100 ml water. Adjust pH to 7 with
NaH2PO4·H2O solution. Add 0.1 % Triton.
4. Ferro-ferricyanide buffer (5 mM) made of K3Fe(CN)6 (0.08 g)
and K4Fe(CN)6 (0.105 g) dissolved in 50 ml Na-phosphate
buffer (see item 3).
5. Dimethylformamide (DMF) solvent.
6. X-Gluc dissolved freshly as 1 mg X-Gluc in 10 μl DMF
( see Note 2).
7. GUS staining buffer made of seven parts Na-phosphate buffer
(see item 3), three parts ferro-ferricyanide buffer (see item 4)
and 10 mg (1 mg/ml) freshly dissolved X-Gluc (see item 6).
8. Ethanol:acetic acid (3:1).
9. 70, 50, and 20 % ethanol.
10. Clearing solution made of 90 ml H2O, 240 g chloral hydrate,
and 30 ml glycerol.
11. Thermo incubator set at 37 °C.
12. Stereomicroscope to monitor the staining.
2.4 Sample 1. Tweezers.

Preparation, Imaging, 2. Microscope slides and cover glasses.
and GUS
3. Transmitted light microscope with attached digital color camera.
Quantification
4. ImageJ 1.41 software (http://rsb.info.nih.gov/ij/).
3 Methods
3.1 Seedling Growth 1. Sterilize the seeds. We soaked the seeds for 1–5 min in 70 %
ethanol.
2. Remove ethanol and allow the seeds to dry well in a laminar
flow hood.
3. Sow the seeds in a Petri dish containing solid plant medium.

4. Stratify the seeds for a minimum of 2 days (up to 4 days) by
keeping the plates (wrapped in aluminum foil) at 4 °C in a
fridge or cold room.
5. Grow the seedlings in a growth chamber or cabinet for 5 days
(see Note 3).
3.2 Auxin Treatment 1. We incubated pPILS2::GUS and pPILS5::GUS seedlings in MS

liquid medium containing 500 nM NAA or the same amount
of DMSO (solvent control) for 12 h.
3.3 Histochemical 1. Fixation (optional) (see Note 1):

Staining with X-Gluc (a) Put about 20 seedlings in 1 ml cold acetone for at least
30 min.
(b) Wash the seedlings at room temperature in 1 ml Na-phosphate
buffer for minimum 30 min (2–3 times). Shake occasionally.
2. Histochemical staining with X-Gluc:
(a) Put or transfer the fixed seedlings into 1 ml prepared staining
buffer (see Subheading 2.3, item 7), wrap the incubation
plate in aluminum foil, and place it at 37 °C.
(b) Observe the samples regularly for optimal staining (see
Note 4). Avoid over-staining the zone of interest. For the
here used pPILS2::GUS and pPILS5::GUS we stained for
6 h and 2 h, respectively.
3. Clearing:
(a) Transfer the stained seedlings into 1 ml mixture of
ethanol:acetic acid (see Subheading 2.3, item 8) until
complete disappearance of chlorophyll.
(b) Wash the seedlings with 1 ml 70 %, 50 %, and 20 % ethanol,
each step for 10 min or longer.
(c) Transfer the seedlings into an appropriate volume of (all
seedlings should be submerged) clearing solution (see
Subheading 2.3, item 10) and keep them overnight or
longer at 4 °C. It is crucial to perform a thorough clearing
for optimal reduction of optical density and because all
residual (non-blue) color needs to be removed to allow its
reliable ImageJ-based quantification.
3.4 Sample 1. Place the overnight-cleared seedlings on microscope slides.

Preparation Add few drops of chloral hydrate as a mounting solution.
and Imaging Carefully place the cover slip on the seedlings.
2. Observe the samples with a light microscope using bright-field
illumination (see Note 5).
3. Take photos of the GUS-stained regions of interest (see Notes 6
and 7).
3.5 ImageJ-Based 1. Prior to quantification, convert the color mode of light micro-
Quantification scope images from RGB to HSB by using ImageJ software
of Histochemical GUS function: Image → Type → HSB stack.
Staining 2. Select Saturation channel in the HSB stack (scroll the bar to
the second position).
3. Measure the intensity of GUS staining in the Saturation channel.
An increase of saturation depicts more “pure” color, while a
decrease denotes a more “washed-out” signal. We thoroughly
cleared the tissue and therefore the color information corre-
lates with the degree of blue staining. Thus, this channel
reflects the intensity of GUS dye present in the sample.
4. Display the results, taking into consideration the biological
process that is being investigated (various ImageJ software
options are available for this). Here, we (1) compared different
cells (Fig. 1c, d), cell files (Fig. 1c, d), or regions (Fig. 1a, b)
of the same sample; (2) compared different samples, such as
NAA-treated and -untreated (Fig. 2a–c); and (3) used various
region of interest (ROI) selections, such as rectangular (Figs. 1a
and 2a, b) or linear (Fig. 1c) to define where to measure the
GUS intensity. Moreover, we described two very basic quanti-
fication methods, focusing on intensity profile [rectangular
profile (see step 5; Fig. 1a, b) or line profile (see step 6; Fig. 1c, d)]
and mean intensity depiction (see step 7; Fig. 2a–c).
5. For quantifying differences in the intensity of GUS staining
observed in the same sample, we used roots of pPILS5::GUS
reporter seedlings that showed weak GUS staining in the meri-
stem and strong GUS staining above the meristem (Fig. 1a).
To depict these differences we performed an intensity profile
of the GUS signal present in a defined region:
(a) Define a rectangular ROI by using “Rectangular selec-
tion” drawing tool from ImageJ toolbar (Fig. 1a).
(b) Plot the intensity values measured within the selected
region: Analyze → Plot profile (Fig. 1b).
(c) To obtain the plot values, press “List” in the Plot window
(see Note 8).
6. To quantify subtle differences (or differences at cellular level)
in GUS staining we used roots of pPILS2::GUS reporter line
and performed an intensity profile of the GUS signal measured
along defined linear ROIs drawn crossing cells or crossing cell
files (Fig. 1c):
(a) Draw lines crossing the specific cell files (blue and gray
lines) by using the ImageJ drawing tool “Straight line
selection” (Fig. 1c).
(b) Plot the values measured along the linear ROIs: Analyze →
Plot profile (Fig. 1d).
Fig. 1 Quantification of GUS histochemical staining in roots of pPILS5::GUS (a and b) and pPILS2::GUS (c and
d) marker lines. (a) RGB (upper) or HSB (lower) color mode images showing histochemical staining of graded
pPILS5::GUS expression in root tip. The blue box in the HSB image depicts the rectangular ROI for measuring
the GUS staining intensities. (b) Graph showing the profile of GUS intensity measured within the rectangular
ROI depicted in the HSB image (a). Note the weak expression of pPILS5::GUS in the meristem and strong
expression above the meristem. (c) RGB (upper) or HSB (lower) color mode images showing cell- and cell file-
specific histochemical staining of pPILS2::GUS expression in root tip. Drawings in HSB image represent linear
ROIs crossing neighboring root epidermal cell files (blue and gray lines), showing distinct GUS staining. (d)
Graph depicts the profile of GUS intensity measured along the linear ROIs depicted in HSB image (c). Note the
gradual increase in pPILS2::GUS expression (blue profile). Moreover, pPILS2::GUS expression is distinct in
neighboring epidermal cell files showing strong (blue profile) and weak (gray profile) staining
(c) To visualize and analyze the signal intensity differences

between the two distinct cell files, copy the measured plot
values from the Plot window (press “List” in the Plot win-
dow) into Excel or any other graphing program of choice
(see Note 9).
7. For quantifying changes in GUS activity of a certain reporter
gene, measure the signal intensity across an entire image or a
defined region. We used pPILS5::GUS seedlings treated with
NAA, and quantified the mean gray value in a selected region of
the root, showing an increase after NAA treatment (Fig. 2a–c):
(a) Define the shape and size of ROI, where the signal should
be measured by using the ImageJ drawing tool “rectangular
selection” (Fig. 2a, b).
(b) Measure the mean gray value of the defined ROI: Analyze
→ Measure.
(c) Display the measured values in graphs (Fig. 2c). Compare
signal intensity between auxin-treated and non-treated
seedlings (see Note 8).
Fig. 2 Quantification of GUS histochemical staining of pPILS5::GUS in roots treated with auxin. (a and b) RGB
(left) or HSB (right) color mode images showing histochemical staining of pPILS5::GUS expression in roots
treated with DMSO (a) or NAA (b). Drawings in HSB images represent identical rectangular ROIs defined in
DMSO-treated (a) and NAA-treated (b) pPILS5::GUS expressing roots. (c) Graph showing the mean GUS inten-
sity measured within the rectangular ROIs depicted in HSB images (a and b). n = 10 seedlings, ***p < 0.0001
4 Notes
1. Fix your seedlings because the activity of the investigated pro-

moter could be responsive to heat or dark stresses.
2. Commercially available X-Gluc salts deviate in their chemical
properties and hence may show distinct tissue perfusion that
could affect staining outcome.
3. Seedlings can be younger or older, depending on the biological
processes and reporter genes that are being investigated.
4. Some GUS reporter markers show strong GUS staining within
minutes while others must be stained much longer.
5. Samples can be stored at 4 °C for several weeks, but preferen-
tially should be imaged immediately.
6. Keep the same magnification and microscope settings during
the same experiment.
7. Image more than ten seedlings per line and experiment. Repeat
the experiment at least three times to confirm the results.
8. Analyze identical ROIs (shape and size) in all images used for
comparison: Edit → Selection → Restore Selection.
9. Use linear ROIs with an identical length throughout the
experiment (Edit → Selection → Restore Selection).
Acknowledgements
We thank Esteban Fernandez for his online posts to scientific ques-

tions and David Whittaker for help with the manuscript. This work
was supported by the Vienna Science and Technology Fund
(WWTF) (Vienna Research Group), Austrian Science Fund (FWF)
(P26568-B16; P26591-B16 (to J.K.-V.) and T-728-B16 (to
E.F.)), and the European Research Council (ERC) (Starting Grant
639478-AuxinER (to J.K.-V.)).
References
1. Jefferson RA, Burgess SM, Hirsh D (1986) reporter of gene expression. Academic, San
beta-Glucuronidase from Escherichia coli as a Diego, pp 47–59
gene-fusion marker. Proc Natl Acad Sci U S A 6. Abramoff MDM, Paulo J, Ram SJ (2004)
83(22):8447–8451 Image processing with ImageJ. Biophotonics
2. Jefferson RA, Kavanagh TA, Bevan MW (1987) Int 11(7):36–42
GUS fusions: beta-glucuronidase as a sensitive 7. Barbez E, Kubes M, Rolcik J, Beziat C, Pencik A,
and versatile gene fusion marker in higher Wang B, Rosquete MR, Zhu J, Dobrev PI, Lee
plants. EMBO J 6(13):3901–3907 Y, Zazimalova E, Petrasek J, Geisler M, Friml J,
3. Jefferson RA (1987) Assaying chimeric genes in Kleine-Vehn J (2012) A novel putative auxin car-
plants: the GUS gene fusion system. Plant Mol rier family regulates intracellular auxin homeosta-
Biol Rep 5:387–405 sis in plants. Nature 485(7396):119–122
4. Jefferson RA (1988) Plant reporter genes: the 8. Barbez E, Kleine-Vehn J (2013) Divide Et
GUS gene fusion system. In: Setlow JK (ed) Impera—cellular auxin compartmentalization.
Genetic engineering, vol 10. Plenum Publishing Curr Opin Plant Biol 16(1):78–84
Corporation, New York 9. Pozhvanov GA, Medvedev SS (2008) Auxin
5. Gallagher SR (1992) Quantitation of GUS quantification based on histochemical staining
activity by fluorometry. In: Gallagher SR (ed) of GUS under the control of auxin-responsive
GUS protocols: using the GUS gene as a promoter. Russ J Plant Physiol 55(5):706–711
Chapter 9
Imaging TCSn::GFP, a Synthetic Cytokinin Reporter,

in Arabidopsis thaliana
Jingchun Liu and Bruno Müller
Abstract
Cytokinins are classical plant hormones that control numerous developmental processes throughout the
plant life cycle. Cytokinin-responsive cells activate transcription via a phospho-relay signaling network.
Type-B nuclear RESPONSE REGULATOR (RR) proteins mediate transcriptional activation as the final
step in the signaling cascade. They bind to promoters of immediate-early target genes via a conserved Myb-
related DNA-binding domain. To monitor transcriptional activation in response to a cytokinin stimulus, we
have constructed a synthetic promoter, TCS (two-component signaling sensor) that harbors the concate-
merized binding motifs for activated type-B RR in an optimized configuration. Here, we describe our
protocols for imaging TCSn::GFP expression in transgenic Arabidopsis plants. The use of the fluorescent
reporter GFP allows the visualization of cytokinin-responding cells by fluorescent microscopy without the
need for tissue processing steps, or staining reactions. This method is fast and with a low risk of artifacts.
However, since cytokinin signaling integrates various environmental information including light, nutrient
status, and biotic and abiotic stress, special care needs to be devoted to the control of growth conditions.
Key words Cytokinin signaling, Synthetic reporter, Fluorescent microscopy, Plant development,
Plant physiology, Arabidopsis thaliana
1 Introduction
Development of a multicellular organism relies on differential gene

expression triggered by instructive signals. In plants, chemical hor-
mones such as cytokinins are key signals to govern cell specification
and growth of selected cells, as well as to mediate the response to
environmental cues ranging from light and stress to nodulation [1,
2]. Cytokinins are a class of adenine N6-substituted organic mol-
ecules that trigger transcriptional changes in responding cells. The
production of active cytokinins, the nucleobases, involves a num-
ber of enzymatic processes that occur in different organs, tissues,
and cellular compartments. Cytokinins are then subjected to trans-
port, modifications, and degradation. Each of these processes is
regulated, leading to tight control of spatiotemporal distribution
of active cytokinins in planta. Cytokinin-responsive cells activate
81
82 Jingchun Liu and Bruno Müller
transcription via a phospho-relay signaling network that involves

the following steps: cytokinins bind to the CHASE (cyclases/histi-
dine kinase-associated sensory extracellular) domains of transmem-
brane hybrid kinases, which autophosphorylate at a conserved
histidine. ARABIDOPSIS HISTIDINE TRANSFER PROTEINS
(AHP) proteins shuttling between cytosol and nucleus then trans-
fer the phosphoryl group from receptors to the nuclear
ARABIDOPSIS RESPONSE REGULATOR (ARR) proteins.
Upon phosphorylation, type-B nuclear RR proteins mediate tran-
scriptional activation as the final step in the signaling cascade. They
bind to promoters of immediate-early target genes via a Myb-
related DNA-binding domain, which is conserved from algae to
flowering plants [3], indicating that type-B RR from different
plant species recognize common DNA-binding motifs.
To monitor transcriptional activation in response to a cytokinin
stimulus, we have constructed a synthetic promoter, TCS (two-
component signalling sensor), that harbors the concatemerized bind-
ing motifs for activated type-B RR in an optimized configuration.
Combined with a minimal promoter and a reporter gene, TCS-based
reporters allow the specific and sensitive detection of activated type-B
RR [4, 5]. By now, TCS-based reporters in planta have been widely
used to visualize the cytokinin response in different developmental
contexts, including the female gametophyte [6], gynoecium [7],
seed [4, 8], root [9, 10], shoot meristem [11], biotic stress [12], and
many others. In combination with short-term treatments or induc-
ible transgenes, the use of the synthetic reporter allows studying
changes in the immediate-early responses to cytokinin at cellular res-
olution, before secondary effects mask the phenotypes. Crosses with
permanent mutants can also reveal relevant information. The first
version of the TCS reporter [4] suffered from silencing, an issue that
was corrected in the second version, TCSn [5].
In addition, mesophyll protoplast transiently transfected with
TCS::LUCIFERASE (LUC) reporters can serve as a cellular model
for cytokinin-responsive cells, with the LUC activities representing
the quantitative cytokinin response [4, 5, 10]. This system is ideal
to rapidly screen high numbers of candidate genes and different
treatments for potential functions in cytokinin signaling. Readers
interested in applying this method are referred to a detailed proto-
col describing transient transfection assays [13]. Here, we describe
our protocols for imaging TCSn::GFP expression in various organs
derived from transgenic Arabidopsis plants and seedlings using
confocal microscopy. The choice of a fluorescent reporters such as
GFP is advantageous due to the visualization of cytokinin-
responding cells without the need for tissue processing steps or
staining reactions. Thus, the method is fast and the risk of artifacts
is low. However, since cytokinin signaling integrates various envi-
ronmental information including light, nutrient status, and biotic
and abiotic stress, special care needs to be devoted to control the
growth conditions (see Fig. 1A–H).
TCSn::GFP Imaging 83
Fig. 1 Parameters influencing TCSn::GFP expression in the Arabidopsis seedling root apices. Confocal micro-
graphs showing TCSn::GFP expression in the root tip of a representative seedling grown at the indicated condi-
tion. If not otherwise noted, seedlings were germinated on vertical plates with solid medium as listed in
Subheading 2, and grown for 7 days at 12-h light at an intensity 90 μmol/m2/s at 22 °C during the day and at
18 °C during the night. Of each condition, five seedlings were analysed. (A) Short-day conditions (8-h light) and
(B) long-day conditions (16-h light). (C) Seedlings grown on medium without sucrose and (D) medium contain-
ing high amounts of sucrose, and on medium without (E) or with (F) 100 mM NaCl to simulate osmotic stress.
(G) 6-day-old seedlings compared to (H) 9-day-old seedlings. (I–L) Seedlings were transferred to liquid
medium after 7 days and incubated overnight with indicated hormones at a concentration of 1 μM. (I) Mock-
treated, (J) trans-zeatin (tZ), (K) isopentenyl-adenine (iP), and (L) indole acetic acid (IAA)-treated samples.
Arrows point to changes in GFP expression. Scale bar 50 μm
2 Materials
Make all solutions using ultrapure water (with resistivity greater

than 18 mΩ cm) at room temperature.
2.1 Media, Reagents, 1. Liquid growth medium: Combine half-strength MS (Murashige

and Disposables and Skoog) basal salt mixture without vitamins (2.15 g per
liter) with 1 % (wt/vol) sucrose, and adjust to pH 5.7 using
1 M KOH. Autoclave for 20 min at 15 psi, at 121 °C.
2. Solid growth medium: Add phytagar at a concentration of
0.8 % (wt/vol) as a gelling agent to liquid growth medium
(see Note 1). Autoclave for 20 min at 15 psi, at 121 °C. Mix
well, and place in 60 °C oven. When the medium has reached
60 °C, pour 50 ml of medium to each square Petri dish in a
laminar flow hood.
3. Seed sterilization solution. Prepare 5 % (vol/vol) sodium hypo-
chlorite and 0.05 % (vol/vol) Tween 20 in water.
4. Standard soil substrate type ED 73.
5. Square plastic pots, 80 mm diameter.
6. Square Petri dishes, 120 × 120 × 15 mm.
7. 6-Well tissue culture plates.
8. Agarose, low-melting temperature, forms gel at <30 °C
(Sigma-Aldrich, cat. no. A9045-10G).
9. Mounting medium for shoot apical meristems: Weigh 0.1 g of
low-melting-temperature agarose in a 15 ml tube to prepare a
1 % (wt/vol) solution. Add 10 ml of liquid growth medium,
mix it thoroughly by vortexing, and place the tube in a water
bath set at 70 °C. Incubate for at least 15 min with occasional
vortexing until the agarose has dissolved. Alternatively, use a
microwave oven to melt the mounting solution. Once the aga-
rose is dissolved and the solution is clear and homogeneous,
set the heating block to 40 °C and allow agarose to cool down
to this temperature. Keep the agarose solution at 40 °C to use
for mounting floral meristems. Dissolved aliquots may be
stored for later use at 4 °C for 1–2 months.
10. 1 ml Disposable syringe with 30 G needle.
11. Microscope slides.
12. 5-Well pattern printed microscope slides (Tekdon 6-101,
Myakka City, Florida, USA).
13. 24 × 50 mm and 18 × 18 mm (0.13–0.16 mm) cover slips.
2.2 Equipment 1. Forceps (Dumont 0208-5-PS, Actimed SA, Saint-Sulpice,

Switzerland) (see Note 2).
2. Greenhouse and growth chambers with controlled light and
temperature conditions.
3. Sterile laminar flow hood.
4. Autoclave.
5. Heated water bath.
6. Stereomicroscope (e.g., Leica MZFLIII).
7. Confocal microscope, equipped with adequate lenses. A Leica
SP5 II, quipped with Leica HC PL APO CS2 20× 0.75 IMM,
HCX PL APO CS 40× 1.3, and Leica HC PL APO CS2 63×
1.3 GLYC lenses were used in this study.
8. Imaging acquisition software (Leica Application Suite v2.83).
9. 3D-rendering software Imaris 8.1.1 (Bitplane, Zurich,

Switzerland).
10. Adobe Photoshop CC, Adobe Illustrator CC.
3 Methods
3.1 Seed 1. Place seeds from transgenic plants of Arabidopsis thaliana,

Sterilization ecotype Columbia (Col-0), stably expressing a TCSn::GFP
and Preparation reporter transgene [5] (see Note 3) into a 1.5 ml microcentri-
fuge tube and add 1 ml of sterilization solution. Work in the
laminar flow hood.
2. Wait for 15 min, let the seeds sink down to the bottom of the
tube, and remove the supernatant.
3. Wash the seeds 3–4 times with sterile water, and wait each time
till seeds sink to bottom before removing supernatant.
4. To break seed dormancy and stimulate synchronized germina-
tion, stratify seeds for 72 h at 4 °C.
5. Resuspend the seeds in 1 ml sterilized 0.1 % agarose solution in
a laminar flow hood. The seeds should float in the solution at
a density that allows dispensing individual seeds by pipetting.
3.2 Growing 1. Place about 20 seeds evenly spaced on a line, about 2 cm from
Seedlings on MS the edge of a square 120 × 120 mm Petri dish with appropriate
medium solid growth medium using a sterile pipette tip.
2. Put the plates vertically into a plant growth chamber providing
the desired light and temperature conditions, typically 12-h
light at an intensity of 75–100 μmol/m2/s at 22 °C during the
day and at 18 °C during the night (see Note 4).
3. Grow the seedlings for the desired time, usually between 3 and
10 days.
4. Optional: Add 3 ml of liquid growth medium to each well of a
6-well tissue culture plate. Carefully transfer five to seven seed-
lings with fine forceps to each well. Hormones and other signals
at various concentrations may be added to the liquid growth
medium to test for effects on TCSn::GFP expression. Typically,
overnight incubations with signals at different concentrations
will be informative. As positive controls, include cytokinins at
1 μM to stimulate TCSn::GFP expression, and as negative
controls, include a mock-treated sample (see Fig. 1I–L).
3.3 Growing Plants 1. Fill pots with soil to the rim, and compress lightly. Wet the soil
on Soil with tap water (see Note 5).
2. Dispense individual seeds directly on soil using a pipette tip,
typically one or two seeds per 8 cm plastic pot.
3. Transfer the tray to a greenhouse or growth chamber that

provides the desired light, temperature, and humidity condi-
tions (see Note 6).
4. Cover the tray containing the pots with a transparent lid to
maintain high humidity for about 4 days till cotyledons or ger-
minating seedlings are fully expanded.
5. Remove the lid. Water upon need; avoid drought or water
stress (see Note 7).
6. Grow until the desired organs have fully developed (see Note 8).
3.4 Mounting Organs 1. Mounting of seedlings under stereomicroscope: Dispense

and Seedlings 200 μl of liquid growth medium on a microscope slide. Using
for Imaging fine dissection forceps, gently grasp the seedlings from the agar
plates, transfer them onto a microscopic slide (see Note 9),
cover with a 24 × 50 mm cover slip, and immediately proceed
to imaging (see Note 10).
2. Preparing and mounting female gametophytes under the
stereomicroscope.
Harvest pistils from flower and place it on microscope
slide. Slit open the pistil replum on both sides using injection
needle and forceps. Add the appropriate amount of liquid
medium, about 50 μl, and cover with a 24 × 50 mm cover slip.
Proceed immediately to imaging with the confocal microscope
(see Note 10).
3. Dissecting seeds to recover and mount embryos under
stereomicroscope.
Collect a silique (see Note 11), and place on a 5-well pat-
tern printed microscope slide. Add 10 μl of liquid medium to
the silique and also to a neighboring printed well. Slit open
the silique with the forceps. Transfer about three to four seeds
into the droplet previously placed to the neighboring well.
Dissect embryo out of the seed coat (see Note 12). Cover
with a 18 × 18 mm cover slip. Immediately proceed to imaging
(see Note 10).
4. Preparing and mounting shoot apical meristems under
stereomicroscope:
Remove an inflorescence from the plant using forceps.
Place on a microscope slide. It is necessary to remove the older,
larger flower buds to expose the floral meristems. To do so, cut
as close to the base of the meristem as possible, perpendicular
to the shoot axis (see Note 13). If necessary, remove additional
older flower buds that block the view to meristem. Add a drop
of warm and liquid mounting medium for shoot meristems to
a slide. Transfer the dressed meristem to the drop. The meri-
stem should be completely immersed without air bubbles,
and with the meristem facing up. Add a 18 × 18 mm cover slip.
Do not push down. By gently repositioning the cover slip, you

may also optimize the meristem position. In the meantime, the
mounting medium will have cooled and the low-temperature
melting agarose solidified, fixing the meristem in the desired
position. Immediately proceed to imaging (see Note 10).
3.5 Confocal Laser Seedling root tips may be recorded using a 20× immersion lens to
Scanning Microscopy obtain an overview of the root apex as shown in Fig. 1. The shoot
Imaging (See Note 14) meristem shown in Fig. 2C was recorded with 40× oil lens, and
female gametophytes or embryos (see Fig. 2A, B) were recorded
with a 63× immersion lens. Reduce laser power to minimize photo-
damage that may lead to autofluorescence. The window of the
detector for GFP emission was set from 510 to 550 nm. Transmitted
light was captured in parallel in transmitted light mode for informa-
tion about cell outlines and morphology. Tissues were imaged in
three dimensions, and interval of sections, also called the axial sam-
pling rate, was chosen to fulfil the Nyquist criterion [14]. When
comparing and presenting data from different samples, it is impor-
tant to include image data representing equivalent regions to avoid
bias. Generally, it is preferable to choose parameters that provide fast
Fig. 2 TCSn::GFP expression (A) at an early stage of female gametophyte development, (B) heart stage, and (C)
in the floral meristem. The GFP signal is shown in green, while transmitted light is shown in grey to outline the
morphology of the organ. Scale bar 10 μm
scanning and minimal bleaching, but may produce some more noise
compared to slow scanning modes that provide higher resolution
and less noise but lead to bleaching. Stacks of recorded confocal
sections may be processed by Imaris software for measurements,
projections, and 3D reconstructions. Final images may then be
trimmed and enhanced in Adobe Photoshop. Apply all image pro-
cessing equally to all samples and the whole area. Composite images
may be produced and assembled using Adobe Illustrator.
4 Notes
1. Depending on the biological question, variations to the com-

position of the solid medium may be introduced. This includes
addition of NaCl to simulate salt stress (see Fig. 1E, F), variation
in sucrose (see Fig. 1C, D) or nitrate content to vary nutrient
conditions, or the addition of signals such as phytohormones.
2. These forceps are very convenient for dissecting and handling
small organs such as embryos or shoot meristems, in conjunc-
tion with needles. Be careful in handling the forceps, as they
are very fine and delicate. The tips will easily bend or break
upon touching objects.
3. Depending on the biological question, different mutants or
transgenic plants may be crossed to the TCSn::GFP reporter
line, and used as experimental conditions to be compared against
the reporter transgene in wild-type genetic background.
4. Age and light conditions may affect TCSn::GFP expression; see
Fig. 1A, B, G, H. Control these parameters for consistent
results.
5. To restrict the growth of insect larvae of various species, toxin
formulations of Bacillus thuringiensis may be added to the
water; we use SolBac (Andermatt Biocontrol, Grossdietwil,
Switzerland).
6. As illustrated in Fig. 1, different environmental conditions (light,
nutrition, stress) may affect TCSn::GFP expression, which
emphasizes the need for diligent control of all parameters.
7. Water consumption depends on growth stage, which usually
cannot be accounted for by automatic watering systems that
run on fixed schedules. It is recommended to resort to manual
watering.
8. Avoid using organs from plants showing senescence, or from
stressed plants, unless you are interested in studying the
influence of age or stress.
9. Cut off and remove shoot, or root, respectively, depending on
which tissue will be imaged to facilitate flat mounting.
10. An advantage of this protocol is the use of fresh and living tissue.
Therefore, refrain from accumulating samples but work quickly
through the different steps for each sample to minimize the
risk of signal loss and artifacts due to dying cells.
11. Embryo stages correlate to silique age. For guidance, we found
that the fourth silique, counted from the top, without any
sepal or petal remnants typically contains embryos ranging
from late globular to transition stage, and the fifth silique con-
tains embryos around the heart stage.
12. This procedure needs practice. Embryos older than heart stage
are relatively easy to release from the seed coat, while young
embryos are tricky, especially with the suspensor left intact.
In our experience it is best to use only one or two well-placed
cuts to expose the embryo, and then gently grasp the embryo
with the tips of the forceps, relying on the cohesive force of the
liquid medium rather than on pressure from the forceps tips to
free the embryo from the coat.
13. If cut too far away from the base, it cannot be easily corrected
later.
14. For each tissue more sophisticated protocols allowing deeper
tissue penetrance, or co-staining with multiple markers, are
available. However, these protocols often require special equip-
ment, or additional processing steps.
References
1. Kieber JJ, Schaller GE (2014) Cytokinins. Mayo VM, de Folter S (2012) The role of
Arabidopsis Book 12:e0168 cytokinin during Arabidopsis gynoecia and
2. Hwang I, Sheen J, Müller B (2012) Cytokinin fruit morphogenesis and patterning. Plant
signaling networks. Annu Rev Plant Biol J 72:222–234
63:353–380 8. De Rybel B, Adibi M, Breda AS, Wendrich JR,
3. Pils B, Heyl A (2009) Unraveling the evolution Smit ME, Novák O, Yamaguchi N, Yoshida S,
of cytokinin signaling. Plant Physiol 151: Van Isterdael G, Palovaara J, Nijsse B,
782–791 Boekschoten MV, Hooiveld G, Beeckman T,
4. Müller B, Sheen J (2008) Cytokinin and auxin Wagner D, Ljung K, Fleck C, Weijers D (2014)
interaction in root stem-cell specification during Plant development. Integration of growth and
early embryogenesis. Nature 453:1094–1097 patterning during vascular tissue formation in
Arabidopsis. Science 345:1255215
5. Zürcher E, Tavor-Deslex D, Lituiev D, Enkerli
K, Tarr PT, Müller B (2013) A robust and 9. Antoniadi I, Plačková L, Simonovik B, Doležal
sensitive synthetic sensor to monitor the tran- K, Turnbull C, Ljung K, Novák O (2015)
scriptional output of the cytokinin signaling Cell-type-specific cytokinin distribution within
network in planta. Plant Physiol the Arabidopsis primary root apex. Plant Cell
161:1066–1075 27:1955–1967
6. Bencivenga S, Simonini S, Benková E, 10. Bielach A, Václavíková K, Marhavý P, Duclercq
Colombo L (2012) The transcription factors J, Cuesta C, Müller B, Grunewald W, Tarkowski
BEL1 and SPL are required for cytokinin and P, Benková E (2012) Spatiotemporal regula-
auxin signaling during ovule development in tion of Arabidopsis lateral root organogenesis
Arabidopsis. Plant Cell 24:2886–2897 by cytokinin. Plant Cell 24:3967–3981
7. Marsch-Martinez N, Ramos-Cruz D, Irepan 11. Chickarmane VS, Gordon SP, Tarr PT, Heisler
Reyes-Olalde J, Lozano-Sotomayor P, Zuniga- MG, Meyerowitz EM (2012) Cytokinin
signaling as a positional cue for patterning the cell division and orchestrates feeding site for-
apical-basal axis of the growing Arabidopsis mation in host plants. Proc Natl Acad Sci U S
shoot meristem. Proc Natl Acad Sci U S A A 112:12669–12674
109:4002–4007 13. Yoo SD, Cho YH, Sheen J (2007) Arabidopsis
12. Siddique S, Radakovic ZS, De La Torre CM, mesophyll protoplasts: a versatile cell system
Chronis D, Novák O, Ramireddy E, Holbein J, for transient gene expression analysis. Nat
Matera C, Hütten M, Gutbrod P, Anjam MS, Protoc 2:1565–1572
Rozanska E, Habash S, Elashry A, Sobczak M, 14. Pawley JB (2006) Points, pixels and gray levels:
Kakimoto T, Strnad M, Schmülling T, digitzing image data. In: Pawley JB (ed)
Mitchum MG, Grundler FM (2015) A para- Handbook of biological confocal microscopy.
sitic nematode releases cytokinin that controls Springer, New York
Chapter 10
Highlighting Gibberellins Accumulation Sites in Arabidopsis

thaliana Root Using Fluorescently Labeled Gibberellins
Hilla Schayek, Eilon Shani, and Roy Weinstain
Abstract
The physical location of plant hormones is an important factor in maintaining their proper metabolism,
perception, and mediated developmental responses. Thus, unveiling plant hormones dynamics at the
molecule’s level is essential for a comprehensive, detailed understanding of both their functions and the
regulative mechanisms they are subjected to. Here, we describe the use of fluorescently labeled, bioactive
gibberellins (GAs) to highlight the dynamic distribution and accumulation sites of bioactive GAs in
Arabidopsis thaliana roots by confocal microscopy.
Key words Plant hormones, Gibberellins, Fluorescent labeling, Fluorescence microscopy, Chemical
biology
1 Introduction
Gibberellins are a class of tetracyclic diterpene carboxylic acids,

functioning as plant hormones (phytohormones) and influencing a
range of developmental processes including dormancy, germina-
tion, root and shoot elongation, and flowering [1, 2]. Over the
years, more than 130 GAs have been identified, of which only a
handful are bioactive (GA1, 3, 4, 7). Plants exert tight regulation over
their GA response pathways at multiple levels including biosynthe-
sis, metabolism, and perception [3, 4]. Clearly, to exert their
intended roles, GA must be present in correct quantity, at the cor-
rect location and in the correct time. To understand how plants
synchronize spatial localization, concentration, and timing of GA
at their sites of action or metabolism, it is crucial to understand GA
dynamics at the molecule’s level.
Fluorescence microscopy is a minimally invasive technique that
enables high-resolution, real-time tracking of fluorescent molecules
in complex biological environments [5, 6]. The specificity and sensi-
tivity of fluorescence microscopy allow the precise location of intracel-
lular components, which can then be monitored over time, as well as
91
92 H. Schayek et al.
understanding of their associated diffusion coefficients, transport

characteristics, and interactions with other biomolecules. As most
biological molecules of interest do not inherently fluoresce, their
labeling with appropriate fluorescent tags is a perquisite for detec-
tion by fluorescence microscopy. Whereas molecular biology tech-
niques are applied to label proteins of interest with fluorescent
protein tags [7], synthetic chemistry provides access to fluorescent
labeling of small bioactive molecules [8, 9], such as plant hormones,
by small-molecule fluorophores [10, 11]. We, and others, have pre-
viously demonstrated the substantial utility of synthetic labeling
techniques in studying the location, and involvement in dynamic
processes, of plant hormones including GA [12], auxin [13], brassi-
nosteroids [14], and strigolactones [15]. Here, we describe how the
bioactive, fluorescently labeled GAs can be used to highlight the
dynamic distribution and accumulation sites of bioactive gibberellins
in Arabidopsis thaliana root.
The synthetic procedures utilized to label GA3 and GA4 with
the small-molecule fluorophore fluorescein, as well as a compre-
hensive analysis of the labeled GAs, termed GA3-Fl and GA4-Fl,
bioactivity retention, have been reported before [12] and are not
described in this method. Application of GA-Fl to Arabidopsis
wild-type seedlings, through either solid or liquid medium, leads
to uptake of the fluorescent GA into the plant and is followed by
movement through the root, resulting in specific accumulation in
the root elongating endodermal cells. The accumulation pattern of
GA-Fl can be visualized by confocal microscopy and target cells
can be identified by their location in the root. Co-staining of the
root with propidium iodide, to highlight cell walls, does not inter-
fere with GA-Fl signal detection and assist in observing root mor-
phology. The method described herein can be easily adjusted to
visualize GA-Fl distribution patterns under varying developmental
and environmental conditions and in plant species beyond
Arabidopsis. A short summary of the method’s advantages and
disadvantages is presented in Table 1.
Table 1
Pros and cons for live imaging of fluorescently labeled hormones in the plant
Pros Cons
Easily transferable to plant species beyond Exogenous application of the molecule, thus differ
Arabidopsis from endogenous sites of biosynthesis
Directly monitoring the labeled hormone The hormone is bioactive but structurally modified,
(opposed to monitoring genetic markers, thus might differ with respect to transport
proteins involved in perception and response) parameters (charge, crossing membranes, etc.)
No requirement for transgenes
Live imaging
High spatial resolution (sub-cellular)
Fluorescent Gibberellins 93
2 Materials
Carry out all procedures at room temperature, unless otherwise

specified.
2.1 GA-Fl Stock 1. Prior to preparation, cover all 0.6/1.5 mL Eppendorf tubes
Solution (10 mM) with tin foil (see Note 1).
2. Dissolve 1.5 mg of solid GA3-Fl in 180 μL dimethyl sulfoxide
(DMSO) to a final concentration of 10 mM, OR 1.5 mg of
solid GA4-Fl in 183 μL DMSO to a final concentration of
10 mM (for calculation, see Note 2).
3. Vortex to full dissolution. Store at −20 °C (see Note 3).
2.2 GA-Fl Liquid 1. Prepare fresh GA-Fl liquid work solution for each
Work Solution (5 μM) experiment.
2. Wrap 15 mL tube with tin foil.
3. Add ddH2O, 3 mL and GA-Fl stock solution, 1.5 μL.
4. Vortex well.
2.3 GA-Fl MS Agar 1. Prepare plant growth medium: 0.5 g/L MES (2-(N-morpholino)
Plates (5 μM) ethanesulfonic acid), 2.3 g/L Murashige and Skoog salt (MS),
10 g/L sucrose, 5 g/L plant agar (final pH 5.7).
2. Prepare the medium in a bottle or a flask (see Note 4). Stir the
mixture using a magnetic stirrer. Once the mixture is homo-
genic, measure its pH using a pH meter. Adjust to pH of 5.7
by adding the necessary amount of KOH. Pour the first mix-
ture into the new bottle, along with the plant agar.
3. Autoclave the plant growth medium at a temperature of
120 °C for 30 min. Subsequently, mix well by stirring and
leave to cool off at room temperature. Once the temperature
has dropped to approximately 50 °C (see Note 5), add the
necessary amount of GA-Fl to reach 5 μM. For example, 50 μL
of GA-Fl (10 mM stock solution) into 100 mL of MS agar
media. Mix well. Once the mixture is homogenic, pour it into
12 × 12 × 1.5 cm plates.
4. Keep the plates open in the hood for the medium to dry off
(~20 min, see Note 6). Store plates upside-down at 4 °C in the
dark for up to 1 month.
3 Methods
3.1 Seed Important! The sterilization process must be performed in a fume

Sterilization hood to prevent inhalation of volatile materials.
1. Prepare the amount of Arabidopsis seeds (see Note 7) to be

sterilized in a 1.5 mL Eppendorf tube (see Note 8).
2. Add sodium hypochlorite (NaClO), 33 mL, into a 100 mL
Erlenmeyer flask.
3. Inside a glass desiccator—place the open Eppendorf tube with
the seeds in a stand next to the Erlenmeyer flask containing the
NaClO (see Note 9).
4. Make sure that the fume hood is properly working.
5. Add hydrochloric acid (HCl, concentrated 32 %), 1 mL, into
the Erlenmeyer flask and immediately place the lid. CAUTION:
Addition of hydrochloric acid to hypochlorite solution leads to
immediate generation of highly toxic chlorine gas (Cl2).
6. Set a timer for 2 h (see Note 10).
7. Subsequently, remove the Eppendorf tube from the desiccator
(in the fume hood). Make sure that it is properly closed, to
prevent contamination.
8. Open the tubes in biological hood for 10 min to allow air
exchange.
9. Seeds can be stored for several days (see Note 11).
3.2 Sowing 1. Sow seeds on a plate containing MS growth medium (see

Subheading 2.3 for plated preparation and Subheading 3.1 for
seed sterilization).
2. Cover the plate with tin foil and store at 4 °C for 48 h for
stratification.
3. Place the plates horizontally in an Arabidopsis growth chamber
with the tin foil off.
4. Allow seedlings to grow for 5 days (depending on the develop-
mental system).
3.3 Seedling There are two major administration approaches for applying GA-Fl
Preparation to roots:
1. Liquid (see Subheading 2.2)—Immerse the seedlings in the
GA-Fl liquid work solution for 0.5–3 h, depending on the
developmental question (see Note 12).
2. Agar plates (see Subheading 2.3)—Replace the seedlings on
GA-Fl agar plates for 2–3 h depending on the developmental
question. Make sure that roots are not damaged while replaced
(see Note 13).
3. Prepare a propidium iodide (PI) solution, 10 μg/mL in
ddH2O, in a separate plate.
4. Following GA-Fl application, immerse the seedlings in the PI
solution for 2–4 min. Rinse with ddH2O.
3.4 Specimen To prevent dehydration of the seedlings during imaging—first

Preparation place a drop of ddH2O on the microscope slide. Use a pipette; the
amount of water would vary according to the number of seedlings
to be mounted (typically 80–200 μL for 1–3 seedlings). Place 1–5
seedlings on the wet slide. Gently place a cover slip (cover glass)
over the mounted seedlings. Important: It is crucial not to apply
any pressure to the cover slip, as the seedlings can be easily crushed.
Mount the slide onto the microscope.
3.5 Imaging Using confocal microscopy, illuminate the root at a wavelength

of 488 nm, according to the absorption spectrum of fluorescein.
Fluorescence is expected to be seen mostly in the endodermis
layer of the root’s elongation zone ( see Notes 14 and 15 )
( see Fig. 1).
Fig. 1 Imaging fluorescently labeled GA3 in Arabidopsis thaliana root. (a) Molecular structure of GA3-Fl. (b and c)
Confocal images of fluorescently labeled GA3 (GA3-Fl) in the elongating endodermal cells of roots. Meristematic
zone (b), elongating zone (c). Root treated with GA3-Fl (5 μM, 3 h). GA3-Fl (green), propidium iodide (red)
4 Notes
1. It is important that GA-Fl solutions remain unexposed to light.

Exposure to light will lead to gradual photobleaching of the
fluorescein. Thus, tin foil should be used to cover all tubes and
plates that contain the GA-Fl solutions, throughout the prepa-
ration process and during the experiment itself.
m
2. According to the formula: V = Mw
C
where m (solid GA3-Fl or GA4-Fl) = 1.5 mg
Mw (GA3-Fl) = 834 g/mol OR Mw (GA4-Fl) = 820 g/mol
In order to reach a concentration of 10 mM, a volume of
180 μL DMSO should be added to GA3-Fl and 183 μL DMSO
to GA4-Fl.
3. GA-Fl DMSO stock solutions can be kept at −20 °C for at least
a year.
4. The precise volume of the flask/bottle is not crucial to the
procedure, but at least 2/3 of the volume should remain
empty, to prevent spillage while autoclaving.
5. This could be done by carefully touching the bottle to assess its
temperature, or more accurately, by using a touch-thermometer.
6. Plates could be kept open in the hood for a longer period of
time, ~45 min. This will allow excess water to evaporate.
7. The method is not limited to WT Arabidopsis Col-0 seedlings
and may be applied to mutants, transgenes, ecotypes, and addi-
tional species. However, the protocol listed here is optimized
for Arabidopsis Col-0.
8. This could be done using a folded paper of a sort, while pouring
the seeds onto it and in turn pouring the seeds into the
Eppendorf tube.
9. If needed, it is possible for the flask to be tilted, and for the
tube stand to sit on top of it, to allow enough room.
10. If there are more than 100 μL seeds in the tube, divide the
seeds into several tubes and extend the sterilization to 3 h. Do
not exceed 3 h as it might damage the seeds.
11. Sterilized seeds may be stored for several days at room tem-
perature. After about 2 weeks, seeds tend to lose their vitality.
12. GA-Fl accumulation is temperature dependent. Immersing the
seedlings in 28 °C dramatically increases GA-Fl uptake. This
should be taken into account when deciding on the incubation
times.
13. The seedling should be lifted by gently pulling the cotyledon
upwards and replacing on top of a new agar plate (not more
than 2 s in the air to prevent desiccation responses). Gently drag

the seedling upwards to straighten the root on the new plate.
14. When imaging, use the BF (bright-field) channel in order to
examine the root. If dark blots and no definite outlines of cells
(in the epidermis it is more prominent) are observed, the seed-
ling was probably harmed while placing the cover glass on the
slide. Similarly, patches of PI (red channel) and no definite out-
lines of cells points on root cells death.
15. Fluorescence may also be detectable around the outlines and
on the surface of the root. This happens due to sticking of
GA-Fl to the root’s surface and can be reduced by increasing
washing time or number of washing repetition (Subheading 3.3).
References
1. Achard P, Genschik P (2009) Releasing the 11. Lavis LD, Raines RT (2014) Bright building
brakes of plant growth: how GAs shutdown blocks for chemical biology. ACS Chem Biol
DELLA proteins. J Exp Bot 60:1085–1092 9:855–866
2. Daviere JM, Achard P (2013) Gibberellin signal- 12. Shani E, Weinstain R, Zhang Y, Castillejo C,
ing in plants. Development 140:1147–1151 Kaiserli E, Chory J, Tsien RY, Estelle M (2013)
3. Yamaguchi S (2008) Gibberellin metabolism and Gibberellins accumulate in the elongating
its regulation. Annu Rev Plant Biol 59:225–251 endodermal cells of Arabidopsis root. Proc
4. Sun TP (2010) Gibberellin-GID1-DELLA: a Natl Acad Sci U S A 110:4834–4839
pivotal regulatory module for plant growth and 13. Hayashi K, Nakamura S, Fukunaga S,
development. Plant Physiol 154:567–570 Nishimura T, Jenness MK, Murphy AS, Motose
5. Coling D, Kachar B (2001) Theory and appli- H, Nozaki H, Furutani M, Aoyama T (2014)
cation of fluorescence microscopy. Curr Protoc Auxin transport sites are visualized in planta
Neurosci Chapter 2, Unit 2 1 using fluorescent auxin analogs. Proc Natl
Acad Sci U S A 111:11557–11562
6. Combs CA (2010) Fluorescence microscopy: a
concise guide to current imaging methods. 14. Irani NG, Di Rubbo S, Mylle E, Van den
Curr Protoc Neurosci Chapter 2, Unit2 1 Begin J, Schneider-Pizon J, Hnilikova J, Sisa
M, Buyst D, Vilarrasa-Blasi J, Szatmari AM,
7. Shaner NC, Steinbach PA, Tsien RY (2005) A Van Damme D, Mishev K, Codreanu MC,
guide to choosing fluorescent proteins. Nat Kohout L, Strnad M, Cano-Delgado AI,
Methods 2:905–909 Friml J, Madder A, Russinova E (2012)
8. Sahoo H (2012) Fluorescent labeling tech- Fluorescent castasterone reveals BRI1 signal-
niques in biomolecules: a flashback. RSC Adv ing from the plasma membrane. Nat Chem
2:7017–7029 Biol 8:583–589
9. Jung D, Min K, Jung J, Jang W, Kwon Y 15. Tsuchiya Y, Yoshimura M, Sato Y, Kuwata K,
(2013) Chemical biology-based approaches on Toh S, Holbrook-Smith D, Zhang H,
fluorescent labeling of proteins in live cells. McCourt P, Itami K, Kinoshita T, Hagihara S
Mol Biosyst 9:862–872 (2015) Probing strigolactone receptors in
10. Lavis LD, Raines RT (2008) Bright ideas for Striga hermonthica with fluorescence. Science
chemical biology. ACS Chem Biol 3:142–155 349:864–868
Chapter 11
In Silico Methods for Cell Annotation, Quantification

of Gene Expression, and Cell Geometry at Single-Cell
Resolution Using 3DCellAtlas
Petra Stamm, Soeren Strauss, Thomas D. Montenegro-Johnson,
Richard Smith, and George W. Bassel
Abstract
A comprehensive understanding of plant growth and development requires the integration of the spatial
and temporal dynamics of gene regulatory networks with changes in cellular geometry during 3D organ
growth. 3DCellAtlas is an integrative computational pipeline that semi-automatically identifies cell type
and position within radially symmetric plant organs, and simultaneously quantifies 3D cell anisotropy and
reporter abundance at single-cell resolution. It is a powerful tool that generates digital single-cell cellular
atlases of plant organs and enables 3D cell geometry and reporter abundance (gene/protein/biosensor)
from multiple samples to be integrated at single-cell resolution across whole organs. Here we describe how
to use 3DCellAtlas to process and analyze radially symmetric organs, and to identify cell types and extract
geometric cell data within these 3D cellular datasets. We detail how to use two statistical tools in 3DCellAtlas
to compare cellular geometries, and to analyze reporter abundance at single-cell resolution.
Key words 3DCellAtlas, MorphoGraphX, 3D imaging, 3D image analysis, Cell type identification,
3D anisotropy, Digital single-cell analysis
1 Introduction
In order to understand plant growth and development, it is neces-

sary to analyze events across scales ranging from the whole organ
level to cellular level, and down to molecular processes. Linking
these events in the context of dynamic growing organs requires
measuring changes in geometry and gene/protein abundance
simultaneously. Generating 3D cellular resolution data sets in com-
bination with reporters enables these spatial and temporal associa-
tions to be established. It is these 3D data sets that will allow a
comprehensive analysis of plant growth and development [1].
Recent years have seen advances in such 3D imaging techniques,
and the number of 3D image data sets is steadily increasing [2–5].
99
100 Petra Stamm et al.
Multiple computational approaches have as well been developed to

enable the analysis of such data sets [3, 6–8].
Here we describe a method for the analysis of 3D images of
plant organs using 3DCellAtlas, an integrative computational pipe-
line that has been implemented as an add-on in MorphoGraphX
(www.MorphoGraphX.org; [9]), an open-source software devel-
oped for the analysis of 3D images that has been used in various
studies of plant development (for example [6, 8–10]). The
3DCellAtlas extension [11] provides additional functionality to
this software in the analysis of plant growth and development on a
whole-organ level. It allows for the semi-automatic identification
of cell types in radially symmetric plant organs based on geometry
of the cells without the need for transgenic cell type markers or
reference atlases, as well as the quantification of 3D cellular anisot-
ropy and reporter abundance at single-cell resolution across whole-
plant organs. It enables researchers to pool data from several
samples for analysis, providing great statistical power. Thus, using
3DCellAtlas, cell shapes and sizes can be extracted for individual
cells in a plant sample; these geometric data can be pooled for a set
of samples and compared to another set. 3DCellAtlas is therefore a
powerful tool to quantitatively analyze plant development by
allowing the integration of 3D organ growth with diverse regula-
tory network dynamics.
Starting with the acquisition of confocal image stacks of stained
plant organs or tissues, we describe how to use MorphoGraphX to
process and analyze these 3D images, and how to use the 3DCellAtlas
add-on to identify cell types within the plant organ. We further
describe the use of two statistical tools in MorphoGraphX, (a) to
compare cellular geometries of two sets of samples, and (b) to ana-
lyze reporter abundance within the organs at cellular resolution.
2 Materials
2.1 Samples 1. Plant tissue in which cell walls have been stained with propid-
ium iodide and cleared with chloral hydrate [5]. In principle
any 3D image set of sufficient quality for 3D segmentation
could be used.
2. In these examples, reporter abundance from the GUS enzyme
assays is detected using reflectance on a second channel. Again,
any reporter can be used in the context of 3D segmented cells.
2.2 Image 1. Inverted confocal microscope capable of acquiring high-

Acquisition resolution multi-channel z-stacks, ideally 16-bit images. All
images presented here as examples were acquired using a Zeiss
LSM 710.
2. As plant tissue is best imaged in liquid (in chloral hydrate clear-
ing solution), dishes with a cover glass bottom are best suited
3D Image Analysis with 3DCellAtlas 101
for imaging (for example Lab-Tek® Chambered Borosilicate

Coverglass System, or Greiner Bio-One Cellview™ cell culture
dish with glass bottom).
2.3 Image Analysis 1. Fiji (http://fiji.sc/Fiji) or equivalent software to convert propri-

Software etary microscope format files into single 16-bit multi-page tiffs.
2. MorphoGraphX, which can be downloaded from the develop-
ers’ website: http://www.MorphoGraphX.org.
● Recommended to be run on Linux system—Linux Mint
17 (which is based on Ubuntu 14.04), but a version run-
ning on Microsoft Windows is available.
● Cuda-enabled nVidia graphics card with as much RAM as
possible to handle large image stacks.
● A workstation with 32–64 Gb of RAM is recommended.
3 Methods
3.1 Image Acquire confocal z-stacks of the organ of interest. Particular atten-
Acquisition tion should be paid to achieve the highest possible signal-to-noise
ratio of cell wall to cell interior for ideal segmentation, and using a
thin enough slice interval to attain the best possible resolution
along the z-axis (see Notes 1–3). For Arabidopsis embryo axes, a
suitable image quality is achieved by using a 25× oil objective with
a 0.6 zoom factor and an image resolution of 2048 by approxi-
mately 600 pixels.
3.2 Importing To open a 3D stack in MorphoGraphX, the image needs to be in

Images *.tiff format (see Note 4). If multiple channels have been acquired,
into MorphoGraphX each channel has to be converted into a separate file via the “split
channels” function during the conversion. The *.tiff file can
either be dragged onto the open MorphoGraphX window, or it
can be accessed via Stack > Stack 1 > Main > Open (Fig. 1a).
Functions to process images are accessed via the three tabs on the
right of the main window—“Main,” “View,” and “Process”
(Fig. 1b–d).
In the “View” tab, clipping planes 1, 2, and 3 are tools to slice
into the 3D image. Clip 1 is initially oriented in the y–z-plane, clip
2 in the x–z-plane, and clip 3 in the x–y-plane. Each clipping plane
can be visualized on the image by ticking the box for “Grid,” and
enabled by ticking “Enable.” To control the movement of the clip-
ping planes, make sure that the correct clip is selected in the main
tab under “Control-Key Interaction” (Fig. 1b). Holding down the
Ctrl key and clicking the left mouse button will rotate the plane,
while right-clicking moves the plane, keeping the same orientation.
Move the slider below each clip to adjust the thickness of the clip-
ping plane.
Fig. 1 MorphoGraphX basic functions. (a) The mouse pointer highlights the path to open an image in stack 1.
(b) The main tab with options to choose control key interactions, and to switch between stack 1 and stack 2.
(c) The view tab which allows control of brightness and contrast and interaction with clipping planes. The ori-
entation of each clipping plane can be visualized by ticking “Grid,” and enabled by ticking “Enabled”. (d) The
Process tab which contains all functions to process and analyze stacks and meshes
3.3 3D Segmentation Prior to the 3D segmentation of cells, the image needs to be

of Cells blurred to reduce noise. This is done using “ITK Smoothing
Recursive Gaussian Blur,” in the “Process” tab, under
“Stack” > “ITK” > “Filters” (Fig. 2a). The parameter (Fig. 2b) is
set to a radius of 0.5 μm (see Note 5).
The result of each modification to the main stack will auto-
matically be transferred into the work stack. Stacks can be saved
from either position, and can be copied or transferred from one to
another via tools in the “Process” tab, under “Stack” > “Multi-
stack” (Fig. 2c).
Fig. 2 Close-up of several functions in the Process tab. (a) To blur an image as preparation for segmenting it,
the “ITK Smoothing Recursive Gaussian Blur” is used. (b) The adjustable parameter for this algorithm is the
radius of the blurring, in μm. (c) Tools for copying and swapping main and work stack can be found in the
drop-down menu “Multi-stack”. (d) Tools for segmenting a stack using ITK segmentation algorithms. (e)
Adjustable parameters for the “ITK Watershed Auto Seeded” segmentation process are shown
The blurred stack is then segmented, using the “ITK Watershed

Auto Seeded” tool, in the “Process” tab, under “Stacks”—“ITK”—
“Segmentation” (Fig. 2d). From the adjustable parameters (Fig. 2e)
typically only the level is changed from its default to a value between
250 and 700. This level refers to the threshold signal intensity for
the detection of “edges” in the segmentation process. This value
will have to be adjusted to suit each image stack (depending on
signal intensity, and how the image was acquired on the confocal,
see Notes 6–9).
3.4 Editing of 3D The segmented stack now needs to be visually inspected, and cells
Segmented Cells that have mistakenly been assigned more than one label (“overseg-
mented”) can be corrected by editing (note that undersegmented
cells cannot be edited, see Note 6). The tools for this can be found
in the tool bar of the MorphoGraphX window (Fig. 3); all of these
are used by holding down the Alt key and left-clicking.
The region surrounding the actual image is usually segmented
as one big “box,” and has to be deleted with the scissors tool.
Fig. 3 Tools for editing stacks in MorphoGraphX. The highlighted dropper tool is
used to pick a label within the segmented stack by holding down the “Alt” key
and clicking on a label. The cell number will appear in the bottom left corner of
the MorphoGraphX window. The scissors tool can be used to delete labels, while
the pixel edit tool will delete pixels within a stack regardless of labels. The paint
bucket tool will replace any label with the currently selected one
The dropper icon is used to select a label from the work stack;
the number of the label will appear on the bottom left corner of
the MorphoGraphX window.
To replace another label with the selected one, the paint bucket
tool is selected. Clicking on the label to be changed will replace the
original label with the active selected one.
To edit cells on the inside of the stack, any of the three clipping
planes can be used to move through the image. Progressive changes
during editing of the segmented image stack need to be saved
frequently, as there is no “undo” function in MorphoGraphX.
3.5 Creating the 3D Once the segmented image is fully edited, a cell mesh describing
Cell Mesh the surface of each individual cell is created. The tool for this pro-
cess is called “Marching Cubes 3D,” and can be found in the
“Process” tab, under “Mesh” > “Creation” (Fig. 4a).
Under the parameters (Fig. 4b), “Cube size” determines the
distance between each point in the mesh; this will have to be
adjusted to suit the size of the cells. A cube size of 2 μm is typically
used (see Note 10).
“Min Voxels” determines the minimum size in voxels of a seg-
mented cell to be assigned a surface.
“Smooth passes” is the number of times the mesh is smoothed
during its creation. Repeated smoothing will shrink segmented cell
meshes. This procedure must be used sparingly, if at all (see Note 11).
“Label” refers to the cell ID the mesh is being created on. If a
number is chosen, the marching cubes will only run on this par-
ticular label; if it is “0” it runs on the whole sample.
3.6 Creating A second mesh, which describes the surface of the whole organ to
the Surface Mesh be analyzed, needs to be created; this is to capture the global organ
shape rather than that of individual cells. For this, the original
image is blurred with the “ITK Smoothing Recursive Gaussian
Blur,” at a radius of 5 μm (see Fig. 2a, b), and the surface mesh is
created using the “Marching Cubes Surface” tool, in the “Process”
tab, under “Mesh” > “Creation” (Fig. 4a). The cube size is typi-
cally set to 5 μm for this mesh. The threshold refers to the signal
intensity (Fig. 4c). This will have to be adjusted to suit each image
to get a smooth description of the surface.
3.7 Trimming The end of the surface mesh, where the sample was truncated during
the Open End imaging, needs to be cut off for 3DCellAtlas. Surface meshes of
of the Surface Mesh hypocotyls have two “open ends,” while embryo axes and roots
where the tip is included need to be cut at one end only. To trim the
ends of a surface mesh, use any of the clipping planes. Move the clip-
ping plane to include the end to be cut off and select it by running
“Select Clip Region” in the “Process” tab, under “Mesh” > “Selection.”
The selected region will be highlighted in red (Fig. 5). Pressing the
Delete button on the keyboard will delete the selected region. The
final surface mesh with the end(s) cut off is then saved.
3.8 Creating To create a Bezier curve that describes the geometric center of the
the Bezier Curve radial sample, in the “View” tab, under Cutting Surface, tick
“Draw,” “Grid,” and “Bezier,” and then click “Reset” (Fig. 6a).
This will generate a 2D Bezier plane in the x–y plane with 5 × 5
Fig. 4 Tools and parameters for creating meshes. (a) Algorithms used to create meshes. A mesh describing the
surface of individual cells of a segmented stack is created using the “Marching Cubes 3D” tool, while a mesh
describing the overall surface of the whole organ is generated with the “Marching Cubes Surface” algorithm. (b)
Adjustable parameters for the “Marching Cubes 3D” tool. (c) Adjustable parameters to create a surface mesh with
the “Marching Cubes Surface” tool
handles. To collapse these into a single line, go to “Mesh” > “3D

Cell Atlas” > “Tools” > “Collapse Bezier Points” in the “Process”
tab. This will transform the Bezier plane into a single line with five
handle points. These five handle control points need to be adjusted
to fit the Bezier line to the center. The “Select points in mesh”
(Fig. 6b) tool is used to select handles, by holding down the Alt key
and clicking on, or dragging over, a control handle (Fig. 7). Selected
handles will turn red. These can be moved by simultaneously hold-
ing down the Alt and Shift key while clicking and dragging them to
the required position (see Note 12). The positions of the Bezier
points are stored in MorphoGraphX View (*.mgxv) files, and need
to be saved from the top main menu via “File” > “Save as.”
3.9 Preparing The 3DCellAtlas tool will create a coordinate system through the
the Sample plant organ using a Bezier line through its centrer and a surface
for 3DCellAtlas mesh on its outside. To run 3DCellAtlas, open the Bezier view file
(see Subheading 3.8), load the segmented cell mesh into “Mesh 1”
(see Subheading 3.5), and the organ surface mesh into “Mesh 2”
Fig. 5 Selecting and deleting parts of meshes. Screenshot of the MorphoGraphX window showing a surface
mesh of an Arabidopsis hypocotyl stack. The grid of clip 1 is visible, which has been used to select one end of
the surface mesh. The mouse pointer highlights the function to select the part of the mesh within the clipping
plane. Once selected, the vertices of the mesh will be highlighted in red. Highlighted regions can then be
deleted by pressing the “Delete” key on the keyboard
Fig. 6 Creation of and interaction with a Bezier curve. (a) To create a Bezier plane, “Draw,” “Grid,” and “Bezier” need
to be ticked in the View tab before pressing “Reset.” (b) The resulting Bezier plane is collapsed into a Bezier curve
with five handle points using the “Collapse Bezier Points” tool in the Process tab, under Mesh > Cell Atlas 3D > Tools.
(c) Tools for editing meshes and Beziers. The tool for selecting and moving Bezier handles is circled
Fig. 7 Adjusting the Bezier curve to align with the axis of the imaged plant organ. MorphoGraphX window with
a Bezier curve through the center of an Arabidopsis root. The mouse pointer has been dragged over one of the
handles while holding down the Alt key to select it
(see Subheading 3.7). This can be saved again as the current *.mgxv
file. In this way opening the *.mgxv file will open the cell mesh,
organ surface mesh, and the Bezier curve, all together in the cor-
rect locations.
Highlight a cortical cell at one end of the sample in the seg-
mented mesh using the “Select Connected Area” tool (Fig. 8a,
circled) with a left click while holding down the Alt key. Note that
the “Stack 1” tab needs to be selected in the “Main tab” (thus,
Mesh 1 will be active, instead of Mesh 2). This cortical cell will
identify cell position “1” for the cortical referencing system.
If the organ has two different cell topologies, for example roots
where the root cap splits, or embryo axes which have both radicle
and hypocotyl cellular arrangements, an additional cell on the out-
side of the sample needs to be highlighted (Fig. 8b). This second
outside cell should be selected at the position where the cellular
topology changes. This is not required for organs with a single-cell
topology, for example hypocotyls (Fig. 8c).
3.10 Step 1 In the “Process” tab, under “Mesh” > “Cell Atlas 3D,” “A—Analyze
of 3DCellAtlas: Cells 3D” will calculate all geometric data of the mesh, and deter-
Analyze Cells 3D mine the positions of all cells relative to the Bezier (centre) and the
surface (Fig. 9a). Hereby, the coordinates and lengths along the three
principal directions are calculated for each cell. The three directions
are longitudinal (along the Bezier line), radial (relative location
Fig. 8 Setting up samples to be analyzed using 3DCellAtlas. (a) Tools to interact with meshes in MorphoGraphX.
The highlighted tool is used to select cells within a mesh. (b and c) Screenshots of MorphoGraphX windows
with the setup required for running the 3D Cell Atlas tools: the segmented mesh of the organ in stack 1, the
surface mesh in stack 2, and the Bezier curve describing the center of the axis. (b) Arabidopsis embryo with
two cells highlighted, the first cell in a cortical cell file, and a cell at the position of the topological switch. (c)
Arabidopsis seedling hypocotyl with the first cortical cell highlighted
between the Bezier line and the surface mesh), and circumferential
(angular position within the radially symmetric cross section).
In the parameters (Fig. 9b), a minimum volume to be identi-
fied as a cell can be set. Thus, cells with a smaller volume than this
set value will be ignored. Also, other “malformed” cells will be
ignored and marked as “bad”; those can be visualized using the
process “Select Bad Cells.”
Fig. 9 Cell Atlas 3D tools and algorithms. (a) Close-up of the Process tab with the drop-down menus for the
Cell Atlas 3D tools. (b–g) Adjustable parameter for each function. (b) Parameters for “A—Analyze Cells 3D.”
The Volume Threshold can be set to exclude cells below a certain size from the analysis. (c) Parameters for
“B—Assign Cell Types.” “Has multiple segments” refers to segments within the organ to be analyzed that
have different topologies. (d) Parameters for “C—Assign Columella.” The values for Root Cap and Columella
label can be changed, but must be numbers that have not been used during the assigning of cell types. (e)
Parameters for “D—Assign Cortical Cells.” The value has to be the number assigned to cortical cells. (f)
Parameters for “E—Topological Check.” “Work on Selection” determines which cells will be analyzed; set to
“No” to examine all cells in the sample. A “Threshold Volume” defines the minimum size of cells to be exam-
ined, while “Threshold Wall Area” determines the minimum shared cell wall area between neighboring cells to
be considered as a connection. Labels for root cap, air spaces, and vasculature need to match the numbers
assigned to these cell types. The “Error Limit” is the minimum number of wrong connections (impossible
edges) between cell types for a given cell to be identified as mislabeled. (g) Parameters for “F—Examine
Vasculature.” Here, labels of vasculature, air spaces, and endodermis need to be matched up with the num-
bers assigned for these cell types
3.11 Step 2 Next, run “B—Assign Cell Types” (Fig. 9a). For organs with a
of 3DCellAtlas: Assign single-cell topology, for example hypocotyls, the parameter “Has
Cell Types multiple segments” needs to be “No.” If the organ has two differ-
ent topologies, for example embryo axes and roots, the parameter
needs to be “Yes” (Fig. 9c). Running the process will prompt a
GUI to open (Fig. 10). The drop-down next to “Part of organ,”
with the options “segment 1” and “segment 2,” refers to the parts
of the organ with different topologies, divided at the second
selected cell prior to analysis. This option will be disabled if “B—
Assign Cell Types” was run with the parameter “No” (Fig. 9c). By
changing the selection of the segment (1 or 2), the heatmap will
show all cells located within each respective region. This facilitates
the selection of cell clusters across organs with multiple topolo-
gies. Without this feature, cell clusters of diverse cell types would
overlap.
Fig. 10 Screenshots of the GUI opened upon running “Assign Cell Types.” (a) If an organ with two topologies is
analyzed, these two regions will be separately annotated, and appear as segment 1 and segment 2 in a drop-
down menu next to “Part of organ.” (b) This option is disabled for organs with a single topology. Cell data are
plotted in heatmaps, displaying the radial distance of cells from the central Bezier curve on the x-axis, and a
chosen geometric property of cells along the y-axis. The y-dimension “Y-Dim” can be changed on the drop-
down menu to display either of the three principal cell lengths, to find the best option for ideal spread of cell
clusters. Tick the box next to “Show Cell” to display each cell on the heatmap. “Sigma” is the variance param-
eter for a 2D Gaussian representing each cell’s parameters. A larger sigma will lead to a smoother heatmap
3.12 Step 3 The 2D heatmap GUI generates cell clusters by plotting the radial
of 3DCellAtlas: distance of cells from the central Bezier curve on the x-axis, and a
Selecting Cell Clusters chosen geometric property of cells along the y-axis. The heatmap
(red—high value, blue—low value) is created by using the radial
distance between the Bezier and the surface mesh in the x-dimension
(“X-Dim”), and a user-defined cell length in the y-dimension
“Y-Dim” and summing Gaussians around the coordinates of each
cell. “Y-Dim” can be changed on the drop-down menu to display
the cell length along the three different directions (circumferential,
radial, or longitudinal, see Subheading 3.10). The best option for
ideal spread of cell clusters can be chosen to define cell types accu-
rately. Tick the box next to “Show Cell” to display the data point
of each cell within the heatmap.
The value of “Sigma” is the variance parameter for a 2D
Gaussian representing each cell’s parameters. A larger sigma will
create larger overlap of the “heat” of single cells, which leads to a
smoother heatmap.
In this 2D heatmap GUI, left mouse clicks are used to place
seeds upon cell clusters, and to assign these cell cluster (represent-
ing distinct cell types) unique numbers. It is important to keep
track of labels assigned to cell types, and to make sure that these
cell identity labels match across different segments in samples with
two different topologies. Small white crosses refer to the location
of local maxima in the heatmap; placed seeds are assigned to their
nearest local maximum.
Clicking onto a cluster will place a seed onto it, automatically
assigning this the number “1.” To change the cell identity label a
seed gives to a cluster, hover the mouse over the seed, and press a
number key on the keyboard. A maximum of ten different cell
identities can be assigned (numbers 0 through 9). To move placed
seeds, drag and drop them to the required position. To remove
seeds from the GUI, click and drag them out of the visible area.
We recommend a cell identity numbering system where outer-
most cells are given number 1 and numbers increase towards the
middle with each progressive cell layer:
1. Epidermis
2. Outer cortex
3. Inner cortex
4. Endodermis
5. Vasculature
6. Air spaces
7. Lateral root cap
Additional layers including the columella are assigned higher
numbers (see Subheading 3.13).
The “Auto Clustering” button automatically identifies the

selected number of clusters by placing seeds on the highest peaks
present on the heatmap.
By pressing “Preselect cluster” after having placed a seed onto
one cell cluster, all cells assigned to this cluster will be given this
particular label, and will be removed from the heatmap. This will
also re-scale the heatmap with the remaining cells. Preselecting
clusters can be undone using the button “Reset All.”
Cell type labels for all cells are stored as parent labels, and fol-
lowing GUI selection, cells are false colored based on their parent
label. If cells were mis-annotated, it is possible to return to 2D
Heatmap GUI in “B—Assign Cell Types” and iteratively adjust
seed locations to improve accuracy. Typical distribution of cell
clusters and placement of seeds for an Arabidopsis hypocotyl, root,
and embryo axis are shown in Fig. 11.
3.13 Step 4 For roots and embryos, cells in front of the highlighted first corti-
of 3DCellAtlas: Assign cal cell are not included in the cell type analysis described above.
Columella To assign columella and lateral root cap at this end of the sample,
run “C—Assign Columella.” This will identify cells as root cap
when their relative wall area touching other cells is below a certain
Fig. 11 Heatmaps and locations where annotated seeds are placed when identifying cell clusters in 3DCellAtlas.
(a–c) Arabidopsis hypocotyl. (d–i) Arabidopsis root showing segment 1 in (d–f) and segment 2 in (g–i). (j–o)
Arabidopsis embryo showing segment 1 in (j–l) and segment 2 in (m–o). Yellow labels indicate true cell clus-
ters and red labels those for segmented air spaces
threshold. In the parameters (Fig. 9d), the cell identity value needs
to be changed to the number assigned to these cell types, and not
a number used previously for other cell types.
3.14 Step 5 Next, run “D—Assign Cortical Cells.” With the first cortical cell
of 3DCellAtlas: Assign selected, this will assign the cortical cell referencing system of cell
Cortical Cells positions to all cells in the mesh. Make sure that the value in the
parameters (Fig. 9e) contains the correct label for cortical cells.
3.15 Step 6 Mis-annotated cells can be detected and highlighted by running

of 3DCellAtlas: “E—Topological Check.” Hereby a neighborhood graph is used
Topological Check to determine unlikely cell neighborhoods, which will be labeled as
“wrong connections.” Note that this process requires a sufficient
accuracy of the input data to work correctly, as it depends on a
majority of correct neighborhoods.
In the parameters (Fig. 9f), “Work on Selection” determines which
cells will be analyzed; set to “No” to examine all cells in the sample.
“Threshold Volume” refers to the minimum size of cells to be
examined, while “Threshold Wall Area” defines the minimum
shared cell wall area between neighboring cells to be considered as
a connection.
Make sure that the labels for root cap, air space, and vascula-
ture match the earlier selection.
The “Error Limit” is the minimum number of wrong connec-
tions (impossible edges) between cell types for a given cell to be
identified as mislabeled.
3.16 Step 7 Following the topological check, run “F—Examine Vasculature.”

of 3DCellAtlas: Make sure that the correct parent labels are assigned to vascula-
Examine Vasculature ture, air space, and endodermis (Fig. 9g). This tool will automati-
cally re-annotate vasculature cells that have been mislabeled (see
Note 13).
3.17 Step 8 Once all parent labels have been correctly identified, save all cell
of 3DCellAtlas: Saving data. This is done using “Save Cell Data,” under “Cell Atlas
and Loading Cell Data 3D” > “Tools” (Fig. 9a). This stores geometric data for all cells,
which cells have been highlighted, the locations of each seed placed
in the 2D Heatmap GUI, cell position according to the referenc-
ing system (in the column “associated cortical cell”), and parent
labels associated with each original unique cell ID in the mesh.
Using “Load Cell Data,” all this information can be reloaded onto
the corresponding mesh.
3.18 Step 9 The “Display Cell Data” tool can be used to visualize and validate
of 3DCellAtlas: measured cell data (Fig. 12). These will be displayed in the form of
Displaying Cell Data a false-colored heatmap on the analyzed organ. The drop-down
menu allows the selection of the parameter to visualize.
3.19 Step 10 To analyze differences in 3D cell shape, data saved as described in

of 3DCellAtlas: Subheading 3.18 in csv files for multiple samples and/or time
Statistical Analysis points are placed into separate directories for “control” and
of 3D Cell Anisotropy “treatment” data. Additionally, create a directory in which the out-
put of these statistical analysis is to be placed (see Note 14).
In the “Process” tab, under “Mesh” > “Cell Atlas 3D” > “Statistics,”
run “Cell growth analysis 3D” (Fig. 9a) (see Note 15).
If any of the values for “Control folder,” “Treatment folder,”
and “Output folder” are left blank (Fig. 13), the program will
prompt the user to guide it to each missing directory.
Under “Output Type,” either “Control” or “Treatment” can
be chosen. This refers to the samples that will be used to generate
heatmaps with the result of the analysis.
“Sliding Average” is an optional smoothing function, which
averages the value for each cell based on its neighbors. The win-
dow for this averaging needs to be an odd number, and can be set
under “Window” (see Note 16).
The output of this analysis is a number of text files for each and
every control or treatment sample analyzed, depending on which
has been chosen under “Output Type.”
Per sample, two text files are created. The first contains
the data for the mean ratio (treatment over control) of all geo-
metric measurements for each cell (“[filename].csv”). The sec-
ond file contains the value of elongation factors (E-factors)
(“[filename]_eFactor.csv”). These text files may be imported into
MorphoGraphX as heatmaps to false color their respective seg-
mented cellular organ mesh files.
An additional file named “Data_CI.csv” is also generated as
part of the process output. This file contains the ratio of each
Fig. 12 Drop-down menu for the “Display Cell Data” tool. Any of the geometric
data can be chosen, and will be displayed as false colors on the mesh
Fig. 13 Parameters for the “Growth Analysis 3D” tool. “Control folder” and
“Treatment folder” are directories that contain the raw data of samples to be
analyzed and compared. “Output folder” is the directory in which the results of
this analysis will be saved. Under “Output Type,” either “Control” or “Treatment”
can be chosen, referring to the samples that will be used to generate heatmaps
with the result of the analysis. “Sliding Average” is an optional smoothing func-
tion which averages the value for each cell based on its neighbors. The window
for this averaging needs to be an odd number, and can be set under “Window”
metric for each cell type, at each position, in addition to their

respective 95 % confidence intervals (based on the t-distribution) at
each data point. Plots can be generated to visualize these data by
importing the Data_CI.csv file into a spreadsheet program such as
Excel. Note that the confidence intervals will not be smoothed, if
smoothing has been enabled; the “sliding average” smoothing
function will only be applied to the metric itself.
3.20 Step 11 To statistically analyze the abundance of a reporter within each cell
of 3DCellAtlas: of the organ in question, heatmaps for reporter abundance need to
Statistical Analysis be created for each individual sample. For this, the segmented
of 3D Reporter mesh of a sample is loaded, together with the z-stack of the reporter
Abundance channel of that same sample in the main stack 1.
In the “Process” tab, under “Mesh” > “Heatmap,” run
“Heatmap” (Fig. 14). This will open a window in which Heat Map
Type and Heat Map Visualisation can be chosen from drop-down
menus (Fig. 15). For reporter abundance, Volume and Interior
signal are chosen to quantify reporter concentration.
Ticking the box next to “Signal average” will calculate the
reporter abundance relative to the volume of each cell, so as to
reflect reporter concentration for each cell.
Under “Report to Spreadsheet,” type in a name for the output
file to be created. This will be a csv file, which contains each cell ID
of the sample and its respective reporter abundance.
Fig. 14 Reporter abundance analysis in MorphoGraphX. Screenshot of a MorphoGraphX window with a sample
mesh and reporter z-stack file loaded. The mouse pointer highlights the “Heatmap” function to be used to
calculate reporter abundance in each cell
Fig. 15 Windows displaying options for generating heatmaps. (a) The type of heatmap can be chosen from
Area, Volume, or Walls. (b) The drop-down menu for heatmap visualization offers several options; for reporter
abundance, “Interior signal” is chosen
Save cell data and reporter abundance files for all samples to be
analyzed in a directory. The “3D Reporter Abundance Analysis”
process will recognize pairs of files with the same file name, with
the reporter abundance file name containing an additional string
that can be chosen by the user. In this way, the cell type and posi-
tion information can be matched to the corresponding reporter
abundance file.
In the “Process” tab, under “Mesh” > “Cell Atlas 3D” > Statistics,
run “Reporter abundance analysis.” This will prompt the user to
direct the program to the required directory.
In the parameters (Fig. 16), if the value for “Input folder” is
left blank, the program will prompt the user to guide it to the cor-
rect directory.
To visualize data on a mesh that is not part of the samples to
be analyzed, a separate sample (in the form of its “cell_data.csv”
file) can be set as “Output File.” “Merge With File” will then have
to be set as “Yes.”
Raw data can be optionally filtered in the process by setting
“Upper Filter Type” and/or “Lower Filter Type” to one of the
options. “No Filter” will include all data points, while “Percentage”
or “Value” will cut off data points beyond the given limit. The
Fig. 16 Parameters for the 3D reporter abundance analysis tool. “Input folder” is
the directory that contains cell data and corresponding reporter abundance data
of all samples to be analyzed. To visualize the results on a separate sample
mesh, the “cell_data.csv” file of this mesh can be set as “Output File.” This
requires the “Merge With File” option to be set to “Yes.” “Sliding Average” is an
optional smoothing function which averages the value for each cell based on its
neighbors. The window for this averaging needs to be an odd number, and can
be set under “Sliding Avg Window.” Raw data can be optionally filtered in the
process by setting “Upper Filter Type” and/or “Lower Filter Type” to one of the
options. “No Filter” will include all data points, while “Percentage” or “Value” will
cut off data points beyond the given limit. The limits for each filter can then be
set to a value defined under “Upper Filter Limit” and “Lower Filter Limit”
limits for each filter can then be set to a value defined under “Upper
Filter Limit” and “Lower Filter Limit.”
If the mesh to be used for visualization of these data is loaded
into MorphoGraphX, the results will automatically be loaded onto it.
The data output file, labeled “output_CI.csv,” will automati-
cally be created in the same directory. It contains the mean of the
reporter abundance for each cell type, at each position, in addition
to the confidence intervals at each data point. From there, plots
can be generated to visualize these data as described in the growth
analysis (see Note 17).
3.21 Visualization Any cell data obtained through 3DCellAtlas, be it individual geo-
of Data on a Segmented metric data of an individual sample, growth data from the compari-
Organ Mesh son of two sets of samples, or reporter abundance data from a set
of samples, can be displayed on a segmented cellular organ mesh of
your choice. Cells will thus be false colored according to the value
in the data file. For this, a *.csv file with the respective data is
loaded onto a segmented mesh as a heatmap in the “Process” tab,
under “Mesh” > “Heat Map” > “Load Heat Map” (Fig. 17). It is
important that the data are correctly matched to the cell IDs of the
segmented mesh of choice with regard to cell type and position
Fig. 17 Visualizing data as heatmaps on a segmented mesh. (a) Close-up of the Process tab with the drop-
down menu for the heatmap tools. Highlighted is “Load Heat Map,” which is used to upload a data file (in *.csv
format) to false-colored cells in a segmented cellular organ mesh. (b–d) Examples of false-colored meshes of
an Arabidopsis embryo axis, halved to show cells on the inside. Cells have been false colored to show (b) cell
types, (c) the relative radial distance of each cell from the Bezier line, and (d) cell volume
(see Note 18). “Rescale Heat Map” allows the user to change the
range of the scale displayed. To save a screenshot, the camera tool
is used; image type and resolution can be adjusted according the
user’s preferences.
4 Notes
1. Image stacks need to be stored as 16-bit images; if 8-bit images

are acquired, they will be too dark to see in MorphoGraphX.
This can be overcome, however, by auto-adjusting the transfer
function. In the “Main” tab, click on the color map button for
the stack that is open, and click on “Auto Adjust” in the window
thatpopsup.Inthe“Process”tab,use“Stack” > “Filters” > “Brighten
Darken” with an amount of 16 to convert 12-bit images, or 256
for 8-bit images, then reset the transfer function, and save the
16-bit stack.
2. The pinhole should be as small as practical (depending on sig-
nal intensity) to narrow the optical slice thickness. This will
reduce noise to a minimum, but will also decrease the amount
of fluorescence signal being captured.
3. For best quality and resolution, choose a z-stack interval of the
same size as the x–y resolution in each slice (i.e., if the resolu-
tion is 0.7 μm per pixel in the x–y direction, a slice interval of
0.7 μm is recommended).
4. An extensive user manual and FAQs can be found on the
developers’ website: http://www.MorphoGraphX.org.
5. The radius for blurring the image depends on the magnifica-
tion used during its acquisition; a 0.5 μm radius is typically
used for images acquired with 20 or 25× objectives. A lower
radius will work better for smaller cells, for example images
acquired with 40× or higher objectives.
6. A segmentation threshold that is too high will cause thinner cell
walls with low signal intensity to be missed and cells to be merged,
while a threshold that is too low over-segments cells into numer-
ous segments. It is worth noting that oversegmented cells can be
corrected, while merged cells cannot, so it is advisable to choose a
threshold with minimum to no bleeding of adjacent cells, which
could lead to some cells being oversegmented.
7. If a lot of cells within the image are merged after segmenting
the image (“undersegmented” or “bleeding”), the signal-to-
noise ratio in the image might be too low.
● Likely, the threshold used for segmenting the image has
been too high. Use a lower threshold for segmentation. It
is also possibly to improve image quality by normalizing
the image before blurring (in the “Process” tab, under “St
ack > Filters > Normalization”). This will even out differ-
ences in signal across the sample by narrowing the distri-
bution of signal intensities.
● Bleeding of cells could also be due to the sample not being
clear enough. If that is the case, continue clearing the sam-
ple, with changes of chloral hydrate clearing solution, to
reduce noise inside of cells, and re-image.
● Similarly, if the staining has not worked well, the intensity
from the cell walls could be too low to be sufficiently dis-
tinguishable from noise.
8. If cells are excessively oversegmented after segmentation,
repeat the segmentation using a higher threshold.
9. A common error message in MorphoGraphX upon segment-
ing an image is the following: “Exception occurred during
SingleMethodExecute […]: Number of objects greater than
maximum of output pixel type.” This indicates that the ITK
segmentation has found too many regions to label in 16-bit,
which could be due to the image not having been blurred, or
the radius of the Gaussian blur not having been big enough.
10. A mesh size of 2 μm is typically used. This value describes the
distance between points (vertices) within the mesh. A higher
value will generate a more coarse mesh; this risks misrepresen-
tation of cell sizes, as corners may be cut off. A smaller value
will generate a smoother mesh with more points, which will
increase file size, but is likely to represent the cell sizes and
shapes more accurately.
11. The smoothing operation for meshes in MorphoGraphX uses
the Laplacian smoothing; it repositions vertices to an average
position along a mesh. As a side effect, this will slightly shrink
the mesh, and thus misrepresent cell sizes. This operation
should therefore be used with caution [12].
12. While adjusting the handle points of the Bezier to align with the
axis of the imaged plant organ, it is helpful to reduce the opacity
of the image, and make use of clipping planes to determine the
exact position of the Bezier curve within the 3D image.
13. After cell types have been identified and corrected via the topo-
logical check function, it is still possible that some cells are not
assigned the correct type. It is important to note that the accu-
rate determination of cell types strongly depends on the quality
of the segmentation. Cell types are saved as parent labels. In
the main tab, next to “Surface,” the display can be changed to
show “Nrml” (no color), “Labels” (individual cell IDs),
“Heat” (values based on a heatmap), or “Parents” (cell types).
To edit parent labels, check that “parents” is ticked in the main
tab, select a parent label using the dropper tool amongst the
mesh tools (see Fig. 6c), and change the parent label of another
cell to the selected one using the paint bucket tool.
14. It is important that all files to be analyzed are in csv file format.
If one or all files are not comma delimited, an error message
will occur, and the analysis will not run.
15. During this process, data from multiple samples will be pooled
by cell type, and then associated cortical cell, into bins repre-
senting unique cells. Data will be log transformed, and the
mean for each cell size metric will be calculated. Total growth
is then determined as a ratio of the means of treatment over
control. For more detail of the exact calculations, see ref. 11.
16. If a sliding window average of three is chosen, the value for
each cell will be changed to the average of three associated
cortical cell positions, one cell on either side. Cells in the first
and last position within samples, possessing only a single neigh-
bor, will be reduced to the average of two cells.
17. Reporter concentration values are assigned to the appropriate
cell type and position in this process, and data from multiple
samples will be pooled. The statistical analysis determines the
mean reporter concentration for each cell type and position as
described in Note 15 for growth analysis. The 95 % confidence
intervals for reporter abundance data are computed using
bootstrapping, with a minimum sample size of seven cells.
Thus, should there be less than seven data points available for
one particular cell type at a given position, the confidence
interval will be returned as a value of −1.
18. Each individual cell of the segmented mesh is assigned a cell
type and position; the values of each metric to be displayed for
each of these combinations have to be matched according to
these cell type-cell position combinations. This can be done
quickly using database software, for example Microsoft Access,
to match the cell data file of the segmented cellular mesh to be
used, and the data file to be displayed.
Acknowledgments
G.W.B. was funded by Biotechnology and Biological Sciences

Research Council (BBSRC) Grant BB/L010232/1 and a
Birmingham Research Fellowship. P.S. was funded by BBSRC
grant BB/J017604/1. T.D.M.J. is supported by a Royal
Commission for the Exhibition of 1851 Fellowship. R.S.S. was
funded by Swiss National Science Foundation Interdisciplinary
Project Grant number CR32I3_143833 and the Max Planck
Society.
References
1. Roeder AH et al (2011) Computational mor- Arabidopsis embryo. Proc Natl Acad Sci U S A
phodynamics of plants: integrating develop- 111(23):8685–8690
ment over space and time. Nat Rev Mol Cell 7. Cunha AL, Roeder AH, Meyerowitz EM
Biol 12(4):265–273 (2010) Segmenting the sepal and shoot apical
2. Fernandez R et al (2010) Imaging plant growth meristem of Arabidopsis thaliana. Conf Proc
in 4D: robust tissue reconstruction and lineaging IEEE Eng Med Biol Soc 2010:5338–5342
at cell resolution. Nat Methods 7(7):547–553 8. Yoshida S et al (2014) Genetic control of plant
3. Kierzkowski D et al (2012) Elastic domains development by overriding a geometric divi-
regulate growth and organogenesis in the sion rule. Dev Cell 29(1):75–87
plant shoot apical meristem. Science 9. Barbier de Reuille P et al (2015) MorphoGraphX:
335(6072):1096–1099 a platform for quantifying morphogenesis in
4. Roeder AH et al (2012) A computational image 4D. eLife 4:e05864
analysis glossary for biologists. Development 10. Shapiro BE et al (2015) Analysis of cell divi-
139(17):3071–3080 sion patterns in the Arabidopsis shoot apical
5. Truernit E et al (2008) High-resolution whole- meristem. Proc Natl Acad Sci U S A 112(15):
mount imaging of three-dimensional tissue 4815–4820
organization and gene expression enables the 11. Montenegro-Johnson TD et al (2015) Digital
study of Phloem development and structure in single-cell analysis of plant organ develop-
Arabidopsis. Plant Cell 20(6):1494–1503 ment using 3DCellAtlas. Plant Cell
6. Bassel GW et al (2014) Mechanical constraints 27(4):1018–1033
imposed by 3D cellular geometry and arrange- 12. Bassel GW (2015) Accuracy in quantitative 3D
ment modulate growth patterns in the image analysis. Plant Cell 27(4):950–953
Chapter 12
Analyzing Cell Wall Elasticity After Hormone Treatment:

An Example Using Tobacco BY-2 Cells and Auxin
Siobhan A. Braybrook
Abstract
Atomic force microscopy, and related nano-indentation techniques, is a valuable tool for analyzing the
elastic properties of plant cell walls as they relate to changes in cell wall chemistry, changes in development,
and response to hormones. Within this chapter I will describe a method for analyzing the effect of the
phytohormone auxin on the cell wall elasticity of tobacco BY-2 cells. This general method may be easily
altered for different experimental systems and hormones of interest.
Key words Atomic force microscopy, Auxin, Cell growth, Cell wall, Elasticity
1 Introduction
Cell wall elasticity has been correlated with important developmental

transitions and growth in several plant systems [1–5]. There are sev-
eral methods currently used to measure cell wall elasticity at the cell
and tissue level (reviewed in [6]); one of the most prevalent modern
methods of measuring cell wall elasticity in planta at the cellular level
is atomic force microscopy (AFM) [1, 3–5, 7, 8]. Non-AFM meth-
ods include other nano-indentation devices [9–11] and osmotic
swelling and shrinking of tissues [2, 9]. Previously, AFM-based
indentation has been used to demonstrate that the phytohormone
auxin induces an increase in cell wall elasticity (decrease in the elastic
modulus) of cells near the shoot apex in Arabidopsis thaliana [1]. In
order to simplify the experimental system for this chapter, I will
examine cell wall elasticity in growing tobacco bright yellow 2 (BY-2)
culture cells upon treatment with auxin. Using the simple method
presented here, I have confirmed that auxin treatment increases cell
length in BY-2 cells and decreases cell wall elastic modulus. This is
consistent with data showing that BY-2 cells respond positively, with
respect to growth, when treated with a cell wall-acidifying agent or
the wall-loosening agent expansin [12].
125
126 Siobhan A. Braybrook
This method proposes a two-step process: first, the experi-

menter should determine the concentration of hormone to use
given their experimental question; second, the experimenter should
choose a time point for AFM-based elasticity analysis that further
suits their question and system. Here, a concentration of 10 μM
IAA and a treatment time of 1 h were chosen after initial experi-
ments; a more complex set of treatment times and concentrations
could easily be constructed using this method as a basis.
2 Materials
2.1 Plant Material Tobacco BY-2 cells, non-transgenic, were obtained from the
Leibniz Institute DSMZ-German Collection of Microorganisms
and Cell Cultures (DSMZ, Cat # PC-1181). Cells were cultured
in MSBY media (see below), in the dark at 25 °C, with shaking
at 100 rpm. Cells were sub-cultured weekly into fresh media at a
1/10 dilution. All experiments were performed 1–2 days after
sub-culturing, when cells were growing in their exponential
phase.
2.2 Solutions All solutions should be prepared fresh, using ultrapure water whose
pH has been verified as neutral (see Note 1).
0.55 M Mannitol, for plasmolysis.
10 mM Indole acetic acid (IAA) stock solution: 1.97 mg of IAA
(Sigma, Cat # I5148) is dissolved in 1 mL of ethanol. After
dissolving, ultrapure water is added to 10 mL. Store in the
dark at −20 °C.
MSBY media: Full MS media supplemented with 1 mg/L thia-
mine, 100 mg/L myo-inositol, 0.2 mg/L 2,4-D, 3 % sucrose,
pH 5.7 (see Note 2).
2.3 Equipment Atomic force microscope: The experiments described were per-
and Consumables formed on a JPK Nanowizard 3 (JPK Instruments, DE).
Cantilevers (without tips) of spring constant ~45 N/m (Windsor
Scientific, UK, Cat #TL-NCL-10). Silica beads of ~5 μm radius
were mounted as per [7] (Microspheres-Nanospheres,
Corpuscular, USA, Cat# C-SIO-5.0). Note that tip size and
cantilever stiffness should be chosen carefully with respect to
the question and system.
Small petri dishes coated with adhesive agent (see Note 3).
Pipette and tips.
A dissecting microscope with camera for observation and length
measurements.
250 mL Flasks for sub-culturing and growth of cells.
A sterile bench.
Measuring Hormone-induced Changes in Wall Elasticity 127
3 Methods
3.1 Assessing 1. Use the literature to determine a range of concentration of

an Appropriate hormone to test. You are looking for a hormone concentration
Hormone that is not too high so as to be unrealistic, but also shows con-
Concentration sistent and significant changes in the phenotype you are inter-
and Time of Treatment ested in. Here 1, 10, and 50 μM IAA were tested and 10 μM
(See Note 4) was chosen (see Fig. 1).
2. At the time of sub-culturing, make four extra sub-culture flasks
for this initial experiment. To two flasks of 30 mL MSBY media
add 30 μL of 10 % ethanol (Mock), and to the other two flasks
add 30 μL of 10 mM IAA stock (10 μM IAA treatment). Sub-
culture 1/10 volume of week-old culture into each flask. Grow
the cells for 24 h in culture conditions.
3. After 24 h, sterilely remove a small aliquot of cells and place on
a glass glide with a cover slip. Return the cells to culture condi-
tions. Using a dissecting microscope, take several images of the
slide area. Make sure that you have the correct calibration for
length measurements. Figure 1 shows the effect of 10 μM IAA
on BY-2 cell length after 24 h of treatment.
4. Using ImageJ (open source, ImageJ.net), measure the length
of cells within the images. Select cells at random, covering a
wide area of each image and thus the whole slide area.
5. Repeat these measurements for further time points of interest
(see Note 5). In the experiments here, cell lengths were mea-
Fig. 1 Cell length of BY-2 cells after culture with the auxin IAA. Cell length of randomly chosen individual cells
with a BY-2 cell chain (diagram on left, shading not meant to denote a fixed cell position) was measured after
24 h of growth with mock (diluted ethanol) or 10 μM IAA treatment. Cell length was measured from two repli-
cate flasks per treatment and from 100 cells per replicate. A notched box plot was generated in Matlab
sured at 1, 2, and 3 days. As cell length was increased at all time

points, it was decided to assess the effect of IAA on wall elastic-
ity before 24 h (when a statistically significant phenotype was
observable, Fig. 1a).
3.2 Preparation 1. Using cells 1–2 days post-sub-culture, take a sterile aliquot of
of Cells for AFM 3 mL. Apply the aliquot of cells to the coated petri dish. Allow
the cells to sit for ~5 min, then tip the plate to remove the
media and un-stuck cells, rinse three times with MSBY media
(−2,4D), and then fill the dish halfway with fresh MSBY (−2,4-
D) media. Check that the cells are remaining stuck to the dish
during washes. The media will be replaced with 0.55 M man-
nitol just before the initial AFM experiments begin (see Note
6). Figure 2a shows an outline of the experiment.
3.3 AFM-Based Cell 1. Prepare a cantilever with a 5 μM bead and calibrate as in [7].
Wall Elasticity Different tips may be used for different purposes and experi-
Measurements mental questions, as detailed in [7].
2. Place the petri dish of cells (in 0.55 M mannitol) under the
AFM head.
3. Locate an area of the dish where several cells are well adhered
to the surface and flat (see Note 7). Take a picture of the cells
so that you can locate them again after treatment.
4. Position the AFM tip over a cell-cell junction and perform an
AFM scan. Here I have used a force of 500 nN which yields an
indentation depth of roughly 250–500 nm. Details of AFM
procedure can be found in [7] and a schematic of how the
indentation relates to BY-2 cell dimension in Fig. 2b. Here, a
Fig. 2 Diagram of experimental procedures. (a) The experimental workflow for auxin treatment of BY-2 cells
for AFM indentation. Cells are plasmolyzed in mannitol for an initial AFM scan; they are then recovered and
treated in media for 1 h, then plasmolyzed again, and re-scanned with the AFM. (b) Diagrammatic representa-
tion of a BY-2 cell in cross section with dimensions and scale of AFM indentation. BY-2 cells are usually
~25 μm in diameter and have a wall thickness of ~100 nm [13]. Indentations here were performed with a
5 μm diameter bead to a depth of 250–500 nm
map of 32 × 32 indentations was made over a 30 μm2 area.

Cell-cell junctions are essential as the cross wall will be used for
further analysis; in plasmolyzed cells, cross walls will be evident
as stiffer regions within the scan. You can use the height and
stiffness maps produced during the AFM scan to help ensure
that you are in the right area.
5. Proceed to make cross-wall scans for several cells within your
field of view. It is likely that you will lose some cells between
the subsequent steps, so starting with more than you need is
advisable. Take a picture of each cell you scan so that you can
re-locate it after treatment. Try to keep your time of collection
under 1 h to keep your cells happy.
6. After you have collected enough maps, remove the AFM head
and carefully use a pipette to remove the mannitol. Try not to
move the plate so that the position of cells under the AFM does
not shift too much. Wash the cells once with MSBY media
(−2,4-D), and then fill the dish with 5 mL of MSBY (−2,4-D).
For cells to be mock treated, add 5 μL of 10 % ethanol. For cells
to be auxin treated, add 5 μL of 10 mM IAA stock (to get 10 μM
treatment conditions). Let the cells sit for 1 h (see Note 8).
7. Exchange the treatment media with 0.55 M mannitol (includ-
ing one wash step). Place the plasmolyzed cells under the AFM
head.
8. Using your field of view image from step 3, locate the same
area in your dish.
9. Using your cell images from steps 4 and 5, re-find the same
cells and perform AFM maps over the same cell-cell junctions
using identical settings.
3.4 AFM Data 1. Indentation force maps may be analyzed in several ways.
Analysis Usually, AFM manufacturers supply analysis software. Here, I
have used JPK SPM Data Processing (JPK Instruments,
DE. Version “spm 4.3-10”). During analysis, you will need to
choose whether you will fit models for linear stiffness or a
Hertzian indentation (yielding a Young’s modulus, here “EA”)
[7]. In brief, stiffness fits may be useful when indentation
depth is constant and there is adhesion in the sample; in these
cases the approach indentation should be analyzed. In cases
where more complex indentation behavior is observed, and
where the entire indentation curve is desired for analysis, fit-
ting a Hertzian model is preferable. Here, I have fit a Hertzian
model to the indentation curve in order to obtain an apparent
Young’s modulus (EA) for each map.
2. Select the data from along the cross wall in each map. Use this
data to obtain values of EA along cross walls for each cell junc-
tion before and after treatment. Figure 3a shows the EA value
Fig. 3 Young’s modulus (EA) of BY-2 cell walls before and after treatment. The log of EA (in Pascals) was plotted
as box plots for each equivalent cross wall before (−) and after (+) treatment. IAA-treated cells which showed
no significant changes are shaded in dark grey. One replicate showing an increase in EA (stiffening) after IAA
treatment is shaded in light grey. Boxes are colored by individual cell, with the lighter shade being before treat-
ment and the darker after treatment. It is evident that there is variation in EA between cells and also in the
repose of individual cells to auxin, although the majority of cells responded by decreasing EA (t-test; P < 0.05
(*), P < 0.01 (**))
for each cell in this experiment, before and after treatment.

Here it can be seen that there is a large variance between sam-
ples, even without treatment. For this reason, it may be advis-
able to keep biological samples separate for analysis and to
focus on relative trends. For example, in 13 of the cells treated
with IAA a significant decrease in EA was obtained after 1 h of
treatment (t-test; P < 0.05). There was one sample where the
EA increased after treatment (light grey over-shading), and
four where no difference was observed (dark grey over-shad-

ing). As BY-2 cells are not always perfectly synchronized in
growth, this may represent a change in biological response to
the treatment.
3. For this example, data was collected, analyzed, and graphed in
MatLab.
4 Notes
1. Water pH should be neutral as pH can affect cell wall elasticity

and cell growth.
2. As 2,4-D is an auxin, for treatments cells were washed three
times with media minus 2,4-D before first measurement.
Application of mock (1/1000 dilution of 10 % ethanol) or
10 μM IAA was performed in media without 2,4-D.
3. It is essential that cells do not move during the experimental
procedure. After much testing, it appears as though tissue sec-
tion adhesives (e.g., Biobond Tissue Adhesive, Agar Scientific,
UK) work well. Roughly 10 μL of Biobond adhesive can be
placed in the bottom of a small petri dish. A small piece of
parafilm is then used to spread the adhesive over the whole
plate bottom. The plate is placed in a 50 °C oven to dry for at
least 10 min.
4. When assessing the concentration of hormone to use, it is
important to calibrate your method to your system and ques-
tion. Here, I present a method for BY-2 cells in culture. You
might be interested in how auxin (or another hormone) affects
some other growth process in another system (perhaps gibber-
ellins and hypocotyl elongation). It is essential that you assess
several concentrations for the phenotype you wish to examine,
and also make assessments in time, thus allowing you to wisely
choose the best concentration, and treatment time, for your
AFM experiment.
5. If at all possible, measurement of your phenotype of interest
should be made with as a high a temporal resolution as possible.
For example, in BY-2 cells one could use time-lapse imaging as
in [12] if one had the correct equipment. This would allow for
very accurate choosing of time points for ANF analysis.
6. For the method presented here, it is essential that cells are plas-
molyzed as the indentation depth is more than wall thickness and
thus turgor pressure would affect our measurements. It is impor-
tant to know how quickly your cells plasmolyze for these experi-
ments. I recommend testing this beforehand using a membrane
dye (such as FM 4-64) or transgenic membrane marker, and
monitoring plasmolysis under a confocal microscope. In our
hands, BY-2 cells plasmolyze in 0.55 M mannitol within 5 min.

There are methods which do not require plasmolysis; please refer
to [3, 7, 8].
7. If cells are not flat, there is a strong possibility that they will be
knocked loose during the AFM scanning. You want to ensure
that the cells you begin with will have staying power for the
duration of the experiment.
8. The time of treatment will vary depending on your question.
Here, it was desirable to see if changes in wall elasticity pre-
ceded changes in length and the first experiments revealed that
24 h of treatment showed length changes; thus a 1-h treat-
ment should precede this. You might wish to perform a more
detailed time series for the AFM scans. To do this, simply use
this basic protocol as a starting point but add in more time
steps. It may be efficient to stagger treatments and reads.
Acknowledgements
I would like to thank Cris Kuhlemeier, Theres Imhof, and Alexis

Peaucelle for help with initial experiments with BY-2 cells and
AFM. I would also like to thank my research team at SLCU for
their patience with me while writing. This work was started in the
lab of C. Kuhlemeier, supported by an NSF-IRFP (Award
#0853105; to S.A.B.), and completed at The Sainsbury Laboratory,
University of Cambridge, supported by a Gatsby Career
Development Fellowship (GAT3396/PR4) and the BBSRC (BB.
L002884.1) to S.A.B.
References
1. Braybrook SA, Peaucelle A (2013) Mechano- growth rate between adjacent cells in
chemical aspects of organ formation in Arabidopsis. Cell 149(2):439–451
Arabidopsis thaliana: the relationship between 6. Milani P, Braybrook SA, Boudaoud A (2013)
auxin and pectin. PLoS One 8(3):e57813 Shrinking the hammer: micromechanical
2. Kierzkowski D et al (2012) Elastic domains approaches to morphogenesis. J Exp Bot
regulate growth and organogenesis in the plant 64(15):4651–4662
shoot apical meristem. Science 335(6072): 7. Braybrook SA (2015) Chapter 13—Measuring
1096–1099 the elasticity of plant cells with atomic force
3. Milani P et al (2011) In vivo analysis of local microscopy. In: Ewa KP (ed) Method cell biol-
wall stiffness at the shoot apical meristem in ogy. Academic, New York
Arabidopsis using atomic force microscopy. 8. Sampathkumar A et al (2014) Subcellular and
Plant J 67(6):1116–1123 supracellular mechanical stress prescribes cyto-
4. Peaucelle A et al (2011) Pectin-induced skeleton behavior in Arabidopsis cotyledon
changes in cell wall mechanics underlie organ pavement cells. eLife 3
initiation in Arabidopsis. Curr Biol 21(20): 9. Weber A et al (2015) Measuring the mechani-
1720–1726 cal properties of plant cells by combining
5. Uyttewaal M et al (2012) Mechanical stress micro-indentation with osmotic treatments.
acts via Katanin to amplify differences in J Exp Bot 66(11):3229–3241
10. Routier-Kierzkowska A-L et al (2012) Cellular 12. Link BM, Cosgrove DJ (1998) Acid-growth
Force Microscopy for in vivo measurements of response and α-Expansins in suspension cul-
plant tissue mechanics. Plant Physiol 158(4): tures of Bright Yellow 2 tobacco. Plant Physiol
1514–1522 118(3):907–916
11. Hayot CM et al (2012) Viscoelastic properties 13. Sabba RP, Durso NA, Vaughn KC (1999)
of cell walls of single living plant cells deter- Structural and immunocytochemical characteriza-
mined by dynamic nanoindentation. J Exp Bot tion of the walls of Dichlobenil‐habituated BY‐2
63(7):2525–2540 tobacco cells. Int J Plant Sci 160(2):275–290
Chapter 13
FRET-FLIM for Visualizing and Quantifying Protein

Interactions in Live Plant Cells
Alejandra Freire Rios, Tatyana Radoeva, Bert De Rybel, Dolf Weijers,
and Jan Willem Borst
Abstract
Proteins are the workhorses that control most biological processes in living cells. Although proteins can
accomplish their functions independently, the vast majority of functions require proteins to interact with
other proteins or biomacromolecules. Protein interactions can be investigated through biochemical assays
such as co-immunoprecipitation (co-IP) or Western blot analysis, but such assays lack spatial information.
Here we describe a well-developed imaging method, Förster resonance energy transfer (FRET) analyzed
by fluorescence lifetime imaging microscopy (FLIM), that can be used to visualize protein interactions
with both spatial and temporal resolution in live cells. We demonstrate its use in plant developmental
research by visualizing in vivo dimerization of AUXIN RESPONSE FACTOR (ARF) proteins, mediating
auxin responses.
Key words FRET, FLIM, Protein interactions, Fluorescent proteins
1 Introduction
In the mid-1990s, commercial confocal microscopes were intro-

duced, which has revolutionized the investigation of biomolecules in
living cells [1]. Visualizing biological processes has made great prog-
ress due to significant advances in instrument and detector design as
well as the introduction of genetically encoded fluorescent proteins
in vivo. Although confocal microscopy provides very-high-quality
multicolored images, spatial resolution is limited by the diffraction
of light [2]. Since 2006, several fluorescence imaging techniques
have been developed known as super-resolution imaging technol-
ogy, which provide fine-detailed and high-contrast images to create
images down up to 20 nm spatial resolution [3, 4]. However, these
imaging techniques are still unable to visualize protein interactions
in vivo [5, 6]. Förster resonance energy transfer (FRET) is a phe-
nomenon of radiation-less energy transfer from a fluorescent donor
molecule to an acceptor molecule (either fluorescent or not) through
135
136 Alejandra Freire Rios et al.
dipole-dipole coupling. This process only takes place when the

donor and acceptor molecules are in very close proximity (typically
2–10 nm; [7]). The energy transfer rate is inversely proportional to
the sixth power of the distance between donor and acceptor, making
FRET very sensitive to small changes in distance and useful as a
molecular ruler ([8]; see Fig. 1). In order to accomplish FRET, the
fluorescent donor and acceptor molecules have to fulfill several pre-
requisites: (1) a spectral overlap between donor emission and accep-
tor absorption spectra, (2) close proximity between donor and
acceptor (<10 nm), and (3) finite dipole orientation factor [7].
FRET can be quantified using different imaging methods of which
fluorescence lifetime imaging microscopy (FLIM) is the most robust
and straightforward approach [9]. A fluorescence lifetime can be
defined as the average time a fluorophore remains in the excited
a
Large distance Close proximity
low FRET high FRET
Interaction
Donor Acceptor
b c d
1.0 D only
D+A
Normalized intensity
0.5
0.0
0 10 20
Time (ns)
Fig. 1 Illustration of FRET principle and FRET analysis. (a) Two DNA-binding domains of ARF5 are coupled to
CFP (donor) and YFP (acceptor). Large distance between CFP and YFP results in low FRET whereas high FRET
signals are obtained if ARF domains are in close proximity and correct orientation. FRET can be quantified by
fluorescence lifetime analysis. A fluorescence intensity image (b) shows the nuclear expression of the donor
ARF5-sCFP3A in a protoplast. Per pixel the time-resolved photon distribution is plotted (c) of which the fluo-
rescence lifetime is calculated and shown as a false color-coded fluorescence lifetime image (d). Interaction
of a donor with an acceptor results in a decrease of the fluorescence lifetime (c)
Visualizing Protein Interactions in Plant Protoplasts 137
state before returning to the ground state. The fluorescence lifetime

is an intrinsic property of a fluorophore and is independent of
concentration, absorption by the sample, sample thickness, photo-
bleaching, and laser excitation intensity [10]. However, the fluores-
cence lifetime is very sensitive to the microenvironment of the
fluorophore such as pH, ion or oxygen concentration, molecular
binding, or the proximity of energy acceptors [11], which makes this
method unique for FRET analysis. FRET-FLIM measurements are
based on determination of the fluorescence lifetime of the donor
molecule. The close proximity of acceptor molecules creates an addi-
tional relaxation path of the donor-excited fluorophores resulting in
a decreased donor fluorescence lifetime ([9]; see Fig. 1). The amount
of donor fluorescence lifetime reduction is directly correlated with
the FRET efficiency (E) via E = (1 − τDA/τD) where τDA is the fluores-
cence lifetime of the donor in the presence of acceptor and τD is the
fluorescence lifetime of the donor alone. In order to obtain a com-
prehensive overview and critical essentials concerning the funda-
mentals of FRET and the methods of measurements, we refer to a
detailed description [12].
In this chapter, a detailed description of FRET-FLIM technique
and analysis is presented by showing recent data on the dimeriza-
tion of the transcription factor family AUXIN RESPONSE
FACTORs (ARFs) [13] in living plant cells. ARFs act dependent on
the signaling molecule auxin, which is a small organic molecule
important for cell division, elongation, and differentiation [14].
Specificity in auxin responses is generated through the DNA-
binding domain (DBD) of ARFs. Crystal structures of the ARF5
and ARF1 DBDs were solved showing homodimerization of DBDs
generating cooperative DNA binding, which is critical for in vivo
ARF5 function [13]. ARF5 dimerization was also demonstrated in
Arabidopsis protoplasts using a FRET-FLIM analysis [13]. In this
assay ARF5-sCFP3A and ARF5-sYFP2 fusion proteins were
expressed and donor fluorescence lifetime analysis revealed specific-
ity in homodimerization of ARF5 proteins. Full-length ARF5
homodimerization revealed a decrease of donor fluorescence life-
time resulting in a FRET efficiency of 10 % (see Subheading 3) [13],
whereas mutations within the dimerization domain of the DBD of
ARF5 resulted in FRET efficiencies around background level. These
FRET data excluded the possibility of ARF DBD dimerization
being an artifact of the crystallization process but confirmed that
these interactions take place in vivo. Furthermore, site-directed
mutations and subsequent phenotypic analysis shed light on the
importance of dimerization for the ARF function in vivo.
Here we present a stepwise protocol to perform FRET-FLIM
analysis and show data on ARF5 homodimerization as an example.
Protein expression in plant protoplasts, FRET-FLIM analysis,
quantification of FRET efficiencies, use of proper controls, and dif-
ferent methods to display protein interactions are discussed.
2 Materials
2.1 Cloning 1. Plant expression vector: LIC vectors based on pMON999

of Transient (Monsanto, USA) [15].
Expression Vectors 2. The cDNA of ARF5 was cloned into LIC vectors harboring
either SCFP3A fluorescent protein (sCFP3A) or yellow fluo-
rescent protein (sYFP2) (see Note 1).
3. Constructs were generated using ligase-independent cloning
procedure (see Note 2).
2.2 Protoplast Protocol adapted from Refs. [16] and [17]:

Isolation
1. Plant material: Rosette leaves of Arabidopsis thaliana plants
(ecotype Columbia) grown for ±3 weeks under long-day con-
ditions (16-h light/8-h dark) at ±22 °C.
2. Mannitol solution: 0.4 M Mannitol, 20 mM KCl, and 20 mM
2-(N-morpholino) ethanesulfonic acid (MES) pH 5.7 in
Milli-Q water.
3. Calcium chloride solution: 1 M CaCl2 in Milli-Q water.
4. Enzyme solution: 1 % (w/v) cellulose “Onozuka” R10 (Serva
Electrophoresis, GmbH, Germany) and 0.2 % (w/v)
Macerozyme R10 from Rhizopus sp. (Serva Electrophoresis,
GmbH, Germany) dissolved in mannitol solution; subsequent
addition of CaCl2 to a final concentration of 10 mM.
2.3 Transfection Protocol adapted from Refs. [16] and [17]:

Protocol Adapted
1. PEG/Ca2+ solution: 40 % (w/v) Polyethyleneglycol 4000
from [16] and [17]
(PEG) 0.2 M mannitol, and 100 mM Ca(NO3)2 in Milli-Q
water (see Note 3).
2. W5 solution: 154 mM NaCl, 125 mM CaCl2, 5 mM KCl, and
2 mM MES pH 5.7.
3. W5/glucose solution: W5 solution containing 1 mM glucose.
4. MMg solution: 0.2 M Mannitol, 15 mM MgCl2, and 4 mM
MES pH 5.7.
5. Plastic round-bottom tubes.
6. Microscope 8-well slides
2.4 Confocal 1. Leica SP5X (see Note 4).

Imaging and FLIM 2. B&H SPC 730/830 module (Becker & Hickl, Germany).
3 Methods
3.1 Protoplast 1. All steps are carried out at room temperature unless otherwise
Isolation stated.
2. Rosette leaves of ±3-week-old A. thaliana plants are fixed on

Time tape adhered to the upper epidermis [18].
3. 3 M Magic tape is fixed to the lower epidermis to make a
“Tape-Arabidopsis Sandwich” ([18]; see Note 5).
4. A clean Petri dish (9 cm diameter) is covered with three stripes
of tape covered with leaves sufficient for six independent
transfections.
5. Add 25 mL of enzyme solution, and swirl to dampen all plant
material.
6. The Petri dish is transferred on a platform shaker (shak-
ing ± 60 rpm) and incubated at 27 °C for 20 min (Note 6).
7. Protoplasts are released from the leaf matrix by carefully pipet-
ting the enzyme solution on the taped leaves.
8. Transfer the protoplasts to a clean plastic round-bottom tube.
9. Protoplasts are collected by centrifugation for 3 min at 50 × g
using a tabletop centrifuge; wash once with 5 mL of W5
solution.
3.2 Protoplast 1. All steps are carried out at room temperature.

Transfection 2. Pipette plasmid DNAs in round-bottom tubes; for single trans-
fections 10–20 μg of the respective DNA is required and for
double transfections, 10–15 μg of each construct. Single trans-
fection is used to obtain protoplasts expressing the donor con-
struct only, whereas the transfections of two plasmids yield
protoplasts for the interaction studies.
3. Collect the isolated and washed protoplasts (see Subheading 3.1)
by centrifugation for 3 min at 50 × g in a tabletop centrifuge
and remove the supernatant carefully with a pipette.
4. Resuspend the protoplasts in 1.2 mL of MMg (see Note 6) and
transfer 200 μL aliquots to round-bottom tubes that contain
plasmid DNA already (see Note 7).
5. Add 220 μL of PEG/Ca2+ solution to the protoplasts, mix well
but carefully, and incubate for 5 min (see Note 8).
6. Stop the transfection by addition of 800 μL of W5 solution to
the protoplast solution and collect the protoplasts by centrifu-
gation at 50 × g for 90 s in a tabletop centrifuge.
7. Remove the supernatant with a pipette and wash with 5 mL of
W5 solution.
8. Collect protoplasts by centrifugation at 50 × g for 3 min,
remove the supernatant with a pipette, and resuspend the pro-
toplasts in 1 mL of W5 solution containing 1 mM glucose.
9. Protoplasts can stay in round-bottom tubes incubated at 22 °C
in the dark for 16 h (see Note 9).
3.3 Image 1. In this chapter the FRET-FLIM experimental procedure is

Acquisition optimized for a Leica SP8. Single-photon excitation pulses of
40 Mhz (see Note 10) are generated by a diode-pulsed laser
resulting in excitation pulses of ±10 ps. (see Note 11).
2. Image acquisition is performed using a 60×/1.2 water-
immersion objective.
3. Set objective for optimal glass thickness (see Note 12).
4. Cyan fluorescence emission was selected from 450–495 nm
using spectral detection (see Note 13).
5. Single photons are captured using internal hybrid detector (see
Note 14).
6. Fluorescence images of 128 × 128 pixel size are set for both the
Leica acquisition software and the B&H SPC 830 FLIM mod-
ule (see Note 15).
7. The analog-digital converter (ADC) of TCSPC module is set
at 256 channels (see Note 16).
8. FLIM acquisition is performed using line scanning at 400 Hz
(see Note 17) using an average count rate around 104 photons
per second for an acquisition time of 90 s (see Note 18).
9. Single transfected protoplasts expressing C-terminally sCFP3A-
tagged ARF5 result in the donor fluorescence lifetime required as
reference. To reveal dimer formation of ARF5, protoplasts
expressing ARF5 sCFP3A- and sYFP2-tagged proteins are inves-
tigated (see Note 18).
10. The protoplasts are transferred into an 8-well chamber for
imaging.
3.4 Analysis of FLIM 1. SPCImage is a software package, which is included with the
Data with B&H acquisition card. This software determines the instru-
SPCImage 5.2 mental response function from the rise of the decay curve and
performs data analysis using an exponential model function
(see Note 20).
2. After importing the raw data, a fluorescence intensity image
will appear. A blue crosshair is visible that can be used to point
anywhere in the image for displaying the fluorescence decay of
that pixel.
3. Within the histogram showing the fluorescence decay, the lim-
its in between the fluorescence lifetime analysis should be per-
formed, and can be defined. Typical values are around 2 ns
(before the rising edge) for the left and about 20 ns for the
right border.
4. A threshold of fluorescence intensity can also be set. Pixels will
be excluded from the analysis if the numbers of photons in that
pixel is below that value.
5. In case the number of photons is too low for fluorescence

lifetime analysis, static pixel binning can be applied. This specific
binning option is a procedure where the selected pixel is ana-
lyzed including the photon statistics of neighboring pixels for
calculation of the fluorescence lifetime. The binning factor can
be calculated according to the following formula: binning fac-
tor = (2n + 1)2, where n is the number of pixels. For the FLIM
images shown in Fig. 2, a binning factor of 1 has been applied
(see Note 21).
6. Before fitting the raw data, the number of components for the
underlying exponential fit function has to be defined (here one
or two). All other parameters should remain unfixed indepen-
dently of the fit model.
7. After the fit procedure, the fluorescence lifetimes are calculated
per pixel and displayed as a false-colored image (see Fig. 1d and
Fig 2a, c). During the fitting process, the chi-square value
between model function and data is minimized [19]. For a
good fit, the chi-square value should be around one and the
displayed residuals in the box below the fitted curve should
scatter around zero.
8. To reveal if protein interactions take place one compares donor
fluorescence lifetime in the absence and presence of acceptor.
Single-transfected protoplasts (donor only) are evaluated using
one-component analysis (see Note 22).
9. Double-transfected protoplasts (donor and acceptor) are ana-
lyzed based on a two-component model. The fluorescence life-
times and amplitudes obtained can be assigned to two donor
populations—one transferring energy to an adjacent acceptor
molecule resulting in a reduced fluorescence lifetime, and the
second showing no FRET and hence exhibiting fluorescence
decay kinetics as donor alone (see Note 23).
10. After FLIM analysis, a fluorescence intensity image and a false-
colored donor fluorescence lifetime image (see Fig 2a, c) are
shown. The color distribution of the FLIM image is automati-
cally set. To compare donor-alone versus donor-acceptor sam-
ples it is possible to set the color limits of fluorescence lifetime
(see Fig. 2c, d).
11. The nucleus of the fluorescence intensity image can be selected
using the “region of interest” (ROI) tool. The corresponding
fluorescence lifetime distribution will directly be adapted.
12. After performing the calculations, the mean fluorescence
lifetime, the distribution of the single-pixel values, as well as a
false-color-coded lifetime image will be displayed for the
selected ROI.
a b
2600 tau (ps) 3300

c d
2600 tau (ps) 3300

e
p-value 0.01 0.07 7.7E-7 0.47 4.7E-7 0.06
10
FRET Efficiency (%)
19 15 10 19 27 10 19 19
0
Donor: ARF5 ARF5 ARF5 ARF5 ARF5 ARF5 ARF3 ARF3
G279E G279I delIII/IV G279I
Acceptor: ARF5 ARF5 ARF5 ARF5 ARF5 ARF5 ARF3 YFP

G279A G279E G279I delIII/IV G279I
delIII/IV
Fig. 2 FRET-FLIM analysis of ARF5 homodimerization in plant protoplasts. Donor fluorescence lifetime images of
ARF5-sCFP3A alone (a) and ARF5-sCFP3A in the presence of ARF5-sYFP2 proteins (c). Histograms of fluorescence
lifetime values of the nucleus (b and d) show a change of color (from blue to green) when donor fluorescence
13. In order to determine if a particular protein couple is interacting,

three independent transfections of 10–15 samples of each con-
struct are measured and average fluorescence lifetimes are deter-
mined for all cells. In this example we show a histogram of the
FRET efficiencies of different combinations tested (see Fig 2e).
4 Notes
1. Over the years many variants of the cyan family were devel-
oped. For fluorescence lifetime analysis mTurquoise2 is the
most optimal fluorophore, as it shows the longest fluorescence
lifetime (see more information in [20]).
2. An extended protocol for ligation-independent cloning can be
found in Wendrich and coworkers [21].
3. Heat twice for 6 s at 300 W in a microwave to dissolve PEG.
4. FLIM can be performed on any confocal microscope (or mul-
tiphoton microscope). The only requirement is the photon
registration of the detector, which is coupled to a time-
correlated single-photon counting card.
5. The Time tape supports the top side of the leaf during manipu-
lation, while tearing off the 3 M Magic tape allows easy removal
of the lower epidermal layer and exposes mesophyll cells to cell
wall-digesting enzymes when the leaf is later incubated in an
enzyme solution. The protoplasts released into solution are
collected and washed for further use.
6. Incubation can be performed at lower temperatures but this
will reduce the enzyme activity and duration should be adapted.
7. MMg solution is stressful for protoplasts; therefore the trans-
fection procedure should take place quickly.
8. Use tips with enlarged openings for pipetting to reduce shear
forces.
9. In general, imaging can be performed about 16 h after trans-
fection but this depends on the protein expression. It is recom-
mended to check the expression levels at several time points to
determine the duration for optimal expression.
Fig. 2 (continued) lifetime is reduced. (e) Quantification of in vivo ARF dimerization measured by FRET-FLIM in
protoplasts. Protein interaction is expressed as average FRET efficiency (±SEM), number of protoplasts is indicated
in each bar, and the p-value for significance of difference between mutants and wild type is shown (two-sided
Student’s t test). Mutations within the dimerization domain of the ARF5 DBDs abolished the interactions, resulting
in FRET efficiencies around background level. The donor is indicated in green, and acceptor in red. The histogram
is reproduced from Ref. [13]
10. Pulsed diode lasers have the possibility to vary the repetition
rate of the laser, which is set depending on the fluorescence
lifetime. Rule of thumb is the fluorescence lifetime multiplied
by 5 to know what time window is appropriate.
11. Fluorophores absorbing in the range of 470–670 nm can be
excited using a super-continuum laser, which have typical exci-
tation pulses of 1 ps.
12. This can be achieved by selecting 488 nm laser, setting AOBS
in reflection mode, and selecting regular PMT.
13. In case filters are used, narrow band-pass filters should be used
to avoid contamination of acceptor signal in donor channel.
14. External hybrid (HyD) or MCP-PMT (Hamamatsu Photonics,
Japan) detectors can also be used for time-correlated single-
photon counting experiments. HyD detectors show high
quantum efficiencies, low dark noise, and large dynamic range
and have a typical time transient of 100 ps, which offers oppor-
tunities like fluorescence lifetime imaging, but can also be used
for single-molecule type of experiments like fluorescence cor-
relation spectroscopy. Hamamatsu R3809U MCP photomul-
tiplier suffers in quantum efficiency but exhibits best time
resolution (30 ps).
15. FLIM analysis is optimal at sufficient photon distribution per
pixel. Imaging at higher pixel resolution requires a prolonged
data acquisition time to ensure the detection of a statistically
relevant number of photons.
16. Time resolution of the ADC can be set ranging from 16 to
4096 channels. In case of low photon counts, the ADC can be
set to 64 channels.
17. Scanning should not be slower than 200 Hz and bidirectional
scanning is not appropriate.
18. TCSPC principle only holds when single photons are counted.
Higher count rates can cause so-called pile-up effects, which
affects fluorescence lifetime estimation.
19. In order to obtain information about the expression levels of
the sCFP3A- and sYFP2-tagged proteins a sequential confocal
image of donor and acceptor should be taken. Before a FLIM
measurement, confocal images of donor and acceptor should
be taken to select cells that show similar expression levels.
20. Other analysis software like TIMP developed by [22] or
FLIMfit [23] can be used.
21. In these studies, the effect of binning has been checked
between binning 0 (no pixel binning) and 1. No significant
difference of the average fluorescence lifetime distribution was
observed. However, image acquisition was taking five times
longer in case binning 0 was used for obtaining sufficient num-

ber of photons for data analysis.
22. Donor samples are also evaluated using two-component analy-
sis to check for false positives.
23. Using two-component analysis one can choose to fix the long
lifetime to value found for donor-only sample.
Acknowledgements
This study was supported by the Netherlands Organization for

Scientific Research (NWO; NWO-NSFC grant number 846.11.001
and ECHO grant 711.011.002 to D.W.). B.D.R. was funded by
the Netherlands Organization for Scientific Research (NWO; VIDI
864.13.001) and by The Research Foundation—Flanders (FWO;
Odysseus II G0D0515N and Post-doc grant 12D1815N). The
authors would like to thank Dr. S. Lindhoud for providing us with
Fig. 1. FRET-FLIM experiments were performed on a multimode
confocal microscope supported by an NWO Middelgroot
Investment Grant (721.011.004; J.W.B.).
References
1. Amos WB, White JG (2003) How the confo- 9. Borst J, Visser AJ (2010) Fluorescence lifetime
cal laser scanning microscope entered biologi- imaging microscopy in life sciences. Meas Sci
cal research. Biol Cell 95:335–342 Technol 21
2. Cox G, Sheppard CJ (2004) Practical limits of 10. Lakowicz JR (2006) Principles of fluorescence
resolution in confocal and non-linear micros- spectroscopy. Springer, New York, NY
copy. Microsc Res Tech 63:18–22 11. Valeur B (2002) Molecular fluorescence.
3. Betzig E, Patterson GH, Sougrat R, Lindwasser Principles and applications. Wiley-VCH,
OW, Olenych S, Bonifacino JS, Davidson MW, Weinheim
Lippincott-Schwartz J, Hess HF (2006) 12. Clegg RM (2009) Förster Resonance Energy
Imaging intracellular fluorescent proteins at Transfer – FRET what is it, why do it, and how
nanometer resolution. Science 313:1642–1645 it's done. In: Gadella T (ed) FRET and FLIM
4. Hess ST, Girirajan TP, Mason MD (2006) techniques. Elsevier Science, Amsterdam,
Ultra-high resolution imaging by fluorescence pp 1–57
photoactivation localization microscopy. 13. Boer DR, Freire-Rios A, van den Berg WA,
Biophys J 91:4258–4272 Saaki T, Manfield IW, Kepinski S, Lopez-
5. Bastiaens PI, Pepperkok R (2000) Observing Vidrieo I, Franco-Zorrilla JM, de Vries SC,
proteins in their natural habitat: the living cell. Solano R et al (2014) Structural basis for DNA
Trends Biochem Sci 25:631–637 binding specificity by the auxin-dependent
6. Bücherl CA, Bader A, Westphal AH, Laptenok ARF transcription factors. Cell 156:577–589
SP, Borst JW (2014) FRET-FLIM applications 14. Ljung K (2013) Auxin metabolism and
in plant systems. Protoplasma 251:383–394 homeostasis during plant development.
7. Clegg RM (1996) Fluorescence resonance Development 140:943–950
energy transfer. In: Herman XFWB (ed) 15. De Rybel B, van den Berg W, Lokerse A, Liao
Fluorescence imaging spectroscopy and CY, van Mourik H, Moller B, Peris CL, Weijers
microscopy. Wiley, New York, pp 179–252 D (2011) A versatile set of ligation-
8. Stryer L (1978) Fluorescence energy-transfer independent cloning vectors for functional
as a spectroscopic ruler. Annu Rev Biochem studies in plants. Plant Physiol
47:819–846 156:1292–1299
16. Sheen J (2001) Signal transduction in maize Weeren L, Gadella TW Jr, Royant A (2012)
and Arabidopsis mesophyll protoplasts. Plant Structure-guided evolution of cyan fluorescent
Physiol 127:1466–1475 proteins towards a quantum yield of 93%. Nat
17. Bücherl C, Aker J, de Vries S, Borst JW (2010) Commun 3:751
Probing protein-protein Interactions with 21. Wendrich JR, Liao CY, van den Berg WA, De
FRET-FLIM. Methods Mol Biol Rybel B, Weijers D (2015) Ligation-
655:389–399 independent cloning for plant research.
18. Wu FH, Shen SC, Lee LY, Lee SH, Chan MT, Methods Mol Biol 1284:421–431
Lin CS (2009) Tape-Arabidopsis Sandwich - a 22. Laptenok SP, Snellenburg JJ, Bucherl CA,
simpler Arabidopsis protoplast isolation Konrad KR, Borst JW (2014) Global analysis
method. Plant Methods 5:16 of FRET-FLIM data in live plant cells. Methods
19. Becker W, Bergmann A, Hink MA, Konig K, Mol Biol 1076:481–502
Benndorf K, Biskup C (2004) Fluorescence 23. Warren SC, Margineanu A, Alibhai D, Kelly
lifetime imaging by time-correlated single- DJ, Talbot C, Alexandrov Y, Munro I, Katan
photon counting. Microsc Res Tech M, Dunsby C, French PM (2013) Rapid
63:58–66 global fitting of large fluorescence lifetime
20. Goedhart J, von Stetten D, Noirclerc-Savoye imaging microscopy datasets. PLoS One
M, Lelimousin M, Joosen L, Hink MA, van 8:e70687
Chapter 14
In Vivo Identification of Plant Protein Complexes

Using IP-MS/MS
Jos R. Wendrich, Sjef Boeren, Barbara K. Möller, Dolf Weijers,
and Bert De Rybel
Abstract
Individual proteins often function as part of a protein complex. The identification of interacting proteins
is therefore vital to understand the biological role and function of the studied protein. Here we describe a
method for the in vivo identification of nuclear, cytoplasmic, and membrane-associated protein complexes
from plant tissues using a strategy of immunoprecipitation followed by tandem mass spectrometry. By
performing quantitative mass spectrometry measurements on biological triplicates, relative abundance of
proteins in GFP-tagged complexes compared to background controls can be statistically evaluated to iden-
tify high-confidence interactors. We detail the entire workflow of this approach.
Key words Protein–protein interaction, Arabidopsis, Immunoprecipitation, Tandem mass spectrometry,

Protein complex identification, GFP
1 Introduction
To understand a biological process, one can focus on the function

or activity of the individual proteins that are involved. However,
these proteins often perform their function as part of intricate
complexes, comprising several and/or different proteins bound
and functioning together. Therefore, being able to identify which
proteins interact with one another is a very important step towards
a thorough understanding of their biological function.
Several methods are currently used to test the interaction of
two (or more) selected proteins, including both in vitro and in vivo
techniques like yeast-two/three-hybrid (Y2H, Y3H; [1]), co-
immunoprecipitation (Co-IP) followed by Western blotting [2],
and Förster resonance energy transfer measured by fluorescence
lifetime imaging (FRET-FLIM) (also described in this issue by
Freire Rios et al.; [3]). The drawback of these types of techniques is
that they are limited to test the interaction of the selected proteins,
and often require prior knowledge regarding the interactions or
147
148 Jos R. Wendrich et al.
require all tested proteins to be labeled, limiting the amount of

interactions that can be examined and creating a bias in the mea-
surements. Regardless of this, techniques like FRET-FLIM are
powerful as a means to characterize individual interactions, as they
allow for testing of dynamic interactions at the subcellular level [3].
Here, we describe an optimized protocol for immunoprecipi-
tation (IP) followed by (label-free) tandem mass spectrometry
(MS/MS) that allows the identification of proteins interacting
with a fluorescently labeled bait in a non-biased and (semi-) quan-
titative fashion. This methodology has been successfully used to
identify complexes of multiple types of proteins, including tran-
scription factors and other nuclear proteins [4, 5], membrane-
associated proteins [6], trans-membrane proteins [6], and cytosolic
proteins [7]. When considering IP-MS/MS for the identification
of interacting proteins to a protein of interest, it is important to
keep in mind that this is not a saturated identification method. Due
to the finite amount of peptides that can be measured by the MS
per run, not all interacting proteins may be visible. Also some (pre-
viously known) interacting proteins may be absent from the results
list, due to very low abundance in the sample. It is, therefore, of
vital importance to keep contaminations to a minimum, as these
will fill the MS with unwanted peptides.
This step-by-step protocol for IP-MS/MS, using anti-GFP-
coated magnetic beads, describes the workflow that is used, from
collection of plant material up to the final data analysis (see Fig. 1),
and emphasizes important points for obtaining high purity and
efficiency.
2 Materials
2.1 Plant Material Homozygous transgenic lines should be generated, expressing the
protein of interest tagged with a GFP (or derivative) in the desired
tissue. Generally, 3 g of powdered material is used per replicate in the
IP. As a rule of thumb, for three replicates, 1 mL of seed is needed
when studying 6-day-old whole seedlings. For the control sample,
use either a non-transgenic line or a line expressing only GFP.
2.2 Immuno- 1. Liquid nitrogen.

precipitation 2. Mortar and pestle.
3. Standard 50 mL Falcon tubes.
4. Standard 14 mL Falcon tubes.
5. Extraction buffer without detergent (~60 mL is needed per
sample; EB−):
Tris–HCl, pH7.5 1 M (5 mL), NaCl, 5 M (3 mL), protease
inhibitor complete (two tablets), H2O to 100 mL.
IP-MS/MS for Identification of Protein Complexes in Plants 149
Fig. 1 IP-MS/MS workflow. The protocol starts with collecting plant material, then homogenizing it, and extract-
ing the proteins. The soluble fraction is then separated and immunoprecipitation is performed. Beads and
proteins are subsequently eluted and trypsin digestion of the proteins is done. After cleanup, label-free tandem
MS measurements are performed followed by MaxQuant protein identification, statistical analysis, and data
visualization
6. Extraction buffer with 1 % detergent (~10 mL needed per

sample; EB+):
Nonidet P-40; 0.1 mL, EB−; 9.9 mL.
7. Needle sonicator.
8. 50 mL Centrifuge tubes.
9. Beckman-Coultier Avanti J26 XP—JA 25.50 rotor.
10. Standard tabletop balance, 0.05 g accurate.
11. 40 μm Cell strainer.
12. Anti-GFP-coated magnetic beads (μMACS, Miltenyi).
13. Magnetic μ columns (μMACS, Miltenyi).
14. MultiStand (μMACS, Miltenyi).
15. μMACS Separator (Miltenyi).
16. ABC buffer: 50 mM NH4HCO3 in H2O.
17. Thermoblock.
18. 0.5 mL Protein low-bind tubes.
2.3 Sample 1. ABC buffer: 50 mM NH4HCO3 in H2O.

Preparation Mass 2. Dithiothreitol (DTT).
Spectroscopy
3. Iodoacetamide (IAA).
4. l-Cysteine.
5. Trypsin, sequencing grade (0.5 μg/μL in 1 mM HCl).
6. Trifluoroacetic acid (TFA).
7. Thermoblock.
8. C18 Empore membrane.
9. 1 mm Tissue puncher.
10. Methanol.
11. LiChroprep RP-18.
12. Formic acid.
13. Acetonitrile (AcNi).
14. 0.5 mL Protein low-bind tubes.
15. SpeedVac.
16. Water bath sonicator.
17. Standard tabletop centrifuge.
3 Methods
3.1 Immuno- Day 1:

precipitation
1. Thoroughly grind plant material (both transgenic line and the
control) in liquid nitrogen with mortar and pestle. Transfer the
ground material to a precooled 50 mL tube and place it in
liquid nitrogen.
2. For each of the replicates, weigh the desired amount of plant
powder on a scale into a precooled tube (see Note 1).
3. Store at −80 °C until next day (see Note 2).
Day 2:
1. Precool mortars on ice and prepare extraction buffers.
2. Add weighed material to mortars and add as little EB+ as pos-
sible to solubilize the plant material (see Note 3) and grind
again very thoroughly with mortar and pestle.
3. Transfer the protein extract to a 14 mL tube (if you have more
than 3 mL protein extract, divide over several 14 mL tubes)
and sonicate three times for 15 s on ice with at least 15 s pause
in between (see Note 4).
4. Keep the samples on ice for 30 min to thoroughly extract the
proteins (see Note 5).
5. Dilute the protein extract to 0.2 % NP40 by adding EB− and
transfer the protein extract to centrifuge tubes (see Note 6).
6. Centrifuge for 15 min at 4 °C at 18,000 rpm/39.000 g (see

Note 7). Transfer the supernatant to another centrifuge tube
and centrifuge again for 15 min at 4 °C at 18,000 rpm/39.000 g.
7. Transfer the supernatant through a 40 μm cell strainer into a
new 50 mL tube (see Note 8).
8. Add 50–100 μL anti-GFP μBeads (see Note 9). Mix well by
swirling the tube. Rotate for 2 h at 4 °C (see Note 10).
9. Place μColumn in the magnetic field of the μMACS Separator.
10. Add 200 μL clean EB+ to the column.
11. Prepare EB with 0.1 % NP40.
12. Add 500 μL EB with 0.1 % NP40 to the column.
13. Apply cell lysate onto the column and let the lysate run through
(see Note 11).
14. Prepare ABC buffer and heat 1 mL to 95 °C in a heat block for
later use.
15. Rinse column four times with 200 μL EB with 0.1 % NP40.
16. Rinse column twice with 500 μL ABC buffer to remove all
detergent (see Note 12).
17. Remove the column from the magnet, immediately place a
0.5 mL low-bind Eppendorf tube below the column, and add
50 μL preheated ABC buffer to the column to transfer the
beads into the Eppendorf tube (see Note 13).
18. Transfer column into a new low-bind tube and add another
50 μL preheated ABC buffer to the column (second elution
with ABC, see Note 14) to check whether all beads were trans-
ferred in step 17. Combine with the beads in step 17 when
beads are visible.
19. Store at −20 °C.
3.2 Sample Day 3:

Preparation Mass
1. Add 1 μL 500 mM DTT in ABC buffer to the beads contain-
Spectrometry
ing eluate and incubate for 2 h at 60 °C (see Note 15).
2. Add 1 μL 750 mM IAA in ABC buffer to the samples. Mix well
and incubate for 2 h at room temperature in the dark (see Note 16).
3. Add 1 μL 200 mM l-cysteine in ABC buffer to the samples (see
Note 17). Mix well.
4. Add 1 μL sequence-grade trypsin, mix well, and incubate for
16 h at 20 °C (see Note 18).
Day 4:
1. Using 10 % TFA, adjust the pH of the samples to approxi-
mately 3; check the pH with 0.1 μL of sample on a piece of pH
paper (see Note 19).
2. Use a tissue puncher to cut a small (~1 mm) piece of a C18

Empore membrane and insert it into a polypropylene 200 μL
pipet tip [8].
3. Add 200 μL methanol to the tip with membrane (see Note 20).
4. Prepare a 50 % (v/v) slurry of LiChroprep RP-18 column
material in methanol and apply 4 μL of the slurry to the tip/
μColumn in the methanol.
5. Elute the μColumn and wash once with 100 μL methanol (see
Note 21).
6. Equilibrate the μColumn once with 100 μL 0.1 % formic acid
(see Note 22).
7. Spin down the beads at full speed for 15 min in a standard
tabletop centrifuge.
8. Add sample to the μColumn and elute through (see Note 23).
9. Wash the μColumn once with 100 μL 0.1 % formic acid.
10. Transfer the μColumn to a 0.5 mL protein low-bind tube.
11. Add 50 μL AcNi / 0.1 % formic acid (1:1) and manually elute
the μColumn directly into the 0.5 mL protein low-bind tube
(see Note 24).
12. For LCMS analysis, reduce the AcNi content by putting the
samples with open cap in a SpeedVac at 30–45 °C for ~2 h.
The final volume should be well below 10 μL. Adjust the sam-
ple volume to 50 μL with 0.1 % formic acid (see Note 25).
13. The sample is now ready to be loaded onto an LCMS and can
be stored at −20 °C.
3.3 Tandem Mass 1. Separate peptides by reversed-phase nano liquid chromatogra-

Spectrometry phy connected to the MS via an electrospray interface.
2. Measure MS spectra online by accurate mass spectrometry
(MS) as well as collision-induced dissociation fragmented
MS-MS spectra according to the settings in Tables 1 and 2 (see
Notes 26 and 27).
3.4 MaxQuant 1. Analyze all LC-MS/MS runs obtained with all MS/MS spec-
Protein Identification tra with the database search algorithm Andromeda from
and Statistical MaxQuant 1.3.0.5 [9, 10] in the “Label Free Quantification”
Analysis mode with settings described in Table 3 (see Note 28).
2. Normalized quantitative information (Label Free Quantification
[LFQ] intensity) is obtained for all protein groups identified when
two or more peptides could be properly quantified (see Note 29).
3. Group the replicates together. An optional final filtering can be
performed in the Perseus filtering and statistics software to
improve the statistics by deleting those proteins that are identi-
fied in less than half of the replicates (see Table 4 and Note 30).
Table 1
Chromatography settings
Nano LC Proxeon EASY nLC II

Setup Vented column
Pre-concentration column 0.10 × 32 mm Magic C18AQ 200A 5 μm beads prepared in-house
Analytical column 0.10 × 250 mm Magic C18AQ 200A 3 μm beads prepared in-house
Column temperature Room temperature
Autosampler temperature 7 °C
Sample loading volume and 30 μL at 270 bar (generally 7–8 μL/min) with eluent A
flow
Autosampler wash 2 cycles with 27 μL of 50 % (v/v) acetonitrile/50 % (v/v) 0.1 % formic
acid in water, 100 μL of 0.1 % formic acid in water
Eluent A 5 mL/L acetic acid in LCMS-grade water
Eluent B 5 mL/L acetic acid in LCMS-grade acetonitrile
Measurement gradient (time 0 (8), 50 (33), 53 (50), 58 (50)
in min (% eluent B))
Cleaning gradient (time 0 (20), 10 (80), 15 (80). Run at least one cleaning gradient after each
in min (% eluent B)) measurement gradient
Injection volume 18 μL of sample (measurement gradient) or 15 μL of 50 % (v/v)
acetonitrile/50 % (v/v) 0.1 % formic acid in water
High-voltage connection Apply an electrospray potential of 3.5 kV directly to the eluent via the
stainless steel needle fitted into the vented waste line of a P777
Upchurch microcross that was positioned between the pre-
concentration and analytical column
Pre-column re-equilibration 12 μL Eluent A at 270 bar
Analytical column 3 μL Eluent A at 270 bar
re-equilibration
4. Perform T-tests between groups in Perseus. This not only

results in the p-value (column: -Log t-test p-value), but also
immediately yields the ratio of the group’s average LFQ values
(column: t-test Difference) when applied to two groups.
Significance at a variable false discovery rate (FDR) is judged
by both p-value and ratio, of which the weight is set by a vari-
able S0 value (see Note 31). Both FDR and S0 should be opti-
mized to yield no or only a few proteins significantly different
on the control side of the volcano plot obtained (see Fig. 2).
5. In a successful experiment, the GFP and the bait (tagged) pro-
tein are in the top 10 when the list is sorted by ratio (see Note
32). The high ratio and low p-value should make the differ-
ence between sample and control clearly significant. Depending
Table 2
Mass spectrometry settings
MS LTQ-Orbitrap XL
Measurement Xcalibur 2.1
software
Tuning and From m/z 150–2000 by direct infusion of LTQ ESI calibration solution for
calibration positive mode containing caffeine, peptide MRFA, and Ultramark 1621
according to instructions of the manufacturer
nESI Capillary temperature = 200 °C, capillary voltage =35 V, tube lens voltage = 100 V,
source voltage = 3.5 kV (see high-voltage connection in Table 1)
Tune file FTMS Full Microscans = 1, Ion Trap MSn Microscans = 1, FTMS Full Max Ion
parameters Time = 500 ms, Ion Trap MSn Max Ion Time = 100 ms, FTMS Full AGC
Target = 1,000,000, Ion Trap MSn AGC Target = 10,000
MS detector 58
acquire time
(min)
Start delay (min) 0
Segments 1
Scan events 2
Event 1 Analyzer = FTMS, Mass range = Normal, Scan Range = 380-1400 m/z,
Resolution = 60.000, Scan type = Full, Polarity = Positive, Data type = Profile
Event 2 (data Analyzer = Ion Trap, Mass range = Normal, Scan Rate = Normal, Data
dependent) type = Centroid, Dependent scan On
Lock mass Disabled (FTMS is calibrated every day)
Global dynamic Repeat Count = 1, Repeat duration = 0 s, Exclusion list size = 500, Exclusion
exclusion duration = 45 s, Exclusion mass width Relative to reference mass = 25 ppm (low,
and high)
Early expiration = disabled
Segment Preview mode for FTMS master scans = Enabled, Charge state screening and
Rejection = Enabled, Monoisotopic precursor selection = Enabled, Charge state
+1 and 4 and up = Rejected, Unassigned charge state = Rejected
Scan event Minimum MS signal Threshold for MS2 trigger = 5000, Mass determined from
scan event = 1, Nth most intense ion = Enabled, Analyze top N peaks = 4,
Activation type = CID, Activation Default charge state = 3, Activation Isolation
width (m/z) = 2.0, Normalized collision energy = 30, Activation Q = 0.250,
Activation time (ms) = 15
Mass lists and NOT used
global mass lists
on the starting tissue that is used, some highly abundant pro-

teins like glyceraldehyde-3-phosphate dehydrogenase and
Rubisco can always be observed, but these should give ratios
around one (=0 on a logarithmic scale).
Table 3
Database search and quantification software settings
Software MaxQuant 1.3.0.5

Group-specific Variable modifications = Oxidation (M), Acetyl (Protein N-term), Deamidation (NQ),
parameters Separate variable modifications for first search = disabled, Multiplicity = 1,
Enzyme = Trypsin/P, First search ppm = 20, Main search ppm = 6, Max number of
modifications per peptide = 5, Max. missed cleavages = 2, Max. charge = 7, individual
peptide mass tolerances = Enabled, Type = Standard, Separate enzyme for first
search = disabled
MS/MS ITMS MS/MS tolerance = 0.5 Da, Include contaminants = Enabled
sequences FASTA files = Arabidopsis thaliana Uniprot reference database, Reverse, Special
AAs = KR, Separate file for first search = a contaminants database with 60 proteins
including GFP/YFP/CFP/eYFP and BSA (P02769, bovin serum albumin
precursor), Trypsin (P00760, bovin), Trypsin (P00761, porcin), Keratin K22E
(P35908, human), Keratin K1C9 (P35527, human), Keratin K2C1 (P04264,
human) and Keratin K1CI (P35527, human)
Fixed modifications = Carbamidomethyl (C)
Identification All default with deamidation (NQ) added to the protein quantification peptide list
and Experimental design file should be made and selected
quantification
Misc. Match between runs = Enabled (2 min), Label-free quantification = Enabled (LFQ
min ratio count = 2), iBAQ and Log fit = Enabled
Table 4
Filtering and statistics on identified protein groups
Software Perseus
Selectively load Expression: all”LFQ_name” rows, Categorical annotation: Only identified by
data from the site, Reverse, Contaminant, Textual annotation: Protein IDs, Majority protein
result file IDs, Protein names, Gene names, Fasta headers, (remove “Proteins”); Numerical
annotation: id, Proteins, Peptides, Unique Peptides, iBAQ and any other info you
may need like: sequence coverage (%), Mol. Weight [kDa], Sequence length and/
or PEP score
Extra filtering Filter out “Reverse” and “Only identified by site” proteins. “Only identified by
site” marks those proteins that have been identified by modified peptides only
Filter out proteins identified by only one peptide or no unique peptide
Logarithmic Transform the LFQ and iBAQ to its logarithmic values
transformation
Grouping of Group biological replicate measurements
samples
Optional extra Filter out those protein groups that have less than two LFQ values in at least one
filtering 2 group
Replace NaNs Replace/imputate missing values by a constant that is slightly lower than the
lowest (Log) value measured. This is done to make sensible ratio calculation
possible
T-test Perform a first T-test with FDR = 0.01 and S0 = 1
12
10
-LOG T-test p-value
8
Bait
GFP
0
-4 -3 -2 -1 0 1 2 3 4
LOG protein abundance ratio (sample/control)
Fig. 2 Volcano plot example. Volcano plot showing all identified proteins after filtering and statistical analysis,
with their corresponding protein abundance ratios over the T-test p-value. Depicted in red are proteins that are
significantly different from the control, with the bait and GFP among the top hit list
4 Notes
1. Usually about 1–3 g of ground plant material is used per repli-

cate/sample (for low-expressed proteins use 3 g of tissue and
use only tissue where your protein is expressed to limit the
amount of background proteins in the sample).
2. This is recommended because it ensures that the mortars are
not too cold, which in some cases can lead to complications
with fluidity of the sample.
3. Usually ~8–9 mL of EB+ is needed for 3 g material. Using less
of the buffer in this step will yield more concentrated sample
later in the protocol.
4. Sonication is used to break up cell membranes and release cel-
lular components in the liquid. For the described sonicator,
use 3/4 power and middle tune. Optimization might be
required depending on the sonicator. It is very important to
keep the samples on ice the whole time, especially during soni-
cation, to prevent complex deformation through heat.
5. If desired, a small aliquot can be taken as INPUT sample for
Western blotting.
6. Too high detergent concentrations may increase background
proteins and disrupt MS measurements.
7. Make sure that the tubes are balanced properly (±0.05 g).
8. This ensures that any leftover cell debris is filtered out, as these
could clog the IP μColumns. If desired, a small aliquot can be
taken as supernatant sample for Western blotting.
9. Use 50 μL when 1 g of material is used and 100 μL when 3 g

of material is used. Make sure to resuspend the beads before
pipetting with a cut tip.
10. In some cases it can be advisable to speed up the whole proto-
col. This incubation step can then be decreased to 1 h.
11. The columns are “flow-stop” and do not run dry. If desired, an
aliquot of the flow-through can be taken as sample for Western
blotting.
12. Any detergent present in the final MS sample may interfere
with the chromatography and the MS measurement, as this
may cause peak broadening.
13. Make sure that the “end” of the column is always in contact
with the eluate. This way more beads will be eluted.
14. If desired, a small aliquot can be taken as IP sample for Western
blotting.
15. DTT reduces disulfide bonds in the proteins, leaving them
more exposed for the trypsin digestion later in the protocol.
16. IAA alkylates the proteins, blocking disulfide formation.
17. l-Cysteine stops the alkylation.
18. Do not leave this longer than 16 h, as this may result in chy-
motrypsinic cleavages.
19. Lowering the pH inactivates trypsin, stopping digestion. Avoid
a pH of lower than 2.
20. If some air is trapped between the liquid and the frit/column
material, flick the μColumn to remove.
21. The prepared μColumns can be eluted by hand using a 1 mL
syringe with a rubber-lined tip or using a vacuum pump.
Whatever method used, do not let the columns run dry.
22. Methanol concentration should be brought below 5 % to allow
binding of peptides.
23. Peptides in the sample will bind the column material and deter-
gent and other contaminants will be washed out.
24. Peptides are released from the column material in acetonitrile con-
centrations above 5 %; 50 % is used to limit yield losses in this step.
25. If samples are completely dry, resuspend in 50 μL 0.1 % formic
acid and sonicate several times for 10 s, vortexing in between.
26. A high chromatographic resolution as well as high MS resolu-
tion, accuracy, and sensitivity are required to successfully mea-
sure transcription factor interactors.
27. Similar results can be obtained using instruments from suppli-
ers other than mentioned in the tables when they have com-
parable or improved specifications like some more modern
instruments do.
28. MaxQuant internally re-calibrates MS spectra resulting in MS

deviations which generally lie below 2 ppm. The advantage of
using the “Match between runs” option is that the software
will search for all peptides identified in all MS runs based on
retention time, m/z measured, and isotopic distribution. This
way, quantitative information can be obtained for all peptides
identified in all runs even when the peptide may only have
been selected for fragmentation in a single run.
29. By default a FDR of 1 % is used.
30. This will restrict the results table to only those proteins with a
high reliability.
31. Significant FDR can be set manually. S0 values <1 give more
weight to the p-value and S0 values >1 give more weight to the
ratio, when determining the significance. Generally an S0 of 1
is chosen to give both p-value and ratio an equal weight.
32. Sometimes, GFP is found somewhat lower in the list, probably
due to its compact folding, making it more difficult to digest
with trypsin.
References
1. Ferro E, Trabalzini L (2013) The yeast two- 6. Zwiewka M, Feraru E, Möller B, Hwang I,
hybrid and related methods as powerful tools Feraru MI, Kleine-Vehn J, Weijers D, Friml
to study plant cell signalling. Plant Mol Biol J (2011) The AP-3 adaptor complex is required
83:287–301 for vacuolar function in Arabidopsis. Cell Res
2. Albrecht C, Boutrot F, Segonzac C, 21:1711–1722
Schwessinger B, Gimenez-Ibanez S, Chinchilla 7. Smaczniak C, Li N, Boeren S, America T, van
D, Rathjen JP, de Vries SC, Zipfel C (2012) Dongen W, Goerdayal SS, de Vries S, Angenent
Brassinosteroids inhibit pathogen-associated GC, Kaufmann K (2012) Proteomics-based
molecular pattern-triggered immune signaling identification of low-abundance signaling and
independent of the receptor kinase BAK1. regulatory protein complexes in native plant
Proc Natl Acad Sci U S A 109:303–308 tissues. Nat Protoc 7:2144–2158
3. Bücherl CA, Bader A, Westphal AH, Laptenok 8. Rappsilber J, Mann M, Ishihama Y (2007)
SP, Borst JW (2014) FRET-FLIM applications Protocol for micro-purification, enrichment,
in plant systems. Protoplasma 251:383–394 pre-fractionation and storage of peptides for
4. De Rybel B, Möller B, Yoshida S, Grabowicz I, proteomics using StageTips. Nat Protoc
Barbier de Reuille P, Boeren S, Smith RS, 2:1896–1906
Borst JW, Weijers D (2013) A bHLH complex 9. Cox J, Mann M (2008) MaxQuant enables
controls embryonic vascular tissue establish- high peptide identification rates, individualized
ment and indeterminate growth in Arabidopsis. p.p.b.-range mass accuracies and proteome-
Dev Cell 24:426–437 wide protein quantification. Nat Biotechnol
5. Saiga S, Möller B, Watanabe-Taneda A, Abe M, 26:1367–1372
Weijers D, Komeda Y (2012) Control of embry- 10. Cox J, Neuhauser N, Michalski A, Scheltema RA,
onic meristem initiation in Arabidopsis by Olsen JV, Mann M (2011) Andromeda: a peptide
PHD-finger protein complexes. Development search engine integrated into the MaxQuant
139:1391–1398 environment. J Proteome Res 10:1794–1805
Chapter 15
Assaying Auxin Receptor Activity Using SPR Assays

with F-Box Proteins and Aux/IAA Degrons
Mussa Quareshy, Veselina Uzunova, Justyna M. Prusinska,
and Richard M. Napier
Abstract
The identification of TIR1 as an auxin receptor combined with advanced biophysical instrumentation has
led to the development of real-time activity assays for auxins. Traditionally, molecules have been assessed
for auxinic activity using bioassays, and agrochemical compound discovery continues to be based on “spray
and pray” technologies. Here, we describe the methodology behind an SPR-based assay that uses TIR1
and related F-box proteins with surface plasmon resonance spectrometry for rapid compound screening.
In addition, methods for collecting kinetic binding data and data processing are given so that they may
support programs for rational design of novel auxin ligands.
Key words Surface plasmon resonance (SPR), Biacore, F-box proteins, Auxin binding, Compound
screen, Kinetics
1 Introduction
The study of hormone-active proteins requires robust assays for

binding. Most binding assays are developed for receptor candi-
dates, but both enzymes and transport proteins also carry binding
sites specific for their substrates. Historically, the basis of most
analyses was radiolabeled ligand-binding assays. More recently a
range of biophysical techniques has offered new, label-free tools for
binding analysis and while no single technique can address all
experimental systems, surface plasmon resonance (SPR) has found
widespread utility. It yields definitive kinetic data for binding inter-
actions and may also be used for thermodynamics and high-
throughput compound screening. SPR requires that one partner is
immobilized on a chip surface which can be a limiting factor, but
in most cases the variety of available chip surface chemistries allows
binding reactions to be measured by immobilizing either receptor
or ligand. Ideally kinetics are recorded in both orientations.
159
160 Mussa Quareshy et al.
The most revealing kinetic parameter sought from binding assays

is the affinity of the site for its ligands. Let affinity be KD, the dissocia-
tion constant for the interaction under the stated conditions. The
units are molar and KD is the concentration of ligand at which the
receptor-binding sites are at half occupancy. Radiolabel assays in vari-
ous formats generally yield the affinity. For example, most binding
analyses on auxin-binding protein 1 (ABP1) used radiolabel displace-
ment assays [1], competing radiolabel off the binding site with increas-
ing concentrations of unlabeled auxin to give KD from a Scatchard
analysis. However, label-free assays like SPR are able to provide addi-
tional valuable information. How much information depends some-
what on the availability of pure protein. For example, SPR typically
requires micrograms of protein and can provide binding on- and off-
rates as well as KD, with the off-rate indicating the durability of the
interaction-a useful feature in ligand screening. SPR can also be used
for determining the thermodynamics of binding, measuring changes
in free energy, enthalpy and entropy, as well as affinity to reveal aspects
of the mechanism of binding. We will explain how this experiment is
run, but with more protein available (milligrams) isothermal titration
calorimetry (ITC) will give more definitive thermodynamic values
given that the ITC reaction is entirely in solution.
Many candidate auxin receptors have been listed [1, 2] and
two have been characterized extensively, ABP1, and transport
inhibitor resistant 1 (TIR1) which mediates auxin-inducible
changes in transcription [3, 4]. Both are soluble proteins and
structures have been determined for each, with binding sites empty
and with auxin bound [5, 6]. Structural data are invaluable in
terms of mechanistic insights, but crystallography offers only a
snapshot of a binding interaction taken under extreme conditions.
We describe here an SPR assay developed for TIR1 (and its homo-
log AFB5) [7] in which binding is measured in real time. The assay
not only gives us kinetic data on auxin binding, but it also provides
a platform for novel auxin and anti-auxin discovery.
A schematic of the cascade of interactions involved in TIR1-
mediated auxin perception in vivo is shown in Fig. 1 (see [8] for
more details). The SPR assay mimics this by using the interacting
peptide from the Aux/IAA degron as the “bait” molecule on the
SPR chip surface. Purified TIR1 protein is then passed over the
“bait” with or without auxin. In the presence of an active auxin
TIR1 will bind and the SPR technique interprets this increase in
mass on the chip surface to give a binding sensorgram (Fig. 2).
2 Materials
1. Biacore 2000
2. Maintenance Chip and SA Sensor Chip (streptavidin coated
for coating with biotinylated molecules) (GE Healthcare BR
1006-51, 1000-32).
Assaying a Auxin Binding by SPR 161
Fig. 1 Schematic representation of the cascade of events during auxin perception by TIR1
3. Plastic vials (7 mm, 0.8 mL) with rubber caps, type 3 (GE
Healthcare BR 1002-12, 1005-02).
4. Desorb kit consisting of BIAdesorb solutions 1 and 2 (GE
Healthcare BR 1008-23).
5. Chip cleaning solution: 1 M NaCl, 50 mM NaOH.
6. Surfactant P20 10 % v/v (GE Healthcare BR 1000-54).
7. SPR rinse solution, filtered (0.2 μm) and degassed: 0.05 % v/v
Surfactant P20.
8. SPR buffer, filtered and degassed: 10 mM HEPES pH 7.4,
150 mM NaCl, 3 mM EDTA, 50 μM phytic acid, 1 mM TCEP,
0.05 % v/v Surfactant P20 (see Note 1).
Fig. 2 Experimental setup for the SPR assay based on initial events in the TIR1 binding. A peptide which includes
the Aux/IAA degron motif is immobilized on the SPR chip surface. The peptide is synthesized as a biotin conjugate.
Purified TIR1 is the mobile partner flowing over the chip surface. Where appropriate, auxin is added to the TIR1
protein solution before being passed over the chip. The TIR1-auxin complex then binds to the degron and this
increased mass on the chip surface is recorded in a binding isotherm and presented as a sensorgram
9. Biocytin, 1 mg/mL stock solution in water (stored at −20 °C

in 100 μL aliquots) (Sigma-Aldrich B4261).
10. Biotinylated peptide, 1 mg/mL stock solution in water (stored
at −20 °C in 5 μL aliquots) (our standard is the sequence
derived from the AXR3 (Aux/IAA7) degron, biot-
QVVGWPPVRNYRK (Thermo Fisher Scientific)).
11. Indole-3-acetic acid (IAA): 100 mM stock solution and 10 mM
working solution in dimethyl sulfoxide (DMSO), stored at −20 °C.
12. TIR1/ASK1 purified protein complex in SPR buffer.
3 Methods
The methods discussed below relate to Biacore 2000 and 3000

instruments, but may be adapted for other SPR platforms. The
control software opens as shown below:
From the menu bar Command, Run and Tools are used most
commonly. The software guides users through all the routines and
gives instructions for sample placement, volume requirements, etc.
We recommend that new Biacore users take training on the instru-

ment before their first experiment. However, the instructions given
by the software are comprehensive. For more details see the Biacore
2000 Instrument Handbook.
3.1 Chip Preparation Prior to coating a new sensor chip (or coating additional channels
on a part-used chip), always run a Desorb routine to clean the fluid
lines. Desorb is run using a Maintenance Chip.
1. Place the buffer inlet tubes into a bottle of SPR rinse
solution.
2. Command -> Undock; wait until the undocking routine com-
pletes (the countdown time is shown, see below). Place a
maintenance chip in the holder. Command -> Dock; wait for
docking to complete.
3. You will be asked if you want to Prime the system. Accept the
Prime option. Read the notes in the dialog box.
4. Tools -> Working Tools -> Desorb. Read the warnings dis-
played. You need 3 mL of Desorb solutions 1 and 2 placed in
the nominated rack positions. -> Start.
5. As in 2, undock to remove the maintenance chip and dock a

new SA chip.
6. Replace SPR rinse solution with SPR buffer. As above, select
Prime (see Note 2).
7. Perform a chip conditioning routine using Manual mode and
Chip cleaning solution (see Note 3). Select Run -> Run sens-
orgram. A window entitled Detection opens and prompts a
choice of flow cells. Choose the relevant cells and hit OK (see
Note 4).
8. A Flow dialog box appears. Input a flow rate of 20 μL/min. ->

OK. You will be prompted to set up a .blr file where the results
are saved (see Note 5). As soon as you hit Save the main screen
refreshes (as shown) giving a readout for each channel selected
in 7 (see Note 6). The control buttons for manual injection are
now active and a Command Queue dialog box opens.
9. Set up three consecutive 1-min injections of Cleaning Solution

A (see Note 7). Select -> Inject; a dialog box opens. Position
asks you to nominate the rack position for the cleaning solu-
tion vial. Volume -> 20 μL is recommended. The volume
entered determines the contact time for a given flow rate.
Consumption (the volume of conditioning solution used) is
calculated. This includes system dead volumes plus the 20 μL
wash volume.
10. Hit -> Start Injection. You will be shown the Command Queue.
While the first command is executed (indicated by the cogwheel
icon), add two identical injections to the command queue.
11. After the last injection finishes, hit -> Stop Sensorgram. The
queue is cleaned.
Hit -> OK. An example sensorgram is shown in Fig. 3.
3.2 Coating a Chip The recommended experimental design is to use flow cell 1 as a ref-
with Degron Peptide erence. We will describe a method using flow cell 2 for biotinylated
degron peptide, keeping flow cells 3 and 4 in reserve (see Note 8).
Coating is best run using the wizard, but may also be done with
Manual Injections. You will need to consider surface coating density
for kinetic experiments (see Note 9). Here we give conditions for
high coating densities which are suitable for initial screening work.
1. Run -> Run Application Wizard. Select -> Surface Preparation
-> Start.
2. Select -> Immobilization -> Next.
Fig. 3 Manual sensorgram collected during chip surface cleaning. Three 1-min injections are represented by
three consecutive peaks. Data are shown for flow cell 1 only. You may also visualize sensorgrams for all chan-
nels simultaneously using the menu drop at the top right of the window
3. Sensor chip, select -> SA -> Next.
4. Enter the Ligand Name for each cell to be coated (see Note 10).
5. Enter Injection Time and Flow rate as shown. -> Next.
6. The consumption of solutions is calculated and shown in the

next dialog box. Prepare the appropriate volumes of biocytin
and biotinylated peptide (at 0.05 mg/mL, stock diluted 1 to
20 in water) and place them into the rack positions indicated.
-> Next.
7. Provide a filename when prompted. -> Save. The immobiliza-

tion routine starts.
8. Sensorgrams are shown during the immobilization run (Fig. 4).
9. When finished, a Results box records the amount of ligand
bound for each flow cell/ligand in RUs (see Note 11).
The bound RU represents the difference between the baselines

before and after each immobilization. The report points are marked
by an x (Fig. 4). Cells 1 and 2 are now ready for use.
Fig. 4 Example sensorgrams for automated SA chip coating. For both flow cells it is clear that the signal does
not return to baseline after the final wash step, but reaches a new steady level to show that the surface was
successfully coated. The sensorgram for our degron peptide is typically “wavy”
3.3 TIR1 The purpose of the assay is to test protein for auxin binding activ-
Activity Assay ity. This assay is a simple and effective procedure to confirm
successful purification of active TIR1. Some typical analytical
experiments are described below.
3.3.1 Sample TIR1 protein is co-expressed with ASK1 in insect cells using the
Preparation baculovirus system [7, 9] and purified using affinity chromatogra-
phy. In brief, cells 3 days post-infection are harvested by centrifu-
gation. Cells are lysed, the lysate cleared, and proteins purified by
nickel affinity chromatography. The eluted protein is incubated
overnight with TEV protease (to remove His-MBP fusion tags)

before affinity capture of TIR1 (and associated ASK1) by FLAG
affinity chromatography. Elution is in SPR buffer plus FLAG
peptide.
1. In three labeled vials prepare the following mixes (see Note 12):
Volumes (μL)
Sample IAAa DMSO SPR buffer TIR1b Final volume (μL)
Buffer blank 0 0 400 0 400
TIR1 + IAA 0.5 0 49.5 50 100
TIR1 − IAA 0 0.5 49.5 50 100
a
From a 10 mM DMSO working stock this gives 50 μM IAA and DMSO is diluted
to 0.5 %.
b
TIR1 concentration does not need to be measured at this stage because we are
looking for yes/no answers.
2. Mix samples thoroughly by gentle vortexing, cap, and keep

on ice.
3.3.2 SPR Setup 1. Cease Standby mode -> Stop.

2. If necessary, Dock your SA chip and Prime.
3. Run -> Run Application Wizard. Select -> Binding Analysis ->
Start.
4. Select -> Direct Binding, as we have already prepared the chip.

-> Next.
5. Use Flow Cell(s) -> “2 with 1 as reference.”

6. Flow Rate -> 20 μL/min. The rest of the parameters are left as
defaults. -> Next.
7. Ensure that Run Order is “As Entered” (see Note 13).

8. You need to specify how many times each sample is assayed. In
the column Repl., enter 1 against each sample except the buf-
fer blank (see Note 14), as shown.
-> Next.
After each binding cycle the coated surfaces need to be cleared

of bound analyte, a process referred to as Regeneration. We
use 50 mM NaOH. Use default values for the regeneration
routine.
Regeneration Method -> Single Injection,
Flow Rate - 30 μL/min,
Solution - 50 mM NaOH,
Injection Time - 30 s.
-> Next.
9. Rack Positions for all sample tubes are pre-selected, but you
can change them by drag-and-dropping onto the desired posi-
tions. Make sure that vial contents correspond to positions. ->
Next.
10. This is your last chance to check the vial contents, volumes,
and positions.
Ensure the “Standby Flow After Run” box is checked (this is
the default).
Click -> Start, and enter a filename when prompted.
11. At the end of the run you will see a data summary box with a
set of tabs. Inspect these (especially the baseline drift report),
but you will get a better picture of your data by opening the
.blr file in BIAevaluation software (see Subheading 3.3.3
below).
12. The temperature of the chip and binding experiment will be
25 °C as default. The chip temperature is shown in the lower
right window of the Control software at all times. Each analyte
is warmed to the chip temperature as it passes through the
microfluidic cartridge (see Note 15).
3.3.3 Data Processing Data processing has several goals, including normalization, removal
of anomalies (air bubble spikes, etc.), and baseline adjustment.
Later, additional manipulations are needed for successful kinetic
analysis. Open the BIAevaluation package. This can also be started
from inside BIACORE 2000 Control Software by hitting on the
abacus icon. For more details see the BIAevaluation Software
Handbook.
1. File -> Open, and then navigate to your results file. A dialog
box will enquire which of the collected data you want to pro-
cess and display. We want to see the binding with respect to
flow cell 1.
Open -> Curves only
Value (lower right) -> click on the “2 - 1.” The data files from
all individual flow cells are deselected. You see only Fc2 - Fc1.
Curves -> use the mouse and Ctrl to select all you want to
consider, click -> OK.
2. You will see in the Project window the names, IDs, and Sources
of the data you are going to process.
3. Each curve is given an automatic name consisting of the filename

of the results file, the cell selection (i.e., Fc=2-1), and the order
in which they were run. To edit the name and display or change
features of a binding curve, double left click on the respective
title and work within the Curve Properties dialog box.
The end result will appear as shown below.
4. Select the curves [TIR1 + IAA], [TIR1 − IAA] and one of the
Buffer traces and click on the graph icon button. The graphs
that will open would look as follows. The axes auto-scale and
spikes from injection start and stop points will dominate. We
will cut these from the data after normalizing the graphs.
5. To normalize the curves, right click and drag over the appro-
priate baseline area. Click -> Y-Transform button.
6. A dialog box opens with several choices of transformation.
Select -> Zero at Average of Selection -> Replace Original.
Both curves now start at 0 RU.
7. Cut out the spikes caused during regeneration and/or at the

start and end of injections. Use right mouse button to make a
selection, as above.
Click on Edit -> Cut. Data from all curves will be edited
simultaneously.
Note that this editing requires three separate operations.
The end result should look as follows.
8. If there is some background binding in the absence of IAA,

you may wish to subtract this from all the binding curves to
leave graphs of auxin-dependent binding.
Click on -> Y-Transform and select “Curve-Curve 2.”
Select “TIR1 - IAA” and -> Replace Original.
The end result should look as follows.
9. The data are now processed and clearly show auxin-dependent

TIR1 activity. In this case Rmax is around 200 RU (see Note
16). This rapid assay confirms that the protein is suitable for
further experimentation, such as compound screening or
kinetic analysis (see Notes 17 and 18).
3.4 Compound The binding assay described above is readily adapted for compound
Screening Assay library screening. Purified TIR1 protein is mixed with compounds,
each diluted to 50 μM from 10 mM stocks in DMSO. Buffer
blanks and controls with no auxin are run in every set, as well as a
number of replicates containing 50 μM IAA as positive controls to
help normalize data between protein preparations. The activity of
a compound may be expressed as, e.g., a percentage of binding in
the presence of IAA. Binding values are taken from a report point
taken just before the end of the association phase of the binding
curve. Compound activities are loosely classified as shown in Fig. 5.
3.5 Kinetics A kinetic experiment is needed in order to get values for affinity
and rates of association and dissociation. For kinetic work we need
to ensure that no factor is limiting except one variable, the concen-
tration of analyte. In particular, we need to ensure that there is no
mass transport limitation (see Note 19), and we start by limiting
coating of the chip surface (see Note 9) such that Rmax is 300 RU
or lower (on Biacore 2000 and 3000 instruments). The kinetic
analysis wizard on the Biacore 2000 only allows use of flow cells 1
and 2 (Fc2-1) in a kinetic run (see Note 20).
1. Run -> Run Application Wizard -> Kinetic Analysis.
Fig. 5 Classification of binders based on % of IAA’s maximum binding. Data has

undergone additional processing with Prism version 6.00f for Mac OS X, GraphPad
Software, La Jolla, California, USA, www.graphpad.com
2. -> Concentration Series (see Note 21).

You can choose to run two additional control experiments at
the same time (see Notes 19 and 22).
3. -> Direct Binding.
4. You will be prompted to fill in the concentration series for your

chosen analyte molecule. Input the molar concentrations and
the molecular weight of analyte (see Note 21).
5. Use Flow Cells -> 2, using 1 as reference (see Note 20).
Flow Rate -> 30 μL/min (default) (see Note 19 and 22).
Stabilization Time -> 0 min (default).
Injection Time -> 3 min (default).
Dissociation Time -> 15 min (default; see Note 23).
Run Order -> Random (default)

6. Select the regeneration conditions and sample rack positions
(see steps 8 and 9 in Subheading 3.3.2). Check the sample
summary sheet. -> Start -> Filename.
The wizard runs the experiment and presents you with a sum-
mary table of kinetic parameters, statistics, etc., all based on a
1:1 Langmuir kinetic model. For TIR1 and IAA this should be
appropriate.
7. At the end of the wizard run you will be presented with sum-
mary data windows.
The first summary is the kinetic fit (in blue) to the data (in
red). The kinetic parameters of the fit show that the affinity
was calculated to be KD = 7.4e-8 M (74 nM). In this example
the TIR1 protein has been mixed with saturating IAA (100 μM)
and the protein is titrated in a dilution series against the degron
peptide on the chip. You will see that some repetitions were
built into the experiment. In this case the raw data and model
curves are not perfectly matched. You may use the other report
tabs to explore why this might be.
-> tab Baseline Level.
8. You will see that this report shows how much the baseline
shifts over all the injections for Fc1. In this case there has been
some drift, but this is not systematic, nor large (note RU scale).
-> Mass Transfer.
9. This report shows the results from the optional mass transport
limitation experiment at three flow rates (see Notes 19 and
22). In this case the binding curves almost overlap, but the red
curve is slightly displaced, indicating some reduced binding
early in the injection cycle at this low flow rate. Therefore,
there was some limitation to binding at the slowest flow rate,
but the curves for 15 and 75 μL/min are essentially identical
and so you could choose any flow rate greater than 15 μL/min
for kinetic work on this chip. -> Linked Reactions.
10. Linked reactions are asking whether your binding interaction

changes the reagents by comparing off-rates after different
reaction times. In this example it is difficult to tell and if you
want a more quantitative analysis of the off-rates, this can be
done in BIAevaluation.
11. Your own kinetic analysis of the data in BIAevaluation will
allow you to clean the data, and select the best time windows
for fitting, and the option of selecting alternative kinetic mod-
els. BIAevaluation also allows kinetic analysis using flow cells 3
and 4.
12. Data processing for kinetic experiments in BIAevaluation is
similar to that shown in Subheading 3.3.3 above. Particularly
important is the selection of curves and subtraction of your
chosen blank(s). Create a working set as an Overlay plot. There
are a number of kinetic models that can be applied to data.
TIR1 has a single binding pocket for auxin and so we will use
the 1:1 Langmuir model (see Note 22).
13. Start the kinetic evaluation wizard by choosing Fit -> Kinetics
Simultaneous ka/kd, or using the toolbar button with the sen-
sorgram icon.
14. Unless you have edited the curves beforehand, use Cut to pro-
cess the curves, remove regeneration spikes, etc.
15. Subtract the chosen blank -> Y-Transform and select “Curve-
Curve 2.”
Select your blank sample without the analyte “IAA 0” and ->
Replace Original.
16. You will see a dialog box headed Select Data and a time bar
appears at the top of the sensorgram screen.
The black lines should be dragged to the times of the start of

the injection and the end of the injection. The grey windows in
the bar are then dragged to select the time periods you want to
select for kinetic modeling. Use as much of your data as is rea-
sonable. You will see the kinetic model you have selected dis-
played at the top of the window. You may identify which curve
is which by clicking on it and the identifier will be displayed
with color code at the top right of the window.
-> Next will produce a table in which you will see the concen-
tration data for the analyte (as you entered the data in 3.5.5
above). The software is ready to fit.
17. Hit -> Next -> Filename -> Start. You will see the iterative fits
on screen.
18. At the end of the fit, the first report window shows a table
displaying the fitted parameters with statistics. The Selections
tab will review the data selection (as above). The fitted graphs
are superimposed on the colored data curves. On the Report
page you are seeking a low chi2 value and you may close any fit
and change the data selection windows, datasets, etc. until you
are satisfied with the result.
This example shows an experiment in which IAA was titrated

against a fixed TIR1 protein concentration. The chi2 value of
5.37 is reasonable and the model calculates the affinity KD to
be 2.43e − 6 M (2.4 μM), which reflects the affinity of TIR1 for
IAA on its own (see Note 24).
19. -> Residuals shows how well the data fit across the selected
windows.
20. The residuals from each binding curve and fit are plotted.
Ideally they all fall on the zero line. Here residuals deviate
more as the concentration of analyte rises, and more towards
the start of each phase (association and dissociation). Rmax in
this example is 378 RU and so all residuals are less than 10 %
Rmax, most below 5 %.
3.6 Thermodynamics 1. As noted above, thermodynamic values may be derived from

with SPR SPR experiments although ITC will give values for the reaction
in solution. The thermodynamic experiment is a kinetic experi-
ment run at a series of chip temperatures. Biacore instruments
can all control microfluidic and chip temperatures to within
0.1 °C over the range of 5–45 °C. Given that the transition
from one temperature to the next can take many tens of min-
utes and that kinetic experiments all require long runs, allow
plenty of time for thermodynamic work. Depending on the
availability of your protein you may choose as many tempera-
tures for the runs as you wish, but intervals of 10 °C should
give reasonable data for Van’t Hoff analysis. The wizard will
give summary report windows, as for kinetics above.
3.7 System 1. For all the wizard routines, unless you have unchecked the
Maintenance default box for Standby Flow After Run, the SPR buffer solu-
and Closure tion will be passed over your chip at 5 μL/min for up to 4 days.
The degron peptide chips are stable over many rounds of bind-
ing and regeneration, but once experiments are concluded the
chip is undocked and replaced with a Maintenance Chip.
2. Once undocked, rinse the degron-coated chip with deionized
water and then dry the surface under a stream of dry gas. Store
in a 50 mL Falcon tube at 4 °C.
3. Replace the SPR buffer with SPR rinse solution and rinse or
prime the system. This washes any residual salts from the fluid-
ics. The instrument may be left on Standby with SPR rinse
solution, but if Desorb has not been run for 4 weeks run a
Desorb routine. Ideally, each user will follow their runs with a
Desorb cycle to ensure that the SPR is ready and clean for the
next user. Make sure that Desorb is run only with a Maintenance
Chip docked.
4. Select -> Desorb from Working Tools (see step 4 in
Subheading 3.1).
5. If longer term storage is anticipated, follow Desorb with Sanitize.
4 Notes
1. The surfactant concentration is 10x higher than the suggested

one in the ready-made buffer from GE Healthcare (0.005 % v/v).
We found out that while having no negative effects on binding,
it does improve the smoothness of the signal and reduces bub-
ble generation especially in instruments where fluidics are not
brand new.
2. Always Prime when exchanging the running solution, even
between two batches of SPR buffer. This exchanges the fluid in
the pumps, flushes air bubbles, and washes the chip.
3. We strongly recommend that you condition the new chip
before coating by running this extra cleaning step. It removes
from the binding surface debris left from manufacture.
4. The Detection dialog box allows each cell to be selected inde-
pendently. If you are cleaning flow cells 1 and 2 you select the
option Fc 1-2. The reference subtraction-none means that the
signal from both cells will be displayed separately. The sub-
tracted 2-1 signal is not helpful when cleaning. Sensorgrams
will be recorded only from the flow cells you select.
5. All files with extension .blr contain raw results and can be opened
and post-processed with BIAevaluation software as will be shown.
6. Notice that the two flow cells of a brand new chip give differ-
ent absolute response units (RU) when same fluid is moved
over them. These units are arbitrary. When analyzing the data
what matters is the response difference.
7. The manual run can be adapted for other occasions when man-
ual control over the process is preferable. Select combinations
of commands via the buttons to create appropriate command
queues.
8. Heavy usage with flow cells left uncoated tends to allow debris
to accumulate over them and impair subsequent coating. While
economical to hold open cells 3 and 4, remember that their
performance may deteriorate if held for long periods. If you
know what you want to coat 3 and 4 with, we recommend that
all flow cells are coated at the same time.
9. You may vary surface coating density by diluting the biotinyl-
ated reagent or by reducing coating time. In practice with bio-
tinylated reagents, the affinity between biotin and streptavidin
is so high (10−15 M) the better strategy is dilution. Our bioti-
nylated peptide stocks at 1 mg/mL are diluted 20-fold for
saturating chip surfaces, but diluted up to a further 1000-fold
for kinetic experiments for which mass transport limitation (see
Note 18) needs to be avoided. You may also wish to use the
Surface Preparation wizard and the facility for Target coating.
In this you provide a diluted biotinylated ligand and enter an
RU value that you want to achieve. The software will then sip,
wash, and check as coating progresses, with short injections
used to allow ligand binding up to the target value.
10. The molecule captured on the chip is referred to as the ligand
in Biacore terminology. The molecule in solution which binds
to the immobilized ligand is referred to as the analyte.
11. Although arbitrary, RUs scale directly with mass and you will
see that far more peptide appears immobilized than biocytin.
In fact, more moles of biocytin are immobilized than the pep-
tide, probably because the small biocytin can access more avail-
able sites on the surface.
12. For TIR1 + IAA and TIR1 - IAA we recommend pipetting the
IAA in DMSO first onto the walls of the sample tube followed
by the buffer; this will ensure better mixing.
13. In more advanced analyses you may, e.g., randomize sampling.
14. We recommend that assay runs are preceded by at least three
buffer injections to ensure that the cells are purged and stabi-
lized. The regeneration of the coated surfaces between assay
cycles will remove some loosely associated ligand on the first
runs and a few regeneration cycles settle the surface before
accurate measurements are taken.
15. We recommend that the sample racks are chilled to 10 °C if
you have a series of experiments to run, but for a rapid activity
assessment this may not be necessary. Rack temperature is con-

trolled externally on the Biacore 2000.
16. Rmax is the value at the top of the binding hyperbola and repre-
sents full binding site occupancy. This value is important in
kinetic experiments.
17. The data can also be exported for processing in you favorite
software. Select File, then -> Export.
18. Files from BIAevaluation are saved as .ble files.
19. Mass transport limitation: If the rate of exchange of analyte
from solution with the ligand on the surface is limiting, or
binding is so advanced at the front of the flow cell such that the
analyte concentration flowing over areas of the flow cell behind
the front is depleted, the measured rates of binding will be
affected. This happens, for example, when the coating density
is too high or flow rates are too slow. Even if you follow best
practice with low coating densities we recommend that you
check for mass transport limitation by evaluating association
rates at different flow rates. The kinetic wizard offers this as an
optional extra.
20. The kinetic analysis wizard only allows you to use Fc2-1 in
kinetic screens. In order to use Fc4-3 or 2, 3 and 4 with 1 as
reference, you can set up the method in the binding analysis
wizard, but you will then need to fit kinetic models in the
BIAevaluation software.
21. Analyte concentrations should cover a full range of binding
curves. The kinetic analysis wizard requires a minimum of five
different concentrations of analyte and at least one concentra-
tion in duplicate. A minimum of one zero concentration sam-
ple (blank) should also be included. There is a calculator built
in so that if you enter the MW, entry of either nanomolar or
ng/mL concentrations will calculate the other. If you enter the
highest concentration first, the wizard will automatically sug-
gest values for the next rows when you press Enter to complete
an entry.
22. The Kinetic Analysis dialog box offers you the option of select-
ing two additional experiments, Mass Transport Limitation
and Linked Reactions. If you have already checked that a flow
rate of 30 μL/min will not limit the association rate, you do
not need to run the mass transport option. However, if you
have not checked, we recommend that this option is included.
It will ask for more protein sample (positive control mix) and
will run the binding at 5, 15, and 75 μL/min and report on
these as a cluster at the end of the run. You are looking to see
that there is no difference between association curves at the
higher flow rates. If there is a difference, increase the flow rate
to 75 μL/min. Remember that this will use a lot more of your
analyte sample. The Linked Reactions option will run the

binding of your positive control for three reaction times, 1, 3,
and 20 min. As above, you will have a report box on this at the
end of the run. The curves will be aligned at the point of the
start of the dissociation curve. You are looking to see that asso-
ciation time does not affect dissociation rate. If the dissociation
curves differ you need to consider the possibility of linked
reactions and, hence, more complex binding models than the
1:1 Langmuir. Again, this option requires considerably more
analyte sample.
23. In kinetic analyses a long (15 min) dissociation time is offered
as default. This may be shortened if you know that dissociation
of the bound complex is rapid. Otherwise, the model used to
fit the kinetic parameters needs a lot of data if the off-rate is
slow (gradient is shallow).
24. We show examples of two different kinetic experiments. First
was titrating TIR1 saturated with IAA over the degron. Second
was titrating IAA against TIR1 before exposure to the degron.
Hence the first experiment reflects the affinity of the TIR1-
IAA complex for degron, and the second the affinity of TIR1
for IAA. The former has a higher affinity than the latter, i.e.,
bound IAA as molecular glue has elevated the affinity for the
three-partner interaction over tenfold [7].
Acknowledgement
The authors are grateful for support from BBSRC, UK, Award
BB/L009366.
References
1. Napier RM, David KM, Perrot-Rechenmann C (2007) Mechanism of auxin perception by the
(2002) A short history of auxin-binding pro- TIR1 ubiquitin ligase. Nature 5:640–645
teins. Plant Mol Biol 49:339–348 7. Calderon-Villalobos LI, Lee S, De Oliveira C,
2. Venis MA, Napier RM (1992) Plant hormone Ivetac A, Brandt W, Armitage L, Sheard LB,
receptors: past, present and future. Biochem Tan X, Parry G, Mao H, Zheng N, Napier R,
Soc Trans 20:55–59 Kepinski S, Estelle M (2012) A combinatorial
3. Dharmasiri N, Dharmasiri S, Estelle M (2005) TIR1/AFB-Aux/IAA co-receptor system for
The F-box protein TIR1 is an auxin receptor. differential sensing of auxin. Nat Chem Biol
Nature 26:441–445 1:477–485
4. Kepinski S, Leyser O (2005) The Arabidopsis 8. Woodward AW, Bartel B (2005) Auxin: regu-
F-box protein TIR1 is an auxin receptor. Nature lation, action, and interaction. Ann Bot
26:446–451 95:707–735
5. Woo EJ, Marshall J, Bauly J, Chen JG, Venis M, 9. Lee S, Sundaram S, Armitage L, Evans JP,
Napier RM, Pickersgill RW (2002) Crystal Hawkes T, Kepinski S, Ferro N, Napier RM
structure of auxin-binding protein 1 in complex (2014) Defining binding efficiency and specificity
with auxin. EMBO J 17:2877–2885 of auxins for SCF(TIR1/AFB)-Aux/IAA co-
6. Tan X, Calderon-Villalobos LI, Sharon M, receptor complex formation. ACS Chem Biol
Zheng C, Robinson CV, Estelle M, Zheng M 21:673–682
Chapter 16
Studying Transcription Factor Binding to Specific Genomic

Loci by Chromatin Immunoprecipitation (ChIP)
S. Vinod Kumar and Doris Lucyshyn
Abstract
Plant hormone signaling involves complex transcriptional networks, where transcription factors orchestrate
the control of specific gene expression. These networks include cross talk between hormone signaling path-
ways, and the integration of environmental signals and the developmental program. Understanding how
particular transcription factors respond and integrate specific signals is crucial in order to understand the
basic mechanisms of hormonal signaling and cross talk. Studying transcription factor binding at specific
genomic loci by chromatin immunoprecipitation (ChIP) is therefore a valuable technique in order to ana-
lyze transcriptional regulation. The method is based on cross-linking proteins to DNA, the isolation of
chromatin, and immunoprecipitation of a transcription factor of interest. The attached DNA is then recov-
ered and analyzed by quantitative real-time PCR in order to establish binding sites of the respective tran-
scription factor. Here, we present a relatively simple and short protocol for ChIP on single loci.
Key words Chromatin immunoprecipitation (ChIP), Transcription factor, Gene expression,

Chromatin extraction, Binding site
1 Introduction
Plant hormone perception is mostly implemented by complex

transcriptional networks, where transcriptional regulators play an
important role in orchestrating the downstream control of gene
expression. A well-studied example is phytochrome interacting fac-
tors (PIFs) [1]. They are involved in gibberellin, auxin, brassino-
steroid, and ethylene signaling and respond to environmental
triggers, such as light and temperature [2–9]. Chromatin immuno-
precipitation (ChIP) has played a pivotal role for determining
direct PIF-target genes and hence elucidating the function of these
transcription factors in many studies [1–4, 10].
ChIP is commonly used to assess the binding of transcriptional
regulators to specific genomic loci. The technique involves cross-
linking proteins to DNA and extracting chromatin, which is then
fragmented to achieve relatively small DNA fragments with
193
194 S. Vinod Kumar and Doris Lucyshyn
Fig. 1 ChIP workflow
transcriptional regulators attached. The fragmented and cross-

linked chromatin is then used for immunoprecipitation (IP) using
an antibody directed to an epitope-tagged transcriptional regulator
of interest. After this step, any DNA attached to your protein of
interest is recovered and analyzed for specific genomic regions by
quantitative real-time PCR (see Fig. 1). Several protocols for isolat-
ing chromatin are available. The method described here is based
Chromatin Immunoprecipitation 195
Fig. 2 Principle of ChIP
on enrichment of nuclei by using a sequence of buffers with

different sucrose concentrations during extraction [11]. This will
also separate chloroplasts from nuclei and give relatively pure
chromatin.
Setting up a ChIP experiment requires careful preparation, and
its success will also depend on the right choice of starting material.
The most important considerations are described below; for more
details see for example [11–13]. Here, we describe a simple and
short protocol to perform ChIP on specific targets. For the analysis
of genome-wide protein-DNA interactions like ChIP-seq, cell-
type-specific interactions, or the analysis of locus-specific histone
modifications, see other protocols such as [14–17].
1.1 Starting Material Preparing or choosing a suitable plant line is essential to achieve clear
results. Using a small tag for fusion to your protein of interest, such
as HA or FLAG tag, is a good choice, as they most likely will not
interfere with the activity of your protein of interest, and reliable
good-quality antibodies are available from a range of suppliers.
Using a protein-specific antibody requires extensive optimization
and careful design of negative controls, and might give poor or false-
positive results. Another important factor is the choice of promoter
for your constructs. Placing the epitope-tagged gene of interest
under the endogenous promotor would be ideal and is certainly
worth trying. However, the efficiency of ChIP is often very low, and
using endogenous promoters might not result in protein levels that
are sufficient to get clear results. ChIP for transcriptional regulators
is therefore often performed using over-expression constructs. If
possible, transform your epitope-tagged construct to a mutant
(loss-of-function) background and test for complementation. It is

also advisable to check protein levels by Western blot on plant mate-
rial grown in the same conditions that you are planning for ChIP.
This will give confidence about the plant line, tissue, growth condi-
tions, and antibody before you start with the ChIP protocol.
1.2 Primer Design Design primers suitable for qPCR-based analysis of recovered DNA
after ChIP. PCR fragments should be 100–200 bp in length; frag-
ments longer than 500 bp are not suitable. For primer design,
online tools such as primer3 (http://bioinfo.ut.ee/primer3-0.4.0/)
or primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-
blast/) are very helpful. Use these tools also to look for potential
off-targets and adjust primers accordingly. The PCR fragments
should cover predicted binding sites, and it is recommended to
design several sets of primers covering a region of interest. Negative
control primers are necessary to assess background—choose regions
in the coding sequence or 3′-UTR of your gene of interest (keep at
least 1–2 kb distance to the predicted binding sites), as well as pro-
moters of unrelated genes for negative control fragments. Keep in
mind that transcriptional regulators might bind in coding regions,
and predictions about negative control fragments within the coding
region of your gene of interest are not always accurate.
2 Materials
2.1 Equipment 1. Desiccator.

2. Miracloth.
3. Liquid nitrogen.
4. Mortar and pestle.
5. Sonication device.
6. Photometer or NanoDrop.
7. DNA-electrophoresis equipment.
8. Water bath.
9. Magnetic stand.
10. Rotating wheel (in a cold room).
11. Real-time PCR machine.
2.2 Material 1. Plant growth media.

and Buffers 2. Cross-linking buffer: 0.4 M Sucrose, 10 mM Tris–HCl pH 8,
10 mM MgCl2.
3. Formaldehyde.
4. 2 M Glycine.
5. EB1, extraction buffer 1 (prepare freshly from stock solutions):
0.4 M Sucrose, 10 mM Tris–HCl pH 8, 10 mM MgCl2, 5 mM
2-mercaptoethanol, 0.1 mM PMSF, 1× protease inhibitor mix.
0.25 M Sucrose, 10 mM Tris–HCl pH 8, 10 mM MgCl2,
5 mM 2-mercaptoethanol, 1 % Triton X-100, 0.1 mM PMSF,
1× protease inhibitor mix.
1.7 M Sucrose, 10 mM Tris–HCl pH 8, 10 mM MgCl2, 5 mM
2-mercaptoethanol, 0.1 mM PMSF, 1× protease inhibitor mix.
8. Nuclei lysis buffer: 50 mM Tris–HCl pH 8, 10 mM EDTA, 1×
protease inhibitor mix.
9. Triton X-100.
10. 20 % Chelex® 100 Resin.
11. Elution buffer: 0.1 M NaHCO3, 1 % SDS (prepared freshly,
incubate at 65 °C immediately before use).
12. Proteinase K (10 mg/ml).
13. RNaseA (10 mg/ml).
14. Phenol-chloroform-isoamyl alcohol (PCI, 25:24:1) (optional,
see step 9 of Subheading 3.4).
15. Glycogen 1 μg/μl (optional, see step 9 of Subheading 3.4).
16. 3 M NaOAc (optional, see step 9 of Subheading 3.4).
17. Antibody against epitope-tagged protein.
18. Magnetic beads (Protein A or Protein G beads, according to
the data sheets of your antibody and the magnetic beads).
19. BSA (5 mg/ml).
20. ChIP dilution buffer: 5 mM EDTA, 10 mM Tris–HCl, pH 8,
160 mM NaCl, 1× protease inhibitor mix.
21. Low-salt buffer: 150 mM NaCl, 0.1 % SDS, 1 % Triton X-100,
2 mM EDTA, 20 mM Tris–HCl pH 8.
22. High-salt buffer: 500 mM NaCl, 0.1 % SDS, 1 % Triton X-100,
2 mM EDTA, 20 mM Tris–HCl pH 8.
23. LiCl wash buffer: 0.25 M LiCl, 1 % IGEPAL, 1 % Na-
deoxycholate, 1 mM EDTA, 10 mM Tris–HCl pH 8.
24. TE buffer: 10 mM Tris–HCl pH8, 1 mM EDTA.
25. DNA purification or PCR cleanup kit.
26. SYBR Green real-time PCR master mix.
3 Methods
3.1 Plant Growth 1. Sterilize an amount of seeds that will give at least 2.5 g of
material (see Note 1).
2. Prepare plant growth media.
3. Sow the seeds and stratify for 2–4 days in the dark at 4 °C.
4. Grow plants according to the requirements of your
experiment.
3.2 Harvesting 1. Prepare 60 ml cross-linking buffer for each sample (for cross-
and Fixing linking and rinsing).
2. Punch holes in the lids of 50 ml tubes, aliquot 30 ml of cross-
linking buffer for each sample, and add 0.85 ml of 35 % form-
aldehyde (1 % final concentration).
Caution: Formaldehyde is toxic—work under the fume hood.
3. Harvest 2.5–5 g of plant material directly in the prepared tubes
(see Note 2).
4. Apply vacuum in a desiccator for max. 20 min (see Note 3).
5. To stop cross-linking, add 2.5 ml of 2 M glycine in each tube,
and mix well.
6. Apply vacuum for 5 min.
7. Carefully remove solution from the tubes and rinse well with
cross-linking buffer (without formaldehyde).
8. Drain tissue on lab tissue paper, keeping the tubes inverted for
2–5 min. Do not let the plant material dry.
9. Snap freeze plant material in liquid nitrogen.
Stopping point: The material can be stored at −80 °C at this point
(see Note 4).
3.3 Preparation Caution: 2-Mercaptoethanol is toxic, work in the fume hood.

of Chromatin
1. Grind tissue in liquid nitrogen to a fine powder using mortar
and pestle.
2. Weigh ground material and use roughly same amounts for
chromatin preparation for each sample (see Note 5).
3. Resuspend samples in 25 ml of EB1 (see Note 6).
4. Filter through two layers of miracloth.
5. Centrifuge filtrate at 4000 rpm (1500 × g), 4 °C, for 15 min.
6. Resuspend pellet in 1.5 ml EB2, and incubate on ice for
3–5 min with intermittent swirling.
7. Centrifuge at 14,000 rpm (15800 × g), 4 °C, for 10 min (see
Note 7).
8. Prepare a fresh 2 ml tube containing 500 μl EB3 for each

sample.
9. Resuspend the pellet in 500 μl of EB3 and slowly load on top
of fresh EB3 prepared before.
10. Centrifuge at 14,000 rpm (15800 × g), and 4 °C for 20–30 min
(see Note 8).
11. Completely remove supernatant.
Stopping point: The pellet can be frozen and stored at −80 °C for
months.
3.4 Chromatin 1. Resuspend the crude nuclei pellet in 500 μl nuclei lysis buffer,
Fragmentation avoid bubble formation.
2. Sonicate the samples at 4 °C to shear chromatin to fragments
of 200–1000 bp (the optimal fragment size is 500 bp). This
step will have to be optimized depending on the sonication
device (see Note 9).
3. Prepare an aliquot of sheared chromatin for the control of
fragment size on a gel and DNA quantification: transfer 25 μl
of the chromatin to a fresh tube, add an equal volume of 20 %
Chelex, vortex briefly, and boil for 10 min with intermittent
mixing (Note 10).
4. Bring tubes to room temperature, add 0.1 μl proteinase K
(10 mg/ml), and incubate at 50 °C for 30 min with intermit-
tent swirling.
5. Boil again for 10 min.
6. Spin the samples at maximum speed for 1 min, and recover the
supernatant carefully without taking Chelex.
7. Add 0.1 μl RNaseA (10 mg/ml) and incubate at 37 °C for
20 min.
8. Load samples on a 1.5 % agarose gel (see Note 11).
9. Alternative de-cross-linking protocol (see Note 12): After RNase
treatment at step 7, add 50 μl elution buffer to the chromatin,
mix well, add 4 μl 5 M NaCl, and incubate at 65 °C overnight.
Extract DNA by adding an equal volume of PCI, vortex, and
centrifuge at full speed for 5 min at room temperature. Recover
supernatant, add 5 μl glycogen (1 μg/μl) as carrier and 1/10 of
sample volume 3 M NaOAc, and precipitate with 2.5 vol EtOH
at −20 °C. Wash the pellet once with 70 % ethanol, dissolve the pel-
let in TE, and load on an agarose gel as before.
10. De-cross-linked DNA can be used for quantification using a
spectrophotometer or NanoDrop.
11. If sonication is complete add Triton X-100 to a final concen-
tration of 1 % and centrifuge at 13,000 rpm (18400 × g), 4 °C,
for 10 min.
12. Recover the supernatant and discard the pellet to remove

unsoluble debris.
Stopping point: Chromatin can be stored at −80 °C for months.
Snap freeze in liquid nitrogen and avoid multiple freeze-thaws.
Freeze the input-aliquot separately (see Subheading 3.5.1).
3.5 Immuno- 1. Wash 45 μl magnetic beads for the following three samples:
precipitation (IP) 15 μl beads each for two IPs (technical repeats) and one no-
antibody control (mock) (see Note 13).
3.5.1 Prepare Beads
and Chromatin 2. Block the beads: Add 500 μl of 5 mg/ml BSA in PBS, and
incubate on a rotating wheel for 2 min at 4 °C. Bring the tubes
to the magnetic rack, let beads attach to sides of the tube, dis-
card the buffer, and repeat washing two more times.
3. Wash the beads twice in 1 ml ChIP dilution buffer, remove the
buffer completely, and resuspend the beads in 45 μl ChIP dilu-
tion buffer.
4. Take out 30 μl of washed beads, resuspend in 1 ml ChIP dilu-
tion buffer, and add antibody (2–5 μg per 25 μg of chromatin,
see Note 14).
5. Incubate at least for 5 h or overnight at 4 °C on a rotating
wheel to couple the antibody to the beads. Alternatively this
can be done for 1 h at room temperature.
6. Store remaining 15 μl of washed beads in ChIP dilution buffer
at 4 °C until needed for the mock sample.
7. Wash freshly antibody-coated beads twice with ChIP dilution
buffer.
3.5.2 Immuno- 1. Aliquot chromatin: Use equal amounts (ca. 25 μg) for each IP,
precipitation the no-antibody control, and the input-sample. Store input at
−80 °C until de-cross-linking with the other samples after
immunoprecipitation (see Note 15).
2. Dilute chromatin with ChIP dilution buffer to an SDS concen-
tration of 0.1 %.
3. Mix diluted chromatin with the beads: two samples with 25 μg
chromatin + 15 μl antibody-coated beads, one sample with 25 μg
chromatin + 15 μl beads (mock), and bring samples to a final
volume of ca. 1.5 ml with ChIP dilution buffer (see Note 16).
4. Incubate on a rotating wheel at 4 °C overnight.
3.5.3 Washes For each washing step add 1 ml of respective buffer, invert gently,
incubate on ice for 5 min, and let the beads settle in the magnetic
stand. Remove supernatant carefully, and do not let the beads dry out.
1. Wash twice with low-salt wash buffer.

2. Wash twice with high-salt wash buffer.

3. Wash once with LiCl wash buffer.
4. Wash twice with TE (see Note 17).
5. Resuspend beads in 100 μl TE.
3.5.4 De-cross-linking At this step, include input sample for de-cross-linking.

and DNA Recovery
1. To recover DNA, add 100 μl of 20 % Chelex and boil for
10 min.
2. Bring tubes to room temperature, add 0.1 μl proteinase K
(10 mg/ml), incubate at 50 °C for 30 min, and boil again for
10 min. Mix regularly during incubation steps.
3. Spin the tubes at maximum speed for 1 min, and recover the
supernatant carefully.
4. Optional: In case Chelex particles remain in the samples, purify
recovered DNA using a PCR product cleanup kit. Follow the
instructions of the kit and elute DNA with 100 μl TE. The elu-
ate from this step can be directly used for PCR (see Note 18).
Stopping point: The recovered DNA can be stored at −20 °C.
3.6 qPCR and Data 1. Dilute eluted DNA 1:10 with water and use 5 μl as template in
Analysis 15 μl reactions (see Note 19).
2. For analyzing data using the percentage of input method (rec-
ommended), use the following method:
% of input = 100 × 2Δ(ct(input) − ct(sample)) (see Note 20).
3. For analyzing data using the fold enrichment method:
Fold enrichment = (fraction of inputsample)/(fraction of inputcontrol)
(see Note 20).
4 Notes
1. Any healthy plant tissue is suitable, such as seedlings, roots,

rosette leaves, or flowers, depending on the expression pat-
terns of your transcription factor and your experimental setup.
2. For harvesting, work quickly. Be sure to always harvest at the
same time of day for biological repeats and take potential circa-
dian regulation of your transcription factor into account.
3. Apply vacuum until the tissue sinks and gets translucent. Keep
the exposure to formaldehyde as short as possible and work
quickly during cross-linking.
4. If the material is stored, wrap it in aluminum foil.
5. Make sure that the ground material does not thaw at any stage,
and use spatula chilled in liquid nitrogen while transfer.
6. Adjust the amount of buffer to the amount of ground material.

25 ml of EB1 works well for 2–3 g of ground material.
7. If the pellet looks yellow and clean, proceed with chromatin
fragmentation (Subheading 3.4). If there is a green rim, con-
tinue with the next step.
8. At this step, nuclei will settle in the pellet, while the chloro-
plasts keep floating in the top layer of EB3. This is a crude
separation enriching for nuclei.
9. This is a crucial step for the success of the ChIP protocol, and
should be optimized in order to get reproducible results.
Check fragmentation on an agarose gel. If necessary, reduce or
increase the number of sonication cycles to optimize fragment
size. During sonication it is essential that the samples do not
heat up. Try to avoid foaming, as it might cause uneven
sonication.
10. Always mix Chelex well before pipetting, as it settles very
quickly. Intermittent swirling is crucial to keep Chelex in sus-
pension during incubation steps.
11. DNA fragments should run between 200 and 1000 bp of size,
with the majority of signal at around 500 bp. If the samples do
not migrate into the gel, either sonication or de-cross-linking
was not efficient. Try more sonication cycles and de-cross-
linking by using the alternative protocol described at point
step 9 of Subheading 3.4 of this protocol to determine the
cause of the problem.
12. Chelex offers the advantage of speeding up and simplifying the
de-cross-linking protocol, as well as being nonhazardous com-
pared to classical DNA precipitation methods [18]. If you
encounter problems with de-cross-linking due to Chelex, use
this alternative protocol. Chelex is also not compatible with
sequencing of recovered DNA, as most of it is denatured after
the treatment.
13. See antibody and magnetic bead data sheets for choosing pro-
tein A or G magnetic beads.
Never let the beads dry out; always resuspend before
pipetting.
14. The amount of antibody might have to be optimized—start
with the recommendations of the antibody supplier.
15. In case the amount of chromatin is limiting, keep less for the
input sample (i.e., 1/10 of the total chromatin). Be sure to
take a corresponding dilution factor into account when analyz-
ing the qPCR data.
16. Use safe-lock 2 ml tubes with round bottom.
17. Transferring the beads to fresh tubes between TE washes helps

to reduce background.
18. The alternative de-cross-linking protocol described in step 9
of Subheading 3.4 can be used at this point.
19. Use this dilution as a starting point. If the qPCR does not give
good results, try using more template. In some cases too much
template might also be inhibitory, depending on the purity of
the DNA.
20. For more details and other possibilities of data analysis and the
differences between these methods see Ref. [13].
References
1. Pfeiffer A et al (2014) Combinatorial complex- 10. Hornitschek P et al (2012) Phytochrome inter-

ity in a transcriptionally centered signaling hub acting factors 4 and 5 control seedling growth in
in Arabidopsis. Mol Plant 7(11):1598–1618 changing light conditions by directly controlling
2. de Lucas M et al (2008) A molecular frame- auxin signaling. Plant J 71(5):699–711
work for light and gibberellin control of cell 11. Bowler C et al (2004) Chromatin techniques
elongation. Nature 451(7177):480–484 for plant cells. Plant J 39(5):776–789
3. Franklin KA et al (2011) Phytochrome- 12. Yamaguchi N et al (2014) PROTOCOLS: chro-
interacting factor 4 (PIF4) regulates auxin bio- matin Immunoprecipitation from Arabidopsis
synthesis at high temperature. Proc Natl Acad Tissues. Arabidopsis Book 12:e0170
Sci U S A 108(50):20231–20235 13. Haring M et al (2007) Chromatin immunopre-
4. Kumar SV et al (2012) Transcription factor cipitation: optimization, quantitative analysis
PIF4 controls the thermosensory activation of and data normalization. Plant Methods 3:11
flowering. Nature 484(7393):242–245 14. Deal RB, Henikoff S (2011) The INTACT
5. Bai MY et al (2012) Brassinosteroid, gibberel- method for cell type-specific gene expression
lin and phytochrome impinge on a common and chromatin profiling in Arabidopsis thali-
transcription module in Arabidopsis. Nat Cell ana. Nat Protoc 6(1):56–68
Biol 14(8):810–817 15. Kaufmann K et al (2010) Chromatin immu-
6. Sakuraba Y et al (2014) Phytochrome- noprecipitation (ChIP) of plant transcrip-
interacting transcription factors PIF4 and PIF5 tion factors followed by sequencing
induce leaf senescence in Arabidopsis. Nat (ChIP-SEQ) or hybridization to whole
Commun 5:4636 genome arrays (ChIP-CHIP). Nat Protoc
7. Feng S et al (2008) Coordinated regulation of 5(3):457–472
Arabidopsis thaliana development by light and 16. Song J, Rutjens B, Dean C (2014) Detecting
gibberellins. Nature 451(7177):475–479 histone modifications in plants. Methods Mol
8. Huq E, Quail PH (2002) PIF4, a phytochrome- Biol 1112:165–175
interacting bHLH factor, functions as a nega- 17. Weinhofer I, Kohler C (2014) Endosperm-
tive regulator of phytochrome B signaling in specific chromatin profiling by fluorescence-
Arabidopsis. EMBO J 21(10):2441–2450 activated nuclei sorting and ChIP-on-chip.
9. Koini MA et al (2009) High temperature- Methods Mol Biol 1112:105–115
mediated adaptations in plant architecture 18. Nelson JD et al (2006) Fast chromatin immuno-
require the bHLH transcription factor PIF4. precipitation assay. Nucleic Acids Res 34(1), e2
Curr Biol 19(5):408–413
Chapter 17
Hormone Receptor Glycosylation

Ulrike Vavra, Christiane Veit, and Richard Strasser
Abstract
Glycosylation is essential for all trees of life. N-glycosylation is one of the most common covalent protein
modifications and influences a large variety of cellular processes including protein folding, quality control
and protein-receptor interactions. Despite recent progress in understanding of N-glycan biosynthesis, our
knowledge of N-glycan function on individual plant proteins is still very limited. In this respect, plant
hormone receptors are an interesting group of proteins as several of these proteins are present at distinct
sites in the secretory pathway or at the plasma membrane and have numerous potential N-glycosylation
sites. Identifying and characterization of N-glycan structures on these proteins is essential to investigate the
functional role of this abundant protein modification. Here, a straightforward immunoblot-based approach
is presented that enables the analysis of N-glycosylation on endogenous hormone receptors like the brassi-
nosteroid receptor BRI1.
Key words N-glycosylation, Glycoprotein, Brassinosteroid receptor, Secretory pathway, Endoplasmic

reticulum, Golgi apparatus
1 Introduction
Glycosylation is a ubiquitous protein modification in eukaryotic

cells. Dependent on the linkage of the sugar or oligosaccharide to
the amino acid side chain there are two main types of glycosylation:
N- and O-glycosylation. N-glycosylation and early N-glycan pro-
cessing steps are highly conserved between plants and animals.
N-glycans are crucial for protein folding in the endoplasmic reticu-
lum and for quality control processes including the glycan-mediated
degradation of terminally misfolded glycoproteins [1]. A wide vari-
ety of other functions ranging from the modulation of protein–
receptor interactions and regulation of subcellular targeting to
protein stability have been described in mammals. In plants, com-
paratively little is known about the contribution of distinct N-glycans
to protein function [2]. However, studies with the heavily
N-glycosylated pathogen recognition receptor EFR, for example,
revealed an important role of N-glycosylation and N-glycan-
dependent quality control during plant immunity [3–7]. In contrast
205
206 Ulrike Vavra et al.
to animals, O-glycosylation of proteins is fundamentally different.

Plants lack a functional mucin-type O-glycosylation pathway which
involves the transfer of a single N-acetylgalactosamine (GalNAc)
residue to serine or threonine [8]. Instead, plants can attach galac-
tose to serine and modify hydroxyproline residues in proline-rich
proteins either with small arabinose chains or with larger arabinoga-
lactans [9].
Hormone receptors in the nucleus or cytoplasm and cytoplas-
mic domains of membrane-anchored receptors are typically neither
N-glycosylated nor contain any extensive O-glycan-like modifica-
tions because these proteins and their cytoplasmic regions are
never in contact with the ER/Golgi-located glycosylation machin-
ery. However, there is the possibility for the attachment of a single
N-acetylglucosamine (GlcNAc) to serine or threonine residues of
cytosolic or nuclear proteins. This posttranslational modification is
well described in mammals, where it plays a major role in diverse
signaling reactions [10]. Even though O-linked GlcNAc modifica-
tions have been described for plant proteins including histones
[11], most of the substrate proteins are currently unknown.
Interestingly, one of the two Arabidopsis O-GlcNAc-transferases
(SPINDLY) is also involved in gibberellin and cytokinin depen-
dent processes [12, 13]. Moreover, the recent identification of the
specific nucleotide sugar transporter ROCK1 suggests that an
unknown GlcNAc or GalNAc modification in the ER is important
for regulation of cytokinin levels [14].
Here, we focus on N-glycosylation and N-glycan processing of
hormone receptors in Arabidopsis, because the machinery for this
frequent glycosylation reaction is already very well understood [2]
and N-glycosylation has been clearly demonstrated for some of
them. N-glycosylation is initiated in the lumen of the ER by the
oligosaccharyltransferase (OST) complex. OST recognizes the
Asn-X-Ser/Thr (X can be any amino acid except proline) consen-
sus sequence on nascent polypeptide chains and transfers a preas-
sembled oligosaccharide en bloc to the asparagine of the acceptor
substrate in the lumen of the ER. In addition to this cotranslational
protein modification, which plays a major role for protein folding,
some skipped N-glycosylation sites are glycosylated posttransla-
tionally by a distinct OST complex with a different subunit compo-
sition [15]. As soon as the glycan precursor is transferred,
ER-resident glucosidases and mannosidases remove terminal sugar
residues and the glycoprotein with partially processed N-glycans
can undergo glycan-dependent quality control to ensure that only
fully folded proteins are released to other compartments of the
secretory pathway [1].
Which hormone receptors are possible targets for
N-glycosylation? Primary candidates for this protein modification
are all receptors that reside in the ER or along the secretory
pathway including the plasma membrane and extracellular space.
Glycosylation 207
Auxin transport, binding or metabolic enzymes have been detected

in the ER membrane and the ER lumen [16, 17]. Ethylene percep-
tion is supposed to take place in the ER [18] and studies with fluo-
rescent cytokinin receptor fusions suggest localization in this
organelle [19, 20]. Brassinosteroids are perceived at the plasma
membrane by the receptor BRASSINOSTEROID INSENSITIVE
1 (BRI1) [21]. Likewise, abscisic acid and cytokinin binding may
be mediated by plasma membrane anchored receptors [22, 23].
For many of the aforementioned hormones the N-glycosylation
status of their putative receptors is unknown or not very well inves-
tigated. N-glycosylation of the cytokinin receptor AHK3 has been
reported by Endoglycosidase H (Endo H) digestion of transiently
expressed GFP-AHK3 [19]. AUXIN BINDING PROTEIN1
(ABP1) contains one conserved N-glycosylation site [24] and an
insect cell produced ABP1 was found to be N-glycosylated [25].
In ER-resident auxin-transport related proteins like PILS5 an
N-glycosylation motif is present in one of the ER-lumen facing
loops. However, this motif is not conserved in all PILS proteins
[26] and only found in PILS5 and PILS7. In contrast to most of
these ER-to-plasma membrane located hormone binding proteins,
N-glycosylation of BRI1 is very well established [27–29].
Furthermore, it has been demonstrated that the N-glycans of mis-
folded BRI1 variants play a crucial role during ER-mediated qual-
ity control processes and glycan-dependent ER-associated
degradation (ERAD) [29–32].
In this chapter we describe simple immunoblot-based proto-
cols for the N-glycosylation analysis of hormone receptors. These
procedures are designed to discriminate between glycosylated and
non-glycosylated endogenous proteins without using tagged or
overexpressed proteins. Based on the presence of distinct N-glycan
structures additional information can be gathered about the sub-
cellular localisation and trafficking route of a given plant hormone
receptor. To identify the precise glycosylation profile of these pro-
teins more sophisticated mass spectrometry (MS)-based methods
have to be applied which in most cases require at least partial puri-
fication of the protein of interest (see Note 1). While this is still
challenging for many low expressed or tightly regulated endoge-
nous proteins, advances in MS-based technologies will provide a
more complete picture of plant hormone receptor glycosylation in
the future.
2 Materials
1. Arabidopsis thaliana stt3a-2 seeds [33] can be purchased from

the European Arabidopsis Stock Center (NASC ID: N800052).
The fut11 fut12 seeds [34] can be obtained upon request from
the Strasser group.
2. Mixer mill (e.g., from Retsch) with steel beads.

3. Container with liquid nitrogen.
4. Refrigerated benchtop centrifuge with a relative centrifugal
force of 9400 or higher and a rotor for 1.5/2 ml microcentri-
fuge tubes (such as Eppendorf Centrifuge 5414R).
5. Thermo block.
6. Orbital shaker.
7. Vertical electrophoresis system (such as Mini-Protean,
Bio-Rad).
8. Tank transfer system (such as Mini-Trans Blot, Bio-Rad).
9. Power supply with at least 100 V and 400 mA (such as Power
Pac Universal Power Supply, Bio-Rad).
10. SDS-PAGE (10 %) gels.
11. Tris–glycine–SDS PAGE running buffer (25 mM Tris, 192 mM
glycine, 0.1 % (w/v) SDS).
12. 3× Laemmli sample buffer (187.5 mM Tris pH 6.8, 30 % (w/v)
glycerol, 6 % (w/v) SDS 15 % (v/v) β-mercaptoethanol, 0.15 %
bromophenol blue).
13. Protein standard (we use Prestained protein marker, Fermentas).
14. Tris–glycine–methanol transfer-buffer (25 mM Tris, 192 mM
glycine, 20 % (v/v) methanol).
15. Blotting paper (we use Whatman 3 MM).
16. Blotting membrane: we use Amersham Protran Premium
nitrocellulose membrane (GE Healthcare).
17. 1× PBS: phosphate buffered saline (137 mM NaCl, 2.7 mM
KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4).
18. PBST: 1× PBS + 0.1 % (v/v) Tween 20.
19. 1× TBS: Tris buffered saline (25 mM Tris–HCl, pH 7.5,
150 mM NaCl, 2 mM KCl).
20. TBST: 1× TBS + 0.1 % (v/v) Tween 20.
21. Blocking solution for BRI1 detection on immunoblots:
TBST + 5 % (w/v) skimmed milk powder.
22. Blocking solution for AHK2 detection on immunoblots:
PBST + 3 % (w/v) BSA.
23. Western blotting detection reagent: we use SuperSignal West
PICO Chemiluminescent Substrate (Pierce) or Amersham ECL
Prime Western Blotting Detection Reagent (GE Healthcare).
24. Sensitive films for detection: we use Amersham Hyperfilm
ECL (GE Healthcare).
25. 1× PBS + 1 % (v/v) Triton X-100.
26. Film developer and fixer solutions.
Glycosylation 209
27. Polyclonal anti-BRI1 antibody (Agrisera: AS12 1859): 1:5000

diluted in TBST + 5 % (w/v) skimmed milk powder.
28. Polyclonal anti-AHK2 antibody (Agrisera: AS12 2113):
1:2500 diluted in PBST + 3 % (w/v) BSA.
29. Anti-rabbit IgG-peroxidase (such as Sigma, A0545): 1:50,000
diluted in PBST for detection of AHK2 or 1:5000 diluted in
TBST + 5 % (w/v) skimmed milk powder for BRI1 detection.
30. Endoglycosidase H (Endo H, we use 500,000 U/mL, New
England Biolabs) + Glyco Buffer 3 (New England Biolabs).
31. Peptide-N-Glycosidase F (PNGase F, we use 500,000 U/mL,
New England Biolabs) + 1× Glycoprotein Denaturing Buffer
(New England Biolabs).
32. NP-40 (10 % solution, such as provided by New England
Biolabs).
33. Kifunensine (such as Sigma: K1140): class I α-mannosidase
inhibitor, dissolved in ultrapure water.
3 Methods
3.1 N-glycosylation The prerequisite for monitoring of hormone receptor

Prediction N-glycosylation is the presence of the consensus sequence Asn-X-
Ser/Thr (where X can be any amino acid except proline) within
the protein of interest. N-glycosylation site prediction tools like
NetNGlyc (http://www.cbs.dtu.dk/services/NetNGlyc/) [35]
(see Note 2) provide a list of all potential sites present in a given
amino acid sequence (see Note 3). Examples are shown for
Arabidopsis AHK2 and BRI1 in Figs. 1 and 2, respectively. Using
artificial neural networks and a set of human proteins the NetNGlyc
predictor is trained to discriminate glycosylated from non-
glycosylated sites. This is relevant, since not all N-glycosylation
sites are glycosylated (see Note 4).
Next, it is important to predict the subcellular localization and
topology of the protein. The presence of a signal peptide—e.g., pre-
dicted using SignalP (http://www.cbs.dtu.dk/services/SignalP/)—
indicates targeting to the secretory pathway. The detection of
transmembrane regions (e.g., using TMHMM: http://www.cbs.dtu.
dk/services/TMHMM/ or HMMTOP: http://www.enzim.hu/
hmmtop/) and their position provides information on the accessibil-
ity of certain protein domains for the glycosylation machinery in the
lumen of the ER. While these predictions are informative for some
proteins like BRI1 (Fig. 2) or the proposed abscisic acid receptors
GTG1 and GTG2 [22] which do not have any potential N-glycosylation
site at all (not shown), for others (e.g., AHK2, Fig. 1) the number of
accessible potential N-glycosylation sites is ambiguous. Regardless of
the quality of these predictions, experimental confirmation of the
MSITCELLNLTSKKAKKSSSSDKKWLKKPLFFLILCGSLVIVLVMFLRLGRSQKEETDSCNGEEKVLYRHQNVTRSEIHD
80
LVSLFSDSDQVTSFECHKESSPGMWTNYGITCSLSVRSDKQETRGLPWNLGLGHSISSTSCMCGNLEPILQQPENLEEEN
160
HEEGLEQGLSSYLRNAWWCLILGVLVCHKIYVSHSKARGERKEKVHLQEALAPKKQQQRAQTSSRGAGRWRKNILLLGIL
240
GGVSFSVWWFWDTNEEIIMKRRETLANMCDERARVLQDQFNVSLNHVHALSILVSTFHHGKIPSAIDQRTFEEYTERTNF
320
ERPLTSGVAYALKVPHSEREKFEKEHGWAIKKMETEDQTVVQDCVPENFDPAPIQDEYAPVIFAQETVSHIVSVDMMSGE
400
EDRENILRARASGKGVLTSPFKLLKSNHLGVVLTFAVYDTSLPPDATEEQRVEATIGYLGASYDMPSLVEKLLHQLASKQ
480
TIAVDVYDTTNTSGLIKMYGSEIGDISEQHISSLDFGDPSRNHEMHCRFKHKLPIPWTAITPSILVLVITFLVGYILYEA
560
INRIATVEEDCQKMRELKARAEAADIAKSQFLATVSHEIRTPMNGVLGMLKMLMDTDLDAKQMDYAQTAHGSGKDLTSLI
640
NEVLDQAKIESGRLELENVPFDMRFILDNVSSLLSGKANEKGIELAVYVSSQVPDVVVGDPSRFRQIITNLVGNSIKFTQ
720
ERGHIFISVHLADEVKEPLTIEDAVLKQRLALGCSESGETVSGFPAVNAWGSWKNFKTCYSTESQNSDQIKLLVTVEDTG
800
VGIPVDAQGRIFTPFMQADSSTSRTYGGTGIGLSISKRLVELMQGEMGFVSEPGIGSTFSFTGVFGKAETNTSITKLERF
880
DLAIQEFTGLRALVIDNRNIRAEVTRYELRRLGISADIVSSLRMACTCCISKLENLAMILIDKDAWNKEEFSVLDELFTR
960
SKVTFTRVPKIFLLATSATLTERSEMKSTGLIDEVVIKPLRMSVLICCLQETLVNGKKRQPNRQRRNLGHLLREKQILVV
1040
DDNLVNRRVAEGALKKYGAIVTCVESGKAALAMLKPPHNFDACFMDLQMPEMDGFEATRRVRELEREINKKIASGEVSAE
1120
MFCKFSSWHVPILAMTADVIQATHEECMKCGMDGYVSKPFEEEVLYTAVARFFEPC
TMHMM posterior probabilities for WEBSEQUENCE

1.2
0.8
probability
0.6
0.4
0.2
0
0 200 400 600 800 1000
transmembrane inside outside
Fig. 1 N-glycosylation site and transmembrane domain predictions for AHK2 (At5g35750, Q9C5U2 Uniprot
database, 1176 amino acids). Putative N-glycosylation sites are depicted in blue. Two to four transmembrane
domains are predicted by TMHMM. Based on this prediction the topology and the number of potential
N-glycosylation in luminal (extracellular—marked as “outside”) domains are unclear
N-glycosylation status of an individual protein is always required to

identify novel glycoproteins.
3.2 Monitoring N-glycosylation shows that the hormone receptor is present in the
of Protein secretory pathway or extracellular environment (see Note 5). A simple
Underglycosylation way to monitor N-glycosylation is the analysis of differences in mobility
upon SDS-PAGE and immunoblotting in underglycosylation mutants
which display a reduced N-glycosylation efficiency. The most suitable
Glycosylation 211
MKTFSSFFLSVTTLFFFSFFSLSFQASPSQSLYREIHQLISFKDVLPDKNLLPDWSSNKNPCTFDGVTCRDDKVTSIDLS
80
SKPLNVGFSAVSSSLLSLTGLESLFLSNSHINGSVSGFKCSASLTSLDLSRNSLSGPVTTLTSLGSCSGLKFLNVSSNTL
160
DFPGKVSGGLKLNSLEVLDLSANSISGANVVGWVLSDGCGELKHLAISGNKISGDVDVSRCVNLEFLDVSSNNFSTGIPF
240
LGDCSALQHLDISGNKLSGDFSRAISTCTELKLLNISSNQFVGPIPPLPLKSLQYLSLAENKFTGEIPDFLSGACDTLTG
320
LDLSGNHFYGAVPPFFGSCSLLESLALSSNNFSGELPMDTLLKMRGLKVLDLSFNEFSGELPESLTNLSASLLTLDLSSN
400
NFSGPILPNLCQNPKNTLQELYLQNNGFTGKIPPTLSNCSELVSLHLSFNYLSGTIPSSLGSLSKLRDLKLWLNMLEGEI
480
PQELMYVKTLETLILDFNDLTGEIPSGLSNCTNLNWISLSNNRLTGEIPKWIGRLENLAILKLSNNSFSGNIPAELGDCR
560
SLIWLDLNTNLFNGTIPAAMFKQSGKIAANFIAGKRYVYIKNDGMKKECHGAGNLLEFQGIRSEQLNRLSTRNPCNITSR
640
VYGGHTSPTFDNNGSMMFLDMSYNMLSGYIPKEIGSMPYLFILNLGHNDISGSIPDEVGDLRGLNILDLSSNKLDGRIPQ
720
AMSALTMLTEIDLSNNNLSGPIPEMGQFETFPPAKFLNNPGLCGYPLPRCDPSNADGYAHHQRSHGRRPASLAGSVAMGL
800
LFSFVCIFGLILVGREMRKRRRKKEAELEMYAEGHGNSGDRTANNTNWKLTGVKEALSINLAAFEKPLRKLTFADLLQAT
880
NGFHNDSLIGSGGFGDVYKAILKDGSAVAIKKLIHVSGQGDREFMAEMETIGKIKHRNLVPLLGYCKVGDERLLVYEFMK
960
YGSLEDVLHDPKKAGVKLNWSTRRKIAIGSARGLAFLHHNCSPHIIHRDMKSSNVLLDENLEARVSDFGMARLMSAMDTH
1040
LSVSTLAGTPGYVPPEYYQSFRCSTKGDVYSYGVVLLELLTGKRPTDSPDFGDNNLVGWVKQHAKLRISDVFDPELMKED
1120
PALEIELLQHLKVAVACLDDRAWRRPTMVQVMAMFKEIQAGSGIDSQSTIRSIEDGGFSTIEMVDMSIKEVPEGKL
TMHMM posterior probabilities for WEBSEQUENCE

1.2
0.8
probability
0.6
0.4
0.2
0
0 200 400 600 800 1000
transmembrane inside outside
Fig. 2 N-glycosylation site and transmembrane domain predictions for Arabidopsis BRI1 (At4g39400, O22476
Uniprot database, 1196 amino acids). The 14 Asn-residues from N-glycosylation sites (N-X-S/T, shown in blue)
that face the lumen of the ER and are located in the BRI1 extracellular domain are highlighted in red. Asn-
residues in N-glycosylation sites that face the cytoplasmic side are shown in blue. The first transmembrane
domain prediction overlaps with the signal peptide sequence
Arabidopsis thaliana mutant for this approach is stt3a [33] (see Note
6). This line has a mutation in one of the two catalytic subunits of the
OST complex. As a consequence, glycoproteins can have either no or
a reduced number of N-glycans compared to wild-type plants. A faster
migrating band indicates the presence of underglycosylation suggest-
ing that the protein is N-glycosylated in wild-type plants (Fig. 3a).
Fig. 3 (a) Immunoblot analysis of BRI1 and AHK2 in the stt3a underglycosylation mutant which has a defect in
the catalytic subunit of the oligosaccharyltransferase complex [33]. BRI1 displays a faster migrating band
indicating a reduced number of N-glycans. (b) Col-0 seedlings were incubated with the α-mannosidase inhibi-
tor kifunensine and protein extracts were subjected to immunoblotting with BRI1- or AHK2-specific antibodies.
Due to the blocked N-glycan processing BRI1 displays a shift in mobility. The major N-glycan structures are
indicated. (c) Immunoblot of Endo H digested protein extracts from Col-0 seedlings that were incubated in the
presence or absence of kifunensine. The faster migrating band (<130 kDa) in the Endo H-treated sample
represents BRI1 lacking any N-glycans. No clear shift is observed for AHK2 suggesting that it is not
N-glycosylated. (d) Schematic representation showing the cleavage specificity of Endo H. Note: a single GlcNAc
is still present after Endo H digestion. For description of sugar symbols see Fig. 4
1. Harvest 100 mg stt3a and Col-0 wild-type seedlings (or leaf

material) and transfer to 2 mL Eppendorf tubes containing 2
steel beads (5 mm diameter) per tube.
2. Submerge 2 mL tubes with plant material in container with
liquid nitrogen.
3. Mount 2 mL tubes in mixer mill and run 2 min at 50–60
amplitude.
4. Add 4 μL extraction buffer (e.g., PBST or RIPA) per mg of
leaf material, vortex shortly, transfer liquid into a 1.5 mL tube
and incubate on ice for 15 min, invert tube every 3 min.
5. Centrifuge two times 15 min, 9600 × g at 4 °C, transfer super-
natant each time to a new tube.
Glycosylation 213
6. Mix samples with 3× Laemmli sample buffer and heat to 95 °C

for 5 min.
7. Load SDS-PAGE gel with 10–20 μl and run protein separation
for 1.5 h at 100 V.
8. Soak nitrocellulose membrane, sponges and blotting paper in
transfer buffer.
9. Perform the gel-membrane assembly according to the user
manual from Bio-Rad.
10. Blot for 1 h at 100 V.
11. Disassemble gel-membrane sandwich and carefully rinse the
membrane with ultrapure water.
12. Incubate in blocking solution for 1 h at room temperature.
13. Rinse briefly with PBST (AHK2) or TBST (BRI1).
14. Incubate membrane on a shaker for 1.5 h at room temperature
in the antibody solution.
15. After incubation, wash 4 times 5 min with PBST or TBST.
16. Add second antibody solution to membrane and incubate for
1.5 h at room temperature on shaker.
17. Wash four times 5 min in PBST or TBST.
18. Perform detection using the chemiluminescent substrate.
19. Develop the film.
3.3 Kifunensine Due to the current lack of understanding how the OST complex
Treatment Followed works in plants, it is possible that N-glycosylation is not detected
by Immunoblotting using the stt3a-dependent approach, e.g., when N-glycosylation is
solely dependent on STT3B activity (see Note 7). Therefore another
practicable approach is the use of the N-glycan processing inhibitor
kifunensine. Kifunensine is a class I α-mannosidase inhibitor that
blocks the removal of α-linked mannose residues from oligoman-
nosidic N-glycans [36] (Fig. 4). Since this trimming is absolutely
required for further processing and complex N-glycan formation
[37] all glycoproteins will essentially have the same N-glycan struc-
tures upon kifunensine treatment. The change in N-glycan process-
ing can be seen by a mobility shift towards a slower (due to the
presence of larger oligomannosidic N-glycans) migrating band
(Fig. 3b). Even more important for determination of the glycosyl-
ation status, all N-glycans from kifunensine-treated plants are fully
sensitive to deglycosylation by Endo H (Fig. 3c, d) (see Note 8).
1. Seedlings were grown in 0.5× MS medium (Duchefa) supple-
mented with 0.8 % (w/v) agar and 1 % (w/v) sucrose or on soil
at 22 °C under long day conditions (16 h light/8 h dark).
2. Transfer 10–12-day-old seedlings to liquid 0.5× MS medium
supplemented with 1 % sucrose and 20 μM kifunensine (see
Note 9) .
Fig. 4 Major N-glycan processing pathway in plants. STT3A is one of the two catalytic subunits of the oligosac-
charyltransferase (OST) complex. Kifunensine (Kif) blocks the activity of class I α-mannosidases (MNS1 to
MNS3) resulting in the formation of oligomannosidic N-glycans with Man9GlcNAc2 structures [37]. The sym-
bols for representation of the glycan structures follow the style of the Consortium for Functional Glycomics
(www.functionalglycomics.org). ALG10 α1,2-glucosyltransferase, GCSI α-glucosidase I, GCSII α-glucosidase
II, MNS3 ER-α-mannosidase I, MNS1/MNS2 Golgi-α-mannosidase I, GnTI N-acetylglucosaminyltransferase I,
GMII Golgi-α-mannosidase II, GnTII N-acetylglucosaminyltransferase II, XYLT β1,2-xylosyltransferase, FUT11/
FUT12 core α1,3-fucosyltransferase
3. Incubate at 22 °C with gentle shaking under long day condi-

tions for 24 h (see Note 10).
4. Harvest seedlings, remove excess liquid and extract protein as
described under Subheading 3.2.
5. Perform immunoblotting as described before.
3.4 Endo H Digestion 1. Incubate 22.5 μL of the protein extract (1–20 μg of protein)
of Kifunensine Treated with 2.5 μL 10× glycoprotein denaturing buffer for 10 min at
Plant Material 95 °C, transfer to ice and cool for 5 min (see Note 11).
2. Mix 22.5 μL of the denatured glycoproteins with 3 μL 10×
Glyco 3 buffer (NEB), 3 μL ultrapure water, and 1.5 μL Endo
H (see Note 12). For the control reaction replace Endo H with
ultrapure water.
3. Incubate for 120 min at 37 °C and stop the reaction by heating
to 95 °C for 5 min.
4. Load samples on the SDS-PAGE gel and perform immunob-
lotting, for example, with anti-BRI1 antibody as described
before in Subheading 3.2.
3.5 PNGase F This protocol shows that the protein is N-glycosylated and can
Digestion in Core help to answer the question whether the hormone receptor is traf-
Fucose Deficient ficking through the Golgi. Golgi processed glycoproteins com-
Plants monly carry N-glycans with core α1,3-fucose which are resistant to
PNGase F digestion. A difference in electrophoretic mobility
Glycosylation 215
Fig. 5 (a) PNGase F digestion of BRI1 and AHK2 in Col-0 and in the fut11 fut12 double mutant which is devoid
of core α1,3-fucose {Strasser, 2004}. BRI1 in Col-0 is only partially deglycosylated (only oligomannosidic and
complex N-glycans devoid of core fucose are removed) by PNGase F as visible by the minor shift in mobility
(band >130 kDa). PNGase F treatment of BRI1 in fut11 fut12 results in complete deglycosylation (band
<130 kDa). The major complex N-glycan structure is shown. Again no shift is detectable for AHK2. (b) Schematic
illustration showing the cleavage specificity of PNGase F. Note: PNGase F deaminates the asparagine to aspar-
tic acid. For description of sugar symbols see Fig. 4
between PNGase F-digested wild-type and fut11 fut12 protein

extracts indicates processing in the Golgi (Fig. 5a, b).
1. Perform solubilization and denaturation as described for the

Endo H digestion (see Note 13).
2. Mix 22.5 μL of the denatured glycoproteins with 3 μL 10×
Glyco 2 buffer, 3 μL NP-40, and 1.5 μL PNGase F. For the
control reaction replace PNGase F with ultrapure water.
3. Incubate for 120 min at 37 °C and stop the reaction by heating

to 95 °C for 5 min.
4. Load samples on the SDS-PAGE gel and perform immunob-
lotting for example with anti-BRI1 antibody as described
before in Subheading 3.2.
4 Notes
1. Liquid-chromatography-tandem mass spectrometry-based

analysis of glycoprotein glycosylation is the method of choice
to obtain information on the glycoforms present at a specific
N-glycosylation site [38]. This method which allows also the
simultaneous identification of other posttranslational modifica-
tions has been described in detail recently [39].
2. There are also other N-glycosylation site prediction tools such as
GlycoMine (http://www.structbioinfor.org/Lab/GlycoMine/)
which provide the probability for N-glycosylation.
3. Apart from Asn-X-Ser/Thr there is the rare possibility for the
use of noncanonical glycosylation sites. The best described is
Asn-X-Cys. In addition, on a small number of glycoproteins
from diverse model organisms N-glycosylation of Asn-X-Val
and Asn-Gly has been described in a recent proteomics
approach [40, 41]. While the presence of N-glycosylation at
Asn-X-Cys has been reported for a recombinant protein
expressed in different plant cells [42], the existence of
N-glycosylation at such noncanonical sites remains to be shown
for endogenous plant proteins [41].
4. The sequon is necessary but not sufficient for N-glycosylation.
N-glycosylation efficiency is dependent on many factors involv-
ing the amino acid sequence close to the N-glycosylation site,
the secondary structure, the positioning of the site within the
protein [15] as well as organism-specific factors. The latter are
dependent on the composition and function of the OST com-
plex, which is known to differ between species [43]. Absence or
altered function of a specific OST subunit may cause skipping
of individual N-glycosylation sites resulting in underglycosyl-
ation. Since the OST complex is not very well studied in plants
[44] cell- or tissue-specific variations of N-glycosylation on a
given glycoprotein cannot be ruled out. Organ-specific changes
in N-glycan processing have been reported for the Lewis a epi-
tope formation on complex N-glycans in Arabidopsis [45].
5. N-glycosylation is spatially restricted to the lumen of the
ER. Per definition only proteins that enter the secretory path-
way can be N-glycosylated. Some studies have reported the
presence of N-glycosylated proteins in chloroplasts [46]. These
proteins very likely are transported from the ER/Golgi to
Glycosylation 217
chloroplasts via a largely unknown trafficking route. The pres-

ence of N-glycosylated byproducts of the ERAD pathway in
the cytosol is also possible, but this proteasomal degradation
route which requires retrotranslocation from the ER has not
been clearly shown for endogenous plant glycoproteins.
6. The stt3a mutant grows like wild-type under standard growth
conditions and does not show any severe developmental
phenotype.
7. Arabidopsis STT3A and STT3B are the two different catalytic
subunits of the OST complex. Genetic and biochemical data
suggest that they have overlapping as well as distinct functions
in N-glycosylation. The mechanism of glycosylation site selec-
tion is currently unknown [33]. It is therefore necessary to
experimentally test whether a protein is subjected to STT3A-
dependent glycosylation. Underglycosylation in the stt3b sin-
gle mutant can be tested in the same way.
8. Crucial for this kind of analysis is the removal of all N-glycans
from the glycoprotein. Unfortunately, there is no endoglycosi-
dase available that allows this cleavage. The problem arises from
the presence of core α1,3-linked fucose on complex N-glycans
which blocks the activity of the commonly used deglycosylation
enzyme PNGase F [47]. For subsequent analysis it is therefore
essential to prevent the attachment of the core α1,3-fucose.
Two strategies are shown here: (i) generation of oligomanno-
sidic N-glycans by kifunensine treatment and (ii) analysis of
N-glycosylation in the fut11 fut12 double knockout [34] which
is devoid of active core α1,3-fucosyltransferases.
9. Tunicamycin treatment of plant cells or other plant material is
another possibility to examine the N-glycosylation status of a
protein [7]. However, in our opinion tunicamycin treatment has
several disadvantages: (a) it completely blocks N-glycosylation
and therefore can cause extensive misfolding of glycoproteins
with subsequent triggering of the unfolded protein response in
the ER which eventually leads to cell death. (b) As a consequence
of the misfolding and absence of N-glycans the protein of inter-
est may be more instable or aggregate and the subsequent detec-
tion by immunoblots may be impossible. By contrast, kifunensine
leads only to a modest activation of the unfolded protein response
and glycan-dependent folding is less affected.
10. Kifunensine treatment can be done in different ways.
Germination and growth of seedlings on MS agar supple-
mented with 10 μM kifunensine is possible [37]. Alternatively,
leaves can be detached from soil grown plants, cut into small
pieces and incubated for 24 h in 20 μM kifunensine.
11. Samples in Laemmli buffer can also be used to perform Endo
H and PNGase F digestions.
12. For endoglycosidase treatment it is important to find a balance

between protein denaturation (to enable access of the deglycosyl-
ation enzyme) and protein stability. Prolonged incubation at
37 °C should be avoided to prevent the loss of the glycoprotein.
13. Exoglycosidase digestions are less frequently used to investigate
N-glycosylation of plant glycoproteins. Typically, these enzymes
remove only individual sugar residues. Unless using MS-based
approaches their effect on glycoprotein glycosylation is difficult
to monitor. One useful exception is the digestion of glycopro-
teins with jack bean α-mannosidase. Jack bean α-mannosidase
removes up to eight α-linked mannose residues from oligoman-
nosidic N-glycans and allows to distinguish between complex
and oligomannosidic N-glycans or between oligomannosidic
and glucosylated oligomannosidic N-glycans [30].
Acknowledgements
This work was supported by a grant from the Austrian Science

Fund (FWF): P23906.
References
1. Aebi M, Bernasconi R, Clerc S et al (2010) 8. Strasser R (2012) Challenges in O-glycan

N-glycan structures: recognition and process- engineering of plants. Front Plant Sci 3:218
ing in the ER. Trends Biochem Sci 35:74–82 9. Hijazi M, Velasquez SM, Jamet E et al (2014)
2. Strasser R (2014) Biological significance of An update on post-translational modifica-
complex N-glycans in plants and their impact tions of hydroxyproline-rich glycoproteins:
on plant physiology. Front Plant Sci 5:363 toward a model highlighting their contribu-
3. Lu X, Tintor N, Mentzel T et al (2009) tion to plant cell wall architecture. Front
Uncoupling of sustained MAMP receptor sig- Plant Sci 5:395
naling from early outputs in an Arabidopsis 10. Bond MR, Hanover JA (2015) A little sugar
endoplasmic reticulum glucosidase II allele. goes a long way: the cell biology of O-GlcNAc.
Proc Natl Acad Sci U S A 106:22522–22527 J Cell Biol 208:869–880
4. Nekrasov V, Li J, Batoux M et al (2009) 11. Delporte A, De Zaeytijd J, De Storme N et al
Control of the pattern-recognition receptor (2014) Cell cycle-dependent O-GlcNAc mod-
EFR by an ER protein complex in plant immu- ification of tobacco histones and their interac-
nity. EMBO J 28(21):3428–3438 tion with the tobacco lectin. Plant Physiol
5. Li J, Zhao-Hui C, Batoux M et al (2009) Biochem 83:151–158
Specific ER quality control components 12. Olszewski NE, West CM, Sassi SO et al (2010)
required for biogenesis of the plant innate O-GlcNAc protein modification in plants: evo-
immune receptor EFR. Proc Natl Acad Sci U S lution and function. Biochim Biophys Acta
A 106:15973–15978 1800:49–56
6. Saijo Y, Tintor N, Lu X et al (2009) Receptor 13. Steiner E, Efroni I, Gopalraj M et al (2012)
quality control in the endoplasmic reticulum The Arabidopsis O-linked N-acetylglucosamine
for plant innate immunity. EMBO J 28(21): transferase SPINDLY interacts with class I
3439–3449 TCPs to facilitate cytokinin responses in leaves
7. Häweker H, Rips S, Koiwa H et al (2010) and flowers. Plant Cell 24:96–108
Pattern recognition receptors require 14. Niemann MC, Bartrina I, Ashikov A et al
N-glycosylation to mediate plant immunity. (2015) Arabidopsis ROCK1 transports UDP-
J Biol Chem 285:4629–4636 GlcNAc/UDP-GalNAc and regulates ER pro-
Glycosylation 219
tein quality control and cytokinin activity. Proc 29. Hüttner S, Veit C, Vavra U et al (2014)
Natl Acad Sci U S A 112:291–296 Arabidopsis class I α-mannosidases MNS4 and
15. Shrimal S, Cherepanova NA, Gilmore R MNS5 are involved in endoplasmic
(2015) Cotranslational and posttranslocational reticulum-associated degradation of misfolded
N-glycosylation of proteins in the endoplasmic glycoproteins. Plant Cell 26:1712–1728
reticulum. Semin Cell Dev Biol 41:71–78 30. Jin H, Yan Z, Nam K et al (2007) Allele-
16. Friml J, Jones AR (2010) Endoplasmic reticu- specific suppression of a defective brassino-
lum: the rising compartment in auxin biology. steroid receptor reveals a physiological role of
Plant Physiol 154:458–462 UGGT in ER quality control. Mol Cell
17. Barbez E, Kleine-Vehn J (2013) Divide Et 26:821–830
Impera—cellular auxin compartmentalization. 31. Hong Z, Jin H, Tzfira T et al (2008) Multiple
Curr Opin Plant Biol 16:78–84 mechanism-mediated retention of a defective
18. Qiao H, Shen Z, Huang SS et al (2012) brassinosteroid receptor in the endoplasmic
Processing and subcellular trafficking of reticulum of Arabidopsis. Plant Cell
ER-tethered EIN2 control response to ethyl- 20:3418–3429
ene gas. Science 338:390–393 32. Hong Z, Kajiura H, Su W et al (2012)
19. Caesar K, Thamm AM, Witthöft J et al (2011) Evolutionarily conserved glycan signal to
Evidence for the localization of the Arabidopsis degrade aberrant brassinosteroid receptors in
cytokinin receptors AHK3 and AHK4 in the Arabidopsis. Proc Natl Acad Sci U S A
endoplasmic reticulum. J Exp Bot 109:11437–11442
62:5571–5580 33. Koiwa H, Li F, McCully M, Mendoza I et al
20. Wulfetange K, Lomin SN, Romanov GA et al (2003) The STT3a subunit isoform of the
(2011) The cytokinin receptors of Arabidopsis Arabidopsis oligosaccharyltransferase controls
are located mainly to the endoplasmic reticu- adaptive responses to salt/osmotic stress. Plant
lum. Plant Physiol 156:1808–1818 Cell 15:2273–2284
21. Li J, Chory J (1997) A putative leucine-rich 34. Strasser R, Altmann F, Mach L et al (2004)
repeat receptor kinase involved in brassino- Generation of Arabidopsis thaliana plants with
steroid signal transduction. Cell 90:929–938 complex N-glycans lacking beta1,2-linked
xylose and core alpha1,3-linked fucose. FEBS
22. Pandey S, Nelson DC, Assmann SM (2009) Lett 561:132–136
Two novel GPCR-type G proteins are abscisic
acid receptors in Arabidopsis. Cell 35. Blom N, Sicheritz-Pontén T, Gupta R et al
136:136–148 (2004) Prediction of post-translational glyco-
sylation and phosphorylation of proteins from
23. Kim HJ, Ryu H, Hong SH et al (2006) the amino acid sequence. Proteomics
Cytokinin-mediated control of leaf longevity 4:1633–1649
by AHK3 through phosphorylation of
ARR2 in Arabidopsis. Proc Natl Acad Sci U S 36. Elbein AD, Tropea JE, Mitchell M et al (1990)
A 103:814–819 Kifunensine, a potent inhibitor of the glyco-
protein processing mannosidase I. J Biol Chem
24. Napier RM, David KM, Perrot-Rechenmann 265:15599–15605
C (2002) A short history of auxin-binding
proteins. Plant Mol Biol 49:339–348 37. Liebminger E, Hüttner S, Vavra U et al (2009)
Class I alpha-mannosidases are required for
25. Massotte D, Fleig U, Palme K (1995) N-glycan processing and root development in
Purification and characterization of an auxin- Arabidopsis thaliana. Plant Cell 21:3850–3867
binding protein from Arabidopsis thaliana
expressed in baculovirus-infected insect cells. 38. Liebminger E, Grass J, Jez J et al (2012)
Protein Expr Purif 6:220–227 Myrosinases TGG1 and TGG2 from
Arabidopsis thaliana contain exclusively
26. Feraru E, Vosolsobě S, Feraru MI et al (2012) oligomannosidic N-glycans. Phytochemistry
Evolution and structural diversification of 84:24–30
PILS putative auxin carriers in plants. Front
Plant Sci 3:227 39. Gruber C, Altmann F (2015) Site-specific gly-
cosylation profiling using Liquid
27. She J, Han Z, Kim TW et al (2011) Structural Chromatography-Tandem Mass Spectrometry
insight into brassinosteroid perception by (LC-MS). Methods Mol Biol 1321:407–415
BRI1. Nature 474:472–476
40. Zielinska DF, Gnad F, Wiśniewski JR et al
28. Hothorn M, Belkhadir Y, Dreux M et al (2010) Precision mapping of an in vivo
(2011) Structural basis of steroid hormone N-glycoproteome reveals rigid topological and
perception by the receptor kinase BRI1. sequence constraints. Cell 141:897–907
Nature 474:467–471
41. Zielinska DF, Gnad F, Schropp K et al (2012) 45. Strasser R, Bondili J, Vavra U et al (2007) A
Mapping N-glycosylation sites across seven unique beta1,3-galactosyltransferase is
evolutionarily distant species reveals a diver- indispensable for the biosynthesis of
gent substrate proteome despite a common N-glycans containing Lewis a structures in
core machinery. Mol Cell 46:542–548 Arabidopsis thaliana. Plant Cell 19:
42. Matsui T, Takita E, Sato T et al (2011) 2278–2292
N-glycosylation at noncanonical Asn-X-Cys 46. Villarejo A, Burén S, Larsson S et al (2005)
sequences in plant cells. Glycobiology Evidence for a protein transported through
21:994–999 the secretory pathway en route to the higher
43. Kelleher D, Gilmore R (2006) An evolving plant chloroplast. Nat Cell Biol 7:
view of the eukaryotic oligosaccharyltransfer- 1224–1231
ase. Glycobiology 16:47R–62R 47. Tretter V, Altmann F, März L (1991) Peptide-
44. Farid A, Malinovsky FG, Veit C et al (2013) N4-(N-acetyl-beta-glucosaminyl)asparagine
Specialized roles of the conserved subunit amidase F cannot release glycans with fucose
OST3/6 of the oligosaccharyltransferase com- attached alpha 1-3 to the asparagine-linked
plex in innate immunity and tolerance to abi- N-acetylglucosamine residue. Eur J Biochem
otic stresses. Plant Physiol 162:24–38 199:647–652
Chapter 18
Highly Sensitive Salicylic Acid Quantification

in Milligram Amounts of Plant Tissue
Víctor Carrasco Loba and Stephan Pollmann
Abstract
Technical advances in mass spectrometry constantly raise the bar for analyzing trace amounts of plant
hormones in only very small amounts of tissue. Here, a highly sensitive and accurate method is described
for the quantitative analysis of the plant hormone salicylic acid not only in the model plant Arabidopsis
thaliana but also in other plant species. The presented method is optimized for the working up of as little
as 20 to 50 mg of plant tissue. The discussed protocol and the utilized laboratory equipment facilitate the
implementation of the method into other laboratories that possess access to adequate state-of-the-art gas
chromatography-mass spectrometry (GC-MS) equipment.
Key words Salicylic acid, Electron-impact tandem mass spectrometry (EI-GC-MS/MS), Solid-phase
extraction, Stable isotopes, Derivatization, Plant hormone analysis
1 Introduction
Plant hormones, for instance, auxins, cytokinins, gibberellins,

abscisic acid, salicylic acid, jasmonates, and ethylene, possess piv-
otal functions in regulating plant growth and development [1]. In
this context, salicylic acid plays a particularly important role in the
induction of plant defense responses against a variety of biotic and
abiotic stresses by orchestrating a multitude of morphological,
physiological, and biochemical mechanisms. Among the numerous
functions of salicylic acid its central role in triggering plant immune
responses to pathogens has to be highlighted [2].
By using large-scale molecular genetic approaches, producing
a multitude of publicly available mutant collections and a wealth of
different transgenic plants, very valuable insight into the function
and signaling of plant hormones has been obtained over the last
three decades. However, bearing the essential function of plant
hormones in general and salicylic acid in particular in mind, the
quantitative analysis of those signaling molecules still remains a
challenging task, due to their low abundance in plant tissues.
221
222 Víctor Carrasco Loba and Stephan Pollmann
In order to overcome this problem, modern methods for the

quantitative analysis of such trace compounds nowadays rely on
highly sensitive mass spectrometric techniques. In combination with
the employment of stable-isotope labeled internal standards, mass
spectrometry offers not only the unambiguous identification of
plant hormones but also their absolute quantitation within a given
sample. It has to be remarked that the extraction and pre-purification
protocol presented here is suitable to enrich all kind of acidic plant
hormones at once, which in fact makes multiplex analysis of several
acidic phytohormones, such as auxins, jasmonates, salicylic acid,
abscisic acid, and with modifications of gibberellins, from one sam-
ple possible if necessary GC-MRM-MS/MS equipment is available
[3–5]. Apart from very few entry-level GC-MS machines, modern
GC-MS setups generally possess tandem MS (MS/MS or MSn)
capacities. In brief, working in MS/MS mode refers to the selection
of suitable parent or precursor ions in the first MS step, which are
then fragmented through the collision with a noble gas, most com-
monly argon. The resulting so called fragment, daughter or product
ions are finally recorded in a second MS step. The upside of this
methodology is the substantial improvement of the signal to noise
ratio that offers the possibility to obtain clean spectra for given target
compounds even from very complex samples. On the downside, uti-
lizing a very sensitive technique such as GC-MS renders the estab-
lishment of an efficient pre-purification/sample preparation protocol
mandatory, because primary extracts are generally not considered
suitable for direct assessment by mass spectrometry, even though
exactly this would be desirable. In this respect, it has to be remarked
that there is no one method that facilitates the simultaneous pre-
purification of all plant hormones at once, because of their very dif-
ferent chemical properties. However, mainly because of major
advances in sensitivity, liquid chromatography-mass spectrometry
(LC-MS) moves more and more into the limelight, also in the field
of plant hormone quantification. This is due to the fact that the
LC-MS setup is less prone to malfunction and long-lasting contami-
nation of the hardware when the sample quality is slightly on the low
side. However, it has to be underlined that also in LC-MS compre-
hensive sample preparation protocols are sometimes required to
facilitate analysis of the desired target substances [6–10].
The method presented in this chapter facilitates the quantita-
tive analysis of salicylic acid in small tissue samples. As stated
above, the described purification procedure is also suitable for
several other acidic phytohormones and related substances. The
method has been validated and successfully used for a number of
plant species, for instance Arabidopsis, barley, tobacco, potato,
tomato, corn, and rice. It is, however, not appropriate to purify
basic plant hormones, such as cytokinins and related derivatives.
In this case, the reader is referred to advanced literature for fur-
ther reading [11–16].
Salicylic Acid Detection in Plant Tissue 223
2 Material
2.1 Salicylic Acid Prepare all solvents using reagents in pro analysis or at least
Extraction, Pre- high-performance liquid chromatography (HPLC) grade. If not
purification, directly using the pure compounds, it is recommended to prepare
and Derivatization all solvents freshly and to use them at room temperature (if not
explicitly stated otherwise). The reagents should be stored at
room temperature.
1. Acetic acid.
2. Acetone.
3. Methanol.
4. Diethyl ether, dry, free of peroxides.
5. Chloroform.
6. 2-Propanol.
7. (Trimethylsilyl)diazomethane solution, 2.0 M in diethyl ether,
8. N , O - B i s ( t r i m e t h y l s i l y l ) t r i f l u o r o a c e t a m i d e
(BSTFA) + Trimethylsilyl chloride (TMCS), 99:1 [v/v],
9. Washing solvent for solid-phase extraction: chloroform–2-pro-
panol [2:1, v/v],
10. Elution solvent for solid-phase extraction: diethyl ether + 2 %
acetic acid [v/v],
11. Thermomixer.
12. Aminopropyl solid-phase extraction columns (Chromabond
NH2 shorty 10 mg, Macherey-Nagel GmbH, Düren, Germany).
13. Centrifuge and corresponding rotor for 1.5 ml microcentri-
fuge tubes.
14. SpeedVac concentrator.
15. Extraction vacuum manifold.
16. Membrane vacuum pump stand with 1.5 mbar minimal vacuum.
17. Ball mill with Eppendorf cup adaptor (such as Retsch MM300).
18. Stainless steel balls, 3 mm diameter.
19. Ultrasonic bath (these vary widely in their specifications. We
use a Bansonic 5510E-DTH, Branson Ultrasonics Corporation,
Danbury, CT, USA).
20. 600 μl crimp top conical microvial (Chromacol Uni-VL
Supelco #27312), plus appropriate 8 mm crimp seal PTFE/
rubber for closing.
2.2 Stable-Isotope Stable-isotope labeled salicylic acid, [2H4]-salicylic acid, to be used

Labeled Internal as internal standard in the mass spectrometric analysis is commer-
Standard cially available and can be purchased, e.g., from CDN isotopes
(Pointe-Claire, Canada) or OlChemIm Ltd. (Olomouc, Czech
Republic). There is the possibility to use standards from other ven-

dors (see Note 1) given that the isotopic enrichment of the sub-
stance is higher than 96 %, and the compound does neither decay
nor detectably exchange the heavy isotopes during the sample
preparation and analysis process. Please note that it is highly rec-
ommended that the mass difference between internal standard and
analyte is not less than two AMU.
3 Methods
3.1 Extraction If not explicitly specified otherwise, all steps can be carried out at
of Plant Tissue room temperature.
1. Weigh between 20 to 50 mg of plant tissue (fresh weight) with-
out previous homogenization into a 1.5 ml disposable reaction
tube and immediately add 1 ml of methanol, 30 pmol of [2H4]-
salicylic acid internal standard (see Note 2), and three steel balls.
2. Use a thermomixer to heat the sample to 60 °C and incubate it
under gentle agitation for 3 min, then homogenize by heavy
vortexing or shaking for several minutes. Optimally, a vibrating-
ball micro-mill (Retsch MM300) can be utilized, subjecting the
sample to two rounds of 3 min, each at 30 Hz (see Note 3).
Further extraction will take place at room temperature for at
1 h. The sample should be vortexed or inverted once in a while.
3. Remove the steel balls by using a magnetic tool and centrifuge
the sample (14,000 × g, 1 min) to sediment cell wall debris and
other floating particles. Then, carefully transfer the supernatant
to a new reaction tube, not picking up parts of the pellet. Take
the sample to complete dryness using a SpeedVac set to 60 °C
(see Note 4) and maximum vacuum. Please note that after drying
the sample can be stored at −80 °C until further processing.
3.2 Pre-purification 1. Thoroughly dissolve the crude residue in 50 μl methanol, then

and Enrichment add 200 μl of diethyl ether. Sonify the sample for 5 min in an
of Acidic Compounds ultrasonic bath, before centrifuging it for 5 min at 14,000 × g.
by Solid-Phase 2. During the preparation of the sample, already start with the
Extraction equilibration of the aminopropyl solid-phase extraction col-
umn (see Note 5). To do so, wash the column with 2 × 200 μl
diethyl ether.
3. Transfer the entire sample onto the conditioned column and
let the sample pass the matrix by gravity flow without applying
vacuum to the SPE manifold. This is likely to improve the asso-
ciation of acidic compounds to the NH2-matrix.
4. After passage of the sample, wash the column twice with 200 μl
freshly prepared washing solvent (chloroform:2-propanol).
Clean the tip of the plastic gasket holding the column with
some white laboratory paper.
5. Elute the sample from the matrix into a fresh reaction vessel by
adding 2 × 200 μl of the elution solvent (acidified diethyl ether,
containing 2 % acetic acid [v/v]). Finally, dry the combined
eluates using a SpeedVac concentrator (10 mbar, 60 °C).
Remove residual acetic acid in a gentle stream of nitrogen.
3.3 Derivatization The contained salicylic acid is preferentially analyzed by gas

of the Sample Prior chromatography-mass spectrometry (GC-MS) in form of its methyl
to GC-MS Analysis ester, which is characterized by a substantially lower melting point
and, thus, better vaporization properties during the subsequent
GC-MS analysis.
1. Prepare 10 ml of acetone–methanol (9:1, [v/v]) solution.
When kept in darkness and in an airtight vessel, the mixture can
be stored at room temperature over a longer period of time.
2. To prepare the derivatization solvent for the methylation of
organic acids, mix 220 μl of the acetone–methanol solution with
27 μl diethyl ether and 3 μl of the (trimethylsilyl)diazomethane
solution (see Note 6). The resulting 250 μl are sufficient for the
chemical modification of at least ten samples (see Note 7).
3. Dissolve sample in 25 μl methanol and transfer the solution into
a 600 μl crimp top conical microvial (Chromacol Uni-VL Supelco
#27312). Dry completely in a gentle stream of nitrogen.
4. Add 20 μl of the freshly prepared derivatization solvent to the
sample. Immediately close the vial with a suitable 8 mm crimp
seal with PTFE/rubber septa. Let the sample rest for 30 min at
room temperature before proceeding with the GC-MS analysis.
3.4 Optional In case that handling diazomethane containing substances in the

Derivatization Protocol laboratory is not desired or permitted for whatever reason, salicylic
acid can alternatively and very effectively be chemically modified
by trimethyl silylation.
1. Dissolve sample in 25 μl methanol and transfer the solution
into a 600 μl crimp top conical microvial (Chromacol Uni-VL
Supelco #27312). Dry completely in a gentle stream of
nitrogen.
2. Add 20 μl N,O-Bis(trimethylsilyl)trifluoroacetamide (BSTFA)
+ Trimethylsilyl chloride (TMCS), 99:1 [v/v] to the sample.
Immediately close the vial with a suitable 8 mm crimp seal with
PTFE/rubber septa.
3. Incubate the sample over 30 min at 70 °C before proceeding
with the GC-MS analysis.
3.5 Gas From the derivatized sample, 1 μl aliquots are injected into the
Chromatographic- GC-MS system for gas chromatographic separation and subse-
Mass Spectrometric quent mass spectrometric analysis.
Assessment
3.5.1 GC Setup Injection of the sample was carried out in splitless mode. A pressure
pulse of 25 psi over 1 min was used to force the transfer of com-
pounds from the injector into the column. After that time, the split
was fully opened for 1.5 min before it was set to a split ratio of 20 %
for the remaining run time. The injector temperature was 250 °C
and the column temperature was held at 50 °C for 1.2 min. Then,
it was increased by 30 °C/min to 120 °C, followed by a further
increase to 325 °C by 10 °C/min. Finally, a temperature of 325 °C
was held for another 4 min. Separation was achieved by using a
30 m × 0.25 mm i.d. fused silica capillary column with a chemical
bond 0.25-μm ZB35 stationary phase (Phenomenex, Torrance,
USA) (see Note 8). Helium at a flow rate of 1 mL/min served as
the mobile phase.
3.5.2 Mass Spectrometer The method described here is optimized for triple-quadrupole
Setup mass spectrometers run with positive polarity. The mass spectrom-
eter was operated in EI-MRM mode. The transfer line temperature
was set to 250 °C and the ion source temperature to 200 °C. Ions
were generated with −70 eV at a filament emission current of
80 μA. The dwell time was 175 ms. Argon set at 2.0 mTorr was
used as the collision gas. For both target compounds, derivatized
salicylic acid and derivatized [2H4]-salicylic acid, at least two transi-
tions were recorded. One transition was used as the quantifier ion,
whereas the others served as qualifier ions providing additional
information about the analyte as well as indicating the presence of
possible impurities. The selected precursor ions and corresponding
diagnostic product ions are listed in Table 1.
The amount of the endogenous compound was calculated
from the signal ratio of the unlabeled over the stable isotope-
containing mass fragment observed in the parallel measurements
(Table 1).
4 Notes
1. Stable-isotope labeled salicylic acid is also available from

Cambridge Isotopes (www.isotope.com), Isotec (Sigma-
Aldrich; www.isotec.com, www.sigma-aldrich.com), and
Medical Isotopes (www.medicalisotopes.com).
2. Note that it is mandatory to determine the appropriate amount
of internal standard to be used in preliminary experiments.
The necessary standard content can change for different tissues
and developmental stages, respectively. It is recommended that
the amount of standard added to the sample should neither
exceed the endogenous content of the analyte by five times,
nor should it be less than 1/5 of the respective analyte to be
analyzed.
Table 1
Characteristic precursor and product ions used for the detection of derivatized salicylic acid and its
corresponding internal standard
Retention Scan
Retention time window Precursor Product Collision time Quantifier
Compound time(min) (min) ion (m/z) ion (m/z) energy (%) ion
MeSA 6.31 1 152 120 10 33.33 X
6.31 1 120 91 10 33.33
6.31 1 120 59 15 33.33
2
[ H4]-MeSA 6.31 1 156 124 10 33.33 X
6.31 1 124 95 10 33.33
6.31 1 124 63 25 33.33
SA-diOTMS 10.39 10.39 267 209 10 33.33 X
10.39 10.39 267 149 10 33.33
10.39 10.39 267 73 10 33.33
2
[ H4]-SA- 10.39 10.39 271 213 10 33.33 X
diOTMS
10.39 10.39 271 153 10 33.33
10.39 10.39 271 73 10 33.33
It has to be noted that the retention time for the target compound may moderately shift due to system dependent
parameters. It is recommended to use reference compounds to determine retention times for the actual setup prior to
the analysis (see Note 8)
3. More rigid plant organs, e.g., more lignified tissues such as

stems or seeds, may require longer vibration times. If no
vibrating-ball micro-mill is available, the tissue can alterna-
tively be ground in liquid nitrogen using micro-pistils and
standard reaction tubes. The resulting powder must not thaw
before the addition of the methanol.
4. It is important that the sample is completely dried with no traces
of water left. Unnecessary processing in the SpeedVac concen-
trator, however, should be avoided as overdrying may result in
analyte loss. Given the fact that salicylic acid has its melting point
at approximately 159–161 °C the risk of analyte loss by evapora-
tion is low, though. To our experience, processing times of up to
2 h do not negatively affect analyte contents.
5. It is possible to use commercially available microtips from
other vendors, e.g., from Glygen Corp., Columbia, USA or
Chromacol Ltd., Welwyn Garden City, UK, for solid-phase
extraction. Please note, however, that the presented proto-
col has not been optimized for such type of pipet microtips
and, thus, needs to be adapted in case that usage of micro-
tips is desired.
6. Alternatively, ethereal diazomethane can be used for the meth-
ylation. Ethereal diazomethane can be prepared from N-
nitrosomethylurea recrystallized from methanol.
N-nitrosomethylurea can be obtained from Sigma-Aldrich.
However, in any case, when preparing and/or handling diazo-
methane containing solutions safety precautions should strictly
be obeyed. All steps have to be performed in a well-ventilated
laboratory fume hood.
7. It is recommended to prepare only the amount of derivatiza-
tion solution required for immediate use. The solvent should
only be used freshly prepared.
8. GC-capillary columns with an intermediate polarity similar to
that of the here used ZB-35 column (35 % phenyl- 65 %
dimethylpolysiloxane, low bleeding) are standard for acidic
phytohormone profiling. However, other stationary phases
(e.g., containing 5 % or 50 % phenyl) that provide either lower
or higher polarity can also be used. It has to be noted that such
columns may need a slightly different temperature program
than the one reported here to achieve satisfying separation of
the analytes.
References
1. Davies PJ (2010) Plant hormones. Biosynthesis, of lettuce (Lactuca sativa L.) seeds. Plant
signal transduction, action! 3rd edn. Kluwer J 35:405–417
Academic, London, Revised 7. Ziegler J et al (2014) Simultaneous analysis of
2. Yan S, Dong X (2014) Perception of the plant apolar phytohormones and 1-aminocyclo
immune signal salicylic acid. Curr Opin Plant propan-1-carboxylic acid by high performance
Biol 20:64–68 liquid chromatography/electrospray negative
3. Müller A et al (2002) A multiplex GC-MS/MS ion tandem mass spectrometry via 9-fluorenyl-
technique for the sensitive and quantitative methoxycarbonyl chloride derivatization.
single-run analysis of acidic phytohormones J Chromatogr A 1362:102–109
and related compounds, and its application to 8. Strehmel N et al (2014) Profiling of secondary
Arabidopsis thaliana. Planta 216:44–56 metabolites in root exudates of Arabidopsis
4. Schmelz EA et al (2003) Simultaneous analysis thaliana. Phytochemistry 108:35–46
of phytohormones, phytotoxins, and volatile 9. Wang X et al (2014) Quantitative profiling
organic compounds in plants. Proc Natl Acad method for phytohormones and betaines in
Sci U S A 100:10552–10557 algae by liquid chromatography electrospray
5. Birkemeyer C et al (2003) Comprehensive ionization tandem mass spectrometry. Biomed
chemical derivatization for gas chromatography- Chromatogr 28:275–280
mass spectrometry-based multi-targeted profil- 10. Novák O et al (2012) Tissue-specific profiling
ing of the major phytohormones. J Chromatogr of the Arabidopsis thaliana auxin metabolome.
A 993:89–102 Plant J 72:523–536
6. Chiwocha SD et al (2003) A method for profil- 11. Astot C et al (1998) Precolumn derivatization
ing classes of plant hormones and their metab- and capillary liquid chromatographic/frit-fast
olites using liquid chromatography-electrospray atom bombardment mass spectrometric analy-
ionization tandem mass spectrometry: an anal- sis of cytokinins in Arabidopsis thaliana. J Mass
ysis of hormone regulation of thermodormancy Spectrom 33:892–902
12. Jameson PE et al (2000) Cytokinins. chromatography–tandem mass spectrometry.

Extraction, separation, and analysis. Methods In: Normanly J (ed) High-throughput pheno-
Mol Biol 141:101–121 typing in plants. Humana, New York,
13. Farrow S, Emery RN (2012) Concurrent pro- pp 151–164
filing of indole-3-acetic acid, abscisic acid, and 15. Takei K et al (2003) A method for separation
cytokinins and structurally related purines by and determination of cytokinin nucleotides
high-performance-liquid-chromatography tan- from plant tissues. J Plant Res 116:265–269
dem electrospray mass spectrometry. Plant 16. Novák O et al (2008) Cytokinin profiling in
Methods 8:42 plant tissues using ultra-performance liquid
14. Kojima M, Sakakibara H (2012) Highly sensi- chromatography–electrospray tandem mass
tive high-throughput profiling of six phytohor- spectrometry. Phytochemistry 69:2214–
mones using MS-probe modification and liquid 2224
Chapter 19
High-Resolution Cell-Type Specific Analysis of Cytokinins

in Sorted Root Cell Populations of Arabidopsis thaliana
Ondřej Novák, Ioanna Antoniadi, and Karin Ljung
Abstract
We describe a method combining fluorescence-activated cell sorting (FACS) with one-step miniaturized
isolation and accurate quantification of cytokinins (CKs) using ultra-high performance liquid chromatog-
raphy–tandem mass spectrometry (UHPLC-MS/MS) to measure these phytohormones in specific cell
types of Arabidopsis thaliana roots. The methodology provides information of unprecedented resolution
about spatial distributions of CKs, and thus should facilitate attempts to elucidate regulatory networks
involved in root developmental processes.
Key words Cytokinins, Fluorescence-activated cell sorting (FACS), Solid-phase extraction (SPE),
Liquid chromatography–mass spectrometry (LC-MS), Roots, Arabidopsis, Protoplasts
1 Introduction
Hormonal interactions affect most (if not all) aspects of plant

growth and development. Dissection of the interactive regulatory
networks involved frequently requires accurate quantitative infor-
mation on plants’ hormonal contents at a fine tissue level. Thus,
methods for measuring cell-type specific auxin levels have been
developed and applied [1], but developing corresponding meth-
ods for cytokinin (CK) quantification has proved to be much more
challenging [2]. This is partly because CK levels are at least tenfold
lower than auxin levels, and thus more difficult to measure in small
amounts of tissue. Furthermore, CK metabolite profiles are much
more complex, requiring simultaneous measurements of multiple
compounds. The methodology presented here increases the reso-
lution of CK quantification by permitting targeted isolation of spe-
cific cell populations, efficient pre-concentration of cytokinin
metabolites and their detailed profiling from sets as small as just
20,000 specific cells.
The method involves use of flow cytometry-based fluorescence-
activated cell sorting (FACS), which enables the recognition of
231
232 Ondřej Novák et al.
Fig. 1 Isolation of GFP-expressing protoplasts using flow cytometry. (a) The scatter of the laser beam by the
cells is used to count the cells and measure cell size and intracellular granularity/homogeneity. (b) Cell distri-
bution based on FSC-A (Forward-scattered light) vs SSC-A (Side-scattered light). (c) Dot plot with a gate
encompassing the protoplast population (P1). (d, e) Plots of green fluorescence intensity (FITC-A; excitation
488 nm/emission 530/30 nm) and red fluorescence intensity (PA-Texas Red-A; excitation 488 nm/emission
576/26 nm). (d) Scatter plot of protoplasts from wild-type Arabidopsis plants not expressing GFP with most
cells having a typical autofluorescence pattern (P2 population). (e) Sorting of protoplasts from an Arabidopsis
GFP-expressing line with a gate (P3) set to collect the GFP tail where cells display more green than red fluo-
rescence. (f) Single-parameter histogram displaying relative fluorescence intensity (FITC-A) on the x-axis and
the cell count on the y-axis of the gated protoplast population (P2 blue-GFP−; P3 green-GFP+)
isolated protoplasts of targeted size, granularity, and/or complexity,

followed by their sorting into single-cell units or homogenous
groups according to the presence or absence of internal fluoro-
chromes. A large collection of transgenic plant lines expressing
fluorescent reporter genes, especially green fluorescent protein
(GFP), under specific promoters is now available, enabling the
labeling of spatially well-defined cell types (NASC, the European
Arabidopsis Stock Centre; http://arabidopsis.info/). As shown in
Figs. 1 and 2, GFP-based isolation of protoplasts from an orga-
nized plant tissue of such a transgenic line enables separation of
single cells while maintaining their cellular identities [3, 4]. Cell
populations isolated using flow cytometry into GFP-expressing
(GFP+) and non-GFP-expressing (GFP−) protoplasts (Fig. 1) can
be collected via FACS (Fig. 2) and used to analyze cell type-specific
metabolite profiles ([5–7], see Note 1).
Cell-Type Specific Analysis of Cytokinins 233
Fig. 2 Fluorescence-activated cell sorting (FACS) of plant protoplasts expressing green fluorescent protein
(GFP). (a) Fluorescent image of the root tip of the Arabidopsis thaliana pSCR:GFP line [3]. (b, c) The resulting
GFP-expressing (GFP+; b) and non-GFP-expressing (GFP−; c) protoplasts after digestion of the root with cell
wall-degrading enzymes (FM4-64 staining in magenta; GFP fluorescence in green). (d) Schematic illustration
of fluorescence-activated cell sorting. The cells are individually hit by a laser and consequently light is scat-
tered and fluorescence is emitted. These light signals are perceived by detectors and enable characterization
of each cell according to its size, granularity, complexity, and fluorescence. The desired cell populations are
then selected by gates designed to indicate the GFP− and GFP+ cells to be sorted and collected. Stream vibra-
tion causes the formation of single droplets (each containing a characterized cell), which are separated from
the main stream. An electrical charge is applied to the stream and the newly formed drops become positively
or negatively charged according to the gating strategy while the cells/particles excluded from the gates remain
uncharged. The charged droplets are then deflected to the left or right by charged electrodes into the appropri-
ate GFP+ and GFP− cell collection tubes, while the uncharged droplets go straight to the waste aspirator
The methodology also exploits recent advances in mass spec-

trometry that have sharply increased the sensitivity and selectivity of
final analyses of complex plant samples, particularly ultra-high per-
formance liquid chromatography–electrospray tandem quadrupole
mass spectrometry (UHPLC-MS/MS), which has been widely used
in plant hormone profiling [8]. As shown here, in combination with
recently developed micro-purification techniques for cytokinin isola-
tion, such as multilayer solid-phase microextraction [9], batch
immunoaffinity chromatography [10], magnetic solid-phase
extraction [11], and monolithic molecularly imprinted solid-phase

extraction [12], UHPLC-MS/MS has the potential to provide cell-
specific quantification of CK metabolites.
The methodology presented here also includes a powerful
one-step high-throughput micro-purification (microSPE) approach
for extracting cytokinins from only a few thousand sorted cells. It
is based on the so-called StageTip (Stop and Go Extraction Tip)
technology recently employed for isolating CKs from minute
amounts of fresh plant tissues (1–5 mg) using a combination of
reverse-phase and cation-exchange polymer-based sorbents [8].
Exploiting the high selectivity, affinity and capacity of the selected
phases, we subsequently optimized the microSPE protocol for effi-
cient isolation and enrichment of CKs from a minimal quantity of
protoplasts [2]. Finally, the whole procedure was coupled with a
highly selective and extremely sensitive UHPLC-MS/MS method
for screening 26 isoprenoid CK metabolites. In summary, the pro-
toplast isolation and cell sorting procedure combined with rapid
microSPE and ultrasensitive UHPLC-MS/MS can be used for
high-resolution measurements of CKs in specific Arabidopsis root
cell populations.
2 Materials
All solutions are prepared at room temperature using ultrapure

water (e.g., Milli-Q water with ≤18 M W /cm resistance at 25 °C).
All chemicals and solvents should be at least of analytical reagent
grade.
2.1 Protoplast 1. 1 × MS growth media (3 L): 0.44 % (w/v) Murashige and

Isolation for One FACS Skoog salt mixture, 1 % (w/v) sucrose, 0.5 % (w/v) MES
Experiment (pH 5.7, adjusted with KOH), 1 % (w/v) plant agar.
2. Thirty square petri dishes (120 mm × 120 mm).
3. Thirty sterile square-cut nylon meshes (Sefar Nitex 03-110/47).
4. Sterilized seeds of a transgenic line expressing GFP in a specific
cell population (~9000 seeds or ~500 μL volume measured in
a microcentrifuge tube).
5. 100 mL PB solution: 600 mM mannitol, 2 mM CaCl2, 10 mM
KCl, 2 mM MES, 2 mM MgCl2, 0.1 % (w/v) BSA, pH 5.7 (see
Note 2).
6. 25 mL PBI solution: 45.45 mg pectolyase (0.3 U/mL) and
450 mg cellulysin (45 U/mL) in 25 mL PB solution (see Note 3).
7. Surgical blade and forceps.
8. Aluminum foil.
9. Orbital shaker.
10. 40 μm cell strainer (BD Falcon).

11. Refrigerated swing-out centrifuge.
2.2 Fluorescent 1. 5 L of 0.7 % (w/v) sodium chloride (Sheath Fluid) in sterile

Activated Cell Sorting container.
(FACS) 2. Sterile 500 mL Bottle Top Filter with 75 mm Supor® machV
PES Membrane, 0.2 μm pore diameter (Nalgene® 595-4520
Rapid-Flow™, 45 μm, 33 mm Blue Neck, Thermo Scientific).
3. FACS (BD FACS Aria I Flow Cytometer, BD Biosciences).
4. Software: BD FACSDiva version 6.1.2.
5. FACS Rinse and FACS Clean solutions (Cat. Nos. 340346 and
340345, respectively, BD Biosciences).
6. CS&T and Accudrop beads (BD Biosciences).
7. Cooling system connected to FACS.
8. 100 μm nozzle (P/N 342909, BD Biosciences).
9. Rubber O-Ring (70SL, Apple Rubber Products).
10. 50 μm cup filter (Cat. No. 340631, BD Biosciences).
11. Round-bottom tubes 12 × 75 mm with lid for sample collection
(Cat. No. 352054, BD Falcon) and without lid for sample load-
ing and cleaning processes (Cat. No. 14-961-26, Fisher Scientific).
2.3 Cytokinin The volumes in parenthesis are for 50 samples.

Purification
1. Stable isotope-labeled internal standards (IS) (see Note 4):
Mixture I—1 × 10−8 M (0.01 pmol/μL; 500 μL) of [13C5]cis-
zeatin (cZ), [13C5]trans-zeatin (tZ), [2H3]dihydrozeatin (DHZ),
[2H6]N6-(Δ2-isopentenyl)adenine (iP), [2H5]tZ riboside (tZR),
[2H3]DHZ riboside (DHZR), [2H6]iP riboside (iPR), [2H5]tZ-
7-N-glucoside (tZ7G), [2H6]iP-7-N-glucoside (iP7G):
Mixture II—2 × 10−8 M (0.02 pmol/μL; 500 μL) of [2H5]tZ-
O-glucoside (tZOG), [2H7]DHZ-O-glucoside (DHZOG),
[2H5]tZ-9-N-glucoside (tZ9G), [2H3]DHZ-9-N-glucoside
(DHZ9G), [2H6]iP-9-N-glucoside (iP9G):
Mixture III—1 × 10−7 M (0.1 pmol/μL; 500 μL) of
[ H5]tZR-O-glucoside (tZROG), [2H5]tZR 5′-monophos-
2
phate (tZRMP), [2H3]DHZR 5′-monophosphate (DHZRMP),

and [2H6]iPR 5′-monophosphate (iPRMP).
2. Multi-StageTips (Fig. 3; see Note 5)—microSPE columns
packed with Empore™ High Performance Extraction Disks
(3-layers of each sorbent type C18, SDB-RPS and Cation-SR).
3. Microcentrifuge tube, 2 mL with lid (×2).
4. 100 % acetone (2.5 mL).
5. 100 % methanol (HPLC grade) (5 mL).
Fig. 3 Step-by-step guide for preparing a multi-StageTip. (a) The pipette tip, Empore™ High Performance
Extraction Disk on a petri dish, cutter (blunt-ended syringe needle) and plunger (rod needle); (b) cutting a small
disk, with approximately 1.0 mm diameter and 0.5 mm thickness (I–III, the cutter is gently pressed into the
Empore disk, thus introducing a small disk of the sorbent material into the needle); (c–d) insertion of the disk
into the pipette tip (200 μL) using the cutter and plunger fitted into the needle; (e) placement of an additional
disk onto the first disk; (f) prepared multi-StageTips in a lid of the microcentrifuge tube (2.0 mL)
6. 50 % (v/v) nitric acid (2.5 mL): add 1.92 mL of 65 % HNO3 to

0.58 mL water.
7. 1 M hydrochloric acid (375 μL): add 33 μL of 35 % HCl to
342 μL water.
8. 0.35 M ammonium hydroxide in 60 % methanol (2.5 mL): mix
1.5 mL methanol with 0.91 mL water, then add 0.089 mL
concentrated aqueous ammonia (≥25 % in H2O).
9. Refrigerated centrifuge.
10. Centrifuge vacuum concentrator (e.g., SpeedVac).
11. Screw top autosampler vials (volume 2 mL) and glass inserts
(volume 250 μL).
2.4 UHPLC-MS/MS 1. Acquity UPLC® CSH C18 Column, 130 Å, 1.7 μm,
Method 2.1 × 150 mm (Waters, Milford, MA, USA).
2. Methanol (LC-MS grade).
3. 15 mM ammonium formate (pH 3.95): add 566 μL formic
acid (~98 %, LC-MS grade) to 1 L ultrapure water and adjust
to pH 3.95 with ammonium hydroxide (concentration ≥25 %
in H2O).
4. UHPLC-MS/MS system and associated equipment. The LC
instrument should include at least binary ultrahigh-pressure
pumps, a column heater and an autosampler with temperature-
controlled system. The MS system should include a tandem

quadrupole mass spectrometer.
3 Methods
The initial step is collection of samples (GFP+ and corresponding

GFP− cell populations, see Note 6). The time required for sorting
will depend mainly on the GFP expression pattern and may be
substantial (e.g., 2–3 weeks for 6–10 biological replicates of a
transgenic line expressing GFP in 20 % of the cells, see Note 7).
When the samples have been collected, CK purification using the
high-throughput micropurification process presented here can be
performed within six hours (for up to 48 samples). Finally with the
rapid UHPLC-MS/MS method described in Subheading 3.4, less
than one day is required to analyze 72 samples.
3.1 Material 1. Wrap mesh nylon squares in double aluminum foil and sterilize
Preparation for One at 70 °C for 2–3 h.
Sorting Experiment 2. Prepare, autoclave, and pour 3 L of MS growth medium in
(3–4 Days) square petri dishes (approximately 60 ml/dish). When the
medium has just solidified, transfer the mesh with forceps
under sterile conditions onto its surface (see Note 8). The
mesh should instantly darken slightly as it absorbs moisture
from the medium (the use of mesh in the process is optional).
Store the prepared dishes at least overnight at 4 °C.
3. Sow sterilized seeds of the transgenic GFP-expressing line by
pipetting them in a water suspension, in three dense rows on
the petri dishes containing the MS medium covered with mesh
(approximately 100 seeds per row).
4. Store for 2–3 days in the dark at 4 °C to stimulate and synchro-
nize germination.
5. Transfer the plates to the growth chamber and grow the seed-
lings under long day conditions for 8 days (see Note 9).
3.2 Protoplast Protoplasts are fragile and susceptible to mechanical damage,

Isolation for One Cell therefore gentle handling during all steps and use of a swing-out
Sorting Experiment centrifuge is highly recommended.
(3–4 h) 1. On the ninth day (see Note 10) make fresh PBI buffer and
keep at room temperature in the dark until use. Harvest the
roots rapidly, by opening one petri dish at a time and quickly
snipping each row of seedlings’ roots at the same point (see
Note 11).
2. After snipping each row of roots, gently collect them with the
forceps (see Note 12) and quickly place them in PB buffer,
ensuring that the whole root tissue is covered by the solution.
3. Transfer the PBI buffer into a 100 mL flask and ensure that the
enzymes added are fully dissolved (i.e., there is no precipitate).
After rinsing the harvested roots with PB, remove the excess buffer
and transfer them to the flask containing PBI (see Note 13).
4. Incubate for 2 h in the dark at room temperature, with shaking
at 100 rpm.
5. Filter the protoplasts through a 40-μm cell strainer into a
50 mL Falcon tube (see Note 14). Rinse the incubation flask
and all equipment involved with 25 mL PBS buffer and collect
in the same Falcon tube.
6. Centrifuge (1000 × g, 4 °C) the collected PBS-protoplast solu-
tion (50 mL) for 3 min and carefully discard as much superna-
tant as possible (see Note 15).
7. Gently redissolve protoplasts in 1 mL of 0.7 % w/v sodium
chloride (sheath fluid for cell sorting) and maintain at 4 °C
until loading into the FACS.
3.3 Cell-Sorting There are very few generally applicable parameters in the cell-sorting
(3–4 h) process since most of them depend on the FACS equipment avail-
able, the fluorescent transgenic plant line (e.g., percentage of fluores-
cent cells) and the tissue being sorted (e.g., root cells have lower
autofluorescence levels than shoot cells). Some adjustment of param-
eters may even be required when processing the same plant line and
tissue (biological replicates) on different days (see Note 16).
Therefore, here we describe the process according to our set-up
using a BD FACS Aria I Flow Cytometer (BD Biosciences) with a
100 μm nozzle, as described in Subheading 3.2, focusing on the gen-
eral parameters and tips for a successful sorting experiment. Terms in
italic font are equipment-specific commands for the FACS Aria I
instrument. However, the purpose of the commands is also explained
to facilitate application of the procedures with any system.
Before starting the tissue harvesting (Subheading 3.2, steps
1–3).
1. Turn on the FACS instrument so that the cytometer lasers start
to warm up.
During the enzymatic digestion of cell walls in the root tis-
sue (Subheading 3.2, step 4), which takes 2 h, perform the
following set-up procedures:
2. Perform Long-clean with 70 % EtOH to clean the system (see Note
17).
3. Run fluidics start up twice to ensure complete ethanol removal
from the system.
4. Turn on the stream (sheath fluid: 0.7 % sodium chloride fil-
tered through a 0.2 μm Bottle Top Filter) and run quality con-
trol tests to check that the cytometer is performing adequately
(CS&T application).
5. Let the stream run for 15–30 min to allow the flow to stabilize
while collecting the protoplasts (Subheading 3.2, steps 5–7).
6. Adjust the drop breakoff by fine-tuning the amplitude and fre-
quency settings, which will automatically modify the Gap and
Drop 1 values representing the output of breakoff parameters
(actual values; see Note 18). The optimal breakoff parameters
in our system are: frequency ~30, amplitude ~40–60, Drop
1 ~ 100–300 and Gap ~10.
7. Enter the modified Gap and Drop 1 actual values as the manu-
ally imported target values and turn on the Sweet Spot. This
function promotes stability of the flow and breakoff position
with the given digital parameters (Drop 1 and Gap), via slight
automatic alterations of the amplitude if and when necessary.
8. When the flow has been stabilized via the Sweet Spot function,
set the Drop Delay. This is the exact time required for a cell
within the stream to move from the interrogation point to the
droplet break-off position (see Note 19), as determined by an
Accudrop beads solution.
9. Finally perform a test sort by turning on the voltage to charge
the deflection plates (see Note 20) and the waste drawer so the
main and side streams can be visually observed when entering
the collection tubes (do not touch the deflection plates, there
is high voltage!). If all streams reach the appropriate destina-
tions (waste for the main stream and collection tubes for the
side streams) sorting of the sample can start.
Cell-sorting of GFP+ and GFP− protoplasts:
10. After filtering the redissolved isolated protoplasts (Subheading 3.2,
step 7) through a 0.5 μm cup filter, collect them in a round bot-
tom tube and load them into the FACS. Make sure that the sys-
tem is kept at 4 °C (from the collection chamber to the collection
apparatus) and that the sample is agitated to avoid protoplasts
settling in the bottom of the tube.
11. Set the criteria (gates) for sorting and collecting the desired cell
populations. When the laser hits a cell or particle, at the so-called
interrogation point, the light will be reflected and scattered at all
angles (Fig. 1a). The forward scatter provides information about
the cell’s or particle’s size and the side scatter about its granular-
ity and complexity (see Note 21). A two-dimensional forward-
side scatter plot allows identification of different cell populations
and/or distinction between intact cells and smaller particles (see
Note 22) in the injected sample (Fig. 1b). The protoplasts are
manually selected/gated as the P1 population (Fig. 1c), then
further examined for the presence or absence of an internal
fluorophore (autofluorescence and/or GFP). While a cell is
being interrogated, the laser may also strike an internal fluoro-
phore and then a fluorescent signal is emitted with a specific
wavelength. In such cases a two-dimensional plot showing Red

and Green Fluorescence will present one cell population (non-
GFP fluorescent line, Fig. 1d) or two cell populations (GFP fluo-
rescent line, Fig. 1e). Then the GFP− and GFP+ cells can be
selected in gates P2 and P3, respectively (Fig. 1e, see Note 23).
12. After all the parameters have been set, quality control tests
have been performed and the desired cell populations have
been gated and assigned a collection tube position, sorting can
be initiated. Cell sorting is always performed in high purity
mode and at a flow rate not higher than 6.
13. After 2 h of collection (replacing the collection tubes with new
ones when 200,000 cells or some other desired number of cells
has been collected, see Note 6) freeze the samples in liquid
nitrogen and store them at −80 °C until further analysis.
14. Clean the equipment prior to shutting down by sequentially
running the instrument with FACS Clean, FACS Rinse, and
finally ultrapure water (for 5–10 min with each fluid). Turn off
the stream then remove and clean the nozzle. Once a week also
clean the flow cell and sonicate the nozzle. Finally, perform
Fluidics Shut Down and turn off the FACS.
3.4 Cytokinin The microSPE protocol can be applied after the cell sorting experi-
Purification ments have been completed and all samples containing GFP-
from Sorted Cells (6 h) expressing and non-GFP-expressing cell populations have been
collected. The total number of samples purified by the one-step
isolation protocol depends on the number of positions in the cen-
trifuge used.
1. Prepare two microcentrifuge tubes which will be used for the
loading, washing and elution steps (Subheading 3.4, step 8).
Remove the lids from both tubes and pierce a hole in the cen-
ter of one lid using forceps (the other lid can be discarded).
2. Insert the self-packed multi-StageTip containing C18/SDB-
RPS/Cation-SR layers (see Note 24; Fig. 3) into the hole of
the lid, install it on the tube and place the tube in a centrifuge
(see Note 25).
3. Let the samples thaw on ice (see Note 26) and dilute 1 mL of
isolated protoplasts in 0.7 % NaCl 3:1 (v/v) by adding 333 μL
of H2O. Finally, acidify the samples to pH 2.7 with 1 M HCl
(7.5 μL) to enable CK binding to the Stage Tips.
4. Add 10 μL of working solutions I–III of isotope-labeled CK
standards to each sample (see Note 4).
5. Before loading the sample, activate the multi-StageTip sor-
bents sequentially with 50 μL acetone, methanol, water, 50 %
(v/v) nitric acid, and water (step-by-step centrifugation at 434
× g; 15 min for each solution; keep the centrifuge temperature
at 4 °C) (see Note 24).
6. Bind CK analytes in the diluted samples by sequentially apply-

ing seven 200 μL portions to the in-tip microSPE column and
centrifuging for 30 min at 678 × g and 4 °C (see Note 27).
7. Rinse the tips sequentially with 50 μL of water and methanol (by
loading and centrifuging for 20 min at 525 × g and 4 °C).
8. Collect bound CK analytes in new clean microcentrifuge tubes,
prepared in step 1 (Subheading 3.4), by elution with 50 μL of
0.5 M NH4OH in 60 % MeOH (with centrifugation for 20 min
at 525 × g and 4 °C).
9. Transfer the eluates to chromatography vials with microvol-
ume inserts (150–200 μL recommended).
10. Evaporate to dryness in a SpeedVac concentrator (at 37 °C)
and store at −20 °C until UHPLC-MS/MS analyses.
3.5 Cytokinin CK metabolites can be separated and determined using various

Quantification LC-MS/MS systems. Our protocol was developed for use with a
1290 Infinity Binary LC System coupled to a 6490 Triple Quad
LC/MS System with Jet Stream and Dual Ion Funnel technologies
operating in positive mode (Agilent Technologies). Each technical
parameter should be modified and optimized according to the
apparatus used (see Note 28). However, generally:
1. Dissolve the pre-purified samples on multi-StageTips in 40 μL
of 10 % methanol and place them in an autosampler tray pre-
cooled to 4 °C.
2. Inject 10 μL of each sample onto a reversed-phase column
(Acquity UPLC® BEH C18, 1.7 μm, 2.1 × 150 mm). The col-
umn thermostat should be set at 45 °C (see Note 29).
3. Separate the samples using the following 20 min gradient of
methanol (A) and 15 mM ammonium formate (pH 3.95, B) at
a flow rate of 0.35 mL /min: 0 min, 10:90 (A:B)—10.0 min,
23:77 (A:B)—15.0 min, 36:64 (A:B). At the end of the elution
the column was washed with 100 % A (2 min) and equilibrated
to initial conditions (10:90, A:B) for 3 min. A chromatographic
separation of 26 CK metabolites is shown in Fig. 4.
4. Determine the endogenous CKs in protoplasts by multiple
reaction monitoring (MRM) of the protonated precursor and
appropriate product ions. The MRM transitions, ionization
and collision energies, and retention times, are summarized in
Table 1. The optimized settings for the 6490 Triple Quad
LC/MS System are also specified in Note 29.
5. Process data by mass spectrometry software (e.g., Agilent
MassHunter Workstation Software—Quantitative Analysis, ver-
sion B.05.02) and determine concentrations of the CKs using the
stable isotope dilution method (see Note 30). Further data nor-
malization is performed by calculation of the GFP+/GFP− ratio of
the CK concentrations as described in [1, 2] (see Note 6).
Fig. 4 Representative multi-MRM chromatogram of 26 isoprenoid cytokinins. 100 fmol of each derivative was
injected and separated by UHPLC-MS/MS
4 Notes
1. The FACS samples are maintained at 4 °C, from loading in the

machine to collection of the sorted cells. This will inhibit enzy-
matic processes and thus help ensure that levels of CKs subse-
quently measured correspond to the endogenous levels.
However, appropriate controls should be included, which
depend mainly on the experimental purposes.
2. To a 200 mL beaker add 50 mL of Milli-Q water, 10.93 g
mannitol (mw = 182.17), 0.02 g CaCl2 (mw = 110.99), 0.074 g
KCl (mw = 74.56), 0.0386 g MES (mw = 195.2), 0.04 g MgCl2
(mw = 203.3) and 0.1 % w/v BSA. Stir at room temperature
until all chemicals are dissolved and adjust the pH to 5.7 with
0.1 M KOH. Transfer the buffer into a 100 mL graduated cyl-
inder and add Milli-Q water to 100 mL. The PB solution can
be stored at 4 °C in darkness for up to 5 days.
3. The PB solution should be at room temperature during addi-
tion of the cell wall-degrading enzymes (cellulysin and pectoly-
ase). The resulting PBI buffer should be kept in the dark at all
times.
4. The unlabeled and labeled CK standards can be purchased
from OlChemim Ltd. (Olomouc, Czech Republic, www.
olchemim.cz). Dissolve the compounds in methanol to a final
concentration of 1 mM (stock solutions) and store at
−20 °C. Working solutions (I–III) containing each of the sta-
ble isotope-labeled standards can be made by diluting the stock
solutions of the standards to the following final concentrations
Table 1
MRM transitions, retention times, and optimized instrument settings
MRM transition MRM transition

RT Fragmentor CEc
Compound RTa (min) Precursor Products ISb (min) Precursor Products (V) (V)
tZ 10.15 220.1 136.1 [13C5]tZ 10.14 225.1 141.1 380 19
13
cZ 11.28 220.1 136.1 [ C5]cZ 11.27 225.1 141.1 380 19
2
DHZ 10.91 222.1 136.1 [ H3]DHZ 10.82 225.1 136.1 380 23
2
iP 17.88 204.1 136.1 [ H6]iP 17.79 210.1 136.1 380 16
2
tZR 12.92 352.2 220.1 [ H5]tZR 12.83 357.2 225.1 380 20
cZR 13.68 352.2 220.1 380 20

2
DHZR 13.64 354.2 222.1 [ H3]DHZR 13.58 357.2 225.1 380 22
2
iPR 18.15 336.2 204.1 [ H6]iPR 18.10 342.2 210.1 380 20
2
tZ7G 7.71 382.2 220.1 [ H5]tZ7G 7.61 387.2 225.1 380 21
(±)DHZ7G 8.61/8.92 384.2 222.1 380 23

2
iP7G 13.16 366.2 204.1 [ H6]iP7G 13.05 372.2 210.1 380 21
2
tZ9G 8.93 382.2 220.1 [ H5]tZ9G 8.83 387.2 225.1 380 21
cZ9G 9.63 382.2 220.1 380 21

2
DHZ9G 9.61 384.2 222.1 [ H3] 9.54 387.2 225.1 380 23
DHZ9G
iP9G 16.44 366.2 204.1 [2H6]iP7G 16.37 372.2 210.1 380 22

2
tZOG 9.52 382.2 220.1 [ H5]tZOG 9.42 387.2 225.1 380 21
cZOG 10.35 382.2 220.1 380 18

2
DHZOG 10.97 384.2 222.1 [ H7] 10.80 387.2 225.1 380 21
DHZOG
tZROG 11.96 514.2 382.2 [2H5]tZROG 11.88 519.2 387.2 380 21
cZROG 12.71 514.2 382.2 380 21
DHZROG 13.29 516.2 384.2 380 22
tZRMP 8.36 432.2 220.1 [2H5]tZRMP 8.24 437.2 225.1 380 21
cZRMP 9.01 432.2 220.1 380 21
DHZRMP 8.87 434.2 222.1 [2H3] 8.79 437.2 225.1 380 23

DHZRMP
iPRMP 15.31 416.2 204.1 [2H6]iPRMP 15.22 422.2 210.1 380 22

a
Retention time
b
Internal standards used for stable isotope dilution
c
Collision energy
(of each compound): Mixture I (1 × 10−8 M), Mixture II

(2 × 10−8 M) and Mixture III (1 × 10−7 M).
5. Self-packed multi-StageTips are prepared manually from
Empore™ High Performance Disks (3 M Center, St. Paul, MN,
USA) as described by Rappsilber et al. [13]. Briefly, an Empore
disk is placed on a clean surface (petri dish) then very small
disks (approximately 1.0 mm diameter, 0.5 mm thick) are man-
ually cut out and introduced into a hollow blunt-ended syringe
needle, by gently pressing the needle into the disk. The needle
is then placed in a pipette tip (volume 200 μL) and the small
disk is released and pressed gently into position using a plunger
(rod) fitted into the needle. After removing the cutter and
plunger, a one-layer StageTip has been created. Another two
layers are added in the same manner to produce single-StageT-
ips. To isolate the cytokinin metabolites, multi-StageTips are
employed [9]. These include three layers of each of three com-
mercially available poly-tetrafluoroethylene matrices, contain-
ing: reversed-phase octadecyl-bonded silica (C18) groups,
poly(styrene-divinylbenzene) (SDB) copolymer modified with
sulfonic acid groups to increase hydrophilicity (SDB-RPS Disk),
and ion-exchange sorbent including sulfonic acid as a cation
exchanger (Cation-SR Disk) (Fig. 3).
6. While GFP+ cells are usually of most research interest, the
GFP− cells present in each sorted sample should be collected
and analyzed too, as they provide an internal reference popula-
tion that is crucial for data normalization and interpretation.
7. One to two cell-sorting experiments should be sufficient for
collecting 20,000–200,000 GFP+ cells (enough for cytokinin
analysis) from a transgenic line in which 20 % of the cells are
fluorescent. To obtain a complete data set from cell sorting
experiments 6–10 biological replicates are required, thus two
or three weeks should be allowed for sample collection with
approximately three sorting experiments (days) per week.
8. At this stage bubbles should be eliminated by being pushed
out, or replacing the mesh if necessary, as plants will be unable
to grow on them.
9. Healthy plants give rise to healthy root protoplasts with rela-
tively high resistance to mechanical stress caused by the
required handling for protoplast isolation. This not only
increases the efficiency of the method but also reduces biologi-
cal variation between samples.
10. Cells should be harvested at the same time every day to avoid
possible effects of circadian regulation on the results.
11. Assist the cutting by keeping the cotyledons in place by hold-
ing them gently with the full length of a forceps. This process
is also facilitated by the presence of the mesh and the high sow-
ing density (see Subheading 3.1, steps 2 and 3, respectively).
12. Since the seedlings have been grown at high density the roots
can be collected as a group rather than individually, thereby
reducing the extent of tissue wounding and accelerating the
process.
13. Gentle manual shaking every 30 min during incubation will
improve the efficiency of the protoplast isolation procedure.
14. The undigested tissue will stay on the cell strainer and the col-
lected PBI-protoplasts mixture should be transparent.
15. After centrifugation the protoplasts will accumulate at the bot-
tom of the tube, but they should be very easy to resuspend
(especially if they are in good condition). Therefore, the super-
natant should be slowly pipetted out using a 10 mL glass
pipette, and ca. 0.5 mL of supernatant should be left in the
tube to avoid discarding protoplasts that have already started
to resuspend. The cell wall-digestion enzymes have already
been diluted (1:2) by addition of the 25 mL PB for rinsing
(Subheading 3.2, step 5), and from that point on the samples
will be kept at 4 °C, thus enzymatic activity will be reduced.
16. For example, the flow rate parameter should be adjusted
according to the density of protoplasts isolated and injected in
the FACS, which can vary from day to day. The amplitude
parameter regulating the formation of single droplets during
FACS may also require adjustment, which may affect other
parameters, such as the drop delay.
17. The FACS instrument is cleaned prior to shut down. However
additional rinsing with 70 % ethanol upon start-up further
ensures cleanliness of the machine, which is crucial for its high
quality performance.
18. Drop breakoff is the point at which the first drop is detached
from the main stream. The formation of individual drops at a
stable and optimum physical space within the FACS and a sta-
ble flow is essential for high quality and purity cell-sorting, and
(hence) reproducible data. Frequency and amplitude are
adjusted to create the vibration wave that will provide the opti-
mum droplet breakoff. Amplitude controls the strength of the
steam vibration while the frequency regulates the number of
droplets generated per second, measured in kHz. Drop 1 rep-
resents a random point of the first separated droplet measured
in pixels, but it is required as a known time offset from the
start. Gap is the distance between drops (measured in pixels)
and is used for stream supervision.
19. Knowledge of the stable period between the times when a cell
or particle is being interrogated by the lasers and it reaches the
last attached droplet in the stream (breakoff position) is essen-
tial for the sorting and thus isolation of the desired cell popula-
tion. This is because when a desired interrogated cell arrives in
the last droplet connected to the stream, a charge is applied to

the stream which is retained by the desired droplet/cell after
leaving the stream. This charge facilitates the sorting of the
drop/cell in the correct collection tube through enabling its
attraction by an oppositely charged deflection plate oriented
almost parallel to the main stream.
20. Sorting a desirable cell into a specific collection tube is facili-
tated by applying a charge in the main stream, which is retained
by the detached drop containing the desirable cell for sorting.
The charged drop passes through an electrostatic field created
by two deflection plates that are oriented almost parallel to the
main stream and is then deflected to the left or right according
to the polarity of its charge. Thus, cells contained in oppositely
charged drops will be shunted into side streams moving in
opposite directions and sorted into respectively positioned col-
lection tubes while uncharged drops containing undesirable
cells/particles will not be affected by the electrostatic field and
pass down to the waste aspirator (main stream direction).
21. The light is perceived by detectors, converted to a voltage
pulse and the information can then be visualized in plots on
the computer screen (Fig. 1b–e).
22. In the suspension of protoplasts injected in the FACS “debris”
will also be present. This will include all material apart from the
desired intact protoplasts and may include undigested cell wall
parts or organelles derived from broken protoplasts (Fig. 2c).
23. There may be some overlap between the GFP+ and GFP− pop-
ulations. Therefore, it is essential to inspect and avoid possible
overlap between gates P2 and P3. This can be done by visual-
izing histograms of the populations’ GFP fluorescence as
shown in Fig. 1e.
24. Loading times of individual solutions can vary depending on
experimental conditions and type of centrifuge instrument,
and must be optimized.
25. At the beginning of the experiment, chill the centrifuge to
4 °C (approximately 10 min).
26. Thawing the isolated GFP+ and GFP− protoplasts collected by
FACS (samples snap-frozen as described in Subheading 3.3,
step 13), facilitates complete disruption of the protoplasts,
thus providing access to the endogenous CKs.
27. This is by far the most time-consuming step of the purification
protocol (3.5 h).
28. Reproducible chromatographic separation and retention time
stability is essential for separation of cis- and trans-zeatin iso-
mers, their ribosides, nucleotides as well as O- and N-glucosides
with identical precursor and basic product ions (Table 1).
29. In our system, MS conditions in positive ion mode are as fol-

lows: Drying Gas Temperature, 200 °C; Drying Gas Flow,
16 L/min; Nebulizer Pressure, 35 Psi; Sheath Gas Temperature,
375 °C, Sheath Gas Flow, 12 L/min; Capillary, 3400 V;
Nozzle Voltage, 0 V; Delta iFunnel High/Low Pressure RF,
150/60 V.
30. Calibration curves ranging from 100 amol to 1 pmol were cre-
ated from analyses of CKs metabolites by plotting known con-
centrations of each unlabeled analyte against analyte/IS ratios.
To quantify levels of endogenous CK metabolites in proto-
plasts, their ratios to the appropriate stable isotope-labeled
standards (Table 1) can be determined and further used in final
calculations, based on the known quantity of IS added (see
Note 4).
Acknowledgements
This work was supported by the Ministry of Education, Youth and

Sports of the Czech Republic (the National Program for
Sustainability I No. LO1204 and the “Návrat” program LK21306)
and by the Czech Science Foundation (No. GA14-34792S), the
European Molecular Biology Organization (EMBO ASTF 297-
2013), the Erasmus program (European Commission: Lifelong
Learning Programme), Development-The Company of Biologists
(DEVTF2012), the Plant Fellows programme (http://www.plant-
fellows.ch/), the Swedish Governmental Agency for Innovation
Systems (VINNOVA), and the Swedish Research Council (VR).
References
1. Petersson SV, Johansson AI, Kowalczyk M 5. Pěnčík A, Simonovik B, Petersson SV et al

et al (2009) An auxin gradient and maximum (2013) Regulation of auxin homeostasis and
in the Arabidopsis root apex shown by high- gradients in Arabidopsis roots through the for-
resolution cell-specific analysis of IAA distribu- mation of the IAA catabolite oxIAA. Plant Cell
tion and synthesis. Plant Cell 21:1659–1668 25:3858–3870
2. Antoniadi I, Plačková L, Simonovik B et al 6. Moussaieff A, Rogachev I, Brodsky L et al
(2015) Cell-type specific cytokinin distribu- (2013) High-resolution metabolic mapping of
tion within the Arabidopsis primary root apex. cell types in plant roots. Proc Natl Acad Sci U
Plant Cell 27:1955–1967 S A 110:E1232–E1241
3. Birnbaum K, Shasha DE, Wang JY et al (2003) 7. Petersson SV, Lindén P, Moritz T et al (2015)
A gene expression map of the Arabidopsis root. Cell-type specific metabolic profiling of
Science 302:1956–1960 Arabidopsis thaliana protoplasts as a tool for plant
4. Yadav RK, Girke T, Pasala S et al (2009) Gene systems biology. Metabolomics 11:1679–1689
expression map of the Arabidopsis shoot apical 8. Tarkowská D, Novák O, Floková K et al
meristem stem cell niche. Proc Natl Acad Sci (2014) Quo vadis plant hormone analysis?
U S A 106:4941–4946 Planta 240:55–76
9. Svačinová J, Novák O, Plačková L et al (2012) chromatography-tandem mass spectrometry.

A new approach for cytokinin isolation from J Chromatogr B 891–892:27–25
Arabidopsis tissues using miniaturized purifi- 12. Du FY, Ruan GH, Liang SH et al (2012)
cation: pipette tip solid-phase extraction. Plant Monolithic molecularly imprinted solid-
Methods 8:17 phase extraction for the selective determina-
10. Novák O, Hauserová E, Amakorová P et al tion of trace cytokinins in plant samples with
(2008) Cytokinin profiling in plant tissues liquid chromatography-electrospray tandem
using ultra-performance liquid chromatogra- mass spectrometry. Anal Bioanal Chem
phy–electrospray tandem mass spectrometry. 404:489–501
Phytochemistry 69:2214–2224 13. Rappsilber J, Mann M, Ishihama Y (2007)
11. Liu Z, Cai BD, Feng YQ (2012) Rapid deter- Protocol for micro-purification, enrichment,
mination of endogenous cytokinins in plant pre-fractionation and storage of peptides for
samples by combination of magnetic solid proteomics using StageTips. Nat Protoc
phase extraction with hydrophilic interaction 2:1896–1906
Chapter 20
Hormone Profiling in Plant Tissues

Maren Müller and Sergi Munné-Bosch
Abstract
Plant hormones are for a long time known to act as chemical messengers in the regulation of physiological
processes during a plant’s life cycle, from germination to senescence. Furthermore, plant hormones simul-
taneously coordinate physiological responses to biotic and abiotic stresses. To study the hormonal regula-
tion of physiological processes, three main approaches have been used (1) exogenous application of
hormones, (2) correlative studies through measurements of endogenous hormone levels, and (3) use of
transgenic and/or mutant plants altered in hormone metabolism or signaling. A plant hormone profiling
method is useful to unravel cross talk between hormones and help unravel the hormonal regulation of
physiological processes in studies using any of the aforementioned approaches. However, hormone profil-
ing is still particularly challenging due to their very low abundance in plant tissues. In this chapter, a sensi-
tive, rapid, and accurate method to quantify all the five “classic” classes of plant hormones plus other plant
growth regulators, such as jasmonates, salicylic acid, melatonin, and brassinosteroids is described. The
method includes a fast and simple extraction procedure without time consuming steps as purification or
derivatization, followed by optimized ultrahigh-performance liquid chromatography coupled to electro-
spray ionization-tandem mass spectrometry (UHPLC-MS/MS) analysis. This protocol facilitates the high-
throughput analysis of hormone profiling and is applicable to different plant tissues.
Key words Plant hormones, UPLC/ESI-MS/MS, Auxin, Abscisic acid, Brassinosteroids, Cytokinins,
Ethylene, Gibberellins, Jasmonates, Melatonin, Salicylic acid
1 Introduction
Plant hormones have been classified according to their chemical

structures into five “classic” classes: cytokinins (CKs), auxins [in
particular indole-3-acetic acid (IAA)], abscisic acid (ABA), gibber-
ellins (GAs), ethylene, and several other plant growth regulators,
such as jasmonates [jasmonic acid (JA) and its derivatives 12-oxo-
phyodienoic acid (OPDA) and jasmonic acid-isoleucine (JA-Ile)],
salicylic acid (SA), brassinosteroids (BRs), and melatonin (MEL)
[1, 2]. They are crucial in regulating plant development and plant
responses to various abiotic and biotic challenges [3]. Moreover,
the significance of hormonal physiology in plants is increasing as we
come to understand the interactive cascades of signal transduction
249
250 Maren Müller and Sergi Munné-Bosch
network pathways, mechanisms of tissue sensitivity, genetic bases

for hormonal action, and synergistic and antagonistic interaction
among hormones [4]. Hence, the interest has changed from a spe-
cific plant hormone or a certain group of plant hormones to a mul-
tiple plant hormone profile analysis [5]. Frequently these molecules
act at low levels, but levels can strongly differ depending on plant
tissues. For instance, pollen and seeds contain BRs amounts at the
nanogram/g FW level, while amounts in roots and shoots are at the
picogram/g FW level [6, 7].
In the last 15–20 years many methods have been developed to
quantify plant hormones using techniques such as gas chromatog-
raphy (GC) and liquid chromatography (LC) coupled to mass
spectrometry (MS) [8–11]. Mass spectrometers (triple quadru-
pole, ion trap or time-of-flight) provide a powerful analytical tool
for quantification of plant hormones due to their high sensitivity,
specificity, and reproducibility. Moreover, using the multiple reac-
tion monitoring (MRM) mode increases the sensitivity of the mass
spectrometer and decreases the background noise. However,
depending on the plant tissues, other compounds present in the
matrix can affect analyte ionization during plant hormone analyses,
which must be controlled during an analytical method develop-
ment for quantification of plant hormones.
In this chapter, a method for the sensitive and quantitative
analysis of 20 plant hormones the cytokinins trans-zeatin (tZ),
trans-zeatin riboside (tZR), 2-isopentenyladenine (2iP), and iso-
pentenyladenosine (IPA), the auxin IAA, the ethylene precursor,
1-amino-cyclopropane-1-carboxylic acid (ACC), ABA, SA, the jas-
monates JA, OPDA, and JA-Ile, the gibberellins GA1, GA4, GA9,
GA19, GA20, GA24, MEL, and the brassinosteroids, brassinolide
(BL) and castasterone (CS) is described. This protocol facilitates
the high-throughput analysis of hormones without time-consuming
multiple steps of sample preparation followed by a rapid ultrahigh-
performance liquid chromatography–electrospray ionization tan-
dem mass spectrometry (UPLC/ESI-MS/MS), and has been
successfully used for different plant tissues including leaves, roots,
seeds, flowers, and pollen.
2 Materials
2.1 Chemicals Solvents used for extraction are methanol (MeOH), isopropanol
(iPrOH) and glacial acetic acid. Solvents used for UPLC-MS/MS
analysis are acetonitrile, milli-Q water (18.2 MΩ), and glacial acetic
acid. ABA, ACC, IAA, IPA, JA, tZ, tZR, and SA standards can be
obtained from any supplier of fine chemicals. BL, CS, GA, 2-iP,
JA-Ile, OPDA, and MEL standards, together with the internal stan-
dards [2H6]ABA, [2H4]ACC, [2H3]BL, [2H3]CS, [2H2]GA1, [2H2]
GA4, [2H2]GA9, [2H2]GA19, [2H2]GA20, [2H2]GA24, [2H5]IAA,
Hormone Profiling in Plant Tissues 251
[2H6]2-iP, [2H6]IPA, [2H5]OPDA, [2H5]JA, [2H2]JA-Ile, [2H4]SA,

[2H5] tZ, and [2H5]tZR can be purchased from OlChemim Ltd.
(Olomouc, Czech Republic). [2H4] MEL can be purchased from
Toronto Research Chemicals (Toronto, Ontario, Canada).
2.2 Sample 1. Extraction solvents: MeOH–glacial acetic acid, 99:1 (v/v);

Extraction MeOH–iPrOH–glacial acetic acid, 50:49:1 (v/v) (see Note 1).
2. Internal standard solution: stock solution of [2H6]ABA, [2H4]
ACC, [2H3]BL, [2H3]CS, [2H5]IAA, [2H6]2-iP, [2H6]IPA,
[2H5]JA, [2H4]MEL, [2H4]SA, [2H5] tZ, [2H5]tZR, [2H2]GA1,
[2H2]GA4, [2H2]GA9, [2H2]GA19, [2H2]GA20, [2H2]GA24,
[2H5]OPDA, and [2H2]JA-Ile in MeOH (see Note 2).
3. Liquid nitrogen.
4. Microcentrifuge tubes (1.5 and 2 mL).
5. Mixer mill (MM400, Retsch GmbH, Haan, Germany).
6. Pipettes and pipette tips.
7. Vortex mixer.
8. Ultrasonication bath (Branson 2510, Danbury, CT USA).
9. Centrifuge type, which allows to use 1.5 to 2 mL tubes at 4 °C
and 14,000 × g.
10. 0.22 μm PTFE syringe filters.
11. HPLC glass vials with fused-in insert (300 μL) and caps.
2.3 LC-ESI-MS/MS 1. Quaternary pump Aquity UPLC™ System (Waters, Milford,

Analysis MA USA).
2. Colum C18 Kinetex 50 × 2.1 mm, 1.7 μm (Phenomenx,
Macclesfield, UK).
3. API 3000 triple quadrupole mass spectrometer (PE Sciex,
Concord, Ont., Canada) operating in MRM mode (see Note 3).
4. Water containing 0.05 % glacial acetic acid.
5. Acetonitrile containing 0.05 % glacial acetic acid.
6. Syringe pump (Model 11; Harvard Apparatus, Holliston, MA
USA):
7. Each natural hormone (ABA, ACC, BL, CS, IAA, 2-iP, IPA,
GA1, GA4, GA9, GA19, GA20, GA24, JA, JA-Ile, MEL, OPDA,
SA, tZ, tZR) and isotopically labeled hormones ([2H6]ABA,
[2H4]ACC, [2H3]BL, [2H3]CS, [2H5]IAA, [2H6]2-iP, [2H6]
IPA, [2H2]GA1, [2H2]GA4, [2H2]GA9, [2H2]GA19, [2H2]GA20,
[2H2]GA24, [2H5]JA, [2H2]JA-Ile, [2H4]MEL, [2H5]OPDA,
[2H4]SA, [2H5] tZ, [2H5]tZR) at a concentration of 1 μg mL−1
in MeOH should be infused of a constant flow rate of
15 μL min−1.
2.4 Quantification 1. Calibration solutions: Ten point standard solutions should be

prepared in MeOH, containing each of the authentic standards
(ABA, ACC, BL, CS, GA1, GA4, GA9, GA19, GA20, GA24, IAA,
2-iP, IPA, JA, JA-Ile, MEL, OPDA, SA, tZ, tZR) ranging the
expected physiological concentrations and within the linear
response of the mass spectrometer. For example, ten point
standard solutions from 0.05 to 1000 ng mL−1, however, each
solution should contain a constant amount of internal stan-
dards (see Note 2).
2.5 Validation 1. As the matrix components can affect the analyte stability,
extraction and ionization samples should be spiked with low
levels of standards.
3 Methods
The idea was to provide a protocol that facilitates the high-throughput

analysis of hormone profiling which is applicable to different plant
tissues. The protocol includes a fast and simple extraction procedure
to quantify plant hormones and their stable isotope-labeled standards
without time consuming steps as purification or derivatization fol-
lowed by UHPLC/ESI-MS/MS analysis. In our research group, this
method has been used to quantify hormone profiling in leaves, flow-
ers, fruits, roots, and seeds of different species of a number of plant
families including the model plant Arabidopsis thaliana, Brassicaceae,
Liliaceae, Lamiaceae, Velloziaceae, Arecaceae, or Rosaceae. MS chro-
matograms for plant hormones extracted from different plant tissues
are shown in Fig. 1.
3.1 Hormone 1. Freeze immediately the obtained sample material in liquid

Extraction nitrogen and store it at −80 °C until analysis.
2. Transfer a pool of each sample tissue to a 2 mL microcentri-
fuge tube, add a magnetic bead and grind it using a mixer mill
(see Note 4). Take care that the plant material does not thaw.
3. Weigh 100 mg of frozen powder and transfer it to a 1.5 mL
microcentrifuge tube and take care that the sample tissue does
not thaw (see Note 5).
4. Add 200 μL of extraction solvent containing internal stan-
dards solution to each sample. Mix each solution using vortex
and extract them between 4 and 7 °C using ultrasonication for
30 min.
5. Centrifuge at 14,000 × g for 10 min at 4 °C. Transfer the super-
natant to a 1.5 mL microcentrifuge tube and store it at 4 °C
(see Note 6).
a Endogenous hormones
1.09 100 100 1.48 1.62
100 100
1.25
IAA
SA
MEL 174/130 ABA
137/93
233/174 263/153
% %
% %
Time (min) Time (min)

Time (min) Time (min)
Internal standards
1.09 1.25 1.46 1.61
100 100 100 100
[2H4]MEL [2H4]SA [2H5]IAA

237/178 [2H6]ABA
141/97 179/135 % 269/159
% % %
Time (min) Time (min) Time (min) Time (min)
b Endogenous hormones
100 1.60
100
100 100 3.66
tZR 2.03 GA GA9

4 CS
352/220 331/213 315/271
% % 465/447
% %
2.39
Time (min) Time (min) Time (min)

Time (min)
Internal standards 3.66

2.02 2.38 100
100 1.59 100 100
[2H3]CS
[2H2]GA9 468/450
% [2H2]GA4 % 357/225 %
% [2H5]tZR
357/225 357/225

Time (min)
c Endogenous hormones
1.84 2.02 100 2.58
1.09 100
100 100
JA OPDA
MEL JA-Ile
% 209/59 %
291/170
233/174 % 322/130
%
Time (min)
Internal standards
1.09 1.83 2.02 100 2.58
100 100 100
[2H4]MEL [2H5]JA [2H2]JA-Ile [2H5]OPDA

214/64 296/1710
% 237/178 % %
324/130 %
Time (min) Time (min) Time (min) Time (min)
Fig. 1 Several examples of MS chromatograms for endogenous hormones and internal standards obtained by
UPLC/ESI-MS/MS analysis. (a) MS chromatograms for MEL, SA, IAA, and ABA extracted from 100 mg
Arabidopsis thaliana leaves. (b) MS chromatograms for tZR, GA4, GA9, and CS extracted from 100 mg
Chamaerops humilis var. cerifera seeds. (c) MS chromatograms for MEL, JA, JA-Ile, and OPDA extracted from
100 mg Iris × hollandica tepals
Table 1
LC conditions
Time (min) Solvent A (%) Solvent B (%)

0 99 1
0–4 1 99
4–4.20 99 1
4.20–5 99 1
A water containing 0.05 % glacial acetic acid

B acetonitrile containing 0.05 % glacial acetic acid
6. Re-extract the pellet three times with 100 μL of extraction sol-

vent as described above and combine the supernatants.
7. Mix each combined supernatant using vortex and centrifuge
them at 14,000 × g for 5 min at 4 °C.
8. Filter the supernatants through a 0.22 μm PTFE filter into a
HPLC vial with a fused-in insert for analysis.
3.2 UPLC-MS/MS 1. UPLC conditions: using a C18 Kinetex column (50 × 2.1 mm,
Analysis 1.7 μm) (Phenomenx, Macclesfield, UK); typical UPLC gradi-
ent conditions are listed in Table 1. The flow rate was
0.5 mL min−1, the injection volume was 5 μL and the column
temperature was maintained at 35 °C.
2. The mass spectrometer should operate in multiple reaction
mode (MRM) due to the high selectivity using precursor-to-
product ion transitions.
3. For each authentic standard and internal standard declustering
potential (DP), focusing potential (FP), entrance potential
(EP), collision energy (CE), collision cell exit potential (CXP)
should be optimized to determine their appropriate MRM
transition. Infuse each solution (1 ppm) into the tandem mass
spectrometer using a syringe pump at for instance 15 μl min−1.
In contrast, temperature, nebulizer gas, curtain gas, collision
gas, and the capillary voltage cannot be modified during analy-
sis and average values for the different hormones should be
selected (see Note 7). Optimized parameters for the API 3000
triple quadrupole mass spectrometer (PE Sciex, Concord,
Ont., Canada) are given in Table 2 (see also Note 8).
4. Quantification should be performed by running all calibration
solutions and construct calibration curves using spectrometer
software (e.g., Analyst™ software; PE Sciex, Concord, Ont.,
Canada) for each authentic standard by integrating peak areas
for each analyte at a given concentration and the correspond-
ing internal standard.
Table 2
Optimized UPLC/ESI-MS/MS parameters for quantifying plant hormones
Product Product
MS RTa Precursor iona RTb Precursor ionb Scan CXPa,b
Analytes method (min) iona (m/z) (m/z) ISb (min) ionb (m/z) (m/z) Mode CEa,b (V) (V)
ACC 1 0.26 102 56 [2H4]tZ 0.26 106 60 + 15 15
MEL 1 1.09 233 174 [2H4] 1.09 237 178 + 20 10

MEL
tZ 1 1.25 220 136 [2H5]tZ 1.24 225 137 + 25 15

2
tZR 1 1.60 352 220 [ H5]tZR 1.59 357 225 + 25 15
2
2-iP 1 2.15 204 136 [ H6]2-iP 2.13 210 137 + 35 10
2
IPA 1 2.46 336 204 [ H6]IPA 2.45 342 210 + 25 15
2
BL 1 3.58 481 315 [ H3]BL 3.57 484 315 + 30 35
2
CS 1 3.66 465 447 [ H3]CS 3.66 468 450 + 20 15
2
SA 2 1.25 137 93 [ H4]SA 1.25 141 97 − −20 −15
2
IAA 2 1.48 174 130 [ H5]IAA 1.46 179 135 − −50 −10
2
ABA 2 1.62 263 153 [ H6] 1.61 269 159 − −30 −15
ABA
JA 2 1.84 209 59 [2H5]JA 1.83 214 64 − −40 −15
JA-Ile 2 2.02 322 130 [2H2] 2.02 324 130 − −34 −15
JA-Ile
OPDA 2 2.58 291 165 [2H5] 2.58 296 170 − −35 −15
OPDA
GA1 3 1.24 347 273 [2H2]GA1 1.24 349 275 − −40 −15
2
GA19 3 1.57 361 273 [ H2] 1.57 363 275 − −15/−25 −15
GA19
GA20 3 1.63 331 287 [2H2] 1.62 333 289 − −25 −15
GA20
GA4 3 2.03 331 213 [2H2]GA4 2.02 333 215 − −40 −15
2
GA24 3 2.11 345 257 [ H2] 2.11 347 259 − −50 −15
GA24
GA9 3 2.39 315 271 [2H2]GA9 2.38 317 273 − −50/−40 −15
MS mass spectrometry, RT retention time, IS internal standards, CE collision energy, CXP collision cell exit potential,
V voltage
a
Analytes
b
Internal standards
ACC 1-amino-cyclopropane-1-carboxylic acid, MEL melatonin, tZ trans-zeatin, tZR trans-zeatin riboside, 2iP isopen-
tenyladenine, IPA isopentenyladenosine, BL brassinolide, CS castasterone, SA salicylic acid, IAA índole-3-acetic acid,
ABA abscisic acid, JA jasmonic acid, JA-Ile jasmonic acid-isoleucine, OPDA 12-oxo-phyodienoic acid, GA gibberellin
3.3 Validation The method evaluation should be conducted to demonstrate that

the analytical method is applicable to obtain values, which are close
to the unknown level of the analyte present in plant tissue. At least
selectivity, linearity, limits of detection (LOD) and quantification
(LOQ), recovery, repeatability and reproducibility should be eval-
uated during the validation processes.
1. Selectivity is defined as the ability to separate the analyte from
the other sample components, giving pure, symmetric and
resolved peaks. The method is proposed to be selective in the
case that no additional peaks in the MS/MS spectrum can be
detected for the band that corresponds to the analyte in the
matrix compared to the original standards.
2. Linearity: calibration curves should be prepared in matrix and
indicate good linearity with correlation coefficients (R) > 0.99
(see Note 9).
3. LOD and LOQ: LOD is defined as the lowest concentration of
hormone, which can be distinguished from the noise in blank
and is defined as a signal-to-noise ratio of 3; LOQ is defined as
the lowest concentration of hormone that can be quantified
precisely and accurately and is defined as a signal-to-noise ratio
of 10 (see Note 10).
4. Recovery: spike six plant extracts with internal standard solu-
tion before extraction and six plant extracts with internal stan-
dard solution after extraction and calculate the recovery (%).
5. Repeatability and reproducibility: retention time and peak area
reproducibility and intra- and inter-day precision in matrix
should be determined and evaluated by calculating the relative
standard deviation (% RSD, n = 6). RSD values below 15 % are
acceptable (see Note 11).
6. To evaluate matrix effects analyze calibration curves performed
in the matrix and in solvent and calculate the ratio (see Note 12).
4 Notes
1. Due to the diverse chemical structures, physiological proper-

ties of plant hormones and matrix effects no extraction solvent
has been published to isolate all plant hormones equally well
from all tissue types. MeOH–glacial acetic acid, 99:1 (v/v)
showed good results for plant tissues with high lipid contents
such as seeds, MeOH–iPrOH–glacial acetic acid, 50:49:1
(v/v) showed good results for plant tissues such as leaves and
flowers. Nevertheless, high water contents of plant tissues can
influence extraction efficiency and should also be taken into
account.
2. A stock solution should be prepared with a high concentration

that can be stored at −20 °C and from this stock solution an
internal standard solution can be prepared freshly before use.
The optimal final concentration of the internal standard solu-
tion needs to be assessed empirically but 100 ng mL−1 [2H6]
ABA, [2H4]ACC, [2H3]BL, [2H3]CS, [2H5]IAA, [2H6]2-iP,
[2H6]IPA, [2H5]JA, [2H4]MEL, [2H4]SA, [2H5] tZ, [2H5]tZR,
and 20 ng/ mL [2H2]GA1, [2H2]GA4, [2H2]GA9, [2H2]GA19,
[2H2]GA20, [2H2]GA24, [2H5]OPDA, [2H2]JA-Ile per sample is
a good starting point.
3. The mass spectrometer should operate in MRM mode due to
the high selectivity using precursor-to-product ion transitions
as many compounds can present the same nominal molecular
masses or peaks can overlap.
4. To avoid that plant tissue thaws it is recommended to dip
microcentrifuge tubes and the tube adaptor for the mixer mill
in liquid nitrogen. If the grinding process takes longer than
30 s, the samples should be dipped again in liquid nitrogen.
IMPORTANT: avoid that liquid nitrogen enters in the micro-
centrifuge tubes, because it could destroy them.
5. To avoid the plant tissue from thawing it is recommended to
dip microcentrifuge tubes and spatula in liquid nitrogen before
and after weighing.
6. It is important to transfer the supernatant without any solid
particles.
7. Optimized parameters obtained on the API 3000 triple quad-
rupole spectrometer (PE Sciex, Concord, Ont., Canada): tem-
perature 400 °C, nebulizer gas (N2) 10 (arbitrary units),
curtain gas (N2) 12 (arbitrary units), collision gas (N2) 4 (arbi-
trary units), the capillary voltage 3.5 kV in the positive ion
mode and −3.5 kV in the negative ion mode.
8. Plant hormones were analyzed using three different MS meth-
ods. CKs, ACC, and MEL were analyzed using MS method 1,
ABA, SA, IAA, and jasmonates using MS method 2, and GAs
using MS method 3 and given in Table 2.
9. 100 mg ground samples should be spiked with authentic stan-
dard solutions ranging for example from 0.1–500 ng mL−1 for
CKs, IAA, ACC, GAs, MEL, and BRs; 1–500 ng mL−1 for
ABA, SA, and jasmonates and spiked with constant levels of
internal standards and extracted as described in Subheading 3.1.
10. LOD should be established using matrix samples spiked with
low amounts of standards solution. In the case that none
analyte-free matrix is available LODs should be determined in
solvent.
11. Samples should be performed using three spiking solutions

(low, medium, and high) and extracted as described in
Subheading 3.1.
12. Matrix interferences can affect the efficiency of the analyte ion-
ization resulting in suppressed or enhanced analyte ionization
(decreasing or increasing the response); however, they are not
well understood and seem to vary depending on analyte and
matrix [12].
Acknowledgments
We thank Barbara Simancas and Javier A. Miret for help in extrac-

tion of Arabidopsis thaliana leaves and Iris × hollandica tepals.
References
1. Pan X, Welti R, Wang X (2008) Simultaneous 8. Müller A, Düchting P, Weiler EW (2002) A
quantification of major phytohormones and multiplex GC-MS/MS technique for the sen-
related compounds in crude plant extracts by sitive and quantitative single-run analysis of
liquid chromatography tandem mass spec- acidic phytohormones and related compounds,
trometry. Phytochemistry 69:1773–1781 and its application to Arabidopsis thaliana.
2. Zhang N, Sun Q, Zhang H, Cao Y, Weeda S, Planta 216:44–56
Ren S, Guo Y-D (2015) Roles of melatonin in 9. Chiwocha SDS, Abrams SR, Ambrose SJ et al
abiotic stress resistance in plants. J Exp Bot (2003) A method for profiling different classes
66:647–656 of plant hormones and their metabolites using
3. Davies PJ (2010) The plant hormones: their liquid chromatography-electrospray ionization
nature, occurrence, and functions. Kluwer tandem mass spectrometry: an analysis of
Academic, The Netherlands hormone regulation of thermodormancy of
4. Farnsworth E (2004) Hormones and shifting lettuce (Lactuca sativa L.) seeds. Plant
ecology throughout plant development. J 35:405–417
Ecology 85:5–15 10. Pan X, Welti R, Wang X (2010) Quantitative
5. Van Meulenbrock L, Vanden BJ, Steppe K et al analysis of major plant hormones in crude
(2012) Ultra-high performance liquid chroma- plant extracts by high-performance liquid
tography coupled to high resolution Orbitrab chromatography mass spectrometry. Nat
mass spectrometry for metabolomics profiling Protoc 5:986–992
of the endogenous phytohormonal status of the 11. Müller M, Munné-Bosch S (2011) Rapid and
tomato plant. J Chromatogr A 1260:67–80 sensitive hormonal profiling of complex plant
6. Baiguz A, Tretyn A (2003) The chemical char- samples by liquid chromatography coupled to
acteristic and distribution of brassinosteroids electrospray ionization tandem mass spec-
in plants. Phytochemistry 62:1027–1046 trometry. Plant Methods 7:37
7. Takatsu S (1994) Brassinosteroids: distribu- 12. Almeida Trapp M, De Souza GD, Rodrigues-
tion in plants, bioassays and microanalysis by Filho E et al (2014) Validated method for phy-
gas chromatography-mass spectrometry. tohomone quantification in plants. Front Plant
J Chromatogr A 658:3–15 Sci 5:417
Chapter 21
Use of Xenopus laevis Oocytes to Study Auxin Transport

Astrid Fastner, Birgit Absmanner, and Ulrich Z. Hammes
Abstract
Xenopus laevis oocytes are an expression system that is particularly well suited for the characterization of
membrane transporters. Oocytes possess only very little endogenous transport systems and therefore trans-
porters can be studied with a high signal-to-noise ratio. This book chapter provides the basic methods to
use Xenopus oocytes for the characterization of transporters by radiotracer experiments. While the methods
described here were established to study auxin transport they can easily be adapted to study other hormone
transporters and their substrates.
Key words Xenopus laevis oocytes, Microinjection, Transport, Import export, Radiotracer, Liquid
scintillation
1 Introduction
Transport proteins facilitate the uptake and also the release of

substrates across membranes. The characterization of their bio-
chemical properties is challenging because large parts of the protein
are hydrophobic and reside in the two-dimensional membranes—as
a consequence they cannot be expressed to high levels. Nevertheless,
knowledge about the kinetics of a transport process, the interfer-
ence of inhibitors, or their substrate specificity are key to the under-
standing of many biological processes. The vast majority of plant
transporters have been studied in heterologous expression systems
particularly yeast and Xenopus laevis oocytes.
The use of oocytes as an expression system has been pioneered
more than 40 years ago [1]. It was not until the early 1990s that
plant transporters were studied in Xenopus oocytes [2]. But since
then many plant transporters were characterized in oocytes. Stage
V and VI oocytes are generally considered to be particularly good
expression systems for three reasons: (1) During this stage, the
cells are rich in yolk and as a consequence are among the cells that
have the lowest number of endogenous transport systems present
in their membrane. As many as ~1010–1011 transporters per oocyte
259
260 Astrid Fastner et al.
are found in the oolemma (oocyte plasma membrane) upon

injection of an mRNA. Therefore, transporters can be studied in
many cases background free—this is especially true for charged
substances [3]. (2) For obvious reasons, Xenopus oocytes do not
naturally contain plant-specific proteins. This minimizes possible
interference of transporter-modifying activities that must be postu-
lated in plant expression systems such as plant protoplasts or BY2
cells and—to a lesser extent—also yeast and mammalian cells. (3)
Oocytes are fairly big, with a diameter of about 1 mm. This allows
the injection of defined amounts of substrate into the cell to pro-
duce an outward-directed electrochemical gradient in a substrate-
free solution with an initial internal substrate concentration that is
very similar in each cell. This is a clear advantage over the passive
equilibration of substrate prior to the experiment, which inevitably
leads to larger variations.
Due to these properties and advantages oocytes are mainly
used to perform electrophysiological experiments using two-
electrode voltage clamp (TEVC). Nevertheless, oocytes have also
widely been used to study transport of radioactively labeled sub-
strates. By this approach importers and exporters for plant hor-
mones like gibberellic acid, jasmonate, abscisic acid, as well as
several auxin transporters were characterized in oocytes [4–10]. So
far hormone transporters have not been characterized electrophys-
iologically. This is likely due to the comparatively low transport
rates for these substrates, which are in many cases below the reso-
lution limit of TEVC.
Applying the methods described below, we were able to suc-
cessfully characterize auxin import and export by expression of
transporters with or without modifying kinases in oocytes. In prin-
ciple, the methods can be modified and applied to other transport-
ers and other substrates in the oocyte expression system.
2 Materials
All solutions are prepared with ultrapure nuclease-free water

(Biopak®, Merck-Millipore).
2.1 In Vitro 1. RNase Zap®: RNaseZap® RNase Decontamination Solution

Transcription of cRNA (Thermo Fisher Scientific, Waltham, Massachusetts, USA).
2. Phenol/chloroform/isoamyl alcohol: Roti®-phenol/chloro-
form/isoamyl alcohol, TE buffered, ratio 25:24:1,
pH 7.5–8.0.
3. Natriumacetat: 3 M, pH 5.2.
4. Chloroform: 99.8 %, per analysis, stabilized with amylene.
5. Ethanol: > 99.8 %, absolute.
Use of Xenopus laevis Oocytes to Study Auxin Transport 261
6. mMESSAGE mMACHINE® SP6 Transcription Kit (Thermo

Fisher Scientific, Waltham, Massachusetts, USA).
7. MEGAclear™ Transcription Clean-Up Kit (Thermo Fisher
Scientific, Waltham, Massachusetts, USA).
8. NanoDrop™ 2000c Spectrophotometer or an other photo-
metric device.
2.2 Oocyte Injection 1. Oocytes stages V and VI are manually selected (see Note 1).
with cRNA 2. Barth’s pH 7.4: 88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO3,
10 mM 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic
acid (HEPES), 0.33 mM Ca(NO3)2, 0.41 mM CaCl2, 0.82 mM
MgSO4, pH 7.4 (with NaOH). Filter (0.45 μm) sterilize and
store at 4 °C. A 5× stock solution can be prepared.
3. Gentamycin sulfate: 1000× Stock solution 50 mg/ml, store at
−20 °C.
4. Nanoject II 3.5″ replacement glass capillaries (Drummond
Scientific Company, Broomall, Pennsylvania, USA).
5. P‐97 Flaming/Brown micropipette puller (Sutter Instrument,
Novato, California, USA).
6. Stereomicroscope.
7. Plastic Pasteur pipettes with wide opening (drop volume 62 μl).
8. Disposable glass Pasteur pipettes 230 mm.
9. Pipettor pi-pump® 2500 (Glasfirn Giessen, Giessen, Germany).
10. Injection groove: Made from acrylic glass, dimensions (in cm)
5 × 5 × 0.3. Groove angle 45° (Fig. 1).
11. Nanoject II™ Auto-Nanoliter Injector (Drummond Scientific
Company, Broomall, Pennsylvania, USA).
12. BSA: Bovine serum albumin Fraction V, protease free.
Fig. 1 Schematic representation of an injection groove. (a) Top view. (b) Lateral
view with injection needle
2.3 Flux Experiments 1. Radioactivity: Indole-3-acetic acid [5-3H], solvent ethanol,

with Radioactive concentration 1 mCi/ml, specific activity 25 Ci/mmol,
Labeled Substances 250 μCi/vial, LOT 150522 (American Radiolabeled
Chemicals, St. Louis, Missouri, USA).
2. 3-Indole acetic acid (IAA): M.W. 175.18 g/mol.
3. Barth’s pH 7.4: 88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO3,
acid (HEPES), 0.33 mM Ca(NO3)2, 0.41 mM CaCl2, 0.82 mM
MgSO4, pH 7.4 (with NaOH). Filter (0.45 μm) sterilize and
store at 4 °C. A 5× stock solution can be prepared. Or Ringer:
115 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl2, 1 mM NaHCO3,
acid (HEPES), 1 mM MgCl2, adjust to desired pH (with HCl
or NaOH). A 10× stock solution can be prepared. Filter
(0.45 μm) sterilize and store at 4 °C.
4. 12-Well cell culture plate.
5. Scintillation vial 20 ml.
6. SDS: Sodium lauryl sulfate, 10 % solution.
7. Scintillation cocktail.
8. Liquid scintillation counter.
2.4 Immunoblotting 1. Homogenization buffer: 50 mM Tris(hydroxymethyl)-

and Biotinylation aminomethan (Tris) pH 7.5, 100 mM NaCl, 1 mM EDTA,
Components 1 mM Pefabloc® SC-Protease-Inhibitor, 1 tablet/ml PhosSTOP—
Phosphatase Inhibitor Cocktail Tablet (Roche, Basel, Switzerland).
2. Ultracentrifuge 100.000 g and appropriate vials.
3. PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM
KH2PO4, pH 8.0. A 10× stock solution can be prepared.
Autoclave and store at room temperature.
4. Biotin: EZ-Link® Sulfo-NHS-LC-Biotin (Thermo Fisher
Scientific, Waltham, Massachusetts, USA).
5. Biotin solution: 1 g EZ-Link® Sulfo-NHS-LC-Biotin/ml PBS
pH 8.0 (= ~ 2 mM biotin).
6. Quench buffer: 100 mM Glycine in PBS pH 8.0.
7. Lysis buffer: 150 mM NaCl, 20 mM Tris(hydroxymethyl)-
aminomethan (Tris) pH 7.6, 1 % Triton X-100, 0.5 mM
phenylmethylsulfonyl fluoride (PMSF).
8. Streptavidin–agarose from Streptomyces avidinii, buffered
aqueous solution (Sigma Aldrich, St. Louis, Missouri, USA).
9. 4× SDS-PAGE sample buffer: 250 mM Tris(hydroxymethyl)-
aminomethan (Tris) pH 6.8, 8 % SDS, 40 % glycerol, 20 %
β-mercaptoethanol, 0.004 % bromophenol blue.
10. Specific antibody.
3 Methods
Unless stated otherwise all steps are carried out at room tempera-
ture. For RNA work, gloves and RNase-free and freshly autoclaved
tips and cups are used. Pipettes and workspace should be thor-
oughly cleaned with RNase ZAP®.
3.1 In Vitro 1. Clone the CDS of the gene of interest (the transporter or the
Transcription of cRNA kinase) into the multiple cloning site of the vector backbone
pOO2 [11] (see Note 2).
2. Linearize 20 μg of the vector containing the gene of interest
(see Note 3).
3. Bring the linearization reaction to a volume of 200 μl using
RNase-free water.
4. Extract with one volume phenol/chloroform/isoamyl alcohol
(see Note 4).
5. Transfer the aqueous phase to a new RNase-free reaction cup.
6. Extract twice with one volume of chloroform.
7. Each time, transfer the aqueous phase to a new RNase-free
reaction cup.
8. Add 1/10 volume of sodium acetate and 2.5 volumes of ethanol
to the aqueous phase to precipitate the DNA on ice for 20 min.
9. Centrifuge the mixture at 4 °C for 20 min at 16,000 × g to
recover the precipitate.
10. Remove the liquid and air-dry the pellet.
11. Resuspend the pellet in RNase-free water to a concentration of
0.5 μg/μl.
12. For in vitro transcription the mMESSAGE mMACHINE® SP6
Transcription Kit is used according to the manufacturer’s
instructions (see Note 5).
13. cRNA is purified using the MEGAclear™ Transcription
Clean-Up Kit according to the manufacturer’s instructions (see
Note 6).
14. Determine the yield and purity of the cRNA synthesis using
the NanoDrop.
15. Determine the quality of the cRNA obtained by an analysis on
a 0.8 % agarose-gel. Mix 1 μl of cRNA with RNase-free water
and DNA-loading dye (see Note 7).
3.2 Oocyte Injection 1. Use oocytes stages V and VI for cRNA injection. Inject oocytes
with cRNA the day after surgery.
2. Keep oocytes in sterile petri dishes in BARTH’s pH 7.4 supple-
mented with gentamycin (50 μg/ml) at 16 °C.
3. Pull injection needles from glass capillaries. Break the needles

prior to injection under a stereomicroscope to achieve an aper-
ture of about 20 μm (see Note 8).
4. Prepare cRNA working solutions in the desired concentration
(see Note 9).
5. Transfer oocytes by gravity drop into the injection groove filled
with Barth’s (see Note 10, Fig. 1).
6. Inject each oocyte with max. 50 nl containing the optimal
amount of cRNA or water (negative control) under a stereo
microscope.
7. Transfer the oocytes into a small sterile plastic petri dish contain-
ing Barth’s pH 7.4 supplemented with gentamycin (50 μg/ml).
8. Incubate the oocytes until optimal expression is achieved in
Barth’s pH 7.4 supplemented with gentamycin (50 μg/ml) at
16 °C (see Note 11). Change the buffer every day and discard
dead or damaged oocytes.
3.3 Efflux 1. Prepare and fill one recovery plate (Fig. 2a) and one flux plate
Experiments (Fig. 2b) for each construct to test, as shown in Fig. 2. Fill each
with Radioactive well required with 2 ml Barth’s (see Note 12). Place all the
Labeled Substances “recovery”-plates in the fridge (4 °C) to cool them. Store the
“efflux” plates at room temperature. Before starting the injec-
tion put one “recovery” plate on ice.
2. Prepare the injection needles as described in Subheading 3.2.
3. Prepare substrate for injection (see Note 13).
4. Place the injection groove on an ice-cold metal block.
5. Place the petri dish containing the oocytes on ice.
6. Fill the injection groove with ice-cold Barth’s.
7. Inject ≥10 oocytes per time point with radiolabeled substrate.
Fig. 2 Schematic representation of recovery an flux plates. (a) Recovery plate. This plate is placed on ice.
Oocytes are placed in the top vial to allow closure of the injection spot and diffusion of substrate in the oocyte.
Oocytes are transferred to the wash well by gravity drop and the to the top vial of the flux plate. (b) Flux plate.
This plate is kept at room temperature. Oocytes are allowed to transport substrate for a defined amount of time
and then transferred to the wash solution by gravity drop
8. Transfer the oocytes immediately after injection into ice-cold

Barth’s in the first upper well of the “recovery” plate and keep
them on ice for 10 min.
9. To wash, transfer oocytes by gravity drop into the well below
containing 2 ml of ice-cold Barth’s.
10. For the second wash, transfer oocytes by gravity drop into the
well below containing 2 ml of ice Barth’s.
11. Transfer the oocytes in the first uppermost well of the flux plate
and allow them to efflux for 30 min at room temperature.
12. Proceed with the other time points (15 and 7.5 min) and use
for each time point a new column of the efflux plate (Fig. 2b).
13. After the efflux time is up, wash the oocytes twice by transfer-
ring them to the next well by gravity drop.
14. Transfer oocytes individually into one scintillation vial for each
oocyte.
15. Oocytes representing the 0 min are only put in the recovery
plate for 10 min, washed twice, and immediately separated into
one scintillation vial per oocyte.
16. Lyse the oocytes in the scintillation vials with 100 μl 10 % SDS
and let them burst (app. 10 min).
17. Add 4 ml scintillation cocktail to the lysed oocytes and vortex
vigorously.
18. Determine the amount of radioactivity in each vial by liquid
scintillation counting.
3.4 Import 1. Place 12 oocytes into the first well of a flux plate (Fig. 2b)
Experiments containing 2 ml Ringer/Barth’s containing the desired amount
with Radioactive of labeled substrate (see Note 12).
Labeled Substances 2. Incubate oocytes for an appropriate time, usually between 5
and 120 min depending on the expression level of the
transporter. In order to do kinetics at least four time points are
required!
3. To wash oocytes, transfer them into 2 ml of ice-cold Ringer/
Barth’s by gravity drop.
4. Wash a second time.
5. After the second wash transfer individual oocytes to one scin-
tillation vial per oocyte.
6. Lyse oocytes by addition of 100 μl 10 % SDS.
7. Vortex vigorously to homogenize, and add 4 ml scintillation
cocktail.
8. Determine the amount of accumulated radioactivity by liquid
scintillation counting.
3.5 Immunoblotting 1. Transfer 8–25 oocytes in a reaction cup.

of Microsomal 2. Homogenize in 40 μl homogenization buffer per oocyte with
Fractions a 1 ml pipet tip by trituration, i.e., pipetting up and down vig-
orously (see Note 14).
3. Centrifuge at 2000 × g and 4 °C for 10 min.
4. Transfer the supernatant in a reaction cup suitable for ultracen-
trifugation (see Note 15).
5. Fractionate the solution in an ultracentrifuge at 150,000 × g
and 4 °C for 30 min.
6. Transfer the supernatant, which represents the soluble protein
fraction, in a new reaction cup and store it for further analysis.
7. Resuspend the pellet in 8 μl homogenization buffer supple-
mented with 4 % SDS.
8. Perform SDS-PAGE and Western blot according to standard
protocols (see Note 16).
3.6 Biotinylation In order to be able to compare the amount of transporter in the

of Plasma Membrane membranes between experiments it is necessary to assess the
Proteins amount of transporter in the oocyte plasma membrane.
1. Use 20 oocytes for each construct. As control use water-
injected oocytes (see Note 17).
2. Wash oocytes twice in Barth’s without antibiotics and incubate
for 30 min in Barth’s at 16 °C (see Note 18).
3. Transfer oocytes in a reaction cup.
4. Wash oocytes three times with 1 ml PBS at 4 °C for 1 min.
5. Incubate the oocytes for 15 min in 0.5 ml biotin solution at
room temperature in the dark. Turn the reaction cup slightly
from time to time to evenly biotinylate the oocytes.
6. Wash three times with 1 ml ice-cold PBS for 2 min.
7. Wash once with 1 ml quench buffer.
8. Incubate oocytes with 1 ml quench buffer on ice for 20 min.
9. Wash once with 1 ml ice-cold PBS.
10. To lyse oocytes incubate for 30 min with 100 μl lysis buffer on
ice and homogenize by trituration.
11. Centrifuge at 1500 × g and 4 °C for 15 min.
12. Transfer the supernatant to a new reaction cup (see Note 15).
13. Perform Bradford analysis to determine the protein content.
14. Prepare the streptavidin–agarose. Vortex vigorously before
use. Pipet 50 μl with a cut pipet tip in a reaction cup. Centrifuge
shortly and remove the supernatant. Wash once with 1 ml lysis
buffer and remove supernatant after a centrifugation step.
15. Bind the biotinylation reaction to the streptavidin–agarose

beads. Pipet a defined amount of protein to the washed strep-
tavidin–agarose (see Note 19).
16. Incubate overnight at 4 °C while slightly shaking.
17. Centrifuge at 5000 g and 4 °C for 1 min.
18. Collect the supernatant (not biotinylated fraction).
19. Wash biotin-streptavidin-agarose pellet four times with 1 ml
ice-cold PBS. Remove supernatant after a centrifugation step
(5000 × g, 4 °C, 1 min).
20. Add 20 μl of 4× SDS-PAGE sample buffer and boil at 95 °C at
5 min (see Note 20).
21. Perform SDS-PAGE and Western blot.
22. Use a specific antibody to detect your protein of interest.
4 Notes
1. The quality of the oocytes used is absolutely crucial. The laws

regulating keeping of vertebrates are becoming increasingly
strict. As a consequence, keeping frogs in your own depart-
ment may be complicated and starting the own facility may be
impossible due to regulations. In many universities Xenopus
facilities exist in medical faculties and/or zoology/animal
physiology departments. It is certainly of great advantage to
obtain the oocytes from a department in which oocytes are
routinely used. As an alternative, it is possible—albeit fairly
expensive—to purchase oocytes from private companies.
2. It is necessary to keep in mind that the NcoI site will generate
an ATG start codon. pOO2 contains the 3′-UTR of the the
Xenopus laevis beta globin gene and a poly A sequence behind
the multiple cloning site to stabilize the cRNA transcript [11].
For in vitro transcription the SP6-RNA-polymerase-promotor
is used.
3. Linearization of the vector forces transcript termination. This
increases the yield of transcripts with an optimal size. Cut after
the poly A sequence. MluI or PmlI can be used. If the con-
struct contains both sites introduction of a silent mutation to
eliminate the restriction site is required.
4. This step removes any residual RNase and is absolutely required
because in most Midi/Maxi Kits RNase A is added during the
plasmid purification process.
5. We recommend the optional DNase I treatment to remove the
linearized plasmid DNA.
6. Best recovery rates are achieved when the elution is done with
preheated (95 °C) elution buffer. No second elution is per-
formed. Typically, around 25 ug cRNA is recovered. If higher
cRNA concentrations are required perform a LiCl (provided
with the mMESSAGE mMACHINE® SP6 Transcription Kit)
precipitation according to the manufacturer’s instructions.
7. To ensure that the cRNA is not degraded, standard electropho-
resis on a TAE/TBE gel is generally sufficient. In rare instances
more than one band can be detected. If this is the case a dena-
turing RNA gel electrophoresis is required. In very few cases
the appearance of another band (we never observed more than
one additional band) may be due to internal transcription start
or premature termination. In these instances, the cRNA can be
used but it must be checked subsequently by Western blotting
that only the expected protein is present in oocytes.
8. The settings to pull injection needles depend on the platinum
filament and vary between pullers and filaments. We recom-
mend a rather slow pull with medium heat to yield rather long
needles. Reading the “pipette cookbook” (http://www.sutter.
com/PDFs/pipette_cookbook.pdf) is recommended. We
break the needles open by gently pushing them against the rim
of a petri dish under a stereomicroscope. Opening size is
checked under a microscope. With a little bit of experience this
process becomes very reproducible.
9. This point is critical but entirely empirical and will vary between
transporters. When establishing this method start with 1, 0.5,
and 0.1 μg/μl. If the protein is GFP labeled monitor the
increase in fluorescence. Alternatively perform Western to
check protein abundance. If neither of that is possible: Perform
transport experiments to see if the activity is present and closely
monitor changes depending in expression time. It is going to
be a race between death and optimal expression. The oocyte
membrane becomes instable when too much protein is inserted
and the oocytes rupture. As a result some transporters must be
characterized a day after RNA injection—others will take up to
a week to accumulate in the membrane to an amount that
allows performing the experiment. If more than one protein is
expressed it is necessary to optimize the expression of all pro-
teins. It is mandatory to establish that upon coexpression of
another protein the amount of transporter does not decrease.
In these cases it is necessary to adjust amount of cRNA or
expression time accordingly.
10. To transfer the oocytes—particularly when handling oocytes
injected with radioactivity or which are transferred out of a
radioactive solution—we break glass Pasteur pipettes to yield
an opening of about 3 mm and flame polish the opening to
melt sharp edges. We then insert the pipettes into the pi-pump®
and suck up the oocytes. The oocytes sink to the bottom of the
liquid. If the liquid level is exactly at the opening of the pipette,
the oocytes will fall out when the pipette is attached to the
surface of a liquid. The transfer of liquid is extremely low. We
checked the carryover of radioactivity with the oocytes between
wells (see below) and found no radioactivity in the second wash
solution.
11. Lab incubators that can be used at 16 °C are fairly expensive. As
a considerably cheaper alternative we use a wine cooler that can
be adjusted to 16 °C (purchasing this equipment may lead to
interesting discussions with the university’s accounting staff).
12. Until efflux oocytes are kept cold to minimize loss of activity
during handling and closure of the injection spot. Oocytes are
more stable at pH 7.4. It may become necessary to vary the
external pH. In this case we generally use Ringer solution. This
has historical reasons. In electrophysiology Ringer is used
because it has higher ionic strength. Adjusting pH in Barth’s
will probably work and show no difference. We never tried.
Mainly because “this is how we always do it.”
13. Carefully check pH! The pH inside an oocyte is 7.4 [12].
Substrates must also be adjusted to this pH! Oocytes will burst
when injecting an acid! In case of auxin, start with a dilution of
radioactivity that injection of 50 nl results in an end concentra-
tion of 1 μM radiolabeled auxin inside the oocyte based on an
oocyte volume of 400 nl after injection [13].
14. If the phosphorylation status of the transporter upon coexpres-
sion of a kinase should be analyzed, it is necessary to use phos-
phatase inhibitors. For best results use fresh, not frozen oocytes.
Perform the SDS-PAGE and Western blot immediately.
15. Fat from yolk will be covering the supernatant. Transfer as lit-
tle fat as possible.
16. To solubilize, do not boil membrane proteins in SDS-PAGE
sample buffer prior to PAGE. Incubate at 42 °C for 15 min.
17. Carefully check the pH of all buffers. During the washing steps
it is crucial that the oocytes do not burst. To remove residue
shake the oocytes carefully.
18. The buffers used for biotinylation must not contain primary
amino groups, as biotin reacts with them.
19. Several dilutions of membrane factions must be checked to
determine the binding capacity of streptavidin agarose. Do not
overload and saturate the streptavidin agarose.
20. In this case it is necessary to boil to break the streptavidin from
the biotinylated protein! It may run differently from the pro-
tein that is solubilized at 42 °C.
Acknowledgement
This work was supported by DFG grant HA3469/6-1.
References
1. Gurdon JB, Lane CD, Woodland HR, Marbaix mediated stamen development in Arabidopsis.
G (1971) Use of frog eggs and oocytes for the Nat Commun 6:6095
study of messenger RNA and its translation in 8. Swarup K et al (2008) The auxin influx carrier
living cells. Nature 233:177–182 LAX3 promotes lateral root emergence. Nat
2. Boorer KJ, Forde BG, Leigh RA, Miller AJ Cell Biol 10:946–954
(1992) Functional expression of a plant plasma 9. Yang Y, Hammes UZ, Taylor CG, Schachtman
membrane transporter in Xenopus oocytes. DP, Nielsen E (2006) High-affinity auxin
FEBS Lett 302:166–168 transport by the AUX1 influx carrier protein.
3. Zampighi GA, Kreman M, Boorer KJ, Loo Curr Biol 16:1123–1127
DD, Bezanilla F, Chandy G, Hall JE, Wright 10. Zourelidou M et al (2014) Auxin efflux by
EM (1995) A method for determining the PIN-FORMED proteins is activated by two
unitary functional capacity of cloned channels different protein kinases, D6 PROTEIN
and transporters expressed in Xenopus laevis KINASE and PINOID. ELife 3:e02860
oocytes. J Membr Biol 148:65–78 11. Ludewig U, von Wiren N, Frommer WB
4. Krouk G et al (2010) Nitrate-regulated auxin (2002) Uniport of NH4+ by the root hair
transport by NRT1.1 defines a mechanism for plasma membrane ammonium transporter
nutrient sensing in plants. Dev Cell LeAMT1;1. J Biol Chem 277:13548–13555
18:927–937 12. Broer S, Schneider HP, Broer A, Rahman B,
5. Pellizzaro A et al (2014) The nitrate trans- Hamprecht B, Deitmer JW (1998)
porter MtNPF6.8 (MtNRT1.3) transports Characterization of the monocarboxylate
abscisic acid and mediates nitrate regulation transporter 1 expressed in Xenopus laevis
of primary root growth in Medicago truncat- oocytes by changes in cytosolic pH. Biochem
ula. Plant Physiol 166:2152–2165 J 333(Pt 1):167–174
6. Ranocha P et al (2013) Arabidopsis WAT1 is a 13. Stegen C, Matskevich I, Wagner CA, Paulmichl
vacuolar auxin transport facilitator required M, Lang F, Broer S (2000) Swelling-induced
for auxin homoeostasis. Nat Commun taurine release without chloride channel activ-
4:2625 ity in Xenopus laevis oocytes expressing anion
7. Saito H et al (2015) The jasmonate-responsive channels and transporters. Biochim Biophys
GTR1 transporter is required for gibberellin- Acta 1467:91–100
Chapter 22
Characterizing Auxin Response Circuits in Saccharomyces

cerevisiae by Flow Cytometry
Edith Pierre-Jerome, R. Clay Wright, and Jennifer L. Nemhauser
Abstract
Recapitulation of the nuclear auxin response pathway in Saccharomyces cerevisiae (yeast) provides a means
to functionally assay the contribution of individual signaling components to response dynamics. Here, we
describe a time course assay for characterizing auxin response circuits using flow cytometry. This method
allows for quantitative measurements of the dynamic response of up to 12 circuits (strains) at once. We also
describe a steady-state assay and how to utilize an R package we developed to facilitate data analysis.
Key words Synthetic biology, Budding yeast, Fluorescent reporters, Signaling dynamics
1 Introduction
Auxin influences nearly every aspect of plant growth and develop-

ment. The core auxin signaling pathway consists of only a handful of
components, and yet is capable of eliciting a diverse array of context-
specific responses to auxin. Experimental analysis of this small signal-
ing pathway in planta is confounded by the ubiquity of auxin
response, integration and feedback from other signaling pathways,
and genetic redundancy among the core signaling components. To
test the capacity of the auxin signaling pathway to generate diverse
responses to auxin, we transplanted each component of the pathway
from Arabidopsis to the budding yeast Saccharomyces cerevisiae [1].
A minimal yeast auxin response circuit is composed of a TIR1/
AFB auxin receptor, an Aux/IAA (IAA), a TOPLESS (TPL) co-
repressor, an ARF transcription factor, and an auxin-responsive
promoter driving a fluorescent reporter (Fig. 1). In the absence of
auxin, ARF activity is repressed through interaction with an IAA-
TPL fusion. Auxin treatment promotes interaction between the
IAA and the AFB receptor, an F-box protein, which catalyzes the
ubiquitination and subsequent degradation of the IAA. IAA deg-
radation relieves the repression of the ARF transcription factor
271
272 Edith Pierre-Jerome et al.
SCF Ubiquitin Ligase
AFB Ub
IAA
Ub
Ub
Ub U
b
TPL
IAA
ARF
ARF
x Fluorescent Reporter
Fluorescent Reporter
OFF state auxin ON state
Fig. 1 A yeast auxin response circuit. An AFB auxin receptor, an IAA-TPL repressor, an ARF transcription factor,
and an auxin-responsive promoter driving a fluorescent reporter are sufficient to recapitulate auxin-induced
transcription in yeast
resulting in the induction of the fluorescent reporter. Flow cytom-

etry is used to monitor levels of the fluorescent reporter over time
following auxin treatment.
This user-defined auxin response circuit can be used to gener-
ate numerous circuit variants by substituting variants of any one of
the core components alone or in combination. These component
variants can be derived from the naturally evolved gene family
members or by altering a functional domain within a component
through mutation or truncation [1, 2]. The modularity of the cir-
cuit also facilitates decomposition into sub-modules to focus on
specific aspects of the pathway, such as auxin-induced degradation
of the IAAs. Auxin response circuit composition can therefore be
tailored to suit the need of the researcher. As such, we focus here
on describing a method for the quantitative characterization of
auxin circuits rather than detailing their assembly.
2 Materials
2.1 Yeast Materials 1. Diploid yeast strains containing circuit variants to be assayed
(see Note 1).
2. YPDA agar plates: Add about 700 mL of water to a glass bea-
ker. Add a magnetic stir bar and set on a stir plate. Add each
ingredient while stirring: 10 g yeast extract, 20 g of peptone,
80 mg of adenine. Bring volume up to 900 mL with water. Stir
well, and then split into two 450 mL aliquots in 1 L bottles.
Add 10 g of agar into each bottle. Autoclave for 25 min at
121 °C. After cooling, bring media up to final volume of
Auxin Response Circuits in Yeast 273
500 mL per bottle containing 2 % glucose by supplementing

with 50 mL 20 % sterile filtered glucose. Pour 20 mL molten
YPDA into 95 × 15 mm petri dishes as required.
3. Synthetic Complete (SC) liquid media: Add about 1 L of water
to a glass beaker. Add a magnetic stir bar and set on a stir plate.
Add each ingredient while stirring: 2.55 g of yeast nitrogen
base without amino acids or ammonium sulfate (Sigma); 2.1 g
of yeast synthetic drop-out medium without histidine, leucine,
tryptophan, and uracil (Sigma); 7.5 g ammonium sulfate;
150 mg of histidine; 150 mg of leucine; 150 mg of trypto-
phan; 37.5 mg of uracil; and 120 mg adenine. Bring volume
up to 1.35 L with water. Stir well, and then split into two
675 mL aliquots in 1 L bottles. Autoclave for 25 min at 121 °C.
After cooling, bring each bottle to a final volume of 750 mL by
adding 75 mL 20 % sterile filtered glucose. Store at room tem-
perature and protect liquid media from light.
2.2 Cytometry Assay 1. 17 × 100 mm sterile plastic (translucent polypropylene) culture

Materials tubes with snap cap.
2. 50 mL plastic conical centrifuge tubes.
3. 25 mL serological pipettes.
4. Pipet-Lite™ XLS Adjustable-Spacer Multichannel Manual
Pipette, 100–1200 μl (Rainin).
5. Multiplex culture tube lids (see Note 2).
6. 96-well PCR plates.
7. Indole-3-acetic acid (auxin).
8. 95 % EtOH.
9. 2000 μl Eppendorf™ Deepwell™ Plates 96.
10. 25 mL reagent troughs.
11. 12-channel pipette with 100 μl capacity.
12. AeraSeal™ breathable film seals (Excel Scientific).
2.3 Equipment 1. Flow cytometer with plate capacity configured to detect GFP
or YFP. We currently take fluorescence measurements with a
BD Accuri™ SORP flow cytometer with a CSampler 96-well
plate adapter using an excitation wavelength of 514 nm and an
emission detection filter at 545/35 nm (see Note 3).
2. Benchtop shaker incubator amenable to rapid and repeat sam-
pling of growing yeast cultures such as a MaxQ™ 4000
Benchtop Oribital Shaker (Thermo Scientific).
2.4 Data Analysis We have developed a package called flowTime for the R program-
Software ming language to facilitate data analysis for these assays. This pack-
age will soon be available via http://bioconductor.org (current
and continuing development version available at http://github.

com/wrightrc/flowTime). We suggest using RStudio (http://
www.rstudio.com) to edit and run the R-code for your data analy-
sis. Numerous resources are available on the internet to gain an
introductory knowledge of the R language [3].
3 Methods
3.1 Auxin Time This protocol can be used to assay auxin-induced transcription of
Course Assay up to 12 strains in parallel. This constraint is due to the time
required for auto-sampling of multiple wells by the cytometer as
well as to minimize the time a yeast culture sits in the sampling
plate. We therefore generally limit each read to 12 strains. This
assay can also be used for monitoring auxin-induced degradation
with only minor adjustments (see Note 4).
Day One
1. For each strain to be assayed, inoculate ¼ to ½ of a fresh yeast
colony from a YPDA plate into 3 mL of synthetic complete
(SC) media. Vortex well to mix. Aliquot 100 μl into a 96-well
plate and read on cytometer to estimate cell density (in events
per microliter).
2. Export data as .FCS files and import into R as a flowSet (see
Subheadings 3.3 and 3.4, for example R-code). Create sum-
mary statistics for this flowSet using the summary.cyt func-
tion with the “only” parameter set to “yeast” to subset the
data for all yeast. The value in the “conc” column of the
resulting data frame is the concentration in yeast events per μL
for each sample or flowFrame. Use this value to calculate the
volume of the initial 3 mL inoculant required to prepare the
requisite dilution.
3. Dilute each strain to 0.25 events per μL in 12 mL of SC in a
50 mL conical tube. Mix well by vortexing.
4. Split dilutions into duplicate 5 mL aliquots in sterile plastic
culture tubes and incubate for 16 h at 30 °C with shaking at
220 rpm.
Day Two
5. After 16 h of growth, combine duplicate aliquots to homoge-
nize cultures. Mix gently by pipetting and split back into two
5 mL aliquots.
6. Separate duplicate cultures into two groups of tubes while
maintaining the strain order between groups. This will be the
order samples are read on the cytometer. Return tubes to
shaker and replace individual plastic caps with 3D printed
multiplex tube lids. Use of these lids facilities rapid sampling of

the cultures using the Rainin Adjustable-Spacer Multichannel
Pipette.
7. Set cytometer settings to measure 10,000 events for each well
at a flow rate of 66 μl/min and core size of 22 μm with a time
limit of 1 min. Use these settings for the entire experiment.
8. Transfer 100 μl of each culture from one replicate group to a
96-well PCR plate by expanding the multichannel pipette to
sample from the culture tubes and collapsing to dispense into
the PCR plate. Fill an additional well with 200 μl of sheath
fluid to serve as a wash between readings (see Note 5).
9. Load PCR plate containing samples onto cytometer sampling
arm and run the auto-sampling program to flow samples. The
density of each culture during this initial read should be around
200 events/μl (see Note 6). Repeat steps 8 and 9 for the sec-
ond replicate of cultures.
10. For each strain, mock treat the first replicate measured with
95 % ethanol. Immediately transfer 100 μl of each mock-treated
culture to the next row of the 96-well PCR plate and read
samples on the cytometer to record fluorescence. These mea-
surements serve as the 0-min time point.
11. Treat each strain of the second replicate of cultures with indole-
3-acetic acid (IAA) dissolved in 95 % ethanol to a final concen-
tration of 10 μM. Immediately after auxin addition, transfer
100 μl of each culture to a new 96-well PCR plate. Open a new
file with the settings described in step 7 and measure fluores-
cence for the 0-min time point.
12. Use the same plate and file to acquire measurements for auxin-
treated samples at 10-min intervals. Prepare new plates and a
corresponding cytometer file as needed to acquire 5 h of data.
13. Take measurements for mock-treated samples every hour using
the same plate and file used in step 9.
14. For each plate file, export all wells as a folder of .FCS files.
There should be a separate plate file containing the initial and
mock-treated reads and several files for the auxin-treated
measurements.
3.2 Auxin Steady- Day One

State Assay
1. For each strain to be assayed, inoculate ¼ to ½ of a fresh yeast
colony from a YPDA plate into a well of a 2000 μl Eppendorf™
Deepwell™ Plate 96 containing 500 μl of SC media.
2. Seal plate with a breathable film seal and incubate at 30 °C
with shaking at 375 rpm for 16 h.
Day Two
3. After 16 h of growth, dilute strains 1:200 into a new well con-
taining 500 μl of fresh SC media in a deep 96-well plate (see
Note 7). Prepare two replicate wells for each strain.
4. Return plate containing diluted cultures to the shaker for 2 h
of additional growth at 30 °C with shaking at 375 rpm (see
Note 8).
5. After 2 h, mock (95 % EtOH) treat one replicate well and auxin
(indole-3-acetic acid) treat the other (see Note 8). To assay
more than 12 strains or multiple doses of auxin, treatment
additions must be staggered for every 12 wells. This is impor-
tant to allow time for data acquisition and maintain a consis-
tent treatment time between reads.
6. Return plate to shaker for an additional 4 h of growth (see
Note 8).
7. Set cytometer settings to measure 20,000 events for each well
at a flow rate of 66 μl/min and core size of 22 μm with a time
limit of 1 min.
8. Sample up to 12 strains at once using a multi-channel pipette to
transfer 100 μL of culture from the deep-well plate in the shaker
to a PCR plate. Fill an additional well in the PCR plate with
200 μl of sheath fluid to serve as a wash between readings.
9. Load the PCR plate onto cytometer sampling arm and run the
auto-sampling program to measure fluorescence.
10. Repeat steps 8 and 9 until all strain/treatment combinations
have been measured.
11. Typically, data for at least three independent replicates for each
strain/treatment is collected.
12. Export data as a folder of .FCS files.
3.3 Time Course Here, we demonstrate how to import the .FCS files into R, gate
Data Analysis in R and annotate this data with experimental metadata (e.g., the strain
and treatment for each sample), generate summary statistics for
each sample and time point, and finally plot the time course data
(see Note 9).
3.3.1 Importing 1. Import your flow cytometry data using read.flowSet. Here,
and Annotating Data we will import an example flowSet:
plate1 <- read.flowSet(path = system.file("extdata",
"tc_example/", package = "flowTime"), alter.names = T)
# Next, add plate numbers to the sampleNames
sampleNames(plate1) <- paste(‘1_’,
sampleNames(plate1),
sep = “”)
2. If you have several plates, this code can be repeated and each
plate can be combined (using rbind2) to assemble the full
dataset:
plate2 <- read.flowSet(path = paste(experiment, "_2/", sep = ""),
alter.names = T);
sampleNames(plate2) <- paste("2_", sampleNames(plate2), sep = "");
dat <- rbind2(plate1, plate2)
3. Import the table of metadata for each well. The “sample-

Names” field of the assembled flowSet (“dat” in this exam-
ple) must match that of a unique identifier column in the table
of metadata. In this example and as a default name is the unique
identifier column. One can also create this column from a
flowSet and attach the annotation columns, as demonstrated in
the code below. The order of the unique identifier column
does not matter, as annotateFlowSet will join annotation to
“dat” by matching identifiers, as specified by the “mergeBy”
parameter:
annotation <- read.csv(system.file("extdata", "tc_example.csv",
package = "flowTime"));
sampleNames(dat) #view the sample names
sampleNames(dat) == annotation$id #replace 'id' with the unique
identifier column to test if this column is identical to the
sample names of your flowset
annotation <- cbind(annotation, names = sampleNames(dat)) #If
the sampleNames and unique identifiers are in the correct order
this command will add the sampleNames as the identifier
4. Attach this metadata to the flowSet using the annotateFlow-
Set function:
adat <- annotateFlowSet(dat, annotation);
head(rownames(pData(adat)))
#> [1] "0_A08.FCS" "0_B08.FCS" "0_C08.FCS" "0_D08.FCS"
"0_E08.FCS" "0_F08.FCS"
head(pData(adat))
#>name X strain RD ARF AFB treatment
#> 0_A08.FCS 0_A08.FCS 0_A08 3 TPLRD1 19 AFB2 0
#> 0_B08.FCS 0_B08.FCS 0_B08 3 TPLRD1 19 AFB2 0
#> 0_C08.FCS 0_C08.FCS 0_C08 3 TPLRD1 19 AFB2 0
#> 0_D08.FCS 0_D08.FCS 0_D08 3 TPLRD1 19 AFB2 0
#> 0_E08.FCS 0_E08.FCS 0_E08 3 TPLRD1 19 AFB2 0
#> 0_F08.FCS 0_F08.FCS 0_F08 3 TPLRD1 19 AFB2 0
3.3.2 Compiling 1. Use the summary.cyt function to compile the summary statis-
and Plotting Time Course tics from the raw data in this flowSet. This function will gate
Data each flowFrame in the flowSet and compile and return a
dataframe of summary statistics in the specified channel for

each flowFrame:
# load the gate set for BD Accuri C6 cytometer
loadGates(gatesFile = "C6Gates.RData");
dat_sum <- summary.cyt(adat, ploidy = "diploid", only =
"singlets", channel = "FL1.A")
#> [1] "Gating with diploid gates…"
#> [1] "Summarizing singlets events…"
2. Use the resulting dataframe to plot mean fluorescence for

each mock- and auxin-treated sample over time to generate
induction curves (Fig. 2):
qplot(x = time, y = FL1.Amean, data = dat_sum, linetype =
factor(treatment)) +
geom_line() + xlab("Time post Auxin addition (min)") +
ylab("Reporter Fluorescence (AU)") +
scale_color_discrete(name = expression(paste("Auxin (", mu,
"M)", sep = ""))) +
theme_classic(base_size = 14, base_family = "Arial")
3.4 Steady-State For steady-state assays, the entire dataset (all measured events/
Assay Data Analysis sample) can be readily plotted. Data import and annotation remains
the same.
Fig. 2 Example data from a time course assay. Induction of transcription in

response to auxin is shown for a mock- (solid line) and auxin-treated (dashed
line) circuit
3.4.1 Compiling and 1. Use the steadyState function to gate each flowFrame in the
Plotting Steady-State Data flowSet and compile and return a dataframe of the relevant
data and metadata for each event:
loadGates(gatesFile = "SORPGates.RData")
dat.SS <- steadyState(flowset = adat, ploidy = "diploid", only =
"singlets")
#> [1] "Gating with diploid gates…"
#> [1] "Converting singlets events…"
2. Use the dataframe to plot the fluorescence for all events cap-
tured for each sample (Fig. 3):
p <- ggplot(dat.SS, aes(as.factor(treatment), FL2.A, fill = AFB))
+ geom_boxplot(outlier.size = 0) +
facet_grid(IAA ~ AFB) + theme_classic(base_family = "Arial",
base_size = 16) +
ylim(c(-1000, 10000)) + xlab(expression(paste("Auxin (", mu,
"M)", sep = ""))) +
ylab("Fluorescence (AU)") + theme(legend.position = "none");
p
Fig. 3 Example data from a steady-state assay. Each sub-panel represents a

different circuit built to assay auxin-induced degradation. Box plots display the
fluorescence distribution of IAA protein levels following a dose response assay
4 Notes
1. Strain construction for auxin response circuits has been previ-

ously described [1]. Briefly, we use laboratory strains W814-
29B MATα and W303-1A MATa which are both auxotrophic
for uracil, histidine, tryptophan, and leucine. Each component
of the pathway is singly integrated into the yeast genome at a
different auxotrophic locus. Typically, a TIR1/AFB receptor is
integrated at LEU2 and an IAA-TPL fusion is integrated at
TRP1 in W814-29B MATα. An ARF is integrated at HIS3 and
the auxin-responsive fluorescent reporter (i.e., pIAA19::GFP)
is integrated at URA3 in W303-1A MATa. The two strains can
then be mated to generate a diploid strain containing the entire
auxin circuit. For auxin-induced degradation assays, only intro-
duction of a TIR1/AFB receptor and a fluorescently tagged
IAA is required [2].
2. The .stl design file for the 3D printed multiplex lid is publically
available on Thingiverse (http://www.thingiverse.com/
thing:893490). Each lid can accommodate six culture tubes in
a 6 × 6 or 6 × 12 tube rack.
3. Our published work utilized a standard BD Accuri™ C6 flow
cytometer with a CSampler 96-well plate adapter configured
with a 488 nm laser and an emission detection filter at 533 nm
to measure GFP and YFP fluorescence levels. Our current
514 nm laser configuration optimizes YFP excitation but no
longer excites GFP. We tested several bench-top cytometers for
YFP and GFP detection in our strains and found the Accuri to
be the most sensitive.
4. For auxin-induced degradation assays, follow time course pro-
tocol with the following modifications. The first day, dilute
strains to 0.5 events/μl in 10 mL of SC and split into two
4 mL aliquots. Culture density during the initial reads the next
day should be ~500 events/μl. After treatments, samples are
measured at 10-min intervals for 3 h.
5. A wash well should be prepared for every read containing up
to 12 strains. We typically prepare the full PCR plates contain-
ing a wash well for each row so that only cultures need to be
transferred to the PCR plate at 10-min intervals.
6. A starting density of ~200 events/μl after 16 h is based on the
growth rate of our particular strains. We found this starting
density to be optimal to ensure that yeast are in log phase for
the duration of the experiment. If starting density is too low, it
is best to allow the yeast to grow for an additional 30 min to
an hour as fluorescence readings can be more variable.
7. For haploid strains, dilute cultures 1:100 in order to maintain
the time frame outlined.
8. We typically treat with 10 μM IAA but auxin treatment can

range from 0.05 to 50 μM. Dimethyl sulfoxide can be used as
a solvent in place of 95 % ethanol. To conduct a dose response
assay, replicate wells for each dose must be prepared in step 3.
For auxin-induced degradation assays, treatment should be
added 5 h post-dilution and fluorescence measured after 1 h.
9. To minimize noise in yeast cytometry data we typically present
fluorescence measurements from only singlet yeast cells (i.e.,
cells without a significantly sized bud), as budding increases cell
size and decreases the effective concentration of the circuit
components. Unfortunately, gates isolating populations of
singlet-yeast cells for both haploid and diploid strains are unique
for each yeast strain, shaker setup, and flow cytometer.
Determining the parameters of these gates is currently done by
trial and error in R. More visual programs such as flowJo may
be helpful. See the flowCore documentation [4] for more
information about using R to create gates, and refer to [5] for
an example of how to identify live and singlet cell populations.
References
1. Pierre-Jerome E, Jang SS, Havens KA, Nemhauser 4. Ellis B, Haaland P, Hahne F, Le Meur N,
JL, Klavins E (2014) Recapitulation of the for- Gopalakrishnan N, Spidlen J, Jiang M flowCore:
ward nuclear auxin response pathway in yeast. Proc basic structures for flow cytometry data. R pack-
Natl Acad Sci U S A 111(26):9407–9412 age version 1.36.9. https://www.bioconduc-
2. Havens KA, Guseman JM, Jang SS et al. (2012) t o r. o rg / p a c k a g e s / r e l e a s e / b i o c / h t m l /
A synthetic approach reveals extensive tunability flowCore.html
of auxin signaling. Plant Physiol 160:135–142 5. Baxter SK, Lambert AR, Scharenberg AM,
3. R Core Team (2015) R: a language and envi- Jarjour J (2013) Flow cytometric assays for inter-
ronment for statistical computing. R Foundation rogating LAGLIDADG homing endonuclease
for Statistical Computing, Vienna, Austria, DNAbinding and cleavage properties. Methods
http://www.R-project.org/ Mol Biol 978:45–61
Printed on acid-free paperLife Sciences
INDEX
A cell wall elasticity ................................................128–129

data analysis ........................................................129–131
Abscisic acid (ABA) ......................................... 221, 222, 249 preparation of cells......................................................128
Acidic peptides ................................................................... 27 Auto clustering ................................................................. 113
Adenine N6-substituted organic molecules ........................ 81 Autocrine (self ) signaling ................................................... 19
Adobe Illustrator ................................................................ 88 Automatic scoring .................................................. 58, 66, 70
Adobe Photoshop ......................................................... 62, 88 Auxin ..................... 2, 221, 222, 231, 249, 271–276, 278–281
AFM. See Atomic force microscopy (AFM) binding ............................................................... 160, 169
Agar plates.......................................................................... 94 steady state assay .................................................275–276
Alanine scanning technique................................................ 25 transport ............................................................. 207, 260
Amino acid substitutions, peptides ..................................... 25 treatment ................................................................75, 76
Analyze Cells 3D ..................................................... 109–111 Auxin binding protein1 (ABP1) ............................... 160, 207
Antagonistic peptides ......................................................... 25 Auxin response factors (ARFs) ................. 136, 137, 142–143
Apical hook development ......................................... 1, 3, 5, 7 Axial sampling rate ............................................................. 87
definition ........................................................................6
ethylene levels .................................................................2 B
formation ........................................................................1
Bezier curve ...................................................... 105–108, 112
kinetic analysis ................................................................6
β-Glucuronidase (GUS) ............................................... 73–78
materials and chemicals ..............................................3–5
BIACORE 2000 ......................................160, 162, 163, 174,
methods ......................................................................5–6
180, 187, 189, 190
molecular pathways.........................................................3
BIAevaluation software ............................ 174, 184, 188, 190
monitoring early phases ..................................................7
Binding site ...................................................................... 196
phases .............................................................................1
Biological assays ................................................................. 24
real-time analysis ........................................................2, 7
Biotinylation ..................................................... 262, 267, 269
Arabidopsis ........................................................ 109, 232–234
Brassinosteroid insensitive 1 (BRI1)......................... 207, 209
liquid medium for .........................................................23
Brassinosteroid receptor ................................................... 207
solid medium for.....................................................22–23
Brassinosteroids (BRs).............................................. 249, 250
sterilization of seeds......................................................22
Bright yellow 2 (BY-2) cells .............. 125, 127, 128, 130, 131
studies conducted in .....................................................24
Bright-field (BF) ................................................................ 97
Arabidopsis histidine transfer proteins (AHP) ................... 82
5-Bromo-4-chloro-3-indolyl-beta-d-glucuronide ............. 74
Arabidopsis response regulator (ARR) ............................... 82
Budding yeast ........................................................... 271, 281
Arabidopsis thaliana ........................... 9, 11–17, 29, 30, 47–49,
60, 69, 71, 82, 85, 92, 95, 125, 211, 252, 253 C
grafting .....................................................................9, 12
germination experiment..........................................58, 64 Calibration solution .......................................................... 252
merits ............................................................................57 CCD scanner...................................................................... 52
three-segment graft ................................................14–15 Cell Atlas 3D tools ........................................................... 110
two segment shoot-root graft .................................11–14 Cell clusters .............................................................. 112–114
two-segment self-grafts ................................................16 Cell data ................................................................... 114, 115
two-shoot Y-graft .........................................................16 Cell growth....................................................................... 131
Arduino ............................................................ 37, 38, 43–45 Cell type identification ......................112–114, 117, 119, 122
Area under the curve (AUC) .................................. 59, 68, 71 Cell types .................................................................. 111–112
ARF5................................................ 136–138, 140, 142–143 Cell wall ........................................................... 125, 128–131
ARFs. See Auxin response factors (ARFs) Cell wall elasticity ............................................. 125, 128–129
Atomic force microscopy (AFM) ............. 125, 126, 128, 132 Cells 3D ................................................................... 109–111
DOI 10.1007/978-1-4939-6469-7, © Springer Science+Business Media New York 2017
283
PLANT HORMONES: METHODS AND PROTOCOLS
284 Index
Cell-sorting (3–4 h).................................................. 238–240 saving and loading cell data ........................................114
Chelex ...................................................................... 201, 202 statistical analysis
Chimeric organisms.............................................................. 9 3D cell anisotropy .........................................115–116
ChIP. See Chromatin immunoprecipitation (ChIP) 3D reporter abundance .................................116–119
Chromatin ........................................ 193–195, 198–200, 202 topological check ........................................................114
aliquot .........................................................................200 3D cell anisotropy............................................. 100, 115–116
extraction ....................................................................193 3D cell mesh ..................................................................... 105
fragmentation .....................................................199–200 3D data sets ........................................................................ 99
preparation..........................................................198–199 3D image analysis ....................................................... 99–101
Chromatin immunoprecipitation (ChIP) ............193, 195–197, 3D reporter analysis tool .......................................... 116–119
200, 202 3D segmentation ...................................................... 102–105
principle ......................................................................195 Data interpretation ............................................................. 68
workflow .....................................................................194 Data processing ........................................................ 174–180
CK. See Cytokinin (CK) Data_CI.csv file ................................................................ 116
Classical plant hormones .................................................... 19 DBD. See DNA-binding domain (DBD)
Columella ................................................................. 110, 114 DBP trays ........................................................................... 68
Compound screen............................................. 159, 179, 180 De-cross-linking............................................... 199, 201, 202
Conducting germination tests ...................................... 32–33 Degron peptide..................................166–168, 170, 172, 174
Confocal laser scanning microscopy ............................. 87–88 Deoxyribonucleic acid (DNA)
Confocal microscopy ...........................95, 135, 138, 143–145 fragments .................................................................... 202
Control folder ................................................................... 116 recovery.......................................................................201
Cortical cells ............................................................. 110, 114 Derivatization ................................................... 223, 225, 228
Cotyledons ................................................... 9, 11, 13, 15–17 Digital single-cell analysis ................................ 100, 106, 111
cRNA ............................................................................... 268 Display cell data tool ........................................................ 115
oocyte injection........................................... 261, 263–264 DNA-binding domain (DBD) ......................... 137, 142–143
vitro transcription ....................................... 260–261, 263 Dose–response analysis ....................................................... 24
csv file ............................................................................... 122 Dot-PCB............................................................................ 42
Culturing plants ........................................................... 49–51 Double-transfected protoplasts......................................... 141
Curve-fitting ...................................................................... 68 Dropper tool ..................................................................... 104
Cysteine-rich peptides ........................................................ 20 DSMZ-German collection of microorganisms and cell
Cytokinin (CKs) .........................................81, 221, 222, 231, cultures............................................................ 126
234–236, 242, 249
CHASE........................................................................82 E
class ..............................................................................81 Efflux experiments............................................ 264–265, 269
purification .........................................................240–241 Elasticity ................................................... 125, 128–129, 132
quantification......................................................241–242 Endocrine (systemic) signaling ........................................... 19
signaling .......................................................................82 Endoglycosidase H (Endo H) ...........207, 209, 212–214, 218
stimulus ........................................................................82 Epitope gene .................................................................... 195
Error limit ................................................................ 110, 114
D
Escherichia coli ..................................................................... 73
2,4-D auxin ...................................................................... 131 Ethereal diazomethane ..................................................... 228
2D heatmap GUI ..................................................... 112–114 Ethylene ................................................................... 221, 249
3D CellAtlas .................................................... 100, 115–119 External hybrid (HyD) detectors...................................... 144
add-on ........................................................................100 Extra filtering software ..................................................... 155
Analyze Cells 3D ...............................................109–111
Arabidopsis hypocotyl ..................................................113 F
assign cell types...................................................111–112 FACS. See Fluorescence-activated cell sorting (FACS)
assign columella ..........................................................114 False discovery rate (FDR) ............................................... 153
assign cortical cells ......................................................114 FIJI (ImageJ) ................................................................ 48, 52
cell clusters..........................................................112–114 FLIM. See fluorescence lifetime imaging microscopy
displaying cell data......................................................115 (FLIM)
examine vasculature ....................................................114 Flow cytometry......................................................... 272, 273
merits ..........................................................................100 FlowTime ......................................................................... 273
sample for ...........................................................108–109 Fluorescence ................................................................. 95, 97
Index
285
microscopy ....................................................................82 High-performance liquid chromatography
proteins ............................................................... 135, 138 (HPLC) .................................................... 26, 223
reporters...................................................... 271, 272, 280 Histochemical staining ..................................... 73, 74, 76–79
Fluorescence lifetime imaging microscopy Hormones
(FLIM) ...................... 91, 136–138, 140–145, 147 classical plant ................................................................ 19
Fluorescence-activated cell sorting (FACS)............. 231–235, extraction ............................................................252–254
238, 240, 242, 245, 246 mammalian models.......................................................19
Flux experiments .............................................................. 262 paracrine peptide ..........................................................19
Forceps ....................................................... 10, 11, 13, 15–17 receptors ......................................206, 207, 209, 210, 214
Formation phase ............................................................... 2, 6 HPLC. See High-performance liquid chromatography
Förster resonance energy transfer (FRET) .............. 135–137, (HPLC)
141–143, 145 Hybond membrane................................................. 11–13, 17
FSC-A (Forward-scattered light) vs. SSC-A Hydroxyprolination ............................................................ 25
(Side-scattered light) ...................................... 232 Hypocotyl ........................................ 1–3, 6, 11–17, 29, 30, 35
Fucose deficient plants.............................................. 214–216
fut11 fut12 mutant ............................................................ 215 I
Identification and quantification software ........................ 155
G
Image acquisition....................... 60–61, 64–65, 100–101, 140
GA. See Gibberellins (GA) Image analysis software .............................58, 59, 61, 69, 101
GA-Fl ImageJ software ................. 59, 61, 62, 65–67, 69, 71, 74, 127
application .................................................................... 92 ImageJ-based quantification ......................................... 77–79
effects of light ...............................................................96 Imaris software ................................................................... 88
liquid work solution ......................................................93 Imbibition phase I .............................................................. 70
MS agar plates ..............................................................93 Immunoblotting ....................................... 213–214, 262, 266
pattern ..........................................................................92 Immunoprecipitation (IP) ................ 147–151, 194, 200–201
roots ..............................................................................94 chromatin beads..........................................................200
stock solution ................................................................93 de-cross-linking ..........................................................201
Gas chromatography-mass spectrometry DNA recovery ............................................................201
(GC-MS) ............................... 222, 225–226, 250 washes.................................................................200–201
Gene expression................................................................ 196 Import experiments .......................................................... 265
Genome-wide association mapping........................ 50, 52–54 Import export ................................................................... 260
Genome-wide association studies (GWAS) ................ 47, 48, Indole acetic acid (IAA) .................................. 126, 127, 129,
51–53 130, 271, 275, 279
Germination score .............................................................. 71 Infrared light .................................................................... 3–5
Germinator ....................................................... 58–63, 67, 68 Internal standard solution ................................. 251, 256, 257
curve fitting ..................................................................59 Interstock graft. See Three-segment graft
experiment design .........................................................59 IP. See Immunoprecipitation (IP)
image analysis ...............................................................59 Isothermal titration calorimetry (ITC) ............................. 160
Gibberellins (GAs) ..................................... 91, 221, 222, 249 ITK Smoothing Recursive Gaussian Blur ................ 103, 105
Glycoprotein...................... 206, 210, 211, 213, 214, 216–218 ITK Watershed Auto Seeded ........................................... 103
Glycosylation .................................................................... 205
Golgi apparatus ................................................ 206, 214–216 J
Grafting .......................................................................... 9, 13 Jasmonates ......................................... 221, 222, 249, 250, 257
Green fluorescent protein (GFP) ........................ 82, 87, 148,
153, 156, 232, 233, 273, 280 K
Group-specific parameters software ................................. 155
Karrikins (KARs) ......................................................... 30–35
Growing seedlings .............................................................. 85
Arabidopsis ............................................................... 29–35
Growth analysis 3D tool................................................... 116
germination tests ..............................................32–33
GUS. See β-Glucuronidase (GUS)
hypocotyl elongation assay .......................... 30, 33–35
GWAS. See Genome-wide association studies (GWAS)
methods ............................................................31, 32
H seed germination...............................................30–32
Kifunensine treatment ...................................... 213–214, 217
Heatmaps ................................................................. 117, 119 Kinetic analysis dialogue box ............................................ 190
Hertzian model ................................................................ 129 Kinetics............................................................. 159, 180–187
286 Index
L One cell sorting experiment (3–4 h) ......................... 237–238

One sorting experiment .................................................... 237
Laplacian smoothing ........................................................ 121 Opening phase.................................................................. 2, 6
l-arabinosylation ................................................................. 25 Optimal ImageJ ............................................................ 65–66
LC-ESI-MS/MS ............................................................. 251 Output folder ................................................................... 116
Limits of detection (LOD)....................................... 256, 257 Output type ...................................................................... 116
Limits of quantification (LOQ) ....................................... 256 Output_CI.csv file............................................................ 119
Linked reactions ............................................................... 184
Liquid chromatography-mass spectrometry P
(LC-MS) ........................................ 222, 250, 254
Paracrine (adjacent/nearby tissue) signaling ....................... 19
Liquid growth medium ...................................................... 83
Paracrine peptide hormones ............................................... 19
Liquid scintillation ................................................... 262, 265
Peptides ........................................................................ 19, 20
Load heat map.................................................................. 119
acidic.............................................................................27
Logarithmic transformation software ............................... 155
amino acid monomers...................................................20
Luciferase (LUC) reporters ................................................ 82
analysis of .....................................................................21
M antagonistic...................................................................25
basic ..............................................................................27
Mammalian models ............................................................ 19 cysteine-rich .................................................................20
Manhattan plot................................................................... 53 hormone .......................................................................19
Marching Cubes 3D ................................................. 105, 106 PTMPs (see Posttranslationally modified peptides
Mass spectrometry (MS) ..........................154, 222, 226, 233, (PTMPs))
241, 250, 253, 255, 257 purity ............................................................................26
sample preparation..............................................151–152 signal ............................................................................19
tandem ........................................................................152 uncharged .....................................................................27
Mass spectroscopy (MS)................................... 149–150, 152 use of synthetic .............................................................20
Mass transport limitation ................................................. 190 Petiole ..................................................................... 13, 15, 16
Matlab .............................................................................. 127 Petri dishes ....................................................4, 5, 11, 12, 236
MaxQuant protein identification ............. 149, 152–154, 158 Phosphate-buffered saline (PBS) ..................................... 242
MCP-PMT detectors....................................................... 144 Phytochrome interacting factors (PIFs) ........................... 193
Melatonin (MEL) ............................................................ 249 Phytohormone auxin ........................................................ 125
Micropurification (microSPE) ................................. 234, 240 Phytohormones .......................................................... 19, 222
Microsoft Excel 2003 ......................................................... 62 Pipette cookbook .............................................................. 268
Microsoft Excel 2010 ......................................................... 62 Plant development .............................................................. 81
Min voxels ........................................................................ 105 Plant hormones ......................... 221, 222, 249, 250, 256, 257
Mis-annotated cells .......................................................... 114 Plant material ........................................... 148–150, 156, 214
Misc.software ................................................................... 155 Plant tissue ....................................................... 201, 224, 257
Molecular biology ............................................................... 92 Plasma membrane proteins....................................... 266–267
Molecular pathways .............................................................. 3 PNGase F digestion ................................................. 214–216
MorphoGraphX ................ 100–105, 107–109, 117, 119–121 Posttranslationally modified peptides (PTMPs)........... 20–26
Mounting medium ............................................................. 84 applications .............................................................24–25
MS medium ....................................................................... 85 amino acid substitutions .........................................25
MSBY media ............................................................ 126, 128 biological assay........................................................24
Multiple reaction monitoring (MRM) ............. 241, 250, 254 homologous signaling modules ...............................24
Multi-StageTips ....................................................... 236, 244 modification, peptides.............................................25
caveats.....................................................................25–26
N
contamination .........................................................26
N-acetylglucosamine (GlcNAc) ....................................... 206 nonspecific effects .............................................25–26
NetNGlyc ......................................................................... 209 functional analysis of ....................................................20
N-glycosylation ................. 205–207, 209–211, 213, 216–218 materials .......................................................................21
analysis, peptide effect ............................................21
O reconstitution and dilution, peptide ........................21
O-glycosylation ................................................................ 206 methods ..................................................................21–24
Oligosaccharyltransferase (OST) .................... 206, 211, 213, Arabidopsis ......................................................... 22–23
214, 216, 217 crop plants ..............................................................23
Index
287
direct application ..............................................23–24 Select bad cells.................................................................. 111
dose–response analysis ............................................24 Shoot apical meristems ................................................. 84, 86
initial reconstitution and dilution .....................21–22 Signal sequences ................................................. 19, 111, 112
wheat/tomato/oilseed rape/Brachypodium Signaling peptide ................................................................ 19
seeds..................................................................22 Silique age .......................................................................... 89
Preselect cluster ................................................................ 113 Sliding average ......................................................... 115, 116
Primer design ................................................................... 196 Small white cross .............................................................. 112
Protein complex identification.................................. 148, 156 Smooth passes .................................................................. 105
Protein–protein interaction ..................................... 135, 137, Solid growth medium ......................................................... 83
141–143, 147 Solid-phase extraction (SPE) ................... 223–225, 227, 234
Proteins .............................................................................. 25 Sonication device ...................................................... 199, 202
Proteolytic cleavage ............................................................ 20 Sow seeds............................................................................ 94
Protoplasts .................................232–235, 237–241, 244–247 SPCImage ........................................................................ 140
isolation ..............................................................138–139 Specimen preparation ......................................................... 95
transfection .................................................................139 Spectrum digital camera ....................................................... 3
SPR. See Surface plasmon resonance (SPR)
Q Stable isotope ................................................... 222–224, 226
Quantitative PCR (qPCR) ............................................... 201 StageTip technology ........................................................ 234
Steady-state assay ..................................... 275–276, 279–280
R Steady-state data ...................................................... 279–280
Stereomicroscope ................................................................ 86
Radioactive substance
Streptavidin agarose.......................................... 262, 266, 269
efflux experiments............................................... 264–265
stt3a mutant ...................................................... 211–214, 217
flux experiments..........................................................262
stt3b mutant ...................................................................... 217
import experiments.....................................................265
Surface mesh
Region of interest (ROI) .................................................. 141
create .......................................................................... 105
Replace NaNs software..................................................... 155
trim .............................................................................105
Response units (RU)................................................. 188, 189
Surface plasmon resonance (SPR) ................... 135, 159, 160,
R-language ....................................................................... 274
162, 164, 170, 188
Root traits quantification.............................................. 50, 52
setup ........................................................... 162, 170–174
Roots ................................................................ 237, 238, 245
thermodynamics .........................................................187
Rotary shaker
Synthetic biology .............................................................. 273
Arabidopsis seeds ........................................................... 22
Synthetic complete (SC) .................................................. 274
wheat/tomato/oilseed rape/Brachypodium seeds ............22
Synthetic peptide ................................................................ 27
S Arabidopsis ............................................................... 22–23
crop plants ....................................................................23
Saccharomyces cerevisiae ...................................................... 271 effect on plants .............................................................24
Salicylic acid ..................................................... 221–227, 249 purchasing ....................................................................26
Sample preparation mass spectroscopy ..................... 149–152 usage .............................................................................20
Saturation channel .............................................................. 74 Synthetic reporter ............................................................... 82
Save cell data .................................................................... 114
Seed dormancy ................................................................... 30 T
Seed germination.................................................... 57, 63–67
Tandem mass spectrometry ............................. 148, 149, 152,
assays ............................................................................60
155, 216, 222
color thresholds ............................................................65
TCS. See Two-component signalling (TCS)
experiment
TCSn ...........................................................82, 83, 85, 87, 88
analyze ..............................................................66–67
GFP expression ...................................................... 82, 83,
pictures ...................................................................65
85, 87, 88
preparation..............................................................63
TCSPC ............................................................................ 144
start ...................................................................63–64
Tetracyclic diterpene carboxylic acids ................................. 91
Seed sterilization .................................................... 85, 93–94
Three-segment graft ............................................... 10, 14–15
Seedling growth............................................................ 74–76
Threshold volume..................................................... 110, 114
Seedling preparation ..................................................... 94–95
Threshold wall area .......................................................... 110
Segmentation threshold.................................................... 120
Time course data ...................................................... 274–280
Segmented organ mesh ............................................ 119–120
288 Index
Time-lapse plant phenotyping ......................... 37, 38, 42–45 U

control unit construction ........................................40–41
imaging ...................................................................38–39 UHPLC-MS/MS method ...................................... 233, 234,
microcontroller ................................................. 38, 42–43 236–237, 242
rotating stage ................................................................38 Ultra Fine Micro Knife ...................................................... 16
rotor construction .........................................................39 Ultrahigh-performance liquid chromatography–electrospray
software ........................................................................38 ionization tandem mass spectrometry (UPLC/
stage ........................................................................41–42 ESI-MS/MS) ......................................... 253, 255
tools ..............................................................................39 Ultrahigh-performance liquid chromatography–electrospray
uploading Arduino script ........................................43–44 tandem quadrupole mass spectrometry
ways to use ....................................................................44 (UHPLC-MS/MS) ........................................ 254
Tissue harvesting .............................................................. 238
V
Transcription factors ................................................. 193, 201
Transcriptional fusions ....................................................... 24 Vasculature ............................................................... 110, 114
Transfection protocol........................................................ 138
Transient expression vectors ............................................. 138 W
Transport inhibitor resistant 1 (TIR1) ............. 160, 162, 191 Western blotting ............................................................... 268
Transport proteins ............................................................ 259 Whatman circles........................................................... 11, 12
Treatment folder ............................................................... 116
Triple-quadrupole mass spectrometers ............................. 226 X
T-test software ................................................................. 155
Tunicamycin treatment ..................................................... 217 Xenopus laevis oocytes .............................................. 259, 260,
Two segment shoot-root graft ...................................... 11–14 262–269
Two-component signalling (TCS) ..................................... 82
Y
LUC reporters ..............................................................82
reporter .........................................................................82 Y-Dim ...................................................................... 111, 112
Two-electrode voltage clamp (TEVC) ............................. 260 Yellow fluorescent protein (YFP) ............................. 273, 280
Two-segment self-grafts ............................................... 15, 16 Y-grafts ................................................................... 10, 15, 16
Two-shoot Y-graft .............................................................. 16 Young’s modulus (EA).............................................. 129, 130
Tyrosine sulfation ............................................................... 25 YPDA agar plates ..................................................... 272, 273
View publication stats

2017 Book PlantHormones

Uploaded by

Copyright:

Available Formats

2017 Book PlantHormones

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2017 Book PlantHormones

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Hormone Proﬁling in Plant Tissues

Chapter in Methods in molecular biology (Clifton, N.J.) · November 2017

Maren Müller Sergi Munné-Bosch

SEE PROFILE SEE PROFILE

Oxidative stress as a master regulator of plant trade-offs View project

The user has requested enhancement of the downloaded file.

For further volumes:

Methods and Protocols

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Library of Congress Control Number: 2016955955

© Springer Science+Business Media New York 2017

Printed on acid-free paper

This Humana Press imprint is published by Springer Nature

Vienna, Wien, Austria Jürgen Kleine-Vehn

1 Real-Time Analysis of the Apical Hook Development . . . . . . . . . . . . . . . . . . . 1

14 In Vivo Identification of Plant Protein Complexes Using IP-MS/MS . . . . . . . 147

BIRGIT ABSMANNER • Cell Biology and Plant Biochemistry, University of Regensburg,

HENK W.M. HILHORST • Wageningen Seed Lab, Laboratory of Plant Physiology,

DAJO SMET • Laboratory of Functional Plant Biology, Department of Physiology,

Real-Time Analysis of the Apical Hook Development

To compensate for their sessile lifestyle, plants developed unique

organs, including the apical hook, is differential growth, which,

Despite recent progress in dissecting the regulatory pathways

2 Materials and Chemicals

1. Seeds of Arabidopsis thaliana and mutants of interest.

2.1 Equipment 1. Dark growth chamber to accommodate plate, camera, and IR

or DSLR Remote Pro Multi-Camera software (for one or

The infrared light is fixed in the dark box as indicated at

1. Sterilized seeds are plated on a square Petri dish containing 45 ml

4. The kinetic analysis of the apical hook development is performed

1. During early germination, the seed coat attached to the germi-

embryos are cautiously dissected from seed coats using tweezers.

We thank Herman Höfte, Todor Asenov, Robert Hauschield, and

ethylene-related mutants. Plant Cell 2: 16. Kuhn H, Galston AW (1992) Physiological

Grafting with Arabidopsis thaliana

1. Sterile Arabidopsis seed.

9cm Petri dish

when moving seedlings not to damage the hypocotyl, and root

3.2 Three-Segment 1. Three-segment (or interstock) grafts are set up in a similar

3. After cutting, do not move root segments except for a gentle

3.3 Two-Segment 1. Two-segment self-grafts are essentially the same protocol as

3.4 Two-Shoot 1. Y-grafts involve adding an additional shoot to an intact plant.

1. Using an Ultra Fine Micro Knife is important for success, but

2. Periodically dip the forceps and micro knife in 70 % ethanol and

I thank Elliot Meyerowitz and Raymond Wightman for critical

Tips and Tricks for Exogenous Application of Synthetic

In addition to phytohormones, such as auxin, cytokinin, gibberel-

(e.g., Glycine max RHIZOBIA-INDUCED CLE1 and 2) [5],

2.1 Reconstitution 1. Synthetic peptide (several companies provide these).

2.2 Analysis 1. Seed stocks (wild-type, reporters, and mutants of interest)

3. Reconstitution should always first be attempted in a small vol-

3.2 Sterilization 1. Distribute seeds into microcentrifuge tubes.

3.3 Sterilization of 1. Distribute seeds into 50 ml centrifuge tubes.

5. Add approximately 10–20 surface-sterilized seeds (see

3.6 Inclusion 1. After preparing 100 ml aliquots of ½ MS liquid medium (see

3.7 Direct 1. Dilute peptide to appropriate working concentration—deter-

6. Place in growth chamber until required for treatment.

3.8 Dose–Response 1. Create stocks of peptide at a range of concentrations between

4.1 Determination To determine the effect of a synthetic peptide on plants, it must be