2021 - Book - Phosphoinositides Methods and Protocols

Methods in
Molecular Biology 2251
Roberto J. Botelho Editor
Phosphoino-
sitides
Methods and Protocols
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, UK
For further volumes:

http://www.springer.com/series/7651
For over 35 years, biological scientists have come to rely on the research protocols and
methodologies in the critically acclaimed Methods in Molecular Biology series. The series was
the first to introduce the step-by-step protocols approach that has become the standard in all
biomedical protocol publishing. Each protocol is provided in readily-reproducible step-by-
step fashion, opening with an introductory overview, a list of the materials and reagents
needed to complete the experiment, and followed by a detailed procedure that is supported
with a helpful notes section offering tips and tricks of the trade as well as troubleshooting
advice. These hallmark features were introduced by series editor Dr. John Walker and
constitute the key ingredient in each and every volume of the Methods in Molecular Biology
series. Tested and trusted, comprehensive and reliable, all protocols from the series are
indexed in PubMed.
Phosphoinositides
Methods and Protocols
Edited by
Roberto J. Botelho
Department of Chemistry and Biology and the Graduate Program in Molecular Science, Ryerson
University, Toronto, ON, Canada
Editor
Roberto J. Botelho
Department of Chemistry and Biology
and the Graduate Program in Molecular
Science
Ryerson University
Toronto, ON, Canada
ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-0716-1141-8 ISBN 978-1-0716-1142-5 (eBook)
https://doi.org/10.1007/978-1-0716-1142-5
© Springer Science+Business Media, LLC, part of Springer Nature 2021

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,
expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Humana imprint is published by the registered company Springer Science+Business Media, LLC part of Springer
Nature.
The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.
Preface
As the editor, I am pleased to present the current volume in Methods in Molecular Biology
(MiMB) entitled Phosphoinositides: Methods and Protocols. This book offers 17 chapters
detailing experimental approaches used to investigate phosphoinositide (PtdInsP) regula-
tion and function. One would be hard pressed to overstate the importance of these rare
eukaryotic phospholipids given their broad role in biological processes such as signal
transduction, cell migration and adhesion, cell growth, subcellular organization, and mem-
brane trafficking. These cellular processes ultimately translate into a wide range of physio-
logical functions including development, immunity, cardiovascular function, and
neurobiology. Given these foundational roles in cell and organismal function, it is not
surprising that PtdInsP regulation and function are complex and highly interconnected to
a variety of other processes and pathways. Thus, to dissect PtdInsP regulation and function,
we need a large array of complementary methods.
Based on the phosphorylation pattern of the inositol headgroup, there are seven
PtdInsP species that are interconverted via a variety of lipid kinases, phosphatases, and
phospholipases. The activity of these enzymes enables dynamic synthesis and turnover of
each PtdInsP species within specific membranes or organelles. Thus, we need methods that
can identify and quantify the levels of each PtdInsP species. This can be done biochemically
using metabolic labeling or via mass spectrometry and lipidomics. Lipidomics is also
essential to address a significant challenge in the field, which is that there are more than
seven PtdInsP species when one considers the acyl chains. However, biochemical and mass
spectrometry methods do not readily assess spatiotemporal dynamics of PtdInsPs. We
cannot simply fuse these lipids to green fluorescent proteins (GFP). Instead, the spatiotem-
poral properties of PtdInsPs have been investigated by using protein domains with high
affinity and specificity toward a specific PtdInsP species fused to a fluorescence tag like GFP.
The inception of these tools made it possible to visualize and study PtdInsP distribution and
dynamics in living cells. Yet, this approach also has its own limitations including the
possibility that not all pools of a PtdInsP species are detectable by a specific domain.
Altogether, the combination of complementary biochemical, mass spectrometry, and imag-
ing methods has been instrumental for the detection and quantification of PtdInsP species.
The presence of a PtdInsP species on a membrane bilayer elicits the recruitment of
effector proteins to the host membrane. Typically, effector proteins carry a domain that
exhibits specificity and affinity to that PtdInsP species. This simple model underpins the
common view that PtdInsP regulate organelle identity. To dissect PtdInsP function then,
one needs to manipulate PtdInsP levels. Drug inhibitors, gene deletion, and silencing are
commonly used to disrupt PtdInsP kinases and phosphatases, thus altering the levels of
specific PtdInsPs. However, an elegant approach to acutely disrupt the levels of a PtdInsP
relies on induced dimerization methods. Here, a phosphatase domain that exhibits catalytic
specificity toward a PtdInsP can be acutely targeted to a given membrane via chemical-
induced dimerization. This approach is reversible and allows for subcellular disruption of a
PtdInsP pool. Finally, to understand PtdInsP function, one must identify PtdInsP effectors
and measure their activity. This can be done via different methods including affinity precipi-
tation or co-sedimentation with liposomes, protein insertion within lipid bilayers, and
v
vi Preface
enzymatic assays, or through emerging methods like native mass spectrometry and
microfluidics.
Overall, this volume provides detailed methodology on a variety of complementary
methods that have been instrumental in dissecting the regulation, dynamics, and function of
PtdInsPs. Of significance, both specialist and novice workers will find detailed tips in the
form of “Notes,” which reflect the collective wisdom of the method experts that are so often
essential for experimental reproducibility. On behalf of the contributing authors, I hope
readers of Phosphoinositides: Methods and Protocols find this volume to be an indispensable
resource for their research.
Toronto, ON, Canada Roberto J. Botelho

Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1 Simultaneous Detection of Phosphoinositide Lipids by Radioactive

Metabolic Labeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Noah Steinfeld, Sai Srinivas Panapakkam Giridharan,
Emily J. Kauffman, and Lois S. Weisman
2 Label-Free Quantification of Phosphoinositides in Drosophila
by Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Avishek Ghosh and Padinjat Raghu
3 Profiling of Phosphoinositide Molecular Species in Resting
or Activated Human or Mouse Platelets by a LC-MS Method . . . . . . . . . . . . . . . . 39
Gaëtan Chicanne, Justine Bertrand-Michel, Julien Viaud,
Karim Hnia, Jonathan Clark, and Bernard Payrastre
4 Quantification of Genetically Encoded Lipid Biosensors . . . . . . . . . . . . . . . . . . . . . 55
Rachel C. Wills, Jonathan Pacheco, and Gerald R. V. Hammond
5 Detection of Plasma Membrane Phosphoinositide Dynamics
Using Genetically Encoded Fluorescent Protein Probes. . . . . . . . . . . . . . . . . . . . . . 73
Rebecca Cabral-Dias, Yasmin Awadeh, Roberto J. Botelho,
and Costin N. Antonescu
6 Imaging the Nanoscale Distribution of Phosphoinositides in the Cell
Plasma Membrane with Single-Molecule Localization Super-Resolution
Microscopy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Fan Fan, Chen Ji, and Xuelin Lou
7 Induced Dimerization Tools to Deplete Specific Phosphatidylinositol
Phosphates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Jonathan Pacheco, Rachel C. Wills, and Gerald R. V. Hammond
8 A Real-Time Phosphatidylinositol 4-Phosphate 5-Kinase Assay
Using Fluorescence Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Taki Nishimura
9 Assessing In Situ Phosphoinositide–Protein Interactions Through
Fluorescence Proximity Ligation Assay in Cultured Cells. . . . . . . . . . . . . . . . . . . . . 133
Mo Chen, Hudson T. Horn, Tianmu Wen, Vincent L. Cryns,
and Richard A. Anderson
10 Characterization of Protein–Phospholipid/Membrane Interactions
Using a “Membrane-on-a-Chip” Microfluidic System . . . . . . . . . . . . . . . . . . . . . . . 143
Calvin Yeager, Djoshkun Shengjuler, Simou Sun, Paul S. Cremer,
and Craig E. Cameron
11 Exploring Phosphoinositide Binding Using Native Mass Spectrometry . . . . . . . . 157
Julian Bender and Carla Schmidt
vii
viii Contents
12 Liposome-Based Methods to Study Protein–Phosphoinositide

Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Mélanie Mansat, Mélanie Picot, Gaëtan Chicanne, Virginie Nahoum,
Frédérique Gaits-Iacovoni, Bernard Payrastre, Karim Hnia,
and Julien Viaud
13 Liposome-Based Methods to Study GTPase Activation
by Phosphoinositides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Julien Viaud, Laurie Ceccato, Bernard Payrastre,
and Frédérique Gaits-Iacovoni
14 Liposome Co-sedimentation and Co-flotation Assays to Study
Lipid–Protein Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Yosuke Senju, Pekka Lappalainen, and Hongxia Zhao
15 Characterization of PROPPIN–Phosphoinositide Binding
by Stopped-Flow Fluorescence Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Ángel Pérez-Lara and Reinhard Jahn
16 Fluorescence Assays to Study Membrane Penetration of Proteins . . . . . . . . . . . . . 215
Yosuke Senju and Hongxia Zhao
17 Fluorogenic XY-69 in Lipid Vesicles for Measuring Activity
of Phospholipase C Isozymes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
Adam J. Carr, Edhriz Siraliev-Perez, Weigang Huang,
John Sondek, and Qisheng Zhang
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Contributors
RICHARD A. ANDERSON • University of Wisconsin-Madison, School of Medicine and Public

Health, Madison, WI, USA
COSTIN N. ANTONESCU • Department of Chemistry and Biology and the Graduate Program
in Molecular Science, Ryerson University, Toronto, ON, Canada
YASMIN AWADEH • Department of Chemistry and Biology and the Graduate Program in
Molecular Science, Ryerson University, Toronto, ON, Canada
JULIAN BENDER • Interdisciplinary research center HALOmem, Institute for Biochemistry
and Biotechnology, Charles Tanford Protein Centre, Martin Luther University Halle-
Wittenberg, Halle, Germany
JUSTINE BERTRAND-MICHEL • Institut des Maladies Métaboliques et Cardiovasculaires,
Inserm U1048, Université Toulouse III, Toulouse, France; MetaToul-Lipidomic Facility,
MetaboHUB, Université Toulouse III, Toulouse, France
ROBERTO J. BOTELHO • Department of Chemistry and Biology and the Graduate Program in
Molecular Science, Ryerson University, Toronto, ON, Canada
REBECCA CABRAL-DIAS • Department of Chemistry and Biology and the Graduate Program
in Molecular Science, Ryerson University, Toronto, ON, Canada
CRAIG E. CAMERON • Department of Microbiology and Immunology, University of North
Carolina School of Medicine, Chapel Hill, NC, USA
ADAM J. CARR • Division of Chemical Biology and Medicinal Chemistry, University of North
Carolina at Chapel Hill, Chapel Hill, NC, USA
LAURIE CECCATO • INSERM U1048 and Université Toulouse 3, Toulouse Cedex, France
MO CHEN • University of Wisconsin-Madison, School of Medicine and Public Health,
Madison, WI, USA
GAËTAN CHICANNE • Institut des Maladies Métaboliques et Cardiovasculaires, Inserm
U1048, Université Toulouse III, Toulouse, France
JONATHAN CLARK • Signalling Programme, Babraham Institute, Cambridge, UK
PAUL S. CREMER • Department of Chemistry, The Pennsylvania State University, University
Park, PA, USA
VINCENT L. CRYNS • Department of Medicine, University of Wisconsin Carbone Cancer
Center, University of Wisconsin-Madison, School of Medicine and Public Health, Madison,
WI, USA
FAN FAN • Department of Cell Biology, Neurobiology and Anatomy, Medical College of
Wisconsin, Milwaukee, WI, USA
FRÉDÉRIQUE GAITS-IACOVONI • INSERM U1048 and Université Toulouse 3, Toulouse Cedex,
France
AVISHEK GHOSH • National Centre for Biological Sciences-TIFR, GKVK Campus,
Bangalore, India
SAI SRINIVAS PANAPAKKAM GIRIDHARAN • Life Sciences Institute, University of Michigan,
Ann Arbor, MI, USA
GERALD R. V. HAMMOND • Department of Cell Biology, University of Pittsburgh School of
Medicine, Pittsburgh, PA, USA
KARIM HNIA • Institut des Maladies Métaboliques et Cardiovasculaires, Inserm U1048,
Université Toulouse III, Toulouse, France
ix
x Contributors
HUDSON T. HORN • University of Wisconsin-Madison, School of Medicine and Public Health,

Madison, WI, USA
WEIGANG HUANG • Division of Chemical Biology and Medicinal Chemistry, University of
North Carolina at Chapel Hill, Chapel Hill, NC, USA
REINHARD JAHN • Department of Neurobiology, Max-Planck-Institute for Biophysical
Chemistry, Göttingen, Germany
CHEN JI • Department of Cell Biology, Neurobiology and Anatomy, Medical College of
Wisconsin, Milwaukee, WI, USA; National Institute of Health, Bethesda, MD, USA
EMILY J. KAUFFMAN • Life Sciences Institute, University of Michigan, Ann Arbor, MI, USA
PEKKA LAPPALAINEN • Institute of Biotechnology, University of Helsinki, Helsinki, Finland
XUELIN LOU • Department of Cell Biology, Neurobiology and Anatomy, Medical College of
Wisconsin, Milwaukee, WI, USA
MÉLANIE MANSAT • INSERM U1048 and Université Toulouse 3, Toulouse Cedex, France
VIRGINIE NAHOUM • CNRS, Institut de Pharmacologie et de Biologie Structurale (IPBS),
Toulouse, France; Université de Toulouse, UPS (Université Paul Sabatier), IPBS,
Toulouse, France
TAKI NISHIMURA • MRC Laboratory for Molecular Cell Biology, University College London,
London, UK; Molecular Cell Biology of Autophagy, The Francis Crick Institute,
London, UK
JONATHAN PACHECO • Department of Cell Biology, University of Pittsburgh School of
Medicine, Pittsburgh, PA, USA
BERNARD PAYRASTRE • Institut des Maladies Métaboliques et Cardiovasculaires, Inserm
U1048, Université Toulouse III, Toulouse, France; CHU (Centre Hospitalier
Universitaire) de Toulouse, Laboratoire d’Hématologie, Toulouse Cedex, France
ÁNGEL PÉREZ-LARA • Department of Physical Chemistry, Faculty of Pharmacy, University of
Granada, Granada, Spain
MÉLANIE PICOT • INSERM U1048 and Université Toulouse 3, Toulouse Cedex, France
PADINJAT RAGHU • National Centre for Biological Sciences-TIFR, GKVK Campus,
Bangalore, India
CARLA SCHMIDT • Interdisciplinary research center HALOmem, Institute for Biochemistry
and Biotechnology, Charles Tanford Protein Centre, Martin Luther University Halle-
Wittenberg, Halle, Germany
YOSUKE SENJU • Research Institute for Interdisciplinary Science (RIIS), Okayama
University, Okayama, Japan
DJOSHKUN SHENGJULER • Viral Populations and Pathogenesis Unit, Institut Pasteur,
Paris, France
EDHRIZ SIRALIEV-PEREZ • Department of Biochemistry and Biophysics, University of North
Carolina at Chapel Hill, Chapel Hill, NC, USA
JOHN SONDEK • Department of Biochemistry and Biophysics, University of North Carolina at
Chapel Hill, Chapel Hill, NC, USA; Department of Pharmacology, University of North
Carolina at Chapel Hill, Chapel Hill, NC, USA; The Lineberger Comprehensive Cancer
Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
NOAH STEINFELD • Life Sciences Institute, University of Michigan, Ann Arbor, MI, USA;
Program in Cellular and Molecular Biology, University of Michigan, Ann Arbor, MI, USA
SIMOU SUN • Department of Chemistry, The Pennsylvania State University, University Park,
PA, USA
JULIEN VIAUD • Institut des Maladies Métaboliques et Cardiovasculaires, Inserm U1048,
Université Toulouse III, Toulouse, France
Contributors xi
LOIS S. WEISMAN • Life Sciences Institute, University of Michigan, Ann Arbor, MI, USA;
Department of Cell and Developmental Biology, University of Michigan, Ann Arbor,
MI, USA
TIANMU WEN • University of Wisconsin-Madison, School of Medicine and Public Health,
Madison, WI, USA
RACHEL C. WILLS • Department of Cell Biology, University of Pittsburgh School of Medicine,
Pittsburgh, PA, USA
CALVIN YEAGER • Department of Microbiology and Immunology, University of North
Carolina School of Medicine, Chapel Hill, NC, USA
QISHENG ZHANG • Division of Chemical Biology and Medicinal Chemistry, University of
North Carolina at Chapel Hill, Chapel Hill, NC, USA; Department of Pharmacology,
University of North Carolina at Chapel Hill, Chapel Hill, NC, USA; The Lineberger
Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill,
NC, USA
HONGXIA ZHAO • Faculty of Biological and Environmental Sciences, University of Helsinki,
Helsinki, Finland; School of Life Sciences, Guangxi Normal University, Guangxi, China;
Key Laboratory of Stem Cell and Biopharmaceutical Technology, Guangxi Universities,
Guangxi, China
Chapter 1
Simultaneous Detection of Phosphoinositide Lipids

by Radioactive Metabolic Labeling
Noah Steinfeld, Sai Srinivas Panapakkam Giridharan,
Emily J. Kauffman, and Lois S. Weisman
Abstract
Phosphoinositide (PPI) lipids are a crucial class of low-abundance signaling molecules that regulate many
processes within cells. Methods that enable simultaneous detection of all PPI lipid species provide a
wholistic snapshot of the PPI profile of cells, which is critical for probing PPI biology. Here we describe a
method for the simultaneous measurement of cellular PPI levels by metabolically labeling yeast or mam-
malian cells with myo-3H-inositol, extracting radiolabeled glycerophosphoinositides, and separating lipid
species on an anion exchange column via HPLC.
Key words Phosphatidylinositol, Phosphoinositide lipids, PtdIns, PI3P, PI4P, PI5P, PI3,5P2,
PI4,5P2, PI3,4P2, PI3,4,5P3, Radioactive metabolic labeling
1 Introduction
Phosphorylated phosphatidylinositol signaling lipids (PPI, also

known as phosphoinositides) are a low-abundance class of mole-
cules that control a variety of signal transduction pathways and play
crucial roles in cellular homeostasis. These phosphoinositides are
generated by phosphorylation of phosphatidylinositol (PI) on the
3, 4, or 5 position of its inositol ring in every combination to yield a
total of seven PPI species. All seven species can be detected in
mammalian cell types, whereas only four are readily detectable in
Saccharomyces cerevisiae. The levels of these signaling lipids are
dynamically regulated in cells. For example, a dynamic change is
observed during hyperosmotic shock in S. cerevisiae, which leads to
a transient 20-fold elevation in phosphatidylinositol
3,5-bisphosphate (PI3,5P2) [1]. The changes in PPIs are con-
trolled spatially by the specific localization of lipid kinases and
Noah Steinfeld and Sai Srinivas Panapakkam Giridharan contributed equally to this work.
Roberto J. Botelho (ed.), Phosphoinositides: Methods and Protocols, Methods in Molecular Biology, vol. 2251,
https://doi.org/10.1007/978-1-0716-1142-5_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021
1
2 Noah Steinfeld et al.
phosphatases, and temporally via specific activation of these

enzymes (Fig. 1) [2–6].
PPI lipids regulate various signaling pathways, and various
signaling pathways in turn regulate these lipids. Thus, it is impor-
tant to measure the levels of these lipids. One established method is
to use protein-based probes that bind to a specific PPI species. A
major advantage of these PPI bioprobes is that they reveal the
subcellular localization of the PPI lipid, unlike the metabolic label-
ing approach described here. However, there are not highly specific
bioprobes for every PPI. In addition, there is a concern that an
efficient bioprobe would mask the phosphoinositide head group
and potentially compete with endogenous PPI effectors. Another
concern is that some bioprobes may require both a PPI lipid and
specific protein or cellular context. For example, FYVE domains,
protein domains that specifically bind PI3P, from Fab1 and EEA1
exhibit localizations dissimilar from each other [7].
PPIs exist in a substrate/product relationship with other PPIs,
and thus the levels of selected PPI species are linked. For example,
during hyperosmotic shock, PI3,5P2 sharply increases by 20-fold
[1]. This change in PI3,5P2 also results in a concomitant decrease
in the levels of its precursor PI3P. For this reason, changes in a cell’s
behavior are likely due to changes in more than one PPI. Moreover,
some PPI kinases and phosphatases act on multiple PPI substrates
[8]. Thus, drug treatments or knockouts that affect one PPI kinase
or phosphatase have the potential to affect the levels of multiple
PPIs. Therefore, the ability to simultaneously measure all PPI
species within cells provides a critical snapshot of how specific
stimuli or treatments impact each PPI lipid. Yet, accurately measur-
ing the levels of these PPIs simultaneously has proven difficult.
PPI lipids were first identified from a phosphatide fraction
known as cephalin, defined by their solubility in ether and insolu-
bility in ethyl alcohol [9, 10]. These were first separated from
phosphatidylethanolamine and phosphatidylserine using a method
that took advantage of the lower solubility of PPIs in methanol.
Since then, optimization of protocols to extract and detect PPI
species has culminated in the current method of metabolic labeling
and simultaneous detection of all PPIs by separation on an anion
exchange column.
The method described in this chapter starts with metabolic
labeling with myo-3H-inositol. Ortho(32P)-phosphate can be used
instead [11]. While use of ortho(32P)-phosphate greatly increases
sensitivity to specific molecules, in comparison with myo-3H-inosi-
tol, there is a much larger cohort of phosphate-containing species
that co-purify with PPI. Following metabolic labeling, cells are
lysed by precipitating all cellular macromolecules using perchloric
acid [12, 13]. The precipitated material, including phospholipids, is
deacylated by mild-base hydrolysis using mono-methylamine
(Fig. 2) [14]. The deacylation reaction generates the corresponding
Simultaneous Detection of Phosphoinositide Lipids 3
Fig. 1 Interconversion of PPIs by PPI kinases and phosphatases. PPIs are gener-
ated by the action of phosphoinositide kinases and phosphatases, and these
reactions are confined to membrane subdomains where these phosphoinositide-
modifying enzymes reside (reviewed in [2–6]). These reactions yield mono-
phosphates (phosphatidylinositol 3-phosphate (PI3P), phosphatidylinositol
4-phosphate (PI4P), phosphatidylinositol 5-phosphate (PI5P)), bisphosphates
(phosphatidylinositol 3,4-bisphosphate (PI3,4P2), phosphatidylinositol
3,5-bisphosphate (PI3,5P2), phosphatidylinositol 4,5-bisphosphate (PI4,5P2)), and
trisphosphate (phosphatidylinositol 3,4,5-trisphosphate (PI3,4,5P3)). PI can be
phosphorylated by the class-III PI3-kinase Vps34 to PI3P, by type-II and type-III
PI4-kinases to PI4P, or by PIKfyve to PI5P. PI3P can be phosphorylated by PIKfyve
to PI3,5P2 or dephosphorylated by myotubularin-family proteins (MTMRs) to
PI. PI3,5P2 can be dephosphorylated to PI5P by MTMRs. PI3,5P2 may also be
dephosphorylated to PI3P by Fig4. PI4P can be phosphorylated to PI4,5P2 by PIP5-
kinases. Furthermore, PI4,5P2 can be phosphorylated by Class-I PI3-kinases to
PI3,4,5P3. PI3,4,5P3 can be dephosphorylated to PI3,4P2 by phosphoinositide-5-
phosphatases or dephosphorylated to PI4,5P2 by PTEN. PI4,5P2 can be depho-
sphorylated to PI4P by phosphoinositide-5-phosphatases (INPP5B, INPP5J,
INPP5E, Synaptojanin, OCRL). PI4P can be phosphorylated to PI3,4P2 by Class-II
PI3-kinases or dephosphorylated to PI by SAC1. PI3,4P2 can be dephosphorylated
to PI3P by INPP4 proteins. PI5P can be phosphorylated to PI4,5P2 by PIP4-kinases.
Red circles represent phosphate groups. Hydroxyl groups on the inositol ring as
well as the ester linkages for the fatty acyl chains are not indicated
Fig. 2 The deacylation reaction of PPIs to glycerophosphoinositides. Mild-base

hydrolysis via methylamine results in the cleavage of the fatty acid chains from
the glycero-3-phospho-1-inositol backbone. The resultant phosphorylated glyc-
erol phosphoinositides can be separated on an anion exchange column. *
denotes phosphorylation sites
phosphorylated glycero-3-phospho-1-inositol and fatty acid

amides. Following deacylation, methylamine is evaporated under
vacuum. The phosphorylated glycero-inositol head groups are
solubilized in water. Phosphorylated glycero-inositols are further
isolated from fatty acid amides and other contaminants by phase
separation with butanol/ethyl ether/formic acid. The phosphory-
lated glycero-inositol fraction is present in the aqueous phase, while
fatty acid amides are in the organic phase [14]. This extraction
method is distinct from extraction from membranes using acidified
chloroform/methanol/HCl [15, 16]. In our hands, using the
perchloric acid precipitation protocol described here results in
detection of higher quantities of PPIs [13]. The resultant phos-
phorylated glycero-inositol fraction is then separated on an anion
exchange column and detected with an inline flow-through scintilla-
tion counter. Anion exchange on an HPLC using ammonium phos-
phate dibasic was first used to identify PI3P in fibroblasts
[17]. The elution gradient described here for yeast was developed to
separate the four PPI species in S. cerevisiae [18, 19]. A few different
gradients are used to separate PPIs in mammalian cells [16]. In
addition, either one or two anion exchange columns are used to
separate phosphorylated glycero-inositol species [20, 21]. Better
resolution between PI4P and PI5P peaks is achieved when two
anion exchange columns are used in tandem, although utilizing
two-columns is more expensive and requires more sample volume.
New protocols are being developed to measure phospholipids
using high-performance liquid chromatography–mass spectrome-
try [22]. While this new approach can determine the acyl chain
structure of PPIs, this method fails to distinguish among mono-
phosphorylated PPIs (PI3P vs. PI4P vs. PI5P) and among
bis-phosphorylated PPIs (PI3,5P2 vs. PI4,5P2 vs. PI3,4P2).
The method below describes the metabolic labeling and mea-
surement of PPI lipids in S. cerevisiae and some types of mammalian
cells. This protocol has been modified for use in other organisms
[23, 24]. This protocol is a culmination of over 60 years of chemis-
try and biology research in multiple laboratories and provides a
comprehensive snapshot of the PPI lipids in cells.
2 Materials
2.1 General System 1. Tabletop microcentrifuge.

Requirements 2. Benchtop centrifuge.
3. Mini-beadbeater (BioSpec Products) (yeast protocol only).
4. Samco Scientific 233 disposable plastic transfer pipettes with a
fine tip (yeast protocol only).
5. Heating block.
6. SpeedVac concentrator.
7. HPLC and flow scintillation system (see Note 1).
8. 4.6 250 mm SAX cartridge anion exchange column and
anion guard cartridge (Macmod Analytical) (see Notes 2–4).
Other anion exchange columns may be appropriate.
9. Institutional safety approval (see Note 5).
2.2 Materials 1. myo-inositol,-[2-3H] (PerkinElmer) dissolved in ddH2O with

Common to All proprietary stabilizer PT6-271.
Protocols 2. 4.5% perchloric acid diluted from a 70% stock in ddH2O
(see Note 6). Made fresh for each experiment.
3. 100 mM EDTA, pH 8.0.
4. Methylamine reagent: Mix methylamine (Alfa Aesar (see
Note 7)), methanol, n-butanol, and ddH2O to a final
concentration of 10.7% methylamine, 45.7% methanol, and

11.4% n-butanol. Made fresh for each experiment.
5. Butanol/ethyl ether/formic acid: Mix n-butanol, ethyl ether,
and formic acid in a 20:4:1 ratio. Made fresh for each
experiment.
6. 1 M (NH4)2HPO4, pH 3.8 (adjust the pH with o-
phosphoric acid).
7. Liquid scintillation fluid FlowLogic HS (Lab Logic) for high
salt samples is used in our lab. Other scintillation liquids may be
appropriate.
2.3 Materials 1. 10 Mg2+, Ca2+, NaCl salts: Prepare 41.5 mM magnesium
Specific sulfate, 9.0 mM calcium chloride, and 17.1 mM sodium chlo-
for Saccharomyces ride in 500 mL ddH2O and autoclave.
cerevisiae Protocol 2. 10 potassium salts: Prepare 66.1 mM potassium phosphate
monobasic and 8.6 mM potassium phosphate dibasic in
ddH2O and autoclave.
3. 10 amino acid stock solution: For 500 mL, use the following
mixture of amino acids: 0.1 g L-adenine, 0.1 g L-uracil*, 0.1 g
L-histidine-HCl*, 0.1 g L-arginine-HCl, 0.1 g L-methionine,
0.15 g L-tyrosine, 0.15 g L-leucine*, 0.15 g L-isoleucine,
0.15 g L-lysine-HCl, 0.25 g L-phenylalanine, 0.5 g L-glutamic
acid, 0.5 g L-aspartic acid, 0.750 g L-valine, 2.0 g L-serine,
0.15 g L-cysteine, and 0.15 g L-proline. Amino acids are dis-
solved in 500 mL ddH2O and filter sterilized. To create selec-
tive media, the amino acids that are indicated in the recipe with
* can be omitted.
4. 100 trace elements: Mix 25 mg boric acid, 2 mg cupric
sulfate, 5 mg potassium iodide, 10 mg ferrous sulfate, 20 mg
manganese sulfate, 10 mg sodium molybdate, and 20 mg zinc
sulfate in 500 mL ddH2O and autoclave.
5. 10 ammonium sulfate: Prepare 378 mM ammonium sulfate
in ddH2O and filter sterilize.
6. 40% glucose: Prepare 40% glucose in ddH2O and autoclave.
7. 50 Trp/Thr solution: Mix 10 g threonine and 1 g tryptophan
in 500 mL ddH2O and filter sterilize. Store at 4 C.
8. 100 vitamins: Mix 1 mg D-biotin, 100 mg D-pantothenic
acid-calcium salt, 0.1 mg folic acid, 20 mg niacinamide,
10 mg p-aminobenzoic acid, 20 mg pyridoxine hydrochloride,
10 mg riboflavin, and 20 mg thiamine hydrochloride in
500 mL ddH2O. Filter sterilize and store at 4 C.
9. Inositol-free growth media: For 500 mL, mix stock solutions:
50 mL 10 Mg2+, Ca2+, NaCl salts solution, 50 mL 10
potassium salts, 50 mL 10 amino acid stock solution, 5 mL
100 trace elements solution, 50 mL 10 ammonium sulfate

solution, 25 mL 40% glucose solution, 10 mL 50 Trp/Thr
solution, 5 mL 100 vitamins solution, and 255 mL ddH2O.-
Filter sterilize and store at 4 C.
10. 1.8 M NaCl solution (use to induce hyperosmotic shock).
2.4 Materials 1. Custom Dulbecco’s modified Eagle’s medium (DMEM):

Specific Custom-made DMEM as the base medium and modified to
for Mammalian Cells omit myo-inositol and sodium pyruvate.
2. Custom Neurobasal-A medium: Custom-made Neurobasal-A
as the base medium and modified to omit L-aspartic acid, L-
glutamic acid, myo-inositol, and L-glutamine.
3. Phosphate-buffered saline (PBS), pH 7.4.
4. Dialyzed fetal bovine serum.
5. 1 M HEPES, pH: 7.2 to 7.5.
6. Apo-transferrin: Freshly prepare 4 mg/mL of apo-transferrin
and filter sterilize before adding to the media.
7. Insulin.
8. Penicillin–streptomycin–glutamine.
9. B27.
10. L-glutamine.
11. DMEM–inositol labeling medium: inositol-free DMEM with
10 μCi/mL of myo-3H-inositol, 10% dialyzed fetal bovine
serum, 20 mM HEPES, 5 μg/mL transferrin, 5 μg/mL insu-
lin, and 1 penicillin/streptomycin/glutamine.
12. Neurobasal-A–inositol labeling medium: custom Neurobasal-
A medium with 50 μCi/mL myo-3H-inositol, B27, and L-
glutamate.
13. Plate the cells and prepare the inositol medium as indicated:
MEFcells: Grow MEF cells on 10-cm dishes to 50–60% con-
fluency. Use one dish per sample and prepare 6 mL of
DMEM–inositol labeling medium per dish.
Macrophage: Grow RAW macrophage cells on 35-mm dishes to
20–30% confluency. Use two dishes per sample and prepare
2 mL of DMEM–inositol labeling medium per dish.
Primary neuron culture: Cultured primary rat hippocampal
neurons are grown for 21 days on 35-mm dishes. Plate
600,000 cells per 35-mm dish and use two dishes per
sample and prepare 2 mL of Neurobasal-A–inositol label-
ing medium per dish.
3 Methods
While the details of the yeast and mammalian protocols differ, both
protocols follow the same general steps: metabolic labeling cells
with myo-3H-inositol, harvesting cells followed by addition of per-
chloric acid to precipitate all macromolecules, mild-base hydrolysis
to release and extract phosphorylated glycero-inositol head groups
from the PPI lipids, and separation by anion exchange chromatog-
raphy via HPLC.
3.1 Yeast Protocol 1. Grow a 10-mL starter culture in appropriate media in a shaker
at 24 C, at 200 rpm. Cultures need to reach mid-log phase
3.1.1 Preparing Cells
(5 106 cells/mL) before they are inoculated into myo-3H-i-
for myo-3H-Inositol
nositol-containing media.
Labeling
2. Pellet cells at room temperature at 1700 g for 3 min. Wash
the cell pellet in 4 mL of inositol-free media and pellet again.
3. Inoculate 50-mL disposable centrifuge tubes with an appropri-
ate amount of each yeast strain and grow at 24 C, at 200 rpm.
In addition, to track the growth of the 3H-containing culture,
grow a parallel culture in another 50-mL disposable centrifuge
tube containing 5 mL of inositol-free media without myo-3H-i-
nositol. Enough cells of the starter culture should be added so
that cultures reach mid-log phase (5 106 cells/mL) 16 h
following inoculation (see Note 8).
4. Following inoculation, add 50 μL of 1 μCi/μL myo-3H-inositol
to the appropriate cultures. Loosely tape the lid of the 50-mL
disposable centrifuge tube to allow for airflow. Lids are
required to prevent aerosol contamination. Allow cells to
grow for 16 h in myo-3H-inositol at 24 C, at 200 rpm.
3.1.2 Cell Harvesting 1. Prepare a 2.0-mL screw cap tube for each sample by filling half
and Lysis of it with 0.5-mm zirconia beads and adding 470 μL of 4.5%
perchloric acid. Chill the screw cap tubes and the remaining
4.5% perchloric acid on ice.
2. At the 16-h time point, measure the OD600 of the parallel
cultures that were grown in inositol-free media without
myo-3H-inositol.
3. Pellet the cells incubated with myo-3H-inositol at 1700 g for
3 min at room temperature.
4. Resuspend the cell pellet in 4 mL of inositol-free media and
pellet the cells at 1700 g for 3 min.
5. Resuspend the cell pellet in 100 μL of inositol-free media and
transfer to a microcentrifuge tube.
6. To perform hyperosmotic shock, add 100 μL of 1.8 M NaCl
solution to the samples. Incubate cells for the appropriate time
interval. For control cells not undergoing hyperosmotic shock,

add 100 μL of inositol-free media.
7. Add 800 μL of ice cold 4.5% perchloric acid and incubate on ice
until all samples are ready.
8. Transfer the resultant perchloric acid mixture to the prepared
2-mL screw cap tubes containing zirconia beads.
9. Lyse the cells by placing each tube into a mini-beadbeater and
shaking for 2 min at high speed. Immediately place the cells on
ice for 2 min. Repeat this process two more times for a total
lysis time of 6 min.
10. Using fine tip disposable plastic transfer pipettes, transfer the
lysed cells to new microcentrifuge tubes. Make sure to avoid
the beads. Centrifuge the lysate at 16,000 g for 10 min at
room temperature.
3.1.3 Deacylation of PPI 1. Resuspend each pellet with 1 mL of 100 mM EDTA. Ensure
Lipids that all the cell precipitate is resuspended, including precipitate
on the side of the microcentrifuge tube.
2. Pellet the cells at 16,000 g for 10 min at room temperature.
3. Resuspend the pellet in 50 μL of ddH2O and transfer to a 5-mL
glass vial. It is critical to use glass vials with a rubber-lined cap
that can make a tight seal.
4. Wash the microcentrifuge tube with 1 mL of methylamine
reagent and add it to the appropriate 5-mL glass vial. Heat
the samples at 55 C for 1 h.
5. Allow the tubes to cool for 5 min and transfer to microcentri-
fuge tubes. Dry the samples in a SpeedVac concentrator at
55 C for 4 h under vacuum.
6. Dried samples can be stored at 20 C until the next day. If
samples have been stored at 20 C, thaw the samples on the
bench for 10 min before the next step.
3.1.4 Butanol/Ethyl 1. Resuspend each sample with 300 μL of ddH2O and leave at
Ether/Formic Acid room temperature for 30 min.
Extraction 2. Mix each sample again and vortex for about 30 s. Centrifuge at
16,000 g for 2 min at room temperature.
3. Transfer the supernatant to a new microcentrifuge tube and
add 300 μL of butanol/ethyl ether/formic acid solution. Vor-
tex each tube for 15 s and centrifuge the samples at 16,000 g
for 2 min at room temperature.
4. Transfer the bottom layer (aqueous phase) to a new tube. Add
another 300 μL butanol/ethyl ether/formic acid solution.
Repeat vortex and centrifugation steps as above.
5. Transfer the bottom layer (aqueous phase) to a new tube.
6. Dry the sample in a SpeedVac concentrator at 55 C for 4 h

under vacuum.
7. Samples can be stored at 20 C until loading on HPLC, or
the protocol can be continued. Samples should be stable at
20 C for several months.
3.1.5 HPLC Sample 1. Thaw the samples at room temperature for about 5 min and
Preparation and Anion resuspend the sample in 60 μL of ddH2O. Incubate the sample
Exchange at room temperature for 5 min. Centrifuge the sample at
16,000 g for 5 min to pellet any particulate matter.
2. Load 55 μL into the autosampler vial inserts, avoiding any
particulate matter at the bottom of the tube (see Note 9).
3. Inject 50 μL of sample into the HPLC and separate on a SAX
anion exchange column. Buffer A (ddH2O) and Buffer B (1 M
(NH4)2HPO4, pH 3.8) are used to generate the following
gradients run at a 1-mL/min flow rate: 1% Buffer B for
5 min, 1–20% Buffer B for 44 min, 20–50% Buffer B for
3.75 min, and 50% Buffer B for 8 min. At the end of the run,
re-equilibrate the column with ddH2O at 1 mL/min for
15 min. Radiolabeled eluate can be detected by an inline flow
scintillation analyzer. A 1:2 ratio of eluate to scintillant is used
with a 3-mL/min flow rate. Fractions are analyzed by binning
counts every 6 s. The data are analyzed by Laura-4 software.
4. To quantify scintillation counts from each sample, the raw
counts in each peak are expressed as a percentage of total
phosphatidylinositol-related species, calculated from summa-
tion of the counts of the five glycero-inositol peaks present in
yeast (PI, PI3P, PI4P, PI3,5P2, and PI4,5P2). Background
scintillation counts are calculated from adjacent regions and
subtracted from all peaks. Figure 3 illustrates a typical yeast
elution profile. The average PPI levels in the SEY6210 yeast
strain background are presented in Table 1. It should be noted
that PPI levels may vary between different wild-type yeast
strains.
3.2 Mammalian Cell 1. Grow cells to appropriate confluence, which needs to be deter-
Protocol mined empirically. All cells are grown at 37 C and 5% CO2.
3.2.1 Preparing Cells 2. For all cells described except primary neurons, rinse twice with
for myo-3H-Inositol PBS and incubate for 48 h with DMEM–inositol labeling
Labeling medium. For primary neurons, rinse with inositol-free Neuro-
basal-A medium and incubate with Neurobasal-A–inositol
labeling medium for 24 h.
3. To test the effects of small molecule inhibitors or growth
factors on PPI levels, cells are treated for an appropriate time
during labeling such that the total metabolic labeling time is
Fig. 3 Example of anion exchange chromatography of a yeast sample. Peaks were truncated at 500 cpm to
display all the PPI detected
Table 1
Levels of PPIs in yeast determined by myo-3H-inositol labeling
Yeast
PI 93.45 0.45
PI3P 2.66 0.17
PI4P 2.07 0.16
PI5P Not detectable
PI3,5P2 0.12 0.01
PI3,4P2 Not detectable
PI4,5P2 1.7 0.13
PI3,4,5P3 Not detectable
PPI levels were measured in seven wild-type yeast samples (SEY 6210)
48 h for MEF and macrophages and 24 h for primary cortical

neurons.
4. To test the effect of depleting proteins on PPI levels, cells are
infected with lentivirus expressing shRNA to the protein of
interest. One day after infection, cells are selected with appro-
priate antibiotics. After 2 days of selection, cells are split and
plated with an appropriate amount of cells to achieve 50–60%
confluency 24 h later. MEF cells are then washed and incubated

with inositol labeling medium for 48 h. shRNA depletion
followed by metabolic labeling has not yet been tested in
macrophages or primary neurons.
3.2.2 Cell Harvesting 1. Discard growth media and rinse cells two times at room tem-
and Lysis perature with 5 mL of PBS. Neuron cultures are washed with
PBS containing 1 mM MgCl2 and 0.1 mM CaCl2.
2. Precipitate cells by adding 1 mL of 4.5% perchloric acid. Incu-
bate for 15 min at room temperature, gently rotating plates by
hand every 2 min to prevent cells from drying.
3. Remove cells from the bottom of the dish using a hard-plastic
scraper and transfer to microcentrifuge tubes. Centrifuge at
16,000 g for 2 min at room temperature.
3.2.3 Deacylation of PPI 1. Remove and discard the supernatant. Rinse the pellet with
Lipids 1 mL of 100 mM EDTA and vortex for 15 s. Note that the
pellet may not break completely.
2. Centrifuge at 16,000 g for 2 min at room temperature and
discard the supernatant.
3. Dislodge the pellet in 50 μL of ddH2O. Add 1 mL of methyl-
amine reagent to each tube and dissolve the pellet by pipetting.
Transfer the contents to a 5-mL glass vial. It is critical to use
glass vials with a rubber-lined cap that can make a tight seal.
4. Heat the samples at 55 C for 1 h.
5. Allow the glass vials to cool for 5 min and transfer to micro-
centrifuge tubes. Dry the samples using a SpeedVac concentra-
tor at 55 C for 4 h under vacuum.
6. Dried samples can be stored at 20 C until the next day. If
samples have been stored at 20 C, thaw the samples on the
bench for 10 min before the next step.
3.2.4 Butanol/Ethyl 1. Resuspend each sample in 500 μL of ddH2O. Samples will not
Ether/Formic Acid be fully dissolved. Use a pipette tip to dislodge the pellet and
Extraction vortex each sample for about 30 s. Leave the samples at room
temperature for 30 min to let the pellet dissolve.
2. Centrifuge at 16,000 g for 5 min at room temperature.
3. Transfer the supernatant to a new tube. Add 500 μL of buta-
nol/ethyl ether/formic acid solution and vortex each tube for
15 s. Centrifuge the samples at 16,000 g for 5 min.
4. Transfer the bottom layer (aqueous phase) to a new
microcentrifuge tube.
5. Add another aliquot of 500 μL of butanol/ethyl ether/formic
acid solution and repeat vortex and centrifugation steps.
6. Transfer the bottom layer (aqueous phase) to a new

7. Dry samples using a SpeedVac concentrator at 55 C for 4 h
under vacuum.
8. Samples can be stored at 20 C prior to anion exchange
chromatography by HPLC, or the protocol can be continued.
Samples should be stable at 20 C for several months.
3.2.5 HPLC Sample 1. Thaw the samples at room temperature for about 10 min and
Preparation and Anion resuspend each sample in 60–75 μL of ddH2O. Vortex and
Exchange centrifuge the sample at 16,000 g for 10 min to pellet any
particulate matter.
2. Load 55 μL into the autosampler vial inserts, avoiding any par-
ticulate matter at the bottom of the tube (see Notes 9 and 10).
3. Inject 50 μL of sample into the HPLC and separate on a SAX
anion exchange column. Buffer A (ddH2O) and Buffer B (1 M
(NH4)2HPO4, pH 3.8) are used to generate the following
gradients run at a 1-mL/min flow rate: 0% Buffer B for
5 min, 0–2% Buffer B for 15 min, 2% Buffer B for 80 min,
2–10% Buffer B for 20 min, 10% Buffer B for 65 min, 10–80%
Buffer B for 40 min, 80% Buffer B for 20 min, and 80–100%
Buffer B for 5 min. At the end of the run, re-equilibrate the
column with ddH2O at 1 mL/min for 15 min. Radiolabeled
eluate can be detected by an inline flow scintillation analyzer. A
1:2 ratio of eluate to scintillant is used with a 3-mL/min flow
rate. Fractions are analyzed by binning counts every 6 s. The
data are analyzed by Laura-4 software.
4. To quantify scintillation counts from each sample, the raw
counts in each peak are expressed as a percentage of total
phosphatidylinositol, calculated from summation of the counts
of the eight glycero-inositol peaks (PI, PI3P, PI4P, PI5P,
PI3,5P2, PI4,5P2, PI3,4P2, and PI3,4,5P3). Background scin-
tillation counts are calculated from adjacent regions and sub-
tracted from all peaks. Figure 4 illustrates a typical mammalian
cell elution profile. The average PPI levels in three mammalian
cell types are presented in Table 2.
4 Notes
1. While other HPLC and flow scintillation systems may be

appropriate, our lab uses the Prominence UFLCXR liquid chro-
matograph (Shimadzu), with a SIL-20A autosampler (Shi-
madzu) and a Beta-RAM model 5 (LabLogic) flow detector.
This system requires autosampler vials and vial inserts
Fig. 4 Example of anion exchange chromatography of a mammalian (MEF) sample. Peaks were truncated at
200 cpm to display all the PPI detected
(Shimadzu). Our lab uses Laura-4 software (LabLogic) for

controlling the radiochromatography system.
2. To protect the anion exchange column, change the guard
column and wash the column after running 10 yeast samples
or 4 mammalian samples. Each time the guard column is
changed, the column is washed at a rate of 1 mL/min for 2 h
with ddH2O warmed to 40 C through all inlet lines. After the
wash is completed, run a mock sample (water), bypassing the
inline flow scintillation analyzer, and using the gradient
described for yeast samples, followed by re-equilibration of
the column with ddH2O at 1 mL/min for 15 min.
3. Pressure builds up over repeated usage of the column. Replace-
ment of the column is recommended when the pressure of the
column is no longer reduced after washing or when the peak
separation between PI4P and PI5P is poor. Reversing the
orientation of an older column will also help to relieve the
column pressure and improve peak separation.
4. When the HPLC is not in use, the guard column is changed,
washed as described above, and left in ddH2O.
5. Radiation safety approval from the institution and proper facil-
ities to monitor radioactive contamination and safely dispose of
radioactive waste are necessary before planning this
experiment.
Table 2
Levels of PPIs in mammalian cells determined by myo-3H-inositol labeling
These values were derived from 11 samples from neurons, 5 samples from mouse embryonic fibroblasts (MEF), and
3 samples from macrophages [25, 26]. Each box is shaded according to the levels of each lipid observed in MEF: less than
50% of MEF—dark blue; 50–10% of MEF levels—light blue; within 10% of MEF levels—white; 10–50% higher than
MEF levels—light red; more than 50% higher than MEF levels—dark red
6. Perchloric acid must be stored in a designated acid cabinet

because of its strong oxidizing properties. Radioactive samples
containing perchloric acid must be disposed separately from
the rest of the radioactive waste and neutralized before disposal
using sodium carbonate.
7. Note that our lab has been less successful in deacylation of
phosphoinositides with methylamine from other companies.
8. Yeast tends to grow more slowly in inositol-free media com-
pared to synthetic dropout media, so doubling times in
inositol-free media should be determined. In a wild-type
SEY6210 background, inoculation of 1 106 cells will yield
a culture with a density of approximately 5 106 cells/mL
16 h after inoculation.
9. While this protocol should yield an appropriate amount of

radioactive glycerol phosphoinositides, if there is concern that
a sample does not contain sufficient radioactive material, prior
to analyzing the sample on the HPLC, take 2 μL of sample
resuspended in ddH2O and perform a scintillation count on
the sample using a FlowLogic HS scintillant. To obtain good
accuracy of each PPI peak, the final sum of the peaks should
total 1 to 2 million cpm with not more than 3 million cpm.
10. In mammalian cells, if PI5P and PI4P are not separated into
discrete peaks, loading less sample may help to achieve the
separation. On the other hand, if the PI3,5P2 peak is not visible
above the background signal, loading more sample may help
detection of PI3,5P2. Peak separation varies between different
batches of the anion exchange column. Individual columns
should be tested empirically.
Acknowledgments
This work was supported by NIH grants R01-NS099340-03 and

R01-NS064015-09 to LSW, and LSI Cubed to NS and SSPG. NS
was supported in part by NIH T-32-GM007315.
References
1. Duex JE, Nau JJ, Kauffman EJ, Weisman LS BioEssays 37(3):267–277. https://doi.org/
(2006) Phosphoinositide 5-phosphatase Fig4p 10.1002/bies.201400129
is required for both acute rise and subsequent 6. Wallroth A, Haucke V (2018) Phosphoinosi-
fall in stress-induced phosphatidylinositol tide conversion in endocytosis and the endoly-
3,5-bisphosphate levels. Eukaryot Cell sosomal system. J Biol Chem 293
5(4):723–731. https://doi.org/10.1128/EC. (5):1526–1535. https://doi.org/10.1074/
5.4.723-731.2006 jbc.R117.000629
2. Balla T (2013) Phosphoinositides: tiny lipids 7. Botelho RJ, Efe JA, Teis D, Emr SD (2008)
with giant impact on cell regulation. Physiol Assembly of a Fab1 phosphoinositide kinase
Rev 93(3):1019–1137. https://doi.org/10. signaling complex requires the Fig4 phosphoi-
1152/physrev.00028.2012 nositide phosphatase. Mol Biol Cell 19
3. Dickson EJ, Hille B (2019) Understanding (10):4273–4286. https://doi.org/10.1091/
phosphoinositides: rare, dynamic, and essential mbc.E08-04-0405
membrane phospholipids. Biochem J 476 8. Malek M, Kielkowska A, Chessa T, Anderson
(1):1–23. https://doi.org/10.1042/ KE, Barneda D, Pir P, Nakanishi H, Eguchi S,
BCJ20180022 Koizumi A, Sasaki J, Juvin V, Kiselev VY,
4. Sasaki T, Takasuga S, Sasaki J, Kofuji S, Niewczas I, Gray A, Valayer A,
Eguchi S, Yamazaki M, Suzuki A (2009) Mam- Spensberger D, Imbert M, Felisbino S,
malian phosphoinositide kinases and phospha- Habuchi T, Beinke S, Cosulich S, Le
tases. Prog Lipid Res 48(6):307–343. https:// Novere N, Sasaki T, Clark J, Hawkins PT, Ste-
doi.org/10.1016/j.plipres.2009.06.001 phens LR (2017) PTEN regulates PI(3,4)P2
5. Shisheva A, Sbrissa D, Ikonomov O (2015) signaling downstream of class I PI3K. Mol Cell
Plentiful PtdIns5P from scanty PtdIns(3,5)P2 68(3):566–580 . e510. https://doi.org/10.
or from ample PtdIns? PIKfyve-dependent 1016/j.molcel.2017.09.024
models: evidence and speculation (response 9. Folch J (1949) Brain diphosphoninositide, a
to: DOI 10.1002/bies.201300012). new phosphatide having inositol
metadiphosphate as a constituent. J Biol Chem essential for protein sorting. Science 260
177(2):505–519 (5104):88–91
10. Folch J (1949) Complete fractionation of brain 20. Zhang Y, Zolov SN, Chow CY, Slutsky SG,
cephalin; isolation from it of phosphatidyl ser- Richardson SC, Piper RC, Yang B, Nau JJ,
ine, phosphatidyl ethanolamine, and dipho- Westrick RJ, Morrison SJ, Meisler MH, Weis-
sphoinositide. J Biol Chem 177(2):497–504 man LS (2007) Loss of Vac14, a regulator of
11. Agranoff BW, Murthy P, Seguin EB (1983) the signaling lipid phosphatidylinositol
Thrombin-induced phosphodiesteratic cleav- 3,5-bisphosphate, results in neurodegenera-
age of phosphatidylinositol bisphosphate in tion in mice. Proc Natl Acad Sci U S A 104
human platelets. J Biol Chem 258 (44):17518–17523. https://doi.org/10.
(4):2076–2078 1073/pnas.0702275104
12. Whiteford CC, Best C, Kazlauskas A, Ulug ET 21. Sarkes D, Rameh LE (2010) A novel HPLC-
(1996) D-3 phosphoinositide metabolism in based approach makes possible the spatial char-
cells treated with platelet-derived growth fac- acterization of cellular PtdIns5P and other
tor. Biochem J 319(Pt 3):851–860. https:// phosphoinositides. Biochem J 428
doi.org/10.1042/bj3190851 (3):375–384. https://doi.org/10.1042/
13. Bonangelino CJ, Nau JJ, Duex JE, BJ20100129
Brinkman M, Wurmser AE, Gary JD, Emr 22. Wakelam MJ (2014) The uses and limitations
SD, Weisman LS (2002) Osmotic stress- of the analysis of cellular phosphoinositides by
induced increase of phosphatidylinositol lipidomic and imaging methodologies. Bio-
3,5-bisphosphate requires Vac14p, an activator chim Biophys Acta 1841(8):1102–1107.
of the lipid kinase Fab1p. J Cell Biol 156 https://doi.org/10.1016/j.bbalip.2014.04.
(6):1015–1028. https://doi.org/10.1083/ 005
jcb.200201002 23. Jones DR, Ramirez IB, Lowe M, Divecha N
14. Clarke NG, Dawson RM (1981) Alkaline O (2013) Measurement of phosphoinositides in
leads to N-transacylation. A new method for the zebrafish Danio rerio. Nat Protoc 8
the quantitative deacylation of phospholipids. (6):1058–1072. https://doi.org/10.1038/
Biochem J 195(1):301–306. https://doi.org/ nprot.2013.040
10.1042/bj1950301 24. Kanehara K, Yu CY, Cho Y, Cheong WF,
15. Bligh EG, Dyer WJ (1959) A rapid method of Torta F, Shui G, Wenk MR, Nakamura Y
total lipid extraction and purification. Can J (2015) Arabidopsis AtPLC2 is a primary
Biochem Physiol 37(8):911–917. https://doi. phosphoinositide-specific phospholipase C in
org/10.1139/o59-099 phosphoinositide metabolism and the endo-
16. Hale AT, Clarke BP, York JD (2020) Metabolic plasmic reticulum stress response. PLoS Genet
labeling of inositol phosphates and Phosphati- 11(9):e1005511. https://doi.org/10.1371/
dylinositols in yeast and mammalian cells. journal.pgen.1005511
Methods Mol Biol 2091:83–92. https://doi. 25. McCartney AJ, Zolov SN, Kauffman EJ,
org/10.1007/978-1-0716-0167-9_7 Zhang Y, Strunk BS, Weisman LS, Sutton MA
17. Whitman M, Downes CP, Keeler M, Keller T, (2014) Activity-dependent PI(3,5)P2 synthesis
Cantley L (1988) Type I phosphatidylinositol controls AMPA receptor trafficking during syn-
kinase makes a novel inositol phospholipid, aptic depression. Proc Natl Acad Sci U S A 111
phosphatidylinositol-3-phosphate. Nature (45):E4896–E4905. https://doi.org/10.
332(6165):644–646. https://doi.org/10. 1073/pnas.1411117111
1038/332644a0 26. Samie M, Wang X, Zhang X, Goschka A, Li X,
18. Stack JH, DeWald DB, Takegawa K, Emr SD Cheng X, Gregg E, Azar M, Zhuo Y, Garrity
(1995) Vesicle-mediated protein transport: AG, Gao Q, Slaugenhaupt S, Pickel J, Zolov
regulatory interactions between the Vps15 SN, Weisman LS, Lenk GM, Titus S, Bryant-
protein kinase and the Vps34 PtdIns 3-kinase Genevier M, Southall N, Juan M, Ferrer M, Xu
essential for protein sorting to the vacuole in H (2013) A TRP channel in the lysosome reg-
yeast. J Cell Biol 129(2):321–334 ulates large particle phagocytosis via focal exo-
cytosis. Dev Cell 26(5):511–524. https://doi.
19. Schu PV, Takegawa K, Fry MJ, Stack JH, org/10.1016/j.devcel.2013.08.003
Waterfield MD, Emr SD (1993) Phosphatidy-
linositol 3-kinase encoded by yeast VPS34 gene
Chapter 2
Label-Free Quantification of Phosphoinositides

in Drosophila by Mass Spectrometry
Avishek Ghosh and Padinjat Raghu
Abstract
Phosphoinositides (PIs), the seven phosphorylated derivatives of phosphatidylinositol, are recognized as
key molecules in the control of multiple molecular events in eukaryotic cells. Within cells, PIs are
low-abundance lipids making their detection and quantification challenging. While many methods that
allow radiolabeling and quantification of PIs in the context of cultured cells are available, these are not
useful in the context of in vivo animal models where cell and developmental processes are best studied. In
this chapter, we describe radionuclide-free, mass spectrometry-based methods for the detection and
quantification of PIs from Drosophila tissues in vivo. The use of these methods should facilitate the
discovery of novel modes by which PIs regulate cellular and developmental processes in complex
metazoans.
Key words Phosphoinositides, Drosophila, Mass spectrometry, label-free quantification, In vivo

biochemistry
1 Introduction
Phosphoinositides (PIs) are the phosphorylated derivatives of

phosphatidylinositol that have been phosphorylated at positions
3, 4, and/or 5. Each of the positions can be phosphorylated
individually and in combinations giving rise to a set of seven mole-
cules: three monophosphates-PIPs [phosphatidylinositol
3-phosphate (PI3P), phosphatidylinositol 4-phosphate (PI4P),
and phosphatidylinositol 5-phosphate (PI5P)]; three bispho-
sphates-PIP2s [phosphatidylinositol 4,5-bisphosphate (PI4,5P2),
phosphatidylinositol 3,5-bisphosphate (PI3,5P2), and phosphati-
dylinositol 3,4-bisphosphate(PI3,4P2)], and one trisphosphate
[phosphatidylinositol 3,4,5-trisphosphate (PIP3)] [1]. These PIs
can be generated by the action of an evolutionarily conserved family
of lipid kinases and phosphatases that can selectively add or remove
a phosphate from the inositol headgroup [2]. Given that they are
lipid molecules, the PIs are not diffusible in the cytosol and are
19
20 Avishek Ghosh and Padinjat Raghu
therefore restricted to the membrane at which they are produced.

As a result, each of the seven PIs shows a characteristic distribution
at specific organelle membranes in a eukaryotic cell [3]. PIs regulate
numerous processes in cells, and much of this is achieved by their
ability to bind specifically to proteins in cells, thus regulating their
function [4]. Cellular and developmental processes controlled by
PIs can be studied in vivo, through the use of the metazoan model
organism Drosophila. The Drosophila genome encodes a set of
proteins that includes all of the known mammalian phosphoinosi-
tide kinases and phosphatases as well as orthologs of most of the
known PI-binding proteins [5]. Using the Drosophila model, it is
possible to study in vivo, the regulation of cellular and develop-
mental processes by PIs, many of which are conserved with mam-
malian models. A full understanding of the regulation of such
processes by PIs requires methods to measure and quantify PIs
from Drosophila cells and tissues. Although PIs have many impor-
tant roles in cell biological processes, their absolute levels in tissues
are very low, making their detection and quantification challenging.
To resolve this challenge, we have developed a set of methods to
measure the levels of PIPs, PIP2s, and PIP3 from Drosophila tissues
using highly sensitive liquid chromatography/tandem mass spec-
trometry (LC-MS/MS) techniques. These methods will facilitate
the use of the Drosophila model to investigate the role of PIs in
regulating cellular and developmental processes in a complex
metazoan.
Historically, the detection and quantification of PIs relied on
the ability to label the inositol head group with radioactive phos-
phorous, added by the activity of kinases. However, while this
remains a powerful approach for cultured cells, practical considera-
tions in the use of radionuclides limit its use for the analysis of PIs in
organisms in vivo. Mass spectrometry has emerged as a powerful
method for the detection and quantification of PIs from cells and
tissues without the need for radiolabeling. In this approach, small
molecules are identified on the basis of their unique molecular
mass. However, while using mass spectrometry to quantify PIs,
two key points should be noted:
1. Since the identification of small molecules by mass spectrome-
try relies on their unique mass, in the context of PIs, what this
means is that the three monophosphates PI3P, PI4P, and PI5P
cannot be individually distinguished on the basis of their mass
(as the mass of all three are identical). This also applies to the
three bisphosphorylated species PI(4,5P)2, PI(3,4P)2, and PI
(3,5)P2. However, since PI4P is the most abundant of the PIPs
(>90% of PIPs) and PI(4,5)P2 is the most abundant of the
PIP2s, any measurement of the mass of PIPs and PIP2s will
largely be a reflection of the levels of PI4P and PI(4,5)P2,
respectively.
Quantification of Phosphoinositides by Mass Spectrometry 21
Fig. 1 A representative PI(4,5)P2 molecule on the left with acyl chains R1 and R2. After derivatization with
TMS-diazomethane, the molecule gets methylated at the hydroxyl groups of the phosphate groups. Upon
fragmentation of the methylated parent mass, a methylated neutral head group and a charged diacyl glycerol
fragment are generated. The charged fragment mass can be used to quantify the R1 and R2 acyl chain-
containing phosphoinositide species by the MRM method described in the main text
2. In addition to the inositol headgroup, each PI contains two

fatty acyl chains (R1 and R2 in Fig. 1). The carbon chain length
and desaturation of these acyl chains can show considerable
variation. From the point of view of mass spectrometry, this will
require the identification of a number of species of PIs with
varying acyl chain composition from a tissue. A derivatization
technique originally described by Clark et al. has demonstrated
higher sensitivity of phosphoinositides due to methylation of
phosphate groups on the inositol head groups (described in
Fig. 1) [6]. Consequently, derivatizing biological lipid extracts
by this technique can estimate all of the unique acyl chain phos-
phoinositides by setting up multiple reaction monitoring
(MRM)-based analysis pipeline described in Ghosh et al.
[7]. Briefly, the mass spectrometer is instructed to scan for a
parent mass ion/daughter mass ion (diacylglycerol fragment in
Fig. 1) pair (MRM transition) corresponding to a particular
methylated phosphoinositide species. If the phosphoinositide
corresponding to the abovementioned MRM transition is pres-
ent in the sample, the mass spectrometer will detect and acquire
the signal. The details of performing the experiment and quan-

tifying the signal will be discussed in the subsequent sections.
Another advantage of using these methods over traditional
radioactivity-based methods is the ability to control for extrac-
tion variability by the use of synthetic internal standards
(ISDs). ISDs are designed in a manner such that their molecu-
lar mass is distinguishable from the mass of naturally occurring
lipids from biological material. This allows us to spike known
amounts of ISD during the event of extraction and subse-
quently quantify mass spectrometer-generated signal response
of biological lipids relative to the response of the corresponding
ISD. A total organic phosphate assay is performed in conjuga-
tion with these experiments to normalize for size differences
across biological samples. An overview of the methods
described below is depicted in Fig. 2a.
2 Materials
2.1 Fly Stock Rearing 1. Flies (Drosophila melanogaster).

and Dissections 2. Fly food: corn flour 80 g/L, D-glucose 20 g/L, sucrose 40 g/
L, agar 8 g/L, yeast extract 15 g/L, propionic acid 4 mL,
TEGO (methyl parahydroxybenzoate) 0.7 g/L, orthopho-
sphoric acid 0.6 mL.
3. Filter paper.
4. Forceps.
5. Blade holder.
6. Breakable blades.
7. Butter paper.
8. Dissecting microscopes.
9. Silicone pads to dissect retinae.
10. Illumination controlled incubators containing water in a bea-
ker to maintain humidity.
11. Humidified CO2 connected to fly pads for anesthetizing flies.
2.2 Total Lipid 1. Glass bottles: transparent and amber color.

Isolation 2. Chloroform (see Note 1).
3. Methanol.
4. LC-MS grade water.
5. Phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 mM KCl,
10 mM Na2HPO4, and 1.8 mM KH2PO4, pH 7.4.
6. Low retention tips and microcentrifuge tubes.
7. Homogenization tubes with zirconium oxide beads.
Fig. 2 (a) Schematic showing a typical lipid extraction procedure followed for analysis of phosphoinositides
using LC-MS/MS. (b) Schematic showing the order of events for acquiring data for running the LC-MS/MS
method
8. Phosphoinositide elution buffer (PEB): chloroform/metha-

nol/2.4 M hydrochloric acid in a ratio of 250/500/200 (vol/-
vol/vol). Do not store for more than 3 months.
9. Initial organic mixture: 24 mL methanol:12 mL chloroform.
10. Lower phase wash buffer (LPWS): methanol/1 M hydrochlo-

ric acid/chloroform in a ratio of 235/245/15 (vol/vol/vol).
Do not store for more than 3 months.
11. 37% pure HCl.
12. 100 μM bovine insulin.
2.3 Lipid Internal 1. 17:0|14:1 PE.

Standards (See Note 2) 2. 17:0|20:4 PI4P.
2.3.1 For Retinae 3. 16:0|16:0 d5-PI(3,5)P2.
2.3.2 For Larvae 1. 17:0|20:4 PI(4,5)P2.

2. 17:0|20:4 PI(3,4,5)P3.
2.4 Neomycin 1. Neo-beads™ (Echelon Biosciences) (see Note 3).

Chromatography 2. Phosphoinositide-binding buffer (PBB): chloroform/metha-
nol/water/2 M ammonium formate in a ratio of 100/200/
16.8/3.2 (vol/vol/vol/vol). Do not store for more than
3 months.
3. Phosphoinositide elution buffer (PEB).
2.5 18O PI5P 1. 10 mM Tris–HCl (pH 7.4).

Mass Assay 2. Diethyl ether.
3. 0.5 mM bovine brain-derived phosphatidylserine (PS).
4. GST-PIP4K enzyme (see Note 4).
5. Kinase assay buffer: 100 mM Tris–HCl (pH 7.4), 20 mM
magnesium chloride, 2 mM EGTA, and 140 mM potassium
chloride. Store the buffer at 4 C when not in use; it is stable for
3 months.
18
6. O-ATP (see Note 5).
2.6 Derivatization 1. TMS-diazomethane.

2. Glacial acetic acid.
3. Post-derivatization wash solution: chloroform:methanol:water
8:4:3 (vol/vol/vol) (see Note 6).
2.7 Total Organic 1. High-temperature-resistant phosphate-free glass tubes.

Phosphate Assay 2. Perchloric acid.
3. Ammonium molybdate.
4. Ascorbic acid.
5. Phosphate standard potassium dihydrogen phosphate.
6. 96-well plate.
7. BOD incubator.
2.8 Liquid 1. Column: C4, 300 Å, 1.7 μm, 1 mm 100 mm column
Chromatography (see Note 5).
2. Solvents: solvent A: 0.1% formic acid in water; solvent B: 0.1%
formic acid in acetonitrile. All solvents and chemicals should be
of MS grade.
3. Chromatography amber color glass vials with inserts.
2.9 Lab Equipment 1. Homogenizer instrument (see Note 5).

2. Rotospin instrument.
3. Table-top centrifuge.
4. Vortex machine.
5. Thermo-mixer.
6. Bath sonicator.
7. Dry heat bath.
8. Vacuum centrifugal concentrator.
9. Fume hood.
10. UPLC system (see Note 5).
11. Triple quadrupole mass spectrometer (see Note 5).
3 Methods
3.1 Lipid Extraction Phosphoinositides support a varied range of physiological functions

in most metazoans. In Drosophila, processes such as phototransduc-
tion are supported by phosphoinositides PI4P and PI(4,5)P2.
Therefore, the adult photoreceptor or the retinae provides a good
model tissue to study the same. These cells have a larger proportion
of apical plasma membrane, and it is a fair assumption that more
than 90% of the PIPs and PIP2 measured from these cells will
represent PI4P and PI(4,5)P2 [5, 8, 9]. Therefore, we choose
retinae for the analysis of PIPs and PIP2. In contrast, we choose
third-instar wandering larvae for the analysis of PIP3 and PI5P
because these low-abundance phosphoinositides have been corre-
lated to the maintenance of the physiological processes of growth
and development in Drosophila [7, 10, 11]. All these analyses can be
extended to measure these lipids from other tissues of Drosophila
with standardization of the starting quantity of tissue. The proto-
cols described in this section are best-suited extraction methods for
PIP, PIP2, and PIP3 according to our analysis.
3.1.1 Sample Collection 1. PIP and PIP2 measurement: Collect Drosophila adults of
defined ages in plastic microcentrifuge tubes. Anesthetize flies
by subjecting them to cold shock by keeping the tubes in ice for
15 min.
2. Dissect 25 retinae per sample under conditions of appropriate

illumination (i.e., red or white light) and store in 50 μL 1 PBS
in homogenizer tubes on dry ice until ready for extraction
(see Note 7).
3. PI5P and PIP3 measurement: Identify third-instar wandering
larvae from the walls of the media bottles using protruding
spiracles as a morphological feature.
4. Collect these larvae, wash using 1 PBS, and blot them dry
using tissue paper.
5. Transfer the dried larvae to homogenizer tubes (for PI5P
analysis) or microcentrifuge tubes (for PIP3 analysis) in batches
of five.
3.1.2 Lipid Extraction 1. To the tissue collected in the homogenizer tubes in Subhead-
for Total PIP-, PIP2-, ing 3.1.1, add 200 μL of phosphoinositide elution buffer
and 18O-ATP-Based PI5P (PEB) and homogenize for 10 s; repeat this four times keeping
Mass Assay Measurements the tubes on ice for 2 min between each round of homogeni-
zation (see Note 8).
2. Collect the homogenate in a 1.5-mL low retention microcen-
trifuge tube and wash the tube with 750 μL of PEB and add the
washed content to the 1.5-mL microcentrifuge tube.
3. Place the 1.5-mL microcentrifuge in a bath sonicator, and
sonicate at maximum amplitude for 2 min (see Note 9).
4. To the sonicated sample, add 250 μL of chloroform followed
by internal standards. For PIP and PIP2 analysis, add 0.2 ng of
17:0|14:1 PE and 50 ng of d5-PI(3,5)P2. For PI5P analysis,
add 20 ng of 17:0|20:4 PI(4,5)P2.
5. After this, add 250 μL of LC-MS grade water to the same tube.
6. Vortex for 2 min in the vortex mixer at 250 g.
7. Obtain clean phase separation by centrifuging for 5 min at
1000 g.
8. Puncture through the upper aqueous phase obtained and
transfer the lower phase to a fresh tube. To this, add 1000 μL
of lower phase wash solution (LPWS) and repeat steps 5 and 6.
9. For analysis of PIP and PIP2, proceed with derivatization
(Subheading 3.4). For the analysis of PI5P, dry the sample
obtained at step 8 in a centrifugal concentrator operating
in vacuo and follow Subheading 3.1.3 (seeNote 10).
3.1.3 Phosphoinositide 1. Reconstitute the dried lipids (from Subheading 3.1.2, obtained
Enrichment by Neomycin from step 8) in 30 μL of chloroform and add 950 μL of
Chromatography phosphoinositide-binding buffer (PBB). To this, add 2 mg of
equivalent Neo-beads™ and mix by rotation using a rotospin
instrument operating at 0.001 g for 1 h at room temperature.
2. Spin down the beads at 1000 g for 5 min.

3. Take the upper PBB containing the unbound lipid fraction for
organic phosphate assay (proceed with Subheading 3.2) in a
fresh microcentrifuge tube (can be transferred to phosphate-
free glass tubes later) (see Note 11).
4. Add 1 mL of fresh PBB to the tubes and wash the beads to
remove nonspecifically bound lipids (mix by rotation at 0.001
g) for 15 min.
5. Discard the PBB as much as possible, and to this, add 950 μL of
PEB and mix by rotation at 0.001 g speed for 1 h at room
temperature.
6. Remove the PEB and transfer it to a 1.5-mL microcentrifuge.
To this, add 250 μL of chloroform and 250 μL of LC-MS-
grade water.
7. Vortex for 2 min on a vortex mixer at 250 g.
1000 g.
9. Puncture through the upper aqueous phase and transfer the
lower phase to a new tube. To this, add 20 μL of 0.5 mM PS
and mix, spin, and dry the lipids in a centrifugal concentrator
operating in vacuo for 30–40 min (see Note 12).
10. The lipids are now ready to be prepared for the 18O PI5P mass
assay (Subheading 3.3).
3.1.4 Lipid Extraction 1. Dissect and open up five larvae in 1 PBS and immediately
for Total PIP3 transfer into 37.5 μL of 1 PBS in a 2-mL
2. For insulin stimulation, add 37.5 μL of 100 μM insulin to the
larval samples and incubate the tube on a mix mate shaker for
10 min. For sham stimulation, directly incubate dissected lar-
vae in 75 μL of 1 PBS.
3. At the end of incubation time, add 750 μL of ice-cold initial
organic mixture to stop the reaction. Decant the upper part of
this solution and store it in a 2-mL microcentrifuge tube and
transfer the rest of the solution containing the carcasses into a
homogenization tube.
4. Homogenize for 10 s; repeat this four times, keeping the tubes
on ice for 2 min between each homogenization.
5. Transfer the entire homogenate to a fresh 2-mL microcentri-
fuge tube and wash off the homogenization tube with the
decanted mix kept aside earlier.
6. To the pooled homogenate collected in the tube, add 120 μL
of water, followed by 5 ng of 17:0|20:4 PIP3 internal standard
(ISD). Vortex the mixture for 2 min and add 725 μL of chlo-
roform to it.
7. After this, vortex again for 2 min at 250 g and obtain clean
phase separation by centrifuging for 5 min at 1000 g.
8. Remove 1 mL of the lower organic phase and store it in a fresh
2-mL tube. Add 725 μL of chloroform to the remaining aque-
ous upper phase. Repeat step 7. Again, collect 1 mL of the
organic phase and pool with the previous collection (total of
2 mL). This organic phase will be used for measuring total
organic phosphate (Subheading 3.2).
9. To the aqueous phase obtained at step 8, add 500 μL of the
initial organic mixture followed by 170 μL of 2.4 M HCl and
500 μL of chloroform. Vortex this mixture for 5 min at 250
g and allow it to stand at room temperature for 5 min.
10. Obtain clean phase separation by centrifugation (1500 g,
5 min). Next, collect the lower organic phase into a fresh tube
by piercing through the protein band sitting at the interface.
11. Add 700 μL of LPWS, and vortex the mixture. Obtain clean
phase separation by centrifugation (1500 g, 5 min). Transfer
the resultant lower organic phase completely into a fresh tube
and proceed with derivatization (Subheading 3.4).
3.2 Organic 1. Take 500 μL from flow-through collected in step 3 of Sub-

Phosphate Assay heading 3.1.3 or 800 μL obtained in step 8 of Subheading
3.1.4 into a phosphate-free glass tube as starting material for
this assay.
2. Evaporate the samples to dryness in phosphate-free glass tubes
in a dry heat bath at 90 C (see Note 13).
3. Prepare standard dilutions (0–150 nanomoles, 10 points) in
similar phosphate-free glass tubes and dry them in an oven at
120 C.
4. Cool down the tubes, add 50 μL of 70% perchloric acid, and
mix the contents by vortexing. Heat the tubes at 180 C for
30 min in a dry heat bath (see Note 14).
5. Cool down the tubes, and to it, add 250 μL of water, 50 μL of
2.5% ammonium molybdate, and 50 μL of 10% ascorbic acid.
6. Incubate the tubes in a shaking incubator at 37 C for 1 h.
7. Aliquot 130 μL from each tube onto a 96-well plate and read
the absorbance at 630 nm in a multimode spectrophotometric
reader (see Note 15).
8. Plot the standard curve to obtain linearity. Then, acquire
organic phosphate values of biological samples by fitting the
absorbance values in the linearity equation.
3.3 18O-ATP PI5P 1. To the dried lipid samples from Subheading 3.1.3, add 50 μL
Mass Assay of 10 mM Tris–HCl (pH 7.4) and 50 μL of diethyl ether. Then,
sonicate in a bath sonicator for 2 min at maximum amplitude.
2. Separate the phases by centrifuging for 5 min at 1000 g.
Mark the junction of the phases with a marker pen.
3. Dry the upper ether phase in a centrifugal concentrator
in vacuo for no more than 3 min (see Note 16).
4. Transfer the tubes to ice and store them for 10 min.
5. Prepare a mixture of 2 kinase buffer with 80 μM 18O-ATP
and 2 μg equivalent of GST-PIP4K enzyme (kinase cocktail);
mix well.
6. To the tubes, add 50 μL of kinase cocktail (prepared in step 5)
and incubate for 16 h at 30 C, at constant shaking in a thermo-
mixer.
7. End the reaction by adding 125 μL of 2.4 N HCl, 250 μL of
chloroform, and 250 μL of methanol.
8. Vortex for 2 min in the vortex mixer at 250 g.
1000 g.
10. Transfer the lower phase to a new tube by puncturing through
the protein band at the interface of the two phases. To this, add
500 μL of LPWS and repeat steps 8 and 9.
11. Proceed with derivatization (Subheading 3.4) of the lower
organic phase obtained at the end of step 10.
3.4 Derivatization 1. To the lower organic phase of the samples obtained from the
end of Subheadings 3.1.2, 3.1.4, or 3.3, add 50 μL of 2 M
TMS-diazomethane (see Note 17).
2. Allow the reaction to proceed at room temperature for 10 min
at constant shaking. After 10 min, add 10 μL of glacial acetic
acid to quench the reaction, which is indicated by the disap-
pearance of the yellow color of the solution.
3. Tap the tubes and open them with caution to remove the N2
formed in the quenching reaction. Spin down the contents, and
to this, add 600 μL of post-derivatization wash solution
(see Note 18).
4. Vortex the contents for 2 min in the vortex mixer at 250 g.
5. Discard ~400 μL from the upper phase that will form at the end
of step 4. Repeat step 4.
6. Discard the whole of the upper phase and add 50 μL of 90%
methanol to the lower phase and mix it.
7. Dry the samples for 2 h in a centrifugal concentrator operating
in vacuo.
8. After drying for the above time, the tubes should contain
~20 μL of the remaining sample. To this, add 180 μL of
methanol, mix well, and store at 4 C for up to 2–3 days
prior to LC-MS/MS.
3.5 Liquid 1. Separate samples obtained from step 8 of Subheading 3.4 by

Chromatography–- liquid chromatography using a UPLC system on a C4 column
Tandem Mass of dimensions 300 Å, 1.7 μm, 1 100 mm, using a 45–100%
Spectrometry acetonitrile in water (with 0.1% formic acid) gradient over
10 min (see Table 1 for details) (see Note 5).
2. Ensure that the mass spectrometer software can integrate the
UPLC software before starting the runs.
3. Connect the UPLC system to a triple quadrupole mass spec-
trometer. Inject the eluate from the UPLC system into the
mass spectrometer for analysis.
4. In the sample injection sequence, always start with blank sol-
vent injection (methanol) and keep neat standard samples
intermittently in between biological samples for run quality
control checks (see Note 19).
5. Analyze all samples in the positive mode of the mass spectrom-
eter. Mass spectrometer parameters are as follows: dwell time:
65 ms; CAD (collisionally activated dissociation): 2;
GS1 (source gas 1) and GS2 (source gas 2): 20; CUR (curtain
gas): 37; IS (ESI voltage): 5200; and TEM (source tempera-
ture): 350 C. The phospholipid specific mass spectrometer
parameters are mentioned in Table 2.
6. Tune the mass spectrometer to be used for data acquisition
specifically for the phosphoinositides in discussion (described
as a schematic in Fig. 2b). The various operating parameters of
the triple quadrupole mass spectrometer used in our analysis
Table 1
Liquid chromatography parameters and solvents for the separation of PIs prior to mass spectrometry
Time (in min) % solvent A (0.1% % solvent B (0.1%

1 mm 100 mm column Flow rate (mL/min) formic acid in water) formic acid in acetonitrile)
Initial 0.100 55 45
5 0.100 55 45
10 0.100 0 100
15 0.100 0 100
16 0.100 55 45
20 0.100 55 45
Table 2
Mass spectrometer setting optimized for detecting PIs using the methods described in this chapter
Parameters PIP species PIP2 species PIP3 species PE species

DP (declustering potential) 140 60 35 65
EP (entrance potential) 12 11 11 11
CE (collision energy) 29 37 37 30
CXP (collision cell exit potential) 12 15 15 12
have been described in detail previously and in Table 2 [7, 8,

11]. The MRM parameters for the various species of PIP, PIP2,
and PIP3 in Drosophila have been described in [7, 8, 11] and
consolidated in Tables 3 and 4 (see Note 20).
3.6 Analysis for Peak 1. The data files should be locally available before using the anal-
Integration ysis software provided by the mass spectrometer vendor.
2. At first, create a new quantitation method using the “File” !
“New Quantitation method” option.
3. Browse and select a sample data file and set the parameters for
analysis for peak integration for individual MRMs from here.
4. For peak integration, use default parameters except for “Gauss-
ian smooth width,” set at 5; and “minimum peak height,” set at
values thresholding the signal from blank. Do this for all the
MRMs in the method. Fig. 3 depicts a typical analysis window
of the software used in our analysis.
5. Next, save this quantitation method locally and open a new
results table using “File” ! “New Results table” option.
6. Browse and select the data files to be analyzed and then proceed
with selecting the above-saved quantitation method.
7. Once loaded, proceed to individually curate the chromato-
grams for peak shape, retention time shifts, and signal manually
from the software to finalize the values for the area under the
curve for each biological species (MRM) (Fig. 3).
8. After this, copy the data table onto an excel worksheet and
proceed with calculations.
3.7 Calculations 1. Divide the area under the curve value of each biological PIP
of Normalized and PIP2 species (MRM) by the value of the area under the
Intensities curve of the respective internal standard [17:0|20:4 PI4P for
PIPs and d5-PI(3,5)P2 for PIP2].
3.7.1 For PIP and PIP2
2. Multiply the ratio obtained to the actual amount of ISD added
(in ng) to obtain the estimated amount of biological PIP or
PIP2 analyte.
Table 3
List of MRMs for identifying the individual molecular species of PIs with specific acyl chain
composition
18
O-PIP2
Acyl chain length PIP parent ion PIP2 parent ion PIP3 parent ion parent ion Daughter ion
32:0 933.5 1041.5 1149.5 1047.5 551.5
32:1 931.5 1039.5 1147.5 1045.5 549.5
32:2 929.5 1037.5 1145.5 1043.5 547.5
32:3 927.5 1035.5 1143.5 1041.5 545.5
34:0 961.5 1069.5 1177.5 1075.5 579.5
34:1 959.5 1067.5 1175.5 1073.5 577.5
34:2 957.5 1065.5 1173.5 1071.5 575.5
34:3 955.5 1063.5 1171.5 1069.5 573.5
34:4 953.5 1061.5 1169.5 1067.5 571.5
34:5 951.5 1059.5 1167.5 1065.5 569.5
36:0 989.5 1097.5 1205.5 1103.5 607.5
36:1 987.5 1095.5 1203.5 1101.5 605.5
36:2 985.5 1093.5 1201.5 1099.5 603.5
36:3 983.5 1091.5 1199.5 1097.5 601.5
36:4 981.5 1089.5 1197.5 1095.5 599.5
36:5 979.5 1087.5 1195.5 1093.5 597.5
38:0 1017.5 1125.5 1233.5 1131.5 635.5
38:1 1015.5 1123.5 1231.5 1129.5 633.5
38:2 1013.5 1121.5 1229.5 1127.5 631.5
38:3 1011.5 1119.5 1227.5 1125.5 629.5
38:4 1009.5 1117.5 1225.5 1123.5 627.5
38:5 1007.5 1115.5 1223.5 1121.5 625.5
17:0|20:4 ISD 995.5 1103.5 1211.5 – 613.5
3. Similarly, divide the area under the curve value of each

biological PE species (MRM) by the value of the area under
the curve of the internal standard (17:0|14:1 PE). Multiply the
ratio obtained to the actual amount of ISD added (in ng) to
obtain the estimated amount of biological PE analyte.
4. For normalization, divide the estimated amount of acyl chain
length matched PIP or PIP2 species to the estimated amount of
PE obtained in previous steps.
Table 4
List of MRMs for identifying the major species of PE in Drosophila tissues
Acyl chain length PE parent ion PE daughter ion

34:2 730.5 575.5
34:3 728.5 573.5
36:2 758.5 603.5
36:3 756.5 601.5
17:0|14:1 ISD 690.5 535.5
Fig. 3 A software analysis window depicting the chromatograms of biological samples for the analysis of one
of the MRMs (17:0|20:4 PIP2 ISD). The window shows parameters such as “Gaussian smooth width” and
“minimum peak height,” which are described in the main text
3.7.2 For 18O-PIP2 1. Divide the area under the curve value of each biological
18
(Biological PI5P) O-PIP2 species (MRM) by the value of the area under the
curve of the internal standard [17:0|20:4 PI(4,5)P2].
2. For sample size normalization, divide the above ratio value by the
organic phosphate value obtained in step 8 of Subheading 3.2.
3.7.3 For Total PIP3 1. Divide the area under the curve value of each biological PIP3
species (MRM) by the value of the area under the curve of the
internal standard (17:0|20:4 PIP3).
2. For sample size normalization, divide the above ratio value by the
organic phosphate value obtained in step 8 of Subheading 3.2.
4 Notes
1. Do not store chloroform in large quantities and for longer

times. Also, use amber color glass bottles for storing
it. Chloroform breaks down into toxic phosgene gas upon
exposure to light and oxygen.
2. Commercially available lipid standards usually come in powder
form to be reconstituted upon breaking the seal into vendor-
specified solvents. Always reconstitute the full contents without
measuring it separately in a fine weighing balance and note the
concentration values thus obtained. Weighing portions of
vendor-supplied lipids inadvertently leads to errors in stock
concentrations. Usually, the reconstituted stocks can be stored
in chromatography glass vials at 20 C for 6–12 months
without repeated thawing. Prepare dilutions based on calcula-
tions of corresponding standards and use them repeatedly
across experiments.
3. Reconstitute Neo-beads™ in 150 μL of 50% methanol such
that 2 mg equivalent bead slurry is about 15 μL. Do cut the tip
before aspirating the bead slurry to avoid damaging the beads.
4. The GST-PIP4K recombinant enzyme should be prepared in
advance in the lab as published earlier [12]. The enzyme should
be stored in smaller aliquots in a 80 C freezer until use.
5. These items are proprietary and have been used in our analysis
based on previous work in our and other laboratories [7, 12,
13]. However, if similar products are available from other
vendors, they can be used following proper standardizations.
The list of items and their catalog numbers are as follows:
(1) 18O-ATP, Cambridge Isotope Laboratory (OLM-7858-
20); (2) ACQUITY® UPLC Protein BEH C4 (186005590);
(3) Minilys™ Personal Homogenizer (P000673-MLYS0-A);
(4) Waters Acquity® UPLC I class system; and (5) hybrid triple
quadrupole mass spectrometer (Sciex 6500 Q-Trap®).
6. Keep this ready the day before the derivatization reaction.

Measure and pour the contents in a small glass bottle and
vigorously shake for a minute. Allow it to settle for phase
separation and use the top aqueous layer for post-
derivatization wash.
7. The number of retinae have been standardized based on our fly
culturing conditions and handling. We would recommend
performing a pilot experiment with three different amounts
of starting material to obtain the right sample size based on
the culture conditions.
8. Transfer the retinae tissue after dissection to the homogeniza-
tion tubes using fine paintbrushes. Organize the dissection in a
place with minimum airflow to ensure the retinae to not fall off
the brushes while transferring. For larvae, transfer the dissected
larvae using forceps to microcentrifuge tubes for sham or insu-
lin stimulation with as minimum PBS from the dissection watch
glass as possible.
9. Change the water of the bath sonicator if it is turbid before use.
The water becomes turbid due to microbial growth over
repeated use and decreases the efficiency of the sonication.
10. The sample size normalization for the analysis of PIP and PIP2
from adult retinae is done by estimating phosphatidylethanol-
amine (PE) from the same extracts instead of measuring total
organic phosphate levels. This is based on the fact that (a) our
genetic perturbation does not affect PE levels and (b) total
organic phosphate values for retinae samples usually come out
to be below the detection limit of the assay. Do not store the
samples for PIP and PIP2 analysis until the derivatization step is
complete.
11. While taking the PBB into a microcentrifuge tube, ensure that
glass beads containing the enriched phosphoinositides do not
come in.
12. Ensure that the lipids are not overdried at this step by checking
the status of drying after every 15-min interval within the
drying time.
13. Dry the samples in a fume hood. Even after the samples are
dried to completion, there might be a soft jelly-like deposit on
the bottom of the glass tubes. Do not dry the tubes beyond
this point. It would usually take 15–30 min to dry the organic
samples in a dry heat bath.
14. After adding perchloric acid, mix the dried organic phosphate
lipids to make a homogeneous mixture before keeping for
heating at 180 C. Heat the samples in a chemical fume hood.
15. Do not store the samples at this stage without taking a reading.
Storage for a long duration will lead to precipitate formation.
16. Using a vacuum centrifugal concentrator for a long time can

decrease the efficiency of the condenser cup and might take
longer to dry samples. Note that the vacuum centrifugal con-
centrator should be operated freshly at this step so that it can
function at its maximum potential and dry out the ether phase
above the aqueous phase in exactly 2 min.
17. Use necessary precautions while handling this reagent. Basic
equipment items such as chemical fume hood, protective
masks, and lab coat are necessary. Glacial acetic acid can be
used to neutralize TMS-diazomethane. Therefore, this can be
used for discarding tips and other consumables that have been
in contact with TMS-diazomethane.
18. Perform this step in a fume hood. Note that the microcentri-
fuge tubes should be opened gently at this step, failing which
the contents might spill out due to pressurized N2 in the tubes.
Work in the regular workspace without masks after this step.
19. Neat standards are prepared by derivatizing an equal amount of
ISD used for spiking biological samples during extraction. In
order to do so, derivatize corresponding amounts of ISD, and
process them similarly like biological samples. Keep intermit-
tent injections of corresponding ISD in between biological
samples and check the variability of signal in between multiple
injections of the same ISD at different times to control for MS
run quality. For example, for an experiment to estimate PI5P
levels, derivatize 20 ng 17:0|20:4 PI(4,5)P2 in a separate tube,
and process it with Subheading 3.4 and onward. Use this
reconstituted sample as a quality control neat standard.
20. Tuning parameters is a vendor-specific detail of the mass spec-
trometer that is used for analysis. In this context, the MS
parameters have been tuned in-house using synthetic standards
of PIP, PIP2, and PIP3. The MRMs have been set up after
running each biological sample in a Neutral Loss (NL) Scan
mode available with the mass spectrometer used in our analysis.
The methylated neutral head group of all of the phosphoinosi-
tide species is of fixed mass: 382 Da for PIP, 490 Da for PIP2,
598 Da for PIP3, and 496 Da for 18O-PIP2. Therefore, NL
scans revealed all the precursor masses (Q1) that generate
charged fragments after the loss of the corresponding neutral
masses for each of the species. Additionally, in the case of PIP3
measurements, among all species detected by NL scans, only
those that showed a reproducible increase in levels in the
insulin-stimulated samples were used for quantification. The
table for MRM is prepared based on this NL scan information
and can be manually deduced to be of certain chain length. We
have provided all the MRMs and their corresponding MS
parameters to be used for each of the phosphoinositide analyses
discussed in the methods above in Tables 2, 3 and 4.
Acknowledgments
This work was supported by the National Centre for Biological

Sciences-TIFR and a Wellcome DBT India Alliance Senior Fellow-
ship to PR (IA/S/14/2/501540). We thank Dhananjay Shinde,
Sanjeev Sharma, Sruthi Balakrishnan, and the NCBS mass spec-
trometry Facility for assistance while developing these methods.
References
1. Balla T (2013) Phosphoinositides: tiny lipids 8. Balakrishnan SS, Basu U, Shinde D et al (2018)
with Giant impact on cell regulation. Physiol Regulation of PI4P levels by PI4KIIIα during
Rev 93:1019–1137. https://doi.org/10. G-protein-coupled PLC signaling in drosoph-
1152/physrev.00028.2012 ila photoreceptors. J Cell Sci 131:jcs217257.
2. Sasaki T, Takasuga S, Sasaki J et al (2009) https://doi.org/10.1242/jcs.217257
Mammalian phosphoinositide kinases and 9. Thakur R, Panda A, Coessens E et al (2016)
phosphatases. Prog Lipid Res 48:307–343. Phospholipase D activity couples plasma mem-
https://doi.org/10.1016/j.plipres.2009.06. brane endocytosis with retromer dependent
001 recycling. Elife 5:e18515. https://doi.org/
3. Di Paolo G, De Camilli P (2006) Phosphoino- 10.7554/eLife.18515
sitides in cell regulation and membrane dynam- 10. Gupta A, Toscano S, Trivedi D et al (2013)
ics. Nature 443:651–657. https://doi.org/10. Phosphatidylinositol 5-phosphate 4-kinase
1038/nature05185 (PIP4K) regulates TOR signaling and cell
4. Hammond GRV, Balla T (2015) Polypho- growth during drosophila development. Proc
sphoinositide binding domains: key to inositol Natl Acad Sci U S A 110:5963–5968. https://
lipid biology. Biochim Biophys Acta doi.org/10.1073/pnas.1219333110
1851:746–758. https://doi.org/10.1016/j. 11. Sharma S, Mathre S, Ramya V et al (2019)
bbalip.2015.02.013 Phosphatidylinositol 5 phosphate 4-kinase reg-
5. Balakrishnan SS, Basu U, Raghu P (2015) ulates plasma-membrane PIP 3 turnover and
Phosphoinositide signalling in drosophila. Bio- regulates plasma-membrane PIP 3 turnover
chim Biophys Acta Mol Cell Biol Lipids and insulin signaling. Cell Rep
1851:770–784. https://doi.org/10.1016/j. 27:1979–1990.e7. https://doi.org/10.1016/
bbalip.2014.10.010 j.celrep.2019.04.084
6. Clark J, Anderson KE, Juvin V et al (2011) 12. Jones DR, Ramirez IB-R, Lowe M, Divecha N
Quantification of PtdInsP3 molecular species (2013) Measurement of phosphoinositides in
in cells and tissues by mass spectrometry. Nat the zebrafish Danio rerio. Nat Protoc
Methods 8:267–272. https://doi.org/10. 8:1058–1072. https://doi.org/10.1038/
1038/nmeth.1564 nprot.2013.040
7. Ghosh A, Sharma S, Shinde D et al (2019) A 13. Sharma S, Mathre S, Ramya V et al (2019)
novel mass assay to measure phosphatidylino- Phosphatidylinositol 5 phosphate 4-kinase reg-
sitol-5-phosphate from cells and tissues. Biosci ulates plasma-membrane PIP3 turnover and
Rep 39:707604. https://doi.org/10.1042/ insulin signaling. Cell Rep 27:1979–1990.e7.
BSR20192502 https://doi.org/10.1016/j.celrep.2019.04.
084
Chapter 3
Profiling of Phosphoinositide Molecular Species in Resting

or Activated Human or Mouse Platelets by a LC-MS Method
Gaëtan Chicanne, Justine Bertrand-Michel, Julien Viaud, Karim Hnia,
Jonathan Clark, and Bernard Payrastre
Abstract
Our knowledge of the role and biology of the different phosphoinositides has greatly expanded over recent
years. Reversible phosphorylation by specific kinases and phosphatases of positions 3, 4, and 5 on the
inositol ring is a highly dynamic process playing a critical role in the regulation of the spatiotemporal
recruitment and binding of effector proteins. The specific phosphoinositide kinases and phosphatases are
key players in the control of many cellular functions, including proliferation, survival, intracellular traffick-
ing, or cytoskeleton reorganization. Several of these enzymes are mutated in human diseases. The impact of
the fatty acid composition of phosphoinositides in their function is much less understood. There is an
important molecular diversity in the fatty acid side chains of PI. While stearic and arachidonic fatty acids are
the major acyl species in PIP, PIP2, and PIP3, other fatty acid combinations are also found. The role of these
different molecular species is still unknown, but it is important to quantify these different molecules and
their potential changes during cell stimulation to better characterize this emerging field. Here, we describe a
sensitive high-performance liquid chromatography–mass spectrometry method that we used for the first
time to profile the changes in phosphoinositide molecular species (summed fatty acyl chain profiles) in
human and mouse platelets under resting conditions and following stimulation. This method can be applied
to other hematopoietic primary cells isolated from human or experimental animal models.
Key words Phosphoinositides, Molecular species, Fatty acyl chains, LC-MS, Platelets
1 Introduction
Phosphoinositides (PIs) are a family of seven lipids characterized by

a highly active metabolism under the control of specific PI kinases,
phosphatases, and phospholipases. These low abundant lipids are
known to interact with a variety of proteins and to organize, in a
spatiotemporal manner, the formation of functional protein com-
plexes playing an essential role in cell regulation [1, 2]. Mutations
in the genes encoding for several PI kinases or PI phosphatases that
control the reversible phosphorylation of the inositol head group of
these lipids are directly responsible for human diseases [2]. We
39
40 Gaëtan Chicanne et al.
realize now that there is a more important molecular diversity of PIs

than previously thought, due to their fatty acid composition [3–5]
(Fig. 1). While important changes to the acyl composition of PIs
occur through cycles of deacylation/reacylation under the control
of acyltransferases [6], they may also be modulated by diet. How-
ever, the role of the fatty acid composition of PIs remains poorly
understood and is an emerging research field.
The aim of this chapter is to describe a LC-MS method to
perform a relative quantification of PI, PIP, PIP2, and PIP3 and
their fatty acid composition (number of total carbon and unsatu-
rated bonds corresponding to the summed fatty acyl chains) in
platelets, a human primary blood cell specialized in hemostasis
(Fig. 2). Platelets are anucleated cells, which are the first line of
defense against hemorrhages. The PI metabolism is particularly
active in blood platelets and plays a key role in several aspects of
their activation process [7]. So far, different approaches have been
developed to quantify PIs, including radioactive myo-inositol or
32
Pi cell metabolic labeling followed by HPLC analysis, mass
assay using specific recombinant kinases, and more recently mass
spectrometry (MS)-based measurements [8]. A first approach using
ESI-MS to examine PIs in cell extracts was provided by Wenk et al.
[9], but this method did not allow the measurement of PIP3. Milne
et al. [10] directly analyzed PIs in lipid extracts, but the sensitivity
of the method was limited. Clark et al. published a method [11, 12]
that was more sensitive due to the derivatization of phosphate
(methylation) before MS analysis. Although MS has the great
Fig. 1 Schematic structure of phosphoinositides. Note that the positions 3, 4, and

5 of the inositol head group can be phosphorylated (labeled “P”) by specific
kinases to produce seven different phosphoinositides. Importantly, while stearic/
arachidonic acids are frequently found in polyphosphoinositides, other fatty
acids are also found and may contribute to the diverse functions of these
lipids. Our results [5] show that there is an important diversity in the fatty acid
composition of phosphatidylinositol in human and mouse platelets and that
PI-kinases have a marked preference for specific fatty acyl composition to
produce phosphorylated derivatives of phosphatidylinositol
Profiling of Platelet Phosphoinositides by LC-MS Method 41
Fig. 2 Schematic representation of the LC-MS method. Schematic representation of the LC-MS method used
to analyze phosphoinositides molecular species from resting or activated platelets
advantage of bypassing radiolabeling and resolving the fatty acyl

composition in small amounts of samples [11], quantification of the
different regioisomers of PIP and PIP2 remains challenging [13].
2 Materials
All solvents have to be prepared with ultrapure water (18 MΩ cm at

25 C) and analytical grade reagents. All solvents used are at least of
HPLC grade. Prepare and store all reagents at room temperature
(unless indicated otherwise). All work must be carried out in a
chemical fume hood to avoid inhalation of solvents and reagents.

Trimethylsilyl diazomethane (TMS-D) is toxic by inhalation and
must be handled with particular care; safety procedures should be
considered in case of accidental breakage of the reagent bottle.
Safety glasses, lab coat, and gloves resistant to chemicals must be
worn to protect skin and eyes.
2.1 Platelet 1. Anesthetic solution: mix 1 mL of xylazine 2% and 0.5 mL of

Preparation ketamine 10% with 8.5 mL of NaCl 0.9%.
2. Acid–citrate–dextrose (ACD) anticoagulant solution: dissolve
25 g of trisodium citrate dihydrate, 14 g of citric acid mono-
hydrate, and 20 g of anhydrous D(+)glucose in a final volume of
1 L of distilled water.
3. Mouse HEPES buffer 10: dissolve 1.19 g of HEPES
(4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid, N-
(2-hydroxyethyl)piperazine-N0 -(2-ethanesulfonic acid) sodium
salt), 8 g of NaCl, 0.149 g of KCl, 1 g of NaHCO3, 0.04 g of
NaH2PO4, and 0.2 g of MgCl2·6H2O in a final volume of
100 mL of distilled water.
4. Buffer A: dissolve 27 mg of anhydrous D(+)glucose and
105 mg of BSA in 25 mL of distilled water. Add 3 mL of
mouse HEPES buffer 10, mix, and adjust the pH to 6.7
with HCl. Make up to 30 mL with distilled water and filter
through a 0.45-μm filter.
5. 1 mM prostaglandin I2 sodium salt (PGI2): dissolve 1 mg of
PGI2 in 2.34 mL of 50 mM Tris–HCl, pH 9.36. Freeze at
20 C in 10-μL aliquots.
6. Buffer B: dissolve 27 mg of anhydrous D(+)glucose in 20 mL of
distilled water. Add 3 mL of mouse HEPES buffer 10 and
3 mL of 20 mM CaCl2. Mix and adjust the pH to 7.38 with
HCl. Make up to 30 mL with distilled water and filter through
a 0.45-μm filter.
7. 20 IU/mL apyrase: dissolve 100 IU of apyrase in 0.5 mL of
distilled water. Freeze at 20 C in 20-μL aliquots. Thaw
1 aliquot and add 180 μL of distilled water.
8. Human HEPES buffer 10: dissolve 2.383 g of HEPES
(4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid, N-
(2-hydroxyethyl)piperazine-N0 -(2-ethanesulfonic acid) sodium
salt), 8.181 g of NaCl, 0.373 g of KCl, 0.68 g of KH2PO4, and
0.246 g of MgSO4·7H2O in a final volume of 100 mL of
distilled water.
9. Buffer C: dissolve 45 mg of anhydrous D(+)glucose and
human HEPES buffer 10, mix, and adjust the pH to 6.7
with HCl. Make up to 50 mL with distilled water and filter
through a 0.45-μm filter.
10. Buffer D: dissolve 45 mg of anhydrous D(+)glucose and

human HEPES buffer 10 and 5 mL of 10 mM CaCl2. Mix
and adjust the pH to 7.38 with HCl. Make up to 50 mL with
distilled water and filter through a 0.45-μm filter.
2.2 Platelet 1. 100 IU/mL thrombin: dissolve 100 IU of thrombin from

Activation bovine plasma in 1 mL of distilled water. Freeze at 80 C in
10-μL aliquots.
2. 400 μg/mL collagen-related peptide: dilute 4 mg/mL
collagen-related peptide using water containing 0.06% acetic
acid. 4.25 μl per stimulation will be needed. Use immediately.
3. Quench solution: mix 48.4 mL of methanol with 24.2 mL of
chloroform and 2.355 mL of 1 N HCl.
4. Cooled 1 N HCl: mix 8.3 mL of 37% hydrochloric acid (12 N)
with 91.7 mL of H2O (see Note 1).
2.3 Lipid Preparation 1. Internal standards: 17:0-14:1 PI (Avanti Polar Lipids), 17:0-
20:4 PI4P (Avanti Polar Lipids), 17:0-20:4 PI(4,5)P2 (Avanti
Polar Lipids), and 17:0-20:4 PI(3,4,5)P3 (Avanti Polar Lipids)
serve as the internal standards.
2. Internal standards stock solutions: dissolve each standard pow-
der in a glass vial to 1 μg/μL using methanol and store at
20 C. Dilute each standard in a glass vial to 100 ng/μL
using methanol and store at 80 C.
3. Internal standards mix solution PI/PIP/PIP2: mix 10 μL of
17:0-14:1 PI at 100 ng/μL, 17:0-20:4 PI4P at 100 ng/μL,
and 17:0-20:4 PI(4,5)P2 at 100 ng/μL with 970 μL methanol
to obtain a final concentration of 1 ng/μL per standard (see
Note 2).
4. Internal standard solution PIP3: mix 10 μL of 17:0-20:4 PI
(3,4,5)P3 at 100 ng/μL with 990 μL of methanol to obtain a
final concentration of 1 ng/μL (see Note 2).
5. 2 N HCl: mix 16.7 mL of 37% hydrochloric acid (12 N) with
83.3 mL of H2O (see Note 1).
6. Pre-derivatization solution: mix 120 mL of methanol with
240 mL of chloroform and 90 mL of 0.01 N HCl (see Note 3).
2.4 Sample 1. Neutralizing solution: mix 50 mL of methanol with 12.5 mL of

Derivatization acetic acid.
2. Post-derivatization solution: mix 120 mL of methanol with
240 mL of chloroform and 90 mL of H2O (see Note 4).
3. 2 M trimethylsilyl diazomethane (TMS-D) in hexane (see
Note 5).
4. Syringe and needle: use a 1-mL syringe with a 25G 5/800

needle.
5. Glass tube: use 6-mL borosilicate glass, round-bottom, rimless,
12.25 75 0.80 mm.
6. Methanol:H2O (9:1, v/v): mix 90 mL of methanol with 10 mL
of H2O.
7. Glass screw top vials: 2 mL, 32 12 mm, 9 mm clear glass
screw thread vials.
8. Glass inserts for screw top vials: 0.1-mL micro insert,
29 5.7 mm, clear glass.
9. Caps for screw top vials: 9 mm combination seal PP short
thread caps.
2.5 Liquid Chromato- 1. LC-MS triple quadrupole (TQ): for example, a high-
graphy–Mass performance liquid chromatography (HPLC) system Agilent
Spectrometry 1290 Infinity equipped with a thermostated autosampler, a
binary pump, and a column oven coupled to TQ mass spec-
trometer Agilent 6460 (Agilent Technologies, USA) equipped
with electrospray ionization (ESI) can be used (see Note 6).
2. Liquid chromatography column: use Acquity UPLC BEH300-
C4 (100 1.0 mm, 1.7 μm) (Waters) maintained at 22 C (see
Note 7).
3. Buffer A: mix 1 mL of formic acid with 999 mL of water.
4. Buffer B: mix 1 mL of formic acid with 999 mL of HPLC-
grade acetonitrile.
3 Methods
3.1 Washed Mouse 1. Anesthetize 8-week-old mice (~20 g) by injection of 100 μL of

Platelet Preparation anesthetic (see Note 8).
2. Collect whole blood from the inferior vena cava of anesthetized
mice using a 25-gauge needle mounted on a syringe containing
100 μL of ACD (1 volume of ACD for 9 volumes of blood) (see
Notes 9 and 10).
3. Transfer in a 2-mL tube and add 1 volume of buffer A.
4. Up-and-down homogenization by hand.
5. Centrifuge at 300 g for 4 min at room temperature (low
brake).
6. Collect the platelet-rich plasma (PRP, upper phase) in a new
2-mL tube and add 0.5 μL of 1 mM PGI2 per 1 mL of PRP
(final concentration of 500 nM).
7. Up-and-down homogenization by hand.

brake).
9. Discard the supernatant.
10. Gently resuspend the platelet pellet in 500 μL of buffer B in the
presence of 0.02 IU/mL of apyrase.
11. Count platelets using a hematological cell counter (see Note
11) and adjust the volume to 1 108 platelets in 170 μL using
buffer B in the presence of 0.02 IU/mL of apyrase.
12. Rest for 45 min at 37 C.
3.2 Washed Human 1. Collect 25 mL of blood from healthy human donors according
Platelet Preparation to local and national regulations into a 50-mL centrifuge tube
containing 1 volume of ACD for 6 volumes of blood.
brake).
3. Collect the platelet-rich plasma (PRP, upper phase) in a new
50-mL centrifugation tube and add 0.5 μL of 1 mM PGI2 per
1 mL of PRP (final concentration of 500 nM) and 2 μL of
5000 U/mL heparin per 1 mL of PRP (final concentration of
10 U/mL).
brake).
5. Discard the supernatant.
6. Gently resuspend the platelet pellet in 5 mL of human buffer A
containing 2.5 μL of 1 mM PGI2.
7. Add 15 mL of human buffer A containing 7.5 μL of
1 mM PGI2.
8. After 10 min of incubation at 37 C, count platelets using a
hematological cell counter.
brake).
10. Gently resuspend the platelet pellet in human buffer B contain-
ing 0.04 IU/mL of apyrase at 1 108 platelets/170 μL (see
Note 12).
11. Rest for 30 min at 37 C.
3.3 Platelet 1. Transfer 1 108 washed blood platelets (170 μL) in a 2-mL
Activation tube (see Note 13).
2. Stimulate platelets by addition of physiological agonists (for
instance, thrombin at 0.5 IU/mL or collagen-related peptide
at 10 μg/mL) at 37 C during appropriate times (1–15 min, for
instance) [5].
3. In the case of direct sample processing, add 750 μL of quench

solution (see Note 14) and proceed to Subheading 3.4.
4. In the case of a step of freezing platelets, add 500 μL of cooled
1 N HCl.
5. Centrifuge the sample for 30 s at 10,000 g at 4 C.
6. Remove the supernatant and snap freeze the pellet in liquid
nitrogen and store it at 80 C (see Notes 15 and 16).
7. Resuspend in 170 μL of H2O.
8. Sonicate using a probe sonicator (130 W) in an ice bath (7 s,
output 30%).
9. Add 750 μL of quench solution.
3.4 Lipid Preparation 1. Vigorously mix 3 by manual agitation, each 10 s.

2. Add 20 μL of internal standards mix solution PI/PIP/PIP2
and 20 μL of internal standard solution PIP3 (20 ng of each)
(see Note 17).
3. Vigorously mix 3 by manual agitation, each 10 s.
4. Leave to stand 5 min at room temperature.
5. Add 725 μL of CHCl3 (see Note 18) and then 170 μL of
2 N HCl.
7. Centrifuge the sample for 3 min at 1500 g at room tempera-
ture (see Note 19).
8. Collect the lower organic phase (see Note 20) and transfer into
new 2-mL tubes containing 708 μL of the upper aqueous phase
from pre-derivatization solution (see Note 18).
10. Centrifuge the sample for 3 min at 1500 g at room
temperature.
11. Carefully collect the lower organic phase into a fresh 2-mL
tube (see Note 20).
3.5 Sample 1. Transfer the required amount of TMS-D in a 6-mL glass tube
Derivatization using a syringe and needle (see Note 21).
2. Wash the syringe and the needle by several flushes in neutraliz-
ing solution.
3. Add 50 μL of TMS-D to each sample and mix by pipetting up
and down (see Note 22).
4. Wash tips by several flushes in neutralizing solution.
5. Keep the tubes opened in the fume hood (nitrogen is generated
by the reaction) and incubate for a maximum of 10 min at
room temperature.
6. Add slowly 6 μL of glacial acetic acid to each tube (this destroys

excess reagent with the release of nitrogen gas).
7. Mix by pipetting from top to bottom and keep the tubes
opened (see Note 23).
8. Add 700 μL of the upper aqueous phase from post-
derivatization wash solution (see Note 18).
10. Centrifuge for 3 min at 1500 g at room temperature.
11. Collect the lower organic phase and transfer into new 2-mL
tubes containing 700 μL of the upper aqueous phase from
post-derivatization wash solution (see Note 18).
13. Centrifuge for 3 min at 1500 g at room temperature.
14. Collect the lower organic phase in a 6-mL glass tube.
15. Add 100 μL of methanol:H2O (9:1, v/v).
16. Dry samples under a stream of N2 at room temperature until
almost dry (see Note 24).
17. Immediately, dissolve samples in 80 μL of methanol.
18. Sonicate for 30 s in a bath sonicator (100 W).
19. Add 20 μL of H2O.
20. Sonicate for 30 s in a bath sonicator (100 W).
21. Transfer in a glass vial with insert.
22. Dry samples under a stream of N2 at room temperature until
almost dry (see Note 24).
23. Add 20 μL of methanol (see Note 25).
3.6 Relative 1. The source parameters need to be optimized on each instru-

Quantification Using ment. On the Agilent 6460, the parameters we used are as
MS follows: source temperature ¼ 250 C, nebulizer gas (nitrogen)
flow rate ¼ 8 L/min, sheath gas temperature ¼ 150 C, sheath
gas (nitrogen) flow rate ¼ 3 L/min, and the spray voltage:
+4000 V. The collision energy optimums for derived PIPs were
30 eV (see Note 26).
2. Establish the separation gradient as follows: 45% of buffer B at
0 min, 100% of buffer B at 5 min, 100% of buffer B at 8 min,
45% of buffer B at 9 min, and 45% of buffer B at 10 min with a
flow rate of 0.1 mL/min.
3. Place the samples in an autosampler at 8 C and inject with an
injection volume of 10 μL.
4. Perform electrospray ionization (ESI) in positive ion mode.
5. Perform analyses in Selected Reaction Monitoring detection
mode (SRM) using nitrogen as collision gas with the transi-
tions, as shown in Table 1 (see Note 27).
Table 1
Single reaction monitoring transition
Molecules Precursor ion (m/z) Product ion (m/z)

PI
PI 14:1-17:0 (ISTD) 809.5 535.5
PI 32:0 825.5 551.5
PI 32:1 823.5 549.5
PI 32:2 821.5 547.5
PI 34:0 853.5 579.5
PI 34:1 851.5 577.5
PI 34:2 849.5 575.5
PI 36:0 881.5 607.5
PI 36:1 879.5 605.5
PI 36:2 877.5 603.5
PI 36:3 875.5 601.5
PI 36:4 873.5 599.5
PI 38:0 909.5 635.5
PI 38:1 907.5 633.5
PI 38:2 905.5 631.5
PI 38:3 903.5 629.5
PI 38:4 901.5 627.5
PI 38:5 899.5 625.5
PIP
PIP 17:0–20:4 (ISTD) 995.5 613.5
PIP 32:0 933.5 551.5
PIP 32:1 931.5 549.5
PIP 32:2 929.5 547.5
PIP 34:0 961.5 579.5
PIP 34:1 959.5 577.5
PIP 34:2 957.5 575.5
PIP 36:0 989.5 607.5
PIP 36:1 987.5 605.5
PIP 36:2 985.5 603.5
PIP 36:3 983.5 601.5
(continued)
Table 1
(continued)

PIP 36:4 981.5 599.5
PIP 38:0 1017.5 635.5
PIP 38:1 1015.5 633.5
PIP 38:2 1013.5 631.5
PIP 38:3 1011.5 629.5
PIP 38:4 1009.5 627.5
PIP 38:5 1007.5 625.5
PIP2
PIP2 17:0-20:4 (ISTD) 1103.5 613.5
PIP2 32:0 1041.5 551.5
PIP2 32:1 1039.5 549.5
PIP2 32:2 1037.5 547.5
PIP2 34:0 1069.5 579.5
PIP2 34:1 1067.5 577.5
PIP2 34:2 1065.5 575.5
PIP2 36:0 1097.5 607.5
PIP2 36:1 1095.5 605.5
PIP2 36:2 1093.5 603.5
PIP236:3 1091.5 601.5
PIP2 36:4 1089.5 599.5
PIP2 38:0 1125.5 635.5
PIP2 38:1 1123.5 633.5
PIP2 38:2 1121.5 631.5
PIP2 38:3 1119.5 629.5
PIP2 38:4 1117.5 627.5
PIP2 38:5 1115.5 625.5
PIP3
PIP3 17:0-20:4 (ISTD) 1211.5 613.5
PIP3 32:0 1149.5 551.5
PIP3 32:1 1147.5 549.5
PIP3 32:2 1145.5 547.5
(continued)
Table 1
(continued)

PIP3 34:0 1177.5 579.5
PIP3 34:1 1175.5 577.5
PIP3 34:2 1173.5 575.5
PIP3 36:0 1205.5 607.5
PIP3 36:1 1203.5 605.5
PIP3 36:2 1201.5 603.5
PIP3 36:3 1199.5 601.5
PIP3 36:4 1197.5 599.5
PIP3 38:0 1233.5 635.5
PIP3 38:1 1231.5 633.5
PIP3 38:2 1229.5 631.5
PIP338:3 1227.5 629.5
PIP3 38:4 1225.5 627.5
PIP3 38:5 1223.5 625.5
The table presents all the single reaction monitoring transition used of each molecule with the precursor ion and product
ion of each transition. ISTD: Internal Standard
6. Finally, perform peak detection, integration, and quantitative

analysis using MassHunter QqQ Quantitative analysis software
(Agilent Technologies) and Microsoft Excel software. The rel-
ative quantification of each species is given by the ratio between
areas of the molecule of interest versus the area of the internal
standard of the corresponding family. This value is then nor-
malized (divided) by the amount of platelets.
4 Notes
1. HCl solution should be prepared in a hood with HCl slowly

being added to H2O.
2. Store at 20 C, and prior to each use, incubate at room
temperature for 10 min and sonicate for 30 s in a bath
sonicator.
3. After mixing, leave to stand at room temperature. Two phases
will be formed, and use the upper phase, also called the upper
aqueous phase, from pre-derivatization solution.
4. After mixing, leave to stand at room temperature. Two phases

will be formed, and use the upper phase, also called the upper
aqueous phase, from post-derivatization solution.
5. Store the solution under nitrogen and keep it under the chem-
ical fume hood.
6. The analysis has to be run on any triple quadrupole detector
coupled to HPLC.
7. The method described is optimized on a BEH300-C4 column,
but it can be adapted on other C4 columns.
8. Local and national regulations for animal care, anesthesia, and
removal of blood should be followed at all times.
9. The quantity of blood obtained from a mouse weighing 20 g is
about 1 mL.
10. Use one 2-mL tube per mouse and do not pool blood from
several mouses.
11. A mouse blood volume of 1 mL can provide 500 μL of washed
platelets at 1 109 platelets per mL.
12. A human blood volume of 25 mL can provide between
10 108 and 20 108 platelets.
13. Use a 2-mL safe-lock polypropylene tube to avoid leaks.
14. If you add quench solution, you must immediately extract and
derivatize the sample.
15. Liquid nitrogen should be manipulated under natural or
mechanical ventilation to prevent oxygen-deficient atmo-
spheres below 19.5%. Use loose-fitting thermal insulated or
leather gloves for the skin protection and tongs to immerge
and withdraw objects in liquid nitrogen. Be careful to minimize
boiling and splashing by handling the liquid nitrogen slowly.
16. Store the sample at 80 C for a maximum period of 1 month.
17. It is recommended to make two additional control samples:
(1) with all the internal standards and without biological sam-
ple and (2) without internal standards and without biological
sample.
18. Before pipetting CHCl3 or organic phase, equilibrate the tips
with CHCl3 or organic phase by pipetting up and down before
to avoid dripping.
19. The aqueous and organic phases separate with a protein band
sitting at the interface. The organic phase containing the lipids
is at the bottom.
20. To collect the lower phase, use 1-mL tips and expel the major-
ity of the air while crossing the aqueous phase as well as the
protein interfering. Be careful not to take the aqueous phase
and protein.
21. TMS-D is toxic by inhalation and can explosively release N2

gas. Personal protective equipment is required (safety glasses,
gloves, and using a fume hood). Clean all tips, syringes, and
needles in contact with TMS-D by aspirating and flushing in
neutralizing solution. For more details, please read its safety
data sheet.
22. The addition of TMS-D gives a yellow solution.
23. The addition of glacial acetic acid removes the sample’s yellow
color and generates bubbles.
24. Stop drying until ~10 μL remaining, and it will take around
30 min. Overdrying can cause loss of lipid phosphate ester
migration and it is difficult to resuspend lipids.
25. Samples can be stored at +4 C for a few hours or 80 C for
longer.
26. The source parameters described are optimized for ESI source
Agilent 6460 triple quadrupole MS; they will need to be
adapted for another system.
27. Make sure that the dwell time of your SRM method is adapted
to the number of transitions recorded. To improve the sensi-
tivity, the method could be focused on one particular
phosphoinositide.
Acknowledgments
We thank all the members of the Inserm U1048-Team 11 and

MetaToul-Lipidomic facility. This work was supported by
INSERM, by MetaboHUB (ANR-11-INBS-0010), and by the
Fondation pour la Recherche Médicale (DEQ20170336737). BP
is a scholar of the Institut Universitaire de France.
References
1. Balla T (2013) Phosphoinositides: tiny lipids BioEssays 39(12). https://doi.org/10.1002/
with giant impact on cell regulation. Physiol bies.201700121
Rev 93(3):1019–1137. https://doi.org/10. 5. Mujalli A, Chicanne G, Bertrand-Michel J et al
1152/physrev.00028.2012 (2018) Profiling of phosphoinositide molecu-
2. Viaud J, Mansour R, Antkowiak A et al (2016) lar species in human and mouse platelets iden-
Phosphoinositides: important lipids in the tifies new species increasing following
coordination of cell dynamics. Biochimie stimulation. Biochim Biophys Acta Mol Cell
125:250–258. https://doi.org/10.1016/j.bio Biol Lipids 1863(9):1121–1131. https://doi.
chi.2015.09.005 org/10.1016/j.bbalip.2018.06.009
3. Hammond GR, Balla T (2014) A tail of new 6. Bone LN, Dayam RM, Lee M et al (2017) The
lipids. EMBO J 33(19):2140–2141. https:// acyltransferase LYCAT controls specific phos-
doi.org/10.15252/embj.201489773 phoinositides and related membrane traffic.
4. Choy CH, Han BK, Botelho RJ (2017) Phos- Mol Biol Cell 28(1):161–172. https://doi.
phoinositide diversity, distribution, and effec- org/10.1091/mbc.E16-09-0668
tor function: stepping out of the box.
7. Min SH, Abrams CS (2013) Regulation of 46(8):1796–1802. https://doi.org/10.1194/

platelet plug formation by phosphoinositide jlr.D500010-JLR200
metabolism. Blood 122(8):1358–1365. 11. Clark J, Anderson KE, Juvin V et al (2011)
https://doi.org/10.1182/blood-2013-05- Quantification of PtdInsP3 molecular species
427716 in cells and tissues by mass spectrometry. Nat
8. Wakelam MJ, Clark J (2011) Methods for ana- Methods 8(3):267–272. https://doi.org/10.
lyzing phosphoinositides using mass spectrom- 1038/nmeth.1564
etry. Biochim Biophys Acta 1811 12. Kielkowska A, Niewczas I, Anderson KE et al
(11):758–762. https://doi.org/10.1016/j. (2014) A new approach to measuring phos-
bbalip.2011.09.004 phoinositides in cells by mass spectrometry.
9. Wenk MR, Lucast L, Di Paolo G et al (2003) Adv Biol Regul 54:131–141. https://doi.
Phosphoinositide profiling in complex lipid org/10.1016/j.jbior.2013.09.001
mixtures using electrospray ionization mass 13. Wang C, Palavicini JP, Wang M et al (2016)
spectrometry. Nat Biotechnol 21(7):813–817. Comprehensive and quantitative analysis of
https://doi.org/10.1038/nbt837 Polyphosphoinositide species by shotgun Lipi-
10. Milne SB, Ivanova PT, DeCamp D et al (2005) domics revealed their alterations in db/db
A targeted mass spectrometric analysis of phos- mouse brain. Anal Chem 88
phatidylinositol phosphate species. J Lipid Res (24):12137–12144. https://doi.org/10.
1021/acs.analchem.6b02947
Chapter 4
Quantification of Genetically Encoded Lipid Biosensors

Rachel C. Wills, Jonathan Pacheco, and Gerald R. V. Hammond
Abstract
Lipids, like phosphoinositides, can be visualized in living cells in real time using genetically encoded
biosensors and fluorescence microscopy. Sensor localization can be quantified by determining the fluores-
cence intensity of each fluorophore. Enrichment of lipids at membranes can be determined by generating
and applying an organelle-specific binary mask. In this chapter, we provide a detailed list of reagents and
methods to visualize and quantify relative lipid levels. Applying this approach, changes in lipid levels can be
assessed in cases when lipid metabolizing enzymes are mutated or otherwise altered.
Key words Biosensors, Lipid-binding domain, Confocal microscopy, TIRF microscopy
1 Introduction
1.1 Phospholipids Lipids are central molecules in cell biology and serve critical roles in
energy storage and membrane formation and as regulators of vari-
ous cellular activities [1]. There are a variety of lipid species that
provide distinct functional properties to membranes [1, 2]. Phos-
phatidylinositol (PtdIns) is a unique phospholipid possessing a myo-
inositol headgroup, which can be reversibly phosphorylated in up
to three different positions, generating seven different polypho-
sphoinositides (PPIn) [3]. The specific lipid makeup of membranes
provides unique identities to various organelles, and this results in
exquisite spatial and temporal regulation of cellular processes
[3, 4]. As such, a great deal of effort has gone into understanding
their function, cellular distribution, and metabolism [5]. A variety
of methods have emerged with which to quantify lipid levels and to
observe the enzymes responsible for their metabolism [3, 6–10].
1.2 Detection Lipid species have been detected and quantified in cells primarily via
of Lipids in Cells biochemical approaches. These include mass spectroscopy (MS)
[11–13], high-performance liquid chromatography (HPLC) [14],
and thin-layer chromatography (TLC) [15–17]. After the chemical
extraction of lipids from cells, these techniques can be used to
55
56 Rachel C. Wills et al.
determine the quantity of various lipid species. However, these

measurements are obtained from a population of cells at a single
time point and therefore do not provide dynamic, temporal infor-
mation and provide a measurement averaged across a population.
Microscopy techniques, on the other hand, allow for the real-time
visualization of lipid dynamics in single cells. These approaches
include total internal reflection fluorescence microscopy
(TIRFM), confocal microscopy, and super-resolution imaging
techniques. Additionally, microscopy allows for the observation of
single-cell dynamics over time, thus providing both spatial and
temporal resolution [7].
Using microscopy to visualize lipids presents a set of challenges.
Visualizing proteins is relatively easy and can be done by generating
fusions of the protein of interest with fluorescent proteins, such as
GFP or mCherry [8, 10, 18–21]. Utilizing fluorescently tagged
proteins that efficiently and selectively bind lipids headgroups has
proven effective in the visualization of specific lipid species [20–
22]. These fusion proteins have been termed biosensors. This
chapter will focus exclusively on qualitatively and quantitatively
observing lipids via these genetically encoded lipid biosensors.
Our group has come up with three crucial criteria that should
be considered when determining the validity of a biosensor
[8]. First, is the probe specific for the target lipid? Second, is the
biosensor’s localization dependent on that lipid? And third, if
dependent, is the lipid alone sufficient to localize the biosensor?
Table 1 provides an updated list of lipid-binding domains and
illustrates whether each sensor meets the previously mentioned
criteria [8, 23]. Biosensors, however, are not to be used as a binary
means of determining lipid localization or definitive lipid amounts.
The use of biosensors allows for the visualization of lipid local-
ization and relative abundance in living cells and in real time. The
method of performing these experiments involves the transfection
of plasmid DNA, containing the biosensor of interest, into cells.
Cells can then be visualized so as to determine the localization of
lipids without manipulation, or alternatively, lipid levels can be
manipulated using enzymes, growth factors, or other chemical
treatments. The use of biosensors for the detection and observation
of lipids presents a number of caveats [8]. These have been dis-
cussed in great detail in recent reviews [8, 23].
1.3 Quantification The use and visualization of lipid biosensors via microscopy pro-
of Biosensors in Cells vides a real-time single-cell approach to monitor changes in fluo-
rescence intensity. Images themselves are not able to be used as
statistical data, but that does not mean the images cannot be used
to obtain this type of quantitative data. Measuring the fluorescent
intensity of biosensor directly corresponds to the amount of sensor
bound and can indicate the relative amount of lipid in a membrane.
For example, using a PtdIns (4,5)P2 probe would result in a high
Table 1
Genetically encoded lipid biosensors and whether or not each sensor meets criteria that define an accurate biosensor (see Subheading 1 for details and references)
Cellular localization
Lipid
Lipid Biosensor Intracellular localization Affinity Lipid specific? dependent? Lipid sufficient? References
Chol D4-PFO and PM (exoplasmic > cytosolic) 2–30 mol% ✔ ✔ ? [26–28]

Muts
SM Lysenin PM (exoplasmic) KD ~ 5 nM ✔ ✔ ? [29–31]
PA PASS/2xPABD PM, ER ? ✘—Binds PtdIns(4,5)P2 weakly ✔ ? [32, 33]
PS C2-lactadherin PM, endosomes, TGN KD ~ 0.5 μM ✔ ✔ ✔ [34–36]
DAG C1ab-PKCE Golgi, ER, NE KD ~ 10 nM ✔ ✔ ? [37, 38]
C1ab-PDK1 Golgi Ki ~ 0.2 μM ✔ ✔ ✔ [39, 40]
PtdIns ? Golgi, lysosome, endosome, NA NA NA NA [41, 42]

mitochondria, peroxisome, ER, PM
PtdIns3P FYVE-Hrsx2 EE KD ~ 2.5 μM ✔ ✔ ? [43–46]
FYVE-EEA1 EE KD ~ 45 nM ✔ ✔ ? [43, 44, 47]
PX-p40phox EE KD ~ 5 μM ✔ ✔ ? [48–50]
PtdIns4P P4M-SidM PM, Golgi, endosomes KD ~ 1 μM or ✔ ✔ ✔ [22, 51, 52]

(P4Mx1) ~18.2 nM FL
P4M-SidMx2 PM, Golgi, endosomes KD < 1 ✔ ✔ ✔ [22, 53]

(P4Mx2)
P4C-SidC (P4C) PM, Golgi, endosomes KD ~ 250 nM ✔ ✔ ✔ [54–56]
N-PH-ORP5, PM KD ~ 5 μM for ✘—Binds PtdIns(4,5)P2 and PtdIns4P ✔ ✘—Requires [57–59]

N-PH-ORP8 PtdIns (4,5)P2 PtdIns(4,5)P2
Golgi, PM KD ~ 2–4 μM ✘—Binds PtdIns(4,5)P2 ✔ [60–62]
(continued)
Table 1
(continued)
Cellular localization
Lipid
Lipid Biosensor Intracellular localization Affinity Lipid specific? dependent? Lipid sufficient? References
PH-OSBP, ✘—Requires
PH-FAPP1 Arf1
PtdIns5P 3xPDH (ING2) PM, nucleus ? ✘—Binds PtdIns3P ✔ ✔ [63, 64]
PtdIns PH-TAPP1-CT PM KD ~ 80 nM ✔ ✔ ✔ [65–69]

(3,4)
P2
TAPP1 cPHx1 PM KD ~ 80 nM ✔ ✔ ✔ [67]
TAPP1 cPHx2 PM KD < 1 ✔ ✔ ✔ [67, 70–73]
TAPP1 cPHx3 PM KD < 2 ✔ ✔ ✔ [67, 70–73]
PtdIns MLN1-Nx2 EE, LE, lysosomes KD ~ 5.6 μM ✔ ✔/✘ ✔/✘ [74, 75]
(3,5)
P2
PtdIns ENTH/ANTH PM KD ~ 5 μM ✘—Binds PtdIns(3,4,5)P3 ✔ ? [76–78]

(4,5)
P2
PH-PLCd1 PM KD ~ 2 μM ✘—Binds Ins(1,4,5)P3 ~ 20-fold more ✔ ✔ [79–84]

tightly than PtdIns(4,5)P2
PH-PLCd4 PM KD > PH-PLCd1 ✘—Binds Ins(1,4,5)P3 ✔ ✔ [8, 85]
TubbyC PM KD > PH-PLCd1 ✘—Binds PtdIns(3,4)P2 and PtdIns ✔ ✔ [8, 86–88]

(3,4,5)P3
TubbyCR332H PM KD > tubby ✘—Binds PtdIns(3,4)P2 and PtdIns ✔ ✔ [86]

(3,4,5)P3
PtdIns PH-Akt PM KD ~ 590 nM ✘—Binds PtdIns(3,4)P2 and Ins(1,3,4,5) ✔ ? [69, 89, 90]
(3,4,5) P4
P3
PH-Btk PM KD ~ 80 nM ✘—Binds Ins(1,3,4,5)P4 ✔ ? [69, 91–94]

2G
PH-GRP1 PM KD ~ 170 nM ✘—Binds Ins(1,3,4,5)P4 ✔ ✘—Binds [69, 95–101]
PH-ARNO2G Arf/Arl
PH-ARNO2G- PM KD ~ 170 nM ✘—Binds Ins(1,3,4,5)P4 ✔ ✔ [69, 73, 97,

I303E
x2 100]
level of fluorescence at the plasma membrane (PM), visualized by

either TIRFM or confocal microcopy. Using tools (see subsequent
chapters) to deplete PtdIns(4,5)P2 levels would result in a much
lower membrane-associated fluorescence.
The quantification of fluorescent images has been done in a
number of ways. Our lab performs analysis on data collected from
TIRFM and from confocal microscopy, depending on the experi-
ment being performed. TIRFM is a particularly efficient method of
measuring changes in lipid level at the PM due to selective illumi-
nation of the bottom most ~100 nm of the cell. Confocal micros-
copy has been utilized for assaying lipid levels at various membrane
compartments, including the PM.
For TIRFM analysis, the fluorescence intensity is measured
within the cell over time, and then the average pixel intensity of
each frame is normalized to the average pixel intensity before
treatment (Ft/Fpre). For confocal microscopy, the organelle of
interest is used to generate a mask in which the biosensor intensity
can be measured. Once the mask is generated, the biosensor pixel
intensity is measured in the mask relative to the intensity in the rest
of a region of interest (ROI) that encompasses the whole cell,
giving an organelle/cell ratio. Similarly, this can be done by mea-
suring the intensity within the mask relative to a secondary ROI
drawn in the cytoplasm yielding an organelle/cytosol ratio.
2 Materials
2.1 Tissue Culture 1. Glass-bottom culture dishes. 35 mm, #1.5 glass-bottom

dishes, with 200 mm glass aperture (In vitro Scientific).
2. Extracellular matrix for coating dishes. Fibronectin solution,
1 mg/mL in DDiH2O.
3. Extracellular matrix for coating dishes. Entactin–Collagen IV–
Laminin (ECL) Cell Attachment Matrix.
4. Phosphate-buffered saline (PBS). Purchased directly from
suppliers.
5. Cell dissociation reagent. Trypsin or TrypLE.
6. Complete cell culture media. DMEM (low glucose), 10% fetal
bovine serum (FBS), 100 U/mL penicillin, 100 U/mL strep-
tomycin, 1:1000 chemically defined lipid supplement (CDLS).
7. HeLa cells. The HeLa cervical cancer immortalized cell line can
be purchased from the American Type Culture Collection
(ATCC).
8. HEK293A cells. The HEK293A embryonic kidney immorta-
lized cell line can be purchased from ThermoFisher Scientific.
Lipid Biosensors and their Quantification 61
9. Transfection reagent. Cationic liposomal transfection reagent,

such as Lipofectamine 2000.
10. Reduced serum medium for transfection. Minimal essential
medium (MEM), such as Opti-MEM.
11. Plasmids for the expression of lipid sensors are readily available
at Addgene (https://www.addgene.org/). The different
options are listed in Table 1.
2.2 Microscopy 1. Complete HEPES Imaging Media (CHIM). Phenol red-free

DMEM (FluorBrite), 10% FBS, 2 mM GlutaMax, 25 mM
HEPES, pH 7.4, 1:1000 CDLS.
2. Plasma membrane stain. 1:5000 CellMask deep red.
3. Microscope. TIRF microscope, such as Nikon TiE with TIRF
illuminator arm with 100 1.45 NA plan-apochromatic
objective.
4. Microscope. Confocal microscope, such as Nikon TiE A1R
with 100 1.45 NA plan-apochromatic objective.
2.3 Image Analysis 1. Image processing software package. ImageJ or Fiji Image Anal-
ysis software [24, 25] (see Note 1).
2. Statistics software package. Excel or Prism 8—GraphPad
Prism.
3 Methods
3.1 Seeding Cells For each glass-bottom culture dish, coat glass with either FN or
ECL (see Note 2).
3.1.1 Coating Dishes 1. Dilute 5 μL of 1 mg/mL stock FN into 500 μL of DDiH2O

with FN and add to the glass inset at the bottom of the 35-mm dish.
2. Incubate for 30–60 min in an incubator.
3. Aspirate DDiH2O and FN and add 2 mL of complete DMEM,
and return to the incubator to continue to warm.
3.1.2 Coating Dishes 1. Dilute 10 μL of 1 mg/mL stock ECL into 500 μL of DMEM
with ECL and add to the glass inset at the bottom of the 35-mm dish.
2. Incubate for 60 min in an incubator.
3. Aspirate DMEM and ECL and add 2 mL of complete DMEM,
and return to the incubator to continue to warm.
3.2 Splitting 1. Using a sterile aspirating pipette, aspirate media from a conflu-
and Seeding Cells ent T75 flask (see Note 3).
2. Wash the adherent layer of cells once with 10 mL of PBS (see
Note 4).
Table 2
Suggested cell suspension volumes to seed cells for transfections at various timepoints
Volume of Ending Starting

suspension Days post-seeding Confluence at surface area surface area Resuspension
(μL) for transfection seeding (%) (cm2) (cm2) volume (mL)
80 2 12.5 9.6 75 5
160 1 25 9.6 75 5
320 0 (same day) 50 9.6 75 5
3. Add 1 mL of TrypLE and place the flask in an incubator for

1–5 min until cells have detached from the surface of the flask.
4. Add 4 mL of complete DMEM to neutralize the TrypLE and
to resuspend the cells in a total volume of 5 mL.
5. Seed cells in previously coated dishes at an appropriate volume
(see Note 5). For a next day transfection, seed 160 μL of cell
suspension, or seed cells at 25%.
6. Allow cells to settle and adhere for >1 h before transfecting.
Table 2 shows an overview of the cell suspension volumes
for variations in seeding before transfection.
3.3 Transfect Cells Per 35-mm glass-bottom dish:

1. Dilute 3.1 μL of Lipofectamine 2000 into 96.9 μL of Opti-
MEM in a 1.7-mL tube and mix. Generate a master mix for
multiple transfections.
2. Dilute up to 1 μg total plasmid DNA (10 μL total) into a total
of 100 μL Opti-MEM in a 1.7-mL tube and mix (see Note 6).
Again, generate a master mix for multiple transfections. Below
is an example of a single transfection mix for EGFP with
biosensors for PtdIns4P and for PtdIns(4,5)P2 (see Note 7).
0.25 μg EGFP-C1.
0.25 μg mCherry-TubbyR332H.
0.25 μg TagBFP2-P4Mx1 (see Note 8).
92.5 μL Opti-MEM (see Note 9).
3. Combine the 100-μL diluted Lipofectamine 2000 and the
100-μL diluted DNA, and mix gently by flicking. Incubate
for ~15 min at room temperature to allow for the formation
of lipid–DNA complexes.
4. Add the total 200-μL volume of lipid–DNA complex to the
pre-seeded 35-mm glass-bottom dish.
5. After ~4 h, aspirate the transfection mix off of the cells and

replace with 2 mL of fresh, prewarmed, complete DMEM.
6. For transient transfections, visualize cells 16–48 h after
transfection.
3.4 Microscope 1. Use appropriate excitation for each fluorescent protein. For
Setup and Imaging example, we use a fiber-coupled device with 405-nm (for
TagBFP2), 488-nm (for EGFP), 561-nm (for mCherry), and
3.4.1 TIRF Microscopy
640-nm (for iRFP) laser lines to excite the indicated fluorescent
protein.
2. Collect fluorescence emission with appropriate bandwidth fil-
ters. For example, blue and yellow/orange channels are imaged
through a dual-pass 420–480 nm and 570–620 nm Chroma
filter, respectively, and green and far/infrared are imaged
through a dual-pass 505–550 nm and 650–850 nm Chroma
filter, respectively.
3. Optimize the incidence angle of the illuminating beam to
greater than the critical angle, pixel binning, excitation inten-
sity, and/or camera exposure time to maximize the signal-to-
noise ratio in images and minimize phototoxicity (i.e., set the
minimal excitation power and exposure time) (see Note 10).
4. A motorized XY positioning stage is used to register up to
30 fields during the recording period (see Note 11).
5. Configure software acquisition to take one image over the
30 marked XY positions with no lag between images (see
Note 12).
3.4.2 Confocal 1. Use appropriate excitation for each fluorescent protein. For
Microscopy example, we use a fiber-coupled device with 405-nm (for
TagBFP2), 488-nm (for EGFP), 561-nm (for mCherry), and
640-nm (for iRFP) laser lines to excite the indicated fluorescent
protein.
2. Collect fluorescence emission with appropriate bandwidth fil-
ters. We use Chroma filters with 425–475 nm (for TagBFP2),
500–550 nm (for EGFP), 570–620 nm (for mCherry), and
663–737 nm (for iRFP). Differential interference contrast
(DIC) is used on a transmitted light channel for confocal
imaging.
3. Set confocal pinhole at diameter for maximal axial resolution
based on the longest wavelength channel (e.g., 1.2 Airy disc
size on our Nikon A1R).
4. Optimize confocal scan speed, line averaging, detector gain,
excitation intensity, and/or camera exposure time for maximal
signal-to-noise ratio in images with minimal phototoxicity (i.e.,
minimal excitation power and exposure time possible) (see

Note 13).
5. A motorized stage is used to register up to 30 fields during the
whole recording period (see Note 11).
6. Configure software acquisition to take one image over the
30 marked XY positions with no lag between images.
3.4.3 Imaging each 1. Prewarm CHIM media (2.5 mL per dish) (see Note 14).
35-mm Dish 2. Prepare CellMask (500 μL per dish at 1:5000 in prewarmed
CHIM, i.e., 0.1 μL per dish).
3. Aspirate media off dish and add 500 μL of prediluted CellMask
in CHIM. Allow to incubate for 3 min.
4. Aspirate the CellMask mix and replace with 2 mL of
fresh CHIM.
5. Apply immersion oil on the 100 plan-apochromatic, 1.45 NA
objective lens.
6. Mount the dish on the microscope.
7. Find the focus plane and look for healthy cells showing fluores-
cent signals in the channels required (see Note 15).
8. Save up to 30 positions for acquisition during the experiment.
9. Optimize the scan path and record the data to a file location of
choice.
3.5 Data Analysis The above microscopy experiment results in images at single data
points, which can be used to determine changes in lipid levels due
to chronic effects, such as expression changes in lipid metabolizing
enzymes. These differences can be observed at the plasma mem-
brane using TIRFM or other membranes using confocal micros-
copy. We note that the quantification of data from TIRF microcopy
is better suited for time-lapse imaging; refer to chap. 7 for a detailed
protocol. The following analysis best applies for images acquired by
confocal microscopy.
1. Import files into ImageJ (or its cell biology optimized build,
Fiji) using the Laboratory for Optical and Computational
Instrumentation (LOCI) Bio-Formats importer [24, 25]
(see Note 16).
2. Draw ROIs in the iRFP channel, around the CellMask-stained
PM, of each cell to be analyzed using the Freehand selection
tool (see Note 17).
3. Select the image(s) to be used to generate the mask and dupli-
cate this image three separate times (Image ! Duplicate).
Rename the starting image of I0 (Image ! Rename). Rename
the other three duplicates I1, I2, and I3.
4. Each of these images will be processed using a Gaussian blur

filter with a progressively larger radius determined by the point
spread function (PSF) of the channel used to generate the mask
(Process ! Filters ! Gaussian Blur).
PSF ¼ Wavelength=2 NA ¼ 640=2 1:45 ¼ 0:22 μm

5. The radius for the filter in I1 ¼ 0.22 μm, I2 ¼ 0.44 μm, and
I3 ¼ 0.66 μm.
6. I0–I3 can then be used to generate wavelets 1–3.
7. Using the image calculator function, generate the three wave-
lets by subtracting each smoothed image from the previous
image (Process ! Image Calculator).
I0–I1 ¼ W1
I1–I2 ¼ W2
I2–I3 ¼ W3
8. Determine the standard deviation (SD) of each resulting wave-
let and threshold each by 0.5*SD (Analyze ! Measure).
9. Generate the product of the three wavelets to generate a fil-
tered image (Process ! Image Calculator).
W 1 * W2 * W3
10. Convert this image to a binary Mask by converting to 32-bit
and dividing the image by itself (Image ! Type ! 32-bit !
Process ! Image Calculator). Set the brightness and contrast
on this image between 0 and 1 (Image ! Adjust ! Bright-
ness/Contrast ! Set).
11. Save this as Mask.
12. Measure and determine the mean intensity of all ROIs in each
channel (Analyze ! Measure).
13. Normalize the intensity of each ROI to the mean intensity of
each channel.
14. Measure the intensity of each channel within the mask by
multiplying the image by the mask (Process ! Image
Calculator).
15. Calculate the ratio of intensity within the mask relative to the
intensity in the ROI (organelle/cyt or PM/cyt here).
16. Copy and paste data into GraphPad Prism or other data analy-
sis software and format as desired.
17. This process is illustrated in Fig. 1 (see Note 18).
18. The data collected from this process are then illustrated in
Fig. 2.
Fig. 1 Generation of a binary PM mask. Step 1 illustrates a ROI applied to the original image (I0). Images I1–I3
have had the Gaussian blur filter applied at increasing radii (In x pfs). Step 2 shows the generation of
thresholded wavelets (W1–W3) from each pair of smoothed (I0–I3) images. Step three shows the final wavelet
decomposition to generate the binary mask with the background cleared to reduce noise
Fig. 2 Analysis of PtdIns(4,5)P2 biosensor distribution with the aid of a binary mask. A PM stain, CellMask deep
red, is used to generate a mask of the PM signal, as shown in Fig. 1. The cell is also expressing a low-affinity
PtdIns(4,5)P2 biosensor, TubbycR332H-mCherry. This low-affinity biosensor is barely enriched at the PM (see
inset in TubbycR332H-mCh image). Applying the binary mask allows the intensity of the biosensor to be easily
measured at the PM and compared (as an intensity ratio) with the average intensity across the whole cell’s
optical section, or else a region of cytosol (Cyt)
4 Notes
1. Fiji is a free version of ImageJ2, which includes the LOCI

Bio-Formats importer. The Fiji software is also regularly
updated using a built-in macro.
2. Dishes can be coated with FN for most cells; however, ECL
coating should be used for poorly adherent HEK293A cells.
3. To ensure reproducibility of experiments, it is recommended
that cells have been passaged no more than 30 times.
4. Prewarm complete DMEM, PBS, and TrypLE to 37 C for
>20 min before culturing cells. DMEM can be prewarmed in a
vented tissue culture flask in the incubator to allow for gas
exchange.
5. Consider a doubling time of 24 h for cells, recommended
seeding density of 50% for a same day transfection, or a seeding
density of 25% for a next day transfection. Volume of cell
suspension per dish ¼ (confluence at seeding/100) (dish
surface area/surface area of flask) resuspension volume (see
Table 2).
6. The volume of DNA can be controlled by diluting all plasmids
to 100 ng/μL before storing and using them.
7. Transfecting high concentrations of lipid biosensor has the
potential to sequester the target lipid and effect normal cellular
interactions. As such, biosensors should be expressed at the
lowest concentration possible. This may require optimization
depending on cell type and promoter driving expression of
biosensor protein.
8. Appropriate spectral separation is obtained by using these fluo-
rescent proteins.
9. This plasmid mix does not contain an iRFP-labeled PM marker.
Cell Mask Deep Red will be used as a PM marker in this setup.
However, an iRFP-tagged PM marker (iRFP-CAAX) could
easily be included and transfected alongside the other plasmids.
10. TIRF image acquisition parameters should be made to obtain
the greatest sensitivity, resolution, and acquisition speed. The
settings used in our lab are TIRF slider ¼ 2150, 2 2 pixel
binning on our Andor Zyla 5.5 sCMOS camera (giving
~130 nm image pixel size through the 100 objective), <5%
excitation, and 100 ms exposure time. These parameters maxi-
mize sensitivity and resolution and minimize photobleaching
of the fluorophores.
11. This setup is for cells taken at a single time point; refer to chap.
7 for a detailed setup for time-lapse imaging.
12. The time to acquire each image depends on the optimization of

the settings in Note 10. The higher the signal-to-noise ratio
and resolution, the slower the acquisition speed. The condi-
tions outlined above result in about a 3-s acquisition time per
image, or about a 90-s total acquisition time for 30 images.
13. Again, image acquisition parameters for confocal should be
made to get the greatest sensitivity, resolution, and acquisition
speed. The settings used in our lab are resonant scan mode, line
average of 8 or 16 frames, and gain and excitation varying
between different fluorophores, but these should be set to
the lowest value possible. These settings maximize sensitivity
and resolution and minimize photobleaching of the
fluorophores.
14. This DMEM-based formulation lacks phenol red and therefore
has low background to increase the signal-to-noise ratio on
fluorophores.
15. In this case, healthy cells are defined by morphology visualized
in DIC. Cells should be fully adherent to the dish and fully
spread. Avoid imaging cells that have rounded up or those with
blebbing or protrusions around the membrane, as these are
dead or dying cells.
16. LOCI developed Bio-Formats, which allows for ImageJ or Fiji
to open all file formats, including proprietary formats from
specific microscope vendors.
17. Draw each ROI around a maximum intensity projection of all
frames in the time-lapse to capture the maximum PM footprint
for the generation of the PM mask in each frame.
18. This process can be recorded into an automated macro using
the ImageJ IJ1 macro language.
Acknowledgments
This work was supported by a National Institutes of Health grant

(1R35GM119412-01) to G.R.V.H.
References
1. van Meer G, Voelker DR, Feigenson GW 4. Hammond GR, Fischer MJ, Anderson KE
(2008) Membrane lipids: where they are and et al (2012) PI4P and PI(4,5)P2 are essential
how they behave. Nat Rev Mol Cell Bio but independent lipid determinants of mem-
9:112–124 brane identity. Science 337:727–730
2. van Meer G, de Kroon AI (2011) Lipid map 5. Dickson EJ, Hille B (2019) Understanding
of the mammalian cell. J Cell Sci 124:5–8 phosphoinositides: rare, dynamic, and essen-
3. Balla T (2013) Phosphoinositides: tiny lipids tial membrane phospholipids. Biochem J
with Giant impact on cell regulation. Physiol 476:1–23
Rev 93:1019–1137
6. Balla T, Szentpetery Z, Kim Y (2009) Phos- C δ1 and p130. J Biol Chem

phoinositide signaling: new tools and 277:27412–27422
insights. Physiology 24:231–244 19. Balla T, Várnai P (2002) Visualizing cellular
7. Maekawa M, Fairn G (2014) Molecular phosphoinositide pools with GFP-fused pro-
probes to visualize the location, organization tein-modules. Sci Stke 2002:pl3–pl3
and dynamics of lipids. J Cell Sci 20. Lemmon MA, Ferguson KM, Abrams CS
127:4801–4812 (2002) Pleckstrin homology domains and
8. Hammond G, Balla T (2015) Polyphosphoi- the cytoskeleton. FEBS Lett 513:71–76
nositide binding domains: key to inositol lipid 21. Lemmon MA (2003) Phosphoinositide rec-
biology. Biochim Biophys Acta ognition domains. Traffic 4:201–213
1851:746–758 22. Hammond G, Machner MP, Balla T (2014) A
9. Tóth JT, Gulyás G, Tóth DJ et al (2016) novel probe for phosphatidylinositol
BRET-monitoring of the dynamic changes of 4-phosphate reveals multiple pools beyond
inositol lipid pools in living cells reveals a the Golgi. J Cell Biol 205:113–126
PKC-dependent PtdIns4P increase upon 23. Wills RC, Goulden BD, Hammond GR
EGF and M3 receptor activation. Biochim (2018) Genetically encoded lipid biosensors.
Biophys Acta 1861:177–187 Mol Biol Cell 29:1526–1532
10. Várnai P, Gulyás G, Tóth DJ et al (2017) 24. Schindelin J, Arganda-Carreras I, Frise E et al
Quantifying lipid changes in various mem- (2012) Fiji: an open-source platform for
brane compartments using lipid binding pro- biological-image analysis. Nat Methods
tein domains. Cell Calcium 64:72–82 9:676–682
11. Ivanova PT, Cerda BA, Horn DM et al (2001) 25. Linkert M, Rueden CT, Allan C et al (2010)
Electrospray ionization mass spectrometry Metadata matters: access to image data in the
analysis of changes in phospholipids in real world. J Cell Biol 189:777–782
RBL-2H3 mastocytoma cells during degran-
ulation. Proc Natl Acad Sci U S A 26. Shimada Y, Maruya M, Iwashita S et al (2002)
98:7152–7157 The C-terminal domain of perfringolysin O is
an essential cholesterol-binding unit targeting
12. Wenk MR, Lucast L, Paolo G et al (2003) to cholesterol-rich microdomains. Eur J Bio-
Phosphoinositide profiling in complex lipid chem 269:6195–6203
mixtures using electrospray ionization mass
spectrometry. Nat Biotechnol 21:813–817 27. Maekawa M, Fairn GD (2015) Complemen-
tary probes reveal that phosphatidylserine is
13. Milne SB, Ivanova PT, DeCamp D et al required for the proper transbilayer distribu-
(2005) A targeted mass spectrometric analysis tion of cholesterol. J Cell Sci 128:1422–1433
of phosphatidylinositol phosphate species. J
Lipid Res 46:1796–1802 28. Liu S-L, Sheng R, Jung J et al (2016) Orthog-
onal lipid sensors identify transbilayer asym-
14. Alter CA, Wolf BA (1995) Identification of metry of plasma membrane cholesterol. Nat
phosphatidylinositol 3,4,5-trisphosphate in Chem Biol 13:268–274
pancreatic islets and insulin-secreting β-cells.
Biochem Biophys Res Commun 29. Yamaji A, Sekizawa Y, Emoto K et al (1998)
208:190–197 Lysenin, a novel sphingomyelin-specific bind-
ing protein. J Biol Chem 273:5300–5306
15. Hokin-Neaverson M, Sadeghian K (1976)
Separation of [3H]inositol monophosphates 30. Kiyokawa E, Baba T, Otsuka N et al (2005)
and [3H]inositol on silica gel glass-fiber Spatial and functional heterogeneity of
sheets. J Chromatogr A 120:502–505 sphingolipid-rich membrane domains. J Biol
Chem 280:24072–24084
16. Natarajan V, Schmid HH (1987) Inositol
phospholipid hydrolysis by rat sciatic nerve 31. Abe M, Makino A, Hullin-Matsuda F et al
phospholipase C. J Neurochem (2012) A role for sphingomyelin-rich lipid
49:1878–1887 domains in the accumulation of phosphatidy-
linositol-4,5-bisphosphate to the cleavage fur-
17. Hegewald H (1996) One-dimensional thin- row during cytokinesis. Mol Cell Biol
layer chromatography of all known D-3 and 32:1396–1407
D-4 isomers of phosphoinositides. Anal Bio-
chem 242:152–155 32. Bohdanowicz M, Schlam D, Hermansson M
et al (2013) Phosphatidic acid is required for
18. Várnai P, Lin X, Lee S et al (2002) Inositol the constitutive ruffling and macropinocytosis
lipid binding and membrane localization of of phagocytes. Mol Biol Cell 24:1700–1712
isolated Pleckstrin homology (PH) domains
studies on the pH domains of phospholipase 33. Zhang F, Wang Z, Lu M et al (2014) Tempo-
ral production of the signaling lipid
phosphatidic acid by phospholipase D2 deter- 46. Sankaran VG, Klein DE, Chdeva M et al
mines the output of extracellular signal- (2001) High-affinity binding of a FYVE
regulated kinase signaling in cancer cells. domain to phosphatidylinositol 3-phosphate
Mol Cell Biol 34:84–95 requires intact phospholipid but not FYVE
34. Yeung T, Gilbert GE, Shi J et al (2008) Mem- domain oligomerization {. Biochemistry
brane phosphatidylserine regulates surface 40:8581–8587
charge and protein localization. Science 47. Gaullier J-M, Rønning E, Gillooly DJ et al
319:210–213 (2000) Interaction of the EEA1 FYVE finger
35. Maeda K, Anand K, Chiapparino A et al with phosphatidylinositol 3-phosphate and
(2013) Interactome map uncovers phosphati- early endosomes. Role of conserved residues.
dylserine transport by oxysterol-binding pro- J Biol Chem 275:24595–24600
teins. Nature 501:nature12430 48. Bravo J, Karathanassis D, Pacold CM et al
36. Vecchio K, Stahelin RV (2018) Investigation (2001) The crystal structure of the PX
of the phosphatidylserine binding properties domain from p40phox bound to phosphati-
of the lipid biosensor, Lactadherin C2 dylinositol 3-phosphate. Mol Cell 8:829–839
(LactC2), in different membrane environ- 49. Ellson CD, Gobert-Gosse S, Anderson KE
ments. J Bioenerg Biomembr 50:1–10 et al (2001) PtdIns(3)P regulates the neutro-
37. Stahelin RV, Digman MA, Medkova M et al phil oxidase complex by binding to the PX
(2005) Diacylglycerol-induced membrane domain of p40phox. Nat Cell Biol 3:679–682
targeting and activation of protein kinase Cϵ 50. Kanai F, Liu H, Field SJ et al (2001) The PX
mechanistic differences between protein domains of p47phox and p40phox bind to
kinases Cδ and Cϵ. J Biol Chem lipid products of PI(3)K. Nat Cell Biol
280:19784–19793 3:675–678
38. Domart M-C, Hobday TM, Peddie CJ et al 51. Brombacher E, Urwyler S, Ragaz C et al
(2012) Acute manipulation of diacylglycerol (2009) Rab1 guanine nucleotide exchange
reveals roles in nuclear envelope assembly & factor SidM is a major phosphatidylinositol
endoplasmic reticulum morphology. PLoS 4-phosphate-binding effector protein of
One 7:e51150 legionella pneumophila. J Biol Chem
39. Chen J, Deng F, Li J et al (2008) Selective 284:4846–4856
binding of phorbol esters and diacylglycerol 52. Schoebel S, Blankenfeldt W, Goody RS et al
by individual C1 domains of the PKD family. (2010) High-affinity binding of phosphatidy-
Biochem J 411:333–342 linositol 4-phosphate by legionella pneumo-
40. Kim Y, Guzman-Hernandez M, Balla T phila DrrA. EMBO Rep 11:598–604
(2011) A highly dynamic ER-derived phos- 53. Levin R, Hammond GR, Balla T et al (2016)
phatidylinositol-synthesizing organelle Multiphasic dynamics of phosphatidylinositol
supplies phosphoinositides to cellular mem- 4-phosphate during phagocytosis. Mol Biol
branes. Dev Cell 21:813–824 Cell 28:128–140
41. Zewe J, Miller A, Sangappa S, et al (2019) 54. Dolinsky S, Haneburger I, Cichy A et al
Probing the subcellular distribution of phos- (2014) The legionella longbeachae Icm/dot
phatidylinositol reveals a surprising lack at the substrate SidC selectively binds phosphatidy-
plasma membrane. Biorxiv 677039 linositol 4-phosphate with Nanomolar affinity
42. Pemberton JG, Kim Y, Sengupta N, et al and promotes pathogen vacuole-endoplasmic
(2019) Defining the subcellular distribution reticulum. Interactions 82:4021–4033
and metabolic channeling of phosphatidylino- 55. Weber S, Wagner M, Hilbi H (2014) Live-cell
sitol. Biorxiv 677229 imaging of phosphoinositide dynamics and
43. Burd CG, Emr SD (1998) membrane architecture during legionella
Phosphatidylinositol(3)-phosphate signaling infection. MBio 5:e00839–e00813
mediated by specific binding to RING FYVE 56. Zewe JP, Wills RC, Sangappa S et al (2018)
domains. Mol Cell 2:157–162 SAC1 degrades its lipid substrate PtdIns4P in
44. Gaullier J-M, Simonsen A, D’Arrigo A et al the endoplasmic reticulum to maintain a steep
(1998) FYVE fingers bind PtdIns(3)P. Nature chemical gradient with donor membranes.
394:432–433 Elife 7:e35588
45. Gillooly DJ, Morrow IC, Lindsay M et al 57. Chung J, Torta F, Masai K et al (2015) PI4P/
(2000) Localization of phosphatidylinositol phosphatidylserine countertransport at
3-phosphate in yeast and mammalian cells. ORP5- and ORP8-mediated ER–plasma
EMBO J 19:4577–4588 membrane contacts. Science 349:428–432
58. Ghai R, Du X, Wang H et al (2017) ORP5 69. Manna D, Albanese A, Park W et al (2007)
and ORP8 bind phosphatidylinositol-4, Mechanistic basis of differential cellular
5-biphosphate (PtdIns(4,5) P 2) and regulate responses of phosphatidylinositol 3,4-bispho-
its level at the plasma membrane. Nat Com- sphate- and phosphatidylinositol 3,4,5-tri-
mun 8:757 sphosphate-binding Pleckstrin homology
59. Sohn M, Korzeniowski M, Zewe JP et al domains. J Biol Chem 282:32093–32105
(2018) PI(4,5)P2 controls plasma membrane 70. Oikawa T, Itoh T, Takenawa T (2008)
PI4P and PS levels via ORP5/8 recruitment Sequential signals toward podosome forma-
to ER–PM contact sites. J Cell Biol 217 tion in NIH-src cells. J Cell Biol
(5):1797–1813 182:157–169
60. Levine TP, Munro S (2002) Targeting of 71. Posor Y, Eichhorn-Gruenig M, Puchkov D
Golgi-specific Pleckstrin homology domains et al (2013) Spatiotemporal control of endo-
involves both PtdIns 4-kinase-dependent cytosis by phosphatidylinositol-3,4-bispho-
and -independent components. Curr Biol sphate. Nature 499:233–237
12:695–704 72. He K, Robert Marsland SU III et al (2017)
61. Szentpetery Z, Várnai P, Balla T (2010) Acute Dynamics of phosphoinositide conversion in
manipulation of Golgi phosphoinositides to clathrin-mediated endocytic traffic. Nature
assess their importance in cellular trafficking 552:410
and signaling. Proc Natl Acad Sci U S A 73. Goulden BD, Pacheco J, Dull A et al (2018) A
107:8225–8230 high-avidity biosensor reveals plasma mem-
62. Lenoir M, Grzybek M, Majkowski M et al brane PI(3,4)P2 is predominantly a class I
(2015) Structural basis of dynamic membrane PI3K signaling product. J Cell Biol 218:
recognition by trans-Golgi network specific jcb.201809026
FAPP proteins. J Mol Biol 427:966–981 74. Li X, Wang X, Zhang X et al (2013) Geneti-
63. Gozani O, Karuman P, Jones DR et al (2003) cally encoded fluorescent probe to visualize
The PHD finger of the chromatin-associated intracellular phosphatidylinositol
protein ING2 functions as a nuclear phos- 3,5-bisphosphate localization and dynamics.
phoinositide receptor. Cell 114:99–111 Proc Natl Acad Sci U S A 110:21165–21170
64. Pendaries C, Tronchère H, Arbibe L et al 75. Hammond G, Takasuga S, Sasaki T et al
(2006) PtdIns(5)P activates the host cell (2015) The ML1Nx2 phosphatidylinositol
PI3-kinase/Akt pathway during Shigella flex- 3,5-bisphosphate probe shows poor selectiv-
neri infection. EMBO J 25:1024–1034 ity in cells. PLoS One 10:1–13
65. Dowler S, Currie RA, Campbell DG et al 76. Ford MG, Pearse BM, Higgins MK et al
(2000) Identification of pleckstrin-homol- (2001) Simultaneous binding of PtdIns(4,5)
ogy-domain-containing proteins with novel P2 and Clathrin by AP180 in the nucleation of
phosphoinositide-binding specificities. Bio- Clathrin lattices on membranes. Science
chem J 351:19–31 291:1051–1055
66. Thomas CC, Dowler S, Deak M et al (2001) 77. Itoh T, Koshiba S, Kigawa T et al (2001) Role
Crystal structure of the phosphatidylinositol of the ENTH domain in Phosphatidylinosi-
3,4-bisphosphate-binding pleckstrin homol- tol-4,5-bisphosphate binding and endocyto-
ogy (PH) domain of tandem PH-domain- sis. Science 291:1047–1051
containing protein 1 (TAPP1): molecular 78. Yoon Y, Lee PJ, Kurilova S et al (2011) In situ
basis of lipid specificity. Biochem J quantitative imaging of cellular lipids using
358:287–294 molecular sensors. Nat Chem 3:868–874
67. Marshall AJ, Krahn AK, Ma K et al (2002) 79. Garcia P, Gupta R, Shah S et al (1995) The
TAPP1 and TAPP2 are targets of phosphati- pleckstrin homology domain of phospholi-
dylinositol 3-kinase signaling in B cells: sus- pase C-.delta.1 binds with high affinity to
tained plasma membrane recruitment phosphatidylinositol 4,5-bisphosphate in
triggered by the B-cell antigen receptor. Mol bilayer membranes. Biochemistry
Cell Biol 22:5479–5491 34:16228–16234
68. Kimber WA, Trinkle-mulcahy L, Cheung PC 80. Lemmon M, Ferguson K, O’Brien R et al
et al (2002) Evidence that the tandem-pleck- (1995) Specific and high-affinity binding of
strin-homology-domain-containing protein inositol phosphates to an isolated pleckstrin
TAPP1 interacts with Ptd(3,4)P2 and the homology domain. Proc Natl Acad Sci U S A
multi-PDZ-domain-containing protein 92:10472–10476
MUPP1 in vivo. Biochem J 361:525–536 81. Stauffer TP, Ahn S, Meyer T (1998)
Receptor-induced transient reduction in
plasma membrane PtdIns(4,5)P2 concentra- Tetrakisphosphate binding capacity. J Biol

tion monitored in living cells. Curr Biol Chem 271:30303–30306
8:343–346 92. Salim K, Bottomley M, Querfurth E et al
82. Várnai P, Balla T (1998) Visualization of (1996) Distinct specificity in the recognition
Phosphoinositides that bind Pleckstrin of phosphoinositides by the pleckstrin homol-
homology domains: calcium- and agonist- ogy domains of dynamin and Bruton’s tyro-
induced dynamic changes and relationship to sine kinase. EMBO J 15:6241–6250
Myo-[3H]inositol-labeled phosphoinositide 93. Rameh LE, Arvidsson A, Carraway KL et al
pools. J Cell Biol 143:501–510 (1997) A comparative analysis of the phos-
83. Hirose K, Kadowaki S, Tanabe M et al (1999) phoinositide binding specificity of Pleckstrin
Spatiotemporal dynamics of inositol 1,4,5- homology domains. J Biol Chem
trisphosphate that underlies complex Ca2+ 272:22059–22066
mobilization patterns. Science 94. Kontos CD, Stauffer TP, Yang W-P et al
284:1527–1530 (1998) Tyrosine 1101 of Tie2 is the major
84. Suh B, Inoue T, Meyer T et al (2006) Rapid site of association of p85 and is required for
chemically induced changes of PtdIns(4,5)P2 activation of phosphatidylinositol 3-kinase
gate KCNQ ion channels. Science and Akt. Mol Cell Biol 18:4131–4140
314:1454–1457 95. Klarlund JK, Guilherme A, Holik JJ et al
85. Lee S, Várnai P, Balla A et al (2004) The (1997) Signaling by Phosphoinositide-3,4,5-
Pleckstrin homology domain of trisphosphate through proteins containing
phosphoinositide-specific phospholipase Cδ4 Pleckstrin and Sec7 homology domains. Sci-
is not a critical determinant of the membrane ence 275:1927–1930
localization of the enzyme. J Biol Chem 96. Venkateswarlu K, Oatey PB, Tavaré JM et al
279:24362–24371 (1998) Insulin-dependent translocation of
86. Quinn K, Behe P, Tinker A (2008) Monitor- ARNO to the plasma membrane of adipocytes
ing changes in membrane phosphatidylinosi- requires phosphatidylinositol 3-kinase. Curr
tol 4,5-bisphosphate in living cells using a Biol 8:463–466
domain from the transcription factor tubby. J 97. Cronin TC, DiNitto JP, Czech MP et al
Physiol 586:2855–2871 (2004) Structural determinants of phosphoi-
87. Halaszovich CR, Schreiber DN, Oliver D nositide selectivity in splice variants of Grp1
(2009) Ci-VSP is a depolarization-activated family PH domains. EMBO J 23:3711–3720
phosphatidylinositol-4,5-bisphosphate and 98. Gray A, Kaay J, Downes PC (1999) The
phosphatidylinositol-3,4,5-trisphosphate 5- pleckstrin homology domains of protein
0
-phosphatase. J Biol Chem 284:2106–2113 kinase B and GRP1 (general receptor for
88. Szentpetery Z, Balla A, Kim YJ et al (2009) phosphoinositides-1) are sensitive and selec-
Live cell imaging with protein domains capa- tive probes for the cellular detection of phos-
ble of recognizing phosphatidylinositol phatidylinositol 3,4-bisphosphate and/or
4,5-bisphosphate; a comparative study. BMC phosphatidylinositol 3,4,5-trisphosphate
Cell Biol 10:67 in vivo. Biochem J 344:929–936
89. Frech M, Andjelkovic M, Ingley E et al 99. Cohen L, Honda A, Varnai P et al (2007)
(1997) High affinity binding of inositol phos- Active Arf6 recruits ARNO/cytohesin GEFs
phates and Phosphoinositides to the Pleck- to the PM by binding their PH domains. Mol
strin homology domain of RAC/protein Biol Cell 18:2244–2253
kinase B and their influence on kinase activity. 100. Hofmann I, Thompson A, Nderson C et al
J Biol Chem 272:8474–8481 (2007) The Arl4 family of small G proteins
90. Watton SJ, Downward J (1999) Akt/PKB can recruit the Cytohesin Arf6 exchange fac-
localisation and 30 phosphoinositide genera- tors to the plasma membrane. Curr Biol
tion at sites of epithelial cell–matrix and cell–- 17:711–716
cell interaction. Curr Biol 9:433–436 101. Li C-C, Chiang T-C, Wu T-S et al (2007)
91. Fukuda M, Kojima T, Kabayama H et al ARL4D recruits Cytohesin-2/ARNO to
(1996) Mutation of the Pleckstrin homology modulate actin remodeling. Mol Biol Cell
domain of Bruton’s tyrosine kinase in immu- 18:4420–4437
nodeficiency impaired inositol 1,3,4,5-
Chapter 5
Detection of Plasma Membrane Phosphoinositide Dynamics

Using Genetically Encoded Fluorescent Protein Probes
Rebecca Cabral-Dias, Yasmin Awadeh, Roberto J. Botelho,
and Costin N. Antonescu
Abstract
The dynamic phosphorylation of phosphatidylinositol produces seven distinct but interconvertible phos-
phatidylinositol phosphates (PIPs). Each PIP exhibits specific enrichment in a subset of membrane com-
partments as a result of dynamic phosphorylation and dephosphorylation by lipid kinases and phosphatases,
and/or by vesicle-mediated transport. Several PIPs are found within the plasma membrane, such as
phosphatidylinositol-4-phosphate [PI(4)P], phosphatidylinositol-4,5-bisphosphate [PI(4,5)P2], phospha-
tidylinositol-3,4-bisphosphate [PI(3,4)P2], and phosphatidylinositol-3,4,5-trisphosphate (PIP3), and
these control many aspects of cell physiology, including receptor signaling and membrane traffic. As a
result, measurement of the cell surface abundance of these PIPs is a valuable resource to allow understand-
ing of the regulation and function of these cell surface lipids. Here, we describe methods based on
quantification of the localization of genetically encoded fluorescent PIP probes to the cell surface by either
spinning disc confocal microscopy or total internal reflection fluorescence microscopy that allow detection
of changes in cell surface levels of PI(4,5)P2, PI(3,4)P2, and PIP3. These methods can also be applied to
the measurement of other PIPs or lipid species at the cell surface, and thus represent a useful resource for
the study of the cell biology of PIPs.
Key words Lipids, Membrane, Microscopy, Biosensors, Fluorescence, Transfection, Signaling
1 Introduction
1.1 Phosphoinositide Phosphatidylinositol phosphates (PIPs) or phosphoinositides are a

Signaling collection of phosphorylated derivatives of phosphatidylinositol
that have many critical roles in the regulation of cell physiology
[1–3]. PIPs have well-established functions to regulate various
signal transduction pathways, including signals elicited by many
growth factors, hormones, and cytokines as well as critical roles in
the regulation of vesicle-mediated membrane traffic and membrane
remodeling [1, 2, 4–8]. Additional roles of PIPs have also emerged,
such as control of gene expression and chromatin organization in
the nucleus [9, 10], regulation of lipid transfer at membrane
73
74 Rebecca Cabral-Dias et al.
contact sites [11, 12], and the regulation of ion channels [13–
17]. These collective efforts have revealed the central role that
PIPs play in a multitude of cellular processes and how disruptions
in PIP regulation can contribute to diseases such as cancer and
metabolic disorders [1–3, 8].
There are seven species of PIPs as defined by single or combi-
nations of phosphorylation of either the third, fourth, or fifth
position on the inositol ring in the headgroup. These PIPs are
generated by a highly regulated cohort of lipid kinases and phos-
phatases that collectively ensure proper synthesis, turnover, and
compartmentalization of each PIP species [2, 17]. Past models
envisioned each PIP species as exclusively localized to specific
membrane compartments; for example, phosphatidylinositol-4,5-
bisphosphate [PI(4,5)P2] was thought to exist at the plasma mem-
brane, while phosphatidylinositol-4-phosphate [PI(4)P] was recog-
nized as present in the trans-Golgi network and the plasma
membrane [18]. However, a more nuanced view of PIP spatiotem-
poral distribution has emerged, whereby a given PIP can exist in
multiple compartments [1]. As such, PI(4)P has now been detected
on lysosomes, the plasma membrane, and phagosomes at different
stages of maturation and seems to drive the dynamics of membrane
contact sites involving the endoplasmic reticulum [1, 17, 19–
22]. In comparison, PI(4,5)P2 was observed on autolysosomes
undergoing lysosome generation and even in the nucleus [23–
25]. Moreover, PIP spatial distribution is subject to dynamic
changes, whereby a PIP may exist transiently in a given membrane
or membrane subdomain [1, 17, 26]. Thus, the regulation of each
PIP is a carefully orchestrated process that achieves selective spatio-
temporal production of specific PIPs in each organellar compart-
ment or subdomain. Advancing our understanding of PIPs thus
requires methods to measure PIP dynamics, as well as to detect
compartment-specific PIP pools.
1.2 Fluorescent To assess the spatiotemporal dynamics of a specific PIP, genetically

Protein Biosensors encoded protein probes are commonly employed [27, 28]. In gen-
for PIPs eral, these probes comprise a protein domain that exhibits high
specificity and good affinity for a given PIP, either as a single copy or
as multiple tandem copies, fused to a fluorescent protein (Fig. 1a).
As such, these engineered protein biosensors require transfection of
plasmids and expression of the proteins within cell types of interest,
followed by fluorescence microscopy, and then quantification of
changes in fluorescence intensity and subcellular distribution of
the probe to a specific compartment such as the plasma membrane
[27, 28]. Several recent articles have reviewed the collection of
genetically encoded probes available to monitor the levels and
localization of several PIPs [28–30].
Here, we describe the use of genetically encoded probes to
detect and quantify changes in PI(4,5)P2 and poly-3-PIPs
Detection of Plasma Membrane Phosphoinositides 75
A
lipid bilayer (PM)
PIP headgroup
PH
FP
B C
400-700 nm
~100 nm
Fig. 1 Measuring the fluorescence of PIP biosensors in cells with different morphologies. (a) Shown is a
diagram of a genetically encoded fluorescent PIP-binding probe, composed of a fluorescent protein (FP) such
as eGFP or mCherry fused to a PH domain with specific PIP-binding properties. In this diagram, the PIP
headgroup is shown exaggerated relative to the plasma membrane (PM) lipid bilayer to highlight the
interaction between the PIP headgroup and the PH domain. (b) The relative cell surface localization of
fluorescent probes in cells with a rounded morphology can be commonly performed using spinning disc
confocal microscopy, as axial resolution of an optical section (400–700 nm, shown in blue) allows the cell
surface to appear as a rim along the cell periphery that is readily distinguished from bulk cytoplasm. (c) The
relative cell surface localization of fluorescent probes in cells with a flat or well-spread morphology can be
performed using total internal reflection fluorescence (TIRF) microscopy as the restricted illumination of
~100 nm proximal to the coverslip (shown in blue) corresponds to a volume highly enriched in the plasma
membrane. Thus, TIRF microscopy is effective at scoring changes in cell recruitment of various PIP biosensors
[phosphatidylinositol-3,4-bisphosphate, PI(3,4)P2, or phosphati-

dylinositol-3,4,5-trisphosphate, PIP3] at the plasma membrane
using quantitative microscopy methods. The detection of PI(4,5)
P2 is commonly accomplished using the PH domain of PLCδ1 [31–
33]. This probe has high specificity and affinity for PI(4,5)P2 yet
has the disadvantage of also having high affinity for IP3
[27, 32]. Additional protein biosensors based on the C-terminal
domain of Tubby have also been described that bind to PI(4,5)P2
but not the soluble IP3 [27, 34, 35]. As for detection of 3-poly-
PIPs, a fluorescent protein chimera of the PH domain of Akt is
typically used [27, 36, 37]. Of note, this probe is often used as a
PIP3 reporter, but it also reports PI(3,4)P2 with comparable affin-
ity, making interpretations regarding changes in PIP3 challenging
[27, 28, 38]. Alternative domains that exhibit selective binding and
detection of PIP3 include the PH domains of Btk [38, 39] or
Myosin X [40], while a biosensor based on tandem copies of the
C-terminal PH domains of TAPP1 was recently engineered as a
higher fidelity sensor for PI(3,4)P2 [41]. Thus, it may be important

to employ complementary biosensors to draw more robust conclu-
sions about PIP dynamics under given experimental conditions.
The levels of PIP3 and PI(3,4)P2 are low in resting cells and can
be triggered by a variety of signals such as growth factor stimulation
[42]. As an example, here we describe the measurement of PIP3 and
PI(3,4)P2 levels at the cell surface by monitoring the localization of
the poly-3-PIP reporter eGFP-PH(AKT) following epidermal
growth factor (EGF) stimulation [43]. In contrast, PI(4,5)P2 is
constitutively present at the plasma membrane but can undergo
acute depletion upon activation of certain signals that trigger
hydrolysis of PI(4,5)P2 via phospholipase C (PLC). As an example,
we describe the measurement of PI(4,5)P2 by the detection of
mCherry-PH(PLCδ1) upon treatment with ionomycin, a calcium
ionophore that elicits loss of PI(4,5)P2 via PLC activation [44]. We
also describe a method in ARPE-19 cells relying on cell fixation and
TIRF-M to detect the cell surface levels of PI(4,5)P2 based on
detection of eGFP-PH(PLCδ1).
1.3 Microscopy In this chapter, we describe two microscopy-based methods that

and Image Analysis quantify PIPs on the plasma membrane. In one method, we use
spinning disc confocal microscopy to measure the fluorescence of
PIP biosensors at the cell periphery, which appears as a rim of
fluorescence around the cell, versus cytosolic signal. This approach
is commonly used in cells that have a more rounded morphology, as
this allows resolution of the plasma membrane-proximal signals
that appear along the cell periphery from bulk cytoplasm ([45];
Fig. 1b). The second approach is based on the detection of PIPs at
or near the cell surface using total internal reflection fluorescence
(TIRF) microscopy, which selectively illuminates fluorophores
within ~100 nm of the coverslip, which corresponds to a volume
that is highly enriched in plasma membrane versus other compo-
nents of the cytosol ([28, 46, 47]; Fig. 1c). This TIRF-based
method is effective at scoring changes in plasma membrane recruit-
ment of various PIP probes in well-spread or flat cells in which the
axial resolution of a confocal microscope does not allow one to
discern the cytosol and ventral/dorsal plasma membrane in differ-
ent optical sections.
The TIRF field also reports changes in bulk membrane and/or
cell morphology, which need to be taken into consideration
[47]. For instance, the apparent localization of a PIP-binding
probe to the plasma membrane may also be impacted by large
changes in membrane traffic like bulk endocytosis or exocytosis.
In addition, when considering that illumination of fluorophores in
TIRF microscopy is highly sensitive to the distance of fluorophores
relative to the coverslip [47], it is important to resolve if any

treatments examined result in changes in cell attachment or prox-
imity to the coverslip. Each of these examples can result in apparent
changes in the plasma membrane localization of a genetically
encoded probe to detect a specific PIP that does not entirely reflect
changes in the levels of that specific PIP in the plasma membrane.
Thus, to account for bulk membrane changes or alterations to cell
morphology and attachment, measurement of a plasma membrane
marker is required. Many such probes have been described, includ-
ing the CAAX farnesylation motif from H-Ras fused to a fluores-
cent protein ([48], e.g., eGFP-CAAX) or a fluorescent protein
conjugated to a GPI anchor ([49], e.g., eGFP-GPI). We describe
the use of mCherry-GPI probe or eGFP-CAAX to distinguish
between changes in apparent cell surface localization. Such probes
can be readily used to demark the plasma membrane and ascertain if
changes in bulk plasma membrane traffic and/or changes in cell
morphology occur in a given condition or treatment. When
co-expressed with the PIP biosensor, a ratio of the change of
fluorescence within the two channels (e.g., at the plasma mem-
brane) upon a specific treatment allows determination of to what
extent changes in cell surface fluorescence reflect changes in PIP
levels versus broader membrane or cellular alterations.
Here, we provide detailed methods for the culture (Subheading
3.1) and transfection (Subheading 3.2) of cells with various geneti-
cally encoded fluorescent probes in preparation for imaging experi-
ments. We then describe the treatment, live-cell spinning disc
confocal microscopy imaging (Subheading 3.3.1), and analysis
(Subheading 3.4.1) methods to measure changes in plasma mem-
brane 3-poly-PIPs (using eGFP-PH(AKT) probe and mCherry-
GPI plasma membrane marker, Fig. 2a, c) and PI(4,5)P2 (using
mCherry-PH(PLCδ1) and eGFP-GPI plasma membrane marker,
Fig. 2b, d) in MDA-MB-231 cells upon treatment with EGF and
ionomycin, respectively. MDA-MB-231 cells exhibit a rounded
morphology suitable for detection of plasma membrane localiza-
tion of fluorescent biosensors by this method. Then, we provide a
detailed method for treatment and imaging using TIRF and wide-
field fluorescence microscopy (Subheading 3.3.2) followed by anal-
ysis (Subheading 3.4.2) for the measurement of plasma membrane
PI(4,5)P2 using eGFP-PH(PLCδ1) or the eGFP-CAAX plasma
membrane marker in ARPE-19 cells in fixed cell samples (Fig. 3).
ARPE-19 cells exhibit a flat and well-spread morphology suitable
for detection of plasma membrane localization of fluorescent bio-
sensors by this method.
time of EGF addition time of ionomycin addition

A -45s 60s 195s 285s C -45s 30s 105s 285s
mCherry-GPI
eGFP-GPI
mCh-PH (PLcδ1)
eGFP-PH (AKt)
mCherry-GPI
eGFP-GPI
mCh-PH (PLcδ1)
eGFP-PH (AKt)
B 2.0 3.5 D 3.5 eGFP-GPI

the plasma membrane
the plasma membrane
mCherry-PH (PLCδ1)
mCherry-GPI within
3.0
eGFP-PH-Akt within
Probe enrichment
1.8 3.0
2.5
1.6 in the PM
2.5 2.0
1.4
1.5
2.0
1.2
1.0
1.5
1.0 0.5
-100 0 100 200 300 400 -100 0 100 200 300 400 0 100 200 300
time of EGF addition (s) time of EGF addition (s) time of ionomycin addition (s)
Fig. 2 Detection of changes in cell surface 3-polyPIPs or PI(4,5)P2 in MDA-MB-231 cells using live-cell
spinning disc confocal microscopy. (a) MDA-MB-231 cells were co-transfected with plasmids encoding
mCherry-GPI and eGFP-PH (Akt), as described in Subheading 3.2. On the day of the experiment, measurement
of cell surface localization of each probe upon addition of 20/mL EGF at the indicated time was performed as
described in Subheading 3.3.1 and quantified as in Subheading 3.4.1. Shown are representative fluorescence
micrographs of mCherry-GPI and eGFP-PH(Akt) fluorescence (upper panels). Also shown (lower panels) are the
same images overlaid with ROIs selected for quantification (yellow: cytosolic; red: cell surface), as described
in Subheading 3.4.1. Scale bar ¼ 20 μm. (b) The relative cell surface localization of each probe at indicated
times of EGF (20 ng/mL) addition, determined as in Subheading 3.4.1, is shown as mean SE of independent
measurement in 3 cells. (c) MDA-MB-231 cells were co-transfected with plasmids encoding eGFP-GPI and
mCherry-PH(PLCδ1), as described in Subheading 3.2. On the day of the experiment, measurement of cell
surface localization of each probe upon addition of 10 μM ionomycin at the indicated time was performed as
described in Subheading 3.3.1 and quantified as in Subheading 3.4.1. Shown are representative fluorescence
micrographs of eGFP-GPI and mCherry-PH(PLCδ1) fluorescence (upper panels). Also shown (lower panels) are
the same images overlaid with ROIs selected for quantification (red: cell surface, yellow: cytosolic), as
described in Subheading 3.4.1. (d) The relative cell surface localization of each probe at indicated times of
ionomycin addition, determined as in Subheading 3.4.1, is shown as mean SE of independent measurement
in 3 cells. Scale bar ¼ 20 μm
A Ionomycin (min)
0 min
Ionomycin (min)
5 min 10 min
0 min 5 min 10 min
TIRF
TIRF
eGFP-PH (PLCδ1)
eGFP-CAAX
TIRF
TIRF
eGFP-PH (PLCδ1)
eGFP-CAAX
B 1.5 1.5
eGFP-CAAX probe
TIRF/EPI intensity
TIRF/EPI intensity
GFP-PH (PLCδ1)
1.0 1.0
0.5 0.5
0.0 0.0
0 5 10 15 20 0 5 10
ionomycin (min) ionomycin (min)
Fig. 3 Detection of changes in cell surface PI(4,5)P2 in ARPE-19 cells using TIRF/widefield fluorescence
microscopy in fixed cell samples. ARPE-19 cells were transfected with plasmids encoding eGFP-PH(PLCδ1) or
eGFP-CAAX, as described in Subheading 3.2. On the day of the experiment, some cells were treated with
10 μM ionomycin for the indicated times, fixed, and subjected to TIRF/widefield fluorescence microscopy as in
Subheading 3.3.2 and quantified as in Subheading 3.4.2. (a) Shown are representative fluorescence micro-
graphs, showing eGFP-PH(PLCδ1) (left panels) or eGFP-CAAX (right panels). In addition, overlays of ROIs
selected for quantification are also shown (lower panels) (yellow: cell outline, red: background). (b) The
relative cell surface localization of each probe at indicated times of ionomycin addition, determined as in
Subheading 3.3.2, by determining the TIRF to widefield epifluorescence ratio (TIRF/EPI) is shown as mean SE
of measurements in >10 cells per condition. Scale bar ¼ 20 μm
2 Materials
2.1 DNA Constructs 1. Plasmids encoding fluorescent probes to detect specific PIPs,
of PIP Biosensor such as eGFP-PH(PLCδ1), eGFP-PH(AKT), or mCherry-PH
Probes for PI(4,5)P2 (PLCδ1) (see Notes 1 and 2).
and PIP3/PI(3,4)P2 2. Plasmids encoding fluorescent plasma membrane markers,
such as mCherry-GPI or eGFP-CAAX (see Notes 1 and 2).
2.2 Tissue Culture 1. Tissue culture flasks (see Note 3).

2. Transfectable cell lines such as MDA-MD-231 (round mor-
phology, Fig. 1b) or ARPE-19 (flat morphology, Fig. 1c).
3. Complete cell culture media: DMEM/F12 (high glucose),
supplemented with 10% fetal bovine serum, 100 U/mL peni-
cillin, and 100 μg/mL streptomycin.
4. Dulbecco’s phosphate-buffered saline (dPBS).
5. Cell dissociation reagent, such as solution of 0.25% trypsin
supplemented with EDTA.
6. Square coverslips (#1.5, 22 mm 22 mm) (see Note 4).
7. Serum-starvation media: DMEM (low glucose) with 15 mM HEPES.
8. Transfection reagents (see Note 5).
9. Opti-MEM media for transfection.
2.3 Microscopy 1. Serum-free imaging media: Phenol red-free DMEM/F12 (low

Reagents glucose) with 15 mM HEPES.
2. EGF working stock solution: 20–40 ng/mL human EGF
diluted in serum-free imaging media (see Note 16).
3. Ionomycin working stock solution: 10–20 μM ionomycin from
Streptomyces conglobatus, diluted in serum-free imaging media
(see Note 17).
2.4 Immuno- 1. PBS: 2.7 mM KCl, 1.5 mM KH2PO4, 136.9 mM NaCl,

fluorescence Fixation 8.9 mM Na2HPO4, pH 7.4.
Preparation 2. Fixative: 4% paraformaldehyde (PFA) diluted in PBS.
3. Quenching solution: 100 mM glycine in PBS.
2.5 Microscopy 1. Spinning disc confocal microscope (SDCM) system (for Sub-
and Image Analysis heading 3.3.1) and/or combination total internal reflection
Infrastructure fluorescence (TIRF) microscope with widefield epifluorescence
capabilities (for Subheading 3.3.2).
2. A coverslip chamber compatible with the imaging microscope
stage and coverslips used (see Note 11).
3. Live-cell imaging chamber for temperature and/or CO2 control
during image acquisition (for live-cell imaging applications).
4. Microscope control to enable two fluorescence channel acqui-
sition during time-lapse. Ideally, stage is motorized and can
collect data from multiple locations.
5. Laser lines at 488 nm and 561 nm (or similar, for green and red
channels, respectively).
6. Emission filters compatible with fluorophores used, such as at
527/30 nm (green) and 630/75 nm (red).
7. Objective suitable for spinning disc confocal (Subheading
3.3.1) and/or TIRF (Subheading 3.3.2) microscopy, such as
63/1.49 NA objective with a 1.8 camera relay (total mag-
nification 108).
8. High-resolution sCMOS camera or similar.
9. Image analysis software such as ImageJ (National Institutes of
Health), version 1.52p or newer [50].
3 Methods
3.1 Cell Culture 1. Aspirate media from a confluent T75 flask with a sterile pipette
and wash cells three times with 10 mL of dPBS (see Note 6).
2. Aspirate dPBS and add 1.5 mL of 0.25% trypsin–EDTA
solution.
3. Place the flask in an incubator at 37 C with 5% CO2 for
2–5 min, or until cells have detached from the surface of the
flask.
4. Add 8.5 mL of complete cell culture media to neutralize the
trypsin–EDTA solution, and aspirate cells by gently pipetting.
5. Seed cells in a 6-well plate on square coverslips at an appropri-
ate volume (see Note 7), passing a fraction of cell suspension to
a new T75 flask as needed.
6. Allow cells to adhere to coverslips before transfection (see Note 8).
3.2 Transfection Per 1 well in a 6-well plate:

1. Dilute 3 μL of lipofectamine 2000 transfection reagent into
97 μL of Opti-MEM media in a sterile 1.5-mL microfuge tube.
Mix by trituration.
2. Dilute 0.5 μg of the PIP reporter plasmid DNA and 0.5 μg of
the plasma membrane reporter plasmid DNA (see Note 9) into
a total of 100 μL of Opti-MEM in a 1.5-mL microfuge tube.
Mix by trituration.
3. Combine the diluted lipofectamine 2000 and DNA solution
(from steps 1 and 2), and vortex for a minimum of 30 s.
4. Incubate the mixture for 15 min at room temperature to allow
for the formation of lipid–DNA complexes.
5. Wash cells 3 in PBS and replace media with 1.8 mL of pre-
warmed Opti-MEM media in each well.
6. Pipette 200 μL dropwise of lipid–DNA complex to Opti-MEM
media in each well (step 5), making the total well volume to 2.0 mL.
7. Incubate cells with transfection mixture for 4–6 h at 37 C with

5% CO2.
8. Aspirate transfection mixture, wash 3 in PBS, and replace
media with prewarmed complete cell culture media.
9. Incubate cells for an additional 16–20 h (total 20–24 h since
starting step 7) before visualizing cells.
3.3 Detection 1. Prewarm serum-free imaging media (see Note 10).

of PI(4,5)P2 and PI(3,4) 2. Wash transfected cells in PBS and incubate cells (MDA-MB-
P2/PIP3 Dynamics 231 cells are used here) in serum-free imaging media at 37 C
3.3.1 Measurement with 5% CO2 for 1 h.
of Cell Surface PIP Levels 3. Prepare a live-cell chamber suitable for coverslip and add
Using SDCM in Live Cells serum-free imaging media (see Note 11).
4. Remove debris on the outer surface of the coverslip in the live-
cell chamber with 70% ethanol and a Kim Wipe.
5. Apply immersion oil on the objective lens (see Note 12).
6. Mount the live-cell chamber on the microscope stage.
7. Find the focus plane and look for viable cells that are transiently
transfected with the fluorescent signals in the channels required
(see Note 13).
8. Prewarm drug or growth factor working solution (e.g., EGF or
ionomycin).
9. Begin live-cell imaging of cells by acquiring images in required
channels (see Note 14).
10. Acquire fluorescence frames sequentially and near-simultaneously
with a 10 s framerate. Acquire at least 3 frames of each channel in
basal condition before proceeding to treatment, as below.
11. Treat cells without interrupting image acquisition (see Note
15), while noting the time of stimulus addition:
(a) For detection of PIP3/PI(3,4)P2: For example, add pre-
warmed EGF working solution to cells (in the live-cell
chamber) to achieve a final concentration of 20 ng/mL
(see Note 16). Sample images obtained by this method are
shown in Fig. 2a (top panels).
(b) For detection of PI(4,5)P2: For example, add prewarmed
ionomycin solution to cells (in the live-cell chamber) to
achieve a final concentration of 10 μM (see Note 17).
Sample images obtained by this method are shown in
Fig. 2c (top panels).
12. Continue to image cells for 5 min, or until the time-lapse
movie is complete.
13. Export files as TIFF or other lossless format. Sample images at
various times of EGF or ionomycin treatment are shown in
Fig. 2a, c (top panels).
3.3.2 Measurement 1. Prewarm serum-free imaging media and ionomycin working

of Cell Surface PIP Levels solution.
Using TIRF and Widefield 2. Wash transfected cells in PBS and incubate cells (ARPE-19 cells
Epifluorescence are used here) in serum-free imaging media at 37 C with 5%
Microscopy in Fixed Cells CO2 for 1 h.
3. Treat some cells with media containing 10 μM of ionomycin
and incubate for 5 min (and/or other experimental time-
points) at 37 C with 5% CO2, while retaining (a) separate
well(s) untreated (control).
4. Immediately following ionomycin treatment, place cells on ice
and wash cells with ice-cold PBS three times.
5. Aspirate PBS and fix cells with ice-cold fixative solution (see
Note 18).
6. Incubate cells on ice for 30 min in the dark.
7. Aspirate the fixative solution from each well and quickly wash
the cells once with quenching solution (see Note 19).
8. Following the wash, incubate cells in quenching solution for
20 min at room temperature in the dark.
9. Aspirate quenching solution and wash cells with PBS three
times. Leave cells in PBS for immediate TIRF imaging. If
necessary, slides can be stored to be imaged at a later time (see
Note 20).
10. Prepare the coverslip chamber by immersing cells in room
temperature PBS (see Note 21) and wiping the exterior side
of the coverslip with 70% ethanol.
11. Wipe the objective lens with 70% ethanol and apply immersion
oil on the objective lens.
12. Mount the coverslip chamber on the microscope stage and find
the focus plane (see Note 22).
13. Set TIRF depth (<100 nm), appropriate excitation, filters, and
exposure in each fluorescent channel (see Note 14) and acquire
TIRF and widefield epifluorescence images for each channel.
14. Export files as TIFF or other lossless format. Image sample
pairs for each channel (TIRF and widefield epifluorescence), as
shown in Fig. 3a (top panels).
3.4 Fluorescence This quantification measures the relative localization of a specific

Microscopy Image PIP probe, e.g., either mCherry-PH(PLCδ1) or eGFP-PH(Akt), as
Analysis described, as well as the paired plasma membrane probe, e.g.,
mCherry-GPI or eGFP-GPI, expressed in the same cell (Fig. 2).
3.4.1 Quantification
of PI(4,5)P2 and PI(3,4)P2/ 1. Import image files into ImageJ as a time-lapse image stack (see
PIP3 Levels from SDCM Note 23).
Time-Lapse Image Series 2. Draw regions of interest (ROIs) using the straight-line tool
(e.g., set to 4 pixels wide, see Fig. 2a, b, lower panels) (see Note
24).
3. Using the method in step 2, for each cell, identify a total of

3 ROIs for each of the (A) cell surface (this is the outer “rim” of
the cell), (B) cell cytosol and (C) the extracellular space using
the first frame prior to viewing the time-lapse (see Fig. 2a, b,
bottom panels) (see Note 25).
4. For each ROI (one at a time), measure the mean pixel fluores-
cence intensity in each image in the time-lapse image series by
selecting the Analyze ! Plot Profile function, and then select-
ing the “List” option. Each row represents the measurement of
mean pixel intensity within the ROI in a sequential image in the
time-lapse. Export the data to a spreadsheet or other software.
As a result, for each frame in the time-lapse, each cell will have
three measurements for cell surface (A1, A2, A3), three for
cytosol (B1, B2, B3), and three for extracellular space (C1, C2,
C3).
5. To obtain a cell surface/cytosol ratio of each fluorescent probe
in each cell, in each frame of the time-lapse, use the values in
step 4 in the formula below:
cell surface ðmeanðA1 , A 2 , A 3 Þ meanðC1 , C2 , C3 ÞÞ

¼
cytosol ðmeanðB1 , B2 , B3 Þ meanðC1 , C2 , C3 ÞÞ
6. Repeat steps 4 and 5 for all cells within a time-lapse series.
Sample data are shown in Fig. 2b, d.
3.4.2 Quantification 1. Import files into ImageJ (see Note 23).

of PI(4,5)P2 Localization 2. Draw ROIs using the free-hand selection tool. To do so, trace
from TIRF and Widefield the outline of each cell (for example, Cell1, Fig. 3a, b, bottom
Epifluorescence panels, yellow dashed line) and 3 ROIs within the extracellular
Microscopy Images space of the image (for example, BG1-3, Fig. 3a, b, bottom
panels, red dashed line) (see Note 24).
3. Within these ROIs, measure the mean pixel fluorescence inten-
sity in both the TIRF and widefield epifluorescence (EPI)
images. To do so, within the ROI manager, select all ROIs to
be measured and click “Measure.” This will generate values for
Cell1TIRF, BG1TIRF, etc. in the TIRF channel and Cell1EPI,
BG1EPI, etc. in the widefield epifluorescence channel. Export
the data to a spreadsheet or other software.
4. For each cell, measure the TIRF/EPI ratio that reflects the cell
surface localization of each fluorescent probe by using the
following formula:
TIRF Cell1TIRF MeanðBG1TIRF, BG2TIRF, BG3TIRFÞ

¼
EPI Cell1EPI MeanðBG1EPI, BG2EPI, BG3EPIÞ
5. Repeat steps 2–4 for each image in an experiment. Sample data
are shown in Fig. 3b.
4 Notes
1. Addgene plasmids # 51407, 127812, 51465, 36075, and

86056.
2. Plasmids should be purified using cell culture transfection-
compatible methods, such as endotoxin-free plasmid maxi- or
midi-prep kits.
3. Cells are grown in T75 flasks and subcultured according to
ATCC guidelines [44, 51].
4. Commercially prepared coverslips may be coated with debris or
grease as a result of their manufacture [52]; thus, coverslips
should be prepared by acid washing the coverslips in 1 M HCl
at 50–60 C for 4–16 h prior to seeding cells [52, 53].
5. This method describes the use of Opti-MEM reduced serum
media (Gibco) and lipofectamine 2000 (ThermoFisher Scien-
tific); however, other reagents and media may be optimized
and used for transfection.
6. DMEM/F12 complete media and PBS are prewarmed to
37 C prior to culturing cells. Trypsin–EDTA solution is used
at room temperature.
7. Recommended seeding density is ~40–50% on the day of
transfection or a seeding density of 25% for next day transfec-
tion. Be aware of the doubling time of your cells and adjust
seeding density as needed. If needed, in addition to cells seeded
in a 6-well plate or similar, passage a fraction of cells to a new
T75 flask.
8. Cells are left overnight (~24 h) for optimal adhesion to cover-
slips prior to transfection.
9. Cells are co-transfected with two plasmids: one PIP reporter
plasmid (e.g., eGFP-PH(PLCδ) or eGFP-PH(AKT)) and one
plasma membrane reporter plasmid (e.g., mCherry-GPI or
mCherry-CAAX). Transfection may need to be optimized for
each cell type used in this assay.
10. Phenol red-free DMEM media is preferred as phenol red is
fluorescent when excited at ~450 nm and can interfere with
fluorescence detection of cyan and green fluorophores. Also,
serum starvation (~1 h) is recommended to allow cells to
exhibit low basal levels of specific signals, in particular, cell
surface PIP3 and PI(3,4)P2, as this best allows detection of
gains in the levels of these lipids elicited by acute treatments
(e.g., with specific growth factors such as EGF).
11. For live-cell imaging, 22 22 coverslips and corresponding
Chamlide CMS (Quorum) for square coverslips were used
here; however, other coverslips and matching chambers can
be used instead. During live-cell imaging, maintain cells at
constant 37 C with 5% CO2 in serum-free imaging media. Use

an initial volume of serum-free imaging media that is <50% of
the volume of the chamber (e.g., 200 μL for a 500-μL volume
chamber).
12. Prior to adding the immersion oil, wipe the lens with 70%
ethanol and a Kim Wipe. Ensure that the objective is
completely dry before applying immersion oil. This prepara-
tion step is necessary to avoid forming an oil–water emulsion
that leads to compromised image quality that manifests as
shade or fog in the image.
13. For MDA-MD-231 or ARPE-19 cells, viable cells typically
appear well spread and adherent to the coverslip. Avoid cells
that are rounded, partially adherent, or blebbing, as these may
not be viable.
14. Use the appropriate excitation illumination and filter(s) for
each fluorescent protein. For example, for imaging mCherry
and similar fluorophores (red channel), use 561-nm excitation
light and 630/75-nm emission filters, or similar, and for imag-
ing eGFP or similar fluorophores (green channel), use 488-nm
excitation light and 527/30-nm emission filters, or similar.
The exposure used to excite probes can be adjusted between
each experiment yet are typically ~50–100 ms (exposure range
will vary based on microscope and camera specifications).
Imaging conditions must be identical for all images acquired
in a single experiment.
15. When imaging cells involving stimulation with a drug or
growth factor, it is essential to have the working treatment
solution prepared prior to starting imaging. In addition, pre-
warm this solution to 37 C so that it is ready to be added to
the live-cell chamber. Prior to adding the stimulant, remove
the live-cell chamber lid and keep off for the duration of
imaging as replacing it after adding treatment could alter imag-
ing quality and focus.
16. For example, if in Subheading 3.3.1 cells are being imaged in
200 μL of serum-free imaging media in the imaging chamber,
add 200 μL of working stock EGF solution (40 ng/mL EGF) for
a final volume of 400 μL of media at a concentration of 20 ng/
mL EGF.
17. For example, if in Subheading 3.3.1 cells are being imaged in
200 μL of serum-free imaging media in the imaging chamber,
add 200 μL of working stock ionomycin solution (20 μM iono-
mycin) for a final volume of 400 μL of media at a concentration
of 10 μM ionomycin.
18. For a 6-well plate, add 1 mL of the fixation solution (4% PFA)
to each well. Adjust volume accordingly to your experimental
plate size.
19. For a 6-well plate, add 2 mL of the quenching solution

(100 mM glycine) to each well. Adjust volume accordingly to
your experimental plate size.
20. At this stage, coverslips can be stored for 24–48 h at 4 C and in
the dark after wrapping the plate in tinfoil.
21. Coverslips should not be mounted in commercially available
mounting medium for imaging by TIRF microscopy. Instead,
TIRF microscopy imaging of coverslips can be done by using
an imaging chamber (e.g., Chamlide CMS for square cover-
slips) with PBS as the imaging media. Fixed cells must always be
immersed in an imaging media to prevent drying and thus
damage to samples.
22. To achieve TIRF microscopy, you must use a combination of
imaging media, such as PBS, and coverslips that have a suffi-
cient difference in refractive indices that allows total internal
reflection [47].
23. Analysis is done using ImageJ software (National Institutes of
Health, Bethesda, MD) [50] or its Fiji distribution [54] that
contains many additional features including additional func-
tionality for opening a wide range of file types.
24. In ImageJ, save ROIs by the keystroke “t,” or by selecting
Analyze!Tools!ROI Manager, and clicking “Add.”
25. To limit variability and bias, select three different ROIs for each
of the three different areas (cell surface, cytosol, extracellular
space) prior to viewing the rest of the images in the time-lapse.
If a cell shifts during image acquisition, adjust the ROI position
such that it shifts in space to maintain position (e.g., at cell
surface) or add a new ROI in a region of the cell that does not
exhibit movement.
References
1. Choy CH, Han BK, Botelho RJ (2017) Phos- 5. Mccartney AJ, Zhang Y, Weisman LS (2014)
phoinositide diversity, distribution, and effec- Phosphatidylinositol 3,5-bisphosphate: low
tor function: stepping out of the box. abundance, high significance. BioEssays
BioEssays 39 36:52–64
2. Balla T (2013) Phosphoinositides: tiny lipids 6. De Matteis MA, Wilson C, D’Angelo G (2013)
with giant impact on cell regulation. Physiol Phosphatidylinositol-4-phosphate: the Golgi
Rev 93:1019–1137. https://doi.org/10. and beyond. BioEssays 35:612–622. https://
1152/physrev.00028.2012 doi.org/10.1002/bies.201200180
3. Hammond GRV, Burke JE (2020) Novel roles 7. Hawkins PT, Stephens LR (2015) PI3K signal-
of phosphoinositides in signaling, lipid trans- ling in inflammation. Biochim Biophys Acta
port, and disease. Curr Opin Cell Biol Mol Cell Biol Lipids 1851:882–897
63:57–67 8. Hirsch E, Gulluni F, Martini M (2020) Phos-
4. Schink KO, Raiborg C, Stenmark H (2013) phoinositides in cell proliferation and metabo-
Phosphatidylinositol 3-phosphate, a lipid that lism. Adv Biol Regul 75:100693. https://doi.
regulates membrane dynamics, protein sorting org/10.1016/j.jbior.2020.100693
and cell signalling. BioEssays 35:900–912.
https://doi.org/10.1002/bies.201300064
9. Han BK, Emr SD (2011) Phosphoinositide [PI Cell Biol 219:e201905162. https://doi.org/
(3,5)P2] lipiddependent regulation of the gen- 10.1083/jcb.201905162
eral transcriptional regulator Tup1. Genes Dev 21. Quon E, Sere YY, Chauhan N et al (2018)
25:984–995. https://doi.org/10.1101/gad. Endoplasmic reticulum-plasma membrane
1998611 contact sites integrate sterol and phospholipid
10. Stijf-Bultsma Y, Sommer L, Tauber M et al regulation. PLoS Biol 16:e2003864. https://
(2015) The basal transcription complex com- doi.org/10.1371/journal.pbio.2003864
ponent TAF3 transduces changes in nuclear 22. Sridhar S, Patel B, Aphkhazava D et al (2013)
phosphoinositides into transcriptional output. The lipid kinase PI4KIIIβ preserves lysosomal
Mol Cell 58:453–467. https://doi.org/10. identity. EMBO J 32:324–339. https://doi.
1016/j.molcel.2015.03.009 org/10.1038/emboj.2012.341
11. Mesmin B, Antonny B (2016) The counterflow 23. Rong Y, Liu M, Ma L et al (2012) Clathrin and
transport of sterols and PI4P. Biochim Biophys phosphatidylinositol-4,5-bisphosphate regu-
Acta 1861:940–951. https://doi.org/10. late autophagic lysosome reformation. Nat
1016/j.bbalip.2016.02.024 Cell Biol 14:924
12. Mesmin B, Bigay J, Moser von Filseck J et al 24. Tan X, Thapa N, Choi S, Anderson RA (2015)
(2013) A four-step cycle driven by PI(4)P Emerging roles of PtdIns(4,5)P2 - beyond the
hydrolysis directs sterol/PI(4)P exchange by plasma membrane. J Cell Sci 128:4047–4056
the ER-Golgi tether OSBP. Cell 25. Choi S, Chen M, Cryns VL, Anderson RA
155:830–843. https://doi.org/10.1016/j. (2019) A nuclear phosphoinositide kinase
cell.2013.09.056 complex regulates p53. Nat Cell Biol
13. Bousova K, Jirku M, Bumba L et al (2015) 21:462–475. https://doi.org/10.1038/
PIP2 and PIP3 interact with N-terminus s41556-019-0297-2
region of TRPM4 channel. Biophys Chem 26. Dong R, Saheki Y, Swarup S et al (2016)
205:24–32. https://doi.org/10.1016/j.bpc. Endosome-ER contacts control actin nucle-
2015.06.004 ation and Retromer function through
14. Dong X, Shen D, Wang X et al (2010) PI(3,5)P VAP-dependent regulation of PI4P. Cell
(2) controls membrane trafficking by direct 166:408–423. https://doi.org/10.1016/j.
activation of mucolipin Ca(2+) release channels cell.2016.06.037
in the endolysosome. Nat Commun 1:38. 27. Hammond GRV, Balla T (2015) Polypho-
https://doi.org/10.1038/ncomms1037 sphoinositide binding domains: key to inositol
15. Hansen SB (2015) Lipid agonism: the PIP2 lipid biology. Biochim Biophys Acta Mol Cell
paradigm of ligand-gated ion channels. Bio- Biol Lipids 1851:746–758
chim Biophys Acta Mol Cell Biol Lipids 28. Várnai P, Gulyás GG, Tóth DJ et al (2016)
1851:620–628 Quantifying lipid changes in various membrane
16. Huang CL, Feng SY, Hilgemann DW (1998) compartments using lipid binding protein
Direct activation of inward rectifier potassium domains. Cell Calcium 64:72–82. https://doi.
channels by PIP2 and its stabilization by G beta org/10.1016/j.ceca.2016.12.008
gamma. Nature 391:803–806. https://doi. 29. Idevall-Hagren O, De Camilli P (2015) Detec-
org/10.1038/35882 tion and manipulation of phosphoinositides.
17. Dickson EJ, Hille B (2019) Understanding Biochim Biophys Acta Mol Cell Biol Lipids
phosphoinositides: rare, dynamic, and essential 1851:736–745
membrane phospholipids. Biochem J 30. Wills RC, Goulden BD, Hammond GRV
476:1–23 (2018) Genetically encoded lipid biosensors.
18. Behnia R, Munro S (2005) Organelle identity Mol Biol Cell 29:1526–1532. https://doi.
and the signposts for membrane traffic. Nature org/10.1091/mbc.E17-12-0738
438:597–604. https://doi.org/10.1038/ 31. Stauffer TP, Ahn S, Meyer T (1998) Receptor-
nature04397 induced transient reduction in plasma mem-
19. Hammond GRV, Machner MP, Balla T (2014) brane PtdIns(4,5)P2 concentration monitored
A novel probe for phosphatidylinositol in living cells. Curr Biol 8:343–346. https://
4-phosphate reveals multiple pools beyond doi.org/10.1016/S0960-9822(98)70135-6
the Golgi. J Cell Biol 205:113–126. https:// 32. Lemmon MA, Ferguson KM, O’Brien R et al
doi.org/10.1083/jcb.201312072 (1995) Specific and high-affinity binding of
20. Du X, Zhou L, Aw YC et al (2020) ORP5 inositol phosphates to an isolated pleckstrin
localizes to ER-lipid droplet contacts and reg- homology domain. Proc Natl Acad Sci U S A
ulates the level of PI(4)P on lipid droplets. J 92:10472–10476. https://doi.org/10.1073/
pnas.92.23.10472
33. Várnai P, Balla T (1998) Visualization of phos- phosphorylation requires clathrin or ErbB2
phoinositides that bind pleckstrin homology but not receptor endocytosis. Mol Biol Cell
domains: calcium- and agonist-induced 26:3504–3519. https://doi.org/10.1091/
dynamic changes and relationship to mbc.E14-09-1412
myo-[3H]inositol-labeled phosphoinositide 44. Delos Santos R, Bautista S, Bone L et al (2017)
pools. J Cell Biol 143:501–510 Selective control of clathrin- mediated endocy-
34. Santagata S (2001) G-protein signaling tosis and clathrin-dependent signaling by phos-
through tubby proteins. Science pholipase C and Ca2+ signals. Mol Biol Cell
292:2041–2050. https://doi.org/10.1126/ 28:2802–2818
science.1061233 45. Botelho RJ, Teruel M, Dierckman R et al
35. Field SJ, Madson N, Kerr ML et al (2005) (2000) Localized biphasic changes in phospha-
PtdIns(4,5)P2 functions at the cleavage furrow tidylinositol-4,5-bisphosphate at sites of
during cytokinesis. Curr Biol 15:1407–1412. phagocytosis. J Cell Biol 151:1353–1368
https://doi.org/10.1016/j.cub.2005.06.059 46. Fish KN (2009) Total internal reflection fluo-
36. Servant G, Weiner OD, Herzmark P et al rescence (TIRF) microscopy. Curr Protoc
(2000) Polarization of chemoattractant recep- Cytom Chapter 12:Unit12.18. https://doi.
tor signaling during neutrophil chemotaxis. org/10.1002/0471142956.cy1218s50
Science 287:1037–1040. https://doi.org/10. 47. Mattheyses AL, Simon SM, Rappoport JZ
1126/science.287.5455.1037 (2010) Imaging with total internal reflection
37. Watton SJ, Downward J (1999) Akt/PKB fluorescence microscopy for the cell biologist.
localisation and 30 phosphoinositide generation J Cell Sci 123:3621–3628. https://doi.org/
at sites of epithelial cell-matrix and cell-cell 10.1242/jcs.056218
interaction. Curr Biol 9:433–436. https:// 48. Madugula V, Lu L (2016) A ternary complex
doi.org/10.1016/s0960-9822(99)80192-4 comprising transportin1, Rab8 and the ciliary
38. Manna D, Albanese A, Wei SP, Cho W (2007) targeting signal directs proteins to ciliary mem-
Mechanistic basis of differential cellular branes. J Cell Sci 129:3922–3934. https://doi.
responses of phosphatidylinositol 3,4-bispho- org/10.1242/jcs.194019
sphate- and phosphatidylinositol 3,4,5-tri- 49. Legler DF, Doucey MA, Schneider P et al
sphosphate-binding pleckstrin homology (2005) Differential insertion of GPI-anchored
domains. J Biol Chem 282:32093–32105. GFPs into lipid rafts of live cells. FASEB J
https://doi.org/10.1074/jbc.M703517200 19:73–75. https://doi.org/10.1096/fj.03-
39. Várnai P, Rother KI, Balla T (1999) Phospha- 1338fje
tidylinositol 3-kinase-dependent membrane 50. Schneider CA, Rasband WS, Eliceiri KW
association of the Bruton’s tyrosine kinase (2012) NIH image to ImageJ: 25 years of
pleckstrin homology domain visualized in sin- image analysis. Nat Methods 9:671–675
gle living cells. J Biol Chem 51. Fekri F, Abousawan J, Bautista S et al (2019)
274:10983–10989. https://doi.org/10. Targeted enhancement of flotillin-dependent
1074/jbc.274.16.10983 endocytosis augments cellular uptake and
40. Lu Q, Yu J, Yan J et al (2011) Structural basis impact of cytotoxic drugs. Sci Rep 9:17768.
of the myosin X PH1 N-PH2-PH1 C tandem https://doi.org/10.1038/s41598-019-
as a specific and acute cellular PI(3,4,5)P 3 sen- 54062-9
sor. Mol Biol Cell 22:4268–4278. https://doi. 52. Fischer AH, Jacobson KA, Rose J, Zeller R
org/10.1091/mbc.E11-04-0354 (2008) Preparation of slides and coverslips for
41. Goulden BD, Pacheco J, Dull A et al (2018) A microscopy. CSH Protoc 2008:pdb.prot4988.
high-avidity biosensor reveals plasma mem- https://doi.org/10.1101/PDB.PROT4988
brane PI(3,4)P2 is predominantly a class I 53. Lucarelli S, Delos Santos RC, Antonescu CN
PI3K signaling product. J Cell Biol 218 (2017) Measurement of epidermal growth fac-
(3):1066–1079. https://doi.org/10.1083/ tor receptor-derived signals within plasma
JCB.201809026 membrane clathrin structures. Methods Mol
42. Sugiyama MG, Fairn GD, Antonescu CN Biol 1652:191–225
(2019) Akt-ing up just about everywhere: 54. Schindelin J, Arganda-Carreras I, Frise E et al
compartment-specific Akt activation and func- (2012) Fiji: an open-source platform for
tion in receptor tyrosine kinase signaling. Front biological-image analysis. Nat Methods
Cell Dev Biol 7:70. https://doi.org/10.3389/ 9:676–682
FCELL.2019.00070
43. Garay C, Judge G, Lucarelli S et al (2015)
Epidermal growth factor-stimulated Akt
Chapter 6
Imaging the Nanoscale Distribution of Phosphoinositides

in the Cell Plasma Membrane with Single-Molecule
Localization Super-Resolution Microscopy
Fan Fan, Chen Ji, and Xuelin Lou
Abstract
Phosphoinositides make up only a small fraction of cellular phospholipids yet control cell function in a
fundamental manner. Through protein interactions, phosphoinositides define cellular organelle identity
and regulate protein function and organization and recruitment at the cytosol–membrane interface. As a
result, perturbations on phosphoinositide metabolism alter cell physiology and lead to a wide range of
human diseases, including cancer and diabetes. Among seven phosphoinositide members, phosphatidyli-
nositol 4,5-bisphosphate (PtdIns(4,5)P2, also known as PI(4,5)P2 or PIP2) is abundant in the plasma
membrane. Besides its role in the second messenger pathway of phospholipase C that cleaves PtdIns(4,5)P2
to form diacylglycerol and inositol-1,4,5-trisphosphate (IP3), PtdIns(4,5)P2 regulates membrane traffick-
ing and the function of the cytoskeleton, ion channels, and transporters. The nanoscale organization of
PtdIns(4,5)P2 in the plasma membrane becomes essential to understand cellular signaling specificity in time
and space. Here, we describe a single-molecule method to visualize the nanoscale distribution of PtdIns
(4,5)P2 in the plasma membrane by using super-resolution microscopy and the dual-color fluorescent
probes based on the PLCδ1 pleckstrin homology (PH) domain. This approach can be extended to image
other phosphoinositides by changing the specific probes.
Key words Membrane lipid, TIRFM, PALM, dSTORM, PtdIns(4,5)P2, INS-1 cells, Insulin secre-
tion, Exocytosis, Endocytosis, Cell signaling
1 Introduction
The nanoscale organization of the plasma membrane (PM) of

eukaryotic cells is not fully understood. It is clear that the PM
composes a complex topographic structure of proteins and lipids
that are mingled with protein clusters [1, 2] or lipid rafts [3]. These
types of molecule organizations facilitate the efficiency, fidelity, and
specificity of cellular signaling in time and space [3]. However, it is
unclear whether a similar organization pattern applies to phosphoi-
nositides (PIs) in the PM.
91
92 Fan Fan et al.
Phosphoinositides include a family of seven members that are

derived from reversible phosphorylation of the inositol ring at
position 3, 4, or/and 5 via PI kinases and phosphatases
[4, 5]. These PIs are compartmentalized in live cells with dynamic
equilibrium. PtdIns(4,5)P2 predominantly locates in the inner leaf-
let of the PM and participates in a variety of cellular processes [4, 6],
including signal transduction, actin dynamics, membrane traffick-
ing, and ion channel activity. PtdIns(4,5)P2 is produced through
PtdIns4P 5-kinase (PIP5K) activation and can be converted into
other PIs, as well as the second messenger inositol 1,4,5-
trisphosphate (IP3) and diacylglycerol [7]. Our previous work
demonstrates important function of PtdIns(4,5)P2 associated sig-
naling in synaptic transmission [8–10] and insulin secretion
[11]. PtdIns4P is synthesized by PI4KIIIα in the PM
[12, 13]. PtdIns4P and PtdIns(4,5)P2 determine the PM proper-
ties and subcellular localization of many PM proteins [13, 14]. In
contrast, PI(3,4,5)P3 has a lower level in the PM, but it also affects
actin dynamics, phagocytosis, and hormone secretion, among other
functions [15, 16].
PtdIns(4,5)P2 interacts with proteins through its negative
charges and structured domains, such as pleckstrin homology
(PH), ENTH/ANTH, FERM, FYVE, PX, and TARF domains
[17]. The development of a high-affinity PtdIns(4,5)P2 biosensor
(Kd ¼ 2 μM) based on the PHPLCδ1 domain [18] allows visualizing
PtdIns(4,5)P2 distribution and dynamics in live cells. However, the
spatial organization of PtdIns(4,5)P2 in the PM appears diverse and
controversial. Inconsistent results have been reported using differ-
ent cells and experimental systems, ranging from uniform distribu-
tion [19–21], large patches [22, 23], to small dense puncta [23–
26]. At the nanoscale level, the work using stimulated-emission
depletion (STED) microscopy reported the presence of abundant
dense PtdIns(4,5)P2 clusters (~73 nm in diameter) in PC12 PM
sheets [27], and such clusters drive syntaxin-1A clustering. In
contrast, another study using the rapid freezing electron micros-
copy (EM) reported loose microdomains of PtdIns(4,5)P2 but no
dense clusters [28, 29]. To better understand this issue, we have
developed a different method: single-molecule approach. We com-
bine it with the cell PM sheet method adapted from the previous
work [30, 31] to study PI nanoscale distribution in INS-1 cell
membrane [32, 33].
Single-molecule localization microscopy (SMLM) significantly
improves the spatial resolution of light microscopy because it marks
accurate coordinates of molecules when their fluorescent probes are
photoactivated or photoconverted at well-separated locations and
times, thus offering a better spatial resolution than structured-
illuminated microscopy (SIM). The use of genetic proteins allows
specific labeling, and the stunning spatial resolution makes SMLM
a powerful tool to study cell biology and macromolecule complexes
Single-Molecule Localization Microscopy of Phosphoinositides 93
[34–40]. The method offers a new way to study the subcellular

organization of PIs using optical microscopy, which avoids harsh
sample processing in conventional EM that can distort normal lipid
organization [28]. We apply the SMLM super-resolution method
to PtdIns(4,5)P2 in cellular PM, and a similar method may be
adapted to study the nanoscale distribution of other PIs.
2 Materials
2.1 DNA Constructs C1 and N1 cloning vectors (Clontech), PAmCherry1 [41],

of Photoswitchable PI mEos3.1 [38], and the following plasmids subcloned in our previ-
Probes for PALM ous work [32] including iRFP-PAmCherry1, iRFP-mEos3.1,
iRFP-PAmCherry1-PHPLCδ1, iRFP-PAmCherry1-PHosh2, and
iRFP-PAmCherry1-PHGRP1.
2.2 Cell Culture 1. Insulin-secreting INS-1 832/13 cells (from Dr. Christopher
B. Newgard, Duke University), passages 50–65. Other cells
grown on coverslips may be used similarly.
2. RPMI-1640 culture medium: 11.1 mM glucose, 10 mM
HEPES, 10 mM glutamine, 10 mM sodium pyruvate, 50 μM
β-mercaptoethanol, and 10% FBS (Atlanta Biological).
3. Round coverslips (#1.5, 18 mm in diameter).
4. Fibronectin, 30 μg/mL.
5. Transfection reagents. We use lipofectamine 3000 (Life Tech-
nologies) here; other similar reagents or transfection methods
may be used after optimization.
2.3 Membrane Sheet 1. 500 μg/mL poly-D-lysine.

Preparation 2. PBS buffer.
3. 1 mM EGTA.
4. A metal plate: We used a one-foot square metal plate to main-
tain the low temperature of the plasma membrane sheets; other
methods that keep samples at the constant temperature
may work.
5. Fixative: 4% paraformaldehyde (PFA), 0.2% glutaraldehyde
in PBS.
2.4 Primary 1. PBS.

and Secondary 2. 50 mM NH4Cl in PBS.
Antibodies for dSTORM
3. 0.1% sodium borohydride in PBS.
4. Blocking buffer: 5% (v/v) normal goat serum, 5% (v/v) BSA,
and 50 mM NH4Cl in PBS.
94 Fan Fan et al.
5. Primary antibody for PtdIns(4,5)P2 (Santa Cruz Biotechnol-

ogy, 1:300 dilution).
6. Secondary F(ab0 )2-goat anti-mouse, Alexa Fluor 647 (Invitro-
gen, 1:300).
2.5 Nanoscopy: 1. TetraSpeck beads (200-nm diameter, Invitrogen).

PALM and dSTORM 2. Imaging chambers with and without temperature control.
3. Dual-channel temperature controller.
4. Extracellular buffer for live-cell imaging: 135 mM NaCl,
5.6 mM KCl, 2.6 mM CaCl2, 1.2 mM MgCl2, 3 mM glucose,
and 20 mM HEPES (pH adjusted to 7.3 with NaOH).
5. dSTORM imaging buffer: 50 mM Tris–HCl, 10 mM NaCl,
10% glucose, 0.8 mg/mL glucose oxidase, 40 μg/mL catalase,
and 10 mM cysteamine (MEA) (Sigma), pH 8.0.
6. SMLM system. We have built the SMLM based on a Nikon
Ti-E Eclipse inverted microscope. It is capable of TIRFM and
spinning disk confocal imaging besides SMLM (PALM and
dSTORM). The system has the Ti-ND6-PFS Perfect Focus
and multiple objectives (CFI Plan Fluor 4; NA 0.13, WD
17.2 mm; CFI Plan APO λ 20, NA 0.75, WD 1.00 mm; Plan
APO VC 60 oil, NA 1.4, WD 0.13 mm; and APO 100 oil,
NA 1.49, WD 0.12 mm), a spinning disk (CSU X-1,
10,000 rpm, Yokogawa), a Ti-TIRF motorized illuminator
unit, and an Agilent MLC400 high-power monolithic laser
combiner SP with 405-, 488-, 561-, and 642-nm lasers. The
EMCCD (iXon X3 DU897, Andor) and Neo-sCMOS cameras
(Andor, 6.5 6.5 μm pixel size) are installed at the left and
right imaging ports to acquire confocal and TIRFM/SMLM
images, respectively, under the control of NIS-Elements AR
software. Other systems with similar functionalities may
be used.
7. MATLAB and MLE_GPU algorithm [32]. Other similar algo-
rithms and software with single-molecule localization function
may be used similarly.
3 Methods
3.1 Cloning 1. PCR cloning. We focused on the PALM probe of PtdIns(4,5)

the Dual-Color PALM P2 with PHPLCδ1 [18, 42] (from Tamas Balla at NIH) here, but
Probes other single-molecule probes for PI4P and PI(3,4,5)P3 can also
be generated.
2. Photoactivatable PAmCherry1 and photoswitchable mEos3.1
are PCR-amplified, which are used to replace enhanced GFP
(EGFP) in C1 and N1 cloning vectors (Clontech) with AgeI

and BglII and AgeI and NotI, respectively (see Note 1).
3. iRFP is PCR-amplified and inserted before the PAmCherry1
with NheI and AgeI to make the iRFP-PAmCherry1-C1 vector
(see Notes 2 and 3).
4. PHPLCδ1, PHosh2, and PHGRP1 domains (from Pietro DeCa-
milli at Yale) can be PCR-amplified and inserted after iRFP-
PAmCherry1 or mEos3.1 vector with BglII and BamHI to
generate the new PALM probes iRFP-PAmCherry1-PHPLCδ1,
iRFP-PAmCherry1-PHosh2, iRFP-PAmCherry1-PHGRP1, and
mEos3.1-PHPLCδ1. We have generated these probes in our
previous work [32, 33].
3.2 INS-1 Cell 1. Grow INS-1 cells on #1.5 (18 mm, round) coverslips pre-
Culture and DNA coated with 30 μg/mL of fibronectin with the standard culture
Transfection protocol [43]. Wait till cells reach ~60% confluence.
2. Transfect INS-1 cells with iRFP-PAmCherry1-PHPLCδ1 DNA
using Lipofectamine-3000 following the manufacturer’s pro-
tocol. iRFP-PAmCherry1 with ΔCMV promoter is transfected
in parallel as a control probe.
3. Continue growing cells for another 48 h till fluorescence
probes are visible. Use them for live-cell imaging or membrane
sheet generation (see Note 4). Figure 1a shows a typical distri-
bution of fluorescence probes in live INS-1 cells under TIRFM.
3.3 PM Sheet 1. To prepare coverslips, coat coverslips on only one side with
Generation ~0.5–1 mL of 500 μg/mL poly-D-lysine (PDL; diluted in
and Fixation dH2O) for 1–2 h, drain excessive PDL by placing a tissue
paper at coverslip edge, place the coverslips on the metal
plate, and keep it at 4 C in a refrigerator for later use.
2. Wash pre-transfected INS-1 cells three times with ~0.5–1 mL
cold (4 C) PBS containing 1 mM EGTA and carefully drain
PBS with tissue paper.
3. Place these coverslips with cells on the PDL-coated coverslips
on the pre-chilled metal plate with tweezers and keep the cells
and plate back in the refrigerator (4 C) for 8 min (see Note 5)
to allow cell attachment (Fig. 2).
4. Take out the metal plate from the refrigerator and gently peel
off the top coverslip using tweezers. This step produces a thin
layer of PM sheet on the PDL-coated coverslip.
5. Gently wash the PM sheets with ~0.5–1 mL of ice-cold PBS,
and then fix with ~0.5–1 mL of ice-cold fixative for 15 min at
4 C (see Notes 6 and 7).
6. Image the PM sheets or leave them in ~0.5–1 mL of PBS at
4 C for short storage in the dark (see Note 8).
96 Fan Fan et al.
Fig. 1 PI(4,5)P2 spatial organization in INS-1 cell PM. TIRFM images of the PM are taken at four different
conditions below: (a) intact PM of two live INS-1 cells expressing EGFP-PHPLCδ1 at low and elevated levels; (b)
the PM sheet fixed at 4 C with 4% PFA + 0.2% GA and immunostained with PI(4,5)P2 antibody; (c) the PM
sheet expressing iRFP-PAmCherry1-PHPLCδ1; and (d) the PM sheet fixed at room temperature with 4% PFA and
immunostained with PI(4,5)P2 antibody. Note that the images in (b and c) (but not d) recapitulate the spatial
distribution of fluorescence probes in live cells in (a). Scale bars: 5 μm in (a), 3 μm in (b–d)
3.4 PALM Imaging 1. Add TetraSpeck beads (prediluted with PBS containing extra
of the PM Sheets 50 mM MgCl2) to the imaging chamber, let them settle down
for 10 min on the coverslip, and then wash away those mobile
beads with PBS. Other drift-correction markers similar to Tet-
raSpeck beads may work as well.
2. Place the chamber onto the SMLM sample stage, and find the
PM sheets expressing PH probes in the iRFP-channel (642 nm
excitation, 700/75 nm emission) under the TIRF mode
(100 lens, NA ¼ 1.49).
3. Adjust the membrane sheets position so that the image area has
one or two TetraSpeck beads nearby (see Note 9).
4. Draw an imaging area covering both PM sheets and beads and
take an iRFP TIRF image (642 nm excitation, 700/75 nm
emission) as the reference of future imaging reconstruction.
Fig. 2 The scheme of membrane sheet preparation and PALM imaging with the dual-color probe in INS-1 cells.
(a) The membrane preparation cartoon. Place the cells on the coverslip (with cells facing down) onto a
PDL-coated coverslip (a), and after cell attachment for 7–10 min at 4 C, peel off the top coverslips and
produce the PM sheet (b), followed with fixation and super-resolution imaging (c). (b) The dual-color PtdIns
(4,5)P2 probe expressed in INS-1 cells for TIRFM and PALM. The peak wavelengths of excitation and emission
are indicated. iRFP is used to identify the cells with probe expression under TIRFM, without consuming
PAmCherry1 molecules before PALM acquisition
5. Switch to PAmCherry1-channel (561 nm excitation,

600/50 nm emission), and bleach the background fluores-
cence from the membrane using the full intensity of 561-nm
laser for 10–20 s.
6. Set imaging parameters such as camera binning (2 2, effec-
tive pixel size 130 130 nm) and exposure time, and the fast
acquisition protocol “ND sequence acquisition.”
7. Start imaging PAmCherry1 (561-nm laser at 100% power,
600/50 nm emission) at 50 ms/frame with simultaneous
405-nm laser excitation at low intensity to activate PAm-
Cherry1. The low 405 activation power is key to achieve better
imaging resolution (see Notes 10 and 11). All the images are
collected continuously.
8. The acquisition continues till PAmCherry1 signal is no longer
activated, with 10,000–40,000 images collected from each
INS-1 cell.
3.5 PALM Imaging This is performed as described above for the PM sheets with minor
in Live Cells changes.
1. Set up imaging temperature. Maintained cells at 35 C with the
temperature controller under constant perfusion of the extra-
cellular buffer.
98 Fan Fan et al.
2. Add TetraSpeck beads in the imaging chamber for 10 min as

described above.
3. Find INS-1 cells expressing mEOS3.1-PHPLCδ1 under TIRFM
at the green channel (488 nm excitation, 525/50 nm emission)
(see Note 12) and take an image with TetraSpeck beads.
4. Switch to the red channel (561 nm excitation, 600/50 nm
emission), bleach background fluorescence of the membrane
with the 561-nm laser for 10–20 s, and acquire PALM images
(561 nm excitation, 600/50 nm emission) with optimal
405 nm excitation (1% of full power) at 10 ms/frame (see
Notes 13–15).
3.6 Immunostaining 1. Prepare fixed membrane sheets from normal INS-1 cells at 4 C
the PM Sheets as described above (Subheading 3.3).
and dSTORM Imaging 2. Wash the coverslips three times with ice-cold PBS containing
50 mM NH4Cl, quench with 0.1% sodium borohydride in PBS
for 7 min, and wash with PBS w/o 50 mM NH4Cl.
3. Block the samples with blocking buffer.
4. Add the primary PtdIns(4,5)P2 antibody (1:300 dilution) in
blocking solution for 1 h. Wash three times (10 min each) with
PBS containing 50 mM NH4Cl.
5. Add the secondary F(ab0 )2-goat anti-mouse, Alexa Fluor
647 (Invitrogen, 1:300) in blocking buffer for 1 h, and then
wash three times. Other far-red fluorophores may be used, but
we choose Alexa Fluor 647 here because it has been well-
characterized and proven in SMLM application.
6. Fix samples with fixative for 15 min, and wash three times
(7 min each time), followed with dSTORM imaging. Other
samples are temporally kept in PBS at 4 C.
7. To perform dSTORM using the SMLM system next, prepare
fresh dSTORM imaging buffer.
8. Find the PM sheets and TetraSpeck beads, adjust acquisition
parameters and illumination angle (slightly larger than the
TIRF angle), and add dSTORM buffer to the samples.
9. Bleach the samples with the full intensity of a 642-nm laser, and
then adjust the UV laser to the optimal intensity for AF-647
photoactivation.
10. Collect images at 40 ms/frame (642 nm excitation,
700/75 nm emission) when individual fluorescent molecules
are spatially isolated. Approximately, 30,000 images are col-
lected from individual cells for reconstruction.
1. Transfer image files to the imaging analysis station, and convert

image stacks from “.nd2” to “.tiff” format. Launch the
18 Diameter=
383.3 ± 14.3nm
Microdomain num (#)

12
175
225
275
325
375
425
475
525
575
625
More
Patch size (nm)
Fig. 3 The single-molecule localization microscopy reveals the nanoscale distribution of PI(4,5)P2 probes in
the INS-1 cell PM. (a, b) TIRFM and PALM images of the INS-1 cell PM sheet expressing iRFP-PAmCherry1-
PHPLCδ1. (c1, c2) Enlarged views in (b). Note the homogeneous PI(4,5)P2 probe distribution in the major PM
area and some sparse microdomains (arrows) enriched with PI(4,5)P2 probes. (d) Enlarged view of the boxed
area in (C1). (d) The histogram of PI(4,5)P2 microdomain size in INS-1 membrane sheets (383.3 14.3 nm in
diameter). Scale bar: (a, b): 3 μm; (c1, c2): 500 nm; (d): 200 nm (Adapted from Ji et al., [32])
3.7 Point imaging reconstruction program customer-written in

Localization, Image MATLAB as described previously [32, 38].
Process, 2. Localize individual molecules from each image by a wavelet
and Reconstruction transform algorithm in a 7 7 pixel area. The precise locations
for Super-Resolution are determined by the local maximum of each fluorescence
Imaging point with a mask of 5 5 pixels after 2-D gaussian fitting
using the MLE_GPU algorithm [44]. Optimize the parameters
for point detection by visual inspection before the reconstruc-
tion (see Note 16). Minor drift is corrected with TetraSpeck
beads. A typical PALM image of the fixed membrane sheets is
shown in Fig. 3.
3. For PALM imaging of the fixed PM sheets expressing iRFP-
PAmCherry1-PHPLCδ1, all the image stacks from the same
sample are reconstructed and combined into a single super-
resolution image. For live-cell PALM, the image stack of
mEOS3.1-PHPLCδ1 is separated into several 1000-frame
stacks, and each stack is reconstructed as a single PALM
100 Fan Fan et al.
image (see Note 17). This produces a temporal resolution of

10 s in the time series of PALM images.
4. After the image reconstruction, quantify molecule density and
microdomain density and size with the home-written program
in MATLAB [32]. Cluster analysis is performed with pair-
correlation [45].
4 Notes
1. Both PAmCherry1 and mEos3.1 are used as PALM probes for

different purposes. The former shows a negligible level of
photoblinking and excellent monomeric property in cells
[46], and thus it is a good option for quantitative PALM.
The dual-color PAmCherry1 probe we generated shows typical
monomeric behavior with little photoblinking in cells [32],
suggesting that this probe in cellular environment preserves
the photophysical and biochemical properties of single PAm-
Cherry1 alone [36, 46, 47] and thus avoids overcounting and
probe oligomerization [32].
2. iRFP identifies the cells with PALM probe expression. This is
performed at the 647-channel to avoid activating PAmCherry1
or converting mEos3.1 (Fig. 2b) and reserve all the PALM
probe molecules for quantitative PALM imaging.
3. We have used a truncated CMV promoter (ΔCMV) [48] to
express iRFP-PAmCherry1 at a low level to avoid the probe
molecule oligomerization (as the single-molecule control). It is
suitable to characterize the single-molecule property (e.g.,
photo blinking and monomer distribution) of the dual-color
probe in a cellular environment.
4. Live-cell imaging must be acquired at a speed before the objec-
tives of interest move during acquisition. We used 200 ms/
frame to image the probes under TIRFM.
5. The incubation time is 7–10 min for cells to attach on the
PDL-coated surface. A longer duration should be avoided
since the refrigerator environment is dry.
6. PtdIns(4,5)P2 is sensitive to fixation conditions, such as fixa-
tion composition and temperature [32]. The sample fixation
conditions commonly used in immunocytochemistry can dis-
rupt PtdIns(4,5)P2 distribution and compromise data interpre-
tation (Fig. 1d). By comparing different fixation conditions in
INS-1 cells, we find that the fixation condition we used here
(4% PFA + 0.2% GA at 4 C) better preserves PtdIns(4,5)P2
distribution (Fig. 1b, c), where cells recapitulate their patterns
in live cells (Fig. 1a) and no adverse effect is seen. In addition,
the use of membrane sheets avoids detergents (e.g., Triton
X-100); the latter can induce artificial molecular clustering in

the PM [49].
7. Paraformaldehyde and glutaraldehyde are toxic. Handle them
with caution in a fume hood with the protection of skin
and eyes.
8. We recommend performing imaging experiments on the same
day after the PM sheet fixation for the optimal results. Despite
fixed, the bound PH probes on the PM may gradually dissoci-
ate from the membrane after long-term storage (days or
weeks).
9. This is needed later for drift correction during imaging recon-
struction. The mechanical stability of the imaging system is
critical for better resolution. Even with special attention and
effort, the drift correction is often needed to achieve the best
results.
10. We use 405-nm laser at 0.1–1% of full intensity. This number
may vary depending on the exact imaging system and fluores-
cence probes used. Ideally, it should activate probes at optimal
illumination so that the activated molecule density is high but
still isolated spatially from each other in each image. Increase
this number gradually to compensate for the decrease of avail-
able probes for photoactivation/conversion.
11. The density of the PALM probe is critical for the nanoscale
resolution. Changing the expression level of the PALM probe
shows comparable results in the PM sheets, suggesting the
excessive number of PALM probes in INS-1 cells. As a general
drawback of the PH domain-based probes, however, it remains
possible that the probes cannot access all the targets (e.g., those
bound PIs by some sequestering proteins) due to space hin-
drance or competition in cells.
12. The green fluorescence from mEos3.1 is used for cell identifi-
cation, without inducing photoconversion. We tune a TIRF
laser to a narrow illumination angle (controlled by the TIRF
angle panel) for live-cell imaging, and the resulting short pen-
etration depth under TIRFM minimizes the fluorescence from
a small number of unbound mEos3 probes in the cytosol.
13. The high-speed imaging is critical for live-cell imaging.
Although it is not an issue for confocal, TIRFM, or instant
SIM imaging, it is challenging for PALM, especially if objects
move fast. PALM needs a sufficient number of images before
object moves.
14. Since the number of unconverted mEOS3.1-PHPLCδ1 mole-
cules decreases during acquisition, one can increase the
405-nm laser power gradually to bring up the single-molecule
density.
102 Fan Fan et al.
15. The duration of the acquisition depends on the imaging sys-

tem, probe density, and dynamics, as well as cell tolerance to
phototoxicity. We saw no sign of photodamage (e.g., cell mor-
phology changes) during 5-min PALM acquisition. We limit
each image stack to less than 4 GB, a limit for the later
MATLAB analysis.
16. The point detection is critical for SMLM, both for the best
resolution and avoiding overcount artifact. The program
includes multiple parameters to optimize this process. We set
the neighborhood distance as 65 nm (half effective pixel
width); the fluorescence points within this distance are com-
bined and fitted as a single molecule. We set gap frame as
26 frames (1.3 s) for PAmCherry1 imaging [50] so that indi-
vidual events that occur within the gap frame and neighboring
distance are combined as a single molecule. Proper choosing of
these parameters not only impacts imaging resolution but also
avoids data misinterpretation such as the artificial clusters
resulting from overcounting [47]. The exact parameters
depend primarily on the single-molecule photophysics of
probes.
17. During high-speed PALM imaging in live cells, a small number
of activated probes may move or dissociate from PtdIns(4,5)P2
in the PM before bleaching and thus lead to oversampling. To
account for this issue, we set neighboring distance as 130 nm,
instead of 65 nm, for imaging analysis, which combines indi-
vidual molecule signals in this region as a single event.
Acknowledgments
This work is supported partially by the National Institutes of

Health (NIH) (R01DK093953 and R21NS101584) and the
grant AAB1425-135-A5362. We thank Tamas Balla at NIH and
Pietro De Camilli at Yale for the PH-domain constructs and
Dr. Christopher Newgard at Duke for INS-1 832/13 cells.
References
1. Engelman DM (2005) Membranes are more 4. Di Paolo G, De Camilli P (2006) Phosphoino-

mosaic than fluid. Nature 438(7068):578–580 sitides in cell regulation and membrane dynam-
2. Spira F, Mueller NS, Beck G, von Olshausen P, ics. Nature 443(7112):651–657
Beig J, Wedlich-Soldner R (2012) Patchwork 5. Vicinanza M, D’Angelo G, Di Campli A, De
organization of the yeast plasma membrane Matteis MA (2008) Function and dysfunction
into numerous coexisting domains. Nat Cell of the PI system in membrane trafficking.
Biol 14(6):640–648 EMBO J 27(19):2457–2470
3. Lingwood D, Simons K (2010) Lipid rafts as a 6. Balla T (2013) Phosphoinositides: tiny lipids
membrane-organizing principle. Science 327 with giant impact on cell regulation. Physiol
(5961):46–50 Rev 93(3):1019–1137
7. Berridge MJ, Irvine RF (1984) Inositol tri- 19. Milosevic I, Sorensen JB, Lang T, Krauss M,
sphosphate, a novel second messenger in cellu- Nagy G, Haucke V, Jahn R, Neher E (2005)
lar signal transduction. Nature 312 Plasmalemmal phosphatidylinositol-4,5-
(5992):315–321 bisphosphate level regulates the releasable vesi-
8. Lou X, Scheuss V, Schneggenburger R (2005) cle pool size in chromaffin cells. J Neurosci 25
Allosteric modulation of the presynaptic Ca2+ (10):2557–2565
sensor for vesicle fusion. Nature 435 20. van Rheenen J, Jalink K (2002) Agonist-
(7041):497–501 induced PIP(2) hydrolysis inhibits cortical
9. Korogod N, Lou X, Schneggenburger R actin dynamics: regulation at a global but not
(2007) Posttetanic potentiation critically at a micrometer scale. Mol Biol Cell 13
depends on an enhanced Ca(2+) sensitivity of (9):3257–3267
vesicle fusion mediated by presynaptic PKC. 21. Hammond G, Schiavo G, Irvine R (2009)
Proc Natl Acad Sci U S A 104 Immunocytochemical techniques reveal multi-
(40):15923–15928 ple, distinct cellular pools of PtdIns4P and
10. Lou X, Korogod N, Brose N, Schneggenbur- PtdIns (4, 5) P2. Biochem J 422:23–35
ger R (2008) Phorbol esters modulate sponta- 22. Huang S, Lifshitz L, Patki-Kamath V, Tuft R,
neous and Ca2+evoked transmitter release via Fogarty K, Czech MP (2004) Phosphatidyli-
acting on both Munc13 and protein kinase C. J nositol-4,5-bisphosphate-rich plasma mem-
Neurosci 28(33):8257–8267 brane patches organize active zones of
11. Ji C, Fan F, Lou X (2017) Vesicle docking is a endocytosis and ruffling in cultured adipocytes.
key target of local PI(4,5)P2 metabolism in the Mol Cell Biol 24(20):9102–9123
secretory pathway of INS-1 cells. Cell Rep 20 23. James DJ, Khodthong C, Kowalchyk JA, Mar-
(6):1409–1421 tin TF (2008) Phosphatidylinositol
12. Balla A, Kim YJ, Varnai P, Szentpetery Z, 4,5-bisphosphate regulates SNARE-dependent
Knight Z, Shokat KM, Balla T (2008) Mainte- membrane fusion. J Cell Biol 182(2):355–366
nance of hormone-sensitive phosphoinositide 24. Laux T, Fukami K, Thelen M, Golub T,
pools in the plasma membrane requires phos- Frey D, Caroni P (2000) GAP43, MARCKS,
phatidylinositol 4-kinase IIIalpha. Mol Biol and CAP23 modulate PI(4,5)P(2) at plasma-
Cell 19(2):711–721 lemmal rafts, and regulate cell cortex actin
13. Nakatsu F, Baskin JM, Chung J, Tanner LB, dynamics through a common mechanism. J
Shui G, Lee SY, Pirruccello M, Hao M, Ingolia Cell Biol 149(7):1455–1472
NT, Wenk MR, De Camilli P (2012) PtdIns4P 25. Aoyagi K, Sugaya T, Umeda M, Yamamoto S,
synthesis by PI4KIIIalpha at the plasma mem- Terakawa S, Takahashi M (2005) The activa-
brane and its impact on plasma membrane tion of exocytotic sites by the formation of
identity. J Cell Biol 199(6):1003–1016 phosphatidylinositol 4,5-bisphosphate micro-
14. Hammond GR, Fischer MJ, Anderson KE, domains at syntaxin clusters. J Biol Chem 280
Holdich J, Koteci A, Balla T, Irvine RF (17):17346–17352
(2012) PI4P and PI(4,5)P2 are essential but 26. Kabachinski G, Yamaga M, Kielar-Grevstad
independent lipid determinants of membrane DM, Bruinsma S, Martin TF (2014) CAPS
identity. Science 337(6095):727–730 and Munc13 utilize distinct PIP2-linked
15. Czech MP (2000) PIP2 and PIP3: complex mechanisms to promote vesicle exocytosis.
roles at the cell surface. Cell 100(6):603–606 Mol Biol Cell 25(4):508–521
16. Czech MP (2003) Dynamics of phosphoinosi- 27. van den Bogaart G, Meyenberg K, Risselada
tides in membrane retrieval and insertion. HJ, Amin H, Willig KI, Hubrich BE, Dier M,
Annu Rev Physiol 65:791–815 Hell SW, Grubmuller H, Diederichsen U, Jahn
17. Hammond GR, Balla T (2015) Polyphosphoi- R (2011) Membrane protein sequestering by
nositide binding domains: key to inositol lipid ionic protein-lipid interactions. Nature 479
biology. Biochim Biophys Acta 1851 (7374):552–555
(6):746–758 28. van Rheenen J, Achame EM, Janssen H,
18. Garcia P, Gupta R, Shah S, Morris AJ, Rudge Calafat J, Jalink K (2005) PIP2 signaling in
SA, Scarlata S, Petrova V, McLaughlin S, lipid domains: a critical re-evaluation. EMBO
Rebecchi MJ (1995) The pleckstrin homology J 24(9):1664–1673
domain of phospholipase C-delta 1 binds with 29. Sato K, Ernstrom GG, Watanabe S, Weimer
high affinity to phosphatidylinositol RM, Chen CH, Sato M, Siddiqui A, Jorgensen
4,5-bisphosphate in bilayer membranes. Bio- EM, Grant BD (2009) Differential require-
chemistry 34(49):16228–16234 ments for clathrin in receptor-mediated endo-
cytosis and maintenance of synaptic vesicle
104 Fan Fan et al.
pools. Proc Natl Acad Sci U S A 106 nanoscopy using sCMOS camera-specific sin-
(4):1139–1144 gle-molecule localization algorithms. Nat
30. Sanan DA, Anderson R (1991) Simultaneous Methods 10(7):653–658
visualization of LDL receptor distribution and 40. Sengupta P, Van Engelenburg S, Lippincott-
clathrin lattices on membranes torn from the Schwartz J (2012) Visualizing cell structure
upper surface of cultured cells. J Histochem and function with point-localization superreso-
Cytochem 39(8):1017–1024 lution imaging. Dev Cell 23(6):1092–1102
31. Morone N, Fujiwara T, Murase K, Kasai RS, 41. Subach FV, Patterson GH, Manley S, Gillette
Ike H, Yuasa S, Usukura J, Kusumi A (2006) JM, Lippincott-Schwartz J, Verkhusha VV
Three-dimensional reconstruction of the mem- (2009) Photoactivatable mCherry for high-
brane skeleton at the plasma membrane inter- resolution two-color fluorescence microscopy.
face by electron tomography. J Cell Biol 174 Nat Methods 6(2):153–159
(6):851–862 42. Lemmon MA, Ferguson KM, O’Brien R, Sigler
32. Ji C, Zhang Y, Xu P, Xu T, Lou X (2015) PB, Schlessinger J (1995) Specific and high-
Nanoscale landscape of Phosphoinositides affinity binding of inositol phosphates to an
revealed by the specific PH-domains using isolated pleckstrin homology domain. Proc
single-molecule super-resolution imaging in Natl Acad Sci 92(23):10472–10476
the plasma membrane. J Biol Chem 290 43. Dyachok O, Isakov Y, Sågetorp J, Tengholm A
(45):26978–26993 (2006) Oscillations of cyclic AMP in hormone-
33. Ji C, Lou X (2016) Single-molecule super-res- stimulated insulin-secreting β-cells. Nature
olution imaging of phosphatidylinositol 439(7074):349–352
4,5-bisphosphate in the plasma membrane 44. Smith CS, Joseph N, Rieger B, Lidke KA
with novel fluorescent probes. J Vis Exp (2010) Fast, single-molecule localization that
116:54466 achieves theoretically minimum uncertainty.
34. Szymborska A, de Marco A, Daigle N, Cordes Nat Methods 7(5):373–375
VC, Briggs JA, Ellenberg J (2013) Nuclear 45. Sengupta P, Jovanovic-Talisman T, Skoko D,
pore scaffold structure analyzed by super- Renz M, Veatch SL, Lippincott-Schwartz J
resolution microscopy and particle averaging. (2011) Probing protein heterogeneity in the
Science 341(6146):655–658 plasma membrane using PALM and pair corre-
35. Betzig E, Patterson GH, Sougrat R, Lindwas- lation analysis. Nat Methods 8(11):969–975
ser OW, Olenych S, Bonifacino JS, Davidson 46. Durisic N, Laparra-Cuervo L, Sandoval-
MW, Lippincott-Schwartz J, Hess HF (2006) Álvarez Á, Borbely JS, Lakadamyali M (2014)
Imaging intracellular fluorescent proteins at Single-molecule evaluation of fluorescent pro-
nanometer resolution. Science 313 tein photoactivation efficiency using an in vivo
(5793):1642–1645 nanotemplate. Nat Methods 11(2):156–162
36. Subach FV, Patterson GH, Manley S, Gillette 47. Nan X, Collisson EA, Lewis S, Huang J, Tam-
JM, Lippincott-Schwartz J, Verkhusha VV güney TM, Liphardt JT, McCormick F, Gray
(2009) Photoactivatable mCherry for high- JW, Chu S (2013) Single-molecule superreso-
resolution two-color fluorescence microscopy. lution imaging allows quantitative analysis of
Nat Methods 6(2):153–159 RAF multimer formation and signaling. Proc
37. Pertsinidis A, Mukherjee K, Sharma M, Pang Natl Acad Sci U S A 110(46):18519–18524
ZP, Park SR, Zhang Y, Brunger AT, Sudhof 48. Barg S, Knowles MK, Chen X, Midorikawa M,
TC, Chu S (2013) Ultrahigh-resolution imag- Almers W (2010) Syntaxin clusters assemble
ing reveals formation of neuronal SNARE/ reversibly at sites of secretory granules in live
Munc18 complexes in situ. Proc Natl Acad Sci cells. Proc Natl Acad Sci U S A 107
U S A 110(30):E2812–E2820 (48):20804–20809
38. Zhang M, Chang H, Zhang Y, Yu J, Wu L, 49. Heerklotz H (2002) Triton promotes domain
Ji W, Chen J, Liu B, Lu J, Liu Y, Zhang J, formation in lipid raft mixtures. Biophys J 83
Xu P, Xu T (2012) Rational design of true (5):2693–2701
monomeric and bright photoactivatable fluo- 50. Alabi AA, Tsien RW (2013) Perspectives on
rescent proteins. Nat Methods 9(7):727–729 kiss-and-run: role in exocytosis, endocytosis,
39. Huang F, Hartwich TM, Rivera-Molina FE, and neurotransmission. Annu Rev Physiol
Lin Y, Duim WC, Long JJ, Uchil PD, Myers 75:393–422
JR, Baird MA, Mothes W, Davidson MW,
Toomre D, Bewersdorf J (2013) Video-rate
Chapter 7
Induced Dimerization Tools to Deplete Specific

Phosphatidylinositol Phosphates
Jonathan Pacheco, Rachel C. Wills, and Gerald R. V. Hammond
Abstract
Chemical dimerization systems have been used to drive acute depletion of polyphosphoinsitides (PPIns).
They do so by inducing subcellular localization of enzymes that catabolize PPIns. By using this approach, all
seven PPIns can be depleted in living cells and in real time. The rapid permeation of dimerizer agents and
the specific expression of recruiter proteins confer great spatial and temporal resolution with minimal cell
perturbation. In this chapter, we provide detailed instructions to monitor and induce depletion of PPIns in
live cells.
Key words Lipid-binding domain, Dimerization, Rapamycin, Kinase, Phosphatase
1 Introduction
Polyphosphoinositides (PPIns) are central regulators of membrane

function in eukaryotic cells [1], executing their functions by bind-
ing and thereby regulating scores of cytosolic and membrane pro-
teins. Dissecting PPIn function requires either identifying these
proteins and studying function through traditional molecular
genetic tools to manipulate protein expression, or modulating the
lipids themselves in an experimental capacity. This latter approach
has proved challenging. PPIns can be sequestered by overexpres-
sing PPIn-binding proteins, though this is often harder than typi-
cally realized [2]. Alternatively, enzymes that synthesize or degrade
the lipids can be overexpressed or knocked-down, though this
causes chronic effects over several hours. Chemically induced
dimerization approaches (and the more recent photoinduced
dimerization) provide a powerful approach to induce rapid removal
of lipids from precise intracellular locations, allowing acute effects
on downstream biology to be interrogated [3, 4]. In this chapter,
we describe such an approach using the FKBP/FRB rapamycin-
inducible dimerization system.
105
106 Jonathan Pacheco et al.
1.1 Chemical The origins of chemical dimerization systems come from pioneer-
Dimerization Systems ing work from Crabtree and co-workers, who rationalized that
several receptors at the plasma membrane are activated when form-
ing dimers. Crabtree’s group successfully activated T-cell signaling
by forcing dimer formation of cytosolic portions of T-cell receptors
fused to FKBP12, an immunophilin that binds immunosuppressant
drugs such as cyclosporine A (CsA), rapamycin, or FK506 (which
gave the name to these receptors: FK506-Binding Protein) [5].
The extended use of immunophilins as a heterodimerization
system was adopted thanks to Schreiber’s work, which showed that
FKBP12 interacts with a minimal portion of the mechanistic Target
of Rapamycin (mTOR), called FKBP12 Rapamycin-Binding (FRB)
domain [6]. The use of FRB domain as an FKBP partner presented
a major improvement in comparison with previous systems. The
main advantages are the size, since FRB comprises only 90 amino
acid residues and 11 kDa instead of 2549 amino acid residues and
289 kDa of the whole human mTOR [6]. FRB also binds FKBP12
with nanomolar binding affinity when rapamycin is used as a dimer-
izing agent. FRB and FKBP12 do not interact in the absence of
rapamycin, and FRB shows a modest interaction with rapamycin by
itself (Kd ¼ 26 μM) [7].
1.2 Rapamycin A whole set of ligands are used to induce dimerization. Most
to Induce Dimerization desired attributes for ligands are high permeability, low molecular
weight, high affinity, and high specificity. All of them exhibit mini-
mal secondary effects with rapid induction of dimerization. Selec-
tion of the best ligand must be based on the experimental design to
contrast the advantages and disadvantages of available drugs. For
example, rapamycin is the most commonly used compound for
experiments on live cells with measurable effects in short periods
of time [8]. However, rapamycin is a potent teratogenic agent in
mouse embryos, precluding its use for long incubation times [9]. In
addition, rapamycin binds to FKBP12 via a pipecolate moiety. The
immunosuppressive effect of rapamycin occurs through inhibition
of mTOR, a serine/threonine protein kinase that regulates cell
growth and proliferation [10, 11]. As consequence, the rapamy-
cin–FKBP12 complex stops cell cycle progression in the G1/S
phase [12], preventing clonal expansion and effector functions of
T-cells [13].
In addition to mTOR, FKBP12 also interacts and regulates
calcineurin, ryanodine receptors [14], TGF-β [15], and inositol
[1, 4, 5]-triphosphate (IP3) receptors [16] and presents rotamase
activity that interconverts cis–trans isomerization on proline resi-
dues at specific peptides [17]. Much of these interactions are pro-
duced in the presence of FK506 and not by rapamycin. The
extensive literature concerning this topic should be considered to
avoid potential side effects [15, 16, 18–23]. In order to reduce the
risk of secondary effects, several synthetic ligands based on
Tools to Deplete Phosphatidylinositol Phosphate 107
rapamycin, generically called “rapalogs,” have been synthesized

[8, 24–29]. Whatever the drug employed, suitable control experi-
ments must be performed to ensure that biological effects are due
to manipulation of intended enzymatic activity, and not the dimer-
izer or FKBP12/FRB proteins or combinations thereof. Moreover,
another way to induce dimerization can bring versatility, such as the
use of photoactivatable and reversible dimerizers, with many
options available commercially [30–33].
2 Materials
Generally speaking, this protocol uses a dimerization system to

target localization of a phosphatase to deplete a specific pool of
PPIns (Fig. 1). A membrane-bound recruiter carrying an FRB
domain is used to target the enzyme fused with the FKBP domain.
Simultaneously, a lipid-binding domain fused to a fluorescent pro-
tein can be used to observe how lipid levels change at the specific
cellular location. A prototypical protocol of depletion of PtdIns
(4,5)P2 from the plasma membrane (PM) is described here. How-
ever, lipid biosensors, enzymes, and membrane recruiters are inter-
changeable according to the experimenter’s scope (Table 1).
2.1 Preparation A wide range of phosphatases and biosensors for PPIns fused with a
of Plasmids full spectrum of fluorescent protein are available from Addgene.
For detailed directions about plasmid manipulations, refer to
Chap. 4 (see Note 1).
As an example, PtdIns(4,5)P2 depletion is described in this
protocol by using the following plasmids:
1. piRFP-N1-Lyn11-FRB (recruiter).
2. pTagBFP2-N1-Tubby (lipid biosensor).
3. pmCherry-C1-FKBP-INPP5E (phosphatase that hydrolyzes
5-phosphate of PtdIns(4,5)P2).
2.2 Cell Culture Standard cell culture and transfection materials are used:
and Transfection
1. TrpLE (Life Technologies 12604039) for cell dissociation.
2. Low glucose DMEM (Life Technologies) supplemented with
penicillin (100 U/mL), streptomycin (100 μg/mL), 10% of
heat-inactivated fetal bovine serum (FBS), and 0.1% chemically
defined lipid supplement (Life Technologies).
3. T75 flask (VWR International) and 35-mm glass-bottom
dishes with 20-mm glass aperture (In Vitro Scientific).
4. Opti-MEM (Life Technologies).
5. Human Plasma Fibronectin (Life Technologies).
Fig. 1 Target localization at different cellular compartments. (a) Targeted recruitment of chimeric phosphatase
PJ-Sac to specific subcellular localization. Giantin, Lamp1, and AKAP1 are membrane anchor proteins that
target localization to the Golgi apparatus, lysosomes, and mitochondria, respectively. (b) Localized depletion
of PtdIns(4,5)P2 from the PM after addition of rapamycin. Lyn11-FRB-iRFP is used to drag the phosphatase
INPP5E to the PM. Acute depletion of PtdIns(4,5)P2 is registered simultaneously with the biosensor PH-PLCδ1-
GFP. Scale bars in A and B ¼ 10 μm
6. Phosphate-buffered saline (PBS).

7. Lipofectamine 2000.
8. Cell culture incubator at 37 in a humidified atmosphere with
5% CO2.
Table 1
Representative phosphoinositide phosphatases: cellular function and experimental use
Lipid Cellular
Categorization Type Phosphatase substrate Product location Recruitment tested Dysregulation effect References
3- Voltage TPIP PtdIns PtdIns ER and Golgi Voltage sensing Not reported [34]
phosphatases sensing (3,4,5)P3 (4,5)P2 domain is used to
phosphatase produce chimeras
activated by
depolarization
Myotubularins MTM and PtdIns(3,5) PtdIns(5) PM ✓ l MTM1/ is lethal at [35, 36]
MTMR P2 and P and an early stage
PtdIns(3) PtdIns l Loss of MTM2R2
P causes neuropathies
PTEN PtdIns PtdIns PM, cytosol, ✓ l Increment of cell [37, 38]
(3,4,5)P3 (4,5)P2 Golgi, and proliferation and
and PtdIns and nucleus migration
(3,4)P2 PtdIns l PTEN/ is lethal at
(4)P an early stage

l Dysregulation of PTEN
produces cancer and

Cowden disease
4-phosphatases Sac1 PtdIns(4)P PtdIns ER and Golgi ✓ l Golgi membranes and [39–41]
and PtdIns mitotic spindles
(3)P disruption
l Sac1/ is lethal in
mice
INPP4A/B PtdIns(3,4) PtdIns(3) Cytoplasm ✓ l INPP4A: Asthma [38, 42]
P2 P and early l INPP4B: Cancer
endosomes
Tools to Deplete Phosphatidylinositol Phosphate
(continued)
109
Table 1
110
(continued)
Lipid Cellular
Categorization Type Phosphatase substrate Product location Recruitment tested Dysregulation effect References
5-phosphatases II ORCL PtdIns(4,5) PtdIns(4) Golgi, ✓ l Lowe syndrome [43, 44]
P2, PtdIns P, endosomes,
(3,4,5)P3, PtdIns and
and PtdIns (3,4)P2, lysosomes
(3,5)P2 and
Jonathan Pacheco et al.
PtdIns
(3)P
II Synaptojanins PtdIns PtdIns Synaptic ✓ l Synj1/ is lethal at an [45, 46]
(3,4,5)P3 (3,4)P2 terminals, early stage
and PtdIns and mito- l Accumulation of
(3,5)P2 PtdIns chondria Clathrin-coated

(3)P vesicles
l Upregulated expression
of Synj1 is associated
with Down’s
syndrome,
schizophrenia, and
bipolar disorder
II INPP5B PtdIns(4,5) PtdIns(4) Phagosomes Not tested l Infertility [42, 47,
P2 and P and and 48]
PtdIns PtdIns endosomes
(3,4,5)P3 (3,4)P2
II SKIP PtdIns(4,5) PtdIns(4) ER and PM Not tested l Suppression of insulin [42, 49]
P2 and P and sensitivity in skeletal
PtdIns PtdIns muscle
(3,4,5)P3 (3,4)P2 l SKIP/ is lethal in
mice
III SHIP1 PtdIns PtdIns PM and Not tested l Leukemia [50]
(3,4,5)P3 (3,4)P2 nucleus l Acute myelogenous
and PtdIns and
(4,5)P2 PtdIns
(4)P
SHIP2 PtdIns(4,5) PtdIns(4) PM and Not tested l Diabetes [46, 51]
P2, PtdIns P, cytoplasm
(3,4,5)P3, PtdIns
and PtdIns (3,4)P2,
(3,5)P2 and
PtdIns
(3)P
IV INPP5E PtdIns PtdIns Cilia and ✓ l Polycystic kidney [52–54]
(3,4,5)P3 (3,4)P2 centriole disease and ciliopathy
and PtdIns and syndromes
(4,5)P2 PtdIns
(4)P
Tools to Deplete Phosphatidylinositol Phosphate
111
2.3 Live Cell Imaging The steps provided below describe an experiment performed on a
confocal microscope. However, this protocol is also applicable for
the TIRF microscope, depending on the scope of the experiment.
1. FluoroBrite DMEM (Life Technologies) supplemented with
10% of heat-inactivated fetal bovine serum, 25 mM HEPES
(pH 7.4), and 0.1% chemically defined lipid supplement (see
Note 2).
2. 1 mM rapamycin in DMSO. Store in 20-μL aliquots at 20 C.
3. Confocal microscope equipped with at least three laser lines.
Experiments are optimized for a Nikon Eclipse TiE inverted
microscope equipped with a 100, plan apochromatic, 1.45
NA oil-immersion objective lens and a Nikon A1R confocal
scan head with a laser unit (LU-NV) containing four laser lines
(405, 488, 561, and 640 nm).
4. Motorized stage.
5. Image acquisition software such as Elements (Nikon).
2.4 Image Analysis 1. Open-access image analysis package Fiji.

2. Analysis software such as GraphPad Prism.
3 Methods
Instructions to maintain cell culture and transfection procedures

are standard. Particular indications are provided in Chap. 4.
3.1 Transfection Per 35-mm glass-bottom dish at ~50% confluence:
1. Dilute 3.1 μL of lipofectamine 2000 into 96.9 μL of Opti-
MEM in a sterile 1.5-mL plastic tube and mix.
2. Dilute up to 1 μg of total plasmid DNA into a total of 100 μL of
Opti-MEM in a 1.5-mL plastic tube and mix. As a preliminary
experiment, begin with the following masses of DNA (see Note
3):
(a) 300 ng of piRFP-N1-Lyn11-FRB.
(b) 300 ng of pmCherry-C1-FKBP-INPP5E.
(c) 400 ng of pTagBFP2-N1-Tubby (see Note 4).
3. Combine tubes containing plasmids and lipofectamine
dilution.
4. Incubate for 20 min at room temperature to allow lipid–DNA
complex formation.
5. Gently add complex lipid–DNA to cells and mix by swirling
the dish.
6. Incubate for 4 h and replace the transfection mix with 2 mL of
fresh DMEM.
7. Cells are imaged 18–26 h after transfection.
3.2 Microscope 1. Use a fiber-coupled laser device equipped with 405-nm (for
Setup TagBFP2), 561-nm (for mCherry), and 640-nm (for iRFP)
laser lines to excite the indicated fluorescent protein (see Note
5).
2. Collect fluorescence emission with bandwidths at 425–475 nm
(for TagBFP2), 570–620 nm (for mCherry), and 663–737 nm
(for iRFP). Differential interference contrast (DIC) is used on a
transmitted light channel (see Note 6).
3. Our experiments are optimized with the Nikon A1R laser
scanning confocal microscope. If available, use the resonant
mode of A1R confocal scan head (see Note 7).
4. Set pinhole at 1.2 Airy disc size based on the longest wave-
length channel (see Note 8).
5. A motorized stage is used to register up to 16 fields for each
time frame during the whole recording (see Note 9).
6. Our experiments are optimized with the acquisition software
“Elements” by Nikon (see Note 10).
7. Configure software acquisition to take a sequence of frames
each 30 s for 17 m (see Note 11).
3.3 Time-Lapse 1. Turn on a microscope at least 20 min before starting the

Imaging and experiment to allow stabilization of the optical system.
Rapamycin Addition 2. Prepare a fresh dilution of rapamycin at 5 μM in FluorBrite
DMEM (see Note 12).
3. Clean 100 objective lens with appropriate wipes and cleaning
solution.
4. Apply immersion oil on the 100 plan apochromatic, 1.45 NA
objective lens.
5. Before placing the dish on the microscope, remove growth
medium and rinse once with 1 mL of pre-warmed
FluoroBrite DMEM.
6. Add 1.6 mL of FluroBrite DMEM and mount the dish on the
microscope.
7. Find the focus plane and look for cells showing fluorescent
signal in the three channels (see Note 13).
8. Save the number of positions to register during the experiment
(see Note 14).
9. Start acquisition allowing 2 min of baseline before inducing
recruitment of pmCherry-C1-FKBP-INPP5E with rapamycin
(see Note 15).
10. Add 400 μL of 5 μM rapamycin (see Note 16).
11. Continue acquisition and save experiment.
3.4 Data Analysis Using the above plasmids and protocol, depletion of PtdIns(4,5)P2
on the PM can be performed. In addition, the strategy described
above leaves the green channel to record an additional fluorescent
reporter (e.g., a second GFP-tagged lipid biosensor). Using these
image data, we can obtain fluorescence intensity measurements on a
given cellular location for each channel. These selective measure-
ments over a cellular compartment of interest are done by produc-
ing a binary mask. Usually, the recruiter channel is used to
construct the mask.
3.4.1 Analysis of Images 1. Import images into the open-access image analysis platform Fiji
Obtained by Confocal (see Note 17).
Microscopy 2. Single series are analyzed, first producing a binary mask using
the recruiter channel (piRFP-N1-Lyn11-FRB) for each frame
along the whole time-series (see Note 18).
3. Draw regions of interest (ROIs) containing the whole cell (see
Note 19).
4. Measure fluorescence intensity on each mask at a specific time
frame for all channels, including the recruiter channel (see Note
20).
5. Normalize signal between experiments by using the ratio of
fluorescence intensity raw data divided by the fluorescence at
the first frame for the same cell.
6. Expected results must show an increase in fluorescence inten-
sity of pmCherry-C1-FKBP-INPP5E after inducing dimeriza-
tion with rapamycin. As consequence, a decrease of the signal
of pTagBFP2-N1-Tubby must occur and no changes on
piRFP-N1-Lyn11-FRB are expected to happen.
3.4.2 Analysis of Images 1. Import files into Fiji using the LOCI Bio-Formats importer.
Obtained by TIRF 2. Draw an ROI in the background of the image(s) to be
Microscopy analyzed.
3. Measure the intensity of the background ROI.
4. Subtract the background data from the images.
5. Draw ROIs around labeled cells in each frame (see Note 21).
6. Measure the intensity within the ROI in each channel at each
timepoint.
7. Copy and paste data into GraphPad Prism and format as
desired.
4 Notes
1. Keep in mind the topology of protein of interest. The FKBP-

tagged protein must be cytosolic. If you are working with an
enzyme that has a membrane-interacting domain, it will be
necessary to remove it and test that the enzymatic function of
protein remains unaltered. Verify plasmids by DNA sequencing
before using them.
2. Experiments were optimized for this reagent. Fluorobrite is a
DMEM-based formulation presenting low background to
enhance signal-to-noise ratio on fluorophores.
3. The amount of FRB expression is limiting to recruit FKBP
(Fig. 2). For this reason, the first time the experiment is run,
different ratios of FRB/FKBP expression should be tested.
Because transfection efficiency is variable for each plasmid and
several factors affect the expression of the protein of interest,
this step is experimentally determined by changing plasmid
concentrations during transfection (Fig. 2).
Fig. 2 Rapamycin induces recruitment at different efficiencies of FRB-FKBP ratios. Plasmid encoding Lyn11-
FRB-iRFP transfected at a ratio of 1.4:1 with plasmid encoding mCherry-FKBP produces little apparent
localization change. When the plasmid encoding Lyn11-FRB-iRFP is transfected at a ratio of 3:1 to the
plasmid encoding mCherry-FKBP instead, then rapamycin induces clear localization of mCherry at the PM.
Scale Bar ¼ 10 μm
4. An appropriate spectral separation is performed by using these

three fluorescent proteins. In addition, this configuration
allows an extra emission window to add another fluorophore
in the green channel.
5. Under this configuration, a 488-nm laser line is available to
excite another fluorophore. This excitation is most commonly
used to excite GFP with an emission of 500–550 nm.
6. To avoid crosstalk, images are sequentially acquired by avoid-
ing spectral overlapping: TagBFP2 followed by mCherry, then
GFP, and lastly iRFP.
7. Other microscope systems equipped with similar settings as
mentioned above can be used.
8. Increasing Airy disc size increases signal but decreases both Z
and X–Y resolution.
9. A total of 8 or 16 frames are averaged to improve the signal to
noise ratio. This step is optional, but registering more than one
field per dish makes for more efficient and robust data
gathering and statistically tractable results. It also minimizes
cell variability effects and obtains more reliable data when
single treatment experiments are performed, as occurs with
irreversible chemical dimerizers. The number of positions is
determined as the maximum number of fields accommodated
by the time frame.
10. Other systems and proprietary image acquisition software may
be used, though the user may need to optimize.
11. High affinity and fast permeability of rapamycin induces dimer-
ization in the order of seconds. In this way, kinetics of deple-
tion of PPIns can be recorded in 5 min, with 2 min of baseline
signal. For custom experiments, establish time-lapses by con-
sidering the time frame of changes that will be observed. This
later point is determined in pilot experiments.
12. Final concentration of rapamycin in cells is 1 μM. A dilution
1:5 is used at the moment of the experiment.
13. Set confocal Z-axis position on the equatorial section of cells to
reduce out-of-focus changes during acquisition. In addition,
the maximal signal on the PM is recorded at this z-section.
14. Increase movement efficiency of motorized stage by using a
looping pattern.
15. Pre-addition frames will be used to normalize baseline
correction.
16. Add rapamycin gently after the fifth frame capture.
17. Use LOCI’s Bio-Formats plugin to open images on a standard
ImageJ compatible format. A series containing information of
three or four channels will be open. Each series represents one
field of acquisition along the time.
18. Detailed instructions about the mask constructions are

provided in Chap. 4.
19. Measurements of total fluorescent intensity are performed on
the respective mask for each frame; ROIs delimit signal by cell
in order to get an average signal of all the cells.
20. Basic protocol registers the depletion of specific PPIns on the
recruiter localization, but another subcellular region can be
monitored simultaneously, taking advantage of additional
channels.
21. Draw each ROI around a minimum intensity projection to
capture PM footprint, which is always in contact with the
glass-bottom dish.
Acknowledgments
This work was supported by National Institutes of Health grant

1R35GM119412-01 (to G.R.V. Hammond).
References
1. Dickson EJ, Hille B (2019) Understanding residue. Proc Natl Acad Sci U S A 92
phosphoinositides: rare, dynamic, and essential (11):4947–4951. https://doi.org/10.1073/
membrane phospholipids. Biochem J 476 pnas.92.11.4947
(1):1–23. https://doi.org/10.1042/ 7. Banaszynski LA, Liu CW, Wandless TJ (2005)
BCJ20180022 Characterization of the FKBP.Rapamycin.FRB
2. Wills RC, Goulden BD, Hammond GRV ternary complex. J Am Chem Soc 127
(2018) Genetically encoded lipid biosensors. (13):4715–4721. https://doi.org/10.1021/
Mol Biol Cell 29(13):1526–1532. https:// ja043277y
doi.org/10.1091/mbc.E17-12-0738 8. Bayle JH, Grimley JS, Stankunas K, Gestwicki
3. Fegan A, White B, Carlson JC, Wagner CR JE, Wandless TJ, Crabtree GR (2006) Rapa-
(2010) Chemically controlled protein assem- mycin analogs with differential binding speci-
bly: techniques and applications. Chem Rev ficity permit orthogonal control of protein
110(6):3315–3336. https://doi.org/10. activity. Chem Biol 13(1):99–107. https://
1021/cr8002888 doi.org/10.1016/j.chembiol.2005.10.017
4. DeRose R, Miyamoto T, Inoue T (2013) 9. Hentges KE, Sirry B, Gingeras AC,
Manipulating signaling at will: chemically- Sarbassov D, Sonenberg N, Sabatini D, Peter-
inducible dimerization (CID) techniques son AS (2001) FRAP/mTOR is required for
resolve problems in cell biology. Pflugers Arch proliferation and patterning during embryonic
465(3):409–417. https://doi.org/10.1007/ development in the mouse. Proc Natl Acad Sci
s00424-012-1208-6 U S A 98(24):13796–13801. https://doi.org/
5. Spencer DM, Wandless TJ, Schreiber SL, Crab- 10.1073/pnas.241184198
tree GR (1993) Controlling signal transduc- 10. Brown EJ, Albers MW, Shin TB, Ichikawa K,
tion with synthetic ligands. Science 262 Keith CT, Lane WS, Schreiber SL (1994) A
(5136):1019–1024. https://doi.org/10. mammalian protein targeted by G1-arresting
1126/science.769436510.1126/science. rapamycin-receptor complex. Nature 369
7694365 (6483):756–758. https://doi.org/10.1038/
6. Chen J, Zheng XF, Brown EJ, Schreiber SL 369756a0
(1995) Identification of an 11-kDa FKBP12- 11. Sabatini DM, Erdjument-Bromage H, Lui M,
rapamycin-binding domain within the Tempst P, Snyder SH (1994) RAFT1: a mam-
289-kDa FKBP12-rapamycin-associated pro- malian protein that binds to FKBP12 in a
tein and characterization of a critical serine rapamycin-dependent fashion and is
homologous to yeast TORs. Cell 78(1):35–43. 21. Cameron AM, Steiner JP, Roskams AJ, Ali SM,
https://doi.org/10.1016/0092-8674(94) Ronnett GV, Snyder SH (1995) Calcineurin
90570-3 associated with the inositol 1,4,5-trisphosphate
12. Albers MW, Williams RT, Brown EJ, Tanaka A, receptor-FKBP12 complex modulates Ca2+
Hall FL, Schreiber SL (1993) FKBP-rapamycin flux. Cell 83(3):463–472. https://doi.org/
inhibits a cyclin-dependent kinase activity and a 10.1016/0092-8674(95)90124-8
cyclin D1-Cdk association in early G1 of an 22. Shin DW, Pan Z, Bandyopadhyay A, Bhat MB,
osteosarcoma cell line. J Biol Chem 268 Kim DH, Ma J (2002) Ca(2+)-dependent
(30):22825–22829 interaction between FKBP12 and calcineurin
13. Dumont FJ, Su Q (1996) Mechanism of action regulates activity of the Ca(2+) release channel
of the immunosuppressant rapamycin. Life Sci in skeletal muscle. Biophys J 83
58(5):373–395. https://doi.org/10.1016/ (5):2539–2549. https://doi.org/10.1016/
0024-3205(95)02233-3 s0006-3495(02)75265-x
14. Jayaraman T, Brillantes AM, Timerman AP, 23. Wang T, Li BY, Danielson PD, Shah PC,
Fleischer S, Erdjument-Bromage H, Rockwell S, Lechleider RJ, Martin J,
Tempst P, Marks AR (1992) FK506 binding Manganaro T, Donahoe PK (1996) The
protein associated with the calcium release immunophilin FKBP12 functions as a common
channel (ryanodine receptor). J Biol Chem inhibitor of the TGF beta family type I recep-
267(14):9474–9477 tors. Cell 86(3):435–444. https://doi.org/10.
15. Wang T, Donahoe PK, Zervos AS (1994) Spe- 1016/s0092-8674(00)80116-6
cific interaction of type I receptors of the 24. Pollock R, Issner R, Zoller K, Natesan S, Rivera
TGF-beta family with the immunophilin VM, Clackson T (2000) Delivery of a stringent
FKBP-12. Science 265(5172):674–676. dimerizer-regulated gene expression system in
https://doi.org/10.1126/science.7518616 a single retroviral vector. Proc Natl Acad Sci U
16. Cameron AM, Steiner JP, Sabatini DM, Kaplin S A 97(24):13221–13226. https://doi.org/
AI, Walensky LD, Snyder SH (1995) Immuno- 10.1073/pnas.230446297
philin FK506 binding protein associated with 25. Terrillon S, Bouvier M (2004) Receptor
inositol 1,4,5-trisphosphate receptor modu- activity-independent recruitment of βarrestin2
lates calcium flux. Proc Natl Acad Sci U S A reveals specific signalling modes. EMBO J
92(5):1784–1788 20-23:3950–3961. https://doi.org/10.
17. Fischer G, Wittmann-Liebold B, Lang K, 1038/sj.emboj.7600387
Kiefhaber T, Schmid FX (1989) Cyclophilin 26. Muthuswamy SK, Gilman M, Brugge JS
and peptidyl-prolyl cis-trans isomerase are (1999) Controlled dimerization of ErbB recep-
probably identical proteins. Nature 337 tors provides evidence for differential signaling
(6206):476–478. https://doi.org/10.1038/ by homo- and heterodimers. Mol Cell Biol 19
337476a0 (10):6845–6857. https://doi.org/10.1128/
18. Brillantes AB, Ondrias K, Scott A, Kobrinsky E, mcb.19.10.6845
Ondriasova E, Moschella MC, Jayaraman T, 27. Amara JF, Clackson T, Rivera VM, Guo T,
Landers M, Ehrlich BE, Marks AR (1994) Sta- Keenan T, Natesan S, Pollock R, Yang W,
bilization of calcium release channel (ryano- Courage NL, Holt DA, Gilman M (1997) A
dine receptor) function by FK506-binding versatile synthetic dimerizer for the regulation
protein. Cell 77(4):513–523. https://doi. of protein-protein interactions. Proc Natl Acad
org/10.1016/0092-8674(94)90214-3 Sci U S A 94(20):10618–10623. https://doi.
19. Ahern GP, Junankar PR, Dulhunty AF (1994) org/10.1073/pnas.94.20.10618
Single channel activity of the ryanodine recep- 28. Liberles SD, Diver ST, Austin DJ, Schreiber SL
tor calcium release channel is modulated by (1997) Inducible gene expression and protein
FK-506. FEBS Lett 352(3):369–374. translocation using nontoxic ligands identified
https://doi.org/10.1016/0014-5793(94) by a mammalian three-hybrid screen. Proc Natl
01001-3 Acad Sci U S A 94(15):7825–7830. https://
20. Cameron AM, Nucifora FC Jr, Fung ET, doi.org/10.1073/pnas.94.15.7825
Livingston DJ, Aldape RA, Ross CA, Snyder 29. Pollock R, Giel M, Linher K, Clackson T
SH (1997) FKBP12 binds the inositol 1,4,5- (2002) Regulation of endogenous gene expres-
trisphosphate receptor at leucine-proline sion with a small-molecule dimerizer. Nat Bio-
(1400-1401) and anchors calcineurin to this technol 20(7):729–733. https://doi.org/10.
FK506-like domain. J Biol Chem 272 1038/nbt0702-729
(44):27582–27588. https://doi.org/10. 30. Briesewitz R, Ray GT, Wandless TJ, Crabtree
1074/jbc.272.44.27582 GR (1999) Affinity modulation of small-
molecule ligands by borrowing endogenous 40. Liu Y, Boukhelifa M, Tribble E, Morin-

protein surfaces. Proc Natl Acad Sci U S A 96 Kensicki E, Uetrecht A, Bear JE, Bankaitis VA
(5):1953–1958. https://doi.org/10.1073/ (2008) The Sac1 phosphoinositide phospha-
pnas.96.5.1953 tase regulates Golgi membrane morphology
31. Feng S, Laketa V, Stein F, Rutkowska A, and mitotic spindle organization in mammals.
MacNamara A, Depner S, Klingmuller U, Mol Biol Cell 19(7):3080–3096. https://doi.
Saez-Rodriguez J, Schultz C (2014) A rapidly org/10.1091/mbc.E07-12-129010.1091/
reversible chemical dimerizer system to study mbc.e07-12-1290
lipid signaling in living cells. Angew Chem Int 41. Liu Y, Boukhelifa M, Tribble E, Bankaitis VA
Ed Engl 53(26):6720–6723. https://doi.org/ (2009) Functional studies of the mammalian
10.1002/anie.201402294 Sac1 phosphoinositide phosphatase. Adv
32. Schifferer M, Feng S, Stein F, Schultz C (2017) Enzym Regul 49(1):75–86. https://doi.org/
Reversible chemical dimerization by rCD1. 10.1016/j.advenzreg.2009.01.006
Methods Enzymol 583:173–195. https://doi. 42. Sasaki T, Takasuga S, Sasaki J, Kofuji S,
org/10.1016/bs.mie.2016.10.035 Eguchi S, Yamazaki M, Suzuki A (2009) Mam-
33. Umeda N, Ueno T, Pohlmeyer C, Nagano T, malian phosphoinositide kinases and phospha-
Inoue T (2011) A photocleavable rapamycin tases. Prog Lipid Res 48(6):307–343. https://
conjugate for spatiotemporal control of small doi.org/10.1016/j.plipres.2009.06.001
GTPase activity. J Am Chem Soc 133 43. Attree O, Olivos IM, Okabe I, Bailey LC, Nel-
(1):12–14. https://doi.org/10.1021/ son DL, Lewis RA, McInnes RR, Nussbaum
ja108258d RL (1992) The Lowe’s oculocerebrorenal syn-
34. Murata Y, Iwasaki H, Sasaki M, Inaba K, Oka- drome gene encodes a protein highly homolo-
mura Y (2005) Phosphoinositide phosphatase gous to inositol polyphosphate-5-phosphatase.
activity coupled to an intrinsic voltage sensor. Nature 358(6383):239–242. https://doi.org/
Nature 435(7046):1239–1243. https://doi. 10.1038/358239a0
org/10.1038/nature03650 44. Idevall-Hagren O, Dickson EJ, Hille B,
35. Laporte J, Hu LJ, Kretz C, Mandel JL, Toomre DK, De Camilli P (2012) Optogenetic
Kioschis P, Coy JF, Klauck SM, Poustka A, control of phosphoinositide metabolism. Proc
Dahl N (1996) A gene mutated in X-linked Natl Acad Sci U S A 109(35):E2316–E2323.
myotubular myopathy defines a new putative https://doi.org/10.1073/pnas.1211305109
tyrosine phosphatase family conserved in yeast. 45. Chang-Ileto B, Frere SG, Chan RB, Voronov
Nat Genet 13(2):175–182. https://doi.org/ SV, Roux A, Di Paolo G (2011) Synaptojanin
10.1038/ng0696-175 1-mediated PI(4,5)P2 hydrolysis is modulated
36. Fili N, Calleja V, Woscholski R, Parker PJ, Lar- by membrane curvature and facilitates mem-
ijani B (2006) Compartmental signal modula- brane fission. Dev Cell 20(2):206–218.
tion: endosomal phosphatidylinositol https://doi.org/10.1016/j.devcel.2010.12.
3-phosphate controls endosome morphology 008
and selective cargo sorting. Proc Natl Acad 46. Dyson JM, Fedele CG, Davies EM,
Sci U S A 103(42):15473–15478. https:// Becanovic J, Mitchell CA (2012) Phosphoino-
doi.org/10.1073/pnas.0607040103 sitide phosphatases: just as important as the
37. Leslie NR, Downes CP (2004) PTEN func- kinases. Subcell Biochem 58:215–279.
tion: how normal cells control it and tumour https://doi.org/10.1007/978-94-007-3012-
cells lose it. Biochem J 382(Pt 1):1–11. 0_7
https://doi.org/10.1042/bj20040825 47. Bohdanowicz M, Balkin DM, De Camilli P,
38. Goulden BD, Pacheco J, Dull A, Zewe JP, Grinstein S (2012) Recruitment of OCRL
Deiters A, Hammond GRV (2019) A high- and Inpp5B to phagosomes by Rab5 and
avidity biosensor reveals plasma membrane PI APPL1 depletes phosphoinositides and attenu-
(3,4)P2 is predominantly a class I PI3K signal- ates Akt signaling. Mol Biol Cell 23
ing product. J Cell Biol 218(3):1066–1079. (1):176–187. https://doi.org/10.1091/mbc.
https://doi.org/10.1083/jcb.201809026 E11-06-0489
39. Zewe JP, Wills RC, Sangappa S, Goulden BD, 48. Williams C, Choudhury R, McKenzie E, Lowe
Hammond GR (2018) SAC1 degrades its lipid M (2007) Targeting of the type II inositol
substrate PtdIns4P in the endoplasmic reticu- polyphosphate 5-phosphatase INPP5B to the
lum to maintain a steep chemical gradient with early secretory pathway. J Cell Sci 120
donor membranes. Elife 7:e35588. https:// (Pt 22):3941–3951. https://doi.org/10.
doi.org/10.7554/eLife.35588 1242/jcs.014423
49. Ijuin T, Yu YE, Mizutani K, Pao A, Tateya S, Riley JH, Perez-Morga D, Woods CG, Schur-
Tamori Y, Bradley A, Takenawa T (2008) mans S (2009) INPP5E mutations cause pri-
Increased insulin action in SKIP heterozygous mary cilium signaling defects, ciliary instability
knockout mice. Mol Cell Biol 28 and ciliopathies in human and mouse. Nat
(17):5184–5195. https://doi.org/10.1128/ Genet 41(9):1027–1031. https://doi.org/10.
mcb.01990-06 1038/ng.427
50. Luo JM, Yoshida H, Komura S, Ohishi N, 53. Bielas SL, Silhavy JL, Brancati F, Kisseleva MV,
Pan L, Shigeno K, Hanamura I, Miura K, Al-Gazali L, Sztriha L, Bayoumi RA, Zaki MS,
Iida S, Ueda R, Naoe T, Akao Y, Ohno R, Abdel-Aleem A, Rosti RO, Kayserili H,
Ohnishi K (2003) Possible dominant-negative Swistun D, Scott LC, Bertini E,
mutation of the SHIP gene in acute myeloid Boltshauser E, Fazzi E, Travaglini L, Field SJ,
leukemia. Leukemia 17(1):1–8. https://doi. Gayral S, Jacoby M, Schurmans S,
org/10.1038/sj.leu.2402725 Dallapiccola B, Majerus PW, Valente EM,
51. Marion E, Kaisaki PJ, Pouillon V, Gueydan C, Gleeson JG (2009) Mutations in INPP5E,
Levy JC, Bodson A, Krzentowski G, Daubresse encoding inositol polyphosphate-5-
JC, Mockel J, Behrends J, Servais G, Szpirer C, phosphatase E, link phosphatidyl inositol sig-
Kruys V, Gauguier D, Schurmans S (2002) The naling to the ciliopathies. Nat Genet 41
gene INPPL1, encoding the lipid phosphatase (9):1032–1036. https://doi.org/10.1038/
SHIP2, is a candidate for type 2 diabetes in rat ng.423
and man. Diabetes 51(7):2012–2017. https:// 54. Hammond GR, Machner MP, Balla T (2014) A
doi.org/10.2337/diabetes.51.7.2012 novel probe for phosphatidylinositol
52. Jacoby M, Cox JJ, Gayral S, Hampshire DJ, 4-phosphate reveals multiple pools beyond
Ayub M, Blockmans M, Pernot E, Kisseleva the Golgi. J Cell Biol 205(1):113–126.
MV, Compere P, Schiffmann SN, Gergely F, https://doi.org/10.1083/jcb.201312072
Chapter 8
A Real-Time Phosphatidylinositol 4-Phosphate 5-Kinase

Assay Using Fluorescence Spectroscopy
Taki Nishimura
Abstract
Phosphatidylinositol 4-phosphate 5-kinase (PIP5K) is an enzyme that converts phosphatidylinositol
4-phosphate [PI4P] to phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2]. PIP5K plays a key role in the
regulation of vesicular transport, cytoskeleton reorganization, and cell division. In general, to investigate an
enzymatic activity of PIP5K, the amount of incorporated [P32] ATP into PI(4,5)P2 fraction is measured in
in vitro reconstitution experiments. However, tools to monitor dynamic changes in its activity in real time
have been lacking. Recently, we have developed a novel PIP5K assay using fluorescence spectroscopy.
Compared to conventional methods in which lipids extraction steps are needed, our method is easy and
quick to perform and enables a real-time analysis. This chapter provides a protocol to set up and perform the
novel PIP5K assay we have recently established.
Key words Phosphatidylinositol 4-phosphate 5-kinase, Phosphatidylinositol 4-phosphate, Phospha-

tidylinositol (4,5)-bisphosphate, Pleckstrin homology domain, 7-nitrobenz-2-oxa-1,3-diazol-4-yl
1 Introduction
Phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2] is a key lipid in

the plasma membrane (PM) and controls fundamental PM func-
tions including exocytosis, endocytosis, cytoskeleton dynamics,
cytokinesis, and ion fluxes [1]. PI(4,5)P2 regulates membrane tar-
geting of various effector proteins via its electrostatic interaction
with lipid-binding domains, such as the Pleckstrin homology
(PH) domain and the Phox homology (PX) domain [1]. In addi-
tion, PI(4,5)P2 modulates ligand sensitivity of many ion channels
by binding to transmembrane domains and causing a conforma-
tional change of ion channels [2, 3]. Therefore, it is thought that PI
(4,5)P2 metabolism should be strictly regulated to maintain the PM
homeostasis.
Several different techniques are utilized to analyze PI(4,5)P2
metabolism. In particular, a fluorescence-tagged PHPLCδ, a PI(4,5)
P2 probe, has been widely used to monitor dynamic changes in PI
121
122 Taki Nishimura
PIP5K
530nm
PI4P PI4,5P2
NBD-
PHPLCδ
Em 530nm
Fig. 1 Scheme for the real-time PIP5K assay. PIP5K and PI4P liposomes are mixed with a PI(4,5)P2 sensor
NBD-PHPLCδ. After addition of ATP, PIP5K converts PI4P to PI(4,5)P2 on liposomes, followed by binding of
NBD-PHPLCδ to liposomes containing PI(4,5)P2. As NBD is designed to be located at the interface between
PHPLCδ and membranes, its fluorescence intensity is increased in response to PI(4,5)P2 production
(4,5)P2 levels and distribution in cells [4]. In contrast, a method to

measure PI(4,5)P2 levels using in vitro reconstitution experiments
is not conveniently designed. To analyze newly synthesized PI(4,5)
P2 during an in vitro reaction, PIP5K reaction is performed using
[32P] ATP, and then, lipid fractions are needed to be extracted to
measure the incorporation of [32P] ATP into PI(4,5)P2 fraction.
Several steps are required to obtain a final output [5].
Recent studies have shown that 7-nitrobenz-2-oxa-1,3-diazol-
4-yl (NBD)-labeled lipid-binding domains are useful to measure
lipid levels on liposomes without the need for lipid extraction
[6, 7]. As NBD displays a blue shift accompanied by an enhance-
ment of fluorescence quantum yield in a hydrophobic environment,
NBD fluorescence intensity is increased when NBD-labeled lipid-
binding domains interact with liposomes. By taking advantage of
this technique, we have recently designed an NBD-labeled PHPLCδ
to monitor PI(4,5)P2 levels on liposomes and applied it to an
in vitro PIP5K assay for a real-time analysis (Fig. 1) [8]. Here, I
will describe the procedures that are performed for a real-time
PIP5K assay in vitro.
2 Materials
2.1 Preparation of 1. Plasmid: pET21b(+)-zPIP5K1Aα(49-431aa) (see Note 1) [9].

Zebrafish PIP5K 2. E. coli competent cells: BL21 (DE3) pLysS.
(zPIP5K) Recombinant
Protein
A Real-Time PIP5K Assay 123
3. 2 YT medium: 10% yeast extract (w/v), 16% tryptone (w/v),

5% NaCl (w/v) in Milli-Q water (ddH2O). Make 1 L, storage
at 4 C for 6 months.
4. Kanamycin sulfate: 50 mg/mL stock solution in ddH2O, fil-
tered 0.22 μm. Make 10 mL, storage as 1-mL aliquots at
20 C for up to 6 months.
5. Dithiothreitol (DTT): 1 M stock solution in ddH2O. Make
1 mL, storage as 0.2-mL aliquots at 20 C for up to
6 months.
6. Imidazole stock solution: 2 M imidazole, 500 mM NaCl,
20 mM Tris–HCl, pH 7.5 in ddH2O. Make 5 mL, storage at
4 C for 1 month.
7. Pre-wash buffer for His-tagged protein: 50 mM Tris–HCl,
pH 6.8, 300 mM NaCl, 1 mM DTT, 10 mM imidazole in
ddH2O. Make 40 mL, prepare just prior to use.
8. Complete EDTA-free protease inhibitor.
9. AEBSF protease inhibitor: 0.1 M stock solution in ddH2O.-
Make 1 mL, storage as 0.2-mL aliquots at 20 C for up to
6 months.
10. Homogenization buffer for His-tagged protein: 50 mM Tris–
HCl, pH 6.8, 300 mM NaCl, 1 mM DTT, 10 mM imidazole,
0.1 mM AEBSF, complete protease inhibitor in ddH2O. Make
20 mL, prepare just prior to use.
11. Branson Sonifier 250 (Branson).
12. Ni-NTA agarose.
13. Isopropyl-β-thiogalactopyranoside (IPTG): 1 M stock solution
in ddH2O, filtered 0.22 μm. Make 20 mL, storage as 1-mL
aliquots at 20 C for up to 6 months.
14. Dialysis buffer: 50 mM Tris–HCl, pH 6.8, 150 mM NaCl,
1 mM DTT in ddH2O. Make 1.5 L, prepare just prior to use.
15. Slide-A-Lyzer Dialysis Cassettes, 3.5 K MWCO (Thermo
Fisher Scientific).
16. Storage buffer: 50 mM Tris–HCl, pH 6.8, 150 mM NaCl,
2 mM DTT, 50% glycerol in ddH2O. Make 300 mL, prepare
just prior to use.
2.2 Preparation of 1. Plasmids: pGEX6P-1-PHPLCδV58C [8].

PHPLCδV58C 2. E. coli competent cells: BL21 (DE3) pLysS.
Recombinant Protein
3. 2 YT medium: 10% yeast extract (w/v), 16% tryptone (w/v),
5% NaCl (w/v) in Milli-Q water (ddH2O). Make 1 L, storage
at 4 C for 6 months.
124 Taki Nishimura
4. Ampicillin sodium salt: 100 mg/mL stock solution in ddH2O,

filtered 0.22 μm. Make 10 mL, storage as 1-mL aliquots at
20 C for up to 6 months.
5. Dithiothreitol (DTT): 1 M stock solution in ddH2O. Make
1 mL, storage as 0.2-mL aliquots at 20 C for up to
6 months.
6. Pre-wash buffer for GST-tagged protein: 50 mM Tris–HCl,
pH 6.8, 300 mM NaCl, 1 mM DTT in ddH2O. Make 40 mL,
prepare just prior to use.
7. Complete EDTA-free protease inhibitor.
8. AEBSF protease inhibitor: 0.1 M stock solution in ddH2O.-
Make 1 mL, storage as 0.2-mL aliquots at 20 C for up to
6 months.
9. Homogenization buffer for GST-tagged protein: 50 mM Tris–
HCl, pH 6.8, 300 mM NaCl, 1 mM DTT, 0.1 mM AEBSF,
complete protease inhibitor in ddH2O. Make 20 mL, prepare
just prior to use.
11. Glutathione Sepharose 4B (GE Healthcare).
12. PreScission protease (GE Healthcare).
13. Isopropyl-β-thiogalactopyranoside (IPTG): 1 M stock solution
in ddH2O, filtered 0.22 μm. Make 20 mL, storage as 1-mL
aliquots at 20 C for up to 6 months.
14. Dialysis buffer without DTT: 50 mM Tris–HCl, pH 6.8,
150 mM NaCl in ddH2O. Make 1.5 L, prepare just prior
to use.
15. Slide-A-Lyzer Dialysis Cassettes, 3.5 K MWCO (Thermo
Fisher Scientific).
2.3 NBD Labeling of 1. IANBD-amide (N,N0 -dimethyl-N-(lodoacetyl)-N0 -(7-nitr-

PHPLCδV58C obenz-2-oxa- 1,3-diazol-4-yl)ethylenediamine) (Invitrogen).
Recombinant Protein 2. L-Cysteine hydrochloride: 100 mM solution in ddH2O. Make
1 mL, prepare just prior to use.
3. Dialysis buffer: 50 mM Tris–HCl, pH 6.8, 150 mM NaCl,
1 mM DTT in ddH2O. Make 1.5 L, prepare just prior to use.
2.4 Liposome 1. POPC (Avanti Polar Lipids): 25 mg/mL chloroform solution.

Preparation Store at 20 C for up to 6 months.
2. DOPS (Avanti Polar Lipids): 10 mg/mL chloroform solution.
Store at 20 C for up to 6 months.
3. Cholesterol (Avanti Polar Lipids): 10 mg/mL chloroform
solution. Store at 20 C for up to 6 months.
4. Brain PI4P (Avanti Polar Lipids): 1 mg/mL chloroform/
methanol/water solution. Store at 20 C for up to 6 months.
5. Brain PI(4,5)P2 (Avanti Polar Lipids): 1 mg/mL chloroform/

methanol solution. Store at 20 C for up to 6 months.
6. 0.5 M EDTA.
7. 0.5 M EGTA.
8. Buffer A: 50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1 mM
DTT, 0.2 mM EDTA, and 1 mM EGTA.
9. Nitrogen gas.
10. Liquid nitrogen.
11. Water bath.
12. Ultrasonic cleaning bath USC100T (VWR).
2.5 Evaluation of 1. Fluoromax spectrometer (HORIBA Scientific).

NBD-PHPLCδ Sensitivity 2. 200-μL quartz cell (Starnacells, 3–3.30-Q-3).
and Real-Time PIP5K
3. Adaptor (Starna Cells, FCA3.30).
Activity
4. GraphPad Prism 8 (GraphPad Software).
5. ATP solution, Tris buffered.
3 Methods
3.1 Preparation of 1. Transform BL21 (DE3) pLysS with zPIP5K plasmid, spread
zPIP5K Recombinant the bacteria on an LB agar plate containing 50 μg/mL of
Protein kanamycin, and incubate the plate at 37 C overnight.
2. Inoculate a single bacterial colony in 5 mL of 2 YT medium
containing kanamycin in a 15-mL tube and shake at 37 C
overnight.
3. Dilute the overnight culture 1:500 to a final volume of 0.5–2 L
of 2 YT medium containing kanamycin and shake the flask at
37 C until the cells reach OD600 of 0.6.
4. Cool down the cell culture and induce protein expression by
adding IPTG (final conc. 0.20 mM) and shake the flask at
22 C overnight.
5. Centrifuge cells at 3500 g for 10 min at 4 C and store at
80 C until use.
6. Resuspend cell pellets in 5 mL of pre-wash buffer.
7. Centrifuge cells at 3500 g for 10 min at 4 C and discard the
supernatant.
8. Resuspend cell pellets in 5 mL of pre-wash buffer again.
supernatant.
10. Resuspend cell pellets in 5 mL of homogenization buffer.
126 Taki Nishimura
11. Disrupt the cells by sonication for 5 min (Branson Sonifier

250, max amplitude, 10 s ON and 20 s OFF pulse) on ice
water (see Note 2).
12. Centrifuge the cell homogenates at 20,800 g for 30 min at
4 C to remove cell debris.
13. Collect the supernatant (~5 mL) and mix with 500 μL of
Ni-NTA agarose (50% slurry).
14. Rotate for 2 h at 4 C.
15. Centrifuge at 1700 g for 2 min at 4 C and discard the
supernatant. Resuspend the agarose in 500 μL of homogeniza-
tion buffer and rotate for 10 min at 4 C. Repeat this washing
step once more.
16. Wash the agarose by using 500 μL of homogenization buffer
containing 20 mM imidazole twice.
17. Wash the agarose by using 500 μL of homogenization buffer
containing 40 mM imidazole twice.
18. Resuspend the agarose in 400 μL of homogenization buffer
containing 80 mM imidazole and rotate for 10 min at 4 C.
19. Collect the supernatant. Repeat this elution step once more.
containing 160 mM imidazole and rotate for 10 min at 4 C.
21. Collect the supernatant. Repeat this elution step once more.
22. Analyze the eluate fractions by SDS-PAGE to check
the amount and purity of zPIP5K recombinant protein (see
Note 1).
23. Pool appropriate fractions and dialyze against 500 mL of dialy-
sis buffer using a dialysis cassette at 4 C.
24. Change dialysis buffer (500 mL) twice.
25. Change dialysis buffer to storage buffer and dialyze at 4 C
overnight.
26. Recover the dialyzed proteins and store the aliquots at 80 C
for up to 6 months. Once thawed, store at 20 C for up to
2 weeks.
3.2 Preparation of 1. Transform BL21 (DE3) pLysS with pGEX-6P-1-PHPLCδV58C

PHPLCδV58C plasmid, spread the bacteria on an LB agar plate containing
Recombinant Protein 100 μg/mL of ampicillin, and incubate the plate at 37 C
overnight.
2. Inoculate a single bacterial colony in 5 mL of 2 YT medium
containing ampicillin in a 15-mL tube and shake at 37 C
overnight.
3. Dilute the overnight culture 1:500 to a final volume of 0.5–2 L

of 2 YT medium containing ampicillin, and shake the flask at
37 C until the cells reach OD600 of 0.6.
4. Cool down the cell culture, induce protein expression by add-
ing IPTG (final conc. 0.20 mM), and shake the flask at 22 C
overnight.
5. Centrifuge cells at 3500 g for 10 min at 4 C and store at
80 C until use.
6. Resuspend cell pellets in 5 mL of pre-wash buffer.
supernatant.
8. Resuspend cell pellets in 5 mL of pre-wash buffer again.
supernatant.
10. Resuspend cell pellets in 5 mL of homogenization buffer.
11. Disrupt the cells by sonication for 5 min (Branson Sonifier
250, max amplitude, 10 s ON and 20 s OFF pulse) on ice
water (see Note 2).
12. Centrifuge the cell homogenates at 20,800 g for 30 min at
4 C to remove cell debris.
13. Collect the supernatant (~5 mL) and mix with 500 μL of
Glutathione Sepharose 4B (50% slurry) (see Note 3).
14. Rotate for 2 h at 4 C.
15. Centrifuge at 1700 g for 2 min at 4 C and discard the
supernatant. Resuspend the agarose in 500 μL of homogeniza-
tion buffer and rotate for 10 min at 4 C. Repeat this washing
step 4 times more.
16. Resuspend the sepharose in 400 μL of homogenization buffer
containing PreScission protease (final conc. 0.1 U/μL) (see
Note 4) and rotate overnight at 4 C.
17. Centrifuge at 1700 g for 2 min at 4 C and collect the
supernatant.
and rotate for 10 min at 4 C. Repeat this step two times more.
19. Analyze the eluate fractions by SDS-PAGE to check the
amount and purity of PHPLCδV58C recombinant protein.
20. Pool appropriate fractions and dialyze against 500 mL of dialy-
sis buffer without DTT using a dialysis cassette at 4 C.
21. Change dialysis buffer (500 mL) twice.
22. Recover the dialyzed proteins.
128 Taki Nishimura
3.3 NBD Labeling of 1. Mix 400 μL of the dialyzed PHPLCδV58C recombinant proteins
PHPLCδV58C with 100 μL of IANBD-amide (more than ten-fold excess in a
Recombinant Protein mol ratio) and incubate overnight at 4 C (protecting samples
from the light).
2. Add L-cysteine (final conc. 4 mM) to quench the reaction.
3. Dialyze against 500 mL of dialysis buffer to remove residual
IANBD-amide.
4. Change dialysis buffer (500 mL) twice and recover the dialyzed
NBD-PHPLCδ proteins.
5. Mix NBD-PHPLCδ proteins with an equal volume of glycerol
and store at 80 C until use.
3.4 Liposome 1. Mix the desired amount of lipids in a glass tube (see Note 5).
Preparation 2. Evaporate the organic solvent under a nitrogen gas stream.
3. Put the glass tube in a vacuum chamber for 2 h to completely
remove the organic solvent.
4. Add buffer A to reach a final concentration of 400 μM total
lipids and (optional) incubate for 30 min at room temperature
for efficient hydration.
5. Vortex thoroughly.
6. Incubate at 50 C for 5 min in a water bath (see Note 6) and
then vortex thoroughly. If the lipid film is still attached to the
wall of the tube, repeat this step.
7. Transfer to a 1.5-mL microcentrifuge tube.
8. Place the tube in liquid nitrogen for 1 min and then 3 min in a
50- C water bath (see Note 6). Vortex thoroughly.
9. Repeat this freeze-thaw cycles ten times.
10. Sonicate the suspension with a bath sonicator (10 s ON and
10 s OFF pulse, ten times) or with a tip sonicator (amplitude
50%, 10 s ON and 10 s OFF pulse, for 10 min) on ice water (see
Note 7). Keep the liposomes at 4 C and use within 1 day.
3.5 Evaluation of 1. Take 150 μL of 0 mol% PI4,5P2 liposomes into a 200-μL

NBD-PHPLCδ Sensitivity quartz cell.
2. Add 3 μL of NBD-PHPLCδ (final conc. 400 nM) and 30 μL of
10 mM MgCl2 (final conc. ~2 mM).
3. After 5 min, record NBD fluorescence (ex/em 468 nm/
530 nm, 5 nm bandwidths) using a Fluoromax spectrometer
(see Note 8).
4. Repeat these steps using liposomes containing different
amounts of PI4,5P2.
ΔEm530 (x103)
4
0
0.25 0.5 0.75 1.0
PI(4,5)P2 (mol%)
-2
Fig. 2 Changes in NBD-PHPLCδ fluorescence intensity in response to PI(4,5)P2

levels on liposomes. Liposomes (POPC/DOPS/cholesterol/PI(4,5)
P2 ¼ 70 X:10:20:X) were incubated with NBD-PHPLCδ. After 5 min, NBD
fluorescence was recorded by using fluorescence spectroscopy. Data represent
mean SEM (n ¼ 3). Note that the increase of NBD fluorescence intensity was
correlated with that of PI(4,5)P2 levels
5. Calculate an increase (ΔEm530) in signal of NBD fluorescence

by subtracting a background signal (0 mol% PI4,5P2 lipo-
somes; Fig. 2; see Note 9).
3.6 Real-Time 1. Take 150 μL of 1 mol% PI4P liposomes into a 200-μL

PIP5K Assay quartz cell.
2. Add zPIP5K (final conc. 60–150 nM) and 3 μL of
NBD-PHPLCδ (final conc. 400 nM).
3. Set the quartz cell in a Fluoromax spectrometer (see Note 8).
4. After 5 min, add 15 μL of 20 mM MgCl2 (final conc. ~2 mM).
5. At 2 min later, start to record NBD fluorescence (ex/em
468 nm/530 nm, 5 nm bandwidths) (time ¼ 0) and then
quickly add 15 μL of 800 μM ATP solution (final conc.
~80 μM) (see Note 10).
6. Record NBD fluorescence every second (from time ¼ 5 s to
time ¼ 300 s; Fig. 3a, red).
7. Repeat these steps using 0 mol% PI4P liposomes to measure a
background signal (Fig. 3a, blue).
8. Calculate an increase in signal of NBD fluorescence in the
presence of 1 mol% PI4P liposome (ΔEm530, raw data) from
that measured before ATP addition (time ¼ 0) (Fig. 3b, red).
130 Taki Nishimura
a b
1 mol% PI4P liposome 1 mol% PI4P liposome
16 0 mol% PI4P liposome 6 0 mol% PI4P liposome
14
ΔEm530 (x103)
12 4
Em530 (x103)
10
8 2
6
4 0
2
0 30 60 90 120 150 180
-2
sec
0 30 60 90 120 150 180
sec
c
6
1 mol% PI4P liposome
ΔEm530 (x103)
0
0 30 60 90 120 150 180
sec
Fig. 3 A result of real-time PIP5K assay. Liposomes were mixed with NBD-PHPLCδ and zPIP5K. Prior to
(time ¼ 0 s) and after addition of ATP, NBD fluorescence was recorded by using fluorescence spectroscopy.
(a) Raw data of NBD fluorescence intensity. (b) An increase in signal of NBD fluorescence by addition of ATP.
(c) An increase in signal of NBD fluorescence normalized by subtracting the baseline. Data represent
mean SEM (n ¼ 3)
9. Calculate an increase in signal of NBD fluorescence in the

presence of 0 mol% PI4P liposome (ΔEm530, background)
from that measured before ATP addition (time ¼ 0)
(Fig. 3b, blue).
10. Calculate the NBD signal increase dependently of PIP5K reac-
tion by using this equation (Fig. 3c):
ΔEm530 ¼ ΔEm530, raw data ΔEm530, background

11. Analyze data by using Spreadsheet software or GraphPad Prism
(Fig. 3; see Note 11).
4 Notes
1. The protein expression and solubility of the zPIP5K catalytic

domain are high enough for a large-scale production [9]. If
other PIP5Ks are used, cultivation temperature and the con-

centration of IPTG for protein induction need to be optimized
in order to maximize the amount of soluble protein.
2. Sonication time and power intensity need to be optimized
when using other sonicators.
3. GST-Accept (Nacalai Tesque) can be used as well.
4. As PreScission is a fusion protein of human rhinovirus
3C-protease and GST, it allows for cleavage and removal of
GST tags in a one-step process. Similar reagents are available
from other commercial sources.
5. As lipids are supplied as an organic solution that could leach
impurities out of plastics, it is recommended to use glass. The
information can be checked on this web page: https://
avantilipids.com/tech-support/storage-handling-of-lipids.
6. Set to a temperature above the phase transition temperature of
the chosen lipids. The information can be checked on this web
page: https://avantilipids.com/tech-support/physical-pro
perties/phase-transition-temps/.
7. Use a tip sonicator if smaller size liposomes (no more than
80 nm) are required. Sonicate the suspension for 10 min
(amplitude 50%, 10 s ON and 10 s OFF pulse) on ice water
and then centrifuge the suspension at 14,000 g for 10 min to
remove aggregates.
8. Alternatively, a fluorescence spectroscope from other manufac-
turers (Shimadzu, JASCO) is also available.
9. NBD-PHPLCδ can easily be damaged or destroyed. To avoid
this, aliquots should be frozen and thawed once, with any
remainder kept at 20 C for up to 1 week. Also, it will be
important to ensure that NBD-PHPLCδ can detect PI(4,5)P2
on liposomes using liposomes containing different amounts of
PI(4,5)P2.
10. Take 15 μL of ATP solution to a new 1.5-mL tube. It is taken
by using a P200 pipette set to 100 μL, then added to the quartz
cell, and gently mixed by pipetting up and down.
11. To confirm PIP5K activity, ADP-Glo kinase assay (Promega) is
useful as a complementary analysis.
Acknowledgments
I thank Christopher J. Stefan for giving me a great opportunity to

write this method article. This new technique I have shown in this
chapter has been established when I had worked as a postdoctoral
researcher in the Stefan lab at UCL. I thank members of the Tooze
lab at the Crick Institute for helpful discussions. In particular, Javier
132 Taki Nishimura
H. Hervás gave helpful suggestions regarding liposome prepara-

tion. Taki Nishimura was supported by JSPS Postdoctoral Fellow-
ships for Research Abroad and the Osamu Hayaishi Scholarship for
Study Abroad.
References
1. Balla T (2013) Phosphoinositides: tiny lipids step cycle driven by PI(4)P hydrolysis directs
with giant impact on cell regulation. Physiol sterol/PI(4)P exchange by the ER-Golgi tether
Rev 93(3):1019–1137. https://doi.org/10. OSBP. Cell 155(4):830–843. https://doi.org/
1152/physrev.00028.2012 10.1016/j.cell.2013.09.056
2. Hille B, Dickson EJ, Kruse M, Vivas O, Suh BC 7. Moser von Filseck J, Copic A, Delfosse V,
(2015) Phosphoinositides regulate ion channels. Vanni S, Jackson CL, Bourguet W, Drin G
Biochim Biophys Acta 1851(6):844–856. (2015) Intracellular transport. Phosphatidylser-
https://doi.org/10.1016/j.bbalip.2014.09. ine transport by ORP/Osh proteins is driven by
010 phosphatidylinositol 4-phosphate. Science 349
3. Hansen SB (2015) Lipid agonism: the PIP2 (6246):432–436. https://doi.org/10.1126/sci
paradigm of ligand-gated ion channels. Biochim ence.aab1346
Biophys Acta 1851(5):620–628. https://doi. 8. Nishimura T, Gecht M, Covino R, Hummer G,
org/10.1016/j.bbalip.2015.01.011 Surma MA, Klose C, Arai H, Kono N, Stefan CJ
4. Hammond GR, Balla T (2015) Polyphosphoi- (2019) Osh proteins control nanoscale lipid
nositide binding domains: key to inositol lipid organization necessary for PI(4,5)P2 synthesis.
biology. Biochim Biophys Acta 1851 Mol Cell 75(5):1043–1057.e8. https://doi.
(6):746–758. https://doi.org/10.1016/j. org/10.1016/j.molcel.2019.06.037
bbalip.2015.02.013 9. Hu J, Yuan Q, Kang X, Qin Y, Li L, Ha Y, Wu D
5. Kunz J, Wilson MP, Kisseleva M, Hurley JH, (2015) Resolution of structure of PIP5K1A
Majerus PW, Anderson RA (2000) The activa- reveals molecular mechanism for its regulation
tion loop of phosphatidylinositol phosphate by dimerization and dishevelled. Nat Commun
kinases determines signaling specificity. Mol 6:8205. https://doi.org/10.1038/
Cell 5(1):1–11 ncomms9205
6. Mesmin B, Bigay J, Moser von Filseck J, Lacas-
Gervais S, Drin G, Antonny B (2013) A four-
Chapter 9
Assessing In Situ Phosphoinositide–Protein Interactions

Through Fluorescence Proximity Ligation Assay in Cultured
Cells
Mo Chen, Hudson T. Horn, Tianmu Wen, Vincent L. Cryns,
and Richard A. Anderson
Abstract
Proximity ligation assay (PLA) is a well-established method for detecting in situ interactions between two
epitopes with high resolution and specificity. Notably, PLA is not only a robust method for studying
protein–protein interaction but also an efficient approach to characterize and validate protein posttransla-
tional modifications (PTM) using one antibody against the core protein and one against the PTM residue.
Therefore, it could be applied as a powerful approach to detect specific interactions of endogenous
phosphoinositides and their binding proteins within cells. Importantly, we have specifically detected the
PLA signal between PtdIns(4,5)P2 and its binding effector p53 in the nucleus. This cutting-edge method
fully complements other conventional approaches for studying phosphoinositide–protein interactions and
provides important localization signals and robust quantitation of the detected interactions. Here, we
present the PLA fluorescence protocol for detecting in situ phosphoinositide–protein interactions in
cultured cells and is semiquantitative for interactions that are regulated by cellular signaling.
Key words Proximity ligation assay, Posttranslational modification, Nuclear localization, Phosphoi-
nositide–protein interaction, PtdIns(4,5)P2, p53
1 Introduction
Proximity ligation assay (PLA) is a robust method for detecting and

quantifying interactions between two epitopes with high resolution
(<40 nm, traditionally considered as direct interaction) and speci-
ficity because interactions between endogenous proteins are
detected in their cellular context at physiological expression levels
[1, 2]. Since its development by Fredriksson et al. in 2002 [3], PLA
has been increasingly used to detect the interaction between two
proteins [4–8]. In addition to those studies, we have also applied
PLA for validating protein–protein interactions suggested by tradi-
tional methods, including pull-down assay followed by mass spec-
trometry, co-immunoprecipitation, in vitro protein binding assay,
133
134 Mo Chen et al.
enzyme-linked immunosorbent assay (ELISA), and protein–pro-

tein colocalization post-immunofluorescence staining [9–11].
Notably, PLA is not only a robust method for studying pro-
tein–protein interactions but also an efficient approach to charac-
terize and quantify protein posttranslational modifications (PTM)
using one antibody against the core protein and one against the
PTM residue. For example, the covalent modification of proteins
can be studied in situ owing to the dual recognition format
provided by PLA [12]. Therefore, it could be applied as a powerful
approach to detect specific interaction of endogenous phosphoino-
sitides and their binding proteins within cells. Importantly, we have
first introduced PLA into the field of phosphoinositide signaling by
specifically detecting the PLA signal between PtdIns(4,5)P2 and its
binding effector p53 in the nucleus, which was enhanced by the
genotoxic agent cisplatin and diminished by deletion of PIPKIα,
the kinase responsible for PtdIns(4,5)P2 generation [13–15]. This
cutting-edge method fully complements other conventional
approaches for studying phosphoinositide–protein interactions,
such as lipid strip assay and liposome sedimentation assay, and
provides semiquantitative subcellular localization of the detected
interactions.
Here, we present the PLA protocol, modified from the Duo-
link® Proximity Ligation Assay procedure (Millipore Sigma), the
only commercial resource currently available, for detecting the in
situ phosphoinositide–protein interactions in the nucleus (Fig. 1).
Briefly, cultured cells are fixed, permeabilized, and blocked as per
traditional immunofluorescence staining procedure. Next, two pri-
mary antibodies raised in different species are used to detect a
specific phosphoinositide and its potential binding effector. A pair
of PLA probes, oligonucleotide-labeled secondary antibodies raised
in corresponding species, then binds to the primary antibodies.
Only PLA probes located in close proximity (less than 40 nm) are
able to be joined by the hybridizing connector oligos and ligase to
form a closed circular DNA template, which is required for rolling-
circle amplification (RCA). The PLA probe then functions as the
primer for DNA polymerase to generate concatemeric sequences
during RCA. This reaction results in up to 1000-fold amplification
of the signal, thereby enabling in situ detection of phosphoinosi-
tide–protein interaction. Lastly, fluorophore-labeled oligos hybrid-
ize to the complementary repeating sequences in the amplicon.
These PLA signals are visualized as discrete spots by fluorescence
microscopy that can be quantified by NIH ImageJ analysis to
provide precise intracellular localization of the phosphoinositide–
protein interaction.
PLA of Phosphoinositide-Protein Complex 135
Fig. 1 Schematic illustration of protein–phosphoinositide PLA reaction. First, two primary antibodies recognize
the specific epitopes of the protein–phosphoinositide (PI) complex in the cell. Then, secondary antibodies
coupled with oligonucleotides (PLA probes) bind to the primary antibodies. Next, the connector oligos join the
PLA probes located in close proximity and become ligated. The resulting circular, closed DNA template
becomes amplified by the DNA polymerase. Complementary detection oligos conjugated with fluorochromes
hybridize to repeating sequences in the amplicons. Lastly, PLA signals are detected by fluorescent microscopy
as discrete punctate foci and provide the intracellular localization of the protein–PI complex. The example
image shows the PLA signals of p53-PtdIns(4,5)P2 complex (red) located at the nucleus (DAPI, blue) of
MDA-MB-231 cells
2 Materials
1. Microscope cover glass (22 22 mm) (see Note 1).

2. 6-well plate.
3. 70% ethanol.
4. Phosphate-buffered saline (PBS) (137 mM NaCl, 10 mM
phosphate, 2.7 mM KCl, pH 7.4).
5. Coating solution: dissolve 0.2% gelatin in ddH2O, incubate in
60–70 C water bath for 1 h to make sure complete dissolution,
filter with a 0.22-μm filter before use, and store at 4 C.
136 Mo Chen et al.
6. 4% paraformaldehyde (PFA) solution in PBS (4% PFA).

7. 0.3% Triton X-100 in PBS.
8. Humidity chamber: 15-cm cell culture dish with damp tissue
on the bottom, covered by PARAFILM to create a clean dry
surface and prewarmed to 37 C before use.
9. Blocking buffer: 1% bovine serum albumin (BSA) in PBS, store
at 4 C.
10. Buffer A: 0.01 M Tris, 0.15 M NaCl, 0.05% Tween 20, pH 7.4;
filter with a 0.22-μm filter before use and store at 4 C. Bring
the solution to room temperature before use.
11. Buffer B: 0.2 M Tris, 0.1 M NaCl, pH 7.5; filter with a 0.22-μ
m filter before use and store at 4 C. Bring the solution to
room temperature before use.
12. Buffer C: Mix 10 ml of buffer B and 990 ml of high purity
water and store at 4 C. Bring the solution to room tempera-
ture before use.
13. Primary antibodies against the phosphoinositides or proteins
of interest (see Note 2).
14. PLA PLUS probe: Duolink® in situ PLA Probe anti-Rabbit
PLUS (Millipore Sigma).
15. PLA MINUS probe: Duolink® in situ PLA Probe anti-Mouse
MINUS (Millipore Sigma).
16. Antibody diluent: provided in the above Duolink® in situ PLA
Probes (Millipore Sigma).
17. Duolink® in situ detection reagents Red kit (Millipore Sigma).
(a) 5 ligation buffer (see Note 3).
(b) Ligase.
(c) 5 amplification buffer (see Note 3).
(d) Polymerase.
18. 40 ,6-Diamidino-2-phenylindole (DAPI)-containing mounting
medium.
19. Glass microscope slides.
20. Nail polish.
21. Incubator at 37 C.
22. Freeze block for enzymes.
23. Shaker.
24. Water bath.
25. Fume hood.
26. Fluorescence microscope.
27. Analysis software (such as NIH ImageJ).
3 Methods
3.1 Cell Culture 1. Place a microscope cover glass into each well of a 6-well plate.
and Cover Glass 2. Add 2 ml of 70% ethanol to each well and incubate for 10 min.
Preparation
3. Remove the 70% ethanol and wash the wells with 3 ml of PBS
three times.
4. Add 2 ml of the coating solution to each well and incubate
overnight at 4 C.
5. Remove the coating solution, and seed and treat cells with
desired experimental conditions to reach 50% confluency
before fixation.
3.2 Prepare Cells 1. Remove the media from the cells and wash the cells once
for PLA with PBS.
2. In a fume hood, add 1 ml of 4% PFA in PBS to each well and
put on a rocker at low speed for 30 min at room temperature.
3. In a fume hood, remove the 4% PFA in PBS and wash the cells
three times with 2 ml of PBS.
4. Add 500 μl of 0.3% Triton X-100 in PBS to each well to
permeabilize cells for 10 min at room temperature with gentle
rocking (see Note 4).
5. Remove the 0.3% Triton X-100 in PBS and wash three times
with 2 ml of PBS.
6. Add 500 μl of blocking buffer to each well for 1 h at room
temperature with gentle rocking to block the cells.
7. Prepare primary antibody mixes by diluting the primary anti-
bodies, one from mouse and one from rabbit, in 40 μl of
blocking buffer per cover glass at a 1:50 to 1:300 dilution
depending on the antibody for each reaction.
8. Add the 40-μl primary antibody mixes on top of the PARAF-
ILM in the humidity chamber.
9. Take out the cover glasses from the wells and place them cell
side down on the antibody mixes in the humidity chamber so
that the cells are touching the mixture (see Notes 5 and 6).
10. Cover the humidity chamber and incubate at 4 C overnight
(see Note 7).
11. Place the cover glasses back into the 6-well plate with the cell
side facing up and wash three times with 2 ml of PBS.
3.3 Prepare PLA 1. Dilute the two PLA probes in 1:5 dilutions with the antibody
Probes diluent provided by the detection reagents kit to a total volume
of 40 μl per cover glass (8 μl PLUS probe and 8 μl MINUS
probe in 24 μl antibody diluent per cover glass).
138 Mo Chen et al.
2. Add the 40-μl mixes on top of the PARAFILM in the humidity

chamber.
3. Take out the cover glasses from the wells and place them cell
side down on the antibody mixes in the humidity chamber so
that the cells are touching the mixture (see Notes 5 and 6).
4. Cover the humidity chamber and transfer it into the 37 C
incubator and incubate for 1 h.
3.4 Ligation 1. Dilute the ligation buffer in high purity water in a 1:5 dilution
to a total volume of 40 μl per cover glass (8 μl ligation buffer in
32 μl of high purity water per cover glass).
2. Place the cover glasses back into the 6-well plate with the cell
sides facing up.
3. Wash the cover glasses for 5 min twice in buffer A at room
temperature with gentle rocking.
4. Clean the PARAFILM surface in the humidity chamber with
70% ethanol and high purity water and let dry.
5. Add ligase to the diluted ligation buffer in a 1:40 dilution to a
total volume of 40 μl per cover glass (1 μl ligase in 39 μl diluted
ligation buffer per cover glass) (see Note 8).
6. Add the 40 μl ligation solutions to the surface of the PARAF-
ILM in the humidity chamber and place the cover glasses on
top with the cell side facing down after tapping off excess wash
buffer (see Notes 5 and 6).
7. Incubate in a 37 C incubator for 30 min.
3.5 Amplification 1. Dilute the amplification buffer in high purity water in a 1:5
dilution to a total volume of 40 μl per cover glass (8 μl amplifi-
cation buffer in 32 μl high purity water per cover glass).
2. Place the cover glasses back into the 6-well plate with the cells
facing up and wash the glasses for 5 min twice in buffer A with
gentle rocking at room temperature.
3. Clean the PARAFILM surface in the humidity chamber with
70% ethanol and high purity water and let dry.
4. Add the polymerase to the diluted amplification buffer in a
1:80 dilution to a total volume of 40 μl per cover glass (0.5 μl
polymerase in 39.5 μl of diluted amplification buffer per cover
glass) (see Note 9).
5. Add 40 μl of amplification solutions to the surface of the
PARAFILM in the humidity chamber and add the cover glasses
to them cell side down after tapping off excess buffer A (see
Notes 5 and 6).
6. Incubate in a 37 C incubator for 100 min.
3.6 Final Washing 1. Place the cover glasses back into the 6-well plate with the cells
facing up and wash with buffer B for 10 min twice with gentle
rocking at room temperature (see Note 10).
2. Wash the cover glasses with buffer C for 1 min with gentle
rocking at room temperature (see Note 11).
3.7 Mounting 1. Take out the cover glasses and place them on a dry tissue with
cells facing up.
2. Loosely cover them with aluminum foil and let dry for 20 min
(see Note 12).
3. Add 30 μl of DAPI-containing mounting medium onto a glass
microscope slide.
4. Place the cover glasses on top of the mounting medium with
cells facing down (see Note 13).
5. Seal the edges of the cover glasses by applying nail polish to
them and dry the applied nail polish at room temperature with
the slides protected from light (see Note 14).
6. Store the microscope slides at 4 C prior to imaging.
3.8 Imaging 1. Acquire images of the slides using a fluorescence microscope.

Acquisition 2. Make sure you use appropriate filters for the PLA channel. The
and Quantification fluorophore in the Amplification Red has an excitation/emis-
sion wavelength of 594 nm/624 nm and can be detected using
the same filter as for Texas Red. Alternative detection channels
are also available, such as Amplification Green, Orange, and Far
Red reagents from Millipore Sigma, the only commercial
resource for PLA reagent currently available.
3. Export PLA and DAPI channels separately.
4. The PLA signal can be counted either manually (if the number
is not high) or by analysis software such as NIH ImageJ.
5. When using ImageJ for PLA signal quantification, load the
corresponding PLA and DAPI channels to ImageJ by using
File>Open> function.
6. Click the opened images, remove the background by using
Image>Adjust>Brightness/Contrast function. Drag the
adjusting bars for Maximum/Minimum/Brightness/Contrast
until the PLA signals are recognized as fluorescent puncta. An
individual signal is of sub-micrometer size.
7. To create the merged image, overlay the PLA and DAPI chan-
nels by using Image>Color>Merge Channels function.
Remember to check the box for keeping source images.
8. Use the DAPI channel as the reference to quantify PLA signals
per cell.
140 Mo Chen et al.
9. To quantify the PLA signal, select the opened PLA channel,

adjust the threshold using Image>Adjust>Color Threshold
function, choose the default threshold method, and click select.
Then, use the Analyze>Analyze Particles function to get the
PLA signal count.
10. To count the PLA signal in each cell, crop the image before
analyzing particles.
4 Notes
1. Commercially available cover glasses come in a variety of thick-

nesses, but the thickness closest to 0.17 mm is a No. 1.5
coverslip. For the newer, high- to super-resolution optical
microscopes, a No. 1.5H (high performance) is the required
thickness. Using the incorrect thickness for cover glasses can
greatly reduce the ability to maximize information from
prepared samples using an optical microscope.
2. Use verified phosphoinositide/protein antibodies suitable for
immunofluorescence staining in the fluorescence PLA proce-
dure. For example, the anti-PtdIns(4,5)P2 antibody (Echelon
Biosciences) used in our recent study to show the PtdIns(4,5)
P2-p53 interaction through PLA was verified for immunofluo-
rescence staining, which detected nuclear PtdIns(4,5)P2 in
nuclear speckles, nucleoplasmic foci, and nucleoli and was neu-
tralized by preincubation with PtdIns(4,5)P2 [13–15]. Also,
the two selected primary antibodies must be raised from differ-
ent species, such as one from mouse and one from rabbit, and
correspond to the subsequently used PLA probes.
3. Aliquot the 5 ligation buffer and 5 amplification buffer into
50-μl per tubes and keep those tubes at 20 C before use to
avoid multiple freeze-and-thaw cycles. Wrap all the 5 amplifi-
cation buffer tubes with aluminum foil to protect them from
light.
4. Permeabilization with a detergent such as Triton X-100 or
NP-40, which could penetrate the nuclear membrane, is
required for detecting the PLA signal of interacting phosphoi-
nositide–protein in the nucleus.
5. When taking out the cover glasses from the 6-well plate for
incubation in the humidity chamber, save the 6-well plate for
subsequent washings.
6. Make sure the reaction area corresponds to the reaction vol-
ume. The droplet must cover the reaction area.
7. If no or few signals in positive samples occur due to insufficient
binding of primary antibodies, try to enhance the incubation
condition of primary antibodies by incubating the cover glass
in a humidity chamber at room temperature overnight or in a

37 C incubator for 1 h. This all depends on the specific
antibodies used.
8. Keep the ligase in a freezer block (20 C), and only add it
immediately before the addition to the sample.
9. Keep the polymerase in a freezer block (20 C), and only add
it immediately before the reaction. The amplification buffer is
light-sensitive, so protect all amplification buffer solutions
from light.
10. Keep cover glasses protected from light while washing.
11. Additional staining steps for extra proteins/lipids of interest or
subcellular/subnuclear structures using fluorophore direct-
conjugated antibodies or dyes, such as fluorophore-conjugated
antibodies against SC-35 for nuclear speckles and phosphati-
dylserine (PS) for nuclear envelope [13], can be added from
this step by following conventional immunofluorescence stain-
ing procedure. Make sure they are conjugated with different
fluorophores from the one used for PLA detection.
12. Make sure the cover glasses are dry before mounting.
13. Use approximately 30 μl of mounting media for each cover
glass. The mounting media is viscous, so cut the pipette tip
before pipetting the mounting media. Apply the mounting
media to the slide, hold the cover glass with the side of the
cells facing the mounting media, tilt the cover glass, touch one
side to the mounting media, slowly place the cover glass down,
and allow the mounting media to spread throughout the entire
cover glass. Try to avoid bubbles. If there are bubbles, gently
push or rotate the cover glass to slide off the microscope slide
to release those bubbles. After bubbles are removed, push or
rotate the cover glass back. Then, gently remove excessive
mounting media at the edge of the cover glass using tissue
paper.
14. Apply half of the nail polish on the cover glass and half of the
nail polish on the slides. Leave no gap to join every side of the
cover glass and slide, and make sure the cover glass is
completely sealed by the nail polish onto the slides. This can
be repeated once when the first-time applied nail polish is dry.
Also, protect the slides from light while drying them at room
temperature.
References
1. Weibrecht I, Leuchowius KJ, Clausson CM Rev Proteomics 7(3):401–409. https://doi.
et al (2010) Proximity ligation assays: a recent org/10.1586/epr.10.10
addition to the proteomics toolbox. Expert 2. Bagchi S, Fredriksson R, Wallen-Mackenzie A
(2015) In situ proximity ligation assay (PLA).
142 Mo Chen et al.
Methods Mol Biol 1318:149–159. https:// 9. Chen M, Zhang Y, Yu VC et al (2014) Isthmin

doi.org/10.1007/978-1-4939-2742-5_15 targets cell-surface GRP78 and triggers apo-
3. Fredriksson S, Gullberg M, Jarvius J et al ptosis via induction of mitochondrial dysfunc-
(2002) Protein detection using proximity- tion. Cell Death Differ 21(5):797–810.
dependent DNA ligation assays. Nat Biotech- https://doi.org/10.1038/cdd.2014.3
nol 20(5):473–477. https://doi.org/10. 10. Chen M, Qiu T, Wu J et al (2018) Extracellular
1038/nbt0502-473 anti-angiogenic proteins augment an endoso-
4. Iwabuchi E, Miki Y, Ono K et al (2017) In situ mal protein trafficking pathway to reach mito-
detection of estrogen receptor dimers in breast chondria and execute apoptosis in HUVECs.
carcinoma cells in archival materials using prox- Cell Death Differ 25(11):1905–1920. https://
imity ligation assay (PLA). J Steroid Biochem doi.org/10.1038/s41418-018-0092-9
165:159–169. https://doi.org/10.1016/j. 11. Chen M, Choi S, Jung O et al (2019) The
jsbmb.2016.05.022 specificity of EGF-stimulated IQGAP1 scaffold
5. Thymiakou E, Episkopou V (2011) Detection towards the PI3K-Akt pathway is defined by
of signaling effector-complexes downstream of the IQ3 motif. Sci Rep 9(1):9126. https://
BMP4 using in situ PLA, a proximity ligation doi.org/10.1038/s41598-019-45671-5
assay. J Vis Exp 49:UNSP e2631. https://doi. 12. Sehat B, Tofigh A, Lin Y et al (2010) SUMOy-
org/10.3791/2631 lation mediates the nuclear translocation and
6. Karamouzis M, Dalagiorgou G, Georgopoulou signaling of the IGF-1 receptor. Sci Signal 3
U et al (2015) Proximity ligation assay (PLA) (108):ra10. https://doi.org/10.1126/
to identify HER2-negative breast carcinomas scisignal.2000628
responding in HER-3 targeting agents. J Clin 13. Choi S, Chen M, Cryns VL et al (2019) A
Oncol 33(15) nuclear phosphoinositide kinase complex reg-
7. Bedzhov I, Stemmler MP (2015) Applying the ulates p53. Nat Cell Biol 21(4):462–475.
proximity ligation assay (PLA) to mouse pre- https://doi.org/10.1038/s41556-019-0297-
implantation embryos for identifying protein- 2
protein interactions in situ. Methods Mol Biol 14. Kalasova I, Faberova V, Kalendova A et al
1233:57–64. https://doi.org/10.1007/978- (2016) Tools for visualization of phosphoino-
1-4939-1789-1_6 sitides in the cell nucleus. Histochem Cell Biol
8. Maszczak-Seneczko D, Sosicka P, Olczak T 145(4):485–496. https://doi.org/10.1007/
et al (2016) In situ proximity ligation assay s00418-016-1409-8
(PLA) analysis of protein complexes formed 15. Chen M, Wen T, Horn HT et al (2020) The
between Golgi-resident, glycosylation-related nuclear phosphoinositide response to stress.
transporters and transferases in adherent mam- Cell Cycle 19(3):268–289. https://doi.org/
malian cell cultures. Methods Mol Biol 10.1080/15384101.2019.1711316
1496:133–143. https://doi.org/10.1007/
978-1-4939-6463-5_11
Chapter 10
Characterization of Protein–Phospholipid/Membrane
Interactions Using a “Membrane-on-a-Chip” Microfluidic
System
Calvin Yeager, Djoshkun Shengjuler, Simou Sun, Paul S. Cremer,
and Craig E. Cameron
Abstract
It is now clear that organelles of a mammalian cell can be distinguished by phospholipid profiles, both as
ratios of common phospholipids and by the absence or presence of certain phospholipids. Organelle-
specific phospholipids can be used to provide a specific shape and fluidity to the membrane and/or used
to recruit and/or traffic proteins to the appropriate subcellular location and to restrict protein function to
this location. Studying the interactions of proteins with specific phospholipids using soluble derivatives in
isolation does not always provide useful information because the context in which the headgroups are
presented almost always matters. Our laboratory has shown this circumstance to be the case for a viral
protein binding to phosphoinositides in solution and in membranes. The system we have developed to
study protein–phospholipid interactions in the context of a membrane benefits from the creation of tailored
membranes in a channel of a microfluidic device, with a fluorescent lipid in the membrane serving as an
indirect reporter of protein binding. This system is amenable to the study of myriad interactions occurring
at a membrane surface as long as a net change in surface charge occurs in response to the binding event of
interest.
Key words Supported lipid bilayer, Pleckstrin Homology domain, pH modulation, Fluorescence,
Phosphoinositides, Phosphatidylinositol 4-phosphate, Microfluidics, Label-free
1 Introduction
Despite the large number of cellular processes that rely on interac-

tions with lipid membrane surfaces, there are few ways to model
these interactions in vitro. It is well known that molecular cofactors,
such as small molecules, divalent metal ions, and certain proteins
bearing membrane-binding domains such as pleckstrin homology
domains, can interact with lipid headgroups in biological mem-
branes [1–3]. While techniques such as isothermal calorimetry
(ITC), nuclear magnetic resonance (NMR) spectroscopy, and sur-
face plasmon resonance (SPR) are valuable tools to study these
143
144 Calvin Yeager et al.
Fig. 1 The membrane-on-a-chip device is a microfluidic platform. (a) The device is composed of three major
components: a PDMS block with an imprinted micropattern; a borosilicate glass coverslip; and a microscope
slide. Holes are punched in the PDMS block for inlet and outlet tubing. (b) The micropattern provides eight
microchannels for establishment of supported lipid bilayers (SLBs). Inlet and outlet holes should be punched at
the black circles. (c) The microfluidic device loaded onto an epifluorescence microscope. Outlet tubing feeds
into a petri dish; inlet tubing brings in running buffer from gravity flow. The microchannels are capable of being
fluorescently excited from an inverted microscope setup
interactions, they are not suitable for every situation. For example,
these techniques can require large quantities of protein, expensive
instrumentation, and special training that can be highly prohibitive
[4, 5]. Therefore, we have developed an on-chip microfluidic assay
dubbed “membrane-on-a-chip” capable of detecting analyte or
protein interactions with a lipid bilayer that is facile, inexpensive,
and highly sensitive and requires few specialized components
(Fig. 1).
By rupturing small unilamellar vesicles (SUVs) composed of
lipids of interest upon a borosilicate glass within a polydimethylsi-
loxane (PDMS) chip, the membrane-on-a-chip device can detect
membrane–analyte interactions by using a pH-sensitive fluoro-
phore, ortho-sulforhodamine B (Fig. 2). This fluorescent dye is
conjugated to a phosphatidylethanolamine lipid head group and
therefore sits in the plane of the bilayer and is solvent accessible
[6]. The probe is highly fluorescent at a low pH and quenched at a
high pH, with a pKa dependent on the other lipid constituents of the
bilayer but generally between 6.0 and 7.5 (Fig. 3) [6, 7]. Molecular
cofactors or proteins can interact with the supported lipid bilayer
(SLB). The charged analyte comes into close contact with the
membrane and will carry counterions from the bulk solution that
could impact the local pH of the SLB–solution interface [8]. This
local pH change will result in a change in the fluorescence signal of
the probe, which can be plotted against the concentration of the
analyte (Fig. 4). From this plot, an apparent dissociation constant
(Kd,app) can be extrapolated from the fitted data. This technique has
previously been validated studying SLB interactions with small
molecules, divalent metals, phospholipase Cδ1 Pleckstrin Homol-
ogy (PLC-δ1 PH) domain, and poliovirus 3C [6, 9, 10].
“Membrane-on-a-Chip” 145
Fig. 2 Small unilamellar vesicles (SUVs) can be formed from lipids of interest. (a) Our experiments contain
phosphatidylcholine (POPC), phosphatidylinositol-4-phosphate (PI4P), and ortho-sulforhodamine B conjugated
to phosphatidylethanolamine (oSRB-POPE). Depending on the user’s need, SUVs from other lipid mixtures can
be used. (b) Measurement of hydrodynamic radius of 92 mol% POPC, 7.5 mol% PI4P, and 0.5 mol% oSRB-
POPE vesicles. Dynamic light scattering (DLS) can be used to assure SUVs within their associated hydrody-
namic radius size range of 15–90 nm
Fig. 3 The oSRB-POPE probe is fluorescent and pH sensitive. (a) Primary data of the microfluidic channels. All
channels in the “Before” image are at pH 7.0. A range of different pH buffers are introduced to those channels
to highlight the dynamic range of 0.5 mol% oSRB-POPE in the “After” image. (b) Fluorescence quantitation of
panel a with a line tool. (c) Plotted absolute fluorescence versus pH values. The pKa of the oSRB-POPE probe in
92 mol% POPC, 7.5 mol% PI4P, and 0.5 mol% oSRB-POPE is 6.5
Fig. 4 The membrane-on-a-chip can detect protein–membrane interactions. (a) Protein is added to the
microchannels, and it interacts with lipids in the bilayer. The protein carries counterions, which will interact
with oSRB-POPE and impact the fluorescence values of the channel. (b) Increasing concentrations of a PV 3C
derivative bind to saturation on a 92 mol% POPC, 7.5 mol% PI4P, and 0.5 mol% oSRB-POPE membrane. (c)
Plotted values from panel b and Langmuir isotherm fit. The Kd,app was determined to be 3.0 0.2 μM
2 Materials
2.1 Preparation 1. 0.10-μm track-etched polycarbonate membranes (Whatman).

of Small Unilamellar 2. Extruder: The extruder required depends on the scale of SUVs
Vesicles (SUVs) to be produced. Model options presented are from both Avanti
(Mini-Extruder, Avanti), which can hold a maximum volume

of 1 mL, and Transferra Nanosciences (LIPEX Extruder,
Transferra T.001), which can hold 1.5–10 mL.
3. Glass scintillation vials: 20 mL capacity.
4. Lipids: We describe the use of 10 mg mL1 1-palmitoyl-2-
oleoyl-sn-glycero-3-phosphocholine (POPC) and
1 mg mL1 L-α-phosphatidylinositol-4-phosphate (PI4P).
Those and other lipids can be purchased from Avanti Lipids
or another supplier. When not used, lipids should be stored
under an inert gas such as argon or nitrogen.
5. Liquid nitrogen or ethanol–dry ice baths: To lower the tem-
perature of the lipid mixture below the transition temperature
for freeze–thaw cycles. Use 200 proof ethanol when making an
ethanol–dry ice bath.
6. pH meter with micro-pH probe.
7. pH-sensitive probe (oSRB-POPE): Typical stock concentra-
tions range 100 –300 μM in chloroform. oSRB-POPE is pro-
duced from lissamine rhodamine B sulfonyl chloride
(ThermoFisher Scientific). Conjugation to the phosphatidyl-
ethanolamine lipid and analytical separation of the ortho isomer
have been fully detailed previously [6].
8. Polypropylene tubes: 1.7 mL capacity, hinge-style caps.
9. Running buffer: 20 mM N-(2-hydroxyethyl)-piperazine-N-
0
-(2-ethanesulfonic acid) (HEPES), 100 mM sodium chloride
(NaCl) in deionized water. Adjust pH to 7.0 with 10 N sodium
hydroxide (NaOH). Filter sterilize with a 0.2-μm filter.
10. Ultrasonic bath: While we have used the Aquasonic 250D
(VWR), other ultrasonic baths can be used.
2.2 Fabrication of 1. Biopsy punch: A biopsy punch with a 1-mm diameter (Miltex).
Polydimethylsiloxane 2. Dry oven: Also referred to as gravity convection ovens.
(PDMS) Devices
3. PDMS: Elastomer kit of polydimethylsiloxane (Dow Corning).
4. Silicon mold preparation: Silicon wafers can be purchased from
University Wafer.
5. SU-8 50 photoresist (MicroChem Corp.).
6. SU-8 developer (MicroChem Corp.).
7. The design file containing the patterns of the microchannels is
provided in the paper published in Journal of Visualized Experi-
ments [10] and can be used for lithography, or provided to a
commercial lithography source [11–15].
2.3 Preparation 1. 7 cleaning solution: While we have purchased this detergent

of Glass Coverslips from MP Biomedicals, other detergents can be used.
2. Borosilicate glass dish: Recommended proportions are a diam-

eter height of 190 100 mm.
3. Deionized water.
4. Glass coverslips: Rectangular glass coverslips measuring
22 40 0.16–0.19 mm made of borosilicate glass.
5. Hot plate.
6. Kiln: The kiln used for treatment of glass coverslips can be
purchased from Paragon Industries (SC-2).
7. Nitrogen gas.
2.4 Device Assembly 1. Epifluorescence microscope: The user’s microscope setup may
and Preparation vary. To support this experimental technique, the microscope
must be outfitted with an Alexa 568 filter set, an image proces-
sing software capable of reading fluorescence intensity over
distance, and the ability to export those data into spreadsheet
software. To reduce user interaction, it would be beneficial to
have a microscope capable of automatically taking a series of
images over a fixed time (“time-lapse”).
2. Microscope slide: Glass. Recommended dimensions are
75 25 1 mm.
3. Nitrogen gas.
4. Plasma etching system (low-cost plasma cleaner), two-stage
direct drive oil vacuum pump, O2 service [Krytox charged]
(PlasmaEtch).
2.5 Preparation 1. Adjustable height platform: It is useful to adjust the height

of Supported Lipid difference between the running buffer and the microscope
Bilayers (SLBs) platform to change the flow rate by gravity. An easy way to do
this is with an adjustable height platform, which can be pur-
chased under the name “Lab Jack” from FisherScientific.
2. Hydrochloric acid (HCl): 0.2 N in deionized water.
3. Microcentrifuge tubes: 0.65 mL capacity.
4. Tubing: Tubing with an internal diameter of 0.020–0.026 in.
(0.051–0.066 cm) and a wall thickness of 0.010 in.
(0.025 cm).
2.6 Data Preparation 1. Data analysis: While we have had success using GraphPad Prism
and Analysis (v8.2.1) from GraphPad Software, other software can be used.
2. Dialysis buffer: The final buffer solution used for production of
your protein of interest but excluding the protein. For exam-
ple, our dialysis buffer for poliovirus 3C was composed of
20 mM HEPES, pH 7.0, 20% glycerol, 100 mM NaCl, and
1 mM β-mercaptoethanol.
3. Microscope software: While we have had success using AxioVi-

sion LE64 v.4.9.1.0 software from Carl Zeiss Microscopy,
other software can be used.
3 Methods
3.1 Preparation 1. Calculate volumes of phosphatidylcholine (POPC), phosphati-

of Rupturable Small dylinositol-4-phosphate (PI4P), and ortho-sulforhodamine B
Unilamellar Vesicles conjugated to phosphatidylethanolamine (oSRB-POPE) to be
(SUVs) mixed for appropriate mol percentage values at the desired final
volume (see Note 1).
2. Transfer these volumes into a 20-mL scintillation vial and dry
them using a stream of nitrogen gas. Desiccate the dried lipid
film in a desiccator under vacuum for 3 h at 500 Torr to
remove residual organic solvent.
3. Add running buffer to rehydrate the lipid film to the desired
final volume. At this point, the resulting solution will be a
mixture of multilamellar vesicles and will be turbid. Insert the
scintillation vial into an ultrasonic bath (sonicator) at a fre-
quency of 35 kHz for 30 min at room temperature. The
solution should be less turbid than before.
4. Freeze-thaw the vesicle suspension with an ethanol–dry ice
bath (or liquid nitrogen) and a water bath set to 40 C for a
total of ten times. If the solution does not lack turbidity by the
end of the freeze-thaw cycles, the vesicle suspension can be
centrifuged to assure removal of nonunilamellar vesicles
[16, 17].
5. Perform extrusion with the vesicle suspension ten times using a
0.10-μm track-etched polycarbonate membrane using your
choice of lipid extruder (see Note 2). The resultant SUVs
should not be turbid by eye [18]. Confirm the diameter of
SUVs using dynamic light scattering (DLS) (Fig. 2b) (see Note
3).

6. Store SUVs at 4 C in an aluminum foil covered
polypropylene tube.
3.2 Fabrication of 1. Mix the PDMS kit’s prepolymer and curing agent at a 10:1
Polydimethylsiloxane (w/w) ratio in a plastic cup. Mix vigorously for 5 min with a
(PDMS) Devices scoopula or glass stir bar.
2. Degas for 10 min at 500 Torr in a desiccator.
3. Pour the degassed solution into a prepared mold, including the
SU-8 micropattern, to a final height of 0.5 cm (see Note 4).
Degas the mold containing the PDMS solution a final time
until there are no air pockets remaining in or on the PDMS
surface. Cure the PDMS in an oven at 60 C for 30–60 min.
4. Remove the PDMS that covered the SU-8 micropattern with a

razor blade or scalpel. Cut out each device individually so that
the micropattern is preserved, or to a final size of
2 3 0.5 cm (Fig. 1a). Use a 1.0-mm hole diameter biopsy
punch to punch holes at the edges of the 8 inlets and 8 outlets
(Fig. 1b).
5. Wrap devices in scotch tape to protect them from dust and
store in a plastic container until use.
3.3 Preparation 1. Prepare a 1 solution from the 7 detergent using deionized

of Glass Coverslips water in a borosilicate glass dish. Heat the dish to 95 C on a
hot plate until the cloudy solution clears (see Note 5).
2. Arrange coverslips onto a ceramic staining rack, leaving a space
between each coverslip to prevent sticking from further manip-
ulations. Submerge the staining rack and coverslips in the 1
detergent solution for 1 h (see Note 6).
3. Remove the coverslips and staining rack and wash them gently
with 1 L of deionized water to remove remaining detergent.
Treat the coverslip with nitrogen gas until fully dry.
4. Place coverslips into a kiln for 6 h at 550 C (see Note 7).
5. Store coverslips in a plastic container (or their original con-
tainer) to prevent dust adhesion.
3.4 Device Assembly 1. Wash the surface, inlets, and outlets of the PDMS block with
deionized water. Dry the block with nitrogen gas. Use nitrogen
gas to blow away any dust particles from a single glass coverslip.
2. Insert a glass coverslip (prepared from Subheading 3.3) and
PDMS block into a plasma oxygen system chamber. Expose
with oxygen plasma for 45 s at a power setting of 75 W, an
oxygen flow speed of 10 cm3 min1, and a vacuum strength of
200 mTorr.
3. Place the PDMS block, patterned side down, onto the glass
coverslip to form the PDMS chip (see Note 8). Adjust so that
the block is in the center of the coverslip. Place the chip onto a
hot plate preheated to 100 C for 3 min to finalize the bond-
ing. Tape the device onto a glass microscope slide for simple
mounting into the microscope. Remove any further dust with
scotch tape.
4. Proceed immediately to Subheading 3.5 (see Note 9).
3.5 Preparation 1. Transfer 50 μL of prepared SUVs from Subheading 3.1 to a

of Supported Lipid 0.65-mL microcentrifuge tube. Acidify the vesicles to pH 3.2
Bilayers (SLBs) by adding 3.2 μL of 0.2 N HCl (see Note 10).
2. Pipette 5 μL of the acidified SUVs into each channel through
the inlet; the force required to push the SUVs through the
channel may vary by device. When detaching the pipette, leave

the tip anchored in the chip inlets. Incubate the device at room
temperature for 10 min.
3. Insert precut sets of outlet tubing, 8 cm long with an internal
diameter of 0.05 cm, into each outlet channel (see Note 11).
4. Take the chip, now containing tips from vesicle insertion and
outlet tubing, and tape its microscope slide base down to the
microscope stage.
5. Use an adjustable height platform to create a platform 20 cm
higher than the microscope stage. Fill a conical tube with
25–50 mL of running buffer and submerge one end of 8 precut
inlet tubing pieces (see Note 12). Tape the tubing to the
conical tube to make sure that it will not become dislodged.
6. Use a 1-mL syringe to draw running buffer from the conical
tube through the free end of the tubing. Remove the attached
pipette tip from the inlet of the chip, detach the syringe from
the inlet tubing, and use forceps to insert the inlet tubing into
the inlet (see Note 13). Repeat for all remaining inlet tubing.
7. Use the microscope’s software to set the objective (see Note
14). Use a “red” filter set (excitation/emission at 576/603 nm
or similar) to visualize the oSRB-POPE fluorophore in the lipid
bilayer. Bring the lipid bilayer into focus with the microscopes
focusing knobs. Minimize fluorescence exposure to the SLB
(see Note 15).
8. Use the microscope software to adjust exposure time for cap-
tured images to be “200 ms.” If possible, use the “time-lapse”
setting for your microscope software. Otherwise, manually take
images once every 5 min until fully equilibrated. Use the
microscope software to measure the fluorescence intensity
across a line. Draw the line measurement over an equivalent
section of all eight channels to establish the fluorescence inten-
sity (see Note 16). The fluorescence intensity should equili-
brate over time and be relatively similar per channel.
9. Save the image file(s).
3.6 Equilibrating 1. If your protein of interest was prepared using a buffer other
the On-Chip SLBs than the running buffer used in Subheading 3.5, it is necessary
with Dialysis Buffer to equilibrate the SLBs with that buffer (herein called dialysis
buffer) (see Note 17). Otherwise, move on to Subheading 3.7.
2. Pull the first inlet tubing out of the chip. Free the affixed end
previously taped to the conical tube in Subheading 3.5, step 5
and place it into a conical tube filled with dialysis buffer. Insert
a 1-mL syringe into the inlet tubing and exchange the running
buffer with dialysis buffer. Insert the inlet tubing into the chip
similarly to Subheading 3.5, step 4. Repeat for all remaining

inlet tubing.
3. Bring the channels into focus. Use a time-lapse or manual
capture to measure the fluorescence per channel until fully
equilibrated, similarly to Subheading 3.5, step 6.
3.7 Addition 1. Lower the height-adjustable platform so that there is no differ-

of Protein to On-Chip ence in height between the stage and the buffer flowing
SLBs through the inlet tube; this should completely stop the
gravity flow.
2. Prepare dilutions of the protein of interest. Herein, we demon-
strate using poliovirus 3C (PV 3C). We used the following PV
3C concentrations: 0, 1, 2, 5, 10, and 20 μM with a total
volume of 200 μL each (see Note 18).
3. Remove the first outlet tubing from the chip (see Note 19).
After removing the tubing, draw 200 μL of protein solution
into a pipette. Gently place the tip of the pipette into the outlet
and detach the pipette from the tip (see Note 20). Repeat for all
remaining outlets.
4. Lower the inlet tubing to approximately 10 cm lower than the
microscope stage, making sure to disconnect the taped end
from the buffer (see Note 21). Gravity will flow protein
through the pipette tips over the SLBs out of the device.
5. Adjust the channels into focus. Use a time-lapse or manual
capture to measure the fluorescence per channel until equili-
brated, similarly to Subheading 3.5, step 6.
3.8 Data Preparation 1. Extract the line-tool fluorescence intensity data from the final,
and Analysis equilibrated image after addition of the running buffer, dialysis
buffer (if necessary), and protein solutions (see Note 22). Use
your microscope software to export the fluorescence intensity
per pixel (or per unit length) into spreadsheet software such as
Microsoft Excel.
2. Average 10 or more data points from each channel before and
post addition of the protein of interest. Apply that data points
into the following formula:
F Protein
Relative Fluoresence ¼ ð1Þ
F0
where FProtein is the average fluorescence value over the
averaged data points after equilibration with a protein solution
and F0 is the average fluorescence value over the averaged data
points before addition of protein but after the addition of

running buffer (or dialysis buffer).
3. Repeat step 2 for each channel in your device. Plot the concen-
tration of protein added into the chip against the relative
fluorescence obtained (Fig. 4) and fit to a Langmuir isotherm:
ΔF max ½Protein
ΔF ¼ ð2Þ
K d,app þ ½Protein
where ΔF represents the change in fluorescence in any
sample relative to the maximum fluorescence change observed
(ΔFmax). Kd,app represents the apparent dissociation constant
equal to the protein concentration ([Protein]) at which
0.5ΔFmax is achieved.
4 Notes
1. These volumes will depend on the desired lipid composition,

concentration of lipids, and the extruder to be used. Common
lipid compositions for our use include (a) 92 mol% POPC,
7.5 mol% PI4P, and 0.5 mol% oSRB-POPE; or (b) 99.5 mol%
POPC and 0.5 mol% oSRB-POPE. To make 1 mL of 0.5 mg/
mL 92 mol% 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocho-
line (POPC), 7.5 mol% L-α-phosphatidylinositol-4-phosphate
(PI4P), and 0.5 mol% ortho-sulforhodamine B-1-palmitoyl-2-
oleoyl-sn-glycero-3-phosphoethanolamine (oSRB-POPE), use
0.448 mg of POPC, 0.048 mg of PI4P, and 3.2 103 ng of
oSRB-POPE.
2. The Avanti Mini-Extruder (Avanti 610,000) can hold a maxi-
mum volume of 1 mL while the LIPEX Extruder (Transferra
Nanosciences T.001) can hold 1.5–10 mL.
3. Typical size ranges for SUVs are between 10 and 100 nm in
diameter. SUVs are not stable indefinitely and will spontane-
ously fuse if dropped below the phase transition temperature,
which varies on the lipid composition [18, 19].
4. SU-8 photoresist protocols were performed in a nanofabrica-
tion facility to produce the micropattern. The silicon wafer
mold can be taped to the bottom of a 10-cm petri dish and
filled with PDMS to provide a flat surface for device creation.
Prevent PDMS from pooling under the silicon mold to assure
even curing.
5. Exercise caution with the hot detergent. Use tools to extract
and place the ceramic staining rack.
6. The detergent solution may evaporate over the hour-long sub-
mersion time. Add deionized water as needed to keep the
detergent at a 1 concentration. It may help to mark the liquid

level at the beginning of Subheading 3.3, step 1.
7. To assure the formation of SLBs, it is critical to use a kiln to
heat the glass coverslips to smooth the molecular surface of the
silicon dioxide.
8. Gently lower one side and then the second side of the PDMS
block onto the glass. This will provide more flexibility in repo-
sitioning the block before permanent attachment to the glass.
9. Plasma oxygen treatment leaves the borosilicate glass hydro-
philic, which enhances rupturing of the SUVs. This hydrophi-
licity lasts for only a few hours, so it is important to fabricate the
chip directly before using it for an experiment.
10. If the SUVs are not properly acidified, they will not rupture as
efficiently. This is one of the first steps to troubleshoot if SLBs
are not forming properly. Test the pH of the vesicle solution
with a micro-pH meter or remake the 0.2 N HCl solution.
11. All of the outlet tubing should be gently curved to the same
degree to assure the same flow rate from each channel.
12. The length of the inlet tubing will vary depending on the
microscope setup. Our use requires an inlet tubing length of
60 cm.
13. It is critical to not allow air into the microfluidic chip; if an air
bubble is trapped within a channel, the SLB is prone to peeling
of the channel, rendering it unusable. To prevent this, a good
technique is to allow for a drop of the current buffer to be
dripped onto the inlet and then to insert the tubing.
14. Each microscope is different, and the setup will vary. Our
Axiovert 200 M epifluorescence microscope allows us to see
the entire microfluidic chip at a 4 objective, with better
resolution using a 10 objective.
15. The fluorophore (oSRB-POPE) will photobleach upon exces-
sive fluorescence exposure. Limit exposure whenever possible.
16. Measurement techniques will vary with different microscopes
and software. The goal is to be able to measure the intensity of
the fluorescence over a distance measurement (pixels or
micrometers).
17. oSRB-POPE fluorescence is changed by local pH alterations;
therefore, the buffer composition of analyte solutions matters.
Adding charged species (e.g., Mg2+, Ca2+) will alter the maxi-
mum fluorescence of the channels, which could result in a false
positive for interactions of the analyte [9, 10, 20, 21]. The
concentration of nonanalyte molecules (like divalent cations,
proteins other than the protein of interest) should remain
constant between channel equilibration steps, as changes in
concentration will depress the overall fluorescence intensity

and mimic an analyte–SLB interaction.
18. 200 μL is a large enough volume to keep the channel from
being exposed to air for at least 30 min at the recommended
heights (see Note 21). Depending on the analyte of interest,
this period is usually long enough to reach equilibrium for
analyte binding to the SLB.
19. To prevent introducing an air bubble when inserting the
pipette tip, drip a drop of buffer from the outlet tubing onto
the outlet. Additionally, be aware that the chip will now be
running in the reverse direction of previous steps of the proto-
col. Protein will be entering through the outlet holes and
flowing out of the chip through the inlet holes and tubing.
We do not change the designations: despite buffer and protein
flowing out of the inlet tubing, we still call it the inlet tubing
for Subheading 3.7. We have previously published a paper in
Journal of Visual Experiments where a video of this process is
available [10].
20. Do not touch the plunger of the pipette. The protein will move
through the channel by gravity flow. Forcing the protein into
the channel risks collapsing the SLBs.
21. Adjust the height difference as necessary to find an appropriate
flow rate. Approximately 10 cm will let the protein flow
through the channel for around 30 min. It may be desirable
to tape the inlet tubing to the inside of a waste beaker so that
buffer/protein mixtures do not pool near the microscope.
22. When collecting data, be sure that the measurement lengths are
the same for different images (e.g., 1000 pixels for running
buffer, dialysis buffer, and protein solution images).
References
1. Lemmon MA (2008) Membrane recognition 5. Scott JL, Musselman CA, Adu-Gyamfi E,
by phospholipid-binding domains. Nat Rev Kutateladze TG, Stahelin RV (2012) Emerging
Mol Cell Biol 9(2):99–111. https://doi.org/ methodologies to investigate lipid-protein
10.1038/nrm2328 interactions. Integr Biol (Camb) 4
2. Kutateladze TG (2010) Translation of the (3):247–258. https://doi.org/10.1039/
phosphoinositide code by PI effectors. Nat c2ib00143h
Chem Biol 6(7):507–513. https://doi.org/ 6. Huang D, Zhao T, Xu W, Yang T, Cremer PS
10.1038/nchembio.390 (2013) Sensing small molecule interactions
3. Jung H, Robison AD, Cremer PS (2009) with lipid membranes by local pH modulation.
Detecting protein-ligand binding on sup- Anal Chem 85(21):10240–10248. https://
ported bilayers by local pH modulation. J Am doi.org/10.1021/ac401955t
Chem Soc 131(3):1006–1014. https://doi. 7. Sun S, Sendecki AM, Pullanchery S, Huang D,
org/10.1021/ja804542p Yang T, Cremer PS (2018) Multistep interac-
4. Narayan K, Lemmon MA (2006) Determining tions between ibuprofen and lipid membranes.
selectivity of phosphoinositide-binding Langmuir 34(36):10782–10792. https://doi.
domains. Methods 39(2):122–133. https:// org/10.1021/acs.langmuir.8b01878
doi.org/10.1016/j.ymeth.2006.05.006
8. Best QA, Xu R, McCarroll ME, Wang L, Dyer printed microfluidic devices: enablers and bar-
DJ (2010) Design and investigation of a series riers. Lab Chip 16(11):1993–2013. https://
of rhodamine-based fluorescent probes for doi.org/10.1039/c6lc00284f
optical measurements of pH. Org Lett 12 16. Axmann M, Schutz GJ, Huppa JB (2015) Sin-
(14):3219–3221. https://doi.org/10.1021/ gle molecule fluorescence microscopy on pla-
ol1011967 nar supported bilayers. J Vis Exp 105:e53158.
9. Shengjuler D, Chan YM, Sun S, Moustafa IM, https://doi.org/10.3791/53158
Li Z-L, Gohara DW, Buck M, Cremer PS, 17. Barenholz Y, Gibbes D, Litman BJ, Goll J,
Boehr DD, Cameron CE (2017) The Thompson TE, Carlson FD (1977) A simple
RNA-binding site of poliovirus 3C protein method for the preparation of homogeneous
doubles as a phosphoinositide-binding phospholipid vesicles. Biochemistry 16
domain. Structure 25(12):1875–1886.e1877. (12):2806–2810. https://doi.org/10.1021/
https://doi.org/10.1016/j.str.2017.11.001 bi00631a035
10. Shengjuler D, Sun S, Cremer PS, Cameron CE 18. Braunger JA, Kramer C, Morick D, Steinem C
(2017) PIP-on-a-chip: A label-free study of (2013) Solid supported membranes doped
protein-phosphoinositide interactions. J Vis with PIP2: influence of ionic strength and pH
Exp 125:55869. https://doi.org/10.3791/ on bilayer formation and membrane organiza-
55869 tion. Langmuir 29(46):14204–14213.
11. Lee J, Choi K-H, Yoo K (2014) Innovative https://doi.org/10.1021/la402646k
SU-8 lithography techniques and their applica- 19. van Paridon PA, de Kruijff B, Ouwerkerk R,
tions. Micromachines 6(1):1–18. https://doi. Wirtz KW (1986) Polyphosphoinositides
org/10.3390/mi6010001 undergo charge neutralization in the physio-
12. Poyton MF, Sendecki AM, Cong X, Cremer PS logical pH range: a 31P-NMR study. Biochim
(2016) Cu(2+) binds to phosphatidylethanol- Biophys Acta 877(1):216–219. https://doi.
amine and increases oxidation in lipid mem- org/10.1016/0005-2760(86)90137-2
branes. J Am Chem Soc 138(5):1584–1590. 20. Robison AD, Huang D, Jung H, Cremer PS
https://doi.org/10.1021/jacs.5b11561 (2013) Fluorescence modulation sensing of
13. Karasek P, Grym J, Roth M, Planeta J, Foret F positively and negatively charged proteins on
(2015) Etching of glass microchips with super- lipid bilayers. Biointerphases 8(1):1. https://
critical water. Lab Chip 15(1):311–318. doi.org/10.1186/1559-4106-8-1
https://doi.org/10.1039/c4lc00843j 21. Banerjee S, Aponte-Diaz D, Yeager C, Sharma
14. Thomas MS, Millare B, Clift JM, Bao D, SD, Ning G, Oh HS, Han Q, Umeda M,
Hong C, Vullev VI (2010) Print-and-peel fab- Hara Y, Wang RYL, Cameron CE (2018)
rication for microfluidics: what’s in it for bio- Hijacking of multiple phospholipid biosyn-
medical applications? Ann Biomed Eng 38 thetic pathways and induction of membrane
(1):21–32. https://doi.org/10.1007/ biogenesis by a picornaviral 3CD protein.
s10439-009-9831-x PLoS Pathog 14(5):e1007086. https://doi.
15. Waheed S, Cabot JM, Macdonald NP, Lewis T, org/10.1371/journal.ppat.1007086
Guijt RM, Paull B, Breadmore MC (2016) 3D
Chapter 11
Exploring Phosphoinositide Binding Using Native Mass

Spectrometry
Julian Bender and Carla Schmidt
Abstract
Phosphoinositides interact with proteins to fulfill various functions in the cell. In many cases, they specifi-
cally recruit peripheral membrane proteins to biological membranes. The analysis of their interactions with
proteins is therefore essential for understanding the underlying processes. Native mass spectrometry
(MS) preserves noncovalent interactions in the gas phase of a mass spectrometer and is therefore well-
suited to study protein–phosphoinositide interactions. In this protocol, we describe the application of
native MS to integral and peripheral membrane proteins and their interactions with lipids. We discuss
sample and instrumental requirements, the realization of experiments, and the data analysis workflow. We
further describe a biochemical assay to proof interactions of peripheral membrane proteins with lipids.
Key words Native mass spectrometry, Membrane proteins, Phosphoinositides, Protein–phosphoino-

sitide interactions
1 Introduction
1.1 Investigating Biological membranes of cells or organelles are diverse in their lipid
the Biological Function composition. As the lipid composition of the membranes affects
of Phosphoinositides their biophysical properties, including curvature, domain forma-
tion, or protein–membrane interactions, this lipid diversity results
in a variety of membrane characteristics. Variation in lipid composi-
tion is introduced by sorting processes of the ER and Golgi appa-
ratus, but also by chemical modification of the lipids in the
membrane. An example is phosphorylation of phosphatidylinositol
(PI) at positions 3, 4, or 5 of the inositol ring resulting in different
PI phosphate modifications (PIPs), including poly-phosphorylated
variants (PIP2 and PIP3 species) [1, 2].
Besides their role for signal transduction through formation of
second messengers by degradation into soluble inositol-1,4,5-tri-
sphosphate and membrane-bound diacylglycerol, PIPs mediate
protein–membrane interactions. Importantly, the various PIPs
located in different organelle membranes function as a molecular
157
158 Julian Bender and Carla Schmidt
tag for recruiting a specific subset of cytosolic proteins to the

respective membranes. These interactions are often formed
through specific binding modules, for instance, PH, C2, Phox,
Fyve, Dix, and ENTH domains, as well as short basic amino acid-
rich sequences [3, 4]. In addition, PIPs interact directly with
intrinsic membrane proteins in the membrane [5]. As a conse-
quence, PIPs are key players of transport processes such as exocy-
tosis or vesicle trafficking. The analysis of protein–PIP interactions
is therefore essential for understanding signaling processes in
the cell.
Several techniques have been developed for screening protein–
phosphoinositide interactions. A simple and straightforward
approach is the liposome floatation assay, which separates protein-
bound liposomes from free protein through density-gradient cen-
trifugation [6]. By varying the lipid composition of the liposomes
or introducing mutations in the protein sequence, this assay has
successfully been employed to discriminate between PIP-binding
events [7] as well as defining the PIP-binding pocket of
proteins [8].
In addition to targeted approaches, mass spectrometry (MS) is
a fast and reliable technique for the identification of proteins or
lipids from large-scale experiments [9, 10]. It is therefore well-
suited to study protein–PIP interactions and, importantly, to
uncover novel binding partners. As an example, the PI(3,5)P2
and PI(4,5)P2 interactomes were identified by MS following affin-
ity purification of specifically interacting proteins [11]. Similar to
proteins, the class and fatty acyl chain composition of a lipid can be
determined by identifying its exact mass together with characteris-
tic fragment ions obtained from collisional dissociation in the gas
phase of a mass spectrometer [12]. In this way, the profiles of
low-abundant phosphoinositides in the mammalian brain were
monitored after their extraction from brain tissue [13]. In addition
to simple MS-based identification, there are many structural MS
techniques to study the arrangements of proteins and lipids
[14]. One of these techniques is “native” MS, which will be dis-
cussed in detail in this protocol.
1.2 MS of Intact Native MS preserves noncovalent interactions in the gas phase of a

Proteins and Protein– mass spectrometer and therefore allows studying intact protein
ligand Complexes (ligand) complexes under near-physiological conditions [15–
17]. By determining the masses of the intact complexes, the sub-
unit stoichiometry, the formation of subcomplexes, or the oligo-
meric states of the proteins can directly be assessed (Fig. 1). Ligand
binding and, when combined with advanced techniques such as ion
mobility, conformational changes and complex stability are also
accessible through native MS [18, 19]. Consequently, many
biological questions can be addressed with native MS. In the
Exploring Phosphoinositide Binding Using Native Mass Spectrometry 159
Fig. 1 Applications of native MS for the analysis of protein–lipid complexes. Native MS can be used to probe
conformational changes (in combination with ion mobility), complex stoichiometries (by determining the exact
mass of complex), oligomeric state (by simultaneously monitoring all oligomeric species), lipid identity
(by determining the mass of attached lipids), and protein–lipid stoichiometry (by observing adduct peaks)
following sections, we will exemplify prerequisites and data inter-

pretation for native MS of membrane proteins and lipid binding.
1.2.1 Requirements Before analyzing a protein or protein complex by native MS, some
for Native MS prerequisites have to be fulfilled. These mainly include (1) a volatile
buffer system maintaining the native structure of the proteins and
avoiding unfolding during transfer into the gas phase, (2) a soft
ionization technique allowing ionization of the intact protein
assemblies and preventing dissociation and fragmentation of indi-
vidual subunits, and (3) instrument modifications facilitating
smooth transition of the protein complexes into the gas phase.
1. A suitable buffer system for native MS is volatile and mimics the
solution properties under physiological conditions (i.e., salt
strength). This is usually achieved by using aqueous
ammonium acetate solution or similar components at physio-

logical pH [20]. Organic solvents and strong acids, which are
commonly employed in conventional MS, should be avoided as
they cause unfolding of the proteins and dissociation of the
subunits [21].
2. During native MS, ionization is often achieved using nano-
electrospray ionization (nano-ESI). For this, the analyte solu-
tion is subjected to a high electric field leading to formation of
charged droplets that continuously shrink due to solvent evap-
oration on their way to the mass spectrometer. Using a capillary
with a small diameter delivers low flow-rates in the nanoliter
range and consequently causes formation of smaller droplets.
Employing nano-ESI instead of conventional ESI improves
spectral quality by reducing salt adducts and at the same time
sample consumption. Other ionization techniques for proteins
include laser-induced liquid bead ion desorption (LILBID)
[22] and matrix-assisted laser desorption ionization (MALDI)
[23, 24], even though the latter is rarely used for intact protein
complexes.
3. The requirements for native MS instrumentation are twofold:
First, the transmission of the high-mass ions of intact protein
complexes has to be optimized, and second, noncovalent inter-
actions have to be maintained. This is mainly achieved by
increasing the gas pressure at the different pumping stages of
the mass spectrometer, thereby establishing a pressure gradient
that gently transfers the ions into the high vacuum of the mass
spectrometer. Importantly, the high pressure in the source
region and the early vacuum stages induces collisional focusing
of the ions so that high-mass ions are stabilized on appropriate
trajectories in the instrument. Due to their theoretically unlim-
ited mass range, ToF mass analyzers are commonly used for
native MS experiments. Combined with quadrupole mass ana-
lyzers with reduced frequency, Q-ToF hybrid instruments
allow selecting and transmitting high-mass ions and are there-
fore traditionally employed for native MS [25]. More recently,
specialized quadrupole-orbitrap instruments allowing high-
resolution analyses were developed [26].
1.2.2 Spectrum Ion signals (“peaks”) detected in MS experiments have two main
Annotation/Analysis characteristics: the mass-to-charge (m/z) ratio and the signal inten-
sity. The number of charges acquired by an analyte molecule from
the solvent during ionization strongly depends on the size of the
molecule. In contrast to peptides, which are usually observed with
one or two charge states, analyte ions generated by nano-ESI in
native MS experiments often appear as a series of multiple charge
states with intensities following Gaussian distribution (see Fig. 2).
These series are also indicative for the folding state of proteins;
Fig. 2 Native mass spectrum of a peripheral membrane protein interacting with PIP2. Top spectrum: A charge
state series (13+ to 17+) of the protein is observed. The apo form (yellow) of the protein as well as binding of
one (light red) or two (dark red) PIP2 molecules is shown. Masses of the apo state as well as the lipid-bound
states are obtained by deconvolution of each Gaussian-shaped peak series (yellow, light, and dark red dashed
lines). The mass differences between the populations correspond to the mass of the attached lipid. In the low
m/z region, a DDM dimer (m/z 1021.6) and a Na+-adduct (m/z 1043.5) are observed. Negatively charged PIP2
is not observed in positive-ion mode. Bottom spectrum: Upon addition of phosphatidylcholine, only the apo
form of Atg18 is observed. Unbound phosphatidylcholine is observed in the low-m/z region (m/z 760.6).
Example protein used here: core autophagy protein Atg18; lipids: brain PI(4,5)P2 (top spectrum) and POPC
(bottom spectrum)
folded proteins show a narrow distribution of few charge states,

while unfolded proteins occur in a variety of states with different
surface areas resulting in broader charge state distributions.
One of the major advantages of MS over other techniques is the
possibility to study heterogeneous complex mixtures. In these
cases, charge state distributions of different complexes and sub-
complexes overlap in the mass spectra. Nonetheless, at sufficient
resolution, these charge state distributions can be assigned to the
complexes. Simulating theoretical charge state series confirms the
assignments. There are several software tools available for auto-
mated and semiautomated complex assignments and simulations
[27–29].
Similar to proteomics, tandem MS experiments can also be

performed in native MS. For this, a charge state of an intact protein
complex is selected in the quadrupole (see above), and collisional
energy is increased. During dissociation, a peripheral subunit
undergoes unfolding and is ejected as a highly charged monomer.
The remaining parent complex, the so-called “stripped” complex,
retains the residual charges. As the number of charges is reduced for
the stripped complex when compared with the same subcomplex
ionized from solution, it appears at higher m/z.
1.2.3 Determining Ligand As proteins maintain their noncovalent interactions during native
Binding by MS MS, it is also possible that they maintain interactions with their
ligands. This is usually observed by a shift to higher masses or by
mass adducts resulting in additional peaks flanking the individual
charge states toward higher m/z (Fig. 2). The mass difference
between the apo form of the complex (i.e., the assigned complex
without ligands) and the ligand-bound forms provides information
on the identity of the ligand. Nonetheless, unambiguous identifi-
cation of the ligand requires additional experiments.
Addition of potential ligands to a protein solution followed by
native MS is consequently an appropriate procedure to test and
verify ligand binding to proteins in vitro. Note that not only the
presence of bound ligands but also the protein–ligand binding
stoichiometries are obtained from these experiments. When
performing quantitative titration experiments of the ligands, native
MS further allows determining kinetic characteristics such as disso-
ciation constants [30, 31].
1.3 MS of Membrane Membrane proteins, in contrast to their soluble counterparts,

Proteins require special conditions to remain soluble, properly folded, and
active in aqueous solutions. Most commonly, this is achieved by
using detergents which provide a hydrophobic environment similar
to biological membranes. In general, there is no universal detergent
for membrane proteins, and, therefore, detergents have to be
screened for every individual membrane protein to ensure that
structure and activity are retained after solubilization. In some
cases, detergents interfere with native MS analysis; membrane pro-
teins are therefore usually purified in their preferred detergents and
are transferred into MS-compatible detergents prior to MS
analysis [32].
Similar to soluble proteins, native MS allows transferring mem-
brane proteins encapsulated in a detergent micelle into the gas
phase of the mass spectrometer. As masses and sizes of micelles
are very heterogeneous, protein–micelle complexes can rarely be
resolved and mostly result in broad signals. However, the detergent
micelle can be removed by collisional activation inside the mass
spectrometer, releasing the intact membrane protein complex [33].
In addition to detergent micelles, other membrane mimetics
providing a hydrophobic environment for membrane proteins in
aqueous solution have been developed. These include amphipols,

bicelles, nanodiscs, and styrene–maleic acid copolymers, which have
recently been introduced for native MS analysis and show great
promise for future studies [34–36].
Some lipids bind tightly to membrane proteins and are impor-
tant for their structure. These lipids often co-purify with the pro-
teins and, when using native MS, remain attached to the protein
after release from the detergent micelle or membrane mimic into
the gas phase [37–39]. As described for ligand binding (see above),
these lipids are often observed as adduct peaks, and the mass of
lipid-bound complex is shifted by the mass of the lipid (Fig. 2).
However, naturally occurring phospholipids are usually similar in
mass and therefore difficult to distinguish by native MS. Therefore,
lipids co-purified with membrane proteins are often identified after
extraction with organic solvents followed by lipidomics experi-
ments (Fig. 3) [40, 41].
Fig. 3 Schematic workflow for protein–lipid analysis by native MS. Lipids co-purified with proteins are
identified by combining native MS and lipidomics analysis following extraction of the lipids (lhs). Lipid
specificity can be assessed by incubating recombinantly expressed or delipidated proteins with a lipid mixture
of various lipid classes followed by native MS (rhs). Lipid affinity can be obtained from lipid titration
experiments using native MS. Lipid binding to peripheral membrane proteins can be confirmed by floatation
assays
Once the interacting lipids are identified, their importance for

membrane proteins can be assessed. As an example, the effect of
lipids onto the oligomeric state of membrane transporters was
examined in the presence or absence of different lipid species
[42, 43]. Furthermore, by varying the lipid concentration, binding
constants and thereby binding specificity can be determined
[44]. Similarly, lipid binding to peripheral membrane proteins
interacting with biological membranes can be assessed by native
MS. For this, intact phospholipids or lipid head groups are added in
solution or together with detergent micelles [45]. A simple proof
for interactions between peripheral membrane proteins and the
lipids is the liposome floatation assay (Fig. 3).
2 Materials
2.1 Sample 1. 1 M ammonium hydroxide solution.

Preparation 2. 7.5 M ammonium acetate solution.
3. Acetic acid (glacial, 99.8%).
4. 0.5-mL centrifugal filters containing different filtration mem-
branes (e.g., regenerated cellulose or polyethersulfone).
5. Small-scale size-exclusion chromatography spin columns
(0.8 mL bed volume).
6. Detergents (see [32] for a list of MS-compatible detergents).
7. Refrigerated centrifuge.
8. UV spectrophotometer.
2.2 Native MS 1. 0.5–20 μL long and flexible gel loading tips.

2. 100 mg/mL cesium iodide solution.
3. AA tweezers.
4. Argon (purity: N5.0).
5. Borosilicate glass capillaries, e.g., 1.0 mm outside diameter
(OD) 0.78 mm inside diameter (ID) or 1.2 mm
OD 0.69 mm ID.
6. Capillary puller with a heated filament (e.g., Sutter Instrument
Co., Model P-1000).
7. Ceramic cutter.
8. Conductive elastomer for nanospray probe.
9. Double-sided adhesive tape.
10. Petri dish (approx. 9 cm diameter).
11. Mass spectrometer modified for the transmission of high-mass
ions; e.g., Synapt G1 with high-mass modifications (Waters,
high-mass modifications, MS Vision). Mass spectrometers
from other vendors are available, but the protocol has to be

optimized for these systems.
12. Nitrogen (purity: >N4.8).
13. Optical stereomicroscope.
14. Software for analysis of mass spectra: Massign [28], UniDec
[27], and SUMMIT [46].
15. Sputter coater (e.g., Quorum Q150R S).
16. Vendor-specific software (e.g., Waters MassLynx 4.1).
17. Lipids of various classes (e.g., PIP, PIP2, PE, PG, PS)
(purity > 99%).
2.3 Floatation Assay 1. 0.75 M sucrose solution in protein buffer.

2. 2.5 M sucrose solution in protein buffer.
3. Gel electrophoresis materials (sample buffer, running buffer,
polyacrylamide gels, molecular weight marker, gel electropho-
resis apparatus, power supply, etc.).
4. Protein buffer (e.g., HEPES, Tris, or phosphate buffers, see
Note 1).
5. Ultracentrifuge equipped with a spin-out rotor and compatible
tubes for centrifugation up to 270,000 g and a sample
volume of 4 mL.
3 Methods
3.1 Sample Purify the peripheral or integral membrane protein of interest using
Preparation established protocols. Proteins from various sources can be ana-
lyzed by native MS (see Notes 2 and 3).
Prior to MS analysis, the protein of interest is transferred into
ammonium acetate solution. This is usually achieved by size-
exclusion chromatography using small-scale size-exclusion chroma-
tography columns. However, some proteins interact with the col-
umn material, or small-scale size-exclusion chromatography is not
sufficient to remove buffer salts. In these cases, the use of centrifu-
gal filters is recommended.
3.1.1 Small-Scale 1. Prepare 0.1–1 M ammonium acetate solution. For membrane

Size-Exclusion proteins, supply ammonium acetate solution with 2 the crit-
Chromatography ical micelle concentration (cmc) of detergent (see Notes 3–5).
2. Remove air from the gel bed of a size-exclusion spin column.
3. Place the column into a collection tube (supplied with the
columns). Remove storage buffer by centrifugation at
1,000 g and 4 C for 2 min. Discard the flow-through.
4. Add 500 μL of ammonium acetate solution to the column.

Centrifuge at 1,000 g and 4 C for 1 min. Discard the flow-
through.
5. Repeat step 4 three times to achieve complete buffer exchange.
6. Add 20–50 μL of your protein sample at a concentration of
10–20 μM to the column. Transfer the column to a fresh
collection tube and centrifuge for 4 min at 1,000 g and
4 C. Collect the flow-through containing the protein.
7. Determine the protein concentration of the flow-through by
UV spectroscopy. Adjust the concentration to approximately
10 μM using ammonium acetate solution prior to MS analysis.
3.1.2 Centrifugal Filters 1. Prepare 0.1–1 M ammonium acetate solution. For membrane
proteins, supply ammonium acetate solution with 2 cmc of
detergent (see Notes 3–5).
2. Add 200 μL of ammonium acetate buffer to the filter of a
filtration device with appropriate molecular weight cutoff (see
Note 7) to prevent adsorption of the protein to the dry filter
membrane.
3. Add the protein of interest and ammonium acetate solution to
a final volume of 500 μL.
4. Centrifuge at a maximum speed according to the manufac-
turer’s protocol until a volume of approximately 50 μL is
retained.
5. Add 450 μL of ammonium acetate solution to the filter and
repeat the centrifugation step at least four times (see Note 8).
Collect the remaining protein solution at a final volume of
approximately 50 μL.
6. Determine the protein concentration of the remaining protein
solution by UV spectroscopy. Adjust the concentration to less
than 10 μM using ammonium acetate solution prior to MS
analysis.
3.2 Native Mass 1. Insert borosilicate capillary into the capillary puller and start
Spectrometry the pulling process (see Note 9). The following parameters can
be used as a starting point for optimization of pulling ESI
3.2.1 Preparation of ESI
emitters:
Emitters
Line Heat Pull Velocity Delay Pressure Ramp

1x1 549 90 70 70 200 549
2. Collect and fix pulled emitters in a petri dish using double-

sided adhesive tape.
3. Insert the petri dish into a sputter coater and start the coating
process using the manufacturer’s settings (see Note 10).
4. Insert a coated emitter into the capillary holder of the instru-
ment and shorten the end using a ceramic cutter. Use a micro-
scope to open the tip using tweezers. Check the length and
shape of the tip and, if necessary, cut the tip with tweezers.
3.2.2 Native MS Quadrupole time-of-flight (Q-ToF) or quadrupole-orbitrap instru-

of Soluble Proteins ments are commonly used for native MS. Here, we describe the
procedure using a Waters Synapt G1 Q-ToF instrument in detail.
1. Prepare the instrument by increasing the detector voltage to a
value that is sufficient to gain maximum signal (see Note 11).
2. Switch collision gas on and increase the pressure in the source
region to 4–8 mbar. Usually nitrogen is employed as
collision gas.
3. Insert an emitter (see Subheading 3.2.1) into the capillary
holder containing a conducting elastomer and load 3–5 μL of
protein sample in ammonium acetate solution (see Subheading
3.1).
4. Mount the capillary holder and apply backing pressure to move
the liquid to the tip of the emitter. Reduce pressure once a
droplet emerges from the emitter tip.
5. Move the capillary toward the cone and apply a capillary volt-
age of 1.3 to 1.7 kV (the droplet should disappear). Start the
acquisition at the instrument and monitor acquired mass spec-
tra (see Notes 12–14).
6. Tune instrument parameters so that peaks are observed and
spectrum quality is maximized (see Note 15). Typical starting
parameters for a Synapt G1 Q-ToF instrument are 80 V sam-
pling cone voltage, 5 V extraction cone voltage, and 10 V
collision energy (trap and transfer cell).
7. Acquire several scans with optimized parameters and combine
multiple scans to improve the signal-to-noise ratio of the com-
bined mass spectrum. Smooth mass spectrum if required.
3.2.3 Native MS 1. Prepare the instrument following steps 1–5 of Subheading

of Membrane Proteins 3.2.2.
2. Tune the instrument parameters to observe peaks. For mem-
brane proteins, a higher collision energy is usually required
when compared with soluble proteins (Subheading 3.2.2,
step 1). Typical starting parameters for a Synapt G1 Q-ToF
instrument are 180 V sampling cone voltage, 190 V collision
energy, and 5 V extraction cone voltage.
3. Fine-tune instrument settings by reducing collision energy to
observe adduct peaks (see Notes 16 and 17).
4. Acquire several scans using optimized parameters to obtain

charge state series of proteins with and without adduct peaks.
5. For identification of attached lipids, select a protein–adduct
peak in the quadrupole and increase collision energy (tandem
MS). Monitor the low-m/z region (500–1300 m/z). The mass
of observed low-mass species is indicative of the attached lipids
retained during purification (see Note 15).
3.2.4 Determining Lipid To confirm lipid binding and to monitor lipid-induced effects,
Specificity lipids can be added to lipid-free or delipidated protein (see Note
18). This approach is applicable to peripheral and integral mem-
brane proteins.
1. Prepare aqueous solutions of specific lipids (e.g., PIP, PIP2,
PE, PG, PS, etc.) with defined concentration (see Note 19).
2. Mix the protein of interest in ammonium acetate solution with
defined molar ratios of each lipid individually.
3. Record mass spectra for each condition as described in Sub-
heading 3.2.3.
4. Vary the protein-to-lipid ratio and record mass spectra.
5. Combine various lipid classes and species at an equimolar ratio
and add them to the protein solution. Record mass spectra for
different mixing ratios.
3.2.5 Calibration of Mass 1. Use 3–5 μL of 100 mg/mL cesium iodide solution and record
Spectra mass spectra under the same conditions (pressure of source
region, temperature, etc.) as for protein analysis (Subheadings
3.2.2 and 3.2.3).
2. Optimize the spectrum to obtain cesium iodide clusters in the
same m/z-range recorded during protein analysis (e.g., by
decreasing the collision energy).
3. Record and combine multiple scans to obtain a mass spectrum.
4. Using the MassLynx instrument software, assign peaks of the
acquired spectrum to theoretical peaks of the calibrant. Use
theoretical mass spectra provided by MassLynx. Assign peaks
manually or using an automated peak assignment. Save the
calibration file.
5. Apply the calibration to every raw file acquired under the same
conditions.
3.3 MS Data Analysis Calculation of masses from native MS spectra is based on the fact
that neighboring charge states of a protein differ by the mass and
3.3.1 Determining
the charge of one proton. For manual assignment, follow these
the Mass of a Protein or
steps:
Protein Complex
Fig. 4 Schematic representation of a simplified mass spectrum. A charge state

series (z ¼ 1 to z ¼ 4) is shown. The m/z values (u1 to u4) result from the mass
(m ¼ 10) and the mass ( p) and charge (z) of the acquired protons see
Subheading 3.3.1 for details
1. Select an m/z from each peak. If the peak is symmetric, select

the center of the peak. If the peak has multiple maxima (for
instance, due to salt adducts), select the first maximum.
2. The m/z value of an ion (ui) with charge zi originates from the
mass of the analyte (m) and the mass ( p) and charge (e) of the
number (zi) of acquired protons during ESI.
m þ zip
ui ¼
zie
3. The neighboring peak ui + 1 of the same charge state distribu-
tion (at lower m/z, see Fig. 4) acquired one additional proton
and shows an m/z of.
m þ z iþ1 p m þ ðz i þ 1Þp
uiþ1 ¼ ¼
z iþ1 e ðz i þ 1Þe
4. As both peaks (steps 2 and 3) correspond to charge states of a
complex with the same mass m, the charge zi of the peak ui can
be calculated. The mass of a proton (in Da) divided by the
charge of a proton is approximately one, allowing to simplify
the term:
uiþ1 p=e p=e uiþ1 1 uiþ1

zi ¼ ¼ ffi
ui uiþ1 ui uiþ1 ui uiþ1
5. Calculate the mass m of the protein complex using the calcu-
lated charge of ui.
m ¼ ui z i
6. Repeat steps 1–4 for other peak pairs and calculate the average
mass of the protein complex and its standard deviation.
Alternatively, use available software tools for spectrum

deconvolution and annotation. Calculate masses of protein
complexes and protein–lipid complexes. Available software
tools are MaxEnt (incorporated in MassLynx) [47], Massign
[28], and UniDec [27].
3.3.2 Calculation 1. Assign masses to all peak series in the mass spectra recorded for
of Lipid-Binding Kinetics every protein-to-lipid ratio (Subheading 3.2.4).
2. For every mass spectrum, determine the sum of intensities (i.e.,
the sum of peak areas) of the peaks corresponding to the same
charge distribution.
3. For normalization, divide intensities of lipid-bound proteins by
intensity of the lipid-free protein. The lipid with the highest
normalized intensity is a good candidate for further interaction
studies.
4. For titration series of different protein-to-lipid ratios, plot the
normalized intensities for every protein–lipid complex. Calcu-
late association constants from the curves.
3.4 Confirmation 1. Prepare liposomes containing lipids that potentially interact

of Lipid Interactions with the protein of interest (detailed protocols for liposome
of Peripheral Proteins preparation can be found in this volume of Methods in Molec-
by Floatation Assay ular Biology). Use a suitable protein buffer (e.g., HEPES, Tris,
or phosphate buffers).
2. Mix liposomes with the protein of interest solubilized in the
same buffer and incubate for 1 h at room temperature or 4 C.
3. Adjust sucrose concentration of the sample to a final concen-
tration of 1 M using 2.5 M sucrose stock solution.
4. Transfer the solution into an ultracentrifugation tube.
5. Overlay with ¾ vol. of 0.75 M sucrose solubilized in buffer.
Pipette gently along the tube walls without mixing the sucrose
solutions.
6. Overlay with 1/10 vol. of buffer without sucrose.
7. Centrifuge for 2 h at 268,000 g (see Note 20).
8. Remove the top layer without sucrose and take an aliquot. Take
another aliquot from the bottom layer.
9. Perform gel electrophoresis of top and bottom fractions. Pro-
tein bound to liposomes is observed in the top fraction.
4 Notes
1. Generally, phosphate or HEPES buffer is preferred over Tris as

the latter readily forms adducts in the gas phase [21]. Moreover,
the effect of the buffer on the protein should be considered.
For example, purification of a PIP-interacting protein in phos-

phate buffer might hamper the binding to the lipid.
2. Phosphate or HEPES buffers are less prone to adduct forma-
tion than Tris buffers [21] and are therefore better suited for
native MS and should be used in the last purification steps, if
possible.
3. Native MS-compatible detergents have to be screened for every
individual protein. The use of nonionic detergents such as
glucosides, maltosides, and polyoxyethylene glycols is generally
advised for the purification of membrane proteins for native
MS [32]. A detergent concentration of at least 2 cmc should
be used to guarantee protein stability. Nonetheless, stability
and activity of the protein should be verified for each detergent.
In some cases, the available collision energy of the mass spec-
trometer is not sufficient to remove the detergent micelle by
collision-induced dissociation (CID). In these cases, the deter-
gent should be exchanged prior to native MS analysis.
4. The addition of detergent to ammonium acetate buffer
depends on the protein under investigation. Peripheral mem-
brane proteins usually do not require detergent, whereas mem-
brane proteins might require intact micelles. However, in some
cases, mass spectra with a better resolution are obtained when
buffer exchange is performed without additional detergent in
the ammonium acetate solution.
5. Ammonium acetate solution can be stored at 4 C for up to
14 days.
6. The CMC of a detergent can be determined experimentally, for
example, by measuring the electrical conductivity of a deter-
gent solution at increasing detergent concentration. The CMC
is the concentration at which the slope of the plot decreases as
new detergent molecules are mainly recruited into micelles. For
many MS-related detergents, the CMC is described in [32].
7. Choose a molecular weight cutoff smaller than the size of the
protein or protein complex under investigation. Note that
using a smaller molecular weight cutoff increases centrifuga-
tion time and might increase the detergent concentration (see
also Note 6).
8. Depending on the micelle size of the detergent (see Note 21),
filtration devices might increase the detergent concentration. If
the detergent is concentrated during filtration, detergent-
containing ammonium acetate solution might only be
employed for the first step and following steps might be per-
formed with ammonium acetate omitting the detergent.
9. The tip shape of the ESI emitter is crucial for native MS
analysis. It should be narrow at the front and not too flexible.
Depending on the preferred tip shape, they might require

opening using tweezers before ESI. The tip should have a
small inner diameter to allow sample flow in the nanoliter
range.
10. The coating settings depend on the instrument and metal
target used. For a Quorum Q150R S sputter coater, the fol-
lowing setting might be used: 40 mA sputter current, 300 s
sputter time, and tooling factor: (1) Commonly used metals
for coating ESI emitters are gold, platinum, and palladium.
Note that platinum and palladium allow applying higher capil-
lary voltages. Emitters can be prepared in-house as described
(Subheading 3.2.1) or commercially purchased.
11. For instruments containing multichannel plate detectors, the
lowest detector voltage at which the signal is maximally ampli-
fied is determined during maintenance and is increased with
the aging of the detector.
12. If no peaks are observed, increase capillary voltage or backing
pressure. Note that the best resolution is usually obtained with
low backing pressure.
13. The individual tip shape, the cutting of the tip, and the coating
of every single ESI emitter might influence the obtained mass
spectra. Try emitters of different shape and coating to obtain
the best results.
14. If mass spectra are insufficiently resolved, vary the distance
between cone and capillary, change the capillary voltage,
reduce the backing pressure, or increase collisional energy.
Note that increasing activation energy helps cleaning analyte
ions from unspecific adducts but also dissociates protein com-
plexes or ligands.
15. If tuning instrument parameters does not increase spectral
quality, use a different sample preparation protocol (e.g., buffer
exchange, Subheading 3.1). In some cases, buffer components
added during purification might hamper MS analysis. Opti-
mize purification conditions if mass spectra show noise signals
from buffer components. The purity of a protein complex
might also affect the analysis.
16. Collision energy should be raised to remove salt adducts, how-
ever, should be kept low enough to retain attached lipids. An
optimization protocol for membrane protein analysis is
provided in [32].
17. Negatively charged lipids might not be observed in positive-
ion mode, which is usually employed for native MS of proteins.
To confirm lipid identity, extract the lipids using organic sol-
vents and identify the lipid species by lipidomics [40, 41] or
chapters included in this Methods in Molecular Biology
volume.
18. Protein delipidation can be performed using harsh detergents

such as Triton X-100 [32, 42, 48].
19. The concentration of phospholipids in aqueous solutions can
be quantified by reducing the phospholipids to ashes, resus-
pension in a defined volume, and determining the concentra-
tion of inorganic phosphate, e.g., by addition of ammonium
heptamolybdate.
20. If the protein is stable under these conditions, centrifugation
can also be carried out overnight.
21. Detergent micelle size can be determined experimentally, for
example, by dynamic light scattering. However, many vendors
provide information on the micelle sizes of their detergents.
Acknowledgments
We thank Dr. Karin Kühnel for providing Atg18 and Melissa Frick
for critically reading this protocol. We acknowledge funding from
the Federal Ministry for Education and Research (BMBF, ZIK
program, 03Z22HN22), the European Regional Development
Funds (EFRE, ZS/2016/04/78115), and the MLU Halle-
Wittenberg. CS and JB further acknowledge funding from the
German Research Foundation (DFG, RTG 2467 “Intrinsically Dis-
ordered Proteins – Molecular Principles, Cellular Functions, and
Diseases”, project number 391498659). JB acknowledges funding
from the Studienstiftung des deutschen Volkes.
References
1. McCartney AJ, Zhang Y, Weisman LS (2014) 7. Busse RA, Scacioc A, Hernandez JM et al
Phosphatidylinositol 3,5-bisphosphate: low (2013) Qualitative and quantitative characteri-
abundance, high significance. BioEssays 36 zation of protein-phosphoinositide interac-
(1):52–64 tions with liposome-based methods.
2. York JD (2006) Regulation of nuclear pro- Autophagy 9(5):770–777
cesses by inositol polyphosphates. Biochim 8. Krick R, Busse RA, Scacioc A et al (2012)
Biophys Acta 1761(5-6):552–559 Structural and functional characterization of
3. Hurley JH, Meyer T (2001) Subcellular target- the two phosphoinositide binding sites of
ing by membrane lipids. Curr Opin Cell Biol PROPPINs, a beta-propeller protein family.
13(2):146–152 Proc Natl Acad Sci U S A 109(30):
4. Lemmon MA (2003) Phosphoinositide recog- E2042–E2049
nition domains. Traffic 4(4):201–213 9. Aebersold R, Mann M (2016) Mass-
5. Hamilton PJ, Belovich AN, Khelashvili G et al spectrometric exploration of proteome struc-
(2014) PIP2 regulates psychostimulant beha- ture and function. Nature 537(7620):347–355
viors through its interaction with a membrane 10. Shevchenko A, Simons K (2010) Lipidomics:
protein. Nat Chem Biol 10(7):582–589 coming to grips with lipid diversity. Nat Rev
6. Busse RA, Scacioc A, Schalk AM et al (2016) Mol Cell Biol 11(8):593–598
Analyzing protein-phosphoinositide interac- 11. Catimel B, Schieber C, Condron M et al
tions with liposome flotation assays. Methods (2008) The PI(3,5)P2 and PI(4,5)P2 interac-
Mol Biol 1376:155–162 tomes. J Proteome Res 7(12):5295–5313
12. Pulfer M, Murphy RC (2003) Electrospray macromolecular assemblies. Anal Chem 74

mass spectrometry of phospholipids. Mass (6):1402–1407
Spectrom Rev 22(5):332–364 26. Dyachenko A, Wang G, Belov M et al (2015)
13. Wenk MR, Lucast L, Paolo GD et al (2003) Tandem native mass-spectrometry on
Phosphoinositide profiling in complex lipid antibody-drug conjugates and submillion Da
mixtures using electrospray ionization mass antibody-antigen protein assemblies on an
spectrometry. Nat Biotechnol 21(7):813–817 Orbitrap EMR equipped with a high-mass
14. Bender J, Schmidt C (2019) Mass spectrome- quadrupole mass selector. Anal Chem 87
try of membrane protein complexes. Biol (12):6095–6102
Chem 400(7):813–829 27. Marty MT, Baldwin AJ, Marklund EG et al
15. Allison TM, Bechara C (2019) Structural mass (2015) Bayesian deconvolution of mass and
spectrometry comes of age: new insight into ion mobility spectra: from binary interactions
protein structure, function and interactions. to polydisperse ensembles. Anal Chem 87
Biochem Soc Trans 47(1):317–327 (8):4370–4376
16. Leney AC, Heck AJR (2017) Native mass spec- 28. Morgner N, Robinson CV (2012) Massign: an
trometry: what is in the name? J Am Soc Mass assignment strategy for maximizing informa-
Spectrom 28(1):5–13 tion from the mass spectra of heterogeneous
17. Mehmood S, Allison TM, Robinson CV protein assemblies. Anal Chem 84
(2015) Mass spectrometry of protein com- (6):2939–2948
plexes: from origins to applications. Annu Rev 29. Tseng YH, Uetrecht C, Yang SC et al (2013)
Phys Chem 66:453–474 Game-theory-based search engine to automate
18. Schiffrin B, Calabrese AN, Devine PWA et al the mass assignment in complex native electro-
(2016) Skp is a multivalent chaperone of outer- spray mass spectra. Anal Chem 85
membrane proteins. Nat Struct Mol Biol 23 (23):11275–11283
(9):786–793 30. Hopper JT, Robinson CV (2014) Mass spec-
19. Laganowsky A, Reading E, Allison TM et al trometry quantifies protein interactions--from
(2014) Membrane proteins bind lipids selec- molecular chaperones to membrane porins.
tively to modulate their structure and function. Angew Chem Int Ed Engl 53
Nature 510(7503):172–175 (51):14002–14015
20. Konermann L (2017) Addressing a common 31. Jecklin MC, Schauer S, Dumelin CE et al
misconception: ammonium acetate as neutral (2009) Label-free determination of protein-
pH "buffer" for native electrospray mass spec- ligand binding constants using mass spectrom-
trometry. J Am Soc Mass Spectrom 28 etry and validation using surface plasmon reso-
(9):1827–1835 nance and isothermal titration calorimetry. J
Mol Recognit 22(4):319–329
21. Hernández H, Robinson CV (2007) Deter-
mining the stoichiometry and interactions of 32. Laganowsky A, Reading E, Hopper JTS et al
macromolecular assemblies from mass spec- (2013) Mass spectrometry of intact membrane
trometry. Nat Protoc 2(3):715–726 protein complexes. Nat Protoc 8(4):639–651
22. Morgner N, Barth HD, Brutschy B (2005) A 33. Barrera NP, Di Bartolo N, Booth PJ et al
new way to detect noncovalently bonded com- (2008) Micelles protect membrane complexes
plexes of biomolecules from liquid micro- from solution to vacuum. Science 321
droplets by laser mass spectrometry. Aust J (5886):243–246
Chem 59:109–114 34. Hellwig N, Peetz O, Ahdash Z et al (2018)
23. Kohler M, Neff C, Perez C et al (2018) Bind- Native mass spectrometry goes more native:
ing specificities of nanobody·membrane pro- investigation of membrane protein complexes
tein complexes obtained from chemical cross- directly from SMALPs. Chem Commun 54
linking and high-mass MALDI mass spectrom- (97):13702–13705
etry. Anal Chem 90(8):5306–5313 35. Hopper JTS, Yu YT-C, Li D et al (2013)
24. Lehmann E, Zenobi R (1998) Detection of Detergent-free mass spectrometry of mem-
specific noncovalent zinc finger peptide- brane protein complexes. Nat Methods 10
oligodeoxynucleotide complexes by matrix- (12):1206–1208
assisted laser desorption/ionization mass spec- 36. Watkinson TG, Calabrese AN, Giusti F et al
trometry. Angew Chem Int Ed Engl 37 (2015) Systematic analysis of the use of amphi-
(24):3430–3432 pathic polymers for studies of outer membrane
25. Sobott F, Hernández H, McCammon MG et al proteins using mass spectrometry. Int J Mass
(2002) A tandem mass spectrometer for Spectrom 391:54–61
improved transmission and analysis of large
37. Zhou M, Morgner N, Barrera NP et al (2011) 43. Pyle E, Kalli AC, Amillis S et al (2018) Struc-
Mass spectrometry of intact V-type ATPases tural lipids enable the formation of functional
reveals bound lipids and the effects of nucleo- oligomers of the eukaryotic purine symporter
tide binding. Science 334(6054):380–385 UapA. Cell Chem Biol 25(7):840–848. e4
38. Cong X, Liu Y, Liu W et al (2016) Determining 44. Marcoux J, Wang SC, Politis A et al (2013)
membrane protein-lipid binding thermody- Mass spectrometry reveals synergistic effects
namics using native mass spectrometry. J Am of nucleotides, lipids, and drugs binding to a
Chem Soc 138(13):4346–4349 multidrug resistance efflux pump. Proc Natl
39. Liko I, Degiacomi MT, Mohammed S et al Acad Sci U S A 110(24):9704–9709
(2016) Dimer interface of bovine cytochrome 45. Landreh M, Costeira-Paulo J, Gault J et al
c oxidase is influenced by local posttranslational (2017) Effects of detergent micelles on lipid
modifications and lipid binding. Proc Natl binding to proteins in electrospray ionization
Acad Sci 108(53):201600354 mass spectrometry. Anal Chem 89
40. Frick M, Schmidt C (2019) Mass (14):7425–7430
spectrometry-A versatile tool for characterising 46. Taverner T, Hernandez H, Sharon M et al
the lipid environment of membrane protein (2008) Subunit architecture of intact protein
assemblies. Chem Phys Lipids 221:145–157 complexes from mass spectrometry and homol-
41. Gupta K, Li J, Liko I et al (2018) Identifying ogy modeling. Acc Chem Res 41(5):617–627
key membrane protein lipid interactions using 47. Ferrige AG, Seddon MJ, Jarvis S (1991) Maxi-
mass spectrometry. Nat Protoc 13 mum entropy deconvolution in electrospray
(5):1106–1120 mass spectrometry. Rapid Commun Mass
42. Pyle E, Guo C, Hofmann T et al (2019) Spectrom 5
Protein-lipid interactions stabilise the oligo- 48. Bechara C, Noll A, Morgner N et al (2015) A
meric state of Bor1p from Saccharomyces cere- subset of annular lipids is linked to the flippase
visiae. Anal Chem 91(20):13071–13079 activity of an ABC transporter. Nat Chem 7
(3):255–262
Chapter 12
Liposome-Based Methods to Study

Protein–Phosphoinositide Interaction
Mélanie Mansat, Mélanie Picot, Gaëtan Chicanne, Virginie Nahoum,
Frédérique Gaits-Iacovoni, Bernard Payrastre, Karim Hnia,
and Julien Viaud
Abstract
Following their generation by lipid kinases and phosphatases, phosphoinositides regulate important
biological processes such as cytoskeleton rearrangement, membrane remodeling/trafficking, and gene
expression through the interaction of their phosphorylated inositol head group with a variety of protein
domains such as PH, PX, and FYVE. Therefore, it is important to determine the specificity of phosphoi-
nositides toward effector proteins to understand their impact on cellular physiology. Several methods have
been developed to identify and characterize phosphoinositide effectors, and liposomes-based methods are
preferred because the phosphoinositides are incorporated in a membrane, the composition of which can
mimic cellular membranes. In this report, we describe the experimental setup for liposome flotation assay
and a recently developed method called protein–lipid interaction by fluorescence (PLIF) for the characteri-
zation of phosphoinositide-binding specificities of proteins.
Key words Phosphoinositide, Protein–lipid interaction, Liposome, Liposome flotation, Protein–lipid

interaction by fluorescence
1 Introduction
Phosphoinositide (PIP) [1]-binding specificity can be analyzed by

different techniques such as lipid–protein overlay assay, isothermal
titration calorimetry (ITC), surface plasmon resonance (SPR), lipo-
some sedimentation or flotation, and more recently by protein–
lipid interaction by fluorescence (PLIF) [2–5]. The liposome-based
methods are considered more robust because PIPs are presented in
a membrane with a defined composition that can mimic cellular
compartments and are less prone to detect false-positive interac-
tions. Some of them are quantitative (SPR, ITC) but require either
expensive materials and/or high amounts of lipids and proteins and
are used as a second approach to determine affinities and thermo-
dynamic parameters. As a first approach, liposome sedimentation or
177
178 Mélanie Mansat et al.
flotation is often used (Fig. 1), and liposome flotation is preferred

to avoid false-positive or overestimated interaction, considering
that some proteins have tendency to precipitate [4].
Briefly, in the flotation assay, the protein of interest is incubated
with liposomes that contain a phospholipid matrix (POPC, POPE,
and fluorescent PE that is useful to detect the liposomes) and each
different PIP. Following an incubation period, a sucrose gradient is
applied that will separate liposome-bound proteins (top
low-density fraction) from unbound proteins after ultracentrifuga-
tion. Then, fractions are collected and analyzed by SDS-PAGE
followed by western blot analysis or Coomassie blue staining
(Fig. 1). The advantage of this assay is the use of liposomes of
various sizes (obtained by sequential extrusion) that allows study-
ing the effect of membrane curvature on protein binding. How-
ever, this method is time-consuming and allows the simultaneous
processing of only few experimental samples.
More recently, we developed the PLIF assay that lies on the use
of 96-well plates to screen protein–PIP specificity in less than 1 h
with access to relative affinities [5]. PLIF is based on the use of
glutathione-coated plates that will pull down glutathione
S-transferase (GST)–tagged proteins and interacting
PIP-containing fluorescent liposomes (Fig. 2). Following a single
wash, PIP specificity is simply measured through fluorescence
reading (Fig. 2). Despite that this assay requires the use of
GST-tagged recombinant proteins, the PLIF strategy has signifi-
cant advantages compared to other methods in terms of execution
speed and experimental setup (Fig. 2).
Fig. 1 Experimental principles of liposome sedimentation and flotation assays. Liposomes containing a given
PIP are incubated with recombinant PIP-binding domain or protein and subjected to sedimentation or flotation
using sucrose gradient. The binding reaction is then analyzed by SDS-PAGE followed by Coomassie blue
(CB) staining or western blot detection
Protein-Phosphoinositide Interaction 179
Fig. 2 Experimental principles of the PLIF assay. Each GST-tagged protein is incubated on glutathione-coated
plates with fluorescent liposome preparations containing a given PIP. The binding reaction is proportional to
fluorescent signal intensity of the bound liposomes
2 Materials
2.1 Liposome Liposomes are composed of phospholipids (POPC and POPE) and
Preparation a fluorescent lipid (carboxyfluorescein-PE) (molar ratio: 70:28:2)
to follow the liposomes, with or without 1% of a given PIP taken at
the expense of POPC (see Note 1).
1. 10 mg/mL 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocho-
line (POPC) in chloroform/methanol in a ratio of 1/1 (v/v).
Store at 20 C up to 6 months (see Note 2).
2. 10 mg/mL 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoetha-
nolamine (POPE) in chloroform/methanol in a ratio of 1/1
(v/v). Store at 20 C up to 6 months (see Note 2).
3. 1 mg/mL carboxyfluorescein-phosphoethanolamine (CF-PE).
Store at 20 C up to 6 months (see Note 2).
4. 1 mM PtdIns3P, PtdIns4P, PtdIns5P, PtdIns(3,4)P2, PtdIns
(3,5)P2, PtdIns(4,5)P2, and PtdIns(3,4,5)P3 in chloroform/
methanol in a ratio of 1/1 (v/v). Store at 20 C up to
6 months (see Note 2).
5. Liposome resuspension buffer: PBS containing 150 mM potas-
sium acetate. Prepare 5 mL (see Note 3).
6. Nitrogen evaporator.
7. Vacuum dessicator.
8. Water bath sonicator.
2.2 Liposome 1. 1 mM fluorescent PIP-containing liposomes (see Subheading

Flotation 2.1).
2. Ultracentrifuge that can adapt ~1 mL thickwall polycarbonate
tubes.
3. 60% and 25% sucrose solutions in PBS (w/o Ca2+/Mg2+):
dissolve the sucrose in PBS and check density with a
refractometer and adjust with PBS. To make the 60% sucrose

solution, it is easier to check a twofold dilution of a sample with
the refractometer and to adjust the mother solution until a 30%
solution is obtained. Prepare ~10 mL of each solution. Sucrose
stock solutions can be stored at 4 C (see Note 3).
4. Portable UV lamp.
2.3 Protein–Lipid 1. GST-tag recombinant protein (well dialyzed, see Note 4). 8 μg
Interaction by of protein is necessary for each experiment and depending on
Fluorescence (PLIF) the number of samples.
2. 1 mM fluorescent phosphoinositide-containing liposomes (see
Subheading 2.1).
3. Liposome-binding buffer: PBS (w/o Ca2+/Mg2+) containing
150 mM potassium acetate and 1 mM MgCl2. Prepare 50 mL
(see Note 3).
4. Liposome lysis buffer: PBS (w/o Ca2+/Mg2+) containing 1%
(v/v) Triton X-100. Prepare 50 mL.
5. Glutathione-coated plates, clear, 8-well strip (Thermo Scien-
tific Pierce).
6. Multichannel pipette (for 100 and 200 μL pipetting).
7. Reagent reservoir for multichannel pipette.
8. Fluorescence plate reader for 96-well microtiter plate.
3 Methods
3.1 Liposome This protocol describes how to prepare liposomes that can be used
Preparation for liposome flotation or PLIF assay. During lipid mixing, to limit
solvent evaporation, keep the lipid stocks and prepare lipid mixtures
on ice.
1. 200 μL of 1 mM liposome suspension composed of POPC:
POPE:CF-PE (molar ratio: 70:28:2) and POPC:POPE:CF-
PE:PIP (molar ratio: 69:28:2:1) is a good starting point for
multiple experiments during 2 days. In glass tubes labeled for
each PIP, add the different phospholipids (POPC, POPE,
CF-PE PIP) in 200 μL final volume of chloroform. For
control liposomes, in 181 μL of chloroform, add 10.6 μL of
POPC at 10 mg/mL, 4 μL of POPE at 10 mg/mL, and
4.55 μL of CF-PE at 1 mg/mL. For PIP-containing liposomes,
in 179 μL of chloroform, add 10.5 μL of POPC at 10 mg/mL,
4 μL of POPE at 10 mg/mL, 4.55 μL of CF-PE at 1 mg/mL,
and 2 μL of PIP at 1 mM.
2. Incubate the glass tubes for 10 min at 37 C to homogenize
lipid mixtures.
3. Evaporate the solvents for 30 min under nitrogen to form a

dry film.
4. Place the glass tubes in the vacuum chamber overnight to
eliminate traces of solvent (see Note 5).
5. Resuspend the lipids in 200 μL of liposome resuspension buffer
preheated at 37 C to give 1 mM lipid suspensions.
6. Place the glass tubes for 1 h in a 37 C water bath and apply
vigorous vortexing every 10 min.
7. Subject the lipid suspensions to 5 cycles of freezing and thaw-
ing using liquid nitrogen and a 37 C water bath. Centrifuge at
100 g for 1 min and transfer the lipid mixtures to 1.5 mL
microcentrifuge tubes (see Note 6). After this stage, lipids can
be stored at 80 C for up to 1 month.
8. Thaw the lipid suspensions and sonicate for 60 min in a water
bath sonicator filled with ice. The lipid suspension should be
clear; otherwise, sonicate for a few more minutes.
Using this protocol, liposomes are around 50 nm in dia-
meters (as measured by dynamic light scattering). Liposomes
can be kept for 2–3 days at 4 C, protected from light (see Note
7).
3.2 Liposome 1. In thickwall polycarbonate tubes, mix 1–10 μg of the recombi-

Flotation nant protein of interest with 20 μL of each liposome mixture in
a final volume of 100 μL of PBS (w/o Ca2+/Mg2+) (this
corresponds to a 5 μM concentration of exposed PIPs) (see
Note 8).
2. After a 20 min incubation at room temperature, add 100 μL of
60% sucrose in PBS to yield a 30% final sucrose concentration.
Mix gently to have a homogeneous solution.
3. Overlay with 250 μL of 25% sucrose and 50 μL of PBS (w/o
Ca2+/Mg2+) on top without disturbing the layers (see Note 9).
4. Centrifuge at 200,000 g for 1 h at 20 C. During centrifu-
gation, the fluorescent liposomes migrate to the top buffer
layer.
5. Using a portable UV lamp, check the presence of the liposomes
on the top buffer layer and collect the fractions from bottom to
top using a Hamilton syringe. Fractions volume: bottom
200 μL, middle 200 μL, and top 100 μL.
6. Analyze 20–30 μL of each fraction by SDS-PAGE followed by
western blot analysis or Coomassie staining.
3.3 Protein–Lipid 1. Thaw your GST-tag recombinant proteins of interest (keep on

Interaction by ice) and place your liposomes on the bench protected from the
Fluorescence (PLIF) light (using aluminum foil, for example) (see Note 10).
2. Fill three reagent reservoirs for multichannel pipette with

50 mL of PBS (w/o Ca2+/Mg2+), liposome-binding buffer,
and liposome lysis buffer (room temperature).
3. Remove the number of strips required from glutathione-coated
plate and place in strip holder. One 8-well strip is necessary to
test one recombinant protein with each liposome (control,
PtdIns3P, PtdIns4P, PtdIns5P, PtdIns(3,4)P2, PtdIns(3,5)P2,
PtdIns(4,5)P2, PtdIns(3,4,5)P3). Return remaining strips to
storage.
4. Add 200 μL of PBS (w/o Ca2+/Mg2+) in each well using a
multichannel pipette and incubate for 2 min on an orbital
microplate shaker, 300 rpm. Remove all solutions from the
wells with a vigorous flick of the plate and series of pats onto
paper towels (5–7 hard pats). Repeat the washes twice.
5. In each well, in a 100 μL final volume of liposome-binding
buffer, add 1 μg of GST-tag recombinant protein diluted in
liposome-binding buffer and 5 μL of fluorescent
PIP-containing liposomes (this corresponds to a 250 nM con-
centration of exposed PIPs) and incubate on an orbital micro-
plate shaker at 300 rpm for 20 min (see Note 11).
6. Flick out the solutions from the wells and wash once with
200 μL of liposome-binding buffer using the multichannel
pipette. Remove all solutions from the wells with a vigorous
flick of the plate and series of pats onto paper towels (5–7
hard pats).
7. Add 100 μL of liposome lysis buffer using a multichannel
pipette and incubate on an orbital microplate shaker for
5 min at 300 rpm.
8. Transfer 100 μL from each well in a new white 96-well microti-
ter plate using a multichannel pipette. Avoid making bubbles,
which may alter the fluorescence measurement. The transfer in
a new white plate is necessary because well strips are transparent
and not usable for fluorescence reading.
9. Read the fluorescence with a fluorescence plate reader for the
96-well microtiter plate (λex, 485; λem, 518 nm) and plot the
results with the name of each PIP on abscissa axis and the
fluorescence on the ordinate axis (see Note 12).
4 Notes
1. Hydration of lipids is an important step for liposomes forma-

tion. Thus, it is important to use phospholipids with low
transition temperatures (Tc or Tm) such as POPC and POPE.
Other phospholipids can also be included in liposome
composition such as POPS and cholesterol and with different

ratios to mimic specific cellular compartment, for example.
2. Phospholipids in organic solution should be stored in a glass
container layered with nitrogen. Saturate the atmosphere with
nitrogen each time you open a tube to avoid lipid oxidation and
wrap with parafilm.
3. Other buffers than PBS can be used such as HEPES-buffered
saline. This is important for low-affinity interactions for which
the phosphate present in PBS can act as a competitor for PIP
binding. Protein–phosphoinositide interaction can also be
dependent on the pH mainly through protonation of histidine
residues.
4. It is very important that your recombinant protein of interest is
well dialyzed to eliminate the reduced glutathione used to elute
your protein during the purification steps. This is because the
PLIF assay involves the interaction of proteins with
glutathione-coated wells.
5. It is very important to eliminate all solvent traces using the
vacuum chamber because solvent can interfere with liposome
formation.
6. Liposome formation by sonication is much more efficient in
microcentrifuge tubes than glass tubes.
7. As an alternative, liposomes can be formed using an extruder to
produce liposomes of different sizes.
8. The use of negative (for example, GST when a GST-tagged
protein is the protein of interest) and positive controls (for
example, GST-FYVE (HRS), which interacts with PtdIns3P;
GST-PH (PLCδ1), which interacts with PtdIns(4,5)P2); and
GST-PH (GRP1), which interacts with PtdIns(3,4,5)P3) is
highly recommended.
9. The use of gel-loading tips for this task can be very helpful.
10. As for liposome flotation, the use of positive controls (for
example, GST-FYVE (HRS), which interacts with PtdIns3P;
GST-PH (PLCδ1), which interacts with PtdIns(4,5)P2); and
GST-PH (GRP1), which interacts with PtdIns(3,4,5)P3) is
highly recommended.
11. By doing the experiment with increasing concentration of
liposomes, one can have access to relative affinity determina-
tion. Fluorescence values obtained for each concentration and
for each PIP tested are then plotted and fitted using one-site
binding (hyperbola) to obtain affinity values [5].
12. If no fluorescence signal was obtained whereas your positive
control gave expected results, this can be due to low affinity of
your protein for PIP. In this case, you can increase the
concentration of liposomes in your experiments to increase the

concentration of exposed PIPs or increase the percentage of
PIPs in your liposomes to increase the concentration range. If
high background signal is obtained, this is due to an
inefficient wash.
Acknowledgments
This work was supported by grants from Inserm, Fondation pour la

Recherche Médicale (DEQ20170336737), E-Rare (TREAT-
MTMs), and AFM-Téléthon (#20879 to K.H.). M.P. is a recipient
of a PhD fellowship from AFM-Téléthon (#22115). B.P. is a
scholar of the Institut Universitaire de France.
References
1. Viaud J, Mansour R, Antkowiak A et al (2016) domains. Methods 39(2):122–133. http://sci-

Phosphoinositides: important lipids in the coor- hub.tw/10.1016/j.ymeth.2006.05.006
dination of cell dynamics. Biochimie 4. Bigay J, Antonny B (2005) Real-time assays for
125:250–258. http://sci-hub.tw/10.1016/j. the assembly-disassembly cycle of COP coats on
biochi.2015.09.005 liposomes of defined size. Methods Enzymol
2. Dowler S, Kular G, Alessi DR (2002) Protein 404:95–107. http://sci-hub.tw/10.1016/
lipid overlay assay. Sci STKE 2002(129):pl6. S0076-6879(05)04010-3
http://sci-hub.tw/10.1126/stke.2002.129. 5. Ceccato L, Chicanne G, Nahoum V et al (2016)
pl6 PLIF: a rapid, accurate method to detect and
3. Narayan K, Lemmon MA (2006) Determining quantitatively assess protein-lipid interactions.
selectivity of phosphoinositide-binding Sci Signal 9(421):rs2. http://sci-hub.tw/
10.1126/scisignal.aad4337
Chapter 13
Liposome-Based Methods to Study GTPase Activation by

Phosphoinositides
Julien Viaud, Laurie Ceccato, Bernard Payrastre, and
Frédérique Gaits-Iacovoni
Abstract
Phosphoinositides (PIPs) are lipid messengers with different functions according to their localization. After
their local production by the action of lipid kinases or phosphatases, PIPs regulate various biological
processes such as cytoskeleton rearrangement, membrane remodeling/trafficking, or gene expression
through binding of their phosphorylated inositol head group with different protein domains such as PH,
PX, and FYVE. It is well known that PIPs regulate the activity of small GTPases by interacting with and
activating Guanyl-nucleotide Exchange Factor (GEF) proteins through specific domains such as the ones
mentioned above. However, most of the in vitro assays to assess the activation of GTPases focus on the
GTPase only and neglect the fact that co-activators, such as membranes and protein activators, have a
significant effect in vivo. Herein, we describe not only the classical protein–lipid overlay and liposome
sedimentation methods but also an assay we have developed, which contains three partners: a liposome
which composition reproduces the membrane of the target of the GTPase, the recombinant specific
DH-(PIP affinity) GEF domain, and the recombinant GTPase to be tested by different PIPs. This assay
allows us to clearly quantify the GTPase activation.
Key words Phosphoinositide, GTPase–liposome interaction, GTPase exchange factors, Liposome
1 Introduction
Small GTPases of the Rho family are involved in many biological

processes such as adhesion, migration, invasion, remodeling of the
cytoskeleton, cell division, secretion, membrane trafficking, and so
on [1]. They are tightly regulated and cycle between an active state,
in which they are loaded with GTP upon activation by their positive
regulators, the GEFs (Guanyl-nucleotide Exchange Factors), lead-
ing to interaction with cellular targets. They are inactive in terms of
signal transduction when bound to GDP since their enzymatic
activity is under the influence of GAPs (GTPase-activating pro-
teins) to which they also bind. Shifting from an active to inactive
form is a rapid and localized event difficult to study both in vivo and
Roberto J. Botelho (ed.), Phosphoinositides: Methods and Protocols , Methods in Molecular Biology, vol. 2251,
185
186 Julien Viaud et al.
in vitro due to the complexity of the mandatory interactions of lipid

and protein partners involved. They have been demonstrated to be
activated by members of the phosphoinositide (PIPs) family via
binding to those lipids second messengers using specific domains
such as PH domains often found in the GEFs [2]. This implies that
the composition of the membranes to which the GTPases are
associated when activated presents a very specific composition
that is detrimental in their timely local activation in vivo. Since
the binding of GEF–PH (Pleckstrin homology) domains with the
regulating PIs is very elusive and not very strong, it has been
difficult to set up robust assays to study their activation in vitro
[3–5]. Herein, we describe two assays that have been used by
several teams. The first is the lipid–protein overlay assay (so-called
Fat blot), which has given interesting clues but has the major
problem of presenting a pure solution of PIs spotted on a nitrocel-
lulose membrane [6]. This is very far from the cell reality since the
lipids are not arranged in bilayers, are on a flat surface, leaving the
curvature element off the equation, and are not in the specific
environment of the different types of membranes and compart-
ments containing a defined mix of different phospholipids (e.g.,
PC (phosphatidylcholine), PE (phosphatidylethanolamine), PS
(phosphatidylserine)), thereby leading to possible false-positive
data that might prove irrelevant in vivo. The second is the liposome
pull-down assay. In this assay, the lipid composition of membranes
and their curvature can be mimicked in vitro to match a chosen
subcellular compartment. However, liposomes do not include
important regulatory protein partners or cofactors since those are
difficult to include in a hydrophobic environment [7]. To bypass
those difficulties, we have developed a technique that defines a
specific membrane composition, including protein components
[8]. It contains the PIP of choice, the GTPase which is actually
anchored to the membrane by the use of a recombinant His-tagged
GTPase loaded with GDP-BODIPY, and DSG-NTA(Ni) included
in the liposomes (Fig. 1a). Then, the DH-PH domain protein
(DH is the active domain of the GEF, while PH binds to a specific
PIP) and the GTPase are added to a lipid environment that mimics
the in vivo biological membrane. We can then observe changes to
GTPase activation by fluorimetry while manipulating several factors
such as lipid composition, liposome curvature, and the requirement
of a given GEF and its conditions of action (Fig. 1b). Importantly,
this technique can be set up in most of biology laboratories.
2 Materials
Liposomes are composed of phospholipids POPC:POPE:DSG-

NTA(Ni) (molar ratio: 67:30:3) with or without 6% of a given
PIP taken at the expense of POPC (see Note 1).
GTPases Activation by Phosphoinositides 187
120
spontaneous exchange
Fluorescence (% of time 0)
100
EDTA
80
Tiam1 (PC:PE:DGS (Ni))
60
Tiam1
40 (PC:PE:DGS(Ni):PI3P)
Tiam1
Tiam1
0 500 1000 1500 2000 2500
Time (Sec)
Fig. 1 Liposomes and GEF-based in vitro assay to monitor the activation of Rac1. (a) The assay was performed
on Ni2+-NTA-containing liposomes (with different PIP species) bound to Rac-His loaded with BODIPY FL-GDP.
Recombinant Tiam1 DH-PHc was added, and the nucleotide exchange reaction was started by adding an
excess of GTP. The exchange reaction was followed by fluorospectroscopy. (b) Recombinant Tiam1 DH-PHc
was preincubated with the liposome-binding Rac1-His-GDP, and the exchange reaction (BODIPY FL-GDP/GTP)
was started by adding GTP. The data were fit as a single exponential decay
2.1 Liposome 1. Lipid stock solutions are in chloroform/methanol in a ratio of

Preparation for Rac 1/1 (v/v), in small glass tubes. 1-Palmitoyl-2-oleoyl-sn-gly-
Activation Assay cero-3-phosphocholine (POPC) (Avanti) and 1-palmitoyl-2-
oleoyl-sn-glycero-3-phosphoethanolamine (POPE) (Avanti) at

a concentration of 10 mg/mL, 1,2-dioleoyl-sn-glycero-3[(N-
(5-amino-1carboxypentyl) iminodiacetic acid)succinyl] (Nickel
salt) at a concentration of 1 mg/mL, and PI3P (Echelon),
PI4P (Avanti), or PI5P (Echelon) (see Note 2). Saturate the
atmosphere with nitrogen each time you open a tube to avoid
lipid oxidation and wrap with parafilm. Store up to 6 months at
20 C.
2. Liposome resuspension buffer: 50 mM HEPES, pH 7.5,
100 mM NaCl, 200 mM sucrose, and 5 mM EGTA in
ddH2O. Prepare 5 mL (see Note 3) or enough to resuspend
the lipid mixtures in 200 μL for the experiments.
3. Nitrogen evaporator.
4. Vacuum desiccator.
5. Water bath sonicator.
6. Chemical fume hood for solvents.
2.2 Rac1-His 1. Competent BL21 (DE3) E. coli bacteria (see Note 4).
Purification 2. Rac1 resuspension buffer: 50 mM HEPES, pH 7.5, 150 mM
NaCl, 10 mM imidazole, 100 μM GDP, 5 mM MgCl2,
0.5 mg/mL lyzozyme, 1 μM aprotinin, 10 μM leupeptin,
1 mM PMSF, 10% glycerol, 1% Triton X-100 in ddH2O.
3. Rac1 washing buffer: 50 mM HEPES, pH 7.5, 150 mM NaCl,
40 mM imidazole, 10 μM GDP, 5 mM MgCl2 in ddH2O.
4. Rac1 elution buffer: 50 mM HEPES, pH 7.5, 140 mM NaCl,
400 mM imidazole, 10 μM GDP, 5 mM MgCl2 in ddH2O.
5. Rac1 dialysis buffer: 50 mM HEPES, pH 7.5, 150 mM NaCl,
2 μM GDP in ddH2O.
6. Rotating dry incubator.
7. Big centrifuge for 500-mL plastic bottles.
8. Sonicator.
9. Sterile double flux hood to manipulate bacteria in a sterile
environment.
10. Vivaspin 6 conical concentrators.
11. TALON on sepharose beads slurry.
12. BODIPY-GDP.
2.3 Liposome 1. Rac1-liposome resuspension buffer: 50 mM HEPES, pH 7.5,

Preparation and Rac1- 100 mM NaCl, 200 mM sucrose, 5 mM EGTA.
His Binding 2. Rac1-His binding buffer: 50 mM HEPES, pH 7.5, 120 mM
potassium acetate, 1 mM MgCl2, and 1 mM DTT in ddH2O.
2.4 Tiam DH-PHc 1. Rotating dry incubator.

Recombinant Domain 2. Big centrifuge for 500-mL plastic bottles.
Production
3. Sonicator.
4. Sterile double flux hood to manipulate bacteria in a sterile
environment.
5. Vivaspin 6 conical concentrators.
6. Sepharose 4B–glutathione slurry.
2.5 DH-PH Exchange 1. HKM buffer: 50 mM HEPES, pH 7.2, 120 mM K-acetate,

Factor Activity 1 mM MgCl2 in ddH2O; store at 4 C.
2. Exchange assay buffer: 50 mM Tris, pH 7.5, 100 mM NaCl,
10 mM MgCl2.
3. 96-well black microplates.
4. Varioskan flash spectrofluorometer.
2.6 Liposome 1. Liposome sedimentation resuspension buffer: 10 mM HEPES,

Sedimentation Assay pH 7.7, 100 mM NaCl, 200 mM sucrose, 5 mM EGTA.
2. Rac1-liposome binding buffer: 50 mM Tris, pH 7.5, 100 mM
NaCl, and 10 mM MgCl2.
3. Immobilon P membrane.
4. Horseradish peroxidase-coupled secondary antibodies.
5. X-ray film developer or ChemiDoc fluorescence camera.
2.7 Lipid–Protein 1. Recombinant GST-tagged Rac1 proteins were expressed in

Overlay Assay (Fat E. coli strain BL21(DE3) as described above. They were then
Blots) purified as GST-Rac1 by affinity chromatography using gluta-
thione Sepharose 4B beads, eluted with 10 mM reduced gluta-
thione in ddH2O, and dialyzed against PBS, pH 7.4.
2. PIP strips (Echelon) containing increasing amounts of
various PIPs.
3. Anti-GST antibody.
4. Horseradish peroxidase-coupled secondary antibodies.
5. Immobilon P membrane.
6. X-ray film developer or ChemiDoc fluorescence camera.
3 Methods
3.1 Rac-His 1. Rac1 is subcloned in the pET-28a vector (6 his in C-terminal).

Purification 2. 5 mL of preculture of a colony of BL21 (DE3) E. coli is
cultured in 1 L culture O/N at 37 C under agitation.
3. Centrifuge two tubes of 500 mL at 6000 g for 20 min.
4. Resuspend both pellets in 30 mL of Rac1 resuspension buffer.

5. After strong sonication and centrifugation for 30 min at
10,000 g at 4 C, add 500 μL of TALON slurry to the
supernatant and incubate for 3 h at 4 C rotating.
6. Wash five times with Rac1 washing buffer.
7. Elute the bound Rac1-His six times in 500 μL of Rac1 elution
buffer.
8. Dialyze 3 mL of the eluate in 2 L of Rac1 dialysis buffer twice in
fresh dialysis buffer for 2 h each (see Note 5).
9. Post-dialysis, the eluate is concentrated on Vivaspin 6 centrifu-
gal concentrators (molecular weight cutoff (MWCO), 10 kDa)
up to a 1-mL solution; glycerol (10% final) is added.
10. Samples are aliquoted (about 100 μL routinely), flash frozen in
liquid nitrogen, and stored at 80 C.
3.2 Charging of 1. Mix in 550 μM final, 161 nM Rac1-His, 300 nM BODIPY-

Recombinant Rac1-His GDP, and 2 mM EDTA in ddH2O.
with BODIPY-GDP 2. Incubate for 20 min at 30 C.
3. Terminate the reaction by adding 5 mM MgCl2.
4. Change the buffer by using Vivaspin (MWCO 10 kDa) for
PBS, pH 7.5.
3.3 Liposome 1. After mixing the different lipids required for your experiment,
Preparation and Rac1- incubate the glass tubes for 10 min at 37 C to homogenize
His Binding lipid mixtures.
2. Evaporate the solvents for 30 min under nitrogen to form a
dry film.
3. Place the glass tubes in the vacuum chamber overnight to
eliminate traces of solvent (see Note 3).
4. Resuspend the lipid in 200 μL of Rac1-liposome resuspension
buffer preheated at 37 C to give 1-mM lipid suspensions.
5. Place the glass tubes for 1 h in a 37 C water bath with vigorous
vortexing every 10 min.
6. Subject the lipid suspensions to 5 cycles of freezing and thaw-
ing using liquid nitrogen and a 37 C water bath. After a brief
centrifugation, transfer the lipid mixtures in 1.5-mL microcen-
trifuge tubes.
7. Leave the mix for 10 min to equilibrate at room temperature.
8. Extrude 19 times through a 0.1-μM polycarbonate filter using
a hand extruder to generate liposomes of about 100 nm of
diameter (see Note 6).
9. Liposomes were kept at 4 C and used within 4 days of their
preparation.
10. Increasing concentrations of Rac1-His were incubated with

300 μM liposomes containing POPC:POPE:DSG-NTA(Ni)
(molar ratio: 67:30:3) with or without 6% of a given PIP
(prepared as described in Subheading 2.1) for 20 min at
room temperature in Rac1-His binding buffer.
11. Liposomes were then pelleted by centrifugation (1 h,
16,000 g, 20 C).
12. The fluorescent liposome pellet was visualized using a
UV lamp.
13. Supernatant and pellet fractions were analyzed by SDS-PAGE
and Coomassie staining.
14. They were quantified against a range of known amounts of
bovine serum albumin (BSA).
3.4 Tiam DH-PHc 1. The C terminal DH-PH domain of Tiam1 was subcloned in
Recombinant Domain pGEX-2T (contains the GST (glutathione S transferase) tag).
Production 2. Production of the GST-DH-PHc protein was obtained by an
ON subculture of a colony of plasmid containing BL21 in
15 mL of LB at 37 C under agitation.
3. 1 L of LB was seeded with the preculture and cultivated under
the same conditions up to OD600 at 0.6 < OD < 0.9.
4. Bacteria were pelleted as described above in Subheading 2.2
and resuspended in 9 mL of PBS (phosphate-buffered saline,
pH 7.4) containing 1 mg/mL lysozyme.
5. After 30 min on ice, the solution was sonicated 4 30 s, and
2 mL of 10% Triton X100 and some antiprotease cocktail were
added.
6. After another 30 min on ice, the samples were spinned at
10,000 g for 30 min at 4 C.
7. The supernatant was transferred to a conical 50-mL tube and
500 μL of sepharose 4B–glutathione 50% slurry and rotated for
3 h at 4 C.
8. Beads were centrifugated and washed twice with 30 mL of PBS
containing 0.1% Tween and then once with PBS alone.
9. The GST tag from the GST-DH-PHc recombinant protein was
then cleaved by thrombin.
10. Beads were incubated at RT with 15 units of thrombin from
bovine plasma for 1 h.
11. Thrombin was then inactivated in the supernatant post-
centrifugation of the beads with 200 μL of p-aminobenzami-
dine (prepared in ddH2O).
12. The buffer of the eluate was then changed on Vivaspin MWCO
10 kDa as described in Subheading 2.2 to PBS containing 10%
glycerol, aliquoted, flash frozen, and stored at 80 C.
3.5 Exchange Factor 1. 300 μM liposomes PC:PE:DGS-NTA(Ni) (67:30:3) or PC:PE:

Activity DGS-NTA(Ni):PtdInsP (61:30:3:6) were incubated with
1 μM Rac-His (loaded with BODIPY FL-GDP) and diluted
appropriately in HKM buffer in 100 μL final of exchange assay
buffer for 30 min at 30 C.
2. Then, 100 nM Tiam1 DH-PHc (GST tag was cleaved with
thrombin; see Subheading 3.4) was added for an additional
10 min.
3. Reaction was started by adding 400 μM GTP and carried out at
30 C.
4. Activation of Rac1 was monitored by fluorescence of released
BODIPY (488/535 nm).
5. Fluorescence intensity was recorded every 15 s for 40 min
using a Varioskan flash spectrofluorometer (see Note 7).
3.6 Liposome 1. The liposomes were prepared as described in Subheading 2.1

Sedimentation Assay but contained different ratios of lipids PC:PE (50:50) or PC:
PE:PIP(n) (45:45:10).
2. Briefly, they were prepared by mixing chloroform stocks of the
various lipids, evaporating chloroform under nitrogen flow, and
then resuspending the dryed lipids with liposome resuspension
buffer using a bath sonicator.
3. Liposome sedimentation assays were conducted using 300 μM
liposomes and 2 μg of GST fusion Rac1 in Rac1-liposome
binding buffer.
4. After 20 min of incubation at room temperature, one wash was
performed in PBS.
5. All centrifugation steps were done at 16,000 g and 4 C.
Supernatant and pellet fractions were analyzed by SDS-PAGE
and western blotting on immobilon P.
6. Binding of the bound GST-Rac1 was detected after SDS-PAGE
and western blotting using anti-GST antibody.
7. Immunoreactive bands were detected by chemiluminescence
using either medical X-ray film or ChemiDoc with the Super-
Signal detection system. Quantification was done using ImageJ
or ChemiDoc Image lab (see Note 8).
3.7 Lipid–Protein 1. PIP strips (Echelon Biosciences Inc) and in-house membranes
Overlay Assay (Fat were used as previously described [8]. Membranes were
Blots) blocked in TBS or PBS containing 3 mg/mL of FAT-free
BSA and 0.1% Tween 20.
2. Incubations were done with 0.5 μg/mL of purified GST-Rac1

protein in blocking buffer TBS containing 0.1% Tween 20.
3. Detection of the bound GST-Rac1 was done with an anti-GST
antibody after incubation at room temperature for 1–2 h.
4. After four washes with the blocking buffer, binding of the
domains to the PIPs was detected by the anti-GST western
blotting on Immobilon-P membranes as described in Subhead-
ing 3.6 (see Notes 9 and 10).
4 Notes
1. Hydration of lipids is an important step for liposome forma-

tion. Thus, it is important to use phospholipids with low
transition temperatures (Tc or Tm) such as POPC and POPE.
Other phospholipids can also be used to form liposomes such
as POPS or cholesterol. Different ratios can then be applied to
mimic a specific cellular compartment.
2. We have tested different lipid providers for all our lipid-
containing experiments, and clearly, they should be bought as
indicated, either from Avanti or from Echelon, since the pro-
tocols have been optimized and successful with lipids from
these companies.
3. It is very important to eliminate all solvent traces using the
vacuum chamber because the solvent interferes with liposome
formation.
4. The BL21(DE3) E. coli strain has been optimized for exoge-
nous protein productions since it expresses the mRNA-
encoding protein of the DE3 bacteriophage that is activated
upon IPTG induction to produce the target protein on the
expression vector. Moreover, it has a better tolerance for exog-
enous proteins expressed in big amounts since it has been
mutated to form less inclusion bodies from which it is hardly
possible to recover functional recombinant proteins and is also
defective for genes encoding proteases that would degrade
overexpressed proteins.
5. The recombinant protein should be well dialyzed to eliminate
the imidazole used for elution because it could be detrimental
to your reaction later on by binding histidine residues on
proteins.
6. Liposomes can be formed using an extruder membrane with
different size holes to produce liposomes of different sizes.
7. If a very weak fluorescence decrease is obtained while the
positive control (adding EDTA) gives a good decreasing
curve, one can suspect a low affinity of the GEF for the tested
PIP. In this case, try increasing the concentration of liposomes

in the experiments to increase the concentration of exposed
PIPs or increase the percentage of PIPs in the liposomes to
increase the concentration range.
8. For liposome sedimentation, the use of positive controls (e.g.,
GST-FYVE (HRS), which interacts with PI3P; GST-PH
(PLCδ1), which interacts with PI(4,5)P2); and GST-PH
(GRP1), which interacts with PI(3,4,5)P3) is highly
recommended.
9. If all the lipids show a binding of GST-Rac1 by lipid overly
assay, use PIP strips presenting different concentrations of the
lipids. You can also decrease the amount of recombinant
GST-Rac1 added to the reaction, or use an incubation buffer
slightly more stringent (0.1%NP40, for example).
10. The use of negative controls (e.g., GST alone) and positive
controls (e.g., GST-FYVE (HRS), which interacts with PI3P;
GST-PH (PLCδ1), which interacts with PI(4,5)P2); and
GST-PH (GRP1), which interacts with PI(3,4,5)P3) is highly
recommended.
Acknowledgments
This work was supported by grants from Inserm, Fondation pour la

Recherche Médicale (DEQ20170336737), La ligue contre le Can-
cer (RPV17013BBA), and the ARC Foundation against cancer
(RAC17004BBA) attributed to F. Gaits-Iacovoni.
References
1. Heasman SJ, Ridley AJ (2008) Mammalian Rho 5. Viaud J, Mansour R, Antkowiak A et al (2016)
GTPases: new insights into their functions from Phosphoinositides: important lipids in the coor-
in vivo studies. Nat Rev Mol Cell Biol 9 dination of cell dynamics. Biochimie
(9):690–701. https://doi.org/10.1038/ 125:250–258. https://doi.org/10.1016/j.bio
nrm2476 chi.2015.09.005
2. Viaud J, Gaits-Iacovoni F, Payrastre B (2012) 6. Dowler S, Kular G, Alessi DR (2002) Protein
Regulation of the DH-PH tandem of guanine lipid overlay assay. Sci STKE 129:pl6. https://
nucleotide exchange factor for Rho GTPases by doi.org/10.1126/stke.2002.129.pl6
phosphoinositides. Adv Biol Regul 52 7. Narayan K, Lemmon MA (2006) Determining
(2):303–314. https://doi.org/10.1016/j.jbior. selectivity of phosphoinositide-binding
2012.04.001 domains. Methods 39(2):122–133. https://
3. Lemmon MA (2008) Membrane recognition by doi.org/10.1016/j.ymeth.2006.05.006
phospholipid-binding domains. Nat Rev Mol 8. Viaud J, Lagarrigue F, Ramel D et al (2014)
Cell Biol 9(2):99–111. https://doi.org/10. Phosphatidylinositol 5-phosphate regulates
1038/nrm2328 invasion through binding and activation of
4. Cook DR, Rossman KL, Der CJ (2014) Rho Tiam1. Nat Commun 5:4080. https://doi.
guanine nucleotide exchange factors: regulators org/10.1038/ncomms5080
of Rho GTPase activity in development and dis-
ease. Oncogene 33(31):4021–4035. https://
doi.org/10.1038/onc.2013.362
Chapter 14
Liposome Co-sedimentation and Co-flotation Assays

to Study Lipid–Protein Interactions
Yosuke Senju, Pekka Lappalainen, and Hongxia Zhao
Abstract
A large proportion of proteins are expected to interact with cellular membranes to carry out their
physiological functions in processes such as membrane transport, morphogenesis, cytoskeletal organiza-
tion, and signal transduction. The recruitment of proteins at the membrane–cytoplasm interface and their
activities are precisely regulated by phosphoinositides, which are negatively charged phospholipids found
on the cytoplasmic leaflet of cellular membranes and play critical roles in membrane homeostasis and
cellular signaling. Thus, it is important to reveal which proteins interact with phosphoinositides and to
elucidate the underlying mechanisms. Here, we present two standard in vitro methods, liposome
co-sedimentation and co-flotation assays, to study lipid–protein interactions. Liposomes can mimic various
biological membranes in these assays because their lipid compositions and concentrations can be varied.
Thus, in addition to mechanisms of lipid–protein interactions, these methods provide information on the
possible specificities of proteins toward certain lipids such as specific phosphoinositide species and can hence
shed light on the roles of membrane interactions on the functions of membrane-associated proteins.
Key words Phosphoinositides, Liposome, Lipid–protein interactions, Phospholipids, Binding mode
1 Introduction
Many proteins are proposed to play their physiological functions

through binding to cellular membranes. The plasma and intracellu-
lar membranes have different lipid compositions that enable locali-
zation of specific proteins to distinct subcellular compartments
[1, 2]. For example, one of the phosphoinositides, phosphatidyli-
nositol 4,5-bisphosphate [PI(4,5)P2], is abundant at the plasma
membrane, where it regulates the functions of several actin-binding
proteins that control the dynamics of the actin cytoskeleton
[3, 4]. Interactions of proteins with PI(4,5)P2 may affect their
association with binding partners or induce conformational
changes in proteins. Such lipid-induced conformational changes
can, for example, release the auto-inhibited protein structures as
demonstrated in the case of actin-regulatory proteins N-WASP and
195
196 Yosuke Senju et al.
Ezrin/Radixin/Moesin [ERM] [5–7]. Furthermore, certain PI

(4,5)P2-binding proteins, including the Bin/Amphiphysin/Rvs
(BAR) domain proteins, can reciprocally regulate membrane prop-
erties such as fluidity, curvature, tension, and lateral diffusion of
lipids [8–10]. Therefore, it is crucial to understand the lipid-
binding affinities, specificities, and mechanisms of proteins.
Here, we present two basic in vitro methods to study lipid–
protein interactions. One is a liposome co-sedimentation assay,
which is a simple method to identify membrane-binding proteins
through sedimentation of the protein-bound liposomes to separate
them from the unbound proteins using ultracentrifugation.
Membrane-binding proteins or protein domains can be identified
by this approach, and their membrane-binding affinities as well as
possible lipid specificities can be uncovered by modifying the con-
centrations and lipid compositions of the vesicles. In addition,
membrane-binding modes (electrostatic vs. hydrophobic,
cooperative vs. noncooperative) of proteins can be determined by
altering the salt concentration in the reaction buffer and the molar
ratio of phosphoinositides. The other method is liposome co-flota-
tion assay, which can be applied to study protein–lipid interactions
in cases where the protein of interest forms large oligomers or
aggregates during the co-sedimentation assay [11]. In the liposome
co-flotation assay, the membrane-binding proteins float with lipo-
somes in a sucrose gradient during centrifugation. Taken together,
in vitro liposome co-sedimentation and co-flotation assays are use-
ful tools to identify membrane-binding proteins and to provide
mechanistic insights into these interactions.
2 Materials
2.1 Phospholipids 1-Palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC) is dis-

solved in chloroform at 4.39 mM. 1-Palmitoyl-2-oleoyl-sn-gly-
cero-3-phosphoethanolamine (POPE) is dissolved in chloroform
at 4.64 mM. 1-Palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine
(POPS) is dissolved in chloroform at 4 mM. PI(4,5)P2 is dissolved
in chloroform:methanol:water (20:9:1) at 0.57 mM. Rhodamine
DHPE is dissolved in chloroform at 0.25 mM. The phospholipids
should be purged with an inert gas (e.g., N2) to prevent oxidation
and stored in glass tubes at 20 C (see Note 1).
2.2 Stock Solutions 1. Lysis buffer: 20 mM Tris–HCl (pH 7.5), 150 mM sodium
chloride (NaCl), 0.1% Triton X-100, 1 mM ethylenediamine-
tetraacetic acid (EDTA), 1 mM phenylmethanesulfonyl fluo-
ride (PMSF), and 1 mM dithiothreitol (DTT).
2. Cleavage buffer: 50 mM Tris–HCl (pH 7.5), 150 mM NaCl,
1 mM EDTA, and 1 mM DTT.
In vitro Assays to Study Lipid-Protein Interactions 197
3. Binding buffer (HEPES-buffered saline, HBS): 20 mM

HEPES (pH 7.5) and 100 mM NaCl (see Notes 2 and 3).
4. HBS containing 0.3 M, 60%, and 25% (w/v) sucrose.
5. Laemmli sample buffer (4 concentration): 8% SDS, 40% glyc-
erol, 20% 2-mercaptoethanol, 0.02% bromophenol blue, and
0.25 M Tris–HCl, pH 6.8.
2.3 Reagents and 1. Glutathione S-transferase (GST) or His-tagged vector.

Equipment for Protein 2. BL21(DE3) Competent E. coli.
Expression and
3. Glutathione agarose.
Purification
4. LB medium.
5. Ampicillin (100 mg/mL) in sterile distilled/deionized water
stored at 20 C.
6. Isopropyl β-D-1-thiogalactopyranoside (IPTG) (0.1 M) in ster-
ile distilled/deionized water stored at 20 C.
7. Sonicator.
8. Protease.
9. Chromatography.
10. Gel filtration column.
11. Centrifugal protein concentrators.
12. Ni-NTA agarose.
2.4 Reagents and 1. Methanol.

Equipment for Co- 2. Chloroform.
sedimentation and Co-
3. Glass tubes.
flotation Assays
4. Glass syringes.
5. Nitrogen gas.
6. Pressured gas blowing concentrator.
7. Vacuum concentrator.
8. Vortex mixer.
9. Tabletop ultracentrifuge.
10. Fixed-angle rotor (20 0.2 mL).
11. Thickwall polypropylene tube (230 μL).
12. Swinging-bucket rotor (4 2.2 mL).
13. Thinwall polypropylene tube (2.2 mL).
14. Microvolume UV–Vis spectrophotometer.
15. Microcentrifuge tubes, 1.5 mL.
16. Heat block.
17. Pre-cast or “home-made” gels for sodium dodecyl sulfate-

polyacrylamide gel electrophoresis (SDS-PAGE).
18. Coomassie brilliant blue (CBB) [0.25% (w/v)] in 40% (v/v)
methanol and 7% (v/v) acetic acid.
19. Gel imaging analysis software.
20. Fluorescence spectrometer.
3 Methods
3.1 Protein 1. Prepare a construct by subcloning a gene encoding a protein of

Expression and interest into a GST-tagged vector (see Note 4).
Purification (1L 2. Transform the construct into BL21 (DE3) competent cells (see
Culture) Note 5).
3. Pick a single colony and inoculate it into 10 mL of LB medium
containing 100 μg/mL of ampicillin (or other antibiotics
depending on vectors) and incubate at 37 C with shaking at
200 rpm overnight. Then, add the entire volume to 1000 mL
of LB medium containing 100 μg/mL of ampicillin.
4. Incubate the culture medium at 37 C with shaking until the
optical density at 600 (OD600) reaches 0.6. Then, induce
protein expression by adding IPTG to a final concentration of
0.2 mM (see Note 6).
5. After overnight induction at 16 C, harvest the cells by centri-
fugation at 3000 g for 10 min at 4 C (see Note 7).
6. Resuspend the cells in ice-cold 40 mL of lysis buffer and lyse
the cells by sonication (10 s 6) on ice (see Note 8).
7. Centrifuge the lysed cells at 12,000 g for 20 min at 4 C and
collect the supernatant.
8. Equilibrate 500 μL (bed volume) of glutathione agarose by
washing twice with lysis buffer.
9. Add the equilibrated glutathione agarose to the cell lysate and
incubate for 1 h at 4 C with gentle rotation.
10. Wash the protein-bound glutathione agarose four times with
ice-cold lysis buffer.
11. Wash the protein-bound glutathione agarose once with
ice-cold cleavage buffer.
12. Add protease to ice-cold cleavage buffer, then add this to the
protein-bound glutathione agarose, and incubate at 4 C over-
night with rotation (see Note 9).
13. Centrifuge the solution at 500 g for 5 min at 4 C to pellet
the glutathione agarose and carefully transfer the supernatant
(eluted fraction) to a microcentrifuge tube.
14. Equilibrate a gel filtration column with the buffer (20 mM

Tris–HCl and 150 mM NaCl, pH 7.5, 1 mM PMSF, and
1 mM DTT) (see Notes 10 and 11).
15. Load the eluate onto the gel filtration column with an FPLC
system. Confirm the purity of the protein in the peak fractions
(the fraction volume could be, for example, 0.5–3 mL depend-
ing on various columns and proteins) by SDS-PAGE followed
by CBB staining.
16. Pool the pure proteins and concentrate with centrifugal protein
concentrators. The final concentration of proteins can be
100 μM in the buffer (20 mM Tris–HCl and 150 mM NaCl,
pH 7.5, 1 mM PMSF, and 1 mM DTT). Aliquot the proteins,
freeze in liquid nitrogen, and store them at 80 C.
3.2 Preparation of 1. In a fume hood, add each lipid from lipid stocks in organic
Multilamellar Vesicles solvent (e.g., PC, PE, PS, and PI(4,5)P2) including fluores-
(MLVs) cently labeled lipids into glass tubes using glass syringes accord-
ing to the desired lipid composition (see Note 12). For
example, the lipid composition is POPC:POPE:POPS:PI(4,5)
P2:rhodamine DHPE (50:19.5:20:10:0.5, mol/mol), and the
concentration is 1 mM.
2. Evaporate solvents such as chloroform and methanol under a
stream of nitrogen gas in the fume hood, and then remove the
remaining organic solvent with a vacuum concentrator for 2 h.
3. Add binding buffer to obtain a final lipid concentration of
1 mM, for example. Hydrate lipids for at least 1 h with vortex-
ing at room temperature (see Note 13) to generate MLVs.
3.3 Liposome Co- 1. Ultracentrifuge protein solutions at 436,000 g for 30 min at

sedimentation Assay 4 C using a pre-cooled fixed-angle centrifuge rotor to get rid
(See Fig. 1) of protein aggregates, and carefully collect the supernatant.
2. Measure the protein concentrations again with a microvolume
UV-Vis spectrophotometer.
3. Pipette binding buffer so that the total volume will be 50 μL
into the ultracentrifugation tubes. The buffer volume is 50 μL
total volume minus the volumes of protein and liposomes.
4. Pipette proteins into the ultracentrifugation tubes at a final
concentration of 1 μM, for example (see Notes 14 and 15).
5. Pipette liposomes into the ultracentrifugation tubes at a final
concentration of 0.5 mM, for example. A negative control
(only protein in a buffer without any liposomes) is important
to include in all assays.
6. Mix the solutions by pipetting gently to avoid disturbing the
liposomes and incubate for 30 min at room temperature.
Fig. 1 Liposome co-sedimentation assay to study protein–phosphoinositide

interactions. (a) Schematic representation of liposome co-sedimentation
assay. (b) (Upper) Images of CBB-stained SDS-PAGE gels after the liposome
co-sedimentation assay of Moesin FERM domain. S and P indicate the superna-
tant fraction, and the pellet fraction containing the protein-bound liposomes,
respectively. The lipid composition was POPC:POPE:POPS:PI(4,5)P2:rhodamine
DHPE (50:19.5:20:10:0.5, mol/mol). The concentration of Moesin FERM domain
was 1 μM, and that of liposomes was 1 mM. (Lower) Salt sensitivity of the
interaction of Moesin FERM domain with liposomes. The NaCl concentration was
varied at 0 mM, 100 mM, and 400 mM. The concentration of Moesin FERM
domain was 1 μM, and that of liposomes was 1 mM [12]
7. Place the tubes in the fixed-angle rotor, ensuring that the tubes
are balanced properly.
8. Ultracentrifuge at 436,000 g for 30 min at 20 C.
9. Immediately after the ultracentrifugation, carefully remove the
tubes from the rotor to avoid disturbing the pellets.
10. Transfer the supernatant to new microcentrifuge tubes (see
Note 16).
11. Resuspend the pellets in 50 μL of binding buffer and transfer to
new microcentrifuge tubes (see Notes 17 and 18).
12. Add 4 Laemmli sample buffer to the supernatant and pellets.
13. Incubate samples on a heat block at 95 C for 5 min.
14. Perform SDS-PAGE followed by CBB staining.
15. Image SDS-PAGE gels and quantify the intensity of each band
with a gel imaging analysis software (see Note 19).
16. The binding affinities of the proteins of interest are estimated
by calculating the band intensities using the equation pellet/
(supernatant + pellet) when the assay is performed by keeping
the protein concentration constant and varying the concentra-
tion of lipids.
3.4 Liposome 1. Ultracentrifuge protein solutions at 436,000 g for 30 min at

Co-flotation Assay 4 C using a pre-cooled fixed-angle rotor to get rid of protein
(See Fig. 2) aggregates.
2. Prepare MLVs according to the method described in Subhead-
ing 3.2 by hydrating with the HBS containing 0.3 M sucrose.
3. Add protein to 150 μL of MLV solution (e.g., 1 mM) in the
ultracentrifugation tubes at a final concentration of 1 μM, for
example. Gently mix and incubate for 10 min at room
temperature.
4. Add 100 μL of HBS containing 60% sucrose and gently mix
the solution. This will provide a sucrose concentration of 30%.
5. Carefully overlay (drop by drop against the wall of the tube)
200 μL of HBS containing 25% sucrose on top of the above
solution.
6. Carefully overlay (drop by drop against the wall of the tube)
200 μL of HBS on top of the above solution.
7. Centrifuge samples at 259,000 g for 30 min using a
swinging-bucket rotor.
Fig. 2 Liposome co-flotation assay to study protein–phosphoinositide interac-

tions. (a) Schematic representation of a liposome co-flotation assay. Sucrose
gradient was generated by ultracentrifugation to separate protein-bound lipo-
somes (fraction 2). During centrifugation, the membrane-associated proteins
float with liposomes in a sucrose gradient. (b) Each fraction from the co-flotation
assay of Moesin FERM domain was subjected to SDS-PAGE, followed by CBB
staining. Fraction 2 contained protein-bound liposomes, which was confirmed by
fluorescence spectroscopy [12]
8. Carefully collect 100 μL of each fraction from top to bottom

into microcentrifuge tubes.
9. Confirm the lipid fractions with the fluorescence spectrometer.
Take a few microliters (2 μL, for example) from each fraction
and dilute with HBS to a total volume of 100 μL (in cuvettes).
Set the excitation and emission wavelengths at 560 and
580 nm, respectively, and measure the rhodamine fluorescence
in each fraction.
10. Add 4 Laemmli sample buffer to each fraction.
11. Incubate samples on a heat block at 95 C for 5 min.
12. Carry out SDS-PAGE followed by CBB staining (see Notes
18).
13. Image SDS-PAGE gels and quantify the intensities of each
band with a gel imaging analysis software.
4 Notes
1. The glassware, but not plastic tubes, should be used to handle

the organic solvents such as methanol and chloroform.
2. Salt concentrations of the binding buffer can be varied to study
whether the membrane-binding modes of proteins depend on
the electrostatic interactions or not. When the membrane bind-
ing of the proteins depends on the electrostatic interactions,
the affinities of proteins to the membranes decrease with the
increase in salt concentration in the binding buffer. This will be
useful to study whether proteins bind to negatively charged
head groups of phosphoinositides. Special attention should be
paid to make sure that the protein of interest does not aggre-
gate (solution becomes cloudy) when the salt concentration in
the binding buffer decreases.
3. The pH of the binding buffer can be varied to examine whether
the membrane binding of the proteins is pH dependent that
can affect the subcellular localizations of membrane-binding
proteins.
4. His-tagged proteins can be used as well using Ni-NTA agarose
resin if the GST-tag is too large to express proteins of interest.
5. If the proteins of interest are difficult to express, other bacterial
strains such as Rosetta (DE3) can be used.
6. Optimize the IPTG concentration to improve protein
expressions.
7. The time and temperature of expression can be optimized.
8. If the proteins tend to be insoluble, optimize buffer conditions
such as salt concentrations and pH.
9. Proteases can remove GST-tags from fusion proteins.

10. Choose a gel filtration column that is appropriate for the
molecular weight of the proteins under investigation.
11. Alternatively, a cation or anion exchange column (depending
on the theoretical pI of the proteins of interest) can be used to
improve protein purity.
12. Various lipid compositions, including phosphatidylinositol
monophosphate [PI(3)P, PI(4)P, PI(5)P], diphosphate [PI
(3,4)P2, PI(3,5)P2, PI(4,5)P2], triphosphate [PI(3,4,5)P3],
and different phosphoinositide concentrations, can be used
by mimicking the lipid compositions of distinct cellular mem-
branes to elucidate the subcellular localizations of
proteins [12].
13. The temperature for lipid hydration should be higher than the
lipid phase transition temperature of the desired lipid compo-
sition, which can be measured, for example, by differential
scanning calorimetry.
14. Recombinant proteins purified from E. coli should be as pure as
possible, and the GST- or His-tags should be removed to avoid
nonspecific binding or interference in membrane binding of
proteins.
15. The protein concentrations can be varied; however, special
attention should be paid to avoid oversaturation of membrane
binding.
16. Special attention should be paid not to disturb or pipette the
pellets when collecting the supernatant. Gel-loading pipette
tips can be used.
17. Proteins tend to be absorbed onto tube surfaces. Low adsorp-
tion microtubes that reduce the adsorption of proteins can
be used.
18. The addition of fluorescently labeled lipids such as 0.5% (mol/-
mol) rhodamine DHPE allows the visualization of liposome
fractions in the tubes.
19. This can be done with western blotting if affinity/recovery is
weaker and if antibodies are available.
Acknowledgments
This work was supported by grants from the Academy of Finland

(H.Z. and P.L.), Jane and Aatos Erkko Foundation (H.Z.),
Guangxi distinguished expert funding (H.Z.), FY 2015 Researcher
Exchange Program between JSPS and AF (Y.S.), Astellas Founda-
tion for Research on Metabolic Disorders (Y.S.), The Scandinavia-
Japan Sasakawa Foundation (Y.S.), The Ichiro Kanehara Founda-
tion for the Promotion of Medical Sciences and Medical Care (Y.
S.), The Association for Fordays Self-Reliance Support in Japan (Y.

S.), The Futaba Research Grant Program of the Futaba Foundation
(Y.S.), The NOVARTIS Foundation (Japan) for the Promotion of
Science (Y.S.), Okayama Foundation for Science and Technology
(Y.S.), Wesco Scientific Promotion Foundation (Y.S.), and JSPS
KAKENHI Grant Numbers JP19K23727, JP20K06589 (Y.S.).
References
1. Roth MG (2004) Phosphoinositides in consti- act as a sensor of PIP(2) density. Mol Cell
tutive membrane traffic. Physiol Rev 17:181–191
84:699–730 8. Zhao H, Michelot A, Koskela EV et al (2013)
2. Picas L, Gaits-Iacovoni F, Goud B (2016) The Membrane-sculpting BAR domains generate
emerging role of phosphoinositide clustering stable lipid microdomains. Cell Rep
in intracellular trafficking and signal transduc- 4:1213–1223
tion. F1000Res 5 9. Prévost C, Zhao H, Manzi J et al (2015)
3. Saarikangas J, Zhao H, Pyk€al€ainen A et al IRSp53 senses negative membrane curvature
(2009) Molecular mechanisms of membrane and phase separates along membrane tubules.
deformation by I-BAR domain proteins. Curr Nat Commun 6:8529
Biol 19:95–107 10. Simunovic M, Manneville J-B, Renard H-F
4. Senju Y, Lappalainen P (2019) Regulation of et al (2017) Friction mediates scission of tubu-
actin dynamics by PI(4,5)P2 in cell migration lar membranes Scaffolded by BAR proteins.
and endocytosis. Curr Opin Cell Biol 56:7–13 Cell 170:172–184.e11
5. Rohatgi R, Ho HY, Kirschner MW (2000) 11. Pyk€al€ainen A, Boczkowska M, Zhao H et al
Mechanism of N-WASP activation by CDC42 (2011) Pinkbar is an epithelial-specific BAR
and phosphatidylinositol 4, 5-bisphosphate. J domain protein that generates planar mem-
Cell Biol 150:1299–1310 brane structures. Nat Struct Mol Biol
6. Hamada K, Shimizu T, Matsui T et al (2000) 18:902–907
Structural basis of the membrane-targeting and 12. Senju Y, Kalimeri M, Koskela EV et al (2017)
unmasking mechanisms of the radixin FERM Mechanistic principles underlying regulation of
domain. EMBO J 19:4449–4462 the actin cytoskeleton by phosphoinositides.
7. Papayannopoulos V, Co C, Prehoda KE et al Proc Natl Acad Sci U S A 114:E8977–E8986
(2005) A polybasic motif allows N-WASP to
Chapter 15
Characterization of PROPPIN–Phosphoinositide Binding

by Stopped-Flow Fluorescence Spectroscopy
Ángel Pérez-Lara and Reinhard Jahn
Abstract
PROPPINs (β-propellers that bind polyphosphoinositides) are a protein family that binds preferentially
phosphatidylinositol 3-phosphate (PtdIns(3)P) and phosphatidylinositol 3,5-bisphosphate (PtdIns(3,5)
P2) via its FRRG motif. PROPPINs are involved in autophagic functions, but their molecular mechanism is
still elusive. To unravel the molecular mechanism of PROPPINs, it is essential to understand the PROP-
PIN–phosphoinositide binding. Here, we describe a protocol to study the kinetics of the PROPPIN–pho-
sphoinositide binding using a fluorescence resonance energy transfer (FRET) stopped-flow approach. We
use FRET between fluorophore-labeled protein and fluorophore-labeled liposomes, monitoring the
increase of the acceptor emission in labeled liposomes after the protein–membrane binding. Through this
approach, we studied the kinetics of the PROPPIN Atg18 (Autophagy-related protein 18) from Pichia
angusta (PaAtg18) and a mutant of its FRRG motif, called FTTG mutant. Stopped-flow experiments
demonstrated that the main function of the FRRG motif is to retain, instead of to drive, Atg18 to the
membrane, decreasing the Atg18 dissociation rate. Furthermore, this method is suitable for the study of
other PI-binding proteins.
Key words Protein–phosphoinositide binding, Kinetics, Stopped-flow, PROPPIN, Liposome
1 Introduction
Phosphoinositides (PIs) are phospholipids derived from phospha-

tidylinositol (PtdIns) involved in the eukaryotic cell signaling as
lipid regulators (for a review, see [1]). The inositol ring of PtdIns
can be reversibly phosphorylated on positions 3, 4, and 5, generat-
ing seven different polyphophoinositides (PPIs) that work as a “PI
code”. This “PI code” regulates protein binding or/and activation,
determining their localization within the cell and activity in time
and space [2, 3]. A major breakthrough in understanding this “PI
code” was the identification of protein domains that bind specific
PPIs in a wide range of affinity and selectivity [3, 4].
PROPPINs (β-propellers that bind polyphosphoinositides)
represent a family of proteins involved in autophagosome
205
206 Ángel Pérez-Lara and Reinhard Jahn
biogenesis [5] that fold as seven-bladed β-propellers. They bind

specifically to phosphatidylinositol 3-phosphate (PtdIns(3)P) and
phosphatidylinositol 3,5-bisphosphate (PtdIns(3,5)P2) via a con-
served FRRG motif [6, 7], which defines two PI-binding sites
[8, 9]. They are conserved from yeast to humans. In yeast, there
are three highly homologous PROPPINs—Atg18, Atg21, and
Hsv2 (homologous with swollen vacuole phenotype 2) with differ-
ent autophagic subtype specificities [5, 9]. Atg18 functions in both
macroautophagy and the selective Cvt-pathway, while the function
of Atg21 is restricted to the Cvt-pathway. In contrast, Hsv2 is
involved in micronucleophagy [10]. Atg18 forms part of the
Atg2–Atg18 complex and Atg21 recruits the Atg12–Atg5–Atg16
complex to the phagophore assembly site (PAS) [5], but little is
known about the function of Hsv2. Although the crystal structures
of Hsv2 [9, 8], Atg18 [11], and Atg21 [12] have been solved, the
membrane-binding mechanism of these PROPPINs remains
unclear.
Here, we describe a novel method for determining the Atg18–
membrane binding kinetics using rapid stopped-flow fluorescence
spectroscopy. Protein–membrane binding is monitored by the
increase of the acceptor emission fluorescence after the binding
between proteins and liposomes. Upon binding, FRET between
acceptor dye-labeled liposomes and donor dye-labeled proteins
take place, decreasing donor emission fluorescence and increasing
acceptor emission fluorescence. This method has been widely used
to study the kinetics of PI-binding proteins such as protein kinase C
(PKC) [13, 14] and synaptotagmin-1 (syt-1) [15–17], among
many others. Using this method, a mechanistic view of the pro-
tein–membrane binding can be obtained that cannot be achieved
by other methods. This technique measures the rate of the binding
of two reactants, protein and liposomes in this case, which are
rapidly mixed in a stopped-flow fluorescence spectrophotometer.
In this protocol, we use the PROPPIN Atg18 from Pichia angusta
(PaAtg18) to study the PROPPIN–phosphoinositide binding.
Additionally, we designed different single cysteine mutants for
labeling with different fluorescence dyes to monitor the kinetics
of the PaAtg18–membrane binding by recording the increase of
fluorescence during these processes.
2 Materials
1. A stopped-flow fluorescence spectrophotometer with a cham-

ber linked to a thermostat is required. Our experiments are
carried out using the Applied Photophysics Π* 180, Applied
Photophysics SX20MV, and Applied Photophysics SX18MV
stopped-flow spectrophotometers.
Stopped-Flow Fluorescence Spectroscopy 207
2. Stock solutions of membrane lipids (see Note 1): Stocks of

10 mg/mL DOPC (1,2-dioleoyl-sn-glycero-3-phosphocho-
line) and DOPE (1,2-dioleoyl-sn-glycero-3-phosphoethanola-
mine) in chloroform. Stocks of 1 mg/mL PtdIns(3,5)P2
(1,2-dioleoyl-sn-glycero-3-phospho-(10 -myo-inositol-30 ,5-
0
-bisphosphate) and TR-PE (TexasRed 1,2-dihexadecanoyl-sn-
glycero-3-phosphoethanolamine) in chloroform. Then, store
lipid stocks in amber glass vials at 20 C.
3. 2 M KCl (potassium chloride): Dissolve 74.55 g in 0.5 L of
ultrapure water. Prepare all solutions using ultrapure water
(18 MΩ-cm at 25 C) and calibrated flasks.
4. 1 M HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic
acid): Dissolve 119.15 g in 450 mL of ultrapure water and
adjust pH to 7.4. Then, dilute to 500 mL and check pH again.
5. 0.5 M TCEP (Tris(2-carboxyethyl)phosphine hydrochloride):
Dissolve 1.44 g in 10 mL of ultrapure water.
6. 0.4 mM phosphate standard solution: Dissolve 54.4 mg
KH2PO4 in 1 L ultrapure water.
7. Perchloric acid solution: 70% HClO4 (perchloric acid).
8. Molybdate reagent: Prepare 1 L of molybdate reagent solution
adding 2.2 g of (NH4)6Mo7O24·4H2O and 14.3 mL of con-
centrated H2SO4. Add concentrated H2SO4 to the aqueous
solution with caution and then add ultrapure water until 1 L.
9. Ascorbic acid solution: Prepare a 10% (w/v) solution of
ascorbic acid.
10. Purified single cysteine mutants of the PaAtg18 protein.
Expression and purification of these mutant proteins were
previously described in detail [11].
11. Alexa Fluor aliquots: Prepare an Alexa Fluor 488 C5 Malei-
mide (xanthylium, 3,6-diamino-9-[2-carboxy-4(or5)-
[[[5-(2,5-dihydro-2,5-dioxo-1H-pyrrol-1-yl)pentyl]amino]
carbonyl]phenyl]-4,5-disulfo-, inner salt, monosodium salt)
stock of 10 mg/mL in DMSO (dimethyl sulfoxide), for specific
labeling of the cysteine mutants. Divide the resuspended dye
into suitable aliquots, followed by lyophilization and storage at
80 C.
12. Desalting columns: Columns containing Sephadex G-25 resin
to remove free dye from labeled proteins.
13. A mini-extruder, 10-mm filter supports, and 0.1-μm polycar-
bonate membranes for the preparation of liposomes.
14. Assay buffer: 150 mM KCl, 20 mM HEPES, pH 7.4, and
0.1 mM TCEP. Adjust the pH to 7.4 with NaOH.
15. A visible light spectrophotometer.
16. Phosphate-free glass tubes: Glass tubes have to be washed

extensively using ultrapure water in order to remove any trace
of phosphate from detergent.
3 Methods
3.1 Preparation Large unilamellar vesicles (LUV) composed of DOPC/DOPE/

of Liposomes TR-PE/PtdIns(3,5)P2 (79:18:2:1, molar ratio) were used for
studying the kinetics of PaAtg18–phosphoinositide binding.
1. Mix chloroform lipid stocks at the desired proportions in
phosphate-free glass tubes followed by drying under a stream
of oxygen-free nitrogen. Then, remove remaining traces of
chloroform in a vacuum desiccator for at least 3 h (see Note 2).
2. Resuspend the dried lipid film in assay buffer by vigorous
vortexing until the lipid film is completely removed (a couple
of minutes should be enough) from the walls of the glass tube.
3. Prepare LUVs of around 140 nm diameter by extrusion of the
phospholipid suspensions through a 0.1-μm polycarbonate
membrane using a mini-extruder according to the manufac-
turer’s instructions.
3.2 Quantification Quantify the phospholipid-bound phosphate of the prepared LUVs

of Phospholipid-Bound in order to determine the phospholipid concentration of them (see
Phosphate Note 3).
1. Prepare phosphate standards ranging 20–160 nmol using the
phosphate standard solution in phosphate-free glass tubes (see
Table 1).
2. Prepare sample tubes adding an estimated amount of the LUV
suspension ranging 20–160 nmol of phospholipids in
phosphate-free glass tubes (see Note 4).
3. Add 0.4 mL of perchloric acid solution to each tube, vortex
briefly to mix, and then transfer to a heating block, followed by
incubation at 180 C for at least 1 h.
Table 1
Volume of 0.4 mM phosphate standard solution used, and corresponding phosphate total amount, to
prepare the phosphate standards
Tube number 1 2 3 4 5 6 7
Volume of phosphate standard solution (μL) 0 50 100 150 200 250 400
nmoles of phosphate 0 20 40 60 80 100 160
4. Remove the tubes from the heating block and let them cool to
room temperature (RT). To each tube, add 4 mL of molybdate
reagent and 0.5 mL of freshly prepared ascorbic acid solution.
Mix vigorously and then incubate in a water bath with boiling
water for 5–10 min.
5. After cooling to RT, measure absorbance at 812 nm using a
visible light spectrophotometer.
6. Use the phosphate standards to estimate the phosphate con-
centration of the liposome samples, which corresponds to the
phospholipid concentration in the liposome sample.
3.3 Protein Labeling 1. Resuspend Alexa Fluor aliquots in DMSO to 10 mg/mL (see
Note 5).
2. Incubate the single Cys mutants (50–100 μM) with a fivefold
molar excess of the dye while gently shaking overnight at 4 C
in a cold room (see Note 6).
3. Remove unreacted free dye from the labeled protein by size
exclusion chromatography using a desalting column according
to the manufacturer’s instruction. After collecting the labeled
protein, determine the protein concentration (see Note 7).
3.4 PROPPIN– 1. Clean the stopped-flow apparatus thoroughly before using

Liposome detergent-containing solutions (2% Hellmanex® III or 10%
Binding Assay SDS are suitable detergent solutions for this washing step).
Then, flush the plumbing lines extensively with water (over
100 mL) and finally with assay buffer (over 20 mL).
2. Adjust the concentration of Alexa Fluor 488-labeled Atg18 to
0.5 μM in assay buffer.
3. Prepare a liposome solution in assay buffer. The liposome
solution should contain the target phosphoinositide at least
ten-fold higher concentration than the labeled protein in
order to ensure pseudo-first-order conditions.
4. Load the drive syringes of the stopped-flow apparatus with the
protein solution and the liposome solution, respectively. Allow
at 5 min for temperature equilibration before triggering rapid
mixing.
5. Set the excitation wavelength for Alexa Fluor 488-labeled
Atg18 to 495 nm and set up the fluorescence detector module
with a 610-nm cutoff filter to monitor the emission fluores-
cence of the TR-PE. To optimize instrument settings, certain
guidelines may need to be taken into account (see Note 8).
6. Then, prime the instrument driving the system two or three
times.
7. Trigger and record the time course of the reaction after rapid
mixing. Repeat the experiment three to five times.
8. Repeat from steps 3–6 using samples with different total lipid
concentrations ranging 100–1000 μM in a syringe (see
Note 9).
3.5 Data Analysis 1. Fit the fluorescence traces (Fig. 1a, b) with an exponential
equation as follows (Eq. 1):
X
n
F ðt Þ ¼ F 0 þ A obsðiÞ ekobsðiÞt ð1Þ
i¼1
where F(t) is the observed fluorescence at time t, F0 is the

final fluorescence, Aobs(i) represents the amplitude of the fluo-
rescence traces, and kobs(i) is the observed rate constant for the
ith of n phases (see Note 10).
2. Test for the accuracy of the data fitting by analyzing the residual
plots of the fitting (see Note 11). In this case, stopped-flow
traces are fitted using a three-exponential equation, yielding up
to three kobs. It is noteworthy that kobs2 and kobs3 are discarded
for data analysis due to vesicle aggregation during the time
course of the measurements [11].
3. Then, plot the observed rate constants as a function of the
accessible lipid concentration (Fig. 1c), assuming that only
60% of the total lipid concentration was accessible to the pro-
tein (outer vs. inner surface of the liposomes), and fit it with the
linear equation as follows (Eq. 2):
kobs ¼ kon ½total accesible lipid þ koff ð2Þ
where kon (the slope) is the second-order association rate
constant (in M1 s1) and koff (the y-intercept) is the apparent
first-order rate constant (in s1).
Fig. 1 Characterization of PaAtg18–liposome binding by stopped-flow fluorescence resonance energy transfer

(FRET). Fluorescence time courses of Texas Red emission (a and b) at different accessible lipid concentrations
are shown for S448C Atg18 (a) and S448C FTTG Atg18 (b). (c) Plots showing kobs1 as a function of the total
accessible lipid concentration. The solid and dashed lines show the linear fit using Eq. (2). The y-intercept
provides the dissociation rate constant koff, and the slope yields the association rate constant kon. (Modified
from [11] and under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/
licenses/by/4.0/))
4 Notes
1. Lipid stocks in chloroform need to be kept on ice at all times in

order to avoid solvent evaporation. To ensure accurate pipet-
ting, we use glass syringes.
2. In our case, to get a final volume of 5 mL at 1 mM total lipid,
we mixed 310.5 μL DOPC (10 mg/mL stock), 67 μL DOPE
(10 mg/mL stock), 310.5 μL PtdIns(3,5)P2 (1 mg/mL
stock), and 138.1 μL TR-PE (1 mg/mL stock). Nevertheless,
these volumes can change depending on the desired molar ratio
of the different lipids used and the desired final volume and
concentration of the needed liposomes.
3. Lipid concentrations are monitored indirectly using a phos-
phate determination method [18] (see Subheading 3.2). Note
that it is not sufficient to estimate the phospholipid concentra-
tions in the final liposome fraction solely by extrapolating from
the amount of material used as input. There are significant
losses during the liposome preparation, and thus it is essential.
We must take into account that while DOPC and DOPE only
have a phosphate group per phospholipid, PtdIns(3,5)P2 has
three phosphate groups per phospholipid.
4. Taking into account that we aimed to get a final total lipid
concentration of 1 mM, in order to get a total phosphate
amount into the phosphate standard range (20–160 nmol),
we have to add to the sample tubes between 20 and 160 μL
of LUV suspension.
5. Although our assay was optimized for Alexa Fluor 488, other
green-fluorophore dyes have been used successfully as IANBD
(N,N0 -dimethyl-N-(Iodoacetyl)-N0 -(7-nitrobenz-2-oxa-1,3-
diazol-4-yl)ethylenediamine) and Oregon Green (20 ,7-
0
-dichlorofluorescein maleimide).
6. Avoid light exposure during the incubation in order to reduce
the photobleaching of the dye.
7. Different methods for protein concentration determination
can be used. The simplest one is to measure the absorbance at
280 nm, but other colorimetric and fluorescent methods have
been developed. In order to select one particular method
among the different available ones, we have to consider the
possible interferences of the dye with the method of protein
concentration determination.
8. Although this assay monitors the FRET between Alexa Fluor
488 and Texas Red fluorophores, it is possible to use different
FRET pairs. As mentioned above, we have used other “green”
dyes (Oregon Green and NBD) instead of Alexa Fluor 488.
Additionally, TR-PE can be substituted by Rhod-PE
(1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissa-
mine rhodamine B sulfonyl) as acceptor-labeled lipid in lipo-
somes. Furthermore, similar assays have been carried out in our
lab recording the FRET between Trp residues of the unlabeled
protein and Dansyl-PE (1,2-dioleoyl-sn-glycero-3-phos-
phoethanolamine-N-(5-dimethylamino-1-
naphthalenesulfonyl)) or Marina Blue-PE (Marina Blue
1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine).
Thus, depending on the FRET pair used, different experimen-
tal settings have to be optimized as excitation wavelength for
the donor dye and the cutoff filter used to monitor the emis-
sion fluorescence of the acceptor-labeled lipid.
Additionally, voltage of the photomultiplier detector must
be optimized to ensure the dynamic range of the signal and to
achieve a suitable signal-to-noise ratio. The slit width on the
monochromator for excitation has to be adjusted to get
enough light detection without photobleaching.
9. The different liposome solutions are prepared by dilution of
the first liposome solution used. After triggering the rapid
mixing and acquiring the time courses, liposomes are recovered
from the drive syringe and diluted, ensuring pseudo-first-order
conditions.
10. We used an OriginLab software (Northampton,
Massachusetts, USA) for the curve fitting, but any other soft-
ware could be used. Select the nonlinear curve fit and user-
defined tabs. Write the appropriate script for the exponential
equation needed. For instance:
y ¼ y0 + A1*exp(k1*) for a monophasic exponential,

y ¼ y0 + A1*exp(k1*) + A2*exp(k2*) for a biphasic
exponential,
y ¼ y0 + A1*exp(k1*) + A2*exp(k2*) + A3*exp(k3*)
for a triphasic exponential.
y is the dependent variable corresponding to the emission

fluorescence (F(t)); x is the independent variable
corresponding to the time (t); and y0, Ai, and ki are the differ-
ent parameters calculated by the fitting for the i phase of the
different exponential equations. y0 is the final fluorescence,
A represents the amplitude of the fluorescence traces, and k is
the observed rate constant.
11. Improvement in curve fitting by the addition of a second or

third exponential term must be interpreted with caution
because the inclusion of additional phases always improves
the fitting to a given equation. Thus, significant improved
fitting to mono-, bi-, or tri-exponential suggests a one-, two-,
a b c
0.8 0.8 0.8
monophasic biphasic triphasic
Fluorescence (A.U.)
Fluorescence (A.U.)
Fluorescence (A.U.)
0.6 0.6 0.6
0.4 0.4 0.4
0.2 0.2 0.2
0.0 0.0 0.0
0.01 0.1 1 10 100 0.01 0.1 1 10 100 0.01 0.1 1 10 100

time (s) time (s) time (s)
c d e
0.08 0.08 0.08
0.04 0.04 0.04

Residual
Residual
Residual
0.00 0.00 0.00
-0.04 -0.04 -0.04
-0.08 -0.08 -0.08
0.01 0.1 1 10 100 0.01 0.1 1 10 100 0.01 0.1 1 10 100

time (s) time (s) time (s)
Fig. 2 Examples of curve fittings to a monophasic (a, d), biphasic (b, e), or triphasic (c, f) exponential. Solid red
lines show the monophasic (a), biphasic (b), and triphasic (c) exponential fits for the same fluorescence time
course (black). Representative residual plots from monophasic (d), biphasic (e), and triphasic (f) exponential
fits of the fluorescence time course used in panels a, b, and c. Note that better goodness of the fit using the
tri-exponential equation does not necessarily mean three different binding processes. Fitted parameters have
to be reasonable (e.g., numbers of phases and rate constants must be consistent with the time scale of data
collection) (see Note 11)
or three-step process, respectively, but it is not necessarily true.

For instance, protein–membrane binding processes are nor-
mally complete within a few seconds. Thus, fluorescence time
courses involving over tens of seconds to reach the plateau
(Fig. 2) suggest liposome aggregation. Aggregation leads to
slow observed constant rates, which are discarded for the data
analysis. Furthermore, previous knowledge of molecular mech-
anism of the protein–membrane binding would be very useful
in order to interpret the results correctly and to deduce a
proper membrane-binding model (see [19] for a more detailed
guideline).
Acknowledgments
We are indebted to Dr. Karin Kühnel for her generous support and
advice. We thank Dr. Bruno Ramos-Molina and Dr. Manuel Rosa-
Garrido for advice and discussions. This work was supported by a
NIH Grant GB10440.155740 and Max Planck Gesellschaft finan-
cial support (Grant PSBICH11200) to R.J.
References
1. Balla T (2013) Phosphoinositides: tiny lipids of Atg18, Atg21 and Ygr223c. Autophagy
with Giant impact on cell regulation. Physiol 4:896–910
Rev 93:1019–1137 11. Scacioc A, Schmidt C, Hofmann T, Urlaub H,
2. Dickson EJ, Hille B (2019) Understanding Kuhnel K, Perez-Lara A (2017) Structure
phosphoinositides: rare, dynamic, and essential based biophysical characterization of the
membrane phospholipids. Biochem J PROPPIN Atg18 shows Atg18 oligomeriza-
476:1–23 tion upon membrane binding. Sci Rep
3. Kutateladze TG (2012) Molecular analysis of 7:14008
protein-phosphoinositide interactions. Curr 12. Metje J (2017) Structural characterization of
Top Microbiol 362:111–126 autophagy related protein complexes. Univer-
4. Hammond GRV, Balla T (2015) Polypho- sity of Göttingen - Georg-August-Universit€at
sphoinositide binding domains: key to inositol Göttingen, Göttingen
lipid biology. BBA-Mol Cell Biol L 13. Corbin JA, Evans JH, Landgraf KE, Falke JJ
1851:1283–1283 (2007) Mechanism of specific membrane tar-
5. Farre JC, Subramani S (2016) Mechanistic geting by C2 domains: localized pools of target
insights into selective autophagy pathways: les- lipids enhance Ca2+ affinity. Biochemistry
sons from yeast. Nat Rev Mol Cell Bio 46:4322–4336
17:537–552 14. Nalefski EA, Newton AC (2001) Membrane
6. Busse RA, Scacioc A, Krick R, Perez-Lara A, binding kinetics of protein kinase C beta II
Thumm M, Kuhnel K (2015) Characterization mediated by the C2 domain. Biochemistry
of PROPPIN-phosphoinositide binding and 40:13216–13229
role of loop 6CD in PROPPIN-membrane 15. Perez-Lara A, Thapa A, Nyenhuis SB, Nyen-
binding. Biophys J 108:2223–2234 huis DA, Halder P, Tietzel M, Tittmann K,
7. Krick R, Tolstrup J, Appelles A, Henke S, Cafiso DS, Jahn R (2016) PtdInsP(2) and
Thumm M (2006) The relevance of the PtdSer cooperate to trap synaptotagmin-1 to
phosphatidylinositolphosphat-binding motif the plasma membrane in the presence of cal-
FRRGT of Atg18 and Atg21 for the Cvt path- cium. Elife 5:e15886
way and autophagy. FEBS Lett 16. Bai JH, Tucker WC, Chapman ER (2004)
580:4632–4638 PIP2 increases the speed of response of synap-
8. Baskaran S, Ragusa MJ, Boura E, Hurley JH totagmin and steers its membrane-penetration
(2012) Two-site recognition of phosphatidyli- activity toward the plasma membrane. Nat
nositol 3-phosphate by PROPPINs in autop- Struct Mol Biol 11:36–44
hagy. Mol Cell 47:339–348 17. Bai JH, Wang P, Chapman ER (2002) C2A
9. Krick R, Busse RA, Scacioc A, Stephan M, activates a cryptic Ca2+triggered membrane
Janshoff A, Thumm M, Kuhnel K (2012) penetration activity within the C2B domain of
Structural and functional characterization of synaptotagmin I. Proc Natl Acad Sci U S A
the two phosphoinositide binding sites of 99:1665–1670
PROPPINs, a beta-propeller protein family. 18. Boettcher C, Pries C, Vangent CM (1961) A
Proc Natl Acad Sci U S A 109:E2042–E2049 rapid and sensitive sub-micro phosphorus
10. Krick R, Henke S, Tolstrup J, Thumm M determination. Anal Chim Acta 24:203–204
(2008) Dissecting the localization and function 19. Bernasconi CF (1976) Relaxation kinetics.
Academic Press, New York
Chapter 16
Fluorescence Assays to Study Membrane Penetration

of Proteins
Yosuke Senju and Hongxia Zhao
Abstract
Phosphoinositides play important roles in the regulation of protein recruitment at specialized membrane
domains, protein activity, and membrane dynamics. Phosphoinositide–protein interplay occurs via multiple
mechanisms and proteins associate with membranes through different binding patterns. Determinations of
membrane-binding mode and membrane penetration depth of proteins in lipid bilayer are thus important
steps in characterizing the molecular mechanisms of membrane–protein interactions. Here, we show two
standard in vitro assays using liposomes, diphenylhexatriene (DPH) anisotropy, and fluorescence quench-
ing by brominated lipids to determine membrane penetration of proteins into lipid bilayer. These methods
will provide useful tools to study membrane–protein association and uncover molecular details of protein–-
lipid interplay, which are important for understanding biological functions of membrane-associated pro-
teins and membrane dynamics.
Key words Liposome, Lipid–protein interactions, Membrane insertion, Membrane penetration,

Anisotropy, Phosphoinositides
1 Introduction
The plasma and intracellular membranes have important lipid reg-

ulators, such as phosphoinositides, which recruit specific peripheral
membrane-binding proteins to the subcellular compartments for
many essential cellular processes [1–4]. Phosphoinositide–protein
interactions play critical roles in membrane homeostasis and cellular
signaling [3, 4]. Association of proteins with lipids can lead to
changes in membrane physicochemical properties such as fluidity,
tension, curvature, and reorganization of membrane components.
For example, some of the Bin/Amphiphysin/Rvs (BAR) domain
proteins can generate membrane curvatures by penetrating an
amphipathic helix into the lipid bilayer and induce formation of
stable phosphoinositide–protein microdomains by lipid clustering
[5–8]. On the other hand, arrangement of proteins in membranes
may affect their conformation, biological activity, thermodynamic
215
216 Yosuke Senju and Hongxia Zhao
stability, and binding of surrounding macromolecules. Determina-

tion of precise spatial positions of proteins in lipid bilayer is critical
for understanding the underlying mechanisms of protein–mem-
brane interactions. However, biological membranes have complex
and varied compositions of lipids and proteins, which make in vivo
study of the protein–membrane interactions very challenging.
Thus, we introduce two in vitro assays using unilamellar vesicles
to study membrane-binding patterns of peripheral membrane pro-
teins. One is 1,6-diphenyl-1,3,5-hexatriene (DPH) anisotropy
assay, which is a simple method to characterize membrane insertion
of proteins in lipid bilayer [9]. DPH is a probe extensively used to
determine the structural and dynamical features of lipid bilayers. In
particular, DPH fluorescence anisotropy is a parameter interpreted
as membrane microviscosity or fluidity that is sensitive to the order
of lipid acyl chains. Penetration of membrane-bound proteins into
lipid bilayers generally decreases the acyl chain order and the rota-
tional motion of DPH molecules, resulting in a rise in DPH anisot-
ropy. The other method is fluorescence quenching by
phospholipids labeled with the fluorescence quencher bromine
atoms at different positions of the lipid acyl chains. Depth-
dependent quenching is applied to estimate membrane penetration
depth of intrinsic tryptophan or labeled fluorophore of membrane-
bound proteins in lipid bilayer [6, 10–12]. By using these probes,
one can characterize membrane orientation, topology, and folding
of a membrane protein upon membrane binding. Taken together,
these in vitro fluorescence assays using liposomes as model mem-
branes will provide new insights into fundamental principles and
molecular details of lipid–protein associations.
2 Materials
2.1 Phospholipids 1. 1-Palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC) is

dissolved in chloroform at 4.39 mM. 1-Palmitoyl-2-oleoyl-sn-
glycero-3-phosphoethanolamine (POPE) is dissolved in chlo-
roform at 4.64 mM. 1-Palmitoyl-2-oleoyl-sn-glycero-3-phos-
pho-L-serine (POPS) is dissolved in cholorofrom at 4 mM. PI
(4,5)P2 is dissolved in chloroform:methanol:water (20:9:1,
v/v) at 0.57 mM. 1-Palmitoyl-2-(6,7-dibromo)stearoyl-sn-
glycero-3-phosphocholine, 1-palmitoyl-2-(9,10-dibromo)
stearoyl-sn-glycero-3-phosphocholine, and 1-palmitoyl-2-
(11,12-dibromo)stearoyl-sn-glycero-3-phosphocholine are
dissolved in chloroform at 3.6 mM. Diphenylhexatriene
(DPH) is dissolved in dimethylformamide at 0.87 mM. The
phospholipids should be purged with an inert gas to prevent
oxidation and stored in glass tubes at 20 C.
Analysis of Membrane Penetration of Proteins 217
2.2 Stock Solutions HEPES-buffered saline (HBS): 20 mM HEPES, pH 7.5,

100 mM NaCl.
2.3 Reagents 1. Methanol.

and Equipment 2. Chloroform.
3. Glass tubes.
4. Hamilton syringe.
5. Nitrogen gas.
6. Pressured gas blowing concentrator.
7. Vacuum concentrator.
8. Vortex mixer.
9. Mini-extruder (Avanti Polar Lipids).
10. Filter supports 10 mm (Avanti Polar Lipids).
11. Polycarbonate membranes, 0.1 μm (Avanti Polar Lipids).
12. Gas-tight syringe, 250 μL (Avanti Polar Lipids).
13. Fluorescence spectrometer equipped with automated excita-
tion and emission polarizer (horizontal and vertical), and a
temperature control system.
14. Quartz glass cuvettes.
3 Methods
3.1 DPH Anisotropy 1. To prepare multilamellar vesicles (MLVs), add each lipid from
Assay (See Fig. 1) lipid stocks dissolved in organic solvent (e.g., PC, PE, PS, and
PI(4,5)P2) and 0.002 (mol/mol) DPH into glass tubes with
Hamilton syringes in a fume hood according to the desired
lipid composition of study. The concentration of total lipids
could be 1 mM, for example (see Notes 1 and 2).
2. Evaporate the organic solvents such as chloroform and metha-
nol under a stream of nitrogen gas in a fume hood. Dry the
remaining organic solvents by a vacuum concentrator from 3 h
to overnight.
3. To fit the volume of syringes for mini-extruder, add 0.2 to
1 mL of HBS to obtain a final total lipid concentration of
1 mM, for example, and hydrate the lipids for at least 1 h
with vortexing/shaking at room temperature (see Note 3) to
generate MLVs.
4. To prepare large unilamellar vesicles (LUVs), assemble the
mini-extruder according to the manufacturer’s instructions
and extrude the MLVs through a polycarbonate filter
(100-nm pore size) using the mini-extruder (see Notes 4–6).
Fig. 1 DPH anisotropy assay for detecting penetration of proteins into the
hydrophobic core of lipid bilayers. (a) A linearly polarized excitation light
generated by a vertical polarizer excites DPH molecules efficiently aligned
parallel to the acyl chains in the lipid bilayers. Fluorescence intensities of DPH
are measured with both vertical and parallel polarizers. Using these parameters,
the fluorescence anisotropy (r) can be calculated by the equation: r ¼ (IH IV)/
(IH + 2IV), where the direction of fluorescence polarization is both parallel (IH) and
perpendicular (IV) to that of the excitation beam. When DPH molecules exhibit
isotropic motion (same values when measured in different directions) in a
membrane with fluid phase, it leads to fluorescence depolarization, and as a
result, fluorescence anisotropy shows low values. However, when membrane-
binding proteins penetrate into lipid bilayers, the immersed proteins constrain
the motion of DPH molecules. Thus, the polarized light excites DPH molecules
aligned to the acyl chains more efficiently, resulting in an increase in
fluorescence anisotropy. (b) DPH anisotropy increased in the presence of the
I-BAR domain of missing in metastasis protein (MIM) in a dose-dependent
5. Mix 4 μL LUVs with 96 μL of HBS in a quartz glass cuvette,

providing a final lipid concentration of 40 μM.
6. Measure DPH anisotropy using a fluorescence spectrometer.
Set the excitation wavelength at 360 nm and the emission
wavelength at 450 nm. All measurements are performed at
desired temperature using temperature-controlled cuvette
holders and a circulating water bath (see Note 7).
7. Add proteins (the optimal protein volume is between 0.2 and
1 μL) into the lipid solutions in a step-wise manner (e.g., 0.5,
1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, and 5 μM concentrations), and
measure the DPH anisotropy with a fluorescence spectrometer
(see Notes 8 and 9).
8. Repeat the measurements three times using the same liposome
sample and average the values of DPH anisotropy.
9. Prepare new liposomes and repeat the experiments three times.
Calculate the average of DPH anisotropy and standard
deviation.
3.2 Fluorescence 1. Prepare large unilamellar vesicles as described above.

Quenching by Brominated-PC lipids (for example, (6,7)-, (9,10)-, and
Brominated Lipids (11,12)-Br2-PC) are added to 30% of total lipid concentration
(See Fig. 2) while maintaining the same molar ratio of PC with other lipids
(see Note 10).
2. Prepare protein solution in HBS (see Note 11).
3. Measure tryptophan fluorescence with a fluorescence spec-
trometer. The tryptophan residue is excited at 280 nm, and
emission spectra are recorded from 300 to 450 nm, averaging
three scans from the same sample. All measurements are per-
formed at desired temperature using temperature-controlled
cuvette holders and a circulating water bath.
4. Titrate liposomes into protein solution and record tryptophan
spectra. Prior to measurement, protein-vesicle solutions are
incubated for at least 15 min. Tryptophan spectra are recorded
in the absence and presence of brominated phospholipids, for
example, (6,7)-, (9,10)-, and (11,12)-Br2-PC (see Note 12).
For all spectra measured in the presence of vesicles, lipid back-
ground scans at indicated lipid concentrations are collected and
subtracted. A phospholipid-to-protein molar ratio (L/P) of
200–400 is used (see Note 13).
ä
Fig. 1 (continued) manner, indicating penetration of MIM I-BAR domain into the
lipid bilayer. In contrast, the actin-binding protein profilin-1 did not show an
obvious increase in DPH anisotropy, suggesting that no prominent hydrophobic
interactions occurred. The lipid composition was POPC:POPE:POPS:PI(4,5)P2:
DPH (50:20:20:10:0.002, mol/mol) [15]
Fig. 2 Illustration of brominated PC shown with positions of bromides (blue circles) and their corresponding
distances from the bilayer center. hm is the mean insertion depth of tryptophan (Trp) of a protein in lipid
bilayer. The fluorescence quenching efficiency of Trp by phospholipids labeled with the fluorescence quencher
bromine atoms at different positions of the lipid acyl chains depends on the distance between Trp and bromine
atoms. Thus, depth-dependent quenching is applied to assess membrane penetration depth (hm) of intrinsic
tryptophan or labeled fluorophore of membrane-binding proteins in lipid bilayer
5. The logarithm of the ratio of steady-state Trp emission (see

Note 14) measured in the absence (F0) and presence (Fh) of
the brominated lipids is plotted as a function of distance from
the bilayer center (see Note 15). The insertion depth of Trp is
obtained by the distribution analysis [10–12] via fitting Gauss-
ian distribution using Eq. (1) and data analysis software (see
Note 16).
Eq. (1) (see Note 17).
F0 S ðhh m Þ2
ln ñc ðh Þ ¼ pffiffiffiffiffiffi ñe 2σ2 ð1Þ
F ðh Þ σ 2π
where F0 represents tryptophan fluorescence intensity in
the absence of quencher, F(h) is tryptophan fluorescence inten-
sity in the presence of quencher, and c(h) is the difference in
concentration of different quenchers. When equal concentra-
tions of the Br-lipids are used, the value for c(h) is 1 [11]. h is
the average bromine distance from a bilayer center (see Note
15). hm (mean insertion depth of tryptophan, see Note 18), σ
(dispersion, see Note 19), and S (area of the quenching profile)
describe the distribution of the quenched population.
4 Notes
1. Glassware but not plastics should be used to handle the organic

solvents.
2. According to the purpose of study, various phospholipids can
be used to make model membranes according to the lipid
compositions of distinct cellular membranes to elucidate inter-
actions of proteins with lipid bilayers, as well as their physio-
logical relevancies and subcellular localizations.
3. The temperature for lipid hydration should be higher than the
lipid phase transition temperature, which can be measured, for
example, by differential scanning calorimetry.
4. The temperature for extrusion should be higher than the lipid
phase transition temperature. LUV solutions will become
transparent.
5. Make sure that the Hamilton syringes are tightened properly to
prevent leakage of the solutions.
6. Avoid generating air bubbles during extrusions by pushing the
syringe slowly.
7. The temperature has to be kept constant during the measure-
ment as fluctuations in temperature affect membrane fluidity.
The temperature can be altered according to experimental
design.
8. Recombinant proteins are purified from E. coli as pure as possi-
ble, and remove the tags to avoid nonspecific binding or inter-
ference in membrane binding of the proteins.
9. The concentrations of the protein stock solutions should be
high enough to minimize the effects of total volume changes.
The volume of added protein is better not over 5% of the total
volume.
10. Brominated-PC lipids can be used to a final concentration of
20–50%.
11. Proteins labeled with fluorescent probes such as Alexa Fluor
dyes can be used instead of intrinsic tryptophan. The concen-
tration of protein in HBS depends on the fluorescence intensity
of tryptophan or labeled fluorophore. Typically, 0.5–5 μM
protein is used.
12. Other bromolipids or spin-labeled lipids can be used. More
data points will give a better and accurate fitting.
13. Use the condition that all proteins are bound to membranes.
The stoichiometry of membrane binding can be determined
using the vesicle co-sedimentation assay.
14. Tryptophan emission maximum is taken from the peak of the

spectrum.
15. The average bromine distances from the bilayer center based
on X-ray diffraction are 11, 8.3, and 6.5 Å for (6,7)-, (9,10)-,
and (11,12)-Br2-PC, respectively [13].
16. Other similar fitting software can be used. Alternatively, the
parallax method can be used to calculate the penetration depth
of proteins in lipid bilayer [11, 14]. It is important to note the
limitations of the method [11]. The parallax method operates
on only two fitting parameters: mean depth (hm) and radius of
quenching (Rc). The area and width of the quenching profile
are coupled to a single parameter Rc and cannot be changed
independently. Thus, a third fitting parameter, f, is introduced
for uncoupling of area and width of the depth-dependent
fluorescence quenching profile to make the data more
adequate.
17. If protein inserts deep into the membrane, which can be esti-
mated from the quenching profile, quenching coming from
both membrane leaflets should be considered, and a different
equation should be used for fitting [11]:
F0
ln ñc ðh Þ ¼ G ðh h m , S, σ Þ þ G ðh þ h m , S, σ Þ
F ðh Þ
18. The distance is calculated from the bilayer center [13].
19. The width at half maximum of the quenching profile.
Acknowledgments
This work was supported by grants from the Academy of Finland

(H.Z.), Jane and Aatos Erkko Foundation (H.Z.), Guangxi distin-
guished expert funding (H.Z.), FY 2015 Researcher Exchange
Program between JSPS and AF (Y.S.), Astellas Foundation for
Research on Metabolic Disorders (Y.S.), The Scandinavia-Japan
Sasakawa Foundation (Y.S.), The Ichiro Kanehara Foundation for
the Promotion of Medical Sciences and Medical Care (Y.S.), The
Association for Fordays Self-Reliance Support in Japan (Y.S.), The
Futaba Research Grant Program of the Futaba Foundation (Y.S.),
The NOVARTIS Foundation (Japan) for the Promotion of Science
(Y.S.), Okayama Foundation for Science and Technology (Y.S.),
Wesco Scientific Promotion Foundation (Y.S.), and JSPS
KAKENHI Grant Numbers JP19K23727, JP20K06589 (Y.S.).
References
1. van Meer G, Voelker DR, Feigenson GW 9. Zhao H, Lappalainen P (2012) A simple guide
(2008) Membrane lipids: where they are and to biochemical approaches for analyzing
how they behave. Nat Rev Mol Cell Biol protein-lipid interactions. Mol Biol Cell
9:112–124 23:2823–2830
2. van Meer G, de Kroon AIPM (2011) Lipid 10. Ladokhin AS (1997) Distribution analysis of
map of the mammalian cell. J Cell Sci 124:5–8 depth-dependent fluorescence quenching in
3. Balla T (2013) Phosphoinositides: tiny lipids membranes: a practical guide. Meth Enzymol
with giant impact on cell regulation. Physiol 278:462–473
Rev 93:1019–1137 11. Ladokhin AS (1999) Analysis of protein and
4. Saarikangas J, Zhao H, Lappalainen P (2010) peptide penetration into membranes by
Regulation of the actin cytoskeleton-plasma depth-dependent fluorescence quenching: the-
membrane interplay by phosphoinositides. oretical considerations. Biophys J 76:946–955
Physiol Rev 90:259–289 12. Ladokhin AS (2014) Measuring membrane
5. Zhao H, Michelot A, Koskela EV et al (2013) penetration with depth-dependent fluores-
Membrane-sculpting BAR domains generate cence quenching: distribution analysis is com-
stable lipid microdomains. Cell Rep ing of age. Biochim Biophys Acta
4:1213–1223 1838:2289–2295
6. Saarikangas J, Zhao H, Pyk€al€ainen A et al 13. McIntosh TJ, Holloway PW (1987) Determi-
(2009) Molecular mechanisms of membrane nation of the depth of bromine atoms in
deformation by I-BAR domain proteins. Curr bilayers formed from bromolipid probes. Bio-
Biol 19:95–107 chemistry 26:1783–1788
7. Drin G, Antonny B (2010) Amphipathic heli- 14. Chattopadhyay A, London E (1987) Parallax
ces and membrane curvature. FEBS Lett method for direct measurement of membrane
584:1840–1847 penetration depth utilizing fluorescence
8. Madsen KL, Bhatia VK, Gether U, Stamou D quenching by spin-labeled phospholipids. Bio-
(2010) BAR domains, amphipathic helices and chemistry 26:39–45
membrane-anchored proteins use the same 15. Senju Y, Kalimeri M, Koskela EV et al (2017)
mechanism to sense membrane curvature. Mechanistic principles underlying regulation of
FEBS Lett 584:1848–1855 the actin cytoskeleton by phosphoinositides.
Proc Natl Acad Sci U S A 114:E8977–E8986
Chapter 17
Fluorogenic XY-69 in Lipid Vesicles for Measuring Activity

of Phospholipase C Isozymes
Adam J. Carr, Edhriz Siraliev-Perez, Weigang Huang, John Sondek,
and Qisheng Zhang
Abstract
Mammalian phospholipase C (PLC) isozymes are major signaling nodes that regulate a wide range of
cellular processes. Dysregulation of PLC activity has been associated with a growing list of human diseases
such as cancer and Alzheimer’s disease. However, methods to directly and continuously monitor PLC
activity at membranes with high sensitivity and throughput are still lacking. We have developed XY-69, a
fluorogenic PIP2 analog, which can be efficiently hydrolyzed by PLC isozymes either in solution or at
membranes. Here, we describe the optimized assay conditions and protocol to measure the activity of
PLC-γ1 (D1165H) with XY-69 in lipid vesicles. The described protocol also applies to other PLC isozymes.
Key words Phospholipase C, phosphatidylinositol 4,5-bisphosphate, Lipid vesicles, Enzymatic assay,

Fluorogenic reporter, High-throughput screen
1 Introduction
The 13 mammalian PLC isozymes are divided into 6 subgroups (-β,

-γ, -δ, -ε, -ζ, and -η) based on sequence similarity and share a
conserved catalytic core [1]. Due to autoinhibition, the basal activ-
ities of PLCs are minimal in cells [2, 3]. When activated, PLCs
hydrolyze the minor membrane phospholipid phosphatidylinositol
4,5-bisphosphate (PIP2) at the inner leaflet of the plasma mem-
brane to generate the second messengers 1,2-diacylglycerol (DAG)
and inositol 1,4,5-trisphosphate (IP3). These second messengers
activate protein kinase C (PKC) isozymes and promote the release
of intracellular Ca2+ stores, respectively [4]. The consumption of
PIP2 also contributes to changes in the functions of membrane
proteins such as ion channels [5]. Consequently, PLCs are impor-
tant signaling proteins that regulate diverse cellular processes,
including proliferation, migration, and nerve conductance. Con-
versely, aberrant regulation of PLCs contributes to various diseases
such as cancer [6–8], Alzheimer’s disease [9, 10], and rheumatoid
225
226 Adam J. Carr et al.
arthritis [11, 12]. PLCs, particularly the PLC-γ isozymes, have

emerged as promising therapeutic targets.
Despite extensive studies on PLCs, it remains challenging to
continuously and directly monitor PLC activity at membranes with
high sensitivity and throughput. Traditionally, the phospholipase
activities of PLCs have been quantified using radiolabeled sub-
strates [13–16]. However, these formats have substantial limita-
tions. Primarily, they do not allow for continuous monitoring of
PLCs. In addition, the use of radioactive materials requires special
training, handling, storage, equipment, and facilities that are less
common than in previous decades. Indeed, radioactive PIP2 has
recently been discontinued by all chemical vendors, and it is expen-
sive to obtain such reagents through custom synthesis. Several
chromogenic or fluorogenic PIP2 analogs have been developed to
continuously monitor PLC activity [17–21]. However, most of
these compounds are either inefficient substrates of PLCs or have
other issues such as requiring a secondary enzymatic reaction not
related to PLCs. Importantly, none of these compounds reliably
report PLC activity at the membranes.
We created XY-69 as a membrane-associated PIP2 analog that
works as a selective fluorogenic reporter of PLC activity
[22]. XY-69 contains fluorescein as a fluorophore and
4-(dimethylaminoazo)benzene-4-carboxylic (DABCYL) acid as a
dark quencher that absorbs emission energy from fluorescein and
dissipates it as heat. PLC isozymes hydrolyze XY-69 to separate
DABCYL from fluorescein, leading to a large increase in fluores-
cence intensity that is readily monitored in real time. The assay
works both in detergent micelles and in lipid vesicles that more
closely resemble biological membranes and avoids the quench and
workup steps otherwise needed in radioactivity-based assays. Fur-
thermore, XY-69 functions in a microplate format, which enables
high-throughput discovery of molecules that modulate PLC activ-
ity. When presented in lipid vesicles, XY-69 captures the
membrane-dependent activation of PLC activity by Gαq [22], the
subunit of G-protein Gq that is known to activate PLC-β isozymes
[14, 23, 24], or by conformational changes due to mutations
[25]. Consequently, XY-69 is a compelling replacement for radio-
active PIP2 to monitor PLC activity with the added advantages of
continuous reaction monitoring and high-throughput. Assays
based on XY-69 are readily applicable to both purified PLC iso-
zymes and lysates from cell lines or animal tissues.
To further optimize the assays based on XY-69 to measure PLC
activity, we have investigated the impact of varying lipid composi-
tion, pH, free Ca2+ concentration, and vesicle size on assay perfor-
mance. Here, we describe optimized assay conditions and a detailed
protocol to measure the activity of purified, cancer-associated
mutant PLC-γ1 (D1165H) [8] with XY-69 embedded in lipid
Measuring PLC Activity in Lipid Vesicles 227
vesicles. The described protocol also applies to other PLC isozymes

and may be modified to include kinases, effectors, peptides, and
small molecules in the assay.
2 Materials
2.1 Equipment 1. Basic pH meter.

and Supplies 2. 0.45-μm polyethersulfone syringe filters.
3. 12 75 mm borosilicate culture tubes.
4. Dry nitrogen (N2) line equipped with a plastic hose and glass
Pasteur pipette.
5. Vacuum pump.
6. Black 384-shallow well microplate.
7. Ultrasonic dismembrator.
8. 5/6400 probe microtip for sonicator.
9. Multimode plate reader.
10. Ring stand.
11. Dynamic light scattering instrument.
12. 40-μL plastic cuvettes with stoppers.
13. Water bath.
14. Dewar flask.
15. Septum.
16. Pipette.
17. Pipette tips.
18. Syringe.
19. Vortex mixer.
2.2 Lipids and Other 1. 10 mg/mL (13.2 mM) L-α-phosphatidylethanolamine (PE;

Chemicals liver, bovine) stock solution in chloroform (CHCl3), stored at
20 C under N2.
2. 1 mg/mL (912 μM) L-α-phosphatidylinositol-4,5-bispho-
sphate (PIP2; brain, porcine; ammonium salt) stock solution
in CHCl3:methanol (MeOH):H2O (20:9:1; v/v/v), stored at
20 C under N2.
3. 210 μM XY-69 stock solution in H2O, stored at 80 C under
N2, wrapped in aluminum foil.
4. HPLC-grade MeOH.
5. Acetone.
6. Sodium hydroxide (NaOH).
7. 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid
(HEPES).
8. Potassium chloride (KCl).
9. Ethylene glycol-bis(2-aminoethylether)-N,N,N0 ,N0 -tetraacetic
acid (EGTA).
10. Calcium chloride (CaCl2).
11. 1,4-Dithiothreitol (DTT).
2.3 Proteins 1. PLC-γ1 (21-1215) referred to as PLC-γ1 (WT).

2. PLC-γ1 (21-1215) D1165H referred to as PLC-γ1
(D1165H).
3. Fatty acid-free bovine serum albumin (FAF BSA).
3 Methods
3.1 Overview The final assay is carried out in a 384-well plate at a volume of
12 μL. The assay mixture contains HEPES (33.9 mM, pH 7.4),
KCl (70 mM), EGTA (3 mM), CaCl2 (2.35 mM), DTT (2 mM),
FAF BSA (0.17 mg/mL), PE (192 μM), PIP2 (48 μM), XY-69
(0.5 μM), and the desired concentration of PLC isozyme (see Note
1). Under these conditions, the free Ca2+ concentration is approxi-
mately 390 nM (see Notes 2 and 3). The assay takes approximately
3 h from reagent setup to completion of data acquisition.
3.2 Preparation 1. Prepare stock solution A (see Note 4) containing 120 mM

of Buffers HEPES (pH 7.4) (see Note 5), 420 mM KCl, 18 mM EGTA,
and 14.1 mM CaCl2.
2. Make 1 mL of lipid vesicle assay buffer (LVAB) fresh for the
assay by adding 12 μL of 1 M DTT (stored at 20 C) to
988 μL of the stock solution A.
3. Make 1.2 mL of PLC dilution buffer (PLCDB) by mixing
200 μL of LVAB, 120 μL of 10 mg/mL FAF BSA, and
880 μL of H2O (see Note 6).
4. Make lipid sonication buffer containing 20 mM HEPES
(pH 7.4).
3.3 Generation 1. Briefly vortex and spin-down a thawed aliquot of XY-69 stock
of Lipid Films and Lipid solution (210 μM in H2O), and then transfer 1.5 μL to the
Vesicles bottom of a borosilicate culture tube.
2. Dry with a gentle stream of N2 for 5 min.
3. Allow PIP2 and PE stock solutions to equilibrate to room
temperature, and then vortex briefly.
4. Transfer 34.1 μL of PIP2 stock solution (912 μM) to the

bottom of the tube containing XY-69. Tap the tube gently to
re-dissolve the XY-69 flake.
5. Add 9.4 μL of PE stock solution (13.2 mM) into the PIP2/XY-
69 solution, followed by 20 μL of HPLC-grade MeOH (see
Note 7). Dry this mixture down to an opaque film on the
bottom of the tube using a gentle stream of N2 (see Note 8).
6. Leave the lipid film (see Note 9) under a gentle stream of N2 for
at least another 60 min to remove the bulk of organic solvent,
and then remove remaining traces of solvent by applying high
vacuum (0.5 mTorr) for at least 60 min (see Note 10).
7. Add 450 μL of lipid sonication buffer to the dry lipid film, and
then clamp the tube securely to a ring stand (see Note 11).
8. Submerge the probe tip of a sonicator halfway down into the
solution. The probe tip should be positioned vertically and
centered in the tube (see Note 12).
9. Place the tube in a Dewar flask containing iced water (ice bath)
and cool the solution for 45 s.
10. Pulse the sonicator at 20% output for three cycles of 5 s ON,
15 s OFF (see Note 13).
11. Remove the tube from the ice bath and cover the opening with
a septum. Confirm that the solution is free of particulate matter
and that no lipid film remains on the bottom of the tube. Allow
the solution to equilibrate to room temperature for at least
15 min.
12. Subject the lipid vesicles to freeze-thaw (see Note 14) by
immersing the tube in a Dewar flask containing acetone-dry
ice (dry ice bath, 78 C) for 40 s with gentle shaking, fol-
lowed by immersing in a water bath (35 C) for 70 s with
gentle shaking. Repeat this process for a total of six cycles.
13. Add 90 μL of LVAB directly to the solution at room tempera-
ture, and then vortex on a medium intensity setting for 20 s.
14. Protect the resulting lipid vesicle solution (LVS) from light and
store at room temperature (~20 C) until needed for the assay.
This solution is calculated to give 1.2 times the desired con-
centration of lipids in the final assay since 10 μL of LVS will be
mixed with 2 μL of PLC solution to make a 12-μL reaction
mixture.
3.4 Measurement 1. Dilute PLC proteins using ice-cold PLCDB to six times the
of PLC Activity final concentration in the assay since PLC solution composes
2 μL out of the final 12 μL in the reaction (see Note 15).
2. Keep PLC solutions on ice until needed for the assay.
3. Warm up the plate reader and allow it to equilibrate to assay

temperature (see Note 16).
4. Set the excitation wavelength at 485 nm and emission detec-
tion wavelength at 520 nm by using a FITC 485 excitation
filter and 520 emission filter (see Note 17).
5. To perform the assay in quadruplicate, distribute 2 μL aliquots
of PLC solution of a given concentration into four sequential
wells of a 384-well microplate (see Note 18). Repeat for each
desired concentration of enzyme. Blank PLCDB can be used as
a PLC-free or “BSA-only” control.
6. Add 10 μL of LVS into each well containing PLC solution to
initiate the assay and immediately insert the assay plate into the
plate reader for measurement (see Note 19).
7. Record fluorescence at required time points at 60-s intervals for
20–60 min (see Note 20).
8. Use the remaining LVS for dynamic light scattering (DLS)
analysis.
9. Add 8 μL of blank PLCDB to a cuvette, followed by 40 μL of
LVS and mix gently with a pipette. The resulting mixture
contains the same buffer and lipid component concentrations
as the microplate PLC assay. Cap the cuvette and insert the
sample into the DLS instrument for measurement at a temper-
ature matching the relevant PLC assay conditions (see Note
21).
4 Notes
1. The parameters presented here are optimal for the range of

1–100 nM PLC-γ1 (WT) and 0.1–50 nM PLC-γ1 (D1165H).
2. The concentration of free calcium ion is calculated using the
Maxchelator program, which accounts for temperature, pH,
and ionic strength of the assay mixture, as well as the concen-
trations of EGTA and total calcium (CaCl2).
3. All PLC isozymes are calcium dependent, and a calcium ion
binds to the active site within the TIM barrel. Therefore, the
concentration of free Ca2+ is an essential component for signal
generation in the XY-69 assay. Figure 1b shows that with
1.0 nM PLC-γ1 (D1165H), 100 nM free Ca2+ (this is near
basal cytosolic levels) [26] produces a modest rate of hydroly-
sis. However, rates of hydrolysis plateau at concentrations of
free Ca2+ greater than 400 nM. These results are consistent
with a binding affinity (kd) of ~250 nM for CaCl2 in PLC-γ1
(WT). At 100 μM free Ca2+, the assay reagents precipitated,
which terminated PLC activity.
Fig. 1 Effects of pH and free Ca2+ concentration on hydrolysis of XY-69 by PLC-γ1 (D1165H). (a) pH effect.
Lipid vesicles consisting of PE (192 μM), PIP2 (48 μM), and XY-69 (0.5 μM) in assay buffers with varying pH
were added to PLC-γ1 (D1165H) (1 nM) at 20 C or 30 C. The initial velocity of XY-69 hydrolysis was
measured and plotted against pH. The free Ca2+ concentration was 390 nM. (b) Effect of free Ca2+
concentration. The initial velocity of XY-69 (0.5 μM) hydrolysis by PLC-γ1 (D1165H) (1 nM) was measured
with varying concentrations of free Ca2+ in the assay buffer (pH 7.4) at 20 C or 30 C
4. All stock solutions are made with Milli-Q-grade water, adjusted

to the desired pH using 8 M NaOH, filtered through a 0.45-μ
m filter, and stored at 4 C.
5. The pH of the XY-69 assay buffer controls the protonation
state of key amino acid residues and influences the concentra-
tion of free Ca2+ ions, thereby regulating PLC activity. PLC-γ
isozymes are typically purified using HEPES buffer at a pH of
7.4. We have examined PLC-γ1 activity in 34 mM HEPES
across a pH range of 6.5 to 8.5 at 20 C and 30 C with the
results shown in Fig. 1a. These data suggest an optimal assay
buffer pH of 7.4 0.2, and that deviating from this range
causes precipitous loss of PLC-γ1 activity.
6. The final composition of PLCDB includes 20 mM HEPES
(pH 7.4), 70 mM KCl, 3 mM EGTA, 2.35 mM CaCl2,
2 mM DTT, and 1 mg/mL FAF BSA. We recommend making
PLCDB fresh for the assay and cool the solution on ice prior to
making PLC dilutions.
7. The addition of MeOH leads to a more even, homogeneous
film as the volume is reduced under the N2 stream. Warmth
from the hands also helps to prevent early precipitation of lipids
due to evaporative cooling.
8. The goal is to maintain XY-69, PE, and PIP2 on the bottom of

the tube where the resulting film will be most efficiently
reached by energy from the sonication probe tip. We recom-
mend drying the film gently to avoid blowing lipid solutions up
the sides of the tube and rotate the tube slowly while drying in
order to form an even layer.
9. We have investigated the effects of the composition of the lipid
film on the assay. The assay tolerates a wide range of vesicles
with different lipid compositions. We obtained the best data
quality using lipid vesicles with a final PE/PIP2 content of
220/20 or 192/48 μM.
10. We recommend the use of high vacuum when drying lipid
films, but if one is not available, then leave the lipid films
under streaming N2 for at least 2 h.
11. The outcome of lipid vesicle formation by probe sonication
depends on several factors, including solution volume, temper-
ature, vessel size and shape, lipid concentration, transducer
percent output, pulse time, probe tip immersion depth, and
probe tip diameter. Due to the sensitivity of this assay to the
properties of lipid vesicles, deviation from the recommended
parameters may affect assay performance and require
re-optimization.
12. If probe tip placement is too high, then foaming of the solution
may occur. If probe placement is too low, inadequate circula-
tion of contents may cause a problem. Review the manufac-
turer’s instructions for additional tips on optimizing this step.
13. If the scale of LVS is changed, this step may require
re-optimization of tube size, sonication power output, sonica-
tion time, and/or probe tip depth.
14. The freeze–thaw step increases the size of lipid vesicle.
15. When the solutions containing PLC isozymes are mixed, we
recommend gentle pipette mixing or inversion only.
16. This assay has been successfully performed in our laboratory at
temperatures ranging from 20 C to 37 C. Running the assay
at room temperature offers simplicity, and PLC activity gener-
ally increases with temperature under the listed conditions
without detriment to data quality.
17. Other default settings may need to be optimized depending on
the type of microplate and plate reader used.
18. The use of multichannel pipettes is recommended for mini-
mizing error and reducing the amount of time it takes to
assemble the assay. Otherwise, the experimenter should work
quickly to avoid losing early data points on highly active sam-
ples. Keep any microplates and feeder solutions covered as
much as practical to exclude dust, minimize evaporation, and
protect fluorescent molecules from light.
Fig. 2 Phospholipase C activity of PLC-γ1 (D1165H) in lipid vesicles. (a) XY-69 (0.5 μM) was embedded into
lipid vesicles containing PE (200 μM) and PIP2 (20 μM) and added to PLC-γ1 (D1165H) with indicated
concentration to initiate the hydrolysis. The reaction progression at 20 C was monitored continuously for
60 min by fluorescence (λex/λem ¼ 485/520 nm). The final assay buffer has a pH of 7.4 and free Ca2+
concentration at 390 nM. (b) The concentration of PE and PIP2 in lipid vesicles was 192 and 48 μM,
respectively. The experiment was run similarly as described in (a)
19. The addition of 10 μL of LVS into 2 μL of PLC solution should

provide adequate mixing of assay components. A single gentle
pipette mix of each well may be useful for homogenizing more
complex samples. Additionally, gentle tapping on the side of
the microplate may help even out all menisci prior to placing in
the reader.
20. The representative enzymatic reaction progression curves with
different concentrations of PLC-γ1 (D1165H) in two lipid
vesicle conditions are shown in Fig. 2.
21. Membrane curvature and vesicle size have been shown to
influence vesicle binding [27], domain conformation [28],
and catalytic activity [29, 30] of phospholipase enzymes. Con-
sequently, we investigated the effects of vesicle size on the
XY-69 based assay. Lipid vesicles generated from probe sonica-
tion alone have Z-average size (Zav) in the range of
150–160 nm (Fig. 3a). We noticed that repeated freezing and
thawing (freeze–thaw) of sonicated lipids produces larger vesi-
cles ranging from 180 to 250 nm (Fig. 3b). The autoinhibited,
wild-type PLC-γ1 has significantly higher activity in smaller-
sized lipid vesicles (Fig. 3c) than in larger-sized ones (Fig. 3d).
The relative activity of constitutively active mutant PLC-γ1
(D1165H) to that of wild-type enzyme is approximately two-
fold in vesicles prepared from probe sonication alone. This
ratio increased to nine-fold when lipid vesicles were prepared
Fig. 3 Size of lipid vesicle affects the relative rate of XY-69 hydrolysis by PLC-γ1 (WT) and PLC-γ1 (D1165H).
(a, b) The Z-average (Zav) particle size of lipid vesicles was measured by dynamic light scattering (DLS). The %
intensity of total scattered light was plotted versus particle size of lipid vesicles, which were formed either by
probe sonication only (a) or by first probe sonication and then freeze–thaw (b). (c, d) XY-69 (0.5 μM) hydrolysis
by PLC-γ1 (WT) or PLC-γ1 (D1165H) at 20 C was monitored by fluorescence (λex/λem ¼ 485/520 nm). Lipid
vesicles were formed either by probe sonication only (c) or by first probe sonication and then freeze–thaw (d).
All lipid vesicles contain PE (192 μM) and PIP2 (48 μM). The final pH of the assay buffer was 7.4, and the free
Ca2+ concentration was 390 nM
by first probe sonication and then freeze–thaw. Therefore,

membrane-dependent regulation of PLC activity due to exter-
nal stimulations or mutations is better monitored with XY-69
embedded in larger lipid vesicles.
Acknowledgments
This work was supported by the National Institutes of Health

(GM057391 to J. S. and CA177993 to Q. Z.) and the North
Carolina Biotechnology Center (TEG-2018-1505 to Q. Z.). We
thank Dr. Xiaoyang Wang for providing XY-69 and the Center for
Integrative Chemical Biology and Drug Discovery (CICBDD) at
the UNC Eshelman School of Pharmacy for the access to a micro-
plate reader.
References
1. Harden TK, Sondek J (2006) Regulation of 12. Afroz S, Giddaluru J, Vishwakarma S et al
phospholipase C isozymes by Ras superfamily (2017) A comprehensive gene expression
GTPases. Annu Rev Pharmacol Toxicol meta-analysis identifies novel immune signa-
46:355–379. https://doi.org/10.1146/ tures in rheumatoid arthritis patients. Front
annurev.pharmtox.46.120604.141223 Immunol 8:74. https://doi.org/10.3389/
2. Hicks SN, Jezyk MR, Gershburg S et al (2008) fimmu.2017.00074
General and versatile autoinhibition of PLC 13. Waldo GL, Ricks TK, Hicks SN et al (2010)
isozymes. Mol Cell 31:383–394. https://doi. Kinetic scaffolding mediated by a phospholi-
org/10.1016/j.molcel.2008.06.018 pase C-β and Gq signaling complex. Science
3. Hajicek N, Charpentier TH, Rush JR et al 330:974–980. https://doi.org/10.1126/sci
(2013) Autoinhibition and phosphorylation- ence.1193438
induced activation of phospholipase C-γ iso- 14. Smrcka AV, Hepler JR, Brown KO et al (1991)
zymes. Biochemistry 52:4810–4819. https:// Regulation of polyphosphoinositide-specific
doi.org/10.1021/bi400433b phospholipase C activity by purified
4. Rhee SG (2001) Regulation of Gq. Science 251:804–807. https://doi.org/
phosphoinositide-specific phospholipase 10.1126/science.1846707
C. Annu Rev Biochem 70:281–312. https:// 15. Seifert JP, Snyder JT, Sondek J et al (2006)
doi.org/10.1146/annurev.biochem.70.1.281 Direct activation of purified phospholipase
5. Balla T (2013) Phosphoinositides: tiny lipids C-ε by RhoA studied in reconstituted phos-
with giant impact on cell regulation. Physiol pholipid vesicles. Methods Enzymol
Rev 93:1019–1137. https://doi.org/10. 406:260–271. https://doi.org/10.1016/
1152/physrev.00028.2012 S0076-6879(06)06019-8
6. Koss H, Bunney TD, Behjati S et al (2014) 16. Litosch I (2000) Regulation of phospholipase
Dysfunction of phospholipase C-γ in immune C-β1 activity by phosphatidic acid. Biochemis-
disorders and cancer. Trends Biochem Sci try 39:7736–7743. https://doi.org/10.1021/
39:603–611. https://doi.org/10.1016/j.tibs. bi000022y
2014.09.004 17. Liu Y, Mihai C, Kubiak R et al (2007) Phos-
7. Martins M, McCarthy A, Baxendale R et al phorothiolate analogues of phosphatidylinosi-
(2014) Tumor suppressor role of phospholi- tos as assay substrates for phospholipase
pase C-ε in Ras-triggered cancers. Proc Natl C. Chembiochem 8:1430–1439. https://doi.
Acad Sci U S A 111:4239–4244. https://doi. org/10.1002/cbic.200700061
org/10.1073/pnas.1311500111 18. Rukavishnikov AV, Zaikova TO, Birrell GB
8. Kataoka K, Nagata Y, Kitanaka A et al (2015) et al (1999) Synthesis of a new fluorogenic
Integrated molecular analysis of adult T cell substrate for the continuous assay of mamma-
leukemia/lymphoma. Nat Genet lian phosphoinositide-specific phospholipase
47:1304–1315. https://doi.org/10.1038/ C. Bioorg Med Chem Lett 9:1133–1136.
ng.3415 https://doi.org/10.1016/s0960-894x(99)
9. Sims R, van der Lee SJ, Naj AC et al (2017) 00166-3
Rare coding variants in PLCG2, ABI3, and 19. Zaikova TO, Rukavishnikov AV, Birrell GB
TREM2 implicate microglial-mediated innate et al (2001) Synthesis of fluorogenic substrates
immunity in Alzheimer’s disease. Nat Genet for continuous assay of phosphatidylinositol-
49:1373–1384. https://doi.org/10.1038/ specific phospholipase C. Bioconjug Chem
ng.3916 12:307–313. https://doi.org/10.1021/
10. Magno L, Lessard CB, Martins M et al (2019) bc0001138
Alzheimer’s disease phospholipase C-gamma- 20. Rose TM, Prestwich GD (2006) Synthesis and
2 (PLCG2) protective variant is a functional evaluation of fluorogenic substrates for phos-
hypermorph. Alzheimers Res Ther 11:16. pholipase D and phospholipase C. Org Lett
https://doi.org/10.1186/s13195-019-0469- 8:2575–2578. https://doi.org/10.1021/
0 ol060773d
11. Cremasco V, Graham DB, Novack DV et al 21. Huang W, Hicks SN, Sondek J et al (2011) A
(2008) Vav/phospholipase C-γ2-mediated fluorogenic, small molecule reporter for mam-
control of a neutrophil-dependent murine malian phospholipase C isozymes. ACS Chem
model of rheumatoid arthritis. Arthritis Biol 6:223–228. https://doi.org/10.1021/
Rheum 58:2712–2722. https://doi.org/10. cb100308n
1002/art.23757
22. Huang WG, Wang XY, Endo-Streeter S et al 27. Wehbi H, Feng J, Kolbeck J et al (2003) Inves-
(2018) A membrane-associated, fluorogenic tigating the interfacial binding of bacterial
reporter for mammalian phospholipase C iso- phosphatidylinositol-specific phospholipase
zymes. J Biol Chem 293:1728–1735. https:// C. Biochemistry 42:9374–9382. https://doi.
doi.org/10.1074/jbc.RA117.000926 org/10.1021/bi034195+
23. Waldo GL, Boyer JL, Morris AJ et al (1991) 28. Uekama N, Aoki T, Maruoka T et al (2009)
Purification of an AlF4- and G-protein β- Influence of membrane curvature on the struc-
γ-subunit-regulated phospholipase ture of the membrane-associated pleckstrin
C-activating protein. J Biol Chem homology domain of phospholipase C-δ1. Bio-
266:14217–14225 chim Biophys Acta 1788:2575–2583. https://
24. Harden TK, Morris AJ, Waldo GL et al (1991) doi.org/10.1016/j.bbamem.2009.10.009
Avian G-protein-regulated phospholipase 29. Ahyayauch H, Villar AV, Alonso A et al (2005)
C. Biochem Soc Trans 19:342–346. https:// Modulation of PI-specific phospholipase C by
doi.org/10.1042/bst0190342 membrane curvature and molecular order. Bio-
25. Hajicek N, Keith NC, Siraliev-Perez E et al chemistry 44:11592–11600. https://doi.org/
(2019) Structural basis for the activation of 10.1021/bi050715k
PLC-γ isozymes by phosphorylation and 30. Burack WR, Biltonen RL (1994) Lipid bilayer
cancer-associated mutations. Elife 8:e51700. heterogeneities and modulation of phospholi-
https://doi.org/10.7554/eLife.51700 pase A2 activity. Chem Phys Lipids
26. Clapham DE (2007) Calcium signaling. Cell 73:209–222. https://doi.org/10.1016/
131:1047–1058. https://doi.org/10.1016/j. 0009-3084(94)90182-1
cell.2007.11.028
INDEX
A F
Actin........................................................................ 92, 195 Fabrication................................................... 147, 149, 150
Acyl chains ...........................................5, 21, 32, 158, 216 Fatty acid .....................................................................4, 40
Aggregation .......................................................... 210, 213 Fluorescence colocalization .......................................... 134
Anisotropy ................................................... 216, 217, 219 Fluorescence microscopy ........................................ 56, 74,
77, 83–84, 134
B Fluorescence quenching ..................................... 216, 219,
Biosensors ....................................................55–68, 74–77, 220, 222
80, 92, 107, 108, 114 Fluorescence spectroscopy.................................. 121–131,
201, 205–213
C Fluorescent lipids .......................................................... 179
Fractionation ................................................................2, 4,
Calcium...................................................... 6, 76, 228, 230 10, 13, 122
Calibration..................................................................... 168 Fluorescent proteins..........................................56, 63, 67,
Cell culture ........................................................ 60, 80–82, 73–87, 107, 113, 116
85, 93, 95, 107, 108, 112, 125, 127, 136, 137
Cell lines ...........................................................60, 80, 226 G
Centrifugation .............................................. 9, 11, 28, 45,
158, 165, 166, 171, 173, 181, 189–192, 196, 198 Genetically encoded biosensors................................55–68
Glutathione S-transferase (GST)............................ 24, 29,
Chemical dimerization.................................................. 106
Chromatography .......................................... 8, 13, 25, 30, 34, 124, 131, 178–183, 189, 191–194, 197, 198,
34, 44, 55, 163, 165, 166, 189, 197, 209 202, 203
Green fluorescent proteins (GFP).......................v, 56, 94,
Co-flotation .......................................................... 195–203
Confocal microscopy .................................. 56, 60, 63, 64 108, 114, 116
Co-sedimentation ........................................195–203, 221 GTPases ................................................................ 185–194
Cytoskeleton................................................ 121, 185, 195
H
D High performance liquid chromatography
(HPLC).................................................. 4, 5, 8–10,
Deacylation............................................2, 4, 9, 11, 15, 40
Derivatization ....................................................21, 23, 26, 13, 14, 16, 40, 41, 44, 51, 55, 227, 229
28, 29, 35, 40, 43 His-tag ........................................123, 186, 197, 202, 203
Device assembly ................................................... 148, 150
I
Device fabrication ........................................147, 149–150
DH domain .......................................................... 186, 191 Image analysis.......................................................... 61, 80,
DPH anisotropy .......................................... 216, 217, 219 81, 112, 114
Drosophila melanogaster.................................................. 22 Image processing.................................................... 61, 148
Image quantification ............................. 56, 60, 61, 63–68
E Image reconstruction...................................................... 99
Enzyme activation ............................................................. 2 Induced dimerization........................................... 105–117
Enzyme regulation .............................................. 1, 19, 74, Inositols .............................................................1, 7, 8, 10,
12, 19–21, 39, 74, 92, 106, 157, 205, 225
76, 225–234
Enzymes..................................................2, 23, 29, 34, 55, In situ.................................................................... 133–141
56, 64, 105, 107, 115, 136, 230, 233 Internal standards..............................................22, 23, 26,
27, 31, 32, 34, 43, 46, 50, 51
Epifluorescence microscopy......................................83, 84
https://doi.org/10.1007/978-1-0716-1142-5, © Springer Science+Business Media, LLC, part of Springer Nature 2021
237
PHOSPHOINOSITIDES: METHODS AND PROTOCOLS
238 Index
K Multilamellar vesicles (MLVs).................... 149, 199, 217
Myo-3H-inositol.....................................2, 7, 8, 10, 11, 15
Kinetics ................................................116, 161, 206, 208
N
L
Nanoscale.................................................................91–102
Label free ...................................................................19–36 Native mass spectrometry.................................... 157–173
Labelling ........................................................................ 207
Large unilamellar vesicles O
(LUVs) ............................................. 208, 217, 219
Laser scanning confocal microscopy ............................ 113 Organelles............................................20, 55, 60, 65, 157
Ligand binding............................................ 158, 161, 163
P
Ligands ................................................................ 106, 121,
158–161, 172 PALM ...................................................... 93–99, 101, 102
Lipid extraction ................................................ 25–27, 122 PCR.................................................................................. 94
Lipid kinases ......................................................... 1, 19, 74 Phosphatidylinositol............................................. 105–116
Lipidomics ............................................................ 163, 172 Phosphatidylinositol-3-phosphate (PI3P) ..................2, 4,
Lipid phosphatase .................................................... v, 2, 3, 5, 10, 13, 19, 188, 194
19, 109, 110 Phosphatidylinositol-3,4-bisphosphate ...................19, 75
Lipid preparation..................................................... 43, 46, Phosphatidylinositol-3,4,5-trisphosphate
148, 150, 170 (PIP3) .............................................. 19, 20, 25–27,
Lipid profiling ......................................................... 39, 40, 31, 34, 36, 40, 43, 46, 75, 76, 80, 82–85, 157
43, 46, 51, 52 Phosphatidylinositol-3,5-bisphosphate (PtdIns(3,5)
Lipid-protein overlay assays................................ 177, 186, P2 ................................................................. 19, 206
189, 192–193 Phosphatidylinositol-4-phosphate [PI(4)P] .................. 74
Liposome flotation.......................................178–181, 183 Phosphatidylinositol-4,5-bisphosphate
Liposomes............................................................ 122, 124, [PI(4,5)P2].......................................................... 74
125, 128–132, 158, 164, 169, 178–184, Phosphatidylinositol-5-phosphate (PI5P) ................5, 13,
186–194, 196, 199, 200, 202, 203, 206, 207, 14, 16, 19, 20, 23, 25–27, 29, 34, 36, 188
209–213, 216, 219 Phosphatidylinositols ................................... 1, 13, 19, 55,
Liposome sedimentation ............................ 134, 177, 194 73, 105–117, 121, 157, 195, 203, 205, 206, 225
Liquid chromatography (LC)..................... 20, 30, 44, 55 Phosphoinositides (PIs) .............................................1–16,
Live-cell imaging ..................................................... 80, 82, 19–36, 39–52, 73–85, 91–102, 109, 134, 136,
85, 94, 95, 99, 101 140, 157–173, 183, 185–196, 202, 203, 205,
209, 215
M Phospholipase C (PLC) ...........................................75–77,
Mammalian cells........................................ 1, 5, 10–13, 16 80, 83, 85, 225–234
Mass assay .................................................... 26, 27, 29, 40 Phospholipases ............................................ 144, 226, 233
Mass spectrometry (MS)..................................... 5, 19–37, Phospholipids ........................................................ 2, 5, 55,
40, 44, 47, 50, 52, 55, 158–169, 171, 172 163, 164, 173, 178–180, 182, 183, 186, 193,
Membrane ........................................................... 4, 20, 55, 196, 205, 208, 209, 211, 216, 219, 221, 225
56, 60, 64, 68, 73–75, 77, 92, 97, 98, 101, 105, Plasma membrane (PM) ................................... 25, 57–61,
107, 108, 115, 121, 140, 143–155, 157, 158, 64–68, 73–87, 91–102, 106–108, 114–117, 121,
161, 163, 164, 166, 167, 177, 178, 185, 186, 195, 225
189, 192, 193, 195, 196, 202, 203, 206–208, Plasmids .............................................................56, 61, 62,
213, 215–217, 221, 222, 225, 226, 233 67, 74, 80, 81, 85, 93, 107, 112, 114, 115, 122,
Membrane insertion...................................................... 216 125, 126, 191
Membrane-on-a-chip ........................................... 143–155 Platelets......................................................................39–52
Membrane penetration ........................................ 215–222 Pleckstrin homology (PH) domain.............................. 121
Membrane proteins............................................. 105, 158, Post-translational modification (PTM)........................ 134
159, 161, 163–168, 171, 172, 216, 225 Protein expressions ............................................. 105, 125,
Membrane sheets ..................................93, 95, 96, 98, 99 127, 130, 197–199, 202
Metabolic labeling................................................ 1–16, 40 Protien-ligand affinity .......................................... 158, 162
Microfluidics......................................................... 143–155 Protein-lipid interaction by fluorescence
Mouse embryonic fibroblasts (MEF).................. 7, 11, 12 (PLIF) ......................................177, 178, 180–183
PHOSPHOINOSITIDES: METHODS AND PROTOCOLS
Index 239
Protein-lipid interactions .............................180–182, 196 Signal transduction .............................1, 73, 92, 157, 185
Protein purification ............................................. 158, 168, Single-molecule ................................ 92, 94, 99, 101, 102
171, 183, 197–199, 207 Single-molecule localization...................................91–102
Proteins2, 11, 20, 28, 29, 34, 39, 51, 56, 67, 74, 75, 91, Small unilamellar vesicles (SUVs) ...................... 144, 146,
92, 101, 105–108, 115, 121, 123, 124, 126–128, 149, 150, 153, 154
131, 133, 134, 136, 140, 141, 143, 144, 148, Soluble proteins .......................................... 131, 161, 167
151–155, 158–163, 165–173, 177, 178, Spectroscopy...................................................55, 143, 166
180–183, 185, 186, 189, 191, 193, 195–200, Spinning disc confocal microscopy
202, 203, 205–207, 209–212, 215–222, 225, (SDCM) .....................................75, 77, 80, 82, 83
228, 229 Stopped-flow fluorescence spectroscopy ............ 205–213
Proximity ligation assay (PLA) ............................ 133–141 Super resolution microscopy ..................................91–102
Supported lipid bilayers (SLBs)................. 148, 150–152,
Q 154, 155
Quantification ................................................... 19–36, 40,
T
41, 47, 50, 56, 60, 64, 74, 83, 84, 139, 140, 192,
208, 209 Time-lapse imaging...................................................64, 67
Quenching.........................................................29, 80, 83, Total internal reflection fluorescence (TIRF)
87, 216, 219, 220, 222 microscopy ........................................................... 75
Total organic phosphate .................................... 22, 28, 35
R Transfections ..................................................... 56, 61–63,
67, 74, 77, 80–82, 85, 93, 95, 107, 112, 115
Radioactive labelling ......................................1–16, 20, 40
Radioactive probes ............................................................ 2
V
Radioactivity .......................................... 1–16, 20, 40, 226
Rapamycin ...........................................106, 108, 112–116 Vesicles ................................................................. 149–151,
Real-time ........................................................56, 121–131 154, 158, 196, 210, 216, 219, 221, 226, 228,
Recombinant protein expression......................... 125–127 229, 232–234
S Y
Sedimentation ............................................. 189, 192, 196 Yeast ................................... 5, 8–10, 14, 15, 22, 123, 206

2021 - Book - Phosphoinositides Methods and Protocols

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2021 - Book - Phosphoinositides Methods and Protocols

Uploaded by

Copyright:

Available Formats

Methods in

Molecular Biology 2251

Roberto J. Botelho Editor

For further volumes:

Methods and Protocols

ISSN 1064-3745 ISSN 1940-6029 (electronic)

© Springer Science+Business Media, LLC, part of Springer Nature 2021

Toronto, ON, Canada Roberto J. Botelho

1 Simultaneous Detection of Phosphoinositide Lipids by Radioactive

12 Liposome-Based Methods to Study Protein–Phosphoinositide

RICHARD A. ANDERSON • University of Wisconsin-Madison, School of Medicine and Public

HUDSON T. HORN • University of Wisconsin-Madison, School of Medicine and Public Health,

Simultaneous Detection of Phosphoinositide Lipids

Phosphorylated phosphatidylinositol signaling lipids (PPI, also

phosphatases, and temporally via specific activation of these

Fig. 2 The deacylation reaction of PPIs to glycerophosphoinositides. Mild-base

phosphorylated glycero-3-phospho-1-inositol and fatty acid

2.1 General System 1. Tabletop microcentrifuge.

2.2 Materials 1. myo-inositol,-[2-3H] (PerkinElmer) dissolved in ddH2O with

concentration of 10.7% methylamine, 45.7% methanol, and

100 trace elements solution, 50 mL 10 ammonium sulfate

2.4 Materials 1. Custom Dulbecco’s modified Eagle’s medium (DMEM):

interval. For control cells not undergoing hyperosmotic shock,

6. Dry the sample in a SpeedVac concentrator at 55  C for 4 h

48 h for MEF and macrophages and 24 h for primary cortical

confluency 24 h later. MEF cells are then washed and incubated

6. Transfer the bottom layer (aqueous phase) to a new

1. While other HPLC and flow scintillation systems may be

(Shimadzu). Our lab uses Laura-4 software (LabLogic) for

6. Perchloric acid must be stored in a designated acid cabinet

9. While this protocol should yield an appropriate amount of

This work was supported by NIH grants R01-NS099340-03 and

Label-Free Quantification of Phosphoinositides

Key words Phosphoinositides, Drosophila, Mass spectrometry, label-free quantification, In vivo

Phosphoinositides (PIs) are the phosphorylated derivatives of

therefore restricted to the membrane at which they are produced.

2. In addition to the inositol headgroup, each PI contains two

the signal. The details of performing the experiment and quan-

2.1 Fly Stock Rearing 1. Flies (Drosophila melanogaster).

2.2 Total Lipid 1. Glass bottles: transparent and amber color.

8. Phosphoinositide elution buffer (PEB): chloroform/metha-

10. Lower phase wash buffer (LPWS): methanol/1 M hydrochlo-

2.3 Lipid Internal 1. 17:0|14:1 PE.

2.3.2 For Larvae 1. 17:0|20:4 PI(4,5)P2.

2.4 Neomycin 1. Neo-beads™ (Echelon Biosciences) (see Note 3).

2.5 18O PI5P 1. 10 mM Tris–HCl (pH 7.4).

2.6 Derivatization 1. TMS-diazomethane.

2.7 Total Organic 1. High-temperature-resistant phosphate-free glass tubes.

2.9 Lab Equipment 1. Homogenizer instrument (see Note 5).

3.1 Lipid Extraction Phosphoinositides support a varied range of physiological functions

2. Dissect 25 retinae per sample under conditions of appropriate

2. Spin down the beads at 1000 g for 5 min.

3.2 Organic 1. Take 500 μL from flow-through collected in step 3 of Sub-

3.5 Liquid 1. Separate samples obtained from step 8 of Subheading 3.4 by

Time (in min) % solvent A (0.1% % solvent B (0.1%

Parameters PIP species PIP2 species PIP3 species PE species

have been described in detail previously and in Table 2 [7, 8,

3. Similarly, divide the area under the curve value of each

Acyl chain length PE parent ion PE daughter ion

1. Do not store chloroform in large quantities and for longer

6. Keep this ready the day before the derivatization reaction.

16. Using a vacuum centrifugal concentrator for a long time can

This work was supported by the National Centre for Biological

6. Dry the sample in a SpeedVac concentrator at 55 C for 4 h