Methods in

Molecular Biology 1722

Kenneth R. Boheler
Rebekah L. Gundry Editors

The
Surfaceome
Methods and Protocols
METHODS IN MOLECULAR BIOLOGY

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

For further volumes:

http://www.springer.com/series/7651
 The Surfaceome

Methods and Protocols

Edited by

Kenneth R. Boheler
School of Biomedical Sciences, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong,
SAR, China; Stem Cell & Regenerative Medicine Consortium, School of Biomedical Sciences,
LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, SAR, China

Rebekah L. Gundry
Department of Biochemistry, Medical College of Wisconsin, Milwaukee, WI, USA
Editors
Kenneth R. Boheler Rebekah L. Gundry
School of Biomedical Sciences Department of Biochemistry
LKS Faculty of Medicine Medical College of Wisconsin
The University of Hong Kong Milwaukee, WI, USA
Hong Kong, SAR, China
Stem Cell & Regenerative Medicine
Consortium, School of Biomedical Sciences
LKS Faculty of Medicine
The University of Hong Kong
Hong Kong, SAR, China

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-7551-8 ISBN 978-1-4939-7553-2 (eBook)
https://doi.org/10.1007/978-1-4939-7553-2
Library of Congress Control Number: 2017960803

© Springer Science+Business Media, LLC 2018

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

Printed on acid-free paper

This Humana Press imprint is published by Springer Nature

The registered company is Springer Science+Business Media, LLC
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface

The plasma membrane is the gateway through which cells sense and respond to their
microenvironment. Critical to this process are cell surface proteins that span (transmem-
brane) or are anchored/embedded in the plasma membrane. Cell surface proteins perform
diverse functions, including nutrient and ion transport, intra- and intercellular communica-
tion, receptor signaling, and enzymatic reactions. Altogether, the collection of proteins that
reside at the cell surface (i.e., surfaceome) facilitates interactions with pathogens, binding of
chemical messengers, and transmission of signaling cascades, and it is required for cell
migration, adhesion, and survival. Surfaceome content, including protein identity and
modifications, differs among cell types and is dynamic during development and disease
states. For these reasons, and the fact that cell surface proteins are accessible, the surfaceome
is a rich source of drug and immunotherapy targets and contains unique markers that can be
used to identify cell types, disease states, and cellular phenotypes. Despite their critical
functions in health and disease, cell surface proteins have historically been understudied in
most cell and tissue types. This is due, in part, to the challenges posed by their relatively low
abundance when compared to intracellular proteins, their hydrophobic nature, and the
difficulty in biophysically purifying plasma membrane proteins without contamination
from intracellular membrane components. Moreover, high-quality antibodies are currently
available for a limited subset of cell surface proteins.
Considering these challenges, the development and dissemination of modern methods
and technologies that enable the study of cell surface proteins will undoubtedly advance a
broad range of research efforts, including our understanding of cellular differentiation and
development, host-pathogen interactions, and metastatic processes, and will lead to the
development of new treatments for disease. In this volume of Methods in Molecular
Biology: The Surfacome, we have assembled 19 chapters that cover a variety of methods
ranging from molecular and cellular biology to proteomics to bioinformatics. The overall
aim of this edition is to provide state-of-the-art techniques and tools to assess the surfa-
ceome content, modifications, and function. The volume does not include standard
approaches extensively reviewed elsewhere, nor does it include methods to analyze lipids
and glycans, which are key components of the plasma membrane and worthy of separate
volumes dedicated to their study. While most of the methods described in this volume are
generally applicable to any cell type, some chapters focus on specific cell types and/or
specific molecule classes of interest. These latter chapters are designed to illustrate the
application of these procedures and protocols in defined systems, but the approaches should
be applicable across a broad range of cells. Altogether, we hope this collection of methods
will facilitate the study of cell surface protein biology and function and lead to the discovery
of new drug and immunotherapy targets for treating disease and new immunophenotyping
markers for studying cellular function, differentiation, and disease. The chapters are
arranged in four parts, beginning with discovery-based and then targeted strategies for
cataloging surfaceome content, moving to functional assays for specific protein and cell
types, and ending with computational approaches.
Part I focuses on discovery-based approaches for cataloging surfaceome content and
includes methods to analyze the surfaceome of bacteria, avian embryos, and mammalian
systems. Chapters in this part focus on modern proteomic methods that offer the ability to

v
vi Preface

specifically target cell surface proteins with limited interference from intracellular membrane
proteins. These include surface membrane protein enrichment techniques, using proteases
to “shave” proteins from the surface of bacteria to identify surface-exposed proteins, and
exploiting the avian system to study developmental changes in cell surface proteins, includ-
ing bioinformatics-based techniques to translate to human orthologs. Subsequent chapters
describe the Cell Surface Capture Technology, a targeted analytical approach to specifically
identify cell surface N-glycoproteins, the use of iron oxide nanoparticles to enrich plasma
membrane proteins, and methods to profile secreted proteins and exosomes in cell culture, a
topic that has recently gained attention across a variety of research disciplines.
Part II focuses on targeted approaches to analyze the surfaceome. The chapters in this
part include methods to overexpress specific targets in Sf9 cells and an approach to generate
bispecific antibodies that are valuable for targeting cancer and somatic cells. Also included is
a tutorial chapter on flow cytometry and its application to immunophenotyping to assist
novices in their pursuit of surface proteins. The last chapter in this part provides an example
of how ELISA and flow cytometry are applied to detecting the G protein-coupled receptor
CXCR4, a strategy particularly valuable for investigators interested in G proteins and in drug
repurposing.
Part III focuses on cell-based functional analyses. This part begins with a review on
voltage-dependent sodium channels and methods for high content electrophysiological
analyses. Methods are then described for the evaluation of vascular endothelial cell functions
and approaches to study signal transduction of surface receptor tyrosine kinase in neurons. A
comprehensive analysis of cell polarity, using retinal pigmented epithelium as a model
system, is then described, including techniques for immunostaining for apical and basolat-
eral membrane markers, polarized cytokine secretion, fluid transport, phagocytosis, and
identification of plasma membrane proteins through cell surface capturing technologies as
described in the first part. This part finishes with a description of methods that take
advantage of extracellular matrix components to capture mesenchymal stromal cells under
flow, model disease states, and ultimately analyze cell-matrix interactions through the use of
3D microtissues.
Part IV focuses on computational approaches in surfaceome studies and describes a new
web-based platform, Targets-search, that incorporates information from a variety of sources
including the Cell Surface Protein Atlas and online drug databases, to facilitate identification
of surface proteins that are informative for a particular cell type or disease and known drugs
that interact with these proteins.
In closing, we would like to thank Springer for its support, dedication to this project,
and patience in developing this book. We also wish to especially thank all of the authors for
their time, energy, and valuable contributions. With their efforts, we have assembled what
we hope will be a valuable resource for those research laboratories working to advance the
study of surface protein biology.

Hong Kong, SAR, China Kenneth R. Boheler

Milwaukee, WI, USA Rebekah L. Gundry
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

PART I DISCOVERY-BASED APPROACHES TO SURFACEOME CONTENT

1 Surfaceome Analysis Protocol for the Identification of Novel
Bordetella pertussis Antigens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Yulanda M. Williamson, Jennifer Whitmon, Rolieria West-Deadwyler,
Hercules Moura, Adrian R. Woolfitt, Jon Rees, David M. Schieltz,
and John R. Barr
2 “Shaving” Live Bacterial Cells with Proteases for Proteomic
Analysis of Surface Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Manuel J. Rodrı́guez-Ortega
3 Methods for Mapping the Extracellular and Membrane
Proteome in the Avian Embryo, and Identification of Putative
Vascular Targets or Endothelial Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Witold W. Kilarski, John Herbert, and Andreas Bikfalvi
4 Mass Spectrometry-Based Identification of Extracellular Domains
of Cell Surface N-Glycoproteins: Defining the Accessible Surfaceome
for Immunophenotyping Stem Cells and Their Derivatives. . . . . . . . . . . . . . . . . . . 57
Chelsea M. Fujinaka, Matthew Waas, and Rebekah L. Gundry
5 Application of Higher Density Iron Oxide Nanoparticle Pellicles
to Enrich the Plasma Membrane and Its Proteome from Cells
in Suspension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Rebecca L. Rose, Waeowalee Choksawangkarn, and Catherine Fenselau
6 Proteomic Profiling of Secreted Proteins, Exosomes,
and Microvesicles in Cell Culture Conditioned Media . . . . . . . . . . . . . . . . . . . . . . . 91
Ankit Sinha, Simona Principe, Javier Alfaro, Alex Ignatchenko,
Vladimir Ignatchenko, and Thomas Kislinger

PART II TARGETED APPROACHES FOR SURFACEOME CONTENT

7 Cloning, Expression, and Purification of the Glycosylated
Transmembrane Protein, Cation-Dependent Mannose
6-Phosphate Receptor, from Sf9 Cells Using the Baculovirus System . . . . . . . . . . 105
Linda J. Olson and Nancy M. Dahms
8 Bispecific Antibody Armed T Cells to Target Cancer Cells . . . . . . . . . . . . . . . . . . . 117
Archana Thakur, Lawrence G. Lum, and Sandeep Mittal
9 Immunophenotyping of Live Human Pluripotent Stem Cells
by Flow Cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Daniel R. Riordon and Kenneth R. Boheler

vii
viii Contents

10 Detecting Cell Surface Expression of the G Protein-Coupled

Receptor CXCR4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Amanda M. Nevins and Adriano Marchese

PART III CELL-BASED FUNCTIONAL ANALYSES RELATED

TO SURFACEOME CONTENT

11 NaV Channels: Assaying Biosynthesis, Trafficking, Function. . . . . . . . . . . . . . . . . . 167

Gordon F. Tomaselli and Federica Farinelli
12 High-Content Electrophysiological Analysis of Human
Pluripotent Stem Cell-Derived Cardiomyocytes (hPSC-CMs) . . . . . . . . . . . . . . . . 185
Chi-Wing Kong, Lin Geng, and Ronald A. Li
13 Methods for Evaluation of Vascular Endothelial Cell Function
with Transient Receptor Potential (TRP) Channel Drugs . . . . . . . . . . . . . . . . . . . . 195
Yung Wui Tjong and Xiaoqiang Yao
14 Methods to Study the Signal Transduction of the Surface
Receptor Tyrosine Kinase TrkB in Neurons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Kwok-On Lai and Nancy Y. Ip
15 Polarized Human Retinal Pigment Epithelium Exhibits
Distinct Surface Proteome on Apical and Basal Plasma Membranes . . . . . . . . . . . 223
Vladimir Khristov, Qin Wan, Ruchi Sharma, Mostafa Lotfi,
Arvydas Maminishkis, and Kapil Bharti
16 Extracellular Matrix Molecule-Based Capture of Mesenchymal
Stromal Cells Under Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Teresa Massam-Wu, Stuart A. Cain, and Cay M. Kielty
17 Generation of Induced Pluripotent Stem Cells from Patients
with COL3A1 Mutations and Differentiation to Smooth
Muscle Cells for ECM-Surfaceome Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Jiaozi He, Zhihui Weng, Stanley Chun Ming Wu,
and Kenneth R. Boheler
18 Fabrication and Mechanical Properties Measurements of 3D
Microtissues for the Study of Cell–Matrix Interactions . . . . . . . . . . . . . . . . . . . . . . 303
Prasenjit Bose, Chen Yu Huang, Jeroen Eyckmans,
Christopher S. Chen, and Daniel H. Reich

PART IV COMPUTATIONAL APPROACHES IN SURFACEOME STUDIES

19 Discovery of Surface Target Proteins Linking Drugs, Molecular Markers,

Gene Regulation, Protein Networks, and Disease by Using a
Web-Based Platform Targets-search. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
Bin Yan, Panwen Wang, Junwen Wang, and Kenneth R. Boheler

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Contributors

JAVIER ALFARO  Department of Medical Biophysics, University of Toronto, Toronto, ON,

Canada; Princess Margaret Cancer Centre, Toronto, ON, Canada
JOHN R. BARR  Division of Laboratory Sciences, National Center for Environmental Health,
Centers for Disease Control and Prevention, Chamblee, GA, USA
KAPIL BHARTI  Unit on Ocular and Stem Cell Translational Research, National Eye
Institute, National Institute of Health, Bethesda, MD, USA
ANDREAS BIKFALVI  Angiogenesis and Tumor Microenvironment Laboratory, INSERM
U1029, Pessac, France; Angiogenesis and Tumor Microenvironment Laboratory,
University Bordeaux, Pessac, France
KENNETH R. BOHELER  School of Biomedical Sciences, LKS Faculty of Medicine, The
University of Hong Kong, Hong Kong, SAR, China; Stem Cell & Regenerative Medicine
Consortium, School of Biomedical Sciences, LKS Faculty of Medicine, The University of
Hong Kong, Hong Kong, SAR, China
PRASENJIT BOSE  Department of Physics and Astronomy, Johns Hopkins University,
Baltimore, MD, USA
STUART A. CAIN  Wellcome Trust Centre for Cell-Matrix Research, Faculty of Life Biology,
Medicine and Health, University of Manchester, Manchester, UK
CHRISTOPHER S. CHEN  Department of Biomedical Engineering, Biological Design Center,
Boston University, Boston, MA, USA; The Wyss Institute for Biologically Inspired
Engineering, Harvard University, Boston, MA, USA
WAEOWALEE CHOKSAWANGKARN  Department of Biochemistry, Faculty of Science, Burapha
University, Mueang, Chonburi, Thailand
NANCY M. DAHMS  Department of Biochemistry, Medical College of Wisconsin, Milwaukee,
WI, USA
JEROEN EYCKMANS  Department of Biomedical Engineering, Biological Design Center,
Boston University, Boston, MA, USA; The Wyss Institute for Biologically Inspired
Engineering, Harvard University, Boston, MA, USA
FEDERICA FARINELLI  Division of Cardiology, Department of Medicine, Johns Hopkins
University, Baltimore, MD, USA
CATHERINE FENSELAU  Department of Chemistry and Biochemistry, University of Maryland,
College Park, MD, USA
CHELSEA M. FUJINAKA  Department of Biochemistry, Medical College of Wisconsin,
Milwaukee, WI, USA
LIN GENG  Department of Paediatrics and Adolescent Medicine, School of Biomedical
Sciences, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, SAR, China
REBEKAH L. GUNDRY  Department of Biochemistry, Medical College of Wisconsin,
Milwaukee, WI, USA
JIAOZI HE  School of Biomedical Sciences, LKS Faculty of Medicine, The University of Hong
Kong, Hong Kong, SAR, China
JOHN HERBERT  Institute of Integrative Biology, University of Liverpool, Liverpool, UK
CHEN YU HUANG  Department of Physics and Astronomy, Johns Hopkins University,
Baltimore, MD, USA
ALEX IGNATCHENKO  Princess Margaret Cancer Centre, Toronto, ON, Canada

ix
x Contributors

VLADIMIR IGNATCHENKO  Princess Margaret Cancer Centre, Toronto, ON, Canada

NANCY Y. IP  Division of Life Science, Molecular Neuroscience Center and State Key
Laboratory of Molecular Neuroscience, The Hong Kong University of Science and
Technology, Hong Kong, SAR, China
VLADIMIR KHRISTOV  Section on Epithelial and Retinal Physiology and Disease, National Eye
Institute, National Institutes of Health, Bethesda, MD, USA
CAY M. KIELTY  Wellcome Trust Centre for Cell-Matrix Research, Faculty of Life Biology,
Medicine and Health, University of Manchester, Manchester, UK
WITOLD W. KILARSKI  Institute for Molecular Engineering, The University of Chicago,
Chicago, IL, USA
THOMAS KISLINGER  Department of Medical Biophysics, University of Toronto, Toronto, ON,
Canada
CHI-WING KONG  Stem Cell & Regenerative Medicine Consortium, Department of
Paediatrics and Adolescent Medicine, School of Biomedical Sciences, LKS Faculty of
Medicine, The University of Hong Kong, Hong Kong, SAR, China
KWOK-ON LAI  School of Biomedical Sciences, LKS Faculty of Medicine, State Key Laboratory
of Brain and Cognitive Sciences, The University of Hong Kong, Hong Kong, SAR, China
RONALD A. LI  Dr. Li Dak-Sum Research Centre, University of Hong Kong, Hong Kong,
SAR, China; Ming-Wai Lau Centre for Reparative Medicine, Karolinska Institutet,
Stockholm, Sweden
MOSTAFA LOTFI  Section on Epithelial and Retinal Physiology and Disease, National Eye
Institute, National Institutes of Health, Bethesda, MD, USA
LAWRENCE G. LUM  Division of Hematology/Oncology, Department of Medicine, University
of Virginia Cancer Center, Charlottesville, VA, USA
ARVYDAS MAMINISHKIS  Section on Epithelial and Retinal Physiology and Disease, National
Eye Institute, National Institutes of Health, Bethesda, MD, USA
ADRIANO MARCHESE  Department of Biochemistry, Medical College of Wisconsin, Milwaukee,
WI, USA
TERESA MASSAM-WU  Wellcome Trust Centre for Cell-Matrix Research, Faculty of Life
Biology, Medicine and Health, University of Manchester, Manchester, UK
SANDEEP MITTAL  Department of Neurosurgery, Karmanos Cancer Institute, Wayne State
University, Detroit, MI, USA; Department of Oncology, Karmanos Cancer Institute,
Wayne State University, Detroit, MI, USA
HERCULES MOURA  Division of Laboratory Sciences, National Center for Environmental
Health, Centers for Disease Control and Prevention, Chamblee, GA, USA
AMANDA M. NEVINS  Department of Biochemistry, Medical College of Wisconsin, Milwaukee,
WI, USA
LINDA J. OLSON  Department of Biochemistry, Medical College of Wisconsin, Milwaukee,
WI, USA
SIMONA PRINCIPE  Princess Margaret Cancer Centre, Toronto, ON, Canada
JON REES  Division of Laboratory Sciences, National Center for Environmental Health,
Centers for Disease Control and Prevention, Chamblee, GA, USA
DANIEL H. REICH  Department of Physics and Astronomy, Johns Hopkins University,
Baltimore, MD, USA
DANIEL R. RIORDON  Laboratory of Cardiovascular Sciences, National Institute on Aging,
National Institutes of Health, Baltimore, MD, USA
MANUEL J. RODRÍGUEZ-ORTEGA  Departamento de Bioquı́mica y Biologı́a Molecular,
Universidad de Cordoba, Cordoba, Spain
 Contributors xi

REBECCA L. ROSE  Department of Chemistry and Biochemistry, University of Maryland,

College Park, MD, USA
DAVID M. SCHIELTZ  Division of Laboratory Sciences, National Center for Environmental
Health, Centers for Disease Control and Prevention, Chamblee, GA, USA
RUCHI SHARMA  Unit on Ocular and Stem Cell Translational Research, National Eye
Institute, National Institute of Health, Bethesda, MD, USA
ANKIT SINHA  Department of Medical Biophysics, University of Toronto, Toronto, ON,
Canada; Princess Margaret Cancer Centre, Toronto, ON, Canada
ARCHANA THAKUR  Division of Hematology/Oncology, Department of Medicine, University
of Virginia Cancer Center, Charlottesville, VA, USA
YUNG WUI TJONG  The HKU School of Professional and Continuing Education, Po Leung
Kuk Stanley Ho Community College, Hong Kong, China
GORDON F. TOMASELLI  Division of Cardiology, Department of Medicine, Johns Hopkins
University, Baltimore, MD, USA
MATTHEW WAAS  Department of Biochemistry, Medical College of Wisconsin, Milwaukee,
WI, USA
QIN WAN  Section on Epithelial and Retinal Physiology and Disease, National Eye Institute,
National Institutes of Health, Bethesda, MD, USA
JUNWEN WANG  Centre of Genomics Sciences, LKS Faculty of Medicine, The University of
Hong Kong, Hong Kong, China; Department of Health Sciences Research, Center for
Individualized Medicine, Mayo Clinic, Scottsdale, AZ, USA; Department of Biomedical
Informatics, Arizona State University, Scottsdale, AZ, USA
PANWEN WANG  Department of Health Sciences Research, Center for Individualized
Medicine, Mayo Clinic, Scottsdale, AZ, USA; Department of Biomedical Informatics,
Arizona State University, Scottsdale, AZ, USA
ZHIHUI WENG  School of Biomedical Sciences, LKS Faculty of Medicine, The University of
Hong Kong, Hong Kong, SAR, China
ROLIERIA WEST-DEADWYLER  Division of Bacterial Diseases, National Center for
Immunizations and Respiratory Diseases, Centers for Disease Control and Prevention,
Atlanta, GA, USA
JENNIFER WHITMON  Division of Bacterial Diseases, National Center for Immunizations
and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, GA, USA
YULANDA M. WILLIAMSON  Division of Laboratory Sciences, National Center for
Environmental Health, Centers for Disease Control and Prevention, Chamblee, GA, USA
ADRIAN R. WOOLFITT  Division of Laboratory Sciences, National Center for Environmental
Health, Centers for Disease Control and Prevention, Chamblee, GA, USA
STANLEY CHUN MING WU  School of Biomedical Sciences, LKS Faculty of Medicine,
The University of Hong Kong, Hong Kong, SAR, China
BIN YAN  School of Biomedical Sciences, LKS Faculty of Medicine, The University of Hong
Kong, Hong Kong, SAR, China; Centre of Genomics Sciences, LKS Faculty of Medicine,
The University of Hong Kong, Hong Kong, China
XIAOQIANG YAO  Li Ka Shing Institute of Health Sciences, School of Biomedical Sciences,
Chinese University of Hong Kong, Hong Kong, SAR, China
 Part I

Discovery-Based Approaches to Surfaceome Content

 Chapter 1

Surfaceome Analysis Protocol for the Identification of Novel

Bordetella pertussis Antigens
Yulanda M. Williamson, Jennifer Whitmon, Rolieria West-Deadwyler,
Hercules Moura, Adrian R. Woolfitt, Jon Rees, David M. Schieltz,
and John R. Barr

Abstract
The bacterial surfaceome, comprising outer membrane-sorted and/or associated (i.e., cell transporters),
cell surface-exposed (i.e., adhesins) and extracellularly secreted proteins (i.e., toxins), has been character-
ized in bacterial pathogens, such as Bordetella pertussis (Bp) to provide information for use in development
of diagnostic and prevention strategies. This protein subset has clinical significance, as these bacterial
proteins are often associated with attachment to host cells, microbial pathogenesis and antibody-mediated
immunity. Here we describe classical surface membrane protein enrichment techniques, followed by
proteomic methodologies, such as gel-free protein separation and antibody-affinity capture technologies
in combination with nano-liquid chromatography mass spectrometry, for the identification and characteri-
zation of Bp surfaceome proteins.

Key words Carbonate extraction, Membrane proteins, Sample solubilization, Antibody affinity,
Immunoprecipitation, Gel-free, Mass spectrometry, nLC-MS/MS, Bioinformatics

1 Introduction

Bordetella pertussis (Bp) is the causative agent of pertussis (whoop-

ing cough), a highly communicable respiratory infection [1]. Char-
acterization of the Bp surfaceomes [2–8], a subset of proteins
localized, sorted and either exposed on the bacterial surface or
ultimately secreted into the extracellular milieu, is important.
Many of these “surface” proteins are known to be biologically
associated with microbial pathogenesis and humoral immune
response [4, 5]. Once characterized at the gene and protein level,
they could serve as biomarkers for strain differentiation and rapid
microbial detection of clinically relevant matrices. Even more, the
proteins could be candidates for therapeutic targets and novel/or
improved vaccine development.

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_1, © Springer Science+Business Media, LLC 2018

3
4 Yulanda M. Williamson et al.

Traditional tools such as, sodium dodecyl sulfate–polyacryla-

mide gel electrophoresis (SDS-PAGE) have been used through the
years to separate and resolve proteins based on size [9]. This
approach is valuable as it provides a visible snapshot of protein
integrity and separation over a wide molecular weight range, but
for protein identification, requires protein band excision, in-gel
protein extraction and enzymatic digestion, followed by mass spec-
trometry (MS) and database search analysis. To determine those
proteins associated with immunity, additional steps subsequent to
SDS-PAGE must be inserted. These include protein transfer to
membranes, detection with a specific antibody or immune serum,
followed finally by digestion and MS analysis. These traditional
methods or approaches are lengthy and arduous. Conversely,
newer methods which encompass gel-free enzymatic digestions of
the total cellular proteome or enriched subproteome (i.e., surfa-
ceome) in parallel with MS can achieve the same objective (protein
identification), and are less labor intensive. Additionally, contem-
porary advanced antibody-affinity magnetic bead capture technol-
ogies can be utilized to pull-down proteins specifically associated
with immunoreactivity, followed by enzymatic digestion and MS
for putative antigen identification. Protein complexes with direct or
indirect biological association with the pull-downed antigen can be
examined as well. Enriched outer membrane protein extraction,
gel-free direct protein enzymatic cleavage and peptide separation,
immunoproteome profiling, mass spectrometric approaches and
bioinformatics are important methodologies, technologies and
analytical tools which we have used to examine bacterial surfa-
ceomes and will be highlighted in this chapter.
In this chapter, we describe methods for the enrichment of
outer membranes (3.1–3.2), antibody affinity capture (3.3–3.4)
and outer protein membrane identification (3.5–3.6) in Subhead-
ings 3.1 through 3.6. Enriched membrane fractions (EMF)
(3.1–3.2) are obtained and the proteins solubilized using the car-
bonate extraction method based on Molloy et al. [10] with some
modifications. This procedure results in samples containing not
only membrane proteins but also cytosolic and ribosomal proteins.
For gel-free protein identification, EMF or EMF-antibody bead
complexes are denatured with an acid-cleavable detergent, and
enzymatically digested. The tryptic peptides are chromatographi-
cally (LC) separated based on hydrophobicity and peptide elution
at a specific retention time followed by electrospray ionization
tandem mass spectrometry (ESI-MS/MS). Mass spectrum data
sets are searched against a Mascot Bp protein database library,
followed by Scaffold data packaging in which identified proteins
are scored a percent probability value. Predicted subcellular locali-
zations (PSORTb) and gene function associations (Kegg identi-
fiers) can also be performed for identified proteins.
 Identification of Novel Bordetella pertussis Antigens 5

Immune profiling assesses the putative antigenic potential of

proteins. Indirect antibody affinity magnetic bead capture technol-
ogies are a tool used to achieve this goal. Alternatively as a direct
capture approach, protein G Dynabeads® can be conjugated with
antibodies, cross-linked, and subsequently complexed with Bp T
EMF. Rabbit polyclonal antisera generated against Bp T was used to
capture known and novel Bordetella species antigenic proteins. An
indirect immunocapture method was implemented, first, incubat-
ing rabbit anti-Bp T antibodies with Bp T EMF, followed by the
addition of these Bp antibody–antigen/protein complexes with
protein G Dynabeads®. Variable nonionic detergent concentra-
tions were used as a means to enhance the immunoprecipitated
complex specificity between the antibody (ies) and the captured Bp
T EMF antigenic protein(s).
The methodologies presented to identify outer protein mem-
branes (Subheadings 3.5 and 3.6) are related to reverse phase liquid
chromatography and mass spectrometry. These technologies are
intended for individuals with mid-level expertise in the fields of
separation science and analytical chemistry. However, low or no
previous instrument expertise or experience should deter the user.
Communication with the instrument manufacturer or vendor is a
fruitful resource for assistance with instrument set-up, method
development and trouble-shooting. Additionally, instrument man-
ufacturers periodically offer free webinars highlighting instrument
use and method development, intended for novice or experts.
Lastly, a core proteomics and mass spectrometry facility at the
researchers’ institution would also be an excellent source for tech-
nical insight.

2 Materials

2.1 Enriched Outer 1. Bordet-Gengou agar plates (Becton, Dickinson and Company,
Membrane Isolation Sparks, MD, USA).
Method 2. Stainer-Scholte media (SS- or SSM) (see Note 1): To prepare
1 L of SSM, add the following to
900 mL of water: 10 g
casamino acids, 10.7 g L-glutamic acid (Na salt), 0.24 g L-
proline, 2.5 g NaCl, 0.5 g KH2PO4, 0.2 g KCl, 0.1 g
MgCl2·6H2O, 0.02 g CaCl2, 6.1 g Tris·Cl and 2,6-o-
dimethyl-beta-cyclodextrin (heptakis). Adjust the pH to 7.6
using concentrated HCl, and next add water up to 1 L (broth
should be light yellow in color). Sterilize by autoclaving and
dispense 100 mL aliquots into 1000 mL screw-capped flasks.
The media can be stored up to a month at room temperature.
3. Carbonate buffer solution: ice-cold 100 mM sodium carbonate
(pH 11) (see Note 2).
4. Wash solution: 50 mM Tris–HCl buffer (pH 8) (see Note 3).
6 Yulanda M. Williamson et al.

5. Membrane solution buffer: 7 M urea, 2 M thiourea, 2% w/v

3-((3-cholamidopropyl) dimethylammonio)-1-propanesulfo-
nate) (CHAPS), 10% isopropanol, protease inhibitor mix
(10 μL/mL), 100 mM 1, 4-dithiothreitol (DTT) (see Note
4): A 25 mL stock solution should be made. Add approxi-
mately 12 mL of water (urea takes up about 50% of the vol-
ume), then add 10.5 g of urea sparingly, stir until dissolved.
Gently heat the urea and water solution but do not exceed
30  C. Next add 3.8 g of thiourea, 2% CHAPS, 10% isopropa-
nol, protease inhibitor mix and DTT which are fully compatible
with urea. The solution should be made up to 25 mL and
stored at 20  C in 2.5 mL aliquots.
6. Protease Inhibitor Mix: 100 solution (Sigma-Aldrich Chem-
ical Company, St. Louis, MO, USA) (see Note 5).
7. Benzonase: Benzonase® Nuclease, ultrapure 250 unit/μL
(Sigma-Aldrich Chemical Company).
8. Incubator with shaker similar to New Brunswick™ S41i
(Eppendorf, Hauppauge, NY, USA).
9. French press (similar to FA-078A Thermo Electron, Waltham,
MA, USA) (see Note 6).
10. Ultracentrifuge with rotor and tubes capable of reaching
115,000  g (Beckman Type 70.1 Ti or equivalent, Beckman
Coulter, Inc., Indianapolis, IN, USA).
11. 2-D Quantification Kit (GE Healthcare Life Sciences, Marlbor-
ough, MA, USA).
12. 100 mL and 1000 mL screw capped flasks (Fisher Scientific,
Pittsburgh, PA, USA).
13. 50 mL screw-capped polypropylene tubes (Fisher Scientific).
14. 1.5 mL Eppendorf™ Snap Cap Microcentrifuge Tubes.

2.2 Antibody Affinity 1. New Zealand White Rabbits.

Capture Technology 2. BD Vacutainer with K2EDTA (Fisher Scientific).
3. BD General Use 25 Gauge Hypodermic Needle (Fisher
Scientific).
4. BD Disposable Syringes with Luer-Lok™ Tips (Fisher
Scientific).
5. Purification of rabbit sera: Melon Gel® IgG Purification Kit
(Thermo Scientific, Rockford, IL, USA).
6. Determination of IgG concentrations: Easy-Titer® IgG Assay
Kit (Thermo Scientific).
7. Protein G Dynabeads® (Life Technologies, Grand Island, NY).
8. Phosphate Buffered Saline (PBS), pH 7.2.
9. Nonidet P-40 (NP-40) (Sigma-Aldrich Chemical Company).
 Identification of Novel Bordetella pertussis Antigens 7

10. Phosphate citrate buffer (PCB) (Product number: P4809-

100TAB—Sigma-Aldrich Chemical Company). Prepare
according to the manufacturer’s instructions.
11. Hula mixer® (Life Technologies).
12. DynaMag™-2 Magnet (Life Technologies).

2.3 Outer Membrane 1. Trypsin (modified sequencing grade) (Promega Corporation,

Protein Identification Madison, WI, USA).
2. Rapigest (RG): 0.1% rapigest (acid-cleavable denaturant)
(Waters Corporation, Milford, MA, USA) prepared in 50 mM
ammonium bicarbonate and 1 mM calcium chloride (pH 10).
3. 50 mM ammonium bicarbonate, 1 mM calcium chloride
(pH 10). Prepare a 1 L solution based on the following:
50 mM (0.050 M—mol/L) ammonium bicarbonate, (NH4)
HCO3 (MW ¼ 79.06 g/mol) (add 4 g) and 1 mM (0.001 M—
mol/L) calcium chloride, CaCl2 (MW ¼ 110.98 g/mol) (add
0.1 g). Add the appropriate amounts to ~900 mL of distilled
water, adjust to pH 10 with concentrated ammonium hydrox-
ide and bring the volume up to 1 L (see Note 7).
4. 1 M HCl (Sigma-Aldrich Chemical Company).
5. 0.1% formic acid in water (a), 50% acetonitrile solution
(b) (Product number: (a) LS118–4 (b) A955–4) (Fisher Scien-
tific) (see Note 8).
6. Tabletop microcentrifuge rotating up to 10,000  g.
7. Refrigerated CentriVap concentrator (Labconco, Kansas City,
MO, USA).
8. Incubator (see Note 9).
9. Capped vials (to be placed in an LC-autosampler).
10. Nano liquid chromatography (nLC) system.
11. Electrospray ionization (ESI) tandem mass spectrometer (MS).
12. C18 packed silica capillary column (see Note 10).

3 Methods

3.1 Enriched Outer 1. Grow Bp Tohama I (T) strain on Bordet-Gengou agar and then
Membrane Isolation subculture into 1000 mL screw-capped flasks containing
Method: Cell 100 mL of SS-liquid media (see Note 1).
Harvesting 2. Cultures were maintained at 36  C with shaking (200 rpm)
until the OD650nm reaches 0.5–1.0.
3. Collect Bp bacterial cells from 100 mL of liquid culture by
centrifugation at 8000  g for 30 min at 4  C.
8 Yulanda M. Williamson et al.

4. Discard the media and resuspend the cells in 5 mL of 50 mM

Tris–HCl buffer (pH 8.0) with gentle pipetting.
5. Collect the cells by centrifugation for 20 min at 8000  g.
6. Discard the media and repeat wash and centrifugation steps.
7. After the second wash, immediately add 100 μL of protease
inhibitor mix and 1 μL of benzonase to the 5 mL cell mixture,
resuspended in 5 mL of 50 mM Tris–HCl buffer (pH 8.0).
8. Rupture the cells by one passage through a cold French press at
20,000 psi (see Note 6).
9. Discard any unbroken cells by centrifugation at 8000  g for
20 min at 4  C.
10. Keep the supernatant for carbonate extraction (see Note 11).

3.2 Carbonate 1. Add the supernatant directly to 50 mL of ice cold carbonate

Extraction buffer solution. Slowly stir the solution for 2 h at 4  C (see
Note 12).
2. The cell membranes are collected by ultracentrifugation of the
carbonate extraction solution using a Beckman Type 70.1 Ti
rotor at 115,000  g for 1 h at 4  C.
3. Discard the supernatant and add 20 μL of protease inhibitor
mix to each membrane pellet.
4. Wash each membrane pellet in 5 mL of the wash solution by
gentle pipette mixing with a 10 mL pipette (see Note 13).
5. Collect the membrane pellet by ultracentrifugation as
described in step 2 above.
6. Collect washed membranes and transfer them to 1.5 mL
Eppendorf™ tubes with a 10 mL pipette.
7. Solubilize the membranes for 1 h in 500 μL of membrane
solution buffer (see Note 14).
8. Determine the protein concentration using a 2-D Quantifica-
tion kit (see Note 15).

3.3 Antibody Affinity 1. All experiments are performed in accordance to project specific
Capture Technology: animal protocol # 1642, “Production of Antibodies to Borde-
Preparation tella pertussis in New Zealand White Rabbits” approved by the
and Purification Institutional Animal Care and Use Committee (IACUC), Cen-
of Rabbit Anti-Bp ters for Disease Control and Prevention, Atlanta,
Immune Sera Georgia, USA.
2. For the antibody affinity capture technology, carry out all
procedures at room temperature unless otherwise specified.
3. Immunize New Zealand white rabbits, intraperitoneally, with
1  107 cfu Bp T strain; administer a comparable dosage every
2 weeks for 6 weeks (see Note 16).
 Identification of Novel Bordetella pertussis Antigens 9

4. Bleeds are performed via a rabbit ear artery. The sampling site is
disinfected (marginal ear artery area is wiped with an alcohol
swab), and a 25 gauge butterfly needle-capped syringe (5 cc)
inserted into the artery as distally (toward the tip of the ear) as
possible.
5. Approximately 25 mL of blood is drawn from each rabbit
biweekly and dispensed in a K2EDTA-containing vacutainer.
6. Blood was collected before the first immunization (pre-bleed)
representing the normal rabbit sera negative control (NRS) and
after the third immunization for the rabbit anti-Bp immune
sera experiments (see Note 16).
7. After the third immunization, collect all remaining blood (via
exsanguination) from rabbits then euthanize and sacrifice (see
Note 17).
8. Prepare Bp immune sera by centrifuging vacutainers containing
blood into the Beckman Coulter, Inc. centrifuge for 20 min at
1200  g (see Note 18).
9. Purify rabbit anti-Bp immune sera using the Melon Gel® IgG
purification kit.
10. Determine the IgG concentrations using Easy-Titer® IgG
assay kit (see Note 19).

3.4 Rabbit Immuno- 1. Determine the optimal concentrations of antigen (Ag) and
precipitation (IP) antibody (Ab) for rabbit IP (see Note 20).
2. In an Eppendorf tube, combine purified IgG (2 μg) rabbit
antisera (Ab) with Bp T EMF (80 μg) (Ag) then adjust to a
final volume of 200 μL using x% NP40/PBS solution.
(x ¼ 0.22%, 0.45%, or 0.90% v/v). The detergent concentra-
tion should be optimized (see Note 21), and the concentration
that yields the largest quantity of target proteins should be used
for final data analysis.
3. Prepare two Eppendorf™ tubes to be used as negative controls
in a similar manner.
4. Combine purified NRS (2 μg) and Bp T EMF (80 μg) in one
tube and place Bp T EMF (80 μg) alone in the second tube (see
Note 22).
5. Place all samples (Bp T immune complex, NRS, Bp T EMF
only) on a Hula mixer® and allow them to rotate at 10 rpm at
room temperature for 1.5 h (see Note 23).
6. During the 1.5 h incubation period, wash three aliquots con-
taining 200 μL Protein G Dynabeads® (per sample, ~4  108
beads) with PCB. Resuspend each aliquot in 400 μL of PCB,
place each sample on the magnetic rack, and remove the super-
natant. Repeat 2. After the final wash, leave the bound Dyna-
beads® on the magnetic rack.
10 Yulanda M. Williamson et al.

7. After the 1.5 h incubation, add the Ag-Ab complexes to 200 μL

of bound Dynabeads® then resuspend the mixture with a
1000 mL pipette.
8. Draw up the mixture with the 1000 mL pipette, release back
into the tube then repeat twice.
9. Incubate for 1 h at room temperature on Hula Mixer® and
allow them to rotate at 10 rpm.
10. Begin NP40 detergent washes on all Ag-Ab complex bound-
bead samples. Place bead bound Ag-Ab sample in 200 μL x%
NP40/PBS detergent (use the same detergent from step 1).
11. Place each sample on the Hula Mixer® for 5 min, put them on
the magnetic rack, and remove the supernatant.
12. Repeat 2. After the final wash, leave the bound Dynabeads®
on the magnetic rack.
13. Perform two additional washes with 100 μL 0.01 M PBS,
pH 7.2 (using magnetic rack in the same way as described in
steps 4 and 6). Then resuspend the Ag-Ab complex bound-
bead samples in 100 μL of PBS (see Note 24).
14. The rabbit IP rapid method coupled with nLC-ESI MS/MS is
shown in Fig. 1 and the final method is also published as a
Technical Note [7].
15. The technique can be approached using the direct or indirect
immunocapture method (Fig. 2).

3.5 Outer Membrane 1. EMF (10 μg) or EMF-antibody (IP) complexes (50 μL) are
Protein Identification: treated with 0.1% RG (10 μL) and heated at 100  C for 5 min
Protein Digestion and cooled at 4  C for 5 min.
2. After incubation, the protein containing tubes are briefly spun
at 10,000  g. The denatured proteins are digested with
sequence-grade trypsin (10 μg) and incubated at 37  C over-
night (18 h) (see Note 9).
3. To inactivate the RG, 1 M HCl (final concentration 175 mM)
is added to each tube and incubated for 30 min at 37  C,
followed by centrifugation at 10,000  g (15 min) (see Note
25).
4. The supernatant containing antibody-captured EMF tryptic
peptides is transferred to a fresh Eppendorf tube and dried
(see Note 26) to concentrate the peptide pool.
5. Equal volumes of tryptic peptides and a 0.1% formic acid, 50%
acetonitrile solution are transferred to vials, capped and stored
at 70  C until needed (see Note 27).
 Identification of Novel Bordetella pertussis Antigens 11

Rabbit IP method coupled with nLC-ESI MS/MS workflow

= Antigen (Ag)

= Antibody (Ab) ELUTE

= Protein G DynaBeads®

Ag and Ab bind to form Protein G DynaBeads® Ag-Ab complexes bind to

Immune (Ag-Ab) complexes Protein G DynaBeads®
+ x% NP40

>95% PROTEIN IDENTIFICATION

TRYPSIN
DIGESTION
KEGG & PSORTb SCAFFOLD ANALYSIS

BIOINFORMATICS

Peptides analyzed via

LTQ-Orbitrap Mass Spectrometer

Fig. 1 Rabbit immunoprecipitation (IP) method coupled with nLC-ESI MS/MS workflow. Schematic of how IP
sample was isolated then analyzed by mass spectrometry, scaffold analysis, KEGG, and PSORTb

Fig. 2 Direct and indirect IP methods. Invitrogen™ Protein G Dynabeads® includes a Direct and an Indirect IP
method: In the direct method, the antibody and Protein G Dynabeads® are incubated, cross-linked, then the
antigen is added. In the indirect method, the antigen and antibody are added, incubated, and the Protein G
Dynabeads® are added later
12 Yulanda M. Williamson et al.

3.6 Liquid 1. Protein tryptic digests are loaded on a C18 packed silica capil-
Chromatography, ESI lary column and separated using a nanoflow reverse phase
Tandem Mass gradient at 400 nL/min (see Notes 28 and 29) [7, 8].
Spectrometry, 2. The peptides are separated by the reverse phase column and
and Bioinformatics ionized before being transferred to the mass spectrometer, for
MS and MS/MS acquisition (see Note 30) [7, 8].
3. For protein identification, mass spectrum raw files are extracted
and searched using a database search algorithm using specified
parameters (see Note 31).
4. MS/MS peptide and protein identifications are validated using
Scaffold (see Note 32). See Fig. 3 for a visual example of an
MS/MS spectrum profile [13, 14].
5. Bioinformatic tools can be further utilized to categorize iden-
tified proteins based on gene ontology/function or predicted
subcellular localization (see Note 33) [6, 15].

4 Notes

1. SSM is prepared according to the method based on Hulbert

(2009) [11] with minor modifications. For instance, 1 g of
heptakis is added to the original SS-broth components not as a
separate supplement. Also, prior to inoculating the broth,
1 mL of a 100 SS-supplement is added to our media stocks;
100 mL in 1000 mL screw-capped flasks. Additionally, the
large flasks are needed to provide adequate aeration for Bp
cultures to grow. One batch is grown to yield the desired
amount of protein. For 10 mL of SS-supplement, add the
following to a 15-mL conical tube: 40 mg L-cysteine, 10 mg
FeSO4·7H2O, 4.0 mg niacin (nicotinic acid), 150 mg glutathi-
one, 0.4 g ascorbic acid. Bring to 10 mL with water and vortex
well to mix. Filter-sterilize using a 0.22-μm filter and dispense
into 1.5-mL tubes and store up to 1 year at 20  C. Prior to
inoculating SS-broth, add 10 μL of 100 supplements per
1 mL broth. Of note, safety procedures should be taken when
handling large volumes of bacterial cultures.
2. The amount of carbonate stock solution prepared is 100 mL
and the pH is approximately 11. There is no need to adjust the
pH of the solution by titration. However, the solution must be
kept at 4  C for at least 1 h before use.
3. The approximate volume to prepare for the wash solution is
1000 mL. However, more buffer can be prepared based on the
number of bacterial cultures that will need to be washed.
4. Please make the membrane solution buffer in advance of your
work as it takes a long time to prepare. This solution contains
chemicals that are hazardous. Please review the safety data
 Identification of Novel Bordetella pertussis Antigens 13

Fig. 3 Sequence of a known Bordetella membrane surface protein, virulence factor and immunogen,
filamentous hemagglutinin (FHA); and the scaffold mass spectrum b and y ion profiles for the tryptic peptide
14 Yulanda M. Williamson et al.

sheets (SDS) for these chemicals prior to making up the solu-

tions and use appropriate PPE. Hazardous materials should be
weighed in a chemical fume hood and materials disposed of
according to appropriate regulations.
5. Protease inhibitor mix is used throughout the procedure in the
membrane solution buffer and added directly to the bacterial
cells (along with benzonase) after the wash step and before the
cell membranes are solubilized. This substance inhibits general
degradation of proteins in cell extracts by endogenous
proteases.
6. The French pressure call should be prechilled at 4  C. The
bacterial cells to be applied to the apparatus should be kept
on ice. After adding the sample, and bringing the cell under the
desired pressure, adjust the outlet flow rate to about one drop
per second. Collect the cell lysate in a 50 mL conical tube
on ice.
7. Refer to the SDS for chemical hazard information and wear
appropriate personal protective equipment (PPE) as a safety
precaution.
8. Prepare a 1 L solution based on a 1 to 1 ratio: 500 mL 0.1%
formic acid in water plus 500 mL 100% acetonitrile. Refer to
the SDS for chemical hazard information and wear appropriate
PPE as a safety precaution.
9. Any type of incubator set at the specified temperature is accept-
able. Alternatively, a PCR machine programmed at the appro-
priate temperature and duration can be used as an incubation
source.
10. The C18 column used for this proteomic analysis was prepared
in-house in which fused silica capillary columns were packed
with reverse phase packing material. Alternatively, prepacked
C18 columns can be purchased from various vendors (manu-
facturers) of your choice.
11. The pellet is normally discarded at this point. During the early
stages of performing the carbonate extraction method, it may
be a good idea to save the pellet at 4  C until the extraction
method and protein quantification is completed to determine if
another passage with the French press needs to be repeated due
to low protein yield. After the final centrifugation, after pas-
saging the cells with the French press, in order to remove any
ä

Fig. 3 (continued) GALALDGGAGVALQSAK digested from this protein. (a) Yellow highlights represent addi-
tional FHA MS/MS detected tryptic peptides identified within the FHA amino acid sequence (partial sequence
snapshot—amino acid 1–1150). Green highlights in the sequence represent deamidation of the amino acid—
glutamine (Q). MS/MS FHA detected peptide/protein identification corresponding to (b) Gel-free Bp T EMF and
(c) Indirect Bp T EMF-IP capture
 Identification of Novel Bordetella pertussis Antigens 15

unbroken cells, immediately proceed to the carbonate extrac-

tion step by adding the supernatant to the ice-cold sodium
carbonate solution.
12. The solution was directly poured into 100 mL screw-capped
flasks and slowly stirred on a stir plate placed in a cold room for
2 h. Using an ice bath for 1 h may also be a sufficient method
for this step. There are two stopping points in the procedure.
After the carbonate extraction step is completed, the ice-cold
carbonate solution can be transferred to Beckman centrifuge
tubes and stored overnight at 4  C before proceeding to the
ultracentrifugation step. The second stopping point is after the
ultracentrifugation step in which membranes are collected.
Membranes can be stored in minimal Tris–HCl buffer between
200 and 500 μL with liquid covering the membrane pellet.
13. Gently resuspend the membrane pellets in wash solution. The
wash steps are important because they remove the excess
sodium carbonate that may interfere with downstream applica-
tions. After washing, samples can be stored in minimal
Tris–HCl buffer overnight at 20  C.
14. The solubilization step may take longer than 1 h depending on
the type of bacteria used. To aid in the solubilization process,
the membranes may have to be split into two tubes. Addition-
ally, if the membranes are difficult to solubilize, the tube may
have to be gently swirled or rotated. Gentle heating, not to
exceed 30  C, can also be used. Temperatures above 30  C may
be detrimental to the urea in the membrane solution buffer.
15. The protein concentration needs to be determined using a
protein assay kit that is compatible with the reagents in the
solubilization buffer, such as the 2-D Quantification Kit. The
carbonate protein extraction method yields a protein concen-
tration of 5–20 μg/μL per 100 mL of bacterial culture. In
order to remove residual salts prior to performing downstream
applications, desalt the membrane proteins with spin columns,
such as Pierce spin columns (Thermo Fisher Scientific, Pitts-
burgh, PA, USA). Aliquots of 100 μL can be stored at 80  C
until further use.
16. Rabbit anti-Bp immune sera are produced using irradiated
bacteria. In order to preserve bacterial surface structures
while blocking bacterial replication and infectivity, the Bp T
strain is irradiated using 5  106 rad of gamma irradiation.
Rabbits are bled from the marginal ear artery using a 25 gauge
butterfly needle and 5 cc syringe. Acceptable limits are a 7.5%
collection of the rabbit’s total blood volume weekly. Multiple
bleeds should not exceed 3 per week and no more than 25 mL
per week (single bleed of no more than 10 mL). Multiple
bleeds should not exceed the single sample limit of 1% daily,
16 Yulanda M. Williamson et al.

7.5% weekly, or 15% monthly. If multiple bleeds occur using a

single week approach, with the single sample limit of 15%, then
the animal should be allowed to recover for a 30-day recovery
period. Please ensure the rabbits weigh at least 8 lb or more as
weight can be used to determine the total percentage of blood
loss. Furthermore, according to IACUC (CDC), the estimated
circulating blood volume (CBV) of rabbits (R) is ~55 mL/kg.
The following is an example of how we determine a single
sample limit using the equation:

ðCBVÞ  ðweight of R Þ ¼ ðblood volumeÞ  ð%percentage  0:01Þ

55 mL=kg ðCBVÞ  5 kg ðweight of R Þ ¼ 275 mL blood  0:075
ðor 7:5%  0:01Þ ¼ 20:6 mL blood per week

For further guidance pertaining to antibody production using

rabbits (i.e., immunization schedule, blood collection, etc.),
refer to your institutions IACUC or an external animal care
facility.
17. Following the terminal bleed collection using K2EDTA-con-
taining vacutainers (after 6 weeks or if the animals were found
to be moribund—very sick or close to death), the animals were
euthanized.
18. There are several centrifuges that can be used to obtain serum
from blood. Follow the manufacturer’s standard centrifugation
recommendations.
19. Due to the production of polyclonal antibodies, rabbit serum is
highly reactive. Even after several attempts to use diluted rabbit
sera, these IP experiments were not successful until rabbit
antiserum was purified. Importantly, purified IgG and rabbit
sera can be preserved at 4  C for 5–7 days. Therefore, only
purify the amount of IgG that is needed for IP experiments and
store any remaining rabbit antiserum at 20  C or below until
needed for experimentation.
20. In preliminary experiments, variable concentrations of EMF
(Ag) and IgG (Ab) are tested to determine the optimal
Ag-Ab ratios for the IP procedure. Since, all EMF and rabbit
immune sera samples are not exactly the same; the optimal
concentrations may differ and should be determined for each
experiment. Also, use of NP-40 detergent is crucial for success-
ful immunoprecipitation experiments and in achieving benefi-
cial results [12]. Thus, several concentrations of NP-40/PBS
are tested before determining optimal concentration for our
rabbit IP experiment.
21. As stated in West et al. [7], each IP sample should be analyzed
by SDS-PAGE and Western blot. The gels and blots of IP
samples display poor quality and high backgound signal due
 Identification of Novel Bordetella pertussis Antigens 17

to many nonspecific proteins. Since the goal of IP is to isolate

target proteins while minimizing nonspecific binding of pro-
teins, optimization of detergent concentration is critical to
improve the outcome of IP experiments. Strategies from
Yang et al. [12] were utilized to determine optimal concentra-
tion of NP-40/PBS for rabbit IP experiments. According to
Yang et al. [12], IP experiments with higher initial protein
concentrations have an increased tolerance to detergent. Simi-
larly, Yang et al. states that when IP sample contains low
protein concentrations, a low detergent concentration (i.e.,
0.05% NP-40) is helpful to isolate target proteins and minimize
nonspecific bonding. In order to determine the optimal
NP-40/PBS for Rabbit IP, protein amounts are determined
and kept constant. Various concentrations of NP-40/PBS
detergent then are added to each immunoprecipitated sample.
As an example, the following equation was utilized to calculate
the final concentration of a 0.90% NP-40/PBS solution:
C1V1 ¼ C2V2
(10% NP-40 stock detergent)  (V1) ¼ (0.90% NP-40 final
concentration  0.01)  (10 mL PBS  final volume)
(0.10)  (x) ¼ (0.009) (10 mL)
C2 ¼ 0.9 mL or 900 μL of 10% NP-40 detergent
Add 900 μL of 10% NP-40 detergent to 9.1 mL PBS
22. Negative controls are used to account for nonspecific binding
to Dynabeads®.
23. The Hula mixer® is a valuable tool to achieve uniform sample
distribution. It works by 360 rotation, up to 90 tilting, and
up to 5 vibration. For our experiments, the Hula mixer® is set
to the orbital for 10 s at 10 rpm, 45 reciprocal for 5 s, and 5
vibration for 5 s. This setting can be adjusted for each experi-
ment, as necessary. Also, when samples were rotating on the
Hula Mixer®, parafilm is wrapped on the top of the tubes so
that samples would not leak out.
24. This protocol describes an indirect IP method (Fig. 1) which
allows binding of solubilized antigen to antibody prior to
addition of protein binding Protein G Dynabeads® for precip-
itation. This is the method that maximized binding of all
reactants for our rabbit IP experiments but there is a direct
method that can also be used for IP. In the direct method, the
Protein G Dynabeads® beads are added directly to Ag and Ab
(Fig. 2) [7].
25. Prior to RG inactivation, EMF-antibody complexes are mag-
netically stabilized and the clear supernatant containing tryptic
peptides is removed and transferred to a fresh tube.
18 Yulanda M. Williamson et al.

26. After centrifugation a very small white pellet may be present.

Carefully remove the supernatant (step 3) to avoid removal of
the precipitate, as this may hinder chromatographic separation
in downstream steps. The supernatants are dried to completion
(~30 min) using a refrigerated centrivap or a speed vacuum
could be used as an alternative. 20 μL of sterile distilled water is
added to the tubes, which are then resuspended, mixed on a
vortex mixer and briefly (10 s) spun using a tabletop
microcentrifuge.
27. Based on maximal nLC autosampler vial volume, for instance,
10 μL tryptic peptides plus 10 μL 0.1% formic acid, 50%
acetonitrile could be used. Please review chemical safety data
sheets (SDS) prior to use and wear appropriate PPE.
28. Reverse phase nLC is a type of separation involving two phases:
stationary and mobile. The stationary phase comprises a glass
or metal column containing porous material such as silica. In
brief, an injected sample will flow (nL/min) and be transported
through the mobile phase and bind to the column. A gradient
is generally implemented in which at the start of the run, the
mobile phase composition is more aqueous (i.e., 95% water).
As the run proceeds, the amount of organic solvent (i.e.,
acetonitrile) increases. In essence, at the beginning, hydro-
philic tryptic peptides will elute off the column easily, while
hydrophobic peptides interact stronger with the column, and
will require longer periods of time for elution. For instance,
surface-membrane embedded proteins, which are likely more
hydrophobic in nature, will tend to “stick” or be retained to
the column longer and elute off the column later in the run.
29. Solvents used for gradient elution are variable, but suggested as
follows; solvent A: 0.1% formic acid in water, and solvent B:
0.1% formic acid in acetonitrile. Gradient profiles are previ-
ously used are as follows:

Total time (min) Solvent A (%) Solvent B (%)

5 95 5 Hold
100 70 30 Ramp up
5 10 90 Ramp up
2 10 90 Hold
2 95 5 Ramp down
20 95 5 Equilibration

The nLC system utilized is flexible and researcher-dependent,

in which the elution method is transferrable or a comparable
method can be created.
 Identification of Novel Bordetella pertussis Antigens 19

30. Ion source voltages for capillary columns are at flow rates of
300–600 nL/min and are generally between 1.7 and 2.3 kV.
The mass spectrometer was programmed to perform data-
dependent acquisition by scanning the mass range from m/z
400 to 1600 at a nominal resolution setting of 60,000 for
parent ion acquisition. For MS/MS analysis the mass spec-
trometer chose the top 16 most intense ions with two or
more charges. Singly charged ions were rejected for MS/MS
as these ions are likely due to detergents or other sample
additives. The tandem mass spectrometer implemented ulti-
mately is researcher-dependent, but one that incorporates a
hybrid mass spectrometer such as the LTQ-Orbitrap instru-
mental systems is ideal for surfaceome protein identification
and biomarker discovery.
31. Mascot Distiller (data extraction) and Daemon (mass spectrum
matching) (Matrix Science, London, UK; version 2.2.1.0) is
preferred. Suggested parameters for the mascot search include
digestion agent (trypsin), missed cleavages (two); fragment ion
tolerance mass (0.50 Da), precursor ion tolerance (200 ppm),
variable modification (oxidation). The database utilized is
researcher-dependent, but in this instance an NCBI nonredun-
dant database specific for the organism/typed-strain of choice
(Bp) was implemented.
32. Scaffold_4.0.6.1 (or current version) is preferred. Proteome
Software, Portland, OR, USA.
33. PSORTb can be used to detect subcellular localization of Bp
EMF proteins http://www.psort.org/psortb/. Utilize KEGG
identifiers in conjunction with NCBI gi accession numbers to
discover functions of each of the identified proteins.

Acknowledgments

The authors would like to thank Drs. Jacquelyn Sampson, Eddie

Ades, Maria L. Tondella, and George Carlone at the Centers for
Disease Control and Prevention for their insight and providing
materials associated with the described research protocols.
Disclaimer
References in this chapter to any specific commercial products,
process, service, manufacturer, or company do not constitute an
endorsement or a recommendation by the US Government or the
Centers for Disease Control and Prevention. The described proto-
col and suggested notes in this report are those of the authors and
do not necessarily represent the views of CDC.
20 Yulanda M. Williamson et al.

References
1. Bordet J, Gengou O (1906) Le microbe de la (2012) A gel-free proteomic-based method
coqueluche. Ann Inst Pasteur 20:48–68 for the characterization of Bordetella pertussis
2. Hulbert RR, Cotter PA (2009) Laboratory clinical isolates. J Microbiol Methods
maintenance of Bordetella pertussis. Curr Pro- 90(2):119–133
toc Microbiol 15(4B):1.1–1.9. https://doi. 9. Laemmli UK (1970) Cleavage of structural
org/10.1002/9780471729259.mc04b01s15 proteins during the assembly of the head of
3. Parkhill J, Sebaihia M, Preston A, Murphy LD, bacteriophage t4. Nature 227:685–689
Thomson N, Harris DE et al (2003) Compar- 10. Molloy MP (2008) Isolation of bacterial cell
ative analysis of the genome sequences of Bor- proteins using carbonate extraction. Methods
detella pertussis, Bordetella parapertussis and Mol Biol 424:397–401
Bordetella bronchioseptica. Nat Genet 11. Hulbert RR, Cotter PA (2009) Laboratory
35:32–40 maintenance of Bordetella pertussis. Curr Pro-
4. Matoo S, Cherry JD (2005) Molecular patho- toc Microbiol 15:4B.1.1–4B.1.9
genesis, epidemiology, and clinical manifesta- 12. Yang L, Zhang H, Bruce JE (2009) Optimiz-
tions of respiratory infections due to Bordetella ing the detergent concentration conditions for
pertussis and other Bordetella subspecies. Clin immunoprecipitation (IP) coupled with
Microbiol Rev 18(2):326–382 LC-MS/MS identification of interacting pro-
5. Shrivastava R, Miller JF (2009) Virulence fac- teins. Analyst 134:755
tor secretion and translocation by Bordetella 13. Keller A, Nesvizhskii AI, Kolker E, Aebersold R
species. Curr Opin Microbiol 12(1):88–93 (2002) Empirical statistical model to estimate
6. Tefon B, Maass S, Ozcengiz E, Becher D, the accuracy of peptide identifications made by
Hecker M, Ozcengiz G (2011) A comprehen- MS/MS and database search. Anal Chem
sive analysis of Bordetella pertussis surface pro- 74:5383–5392
teome and identification of new immunogenic 14. Nesvizhskii AI, Keller A, Kolker E, Aebersold R
proteins. Vaccine 29(19):3583–3595 (2003) A statistical model for identifying pro-
7. West R, Whitmon J, Williamson YM, teins by tandem mass spectrometry. Anal Chem
Moura H, Nelson M, Melnick N et al (2012) 75:4646–4658
A rapid method for capture and identification 15. Yu NY, Wagner JR, Laird MR, Mellj G, Rey S,
of immunogenic proteins in Bordetella pertussis Lo R et al (2010) PSORTb 3.0: improved
enriched membranes fractions: a fast-track protein subcellular localization prediction
strategy applicable to other microorganisms. J with refined localization subcategories and pre-
Proteome 75(6):1966–1972 dictive capabilities for all prokaryotes. Bioinfor-
8. Williamson YM, Moura H, Simmons K, matics 26(13):1608–1615
Whitmon J, Melnick N, Woolfitt A et al
 Chapter 2

“Shaving” Live Bacterial Cells with Proteases for Proteomic

Analysis of Surface Proteins
Manuel J. Rodrı́guez-Ortega

Abstract
Surface proteins are essential molecules for the interplay between cells and the environment. They partici-
pate in many biological processes including transport, adhesion, cell–cell recognition, signaling, and other
cell interactions. In pathogenic microorganisms, these molecules may act as virulence or cytotoxicity
factors. Analyzing the set of surface proteins is critical to understand these processes and to identify possible
targets that can be the starting point for other studies or discoveries (e.g., vaccines or diagnostics). Here I
describe a proteomic procedure to identify in a fast and reliable way a set of surface-exposed proteins in
bacteria, the methodology of which can be adapted to other biological systems (unicellular fungi, parasites).
The protocol presented here involves “shaving” the cells cultured in broth with proteases followed by liquid
chromatography–tandem mass spectrometry (LC/MS/MS) and analysis of the generated peptides. This
method overcomes some important limitations of the first-generation, gel based proteomics techniques,
and the “shaving” approach allows one to identify which domains from identified proteins are more
accessible to proteases. These identified proteins have the highest potential to be recognized by antibodies,
and thus permits the identification of potential epitopes or antigens.

Key words Surface proteins, “Shaving”, Proteomics, Proteases, Antigens, Epitopes

1 Introduction

Since the mid-1990s, the availability of genome sequences has

made possible the advent of “omics” technologies (i.e., massive
analysis platforms that allow the identification of hundreds or
thousands of biomolecules from one sample). To understand
what occurs at or through biological surfaces (cell membranes,
cell walls, teguments, etc.), proteomic-based approaches provide
invaluable tools capable of providing a snapshot of proteins parti-
cipating in the interaction between cells and their environment.
Once identified, this information helps unravel cell functions and
mechanisms that may involve transport, adhesion, cell–cell recog-
nition, signaling, and others [1]. When studying pathogenic
microorganisms, the analysis of surface proteins is key to the

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_2, © Springer Science+Business Media, LLC 2018

21
22 Manuel J. Rodrı́guez-Ortega

identification of potential targets for drugs or for vaccine or diag-

nostic candidate discovery, as these molecules have the highest
chances to raise an effective immune response [2].
In the first decade of the proteomics era, the study of surface
proteins was mainly approached using gel-based protein separation
followed by matrix-assisted laser desorption/ionization-time of
flight mass spectrometry (MALDI-TOF MS) analysis, which has
been called “first generation proteomics.” This allowed the descrip-
tion of many surface/membrane proteomes. Researchers recog-
nized several important limitations associated with this approach.
There was a general underrepresentation of surface/membrane
proteins in 2-dimension (2-D) gels, mainly due to the fact that
(1) these proteins are generally synthesized in relatively low copy
numbers, compared to other cellular compartments (especially
when compared to cytoplasmic proteins), and (2) many of them
are insoluble, particularly those having transmembrane spanning
domains [3]. In addition, the protein identification from this work-
flow misses information about possible discrepancies/concor-
dances between experimental and predicted topology, which is
very important for projects requiring epitope or antigen descrip-
tions for drug or vaccine discovery.
Second-generation proteomic-based approaches do not require
gels for the identification of hundreds or thousands of proteins/
peptides in a liquid sample, using principally LC/MS/MS plat-
forms. In 2003, a simple and smart strategy was reported that
facilitated the identification of membrane proteins and the topo-
logical characterization of domains on both sides of a biological
membrane [4]. Inspired by this idea, Rodrı́guez-Ortega and cow-
orkers set up a procedure to identify, in a fast and reliable way,
surface proteins of pathogenic bacteria, initially for vaccine discov-
ery purposes. This was applied to the gram-positive bacterium
Streptococcus pyogenes [5]. The initial protocol consists of “shaving”
the surface of live cells cultured in a broth with two different
proteases: trypsin (the enzyme used in >99% of proteomic proto-
cols), which specifically cleaves the C-terminal of arginine and lysine
amino acid residues, and Proteinase K, a broad-spectrum, relatively
nonspecific serine endopeptidase. The latter enzyme allowed the
identification of pilin proteins, which are resistant to trypsin diges-
tion. The mixture of peptides generated (called the “surfaceome”
or “surfome”) is then subject to cleaning and concentration with
chromatographic cartridges to remove salts and sucrose present in
the digestion buffer, followed by LC/MS/MS analysis. An impor-
tant factor is the control of cell lysis, as this helps avoid an excess of
predicted cytoplasmic proteins in the “surfome.” In the initial
procedures, this step was checked by colony forming unit (CFU)
counting, although flow cytometry can also be used to assess the
cell viability [6].
 Protease Shaving of Bacterial Cells 23

The procedure described here allows the identification of the

most protease accessible protein regions, which was not possible
with first-generation proteomics approaches. These “hot zones”
can be used to determine potential epitopes for antibody recogni-
tion or to design potentially exposed polypeptides for vaccine or
even diagnostics projects [7, 8]. The protocol is generally applicable
to gram-positive bacteria; however, it can be applied to bacteria that
are more labile, like the gram negatives, through the modification
of several steps (digestion buffer, protease digestion time, etc.). The
protocol has been successfully applied, as well, to other biological
systems like yeast or parasites (for an extensive review, see [9]).
Additional experimental variations may be used to improve the
procedure for certain species: redigestion of “surfome” fractions
to reduce the presence of protease missed cleavages; the digestion
of large peptides that otherwise would be undetected by the MS
instruments [10, 11]; and the use of immobilized proteases, for
very labile species [12, 13].

2 Materials

Prepare all solutions using ultrapure water and analytical grade

reagents. Prepare and store all reagents at room temperature
(unless indicated otherwise). Sodium azide does not need to be
added to the reagents.

2.1 Shaving Protocol 1. Culture broth (see Note 1).

Components 2. Sterile material for microbiological culture (disposable pip-
ettes, flasks, etc.).
3. Falcon-type tubes (50 mL volume).
4. Wash buffer: prepare phosphate-buffer saline solution (PBS),
pH 7.4. For 1 L solution, add 8.181 g NaCl, 0.2 g KCl, 2.68 g
Na2HPO4, and 0.245 g NaH2PO4 to a 1 L graduated cylinder
or a glass beaker. Dissolve salts in 800 mL of distilled water,
adjust the pH to 7.4 with HCl, and then bring to a final volume
of 1 L. Sterilize it by filtering or autoclaving.
5. Trypsin digestion buffer: PBS/30% sucrose, pH 7.4. Prepare it
by dissolving 30 g sucrose in 80 mL PBS, pH 7.4, and then
bring to a final volume of 100 mL. Sterilize it by filtering (see
Note 2).
6. 1.5 mL low-binding tubes (Eppendorf® Protein LoBind
microcentrifuge tubes).
7. Sequencing grade modified trypsin (Promega, Madison, WI,
USA) (see Note 3). Store trypsin at
20  C.
24 Manuel J. Rodrı́guez-Ortega

8. 0.22-μm pore-size filters (Millex®-GV PVDF 0.22-μm filter

units, 13 mm diameter (EMD Millipore, Billerica, MA, USA)
(see Note 4).
9. Top-down agitation rotor.

2.2 Cleaning 1. Oasis® HLB cartridges, 1 cc (Waters Corporation, Milford,

of the “Surfome” MA, USA).
2. Prepare 10% formic acid by the addition of 10 mL of 100%
formic acid to 100 mL distilled water. Store it in a glass flask.
3. Equilibration solution for Oasis® HLB cartridges: prepare 80%
acetonitrile (HPLC grade) by mixing 80 mL acetonitrile with
20 mL distilled water.
4. Wash solution for Oasis® HLB cartridges: prepare 2% acetoni-
trile (HPLC grade)/0.1% formic acid by mixing 2 mL acetoni-
trile with 97 mL distilled water, and adding 1 mL of 10%
formic acid.
5. Elution solutions for Oasis® HLB cartridges: prepare 10%, 20%
and 50% acetonitrile (HPLC grade)/0.1% formic acid by mix-
ing 10, 20 or 50 mL acetonitrile with 89, 79 or 49 mL distilled
water, respectively, followed by the addition of 1 mL of 10%
formic acid.
6. Speed-vacuum concentrator system.
7. Vacuum manifold and vacuum pump, or 5 mL automatic
pipette.

3 Methods

Be cautious when working with pathogenic microorganisms. Take

appropriate precautions and wear personal protective equipment
(e.g., lab coat, latex gloves, and goggles). Be sure to fulfill the
biological safety standards in terms of protection levels according
to an organism’s risk. The protocol described here is adapted from
the original one described for the first time for Streptococcus
pyogenes [5].

3.1 “Shaving” 1. Grow the bacterial culture to the desired OD600, normally
Protocol for Gram- corresponding to mid-exponential phase (see Notes 5 and 6).
Positive Bacteria 2. Pellet the bacteria normally by centrifugation at 3500  g,
10 min, 4  C.
3. Resuspend the pelleted bacteria in PBS (see Note 7). Repellet
the bacteria as in step 2.
4. Repeat step 3 two more times.
5. Resuspend the bacterial pellet in PBS/30% sucrose, pH 7.4 in a
1.5 mL low-binding tube at a ratio of 800 μL buffer per each
100 mL of initial bacterial culture.
 Protease Shaving of Bacterial Cells 25

6. Add 5 μg/mL of sequencing grade, modified trypsin.

7. Incubate the suspension for 30 min at 37  C with top-down
agitation (see Notes 8 and 9).
8. Pellet the “shaved” bacteria by centrifugation at 3500  g,
10 min, 4  C and recover the supernatant containing the
peptide fraction (“surfome”) in a clean, sterile
low-binding tube.
9. Filter the supernatant (“surfome”) with a 0.22-μm pore-size
filter (see Note 10).
10. Optionally, if the trypsin digestion has not worked well (e.g.,
too many large peptides with many trypsin missed cleavage
sites), redigest the “surfome” with 2 μg trypsin overnight at
37  C with top-down agitation (see Note 11).

3.2 Cleaning Cleaning can be done with a vacuum manifold system or by push-
the “Surfome” ing the liquids through cartridge resin with a 5 mL tip pipette,
with Oasis® HLB using a 5 mL pipette for generating pressure (see Note 12). Accord-
Extraction Cartridges ing to the manufacturer’s instructions, the use of the cartridges
involves the following steps: equilibration, sample loading, wash-
ing, and elution.
1. Equilibrate Oasis® HLB extraction cartridges with 0.6 mL of
80% acetonitrile.
2. Add 0.6 mL of 0.1% formic acid.
3. Load the sample (in our hands, loading 150 μL of the “sur-
fome” of any streptococcal species mixed with 450 μL PBS
works well) (see Note 13).
4. Wash the sample twice with 0.6 mL of 2% acetonitrile/0.1%
formic acid.
5. Elution in three steps with 0.6 mL of each of the following
solutions (see Note 14):
(a) 10% acetonitrile/0.1% formic acid
(b) 20% acetonitrile/0.1% formic acid
(c) 50% acetonitrile/0.1% formic acid
6. Dry in a speed-vacuum system (see Note 15).
7. Resuspend the pellet in 100 μL 2% acetonitrile/0.1% formic
acid (you can divide it among three tubes, or resuspend the first
tube, then transfer the volume to the second one and resus-
pend, and then transfer it to the third one and resuspend).
Keep the sample in a low-binding tube.
8. At this stage, the sample is ready for MS/MS analysis (see
Note 16). Otherwise, it can be stored at
20  C for some
months (see Notes 17 and 18).
26 Manuel J. Rodrı́guez-Ortega

4 Notes

1. The protocol described here is adapted from [5], which can be

generally applied to most gram-positive bacteria. For microor-
ganisms of the Streptococcus genus, it works well using a com-
plex broth like Todd-Hewitt. However, Olaya-Abril et al. [6]
used a chemically defined medium supplemented with ethanol-
amine to reduce cell lysis. The occurrence of cell lysis is a
potentially major drawback of this proteomic procedure. If
the lysis is too extensive, nonspecific peptides may mask the
desired targets of this protocol, i.e., surface-attached proteins
will be hidden and/or underestimated within the vast amount
of peptides in the sample. Choosing the right culture medium
is very important to assure the success of this protocol, and the
medium may vary for each organism. Cell lysis can be assessed
by counting CFUs [5] or by flow cytometry [6], which requires
more complex equipment, but is more precise.
2. As indicated in Note 1, a critical aspect of this protocol is the
control of cell lysis. This can be achieved, in part, through the
use of isotonic buffers when handling microorganisms, espe-
cially at the protease digestion step. For most gram-positive
bacteria, digestion is performed using an isotonic buffer in PBS
with a high concentration of sugars (sucrose, raffinose). The
pH is adjusted to 7.4 to be near the optimal value of the
trypsin. However, if other proteases are used, changes may be
made to the buffer composition and/or the pH. This concept
can be illustrated by experiments with proteinase K, a nonspe-
cific protease. This enzyme has a high turnover number, so if its
activity is not controlled, it may cleave all substrates very rap-
idly. One way to control the activity is either to lower the pH to
reduce its activity or to avoid/reduce buffer Ca2+, which nor-
mally acts as an activator. Rodrı́guez-Ortega et al. [5] deter-
mined empirically that a buffer consisting of PBS/30% sucrose,
pH 6.0, without added Ca2+ works well for a 20–30 min
digestion.
3. As previously described in Note 2, other proteases can be used
according to the purpose and/or protein target(s). If the tar-
gets are recalcitrant, as seen with trypsin-resistant proteins or
proteins with potentially hidden specific cleavage sites, nonspe-
cific proteases like proteinase K can be used. It has also been
described for the use of immobilized enzymes (e.g., trypsin on
agarose beads), which theoretically will reduce cell disruption
because of reduced protease penetration into the cell wall.
Immobilized enzymes also reduce the numbers of generated
peptides, as they have lower turnover numbers [12, 13].
 Protease Shaving of Bacterial Cells 27

4. Filters of 4 mm diameter may be used, but they clog more easily

and must be changed during the filtration operation. This may
lead to loss of some sample amounts.
5. Ideally at the mid-exponential phase, bacteria are in the most
active division phase with minimal cell death. Many gram-
positive bacteria express a lot of surface proteinsduring this
phase. The optimal moment of the growth phase in which
cells are harvested will depend on the organism and/or the
research purpose. The protocol described here is well adapted
for Streptococcus sp. At the mid-exponential phase
(OD600 ¼ 0.25–0.30), the cell concentration is approximately
108 cells/mL. In 100 mL cultures there is enough material for
recovering a high amount of peptides. Nevertheless, it has been
proven with bacteria from 25 mL culture that there is still
sufficient material for protease digestion and peptide recovery.
6. Follow specific safety recommendations when handling patho-
genic microorganisms.
7. Avoid resuspending the pelleted bacteria by pipetting exces-
sively, as this could break the cells. It is better to vortex gently.
8. The digestion time with the protease may influence cell viabil-
ity. If it is too long, it can favor cell lysis. Therefore, it should be
set up empirically. For most gram-positive organisms, 30 min
gives a good yield in terms of peptide recovery, without
compromising significantly cell viability. A study on the effect
of trypsin digestion time on peptide yield and cell lysis is
available in [6].
9. A top-down agitation rotor may be used to improve protease-
cell contact during digestion, especially when using narrow
tubes. By rotating at low rpm, the bacterial suspension will be
in continuous movement within the tube and thus avoid
settling of cells at the bottom of the tube, which would prevent
contact with the protease. The rotor may be placed in a cham-
ber at 37  C.
10. As the digestion volume is normally less than 1 mL, a 1-mL
syringe is recommended to filter the “surfome” fraction. Push
the syringe piston slowly to avoid that the filter is dislodged
from the syringe if it is clogged. This is quite probable if using
4-mm diameter filters. Clogging may be due to an incomplete
removal of cells and subsequent cell contamination of the
supernatant. To avoid this, the “surfome” fraction can be
centrifuged.
11. If the LC/MS/MS analysis contains too many large peptides
with consensus trypsin sites that were not cleaved, the trypsin
added to the solution may have not worked adequately
[10]. Under these circumstances, a redigestion of the primary
“surfome” can be performed. For this, 2 μg of trypsin are
28 Manuel J. Rodrı́guez-Ortega

added and the “surfome” is incubated for a minimum of 2 h

(or overnight). This improves significantly the digestion yield.
12. The easiest way to make the liquids pass through the cartridge
resin is by creating a negative pressure from the bottom side of
the cartridge, using a vacuum pump coupled to a manifold
system, on which the cartridges are placed. This operation
takes only a few seconds for liquid passing through the resin.
If this system is not available, it can be done using a 5 mL
automatic pipette. The tip of the pipette is placed on the upper
side of the cartridge, and the piston is slowly pushed to make
the liquid pass through the resin.
13. As previously indicated in the notes, the amount of sample to
load in the resin must be determined empirically. To the
author’s knowledge, for all the streptococcal species 150 μL
of the “surfome” contains sufficient peptide material to be
detected by LC/MS/MS. But, as expected, factors like the
efficiency of the protease digestion, the amount of bacterial
cells, and other parameters, may require modifications to this
volume. Although the cartridges have a high retention capacity,
the sample flow-through can be passed again through the resin.
14. According to the manufacturer’s instructions, one elution step
with a high concentration of organic solvent is enough to elute
the retained molecules. However, the author has observed that
a second elution with the same solvent concentration still
elutes peptides. This may be due to the small volume of the
resin and the low interaction time between the resin and the
solvent. A clear improvement is achieved by carrying out three
elution steps, as described. The three obtained volumes may be
kept separately or mixed together.
15. Be patient. It may take a few hours to completely evaporate the
volume, especially for the elution fraction at 10% acetonitrile.
16. For all the streptococcal species with which the author has
experience, there is enough peptide material in the resus-
pended “surfome” after cleaning with the cartridges to be
detected by LC/MS/MS. Moreover, the sample can be diluted
without losing detection capacity.
17. Use low-binding tubes. It has been observed that, when using
normal, non low-binding tubes, the number of different pep-
tides identified by LC/MS/MS dramatically decreases after
some months.
18. To be sure that the cleaning process with the Oasis® HLB
cartridge has worked well, keep both the sample flow-through
and the wash fractions, and vacuum-concentrate them,
together with the elution fractions. In there are no peptides
in the eluted fraction(s), the presence of peptides can be
quickly checked in the sample flow-through and the wash
fraction by MALDI-TOF MS.
 Protease Shaving of Bacterial Cells 29

References
1. Navarre WW, Schneewind O (1999) Surface protein antigens in children. J Proteome
proteins of gram-positive bacteria and mechan- 126:228–233
isms of their targeting to the cell wall envelope. 8. Olaya-Abril A, Jimenez-Munguia I, Gomez-
Microbiol Mol Biol Rev 63(1):174–229 Gascon L, Obando I, Rodriguez-Ortega MJ
2. Grandi G (2006) Genomics and proteomics in (2015) A pneumococcal protein array as a plat-
reverse vaccines. Methods Biochem Anal form to discover serodiagnostic antigens
49:379–393 against infection. Mol Cell Proteomics 14
3. Rabilloud T, Chevallet M, Luche S, Lelong C (10):2591–2608
(2010) Two-dimensional gel electrophoresis in 9. Olaya-Abril A, Jimenez-Munguia I, Gomez-
proteomics: past, present and future. J Prote- Gascon L, Rodriguez-Ortega MJ (2014) Sur-
ome 73(11):2064–2077 fomics: shaving live organisms for a fast pro-
4. Wu CC, MacCoss MJ, Howell KE, Yates JR teomic identification of surface proteins. J
3rd (2003) A method for the comprehensive Proteome 97:164–176
proteomic analysis of membrane proteins. Nat 10. Doro F, Liberatori S, Rodriguez-Ortega JJ,
Biotechnol 21(5):532–538 Rinaudo CD, Rosini R, Mora M et al (2009)
5. Rodriguez-Ortega MJ, Norais N, Bensi G, Surfome analysis as a fast track to vaccine discov-
Liberatori G, Capo S, Mora M et al (2006) ery: identification of a novel protective antigen
Characterization and identification of vaccine for Group B Streptococcus hypervirulent strain
candidate proteins through analysis of the COH1. Mol Cell Proteomics 8(7):1728–1737
group A Streptococcus surface proteome. Nat 11. Gomez-Gascon L, Luque I, Olaya-Abril A,
Biotechnol 24(2):191–197 Jimenez-Munguia I, Orbegozo-Medina RA,
6. Olaya-Abril A, Gomez-Gascon L, Jimenez- Peralbo E, Tarradas C, Rodriguez-Ortega MJ
Munguia I, Obando I, Rodriguez-Ortega (2012) Exploring the pan-surfome of Strepto-
MM (2012) Another turn of the screw in coccus suis: looking for common protein anti-
shaving Gram-positive bacteria: optimization gens. J Proteome 75(18):5654–5666
of proteomics surface protein identification in 12. Bohle LA, Riaz T, Egge-Jacobsen W, Skaugen M,
Streptococcus pneumoniae. J Proteome 75 Busk OL, Eijsink VG, Mathiesen G (2011) Iden-
(12):3733–3746 tification of surface proteins in Enterococcus fae-
7. Jimenez-Munguia I, van Wamei WJ, Olaya- calis V583. BMC Genomics 12:135
Abril A, Garcia-Cabrera E, Rodriguez-Ortega 13. Tjalsma H, Lambooy L, Hermans PW, Swin-
MM, Obando I (2015) Proteomics-driven kels DW (2008) Shedding & shaving: disclo-
design of a multiplex bead-based platform to sure of proteomic expressions on a bacterial
assess natural IgG antibodies to pneumococcal face. Proteomics 8(7):1415–1428
 Chapter 3

Methods for Mapping the Extracellular and Membrane

Proteome in the Avian Embryo, and Identification of Putative
Vascular Targets or Endothelial Genes
Witold W. Kilarski, John Herbert, and Andreas Bikfalvi

Abstract
We present a protocol for the specific labeling and isolation of proteins from the membrane surface of
endothelial cells and the surrounding extracellular matrix of organs, experimental wounds and tumors using
chicken embryos. Proteins are deglycosylated on streptavidin resin and purified after gentle elution and
trypsin digestion. Peptides are analyzed by spectroscopy and reverse proteomic fingerprinting. The major
advantages of this protocol include reductions in both the background and overrepresentation of single
proteins that would otherwise mask less well-represented proteins in the mass spectroscopy analysis. We also
present methods to identify putative vascular and endothelial cell targets from isolated chicken membranes
and extracellular proteins. The use of human genome and transcriptome data facilitates this analysis.
Human orthologs of isolated chicken proteins are identified using best hit BLAST searches against the
Human Reference Sequence Database. The expression of Human orthologs is then assessed for endothelial
and non-endothelial cell enrichment using second generation RNA-seq sequenced libraries. Scanning of
the published literature then provides a ranking score of those genes most likely involved in cancer or having
a link to angiogenesis.

Key words Proteomics chick embryo, Membrane and matrix proteome, Chorio-allantoic membrane,
Mass spectrometry, Protein mapping, In vivo biotinylation

1 Introduction

Vascular endothelium and its supportive basement membrane (sur-

fectome) encounter dynamic changes in blood nutrient concentra-
tion and variations in regulatory factor control. The endothelium
also serves as the frontline defense against systemic invasion of
pathogens or metastasis, but under certain situations serves as an
entry point for blood-borne parasites, like viruses, bacteria, or
metastatic tumors. The ability of endothelium and basement

Electronic supplementary material: The online version of this chapter (https://doi.org/10.1007/978-1-4939-

7553-2_3) contains supplementary material, which is available to authorized users.

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_3, © Springer Science+Business Media, LLC 2018

31
32 Witold W. Kilarski et al.

membrane to control key physiological features of organs during

hemostasis or to affect targeted interventions aimed at treating
tissue pathologies during inflammation or metastasis requires an
understanding of the molecular profiles and the dynamic changes of
surfectomes among different vascular beds [1, 2].
The avian embryo is a model organism used to study ontoge-
netic processes, and it is ideally suited for vascular studies. The
quail-chick chimera was developed to study the migration and the
fate of cell populations of intact embryos and to elucidate the
origins and fate of neural crest cells. It has been essential to the
discovery of hemangioblasts and to a better understanding of neu-
ral tube patterning [3, 4]. The virtually two-dimensional vascula-
ture of the developing embryo permits direct and nontraumatic
access to extraembryonic chicken chorioallantoic membrane
(CAM) for the investigation of mechanisms of angiogenesis during
development [5]. In addition, implantation of fibrin gel on CAM
can mimic vascularization of provisional matrix (a deposit of extra-
cellular matrix molecules that support morphogenesis) during
wound healing [6]. Mature CAM can also be used in a nondeve-
lopmental context as a highly sensitive system for testing factors
that induce blood vessel permeability [7]. Using this model system,
we discovered a postdevelopmental growth mechanism utilized by
blood vessels during wound healing and tumorigenesis [8]. We
have also successfully used CAM to compare the surfaceome of
blood vessels in normal organs with the surfaceome of newly
formed vessels in healing wounds and glioblastoma tumor [2].
Completion of the chicken genome sequence has facilitated the
identification and classification of specific protein groups
[9, 10]. Using the principle of reverse engineering, specific groups
of proteins can be isolated from a biological system by any bio-
chemical mean preferable. Following enzymatic cleavage of pro-
teins into smaller peptides [11], molecular masses of the peptide
pool are analyzed by high-resolution mass spectrometry. The origi-
nal proteins can be identified from the peptides in a quantitative
manner after comparison to known genetic coding material. Rybak
et al. [11], however, showed that the specific isolation and propor-
tional enrichment of a poorly represented protein is a major obsta-
cle to this approach. As with other quantitative methods,
overrepresentation of a few major proteins can mask the presence
of proteins that exist in scarce numbers (i.e., regulatory factors like
receptors or their matrix-bound ligands).
We have tried to resolve this problem by increasing the number
of recognizable peptides available for peptide fingerprinting. Using
an approach analogous to the method published by Kamada et al.
[12], N-linked sugar antennas are removed by enzymatic digestion.
Cleavage of large sugar antennas from protein surfaces removes
steric hindrance and allows unrestrained access of proteolytic
enzymes to the protein backbone [13]. This results in an increase
 Mapping the Avian Extracellular and Membrane Proteome 33

in the total number of peptides that can be recognized by their

molecular mass. We additionally cleave off the sialic acid from the
ends of S-linked antennas [14]. This further reduces the sugar
charge and minimizes the repellent effects of sugars on proteolytic
enzymes to increase the number of peptides available for mass
spectroscopic analysis. Avoiding harsh conditions during protein
elution from the resin or digestion of the entire resin with bound
proteins also reduces analytical background. These procedures in
addition to the isolation of non-specifically bound peptides increase
the peptide pool from the resin protein (in our case
non-glycosylated streptavidin), which may otherwise mask less
represented proteins in the sample.
The experimental procedures described here can be divided
into a number of independent steps: (1) Embryo culture; (2) Perfu-
sion and biotinylation of chicken embryos and its extraembryonic
vasculature; (3) Lysis of selected tissues or organs and purification
of biotinylated proteins; (4) Proteolytic digestion of proteins and
deglycosylation of proteins; (5) Elution of biotinylated proteins;
and (6) Bioinformatic and statistical identification of original pro-
teins from known genetic database after calculation of nucleotide
sequence from acquired peptide masses. Mass spectrometry of pep-
tides obtained from proteolytic digestion of proteins and measure-
ment of peptide masses and their relative representation within the
sample are not described in this chapter, as it is anticipated that
users of this protocol will interact with a mass spectrometry core
facility (or similar) to obtain these data.

2 Materials

2.1 Materials 1. Fertilized Brown Leghorn eggs (local sources).

for Embryo Culture 2. Chicken egg incubator set to 38
C and with humidity set to
75% (e.g., Grumbach incubator; Grumbach GmbH, M€ ucke,
Germany).
3. Cell culture grade, 10 cm petri dishes.
4. 100 mM calcium carbonate slurry in water.
5. Electron microscopy grade nylon grid: 250 or 300 μm grid size
mesh (Fisher Scientific, Pittsburgh, PA, USA).
6. Tumor cell cultivation media: 4.5 g/L glucose DMEM with
20 mM HEPES, containing 10% fetal bovine serum (FBS).
7. Confluent monolayer of tumor cells (e.g., glioblastoma, U87
cell line; American Type Culture Collection (ATCC), Mana-
ssas, VA, USA).
8. Cell culture-grade phosphate-buffer saline, pH 7.5 (PBS).
34 Witold W. Kilarski et al.

2.2 Perfusion 1. Anesthesia: 12.5 mg/mL ketamine and 4 mg/mL of xylazine

and Biotinylation solution. Xylazine and ketamine are drugs that can be pur-
Reagents chased from the pharmacy according to local regulations.
2. H-Ringer solution. The H-Ringer solutions used in this proto-
col are prepared either with or without heparin and supple-
mented with glucose Ringer’s solution buffered. For the
standard H-Ringer solution, prepare 20 mM (4-(2-hydro-
xyethyl)-1-piperazineethanesulfonic acid) (HEPES), 115 mM
NaCl, 2.5 mM KCl, 1 mM MgSO4, 1.8 mM CaCl2, 2 mg/mL
glucose, 100 IU heparin (final osmolarity 320 mOsM, pH 7.5)
(B. Braun Medical AG, Sempach, Germany).
3. AutoMate In Vivo Manual Gravity Perfusion system (Braintree
Scientific, Braintree, MA, USA).
4. Biotinylation reagent solution: Prepare a fresh solution (up to
1 h before the procedure) containing 15 mg of N-hydroxysul-
fosuccinimide ester of biotin (sulfo-NHS-LC-LC-biotin)
(ThermoFisher Scientific, Waltham, MA, USA) dissolved in
H-Ringer (without heparin).
5. Quenching solution: Prepare Tris-glycine buffer containing
320 mM glycine (Sigma Aldrich Chemical Company, St. Louis,
MO, USA) and 15 mM tris(hydroxymethyl)aminomethane
(Tris) (Sigma) and adjust pH to 7.4 with HCl.

2.3 Organ Lysis 1. Lysis buffer: Prepare a solution composed of 1% sodium dode-
and Purification cyl sulfate (SDS), 0.1% Triton X-100 (Sigma), 5 mM ethylene-
of Biotinylated diaminetetraacetic acid (EDTA) (Sigma), 10 mM dithiothreitol
Peptides (DTT), 20 mM Tris, 150 mM NaCl and adjust pH to 7.5 with
HCl, and add 2 concentrated inhibitor cocktail (see Note 1).
2. Sonicator-homogenizer.
3. Washing buffer: Prepare a solution composed of 1% SDS, 0.1%
Triton X-100 (Sigma), 5 mM EDTA (Sigma), 10 mM iodoa-
cetamide, 20 mM Tris, 150 mM NaCl and adjust the pH to 7.5
with HCl.
4. 1 M iodoacetamide in water (Bio-Rad Laboratories, Hercules,
CA, USA).
5. Streptavidin (ST)-sepharose beads (High Performance)
(GE Healthcare Life Sciences, Marlborough, MA, USA) (see
Note 2).
6. Streptavidin-sepharose washing column: 10 mL syringe,
clogged with cotton gauze secured with perforated socket
and placed in 50 mL tube.
7. Sepharose washing buffer: Prepare a solution composed of
0.1% SDS, 20 mM Tris, 150 mM NaCl and adjust the pH to
7.5 with HCl.
 Mapping the Avian Extracellular and Membrane Proteome 35

8. Repel Silane (dimethyl dichlorosilane) (see Note 3).

9. Build a silanized washing column (WC) by securely attaching
an 18 Gauge needle with protective plastic cover to a 10 mL
syringe but with the plunger removed. Both, syringe outlet and
needle should be clogged using cotton gauze. Four holes
should be made in the bottom of the plastic protective cover
of the 18 Gauge needle using either another needle or scissors.
10. Tris-buffered saline (TBS): Prepare a solution of 20 mM Tris,
150 mM NaCl and adjust the pH to 7.5 with HCl.
11. Construct and elution column (EC) made from a 1 mL syringe
connected to the needle (Gauge 28) clogged with cotton
gauze.

2.4 Proteolytic 1. PNGase F (Sigma).

Digestion of Proteins 2. Sialidase (Neuramidase from Arthrobacter ureafaciens)
and Deglycosylation (Sigma).
of Proteins
3. 200 mM, pH 5.0 phosphate buffer.
4. Ultrapure water of 18 or more megohms resistance at 25
C
(see Note 4).
5. Biotin Stock Solution: 100 mM biotin stock dissolved in ultra-
pure dimethyl sulfoxide (DMSO).
6. Modified zinc fixative: This osmolarity-corrected zinc fixative
(Zn-fix) solution consists of 4.5 mM CaCl2, 52 mM ZnCl2,
32 mM Zn(CF3COO)2, 2 mM Tris, 38 mM glycine at pH 6.5,
and an osmolarity of 340 mOsm/L (see Note 5).
7. Standard immunohistochemical (IHC) setup suitable for pre-
paring fixed tissues/sections, and incubations with appropriate
antibodies that bind to specific antigens. The presence/loca-
tion of the antigen is then determined using an appropriate
substrate or chromogen to generate colored deposits/signals at
the sites of antibody–antigen binding.
8. Phosphate buffer: 200 mM, pH 5.0. Adjust the pH of solution
by using dilute orthophosphoric acid.

2.5 Miscellaneous 1. Biological safety cabinet (BSC).

Equipment 2. Centrifuge capable of holding 15 and 50 mL conical tubes.
and Supplies
3. Microcentrifuge.
4. Scapel with blade.
5. 15 and 50 mL conical tubes.
6. Polystyrene tub.
7. 18 cm long blunt forceps (Allgaier Instrumente GmbH, Frit-
tlingen, Germany).
8. Curved hemostat forceps (Allgaier).
36 Witold W. Kilarski et al.

9. Fine curved forceps (Allgaier).

10. Sangofix, cannula tube (B. Braun Melsungen AG, Melsungen,
Germany).
11. Syringes and 0.45 μm syringe filter.
12. 10 μL pipette tips (sterile).
13. Luer-locked 10 and 1 mL syringes (B. Braun).
14. 18 Gauge needle.
15. Gauze.
16. Tube rotator.
17. Polypropylene microfuge tubes.

3 Methods

Obtain all necessary permissions to work with vertebrate embryos

according to your local laws and follow all guidelines established by
your institution.

3.1 Embryo Culture 1. Fertilized brown leghorn eggs should be obtained from a local
farmer.
2. On the day of delivery place eggs horizontally inside the incu-
bator and culture without tilting for 3 days at 38
C and 75%
humidity (see Note 6).
3. On day 4, gently move eggs to a biological safety cabinet (BSC).
4. Crack the shell by hitting the bottom side of the egg on the edge
of an open petri culture dish using a single decisive motion.
5. Gently open the bottom side of the shell and transfer the
contents into the dish, taking special care not to break the
egg yolk (see Notes 7 and 8). Discard the shell and close the
lid. Dishes where the yolk is broken after embryo transfer from
the egg should be immediately discarded as biological waste.
6. Gently and without unnecessary shaking, return the closed
petri dish containing the unbroken egg yolk to the incubator.
7. Continue embryo culture for an additional 7 days.
8. During the incubation period, check the embryos every 3 days
to eliminate dead embryos. This can be determined with the
naked eye while observing the embryos in the incubator with-
out opening the dish, but avoid unnecessary plate movements
during the embryo examinations. Dead embryos loose
completely the red appearance of blood vessels in the yolk
and later CAM. Gently remove and discard dead embryos.
9. To prevent embryo deaths after day 10, 0.5 mL of a 100 mM
calcium carbonate slurry (in water) can be added onto the
embryo CAM. The addition of this slurry to the embryo serves
 Mapping the Avian Extracellular and Membrane Proteome 37

to supplement the calcium supply that is lost upon removal of

the shell (see Note 9).
3.1.1 Optional Wound 1. On day 10, CAM injuries can be made by multiple parallel and
Healing on the CAM superficial scalpel cuts (non-penetrable) across a 1 cm2 surface
area, followed by gentle epithelium scraping of the injured
CAM; however, minimize the time that an embryo is kept
outside the incubator and perform all procedures in a BSC.
2. The wound can be localized anywhere on the CAM; however,
locations close to the CAM edge should be avoided as, due to
embryo movements, the nylon mesh may move under
the CAM.
3. The cuts are made by moving the scalpel back and forth above
the area to be injured. The scalpel should only touch the CAM
surface. Do not push the scalpel into the CAM as this can cut
open the CAM. In case the CAM is cut and major vessel bleeds
extensively and embryo has to be discarded.
4. The wound area should be covered immediately with a 1.5 cm2
square electron microscopy nylon mesh.
5. Using the same procedure, a second wound can be made at
least 2 cm from the initial mesh placement.
6. Return dishes to the incubator for an additional 6 days.

3.1.2 Optional Tumor 1. Wash confluent monolayer of tumor cells (e.g., glioblastoma,
Cells Implantation U87 cell line) twice with PBS and trypsinize them for 3 min
at RT.
2. Cells should be then washed twice in cultivation media (4.5 g/
L glucose DMEM with 20 mM HEPES) containing 10%
serum, followed by a third wash with PBS.
3. Dissociate the cells using a trypsin-based solution according to
the protocols provided by ATCC.
4. After centrifugation (3 min at 4 oC, 800  g), the media/
supernatant should be removed by aspiration and discarded.
5. The pellet slurry should be applied directly onto CAM wounds
covered with nylon grid as described above.
6. Return dishes with embryos to the incubator for additional
6 days.

3.1.3 Perfusion and In 1. All procedures can be performed on the bench, and all solu-
Vivo Biotinylation tions should be prewarmed to 37
C.
of Endothelial Surfectome 2. On day 16 of incubation, embryos should be anesthetized by
of Chicken Embryo Organs application of 500 μL of the Anesthesia solution (assuming an
egg weight of approximately 25 g) directly on the CAM surface
(see Note 10).
38 Witold W. Kilarski et al.

3. Approximately 5 min later when the embryo is fully anesthe-

tized, the petri dish containing one embryo should be plunged
into a polystyrene tub containing 2 L of PBS at 37
C. Only
one embryo at the time can be submerged in warm PBS in the
polystyrene tub. Fresh prewarmed PBS should be poured into
the tub for each new embryo.
4. The dish and embryo should be fully submerged, after which
time the whole embryo surrounded by the CAM should be
carefully detached from the plastic.
5. The CAM almost completely encloses the whole embryo,
except at one location where CAM membranes do not fuse
(“CAM opening”). The following procedure moves the
embryo outside of the CAM membrane.
6. Insert blunt forceps through the natural opening in the floating
CAM. Freely floating embryos can then be gently grabbed with
blunt forceps and pulled inside-out through an opening in the
CAM. This procedure exposes the amniotic sac with the
enclosed embryo and separates it from the CAM.
7. At this time, the CAM and the embryo are still connected by an
artery and two veins. No major bleeding in the CAM or
embryo should be observed (only capillary bleeding is an
acceptable type of injury).
8. The PBS should be decanted gently while holding the embryo
and CAM. The amniotic sac should be cut with scissors.
9. Embryo wings and legs should be fixed to the bottom of the
tub with needles that bend the chicken appendages and
immobilize them.
10. The embryo chest should be opened using fine scissors with a
single axial cut under the keel and two parallel cuts through the
ribs along the median body plane.
11. The right pulmonary artery (see Note 11) should be cannu-
lated with a 10 μL plastic tip inserted into the cannula tube and
fixed in place with a curved hemostatic forceps. The heart aorta
should be blocked with another clamp.
12. Whole blood from the CAM, yolk, and embryo circulation is
washed away for 15 min with 15 mL of 37
C H-Ringer’s
solution in container A using gravity enforced flow (Ringer’s
buffer on 130 cm (100 mm Hg) at the beginning of the
perfusion) and a perfusion device (AutoMate In Vivo Manual
Gravity Perfusion system). The speed of flow is approximately
2 mL/min.
13. Attach the perfusion device to a container B holding 15 mL
freshly prepared biotinylation reagent solution and perfuse this
solution into the pulmonary artery over a period of 15 min
(speed of flow is approximately 1 mL/min).
 Mapping the Avian Extracellular and Membrane Proteome 39

14. Residual succinimide reagent is quenched by injection of

15 mL of Tris-glycine buffer (without heparin) for 15 min at
room temperature using container A of the perfusion device.
15. The amine buffer is then eluted with room temperature
H-Ringer solution supplemented with 15 mL of protease
inhibitor cocktail for 15 min.
16. The healing CAM, tumor CAM and embryonic organs of
choice (e.g., CAM, small intestine, liver, and kidney) should
be dissected free and Zn-fixed [15] for IHC or snap-frozen in
liquid nitrogen for further purification of biotinylated proteins
(see Note 5).

3.2 Organ Lysis 1. Pool selected organs from 5 to 15 biotinylated embryos. Use
and Purification 15 or 50 mL conical tubes, depending on organ size.
of Biotinylated 2. Weigh the pooled organs and add 40 μL of lysis buffer per
Proteins milligram of wet tissue. Because tissue with lysis buffer heats
during sonication, tissue together with lysis buffer should not
exceed 10% of maximum tube volume.
3. Sonicate the mixture at room temperature in a sonicator-
homogenizer at medium amplitude settings until it becomes
frothy
4. Boil samples for 15 min at 95
C and after cooling to room
temperature, add 1 M iodoacetamide (to acetylate and perma-
nently protect free thiols) to a final concentration of 20 mM
and leave at room temperature for an additional 30 min.
5. Cool sample at room temperature, but do not place sample on
ice as SDS precipitates below 15
C.
6. Spin sample at 14,000  g for 5 min at room temperature.
7. Collect the soluble fraction and filter through a 0.45 μm
syringe filter (see Note 12).
8. Mix 1 mL of lysate with 100 μL of streptavidin (ST)-linked
beads prewashed twice with lysis buffer supplemented with
iodoacetamide at a final concentration of 20 mM.
9. Incubate with constant rotation (on a Tube Rotator) at room
temperature for 1 h.
10. Quick spin the slurry at 1000  g for 3 min and transfer the
beads with washing buffer to a new 15 mL conical tube.
11. Perform two additional washes with 15 mL washing buffer by
spinning the slurry for 3 min at 1000  g.
12. From this point all equipment used for the washing and elution
procedures of biotinylated proteins must be silanized or the
plasticware should be made of materials characterized by
low-binding affinity.
13. Transfer the bead slurry to a WC using SDS washing buffer.
40 Witold W. Kilarski et al.

14. Place the WC in a 50 mL conical tube and pour 12 mL (maxi-

mum volume of 10 mL syringe) of washing buffer into the
syringe. Centrifugate at 2000  g for 5 min at room
temperature.
15. Repeat the washing step six more times.
16. Perform three additional washes with TBS using the same
conditions as indicated in step 14 (above).

3.3 Enzymatic 1. Washed beads should be transferred to a 2 mL polypropylene

Digestion microfuge tube and TBS pH 7.4 added to final volume of
and Deglycosylation 0.5 mL.
2. The enzyme PNGase F (5 U/mL final), which removes
N-linked sugar antennas, should be added to the slurry and
incubated for 24 h at room temperature with constant rotation.
3. Pellet the slurry by pulse centrifugation and discard the
supernatant.
4. Wash the slurry three times with 500 μL phosphate buffer
(200 mM, pH 5.0). Pulse-centrifuge after each wash and dis-
card the supernatant.
5. Sialidase, which removes negatively charged terminal sialic acid
from N- and O-linked antennas, should be added to a final
concentration of 5 U/ml in phosphate buffer and incubated
for 24 h at room temperature with constant rotation in the
same final volume.
6. Transfer the beads in phosphate buffer back to the WC. Wash
three times with TBS and once with ultrapure water at room
temperature.
7. Transfer the beads in water to an elution column (EC). Water is
removed from the beads by slowly pushing the plunger into the
syringe until no more eluent is seen.

3.4 Elution 1. Elute biotinylated proteins with water heated to 70
C accord-

of Biotinylated ing to the technique of Holmberg et al. [16] (see Note 13).
Proteins 2. Water should be prewarmed to 70
C. A volume of water equal
to that of the bead volume should be drawn into the syringe by
placing the syringe into the prewarmed water and pulling back
on the plunger.
3. Place the syringe in a water bath at 70
C for 3 min.
4. Attach silanized tubing to the syringe and elute the biotinylated
proteins by depressing the plunger.
5. Repeat this procedure from steps 1 to 3 at least ten times.
6. Pool all eluants and concentrate the proteins using either a
Speedvac for electrophoresis or frozen and lyophilized for
mass spectrometric analysis.
 Mapping the Avian Extracellular and Membrane Proteome 41

3.4.1 Histochemistry 1. Histochemistry and Western blotting of purified biotinylated

and Western Blotting proteins should be performed according to a standard labora-
(Optional) tory protocols. Western blot for major intracellular proteins
(e.g., actin or tubulin) can serve as a control for extracellular
specificity of the biotinylation reaction (see Note 14).

3.5 Bioinformatics 1. We recommend that the user work with individuals who are
Procedures to Identify adept in the use of databases to identify endothelium enriched
Endothelium Enriched genes and vascular targets.
Genes and Potential 2. We also provide protocols in the Supplement that are intended
Vascular Targets as a guide for someone with limited bioinformatics experience
who is willing to learn and use a command line based Linux
operating system.
3. The bioinformatics protocols involve four main themes, all of
which are provided in the Supplement to this chapter.
(a) Human ortholog identification of chicken proteins.
(b) Endothelial gene expression assessment of human
orthologs.
(c) Literature abstract scanning and the generation of a rank-
ing score.
(d) Combining results and investigation of the best candidate
genes.
4. These methods were performed using the Ubuntu 12.04.5
OS. It is assumed that a user has access to such a machine
(or similar) and is able to navigate directories and run simple
commands.
5. For further information, tutorials are available online for
Linux, including one at http://www.ee.surrey.ac.uk/Teach
ing/Unix/.

4 Notes

1. Triton X-100 is necessary for dissolving SDS precipitates that

form when a potassium-rich cell content is released.
2. Streptavidin-sepharose beads differ in quality depending on the
vender as some preparations “leak” streptavidin into the buffer.
We have found that High Performance Streptavidin-sepharose
from GE Healthcare is optimal for our applications.
3. All tubes and tips that are in contact with the streptavidin beads
and the eluted biotinylated proteins have to be of low-binding
quality or should be silanized with Repel Silane to minimize
unspecific binding of proteins to plasticware.
4. It is important that water used to purify proteins is of the
highest purity. Ultrapure water should be aliquoted and frozen
42 Witold W. Kilarski et al.

to avoid prolong exposure to air CO2 that will lead to

acidification.
5. Zinc fixative is the choice in our lab, but any fixation method
that preserves biotin-protein conjugates can be used.
6. Do not clean egg shells with detergents or ethanol as this
treatment removes the protective wax layer from the shell
surface and can lead to infection and death of the developing
egg. Clean all dirty eggs with wet tissue or exclude any egg that
cannot be cleaned easily.
7. White in 1974 described an alternative method that helps avoid
laborious techniques involved with the initial transfer of the
embryo from the shell to the ex ova culture [17].
8. The embryo is less dense than the yolk and egg white and will
always locate itself to the top surface of an egg. When eggs are
incubated in an horizontal position, hit the shell to open it at its
ventral side, as the embryo is located at the dorsal side. This
minimizes the chances of breaking the yolk and injuring the
embryo.
9. Addition of calcium carbonate supports proper bone formation
and development of central nervous system of the embryo. It
will also lead to intense embryo movements and, as a conse-
quence, to massive injuries to the extraembryonic membranes
and embryo death. These embryo movements can be stopped
by daily addition 250 μL of 1 mg/mL pancuronium bromide
(muscle relaxant, Pavulon®) onto the CAM for the course of
the culture starting at day 10 of the incubation. It is worth
noting that low molecular drugs are quickly adsorbed by CAM
vasculature and result in immediate effects on the embryo.
10. To ensure that the embryo enters full anesthesia, pancuronium
bromide (if used) should be omitted from the culture on the
day of the experiment. Deep anesthesia is required before
perfusion of the embryo, which can be visually assured by
cessation of all embryo movements.
11. This is a classic procedure for the intracardiac perfusion. The
vast majority of fluid injected into the right pulmonary artery
bypass embryo fluid-fill (dysfunctional lungs) through connec-
tions of lung arteries with the proximal descending aorta via
vessels called ductus arteriosa. We found that injecting Ringer’s
solution in the left ventricle and opening of the right atrium for
the blood to be flushed from the embryo vasculature (standard
procedure for perfusion of adult animals) is not an efficient
method for embryo perfusion as most of the infused fluid will
be shunted directly to right atrium via ductus arteriosa. Only
injection of fluids into the right aortic arch (or right ventricle
and cutting open right atrium) allows successful perfusion of
the whole embryo and its extraembryonic tissues (CAM).
 Mapping the Avian Extracellular and Membrane Proteome 43

When injecting H-Ringer’s solution directly into the right

pulmonary artery (with the heart completely cut off), the left
artery should be clamped. Direct injection into arteries
(instead of heart atrium) is preferable when embryo cultures
are performed without calcium supplementation, as heart tis-
sues are too fragile for the perfusion.
12. Filtration is necessary to remove insoluble material that has
lower density than tissue lysate.
13. Because of an increase in the Tm of streptavidin denaturation
from 75
C in absence of biotin to 112
C at full biotin
saturation (increase in streptavidin tetramer stability in pres-
ence of biotin) [18], 100 μM biotin is added to the water
before elution of biotinylated proteins. This prevents dissocia-
tion of streptavidin monomers and leaking of streptavidin to
elute.
14. Sulfo-NHS-LC-LC-biotin is membrane impenetrable. Using
our biotinylation protocol, we could detect only traces of
intracellular proteins after surfectome biotinylation in normal
tissues and organs, and in granulation tissues present in
wounds. This was in contrast to glioblastoma cultured on the
CAM, where in healthy tissues we have detected large number
of intracellular proteins. This is most likely due to high level of
necrosis of the tumor cells, resulting in the presence of intra-
cellular proteins in extracellular compartments and in highly
permeable tumor blood vessels that can leak biotinylation
reagent into the extracellular matrix where it can react with
proteins that are normally not present outside cellular plasma
membrane. Hence caution is necessary when analyzing data
from pathological condition (e.g., tumors) where high vessels
permeability is associated with high level of necrosis.

Acknowledgment

This work was supported by INSERM, the “Association de la

Recherche sur le Cancer” (ARC) and the “Ligue Contre le Cancer”
to A.B. This work was supported also in part by grants from the
Polish-Swiss Research Programme (PSPB-057/2010).

Supplementary Materials

Bioinformatics Mass spectrum peak lists are assigned to Uniprot [19] or Refseq
Procedures to Identify [20], chicken proteins using the Mascot Daemon software package
Endothelial Enriched (Matrix Science). FASTA sequences of Uniprot and Refseq proteins
Genes are the starting data. To facilitate identification of endothelial and
putative vascular target genes, the superior annotation and data of
44 Witold W. Kilarski et al.

the human transcriptome and genome are utilized. The first step in
the procedure is to assign putative human orthologs to the chicken
proteins. This is done using the BLAST database search alignment
algorithm [21] as performed previously [22]. This methods finds
the best Human protein hit to a chicken protein, using a stringent
expectation value cutoff of 1e6. This can be accomplished as
follows:
– Identification of chicken proteins human ortholog

1. Make a new directory where this analysis can be carried out;

it can be any name but endo_searches is used here. Type the
commands;
mkdir endo_searches;
cd endo_searches

This stage is very important; do not use a directory where

you have existing files as we will clean up as we go along. If
the “rm” command is mistyped, causing all files to be lost,
you only have start this tutorial again and no important files
are lost.
2. Download and install the command line version of BLAST
appropriate to the computer architecture; (ftp://ftp.ncbi.
nlm.nih.gov/blast/executables/blastþ/LATEST/).
3. Follow the installation guide provided in the BLAST docu-
mentation. The programs “makeblastdb” and “blastp” are
the programs used in this protocol. For instance, you can
copy the download link from the BLAST download page by
right-clicking in either the Google Chrome or Firefox web
browsers; then enter “wget --content-disposition”, then
paste the copied link into the Linux command prompt.
The full command will appear as;
wget --content-disposition ftp://ftp.ncbi.nlm.nih.gov/
blast/executables/blastþ/2.2.29/ncbi-blast-2.2.29-
þ-x64-linux.tar.gz

4. Once downloaded, decompress BLAST by typing;

tar --zxvf ncbi-blast-2.2.29þ-x64-linux.tar.gz

Then you will be able to change into the BLAST direc-

tory by typing;
cd ncbi-blast-2.2.29þ/bin/

This “bin” location has the makeblastdb and blastp

programs. Typing
pwd
 Mapping the Avian Extracellular and Membrane Proteome 45

will print the full path to the location on the file system (pwd
means present working directory) where the programs are,
and this location should be entered into the accompanying
Perl script.
5. Download and unzip the supplementary files that accom-
pany this article. Edit these paths into the Perl script
provided, called “find_human_chick_ortholog.pl”. The
Ubuntu path variable can also be adjusted to include the
BLAST bin location, by typing
export PATH=$PATH:‘pwd‘

Change directory back to endo_searches directory with;

cd ../../

The current directory size now is ~705Mb, so remove the

download file with the command;
rm ncbi-blast-2.2.29+-x64-linux.tar.gz

6. To test BLAST is working at the command line, type

makeblastdb –h
blastp –h

These commands should produce a printed message

that describes the different command line switches available.
If a “No command” warning message appears, then either
BLAST is not installed correctly or the path to the execu-
tables is incorrect.
7. The Reference Sequence database of Human proteins is
downloaded from the NCBI [23]; both the FASTA and
Genbank Flat Files from the following link (ftp://ftp.ncbi.
nlm.nih.gov/refseq/H_sapiens/mRNA_Prot/). The
FASTA sequences are used in the BLAST searches and the
flat file is used to collect gene symbols. These files are first
decompressed using gunzip.
8. Download files with the two following commands;
wget --content-disposition ftp://ftp.ncbi.nlm.nih.gov/
refseq/H_sapiens/mRNA_Prot/human*protein.faa.gz

wget --content-disposition ftp://ftp.ncbi.nlm.nih.gov/

refseq/H_sapiens/mRNA_Prot/human*protein.gpff.gz

Decompress with
zcat human*protein.faa.gz > human.protein.faa
zcat human*protein.gpff.gz > human.protein.gpff

The size of the data now is ~6.4Gb, so delete any files not
needed with the command;
rm human*protein.faa.gz human*protein.gpff.gz
46 Witold W. Kilarski et al.

The Perl script find_human_chick_ortholog.pl

(that you copied from the supplementary files) performs
file formatting, the BLAST searches and some post proces-
sing. It is run using four command line switches or names of
files.
(a) The name of the FASTA file for human Refseq proteins
(human.protein.faa)
(b) The name of the Genbank Flat File containing gene
symbols (human.protein.gpff)
(c) The name of chicken proteins in FASTA format
(chick_proteins.fa supplied example)
(d) The name of the results file
9. Copy all the supplementary files into the endo_searches
directory made earlier.
If you are in the endo_searches directory;
cp /location/of/SupplementaryFilesDirectory/* ./.

The Perl script will need to be “executable”, type;

chmod þx find_human_chick_ortholog.pl

Use the following command to find human chicken

orthologs;
(nohup ./find_human_chick_ortholog.pl human.protein.faa
human.protein.gpff chick_proteins.fa chick.out )>&
stderrs

this will produce a tab delimited file with chicken protein

accession number in column 1 and the human gene symbols
(orthologs) in column 2, and product information in col-
umn 3. The human gene symbol will be used in the
subsequent analysis to ascertain whether genes are enriched
in endothelial cells or not. If an error is produced, it could
be the Perl “Shebang” line needs altering. Type;
which perl

this should produce where Perl is installed. The first line of

the Perl script should have this location, with a “#!” char-
acters at the beginning, like;
#!/usr/bin/perl –w

In summary, a best hit BLAST searching approach is

used to find a Human ortholog of a chicken protein. In
most cases, using the 1e–6 cutoff, orthologs will be evolu-
tionary related as BLAST approaches have proven effective
[24]. These human genes are then applied in the following
expression analysis to determine endothelial enrichment.
 Mapping the Avian Extracellular and Membrane Proteome 47

– Endothelial gene expression assessment of Human orthologs

In previous studies, Sanger sequenced cDNA libraries of
endothelial and non-endothelial cells were compared to find
genes with an endothelial enriched expression profile
[25, 26]. Since that work, second generation sequencing (Illu-
mina [27], 454 (http://www.454.com/) and Solid (http://
www.appliedbiosystems.com/)) projects have generated a new
wave of gene expression data that can be mined for endothelial
enrichment. Accordingly, RNAseq libraries of second generation
sequencing are mined to find endothelial preferentially
expressed genes.
Samples of endothelial cell and non-endothelial cell expres-
sion data are sought by mining public expression resources.
Specifically the Gene Expression Omnibus (GEO, [28]) and
Sequence Read Archive (SRA, http://www.ncbi.nlm.nih.gov/
sra) databases are searched by “eye” to find contrasting expres-
sion library sets. We want to find human genes that show a high
level of expression in endothelial cells and a low level in other cell
types. For this protocol, and as an example, two RNAseq data
sets were found and utilized.
Differential expression of RNAseq data can be accomplished
using a variety of published software. For instance bowtie,
tophat, and cufflinks [29–31] have been published to help biol-
ogists map RNAseq reads onto a reference genome and quanti-
tate differential expression between different tissue types, as well
as construct novel transcripts. In contrast, Express [32] can be
used to measure differential expression using a novel transcrip-
tome (no genome being available). Here a basic guide is given
on how to perform a differential gene expression analysis on
human genome data using STAR [33], FeatureCounts [34], and
edgeR [35] algorithms. These programs enable a quick analysis,
and a basic protocol is described as follows:
1. RNAseq libraries are RNA libraries similar to cDNA libraries
(polyA selected RNA or Ribo RNA minus enriched RNA),
and both give a measure of the level of RNA abundance.
The reads for RNAseq are shorter than those of cDNA
libraries but their transcriptome coverage is far greater.
Investigating gene expression repositories (http://www.
ncbi.nlm.nih.gov/sra/) is done to identify appropriate
gene expression data. In this analysis, the following acces-
sion numbers for endothelial libraries were found;
GSM1386279: HUVEC 1; Homo sapiens; RNA-Seq
(SRR1286927), GSM1386280: HUVEC 2; Homo sapiens;
RNA-Seq(SRR1286928) and GSM1386281: HUVEC 3;
Homo sapiens; RNA-Seq(SRR1286929). More informa-
tion on this study can be found at; http://www.ncbi.nlm.
nih.gov/Traces/sra/?study¼SRP041988). Specifically,
three biological replicate RNA-seq libraries of Human
48 Witold W. Kilarski et al.

Umbilical Vein Endothelial Cells were used. In a similar

manner, three biological replicates of Ovarian cancer cell
lines were chosen as a contrast data set (Study was http://
www.ncbi.nlm.nih.gov/Traces/sra/?study¼ERP000710
and accessions of the three libraries are; ERR035544,
ERR035538 and ERR035534).
2. Downloading SRA RNA-seq libraries. These files are large
and come in a custom SRA format with an .sra string
appended to the accession numbers; e.g., ERR035544.sra.
This can be downloaded by selecting the FTP link, from the
“Run” row of the following page; http://trace.ncbi.nlm.
nih.gov/Traces/sra/?run¼ERR035544. The other repli-
cates are downloaded similarly, by modifying this URL.
ERR035538.sra can be downloaded from; http://trace.
ncbi.nlm.nih.gov/Traces/sra/?run¼ERR035538, and so
on for the other accessions. An example command are;
wget --content-disposition ftp-trace.ncbi.nlm.nih.gov/
sra/sra-instant/reads/ByRun/sra/ERR/ERR035/ERR035534/
ERR035534.sra
wget --content-disposition ftp-trace.ncbi.nlm.nih.gov/
sra/sra-instant/reads/ByRun/sra/ERR/ERR035/ERR035538/
ERR035538.sra
wget --content-disposition ftp-trace.ncbi.nlm.nih.gov/
sra/sra-instant/reads/ByRun/sra/ERR/ERR035/ERR035544/
ERR035544.sra

and for endothelial samples, commands of;

wget --content-disposition ftp-trace.ncbi.nlm.nih.gov/
sra/sra-instant/reads/ByRun/sra/SRR/SRR128/SRR1286927/
SRR1286927.sra
wget --content-disposition ftp-trace.ncbi.nlm.nih.gov/
sra/sra-instant/reads/ByRun/sra/SRR/SRR128/SRR1286928/
SRR1286928.sra
wget --content-disposition ftp-trace.ncbi.nlm.nih.gov/
sra/sra-instant/reads/ByRun/sra/SRR/SRR128/SRR1286929/
SRR1286929.sra

3. Installation of the SRA toolkit. The SRA files need to be

uncompressed and split into separate pairs of the paired end
sequencing. Note; for the gene expression analysis done
here only the first/left reads of the paired ends are used.
The toolkit is available from http://www.ncbi.nlm.nih.gov/
Traces/sra/?view¼software. As this was done on a 64 bit
Ubuntu OS, the command
wget --content-disposition http://ftp-trace.ncbi.nlm.
nih.gov/sra/sdk/2.4.0-1/sratoolkit.2.4.0-1-ubuntu64.
tar.gz
 Mapping the Avian Extracellular and Membrane Proteome 49

was run for download, and decompression is performed

tar –zxvf sratoolkit.2.4.0-1-ubuntu64.tar.gz

An export path command is then executed so that the

appropriate programs can be run from any directory
location;
export PATH=$PATH:‘pwd‘/sratoolkit.2.4.0-1-ubuntu64/bin

Now typing fastq-dump should produce a short usage

message; adding a help flag will print further options for this
program (fastq-dump –h).
Let’s delete the download file;
rm sratoolkit.2.4.0-1-ubuntu64.tar.gz

4. Decompression of .SRA files. For each of the six .sra files,

use the following;
fastq-dump --split-files ERR035544.sra

or to do all files at once;

(nohup fastq-dump --split-files *.sra > stdout )>&
stderr

This command will take some time to finish.

After finishing, the current directory size will have
grown to a big at ~102Gb.
We can delete the files we no longer need, with the
commands;
rm *_2.fastq

Note, be careful to type or copy this command exactly

how it is written. After running this command these
commands, we should have 6 fastq files: SRR1286929_1.
fastq, SRR1286928_1.fastq, SRR1286927_1.fastq,
ERR035544_1.fastq, ERR035538_1.fastq and
ERR035534_1.fastq. Now three biological replicates of
HUVEC and three biological replicates of an ovarian cancer
cell line should remain and these are the endothelial and
non-endothelial RNA-seq pools of data sets to be
compared.
5. Preparation of mapping reads to genome. Each of the
FASTQ file contains sequence reads of RNA that are
assigned to a gene based on genome mapping position
using the STAR read mapping algorithm. The software,
FASTA files of the human genome and a Human Reference
Sequence project GTF file are needed.
(a) Human genome FASTA files. The hg19 version of the
human genome is used in this protocol and can be
downloaded from UCSC Genome Browser FTP site
50 Witold W. Kilarski et al.

(http://hgdownload.soe.ucsc.edu/goldenPath/hg19/
bigZips). Specifically, the FASTA sequences of chromo-
somes are downloaded with
wget --content-disposition http://hgdownload.soe.
ucsc.edu/goldenPath/hg19/bigZips/chromFa.tar.gz

These are decompressed using

tar –zxvf chromFa.tar.gz

The base chromosomes are put into a single file

using the concatenate command;
cat chr1.fa chr2.fa chr3.fa chr4.fa chr5.fa chr6.fa
chr7.fa chr8.fa chr9.fa chr10.fa chr11.fa chr12.fa
chr13.fa chr14.fa chr15.fa chr16.fa chr17.fa chr18.
fa chr19.fa chr20.fa chr21.fa chr22.fa chrX.fa
chrY.fa chrM.fa > hg19.fa;

files no longer needed are deleted

rm chr*.fa chromFa.tar.gz;

(b) Downloading and compiling STAR.

wget --content-disposition https://github.com/
alexdobin/STAR/archive/STAR_2.4.0d.tar.gz . Decom-
press, change directory to STAR_2.4.0d and compile
by typing;
tar –zxvf STAR-STAR_2.4.0d.tar.gz;
cd STAR-STAR_2.4.0d;
make;
cd ..

Remove the download file;

rm STAR-STAR_2.4.0d.tar.gz

A “STAR” program should have been created and

this can be put in the path with
export PATH=$PATH:‘pwd‘/STAR-STAR_2.4.0d/

(c) Formatting the Human genome FASTA for STAR. An

index is built using a Reference sequence project GTF
file (splice junction database), to facilitate mapping of
reads, particularly those that span introns (the GTF file
is provided and is named RefseqHG19_gtf_gene). First
make a new directory in endo_searches directory, called
HumanGenomeDir, with the command
mkdir HumanGenomeDir

Then to build the index, use the command

 Mapping the Avian Extracellular and Membrane Proteome 51

(nohup STAR --runMode genomeGenerate --genomeDir ./

HumanGenomeDir --genomeFastaFiles ./hg19.fa --
sjdbGTFfile ./RefseqHG19_gtf_gene --runThreadN 100
--sjdbOverhang 100 > stdout )>& stderr

Please see the STAR manual for further informa-

tion on the chosen command line options (http://rna-
star.googlecode.com/files/STARmanual_2.1.4.pdf).
You may need a smaller number of threads (run-
ThreadN), and this command can take a bit of time to
complete.
6. Mapping of reads onto the Human genome. Each FASTQ
file can be mapped against the Human genome using STAR
with the command;
STAR --genomeDir ./HumanGenomeDir --readFilesIn
ERR035544_1.fastq --runThreadN 100 --outFileNamePrefix
ERR035544;

STAR --genomeDir ./HumanGenomeDir --readFilesIn

ERR035534_1.fastq --runThreadN 100 --outFileNamePrefix
ERR035534;

STAR --genomeDir ./HumanGenomeDir --readFilesIn

ERR035538_1.fastq --runThreadN 100 --outFileNamePrefix
ERR035538;

STAR --genomeDir ./HumanGenomeDir --readFilesIn

SRR1286928_1.fastq --runThreadN 100 --outFileNamePrefix
SRR1286928;

STAR --genomeDir ./HumanGenomeDir --readFilesIn

SRR1286927_1.fastq --runThreadN 100 --outFileNamePrefix
SRR1286927;

STAR --genomeDir ./HumanGenomeDir --readFilesIn

SRR1286929_1.fastq --runThreadN 100 --outFileNamePrefix
SRR1286929;

7. Installing and running FeatureCounts to determine raw

gene counts. The .sam files produced in the previous step
have to be processed to find which exons of a gene reads
overlap with; assigning reads to genes and the genes raw
counts. Several programs are available for this and Feature-
Counts form the Subread suite was chosen. The latest ver-
sion is downloaded with;
wget --content-disposition http://sourceforge.net/pro-
jects/subread/files/subread-1.4.5-p1/subread-1.4.5-p1-
Linux-x86_64.tar.gz/download
52 Witold W. Kilarski et al.

and decompressed with the command;

tar –zxvf subread-1.4.5-p1-Linux-x86_64.tar.gz
cd subread-1.4.5-p1-Linux-x86_64/bin
export PATH=$PATH:‘pwd‘
cd ../..
Raw counts of genes can be generated with the following
command;
featureCounts -T 28 -t exon -g gene_id -a Re-
fseqHG19_gtf_gene -o Endo_counts.txt ERR035544Aligned.
out.sam ERR035534Aligned.out.sam ERR035538Aligned.out.
sam SRR1286928Aligned.out.sam SRR1286927Aligned.out.sam
SRR1286929Aligned.out.sam

A table of counts is produced, in the Endo_counts.txt

file, which is the input to R and edgeR programs.
The size of the data we now have is ~122Gb. We can
remove the download file and others we don’t need;
rm subread-1.4.5-p1-Linux-x86_64.tar.gz SRR12* ERR03*
STAR-STAR_2.4.0d.tar.gz

8. Using R and the edgeR package to define differentially

expressed genes. The R environment has been developed
for performing statistical analysis and producing associated
graphics. It can be installed following instructions for your
OS form Cran; a guide to installing R on Ubuntu is found at
http://cran.r-project.org/bin/linux/ubuntu/README.
Once R has been installed, type
R
at the command line to get an interactive R session
running. EdgeR is installed by using the commands,
pasted into the R console/command prompt (you may need
super user privileges to do this);
source("http://bioconductor.org/biocLite.R");
biocLite("edgeR")
EdgeR will be installed. It may ask you to update some
packages, type n if you don’t want to If super user is
needed for edgeR installation, quit R with;
q()
Type n if asked to save the session
Start R as the super user
sudo R
It will ask for the super user password (your system
admin should know this)
Reenter the above commands to install edgeR, then quit
R and restart with;
R
and then the differential expression analysis is run
# Load in the edgeR library
 Mapping the Avian Extracellular and Membrane Proteome 53

library(edgeR)
# Read in the counts table that was generated by
FeatureCounts into a data frame
dframe<-read.table("Endo_counts.txt", sep¼"\t", head-
er¼T, row.names¼1 skip=1)
# Make a data matrix from the data frame that only
contains the raw read to transcript counts (removes data
not needed)
datamat<-data.matrix(dframe[,-c(1:5)])
# Remove very low expressed genes, with anything >¼ raw
count of 20 across the six samples
counts_small <- datamat[rowSums(datamat)>20,]
# Create a DGElist object that holds the data.
dob <- DGEList(counts ¼ counts_small, group¼c(rep
("nonEndo",3),rep("Endo",3)))
# Calculate the normalization factors to scale the raw
counts of varying library sizes
dob <- calcNormFactors(dob)
# Estimates the common dispersion parameter (see the
edgeR documentation)
dob <- estimateCommonDisp(dob,verbose¼T)
# This function uses empirical Bayes to assess the
tagwise negative binomial dispersions (see the edgeR
manual)
dob <- estimateTagwiseDisp(dob)
# Perform endothelial versus non-endothelial library
contrast, so up regulated genes are enriched in en-
dothelial cells.
diff_ex <- exactTest(dob, pair¼c("nonEndo", "Endo"))
# Make a table of all the differentially expressed
genes results
out_table<-topTags(diff_ex,n¼Inf)[[1]]
write.table(out_table, sep¼"\t", quote¼F, fi-
le¼"Endo_output.exact.test.txt", col.names¼NA)
finish the R session by typing;
q()
type n if prompted to save the session

– Literature abstract scanning and the generation of a ranking

score

1. PubMed abstracts of the human orthologs are scanned for

endothelial and angiogenesis related keywords. Any matches
are totaled and divided by the total number of publications
for that gene to produce a percentage score, giving a rough
indication to a possible endothelial/vascular biological role.
The keywords are as follows: “vascular,” “tumour,”
“tumor,” “angiogenic,” “angiogenesis,” “neovasculariza-
tion,” “neovascularisation,” “vasculogenesis,” “hypoxia,”
54 Witold W. Kilarski et al.

“endoth,” and “VEGF.” A setup script is run to prepare files

needed for the PubMed literature abstract scanning.
chmod þx make_pubmed_hash.pl;
chmod þx pubmed_runner.pl
chmod þx pubmed_download.bash;
./pubmed_download.bash

2. Run the abstract scanning, preferably overnight as it can

take a long time for genes with a high number of
publications.
(nohup ./pubmed_runner.pl chick.out Endo_output.exact.
test.txt > final_results.txt )>& stderrs

Note; only run pubmed_runner.pl once at a time, or the

NCBI could block access to the computers IP address.
– Combining results and investigation of the best candidate genes
The results of pubmed_runner.pl contain the orthologs,
gene expression data and abstract scanning results, combined
into a single result file (final_results.txt). This is a tab delimited
file that can be opened in a spreadsheet program. Results can be
sorted by either logFC of endothelial to ovarian cancer cell line
expression or by the endothelial Rank percentage score, which is
a rough measure of the proportion of abstracts containing an
endothelial or angiogenic related keyword. From the example
chicken protein file provided (chick_proteins.fa, with eight pro-
teins), orthologs were found for all proteins. Ranking by logFC;
ECSCR, GPR116 (now ADGRF5) and PODXL all have >4
fold upregulation in endothelial cells vs. ovarian cancer cells
(with an FDR < 0.01). ECSCR also has the highest rank score
of 80%, with 12 out of the 15 publication abstracts containing
endothelial, vascular, or cancer-related keywords. Based on this
small analysis of very few chicken proteins, ECSCR could be
worth further investigation as an endothelial enriched gene,
which could have a role in angiogenesis and be a putative vascu-
lar target.

References

1. Rice JJ, Gerwins P, Kilarski WW (2012) 3. Dieterlen-Lievre F (1975) On the origin of

Mechanisms of angiogenesis: perspectives haemopoietic stem cells in the avian embryo:
from antiangiogenic tumor therapies. Curr an experimental approach. J Embryol Exp
Angiogenes 1:139–147 Morphol 33:607–619
2. Soulet F, Kilarski WW, Roux-Dalvai F, Herbert 4. Balaban E, Teillet MA, Le Douarin N (1988)
JMJ, Sacewicz I, Mouton-Barbosa E, Application of the quail-chick chimera system
Bicknell R, Lalor P et al (2013) Mapping the to the study of brain development and behav-
extracellular and membrane proteome asso- ior. Science 241:1339–1342
ciated with the vasculature and the stroma in 5. Djonov VG, Galli AB, Burri PH (2000) Intus-
the embryo. Mol Cell Proteomics susceptive arborization contributes to vascular
12:2293–2312 tree formation in the chick chorio-allantoic
membrane. Anat Embryol (Berl) 202:347–357
 Mapping the Avian Extracellular and Membrane Proteome 55

6. Kilarski W, Petersson L, Fuchs P, Zielinski M, 19. Magrane M, Consortium, U (2011) UniProt

Gerwins P (2012) An in vivo neovasculariza- Knowledgebase: a hub of integrated protein
tion assay for screening regulators of angiogen- data. Database (Oxford) 2011:bar009
esis and assessing their effects on pre-existing 20. Pruitt KD, Brown GR, Hiatt SM, Thibaud-
vessels. Angiogenesis 15:643–655 Nissen F, Astashyn A, Ermolaeva O, Farrell
7. Martino MM, Briquez PS, Guc E, Tortelli F, CM, Hart J et al (2014) RefSeq: an update on
Kilarski WW, Metzger S, Rice JJ, Kuhn GA et al mammalian reference sequences. Nucleic Acids
(2014) Growth factors engineered for super- Res 42(Database issue):D756–D763
affinity to the extracellular matrix enhance tis- 21. Altschul SF, Gish W, Miller W, Myers EW, Lip-
sue healing. Science 343:885–888 man DJ (1990) Basic local alignment search
8. Kilarski WW, Samolov B, Petersson L, tool. J Mol Biol 215:403–410
Kvanta A, Gerwins P (2009) Biomechanical 22. Herbert JM, Buffa FM, Vorschmitt H,
regulation of blood vessel growth during tissue Egginton S, Bicknell R (2009) A new proce-
vascularization. Nat Med 15:657–664 dure for determining the genetic basis of a
9. Clauser KR, Baker P, Burlingame AL (1999) physiological process in a non-model species,
Role of accurate mass measurement (þ/ illustrated by cold induced angiogenesis in the
10 ppm) in protein identification strategies carp. BMC Genomics 10:490
employing MS or MS/MS and database 23. Wheeler DL, Church DM, Lash AE, Leipe
searching. Anal Chem 71:2871–2882 DD, Madden TL, Pontius JU, Schuler GD,
10. Stern CD (2005) The chick; a great model Schriml LM et al (2001) Database resources
system becomes even greater. Dev Cell 8:9–17 of the National Center for Biotechnology
11. Rybak JN, Ettorre A, Kaissling B, Giavazzi R, Information. Nucleic Acids Res 29:11–16
Neri D, Elia G (2005) In vivo protein biotiny- 24. Altenhoff AM, Dessimoz C (2009) Phyloge-
lation for identification of organ-specific anti- netic and functional assessment of orthologs
gens accessible from the vasculature. Nat inference projects and methods. PLoS Comput
Methods 2:291–298 Biol 5:e1000262
12. Kamada H, Fugmann T, Neri D, Roesli C 25. Huminiecki L, Bicknell R (2000) In silico clon-
(2009) Improved protein sequence coverage ing of novel endothelial-specific genes.
by on resin deglycosylation and cysteine modi- Genome Res 10:1796–1806
fication for biomarker discovery. Proteomics 26. Herbert JM, Stekel D, Sanderson S, Heath VL,
9:783–787 Bicknell R (2008) A novel method of differen-
13. Sola RJ, Griebenow K (2009) Effects of glyco- tial gene expression analysis using multiple
sylation on the stability of protein pharmaceu- cDNA libraries applied to the identification of
ticals. J Pharm Sci 98:1223–1245 tumour endothelial genes. BMC Genomics
14. Gramer MJ, Goochee CF, Chock VY, Brous- 9:153
seau DT, Sliwkowski MB (1995) Removal of 27. Bennett S (2004) Solexa Ltd. Pharmacoge-
sialic acid from a glycoprotein in CHO cell nomics 5:433–438
culture supernatant by action of an extracellular 28. Edgar R, Domrachev M, Lash AE (2002) Gene
CHO cell sialidase. Biotechnology 13:692–698 Expression Omnibus: NCBI gene expression
15. Kilarski WW, Muchowicz A, Wachowska M, and hybridization array data repository.
Me˛żyk-Kopeć R, Golab J, Swartz MA, Nucleic Acids Res 30:207–210
Nowak-Sliwinska P (2014) Optimization and 29. Trapnell C, Pachter L, Salzberg SL (2009)
regeneration kinetics of lymphatic-specific pho- TopHat: discovering splice junctions with
todynamic therapy in the mouse dermis. RNA-Seq. Bioinformatics 25:1105–1111
Angiogenesis 17:347–357 30. Trapnell C, Williams BA, Pertea G,
16. Holmberg A, Blomstergren A, Nord O, Mortazavi A, Kwan G, van Baren MJ, Salzberg
Lukacs M, Lundeberg J, Uhlen M (2005) SL, Wold BJ, Pachter L (2010) Transcript
The biotin-streptavidin interaction can be assembly and quantification by RNA-Seq
reversibly broken using water at elevated tem- reveals unannotated transcripts and isoform
peratures. Electrophoresis 26:501–510 switching during cell differentiation. Nat Bio-
17. White PT (1974) Experimental studies on the technol 28:511–515
circulatory system of the late chick embryo. J 31. Langmead B, Salzberg SL (2012) Fast gapped-
Exp Biol 61:571–592 read alignment with Bowtie 2. Nat Methods
18. Gonzalez M, Argarana CE, Fidelio GD (1999) 9:357–359
Extremely high thermal stability of streptavidin 32. Roberts A, Pachter L (2013) Streaming frag-
and avidin upon biotin binding. Biomol Eng ment assignment for real-time analysis of
16:67–72
56 Witold W. Kilarski et al.

sequencing experiments. Nat Methods for assigning sequence reads to genomic fea-
10:71–73 tures. Bioinformatics 30:923–930
33. Dobin A, Davis CA, Schlesinger F, Drenkow J, 35. Robinson MD, McCarthy DJ, Smyth GK
Zaleski C, Jha S, Batut P, Chaisson M, Gingeras (2010) edgeR: a Bioconductor package for dif-
TR (2013) STAR: ultrafast universal RNA-seq ferential expression analysis of digital gene
aligner. Bioinformatics 29:15–21 expression data. Bioinformatics 26:139–140
34. Liao Y, Smyth GK, Shi W (2014) feature-
Counts: an efficient general purpose program
 Chapter 4

Mass Spectrometry-Based Identification of Extracellular

Domains of Cell Surface N-Glycoproteins: Defining
the Accessible Surfaceome for Immunophenotyping Stem
Cells and Their Derivatives
Chelsea M. Fujinaka, Matthew Waas, and Rebekah L. Gundry

Abstract
Human stem cells and their progeny are valuable for a variety of research applications and have the potential
to revolutionize approaches to regenerative medicine. However, we currently have limited tools to permit
live isolation of homogeneous populations of cells apt for mechanistic studies or cellular therapies. While
these challenges can be overcome through the use of immunophenotyping based on accessible cell surface
markers, the success of this process depends on the availability of reliable antibodies and well-characterized
markers, which are lacking for most stem cell lineages. This chapter outlines an iterative process for the
development of new cell surface marker barcodes for identifying and selecting stem cell derived progeny of
specific cell types, subtypes, and maturation stages, where antibody-independent identification of cell
surface proteins is achieved using a modern chemoproteomic approach to specifically identify
N-glycoproteins localized to the cell surface. By taking advantage of a large repository of available cell
surfaceome data, proteins that are unlikely to confer cell type specificity can be rapidly eliminated from
consideration. Subsequently, targeted quantitation by mass spectrometry can be used to refine candidates of
interest, and a bioinformatic visualization tool is key to mapping experimental data to candidate protein
sequences for the purpose of epitope selection during the antibody development phase. Overall, the process
of developing cell surface barcodes for immunophenotyping is iterative and can include multiple rounds of
discovery, refinement, and validation depending on the phenotypic resolution required.

Key words N-glycoproteins, Proteomics, Cell surface, Cell surface protein atlas, Immunophenotyp-
ing, Barcode, Mass spectrometry, Epitope selection, Extracellular domain

1 Introduction

Human stem cells and their progeny are valuable for a variety of
research applications and have the potential to revolutionize thera-
peutic options for treating intractable diseases. Specifically, stem
cells and their progeny can be used to study molecular dynamics
during very early stages of human development that are impossible
to study in vivo [1] and to study potentially toxic effects of

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_4, © Springer Science+Business Media, LLC 2018

57
58 Chelsea M. Fujinaka et al.

compounds in preclinical studies [2, 3]. It may also become possi-

ble to use stem cell progeny to repair damaged tissue and improve
organ function by transplanting cells, tissue constructs, or secreted
factors [4, 5]. Furthermore, patient-specific cells can be used to
study cellular phenotypes and molecular mechanisms resulting
from specific genetic mutations and may promote the advancement
toward patient-specific medical therapy (i.e., personalized medi-
cine). Despite promising preliminary results across a broad spec-
trum of lineages and applications, we currently have limited tools to
permit the live isolation of homogeneous cell populations for the
majority of stem cells or their progeny destined for mechanistic
studies or cellular therapies.
Immunophenotyping is an approach where antibody-based
detection of cell surface makers is used to identify, select, and
quantify populations of cells with defined functional phenotypes.
For example, immunophenotyping based on extracellular epitopes
has enabled reproducible, nonmutagenic, and high-throughput live
cell isolation and assessments of hematopoietic stem cell lineages
[6, 7], but it has been underutilized in most other stem cell types.
Once developed, immunophenotyping can be used both for live
cell identification in quality control processes to assess cell culture
purity and for live cell sorting of desired subpopulations. This
strategy has tremendous potential to overcome current limitations
and enable the isolation of homogeneous cell-type and maturation-
stage specific populations of stem cells and their progeny. However,
the success of this process depends on the availability of reliable
antibodies and well-characterized markers, which are lacking for
the majority of stem cell lineages. The lack of suitable markers can,
in part, be addressed through antibody screening although there is
limited availability of high quality antibodies to surface accessible
epitopes. While approximately 370 antigens are targeted by mono-
clonal antibodies to cluster of differentiation (CD) molecules [8],
this represents only a fraction of all molecules present on the cell
surface (i.e., the surfaceome) and CD molecule screening has had
limited success for nonhematopoietic stem cells [9–11]. Alternative
approaches have been employed to evaluate the cell surface protein
landscape based on transcriptomics [12], and physical and affinity
enrichment strategies coupled with proteomics [13, 14], although
each has limitations. Transcript abundance does not always corre-
late with protein abundance on the cell surface, transcriptomic data
fail to inform subcellular localization, and the precise location of
many membrane proteins are yet unannotated. This is further
confounded by the fact that membrane proteins can be sequestered
within the cell and only cycled to the cell surface in response to
developmental events or cell signaling (e.g., aquaporins in kidney)
[15]. Historically, direct identification and assessment of cell sur-
face proteins using proteomic approaches has been challenging due
to the relatively low abundance, hydrophobicity, and difficulty of
 Cell Surface Capture Technology 59

purifying plasma membrane proteins away from intracellular mem-

branes by biophysical methods. However, modern chemoproteo-
mic strategies are overcoming these challenges and providing
previously unprecedented views of the surfaceome of a variety of
mammalian cell types [16–20]. When coupled with contemporary
bioinformatics tools for comparing surfaceomes among cell types
and mapping details of extracellular domains of surfaceome pro-
teins, this approach can efficiently identify new markers and facili-
tate antibody development for the purpose of live cell
immunophenotyping.
This chapter details the steps involved in the discovery and
refinement phases of an iterative process (Fig. 1) for the develop-
ment of new cell surface marker barcodes for identifying and select-
ing stem cell derived progeny of specific cell types, subtypes, and
maturation stages, where antibody-independent identification of
cell surface proteins is achieved using a modern chemoproteomic
approach to specifically identify N-glycoproteins localized to the
cell surface. By taking advantage of a large repository of cell surfa-
ceome data already available, proteins that are unlikely to provide
any cell type specificity can be rapidly eliminated from consider-
ation. Subsequently, targeted quantitation by mass spectrometry
can be used to refine candidates of interest, and a bioinformatic
visualization tool is key to mapping experimental data to candidate
protein sequences for the purpose of epitope selection for antibody
development. Overall, the process of developing cell surface bar-
codes is iterative and can include multiple rounds of discovery,
refinement, and validation depending on the phenotypic resolution
required.
In the first step of the process, the Cell Surface Capture (CSC)-
Technology is used to identify extracellular epitopes of cell surface
N-glycoproteins. Specifically, protein identity, transmembrane ori-
entation and glycosylation site occupancy are simultaneously deter-
mined in a CSC-Technology experiment. The approach, originally
developed by Wollscheid and colleagues [21], takes advantage of
the prediction that >90% of cell surface proteins are glycosylated
[22, 23]. Experimentally, extracellular oligosaccharides on live cells
are oxidized and biotinylated using membrane-impermeable
reagents. Proteins are enzymatically digested and the resulting
biotinylated glycopeptides are enriched using streptavidin beads.
Through the actions of the enzyme peptide-N-glycosidase F
(PNGaseF), which cleaves between the innermost GlcNAc and
asparagine residues, deglycosylated peptides are released and ana-
lyzed by high mass accuracy mass spectrometry. During the data
analysis process, cell surface proteins are confidently identified by
the signature deamidation (+0.984016 Da) observed on the aspar-
agine residue within the conserved N-glycosylation sequence motif
(Asn-x-Ser/Thr/Cys (NxS/T/C; where x is any amino acid except
proline). Overall, the CSC-Technology generates a highly specific
60 Chelsea M. Fujinaka et al.

Fig. 1 Iterative process for identifying extracellular epitopes and developing cell surface marker barcodes for
live immunophenotyping. In the discovery phase, combining the CSC-Technology with the Cell Surface Protein
Atlas provides a rapid method for identifying proteins that may be informative for identifying or selecting a
specific cell type. Once putative candidates are selected, the refinement phase includes antibody-independent
quantitation by targeted mass spectrometry to further refine those markers that are potentially most
informative. Subsequently, Protter is used to visualize extracellular domains and guide epitope selection for
antibody development. In the validation phase, monoclonal antibodies are validated and functional and
molecular measurements can be performed on the cells identified by the antibodies in order to correlate
the functional phenotype with the cell surface phenotype

“snapshot” of the cell surfaceome at a particular time or stage.

When performed in a comparative manner, it is possible to use
the CSC-Technology to identify proteins that are cycled to the
surface upon a specific stimulus or experimental condition. The
CSC-Technology is thereby advantageous over predictive
approaches or those that rely on more generic membrane protein
enrichment strategies.
Extracellular epitopes from hundreds of cell surface N-glyco-
proteins can be identified in a single CSC-Technology experiment,
and when compared with other datasets, the output facilitates the
 Cell Surface Capture Technology 61

identification of novel and ubiquitous cell surface protein markers.

Given the power of this approach, CSC-Technology data from over
80 mouse and human cell types (normal and diseased) have now
been deposited into a publicly accessible repository, the Cell Sur-
face Protein Atlas (CSPA; http://wlab.ethz.ch/cspa/) [24]. When
these data are analyzed comparatively, classification of a particular
protein as routinely vs. rarely observed among cell types can be
made immediately. Although the presence of a protein among
disparate cell types may not preclude its utility for a specific context
in vitro, this type of analysis can rapidly focus a candidate list from
hundreds to tens in cases where a cell type specific marker is desired.
In cases where proteins unique to a single cell type are rare, it is
expected that panels of surface markers will be necessary for ade-
quate barcoding and sorting well-defined cell populations [24, 25],
and in this context the CSPA is also valuable for highlighting
potential marker combinations that can be further examined in
the next step (Fig. 1).
Once proteins of interest are identified, contemporary targeted
mass spectrometry-based approaches for quantifying hundreds of
targets in a single experiment [26] are critical to further refining
newly discovered targets. Although antibody-based approaches
such as immunofluorescence imaging and flow cytometry offer
the ability to monitor protein abundance levels and can inform
localization, they require validated and specific antibodies that are
often not available for cell surface proteins. Moreover, determining
optimal experimental conditions for each cell type of interest and
validating antibody specificity requires significant time and effort.
As with all antibody-based methods, even if highly specific antibo-
dies are available, they require the epitope to be accessible, which
may not be predictable among cell types, as unexpected conforma-
tional changes and post-translational modifications can mask epi-
tope availability. For these reasons, we use targeted mass
spectrometry-based approaches for simultaneously quantifying
tens to hundreds of proteins of interest to further refine the candi-
dates prior to downstream antibody-based studies. Ultimately, high
quality antibodies (e.g., recombinant) that specifically recognize
extracellular epitopes on live cells are necessary to advance immu-
nophenotyping and sorting of stem cell subpopulations in a way
that is analogous to the isolation of therapeutically relevant hema-
topoietic stem cell populations [6]. For this purpose, the
CSC-Technology is especially advantageous as the experimental
output confirms the occupancy of individual N-glycosylation sites
of identified proteins and thereby confirms extracellular domains.
In the final step of the process outlined in this chapter, data from
the CSC-Technology are imported into the visualization tool Prot-
ter [27]. This approach can be used to confirm or correct predicted
protein orientations and highlight occupied N-glycosites and other
known modifications and sequence features that can be avoided
62 Chelsea M. Fujinaka et al.

during epitope selection. Additional bioinformatics analyses of

identified glycoproteins can be accomplished using Targets-search
as described in Chapter 19 of this book by Yan et al., In the
validation phase, which is beyond the scope of this chapter and is
also the most time-consuming, antibodies are validated for epitope
and cell type specificity, and extensive functional analyses of the cells
identified and selected by cell surface markers are performed. When
successful, the end result is a barcode of cell surface proteins that
can be correlated with specific cellular functions.
Overall, the iterative seven-step workflow described here is
efficient for the identification of cell surface N-glycoproteins that
can be used for live cell immunophenotyping. The steps described
in this chapter include (Subheading 3.1) cell collection; (Subhead-
ing 3.2) oxidation and biotinylation; (Subheading 3.3) cell lysis and
membrane enrichment; (Subheading 3.4) tryptic digestion; (Sub-
heading 3.5) glycopeptide capture and elution; (Subheading 3.6)
desalting and concentration of peptides; and (Subheading 3.7)
mass spectrometry with data analysis. In Subheading 3.8, we
show how to prioritize epitopes for marker and antibody develop-
ment. We have used this strategy to identify markers and to gener-
ate new antibodies for a variety of stem cells and their progeny,
including: selecting pluripotent stem cell-derived hepatocytes [28]
and cardiomyocytes ([29] and in preparation); isolating induced
pluripotent stem cells reprogrammed from mouse fibroblasts [30];
and identifying proteins in human pluripotent stem cells that can be
targeted by small molecules for the purpose of eliminating tumori-
genic cells [25]. Although our previous work and this chapter have
focused on stem cells, the approach is equally applicable to other
mammalian cell types, as demonstrated by others [16, 20, 24,
31–33], and variations of this approach that rely on cysteine or
lysine containing peptides have also been described [31].

2 Materials

Prepare all solutions with ultrapure water and analytical grade

reagents, including liquid chromatography–mass spectrometry
(LC-MS) and high-performance liquid chromatography (HPLC)
grade water. Low-retention tubes should be used when possible.

2.1 Cell Collection 1. Phosphate buffered saline (PBS, 1
): 1
PBS pH 7.4. Add
900 mL water to 100 mL 10
PBS, pH 7.4 (Quality
Biological, Gaithersburg, MD, USA). Store at 4  C.
2. Liberase TH dissociation solution: Liberase TH (Roche Diag-
nostics, Indianapolis, IN, USA) (see Note 1).
3. PBS with 0.1% fetal bovine serum (FBS): 1
PBS pH 7.4, 0.1%
(v/v) FBS. Add 1 mL FBS (Sigma-Aldrich Chemical Company,
 Cell Surface Capture Technology 63

St. Louis, MO, USA) to 100 mL 10
PBS and fill to 1 L with
water. Store at 4  C.
4. 15 and 50 mL conical tubes.
5. Low speed centrifuge capable of holding 15 and 50 mL conical
tubes.

2.2 Oxidation 1. Labeling buffer: 1
PBS, 0.1% (v/v) FBS, pH 6.5. Add 500 μL
and Biotinylation FBS to 50 mL 10
PBS, fill to 500 mL with water, and adjust
of Cell Surface the pH to 6.5 using 85% (w/v) H3PO4. Store at 4  C (see
Oligosaccharides Notes 2 and 3).
2. NaIO4 stock solution: 195 mM NaIO4 in labeling buffer.
Weigh 0.01 g NaIO4 (Sigma-Aldrich) in a microcentrifuge
tube, add 240 μL labeling buffer and vortex until dissolved
(see Note 2).
3. Biocytin hydrazide, 25 mg aliquots (Biotium, Hayward, CA,
USA) (see Note 4).
4. Laboratory rocker.

2.3 Cell Lysis 1. Hypotonic lysis buffer: 10 mM Tris–HCl pH 7.5, 0.5 mM

and Membrane MgCl2. Combine 5 mL 1 M Tris–HCl pH 7.5 and 2.5 mL
Enrichment 100 mM MgCl2. Fill to 500 mL and store at 4  C.
2. GentleMACS M tubes and GentleMACS Dissociator (Miltenyi
Biotec, San Diego, CA, USA).
3. Membrane prep buffer: 280 mM sucrose, 50 mM MES
pH 6.5, 450 mM NaCl, 10 mM MgCl2. Combine 47.92 g
sucrose (Sigma-Aldrich), 25 mL 1M 2-(N-morpholino)-etha-
nesulfonic acid (MES hydrate)-NaOH pH 6.5 (Boston Bio-
Products, Ashland, MA, USA), 45 mL 5 M NaCl (Sigma-
Aldrich), 50 mL 100 mM MgCl2. Fill to 500 mL with water
and store at 4  C.
4. Optima LE-80K ultracentrifuge, SW41 Ti rotor, and
9/16
3½ inch Ultra-clear tubes (Beckman Coulter, Brea,
CA, USA) or equivalent.
5. Thermomixer R with block for 15 mL conical tubes (Eppen-
dorf, Hauppauge, NY, USA).

2.4 Tryptic Digestion 1. 100 mM NH4HCO3 buffer: 100 mM NH4HCO3. Weigh

0.39 g NH4HCO3 (Sigma-Aldrich) and add water to 50 mL
(see Note 2).
2. 1% AALS-I stock solution: 1% (w/v) Anionic acid labile
surfactant-I (AALS-I) (Protea Biosciences, Inc., Morgantown,
WV, USA). Dissolve 5 mg AALS-I in 500 μL LC-MS grade
water. Store at 4  C (see Note 5).
64 Chelsea M. Fujinaka et al.

3. Tris(2-carboxyethyl) phosphine (TCEP) stock solution:

100 mM TCEP. Weigh 0.29 g TCEP (Sigma-Aldrich) and
add water to 10 mL. Store 0.5 mL aliquots at 20  C.
4. Iodoacetamide stock solution: 100 mM iodoacetamide. Weigh
0.184 g iodoacetamide (Sigma-Aldrich) and add water to
10 mL (see Note 2).
5. Sequencing grade-modified trypsin (Promega, Madison, WI,
USA). Store at 80  C.
6. Six inch Pasteur glass pipette.
7. Parafilm.

2.5 Glycopeptide 1. 1 mL Mobicol columns with 35 μm pore, large filters (MoBi-

Capture and Elution Tec, Goettingen, Germany).
2. Vac-man laboratory vacuum manifold (Promega) and vacuum
pump or in house vacuum line.
3. Streptavidin Plus UltraLink Resin (Pierce, Rockford, IL, USA).
Store at 4  C.
4. 20 mL BD Luer-Lok Disposable Syringes without needles
(BD Biosciences, Franklin Lakes, NJ, USA).
5. Luer-lock caps (MoBiTec, Goettingen, Germany).
6. One-way luer-lock stopcocks (Promega).
7. Invitrosol stock solution: 0.1
Invitrosol in 100 mM
NH4HCO3. Combine 500 μL 5
Invitrosol (ThermoFisher
Scientific, Waltham, MA, USA) and 24.5 mL 100 mM
NH4HCO3. Store stock solution in 500 μL aliquots at 20  C.
8. 5 M NaCl solution (ThermoFisher Scientific).
9. 100 mM Na2CO3 solution: 100 mM Na2CO3. Add 1 g
Na2CO3 to 50 mL water and vortex until dissolved.
10. 80% isopropanol: 80% (v/v) 2-propanol. Mix 400 mL
2-propanol and 100 mL water.
11. 10 units/μL peptide N-deglycosylase F, glycerol-free
(PNGase F) (Promega).
12. 0.2% TFA solution: 0.2% (v/v) trifluoroacetic acid (TFA).
Combine 2 mL TFA in 1 L LC-MS grade water. Store in the
dark at room temperature.
13. End over end Shaker.

2.6 Desalting 1. Hydrophobic packing material C18 Micro SpinColumns with

and Concentration 5–50 μg capacity (Harvard Apparatus, Holliston, MA, USA).
of Peptides 2. 0.1% TFA solution: 0.1% (v/v). Combine 1 mL TFA in 1 L
LC-MS grade water.
3. Elution solution: 70% (v/v) LC-MS grade acetonitrile with
0.1% formic acid, 30% (v/v) LC-MS grade water with 0.1%
 Cell Surface Capture Technology 65

formic acid. For 100 mL of solution, combine 70 mL acetoni-

trile with 0.1% formic acid and 30 mL water with 0.1%
formic acid.
4. Savant SpeedVac concentrator and UVS400 vacuum pump
(ThermoFisher Scientific).
5. 0.5 and 1.5 mL low-binding microcentrifuge tube (Thermo-
Fisher Scientific).

2.7 Mass 1. Nanospray liquid chromatography-tandem mass spectrometry

Spectrometry (nLC-MS/MS).
Instrumentation 2. Reprosil-PUR C18-AQ, 3 μm, 120 Å (Dr. Maisch; Ammer-
and Analysis Software buch, Germany) or PicoChip (New Objective, Woburn, MA).
3. Eksigent nanoLC Ultra 2D system (Eksigent Technologies,
Dublin, CA, USA) or comparable.
4. Q Exactive hybrid quadrupole orbitrap (ThermoFisher Scien-
tific) or other high mass accuracy instrumentation.
5. Proteome Discoverer 2.1 (ThermoFisher Scientific) with
SequestHT and MSAmanda search algorithms.
6. Percolator [34] and ptmRS [35] are publicly available bioinfor-
matic tools for post-search peptide scoring and post-transla-
tional modification localization, respectively. These can be used
as nodes within Proteome Discoverer.
7. SPSS Statistics (IBM) or other software package capable of
hierarchical clustering.

3 Methods

3.1 Collect Cells 1. Begin with a culture of approximately 108 cells (see Note 6). If
and Oxidize working with adherent cells, wash cells with room temperature
Extracellular PBS to remove as many dead cells as possible. If working with
Oligosaccharides cells in suspension, proceed to step 3.
2. Detach cells from the culture plate using a trypsin-free solu-
tion. The choice of this solution may vary by cell type, and can
include collagenases, enzyme free, and/or EDTA solutions.
Scraping may be possible, but this must be first tested with a
small aliquot of cells, and cells should be stained with trypan
blue to evaluate plasma membrane integrity (see Notes 1
and 7).
3. Transfer cells and solution into a 50 mL conical tube. Centri-
fuge at 500
g at 4  C for 3 min to collect cells.
4. Aspirate liquid and gently resuspend cells by first tapping
against benchtop to loosen the pellet. Add 10 mL of cold
PBS with 0.1% FBS and pipette up/down three times (pipette
tip close to wall of tube to break up clumps). Remove a 50 μL
66 Chelsea M. Fujinaka et al.

aliquot for counting, then fill conical tube to 50 mL with PBS

with 0.1% FBS. Centrifuge 500
g at 4  C for 3 min to collect
cells.
5. Aspirate liquid and gently resuspend cells by first tapping
against benchtop to loosen the pellet. Add 20 mL cold labeling
buffer. Keep tube on ice.
6. Add 100 μL NaIO4 stock solution to cells/labeling buffer
solution. Place cells on a very slow rocker in the dark for
15 min on ice or at 4  C.
7. Add cold labeling buffer up to 50 mL to dilute solution.
Centrifuge 500
g at 4  C for 3 min to collect cells.
8. Aspirate liquid. Gently resuspend cells by first tapping against
benchtop to loosen the pellet. Add cold labeling buffer up to
50 mL. Centrifuge at 500
g at 4  C for 3 min to collect cells.
Repeat for a total of two washes to remove NaIO4.

3.2 Biotinylate 1. Aspirate liquid. Gently resuspend cells by first tapping against
Extracellular the benchtop to loosen the pellet. Add 4 mL cold labeling
Oligosaccharides buffer.
2. Add 1 mL cold labeling buffer to one vial of biocytin hydrazide
(25 mg). Vortex briefly until resuspended. Add all of this to the
cell solution. Use 1 mL of cold labeling buffer to rinse the
biocytin hydrazide vial and subsequently add this to the cell
suspension. The final concentration will be approximately
10 mM biocytin hydrazide (5–10 mM is optimal).
3. Place tube in ice on a rocker to agitate slowly for 60 min.
4. Add PBS with 0.1% FBS up to 50 mL and invert several times
to mix. Centrifuge 500
g at 4  C for 3 min to collect cells.
5. Aspirate liquid. Gently resuspend cells by first tapping against
benchtop to loosen the pellet. Add PBS with 0.1% FBS up to
50 mL. Centrifuge at 500
g at 4  C for 3 min to collect cells.
Repeat for a total of two washes.
6. At this step, either flash-freeze and store at 80  C for short
term storage or proceed with next step directly.

3.3 Lyse Cells, 1. Aspirate PBS, and ensure that all the PBS is removed. Resus-
Remove Nuclei, pend cells in 4 mL hypotonic lysis buffer. Set on ice for 10 min
and Enrich Membranes to swell the cells making it easier to lyse them.
2. Using a serological pipette, transfer cell solution to Orange
Tube (M tube) for GentleMacs. Homogenize cells twice,
using method D01 both times (see Note 8).
3. Centrifuge GentleMacs tube at 800
g at 4  C for 10 min to
remove nuclei/debris. Transfer supernatant to a new 15 mL
conical tube, and save on ice until step 5.
 Cell Surface Capture Technology 67

4. Using a serological pipette, add 2 mL hypotonic lysis buffer to

the pellet remaining in the GentleMacs tube. Homogenize cell
pellet twice, using method D01 both times (see Note 8).
5. Centrifuge both the GentleMacs tube and the 15 mL conical
tube with the supernatant from step 3 at 800
g at 4  C for
10 min to remove remaining nuclei/debris. Combine the
supernatant from the GentleMacs tube and the supernatant
from the 15 mL conical into an ultracentrifuge tube. The
total volume should be 6 mL.
6. Add an equal volume (6 mL) of membrane preparation buffer.
Balance a second tube to within 0.01 g using membrane prep-
aration buffer. Centrifuge in an ultracentrifuge at 210,000
g
at 4  C overnight (18 h) in an SW41 Ti rotor.
7. Aspirate liquid. Add 200 μL membrane wash buffer to the
pellet. Vortex in thermomixer at 750 rpm at 4  C for 30 min.
8. Add hypotonic lysis buffer until the ultracentrifuge tube is full
and then balance a second tube to within 0.01 g.
9. Centrifuge in ultracentrifuge at 210,000
g at 4  C for
30 min. Aspirate liquid.

3.4 Tryptic Digestion 1. Add 300 μL of 100 mM ammonium bicarbonate, 40 μL 1%

AALS-I MS-compatible surfactant (to make final ~0.1% AALS-
I solution), 25 μL 100 mM TCEP (5 mM final) to the ultra-
centrifuge tube. Place tube on thermomixer at 37  C at
750 rpm for 30 min to reduce proteins.
2. Add 45 μL 100 mM iodoacetamide (10 mM final). Place on
thermomixer, covered with Parafilm for 30 min to alkylate
proteins in the dark.
3. Add the contents of one vial (20 μg) of trypsin. Check pH and
adjust to ~8.5 if necessary. Place on thermomixer overnight (see
Note 9).
4. Lower the sample pH to 2 by adding one drop of 85% w/v
phosphoric acid using a glass Pasteur pipette (~5 μL) to inacti-
vate trypsin and precipitate lipids. The solution will become
turbid.
5. Centrifuge at 16,000
g for 10 min to remove any nondi-
gested material. Typically, an “oily” pellet will be observed. Use
only supernatant to carry on to Subheading 3.5, step 4.

3.5 Glycopeptide 1. Place an empty Mobicol column on the Vacman, using the
Capture and Elution stopcock as shown in Fig. 2a. Rinse Mobicol with
5
500 μL acetonitrile to remove polymer contamination,
then rinse with 5
500 μL 100 mM ammonium bicarbonate.
2. Add 350 μL streptavidin bead slurry to Mobicol column. Wash
beads with 3
500 μL 100 mM ammonium bicarbonate.
68 Chelsea M. Fujinaka et al.

Fig. 2 Vacuum manifold setup. Shown are the configurations for using the Mobicol (A, B) for glycopeptide
capture and washing and the MicroSpin column (C) used for peptide desalting and concentration

3. Remove Mobicol from Vacman. Insert plug into bottom of

Mobicol and ensure it is snug.
4. Transfer peptide mixture to the Mobicol containing streptavi-
din beads. Do not add any of the pellet if there is one. Replace
screw cap and ensure end plug is snug. Wrap entire tube in
parafilm—wrap bottom to top to ensure plug remains tight.
Incubate 1 h at room temperature with end-over-end rotation.
5. Prewash three Mobicol spin columns by following step 1. Add
200 μL water to Mobicol and store temporarily to avoid drying
out the membrane until use.
6. Assemble washing apparatus by attaching Mobicol to syringe
via luer lock adaptor as shown in Fig. 2b. Transfer peptide/
bead slurry to Mobicol column #1. Each time the beads are
transferred to a new Mobicol, rinse the previous Mobicol with
3
300 μL of the solution being used in the subsequent
washing step to ensure all beads are transferred.
7. Wash beads with 10 mL 0.01
Invitrosol in 100 mM ammo-
nium bicarbonate.
8. Change Mobicol (Mobicol #2). Wash beads with 15 mL of 5 M
NaCl, then 15 mL of 100 mM sodium carbonate, then 15 mL
of 80% isopropanol.
9. Change Mobicol (Mobicol #3). Wash beads with 15 mL of
100 mM ammonium bicarbonate. Remove Mobicol from Vac-
man. Insert plug into bottom of Mobicol and ensure it is snug.
 Cell Surface Capture Technology 69

10. Add 350 μL of 100 mM ammonium bicarbonate to beads in

Mobicol, followed by the addition of 3 μL PNGaseF. Replace
screw cap and ensure end plug is snug. Wrap entire tube in
parafilm, wrapping from bottom to top to ensure plug remains
tight. Incubate at 37  C overnight with end-over-end rotation.
11. Unplug Mobicol and place into clean 1.5 mL microfuge tube.
Centrifuge at 1000
g for 30 s to collect liquid which contains
cleaved peptides.
12. Add 350 μL 0.2% TFA to beads and vortex gently to resuspend
beads. Place Mobicol into clean 1.5 mL microfuge tube. Cen-
trifuge at 1000
g for 30 s to collect liquid.
13. Combine liquid from steps 11 and 12 into a single 1.5 mL
microfuge tube.

3.6 Desalting 1. Cut off approximately 25 mm of a P200 pipette tip and use as
and Concentrating an adaptor to place the MicroSpin column on the vacuum
manifold (Fig. 2c).
2. Rinse column with 5
200 μL 100% HPLC grade acetonitrile
then 10
200 μL HPLC grade 0.1% TFA.
3. Place column into a clean 1.5 mL low-binding
microcentrifuge tube.
4. Add 200 μL of peptide sample to the top of the column. Do
not add any pellet if there is one. Using a P200 pipette, slowly
“push” sample through the column, from top to bottom.
Repeat until all of the sample is through the column twice.
5. Place column on vacuum manifold. Rinse column with
10
200 μL HPLC grade 0.1% TFA, then 10
200 μL
HPLC grade water.
6. Place column into a clean 0.5 mL low-binding microcentrifuge
tube. Add 100 μL of elution solution. Use a P200 pipette,
slowly “push” sample through the column, from top to bot-
tom. Centrifuge tube/column at 1000
g for 30 s to collect
remaining eluent.
7. Place the sample in a SpeedVac or similar instrument until all of
the liquid has evaporated.
8. Resuspend peptides in 12 μL LC-MS grade 0.1% formic acid in
water.

3.7 Mass 1. Analyze peptides by nanospray liquid chromatography tandem

Spectrometry and Data mass spectrometry (LC-MS/MS). We typically use a 3 cm trap
Analysis column followed by a 10 cm analytical column, both 75 μm
inner diameter packed with Reprosil-PUR C18-AQ, 3 μm,
120 Å, either packed in house or in a prepacked PicoChip
(New Objective, Woburn, MA). We typically perform two
technical replicate injections of 5 μL peptides using a gradient
70 Chelsea M. Fujinaka et al.

of mobile phase A (water, 0.1% formic acid) and mobile phase B

(acetonitrile, 0.1% formic acid) of 3–20% B over 55 min,
20–30% B over 15 min, 30–50% B over 4 min, 50–98% B
over 4 min at a flow rate of 300 nL/min (Eksigent nanoLC
Ultra 2D system). Comparable chromatographic systems and
settings will work equally well.
2. Data are best acquired using a high mass accuracy mass spec-
trometer with low limits of detection. In our case, this is a Q
Exactive hybrid quadrupole orbitrap. The instrument is oper-
ated in data dependent mode to automatically switch between
full scan MS and MS/MS acquisition. Full scan MS spectra are
acquired in the Orbitrap with 70,000 resolution, accumulation
of ions to a 1
106 target value based on predictive AGC, and
dynamic exclusion is set to 30 s. The 15 most intense multiply
charged ions (2  z  8) are sequentially isolated and fragmen-
ted in the octopole collision cell by higher energy collisional
dissociation with a target value of 1
104 ions, maximum
injection time of 120 ms and 35,000 resolution. Typical con-
ditions are as follows: spray voltage of 2 kV, no sheath and
auxillary gas flow, heated capillary temperature 275  C, nor-
malized collision energy 25%, underfill ratio set to 0.1%, 2.0 m/
z isolation window. Comparable instrumentation that offers
high mass accuracy and low limits of detection will work
equally well.
3. Raw data files from the MS instrument are searched via Pro-
teomeDiscoverer using SequestHT and MSAmanda [36]
search algorithms, Percolator [34] for post-search validation
and ptmRS [35] for added confidence in localizing the site of
deamidation (indicating the occupied N-glycosite), which can
be especially helpful for peptides containing two asparagine
residues (see Note 10). The search is performed against the
SwissProt version of the appropriate organism database from
UniProt (http://www.uniprot.org/). Search settings include
the following: semitryptic digestion, MS1 tolerance of
10 ppm, MS2 tolerance of 0.02 Da, static modification of
57.021464 Da on cysteines representing carbamidomethyla-
tion from iodoacetamide treatment, dynamic modifications of
15.995 Da on methionine representing oxidation, and
0.984016 Da on asparagines representing deamidation result-
ing from cleavage of the N-linked oligosaccharide from the
peptide backbone via PNGaseF treatment.
4. If the experiment is successful, 200–400 cell surface N-glyco-
proteins will be identified. Overall, more than 85% of the pep-
tides identified should contain a deamidation at the asparagine
within the sequence motif NxS/T/C (Fig. 3a) (see Note 11).
 Cell Surface Capture Technology 71

Fig. 3 Exploiting the Cell Surface Protein Atlas to identify proteins of interest. (a) Annotated MS/MS spectrum
of a representative peptide from Integrin Beta 1 that was identified in human pluripotent stem cell derived
72 Chelsea M. Fujinaka et al.

5. Export high confidence identifications from ProteomeDiscov-

erer or ProteinProphet as a .protXML file, which is the file
format used in downstream analyses.

3.8 Prioritization 1. To begin the selection of candidate makers that may be infor-
of Epitopes for Marker mative for a particular cell type, we utilize ProteinCenter to
and Antibody archive all data generated by the CSC-Technology. This work-
Development flow enables comparisons of cell surface N-glycoproteins
observed among various experimental conditions (e.g., cell
types, organ types, and developmental stages). We maintain
an in-house repository of data from three sources: data gener-
ated in-house, data from the Cell Surface Protein Atlas (Atlas
[24]; downloaded from http://wlab.ethz.ch/cspa/), and data
from other publications that may not be contained in the Atlas.
While proteins of interest for distinguishing cellular pheno-
types can be identified by directly including the desired cell
types for comparison within a quantitative CSC-Technology
experiment, the Atlas may still add value as a freely available
repository populated with data from a variety of cell types.
2. To establish a custom database of cell surface N-glycoproteins,
data from the Atlas and other publications are imported into
ProteinCenter via .csv file format. Data generated in-house are
imported as prot.xml files.
3. Once all data are imported into ProteinCenter, select the
desired datasets and create a comparison. At this stage, con-
taminating proteins identified in the experiment can be elimi-
nated by data filtering to keep only those proteins identified by
peptides containing a deamidated asparagine within the con-
served sequence motif (NxS/T/C) for N-glycosylation. To
eliminate redundancy and enable comparisons among datasets
that may have been generated using different protein data-
bases, cluster data within the comparison by “indistinguishable
proteins,” which will group proteins that cannot be

Fig. 3 (continued) cardiomyocytes via the CSC-Technology. One identified site of N-glycosylation (N115) within
the sequence motif is indicated. Spectrum generated using Proteome Discoverer 2.1 (ThermoFisher Scien-
tific). Inset table shows m/z values for fragment ions predicted to result from the proposed peptide sequence
NPCTSEQNCTSPFSYK. Values in red or blue were observed in the MS/MS spectrum. (b) Screenshot of
ProteinCenter repository of CSC-Technology data for 59 human cell types. Color denotes that a protein (one
protein/row) was identified in a particular cell type (one cell type/column). Data are shown for 20 cell surface
N-glycoproteins, including those ubiquitously observed across most cell types, and those that are observed
less commonly among those cell types represented in the Atlas. (c) Hierarchical clustering of 15 human cell
types using 1323 cell surface N-glycoproteins identified via CSC-Technology can successfully cluster similar
cell types
 Cell Surface Capture Technology 73

distinguished based on the peptide evidence they share. This

comparison provides a rapid strategy to eliminate proteins
widely observed across many cell types, and can highlight
proteins observed in only a few cell types (Fig. 3b).
4. In addition to the binary comparisons described above, hierar-
chical clustering of the Atlas data can be used to identify
combinations of proteins that may be informative for a partic-
ular cell type. An example is shown in Fig. 3c, where hierarchi-
cal clustering was performed using SPSS Statistics V22 using
binary representation (1 ¼ detected, 0 ¼ not detected) of 1323
proteins from 15 human cell types. The furthest neighbor (i.e.,
complete linkage) cluster method was used in conjunction with
the Dice similarity measure for determination of clusters. The
Dice measure was used because of its exclusion of joint
absences from consideration and the double weight it applies
to matches. Together, these features are appropriate to address
the limitations of working with discovery-based nonquantita-
tive proteomics data (i.e., “not observed” does not confirm
absence) and this strategy can reveal proteins potentially infor-
mative for distinguishing among protein groups. For example,
the clustering result in Fig. 3c shows pluripotent stem cells
cluster closely to their derivatives and to fibroblasts (the pre-
cursor from which the induced pluripotent stem cells were
derived). Of the proteins of interest revealed by this strategy,
we have previously used flow cytometry to confirm 12 as either
positive or negative markers of pluripotency when compared to
fibroblasts [25].
5. Proteins of interest identified in the data dependent discovery
process described above can be quantified using targeted mass
spectrometry approaches (e.g., multiple reaction monitoring,
parallel reaction monitoring) to further refine the relative or
absolute quantity among cell types, subtypes, and/or develop-
mental stages in an antibody independent manner. This step
can be especially important if it is necessary to analyze cell types
that cannot be obtained in sufficient quantity to permit
discovery-based analyses (e.g., <1E6 cells). An example of
how parallel reaction monitoring was used to quantify integrin
beta 1 and glucose transporter 3 simultaneously across three
time points of human pluripotent stem cell-derived cardiomyo-
cyte differentiation is shown in Fig. 4a. Similar approaches can
be performed to quantify tens to hundreds of targets within a
single experiment and can be performed using whole cell lysate,
membrane enriched fractions, or N-glycopeptides captured via
the CSC-Technology (see Note 12).
6. When selecting epitopes to be used for generating antibodies
for the purpose of live cell analyses, it is important to select
epitopes from the extracellular domain so that the antibodies
74 Chelsea M. Fujinaka et al.

Fig. 4 CSC-Technology data can inform subsequent quantitative studies and antibody development. (a)
Quantitative data for two cell surface proteins, Integrin Beta 1 and Glucose Transporter 3 across three time
points of human pluripotent stem cell derived cardiomyocyte differentiation. Quantitation was achieved using
parallel reaction monitoring to target both proteins in the same experiment and data were analyzed using
Skyline [37]. Top row shows fragment ion chromatograms used for quantitation, and integrated data are
summarized in the chart below. (b) Illustrations of the transmembrane topology of Glypican-1 and Integrin
Beta 1 highlighting the extracellular peptides identified by CSC-Technology, including sites of N-glycosylation
as well as additional annotated sequence features that may be considered during epitope selection for
antibody development. Image generated using Protter

can access the epitopes without permeabilization of the plasma

membrane. In this case, the CSC-Technology data are valuable
as the N-glycopeptides identified in the approach are derived
from the extracellular domain and these data can be used to
verify the transmembrane orientation of the protein. Moreover,
 Cell Surface Capture Technology 75

these data can reveal occupied N-glycosites that can be avoided

during epitope selection. To visualize the N-glycopeptides
identified in the experiment along with annotated protein fea-
tures, upload the prot.xml file to Protter (http://wlab.ethz.ch/
protter/start/) by choosing the option to “load a proteomics
result file.” All peptides identified in the experiment will be
mapped to their respective proteins and can be rapidly viewed
in the context of known modifications and other sequence
features (Fig. 4b).

4 Notes

1. Liberase TH is an example of a trypsin-free cell dissociation

solution for detaching adherent cells from plates and dishes.
The solution and protocol used varies by cell type. It is impor-
tant that the dissociation solution does not contain trypsin, as
this will cleave extracellular protein domains. Acceptable solu-
tions can include enzyme-free dissociation solutions, EDTA
solutions, and/or those that contain collagenases. While it is
possible to perform the oxidation and biotinylation on cells
adhered to the culture dish, this strategy requires a significantly
larger volume of the biotinylation solution and thus is more
expensive.
2. Prepare immediately before use.
3. Typically ~5 drops of H3PO4 with a glass Pasteur pipette are
necessary to reach the correct pH.
4. While alternative quantities are available, the compound is
difficult to transfer and weigh precisely due to static electricity.
Therefore, we use preweighed single-use aliquots.
5. It is important to use a mass spectrometry compatible surfac-
tant to aid in the solubilization of membrane proteins. Alterna-
tive surfactants are available but should be tested for suitability.
6. It is possible to use fewer cells, such as 10–50
106, but the
number of proteins identified will be fewer. We typically iden-
tify 300–500 proteins if starting with 108 cells, 200–250 pro-
teins if starting with 50
106 cells.
7. Plasma membrane integrity is critical to ensure that the oxida-
tion/biotinylation only occurs on extracellular domains.
8. The time and speed for disruption may need to be adjusted
based on cell type. The goal is to obtain a white nuclear pellet
with a slightly cloudy supernatant.
9. Depending on the amount of starting material, it may be
beneficial to add another aliquot of trypsin if a relatively clear
solution is not obtained after the first round of digestion.
76 Chelsea M. Fujinaka et al.

10. If more than 20% of the peptides identified are contaminants

(i.e., do not contain a deamidation within the sequence motif),
this is typically due to insufficient washing of the streptavidin
beads after the glycopeptide capture. To further reduce non-
specific binding, streptavidin beads may be transferred to a new
Mobicol repeatedly throughout the washing steps and volumes
of washing solutions may be increased.
11. Notably, the Trans-Proteomic Pipeline, using Comet [38] and
X!Tandem [39] algorithms followed by PeptideProphet [40]
and ProteinProphet [41] also works well.
12. The choice for sample type used for quantitation can depend
on several factors. Whole cell lysate or biophysically enriched
membrane fractions can be advantageous as the sample prepa-
ration is simpler and more peptides from across the protein can
be used for quantitation (i.e., not only the formerly glycosy-
lated N-glycopeptide). However, this type of preparation does
not directly provide evidence of surface localization. The
CSC-Technology can be used and will provide evidence of
localization, but the peptides available for quantitation are
restricted to those formerly glycosylated N-glycopeptides.

Acknowledgments

This work was supported by National Institutes of Health grants

R01HL126785 and R01HL134010 and the Paul G. Allen Family
Foundation (Grant Award 11715).

References
1. Mummery CL, Zhang J, Ng ES, Elliott DA, non-human primate hearts. Nature 510
Elefanty AG, Kamp TJ (2012) Differentiation (7504):273–277
of human embryonic stem cells and induced 5. Schwartz SD, Hubschman JP, Heilwell G,
pluripotent stem cells to cardiomyocytes: a Franco-Cardenas V, Pan CK, Ostrick RM et al
methods overview. Circ Res 111(3):344–358 (2012) Embryonic stem cell trials for macular
2. Braam SR, Tertoolen L, van de Stolpe A, degeneration: a preliminary report. Lancet 379
Meyer T, Passier R, Mummery CL (2010) Pre- (9817):713–720
diction of drug-induced cardiotoxicity using 6. Bryder D, Rossi DJ, Weissman IL (2006)
human embryonic stem cell-derived cardio- Hematopoietic stem cells: the paradigmatic
myocytes. Stem Cell Res 4(2):107–116 tissue-specific stem cell. Am J Pathol 169
3. Liang P, Lan F, Lee AS, Gong T, Sanchez- (2):338–346
Freire V, Wang Y et al (2013) Drug screening 7. Shizuru JA, Negrin RS, Weissman IL (2005)
using a library of human induced pluripotent Hematopoietic stem and progenitor cells: clin-
stem cell-derived cardiomyocytes reveals ical and preclinical regeneration of the hema-
disease-specific patterns of cardiotoxicity. Cir- tolymphoid system. Annu Rev Med
culation 127(16):1677–1691 56:509–538
4. Chong JJ, Yang X, Don CW, Minami E, Liu 8. Clark G, Stockinger H, Balderas R, van Zelm
YW, Weyers JJ et al (2014) Human embryonic- MC, Zola H, Hart D, Engel P (2016) Nomen-
stem-cell-derived cardiomyocytes regenerate clature of CD molecules from the tenth human
 Cell Surface Capture Technology 77

leucocyte differentiation antigen workshop. 20. Ducret A, Kux van Geijtenbeek S, Roder D,
Clin Transl Immunol 5(1):e57 Simon S, Chin D, Berrera M et al (2015) Iden-
9. Turac G, Hindley CJ, Thomas R, Davis JA, tification of six cell surface proteins for specific
Deleidi M, Gasser T, Karaoz E, Pruszak J liver targeting. Proteomics Clin Appl 9
(2013) Combined flow cytometric analysis of (7–8):651–661
surface and intracellular antigens reveals sur- 21. Wollscheid B, Bausch-Fluck D, Henderson C,
face molecule markers of human neuropoiesis. O’Brien R, Bibel M, Schiess R, Aebersold R,
PLoS One 8(6):e68519 Watts JD (2009) Mass-spectrometric identifi-
10. Dubois NC, Craft AM, Sharma P, Elliott DA, cation and relative quantification of N-linked
Stanley EG, Elefanty AG, Gramolini A, Keller cell surface glycoproteins. Nat Biotechnol 27
G (2011) SIRPA is a specific cell-surface (4):378–386
marker for isolating cardiomyocytes derived 22. Apweiler R, Hermjakob H, Sharon N (1999)
from human pluripotent stem cells. Nat Bio- On the frequency of protein glycosylation, as
technol 29(11):1011–1018 deduced from analysis of the SWISS-PROT
11. Kelly OG, Chan MY, Martinson LA, Kadoya K, database. Biochim Biophys Acta 1473(1):4–8
Ostertag TM, Ross KG et al (2011) Cell- 23. Gahmberg CG, Tolvanen M (1996) Why
surface markers for the isolation of pancreatic mammalian cell surface proteins are glycopro-
cell types derived from human embryonic stem teins. Trends Biochem Sci 21(8):308–311
cells. Nat Biotechnol 29(8):750–756 24. Bausch-Fluck D, Hofmann A, Bock T, Frei AP,
12. Bye CR, Jonsson ME, Bjorklund A, Parish CL, Cerciello F, Jacobs A et al (2015) A mass
Thompson LH (2015) Transcriptome analysis spectrometric-derived cell surface protein
reveals transmembrane targets on transplanta- atlas. PLoS One 10(3):e0121314
ble midbrain dopamine progenitors. Proc Natl 25. Boheler KR, Bhattacharya S, Kropp EM,
Acad Sci U S A 112(15):E1946–E1955 Chuppa S, Riordon DR, Bausch-Fluck D et al
13. Cordwell SJ, Thingholm TE (2010) Technol- (2014) A human pluripotent stem cell surface
ogies for plasma membrane proteomics. Prote- N-glycoproteome resource reveals markers,
omics 10(4):611–627 extracellular epitopes, and drug targets. Stem
14. Elschenbroich S, Kim Y, Medin JA, Kislinger T Cell Rep 3(1):185–203
(2010) Isolation of cell surface proteins for 26. Domon B, Gallien S (2015) Recent advances in
mass spectrometry-based proteomics. Expert targeted proteomics for clinical applications.
Rev Proteomics 7(1):141–154 Proteomics Clin Appl 9(3–4):423–431
15. Kitchen P, Day RE, Salman MM, Conner MT, 27. Omasits U, Ahrens CH, Muller S, Wollscheid B
Bill RM, Conner AC (2015) Beyond water (2014) Protter: interactive protein feature visu-
homeostasis: diverse functional roles of mam- alization and integration with experimental
malian aquaporins. Biochim Biophys Acta proteomic data. Bioinformatics 30
1850(12):2410–2421 (6):884–886
16. Danzer C, Eckhardt K, Schmidt A, 28. Mallanna SK, Cayo MA, Twaroski K, Gundry
Fankhauser N, Ribrioux S, Wollscheid B et al RL, Duncan SA (2016) Mapping the cell-
(2012) Comprehensive description of the surface N-glycoproteome of human hepato-
N-glycoproteome of mouse pancreatic beta- cytes reveals markers for selecting a homoge-
cells and human islets. J Proteome Res 11 neous population of iPSC-derived hepatocytes.
(3):1598–1608 Stem Cell Rep 7(3):543–556
17. Tyleckova J, Valekova I, Zizkova M, 29. Kropp EM, Bhattacharya S, Waas M, Chuppa
Rakocyova M, Marsala S, Marsala M, Gadher SL, Hadjantonakis AK, Boheler KR, Gundry
SJ, Kovarova H (2016) Surface RL (2014) N-glycoprotein surfaceomes of
N-glycoproteome patterns reveal key proteins four developmentally distinct mouse cell
of neuronal differentiation. J Proteome types. Proteomics Clin Appl 8(7–8):603–609
132:13–20 30. Gundry RL, Riordon DR, Tarasova Y,
18. Sun B, Ma L, Yan X, Lee D, Alexander V, Chuppa S, Bhattacharya S, Juhasz O et al
Hohmann LJ et al (2013) N-glycoproteome (2012) A cell surfaceome map for immunophe-
of E14.Tg2a mouse embryonic stem cells. notyping and sorting pluripotent stem cells.
PLoS One 8(2):e55722 Mol Cell Proteomics 11(8):303–316
19. Hofmann A, Thiesler T, Gerrits B, Behnke S, 31. Bausch-Fluck D, Hofmann A, Wollscheid B
Sobotzki N, Omasits U et al (2015) Surfa- (2012) Cell surface capturing technologies for
ceome of classical Hodgkin and non-Hodgkin the surfaceome discovery of hepatocytes.
lymphoma. Proteomics Clin Appl 9 Methods Mol Biol 909:1–16
(7–8):661–670
78 Chelsea M. Fujinaka et al.

32. Hofmann A, Gerrits B, Schmidt A, Bock T, 37. MacLean B, Tomazela DM, Shulman N,
Bausch-Fluck D, Aebersold R, Wollscheid B Chambers M, Finney GL, Frewen B et al
(2010) Proteomic cell surface phenotyping of (2010) Skyline: an open source document edi-
differentiating acute myeloid leukemia cells. tor for creating and analyzing targeted proteo-
Blood 116(13):e26–e34 mics experiments. Bioinformatics 26
33. Schiess R, Mueller LN, Schmidt A, Mueller M, (7):966–968
Wollscheid B, Aebersold R (2009) Analysis of 38. Eng JK, Jahan TA, Hoopmann MR (2013)
cell surface proteome changes via label-free, Comet: an open-source MS/MS sequence
quantitative mass spectrometry. Mol Cell Pro- database search tool. Proteomics 13(1):22–24
teomics 8(4):624–638 39. Craig R, Beavis RC (2004) TANDEM: match-
34. Kall L, Canterbury JD, Weston J, Noble WS, ing proteins with tandem mass spectra. Bioin-
MacCoss MJ (2007) Semi-supervised learning formatics 20(9):1466–1467
for peptide identification from shotgun prote- 40. Keller A, Nesvizhskii AI, Kolker E, Aebersold R
omics datasets. Nat Methods 4(11):923–925 (2002) Empirical statistical model to estimate
35. Taus T, Kocher T, Pichler P, Paschke C, the accuracy of peptide identifications made by
Schmidt A, Henrich C, Mechtler K (2011) MS/MS and database search. Anal Chem 74
Universal and confident phosphorylation site (20):5383–5392
localization using phosphoRS. J Proteome 41. Nesvizhskii AI, Keller A, Kolker E, Aebersold R
Res 10(12):5354–5362 (2003) A statistical model for identifying pro-
36. Dorfer V, Pichler P, Stranzl T, Stadlmann J, teins by tandem mass spectrometry. Anal Chem
Taus T, Winkler S, Mechtler K (2014) MS 75(17):4646–4658
Amanda, a universal identification algorithm
optimized for high accuracy tandem mass spec-
tra. J Proteome Res 13(8):3679–3684
 Chapter 5

Application of Higher Density Iron Oxide Nanoparticle

Pellicles to Enrich the Plasma Membrane and Its Proteome
from Cells in Suspension
Rebecca L. Rose, Waeowalee Choksawangkarn, and Catherine Fenselau

Abstract
Enrichment of the plasma membrane represents one valuable method to characterize the surfaceome, along
with other plasma membrane and structural proteins. Currently, the overlapping densities of many subcel-
lular organelles hinder enrichment of the plasma membrane by centrifugation. However, external access to
the plasma membrane of intact cells allows the attachment of a nanoparticle pellicle to enhance its density
and facilitate enrichment. We describe the synthesis of iron oxide nanoparticles, attachment of the pellicle to
suspended cells, and recovery of plasma membrane proteins for proteomic analysis.

Key words Plasma membrane, Nanoparticles, Pellicles, Proteomics, HPLC-MS/MS

1 Introduction

The plasma membrane (PM) provides access to transmembrane

proteins and glycoproteins that extend onto the surface, as well as
membrane proteins that extend into the cytoplasm, cytoskeletal
proteins and, depending on the method, secreted proteins. The
ability to enrich the plasma membrane from other cellular compo-
nents represents an excellent approach to study the surfaceome of
different cell types. This chapter presents a procedure for enriching
the plasma membrane from intact cells by attaching a nanoparticle
pellicle that increases its density relative to that of other subcellular
organelles. This facilitates the separation by centrifugation of the
plasma membrane from other lysed cellular components. Chaney
and Jacobson first reported the use of a nanoparticle pellicle using
an in-house preparation of cationic colloidal silica [1, 2]. Following
attachment of the cationic nanoparticles to the anionic eukaryotic
cell surface by ionic attraction, the pellicle is stabilized by cross-
linking with poly(acrylic) acid (PAA). The pellicle method was
incorporated into proteomic strategies in 2004, using commercially

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_5, © Springer Science+Business Media, LLC 2018

79
80 Rebecca L. Rose et al.

available colloidal silica [3] and the original Jacobson preparation

[4]. Recently a comparison has been made of pellicles comprising
nanoparticles of differing densities [5]. Iron oxide particles coated
with cationic alumina [6] were found to provide better enrichment
than the lighter colloidal silica particles, and form the basis of the
procedure presented here for enrichment of the plasma membrane
from cells in suspension and, thereby, cells from tissue [7].
In brief, eukaryotic cells grown in suspension are washed with
the plasma membrane coating buffer A (PMCBA), pH 5.3, to
remove cell debris. Alumina-coated iron oxide nanoparticles are
synthesized using polyol procedures and dispersed in PMCBA. A
solution of suspended cells is added dropwise to the nanoparticle
solution to ensure complete coating of the cationic nanoparticle on
the anionic cell surface. After removing excess nanoparticles, the
unbound charges of the nanoparticles are neutralized by anionic
PAA solution in PMCBA, which also provides higher strength to
the plasma membrane–nanoparticle pellicles by cross-linking of the
polymer around the cells. The coated cells were then resuspended
in hypotonic buffer and lysed by nitrogen cavitation. After lysis,
open sheets of plasma membrane–nanoparticle pellicles can be
produced as observed under the scanning electron microscope.
The coated plasma membrane possesses higher density than other
cellular organelles and thus can be separated by low-speed centrifu-
gation (Table 1). The integral plasma membrane proteins are
extracted from the pellicles using microwave-assisted method in
the presence of sodium dodecyl sulfate. The extracted proteins
can be subjected to identification or quantification using bottom-
up proteomic analysis procedures.

Table 1
Densities of subcellular organelles (adapted from ref. 8)

Subcellular component Density (g/cm3)

Nucleus 1.4
Endoplasmic reticulum 1.2
Golgi apparatus 1.14
Mitochondria 1.1
Ribosomes 1.6
Lysosomes 1.1
Plasma membrane 1.12
 Nanoparticles to Enrich Plasma Membrane Proteins 81

2 Materials

2.1 Iron Oxide 1. For the preparation of iron oxide nanoparticles prepare
Nanoparticle 0.5 mM FeCl3·6H2O and 1.5 mM sodium acetate
Construction (C2H3NaO2) in 25 mL of propylene glycol.
2. For coating the iron oxide particles with alumina, prepare
0.01 M Al(NO3)3/0.1 M K(NO3) as a 10 mL aqueous
solution.
3. Glassware required: 100 mL round bottom flask with a 24/40
ground glass joint (catalog number UX-34706-10; Cole-
Parmer, Vernon Hills, IL, USA) connected to the condenser
via a female 24/40 ground glass joint.
4. Centrifugal Filters (Cat No. UFC500396, Millipore, Billerica,
MA, USA).

2.2 Pellicle 1. Plasma Membrane Coating Buffer A (PMCBA): 800 mM sor-

Construction and PM bitol, 20 mM 2-(N-morpholino)ethanesulfonic acid buffer
Analysis (MES), 150 mM NaCl, pH 5.3. Solutions of 10 M NaOH or
concentrated HCl are recommended to adjust pH of PMCBA.
2. 10 mg/mL poly(acrylic) acid (PAA) in PMCBA, pH 6–6.5.
After adding PAA, the pH of PMCBA will be decreased, and a
portion of 10 M NaOH solution is needed to adjust the pH to
be 6.0–6.5.
3. Lysis buffer: 2.5 mM imidazole with protease inhibitor cocktail
(catalog number P8340-1ML) (Sigma-Aldrich, St. Louis, MO,
USA).
4. 1 M Na2CO3, pH 11.4.
5. 1 M KCl.
6. Protein extraction buffer: 2% SDS, 62.5 mM Tris–HCl, and 5%
β-mercaptoethanol.
7. Resolubilization buffer: 8 M urea, 50 mM NH4HCO3.
8. 20 mM dithiothreitol (DTT) in 50 mM NH4HCO3.
9. 40 mM iodoacetamide (IAA) in 50 mM NH4HCO3.
10. 0.1% formic acid.
11. Endoproteinase Lys-C, sequencing grade (catalog number
V1071) (Promega, Madison, WI, USA).
12. Trypsin, mass spectrometry grade (catalog number V5280;
Promega).
13. C18 spin column (Pierce™, Thermo Fisher Scientific, San
Jose, CA, USA).
82 Rebecca L. Rose et al.

2.3 Tissue Culture 1. Human multiple myeloma cells (ATCC RPMI 8226).
Supplies 2. Tissue culture incubator with CO2 regulator.
3. Refrigerated tissue culture compatible centrifuge for 15 and
50 mL tubes.
4. 75 cm2 cell culture flask.
5. RPMI 1640 medium.
6. Fetal Bovine Solution (FBS).
7. Hemocytometer.
8. Light microscope (200 magnification).
9. 50 mL conical tube.
10. 3 mL syringe with 21G needle.

2.4 Other 1. Ultrasonic bath sonicator.

2. Cavitation chamber (4635 Cell Disruption Vessel, Parr Instru-
ment Company, Moline, IL, USA).
3. Microwave compatible glass tubes and stir bar.
4. Reducing Agent and Detergent Compatible assay, RC DC™
(catalog number 500–0122; Bio-Rad, Hercules, CA, USA).
5. Water bath at 56  C.
6. Lyophilizer (see Note 1).
7. Microwave (CEM Corporation, Matthews, NC, USA).

3 Methods

3.1 Iron Oxide 1. To begin the nanoparticle preparation dilute FeCl3·6H2O and
Nanoparticle sodium acetate in 25 mL of propylene glycol to a final concen-
Construction [6] tration of 0.5 mM and 1.5 mM, respectively.
(See Fig. 1) 2. Add 500 μL of deionized water to this mixture and reflux for
12 h in a 100 mL round bottom flask at room temperature; be
sure to fit a condenser to the round bottom flask via a female
24/40 ground glass joint. Particles will form (see Note 2).
3. Collect and wash the particles four times with water via centri-
fugation (8500  g for 30 s) in Millipore centrifugal filter.
4. Resuspend the particles in 10 mL of 0.01 M Al(NO3)3/0.1 M
K(NO3) for 12 h at room temperature on a rocking platform.
5. Collect the particles via centrifugation (8500  g for 10 min),
discard the supernatant, and resuspend the pellet in PMCBA
for use in cell coating and pellicle construction.
 Nanoparticles to Enrich Plasma Membrane Proteins 83

Fig. 1 Transmission electron microscope (TEM) image of alumina-coated iron

oxide nanoparticles (modified from ref. 5). A white scale bar represents 50 nm in
length. An average diameter of the nanoparticles is approximately 17  6 nm.
Individual nanoparticles are shown in gray. Since the TEM sample preparation
requires drying process, some nanoparticle aggregates could be observed.
Darker particles could result from different layers of the nanoparticles as a
part of the aggregates

3.2 Coating 1. Human multiple myeloma cells were cultured in suspension in

of Multiple Myeloma T-75 flasks at 37  C in RPMI 1640 medium containing 10%
Cells fetal bovine serum under a 5% CO2 atmosphere. The cell
with Nanoparticle density should be maintained between 5  105 and 2  106
Pellicles (Figs. 2 and 3, cells/mL. Do not allow the cell density to exceed 2–3  106
See Note 3) cells/mL. Approximately, 5–10 flasks containing 20 mL media
are required for one experiment.
2. To estimate cell culture concentration, remove a sample and
count the cells with a hemocytometer. Use the counts to
determine the volume of culture needed to collect 6  107
cells.
3. To collect cells, each cell culture flask should be poured into a
sterile 50 mL conical tube utilizing aseptic technique (see Note 4).
4. Pellet the 6  107 cells via centrifugation at 900  g for 5 min.
The temperature is controlled to be 4 C (see Notes 5 and 6).
5. Wash cells with cold PMCBA buffer three times by resuspend-
ing the cell pellet gently by trituration with a 10 mL automatic
pipet. After each wash, repellet the cells via centrifugation at
900  g for 5 min (see Note 7).
6. After the final wash, resuspend the cell pellet in 2 mL of
PMCBA buffer and use a syringe to add the suspension drop-
wise to the 10% (w/v) suspension of cationic nanoparticles
prepared under Subheading 3.1 (see Notes 8 and 9).
84 Rebecca L. Rose et al.

Fig. 2 Pellicle construction with alumina-coated iron oxide nanoparticles. Nega-

tively charged intact cells are added to a solution of positively charged iron oxide
nanoparticles to coat the cells via electrostatic interaction. Then coated cells are
cross-linked with anionic PAA to neutralize excess charges on the surfaces. After
that, coated cells are lysed by nitrogen cavitation to produce plasma membrane
pellicle fragments. The pellicles are isolated by low-speed centrifugation and
subjected to protein extraction, followed by bottom-up proteomic analysis

7. Incubate the suspension for 15 min at 4  C with gentle

rocking.
8. After the incubation, repeat the washing outlined in step 5
using fresh PMCBA buffer. This process removes any excess
nanoparticle suspension.
9. Resuspend the coated cells in 2 mL PMCBA buffer and add
dropwise to a solution of 10 mg/mL PAA for cross-linking.
This step should be carried out in the same manner as step 6.
10. Incubate the cross-linking reaction for 15 min at 4  C with
rocking.
11. To collect coated cells spin at 900  g for 5 min.
12. Perform three washes with PMCBA as outlined in step 5.
13. After the final wash, resuspend the cell pellet in 10 mL of lysis
buffer and incubate at 4  C for 30 min with gentle rocking.
This causes cell swelling for efficient cavitation.
14. Perform N2 cavitation within cavitation chamber for 30 min at
1500 psi (see Notes 10 and 11).
15. Visually confirm cavitation by observing a small aliquot of the
sample under a light microscope (see Note 12).
 Nanoparticles to Enrich Plasma Membrane Proteins 85

Fig. 3 Scanning electron microscope images. (a) Human multiple myeloma cell, (b) cell coated with alumina-
coated iron oxide nanoparticles, (c) PAA cross-linked coated cell, (d) fragment of pellicle-coated plasma
membrane (modified from ref. 5)

16. Collect pellicle fragments using low speed centrifugation at

100  g for 7 min at 4  C (see Note 13).
17. Wash the pellicle three times with 1 M Na2CO3 (pH 11.4) via
low-speed centrifugation at 100  g for 7 min at 4  C to
remove soluble proteins and randomly attached proteins from
the pellicles. The plasma membrane nanoparticle pellicles will
be pelleted by centrifugation, while supernatant containing
other organelles should be discarded.
18. Repeat washing steps three times with 1 M KCl to further
remove randomly attached proteins from the pellicles.

3.3 Protein 1. Resuspend the pellicle pellet in 200 μL of protein extraction

Extraction buffer and place samples into microwave compatible glass
tubes, with a small stir bar.
2. Microwave samples for 5 min at 100  C with gentle stirring
while irradiating with 300 W, using a CEM Discover
Microwave.
86 Rebecca L. Rose et al.

3. Remove samples from microwave and transfer to a 1.5 mL

Eppendorf tube (see Note 14).
4. Spin samples at 100  g for 7 min at room temperature to
pellet the pellicle/membrane fragments. Following pellet for-
mation, collect the supernatant as the first protein extract.
5. Resuspend the pellicle/membrane fragments in Protein extrac-
tion buffer and transfer back to the microwave compatible tube
and repeat the extraction two additional times to get the sec-
ond and third protein extracts (see Notes 15 and 16).
6. Perform a Reducing Agent and Detergent Compatible assay on
each protein extract to determine protein concentration, and
combine fractions of the extracts as needed to obtain 100 μg of
protein.

3.4 Preparation 1. Precipitate 100 μg of protein from the extraction sample using
of Samples for High a standard chloroform–methanol procedure (see Note 18).
Performance Liquid 2. Resuspend the protein pellet in 100 μL 8 M urea and 50 mM
Chromatography NH4HCO3.
(HPLC)-Mass 3. Reduce protein sulfhydryl groups by adding 2 μL of 1 M DTT
Spectrometry stock solution for a final concentration of 20 mM DTT. Incu-
(See Note 17) bate the sample at 56  C for 20 min in a water bath.
4. Alkylate protein thiol groups by adding 4 μL of the 1 M IAA
stock solution to the sample for a final concentration of 40 mM
IAA. Incubate at room temperature for 20 min.
5. Add protease Lys-C to a final concentration of 1:50 enzyme–-
substrate and incubate at 37  C for 3 h.
6. Following Lys-C digestion dilute the approximately 8 M urea
solution to 1.6 M urea with 50 mM NH4HCO3.
7. Add trypsin to a final concentration of 1:25 enzyme–substrate
and incubate at 37  C for 16 h.
8. Desalt and concentrate samples using C18 spin columns.
Recover peptides in 85% acetonitrile (see Notes 19 and 20).
9. Freeze samples at 80  C to prepare for lyophilization.
10. Lyophilize samples until the final volume is negligible and
resuspend in 100 μL of 0.1% formic acid for analysis via
HPLC interfaced to a tandem mass spectrometer.

3.5 Analysis 1. The LC gradient utilized should be appropriate for the

of Samples expected complexity of the sample. For analysis of the multiple
(See Note 21) myeloma PM 15 μg of sample was utilized per injection.
2. To ensure thorough desalting, load samples for 10 min at a flow
rate of 10 μL/min onto the trapping column desired. For
analysis of the sample discussed, an Acclaim PepMap
300 C18 column was used.
 Nanoparticles to Enrich Plasma Membrane Proteins 87

3. Samples should be analyzed using a column of the same chem-

istry as the trapping column, such as a Grace Vydac Everest
C18 column.
4. The HPLC gradient should contain a linear phase. In this case,
a 90 min gradient from 10–60% solvent B (90–40% solvent A)
was used, followed by 20 min from 60 to 85% solvent B
(40–15% solvent A). Solvent B contains 97.5% acetonitrile,
2.5% water and 0.1% formic acid; while solvent A contains
97.5% water, 2.5% acetonitrile and 0.1% formic acid.
5. Mass spectrometric parameters should be adjusted to desired
purpose of sample analysis.

4 Notes

1. A centrifugal evaporator can be used instead of a lyophilizer.

2. During the reflux, the solution should be boiled, and the color
of the solution should change from yellow to dark orange.
When the particle forms, changes in the color of the solution
should be observed.
3. This protocol for iron oxide nanoparticle pellicle experiment is
suitable for suspended cells either from cell culture or from
blood sample. It is also applicable for other types of mammalian
cells or tissues; however, the procedure should be optimized.
Since each type of cells possesses different degrees of negative
charge on the surface, you should bring an aliquot of samples
after steps 5, 8, 12 and 14 to observe the cell morphology,
coating ability, cross-linking ability and cell lysis efficiency by
using the scanning electron microscope. If incomplete coating
of the nanoparticle or PAA is observed, the concentration of
nanoparticles or PAA and coating time should be optimized. If
intact cells are presented after cell lysis, the pressure for nitro-
gen cavitation should be increased.
4. Care should be taken to maintain a sterile environment when
transferring cultured cells to conical tubes to ensure minimal
environmental contamination.
5. Every step should be performed at 4  C to avoid the risk of
internalizing nanoparticles.
6. If cells do not readily pellet under the experimental conditions,
increase spin times rather than speeds. This minimizes the risk
of pelleting debris present in the medium.
7. Resuspension of cell pellets at any point in the procedure
should be done via gentle pipetting rather than vortexing, in
order to avoid premature lysis.
88 Rebecca L. Rose et al.

8. Before coating the cells, the nanoparticle suspension should be

sonicated and vortexed to ensure good dispersion of the nano-
particle in solution.
9. When adding cell solutions dropwise to either cationic particle
suspensions or PAA solutions, care should be taken to add
samples slowly. This allows thorough mixing and thus more
efficient and complete coating and cross-linking. Use of 3 mL
syringes with 21G needles is recommended for these additions.
The inner diameter of the syringe needle should be larger than
the cell diameter to avoid cell lysis.
10. During lysis it is often beneficial to add empty conical tubes to
the cavitation chamber to minimize the risk of sample tube
movement (and thus sample loss) with changing pressures.
11. Nitrogen cavitation is the chosen method for this example
because it uses physical forces, rather than chemical reactions,
to lyse the cells. This minimizes the risk of pellicle disruption. If
other lysis methods are attempted, the fragile nature of the
pellicle interaction should be considered.
12. When cell lysis is achieved by nitrogen cavitation, small frag-
ments and debris should be observed under the light micro-
scope. In case intact cells are presented, higher pressure of the
cavitation (1800 psi) is recommended.
13. It is not uncommon for the pellicle to precipitate while the
sample is resting on the bench.
14. When extracting proteins from the pellicle take care when
removing samples from the microwave as glass vials may retain
heat for an extended period following microwaving.
15. Using protein quantitation (RC DC protein assay), we observe
that recovery falls off significantly after three extractions.
16. For a new type of sample it may be wise to determine the
optimal number of extractions required to recover all the pro-
tein from the pellicle fragments. After each time of protein
extraction in the microwave, 25 μL of the extracts should be
subjected to RC DC protein assay to determine the concentra-
tion. If the concentration drastically decreases from the previ-
ous extraction, subsequent extraction is not necessary.
17. It is recommended for a novice to consult with a trained mass
spectroscopist to perform the sample preparation, since the
membrane peptides are generally hydrophobic and may require
additional procedures for different type of samples.
18. To perform a protein precipitation using chloroform/metha-
nol procedure, one volume of sample is mixed with four
volumes of methanol. Then, one volume of chloroform is
added to the mixture, followed by three volumes of deionized
H2O. After vigorously vortexing, the solution should become
 Nanoparticles to Enrich Plasma Membrane Proteins 89

cloudy. The mixture is subjected to centrifugation at

14,000  g for 2 min at room temperature. The white protein
precipitate should be observed as an interface between the top
aqueous layer and the bottom organic solvent layer. The top
layer should be removed by gentle pipetting to avoid disturb-
ing the protein precipitate. Then, four volumes of methanol
are added to wash the pellet, followed by centrifugation at
14,000  g for 2 min at room temperature. The protein
precipitate should be observed as a white pellet, and the super-
natant can be discarded. Methanol can be removed from the
pellet by air-dry or using a vacuum centrifugation.
19. To desalt and concentrate the peptide samples, it is recom-
mended to follow the protocol from manufacturer (Pierce™,
Thermo Fisher Scientific, San Jose, CA, USA). Briefly, the spin
column is activated using 400 μL of 50% acetonitrile and then
equilibrated with 400 μL of 0.5% trifluoroacetic acid in 5%
acetonitrile. Sample buffer should be adjusted to contain the
final concentration of 0.5% trifluoroacetic acid in 5% acetoni-
trile. After column equilibration, sample is applied on the resin
bed and centrifuged at 1500  g for 1 min. At this point,
peptides are bound to the C18 resin, and salt can be removed
by washing several times with 0.5% trifluoroacetic acid in 5%
acetonitrile solution. After washing, the peptides are eluted
using 40 μL of 0.1% formic acid in 85% acetonitrile. The eluted
solution is then frozen to prepare for lyophilization.
20. When performing a C18 column cleanup it is not
uncommon to observe a colored layer within the top of the
column. These are a small number of residual nanoparticles
that manage to remain in solution during extraction and diges-
tion. They do not interfere with column desalting and sample
concentration.
21. For the analysis of the sample using HPLC-MS/MS, it is
recommended to consult with professional mass spectrosco-
pists or staff at a proteomic core facility. The HPLC instrument
and mass spectrometer could be different from what is men-
tioned in this chapter. However, it is recommended to use a
nano-HPLC system to increase the sensitivity of the separation,
since many membrane proteins are low abundant. Analytical
column used to separate peptides should be reverse-phase; C18
is recommended. In case you work with highly hydrophobic
proteins, column heater may be required to elute highly hydro-
phobic peptide from the column. Multiple injections and anal-
ysis of the peptide mixture by LC-MS/MS will provide
additional peptide and protein identifications.
90 Rebecca L. Rose et al.

Acknowledgment

This developmental research was supported by a grant from the US

National Institutes of Health, GM021248.

References

1. Chaney LK, Jacobson BS (1983) Coating cells 5. Choksawangkarn W, Kim SK, Cannon JR,
with colloidal silica for high yield isolation of Edwards NJ, Lee SB, Fenselau C (2013) Enrich-
plasma membrane sheets and identification of ment of plasma membrane proteins using nano-
transmembrane proteins. J Biol Chem particle pellicles: comparison between silica and
258:10062–10072 higher density nanoparticles. JProteome Res
2. Stolz DB, Jacobson BS (1992) Examination of 12:1134–1141
transcellular protein polarity of bovine aortic 6. Bai X, Son SJ, Zhang S, Liu W, Frank JA,
endothelial cells in vitro using the cationic col- Venkatesan T, Lee SB (2008) Synthesis of super-
loidal silica microbead membrane-isolation pro- paramagnetic nanotubes as MRI contrast agents
cedure. J Cell Sci 103:39–51 and for cell labeling. Nanomedicine 3:163–166
3. Rabar AM, Fenselau C (2004) Integration of 7. Choksawangkarn W, Graham LM, Burke M, Lee
Jacobson’s pellicle method into proteomic stra- SB, Ostrand-Rosenberg S, Fenselau C, Edwards
tegies for plasma membrane proteins. J Prote- N (2016) Peptide-based systems analysis of
ome Res 3:1267–1277 inflammation induced myeloid-derived suppres-
4. Durr E, Yu J, Krasinska K, Carver LA, Yates JR, sor cells reveals diverse signaling pathways. Pro-
Testa JE, Oh P, Schnitzer J (2004) Direct pro- teomics 16:1881–1888
teomic mapping of the lung microvascular endo- 8. Lodish H, Berk A, Zipursky SL, Matsudaira P,
thelial cell surface in vivo and in cell culture. Nat Baltimore D, Darnell J (2000) Molecular cell
Biotechnol 22:985–992 biology, 4th edn. W.H. Freeman, New York
 Chapter 6

Proteomic Profiling of Secreted Proteins, Exosomes,

and Microvesicles in Cell Culture Conditioned Media
Ankit Sinha, Simona Principe, Javier Alfaro, Alex Ignatchenko,
Vladimir Ignatchenko, and Thomas Kislinger

Abstract
Secreted proteins are of tremendous biological interest since they can act as ligands for receptors to activate
downstream signalling cascades or be used as biomarkers if altered abundance is correlated with a specific
pathological state. Proteins can be secreted either as soluble molecules or as part of extracellular vesicles
(i.e., exosomes or microvesicles). The complete proteomic profiling of secretomes requires analysis of
secreted proteins and extracellular vesicles. Hence, the method described here enriches for microvesicles,
exosomes, and secreted proteins from conditioned media using differential centrifugation. The three
fractions are then analyzed by mass spectrometry-based proteomics for in-depth characterization and
comparison of the protein secretome of cell lines.

Key words Extracellular vesicles, Exosomes, Microvesicles, Secreted proteins, Secretome, Proteo-
mics, Mass spectrometry

1 Introduction

The development of intercellular communication marked a major

milestone in the evolution of multicellular organisms. Extracellular
signalling molecules allow secreting cells to communicate with
target cells located nearby (paracrine signalling) or at distant sites
(endocrine signalling). The target cells respond by altering meta-
bolic activity or intracellular pathways. Intercellular communication
is of key interest in studying cancer biology. Pre-malignant cells
accumulate mutations in oncogenes and tumor suppressor genes,
that enable unregulated growth and escape from programmed cell
death [1]. For malignant progression, cancer cells use secreted
factors to modify their microenvironment to promote angiogene-
sis, immune evasion, uncontrolled growth, and metastasis
[2]. Secreted factors from cells can be classified based on molecular
composition such as lipids, RNAs (mRNA, miRNA, siRNA, and
other ncRNA), DNA and proteins. Integrated studies of the

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_6, © Springer Science+Business Media, LLC 2018

91
92 Ankit Sinha et al.

secretome, including the vesiculome (extracellular vesicles-based

intercellular communication) hold great interest in cancer biology
as they can aid in deconvoluting mechanisms of disease progres-
sion, immune regulation, mediation of metastasis, and biomarkers
for cancer detection [3–7].
Although not strictly defined by molecular composition, extra-
cellular vesicles (EVs) are spherical structures composed of a lipid
bilayer membrane that encloses an organelle-free cytosol that are
secreted by cells. EVs have emerged as a functionally important
mechanism for transferring lipids, RNA, DNA, and proteins as
cargo for intercellular communication [8–10]. Secreted EVs are
classified into three major groups based on biogenesis and physical
dimensions. Exosomes are the smallest EVs and their biogenesis
involves inward-budding of the maturing endosomes, while captur-
ing cellular content, to become intraluminal vesicles (ILV) within
endosomes. These ILV range between 30 and 100 nm in diameter,
and endosomes containing ILVs are referred to as multivesicular
bodies (MVBs). The fusion of the newly formed MVBs with the
plasma membrane leads to exocytosis of ILV into the extracellular
space. These secreted ILVs are now termed as exosomes
[8, 11]. Microvesicles (MVs) are medium-sized EVs, that range
between 100 and 1000 nm and their biogenesis involves protrusion
and subsequent outward budding of cellular plasma membranes,
followed by release of spherical vesicles [12]. Lastly, apoptotic
bodies are vesicles that form during fragmentation of cells under-
going apoptosis. Apoptotic bodies have a wide range of molecular
cargo (organelles to DNA fragments to cytosolic contents) and a
wide range of diameters, 50–5000 nm.
In the past, biomarker discovery was the major source of scien-
tific interest in EV research (particularly exosomes), but several
recent high-profile publications have reported on a multitude of
mechanistic functions of EVs in cancer progression which has
increased its appeal to aspects of cell signalling and cancer biology
[13, 14]. Tumor cells secrete extracellular vesicles as a mechanism
of immune suppression and tumor escape [15, 16]. Increased
breast cancer cell motility was shown to be aided by activation of
the Wnt-PCP pathway through exosomes derived from cancer-
associated fibroblasts [17]. Shimoda et al. have shown that expres-
sion of the metalloproteinase ADAM10 in exosomes derived from
cancer-associated fibroblasts enhance motility of breast cancer cell
lines through the Notch receptor–GTPase RhoA axis [18]. In con-
junction, MVs are also involved in horizontal transfer of biomole-
cules between cells [19]. MVs isolated from lung primary cells
induced epigenetic changes in bone marrow cells in mice, enhanc-
ing expression of genes specific to pulmonary epithelial cells
[20]. Considering that different secreted factors are released by
cells as a response to altered cellular states (e.g., neoplastic transfor-
mation or viral infection), profiling of secretomes holds potential
 Profiling of Secreted Proteins, Exosomes and Microvesicles 93

Fig. 1 Workflow for the proteomic characterization of MVs, exosomes, and secreted proteins from CM. (a)
Conditioned media is obtained through 48 h incubation in serum-free media. (b) MVs, exosomes, and secreted
proteins are obtained through differential centrifugation of the CM. (c) Protein content of all fractions is
analyzed using LC-MS/MS based proteomics

for identifying novel biomarkers or to obtain a better understand-

ing of the mechanisms involved in mediating intercellular commu-
nication within the tumor microenvironment.
This chapter has two key areas of emphasis. Firstly, the isolation
of secreted proteins, exosomes, and MVs present in conditioned
media (CM) of cultured cell lines. Secondly is the use of a 2,2,2-
trifluoroethanol (TFE)-based sample preparation strategy for
in-depth mass spectrometry (MS) driven proteomic characteriza-
tion of the fractioned secretome [21]. The described method uti-
lizes differential centrifugation to enrich for MVs, exosomes, and
secreted proteins as shown in Fig. 1. The specific steps described in
this chapter involve mammalian cell culture and conditioned
medium (Subheading 3.1), exosome isolation (Subheading 3.2),
TFE based digestion of proteins (Subheading 3.3), preparation of
samples for mass spectroscopy (Subheading 3.4), and liquid chro-
matography and mass spectrometry (Subheading 3.5). The method
described can be executed in any biochemistry laboratory with
access to an ultra-centrifuge.
Isolation of EVs from CM can be performed using various
complimentary techniques. Size exclusion chromatography can be
used for enriching EVs by collecting elution volumes appertaining
to defined molecular sizes. Although this approach can fractionate
vesicles differing in molecular size, secreted proteins would be
diluted across large numbers of fractions which increases the com-
plexity of subsequent analysis, such as the requirement of additional
concentrating steps. Another strategy involves precipitation of EVs
94 Ankit Sinha et al.

using polyethyleneglycol (PEG). Although less time-consuming,

downstream MS-based proteomic analysis of secreted proteins
might be complicated due to interference from PEG. Immunoaffi-
nity isolation of EVs is an attractive alternative that allows for
isolation of EV subpopulations based on the expression of specific
surface antigens. As an example, immunoaffinity based isolation of
exosomes can be performed using magnetic beads conjugated with
antibodies against CD9, CD63, and CD81; all three proteins
belong to the tetraspanin family of transmembrane proteins and
are abundant in exosomes [7, 22]. Similarly, immunoaffinity based
isolation of MVs can be performed using antibodies against PSGL1
and β1-integrin [22]. A consideration of this approach is that it
kinetically favors MVs and exosome particles with greater abun-
dance of the targeted surface antigen. Hence, heterogeneity of MV
or exosome populations is lost as the method is unlikely to enrich
for EVs with lower abundance of the targeted antigen. Although,
under certain experimental conditions, the analyses of subpopula-
tion of EVs might be of particular interest. The last method,
differential ultracentrifugation sediments EVs based on their differ-
ing size and density. Hence, MVs, exosomes, and secreted proteins
are obtained from the CM without additional sample modification
or increases in complexity. Purity of isolation can be increased using
sucrose or iodixanol density gradient ultracentrifugation [23].
In a protocol frequently used in our lab (and described here),
CM from cultured cells (primary cells or immortalized cell lines) is
collected without perturbation of the adherent cells. The CM is first
centrifuged twice at low-speeds for removal of non-adherent cells
and cellular debris. Subsequently, the supernatant is concentrated
using centrifugal filters and afterward the concentrated CM is spun
at high-speed for pelleting of MVs. The remaining supernatant is
then centrifuged at ultra-speed for pelleting exosomes. The result-
ing supernatant from the final centrifugation represents the
EV-depleted CM, and hence contains secreted or shed proteins
(although some proteins might be lost in the concentration step).
Secretome/vesiculome proteins and their relative abundances in
EVs and CM can be characterized using bottom-up MS-based
shotgun proteomics.
An inherent limitation of this protocol is the requirement for
serum-free conditioned media. If the cells under investigation can
survive without serum for 48 h, this would be the most ideal
condition. However, if serum-starvation is deleterious to viability,
the duration of serum-free condition could be either reduced,
serum concentration could be lowered (0.1–1%) or EV-depleted
serum could be used (see Subheading 4). However, the presence of
any amount of serum will skew the proteomic analysis of secreted
cellular proteins present in CM [24]. This problem can be resolved
using a click-chemistry based method as introduced by Eichelbaum
et al. [25]. Here, cells were depleted of methionine, arginine and
 Profiling of Secreted Proteins, Exosomes and Microvesicles 95

lysine and subsequently cultured in CM with L-azidohomoalaine

(AZA, a methionine analog with an azido group), heavy labeled L-
arginine and L-lysine. Newly synthesized secreted proteins contain
incorporation of AZA, and can be enriched from CM using alkyne
labeled agarose resin, followed by sample preparation for MS.
In conclusion, intercellular signaling is one hallmark of multi-
cellular organisms and crucially involved in all biological processes,
including disease progression. Profiling strategies with an emphasis
on studying secreted factors can aid in the discovery of biomarkers
and improve our understanding of cellular activation pathways and
therapeutic targets. Here, we present a method in five steps for
profiling secreted proteins, and proteins present in exosomes and
MVs from CM.

2 Materials

2.1 Reagents Sterile 1. 2,2,2-Trifluoroethanol (TFE).

Reagents, Media, or 2. Mass spectrometry grade Trypsin/Lys-C protease mix (Pro-
Solutions Such mega Corporation, Madison, WI, USA).
as Phosphate Buffered
3. Mass spectrometry grade trifluoroacetic acid (TFA).
Saline (PBS, Room
Temperature and 4
C)
4. Mass spectrometry grade solvents (water, acetonitrile, and
formic acid).
and Appropriate
Tissue Culture Media
Without Phenol Red
(See Note 1)

2.2 Mammalian Cell 1. Ten 15 cm diameter tissue culture dishes.

Culture 2. Secretome media: appropriate tissue culture media (without
and Conditioned phenol red) supplemented with 100 U/mL penicillin–strepto-
Medium mycin, 2 mM L-glutamine (see Note 2).
3. Several 15 and 50 mL conical tubes.

2.3 Exosome 1. Refrigerated centrifuge capable of reaching 300  g, 2000  g,

Isolation and 10,000  g (e.g., Eppendorf 5810R).
2. 15 mL Centrifugal filters with 3 kDa MWCO (catalog number
UFC900396) (EMD Millipore, Ontario, Canada).
3. 10 mL ultracentrifuge polycarbonate tube (product number
355603; Beckman Coulter Inc., Indianapolis, IN, USA).
4. Ultracentrifuge capable of 120,000  g (e.g., Type 70.1 Ti,
Beckman Coulter).

2.4 TFE Based 1. Denaturation buffer: 50% (v/v) TFE in PBS (pH 7.4).
Digestion of Proteins 2. Bicinchoninic acid (BCA) assay for protein determination.
96 Ankit Sinha et al.

3. 500 mM dithiothreitol (DTT).

4. Freshly prepared 1 M iodoacetamide (IAA).
5. Freshly prepared 100 mM Ammonium bicarbonate buffer
(pH 8.0). The pH of this solution is naturally 8.0. Nevertheless
we recommend to confirm by Litmus paper.
6. 2 M CaCl2.
7. Incubator set at 37
C.

2.5 Preparation 1. Desalting tips: OMIX C18 pipette tips (Agilent Technologies,
of Samples for Mass Santa Clara, CA, USA).
Spectrometry 2. Vacuum centrifuge.
3. Priming solution: 80% acetonitrile and 0.1% formic acid in
water.
4. Washing solution: 0.1% formic acid in water.
5. Elution solution: 60% acetonitrile and 0.1% formic acid in
water.

2.6 Liquid 1. Nano-flow ultra-performance liquid chromatography system

Chromatography (such as Easy-nLC 1000, Thermo Scientific).
and Mass 2. High-resolution Orbitrap tandem mass spectrometer (QExac-
Spectrometry tive, ThermoFisher Scientific, Carlsbad, CA, USA).
3. Buffer A: MS grade water with 0.1% MS formic acid.
4. Buffer B: MS grade acetonitrile with 0.1% MS formic acid.

3 Methods

3.1 Obtaining The suggested volumes are for monolayer adherent cells growing in
Conditioned Media 15 cm dishes. All cells are incubated at 37
C in 5% CO2 condition.
1. Culture cells in ten 15-cm dishes using appropriate cell culture
medium, beginning with a confluence of 30–40% (see Note 3).
2. Cells should be cultured to a confluence level where they can be
projected to reach 70–80% confluence within the next 48 h (see
Note 4).
3. Once the adhered cells have reached the predetermined con-
fluence level; aspirate and discard culture media and gently
wash the cells three times with 10 mL of room
temperature PBS.
4. Add 20 mL of secretome media to each plate and incubate at
37
C for 48 h.
5. After 48 h, carefully aspirate and pool together the conditioned
media (using a 25 mL serological pipette) from the ten plates in
four 50 mL conical tubes. Store conditioned media in an ice
bath at 4
C.
 Profiling of Secreted Proteins, Exosomes and Microvesicles 97

3.2 Differential 1. Centrifuge the conditioned media at 300  g for 10 min at

Centrifugation 4
C for pelleting nonadherent cells.
of Conditioned 2. Aspirate the supernatant and transfer to a new set of 50 mL
Medium conical tubes. Centrifuge at 2000  g for 20 min at 4
C for
pelleting cellular debris.
3. Concentrate the supernatant conditioned medium using Ami-
con 3 kDa cut-off filters. Prime centrifugal filters using PBS and
then filter concentrate the supernatant conditioned media from
200 mL to ~9 mL, according to the manufacturer’s instruc-
tions. The total concentration step may take 2–3 h, but can be
reduced by use of multiple spin filters in parallel.
4. Centrifuge the filter concentrate at 10,000  g for 30 min at
4
C to collect microvescicles. The pellet contains the MV
enriched fraction. Resuspend the MV pellet in appropriate
buffer based on downstream application. For proteomics,
resuspend the MV pellet twice in 75 μL of denaturation buffer
and transfer to a 1.5 mL micro-tube (see Note 5).
5. Transfer the supernatant to an ultracentrifuge tube.
6. Ultracentrifuge at 120,000  g at 4
C for 120 min.
7. Carefully aspirate and transfer the supernatant conditioned
media to a 15 mL conical tube and store at 4
C. This is the
EV depleted, secreted protein fraction. Using a waterproof
marker, circle the region on the tube that denotes the exosome
pellet.
8. Resuspend the pellet in the ultracentrifuge tube using ice-cold
PBS (see Note 6).
9. Ultracentrifuge at 120,000  g at 4
C for 120 min.
10. Carefully aspirate the supernatant PBS and discard. The
exosome-enriched pellet should be present in the marked
region on the tube. Resuspend the exosome pellet twice in
75 μL of denaturation buffer and transfer to a 1.5 mL
microtube.

3.3 Digestion 1. Perform BCA assay to determine the total amount of protein
of Proteins present in the MVs, exosome and conditioned media fractions.
2. Aliquot the conditioned media based on protein amount
equating to total protein in the exosome and MV fraction (see
Note 7).
3. Add an equal volume of TFE to the conditioned media (see
Note 8).
4. Heat denature all the samples at 60
C for 120 min.
5. Reduce proteins by adding DTT to a final concentration of
5 mM. Incubate at 60
C for 30 min.
98 Ankit Sinha et al.

6. Allow the samples to cool to room temperature and add IAA to

a final concentration of 25 mM. Alkylation should proceed for
30 min at room temperature in the dark.
7. Dilute samples to five time the volume using 100 mM ammo-
nium bicarbonate.
8. Add CaCl2 to a final concentration of 2 mM.
9. Add Trypsin/Lys-C mix for enzymatic digestion of protein at
ratio of 1:50 (trypsin: total protein amount).
10. Digest overnight at 37
C with constant rotation.
11. Stop digestion by adding TFA to final concentration of 2%.
12. Centrifuge samples at >10,000  g in benchtop centrifuge for
15 min.
13. Transfer the supernatant to a new microtube and discard the
pellet (if present).

3.4 Desalting 1. Vacuum-concentrate the supernatants to a volume of ~200 μL

of Peptides (see Note 9).
2. Prime the desalting tips using 250 μL of priming solution.
3. Equilibrate the desalting tips using 250 μL of wash solution.
4. Bind samples on the desalting tip.
5. Wash the desalting tips using 500 μL wash solution.
6. Elute the peptides from the desalting tips using 100 μL of
elution solution.
7. Concentrate the peptides in a vacuum concentrator to a semi-
dry state (approximately 2–5 μL).

3.5 Analysis Below we describe the proteomics approach that has provided good
of Peptides Using results in our hands. Briefly, in bottom-up MS driven shotgun
Tandem Mass proteomics, digested peptides are analyzed using liquid chroma-
Spectrometry tography coupled to tandem mass spectrometry (LC-MS/MS).
This involves separation of peptides based on hydrophobicity (use
of reverse phase is highly recommended) using a nano-flow ultra-
performance liquid chromatography system (such as Easy-nLC
1000) coupled to a heated 50-cm column packed with 2-μm Pep-
Map C18 particles. The liquid chromatography system is
connected to a high-resolution Orbitrap tandem mass spectrometer
(such as QExactive). Considering that LC-MS/MS systems from
different vendors will have their own specific settings, the method
described below should be used as guidelines.
1. Add 10–13 μL of Buffer A to reach a sample final volume of
~15 μL.
2. Measure total peptide amount using a NanoDrop.
 Profiling of Secreted Proteins, Exosomes and Microvesicles 99

3. For mass spectrometry analysis, inject an equal amount of

peptides for all samples. Ideal range is usually between 0.75
and 1.5 μg of peptides per injection.
4. Reverse phase liquid chromatography is performed at a flow
rate of 250 nL/min in a split-less system with column temper-
ature of 50
C (in our case without using a precolumn; i.e.,
one-column setup). The steps in the chromatography gradient
consisted of:
(a) 0–5% B in 5 min
(b) 5–30% B in 230 min
(c) 30–95% B in 5 min
(d) Hold 95% B for 10 min
5. Ideally, a high-resolution mass spectrometer should be utilized,
such as a QExactive used for our experiments. The MS system
operated in a top-10 data dependent acquisition method, here
a single MS scan was performed at 70,000 resolution followed
by ten MS/MS scans at 17,500 resolution with dynamic exclu-
sion of 20 s.
6. All raw data from MS are analyzed using MaxQuant [26],
which is best suited for Thermo Scientific instruments. Data
from MS systems from other vendors can be analyzed using
additional open access workflows such as Trans-Proteomic
Pipeline or Peaks [27, 28]. False discovery of peptide and
protein identifications is controlled using a target-decoy
approach with false discovery rate set to a defined rate of 1%.
Protein groups identified with at least two or more peptides are
carried forward for subsequence analysis.
7. Statistical analyses such as variance, protein ratio, and visualiza-
tion of expression profiles are performed using the R software
environment (http://www.r-project.org/).

4 Notes

In our hands, this protocol has produced data with high precision
and robustness. We recently used this exosome isolation method
for quantitative comparisons across various ovarian cancer cell lines
[21]. We strongly advise the use of MS grade reagents for Subhead-
ing 3.3 and onward. The following notes are provided as a sugges-
tion for users of this protocol.
1. Phenol red is used as a pH indicator in cell culture media. Cells
are cultured in media without phenol red for only 48 h, and
hence pH should not drastically change over such a short time
period.
100 Ankit Sinha et al.

2. For isolation of secreted proteins present in the conditioned

media, as well as exosomes and MVs, it is advised to use cell
culture medium without fetal bovine serum (FBS). First of all,
abundant FBS proteins will mask the signal of proteins secreted
from the cells. Secondly, FBS is derived from bovine serum and
contains a high abundance of EVs. These EVs can interfere or
cause significant background issues when studying exosomes or
MVs secreted from standard culture conditions. In such cases
EV-depleted serum can be used. EV-depleted serum can be
prepared by ultracentrifugation at 120,000  g overnight at
4
C. The resulting pellet will contain exosomes, MVs, and
other vesicles, and the supernatant can be considered
EV-depleted serum.
3. Initial cell cultivation should be performed in complete media
supplemented with FBS.
4. The incubation of cells in the secretome media should be
initiated at a confluence level where the cells will not exceed
70–80% confluence in the next 48 h. Doubling time of cells
should be considered to estimate the initial confluence level for
starting the 48 h incubation.
5. We recommend resuspending the pellet in 75 μL of denatur-
ation buffer, and use of another 75 μL to rinse any remaining
pellet off the bottom of the tube.
6. Washing EV pellets in PBS minimizes carryover of contaminat-
ing proteins.
7. For digestion of proteins, it is recommended to start with equal
protein amounts for all three (MVs, exosomes, and secreted
proteins) fractions, if a comparison is desired. The amount of
protein to use should be determined by the protein-limiting
sample. If the aliquot of the conditioned media exceeds
150 μL, we highly recommend further concentration using
centrifugal filters, until the volume required for appropriate
protein amount reaches approximately 150 μL or less. This
will prevent the use of highly diluted samples (> 1 mL) for
proteolysis.
8. The secreted proteins are already present in aqueous buffer,
and hence addition of an equal volume of 100% TFE is suffi-
cient for denaturing proteins.
9. The maximum volume for the desalting tips is 200 μL. Hence,
the solvent in the samples need to be evaporated (using a
vacuum concentrator) if it exceeds 200 μL. Every step
(priming, equilibrating, binding, and wash and elution) is per-
formed in ten cycles, where each cycle comprises of aspirate and
dispense step.
 Profiling of Secreted Proteins, Exosomes and Microvesicles 101

Acknowledgment

This study was funded by the Canadian Institutes of Health

Research (CIHR) to T.K. (MOP-133615). A.S. was supported
through a CIHR Doctoral Award. S.P. was supported by the
CIHR Terry Fox Foundation Strategic Training Initiative for
Excellence in Radiation Research in the 21st Century. A.I. was
supported in parts by the PMH Head and Neck Translational
group. T.K. is supported by the Canada Research Chair Program.
Support is also provided from the Campbell Family Institute for
Cancer Research and the Ministry of Health and Long-term
Planning.

References
1. Hanahan D, Weinberg RA (2011) Hallmarks 10. Yoon YJ, Kim OY, Gho YS (2014) Extracellular
of cancer: the next generation. Cell vesicles as emerging intercellular communica-
144:646–674 somes. BMB Rep 47:531–539
2. Tlsty TD, Coussens LM (2006) Tumor stroma 11. Beach A, Zhang HG, Ratajczak MZ, Kakar SS
and regulation of cancer development. Annu (2014) Exosomes: an overview of biogenesis,
Rev Pathol 1:119–150 composition and role in ovarian cancer. J Ovar-
3. Schiarea S, Solinas G, Allavena P, Scigliuolo ian Res 7(1):14
GM, Bagnati R, Fanelli R, Chiabrando C 12. Raposo G, Stoorvogel W (2013) Extracellular
(2010) Secretome analysis of multiple pancre- vesicles: exosomes, microvesicles, and friends. J
atic cancer cell lines reveals perturbations of key Cell Biol 200:373–383
functional networks. J Proteome Res 13. Saleem SN, Abdel-Mageed AB (2014) Tumor-
9:4376–4392 derived exosomes in oncogenic reprogram-
4. Xu BJ, Yan W, Jovanovic B, An AQ, Cheng N, ming and cancer progression. Cell Mol Life
Aakre ME et al (2010) Quantitative analysis of Sci 72:1–10
the secretome of TGF-beta signaling-deficient 14. Peinado H, Aleckovic M, Lavotshkin S,
mammary fibroblasts. Proteomics Matei I, Costa-Silva B, Moreno-Bueno G et al
10:2458–2470 (2012) Melanoma exosomes educate bone
5. Mathias RA, Wang B, Ji H, Kapp EA, Moritz marrow progenitor cells toward a
RL, Zhu HJ, Simpson RJ (2009) Secretome- pro-metastatic phenotype through MET. Nat
based proteomic profiling of Ras-transformed Med 18:883–891
MDCK cells reveals extracellular modulators of 15. Lu M, Huang B, Hanash SM, Onuchic JN,
epithelial-mesenchymal transition. J Proteome Ben-Jacob E (2014) Modeling putative thera-
Res 8:2827–2837 peutic implications of exosome exchange
6. Luo X, Liu Y, Wang R, Hu H, Zeng R, Chen H between tumor and immune cells. Proc Natl
(2011) A high-quality secretome of A549 cells Acad Sci U S A 7:E4165–E4174
aided the discovery of C4b-binding protein as a 16. Lundholm M, Schroder M, Nagaeva O,
novel serum biomarker for non-small cell lung Baranov V, Widmark A, Mincheva-Nilsson L,
cancer. J Proteome 74:528–538 Wikstrom P (2014) Prostate tumor-derived
7. Oksvold MP, Neurauter A, Pedersen KW exosomes down-regulate NKG2D expression
(2015) Magnetic bead-based isolation of exo- on natural killer cells and CD8þ T cells: mech-
somes. Methods Mol Biol 1218:465–481 anism of immune evasion. PLoS One 9:
8. Thery C, Zitvogel L, Amigorena S (2002) Exo- e108925
somes: composition, biogenesis and function. 17. Luga V, Zhang L, Viloria-Petit AM, Ogunjimi
Nat Rev Immunol 2:569–579 AA, Inanlou MR, Chiu E et al (2012) Exo-
9. Huang X, Yuan T, Tschannen M, Sun Z, somes mediate stromal mobilization of auto-
Jacob H, Du M et al (2013) Characterization crine Wnt-PCP signaling in breast cancer cell
of human plasma-derived exosomal RNAs by migration. Cell 151:1542–1556
deep sequencing. BMC Genomics 14:319 18. Shimoda M, Principe S, Jackson HW, Luga V,
Fang H, Molyneux SD et al (2014) Loss of the
102 Ankit Sinha et al.

Timp gene family is sufficient for the acquisi- 23. Baietti MF, Zhang Z, Mortier E, Melchior A,
tion of the CAF-like cell state. Nat Cell Biol Degeest G, Geeraerts A et al (2012) Syndecan-
16:889–901 syntenin-ALIX regulates the biogenesis of exo-
19. Valenti R, Huber V, Filipazzi P, Pilla L, somes. Nat Cell Biol 14:677–685
Sovena G, Villa A et al (2006) Human tumor- 24. Ebert MP, Korc M, Malfertheiner P, Rocken C
released microvesicles promote the differentia- (2006) Advances, challenges, and limitations in
tion of myeloid cells with transforming growth serum-proteome-based cancer diagnosis. J
factor-beta-mediated suppressive activity on T Proteome Res 5:19–25
lymphocytes. Cancer Res 66:9290–9298 25. Eichelbaum K, Winter M, Berriel Diaz M,
20. Aliotta JM, Sanchez-Guijo FM, Dooner GJ, Herzig S, Krijgsveld J (2012) Selective enrich-
Johnson KW, Dooner MS, Greer KA et al ment of newly synthesized proteins for quanti-
(2007) Alteration of marrow cell gene expres- tative secretome analysis. Nat Biotechnol
sion, protein production, and engraftment into 30:984–990
lung by lung-derived microvesicles: a novel 26. Cox J, Mann M (2008) MaxQuant enables
mechanism for phenotype modulation. Stem high peptide identification rates, individualized
Cells 25:2245–2256 p.p.b.-range mass accuracies and proteome-
21. Sinha A, Ignatchenko V, Ignatchenko A, wide protein quantification. Nat Biotechnol
Mejia-Guerrero S, Kislinger T (2014) 26:1367–1372
In-depth proteomic analyses of ovarian cancer 27. Deutsch EW, Mendoza L, Shteynberg D,
cell line exosomes reveals differential enrich- Farrah T, Lam H, Tasman N et al (2010) A
ment of functional categories compared to the guided tour of the Trans-Proteomic Pipeline.
NCI 60 proteome. Biochem Biophys Res Proteomics 10:1150–1159
Commun 445:694–701 28. Zhang J, Xin L, Shan B, Chen W, Xie M, Yuen
22. Lee TH, D’Asti E, Magnus N, Al-Nedawi K, D et al (2012) PEAKS DB: de novo sequencing
Meehan B, Rak J (2011) Microvesicles as med- assisted database search for sensitive and accu-
iators of intercellular communication in can- rate peptide identification. Mol Cell Proteo-
cer--the emerging science of cellular ‘debris’. mics 11:M111.010587
Semin Immunopathol 33:455–467
 Part II

Targeted Approaches for Surfaceome Content

 Chapter 7

Cloning, Expression, and Purification of the Glycosylated

Transmembrane Protein, Cation-Dependent Mannose
6-Phosphate Receptor, from Sf9 Cells Using
the Baculovirus System
Linda J. Olson and Nancy M. Dahms

Abstract
The cation-dependent mannose 6-phosphate receptor (CD-MPR) is a single-pass type I membrane protein.
This protein functions to transport lysosomal enzymes displaying phosphomannosyl residues from the
Golgi complex and the cell surface to the lysosome. This glycosylated protein contains three disulfide
bridges in its 159-residue extracytoplasmic domain. One of the problems with studying eukaryotic mem-
brane proteins is isolating sufficient quantities. Structural studies typically require several hundred milli-
grams of highly purified protein. Here we present a method to isolate milligram quantities of CD-MPR/
Asn81 suitable for structural studies.

Key words CD-MPR, Transmembrane protein, Sf9 cells, Purification, Glycoprotein

1 Introduction

The P-type lectin family of receptors is comprised of the cation-

independent mannose 6-phosphate receptor (CI-MPR) as well as
the significantly smaller cation-dependent mannose 6-phosphate
receptor (CD-MPR) [1]. These transmembrane receptors are
involved not only in delivering newly synthesized lysosomal
enzymes to the lysosome but are also themselves trafficked to the
cell surface where they bind to proteins displaying
phosphomannosyl-modified carbohydrates. These receptors are
trafficked throughout the cell by several sorting signals present in
their cytoplasmic tails [2]. Although multiple structural studies
have been conducted on the soluble cytosolic region of
CD-MPR, studies on the full-length protein have been limited.
Insect cells, and in particular Sf9 cells, have long been used to
express recombinant glycosylated proteins. We have now used this
expression system to generate the full-length transmembrane

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_7, © Springer Science+Business Media, LLC 2018

105
106 Linda J. Olson and Nancy M. Dahms

protein, CD-MPR, at high enough quality and in quantities suffi-

cient to be utilized in structural studies. To aid in crystallization,
the same single glycosylation site mutant, termed CD-MPR/
Asn81, used to obtain the soluble structure of the receptor has
been maintained [3].
In this chapter, we briefly describe the modification of a stan-
dard transfer vector so that high levels of expression can be achieved
in Sf9 cells. This information is provided to users who may not have
access to an appropriate vector suitable for targeting recombinant
proteins through the secretory pathway in Sf9 cells. The
subsequent cloning step, which should be standard in any molecu-
lar biology laboratory, is only briefly described and serves as an
introduction to novices planning on using this system. We then
describe the use of the baculovirus system to generate functional
recombinant CD-MPR in quantities sufficient for structural stud-
ies. All steps involving Sf9 cells are described in detail. The result is a
protein that is glycosylated and retained in the membrane. As
presented here, the protocol is divided into seven steps: cloning
the gene of interest (Subheading 3.1); culturing (Subheading 3.2)
and transfection of Sf9 cells (Subheading 3.3); generation (Sub-
heading 3.4) and titering (Subheading 3.5) of viral stocks; expres-
sion and purification of the full-length protein (Subheading 3.6);
and characterization of the purified protein. This recombinant
approach, while described specifically for CD-MPR, can be applied
with modifications to other glycosylated proteins that are trafficked
to the cell surface.

2 Materials

2.1 Instruments 1. MaxQ™ 4000 orbital benchtop shaker (Thermo Fisher Scien-
and Other General tific, Waltham, MA, USA) or similar.
Materials 2. Standard tissue culture hood.
3. Inverted microscope.
4. EVOS® FL Auto Imaging System.
5. ÄKTA Start chromatography system (GE Healthcare Life
Sciences, Pittsburgh, PA, USA).
6. NanoDrop 2000c (Thermo Fisher Scientific, Waltham, MA,
USA).
7. Branson 450 or similar sonicator .
8. Low speed centrifuge capable of holding a JLA8.1000 rotor
and a JA-20 rotor or similar (Beckman Coulter, Indianapolis,
IN, USA).
9. Ultracentrifuge capable of holding a Ti-75 rotor or similar
(Beckman Coulter).
10. 20 mL syringe and large bore (20G) needle.
 Transmembrane Proteins from Sf9 Cells 107

2.2 Insect Cells Sf9 Expression Systems (Expression Systems, Davis, CA, USA)
(www.ExpressionSystems.com) (see Note 1).

2.3 Media 1. ESF 921 Medium: This is a protein-free insect cell medium
and Transfection appropriate for cell propagation as well as protein production
Reagents (www.ExpressionSystems.com).
2. BestBac 2.0 v-cath/chiA Deleted Baculovirus Cotransfection
Kit: This kit contains the viral DNA backbone, transfection
medium and transfection reagent (www.ExpressionSystems.
com).

2.4 Vectors 1. pVL1392: This transfer vector may be obtained from several
and Plasmid sources including Expression Systems (www.
Purification expressionsystems.com).
2. PureYield™ Plasmid Miniprep System (Promega, Madison,
WI, USA): Transfection efficiency is dependent on quality of
the plasmid DNA (see Note 2).

2.5 Plasticware 1. Propagation and protein expression were done in Corning®

for Cell Transfection, disposable sterile bottles.
Propagation, 2. Transfections and immunofluorescent studies were done in
and Protein Expression 24-well Falcon™ plates.

2.6 Chemicals 1. Lysis buffer: 20 mM Tris base pH 7.8 at 4
C, 300 mM NaCl,

and Miscellaneous 0.1 mM phenylmethylsulfonyl fluoride (PMSF).
Reagents 2. Membrane wash buffer: 20 mM Tris base pH 7.8 at 4
C,
150 mM NaCl.
3. Membrane solubilization buffer: 20 mM Tris base pH 7.8 at
4
C, 150 mM NaCl, 0.1% sodium deoxycholate, 1% Triton
X-100.
4. Ni-wash and elution buffers: 20 mM Tris base pH 7.8 at 4
C,
150 mM NaCl, 0.05% Triton X-100 and either 20, 50, 100, or
400 mM imidazole pH 7.5.
5. HisTrap™ excel 1 mL columns prepacked with Ni Sepharose
excel affinity media for capture and purification of histidine-
tagged proteins (GE Healthcare Life Sciences, Pittsburgh, PA,
USA) (www.gelifesciences.com).
6. Bio-Rad protein assay solution (www.bio-rad.com) (Bio-Rad,
Hercules, CA, USA).
7. Borosilicate glass tubes, 13  100 mm or similar.
8. Phosphate buffered saline (PBS): 137 mM NaCl, 2.7 mM KCl,
1.5 mM KH2PO4, 15 mM Na2HPO4, adjust pH to 7.4 with
KH2PO4.
9. PBS with 0.1% bovine serum albumin (BSA): add 0.1% (w/v)
BSA to PBS.
108 Linda J. Olson and Nancy M. Dahms

10. CD-MPR antibody: provided by NMD [4].

11. Chicken anti-rabbit Alexa Fluor® 594 antibody (Thermo
Fisher Scientific).

3 Methods

3.1 Cloning Gene Although companies have developed different vectors for expres-
of Interest sing foreign proteins in Sf9 cells, we have chosen to remain with
one of the early transfer vectors, pVL1392. This transfer vector
3.1.1 Modification contains features for growth and maintenance in standard E. coli
of pVL1392 strains such as DH5α and JM109. This 9.8 kb plasmid contains the
pMB1 origin of replication from pUC19 for high copy number
growth. Sandwiched between the 50 and 30 -viral sequences required
for homologous recombination is the very strong polyhedron
promoter, which drives recombinant protein expression. This vec-
tor is lacking one key feature needed for recombinant protein
expression: a cleavable signal sequence which directs the recombi-
nant protein to the endoplasmic reticulum for entry into the secre-
tory pathway, an essential targeting signal for secreted and
membrane-bound proteins. To overcome this limitation, we have
chosen to use the honeybee melittin signal peptide: 5-
0
-MKFLVNVALVFMVVYISYIYA-30 (Fig. 1) [6].

3.1.2 Cloning CD-MPR The DNA encoding the signal peptide was synthesized and
into Modified pVL1392 digested as indicated in Fig. 1. Following successful incorporation
of the DNA encoding the signal sequence, the plasmid was digested
with EcoRI and XbaI and treated with FAST-alkaline phosphatase
(AP). The purified PCR product encoding CD-MPR/Asn81 was
also digested with the above two restriction enzymes. The digests
were subjected to standard electrophoresis on a 1% low melting
point agarose gel and the desired bands excised. Standard cloning
procedures were followed and final clones were subjected to DNA
sequencing to verify the sequence.

Fig. 1 Schematic diagram depicting modifications made to the standard pVL1392 transfer vector. The DNA
sequence encoding the melittin signal peptide (translated DNA sequence shown in red) along with a octa-His
tag and GS from an added restriction site (blue) and CD-MPR (black) has been modified to remove
glycosylation at all but one (N81Q) position that was previously determined to be critical for proper folding [5]
 Transmembrane Proteins from Sf9 Cells 109

3.2 Culturing Sf9 Sf9 cells obtained in suspension culture were propagated and fro-
Insect Cells zen as directed by the company. Cells were propagated by thawing a
1 mL vial of 50  106 cells at 37
C and diluting the cells into a
250 mL sterile storage bottle containing 30 mL ESF 921 medium.
1. Cells are grown at 27
C in suspension cultures while shaking at
140 rpm in a MaxQ™ 4000 orbital benchtop shaker.
2. Cells were seeded in ESF 921 medium at a density of
0.75–1  106cells/mL in either 250 mL or 1 L sterile Corn-
ing® storage bottles. The 250 mL bottles contained
30–100 mL of liquid culture while the 1 L bottles contained
400 mL suspension culture (see Note 3).
3. Cells were passaged when they reached a density of
4–6  106cells/mL (typically every 2–3 days).

3.3 Transfection The transfection protocol is based on the published protocol from
of Sf9 Cells Expression Systems with modifications to utilize a 24-well Falcon
plate. All steps are carried out in a standard tissue culture hood
using sterile technique.
1. Seed cells at a density of 1.8  105 cells/well of a 24-well
Falcon plate. Allow the cells to adhere to the plate ~30 min
while transfection solutions are made (see Note 4).
2. Combine gently in a polypropylene tube (solution A) 0.4 μg of
CD-MPR/pVL1392, 1 μL BestBac 2.0, v-cath/chiA Deleted
linearized viral DNA, 20 μL Transfection Medium.
3. Combine together in a polypropylene tube (solution B) 1–2 μL
supplied transfection reagent and 20 μL Transfection Medium.
4. Let the solutions sit for 5 min in the hood.
5. Combine solution B into A and incubate in the tissue culture
hood for ~30 min at room temperature.
6. Add 160 μL Transfection Medium to the tube containing the
combined solutions A and B.
7. Gently remove the medium from the plated adherent cells and
replace with the solution from step 6.
8. Place tray in a tight-sealing, small plastic container and incubate
at 27
C 4–5 h (see Note 5).
9. Remove the transfection mix and feed cells with 600 μL ESF
921.
10. Place tray back in plastic container being sure it is sealed and
incubate at 27
C without shaking.
11. Collect medium containing P-0 virus after 5 days, centrifuge
5 min at 500  g to remove any cells (see Note 6).
110 Linda J. Olson and Nancy M. Dahms

3.4 Generation 1. Seed the cells at a density of 1  106cells/mL in 30 mL of ESF

of Virus Stocks 921 medium.
2. Inoculate these seeded cells with the 600 μL P-0 viral stock
from step 11 above.
3. Incubate with shaking (140 rpm) at 27
C 4 days.
4. Cells should stop multiplying before day 3 (4  106cells/mL)
and should appear larger in size compared to noninfected cells.
5. On day 4, count cells and take note of cell appearance. Cell
counts should be between 3 and 6  106 cells/mL depending
on the transfection efficiency (see Note 7).
6. Pellet cells using a swinging bucket centrifuge (500  g for
5 min) in a sterile 50 mL conical bottom centrifuge tube.
7. Sterile filter the supernatant. This supernatant is the P-1 virus
stock and is stored at 4
C in the dark.

3.5 Titering of Virus Determination of the optimal infection conditions for protein
Stocks expression is critical. The amount of virus stock required to infect
a given number of cells will vary from batch to batch of virus as well
as with the age of the viral stock. The amount of virus stock needed
to infect a given number of Sf9 cells is determined by infecting a
series of 30 mL cultures each containing 1  106 cells/mL.
1. Seed four 250 mL Corning® bottles with 30 mL of culture
containing 1  106 cells/mL.
2. Leave one bottle uninfected as a control and infect the other
three bottles with increasing volumes of virus stock: 25, 50 and
250 μL is typical.
3. Count the cells 3 and 4 days post infection (see Note 3) and
determine which flask(s) did not grow after infection. This
determination will indicate how much virus needs to be
added to arrest cell growth (see Note 8).

3.6 Expression The key to protein production in large quantities is having a large
and Purification of Full number of cells to infect at the same time (see Note 9). This
Length CD-MPR procedure involves two independent steps involving infection of
Sf9 cells and protein purification.
1. For the infection of Sf9 cells, count cells and infect with the
amount of virus determined above to allow one doubling (see
Note 10).
2. Cultures are grown as before with shaking at 140 rpm at 27
C.
It is important to monitor cell density, determined through
counting, until the culture has stopped growing.
3. Cells are ready for harvest 48 h after they have stopped divid-
ing, which typically occurs on the third day post-infection.
 Transmembrane Proteins from Sf9 Cells 111

Fig. 2 Immunodetection of CD-MPR in Sf9 cells following infection with recombinant baculovirus. Infected and
uninfected control Sf9 cells grown in suspension culture are allowed to adhere 30 min to a 24-well plate. After
attachment, cells are washed with PBS and fixed by incubation on ice for 20 min in the presence of 10%
formalin in PBS. Cells were rinsed with PBS plus 0.1% BSA before being incubated overnight at 4
C in the
dark in a solution of PBS plus 0.1% BSA and a 1:100 dilution of anti-CD-MPR (NMD) [4]. Cells were then rinsed
with PBS plus 0.1% BSA two times prior to incubation with a 1:500 chicken anti-rabbit Alexa 864 antibody.
Control, uninfected cells showed no detectable levels of fluorescence and are not shown. Images were taken
using a 40 objective, the bar representing 100 μm

Infected cells appear larger, but cell death should be less than
10% (Fig. 2).
4. Cells are harvested by centrifugation at 1000  g for 15 min in
a JLA8.1000 rotor. Cell pellets can be either frozen at 80
C
for future use or processed immediately.
5. For protein purification, cell pellets from 1.2 L of Sf9 culture
are resuspended in 150 mL of cold lysis buffer: 20 mM Tris,
pH 7.8 at 4
C, 300 mM NaCl, and 0.1 mM PMSF.
6. Cells are lysed by three 10 s bursts of sonication on ice (see
Note 11).
7. Cell debris is removed by centrifugation at 500  g for 10 min
in a JA-20 rotor.
8. The supernatants are collected and centrifuged at 80,000  g
for 60 min in a Ti-75 rotor (Beckman Coulter).
9. The membrane pellets are resuspended in 100 mL of wash
buffer: 20 mM Tris, pH 7.5 at 22
C, 150 mM NaCl.
112 Linda J. Olson and Nancy M. Dahms

10. The pellets are centrifuged again as in step 4 above.

11. The membrane pellets are resuspended in 100 mL of fresh cold
(4
C) wash buffer (see step 5). To ensure pellets are fully
resuspended, the solution is passed several times through a
large bore (20G) needle attached to a 20 mL syringe.
12. Triton X-100 (10%) stock solution is added to a final concen-
tration of 1% (v/v) along with 0.1% (w/v) of a 20% sodium
deoxycholate solution. The mixture is stirred in a cold room at
4
C for 1 h.
13. After 1 h, half (50 mL) of the mixture from step 12 is trans-
ferred to centrifuge tubes.
14. The membrane extracts are collected by centrifugation at
12,000  g in a JA20 rotor.
15. The extracts are loaded using a ÄKTA Start system at 2 mL/
min on a 1 mL HisTrap excel column preequilibrated with
10 column volumes of buffer: 20 mM Tris, pH 7.5 at 22
C,
150 mM NaCl, 0.05% Triton X-100.
16. The column is washed until the baseline returned to 0 milliab-
sorbance units (mAU).
17. The column is then washed with buffer containing increasing
concentrations of imidazole: 25, 50, 100, and 400 mM in
20 mM Tris, pH 7.5 at 22
C, 150 mM NaCl, 0.05% Triton
X-100. The baseline is allowed to approach 0 mAU before the
next concentration of imidazole was introduced.
18. The remaining 50 mL of membrane solution is allowed to stir
overnight (~18 h) at 4
C before being subjected to steps
13–17 above.

3.7 Preliminary We have used several standard techniques to evaluate if the length
Characterization of solubilization and subsequent longer exposure to higher deter-
of Purified Protein gent concentrations had any effect on the quality or quantity of the
transmembrane protein, CD-MPR/Asn81, being isolated. As
described here, four steps are involved: (1) SDS-PAGE analysis,
(2) Protein quantification, (3) Treatment with PNGaseF, and
(4) SPR analysis of binding affinity.
1. Following elution from the HisTrap excel Ni column, the
results of the purification are analyzed by SDS-PAGE. As this
is a standard laboratory protocol, details are not provided.
2. Proteins were visualized after staining with Coomassie Blue
(Fig. 3) (see Note 12).
3. For Protein Quantification, the Bio-Rad protein assay solution
is used to measure total protein concentration.
4. For this, water (800 μL) is aliquoted into borosilicate glass tubes.
5. An aliquot (10 μL) of the100 mM eluate (Subheading 3.6, step
17) is added to the water.
 Transmembrane Proteins from Sf9 Cells 113

Fig. 3 Purification of full-length CD-MPR. Ni-NTA column purification results

comparing solubilization for 1 h at 4
C compared to overnight (~18 h). A 12%
resolving/4% stacking SDS-PAGE gel was loaded with 4 μL of unstained
Benchmark™ Protein Ladder (lane 1) and 12 μL of the corresponding eluates:
buffer wash (lane 2), 20 mM imidazole (lane 3), 50 mM imidazole (lane 4),
100 mM imidazole (lane 5), and 400 mM imidazole (lane 6). Monomer (~28 kDa)
and dimer (~60 kDa) species are marked with arrows

6. Bio-Rad Protein Assay solution (200 μL) is added to each

sample followed by vortexing of the sample.
7. Samples are read on a NanoDrop 2000c spectrophotometer.
8. Concentrations are determined relative to a standard curve
determined from known amounts of bovine serum albumin.
9. The amount of total protein obtained from 1 h solubilization
compared to overnight should be comparable if solubilization
was complete within an hour (1.6 mg and 2.1 mg, 1 h versus
overnight).
10. To verify the presence of an N-linked oligosaccharide, protein
from the 100 mM imidazole eluates are incubated with PNGa-
seF (follow manufacturer’s recommendation for amount) over-
night at room temperature.
11. Samples are subjected to SDS-PAGE (4%/12%) and gels
stained with Coomassie Blue (Fig. 4) (see Note 12).
12. Surface Plasmon Resonance (SPR) Analysis of Binding Affinity
should then be performed, as CD-MPR binds to N-glycans
containing terminal mannose 6-phosphate residue(s) to traffic
lysosomal enzymes.
13. Following previously published procedures, SPR studies can be
used to characterize the interaction between CD-MPR and the
lysosomal enzyme acid α-glucosidase (GAA) (Fig. 5)
[7]. Details are not provided, as this technique may be more
apt for this glycoprotein.
114 Linda J. Olson and Nancy M. Dahms

Fig. 4 PNGaseF treatment of proteins to detect the presence of an N-linked

oligosaccharide. An aliquot of each 100 mM imidazole eluate was incubated
overnight at room temperature in the absence or presence of PNGaseF. The
cleavage reaction products were then resolved on a 12% resolving/4% stacking
SDS-PAGE gel. For reference, PNGaseF is loaded in the right lane marked with
an asterisk. Monomer (~28 kDa) and dimer (~60 kDa) species of CD-MPR are
indicated with arrows

Fig. 5 SPR studies characterizing the interaction between CD-MPR/Asn81 and GAA. Increasing concentrations
(10, 20, 40, 80, 120, 200, and 400 nM) of the CD-MPR/Asn81 protein are flowed over the GAA monoester
coupled surface for 120 s followed by buffer and regeneration with 10 mM M6P. The response at equilibrium
(Req) of the sensorgrams are plotted versus the concentration of protein and fit to a 1:1 binding isotherm
(inset). Data from the 1 h detergent solubilization protocol are shown. The time of solubilization with 1% Triton
X-100/0.1% sodium deoxycholate did not significantly affect binding affinity, and resulted in similar kD values
of 100 and 81 nM, respectively, for 1 h and overnight incubation times
 Transmembrane Proteins from Sf9 Cells 115

4 Notes

1. Sf9 cells isolated from the ovarian tissue of the fall army worm,
Spodoptera frugiperda, were selected for their faster growth rate
and ability to reach higher densities than some of the other
established lepidopteran insect cell lines. This cell line is able to
be maintained in monolayer cultures for transfections as well as
in shake culture for protein production. They are suitable for
use in all phases of protein generation from transfection, to
virus propagation, to protein production [8].
2. This kit from Promega has provided very reliable results.
3. Cells are counted using a standard hemocytometer. Cells are
split on Mondays and Wednesdays at 1  106 cells/mL. Cells
are split at a lower density of 0.75  106 cells/mL on Fridays to
avoid overgrowth by Monday.
4. In order to ensure cells are at ~75% confluence, plate wells at
the calculated number of cells and also 1.5 and 2 times the
calculated number of cells.
5. If nonshaking incubator is not available to incubate the plate, a
shaking floor incubator can be utilized provided some type of
stable, nonmoving platform can be installed. In the MaxQ™
4000 orbital benchtop shaker, enough stationary floor space is
provided to rest the legs of a homemade metal shelf. This shelf
is tall enough to allow bottles to rest on the shaking platform
underneath.
6. Failure to remove cells may result in higher cell densities and
lower infection rates.
7. In our hands, viral stocks grown in this manner and having
these final cell counts will yield a virus stock of 1  109 virus
units/mL.
8. Also note which flask doubled in cell number since it is often
advantageous to allow the cells to double once after infection in
order to increase cell numbers and accordingly protein yields.
9. If six bottles, each with 400 mL of cells at 4  106 cells/mL,
are needed on Friday and today is Monday, counting back,
halving the volume each day until the current day will deter-
mine how many cells are needed (Friday: 6  400, Thursday:
3  400, Wednesday: 1.5  400, Tuesday: 300 mL, Monday:
150 mL at 4  106 cells/mL today).
10. If your 25 μL virus inoculum resulted in no cell growth then
(30 mL (1  106 cells/mL))/0.025 mL ¼ 1.25  109 infect-
ing units/mL.
11. The extent of cell disruption was evaluated by pipetting 20 μL
onto a transparent surface and viewing by a light microscope.
116 Linda J. Olson and Nancy M. Dahms

12. Samples were resolved on a standard SDS-PAGE gel using a

12% resolving gel overlaid with a 4% stacking gel. Gels were
electrophoresed at 190 V for 45 min in standard running
buffer (24.8 mM Tris base, 192 mM glycine, 0.1% SDS).

Acknowledgment

This work supported by NIH grant R01 DK042667 to NMD.

References

1. Dahms NM, Hancock MK (2002) P-type lec- cation-dependent mannose 6-phosphate recep-
tins. Biochim Biophys Acta 1572:317–340 tor: effects on ligand binding, intracellular tar-
2. Ghosh P, Dahms NM, Kornfeld S (2003) Man- getting and association with binding
nose 6-phosphate receptors: new twists in the immunoglobulin protein. Biochem J
tale. Nat Rev Mol Cell Biol 4:202–213 295:841–848
3. Roberts DL, Weix DJ, Dahms NM, Kim J-JP 6. Tessier DC, Thomas DY, Khouri HE, Laliberte
(1998) Molecular basis of lysosomal enzyme VY (1991) Enhanced secretion from insect cells
recognition: three-dimensional structure of the of a forgein protein fused to the honeybee melit-
cation-dependent mannose 6-phosphate recep- tin signal peptide. Gene 98:177–183
tor. Cell 93:639–648 7. Chavez CA, Bohnsack RN, Kudo M, Gotschall
4. Dahms NM, Brzycki-Wessell MA (1995) RR, Canfield WM, Dahms NM (2007) Domain
Expression and characterization of functional 5 of the cation-independent mannose
bovine cation-dependent mannose 6-phosphate 6-phosphate receptor preferentially binds phos-
receptors in baculovirus-infected insect cells. phodiesters (mannose 6-phosphate N-acetylglu-
Arch Biochem Biophys 317:497–503 cosamine ester). Biochemistry 46:12604–12617
5. Zhang Y, Dahms NM (1993) Site-directed 8. Jarvis D (2009) Baculovirus-insect cell expres-
removal of N-glycosylation sites in the bovine sion systems. Methods Ezymol 463:191–222
 Chapter 8

Bispecific Antibody Armed T Cells to Target Cancer Cells

Archana Thakur, Lawrence G. Lum, and Sandeep Mittal

Abstract
The common strategy for making bispecific antibodies (BsAbs) involves combining the variable domains of
the desired monoclonal antibodies (mAbs) into a single bispecific structure. Bispecific immunotherapeutics
has generated many different formats of BsAbs including chemical heteroconjugation of two complete
molecules or fragments of monoclonal antibodies, quadroma, F(ab)2, diabodies, tandem diabodies, and
single-chain antibodies (scFv). This chapter describes the process of generating activated T cells and arming
T cells with heteroconjugated BsAbs to target cancer cells.

Key words Bispecific antibodies, Monoclonal antibodies, Immunotherapy, Armed T cells, Cancer

1 Introduction

Monoclonal antibodies (mAbs) are an important class of protein-

based drugs. Antibody-based therapy for cancer is one of the most
successful strategies for treating patients with both hematological
and nonhematological malignancies [1–7]. Since the success of
mAbs for cancer therapy, various strategies have been employed
to manipulate antibodies for enhancing their therapeutic efficacy.
One of these strategies is the bispecific antibody (BsAb) approach,
which originates from the concept of targeting or blocking two
molecular components simultaneously with a single BsAb mole-
cule. In the setting of T cells as effectors for targeting cancer cells,
one antibody is against membrane bound CD3 on T cells (anti-
CD3) and the second antibody is against tumor-associated antigen
(anti-TAA) on cancer cells [8–11]. BsAbs can be produced by
various methods ranging from simple chemistry of heteroconjuga-
tion to complex hybrid hybridomas [12] or recombinant technol-
ogy [13]. There are various formats of BsAbs, and of these
quadromas, F(ab)2, diabodies, tandem diabodies and single-chain
antibody (scFv)-based constructs are the best characterized
[14–18]. Furthermore, BsAbs can be combined with drugs, pro-
drugs, radionuclides, toxins, antivascular agents, DNA, and

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_8, © Springer Science+Business Media, LLC 2018

117
118 Archana Thakur et al.

enzymes, and can be infused alone or after coating effector cells to

redirect their cytolytic activity to TAA on malignant cells or injured
tissue. Anti-CD64
anti-HER2 BsAb (designated MDX-H210)
was used to induce remissions in patients with HER2þ advanced
breast cancer by redirecting CD64 monocytes, macrophages, and
neutrophils to HER2þ breast cancer cells [19, 20]. Activated T
cells (ATC) and anti-CD3
anti-TAA BsAb have been used to
target glioma [21, 22], neuroblastoma [23], ovarian carcinoma
[24, 25], renal cancer [26], and lymphomas [27–30]. Engagement
of the cell surface localized T cell receptor (CD3 epsilon chain) by
BsAbs can induce T cell activation, cytotoxicity, cytokine/chemo-
kine secretion, and proliferation.
In this chapter, we describe the generation of BsAbs in six steps.
In Subheading 3.1, we describe the generation and cryopreserva-
tion of activated T cells, and in Subheading 3.2 the arming of
activated T cells. Subheading 3.3 describes flow cytometry based
analyses to detect BsAb binding to effector cells, followed in Sub-
heading 3.4 methods to detect dual binding specificity of OKT3 by
anti-Her2 BsAb. In Subheadings 3.5 and 3.6, we describe in vitro
cytotoxicity assays to assess the function of BsAb armed ATC
against adherent cells and against nonadherent cells, respectively.

2 Materials

All steps should be performed in a biosafety cabinet; all media,

buffers, solutions, and plasticware (e.g., pipettes, 96-well round-
bottom plate, and 96-well flat-bottom microtiter plate) that come
in contact with the cells should be handled with proper sterile
technique. Ensure proper waste disposal regulations for tissue cul-
ture and chemical waste materials.

2.1 Tumor Cell Lines 1. Obtain SK-BR-3 (see Note 1), a Her2/neu-positive breast
adenocarcinoma cancer lines and Raji (see Note 2), and a
Her2/neu-negative Burkitt’s lymphoma (as a negative control)
from a commercial source (e.g., ATCC, Rockville, MD, USA).
2. Prepare Iscove’s modified Dulbecco’s medium (IMDM) (Life
Technologies, Grand Island, NY, USA) to culture and maintain
the SK-BR-3 cell line by adding 2 mM L-glutamine, 10% fetal
bovine serum (FBS), and 1% penicillin/streptomycin
antibiotics.
3. Prepare RPMI-1640 medium (Life Technologies, Grand
Island, NY, USA), culture and maintain the Raji cell line by
adding 2 mM L-glutamine, 10% fetal bovine serum (FBS)
(BioWhittaker, Walkersville, MD, USA), and 1% penicillin/
streptomycin antibiotics.
4. Trypsin–EDTA solution (0.05%) (Life Technologies).
 Bispecific Antibodies 119

2.2 T-Cell Activation, 1. Obtain Ficoll-Hypaque for isolation of peripheral blood mono-
Expansion nuclear cells by density gradient centrifugation.
and Cryopreservation 2. Dilute whole blood 1:1 with phosphate buffered saline (PBS),
pH 7.4.
3. Prepare Complete RPMI-1640 medium (Life Technologies)
by supplementing with 10% FBS, 2 mM L-glutamine, 1% peni-
cillin/streptomycin and store it at 4  C.
4. Prepare 100 μg/mL Muromonab-CD3 (OKT3) stock in ster-
ile water and store it at 4  C.
5. Prepare 1
106 IU/mL interleukin-2 (IL-2) stock (Chiron,
Emeryville, CA, USA) and (see Note 3) in sterile water and
store it at 4  C.
6. Obtain trypan blue to determine viability by trypan blue
exclusion.
7. Obtain cell culture grade dimethyl sulfoxide (DMSO) for
freezing.

2.3 Antibodies 1. Obtain anti-mouse IgG2a (Caltag Laboratories, Burlingame,

and Buffers for Flow CA, USA), anti-human CD3, anti-human CD4, and anti-
Cytometry to Assess human CD8 mAbs conjugated to either fluorescein isothiocya-
BsAb Binding nate (FITC) or phycoerythrin (PE) (Becton Dickinson, San
Diego, CA, USA).
2. Prepare flow cytometry wash buffer by adding 1% bovine
serum albumin (BSA) in PBS, pH 7.4.
3. Prepare flow cytometry fix buffer by adding 2% paraformalde-
hyde in PBS, pH 7.4.
4. Flow Cytometer (FACSCalibur, Becton Dickinson, San Diego,
CA, USA).

2.4 51Cr Release 1. Target cells: SK-BR-3 or Raji cells.

Assay to Access BsAb 2. Effector cells: Activated T Cells (ATCs) generated from normal
Function healthy volunteer peripheral blood mononuclear cells
(PBMC).
3. Chromium (51Cr) (PerkinElmer).
4. Complete medium: Iscove’s Modified Dulbecco’s Medium
(IMDM) supplemented with 10% heat-inactivated fetal bovine
serum (FBS).
5. Complete medium: RPMI-1640 medium supplemented with
10% heat-inactivated FCS.
6. Liquid scintillation counter like the COBRA-II gamma
counter (Packard, Downers Grove, IL).
7. Centrifuge capable of handling a microtiter plate.
120 Archana Thakur et al.

3 Methods

3.1 Generation 1. Slowly overlay (see Note 4) 1:1 PBS diluted blood from a
and Cryopreservation normal healthy donor on Ficoll (one part Ficoll and two parts
of Activated T Cells diluted blood) to isolate PBMC by Ficoll-Hypaque density
gradient centrifugation.
2. Centrifuge for 20 min (400
g) with the brake OFF.
3. Harvest PBMC (See Note 5) at the interface of the PBS and
Ficoll layer into a fresh tube.
4. Add PBS to fill the tube to wash the cells.
5. Centrifuge cell suspension 4–5 min (300–400
g) at 4  C,
discard supernatant.
6. Resuspend the cell pellet in complete RPMI-1640 and perform
a cell count (See Note 6) and viability analysis.
7. Culture PBMC in RPMI-1640 into culture flasks adjusting
concentration 1
106 cells/mL.
8. Activate T cells in the PBMC with 20 ng/mL OKT3 from
stock OKT3 solution (100 μg/mL).
9. Bring IL-2 concentration to 100 IU/mL in the final volume.
10. Expand for 14 days (see Note 7) in the presence of 100 IU/mL
of IL-2 in RPMI-1640 supplemented with 10% FCS
[7, 8]. Adjust cell concentration to 1
106 cells/mL along
with adding 100 IU/mL of IL-2 as T cells proliferate.
11. On day 14, collect ATC in 50 mL conical tubes and centrifuge
at 400
g for 10 min at room temperature to pellet the cells,
decant culture supernatant.
12. Cryopreserve ATC by control rate freezing in medium contain-
ing 10% DMSO and 90% FBS.

3.2 Arming 1. Resuspend ATC in 5 mL fresh RPMI-1640 supplemented with

of Activated T Cells 10% FBS.
2. Count ATC, and incubate 1
106 ATC (for arming) with
50 ng of BsAb (anti-CD3
anti-Her2) for 1 h at 4  C (see
Note 8, Fig. 1).
3. Wash ATC twice in complete RPMI-1640 supplemented with
10% FBS.
4. Resuspend in 100–150 μL in flow buffer to detect BsAb bind-
ing by flow cytometry.

3.3 Flow Cytometry 1. Take 1
106 activated T cells in 100–150 μL of 1% BSA in PBS
to Detect BsAb Binding buffer in two separate 5 mL polystyrene round bottom flow
to Effector Cells tube and add 10 μL of each FITC or PE-conjugated anti-CD3
and anti-CD4 antibody in one tube and 10 μL of each FITC or
 Bispecific Antibodies 121

Fig. 1 Targeted killing of cells by bispecific antibody armed T cells. This figure illustrates heteroconjugation of
anti-CD3 antibody with anti-Her2 antibody to produce bispecific antibody (BsAb) capable of binding to
activated T-cells (ATC) and targeted killing of tumor

PE-conjugated anti-CD3 and anti-CD8 antibodies in a second

flow tube.
2. Incubate FITC or PE labeled ATC at 4  C for 30 min.
3. Wash (2
) labeled cell by adding 2 mL of 1% BSA in PBS.
4. Resuspend the labeled cells in 500 μL of fix buffer (2% parafor-
maldehyde in PBS).
5. Analyze cell populations on forward scatter vs side scatter and
gate to exclude dead cells on a flow cytometer.

3.4 Detection of Dual 1. Detach SK-BR-3 (an adherent cell line) cells from the plate
Binding Specificity either by scraping or by enzymatic dissociation using trypsin.
of OKT3
anti-Her2 (a) By scraping: Remove culture supernatant, wash with 1%
Bispecific Antibody BSA in PBS then use scraper to dislodge the adherent
cells.
(b) By trypsinization: Add 1–5 mL 0.05% trypsin–EDTA
solution depending on the size of the culture flask, for
25 cm flask add 1 mL 0.05% trypsin–EDTA solution;
incubate at 37  C for 3 min followed by adding 10%
FBS in PBS to stop enzymatic reaction.
2. Collect cells in a conical tube, centrifuge at 450
g for 5 min.
3. Re-suspend pellet in 1 mL of flow buffer; count and adjust cell
concentration to 1
106 cells in 100 μL flow buffer.
122 Archana Thakur et al.

Fig. 2 Flow cytometry analysis of BsAB binding. (a) Binding of ATC armed with OKT3
anti-Her2 to ATC. A
total of 1
106 ATC were armed with 50 ng of OKT3
anti-Her2BsAb. Antibody bound to the surface of ATC
was detected with directly conjugated goat anti-mouse specific antibodies and analyzed by flow cytometry as
percent positive cells, 96% ATC show binding to BsAb. (b) A total of 1
106 SK-BR-3 were armed with 50 ng
of OKT3
anti-Her2BsAb. Antibody bound to the surface of SK-BR-3 was detected with directly conjugated
goat anti-mouse specific antibody and analyzed by flow cytometry show 92% binding of BsAb

4. Incubate 1
106 ATC or 1
106 SK-BR-3 cells in

100–150 μL PBS containing 1% BSA in 5 mL polystyrene
round bottom flow tube with 100 ng of OKT3
anti-Her2
BsAb for 1 h at 4  C.
5. Wash twice with PBS containing 1% BSA.
6. Resuspend cells in 100–150 μL PBS with 1% BSA, and add
10 μL of PE-conjugated goat anti-mouse IgG2a (see Note 9) to
detect binding of OKT3
anti-Her2 BsAb to effector T cells
or target cancer cells.
7. Incubate for 30 min in the dark at 4  C.
8. Detect binding by analyzing on flow cytometer (see Fig. 2).

3.5 In Vitro 1. Resuspend target SK-BR-3 cells at 0.4
106 cells/mL (total of
Cytotoxicity Assay 10 mL to make one 96-well plate.
to Assess the Function 2. Plate target cells in triplicate at 4
104 target cells/well in
of BsAb Armed ATC 100 μL volume in a flat-bottomed microtiter plate and incubate
Against Adherent Cells overnight at 37  C.
3. Prepare 100 μCi of 51Cr (see Note 10) containing culture
medium (20 μL of stock solution at 5 mCi/mL) and replace
the culture media in the SK-BR-3 plate after overnight incuba-
tion with 51Cr containing media and incubate the plate for
additional 4 h at 37  C.
4. Decant media in appropriate disposal container; wash cells
2
and add 100 μL complete IMDM media in each well.
 Bispecific Antibodies 123

5. Add ATC armed (see Note 11) with OKT3
anti-Her2 BsAb
as effector cells in 100 μL volume at effector to target ratios
(E/T) of 25:1, 12.5:1, 6.25:1, and 3.12:1. Incubate overnight
at 37  C. Set-up spontaneous and total release controls in each
plate.
6. Spontaneous Release: Incubate target cells alone (replace effec-
tor cells by adding 100 μL of medium in triplicate).
7. Total Release: Incubate target cells with 2% SDS (replace effec-
tor cells by 100 μL of 2% SDS).
8. After completion of incubation, centrifuge plate and collect
100 μL aliquot of the supernatant to detect radioactivity on a
COBRA-II gamma counter, output of radioactivity is displayed
in counts per minute (cpm).
9. Calculate the percentage of specific lysis using the standard
formula [(experimental  spontaneous release)/(maximum
load  spontaneous release)
100] and expressed as the
mean of triplicate samples.

3.6 In Vitro 1. For nonadherent Raji cells, label 4
106 cells with 100 μCi of
Cytotoxicity Assay
51
Cr (20 μL of stock solution at 5 mCi/mL) for 1 h at 37  C.
to Assess the Function 2. Wash cells 2
and re-suspend in RPMI-1640 medium.
of BsAb Armed ATC
3. Add 104 target cells/well of a round-bottom 96-well plate in
Against Nonadherent 100 μL RPMI-1640.
Cells
4. Add ATC armed (see Note 11) with OKT3
anti-CD20 BsAb
as an effector cells in 100 μL volume at effector to target ratios
(E/T) of 25:1, 12.5:1, 6.25:1, and 3.12:1.
5. Incubate plate for 4 h at 37  C. Set-up spontaneous and total
release controls in each plate. And follow steps 6–9 as above for
cytotoxicity against adherent targets.

4 Notes

1. After thawing, culture cells at least 3–4 passages before using

SK-BR3 or Raji cell lines. SK-BR-3 is an adherent cell line; cells
should be passed when 80–90% confluence. For adherent cul-
tures, the cells are detached using a protease, such as trypsin,
and/or a chelating agent, such as EDTA, and sub culture is
known as passaging.
2. Raji is a nonadherent cell line. For cells that grow in suspen-
sion, the culture is split into new culture vessels and should be
maintained at 1–2
106/mL concentration.
3. IL-2 stock is stored at 80  C, once thawed and diluted it can
be stored at 4  C for at least 2 weeks.
124 Archana Thakur et al.

4. It is important to overlay 1:1 PBS diluted blood gently and

carefully without letting it mix with the Ficoll (clear fluid), and
to centrifuge with the brake off.
5. During harvest of the interface (buffy coat), there should be
minimal if any mixing of the two liquids. Try to avoid Ficoll and
diluted plasma as much as possible. Wash twice with PBS to
remove any contaminating Ficoll and platelets/plasma pro-
teins. Count cells with hemocytometer and trypan blue.
6. If PBMC contain red blood cells then RBC Lysis Buffer (from
commercial source) can be used to lyse the red blood cells by
re-suspending the pellet in 3–10 mL of 1
RBC Lysis Buffer.
Incubate for 4–5 min at room temperature, and stop the reac-
tion by diluting the 1
RBC Lysis Buffer with 20–30 mL of
1
PBS. Immediately centrifuge at 400
g for 5 min at room
temperature. Re-suspend the pellet in the appropriate volume
of Flow Cytometry Staining Buffer. A small number of residual
red cells does not interfere with any assays and can be gated out
from flow cytometric analysis.
7. During T cell activation and expansion, cells are counted every
other day and cell concentration is adjusted at 1
106 cell/
mL. IL-2 is also adjusted at the concentration of 100 IU/106
cells, and culture is continued up to 14 days.
8. Antibody binding kinetics are temperature dependent. At
room temperature, cells can be incubated for 15–20 min.
While on ice, cells may require longer incubation times (1 h
at 4  C).
9. Since OKT3 is anti-mouse CD3 IgG2a antibody, detection of
bispecific antibody using FITC or PE-conjugated goat anti-
mouse IgG2a works very well.
10. The dose of chromium necessary to label a cell type can be
titrated by ensuring that the spontaneous release is less than
10% of maximum chromium release.
11. Determine the optimal arming dose of BsAb by dose titration
studies at E/T from 3.125:1 to 25:1. BsAb arming doses can
range from 0.5, 5.0, 50 to 500 ng/million ATC.

References
1. Baselga J, Tripathy D, Mendelsohn J, Multinational study of the efficacy and safety
Baughman S, Benz CC, Dantis L et al (1996) of humanized anti-HER2 monoclonal anti-
Phase II study of weekly intravenous recombi- body in women who have HER2-
nant humanized anti-p185HER2 monoclonal overexpressing metastatic breast cancer that
antibody in patients with HER2/neu- has progressed after chemotherapy for meta-
overexpressing metastatic breast cancer. J Clin static disease. J Clin Oncol 17:2639–2648
Oncol 14:737–744 3. Lundin J, Kimby E, Bjorkholm M, Broliden
2. Cobleigh MA, Vogel CL, Tripathy D, Robert PA, Celsing F, Hjalmar V et al (2002) Phase
NJ, Scholl S, Fehrenbacher L et al (1999) II trial of subcutaneous anti-CD52
 Bispecific Antibodies 125

monoclonal antibody alemtuzumab (2002) Extremely potent, rapid and

(Campath-1H) as first-line treatment for costimulation-independent cytotoxic T-cell
patients with B-cell chronic lymphocytic leuke- response against lymphoma cells catalyzed by
mia (B-CLL). Blood 100:768–773 a single-chain bispecific antibody. Int J Cancer
4. Maloney DG, Liles TM, Czerwinski DK, 100:690–697
Waldichuk C, Rosenberg J, Grillo-Lopez A, 14. Rodrigues ML, Shalaby MR, Werther W,
Levy R (1994) Phase I clinical trial using esca- Presta L, Carter P (1992) Engineering a huma-
lating single-dose infusion of chimeric anti- nized bispecific F(ab9)2 fragment for improved
CD20 monoclonal antibody (IDEC-C2B8) in binding to T cells. Int J Cancer 7:45–50
patients with recurrent B-cell lymphoma. 15. Asano R, Ikoma K, Kawaguchi H, Ishiyama Y,
Blood 84:2457–2466 Nakanishi T, Umetsu M et al (2010) Applica-
5. McLaughlin P, Grillo-Lopez AJ, Link BK, tion of the Fc fusion format to generate
Levy R, Czuczman MS, Williams ME et al tag-free bi-specific diabodies. FEBS J
(1998) Rituximab chimeric anti-CD20 mono- 277:477–487
clonal antibody therapy for relapsed indolent 16. Kufer P, Lutterbuse R, Baeuerle PA (2004) A
lymphoma: half of patients respond to a four- revival of bispecific antibodies. Trends Biotech-
dose treatment program. J Clin Oncol nol 22:238–244
16:2825–2833 17. Lazar GA, Dang W, Karki S, Vafa O, Peng JS,
6. Miller RA, Maloney DG, Warnke R, Levy R Hyun L et al (2006) Engineered antibody Fc
(1982) Treatment of B-cell lymphoma with variants with enhanced effector function. Proc
monoclonal anti-idiotype antibody. N Engl J Natl Acad Sci U S A 103:4005–4010
Med 306:517–522 18. Goldenberg DM, Rossi EA, Sharkey RM,
7. Yang JC, Haworth L, Sherry RM, Hwu P, McBride WJ, Chang CH (2008) Multifunc-
Schwartzentruber DJ, Topalian SL et al tional antibodies by the Dock-and-Lock
(2003) A randomized trial of bevacizumab, an method for improved cancer imaging and ther-
anti-vascular endothelial growth factor anti- apy by pretargeting. J Nucl Med 49:158–163
body, for metastatic renal cancer. N Engl J 19. Valone FH, Kaufman PA, Guyre PM, Lewis
Med 349:427–434 LD, Memoli V, Deo Y et al (1995) Phase
8. Riedle S, Rosel M, Zoller M (1998) In vivo Ia/Ib trial of bispecific antibody MDX-210 in
activation and expansion of T cells by a bispe- patients with advanced breast or ovarian cancer
cific antibody abolishes metastasis formation of that overexpresses the proto-oncogene
human melanoma cells in SCID mice. Int J HER-2/neu. J Clin Oncol 13:2281–2292
Cancer 75:908–918 20. Fury MG, Lipton A, Smith KM, Winston CB,
9. Renner C, Pfreundschuh M (1995) Tumor Pfister DG (2008) A phase-I trial of the epider-
therapy by immune recruitment with bispecific mal growth factor receptor directed bispecific
antibodies. Immunol Rev 145:179–209 antibody MDX-447 without and with recom-
10. Donohue JH, Ramsey PS, Kerr LA, Segal DM, binant human granulocyte-colony stimulating
McKean DJ (1990) Enhanced in vitro lysis of factor in patients with advanced solid tumors.
human ovarian carcinomas with activated Cancer Immunol Immunother 57:155–163
peripheral blood lymphocytes and bifunctional 21. Bonino LD, De Monte LB, Sapnoli GC,
immune heteroaggregates. Cancer Res Vola R, Mariani M, Barone D et al (1995)
50:6508–6514 Bispecific monoclonal antibody anti-CD3 x
11. Hombach A, Tillmann T, Jensen M, Heuser C, antitenascin: an immunotherapeutic agent for
Sircar R, Diehl V, Kruis W, Pohl C (1997) human glioma. Int J Cancer 61:509–515
Specific activation of resting T cells against 22. Zitron IM, Thakur A, Norkina O, Barger GR,
tumour cells by bispecific antibodies and Lum LG, Mittal S (2013) Targeting and killing
CD28-mediated costimulation is accompanied of glioblastoma with activated T cells armed
by Th1 differentiation and recruitment of with bispecific antibodies. BMC Cancer 13
MHC-independent cytotoxicity. Clin Exp (83):2013
Immunol 108:352–357 23. Yankelevich M, Kondadasula SV, Thakur A,
12. Lindhofer H, Mocikat R, Steipe B, Thierfelder Buck S, Cheung NK, Lum LG (2012) Anti-
S (1995) Preferential species-restricted heavy/ CD3
anti-GD2 bispecific antibody redirects
light chain pairing in rat/mouse quadromas. T-cell cytolytic activity to neuroblastoma tar-
Implications for a single-step purification of gets. Pediatr Blood Cancer 59:1198–1205
bispecific antibodies. J Immunol 155:219–225 24. Canevari S, Stoter G, Arienti F, Bolis G, Col-
13. Dreier T, Lorenczewski G, Brandl C, naghi MI, Di Re EM et al (1995) Regression of
Hoffmann P, Syring U, Hanakam F et al advanced ovarian carcinoma by intraperitoneal
126 Archana Thakur et al.

treatment with autologous T lymphocytes T cells and low dose IL-2, after stem cell trans-
retargeted by a bispecific monoclonal antibody. plant for high risk non-Hodgkin’s lymphoma: a
J Natl Cancer Inst 87:1463–1469 pilot study. Bone Marrow Transplant 49
25. Bolhuis RLH, Lamers CHJ, Goey SH, Egger- (73–9):2014
mont AMM, Trimbos JBMZ, Stoter G et al 28. Lum LG, Thakur A, Liu Q, Deol A,
(1992) Adoptive immunotherapy of ovarian Al-Kadhimi Z, Ayash L et al (2013) CD20-
carcinoma with BSMAb-targeted lymphocytes: targeted T cells after stem cell transplantation
a multicenter study. Int J Cancer 7:78–81 for non-Hodgkin lymphoma. Biol Blood Mar-
26. Renner C, Bauer S, Sahin U, Jung W, van row Transplant 19:925–933
Lier R, Jacobs G, Held G, Pfreundschuh M 29. Baeuerle PA, Kufer P, Lutterbuse R (2003)
(1996) Cure of disseminated xenografted Bispecific antibodies for polyclonal T-cell
human Hodgkin’s tumors by bispecific mono- engagement. Curr Opin Mol Ther 5:413–419
clonal antibodies and human T cells: the role of 30. Loffler A, Kufer P, Lutterbuse R, Zettl F, Dan-
human T-cell subsets in a preclinical model. iel PT, Schwenkenbecher JM et al (2000) A
Blood 87:2930–2937 recombinant bispecific single-chain antibody,
27. Lum LG, Thakur A, Pray C, Kouttab N, CD19 x CD3, induces rapid and high
Abedi M, Deol A, Colaiace WM, Rathore R lymphoma-directed cytotoxicity by unstimu-
(2014) Multiple infusions of CD20-targeted lated T lymphocytes. Blood 95:2098–2103
 Chapter 9

Immunophenotyping of Live Human Pluripotent Stem Cells

by Flow Cytometry
Daniel R. Riordon and Kenneth R. Boheler

Abstract
Human pluripotent stem cells (hPSCs) have great potential for use in regenerative medicine and cell
replacement therapies; however, prior to clinical application, cultured cell populations need to be screened
to ensure the quality of the culture, as well as the capacity of these pluripotent cells to differentiate into
desired cell types. Flow cytometry, utilizing antibodies recognizing targets restricted to the hPSC surfa-
ceome, offers an invaluable tool for high-throughput validation of hPSC lines. Here we describe the
immunophenotyping of live human embryonic stem cell (hESC, H9) and human induced pluripotent
stem cell (hiPSC, KB3) lines by flow cytometry using a panel of antibodies identified as either stem cell
reference markers (CD90, EpCam) or reported as being prevalent or restricted (c-Kit, HPI-1, Integrin α6,
Semaphorin-6A) to these cells. The protocols described here with hPSCs are also applicable to differen-
tiated hPSC progeny and should be instrumental in the immunophenotyping and isolation of well-defined
homogeneous cell populations useful in regenerative medicine.

Key words Embryonic stem cells, Induced pluripotent stem cells, Surfaceome, Immunophenotyping,
Flow cytometry, Antibody

1 Introduction

Embryonic stem (ES) cells are characterized by their pluripotency

and self-renewing properties. It is because of these properties that
the establishment of the first human embryonic stem cell (hESC)
lines in 1998 [1] garnered much enthusiasm for their significant
potential use in research and regenerative medicine. Clinical appli-
cation of these cells, however, is limited by concerns surrounding
the immunogenicity of the donor cells, and while the more recent
development of hESCs derived by somatic cell nuclear transfer
(ntES cells) overcomes the potential for immune rejection [2],
there remains both the ethical controversy surrounding the use of
human embryos as well as the issue regarding the limited supply of
donor embryos.

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_9, © Springer Science+Business Media, LLC 2018

127
128 Daniel R. Riordon and Kenneth R. Boheler

Human induced pluripotent stem cells (hiPSC), which are

derived from somatic cells through the ectopic expression of a spe-
cific set of transcription factors [3–5] and share the ESC properties of
self-renewal and pluripotency, have overcome many of the limita-
tions associated with hESC. Despite the advantages, there remain
many obstacles to overcome prior to using hiPSC in the treatment of
disease. Variability in the differentiation potential of hiPSC com-
pared to hESC has been reported [6], and could be related to
“epigenetic memory” from the somatic cell of origin [7] or the result
of residual transgene expression [8]. Furthermore, recent compar-
isons of hESC and hiPSC transcriptional profiles at the single cell
level demonstrate greater heterogeneity in gene expression levels in
hiPSC, which may ultimately underlie the differentiation capabilities
of these cells [9]. These factors likely contribute to the heterogeneity
of differentiating human pluripotent stem cells (hPSCs) or the pres-
ence of contaminating undifferentiated cells found in many hPSC
cultures. Therefore, before any hPSC line can be deemed acceptable
for use in clinical settings, it is critical that cells be evaluated for their
ability to differentiate into desired cells types, as well as assessing the
overall quality and purity of differentiated cultures. Several methods
exist for identification and characterization of PSCs, including
single-nucleotide polymorphism (SNP) analysis, epigenetic profiling,
immunocytochemistry, PCR, western blotting, in vitro differentia-
tion, and teratoma formation; however, the majority of these meth-
ods require either the destruction or alteration (fixation) of the cell,
or take days to complete, precluding the ability to use identified
populations in downstream applications.
Immunophenotyping of live cells by flow cytometry, utilizing
targets found on the cell surface, provides a high-throughput,
nonmutagenic, and reproducible method for validation of hESC
and hiPSC cultures, and eventually for the removal of potential
contaminating hPSCs that would potentially lead to tumor forma-
tion after transplantation. Current state-of-the-art cytometric eva-
luations allow one to define distinct cell population using two
physical parameters. The first parameter includes both forward
scattered (FSC) and side scatter light (SSC), while the second
parameter involves fluorescence, usually following some sort of
cell labeling through the use of a dye or use of a fluorescently
labeled antibody. The number of fluorescence parameters can
range from 0 to >12 depending on the equipment available, with
the latter requiring high-end multicolor flow cytometry. Each fluo-
rescence parameter can be used to independently measure a func-
tion or the presence of a protein. When these parameters are
combined, the measurements can be used to immunophenotype
and sort selected cell populations with known functions.
Although a number of kits are available to characterize undif-
ferentiated hPSC, most rely on antibodies against SSEA-1 (mouse),
SSEA-3 (rat), and TRA-1-81 (human). SSEA-1 is absent from
 Flow Cytometry and Immunophenotyping 129

hPSCs, while SSEA-3 (or SSEA-4) is not wholly specific to hPSC.

TRA-1-81 (or TRA-1-60) requires fixation, which limits their use
for sorting and selecting authentic hiPSCs for cultivation and
expansion. SSEA-5 has also been reported as useful for removal of
teratoma-forming cells as part of a surface antibody panel [10]. A
recent publication of a human pluripotent stem cell surface prote-
ome (surfaceome), which compared hESCs and hiPSC against
human fibroblasts and 50 additional cell types, identified >30
positive and negative markers for hPSCs and found an additional
>100 proteins of interest for hPSCs [11]. This resource allows for
the selection of a panel of markers for the identification of newly
derived hPSC populations with greater purity, potency or with
enhanced lineage specific differentiation potential.
The protocols described here demonstrate the immunopheno-
typing of KB3 hiPS cells cultured in feeder free, defined media
conditions, in comparison to an established hESC line (H9) using
live cell antibody labeling and analysis by flow cytometry. Cells were
probed for the reference stem cell markers CD90 and EpCam
(CD326), as well as for stem cell prevalent/restricted markers
c-Kit (CD117) (hematopoietic), HPI-1 (neural progenitor cell)
[12], Semaphorin-6A [11] and Integrin α6 (CD49f) (hematopoie-
tic [13] and mesenchymal stem cells [14] multipotency markers).
Although polychromatic flow cytometry is not described, the anti-
bodies used, if conjugated with appropriate fluorophores, can be
employed with multicolor flow cytometric technologies. The
approaches described here are valid for hPSCs, but these protocols
can be adapted for the analysis of any mammalian cell line or any
differentiated hPSC progeny, provided that informative epitopes
are known and antibodies are available that are suitable for the
species and cell type being analyzed.

2 Materials

All solutions should be prepared using cell culture grade reagents

and sterile supplies. All procedures should be performed using
aseptic techniques in a Biological Safety Cabinet/Tissue Culture
Hood. If analyses of cells without sorting are performed, then cell
preparation after cultivation and flow cytometry can be performed
using nonsterile techniques.

2.1 Coating Cell 1. Microcentrifuge tubes, 1.5 mL, sterile.

Culture Plates 2. DMEM/F-12 (with L-Glutamine and HEPES) culture
with hESC Qualified medium (ThermoFisher Scientific, Waltham, MA, USA)
Matrix chilled to 4
C.
3. hESC qualified matrix coating stock solution: Corning®
Matrigel® hESC-qualified Matrix (catalog number 354277)
130 Daniel R. Riordon and Kenneth R. Boheler

(Corning Incorporated, Corning, NY, USA) (see Note 1).

Thaw vial of hESC-qualified matrix overnight in 4
C refriger-
ator on ice. Swirl vial to ensure material is evenly dispersed and
dispense into single use aliquots (see Note 2) in chilled 1.5 mL
microcentrifuge tubes on ice, switching pipette tips frequently
to prevent clogging. Store the aliquots at 20 or 80
C.
Alternatively use Geltrex® LDEV-Free hESC-qualified matrix
(Catalog number A1413202; ThermoFisher Scientific, Wal-
tham, MA, USA). Allow a bottle of growth factor-reduced
Geltrex® to thaw at 4
C overnight. Aliquot (see Note 2) and
store at 20
C. Thaw at 4
C, thawed vials may be stored at
4
C until needed.
4. Culture dishes: 6-well, 100 mm, sterile (Corning).
5. Conical centrifuge tubes: 50 mL, sterile (Corning, Falcon
Brand).
6. Serological Pipettes: 5 and 10 mL.
7. P10, P20, P200, and P1000 Pipettes (e.g., Gilman).
8. Pipette tips.

2.2 Passaging 1. H9 (WA09) (WiCell, Madison, WI, USA) hESCs cultured in

and Maintenance 6-well plates. Any established hESC line can be used.
of Undifferentiated 2. KB-3 hiPSCs cultured in 6-well plates. Any established hiPSC
hPSCs in Monolayer line can be used.
Culture 3. hESC qualified matrix coated 100 mm culture dishes.
4. Conical centrifuge tubes: 15 mL.
5. FACS (Fluorescence-activated cell sorting) tubes or equivalent:
Polystyrene round bottom tube (5 mL) with 35 μm nylon
mesh cell strainer cap (catalog number 352235) (Corning
Incorporated, Corning, NY, USA).
6. StemPro® Accutase® Cell Dissociation Solution (Catalog
number A1110501) (ThermoFisher Scientific, Waltham, MA,
USA). Thaw Accutase® at 4
C overnight. Use Accutase®
within 2 months, if stored at 4
C. Otherwise, aliquot 10 mL
into 15 mL conical centrifuge tubes and store at 20
C (see
Note 3).
7. Trypan Blue Stain (0.4%) (catalog number T8154) (Sigma-
Aldrich Corp, St. Louis, MO, USA).
8. Ultrapure water.
9. Dulbecco’s PBS (DPBS) without Calcium and Magnesium,
pH 7.4 (catalog number 14190250) (ThermoFisher Scientific,
Waltham, MA, USA).
10. Borosilicate glass disposable 900 pipette.
 Flow Cytometry and Immunophenotyping 131

11. Cotton-plugged borosilicate glass disposable 900 pipette

with bulb.
12. Rho kinase (ROCK) inhibitor Y-27632 stock solution
(10 mM) (STEMCELL Technologies, Vancouver, Canada):
In 1.5 mL microcentrifuge tube, dissolve 1 mg ROCK inhibi-
tor in 312.5 μL DPBS. Store in 250 μL aliquots at 20
C.
13. Essential 8 Medium (E8™ Medium) (catalog number
A1517001) (ThermoFisher Scientific, Waltham, MA, USA)
(see Note 4).
14. E8™ Medium with 1 ROCK Inhibitor: Add 1.25 μL ROCK
Inhibitor stock solution per mL of E8™ Media (see Note 5).
Mix well.
15. Vacuum Aspiration System.
16. Automated cell counter or manual hemocytometer.
17. Centrifuge for 15 mL conical tubes and 5 mL FACS tubes with
cooling (4
C) capability.
18. 5% CO2, 37
C, humidified incubator.

2.3 Harvesting 1. 100 mm plate of hESC/hiPSC grown to 70–80% confluency.

hPSCs for Antibody 2. Phosphate Buffered Saline (PBS), pH 7.4, without calcium,
Labeling without magnesium.
3. Fetal Bovine Serum (FBS) (Atlanta Biologicals, Flowery
Branch, GA, USA).
4. Cell Wash Solution: 1 PBS containing 1% FBS Cell Wash
Solution. Add 5 mL FBS to 500 mL PBS. Store solution at
4
C for up to 2 weeks.
5. Enzyme Free Cell Dissociation Solution (catalog number
S-014-B; EMD Millipore, Billerica, MA, USA).
6. Falcon® Cell Strainers: 40 μm nylon mesh cell strainer (catalog
number 352340; Corning Incorporated, Corning, NY, USA).
7. Conical Centrifuge Tubes: 50 mL, sterile (Corning, Fisher
Brand)

2.4 Titration 1. Stock Solution of antibody to be titrated (fluorochrome-

of Antibodies conjugated or unconjugated) (see Note 6).
2. Stock Solution of fluorochrome-conjugated Secondary
Antibody.
3. Isotype Control antibodies (see Note 7).
4. Target Cells for antibody titration (H9 (WA09) hESCs or KB-3
hiPSCs).
5. Cell Staining Solution for Flow Cytometry: 1 PBS containing
5% FBS. Add 5 mL FBS to 500 mL PBS. Store at 4
C for up to
2 weeks.
132 Daniel R. Riordon and Kenneth R. Boheler

Table 1
List of antibodies (primary and secondary) used in the current protocols

Amount of Amount of
primary secondary
antibody per antibody per
Antigen Isotype 1  106 cells Secondary 1  106 cells
(clone) Fluorochrome Company control in 100 μL anti body in 100 μL
CD49f PE eBioScience Rat 0.13 μg N/A N/A
(GoH3) IgG2a
CD90 PE BD Mouse 4.0 μg N/A N/A
Pharmingen IgG1
CD117 PE Bio Legend Mouse 0.25 μg N/A N/A
(104D2) IgG1
CD326 Alexa Fluor Bio Legend Mouse 0.25 μg N/A N/A
(9C4) 647 IgG2b
HAI-1 PE eBioScience Mouse 0.25 μg N/A N/A
(9B10) IgG2a
SEMA6A R&D Systems Goat 0.5 μg Donkey 0.125 μg
IgG anti-goat
IgG-PE

6. Round-Bottom polystyrene tubes: 5 mL.

2.5 Antibody 1. Antibodies. See Table 1 for antibodies used in the current
Labeling of Cell protocol.
Surface Antigens 2. Isotype Controls. See Table 1 for Isotype controls used in the
current protocol.
3. Borosilicate glass disposable 900 pipette.
4. Phosphate Buffered Saline (PBS), pH 7.4, without calcium,
without magnesium.
5. Cell Staining Solution for Flow Cytometry (see item 5 of
Subheading 2.4).
6. Round-Bottom polystyrene tube: 5 mL.
7. Serological Pipettes: 5 and 10 mL.
8. Human Trustain FcX™ Fc Blocking Solution (Catalog number
422301; BioLegend, San Diego, CA, USA).
9. Platform Rocker.
10. 10% normal goat serum in 1 PBS Secondary Antibody Block-
ing Solution (Catalog number 50062Z; ThermoFisher Scien-
tific, Waltham, MA, USA).
 Flow Cytometry and Immunophenotyping 133

2.6 Preparation 1. Hanks Balanced Salt Solution (without calcium or magnesium)

of Cells for Flow (HBSS).
Cytometry 2. Cell Maintenance Solution for Flow Cytometry: Add 5 mL
FBS to 95 mL HBSS. Store at 4
C.
3. FACS Tubes (see 2.2 above).
4. FBS.

2.7 Flow Cytometry 1. The current protocol used a FACSCanto II flow cytometer
Analysis (BD Biosciences, Franklin Lakes, NJ, USA).
2. Single peak Rainbow beads (catalog number RFP-30-5A;
Spherotech Inc., Green Oaks, IL, USA).
3. Compensation beads (catalog number 557640; Beckton Dick-
inson Immunocytometry Systems).
4. 0.5% paraformaldehyde (PFA).

3 Methods

The protocols described here are applicable for both flow cytome-
try analysis and sorting of hPSC (see Note 8). All procedures should
be performed using aseptic techniques in a Biological Safety Cabi-
net/Tissue Culture Hood unless otherwise noted.

3.1 Coating Cell 1. Slowly thaw 1 aliquot of hESC qualified matrix (Matrigel®) at
Culture Plates 4
C for 30 min.
with hESC Qualified 2. Transfer 25 mL of cold DMEM/F12 medium to a 50 mL
Matrix centrifuge tube on ice.
3. Add an aliquot of hESC qualified matrix to 25 mL of chilled
DMEM/F12 medium and mix well immediately before coat-
ing plates (see Note 9).
4. Add 2 mL hESC-qualified Matrix in culture medium per
9.5 cm2 culture well (one well of a 6-well plate) growth areas.
5. Swirl plate to ensure even coating, and store at 4
C for Matri-
gel or at 37
C for Geltrex. Before using, allow hESC-qualified
Matrix coated plates to set, covered, for 30 min at room
temperature (see Notes 10 and 11).
6. Immediately prior to plating cells, aspirate the excess hESC
qualified matrix (see Note 12).

3.2 Passaging 1. Add 1 mL or 7.5 mL prewarmed E8™ medium containing

Undifferentiated 2 ROCK inhibitor to the selected number of wells of a 6-well
hPSCs/hiPSC plate or a 100 mm plate that has been coated with hESC
in Monolayer Culture qualified matrix (see Note 13).
134 Daniel R. Riordon and Kenneth R. Boheler

2. Thaw sufficient Accutase® (see Note 14) to passage cells and

warm to room temperature (see Note 15).
3. Observe cells under a brightfield microscope. Identify and if
necessary remove regions of differentiation by scraping with a
pipette tip or by aspiration, using a new tip for each plate to
reduce the possibility of cross-contamination (see Note 16).
4. Aspirate medium from culture plates using sterile Pasteur
pipette and vacuum aspiration system.
5. For a single well or a 100 mm culture dish, wash cells twice with
2 or 10 mL of 1 DPBS (room temperature), respectively.
6. Add 1 or 4 mL of Accutase® to each well or 100 mm culture
dish, respectively and leave undisturbed for 3–7 min until
colony boundaries appear folded back and show signs of
becoming less well packed.
7. Using a pipetteman (P1000) or a cotton-plugged glass pipette
with bulb, dislodge cells with very gentle pipetting and transfer
to a 15 or 50 mL conical tube containing DMEM/F-12
medium, prewarmed to room temperature (see Note 17).
Gently triturate suspension to generate single cells.
8. Transfer a 10 μL aliquot of cell solution to a 1.5 mL micro-
centrifuge tube and mix with 10 μL trypan blue. Count cells
(see Note 18). Use total cell counts for all applications going
forward.
9. Collect remainder of cells by centrifugation at 130  g for
5 min at room temperature (see Note 19).
10. Aspirate media and resuspend the cell pellet in DMEM/F12 or
E8™ medium, prewarmed to 37
C, to an optimal concentra-
tion of 0.1–1.0  106 cells per mL (see Note 20).
11. Add up to 1.0 or 7.5 mL of the cell suspension to each well or
100 mm tissue culture plate containing E8™ medium plus 2x
ROCK inhibitor prepared in step 1. Optimally, the volume of
the cell suspension being added is less than the volume in the
previously prepared plate; therefore, the final volume should be
brought to 2.0 mL per well, or 15 mL per 100 mm dish, such
that the ROCK inhibitor is at a final concentration of 1.
Before placing the cells in the incubator, gently move in a
front-to-back and side-to-side motion to uniformly disperse
cells across the well (see Note 21).
12. Incubate cells at 5% CO2, 37
C in humidified incubator, and
replace the ROCK inhibitor containing medium within
18–24 h of plating with E8™ culture medium without
ROCK inhibitor. The volume to add is 2 or 15 mL per well
or 100 mm culture dish, respectively.
 Flow Cytometry and Immunophenotyping 135

13. Replace media daily with E8™ culture medium without

ROCK inhibitor.
14. Harvest cells when plates reach 70–80% confluency.

3.3 Harvesting 1. For antibody labeling, we use primarily 100 mm cultures of

hESCs and hiPSCs hPSCs; however, for simple tests, there are usually enough cells
for Antibody Labeling (1–3  106) in a single well of a 6-well plate.
2. Examine cells for morphological signs of differentiation using a
brightfield microscope and, as needed, remove differentiated
cells as described in step 3 of Subheading 3.2.
3. Aspirate the growth medium from the culture plates. Wash cells
twice with 2 mL cold Cell Wash Solution
4. Aspirate Cell Wash Solution then add cold Millipore “cell
dissociation solution” to cover cells (3–4 mL/ 100 mm plate)
(see Note 22).
5. Incubate dish at 4
C for 20 min, on a rocker to maximize cell
dissociation (see Note 23).
6. Using a cotton-plugged borosilicate glass disposable 900 pipette
with bulb, dislodge cells with gentle pipetting and transfer to a
15 mL conical tube on ice.
7. Collect cells by centrifugation at 200  g for 5 min at 4
C.
8. Aspirate the supernatant.
9. Resuspend cells in 10.1 mL Cell Wash solution using a 10 mL
serological pipette with repeated gentle trituration to break up
cell clumps and ensure a single cell suspension.
10. Using a 10 mL serological pipette, pass cells through a 40 μm
nylon mesh cell strainer fitted to the top of a 50 mL conical
tube (see Note 24).
11. Transfer a 100 μL aliquot of cell solution to a 1.5 mL micro-
centrifuge tube and mix with 100 μL trypan blue. Count cells
as described in step 6 of Subheading 3.2. Use total cell counts
for all applications going forward.
12. Collect cells by centrifugation at 200  g for 5 min at 4
C.

3.4 Titration All steps should be performed on ice and samples protected from
of Antibodies light.
for Percent Positive For Fluorochrome-conjugated Primary Antibodies:
Measurements 1. Determine the concentration and volume of the antibody stock
solutions and recommended antibody concentration for use in
flow cytometry analysis from the manufacturer’s product data
sheet (see Note 25).
2. For each antibody and isotype control, number 6–8 microcen-
trifuge tubes and place on ice.
136 Daniel R. Riordon and Kenneth R. Boheler

Table 2
Antibody serial dilution scheme

Volume and
Cell wash source of Working Working stock Ab staining Concentration in
Tube solution primary Ab stock concentration reaction volume staining reaction
no. volume (μL) (μL) dilution (μg/mL) (μL/106 cells) (μg/106 cells)
1 0 30 of stock 0 200 20 4.0
2 20 20 of tube 2 100 20 2.0
1 dilution
3 20 20 of tube 4 50 20 1.0
2 dilution
4 20 20 of tube 8 25 20 0.5
3 dilution
5 20 20 of tube 16 12.5 20 0.25
4 dilution
6 20 20 of tube 32 6.25 20 0.125
5 dilution
7 20 20 of tube 64 3.125 20 0.0625
6 dilution
8 20 20 of tube 128 1.5625 20 0.03125
7 dilution
Example of a serial dilution scheme for a PE conjugated CD90 antibody with a master stock concentration of 0.2 mg/mL
for staining 1  105 cells

3. Using the manufacturer’s recommended antibody concentra-

tion as a guide, serial dilute antibody to create several stock
antibody concentrations. Begin dilutions at slightly above the
recommended concentration. Table 2 provides an example of a
twofold serial dilution scheme for a PE conjugated CD90
antibody with a master stock antibody concentration of
0.2 mg/mL (see Note 26).
4. Prepare the first working stock antibody solution by pipetting
30 μL of master stock antibody into a labeled microcentrifuge
tube (Tube No. 1) on ice.
5. Perform 6–8 twofold serial dilutions from the highest concen-
tration of the working stock antibody. Pipette 20 μL from the
working stock into microfuge tube containing 20 μL Cell
Staining Solution. Gently vortex the tube followed by a quick
spin in a microfuge. Repeat for subsequent dilutions until series
is complete.
6. Label cells with antibody and prepare for flow cytometry anal-
ysis as described in Subheadings 3.5 and 3.6. Include tubes for
 Flow Cytometry and Immunophenotyping 137

each antibody to be titrated as well as for unstained and isotype

controls (see Note 27).
7. Refer to Subheading 3.8 for determination of the optimal
antibody concentration.

3.5 Antibody All steps should be performed on ice and samples protected from
Labeling of Cell light.
Surface Antigens
1. Resuspend cells prepared in Subheading 3.3 in cold Wash
on Live Cells Solution so that final concentration of cells is 1  106 total
cells in 95 μL. Using a P200 pipette, gently triturate to disag-
gregate cells.
2. Block cells by adding 5 μL Human Trustain FcX™ Fc Blocking
Solution for every 1  106 cells and mix by gently flicking tube
with finger, then incubate for 10 min on ice, gently rocking.
3. Gently triturate cell solution using P200 pipette to ensure
homogeneous mixture of cells then aliquot 1  106 cells
(100 μL) per tube into 5 mL round bottom polystyrene tubes
on ice (see Note 28).
4. Add primary antibody or isotype control at its optimal concen-
tration (as determined in Subheadings 3.4 and 3.8) then incu-
bate for 60 min on ice with gentle rocking (see Note 29).
Include an unstained control.
5. Add 3 mL cold Wash Buffer, then collect cells by centrifugation
at 200  g for 5 min at 4
C.
6. Aspirate solution being careful not to disturb cell pellet.
7. Repeat washing steps 5 and 6 for a total of two washes follow-
ing antibody labeling.
8. If a primary antibody directly conjugated to a fluorochrome is
used, proceed directly to Subheading 3.6. Continue as follows
for labeling with a secondary antibody conjugated to a
fluorochrome.
9. Resuspend cells in 100 μL secondary antibody blocking solu-
tion using a P200 pipette with gentle trituration.
10. Add secondary antibody, gently tap tube to mix, and then
incubate for 30 min on ice, gently rocking.
11. Add 3 mL cold Wash Buffer, then collect cells by centrifugation
at 200  g for 5 min at 4
C.
12. Aspirate solution being careful not to disturb cell pellet
13. Repeat washing steps 11 and 12 for a total of two washes after
secondary antibody labeling.

All steps should be performed on ice and samples protected from

light.
138 Daniel R. Riordon and Kenneth R. Boheler

3.6 Preparation 1. Resuspend cells prepared in 400 μL cold Cell Maintenance

of Cells for Flow Solution. Using a P1000 pipette, gently triturate to disaggre-
Cytometry gate cells.
2. Prewet the 35 μm nylon mesh cell-strainer cap on 5 mL FACS
tube with 50 μL cell maintenance solution (see Note 30). Keep
tubes on ice.
3. Transfer cell solution to cell strainer cap and allow to pass across
the mesh and drop to the bottom of tube by gravity (see Note 31).
4. Rinse strainer with 250 μL cell maintenance solution.
5. Keep cells on ice and protected from light until analyzed by
flow cytometry.

3.7 Flow Cytometry 1. Prior to running samples, alignment beads and calibration
Analysis standards should be run to ensure that the data are reproduc-
ible from day to day. This may be routinely run by an experi-
enced user or in a core facility (if so, go to step 16 below), but if
not, then beads should be run to ensure good instrument
performance.
2. Place diluted single peak Rainbow beads onto the sample inser-
tion tube and initiate data acquisition.
3. While observing FSC versus SSC and all other fluorescence
parameter combinations, adjust the instrument according to
the manufacturer’s instructions to ensure the narrowest CV
with the highest signal intensity. Tolerance ranges are estab-
lished for the coefficient of variance and fluorescent intensity as
well as all instrument parameters.
4. Adjust the voltages for each photomultiplier tube (PMT) to
achieve the predetermined intensity levels for the bead
population.
5. Collect and save all single parameter histograms for subsequent
analyses.
6. The signal-to-background ratio (S/B) is obtained by using the
eight peak beads (i.e., the median channel of each peak) divided
by the median channel. Each PMT will have a characteristic
S/B, and the correct tolerance plotted against time should be
in the range of 10%.
7. If multiple fluorophores are run, then compensation beads
(mouse anti-K beads) will also need to be run as described in
https://www.bdbiosciences.com/documents/BD_
FACSDiva_setup_system.pdf.
8. Latex beads coated with anti-mouse K antibody are used with
each antibody conjugate to determine the compensation
matrix for polychromatic flow cytometry. Into a conical tube
add 40 μL of compensation beads and the volume of previously
 Flow Cytometry and Immunophenotyping 139

titered antibody conjugate (this is done for each antibody).

Dilute to 100 μL with cell wash solution.
9. Incubate at room temperature for 15 min.
10. Wash once with cell wash solution.
11. Remove the supernatant and resuspend in 250 μL of cell wash
solution.
12. Vortex and add 150 μL of 0.5% PFA.
13. Acquire each compensation control tube and the unstained
bead control using the previously defined voltage settings.
14. Set the automated compensation matrix. Compensation
should be rechecked regulatory by acquiring cell samples
stained with combinations of antibody conjugates.
15. Once established, it is inappropriate to alter these voltages
during data acquisition among samples.
16. For experimental data acquisition, gently vortex or triturate
each biological sample immediately prior to placement onto
sample insertion tube to ensure dispersion of cells and the
absence of aggregates.
17. For each cell type, load the unstained control or preferably the
isotype control sample onto the flow cytometer and optimize
forward and side scatter voltage settings (see Note 32). For
statistical analysis, we use forward scatter (FSC) (abscissa) and
side scatter (SSC) (ordinate) to gate viable, single cell events

Fig. 1 Gating strategy for selective live cells. Dot blots showing light scatter profiles of hiPSC and hESC gated for
the selection of live cells and exclusion of debris. In this instance, cells have not been stained for viability, and all
cells are included, excluding those with a very low SSC value. These likely correspond to dead or damaged cells.
50,000 events were collected and the gated population used for determination of cells positive for marker proteins
140 Daniel R. Riordon and Kenneth R. Boheler

and eliminate debris and cell aggregates (see Fig. 1) (see Note
33). If there are multiple isotype controls we begin with
unstained controls and then optimize settings to maximize
data acquisition of all isotype controls tested.
18. Ideally, the scatter plots should show an equal distribution of
cells distributed along a 45
angle relative to the ordinate and
abscissa.
19. Visualize the cell distribution of isotype controls and
fluorochrome-labeled cells to determine threshold values to
limit the number of events acquired by the flow cytometer. If
using total counts, then the threshold value is low and most
events will be acquired; however, population gating may be
useful to eliminate events such as cell debris and dead cells (see
Note 34).
20. Adjust the voltages to ensure that the signals with the isotype
control can be readily visualized. Ideally, the peak signal will be
well defined when viewed as a histogram, and it will have an
even distribution, and a signal on the Fluorochrome log scale
that is minimal.
21. Using fluorochrome-labeled cells, check to see if there is a
signal above that seen with the isotype control. Normally, we
test each antibody singly on a specific cell type before attempt-
ing any multilabeling experiments. Establish optimum baseline
PMT gains (see Note 35) to maximize resolution. Use the
minimum intensity required to achieve a histogram of fluores-
cence, which clearly displays both left and right edges of the
peak(s).
22. Record settings, as this will be valuable for future repeat experi-
ments. We recommend multiple repeats with an n > 4 for each
cell surface protein examined.
23. Once these thresholds have been determined, maintain the
laser voltage settings of each fluorochrome when analyzing
each corresponding antibody labeled sample. It is inappropri-
ate to alter these voltages during data acquisition among sam-
ples, either when determining optimal antibody dilutions or
when performing analyses.
24. Collect a minimum of 10,000 events; however, a higher acqui-
sition may be needed for multicolor analyses.
25. Acquire data using flow cytometry for each antibody tested;
however, it may be necessary to adjust laser voltage settings for
each fluorochrome using the appropriate isotype controls.
26. If multicolor parameters are assessed, then the voltage settings
should be set to maximize data acquisition.
27. If displaying multiple parameters with multiple fluorochromes,
then 2D, 3D, and other plots may be necessary for analyzing
 Flow Cytometry and Immunophenotyping 141

Fig. 2 Titration of CD90-PE antibody using H9 hESC. The data shown here reflect titration results from pilot
tests with three independent antibodies. Data from the antibody showing the best results are shown, but at
three concentrations. Histograms of CD90 in H9 cells illustrating that the optimal signal to noise ratio (S/N) is
achieved using 4.0 μg antibody. MFI median fluorescence intensity, S/N MFI positive/MFI negative

data. Quenching must be considered, and settings must be

based on the absorption spectra of fluorochromes (see Note 36).
3.8 For Antibody 1. Display histogram for each dilution and negative controls (see
Titration Fig. 2).
Determinations 2. Determine the total counts or median fluorescent intensity
(MFI) of both positive (Signal) and negative (Noise) for all
samples.
3. Calculate the signal to noise ratio by dividing the MFI value for
positive cells by that for the negative cells.
4. In the final evaluations, choose an antibody for all subsequent
analyses at the concentration that gives the highest Signal to
Noise ratio for the best discrimination between positive and
negative cells with the least amount of added antibody.
5. Alternatively, for quantitation purposes, where saturation of
the target protein is necessary to achieve accurate measure-
ments, the antibody should be used at the concentration
where antibody saturation is achieved without significantly
increasing nonspecific binding, as indicated by shifts in fluores-
cence in isotypes and negative controls (see Note 37).

3.9 Analysis of Data 1. Data will need to be exported and analyzed using software.
There are a number of software packages that can be used, and
we recommend that you discuss with your Flow Cytometry
Core, which package may be apt. Examples include FCS
Express, FloJo, and free software from Flowing Software.
142 Daniel R. Riordon and Kenneth R. Boheler

2. For data analyses in the histogram mode, adjust the gates so

that less than 2% of the signals/events from the isotype controls
are above the negative control gate. All signals falling below
this setting/gate will be considered as negatives, while those
signals located above this value will be considered positive.
When viewed as a histogram (ordinate—counts (total, FSC,
SSC); abscissa—Fluorochrome Log scale (channel)), only sig-
nals with intensities equal to or greater than the threshold
channel value will be processed.
3. Determine the percent positive cells within the gated popula-
tion by generating a univariate histogram and using the isotype
control to gate the negative population to include >98% of the
events, with staining greater than this in the antibody stained
sample considered positive (see Fig. 3a, b).

4 Notes

1. Although Matrigel® in conjunction with E8™ medium has

been the most widely used culture system for hiPSC mainte-
nance, the optimal coating matrix should be tested whenever
developing or characterizing new hiPS cell lines. Alternative
coating matrices, such as Geltrex®, Vitronectin XF™, or BD™
Laminin/Entactin Complex High Concentration are available
if a fully chemical defined system is required.
2. The dilution factor for creating the correct concentration of
hESC-qualified matrix for plate coating is calculated for each
lot of Matrigel® or Geltrex® LDEV-Free hESC-qualified
matrix based on the product protein concentration, and there-
fore, the user should refer to the manufacturer’s lot specific
Certificate of Analysis for product protein concentrations and
follow the supplier’s instructions for generating appropriate
dilutions and single use aliquots.
3. Dispase can also be used (though in our hands, it has proven
less good than accutase) or cells can be passaged using EDTA
(EDTA dissociation buffer: 500 μL 0.5 M EDTA and 0.9 g
NaCl in 500 mL DPBS). Refer to Notes 14 and 22 for further
information regarding the use of EDTA dissociation in this
protocol.
4. E8, or Essential 8™ medium, is a simplified medium originally
developed by Chen et al. [15] for the culturing of hPSC and
hiPSC in a feeder free, chemically defined culture system.
TeSR™-E8™ is the commercially manufactured version of
the E8 formulation made by STEMCELL Technologies and
it can also be used. Other media, like NutriStem, which is both
serum free and xenofree, can also be used.
 Flow Cytometry and Immunophenotyping 143

A) Reference Stem Cell Markers

94.9% 97.1%

Count

Count
CD90

85.7% 86.0%

Count
Count

EpCam (CD326)

hiPSC(KB3) hESC (H9)

B) Posive Markers of Pluripotency

80.2% 84.9%
Count
Count

C-Kit (CD117)

49.5% 89.7%
Count
Count

Integrin α6 (CD49f)

74.9% 90.2%
Count
Count

SEMA6A

84.1% 90.8%
Count
Count

HAI-1

Isotype Control
Anbody
hiPSC(KB3) hESC (H9)

Fig. 3 Characterization of hiPSC and hESC by flow cytometry. Histograms of live KB3 hiPSC and H9 hESC
showing populations positive for (a) Stem cell reference markers and (b) markers of pluripotency
144 Daniel R. Riordon and Kenneth R. Boheler

5. Adjust volume to ensure the final concentration of ROCK

inhibitor is 10 μM following the addition of cells/medium.
6. We routinely use Phycoerythrin (PE)-conjugated antibodies
for most simple fluorescence analyses, however, when multiple
antibodies are needed, appropriate secondary fluorochrome-
conjugated antibodies must be used or primary antibodies
should be conjugated in the laboratory. Unconjugated mono-
clonal antibodies can be purchased in bulk from many manu-
facturers and subsequently conjugated with selected
fluorophores, ranging from fluorescein (FITC) to Texas Red
to Cascade Blue. All conjugations should be performed as
detailed in (http://www.drmr.com/abcon/). Antibody conju-
gation is a fairly rapid procedure that can be accomplished in
2–3 h.
7. An isotype control is an antibody of the same species, class
(heavy and light chains), and has the same fluorochrome
(e.g., PE) as the target primary antibody, but which lacks
specificity. The ideal isotype controls will also have the same
number of fluorescent molecules (fluorochrome–protein (F:P)
ratio). Isotype controls are necessary negative controls used to
determine the specific antibody signal by allowing the subtrac-
tion of nonspecific antibody binding (background) from the
total positive signal.
8. This protocol can be readily employed with most standard flow
cytometers, including those that permit cell sorting. In the case
of the latter, ROCK inhibitor should be added to the cells at all
stages to ensure cell viability after plating. Failure to add ROCK
inhibitor will adversely affect cell survival.
9. One aliquot of Matrigel® hESC-qualified Matrix in 25 mL of
DMEM/F-12 is sufficient to coat three 100 mm dishes (8 mL/
dish) or four 6-well plates (1 mL/well). It is important to
realize that different batches of Matrigel® or Geltrex® (see
Note 11) may require different volumes for coating of plates
and plating of cells. This needs to be determined empirically,
but we usually start with the recommended dilution; however,
we have on occasion been able to use a twofold greater dilution
or needed to use up to fourfold greater amounts of the matrix
for optimal hPSC growth.
10. Coating can also be accomplished at 30 min in 37
C incuba-
tor. Make sure incubation surfaces are level to achieve evenly
coated plates.
11. As an alternative to Matrigel®, add 250 μL of Geltrex® to
50 mL of cold (4
C) DMEM/F-12 (a 1:200 dilution) and
mix thoroughly. It is also important to realize that the amount
of Geltrex® or Matrigel® required for optimal growth varies
from 1:50 to 1:400 dilutions, and must be determined
 Flow Cytometry and Immunophenotyping 145

empirically. Plate Geltrex® onto tissue culture plates (as was

done for the Matrigel®), allow to polymerize at 37
C for at
least 1 h, and store, sealed with Parafilm™, at 37
C for up to
2 weeks.
12. Matrigel®-coated plates can be used immediately, or if desired,
stored at 4
C immediately after coating for up to 1 week with
addition of 1 mL per well of a 6-well plate or 7 mL per 100 mm
culture plate of DMEM/F-12 media and sealed with Paraf-
ilm® (to prevent dehydration). Previously stored plates should
be equilibrated to room temperature for 60 min prior to use.
Plates that have uneven coating or where the Matrigel® has
evaporated should not be used.
13. 6-Well plates are used for routine passaging and expansion,
while 100 mm plates are used for cell sorting following immu-
nostaining. Each 100 mm plate of cells should be sufficient for
10–15 antibody labeling reactions required for flow cytometry
analysis, and therefore, the total number of plates prepared
should be adjusted to reflect the scope of the experiment.
Alternatively, for cell sorting experiments, 1–2 100 mm plates
of cells should be prepared for each antibody labeling required
to ensure sufficient numbers of live cells for subsequent cultur-
ing and any potential downstream experiments.
14. It is also possible to use an EDTA-based procedure in conjunc-
tion with E8™ medium for routine passaging of hPSCs. This
method has the advantage of shorter protocol times and may
be preferred for high throughput experiments or when an
enzyme-free method is necessary. See Beers et al. [16] for a
detailed protocol.
15. Thawed aliquots of Accutase® can be used immediately or
stored at 2–8
C for up to 2 weeks. Do not refreeze.
16. The current protocol assumes an already established culture of
hESCs or hiPSCs. For initial thawing and seeding of hESCs
and hiPSCs, see supplier recommended procedures. Cell lines
should be well established (>p20) and exhibit a homogeneous
morphology with less than 10–20% differentiation in high
quality cultures. Also, when cultivating cells in E8™ medium,
there is normally very little overt differentiation. If there is
obvious differentiation, then it may be better to repeat the
experiment and ensure optimal growth conditions.
17. Use a new pipette or tip for each well/plate to reduce the
chance of cross-contamination. For large scale experiments,
cells from multiple wells/plates can be combined into single
15 or 50 mL conical centrifuge tube. When combining cells
from multiple plates, the ratio of enzyme/cells solution to
DMEM/F-12 should remain at a minimum of 1:1 to ensure
proper inactivation of the cell detachment solution.
146 Daniel R. Riordon and Kenneth R. Boheler

18. Use either automated or manual hemocytometer. The current

protocol has typically yielded 70–85% viability when using an
automated hemocytometer.
19. To save time, cells can be centrifuged during cell counting.
20. Depending on the cell line, we plate anywhere from 70 to
150,000 cells/well for use in 3–4 days. Higher numbers can
be used, however, we recommend cells be left in culture for
3–4 days before harvest; therefore, initial plating numbers will
need to be adjusted accordingly to attain the desired cell den-
sity prior to harvesting cells for antibody labeling.
21. Do not use a circular motion as cells will accumulated at the
center of the plate.
22. Use of EDTA dissociation solutions are not optimal for this
step as cells will clump together once calcium is added back,
affecting downstream labeling of cells.
23. Cell Dissociation incubation time is dependent on the cell type
and should be determined empirically, using the shortest pos-
sible time to generate optimal combination of single cell sus-
pension and high cell viability (at least 70–80%). Cells are
typically ready to harvest when cell boundaries begin to
round-up and colonies become less well packed.
24. Passing cells through a filter can improve downstream antibody
labeling by eliminating cell aggregates and ensuring only single
cells are labeled.
25. Optimizing the antibody concentration is essential for reduc-
ing nonspecific antibody binding and allowing for the best
discrimination between positive and negative results. For each
cell type and antibody used, an antibody titration assay should
be performed. Titrations should be performed in the same
conditions in which you plan to use the antibody, and an
appropriate isotype control should be included for each anti-
body tested.
26. Some manufacturers provide the recommended concentration
as a volume per number of cells (e.g., 10 μL/106 cells). In such
a case, begin the titration using 2–4 the recommended vol-
ume and prepare serial (1/2) dilutions in Cell Staining Solu-
tion as described.
27. Isotype controls should be added to samples at the same con-
centration as that of the test antibody.
28. It is not necessary to wash the cells between blocking and
immunodetection steps.
29. For optimal results, ensure primary antibody and corresponding
isotype controls are run at the same concentrations.
 Flow Cytometry and Immunophenotyping 147

30. The small volume may take several minutes to wet the entire
area of the strainer. This step can be performed during final
wash/centrifugation steps.
31. Straining of cells is necessary to eliminate large aggregates that
could clog the flow cytometer. Tubes can be tapped gently on
the benchtop to aid the collection of cell solution through
strainer and into the tube. Return cells to ice as quickly as
possible.
32. Forward-scattered light (FSC) is proportional to cell surface
area or cell size. Side-scattered light (SSC) is proportional to
cell granularity or internal complexity.
33. Dead cells can nonspecifically bind many antibody conjugates.
These signals can erroneously be counted as positively labeled
cells; therefore, gating strategies should be used to gate out
these cells. Intercalated dyes like propidium iodide (PI) and live
cell assays like Calcein AM can be useful for identifying and
gating dead cells.
34. Preliminary tests should be run to determine whether a cell
type can be effectively labeled using a specific antibody. For this
type of analysis, we usually use the cells from a single well of a
6-well plate and test the antibody using the manufacturer’s
recommendations. Antibody dilutions must then be per-
formed with more cells to optimize conditions (see Subhead-
ings 3.4 and 3.8).
35. Optimal baseline PMT gains need to be established empirically
at the outset, and should be determined based on the advice of
a FC Core facility or with the help of someone who is well
acquainted with flow cytometry.
36. It is possible for multiple parameters/signals to be measured;
however, if multiple fluorochromes are being assessed simulta-
neously, then quenching must be considered. Contact a FC
Core or experienced user for assistance.
37. If using a fluorochrome-conjugated secondary antibody, it is
best to titrate both the neat, primary antibody, as well as the
fluorochrome-conjugated secondary antibody. Begin by label-
ing the cells with the primary antibody at the manufacturer’s
recommended concentration (typically 1.25–20 ng/mL final
dilution) and then perform a series of secondary antibody
dilutions similar to the method outlined for the titration of
primary antibodies. Once the optimal concentration has been
determined for the secondary antibody, perform a titration of
the primary antibody as described. The listed concentration of
a fluorochrome-conjugated antibody includes both the anti-
body as well as the fluorochrome; therefore, the concentration
of each antibody required will likely be less.
148 Daniel R. Riordon and Kenneth R. Boheler

Acknowledgments

This research was supported by the Intramural Research Program

of the NIH, National Institute on Aging, by the Research Grants
Council of Hong Kong Theme-based Research Scheme T13-706/
11, and the Hong Kong Research Grant Committee General
Research Fund (Project number 17100214). We thank Robert
Wersto and the NIA FC Core Facility for assistance with flow
cytometry.

References
1. Thomson JA, Itskovitz-Eldor J, Shapiro SS, 8. Chin MH, Mason MJ, Xie W, Volinia S,
Waknitz MA, Swiergiel JJ, Marshall VS, Jones Singer M, Peterson C, Ambartsumyan G,
JM (1998) Embryonic stem cell lines derived Aimiuwu O, Richter L, Zhang J et al (2009)
from human blastocysts. Science Induced pluripotent stem cells and embryonic
282:1145–1147 stem cells are distinguished by gene expression
2. Tachibana M, Amato P, Sparman M, Gutierrez signatures. Cell Stem Cell 5:111–123
NM, Tippner-Hedges R, Ma H, Kang E, 9. Narsinh KH, Sun N, Sanchez-Freire V, Lee AS,
Fulati A, Lee HS, Sritanaudomchai H et al Almeida P, Hu S, Jan T, Wilson KD, Leong D,
(2013) Human embryonic stem cells derived Rosenberg J et al (2011) Single cell transcrip-
by somatic cell nuclear transfer. Cell tional profiling reveals heterogeneity of human
153:1228–1238 induced pluripotent stem cells. J Clin Invest
3. Takahashi K, Tanabe K, Ohnuki M, Narita M, 121:1217–1221
Ichisaka T, Tomoda K, Yamanaka S (2007) 10. Tang C, Lee AS, Volkmer JP, Sahoo D, Nag D,
Induction of pluripotent stem cells from adult Mosley AR, Inlay MA, Ardehali R, Chavez SL,
human fibroblasts by defined factors. Cell Pera RR et al (2011) An antibody against
131:861–872 SSEA-5 glycan on human pluripotent stem
4. Park IH, Zhao R, West JA, Yabuuchi A, cells enables removal of teratoma-forming
Huo H, Ince TA, Lerou PH, Lensch MW, cells. Nat Biotechnol 29:829–834
Daley GQ (2008) Reprogramming of human 11. Boheler KR, Bhattacharya S, Kropp EM,
somatic cells to pluripotency with defined fac- Chuppa S, Riordon DR, Bausch-Fluck D, Bur-
tors. Nature 451:141–146 ridge PW, Wu JC, Wersto RP, Chan GC et al
5. Yu J, Vodyanik MA, Smuga-Otto K, (2014) A human pluripotent stem cell surface
Antosiewicz-Bourget J, Frane JL, Tian S, Nie J, N-glycoproteome resource reveals markers,
Jonsdottir GA, Ruotti V, Stewart R et al (2007) extracellular epitopes, and drug targets. Stem
Induced pluripotent stem cell lines derived from Cell Rep 3:185–203
human somatic cells. Science 318:1917–1920 12. Koivuniemi R, Makela J, Hokkanen ME,
6. Narsinh KH, Plews J, Wu JC (2011) Compari- Bruelle C, Ho TH, Ola R, Korhonen L,
son of human induced pluripotent and embry- Schroder J, Kataoka H, Lindholm D (2013)
onic stem cells: fraternal or identical twins? Mol Hepatocyte growth factor activator inhibitor-
Ther 19:635–638 1 is induced by bone morphogenetic proteins
7. Ghosh Z, Wilson KD, Wu Y, Hu S, and regulates proliferation and cell fate of neu-
Quertermous T, Wu JC (2010) Persistent ral progenitor cells. PLoS One 8:e56117
donor cell gene expression among human 13. van Galen P, Kreso A, Mbong N, Kent DG,
induced pluripotent stem cells contributes to Fitzmaurice T, Chambers JE, Xie S, Laurenti E,
differences with human embryonic stem cells. Hermans K, Eppert K et al (2014) The
PLoS One 5:e8975 unfolded protein response governs integrity of
 Flow Cytometry and Immunophenotyping 149

the haematopoietic stem-cell pool during Howden SE, Diol NR, Propson NE et al
stress. Nature 510:268–272 (2011) Chemically defined conditions for
14. Yu KR, Yang SR, Jung JW, Kim H, Ko K, Han human iPSC derivation and culture. Nat Meth-
DW, Park SB, Choi SW, Kang SK, Scholer H ods 8:424–429
et al (2012) CD49f enhances multipotency and 16. Beers J, Gulbranson DR, George N, Siniscalchi
maintains stemness through the direct regula- LI, Jones J, Thomson JA, Chen G (2012) Pas-
tion of OCT4 and SOX2. Stem Cells saging and colony expansion of human plurip-
30:876–887 otent stem cells by enzyme-free dissociation in
15. Chen G, Gulbranson DR, Hou Z, Bolin JM, chemically defined culture conditions. Nat Pro-
Ruotti V, Probasco MD, Smuga-Otto K, toc 7:2029–2040
 Chapter 10

Detecting Cell Surface Expression of the G Protein-Coupled

Receptor CXCR4
Amanda M. Nevins and Adriano Marchese

Abstract
G protein-coupled receptors (GPCRs) are cell surface receptors that relay extracellular signals to the inside
of the cells. C-X-C chemokine receptor 4 (CXCR4) is a GPCR that undergoes receptor internalization and
recycling upon stimulation with its cognate ligand, C-X-C chemokine 12 (CXCL12). Using this receptor/
ligand pair we describe the use of two techniques, enzyme-linked immunosorbent assay (ELISA) and flow
cytometry, widely used to quantify GPCR internalization from the plasma membrane and its return to the
cell surface by recycling.

Key words G Protein-coupled receptor (GPCR), C-X-C chemokine receptor 4 (CXCR4), C-X-C
chemokine ligand 12 (CXCL12), Receptor internalization, ELISA, Flow cytometry

1 Introduction

G protein-coupled receptors (GPCRs) represent the largest class of

cell surface receptors [1, 2]. Approximately 800 GPCRs have been
identified in humans, the largest being the rhodopsin family, and
their number within our genome is a reflection of their importance
in human physiology [2]. GPCRs are often involved in disease, and
as a consequence, are the targets of over 40% of drugs currently on
the market [1, 3]. Comprised of an extracellular N-terminus, seven
transmembrane spanning α-helices, alternating extracellular and
intracellular loops, and an intracellular C-terminus, GPCRs func-
tion as complex signaling switchboards relaying information from
the outside to the inside of cells [4]. Upon activation by a diverse
range of stimuli, GPCRs transduce signals via conformational
changes propagated through their transmembrane helices to intra-
cellular molecules linked to various signaling cascades [3, 5, 6]. To
ensure that signals are of the appropriate magnitude and duration,
receptors are immediately uncoupled from intracellular signaling
pathways, contributing to signal termination in a process known as
homologous desensitization [2, 4]. Activated GPCRs are removed

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_10, © Springer Science+Business Media, LLC 2018

151
152 Amanda M. Nevins and Adriano Marchese

or internalized from the cell surface via endocytosis [7]. For most
GPCRs, a large fraction of the internalized receptor is rapidly
recycled to the plasma membrane leading to the resensitization of
signaling [8–11].
The GPCR that we use here, as an example of internalization
and recycling, is the C-X-C chemokine receptor 4 (CXCR4). Che-
mokine receptors constitute the largest branch of the γ subfamily of
rhodopsin-like GPCRs and, along with their ligands, chemotactic
cytokines, are involved in the direction of leukocyte trafficking
[3, 12–14]. Stimulation of CXCR4 by its cognate ligand,
CXC-chemokine ligand 12 (CXCL12), plays a role in cancer metas-
tasis and progression, HIV infection, and WHIM syndrome [3, 12,
15–18]. Monitoring the internalization and recycling of CXCR4
following CXCL12 stimulation using techniques targeted at the
quantification of cell surface receptor expression is one way to assess
the role of receptor trafficking and recycling in cellular fate. An
enzyme-linked immunosorbent assay (ELISA) is a plate-based tech-
nique that employs an enzyme, like alkaline phosphatase, to detect
an immobilized antigen after incubation with a substrate, yielding a
measurable product [19]. Flow cytometry detects fluorescence
emitted from cell-bound fluorophores upon excitation as they
pass in front of a light source [20]. In this chapter, these techni-
ques, ELISA (1.1 and 1.3) and flow cytometry (1.2), are described
for the quantification of CXCL12-stimulated surface-CXCR4
internalization and recycling.

2 Materials

2.1 Cell Lines 1. Human embryonic kidney cells (HEK293) (Microbix, Tor-
onto, CA) stably expressing tagged CXCR4 [9].
2. Human cervical cancer cells with high levels of endogenous
CXCR4 expression (HeLa, ATCC, Manassas, VA, USA).

2.2 Cell Culture 1. Fetal Bovine Serum (FBS).

Materials 2. Dulbecco’s Modified Eagle Medium (DMEM) with high glu-
and Reagents cose (4500 mg/mL) supplemented with L-glutamine, and
NaHCO3.
3. Complete DMEM: 10% FBS-supplemented DMEM.
4. 4-(2-hydroxyethyl)-piperazine ethane sulfonic acid (HEPES).
5. Incomplete DMEM: DMEM supplemented with 20 mM
HEPES.
6. 0.05% Trypsin—EDTA.
7. Phosphate-Buffered Saline (PBS) without calcium and magne-
sium (Ca+2/Mg+2-free PBS).
 Cell Surface Expression of CXCR4 153

8. Bovine Serum Albumin (BSA).

9. Poly-L-Lysine (PLL).
10. Tissue culture dishes—10- and 6-cm (Falcon), 6-well plates
(Eppendorf), 24-well plates (Falcon).

2.3 ELISA Materials 1. Assay medium: DMEM, 0.1% BSA, 20 mM HEPES, 1 mM Ca+2.
and Reagents 2. 100 mM Ca+2 stock.
3. Ca+2/Mg+2-free PBS.
4. PBS supplemented with 1 mM Ca+2.
5. Monoclonal mouse M1 anti-FLAG® antibody (Sigma-Aldrich,
St. Louis, MO, USA).
6. Ethylenediaminetetraacetic acid (EDTA).
7. 10 μM stock of CXCL12 (CXCR4 ligand/agonist) (Protein
Foundry, Milwaukee, WI, USA) (see Note 1).
8. AMD3100—CAS 155148–31-5 (CXCR4 antagonist) (Sigma-
Aldrich).
9. Fixing solution: 3.7% paraformaldehyde (PFA) in PBS.
10. Alkaline phosphatase-conjugated goat anti-mouse antibody
(Sigma-Aldrich).
11. P-Nitrophenyl phosphate (developing solution) (Bio-Rad
Laboratories, Hercules, CA, USA).
12. Diethanolamine buffer (Bio-Rad Laboratories).
13. 0.4 M NaOH.
14. Plate reader to measure absorbance. We use the FlexStation
3 multi-mode plate reader from Molecular Devices (Sunnyvale,
CA, USA).

2.4 Flow Cytometry- 1. CellStripper® (Fisher Scientific Co., Pittsburgh, PA, USA).
Specific Material 2. Trypan Blue.
and Reagents
3. Flow buffer: PBS supplemented with 0.1% BSA.
4. PE-conjugated anti-CXCR4 (Catalog No: #306506 (clone
12G5); BioLegend, San Diego, CA, USA).
5. IgG2a κ-isotype control (Catalog No: #400212; BioLegend,
San Diego, CA, USA).
6. Fixing solution: 3.7% paraformaldehyde (PFA) in PBS.
7. 10 μM stock of CXCL12 (CXCR4 ligand/agonist) (see Note 1).
8. Round bottom test tubes suitable for flow cytometry
(BD Bioscience, Franklin Lakes, NJ, USA).
9. Our flow cytometry facility is equipped with a FACSCalibur
flow cytometer and FlowJo software (BD Biosciences, San Jose,
CA, USA).
154 Amanda M. Nevins and Adriano Marchese

3 Methods

3.1 CXCR4 1. HEK293 cells stably expressing FLAG-tagged CXCR4 are

Internalization: ELISA maintained in 10-cm dishes containing 10 mL complete
DMEM prepared as described in Subheading 2.2, at 37
C in
a humidified atmosphere composed of 5% CO2 to a confluency
of 80–90% (see Note 2).
2. Passage cells into PLL (Poly-L-Lysine) coated 24-well plates
(or dishes) (see Note 3). Incubate at 37
C for an additional
24 h allowing cells to reach 100% confluency. Assessment of
CXCR4-internalization requires the following four conditions:
(a) time (t) ¼ 0 (T0); (b) t ¼ 0, isotype control antibody only as
background control (TBG), and (c/d) internalization (I) for
45 min (I45) for vehicle and CXCL12 treated cells. Each con-
dition should be performed in triplicate for 18 total wells (see
Note 4).
3. Wash cells once with 500 μL warm with incomplete DMEM.
Replace with 500 μL warm, incomplete DMEM for at least 3 h
(see Note 5).
4. Following serum starvation, place the 24-well plate(s) on ice
and aspirate the incomplete DMEM. Wash cells twice with
500 μL ice-cold (4
C) assay medium prepared as specified in
Subheading 2.3. Replace with 500 μL fresh, ice-cold (4
C)
assay medium and incubate on ice for 15 min (see Note 6).
5. Cell surface CXCR4 is labeled with the calcium-dependent M1
anti-FLAG® monoclonal antibody. Aspirate medium and
replace with the newly prepared medium containing the anti-
body. Add the antibody in a dilution of 1:100–250 μL of
ice-cold (4
C) assay medium. Incubate on ice for 1 h (see
Note 7).
6. Aspirate the medium containing the M1 anti-FLAG® mono-
clonal antibody and wash twice with 500 μL ice-cold (4
C)
assay medium.
7. Aspirate medium from the I45 wells and apply warm assay
medium containing vehicle (PBS þ 0.1% BSA) or 50 nM
CXCL12 (CXCR4 agonist) (see Note 8). Incubate treated
cells at 37
C for 45 min.
8. During this incubation aspirate medium from the T0 and TBG
wells and wash cells twice with ice-cold (4
C) assay medium.
Treat cells with 500 μL of 3.7% paraformaldehyde (PFA) in
PBS (fixing solution). Incubate plates at room temperature
(RT) for 5–10 min (see Note 9).
9. Following fixation, wash cells twice with 500 μL PBS contain-
ing 1 mM Ca+2. Replace with 500 μL fresh PBS (þ1 mM Ca+2)
 Cell Surface Expression of CXCR4 155

after this wash, and leave plates on ice to process in parallel with
the internalization (I45) wells.
10. Following the 45 min incubation with either vehicle or
CXCL12, aspirate medium from the internalization (I45)
wells. Fix and wash cells in the I45 wells as described for control
wells in steps 8 and 9. Cover with aluminum foil to prevent
exposure to light and leave on ice to process in parallel with the
I45 wells.
11. After washes incubate each plate with 300 μL of alkaline
phosphatase-conjugated goat anti-mouse antibody (see Note
10) for 1 h at RT.
12. Post-incubation, wash cells three times with 500 μL PBS
(þ1 mM Ca+2). Incubate plates with 250 μL developing solu-
tion that has been diluted in diethanolamine buffer for
5–15 min (see Note 11).
13. Quench the reactions of each well with 100 μL of 0.4 M
NaOH. Take 100 μL aliquots of each condition and measure
the absorbance at 405 nm using a plate reader.
14. Calculate the proportion of receptor internalized by dividing
the amount of surface receptor by the total number of receptor
present at the cell surface prior to treatment with CXCL12
according to the following formula:

Proportion of receptor internalized

¼ 1  ½ðI 45  T BG Þ  ðT 0  T BG Þ,

where T0 is equal to the total signal at time t ¼ 0, TBG is the

isotype only background control and I45 is the surface signal
remaining 45 min after either vehicle or CXCL12-stimulated
internalization. The percentage of receptor internalization can
be calculated by multiplying the result from the above formula
by 100.

3.2 CXCR4 Surface 1. Culture HeLa cells in 10-cm dishes as described in Subheading
Expression 3.1 until 90–95% confluent (see Note 2).
and Internalization: 2. Wash cells once with 10 mL incomplete DMEM. Subsequently,
Flow Cytometry serum-starve the cells in 10 mL of fresh incomplete DMEM for
at least 3 h (see Note 5).
3. Wash cells twice with 10 mL PBS and detach cells from plates
using a nonenzymatic dissociation buffer such as CellStripper®
(Corning) (see Note 12). Aspirate medium, add 2 mL of
CellStripper® and incubate cells at 37
C for 10 min or until
detached (see Note 13).
4. Add 8 mL cold (4
C) (see Note 6) flow buffer prepared as
specified in Subheading 2.4 and transfer cells to a conical
156 Amanda M. Nevins and Adriano Marchese

centrifuge tube. Pellet cells by centrifugation at 800  g for

3 min (see Note 14).
5. Resuspend cells in 2 mL cold (4
C) flow buffer. Count cells
using trypan blue staining and a hemacytometer. We typically
use an automated cell counter. Transfer approximately 5  105
cells per sample and condition [20] to a 5 mL round bottomed
test tube. Wash cells twice with 500 μL cold (4
C) flow buffer,
pelleting as described in step 4.
6. Resuspend cells in 500 μL warm (37
C) flow buffer and
incubate for 15 min at 37
C (see Note 15).
7. During the wash steps and incubation with ligand, prepare
antibody dilutions in cold (4
C) flow buffer and keep on ice
until use (see Note 16). Choosing the primary antibody is a
crucial step in experimental design (see Note 17). There are
multiple antibodies available that recognize several distinct
epitopes on CXCR4. In this analysis a 1:100 dilution of
PE-conjugated anti-CXCR4 was used (see Notes 18 and 19).
The presence of post-translational modifications on the recep-
tor including, but not limited to, sulfation (see Note 20) and
glycosylation [21] must also be considered as these can prevent
epitope recognition by monoclonal antibodies [21, 22]. Finally,
the fluorochrome you choose should have both a high quan-
tum yield and resistance to photobleaching (see Note 21).
8. Treat cells with 50 nM (final concentration) (see Note 20) of
CXCL12 or vehicle (PBS þ 0.1% BSA) for 2, 5, 10, 20, and
60 min (see Note 22).
9. After 60 min (or final time point) (see Note 23), wash cells
twice with 4 mL of cold (4
C) flow buffer, pelleting as
described in step 4. Resuspend cells in 500 μL of cold (4
C)
flow buffer containing antibody dilutions. Cover the samples
with aluminum foil in order to protect fluorochromes from
light and any subsequent photobleaching. Incubate samples
on ice for 20 min.
10. Following the incubation, wash cells twice with 4 mL of cold
(4
C) flow buffer, pelleting as described in step 4 while
limiting exposure to light. Resuspend cells in 500 μL of 3.7%
PFA (see Note 24) maintaining a concentration of 1  106
cells/mL in each sample.
11. After fixation, wash samples in 500 μL of cold (4
C) flow
buffer and run cell-associated fluorescence analysis immedi-
ately (see Note 25) or store samples at 4
C until use (see
Note 26).
12. Raw data is analyzed using a software package such as FlowJo.
The percent of internalized receptor is calculated using the
geometric mean of PE fluorescence intensity [10, 23].
 Cell Surface Expression of CXCR4 157

3.3 CXCR4 Receptor 1. Maintain HEK293 cells stably expressing FLAG-tagged

Recycling: ELISA CXCR4 in 10-cm dishes containing 10 mL complete DMEM
prepared as described in Subheading 2.2, at 37
C in a humi-
dified atmosphere composed of 5% CO2 to a confluency of
90–95% (see Note 2).
2. Passage cells into PLL coated 24-well plates (see Note 3).
Incubate at 37
C for an additional 24 h such that cells reach
90–95% confluency. Assessment of CXCR4-recycling requires
the following eight conditions: (a) time (t) ¼ 0 (T0); (b) t ¼ 0,
isotype only as background control (TBG); (c/d) internaliza-
tion (I) for 45 min (I45), for vehicle and CXCL12 treated cells
(see Note 27); (e/f ) recycling (R) at t ¼ 0 post-internalization
(R0), for vehicle and CXCL12; and (g/h) recycling after
60 min (R60), for vehicle and CXCL12 treated cells (see Note
28). Each condition should be performed in triplicate for
24 total wells (see Note 4).
3. Wash cells once with 500 μL warm, incomplete DMEM.
Replace with 500 μL warm, incomplete DMEM for at least
3 h (see Note 5).
4. Following serum starvation, place the 24-well plate(s) on ice
and aspirate the incomplete DMEM. Wash cells twice with
500 μL ice-cold (4
C) assay medium prepared as specified in
Subheading 2.3. Replace with 500 μL fresh, ice-cold (4
C)
assay medium and incubate on ice for 15 min (see Note 6).
5. Cell surface CXCR4 is labeled with the calcium-dependent M1
anti-FLAG® monoclonal antibody. Aspirate medium and replace
with the newly prepared medium containing the antibody. Add
the antibody in a dilution of 1:100 to 250 μL of ice-cold (4
C)
assay medium. Incubate on ice for 1 h (see Note 7).
6. Aspirate the medium containing the M1 anti-FLAG® mono-
clonal antibody and wash twice with 500 μL ice-cold (4
C)
assay medium.
7. Aspirate medium from the I45, R0, and R60 wells, and then
apply warm assay medium containing vehicle (PBS þ 0.1%
BSA) or 50 nM CXCL12 (CXCR4 agonist) (see Note 8).
Incubate treated cells at 37
C for 45 min.
8. During this incubation, aspirate medium from the T0 and TBG
wells and wash cells twice with ice-cold (4
C) assay medium.
Treat cells with 500 μL of 3.7% fixing solution. Incubate plates
at room temperature for 5–10 min (see Note 9).
9. Aspirate the fixing solution, wash cells twice with 500 μL PBS
containing 1 mM Ca+2. Replace with 500 μL fresh PBS
(þ1 mM Ca+2) after this wash, cover plates with aluminum
foil to prevent light exposure and leave plates on ice to process
158 Amanda M. Nevins and Adriano Marchese

in parallel with the internalization (I45) and recycling (R0 and

R60) wells.
10. Following the 45 min incubation with either vehicle or
CXCL12, aspirate medium from I45. Fix and wash cells in the
I45 wells as described for control wells in steps 8 and 9. Cover
with aluminum foil to prevent exposure to light and leave on
ice to process in parallel with the recycling (R0 and R60) wells.
11. For the R0 and R60 wells, the remaining surface antibody
bound to uninternalized receptor must be removed. Wash
each plate tree times with 500 μL Ca+2/Mg+2-free PBS con-
taining 0.04% EDTA (see Note 29).
12. The R0 wells should be processed as described above in steps
8 and 9.
13. In order to monitor receptor recycling, treat cells in the R60
wells with 500 μL DMEM supplemented with 1 mM Ca+2 and
10 μM AMD3100 (see Note 30). Incubate cells in this medium
for 60 min at 37
C.
14. Wash R60 wells with 500 μL PBS (þ1 mM Ca+2), then apply
fixing solution for 5 min on ice. Following fixation, wash all
plates three times with 500 μL PBS (þ1 mM Ca+2).
15. After washes, incubate all wells with 300 μL of alkaline
phosphatase-conjugated goat anti-mouse antibody (see Note
10) for 1 h at RT.
16. Post-incubation, wash cells three times with 500 μL PBS
(þ1 mM Ca+2). Incubate plates with 250 μL developing solu-
tion that has been diluted in diethanolamine buffer for
5–15 min (see Note 11).
17. Quench reactions with 100 μL of 0.4 M NaOH. Take 100 μL
aliquots of each condition and measure the absorbance at
405 nm using the “well scan” setting of a FlexStation 3.
18. Calculate the proportion receptor internalized by dividing the
amount of remaining surface receptor by the total number of
receptor present at the cell surface prior to treatment with
CXCL12 according to the following formula:

Proportion of receptor internalized

¼ 1  ½ðI 45  T BG Þ  ðT 0  T BG Þ,

where T0 is equal to the total signal at time t ¼ 0, TBG is the 2
-

only background control and I45 is the surface signal remaining
45 minutes after either vehicle or CXCL12-stimulated
internalization.
19. Using the determined proportion of internalized receptor,
calculate the percentage of receptor recycling by dividing the
proportion of internalized receptor by the amount surface
 Cell Surface Expression of CXCR4 159

receptor recovered post-incubation according to the following

formula:

%Receptor Recycling ¼ ðR60  R0 Þ

 f1  ½ðI 45  T BG Þ  ðT 0  T BG Þg,

where R60 is equal to the total signal recovered 105 min

(45 min stimulation þ 60 min recovery) after the initiation of
either vehicle or CXCL12-stimulated internalization, R0 is
equal to the signal remaining on t ¼ 0 after removal of the
Ca+2-dependent 1
antibody. This is a percentage of the inter-
nalized receptor that has recycled to the cell surface.

4 Notes

1. CXCL12 was purchased from Protein Foundry

(proteinfoundry.com). Protein Foundry produces recombi-
nant chemokines using rigorous production and quality con-
trol methods [24] to ensure the highest standards of product
quality and reproducibility in all research uses.
2. Ten-centimeter dishes work well when larger volumes of cells
are required, and 6-well plates (Eppendorf) or 6-cm dishes are
a good alternative for smaller scale analyses. For a 6-well plate
use 2 mL of PBS and 2 mL of the appropriate medium per well.
3. Twenty-four-well plates are coated with PLL in house. Briefly,
500 μL of a 1 μg/mL PLL stock is added to each well. After
15 min aspirate PLL from wells and let dry for 1 h. Before use
wash three times with PBS and let dry.
4. The number of conditions can change depending on altera-
tions to the cells including knockdowns and transfections;
however, the internalization and recycling portions of the
experiment monitors CXCR4 cell surface expression as dia-
gramed in Fig. 1. The amount of internalized receptor is
expressed as a proportion of the initial surface labeling of the
receptor (t ¼ 0).
5. Depending on the receptor type, cells are serum-starved to
minimize basal receptor activity [25]. We prefer to serum-
starve for 3–4 h. This can vary depending upon the cell type
or mechanism of receptor internalization. The researcher
should determine this empirically.
6. Endocytosis does not occur at 4
C. Keeping cells at this
temperature ensures cell surface labeling of the receptor and
results in synchronous activation of receptor internalization
upon return to 37
C [26]. Therefore, it is important that
samples remain on ice at all times and only ice-cold (4
C)
160 Amanda M. Nevins and Adriano Marchese

Fig. 1 Schematic representation of the conditions and experimental course for

the recycling experiment. In this example, the orange curve represents data from
the ELISA for internalization and the blue curve shows recycling of CXCR4. The
asterisk represents the washes stripping the remaining primary antibody from
the cell surface. The conditions denoted (T0, TBG, I45, R0, and R60) are the
minimum number of conditions monitored in the experiment

solutions are used. Pelleting should be done in a centrifuge

cooled to 4
C.
7. The M1 anti-FLAG® monoclonal antibody labels cell surface
FLAG-CXCR4 in a calcium-dependent manner and requires
the presence of at least 1 mM Ca+2 during all incubation and
wash steps.
8. Treatment with the agonist, in this case CXCL12, promotes
the internalization of the receptor/antibody complex [17].
9. Fixing the control wells (T0 and TBG) during the 45 min incu-
bation with vehicle/CXCL12 prevents constitutive receptor
internalization or loss of antibody binding.
10. Prepare antibody in a 1:1000 dilution in PBS þ 1% BSA. The
antibody dilution will have to be determined empirically for
each receptor and receptor expression system.
11. The alkaline phosphatase conjugated to the antibody enzymat-
ically processes p-nitrophenyl phosphate to p-nitrophenol in
the presence of diethanolamine resulting in the color change
monitored as the assay output [19]. The time required to
obtain a strong signal will depend on the expression of the
receptor and the efficiency of recycling. Care must be taken to
ensure that the signals obtained fall within the linear range of
your instrument to ensure accuracy.
 Cell Surface Expression of CXCR4 161

12. The use of a nonenzymatic dissociation buffer, rather than an

enzyme-based buffer prevents the digestion of the extracellular
domains of CXCR4 [26]. Digestion of these domains can
result in the removal of antibody epitopes and thereby impact
antibody labeling of the receptor or in some cases alter receptor
function in response to ligand stimulation. Enzymatic agents
can be used when incubations are long enough to allow for new
receptor synthesis to occur.
13. For more difficult cell lines, tapping the side of the dishes may
help to dislodge more cells. Gently break up clumps of cells by
repeat pipetting; clumped cells will not give accurate readings
by flow cytometry as they will count as a single event in the
cytometer [20].
14. Depending on the cell type, a lower RCF (relative centrifugal
force) may result in less stress on the cells. Cells must remain
intact, therefore, DO NOT exceed 1000  g; greater speeds
will exert sufficient force to damage cell membranes.
15. Resuspend two samples in cold (4
C) flow buffer to use as
controls, with and without ligand treatment. Although recep-
tor internalization is generally regarded as a ligand-dependent
event, basal levels of constitutive internalization and recycling
are possible [27, 28].
16. NaN3 is often added to the antibody dilutions prior to flow
analysis. If further functional assays are planned using the
sorted cells it should not be included in the primary antibody
buffer as it is also known to irreversibly inhibit the electron
transport chain [29].
17. The antibodies chosen should recognize epitopes that are dis-
tinct from ligand binding sites, in order to ensure that only
internalization is being monitored rather than other factors
such as epitope occupancy or masking by the ligand or receptor
activation [21].
18. The 12G5 monoclonal anti-CXCR4 antibody recognizes resi-
dues in the second extracellular loop (ECL2) of the receptor
[30]. ECL2 is a vital part of CXCL12 binding, recognition and
activation [12] and is necessary for CXCR4 to function as an
HIV coreceptor [16]. It has been shown that CXCL12 com-
petes with 12G5 for receptor binding, while AMD3100 blocks
antibody binding completely [22, 30].
19. As an alternative to the 12G5 monoclonal anti-CXCR4 anti-
body, fluorescently conjugated versions of the 4G10 (Santa
Cruz, sc-53,534) or 2B11 (BD Biosciences #551852) anti-
CXCR4 antibodies may be used. Each of these antibodies
recognizes the N-terminus of CXCR4 [17, 30] avoiding
some of the problems seen with 12G5, including competition
for CXCL12 binding [22].
162 Amanda M. Nevins and Adriano Marchese

20. Sulfation of the CXCR4 N-terminus is known to play a vital

role in CXCL12 binding and recognition [31–33]. Using an
antibody that obscures these moieties could inhibit CXCL12
binding and downstream CXCR4 activation. Alternatively this
posttranslational modification could also prevent antibody
binding and recognition [22].
21. The most commonly available fluorescently labeled anti-
CXCR4 antibodies are phycoerythrin (PE) conjugated. PE
has a maximum excitation at 565 nm, however, if the cyt-
ometer used is equipped with a blue laser (488 nM) rather
than yellow/green (561 nM) this fluorophore can be excited at
a lower wavelength resulting in reduced signal brightness at the
emission wavelength of 578 nM. If this is a concern, Alexa or
Brilliant Violet dyes could be employed instead [21, 22].
22. Using 500 μL sample volume described, a final concentration
of 50 nM CXCL12 is acquired by adding 2.5 μL of a 10 μM
stock directly to the appropriate tube. The required concentra-
tion of chemokine may need optimization. Some chemokines
induce internalization more readily at higher concentrations
(100–200 nM) [21].
23. Have individual time points prepared to end at the same time
(longest to shortest) so that samples can be processed in
parallel.
24. Returning the cells to 4
C stops any further internalization.
25. Set up the cytometer to count 10,000 events per
condition [26].
26. Samples can be stored in the dark at 4
C for 2–3 days after
fixing. If desired, prior to analysis transfer samples to mini
FACS (Fluorescence-activated cell sorting) tubes
(BD Bioscience).
27. I45 refers to wells that will be treated the same as in section 3.1
for the ELISA measuring CXCR4 internalization only. This is
because in order to calculate the percentage of receptor recy-
cling the initial amount of receptor internalization needs to be
known.
28. Recycling rates can differ between receptors, however, for
CXCR4 maximum receptor recycling was found to occur
after 60 minutes [9].
29. EDTA chelates metal ions; in this case Ca+2, resulting in the
uncoupling of the calcium-dependent M1 antibody from any
remaining surface receptors. Successive washes should be suffi-
cient to remove all bound M1 antibody. However, this must be
determined empirically. It is possible that prolonged incuba-
tions may be necessary to complete remove bound antibody,
which is essential to accurately calculate receptor recycling.
 Cell Surface Expression of CXCR4 163

30. AMD3100 is a CXCR4 antagonist, which is used to prevent

the binding of any residual CXCL12 in the medium that may
not have been removed by the washing step. It also serves to
compete CXCL12 off of any receptors recycled with ligand
(CXCL12) still bound. It should be noted that AMD3100
can bind to ACKR3, an atypical chemokine receptor that is
also a receptor for CXCL12 [34]. AMD3100 could theoreti-
cally promote cointernalization of CXCR4-ACKR3 heterodi-
mers and thereby limit the amount of CXCR4 that is recycled
and available at the plasma membrane.

Acknowledgment

This work was supported by NIH grant GM106727 (A.M.).

References

1. Fredriksson R, Lagerstrom MC, Lundin LG, 8. Bhandari D, Trejo J, Benovic JL, Marchese A
Schioth HB (2003) The G-protein-coupled (2007) Arrestin-2 interacts with the ubiquitin-
receptors in the human genome form five protein isopeptide ligase atrophin-interacting
main families. Phylogenetic analysis, paralogon protein 4 and mediates endosomal sorting of
groups, and fingerprints. Mol Pharmacol the chemokine receptor CXCR4. J Biol Chem
63:1256–1272 282:36971–36979
2. Tesmer JJ (2016) Hitchhiking on the heptahe- 9. Malik R, Marchese A (2010) Arrestin-2-
lical highway: structure and function of 7TM interacts with the endosomal sorting complex
receptor complexes. Nat Rev Mol Cell Biol required for transport machinery to modulate
17:439–450 endosomal sorting of CXCR4. Mol Biol Cell
3. Bachelerie F, Ben-Baruch A, Burkhardt AM, 21:2529–2541
Combadiere C, Farber JM, Graham GJ, 10. Malik R, Soh UJ, Trejo J, Marchese A (2012)
Horuk R, Sparre-Ulrich AH et al (2014) Inter- Novel roles for the E3 ubiquitin ligase
national Union of Basic and Clinical Pharma- atrophin-interacting protein 4 and signal trans-
cology. [corrected]. LXXXIX. Update on the duction adaptor molecule 1 in G protein-
extended family of chemokine receptors and coupled receptor signaling. J Biol Chem
introducing a new nomenclature for atypical 287:9013–9027
chemokine receptors. Pharmacol Rev 66:1–79 11. Marchese A, Benovic JL (2001) Agonist-
4. Van Rhee AM, Jacobson KA (1996) Molecular promoted ubiquitination of the G protein-
architechture of G protein-coupled receptors. coupled receptor CXCR4 mediates lysosomal
Drug Dev Res 37(1):1–38 sorting. J Biol Chem 276:45509–45512
5. Rajagopal S, Ahn S, Rominger DH, Gowen- 12. Chevigne A, Fievez V, Szpakowska M,
MacDonald W, Lam CM, Dewire SM, Violin Fischer A, Counson M, Plesseria JM, Schmit
JD, Lefkowitz RJ (2011) Quantifying ligand JC, Deroo S (2014) Neutralising properties of
bias at seven-transmembrane receptors. Mol peptides derived from CXCR4 extracellular
Pharmacol 80:367–377 loops towards CXCL12 binding and HIV-1
6. Whalen EJ, Rajagopal S, Lefkowitz RJ (2011) infection. Biochim Biophys Acta
Therapeutic potential of beta-arrestin- and G 1843:1031–1041
protein-biased agonists. Trends Mol Med 13. Moser B, Willimann K (2004) Chemokines:
17:126–139 role in inflammation and immune surveillance.
7. Neel NF, Schutyser E, Sai J, Fan GH, Rich- Ann Rheum Dis 63(Suppl 2):ii84–ii89
mond A (2005) Chemokine receptor internali- 14. Moser B, Wolf M, Walz A, Loetscher P (2004)
zation and intracellular trafficking. Cytokine Chemokines: multiple levels of leukocyte
Growth Factor Rev 16:637–658 migration control. Trends Immunol 25:75–84
164 Amanda M. Nevins and Adriano Marchese

15. Balabanian K, Lagane B, Pablos JL, Laurent L, 25. Pinilla-Macua I, Sorkin A (2015) Methods to
Planchenault T, Verola O, Lebbe C, Kerob D study endocytic trafficking of the EGF recep-
et al (2005) WHIM syndromes with different tor. Methods Cell Biol 130:347–367
genetic anomalies are accounted for by 26. Kershaw T, Wavre-Shapton ST, Signoret N,
impaired CXCR4 desensitization to CXCL12. Marsh M (2009) Analysis of chemokine recep-
Blood 105:2449–2457 tor endocytosis and intracellular trafficking.
16. Brelot A, Heveker N, Adema K, Hosie MJ, Methods Enzymol 460:357–377
Willett B, Alizon M (1999) Effect of mutations 27. Bonecchi R, Savino B, Borroni EM,
in the second extracellular loop of CXCR4 on Mantovani A, Locati M (2010) Chemokine
its utilization by human and feline immunode- decoy receptors: structure-function and
ficiency viruses. J Virol 73:2576–2586 biological properties. Curr Top Microbiol
17. Forster R, Kremmer E, Schubel A, Breitfeld D, Immunol 341:15–36
Kleinschmidt A, Nerl C, Bernhardt G, Lipp M 28. Borroni EM, Mantovani A, Locati M, Bone-
(1998) Intracellular and surface expression of cchi R (2010) Chemokine receptors intracellu-
the HIV-1 coreceptor CXCR4/fusin on vari- lar trafficking. Pharmacol Ther 127:1–8
ous leukocyte subsets: rapid internalization and 29. Wilson DF, Chance B (1967) Azide inhibition
recycling upon activation. J Immunol of mitochondrial electron transport. I. The aer-
160:1522–1531 obic steady state of succinate oxidation. Bio-
18. Ramsey DM, McAlpine SR (2013) Halting chim Biophys Acta 131:421–430
metastasis through CXCR4 inhibition. Bioorg 30. Carnec X, Quan L, Olson WC, Hazan U, Dra-
Med Chem Lett 23:20–25 gic T (2005) Anti-CXCR4 monoclonal antibo-
19. Engvall E (1980) Enzyme immunoassay dies recognizing overlapping epitopes differ
ELISA and EMIT. Methods Enzymol significantly in their ability to inhibit entry of
70:419–439 human immunodeficiency virus type 1. J Virol
20. Givan AL (2011) Flow cytometry: an introduc- 79:1930–1933
tion. Methods Mol Biol 699:1–29 31. Veldkamp CT, Seibert C, Peterson FC, De la
21. Anselmo A, Mazzon C, Borroni EM, Cruz NB, Haugner JC III, Basnet H, Sakmar
Bonecchi R, Graham GJ, Locati M (2014) TP, Volkman BF (2008) Structural basis of
Flow cytometry applications for the analysis of CXCR4 sulfotyrosine recognition by the che-
chemokine receptor expression and function. mokine SDF-1/CXCL12. Sci Signal 1:ra4
Cytometry A 85:292–301 32. Veldkamp CT, Ziarek JJ, Peterson FC, Chen Y,
22. Sloane AJ, Raso V, Dimitrov DS, Xiao X, Volkman BF (2010) Targeting SDF-1/
Deo S, Muljadi N, Restuccia D, Turville S CXCL12 with a ligand that prevents activation
et al (2005) Marked structural and functional of CXCR4 through structure-based drug
heterogeneity in CXCR4: separation of HIV-1 design. J Am Chem Soc 132:7242–7243
and SDF-1alpha responses. Immunol Cell Biol 33. Ziarek JJ, Getschman AE, Butler SJ, Taleski D,
83:129–143 Stephens B, Kufareva I, Handel TM, Payne RJ,
23. Alekhina O, Marchese A (2016) Beta-arrestin1 Volkman BF (2013) Sulfopeptide probes of the
and signal-transducing adaptor molecule CXCR4/CXCL12 interface reveal oligomer-
1 (STAM1) cooperate to promote focal adhe- specific contacts and chemokine allostery. ACS
sion kinase autophosphorylation and chemo- Chem Biol 8:1955–1963
taxis via the chemokine receptor CXCR4. J 34. Kalatskaya I, Berchiche YA, Gravel S, Limberg
Biol Chem 2016:M116.757138 BJ, Rosenbaum JS, Heveker N (2009)
24. Veldkamp CT, Koplinski CA, Jensen DR, AMD3100 is a CXCR7 ligand with allosteric
Peterson FC, Smits KM, Smith BL, Johnson agonist properties. Mol Pharmacol
SK, Lettieri C et al (2016) Production of 75:1240–1247
recombinant chemokines and validation of
refolding. Methods Enzymol 570:539–565
 Part III

Cell-Based Functional Analyses Related to Surfaceome

Content
 Chapter 11

NaV Channels: Assaying Biosynthesis, Trafficking, Function

Gordon F. Tomaselli and Federica Farinelli

Abstract
Integral to the cell surface is channels, pumps, and exchanger proteins that facilitate the movement of ions
across the membrane. Ion channels facilitate the passive movement of ions down an electrochemical
gradient. Ion pumps actively use energy to actively translocate ions, often against concentration or voltage
gradients, while ion exchangers utilize energy to couple the transport of different ion species such that one
ion moves down its gradient and the released free energy is used to drive the movement of a different ion
against its electrochemical gradient. Some ion pumps and exchangers may be electrogenic, i.e., the ion
transport they support is not electrically neutral and generates a current. Functions of these pore-forming
membrane proteins include the establishment of membrane potentials, gating of ions flows across the cell
membrane to elicit action potentials and other electrical signals, as well as the regulation of cell volumes.
The major forms of ion channels include voltage-, ligand-, and signal-gated channels. In this review, we
describe mammalian voltage dependent Na (NaV) channels.

Key words Ion channel, Electrophysiology, Stem cells, Optical recording, Seizure, Arrhythmia

1 Active Membrane Properties of Excitable Tissues

The currents that underlie biological excitability are carried by

complex transmembrane glycoproteins including ion channels,
exchangers and pumps. Channels facilitate the apparently incon-
gruous rapid (>106 ions/s) yet highly selective flux of ions across
the lipid bilayer down their respective electrochemical gradients. All
self-respecting ion channels exhibit two essential properties: gating
and selective permeation. Gating is the opening and closing of the
channel pore in response to a specific biological stimulus. The
stimuli that produce channel gating in excitable tissues include
changes in transmembrane voltage, ligand binding, or mechanical
stress or deformation. Ion selectivity is in part determined by
molecular sieving, and perhaps more importantly, by different
energetic strategies for transiently binding the permeant ion in
the pore.
Ion channels, once open facilitate the passive movement of ions
down an electrochemical gradient. In contrast ion pumps use

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_11, © Springer Science+Business Media, LLC 2018

167
168 Gordon F. Tomaselli and Federica Farinelli

energy to actively translocate ions, often against concentration or

voltage gradients. Ion exchangers utilize energy to couple the
transport of different ion species such that one ion moves down
its gradient and the released free energy is used to drive the move-
ment of a different ion against its electrochemical gradient. Some
ion pumps and exchangers may be electrogenic, the ion transport
they support is not electrically neutral and generates a current. This
review focuses on mammalian voltage dependent Na (NaV) chan-
nels. Specifically, we describe structure–function relationships, Na
channelopathies and remodeling, and then describe structural and
functional readouts. While we focus on studies in heart, the princi-
ples described here are applicable to other tissues containing volt-
age dependent Na channels.

2 Voltage-Dependent Sodium Channels (NaV)

2.1 Voltage-gated ion channels (VGICs) are abundant in the heart,

Structure–Function skeletal muscle and neuronal tissue. VGICs share common structural
Relationships themes: modular architecture; an ion-selective pore with highly con-
served pore-lining residues for channels with similar selectivity; a
common gating strategy using a charged membrane voltage sensor;
and auxiliary subunits that regulate trafficking and function. NaV
channels are among the most abundant ion channels in excitable
tissues such as working cardiac muscle (~100,000 copies/myocyte)
[1] and carry the major inward current (INa) for excitation. The NaV
channel was the first voltage-gated ion channel to be cloned [2]. The
NaV channel transmembrane domain sequence is highly conserved
from the eel electroplax to humans [3–7]. The pore-forming α
subunit is the only subunit required for current generation and
determines the ion selectivity and conductance properties. The α
subunit is comprised of four internally homologous domains
(labeled I–IV), each of which contains six transmembrane segments;
the peptide linkers between the fifth (S5) and sixth (S6) membrane
spanning repeats or pore (P) segments of each domain come
together to form the ion selective pore [8]. The S5-S6 linkers in
each domain has unique amino acid sequences and the structural
basis of permeation differs fundamentally from that of K channels or
prokaryotic Na channels [9–11] in which four identical P segments
can come together to form a ion-selective pore. The S1–S4 trans-
membrane domains and in particular the positively charged S4
membrane spanning repeat form the voltage sensing domain of the
channel. It is the movement of the voltage sensor upon depolariza-
tion that initiates a conformational change in the channel that lays
bare the pore and permits ion flux. Sustained depolarization results
in a distinct type of channel closure called inactivation that limits ion
movement and is part mediated by cytoplasmic regions of the chan-
nel [12–14]. Inactivated channels do not readily open and recovery
 NaV Channels - Review 169

N Na+
I II III IV S
S
N
β1
+
+

N C
N EFL C
Ankyrin G C
L/S N4L IQ
syn
C

Fig. 1 Topology of the Na channels (left). The domains I–IV wrap around a central ion permeable pore (right)

from inactivation requires repolarization of the membrane allowing

the channel to transit into a closed but activatable state.
There are nine members (α subunits) of the NaV1 subfamily of
sodium channels that exhibit tissue specific expression. The
NaV1.1–1.3 channels (encoded by SCN1A–SCN3A) are predomi-
nantly expressed in the central nervous system (CNS), NaV1.4
(SCN4A) in skeletal muscle, NaV1.5 (SCN5A) in the heart,
NaV1.6 (SCN8A) in the CNS and peripheral nervous system
(PNS), and NaV1.7, 1.8, and 1.9 (SCN9A, SCN10A, SCN11A)
are mostly expressed in the PNS.
Mammalian NaV channels are comprised of a single α and one
or two β subunits (as in Fig. 1). Four genes encode five different β
subunit proteins (NaVβ1, 1B, 2, 3, and 4, SCN1B–SCN4B), each
with a single membrane-spanning domain (type 1 topology) and a
large extracellular V-like immunoglobulin fold often found in cell
adhesion molecules [15]. The only exception is the NaVβ1B splice
variant that is secreted and lacks the small carboxyl terminal cyto-
plasmic domains characteristic of the other isoforms. The α and β
subunits interact by two mechanisms, β1 and β3 subunits interact
noncovalently with the N- and C-termini of the α subunit [16, 17],
in contrast β2 and β4 subunits are linked by a disulfide bond to the
α subunit formed by a cysteine in the N-terminal Ig domain of the
beta subunits [18–20]. Beta subunits are present in wide range of
tissues and a general function is to promote the expression and
specific subcellular localization of NaV α subunits [15]. The struc-
ture of NaV β subunits is similar to that of classes of cell adhesion
molecules (CAMs) [21] and β subunits promote adhesion and can
do so even in the absence of NaV α subunits. The distinct kinetic,
voltage dependent and pharmacological properties of NaV channels
are dependent on the specific α-β subunit combination and cell
expression systems [22].
NaV channels exist in one of three classes of conformational
states (closed, open, or inactivated) with distinct voltage- and time-
dependent rate constants for transitions between states. In response
170 Gordon F. Tomaselli and Federica Farinelli

to a depolarization, charged residues in each S4 segment act as a

voltage sensor [23], physically move within the membrane [24] and
allow ion flux through the pore. The molecular structure of the
activation gate in eukaryotic NaV channels is not precisely defined
but evidence implicates the S6 segment [25–27]. The closed and
inactivated states are conformationally distinct, nonconducting
states of NaV channels. Maintained depolarization causes open
channels to inactivate in a process coupled to activation
[12, 28]. Fast inactivation is mediated at least in part by a cytoplas-
mic linker between domains III and IV [12, 13, 29] that may
function as a hinged lid docking onto a receptor formed by cyto-
plasmic of domains of the channel [30]. Slower inactivated states
are defined by their rates of recovery which can occur over tens of
seconds or longer [31] with structural determinants that are incom-
pletely defined [32].
Isoforms of NaV1 channels differ in length with long and short
variants. For example, compared to the skeletal muscle channel
(NaV1.4), neuronal (NaV1.1–1.3 and cardiac (NaV1.5) isoforms
have a larger intracellular linker between domains I and II that
contains consensus sites for phosphorylation by cyclic adenosine
monophosphate (cAMP)-dependent protein kinase A (PKA) and
Ca+2 calmodulin dependent kinase (CaMKII). PKA-mediated
phosphorylation of NaV1 channels may alter the trafficking, con-
ductance and gating of channels depending upon the channel
isoform and expression system [33–37]. In contrast to PKA, pro-
tein kinase C (PKC) modulates function of all mammalian Na
channel isoforms. PKC phosphorylates a highly conserved serine
in the III–IV linker to decrease the maximal channel conductance
and alter gating in an isoform-specific fashion [38, 39] and has been
implicated in muscarinic modulation of Na currents [40].
In normal hearts, about 1–2% of the Na channels exhibit slow
current decay resulting in a persistent/late (INa-L) that contributes
to AP plateau. The amplitude of INa-L is increased in ischemic,
hypertrophied or failing ventricular myocardium as well as under
conditions of acute oxidant stress [41, 42]. Significant increases in
INa-L may contribute to altered repolarization [43] and can result in
[Na+]i and [Ca+2]i loading and consequent depolarization of the
membrane potential, slower dV/dt (phase 0) of the AP, longer
APD, and increased risk of arrhythmogenic early after depolariza-
tions (EADs) and delayed afterdepolarizations (DADs). NaV1
phosphorylation by CaMKII has been implicated in Na current
remodeling in the diseased heart and in particular the increase in
INa-L [44–50]. CaMKII-induced alterations in NaV1.5 function
represents one of several links between NaV1 channels and regula-
tion of [Ca+2]i in the heart.
 NaV Channels - Review 171

2.2 Na Mutations in Na channels were among the first molecularly char-

Channelopathies acterized human ion channel diseases [51]. Mutations in α and β
and Remodeling subunits have been associated with a wide array of human diseases.
Genes encoding sodium channel α subunits that are expressed in
the central nervous system (e.g., NaV1.1, 1.2, 1.3, 1.6) have been
associated with seizure disorders and cognitive impairment
[52]. Missense mutations in NaV1.7, 1.8, 1.9 prominently
expressed in the PNS have been associated with pain-related dis-
orders [53]. SCN4A (NaV1.4) harbors mutations that cause myo-
tonias and periodic paralyses [54]. Rare allelic variants in SCN5A
encoding NaV1.5, have been linked to long QT syndrome type
3 (LQTS3), Brugada syndrome (BrS), progressive cardiac conduc-
tion disease (PCCD), dilated cardiomyopathy (DCM), sick sinus
syndrome (SSS), atrial fibrillation (AF), sudden infant death syn-
drome (SIDS), and overlap syndromes (for review see [55]).
Arrhythmic disorders with complex phenotypes have been asso-
ciated with mutations in both α and β Na channel subunits
[56–61]. Similarly, mutations in NaV β subunits have been asso-
ciated with epilepsy, a variety of cardiac arrhythmias and predispo-
sition to sudden death and pain disorders [15]. The role of beta
subunits in the regulation of α subunit function and cell adhesion
has led to the recognition of the association of genetic variants with
demyelinating and neurodegenerative diseases either through
altered cell contact or dysregulation of channel function [15, 62].
Altered expression of beta subunits has been observed in a number
of cancers and may serve as biomarkers and possibly therapeutic
targets although their role in pathogenesis is yet to be defined.

3 Methods

3.1 Structural Ion current recording, cellular and tissue electrophysiology gener-
and Functional ate a rich tapestry of data that in combination with molecular and
Readouts protein chemical analyses facilitates a detailed functional assessment
of ion channels. These studies are essential to our understanding
the roles of ion channels in excitation, contraction, secretion, and
other essential biological functions. Moreover, such methods allow
for the study of derangements in channel function and the contri-
bution of such derangements to disease pathogenesis in rare inher-
ited traits and more common acquired diseases. Finally ion channels
are the targets of a number of drugs and toxins that are used for
everything from treatment of disease to biological warfare. Chan-
nels are attractive targets for treatment of disease but carry a serious
liability of potentially life-threatening side effects.
172 Gordon F. Tomaselli and Federica Farinelli

3.2 Channel The cornerstone of the functional assessment of ion channels is

Biosynthesis recording of membrane currents or an integrated cellular response
and Trafficking such as action potentials (AP) or calcium transients (CaT). In order
to assess the function of an NaV1 channel it must be synthesized,
assembled with the proper associated proteins, transported and
inserted into the plasma membrane. As with any other protein a
number of methods are used to interrogate the synthesis of a
functional channel. The human genome encodes for over
140 known ion channel proteins, nine are NaV1α and four NaV1β
subunits. Molecular methods such as polymerase chain reaction
(PCR) are readily used to assess the level of NaV1 mRNA transcripts
as well as alternative splice variants. It is increasingly evident that in
some forms of cardiac and neurological disease, regulation at the
level of the transcript by either RNA editing or alternative splicing
may be important in the phenotypic expression of disease. Further-
more, modification of mRNA as has been exploited in models of
muscular dystrophy may provide a future pathway for development
of therapeutics [63, 64].
Proper trafficking and selective localization of NaV channels is
essential to normal tissue function. Most ion channel and mem-
brane bound proteins are glycosylated in the ER and modified in
the ER and Golgi network. A number of associated proteins are
important in promoting transport of channel proteins to the sur-
face membrane. Notably β subunits [15] and in the case of neuronal
NaV1 isoforms annexin II light chain (p11) and contactin increase
plasma membrane expression.
Subcellular localization of ion channels is important in a num-
ber of excitable tissues. In myelinated neurons that exhibit saltatory
conduction of action potentials along nodes of Ranvier, ion chan-
nels and transporters are concentrated at the axonal initial segment
(AIS) and the nodes of Ranvier. A number of methods have been
used such as fluorescence photobleaching recovery (FRP) of
labeled, NaV-specific neurotoxins (e.g., α and β scorpion toxins,
tetrodotoxin [TTX], saxitoxin [STX]) to assess the mobility of
channels in regions of the neuronal membrane. Channels are freely
diffusible in the cell soma, but mobility is significantly restricted in
the AIS and at the nodes of Ranvier [65]. An ankyrin G binding
motif of the form [V/A]P[I/L]AxxE[S/D]D in the II–III linker of
NaV1 channels is sufficient to restrict the localization of channels to
the AIS [66]. Ankyrin G is tethered to the actin cytoskeleton
through binding to βIV spectrin [67]. A similar strategy is operative
for localization of channels at the nodes of Ranvier but glial cells
that produce myelin are also required. Trafficking is important in
cardiac tissues where NaV1.5 is the predominant isoform and in fact
failure to exit the ER may be associated with inherited arrhythmias
as seen in Brugada syndrome [68]. The details of the subcellular
localization of NaV1.5 channels in cardiac ventricular myocytes
have more recently emerged. Two distinct pools of NaV1.5
 NaV Channels - Review 173

LATERAL MEMBRANE

N
SIV
Ank G

PDZ
Syntrophin
Dystrophin

INTERCALATED DISK

N
SIV
Ank G
SAP97
PDZ
PDZ
PDZ

Plakophilin-2

Fig. 2 Schematic of the proteins involved in localization of Na channels at the lateral membrane (top) or the
intercalated disk (bottom). A different constellation of proteins with PDZ binding domains and ankyrin G
binding proteins. Modified from Petitprez et al., 2011

channels exist in cardiac myocytes one that resides at the interca-

lated disks (ID) at cell ends and another associated with the lateral
and T-tubular membranes (Fig. 2). In a fashion analogous to other
neuronal NaV1 channel localization at the AIS, NaV1.5 channels at
the ID seem to be organized by ankyrin-G binding [69], channels
localized to the lateral membrane compartment are complexed with
dystrophin/syntrophin [70]. The role of membrane-associated
guanylate kinase (MAGUKs) like SAP97 is controversial. It is likely
that in addition to differences in the macromolecular complexes,
the function of these pools of NaV1 channels differs and may be
differentially remodeled in structural heart disease.
174 Gordon F. Tomaselli and Federica Farinelli

Fig. 3 Total internal reflectance fluorescence (TIRF, left) and superresolution fluorescence microscopy (SRFM,
center) localization of NaV1.5 channels in cardiac myocytes. High magnification SFRM image (right) from
inside the white box shows co-localization of NaV1.5 and N-cadherin. The scale bar in the two images on the
right and that on the left are 5 mm and 1 mm, respectively. Image from Agullo-Pascual et al., 2014

Traditionally studies of trafficking and subcellular localization

have used fluorescence staining and confocal microscopy with opti-
cal sectioning and electron microscopy of cells and tissues. Higher
resolution methods such as multicolor superresolution localization
microscopy use dual laser excitation and either total internal reflec-
tance (TIRF) or highly inclined illumination to reduce out of plane
fluorescence and improve the signal-to-noise [71, 72]. Such meth-
ods hold the promise of more refined understanding of the struc-
ture and assembly of macromolecular channel complexes (Fig. 3).

3.3 Electrical The main function of ion channels is to signal by way of modulating
and Optical Recording transmembrane voltage. The ability to directly record currents from
cells and tissues provides an unmatched tool for characterizing
channel function and correlating functional features with molecular
structure. Moreover, voltage recording using electrodes or optical
dyes affords a means to study the function of a channel or channels
in the context of other electrogenic molecules.

3.4 Voltage Early voltage recording in the squid giant axon attested to the
and Patch Clamp importance of the transmembrane movement of Na+ in the gener-
Recording ation and conduction of the action potential [73]. Arguably the
single most importance advance for the quantitative study of ionic
currents was the voltage clamp developed over 65 years ago
[74–76] and a cornerstone method still used today. Using voltage
clamp recording with the proper ionic conditions currents carried
by different ions (Na+, K+, etc.) can be studied in detail. In this
method the experimenter controls the voltage across the membrane
and a feedback amplifier injects current to keep the membrane
voltage constant (Fig. 4a). Clamping the membrane voltage at
 NaV Channels - Review 175

A B

50
a1 a2 V (mV)
0
-160 -110 -60 -10 40
-10 mV
-50
1 nA

-90 3 ms
5 mS -100

-150

I (nA)
i = -1.9 pA
-200
1.2

I/Imax
1

a3 a4 0.8

0.6

0.4
1nA
0.2
2ms
V (mV)
0
-150 -100 -50 0

Fig. 4 Electrophysiological recording of Na currents. (a). Configuration for whole cell current recording. (a1)
Whole cell recordings of Na currents expressed in mammalian tissue culture cells held at a holding voltage of

120 mV and stepped to a range of depolarizing test voltages. (a2) Current-voltage (IV) relationship, peak
currents plotted over a range of test voltages. (a3) Currents recorded at a test voltage of
20 mV over a range of
holding voltages to generate (a3), the steady-state inactivation curve. (b). Configuration for cell-attached
current recording (top) and exemplar single Na channels expressed in a membrane patch of an HEK-293 cell

different levels leads to a change in the activities of ion channels that

generate transmembrane currents, injection of current by the elec-
tronics holds the membrane voltage constant and allow for quanti-
fication of the magnitude and time dependence of the currents that
are active.
Another major advance in the study of ion channels was the
development of the high resistance, low noise membrane seal (giga-
seal) allowing for high fidelity recording from small patches of
membranes using glass pipettes (patch clamp recording)
[77]. Patch clamp methods are versatile and configurations of the
membrane seal formed allow for recording from patches of mem-
branes on the cell surface (cell-attached), excised from the cell
membrane with the external surface of the membrane facing into
the pipette (inside-out) or facing the bath (outside-out) and finally
rupture of the membrane patch allows for access to the cell interior
176 Gordon F. Tomaselli and Federica Farinelli

A B

5 ms
C 20 mV 10 pA/pF
200 ms

0
control E-4031
20 mV

200 ms

Fig. 5 Na channels in adult ventricular myocytes. (a) Adult ventricular cardiomyocyte, the scale bar is 50 mm.
(b) Action potential (AP) recorded from an adult ventricular myocyte, the rapid upstroke is mediated by Na
currents. (c) The effect of a potassium channel blocking drug, E4031 on ventricular APs

in the whole-cell configuration (Fig. 4b). This powerful methodol-

ogy allows for the recording of current through single channel
proteins. In addition, current injection into excitable cells that
have been patch clamped allow for the recording of action potential
using a voltage follower and a recording mode referred to as current
clamping. In contemporary cellular electrophysiology, patch clamp
recording has replaced most other forms of voltage clamp record-
ing. This section will focus on patch clamp for recording of ionic
current or cellular action potentials.
NaV channels are often the most abundant ion channels in
excitable cells and tissues and can be studied in the native cell
context. The confounders when using patch clamp recording of
NaV currents in native cells is that they of course are not the only
currents present so that specific ionic conditions and judicious use
of blockers of other currents are required to isolate, quantify and
biophysically characterize NaV currents. As importantly the size and
speed of NaV currents push the limit of the temporal fidelity of
typical whole-cell recording configurations as such the currents are
often studied at low temperature with low concentrations of Na+ in
the bath. This is a significant concern in cardiac ventricular myo-
cytes where the cells are large (100  30 μm) and the currents are
big (Fig. 5). Typically Na currents are recorded at room tempera-
ture with 5–10 mM Na+ in the extracellular recording solution
rather than physiological concentrations that exceed 130 mM.
Among the virtues of recording the currents through ion channels
 NaV Channels - Review 177

in their native cellular environment is that ancillary proteins that

may modulate function are likely to be present and regulation by
biologically relevant signaling systems may be possible. A caveat is
that recording macroscopic currents is typically performed in the
whole-cell configuration; in this case the cell contents are
exchanged or dialyzed with the contents of the patch pipette
(Fig. 4a). This allows for control of the intracellular environment
but not recapitulation of the cytosolic composition of the cell. A
modification of the whole-cell configuration, called the perforated
patch that does not rupture the membrane but instead permeabi-
lizes it by adding amphotericin or nystatin to the patch solution,
compromises the control of the membrane voltage but allows larger
molecules to remain in the cytosol during recording. This configu-
ration is used for current injection and action potential
(AP) recording (Fig. 5) and when studying highly regulated slower
currents, but less often when recording fast currents like those
through NaV channels. AP recording allows for the study of the
impact of ionic currents on an integrated cellular electrophysiologi-
cal response. For example, evaluating the impact of a pharmacolog-
ical modulator of an ionic current on the AP profile and dynamics
(Fig. 5c).
The cloning of ion channels has allowed for the study of
currents in isolation by expression in heterologous systems. The
cellular platforms for heterologous expression include mammalian
cells, Xenopus oocytes, and insect cells. Cloning of ion channel
complementary DNAs into expression vectors allows for introduc-
tion of the channel DNA or RNA into cells where channel proteins
can be synthesized, trafficked, and expressed in the cell membrane.
Strong constitutive promoters may allow for high levels of channel
expression that facilitate single channel recording. Generally mam-
malian cells used for expression are smaller that cells from excitable
tissues, this may technically improve the quality of the voltage
clamp, a feature that may be particularly important in the case of
NaV currents. This approach allows for expression of not only wild-
type channels but proteins engineered to contain mutations to
mimic disease or interrogate aspects of channel structure or func-
tion. Serial replacement of amino acids in channel proteins has been
used to interrogate the structural basis of channel function. The
limitations of this approach are that the channels are expressed in a
nonnative cell background and differences in function may be the
result of the expression background. Detailed studies of regional
expression of channels in cells are generally not possible in culture
systems and the impact of channel function on cellular biology may
not be possible. Signaling systems that modulate channel function
in native cells may not be present in cultured cells used for heterol-
ogous expression, limiting the study of physiological regulation of
channel in such systems.
In an effort to recreate more physiological systems, heterolo-
gous expression can also be used in primary tissue culture cells to
178 Gordon F. Tomaselli and Federica Farinelli

A Non-transduced WT NaV1.5
150
6 200 6

150 100

0 0
100
50
50
-6 -6
CV = 20.2 cm/s CV = 28.1 cm/s
0 0
-5 0 5 -5 0 5
B
200 ms

NT APD80 = 223 ± 6 ms NaV1.5 APD80 = 168 ± 9 ms

Fig. 6 Neonatal rat ventricular myocytes (NRVM) in monolayer culture. (a) Isochronal maps of 2D monolayer
cultures of nontransduced (left) and NRVMs infected with a lentivirus expressing human NaV1.5 channels. (b)
The corresponding optical action potentials (APs). Infecting cells with Na channels hastens conduction in
monolayer cultures

mimic the native cell environment. The challenges include effi-

ciently transducing primary culture cells, the presence of native
currents and changes in the cellular phenotype with time in culture.
Typically viral expression constructs are created to improve the
overall efficiency of expression in primary cells, if substantial endog-
enous current is present that confounds the interpretation of the
experiments, the endogenous currents can be silenced by RNA
inhibition using small interfering RNAs (siRNA) or short hairpins
(shRNAs), [78] which is facilitated if the species isoform of the
channel differs from that of the host cell [79].
Transduced native cells, particularly from the heart will form an
electrically connected syncytium in culture that facilitates the study
of networks of excitable cells. The impact of expressed channel
variants on tissue properties such as automaticity, conduction,
refractoriness and arrhythmia induction can be studied in these
preparations. The cells can be cultured on multielectrode arrays
(MEA) for standard electrical recording or stained this voltage
sensitive dyes for optical recording (Fig. 6).
Remarkable advances in the understanding and use of progeni-
tor cells have produced major changes in the approach to human
disease. Induced pluripotent stem cells (iPSCs) have driven a para-
digm shift in the modeling of human disease [80, 81]; the ability to
reprogram and redifferentiate patient-specific cells holds the prom-
ise of enhanced understanding of disease mechanisms, patient-
 NaV Channels - Review 179

Fig. 7 Optical action potentials (APs) and isochronal maps recorded in monolayers of control hiPSC-VCMs after
29 days in culture paced from a point source. The optical APs (left) shown are recorded from the black boxes in
the voltage maps (right). Slowing of conduction occurs at faster pacing rates. Loss of 1:1 capture occurred at
pacing CL < 500 ms

specific predictive pharmacology/toxicology, improved cell ther-

apy, and ultimately regenerative medicine. The same methods to
study ionic currents in primary cells in culture apply to human
iPSC-derived tissue cells. The virtues of human iPSC-derived tissue
cells include the ability to study the currents in human cells that
would not normally be available (e.g., heart cells and neurons), the
cells are readily manipulated to introduce genes of interest and in
the case of an inherited channelopathy, the mutant channel can be
studied in the native cell context and the platform offers the possi-
bility of correction of the mutation by genome editing. There are
however significant limitations that include the phenotypic hetero-
geneity of iPSC-derived cells and perhaps the most significant
barrier is the relative immaturity of iPSCs even with extended
time in culture which limits their ability to replicate normal adult
cell/tissue physiology. Despite these limitations, hiPSC derived
cells offer an opportunity to study NaV currents in tissue-type
specific cells derived from humans at both the individual cell level
and in 2D and 3D cultures (Fig. 7). Human iPSCs have already
been extensively used to model human neurological, cardiac and
muscular diseases.

3.5 High-Throughput The introduction of a planar glass electrode rather than a pipette
Patch Clamp over a decade ago [82] ushered in the era of high-throughput
Recording automated patch clamp recording [83]. The planar patch
approaches allow for more automated ion current studies than
180 Gordon F. Tomaselli and Federica Farinelli

conventional patch clamp recording and parallelization, that is, the

use of a number of apertures in the recording platform to allow for
a dramatic increase in the number of simultaneous recordings.
There are a handful of commercially available systems (SyncroPatch
384 Patch Engine, Nanion Technologies; Q Patch HT/HTX,
Sophion; IonWorks Barracuda, Molecular Dynamics Systems; Ion-
Flux HT, Fluxion Bioscience; DynaflowHT Cellecticon) that sup-
port different degrees of automation and parallelization. Cell lines
are most commonly used in these systems although most will now
accommodate stem cells and primary cells. The systems can record
fast currents like those through NaV1 channels but remain expen-
sive with most uptake of this technology in the pharmaceutical
industry and in fewer large academic laboratories.

4 Conclusions

Ionic currents underlie biological excitability and critical physiolog-

ical functions as diverse cognition, excitation and contraction of
muscle, and hormonal secretion. Voltage dependent Na channels
are the most abundant channels in excitable tissues such as nerve
and heart and skeletal muscles. Mutations in voltage dependent ion
channels have been associated with a number of human diseases.
These channels serve as targets for drugs that have been developed
to treat a number of human maladies and diseases including epi-
lepsy, cardiac arrhythmias, headache, pain syndromes, and neuro-
muscular disorders. The cell surface expression, biological function,
amenability to expression in heterologous cells and tissues facilitate
detailed studies of the complex membrane proteins. A varied set of
approaches have been developed to study the expression, traffick-
ing, and function of ion channels at a number of levels of integra-
tion from the single molecule to the intact organism.

References
1. Makielski JC, Sheets MF, Hanck DA, January channel expressed by sensory neurons. Nature
CT, Fozzard HA (1987) Sodium current in 379:257–262
voltage clamped internally perfused canine car- 5. George AL Jr, Komisarof J, Kallen RG, Barchi
diac Purkinje cells. Biophys J 52:1–11 RL (1992) Primary structure of the adult
2. Noda M, Shimizu S, Tanabe T, Takai T, human skeletal muscle voltage-dependent
Kayano T, Ikeda T, Takahashi H, Nakayama sodium channel. Ann Neurol 31:131–137
H et al (1984) Primary structure of Electroph- 6. Gellens ME, George AL Jr, Chen LQ,
orus electricus sodium channel deduced from Chahine M, Horn R, Barchi RL, Kallen RG
cDNA sequence. Nature 312:121–127 (1992) Primary structure and functional
3. Dib-Hajj SD, Tyrrell L, Black JA, Waxman SG expression of the human cardiac tetrodotoxin-
(1998) NaN, a novel voltage-gated Na channel, insensitive voltage-dependent sodium channel.
is expressed preferentially in peripheral sensory Proc Natl Acad Sci U S A 89:554–558
neurons and down-regulated after axotomy. 7. Ahmed CM, Ware DH, Lee SC, Patten CD,
Proc Natl Acad Sci U S A 95:8963–8968 Ferrer-Montiel AV, Schinder AF, McPherson
4. Akopian AN, Sivilotti L, Wood JN (1996) A JD, Wagner-McPherson CB et al (1992) Pri-
tetrodotoxin-resistant voltage-gated sodium mary structure, chromosomal localization, and
 NaV Channels - Review 181

functional expression of a voltage-gated subunit with similarity to beta2. J Neurosci

sodium channel from human brain. Proc Natl 23:7577–7585
Acad Sci U S A 89:8220–8224 20. Chen C, Calhoun JD, Zhang Y, Lopez-
8. Yellen G, Jurman ME, Abramson T, MacKin- Santiago L, Zhou N, Davis TH, Salzer JL,
non R (1991) Mutations affecting internal TEA Isom LL (2012) Identification of the cysteine
blockade identify the probable pore-forming residue responsible for disulfide linkage of Naþ
region of a Kþ channel. Science 251:939–942 channel alpha and beta2 subunits. J Biol Chem
9. Payandeh J, Scheuer T, Zheng N, Catterall WA 287:39061–39069
(2011) The crystal structure of a voltage-gated 21. Isom LL, Catterall WA (1996) Naþ channel
sodium channel. Nature 475:353–358 subunits and Ig domains. Nature 383:307–308
10. McCusker EC, Bagneris C, Naylor CE, Cole 22. Calhoun JD, Isom LL (2014) The role of non-
AR, D’Avanzo N, Nichols CG, Wallace BA pore-forming beta subunits in physiology and
(2012) Structure of a bacterial voltage-gated pathophysiology of voltage-gated sodium
sodium channel pore reveals mechanisms of channels. Handb Exp Pharmacol 221:51–89
opening and closing. Nat Commun 3:1102 23. Bezanilla F (2008) How membrane proteins
11. Zhang X, Ren W, DeCaen P, Yan C, Tao X, sense voltage. Nat Rev Mol Cell Biol
Tang L, Wang J, Hasegawa K et al (2012) 9:323–332
Crystal structure of an orthologue of the 24. Yang N, George AL Jr, Horn R (1996) Molec-
NaChBac voltage-gated sodium channel. ular basis of charge movement in voltage-gated
Nature 486:130–134 sodium channels. Neuron 16:113–122
12. Stuhmer W, Conti F, Suzuki H, Wang XD, 25. Zhao Y, Scheuer T, Catterall WA (2004)
Noda M, Yahagi N, Kubo H, Numa S (1989) Reversed voltage-dependent gating of a bacte-
Structural parts involved in activation and inac- rial sodium channel with proline substitutions
tivation of the sodium channel. Nature in the S6 transmembrane segment. Proc Natl
339:597–603 Acad Sci U S A 101:17873–17878
13. West JW, Patton DE, Scheuer T, Wang Y, 26. Yang Y, Estacion M, Dib-Hajj SD, Waxman SG
Goldin AL, Catterall WA (1992) A cluster of (2013) Molecular architecture of a sodium
hydrophobic amino acid residues required for channel S6 helix: radial tuning of the voltage-
fast Na(þ)-channel inactivation. Proc Natl gated sodium channel 1.7 activation gate. J
Acad Sci U S A 89:10910–10914 Biol Chem 288:13741–13747
14. Patton DE, West JW, Catterall WA, Goldin AL 27. Shaya D, Findeisen F, Abderemane-Ali F,
(1992) Amino acid residues required for fast Arrigoni C, Wong S, Nurva SR,
Na(þ)-channel inactivation: charge neutraliza- Loussouarn G, Minor DL Jr (2014) Structure
tions and deletions in the III–IV linker. Proc of a prokaryotic sodium channel pore reveals
Natl Acad Sci U S A 89:10905–10909 essential gating elements and an outer ion
15. O’Malley HA, Isom LL (2015) Sodium chan- binding site common to eukaryotic channels.
nel beta subunits: emerging targets in channe- J Mol Biol 426:467–483
lopathies. Annu Rev Physiol 77:481–504 28. Chen LQ, Santarelli V, Horn R, Kallen RG
16. Meadows L, Malhotra JD, Stetzer A, Isom LL, (1996) A unique role for the S4 segment of
Ragsdale DS (2001) The intracellular segment domain 4 in the inactivation of sodium chan-
of the sodium channel beta 1 subunit is nels. J Gen Physiol 108:549–556
required for its efficient association with the 29. Patton DE, West JW, Catterall WA, Goldin AL
channel alpha subunit. J Neurochem (1993) A peptide segment critical for sodium
76:1871–1878 channel inactivation functions as an inactiva-
17. McCormick KA, Isom LL, Ragsdale D, tion gate in a potassium channel. Neuron
Smith D, Scheuer T, Catterall WA (1998) 11:967–974
Molecular determinants of Naþ channel func- 30. McPhee JC, Ragsdale DS, Scheuer T, Catterall
tion in the extracellular domain of the beta1 WA (1998) A critical role for the S4-S5 intra-
subunit. J Biol Chem 273:3954–3962 cellular loop in domain IV of the sodium chan-
18. Gilchrist J, Das S, Van Petegem F, Bosmans F nel alpha-subunit in fast inactivation. J Biol
(2013) Crystallographic insights into sodium- Chem 273:1121–1129
channel modulation by the beta4 subunit. Proc 31. Chandler WK, Meves H (1970) Sodium and
Natl Acad Sci U S A 110:E5016–E5024 potassium currents in squid axons perfused
19. Yu FH, Westenbroek RE, Silos-Santiago I, with fluoride solutions. J Physiol 211:623–652
McCormick KA, Lawson D, Ge P, Ferriera H, 32. Silva J (2014) Slow inactivation of Na(þ) chan-
Lilly J, DiStefano PS et al (2003) Sodium chan- nels. Handb Exp Pharmacol 221:33–49
nel beta4, a new disulfide-linked auxiliary
182 Gordon F. Tomaselli and Federica Farinelli

33. Zhou J, Shin HG, Yi J, Shen W, Williams CP, Ca/calmodulin kinase IIdelta is required for
Murray KT (2002) Phosphorylation and puta- late I(Na) augmentation leading to cellular Na
tive ER retention signals are required for pro- and Ca overload. Circ Res 108:555–565
tein kinase A-mediated potentiation of cardiac 45. Wagner S, Dybkova N, Rasenack EC,
sodium current. Circ Res 91:540–546 Jacobshagen C, Fabritz L, Kirchhof P, Maier
34. Smith RD, Goldin AL (1997) Phosphorylation SK, Zhang T et al (2006) Ca2þ/calmodulin-
at a single site in the rat brain sodium channel is dependent protein kinase II regulates cardiac
necessary and sufficient for current reduction Naþ channels. J Clin Invest 116:3127–3138
by protein kinase A. J Neurosci 17:6086–6093 46. Ma J, Luo A, Wu L, Wan W, Zhang P, Ren Z,
35. Li M, West JW, Numann R, Murphy BJ, Zhang S, Qian C, Shryock JC, Belardinelli L
Scheuer T, Catterall WA (1993) Convergent (2012) Calmodulin kinase II and protein
regulation of sodium channels by protein kinase C mediate the effect of increased intra-
kinase C and cAMP-dependent protein kinase. cellular calcium to augment late sodium cur-
Science 261:1439–1442 rent in rabbit ventricular myocytes. Am J
36. Hallaq H, Yang Z, Viswanathan PC, Fukuda K, Physiol Cell Physiol 302:C1141–C1151
Shen W, Wang DW, Wells KS, Zhou J, Yi J, 47. Koval OM, Snyder JS, Wolf RM, Pavlovicz RE,
Murray KT (2006) Quantitation of protein Glynn P, Curran J, Leymaster ND, Dun W et al
kinase A-mediated trafficking of cardiac sodium (2012) Ca2þ/calmodulin-dependent protein
channels in living cells. Cardiovasc Res kinase II-based regulation of voltage-gated
72:250–261 Naþ channel in cardiac disease. Circulation
37. Frohnwieser B, Chen LQ, Schreibmayer W, 126:2084–2094
Kallen RG (1997) Modulation of the human 48. Herren AW, Weber DM, Rigor RR, Margulies
cardiac sodium channel alpha-subunit by KB, Phinney BS, Bers DM (2015) CaMKII
cAMP-dependent protein kinase and the phosphorylation of NaV1.5: novel in vitro
responsible sequence domain. J Physiol 498 sites identified by mass spectrometry and
(Pt 2):309–318 reduced S516 phosphorylation in human
38. Hallaq H, Wang DW, Kunic JD, George AL Jr, heart failure. J Proteome Res 14:2298–2311
Wells KS, Murray KT (2012) Activation of pro- 49. Aiba T, Hesketh GG, Liu T, Carlisle R, Villa-
tein kinase C alters the intracellular distribution Abrille MC, O’Rourke B, Akar FG, Tomaselli
and mobility of cardiac Naþ channels. Am J GF (2010) Naþ channel regulation by Ca2þ/
Physiol Heart Circ Physiol 302:H782–H789 calmodulin and Ca2þ/calmodulin-dependent
39. Frohnwieser B, Weigl L, Schreibmayer W protein kinase II in guinea-pig ventricular myo-
(1995) Modulation of cardiac sodium channel cytes. Cardiovasc Res 85:454–463
isoform by cyclic AMP dependent protein 50. Aiba T, Barth AS, Hesketh GG, Hashambhoy
kinase does not depend on phosphorylation of YL, Chakir K, Tunin RS, Greenstein JL, Win-
serine 1504 in the cytosolic loop interconnect- slow RL et al (2013) Cardiac resynchronization
ing transmembrane domains III and therapy improves altered Na channel gating in
IV. Pflugers Arch 430:751–753 canine model of dyssynchronous heart failure.
40. Cantrell AR, Ma JY, Scheuer T, Catterall WA Circ Arrhythm Electrophysiol 6:546–554
(1996) Muscarinic modulation of sodium cur- 51. Ptacek LJ, George AL Jr, Griggs RC, Tawil R,
rent by activation of protein kinase C in rat Kallen RG, Barchi RL, Robertson M, Leppert
hippocampal neurons. Neuron 16:1019–1026 MF (1991) Identification of a mutation in the
41. Hammarstrom AK, Gage PW (1998) Inhibi- gene causing hyperkalemic periodic paralysis.
tion of oxidative metabolism increases persis- Cell 67:1021–1027
tent sodium current in rat CA1 hippocampal 52. Kumar D, Ambasta RK, Kumar P (2014)
neurons. J Physiol 510(Pt 3):735–741 Mutational consequences of aberrant ion chan-
42. Ju YK, Saint DA, Gage PW (1996) Hypoxia nels in neurological disorders. J Membr Biol
increases persistent sodium current in rat ven- 247:1083–1127
tricular myocytes. J Physiol 497 53. Waxman SG, Merkies IS, Gerrits MM,
(Pt 2):337–347 Dib-Hajj SD, Lauria G, Cox JJ, Wood JN,
43. Saint DA (2006) The role of the persistent Na Woods CG et al (2014) Sodium channel
(þ) current during cardiac ischemia and hyp- genes in pain-related disorders: phenotype-
oxia. J Cardiovasc Electrophysiol 17(Suppl 1): genotype associations and recommendations
S96–S103 for clinical use. Lancet Neurol 13:1152–1160
44. Wagner S, Ruff HM, Weber SL, Bellmann S, 54. Suetterlin K, Mannikko R, Hanna MG (2014)
Sowa T, Schulte T, Anderson ME, Grandi E Muscle channelopathies: recent advances in
et al (2011) Reactive oxygen species-activated
 NaV Channels - Review 183

genetics, pathophysiology and therapy. Curr 65. Angelides KJ, Elmer LW, Loftus D, Elson E
Opin Neurol 27:583–590 (1988) Distribution and lateral mobility of
55. Veerman CC, Wilde AA, Lodder EM (2015) voltage-dependent sodium channels in neu-
The cardiac sodium channel gene SCN5A and rons. J Cell Biol 106:1911–1925
its gene product NaV1.5: role in physiology 66. Garrido JJ, Giraud P, Carlier E, Fernandes F,
and pathophysiology. Gene 573:177–187 Moussif A, Fache MP, Debanne D, Dargent B
56. Tan BH, Iturralde-Torres P, Medeiros- (2003) A targeting motif involved in sodium
Domingo A, Nava S, Tester DJ, Valdivia CR, channel clustering at the axonal initial segment.
Tusie-Luna T, Ackerman MJ, Makielski JC Science 300:2091–2094
(2007) A novel C-terminal truncation SCN5A 67. Bennett V, Baines AJ (2001) Spectrin and
mutation from a patient with sick sinus syn- ankyrin-based pathways: metazoan inventions
drome, conduction disorder and ventricular for integrating cells into tissues. Physiol Rev
tachycardia. Cardiovasc Res 76:409–417 81:1353–1392
57. Medeiros-Domingo A, Kaku T, Tester DJ, 68. Aiba T, Farinelli F, Kostecki G, Hesketh GG,
Iturralde-Torres P, Itty A, Ye B, Valdivia C, Edwards D, Biswas S, Tung L, Tomaselli GF
Ueda K et al (2007) SCN4B-encoded sodium (2014) A mutation causing brugada syndrome
channel beta4 subunit in congenital long-QT identifies a mechanism for altered autonomic
syndrome. Circulation 116:134–142 and oxidant regulation of cardiac sodium cur-
58. Grant AO, Carboni MP, Neplioueva V, Starmer rents. Circ Cardiovasc Genet 7:249–256
CF, Memmi M, Napolitano C, Priori S (2002) 69. Makara MA, Curran J, Little SC, Musa H,
Long QT syndrome, Brugada syndrome, and Polina I, Smith SA, Wright PJ, Unudurthi SD
conduction system disease are linked to a single et al (2014) Ankyrin-G coordinates interca-
sodium channel mutation. J Clin Invest lated disc signaling platform to regulate cardiac
110:1201–1209 excitability in vivo. Circ Res 115:929–938
59. Lupoglazoff JM, Cheav T, Baroudi G, 70. Shy D, Gillet L, Ogrodnik J, Albesa M, Verkerk
Berthet M, Denjoy I, Cauchemez B, AO, Wolswinkel R, Rougier JS, Barc J et al
Extramiana F, Chahine M, Guicheney P (2014) PDZ domain-binding motif regulates
(2001) Homozygous SCN5A mutation in cardiomyocyte compartment-specific NaV1.5
long-QT syndrome with functional two-to- channel expression and function. Circulation
one atrioventricular block. Circ Res 89: 130:147–160
E16–E21 71. Agullo-Pascual E, Reid DA, Keegan S,
60. Kyndt F, Probst V, Potet F, Demolombe S, Sidhu M, Fenyo D, Rothenberg E, Delmar M
Chevallier JC, Baro I, Moisan JP, Boisseau P (2013) Super-resolution fluorescence micros-
et al (2001) Novel SCN5A mutation leading copy of the cardiac connexome reveals
either to isolated cardiac conduction defect or plakophilin-2 inside the connexin43 plaque.
Brugada syndrome in a large French family. Cardiovasc Res 100:231–240
Circulation 104:3081–3086 72. Agullo-Pascual E, Lin X, Leo-Macias A,
61. Bezzina C, Veldkamp MW, van Den Berg MP, Zhang M, Liang FX, Li Z, Pfenniger A, Lub-
Postma AV, Rook MB, Viersma JW, van Lan- kemeier I et al (2014) Super-resolution imag-
gen IM, Tan-Sindhunata G et al (1999) A sin- ing reveals that loss of the C-terminus of
gle Na(þ) channel mutation causing both connexin43 limits microtubule plus-end cap-
long-QT and Brugada syndromes. Circ Res ture and NaV1.5 localization at the intercalated
85:1206–1213 disc. Cardiovasc Res 104:371–381
62. Waxman SG (2006) Axonal conduction and 73. Hodgkin AL, Katz B (1949) The effect of
injury in multiple sclerosis: the role of sodium sodium ions on the electrical activity of giant
channels. Nat Rev Neurosci 7:932–941 axon of the squid. J Physiol 108:37–77
63. Long C, McAnally JR, Shelton JM, Mireault 74. Marmont G (1949) Studies on the axon mem-
AA, Bassel-Duby R, Olson EN (2014) Preven- brane; a new method. J Cell Physiol
tion of muscular dystrophy in mice by 34:351–382
CRISPR/Cas9-mediated editing of germline 75. Hodgkin AL, Huxley AF, Katz B (1949) Ionic
DNA. Science 345:1184–1188 currents underlying activity in the giant axon of
64. Long C, Amoasii L, Mireault AA, McAnally JR, the squid. Arch Sci Physiol 3:129–150
Li H, Sanchez-Ortiz E, Bhattacharyya S, Shel- 76. Cole KS (1949) Dynamic electrical character-
ton JM et al (2016) Postnatal genome editing istics of the squid axon membrane. Arch Sci
partially restores dystrophin expression in a Physiol 3:253–258
mouse model of muscular dystrophy. Science 77. Hamill OP, Marty A, Neher E, Sakmann B,
351:400–403 Sigworth FJ (1981) Improved patch-clamp
184 Gordon F. Tomaselli and Federica Farinelli

techniques for high-resolution current record- and adult fibroblast cultures by defined factors.
ing from cells and cell-free membrane patches. Cell 126:663–676
Pflugers Arch 391:85–100 81. Takahashi K, Tanabe K, Ohnuki M, Narita M,
78. Dykxhoorn DM, Lieberman J (2005) The Ichisaka T, Tomoda K, Yamanaka S (2007)
silent revolution: RNA interference as basic Induction of pluripotent stem cells from adult
biology, research tool, and therapeutic. Annu human fibroblasts by defined factors. Cell
Rev Med 56:401–423 131:861–872
79. Deschenes I, Armoundas AA, Jones SP, Toma- 82. Fertig N, Blick RH, Behrends JC (2002)
selli GF (2008) Post-transcriptional gene Whole cell patch clamp recording performed
silencing of KChIP2 and Navbeta1 in neonatal on a planar glass chip. Biophys J
rat cardiac myocytes reveals a functional associ- 82:3056–3062
ation between Na and Ito currents. J Mol Cell 83. Farre C, Fertig N (2012) HTS techniques for
Cardiol 45:336–346 patch clamp-based ion channel screening -
80. Takahashi K, Yamanaka S (2006) Induction of advances and economy. Expert Opin Drug Dis-
pluripotent stem cells from mouse embryonic cov 7:515–524
 Chapter 12

High-Content Electrophysiological Analysis of Human

Pluripotent Stem Cell-Derived Cardiomyocytes (hPSC-CMs)
Chi-Wing Kong, Lin Geng, and Ronald A. Li

Abstract
Considerable interest has been raised to develop human pluripotent stem cell-derived cardiomyocytes
(hPSC-CMs) as a model for drug discovery and cardiotoxicity screening. High-content electrophysiological
analysis of currents generated by transmembrane cell surface ion channels has been pursued to complement
such emerging applications. Here we describe practical procedures and considerations for accomplishing
successful assays of hPSC-CMs using an automated planar patch-clamp system.

Key words hPSC-CMs, High-content automated electrophysiology, Drug discovery, Cardiotoxicity

1 Introduction

Human pluripotent stem cells (hPSCs), including embryonic stem

cells (ESCs) and induced pluripotent stem cell (iPSCs), can self-
renew indefinitely while maintaining the ability to differentiate into
virtually all derivatives of the three embryonic germ layers
[1, 2]. These cells hold great promise as a potential unlimited cell
source for cell replacement therapies [3]. Patient-specific hiPSC
technologies have also been acknowledged for their use in human
disease modeling, and drug and toxicity screening platforms discov-
ery [2, 4–7]. Indeed, highly efficient protocols are available for
generating hPSC-CMs at high yield and purity [8–10], making
them superior to conventional aneuploidy cell lines such as the
Chinese hamster ovary (CHO) and human embryonic kidney
(HEK) cells that have been heterologously modified to express a
single cardiac ion channel type (e.g., HERG) as are currently being
used in the pharmaceutical industry. High-content analyses of
surface protein channels are in need to meet such emerging applica-
tions of the hPSC-CMs. To dissect the contribution of various
transmembrane cell surface ion channels, their biology and
responses to pharmacological reagents in hPSC-CMs, manual

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_12, © Springer Science+Business Media, LLC 2018

185
186 Chi-Wing Kong et al.

electrophysiological patch-clamp assay, which allows for practical

isolation of specific current via manipulation of electrical and phar-
macological conditions, has been considered as a gold standard
since its original development by Neher and Sakmann in 1980
[11]. However, the conventional technique is time-consuming,
labour-intensive, and highly dependent on skillful practitioners. As
such, automated patch-clamp platforms have been developed to
improve the assay throughput. Currently, applications are largely
restricted to CHO and HEK and a few other primary cells. In this
work, we describe practical procedures such as the preparation of
cell samples and loading density that we have developed for achiev-
ing high-content functional characterization of cardiac ionic cur-
rents in hPSC-CMs, with an automated planar patch clamp system
(PatchXpress 7000A, Molecular device). The CMs were differen-
tiated using a protocol that enables the generation of relatively pure
ventricular derivatives from various hPSC cell lines [10]. In this
chapter we describe the isolation of the cells (Subheading 3.1), the
preparation of a single cell suspension (Subheading 3.2) and the
automated patch clamping technique using a high throughput
apparatus (Subheading 3.3). Although the data obtained from car-
diomyocytes derived from a single hESC line are provided here as an
illustrative example, the same protocol has been tested on CMs
derived from other hPSC lines, with comparable performances.
The approach described here should also be applicable to other
cell types (e.g., neuronal cells) following standardization and testing
to determine optimal conditions for automated patch-clamp assays.

2 Materials

2.1 Human Human embryonic stem cell (hESC) line, HES2 (ESI, passages
Embryonic Stem Cell 35–55), was maintained in its pluripotent state and subsequently
Derived differentiated into cardiomyocytes with our published protocol for
Cardiomyocytes (CM) ventricular specification [10]. Although the differentiation proto-
col is not described here, the differentiated cardiospheres typically
contain >90% Troponin T type 2 (cardiac) (TNNT2+) cells that can
be isolated following enzymatic digestion.

2.2 hESC- 1. Fetal bovine serum (FBS).

Cardiosphere 2. PBSþ/þ (with calcium and magnesium).
Dissociation Solution
3. PBS/ (lacking calcium and magnesium).
4. 1 mg/mL Collagenase IV in 10% FBS containing PBSþ/þ,
filtered sterilized.
5. 10 mg/mL DNase I (Sigma) in 0.15 M NaCl.
6. 0.05% Trypsin–EDTA.
All enzymes are stored as frozen aliquots and freshly thawed
before use.
 High-content Electrophysiological Analyses 187

2.3 Isolated hESC- 1. Culture medium: high-glucose, DMEM-based medium sup-

CM Culture: Isolated plemented with 5% heat-inactivated fetal bovine serum, 1%
hESC-CMs Are MEM nonessential amino acid solution (100), and 1% Glu-
Cultured on Standard taMAX™. The culture medium was filtered sterilized and
Tissue Culture Treated stored at 4  C before use.
Surface 10-cm Dish 2. Trypsin–EDTA neutralizing medium: 20% FBS-containing
(Corning Life Science, DMEM-based medium.
Corning, MA, USA)

2.4 Patch Clamp All solutions were prepared in double distilled water and filtered to
Recording Solutions remove impurities right after preparation. For convenience, all
extra- and intra-cellular solutions can be stored as aliquots at
20  C for 3 months (see Note 1).
1. Sealing solution containing (in mM) 150 NaCl, 4 KCl, 1.2
CaCl2, 1 MgCl2, and 10 HEPES (4-(2-hydroxyethyl)-1-piper-
azineethanesulfonic acid). The pH was adjusted to 7.4 with
NaOH. Freshly prepared sealing solution can be stored at room
temperature, showing consistent performance when used
within 3 days (see Note 2).
2. Sodium current (INa): The extracellular solution contained
(in mM) 50 NaCl, 110 K-aspartate, 1.8 CaCl2, 1 MgCl2,
10 D-Glucose, 10 HEPES, and 0.001 Nifedipine. The pH
was adjusted to 7.4 with CsOH. The intracellular solution
contained (in mM): 135 CsCl, 10 NaCl, 2 CaCl2, 5 EGTA,
10 HEPES, and 5 MgATP, and the pH was adjusted to 7.2
with CsOH.
3. L-type Ca2+ current (ICaL): The extracellular solution
contained (in mM): 160 tetraethylammonium chloride
(TEA-Cl), 1 MgCl2, 5 CaCl2, 10 D-Glucose, and 10 HEPES.
The pH was adjusted to 7.4 with CsOH. The intracellular
solution contained (in mM): 145 CsCl, 5 NaCl, 2 CaCl2,
5 MgATP, 10 HEPES, and 5 EGTA, the pH was adjusted to
7.2 with CsOH. Tetrodotoxin (TTX, 0.01 mM) and
4-aminopyridine (4-AP, 2 mM) were included in the external
bath during recording.
4. Rapid delayed rectifier potassium current (IKr): The extracellu-
lar solution contained (in mM): 140 KCl, 15 NaCl, 1 MgCl2,
1.2 CaCl2, 0.002 Nifedipine, and 10 HEPES, and the pH was
adjusted to 7.4 with NaOH. The intracellular solution
contained (in mM): 110 K-aspartate, 20 KCl, 5 MgATP,
1 EGTA, 1 MgCl2, and 10 HEPES, 5 Na2-phosphocreatine,
and 0.1 NaGTP. The pH was adjusted to 7.4 with KOH.
188 Chi-Wing Kong et al.

2.5 SealChip 1. SealChip16 cartridges (AVIVA BioSciences Corporation, San

Diego CA, USA). These cartridges should be stored at 4  C
according to the manufacturer’s instructions (see Note 3).

2.6 General Supplies 1. 5 mL serological pipette.

and Equipment 2. 15 mL conical tubes.
3. Low speed centrifuge that can hold 15 mL conical tubes.
4. P1000 pipette.
5. Shaking water bath at 37  C or equivalent.
6. Low power microscope for observing cells.
7. 40-μm Nylon cell strainer.
8. 10-cm culture dish.
9. 5% CO2 compatible tissue culture incubator.
10. PatchXpress 7000A automated patch clamp system or similar.
11. PatchXpress Commander 2.0 (Molecular Devices, Sunnyvale,
CA, USA).
12. DataXpress software (Molecular Devices).
13. 1 μM E4031.

3 Methods

In the conventional manual patch-clamp method, cell samples are

usually attached on a matrix-supported cover glass. When using the
automated planar patch-clamp technique, cells in suspension are
placed in a recording chamber.

3.1 Isolation 1. Using a 5 mL serological pipette, transfer hESC-cardiospheres

of hESC-CMs from (25–30 days post-differentiation) in suspension to a 15 mL
hESC-Cardiospheres tube and let the clusters settle to the bottom.
2. Aspirate supernatant carefully and wash cardiospheres once
with PBSþ/þ.
3. Once the hESC-cardiospheres are settled to the bottom again,
aspirate PBSþ/þ carefully.
4. Add freshly thawed Collagenase IV solution supplemented
with 50 μg/mL DNase I (freshly thaw) to the cardiospheres.
Tap to mix. Shake the hESC-cardiospheres suspending in the
digestion mix at 37  C for 30 min.
5. After the Collagenase digestion step, hESC-cardiospheres
remain as clusters though the outline should appear “loos-
ened.” Allow hESC-cardiospheres to settle to the bottom
before removing the supernatant.
 High-content Electrophysiological Analyses 189

6. Wash the hESC-cardiospheres once with PBS/ and either

allow the cells to settle by gravity or spin at 100  g for 1 min,
then aspirate the supernatant.
7. Add 0.05% Trypsin–EDTA (freshly thaw from frozen) and
shake at 37  C for 5–7 min.
8. Triturate gently with a P1000 pipette to dissociate the clusters
into single-cell suspension. This usually takes about 20 pipet-
tings to accomplish. Visually verify the successful dissociation
of the majority of clusters with a microscope (the maximum
dwelling time in Trypsin–EDTA should be limited to 10 min).
9. Add 3 20% FBS-containing DMEM-based medium and mix
to stop the Trypsin-EDTA reaction.
10. Filter the cell suspension through a 40-μm cell strainer to
remove incompletely digested clusters and any sticky debris.
11. Spin at 300  g for 3 min, aspirate the supernatant.
12. Resuspend the isolated cells in hESC-CM culture medium and
seed cells onto a 10-cm culture dish at low density of ~0.3 M
cells per dish (see Note 4).
13. Refresh medium on the second day and then every other day
until day 3–5 post-seeding (see Note 5).

3.2 Preparation On day 3–5 post-seeding, the hESC-CMs on attachment culture

of Single hESC-CMs should be at a density as shown in Fig. 1. Right before automated
Suspension

Fig. 1 Isolated HES2-CMs cultured at low density (day-5 post-seeding) for

optimal automated patch clamp assay. A phase contrast image showing the
culture condition of the isolated HES2-CMs prepared with the procedures
described
190 Chi-Wing Kong et al.

patch-clamping, the hESC-CMs are detached with the following

procedure:
1. Wash the hESC-CMs culture with 6 mL PBS/ once.
2. Add 1.5 mL 0.05% Trypsin-EDTA (see Note 6) and incubate at
37  C.
3. Monitor the cells’ condition under microscope and stop the
Trypsin-EDTA reaction by adding 5 mL of 20%
FBS-containing, DMEM-based medium when ~90% cells
have been detached (see Note 7).
4. Gently transfer the cell suspension into a 15-mL tube.
5. Spin at 300  g for 3 min and aspirate supernatant.
6. Resuspend cells in 5 mL of hESC-CMs culture medium and
incubate cells at 37  C, 5% CO2 for 15–20 min recovery (see
Note 8).
7. Spin at 300  g for 3 min.
8. Remove the supernatant as completely as possible (see Note 9).
9. Resuspend hESC-CMs in 80 μL sealing solution (Cell density:
~3k/μL). Cells are ready for dispensing into the recording
chamber of the SealChip16 which has already been slowly equi-
librated to room temperature (see Note 10). Assay should be
performed as soon as possible (see Note 11).

3.3 Automated Patch The detailed operation procedures of the PatchXpress 7000A or
Clamping equivalent, and data analysis can be found in the user manual
provided by the manufacturer of the high-content electrophysiology
system employed by users. The PatchXpress 7000A automated patch
clamp system employed here is equipped with PatchXpress Com-
mander 2.0. All the patch settings are cell type-specific and require
empirical optimizations (see Note 12) as needed. Table 1 sum-
marizes performance indices such as the cell-detection rate, the
whole-cell rate, the average giga-seal resistance and membrane resis-
tance in this assay of HES2-CMs.
On average, our assays of cardiac ion currents in hESC-CMs
started ~20 min after cell recovery in the sealing solution.
1. Whole-cell currents were sampled at 10 kHz and low-pass
filtered at 2 kHz.
2. Data were transferred automatically into a database and can be
analyzed using the DataXpress software.
3. Figure 2 summarizes the characteristics of three ionic currents
(INa, ICaL, IKr) recorded in our assay of HES2-CMs.
4. The voltage protocols applied to probe for the three cardiac
ionic currents illustrated as examples are detailed as below:
 High-content Electrophysiological Analyses 191

Table 1
Summary of performance indices and cell parameters in our HES2-derived cardiomyocytes assay

Cell detection Whole-cell Current detection Giga-seal Membrane

rate (%) ratea (%) rate (%) resistance (MΩ) resistance (MΩ)
99 63 48 2605  118 2544  465
a
Five Sealchips were used in this set of experiment probing for INa, ICaL, IKr in the HES2-CMs prepared as mentioned in
the described protocol. 79 out of 80 channels (16 channels  5 chips) were available for use after the initial system run-in.
The giga-seal and membrane resistance are Mean  SEM of the cells achieved successful whole-cell condition, which is
defined as stable membrane resistance >300 MΩ and access resistance <10 MΩ.

a Voltage (mV)
1.0 1.0
Current density (pA/pF)

-120-100 -80 -60 -40 -20 20 40 60 0.8

-20

G/Gmax
-40

I/Imax
0.6
50 ms 0.5
50 mV -60
0.4
-80
-100 0.2
200 pA
–120 mV -120 0.0 0.0
1 ms
-140 -100 -80 -60 -40 -20 0
Voltage (mV)
b Voltage (mV)
1.0 1.0
Current density (pA/pF)

-40 -20 20 40 60 0.8 0.8

G/Gmax
0.6
I/Imax

0.6
-5
100 ms 0.4 0.4
60 mV
-10 0.2 0.2
200 pA 0.0 0.0
–50 mV 10 ms -15 -60 -40 -20 0 20
Voltage (mV)

Voltage (mV)
c
[K+]o 140mM/[Na+]o 15mM
Tail current density (pA/pF)

-150 -100 -50 50

IKr -10
500 ms
60 mV
-20

-60 mV -110 mV
-30
-160 mV

Fig. 2 Characteristics of three cardiac ionic currents (INa, ICaL, IKr) recorded in our automated patch clamp
assay of HES2-CMs. The fidelity of the currents recorded with the automated patch clamp assay described is
illustrated. Left: Representative (a) INa and (b) ICaL traces elicited by voltage protocol as shown in the inset.
Middle: peak I–V plots of the two currents. Right: Steady-state inactivation and activation relations. (n ¼ 7 for
both currents). (c) Left: representative IKr traces after subtraction of E4031-insensitive current, elicited by
voltage protocol as shown in the inset. Right: activation relation of IKr (n ¼ 11). Adapted from Weng et al.
(Reprinted with permission from STEM CELLS & DEVELOPMENT Volume 23, Issue 14, 2014, pp. 1704–1716,
published by Mary Ann Liebert, Inc., New Rochelle, NY)
192 Chi-Wing Kong et al.

INa: To elicit INa, the membrane potential was held at

120 mV and voltage steps from 120 to 50 mV in
10 mV-increments were applied for 50 ms. To measure
INa steady-state inactivation, 50 ms-long prepulses from
120 to 10 mV in 10 mV-increments, followed by
30 ms-long pulses of 10 mV were used to elicit INa.
ICaL: To elicit ICaL, the membrane potential was initially held at
60 mV. A 500 ms-long prepulse was applied to 50 mV
to voltage inactivate the Na+ channels and, if any, the
T-type Ca2+ channels. This step was followed by a
100 ms-long test pulse of between 50 and 60 mV in
10 mV-increments to activate the ICaL. To measure ICaL
steady-state inactivation, 500 ms-long prepulses to voltage
steps between 50 and 10 mV in 10 mV-increments fol-
lowed by 100 ms-long pulses to 0 mV to elicit ICaL. The
measured current amplitude was normalized to the peak
ICaL obtained with the 50 mV-prepulse and fitted to the
Boltzmann function.
IKr: To elicit the IKr, the membrane potential was held at
80 mV, a 1 s-long prepulse to between 150 and
60 mV in 10 mV-increments was applied, followed by a
110 mV pulse for 500 ms to elicit large amplitude inward
tail current that would gradually decayed. The IKr was
defined as E4031-sensitive (1 μM) current.

4 Notes

1. Freshly prepared, filtered-clarified solutions are best for consis-

tent assays in the automated system. For convenience, extra-
and intracellular solutions aliquoted and stored at 20  C also
provided reasonable performance within 3 months in our
hands. Degassing will also improve the assay consistency. Intra-
cellular solutions containing ATP must be thawed on ice to
slow decomposition of ATP. The thawed extra- and intra-
cellular solutions can be stored temporarily at 4  C for con-
sumption within 3 days. The solution from fridge should be
brought to room temperature slowly before use.
2. High success rate of giga-seal formation (84%, n ¼ 42/50) can
be achieved in our hands by using the sealing solution sug-
gested. Other extracellular solutions having good sealing per-
formance with manual patch-clamp may also be used
alternatively. It is a good practice to degas the sealing solution
to minimize the formation of tiny air bubbles. Degassing can
be performed on the day when the solution is prepared. The
sealing solution can be kept in a tightly capped container at
room temperature and consumed within 3 days for consistent
performance.
 High-content Electrophysiological Analyses 193

3. The SealChip16 are coated with a substrate designed for opti-

mized giga-seal formation. Their giga-seal-forming perfor-
mance declines quickly after expiration (3 months after
dispatch according to the manufacturer). 10 min ultrasonica-
tion of the expired SealChip16 in its original container can
partially recover their performance. Reuse of the SealChip16 is
not recommended.
4. Isolation of hPSC-cardiospheres into single hPSC-CMs for
attachment culture is a crucial step allowing recovery of the
cardiomyocytes obtained from the enzymatic treatment. The
recovery improves giga-seal formation during automated patch
clamp assay. The single cells isolated should be cultured at low
density (~4k/cm2) to avoid formation of cardiac muscle patch
before automated patch assay (Fig. 1).
5. Additional wash during daily refreshment helps remove cell
debris or cells that have been damaged by the enzymatic treat-
ment during cardiospheres dissociation. A hESC-CM suspen-
sion containing debris-free single cells mainly are very
important for successful giga-seal formation.
6. Other commercially available cell dissociation reagents of dif-
ferent dissociation mechanisms or strength, including Accu-
tase, TrypLE, and 0.5 mM EDTA, had been attempted but
were found to adversely affect giga-seal formation.
7. The detachment of hESC-CMs right before an assay is critical.
To detach the replated hESC-CMs after 3–5 days recovery on
attachment culture, usually 4.0  0.5 min, 0.05% Trypsin/
EDTA treatment was sufficient. Over-digestion or excessive
mechanical trituration of the cells remaining attached dramati-
cally affects the successfulness of assay.
8. The success rate of giga-seal formation declines dramatically if
the recovery time is >30 min for our hESC-CMs.
9. The hESC-CMs culture medium contains 5% FBS, which was
found to reduce the success rate of giga-seal formation if not
adequately removed.
10. The recording chambers are prone to tiny gas bubble forma-
tion if the surrounding temperature increases too quickly. The
existence of gas bubbles will render the recording channel
nonfunctional.
11. The hPSC-CMs suspended in the sealing solution should be
dispensed into the recording chamber within 2–4 min (~9k
cells per chamber) and then the operator should proceed with
the assay immediately. Prolonged delay will reduce the success
rate of giga-seal formation.
12. The manufacturer has provided in their manual sample settings
for optimized patch performance of a few established cell lines.
194 Chi-Wing Kong et al.

Several patch parameter settings optimized for our assay of

various hPSC-CMs, with similar success rates, are provided as
initial guidelines for reference: (a) During cell detection, cell
attraction pressure was set at 45 mmHg; (b) pressure ramps
were applied at 2 mmHg/s for 10 s while peak pressure was
set at 20 mmHg for a hold duration of 1 s, the interval
between pulses was 12 s during sealing; (c) pressure ramps
was applied at 5 mmHg/s for 20 s while peak pressure was
set at 100 mmHg for a hold duration of 1 s, the interval
between pulses was set at 6 s during whole-cell formation.

Acknowledgements

This work was supported by the Theme-based Research Scheme

[T13-706/11] of Hong Kong Research Grant Council.

References
1. Thomson JA, Itskovitz-Eldor J, Shapiro SS, in human pluripotent stem cells. BioEssays 35
Waknitz MA, Swiergiel JJ, Marshall VS, Jones (3):281–298
JM (1998) Embryonic stem cell lines derived 7. Li RA (2012) The use of induced pluripotent
from human blastocysts. Science 282 stem cells for disease modeling: what are the
(5391):1145–1147 promises and hurdles? Drug Discov Today Dis
2. Takahashi K, Tanabe K, Ohnuki M, Narita M, Model 9(4):e143–e145
Ichisaka T, Tomoda K, Yamanaka S (2007) 8. Burridge PW, Thompson S, Millrod MA,
Induction of pluripotent stem cells from adult Weinberg S, Yuan X, Peters A et al (2011) A
human fibroblasts by defined factors. Cell 131 universal system for highly efficient cardiac dif-
(5):861–872 ferentiation of human induced pluripotent
3. Fox IJ, Daley GQ, Goldman SA, Huard J, stem cells that eliminates interline variability.
Kamp TJ, Trucco M (2014) Stem cell therapy. PLoS One 6(4):e18293
Use of differentiated pluripotent stem cells as 9. Lian X, Zhang J, Azarin SM, Zhu K, Hazeltine
replacement therapy for treating disease. Sci- LB, Bao X et al (2012) Directed cardiomyocyte
ence 345(6199):1247391 differentiation from human pluripotent stem
4. Yu J, Vodyanik MA, Smuga-Otto K, cells by modulating Wnt/beta-catenin signal-
Antosiewicz-Bourget J, Frane JL, Tian S et al ing under fully defined conditions. Nat Protoc
(2007) Induced pluripotent stem cell lines 8(1):162–175
derived from human somatic cells. Science 10. Weng Z, Kong CW, Ren L, Karakikes I,
318(5858):1917–1920 Geng L, He J et al (2014) A simple, cost-
5. Belmonte JC, Ellis J, Hochedlinger K, Yama- effective but highly efficient system for deriving
naka S (2009) Induced pluripotent stem cells ventricular cardiomyocytes from human plurip-
and reprogramming: seeing the science otent stem cells. Stem Cells Dev 23
through the hype. Nat Rev Genet 10 (14):1704–1716
(12):878–883 11. Hamill OP, Marty A, Neher E, Sakmann B,
6. Rajamohan D, Matsa E, Kalra S, Crutchley J, Sigworth FJ (1981) Improved patch-clamp
Patel A, George V, Denning C (2013) Current techniques for high-resolution current record-
status of drug screening and disease modelling ing from cells and cell-free membrane patches.
Pflugers Arch 391(2):85–100
 Chapter 13

Methods for Evaluation of Vascular Endothelial Cell

Function with Transient Receptor Potential (TRP) Channel
Drugs
Yung Wui Tjong and Xiaoqiang Yao

Abstract
Vascular endothelial transient potential (TRP) channels, located mostly on the plasma membrane of cells,
are critical in regulatory and pathophysiological circumstances. The objective of this chapter is to describe
several well-established approaches, ranging from function to molecular assays, to investigate the mecha-
nistic role of TRP channels in vascular endothelial cells. We show experimental procedures and representa-
tive figures on the following methods: (1) Isolation and culture of vascular endothelial cells,
(2) examination of electrophysiological activity of TRP channel by patch-clamping with whole-cell config-
uration and its function in vascular tone and blood flow by isometric tension and isobaric diameter
measurements, and Laser Doppler flowmetry, (3) detection of TRP channel-mediated intracellular Ca2+
imaging by using fluorescent microscopy, and (4) determination of TRP channel interaction by coimmu-
noprecipitation, double immunofluorescence staining and Förster resonance energy transfer (FRET)
detection.

Key words Electrophysiology, Intracellular calcium concentration, Protein interaction, Transient

receptor potential channel, Vascular endothelial cell

1 Introduction

Transient receptor potential (TRP) channel is a cation channel

serving essential regulatory roles in diverse physiological functions
including thermosensation [1], signal transduction [2], homeosta-
sis, etc. [3]. Mammalian TRP channel can be subdivided into six
main transmembrane protein families according to amino acid
sequence homology and functions: TRPC (canonical), TRPV
(vanilloid), TRPA (ankyrin), TRPML (mucolipin), and TRPP
(polycystin) [4]. They have common primary structural features
that are composed of six putative transmembrane domains with a
hydrophobic cation-permeable pore region between domains
5 (S5) and 6 (S6), and both carboxyl and amino intracellular

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_13, © Springer Science+Business Media, LLC 2018

195
196 Yung Wui Tjong and Xiaoqiang Yao

loops [5]. Although most TRP channels are nonselective to cations,

the permeability ratio varies among individual members [6]. With
the exception of TRPV5 and TRPV6, most of Ca2+ permeable TRP
channels are poorly selective to Ca2+. TRPM4 and TRPM5 are only
permeable to monovalent ions [6].
TRP channels are abundant in endothelium involved in the
regulation of vascular tone, permeability, angiogenesis and other
functions [7]. Dysfunction of endothelial TRP channels, in con-
trast, may be a causative factor contributing to several cardiovascu-
lar diseases such as hypertension, atherosclerosis, myocardial
infraction and heart failure [8–10]. The function of TRP channel
in vascular endothelial cells can be studied by several techniques
including patch clamping, fluorescent cytosolic Ca2+ measurement,
isometric vessel and isobaric diameter measurement. Patch clamp-
ing technique is commonly used to examine the electrophysiologi-
cal properties of functional channels [11]. The TRP-mediated
pathway in the regulation of blood flow in vascular endothelium
can be determined by Laser Doppler flowmetry [12] and patch-
clamp technique. TRP channel is partly involved in Ca2+ influx and
Ca2+-related signaling cascades play a key role in vascular contrac-
tion/relaxation. Fluorescent probes enable the quantitative analysis
of changes in [Ca2+]i concentration [13]. Isometric tension and/or
isobaric diameter of blood vessels in vitro using wire myograph and
pressure myograph instruments, respectively can be used to exam-
ine the role of TRP channel in the regulation of vascular tone.
Functional TRP channels may form as homodimers or hetero-
dimers [13, 14]. TRP channels may also interact with other pro-
teins to form signaling complexes that are involved in the
regulation of vascular function [15]. The channel–protein interac-
tion can be determined by coimmunoprecipitation that utilizes an
antibody to target a bait protein. The antigen–antibody complex is
then bound to protein A agarose, while irrelevant proteins are
washed out. The target protein in the bait protein complex can be
determined by western blotting [16]. The protein interaction can
also be determined by binding specific fluorescent dye-labeled
antibodies to their specific targets in the cell and their distribution
can be visualized under fluorescent microscopy. The third method
for studying the protein-protein interaction is Förster resonance
energy transfer (FRET). Its principle based on a distance-
dependent interaction between the electronic excited states of
two chromophores, is useful method to study protein-protein
interaction [17]. Cyan fluorescent protein (CFP)–yellow fluores-
cence protein (YFP) pair is one of the most popular donor–acceptor
pairs for this biological approach [18].
 Evaluating Vascular EC Function with TRP Channel Drugs 197

2 Materials

2.1 Instrumentation 1. Whole-cell patch clamping: EPC-10 patch amplifier (HEKA,

Lambrecht/Pfalz), P-97 micropipette puller (Sutter Instru-
ment, Novato, CA), Microscope (Nikon Eclipse TS100,
Tokyo). Software: PulseFit (HEKA, Lambrecht/pfalz).
2. Fluorescence measurement of [Ca2+]i: FV1000 confocal imaging
system (Olympus, Tokyo, Japan), Inverted microscope (Olym-
pus IX81, Olympus, Tokyo, Japan). Software: MetaFluor Ana-
lyst (Molecular Devices, Sunnyvale, CA). Calcium Calibration
Buffer Kit (Molecular Probes, Eugene, OR, USA).
3. Laser Doppler flowmetry: Laser Doppler perfusion imager
(amoorFLPI full-field image, Moor Instruments, Devon, UK).
4. Isometric tension measurement: Thin stainless steel holders
(DMT, Aarhus, Denmark), Isometric tension myograph
(Model 610M, DMT, Aarhus, Denmark). Software: Powerlab
(AD instrument, Sydney, Australia) and LabChart
(AD Instruments, Sydney, Australia).
5. Isobaric diameter measurement: Pressure myograph chamber
with charge-coupled device CCD camera (Model 110P, Danish
Myotechnology, Aarhus, Denmark). Software: Myoview
(GE Healthcare, Piscataway, NJ, USA).
6. Protein concentration determination: Lowry Assay for protein
quantification (Thermo Scientific, Rockford, IL, USA), Epoch
Microplate Spectrophotometer (BioTek, Winooski, VT, USA).
7. Protein electrophoresis: Mini-PROTEIN Tetra System
(Bio-Rad, Hercules, CA, USA) equipped with a Power Pac™
Basic Power Supply (Bio-Rad, Hercules, CA, USA). 2
con-
centrate Laemmli sample buffer (Sigma Chemical Company,
St. Louis, MO, USA).
8. Protein transfer from within SDS-PAGE gel on a PVDF mem-
brane: Trans-Blot SD Semi-Dry Transfer Cell (Bio-Rad, Her-
cules, CA, USA).
9. Protein bands detection reagent: ECL western blotting detec-
tion reagent (GE Healthcare, Pittsburgh, PA, USA).
10. Protein bands detection: FluroChem 8000 system (ProteinSim-
ple, Santa Clara, CA, USA).
11. FRET detection: Olympus IX 81 microscope (Olympus, Tokyo,
Japan) equipped with a CCD camera and three-cube FRET
filter set including (excitation; dichoric; emission): YFP
(S500/20 nm; Q515lp; S535/30 nm); FRET (S430/25 nm;
455dclp; S535/30 nm); and CFP (S430/25 nm; 455dclp;
S470/30 nm) (Olympus, Tokyo, Japan).
198 Yung Wui Tjong and Xiaoqiang Yao

2.2 Cell Culture 1. Endothelial basic medium with 0.02% collagenase IA: Dilute
Components 2 mg of collagenase IA (Sigma-Aldrich) in 10 mL of endothe-
lial basic medium (Lonza, Walkersville, MD, USA).
2. Endothelial growth medium with 1% bovine brain extract:
Dilute 0.05 g bovine brain extract in 5 mL of endothelial
growth medium and add 50 μL of 100
Penicillin–Streptomy-
cin (100 U/mL penicillin and 100 μg/mL streptomycin)
(Invitrogen, Carlsbad, CA, USA).
3. PBS: 80 mM of Na2HPO4, 20 mM of NaH2PO4, 100 mM of
NaCl. Weigh reagents and transfer to the beaker and followed
by adding water up to 900 mL. Mix and adjust pH with HCl to
pH 7.4. Transfer the solution to the 1-L graduated cylinder
and make up to 1 L with H2O and store at 4  C.
4. Reagents and equipment: 25-cm2 tissue culture flasks, CO2
tissue culture incubator.

2.3 Whole-Cell Patch 1. Pipette solution: 120 mM of CsCl, 1 mM of CaCl2, 5 mM of

Clamp Components MgCl2, 5 mM of Na2ATP, 10 mM of TEA (Tetraethylammo-
nium), 11 mM of EGTA (Ethylene glycol tetraacetic acid),
10 mM of HEPES. Add 500 mL of water to a 1-L glass beaker.
Weigh reagents and transfer to the beaker followed by the
addition of water up to 900 mL. Mix and adjust with HCl to
pH 7.3. Transfer the solution to the 1-L graduated cylinder
and make up to 1 L with H2O and store at 4  C.
2. Bath solution: 135 mM of NaCl, 5 mM of CsCl, 2 mM of
CaCl2, 1 mM MgCl2, 10 mM of HEPES, and 10 mM of
glucose. Weigh the reagents and prepare a 1-L solution as in
the previous step and store at 4  C.

2.4 Fluorescent 1. Fluroscent dyes (or probes): Prepare 0.02% pluronic F127 by
Measurement adding 0.2 mg of pluronic 127 to 1 mL of H2O.
of Intracellular 2. Normal physiological salt solution (NPSS): 140 mM NaCl
Calcium Concentration (Sigma-Aldrich), 1 mL KCl (Sigma-Aldrich), 1 mM of CaCl2
([Ca2+]i) (Sigma-Aldrich), 1 mM of MgCl2 (Sigma-Aldrich), 10 mM
glucose (Sigma-Aldrich), and 5 mL HEPES (Sigma-Aldrich).
Add 500 mL of water to a 1-L glass beaker. Weigh reagents and
transfer to the beaker followed by the addition of water up to
900 mL. Mix and adjust with HCl to pH 7.4. Transfer the
solution to the 1-L graduated cylinder and make up to 1 L with
H2O and store at 4  C.

2.5 Laser Doppler 1. Krebs solution: 126 mM of NaCl (Sigma-Aldrich), 0.25 mM of

Flowmetry KCl (Sigma-Aldrich), 0.25 mM of NaH2PO4 (Sigma-Aldrich),
Components 0.12 mM of NaH2PO4 (Sigma-Aldrich), 0.12 mM of MgCl2
(Sigma-Aldrich), 0.25 mM CaCl2 (Sigma-Aldrich). Weigh
reagents and transfer to a beaker followed by the addition of
 Evaluating Vascular EC Function with TRP Channel Drugs 199

water up to 900 mL. Mix and adjust with HCl to pH 7.4.

Transfer the solution to the 1-L graduated cylinder and make
up to 1 L with H2O and store at 4  C.

2.6 Isometric 1. Krebs–Henseleit solution: 118 mM NaCl (Sigma-Aldrich),

Tension Measurement 4.7 mM KCl (Sigma-Aldrich), 2.5 mM CaCl2 (Sigma-Aldrich),
Components 1.2 mM KH2PO4 (Sigma-Aldrich), 1.2 mM MgSO47H2O
(Sigma-Aldrich), 25.2 mM NaHCO3 (Sigma-Aldrich), and
11.1 mM glucose (Sigma-Aldrich). Add 500 mL of water to a
1-L glass beaker. Weigh reagents and transfer to the beaker
followed by the addition of water up to 900 mL. Mix and
adjust with HCl to pH 7.4. Transfer the solution to the 1-L
graduated cylinder and make up to 1 L with H2O and store at
4  C.
2. 60 mM K+ solution: 58 mM NaCl (Sigma-Aldrich), 64.7 mM
KCl (Sigma Adrich), 2.5 mM CaCl2 (Sigma-Aldrich), 1.2 mM
KH2PO4 (Sigma-Aldrich), 1.2 mM MgSO47H2O (Sigma-
Aldrich), 25.2 mM NaHCO3 (Sigma-Aldrich) and 11.1 mM
glucose (Sigma-Aldrich). Weigh reagents and transfer to a
beaker followed by the addition of water up to 900 mL. Mix
and adjust pH with HCl to pH 7.4. Weigh the reagent and
prepare a 1-L solution as in previous step and store at 4  C.

2.7 Western Blotting 1. Cell lysis buffer: 50 mM Tris–HCl (Sigma-Aldrich), 150 mM

and Coimmuno- NaCl (Sigma-Aldrich), 50 mM NaF (Sigma-Aldrich), 1.5%
precipitation Tergitol-type NP-40 (Sigma-Aldrich), 0.5% sodium deoxycho-
Components late (Sigma-Aldrich), and cOmplete™ Protease Inhibitor
Cocktail tablet (Roche). Weigh reagents and transfer to the
beaker followed by the addition of water up to 900 mL. Mix
and adjust pH with HCl to pH 7.5. Transfer the solution to the
1-L graduated cylinder and make up to 1 L with H2O and store
at 4  C.
2. Low assay reagents for protein concentration calibration: Protein
samples were added with Reagent A, B and S (DC™ Protein
assay reagent, Bio-Rad Laboratories) and protein concentra-
tion determined following the manufacturer’s instructions. In
brief, 3–5 dilutions of a protein standard are prepared from
0.2 g/mL to 1.5 mg/mL protein. Both standards and protein
samples are added with reagent A. Add reagent B into each
samples and vortex immediately. Record the absorbance at
750 nm.
3. PBS: 80 mM of Na2HPO4, 20 mM of NaH2PO4, 100 mM of
NaCl. Weigh reagents and transfer to the beaker and followed
by the addition of water up to 900 mL. Mix and adjust pH with
HCl to pH 7.4. Transfer the solution to the 1-L graduated
cylinder and make up to 1 L with H2O and store at 4  C.
200 Yung Wui Tjong and Xiaoqiang Yao

4. PBST: 80 mM of Na2HPO4, 20 mM of NaH2PO4, 100 mM of

NaCl, 0.1% of Tween 20. Weigh reagents and transfer to a
beaker followed by the addition of water up to 900 mL.
Weigh the reagent and prepare a 1-L solution as in previous
step and store at 4  C.
5. Resolving gel buffer: 1.5 mM of Tris–HCl, 2% of APS (ammo-
nium persulfate), 7.5% Sodium dodecyl sulfate (SDS), and N,
N,N0 ,N0 -Tetramethylethylenediamine (TEMED). Mix reagent
with water and adjust pH to pH 8.8.
6. Stacking gel buffer: 1.0 M of Tris–HCl, 2% APS, 10% SDS, and
TEMED. Mix reagents with water and adjust pH to 6.8.
7. Running buffer: 25 mM of Tris–HCl, 250 mM of glycine, and
0.1% of SDS. Weigh reagents and transfer to a beaker followed
by the addition of water up to 900 mL. Mix and adjust to
pH 8.3. Transfer the solution to the 1-L graduated cylinder
and make up to 1 L with H2O and store at 4  C.
8. Transfer buffer: 1.25 mM of Tris–HCl, 192 mM of glycine,
10% (v/v) methanol. Weigh reagents and transfer to a beaker
followed by the addition of water up to 900 mL. Weigh the
reagent and prepare a 1-L solution as in previous step and store
at 4  C.
9. ECL western blotting detection reagents (GE Healthcare).
10. Protein A agarose suspension (Roche).

3 Methods

All animal experimental procedures should be approved by local

animal regulatory authorities and should abide by the US National
Institute of Health regulatory guidelines found in NIH publication
No. 8523.

3.1 Preparation 1. Prepare 10 mL of endothelial basic medium and endothelial

of Rat Mesenteric growth medium (see Subheading 2.2, items 1 and 2).
Arterial Endothelial 2. Sprague-Dawley rat (ca. 250–300 g) is anesthetized. A toe
Cells (MAECs) Culture pinch was applied to each anesthetized rat to determine
whether the withdrawal reflex was present. This strain of rat is
commonly used for animal experiments. A toe pinch is a usual
practice to examine the intensity or stage of anesthetic effect on
rat; the abdomen is dissected and the heart perfused with PBS
solution (see Subheading 2.2, item 3) to remove circulating
blood from blood vessels by cannulation of the aorta.
3. Dissect and remove the small intestine and other viscera and
excise all the vein branches.
 Evaluating Vascular EC Function with TRP Channel Drugs 201

4. The remaining arterial branches are digested with 10 mL of

0.02% collagenase in endothelial basic medium for 45 min at
37  C.
5. Centrifuge at 1600
g for 5 min at room temperature.
6. Resuspend pelleted cells in 5 mL of endothelial growth
medium.
7. Place the resuspended cells in a 25-cm2 culture flask.
8. Remove the nonadherent cells after 1 h.
9. Culture the adherent cells at 37  C in a 5% CO2 humidified
incubator.

3.2 Patch Clamp 1. Prepare pipette solution and bath solution (see Subheading 2.3,
with Whole-Cell items 1 and 2).
Configuration 2. Seed the cells on the coverslips.
3. Place a coverslip with adherent cells into the special flow cham-
ber that is exposed to steady laminar flow with the flow rate at
2 mL/min. The distance between the flow inlet and the cells is
maintained constant. Shear stress is estimated at the range of
0.5–1 dyn/cm2.
4. Prepare glass micropipette by pulling it with P-97 micropipette
puller (see Subheading 2.1, item 1). Fill the recording micropi-
pette with bath solution.
5. Bring the recording micropipette onto the cell until the pipette
tip is in contact with the cell membrane.
6. Apply the suction through the pipette pressure tubing to make
a gigaseal on the cell membrane (see Note 1).
7. Attain the whole-cell configuration by rupturing the mem-
brane patch within the micropipette. Compensate cell capaci-
tance and series resistance using the controls on the amplifier.
8. Apply successive voltage pluses from 80 mV to þ80 mV for
100 ms duration and record the whole cell current density
(pA/pF) by using EPC-10 patch clamp amplifier (Fig. 1).
9. The whole-cell current data is analysed with PulseFit software
(see Subheading 2.1, item 1) (Fig. 1).

3.3 Fluorescence 1. Prepare 0.02% of pluronic F127 and normal physiological salt
Measurement solution (NPSS) (see Subheading 2.4, items 1 and 2).
of Intracellular 2. Seed the cells on the coverslips and load with 10 mM of the
Calcium Concentration fluorescence probe (see Subheading 2.4, item 2) for 30 min
([Ca2+]i) with NPSS solution in the dark at 37  C.
3. Place the coverslip with cells into the recording chamber filled
with NPSS solution and place the recording chamber on the
stage of an inverted microscope (see Subheading 2.1, item 2).
202 Yung Wui Tjong and Xiaoqiang Yao

Fig. 1 Flow-induced whole-cell current in HEK293 cells coexpressing TRPV4,

TRPC1, and TRPP2. Representative trace for time course of flow-stimulated whole
cell current of primary cultured rat MAECs before (resting) and after flow [19]

Fig. 2 Representative traces for the potentiation of flow-induced Ca2+ influx in

human umbilical vein endothelial cells [20]. [Ca2]i increase by TG in HUVECs. TG,
thapsigargin, 4 μmol/L, was given for 15 min; BFA, brefeldin A, 5 μmol/L, was
given for 30 min before TG

4. Record the [Ca2+]i fluorescence using a confocal system at

excitation wavelength of 488 nm or a fluorescence imaging
system at excitation wavelength 340 and 380 nm (see Subhead-
ing 2.1, item 2).
5. Analyze the data by Fluoview FV1000 software or MetaFluor
Analyst software (see Subheading 2.1, item 2). Changes in [Ca2
+
]i are indicated by ratio of the fluorescence intensity relative to
the value before stimulation (flow or chemical challenge
(Fig. 2) or the changes of the ration of the fluorescence
under 340 nm relative to the fluorescence under 380 nm
(F340/F380) (Fig. 3).
6. Convert the Fura-2 ratio F340/F380 to [Ca2+]i based on the
calibration using calcium calibration buffer kit, if necessary
(Fig. 3).
 Evaluating Vascular EC Function with TRP Channel Drugs 203

400 BK (200 nM)

300 Ca2+ (2 mM)

[Ca2+]i (nM)
200

100

0
0 400 800 1200 1600 2000
Time (seconds)

Fig. 3 Bradykinin-induced Ca2+ entry [21]. A representative trace of Fura-

2 fluorescence in rat aortic endothelial cells bathed in 0Ca2+-PSS in response
to bradykinin (200 nM) challenge

3.4 Laser Doppler 1. SD rats (c.a. 280–320 g) are anaesthetized with ketamine
Flowmetry (35 mg/kg) plus xylazine (7 mg/kg).
2. After performing a midline laparotomy, approximately
two-thirds of the rat mesenteric arterial bed is gently placed
into a petri-dish chamber and bathed in Krebs solution.
3. The local blood perfusion of the rat mesenteric arterial bed is
assessed with a Laser Doppler perfusion system (see Subheading
2.1, item 3) (Fig. 4).
4. The acquisition was made in high resolution mode with 5 s
interval. The pixel resolution of image is 760
568. The
digital color-coded images are analyzed to quantify blood
flow in the region from mesenteric vascular beds.
5. Blood pressure of rats is simultaneously monitored through a
pressure transducer inserted in common carotid artery. When
needed, the bathing solution is changed to ones that contain
4 μM 4α-PDD (Fig. 4).

3.5 Isometric 1. Sacrifice the male C57BL mice (~5 weeks old) by cervical
Tension Measurement dislocation.
2. Isolate the thoracic aorta and place it into an ice-chilled Kreb-
s–Henseleit solution (see Subheading 2.6, item 1) bubbled
with 95% O2 and 5% CO2 gas mixture.
3. Remove the fat and peripheral tissues under a dissection
microscope.
4. Cut the aorta into 2 mm segments.
5. Mount the aortic rings onto two thin stainless steel holders
(supplied by isometric tension myograph; see Subheading 2.1,
204 Yung Wui Tjong and Xiaoqiang Yao

Fig. 4 Role of TRPV4-KCa2.3 pathway in the control of local blood flow in mesenteric bed ex vivo. Representa-
tive images of Laser Doppler studies in rat mesenteric arteries from the fourth-order to the end in response to
4α-PDD

item 4) in 5 mL organ baths containing Krebs–Henseleit solu-

tion bubbled with 95% O2 and 5% CO2 at 37  C.
6. Equilibrate the aortic rings for about 30 min (see Note 2).
7. Add 60 mM K+ solution to test the contractile function of the
aortic rings (see Subheading 2.6, item 2).
8. Remove the bath solution by washing with Krebs–Henseleit
solution twice.
9. Repeat steps 7 and 8.
10. Preconstrict the aortic rings with 10 μM phenylephrine to
achieve a sustained contraction.
11. Wash out with Krebs–Henseleit solution for three times.
12. Repeat step 10.
13. Add cumulative chemicals into the bath solution to test its
relaxation effect.
14. Acquire and analyze data by PowerLab and LabChart (see
Subheading 2.1, item 4) (Fig. 5).

3.6 Isobaric 1. Sacrifice the Sprague-Dawley rats by inhalation of CO2.

Diameter 2. Remove the ileum and immerse the mesentery in Krebs–Hen-
Measurement seleit solution (see Subheading 2.6, item 1) bubbled with 95%
O2 and 5% CO2.
 Evaluating Vascular EC Function with TRP Channel Drugs 205

SNAP

100nM 300nM 1µM 300µM

10µM Phe

2.5 mN

5 min

Fig. 5 Representative time courses of isometric tension in isolated mouse aortic

segments in response to cumulatively increasing concentrations of SNAP applied
to the bath [22]. The aortic segments (~2 mm in length) were preconstricted
with 10 μM phenylephrine (Phe). SNAP S-nitroso-N-acetylpenicillamine

3. Dissect the third- or fourth-order mesenteric artery

(~2–3 mm long).
4. Transfer the artery to a pressure myograph chamber filled with
oxygenated Krebs–Henseleit solution at 37  C (see Subheading
2.1, item 5).
5. Cannulate one glass micropipette (tip diameter ~125 μm) into
the proximal part and the other into the distal end of the artery
and secure with two fine nylon sutures.
6. Connect both cannulation pipettes to independent reservoirs
set at the same height and solution level to ensure there is
no flow.
7. Set the intraluminal pressure to 50 mmHg, and equilibrate the
artery in oxygenated Krebs–Henseleit solution for 30 min at
37  C (flow rate ~2–3 mL/min).
8. Pressurize the artery to 80 mmHg, and apply a longitudinal
force to stretch the vessel until it appears straight and then by
an extra 10%.
9. Reduce the pressure down to 50 mmHg, and incubate the
vessel for another 10 min prior to experimentation.
10. Monitor the artery by a charge-coupled device camera (video
camera module) attached to a light inverted microscope (see
Subheading 2.1, item 5).
11. Analyze the external diameter of the vessel and luminal pres-
sure by MyoView software (see Subheading 2.1, item 5)
(Fig. 6).

3.7 Western Blotting 1. Calibrate protein samples to equal amounts based on the stan-
dard curve obtained by Lowry assay (see Subheading 2.1, item 6).
206 Yung Wui Tjong and Xiaoqiang Yao

a Phe b Phe
Flow
Vessel diameter (µm) 400

Vessel diameter (µm)

400 Flow
300
300
200
200
100 5 min
2 min
ACh 100
0 ACh

Preimmune treated T1E3 treated

Fig. 6 Representative traces showing the effect of T1E3, a TRPC1 blocking antibody, on flow-induced vascular
isobaric diameter changes in isolated mice mesenteric arteries [13]. Arteries were preincubated with
preimmune IgG (1:50) or T1E3 (1:50) overnight. The solid bar on the top of the trace indicates the period
when intraluminal flow (Krebs solution with 1% BSA) was applied. The arteries were preconstricted with
phenylephrine (Phe)

2. Prepare a SDS-polyacrylamide gel. Pour resolving gel buffer to

the glass plate and allow to polymerize (see Subheading 2.7,
item 8). Pour stacking gel buffer to remaining space between
the glass plates and insert comb and let polymerize (see Sub-
heading 2.7, items 9 and 10).
3. Equal amount of proteins are mixed with Laemmli sample
buffer, incubated at 90  C for 5 min and loaded onto the gel
with ~20 mg of protein per lane of a polyacrylamide gel.
Proteins are separated on 7.5% SDS polyacrylamide gel with
100 V constant voltage (see Subheading 2.1, item 7). Run until
the front of the bromophenol dye present in the buffer is
approximately 1 cm from the gel bottom (50–90 min).
4. Prewet a PVDF (polyvinylidene-difluoride) membrane in
methanol and transfer buffer (see Subheading 2.7, item 11).
The gels were transferred to the PVDF membrane using Trans-
blot SD semidry electrophoretic transfer cell for 30 min (see
Subheading 2.1, item 8).
5. Immerse the PVDF membrane in a blocking solution contain-
ing 5% nonfat milk and 0.1% Tween 20 in PBS at room tem-
perature with constant shaking for 1 h (see Note 3).
6. The membranes are incubated with primary antibody (1:500)
overnight at 4  C.
7. Rinse the membrane three times for 5 min each with 15 mL of
PBST to remove unbound primary antibody (see Subheading
2.7, item 7).
 Evaluating Vascular EC Function with TRP Channel Drugs 207

8. Incubate the membrane with secondary antibody (1:1000 or as

recommended by the company) conjugated to horseradish
peroxidase at room temperature for 1 h.
9. Rinse the membrane three times for 5 min each with 15 mL
PBST to remove any unbound secondary antibody.
10. Incubate with ECL western blotting detection reagents for
5 min at room temperature (see Subheading 2.1, item 9).
11. Expose the membrane to X-ray film.
12. Intensity of the protein blotting bands is detected by
FluorChem 8000 system (see Subheading 2.1, item 10).

3.8 Coimmuno- 1. Incubate extracted protein sample (~800 μg) with 50 μL of

precipitation protein A agarose suspension (see Subheading 2.7, item 13)
(see Note 4) and incubate at 4  C on a rocking platform for 3 h.
2. Pellet agarose beads by centrifuging in a microcentrifuge
(12,000
g) at 4  C for 2 min, followed by transferring
supernatant to a fresh tube.
3. Add 7 mg of pulling antibody (or preimmune IgG as negative
control (see Note 5) to the sample and incubate at 4  C on a
rocking platform for 2 h.
4. Add 100 μL protein A agarose suspension to the mixture an
incubate overnight at 4  C on rocking platform.
5. Centrifuge (12,000
g) at 4  C for 2 min and collect agar-
ose–antibody–antigen complexes. Discard the supernatant.
6. Resuspend the pellet in 1 mL of lysis buffer (see Subheading 2.7,
item 1) and incubate for 30 min at 4  C on a rocking platform.
7. Pellet the beads again and discard supernatant.
8. Repeat steps 6 and 7 twice.
9. Resuspend pellet in 25 μL of gel-loading buffer.
10. Denature proteins by heating the sample at 95  C for 5 min.
11. Centrifuge the suspension (12,000
g) at 4  C for 2 min.
12. Analyze the supernatant by gel electrophoresis and western
blotting (see Subheadings 2.1, item 9 and 3.6) (Fig. 7).

3.9 Förster 1. Seed cells on uncoated coverslips.

Resonance Energy 2. Transfect the following fusion proteins into cultured mamma-
Transfer (FRET) lian cells:
Detection Component (a) CFP fused to YGP as positive control.
(b) Unfused, free CFP and unfused, free YFP as negative
control.
(c) Protein 1-CYP and Protein 2-YFP.
(d) Protein 1-YFP and Protein 2-CFP.
208 Yung Wui Tjong and Xiaoqiang Yao

a b
IP: pre-immune Anti-C1 IP: pre-immune Anti-V4
IB: Anti-V4 Anti-V4 IB: Anti-C1 Anti-C1
KDa KDa
148- 148-

98-
98-

Fig. 7 Representative images of coimmunoprecipitation followed by

immunoblots in primary cultured MAECs [13]. The pulling and blotting
antibodies are indicated. Control immunoprecipitation was performed using
the preimmune IgG (labeled as preimmune). Anti-C1 indicates anti-TRPC1;
anti-V4, anti-TRPV4; IB, immunoblot; IP, immunoprecipitation

3. Place the coverslips with the cells into a chamber 12–24 h after
transfection.
4. Mount the chamber on an inverted microscope equipped with
a CCD camera and three-cube FRET filter set.
5. Subtract the average background signal.
6. Capture the fluorescence images of the transfected cells at
CFP-, YFP-, and FRET-channels respectively.
7. Calculate the FRET ratio (FR) (Fig. 8) by the equation shown
as follows:
FR ¼ FAD/FA ¼ [SFRET(DA)  RD1
SCFP(DA)]/
RA1
[SYFP(DA)  RD2
SCFP(DA)]
where FAD represents the total YFP emission with 430/425-
nm excitation, and FA represents the direct YFP emission with
500/520-nm excitation. In SCUBE(SPECIMEN), CUBE indi-
cates the filter cube (CFP, YFP, or FRET), and SPECIMEN
indicates whether the cell is expressing donor (D, CFP), accep-
tor (A, YFP), or both (DA). RD1 ¼ SFRET(D)/SCFP(D),
RD2 ¼ SYFP(D)/SCFP(D), and RA1 ¼ SFRET(A)/SYFP(A) are
predetermined constants that require measurement of the
bleed-through of the emission of only CFP- or YFP-tagged
molecules into the FRET channel and the emission of only
CFP-tagged molecules into the YFP channel.

4 Notes

1. In patch clamp, the initial seal membrane resistance must

achieve 1 GΩ.
 Evaluating Vascular EC Function with TRP Channel Drugs 209

Fig. 8 FRET detection for the interaction between TRPV4 and TRPC1. Horizontal
axes indicate FRET ratio of living cells expressing the indicated constructs
[13]. Each point represents the FRET ratio of a single cell. The red lines and
error bars indicate the average FRET ratio values and SE. When the FRET ratio is
1, there is no FRET; when the FRET ratio is greater than 1, there is FRET. Data are
given as mean  SE (n ¼ 50–81). GIRK G-protein-activated inwardly rectifying
K+ channels

2. In myograph, determine the passive tension carefully as too

higher tension may cause the vessel ring injury.
3. Always wear gloves when handling the PVDF membrane, oth-
erwise it may be contaminated or damaged.
4. Preimmune IgG is served as control if antibodies are used for
experiments.
5. Check the affinities of protein A agarose for various IgG sub-
classes before use. Consider other kind of protein agarose (e.g.,
protein G agarose) if the binding capacity of protein A agarose
is low for some certain species.

Acknowledgments

This work was supported by grants from the Hong Kong Research
Grant Committee TBRS T13-706/11, AoE/M-05/12,
CUHK478413 and by the China National Science Foundation
31470912.
210 Yung Wui Tjong and Xiaoqiang Yao

References
1. Vay L, Gu C, McNaughton PA (2012) The 12. Minamitani H, Okada E (1993) Microscopic
thermo-TRP ion channel family: properties laser Doppler velocimeter measuring blood
and therapeutic implications. Br J Pharmacol velocity in single microvessel. Keio J Med
165:787–801 42:186–190
2. Hecquet CM, Zhang M, Mittal M, Vogel SM, 13. Ma X, Qiu S, Luo J, Ma Y, Ngai CY et al (2010)
Di A et al (2014) Cooperative interaction of trp Functional role of vanilloid transient receptor
melastatin channel transient receptor potential potential 4-canonical transient receptor poten-
(TRPM2) with its splice variant TRPM2 short tial 1 complex in flow-induced Ca2þ influx.
variant is essential for endothelial cell apopto- Arterioscler Thromb Vasc Biol 30:851–858
sis. Circ Res 114:469–479 14. Kobori T, Smith GD, Sandford R, Edwardson
3. Brayden JE, Earley S, Nelson MT, Reading S JM (2009) The transient receptor potential
(2008) Transient receptor potential (TRP) channels TRPP2 and TRPC1 form a heterote-
channels, vascular tone and autoregulation of tramer with a 2:2 stoichiometry and an alter-
cerebral blood flow. Clin Exp Pharmacol nating subunit arrangement. J Biol Chem
Physiol 35:1116–1120 284:35507–35513
4. Holzer P, Izzo AA (2014) The pharmacology 15. Shen B, Cheng KT, Leung YK, Kwok YC,
of TRP channels. Br J Pharmacol Kwan HY et al (2008) Epinephrine-induced
171:2469–2473 Ca2þ influx in vascular endothelial cells is
5. Hellmich UA, Gaudet R (2014) Structural mediated by CNGA2 channels. J Mol Cell Car-
biology of TRP channels. Handb Exp Pharma- diol 45:437–445
col 223:963–990 16. Masters SC (2004) Co-immunoprecipitation
6. Owsianik G, Talavera K, Voets T, Nilius B from transfected cells. Methods Mol Biol
(2006) Permeation and selectivity of TRP 261:337–350
channels. Annu Rev Physiol 68:685–717 17. Herman B, Krishnan RV, Centonze VE (2004)
7. Kwan HY, Huang Y, Yao X (2007) TRP chan- Microscopic analysis of fluorescence resonance
nels in endothelial function and dysfunction. energy transfer (FRET). Methods Mol Biol
Biochim Biophys Acta 1772:907–914 261:351–370
8. Mathar I, Vennekens R, Meissner M, Kees F, 18. Latif R, Graves P (2000) Fluorescent probes:
Van der Mieren G et al (2010) Increased cate- looking backward and looking forward. Thy-
cholamine secretion contributes to hyperten- roid 10:407–412
sion in TRPM4-deficient mice. J Clin Invest 19. Du J, Ma X, Shen B, Huang Y, Birnbaumer L
120:3267–3279 et al (2014) TRPV4, TRPC1, and TRPP2
9. Lo CY, Tjong YW, Ho JC, Siu CW, Cheung SY assemble to form a flow-sensitive heteromeric
et al (2014) An upregulation in the expression channel. FASEB J 28:4677–4685
of vanilloid transient potential channels 20. Ma X, Cao J, Luo J, Nilius B, Huang Y et al
2 enhances hypotonicity-induced cytosolic Ca (2010) Depletion of intracellular Ca2þ stores
(2)(þ) rise in human induced pluripotent stem stimulates the translocation of vanilloid tran-
cell model of Hutchinson-Gillford Progeria. sient receptor potential 4-c1 heteromeric chan-
PLoS One 9:e87273 nels to the plasma membrane. Arterioscler
10. Entin-Meer M, Levy R, Goryainov P, Landa N, Thromb Vasc Biol 30:2249–2255
Barshack I et al (2014) The transient receptor 21. Leung PC, Cheng KT, Liu C, Cheung WT,
potential vanilloid 2 cation channel is abundant Kwan HY et al (2006) Mechanism of
in macrophages accumulating at the peri- non-capacitative Ca2þ influx in response to
infarct zone and may enhance their migration bradykinin in vascular endothelial cells. J Vasc
capacity towards injured cardiomyocytes fol- Res 43:367–376
lowing myocardial infarction. PLoS One 9: 22. Wong CO, Sukumar P, Beech DJ, Yao X
e105055 (2010) Nitric oxide lacks direct effect on
11. Hamill OP, Marty A, Neher E, Sakmann B, TRPC5 channels but suppresses endogenous
Sigworth FJ (1981) Improved patch-clamp TRPC5-containing channels in endothelial
techniques for high-resolution current record- cells. Pflugers Arch 460:121–130
ing from cells and cell-free membrane patches.
Pflugers Arch 391:85–100
 Chapter 14

Methods to Study the Signal Transduction of the Surface

Receptor Tyrosine Kinase TrkB in Neurons
Kwok-On Lai and Nancy Y. Ip

Abstract
Receptor tyrosine kinases (RTK) belong to a major class of cell surface receptors that transduce extracellular
signals to elicit diverse intracellular responses. Upon binding to specific ligand, the RTKs become dimerized
and autophosphorylated at tyrosine residues. This creates binding sites to recruit specific signaling inter-
mediates and hence trigger distinct signaling events. The cellular response to a given RTK may be modified
through the regulation of membrane insertion and receptor internalization. Here we use Trk receptor and
its ligand, the neurotrophin brain-derived neurotrophic factor (BDNF), as an example to illustrate the
approaches (coimmunoprecipitation and biotinylation) to study the surface expression and signal transduc-
tion mediated by this class of RTK in the nervous system.

Key words Kinase, Neuron, Phosphorylation, Signal transduction, Surface expression

1 Introduction

Neurotrophin is a family of homologous proteins that play pivotal

roles in multiple processes of nervous system development and
function. Neurotrophins are dimeric, basic proteins that are pro-
duced as precursors (the proneurotrophins), but are proteolytically
cleaved to generate mature proteins that contain about 120 amino
acids. The prototypic member of the neurotrophin family, called
nerve growth factor (NGF), was discovered and purified by Rita
Levi-Montalcini in the 1950s. Since then, three more members of
the family, including brain derived neurotrophic factor (BDNF),
neurotrphin-3 (NT-3), and neurotrophin-4 (NT-4), were isolated
[1]. Additional members (NT-6 and NT-7) were identified in fish,
although their corresponding orthologues are not found in
mammals [2].
As their names imply, neurotrophins were first demonstrated to
act as a target-derived trophic factor for the survival of neurons in
the peripheral nervous system. According to the neurotrophic
hypothesis, the amount of innervation a target receives is

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_14, © Springer Science+Business Media, LLC 2018

211
212 Kwok-On Lai and Nancy Y. Ip

determined by the amount of trophic factors released by the target

itself. However, it has now become clear that neurotrophins are
involved in multiple functions at different stages of neural develop-
ment, and are essential in both the central and peripheral nervous
systems. In addition to being a survival factor, neurotrophins also
promote neuronal differentiation, synapse formation and matura-
tion during brain development. Multiple lines of evidence establish
BDNF as a key growth factor that mediates synaptic plasticity in the
adult brain, which is crucial for cognitive functions such as learning
and memory [3, 4]. Because of their importance in neuronal sur-
vival and synaptic function, there is enormous interest in develop-
ing small molecules that exert neurotrophin function as a potential
therapeutic agent against neurological diseases [5].

1.1 Neurotrophin Neurotrophins elicit their effects on neurons through binding to

Receptors (NTR) two distinct classes of surface receptors [6]. The low affinity recep-
tor p75NTR, which belongs to the tumor necrosis factor receptor
superfamily, preferentially binds to the proneurotrophins while it
interacts with the mature neurotrophins only with low affinity
(dissociation constant of about l0
9 M [7]). Although its intracel-
lular domain does not contain catalytic activity, p75NTR can inter-
act with specific signaling proteins that mediate the effect of the
receptor upon activation by the neurotrophins. On the other hand,
receptor tyrosine kinases called “Trk” offer the high affinity bind-
ing sites (dissociation constant of about 10
11 M) for neurotro-
phins [8]. The Trk receptor is structurally similar to the receptors of
other growth factors such as fibroblast growth factor (FGF) and
platelet derived growth factor (PDGF). Upon binding to the neu-
rotrophin, the Trk receptors become dimerized and auto-
phosphorylated. Three Trk receptors have been identified, and
each neurotrophin exhibits receptor specificity towards a particular
Trk receptor: NGF is the preferential ligand of TrkA; BDNF and
NT-4 both activate TrkB; and NT-3 mainly acts on TrkC. Different
Trk receptors are expressed in different neurons, and mice lacking
each of the Trk receptors display different phenotypes in the ner-
vous system. This indicates that the four neurotrophins can indeed
perform nonredundant functions through interaction with differ-
ent Trk receptors.
Many functions of the mature forms of the neurotrophins are
mediated by the Trk receptor, while the p75NTR may play a
modulatory role in regulating neurotrophin binding to Trk recep-
tor (p75NTR may act by itself independently of the Trk receptor
after binding to the proneurotrophins, such as during the induction
of neuronal death). The remaining part of the chapter focuses on
the Trk receptor, and briefly summarizes its signal transduction
pathways and the regulation of its surface expression. For more
detailed signal transduction of Trk receptor, readers may refer to
some excellent reviews on this topic [6, 9].
 Surface Receptor Tyrosine Kinase TrkB in Neurons 213

1.2 Signal After binding to the neurotrophin, Trk undergoes phosphorylation

Transduction to trigger multiple signaling pathways. There are ten evolutionarily
of Neurotrophins: The conserved tyrosine residues in the Trk receptor intracellular
Trk Receptors domain. The two most well-studied tyrosine residues that undergo
ligand-dependent phosphorylation are Y490 and Y785 of human
TrkA (equivalent to Y515 and Y816 of mouse TrkB), which create
docking sites after auto-phosphorylation for recruitment of specific
signaling molecules. As a consequence of Y490 and Y785 phos-
phorylation, three major signaling cascades can be activated: the
Ras-ERK pathway, the PI3K-AKT pathway, and the PLCγ-Ca2+
pathway.
Phosphorylation of Trk receptor at Y490 in the juxtamembrane
region recruits binding and phosphorylation of the adaptor protein
Shc, which in turn triggers the activation of Ras-ERK pathway via
another adaptor protein Grb2 and the Ras exchange factor son of
sevenless (SOS). Phosphorylated Y490 on Trk receptor also
recruits another adaptor, Frs2. This leads to prolonged activation
of ERK through the GTPase Rap1 and the protein kinase B-Raf. A
second signaling cascade triggered after Ras activation is the PI3K-
AKT pathway. Activation of PI3K leads to generation of
3-phosphoinositides, which activates the kinases PDPK1 and
AKT. Activation of the ERK pathway is crucial for neuronal differ-
entiation, whereas the PI3K-AKT pathway promotes neuronal sur-
vival. One possible explanation for their differential actions is that
the two signaling pathways may signal to the nucleus to regulate
transcription of distinct subsets of genes [6]. In addition to the
control of gene expression, PI3K can regulate protein synthesis
through the mTOR signaling pathway. PI3K and Ras also activate
the Rho GTPase Rac1, which is involved in actin cytoskeleton
dynamics and is crucial in cellular events such as axon guidance or
dendritic spine growth and formation.
Phosphorylation of Y785 near the C-terminus of Trk receptor
triggers a different signaling cascade, which is initiated by the recruit-
ment and phosphorylation of PLC-γ1. The activated PLC-γ1 hydro-
lyzes phosphatidylinositol-4,5-bisphosphate to produce
diacylglycerol (DAG) and inositol-1,4,5-trisphosphate (IP3), which
activates PKC and increases Ca2+ concentration through internal
stores, respectively. The increased Ca2+ level also activates various
Ca2+/Calmodulin-dependent protein kinases (CaMKs). One of
them is CaMKIV, which phosphorylates the transcription factor
CREB to regulate gene transcription for long-lasting synaptic plastic-
ity. Mice harboring a mutation of TrkB at Y816 (tyrosine substituted
to phenylalanine and therefore becomes phosphorylation-deficient in
that particular site), which is equivalent to Y785 of TrkA, have
impaired PLC-γ1 signaling. As a result, the mice exhibit defects in
long-term potentiation (LTP), a learning-related form of synaptic
plasticity, as well as deficits in hippocampus-dependent learning and
memory [10, 11]. Interestingly, LTP in the hippocampus of Y515F
214 Kwok-On Lai and Nancy Y. Ip

mutant mice (which is equivalent to Y490 of TrkA) is not affected.

Studies on these site-specific phosphorylation-deficient knock-in
mice therefore demonstrate that different signaling cascades down-
stream of the Trk receptor are involved in specific cellular events in
neurons.
In addition to the well-studied autophosphorylation of tyrosine
residues, certain RTKs have been known to undergo serine/threo-
nine phosphorylation after ligand binding [12]. However, the
physiological significance of the serine/threonine phosphorylation
of RTK is not clear. Upon BDNF binding, TrkB activates the
proline-directed serine/threonine kinase Cdk5, which then phos-
phorylates TrkB at S418 in the juxtamembrane region
[13]. Through the generation of the serine phosphorylation-
deficient knock-in mice (TrkB-S478A), it was found that TrkB
S478 phosphorylation is crucial for regulating synaptic function
in the adult brain. The formation and maturation of dendritic
spines induced by BDNF is abolished in TrkB-S478A hippocampal
neurons, which also show reduced dendritic spine enlargement
after glutamate stimulation, indicating its importance in activity-
dependent structural plasticity of the synapse [14]. BDNF induces
tyrosine phosphorylation of the Rac1 guanine nucleotide exchange
factor TIAM1 that is crucial for actin cytoskeleton dynamics, and
the interaction and phosphorylation of TIAM1 by TrkB depends
on its S478 phosphorylation [14, 15]. As a result, neurons lacking
TrkB-S478 phosphorylation show impaired Rac1 activation and
phosphorylation of the Rac1 effector PAK after BDNF or NMDA
receptor activation. Notably, LTP and spatial memory in the Morris
water maze is impaired in the TrkB-S478A knock-in mice. There-
fore, activation of Rac1 by TrkB requires not only BDNF-induced
tyrosine phosphorylation but also Cdk5-mediated serine phosphor-
ylation of the receptor.

1.3 Regulation of Trk The number of RTK proteins expressed on cell surface is a crucial
Receptor Surface determinant of the cellular responsiveness to its cognate ligand, and
Expression the surface expression of RTK can be regulated at the level of
membrane insertion and receptor internalization. RTK internaliza-
tion is an important event following ligand binding, and can affect
ligand function by down-regulating the receptor signaling and/or
interaction with specific signaling proteins in distinct cellular com-
partments. For example, it was found that enhancing the rate of
NGF-TrkA internalization or disrupting TrkA endocytosis can dif-
ferentially affect the trophic and differentiation action of NGF on
PC12 cells [16, 17]. Interestingly, both the membrane insertion
and BDNF-induced internalization of TrkB is facilitated by neuro-
nal activity [18, 19]. This suggests that the response to BDNF is
greater in active neurons or synapses, and is consistent with the
 Surface Receptor Tyrosine Kinase TrkB in Neurons 215

notion that BDNF-TrkB signaling is important for activity-

dependent synaptic plasticity. The increased insertion of TrkB by
neuronal activity depends on its phosphorylation at Ser-478 by
Cdk5 [20], although the basal surface expression of TrkB is not
affected in neurons lacking Ser-478 phosphorylation [14].
In the following sections, we will describe the protocols to
examine biochemically the BDNF-induced phosphorylation of sig-
naling proteins (Subheading 3.1). Moreover, coimmunoprecipita-
tion detects the recruitment of specific signaling proteins to the Trk
receptor in cells (Subheading 3.2) and in tissue lysates (Subheading
3.3) (Fig. 1). The surface expression of Trk receptors in neuron can
be determined biochemically using biotinylation (Subheading 3.4).
Although the methods mainly use primary dissociated cortical
neurons [21], these protocols can be generally applied to study
other RTKs in neuronal and nonneuronal cells.

Fig. 1 Schematic diagram illustrating the coimmunoprecipitation of TrkB and the signaling protein TIAM1, a
guanine nucleotide exchange factor (GEF) for the small GTPase Rac1. Upon treatment of cortical neurons with
BDNF, TIAM1 is recruited to TrkB and this interaction depends on phosphorylation of TrkB at Ser-478 by the
serine/threonine kinase Cdk5 [14]. The interaction between TrkB and TIAM1 can be demonstrated by
immunoprecipitation by anti-TIAM1 antibody. After cell lysis in a mild detergent, the associated TrkB can
be coimmunoprecipitated with TIAM1 by the antibody. After captured by the Protein A sepharose beads, the
association between the two proteins is disrupted by elution with sample buffer and boiling. TrkB and TIAM1
will be separated by SDS-PAGE based on difference in molecular weight, and Western blotting with antibodies
against TrkB and TIAM1 will detect two different bands. In cortical neurons carrying a S478A mutation in
which the serine residue is substituted to alanine and hence deficient in phosphorylation, TIAM1 fails to
interact with TrkB, and the receptor will therefore not be captured by the TIAM1 antibody and Protein A
Sepharose beads. As a result, the band corresponding to TrkB will be absent in the Western blot
216 Kwok-On Lai and Nancy Y. Ip

2 Materials

2.1 Reagents 1. Medium for neurons: Primary cortical neurons were cultured
in Neurobasal medium (ThermoFisher Scientific, Carlsbad,
CA, USA) supplemented with 2% B27 (ThermoFisher Scien-
tific), 1 mM L-glutamine (ThermoFisher Scientific) and 10 mM
D-glucose (Sigma-Aldrich Chemical Company, St. Louis, MO,
USA). HEK-293T cells were cultured in Minimal Essential
Medium (MEM) (ThermoFisher Scientific) þ 10% fetal bovine
serum (FBS, ThermoFisher Scientific).
2. Dulbecco’s phosphate-buffered saline (DPBS) (ThermoFisher
Scientific).
3. Bovine serum albumin (BSA; Sigma-Aldrich).
4. Brain-derived neurotrophic factor (BDNF; PeproTech, Rocky
Hill, NH, USA).
5. Lipofectamine plus (ThermoFisher Scientific).
6. Bradford reagent (Bio-Rad, Hercules, CA, USA).
7. Buffer A for coimmunoprecipitation: 20 mM Tris, 50 mM
NaCl, 1 mM EDTA, 1 mM NaF, 0.5% Nonidet P-40
(NP-40) (v/v). A 2 solution is prepared.
8. Radioimmunoprecipitation assay (RIPA) buffer: 12 mM
sodium deoxycholate, 0.1% SDS; 1% NP-40 in Dulbecco’s
phosphate-buffered saline (DPBS). A 2 solution was
prepared, which was stored at 4  C. Dilute to 1 by DPBS
before use.
9. Protease and phosphatase inhibitors (purchased from Sigma-
Aldrich except sodium fluoride):
(a) Soybean trypsin inhibitor.
(b) Leupeptin.
(c) Aprotinin.
(d) Antipain.
(e) Okadaic acid.
(f) Benzamidine.
(g) Sodium orthovanadate.
(h) Phenylmethylsulfonyl fluoride (PMSF).
(i) Sodium fluoride. (A-Tech Global Science Limited, Hong
Kong, SAR China).
(j) Beta-glycerol phosphate.
10. To make RIPA plus protease and phosphatase inhibitors, pre-
pare the RIPA buffer with the following concentrations of
inhibitors:10 μg/mL soybean trypsin inhibitor, 10 μg/mL
leupeptin, 10 μg/mL aprotinin, 2 μg/mL antipain, 30 nM
 Surface Receptor Tyrosine Kinase TrkB in Neurons 217

okadaic acid, 5 mM benzamidine, 1 mM sodium orthovana-

date, 1 mM PMSF, 1 mM sodium fluoride, 100 mM beta-
glycerophosphate.
11. SDS gel mix component (component volume depends on the
percentage of gel required):
(a) Acrylamide–Bis-acrylamide solution, 40% (w/v) (Bio
Basic Canada Inc., Markham, ON, CAN).
(b) 3 M Tris–HCl pH 8.9 (separating)/pH 6.8 (stacking)
(Affymetrix, Santa Clara, CA, USA).
(c) 20% sodium dodecyl sulfate (SDS) solution.
(d) 20% ammonium persulfate (APS) solution (Sigma-
Aldrich).
(e) Temed (Sigma-Aldrich).
12. Prestained protein marker (Cell Signaling Technology, Dan-
vers, MA, USA).
13. SDS sample buffer (5) was used for eluting proteins after
immunoprecipitation: [300 mM Tris–HCl buffer (pH 6.8),
10% SDS (w/v), 25% beta-mercaptoethanol (v/v), 50% glyc-
erol (v/v), 0.05% bromophenol blue (w/v)]. Aliquots were
stored at
20  C.
14. Wash buffer for Western blot (TBST): 20 mM Tris, 0.14 M
NaCl, 0.1% Tween 20, pH 7.6.
15. ECL (Thermo Scientific).
16. Sulfo-NHS-LC-Biotin kit (ThermoFisher Scientific).
17. Streptavidin beads (ThermoFisher Scientific).
18. Protein A&G Sepharose beads (GE Healthcare Life Sciences,
Marlborough, MA, USA).
19. Tris–HCl, 50 mM, pH 7.4 (prepare 1 M stock solution and use
sodium hydroxide to adjust the pH).

2.2 Tissue Culture 1. 293T (ATCC), primary rat cortical neurons (see Note 1).
Supplies 2. 60 and 100 mm tissue culture dishes.
3. Bel-Art Cell scrapers (VWR, Randor, PA, USA).
4. Eppendorf tubes (1.5 mL).

2.3 Antibodies Commercially available antibodies used were: TrkB (Western blot-
ting, BD Biosciences); TIAM1, phospho-S6K-Ser411 (Santa
Cruz); Phospho-TrkA (Y490), phospho-S6K-Thr-389, phospho-
S6-S235/236, S6K, S6 ribosomal protein were from Cell Signaling
Technology; secondary antibodies for Western blotting were horse-
radish peroxidase-conjugated goat antibodies to rabbit or mouse
and were purchased from Cell Signaling Technology.
218 Kwok-On Lai and Nancy Y. Ip

2.4 Equipment 1. CO2 Tissue culture incubator (NuAire, Plymouth, MN, USA).
2. Refrigerated microcentrifuge with speed up to 16,000  g.
3. Equipment to run SDS-PAGE gels (Gel caster: GE Healthcare
Life Sciences, Marlborough, MA, USA; SDS gel tank: Hoefer
Inc., Hollison, MA, USA; Protein transfer: Bio-Rad).
4. Cold room holding a Laboratory rocker (use cold room instead
of refrigerator).
5. Laboratory rocker.
6. Wheaton overhead stirrer with a 15 mL tissue grinder.
7. Spectrophotometer or microplate reader measuring absor-
bance at 595 nm.
8. pH meter.
9. Power Supply (similar to a MINI-300 or equivalent).
10. Weighing scale.

3 Methods

3.1 Phosphorylation 1. Primary rat cortical neurons (1.5  106 cells) were cultured
of Signaling Proteins onto 60 mm culture dishes. BDNF treatment was performed at
in Cortical Neurons 13–14 DIV (see Note 1).
2. Aspirate the medium, add 4 mL serum-free neurobasal
medium to each dish and incubate at 37  C for 2 h. This
starvation can lower the baseline phosphorylation of the recep-
tor and signaling proteins and make it easier to see the induc-
tion by BDNF.
3. Dilute BDNF in 0.1% BSA (in DPBS) to 100 ng/μL. Add 4 μL
BDNF or 0.1% BSA (control) for 5–30 min.
4. To extract total proteins, aspirate the medium. Immediately
put the dish on ice. Add 4 mL ice-cold DPBS.
5. Aspirate the DPBS. Add 0.25 mL ice-cold 1 RIPA plus
protease and phosphatase inhibitors. Collect the lysate using
cell scrapper, and transfer the lysate onto eppendorf tube.
6. Rock the lysate in cold room for 45 min.
7. Centrifuge the sample at 16,000  g at 4  C for 10 min.
8. Collect the supernatant. Measure the protein concentration by
Bradford reagent. The concentration of proteins should be
about 2 μg/μL.
9. Add 5 SDS sample buffer to the lysate. Boil the samples at
>95  C for 6 min right before SDS-PAGE (see Note 2).
 Surface Receptor Tyrosine Kinase TrkB in Neurons 219

3.2 Coimmuno- Interaction between RTK and signaling proteins can be examined
precipitation Between either in transfected 293T cells, in primary rat cortical neurons, or
RTK and Signaling in whole brain homogenate. Both 293T cells and primary rat
Proteins (Fig. 1) cortical neurons (5  106 cortical neurons) were cultured onto
100 mm dishes. For transfected 293T cells, transfection was per-
formed using Lipofectamine plus. Cell lysate was collected 24 h
after transfection.
1. To extract total proteins from 293T cells or cortical neurons,
aspirate the medium. Immediately put the dish on ice. Add
4 mL ice-cold DPBS.
2. Aspirate the DPBS. Add 0.6 mL ice-cold 1 Buffer A that
contains the various protease and phosphatase inhibitors listed
in Subheading 3.1, step 5. Then follow steps 6–8 of Subhead-
ing 3.1 above.
3. Coimmunoprecipitation of HEK293T cell or whole brain
lysate was performed in Buffer A. Lysate (1 mg for
HEK293T cells, 2 mg for cultured cortical neurons) was
diluted to 600 μL with buffer A, and incubated with the
corresponding antibody (1–2 μg) in cold room rocking
overnight.
4. Gently vortex the protein A–Sepharose (if the antibody for
immunoprecipitation is polyclonal) or protein G–Sepharose
(if the antibody for immunoprecipitation is monoclonal) for
1 min. Aliquot 40 μL of the sepharose to each eppendorf tube.
Wash the sepharose 3 times with 0.6 mL DPBS. Centrifuge at
600  g at 4  C for 1 min between each wash.
5. Add the cell lysate in step 3 to the sepharose, and incubate the
samples in a cold room with constant rocking for 1 h.
6. Wash the sepharose with 0.6 mL buffer A (plus the protease
and phosphatase inhibitors) three times. Centrifuge at 600  g
at 4  C for 1 min between each wash.
7. After the last wash (see Note 3), resuspend the sepharose beads
with 35 μL 2 SDS sample buffer. Boil the samples at >95  C
for 6 min. Centrifuge at 16,000  g at room temperature for
1 min, and transfer the samples to a new eppendorf tube. Run
the samples in SDS-PAGE.
The coimmunoprecipitated proteins can be detected by West-
ern blot (see Notes 4 and 5).

3.3 Coimmuno- 1. If brain homogenate is used as the starting material, weigh the
precipitation Between mouse brains and place the tissues into the chilled 15 mL glass
RTK and Signaling tube for tissue grinder (see Note 6). Add 3 volumes (e.g., 3 mL
Proteins in Whole to 1 g brain tissues) of ice-cold DPBS that contains the protease
Brain Lysate and phosphatase inhibitors as indicated in Subheading 3.1,
step 5.
220 Kwok-On Lai and Nancy Y. Ip

2. Homogenize the tissues for 1 min, then chill on ice for 1 min;
repeat the homogenization step two more times.
3. Centrifuge at 1000  g at 4  C for 5 min.
4. Collect supernatant. Add equal volume of 2 buffer A with
protease and phosphatase inhibitors.
5. Rocking for 45 min at 4  C.
6. Spin down at 16,000  g for 10 min at 4  C.
7. Collect supernatant (homogenate). The homogenate (2 mg
proteins) is subjected to coimmunoprecipitation as in steps
3–7 of Subheading 3.2.

3.4 Surface Protein 1. Dissolve Sulfo-NHS-LC-Biotin in DPBS to 1 mg/mL. Place it

as Determined by on ice. Precool the 50 mM Tris solution (pH 7.4) and DPBS.
Biotinylation 2. Aspirate medium, wash cells with ice-cold DPBS.
3. Add 2 mg biotin to cells, and incubate at 4  C for 30 min.
4. Remove the biotin solution, and add 5 mL 50 mM Tris solu-
tion (pH 7.4) to stop the reaction. Incubate at 4  C for 3 min.
5. Aspirate and wash cells once with DPBS.
6. Lyse cells by 1 RIPA (plus protease and phosphatase inhibi-
tors as listed in the Subheading 2.1). Put the dish of cells in
cold room and shake for 30 min.
7. Collect lysate by cell scrapper, and transfer the lysate to 1.5 mL
eppendorf tube. Centrifuge at 16,000  g at 4  C for 10 min.
8. Perform protein assay using Bradford reagent.
9. Wash the Streptavidin magnetic beads (40 μL) three times
by DPBS.
10. Remove most of the DPBS, and add 500 mg lysate to the beads
(dilute the lysate to 500 μL). Rock the tube in a cold room
overnight.
11. Wash the beads with 0.5 mL RIPA plus protease and phospha-
tase inhibitors three times. After each wash, centrifuge at
600  g at 4  C for 30 s.
12. After the final wash, remove residual wash buffer. Elute the
proteins by adding 2 sample buffer in a 1.5 mL eppendorf
tube. Boil the samples for 6 min. Spin the beads and transfer
the sample to a new eppendorf tube.
13. Load the samples to 6% SDS-PAGE. Total lysate was loaded in
separate lanes as control.
14. Western blotting with the antibody against the RTK of interest
(see Subheading 2 for choices) will determine the surface versus
total expression of the RTK.
 Surface Receptor Tyrosine Kinase TrkB in Neurons 221

4 Notes

1. Primary cortical neurons were prepared from day 18 rat

embryos [21]. Neurons were cultured onto 60 or 100 mm
dishes for 14–16 days before treatment and cell lysis.
2. The samples may be stored at
70  C, although for the exami-
nation of phosphorylated proteins, it is recommended to run
SDS-PAGE on the same day without freeze-thawing the
samples.
3. It is desirable to remove as much residual wash buffer as possi-
ble before eluting the proteins but at the same time without
considerable loss of the sepharose beads. After removing much
of the wash buffer from the last wash, centrifuge one more time
at 600  g, and remove the residual wash buffer using a P200
pipetman (with the pipet tip touching gently at the bottom of
the eppendorf tube).
4. The protein samples were separated by polyacrylamide gel, and
transferred onto PVDF membranes, followed by blocking with
5% skim milk in TBST for 1 h at room temperature (RT) with
rocking. The membrane was then incubated with primary anti-
body overnight in cold room with rocking. The next day after
washing the membrane three times with TBST, the membrane
was incubated with HRP-conjugated secondary antibody
diluted in 5% skim milk in TBST for 1 h at RT with rocking.
The membrane was washed with TBST for five times (20 min
first, followed by four times with 10 min interval) and the HRP
signal was detected by ECL.
5. For the separation of Trk receptors (~120 kDa) in SDS-PAGE,
a 6% gel is recommended. SDS running buffer contains the
following components: 24.8 mM Tris, 0.19 M glycine,
0.1% SDS.
6. The mouse brains may be dissected beforehand and quickly
frozen by liquid nitrogen. Put the tissues in 14 mL disposable
Falcon tubes containing liquid nitrogen and stand on dry ice.
Make sure to cap the tubes only when all the liquid nitrogen is
gone. Store the tissues in
70  C freezer.

Acknowledgment

We thank Leonard Wing-Hong Cheung for the schematic diagram.

We also thank the Research Grants Council of Hong Kong [Gen-
eral Research Fund (GRF) 16100814, 17135816, and Early Career
Scheme (ECS) 27119715] and the HKU seed funding programme
for basic research (201407159004 and 201511159170).
222 Kwok-On Lai and Nancy Y. Ip

References
1. Ip NY, Yancopoulos GD (1996) The neurotro- signal transduction pathways. Annu Rev Cell
phins and CNTF: two families of collaborative Biol 10:251–337
neurotrophic factors. Ann Rev Neurosci 13. Cheung ZH, Chin WH, Chen Y, Ng YP, Ip NY
19:491–515 (2007) Cdk5 is involved in BDNF-stimulated
2. Lai KO, Fu WY, Ip FC, Ip NY (1998) Cloning dendritic growth in hippocampal neurons.
and expression of a novel neurotrophin, NT-7, PLoS Biol 5:e63
from carp. Mol Cell Neurosci 11:64–76 14. Lai KO, Wong AS, Cheung MC, Xu P, Liang Z,
3. Lu Y, Christian K, Lu B (2008) BDNF: a key Lok KC et al (2012) TrkB phosphorylation by
regulator for protein synthesis-dependent LTP Cdk5 is required for activity-dependent struc-
and long-term memory? Neurobiol Learn tural plasticity and spatial memory. Nat Neu-
Mem 89:312–323 rosci 15:1506–1515
4. Park H, Poo MM (2013) Neurotrophin regu- 15. Miyamoto Y, Yamauchi J, Tanoue A, Wu C,
lation of neural circuit development and func- Mobley WC (2006) TrkB binds and tyrosine-
tion. Nat Rev Neurosci 14:7–23 phosphorylates Tiam1, leading to activation of
5. Longo FM, Massa SM (2013) Small-molecule Rac1 and induction of changes in cellular mor-
modulation of neurotrophin receptors: a strat- phology. Proc Natl Acad Sci U S A
egy for the treatment of neurological disease. 103:10444–10449
Nat Rev Drug Discov 12:507–525 16. Saragovi HU, Zheng W, Maliartchouk S,
6. Reichardt LF (2006) Neurotrophin-regulated DiGugliemo GM, Mawal YR, Kamen A et al
signalling pathways. Philos Trans R Soc Lond B (1998) A TrkA-selective, fast internalizing
Biol Sci 361:1545–1564 nerve growth factor-antibody complex induces
7. Benedetti M, Levi A, Chao MV (1993) Differ- trophic but not neuritogenic signals. J Biol
ential expression of nerve growth factor recep- Chem 273:34933–34940
tors leads to altered binding affinity and 17. Zhang Y, Moheban DB, Conway BR,
neurotrophin responsiveness. Proc Natl Acad Bhattacharyya A, Segal RA (2000) Cell surface
Sci U S A 90:7859–7863 Trk receptors mediate NGF-induced survival
8. Kaplan DR, Hempstead BL, Martin-Zanca D, while internalized receptors regulate
Chao MV, Parada LF (1991) The trk proto- NGF-induced differentiation. J Neurosci
oncogene product: a signal transducing recep- 20:5671–5678
tor for nerve growth factor. Science 18. Du J, Feng L, Yang F, Lu B (2000) Activity-
252:554–558 and Ca(2þ)-dependent modulation of surface
9. Minichiello L (2009) TrkB signalling pathways expression of brain-derived neurotrophic factor
in LTP and learning. Nat Rev Neurosci receptors in hippocampal neurons. J Cell Biol
10:850–860 150:1423–1434
10. Minichiello L, Calella AM, Medina DL, 19. Du J, Feng L, Zaitsev E, Je HS, Liu XW, Lu B
Bonhoeffer T, Klein R, Korte M (2002) Mech- (2003) Regulation of TrkB receptor tyrosine
anism of TrkB-mediated hippocampal long- kinase and its internalization by neuronal activ-
term potentiation. Neuron 36:121–137 ity and Ca2þ influx. J Cell Biol 163:385–395
11. Gruart A, Sciarretta C, Valenzuela- 20. Zhao L, Sheng AL, Huang SH, Yin YX,
Harrington M, Delgado-Garcia JM, Mini- Chen B, Li XZ et al (2009) Mechanism under-
chiello L (2007) Mutation at the TrkB PLC lying activity-dependent insertion of TrkB into
{gamma}-docking site affects hippocampal the neuronal surface. J Cell Sci
LTP and associative learning in conscious 122:3123–3136
mice. Learn Mem 14:54–62 21. Zeitelhofer M, Vessey JP, Xie Y, Tubing F,
12. van der Geer P, Hunter T, Lindberg RA (1994) Thomas S, Kiebler M et al (2007) High-
Receptor protein-tyrosine kinases and their efficiency transfection of mammalian neurons
via nucleofection. Nat Protoc 2:1692–1704
 Chapter 15

Polarized Human Retinal Pigment Epithelium Exhibits

Distinct Surface Proteome on Apical and Basal Plasma
Membranes
Vladimir Khristov, Qin Wan, Ruchi Sharma, Mostafa Lotfi,
Arvydas Maminishkis, and Kapil Bharti

Abstract
Surface proteins localized on the apical and basal plasma membranes are required for a cell to sense its
environment and relay changes in ionic, cytokine, chemokine, and hormone levels to the inside of the cell.
In a polarized cell, surface proteins are differentially localized on the apical or the basolateral sides of the
cell. The retinal pigment epithelium (RPE) is an example of a polarized cell that performs a variety of
functions that are dependent on its polarized state including trafficking of ions, fluid, and metabolites across
the RPE monolayer. These functions are absolutely crucial for maintaining the health and integrity of
adjacent photoreceptors, the photosensitive cells of the retina. Here we present a series of approaches to
identify and validate the polarization state of cultured primary human RPE cells using immunostaining for
RPE apical/basolateral markers, polarized cytokine secretion, electrophysiology, fluid transport, phagocy-
tosis, and identification of plasma membrane proteins through cell surface capturing technology. These
approaches are currently being used to validate the polarized state and the epithelial phenotype of human
induced pluripotent stem (iPS) cell derived RPE cells. This work provides the basis for developing an
autologous cell therapy for age-related macular degeneration using patient specific iPS cell derived RPE.

Key words Retinal pigment epithelium, Polarization, Monolayer, Immunocytochemistry, Cytokine

secretion, Electrophysiology, Fluid transport, Phagocytosis, Cell surface capturing

1 Introduction

The retinal pigment epithelium (RPE) is a monolayer of highly

polarized pigmented cells situated between the neural retina and
the choroid capillaries, and it plays a critical role in the maintenance
of visual function [1]. The RPE helps regulate the volume and
chemical composition of the subretinal space surrounding photo-
receptor outer segments and serves as the outer blood-retinal bar-
rier for the eye. It plays a number of crucial roles to support the
photoreceptor function including transportation of nutrients from
blood to the photoreceptors, maintenance of constant ion

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_15, © Springer Science+Business Media, LLC 2018

223
224 Vladimir Khristov et al.

composition in the subretinal space, transport of water from the

subretinal space to the blood, polarized secretion of growth factors,
and phagocytosis of the outer segments of the photoreceptors
[1–3]. A highly polarized distribution of membrane ion channels,
transporters, and receptors on apical and basolateral membranes of
the cell are essential for maintaining these functions in RPE mono-
layers. The disruption of polarized RPE structure or function often
contributes to the pathogenesis of several blinding disorders of the
eye [4–6].
In order to verify RPE polarization and identify apical and
basolateral selective cell surface markers, primary human fetal
RPE (hfRPE) cells were grown on semipermeable transwell mem-
branes to obtain confluent polarized monolayers using previously
described protocols [7, 8] (Subheading 3.1). Immunostaining of
cells for well-known RPE apical and basolateral markers gave a
preliminary indication that the cells formed a polarized monolayer.
This was confirmed with transmission electron microscopy (TEM)
imaging which showed apical and basolateral-specific structures
differentially distributed toward the two membranes (Subheadings
3.2.1 and 3.2.2). Functional polarization of RPE monolayer was
then verified by measuring electrical responses evoked by the acti-
vation of apically or basally located channels or receptors, polarized
secretion of the VEGF cytokine into the basal medium, fluid trans-
port from the apical to the basal side, and apical phagocytosis of
photoreceptor outer segments (Subheadings 3.3.1–3.3.4). Taken
together, these assays are strongly indicative of a fully polarized
RPE monolayer that mirrors the function of native RPE tissue.
Cell surface capturing (CSC) technology (Subheading 3.4.1) was
used to perform a comprehensive and unbiased analysis of RPE
surface proteome. A brief description of the mass spectrometry and
data analysis is provided (Subheadings 3.4.2 and 3.4.3); however,
the reader is referred to another chapter in this book by Fujinaka
et al. (Chapter 4) for more in-depth analyses. The CSC results
confirmed the polarization of the RPE monolayer by detecting
the presence of known RPE proteins on the apical and basolateral
membranes. CSC also identified some novel cell surface proteins
that are being used to analyze signaling pathways previously not
known to be involved in RPE physiology and pathology. CSC
technology followed by validation (Subheading 3.4.4) through
immunocytochemistry, western blot, and functional analysis repre-
sents an excellent tool to generate a comprehensive atlas of cell
surface proteins which could serve as a valuable pool for the devel-
opment of novel diagnostic or therapeutic reagents.
 Polarized State of Human RPE 225

2 Materials

2.1 Cell Culture 1. Antibiotic–antimycotic solution diluted to 10
(catalog num-

ber 15240-096) (Invitrogen) plus 1 mg/mL gentamicin (cata-
log number 15710072) (ThermoFisher Scientific, Carlsbad,
CA, USA).
2. Hank’s balanced salt solution (HBSS) Ca2+/Mg2+ free.
3. 2 U/mL Dispase-I solution (Roche Diagnostics, Indianapolis,
IN, USA).
4. MEM-alpha modified medium (MEMα) cell culture media
(Sigma-Aldrich Chemical Company, St. Louis, MO, USA).
5. Fetal bovine serum (treated at 56  C for 30 min).
6. RPE medium (see Table 1)—MEMα is used as the base medium
to prepare 5% and 15% serum-containing media for culturing
of RPE cells [7, 8]. Medium should be filter-sterilized.
7. 0.25% Trypsin–EDTA solution (catalog number 25200-056)
(ThermoFisher Scientific, Grand Island, NY, USA).
8. Human extracellular matrix. Add 10 μg to 200 μL HBSS per
transwell (catalog number 354237) (BD Biosciences, Franklin
Lakes, NJ, USA).
9. Primaria flasks T25 (catalog number 353808) (Corning Life
Science, Corning, NY, USA).
10. 75 mm transwells (Catalog number 3419) (Corning Life
Science).

Table 1
Human fetal RPE medium components for preparation of 500 mL medium

Name Source (Catalog #) Amount Concentration Storage

MEM, alpha modification Sigma (M-4526) 500 mL þ4  C
N2 supplement Gibco (17502-048) 5 mL 1:100 mL/mL 80  C
Penicillin-streptomycin Gibco 15140-148 5 mL 1:100 mL/mL 20  C
GlutaMax-I Gibco 35050 5 mL 1:100 mL/mL 20  C
Non-essential amino acids Sigma M-7145 5 mL 1:100 mL/mL þ4  C
Taurinea Sigma T-0625 125 mg 250 mg/L RT
Hydrocortisone a
Sigma H-0396 10 μg 20 μg/L RT
Triiodo-thyronin a
Sigma (T-5516) 0.0065 μg 0.013 μg/L 20  C
Fetal bovine serumb Atlanta Biologicals 25 or 75 mL 5% or 15% 80  C
a
Taurine, Hydrocortisone and Triiodo-thyronin (THT) were prepared together by dissolving in 1–1.5 mL DMEM;
multiple aliquots are made and stored at 80  C to simplify the preparation of the culture medium
b
Fetal bovine serum (FBS) needs to be heat inactivated (56  C for 1 h) prior to use
226 Vladimir Khristov et al.

11. 12 mm transwells (Catalog number 3401) (Corning Life

Science).
12. 5 and 10 mL tissue culture sterile Pipettes.
13. Water bath.

2.2 Immunocy- 1. Fixation solution: 2% paraformaldehyde (Electron Microscopy

tochemistry Sciences, Hatfield, PA) in phosphate buffered saline (PBS) (Ca2
+
/Mg2+ free) (catalog number 10010023) (ThermoFisher Sci-
entific, Pittsburgh, PA) (see Note 1).
2. Wash solution: 0.5% Tween-20 (Sigma Chemical Company,
St. Louis, MO, USA) in PBS.
3. Blocking and Permeabilization Solution (ICC buffer): 0.5%
bovine serum albumin (BSA) (MP Biomedicals, Santa Ana,
CA, USA), 0.5% Tween 20, 0.005% sodium azide, 0.1% Triton
X-100 (all from Sigma-Aldrich) in PBS (see Note 1).
4. Fluoromount-G (SouthernBiotech, Birmingham, AL, USA).
5. 75
25 mm Superfrost Plus Slides (Daigger, Vernon Hills, IL,
USA), Gold Seal 24
50 mm No. 1.5 coverslips (Thermo
Scientific, Waltham, MA, USA).
6. Antibodies: Rabbit anti-OA1 (catalog number OSR00146W;
ThermoFisher Scientific), Rabbit anti-SLC5A12 (catalog num-
ber PA5-23834; ThermoFisher Scientific), Rabbit anti-
GPNMB (catalog number PA5-27874; ThermoFisher Scien-
tific), Phalloidin-Alexa 488, Donkey anti-Rabbit-Alexa
555, Hoechst (Life Technologies, Carlsbad, CA, USA).
7. Tissue-Tek OCT Cryosectioning Medium (catalog number
25608-930; VWR, Randor, PA, USA).
8. Leica CM1900 Cryostat (catalog number CM1900) (Leica,
Wetzlar, Germany) or similar.

2.3 Transmission 1. EM grade fixation solution: 2.5% glutaraldehyde in 0.1 M

Electron Microscopy phosphate buffer.
(TEM) 2. 1% osmium tetroxide in 0.1 M phosphate buffer.
3. 30%, 50%, 70%, 90%, and 100% acetone solutions for
dehydration.
4. TEM embedding resin (Sigma-Aldrich).
5. Leica RM2235 manual microtome or equivalent.

2.4 Electro- 1. Ringer’s solution: The control Ringer’s solution used for apical
physiology and basal bath perfusion has the following composition
(in mM): NaCl 116.5, NaHCO3 23, KCl 5, MgCl2 0.5, Glu-
cose 5, Taurine 2, CaCl2 1.8, and sucrose 10. Prior to use, all
solutions are bubbled continuously with 8% CO2
(pH 7.4  0.05), and maintained at 37  1.0  C.
 Polarized State of Human RPE 227

2. Voltage-Current Clamp (VCC 600; Physiologic Instruments,

Houston, TX, USA).
€
3. Modified Ussing chamber (Custom-made).
4. Accumet calomel electrode (catalog number: 13-620-52;
ThermoFisher Scientific).
5. Agar bridges: make 3.5–4% Agar gel (catalog number A7921)
(Sigma-Aldrich) in Ringer’s solution and use a 20 mL syringe
to inject the Agar gel into the PE160 Polyethylene tubing
(Catalog number: 427431) (Becton Dickinson & Co, Franklin
Lakes, NJ, USA) (see Note 2). Keep agar bridges in 4  C Ring-
er’s solution and use within 1 month.
6. Branson Ultrasonic bath (Hach, Loveland, Colorado, USA).
7. Skin biopsy punch: 7 mm hole punch (Acuderm Inc., Lauder-
dale, FL, USA).
8. Water Circulator and water bath (Lauda E100 & E200, Del-
ran, NJ, USA).

2.5 Polarized 1. Luminex xMAP magnetic bead assay kit (Millipore, Billerica,
Cytokine Secretion MA, USA).
Measurement 2. Luminex 200 analyzer (Luminex, Austin, TX, USA).
3. Bioplex manager TM 6.0 (Bio-Rad, Hercules, CA, USA).

2.6 Vectorial Fluid 1. 11 mm Trephine tissue hole punch (Lines Industries, Concord,
Transport CA, USA).
Measurement 2. 205 μm thickness Nylon mesh (SEFAR, Buffalo, NY, USA).
3. Capacitive probe (MT Instruments, Latham, NY, USA).
4. Agar bridges: 4.5% MEMα-agar bridges were used (catalog
number A7002) (Sigma, St. Louis, MO, USA) (see Note 2).
€
5. Modified Ussing chamber (custom-made).
€
6. Ussing chamber water jacket (custom-made).
7. Voltage-Current Clamp (VCC600; Physiologic Instruments).

2.7 Phagocytosis 1. Bovine Photoreceptor Outer Segments (POS) (catalog num-

Assay ber 98740) (InVision BioResources, Seattle, WA, USA).
2. Sucrose-Sodium Bicarbonate buffer composed of 10% sucrose
and 0.1 M sodium bicarbonate. Adjust pH with 1.0 N NaOH
to 8.3.
3. Bicinchoninic acid (BCA) assay kit (ThermoFisher Scientific).
4. pH-sensitive pHrodo dye (catalog number P36600) (Life
Technologies).
5. Milk fat globule-EGF factor 8 protein (MFG-E8) (R&D Sys-
tems, Minneapolis, MN, USA).
228 Vladimir Khristov et al.

6. 1
PBS.
7. 0.25% Trypsin–EDTA (ThermoFisher Scientific).
8. Fluorescence-activated cell sorting (FACS) buffer composed of
0.1% BSA (MP Biomedicals) in 1
PBS.
9. 40 μm cell strainer (BD Biosciences).
10. 40 ,6-diamidine-20 -phenylindole dihydrochloride (DAPI) solu-
tion (ThermoFisher Scientific).
11. Zeiss 780 Confocal Microscope or equivalent fluorescence
microscope with 20
magnification objective.

2.8 Cell Surface 1. Cell scraper.

Capturing (CSC) 2. Tissue glass Teflon Dounce homogenizer capable of holding at
Technology (See Note 3) least a volume 5 mL.
3. Thermomixer C (Eppendorf, Hauppauge, NY, USA) or equiv-
alent for microcentrifuge tubes or ultracentrifuge tubes.
4. Speed Vac.
5. Labeling buffer: Add 0.1 mL of FBS to 99 mL of 1
PBS and
adjust pH to 6.5 with phosphoric acid and bring to a final
volume of 100 mL. This solution should be made on the day
of the experiment, placed on ice.
6. Dissolve 0.033 g sodium meta-periodate (Pierce, Rockford, IL,
USA) in 1 mL labeling buffer immediately before use. Vortex
to dissolve.
7. Dissolve 30 mg biocytin hydrazide (Biotium, Hayward, CA,
USA) in 1 mL labeling buffer.
8. Hypotonic Lysis buffer: 10 mM Tris pH 7.5, 0.5 mM MgCl2.
9. Membrane Wash buffer: 25 mM Na2CO3.
10. 1 M Stock MES pH 6: Add 97.6 g of MES to approximately
490 mL of H2O and adjust pH with NaOH. The final volume
should be adjusted to 500 mL.
11. Membrane Prep buffer: 280 mM sucrose, 50 mM MES
pH 6.5, 450 mM NaCl, 10 mM MgCl2. For 500 mL, combine
47.92 g sucrose, 25 mL 1 M MES stock pH 6.5, 45 mL 5 M
NaCl, 50 mL 100 mM MgCl2, and H2O.
12. 100 mM NH4HCO3.
13. 100 mM Tris(2-carboxyethyl) phosphine (TCEP). Dissolve
0.2867 g into 10 mL water. Divide into 0.5 mL aliquots and
store at 20  C. (catalog number C4706) (Sigma-Aldrich).
14. 1% (v/v) Rapigest. Dissolve a 10 mg vial of Rapigest in 1 mL
H2O. Store at 4  C (catalog number 186002122) (Waters
Corporation, Milford, MA, USA).
15. 10 mM iodoacetamide in H2O prepared fresh.
 Polarized State of Human RPE 229

16. Lys-C (catalog number V1071) (Promega, Madison, WI,

USA).
17. Proteomics grade trypsin (catalog number V5280) (Promega).
18. 450 μL bead slurry of UltraLink Immobilized Streptavidin
PLUS (Pierce).
19. MoBiCol spin columns 10 μm/30 μm pore size (catalog num-
ber M1002S) (MoBiTec GbmH, Goettingen, Germany).
20. C 18 MicroSpin™ column (Harvard Apparatus, Holliston,
MA, USA).
21. 0.05% Triton X-100 (Sigma-Aldrich) in 100 mM NH4HCO3.
22. 5 M NaCl.
23. 100 mM NH4HCO3.
24. 100 mM Na2CO3.
25. 80% isopropanol.
26. 500 units glycerol-free endoproteinase PNGaseF (New Eng-
land Biolabs, Ipswich, MA, USA).
27. Trifluoroacetic acid (TFA).
28. 90% acetonitrile–0.1% formic acid in H2O.

2.9 Subcellular 1. Lysis Buffer: Radioimmunoprecipitation assay (RIPA) buffer,

Fractionation 1 mM dithiothreitol (DTT; Sigma-Aldrich), and cOmplete
ULTRA Protease Inhibitor cocktail (catalog number
05892970001; Roche, Basel, Switzerland).
2. Subcellular fractionation buffer composed of 250 mM Sucrose,
20 mM HEPES (pH 7.4), 10 mM KCl, 1.5 mM MgCl2, 1 mM
EDTA, 1 mM EGTA, 1 mM DTT, and cOmplete ULTRA
Protease Inhibitor cocktail, using 1.0 M HCl to adjust pH to
7.4 (see Note 4).
3. Nuclear buffer: lysis buffer with 10% glycerol (ThermoFisher
Scientific) and 0.1% SDS (Bio-Rad), 1 mM DTT, and cOm-
plete ULTRA Protease Inhibitor cocktail.
4. 25 G needle (BD Medical Supplies, Franklin Lakes, NJ).
5. Tabletop microcentrifuge (Eppendorf 5417R or similar).
6. Ultracentrifuge capable of 100,000
g.
7. 1.5 mL microcentrifuge tubes.
8. Ultracentrifuge tubes.

2.10 Western Blot 1. Trans-Blot Turbo RTA Transfer Kit, LF PVDF (catalog num-
ber 1704272) (Bio-Rad).
2. Trans-Blot Turbo Transfer System (catalog number 1704150)
(Bio-Rad).
3. Criterion Cell (catalog number 1656001XTU) (Bio-Rad).
230 Vladimir Khristov et al.

4. Criterion TGX 12þ2 Gel (catalog number 5671103)

(Bio-Rad).
5. Antibodies: Rabbit anti-GRIK1 (catalog number PA5-20448),
Rabbit anti-GPR143 (catalog number OSR00146W) (Ther-
moFisher Scientific, Waltham, MA).

3 Methods

3.1 Maintaining 1. Start with RPE cells maintained in primaria flasks in an incuba-
and Splitting RPE cells tor (37  C, 5% CO2).
(See Note 5) [7–9] 2. To split cells from single 25 cm2 primaria flask, trypsinize cells
in 5 mL trypsin–EDTA for 20 min at 37  C, centrifuge at
400
g for 5 min, and resuspend the cells in 10 mL 15%
FBS containing RPE medium.
3. Put cell suspension into two 25 cm2 primaria flasks and keep
the flask in an incubator (37  C, 5% CO2) overnight.
4. Replace the medium after 1 day with 5% FBS containing RPE
medium. Change medium every 2 days.
5. After 4–5 weeks, trypsinize cells in 5 mL trypsin–EDTA for
20 min at 37  C, centrifuge at 400
g for 5 min, and resus-
pend the cells in 15% FBS containing RPE medium.
6. In some cases, repeat the trypsinization procedure a second
time to collect the cells that do not detach after the first
trypsinization.
7. Before seeding, coat 12 or 75 mm transwells with human
extracellular matrix (100 μg in 2 mL HBSS per well) and
expose them to UV light for 2 h in the hood.
8. Seed the cells onto clear cell culture inserts at a density of
1.5
105 cells per 12 mm diameter transwell for functional
analysis or at 1.0–1.5
107 per 75 mm diameter transwell for
CSC analysis.
9. The same protocol (excluding coating with ECM) is used to
culture cells on the flasks to generate the P1 population of cells
(see Note 6).
10. RPE cells are used for experiments when they reach a total
tissue resistance of >300 Ω cm2 (see Subheading 3.3.1) and
are uniformly pigmented.

3.2 Ensuring RPE A polarized confluent RPE monolayer has structural features (such
Monolayer Polarization as apical processes/basal infoldings, apically localized pigment
granules and basally located nuclei) and special characteristic prop-
erties (such as polarized distribution of certain ion channels, trans-
porters, and receptors on RPE apical or basolateral membranes to
mediate specific electrical responses, ability of the monolayer to
 Polarized State of Human RPE 231

secrete cytokines in a polarized fashion, and the ability of the

monolayer to transport water from the apical side to the basal
side). Ensuring that polarized RPE monolayers have these proper-
ties is essential for subsequent experiments designed to identify
apical versus basal plasma membrane localized proteins.

3.2.1 Ultrastructural A polarized monolayer of RPE cells exhibits well-defined structural

Analysis characteristics. Apical processes that interact with photoreceptor
outer segments, apically localized melanosomes absorb stray light,
basally located nuclei and basal infoldings are structures that can be
clearly observed using Transmission Electron Microscopy (TEM)
(Fig. 1) [7].
1. Fix RPE cells grown on transwell insert in EM-fixation solution
for 1 h at 4 C.
2. Perform post-fixation in 1% osmium tetroxide in 0.1 M phos-
phate buffer for 1–2 h at 4  C.
3. Dehydrate samples in 30%, 50%, 70%, 90%, and 100% acetone
for 15 min each.

Fig. 1 TEM micrograph of hfRPE monolayer grown on transwell membrane. A

polarized monolayer of hfRPE cells is observed as indicated by apical processes,
apically localized melanosomes, and nuclei located close to the basolateral
membrane
232 Vladimir Khristov et al.

4. Embed in TEM sectioning resin according to manufacturer

specifications.
5. Section at 50–100 nm thickness using Leica RM2235, and
image via TEM.

3.2.2 RPE Protein Marker Differentially localized proteins provide specialized functions in
Immunostaining different compartments of a cell (see Note 7). For instance,
EZRIN an apically located protein is a structural component of
RPE apical processes [10] and COLLAGEN IV a basally located
protein is a part of RPE basement membrane [11]. Figure 2 shows
expected localization of EZRIN and COLLAGEN IV in a mono-
layer of primary RPE cells.
1. Rinse RPE cells grown as a polarized monolayer on 12 mm
transwell three times in 1
PBS for 1 min each.
2. Incubate in fixation solution for 15 min exactly.
3. Rinse three times in 1
PBS for 1 min each.
4. Embed RPE monolayer in OCT cryosectioning medium based
on manufacturer specifications.
5. Make 10 μm cryosections on cryostat following manufacturer
instructions, and mount on slides.
6. Incubate in blocking and permeabilization solution 1 h at RT.
7. Add primary antibody (diluted 1:100–1:1000) diluted in
blocking and permeabilization solution, and incubate over-
night at 4  C.
8. Wash three times in washing solution for 15 min each.
9. Add fluorophore-conjugated secondary antibody diluted in
ICC buffer, and incubate for 1 h at room temperature in the
dark, with gentle shaking.

Fig. 2 Polarized distribution of known RPE markers in the monolayer. Immunofluorescence localization of RPE
basolateral marker Collagen IV (a) and apical marker Ezrin (b) in primary cultures hfRPE grown on transwell
inserts. DAPI (blue) labels the nuclei located close to the basolateral membrane
 Polarized State of Human RPE 233

10. Wash three times in washing solution for 15 min each, with
gentle shaking.
11. Mount samples on glass slides using Fluoromount-G and ana-
lyze via microscopy.

3.3 Functional Fully polarized RPE monolayers have characteristic functional

Analysis responses that are not present in nonpolarized cells. Tight junctions
located on apical-lateral sides of the cells seal the paracellular path-
way between neighboring cells forming transepithelial resistance
across the RPE monolayer that can be measured by passing current
through the monolayer. Tight junction formation isolates apical
and basolateral membrane compartments and leads to localization
of different channels and receptors on RPE apical or basolateral
membranes. Differential activity of these transporters and channels
can be measured as differential resting membrane potential on
apical and basal membranes [12]. Moreover, certain cytokines like
vascular endothelial growth factor (VEGF) is preferentially secreted
toward RPE basal side while pigment epithelium-derived factor
(PEDF) is preferentially secreted toward the apical side [13].

3.3.1 Functional 1. Take out a primary cultured hfRPE transwell and cut out a
Validation of Cell Surface piece of RPE monolayer with a 7 mm skin biopsy punch.
Receptors and Channels 2. Mount the RPE monolayer on a modified Ussing € chamber (see
Using Electrophysiology Note 8) [12, 14, 15]. Perfuse Ringer’s solution (pH 7.4,
37  C) to the apical and basal baths of the chamber separately.
3. Connect the calomel electrodes in series with agar bridges to
Ringer’s solution of the chamber to make electrical contact
with each bathing solution.
4. Measure the transepithelial potential (TEP) of the tissue under
CURRENT mode of the Voltage-Current Clamp (see Note 9).
5. Pass 2–4 μA current pulse (bipolar, with a period of 3 s applied
at 40 s intervals) across the tissue mounted in the Ussing€
chamber, measure the resultant changes in TEP to obtain the
total tissue resistance (RT).
6. Dissolve ATP or ClC2 channel activator (lubiprostone) in
Ringer’s solution to reach the final working concentration,
and perfuse it into either the apical or basal bath to detect the
activity of P2Y2 receptors or ClC2 channels on the RPE mem-
brane (see Figs. 3 and 4).

3.3.2 Polarized Cytokine RPE is located between the neural retina and the choroid. These
Secretion Measurement three different cell layers form the homeostatic unit at the back of
the eye to maintain visual function. The polarized secretion of
VEGF on the basal side and PEDF on the apical side by RPE
monolayer is required for the maintenance of the health and integ-
rity of choroid and retina respectively [13, 16] (see Note 10). The
234 Vladimir Khristov et al.

Fig. 3 Differential electrical responses evoked by apical or basal application of ATP. The continuous trace
represents transepithelial potential (TEP) and open squares represent total tissue resistance (RT). The apical or
basal bath perfusion of 100 μM ATP is indicated by black bars above the graph. A time scale bar is located at
the bottom of the graph. In the native tissue, ATP acts on purinergic P2Y2 receptors on RPE apical membrane,
increasing intracellular IP3 and thus stimulating Ca2+ release from the endoplasmic reticulum, to cause
activation of basolateral membrane Ca2+-activated Cl channels followed by a decrease in apical membrane
K+ conductance, thus leading to electrical responses across the RPE monolayer [14]. Consistent with this,
upon the apical application in the primary cultures RPE, ATP produced a dramatic biphasic response in TEP
and a markedly decrease in RT; but induced much smaller TEP responses with little RT changes after basal
perfusion, suggesting the majority of P2Y2 receptors are located on RPE apical membrane

Fig. 4 Differential electrical responses evoked by apical or basal application of the ClC2 channel activator. The
ClC2 activator (10 μM) produced a similar increase in TEP and a decrease in RT upon either the apical or basal
bath application, suggesting the basal localization of ClC2 channels in RPE. The activation of ClC2 on the RPE
basal membrane leads to the efflux of intracellular Cl and depolarization of RPE basolateral membrane, thus
causing TEP increase and RT drop. The larger effect induced by apical application may be due to the much
larger RPE apical surface area
 Polarized State of Human RPE 235

6000
APICAL BASAL

VEGF levels in media (pg/ml)

5000

4000

3000

2000

1000

0
Line A Line B Line C

Fig. 5 Polarized cytokine secretion to determine the polarity of iPSC-derived RPE monolayer. RPE from three
different RPE samples, namely Line A, B, and C, were tested for VEGF secretion on apical and basal side. RPE
monolayers were fed with fresh media 24 h before the media collection. The collected media was spun down
and used for the analysis. The results demonstrate the RPE monolayers grown in vitro secrete the VEGF
cytokine in a manner replicating in vivo function

secretion profiling of both the cytokines by iPSC-derived RPE is

determined using Luminex xMAP (Milliplex™ MAP) technology
based magnetic bead assay kit from Millipore. With good RPE
preparations, total VEGF and PEDF secretions on apical and
basal side can be observed within normal ranges for primary
hfRPE (Fig. 5).
1. Collect media from apical and basal sides and spin separately at
10,000
g for 5 min.
2. Take 150 μL of media from each sample.
3. Measure the total protein concentration for each sample using
a BCA assay kit according to the manufacturer’s
recommendations.
4. Collect cytokine concentration data using Luminex 200 ana-
lyzer using Bioplex ManagerTM 6.0.
5. Normalize experimental values to total protein concentration
measured in step 3 (see Note 10).

3.3.3 Vectorial Fluid An essential function of the RPE monolayer is unidirectional trans-
Transport Measurement port of water from the subretinal space to the choriocapillaris. This
€
fluid flow (JV) can be measured (Fig. 6) using a modified Ussing
chamber as described previously [17, 18].
1. Set incubator temperature to 39  C, CO2 to 5% and relative
humidity to 50%.
€
2. Set Ussing chamber water jacket to 36  C.
3. Place all solutions to be used in an incubator at least 1 h ahead
of time to equilibrate the CO2 and temperature conditions.
236 Vladimir Khristov et al.

Fig. 6 Fluid transport response induced by apical application of ATP. Top panel: Transepithelial fluid transport
(JV). Bottom panel: Black trace represents transepithelial potential (TEP) and red trace represents transe-
pithelial resistance (RT). Application of ATP is indicated by the black bar above the top graph. The apical
(AP) application of 100 μM ATP activated the P2Y2 receptors on RPE apical membrane, which resulted in a
reversible increase in apical-to-basal fluid adsorption. Time scale bar indicated 50 min

4. Use a trephine tissue punch to cut out the Transwell membrane

containing the RPE monolayer from the insert.
€
5. Mount cells vertically in custom designed modified Ussing
chamber using nylon mesh for support.
6. Fill both sides of chamber with MEM alpha media.
7. Record fluid absorption (JV) using the capacitive probe tech-
nique to detect changes in fluid level on the apical and basal
sides of the mounted tissue.
8. Record transepithelial potential (TEP) by voltage-current
clamp (also see Note 9).
9. Record transepithelial resistance (RT) by applying 2 μA pulse at
30 s intervals.

3.3.4 Phagocytosis RPE cells are among the most active phagocytic cells in the human
Assay body. Primary RPE should possess the same phagocytic machinery
as native RPE. Apically localized phagocytic receptors allow healthy
RPE phagocytose photoreceptor outer segments (POS), remove
metabolic waste and recycle visual pigment. The ability of RPE cells
to phagocytose POS in vitro is indicative of a healthy, polarized
monolayer [16, 19] (Fig. 7).
 Polarized State of Human RPE 237

Fig. 7 Phagocytic capability of polarized primary hfRPE monolayer. Fluorescently labeled photoreceptor outer
segments (POS) were observed in hfRPE cells indicating its normal phagocytic capability

1. Thaw the frozen pellets of bovine POS at 4  C and centrifuge at

100,000
g for 10 min twice in 10% sucrose, 0.1 M sodium
bicarbonate buffer.
2. Wash the POS twice in the same buffer before they were
counted.
3. Determine protein concentration using BCA assay kit accord-
ing to the manufacturer’s instructions.
4. Label POS with pHrodo dye using manufacturer’s
instructions.
5. Divide iPSC derived RPE transwells into four groups: UNFED
(without POS), FED (with POS segments, Low and High),
negative Control (4C), positive Control (MFG-E8 addition).
6. Add POS to apical side of transwell containing iPSC derived
RPE (feeding 5 POS particles/RPE cell; estimating 500,000
cells per 12 mm Transwell) and incubate for 4 h at 37  C.
7. After 4 h of incubation, wash cells thoroughly with PBS 3–4 times.
8. Punch the transwells out and mount them on slide in fluoro-
mountG mounting medium.
9. Seal the coverslip and image the samples under Zeiss confocal
microscope at 20
magnification.

3.4 Cell Surface Extracellular, N-linked glycoproteins were labeled and captured
Capturing (CSC) using CSC technology [20–22] (see Note 11). The
Technology CSC-technology uses meta-sodium periodate to oxidize carbohy-
drates on intact cells and followed by covalent chemical labeling of
oxidized carbohydrate-containing proteins with biotin for affinity
enrichment. To selectively identify apical or basolateral surface
238 Vladimir Khristov et al.

proteins, this process is specifically done on those two sides of a

confluent, electrically and mechanically intact RPE monolayer in a
transwell dish. After biotinylation, cells are lysed, and membrane
fractions are isolated. N-linked peptides are released by peptide-N-
glycosidase F and subjected to mass spectrometry to identify labeled
proteins based on peptide sequencing. Peptide tags are aligned
against online protein sequence databases to identify respective pro-
teins. To validate that the CSC technique works well on RPE cells, we
initially looked for the presence of signature proteins previously
identified on RPE cells [23]. Our analysis identified a wide variety
of previously known RPE surface proteins including aquaporin-1,
interferon gamma receptor 1, insulin-like growth factor 1 receptor,
tyrosinase protein kinase MER, membrane frizzled related protein,
sodium bicarbonate cotransporter 3, platelet-derived growth factor
receptor beta, and transferrin receptor 1 [24–34]. In addition, we
have identified a number of previously unreported proteins present
on RPE membrane surface including members of the
ATP-dependent transporter family, several cluster of differentiation
(CD) receptors, integrins, ephrins, sodium/potassium transporting
ATPase subunits, and members of the cadherin superfamily. Bioin-
formatics analysis using predicted/known transmembrane domains,
mRNA expression, and protein localization allowed the selection of a
small subset of proteins of particular interest. The localization of these
proteins was validated by immunostaining (Figs. 8–10), and western
blot (Figs. 11 and 12).

Fig. 8 Immunostaining of a G-protein coupled receptor RPE marker. Immunofluorescence localization of a

known RPE marker, OA1 (red) in cultures hfRPE (a) and fibroblasts (b). In each figure, the Central panel is an en
face view of the cell culture insert, shown as the maximum-intensity projection through the z-axis. Images of
the cross section through the Z-plane are shown in the top and the right panels. F-actin (green) was used to
visualize the cytoskeleton and cell boundary in the cell; DAPI (blue) labels the nuclei located close to the
basolateral membrane
 Polarized State of Human RPE 239

Fig. 9 Immunostaining of a transmembrane protein RPE marker. Immunofluorescence localization of a known

RPE marker, GPNMB (red), F-actin (green) in cultures hfRPE (a) and fibroblasts (b)

Fig. 10 Immunostaining of a novel RPE cell maker. Immunofluorescence localization of a novel RPE marker,
SLC5A12 (red) and F-actin (green) in cultured hfRPE (a) and fibroblasts (b)

Fig. 11 Western blot of a novel RPE marker GRIK1 is enriched in the membrane and cytosolic fractions of
hfRPE cells. (A) Nuclear wash buffer, (B) nuclear fraction, (C) mitochondrial fraction, (D) cytosolic fraction,
(E) membrane fraction
240 Vladimir Khristov et al.

Fig. 12 Western blot of known RPE marker GPR143 is enriched in the membrane
fraction of hfRPE cells. (A) Nuclear wash buffer, (B) nuclear fraction,
(C) mitochondrial fraction, (D) cytosolic fraction, (E) membrane fraction

3.4.1 Cell Surface 1. Labeling of human fetal RPE cells is performed on either the
Capturing (See Note 12) apical or the basal plasma membranes.
2. Aspirate the cultivation medium and wash both the apical and
basolateral membranes of RPE cells in transwells with labeling
buffer. Repeat twice to remove debris and as many dead cells as
possible.
3. Oxidize either the apical or the basal membrane by treating the
designated membrane with 0.2 mg/mL sodium metaperiodate
for 15 min at 4  C in the dark with gentle rotation. Incubate
the other membrane with labeling buffer (see Note 13).
4. Wash cells twice to remove periodate solution. Add 4 mg/mL
biocytin hydrazide or labeling buffer to the respective mem-
branes of the RPE cells for 1 h at 4  C.
5. Gently wash the cells twice with labeling buffer. Collect all RPE
cells by gently scraping the transwell with a cell scraper. Add
additional labeling buffer and continue scraping until all cells
have been isolated.
6. Scraped cells should be transferred to a 50 mL Falcon tube
using a sterile Pasteur pipette.
7. Centrifuge the cells at 800
g. Discard supernatant and wash
with labeling buffer. Recentrifuge.
8. Aspirate the supernatant and resuspend the pellet in 4 mL of
hypotonic lysis buffer on ice. Incubate 15 min.
9. Transfer cells to a Dounce homogenizer and gently homoge-
nize on ice through 35 strokes of the teflon pestle.
10. Transfer the homogenized cells to a 15 mL polypropylene tube
on ice and centrifuge at 800
g for 10 min. Remove tube and
place on ice (see Note 14).
11. Transfer the supernatant, without disturbing the pellet, to a
new tube. Resuspend the pellet in an additional 2 mL hypo-
tonic lysis buffer and rehomogenize the resuspended pellet on
ice with 35 strokes of the teflon pestle (see Note 15).
12. Repeat steps 10 and 11. After the third homogenization and
centrifugation, it usually becomes possible to distinguish
between the cell pellet and the supernatant.
 Polarized State of Human RPE 241

13. Combine the supernatants (~6 mL total), and add an equal

volume of Membrane Prep buffer. Mix and leave on ice for
10 min. Transfer to an ultracentrifuge tube and centrifuge the
supernatant at 210,000
g for 16 h at 4  C to collect the
membranes.
14. Remove all the supernatant from the ultracentrifuge tube and
add 200 μL of Membrane wash buffer to pellet. Incubate for
30 min at 4  C in a Thermomixer (750 rpm) or equivalent to
disrupt peripheral protein interactions (see Note 16).
15. Add Hypotonic Lysis buffer to the top of the tube and centri-
fuge at 210,000
g for 30–60 min at 4  C.
16. After centrifugation, discard all the supernatant, as the pH of
any residual liquid may affect the subsequent step.
17. Resuspend the membrane pellet in 300 μL 25 mM
NH4HCO3. Transfer to a 1.5 mL microfuge tube. Add
another 100 μL of 25 mM NH4HCO3 to the pellet to gather
any remaining membranes. Make sure that none of the pellet
sticks to the pipette following transfer to the microfuge tube.
18. Add Rapidgest to the cell pellet in the microfuge tube at a final
concentration of 0.1%.
19. Add Tris(2-carboxyethyl) phosphine (TCEP) to a final concen-
tration of 5 mM.
20. Incubate in the Thermomixer (750 rpm) for 30 min at 37  C.
21. Alkylate the membranes by the addition of iodoacetamide at a
final concentration of 10 mM for 30 min at 37  C in the
thermomixer.
22. Add 1 μg Lys-C and incubate for 4–6 h at 37  C in the
Thermomixer with shaking.
23. Add 20 μg proteomics grade trypsin to the solution at 37  C
for 16–24 h. Then repeat by a second addition of 20 μg prote-
omics grade trypsin for 24 h (see Note 17).
24. Inactivate proteases by heating to 100  C for 10 min or by the
addition of one drop of phosphoric acid, which will reduce the
pH to <2 (see Note 18).
25. Centrifuge the samples at 16,000
g for 10 min at room
temperature to remove particulates.
26. Perform the glycoprotein capture as described in Chapter 4
using 450 μL bead slurry of UltraLink Immobilized Streptavi-
din PLUS added to a MoBiCol spin columns.
27. Wash MoBiCol 5
with 100 mM NH4HCO3.
28. Add the sample to the prepared spin columns and incubate
with end over end rotation for 1 h at 25  C.
242 Vladimir Khristov et al.

29. Wash the beads sequentially with 10 mL of 0.05% Triton

X-100 in 100 mM NH4HCO3, 15 mL of 5 M NaCl twice,
15 mL 100 mM Na2CO3, and 15 mL 80% isopropanol to
remove nonspecific peptides and lipids.
30. Resuspend the beads in 450 μL of 100 mM NH4HCO3 and
add 500 units glycerol-free endoproteinase PNGaseF. Incubate
at 37  C for 16 h with end-over-end rotation to release the
peptides from the beads.
31. Spin off the liquid into a fresh tube. It is important to note that
the liquid now contains the cleaved peptides.
32. Add 450 μL of 0.1% TFA to the MoBiCol (Plug bottom and
cap) and rotate end over end for 5 min. Centrifuge at 2000
g
to collect the eluent and combine with the sample obtained in
step 31.
33. Utilize the C18 UltraMicroSpin column according to
Chapter 4 to clean the peptides prior to mass spectometry.
34. Collect the peptides in a clean microfuge tube by the addition
of 60 μL 90% acetonitrile–0.1% formic acid.
35. Dry the sample using a Speed Vac.

3.4.2 Mass 1. For each biological sample, analyze three technical replicates by
Spectrometry, Database liquid chromatography tandem mass spectrometry (LC-MS/
Search, and Data MS) on an LTQ-Orbitrap Velos or equivalent as previously
Processing reported [35, 36].
2. Use three independent search engines, Sequest, X!Tandem
[37], and Morpheus [38], followed by manual validation of
spectra as previously described [35]. Compare PSC (Protein
surface classification) to non-PSC surface proteomes and incor-
poration of transcriptome data using ProteinCenter™.
3. Confirm the unambiguous identification of an occupied glyco-
site based on the identified peptide meeting the following three
criteria: (1) peptide was captured by streptavidin beads, indi-
cating it was originally attached to a biotin-labeled oligosaccha-
ride structure, (2) the capture peptide contains a deamidation
(0.98 Da shift; conversion of N!D), and (3) the deamidation
occurred at asparagine within the N-glycosylation consensus
amino acid sequence motif (NxS/T). These criteria allow for
the efficient filtering of protein “contaminants” derived from
intracellular membranes and determination of the orientation
of transmembrane proteins within the membrane. Proteins that
do not meet these criteria are not analyzed further.

3.4.3 Bioinformatics 1. The above approach when applied to cultured human fetal RPE
to Identify Potential RPE cells has yielded over 1800 cell surface proteins. Of these,
Surface Markers 488 proteins have been detected only on the apical membrane,
 Polarized State of Human RPE 243

415 only on the basolateral membrane, while 979 were present

on both membranes (see Note 19).
2. In order to identify likely candidates for RPE-selective cell
surface markers, ascertain the following information for each
protein of interest: (1) presence and number of transmembrane
domains (PubMed protein database), (2) relative expression in
tissue of interest based on mRNA data (GEO database),
(3) protein identity (e.g., solute carriers, receptors, transpor-
ters, CD antigens, glycoproteins, enzymes) (PubMed protein
database).
3. Use positive and negative selection criteria to narrow putative
protein list.
(a) The RPE-gene signature dataset [23].
(b) Positively enriched in RPE as compared to fibroblasts.
(c) Expressed sequence tag (EST) profile data can be used to
positively select genes in expressed exclusively in adult
stages of development.
(d) Genes expressed in both early and late developmental
stages were negatively selected as they are likely not spe-
cific to a mature phenotype of RPE cells.

3.4.4 Validation of Cell 1. Scrape and resuspend cells from one T25 flask of primary
Surface Markers Through human cells using 1 mL Lysis Buffer.
Western Blot 2. Ensure complete cell lysis by passing the lysate through a 25 G
needle 100 times using a 1 mL syringe.
3. Leave lysate containing tube on ice for 20 min.
4. Centrifuge the nuclear pellet (P1) at 720
g for 5 min. The
nuclear pellet should be washed once by adding 500 μL frac-
tionation buffer again.
5. Disperse the pellet with a pipette and pass through a 25 G
needle 10 times.
6. Centrifuge again at 720
g for 10 min. Remove the fraction-
ation buffer and resuspend the nuclear pellet in the nuclear
buffer.
7. Sonicate the nuclear pellet briefly (3 s) on ice at a power setting
of 2-continuous to fractionate chromatin.
8. Remove the supernatant and place in a fresh 1.5 mL
microcentrifuge tube.
9. Centrifuge the supernatant again at 10,000
g. Move the
supernatant containing the cytosolic and membrane fractions
to a fresh 1.5 mL microcentrifuge tube.
244 Vladimir Khristov et al.

10. If the mitochondrial fraction is desired, save the pellet from

step 6, wash as with the nuclear pellet and resuspend in the
nuclear buffer.
11. Separate the cytosolic and membrane fractions by centrifuga-
tion of the supernatant at 100,000
g for 1 h. Resuspend the
membrane pellet in 100 μL nuclear buffer.
12. Analyze subcellular fractions via western blot (Figs. 11 and 12).

4 Notes

1. Paraformaldehyde used for sample fixation for immunostaining is

a known human carcinogen and potential reproductive hazard.
Sodium azide used in immunostaining is highly toxic and a
known mutagen.
2. When making the agar bridge, make sure to remove all the air
bubbles from the agar gel with ultrasonic bath.
3. For the CSC experiments, all reagents must be proteomics
grade and handled with great care. Gloves should be worn at
all times to prevent protein contamination of the samples, even
during reagent preparation. High purity H2O is required, and
all reagents should be dedicated for proteomics to avoid poten-
tial contaminations. Additional details on reagent preparation
can also be found in Chapter 4.
4. Always add DTT and protease inhibitors fresh to cell fraction-
ation buffers.
5. The following kinds of RPE cells can be used: primary human
fetal RPE (see refs. 7, 8, and for detailed procedures on how to
obtain and process primary human tissue for RPE culture);
pluripotent stem cell derived RPE (see ref. 9 for a detailed
protocol for pluripotent stem cell derived RPE differentiation).
RPE cells can replaced with any other epithelial cell type.
6. The number of times RPE cells can be passaged or frozen
without loss of RPE characteristics depends on RPE cell type.
In general, primary cultured and iPSC-derived RPE are only
passaged once.
7. The optimal immunostaining conditions for each antibody
have to be optimized. In some cases, antigen retrieval must
be used when immunostaining fixed tissues. The verification of
surface markers by immunostaining and western blot utilizes
large numbers of antibodies, it is time consuming and
expensive.
8. The punched RPE monolayer was mounted apical-side up
€
between two halves of the modified Ussing chamber designed
to maximize the rate of solution turnover in the apical and basal
 Polarized State of Human RPE 245

baths (10–20 chamber volumes per minute). The apical and

€
basal bath are completely separated in the Ussing chamber. The
area of tissue exposed to the apical and basal bathing solutions
was 3 cm in diameter. The temperature of the Ringer solution
in the chamber was maintained at 36  1  C using custom-
made heated water circulating system.
9. The CURRENT mode was used in the Voltage-Current Clamp
for both Electrophysiology and fluid transport experiments to
measure the transepithelial potential (TEP); Pass 2–4 μA bipo-
€
lar current pulse across the tissue mounted in the Ussing cham-
ber to measure the resultant changes in TEP to obtain the total
tissue resistance (RT).
10. The cytokine secretion should be quantitatively different on
the basal versus apical sides; usually VEGF is preferentially
secreted to the basal side (basal-to-apical ratio is at least two-
fold) and PEDF is preferentially secreted to the apical side.
11. The CSC technology captures N-glycosylated proteins on the
cell surface, including those that have only one N-glycosylation
site. One limitation of the procedure is that O-glycosylated and
nonglycosylated proteins are not captured.
12. Expanded details of the CSC-technology can be found in
Chapter 4; however, the analysis of polarized cells requires
labeling of only one plasma membrane and that the protocol
described here is for RPE cells and by extension any other
polarized cell cultivated on transwells.
13. Optimal concentration of sodium metaperiodate may vary
among cell types. 1 mM has worked well for fetal RPE cells;
however, for other RPE cells the optimal concentration of the
oxidizer must be determined empirically.
14. Due to the cell pigment, it is difficult to discriminate between
the pellet and the supernatant.
15. It is fine to leave some of the supernatant from the first
homogenization. The aim is to lyse the cells, and until the
pigment is diluted, it is almost impossible to distinguish
between the supernatant and the pellet generated from fetal
RPE cells.
16. If more than one ultracentrifuge tube has been used, it is
possible to combine all of the pellets at this step into a
single tube.
17. The trypsin reaction can proceed over the course of an entire
weekend.
18. The solution may become milky, but this is normal.
19. As the CSC-technology is not quantitative and is a sampling
based technique, it is impossible at this stage to know
246 Vladimir Khristov et al.

unequivocally whether the apical or basal membrane proteins

are restricted to only one of these membranes.

Acknowledgment

This work was supported by National Eye Institute Intramural

funds.

References
1. Strauss O (2005) The retinal pigment epithe- 11. Braisted JE, Essman TF, Raymond PA (1994)
lium in visual function. Physiol Rev 85 Selective regeneration of photoreceptors in
(3):845–881 goldfish retina. Development 120
2. Sparrow JR, Hicks D, Hamel CP (2010) The (9):2409–2419
retinal pigment epithelium in health and dis- 12. Baron R, Joseph RD, Owen T, Tennyson J,
ease. Curr Mol Med 10(9):802–823 Miller S, Ballester GE (1991) Imaging Jupiter’s
3. Chiba C (2014) The retinal pigment epithe- aurorae from H3þ emissions in the 3-4 micro-
lium: an important player of retinal disorders meters band. Nature 353:539–542
and regeneration. Exp Eye Res 123:107–114 13. Sonoda S, Sreekumar PG, Kase S, Spee C, Ryan
4. Reichhart N, Strauss O (2014) Ion channels SJ, Kannan R, Hinton DR (2010) Attainment
and transporters of the retinal pigment epithe- of polarity promotes growth factor secretion by
lium. Exp Eye Res 126:27–37 retinal pigment epithelial cells: relevance to
5. Lehmann GL, Benedicto I, Philp NJ, age-related macular degeneration. Aging
Rodriguez-Boulan E (2014) Plasma mem- (Albany NY) 2(1):28–42
brane protein polarity and trafficking in RPE 14. Maminishkis A, Jalickee S, Blaug SA, Rymer J,
cells: past, present and future. Exp Eye Res Yerxa BR, Peterson WM, Miller SS (2002) The
126:5–15 P2Y(2) receptor agonist INS37217 stimulates
6. Kay P, Yang YC, Paraoan L (2013) Directional RPE fluid transport in vitro and retinal reat-
protein secretion by the retinal pigment epithe- tachment in rat. Invest Ophthalmol Vis Sci 43
lium: roles in retinal health and the develop- (11):3555–3566
ment of age-related macular degeneration. J 15. Peterson WM, Meggyesy C, Yu K, Miller SS
Cell Mol Med 17(7):833–843 (1997) Extracellular ATP activates calcium sig-
7. Maminishkis A, Chen S, Jalickee S, Banzon T, naling, ion, and fluid transport in retinal pig-
Shi G, Wang FE et al (2006) Confluent mono- ment epithelium. J Neurosci 17(7):2324–2337
layers of cultured human fetal retinal pigment 16. Kevany BM, Palczewski K (2010) Phagocytosis
epithelium exhibit morphology and physiology of retinal rod and cone photoreceptors. Physi-
of native tissue. Invest Ophthalmol Vis Sci 47 ology (Bethesda) 25(1):8–15
(8):3612–3624 17. Li R, Maminishkis A, Banzon T, Wan Q,
8. Maminishkis A, Miller SS (2010) Experimental Jalickee S, Chen S, Miller SS (2009) IFN
models for study of retinal pigment epithelial {gamma} regulates retinal pigment epithelial
physiology and pathophysiology. J Vis Exp fluid transport. Am J Physiol Cell Physiol 297
45:2032 (6):C1452–C1465
9. Miyagishima KJ, Wan Q, Corneo B, Sharma R, 18. Li R, Wen R, Banzon T, Maminishkis A, Miller
Lotfi MR, Boles NC et al (2016) In pursuit of SS (2011) CNTF mediates neurotrophic factor
authenticity: induced pluripotent stem cell- secretion and fluid absorption in human retinal
derived retinal pigment epithelium for clinical pigment epithelium. PLoS One 6(9):e23148
applications. Stem Cells Transl Med 5 19. Mazzoni F, Safa H, Finnemann SC (2014)
(11):1562–1574 Understanding photoreceptor outer segment
10. Kivela T, Jaaskelainen J, Vaheri A, Carpen O phagocytosis: use and utility of RPE cells in
(2000) Ezrin, a membrane-organizing protein, culture. Exp Eye Res 126:51–60
as a polarization marker of the retinal pigment 20. Boheler KR, Bhattacharya S, Kropp EM,
epithelium in vertebrates. Cell Tissue Res 301 Chuppa S, Riordon DR, Bausch-Fluck D et al
(2):217–223 (2014) A human pluripotent stem cell surface
N-glycoproteome resource reveals markers,
 Polarized State of Human RPE 247

extracellular epitopes, and drug targets. Stem 30. Chen YY, Lu QJ, Lu QX, Wang NL (2011)
Cell Rep 3(1):185–203 Gene expression profile changes caused by the
21. Doerr A (2009) Snapshots of the cell surface. dysfunction of Mer during retinal pigment epi-
Nat Methods 6(6):401–401 thelium phagocytosis. Chin Med J 124
22. Wollscheid B, Bausch-Fluck D, Henderson C, (8):1145–1155
O’Brien R, Bibel M, Schiess R, Aebersold R, 31. Mandal MN, Vasireddy V, Jablonski MM,
Watts JD (2009) Mass-spectrometric identifi- Wang X, Heckenlively JR, Hughes BA, Reddy
cation and relative quantification of N-linked GB, Ayyagari R (2006) Spatial and temporal
cell surface glycoproteins. Nat Biotechnol 27 expression of MFRP and its interaction with
(4):378–386 CTRP5. Invest Ophthalmol Vis Sci 47
23. Strunnikova NV, Maminishkis A, Barb JJ, (12):5514–5521
Wang F, Zhi C, Sergeev Y et al (2010) Tran- 32. Lin H, Kenyon E, Miller SS (1992)
scriptome analysis and molecular signature of Na-dependent pHi regulatory mechanisms in
human retinal pigment epithelium. Hum Mol native human retinal pigment epithelium.
Genet 19(12):2468–2486 Invest Ophthalmol Vis Sci 33(13):3528–3538
24. Surace EM, Angeletti B, Ballabio A, Marigo V 33. Hollborn M, Petto C, Steffen A, Trettner S,
(2000) Expression pattern of the ocular albi- Bendig A, Wiedemann P, Bringmann A, Kohen
nism type 1 (Oa1) gene in the murine retinal L (2009) Effects of thrombin on RPE cells are
pigment epithelium. Invest Ophthalmol Vis Sci mediated by transactivation of growth factor
41(13):4333–4337 receptors. Invest Ophthalmol Vis Sci 50
25. Bachner D, Schroder D, Gross G (2002) (9):4452–4459
mRNA expression of the murine glycoprotein 34. Sugasawa K, Deguchi J, Okami T,
(transmembrane) nmb (Gpnmb) gene is linked Yamamoto A, Omori K, Uyama M, Tashiro Y
to the developing retinal pigment epithelium (1994) Immunocytochemical analyses of distri-
and iris. Brain Res Gene Expr Patterns 1 butions of Na, K-ATPase and GLUT1, insulin
(3–4):159–165 and transferrin receptors in the developing ret-
26. Adijanto J, Banzon T, Jalickee S, Wang NS, inal pigment epithelial cells. Cell Struct Funct
Miller SS (2009) CO2-induced ion and fluid 19(1):21–28
transport in human retinal pigment epithelium. 35. Gundry RL, Riordon DR, Tarasova Y,
J Gen Physiol 133(6):603–622 Chuppa S, Bhattacharya S, Juhasz O et al
27. Young A, Powelson EB, Whitney IE, Raven (2012) A cell surfaceome map for immunophe-
MA, Nusinowitz S, Jiang M et al (2008) notyping and sorting pluripotent stem cells.
Involvement of OA1, an intracellular GPCR, Mol Cell Proteomics 11(8):303–316
and G alpha i3, its binding protein, in melano- 36. Gundry RL, White MY, Murray CI, Kane LA,
somal biogenesis and optic pathway formation. Fu Q, Stanley BA, Van Eyk JE (2009) Prepara-
Invest Ophthalmol Vis Sci 49(7):3245–3252 tion of proteins and peptides for mass spec-
28. Stamer WD, Bok D, Hu J, Jaffe GJ, McKay BS trometry analysis in a bottom-up proteomics
(2003) Aquaporin-1 channels in human retinal workflow. Curr Protoc Mol Biol Chapter 10:
pigment epithelium: role in transepithelial Unit10.25
water movement. Invest Ophthalmol Vis Sci 37. Craig R, Beavis RC (2004) TANDEM: match-
44(6):2803–2808 ing proteins with tandem mass spectra. Bioin-
29. Lambooij AC, van Wely KH, Lindenbergh- formatics 20(9):1466–1467
Kortleve DJ, Kuijpers RW, Kliffen M, Mooy 38. Wenger CD, Coon JJ (2013) A proteomics
CM (2003) Insulin-like growth factor-I and search algorithm specifically designed for
its receptor in neovascular age-related macular high-resolution tandem mass spectra. J Prote-
degeneration. Invest Ophthalmol Vis Sci 44 ome Res 12(3):1377–1386
(5):2192–2198
 Chapter 16

Extracellular Matrix Molecule-Based Capture

of Mesenchymal Stromal Cells Under Flow
Teresa Massam-Wu, Stuart A. Cain, and Cay M. Kielty

Abstract
We present a method to capture mesenchymal stromal cells (MSCs) by adhesion to extracellular matrix
(ECM) molecules under flow conditions. The technique simulates a physiological system and exploits the
natural biological interactions of cells, through integrin receptors, with their ECM. The system offers an
insight into how MSCs could be targeted/localized to the site of interest (graft) following intravenous
injection.

Key words Fibronectin, Laminin, Adhesion, Mesenchymal stromal cell

1 Introduction

This method provides a new approach to capture MSCs that

exploits adhesive extracellular matrix (ECM) molecules and their
specific interactions with integrin receptors. The method has the
potential to select MSC subpopulations in flow according to their
integrin receptor complement and the specific ECM coatings
selected for capture. It can also be exploited to seed MSCs, which
can express endothelial markers [1], on defined ECM-coated scaf-
folds in flow, e.g., for preconditioning vascular grafts.
The method described here takes place in seven steps: (1) Clas-
sification of MSCs by flow cytometry; (2) live cell imaging of MSCs
captured by laminin-521 and FN III 7–14 under flow conditions;
(3) DAPI staining of immobilized cells; (4) quantitation of immo-
bilized cells; (5) RNA extraction; (6) generating AVI videos of
images; and (7) analysis of captured cells. Each step is described in
detail and Notes are provided to assist the reader with technical
hurdles.

Electronic supplementary material: The online version of this chapter (https://doi.org/10.1007/978-1-4939-

7553-2_16) contains supplementary material, which is available to authorized users.

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_16, © Springer Science+Business Media, LLC 2018

249
250 Teresa Massam-Wu et al.

2 Materials

2.1 Ibidi Equipment 1. ibidi Pump System (catalog number 10902) (ibidi GmbH,
Munich, Germany).
2. Fluidic unit (catalog number 10903) (ibidi).
3. ibidi Pump (catalog number 10905) (ibidi).
4. PumpControl software (catalog number 10908) (ibidi).

2.2 Ibidi Labware 1. ibidi μ-slides I 0.6 Luer, tissue culture treated and sterilised
(catalog number 80186) (ibidi).
2. Perfusion set WHITE (catalog number 10963) (ibidi) (see
Note 1).
3. Dehydrated silica beads (see Note 2).

2.3 Microscopes 1. Live cell imaging microscope. Images were acquired on a Zeiss
and Camera Axiovert 200M live cell-imaging microscope (Carl Zeiss AG,
Oberkochen, Germany) under brightfield with an LED light
source (Zeiss Colibri) using a 20
/0.30 LD A-Plan objective
(Zeiss). Images were taken at 500 ms intervals on a CoolSnap
HQ2 CCD camera (Photometrics).
2. High content imaging system Pathway Bioimager 855 (Beck-
ton Dickenson and Company, Sparks, MD, USA). Images were
acquired on a Pathway Bioimager 855 (Beckton Dickenson)
with laser autofocus using a 4
/0.13 UPlan FLN objective
and the following filter setup: DAPI Ex. 360/10, Di. 400D
CLP, Em. 435 LP. Images were collected from the entire chan-
nel on an Orca ER camera (Hammamatsu Photonics, Hama-
matsu, Japan) with an offset from the channel center and a
montage of 3
26 was created without gaps. An exposure time
of 0.6 ms and threshold masks were applied to each image
using the automatic feature of the software. The images were
then processed and analyzed through the BD Data Explorer
software.

2.4 Other Equipment 1. CASY cell counter (Roche Diagnostics, Indianapolis, IN, USA)
(see Note 3).
2. Vacuum desiccator.
3. NanoDrop 2000 spectrophotometer (ThermoFisher Scientific,
Carlsbad, CA, USA).
4. HiSeq2500 System (Illumina, San Diego, CA, USA).
5. Beckman Coulter Cyan ADP (Beckman Coulter, Inc., Brea,
CA, USA). Samples were analyzed on a Beckman Coulter Cyan
ADP running Summit software version 4.3. Both Alexa Fluor
488 and Phycoerythrin were excited using the 488 nm laser.
 ECM-Based Capture of MSCs 251

Alexa Fluor 488 was detected using a 530/40 nm bandpass

filter and Phycoerythrin was detected using a 575/25 nm
bandpass filter. The gating used was forward scatter versus
pulse width to select single cells and forward scatter versus
side scatter to exclude dead cells and debris. Data were col-
lected on 10,000 gated events.

2.5 Image Analysis 1. Download Fiji image analysis software from http://fiji.sc/Fiji.
Software [2] 2. Download a media player such as VLC Media player from
http://www.videolan.org/vlc/index.en_GB.html.

2.6 Mesenchymal 1. Human MSCs (Catalog number PT-2501) (Lonza, Inc., Allen-
Stem Cells (MSCs) dale, NH, USA).

2.7 Extracellular 1. Recombinant histidine tagged FN III 7–14 molecular frag-

Matrix (ECM) ment was expressed in 293-EBNA cells using a modified
Molecules— pCEP-His vector, and purified as previously described by
Recombinant [3, 4].
Fibronectin Type III 2. Laminin-521 (catalog number LN-521) (Biolamina AB, Sund-
Repeats 7–14 (FN III byberg, Sweden).
7–14) and Laminin-521

2.8 Reagents 1. Phosphate-buffered saline (PBS; 1
): 137 mM NaCl, 2.7 mM

KCl, 1.47 mM KH2PO4, 8.1 mM Na2HPO4, pH 7.4.
Optional 1 mM CaCl2·2H2O, 0.5 mM MgCl2·6H2O (see
Note 4).
2. Bovine serum albumin (BSA) Fraction V (catalog number
A3059) (Sigma-Aldrich Chemical Company, St. Louis, MO,
USA): 10 mg/mL in 1
PBS without CaCl2/MgCl2.
3. MesenPRO RS Medium (catalog number 12746-012) (Life
Technologies, Grand Island, NY, USA).
4. StemPro Accutase (catalog number A11105-01) (Life
Technologies).
5. 4% Paraformaldehyde (PFA): Prepare 4% PFA in a fume hood.
Add 4 g PFA powder to 80 mL 1
PBS in a conical flask. Heat
to approximately 60  C with moderate stirring. Add 1 N
NaOH dropwise to clear the solution. Allow to cool, and filter
the solution. Check the pH and adjust pH to 6.9 with a few
drops of dilute HCl. Top up to 100 mL with 1
PBS. Aliquot
into 15 mL tubes, label and date. Solution can be kept at 4  C
for up to 1 month or at 20  C for long-term storage.
6. 40 ,6-Diamidino-2-Phenylindole (DAPI) nucleic acid stain (cat-
alog number D1306) (Life Technologies): 1:10,000 in PBS
with CaCl2/MgCl2.
252 Teresa Massam-Wu et al.

7. ProLong Gold antifade reagent without DAPI (catalog num-

ber 36934) (Life Technologies).
8. ReliaPrep RNA Cell Miniprep System (catalog number Z6011)
(Promega).

2.9 Antibodies MSCs for flow cytometry were labeled with either unconjugated
primary antibodies:
1. Mouse IgG2b κ Isotype Control (catalog number 555743).
2. CD14 [MP9] (catalog number 347490).
3. CD45 [2D1] (catalog number 347460).
4. CD73 [AD2] (catalog number 550256).
5. CD166 (catalog number 559260) (BD Pharmingen).
6. CD34 (catalog number CBL555) (Chemicon).
7. CD90 [5E10] (catalog number 328102) (Biolegend) and sec-
ondary Alexa Fluor 488 Donkey anti-Mouse IgG antibody
(catalog number A-21202) (Life Technologies)
(or)
8. Primary Stro-1 antibodies (catalog number MAB1038) (R&D
Systems) and secondary Alexa Fluor 488 Goat anti-Mouse IgM
antibody (catalog number A-21042) (Life Technologies)
or
9. Directly conjugated Phycoerythrin antibodies: PE Mouse
IgG2b κ Isotype Control (catalog number), CD29 (catalog
number 555443), CD44 [G44-26] (catalog number 555479)
(BD Pharmingen), CD105 [166707] (catalog number
FAB10971P) (R&D Systems).

3 Methods

3.1 Classification Classification conforms to the criteria defined by the Mesenchymal

of MSCs by Flow and Tissue Stem Cell Committee of the International Society for
Cytometry Cellular Therapy [5].
1. Culture MSCs on 5% (v/v) gelatin coated tissue culture flasks
in MesenPRO medium (þ2%(v/v) growth supplement, 1%
(v/v) penicillin-streptomycin, 1% (v/v) L-glutamine) until
80% confluent.
2. Dissociate MSCs using StemPro Accutase, neutralize and har-
vest cell pellet by centrifugation at 250–450
g for 4 min at
room temperature.
3. Resuspend cell pellet in 7 mL media containing MesenPRO
media (þ2% growth supplement, 1% penicillin-streptomycin,
1% L-glutamine), transfer to a 15 mL polystyrene tube and
 ECM-Based Capture of MSCs 253

incubate upright at 37  C for 30 min to allow cell surface

receptors to recover.
4. From this point onward, keep everything ice-cold and use PBS
without CaCl2/MgCl2.
5. Harvest cells at 250–450
g for 4 min and resuspend cell pellet
in 1 mL ice-cold 1
PBS/0.5% (w/v) BSA.
6. Repeat step 4 a further two times.
7. Count cells and aliquot 7.5
104 cells into a 96 well tissue-
culture treated plate (see Note 5).
8. Harvest cell pellets by centrifugation at 250–450
g for 3 min
and resuspend in 100 μL 1
PBS/0.5% (w/v) BSA.
9. Add unconjugated primary antibody at 1:100–500 dilutions in
1
PBS/0.5% (w/v) BSA and incubate on ice for 1 h.
10. Harvest cells by centrifugation at 250–450
g for 3 min.
11. Wash cell pellet by adding 100 μL ice-cold PBS/0.5% (w/v)
BSA (see Note 6).
12. Repeat steps 10 and 11 three times.
13. Add conjugated primary antibody or secondary antibody at a
1:200 dilution in PBS/0.5% (w/v) BSA and incubate on ice for
45 min.
14. Centrifuge cells at 250–450
g for 3 min and resuspend cell
pellet in 200 μL ice-cold PBS/0.5% (w/v) BSA).
15. Repeat step 14 three times.
16. Resuspend pellet in a final volume of 400 μL ice cold
PBS/0.5% BSA and transfer to a 1.5 mL microtube (see Note
7).
17. Store samples at 4  C until ready for flow cytometry.
18. Run samples in a core facility or under the guidance of a
knowledgeable flow cytometrist. Once data are acquired, anal-
ysis can be performed. An example of anticipated data is shown
in Fig. 1.

3.2 Live Cell Imaging All experiments conducted at 37  C unless otherwise stated.
of MSCs Captured by
1. Dehydrate the silica beads in the drying bottle (orange) (see
Laminin-521 and FN III Note 2).
7–14 Under Flow
Conditions
2. Coat μ-slides I (600 μm channel height) Luer with 150 μL of
100 nM ECM molecules in PBS with or without CaCl2/
MgCl2 (see Note 4) overnight at 4  C (see Note 8).
3. Block unbound surfaces by gently exchanging the protein
solution with 10 mg/mL BSA for 1 h at room temperature
(see Note 9).
254 Teresa Massam-Wu et al.

A B

Alexa Fluor 488 -

Alexa Fluor PE- conjugated Control
Negave Posive conjugated
488- CD73
markers conjugated markers Control
CD90
Control CD29
CD44 CD166
CD14
CD34 CD105 Stro-1
CD45

C
BODIPY DAPI Merged
Control
Adipogenesis

D Osteocalcin DAPI Merged

Control
Osteogenesis
 ECM-Based Capture of MSCs 255

4. Wash once with 250 μL PBS without CaCl2/MgCl2 and leave

in this solution until ready to use.
5. Degas 50 mL PBS and 50 mL medium in a vacuum desiccator
for 30 min prior to use (see Note 10).
6. Prewarm the PBS, serum free media, perfusion set, fluidic unit
and live cell imaging system in advance at 37  C, and where
possible in 5% CO2, for 2 h prior to experiment (see Note 11).
7. Set the live cell imaging microscope to focus at the center of the
channel using a 20
objective under brightfield (see Note 12).
8. Prepare the acquisition software to capture images at 500 ms
intervals.
9. Mount the perfusion set onto the fluidic unit as described by
ibidi [6], but do not attach the channel slide (see Fig. 2).
10. Fill and equilibrate perfusion set with 11.7 mL degassed serum
free medium (see Note 13).
11. Prepare a cell suspension of 1
106 MSCs by dissociation with
StemPro Accutase (see Note 14).
12. Resuspend cell suspension in 1 mL serum free MesenPRO
media and incubate upright at 37  C for at least 30 min to
allow cell surface receptors to recover.
13. Add 1
106 of the MSC suspension to the perfusion set and
equilibrate using the manual control panel on the PumpCon-
trol software. Apply a low pressure and alternate the flow
direction by switching the valves to distribute the cells.
14. Ensure the PumpControl software is not running, clamp both
tubes of the perfusion set near the connectors and attach the
channel slide (see Note 15).
15. Set a flow rate of 0.65 mL/min (see Note 16) under positive
pressure, ensure the correct perfusion set color and slide are
selected in the PumpControl software, set timer to 10 min,
release clamp and start image acquisition and run the program.
16. Acquire images for 10 min.

ä
Fig. 1 Classification of MSCs. Human mesenchymal stem cells (MSCs) expressing (a) negative (CD14, CD34,
and CD45) and (b) positive markers (CD29, CD44, CD73, CD90, CD105, CD166, and Stro-1) by flow cytometry
that conform to the criteria defined by the Mesenchymal and Tissue Stem Cell Committee of the International
Society for Cellular Therapy [5]. IgG1 and IgM control peaks are highlighted in red. (c) Adipogenic and (d)
Osteogenic differentiation of human mesenchymal stromal cells (MSCs) in vitro for 28 days using the human
MSC functional identification kit (catalog number SC006) (R&D Systems). Adipogenic MSCs were stained with
BODIPY 493/503 (green), Osteogenic MSCs with osteocalcin (green) and nuclei stained with DAPI (blue).
Un-treated Human MSCs were cultured in basal medium for 28 days. Scale bar ¼ 50 mm. Representative data
and images from two independent experiments
256 Teresa Massam-Wu et al.

positive pressure

Direction of flow

Extracellular matrix
Mesenchymal stem
molecule coated
cells in suspension
surface

Fig. 2 Schematic illustration of the ibidi fluidic pump system under positive pressure. Adapted from ibidi

17. Reattach the clamp to the tubing and disconnect the slide from
the perfusion system.
18. Remove unattached cells in the channel by gently washing once
with 500 μL PBS without CaCl2/MgCl2 (see Note 9).

3.3 DAPI Staining 1. Immobilize cells with 200 μL 4% paraformaldehyde (PFA) at

of Immobilized Cells room temperature for 20 min.
(see Fig. 3a) 2. Stain nuclei of cells with DAPI for 1 h at room temperature.
3. Add 150 μL ProLong Gold antifade reagent without DAPI to
preserve cells.

3.4 Quantitation 1. Image the number of cells attached imaged using a BD Path-
of Immobilized way microscope with a 4
objective (see Note 17).
Cells (see Fig. 3b) 2. Quantify the number of cells attached using Fiji image analysis
software.
3. Open the TIFF image in Fiji.
4. Go to IMAGE then ADJUST the THRESHOLD.
5. Slide bar to the LEFT or RIGHT to ensure nuclei are distinct
then go to APPLY.
 ECM-Based Capture of MSCs 257

Fig. 3 Quantitation of mesenchymal stromal cell (MSC) attachment on the entire

length of the channel slide with 100 nM of immobilized FN III 7–14 or Laminin-
521. (a) A representative image of immobilized MSCs with DAPI stained nuclei on
(i) 100 nM FN III 7–14 and (ii) 100 nM Laminin-521. (b) Quantitation of the total
number of cells immobilized on an entire channel. N ¼ 2. A greater number of
MSCs are captured and attached on laminin-521 immobilized channels. Statis-
tical analysis using unpaired t-tests where *P < 0.05

6. Go to PROCESS then BINARY then MAKE BINARY.

7. Go to PROCESS then BINARY then WATERSHED (see
Note 18).
8. Go to ANALYZE the ANALYZE PARTICLES.
9. Input the following parameters: Size (inch^2) of 10–1000,
select “Pixel units,” Circularity of 0.40–1.00, show “Overlay
Outlines,” select “Display results,” select “Clear results” and
select “Include holes”.

3.5 RNA Extraction 1. Extract the RNA as per manufacturer’s instructions (see
Note 19).
2. Quantify and assess the purity of the extracted RNA sample
using a NanoDrop spectrophotometer (see Note 20).

3.6 Generating AVI Videos are generated from images from the first 7 min of a 10 min
Videos of Images experiment and are 2.5
the speed of real time.
Captured in a 10 min
1. Open Fiji imaging software.
Experiment
2. DRAG and DROP file of images onto  symbol.
258 Teresa Massam-Wu et al.

3. TICK ‘use virtual stack’ then OK.

4. Go to FILE, NEW then SCRIPT.
5. Cut and paste the command file below.

run("Re-order Hyperstack ...", "channels¼[Channels (c)] slices¼

[Frames (t)]
frames¼[Slices (z)]");
run("8-bit");
run("Grays");
run("Time Stamper", "starting¼0 interval¼0.5 x¼5 y¼22
font¼18 ’00 decimal¼0 anti-aliased or¼[]");
setFont("SansSerif", 18, " antialiased");
makeText("Condition", 5, 490);
run("Draw", "stack");
run("AVI... ", "compression¼PNG frame¼5 save¼[]");

6. Go to LANGUAGE then select IJ1 MACRO.

7. Relabel “Condition” to text of choice (see Note 21).
8. Change “frame ¼ ” to number of choice depending on speed of
video required (see Note 22).
9. Select RUN.
10. Save the AVI file.
11. Open the AVI file with an appropriate media player such as
VLC media player.

3.7 Analysis Cells captured can be analyzed by quantitative polymerase chain

of Captured Cells reaction (qRT-PCR), RNA-sequencing (RNA-Seq) and at the pro-
tein level, using standard protocols.

4 Notes

1. Select perfusion set color appropriate to the shear stress rate

required for your model system.
2. Silica beads turn white when saturated with moisture. To dehy-
drate: place Silica beads in a 100 mL glass Duran bottle and put
into a drying oven at 120  C for at least 8 h. Beads will turn
orange. Allow beads to cool to room temperature.
3. A hemocytometer will suffice.
4. Addition of CaCl2 is dependent on protein requirement for
calcium.
5. Include sufficient aliquots for controls.
 ECM-Based Capture of MSCs 259

6. Do not disrupt pellet by trituration as the addition of the PBS

will be sufficient to disrupt the pellet.
7. Fix cells in 1% formaldehyde/PBS (without CaCl2/MgCl2)
and store at 4  C for up to 2 weeks.
8. Set up slides in multiples where necessary if RNA extraction is
required due to the low concentration of RNA obtained from
the low cell density.
9. Do not aspirate the channel dry. Leave solution in channel and
aspirate by using a 1 mL syringe to push liquid through chan-
nel. This minimizes air bubble formation.
10. Unscrew caps on media vessels to prevent implosion during
vacuum.
11. This helps to reduce air bubble formation in the tubing during
operation.
12. A line in the channel can be seen under the microscope to mark
the center of the channel.
13. If sterile conditions are required, the perfusion set mounted on
the fluidic unit can be disconnected from the pump and filled
with medium in a biological safety cabinet.
14. This is a gentle dissociation reagent that helps maintain cell
surface receptors integrity.
15. Ensure the reservoirs of the slide are filled completely to pre-
vent air bubbles entering the system.
16. Select a flow rate appropriate to the shear stress rate required.
17. Cells attach to the top and bottom of the chamber so ensure
both Z planes are analyzed.
18. Watershed segmentation is a way of automatically separating
particles that touch.
19. RNA should be pooled from three or more channels slides in
order to have sufficient RNA for cDNA synthesis.
20. The ratio of absorbance at 260 and 280 nm of ~2.0 is generally
accepted as “pure” for RNA.
21. Include information on cell type, protein and channel height.
22. For example, if each frame is 500 ms, then changing this
parameter to 10 is equivalent to 5
the speed of real time.

Acknowledgment

This work was supported by the UK Engineering and Physical

Sciences Research Council (EPSRC). The Bioimaging Facility
microscopes used in this study were purchased with grants from
the UK Biotechnology and Biological Sciences Research Council
260 Teresa Massam-Wu et al.

(BBSRC) and the Wellcome Trust, and the University of Manche-

ster Strategic Fund. The Beckman Coulter Cyan ADP used in this
study was purchased with a grant from Astra Zeneca. Special thanks
to Dr. Peter March for his help with the microscopy, Dr. Christoph
Ballestrem and Dr. Alex Carisey for their help with the ibidi pump
system, and Mr. Mike Jackson for his help with the flow cytometry.

References
1. Ball SG, Worthington JJ, Canfield AE, Merry (2005) Homotypic fibrillin-1 interactions in
CLR, Kielty CM (2014) Mesenchymal stromal microfibril assembly. J Biol Chem 280
cells: inhibiting PDGF receptors or depleting (6):5013–5021
fibronectin induces mesodermal progenitors 5. Dominici M, Le Blanc K, Mueller I, Slaper-
with endothelial potential. Stem Cells 32 Cortenbach I, Marini FC, Krause DS et al
(3):694–705 (2006) Minimal criteria for defining multipotent
2. Schindelin J, Arganda-Carreras I, Frise E, mesenchymal stromal cells. The International
Kaynig V, Longair M, Pietzsch S et al (2012) Society for Cellular Therapy position statement.
Fiji: an open-source platform for biological- Cytotherapy 8(4):315–317
image analysis. Nat Methods 9(7):676–682 6. Kroutvar M, Guttenberg Z (2013) Ibidi Pump
3. Cain SA, Baldock C, Gallagher J, Morgan A, Bax System Instructions version 1.5.1, revision
DV, Weiss AS, Shuttleworth CA, Kielty CM 3. http://ibidi.com/fileadmin/products/
(2005) Fibrillin-1 interactions with heparin: instruments/I_1090X_PumpSystem/IN_
implications for microfibril and elastic fiber 1090X_pump_system.pdf. Accessed 09 Jan
assembly. J Biol Chem 280(34):30526–30537 2015
4. Marson A, Rock MJ, Cain SA, Freeman LJ,
Morgan A, Mellody K, Shuttleworth CA et al
 Chapter 17

Generation of Induced Pluripotent Stem Cells from Patients

with COL3A1 Mutations and Differentiation to Smooth
Muscle Cells for ECM-Surfaceome Analyses
Jiaozi He, Zhihui Weng, Stanley Chun Ming Wu, and Kenneth R. Boheler

Abstract
Use of experimentally derived induced pluripotent stem cells (iPSCs) has led to the development of cell
models for differentiation, drug testing and understanding disease pathogenesis. For these models to be
informative, reprogrammed cell lines need to be adequately characterized and shown to preserve all of the
critical characteristics of pluripotency and differentiation. Here, we report a detailed protocol for the
generation of iPSCs from human fibroblasts containing mutations in COL3A1 using a Sendai virus
mediated integration-free reprogramming approach. We describe how to characterize the putative iPSCs
in vivo and in vitro to ensure potency and differentiation potential. As an example of how these mutations
may affect cell surface and extracellular matrix (ECM) interactions, we provide protocols for the differenti-
ation of these cells into smooth muscle cells to illustrate how different cell types may display cell autono-
mous differences in collagen receptors that may affect their phenotype. These cells, when applied to
mechanical model systems (see Chapter 18 by Bose et al.) facilitate an assessment of stiffness and stress-
strain relationships useful for understanding how extracellular matrix dysfunction and its interactions with
surface proteins contribute to disease processes.

Key words Induced pluripotent stem cells, Reprogramming, Differentiation, Smooth muscle cells,
Collagen, Integrin

1 Introduction

Experimentally derived induced pluripotent stem cells (iPSC) with

the capacity to self-renew indefinitely and to differentiate into all
somatic cell types has led to the development of unique cell models
for drug testing and for understanding disease pathogenesis. The
generation of iPSC was first described in mouse in 2006 [1] and in
human in 2007 [2] through the forced expression of four transcrip-
tion factors. Generation of early iPSC lines depended on the use of
lentiviral vectors, which through integration could disrupt the

Jiaozi He and Zhihui Weng contributed equally to this work.

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_17, © Springer Science+Business Media, LLC 2018

261
262 Jiaozi He et al.

genome and cause other genetic abnormalities [3–5]. Moreover,

the use of c-MYC, which can be protumorigenic, is thought to be
detrimental to the long-term use of these cells [6]. Protocols that
do not rely on integrating vectors were developed to overcome
limitations associated with insertion events [7–13]. Of particular
interest is the use of modified RNAs and Sendai virus for repro-
gramming, as both are nonintegrating and may be apt for use using
good manufacturing practice (GMP) conditions suitable for thera-
peutic applications. While technological advances have been made
for the generation of iPSC lines and the potential treatment of
human conditions, human iPSC are still used mainly for basic
research and in vitro disease modeling.
The study of human diseases caused by gene defects in human
iPSC, where animal models may inadequately mimic the human
condition, is particularly valuable when trying to unravel disease
processes. One such disease is Ehlers–Danlos Syndrome (EDS), a
heritable disease of connective tissues involving collagen type III
alpha 1 (COL3A1), collagen type V alpha 1 (COL5A1), tenascin-X,
and to a lesser extent collagen type I alpha 1 (COL1A), lysyl
hydroxlase, and ADAM metallopeptidase. These diseases can
cause major defects in skin, musculoskeletal and cardiovascular
tissues, and depending on the mutations, cause hypermobility, eye
defects, scoliosis (spine), digestive disorders, intestinal and uterine
fragility, and cerebrospinal fluid leaks [14–17]. The most severe
form of this disease is caused by mutations in the gene encoding
COL3A1 [18, 19].
Mutations in COL3A1 lead to the development of vascular
EDS (vEDS), which is characterized by altered synthesis and struc-
ture of the pro α1 (III) chain of collagen type III to cause vascular
wall weakness. Patients with this disease are susceptible to aneur-
ysms, vascular tears and other defects. Here, we report a detailed
protocol for the generation of iPSCs from human fibroblasts con-
taining mutations in COL3A1 using a Sendai virus mediated
integration-free reprogramming approach, modified from a proto-
col originally described by Daheron and D’Souza [20], and opti-
mized for skin fibroblasts. Methods are described showing how to
cultivate (Subheading 3.1) and reprogram fibroblasts (Subheading
3.2), and how to select (Subheading 3.3) and freeze (Subheading
3.4) putatively reprogrammed cells. This is followed by character-
izations of putative hiPSC clones, including analyses of undifferen-
tiated cells (Subheading 3.5: PCR analysis; Subheading 3.6:
immunocytochemistry), differentiated cells (Subheading 3.7: non-
directed differentiation, differentiation to specific lineages and ter-
atoma assays); directed differentiation to paraxial smooth muscle
cells (SMCs) (Subheading 3.8); and then transcript analyses of
surface protein and extracellular matrix proteins (Subheading
3.9). Our modified Cheung et al. differentiation protocol permits
generation of smooth muscle cells originally cultivated in E8
 Human iPS Cells to Model 263

medium [21], and the resulting cells are useful for the study of
collagen and cell surface collagen receptors. Depending on the
cultivation medium, these SMCs can be either proliferative or
contractile. The latter cells, when applied to the three dimensional
model system described in the next chapter of this book (Fabrica-
tion and mechanical properties measurements of 3D microtissues for
the study of cell-matrix interactions), allow for an assessment of
stiffness and stress-strain relationships useful for understanding
how extracellular matrix dysfunction and its interactions with sur-
face proteins contribute to disease processes.

2 Materials

Prepare all solutions using tissue culture grade reagents unless

otherwise indicated. Sterile tissue culture supplies (disposable
plates, serological pipettes, tubes, and tips) should be purchased
from suitable vendors. Maintain aseptic conditions throughout all
stages of reprogramming and differentiation, and dispose of all
plastics and wastes that come into contact with human cells accord-
ing to institute and local safety guidelines.

2.1 Equipment 1. Laminar flow cabinet.

and Basic Supplies 2. 37
C water bath.
3. Heating block capable of reaching 95
C.
4. 4
C refrigerator or cold room.
5. Phase-contrast inverted microscope with low objective lenses
(4, 10, and 20).
6. Fluorescent microscope with filters compatible for use with the
DNA staining dye DAPI and antibody-conjugated fluoro-
phores. Examples include the LSM Carl Zeiss 510 Meta or
Nikon Eclipse TiS microscope.
7. Dissecting scope.
8. Automated cell counting system like the Countess (Thermo-
Fisher Scientific, Waltham, MA, USA) or a hemocytometer.
9. Shaker.
10. Microcentrifuge.
11. Vortex.
12. Liquid Nitrogen storage unit suitable for storing cells in
cryotubes.
13. Quantitative Polymerase Chain Reaction (PCR) Equipment
(6-well format), like the ABI Prism® 7900.
14. NanoDrop Spectrophotometer or other suitable spectrophoto
meter.
264 Jiaozi He et al.

15. Serological pipets (5, 10, and 25 mL).

16. Aspirating pipettes: sterilized 9 in. Pasteur pipettes.
17. Sterile conical polypropylene tubes (15 and 50 mL), similar to
those from Fisher Scientific or Falcon with a screw top cap and
suitable for centrifugation.
18. 0.45 μm filter units suitable for filtering tissue culture reagents.
The membranes should be suitable for use with tissue culture
reagents.
19. Syringe filters (0.45 μm). The membranes should be suitable
for use with tissue culture reagents.
20. A vacuum system that can be used to apply suction to the filter
units.
21. Microcentrifuge tubes (1–2 mL), sterile.
22. PCR tubes or plates, depending on equipment. Appropriate
caps or seals to accompany these tubes or plates should be used.
23. Nunc CryoTubes cryogenic vials, 1 mL (Sigma Aldrich Chem-
ical Company, St. Louis, MO, USA).
24. Glass coverslips or MatTek dishes (MatTek Corporation, Ash-
land, MA, USA).
25. 70% ethanol in a spray bottle.

2.2 Solutions, 1. Phosphate buffered saline (PBS, catalog number 10010-

Reagents and Kits 023; ThermoFisher Scientific).
2. Dulbecco’s phosphate buffered saline (D-PBS) lacking calcium
and magnesium (catalog number 141190-144; ThermoFisher
Scientific).
3. Chemically Defined Lipid Concentrate (CDLC) (catalog num-
ber 11905031; ThermoFisher Scientific). Store at 4
C.
4. Polyvinyl alcohol (PVA) (catalog number P8136; Sigma-
Aldrich).
5. Donkey Serum (catalog number D9663) and Goat Serum
(catalog number G9023) (Sigma-Aldrich).
6. 4% Paraformaldehyde solution (PFA), frozen as aliquots
(in microfuge tubes or 15 mL conical tubes) and stored with
minimal exposure to room air, i.e., minimize the amount of air
space in the tube to prevent oxidation.
7. DAPI (40 ,6-Diamidino-2-Phenylindole, Dihydrochloride). To
make a 5 mg/mL DAPI stock solution, dissolve 10 mg in 2 mL
of double distilled water. As this solution is not very water
soluble, it may take time to completely dissolve the substrate.
The working solution for staining of cells should be diluted in
water to a final concentration of 1 μg/mL and stored at 4
C
for up to 6 months.
 Human iPS Cells to Model 265

8. CytoTune™-iPS 2.0 Sendai Reprogramming Kit (catalog

number A16517; ThermoFisher Scientific).
9. Dimethyl sulfoxide (DMSO), tissue culture grade.
10. HCl/BSA, 4 mM/0.2% (w/v): Add 16.7 μL 37% HCl (12 N)
and 100 mg BSA into 50 mL ddH2O, and filter-sterilize.
11. Y27632 (RI), selective Rho-associated kinase p160 ROCK
inhibitor. To prepare 10 mM stocks, add 3 mL of ultrapure
water to a 10 mg vial. Mix thoroughly. Aliquot the stock
solution into working volumes based on the anticipated use.
Store aliquots at 20
C. Multiple freeze-thaws are not recom-
mended. Once thawed, aliquots may be kept at 4
C for
2 weeks. The final working concentration of the Y27632 is
10 μM in cell culture medium.
12. Recombinant Activin A Protein (Activin A) (catalog number
338-AC-010; R&D Systems, Minneapolis, MN, USA). Briefly
spin solution and reconstitute in ultrapure water to a concen-
tration of 50 μg/mL. This solution can be stored at 4
C for up
to 1 month. Further working dilutions can be made in PBS
containing 2 mg/mL albumin, aliquoted and stored at
20
C.
13. Recombinant Human Bone Morphogenetic Protein-4
(BMP4) (catalog number 314-BP) (R&D Systems), 10 μg/
mL. Dissolve 50 μg of BMP4 into 5 mL of sterile 4 mM HCl
containing at least 0.1% (w/v) BSA. Aliquot and store at
80
C for up to 3 months or 4
C for 1 month.
14. Fibroblast Growth Factor Basic (FGF2) (catalog number
PHG0023) (ThermoFisher Scientific) 10 μg/mL. Dissolve
25 μg into 2.5 mL sterile DPBS containing at least 0.1%
(wt/vol) BSA. Aliquot and store at 80
C for up to 3 months
or 4
C for 1 month. Use within five freeze–thaw cycles.
15. LY 294002 PI3-kinase inhibitor (catalog number V1201; Pro-
mega, Madison, WI, USA). Store dried material at 20
C in
the dark. Allow vial to come to room temperature before
resuspension. To make a 10 mM stock solution, dissolve
1 mg into 290.9 μL DMSO. Aliquot and store at 80
C for
up to 12 months.
16. Recombinant human TGF-β1 (catalog number 100-21C;
PeproTech, Rocky Hill, NH, USA). For a stock concentration
of 2 μg/mL, reconstitute 10 μg into 5 mL sterile HCl/BSA.
Aliquot and store at 80
C for up to 3 months. It is stable at
4
C for 1 month.
17. Recombinant human PDGF-BB (catalog number 100-14B;
PeproTech). For a 10 μg/mL: Reconstitute 10 μg into 1 mL
sterile HCl/BSA. Aliquot and store at 80
C for up to
3 months. It is stable at 4
C for 1 month.
266 Jiaozi He et al.

18. SB431542 TGF-β RI kinase Inhibitor VI (catalog number

1614; Tocris Bioscience, Ellisville, MO, USA), 10 mM. Dis-
solve 5 mg into 1.3 mL DMSO. Aliquot and store at 80
C
for up to 12 months.
19. LDN-193189 (catalog number 6053; ToCris). To make a
10 mM stock solution, add 451.5 mL of DMSO to 2 mg of
LDN-193189 and aliquot. Store solutions at 20
C pro-
tected from light. It is stable for 6 months.
20. IWR-1 (catalog number I0161; Sigma-Aldrich). To make a
10 mM stock solution, dissolve 5 mg in 1.22 mL DMSO.
Store at 20
C in aliquots.
21. Triton X-100. To make a solution of 0.2% Triton X-100, add
200 μL triton X-100 solution to 100 mL D-PBS. Filter using a
0.45 mm filter and make 5 mL aliquots in 15 mL conical tubes.
Store at room temperature, ideally in the dark.
22. Bovine serum albumin (BSA), 2%. Add 2 g of BSA powder to
100 mL of D-PBS.
23. Wash solution. 1% BSA (or serum), 0.1% Triton X-100 in
D-PBS. Dissolve 1 g of BSA powder in 100 mL of D-PBS,
and add 200 μL of Triton X-100. Filter solution through a
0.45 μm filter. Aliquot and freeze at 20
C for long-term
storage.
24. QuantiTect Reverse Transcription Kit (catalog number
205311) (Qiagen Inc., Germantown, MD, USA).
25. KAPA SYBR® Fast qPCR Kit Master Mix (2) ABI Prism
(catalog number KK4603) (Kapa Biosystems, Wilmington,
MA, USA).
26. LookOut® Mycoplasma PCR Detection Kit (catalog number
MP0035) (Sigma-Aldrich).

2.3 Tissue Culture 1. Tissue culture plates and dishes (96-, 24-, 6-well plates), Corn-
Reagents and Media ing brand (Sigma-Aldrich).
2. Ultra-Low attachment 6 well plate for suspension cultures
(catalog number 27145; StemCell Technologies, Cambridge,
MA, USA).
3. Dermal fibroblasts (see Note 1).
4. Fetal Bovine Serum (FBS) .
5. Ham’s F12 Nutrient Mix (catalog number 11765-047; Ther-
moFisher Scientific).
6. Dulbecco’s Modified Eagle Medium (DMEM) with high glu-
cose, GlutaMAX™ supplement, and HEPES (catalog number
10564-011; ThermoFisher Scientific).
 Human iPS Cells to Model 267

7. Dulbecco’s Modified Eagle Medium with high glucose, Gluta-

MAX™ supplement, and HEPES (DGGS) (catalog number
10566-011; ThermoFisher Scientific).
8. Dulbecco’s Modified Eagle Medium:Nutrient Mixture Ham’s
F-12 (DMEM/F12) (catalog number 11330-032; Thermo-
Fisher Scientific).
9. Iscove’s Modified Dulbecco’s Medium (IMDM) (catalog
number 12440053; ThermoFisher Scientific).
10. Essential 8 Medium (E8) with supplement (catalog number
A1517001; ThermoFisher Scientific). Store at 4
C, but before
addition to cells, bring medium to room temperature.
11. Essential 6 Medium (E6) (catalog number A1516401; Ther-
moFisher Scientific).
12. RPMI 1640 Medium (catalog number 11875-093; Thermo-
Fisher Scientific).
13. Fibroblast growth medium (FG): To prepare medium, add
15 mL of FBS to 85 mL of DMEM (see Note 2). Store at 4

C until needed, but before use, bring to room temperature.
14. Smooth muscle cell medium (SMCM) (catalog number
#1101; ScienCell Research Laboratories, Carlsbad, CA, USA).
15. CDM medium (500 mL): Combine 247 mL IMDM, 247 mL
F12 Nutrient Mix þ5 mL CDLC þ 250 μL transferrin þ 20 μL
monothioglycerol (MTG). Filter-sterilize and then add 875 μL
insulin. Store at 4
C for up to 4 weeks.
16. Neurobasal Medium (catalog number 21103049; Thermo-
Fisher Scientific).
17. N2 Medium (100 mL): Combine 96 mL DMEM/F12
medium with 1 mL N2 supplement, 1 mL GlutaMAX, 1 mL
nonessential amino acids, 180 μL β-mercaptoethanol (100 μM
final), and 1 mL of 100 Penicillin/Streptomycin solution.
Filter-sterilize and store at 4
C.
18. B27-Neurobasal Medium (100 mL): Combine 96 mL of Neu-
robasal medium with 2 mL B27 supplement with insulin, 1 mL
GlutaMAX, 1 mL of 100 P/S Solution.
19. N3 medium: To make this, combine equal volumes of N2
Medium with B27-Neurobasal Medium.
20. RPMI þ B27 supplement. Add 10 mL of B27 supplement
(minus insulin or containing insulin) to 490 mL RPMI
medium, and store at 4
C.
21. GlutaMAX (catalog number 35050079; ThermoFisher
Scientific).
268 Jiaozi He et al.

22. TrypLE Express (catalog number 12604-013; ThermoFisher

Scientific). Store at room temperature. There is no need to
pre-warm this solution before addition to cells.
23. StemPro® Accutase® Cell Dissociation Reagent (catalog num-
ber A1110501; ThermoFisher Scientific). Store solution at
20
C until needed. After thawing, Accutase may be stored
at 4
C for up to 2 years. Multiple freeze thaws are not
recommended.
24. EDTA dissociation buffer (EDB). Add 500 μL 0.5 M EDTA to
495.5 mL DPBS, pH 7.4. Store at room temperature.
25. Dispase (1 U/mL) in DMEM/F-12 (catalog number 07923;
StemCell Technologies, Vancouver, Canada). Thaw Dispase at
2–8
C overnight. Aliquot 10 mL into 15 mL conical centrifuge
tubes and store at 20
C. Thawed aliquots can be used imme-
diately or stored at 2–8
C for up to 2 weeks. Do not refreeze.
26. Geltrex LDEV-Free Reduced Growth Factor Basement Mem-
brane Matrix hESC-qualified (catalog number A1413202;
ThermoFisher Scientific), stored at 20 or 80
C. Allow a
bottle of growth factor-reduced Geltrex to thaw at 4
C over-
night. Dilute the Geltrex to 2 mg/mL in DMEM/F12
medium; however, the amount needed for growth of cells
may vary and it will have to be determined empirically.
Although the recommended dilution is 1:100, we routinely
add 250 μL of Geltrex to 50 mL of cold (4
C) DMEM/F-
12 (1:200 dilution) and mix thoroughly on ice. To coat the
plates, add 6 mL/10 cm tissue culture plate or 2 mL into each
well of a 6-well plate. Allow to polymerize at 37
C for at least
1 h. Store the plates at 37
C (usually in the tissue culture
incubator) for up to 2 weeks. Before plating cells, aspirate the
medium (see Note 3).
27. BD Matrigel™ hESC-qualified Matrix (catalog number
354277; BD Biosciences, San Jose, CA, USA).
28. Monothioglycerol (MTG, 11.5 M stock) (catalog number
M6145; Sigma-Aldrich).
29. β-Mercaptoethanol (55 mM stock).
30. Transferrin (30 mg/mL) (catalog number 10652202001;
Sigma-Aldrich). The solution is prepared by dissolving
30 mg/mL of transferrin powder in IMDM and filtering
through a 0.2 μm pore size membrane.
31. Insulin (4 mg/mL; catalog number 12585014; ThermoFisher
Scientific).
32. CDM-PVA (500 mL): Dissolve 0.5 g PVA into 40 mL IMDM:
F12 Nutrient Mix (50:50) by shaking at 95
C, then transfer
into 460 mL CDM. Filter-sterilize. Store at 4
C for up to
4 weeks (see Note 4).
 Human iPS Cells to Model 269

33. 100 Penicillin/Streptomycin (P/S) solution (catalog num-

ber 15140122; ThermoFisher Scientific).
34. Nonessential amino acids (NEAA; catalog number
111400040; ThermoFisher Scientific).
35. 0.1% Gelatin solution (catalog number ES-006b; Fisher Scien-
tific) or dissolve 0.5 g gelatin in 500 mL double distilled H2O
or PBS by heating at 56
C. Sterilize using a vacuum filter unit.
Store at 4
C.
36. B27® Supplement, minus Insulin (catalog number A1895601;
ThermoFisher Scientific).
37. B27® Supplement (50), serum free (catalog number
17504044; ThermoFisher Scientific) and containing insulin.
38. N2 supplement (100) (catalog number 17502001; Thermo-
Fisher Scientific).

2.4 Antibodies 1. Anti-Nanog, goat anti-rabbit polyclonal, IgG antibody (cata-

and Primers log number 80892; Abcam, Cambridge, MA, USA). Working
dilution 1:100.
2. Anti-Sox 2, goat anti-human polyclonal, IgG antibody (catalog
number AF2018; R&D Systems). Working dilution 1:100.
3. Anti-SSEA4, mouse monoclonal IgG3 antibody (catalog num-
ber MAB 1435; R&D Systems). Working dilution 1:100.
4. Anti- TRA-1-60, mouse anti-human monoclonal IgG antibody
(catalog number 16288; Abcam). Working dilution 1:100.
5. Anti-Oct3/4, mouse anti-human monoclonal IgG2b antibody
(catalog number sc-5279; Santa Cruz Biotechnology, Dallas,
TX, USA). Working dilution 1:100.
6. Anti-PAX6, mouse anti-human monoclonal IgG2a antibody
(catalog number sc-53108; Santa Cruz Biotechnology).
7. Anti-Tubulin, beta III isoform, (TUJ1), mouse anti-human
monoclonal IgG1 antibody (catalog number MAB1637;
Sigma Aldrich).
8. Anti-cardiac troponin T (TNNT2), mouse anti-human mono-
clonal IgG1 antibody (catalog number ab8295; Abcam).
9. Anti-KDR,Flk-1 rabbit anti-mouse polyclonal IgG (catalog
number sc-393163; Santa Cruz Biotechnology).
10. Anti-TCF15 (paraxis), rabbit anti-human polyclonal IgG anti-
body (catalog number, sc-514,687; Santa Cruz
Biotechnology).
11. Anti-calponin 1 (CNN1), mouse monoclonal IgG1 antibody
(catalog number C2687; Sigma-Aldrich).
12. Anti-TAGLN (SM22α), rabbit polyclonal IgG antibody (cata-
log number ab14106; Abcam).
270 Jiaozi He et al.

13. Anti-smooth muscle myosin heavy chain 11, rabbit monoclo-

nal IgG antibody (catalog number ab196982; Abcam).
14. Anti-alpha smooth muscle Actin (SMA), mouse monoclonal
IgG2a antibody (catalog number ab7817; Abcam).
15. Anti-FOXA2, rabbit anti-human polyclonal IgG antibody (cat-
alog number 710730; ThermoFisher Scientific).
16. Anti-SOX17, rabbit anti-human polyclonal IgG antibody (cat-
alog number PA5-23352) (ThermoFisher Scientific).
17. Anti-alpha fetal protein (AFP), mouse anti-human monoclonal
IgG1 antibody (catalog number MIA1305; ThermoFisher
Scientific).
18. Secondary antibodies suitable for the isotype listed for each
primary antibody (as needed, excluding any that may be pur-
chased directly conjugated).
19. Useful primers used for characterizing cell RNA/cDNA for
pluripotency and lineage markers are listed in Table 1.

3 Methods

All procedures involving cell cultivation should be performed using

aseptic techniques. When working with CytoTune Sendai virus
(SV), precautionary steps should be taken. Proper biosafety level
containment (Biosafety Level 2) should be maintained to prevent
possible viral contamination of both the user and any research
animals. Users should use Personal Protective Equipment to
avoid any direct contact. All infectious materials should be decon-
taminated before disposal. We recommend the use of bleach fol-
lowed by incineration or sterilization by autoclaving. If user contact
occurs, steps should be taken to wash away the virus immediately
(particularly from eyes) and a physician/safety officer should be
notified. Although the Fusion protein has been deleted from the
CytoTune SV and is nontransmissible, special precautions should
be taken to prevent these viral particles from being exposed to mice
or rats. In rodents, SV is highly contagious and one of the most
important respiratory pathogens in these animals. If exposed
rodents have already been infected with wild type SV, the animals
may be able to make infectious SV particles. In summary, proper
procedures must be in place for disposal of all plastics and reagents
that come into contact with SV infected cells according to local
regulations to avoid possible spread in rats and mice.
Reprogramming

3.1 Cultivation 1. Place 5–15 mL of FG medium plus 1:1000 dilution of RI

of Human Dermal (10 μM final concentration) in a 50 mL tube and warm to
Fibroblasts room temperature.
 Human iPS Cells to Model 271

Table 1
Primer sequences for characterization of putative hiPSCs and Progeny

Primer Sequence Annealing temp. (
C) Length (bp)

Sendai virus transgene
Sed V F: CAGAGGAGCACAGTCTCAGTGTTC 60 ~200
R: TCTCTGAGAGTGCTGCTTATCTGTGT
Pluripotency
Oct4 F: CCTCACTTCACTGCACTTGTA 56 125
R: CAGGTTTTCTTTCCCTAGCT
Sox2 F: ATGTCCCAGCACTACCAGAG 56 110
R: GCACCCCTCCCATTTCCC
Nanog F: GATTTGTGGGCCTGAAGAAA 60 119
R: CAGGGCTGTCCTGAATAAGC
ZFP42 F: GGTGGCATTGGAAATAGCAG 60 148
R: TGCCTAGTGTGCTGGTGGT
Ectoderm
Pax6 F: GTTGGTATCCGGGGACTTC 60 101
R: TCCGTTGGAACTGATGGAGT
SOX1 F: GGAATGGGAGGACAGGATTT 60 113
R: ACTTTTATTTCTCGGCCCGT
OTX2 F: AGAGGAGGTGGCACTGAAAA 60 126
R: GCTGTTGTTGCTGTTGTTGG
GBX2 F: AAAGGCTTCCTGGCCAAAG 60 106
R: TTGACTCGTCTTTCCCTTGC
Mesoderm
Brachyury (T) F: AATTGGTCCAGCCTTGGAA 60 109
R: TGCTCACAGACCACAGGC
Goosecoid F: CAGCAGTGCTCCTGCGT 60 98
R: ACGTTCATGTAGGGCAGCA
Endoderm
FoxA2 F: CGACTGGAGCAGCTACTATGC 60 90
R: TACGTGTTCATGCCGTTCAT
AFP F: AGAGGAGATGTGCTGGATTG 60 110
R: GTGGTCAGTTTGCAGCATTC
Sox17 F: CGAGTTGAGCAAGATGCTGG 60 120
R: TTGTAGTTGGGGTGGTCCTG

(continued)
272 Jiaozi He et al.

Table 1
(continued)

Primer Sequence Annealing temp. (
C) Length (bp)

Housekeeping
GAPDH F: AAGGTGAAGGTCGGAGTCAA 60 108
R: AATGAAGGGGTCATTGATGG
ACTB F: GCACAGAGCCTCGCCTT 60 112
R: CCTTGCACATGCCGGAG

2. Immerse cryovial containing frozen fibroblasts in a 37
C water

bath without submerging the cap. Mix gently in the water bath
until nearly thawed.
3. Once the last vestiges of ice have thawed, dry the exterior of the
tube and spray the outside of the vial with 70% ethanol.
4. Perform all subsequent steps in a laminar flow hood to prevent
contamination.
5. Slowly open and then transfer the thawed fibroblasts into a
sterile 15 mL conical tube using a 5 mL sterile pipette.
6. Slowly add 10 mL of FG medium with RI (see step 1) drop-
wise to cells in the 15-mL conical tube. While adding the
medium, gently move the tube back and forth to mix the
cells. This technique reduces osmotic shock to the cells.
7. Rinse the interior of the vial with 1 mL of FG medium with RI
and add to the 15-mL tube with cells.
8. Centrifuge the cells at 200  g for 5 min. Unless stated other-
wise, all cell centrifugations should be performed at 4
C.
9. Aspirate and discard the supernatant.
10. Resuspend the cell pellet in 2 mL of FG medium plus RI by
gently pipetting the cells up and down in the tube a few times.
11. Slowly add the cell suspension onto one well of an untreated
6-well tissue culture plate. >0.5–1  106 thawed cells per well
of the 6-well plate are recommended (see Note 5).
12. Move the dish side-to-side and back-and-forth to evenly dis-
perse cells across the surface of the wells.
13. Place the dish gently into the 37
C, 5% CO2 incubator
overnight.
14. The next day, replace the medium with fresh complete FG
medium lacking RI.
15. Replace the medium every 2–3 days thereafter until the cells are
approximately 85% confluent.
 Human iPS Cells to Model 273

16. To passage cells, aspirate the medium and wash one time with
either D-PBS or DMEM/F12 medium. Aspirate wash solution
and discard.
17. Add 1 mL of TrypLE and incubate at 37
C for 3–7 min.
Observe until cells are visibly coming off of the plate as large
aggregates when shaken back and forth.
18. Add 1 mL of FG medium, and triturate the mixture to dislodge
any remaining cells. No scraping is necessary.
19. Transfer the cells to a 15 mL conical tube containing 7 mL of
FG medium.
20. Rinse the well with 1 mL of FG medium and transfer this
medium with any remaining cells to the conical tube.
21. Centrifuge cells at 200  g for 5 min.
22. Aspirate and discard the supernatant.
23. Resuspend the cells in FG medium and passage 1:3 onto
untreated six well plates. No RI is needed.
24. Once the cells have been passaged and expanded at least one
time after thawing, the cells can be passaged with TrypLE as in
steps 16–20. Before centrifuging, take an aliquot and count
cells with a hemocytometer.

3.2 Reprogramming 1. On Day 0 (D0), plate 150,000 fibroblasts onto three wells each
of Fibroblasts of an untreated 6-well plate containing 2 mL FG medium (see
to Putative iPSCs Note 6).
(See Fig. 1) 2. On Day 1, remove one set of CytoTune™ 2.0 Sendai virus
tubes from the 80
C freezer.
3. Thaw each tube by first immersing the bottom of the tube in a
37
C water bath for 5–10 s, and then remove the tube from
the water bath and allow it to thaw at room temperature.
4. Once thawed, quick spin the tube and place it on ice. Check
certificate of analysis for # of particles/tube and follow the
companies instructions for infection, i.e., approximately
3  106 CIU/virus.
5. Aspirate medium from fibroblasts and wash one time with
2 mL prewarmed DGGS to remove serum.
6. Take pictures of fibroblasts on D0.
7. Aspirate DGGS medium and add 1 mL prewarmed DGGS to
each well to be reprogrammed, including to a negative control
well where no SV will be added to the fibroblasts.
8. Add CytoTune SV according to the manufacturer’s instruc-
tions directly into fibroblast media (see Note 7).
9. After addition of the virus, move the plate back and forth to
ensure that the virus is well distributed throughout the 1 mL of
274 Jiaozi He et al.

Fig. 1 Generation of putative hiPSC lines from fibroblasts using Sendai virus (Sv). In this figure, fibroblasts are
plated and then infected with Sv or left uninfected at the time points indicated. Within 2–3 days, the
morphology of Sv infected cells is dramatically different than that of uninfected cells. By day 6, considerable
cell death can be observed, and the density sometimes decreases by days 8–9. Around days 12–14, early
signs of colonies can be observed, which are clearly discernible usually around days 17–18. We normally
allow the colonies to grow until days 20–22 before picking the cells and transferring the colonies into 96- or
48-well plates. Once these have expanded, the colonies can be passaged with Accutase or EDTA and
expanded. A typical colony at the second passage (P2) in shown

medium. Rinse all tips and pipettes that come into contact with
virus with a 10% bleach solution and discard.
10. Incubate fibroblasts at 37
C for 24 h.
11. Aspirate medium and feed cells with 2 mL of fresh DGGS/
well. CytoTune contaminated medium should be treated with
10% bleach before disposal, and all pipettes and tips should be
similarly treated or disposed after autoclaving or incineration.
12. Take pictures of fibroblasts on Day 2.
13. On Day 3, add 2 mL of fresh DGGS/well and incubate at
37
C.
14. On Day 4, observe the cells and note any morphological dif-
ferences, particularly relative to the uninfected fibroblasts.
15. On Day 5, add 2 mL of fresh DGGS medium to each well and
incubate at 37
C. From this day forth, medium may need to be
changed every day (see Note 8).
 Human iPS Cells to Model 275

16. On Day 7, add 2 mL of fresh DGGS/well and incubate at

37
C. Prepare at least two 10 cm plates coated with Geltrex/
line of fibroblast being reprogrammed.
17. On Day 8, take pictures of cells. Harvest cells using Accutase
prewarmed to room temperature. For this, aspirate medium
and wash once with 2 mL D-PBS. Discard D-PBS.
18. Add 1 mL of Accutase to each well. Incubate 2–5 min at 37
C
until individual single cells start to round up.
19. Add 1 mL of DGGS medium. Gently triturate solution to
remove cells from the plate’s surface. Scraping of cells is also
possible, but usually this is not needed.
20. Transfer cell suspension to a 15 mL conical tube containing
7 mL of DGGS medium. Rinse wells with 1 mL of DGGS
medium and transfer to conical tube.
21. Count cells, ideally with an automated counting system. Note
and record cell viability (trypan blue exclusion). Uninfected
cells should have undergone a seven- to tenfold expansion.
22. Centrifuge cells at 200  g for 5 min. Resuspend cells in DGGS
medium.
23. Plate 100,000 Accutase digested cells on one 10 cm Geltrex-
coated dish in 8 mL of fresh FGM media. Plate another
150,000 cells on the other 10 cm Geltrex-coated dish in
8 mL of fresh DMEM-FCS media. The final medium volume
for each plate should be 10 mL (see Note 9).
24. Incubate cells at 37
C for 24 h.
25. On Day 9, replace 50% of the medium with E8 medium.
Incubate at 37
C for 24 h (see Note 10).
26. On Day 10, remove the medium, and replace with 100% E8
medium. Incubate at 37
C.
27. On Day 12, aspirate medium and add E8 medium. Incubate at
37
C.
28. On Day 14, repeat step 27.
29. From Day 15, depending on the confluency of the cells,
medium may need to be changed every day.
30. At Days 19–21, pick individual colonies, using aseptic techni-
ques, with a bevelled pipette tip attached to a manual pipette.
The tip can be used to dislodge the colony followed by imme-
diate, but gentle aspiration into the tip. No more than
10–12 μL of fluid containing a colony should be taken
each time.
31. Transfer putative colonies onto Geltrex-coated 24-well plates
containing 0.5 mL E8 plus RI followed by gentle trituration
(see Note 11).
276 Jiaozi He et al.

32. Medium is removed 18–24 h later and replaced with fresh E8

medium (i.e., no RI).
33. After colony isolation, the 10 cm plates can be returned to the
incubator after addition of fresh E8 medium to allow other
colonies to reach sizes suitable for picking (see Note 12).
34. On subsequent days when additional colonies have enlarged,
individual colonies should be isolated as in steps 30–32 and
transferred to Geltrex-coated 24-well plates.
35. Allow isolated colonies grow and expand. Monitor the colony
size daily, and change E8 medium, every 1.5–2 days (if the
colony size or number of colonies is still very small). After
4 days of growth, change medium daily, as putative iPSC lines
grow best in the presence of fresh FGF2.
36. Once a clone has individual or multiple colonies that become
sufficiently large (i.e., multiple colonies of several hundred cells
each or one large colony that has expanded to contain
hundreds of cells), it is useful to manually break up the colony
into multiple fragments and transfer the entire set of cells into a
new 24-well plate coated with Geltrex and containing E8
medium plus RI.
37. Change the medium as in step 32.
38. Subsequent passages should be made using Accutase. Here we
describe cell passaging using Accutase for a single well of a
6-well plate; however, volumes need to be proportionally
reduced for 24-well versus 6-well plates. Both the well volume
and the surface area are almost fivefold greater in the 6-well
plate than in the 24-well plate, so for convenience, just divide
the volumes by 5. Passaging by EDTA is also possible (see
Subheading 3.7, steps 10–15) at this stage.
39. Ideally cells should be expanded onto 24-well (first 1–2 pas-
sages), then 6-well plates by passaging 1:1, 1:3, and 1:6
(or some similar combination) using Accutase or 0.5 mM
EDTA in D-PBS without Ca2+/Mg+ for 7 min at RT (see
Notes 13 and 14).
40. To passage cells in a 6-well plate using Accutase, aspirate
medium and wash each well once with 2 mL D-PBS. Aspirate
and discard the D-PBS.
41. Add 1 mL of Accutase to each well. Incubate for ~3–5 min at
37
C until individual single cells start to be discernible and
cells come off in sheets. If the Accutase is not prewarmed, the
times may need to be prolonged.
42. Add 1 mL of DMEM/F12 medium. Gently triturate four or
five times using a 5 mL pipette (or a P1000 Pipetteman or
equivalent for a 24-well plate) to disrupt aggregates and
remove cells from the plate’s surface.
 Human iPS Cells to Model 277

43. Transfer cell suspension to a 15 mL conical tube containing

7 mL of DMEM/F12 medium. Rinse wells with 1 mL of
DMEM/F12 medium and transfer to the same conical
centrifuge tube.
44. Remove a small aliquot for counting, ideally with an automated
cell counting system. Note and record cell viability (trypan blue
exclusion).
45. Centrifuge cells at 200  g for 5 min.
46. Resuspend cells in E8 medium to the desired concentration.
We usually dilute the cells to an appropriate concentration in
1 mL in E8 medium.
47. Plate 50,000–100,000 Accutase digested cells (1 mL) onto
one Geltrex-coated well of a 6-well plate cm dish containing
1 mL of E8 plus 2 RI.
48. RI-containing medium should be replaced within 18–24 h of
passages to standard cultivation medium.
49. Allow the colonies to grow and expand gradually for the first
2–3 passages or until there are large numbers of colonies
per well.
50. Once cells have expanded, it is useful to freeze cells at this stage
(see Subheading 3.4). This is particularly true when relatively
large numbers of clones from multiple lines of fibroblasts need
testing.

3.3 Selection 1. After 5–8 passages, individual clones (5–10) are picked and
of Reprogrammed expanded clonally from each original isolate, as in Subheading
Cells Lacking Sendai 3.2, steps 34–39, except that medium should be changed
Virus and Mycoplasm daily. The isolation of selected colonies should be based on
morphology, with good colonies showing relatively smooth
edges and a high nucleus to cytoplasmic ratio.
2. Once each colony has been isolated and expanded sufficiently
to passage 1:2 or 1:3, make a duplicate set of plates for each
colony on separate 24-well plates. The first plate will be for
continued cultivation, expansion, freezing and immunostain-
ing if required. The second plate will be RNA isolation or
immunostaining and subsequent analysis for the presence of
Sendai virus.
3. Continue to passage the cultivation plate as needed to maintain
clonal lines.
4. For RNA, wash each nearly confluent well of the duplicate plate
with D-PBS, and lyse according to the manufacturer’s instruc-
tions (see Note 15).
5. Prepare RNA as a solution in ultrapure water (at neutral pH),
quantify using a NanoDrop spectrophotometer, and adjust the
278 Jiaozi He et al.

concentration to ~1 mg/mL. RNA should be stored at 80
C

as an aqueous solution, but avoid multiple freeze-thaw cycles.
6. For genomic DNA elimination and reverse transcription, com-
bine on ice 2 μL of gDNA Wipeout Buffer, 1 μL of 1 mg/mL
RNA, and 11 μL RNase-free ultrapure water. These reagents
are in the QuantiTect RT Kit.
7. As a negative no template control, add ultrapure water in place
of the RNA.
8. Incubate at 42
C for 2–5 min and then place back on ice.
9. Prepare the reverse-transcription master mix on ice according
to the manufacturer’s protocol.
10. Incubate for 15 min at 42
C (or according to the manufac-
turer’s instructions), followed by incubation for 3 min at 95
C
to inactivate the reverse transcriptase (RT). Move to ice.
11. Ideally, the PCR reaction is run immediately; however, this
reaction mix can be stored short-term at 20
C.
12. For quantitative PCR (qPCR), prepare the reverse-
transcription reaction components as suggested by the manu-
facturer in PCR tubes or plates, depending on the equipment
available.
13. Cap or seal the reaction tubes or plates, and briefly centrifuge.
14. Place tubes or plate in the PCR apparatus and activate the
enzyme by heating to 95
C for 3 min.
15. For the amplification, denature the template at 95
C for 20 s;
anneal the primers with the DNA at 60
C for 30 s; and
elongate at 70
C for 1–2 min. Perform these steps for
40 cycles.
16. Denature the template at 95
C for 5 min and store as needed
to run on a gel or if no longer needed discard.
17. When analyzing for the presence of Sendai virus, we find that
amplification signals above 35 cycles (ideally above 38) are
indicative of the absence of these particles.
18. Alternatively, we use immunostaining, which is less sensitive.
This technique is described below in Subheading 3.6.
19. When clones are identified that lack Sendai virus, these clones
are expanded (as in step 2) and tested for the presence of
mycoplasma using a Myocplasma PCR Detection kit according
to the manufacturer’s instructions. Once cells are confirmed to
be both Sendai virus and mycoplasma free, cells are expanded
and frozen (Fig. 2).

3.4 Freezing 1. Label 2 or more cryotubes for each clonal line to be preserved
and Thawing with date, cell line name and passage number.
of Putative iPSC
Clones
 Human iPS Cells to Model 279

Fig. 2 Quality control of selected putative iPSCs: (a) Immunostaining of an early (left) and late (right) passage
putative hiPSC clone stained for Sv. In the early passage colonies, many individual colonies can be observed
that contain Sv. (b) Using a standard mycoplasma PCR reaction with controls, we verified that our clones from
both control (057) and vEDS (0197) hiPSC clones do not contain mycoplasma

2. Prepare E8 medium containing 20% DMSO and E8 medium

containing 2 RI. You will need approximately 0.5 mL of both
of these media for every cryovial.
3. Cultivate putative iPSC clones in 6-well plates to between 70%
and 90% confluency. You will need 2–3 wells for each clonal line
to make a minimum of three frozen stocks (see Note 16).
4. Passage cells with Accutase as described in Subheading 3.2,
steps 41–47. Count cells diluted in 15 mL conical tube prior
to centrifugation. Centrifuge at 200  g for 5 min.
5. After centrifugation, resuspend cells in E8 medium þ RI at a
concentration of ~2–6  106 cells/mL.
6. To this suspension add drop wise with tube agitation an equal
volume of E8 medium þ 20% DMSO. Once added the final
concentration of DMSO will be 10%.
7. Transfer 1 mL of cell suspension to sterile cryotubes and close.
8. Transfer tubes to a Mr Frosty or equivalent and leave overnight
at 80
C.
9. The following day, transfer the cells to a liquid nitrogen
container.
280 Jiaozi He et al.

10. Once the cells have been stored in liquid nitrogen for at least
24 h, we generally thaw one tube to ensure cell recovery.
11. To thaw cells, place cryovial up to the top level of cells in a
37
C water bath without submerging the cap. Mix gently in
the water bath until nearly thawed.
12. Once the last vestige of ice has thawed, dry the exterior of the
tube and spray the outside of the vial with 70% ethanol and
open in the laminar flow hood.
13. Once opened, gently transfer the 1 mL of thawed putative
iPSCs into a sterile 15-mL conical tube using a 5-mL sterile
pipette.
14. Slowly add 8 mL of E8 medium with RI drop-wise to cells in
the 15-mL conical tube. While adding the medium, gently
move the tube back and forth to mix the cells. This technique
reduces osmotic shock to the cells.
15. Rinse the vial with 1 mL of E8 medium with RI and add to the
15-mL tube with cells.
16. Centrifuge the cells at 200  g for 5 min. Aspirate and discard
the supernatant.
17. Resuspend the cell pellet in 2 mL of E8 medium plus RI by very
gently pipetting the cells up and down in the tube a few times.
Transfer to a 6 well plate (1–2  106 cells/well) coated with
Geltrex as described earlier.
18. Place in the 37
C incubator overnight and the following
morning, check to ensure cells have survived. If the cells sur-
vived, then these either can be discarded or the cells expanded
for subsequent analysis. If expansion is required, then replace
the cultivation medium with E8 lacking RI, and passage as
needed (see Note 17).
Characterization of Putative human iPS Cell Lines

3.5 PCR Analysis 1. Using cDNA prepared in Subheading 3.3, steps 6–8, prepare
of Pluripotency the master mix to evaluate the RNA levels of pluripotency
Associated transcription factors (OCT4, SOX2, NANOG, and ZFP42)
Transcription Factors and of lineage restricted cell markers (Ectoderm: PAX6,
SOX1, OTX2, GBX2; Mesoderm: Brachyury (T), Goosecoid;
Endoderm: FOXA2, AFP, SOX17).
2. Perform qPCR as described under Subheading 3.3, steps
10–16, except using the primers specific for these target
genes as described in Table 1.
3. As a control, use RNA prepared from an established hiPSC line
or from a line of human embryonic stem cells.
 Human iPS Cells to Model 281

4. The expression of Oct4, Sox2. Nanog and ZFP42 should be

readily detectable by PCR at levels significantly higher than that
found using no template controls and RNA from fibroblasts.
5. The expression of the lineage-restricted markers should all be
very low or absent (i.e., present at amplification cycles of ~35 or
above) (see Note 18).

3.6 Detection 1. Sterilize glass coverslips by autoclaving in foil or by placing the

of Pluripotency coverslips under UV light for an hour. Alternatively employ
Surface Markers by sterile MatTek dishes.
Immunocytochemistry 2. Place coverslips into a 6-well plate and precoat each coverslip or
(Immunostaining) MatTek dish with Geltrex overnight at 37
C.
3. Passage the cells with Accutase and plate cells at various den-
sities onto the coverslips in the plate or directly onto the
MatTek dishes.
4. Allow the putative iPS cell lines to grow for 2–3 days until
distinct colonies can be observed.
5. Aspirate media from culture, and wash 1 with D-PBS.
6. Fix cells by adding 0.5 mL 4% PFA in D-PBS to the cells
followed by incubation for 20–30 min at 4
C.
7. Aspirate PFA and wash cells three times with ice-cold D-PBS
for 5 min each. Discard the solution in contact with cells
between washes. Cells can be stored overnight or for several
days in D-PBS at 4
C.
8. After removal of the last wash of D-PBS, incubate the fixed cells
with 0.2% Triton X-100 in D-PBS for 10 min to permeabilize
the membranes (see Note 19).
9. To block nonspecific binding, incubate cells with 2% BSA in
D-PBS at 4
C for 30 min to 1 h (see Note 20).
10. Decant blocking solution and incubate cells with primary anti-
body diluted in wash solution (1% BSA (or serum), 0.1% Triton
X-100 in D-PBS) at 4
C overnight in a humidified environ-
ment (i.e., a closed box containing moist paper towels) and
protected from light using antibody dilutions as recommended
by the manufacturer.
11. Remove the primary antibody solution and add wash solution
followed by gentle shaking for 5 min. Repeat for a total of three
washes.
12. Dilute secondary antibody in wash solution according to the
manufacturer’s dilution recommendations and incubate cells
with secondary antibody solution for 1 h at 4
C protected
from light (see Note 21).
13. Aspirate and repeat washing as in step 10 above, while ensur-
ing that the samples are protected from light.
282 Jiaozi He et al.

14. Mount coverslips onto glass slides in mounting medium with

DAPI staining following manufacturer’s instructions. Use nail
polish or other sealant to prevent dehydration. Glass slides can
be stored in the dark at room temperature.
15. Image sample using fluorescent microscopy.
16. Examine the pattern of the surface proteins SSEA4 and
TRA-1-60 on the cells. Both antibodies should show positive
staining on most/all of the cells. If compared against and
appropriate hiPSC control line, the pattern should be similar
(Fig. 3a).
17. Similarly, the presence of endogenous pluripotency transcrip-
tion factor Oct4, Sox2, and Nanog should be present in most/
all of the cells, but in the nucleus.

3.7 Differentiation 1. Examine hiPSCs under a microscope at 70–80% confluency to

of Putative iPSCs ensure minimal morphological signs of differentiation (see
to Examine Potency Note 22).
2. Rinse the well of hiPSCs with DMEM/F12, and discard wash.
3. Passage cells using Accutase as described in Subheading 3.2,
steps 41–47.
4. Dilute cells to a concentration of 36,000/mL.
5. Seed 1 mL of cells (i.e., 4K/cm2) on a Geltrex-coated 6-well
plate containing 1 mL E8 plus 2 RI. Ensure that cells are
evenly distributed across the surface area and place in the
incubator at 37
C overnight (see Note 23).
6. The following morning, change the medium to E8 medium
without RI and return to incubator. Allow cells to grow for
3–4 days to ~80–90% confluency.
Nondirected Differentiation
7. For nondirected differentiation via formation of embryoid
bodies (EBs), the following steps should be taken.
8. Aspirate medium from cultures.
9. Wash cells twice with 2 mL of 1 DPBS (prewarmed to room
temperature), aspirate and discard.
10. Add 1 mL dispase to each well of a 6-well plate, and place at
37
C. Leave undisturbed for 3–5 min until colony boundaries
appear folded back and show signs of becoming less well
packed. If accutase is used in place of dispase, do the same
procedure.
11. Gently triturate with a 5 mL pipette to dislodge the cells. Do
not over pipette, as small clusters/aggregates are preferable to
individual cells (see Note 24).
 Human iPS Cells to Model 283

Fig. 3 Characterization of putative hiPSC clones. (a) Immunostaining for pluripotency markers, showing
nuclear staining of NANOG, and also staining of TRA-1-60 and SSEA4, which should be surface proteins
present on good quality hiPSCs. (b) Example of the neuroepithelial differentiation protocol used to test
differentiation potential of putative hiPSC clones. In this example, the neuroepithelial cells were positive for
PAX6 and Tuj1. Similar analyses were also performed to test for mesodermal and endodermal lineages. (c)
H&E staining of sections generated from a teratoma generated in NOD-SCID mice, showing the presence of
various lineages. Upon full analysis by a pathologist, we were able to show that all three germ layer lineages
were equally represented in these tumors
284 Jiaozi He et al.

12. Transfer to a 15 mL conical tube containing 1 mL DMEM/F-

12, prewarmed to room temperature, to dilute the dispase.
Centrifuge at 200  g for 5 min.
13. Resuspend dissociated colonies in 2 mL of E6 medium and
transfer into one or two wells of a 6-well low attachment plate.
When using accutase, add RI to cells. Otherwise these do not
form aggregates, but with RI, aggregates can be seen within
24 h.
14. Place cells into an incubator at 37
C, 5% CO2.
15. On Day 1, remove the plate from the incubator, and observe
using a phase contrast microscope. EBs should have formed.
16. Tilt the plate to allow the aggregates to settle to one side.
Carefully and slowly, with a P1000 pipet and plugged tip,
aspirate 1 mL of medium from each well, without removal of
the EBs. Refeed the cells with E6 medium.
17. On Day 4, repeat step 16.
18. On Day 6, coat a 24-well plate with 0.5 mL of 0.1% gelatin.
Place at 4
C overnight.
19. On Day 7, remove the gelatin-coated plates and warm to room
temperature for 1 h.
20. Aspirate the gelatin and add 0.5 mL of E6 to each well (see
Note 25).
21. Remove low attachment plates with EBs from the incubator.
Transfer all of the EBs to a 15 mL conical tube and centrifuge
at 200  g.
22. Digest aggregates through the addition of 1–2 mL TrypLE for
5–10 min with mixing. Briefly mix by trituration and allow
digestion to occur for 5 min until EBs are dissociated into small
clumps or individual cells. Add 5 mL DMEM/F12 to dilute
and centrifuge at 200  g for 5 min.
23. Resuspend dissociated cells into E6 plus RI (this is not neces-
sary for all clones, but it does help in some cases) and plate onto
gelatin-coated wells. Allow cells to attach overnight.
24. The following morning, replace medium, and every 3 days
thereafter.
25. On Day 15, cells can be fixed and immunostained for markers
to all three germ layers or used to make RNA suitable for qPCR
(see Table 1).
26. Immunocytochemistry can be performed using antibodies to
early neuronal markers (PAX6, TUJ1), mesodermal markers
(SMA, TNNT2) and endodermal markers (FOXA2, SOX17,
AFP) as described in Subheading 3.6, using the antibodies and
dilutions according to the manufacturer’s recommendations.
 Human iPS Cells to Model 285

27. RNA should also be analysed for the presence of lineage mar-
kers and the absence of pluripotency markers, as described
under Subheading 3.5.
Directed Differentiation to Ectodermal Cells
28. For differentiation of putative human iPSCs to ectodermal
lineages, we perform a modified neural induction step origi-
nally provided to us by Dr. Tilo Kunuth (Edinburgh) to pro-
duce PAX6 positive cells, although any differentiation protocol
that produce ectodermal lineages are suitable (see Note 26).
29. For ectodermal cell differentation, take cells from step 6 and
passage using accutase. Add 1 mL of accutase to the 6-well
plate and incubate for 5–7 min.
30. Add 1 mL of medium to dilute the accutase, followed by
trituration using a 5 mL pipette. Transfer the cells to a 15 mL
conical tube, remove an aliquot for cell counting and then
centrifuge the cells at 300  g for 3 min. Aspirate the medium.
31. Resuspend the cells in a volume of E8 medium containing
~4  105 cells/mL of medium.
32. Add 1 mL of cell suspension to a newly prewarmed plate
containing 1 mL of E8 medium with 2 RI.
33. Change the medium every day for 3 days or until the con-
fluency is ~90–95%.
34. At this time (Day 0), change the medium to 3N medium (a 1:1
mixture of N2 medium and B27-Neurobasal Medium) supple-
mented with 100 nM LDN 193189 and 10 μM SB431542.
Feed daily with 2 mL N3 medium until Day 5.
35. On Day 6, feed cells with N3 medium supplement only with
100 nM LDN 193189 until Day 11. Change medium daily. If
cells begin to peel off from plates, then the cultures may need
additional glutamine. This can be accomplished by doubling
the amount of GlutaMAX in the medium.
36. Cells can be passaged at this time onto glass coverslips and
allowed to plate for 2–3 days or used to make RNA suitable
for qPCR (see Table 1).
37. Immunocytochemistry can be performed using antibodies to
PAX6 and Tuj1 as described in Subheading 3.6, using the
antibodies and dilutions according to the manufacturer.
38. Following staining and microscopy, one expects that >90% of
the cells should be PAX6+ and many of the cells should be
Tuj1+ (Fig. 3b). Alternatively, flow cytometric analyses of
PAX6 or nestin can be performed.
39. RNA should also be analysed for the presence of lineage mar-
kers and the absence of pluripotency markers, as described
under Subheading 3.5.
286 Jiaozi He et al.

Directed Differentiation to Mesodermal Cells

40. For directed differentiation to mesodermal lineages, we usually
perform differentiation to cardiomyocytes. This protocol has
been shown to work with a wide variety of human ES cells and
iPS cell lines. The one variable is the initial plating density.
41. Starting from step 10 above on Day 4 of cultivation, digest the
putative iPSC lines with Accutase as described at Subheading
3.2, steps 41–47.
42. Plate cells at a density of 70,000–130,000 cells (see Note 27) in
2 mL E8 medium plus RI.
43. Change the medium daily to E8 (without RI) for a total of
4 days.
44. After 96 h, aspirate medium, and replace with 2 mL of
RPMI þ B27-insulin supplemented with 6 μM of
CHIR99021. Place cells in an incubator with 5% CO2 at
37
C. This is considered Day 0.
45. At the beginning of Day 2 (48 h later), aspirate the medium
and replace with 2 mL of RPMI þ B27-insulin. Return cells to
incubator.
46. On Day 3, aspirate the medium and replace with 2 mL of
RPMI þ B27-insulin supplemented with 5 μM IWR-1. Return
cells to incubator.
47. On Day 5, aspirate medium and replace with 2 mL
RPMI þ B27-insulin. Return cells to incubator. Medium
should be changed every 2 days.
48. On Day 7 or in some cases Day 8, aspirate medium and replace
with 2 mL RPMI þ B27 þ insulin. Return cells to incubator.
49. Medium should continue to be changed every 2 days, and as
needed, 3 mL can be added, particularly once the cells are very
confluent. If the medium begins to change color, indicative of
low pH, it is wise to change the medium.
50. Spontaneous contractions can usually be seen as early as Day
6 or 7, but usually at Day 8 or 9.
51. On Day 10–12, cells can be passaged onto glass coverslips and
allowed to plate for 2–3 days or used to make RNA suitable for
qPCR (see Table 1).
52. Change medium, 1 h prior to addition of TrypLE.
53. Aspirate the medium, and wash cells with 3 mL D-PBS.
Aspirate wash.
54. Treat cells with 1 mL TrypLE at 37
C for ~3–7 min, until the
majority of cells have detached (see Notes 28 and 29).
 Human iPS Cells to Model 287

55. Triturate the cells gently with a P1000 pipetteman or equiva-

lent and transfer cells to a 15 mL conical centrifuge tube
containing 1 mL RPMI þ B27 þ insulin medium.
56. Rinse the well with 1 mL of RPMI þ B27 þ insulin medium
and combine with the remaining cells in the conical tube.
57. Centrifuge the cells at 200  g for 3 min. Resuspend the cells in
RPMI þ B27 þ insulin medium supplemented with 1 mL/mL
Y27632. Count cells using a Countess or similar.
58. Plate cells in this medium onto Geltrex-coated surfaces (glass
coverslips—see Subheading 3.6, step 1) or plate ~1  106 cells
into one well of a 6-well plate.
59. The medium needs to be changed the following day and cells
cultured in RPMI þ B27 þ insulin. Medium should be
changed every 2–3 days, depending on cell survival and the
culture density.
60. Immunocytochemistry can be performed using antibodies to
cardiac Troponin T (TNNT2) as described in Subheading 3.6,
using dilutions according to the manufacturer’s
recommendations.
61. Following staining and microscopy, one expects that >90% of
the cells should be TNNT2+/cTnT+. Alternatively, flow cyto-
metric analyses of cTnT staining can be performed [22].
62. RNA should also be analysed for the presence of lineage mar-
kers and the absence of pluripotency markers, as described
under Subheading 3.5.
Directed Differentiation to Endodermal Cells
63. For Endoderm directed differentiation, on Day 2 after
plating, replace E8 medium with RPMI þ B27-insulin con-
taining 100 ng/mL Activin A, 10 ng/mL BMP4, and 20 ng/
mL FGF2.
64. Place cells in an incubator with 5% CO2 at 37
C.
65. Culture cells with daily medium changes for 2 days.
66. On Days 3–5, replace the culture medium with RPMI þ B27-
insulin and supplemented medium with 100 ng/mL Activin
A only. Place cells in an incubator with 5% CO2 at 37
C.
67. Change medium daily for 3 days.
68. Cells can be passaged at this time onto glass coverslips and
allowed to seed for 2–3 days or used to prepare RNA.
69. Immunocytochemistry can be performed using antibodies to
FOXA2 and SOX17 as described in Subheading 3.6, and
dilutions according to the manufacturer’s recommendations.
Cells can also be prepared for RNA analyses, as described
earlier.
288 Jiaozi He et al.

70. Following staining and microscopy, most cells should express

markers specific for the anterior definitive endoderm such as
FOXA2 and SOX17 with the minimal presence of pluripo-
tency markers such as OCT4.
71. RNA should also be analysed for the presence of lineage
markers and the absence of pluripotency markers, as described
under Subheading 3.5.
In Vivo Analysis of Potency Through Teratoma Formation (See
Note 30)
72. To generate teratomas, follow steps 3–6 of Subheading 3.7,
and cultivate for 3 days.
73. On Day 3, approximately 2–2.5  106 cells should be present
in 70–80% confluent wells.
74. Before passaging cells for injection, thaw Matrigel either over-
night at 4
C or for at least 1–2 h on ice.
75. Passage cells using Accutase as described in Subheading 3.2,
steps 40–45 (see Note 31).
76. Add DMEM/F12 medium to the plate and triturate to break
the cells into small clusters of 10–50 cells.
77. Centrifuge cells at 300  g for 5 min.
78. Aspirate the supernatant and wash the cells with 5 mL of
D-PBS by gentle swirling of the tube or by very gentle tritura-
tion. Recentrifuge as in step 76.
79. Resuspend the cells in DPBS plus 2 RI (Y27632) at a
cell concentration of 1  106 cells/50–70 μL of DPBS (see
Note 32).
80. On ice, add an equal volume of chilled, liquid Matrigel to the
cell-D-PBS suspension and mix gently. Keep this mixture on ice
at all times.
81. Using a tuberculin syringe with a 28.5 gauge needed, draw the
cells up slowly into the syringe.
82. Using NOD-SCID mice (see Note 33), restrain the animal by
the scruff method while holding the tail with your small finger.
83. Prep the area by disinfecting with 70% ethanol, and then insert
the needle, bevel up, just under the surface of the skin of the
hind limb. The needle should be inserted parallel to the skin and
directed towards the anterior end of the animal (see Note 34).
84. Draw back on the syringe plunger to ensure that no vessel has
been penetrated. Once confirmed, inject the material
(120–140 μL) slowly.
85. Withdraw the needle and return the animal to the cage. The
Matrigel-D-PBS-cell mixture should rapidly solidify at the site
 Human iPS Cells to Model 289

of injection, since the animal’s body temperature should be

37
C.
86. Mice should be monitored daily for up to a week to ensure
against any infections and then weekly.
87. Teratomas can often be palpitated within 3 weeks, but these
should be allowed to increase in size up to 1 cm in diameter to
maximize their ability to be examined histologically.
88. Animals should be sacrificed according to local Animal
Research Committee approved protocols 4–8 weeks after injec-
tion. Teratomas should be explanted and processed for H&E
staining (Fig. 3c). An experienced pathologist should interpret
the slides.
Karyotypes
89. One final assessment of the putative human iPSCs needs to be
performed. This is generally done most easily by shipping cells
to an appropriate provider of this service (e.g., WiCell, Madi-
son, WI). All putative human iPSC lines should have normal
karyotypes with correct numbers of chromosomes and with-
out obvious rearrangements. This can be determined concur-
rently with the early assessment of potency, and any line that
has an abnormal karyotype should not be used further [23].
90. In the end, only those cell lines that readily differentiate to all
three germ layers in vitro, generate robust teratomas with
relatively similar quantities of all three germ layers, and have
normal karyotypes should be considered as authenticated
human iPSC lines.
91. Finally it is worthwhile mentioning that human pluripotent
stem cell lines are subject to genetic and epigenetic changes in
culture that need to be considered when using these cells
[24, 25].

3.8 Directed 1. Human iPSCs should be cultured as described in Subheading

Differentiation 3.7, steps 1–6.
of hPSCs to Paraxial 2. Digest cells with Accutase and seed 55,000 cells (but ranging
Mesoderm-Derived from 40,000–75,000 cells, depending on the hiPSC line) onto
Smooth Muscle Cells a Geltrex (1:200 dilution)-coated 6-well plate in the presence
of E8 medium plus 1RI.
3. Change the medium to E8 within 24 h.
4. Within 48 h after plating, the colony size should contain
50–100 cells per cluster.
5. Once the colonies have reached this approximate size, aspirate
the medium, wash once with CDM-PVA, and add in
CDM-PVA supplemented with 20 ng/mL FGF2, 10 μM
LY294002 and 10 ng/mL BMP4. This is considered as Day 0.
290 Jiaozi He et al.

6. At Day 1.5, the colonies should be less hPSC-like and should

spread. Refresh medium by the addition of CDM-PVA supple-
mented with 20 ng/mL FGF2 and 10 μM LY294002.
7. Between Days 2 and 4, replace the medium and supplements,
particularly if a pH change is observed with the medium (i.e.,
the medium becomes yellow).
8. Based on the enlarged cell size and morphological differences,
all cells in the colonies should be differentiated by Days 4–5.
9. On Day 5, aspirate the medium and change the medium to
CDM-PVA supplemented with 10 ng/mL PDGF-BB and
2 ng/mL TGF-β1 (see Note 35).
10. On Day 8, change the medium.
11. Prepare gelatin-coated plates by covering the wells with 0.1%
gelatin overnight at room temperature.
12. On Day 9, aspirate the gelatin and allow the surface to air-dry
in the hood for 10 min. Cover the surface with FG medium
and place plates/dishes into the incubator at 37
C overnight.
This is important, as we have noted that failure to add FBS
leads to major cell loss (lack of cell attachment) following cell
passaging.
13. On Days 10–12, once the cells are confluent, aspirate the
supernatant. Treat cells with 0.8 mL TrpLE for no more than
3 min in a 37
C incubator. Add 1 mL of FG medium to the
cells, triturate gently, and transfer to a 15 mL conical tube
containing 5 mL of FG medium.
14. Centrifuge at 200  g for 3 min RT.
15. Plate cells onto gelatin-coated wells at a seeding density of
100,000 cells per cm2 in CDM-PVA supplemented with
10 ng/mL PDGF-BB and 2 ng/mL TGF-β1.
16. The following day, ~60–70% cells should attach to the plate.
The medium can be changed at this time to remove floating
cells, and it should be changed every 3 days from this time
forward (or earlier if the medium has a pH change).
17. From Day 18, SMC subtypes will appear spindle or stellate
shaped.
18. From this point, cells can be passaged with TrpLE and seeded
onto gelatin-coated wells at a ratio of 1:3. Cells should be
maintained in SMCM containing the appropriate supplements
provided by the manufacturer. The standard medium contains
2% FBS.
19. To obtain more contractile SMCs, the SMCM needs to be
modified and the amount of FBS reduced. The FBS should
be reduced to ~0.5% to induce MYH11 cells after one addi-
tional passage and growth to 70% confluency. High FBS levels
 Human iPS Cells to Model 291

keep the cells in a proliferative state, while a reduction in FBS

may allow the cells to become more contractile. Importantly,
only early proliferative cells can become contractile, as long-
term passaging and cultivation in high FBS makes the cells
refractory to a transition to contractile cells.
20. Immunocytochemistry or flow cytometry can be performed at
various time points using antibodies to MYH11, SMA, CNN1
and TAGLN as described in Subheading 3.6, but using the
antibodies, fixation conditions, and dilutions according to the
manufacturer (Fig. 4a).
21. Following staining and microscopy, one expects that the major-
ity of cells should be SMA+ (>95%), CNN1+ (>90%) and
TAGLN+ (~90%). The number of MYH11 cells will however
be much lower, and can range from 1–50%. Reducing the FBS
level will increase the amount of MYH11.
22. RNA can also be prepared and analysed for the presence of
smooth muscle markers, as described under Subheading 3.5
(Fig. 4b).

3.9 PCR Analysis 1. For PCR analyses of cell surface proteins or extracellular matrix
of Cell Surface components, we use quantitative PCR techniques, as described
Receptors and ECM in Subheading 3.3, steps 12–17.
Components 2. Many investigators may wish to design primers de novo, and
there are numerous sites and software available for this proce-
dure. To minimize testing of primers, we prefer to use pre-
tested/predesigned primers, which have been deposited to
PrimerDepot for the research community.
3. For selection of predesigned primers, go to PrimerDepot
http://primerdepot.nci.nih.gov. Input the RefSeq ID (such as
NM_181501, for integrin alpha1) or HUGO gene name (such
as ITGA1). This website is for human genes ONLY. For quan-
titative PCR, choose a pair of primers with amplicon sizes
ranging from ~80 to 120 bp and estimated genomic amplicon
size over 1000 bp.
4. Search the location of right and left primers in Ensembl
Genome Browser http://www.ensembl.org/index.html (see
Note 36). The reverse primer must be reverse complement
and visualized in software such as SnapGene Viewer.
(a) Choose primers that span two exons, which make geno-
mic amplification impossible for the desired length.
(b) If (a) is not possible, then choose right and left primers in
different exons.
(c) Alternatively, both (a) and (b) can be ignored, if genomic
DNA is eliminated during the RNA extraction procedure.
292 Jiaozi He et al.

Fig. 4 Analysis of paraxial mesoderm derived smooth muscle cells generated from verified hiPSCs from
control and vEDS patients. (a) Immunostaining of three smooth muscle markers, as indicated in the figure. (b)
Normalized qPCR analysis of transcripts analyzed from a control (2003-071-057, clone 1) and vEDS (0197)
human iPSC lines. In these analyses, we noted a difference in MYH11 transcript abundance between the
control and diseased smooth muscle cells cultivated in the presence of high FBS. (c) Example of analyses
performed on collagen isoforms (COL1A1 and COL3A1) and collagen receptors (not shown) after differentiation
to paraxial mesoderm derived smooth muscle cells. Unlike fibroblasts, COL3A1 was reduced relative to control
lines, suggesting some abnormal regulation of this transcript that might contribute to the phenotype observed
in patients with some vEDS point mutations
 Human iPS Cells to Model 293

5. Make sure there are as few SNPs as possible in the primer

sequences. Click the link above the sequence, and a new page
will show the information of the variant.
(a) If an rs number is assigned, click “view in dbSNP.” In the
new page displaying information of the SNP, scroll down
to check the frequency of this SNP in human populations.
If the minor allele frequency (MAF) is below 0.001 or no
frequency data is showed, this SNP is usually OK for the
primer. Otherwise, SNP with MAF over 0.01 should not
be included as a useable primer.
(b) If a COSM number or a TMP_ESP number is assigned,
instead of an rs number, this variant is not common in
human populations so it is fine for primer design.
6. For primer pairs that pass steps 4 and 5, further tests can be
made to enhance the probability of selecting primers that will
yield reproducible data quickly:
(a) Primer-Blast http://www.ncbi.nlm.nih.gov/tools/
primer-blast/index.cgi?LINK_LOC¼BlastHome. Using
blast refseq mRNA of human, check how many transcripts
the primer pair matches and the amplicon size. Make sure
target transcript(s) can be detected by this primer pair.
(b) In-Silico PCR: http://genome.ucsc.edu/cgi-bin/hgPcr?
command¼start. This test shows how the primer pair
works in human genomic DNA (not cDNA).
7. If acceptable primers from PrimerDepot are not available, go to
PrimerBank http://pga.mgh.harvard.edu/primerbank. Search
for deposited primers by NCBI gene symbol and human in
species. Double-check the gene descriptions. Choose a pair of
primers with amplicon size range 80–120 bp. Do the test in
steps 4–6.
8. Basically, one pair of primers satisfies the requirements for
quantitative PCR.
9. Send the primer sequence to a commercial company, such as
IDT or Invitrogen.
10. Before the primer pair is used for quantitative PCR, it is
recommended to test the amplification conditions using PCR
and gel electrophoresis to see: that
(a) Primer pair works efficiently.
(b) Only one product band yield.
(c) The amplicon size is correct.
11. Examples of primers identified using this procedure using Pri-
merDepot are shown in Table 2, as well as outputs from
fibroblasts and vSMCs derived from hiPSC lines (Fig. 4c). In
the example shown, COL3A1 is less prevalent in differentiated
 294
Jiaozi He et al.

Table 2
Primer sequences designed to analyze collagen and collagen receptors

Gene NCBI Forward PRIMER Reverse PRIMER Start End Product length
ITGA1 NM_181501 CTCACTGTTGTTCTACGCTGC CGGAGAACCAATAAGCACCCA 107 253 147
ITGB1 NM_002211 TTTGAGCAAACACACAGCAA GAGTCGCGGAACAGCAG 66 189 124
ITGA11 NM_012211 ACTCAACCTGGGAAGGGTCA GCTCCCACACTCATGAGACC 342 480 139
DDR2 NM_006182 AGATAGGCAGCAGCAGGAAC AGAGGCCAGCTGTTTTTGAG 93 228 136
COL1A1 NM_000088 CCCCGAGGCTCTGAAGGTC GGAGCACCATTGGCACCTTT 1204 1334 131
COL3A1 NM_000090 GCAGGGTCTCCTGGTTCAAA CGGGACCCATTTCGCCTTTA 1183 1318 136
 Human iPS Cells to Model 295

Paraxial mesoderm derived smooth muscle cells from hiPSC

lines derived from patients with vEDS than in the control line.

4 Notes

1. Fibroblasts can be purchased or prepared from skin biopsies.

For reprogramming, it is preferable to use early passage cells
(less than passage 8).
2. 10% FBS containing medium with normal glucose also works
well, but in our hands, rapidly growing and proliferating fibro-
blasts work best.
3. Matrigel can be used in place of Geltrex for growth and culti-
vation of putative hiPSC lines. Moreover, if making GMP
compatible cell lines, then it is advisable to use either XF
Vitronectin derived from human 293T cells or Laminin-521.
4. We have noted some variability in the solubility of PVA. Some
batches may go into solution by simply adding 0.5 g of PVA to
a cold 500 mL bottle of medium and placing it at 4
C for 72 h,
shaking to mix periodically. The CDM/PVA mix can then be
kept at 4
C until use. Other batches of PVA will not go into
solution at 4
C even after 1 month. If this is the case with your
batch an alternative method can be used. Make a 10% w/v
solution of PVA in cold, sterile distilled water in a glass bottle
by slowly adding the PVA (5 g per 50 mL) to the water, trying
to prevent the formation of clumps. Mix thoroughly and heat
to 85
C for 30–40 min with agitation, a hybridization oven is
ideal for this. Store at 4
C (Information from Sigma-Aldrich).
The latter is probably the most reproducible (and the approach
we now use routinely), but you must then add 5 mL or up to
10 mL of solution to the CDM (which may have a minor
dilutive effect).
5. After thawing, some cells will fail to survive. To ensure that
early passage dermal fibroblast maintain relatively high prolif-
eration rates, we keep fibroblast seeding densities relatively
high. If the cells are plated at a low density, proliferation may
be slow and the cells may be less suitable for reprogramming.
6. Normally, it is best to reprogram 3–5 lines or seeding concen-
trations of fibroblasts at a time. The CytoTune SV loses activity
after thawing, and it is not recommended to refreeze, as effi-
ciency will profoundly decrease. In our hands, the plating of
either 100k or 150k cells worked best for reprogramming, but
this varies depending on the primary fibroblast cell culture.
7. A true example for CytoTune SV addition is given for reference
purposes in Table 3.
8. Daily changes of medium may be required to prevent changes
in pH. If the cells become acidic, the reprogramming
296 Jiaozi He et al.

Table 3
Actual example of CytoTune Sv concentrations used for reprogramming

Recommended
Titer (CIU/mL) transduction CTL-0182 Well CTL-0182 Well CTL-0182 Well
transduction volume in μL A 1.5  105 B 1.5  105 C 1.5  105
Component volume for 106 cells cells plated cells plated cells plated
CytoTune™ 1.1  108 45 25 18 0
2.0 human
KOS
CytoTune™ 1.1  108 45 25 18 0
2.0 human
c-Myc
CytoTune™ 1.1  108 27 15 10.8 0
2.0
Human Klf4
Example of the volumes of CytoTune™2.0 added to each well of fibroblasts for reprogramming. In this example,
fibroblast line CTL-0182 was plated at a density of 150,000 cells. The CytoTune SV is added at two concentrations for
reprogramming and we run one well as a negative control. When observing whether the infections have been successful,
visual comparisons with the uninfected cells are useful. KOS—Klf4, Oct4, Sox2. Volumes are in microliter

conditions and survival of partially reprogrammed cells may be

compromised.
9. The company suggests plating up to 200,000 cells at this stage,
but in our hands, the colonies do not look as good, and the
plating density becomes rather high, making it more difficult to
identify and isolate putative colonies. “Better” putative iPSC
colonies were noted at lower plating cell densities.
10. Do not change medium to E8 media too soon (before D8).
According to the company, this puts further stress on the cells
and may limit the number of potentially reprogrammed cells.
11. Isolation of colonies can be challenging particularly for
novices. We typically use a P20 Eppendorf pipetteman with a
bevelled tip to isolate colonies. We also like to use an EVOS XL
Core Cell Imaging System placed inside a tissue culture hood,
but any other suitable equipment can be used. The EVOS
system helps maintain aseptic conditions, and the large screen
facilitates the identification of putative colonies (ideally con-
taining several hundred or more cells/colony) for picking and
transfer to a 48- or 24-well plate. After transfer, we triturate the
colony gently such that it ideally has broken into at least 3–10
sections, for expansion.
12. It should be noted, that picking of colonies from 10 cm plates
may be incomplete. Residual cells can grow to reform colonies.
As these regrowths would be duplicates of existing clones, we
usually limit our isolation of colonies to a window of 2–3 days.
 Human iPS Cells to Model 297

13. Passage with EDTA is a fast and easy way to passage putative
human iPS cells. Cells must, however, be 70–85% confluent, as
densities above 90% are difficult to passage. Through the che-
lator activity of EDTA, the removal of calcium and magnesium
from cells causes adhesions between cells to be lost, allowing
for rapid dissociation, even at room temperature. Dissociations
are not always complete, but for routine passaging, the plating
of small aggregates can be beneficial. The addition of RI is not
required for passaging of cells by EDTA; however, there are
reports that chromosomal aberrations may occur with long-
term cultivation and that these aberrations can be minimized
by the addition of RI. Once lines are well established and stocks
frozen in liquid nitrogen, the addition of RI to iPSC lines may
no longer be required when passaging with EDTA. To passage
cells using EDTA, aspirate medium, followed by addition of
EDTA (2 mL of a 6 well plate) to rinse wells two times at room
temperature. Add EDTA a third time to the well (1 mL) and
incubate the plates at room temperature for 3–5 min for small
aggregates or 7–9 min for single cells. Once cells begin to show
distinct separations within the colonies, remove the EDTA by
aspiration. Add E8 medium and immediately triturate to break
any clumps of cells into small aggregates. The addition of
calcium back to the cells caused rapid reaggregation, so it is
imperative to dissociate the cells quickly. Once cells are disso-
ciated, ideally into small aggregates, transfer to a Geltrex-
coated well and plate with E8 þ RI.
14. Depending on the initial isolation, colony expansion can be
readily achieved within 4–5 passages, but for novices it may
take a few additional passages for successful expansion. This is
not a problem, as the colonies still contain residual SV. With
the CytoTune 2 SV, it usually takes 8–14 passages to eliminate
the SV, and in some cases colony isolation may be required. It is
also important to realize that not every picked putative colony
will be suitable for expansion. In cases where numerous clones
are being expanded, pick those colonies that more resemble
embryonic stem cells, characterized by smooth edges, very
tightly packed cells, and cells with a high nuclear to cytoplasmic
ratio.
15. Isolation of RNA is beyond the scope of this chapter, but many
kits are available that permit high quality RNA isolates (e.g.,
RNeasy Mini Kits from Qiagen).
16. Although a single frozen stock may be acceptable, we recom-
mend making a minimum of three tubes of frozen cells. The
first is to ensure proper freezing, the second is maintained for
storage and thawing, while the third serves as a backup in case
there is contamination of the cells upon thawing. Ideally
1 to 3 million cells/cryotube should be stored; however, we
298 Jiaozi He et al.

have better recovery when >2–3 million cells are frozen/

cryotube.
17. On occasion, frozen hPSC lines do not recover after thawing
after storage in liquid nitrogen. If thawing was unsuccessful the
day after placing cells in liquid nitrogen, then use the extra cells
on the 6-well plate to expand and refreeze according to the
protocol. On rare occasions after long-term storage, cells do
not seem to recovery after thawing. In these cases, try transfer-
ring the entire contents of the cryotube directly to one well of a
6-well Geltrex plate coated with Geltrex and containing 1 mL
of E8 medium plus 2 RI. Allow cells to grow for 4–5 h and
then replace the medium with E8 þ RI overnight. This some-
times helps recovery. If this is unsuccessful, then we have found
it useful to plate frozen cells on feeder layers even if the cells
were never cultivated on feeder layers during their preparation.
18. In some instances, lineage restricted markers may be present or
even prevalent in putative iPS cell clonal lines. This can be
indicative of some overt differentiation or of putative lines
that are not ideal and should not be pursued. Occasionally,
the presence of these markers indicates sub-optimal cultivation
conditions. If the cultures are not handled correctly, with
proper feeding and passaging, some differentiation may
occur. Separately, many early iPSC clonal lines are not uniform.
As selection and expansion increase, the lines become more
robust and more uniform with lower amounts of overt
differentiation.
19. Permeabilization is necessary for the detection of intracellular
and nuclear proteins, unless samples are fixed by acetone.
Surface membrane proteins do not require permeabilization
so long as the epitopes are exposed on the surface. Triton
X-100 can be added to the BSA solution, but we keep these
separate depending on whether we are testing for some surface
proteins or not.
20. A blocking solution made of 5–10% serum from the host
species to which the secondary antibody was raised may be
preferentially to 2% BSA.
21. Stained cells using fluorochrome-conjugated antibodies
require protection from light as this can diminish the signals.
22. If there are overt signs of putative hiPS cell differentiation,
either select colonies and reexpand or passage until the colonies
are of very high quality. The presence of differentiated cells can
adversely affect these analyses.
23. The cell density at the initiation of differentiation can have a
dramatic impact on the differentiation efficiency and may need
to be determined empirically for each cell and clonal line. We
 Human iPS Cells to Model 299

recommend always running a fully characterized human ES or

iPS cell line in parallel for comparative purposes.
24. Use new pipette tip for each well/plate to reduce chance of
cross-contamination. For large scale experiments, cells from
multiple wells/plates can be combined into single 15 or
50 mL conical centrifuge tube. When combining cells from
multiple plates, the ratio of dispase/cells solution to DMEM/
F-12 should remain 1:1 to ensure proper inactivation of the cell
detachment solution.
25. As an alternative, EBs can be transferred on Day 7 to gelatin-
coated 24-well plates. The EBs will attach, and these can be
allowed to grow for another 7 days, so long as the medium is
changed every 2 days. Again on Day 15, these cells can be used
for further characterization through immunocytochemistry,
mRNA extraction followed by detection of differentiation
and pluripotency associated genes, and flow cytometry.
26. As an alternative, one can use the differentiation protocol of
Cheung et al. [26] who generates neuroectoderm-like cells
with a neuroepithelial morphology that can differentiate into
smooth muscle cells. The rationale for these tests is to show
that the cells can readily generate ectodermal lineages using
defined protocols. The more general nondirected protocols are
used for general differentiation.
27. Differentiation to cardiomyocytes is highly dependent on the
initial plating conditions. We usually find that lines differentiate
using an initial plating density of 70,000–130,000 cells/well of
a 6-well plate; however, for some human iPSC lines, we have
had to plate tenfold more cells. An alternative protocol for this
can be found in Bhattacharya et al. [22]
28. The time it takes for cells to detach is highly dependent on the
“age” of the cardiomyocytes (CMs). D15-D30 cells take
around 5–6 min, older cells (D60 or older) requires up to
10 min. CMs can be maintained in culture for at least
120 days if passaged every few weeks.
29. Different enzymes have been tested, including trypsin–EDTA,
TrypLE, Accutase. Each has resulted in >85% viability; how-
ever, trypsin based enzymes require the shortest dissociation
time and yield the highest viability (usually above 90%).
30. Teratoma formation and analysis survey lineage differentiation
capability (i.e., three germ layers) and behavior of stem cell
in vivo. The in vivo ability to form robust teratoma has been
viewed as the gold standard for assessing hiPS cell pluripo-
tency; however, care should be taken in analysing these tumors.
The presence of multiple lineages is likely for most putative iPS
cell lines; however, only robust lines tend to differentiate
300 Jiaozi He et al.

equally well to all three germ layer lineages: ectoderm, endo-

derm and mesoderm.
31. Several different enzymes can be used for dissociation of puta-
tive human iPS cell lines, including TrypLE, dispase, and accu-
tase. It is important to avoid single-cell preparations, as small
aggregates survive better and adhere to the injection site better
than single cells.
32. Teratomas can be formed using a range of cell numbers. Typi-
cally, engraftment of 105–106 cells allows efficient teratoma
formation and growth within 4–8 weeks. The delivery of
greater cell numbers promotes more rapid terotoma forma-
tion. As we are doing subcutaneous injections, higher numbers
are preferable, as the cells do not stay localized as injections to
the testis.
33. Different strains of mice can be used for teratoma formation,
but for putative human iPSCs, immunodeficient rodents (e.g.,
SCID, NOD-SCID, or BC nude) are preferred. Since these
animals are immunodeficient, the mice need to be maintained
in ventilated cages and all handling should take place in a hood
to avoid potential viral or bacterial infections. Check with the
local restrictions on how to maintain and handle these animals.
The rate of terotoma formation differs among different hiPSC
lines and hESCs [27, 28]. Such variations are likely to be
attributed by the heterogeneous gene expressions in hiPSCs
that are reprogrammed by a variety of somatic origins and/or
techniques.
34. We describe a subcutaneous injection, because no anesthetic is
required, and the animal does not suffer any undue stress or
pain. If injections are made in the testis or intramuscularly, then
follow local guidelines to ensure adequate pain alleviation in
accordance with all animal safety guidelines and protocols.
35. At this time, cells can be prepared for flow cytometry and
analysed for the presence of TCF15. If performed, one would
expect that >70% of the cells should be TCF15+.
36. Using the Ensemble Genome Browser, input the correct gene
name, in the left column “Summary”, click “cDNA” in the left
column. After loading, the full cDNA sequence of this tran-
script will be show below. Above the sequence, the letters, such
as *, K, M, Y, indicate possible variants at this nucleotide. “*”
means clinical relevant mutation. Keep this page open.

Acknowledgment

This work was supported by a grant from the Hong Kong Research
Grant Committee General Research Fund (Project number
17100214) and by generous support from the Huey Foundation.
 Human iPS Cells to Model 301

We also thank Ruixia Deng for discussions and help in acquiring

H&E staining of teratoma sections.

References
1. Takahashi K, Yamanaka S (2006) Induction of terminally differentiated circulating T cells.
pluripotent stem cells from mouse embryonic Cell Stem Cell 7:11–14
and adult fibroblast cultures by defined factors. 12. Anokye-Danso F, Anokye-Danso F, Trivedi
Cell 126:663–676 CM, Trivedi CM, Juhr D, Juhr D et al (2011)
2. Takahashi K, Tanabe K, Ohnuki M, Narita M, Highly efficient miRNA-mediated reprogram-
Ichisaka T, Tomoda K et al (2007) Induction ming of mouse and human somatic cells to
of pluripotent stem cells from adult human pluripotency. Cell Stem Cell 8:376–388
fibroblasts by defined factors. Cell 13. Warren L, Manos PD, Ahfeldt T, Loh Y-H,
131:861–872 Li H, Lau F et al (2010) Highly efficient repro-
3. Cahan P, Daley GQ (2013) Origins and impli- gramming to pluripotency and directed differ-
cations of pluripotent stem cell variability and entiation of human cells with synthetic
heterogeneity. Nat Rev Mol Cell Biol modified mRNA. Cell Stem Cell 7:618–630
14:357–368 14. Gazit Y, Jacob G, Grahame R (2016) Ehlers-
4. Sugiura M, Kasama Y, Araki R, Hoki Y, Danlos syndrome-hypermobility type: a much
Sunayama M, Uda M et al (2014) Induced neglected multisystemic disorder. Rambam
pluripotent stem cell generation-associated Maimonides Med J 7:e0034
point mutations arise during the initial stages 15. Kim ST, Brinjikji W, Lanzino G, Kallmes DF
of the conversion of these cells. Stem Cell Rep (2016) Neurovascular manifestations of
2:52–63 connective-tissue diseases: a review. Interv
5. Mekhoubad S, Bock C, de Boer AS, Kiskinis E, Neuroradiol 22:624–637
Meissner A, Eggan K (2012) Erosion of dosage 16. Mak KM, Png CYM, Lee DJ (2016) Type V
compensation impacts human iPSC disease collagen in health, disease, and fibrosis. Anat
modeling. Cell Stem Cell 10:595–609 Rec (Hoboken) 299:613–629
6. Nakagawa M, Koyanagi M, Tanabe K, 17. Hershenfeld SA, Wasim S, McNiven V,
Takahashi K, Ichisaka T, Aoi T et al (2008) Parikh M, Majewski P, Faghfoury H et al
Generation of induced pluripotent stem cells (2016) Psychiatric disorders in Ehlers-Danlos
without Myc from mouse and human fibro- syndrome are frequent, diverse and strongly
blasts. Nat Biotechnol 26:101–106 associated with pain. Rheumatol Int
7. Soldner F, Hockemeyer D, Beard C, Gao Q, 36:341–348
Bell GW, Cook EG et al (2009) Parkinson’s 18. Beridze N, Frishman WH (2012) Vascular
disease patient-derived induced pluripotent Ehlers-Danlos syndrome: pathophysiology,
stem cells free of viral reprogramming factors. diagnosis, and prevention and treatment of its
Cell 136:964–977 complications. Cardiol Rev 20:4–7
8. Sommer CA, Sommer AG, Longmire TA, 19. Frank M, Albuisson J, Ranque B, Golmard L,
Christodoulou C, Thomas DD, Gostissa M Mazzella J-M, Bal-Theoleyre L et al (2015)
et al (2010) Excision of reprogramming trans- The type of variants at the COL3A1 gene
genes improves the differentiation potential of associates with the phenotype and severity of
iPS cells generated with a single excisable vec- vascular Ehlers–Danlos syndrome. Eur J Hum
tor. Stem Cells 28:64–74 Genet 23:1657–1664
9. Kaji K, Norrby K, Paca A, Mileikovsky M, 20. Daheron L, D’Souza S (2012) Blood – SeV
Mohseni P, Woltjen K (2009) Virus-free induc- derived fibroblast generated iPSCs. In: Stem-
tion of pluripotency and subsequent excision of Book [Internet]. Cambridge, MA, Harvard
reprogramming factors. Nature 458:771–775 Stem Cell Institute, pp 2008–2012
10. Woltjen K, Michael IP, Mohseni P, Desai R, 21. Cheung C, Bernardo AS, Trotter MWB, Ped-
Mileikovsky M, H€am€al€ainen R et al (2009) ersen RA, Sinha S (2012) Generation of human
piggyBac transposition reprograms fibroblasts vascular smooth muscle subtypes provides
to induced pluripotent stem cells. Nature insight into embryological origin-dependent
458:766–770 disease susceptibility. Nat Biotechnol
11. Seki T, Yuasa S, Oda M, Egashira T, Yae K, 30:165–173
Kusumoto D et al (2010) Generation of 22. Bhattacharya S, Burridge PW, Kropp EM,
induced pluripotent stem cells from human Chuppa SL, Kwok W-M, Wu JC et al (2014)
302 Jiaozi He et al.

High efficiency differentiation of human plu- instability in the human embryonic stem cell
ripotent stem cells to cardiomyocytes and char- epigenome. Hum Mol Genet 16:1253–1268
acterization by flow cytometry. J Vis Exp 26. Cheung C, Bernardo AS, Pedersen RA, Sinha S
23:52010 (2014) Directed differentiation of embryonic
23. Fazeli A, Liew C-G, Matin MM, Elliott S, Jean- origin-specific vascular smooth muscle sub-
meure LFC, Wright PC et al (2011) Altered types from human pluripotent stem cells. Nat
patterns of differentiation in karyotypically Protoc 9:929–938
abnormal human embryonic stem cells. Int J 27. Stadtfeld M, Apostolou E, Akutsu H,
Dev Biol 55:175–180 Fukuda A, Follett P, Natesan S et al (2010)
24. Draper JS, Smith K, Gokhale P, Moore HD, Aberrant silencing of imprinted genes on chro-
Maltby E, Johnson J et al (2004) Recurrent mosome 12qF1 in mouse induced pluripotent
gain of chromosomes 17q and 12 in cultured stem cells. Nature 465:175–181
human embryonic stem cells. Nat Biotechnol 28. Gutierrez-Aranda I, Ramos-Mejia V, Bueno C,
22:53–54 Munoz-Lopez M, Real PJ, Mácia A et al
25. Allegrucci C, Wu Y-Z, Thurston A, Denning (2010) Human induced pluripotent stem cells
CN, Priddle H, Mummery CL et al (2007) develop teratoma more efficiently and faster
Restriction landmark genome scanning identi- than human embryonic stem cells regardless
fies culture-induced DNA methylation the site of injection. Stem Cells 28:1568–1570
 Chapter 18

Fabrication and Mechanical Properties Measurements of 3D

Microtissues for the Study of Cell–Matrix Interactions
Prasenjit Bose, Chen Yu Huang, Jeroen Eyckmans,
Christopher S. Chen, and Daniel H. Reich

Abstract
Cell interactions with the extracellular matrix (ECM) are critical to cell and tissue functions involving
adhesion, communication, and differentiation. Three-dimensional (3D) in vitro culture systems are an
important approach to mimic in vivo cell–matrix interactions for mechanobiology studies and tissue
engineering applications. This chapter describes the use of engineered microtissues as 3D constructs in
combination with a magnetic tissue gauge (μTUG) system to analyze tissue mechanical properties. The
μTUG system is composed of poly(dimethylsiloxane) (PDMS) microwells with vertical pillars in the wells.
Self-assembled microtissues containing cells and ECM gel can form between the pillars, and generate
mechanical forces that deform the pillars, which provides a readout of those forces. Herein, detailed
procedures for microfabrication of the PDMS μTUG system, seeding and growth of cells with ECM gels
in the microwells, and measurements of the mechanical properties of the resulting microtissues via magnetic
actuation of magnetic sphere-tagged μTUGs are described.

Key words Cell–matrix interactions, Mechanobiology, Engineered microtissues, Microfabrication,

Magnetic actuation

1 Introduction

Critical components of a cell’s surfaceome are the receptors and

transmembrane proteins that determine and control its adhesive
and mechanical interactions with the surrounding extracellular
matrix (ECM) [1–3]. These interactions are coordinated, and reg-
ulate a wide range of cellular functions [4–7]. Increasingly, it has
been recognized that these interactions cannot be faithfully repro-
duced in two-dimensional cell culture, and there is a rapidly devel-
oping need for approaches that mimic more realistically the in-vivo
cellular environment [8, 9].
This chapter describes use of a 3D magnetic microfabricated
tissue gauge (μTUG) system that can be used to study cell-matrix
interactions in engineered three-dimensional microtissue constructs

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_18, © Springer Science+Business Media, LLC 2018

303
304 Prasenjit Bose et al.

Fig. 1 Schematic of microtissue stretching using a magnetic tweezer. (a) Microtissue suspended between two
flexible PDMS micropillars that are deflected by the microtissue’s contractile force. (b) When a magnetic field
is generated by the magnetic tweezer, a magnetic force Fmag is applied to the magnetic sphere bonded on the
top of one of the pillars, and strains are exerted on the microtissue

[10–16]. The essence of the magnetically actuated microfabricated

tissue gauge (μTUG) system is shown in Fig. 1. Cell-laden collagen
gels are formed within poly(dimethylsiloxane) (PDMS) microwells
that contain vertical PDMS pillars in the wells [10]. The microwells
have typical dimensions 800 μm
400 μm by 200 μm deep, and can
contain 100–1000 cells each, depending on the application. As cells
contract the collagen gels, the gels remain anchored by the pillars.
This results in controlled self-assembly of cell–matrix “microtissues”
suspended between the pillars (Fig. 1) [10, 11]. As the pillars are
elastic, the forces of contraction of these microtissues can be moni-
tored via the pillars’ deflections. By incorporating magnetic material
in the pillars, forces can be applied to the tissues [12–14, 17] to
monitor tissue stiffness over time or to measure the dynamic
response of the cells to external mechanical perturbations
[15]. Arrays with >100 μTUG microwells in a standard P35 culture
dish (25 per cm2) can readily be produced, allowing multiple identi-
cal samples or samples with systematically varying properties to be
prepared efficiently.
Procedures for fabrication, tissue seeding and growth, and
tissue mechanical properties measurements are described in this
chapter. Examples are drawn from work on fibroblast and
smooth-muscle cell populated tissues, but the methods described
herein can readily be extended to other cell types. This approach
will also be useful for the evaluation of gene mutations of cell
surface proteins or extracellular matrix components necessary for
cell interactions and structural fidelity of tissues. The methods
described in this section encompass several distinct processes. In
Subheadings 3.1 and 3.2, we describe fabrication of the PDMS
μTUG device arrays. “Master” versions of the devices are first
produced in a thick photoresist on silicon wafers, using multilayer
 Microtissues to Study Cell-Matrix Interactions 305

photolithography techniques. The PDMS μTUG devices are then

made via replica molding from the masters. Briefly, mirror inverses
of the masters are cast in PDMS from the masters, and then these
“negative molds” are used to cast the actual PDMS devices, which
are thus exact copies of the masters. As the original masters are
somewhat fragile, a replica-molding approach is also described to
make more durable copies of the masters in plastic (Smooth-Cast),
which can be used repeatedly without degradation. In Subheadings
3.3–3.5, we describe creation and growth of microtissues in the
μTUG devices. This includes passivation of the devices’ surfaces to
prevent unwanted cell adhesion, preparation of the cell/ECM solu-
tions that will make up the microtissues, insertion of these solutions
into the microwells, and culture techniques to form and sustain the
microtissues. In Subheadings 3.6 and 3.7, we describe data acqui-
sition with the microtissue system. In Subheading 3.6, we discuss
basic contractility measurements, wherein one measures the collec-
tive developed force produced by each microtissue by optically
monitoring the deflections of the flexible pillars between which
the microtissues are suspended. In Subheading 3.7, we describe
approaches to measure the full mechanical properties of the micro-
tissues via stress–strain measurements enabled by dynamic stretch-
ing of the tissue with magnetically applied external forces. Finally in
Subheading 3.8, we describe approaches to analyze the images
resulting from the above protocols to quantify physical properties
of the microtissues. Examples described include the microtissues’
developed force during tissue formation, stress–strain relations, and
active cellular responses to applied forces.

2 Materials

2.1 Equipment 1. Contact mask aligner (Model 200; Optical Associates, San
and Instruments Jose, CA, USA) or equivalent.
2. Programmable spin-coating system like the WS-400B-NPP-
Lite system (Laurell Technologies Corporation, North Wales,
PA, USA).
3. Vacuum desiccator capable of achieving P ¼ 25 Torr pressure
(Bel-Art 230 mm or equivalent).
4. Vacuum pump with ~140 L/min (5.0 Cfm) capacity (model
6912; FJC, Mooresville, NC, USA) or similar.
5. Oven: Set to 65  C for PDMS curing.
6. Hot plates: One in cleanroom for photoresist processing and
one for PDMS processing.
7. Stereo-microscope.
8. Cell culture hood.
306 Prasenjit Bose et al.

9. Cold ice pack: Corning™ XT Starter Ice-free cooler (catalog

number 432015; Corning, Corning, NY, USA).
10. Temperature-controlled centrifuge: e.g., Sorvall RT1 or
Eppendorf 5810R with swing-bucket rotor capable of holding
P35 dishes, and microtiter plate buckets (ThermoFisher Scien-
tific, Grand Island, NY, USA).
11. Epifluorescence microscope: Zeiss Axiovert 200M, Nikon
Eclipse Ti, Nikon TE-2000, or equivalent equipped with 4

and 10
objectives, and a live cell chamber (37  C and 5–10%
CO2). A programmable xyz-stage is helpful but not essential.
12. Fluorescence microscopy camera: CoolSnap HQ (Photo-
metrics, Tucson, AZ, USA) or equivalent.
13. High-speed microscope camera with up to 100 frames/s capa-
bility in 8 bit mode, and at least 1 MPixel image size: Prosilica,
GX (Allied Vision Technologies, Exton, PA, USA) or
equivalent.
14. Magnetic tweezer: homebuilt (see Note 1).
15. Computer Aided Design (CAD) software (AutoCAD) or Illus-
trator (Adobe) for mask design.
16. Computer with multi-channel ADC/DAC card and LabVIEW
software or other instrument control software (catalog number
NI PCIe-6321; National Instruments Corporation, Austin,
TX, USA).
17. Computer with software to control fast camera capable of
recording multi-minute movies at 100 frames/s. StreamPix
(Norpix, Montreal, Quebec, Canada), or equivalent.
18. Programmable bipolar power supply capable of supplying up to
2 A current: (model BOP 50-2; Kepco Power Solutions,
New York City, NY, USA).
19. Hall probe: to monitor field of magnetic tweezer (catalog
number HGT-2100-10; Lakeshore Cryotronics, Westerville,
OH, USA).
20. Image analysis software: Igor Pro (Wavemetrics Inc., Lake
Oswego, OR, USA), Matlab (Mathworks, Natick, MA,
USA), ImageJ (https://imagej.nih.gov/ij/) with Spot tracker
plugin (http://bigwww.epfl.ch/sage/soft/spottracker/).

2.2 Photolithography 1. Mylar masks: patterns printed at 40,000 dots per inch (DPI).
2. Blank glass mask plate: 400
400 to fit in mask aligner (see Note 2).
3. Silicon wafers: h100i Si wafers with 300 diameter and thickness
350–400 μm (see Note 2).
4. Precoated glass plate: 400
400 glass plate precoated with chro-
mium and 0.5 μm of positive photoresist, AZ 1518 (Nanofilm
Inc., Westlake Village, CA, USA).
 Microtissues to Study Cell-Matrix Interactions 307

5. Cleaning agents: acetone, isopropyl alcohol (IPA), ethanol,

deionized water, and low-pressure nitrogen gas (~350 mbar,
5 pounds per square inch (PSI)).
6. Chromium etchant (catalog number 651826; Sigma-Aldrich
Chemical Company, St. Louis, MO, USA).
7. Photoresists: SU-8 2002 (catalog number Y111029; Micro-
Chem Corp, Westborough, MA, USA), SU-8 2050 (catalog
number Y111045; MicroChem Corp), and SU-8 2010 (cata-
log number Y111058; MicroChem Corp) mixed with Micro-
posit S1813 photoresist (catalog number 41280; Shipley Co
Inc., Atlanta GA, USA), 70/30 v/v ratio.
8. Photoresist developers: AZ 400K (1:4 dilution in deionized
water), Propylene glycol monomethyl ether acetate (PGMEA)
SU-8 developer (catalog number Y020100; MicroChem Corp).
9. Precision tweezers and glass petri dishes large enough to hold
Si wafers.

2.3 Replica Molding 1. Poly(dimethyl siloxane) (PDMS) and curing agent Sylgard
and μTUG Device 184 elastomer kit (catalog number 2065662; Dow Corning
Fabrication Corporation, Auburn, MI, USA).
2. Smooth-Cast® 305, Parts A and B (Smooth-On Inc., Macun-
gie, PA, USA).
3. Tridecafluoroctyltrichlorosilane (catalog number 78560-45-9;
UCT Specialties, Bristol, PA, USA).
4. Nickel spheres: Ni powder 74–116 μm (mesh size: 150 þ 200;
catalog number 7440-02-0; Alfa Aesar, Ward Hill, MA, USA).
5. Fluorescent beads: Fluoresbrite YG carboxylate microspheres
2 μm diameter (catalog number 09847-5; Polysciences Inc.,
Warrington, PA, USA) diluted 1:3000 in 100% Ethyl Alcohol
200 proof, Absolute, Undenatured, A.C.S./USP Grade).
6. Plastic culture dishes and plates: P35 (35 mm) culture dishes
(catalog number 353001 or similar), 12-well plates (catalog
number 353043 or similar), 24-well plates (catalog number
353226 or similar; BD Biosciences, Bedford, MA, USA).
7. 44 mm aluminum weighing dishes (catalog number 25433-
016; VWR International, Radnor, PA, USA).
8. Precision nonmagnetic tweezers with 50 μm tips, razor blades,
50 mL centrifuge tubes.

2.4 Microtissue 1. 0.2% Pluronic® F-127: Prepare from mixing Pluronic® F-127
Seeding and Culture powder (catalog number P2443; Sigma-Aldrich Chemical
Company) or 10% Pluronic® F-127 solution (catalog number
P6866; Thermo Fisher Inc.) with deionized water or
Phosphate-Buffered Saline (PBS, pH 7.4) and followed by
308 Prasenjit Bose et al.

sterilizing with 0.22 μm filter (catalog number SCGP00525)

(Merck Millipore KGaA, Darmstadt, Germany).
2. UV sterilizer.
3. 70% Ethyl Alcohol: 100% Ethyl Alcohol (200 Proof, Absolute,
Undenatured, A.C.S./USP Grade) diluted with distilled water.
4. Phosphate-buffered saline (PBS), pH ¼ 7.4 (catalog number
10010-023; ThermoFisher Scientific).
5. Cell dissociation agents: Trypsin (0.05% w/v)-EDTA (2 mM)
(catalog number 25300054; ThermoFisher Scientific), Accu-
tase™ (catalog number AT104; Innovative Cell Technologies,
San Diego, CA, USA) or TrypLE™ Express (catalog number
12604013; ThermoFisher Scientific).
6. Growth medium: Dulbecco’s Modified Eagle Medium (high
glucose) (catalog number 11965092; ThermoFisher Scien-
tific), supplemented with 10% Fetal Bovine Serum or as appro-
priate for cell type in use.
7. Extracellular matrix buffer components: Prepare stock solu-
tions of 1 M NaOH (Sigma-Aldrich Corp), 10
Medium
199 (catalog number 11825015; ThermoFisher Scientific),
5% (w/v) NaHCO3 (catalog number S5761; Sigma-Aldrich
Corp), and 1 M HEPES (4-(2-hydroxyethyl)-1-piperazi-
neethanesulfonic acid) (catalog number 15630-80; Thermo-
Fisher Scientific) diluted in water to 250 mM.
8. Extracellular matrix: Rat Tail type I Collagen (catalog number
354236, 100 mg; Corning®, Corning, NY, USA), human
fibrinogen (catalog number F3879; Sigma-Aldrich Corp),
bovine fibrinogen (catalog number F8630; Sigma-Aldrich
Corp) and other matrix components as needed.
9. Extracellular matrix solution: Calculate the volumes of the
following components to reach their desire final concentrations
and mix them thoroughly. See example in Table 1.
(a) 10
M199: to reach final concentration of 1
in the final
matrix solution.
(b) 5% NaHCO3: to reach final concentration of 0.035%
(w/v).
(c) 250 mM HEPES: to reach final concentration of 10 mM.
(d) S-DI water: to bring the solution to final volume.
(e) Collagen type I: to reach desired concentration.
(f) Fibrin or fibrinogen (if needed): to reach desired
concentration.
(g) 1 M NaOH: adjust the volume ratio of the NaOH to stock
collagen solution to be 0.022:1.
10. Sterilized Deionized Water (S-DI water).
 Microtissues to Study Cell-Matrix Interactions 309

Table 1
Example of creating 2 mL of extracellular matrix solution with 2.5 mg/mL
of collagen and 2 mg/mL of fibrinogen

Component Volume (μL) Final concentration

Water 307.8
10
M199 200 1

HEPES (250 mM) 83 10 mM
NaHCO3 (5% w/v) 14 0.035% (w/v)
NaOH (1 M) 29
Collagen (3.77 mg/mL) 1326.2 2.5 mg/mL
Fibrinogen (100 mg/mL) 40 2 mg/mL
Total 2000

11. Detergent: Triton X-100 (catalog number H5142; Promega

Corporation, Madison, WI, USA).
12. Lids from P35 dishes filled with PDMS (see Note 3).

3 Methods

3.1 Master This section describes fabrication of original “master” versions of

Fabrication μTUG device arrays out of SU-8 photoresist using multilayer pho-
tolithography (see Note 4). Three major steps are involved: Design,
production of chromium-on-glass masks, and fabrication of the
master device. Production of the masks and masters should ideally
be carried out in a cleanroom setting. The master μTUG device is
made from sequential multilayer deposition of photoresists of vary-
ing thickness on a silicon wafer, followed by UV exposure of each
layer and final development in SU-8 developer (see Note 5). The
process flow is shown in Fig. 2.
1. Design patterns for the two masks needed using CAD software,
and print on Mylar sheets at 40,000 DPI (see Note 6). Use one
mask to define the pillars’ stems and the second mask to define
their heads. As negative photoresists are used for device fabrica-
tion, transparent regions in the masks correspond to protruding
or raised features in the masters and the resulting devices. An
example of patterns for a mask set is shown in Fig. 3.
2. To make a chromium-on-glass mask, attach the Mylar mask on
a blank glass plate and expose the AZ1518 coated chromium-
on-glass plate with UV light under a mask aligner using a total
dose of 70 mJ/cm2 (see Note 7).
310 Prasenjit Bose et al.

Fig. 2 Schematics of the fabrication of SU-8 master mold. (a) Deposit photoresist base layer, first layer and
blocking layer on Si wafer. (b) Expose under first layer mask. (c) Deposit second layer of SU-8. (d) Expose
under second layer mask after alignment. (e) Develop with SU-8 developer. (f) Image of an SU-8 master mold
array on a silicon wafer and a magnified image of one microwell

3. Develop the plate for 45 s in 1:4 diluted solution of AZ 400K

in DI water (see Note 8).
4. Immediately wash the plate with DI water and dry with
nitrogen gas.
5. Remove the visible chromium sections on the plate, by immers-
ing it into chromium etchant. Keep it immersed until the
transparent areas through which light is allowed to pass can
be seen by the naked eye (typically 1–2 min).
6. Wash and dry again with DI water and nitrogen gas (see
Note 9).
7. Remove the leftover developed AZ 1518 by washing the plate
in acetone.
8. Clean the plate using IPA and dry with nitrogen gas.
9. The chromium-on-glass mask is now ready to be used for
μTUG master fabrication, which is described in the subsequent
steps (see Note 10).
10. Repeat steps 2–9 for the second mask.
11. Prior to resist deposition, clean the silicon wafer with acetone,
followed by IPA.
12. Dry the wafer at 200  C for 2 min on a hot plate. After drying,
allow it to cool to room temperature (~2 min).
13. Preparation of base layer: Create a 2.4 μm thick layer of SU-8
2002 on the wafer using a spin coater with the following
settings: 10 s at 500 rpm þ 30 s at 2000 rpm (see Note 11).
 Microtissues to Study Cell-Matrix Interactions 311

Fig. 3 Mask set to produce an array of μTUG devices in a P35 culture dish. (a)
Bottom layer mask used to define the pillars’ stems. (b) Top layer mask that
defines the pillars’ heads. A magnified view of one microwell on each mask is
shown. The complimentarily shaded, circumscribed star patterns in the bottom
corners are used for mask and substrate alignment as described in Note 16. The
micro-wells are 800 μm
400 μm in size, and the horizontal pillar-to-pillar
spacing is 500 μm. In the well shown magnified, the pillar stems’ cross-sectional
dimensions are 150 μm
50 μm, and those of the heads’ are 200 μm
90 μm.
This yields pillars with effective spring constants k ¼ 2 μN/μm. The other three
groups of μTUGs (listed top to bottom) had pillar stem dimensions of
150 μm
26 μm, 150
33 μm, and 150 μm
42 μm, pillar head
dimensions of 200 μm
65 μm, 200 μm
75 μm, and 200 μm
80 μm,
and corresponding spring constants k ¼ 0.25 μN/μm, 0.5 μN/μm, and 1 μN/μm,
respectively. All quoted k values assume a 10:1 PDMS mixing ratio (Young’s
modulus ¼ 1.6 MPa)

14. Bake on a hot plate at 95  C for 2 min (soft bake), followed by UV

exposure under a mask aligner with a total dose of 100 mJ/cm2.
15. Do a post exposure bake (PEB) at 95  C for 2 min just after the
exposure (see Note 12).
16. Preparation of first layer: Deposit 130 μm of SU8 2050, using
the following spin coater settings: 60 s at 500 rpm þ 60 s at
900 rpm (see Note 13).
312 Prasenjit Bose et al.

17. Soft bake at 95  C for 3.5 h, and then allow the wafer to cool to
room temperature.
18. Preparation of blocking layer: Deposit a 20 μm thick layer
from a v/v mixture of 30% S1813 in SU-8 2010 on the wafer
via the following spin coater settings: 10 s at 500 rpm þ 30 s at
1000 rpm.
19. Soft bake at 95  C for 30 min. Remove and allow to cool to
room temperature.
20. Once the sample cools to room temperature, expose it under
the mask aligner using the first layer chrome mask. Keep the
total dose at 700 mJ/cm2 and insert a low pass filter in the path
of the light to block out UV rays of wavelength below 350 nm
(see Note 14). The high dose penetrates the blocking layer and
exposes the first layer.
21. Do a post-exposure bake at 95  C for 12 min (see Note 15).
22. Preparation of final layer: Deposit the final layer of 50 μm of
SU-8 2050 using the following spin coater settings: 60 s at
500 rpm þ 60 s at 1500 rpm.
23. Soft bake at 95  C for 3 h.
24. Once the sample cools to room temperature, expose the rest of
the wafer using the second layer mask. Use the mask aligner to
align patterns on the mask to the patterns on the wafer (see
Note 16).
25. The UV light dose at this step should be between
100–110 mJ/cm2 in the presence of the low pass filter (see
Note 17).
26. Follow this exposure with a post exposure bake at 95  C for
12 min. After a few minutes of cooling down, the wafer is ready
for development.
27. Develop the wafer in a glass petri dish filled with PGMEA for
45 min on a stirrer or shaker (see Note 18).
28. Once the development is done, wash the wafer in IPA for 90 s,
and then dry using nitrogen gas (see Note 19).
29. The master fabrication is now complete.

3.2 Replica Molding 1. Generation of negative molds. The ‘negative’ molds required
and PDMS Device to make the final devices are made from PDMS. The process is
Fabrication sketched in Fig. 4a–c.
2. Combine PDMS and curing agent in a 20:1 ratio (see Note
20), and mix thoroughly for 20 min to obtain a uniform
solution.
3. Centrifuge the solution in a 50 mL tube at 720
g (see Note
21) for 1 min to remove any bubbles.
 Microtissues to Study Cell-Matrix Interactions 313

Fig. 4 Schematics of replica molding and PDMS device fabrication. (a) Silanization of SU-8 master molds. (b)
Cast PDMS and cure at 65  C. (c) Peel off PDMS negative mold. (d) Silanization of PDMS negative mold. (e)
Cast PDMS and cure at 65  C. (f) Peel off the PDMS devices

4. Place the μTUG master flat in a large plastic weighing dish, and
add the PDMS þ curing agent solution on top to a depth of
3–5 mm.
5. Insert the entire dish into a vacuum desiccator, and hold at
P ¼ 25 Torr for 1 h to remove all trapped air.
6. Bake in a 65  C convection oven for 24 h to cure the PDMS.
7. Separate the PDMS mold from the master (see Note 22) and
trim away excess PDMS around the features to form squares
~1.75
1.75 cm2.
8. Generation of Smooth-Cast® master copies. To make these
durable masters, attach negative molds (pattern side facing
up) in 44 mm aluminum weighing dishes using double-sided
tape (see Note 23).
9. Prepare a 1:1 volume mixture of Smooth-Cast 305 A and
305 B, mix for 30 s, add on top of the molds to cover them,
and degas for 5 min using a vacuum desiccator at P ¼ 25 Torr
(see Note 24).
10. The Smooth-Cast mixture hardens within 3–4 h at room tem-
perature and the molds can be peeled off (use ethanol for
lubrication) (see Note 25).
11. μTUG device production from molds. The final μTUG devices
are made from PDMS, as outlined in Fig. 4d–f and described in
subsequent steps. The PDMS to curing agent ratio may be
varied to control the pillars’ stiffness.
314 Prasenjit Bose et al.

12. Place a mold made from the masters or the Smooth-Cast copies
in a glass petri dish (pattern side up), clean with ethanol on a
shaker for 5 min, and dry with nitrogen gas (see Note 26).
13. To remove any organic impurities and to make the surface
hydrophilic, clean the molds using an oxygen plasma prior to
each use. We recommend a power of 25 W and pressure of
450 mTorr for 60 s.
14. To “silanize” the molds, place the plasma-cleaned molds in a
vacuum desiccator, together with a few drops of tridecafluor-
octyltrichlorosilane on a separate coverslip.
15. Seal the desiccator under vacuum (P ¼ 25 Torr) and expose the
molds to the silane fumes for at least 12 h. This creates a thin
layer of silane over the PDMS mold to facilitate separation of
the device from the mold.
16. Prepare a mixture of PDMS and curing agent. Together with
the pillar geometry, the PDMS to curing agent ratio deter-
mines the pillar stiffness of the μTUGs [10]. The Young’s
modulus for 20:1, 10:1, and 4:1 mixtures is 0.54, 1.6, and
4 MPa, respectively.
17. To produce μTUG devices in a P35 culture dish, add 2 mL of
PDMS solution to the dish, followed by quick degassing
(10 min at P ¼ 25 Torr).
18. Heat the dish on a hot plate at 65  C for 20 min. This step is
done to harden this bottom layer of PDMS somewhat (see
Note 27).
19. While the bottom layer of PDMS in the petri dish is curing,
pour 1 mL of PDMS on top of the silane-treated mold and
degas (~15 min at P ¼ 25 Torr).
20. Invert the mold and gently push down through the liquid
PDMS and onto the slightly hardened PDMS in the petri dish.
21. Top off the sides of the mold with excess PDMS. It may be
necessary to degas the dish for 2–3 h to remove air bubbles.
22. After all the bubbles are removed, cure the PDMS in a 65  C
oven for 24 h.
23. Once the dishes are cured, the molds are peeled off slowly from
the device. This is done by making incisions along the edge of
the mold device interface from the top using a thin tweezer or a
razor blade, and then prying off the mold from one or multiple
sides using the flat edge of the tweezer or blade. Squirting
ethanol in the fissures is highly recommended as it facilitates
the separation and prevents the μTUG pillars from breaking.
24. To create μTUG devices in 12- or 24-well plates, cut the molds
into smaller cubes with surface area near 1 cm2, and employ the
 Microtissues to Study Cell-Matrix Interactions 315

process described in steps 12–23 above with the following

modifications.
25. Use 1 mL of PDMS solution per well for 12-well plates, and
500 μL per well for 24-well plates.
26. For steps 17–21, to harden the bottom layer of PDMS in 12-
and 24-well plates, place the plates on a hot plate at 65  C for
30 min (see Note 27).
27. If desired, the pillars in the μTUG wells can be labeled with
fluorescent microbeads to aid in tracking their deflections. To
accomplish this, after plasma cleaning and silanizing the molds
(step 14 above), pour a solution of ethanol and fluorescent
beads on top of the molds, degas for 15 min, and centrifuge at
605
g for 3 min to drive the beads into the pillar forms of the
molds (see Note 28).
28. Remove the excess ethanol and dry the molds overnight in a
65  C oven.
29. Resume the device fabrication protocol from step 16 above.
30. Attaching magnetic spheres to pillars. To enable magnetic
actuation of the μTUG pillars, glue magnetic nickel spheres
on top of them. Nickel particles of mesh size 150–200
(74–116 μm) are used for this step.
31. Working under a stereomicroscope, immerse a few such parti-
cles in PDMS solution (10:1 ratio), select for sphericality using
precision tweezers, and place them on top of the desired pillars.
32. Cure the PDMS solution to bond the spheres to the pillars (see
Note 29). For 2-pillar μTUGs, our standard approach is to
bond Ni spheres to one pillar per microwell.

3.3 Passivation 1. Sterilize the devices by placing them in a UV chamber or hood.

of Arrays Alternatively, incubate with 1 mL of 70% ethanol for 10 min,
and dry with nitrogen gas.
2. Add 1 mL of freshly prepared 0.2% Pluronic® F-127 to the
devices.
3. Transfer to desiccator and degas at P ¼ 25 Torr until air
bubbles are visible in the microwells of the devices.
4. Centrifuge the devices at 420
g for 30 s to get rid of bubbles
in the microwells.
5. Transfer the devices to the cell culture hood and let the Pluro-
nic® incubate at room temperature for 10 min (or longer) (see
Note 30) while preparing collagen solution in the seeding
steps below (see Note 31).
6. To fully remove the residual Pluronic®, aspirate the Pluronic®
out of the devices.
7. Rinse once with 1 mL of PBS, and dry with nitrogen gas.
316 Prasenjit Bose et al.

8. Let the devices stand on ice while finishing the preparation of

the cells þ collagen solution.

3.4 Cell Seeding 1. Detach and dissociate the cells by Trypsin, Accutase or TrypLE,
neutralize with fresh media and count the cells (see Note 32.
For fibroblasts, adjust the cell concentration to
~300,000 cells/mL and take 1 mL of cell solution (see Note
33). This volume varies for 12- and 24-well plates, depending
on the device size. Centrifuge at 180
g for 5 min, resuspend
the cells in fresh media and keep the cells on ice).
2. Place all the ECM buffer components and the soluble ECM
(1 M NaOH, 10
M199, 5% NaHCO3, 250 mM HEPES,
S-DI water, collagen and (if needed) fibrin or fibrinogen) on ice.
3. Place μTUG device(s) on a cold ice pack.
4. Determine the desired concentration of matrix components
(collagen, fibrin and/or fibrinogen.) Add calculated volumes
of each component to a 15 mL conical tube (see items 8 and
9 of Subheadings 2.4 and Table 1) and mix thoroughly by
repeatedly and gently pipetting (see Note 34).
5. Add 1 mL of ECM mixture to μTUGs in P35 dish or 250 μL
per well in 12-well or 24-well plates.
6. Place μTUGs along with ice pack in a vacuum desiccator to
degas (2–3 min at P ¼ 25 Torr) (see Note 35), or pipet up and
down swiftly to destroy any air bubbles that may have formed
in the collagen solution.
7. Spin down the cell solution from step 1 at 180
g for 3 min,
and replace the media solution with 500 μL of the ECM
mixture for P35 dishes to reach the desired cell number. This
volume varies for 12- and 24-well plates, depending on the
device size (see Note 36).
8. Transfer the above cell-ECM solution to the μTUGs and mix
by pipetting (see Note 37).
9. Spin the μTUGs at 237
g for 90 s at 9  C (see Note 38) to
pull the cells down into the wells.
10. For P35 dishes, rotate the dishes by 90 and spin at 237
g for
90 s at 9  C to even out effects on the distribution of the
solution due to the tangential acceleration. This additional
spinning step is optional for 12- and 24-well plates.
11. Transfer the μTUGs on ice pack to the cell culture hood.
12. Hold the ice pack slightly tilted towards you.
13. Aspirate the collagen mixture. Start with the aspirator at one of
the top corners of the device, and draw the aspirator horizon-
tally to the other top corner to detach the liquid from the
upper sidewall of the dish. Then gradually slide the aspirator
 Microtissues to Study Cell-Matrix Interactions 317

down along the side, staying away from the microwells (see
Note 39).
14. Invert the dish or plates so that they are now upside down, and
spin at 40–60
g for 15 s at 9  C (see Note 40).
15. Keep the μTUGs inverted on an ice block when transferring
from centrifuge to hood.
16. Add 500 μL of H2O or PBS in the lid of the P35 dish or 3 mL
to 12-well or 24-well plates.
17. Incubate the μTUGs with the lid down (inverted) for
10–15 min at 37  C until the collagen is polymerized (see
Note 41).
18. Aspirate water and gently add medium at the corner (1 mL for
P35 dish and 250 μL per well for 12-well or 24-well plates).
19. Place the dishes or plates in the incubator for culturing to allow
the microtissues to form.

3.5 Culturing 1. For fibroblast microtissues, incubate the microtissues at 37  C

Microtissues with 5% CO2 (see Note 42).
2. Replace medium for fibroblast microtissues every day (see Note
43).
3. Change medium one device at a time. This is important as
PDMS is hydrophobic, and total removal of medium will result
in the tissues drying out rapidly. Therefore, do not aspirate all
of the medium, but leave a thin layer to keep the microtissues
moist.
4. During microtissue growth, when replacing medium, aspirate
medium and gently add an adequate volume of fresh medium
at the corner of the device. Typical volumes for devices in P35
dishes range between 1 to 2 mL.

3.6 Contractility The contractile forces exerted by the microtissues are obtained by
Measurements tracking the deflection of the pillars to which the tissues are
adhered. This is achieved by measuring the positions of the fluores-
cent microbeads on the top of the pillars or by tracking the edges of
the pillars through imaging.
1. To record the initial compaction and formation of the micro-
tissues, which typically takes 12–24 h, mount the P35 dish or
multi-well plate containing the microtissues in a live-cell cham-
ber mounted on top of the microscope stage at 37  C with the
CO2 concentration set between 5–10% depending on the cell
type used (see Note 44).
2. To maximize the ability to resolve the pillars’ motion, choose a
microscope objective so that one microtissue fills the field of
view (typically 4
or 10
). However, make sure that static
318 Prasenjit Bose et al.

features, such as the boundaries of the well, are visible to

provide reference points for the measurement the absolute
deflection of the pillars, and to use in registering successive
images should unwanted horizontal motion occur.
3. Record images for each well in the device every 1–2 h.
4. At each time point, record a white light (phase contrast) image
of each tissue, focused on the tops of the pillars. If the pillars are
labeled with fluorescent beads, record a fluorescent image
as well.
5. For longer time-lapse studies (from 2–3 days to >1 week) that
record the long-term tissue growth and development, culture
the microtissues in a conventional incubator, and bring the
devices to a microscope periodically (see Note 45) to record
single white-light (or white light and fluorescence) images at
12 or 24 h intervals (see Note 46).
6. At the end of the data acquisition, replace the medium by either
trypsin, collagenase or a detergent, such as Triton X-100, for
10 min, and take a final set of images to determine the baseline
position of the pillars for each microtissue.
7. Analyze the data to obtain the force and stress on each micro-
tissue, following the procedures of Subheading 3.8.

3.7 Magnetically The system for obtaining stress–strain measurements of microtis-

Actuated sues via actuation of the magnetic sphere-tagged pillars includes a
Stress–Strain magnetic tweezer and associated electronic controls to generate
Measurements local magnetic fields to apply forces to the magnetic spheres (see
Note 1). The field-dependent magnetic force acting on the nickel
spheres can enable multiple different types of stretching experi-
ments. Quasistatic stretching, stretch-recovery and dynamic
stretching are described, along with associated image analysis and
data reduction methods.
1. Design of the magnetic tweezer system. A magnetic tweezer
consists of a sharpened soft iron core inserted into a solenoid
and mounted on a three-axis micromanipulator (Fig. 5) (see
Note 47).
2. Mount the micromanipulator/tweezer assembly on a fixed
plate mounted above the microscope’s moveable stage. The
tweezer’s sharp pole tip should project through a hole in this
plate and down into the culture dish containing the microtis-
sues, which sits on the moveable stage. This allows the pole tip
of the tweezer to remain in the field of view of the microscope,
while microtissues are successively brought into proximity of
the tip for measurements. Mount a Hall probe at the blunt end
of the iron core away from the sample to monitor the magni-
tude of the field (see Note 48).
 Microtissues to Study Cell-Matrix Interactions 319

Fig. 5 Magnetic tweezer system. (a) shows a schematic and (b) shows a photograph of the unit mounted on a
microscope. The long, tapered rod is an iron core that acts as a magnetic pole tip projecting into a P35 μTUG
sample dish. The tweezer solenoid surrounds the core between the arms of the aluminum bracket on a 3-axis
micromanipulator and is encased in an aluminum heat-sinking block. A second manipulator and coil assembly
(unused in this application) is shown at left in (b) without its core. The image in (b) is reproduced from A. S. Liu,
Ph.D. Thesis, the Johns Hopkins University (2015). Used by permission

3. Electronic control system. Use a bipolar power supply to pro-

vide the electric current that is passed through the solenoid
coils to generate a magnetic field.
4. Control the bipolar power supply with a computer-controlled
digital to analog converter (DAC) card. The Hall probe is read
with an analog to digital converter (ADC) channel in the
same card.
5. Magnetic tweezer positioning and local magnetic force adjust-
ment. Position the tweezer with its tip at the edge of the
microwell under study closest to the magnetic pillar (Fig. 1),
using the microscope’s stage for coarse motion, and the micro-
manipulator for fine alignment (see Notes 49 and 50).
6. Raise the tweezer tip when moving between microwells to
avoid damaging the devices.
7. Magnetic Stretching Protocols and Data Acquisition. The
force on the nickel sphere is given by F ¼ ∇(μ(B)·B), where
μ(B) is the field-dependent magnetic moment of the sphere
induced by the magnetic field B produced by the magnetic
tweezer tip (see Notes 51 and 52). For fields B < 100 mT,
μ ~ B, and hence F ~ B2 [12, 13]. Hence to obtain a linearly
increasing applied p ffiffiffi with time F / t, the magnetic field has
force
to increase as B / t , which is achieved with solenoid currents
pffiffiffi
I / t.
8. The field-dependent magnetic force acting on the nickel
spheres can enable multiple different types of stretching experi-
ments, including Quasistatic stretching, stretch-recovery and
dynamic stretching.
320 Prasenjit Bose et al.

(a) Quasistatic stretching. To acquire basic stress–strain data,

e.g., to determine a microtissue’s elastic modulus, increase
the magnetic field in a stepwise linear fashion. After each
step, record a phase contrast image and, if the pillars are
fluorescently labeled, an epifluorescence image of the
microtissue. One should confirm that there is no creep
of the microtissue during imaging (see Note 53).
(b) Stretch-recovery. To study the cells’ active response to
stretch, for example in smooth muscle microtissues [15],
one may use a triangle-shaped stretch-relaxation protocol
wherein the magnetic force is increased to 25–35 μN over
120 s, and then decreased to zero over a similar time
interval. The microtissues are then observed for an addi-
tional 10–15 min at 5–10 frames per min as they recover
to their baseline states due to the active cytoskeletal
dynamics of the cells (see Note 54).
(c) Dynamic stretching. For higher strain rates, or where
higher temporal resolution is desired, use a fast camera
and record white light images only.

3.8 Data Reduction For both contractility measurements (Subheading 3.6) and stress–-
and Analysis strain measurements (Subheading 3.7), the tissue force is measured
from the deflection of the nonmagnetic pillar(s), using the relation
F ¼ kδ, where k is the pillar’s bending constant, and δ is the
measured pillar deflection. The value of k is calculated for small
deflections from Euler-Bernoulli beam theory. For pillars with
heads, the bending constant is given by, k ¼ a2 ð6EI 3La Þ
, where E is
the Young’s modulus of PDMS, I is the area moment of inertia of
the pillar stem’s cross section, a is the height measured to the center
of the pillar head, and L is the total height of the pillar (see Note 55)
[10]. Strain is measured either from the length change of the
microtissue, or from measurements of the local strain in the central
region of the microtissues via a texture correlation algorithm (see
Note 56) [18].
1. Pillar tracking: fluorescent labeling. When relatively small
numbers of images (<~150) are acquired, and the pillars are
fluorescently labeled, import images into ImageJ and convert
images into a stack.
2. Use the Spot Tracker plugin [19] to analyze the changes of
particle locations at different time point.
3. For larger fluorescently labeled data sets, use automated parti-
cle tracking approaches [20].
4. Pillar tracking: white light. We use custom image analysis
written in the software package Igor Pro (Wavemetrics) to
track the motion of identifiable features, such as the edge of a
moving pillar (see Note 57).
 Microtissues to Study Cell-Matrix Interactions 321

5. Define a region of interest (ROI) for each pillar that encom-

passes the horizontal motion of the feature of interest.
6. Average the pixel intensity values along the y-direction of the
ROI (perpendicular to the motion) and assign a single value to
the corresponding x-position (see Note 58).
7. Fit the positions of the features (edges or peaks) for each frame
to determine the pillars’ motion.
8. Microtissue stress determination. Stress is calculated by divid-
ing the force measured by the nonmagnetic pillar by the cross-
sectional area of the tissue.
9. The cross-sectional area is estimated by finding a linear relation
between the width and thickness of the tissue at its central
position from confocal imaging of selected microtissues.
10. Microtissue strain determination. Rudimentary deformation
is calculated by finding the separation of the current positions
of the pillars and subtracting the original unstretched pillar
separation. Dividing the deformation by original pillar separa-
tion provides the strain induced in the tissue.
11. For more precise strain measurements, a texture correlation
approach may be used [18].
12. Assign a grid of nodes with spacing ~10 μm over the central
region of the microtissue, and track the displacement of each
node in the stretched images by examining the correlation of
the intensity pattern surrounding the node in the stretched and
unstretched configurations.
13. Calculate the strain of each grid element in the stretching
direction as ε ¼ ΔL/L, where ΔL is the distance change
between adjacent nodes.
14. The strain field may be averaged over the central region of the
tissue to report an average strain.
15. The tissues’ elastic moduli are determined from the slopes of
the resulting stress–strain plots.

4 Notes

1. Complete details on the construction of the magnetic tweezer

system are beyond the scope of this methods chapter. The basic
principles involved and a description of key aspects of the
system in the authors’ lab are given in Subheading 3.7. Fabri-
cation of mechanical components and assembly of the system
requires access to a standard machine shop. Further informa-
tion on how to construct such a system may be obtained by
contacting the authors.
322 Prasenjit Bose et al.

2. All photolithography descriptions assume use of a 300 mask

aligner. Adjust accordingly for a different system.
3. Make lids of P35 dishes filled with cured PDMS to use as covers
during time lapse microscopy. Because PDMS is gas permeable
but not water permeable, PDMS-filled lids will prevent drying
out of the samples.
4. A complete guide to photolithography techniques is beyond
the scope of this methods chapter. However, the facilities and
technical expertise needed to support this part of these Meth-
ods should be available at most universities or laboratories, or
through national facilities such as the US National Nanotech-
nology Coordinated Infrastructure network (http://www.
nnci.net).
5. The main challenge in fabricating the μTUGs is that the heads
of the “T-shaped” pillars are typically larger than their stems by
~50 μm in each dimension. This necessitates multilayer photo-
lithography with separate photomasks used to define the pillar
stems and heads, and careful alignment of the corresponding
exposures.
6. Mylar masks are printed by commercial vendors using high-
resolution printers.
7. It is recommended, but not crucial, to make chromium-on-
glass copies of the original Mylar masks for better pattern
transfer to the Si wafer.
8. Stirring the petri dish that contains the developer and the
sample by hand or using a shaker is recommended for uniform
development of samples.
9. Rinse in DI water for 5 s and blow dry with nitrogen gas until
all the water droplets are removed.
10. The shapes in the chrome masks should be accurate to at least
10 μm.
11. Since wafers are typically centered on the axis of rotation of the
photoresist spinner, spinning speeds are most naturally given in
rpm rather than in equivalent “g” forces, as is customary for
centrifugation.
12. The spin speeds, times, and UV exposure doses need to be
calibrated for every new bottle of photoresist and for the
particular mask aligner used.
13. After pouring SU8 2050 on the wafer, it is recommended to
wait for 5–10 min before spinning to release any trapped
bubbles.
14. If UV light below 350 nm is allowed to expose the substrate,
then the features typically end up being somewhat curved
 Microtissues to Study Cell-Matrix Interactions 323

instead of straight. Hence it is recommended to use a low pass

UV filter to block UV rays <350 nm.
15. If the exposure is sufficient, then a visible latent image of the
pattern is seen in the resist within 5–15 s after being placed on
the PEB hotplate.
16. It is recommended to include alignment marks on the masks,
e.g., additional shapes that have distinct symmetries, such as
stars and crosses. This facilitates alignment of features on the
second mask to the corresponding features on the wafer pro-
duced by the first mask.
17. This particular step is exposure sensitive. If the exposure is low,
then the top layer can peel off from the blocking layer during
development. If the exposure is high, then the light can ‘bleed
into’ the first layer. Hence the exact exposure dose needs to be
optimized such that it is enough to get into the blocking layer,
but does not bleed through it.
18. While developing, the speed of the shaker/stirrer has to be low
so that the features do not break from the agitation. Also, the
development times need to be calibrated, as development is
dependent significantly on agitation. The use of an ultrasonic
bath is not recommended as the vibrations can break features in
the patterns.
19. Drying must be done gently to avoid breaking fragile features
in the SU8 structures.
20. 20:1 PDMS to curing agent ratio is used because the standard
10:1 ratio yields molds that are too stiff, leading to pillar heads
tearing off when molds are peeled off μTUG devices.
21. We use rotors A-4-44 (rotor radius ¼ 16.1 cm) A-2-DWP
(rotor radius ¼ 14.7 cm). The relative centrifugal force
(RCF) is related to the rotor radius r and the rotation speed
of the centrifuge. The g force to rpm conversion formula is as
follows:
RCF (g) ¼ 1.118
105
r
(rpm)2, where r is in cm.
22. While removing the mold from the master, it is recommended
to cut the plastic weighing dish off the PDMS mold, and then
slowly pour ethanol into the interface between the master and
mold. This prevents the master from breaking.
23. The SU-8 masters on Si wafers are quite brittle, and tend to
break after a few repetitions of mold fabrication. Thus, stronger
masters should be made from Smooth-Cast®.
24. While degassing, the Smooth-Cast® liquid mixture starts to
boil, and so the vacuum has to be turned on and off a few times
to let the liquid settle.
324 Prasenjit Bose et al.

25. The resulting hardened plastic shapes have the same pattern as
the original masters and can be utilized to make new molds
following the procedure given in step 1 above.
26. Typically, between four to eight molds are treated simulta-
neously to make multiple μTUG devices in P35 dishes.
27. If the PDMS at the bottom of the dish is not cured well, then
the mold will sink all the way to the bottom of the dish when
inverted onto the dish. This makes peeling off the mold from
the completed devices difficult. On the other hand, if the
PDMS is over-cured, then inverting the mold onto the surface
can often lead to trapping of air bubbles.
28. Some optimization of this process is needed, as tracking is
easiest with a small number of isolated beads in each pillar. As
an alternate approach, fluorescent beads can also be glued to
the pillar tops by hand with PDMS after the μTUGs are
fabricated.
29. Dissolution of nickel into cell culture media was measured and
the levels does not show any negative effects on biocompatibil-
ity of the tissues [12].
30. The duration of Pluronic® treatment varies between 1–30 min
depending on the contractility of the cells and the concentra-
tion of ECM. Shorter times can be used when working with
more contractile cells, such as fibroblasts, which start compact-
ing the matrix within a few minutes after polymerization. In
contrast, longer duration Pluronic® treatments are needed for
low contractility cell types, such as smooth muscle cells, that
compact tissues over the course of up to 48 h. For example, it
takes 1–2 min of treatment for 3T3 fibroblasts in 1 mg/mL of
collagen, while 7–8 min treatment is needed for the same cells
in 2.5 mg/mL of collagen. Note that the microtissues may slip
off the pillar caps if the treatment is excessive.
31. Although passivation of the devices and seeding of the cells are
presented as separate methods, we suggest combining both
procedures in order to shorten the time. Typically, we start
with sterilizing the devices for 10–15 min. During this time, we
trypsinize and count the cells, and aliquot the number of cells
needed for the experiment. Then, we treat the devices with
Pluronic®, and during the incubation time we make the colla-
gen solution.
32. Cells may require different dissociation reagents. We use Accu-
tase for SMCs and cardiac fibroblasts, TrypLE for cardiomyo-
cytes, and Trypsin for fibroblasts.
33. In this step, one needs to define the total number of cells
needed for the experiments. For fibroblasts, 1 mL of
300,000 cells/mL is enough for one μTUG array in a P35
 Microtissues to Study Cell-Matrix Interactions 325

dish. For smooth muscle cells, one may require 1 mL of

400,000–500,000 cells/mL. For cardiomyocytes, one may
require 1 mL of 106/mL. For devices in 12-well or 24- well
plates, 250 μL of the cells with the same concentrations can
be used.
34. Solubilized collagen can be buffered in different ways. We use
the formula published by Saltzmann [21]. We use an Excel
spreadsheet to facilitate computation of reagent quantities.
Typical concentrations are 2–3 mg/mL (collagen) and
0.5–1 mg/mL of fibrin or fibrinogen. Tissues with low con-
centrations of collagen are more easily torn apart around the
edges of the pillars during microtissue formation. Fibrin and
fibrinogen are not necessary for fibroblasts, but are typically
used for smooth muscle or cardiac microtissues.
35. The collagen mixture sometimes freezes during degassing.
One can add a small amount of additional collagen mixture
to help it defrost quickly.
36. Cell seeding concentrations vary considerably, depending on
the application. Examples include 4.1 million/mL for 3T3
fibroblasts [12–14] and for smooth muscle cells [15], and
1 million/mL for wound healing assays [16]. Further, one
can adjust the number of cells seeded into each microwell by
the following equation: Total number of cells per μTUG ¼
Cell number per microwell
surface area of microwell
surface area of μTUG
37. Dropping the cell-ECM solution at the edges and also the
center. Hold the p1000 pipette perpendicular to the surface
of the well, and pipette up and down from dish edges to the
dish center to let the cells evenly distributed.
38. Since the ECM solution forms a gel if it warms up, setting a
relatively low temperature (9  C) for the centrifuge will
ensure the ECM solution stays sufficiently cold during
centrifugation.
39. Do not aspirate everything at once. Allow the remainder of the
collagen/cell mixture to run to the bottom of the device while
aspirating. The goal is to avoid having a layer of the mixture
polymerize in between the μTUG microwells.
40. This step allows the cell-ECM solution be slightly lifted to a
position around the heads of the pillars. The spinning time
needs to be optimized so that the solution does not escape
from the microwells. In some cases it may be helpful to do this
step a second time after rotating the device 90 .
41. Check for polymerization of the ECM by putting the tube with
the remaining ECM inside the same incubator. If the ECM
inside the tube is not fully polymerized, allow the μTUGs
incubate for a longer time.
326 Prasenjit Bose et al.

42. We culture microtissues at the same CO2 concentration as used

when the cells are in ordinary 2D culture. For the cells men-
tioned here, all are cultured at 5% CO2, except for bovine
pulmonary artery SMCs, for which we use 10% CO2 [15].
43. For other cell types, such as SMCs, we find that the media can
be changed every other day.
44. While recording time-lapsed images or videos, make sure the
P35 dish or plate containing the microtissues sits firmly in its
holder on the microscope stage. Otherwise, the devices may
shift during the time-lapse measurements, which makes it nec-
essary to register the images (e.g., by aligning successive
images based on the position of static feature) in order to
track the pillars’ deflections accurately.
45. As it is impossible to position the devices identically on the
microscope stage each time images are recorded, including
static features on the devices in the images to provide reference
points to measure the pillar deflections is essential.
46. For such longer term time-lapse studies, we generally do not
monitor the initial formation of the microtissues (steps 1 and 3).
47. We use a commercial relay solenoid. The coil has a resistance
R ~ 18 Ω, and the coil þ core have a self-inductance
L ~ 160 mH. The micromanipulator should have travel of
~2 cm in each direction. We use a manual unit, but an auto-
mated, joystick-driven manipulator could be convenient.
48. In our system, the core is 7.94 mm (5/1600 ) in diameter, and
extends 70 mm out of the solenoid. It tapers to a ~75 μm
diameter tip, which serves as a magnetic pole to concentrate
the local magnetic field near an individual μTUG. The pole tip
projects into the culture media at an angle of ~35 relative to
the plane of the μTUG. The solenoid/core assembly is
attached to the micromanipulator via an aluminum mounting
bracket. This bracket has a U-shaped extension, and each arm
of the U has a hole that fits the core snugly and holds it with set
screws. The solenoid sits on the core between the arms of
the U, and is encased in an aluminum block to provide
heatsinking.
49. The magnetic field depends linearly on the current in the
solenoid, and falls off at least as 1/r2, where r is the distance
from the sphere to the pole tip. This strong localization of the
field ensures that only one μTUG is actuated at a time.
50. The spacing between the magnetic pillar and the tweezer tip
should be more than 50 μm, because, if the magnetic sphere
gets too close to the tweezer tip, then the pillar can be pulled all
the way to the microwell wall, which can damage the
microtissue.
 Microtissues to Study Cell-Matrix Interactions 327

51. To determine the magnetization curves of single nickel sphere,

encase in 0.1 mL of epoxy (e.g., Araldite 502) and measure at
room temperature with a vector vibrating sample magnetome-
ter (VSM) (DMS Model 10; ADE Technologies, Westwood,
MA). For nickel sphere of 100 μm diameter, a field
B ¼ 100 mT yields a magnetic moment of ~0.15 μA m2,
resulting in a magnetic force of ~40 μN [12, 13].
52. The tweezer’s magnetic field can be calibrated with a magnetic
field sensor. We use a magnetic tunnel junction (MTJ) sensor
with the structure CoFeB/MgO/CoFeB, and with an active
circular area 7 μm in diameter, defined by photolithography
and ion beam etching [12, 13].
53. We typically use strain rates of 0.02% per s.
54. It is frequently useful to separate the cell and ECM contribu-
tions to the microtissues’ stiffness and/or dynamic response.
To do so, after the initial measurements of the as-grown tis-
sues, treat with 0.1% Triton-X solution for 10 min to lyse the
cells and remeasure [12, 15].
55. Procedures for calibrating pillar spring constants are given in
Ref. [10]. For large deflections, effective spring constants may
be modeled using finite element software [14].
56. The texture correlation approach is preferable, as nonunifor-
mities in the microtissues’ structure near the pillars can lead to
nonuniform strains near the microtissues’ ends.
57. Igor procedures to measure the pillar displacements are avail-
able from the authors upon request.
58. If the μTUG arrays are lined up so that the motion of the pillars
is only in the x-direction, averaging the pixel intensity values
along the y-direction improves signal-to-noise for tracking
features, such as the pillar edges, that are extended in the y-
direction.

Acknowledgements

This work was supported by NSF Grants CMMI-1463011 (JHU)

and CMMI-1462710 (BU). C.Y.H. acknowledges support from the
Ministry of Science of Technology of Taiwan’s Postdoctoral
Research Abroad Program grant number 105-2917-I-564-003-A1.

References
1. Chen CS, Tan J, Tien J (2004) Mechanotrans- 2. Hynes RO (2002) Integrins: bidirectional,
duction at cell-matrix and cell-cell contacts. allosteric signaling machines. Cell
Annu Rev Biomed Eng 6:275–302 110:673–687
328 Prasenjit Bose et al.

3. Schwartz MA, Ginsberg MH (2002) Networks study collective 3D cell mechanics in long-
and crosstalk: integrin signalling spreads. Nat term cultures (invited). J Appl Phys
Cell Biol 4:E65–E68 115:172616
4. Ciobanasu C, Faivre B, Le Clainche C (2013) 14. Zhao R, Chen CS, Reich DH (2014) Force-
Integrating actin dynamics, mechanotransduc- driven evolution of mesoscale structure in engi-
tion and integrin activation: the multiple func- neered 3D microtissues and the modulation of
tions of actin binding proteins in focal tissue stiffening. Biomaterials 35:5056–5064
adhesions. Eur J Cell Biol 92:339–348 15. Liu AS, Wang H, Copeland CR, Chen CS,
5. Schiller HB, Fassler R (2013) Mechanosensi- Shenoy VB, Reich DH (2016) Matrix visco-
tivity and compositional dynamics of cell- plasticity and its shielding by active mechanics
matrix adhesions. EMBO Rep 14:509–519 in a microtissue model: in situ experiments and
6. Kuo JC (2013) Mechanotransduction at focal mathematical modeling. Sci Rep 6:33919
adhesions: integrating cytoskeletal mechanics 16. Sakar MS, Eyckmans J, Pieters R, Eberli D,
in migrating cells. J Cell Mol Med 17:704–712 Nelson BJ, Chen CS (2016) Cellular forces
7. Wozniak MA, Chen CS (2009) Mechanotrans- and matrix assembly coordinate fibrous tissue
duction in development: a growing role for repair. Nat Commun 7:11036
contractility. Nat Rev Mol Cell Biol 10:34–43 17. Xu F, Zhao R, Liu AS, Metz T, Shi Y, Bose P,
8. Schmeichel KL, Bissell MJ (2003) Modeling Reich DH (2015) A microfabricated magnetic
tissue-specific signaling and organ function in actuation device for mechanical conditioning
three dimensions. J Cell Sci 116:2377–2388 of arrays of 3D microtissues. Lab Chip
9. Shamir ER, Ewald AJ (2014) Three- 15:2496–2503
dimensional organotypic culture: experimental 18. Zhao R, Simmons CA (2012) An improved
models of mammalian biology and disease. Nat texture correlation algorithm to measure sub-
Rev Mol Cell Biol 15:647–664 strate–cytoskeletal network strain transfer
10. Legant WR, Pathak A, Yang MT, Deshpande under large compressive strain. J Biomech
VS, McMeeking RM, Chen CS (2009) Micro- 45:76–82
fabricated tissue gauges to measure and manip- 19. Sage D, Neumann FR, Hediger F, Gasser SM,
ulate forces from 3D microtissues. Proc Natl Unser M (2005) Automatic tracking of indi-
Acad Sci U S A 106:10097–10102 vidual fluorescence particles: application to the
11. Legant WR, Chen CS, Vogel V (2012) Force- study of chromosome dynamics. IEEE Trans
induced fibronectin assembly and matrix remo- Image Process 14:1372–1383
deling in a 3D microtissue model of tissue 20. Crocker JC, Grier DG (1996) Methods of dig-
morphogenesis. Integr Biol 4:1164–1174 ital video microscopy for colloidal studies. J
12. Zhao R, Boudou T, Wang WG, Chen CS, Colloid Interface Sci 179:298–310
Reich DH (2013) Decoupling cell and matrix 21. Saltzman WM, Parkhurst MR, Parsons-
mechanics in engineered microtissues using Wingerter P, Zhu WH (1992) Three-
magnetically actuated microcantilevers. Adv dimensional cell cultures mimic tissues. Ann
Mater 25:1699–1705 N Y Acad Sci 665:259–273
13. Zhao R, Boudou T, Wang WG, Chen CS,
Reich DH (2014) Magnetic approaches to
 Part IV

Computational Approaches in Surfaceome Studies

 Chapter 19

Discovery of Surface Target Proteins Linking Drugs,

Molecular Markers, Gene Regulation, Protein Networks,
and Disease by Using a Web-Based Platform Targets-search
Bin Yan, Panwen Wang, Junwen Wang, and Kenneth R. Boheler

Abstract
Integration and analysis of high content omics data have been critical to the investigation of molecule
interactions (e.g., DNA–protein, protein–protein, chemical–protein) in biological systems. Human pro-
teomic strategies that provide enriched information on cell surface proteins can be utilized for repurposing
of drug targets and discovery of disease biomarkers. Although several published resources have proved
useful to the analysis of these interactions, our newly developed web-based platform Targets-search has the
capability of integrating multiple types of omics data to unravel their association with diverse molecule
interactions and disease. Here, we describe how to use Targets-search, for the integrated and systemic
exploitation of surface proteins to identify potential drug targets, which can further be used to analyze gene
regulation, protein networks, and possible biomarkers for diseases and cancers. To illustrate this process, we
have taken data from Ewing’s sarcoma to identify surface proteins differentially expressed in Ewing’s
sarcoma cells. These surface proteins were then analyzed to determine which ones were known drug
targets. The information suggested putative targets for drug repurposing and subsequent analyses illu-
strated their regulation by the transcription factor EWSR1.

Key words Surface protein, Web-based platform, Targets-search, Drug target, Molecular marker

1 Introduction

Advances in high-throughput technology applications to biomedi-

cal research and biomedicine generate a vast amount of multidi-
mensional omics data. Cell surface proteins are involved in major
cellular functions involving transport, enzymatic activity, signal
transduction, intercellular joining, attachments to the cytoskeleton
and extracellular matrix, and for cell–cell recognition. Identification
of surface proteins on a particular cell type also facilitates the ability
of a researcher to dissect biological complexity and to predict
protein–protein interactions, drug–target networks, cell signaling
cross talk, and other pathways involved in a cell’s phenotype during

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2_19, © Springer Science+Business Media, LLC 2018

331
332 Bin Yan et al.

development or disease. Analysis of proteomics datasets with the

Cell Surface Protein Atlas (CSPA) [1] enables the rapid identifica-
tion of selected cell surface proteins that are potential targets for
diagnostic and therapeutic interventions; however, when compared
with other omics datasets, the CSPA likely underestimates the total
number of surface proteins that need to be evaluated. The CSPA
contains primarily data from N-glycosylated proteins (glyco-CSC,
Cyc-glyco-CSC) and from some extracellularly exposed and con-
formationally available lysines (Lys-CSC). As not all proteins con-
tain these sequences or will be identified using mass spectrometry,
the surfaceome is currently underestimated in the CSPA. Transcrip-
tomic analyses can also be used in the identification of cell surface
proteins, but predictive approaches often rely on algorithms as
opposed to experimentally validated data. Some targets therefore
may be misclassified, as the proteins may be membrane-associated
(e.g., nuclear, mitochondrial, endoplasmic reticulum, other) but
not necessarily part of the surfaceome.
To facilitate surfaceome analyses and specifically to identify
putative drug targets on the surface of cells, we have developed a
web-based platform called Targets-search. This platform is part of an
integrative resource that directly links cell surface proteins identi-
fied either experimentally by Cell Surface Capture (CSC)-
Technology or predicted from bioinformatics, proteomic or tran-
scriptomic analyses [2, 3]. This web-based platform collects pub-
lished and otherwise publicly available data and is designed to serve
as a comprehensive resource center for integrated and systemic
exploitation of multimolecular interactions among drug targets,
gene/protein networks, human diseases, and cancer. Although
databases already exist that focus on one or two interactions, such
as drug–target–disease, transcription factor (TF)–targets, or pro-
tein–protein interactions, these databases generally lack tools capa-
ble of integrating at least three types of information. Comparatively,
our search platform, Targets-search, can assemble five resource types
or ten databases to detect multimolecular interactions with drugs
and chemicals, cell surface proteins, genes, and diseases.
Here, we describe how to use the web-based Targets-search for
extracting and integrating surface genes/proteins, drugs, TFs, pro-
tein networks and cancer biomarkers associated with human dis-
eases with an emphasis on cell surface proteins. To demonstrate the
usefulness of this platform, we have employed two datasets to
identify surface proteins and putative drug targets, as well as
gene/protein and disease linkage. After searching against the multi-
resource datasets, the web-based platform returns a set of output
tables showing all overlaps with these interactions [3].
 Targets-Search for Surface Proteins 333

2 Materials

2.1 Website 1. Targets-search is available at http://www.jjwanglab.org/targets-

and Datasets Analyzed search and will be updated frequently (see Note 1). This web-
based platform is comprised of five types of publicly available
databases or resources: cell surface protein, drug/chemical, dis-
ease/cancer, TF binding targets, and PPIs (Fig. 1).
2. To illustrate the use of Targets-search on normal cells, we have
utilized a publicly available transcriptome-defined catalog of
3702 human genes believed to encode transmembrane proteins
located at the surface of human cells and published by Cunha
et al. [2]. This data set does not include any cell surface proteins
that are anchored or associated with the external side of the
plasma membrane by mechanisms that do not involve trans-
membrane domains (see Note 2).
3. To illustrate the use of Targets-search on diseased cells, we ana-
lyze RNA-seq data to study genes encoding cell surface proteins
that are differentially overexpressed and underexpressed in
Ewing’s sarcoma cells (see Note 2).
4. Excel or similar spreadsheets for data import.

input Genes or
proteins

Search

Database
Surface proteins: Drug-target-disease: Disease and Cancer: Gene regulation: Protein networks
Drugbank OMIM TFBS: transcription
Human surfaceome PPI: protein
TTD Biomarkers factor target genes
Cell surface protein protein
ChemBL Driver genes containing binding
atlas (CSPA) interactions
PharmGKB sites

Interface
Interaction tables Interaction tables with Interaction tables Interaction Interaction
with cell surface drug/chemicals and with cancer tables with tables with
proteins in different targeted pathways detection, diagnosis transcriptional protein
cell types and diseases and prognosis gene regulation networks
Output

Integrations

Fig. 1 Workflow of the web-based Targets-search platform. The platform can integrate information from cell
surface proteins, drug-target, transcriptional regulation, protein networks, and human disease/cancer for data
analysis and interpretation
334 Bin Yan et al.

3 Methods

Targets-search is an open web application and provides user-friendly

access to all existing databases and supports optional queries. The
five resources available in Targets-search include (1) high resolution
human and mouse CSPA information containing CSC-datasets
from 41 human and 31 mouse cell types, including normal and
cancer cells [1] (see Note 3) and a bioinformatics-identified catalog
of 3702 transmembrane proteins located at the surface of human
cells [2] (surfaceome, see Note 4); (2) drug-target-disease
resources derived from the DrugBank database [4, 5] (see Note
5), the Therapeutic Target Database [6, 7] (TTD, see Note 6), the
ChemBL database of bioactive molecules with drug-like properties
and compound bioactivity data against drug targets [8] (see Note
7), and the Pharmacogenetics and Pharmacogenomics Knowledge
Base [9] (PharmGKB, see Note 8); (3) Disease resources derived
from the OMIM (http://www.omim.org) catalog of human genes
and genetic disorders and maintained by NCBI, Curated known
cancer biomarkers (Cancer biomarker), and DriverDB (Cancer
driver, see Note 9) database of cancer driver genes identified by
using data from dbSNP and The Cancer Genome Atlas [10];
(4) Transcription factor (TF) binding target genes (TFBS)
resources based on bioinformatics-predicted conserved TF binding
sites on the gene promoters of human or mouse (see Note 10); and
(5) Protein-protein interactions (PPI) of human and mouse,
derived from public data collected by our previously published
methods [11] (see Note 11). These ten datasets have been stored
in the main user interface to the platform.
To illustrate the utility of this procedure, we have utilized data
from Cunha et al. [2], who identified transmembrane proteins, and
data from Town et al., who evaluated the surfaceome of Ewing’s
sarcoma. In the case of the latter, cell surface proteins are excellent
molecules for targeted therapeutics due to their accessibility.
Ewing’s sarcoma (EWS) is bone or soft tissue sarcomas, and it is
the second most common bone tumor of young adults and adoles-
cents with the mean age 15 years of diagnosis [12]. The major cause
of death is due to limited treatment options and poor prognosis
after EWS diagnosis. The discovery of novel therapeutic targets in
EWS cancer would improve current strategies in clinical need.
However, there are few cancer-specific cell surface proteins that
can be invoked for targeting EWS. We used Targets-search to ana-
lyze and identify these proteins with potentiality of drug targets and
biomarkers.
Targets-search, thus, provides a search engine that allows
optional methods for querying a list of genes/proteins from tran-
scriptomic or proteomic data through selection of the query data-
bases. This flexibility allows the researcher to search for possible
 Targets-Search for Surface Proteins 335

Fig. 2 Input and output interface of the web-based Targets-search platform

molecular interactions or relationships among input genes/pro-

teins (Fig. 2). The output display is in tabular form and results in
multi-useful windows. Each window represents an output derived
from its corresponding built-in database. With “export,” the full
output table can be downloaded in text format. The following
protocol outlines the main procedures for using Targets-search,
and explanations are provided on how to retrieve and download
the results for further investigation.

3.1 General 1. Prepare data to be examined in the form of gene symbols or

Procedure of Usage Protein identifiers. Acceptable formats of gene names or pro-
tein identifiers can be obtained or converted using DAVID or
Biomart (see Note 12).
2. On the Search options page, select the species to be examined
(Human or Mouse).
3. Select the input format, either Gene Symbol or UniProt AC.
4. Directly copy the list of input data (step 1 above) into the
Input Window entitled Genes/Proteins. Targets-search accepts
from one to many genes or proteins. In its current deployment
and under typical server loading conditions, Targets-search does
not have any limit to the number of input genes, but in general
it can process 500–5000 genes or proteins within ~5–20 min.
336 Bin Yan et al.

5. To identify experimentally validated surface proteins from

input genes/proteins, check the box listed “CSPA” on the
Input Window, and query by pressing “search”.
6. Query of the transcriptome-defined 3702 transmembrane pro-
teins identified by Cunha et al. [2] (see Note 4) leads to the
identification of 799 transmembrane proteins that are unequiv-
ocally identified as surface proteins. The cell type in which the
surface protein was identified in the CSPA is shown in the
output list.
7. Data can be transferred in .txt format by hitting “export”,
which is located at the top left region of the Output Windows.
This can be imported into excel or other spreadsheets as
needed.
8. To identify putative drug targets for all putative transmem-
brane surface proteins identified in [2], check one drug-target-
disease datasets (e.g., Drugbank).
9. The output from the Drugbank search identifies 6516 entries
of surface proteins with FDA-approved or putative drug targets
in the Drug Bank. Each gene/UniProtID will be listed in the
output together with the Targets-Name (surface protein), the
Drug (DB_drug), and the Status (DB_Status) of the drug for
therapeutic applications (Approved or Experimental).
10. Data can be transferred to excel in .txt format by hitting
“export”.
11. To look at the individual databases, repeat steps 7–9 with the
other drug datasets (TTD, PharmGKB, and ChemBL), by
sequentially checking each of the boxes individually (and
removing the check from previously checked database e.g.,
Drugbank).
12. Alternatively, all four drug databases can be searched simulta-
neously, by concurrently checking “Drugbank”, “TTD”,
“PharmGKB” and “ChemBL”. After hitting “search,” the
data can be looked at individually using the tabs located at
the top of the pages (Drugbank, TTD, pharmGKB, and
ChemBL).
13. To examine only the CSPA authenticated surface transmem-
brane proteins, remove the Gene list input from [2] in the
Genes/Proteins input box. This can be done by highlighting
all the data and hitting “delete” on the keyboard.
14. Go to the data exported in step 6, and copy and paste the gene
name/UniProtID into the Genes/Proteins input box of Tar-
gets-search.
15. Check the Drugbank, TTD, PharmGKB, and ChemBL boxes
and hit “Search”.
 Targets-Search for Surface Proteins 337

16. Examine the output for potential/putative drug targets. For

Drugbank, this will correspond to 1884 entries; for TTD,
2977 entries; for PharmGKB, 78 entries; and for ChemBL,
193 entries (see Note 13).
17. Data can be transferred to excel in .txt format by hitting
“export”.
18. For simplification purposes, subsequent analyses will be per-
formed on the 78 entries identified from the PharmGKB tar-
gets identified in the list of 799 proteins identified from the
CSPA analysis.
19. Of the 78 drug targets identified in the CSPA-PharmGKB
analysis, only 17 protein targets were identified: ABCB1,
ADRB2, BCHE, EGFR, ERBB2, FGFR1, FGFR3, FLT4,
HLA-A, HLA-B, HLA-DQA1, HLA-DRB1, IL2RA, KIT,
LDLR, MS4A1, and TNFRSF8. Copy and paste this list into
the Genes/Proteins input box.
20. To identify linkage or association with human genetic disease
or cancer, check the boxes “Cancer biomarker”, “Cancer
driver,” and “OMIM.”
21. Hit “Search”. The outputs will show that BCHE, EGFR,
ERBB2, FGFR3, IL2RA and KIT are known cancer biomar-
kers. ABCB1, BCHE, EGFR, ERBB2 and KIT are known
cancer drivers; and ABCB1, ADRB2, BCHE, EGFR, ERBB2,
FGFR1 and FGFR3 are associated with OMIM disorders.
22. To identify possible TF binding sites that may regulate the
expression of these disease associated genes, check “TFBS.”
As an example, only OMIM associated disorders were input
into the Genes/Proteins input box. From this, 601 entries of
TFs that may regulate these seven genes were identified. The
genomic locations of these binding sites are indicated, and a
score showing the probability of consensus sequence binding is
shown.
23. To identify protein-protein interactions, remove the check
from “TFBS” and check “PPI.”
24. Hit “Search.” 842 entries will be observed. As an example of
the output, ABCB1 will be shown to possibly interact with
ABCB6, ABCB7, and 29 other proteins.
25. In summary, after submission of input data, Targets-search
produces a set of output tables showing the results detected
or identified overlaps between the input genes/proteins and
those from the selected databases. Users can choose one or all
of the output tables and retrieve the result. The outputs are
downloadable in text format by pressing the “export” and
performing subsequent analyses.
338 Bin Yan et al.

RNA-seq experiments of Ewing sarcoma (EWS) vs.

Mesenchymal stem cells (MSCs) using human surfaceome

185 up-regulated 316 down-regulated

genes in EWS genes in EWS
• 80 located at Targets-search • 128 located at
plasma membrane plasma membrane
or cell surface or cell surface
• 79 in human cell • 99 in human cell
surfaceome, of 19 surfaceome, of 23
were validated at were validated at
tumor samples tumor samples
• 59 in human CSPA • 137 in human
CSPA

Fig. 3 An example study of Targets-search using surfaceome data of Ewing sarcoma (EWS) vs. Mesenchymal
stem cells (MSCs) from [13]. The results demonstrate that differentially expressed surface proteins not only
have been targeted by approved or testing drugs, acted as cancer biomarkers, but also are regulated by
EWS-related TFs

3.2 Example Study: 1. The first step is to prepare input data (see Fig. 3).
Human Cancer 2. Using published RNA-seq data [13], we identified two input
datasets: 185 genes upregulated in EWS and 316 genes upre-
gulated in MSCs (or downregulated in EWS) with >2.0-fold
changes in abundance and an adjusted p-value below 0.05.
3. Before submitting the input data, we analyzed biological pro-
cesses of the two input gene sets using DAVID. The results
show that 37 of the 185 upregulated and 80 of the 316 down-
regulated genes involve cell surface receptor linked signal trans-
duction events (see Note 14).
4. The second step is to identify how many authenticated surface
proteins are among the 185 upregulated genes in EWS. Copy
and paste these genes to the Input Window, check the box
listed “CSPA” on the Input Window, and search by pressing
“Search.”
5. The Output Window indicates that 59 proteins are detected as
human CSC-surface proteins.
6. Similar to the step above, there were 137 surface proteins
identified from 316 downregulated genes in EWS. These out-
put lists can be exported as .txt files.
 Targets-Search for Surface Proteins 339

7. The third step is to search for human cell surfaceome dataset

(see Note 4). The two input lists were separately submitted to
the Input Window. In the Output Window, we find 79 and
99 genes/proteins that belong to transmembrane surface pro-
teins from the 185 upregulated and the 316 downregulated
genes in EWS, respectively. These output data can be saved in .
txt format by clicking “export.”
8. In the fourth step, based on the two input gene lists, we can
analyze possible interactions with drugs-targets-diseases. After
submitting the 185 upregulated genes, and checking boxes
“Drugbank” and “TTD,” 23 genes in the outputs are reported
as targets of 55 approved drugs listed in the two databases.
These drug targets also correspond to 47 (in Drugbank) and
23 (in TTD) diseases and disorders, respectively. Of note,
14 surface proteins (ALK, ARSA, CA11, CXCR4, DDR1,
FGFR4, FLT4, GRIK5, GRM2, ITGB2, NRP1, PTPRS,
SLC1A4, and SLC7A8) are located at the plasma membrane
or cell surface.
9. By submitting the 316 downregulated input genes, we detect
59 genes as targets of 244 approved drugs. These target
216 (in Drugbank) and 103 (in TTD) diseases and disorders,
respectively. We also determined that 38 of the 58 genes can
encode surface proteins, including seven (PTGS1, GABRE,
EGFR, PDGFRA, SERPINE1, PLAUR, and TUBB) that can
be targeted by at least five approved drugs individually.
10. The fifth step is to search for Cancer driver and biomarker from
the two inputs. Select two boxes “Cancer driver” and “Cancer
biomarker”.
11. After submission, the output interface shows 8 and 16 genes
from the input upregulated and downregulated genes that
were detected as cancer driver genes, respectively. In the upre-
gulated output, LRP1B is predicted as a driver of bladder, head
and neck cancer, and rectal adenocarcinoma. Surface protein
OBSCN is predicted to drive breast cancer and glioblastoma
multiforme. Among the downregulated genes, surface protein
EGFR was detected as a driver of lung cancer, colon cancer or
glioblastoma multiforme. We also detected 14 genes upregu-
lated in EWS as early diagnosis biomarkers, of eight genes
(ALK, CADM1, FGFR4, FXYD6, IGFBP2, PCDH17,
PTPRS, and CXADR) are surface proteins.
12. The sixth step is to identify surface genes in the two input lists
that can be regulated by EWS-related TFs EWSR1, NF-κB, and
STAT3. The database box “TFBS” was selected.
13. After running “Search”, the output includes target genes of all
human TFs that contain conserved TF binding sites on pro-
moters of the input genes.
340 Bin Yan et al.

14. From the output above, we can select the target genes that only
are regulated by the three TFs. The result showed that EWSR1
can bind to 25 promoters of upregulated genes in EWS. Of
these, 12 are surface proteins, and NRP1, CA11, and DDR1
are targets of approved drugs. Furthermore, seven EWSR1
target genes are cotargeted by NF-κB and STAT3 both of
which have been implicated in EWSR1 regulation
[14, 15]. Of note, the surface protein leucine-rich repeat and
Ig domain protein 1 (LINGO1), expressed in over 90% of
EWS, was found to carry antibody conjugated with drugs to
kill Ewing sarcoma cells [13].
15. To link the input genes to protein network, the seventh step is
to check the database box “PPI.” After submission of the two
input lists, the outputs provide a list of all proteins that can be
connected to the input genes. As a result, the derived gene–-
protein interactions can establish protein networks linking the
upregulated genes in EWS and protein modifications.
16. From the outputs generated using the above procedures, Fig. 3
displays a multi-dimensional picture showing how the surface
proteins/genes differentially expressed in EWS link with drug
target, transcriptional gene regulation, protein networks, and
diseases/cancer. The main upregulated (FGFR4, PTPRS, and
ALK) and downregulated (EGFR, SERPINE1, etc.) genes/
proteins in EWS cells are targeted by approved drugs and were
reported as cancer biomarkers. EWSR1 was predicted to bind
to the upregulated NRP1, CA11, and DDR1 that are also
targeted by drugs. This result demonstrates the surface pro-
teins/genes identified by high throughput sequencing in EWS
could be potential targets for drug discovery and reveals their
regulatory components.

4 Notes

1. This database may migrate to a new server in the near future,

but a link will be built to permit direct access.
2. Town et al. (2016) tried to identify antigens that are uniquely
expressed on EWS tumor cells compared with Mesenchymal
stem cells (MSCs) by RNA-seq experiments of human genes
encoding cell surface proteins (the surfaceome). Based on the
published RNA-seq data [13], we identified 185 and 316 upre-
gulated genes with above 2.0-fold change of expression and
adjusted p-value below 0.05 in EWS cells and MSCs
(or downregulated in EWS cells), respectively.
3. Bausch-Fluck et al. (2015) applied the Cell Surface Capture
(CSC) technology to 41 human and 31 mouse cell types to
 Targets-Search for Surface Proteins 341

generate a mass-spectrometry derived Cell Surface Protein

Atlas (CSPA) [1]. This protein atlas provides a broad view of
cellular surfaceome at high resolution, and covers different cell
types, including normal and cancer cells. It includes a com-
bined dataset of 1492 human and 1296 mouse cell surface
glycoproteins, based on experimental evidence for their cell
surface expression on various cell types in human and human.
4. Using bioinformatics tools, Cunha et al. generated a catalog of
3702 transmembrane proteins located at the surface of human
cells (human cell surfaceome) [2]. This data integrated expres-
sion information from a variety of surfaceome genes among
normal tissues and/or differential expression in tumors, impor-
tant characteristics for putative tumor targets. In addition, a
subset of 493 genes was validated by using quantitative real-
time PCR in tumor samples. Thus, this database is able to help
further studies aimed at characterizing tumor targets at the
surface of human cells.
5. The DrugBank database (http://www.drugbank.ca) is a com-
prehensive resource that coordinates both data of detailed drug
(chemical, pharmacological and pharmaceutical) and their tar-
gets (sequence, structure, and pathway) [4, 5]. This database
contains over 7740 drug entries including 1584
FDA-approved (small molecule and biotech) drugs and over
6000 experimental drugs, as well about 1500 targeted path-
ways and diseases.
6. Therapeutic Target Database (TTD) (http://bidd.nus.edu.sg/
group/cjttd/) is a database to collect the known and explored
therapeutic protein and nucleic acid targets, the targeted dis-
ease and the corresponding drugs directed at each of these
targets [6, 7]. This database contains over ~1500 approved,
~1530 clinical and ~12,000 experimental drugs, as well as
~1400 targeted diseases or disorders.
7. ChemBL (https://www.ebi.ac.uk/chembldb/), maintained by
the European Bioinformatics Institute, is a chemical database
of bioactive molecules with drug-like properties and contains
compound bioactivity data against drug targets [8]. We
extracted available data including ~1600 compounds used for
drugs and 460 associated targets.
8. Pharmacogenetics and Pharmacogenomics Knowledge Base
(PharmGKB, http://www.pharmgkb.org) is an integrated
tool for investigating how genetic variation effects drug
response [9]. The database includes drugs, genetic variations,
and clinical data integrated with literature, pathway and proto-
col information. We extracted well-known pharmacogenomic
associations, including 165 FDA or EMA approved drugs and
118 targets.
342 Bin Yan et al.

9. “Cancer driver” gene is defined as a gene whose dysfunction

will cause tumorigenesis. There are many methods to identify
driver genes based on the mutation frequency of an individual
gene compared with the background mutation rate. DriverDB
(http://ngs.ym.edu.tw/driverdb/) is a database which
includes exome-seq data and genetic variation data from
dbSNP and The Cancer Genome Atlas [10]. This database
provides cancer driver genes identified by using these published
bioinformatics algorithms, TCGA mutations and CNV infor-
mation. We extracted a total of 390 driver genes responsible for
15 types of tumor based on at least three bioinformatics
method identification.
10. We employed PWMSCAN to predict TF binding sites
[16]. This method performs computational identification of
the binding sites by scanning promoter sequences of whole
human and mouse genomes using TF binding motifs. The
predicted binding site is then evaluated by the p value through
a permutation-based method, FastPval [17]. Also, the pre-
dicted binding sites can be further filtered based on conserva-
tion scores between human and mouse genomes. Here, we
scanned promoters of human and mouse from 2000 bp
upstream to 500 bp downstream transcriptional start sites
which were downloaded from the UCSC genome browser.
The TF binding target gene database is composed of 595 TFs
in human and 630 TFs in mouse.
11. Proteins that interact with other molecules are the backbone of
protein networks and cellular signaling interconnections.
Recently, we published a web-based server, PTHGRN, for
constructing posttranslational protein-gene regulatory net-
works using PPI and high-throughput Omics data [11]. To
help exploring the protein networks linking with proteins-user-
provided, we provide PPI of human and mouse, derived from
public databases BioGRID, STRING, Dip, HPRD, Intact,
Mint, and Reactome. Users can identify all proteins that can
physically interact with input proteins of interests such as drug
targets, TF targets, etc.
12. Gene or protein IDs can be converted using DAVID (https://
david.ncifcrf.gov/conversion.jsp) or Biomart (http://www.
biomart.org/biomart/martview/65c2ea6c079d1b85820fa5
bbf5af62b5).
13. The number of entries is greater than the number of authenti-
cated transmembrane proteins identified through the initial
CSPA evaluation. This is because of multiple drugs being
identified for the same gene target. For example, ABCB1 in
the PharmGKB-CSPA analysis has two drugs that should tar-
get this surface protein).
 Targets-Search for Surface Proteins 343

14. Using the bioinformatics tool DAVID, we can classify the input
genes to different function categories, gene ontology
(GO) and KEGG pathway. Among the 185 and 316 genes, of
80 and 128 among these genes were found to locate at plasma
membrane or cell surface respectively. By GO biological pro-
cess annotation, there were corresponding 37 and 80 genes
that involve cell surface receptor linked signal transduction.

Acknowledgment

This work was supported by General Research Fund (17100214,

17121414M) and the Theme-based Research Scheme (T13-706/
11) of Hong Kong Research Grant Council, and Seed Funding
Program for Basic Research of the University of Hong Kong
(201606159002, and RCGAS0768448734).

References
1. Bausch-Fluck D, Hofmann A, Bock T, Frei AP, 8. Papadatos G, Overington JP (2014) The
Cerciello F, Jacobs A, Moest H, Omasits U ChEMBL database: a taster for medicinal che-
et al (2015) A mass spectrometric-derived cell mists. Future Med Chem 6:361–364
surface protein atlas. PLoS One 10:e0121314 9. Thorn CF, Klein TE, Altman RB (2010) Phar-
2. da Cunha JP, Galante PA, de Souza JE, de macogenomics and bioinformatics:
Souza RF, Carvalho PM, Ohara DT, Moura PharmGKB. Pharmacogenomics 11:501–505
RP, Oba-Shinja SM et al (2009) Bioinformatics 10. Cheng WC, Chung IF, Chen CY, Sun HJ, Fen
construction of the human cell surfaceome. JJ, Tang WC, Chang TY, Wong TT, Wang HW
Proc Natl Acad Sci U S A 106:16752–16757 (2014) DriverDB: an exome sequencing data-
3. Yan B, Guan D, Poon NYE, Li MJ, Wang P, base for cancer driver gene identification.
Tse WL, Wu SCM, Sham PC, Xia Z, Gundry Nucleic Acids Res 42:D1048–D1054
RL, Wang J, Boheler KR (in preparation). Tar- 11. Guan D, Shao J, Zhao Z, Wang P, Qin J,
getCSP, a comprehensive resource can priori- Deng Y, Boheler KR, Wang J, Yan B (2014)
tize targeting cell surface proteins for exploring PTHGRN: unraveling post-translational hier-
multi-interactions among molecules, drugs, archical gene regulatory networks using PPI,
and diseases ChIP-seq and gene expression data. Nucleic
4. Wishart DS, Knox C, Guo AC, Shrivastava S, Acids Res 42:W130–W136
Hassanali M, Stothard P, Chang Z, Woolsey J 12. Burchill SA (2008) Molecular abnormalities in
(2006) DrugBank: a comprehensive resource Ewing’s sarcoma. Expert Rev Anticancer Ther
for in silico drug discovery and exploration. 8:1675–1687
Nucleic Acids Res 34:D668–D672 13. Town J, Pais H, Harrison S, Stead LF,
5. LawV, KnoxC, Djoumbou Y, JewisonT, Guo Bataille C, Bunjobpol W, Zhang J, Rabbitts
AC, LiuY, Maciejewski A, Arndt D et al (2014) TH (2016) Exploring the surfaceome of
DrugBank 4.0: shedding new light on drug Ewing sarcoma identifies a new and unique
metabolism. Nucleic Acids Res 42: therapeutic target. Proc Natl Acad Sci U S A
D1091–D1097 113:3603–3608
6. Chen X, Ji ZL, Chen YZ (2002) TTD: Thera- 14. Javelaud D, Besancon F (2001) NF-kappa B
peutic Target Database. Nucleic Acids Res activation results in rapid inactivation of JNK
30:412–415 in TNF alpha-treated Ewing sarcoma cells: a
7. Zhu F, Shi Z, Qin C, Tao L, Liu X, Xu F, mechanism for the anti-apoptotic effect of
Zhang L, Song Y et al (2012) Therapeutic NF-kappa B. Oncogene 20:4365–4372
target database update 2012: a resource for 15. Lai R, Navid F, Rodriguez-Galindo C, Liu T,
facilitating target-oriented drug discovery. Fuller CE, Ganti R, Dien J, Dalton J, Billups C,
Nucleic Acids Res 40:D1128–D1136 Khoury JD (2006) STAT3 is activated in a
344 Bin Yan et al.

subset of the Ewing sarcoma family of tumours. 17. Li MJ, Sham PC, Wang J (2010) FastPval: a fast
J Pathol 208:624–632 and memory efficient program to calculate very
16. Levy S, Hannenhalli S (2002) Identification of low P-values from empirical distribution. Bio-
transcription factor binding sites in the human informatics 26:2897–2899
genome sequence. Mamm Genome
13:510–514
 INDEX

A Cell Surface Protein Atlas (CSPA) ...................60, 61, 70,

71, 332, 334, 336–338, 341, 342
Adherent cells ............................65, 73, 94, 96, 109, 118, Cell transporters................................................................ 3
121–123, 201 Channelopathies................................................... 168, 171
Adhesion ................................... v, 21, 169, 171, 297, 305
Chemoproteomic strategies............................................ 59
Anesthesia ........................................................... 34, 37, 42 Chick embryo
Antibody ...................................................v, 281, 298, 340 avian extracellular proteome............................... 31–43
affinity capture................................................ 4, 6, 8, 9
avian membrane proteome ................................. 31–43
development ..............................59, 60, 62, 70, 73–75 chorioallantoic membrane ........................................ 32
Antigen .........................3–19, 22, 58, 94, 117, 132, 136, Cloning ................................................105–115, 168, 177
152, 196, 206, 243, 244, 339
Cluster of differentiation (CD) antibodies .................... 58
Applied forces....................................................... 305, 319 Co-immunoprecipitation ........................... 196, 208, 215,
Armed T cells........................................................ 117–124 216, 219, 220
Arrhythmia ..........................................171, 172, 178, 180
Collagen.....................................232, 262, 263, 292, 304,
308, 309, 313, 316, 317, 324, 325
B
CXC chemokine ligand 12 (CXCL12) ............... 152–163
Bacterial cells ..................................................7, 14, 21–28 CXC chemokine receptor 4 (CXCR4) ............... 151–163
Baculovirus system ............................................... 105–115 Cytokine secretion ..............................227, 233, 235, 245
Barcode ............................................................... 59, 60, 62
Bioinformatics ........................... 4, 12, 41, 43–54, 59, 62, D
238, 241, 332, 334, 341–343 Deglycosylation ........................................... 33, 35, 40, 59
Biotinylation ............................33, 34, 37–39, 41, 43, 62, Differential centrifugation ........................................93, 97
63, 73, 215, 220, 238
Differentiation................................v, 3, 58, 73, 128, 129,
Bispecific antibodies (BsAbs)............................... 117–124 134, 145, 186, 212–214, 261–294, 296–300
Bordetella pertussis....................................................... 3–19 Drug discovery .............................................................. 340
Drug repurposing .................................................... vi, 331
C
Drug target........................ 321–334, 336, 337, 339–342
Cancer...........................48, 49, 54, 91, 92, 99, 117–124, Dynabeads ....................................................5, 6, 9–11, 17
152, 171, 332–334, 337–342
Carbonate extraction .......................................4, 8, 14, 15 E
Cardiomyocytes (CM) ........................... 62, 71, 185–188, Ehlers–Danlos Syndrome (EDS).................................. 262
190, 191, 193, 194, 286, 299, 324, 325
Electrophysiology.................................. vi, 171, 175–177,
Cardiotoxicity................................................................ 185 185–188, 190, 191, 193, 194, 196, 226, 233, 245
Cation-dependent mannose 6-phoshate receptor Embryonic stem cells (ESCs) ..................... 185, 279, 297
(CD-MPR) ..................... 105, 106, 108, 110–113 Endothelium.....................................................31, 41, 196
Cell lineages..................................................................... 58 Engineered microtissues ............................................... 303
Cell-matrix interactions ........................ vi, 263, 303–314, Enriched membrane fractions (EMFs) ......................4, 76
316–327 Enzyme-linked immunosorbent assay (ELISA) .............vi,
Cell surface .................................. v, vi, 57–70, 72–74, 76,
152–156, 158–160, 162
79, 80, 105, 106, 118, 128, 129, 132, 136, 140, Epitope .......................................... 22, 23, 58–62, 70, 73,
147, 151–156, 158–163, 175, 180, 185, 214, 75, 129, 156, 161, 298
224, 228, 229, 233, 237–245, 253, 255, 259,
Epitope selection...................................59, 60, 62, 74, 75
263, 290, 293, 304, 331–334, 338, 339, 343 Exosomes............................................................vi, 91–100
Cell surface capture (CSC)-technology ..................57–76, Extracellular domain .................... 57–70, 73, 74, 76, 161
224, 228, 229, 237–244, 332, 340
Extracellular epitopes ................................................58–61

Kenneth R. Boheler and Rebekah L. Gundry (eds.), The Surfaceome: Methods and Protocols, Methods in Molecular Biology,
vol. 1722, https://doi.org/10.1007/978-1-4939-7553-2, © Springer Science+Business Media, LLC 2018

345
 THE SURFACEOME: METHODS AND PROTOCOLS
346 Index
Extracellular matrix (ECM)........................ 249, 251, 303 K
Extracellular matrix molecule-based capture...... 249–259
Extracellular secreted proteins.......................................... 3 Kegg identifiers ...........................................................4, 19
Extracellular vesicles (EVs).............................. 92–94, 100 Kinase................................. 131, 170, 173, 211–221, 238

F L

Fibroblast......................................... 62, 73, 92, 129, 212, Laminin..................... 142, 249, 251, 253, 255–257, 293
238, 239, 243, 262, 265, 270, 271, 273–277,
M
281, 292, 293, 304, 316, 317, 324, 325
Fibronectin .................................................................... 251 Magnetic actuation ....................................................... 315
Flow cytometry ......................... 22, 26, 61, 73, 118–120, Mass spectrometry (MS).......................... 4, 5, 11, 12, 22,
122, 124, 128, 129, 131–134, 136, 138–141, 32, 57–70, 73, 74, 76, 81, 86, 93, 95, 96, 98, 99,
143, 145, 147, 152, 153, 155, 156, 161, 249, 224, 241, 332, 341
252, 253, 255, 291, 299, 300 ESI-MS/MS ................................................................ 4
Fluid transport ......................vi, 224, 227, 235, 236, 245 HPLC-MS/MS ......................................................... 89
Fluorescence-activated cell sorting (FACS)................130, LC-MS/MS...................................68, 89, 93, 98, 241
131, 133, 138, 162, 228 nLC-MS/MS ................................................. 9, 11, 65
Förster resonance energy transfer (FRET) .................196, Mechanobiology ........................................................... 303
206, 208 Membrane enrichment .............................................62, 63
Membrane protein ...................... 4, 7, 10, 15, 22, 58–60,
G 73, 79, 80, 89, 180, 298
Gel-free systems ..........................................................4, 14 Membrane and matrix proteome ................................... 31
Glycopeptide capture ............................ 62, 64, 67–69, 76 Mesenchymal stromal cells(MSCs) .................... 255, 257,
Glycopeptide elution....................................62, 64, 67–69 BNF–259
Glycoprotein.................................. 62, 79, 113, 167, 237, Microfabrication............................................................ 303
241, 243, 341 Microtissues................................ 263, 303–314, 316–327
G protein-coupled receptor (GPCR).................. 151–163 Microvesicles (MVs)................................................91–100
Molecular marker ................................................. 331–343
H Monoclonal antibodies (mAbs)......................58, 60, 117,
144, 156
High-content automated electrophysiology ............... 185 Monolayer ....................................... 33, 37, 96, 115, 130,
131, 133–135, 178, 179, 223, 224, 230–238, 244
I
Immobilized cells ........................................ 249, 256, 257 N
Immunoaffinity ............................................................... 94 Nanoparticles........................................................vi, 79–89
Immunocytochemistry............................... 128, 224, 226, Neuron vi, 168, 170, 172, 173, 179, 186, 211–221, 284
262, 281, 282, 284, 285, 287, 291, 299 N-glycoproteins.....................................59, 60, 62, 70, 72
Immunohistochemistry (IHC).................................35, 39 N-linked antennas ........................................................... 40
Immunophenotyping................................. 57–70, 73, 74, Non-adherent cells.........................................97, 118, 201
76, 127–134, 136, 138–140, 142, 144–147
Immunoprecipitation (IP) ......................... 9, 11, 16, 208, O
215, 217, 219
Induced pluripotent stem cells (iPSCs) ................. 62, 73, Oligosaccharides.......................................... 59, 63, 65, 66
128, 178, 261–291, 293, 294, 296–300 O-linked antennas ........................................................... 40
Integrin ..................... 71, 73, 74, 94, 129, 238, 249, 290 Optical recording ................................................. 173, 178
Intercellular signaling ..................................................... 95
P
Internalization ...........152, 154–156, 158, 160–162, 214
Intracellular calcium concentration............ 198, 201, 202 Pellicles ......................................................................79–89
Ion channel.........................................167, 168, 171–173, Peptide separation ............................................................. 4
175–177, 180, 185, 224, 230 Peptide-N-glycosidase F (PNGaseF) ..................... 59, 69,
In vitro cytotoxicity assay ........................... 118, 122, 123 70, 111, 113, 114, 229, 242
In vivo biotinylation..................................................37–39 Phagocytosis ........................................224, 227, 236, 237
 THE SURFACEOME: METHODS AND PROTOCOLS
Index 347
Phosphorylation ..................................170, 213–215, 218 Streptococcus pyogenes.................................................22, 24
Plasma membrane (PM) ............................ 43, 59, 65, 73, Stress–strain relations........................................... 263, 305
74, 79–89, 92, 152, 163, 172, 223, 231, 240, Subcellular localization ................................ 4, 12, 19, 58,
245, 333, 339, 343 169, 172, 174
Polarization ..................................................224, 230–233 Surface expression ...................................... 180, 212, 214,
Polymerase chain reaction (PCR) ........................ 14, 108, 215, 341
128, 172, 258, 262–264, 266, 278, 279, 290, Surface protein .................................3, 13, 21–28, 58, 59,
293, 341 61, 62, 74, 140, 185, 220, 224, 237, 238, 241,
Proteases .................................................... 14, 21–28, 241 262, 263, 282, 283, 298, 304, 331–334, 336,
Protein enzymatic cleavage............................................... 4 338–340, 342
Protein interaction ............................ 196, 241, 331, 332, Surfaceome .......................................3–19, 22, 57–70, 73,
334, 337, 340 74, 76, 79, 129, 261–291, 293, 294, 296–300,
Protein mapping.............................................................. 31 303, 332, 334, 341
Protein-protein interactions ...............331, 332, 334, 337 Surfome ........................................................22–25, 27, 28
Proteolytic digestion .................................................33, 35
Proteomics....................................... 5, 22, 23, 58, 73, 75, T
93, 94, 97, 98, 229, 241, 244, 332 Targets-search ..................................................62, 332–337
Purification ....................................... v, 6, 8, 9, 16, 33–35,
Transcriptome .................................................44, 47, 241,
39, 41, 59, 105–115 333, 336
Transfection...............................106, 107, 109, 110, 115,
R
158, 208, 219
Receptor internalization ..................................... 155, 158, Transient receptor potential (TRP) channel...... 195–206,
160–162, 214 208, 209
Receptor tyrosine kinase ...................................... 211–221 Transmembrane protein ...............................79, 105–115,
Reprogramming ......................................... 262, 263, 265, 195, 239, 241, 303, 333, 334, 336, 341, 342
270–273, 275–277, 293 Transmission electron microscopy ............. 224, 226, 231
Retinal pigment epithelium (RPE) ....................BNF–246 Trans-proteomic pipeline..........................................76, 99
RNA ................................47, 49, 92, 172, 177, 178, 249, Tryptic peptides....................................... 4, 10, 14, 17, 18
252, 257–259, 270, 274, 279, 284–287, 290, Tumor cell implantation ................................................. 37
291, 297, 333, 338, 339
RNA-seq data .............................................. 333, 338, 339 U
Ultracentrifugation ..................................... 8, 15, 94, 100
S
Sample solubilization ........................................................ 3 V
Secreted proteins .............................................. 79, 91–100 Vascular endothelial cell......................195–206, 208, 209
Secretome ............................................91, 92, 94–96, 100
Vascular endothelium............................................. 31, 196
Seizure ........................................................................... 171 Voltage dependent Na channels ..................168–171, 180
Sf9 cells ................................................................. 105–115
Shaving ......................................................................21–28 W
Sialidase......................................................................35, 40
Signal transduction .................................... 195, 211–221, Web-based platform ............................................. 332, 333
331, 338, 343 Western ................................. 16, 41, 128, 196, 197, 199,
Smooth muscle cells (SMCs).............................. 261–294, 200, 205–207, 215, 217, 219, 220, 224, 229,
296–300, 324, 325 238, 239, 243, 244
Sodium channels .................................................. 168–171 Wound healing .................................................32, 37, 325
Stem cells .............................................. 57–70, 73, 74, 76,
127–134, 136, 138–140, 142, 144–147, 180,
185, 191, 251, 255, 279, 297, 338, 339