(Methods in Molecular Biology 1323) Rémy Bosselut, Melanie S. Vacchio (Eds.) - T-Cell Development - Methods and Protocols (2016, Humana Press) PDF

Methods in
Molecular Biology 1323
Rémy Bosselut
Melanie S. Vacchio Editors
T-Cell
Development
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
T-Cell Development
Methods and Protocols
Edited by
Rémy Bosselut and Melanie S. Vacchio

Laboratory of Immune Cell Biology, National Institutes of Health,
National Cancer Institute, Bethesda, MD, USA
Editors
Rémy Bosselut Melanie S. Vacchio
Laboratory of Immune Cell Biology Laboratory of Immune Cell Biology
National Institutes of Health National Institutes of Health
National Cancer Institute National Cancer Institute
Bethesda, MD, USA Bethesda, MD, USA
ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-2808-8 ISBN 978-1-4939-2809-5 (eBook)
DOI 10.1007/978-1-4939-2809-5
Library of Congress Control Number: 201506151
Springer New York Heidelberg Dordrecht London

© Springer Science+Business Media New York 2016
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed
to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
Printed on acid-free paper
Humana Press is a brand of Springer

Springer Science+Business Media LLC New York is part of Springer Science+Business Media (www.springer.com)
Preface
Fifty and a few years since the function of the thymus was discovered, the study of T cell
development remains a fascinating part of immunology. This is certainly because T cells,
essential components of immune responses, develop in the thymus. Interactions in the thy-
mus shape the reactivity and function of T cell precursors and are essential to establish a
functional and tolerant immune system. Thus, identifying intrathymic interactions and
understanding their mechanisms are intrinsic parts of T cell immunology. In addition, the
study of T cell (and more broadly lymphocyte) development has exerted broad appeal
beyond the immunological realm. Indeed, throughout their thymic journey, T cell precur-
sors recapitulate key processes involved in metazoan development, including proliferation,
differentiation, death-survival, and migration decisions. Studies of T cell development have
brought decisive insight into critical concepts of modern biology, including somatic DNA
recombination, DNA repair, programmed cell death, and epigenetic gene silencing.
For all its accomplishments and potential, the study of T cell development enjoys the
not-so-enviable reputation of being “complicated.” While this is in part due to the sheer
complexity of T cell development, it certainly also has to do with difficulties in experi-
mental approaches. The objective of the present volume of Methods in Molecular Biology
is to overcome such practical obstacles, by giving simple and accessible experiment pro-
tocols to explore thymus biology. This book is organized in three parts. The first two
chapters offer short reviews on T cell development, for readers with little or no familiarity
with the topic. Both provide basic concepts as a preparation to parsing the protocol chap-
ters, and references for further study. The second part discusses analysis strategies and
presents basic protocols for cell preparation, flow cytometry analyses, and the study of T
cell responses. The last part includes state-of-the-art protocols to explore multiple aspects
of thymocyte biology.
In building this selection, we have tried to put emphasis on three themes. The first is
the emergence and refinement, over the last few years, of a remarkable panoply of in vitro
approaches to study T cell development. While in vivo studies, harnessing the power of
mouse genetics, remain the standard of proof, new strategies offer unprecedented possibili-
ties to investigate thymic biology in a dish. Such approaches, using derivatives of the OP9
stromal line expressing Notch ligands, have truly revolutionized the study of early thymo-
cyte development. In addition, continued refinement of organ culture systems has increased
their power to address later stages of αβ T cell development, including mechanisms of selec-
tion. The present volume includes detailed protocols for three of these approaches: fetal
thymic organ culture, reconstituted organ culture, and culture of thymic “slices.” Future
progress can be expected from attempts at generating thymic epithelial cells in vitro, includ-
ing using “reprogramming” approaches. An important asset of organ culture is to open the
door to live imaging techniques, as described in the chapter on thymic slices, which incor-
porates time and space resolution into the study of thymocyte differentiation. Last but not
least, because in vitro techniques can combine T cell precursors and stromal cells of distinct
genotypes, they offer useful alternatives to complex breeding strategies.
v
vi Preface
The second point we wanted to emphasize is the plasticity of thymocytes, which can
fine-tune intracellular signaling and gene expression to specific environmental conditions.
This is critical during the building of the TCR repertoire: by shifting the repertoire of TCR
specificities being selected, thymocyte plasticity offers the thymic organ a major potential to
compensate signaling or transcriptional changes at the single-cell level. This emphasizes the
importance of genetic approaches analyzing the development of cells carrying a defined
specificity, together with bone marrow chimera techniques allowing the study of cell devel-
opment in a competitive environment. More broadly, even though there are few truly
single-cell assays for the study of T cell development, such cell plasticity should be taken
into consideration when designing experimental approaches.
Last, we have tried to extend the range of protocols presented in this volume beyond
the “workhorse” laboratory mouse. The power of mouse genetics has been instrumental to
T cell immunology, including in defining such essential concepts as MHC restriction and
positive selection. The last decade has seen the realization that thymic functions are present
early on during evolution, actually predating the emergence of the “modern” lymphoid
system of jawed vertebrates. This strengthens the rationale to study T cell development in
other organisms, to bolster concepts generated from mouse studies, especially if they offer
unique opportunities for genetic intervention. The zebrafish embryo is one such powerful
experimental model, allowing for oligonucleotide-directed “knock-down” of gene function
and offering quick and powerful approaches for the study of T cell development. On the
other hand, it is becoming possible, and increasingly important, to use the conceptual
frameworks generated in mouse models as platforms to investigate human T cell develop-
ment, with the objective of understanding how repertoire formation and selection drives T
cell responsiveness in physiological and pathological situations. Three chapters address
these issues, providing approaches to study human thymocytes and to model T cell devel-
opment in humanized mice. New methods for in vivo gene recombination, including
improvements of the bacteria-derived CRISPR system, should increase the power of such
studies in the years to come.
It is our hope that the present volume will be useful to readers whose scientific curiosity
brings them to interrogate the thymus, whether because of its role in shaping the T cell
repertoire or of its ability to encapsulate, in a simple and accessible organ, most of the basic
processes at work in developmental biology. We are grateful to the many authors who
shared their unique expertise to make this volume possible.
Bethesda, MD, USA Rémy Bosselut

Melanie S. Vacchio
Contents
Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
PART I BACKGROUND INFORMATION

1 200 Million Thymocytes and I: A Beginner’s Survival Guide
to T Cell Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Melanie S. Vacchio, Thomas Ciucci, and Rémy Bosselut
2 Development of γδ T Cells, the Special-Force Soldiers
of the Immune System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
David L. Wiest
PART II CORE APPROACHES AND STRATEGIES

3 Genetic Tools to Study T Cell Development . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Thomas Ciucci, Melanie S. Vacchio, and Rémy Bosselut
4 Assessment of T Cell Development by Flow Cytometry. . . . . . . . . . . . . . . . . . 47
Jan Y.M. Lee and Paul E. Love
5 Flow Cytometry Analysis of Thymic Epithelial Cells and Their
Subpopulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Izumi Ohigashi and Yousuke Takahama
6 Identifying the Spatial Relationships of Thymic Stromal
and Thymocyte Subsets by Immunofluorescence Analysis . . . . . . . . . . . . . . . . 75
Virginia Bain and Ellen R. Richie
7 Purification of Thymocyte and T Cell Subsets . . . . . . . . . . . . . . . . . . . . . . . . . 87
Andrea C. Carpenter, Jong Kyong Kim, and Rémy Bosselut
8 Retroviral Transduction of T Cells and T Cell Precursors. . . . . . . . . . . . . . . . . 99
Amie Simmons and José Alberola-Ila
9 Bone Marrow and Fetal Liver Radiation Chimeras. . . . . . . . . . . . . . . . . . . . . . 109
Francis A. Flomerfelt and Ronald E. Gress
10 In Vitro Analyses of T Cell Effector Differentiation . . . . . . . . . . . . . . . . . . . . . 117
Elizabeth A. Wohlfert, Andrea C. Carpenter, Yasmine Belkaid,
and Rémy Bosselut
PART III SPECIFIC TECHNIQUES

11 Studying T Cell Development in Thymic Slices . . . . . . . . . . . . . . . . . . . . . . . . 131
Jenny O. Ross, Heather J. Melichar, Joanna Halkias, and Ellen A. Robey
12 FTOC-Based Analysis of Negative Selection . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Cody A. Cunningham, Emma Teixeiro, and Mark A. Daniels
vii
viii Contents
13 Reconstituted Thymus Organ Culture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Zimu Deng, Haifeng Liu, Jinxiu Rui, and Xiaolong Liu
14 Induction of T Cell Development In Vitro by
Delta-Like (Dll)-Expressing Stromal Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Mahmood Mohtashami, Payam Zarin, and Juan Carlos Zúñiga-Pflücker
15 In Vitro Analysis of Thymocyte Signaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
16 Molecular Analysis of Mouse T Cell Receptor α and β
Gene Rearrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Levi J. Rupp, Liang Chen, Michael S. Krangel, and Craig H. Bassing
17 Intrathymic Injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Sugata Manna and Avinash Bhandoola
18 Analysis of Cell Proliferation and Homeostasis Using EdU Labeling . . . . . . . . 211
19 Characterization and Isolation of Human T Cell Progenitors. . . . . . . . . . . . . . 221
Inge Van de Walle, Karin Davids, and Tom Taghon
20 Approaches to Study Human T Cell Development . . . . . . . . . . . . . . . . . . . . . 239
Anne-Catherine Dolens, Inge Van de Walle, and Tom Taghon
21 Humanized Mice to Study Human T Cell Development . . . . . . . . . . . . . . . . . 253
Sarah Bonte, Sylvia Snauwaert, Stijn Vanhee, Anne-Catherine Dolens,
Tom Taghon, Bart Vandekerckhove, and Tessa Kerre
22 Using the Zebrafish Model to Study T Cell Development . . . . . . . . . . . . . . . . 273
Yong Zhang and David L. Wiest
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Contributors
JOSÉ ALBEROLA-ILA • Immunobiology and Cancer Program, Oklahoma Medical

Research Foundation, Oklahoma City, OK, USA
VIRGINIA BAIN • Department of Epigenetics and Molecular Carcinogenesis, The University
of Texas MD Anderson Cancer Center, Smithville, TX, USA
CRAIG H. BASSING • Division of Cancer Pathobiology, Department of Pathology and
Laboratory Medicine, Center for Childhood Cancer Research, Children’s Hospital of
Philadelphia, Philadelphia, PA, USA; Abramson Family Cancer Research Institute,
Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA;
Cell and Molecular Biology Graduate Group, Perelman School of Medicine at the
University of Pennsylvania, Philadelphia, PA, USA
YASMINE BELKAID • Program in Barrier Immunity and Repair, Mucosal Immunology
Section, Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious
Disease, NIH, Bethesda, MD, USA
AVINASH BHANDOOLA • T-Cell Biology and Development Section, Laboratory of Genome
Integrity, Center for Cancer Research, National Cancer Institute, National Institutes of
Health, Bethesda, MD, USA
SARAH BONTE • The Department of Hematology and Clinical Chemistry, Microbiology and
Immunology, Faculty of Medicine and Health Sciences Ghent University Hospital, Ghent
University, Ghent, Belgium
RÉMY BOSSELUT • Laboratory of Immune Cell Biology, Center for Cancer Research,
National Cancer Institute, National Institutes of Health, Bethesda, MD, USA
ANDREA C. CARPENTER • Laboratory of Immune Cell Biology, Center for Cancer Research,
LIANG CHEN • Department of Immunology, Duke University School of Medicine, Durham,
NC, USA
THOMAS CIUCCI • Laboratory of Immune Cell Biology, Center for Cancer Research,
CODY A. CUNNINGHAM • Molecular Microbiology and Immunology, Center for Cellular
and Molecular Immunology, Columbia, MO, USA; Department of Surgery,
Center for Cellular and Molecular Immunology, Columbia, MO, USA
MARK A. DANIELS • Department of Molecular Microbiology and Immunology, Center for
Cellular and Molecular Immunology, Columbia, MO, USA; Department of Surgery,
KARIN DAVIDS • The Department of Clinical Chemistry, Microbiology and Immunology,
Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium
ZIMU DENG • State Key Laboratory of Cell Biology, Institute of Biochemistry and Cell
Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences,
Shanghai, China
ANNE-CATHERINE DOLENS • The Department of Clinical Chemistry, Microbiology and
Immunology, Faculty of Medicine and Health Sciences Ghent University Hospital,
Ghent University, Ghent, Belgium
ix
x Contributors
FRANCIS A. FLOMERFELT • Experimental Immunology and Transplantation Branch,

RONALD E. GRESS • Experimental Immunology and Transplantation Branch,
JOANNA HALKIAS • Division of Immunology and Pathogenesis, Department of Molecular
and Cell Biology, University of California, Berkeley, CA, USA; Division of Neonatology,
Children’s Hospital and Research
Center Oakland, Oakland, CA, USA
TESSA KERRE • The Department of Clinical Chemistry, Microbiology and Immunology,
Faculty of Medicine and Health Sciences Ghent University Hospital, Ghent University,
Ghent, Belgium
JONG KYONG KIM • State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center,
Sun Yat-sen University, Guangzhou, China
MICHAEL S. KRANGEL • Department of Immunology, Duke University School of Medicine,
Durham, NC, USA
JAN Y.M. LEE • Section for Cellular and Developmental Biology, Program on Genomics of
Differentiation, Eunice Kennedy Shriver National Institute of Child Health and
Human Development, National Institutes of Health, Bethesda, MD, USA
HAIFENG LIU • State Key Laboratory of Cell Biology, Institute of Biochemistry and Cell
Shanghai, China
XIAOLONG LIU • State Key Laboratory of Cell Biology, Institute of Biochemistry and Cell
Shanghai, China
PAUL E. LOVE • Section for Cellular and Developmental Biology, Program on Genomics of
Differentiation, Eunice Kennedy Shriver National Institute of Child Health and
Human Development, National Institutes of Health, Bethesda, MD, USA
SUGATA MANNA • Laboratory of Immune Cell Biology, Center for Cancer Research,
HEATHER J. MELICHAR • Division of Immunology and Pathogenesis, Department of
Molecular and Cell Biology, University of California, Berkeley, CA, USA
MAHMOOD MOHTASHAMI • Department of Immunology, Sunnybrook Research Institute,
University of Toronto, Toronto, ON, Canada
IZUMI OHIGASHI • Division of Experimental Immunology, Institute for Genome Research,
University of Tokushima, Tokushima, Japan
ELLEN R. RICHIE • Department of Epigenetics and Molecular Carcinogenesis,
The University of Texas MD Anderson Cancer Center, Smithville, TX, USA
ELLEN A. ROBEY • Division of Immunology and Pathogenesis, Department of Molecular
and Cell Biology, University of California, Berkeley, CA, USA
JENNY O. ROSS • Division of Immunology and Pathogenesis, Department of Molecular
and Cell Biology, University of California, Berkeley, CA, USA
JINXIU RUI • State Key Laboratory of Cell Biology, Institute of Biochemistry and Cell
Shanghai, China
Contributors xi
LEVI J. RUPP • Division of Cancer Pathobiology, Department of Pathology and Laboratory

Medicine, Center for Childhood Cancer Research, Children’s Hospital of Philadelphia,
Philadelphia, PA, USA; Abramson Family Cancer Research Institute, Perelman School of
Medicine at the University of Pennsylvania, Philadelphia, PA, USA; Cell and Molecular
Biology Graduate Group, Perelman School of Medicine at the University of Pennsylvania,
Philadelphia, PA, USA
AMIE SIMMONS • Immunobiology and Cancer Program, Oklahoma Medical Research
Foundation, Oklahoma City, OK, USA
SYLVIA SNAUWAERT • The Department of Hematology, Faculty of Medicine and Health
Sciences Ghent University Hospital, Ghent University, Ghent, Belgium
TOM TAGHON • The Department of Clinical Chemistry, Microbiology and Immunology,
Ghent, Belgium
YOUSUKE TAKAHAMA • Division of Experimental Immunology, Institute for Genome
Research, University of Tokushima, Tokushima, Japan
EMMA TEIXEIRO • Department of Molecular Microbiology and Immunology, Center for
Cellular and Molecular Immunology, Columbia, MO, USA; Department of Surgery,
MELANIE S. VACCHIO • Laboratory of Immune Cell Biology, Center for Cancer Research,
BART VANDEKERCKHOVE • The Department of Clinical Chemistry, Microbiology and
STIJN VANHEE • The Department of Clinical Chemistry, Microbiology and Immunology,
Ghent, Belgium
INGE VAN DE WALLE • The Department of Clinical Chemistry, Microbiology and
DAVID L. WIEST • Blood Cell Development and Function Program,
Fox Chase Cancer Center, Philadelphia, PA, USA
ELIZABETH A. WOHLFERT • Department of Microbiology and Immunology, School of
Medicine and Biomedical Sciences, University at Buffalo (SUNY), Buffalo, NY, USA
PAYAM ZARIN • Department of Immunology, Sunnybrook Research Institute, University of
Toronto, Toronto, ON, Canada
YONG ZHANG • Blood Cell Development and Function Program, Fox Chase Cancer Center,
Philadelphia, PA, USA
JUAN CARLOS ZÚÑIGA-PFLÜCKER • Department of Immunology, Sunnybrook
Research Institute, University of Toronto, Toronto, ON, Canada
Part I
Background Information
Chapter 1
200 Million Thymocytes and I: A Beginner’s Survival

Guide to T Cell Development
Melanie S. Vacchio, Thomas Ciucci, and Rémy Bosselut
Abstract
T lymphocytes (T cells) are essential for proper adaptive immune responses. They perform a variety of
functions in defenses against pathogens, and notably control, positively or negatively, other cells involved
in immune responses. T cells develop in the thymus from bone marrow-derived precursors. These precur-
sors (thymocytes) proliferate, rearrange the genes encoding subunits of the T cell antigen receptor, which
endow them with their unique antigen specificity, and undergo various degrees of pre-programming for
their functions in immune responses. Thus, analyzing T cell development in the thymus is essential for
understanding their functions in immune responses. In addition, the thymus constitutes an attractive
experimental model to analyze mechanisms of cell proliferation, differentiation and survival, all of which
are involved in thymocyte development. This chapter presents a quick overview of the key events charac-
terizing intrathymic T cell development, as an introduction for readers entering this field of study.
Key words T cells, T cell development, T cell receptor, TCR gene rearrangement, Positive selection,
Negative selection
1 T Cells and T Cell Antigen Recognition
T lymphocytes (T cells) are critical components of the adaptive

immune system and are essential for responses to foreign organ-
isms. Key aspects of T cell responsiveness are set during their devel-
opment in the thymus [1, 2]; thus, studying the development of T
cells provides invaluable insight into their functions, and has
attracted much interest. The following chapter is intended as a
primer for those entering this field of study, unfolding key events,
highlighting checkpoints, and introducing experimental tools to
explore the developmental sequence. We have directed the reader
to in-depth reviews and a few original papers for more information
of each of the points addressed in this overview. While key con-
cepts discussed here are thought valid for both human and mouse
T cell development, the description of specific stages, signals and
Rémy Bosselut and Melanie S. Vacchio (eds.), T-Cell Development: Methods and Protocols, Methods in Molecular Biology,
vol. 1323, DOI 10.1007/978-1-4939-2809-5_1, © Springer Science+Business Media New York 2016
3
4 Melanie S. Vacchio et al.
transcription factors refer to mouse T cell development. Issues

specific to human T cell development are discussed in Chapter 19.
T cells recognize antigens through a receptor (T cell antigen
receptor, TCR) composed of two distinct polypeptide chains [3].
Two possible pairs of chains have been identified: TCRα and TCRβ,
or TCRγ and TCRδ, defining αβ and γδ T cells, respectively. None
of these chains has signaling activity per se. Rather, they are
expressed at the cell surface as a complex with signaling chains
(TCRζ and CD3 complex) [4–6], which generates intracellular sig-
nals upon antigen recognition (Fig. 1a). Expression of the genes
encoding the signaling CD3 chains is a defining attribute of the T
lineage. Signaling by TCR complexes has been extensively investi-
gated, mostly in αβ T cells (but is thought to operate along the
same principles in γδ T cells). It is triggered, upon antigen binding,
by the phosphorylation of tyrosine residues in TCRζ and CD3
chains, which itself recruits kinases that propagates phosphorylation
a
TCR
MHC
peptide CD4
V V
C C
CD3 CD3 P Lck
P
P
Zap70
TCR
P
Fig. 1 T cell receptor structure and signaling. (a) Schematic representation of

TCRαβ complexes, showing each subunit. Rectangular boxes in signaling chains
depict Tyrosine (black ovals)-based ITAM motifs. While there is evidence that
CD3δε interact with TCRα and TCRγε with TCRβ, the architecture of the complex
as represented is speculative. V and C respectively indicate variable and constant
domains of each TCR chain. (b) Schematic depiction of TCR and coreceptor (CD4)
interactions with MHC-II-peptide (filled black circle) complexes, and of early TCR
signaling. Co-engagement of TCR and CD4 by MHC-II (or of TCR and CD8 by
MHC-I, not represented) results in the juxtaposition of Lck molecules and CD3
and TCRζ signaling chains, phosphorylation of tyrosine residues, recruitment and
activation of the Zap70 kinase, which phosphorylates and activates downstream
targets
200 Million Thymocytes and I: A Beginner’s Survival Guide… 5
and intracellular signaling to substrates not physically linked to

TCR complexes (Fig. 1b).
The vast majority of αβ T cells recognize peptide antigens pre-
sented by “classical” class I (MHC-I) or class II (MHC-II) major
histocompatibility complex molecules, and are referred to as “con-
ventional” T cells [5]. MHC molecules are expressed at the cell
surface in association with peptides generated through mechanisms
that differ between MHC-I and MHC-II. MHC-I molecules,
which are ubiquitously expressed, bind peptides resulting from the
proteasome-mediated degradation of proteins synthesized in the
cell; in contrast, MHC-II molecules are expressed mostly on pro-
fessional antigen-presenting cells and bind peptides resulting from
the lysosomal degradation of endocytosed proteins. Antigen rec-
ognition by αβ T cells involves interactions of TCRαβ chains with
both peptide and MHC amino acid residues [7]. An important
implication of this binding mode is that recognition of self (amino
acids from MHC molecules) is an inherent component of αβ T cell
antigen reactivity because it contributes to recognition of foreign
antigens. As discussed below, this has critical consequences for αβ
T cell development in the thymus.
Conventional T cells recognizing MHC-I- or MHC-II-
associated peptides form the vast majority of T cells in peripheral
lymphoid organs and at most effector sites [5]. Each of them
expresses monospecific (“clonotypic”) TCR complexes (i.e. each
receptor complex in a given cell harbors the same uniquely rear-
ranged TCRα and TCRβ sequence); upon recognition of a specific
(“cognate”) MHC-peptide complex, such T cells proliferate and
acquire effector properties (e.g. cytokine production or expression
of specific surface molecules). Even though each T cell carries one
clonotypic TCR species, the multiple specificities they express as a
population underpins the breadth of T cell antigen recognition.
Conventional T cells are distributed into two subsets that differ by
their function and their expression of CD4 or CD8, two surface
glycoproteins that contribute to recognition of MHC molecules
and signaling by T cells (Fig. 1b) [8]. MHC-I-specific cells express
CD8 but not CD4, and typically differentiate into cytotoxic effec-
tors upon activation. MHC-II-specific cells express CD4 but not
CD8, and have a broader functional potential (often designated as
“helper”); in fact, their effector differentiation is to a large extent
determined upon antigen activation by the surrounding cytokine
environment. Unlike in mature T cells, expression of CD4 and
CD8 on thymocytes depends on their developmental stage, and
distinguishes the conventional CD4−CD8− “double negative”
(DN), CD4+CD8+ “double positive” (DP) and CD4+CD8− or
CD4−CD8+ “single positive” (SP) cell subsets, the latter three
being characteristic of the αβ lineage [2].
In addition to conventional T cells, many “non-conventional”
αβ T cell subsets have been identified, often expressing neither
CD4 nor CD8. Among the best characterized are natural killer
(NK) T cells, of which most recognize CD1d-bound lipids,
mucosal-associated invariant T cells (MAIT), which recognize
MR1-bound riboflavin derivatives, and intraepithelial lymphocytes
(IELs) [9–12]. The organization of the γδ T cell population is less
well understood (see Chapter 2). It includes multiple subsets differ-
ing by antigen receptor chain usage, functional properties, and
unique developmental attributes (including their appearance at
specific embryonic or fetal developmental stages). Most such cells
express neither CD4 nor CD8.
2 An Overview of T Cell Development
2.1 A Brief The “objective” of intrathymic T cell development is twofold.

Thymic Tour It first must generate the variety of T cells necessary for immune
responses. This implies a massive expansion of the small popula-
tion of hematopoietic precursors that enter the T lineage, and the
building, by somatic recombination of genes encoding TCR
chains, of a large repertoire of antigen-specific receptors. Second,
for αβ T cells, selection steps (death-survival decisions) ensure
that this repertoire is appropriate to function with antigen-
presenting molecules expressed by each individual. Whether and
how the γδ repertoire is subject to selection is not yet fully under-
stood (see Chapter 2).
Meeting these “objectives” is achieved through a complex
developmental sequence associating differentiation, selection, and
proliferation episodes (Fig. 2). Such events take place in the thy-
mus, a thoracic organ made of bone marrow-derived hematopoietic-
derived cells (including multiple cell types in addition to T cell
precursors) and epithelial cells derived from the third pharyngeal
pouch [13]. The thymus appears during early embryonic develop-
ment, is most active during childhood and involutes after puberty.
Thus, the T cell repertoire itself is mostly constituted during the
early years of life and starts declining in early adulthood. Although
this review focuses on the T cell component of the thymus, other
cell types are essential to its function, perhaps most prominently
thymic epithelial cells [14]. The distribution of T cell precursors
and epithelial cells in the thymus is not homogenous, and its com-
partmentalization is critical to thymic function. The most salient,
almost macroscopic, feature of the thymic architecture is the dis-
tinction between the thymic cortex, in which early T cell develop-
ment takes place, and the medulla, where thymocytes complete
their maturation. Each compartment is characterized by distinct
subtypes of epithelial cells, although it is now clear that both med-
ullary and cortical thymic epithelial cells derive from a common
precursor [15, 16]. This will be discussed again near the end of this
overview.
Thymus Peripheral lymphoid organs
Hematopoietic Committed Repertoire Mature Mature Effector

precursor precursor generation Selection T cell T cell differentiation
Proliferation Proliferation
T lineage Antigen receptor Antigen

specification gene stimulation
and commitment rearrangement
Fig. 2 Schematic outline of T cell development. T cell development can be conceptually separated into four
steps: (i) the T lineage commitment of recent thymic immigrants, which enter the thymus as uncommitted
precursors, (ii) the rearrangement of TCR genes (dependent on Rag gene expression), itself a multistep pro-
cess, of which a critical outcome is the distribution of thymocytes into the αβ or γδ lineage, and (iii) the MHC-
based selection of thymocytes expressing αβ TCRs, during which only cells with appropriate avidity for self
antigens survive and complete their intrathymic differentiation. Most of these cells exit the thymus as “naïve,”
and acquire effector (“helper” or cytotoxic) functions during antigen-induced proliferation in peripheral lym-
phoid organs (see text for details). However, most γδ thymocytes, and a small fraction of αβ lineage cells,
acquire such effector functions in the thymus. Note that there is “chronological” overlap between these con-
ceptual steps during the developmental sequence. This is most striking in the αβ lineage, in which commit-
ment to the αβ lineage follows the rearrangement of TCRβ genes but precedes that of TCRα, and termination
of the latter is mechanistically coupled to MHC-based selection
2.2 Early T Cell The earliest steps of intrathymic development, thought to be

Development: common to all T cells, lead bone marrow-derived hematopoietic
The Common Core precursors to enter the thymus, commit to the T cell lineage and
initiate the rearrangement of genes that encode antigen receptor
chains [17–19]. The migration of such precursors into the thymus
requires their egress from the bone marrow, which we will not
discuss here, and their leaving the circulation in blood vessels that
irrigate the thymic cortico-medullary junction [20]. The latter
process is controlled both by adhesion molecules and chemokines,
including the production by the thymic epithelium of CCL19 and
CCL21, the ligands for CCR7 expressed by colonizing precur-
sors. As a result of their thymic entry, precursors are exposed to
ligands for the Notch1 receptor, specifically Delta-like 4 expressed
on thymic epithelial cells [21, 22]. Both gain and loss-of-function
studies have shown that signaling by Notch1 in T cell precursors
is required for their early development and notably for T-lineage
commitment [23, 24]. Recent studies have identified transcrip-
tion factors that mediate T lineage specification or commitment
[25]. These include HES1 and TCF1, which appear directly
downstream of Notch1, and other transcription factors, such as
Gata3 and Bcl11b [26–28].
In addition to Notch1 signals, early T cell development requires

the cytokine IL-7, which supports the massive proliferation that
takes place prior to rearrangement of TCR genes, and promotes
thymocyte survival in part at least by affecting the balance between
pro- and anti-apoptotic molecules of the Bcl2 family [29].
TCR chain gene rearrangement is the critical event defining
the T lineage. It occurs in distinct steps spreading over several
stages of T cell development. Since the mechanisms of TCR gene
rearrangement [30] are discussed in Chapter 16, only a few basic
principles will be highlighted here. In the germline, non-rearranged,
configuration, each TCR locus contains multiple copies of the
DNA segments [referred to as variable (V), Joining (J) and (for
TCRβ and TCRδ) Diversity (D)] encoding the variable domain of
each TCR chain (Fig. 1a). Such copies, which differ by their
sequence, are arranged in V, D and J clusters within each TCR
locus. Rearrangement involves the cis-joining of such gene seg-
ments within the TCR chain-encoding locus, and joins a given
member of one cluster to a member of another one (i.e. V to J, or
V to D to J). This occurs through excision, mediated by Rag1 and
Rag2 recombinase molecules, of large intervening DNA frag-
ments, followed by ligation of the free chromosomal DNA extrem-
ities. In contrast to Rag1 and Rag2, which are specifically expressed
in lymphocyte precursors, the joining of DNA breaks involves an
ubiquitous DNA repair machinery.
Gene rearrangement generates TCR diversity by two pro-
cesses: (1) combinatorial diversity, as the rearranged allele includes
a single copy of each V, D or J segment, out of the many present in
the germline, and (2) “junctional” diversity, resulting from nucleo-
tide deletions and non-templated additions made during the join-
ing process to the original DNA breaks generated by the
Rag-mediated cleavage. Proper gene rearrangement requires the
stage-specific targeting of the Rag-mediated excision machinery to
the appropriate TCR gene locus. For αβ lineage cells, TCRβ is
rearranged in DN cells and TCRα at the DP stage. Whether spe-
cific signals direct thymocytes to rearrange TCRβ vs. TCRγ and
TCRδ, or whether they “randomly” attempt to rearrange genes
encoding TCRβ or TCRγ and adopt an αβ or γδ fate if either rear-
rangement is productive, remains to be determined.
2.3 Emergence Upon successful completion of TCRβ gene rearrangement, the

of the αβ Lineage resulting TCRβ polypeptides associate with a surrogate TCRα
chain, pre-Tα (instead of TCRα) and CD3 chains to form pre-
TCR complexes [31]. Signaling by such complexes allows thymo-
cytes to clear the checkpoint for TCRβ gene rearrangement, a
processed referred to as β-selection, and causes them to adopt an
αβ fate [32]. Because rearrangement of TCRβ or of TCRγ and
TCRδ occur during the same developmental window, and because
pre-TCR and TCR complexes use common signaling components,
the question arises how developing thymocytes distinguish between

signaling by the pre-TCR, which drives the cells toward an αβ fate,
vs. TCRγδ, which causes them to adopt a γδ fate. The current per-
spective, discussed in more detail in Chapter 2 is that there is no
lineage-instruction signal delivered by either type of receptor.
Rather, there is evidence that intracellular signals elicited by the
pre-TCR are of lower magnitude than those triggered by TCRγδ
complexes, and that developing thymocytes convert such differ-
ences in signaling intensities into αβ or γδ fate commitment signals
[33–35].
Signaling by the pre-TCR fixes αβ lineage commitment by ter-
minating TCRβ gene rearrangement [32]. It also causes the prolif-
eration of αβ precursors and the expression of genes encoding
CD4 and CD8 coreceptors, as well as multiple other gene expres-
sion changes [36]. Each cell going through β-selection gives rise to
multiple daughter DP cells expressing the same TCRβ polypeptide
and actively rearranging TCRα genes, thereby adding to the com-
binatorial diversity generated by antigen gene rearrangement and
maximizing the potential for use of each TCRβ rearrangement.
These cells, which form the bulk of the thymus, are characterized
by low-level expression of surface TCRαβ (characteristically less
than mature αβ T cells). Among these changes are increased expres-
sion of CD27 and CD28 surface antigens, which both can be used
to identify the earliest precursors that have undergone β-selection,
before the onset of CD4 or CD8 expression [37, 38] (Fig. 4).
Several transcription factors, including Gata3, Runx1, E-proteins
E2A and HEB (see below) and RORγt, and multiple signaling
pathways are important for β-selection [2]. Of note, the differen-
tiation of αβ lineage progenitors at the β-selection step retains a
significant dependence on Notch signals [35].
2.4 Selection of αβ T The selection and differentiation of DP thymocytes into mature

Cells: Basic Principles CD4+ and CD8+ T cells has attracted much interest because of the
critical importance of αβ cells in immune responses and their
pathogenic role in autoimmune disease. DP thymocytes, which
reside in the thymic cortex, do not proliferate; thus, even though
the thymus is characterized by rapid cell renewal, it is mainly
populated (85–90 % of thymocytes) by non-dividing cells, a para-
dox explained by the short life of DP thymocytes. The fate of these
cells is primarily dictated by the reactivity of their αβ TCR to self-
peptides bound to MHC molecules expressed on thymic epithelial
cells [39]. Specifically, DP thymocytes whose TCR fails to interact
with self MHC-peptide die in the cortex (by “neglect”) within 3
days of their generation. In contrast, cells which interact with self
MHC-peptide are rescued from cell death (“positive selection”)
and differentiate further. Thymocytes are selected on the basis
of their reactivity to MHC because mature T cells recognize a
complex of MHC and peptide determinants. Because of the high
polymorphism of MHC molecules (i.e. the large number of allelic

variants at loci encoding MHC-I and MHC-II molecules), TCR
specificities generated by Rag-mediated rearrangement must, at
the species level, be able to interact with a broad diversity of MHC
molecules. Consequently, at the individual level, most TCRs bind
poorly, if at all, to the particular MHC molecules expressed by the
thymic cortex, and are therefore “useless.” The function of posi-
tive selection is to select those TCRs with sufficient affinity for self
MHC. Of note, for conventional thymocytes, neither positive
selection per se, nor the differentiation events that accompany or
follow it, are associated with cell proliferation. Thus, these thymo-
cytes undergo no intrathymic division after the β-selection-
associated proliferative boost.
On the other side of the spectrum, a fraction of TCRs will
strongly bind self MHC-peptide complexes, such that their expres-
sion on mature T cells would potentially lead to autoimmune dis-
ease. The “objective” of thymic selection is to eliminate these two
extremes, and only allow cells in the middle of the range to differ-
entiate into mature T cells [39, 40].
2.5 Signaling Positive selection, strictly speaking the rescue from programmed
Positive Selection cell death, results from TCR signaling and involves increased
expression of anti-apoptotic molecules Bcl2 and Mcl1 [2]. It is
associated with differentiation events that quickly follow or coin-
cide with rescue from cell death. These first include the cessation
of TCRα gene rearrangement, at least in part by silencing expres-
sion of Rag1 and Rag2 genes. Positive selection is also accompa-
nied by the up-regulation of CCR7, a receptor for CCL19 and
CCL21 chemokines, both expressed in the thymic medulla but not
in the cortex. Together with the converse change in the expression
of CXCR4, whose ligand (CXCL12, also called SDF1) is preferen-
tially expressed in the cortex, CCR7 expression results in the
migration of thymocytes from the cortex to the medulla [41].
A more distant consequence of TCR signaling is the expression or
IL-7Rα, an event essential to T cell differentiation because it
restores IL-7 responsiveness, which is essential for long-term sur-
vival of mature T cells [42]. In addition, IL-7 is important for the
proper development of CD8+-lineage T cells [43, 44].
How such changes in gene expression are related to positively
selecting TCR signals is not yet fully understood [45]. In thymo-
cytes as in T cells, TCR signaling activates a variety of intracellular
signal transduction cascades, including the Ras-Erk kinase and cal-
cium pathway, both important for positive selection. Nuclear targets
of TCR signals include members of the Egr, AP-1, Ets and NFAT
families, as well as inhibitors (Id2 and I3) of E-box binding proteins
[46]. Because of genetic redundancy, the role of many of these fac-
tors and pathways remains to be fully elucidated, and it is likely that
other, yet to be discovered, factors contribute to positive selection.
2.6 Central The other side of thymic selection is the elimination (or inactivation)
Tolerance of thymocytes carrying TCRs with high avidity for self ligands.
and Negative Selection There is indeed strong genetic evidence for a thymic role in estab-
lishing immune tolerance [47]. Such “central” tolerance (i.e.
established in the thymus) is now understood as having two com-
ponents. The first prevents self-reactive cells from differentiating
into mature T cells, by causing their TCR-induced cell death in the
thymus (a deletion process referred to “negative selection”). The
second causes self-reactive cells to adopt a “regulatory” (Treg)
fate, characterized by and requiring expression of the transcription
factor Foxp3, which endows the cells with suppressive functions
[48, 49]. Both mechanisms appear important for central tolerance,
although their respective contribution is not yet fully understood
[50]. Analyses in mice suggest that deletion mechanisms do not
result in complete elimination of self-reactive cells [51], although
such studies have so far been carried only in experimental models
with clonal or reduced-diversity repertoires, and future studies will
be needed to fully measure the contribution of deletion mecha-
nisms to central tolerance. The genetic evidence for the impor-
tance of Treg cells is compelling, both in mice, in which disruption
of Foxp3 causes a severe autoimmune and inflammatory disease,
and in humans in which Foxp3 mutations are the cause for a rare
but severe autoimmune disease of infancy and childhood called
IPEX syndrome (Immunodysregulation, Polyendocrinopathy, and
Enteropathy, X-linked) [52]. The importance of Treg cell genera-
tion in the thymus is independently underscored by mouse neona-
tal thymectomy experiments; when performed before the
appearance of Treg cells (i.e. by fourth day of age), thymectomy
results in multiple autoimmune manifestations that are prevented
by adoptive transfer of Treg cells [48].
Interactions between thymocytes and medullary epithelial cells
are essential for both arms of central tolerance. That is at least in
part due to the expression, by medullary epithelial cells, of tissue-
specific antigens (epitomized by insulin, whose expression is other-
wise limited to pancreatic islets) [40]. Expression of tissue-specific
antigens in medullary epithelial cells requires the transcriptional
regulator Aire (autoimmune regulator) [47], a multifunctional pro-
tein whose dysfunction causes an autoimmune disease targeting the
endocrine system (Autoimmune Polyendocrinopathy-Candidiasis-
Ectodermal Dystrophy, APECED).
2.7 Positive vs. Because both positive and negative selection depend on TCR sig-
Negative Selection naling, perhaps no question has haunted the field of T cell develop-
Signals ment more than what distinguishes signals determining either
and Mechanisms outcome. While there is no definitive or simple answer to this ques-
tion, we have tried to summarize below current concepts and
perspectives.
MHC-peptide complexes that engage the TCR with high avid-

ity in the thymus promote negative (i.e. cell death) instead of
positive selection [39]. Experiments in a specific TCR transgenic
system (see Chapter 3) suggest that intrinsic attributes of the TCR–
ligand interaction (i.e. independent from the density of each part-
ner) are the critical factor determining the positive vs. negative
selection outcome; in other words, even minute amounts of a
high-affinity ligand will trigger deletion. However, within a very
narrow intermediate-affinity range, increasing ligand density can
switch the outcome of signaling from positive to negative selection
[53]. There is also evidence that high-avidity ligands promote Treg
differentiation [54]. Deciphering the signals that drive a develop-
ing thymocyte toward deletion vs. Treg differentiation is an area of
intense investigation.
Multiple studies have examined the consequences of such avid-
ity differences on thymocyte signaling, and observed differences in
the activation of Erk and Erk-related MAP kinases, whether in the
nature of the kinases being activated, or in the kinetics or topogra-
phy of their activation [45, 55]. In particular, there is evidence that
the conventional MAP kinases Erk1 and Erk2 are required for pos-
itive but not negative selection [56]. Among possible nuclear
effectors of negative selection, recent findings provide new evi-
dence for the involvement of transcription factors of the Nur77
family [57, 58], and suggest that the transcription factor Schnurri-2
contributes to distinguish positive to negative selection signals
[59]. Death effector mechanisms that drive thymocyte deletion
involve the pro-apoptotic Bcl2-family protein Bim, whose activity
is in part redundant with that of the related protein Puma [60].
Inactivation of both genes impairs negative selection and causes
autoimmune manifestations, consistent with the idea of impaired
central tolerance.
Whether negative selection happens in a specific thymic com-
partment has long been debated. It was initially envisioned that,
unlike positive selection which occurs in the cortex, negative selec-
tion takes place in the medulla. This idea is in line with the fact that
medullary epithelial cells express tissue-specific self antigens which
contribute to negative selection. However, recent studies have high-
lighted the possibility of cortical deletion of DP thymocytes. The key
observation is that inactivation of the pro-apoptotic molecule Bim
causes the appearance of large numbers of DP cells with high expres-
sion of genes involved in negative selection (including Nur77) [61];
this suggests that such cells are normally targeted for deletion after
undergoing strong TCR signaling at the DP stage (i.e. in the cortex)
and rescued from cell death upon Bim disruption.
2.8 Differentiation The selection of conventional TCRαβ thymocytes is accompanied

of αβ Lineage Cells by, and possibly mechanistically coupled to the differentiation into
and Thymic Egress either CD4+ or CD8+ lineages, and pre-programming for cytotoxic
or helper differentiation [62, 63]. Recent work has identified

transcription factors important for such lineage differentiation,
including the zinc finger transcription factor Thpok for the CD4
lineage, and the Runx protein Runx3 for the CD8 lineage. These
factors appear to work in a dual negative regulatory loop that
ensures their mutually exclusive expression and commits develop-
ing cells to express either CD4 or CD8 but not both. Other tran-
scription factors, including Gata3, Tox and E-box binding proteins
E2A and HEB, are required for the differentiation of CD4+ T cells,
whereas IRF1, the cytokine signal transducer Stat5 (redundantly
with the related protein Stat6) and Ets1 are important for the dif-
ferentiation of CD8+ T cells [63, 64]. Which signals dictate expres-
sion of these transcription factors, and therefore determines lineage
differentiation, is not yet fully understood.
The last checkpoint in the development of αβ T cell precursors
controls their egress from the thymus. Sphingosine 1 phosphate
(S1P) and its receptor S1pr1 are essential for thymocytes to leave
the thymus for the bloodstream [65]. The transcription factors Klf2
and Foxo1 are important for expression of S1pr1 and other markers
of mature thymocytes [66, 67]; how their expression is determined
by intrathymic ligands remains to be fully understood.
2.9 High-Avidity Even though CD4 and CD8-lineage conventional thymocytes are
Cells and Acquisition functionally “pre-programmed” for helper or cytotoxic differentia-
of Effector Properties tion, they exit the thymus as resting cells, i.e. lacking effector func-
tions. They acquire such functions (e.g. expression of cytotoxic
enzymes or cytokines) only upon antigen stimulation in peripheral
lymphoid organs. In contrast, a small subset of cells become effec-
tors in the thymus [54]. This, to some extent, is the case of Treg
precursors displaying high reactivity against MHC-peptide com-
plexes, which start expressing the key effector transcription factor
Foxp3 in the thymus. However, specific thymocyte subsets acquire
actual effector properties, including NK T cells, characterized by
high-level expression of cytokines or cytotoxic enzymes, in response
to thymic ligands. As in the case of effector differentiation in the
periphery, acquisition of effector functions by thymocytes appears
to be associated with their proliferation. Unlike for conventional
thymocytes, which are selected by MHC-peptide complexes
expressed by the thymic epithelium, selection of and acquisition of
effector properties by such “innate immune” thymocytes involve
ligands (MHC-like molecules and costimulatory molecules)
expressed on DP thymocytes themselves [68].
2.10 Thymic Of all the non-T cell types of the thymus (collectively referred to as
Architecture the stroma), the epithelial component has attracted the most atten-
tion [13, 14]. Its critical role is demonstrated by the thymic aplasia
and massive disruption of T cell development in the nude mouse,
caused by a mutation in the gene encoding the transcription factor
Foxn1 that intrinsically disrupts thymic epithelial cell develop-

ment. Other transcription factors, including HoxA3, have since
been shown to be important for the development of the thymic
epithelial cells.
Medullary and cortical epithelial cells are functionally distinct
and express unique phenotypic markers. We previously mentioned
the role of the medullary epithelium in establishing central toler-
ance; indeed, impairing thymocyte migration to the medulla (by
disrupting the gene encoding CCR7) results in defective central
tolerance and autoimmune manifestations [69]. The last few years
have assigned key functions to the cortical epithelium. Its high-
level expression of Notch ligands is essential to the development of
early T cell precursors. Cortical epithelial cells express specific pro-
teases or proteasome components which differ from those pro-
duced by professional (bone marrow-derived) antigen-presenting
cells, generating unique peptide sets essential to the generation of
a proper T cell repertoire [70–72]. Last, chemokines secreted by
cortical epithelial cells are needed to attract circulating bone mar-
row precursors to the thymus [20]. The thymus cortex has a highly
organized architecture, with specific location for DN thymocytes
at distinct stages of their differentiation, and a complex network of
interactions between DP thymocytes and cortical cells. New
approaches, especially the use of intravital microscopy on thymic
“slices” (see Chapter 11) have started to unveil the underappreci-
ated dynamics of interactions between thymocytes and the thymic
stroma, and the correlations between thymocyte motion and sig-
naling events.
Although their role is not as well characterized, hematopoietic
cells residing in the thymus are important for T cell development.
In particular, cortical macrophages are thought to be critical to
eliminate dead DP thymocytes, whereas medullary dendritic cells
contribute to negative selection.
Mirroring the thymocyte dependence from the thymic epithe-
lium, the development of the thymic epithelium itself requires sig-
nals from thymocytes. The mechanistic bases for such “cross-talk”
have been elucidated in the differentiation of medullary epithelial
cells, whose survival and expression of Aire require ligation of
TNF-family receptors (RANK and TRAIL) by ligands expressed by
CD4 SP thymocytes [41]. This requirement explains the original
observations that the thymic medulla fails to develop in mice car-
rying mutations that prevent the development of mature thymo-
cytes [73, 74]. More recent work has hinted at the possibility that
the differentiation of the cortical epithelium be dependent on DP
thymocytes, through an indirect mechanism that would involve
IL-22 production by epithelial cells [75]. Deciphering the signals
that control the development and homeostasis of thymic epithelial
cells, and the extent of their “cross-talk” with the thymocyte
compartment, are essential goals for strategies aiming a restoring
thymic function in the elderly or in patients whose immune system

has been compromised by cancer radiation or chemotherapy.
3 Tracking Developing Thymocytes
Investigating T cell development typically involves a variety of

approaches, many of them discussed in this book. Among these,
flow cytometry occupies a special place: because of its irreplaceable
contribution to define thymic cell subsets and thymocyte develop-
mental stages, flow cytometric analysis is typically an indispensable
part of any experimental strategy. Thus, the following pages include
a summary of flow-based approaches to explore thymocyte devel-
opment; the reader is referred to Chapters 4 and 20 (for human
cells) for detailed procedures. The simplest and by far most widely
used relies on expression of CD4 and CD8 molecules and distin-
guishes DN, DP, and CD4 and CD8 SP cells. This approach is
useful to assess the overall balance of αβ lineage populations (most
of which express CD4, CD8 or both), but has limited resolution to
dissect selection events, and is not useful to evaluate early differen-
tiation or γδ T cells development.
3.1 Early T Cell Although early thymocytes express neither CD4 nor CD8, it is
Development important to realize that they only form a fraction of DN thymo-
cytes. Thus, identifying early progenitor cells requires additional
gating strategies to eliminate the large number of “other” DN cells
(e.g. γδ T cells, iNK T cells, B cells, etc.). This is easily done by
combining antibodies reacting against each specific lineage (e.g.
TCRγδ, CD19 as a B cell marker) labeled with the same fluoro-
chrome, so that all cells expressing any of these markers can be
electronically excluded from further analyses. This strategy is of
broad applicability to analyze small size subsets as it excludes
aggregates that can form during cell preparation.
DN thymocytes are conventionally subdivided into four sub-
sets (DN1-4) based on the expression of CD44 and CD25 (Figs. 2
and 3), although CD117 (cKit) seems to be a more specific marker
than CD44. Various strategies have been proposed to dissect the
earliest (DN1) subset and identify early T progenitors (ETP)
within the so-called DN1a compartment (CD24− CD117hi) [18].
There is only imperfect matching between such flow-cytometry
“staging” and developmental checkpoints: notably, the T lineage
commitment checkpoint occurs as cells down-regulate CD117.
Similarly, β-selection occurs in DN3 cells, and thymocytes that
have undergone β-selection can usually be identified on the basis of
their greater size, expression of intracellular TCRβ molecules, or
expression of surface CD27 or CD28 (Fig. 3).
Myeloid, NK, TCRβ, TCRγδ

and other lineages rearrangement
Thymic T lineage
γδ
entry commitment
HSC DN1a DN1b DN2a DN2b DN3a
αβ
cKit (CD117) +++ +++ +++ +/– –
CD44 + + + + –
CD25 – – +++ +++ +++
CD24 – +++ ++++ ++++ ++++
Fig. 3 Early T cell development. Early T cell development stages are schematically depicted, using expression
of CD25 and CD44 to define DN1 (CD25−CD44+), DN2 (CD25+CD44+) and DN3 (CD25+CD44−) stages. The DN2
stage is further divided into pre- and post-commitment DN2a and DN2b stages, which can be distinguished
by the reduction in CD117 (cKit) expression. At the DN3 stage, successful completion of TCRβ or TCRγ rear-
rangement distinguishes DN3b cells (not depicted), which can be distinguished from DN3a cells by several
markers (see Fig. 4)
3.2 αβ T Cell Surface markers useful to analyze the development of conventional

Development αβ lineage cells are summarized in Figs. 3 and 4 and in Table 1.
Although the differentiation of CD4+ and CD8+ lineages is a key
outcome of αβ lineage development, expression of CD4 and CD8
molecules is paradoxically of limited interest to distinguish devel-
opmental intermediates during the selection of αβ T cells. This is
because intrathymic TCR signaling, whether MHC-I or MHC-II
induced, can repress expression of Cd8 genes in thymocytes [76].
While such repression is transient only in MHC-I-specific cells
(which eventually re-express CD8 and become CD8 T cells), it
results in both MHC-I and MHC-II-restricted thymocytes acquir-
ing a CD4+CD8int “transitional” surface phenotype. Three surface
markers, TCR complexes (assessed by staining for TCRβ or CD3ε),
CD69 and CD24, provide the best strategies to distinguish selec-
tion intermediates. TCRβ expression is the most reliable surface
marker for selection, whereas CD24 assesses cell maturation. In
most circumstances, CD69 expression provides a faithful readout
of TCR signaling in thymocytes (so that CD69hi cells are those
undergoing TCR signaling). Consequently, a staining protocol
associating CD4, CD8, TCRβ, CD69 and CD24 distinguishes key
cell subsets undergoing selection. It has the additional advantage
of separating an immature CD8 SP subset, which in most mouse
strains contains cells transitioning from the DN4 to DP stage, and
appear as TCRlo CD69− CD24hi. Other markers can be used to
distinguish subsets of SP cells (Table 1). Note that none of these
strategies reliably distinguishes MHC-I from MHC-II-signaled
thymocytes before they have terminated expression of CD4 or
CD4 SP
Treg
β−selection TCRα rearrangement

CD4 SP
MHC II
DN3a DN3b DN4 ISP DP
MHC I
CD8 SP
specific
ligands
(e.g.CD1d)
innate effector
T cells
CD25 ++++ +++ – – – (e.g. iNK T)
CD5 – – lo lo int
CD4 – – – -/lo +++
CD8 – – – ++ +++
CD24 ++++ ++++ ++++ ++++ +++
CD27 – + + + +
CD28 – ++ ++ ++ ++lo
FSC lo hi hi hi
i.c. TCRβ – + + + +
surface TCRβ – – – –/lo lo/int
Fig. 4 αβ T cell development. The key developmental stages of the αβ lineage are schematically depicted.
Commitment to the αβ lineage occurs during β-selection, following successful rearrangement of TCRβ in DN3
cells, and is accompanied by increased cell size (evaluated by flow cytometry on Forward light Scatter [FSC]),
and changes in expression of CD27 and CD28. Subsequent developmental stages and their defining markers
are schematically depicted
Table 1
Surface markers in the study of positive selection
Marker Pre-selection DP Signaled DP Selected SP Mature SP Comments

CD5 +/++ +++ ++++ ++++
CD69 − ++ +++ −/+
TCRβ, CD3ε −/+ +++ +++ +++
IL-7Rα − − +/− +/++
CCR7 − −/+ ++ ++
CXCR4 ++ + − −
CD4 +++ +++ variable +++ or − Depending on
MHC restriction
CD8 ++++ ++++ variable ++++ or − Depending on
MHC restriction
MHC-I ++ − − ++
Qa-2 − − − ++ Strain-specific
CD8 (as mentioned, the TCRhi CD69hi CD4+CD8int subset

includes both MHC-I and MHC-II-restricted cells, and therefore
both CD4 and CD8 T cell precursors) or identifies cells targeted
for negative selection.
4 Concluding Remarks
This brief overview highlights the tremendous progress made over

the past 30 years in elucidating the functions of the thymus and its
role in generating and shaping the T cell repertoire. The combina-
tion of improved investigative approaches, an ever increasing avail-
ability of marker-specific antibodies and the extraordinary versatility
of mouse genetics has helped the field build a detailed map to T
cell development, delineated essential cell–cell interaction and
intracellular signaling pathways. Efforts are underway to fill the
many remaining gaps, and to decipher transcriptional circuits
directing cell differentiation. Future investigations will build on
these advances, and tackle key challenges lying ahead, including
understanding the basis for thymic involution (and elaborating
strategies for thymic regeneration), and further analyses of reper-
toire generation.
References
1. Kisielow P, Von Boehmer H (1995) 8. Singer A, Bosselut R (2004) CD4/CD8 core-
Development and selection of T cells: facts and ceptors in thymocyte development, selection, and
puzzles. Adv Immunol 58:87–209 lineage commitment: analysis of the CD4/CD8
2. Carpenter AC, Bosselut R (2010) Decision lineage decision. Adv Immunol 83:91–131
checkpoints in the thymus. Nat Immunol 11: 9. Bendelac A, Savage PB, Teyton L (2007) The
666–673 biology of NKT cells. Annu Rev Immunol
3. Wucherpfennig KW, Gagnon E, Call MJ, 25:297–336
Huseby ES, Call ME (2010) Structural biology 10. Le Bourhis L, Guerri L, Dusseaux M, Martin E,
of the T-cell receptor: insights into receptor Soudais C, Lantz O (2011) Mucosal-associated
assembly, ligand recognition, and initiation of invariant T cells: unconventional development
signaling. Cold Spring Harb Perspect Biol and function. Trends Immunol 32:212–218
2:a005140 11. Gold MC, Lewinsohn DM (2013)
4. Samelson LE (2002) Signal transduction Co-dependents: MR1-restricted MAIT cells
mediated by the T cell antigen receptor: the and their antimicrobial function. Nat Rev
role of adapter proteins. Annu Rev Immunol Microbiol 11:14–19
20:371–394 12. van Wijk F, Cheroutre H (2009) Intestinal T
5. Reiner S (2008) Peripheral T lymphocyte cells: facing the mucosal immune dilemma
responses and function. In: Paul WE (ed) with synergy and diversity. Semin Immunol
Fundamental immunology. Lippincott 21:130–138
Williams & Wilkins, Philadelphia, pp 407–425 13. Gordon J, Manley NR (2011) Mechanisms of
6. Smith-Garvin JE, Koretzky GA, Jordan MS thymus organogenesis and morphogenesis.
(2009) T cell activation. Annu Rev Immunol Development 138:3865–3878
27:591–619 14. Nitta T, Murata S, Ueno T, Tanaka K,
7. Mazza C, Malissen B (2007) What guides Takahama Y (2008) Thymic microenviron-
MHC-restricted TCR recognition? Semin ments for T-cell repertoire formation. Adv
Immunol 19:225–235 Immunol 99:59–94
15. Rossi SW, Jenkinson WE, Anderson G, 29. Gonzalez-Garcia S, Garcia-Peydro M, Alcain J,
Jenkinson EJ (2006) Clonal analysis reveals a Toribio ML (2012) Notch1 and IL-7 receptor
common progenitor for thymic cortical and signalling in early T-cell development and leukae-
medullary epithelium. Nature 441:988–991 mia. Curr Top Microbiol Immunol 360:47–73
16. Bleul CC, Corbeaux T, Reuter A, Fisch P, 30. Jung D, Alt FW (2004) Unraveling V(D)J
Monting JS, Boehm T (2006) Formation of a recombination; insights into gene regulation.
functional thymus initiated by a postnatal epi- Cell 116:299–311
thelial progenitor cell. Nature 441:992–996 31. von Boehmer H, Aifantis I, Azogui O, Feinberg
17. Yang Q, Jeremiah Bell J, Bhandoola A (2010) J, Saint-Ruf C, Zober C, Garcia C, Buer J
T-cell lineage determination. Immunol Rev (1998) Crucial function of the pre-T-cell
238:12–22 receptor (TCR) in TCR beta selection, TCR
18. Bhandoola A, von Boehmer H, Petrie HT, beta allelic exclusion and alpha beta versus
Zuniga-Pflucker JC (2007) Commitment and gamma delta lineage commitment. Immunol
developmental potential of extrathymic and Rev 165:111–119
intrathymic T cell precursors: plenty to choose 32. von Boehmer H, Aifantis I, Feinberg J, Lechner
from. Immunity 26:678–689 O, Saint-Ruf C, Walter U, Buer J, Azogui O
19. Rothenberg EV (2011) T cell lineage commit- (1999) Pleiotropic changes controlled by the pre-
ment: identity and renunciation. J Immunol T-cell receptor. Curr Opin Immunol 11:135–142
186:6649–6655 33. Hayes SM, Love PE (2006) Strength of signal:
20. Zhang SL, Bhandoola A (2014) Trafficking to a fundamental mechanism for cell fate specifi-
the thymus. Curr Top Microbiol Immunol cation. Immunol Rev 209:170–175
373:87–111 34. Lauritsen JP, Haks MC, Lefebvre JM, Kappes
21. Hozumi K, Mailhos C, Negishi N, Hirano K, DJ, Wiest DL (2006) Recent insights into the
Yahata T, Ando K, Zuklys S, Hollander GA, signals that control alphabeta/gammadelta-
Shima DT, Habu S (2008) Delta-like 4 is indis- lineage fate. Immunol Rev 209:176–190
pensable in thymic environment specific for T 35. Kreslavsky T, Gleimer M, von Boehmer H
cell development. J Exp Med 205:2507–2513 (2010) Alphabeta versus gammadelta lineage
22. Koch U, Fiorini E, Benedito R, Besseyrias V, choice at the first TCR-controlled checkpoint.
Schuster-Gossler K, Pierres M, Manley NR, Curr Opin Immunol 22(2):185–192
Duarte A, Macdonald HR, Radtke F (2008) 36. Mingueneau M, Kreslavsky T, Gray D, Heng
Delta-like 4 is the essential, nonredundant T, Cruse R, Ericson J, Bendall S, Spitzer MH,
ligand for Notch1 during thymic T cell lineage Nolan GP, Kobayashi K, von Boehmer H,
commitment. J Exp Med 205:2515–2523 Mathis D, Benoist C, Best AJ, Knell J, Goldrath
23. Pui JC, Allman D, Xu L, DeRocco S, Karnell A, Jojic V, Koller D, Shay T, Regev A, Cohen
FG, Bakkour S, Lee JY, Kadesch T, Hardy RR, N, Brennan P, Brenner M, Kim F, Rao TN,
Aster JC, Pear WS (1999) Notch1 expression Wagers A, Heng T, Ericson J, Rothamel K,
in early lymphopoiesis influences B versus T Ortiz-Lopez A, Mathis D, Benoist C, Bezman
lineage determination. Immunity 11:299–308 NA, Sun JC, Min-Oo G, Kim CC, Lanier LL,
24. Radtke F, Wilson A, Stark G, Bauer M, van Miller J, Brown B, Merad M, Gautier EL,
Meerwijk J, MacDonald HR, Aguet M (1999) Jakubzick C, Randolph GJ, Monach P, Blair
Deficient T cell fate specification in mice with DA, Dustin ML, Shinton SA, Hardy RR,
an induced inactivation of Notch1. Immunity Laidlaw D, Collins J, Gazit R, Rossi DJ,
10:547–558 Malhotra N, Sylvia K, Kang J, Kreslavsky T,
Fletcher A, Elpek K, Bellemare-Pelletier A,
25. Rothenberg EV (2012) Transcriptional drivers Malhotra D, Turley S (2013) The transcrip-
of the T-cell lineage program. Curr Opin tional landscape of alphabeta T cell differentia-
Immunol 24:132–138 tion. Nat Immunol 14:619–632
26. Rothenberg EV, Moore JE, Yui MA (2008) 37. Williams JA, Hathcock KS, Klug D, Harada Y,
Launching the T-cell-lineage developmental Choudhury B, Allison JP, Abe R, Hodes RJ
programme. Nat Rev Immunol 8:9–21 (2005) Regulated costimulation in the thymus
27. Ho IC, Tai TS, Pai SY (2009) GATA3 and the is critical for T cell development: dysregulated
T-cell lineage: essential functions before and CD28 costimulation can bypass the pre-TCR
after T-helper-2-cell differentiation. Nat Rev checkpoint. J Immunol 175:4199–4207
Immunol 9:125–135 38. Taghon T, Yui MA, Pant R, Diamond RA,
28. Avram D, Califano D (2014) The multifaceted Rothenberg EV (2006) Developmental and
roles of Bcl11b in thymic and peripheral T molecular characterization of emerging beta-
cells: impact on immune diseases. J Immunol and gammadelta-selected pre-T cells in the
193(5):2059–2065 adult mouse thymus. Immunity 24:53–64
39. Starr TK, Jameson SC, Hogquist KA (2003) 53. Daniels MA, Teixeiro E, Gill J, Hausmann B,
Positive and negative selection of T cells. Annu Roubaty D, Holmberg K, Werlen G, Hollander
Rev Immunol 21:139–176 GA, Gascoigne NR, Palmer E (2006) Thymic
40. Klein L, Kyewski B, Allen PM, Hogquist KA selection threshold defined by compartmental-
(2014) Positive and negative selection of the T ization of Ras/MAPK signalling. Nature
cell repertoire: what thymocytes see (and don’t 444:724–729
see). Nat Rev Immunol 14:377–391 54. Stritesky GL, Jameson SC, Hogquist KA
41. Nitta T, Ohigashi I, Nakagawa Y, Takahama Y (2012) Selection of self-reactive T cells in the
(2011) Cytokine crosstalk for thymic medulla thymus. Annu Rev Immunol 30:95–114
formation. Curr Opin Immunol 23:190–197 55. Fu G, Rybakin V, Brzostek J, Paster W, Acuto
42. Boyman O, Krieg C, Homann D, Sprent J O, Gascoigne NR (2014) Fine-tuning T cell
(2012) Homeostatic maintenance of T cells receptor signaling to control T cell develop-
and natural killer cells. Cell Mol Life Sci ment. Trends Immunol 35:311–318
69:1597–1608 56. McGargill MA, Ch’en IL, Katayama CD, Pages
43. McCaughtry TM, Etzensperger R, Alag A, Tai G, Pouyssegur J, Hedrick SM (2009) Cutting
X, Kurtulus S, Park JH, Grinberg A, Love P, edge: extracellular signal-related kinase is not
Feigenbaum L, Erman B, Singer A (2012) required for negative selection of developing T
Conditional deletion of cytokine receptor cells. J Immunol 183:4838–4842
chains reveals that IL-7 and IL-15 specify CD8 57. Fassett MS, Jiang W, D’Alise AM, Mathis D,
cytotoxic lineage fate in the thymus. J Exp Med Benoist C (2012) Nuclear receptor Nr4a1
209:2263–2276 modulates both regulatory T-cell (Treg) differ-
44. Tani-ichi S, Shimba A, Wagatsuma K, Miyachi entiation and clonal deletion. Proc Natl Acad
H, Kitano S, Imai K, Hara T, Ikuta K (2013) Sci U S A 109:3891–3896
Interleukin-7 receptor controls development 58. Sekiya T, Kashiwagi I, Yoshida R, Fukaya T,
and maturation of late stages of thymocyte sub- Morita R, Kimura A, Ichinose H, Metzger D,
populations. Proc Natl Acad Sci U S A 110: Chambon P, Yoshimura A (2013) Nr4a recep-
612–617 tors are essential for thymic regulatory T cell
45. Gascoigne NR, Palmer E (2011) Signaling in thy- development and immune homeostasis. Nat
mic selection. Curr Opin Immunol 23:207–212 Immunol 14:230–237
46. Wang L, Xiong Y, Bosselut R (2010) Tenuous 59. Staton TL, Lazarevic V, Jones DC, Lanser AJ,
paths in unexplored territory: from T cell Takagi T, Ishii S, Glimcher LH (2011)
receptor signaling to effector gene expression Dampening of death pathways by schnurri-2 is
during thymocyte selection. Semin Immunol essential for T-cell development. Nature
22:294–302 472:105–109
47. Mathis D, Benoist C (2009) Aire. Annu Rev 60. Gray DH, Kupresanin F, Berzins SP, Herold
Immunol 27:287–312 MJ, O’Reilly LA, Bouillet P, Strasser A (2012)
48. Sakaguchi S, Yamaguchi T, Nomura T, Ono M The BH3-only proteins Bim and Puma cooper-
(2008) Regulatory T cells and immune tolerate to impose deletional tolerance of organ-
ance. Cell 133:775–787 specific antigens. Immunity 37:451–462
49. Josefowicz SZ, Lu LF, Rudensky AY (2012) 61. Stritesky GL, Xing Y, Erickson JR, Kalekar LA,
Regulatory T cells: mechanisms of differentia- Wang X, Mueller DL, Jameson SC, Hogquist
tion and function. Annu Rev Immunol 30: KA (2013) Murine thymic selection quantified
531–564 using a unique method to capture deleted T
cells. Proc Natl Acad Sci U S A 110:4679–4684
50. Guerau-de-Arellano M, Martinic M, Benoist
C, Mathis D (2009) Neonatal tolerance 62. Wang L, Bosselut R (2009) CD4-CD8 lineage
revisited: a perinatal window for Aire control of differentiation: Thpok-ing into the nucleus.
autoimmunity. J Exp Med 206:1245–1252 J Immunol 183:2903–2910
51. Bouneaud C, Kourilsky P, Bousso P (2000) 63. Xiong Y, Bosselut R (2012) CD4-CD8 differ-
Impact of negative selection on the T cell rep- entiation in the thymus: connecting circuits
ertoire reactive to a self-peptide: a large frac- and building memories. Curr Opin Immunol
tion of T cell clones escapes clonal deletion. 24:139–145
Immunity 13:829–840 64. Xiong Y, Bosselut R (2011) The enigma of
52. Verbsky JW, Chatila TA (2013) Immune dys- CD4-lineage specification. Eur J Immunol
regulation, polyendocrinopathy, enteropathy, 41:568–574
X-linked (IPEX) and IPEX-related disorders: 65. Cyster JG, Schwab SR (2012) Sphingosine-1-
an evolving web of heritable autoimmune dis- phosphate and lymphocyte egress from lym-
eases. Curr Opin Pediatr 25:708–714 phoid organs. Annu Rev Immunol 30:69–94
66. Hart GT, Hogquist KA, Jameson SC (2012) Takahama Y (2010) Thymoproteasome shapes
Kruppel-like factors in lymphocyte biology. J immunocompetent repertoire of CD8+ T cells.
Immunol 188:521–526 Immunity 32:29–40
67. Hedrick SM, Hess Michelini R, Doedens AL, 73. Shores EW, Van Ewijk W, Singer A (1991)
Goldrath AW, Stone EL (2012) FOXO tran- Disorganization and restoration of thymic
scription factors throughout T cell biology. Nat medullary epithelial cells in T cell receptor-
Rev Immunol 12:649–661 negative scid mice: evidence that receptor-
68. Cannons JL, Tangye SG, Schwartzberg PL bearing lymphocytes influence maturation of
(2011) SLAM family receptors and SAP the thymic microenvironment. Eur J Immunol
adaptors in immunity. Annu Rev Immunol 21:1657–1661
29:665–705 74. Philpott KL, Viney JL, Kay G, Rastan S,
69. Nitta T, Nitta S, Lei Y, Lipp M, Takahama Y Gardiner EM, Chae S, Hayday AC, Owen MJ
(2009) CCR7-mediated migration of develop- (1992) Lymphoid development in mice con-
ing thymocytes to the medulla is essential for genitally lacking T cell receptor alpha beta-
negative selection to tissue-restricted antigens. expressing cells. Science 256:1448–1452
Proc Natl Acad Sci U S A 106:17129–17133 75. Dudakov JA, Hanash AM, Jenq RR, Young
70. Nakagawa T, Roth W, Wong P, Nelson A, Farr LF, Ghosh A, Singer NV, West ML, Smith
A, Deussing J, Villadangos JA, Ploegh H, OM, Holland AM, Tsai JJ, Boyd RL, van den
Peters C, Rudensky AY (1998) Cathepsin L: Brink MR (2012) Interleukin-22 drives
critical role in Ii degradation and CD4 T cell endogenous thymic regeneration in mice.
selection in the thymus. Science 280:450–453 Science 336:91–95
71. Klein L, Hinterberger M, Wirnsberger G, 76. Bosselut R, Guinter TI, Sharrow SO, Singer A
Kyewski B (2009) Antigen presentation in the (2003) Unraveling a revealing paradox: why
thymus for positive selection and central toler- major histocompatibility complex I-signaled
ance induction. Nat Rev Immunol 9:833–844 thymocytes “Paradoxically” appear as CD4+8lo
72. Nitta T, Murata S, Sasaki K, Fujii H, Ripen transitional cells during positive selection of
AM, Ishimaru N, Koyasu S, Tanaka K, CD8+ T cells. J Exp Med 197:1709–1719
Chapter 2
Development of γδ T Cells, the Special-Force Soldiers

of the Immune System
David L. Wiest
Abstract
While the functions of αβ T cells in host resistance to pathogen infection are understood in far more detail
than those of γδ lineage T cells, γδ T cells perform critical, essential functions during immune responses
that cannot be compensated by αβ T cells. Accordingly, it is essential to understand how the development
of γδ T cells is controlled so that their generation and function might be manipulated in future for thera-
peutic benefit. This introductory chapter will cover the basic processes that underlie γδ T cell development
in the thymus, as well as the current understanding of how they are controlled.
Key words Gamma-delta T cells, Gamma-delta TCR, V gamma elements, Lineage commitment,
Effector fate
1 Introduction
There are two major T lineages marked by the T cell antigen receptor
(TCR) complexes they express, αβ and γδ. By analogy to the military,
αβ lineage T cells are like conventional soldiers found primarily in
lymphoid organs, which can be thought of as military bases where the
staging of military operations occur. This is consistent with the 7-day
delay required for αβ T cells to mount a primary response to an infec-
tion and travel to the site of the battle. In contrast, γδ T cells make up
a small proportion of T cells in the peripheral lymphoid organs, and
instead predominate in the epithelial tissues that form the inner and
outer surfaces of the body [1–3]. Accordingly, γδ cells are much more
like special-force soldiers found primarily in the “field,” patrolling epi-
thelial barriers and possessing rapid-strike capabilities that do not
require priming. Indeed, γδ cells have been implicated in stress-sur-
veillance by rapidly responding to stress-induced proteins on epithelial
cells (e.g., Rae1 and H60), which is more consistent with an innate-
like mode of function [3]. Nevertheless, γδ T cells are also capable of
staging delayed, more classical adaptive-type responses following
immunization, including the mounting of recall responses consistent
23
24 David L. Wiest
with the generation of memory [4]. As such, γδ T cells combine attri-

butes of adaptive immunity, encoded in their TCR complexes, with
rapid, innate-like capabilities linked to the initiation phase of immune
responses. Although the precise role of γδ T cells in immune responses
remains unclear at present, these cells perform functions that are at
least partially distinct from those of αβ T cells. Indeed, certain bacterial
infections (e.g., Nocardia asteroides) that are normally cleared in wild
type mice are rapidly fatal in mice lacking γδ T cells [5]. Likewise,
resistance of neonates to parasitic infection is critically dependent
upon γδ T cell function [6]. γδ T cells have also been implicated in the
preservation of epithelial barriers and in eradication of cutaneous
malignancies [7–9]. Efforts to exploit the function of γδ T cells thera-
peutically in human disease are now under investigation, as clinical
grade agonists for the human Vγ9Vδ2 γδ T cells are currently being
tested for various infectious diseases and in cancer [10, 11]. In con-
trast to their beneficial effects, γδ T cells have also been found to
contribute to pathologies. IL-17-producing γδ T cells have been
implicated in the pathology of psoriasis [12]. IL-17-producing γδ T
cells have also been found at the borders between colon cancer lesions
and surrounding normal tissue and have been implicated in disease
progression through the recruitment of myeloid-derived suppressor
cells [13]. Nevertheless, despite the growing appreciation of the
importance of γδ T cell function in disease resistance and pathogene-
sis, many important questions regarding their generation and function
remain unanswered.
2 Antigen Recognition by γδ T Cells
While αβ T cells recognize and respond to proteolytically derived

peptide ligands in the context of MHC class I and II, γδ T cells do
not recognize ligand in an MHC-restricted manner and instead rec-
ognize a far more diverse collection of intact, unprocessed ligands
[14]. These include non-classical MHC molecules, heat shock pro-
teins, lipids and stress-induced molecules. In their recognition of
unprocessed, intact ligands, γδ T cells are quite similar to immuno-
globulins (Ig). Consistent with this notion, as is true for Ig, the
lengths of the CDR3 regions of the Vγ and Vδ subunits are quite
diverse, as is expected for a structure that is not constrained in rec-
ognizing cargo presented by a specific presenting element [15].
The ability to recognize intact antigens confers upon γδ T cells
greater flexibility of function.
3 Links Between Vγ Usage and Behavior
γδ T cell development is initiated at the time of seeding of the fetal

thymus by progenitors from fetal liver at embryonic day 13 (E13).
γδ cells represent about 1–2 % of thymocytes (~1000 γδ TCR+
Development of γδ T Cells, the Special-Force Soldiers of the Immune System 25
T cells develop in waves.

Anatomic Uterus, Lung, Lung, Blood, Blood,
Skin
Location: Tongue Spleen, LN Spleen, LN
Raulet: V 3 V 4 V 2 V 1.1
Tonegawa: V 5 V 6 V 4 V 1
T cell development
Thymic
12 14 16 18 20 Birth
Days of Embryonic Development
* Development of the V 7 (Tonegawa)/V 5 (Raulet) subset, which
homes to the gut, is thought to occur extrathymically.
Fig. 1 Timing of developmental waves of Vγ subsets. Schematic of waves of

development of Vγ subsets and the anatomical locations to which they home.
The Tonegawa and Raulet nomenclature systems for the Vγ are indicated.
Development of the Vg7 (Tonegawa)/Vg5 (Raulet) subset, which homes to the
gut, is thought to occur extrathymically
cells per thymic lobe) at E14. Their frequency increases to approxi-

mately 10 % at E16, following which their representation begins to
fall due to the rapid expansion of αβ lineage CD4+CD8+ progeni-
tors. The absolute number of γδ TCR-expressing cells continues to
rise during fetal development, reaching more than 30,000 cells per
lobe at E18 and beyond [16]. γδ development during fetal life is
characterized by waves of progenitors defined by their Vγ usage
(Fig. 1). There are two Vγ nomenclature systems that are in use at
present, one developed by the Raulet lab and the other by the
Tonegawa lab (Fig. 1) [2, 17]. In this chapter, we will be employ-
ing the Raulet nomenclature. The first wave of γδ progenitors from
E13 to 15 is characterized by Vγ3 usage, and is followed succes-
sively by waves defined by Vγ4, Vγ2, and Vγ1.1 (Fig. 1). These
waves are significant as there is a strong correlation between these
waves of developing γδ T cells defined by Vγ usage and their resi-
dence in particular anatomical sites [18, 19] (Fig. 1); however, the
molecular basis for this linkage remains poorly understood. One
exception is the homing of Vγ3+ dendritic epidermal T cells
(DETC) to the epidermis. During selection in the thymus, DETC
progenitors upregulate the chemokine receptor, CCR10, which is
26 David L. Wiest
Table 1
Commercially available anti-V region antibodies to study
mouse γδ T cell subsets
TCR subunit Ab clone Source

Vγ1.1 2.11 Biolegend
Vγ1.1 + Vγ1.2 4B2.9 Biolegend
Vγ2 UC3-10A6 Biolegend/Becton Dickinson
Vγ3 536 Biolegend/Becton Dickinson
Vδ4 GL2 Ebiosciences/Becton Dickinson
Vδ6.3 8F4H7B7 Becton Dickinson
required for their trafficking to the epidermis [20]. Consequently,

it is possible that selection events in the thymus are responsible for
the induction of trafficking molecules that mediate homing of each
wave of γδ progenitors, expressing a particular Vγ subunit, to the
intended anatomical location (e.g., Vγ4 to the genital tract).
Nevertheless, this has not been established for any Vγ subset other
than the Vγ3+ DETC. An alternative explanation is that each par-
ticular Vγ subset expresses γδ TCR complexes with specificity for
ligands expressed in the anatomic location where they reside, and
it is this interaction that is responsible for their homing and reten-
tion. A number of commercially available anti-Vγ and Vδ antibody
reagents have been generated that can be used to monitor the
behavior of the aforementioned subsets (Table 1).
4 Development
Phenotypic characterization of γδ cells during fetal ontogeny: Despite

substantial effort, the only truly unique identifier of γδ T cell pro-
genitors is the γδ TCR itself. γδ lineage progenitors typically
remain CD4−CD8−, although some do acquire expression of
CD8 [21]. Immature γδ lineage progenitors are characterized by
being CD24+ and as they mature they increase expression of
CD45RB and downregulate the expression of CD24 [22, 23].
Upon downregulation of CD24, many γδ T cells in the thymus
have acquired the ability to secrete cytokines including interleukin-17
(IL-17) or interferon-γ (IFN-γ) [24]. Cells with the capability of
producing IFN-γ are typically CD27+, while those that produce
IL-17 are CD27− [25]. An impediment to studying γδ T cell
development is that some CD4−CD8− γδ TCR-expressing
progenitors have not irreversibly committed to the γδ fate, and

can fate switch to the αβ lineage and develop to the CD4+CD8+
stage upon removal from the selecting milieu in the thymus [26].
It should be noted that γδ T cells do not develop through a
CD4+CD8+ intermediate. Consequently, a critical need in the
field is a phenotypic marker that distinguishes CD4−CD8− γδ
TCR-expressing progenitors that have committed to the γδ fate,
from those yet to do so. Indeed, we have recently identified such
a marker, CD73. CD73-expressing CD4−CD8− γδTCR+ progen-
itors remain CD4−CD8− and committed to the γδ fate even upon
removal from the selecting milieu, whereas CD73− progenitors
retain the ability to switch to the αβ fate [24]. Other markers have
been proposed that do not appear to mark γδ lineage commit-
ment. Indeed, through Serial Analysis of Gene Expression (SAGE)
performed by the Hayday lab, a γδ-biased gene signature was
established; however, while this profile is linked to γδ function it
does not mark lineage commitment [27]. Likewise, the transcrip-
tion factor Sox13 has been reported to be highly enriched in some
γδ T cells. Nevertheless, it is now clear that this factor marks only
a subset of γδ lineage cells [28].
Cross-talk between αβ and γδ progenitors: γδ T cells influence the
development of αβ T cells and vice versa. γδ T cell development
precedes that of αβ T cell progenitors by a few days, as the first αβ
progenitors competent to undergo intrathymic selection do not
appear until ~E16. This delay is critical, as prior to this time, the
thymic medulla compartment has not yet emerged in a form capa-
ble of supporting negative selection of autoreactive αβ progenitors.
Importantly, γδ T cell progenitors play an important role in estab-
lishing this capability. Indeed, development of the Vγ3+ DETC
subset begins around E14 and these progenitors promote the gen-
eration of medullary thymic epithelial cells (mTEC) that express
AIRE, and are thus capable of ectopic expression of peripheral
antigens [29]. The ability of DETC progenitors to induce AIRE
expressing mTEC is mediated by the expression of Rank Ligand on
their surface. Lymphoid tissue inducer cells possess the same capa-
bility and also contribute to induction of AIRE expression by
mTEC. Conversely, αβ lineage progenitors also influence the
development of γδ T cells. CD4+8+ αβ lineage thymocytes trans-
condition developing γδ progenitors by presenting cell surface-
bound lymphotoxin β. This is not responsible for γδ lineage
commitment, but has been reported to influence their functional
competence [27, 30].
Control of αβ/γδ lineage commitment: Compelling evidence exists
indicating that the αβ and γδ T cell fates arise from a common pro-
genitor in the thymus, raising the question of how lineage commit-
ment is controlled [31, 32]. While the TCR complexes of these
progenitors, the pre-T cell receptor (pre-TCR) and γδ TCR for the
28 David L. Wiest
αβ and γδ lineages, respectively, certainly influence lineage choice,

the way that they do so remains somewhat controversial. Attempts
to explain the role of the TCR in αβ/γδ lineage commitment have
been distilled into two basic models, stochastic and instructional.
The stochastic model predicts that lineage fate is specified indepen-
dently of TCR expression and that TCR signals serve only to res-
cue viability of already committed progenitors, provided the TCR
isotype matches the preordained lineage fate [33]. Conversely, the
instructional model proposes that TCR signals direct uncommit-
ted precursors to adopt either the αβ or γδ fate [34]. That is, the
signals transduced through the pre-TCR or γδTCR actively specify
the αβ and γδ fates, respectively. These models share the basic idea
that the pre-TCR or γδTCR complexes transduce unique signals
inextricably linked to specification of the αβ and γδ fates, respec-
tively. However, these models are not adequate to explain the sta-
tus of TCR gene rearrangements in αβ and γδ lineage cells, nor do
they appropriately explain the lineage infidelity observed in TCR
transgenic and gene-targeted mice [21]. To address these inconsis-
tencies, a signal strength model was proposed which posits that
strong signaling through a TCR promotes adoption of the γδ lin-
eage, while weaker signals lead to adoption of the αβ lineage, irre-
spective of the isotype of the TCR complex from which those
signals originate [35]. Compelling support for the signal strength
model was provided by the demonstration that thymocytes express-
ing a single γδTCR transgene could adopt either the αβ or γδ fate
upon manipulating the γδTCR to transduce weak or strong TCR
signals, respectively [36, 37].
While the signal strength model is now widely regarded as pro-
viding the best explanation for the role of the TCR complex in
lineage commitment, it remains unclear how the γδTCR complex
transduces the stronger signals required for adoption of the γδ fate.
Ligand stimulation remains a possible mechanism by which the γδ
TCR could produce those stronger signals. The role of ligand
stimulation in the intrathymic selection of αβ T cell progenitors is
well established; however, the role of ligand in regulating the
γδTCR signals that specify the γδ fate remains controversial [38,
39]. Based on the restriction of the chain usage and CDR3
sequences of the γδTCR complex that characterizes the DETC
subset of γδ T cells, it is likely that their development is ligand-
dependent [40]. DETC development requires expression of
Skint1, although whether Skint1 functions as a selecting ligand
remains to be fully established [41]. Analysis of γδTCR Tg model
reactive with the T-10/22 selecting ligand has provided very clear
evidence in support of a role for ligand in their selection [37, 42],
but it is not clear whether the involvement of ligand is a general
phenomenon or only involved in a select few cases. This issue can
only be addressed by either identifying the ligand specificities of a
large number of γδ progenitors, or, alternatively, through the use of
a surrogate for ligand engagement. Interestingly, we have recently
identified a surrogate for ligand engagement, CD73 induction.

CD73 is a TCR-ligand inducible surface protein that is expressed
on a large fraction of thymic γδ T cells, and nearly all peripheral γδ
T cells, indicating that a substantial fraction of developing γδ T cells
have encountered ligand [24].
Acquisition of effector fate during development in the thymus.
Unlike αβ lineage T cells, which exit the thymus in a naïve state
and acquire functional competence in the periphery, many γδ lin-
eage progenitors acquire functional competence prior to exit from
the thymus [18]. γδ T cells have been subdivided into effector
classes based on the cytokines they produce [43]. γδ cells can
adopt an IL-17-producing effector fate linked to the expression of
the transcription factor RORγt, the IFN-γ-producing effector fate
linked to the expression of the transcription factor Egr3, or an
innate effector fate characterized by the expression of the PLZF
transcription factor and the simultaneous production of both
IFN-γ and IL-4 [44]. The molecular processes underlying specifi-
cation of these effector fates remain poorly understood, but there
are two models attempting to explain how this is controlled and
how TCR signaling is involved in the specification process. One
model suggests that effector fate is predetermined and linked to
Vγ usage but not influenced by TCR signaling, while the other model
suggests that effector fate is influenced by TCR signaling (Fig. 2).
Supporting evidence for both models can be found. Indeed, a
Models for specification of T cell effector fate.

Pre-determined Influenced by TCR signaling
ROR t T ROR t IL17

IL17 T
Egr2 T Egr2 IFN

IFN
PLZF T innate PLZF innate
Fig. 2 Models describing the basis for linkage of effector fate to Vγ usage. Many γδ T cells acquire effector
function during development in the thymus. Two models have been advanced to explain how this occurs. The
first is the “pre-determination model,” which suggests that effector fate is pre-programmed and is linked to
Vγ usage by virtue of their developmental timing, but is not influenced by TCR signaling. The second suggests
that effector fate is influenced by TCR signal strength with a gradation of signaling (schematized by blue tri-
angle) ranging from weakest, which specify the IL-17 producing effector fate to the strongest, which specify
the PLZF-expressing innate effector fate associated with co-production of IL-4 and IFNγ
30 David L. Wiest
recent report indicated that immature CD24 high γδ progenitors

exhibited evidence of predetermined effector fate, because these
immature cells expressed elevated levels of the effector fate-specifying
transcription factors listed above, and this was linked to particular
Vγ [45]. For example, Vγ2+ cells, which are associated with IL-17
production, exhibited elevated RORγt levels, while Vγ1.1+ cells,
which are linked to IFN-γ production, exhibited elevated Egr3
levels [45]. Predetermination predicts that this link between the
effector fate-specifying transcription factor and the Vγ region
should be retained throughout maturation. Nevertheless, our
recent analysis suggests that this linkage is severed during matura-
tion, which is inconsistent with predetermination, and suggests
that TCR signaling influences fate [24]. There is additional evi-
dence suggesting that TCR signal strength can influence effector
fate. For example, the γδ subset reactive with the non-classical
MHC-I molecule, H-2T10/22 become IFNγ producers when
exposed to ligand (stronger signal) and IL-17 producers in the
absence of ligand (weaker signal) [46]. Similarly, the DETC subset
of Vγ3-expressing γδ cells adopt the IFNγ-producing effector fate
and home to the skin in response to exposure to presumptive
ligand, Skint1, but become IL-17 producers and are diverted to
the uterus in the absence of Skint1 [44]. Thus, while the role of γδ
TCR signaling in adoption of γ effector fate in the thymus remains
unclear, there is evidence that TCR signaling is able to exert some
influence over this process.
5 Conclusion
For many years, γδ T cells were considered to be few in number

and of questionable importance to host defense; however, in the
last 10–15 years, the importance of γδ T cell function in immune
responses has become abundantly clear, justifying the efforts to
gain insight into how γδ development and specification of effector
fate are controlled. As a result of those efforts an understanding is
beginning to emerge that specification of at least some γδ popula-
tions is dependent on more intense and/or prolonged TCR signals
and that these signals also influence effector fate specification.
Nevertheless, many important questions remain to be addressed.
How is the linkage between Vγ usage and effector fate accomplished?
Why do particular Vγ home to characteristic anatomic locations?
And, finally, it is becoming clear that some γδ T cells emerge from
the thymus capable of rapid innate-like responses, while others
require prior stimulation to acquire effector function [47]. It will
be important to understand the origins of these modes of function
as they almost certainly make distinct contributions to host defense.
References
1. Hayday AC (2000) [gamma][delta] cells: a 16. Lauritsen JP, Haks MC, Lefebvre JM et al
right time and a right place for a conserved (2006) Recent insights into the signals that
third way of protection. Annu Rev Immunol control alphabeta/gammadelta-lineage fate.
18:975–1026 Immunol Rev 209:176–190
2. Carding SR, Egan PJ (2002) Gammadelta T 17. Xiong N, Raulet DH (2007) Development and
cells: functional plasticity and heterogeneity. selection of gammadelta T cells. Immunol Rev
Nat Rev Immunol 2:336–345 215:15–31
3. Vantourout P, Hayday A (2013) Six-of-the-best: 18. Bonneville M, O’Brien RL, Born WK (2010)
unique contributions of gammadelta T cells to Gammadelta T cell effector functions: a blend
immunology. Nat Rev Immunol 13:88–100 of innate programming and acquired plasticity.
4. Sheridan BS, Romagnoli PA, Pham QM et al Nat Rev Immunol 10:467–478
(2013) gammadelta T cells exhibit multifunc- 19. O’Brien RL, Born WK (2010) gammadelta T
tional and protective memory in intestinal tis- cell subsets: a link between TCR and function?
sues. Immunity 39:184–195 Semin Immunol 22:193–198
5. King DP, Hyde DM, Jackson KA et al (1999) 20. Jin Y, Xia M, Saylor CM et al (2010) Cutting
Cutting edge: protective response to pulmo- edge: intrinsic programming of thymic gam-
nary injury requires gamma delta T lympho- madeltaT cells for specific peripheral tissue
cytes. J Immunol 162:5033–5036 localization. J Immunol 185:7156–7160
6. Ramsburg E, Tigelaar R, Craft J et al (2003) 21. Lee SY, Stadanlick J, Kappes DJ et al (2010)
Age-dependent requirement for gammadelta T Towards a molecular understanding of the dif-
cells in the primary but not secondary protec- ferential signals regulating alphabeta/gam-
tive immune response against an intestinal par- madelta T lineage choice. Semin Immunol
asite. J Exp Med 198:1403–1414 22:237–246
7. Girardi M, Oppenheim DE, Steele CR et al 22. Lewis JM, Girardi M, Roberts SJ et al (2006)
(2001) Regulation of cutaneous malignancy by Selection of the cutaneous intraepithelial gam-
{gamma}{delta} T cells. Science 20:20 madelta + T cell repertoire by a thymic stromal
8. Chen Y, Chou K, Fuchs E et al (2002) determinant. Nat Immunol 7:843–850
Protection of the intestinal mucosa by intraepi- 23. Pereira P, Zijlstra M, McMaster J et al (1992)
thelial gamma delta T cells. Proc Natl Acad Sci Blockade of transgenic gamma delta T cell
U S A 99:14338–14343 development in beta 2-microglobulin deficient
9. Sharp LL, Jameson JM, Cauvi G et al (2005) mice. EMBO J 11:25–31
Dendritic epidermal T cells regulate skin homeo- 24. Coffey F, Lee SY, Buus TB et al (2014) The
stasis through local production of insulin-like TCR ligand-inducible expression of CD73
growth factor 1. Nat Immunol 6:73–79 marks gammadelta lineage commitment and a
10. Nedellec S, Bonneville M, Scotet E (2010) metastable intermediate in effector specifica-
Human Vgamma9Vdelta2 T cells: from signals tion. J Exp Med 211:329–343
to functions. Semin Immunol 22:199–206 25. Ribot JC, deBarros A, Pang DJ et al (2009)
11. Meraviglia S, Eberl M, Vermijlen D et al (2010) CD27 is a thymic determinant of the balance
In vivo manipulation of Vgamma9Vdelta2 T between interferon-gamma- and interleukin
cells with zoledronate and low-dose interleukin- 17-producing gammadelta T cell subsets. Nat
2 for immunotherapy of advanced breast cancer Immunol 10:427–436
patients. Clin Exp Immunol 161:290–297 26. Kreslavsky T, Garbe AI, Krueger A et al
12. Cai Y, Fleming C, Yan J (2012) New insights of (2008) T cell receptor-instructed alphabeta
T cells in the pathogenesis of psoriasis. Cell versus gammadelta lineage commitment
Mol Immunol 9:302–309 revealed by single-cell analysis. J Exp Med
13. Wu P, Wu D, Ni C et al (2014) gammadeltaT17 205:1173–1186
cells promote the accumulation and expansion 27. Pennington DJ, Silva-Santos B, Shires J et al
of myeloid-derived suppressor cells in human (2003) The inter-relatedness and interdepen-
colorectal cancer. Immunity 40:785–800 dence of mouse T cell receptor gammadelta+
14. Born WK, Kemal Aydintug M, O’Brien RL and alphabeta+ cells. Nat Immunol 4:
(2013) Diversity of gammadelta T-cell anti- 991–998
gens. Cell Mol Immunol 10:13–20 28. Melichar HJ, Narayan K, Der SD et al (2007)
15. Rock EP, Sibbald PR, Davis MM et al (1994) Regulation of gammadelta versus alphabeta T
CDR3 length in antigen-specific immune lymphocyte differentiation by the transcription
receptors. J Exp Med 179:323–328 factor SOX13. Science 315:230–233
32 David L. Wiest
29. Roberts NA, White AJ, Jenkinson WE et al 39. Meyer C, Zeng X, Chien YH (2010) Ligand
(2012) Rank signaling links the development recognition during thymic development and
of invariant gammadelta T cell progenitors and gammadelta T cell function specification.
Aire(+) medullary epithelium. Immunity 36: Semin Immunol 22:207–213
427–437 40. Havran WL, Allison JP (1988) Developmentally
30. Silva-Santos B, Pennington DJ, Hayday AC ordered appearance of thymocytes expressing
(2005) Lymphotoxin-mediated regulation of different T-cell antigen receptors. Nature 335:
gammadelta cell differentiation by alphabeta T 443–445
cell progenitors. Science 307:925–928 41. Boyden LM, Lewis JM, Barbee SD et al (2008)
31. Petrie HT, Scollay R, Shortman K (1992) Skint1, the prototype of a newly identified
Commitment to the T cell receptor-alpha beta immunoglobulin superfamily gene cluster, pos-
or -gamma delta lineages can occur just prior to itively selects epidermal gammadelta T cells.
the onset of CD4 and CD8 expression among Nat Genet 40:656–662
immature thymocytes. Eur J Immunol 22: 42. Lauritsen JP, Wong GW, Lee SY et al (2009)
2185–2188 Marked induction of the helix-loop-helix pro-
32. Ciofani M, Knowles GC, Wiest DL et al (2006) tein Id3 promotes the gammadelta T cell fate
Stage-specific and differential notch depen- and renders their functional maturation Notch
dency at the alphabeta and gammadelta T lin- independent. Immunity 31:565–575
eage bifurcation. Immunity 25:105–116 43. Prinz I, Silva-Santos B, Pennington DJ (2013)
33. Narayan K, Kang J (2010) Disorderly conduct Functional development of gammadelta T
in gammadelta versus alphabeta T cell lineage cells. Eur J Immunol 43:1988–1994
commitment. Semin Immunol 22:222–227 44. Turchinovich G, Hayday AC (2011) Skint-1
34. Wong GW, Zuniga-Pflucker JC (2010) gam- identifies a common molecular mechanism for
madelta and alphabeta T cell lineage choice: the development of interferon-gamma-
resolution by a stronger sense of being. Semin secreting versus interleukin-17-secreting gam-
Immunol 22:228–236 madelta T cells. Immunity 35:59–68
35. Hayes SM, Love PE (2006) Strength of signal: 45. Narayan K, Sylvia KE, Malhotra N et al (2012)
a fundamental mechanism for cell fate specifica- Intrathymic programming of effector fates in
tion. Immunol Rev 209:170–175 three molecularly distinct gammadelta T cell
36. Hayes SM, Li L, Love PE (2005) TCR signal subtypes. Nat Immunol 13:511–518
strength influences alphabeta/gammadelta lin- 46. Jensen KD, Su X, Shin S et al (2008) Thymic
eage fate. Immunity 22:583–593 selection determines gammadelta T cell effec-
37. Haks MC, Lefebvre JM, Lauritsen JP et al tor fate: antigen-naive cells make interleukin-
(2005) Attenuation of gammadeltaTCR signal- 17 and antigen-experienced cells make
ing efficiently diverts thymocytes to the alpha- interferon gamma. Immunity 29:90–100
beta lineage. Immunity 22:595–606 47. Wencker M, Turchinovich G, Di Marco BR et al
38. Kreslavsky T, von Boehmer H (2010) gam- (2014) Innate-like T cells straddle innate and
madeltaTCR ligands and lineage commitment. adaptive immunity by altering antigen-receptor
Semin Immunol 22:214–221 responsiveness. Nat Immunol 15:80–87
Part II
Core Approaches and Strategies

Chapter 3
Genetic Tools to Study T Cell Development

Thomas Ciucci, Melanie S. Vacchio, and Rémy Bosselut
Abstract
Genetics tools, and especially the ability to enforce, by transgenesis, or disrupt, by homologous recombination,
gene expression in a cell-specific manner, have revolutionized the study of immunology and propelled the
laboratory mouse as the main model to study immune responses. Perhaps more than any other aspect of
immunology, the study of T cell development has benefited from these technologies. This brief chapter
summarizes genetic tools specific to T cell development studies, focusing on mouse strains with lineage-
and stage-specific expression of the Cre recombinase, or expressing unique antigen receptor specificities. It
ends with a broader discussion of strategies to enforce ectopic lineage and stage-specific gene expression.
Key words Cre recombinase, Genetic strategies, Lineage-specific gene disruption, TCR transgenic
mice, Deletion reporter genes
1 Introduction
Perhaps more than any other aspect of immunology, the study of

T cell development is intimately connected to mouse genetics.
Early discoveries elucidating the genetic bases for lymphocyte
responsiveness and tumor rejection proved instrumental for the
emergence of key concepts of T cell biology, including MHC
restriction and T cell selection. The advent of directed homolo-
gous gene recombination has paved the way to the study of mecha-
nisms of antigen receptor rearrangement and has been essential to
decipher signaling and differentiation in developing thymocytes.
Much is expected from new approaches to genetic manipulation,
including those using the bacteria-derived CRISPR-Cas9 system
[1, 2]. This brief chapter provides specific information on genetic
tools and strategies currently available to study mouse T cell devel-
opment; the reader is referred to Chapter 1 for information on
specific developmental stages.
35
36 Thomas Ciucci et al.
2 Lineage- and Stage-Specific Gene Disruption
The development of mouse strains expressing the Cre recombinase

at distinct stages of T cell development, using expression cassettes
specifically active in the T lineage, has considerably facilitated genetic
studies of T cell development. A selection of strains useful to target
Cre activity to specific developmental stages is presented in Table 1.
Several considerations affect the choice of a “deleter” strain. First is
the developmental timing of expression of the Cre recombinase,
schematically indicated in Table 1. It is important to verify the kinet-
ics of gene inactivation on a case-by-case basis. The actual timing of
loss-of-function depends not only on the gene “accessibility” to Cre,
which controls actual DNA excision, but also on downstream param-
eters, such as mRNA and protein half-life if the gene is expressed at
the time of deletion. Deletion “reporters” (e.g. Rosa26 derivatives,
available from commercial repositories) [21] are useful to identify
cells having undergone Cre activity. However, there is often imper-
fect matching between reporter expression and deletion of the gene
of interest, especially at the developmental time where Cre expres-
sion is initiated. Thus, assessing actual expression of the gene of
interest, which can be done by staining (including intracellular stain-
ing) and flow cytometry for an ever greater number of protein-
encoding genes, or by western blotting, remains the standard to
evaluate gene deletion efficiency.
Because of the possibility of off-target effects of Cre, the fre-
quency of which appears to be correlated to the level of Cre pro-
tein expression, breeding strategies should include the generation
of Cre-expressing controls that do not carry the homozygous
floxed allele [9, 22].
The intensive cell proliferation characteristic of early thymo-
cyte development raises issues specific of this stage. If inactivation
of the target gene results in a survival or proliferation disadvantage,
heterogeneity in the timing of deletion may result in compensatory
proliferation of cells (“escapees”) that have not undergone dele-
tion, which in turn masks the phenotypic consequences of dele-
tion. The most appropriate remedy to this situation is to place
mutant cells in competition with wild-type thymocytes, which can
be done in vivo in mixed bone marrow chimeras (see Chapter 9).
In these conditions, the wild-type cells will efficiently compete for
proliferation-driving ligands (e.g. IL-7) and keep in check the
compensatory expansion of the “escapee” population.
3 Fixing TCR Specificity to Analyze Selection Events
The diversity of TCR specificities expressed in αβ lineage thymo-

cytes (typically as many as there are thymocytes at steady state, i.e.
>108 in a typical laboratory mouse) poses daunting challenges for
Genetic Tools to Study T Cell Development 37
Table 1
Cre-expressing strains
Promoter
or enhancer Reference Cre expression Commerciala Comments
Range Onset
Vav1 [ 3] All hematopoietic HSC Jax
cells
Tie2 [4] All hematopoietic HSC Jax Also endothelial
cells cells
Rag1 [5] Rag1-expressing
B and T cell
precursors
Il7r [6] B and T cells BM lymphoid
(IL-7Rα) progenitors
Lck, proximal [7] T lineage cellsc DN2 Taconic
promoterb
(Wilson)
Lck, proximal [8] T lineage cellsc DN3 Jax Cre expression,
promoterb lower and
(Marth) delayed relative
to Wilson strain
[9]. Potential
for incomplete
deletion
CD2 [3] T and B cellsc DN2 Jax
Ptcra [10] T lineage cells DN2a
(Pre-Tα)
Cd4 [ 7] αβ lineage onlyc DN3 to DP, Taconic
therefore
deleting in
all αβ T cell
precursors
Cd8 E(III) [11] αβ lineage only ISP-DP
+
Cd8 E(I) [12, 13] CD8 T cells CD24lo CD8 SP
only thymocytes
Lck (distal) [14] T cells T cells Higher in SP
thymocytes and
T cells
CD2 [15] T cells Mature Expression pattern
thymocytes unrelated to
to Naïve other CD2-
T cells based constructs.
Incomplete in
CD4+ T cells
(continued)
Table 1
(continued)
Promoter
or enhancer Reference Cre expression Commerciala Comments
Range Onset
Cd4-Thpok [16] CD4+ T cells Not active in DP

thymocytes;
strongly specific
for CD4+ T
cells, but
variegated
expression
Ox40 [17, 18] Activated CD4+ T Jax
(Tnfrsf4) cells
Gzmb [19] Activated CD8+ T Jax
cells
Cd4 (ERt2 [20] CD4+ T cells Tamoxifen-
Cre) inducible Cre
a
Jax: Jackson laboratories; Taconic: Taconic Farms
b
Deletion with both Lck-Cre strains, notably that developed by Hennet et al. [8], tends to remain incomplete until the
DP stage, facilitating compensatory expansion of cells that have escaped deletion of genes involved in survival or prolif-
eration [9]
c
See comparison in the study by Shi and Petrie [9]
Abbreviations: DN: CD4−CD8− thymocytes; DP: CD4+CD8+ thymocytes; HSC: hematopoietic stem cells
analyses of T cell differentiation. Despite major progress in

mapping the T cell repertoire of naïve T cells (i.e. those that have
not yet expanded in response to antigenic stimulation) [23], it is
not yet possible to track the development of antigen-specific thy-
mocytes in a truly wild-type thymus. In addition, distinct TCR
gene rearrangement events can give rise to receptors recognizing
the same antigen, often with distinct avidities. The latter property
has important consequences for investigations of gene function in
developing thymocytes. Indeed, compensatory changes in the
selected repertoire can mask the cell-intrinsic impact of genes
involved in TCR-induced signaling or differentiation (e.g. selec-
tion of αβ T cells).
To overcome these limitations, it is possible to force thymocytes
to use a defined TCR specificity. This is traditionally performed by
using mice expressing a transgenic TCR, initially obtained by micro-
injecting cDNA for TCRαβ or TCRγδ chains with known specificity,
driven by appropriate T cell-specific promoters [24]. Multiple such
lines have been developed over the last 25 years. Specific informa-
tion on a few lines that have been extensively used in the study of
intrathymic T cell selection is provided in Tables 2 and 3. However,
Table 2
MHC-I-restricted TCR transgenic strains
Transgene Restriction Antigen Chains Comments References

OT-I H-2Kb Ovalbumin 257–264 Vα2, Vβ5 Extensively studied [25, 27]
(SIINFEKL) for analyses of
thymocyte
responsiveness to
MHC-peptide
ligands [25, 26]
HY H-2Db Smcy (Kdm5d) Vβ8 Recognized by [30]
738–746 clonotype-specific
(KCSRNRQYL) antibody T3.70
[28] [29]
2C H-2Kb Discussed in ref. 31 Vα3, Vβ8 H-2Ld alloreactive. [32, 33]
Recognized by
clonotype-specific
antibody 1B2
In-depth structural
studies of binding
to MHC-peptide
[31]
P14 H-2Db LCMV gp33 33–41 Vα2, Vβ8 Extensively used for [34]
(KAVYNFATC) studies of CD8 T
cell effector
responses.
F5 H-2Db Influenza Virus Vα4, Vβ11 [35]
Nucleoprotein
366–374
(ASNENMDTM)
many more TCR transgenic lines have been generated and used in a
variety of studies of T cell function during physiological and patho-
logical responses (e.g. refs. 45–47).
An important condition to the success of this strategy is that
cells expressing the transgenic TCR do not express other TCR
chains generated by rearrangement of endogenous loci. That is
only partly the case. Expression of a rearranged TCRβ chain
(whether endogenous or transgenic) prevents further rearrange-
ment at the TCRβ loci (a phenomenon called allelic exclusion)
[48] and at the TCRγ loci (isotypic exclusion) [49]. However,
TCRβ allelic exclusion is not perfect; as a result, in many TCR
transgenic strains, small fractions of thymocytes express both the
transgenic and an endogenous TCRβ chain. In addition, there is
no allelic exclusion at the TCRα locus, so that expression of a
transgenic TCRαβ does not guarantee absence of endogenous
TCRα gene rearrangement, which has potential implications for
40
Table 3
MHC-II-restricted TCR transgenic strains
Transgene Restriction Antigen Chains Comments References

OT-II I-Ab Ovalbumin 323–339 Vα2, Vβ5 Core recognition ovalbumin sequence at residues [37]
Thomas Ciucci et al.
(ISQAVHAAHAEINEAGR) 329–337 [36]

DO11.10 I-Ad Ovalbumin 323–339 Vα13, Vβ8 Recognized by clonotype-specific antibody KJ126. [38]
(ISQAVHAAHAEINEAGR) Core recognition ovalbumin sequence at residues
329–337 [36].
Promotes positive selection in H-2b mice (selected
by I-Ab).
AND I-Ek Pigeon Cytochrome c 88–104 Vα11, Vβ3 TCRα and TCRβ chain originated from two distinct [40]
(KAERADLIAYLKQATAK) T cell clones (TCRβ chain identical to 5C.C7
TCR).
Promotes positive selection in H-2b mice (selected
by I-Ab), and more efficiently in mice carrying one
allele of I-Ab and I-Ek (H-2bxk). Promotes negative
selection in mice expressing I-As, and incompletely
in mice homozygous for I-Ek [39].
5C.C7 I-Ek Pigeon Cytochrome c 88–104 Vα11, Vβ3 TCRβ chain identical to AND TCR. [41, 42]
(KAERADLIAYLKQATAK) Promotes positive selection in mice expressing I-Ek
and negative selection in mice expressing I-As. Not
selected by I-Ab [39].
3A9 I-Ak Hen Egg Lysozyme (HEL) 46–61 Vα3, Vβ8 Used in negative selection studies [43] [44]
(NTDGSTDYGILQINSR)
negative selection [50]. The frequency of such “secondary” TCRα

and TCRβ rearrangements depends on the specific TCR transgenic
strain, and can only be evaluated on a case-by-case basis. Introducing
mutations in the Rag1 or Rag2 genes [by intercrossing with lines
carrying such mutations [51, 52], which are available from com-
mercial repositories] prevents any endogenous TCR gene
rearrangement and eliminates the potential for development of
such cells.
While the choice of a specific model is dictated by the question
being examined, one specific approach deserves mention here
because of it usefulness to study CD4+CD8+ (“double positive,”
DP) thymocyte signaling [25, 26]. For MHC I-induced selection,
it is possible to manipulate MHC-I peptide binding in fetal thymic
organ cultures (FTOC, see Chapter 12) when epithelial cells lack
β2 microglobulin molecules or the TAP transporters needed for
MHC-I peptide loading; in such circumstances, addition of pep-
tides and exogenous β2 microglobulin to the FTOC allow partial
reconstitution of MHC-I-peptide complexes and intrathymic sig-
naling. Many studies have used this approach with the OT-I TCR
(see Table 2). Derivatives of this peptide have been generated that
promote positive or negative selection, making this system very
useful to study relationships between intrathymic signaling and
developmental outcomes.
The conventional approach to TCR transgenesis has important
limitations. It is tedious and costly to introduce a TCR transgene
on an otherwise compound mutant background. Because they are
not expressed from the endogenous corresponding loci, most TCR
transgenes are expressed prematurely, resulting in multiple
unwanted effects. Last, the broad expression of transgenic TCRs in
thymocytes disrupts the physiological signaling balance in the
thymus, by generating an artificial situation in which all thymo-
cytes share the same MHC-peptide ligands, which then become
limiting [53]. The latter issue can be dealt with by reducing the
frequency of transgenic precursors using a mixed bone marrow
chimera approach. Strategies to address the first two limitations are
summarized below.
Retroviral transduction can avoid costly and lengthy breeding
of TCR transgenic strains. It involves transducing bone marrow
precursors with a retroviral vector expressing the TCR chains and
a fluorescent marker to identify transduced cells. Vectors express-
ing both chains of a TCR dimer from a monocistronic mRNA are
particularly useful [54]. Retroviral transduction strategies do not
avoid premature TCR expression (see below). Of note, using
recipient mice lacking their own T cells (e.g. deficient for Rag
genes) has the potential to result in spurious developmental effects
that should be carefully considered during experiment design [55].
Premature TCR expression has been one of the most vexing
limitations of TCR transgenes, especially for the study of negative
selection. Indeed, TCRαβ engagement by high-avidity ligands in

early thymocytes results in their lineage redirection toward a γδ-like
fate (even though they continue expressing the transgenic αβ
TCR), and therefore hampers the study of negative selection itself
[56, 57]. More physiological approaches to negative selection have
been developed, based on the Aire-dependent activity of tissue-
specific promoters in medullary epithelial cells. One specific strat-
egy uses two transgenes to express antigen and TCR: a transgenic
TCR specific for a known antigen, and a transgene expressing this
antigen under the control of the rat insulin promoter (RIP), which
in the thymus targets expression of specific antigens to medullary
epithelial cells [43, 45, 58]. In such settings, premature expression
of the TCR transgene has a lesser impact on T cell development as
transgenic thymocytes do not encounter negatively selecting
ligands until they reach the medulla. However, the outcome of
these experiments varies with the level of expression of antigen in
the specific system being used [59].
In addition, strategies have been devised to improve the timing
of TCR expression. The most successful so far relies on the expres-
sion of two separate transgenes, one encoding the TCRβ chain,
and a “conditional” transgene encoding the TCRα chain, whose
expression is normally prevented by a strong “floxed” transcription
termination signal located upstream of the coding sequence [60].
Removing the floxed sequence, using a Cre deleter active in DP
thymocytes (e.g. Cd4-Cre), allows a timely expression of the TCRα
chain, and therefore of TCRαβ complexes. Similar strategies have
been developed for retroviral vectors [61].
References
1. Gupta RM, Musunuru K (2014) Expanding the locations or retroviral insertion during gene
genetic editing tool kit: ZFNs, TALENs, and therapy but has no mandatory role in normal
CRISPR-Cas9. J Clin Invest 124:4154–4161 T-cell development. Mol Cell Biol 23:
2. Mali P, Esvelt KM, Church GM (2013) Cas9 9003–9013
as a versatile tool for engineering biology. Nat 6. Schlenner SM, Madan V, Busch K, Tietz A,
Methods 10:957–963 Laufle C, Costa C, Blum C, Fehling HJ,
3. de Boer J, Williams A, Skavdis G, Harker N, Rodewald HR (2010) Fate mapping reveals
Coles M, Tolaini M, Norton T, Williams K, separate origins of T cells and myeloid lineages
Roderick K, Potocnik AJ, Kioussis D (2003) in the thymus. Immunity 32:426–436
Transgenic mice with hematopoietic and lym- 7. Lee PP, Fitzpatrick DR, Beard C, Jessup HK,
phoid specific expression of Cre. Eur J Lehar S, Makar KW, Perez-Melgosa M,
Immunol 33:314–325 Sweetser MT, Schlissel MS, Nguyen S, Cherry
4. Kisanuki YY, Hammer RE, Miyazaki J, SR, Tsai JH, Tucker SM, Weaver WM, Kelso
Williams SC, Richardson JA, Yanagisawa M A, Jaenisch R, Wilson CB (2001) A critical
(2001) Tie2-Cre transgenic mice: a new model role for Dnmt1 and DNA methylation in T cell
for endothelial cell-lineage analysis in vivo. development, function, and survival. Immunity
Dev Biol 230:230–242 15:763–774
5. McCormack MP, Forster A, Drynan L, Pannell 8. Hennet T, Hagen FK, Tabak LA, Marth JD
R, Rabbitts TH (2003) The LMO2 T-cell (1995) T-cell-specific deletion of a polypeptide
oncogene is activated via chromosomal trans- N-acetylgalactosaminyl-transferase gene by
site-directed recombination. Proc Natl Acad shows its essential function in T(H)1-T(H)2
Sci U S A 92:12070–12074 responses. Nat Immunol 5:1157–1165
9. Shi J, Petrie HT (2012) Activation kinetics 18. Klinger M, Kim JK, Chmura SA, Barczak A,
and off-target effects of thymus-initiated cre Erle DJ, Killeen N (2009) Thymic OX40
transgenes. PLoS One 7, e46590 expression discriminates cells undergoing
10. Luche H, Nageswara Rao T, Kumar S, strong responses to selection ligands. J
Tasdogan A, Beckel F, Blum C, Martins VC, Immunol 182:4581–4589
Rodewald HR, Fehling HJ (2013) In vivo fate 19. Jacob J, Baltimore D (1999) Modelling T-cell
mapping identifies pre-TCRalpha expression memory by genetic marking of memory T cells
as an intra- and extrathymic, but not prethy- in vivo. Nature 399:593–597
mic, marker of T lymphopoiesis. J Exp Med 20. Aghajani K, Keerthivasan S, Yu Y, Gounari F
210:699–714 (2012) Generation of CD4CreER(T(2))
11. Park JH, Adoro S, Guinter T, Erman B, Alag transgenic mice to study development of
AS, Catalfamo M, Kimura MY, Cui Y, Lucas peripheral CD4-T-cells. Genesis 50:908–913
PJ, Gress RE, Kubo M, Hennighausen L, 21. Soriano P (1999) Generalized lacZ expression
Feigenbaum L, Singer A (2010) Signaling by with the ROSA26 Cre reporter strain. Nat
intrathymic cytokines, not T cell antigen Genet 21:70–71
receptors, specifies CD8 lineage choice and 22. Schmidt-Supprian M, Rajewsky K (2007)
promotes the differentiation of cytotoxic-lin- Vagaries of conditional gene targeting. Nat
eage T cells. Nat Immunol 11:257–264 Immunol 8:665–668
12. Zou YR, Sunshine MJ, Taniuchi I, Hatam F, 23. Moon JJ, Chu HH, Pepper M, McSorley SJ,
Killeen N, Littman DR (2001) Epigenetic Jameson SC, Kedl RM, Jenkins MK (2007)
silencing of CD4 in T cells committed to the Naive CD4(+) T cell frequency varies for dif-
cytotoxic lineage. Nat Genet 29:332–336 ferent epitopes and predicts repertoire diver-
13. Maekawa Y, Minato Y, Ishifune C, Kurihara T, sity and response magnitude. Immunity
Kitamura A, Kojima H, Yagita H, Sakata- 27:203–213
Yanagimoto M, Saito T, Taniuchi I, Chiba S, 24. von Boehmer H (1990) Developmental biol-
Sone S, Yasutomo K (2008) Notch2 integrates ogy of T cells in T cell-receptor transgenic
signaling by the transcription factors RBP-J mice. Annu Rev Immunol 8:531–556
and CREB1 to promote T cell cytotoxicity.
Nat Immunol 9:1140–1147 25. Hogquist KA, Jameson SC, Heath WR,
Howard JL, Bevan MJ, Carbone FR (1994) T
14. Zhang DJ, Wang Q, Wei J, Baimukanova G, cell receptor antagonist peptides induce posi-
Buchholz F, Stewart AF, Mao X, Killeen N tive selection. Cell 76:17–27
(2005) Selective expression of the Cre recom-
binase in late-stage thymocytes using the distal 26. Daniels MA, Teixeiro E, Gill J, Hausmann B,
promoter of the Lck gene. J Immunol Roubaty D, Holmberg K, Werlen G, Hollander
174:6725–6731 GA, Gascoigne NR, Palmer E (2006) Thymic
selection threshold defined by compartmental-
15. Vacchio MS, Wang L, Bouladoux N, Carpenter ization of Ras/MAPK signalling. Nature
AC, Xiong Y, Williams LC, Wohlfert E, Song 444:724–729
KD, Belkaid Y, Love PE, Bosselut R (2014) A
ThPOK-LRF transcriptional node maintains the 27. Clarke SR, Barnden M, Kurts C, Carbone FR,
integrity and effector potential of post-thymic Miller JF, Heath WR (2000) Characterization
CD4 T cells. Nat Immunol 15(10):947–956 of the ovalbumin-specific TCR transgenic line
OT-I: MHC elements for positive and nega-
16. Mucida D, Husain MM, Muroi S, van Wijk F, tive selection. Immunol Cell Biol 78:110–117
Shinnakasu R, Naoe Y, Reis BS, Huang Y,
Lambolez F, Docherty M, Attinger A, Shui 28. Markiewicz MA, Girao C, Opferman JT, Sun
JW, Kim G, Lena CJ, Sakaguchi S, Miyamoto J, Hu Q, Agulnik AA, Bishop CE, Thompson
C, Wang P, Atarashi K, Park Y, Nakayama T, CB, Ashton-Rickardt PG (1998) Long-term T
Honda K, Ellmeier W, Kronenberg M, cell memory requires the surface expression of
Taniuchi I, Cheroutre H (2013) Transcriptional self-peptide/major histocompatibility complex
reprogramming of mature CD4(+) helper T molecules. Proc Natl Acad Sci U S A
cells generates distinct MHC class II-restricted 95:3065–3070
cytotoxic T lymphocytes. Nat Immunol 29. Teh HS, Kishi H, Scott B, Von Boehmer H
14:281–289 (1989) Deletion of autospecific T cells in T cell
17. Zhu J, Min B, Hu-Li J, Watson CJ, Grinberg receptor (TCR) transgenic mice spares cells
A, Wang Q, Killeen N, Urban JFJ, Guo L, Paul with normal TCR levels and low levels of CD8
WE (2004) Conditional deletion of Gata3 molecules. J Exp Med 169:795–806
30. Kisielow P, Bluthmann H, Staerz UD, 41. Seder RA, Paul WE, Davis MM, Fazekas de St
Steinmetz M, von Boehmer H (1988) Groth B (1992) The presence of interleukin 4
Tolerance in T-cell-receptor transgenic mice during in vitro priming determines the
involves deletion of nonmature CD4+8+ thy- lymphokine-producing potential of CD4+ T
mocytes. Nature 333:742–746 cells from T cell receptor transgenic mice. J
31. Colf LA, Bankovich AJ, Hanick NA, Exp Med 176:1091–1098
Bowerman NA, Jones LL, Kranz DM, Garcia 42. Fink PJ, Matis LA, McElligott DL, Bookman
KC (2007) How a single T cell receptor recog- M, Hedrick SM (1986) Correlations between
nizes both self and foreign MHC. Cell T-cell specificity and the structure of the anti-
129:135–146 gen receptor. Nature 321:219–226
32. Sha WC, Nelson CA, Newberry RD, Kranz 43. Akkaraju S, Ho WY, Leong D, Canaan K,
DM, Russell JH, Loh DY (1988) Selective Davis MM, Goodnow CC (1997) A range of
expression of an antigen receptor on CD8- CD4 T cell tolerance: partial inactivation to
bearing T lymphocytes in transgenic mice. organ-specific antigen allows nondestructive
Nature 335:271–274 thyroiditis or insulitis. Immunity 7:255–271
33. Garcia KC, Tallquist MD, Pease LR, Brunmark 44. Ho WY, Cooke MP, Goodnow CC, Davis MM
A, Scott CA, Degano M, Stura EA, Peterson (1994) Resting and anergic B cells are defective
PA, Wilson IA, Teyton L (1997) Alphabeta T in CD28-dependent costimulation of naive
cell receptor interactions with syngeneic and CD4+ T cells. J Exp Med 179:1539–1549
allogeneic ligands: affinity measurements and 45. Apostolou I, Von Boehmer H (2004) The
crystallization. Proc Natl Acad Sci U S A TCR-HA, INS-HA transgenic model of
94:13838–13843 autoimmune diabetes: limitations and expecta-
34. Pircher H, Burki K, Lang R, Hengartner H, tions. J Autoimmun 22:111–114
Zinkernagel RM (1989) Tolerance induction in 46. Bettelli E, Pagany M, Weiner HL, Linington
double specific T-cell receptor transgenic mice C, Sobel RA, Kuchroo VK (2003) Myelin oli-
varies with antigen. Nature 342:559–561 godendrocyte glycoprotein-specific T cell
35. Mamalaki C, Elliott J, Norton T, Yannoutsos receptor transgenic mice develop spontaneous
N, Townsend AR, Chandler P, Simpson E, autoimmune optic neuritis. J Exp Med
Kioussis D (1993) Positive and negative selec- 197:1073–1081
tion in transgenic mice expressing a T-cell 47. Katz JD, Wang B, Haskins K, Benoist C,
receptor specific for influenza nucleoprotein Mathis D (1993) Following a diabetogenic T
and endogenous superantigen. Dev Immunol cell from genesis through pathogenesis. Cell
3:159–174 74:1089–1100
36. Robertson JM, Jensen PE, Evavold BD (2000) 48. Uematsu Y, Ryser S, Dembic Z, Borgulya P,
DO11.10 and OT-II T cells recognize a Krimpenfort P, Berns A, von Boehmer H,
C-terminal ovalbumin 323-339 epitope. J Steinmetz M (1988) In transgenic mice the
Immunol 164:4706–4712 introduced functional T cell receptor beta
37. Barnden MJ, Allison J, Heath WR, Carbone gene prevents expression of endogenous beta
FR (1998) Defective TCR expression in transgenes. Cell 52:831–841
genic mice constructed using cDNA-based 49. Fenton RG, Marrack P, Kappler JW, Kanagawa
alpha- and beta-chain genes under the control O, Seidman JG (1988) Isotypic exclusion of
of heterologous regulatory elements. Immunol gamma delta T cell receptors in transgenic
Cell Biol 76:34–40 mice bearing a rearranged beta-chain gene.
38. Murphy KM, Heimberger AB, Loh DY (1990) Science 241:1089–1092
Induction by antigen of intrathymic apoptosis 50. Sarukhan A, Garcia C, Lanoue A, von Boehmer
of CD4+CD8+TCRlo thymocytes in vivo. H (1998) Allelic inclusion of T cell receptor
Science 250:1720–1723 alpha genes poses an autoimmune hazard due
39. Kaye J, Vasquez NJ, Hedrick SM (1992) to low-level expression of autospecific recep-
Involvement of the same region of the T cell tors. Immunity 8:563–570
antigen receptor in thymic selection and for- 51. Mombaerts P, Iacomini J, Johnson RS, Herrup
eign peptide recognition. J Immunol K, Tonegawa S, Papaioannou VE (1992)
148:3342–3353 RAG-1-deficient mice have no mature B and T
40. Kaye J, Hsu ML, Sauron ME, Jameson SC, lymphocytes. Cell 68:869–877
Gascoigne NR, Hedrick SM (1989) Selective 52. Shinkai Y, Rathbun G, Lam KP, Oltz EM,
development of CD4+ T cells in transgenic Stewart V, Mendelsohn M, Charron J, Datta
mice expressing a class II MHC-restricted anti- M, Young F, Stall AM et al (1992) RAG-2-
gen receptor. Nature 341:746–749 deficient mice lack mature lymphocytes owing
to inability to initiate V(D)J rearrangement. 57. Bruno L, Fehling HJ, von Boehmer H (1996)
Cell 68:855–867 The alpha beta T cell receptor can replace the
53. Huesmann M, Scott B, Kisielow P, von gamma delta receptor in the development of
Boehmer H (1991) Kinetics and efficacy of gamma delta lineage cells. Immunity
positive selection in the thymus of normal and 5:343–352
T cell receptor transgenic mice. Cell 58. DiPaolo RJ, Unanue ER (2001) The level of
66:533–540 peptide-MHC complex determines the sus-
54. Bettini ML, Bettini M, Vignali DA (2012) ceptibility to autoimmune diabetes: studies in
T-cell receptor retrogenic mice: a rapid, flexi- HEL transgenic mice. Eur J Immunol 31:
ble alternative to T-cell receptor transgenic 3453–3459
mice. Immunology 136:265–272 59. Byersdorfer CA, Schweitzer GG, Unanue ER
55. Hwang S, Song KD, Lesourne R, Lee J, (2005) Diabetes is predicted by the beta cell level
Pinkhasov J, Li L, El-Khoury D, Love PE of autoantigen. J Immunol 175:4347–4354
(2012) Reduced TCR signaling potential 60. Baldwin TA, Sandau MM, Jameson SC,
impairs negative selection but does not result Hogquist KA (2005) The timing of TCR{alpha}
in autoimmune disease. J Exp Med expression critically influences T cell develop-
209:1781–1795 ment and selection. J Exp Med 202:111–121
56. Takahama Y, Shores EW, Singer A (1992) 61. Turner VM, Gardam S, Brink R (2010)
Negative selection of precursor thymocytes Lineage-specific transgene expression in hema-
before their differentiation into CD4+CD8+ topoietic cells using a Cre-regulated retroviral
cells. Science 258:653–656 vector. J Immunol Methods 360:162–166
Chapter 4
Assessment of T Cell Development by Flow Cytometry

Jan Y.M. Lee and Paul E. Love
Abstract
T cell development is a complex multistep process that requires the coordinated activation of distinct
signaling responses and the regulated progression of developing cells (thymocytes) through key stages
of maturation. Although sophisticated techniques such as fetal thymus organ culture, in vitro thymocyte
culture, and multi-parameter flow cytometric analysis are now widely employed to evaluate thymocyte
maturation by experienced laboratories, defects in T cell development can usually be identified with
more simplified screening methods. Here, we provide a basic protocol for assessment of T cell develop-
ment that will enable laboratories with access to a four parameter flow cytometer to screen mouse
strains, including those generated from embryonic stem cells with targeted gene mutations, for thymo-
cyte maturation defects.
Key words T cell development, Thymocytes, T lymphocytes, Flow cytometry, Antibody staining,
Intracellular staining, αβ T cells, γδ T cells, NKT cells, Regulatory T cells
1 Introduction
T cell development takes place throughout life, beginning during

late embryogenesis and continuing into adulthood. T cell progeni-
tors originate in the fetal liver or in the adult bone marrow then
migrate to the thymus where they mature, acquire functional com-
petence, and eventually leave the thymus and populate peripheral
lymphoid organs [1–3].
Once T progenitor cells reach the thymus they progress
through several discrete stages of development (Fig. 1). The matu-
ration status of the major subset of T cells in humans and mice
[designated αβ T Cell Receptor (αβ TCR)+ T cells] can be gener-
ally assessed by staining for two cell surface molecules (CD4 and
CD8). The most immature thymocytes do not express CD4 or
CD8 and are designated double negative (DN). DN thymocytes
develop into CD4+CD8+ (double positive, DP) thymocytes which
comprise the majority (approximately 80–90 %) of cells in the thy-
mus after going through a transitional CD4−CD8+ Single Positive
47
48 Jan Y.M. Lee and Paul E. Love
a 104
CD4 SP DP
CD4SP
3
10
(4-9%) DP
(84-92%)
102
CD8 SP
CD8SP
(2-3%)
101
DN
(2-4%)
DN
CD4
100
0
10 101 102 103 104
CD8
b
Stage 1 Stage 2
ISP
DN DN
CD4
CD4
CD8 CD8
Stage 4 Stage 3
CD4SP
DP DP
CD8SP
ISP
CD4
CD4
CD8 CD8
Fig. 1 Major stages of thymocyte development. Thymocytes are stained with
anti-CD4 and anti-CD8α. (a) The classic “bird” profile of thymocytes stained with
anti-CD4 and anti-CD8α, showing the four major populations (DN double nega-
tive, DP double positive, CD4 SP CD4 single positive, and CD8 SP CD8 single
positive). Also shown is the expected percentage of each population for a wild-
type mouse thymus. (b) Stage 1: CD4−CD8− (DN double negative); Stage 2:
CD4−CD8+TCRβlow/− (ISP immature CD8 single positive); Stage 3: CD4+CD8+ (DP
double positive); Stage 4: CD4+CD8− (CD4 SP CD4 single positive) or CD4−CD8+
(CD8 SP CD8 single positive). Note that DP thymocytes transition through a
CD4+CD8low intermediate stage before becoming either CD4 SP or CD8 SP
Assessment of T Cell Development by Flow Cytometry 49
a
20000 20000
15000 15000
10000
# of Cells
# of Cells
10000
5000 5000
0 0
100 101 102 103 104 100 101 102 103 104
TCRβ CD5
b CD4-CD8- (DN) CD4+CD8+ (DP) CD4 (SP) CD8 (SP)

Double Negative Double Positive Single Positive Single Positive
20000
800
1200 3000
DN 15000
TCRβ+
600
900
2000 ISP
10000 400
600
1000
300 5000 200
0 0 0 0
# of Cells
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
TCRβ
1000
400
20000 2500
800
2000 300
15000
600
1500
10000 200
400 1000
5000 100
200 500
0 0 0 0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
CD5
Fig. 2 TCRβ and CD5 surface expression on thymocytes. Thymocytes are stained with anti-CD4, anti-CD8α,
anti-TCRβ, and anti-CD5. (a) TCRβ and CD5 levels on total thymocytes. (b) TCRβ and CD5 levels on DN, DP, CD4
SP, and CD8 SP thymocyte populations. The TCRβlow/− cells in the CD8 SP population are ISPs
(ISP) stage. The final stage of development results in downregulation

of either CD4 or CD8 resulting in the generation of “helper”
CD4+CD8− (CD4 Single Positive, CD4 SP) or “cytotoxic”
CD4−CD8+ (CD8 Single Positive, CD8 SP) T cells. Two other
markers that are commonly used to evaluate thymocyte maturation
are TCRβ and CD5 (Fig. 2).
DN thymocytes are typically divided into four sub-groups on
the basis of CD25 and CD44 surface expression that represent
progressive stages of maturation: CD44+CD25− (DN1),
CD44+CD25+ (DN2), CD44−CD25+ (DN3), and CD44−CD25−
(DN4) [4] (Fig. 3). At the earliest DN1 stage, thymocytes contain
a b
104 104
103 103 DN1 DN2

102 102
Lineage
Lineage Mix
101 101
DN4 DN3
Negative
Cells
CD44
100 100
0 200 400 600 800 1000 100 101 102 103 104
Forward Scatter (FSC) CD25
Fig. 3 Stages of double negative (DN) thymocyte development. Thymocytes are stained with anti-CD25, anti-CD44,
and a lineage mix containing the following markers: anti-CD4, anti-CD8α, anti-TCRβ, anti-TCRγδ, anti-CD19,
anti-NK1.1, anti-DX5, anti-Gr1 (Ly6C/G), anti-CD11b, and anti-Ter119 to facilitate gating on DN cells (see Note 1).
All the antibodies used in the lineage mix are conjugated to the same fluorochrome. (a) Shown is a Lineage Mix
vs. Forward Scatter (FSC) plot of total thymocytes and the gate used to analyze DN cells. (b) Gating on thymocytes
that are negative for the lineage mix, the DN subsets are revealed by the CD44 vs. CD25 plot
clonotypic α and β TCR genes in the germline configuration [5].

Rearrangement of the TCRβ locus begins at the DN2 stage and
continues in DN3 cells. Productive (V-(D)-J) rearrangement of
TCRβ results in the surface expression of a precursor form of the
mature αβTCR (designated the pre-TCR) which consists of TCRβ
chain paired with the invariant pre-Tα chain in addition to the
CD3 signal transducing subunits. Expression of and signaling
through the pre-TCR triggers DN3 cells to transition first to the
DN4 stage and then through an transitional Immature
(TCRβlow/−CD24hi) CD8 Single Positive (ISP) stage on their way
to becoming DP thymocytes (Fig. 4) [6, 7]. TCRα locus rear-
rangement begins in DP thymocytes and the productive rearrange-
ment of TCRα results in surface expression of the mature αβTCR
complex. DP thymocytes are subjected to a “selection” process on
the basis of the specificity of their TCR for self-ligands that pro-
motes the survival and differentiation of functionally competent
cells (positive selection) and triggers the death of non-self-reactive
cells (non-selection) as well as overtly auto-reactive cells (negative
selection) [8, 9]. Signaling through the TCR complex on DP thy-
mocytes regulates selection and results in upregulation of the TCR
as well as the surface proteins CD69 and CD5 which can be used
to monitor thymocyte activation and maturation at the DP stage
(Fig. 5). Newly generated CD4 SP and CD8 SP thymocytes are
functionally immature and express high levels of CD24 (Fig. 6).
The end product of the αβ differentiation pathway is the generation
a 104
b 104
CD4 SP
103 103 ISP
ISP+CD8 SP
DP
102 102
CD8 SP
ISP
+
DN 101 101
CD8
CD4
CD4
SP
CD24
100 100
100 101 102 103 104 100 101 102 103 104
CD8 CD8 TCRb
c
100
Iso ISP Iso

80
Ctrl ISP
Ctrl
CD8 SP
# of Cells
60 CD8 SP
40
20
0
100 101 102 103 104 100 101 102 103 104
CD5 TCRb
* Iso Ctrl = Isotype Control
Fig. 4 Immature CD8 single positive thymocytes (ISPs) are intermediates between the DN and DP stages.
Thymocytes are stained with anti-CD4, anti-CD8α, anti-TCRβ, and anti-CD24 (HSA) or anti-CD5. (a) Focusing
on the ISP/CD8 SP gate of the thymus “bird” profile. (b) ISPs can be distinguished from more mature CD8 SP
thymocytes by staining for CD24 and TCRβ. ISPs are CD24hiTCRβlow/− whereas CD8 SP cells are CD24int/
low
TCRβhi (Gated on CD8 SP). (c) ISPs express low surface levels of CD5 and TCRβ compared to more mature
CD8 SP thymocytes
of mature functional TCRβhiCD24low CD4 SP and CD8 SP thymo-

cytes that are exported to the periphery (Fig. 7).
Peripheral CD4 SP and CD8 SP that have not encountered
antigen are referred to as “naïve” and can be identified by their
CD44low CD62Lhi phenotype. After activation by antigen stimula-
tion, peripheral T cells acquire either a central memory (CD44hi
CD62Lhi) or effector memory (CD44hi CD62Llow) phenotype [10]
(Fig. 8).
In addition to “conventional” CD4 SP and CD8 SP T cells,
three less abundant specialized groups of T cells also develop in the
thymus. Natural Killer T (NKT) cells express a restricted αβTCR
that recognizes self or foreign lipids/glycolipids bound to the non-
polymorphic CD1d molecule [11]. NKT cells can be either CD4
SP or DN and are detected by staining with α-galactosylceramide
loaded CD1d tetramers (Fig. 9). Regulatory T cells (Tregs) are
suppressive CD4 SP T cells that maintain tolerance to self antigens
[12]. Tregs in the thymus and periphery can be detected by CD25
a b
CD4 SP
DP 4 3
CD8 SP
1
ISP
DN
+
CD4
TCRβ
CD8 CD69
c 104
d 104
DP gated
103 103
4 3
DP
102 102
101 101
2
1
TCRβ
CD4
100 100
100 101 102 103 104 100 101 102 103 104
CD8 CD69
Fig. 5 Assessment of Double Positive (DP) thymocyte development. Thymocytes are stained with anti-CD4,
anti-CD8α, anti-TCRβ, and anti-CD69. (a) Focusing on the DP population in the thymus “bird” profile. (b) Gating
on the DP population, the four stages of DP development can be visualized by TCRβ vs. CD69 expression: Stage
1, TCRβlow/−CD69low/−; Stage 2, TCRβlowCD69low; Stage 3, TCRβhiCD69hi; Stage 4, TCRβhi CD69low. (c) CD4 vs. CD8
plot of total thymocytes as depicted in (a). (d) TCRβ vs. CD69 plot of DP thymocytes as depicted in (b)
surface expression and by intracellular staining for FoxP3, a

transcription factor essential for their development (Fig. 10).
Finally, γδ T cells arise from immature DN thymocytes that have
productively rearranged the TCRγ and TCRδ loci and that express
a distinct γδTCR complex in lieu of the αβTCR [13]. Most γδ T
cells remain CD4−CD8− and make up a small percentage of periph-
eral T cells (Fig. 11).
a
CD4 SP
DP
CD8 SP
DN
CD4
CD8
b 104 104 104
CD4 SP CD4 SP CD8 SP
10
3
103 103 ISP
102 102 102

CD8 SP
Immature
Immature
101 101 101
Mature Mature
CD24
CD4
0 0 0
10 10 10
0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10 100 101 102 103 104
CD8 TCRβ
c 4
104 104
10
CD4 SP CD4 SP CD8 SP
103 Mature 10 3 Mature
103
2 2
10 10
102
CD8 SP
Immature Immature
101 101
101
TCRβ
CD4
0 100 100
10
100 101 102 103 10
4
10
0
10
1
10
2
10
3
104
100 101 102 103 104
CD8 CD69
Fig. 6 Assessment of Single Positive (SP) thymocyte maturation. Thymocytes are stained with anti-CD4, anti-
CD8α, anti-TCRβ, and anti-CD24. (a) Focusing on the CD4 SP and CD8 SP thymocyte populations. (b) Gating
on either CD4 SP or CD8 SP population, one can differentiate between immature (TCRβhiCD24hi) or mature
(TCRβhiCD24low) SP thymocytes. Note that in the case of CD8 SP, the ISPs are also visible as TCRβlow/−CD24hi
cells. (c) CD69 can also be used as a maturation marker. Gating on CD4 SP or CD8 SP population, immature
SP cells are TCRβhiCD69hi while mature SP cells are TCRβhiCD69low/−
2 Materials
2.1 Removal 1. Refrigerated centrifuge capable of 450 × g (e.g., Sorvall®

and Preparation Legend RT).
of Lymphoid Organs 2. Hemocytometer (for cell counting).
for Flow Cytometry
3. Light microscope.
4. Scissors and forceps for dissection.
a 10
4 b 10
4
CD4 SP T cells
CD4SP 10
3
10
3
2 2
10 10
B cells
B cells,
1 1
MF, NK, CD8SP 10 CD8 10
etc. SP
CD3e
CD4
CD4
0 0
10 10
0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10
CD8 CD8 B220
c 1500 1500
CD4SP CD8SP
1000 1000
500 500
# of Cells
0 0
0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10
TCRb
Fig. 7 T cell population in the periphery. Lymph node cells are stained with anti-CD4, anti-CD8α, anti-CD3ε,
and anti-TCRβ or anti-B220 (CD45R). (a) CD4 SP, CD8 SP, and CD4−CD8− cells in the lymph nodes. (b) Peripherial
T and B cells can be differentiated by the expression of CD3ε and B220, respectively. The majority of CD4−CD8−
cells in the lymph nodes are B cells with a very small population of NKT, macrophages, etc. (c) TCRβ levels are
high on CD4 SP and CD8 SP cell populations
5. Cell Staining Buffer: 1× PBS or 1× HBSS, 0.1 % bovine serum

albumin (Fraction V), 0.01 % Sodium Azide (optional), 4 °C.
6. RPMI 1640 medium, 4 °C.
7. 60 mm × 15 mm tissue culture dishes.
8. 80–100 μm nylon mesh, 1″ × 1″ squares.
9. 15 mL conical tubes.
10. 3 mL syringes, single use only.
2.2 Cell-Surface or 1. FcγIII Receptor (CD16) clone 2.4G2 purified.

Intracellular Staining 2. Various antibodies to detect lymphocyte markers (e.g., anti
with Fluorochrome- CD4, anti-CD8, anti-TCRb, anti-CD24, anti-CD69, anti-
Conjugated Antibodies CD5) or other molecules of interest conjugated with different
fluorochromes appropriate for use on your flow cytometer.
Commonly used fluorochromes include fluorescein isothiocya-
nate (FITC), phycoerythrin (PE), peridinin-chlorophyll-protein
a 4
10
CD4 SP
3
10
2
10
1
10 CD8
SP
CD4
0
10
0 1 2 3 4
10 10 10 10 10
CD8
b 4
CD4 SP 4
CD8 SP
10 10
Central 3 3
Naïve 10 10
Memory
2 2
10 10
Effector 1
10 1
10
Memory
CD62L
CD62L
0 0
10 10
0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10
CD44 CD44
Fig. 8 Naïve and memory CD4 SP and CD8 SP cells in the periphery. Lymph node cells are stained with anti-
CD4, anti-CD8α, anti-CD62L, and anti-CD44. (a) Focusing on CD4 SP and CD8 SP cells in the lymph nodes. (b)
Gating on either CD4 SP or CD8 SP lymph node T cells, the naïve and memory (central and effector, respec-
tively) phenotypes can be differentiated by CD62L vs. CD44
complex (PerCP or PerCy-Cy5.5 tandem dye), and allophyco-

cyanin (APC) although others are also available.
3. 7-aminoactinomycin-D (7AAD) (see Note 3).
4. α-galactosylceramide or PBS-57 complexed to CD1d tetramer
(NIH Tetramer Core Facility).
5. anti-FoxP3 (clone FJK-16s) multiple fluorochrome
conjugations.
6. Flow Cytometer (e.g., BD Biosciences FACSCalibur™) that
detects Forward Scatter (FSC), Side Scatter (SSC), and a mini-
mum of four different fluorochromes.
7. Cell Staining Buffer: 1× PBS or 1× HBSS, 0.1 % bovine serum
albumin (Fraction V), 0.01 % Sodium Azide (optional), 4 °C.
8. FoxP3/Transcription Factor Staining Buffer Set (eBioscience
#00-5523-00).
9. Sample tubes, 5 mL polystyrene round bottom (Falcon
#352052) (see Note 10).
10. 7AAD staining buffer: 0.14 M NaCl, 2.5 mM CaCl2, and
0.01 M HEPES in 1× HBSS or 1× PBS.
a 10
4
CD4 SP
10
3
DP
2
10
CD8 SP
1
10
DN
CD4
0
10
0 1 2 3 4
10 10 10 10 10
CD8
b 4
DN 4 DP 4 CD4 SP 4 CD8 SP
10 10 10 10
3 3 3 3
10 10 10 10
2 2 2 2
10 10 10 10
1 1 1 1
10 10 10 10
CD3e
0 0 0 0
10 10 10 10
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
Empty CD1d Tetramer

4 4 4 4
10 10 10 10
3 3 3 3
10 10 10 10
2 2 2 2
10 10 10 10
1 1 1 1
10 10 10 10
CD3e
0 0 0 0
10 10 10 10
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
a-Gal/Cer CD1d Tetramer

c 1000 d 4 4
10 10
800 3 3
10 10
600
2 2
10 10
400
1 1
10 10
200
CD3e
FSC
CD3e
0 0 0
10 10
0 1 2 3 4
10 10 10 10 10 0 1 2 3 4 0 1 2 3 4
10 10 10 10 10 10 10 10 10 10
CD8/CD19/B220 Empty CD1d a-Gal/Cer CD1d

Tetramer Tetramer
Fig. 9 Detection of Natural Killer T cells (NKT) in the thymus and periphery. Thymocytes are stained with anti-
CD4, anti-CD8α, anti-CD3ε, and α-Gal/Cer CD1d tetramer or “empty” CD1d tetramer. Splenocytes are stained
with anti-CD4, anti-CD3ε, α-Gal/Cer CD1d tetramer or “empty” CD1d tetramer, and an antibody mixture of
anti-CD8α + anti-CD19 + anti-B220. (a) Focusing on all populations of the thymus “bird” profile. (b) CD3ε vs.
α-Gal/Cer CD1d (or “empty” CD1d) tetramer surface expression on DN, DP, CD4 SP, and CD8 SP populations in
the thymus. (c) Gating on CD8, CD19, and B220 negative cells in the spleen. (d) CD3ε vs. α-Gal/Cer CD1d (or
empty CD1d) tetramer surface expression on splenocytes that are CD8/CD19/B220 negative. Expected % of
CD1d+ cells in a C57BL/6 mouse thymus: <1 % in DP and CD8 SP and ~2–3 % in DN and CD4 SP, respectively.
Expected % of CD1d+ cells in the spleen is ~3 %
a CD4 SP CD4 SP
104 104 104
CD4 SP DP CD25+
CD4 SP FoxP3+
103 103 103
FoxP3+
102 102 102
CD8 SP
101 101 101
DN
CD25
CD4
CD4
100 100 100
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
CD8 FoxP3 FoxP3 CD25 -

FoxP3+
b CD4 SP
CD4 SP 104
104 104
CD25+
CD4 SP
103
FoxP3+
103 103
102 102
102
Conv. CD4 SP
CD4 SP FoxP3+ 101
101 CD8 101
SP CD25
CD4
CD4
100 100
100
100 101 102 103 104 100 101 102 103 104
100 101 102 103 104
FoxP3 FoxP3 CD25-

CD8 CD4-CD8- FoxP3+
Fig. 10 Regulatory T cells (Treg) in the thymus and periphery. Thymocytes and lymph node cells are surface
stained with anti-CD4, anti-CD8α, anti-CD25, and then permeabilized and stained intracellularly with anti-
FoxP3 (see Note 2). (a) Focusing on the CD4 SP population on the thymus “bird” profile, FoxP3 expression on
CD4 SP and FoxP3 vs. CD25 expression gated on the CD4 SP population. (b) Focusing on the CD4 SP popula-
tion in peripheral (lymph node) cells, FoxP3 expression on CD4 SP and FoxP3 vs. CD25 expression gated on
the CD4 SP population. Expected % of FoxP3+ cells in the CD4 SP population of a C57BL/6 mouse thymus and
lymph nodes are ~5 % and ~12–18 %, respectively
3 Methods
3.1 Removal 1. Euthanize mouse according to institutional guidelines and

of Lymphoid Organs remove desired lymphoid organs: thymus, spleen and/or
and Preparation lymph nodes. Please refer to J.P. Reeves and P.A. Reeves,
of Cells for Antibody- Removal of Lymphoid Organs, Current Protocols in
Fluorochrome Staining Immunology (1991) 1.9.1–1.9.3 for additional guidance.
2. Place the thymus, spleen, or lymph nodes (LNs) between two
squares of nylon mesh in separate tissue culture dishes that
contain 1 mL of RPMI 1640 medium. For LNs, combine all
a 104 10
4
CD3e+ TCRgd +
10
3
103
2
102 10
Lineage Mix
101
Lineage 10
1
Negative
TCRgd
Cells 100
100
0 1 2 3 4
0 200 400 600 800 1000 10 10 10 10 10
Forward Scatter (FSC) CD3e

b 104 10
4
10
3 103 CD3e+ TCRgd +
102 102
Lineage Mix
101 Lineage 101

Negative
TCRgd
Cells 0
100 10
0 200 400 600 800 1000 100 101 102 103 104
Forward Scatter (FSC) CD3e
Fig. 11 γδ T cells in the thymus and periphery. Thymocytes are stained with anti-CD3ε, anti-TCRγδ, and lin-
eage mix that contains anti-CD4, anti-CD8α, anti-TCRβ, anti-CD19, anti-NK1.1, anti-DX5, anti-Gr1 (Ly6C/G),
anti-CD11b, and anti-Ter119. Lymph node cells are stained with anti-CD3ε, anti-TCRγδ, and a lineage mix that
contains anti-CD4, anti-CD8α, and anti-TCRβ. (a) Gating on thymocytes that are lineage negative, TCRγδ+ cells
can be detected as CD3ε+TCRγδ+. Because of their low numbers, it is advantageous to use a lineage mix and
acquire multiple events on the flow cytometer. (b) Gating on lymph node cells that are lineage negative,
TCRγδ+ lymphocytes can be detected as CD3ε+TCRγδ+. Expected % of TCRγδ+ cells in a C57BL/6 mouse:
~1–2 % in the thymus and ~3 % in the lymph nodes
eight individual LNs (two each of inguinal, brachial, axillary,

and superficial cervical) in one tissue culture dish.
3. Gently tease the cells out of the organs using the flat end of
curved forceps or the plunger end of a 3 mL syringe.
(Recommend forceps for the thymus and 3 mL syringe for the
LNs and spleen). The nylon mesh will trap most of the debris.
4. Wash the dish by adding 5 mL of RPMI 1640 medium and
gently pipet up and down. Dispense the medium containing
the cells into a 15 mL conical tube.
5. Rinse the dish of any remaining cells with another 5 mL of

RPMI 1640 medium. Pipet the medium into the same 15 mL
conical tube for a total of 11 mL.
6. Centrifuge for 5 min at 450 × g at 4 °C.
7. Aspirate or decant excess medium, taking care not to disturb
the cell pellet at the bottom of the tube.
8. Wash cells again with 10 mL of RPMI 1640 medium, centri-
fuge, and aspirate medium.
9. Resuspend the cell pellet in 1 mL of RPMI 1640 medium.
10. Count cells and store at 4 °C for staining.
11. Expected cell numbers for lymphoid organs from one
C57BL/6 mouse (~6–8 weeks old): ~100 million thymocytes,
~80–100 million splenocytes, and ~20 million lymphocytes
(LNs).
3.2 Cell Surface 1. Add one million cells in a sample tube (see Note 4).
Staining 2. Add 2 mL of cell staining buffer to the tube and centrifuge for
with Fluorochrome 5 min at 450 × g, 4 °C.
Conjugated Antibodies
3. Vacuum-aspirate excess buffer, leaving cell pellet undisturbed
at the bottom of the tube. Excess buffer can also be decanted.
4. Stain with antibody mix* (see below) for 30 min at 4 °C in the
dark (see Notes 5 and 6). Taking into account residue buffer
from the cell pellet, final volume of cells AND antibody mix
should be between 50 and 100 μL.
*Composition of Antibody Mix:
10 μL of FcγIII Receptor (CD16) clone 2.4G2 (titered)
10 μL of FITC labeled antibody (titered)
10 μL of PE labeled antibody (titered)
10 μL of PerCp-Cy5.5 antibody (titered)
10 μL of APC antibody (titered)
50 μL Total volume
5. Add 2 mL of cell staining buffer to the tube and centrifuge for

5 min at 450 × g, 4 °C.
6. Vacuum-aspirate excess buffer, leaving cell pellet undisturbed
at the bottom of the tube. Excess buffer can also be decanted.
7. Repeat steps 5 and 6.
8. Resuspend cells in 300 μL of cell staining buffer.
9. Cover with aluminum foil to protect from light and store at
4 °C until ready to run on the flow cytometer (see Notes 7–9).
10. If using 7AAD to exclude dead cells from a sample, start
by washing ~1 million cells with 2 ml of 1× PBS (4 °C).
Centrifuge for 5 min at 450 × g and vacuum-aspirate excess

buffer. Repeat for a total of two washes. Resuspend cell pellet
in 100 μL of 7AAD staining buffer. Add 0.25–1 μg of 7AAD
per sample and incubate for 15 min at RT, protecting the
sample from light. At the end of the incubation, add 400 μL
of the 7AAD staining buffer directly into the sample tube.
Do not wash 7AAD out of the sample tube. Run samples
within 1 hr. Samples should be kept cold during acquisition.
Since 7AAD is detected in the PerCp/PerCp-Cy5.5 (FL3)
channel, leave that channel “open” when surface staining.
11. If running CD1d tetramer in a sample follow cell-staining
instructions steps 1–9, substituting CD1d tetramer for one of
the fluorochrome conjugated antibodies.
3.3 Intracellular 1. Perform surface staining (Subheading 3.2, steps 1–7).

Staining for the FoxP3 2. To the cell pellet, add 100 μL of 1× Fixation buffer (eBiosci-
Transcriptional Factor ence #00-5523-00). Pipet up and down to resuspend the cell
pellet in buffer. Incubate samples for 30 min at room tempera-
ture in the dark (see Note 11).
3. Add 1 mL of 1× permeabilization buffer to the tube, pipet up
and down once to mix, and pellet the cells for 5 min at 450 × g,
4 °C. Aspirate excess buffer. The cells will be translucent.
4. Repeat step 3.
5. Add 1 μg of fluorochrome-conjugated anti-FoxP3 antibody
(eBioscience) to the cells in a total volume of 100 μL of per-
meabilization buffer. Incubate for 30 min at room temperature
in the dark.
6. Add 1 mL of 1× permeabilization buffer to the tube and
centrifuge for 5 min at 450 × g, 4 °C.
7. Aspirate excess buffer off, taking care not to disturb the cell
pellet on the bottom of the tube.
8. Resuspend cells in 300 μL of cell staining buffer.
9. Cover with aluminum foil to protect from light and store at
4 °C until ready to run on the flow cytometer (see Notes 7–9).
4 Notes
1. It is highly recommended to include a lineage mix (Fig. 3)

when analyzing the DN population in the thymus. This allows
for the removal of all lineage positive cells, leaving the lineage
negative cells more visible. This and the acquisition of multiple
events on the flow cytometer allows for the visualization of the
DN1–DN4 profiles. The negative cells are also slightly larger
in size (larger FSC profile) than conventional thymocytes. This

becomes important when setting “live cell” and collection
gates (Fig. 12).
a 4
b 10
4
10
7AAD+ 7AAD+
3
10
3 10
2
10
2 10
1
10
1 10
7AAD
7AAD
7AAD- 0
7AAD-
10
0 10
0 200 400 600 800 1000 0 200 400 600 800 1000
Forward Scatter (FSC) Forward Scatter (FSC)

c
7AAD+ 7AAD - No 7AAD gate
1000 1000 1000
Side Scatter (SSC)
800 800 800
600 600 600
400 400 400
200 200 200
0 0 0
0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000
Forward Scatter (FSC)

d 1000
Live pop
Side Scatter (SSC)
gate
800
600
400
200
0
0 200 400 600 800 1000
Forward Scatter (FSC)

Fig. 12 7AAD and live gating. Thymocytes are stained with 7AAD. (a) Thymocytes stored at 4 °C, measuring
7AAD expression ex vivo. (b) Thymocytes stored at room temperature for 48 h ex vivo, showing 7AAD expres-
sion. (c) 7AAD positive (“dead”) cells are small (low SSC, low FSC profile). 7AAD negative (“live”) cells are
larger (larger FSC profile). Cells that do not have a 7AAD gate applied show a dual population of “dead” cells
(lower FSC) and “live” cells (larger FSC). These thymocytes are stained with 7AAD ex vivo and stored at 4 °C.
(d) The application of a “live” gate on SSC vs. FSC (on the No 7AAD profile from (c))
2. Refer to Subheading 3.3 for intracellular staining of the FoxP3

transcriptional factor. CD25 surface expression can be used as
a substitute marker for regulatory T cells even though not all
CD25 positive cells are FoxP3 positive.
3. 7-amino-actinomycin (7AAD) is a fluorescent reagent used to
exclude nonviable cells that incorporate 7AAD into the GC-
rich regions of double stranded DNA through compromised
plasma membranes. 7AAD is most commonly excited by the
488 nm laser. 7AAD is detected in the FL3 channel (using a
670 nm long pass filter) and can be compensated out of the
FL2 (PE) channel allowing for multi-color flow cytometry.
There are other alternatives to 7AAD. Live/dead or via-
bility stains/dyes are commercially available now in multiple
fluorochromes, allowing for multi-color customization and
are useful with cells that will be fixed and permeabilized. DAPI
(4′,6-diamidino-2-phenylindole) and PI (Propidium Iodide)
are two popular “dyes” used to assess cell viability. DAPI is a
fluorescent stain that can be detected by an ultraviolet
(355 nm) laser when bound to the A-T regions in
DNA. Propidium Iodide also intercolates into the DNA, but
nonspecifically. One advantage to PI is that it can be detected
by the blue (488 nm) laser, which is more commonly found
on flow cytometers compared to an ultraviolet laser. Please
also note that both DAPI and PI can also be used for cell
cycling analysis. For further information, please refer to
Z. Darzynkiewicz, G. Juan and E. Bedner, Determining Cell
Cycle Stages by Flow Cytometry, Current Protocols in Cell
Biology (1999) 8.4.1–8.4.18.
In some cases when 7AAD cannot be included in the anti-
body staining, one can use a “live” gate method where the
majority of “dead” cells are excluded in the Side Scatter (SSC)
vs. Forward Scatter (FSC) profile. This “live” gate is used only
with cells that are stained ex vivo and stored at 4 °C (Fig. 12).
4. To stain more than one million cells in a single sample tube,
scale up the amount of antibody mix as needed. For example,
if staining two million cells, double the antibody mix and so
on. This is very useful when collecting rare cell populations.
5. Titer all fluorochrome conjugated antibodies used for cell
staining. Most antibodies that are commercially obtained can
be used at lower concentrations than recommended by the
vendor. Using less will limit background fluorescence and save
money/reagents in the long run.
6. Because fluorochromes are light sensitive, it is recommended
that all fluorochrome conjugated antibodies are stored in the
dark at 4 °C. Cells labeled with fluorochrome conjugated
antibodies should be covered with aluminum foil at every incu-
bation step. Fluorochromes, especially tandam dyes, can lose
intensity or even dissociate when exposed to light.
7. Unstained, single color, and FMO (fluorescence minus one)

compensation controls should be used to set up the flow
cytometer. This will help with issues pertaining to background
and auto-fluorescence, compensation, negative/positive gate
settings, and expose any issues with the actual antibody stain-
ing procedure. If cells are stained intracellularly (e.g., for
FoxP3), the set of compensation controls should also undergo
the same fixation and permeabilization procedure as the exper-
imental samples. Cells that are fixed/permeabilized exhibit a
different forward scatter (FSC) vs. side scatter (SSC) profile
compared to non-fixed/non-permeabilized cells.
8. To prevent clogging in the flow cytometer by cell samples, fil-
ter cells with the pre-cut nylon mesh right before event acqui-
sition. Alternatively, a 5 mL sample tube that has a “cell-strainer”
cap can also be used to filter out cell clumps (e.g., Falcon
#352235).
9. Fluorochrome conjugated antibody stained cell samples that
are in cell staining buffer can be stored overnight for analysis
by flow cytometry the next day. Although not optimal, this is
sometimes necessary especially when faced with multiple cell
groups and long antibody staining procedures. Sample tubes
should be covered with aluminum foil and stored at 4 °C in the
dark. If samples must be held longer than overnight, samples
can be fixed with a commercial fixation reagent or 2 %
paraformaldehyde.
10. Falcon (#352052) 5 ml tubes fit the majority of flow cytome-
ters from BD Biosciences. If using other flow cytometers,
please refer to their specifications for plastic disposable wear.
11. It is essential to use the fixation/permeabilization kit from
eBioscience formulated for the specific staining of FoxP3.
Using other fix/perm methods might result in sub-optimal or
no staining of FoxP3. Please also note that this fix/perm kit for
FoxP3 can also be used for detection of other nuclear tran-
scription factors. Which nuclear transcription factors will
require testing on the user’s part.
References
1. Anderson G, Moore NC, Owen JJ, Jenkinson entiations. Annu Rev Cell Dev Biol 17:
EJ (1996) Cellular interactions in thymocyte 387–403
development. Annu Rev Immunol 14:73–99 4. Godfrey DI, Kennedy J, Suda T, Zlotnik A
2. Petrie HT (2003) Cell migration and the con- (1993) A developmental pathway involving
trol of post-natal T-cell lymphopoiesis in the four phenotypically and functionally distinct
thymus. Nat Rev Immunol 3(11):859–866 subsets of CD3-CD4-CD8- triple-negative
3. Weissman IL, Anderson DJ, Gage F (2001) adult mouse thymocytes defined by CD44
Stem and progenitor cells: origins, pheno- and CD25 expression. J Immunol 150(10):
types, lineage commitments, and transdiffer- 4244–4252
5. Shortman K, Wu L (1996) Early T lymphocyte 10. Sprent J, Surh CD (2002) T cell memory.
progenitors. Annu Rev Immunol 14:29–47 Annu Rev Immunol 20:551–579
6. Von Boehmer H (1995) Control of T-cell 11. Bendelac A, Savage PB, Teyton L (2007) The
development by the pre-T and alpha beta biology of NKT cells. Annu Rev Immunol
T-cell receptor. Ann N Y Acad Sci 766:52–61 25:297–336
7. Michie AM, Zuniga-Pflucker JC (2002) 12. Sakaguchi S (2004) Naturally arising CD4+
Regulation of thymocyte differentiation: pre- regulatory t cells for immunologic self-
TCR signals and beta-selection. Semin tolerance and negative control of immune
Immunol 14(5):311–323 responses. Annu Rev Immunol 22:
8. Jameson SC, Hogquist KA, Bevan MJ (1995) 531–562
Positive selection of thymocytes. Annu Rev 13. Hayday AC (2000) [gamma][delta] cells: a
Immunol 13:93–126 right time and a right place for a conserved
9. Sebzda E et al (1999) Selection of the T cell third way of protection. Annu Rev Immunol
repertoire. Annu Rev Immunol 17:829–874 18:975–1026
Chapter 5
Flow Cytometry Analysis of Thymic Epithelial Cells

and Their Subpopulations
Izumi Ohigashi and Yousuke Takahama
Abstract
The parenchyma of the thymus is compartmentalized into the cortex and the medulla, which are
constructed by cortical thymic epithelial cells (cortical TECs, cTECs) and medullary thymic epithelial cells
(mTECs), respectively. cTECs and mTECs essentially and differentially regulate the development and
repertoire selection of T cells. Consequently, the biology of T cell development and selection includes the
study of TECs in addition to the study of developing T cells and other hematopoietic cells including den-
dritic cells. In this chapter, we describe the methods for flow cytometric analysis and sorting of TECs and
their subpopulations, including cTECs and mTECs.
Key words Thymic epithelial cell, mTEC, cTEC, Aire, β5t, Flow cytometry, Cell sorting
1 Introduction
The thymic parenchyma provides a three-dimensional thymic

microenvironment that supports the development of T cells.
Cortical thymic epithelial cells (cTECs) are the major stromal com-
ponent of the thymic cortex and support the early stages of T cell
development as well as the positive selection of T cells. On the
other hand, medullary thymic epithelial cells (mTECs) are the major
stromal component of the thymic medulla and support the nega-
tive selection of autoreactive T cells as well as the generation of
regulatory T cells [1–3].
To better understand how T cells are generated and selected
in the thymus, studies that focus on thymic epithelial cells (TECs)
are as important as studies on developing T cells and other hema-
topoietic cells, including dendritic cells. It has been revealed that
a number of molecules expressed in TECs play essential roles in T
cell development and selection. Those molecules include IL-7 [4,
5], DLL-4 [6, 7], β5t-containing thymoproteasome [8, 9], Aire
[10, 11], CCR7 ligand chemokines [12, 13], and self-peptide-
presenting MHC molecules [14].
65
66 Izumi Ohigashi and Yousuke Takahama
In order to further understand the molecular and cellular char-

acteristics of TECs and their subpopulations at the single-cell level,
flow cytometry analysis and cell sorting have been adopted [15,
16]. In this chapter, we will describe methods for the dissociation
of thymic stromal cells (Subheading 3.1). Flow cytometric analysis
of TECs, including mTECs and cTECs, as well as their intracellu-
lar molecules will be discussed in Subheadings 3.2 and 3.3, respec-
tively, and methods for sorting TECs will be described in
Subheading 3.4.
2 Materials
2.1 Dissociation 1. Euthanasia chamber.
of Stromal Cells 2. Scissors and forceps.
from Thymus
3. PBS, pH 7.2.
4. RPMI 1640 medium.
5. Collagenase solution: RPMI 1640 supplemented with 0.125 %
collagenase D and 0.01 % DNase I (see Note 1).
6. PBS + EDTA buffer: PBS supplemented with 2 mM EDTA and
2 % fetal calf serum (FCS).
7. 30-mm dish.
8. 1.5-ml microtubes.
9. 1-ml syringes.
10. 26-gauge needles.
11. 100-μm nylon mesh.
12. Water bath or block incubator.
2.2 Cell Surface 1. 1.5-ml microtubes.

Staining for Flow 2. Fluorochrome-unlabeled anti-FcγR antibody (clone 2.4G2).
Cytometric Analysis
3. Fluorochrome-conjugated antibodies specific for CD45,
CD205, and CD326 (see Note 2).
4. Biotinylated Ulex europaeus agglutinin-1 (UEA-1) (see Note 2).
5. FACS buffer: PBS supplemented with 0.2 % BSA and 0.1 %
NaN3.
6. FACS-PI buffer: FACS buffer supplemented with 0.1 μg/ml
propidium iodide.
8. Flow cytometer.
2.3 Intracellular 1. Antibodies specific for β5t and Aire (see Note 3).
Staining for Flow 2. FACS buffer (see Subheading 2.2).
Cytometric Analysis
3. 2 % Paraformaldehyde: PBS supplemented with 2 %
paraformaldehyde.
Flow Cytometry Analysis of Thymic Epithelial Cells and Their Subpopulations 67
4. Saponin buffer: PBS supplemented with 0.1 % saponin.

6. Flow cytometer.
2.4 Sorting 1. Anti-CD45-conjugated microbeads (Miltenyi Biotec).

of Thymic Epithelial 2. MACS buffer: PBS supplemented with 0.5 % BSA and 2 mM
Cells EDTA.
3. MACS LS column (Miltenyi Biotec).
4. MACS separation unit (Miltenyi Biotec).
5. MACS multistand (Miltenyi Biotec).
6. FACS buffer (see Subheading 2.2).
8. Cell sorter.
3 Methods
3.1 Dissociation 1. Prepare a 30-mm dish containing 1 ml of PBS.

of Stromal Cells 2. Euthanize mouse according to relevant institutional guidelines
from Thymus (see Note 4).
3. Wipe the chest of the mouse with 70 % ethanol, open the chest,
and remove the thymus using a set of scissors and forceps.
4. Transfer the thymus into the 30-mm dish containing
PBS. Where needed, transfer the thymus to another PBS-
containing dish to remove blood.
5. Mince the capsules of the thymus with scissors.
6. Transfer the minced thymus to a 1.5-ml microtube containing
1-ml of RPMI medium.
7. Pipette gently with a wide-bore tip to mechanically release thy-
mocytes (see Note 5).
8. Remove the medium containing thymocytes released from the
thymus.
9. Add 1 ml of RPMI medium.
10. Repeat the gentle pipetting with the wide-bore tip.
11. Remove the medium.
12. Add 1 ml of collagenase solution.
13. Incubate at 37 °C for 15 min. Gently pipette every 5 min dur-
ing the incubation (see Note 6).
14. Collect the released cells, add 1 ml of PBS + EDTA buffer, and
keep on ice.
15. Repeat the collagenase treatment as specified in steps 12–14
(see Note 7).
16. Mix the cells collected at steps 14 and 15.

17. Centrifuge at 200 × g for 5 min at 4 °C and remove the
supernatant.
18. Resuspend cells in 1 ml of PBS + EDTA buffer.
19. Filter the cells through a 100-μm nylon mesh.
3.2 Cell Surface 1. Prepare cell suspension as in Subheading 3.1.

Staining for Flow 2. Centrifuge at 200 × g for 5 min at 4 °C and remove
Cytometric Analysis supernatant.
3. Resuspend cells in 1 ml of FACS buffer.
4. Perform a cell count and transfer 1 × 107 cells into a 1.5-ml
microtube.
5. Centrifuge at 200 × g for 5 min at 4 °C and remove
supernatant.
6. Incubate cells with unlabeled anti-FcγR antibody on ice for
5 min.
7. Add antibodies specific for CD45 and CD326. Where needed,
UEA-1 and anti-CD205 antibody are also added. Incubate the
mixture on ice for 30 min.
8. Add 1 ml of FACS buffer.
9. Centrifuge at 200 × g for 5 min at 4 °C and remove the
supernatant.
10. Where needed, further stain the cells with additional antibod-
ies or secondary antibodies on ice for 30 min.
supernatant.
13. Resuspend cells in 300 μl of FACS-PI buffer.
15. Analyze the cells with a flow cytometer. Representative results

are shown in Fig. 1.
3.3 Intracellular 1. Prepare cell suspension as in Subheading 3.1.

Staining for Flow 2. Perform cell surface staining as in Subheading 3.2 (steps
Cytometric Analysis 1–12).
3. Resuspend cells in 200 μl of 2 % paraformaldehyde.
4. Incubate at room temperature for 30 min.
6. Centrifuge at 200 × g for 5 min at 4 °C and remove superna-
tant (see Note 8).
7. Resuspend cells in 1 ml of saponin buffer.
Fig. 1 Representative flow cytometry profiles of TECs. Collagenase-digested thymus cells from 3-week-old
C57BL/6 mice (as in Subheading 3.1) were stained with antibodies specific for CD326 (PECy7), CD45 (FITC),
and CD205 (APC), as well as with UEA-1 (APCeFluor780) and propidium iodide (PI) (as in Subheading 3.2). (a)
Profiles show CD45 and CD326 expression in PI− total viable cells (left) and CD205 and UEA-1 expression in
CD45−CD326+ PI− viable cells (middle) and CD45+CD326+ PI− viable cells (right) (see Note 11). Numbers in
dot plots indicate frequency within indicated area. Most CD45−CD326+ PI− viable cells consisted of either
_
UEA1–CD205+ cTECs or UEA1+CD205 mTECs. CD45+CD326+ PI− viable cells consisted of thymocyte-associ-
ated cTECs and thymocyte-associated mTECs (see Note 11) as well as many UEA1–CD205– cells. (b) Profiles
show class II MHC (I-Ab) expression (solid lines) and isotype control (shaded histograms) in CD205−UEA-1+
UEA1+CD205– PI− viable mTECs (left) and UEA1–CD205+ CD45−CD326+ PI− viable cTECs (right). The class II
MHC expression indicates that the cells indeed represent TEC subpopulations
8. Centrifuge at 200 × g for 5 min and remove supernatant.

9. Stain cells with antibody specific for the intracellular molecule
on ice for 30 min.
10. Add 1 ml of saponin buffer.
11. Centrifuge at 200 × g for 5 min and remove supernatant.
12. Repeat washing as described in steps 10 and 11 (see Note 9).
Fig. 2 Representative flow cytometry profiles of intracellular molecules expressed

in TECs. The profiles show β5t expression (solid lines in (a)) and Aire expression
(solid lines in (b)) in UEA1+CD205– CD45−CD326+ mTECs (left panels ) and
UEA1–CD205+ CD45−CD326+ cTECs (right panels ). Shaded histograms indicate
profiles obtained by staining with isotype control antibodies

14. Analyze the cells with a flow cytometer. Representative results
are shown in Fig. 2.
3.4 Sorting 1. Prepare cell suspension as in Subheading 3.1.

of Thymic Epithelial 2. Centrifuge at 200 × g for 5 min at 4 °C and remove
Cells supernatant.
3. Resuspend cell pellet in 90 μl of MACS buffer and 10 μl of
anti-CD45-microbeads per 1 × 107 cells. Incubate for 30 min
at 4 °C.
4. Add 1 ml of MACS buffer.

supernatant.
7. Resuspend cell pellet in 1 ml of MACS buffer.
8. Place the MACS separation unit on the MACS multistand.
Expose the MACS LS column to a magnetic field.
9. Rinse the column with 3 ml of MACS buffer.
10. Apply cell suspension onto the column. Collect cells that pass
through the column.
11. Add 3 ml of MACS buffer onto the column and collect cells
that pass through the column.
supernatant.
13. Resuspend cell pellet in 1 ml of FACS buffer.
14. Perform cell surface staining as in Subheading 3.2
(steps 5–14).
15. Sort cells with a cell sorter (see Note 10).
4 Notes
1. Collagenase/dispase (Roche, working concentration at

0.125 %) or liberase TM (Roche, working concentration at 1
unit/ml) may be used instead of Collagenase D. Collagenase/
dispase can dissociate thymic stromal cells more effectively
than Collagenase D. However, the inclusion of these enzymes
weakens subsequent antibody-mediated staining for several
surface molecules, such as CD205.
2. Anti-CD45 and anti-CD326 (EpCAM) antibodies are used for
the detection of leukocytes and epithelial cells, respectively.
Among the TECs, cTECs can be detected by the expression of
CD205 or CD249 (Ly51), whereas mTECs can be detected
by the expression of CD80 or the binding to lectin UEA-1.
PECy5-labeled anti-CD45 antibody, FITC-labeled anti-CD45
antibody, PECy7-labeled anti-CD326 antibody, APC-labeled
anti-CD205 antibody, FITC-labeled CD249 antibody, and
FITC-labeled anti-CD80 antibody can be purchased from
BioLegend. Biotinylated UEA-1 can be purchased from Vector
Laboratories.
3. Anti-β5t and anti-Aire antibodies can be purchased from MBL
and eBioscience, respectively. Saponin buffer should be used to
prepare the working solution.
4. In the collagenase-digested thymus, the frequency of cTECs

exceeds that of mTECs in mice at 2-weeks-old or younger age
but remarkably drops at 3-weeks-old or older age. Therefore,
it is easiest to use mice at 2-weeks-old or younger for the sort-
ing of cTECs.
5. Wide-bore pipette tips (at 1–2 mm diameter) can be made by
cutting 1-ml plastic pipette tips with scissors. The bore size
should not be too large in order to efficiently release thymo-
cytes from the minced thymus.
6. Start pipetting with a wide-bore tip. When the minced thymus
becomes small, change to a regular tip and then a 26G-needle-
attached syringe.
7. Repeat the collagenase treatment until thymus clumps become
invisible. For adult thymus, repeat the treatment two or three
times. For newborn thymus, repeat it once or twice.
8. Cells tend to float after the fixation. When needed, the cells
may be spun at a speed of up to 500 × g.
9. Where needed, cells may be further stained with the secondary
antibodies (as specified in Subheading 3.3, steps 9–11).
10. 100-μm-nozzle should be used for the sorting of TECs.
Collection medium is RPMI supplemented with 30 %
FCS. Before collecting the sorted cells, coat the wall of collec-
tion tube with RPMI 1640 supplemented with 30 % FCS.col-
lection medium.
11. Even after the collagenase treatment, many CD326+ TECs are
still associated with CD45+ thymocytes (Fig. 1a). This tight
association with thymocytes is more prominent in cTECs than
mTECs, reflecting the strong adhesiveness of cTECs to thy-
mocytes [17]. A fraction of cTECs form large multicellular
complexes termed thymic nurse cells (TNCs), which represent
cTECs that completely enclose multiple thymocytes [17]. For
the detection of TNCs by flow cytometry, voltage for the for-
ward scatter channel should be reduced [17]. It should also be
noted that because the majority of cTECs are openly associated
with thymocytes, the depletion of CD45+ cells during the pro-
cedure for the enrichment of TECs (Subheading 3.4) actually
removes the majority of cTECs [17].
References
1. Petrie HT, Zúñiga-Pflücker JC (2007) Zoned 3. Anderson G, Takahama Y (2012) Thymic epi-
out: functional mapping of stromal signaling thelial cells: working class heroes for T cell
microenvironments in the thymus. Annu Rev development and repertoire selection. Trends
Immunol 25:649–679 Immunol 33:256–263
2. Klein L, Hinterberger M, Wirnsberger G, 4. Alves NL, Goff OR, Huntington ND (2009)
Kyewski B (2009) Antigen presentation in the Characterization of the thymic IL-7 niche
thymus for positive selection and central toler- in vivo. Proc Natl Acad Sci U S A 106:
ance induction. Nat Rev Immunol 9:833–844 1512–1517
5. Hara T, Shitara S, Imai K et al (2012) 12. Ueno T, Saito F, Gray DH et al (2004) CCR7
Identification of IL-7 producing cells in primary signals are essential for cortex-medulla migra-
and secondary lymphoid organ using IL7-GFP tion of developing thymocytes. J Exp Med
knock-in mice. J Immunol 189:1577–1584 200:493–505
6. Hozumi K, Mailhos C, Negishi N et al (2008) 13. Liu C, Saito F, Liu Z et al (2006) Coordination
Delta-like 4 is indispensable in thymic environ- between CCR7- and CCR9-mediated chemo-
ment specific for T cell development. J Exp kine signals in prevascular fetal thymus coloni-
Med 205:2507–2513 zation. Blood 108:2531–2539
7. Koch U, Fiorini E, Benedito R et al (2008) 14. Laufer TM, DeKoning J, Markowitz JS et al
Delta-like 4 is the essential, nonredundant (1996) Unopposed positive selection and
ligand for Notch1 during thymic T cell lineage autoreactivity in mice expressing class II MHC
commitment. J Exp Med 205:2515–2523 only on thymic cortex. Nature 383:81–85
8. Murata S, Sasaki K, Kishimoto T et al (2007) 15. Gray DH, Chidgey AP, Boyd RL (2002)
Regulation of CD8+ T cell development by Analysis of thymic stroma cell populations
thymus-specific proteasomes. Science using flow cytometry. J Immunol Methods
316:1349–1353 260:15–28
9. Nitta T, Murata S, Sasaki K et al (2010) 16. Williams KM, Mella H, Lucas PJ et al (2009)
Thymoproteasome shapes immunocompetent Single cell analysis of complex thymus stroma
repertoire of CD8+ T cells. Immunity 32:29–40 cell populations: rapid thymic epithelial prepa-
10. Nagamine K, Peterson P, Scott HS et al (1997) ration characterizes radiation injury. Clin
Positional cloning of the APECED gene. Nat Transl Sci 2:279–285
Genet 17:393–398 17. Nakagawa Y, Ohigashi I, Nitta T et al (2012)
11. Anderson MS, Venanzi ES, Klein L et al Thymic nurse cells provide microenvironment
(2002) Projection of an immunological self for secondary TCRα rearrangement in cortical
shadow within the thymus by the Aire protein. thymocytes. Proc Natl Acad Sci U S A 109:
Science 298:1395–1401 20572–20577
Chapter 6
Identifying the Spatial Relationships of Thymic Stromal

and Thymocyte Subsets by Immunofluorescence Analysis
Virginia Bain and Ellen R. Richie
Abstract
Immunofluorescence analysis of thymic tissue sections is an indispensable technique for visualizing spatial
relationships among thymocyte and stromal cell subsets. The thymus is organized into distinct microenvi-
ronmental zones in which particular thymic epithelial cell (TEC) subsets support specific stages of thymo-
cyte maturation. Conversely, thymocytes and lymphoid tissue inducer cells support functional maturation
of TECs. The composition and organization of TECs change during ontogeny to generate a maximally
functional organ in the young adult. Deterioration of thymic architecture and stromal organization occurs
with age as the thymus undergoes involution. Such changes can be monitored by immunofluorescent
staining of thymic sections obtained at different ages throughout the life-span. Here we describe methods
to generate frozen or paraffin-embedded thymic tissue sections for multicolor immunofluorescence stain-
ing using antibodies to surface and/or cytoplasmic antigens.
Key words Immunofluorescence, Thymus, Thymocytes, Thymic epithelial cells, Thymic

microenvironment
1 Introduction
The thymus is composed of two primary cell types: thymocytes and

stromal cells that form a unique and indispensable microenviron-
ment required for T cell development. Thymic epithelial cells
(TECs) comprise the major type of thymic stromal cells and pro-
vide growth, differentiation, and survival signals to thymocytes.
Reciprocally, thymocyte-derived signals are required for TEC
development. Thus, cross talk between thymocytes and TECs is
essential for development of both cell types (reviewed in [1–4]).
The thymus is spatially organized into thymocyte-dense corti-
cal regions, thymocyte-sparse medullary regions, the corticome-
dullary junction (CMJ), and a subcapsular region. Bone marrow
hematopoietic progenitors enter the thymus via blood vessels at
the CMJ. As thymocytes differentiate from CD4−CD8− precursors
75
76 Virginia Bain and Ellen R. Richie
to the CD4+CD8+ stage, they migrate through the cortex to the

subcapsular region, then reverse course, and traverse the cortex
again where cortical TECs (cTECs) present peptide in the context
of self-MHC to positively select thymocytes with self-MHC-
restricted T cell receptors (TCRs). Positively selected CD4+CD8−
and CD4−CD8+ thymocytes migrate into the medulla where they
encounter medullary TECs (mTECs) that express a diverse array of
tissue-restricted antigens required for deletion of T cells expressing
high-avidity autoreactive TCRs. This chemokine-directed migra-
tion pattern brings thymocytes at successive stages of maturation
into close contact with phenotypically and functionally distinct
subsets of cTECs and mTECs required for development of self-
restricted and self-tolerant T cells [5, 6].
Thymic architecture, organization, and spatial relationships
among thymocytes, TECs, and non-epithelial stromal cells can be
visualized by immunofluorescence analysis of thymic tissue sec-
tions. This chapter outlines thymus embedding and immunofluo-
rescence antibody (IFA) staining techniques that identify TEC and
thymocyte subsets in thymus sections. IFA analysis is an efficient
way to define mutant thymic phenotypes and to assess changes in
the developing and aging thymus. There are three general meth-
ods to prepare the thymus for IFA staining: (1) snap freeze tissue
(referred to as fresh frozen); (2) briefly fix tissue in 4 % PFA, dehy-
drate, and then freeze (referred to as fixed frozen); and (3) fix tis-
sue in 4 % PFA for a longer duration, dehydrate the tissue, infiltrate
with xylenes, and paraffin embed the tissue (referred to as paraffin
embedded). The method of choice is dictated by several factors,
including the objective of the analysis. Better discrimination of
cytoplasmic and surface proteins is obtained using fresh-frozen
thymi as opposed to fixed tissues (Fig. 1). However, for morpho-
logical studies especially hematoxylin and eosin stains, paraffin-
embedded sections provide better discrimination of the structural
characteristics of different cell types than can be obtained using
fresh-frozen tissue. Another important consideration is the sensi-
tivity of different antigenic epitopes to various fixation techniques.
For example, the vascular marker CD31 is extremely sensitive to
PFA fixation, such that IFA analysis is suboptimal using PFA-fixed
tissue. In contrast, the nuclear transcription factor FOXN1 requires
PFA fixation for optimal staining results (Fig. 2).
2 Materials
All reagents can be stored at room temperature unless indicated

otherwise. Solutions are made using ddH2O. Reagents requiring
autoclaving or a fume hood are noted. Please follow proper dis-
posal procedures for waste containing PFA, ethanol, or xylene.
Immunofluorescence on Thymic Sections 77
Fig. 1 Tissue morphology varies with fixation. Keratin 5 and keratin 8 staining of 4-week thymi. (a) Fresh-
frozen tissue is optimal for cytoplasmic markers such as keratin 5 and keratin 8 but is difficult to distinguish
nuclear morphology. (b) Fixed-frozen tissue gives better morphology but has less optimal cytoplasmic stain
and more tissue defects when sectioning. (c) Paraffin-embedded tissue gives clear nuclear morphology and
fewer tissue defects but requires antigen retrieval and has less optimal cytoplasmic or cell surface stain. Scale
bars indicate 50 μm
Fig. 2 Antibodies may require different tissue preparation. (a) Fresh-frozen tissue is required by some cell
surface markers such as the vascular marker CD31. (b) PFA fixation is required by some nuclear markers such
as the TEC-specific transcription factor Foxn1. Scale bars indicate 50 μm and tissue used is 4-week thymi
2.1 Frozen Tissue 1. Kimwipes.

2. Cryomolds (e.g., disposable base molds 15 × 15 × 5 mm).
3. Embedding medium (e.g., OCT—optimal cutting tempera-
ture compound).
4. 10× PBS pH 7.2.
5. 4 % PFA: 4 % paraformaldehyde by weight dissolved in

PBS. Heat to dissolve. Work in a fume hood (see Note 1).
6. 5 % Sucrose: 5 % sucrose by weight dissolved in PBS. Add
0.5 % sodium azide to prevent bacterial growth. Autoclave.
7. 20 % Sucrose: 20 % sucrose by weight dissolved in PBS. Add
0.5 % sodium azide. Autoclave.
8. Acetone.
2.2 Paraffin- 1. Ethanol: 70–90 % ethanol should be mixed by volume with

Embedded Tissue ddH2O (see Note 2).
2. Xylene: Work in a fume hood (see Note 3).
3. Cassettes for embedding (e.g., biopsy cassettes).
4. Paraffin (e.g., tissue embedding medium, melting point
56 °C).
5. Molds for paraffin embedding (e.g., peel-a-way disposable
embedding mold).
6. AR buffer: 10 mM sodium citrate, 0.05 % Tween 20, HCl to
pH 6 (see Note 4).
2.3 Staining 1. Slides (see Note 5).

2. Slide boxes.
3. Slide mailers (see Note 6).
4. Parafilm M: 2 in. × 250 ft cut into 8 cm strips.
5. 1 % Triton-X: 1 % Triton-X by volume dissolved in ddH2O.
6. Donkey serum.
7. Tyramide Signal Amplification kit.
8. 4′,6-Diamidino-2-phenylindole (DAPI).
9. Mounting media (see Note 7).
10. Cover slips (see Note 8).
2.4 Antibodies 1. Primary antibodies: Antibody selection depends on the cell

type of interest (see Note 9).
2. Secondary antibodies: This will depend on the filter sets for
your microscope and the host of the primary antibody. A stan-
dard four-color fluorescent microscope will have filters for blue
(Alexa350), green (Alexa488), red (Alexa594), and far-red
(Alexa647) (see Note 10).
3 Methods
For the purpose of immunostaining, tissue may be fresh frozen,

fixed frozen, or fixed paraffin embedded. Tissue preparation depends
on desired quality of morphology and antibodies to be used.
3.1 Tissue Dissected thymi from embryos or adult mice as well as whole
Preparation for embryos can be flash frozen for cryosectioning. Embryos and adult
Fresh-Frozen Tissue mice should be euthanized according to institutional guidelines.
1. Dab tissue gently on a Kimwipe to remove excess PBS.
2. Place the tissue in a cryomold and cover in OCT. Rotate the
tissue to make certain that it is completely surrounded by OCT
with no pockets of PBS or air.
3. Use forceps to hold the cryomold in liquid nitrogen. Remove
the mold as soon as the entire block has frozen to avoid crack-
ing (see Note 11).
4. Store frozen blocks at −80 °C until ready to cryosection.
5. Move frozen blocks to the cryostat at least 30 min before cut-
ting to allow the blocks to thaw.
6. Trim the block with a single-edged razor blade to reduce cut-
ting area.
7. Mount the blocks to specimen discs using enough OCT to
surround the base of the block on all sides and allow the addi-
tional OCT to completely freeze before sectioning.
8. Section thawed and trimmed blocks at desired thickness
(see Note 12).
9. Allow frozen sections to dry completely at room temperature
for 15 min after the final section on the slide is cut.
10. Store sectioned tissue at −80 °C until ready to stain. Avoid
repeated thawing of the slides.
11. When ready to stain, first allow slides to thaw for 15 min on
the bench top.
12. Fix slides in ice-cold acetone for 10 min (see Note 13).
13. Proceed to Subheading 3.4.
3.2 Tissue Tissue fixed for frozen embedding receives a short PFA fix and
Preparation for does not require antigen retrieval.
Fixed-Frozen Tissue
1. Collect thymi or embryos in PBS and store on ice until ready
to fix.
2. Fix tissue in cold 4 % PFA with duration depending on size of
embryo or thymi (see Note 14).
3. Wash fixed tissue three times for 5 min each in PBS
(see Note 15).
4. Dehydrate tissue in 5 % sucrose/PBS for 1 h at 4 °C. Replace
5 % sucrose/PBS with 15 % sucrose/PBS and leave at 4 °C
overnight.
5. Rinse tissue in OCT (see Note 16).
6. Place tissue in a cryomold and cover with OCT. Freeze on dry
ice (see Note 17).
7. Store at −80 °C until ready to cut. See cryostat-cutting

instructions in Subheading 3.1, steps 5–10.
8. Proceed to Subheading 3.4.
3.3 Tissue Thymi or embryos collected for paraffin embedding require a lon-
Preparation for ger PFA fixation, must be thoroughly dehydrated, and require
Paraffin-Embedded antigen retrieval.
Fixed Tissue 1. Collect thymi or embryos in PBS and store on ice until ready
to fix.
2. Fix tissue in cold 4 % PFA with duration depending on the size
of embryo or thymi (see Note 18).
3. Wash fixed tissue three times for 5 min each.
4. Dehydrate tissue in 70 % ethanol, 90 % ethanol, 95 % ethanol,
and three 100 % ethanol washes at 4 °C. Dehydration time will
vary with size (see Note 19). Tissue can be stored at −20 °C in
the third 100 % ethanol wash.
5. When ready to embed, tissue must be cleared in xylenes. This
will remove fat and allow wax to permeate the tissue. Xylene
washes are also size dependent (see Note 20). Once the tissue
is translucent remove it from xylene, and place it in an embed-
ding cassette (see Note 21).
6. Cassettes containing cleared thymi or embryos are incubated
in three paraffin washes at 65 °C (see Note 22).
7. Place your specimen in a paraffin-embedding mold, fill with
warm paraffin, orient appropriately, and allow the block to
cool. Blocks may be stored at 4 °C.
8. Allow cold blocks to warm to room temperature. Cut off
excess wax from the block to form a square to allow for ribbon-
ing during sectioning.
9. Mount the block in the base of a cassette and clamp it to the
microtome for sectioning.
10. Because paraffin-embedded tissue is thoroughly dehydrated,
cells are condensed and thinner sections should be cut.
Do not exceed 8 μm of paraffin-embedded tissue for
immunofluorescence.
11. Cut sections may be kept in ribbons and floated on a warm
water bath (do not exceed 45 °C).
12. Slides must be thoroughly dried (see Note 23).
13. Dry slides can be stored in slide boxes at 4 °C indefinitely.
14. Allow slides to warm to RT. Use two sets of mailers: one set
used for dewaxing and rehydration and another set used for
antigen retrieval.
15. Fill one set of mailers 2/3rd full with AR buffer and place in a
300 mL beaker filled with 150 mL of water. Heat the water
and buffer to 95 °C. Cover the beaker with foil to avoid excessive
evaporation (see Note 24).
16. Place slides in clean mailers and wash twice for 5 min each in
xylene to remove the paraffin.
17. Remove excess xylene from the tissue by washing twice for
2 min in 100 % ethanol.
18. Rehydrate tissue by 2-min washes in 90, 70, and 30 % ethanol.
Leave slides in a dH2O wash for at least 5 min.
19. Transfer slides from rehydration mailers to AR mailers (see
Note 25).
20. Tissue is antigen retrieved for 30 min at 95 °C. Keep the water
level in the beaker at the level of the slide labels (see Note 26).
21. Allow mailers of slides in hot AR buffer to cool on your bench
for 20 min with the lid open.
3.4 Staining 1. Wash slides in PBS for 5 min.

2. Prepare primary antibody master mix. Dilute antibodies at
the appropriate concentration in 5 % serum of the secondary
host (e.g., donkey serum) and 0.05 % Triton-X in PBS
(see Note 27).
3. Pour off the PBS and lay slides flat in a humid chamber (see
Note 28). Incubate each slide in 100 μL of primary mix.
Gently cover slides in parafilm to help spread antibody evenly
and prevent evaporation.
4. Depending on the antibody incubate overnight at 4 °C or 1 h
at RT in a humid chamber (see Note 29).
5. Gently remove the parafilm. Return slides to mailers and wash
three times in PBS.
6. OPTIONAL STEP: When working with a low-affinity or low-
titer antibody, it is possible to boost the signal with either a
biotin intermediate step or TSA amplification. In the case of
the biotin intermediate, the primary antibody is removed fol-
lowed by an amplification step in which 100 μL of a secondary
biotinylated anti-primary Ig antibody diluted at 1:200 is added
for 30 min. The secondary biotinylated reagent is detected
using a streptavidin reagent conjugated to the fluorochrome of
choice. Briefly wash slides in PBS (see Note 30).
7. Prepare secondary antibody master mix. Dilute secondary Abs
in PBS—1:400 (see Note 31).
8. Incubate with secondary antibody. Apply 100 μL of secondary
antibody mix per slide. Cover in parafilm and incubate in a humid
chamber in the dark for 30 min to 1 h at RT (see Note 32).
9. Wash slides three times in PBS for at least 1 min per wash
(see Note 33).
10. Lay slides flat on a Kimwipe and add at least 100 μL of mounting
media to the bottom edge of each slide (see Note 34).
11. Coverslip each slide and blot the edges of the slide to remove
excess liquid (see Note 35).
12. Slides may be stored in the dark at 4 °C for up to 2 weeks to
preserve fluorescence.
4 Notes
1. PFA will dissolve at temperatures as low as 65 °C; however PFA

heated to boiling will still work for fixation. Dissolved PFA has
a short shelf life—a week at 4 °C or several months at
−20 °C. Dissolved PFA can be filtered using Whatman paper.
2. Ethanol can be reused, although 100 % ethanol steps are best
done with fresh ethanol.
3. For best results use fresh xylene every time when embedding.
Previously used xylene can be used for removing paraffin from
slides.
4. pH is very important. Take note of how much HCl you add
the first time and add a similar amount every time you make
this buffer. Replace AR buffer every month.
5. Pre-prepared silanized slides are commercially available and
will prevent tissue from falling off the slide. Check to make
certain that the coating does not autofluoresce as quality of
pre-prepared slides is variable.
6. Slide mailers come in many shapes and sizes. The optimal
mailer to use for antigen retrieval is the 5-slide mailer with an
end-opening flip-top lid.
7. Mounting media must include an anti-fade reagent.
Commercially available mounting medias differ in price and
duration of signal preservation.
8. Cover slips with dimensions of 24 × 50 mm will cover the non-
labeled portion of the slide. Cover slip thickness is indicated on
the microscope objectives (most objectives indicate 0.17 mm).
Because mounting media adds to the thickness of the cover
slip, #1 thickness is used for 0.17 mm even though the #1
range is 0.13–0.16 mm.
9. Antibodies should be titrated and tested for optimal fixation
conditions. Cytoplasmic and cell surface markers favor fresh
tissue while nuclear markers often require PFA-fixed tissue.
1:200–1:50 is a good starting point for titrations.
10. Microscope filter sets or lasers for confocal microscopy vary.
Determine what wavelengths are detected by the filter sets for
your microscope. In choosing fluorochrome-conjugated
reagents, avoid spectral overlap. For instance if your far-red

filter detects 594 nm, use secondary fluorochrome reagents
that emit at 549 nm in the red channel to avoid nonspecific
spillover. The green and red channels give the strongest signals
and should be reserved for antibodies that give weaker signals.
The blue and far-red channels require a longer exposure and
are best used for antibodies that give strong signals.
11. Alternatively the cryomold can be placed on a weigh boat on
top of the liquid nitrogen to allow for a more gentle freeze to
avoid cracking. When embedding embryos make certain to
orient all embryos the same way.
12. Thickness will vary. I prefer 10 μm sections for immunostain-
ing, but have cut sections as thin as 6 μm when using the same
tissue for multiple stains. Alternating slides while cutting is a
handy trick to allow for multiple sets of stains on serial sections
(Fig. 3). Handle slides with gloves to avoid getting oil on the
slides that will cause sections to wash off later.
13. I prefer to fix slides laid flat in the humidity chamber to avoid
loss of tissue that can occur when fixing slides vertically in a
mailer. Add 1 mL of acetone to cover each slide. Additional
acetone must be added several times due to evaporation.
14. Fixation time is determined by trial and error. A good starting
time for whole embryos is 15 min for embryonic day 11.5
Fig. 3 Alternating slides allows for staining the same tissue with many markers. (a) When cutting through tis-
sue, adjacent sections are similar while further apart sections may have distinct features. (b) Consecutive
sections are placed on alternating slides to allow multiple stains to be performed on the same specimen while
still analyzing many parts of the specimen
(E11.5) or 30 min for E15.5. Adult thymi can be fixed for 30 min
and dissected embryonic thymi require shorter fixation times of
5 min for E12.5–E15.5 and 10 min for E16.5–E18.5. Tissue
which is insufficiently fixed will fail to stain for antibodies requir-
ing fixation while tissue which is over-fixed will autofluoresce.
15. Ten rapid successive PBS washes can be used instead of three
5-min washes to reduce autofluorescence.
16. Place a small amount of OCT in a Petri dish and coat the tissue
thoroughly to remove excess PBS. This will help to reduce
cracks when sectioning.
17. PFA-fixed tissue will crack when snap frozen in liquid nitro-
gen. Use dry ice instead.
18. Aim for the shortest fixation time that allows the tissue to survive
the embedding process. A good starting time for embryos is
30 min for E11.5 or 2 h for E15.5. PFA will not permeate the skin
of embryos older than E15.5 and it must be removed. Fix adult
thymi for 45 min. Tissue that is under-fixed will fall apart during
the embedding process or fail to stain for antibodies requiring fixa-
tion while tissue that is over-fixed will autofluoresce.
19. 20-min washes are sufficient for E11.5 embryos or similar
sized tissue. Older embryos will require longer washes. Adult
thymi should be kept at each stage for at least 1 h. Longer
dehydration washes do not harm the tissue.
20. Adult thymi require two 10-min washes, whereas fetal thymi
require two to three 5-min washes. It is best to check the tissue
frequently while clearing it by holding it to the light and look-
ing for translucency.
21. This is a critical step. Excess time in xylene can lead to brittle
tissue that causes the sections to crumble.
22. Paraffin permeabilization is size dependent. Find the time that
exposes your tissue to the least amount of heat. For E11.5
embryos three 20-min washes are sufficient. For adult thymi
three 2-h washes are necessary.
23. If you are in a hurry, slides may be sufficiently dry as soon as
4 h after sectioning. Spacing sections apart from one another
on the slide will expedite this process. It is best to dry slides
overnight to make certain that no water is left on the slide as
any water under the tissue will cause it to fall off the slide in
future steps.
24. This limits you to 20 slides—four mailers maximum. Processing
more than 20 slides can result in less than ideal antigen retrieval.
25. It is best to pour off the water before transfer. Do this step
quickly to avoid lowering the temperature of the AR buffer.
26. Both time and temperature are critical for antigen retrieval.
This is the most important step in paraffin embedding.
Make certain that the beaker stays full of 95–100 °C water and
remove slides from the water bath at exactly 30 min.
27. Make your antibody mix at N + 1. (If you have 20 slides make
enough mix for 21 slides—2.1 mL of mix.)
28. Your chamber should contain the same liquid you incubate
in—in this case PBS. Cover the base of the slide box with
folded paper towels. Pour at least 25 mL of PBS in each side.
29. Antibodies that give low signal but little background give the
best results with an overnight incubation at 4 °C, while anti-
bodies with higher signal and background give better results
when incubated at RT for a shorter duration.
30. Tyramide streptavidin amplification will substantially boost the
signal but has the potential to significantly increase the back-
ground. If you use a TSA kit to amplify a primary antibody, use
a streptavidin secondary reagent for staining, step 7.
31. Good results can be obtained with highly absorbed secondary
antibodies conjugated with DyLight or Alexa Fluor fluoro-
chromes. Cy3 and Cy5 also give bright signals. FITC and
Texas Red conjugates give weaker signals due in part to the
rapid photobleaching and relatively higher background stain-
ing, respectively.
32. Longer incubations especially at 4 °C give higher background
and rarely amplify the signal.
33. DAPI can be added to the second PBS wash. DAPI concentra-
tion is determined by user preference (I use 1:15,000). Do not
add DAPI if you use an AMCA- or Alexa350-conjugated sec-
ondary reagent.
34. Do not exceed five slides at a time.
35. Lower the cover slip gradually from bottom to top to spread
the mounting media without introducing bubbles on the slide.
Alternatively, place the edge of the cover slip at the bottom of
the slide and allow adhesive forces to move the cover slip. Avoid
bubbles as attempts to remove them may damage your tissue.
References
1. van Ewijk W, Shores EW, Singer A (1994) microenvironment. Front Biosci 16:
Crosstalk in the mouse thymus. Immunol Today 2461–2477
15:214–217 5. Anderson G, Takahama Y (2012) Thymic epi-
2. Gray DH, Ueno T, Chidgey AP et al (2005) thelial cells: working class heroes for T cell
Controlling the thymic microenvironment. development and repertoire selection. Trends
Curr Opin Immunol 17:137–143 Immunol 33(6):256–263
3. Gordon J, Manley NR (2011) Mechanisms of 6. Petrie HT, Zúñiga-Pflücker JC (2007) Zoned
thymus organogenesis and morphogenesis. out: functional mapping of stromal signaling
Development 138:3865–3878 microenvironments in the thymus. Annu Rev
4. Manley NR, Richie ER, Blackburn CC et al Immunol 25:649–679
(2011) Structure and function of the thymic
Chapter 7
Purification of Thymocyte and T Cell Subsets

Andrea C. Carpenter, Jong Kyong Kim, and Rémy Bosselut
Abstract
Many analytical or cell culture procedures require homogeneous starting cell populations that cannot be
obtained directly from organ dissection. Here, we describe two enrichment procedures to achieve this goal
and discuss their respective advantages in specific experimental contexts. Notes in this chapter include
some tips on how to determine the appropriate level of purity (see Note 1).
Key words Single-cell suspension, T cell purification, Magnetic bead depletion, Flow cytometric
sorting
1 Introduction
This chapter describes two procedures for preparing homogeneous

thymocyte and T cell suspensions from organs that contain a het-
erogeneous mixture of cell types. Both techniques rely on the
availability of appropriate antibodies against surface molecules dif-
ferentially expressed on desired and unwanted cells.
The first procedure uses antibody-coated magnetic beads to
retain (“positive selection”) or discard (“negative selection”) cells
expressing specific surface molecules. The positive selection
approach is of broad applicability, as one only needs a single cell-
specific surface antigen, but has the significant disadvantage of
leaving the purified cells coated with beads. This has the potential
to interfere with cell activation or differentiation in culture (anti-
bodies that bind the T cell receptor, for example, are detrimental
for certain procedures as they can activate T cells). In addition, the
presence of cell aggregates tends to limit cell purity, as one desired
cell in the aggregate will positively select the whole aggregate. The
negative selection approach is more complex, because it requires
antibodies against all other cell types present in the starting popu-
lation. However it yields cells “untouched” (i.e., without attached
87
88 Andrea C. Carpenter et al.
antibodies or beads), which is of particular importance if the cells

are to be cultured or tested functionally in subsequent steps.
The second procedure uses flow cytometric sorting of cells
stained with appropriate fluorescent antibodies. As with the bead
purification procedure, antibodies can be used in flow cytometric
sorting to positively or negatively mark cells, and in the latter case
leaves cells untouched. In contrast, positively marked cells carry
fluorescent antibodies that have to be taken into account when
subsequent staining steps are considered, and, as mentioned above,
can contribute to activation or differentiation. The choice of the
sorter is generally dictated by locally available equipment. Except
for experienced users, professional operators generally perform
sorts; prior consultation with operators is highly recommended to
discuss staining and sorting strategies, including gate definitions.
Several experimental considerations guide the choice of one
procedure over the other. The first regards biological consequences
on purified cells. Especially with negative selection, bead purifica-
tion induces only minimal cell change or damage. In contrast,
hydrodynamic stress, electrical charging, and repeated centrifuga-
tions are unavoidable side effects of flow cytometric sorting and
can result in reduced survival and functionality. Time is also an
important consideration. Bead purification is generally faster, par-
ticularly when large cell numbers are desired, as processing time is
not proportional to input size. In contrast, the time required for
sorting is directly proportional to input size, potentially affecting
cell viability for large inputs. Specific situations (e.g., purification
of extremely rare subsets) will require pre-purification steps. Cell
yields are often higher with bead purification than with flow cyto-
metric sorting. Last but not least, cost is a major consideration,
and in many cases tips the scale in favor of bead purification when
it can reach purities consistent with the study objectives. This is
particularly true in applications requiring high cell numbers and
repeated measurements.
However, flow cytometric sorting has decisive advantages. It
offers unequaled flexibility for defining purification criteria. It is
irreplaceable when gradient expression levels (e.g., CD44lo vs.
CD44hi) or co-expression of two marks (e.g., CD4+CD8+ from
CD4−CD8+ or CD4+CD8− thymocytes) defines desired popula-
tions. Despite major progress in bead purification approaches, cell
purity is a second key asset of flow sorting. Although purities
greater than 95 % can be achieved by bead purification with appro-
priate antibodies, they are routinely much lower when the fre-
quency of the target population falls below 10 % of the starting cell
suspension. In contrast, flow cytometric sorting can provide highly
purified cells (>99 %) even when they are at a very low frequency
in the starting population.
Regardless of the procedure chosen, it is important to carefully
delineate the experimental plan well in advance. Estimate the size
Purification of Thymocyte and T Cell Subsets 89
of the starting population (and when applicable the number of

experimental animals) by calculating frequency in the target popu-
lation and expected yield. When working with small cell numbers
or precious samples, it is highly recommended to perform pilot
studies with similar cells that are more easily replaceable. Slower
flow cytometric sorting speeds may be required for infrequent tar-
get populations to increase purity and yield.
Initially a 95–99 % yield may seem ideal; however it is useful to
know the nature of contaminants. After flow sorting, the operator
should “rerun” a small portion of the purified sample, although in
some cases additional stains may be needed. For bead-purified
cells, purity checks are often done by staining an aliquot of the
purified cells and analyzing them by flow cytometry. Antibodies
used for purification, whether positive or negative, can potentially
impede binding of those used for a purity check. Thus, depending
on each specific situation, purities should be verified with antibod-
ies that do not cross-block those used in the purification step.
In this chapter, we provide detailed procedures to prepare
purified T cell subsets (from thymus or peripheral organs) using
either depletion (“negative selection”) or flow cytometry sorting.
These procedures can easily be adapted to specific requirements
and reagents. Protocols for bead purification by positive selection
depend heavily on the specific reagents being used and will there-
fore not be described in detail here. As with every antibody-
mediated experimental procedure, pilot analyses titrating the
amount of antibodies (and related reagents such as beads) are
highly recommended to optimize yield and purity prior to pro-
ceeding with the actual experiment.
2 Materials
All necessary efforts should be made to keep organism contami-

nants out. If the end procedure is sterile, all reagents in purifying T
cells should also be sterile (and the flow cytometry operator should
be notified in advance that the cells must remain sterile during the
sort). Medium should be made just before the experiment. If the
sterility of the starting reagents is in question, sterile filter (≤45 nm
pore size) the final buffer or media before use.
1. Heat-inactivated fetal bovine serum (FBS) (see Note 2).
2. Isolation buffer: 500 mL PBS pH 7.4 (no Ca2+, no Mg2+),
0.1 % FBS (or bovine serum albumin), 2 mM EDTA pH 8.0.
3. Complete medium: RPMI 1640 medium, 10 % FBS, 100 U/
mL penicillin, 0.1 mg/mL streptomycin, 0.292 mg/mL
l-glutamine, 20 mM HEPES (see Note 3).
4. Red cell lysis buffer: 8290 mg/L ammonium chloride, 1 g/L

potassium bicarbonate, 37 mg/L EDTA pH 8.0.
5. Nylon tissue filters (pore size 100 μm, cut to 2 or 4 cm2).
6. 1 or 3 mL syringe pestles.
7. 60 mm petri dishes.
8. 5, 15, and 50 mL polypropylene conical tubes.
9. 5 mL polystyrene round-bottom tubes 12 × 75 mm style
(flow-sorting procedure).
10. 50 μm filters (flow-sorting procedure).
11. Appropriate fluorochrome-conjugated antibodies.
12. Magnetic beads compatible with appropriate purified
antibodies.
13. Rocker.
14. Two curved forceps.
15. Scissors appropriate for dissection.
16. 37 °C incubator with 5 % CO2.
17. Vertical magnet holding 15 or 50 mL conical tube (DynaMag
15, 50, or equivalent).
3 Methods
3.1 Obtaining Minimizing total purification time is critical in this procedure.

Single-Cell Often a trial run with cells is advisable, not only to practice the
Suspensions procedure, but also to get an idea of the efficiency for the specific
application. Depending on the desired final purity and the fragility
of the cell type to be isolated, two to five times as many target cells
should be in the starting population.
1. Prepare the workspace before euthanizing the animal(s). For
each organ, set up one 60 mm petri dish on ice with 3 mL
complete medium and a 2 cm2 piece of nylon filter on the bot-
tom of the dish. Prepare one syringe pestle (1 or 3 mL syringe)
(excluding thymus), plus one 4 cm2 100 μm filter per organ,
and keep them in a clean container. For the thymus, no syringe
pestle is needed.
2. Remove thymus, spleen, and/or lymph nodes (LN)s from
animal(s) (see Note 4 on LN selection). Before adding the thy-
mus to the petri dish with medium, roll it on gauze or clean
tissue paper to remove any blood, and carefully dissect it from
surrounding connective or lymphoid tissue (which tends to
stick to the gauze or paper). Do not pull tissue from the thy-
mus and make sure not to damage the thymus or allow it to
dry. When removing lymph nodes, carefully remove all fat.
3. For the thymus, take two curved forceps and gently tease apart
the organ.
4. For the LNs and spleen, use the rubber end of the syringe
pestle to dissociate the cells from the organ tissue, and ensure
that you have dissociated any observable clumps as well. It is
helpful to use the lid of the dish to store 5 mL additional
medium for step 5. At this point, wash the used end of the
syringe pestle in this medium, by swirling it in the medium;
then discard the pestle.
5. Tilt the dish containing the cells at a 30° angle, draw up the
medium with the free cells, and wash this medium back over the
dish. Gently pipet up the medium from the dish again and run
it through the 4 cm2 100 μm filter into a 15 mL conical tube.
Before passing the medium through the filter, use the pipet tip
to push the filter half way into the top of the tube, and then pull
the pipet tip a few mm back from the filter. Pulling the pipet
back prevents the medium from directly hitting the filter and
splashing out of the tube. Leave the filter balanced in the top of
the conical tube. Next, use the additional 5 mL medium to wash
the dish and pass that through the same filter into the conical
tube. The dish and filter can now be discarded. Pellet the cells at
150 × g, for 5 min at 4 °C, and decant the supernatant.
6. For spleen cell preparation, lyse red blood cells by resuspend-
ing the cell pellet in 2 mL red cell lysing buffer for 2 min on
ice. Then immediately, dilute out the red cell lysing buffer by
adding 8 mL complete medium. Pellet the cells at 150 × g, for
5 min at 4 °C, and decant the supernatant.
7. Resuspend the organs at 10–20 × 106 cells/mL in complete
medium (see Note 5 on expected yields).
8. Count the cells using a dye that distinguishes dead cells (for
example: trypan blue). Calculate the cell concentration and
total number. Set aside an aliquot (106 cells is ideal) from each
starting cell suspension for flow cytometric analysis.
9. In unmanipulated mice, spleen and LN cells can be combined
to maximize the size of the starting population. For T cells,
when large cell numbers are not needed, LN may be a prefer-
able starting point because of the greater frequency of T cells
in LN than in spleen.
3.2 Magnetic Bead Several commercial kits are available to purify target populations
Purification of T Cell (e.g., CD4 or CD8 T cells), most of them using proprietary
Subsets by Depletion reagents and materials (magnets). The following procedure uses
Dynabeads (LifeTech) to negatively select the desired population.
Provided the beads carry the appropriate secondary reagent,
Dynabeads can also be used with user-provided antibodies (see
Note 6). Volumes and bead numbers are given for 107 starting
cells. They should be scaled up proportionally with higher cell

numbers. Up to 5 × 107 cells can be processed in a 15 mL conical
tube; use several tubes or scale up to 50 mL tubes if processing
more cells.
1. Pellet cells to be purified by spinning at 150 × g, for 5 min, at
4 °C.
2. Resuspend cells in 100 μL isolation buffer. If any cell clumping
is observed, refilter before proceeding.
3. Add 20 μL FBS and 20 μL kit antibody.
4. Incubate for 20 min at 2–8 °C (wash beads during this incuba-
tion and keep on ice until needed; see steps 5 and 6).
5. Wash the beads (to eliminate free antibody that might have
been released during storage): first resuspend beads in the vial
by vortexing for 30 s. Pipet 200 μL beads into a new 15 mL
tube and resuspend in equal volume or at least 1 mL isolation
buffer.
6. Put the tube without the lid on magnet for 1 min, remove buf-
fer by pipetting, and discard. Remove tube from magnet and
resuspend beads in the same volume of isolation buffer as the
original bead volume.
7. After antibody incubation, wash cells by adding 2 mL isolation
buffer. Pellet the cells at 150 × g, for 5 min, at 4 °C and decant
the supernatant.
8. Resuspend the cells in 800 μL isolation buffer and add pre-
washed beads.
9. Incubate for 15 min at 25 °C with tilt and rotation on a rocker
at a speed sufficient to keep the beads in suspension.
10. Add 1 mL isolation buffer, and gently pipet up and down five
times with a large-bore pipet (5 mL or similar).
11. Put tube without lid on magnet for 2 min. Beads and bead-
attached cells will be pulled to the tube wall on the magnet
side.
12. Being careful not to dislodge the beads, transfer supernatant
containing non-adherent cells to a new tube. For applications
that require special media or buffers, wash cells with that
medium or buffer as soon as possible after transferring. Discard
beads at this point.
13. Count purified cells to determine recovery and take an aliquot
(if possible, at least 105 cells) to quantify purity by flow cyto-
metric analysis (see Note 7).
14. Purity from magnetic bead isolation depends on the applica-
tion (routinely >85 % for CD4 or CD8 T cell purification
from LN).
3.3 Magnetic Bead A similar procedure used in Subheading 3.2 can be used to increase
Enhancement Pre-flow the frequency of rare target cells prior to flow cytometric sorting.
Cytometric Sorting This is especially useful when starting from high cell numbers
(>108) for which sorting times quickly become prohibitive. In this
case, because the objective is to remove most of the nontarget cells
rather than achieving high purity, significant savings can be achieved
by lowering bead numbers and antibody concentrations. While
pilot experiments are needed to titer down these reagents, a good
starting point is to use 1/8 of the recommended amounts of beads
and antibodies while keeping cell concentration intact and all incu-
bations at 4 °C. This “light” procedure routinely achieves >40 %
cell purity with very little loss of desired cells.
1. Pellet cells to be run through kit by spinning at 150 × g, for
5 min, at 4 °C.
2. Resuspend cells in 100 μL isolation buffer per 107 cells. If any
cell clumping is observed, refilter before proceeding.
3. Add 20 μL FBS cells and 2.5 μL kit antibody.
4. Incubate for 20 min at 2–8 °C (wash beads during this incuba-
tion and keep on ice until needed; see steps 5 and 6).
5. To wash the beads: first resuspend beads in the vial by vortex-
ing for 30 s. Pipet 25 μL beads into a new 15 mL tube and
resuspend in equal volume or at least 1 mL isolation buffer.
6. Put the tube without the lid on magnet (DynaMag 15, 50, or
equivalent) for 1 min, remove buffer by pipetting, and discard.
Remove tube from magnet and resuspend beads in the same
volume of isolation buffer as the original bead volume.
7. After antibody incubation, wash cells by adding 2 mL isolation
buffer. Pellet the cells at 150 × g, for 5 min, at 4 °C and decant
the supernatant.
8. Resuspend cells in 975 μL isolation buffer per 107 cells and add
pre-washed beads.
9. Incubate for 15 min at 4 °C with tilt and rotation on a rocker
at a speed sufficient to keep the beads in suspension. (The 4 °C
incubation reduces efficiency, but better preserves cell viability
in advance of flow cytometric sorting.)
10. Add 1 mL isolation buffer per 107 cells, and gently pipet up
and down five times with a large-bore pipet (5 mL or similar).
11. Put tube without lid on magnet for 2 min.
12. Transfer supernatant containing enhanced homogeneity to a
new tube, and wash with complete medium as soon as possible
after transferring.
13. Count cells to determine recovery and take an aliquot of 105
cells to quantify purity by flow cytometric analysis (see Notes 7
and 8).
14. Purity will be >40 % with very little loss of desired cells.
3.4 Flow Cytometric Before sorting cells, consider doing a pre-depletion with beads (see
Sorting Subheading 3.3), especially when preparing rare cells (<5 % of the
input). Although this increases the pre-sort preparation time, it
often reduces the total cell manipulation time before downstream
procedures. When the target frequency is under 1 %, the pre-
depletion strategy should seriously be considered as the sorter
error rates can substantially decrease purity in this situation (also
see Subheading 3.4.2). The procedure outlined below assumes
basic understanding of flow cytometry procedures and does not
address operation of the cell sorter itself, which is generally left to
a professional operator.
3.4.1 Flow Cytometric 1. Plan ahead. Consult the flow cytometry facility and inquire
Sorting: General Procedure about specifics of available equipment and procedures.
Discuss the choice of fluorochromes and evaluate the time
needed for cell purification and to verify purities. It is also
important to choose a sorting nozzle of appropriate diame-
ter, as a general rule at least five times as large as the cells in
the starting population. For lymphocytes, a 70 μm nozzle
provides adequate speed without damaging the cells.
However, a 100 μm nozzle may be better suited when sorting
larger cells or when using the double-sort procedure as
described in Subheading 3.4.2. For most applications, cells
should be kept cold, and it is important to verify that the
sorter has a chillable (4 °C) sample holder.
2. Resuspend cells to be stained for flow sorting at 2 × 107 cells/mL
complete medium (use 15 mL tube for <6.0 × 107 cells, 50 mL
for >6.0 × 107 cells). Before staining for sort, a titration of the
antibodies is advisable; 0.25 μg antibody per 107 cells is a good
starting point. If using the protocol in Subheading 3.3 prior to
sorting, ensure that the antibodies used for flow cytometric
sorting were not blocked by the antibodies in the bead kit (see
Notes 7–9).
3. Do not forget to prepare cell samples for flow cytometer setup
if needed. Typically set aside 0.5 × 106 cells per compensation
tube. These will be the single-color tubes used to set up the
flow cytometric sorter or analyzer for purity checks.
4. Add appropriate antibodies, mix well, and incubate at 4 °C for
45 min in the dark.
5. Wash with complete medium using four times staining vol-
ume. Pellet the cells at 150 × g, for 5 min, at 4 °C and decant
the supernatant.
6. Resuspend cells to be flow sorted at 2 × 107 cells/mL complete
medium in 5 mL round-bottom polystyrene tubes with caps.
7. Filter the starting population immediately before sorting.

8. Define sorting gates in collaboration with the operator (see
Note 10).
9. Use polypropylene collection tubes that are large enough to
hold the volume collected. For example two to three million
cells will fit in a 5 mL polypropylene round-bottom tube; how-
ever more than that will need a polypropylene 15 mL conical
tube (see Note 11 on plastics in flow cytometric sorting). Add
a “cushion” of 50 % FBS and 50 % complete medium to the
collection tubes, a total of 1 mL for 5 mL tubes and 2 mL for
15 mL tubes.
10. Upon finishing each sorted sample, immediately fill the collec-
tion tube containing the purified cells with complete medium.
If the sorted cells have filled the tube, transfer its contents as
soon as possible to a larger tube and then fill that tube with
complete medium. Diluting the sheath fluid with complete
medium improves cell viability.
11. Take a small sample of purified cells and run in on the sorter to
verify purity. Depending on the target population, it may be
necessary to stain with other markers (e.g., for intracellular
proteins, see Notes 7–9 on staining post-sort).
12. Spin cells at 65 × g, for 30 min, at 4 °C. Resuspend in an appro-
priate volume and count cells (see Note 12).
13. Cells are ready for downstream applications.
3.4.2 Secondary Protocol This procedure uses two consecutive sorts to isolate good purity
for Sorting Populations (>95 %) populations from very rare cells (less than 0.01 % of the
≥0.01 % starting population) if bead isolation procedures are unavailable or
inadequate for pre-sort enrichment (see Subheading 3.3).
1. Perform the primary sort following the procedure defined in
Subheading 3.4.1 through step 10 with the following changes:
(a) Only use a 100 μm nozzle.
(b) In step 8, use a wide gate to select the entire desired popu-
lation and maximize yields. Unwanted cells gated in error
will be removed in the second sort.
2. After step 10, pellet the cells at 150 × g, for 7 min, at 4 °C and
decant the supernatant.
3. Resuspend the cells in a small volume, ≥300 μL complete
medium.
4. Perform the secondary sort, using a tight gate that is slightly
inside the edges of the desired population. This will increase
purity to acceptable levels (>95 %).
5. Follow steps 10–13 in Subheading 3.4.1.
4 Notes
1. It is important to define objectives for target cell purity. High

cell purity (99 % or more) is often necessary when purified cells
will undergo multiple rounds of cell division (e.g., activating T
cells). In this case, preferential expansion of contaminating
cells has the potential of significantly altering experimental
outcomes. Although high cell purity is always desirable, realis-
tic objectives should be defined based on the properties of
expected contaminants and on the dynamic range of assays to
be run on purified cells.
2. FBS should be heat inactivated by incubation at 56 °C for
50 min.
3. HEPES can be used to buffer media outside of a CO2
incubator.
4. Because LNs from different body areas are to be exposed to
different antigenic environments, care should be exercised in
deciding whether to pool populations obtained from distinct
anatomic sites. Typically, in unmanipulated laboratory mice
housed in specific pathogen-free facilities, the major site of
microbial exposure is the gut. Consequently, the frequency of
activated cells is greater in mesenteric LNs that drain intestinal
tissues, and these should generally be processed separately. In
contrast, peripheral LNs (including popliteal, inguinal, axillary,
and cervical) have few germinal-center and activated T cells,
and it is legitimate to pool these populations.
5. Typical cell yields from organs are as follows for a 6–8-week-
old female mouse: thymus, 150–250 × 106; spleen 50–80 × 106;
and LNs 30–40 × 106.
6. Although commercial bead kits typically include specific anti-
body mixes designed for purification of homogeneous cell
populations, it is possible to use beads with user-prepared anti-
bodies. In such cases, it is paramount to verify that the bead-
bound secondary reagent (e.g., streptavidin or anti-IgG) binds
the primary antibodies used for purification. In addition, pilot
experiments should titer each antibody for its ability to remove
undesired cells.
7. When verifying purity of “negative” bead purification, it is
essential to avoid staining with the same antibodies as those
used for purification (or antibodies that cross-block each
other). Otherwise, unwanted cells may escape detection by
staining if the epitope is masked by the purifying antibody.
Commercial kits often use near-saturating levels of anti-CD8α
antibody to purify CD4+ T cells and of anti-CD4 to purify
CD8+ T cells. In such situations, cell purity can be verified
using anti-CD8β, a-non cross-blocking anti-CD4, anti-TCRβ

(H57), anti-MHC class II, an Fc blocking antibody, and a
live/dead discriminating dye (DAPI for example, see Note 8).
Three of the four commonly used anti-CD4 monoclonal anti-
bodies (GK1.5, H129, and RM4-5) cross-block, whereas the
fourth one (RM4-4) does not.
8. Live/dead dyes like DAPI remain in sorted cell suspensions
and could be detrimental. Instead of these membrane-
permeable dyes, use LiveDead fixable dyes that only bind dead
cells and are washed away with the excess antibody. Also con-
sider sorting live cells based on FSC and SSC and then testing
viability post-sort on a small sample used only for purity check.
9. The fluorochromes (especially PE) used for flow sorting can
remain associated with sorted cells for an extended time, unless
they are diluted by proliferation. If additional stains are needed,
make sure to leave open fluorochromes for these stains.
10. Gating strategies obviously depend on each application. In
general, it is advisable to tightly gate the target population.
This slightly reduces the yield but minimizes contaminations.
It is also recommended, when possible, to exclude cells with
unwanted markers by negative gating. Several markers staining
with the same fluorochrome can be combined to exclude mul-
tiple cell types. This also contributes to the exclusion of cell
aggregates (which should be primarily excluded by light
scatter-based doublet discrimination).
11. Polystyrene collection tubes have the potential to build up
static electricity, which can interfere with sorting, and should
be avoided for cell collection.
12. Flow sorters often misreport cell counts, and there is unavoid-
able cell loss during the centrifugation of sorted cells. If pos-
sible, manually recount sorted cells after centrifugation with a
live/dead discriminating dye.
Chapter 8
Retroviral Transduction of T Cells and T Cell Precursors

Amie Simmons and José Alberola-Ila
Abstract
Transduction of lymphoid progenitors with retroviral or lentiviral vectors is a powerful experimental strategy
to tease out the role of a gene or pathway in T cell development via gain-of-function or loss-of-function
strategies. Here we discuss different approaches to use this powerful technology, and present some protocols
that we use to transduce murine HSCs, thymocytes, and lymphoid cell lines with these viral vectors.
Key words Retrovirus, Hematopoietic stem cell, Thymocytes, Bone-marrow chimera, Development
1 Introduction
Transduction of hematopoietic cells with retroviral vectors is an

efficient way to manipulate gene expression during development of
the immune system. Retroviral vectors are easy to manipulate
in the laboratory and provide stable, long-term gene expression in
the infected cells and their progeny because they stably integrate
into the genome. Their major limitation is that they can only effi-
ciently infect and integrate in cycling cells.
Most popular vectors are derived from the murine stem cell virus
(MSCV) [1], a derivative of the Moloney murine leukemia virus
(MoMLV) which can maintain long-term expression in infected cells
and their progeny. While the genomes of these viruses are typically
7–12 kb in size and code for three structural genes, gag, pol, and env,
the vectors consist only of the essential cis-acting elements of the
viruses. gag, pol, and env are provided in trans, either in a stably trans-
fected cell line (see ref. 2) or, as we will discuss in this review, cotrans-
fected with the viral vector in independent plasmids. The separation
of the structural elements in different plasmids increases the safety
of the vectors, since the virus produced is replication defective, and
provides space in the viral backbone to clone the gene of interest
and/or markers to follow the infected cells.
The host range of a retrovirus is determined by its envelope
glycoprotein (Env). The most widely used are the Env proteins of
99
100 Amie Simmons and José Alberola-Ila
ecotropic MLVs. A major advantage (and a limitation) of ecotropic

viruses is that they cannot infect human cells, and therefore can be
used safely, with most inserts, under normal BSL2 laboratory con-
ditions. For infection of human cells, the retrovirus can be pack-
aged with an MLV amphotropic Env. Alternatively, Env proteins
from a different virus can be used to package the retroviral vector.
This process is called pseudotyping, and the most common heter-
ologous Env used is the vesicular stomatitis virus glycoprotein G
(VSV-G). Besides the extension in host range, VSV-G envelope
protein forms very stable viral particles, which can be concentrated
by ultracentrifugation [3]. The use of these modifications signifi-
cantly increases the biosafety concerns (see Note 1).
An alternative to MSCV-based retroviruses is HIV-1- or FIV-
based lentiviral vectors. Their major theoretical advantage over
conventional MSCV-based retroviral vectors is their ability to
transduce nondividing cells, which would make them better to
infect HSCs. While this seems to be the case with human HSCs,
the results with murine HSCs are not so clear [4–6] and it seems
that murine HSCs need to be in G1 for an effective transduction
to occur, and therefore they need to be stimulated in vitro with
cytokines, as is the case for transduction with retroviruses. In our
hands, lentiviral vectors have not been consistently superior to
MSCV-based retroviral vectors for HSC transduction, and tend to
be harder to manipulate in the laboratory.
One limitation of conventional retroviruses and lentiviruses in
some experimental situations is that they express the transgene at
high levels throughout development. If this early expression results
in developmental phenotypes that predate one’s stage of interest,
the approach is not useful. There have been many different retro-
viral designs that try to overcome this limitation. These include
tetracycline-inducible viruses [7], viruses engineered so that the
transgene is turned on by Cre (either by inserting a stop sequence
before the transgene [8, 9], or by inserting the floxed transgene in
an antisense orientation that is irreversibly inverted [10]), or by the
generation of self-inactivating viruses (SIN), where the endoge-
nous LTR is disrupted upon integration, and the transgene is then
expressed by a tissue-specific promoter (see refs. 11–14 for some
examples relevant to T lymphocyte development). The SIN
approach has been performed mostly with lentiviral backbones,
because conventional retroviruses engineered this way suffered
from very low titers. Recent modifications seem to have alleviated
this problem [15, 16] (see Note 2).
For the study of T cell development, the cells that are normally
targeted are hematopoietic stem cells, derived from either bone
marrow or fetal liver, immature thymocytes, and even early DP
thymocytes. In all these cases, retroviral infection provides a fast
and efficient method to induce or inhibit (via shRNA or dominant
negative approaches) the function of a gene of interest in vivo
Retroviral Vectors and T Cell Development 101
(via bone marrow chimeras, or intrathymic injection of infected

cells) or in vitro (via rFTOCs, or using the OP9Δ system [17]).
Infection of hematopoietic stem cells, and adoptive transfer
into irradiated hosts, is a very powerful approach to rapidly test the
function of a gene of interest in development, to perform in vivo
structure-function analysis, or genetic rescue experiments. The
main limitation of this approach is the ability to infect a sufficient
number of long-term reconstituting HSCs (since a very small pro-
portion of them are normally in cycle). Therefore most approaches
involve enrichment of LT-HSCs, and treatments to make them
enter cell cycle, so that they can be infected by the retroviral vector
while not losing their functionality. One approach is to treat the
mice with fluorouracil, which depletes fast-replicating cells in the
bone marrow. This has two effects; it increases the percentage of
LT-HSC in the marrow, and promotes cell cycle entry, as these
cells try to replenish the multipotent progenitors [18, 19]. Then
cells are cultured in media with cytokines and infected (see Note 3).
Another approach is to enrich for HSCs in vitro, using magnetic
depletion with commercially available lineage marker cocktails
(see Note 4), and then stimulate them with the same cytokine
cocktail in vitro before infection. Below we detail our current pro-
tocols to infect HSCs and thymocytes.
2 Materials
2.1 Production 1. 1.5 ml microfuge tubes.

of Retrovirus: 2. 5 and 14 ml polypropylene tubes.
Transient Transfection
3. 293T growth media: Dulbecco’s modified Eagle’s medium
(DMEM), 10 % fetal bovine serum, 2 mM l-glutamine, 100 U/
ml penicillin, 100 μg/ml streptomycin.
4. HEK 293T cells (ATCC, CRL-11268) (see Note 5).
5. 2× HBS: 50 mM Hepes, 10 mM KCl, 12 mM dextrose,
1.5 mM NaH2PO4, 280 mM NaCl, pH 7.05 ± 0.05.
Prepare three or four independent 100 ml batches, filter
sterilize with a 0.2 μm filter, and aliquot in 10 ml/15 ml coni-
cal tubes. Test the four batches with viral vector and packaging
stocks that have worked well previously, and discard those 2×
HBS batches that do not work well (see Note 6).
6. 2 M CaCl2: Prepare a 2 M solution (29.4 g in 100 ml), filter
through a 0.2 μm filter, and store aliquots at −20 °C.
7. Packaging plasmid(s) containing retroviral gag, pol, and env
genes: We routinely use pCL-Eco [20] for retroviral constructs
(e.g., Addgene #12371). Alternatively, gag/pol, and another
envelope such as VSV-G (e.g., Addgene #8454 or #12259),
can be used. For lentivirus we use psPAX2 (e.g., Addgene
#12260) to provide the missing structural elements.
8. Vector plasmid: There are many different vectors that can be

used. As discussed above, for lymphoid cells, and long-term
stable expression, backbones derived from MSCV are pre-
ferred. We use MIGR1, originally described in [21] for overex-
pression, and pBanshee [22] for shRNA expression [23]. For
lentivirus, we normally use FUGW [13] or derivatives.
Prepare plasmid stocks using standard molecular biology tech-
niques (see Note 7), and determine plasmid DNA concentration.
Our laboratory routinely uses PEG purification [24], but other
methods, or commercial kits, also produce good DNA.
2.2 Retroviral 1. Bone marrow growth media (BMGM): RPMI, 10 % fetal

Transduction bovine serum, 2 mM l-glutamine, 100 U/ml penicillin,
100 μg/ml streptomycin, 10 ng/ml IL-3, 10 ng/ml IL-6,
50 ng/ml SCF (see Note 8).
2. 2 % BSA in PBS (sterile).
3. Retronectin-coated plates: Retronectin (see Note 9). 1 mg/ml
in sterile PBS.
Non-tissue culture-treated plates (see Note 10). Dilute
retronectin stock to 12 μg/ml and add 250 μl/well for 24-well
plates, or 1 ml/well for 6-well plates. Incubate for 2 h at room
temperature, or wrap the plates and incubate overnight at
4 °C. Remove the retronectin solution and block the plates
with 2 % BSA (in sterile PBS) for 30 min at room temperature.
Wash twice with PBS. Plates can be stored with PBS at 4 °C.
4. Polybrene.
5. Lipofectamine.
3 Procedures
3.1 Transient 1. Twenty-four hours before transfection plate 0.4–0.8 × 106

Transfection exponentially growing 293T cells per 6 cm Petri dish in a 4 ml
volume of 293T growth media; or 1–2 × 106 cells per 10 cm
dish in 8 ml (see Note 11).
2. On the day of the transfection: have 2× HBS, 2 M CaCl2
solution, and sterile water at room temperature. Use polypro-
pylene tubes (microfuge, 4 or 15 ml):
For 10 cm dish:
(A) Microfuge tube: In a final volume of 500 μl H2O, mix,
7 μg retroviral construct (see Note 12), 7 μg pCL/Eco
(see Note 13), and 62 μl 2 M CaCl2.
(B) 4 or 15 ml tube: 500 μl 2× HBS.
For 6 cm dish:
(A) Microfuge tube: In a final volume of 300 μl H2O, mix 4 μg
retroviral construct, 4 μg pCL/Eco, and 37 μl 2 M CaCl2.
(B) 4 or 15 ml tube: 300 μl 2× HBS.

Take the Petri dish with the 293T cells to be transfected
out of the incubator into the tissue culture hood. Take one
plate out at a time. It is very important to keep the pH of the
media buffered.
Mix (A) and add dropwise to (B) while vortexing (B).
Alternatively, instead of vortexing (B), blow air into (B) using
a Pasteur pipette while adding (A) dropwise. Keep vortexing or
blowing air into (B) for 30 s or so. Immediately add the mix to
the 293Ts distributing it dropwise throughout the plate, and
then swirl dish gently; do not eject liquid into dish because
293T cells will be stripped off. Put dish back into incubator.
Transfect next dish.
3. Incubate cells with transfection mix for 16–20 h, after which
the plates are rinsed once with PBS or media and fed with 3 ml
of fresh media. At this stage the cells are slightly detached and
rinses should be done gently to avoid stripping them off.
4. Supernatant is collected 24 and 48 h after the rinse. Filter
through a 0.45 μm syringe filter to remove cells. It is ideal to
prepare fresh supernatant for every experiment, but we rou-
tinely snap freeze the supernatant in aliquots (keep at least at
−70 º C). The freezing causes a 50 % reduction in viral titer.
GFP expression can be seen 16 h after transfection and is
maximal at 24–48 h. Check transfection efficiency of the 293T
cells by flow cytometry. If the transfection efficiency is not high
(>60 %), you can assume that the virus titer will probably not
be good (see Note 14).
3.2 Transduction In general, the method involves plating cells in 24-well plates
with 2 ml of fresh warm viral supernatant and 4 μg/ml lipo-
fectamine or 5 μg/ml polybrene, centrifuging cells at 20 °C for
1–1.5 h at 460 × g, culturing cells for 1 h at 37 °C, and then
replacing the viral supernatant with media (see Note 15); addi-
tional details are provided below. These approaches yield good
efficiency of infection when coupled with a good virus, e.g.,
>90 % with lymphoid lines, up to 60 % with fetal liver and early
thymocytes, 5–20 % with DP thymocytes, and 15–30 % with
enriched bone marrow cells.
3.2.1 Retroviral 1. Day 0: Cells are plated in BMGM, and cultured for 24–36 h at
Transduction 2 × 106 cells/ml (the yield is approximately 2–5 × 106 cells per
Hematopoietic Cells 5-FU treated mouse).
2. Day 1: Use freshly produced virus, or thaw on ice an aliquot of
previously frozen virus. Calculate how much virus you will
need, depending on how many cells are going to be infected.
For a 6-well plate we use 2 ml viral supernatant and 2.5 × 106
cells/well. For a 24-well plate, we use 0.5 ml virus and 0.5 × 106
cells/well.
3. Add virus to previously prepared retronectin-coated plates.

Incubate at 37 °C for 2 h or centrifuge at 2000 × g for 2 h at
22 °C and then wash with PBS (see Note 16).
4. Add the cells in BMGM and centrifuge at 2000 × g for 5 min,
to facilitate adhesion to the fibronectin. Incubate at 37 °C
overnight.
5. Day 2: Collect the cells (see Note 17). Wash, resuspend in
fresh BMGM, either culture for 48 h or repeat the infection on
fresh virus-loaded retronectin plates (see Note 18), and then
culture for 24–48 h.
6. Day 3: If your virus has a fluorescent marker, infection effi-
ciency can be checked by flow cytometry.
7. Day 4: Adoptively transfer the infected cells into irradiated
hosts following standard bone marrow chimera protocols (see
Note 19).
8. Analyze the bone marrow chimeras 6–12 weeks after transfer
for your phenotype.
3.2.2 Retroviral 1. Resuspend cells from a culture in exponential growth at

Transduction of Lymphoid 5 × 106/ml of fresh media. Mix 100 μl of the cell suspension
Cell Lines with 2 ml of fresh viral S/N, and polybrene at 5 μg/ml, and
dispense in one well, in 24-well plates.
2. Centrifuge plate for 1 h at RT, 450 × g.
3. Incubate cells for 1 h in the incubator (37 °C 5 % CO2).
Replace 1.5 ml of media in each well with fresh media.
Afterwards, cells are fed and split as necessary.
3.2.3 Retroviral 1. Plate 1.5–2 × 106 cells/well in 24-well plates, in 2 ml of retro-

Transduction of Primary viral supernatant, plus Lipofectamine at 4 μg/ml. Spin for
Thymocytes 1–1.5 h, at 450 × g, at RT (see Note 20).
2. Wash cells by aspirating most of the supernatant without dis-
turbing the layer of cells and replace with fresh media.
3. Expression of the transgene (at least when monitored by GFP
expression can be detected after 18–24 h). The thymocytes can
be now used in your model (rFTOC, OP9Δ differentiation,
etc.) (see Note 21).
4 Notes
1. Using oncogenic inserts increases the biosafety requirements.

In any case, approval from the Institutional Biosafety
Committee is required for any work with these vectors, and
different institutions have slightly different requirements.
2. When considering all these alternatives it is important to keep
in mind that retroviral and lentiviral vectors have a limited cargo
capacity, and that the bigger the final construct is, the lower the
titer of the virus we will be able to produce [25]. Retroviruses
bigger than 8 kb, or lentiviruses bigger than 10 kb, are very hard
to work with. Keep this in mind when thinking about SIN len-
tiviruses with a tissue-specific promoter, your gene of interest,
and IRES-GFP to follow infection. Also, systems that work well
with one transgene may not work with other transgenes.
3. Mice are injected I.P. with 250 μg/g body weight of
5-fluorouracil dissolved in PBS (10 mg/ml), and the bone
marrow cells are harvested 4–5 days later.
4. We use the BD Biosciences Mouse Hematopoietic Progenitor
(Stem) Cell Enrichment Set and magnetic depletion, but there
are other equally good commercial alternatives.
5. 293T cells are grown in a 37 °C degree incubator containing
5 % CO2. The cells should not be allowed to become over-
confluent, and should not be split more than 1:5. 293T cells
easily detach from the tissue culture dishes after approximately
30 s of treatment with trypsin at room temperature (0.05 %
trypsin/0.53 mM EDTA). It is important to freeze multiple
(50–100) vials of the 293T cells after first receiving and
expanding them. This will ensure a ready supply of backup vials
to allow for uniform virus production over several years. Go
back to an early freeze and expand again when running low on
backup vials.
6. This is the most finicky reagent in the process. It normally
takes two or three batches of 2× HBS before finding one that
works well for transfections. The ability of the 2× HBS solu-
tion to produce working CaPO4 precipitates deteriorates after
6 months to 1 year, even when the 2× HBS solution is stored
at −20 °C. Prepare and test a fresh batch before the old one is
finished or too old! Despite this, in our experience no com-
mercial reagent works as well as a good homemade batch.
7. It is recommended to grow the viral backbone in RecA-
negative strains, and at 30 °C, to minimize recombination.
8. Other cytokines may also be added to this cocktail such as Flt3
(50 ng/ml) and TPO (25 ng/ml).
9. Matrix proteins such as fibronectin mediate colocalization of
target cells and vector [26]. The highest gene transfer is
obtained with the recombinant protein CH-296 (RetroNectin).
10. This is important. Retronectin binds much better to non-tissue
culture-treated plates [27].
11. Be careful with your 293T cells. Do not use cells that have
been kept in culture for a long time, or your transfection
efficiency will be reduced. It is extremely important that the
cells are not overly clumped and are at the correct density.
It is essential that the cells are extremely healthy prior to
plating. It is recommended to count the cells rather than

estimating the split.
12. The amount of DNA may be increased up to certain limits,
when it starts to be toxic. This is our default start amount.
13. We normally use pCL/eco as the packaging vector. This can be
substituted by other combinations. Please note that if you pseu-
dotype the retrovirus with VSV-G, as discussed, it is important
to titer the amount of VSV-G env plasmid in the transfection.
Too much results in toxicity, cell fusion, and death, and low
virus titers.
14. For most applications that require infection of primary cells,
it is important to obtain a high titer virus. As a general rule,
the bigger the final vector, the lower the titer, but there are
other factors that affect this process that are not well defined.
If your viral backbone expresses GFP or another surface
marker, virus titer is easy to estimate by infecting a receptive,
exponentially growing cell line. In our laboratory we rou-
tinely use 16610D9, a murine thymoma cell line [28], to test
our viral preps.
15. Most primary hematopoietic and lymphoid cells show reduced
viability when cultured for long periods of time with retroviral
supernatant and other infection reagents such as lipids or poly-
brene. Therefore, we use either spin infection or incubation on
virus-loaded retronectin-coated plates as our preferred method
for retroviral transduction because it minimizes the time cells
are exposed to these reagents.
16. Alternatively, cells can be mixed with the viral supernatant,
incubated in the retronectin plates for 4 h, and then washed
and replated with fresh media. Some viral supernatants can
affect HSC survival and function, so we prefer to preabsorb
the virus to retronectin.
17. They will be weakly adherent, so forceful pipeting or gentle
scraping with a rubber policeman is recommended.
18. The second infection increases the infection efficiency, but also
prolongs the amount of time the cells spend in culture, which
may decrease their ability to reconstitute irradiated hosts. We
tend not to do a second infection, unless the virus we are using
is not very good.
19. If the infection efficiency is very low, infected cells can be
enriched by sorting, and then injected alone (or with rescue
bone marrow if their numbers are low). This approach can
salvage some experiments. We prefer to inject both infected
and uninfected cells, because the uninfected cells provide a
good internal control in each reconstituted animal.
20. Use freshly isolated cells and never put on ice for best results.
21. We have used this approach to infect DP thymocytes, and analyze

the role of different signal transduction pathways on GATA-3
induction [29]. In this case it is useful to culture the infected
thymocytes on a OP9Δ layer, because it seems to improve their
viability. For these experiments we use them in experiments
24–36 h after infection.
References
1. Hawley RG, Fong AZ, Burns BF et al (1992) 11. Kowolik CM, Hu J, Yee JK (2001) Locus
Transplantable myeloproliferative disease control region of the human CD2 gene in a
induced in mice by an interleukin 6 retrovirus. lentivirus vector confers position-independent
J Exp Med 176:1149–1163 transgene expression. J Virol 75:4641–4648
2. Swift S, Lorens J, Achacoso P et al (2001) 12. Marodon G, Mouly E, Blair EJ et al (2003)
Rapid production of retroviruses for efficient Specific transgene expression in human and
gene delivery to mammalian cells using 293T mouse CD4+ cells using lentiviral vectors with
cell-based systems. Curr Protoc Immunol regulatory sequences from the CD4 gene.
Chapter 10:Unit 10.17C Blood 101:3416–3423
3. Burns JC, Friedmann T, Driever W et al (1993) 13. Lois C, Hong EJ, Pease S et al (2002) Germline
Vesicular stomatitis virus G glycoprotein pseu- transmission and tissue-specific expression of
dotyped retroviral vectors: concentration to transgenes delivered by lentiviral vectors.
very high titer and efficient gene transfer into Science 295:868–872
mammalian and nonmammalian cells. Proc Natl 14. Romero Z, Torres S, Cobo M et al (2011)
Acad Sci U S A 90:8033–8037 A tissue-specific, activation-inducible, lenti-
4. Mostoslavsky G, Kotton DN, Fabian AJ et al viral vector regulated by human CD40L
(2005) Efficiency of transduction of highly proximal promoter sequences. Gene Ther
purified murine hematopoietic stem cells by 18:364–371
lentiviral and oncoretroviral vectors under 15. Kraunus J, Schaumann DH, Meyer J et al
conditions of minimal in vitro manipulation. (2004) Self-inactivating retroviral vectors with
Mol Ther 11:932–940 improved RNA processing. Gene Ther 11:
5. Schambach A, Bohne J, Chandra S et al (2006) 1568–1578
Equal potency of gammaretroviral and lentivi- 16. Schambach A, Mueller D, Galla M et al (2006)
ral SIN vectors for expression of Overcoming promoter competition in pack-
O6-methylguanine-DNA methyltransferase in aging cells improves production of self-
hematopoietic cells. Mol Ther 13:391–400 inactivating retroviral vectors. Gene Ther 13:
6. Barrette S, Douglas JL, Seidel NE et al (2000) 1524–1533
Lentivirus-based vectors transduce mouse 17. Schmitt TM, Zuniga-Pflucker JC (2002)
hematopoietic stem cells with similar efficiency Induction of T cell development from hemato-
to moloney murine leukemia virus-based poietic progenitor cells by delta-like-1 in vitro.
vectors. Blood 96:3385–3391 Immunity 17:749–756
7. Laurenti E, Barde I, Verp S et al (2010) 18. Lerner C, Harrison DE (1990) 5-Fluorouracil
Inducible gene and shRNA expression in resi- spares hemopoietic stem cells responsible for
dent hematopoietic stem cells in vivo. Stem long-term repopulation. Exp Hematol
Cells 28:1390–1398 18:114–118
8. Tiscornia G, Tergaonkar V, Galimi F et al 19. Harrison DE, Lerner CP (1991) Most primi-
(2004) CRE recombinase-inducible RNA tive hematopoietic stem cells are stimulated to
interference mediated by lentiviral vectors. cycle rapidly after treatment with 5-fluorouracil.
Proc Natl Acad Sci U S A 101:7347–7351 Blood 78:1237–1240
9. Heinonen JE, Mohamed AJ, Nore BF et al 20. Naviaux RK, Costanzi E, Haas M et al (1996)
(2005) Inducible H1 promoter-driven lentiviral The pCL vector system: rapid production of
siRNA expression by Stuffer reporter deletion. helper-free, high-titer, recombinant retrovi-
Oligonucleotides 15:139–144 ruses. J Virol 70:5701–5705
10. Koo BK, Stange DE, Sato T et al (2012) 21. Pear WS, Miller JP, Xu L et al (1998) Efficient
Controlled gene expression in primary Lgr5 and rapid induction of a chronic myelogenous
organoid cultures. Nat Methods 9:81–83 leukemia-like myeloproliferative disease in
mice receiving P210 bcr/abl-transduced bone 26. Moritz T, Patel VP, Williams DA (1994) Bone
marrow. Blood 92:3780–3792 marrow extracellular matrix molecules improve
22. Lee NS, Dohjima T, Bauer G et al (2002) gene transfer into human hematopoietic cells
Expression of small interfering RNAs targeted via retroviral vectors. J Clin Invest 93:
against HIV-1 rev transcripts in human cells. 1451–1457
Nat Biotechnol 20:500–555 27. Tonks A, Tonks AJ, Pearn L et al (2005)
23. Hernandez-Hoyos G, Alberola-Ila J (2005) Optimized retroviral transduction protocol
Analysis of T-cell development by using which preserves the primitive subpopulation of
short interfering RNA to knock down pro- human hematopoietic cells. Biotechnol Prog
tein expression. Methods Enzymol 392: 21:953–958
199–217 28. Bain G, Quong MW, Soloff RS et al (1999)
24. Paithankar KR, Prasad KS (1991) Precipitation Thymocyte maturation is regulated by the
of DNA by polyethylene glycol and ethanol. activity of the helix-loop-helix protein, E47.
Nucleic Acids Res 19:1346 J Exp Med 190:1605–1616
25. Kumar M, Keller B, Makalou N et al (2001) 29. Gimferrer I, Hu T, Simmons A et al (2011)
Systematic determination of the packaging Regulation of GATA-3 expression during CD4
limit of lentiviral vectors. Hum Gene Ther lineage differentiation. J Immunol 186:
12:1893–1905 3892–3898
Chapter 9
Bone Marrow and Fetal Liver Radiation Chimeras

Abstract
Radiation chimeras are prepared by subjecting recipient mice to sublethal or lethal dose of irradiation and
injecting them with hematopoietic stem cells (HSC) from untreated donor mice. HSC can be obtained
from bone marrow or fetal liver. This technique is a powerful tool when coupled with gene targeting
strategies to investigate function of HSCs, thymocyte development, and T cell function. This protocol
describes how to produce bone marrow or fetal liver chimeras.
Key words Bone marrow, Transplantation, Hematopoietic stem cells, Immune reconstitution, T cell
1 Introduction
The presence of hematopoietic stem cells (HSCs) in bone marrow

was inferred from experiments that arose from studies on the effects
of irradiation on mammals early in the nuclear age [1]. HSCs are
defined by their ability to support long-term reconstitution of all
mature blood cell types. Bone marrow transplantation (BMT) has
become a valuable research tool to study immune cell development
and function especially in mouse gene knockout models. In cases
where gene targeting results in late embryonic death, fetal liver cells
can be used as a source of HSC for transplantation in wild-type
adult mice to allow study of the effects of mutations specific to
immune cell development and function. Based on the work done in
animal models, HSC transplantation has evolved into an effective
treatment for hematopoietic diseases and as an increasingly impor-
tant component of anticancer therapy.
HSC transplantation can be used to differentiate between cell
intrinsic and environmental effects on immune cell development and
function. For example, a targeted mutation that causes a severe defect
in T cell development can be traced to lymphoid cells if the pheno-
type is observed in wild-type mice transplanted with bone marrow
from the mutated mouse. Conversely, a defect in thymic epithelial
cell function would be evidenced by defective T cell development in
109
110 Francis A. Flomerfelt and Ronald E. Gress
spite of transfer of wild-type HSC into the mutant mouse. Competitive

transplantation includes a mixture of HSC from different mouse
strains to assess the ability of mutant cells to function in competition.
This technique allows characterization of reduced function due to
mutations that may not be evident in the absence of competition.
In HCS transplantation experiments it is often critical to be able to
differentiate host from donor cells. This can be done using antibodies
that recognize different congenic alleles (such as CD45.1 and
CD45.2), or the use of transgenic markers such as EGFP. These
markers can be used to assess reconstitution using blood samples or
harvested tissues at experimental endpoints. The timing for reconsti-
tution varies but one should expect to be able to observe donor cell
development in the thymus within 2–3 weeks and stable thymic
reconstitution within 8–9 weeks.
2 Materials
Use sterile technique throughout and work in a laminar flow hood

when possible or practical. Dissection of leg bones can be done
outside of hood.
Water bath, ice, centrifuge, microscope, trypan blue, and
hemocytometer or automated cell counter are used in the proce-
dures listed below.
2.1 Host Mice (See 1. C57BL/6 CD45.2 congenic mice 6–12 weeks old, weight at
Note 1) least 20 g (see Notes 2 and 3).
2. Drinking water acidified with hydrochloric acid to pH 2.5–3.0.
3. Antibiotics: Amoxicillin or other.
2.2 Bone Marrow 1. C57BL/6 CD45.1 congenic mice 6–12 weeks old.
Preparation 2. 10 ml syringes tipped with 25 g needles.
3. Flushing media: RPMI 1640, 2 % fetal calf serum, 1 % penicil-
lin/streptomycin, 1 % HEPES, and 1 % l-glutamine.
4. ACK lysing buffer.
2.3 Optional T Cell 1. Wash media: RPMI 1640, 1 % penicillin/streptomycin.

Depletion 2. Low-Tox rabbit or guinea pig complement.
3. Antibodies (clone name): Anti-CD90.2 (HO-13-4), anti-CD8
(83-12-5), anti-CD4 (C3PO) (see Note 4).
4. DNase—2 mg/ml stock.
2.4 Fetal Liver 1. Flushing media (see above).

Preparation 2. Sterilized dissecting tools (curved forceps, tweezers, small and
large scissors), dissecting microscope.
3. Timed pregnant mice days E12.5–18.5 (see Note 5).
Bone Marrow and Fetal Liver Radiation Chimeras 111
2.5 Analysis 1. Flow cytometer and software to analyze data.

2. 5 ml polystyrene tubes required for flow cytometer.
3. FACS buffer: 1× PBS, 0.5–1.0 % BSA, 0.1 % (W/V) sodium
azide (see Note 6).
4. Antibodies (clone name): Anti-CD45.1-FITC (A20), anti-
CD45.2-PE (104), anti-B220-APC (RA3-6B2), anti-CD90-
APC (53-2.1), anti-CD16 (2.4G2) antibodies.
5. Commercial 1-step fix/lyse solution.
3 Methods
3.1 Prepare 1. One week before irradiation, the recipient mice are given acidi-
Host Mice fied drinking water (to prevent growth of Pseudomonas spe-
cies) supplemented with antibiotics such as amoxicillin
(0.5 mg/ml).
3.2 Prepare Bone 1. Prepare 10 ml syringes with 10 ml of flushing media.

Marrow Suspension 2. Sacrifice donor mice following institutional guidelines and
from Donor Mice soak fur with 70 % ethanol. Snip skin at the base of tail and peel
over legs and up the body to contain hair inside the inverted
skin. Dissect long bones (femur, tibia), and strip off tissue
using sterile forceps, scissors, and gauze (see Note 7). Place
stripped bones in Petri dishes containing ice-cold flushing
media and keep on ice.
3. Snip both ends of bones with sharp scissor or bone clippers to
allow easy entry of needle into one end. Use needle to ensure that
bone is open on both ends if needed. The red marrow should be
easily visible in the bone. Insert needle into one end of the bone
and using some force, flush marrow cavity with 2–3 ml of flushing
media into 50 ml conical tubes on ice. In most cases, the marrow
comes out as an intact tube of red tissue.
4. Agitate gently, and then spin marrow cells at 1000 × g for 5 min
at 4 °C.
5. Carefully remove supernatant from the loosely pelleted cells
and resuspend in complete flushing medium using about 1 ml
for each mouse used for marrow collection.
6. Process cells by pipetting up and down several times and then
run through a cell strainer. Use a sterile forceps or 5 ml syringe
plunger to mash marrow clumps and rinse the strainer with
3 ml complete flushing media. Place marrow cells on ice.
7. Remove 5 μl of the marrow cell suspension for counting and
place in an Eppendorf tube. Add 95 μl of ACK lysis buffer to
the 5 μl cell sample in the Eppendorf tube, mix well, and count
the cells.
3.3 Optional T Cell T cell depletion of marrow eliminates resident T cells in the graft
Depletion of Marrow which allows analysis of T cell function of cells derived from
engrafted HSC and prevents graft-versus-host disease (GVHD) if
MHC mismatches are present (see Note 2).
1. Place counted bone marrow cells in 50 ml tube and fill with wash-
ing media, spin down cells, and remove media (see Note 8).
2. Prepare bone marrow cell suspension at 2.0 × 107 cells/ml in
RPMI/PSG without FCS.
3. Add antibodies at predetermined concentrations and mix gently
(see Note 4).
4. Place on ice for 15 min.
5. Wash bone marrow cells with RPMI/PSG without FCS.
Do not resuspend pellet.
6. Add diluted guinea pig complement to bring cell number to
2.0 × 107/ml (see Note 9). Add DNase to 40 μg/ml of cell
suspension. Mix well.
7. Place in 37 °C water bath × 40 min.
8. Wash with RPMI/PSG and resuspend cells in 50 % volume
used in step 1. Count cells and resuspend at 5.0 × 107/ml in
sterile PBS (see Note 10).
3.4 Prepare Liver 1. Harvest E14–E15 embryos and place in flushing media in a
Cell Suspension 15 cm tissue culture dish on ice (see Note 11).
2. Once cooled, rinse embryo in flushing media, place on a piece
of gauze, and euthanize the following institutional guidelines.
Remove a piece of tissue (tail, foot) and freeze for genotyping
if needed (see Note 12).
3. Dissect liver and place in sterile Eppendorf tube with 1 ml
flushing media. Be sure that liver is numbered the same as
genotyping biopsy.
4. Prepare cell suspension by mashing and pipetting the liver up
and down first with a P1000 tip and then with a P200 tip.
5. Allow cells to settle for 5 min on ice. Transfer cell suspension
to a new sterile Eppendorf tube taking care to avoid debris that
has settled to the bottom of the tube (see Note 13).
6. Strain cells with a 40 μM filter. Count viable cells and adjust to
0.25–1 × 107 cells/ml in sterile PBS for injection.
3.5 Inject HSC 1. At least 4–6 h before injection, irradiate recipient mice using
into Mice 9.5 Gy (see Note 14). It is advisable to wait ~2 h after the last
irradiation dose before injecting cells.
2. Inject 1.0 × 107 depleted BM cells in 200 μl sterile PBS in the
tail vein of the irradiated recipient mouse. Alternatively inject
0.5–2 × 106 fetal liver cells in 200 μl sterile PBS (see Note 15).
3. Keep mice on acidified water containing antibiotics and in

autoclaved cages for 2 weeks after injection of stem cell
inoculum.
4. Wait for 1 month before testing for T cell chimerism. Peripheral
T cell reconstitution is complete after approximately 8 weeks.
3.6 Analysis Immune reconstitution can be monitored using blood samples.

of Chimera If bone marrow was T cell-depleted, the presence of donor B cells
Reconstitution can be used to quickly determine if transplant was successful.
1. Obtain 100 μl heparinized blood samples from transplanted
mice and divide into two FACS tubes. For controls, obtain
150 μl blood samples from a CD45.2 mouse and divide into
three FACS tubes (see Note 16). Obtain 50 μl blood from a
CD45.1 mouse and place into one FACS tube. Add 0.5 μl
anti-CD16/CD32 antibody to block antibody binding by the
Fc receptor and incubate on ice for 10 min.
2. Add antibody mixes (CD45.1, CD45.2, and B220 or CD45.1,
CD45.2, and CD90) to blood from transplanted mice at pre-
determined concentrations (0.25–1.0 μl/sample should work).
Add single antibodies to the four control tubes. Incubate on
ice for 15 min.
3. Add 1 ml 1-step fix/lyse solution to each FACS tube, mix gen-
tly, and incubate for 15 min minimum (see Note 17).
4. Spin down cells at 500 × g for 5 min at room temperature.
Remove supernatant and resuspend cells in 1 ml FACS buffer,
spin down again, remove supernatant, and add 500 μl FACS
buffer.
5. Use single-color controls to set up compensation on cytometer
and run samples. If successful, a population of donor CD45.1-
positive cells should be readily apparent. Gate on the donor
cells to determine levels of B cell (B220) and T cell (CD90)
chimerism. Depending on results, proceed to experimental
endpoints (see Note 18).
4 Notes
1. Ensure that institutional guidelines are followed for all animal

housing, handling, procedures, and euthanasia.
2. It is important to match MHC antigens of donor and host
mice to avoid complications due to GVHD or host-versus-
graft disease (HVGD) unless BMT is being used to study these
syndromes in animal models [2].
3. Here we use allelic differences in CD45 to identify host and
donor cells. Another option is the use of a transgenic marker,
such as EGFP expression by donor mouse cells.
4. We use supernatants from hybridoma cultures as an antibody

source. These are IgM antibodies to enhance complement
lysis. Other antibodies and isotypes (IgG) can be used but all
antibodies should be titrated beforehand to optimize efficiency.
Titrations ranging from 0.25 to 5.0 μg/ml should be
sufficient.
5. Timed pregnant mice are commercially available for commonly
used strains. Alternatively, any strain of mice can be used to
produce timed pregnant mice following procedures outlined
by Mader et al. [3].
6. Fetal bovine serum can also be used at 3–10 % instead of
BSA. Sodium azide is used as a preservative; its addition to the
buffer is optional.
7. Bone marrow from humerus, hips, and vertebral bodies can
also be used to increase the number of cells. Vertebral bodies
and hips can be crushed and bone fragments filtered out.
8. Ensure that cells are thoroughly washed with at least 10 volumes
to eliminate serum proteins.
9. Nonspecific cell loss is variable, but can be substantial using
complement treatment. The following suggestions can help to
minimize nonspecific cell lysis. Ensure that cells, centrifuge, and
buffers are kept cool or on ice at all times. Use care to gently
resuspend cell pellets, especially after washes. Antibodies should
be titrated. New lots of complement should be prepared accord-
ing to the specification sheet and dilutions tested for efficient
lysis and minimal nonspecific killing. Typically a 1:5–1:20 dilu-
tion should be sufficient. Low-toxicity rabbit complement may
work better if IgG antibodies are used, or may be used instead
or in combination with guinea pig complement to improve
depletion or reduce unwanted toxicity. There are several reli-
able commercially available alternative methods to effectively
deplete defined cell populations. For example commercially
available kits can be used to deplete various populations using
antibodies coupled with a magnetic bead capture strategy.
Although convenient (especially for small-scale depletions),
these kits and the needed equipment can be costly. For refer-
ence, bone marrow of normal mice should contain about 5 %
T cells and about 12–20 million marrow cells/mouse after T cell
depletion.
10. Use flow cytometry (Subheading 3.6) to assess the efficiency
of T cell depletion using anti-CD90 or similar antibodies.
If needed, another cycle of complement depletion can be done.
If using two cycles, results may be improved if different anti-
bodies are used in each cycle.
11. Hematopoiesis peaks in liver at days E14–E15 making this an
ideal time to harvest HSC. However, HSC can be harvested
from fetal liver from days E12.5–E18.5 if needed [4].
12. Use care to avoid cross contamination between embryos and

avoid contamination with mother’s blood.
13. Fetal liver cells can be frozen at this point in FCS with 10 %
DMSO and stored in liquid nitrogen. To prepare cells for
injection, thaw cells, add 50 ml cold PBS, incubate for 10 min,
spin down cells, and repeat cold PBS wash to eliminate DMSO.
Filter cells, count, and adjust as described in procedure.
14. C57BL/6 mice are relatively resistant to radiation effects and
can be given a single 9.5–10.0 Gy dose. Irradiation can be
divided into two 4.75 Gy treatments 2–3 h apart to reduce
potential radiation injury. Immunodeficient mice such as Rag2
knockout can be injected without irradiation or should only be
given 6.0 Gy. BALB/c mice are radiation sensitive and should
be given two 4.50 Gy doses.
15. Fewer cells can be injected, down to about 1 × 106 per mouse,
but may delay reconstitution. Bone marrow cells can also be
injected retro-orbitally into anesthetized mice. Newborn mice
can be injected intraperitoneally.
16. This procedure works equally well with heparin- or EDTA-
treated blood.
17. Cells can be stored for up to 48 h in the 1-step fix/lyse solution
following the manufacturer’s recommendations.
18. The choices for antibodies to stain cells are great and are dictated
by experimental design. In general, consider using fluoro-
chromes (such as Pacific-Blue) that have minimal overlap in
other channels for gating on donor cells. Preliminary
experiments should be done to optimize fluorochrome choices
and antibody concentrations. Be sure to include appropriate
controls to allow accurate gate setting for data analysis.
Acknowledgement
The Intramural Research Program of the National Cancer Institute

at the National Institutes of Health supports the authors. The
authors have no conflicts of interest to disclose.
References
1. de la Morena MT, Gatti RA (2011) A history of in two strains of genetically engineered mice.
bone marrow transplantation. Hematol Oncol Lab Anim (NY) 38:305–310.
Clin North Am 25:1–15 4. Gudmundsson KO, Stull SW, Keller JR (2012)
2. Hakim F, Fowler DH, Shearer GM, Gress RE Transplantation of mouse fetal liver cells for
(2001) Animal models of acute and chronic analyzing the function of hematopoietic stem
graft-versus-host disease. Curr Protoc and progenitor cells. Methods Mol Biol 879:
Immunol Chapter 4:Unit 4 3 123–133
3. Mader SL, Libal NL, Pritchett-Corning K, Yang R,
Murphy SJ (2009) Refining timed pregnancies
Chapter 10
In Vitro Analyses of T Cell Effector Differentiation

Elizabeth A. Wohlfert, Andrea C. Carpenter, Yasmine Belkaid,
and Rémy Bosselut
Abstract
In vitro culture is an important complement, or substitute, to in vivo approaches in order to study T cell
effector differentiation. Here, we describe culture conditions that generate specific effector cell types by
exposing naïve T cells to appropriate cytokine signals.
Key words T cell differentiation, Tc, Th0, ThN, Th1, Th2, Th17, iTreg
1 Introduction
Effective T cell function is required for protection from invading

pathogens. T cell effector differentiation is determined by several
signals, notably from the innate immune cells, including (1) stimu-
lation through the T cell receptor (TCR), (2) co-stimulation
through the CD28 costimulatory molecule and (3) cytokine expo-
sure that induces acquisition of specific T cell differentiation. Studies
exploring how naïve T cells differentiate into fully functional effector
cells often require assessing the effector potential of these T cells;
protocols for such analyses are presented in this chapter.
During in vivo immune responses, effector T cell differentiation
occurs in secondary lymphoid organs or tissues, and is driven by
architecturally constrained interactions between T cells, antigen-
presenting cells (APC)s and other immune cells. Early stages of
an immune response involve small cell numbers of antigen-specific
T cells, which may be difficult to identify and purify. Thus, it can
be advantageous to reproduce the conditions that promote T cell
effector differentiation in vitro. In vitro differentiation allows
acquisition of a large number of cells, e.g., for biochemical or gene
expression studies. In addition, in vitro studies allow for tighter
control of cytokines and other stimuli offered to T cells. These
in vitro cultures typically include two distinct stimuli. First, a ligand
117
118 Elizabeth A. Wohlfert et al.
Table 1
Effector differentiation in vitro
Conditioning Cytokines (2× Blocking antibodies Cytokines (final Blocking antibodies

mix concentrations) (2× concentration) concentration) (final concentration)
ThN IL-2 (20 ng/ml) – IL-2 (10 ng/ml) –
Tc IL-2 (20 ng/ml) – IL-2 (10 ng/ml) –
Th17 IL-6 (20 ng/ml) Anti-IL-4 (20 μg/ml) IL-6 (10 ng/ml) Anti-IL-4 (10 μg/ml)
TGF-β (5 ng/ml) Anti-IFNγ (20 μg/ml) TGF-β (2.5 ng/ Anti-IFNγ (10 μg/ml)
Anti-IL-12 (20 μg/ml) ml) Anti-IL-12 (10 μg/ml)
+Th0 IL-2 (20 ng/ml) Anti-IL-12 (20 μg/ml) IL-2 (10 ng/ml) Anti-IL-12 (10 μg/ml)
Anti-IFNγ (20 μg/ml) Anti-IFNγ (10 μg/ml)
Anti-IL-4 (20 μg/ml) Anti-IL-4 (10 μg/ml)
IL-12 (20 ng/ml) IL-12 (10 ng/
ml)
IL-4 (20 ng/ml) Anti-IFNγ (20 μg/ml) IL-4 (10 ng/ml) Anti-IFNγ (10 μg/ml)
iTreg TGF-β (5 ng/ml) Anti-IL-4 (20 μg/ml) TGF-β (2.5 ng/ Anti-IL-4 (10 μg/ml)
IL-2 (20 ng/ml) Anti-IFNγ (20 μg/ml) ml) Anti-IFNγ (10 μg/ml)
IL-2 (10 ng/ml)a
a
IL-2 is added at 48 h
for the T cell antigen receptor complex is required to promote

expression of specific cytokine receptors (e.g., for IL-2) and cell
proliferation. This can be a specific peptide antigen if using cells of
defined specificity (e.g., carrying a TCR transgene); in many
instances however, antibodies against TCR or CD3 are used to
mimic antigen stimuli and trigger TCR signaling. In both cases, TCR
stimulation needs to be accompanied by engagement of CD28
costimulatory molecules. The second series of ligands is intended
to direct cytokine gene expression; it includes cytokines and anti-
cytokine antibodies to neutralize the effect of unwanted cytokines.
Appropriate combinations of these reagents typically “polarize”
T cell differentiation into specific effector fates: for CD4 cells, these
generally include Th1 [1], Th2 [1–3], Th17, and inducible (i) T
regulatory (Treg); and for CD8+ cytotoxic T (Tc) cells (see Table 1
and refs. 2, 4, 5). While it is possible to activate highly purified T
cells with the latter reagent combinations, using separately purified
APCs as “feeder” cells for the differentiating effector T cells results
in greater survival for most effector types [6].
During in vitro culture, antigenic stimulation (or its surrogate)
and cytokines drive cell proliferation, which typically starts within
24–36 h after stimulation and continues for 3 or 4 days. Expression
of cytokine genes, and cytokine production (evaluated by ELISA
In Vitro Analyses of T Cell Effector Differentiation 119
Table 2
Effector differentiation in vivo
Cytokine for differentiation Transcription factor Cytokine produced

Th1 IL-12 Tbet IFNγ
Th2 IL-4 Gata3 IL-4
IL-5
IL-13
Th17 IL-6 RORγt IL-17A
TGF-β IL-17F
IL-1 IL-22
IL-21
iTreg TGF-β Foxp3 IL-10
Retinoic acid TGF-β
IL-2
Tc IFNγ Tbet IFNγ
or intracellular cytokine staining) is detected within 3–5 days of

stimulation, depending on the type of cytokine. Similar kinetics are
observed for fate-determining transcription factors.
Although the choice of the starting T cell population is typi-
cally dictated by the specific application, special emphasis must be
placed on separating naïve from antigen-experienced (generally
referred to as “memory”) cells obtained from peripheral lymphoid
organs. In laboratory mice housed under specific pathogen-free
conditions, most spleen and lymph node T cells are “naïve.” That
is, they are directly derived from thymic precursors without having
encountered the antigen that their TCR specifically reacts against
and therefore, they do not express effector (e.g., cytokine) genes.
A simple conceptual example of a naive cell is a lymphocyte carrying
a TCR directed against a virus-derived peptide in a host that has
not been in contact with that particular virus. In contrast, upon
infection or immunization with the cognate antigen, these newly
“antigen-experienced” cells proliferate, and acquire effector prop-
erties or differentiate into memory cells. In unmanipulated labora-
tory mice, cells exhibiting marks of antigen experience are typically
reactive against commensal and environmental antigens.
Regardless of whether they are effector or memory, antigen-
experienced cells have two properties not shared by naïve cells.
They are generally “preprogrammed” to produce specific cytokines
(most antigen-experienced cells in a mouse spleen make IFNγ), and
they produce these cytokines quickly, within hours of antigen recep-
tor triggering. In contrast, naïve cells typically do not produce
effector cytokines until they have undergone multiple rounds of
proliferation, and their effector differentiation is heavily influenced
by the surrounding cytokine milieu (Table 2). Thus, in activation
cultures, the cytokines produced by antigen-experienced cells have

the potential to skew, often towards IFNγ production, the effector
differentiation of the naïve cells. To avoid this potential bias, it is
generally advisable to purify naïve cells for in vitro stimulation cul-
tures. This is all the more necessary when such cells are prepared
from inflammatory contexts, in which the frequency of effector or
memory cells is much higher. For both CD4+ and CD8+ mouse
T cells, CD44 is the most commonly used marker to distinguish
naïve (CD44lo) from antigen experienced (CD44hi) subsets.
Here we delineate protocols designed to evaluate the effector
potential of naïve T cells exposed to conditions that partially mimic
in vivo antigen stimulation. These protocols are designed to differ-
entiate naïve T cells into CD4+ iTregs and CD4+ and CD8+ T effector
cells. Because of the versatility of in vitro cultures, the choice of
appropriate controls is crucial to establish sound conclusions.
We have provided suggestions in Subheading 4 in this regard.
The protocols described below use antibody (anti-CD3)-medi-
ated TCR triggering as a surrogate for antigen stimulation, APCs
(either dendritic cells or T cell-depleted splenocytes), and mixes of
cytokines and anti-cytokine antibodies appropriate for promoting
the differentiation of Th1, Th2, or Th17 CD4+ effectors, Tc CD8+
effectors, or CD4+ iTreg cells. We also provide an alternate protocol
that eliminates the APCs.
2 Materials
2.1 Blocking 1. Anti-CD3e (without azide, unlabeled), Clone 145-2C11.

Antibodies 2. Anti-CD28 (without azide, unlabeled), Clone 37.51.
3. Anti-IFN-γ (without azide, unlabeled), Clone XMG1.2.
4. Anti-IL-4 (without azide, unlabeled), Clone 11B11.
5. Anti-IL-12 p40/p70 (without azide, unlabeled), Clone
C17.8.
2.2 Cytokines 1. Recombinant murine IL-2.

2. Recombinant murine IL-6.
3. Recombinant human TGF-β.
2.3 Beads 1. Mouse pan T (CD90.2) depletion magnetic beads (see Note 1).
2.4 Media 1. Fetal calf serum (FCS) (see Note 2).

Components 2. Phosphate buffered saline (PBS, Ca2+ Mg2+ free) pH 7.4.
3. Iscove’s Modified Dulbecco’s Medium (IMDM) (see Note 3).
4. Culture medium: 10 % FCS, 100 U/ml of penicillin, 0.10 mg/

ml of streptomycin, 0.292 mg/ml of l-glutamine, 10 mM
HEPES (N-2-hydroxyethylpiperazine-N-2-ethane sulfonic
acid) pH range 7.2–7.5 (see Note 4), 55 μM 2-Mercaptoethanol
(see Note 5), 1 mM Sodium pyruvate (see Note 6), 1× MEM
nonessential amino acids (NEAA) (see Note 7), in RPMI 1640.
5. Digestion medium: 100 U/ml of penicillin, 0.10 mg/ml of
streptomycin, 0.292 mg/ml of l-glutamine, 10 mM HEPES
(N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid) pH
range 7.2–7.5 (see Note 4), 0.25 mg/ml of deoxyribonuclease
I from bovine pancreas ≥85 % protein, and 0.5 mg/ml of
Liberase TL (Roche), in RPMI 1640.
6. 500 mM ethylenediamine tetraacetate (EDTA) pH 8.0.
7. Staining medium: PBS (Ca2+Mg2+ free) pH 7.4, 0.1 % bovine
serum albumin (BSA), 2 mM EDTA pH 8.0.
8. Isolation medium: PBS (Ca2+Mg2+ free) pH 7.4, 0.1 % BSA,
2 mM EDTA pH 8.0.
9. Red cell lysis buffer: 8290 mg/l ammonium chloride, 1 g/l
potassium bicarbonate, 37 mg/l EDTA pH 8.0.
2.5 Hardware 1. 96-well flat bottom tissue culture treated plates.

2. 96-well round bottom tissue culture treated plates.
3. Cell scraper.
4. 70 μm cell strainers.
5. Nylon tissue filters (pore size 100 μm, cut to 2 or 4 cm2).
6. 1 or 3 ml syringe pestles.
7. 60 mm petri dishes.
8. 15 and 50 ml conical tubes.
9. Magnetic beads compatible with appropriate purified antibodies.
10. Rocker.
11. Two curved forceps.
12. Scissors appropriate for dissection.
13. 37 °C incubator with 5 % CO2.
14. Vertical magnet holding 15 or 50 ml conical tube (DynaMag
15, 50, or equivalent).
15. 5 ml tuberculin syringe.
2.6 Fluorescent 1. Anti-CD4, Clone RM-4.5.

Antibodies for Flow 2. Anti-CD44, Clone 1M7.
Cell Sorting
3. Anti-CD25, Clone 7D4.
(See Note 8)
4. Anti-CD11c, Clone N418.
5. Anti-MHC Class II (see Note 9).
2.7 Animals 1. Mice of interest for preparation of T cells.

(See Note 10) 2. Mice of interest for preparation of dendritic cells or T-depleted
splenocytes (see Note 11).
3 Methods
Subheadings 3.1–3.3 describe a protocol to stimulate T cells in the

presence of antigen presenting cells (APCs). The protocol involves
three distinct steps: (1) prepare a suspension of purified APCs to
which antibody and cytokines will be added to form the “condi-
tioning mix”; (2) isolate naïve T cells; and (3) set up cultures by
mixing the APC (in the conditioning mix) and T cell suspensions.
Addition of antibody and cytokines to APCs, to form conditioning
mixes, should be performed last, immediately before setting up the
cultures and aliquoted into the culture wells promptly.
Subheading 3.4 presents an alternative protocol for T cell
stimulation on antibody-coated plates, without APCs.
3.1 Preparation Carry out all procedures at 4 °C unless otherwise specified.

of APCs Completing the procedures outlined in the section below produces
a suspension of APCs in 100 μl per 96 well to which the conditioning
mixes are added. Adding the conditioning mixes to the APCs
streamlines the procedure.
3.1.1 Isolation of Splenic For naïve T cells conditioned in Th0 Th1, Th2, Th17, or Tc condi-
CD11c+ DC tions: use a ratio of 25–50 T cells for every DC. iTregs Conditioning
mix should include a 5–25 ratio of T cells for every DC cultured
( see Note 12). The following procedure is for one spleen and
volumes can be scaled proportionally as needed. All work should be
performed under sterile conditions in a tissue culture cabinet.
1. Place 5 ml of Digestion medium into a 60 mm petri dish.
Surgically remove the spleen from a euthanized mouse and
place directly into Digestion medium.
2. While holding the spleen with forceps, inject spleen with the
Digestion medium using a tuberculin syringe. Then, using
scissors, cut spleen into five to ten smaller pieces.
3. Place in 37 °C incubator with 5 % CO2 for 20 min.
4. Add a final concentration of 5 mM EDTA to the dish and incu-
bate five additional minutes to stop the enzymic activity of the
Digestion medium.
5. Transfer spleen pieces and medium to a 50 ml conical tube and
pass through a 70 μm cell strainer. Use cell scraper to remove
any adherent cells and rinse the petri dish thoroughly with
5 ml Staining medium.
2x Conditioning mix
APCs in
100 µL 100 µL
200 µL
96 well plate
Fig. 1 Schematic of procedure
6. Lyse with 1 ml Red cell lysis buffer for 2 min on ice, wash by
filling tube to 10 ml with Staining medium.
7. Resuspend in 5 ml Staining medium and count cells.
8. Refer to Chapter 7 on “Isolating T cell subsets” for staining
cells with fluorescently labeled antibodies and sorting
CD11c+MHC II+.
9. Prepare DC suspensions in Culture medium at a concentration
of 1–2 × 103 cells/100 μl/well for Th0, Th1, Th2, Th17, or Tc
cultures, or of 2–10 × 103 cells/100 μl/well for iTreg cultures
(see introduction to this section and Fig. 1 for more informa-
tion). Refrain from aliquoting cells into wells at this point as
the Conditioning mix will be added to this master mix. Keep
DCs on ice until ready to prepare Conditioning mixes.
10. See Subheading 3.3 for preparing Conditioning mixes.
3.1.2 Irradiated T Carry out all procedures at 4 °C unless otherwise specified and
Cell-Depleted Splenocytes under sterile conditions in a tissue culture cabinet. The ideal ratio
for this type of APC is 5 for every one T cell.
1. See Chapter 7 on “Isolating T cell subsets” for procedure on
isolating a single-cell suspension from a mouse spleen.
2. Immunomagnetic cell separation for depletion of pan T cells.
The following protocol is for a 1 × 108 total splenocyte suspen-
sion, and can be scaled up or down as necessary. While this
protocol utilizes immunomagnetic beads from Dynabead,
users may employ immunomagnetic beads of their choice with
the appropriate protocol.
3. Preparation of Dynabeads: Resuspend beads fully (either vortex
whole vial or place on a tilt rotator for at least 5 min).
4. Remove 1 × 108 beads and wash in equal volume of Isolation

buffer in a fresh tube. Use 1 ml of Isolation buffer to wash if
bead volume is under 1 ml.
5. Place tube without lid in DynaMag-15 (or other magnet suit-
able for bead isolation procedures) for 1 min and discard the
supernatant.
6. Remove tube from magnet and resuspend beads in the original
starting volume of Isolation buffer.
7. Preparation of pan T cell-depleted splenocytes: Spin down
1 × 108 splenocytes in a separate tube and resuspend in 1 ml of
Isolation buffer.
8. Add 1 × 108 pan T cell CD90.2 Dynabeads to cells and incubate
for 30 min at 4 °C.
9. Place tube containing cell and bead mixture without lid onto a
magnet for 2 min.
10. The T cells will bind to the beads and be removed from the
supernatant. Move T-depleted supernatant in a fresh tube,
wash with Culture medium, spin down and resuspend in
Culture medium to count.
11. Irradiate T-depleted splenocytes with 3,000 RADS (see Note
13 on irradiator use).
12. Make a master mix of 2.5 × 105 T cell-depleted splenocytes in
100 μl Culture medium per well (see introduction to this sec-
tion and Fig. 1 for more information). Refrain from aliquoting
cells into wells at this point as the Conditioning mix will be
added to this master mix. Keep DCs on ice until ready to
aliquot for culture.
13. See Subheading 3.3 for preparing Conditioning mixes.
3.2 T Cell Carry out all procedures at 4 °C unless otherwise specified.

Preparation
1. Isolate T cells from spleens and lymph nodes using either flow
cytometry sorting or alternatively, an immunomagnetic bead
selection approach (see Chapter 7 on “Isolating T cell subsets”,
see Note 14 on caveats to using bead selection over flow
sorting.)
2. Resuspend T cells at 5.0 × 104/100 μl in Culture medium.
Keep this on ice until ready to aliquot for culture.
3.3 Conditioning To the APCs prepared in Subheading 3.1, add the following reagents
Mixes to prepare the corresponding 2× Conditioning mixes. These should
be prepared immediately before setting up cultures and kept on ice
until ready to aliquot for culture. Note that all concentrations
given below refer to the concentration in the conditioning mix,
therefore twice as high as in the final culture (see Table 1).
3.3.1 ThN or CD8+ Tc 1. Add: 2 μg/ml anti-CD3, 6 μg/ml anti-CD28, 20 ng/ml IL-2.
Conditioning Mix
3.3.2 Th17 Substituting IMDM for RPMI increases the frequency of IL-17
Conditioning Mix producing cells (see Note 7). The Th17 Conditioning mix does
not contain IL-2.
1. Add: 2 μg/ml anti-CD3, 6 μg/ml anti-CD28, 20 ng/ml IL-6,
5.0 ng/ml TGF-β, 20 μg/ml anti-IL-4, 20 μg/ml anti-IFN-γ,
20 μg/ml anti-IL-12.
3.3.3 Th0 1. Add: 2 μg/ml anti-CD3, 6 μg/ml anti-CD28, 20 ng/ml IL-2,

Conditioning Mix 20 μg/ml anti-IL-12, 20 μg/ml anti-IFNγ, 20 μg/ml
anti-IL-4.

Conditioning Mix 20 ng/ml IL-12, 20 μg/ml anti-IL-4.

Conditioning Mix 20 ng/ml IL-4, 20 μg/ml anti-IL-12, 20 μg/ml anti-IFNγ
(see Note 15).
3.3.6 iTreg 1. Add: 2 μg/ml anti-CD3, 5.0 ng/ml TGF-β, 20 μg/ml anti-
Conditioning Mix IL-4, 20 μg/ml anti-IFNγ (see Note 16). (Retinoic acid can
be added to the culture to enhance the induction of Foxp3,
see Note 17.)
2. At 48 h, add 10 ng/ml IL-2 (final concentration) and incubate
for an additional 24 h. This is earlier than the day 3 culture split
recommended for non-iTreg cultures (see Subheading 3.5).
3.3.7 Culture Set Up 1. Add 100 μl/well of each APC-containing conditioning mix to
wells in a 96 well round bottom tissue culture plate.
2. Add 100 μl T cell suspension (5 × 104 cells) to each APC-
containing well. The final volume should now be 200 μl
per well.
3. Place in a 37 °C incubator with 5 % CO2 for 3 days
(see Subheading 3.5) (see Note 18).
3.4 Alternate Plate-bound anti-CD3/CD28 antibody stimulation is useful when

Protocol: Plate-Bound examining direct effects on T cells without the complications of a
Anti-CD3/CD28 different cell population. Note that anti-CD28 should be omitted for
Stimulation iTreg conditions (see Note 16). T cell preparation in this alternate
protocol is performed as in Subheading 3.2. If using this protocol,
additional anti-CD3/CD28 in the Conditioning mix is not recom-
mended, as the plate-bound anti-CD3/CD28 is sufficient.
3.4.1 Preparation 1. Make a solution of 1 μg/ml of anti-CD3 antibody and 1 μg/ml

of Anti-CD3 and Anti-CD28 anti-CD28 antibody in room temperature PBS.
Coated Plates
2. Immediately aliquot 100 μl into each well.

3. Place at 4 °C overnight or 1 h at 37 °C.
4. Wash plate two times with PBS by gently adding to the side
of the wells and aspirating off. Wash last time in Culture
medium taking care to not let the wells dry out. If necessary,
leave the last Culture medium wash in the well until the T
cells are isolated and the Conditioning mixes are prepared
(see Subheadings 3.2 and 3.3).
3.4.2 Preparation Prepare 2× cell-free conditioning mixes for each stimulation condi-
of APC-Free Conditioning tion by supplementing APC-free Culture medium with cytokines
Mixes and anti-cytokine antibodies (but not soluble anti-CD3 and soluble
anti-CD28) using concentrations indicated in Subheading 3.3 above
(see also Table 1).
3.4.3 Culture Set Up 1. Per well in a 96 well flat bottom tissue culture plate, add
100 μl/well of APC-free conditioning mix and 100 μl T cell
suspension (5 × 104 cells, prepared as in Subheading 3.2), for a
final volume of 200 μl per well.
2. Place in a 37 °C incubator with 5 % CO2 for 72 h (see
Subheading 3.5).
3.5 Split the Culture 1. Incubate for 72 h and then split each 200 μl well into two
at Day 3 100 μl wells and add 100 μl of Culture medium to each
well.
2. Incubate an additional 24 h.
3. Harvest cells for downstream applications.
4 Notes
1. Syngeneic mouse strains express either CD90.1 or CD90.2

allelic isoforms. Beads exist against each isoform and should
be chosen according to the mouse strain used to prepare
APCs.
2. Fetal calf serum should be heat inactivated by incubation at
56 °C for 50 min. Serum from different sources and even dif-
ferent lots can vary in efficacy. Each lot should be tested as it
may differentially support T generation and cytokine produc-
tion. Companies usually offer samples for this purpose.
3. Using IMDM in Th17 culture increases the efficacy of Th17
differentiation. This medium contains natural agonists for aryl
hydrocarbon receptor that support differentiation of Th17
cells [7].
4. HEPES extends media stability outside of a CO2 incubator.
5. 2-Mercaptoethanol reduces oxygen radicals in Culture

medium. This chemical is not stable in solution and should be
added immediately before use.
6. While not essential for cell growth, Sodium pyruvate is benefi-
cial for most effector differentiation and especially recom-
mended for iTreg induction cultures.
7. NEAA are not required but promote cell growth and viability.
8. The fluorochrome choice for antibodies for flow cytometric
sorting of populations should be carefully considered during
planning. These antibodies can remain on cells and interfere
with secondary staining post culture.
9. Choose an antibody that will recognize MHC-II isoforms
expressed in the mouse strain used as a source of CD11c+ DCs.
10. Animals must be housed, handled, and used according to appli-
cable guidelines and after securing required authorizations.
11. APCs (either dendritic cells or T cell-depleted splenocytes)
should come from animals that are sex and strain matched to
the experimental T cells.
12. Optimal ratio of T cells to APC (i.e., DCs or T-depleted spleen)
should be determined for each specific situation. The ratios
suggested here are starting points that may need to be
optimized.
13. Secure required authorization training before irradiator use,
and follow applicable guidelines.
14. Most bead isolation kits for CD4+ or CD8+ T do not remove
CD44hi (memory or effector cells). Upon antigen stimulation
(anti-CD3 exposure in culture), cytokines produced by CD44hi
cells can bias cell differentiation independently from those
included in the conditioning medium. It is therefore preferable
to exclude CD44hi cells from in vitro cultures.
15. Overgrowing the cells will result in a loss of IL-4 production.
Split them earlier if they are turning yellow, or start from lower
cell numbers.
16. iTreg conditions should include polarizing cytokines and
blocking antibodies, but no anti-CD28.
17. Retinoic Acid enhances induction of de novo Foxp3-expressing
cells [8–10].
18. When plating cells, avoid wells at the edge of the plate—these
wells will experience more evaporation. To the surrounding
wells add equal volumes of your final culture volume of sterile
PBS or water to help reduce evaporation over the course of
your cultures.
References
1. Seder RA, Paul WE (1994) Acquisition of pathways of T cell activation. J Exp Med 159:
lymphokine-producing phenotype by CD4+ 881–905
T cells. Annu Rev Immunol 12:635–673 7. Veldhoen M, Hirota K, Christensen J, O’Garra
2. Kupper T, Horowitz M, Lee F, Robb R, Flood A, Stockinger B (2009) Natural agonists for
PM (1987) Autocrine growth of T cells inde- aryl hydrocarbon receptor in culture medium
pendent of interleukin 2: identification of are essential for optimal differentiation of Th17
interleukin 4 (IL 4, BSF-1) as an autocrine T cells. J Exp Med 206:43–49
growth factor for a cloned antigen-specific 8. Mucida D, Park Y, Kim G, Turovskaya O, Scott
helper T cell. J Immunol 138:4280–4287 I, Kronenberg M, Cheroutre H (2007)
3. Swain SL, Weinberg AD, English M, Huston G Reciprocal TH17 and regulatory T cell differ-
(1990) IL-4 directs the development of entiation mediated by retinoic acid. Science
Th2-like helper effectors. J Immunol 145: 317:256–260
3796–3806 9. Coombes JL, Siddiqui KR, Arancibia-Cárcamo
4. Andrus L, Granelli Piperno A, Reich E (1984) CV, Hall J, Sun CM, Belkaid Y, Powrie F (2007)
Cytotoxic T cells both produce and respond to A functionally specialized population of mucosal
interleukin 2. J Exp Med 159:647–652 CD103+ DCs induces Foxp3+ regulatory T cells
5. Zhu J, Yamane H, Paul WE (2010) via a TGF-beta and retinoic acid-dependent
Differentiation of effector CD4 T cell popula- mechanism. J Exp Med 204:1757–1764
tions. Annu Rev Immunol 28:445–489 10. Sun CM, Hall JA, Blank RB, Bouladoux N,
6. Ashwell JD, Defranco AL, Paul WE, Schwartz Oukka M, Mora JR, Belkaid Y (2007) Small
RE et al (1984) Antigen presentation by rest- intestine lamina propria dendritic cells promote
ing B cells. Radiosensitivity of the antigen- de novo generation of Foxp3 T reg cells via
presentation function and two distinct retinoic acid. J Exp Med 204:1775–1785
Part III
Specific Techniques
Chapter 11
Studying T Cell Development in Thymic Slices

Jenny O. Ross, Heather J. Melichar, Joanna Halkias, and Ellen A. Robey
Abstract
Recently, tissue slices have been adapted to study both mouse and human T cell development. Thymic slices
combine and complement the strengths of existing organotypic culture systems to study thymocyte differ-
entiation. Specifically, the thymic slice system allows for high throughput experiments and the ability to
introduce homogenous developmental intermediate populations into an environment with a well-established
cortex and medulla. These qualities make thymic slices a highly versatile and technically accessible model
to study thymocyte development. Here we describe methods to prepare, embed, and slice thymic lobes to
study T cell development in situ.
Key words Thymic slice, Organotypic culture, Vibratome, Thymus, Agarose-embedding
1 Introduction
Thymocytes develop within the highly organized, complex, three-

dimensional microenvironment of the thymus, which many in vitro
models fail to fully recapitulate. Although two-dimensional culture
methods such as antibody cross-linking, peptide-MHC-tetramer
stimulation, and stromal cell co-culture provide important infor-
mation, they do not support efficient positive selection and differ-
entiation. Thymic slices were initially developed to examine the
potential of ex vivo cultured tissue for thymic transplantation in
different clinical settings of T cell dysfunction [1], but are now
being used to study T cell development [2–6]. Prior to the adapta-
tion of the thymic slice system, organotypic culture systems that
fully supported the processes of positive and negative selection
were fetal thymic organ cultures (FTOC) or reaggregate thymic
organ cultures (RTOC) [7, 8]. The development of the thymic
slice model provides the added advantage of maintaining the struc-
tural integrity of mature thymic tissue with well-defined cortical
and medullary microenvironments. Defined thymocyte popula-
tions can be overlaid atop the thymic slices, where they migrate to
their correct anatomical location following entry into the slice [3].
131
132 Jenny O. Ross et al.
Furthermore, the short time required for the entry of different

thymocyte populations after addition to the slice allows for analysis
of thymocyte development within hours of culture establishment.
Additionally, roughly 0.5–2 % of the cells in the thymic slices are
derived from the added thymocytes several hours after overlay. This
relatively low precursor frequency is a more physiological represen-
tation of the precursor frequency within the thymus compared to
other developmental systems that support positive selection. In
this way, development of a relatively synchronized cohort of purified
thymic subsets can be followed over time, and the robust cell
recovery is amenable to efficient analysis of T cell development by
flow cytometry. Thymocytes can also be labeled with fluorescent dyes
for direct observation by live-cell imaging to visualize cell migra-
tion, intracellular calcium changes, and cell death, for example.
The thymic slice model is a highly versatile system that is
currently being used to study positive and negative selection of
developing mouse thymocytes as well as the behavior of developing
human thymic subsets. Owing to the relative ease of thymocyte
entry, the diverse cell types that can be introduced, and the abun-
dance of tissue, this system has the potential to be further expanded
to examine other aspects of thymic development. Here, we outline
the process of generating thymic slices (including thymic harvest,
embedding lobes in agarose, and vibratome-slicing of embedded
tissue), overlaying thymocytes, and dissociating slices for flow cyto-
metric analysis.
2 Materials
2.1 Harvesting 1. Mouse (see Note 1).

a Mouse Thymus 2. 70 % ethanol spray.
for the Generation
3. Tissue pins.
of Thymic Slices
4. Paper towels.
5. Styrofoam board.
6. Small scissors (e.g. Roboz,. RS-5912) or other similar.
7. Blunt forceps (e.g. Roboz, RS-5136) or other similar.
8. Sharp forceps (e.g. Roboz, RS-5047) or other similar.
9. Micro-dissection scissors (e.g. Roboz, RS-5602) or other
similar.
10. 6-cm tissue culture dishes.
11. Phosphate-buffered saline (PBS).
2.2 Embedding 1. A thymus to be embedded [from 2.1 or human thymus fragment

and Vibratome (see Note 2)].
Sectioning a Thymic 2. 4 % Low-melting temperature agarose (e.g. NuSieve GTG
Lobe in Agarose Agarose; Lonza, 50080, or other similar) in Hank’s Balanced
T Cell Development in Thymic Slices 133
Salt Solution (HBSS): 2 g low-melting-point agarose per

50 mL HBSS (see Note 3).
3. 6-Well tissue culture plate.
4. Complete DMEM: DMEM supplemented with 10 % FBS,
2 mM Glutamine, 2 mM Penicillin-Streptomycin, and 50 μM
2-mercaptoethanol.
5. 0.4-μm pore-size organotypic cell culture inserts (BD
Biosciences, cat. no. 353090) (see Note 4).
6. 500 mL beaker.
7. Crushed ice.
8. Tissue molds (e.g. Polysciences, 18986-1) or other similar.
9. Paper towels.
10. Blunt forceps.
11. Vibratome (e.g. 1000 Plus Sectioning System) or other
similar.
12. Vibratome blades, 2 (feather blades; e.g. Leica Biosystems,
39053234, or other similar).
13. Tissue glue (e.g. 3 M Vetbond tissue adhesive, 1469SB) or
other similar.
14. 6-cm tissue culture dishes.
15. PBS.
16. Spatula, bent (spatula, rounded ends, 6.4 mm; e.g. VWR,
57950-000, or other similar) (see Note 5).
2.3 Overlay 1. Purified thymocyte population of interest at 1 × 106 cells/10 μL

of Thymocytes in complete DMEM (see Note 6).
onto Thymic Slices 2. Thymic slices on cell culture inserts (from Subheading 3.2).
3. 37 °C incubator.
4. PBS.
2.4 Dissociation 1. PBS.

of Thymic Slices 2. Microfuge tubes (2/slice).
for Flow Cytometric
3. Thymic slices (as generated in Subheadings 3.2 or 3.3) (see
Analysis
Note 7).
4. Spatula, bent.
5. Microcentrifuge tube sample pestles (e.g. Fisher, 05-559-26,
or other similar).
6. Nylon mesh filter.
3 Methods
3.1 Harvesting 1. Euthanize the mouse by your institution’s approved method.

a Mouse Thymus 2. Soak the mouse with 70 % ethanol.
for the Generation
3. Using tissue pins, fix the mouse to a Styrofoam board covered
of Thymic Slices
with a paper towel.
4. Expose the thymus. To do this, grasp the skin below the ster-
num, making a small longitudinal cut at the bottom of the rib-
cage using the small scissors, without cutting through the
peritoneum. Then, make one long vertical cut through the skin
from the sternum up to the throat followed by two smaller cuts
up to the front legs. Peel back the skin to visualize the ribcage
under the peritoneum. Grasp the sternum with the blunt forceps
and cut along the bottom of the diaphragm using the small
scissors. Cut up through the ribs toward the head on both sides
to avoid large blood vessels under the ribcage, flip the ribcage
toward the head of the mouse, and secure it with a pin.
5. Hold the heart with the blunt forceps and pull caudally. Using
the sharp forceps, carefully dissect the connective tissue to sep-
arate the thymus from the anterior wall of the ribcage. Grasp
the aorta below the thymus and above the heart with the blunt
forceps and, using the micro-dissection scissors, cut through
the attachments below and behind the thymus, cutting the
aorta at a level above the heart last. It is important not to grasp
the thymic tissue itself as tears compromise the tissue and the
thymic slice quality.
6. Submerge the thymus in PBS in a 6-cm dish on ice until ready
to proceed.
7. Place the thymus on a paper towel soaked with PBS. Use the
blunt and sharp forceps to carefully separate the lobes and
remove any remaining connective tissue. Care must be taken
not to grasp the thymus itself to prevent tearing the tissue.
8. Place the individual lobes back into the 6-cm dish while you
prepare to embed them.
3.2 Embedding 1. Prepare the agarose: Add 2 g low-melting-point agarose per

and Vibratome Slicing 50 mL HBSS in a 200 mL Erlenmeyer flask. Microwave on a
a Thymic Lobe low setting until the agarose is completely dissolved, about
in Agarose 2 min. Cover with foil and move to a 55 °C water bath until
needed.
2. Set up the tissue culture plates by adding 1.5 mL of complete
DMEM per well of a 6-well plate, and placing a cell culture
insert in each well.
3. Make an ice water bath: fill a 500 mL beaker with ice and add
water until the ice is a slushy consistency.
Fig. 1 Embedding and preparing thymic slices. (a) Using blunt forceps, grasp connective tissue and transfer
individual thymic lobes into a tissue mold filled half way with 4 % agarose solution. (b) Gently push the lobe
into position within the agarose before the agarose solution sets. (c) Once the lobe is positioned, place the
mold into an ice water bath to set. (d) Example of three lobes embedded in one tissue mold. (e) Trim the excess
agarose around the embedded lobes. Left, thymic lobe embedded horizontally. Right, example of a thymic lobe
embedded vertically. (f) Up to three blocks of embedded tissue can be glued to the vibratome stage and sliced
simultaneously. Use a bent spatula to transfer the thymic slices after each cut as they float free of the vibra-
tome blade. (g) Use a pipet tip to gently push the thymic slice onto a cell culture insert within a 6-well plate.
(h) Use a pipet to remove any liquid in the transwell and around the edges of the thymic slice
4. Fill a tissue mold halfway with 4 % agarose solution and wait

about 30 s to allow the agarose to cool slightly.
5. Grasp the thymic lobe (as isolated in Subheading 3.1) by any
remaining connective tissue and gently blot it on a dry paper
towel to remove excess liquid (see Note 8).
6. Carefully insert the lobe into the agarose using the blunt for-
ceps either vertically to maximize slice number or horizontally
to maximize slice area (see Notes 9 and 10) (Fig. 1a–c).
7. Place the mold in the ice-water bath until the agarose is set
(~5 min, agarose will be opaque and firm to the touch)
(see Note 11) (Fig. 1c).
8. While the agarose solidifies, prepare the vibratome by inserting

a vibratome blade. The following settings should be used for
sectioning thymic tissue: cutting angle 5°, maximum amplitude,
minimum speed.
9. Invert the tissue mold and gently press on its center to release
the agarose-embedded tissue.
10. Using a sharp blade (see Note 12), trim the sides of the embed-
ded thymic tissue, leaving ~2 mm of agarose around the indi-
vidual thymic lobes. It is not necessary to trim the bottom, but
make certain to leave a minimum of 0.5 cm of agarose below
the thymic lobe as the vibratome can only cut to 0.5 cm above
the stage (Fig. 1e).
11. Place the trimmed agarose-embedded tissue into a 6-cm dish
filled with PBS for transport to the vibratome.
12. Dry the agarose-embedded tissue block on a paper towel.
13. Using a pipet tip, apply a small amount of tissue glue to the
vibratome stage directly in front of the blade and place the aga-
rose embedded lobe on top (see Note 13). Make sure the stage
is clean as excess glue and other debris can float off and obstruct
the surface of the slice.
14. Align the vibratome blade with the top of the tissue.
15. Add PBS to fill the vibratome stage and submerge the agarose
blocks and blade.
16. Section thymus at the desired thickness (see Note 14).
17. Retract the blade 200 μm after each slice to prevent dragging
along the surface of the tissue.
18. Transfer slices as they are cut to the 6-well tissue culture plates
by using a pipet tip to gently slide the slice off the spatula and
onto the cell culture insert (Fig. 1f, g) (see Note 15).
3.3 Overlay 1. Remove all excess liquid from the insert (any accumulated PBS
of Thymocytes onto and/or medium) surrounding the thymic slices by pipetting
Thymic Slices carefully to prevent tearing or damaging the agarose surrounding
the tissue (see Note 16) (Fig. 1h).
2. Pipet 10 μL of labeled (either congenic or dye, etc.) thymo-
cytes of interest on top of each slice (see Note 17) carefully
without touching the tissue or the surrounding agarose
(Fig. 2d–f).
3. Incubate slices at 37 °C for 2 h to allow overlaid cells to migrate
into the thymic slices.
4. Rinse the slices with 1 mL/well PBS to wash off any cells that
have not yet migrated into the slice.
5. Continue to incubate slices at 37 °C for the desired length of
time (see Notes 18 and 19).
Fig. 2 Quality control and overlay of thymic slices. (a) Example slices from a vertically (top) versus horizontally
(bottom) embedded thymic lobe. (b) Examples of poor quality slices that will cause overlaid cells to run off. The
top slice has a tear in the agarose and the bottom slice has a chunk of agarose torn free. (c) Example of a tear
in the thymic tissue (top slice) versus a good, intact slice (bottom). After pipetting off excess liquid, inspection
of the thymic tissue itself, as well as the borders of the agarose surrounding the tissue will reveal any slices
with defects. These should be discarded and replaced with good quality slices with clean, clear edges of aga-
rose for optimal overlay. (d) Add 10–20 μL of cell suspension per slice by pipetting the volume so that it forms
a suspended drop. (e) Gently touch the drop of resuspended cells to the center of the thymic slice, without
touching the pipet tip itself to the slice. (f) The overlaid cells will remain on top of the slice
3.4 Dissociation 1. Add 150 μL of PBS to a microfuge tube (1 tube/sample).

of Thymic Slices 2. Add 1 mL of PBS to tissue culture insert containing the thymic
for Flow Cytometric slice to release the slices from contact with the culture insert,
Analysis allowing for ease of transfer.
3. Transfer thymic slice to microfuge tube using the bent spatula.
4. Dissociate slice using tissue sample pestles.

5. Add another 150 μL PBS.
6. Filter the 300 μL volume of cells through a mesh filter into a
new microfuge tube. Cells are ready for analysis and can be
stained for flow cytometry or prepared for other analyses.
4 Notes
1. Vertical embedding of the thymus of a 1–2 month old wild-

type mouse will yield approximately 10–12 400 μm thick slices
that contain both cortex and medulla per lobe. The size of the
thymus and number of slices depends on age, gender, and
genotype.
2. A segment of human thymus (cut to be roughly the same size
as a mouse thymus, ~0.5 cm wide × 1–1.5 cm high) can be
embedded in the same manner as a mouse thymus. However,
connective tissue near the capsule tends to cause the slice to
catch on the blade during slicing. Removal of the connective
tissue does not increase the chances of successful slicing and
may instead lead to tearing of the tissue and generation of
slices of poor quality. Plan on reserving several pieces of tis-
sue for sectioning, in the event that the first segment does
not cut well.
3. The 4 % agarose solution can be stored in a 55 °C water bath
for several days.
4. The tissue culture inserts from BD Biosciences work well for
thymocyte overlay. Several other brands did not maintain the
overlaid liquid on the slice.
5. A spatula with a 90° bend 1 cm from the end works well as a
tool to transfer slices (Fig. 1f).
6. Use labeled thymocytes (for example; fluorescently labeled
through transgenic ubiquitin driven GFP [9], a congenic
marker such as Ly5.1, or vital dye such as CFSE or SNARF) to
distinguish the overlaid population from endogenous thymo-
cytes and/or visualize thymocytes during imaging.
7. Thymic slices can be incubated with or without overlaid cells.
If there are no cells being overlaid, the integrity of the agarose
surrounding the thymic slice does not have to be intact, as
shown (Fig. 2b, c).
8. If the thymic lobe is wet, it may pull out of the agarose block
during slicing.
9. Multiple lobes can be embedded simultaneously side-by-side
(up to 3) (Fig. 1d) if they are inserted vertically.
10. Ensure that the thymic lobe touches the bottom of the mold
while it is being embedded in agarose to provide a 0.5 cm buf-
fer of agarose below the tissue. This will allow for the entire
tissue to be vibratome-sliced.
11. There is approximately a 30 s window after the molds are
placed into the ice water bath during which the placement of
the lobes can be adjusted within the agarose. Use the blunt
forceps to move the thymus into the position desired (Fig. 1b).
12. Vibratome blades can be reused for a couple of experiments.
Discard after they have been used for trimming the blocks.
13. Up to three tissue blocks aligned side-by-side parallel to the
vibratome blade can be sliced simultaneously. Ensure that they
are roughly equivalent in height for best results (Fig. 1f).
14. 400–500 μm thick slices work well for tissue culture and
500–1000 μm thick slices work well for imaging.
15. Place as many slices as will fit per insert without touching each
other or the insert wall. Ensure that there are no air bubbles
below the insert.
16. If the slices are not dried properly, the overlaid cells may spill off.
Any tears or holes in the agarose may also lead to loss of overlaid
cells/fluid volume (Fig. 2a–c). Finally, properly drying the slices
allows for visual inspection of the slice quality. It is a good idea
to cut extra slices and replace any as needed.
17. Overlaying 1 × 106 cells/slice for 2 h prior to washing typically
results in the cells of interest making up 0.5–2 % of the total
thymocytes within the slice. Longer incubations result in a
greater proportion of cells migrating into the tissue slice.
18. If a heterogeneous population of thymocytes is overlaid, har-
vesting an early time-point will establish a baseline of the cells
that migrated into the slices. The quality of the slices deterio-
rates significantly after just a day or two, but development
seems to proceed for several days.
19. Over time, thymocytes leak out of the slices due to the absence
of a capsule.
Acknowledgements
This work was supported by an NIH grant AI064227-07 (to

E.A. Robey), California Institute of Regenerative Medicine clinical
fellowship TG2-01164 (to J. Halkias), and postdoctoral training
grant TG2-01164 (to H.J. Melichar).
References
1. Markert ML, Watson TJ, Kaplan I et al (1997) thymocyte migration in situ. J Clin Invest
The human thymic microenvironment during 123:2131–2142
organ culture. Clin Immunol Immunopathol 6. Dzhagalov IL, Chen KG, Herzmark P et al
82:26–36 (2013) Elimination of self-reactive T cells in the
2. Bhakta NR, Oh DY, Lewis RS (2005) Calcium thymus: a timeline for negative selection. PLoS
oscillations regulate thymocyte motility during Biol. doi:10.1371/journal.pbio.1001566
positive selection in the three-dimensional thy- 7. Anderson G, Partington KM, Jenkinson EJ
mic environment. Nat Immunol 6:143–151 (1998) Differential effects of peptide diversity
3. Ehrlich LI, Oh DY, Weissman IL et al (2009) and stromal cell type in positive and negative
Differential contribution of chemotaxis and sub- selection in the thymus. J Immunol
strate restriction to segregation of immature and 161:6599–6603
mature thymocytes. Immunity 31:986–998 8. Jenkinson EJ, Anderson G, Owen JJ (1992)
4. Le Borgne M, Ladi E, Dzhagalov I et al (2009) Studies on T cell maturation on defined thymic
The impact of negative selection on thymocyte stromal cell populations in vitro. J Exp Med
migration in the medulla. Nat Immunol 176:845–853
10:823–830 9. Schaefer BC, Schaefer ML, Kappler JW et al
5. Halkias J, Melichar HJ, Taylor KT et al (2013) (2001) Observation of antigen-dependent
Opposing chemokine gradients control human CD8þ T-cell/dendritic cell interactions in vivo.
Cell Immunol 214:110–122
Chapter 12
FTOC-Based Analysis of Negative Selection

Abstract
Potentially harmful T cell precursors are removed from the conventional T cell pool by negative selection.
This process can involve the induction of apoptosis, anergy, receptor editing or deviation into a regulatory
T cell lineage. As such this process is essential for the health of an organism through its contribution to
central and peripheral tolerance. While a great deal is known about the process, the precise mechanisms
that regulate negative selection are not clear. Furthermore, the signals that distinguish the different forms
of negative selection are not fully understood. Numerous models exist with the potential to address these
questions in vitro and in vivo. This chapter describes methods of fetal thymic organ culture designed to
analyze the signals that determine these unique cell fates.
Key words FTOC, Negative selection, T cell development, Thymocyte differentiation
1 Introduction
A healthy individual can mount an immune response to exogenous

pathogens while avoiding an autoimmune attack on normal tis-
sues. The ability to distinguish between self and non-self is called
immunological tolerance. T cell central and peripheral tolerance
begins in the thymus with selection [1–3]. T cell selection and
development depends on the ability to distinguish between thymo-
cytes that bear useful TCR from those that express useless or
potentially harmful ones. Thus, the generation of a healthy immune
system depends on the removal or reprogramming of these useless
and harmful cells. Useless thymocytes, those that are unable to
respond to self-peptide MHC, die by neglect. Those that have a
strong response to endogenous self undergo negative selection.
The removal of these potentially auto-reactive T cells is achieved
by the induction of apoptosis, anergy, receptor editing or repro-
gramming into a regulatory T cell lineage [4]. Despite the vast
amount of data that have been generated on this subject, several
questions remain unanswered.
141
142 Cody A. Cunningham et al.
Studies addressing T cell negative selection are inherently

difficult in that a thymocyte’s favorite activity is dying. The thymus
has long been recognized for containing a large number of dying
T cells. Thus, it is difficult to distinguish death by neglect (90 % of
thymocytes) and death by agonist-induced apoptosis (5 % of thy-
mocytes). The need for intact thymic architecture also means there
are no in vitro models that allow one to follow negative selection
from beginning to end. Selection is a multistep process with many
checkpoints. Thus, research in this area is also complicated by mat-
uration-based changes in a thymocytes response to stimulation as
it passes through selection [5–7]. Furthermore, as a T cell matures
it moves through the various thymic compartments each with a
unique environment and population of phenotypically different
thymic epithelium and antigen-presenting cells [8]. Therefore,
work performed on whole thymi from a selecting background may
miss subtle yet fundamentally important information. Furthermore,
the differences between the thymic epithelium of the cortex and
the medulla suggest that there may be multiple ways to induce
apoptosis. Unfortunately, traditional knock out and transgenic
strategies are insufficient as many of the signals that determine
selection fates are also essential for earlier stages of T cell develop-
ment [1–3]. Several in vivo models of selection have provided
invaluable information as to the conditions that lead to negative
selection. However, considering the complications discussed
above, along with the asynchronous nature of in vivo selection,
some of the subtler yet critically important details of the process of
negative selection remain unknown.
Fetal thymic organ culture (FTOC) remains one of the most
informative and manipulatable methods to examine the signals that
distinguish negative selection from positive selection. In addition,
it has the potential to identify the signals that lead to the different
outcomes of agonist selection. As such this is critical for our ability
to understand thymic selection’s contribution to peripheral
tolerance.
The following are examples of manipulations that are easily
used to address the signals of negative selection in FTOC:
– Easily manipulate the selecting ligand strength and
concentration.
– Assess the contribution of “third” signals through the addition
of excess cytokines or antibodies that block or stimulate these
pathways.
– Take advantage of the proliferative burst in DN3/4 thymo-
cytes to retrovirally transduce T cells with various constructs
and test their affect on selection.
In vitro Analysis of Thymocyte Signaling 143
– Use transgenic, tissue-specific and inducible models of antigen

expression to test the unique contribution of the different sub-
sets of thymic epithelium.
– Use transgenic, knockout, and conditional expression systems.
Several considerations need to be kept in mind when analyzing
negative selection. It is difficult to analyze signaling in dead cells.
Dead cells disappear rather quickly, even in FTOC [9, 10]. Thus,
while the loss of thymocytes provides information that is essential
to understand the conditions that lead to negative selection, we
need to catch them before they die to be able to say anything about
where they go and how they got there. Along these lines, it is
important to distinguish death by neglect and death by negative
selection. Thus, we always include both positive selection and non-
selection controls in our analysis of negative selection (We have
had numerous discussions in our lab which one should be called
the positive and the negative controls and finally settled on the
term selection controls as a result).
We use several standard techniques to assess signals and analyze
selection in our system that are further described in Chapter 15.
– For early time points (usually <48 h) immunoblot techniques
can be easily used to determine the expression levels and acti-
vation states of proteins. For these assays we stimulate, pool
and disrupt thymi, lyse, subject to SDS-PAGE, transfer and
immunoblot (shown elsewhere in this manual).
– Flow cytometry can be used for nearly all conditions as long as
there is an antibody that works by intracellular staining. It is
especially useful for late time points (decreasing number of
cells) and assays on small/rare populations of cells that can be
distinguished by a specific phenotypic marker. For these assays
we stimulate, stain surface markers and perform intracellular
staining to distinguish the population of interest. In general,
we use standard kits (BD Phosflow or Cytofix/Cytoperm kits)
and follow their suggested protocols.
– Confocal microscopy provides activation and location informa-
tion that has added to our understanding of the diversity in the
regulation of selection signals [11, 12].
– We assess interactions between different signaling molecules
with IP followed by flow cytometry (IP-FCM). This method
offers the benefit of being able to precipitate low abundance pro-
teins to generate data with a high signal to noise ratio [13–15].
Additionally, this protocol offers the benefit of being able to use
significantly fewer cells for analysis (~1 × 105 for IP-FCM vs.
100 × 106 for IP-Western). For these assays we select our cells
of interest and perform assays whose methods are described
elsewhere in this methods issue.
2 Materials
1. Culture Dishes: We generally use 6- or 12-well plates with 2 ml

of media in each, but small petri dishes also work.
2. Sterile filters and gelfoam:
(a) Mixed cellulose membrane with a 0.45 μm.
(b) Surgical Gel-foam.
(c) ALTERNATIVE: 13 mm Supor-450 membrane filter
(e.g. #60170 from PALL, Life Sciences) (see Note 1).
3. Media:
(a) RPMI 1640 supplemented with 10 % v/v heat-inactivated
FCS, 2 mM glutamine, 10 mM HEPES, 0.5 mg/ml folic
acid, and 0.2 mg/ml glucose, 100 U/ml Penicillin,
100 μg/ml Streptomycin (see Note 2).
(b) Alternatively, FTOC can be done in serum-free media.
For this use RPMI 1640, 100 U/ml Penicillin, 100 μg/
ml Streptomycin, 0.292 mg/ml Glutamine, 1 mM Sodium
pyruvate, 1 % Nutridoma SP (Roche) (see Note 3).
4. Two petri dishes of ice-cold PBS.
5. Two pairs of scissors, one fine.
6. Three pairs of forceps, one regular, one fine curved, one
ultrafine.
7. Sterile 4 × 4 gauze.
8. Several 25 g needles, both 1/2″ and 1–1/2″.
9. FTOC chamber (see Note 4).
10. Dissecting (Stereo) Microscope with a good light source
(see Note 5).
3 Methods
These techniques are adapted from work published by the Hogquist

and Anderson labs [16–19] and from the vast experience and guid-
ance of Barbara Hausmann while I was in the lab of Dr. Ed Palmer.
In general, while there are several selecting and non-selecting
mouse models that can be used for FTOC, we use TCR transgenic,
Rag−/−, β2m−/− mice to perform our assays. They provide a non-
selecting environment and a synchronous population of thymo-
cytes. To induce selection of CD8 T cells we simply add our peptide
of interest and β2m in the FTOC media, check the selection out-
come controls (see Note 6) and analyze our desired experimental
variables [11, 17].
3.1 Timed Matings The days of the week represent an example of the scheduling procedure
used in our lab.
1. Tuesday: Prepare time mating cages by placing a cage divider
with feeders and water for both sides in each cage.
2. Wednesday, morning: Place two to three females on one side
of the divider and one male on the opposite side of the divider
(see Note 7).
3. Friday, late afternoon/evening: Remove cage divider.
4. Saturday, morning (before 10 a.m.): Remove male.
OPTIONAL: check for and record mucous plugs (see Note 8).
5. Saturday (or the day of the plug) is considered day 0.5
(although technically it could be anywhere from 0 to 0.5 days)
(see Note 9).
6. Two weeks from the following Monday (Day 15.5), harvest
fetuses (see below).
3.2 Preparing 1. Boil cellulose filters three times in dH2O. The final time cover
the Culture Dishes with foil and, when cool, transfer into the hood. Place the fil-
ters in medium just before use.
2. Prepare medium.
3. Remove a strip of gel-foam and cut it into three squares (KEEP
STERILE), placing each square in one well. Using one straight
and one curved forceps soak the gel-foam and tease the air
bubbles out.
4. Place one filter (from step 1) on each gel-foam square.
5. Place the plate in a FTOC chamber (see Note 4). Pre-warm in
the incubator.
3.3 Removing Fetal 1. Pregnant mothers are sacrificed according to institutional

Thymic Lobes guidelines.
2. Peel the skin (hide) back from abdomen of pregnant mouse.
3. Make an incision through the abdominal muscle across the
mouse just above the bladder. Continue up each side of the
abdomen to expose the entire abdominal cavity.
4. Excise both horns of the uterine sac and place in a petri dish
with 10 ml of sterile cold PBS.
From this point on everything should be performed in a
horizontal laminar flow hood (or other easily accessibly sterile
hood). Never work with infected (viral or bacterial) tissues or
cells in this hood.
5. Use a sterile scissors to remove the individual embryos from
the amniotic sacs.
6. Remove the placenta and amniotic sac tissue from each embryo
and place fetuses in fresh cold PBS.
7. Wipe down dissecting scope and light source with 70 % ethanol.

The actual removal of thymic lobes is relatively easily from
a day 15.5 embryo. While, there are several methods by which
they can be removed, two of the more common methods are
described below.
8. “Pin” method.
(a) Place petri dish with sterile gauze on the stage of
microscope.
(b) Place the fetus on sterile gauze in petri dish.
(c) Decapitate fetus with forceps or pith the fetus with a 25 g
needle.
(d) Pin the front limbs of the embryo and cut open the chest
to expose the thymic lobes sitting on/above the heart on
either side of the trachea.
9. Thoracic stem method: There is an excellent video of this
method by Jenkinson et al J. Vis. Exp. (2008) at http://www.
jove.com/video/906/preparation-of-2-dguo-treated-thymus-
organ-cultures.
(a) Decapitate the embryo with forceps.
(b) Open the anterior chest cavity by inserting forceps (closed
then open) into the thoracic cavity through the neck.
(c) Use your forceps to reach into the torso and pull out the
entire thoracic tree; heart, lungs and trachea. The thymus
usually comes out with this and can be found on either
side of the trachea, just above the heart (see Note 10).
10. Pluck the lobes out using the ultrafine forceps and place on
filters, taking care not to disrupt the thymic capsule.
11. You can place up to ten lobes per filter disc, place in FTOC
chamber (see Note 4) and culture in 5 % CO2 at 37 °C.
3.4 Feeding We use two methods of feeding. Both are described here.
1. Replace the media in the cultures every day.
(a) To do this, carefully suction off the media with a sterile
Pasteur pipette.
(b) Add 2 ml fresh warm media, carefully basting thymic lobes.
2. Replace the media in the cultures every 2–5 days.
(a) Carefully baste thymic lobes daily.
(b) To replace media, carefully suction off the media with a
sterile Pasteur pipette.
(c) Add 2 ml fresh warm media, carefully basting thymic lobes.
3. Seven days of culture is sufficient for good thymic develop-
ment from an E15.5 lobe. However, they remain viable up to
14 days.
3.5 Analysis To disrupt the lobes:

1. Place on a small nylon mesh.
2. Add media and crush the lobe(s) against the mesh with the
plunger of a 3 cc syringe.
3. On average we recover 2–4 × 105 T cells per individually pro-
cessed lobe. Yields go up to 5–7 × 105 cells/lobe when multi-
ple lobes are pooled during harvesting.
4. Analyze as described in Chapter 15.
4 Notes
1. This brand of filters float nicely in the media (shiny side up)
such that expensive (and sometimes difficult to obtain) surgi-
cal gel-foam is not needed.
2. We often test our FCS in FTOC in advance to minimize
serum-based artifacts of selection.
3. If you wish to add, antibodies, peptides, etc., add them to
media prior to plating out thymi.
4. It is critically important to avoid drying of the thymic lobe. To
do this we use a covered Pyrex baking dish (available from the
local grocery store). We have small holes drilled in the cover
(>5 mm). We place cover of a 6- or 12-well plate in the bot-
tom of the baking dish as a pedestal, and fill the bottom of the
dish with autoclaved dH2O. The culture dish containing the
FTOC are placed on top of the pedestal and placed in the
incubator. See also Jenkinson et al J. Vis. Exp. (2008) for an
alternative apparatus.
5. With practice day 15.5 embryos are easily removed without
the aid of a microscope.
6. It is critically important to have a selection outcome control
for each test condition. For CD8 T cells, we always save a lobe
or partial lobe (minimum of 1 × 104 cells) and stain for TCR,
CD4, CD8α, and CD8β. This is then analyzed by flow cytom-
etry for the following:
(a) The generation of CD8β SP thymocytes is the test of posi-
tive selection.
(b) The lack of the generation of CD8α and CD8β SP thymo-
cytes and the maintenance of a DP population indicates
no selection.
(c) The lack of the generation of CD8β SP thymocytes with
the concomitant loss of DP indicates negative selection.
(d) Finally, the loss of DP with a concomitant lack of CD8β SP

and CD8α SP thymocytes indicates negative selection by
apoptosis and the generation of only CD8α-positive cells
indicates lineage deviation (reprogramming) [11, 20].
7. This procedure helps to induce estrus and increases the effi-
ciency/probability of pregnancies.
8. We generally just wait for the females to show signs of
pregnancy.
9. Some consider this day 1.
10. Thymic lobes sometimes remain nestled in the thoracic cavity
and are easily removed.
Acknowledgements
The authors thank the members of the Daniels and Teixeiro labs
for assistance in compiling these protocols. Funding provided by
grants from the Missouri Mission Enhancement Fund and
University of Missouri Research Board is greatly acknowledged.
References
1. Palmer E (2003) Negative selection–clearing 8. Petrie HT, Zuniga-Pflucker JC (2007) Zoned
out the bad apples from the T-cell repertoire. out: functional mapping of stromal signaling
Nat Rev Immunol 3(5):383–391 microenvironments in the thymus. Annu Rev
2. Starr TK, Jameson SC, Hogquist KA (2003) Immunol 25:649–679
Positive and negative selection of T cells. Annu 9. Borgne ML, Ladi E, Dzhagalov I, Herzmark
Rev Immunol 21:139–176 P, Liao YF, Chakraborty AK, Robey EA (2009)
3. Stritesky GL, Jameson SC, Hogquist KA The impact of negative selection on thymocyte
(2012) Selection of self-reactive T cells in the migration in the medulla. Nat Immunol
thymus. Annu Rev Immunol 30:95–114 10(8):823–830
4. Baldwin TA, Hogquist KA, Jameson SC 10. Dzhagalov IL, Chen KG, Herzmark P, Robey
(2004) The fourth way? Harnessing aggressive EA (2013) Elimination of self-reactive T cells
tendencies in the thymus. J Immunol 173(11): in the thymus: a timeline for negative selec-
6515–6520 tion. PLoS Biol 11(5), e1001566
5. Daniels MA, Devine L, Miller JD, Moser JM, 11. Daniels MA, Teixeiro E, Gill J, Hausmann B,
Lukacher AE, Altman JD, Kavathas P, Roubaty D, Holmberg K, Werlen G, Holländer
Hogquist KA, Jameson SC (2001) CD8 bind- GA, Gascoigne NR, Palmer E (2006) Thymic
ing to MHC class I molecules is influenced by selection threshold defined by compartmental-
T cell maturation and glycosylation. Immunity ization of Ras/MAPK signalling. Nature
15(6):1051–1061 444(7120):724–729
6. Davey GM, Schober SL, Endrizzi BT, Dutcher 12. Daniels MA, Teixeiro E (2010) The persis-
AK, Jameson SC, Hogquist KA (1998) tence of T cell memory. Cell Mol Life Sci
Preselection thymocytes are more sensitive to 67(17):2863–2878
T cell receptor stimulation than mature T cells. 13. Cunningham CA, Knudson KM, Peng BJ,
J Exp Med 188(10):1867–1874 Teixeiro E, Daniels MA (2013) The POSH/
7. Lucas B, Stefanová I, Yasutomo K, Dautigny JIP-1 scaffold network regulates TCR-mediated
N, Germain RN (1999) Divergent changes in JNK1 signals and effector function in CD8 T
the sensitivity of maturing T cells to structurally cells. Eur J Immunol 43(12):3361–3371
related ligands underlies formation of a useful 14. Gil D, Schrum A, Alarcon B, Palmer E (2005)
T cell repertoire. Immunity 10(3):367–376 T cell receptor engagement by peptide – MHC
ligands induces a conformational change in the thymocyte dulling assay. Methods Mol Biol
CD3 complex of thymocytes. J Exp Med 156:219–232
201(4):517–522 18. Hogquist KA, Jameson SC, Bevan MJ (1994)
15. Schrum AG, Gil D, Dopfer EP, Wiest DL, The ligand for positive selection of T lympho-
Turka LA, Schamel WW, Palmer E (2007) cytes in the thymus. Curr Opin Immunol
High-sensitivity detection and quantitative 6(2):273–278
analysis of native protein-protein interactions 19. Travers H, Anderson G, Gentle D, Jenkinson
and multiprotein complexes by flow cytome- E, Girdlestone J (2001) Protocols for high
try. Sci STKE 2007(389):pl2 efficiency, stage-specific retroviral transduc-
16. Anderson G, Jenkinson EJ, Moore NC, Owen tion of murine fetal thymocytes and thymic
JJ (1993) MHC class II-positive epithelium epithelial cells. J Immunol Methods 253(1–2):
and mesenchyme cells are both required for 209–222
T-cell development in the thymus. Nature 20. Lambolez F, Kronenberg M, Cheroutre H
362(6415):70–73 (2007) Thymic differentiation of TCR alpha
17. Hogquist KA (2001) Assays of thymic selec- beta(+) CD8 alpha alpha(+) IELs. Immunol
tion. Fetal thymus organ culture and in vitro Rev 215:178–188
Chapter 13
Reconstituted Thymus Organ Culture

Zimu Deng, Haifeng Liu, Jinxiu Rui, and Xiaolong Liu
Abstract
Reconstituted thymus organ culture is based on fetal thymus organ culture (FTOC). Purified thymocyte
populations, from genetically modified mice or even from other species, are cultured in vitro with thymic
lobes depleted of their endogenous thymocytes (by 2′-deoxyguanosine treatment) to form a new thymus.
This potent and timesaving method is distinct from FTOC, which assesses development of unmodified
thymic lobes, and reaggregate thymic organ culture, in which epithelial cells are separately purified before
being aggregated with thymocytes.
Key words Thymus, Thymocyte, Fetal thymus organ culture, 2′-Deoxyguanosine
1 Introduction
After hematopoietic progenitor cells migrate into thymus, they

interact with stromal cells and undergo a series of differentiation
steps, including V (D) J rearrangement and thymic selection. Fetal
thymus organ culture (FTOC) established by Owen [1] and
Mandel [2] offers a powerful technique to study both T-precursors
and thymic microenvironment during early T lymphocyte develop-
ment. More recently, reaggregate thymus organ culture (RTOC)
[3–6], which allows reconstruction of thymic “aggregates” from
thymocytes and stromal cells of distinct origin (and typically of
distinct genotype), contributed to elucidate molecular and cellular
mechanisms of T cell development [7, 8].
In this chapter, we describe a related method, reconstituted
thymus organ culture [4, 9]. In this procedure, fetal thymus lobes
are treated with 2′-deoxyguanosine in vitro for 5–7 days to deplete
their endogenous T cell progenitors. The depleted lobes are then
offered for colonization by and serve as thymic environment for
purified thymocytes (Fig. 1). This reconstitution method is useful
to study thymocyte development, and especially the development
of retrovirally transduced thymocytes; consequently, a succinct
151
152 Zimu Deng et al.
Fig. 1 Schematic overview of reconstituted thymus organ culture system
protocol for retroviral transduction is provided in Subheading 3.6.

To study stromal cells, the more sophisticated reaggregated thymus
organ culture is preferable.
2 Materials
2.1 Reagents All reagents must be purchased sterile, prepared with sterile
reagents or be sterile-filtered.
1. 1× PBS (Ca2+/Mg2+-free).
2. RPMI-1640 medium: supplemented with 2 mM L-Glutamine,
100 U/ml Penicillin, 0.1 mg/ml Streptomycin, 1× Non-
Essential Amino Acids and 10 mM Hepes-pH 7.4. Add
2-Mercaptoethanol in the medium to a concentration of
0.05 mM before use. Referred to as RPMI medium through-
out. The concentration of FBS varies from 10 to 20 % as indi-
cated in each step.
3. 2′-Deoxyguanosine (dGuo) (e.g. Sigma-Aldrich, D7145) 10×
stock solution: 13.5 mM in 1× PBS (dissolve at 37 °C). Keep
the stock solution at −20 °C.
4. IL7 (e.g. Peprotech, 96-217-17-2): 105 U/ml 1000× stock in
1× PBS, store at −20 °C.
Reconstituted Thymus Organ Culture 153
5. Polybrene (e.g. Sigma-Aldrich, 107689): 10 mg/ml 500×

stock in 1× PBS, store at −20 °C.
6. 75 % Ethanol.
7. DMEM, supplemented with 2 mM L-Glutamine, 100 U/ml
Penicillin, 0.1 mg/ml Streptomycin.
2.2 Equipment 1. Ophthalmic forceps, 10 cm length.

2. Tying forceps, angled platform, 10.5 cm length.
3. Thin forceps, 0.025 × 0.05 mm tip, 11 cm length (e.g. World
Precision Instruments, 50985).
4. Scissors, ophthalmic.
5. Stereo-dissecting Microscope (Olympus, SXZ7).
6. Gelatin sponges, 1 cm2.
7. 1-ml Syringes.
8. Filters, 0.45-μm pore size (e.g. Millipore, HAWG01300).
9. Nylon mesh with 40-μm pores.
10. Petri dishes, 90-, 60- and 30-mm diameter.
11. 72-Well Conical Terasaki Plates (e.g. Electron Microscopy
Sciences, 70439-10).
12. Humidified Box. Fill in a plastic rectangular box (e.g.
Lock&Lock, HPL817) with 20 ml sterilized H2O and place a
Petri dish lid as a support. Holes with 1 cm diameter were
punched on the plastic lid for air diffusion.
13. Carbon dioxide tissue culture incubator, 37 °C, 5 % CO2.
14. Centrifuge.
3 Methods
3.1 Preparation We usually put two female mice and one male in one cage at day
of Timed Pregnancies end (5:00 p.m.). Every female is examined for the presence of a
vaginal plug at 8 am every subsequent morning. The day a plug
detected is counted as Day 1 of gestation. The pregnant mice will
be kept in a new cage till use. See Notes 1–4.
3.2 Isolation 1. Anesthetize a timed pregnant female (gestational age E15-

of Fetuses from E16) mice with CO2, then euthanize by cervical dislocation.
Pregnant Mice See Note 1.
2. Wipe the body with 75 % ethanol and cut open the abdominal
wall to reveal the uterus. Use sterilized scissors and ophthal-
mic forceps to remove the uterus and place it into a 90-mm
Petri dish containing 10 ml 1× PBS. See Note 5.
Fig. 2 Fetal thymus lobes before and after treatment with 2′-deoxyguanosine. (a) Thymus lobes from E15.5
feta mice. (b) An isolated thymus lobe. (c) A lobe cultured with dGuo for 6 days
3. Cut open the uterus and collect the embryos. Swirl the Petri
dish to wash off the blood from the embryos and transfer
them into a 60-mm Petri dish containing 4 ml PBS under the
stereo-dissecting microscope.
4. Hold one embryo in a face-up position with thin forceps on
the head. Using the other thin forceps, pull each foreleg in a
head-to-toe direction, until the chest skin comes off. Two thy-
mus lobes (below the first pair of ribs and above the heart)
should be seen. If not, use thin forceps to snip off the first pair
of ribs under the armpit and open the chest to reveal the lobes
completely.
5. Carefully remove the lobes together using thin forceps and
transfer to a new 60-mm Petri dish containing 4 ml RPMI-
1640 medium with 10 % FBS. See Note 6.
6. Gently swirl the dish to wash off any blood. Use thin forceps
to tear off any connective tissue between the two lobes and
separate them (Fig. 2).
3.3 Elimination 1. Add 0.5 ml of 1 RPMI-1640 medium supplemented with

of Endogenous 10 % FBS and 1.35 mM dGuo to each well of a 24-well plate
Thymocytes [10]. Immerse one piece of sponge (1 cm2) in each well and
wait until the medium soaks the sponge. Place one filter above
each sponge. Make sure no air bubble remains between the
interfaces.
2. Scoop up one lobe each time with tying forceps and transfer it
onto the filter. In a typical experiment, put no more than four
lobes into each well.
3. Put the plate into a humidified box and keep in tissue culture
incubator. See Notes 7 and 8.
4. After 5–7 days, detach the filters from sponges and transfer the
filters into a 10-cm dish containing 10 ml cold 10 % FBS-
supplemented RPMI-1640. See Note 9.
5. Keep the plate on ice and carefully detach the lobes from filters
under a dissecting microscope (Fig. 2). Scoop up the lobes and
transfer them into a new 10-cm dish containing 10 ml fresh
10 % FBS-supplemented RPMI-1640. Swirl the plate and put it
in the incubator for 20 min to diffuse away dGuo. Change the
medium every 20 min, three times. The lobes are ready for the
next step Hanging-Drop reconstitution with T cell precursors.
3.4 Preparation of T 1. Isolate E14-E15 fetal thymus as described in Subheading 3.2.

Cell Progenitors from 2. Drop 100 μl RPMI-1640 medium supplemented with 20 %
Fetal Thymus FBS on the reverse side of the lid of a 30-mm dish [11].
3. Transfer required number of fetal lobes into the drop (usually
put 10 lobes per drop), cover the drop with a piece of nylon
mesh (1 cm2).
4. Bend the needle of a 1-ml syringe to 90° angle. Use ophthal-
mic forceps to hold one side of the mesh to keep it from slip-
ping, and gently press down the mesh with the bent needle.
Gently scrape the mesh back and forth to squeeze the thymo-
cytes out of the lobes.
5. Pipette the thymocyte suspension into a 1.5-ml microfuge
tube containing 1 ml RPMI-1640 medium supplemented
with 20 % FBS and count cells.
6. Fetal thymocytes are ready for use in Subheading 3.5, or can
be purified by cell sorting or retrovirally transduced (see
optional Subheading 3.6).
3.5 Hanging-Drop 1. Centrifuge thymocyte suspension at 170 × g for 7 min at 4 °C.

Culture 2. Discard the supernatant and resuspend the cells with RPMI-
1640 supplemented with 20 % FBS and 100 U/ml IL-7 to a
concentration of (5–100) × 103 cells/ml.
3. Add 20 μl of the thymocyte suspension into each well of
Terasaki plate, and use tying forceps to place one dGuo-treated
lobe (prepared as described in Subheading 3.3) to each well.
4. Cap the plate with its lid and invert. Gently flap the plate to
make sure every lobe is at the bottom of the drop. Avoid any
bubbles in the drops. Place the plate in a humidified box and
incubate in tissue culture incubator.
5. 24 h after reconstitution, set up filter cultures by adding 0.5 ml
of RPMI-1640 supplemented with 20 % FBS and 100 U/ml
IL-7 to wells of a 24-well plate. Immerse one piece of sponge
(1 cm2) in each well and wait until the medium soaks the sponge.
Place one filter above each sponge. Make sure no air bubble
remains between the interfaces.
6. Transfer the lobes to a 10-cm dish containing 10 ml fresh
RPMI-1640 supplemented with 20 % FBS and gently swirl the
plate to wash off cells attaching to the surface of the lobes.
Fig. 3 Viability and GFP expression of progenitor T cells 24 h after pMCS-IRES-

GFP retroviral infection
7. Transfer the lobes onto the filters of the culture well set up in
step 5 above.
8. Incubate the plate in the humidified box in tissue culture incu-
bator for 7–10 days. The introduced thymocytes then can be
assessed by FACS analysis (Fig. 3).
3.6 Optional Step: 1. Seed 1 × 106 Plat E cells [12] in a 60-mm Petri dish containing
Retroviral Transfection 3.5 ml DMEM supplemented with 10 % FBS. Place for
of Thymocytes 18–24 h in tissue culture incubator. See Notes 10 and 11.
2. Transfect 20 μg of retroviral plasmid DNA (e.g. pMCS-IRES-
GFP) using adhesion-assisted lipofection or calcium phos-
phate precipitation method.
3. Seven hours after transfection, remove the supernatants and
add 2 ml fresh DMEM supplemented with 10 % FBS. Collect
culture (retroviral) supernatants 48–72 h after the transfection
and use for infection [13]. Replace with 2 ml fresh DMEM
supplemented with 10 % FBS. See Note 12.
4. Pre-warm up the centrifuge to 32 °C.
5. Filter 500 μl retroviral supernatants with a 0.2-μm filter and
add it in a well of a 24-well plate. Dilute polybrene stock 1:500
into the retroviral supernatants to a final concentration of
20 μg/ml. Add 100–10,000 thymocytes (prepared in
Subheading 3.4) into each well and seal the plate with parafilm.
Spin the plate at 1000 × g in pre-warmed centrifuge for 90 min.
6. After transduction, take an aliquot of thymocytes, culture
overnight and assess cell viability and transduction efficiency
by flow cytometry (Fig. 4).
7. Use the rest of the cells in hanging-drop culture as described
in Subheading 3.5.
Fig. 4 A typical result of RTOC. The pMCS-IRES-GFP retrovirus-infected thymocytes were reconstituted with
dGuo-treated lobes and cultured for 8 days. Cells were isolated and then stained with anti-CD4 and anti-CD8
antibodies
4 Notes
1. Ensure that animal procedures are performed according to

appropriate guidelines and approved by relevant animal care
authorities.
2. To improve frequency of pregnancies, use 8–12 weeks old
females and proven breeder males.
3. Not all females in which a plug is detected are pregnant.
Fetuses can be palpated after gestation day 13.
4. We usually get 8–12 E15 fetuses from one pregnant mouse
(C57BL/6 background). About 10 E15 thymuses provide
1 × 106 T cell progenitors.
5. All procedures must be carried out sterile, if needed in a tissue
culture cabinet. Wipe all instruments with 70 % alcohol every
time before and after use. Make sure to remove any remaining
alcohol before use.
6. Most laboratories use a mouth-controlled glass pipette trans-
ferring the lobes, we use tying forceps handling them. Both
approaches require practice. To keep the thymus lobes inte-
grated, avoid damaging or pinching them with forceps tips of
any forceps. Use the syphonic effect between the tips of for-
ceps to lift the lobes out of liquid.
7. Quality gelatin sponges are required to prevent sponges from
sinking into the medium after 5 or 6 days.
8. During the incubation, check the plate every 2 days to make
sure the filter/sponges are intact and firm. After 5–7 days of
culture, the endogenous thymocytes are eliminated by dGuo.
9. After treatment with dGuo, the lobes collapse (Fig. 2) and
become thinner. Keeping the medium cold makes the detach-
ing process much easier.
10. Ensure that work with recombinant DNA material is per-

formed according to applicable institutional guidelines and
regulations.
11. Plat E cells [12] are a high-efficiency packaging line for retro-
viral transduction. The retroviral gene transfection efficiency
depends on the status of Plat E. The higher GFP expressed in
Plat E, the higher the viral titer of the supernatant.
12. 24 h after the transfection, the high-FSC/low-SSC population
is a good estimate of viable cells (Fig. 4).
Acknowledgements
We thank Graham Anderson (University of Birmingham) for advice

on RTOC.
References
1. Owen JJ, Ritter MA (1969) Tissue interaction fetal liver progenitor cells. J Immunol 156:
in the development of thymus lymphocytes. 3233–3242
J Exp Med 129:431–442 8. Rui J, Liu H, Zhu X, Cui Y, Liu X (2012)
2. Mandel T, Russell PJ (1971) Differentation of Epigenetic silencing of CD8 genes by ThPOK-
foetal mouse thymus. Ultrastructure of organ mediated deacetylation during CD4 T cell dif-
cultures and of subcapsular grafts. Immunology ferentiation. J Immunol 189:1380–1390
21:659–674 9. Anderson G, Jenkinson EJ (2007) Fetal thy-
3. Jenkinson EJ, Van Ewijk W, Owen JJ (1981) mus organ culture. CSH Protoc 2007:pdb
Major histocompatibility complex antigen prot4808
expression on the epithelium of the developing 10. Jenkinson EJ, Franchi LL, Kingston R, Owen
thymus in normal and nude mice. J Exp Med JJ (1982) Effect of deoxyguanosine on lym-
153:280–292 phopoiesis in the developing thymus rudiment
4. Takahama Y (2000) Differentiation of mouse in vitro: application in the production of chi-
thymocytes in fetal thymus organ culture. meric thymus rudiments. Eur J Immunol 12:
Methods Mol Biol 134:37–46 583–587
5. Jenkinson EJ, Anderson G, Owen JJ (1992) 11. Ueno T, Liu C, Nitta T, Takahama Y (2005)
Studies on T cell maturation on defined thy- Development of T-lymphocytes in mouse fetal
mic stromal cell populations in vitro. J Exp thymus organ culture. Methods Mol Biol
Med 176:845–853 290:117–133
6. Watanabe Y, Katsura Y (1993) Development 12. Morita S, Kojima T, Kitamura T (2000) Plat-E:
of T cell receptor alpha beta-bearing T cells in an efficient and stable system for transient pack-
the submersion organ culture of murine fetal aging of retroviruses. Gene Ther 7:1063–1066
thymus at high oxygen concentration. Eur J 13. Travers H, Anderson G, Gentle D, Jenkinson
Immunol 23:200–205 E, Girdlestone J (2001) Protocols for high
7. Tsuda S, Rieke S, Hashimoto Y, Nakauchi H, efficiency, stage-specific retroviral transduction
Takahama Y (1996) Il-7 supports D-J but of murine fetal thymocytes and thymic epithe-
not V-DJ rearrangement of TCR-beta gene in lial cells. J Immunol Methods 253:209–222
Chapter 14
Induction of T Cell Development In Vitro by Delta-Like

(Dll)-Expressing Stromal Cells
Mahmood Mohtashami, Payam Zarin, and Juan Carlos Zúñiga-Pflücker
Abstract
Recreating the thymic microenvironment in vitro poses a great challenge to immunologists. Until recently,
the only approach was to utilize the thymic tissue in its three-dimensional form and to transfer the hema-
topoietic progenitors into this tissue to generate de novo T cells. With the advent of OP9-DL cells (bone
marrow-derived cells that are transduced to express Notch ligand, Delta-like), hematopoietic stem cells
(HSC) could be induced to differentiate into T cells in culture for the first time outside of the thymic tissue
on a monolayer. We, as well as others, asked whether the ability to support T cell development in vitro in
a monolayer is unique to BM-derived OP9 cells, and showed that provision of Delta-like expression to
thymic epithelial cells and fibroblasts also allowed for T cell development. This provides the opportunity
to design an autologous coculture system where the supportive stromal and the hematopoietic compo-
nents are both derived from the same individual, which has obvious clinical implications. In this chapter,
we describe methods for establishing a primary murine dermal fibroblast cell population that is transduced
to express Delta-like 4, and describe the conditions for its coculture with HSCs to support T cell lineage
initiation and expansion, while comparing it to the now classic OP9-DL coculture.
Key words T cell development, Lymphopoiesis, Notch, Delta-like, Fibroblasts, Hematopoietic stem
cells, Stromal cells, IL-7, Stem cell factor
1 Introduction
Ex vivo fetal thymic organ cultures provided the authentic thymic

environment in order to generate T cells through the transfer of
hematopoietic stem cells (HSC) [1, 2]. However, the expensive
and cumbersome nature of this technique propelled immunolo-
gists to seek new methods of generating T cells in culture. It has
been well established that the generation of T cells from hemato-
poietic stem cells (HSC) is dependent on Notch signaling [3].
Based on this insight, our lab established an in vitro coculture
system in which a bone marrow-derived stromal cell line, OP9-DL1,
ectopically expressing Delta-like 1 (Dll1) could effectively and
efficiently support the initiation and expansion of T lineage cells
159
160 M. Mohtashami et al.
from HSCs [4]. The OP9-DL1 cells could mimic the thymus in its
ability to support early T cell development as well as the generation
of mature CD8 single positive (SP) cells [5, 6]. The detailed meth-
ods for generating T cells from mouse and human HSCs using
OP9-DL cells have been published earlier [7]. We also demon-
strated that rather than Dll1, Dll4, which is the Notch ligand phys-
iologically expressed in the thymus, is more efficient at steering
HSCs toward T cell lineage when ectopically expressed at physio-
logical levels on OP9 cells (OP9-DL4) [8].
There are several advantages to working with OP9-DL (OP9
cells expressing Dll1 or Dll4) cells, namely the freeing of the reli-
ance on availability of thymic tissue, the ability to perform clonal
assays and the ease of expansion of T lineage cell populations [9].
However, one limitation is the ability to test genetic differences
within stromal cells supporting T cell development, or to match
the genetic background of stem/progenitor cells to that of the
supporting stromal cells. To circumvent these issues, we and others
recently showed that T cell development can also be supported by
thymic epithelial cells [10, 11] and primary murine dermal fibro-
blasts expressing Dll4 (mFibro-DL4) [12, 13]. Dermal fibroblasts
are easily accessible, and though not as efficient as OP9 cells, these
cells are nevertheless effective in supporting T cell development,
and importantly can be used to establish an autologous T cell
coculture system where both the supporting stromal and the T
cells are derived from the same individual or genetic background.
Here, we present the protocol for (a) establishing a mouse pri-
mary fibroblast, (b) its transduction to express Dll1 or Dll4, (c)
purification of mouse HSCs, and (d) maintenance/expansion of
cocultures and comparison to OP9-DL coculture, also demonstrat-
ing that mFibro-DL1 can also support T cell development (Fig. 1).
2 Materials
2.1 Cell Lines 1. Primary dermal fibroblasts derived from mouse ear biopsies.
and Media 2. Fetal bovine serum (FBS) (see Note 1). Heat-inactivate at
56 °C for 30 min. Store at 4 °C.
3. Phosphate-buffered saline (PBS) without Ca2+/Mg2+.
4. Trypsin 2.5 % (Life Technologies). Dilute with PBS to 0.25 %
solution.
5. Fibroblast media: DMEM supplemented with 10 % (v/v) FBS,
2 mM GlutaMAX, 2 mM 2-mercaptoethanol, Penicillin
100 U/mL–Streptomycin 100 μg/mL.
6. Sterilized 22 mm × 22 mm glass coverslips.
7. Sterilized scissors and forceps.
8. Tissue culture ware (10-cm dishes, 6-well plates), tissue
culture-treated, serological pipettes.
Induction of T Cell Development In Vitro by Delta-Like (Dll)-Expressing Stromal Cells 161
Fig. 1 Cocultures of OP9-DL1, mFibro-GFP, and mFibro-DL1 cell with hematopoietic stem cells. Different
hematopoietic cell lineage differentiation outcomes are obtained from E15 fetal liver-derived HSCs cultured
with OP9-DL1, mFibro-GFP, and mFibro-DL1 cell lines. Cocultures were set up with 103 mouse fetal liver-
derived Lineage− CD117+ Sca1+ HSCs were cultured with the different stromal cell lines as indicated. (a) At
day 6 of coculture, flow cytometric analysis was performed for the expression of CD25, CD44 for T lineage-
specified cells (upper panel) and CD19, CD11b for B and myeloid cells, respectively (lower panels). (b) A frac-
tion of the cocultures were transferred to fresh plates with stromal cells, and at day 12 of coculture, in addition
to CD25 and CD44 (upper panel), the expression of CD4 and CD8 (lower panel) was also examined. Cells were
electronically gated on live (DAPI−) and CD45+ hematopoietic cells. Data are representative of at least three
independent experiments
2.2 Transfection 9. HEK-293 T cells.

and Transduction 10. Viral packaging cell line GP+E-86 cells.
11. pMIG or similar retroviral plasmids with IRES-reporter (e.g.,
GFP) construct containing the coding region of Dll1 or Dll4
cDNA. Plasmids encoding the VSV-Env and GAG/Pol derived
from Moloney murine leukemia virus.
12. CaCl2 2.5 M (in −20 °C freezer).
13. 2× BES solution (50 mM BES (pH 6.95), 280 mM NaCl,
1.5 mM Na2HPO4).
14. 0.2 mm filter units.
2.3 HSC Isolation 15. OP9-DL cells: OP9 cells (Riken repository, Tsukuba, Japan,
and Coculture http://www.rtc.riken.go.jp) retrovirally transduced to express
the coding region for Dll1 or Dll4 cDNA, as previously
reported [8, 14].
16. 40 μm cell strainers.
17. Anti-CD24 mAb: supernatant from J11D.2 hybridoma clone
or purified anti-CD24 mAb.
18. Rabbit complement.
19. Lympholyte®-M Cell Separation Media.
20. 70 μm nylon mesh filters.

21. α-MEM—Minimum Essential Medium Eagle Alpha Modification.
22. Hank’s Buffered Saline Solution.
23. Coculture medium: α-MEM supplemented with 15 % (v/v)
FBS and Penicillin 100 U/mL /Streptomycin 100 μg/mL.
24. Mouse IL-7. Reconstitute at 1 μg/mL in coculture media.
Aliquot and store at −80 °C.
25. Human Flt-3L. Reconstitute at 5 μg/mL in coculture media.
Aliquot and store at −80 °C.
26. Mouse stem cell factor (SCF). Reconstitute at 25 μg/mL in
coculture media. Aliquot and store at −80 °C.
27. CD117-APC, Sca1-PE.
28. Freezing medium: 90 % FBS, 10 % dimethyl sulfoxide, sterile
filtered.
3 Methods
3.1 Primary Mouse Care must be taken to practice sterile culture techniques, especially
Fibroblast Culture for the isolation of primary mouse dermal fibroblasts, as the skin is
a principal source of contamination. The cultures can be prepared
as below [15]:
1. Ear tips are removed from mouse (see Note 4) that is eutha-
nized, in accordance to the institution’s Animal User Protocol
approval, using sterile scissors and forceps. The tissue samples
are then washed with 10 mL 70 % ethanol for 5 min and trans-
ferred to a laminar flow clean bench and rinsed in 10 mL
PBS. The samples are minced (~5 mm2) using sterilized surgi-
cal scissors under PBS and treated with trypsin (0.25 %) for at
least 1 h at 37 ° C.
2. Using fine sterile forceps, the apposed layers of ear epidermis
are pulled away from each other, exposing the dermis. The
peeled away tissues are placed under sterile coverslips in wells
of a 6-well per plate with 2.5 mL of Fibroblast media. The
coverslips are pressed down against the tissue with the hair side
against the coverslip using a sterile pipette tip. This has the
double effect of forcing out air bubbles and of increasing sur-
face interaction between the dermal tissue and the tissue cul-
ture plate. Several wells may be prepared this way.
3. Each well is monitored, and cells are given time to grow out
into the well for 7–10 days. Fibroblast cells migrate from the
tissue to the plate.
4. To establish a cell culture of primary fibroblasts, the media is
removed and 2 mL PBS is used to wash off any remaining
media. Using sterile forceps the coverslips are lifted. PBS is

removed and replaced with 1 mL 0.25 % trypsin followed by its
transfer to 37 °C incubator for 5–7 min. The skin tissue may
remain in the plate to be trypsinized.
5. Following trypsinization, 2 mL of fibroblast media is added to
the wells and the cells are removed from the surface of the
plate by vigorous pipetting. The trypsinized cells and tissue
pieces are passed through a 40 μm cell strainer that is placed on
a 50 mL conical tube containing 10 mL fibroblast media. 3 mL
PBS is used to rinse the plate and added to the filter on the
conical tube.
6. The strainer is discarded and the primary fibroblast cells are
centrifuged at 480 × g for 5 min at 4 °C. The supernatant is
discarded and the pelleted cells are subsequently resuspended
in 10 mL of fibroblast media, and transferred to a 10 cm plate.
7. To expand the primary fibroblast, cells are to be cultured up to
no more than 90 % confluency. Remove the media then add
5 mL PBS to wash off residual media. Remove PBS and incu-
bate with 2.5 mL 0.25 % trypsin for 5–7 min at 37 °C. Following
trypsinization, add 2.5 mL fibroblast media and vigorously
pipette the cells to remove them from the surface of the plate.
Place in a conical tube and rinse the plate with PBS and add to
the contents of the first wash. Pellet the cells at 480 × g for
5 min at 4 °C, resuspend in media, and divide among 10 cm
plates at a 1:4 passage ratio.
3.2 Mouse 1. Retroviral vector pMIG containing the Dll1 or Dll4 cDNA
Fibroblasts coding region tagged with hemagglutinin (HA) peptide [8]
Transduced to Express and an IRES-eGFP (pMIG-DL) are transfected into HEK-
Dll1/ Dll4 293T cells, using CaPO4 transfection, along with plasmids
encoding the VSV-Env and GAG/Pol. Briefly, pMIG-DL
(10 ug), VSV-env (0.3 ug) and GAG/Pol (1 ug) are mixed in
450 uL of TE and 50 uL of CaCl2 before their addition to
500 uL of 2× BES solution, while vigorously mixing by mak-
ing bubbles. The mixture is then incubated at room tempera-
ture for 20 min followed by their addition to a 10 cm plate of
HEK293Ts.
2. After overnight incubation, the medium containing the trans-
fection material is discarded. Cells are washed very gently with
5 mL PBS, and replaced with 7–10 mL fibroblast medium.
3. Within 8 h after replacement with fresh fibroblast medium,
supernatant containing the retroviral particles produced and
released into the medium by HEK293T cells is collected. To
exclude possible HEK293Ts, the supernatant is filtered using
0.2 μm filters. The filtered supernatant is then placed on
50 % confluent GP+E-86 viral packaging cells. The infection
procedure may be repeated once every 4–8 h if more than

one round is required.
4. Seven days after the last infection, GFP+ GP+E-86 cells are
isolated by flow cytometry and expanded in fibroblast media.
These cells are named GP+E-DL1 or GP+E-DL4 (GP+E-DL).
5. GP+E-DL cells are allowed to grow to confluency in 10 cm
plates. Viral particles produced by GP+E-DL cells are collected
within 4–8 h of replacing the medium with fresh fibroblast
medium and filtered using 0.2 μm filters. The supernatant is then
immediately used to transduce primary fibroblast cells at 60 %
confluency in 10 cm plates. Depending on the efficiency of trans-
duction, more than one round of transduction may be done.
6. Two days after transduction, transduced fibroblasts are tested
to determine GFP expression levels using flow cytometry.
7. If positive for GFP, the cells may be passaged into several
10 cm dishes. Cells should be passaged at no more than 90 %
confluency.
8. Seven days after the last transduction, fibroblast cells are sorted
by flow cytometry, based on levels of expression of GFP. Sorted
GFP+ fibroblast cells are named mFibro-DL1 or mFibro-DL4
(mFibro-DL) and its population expanded using fibroblast
media as described above in Subheading 3.1, step 7.
9. To prepare mFibro-DL for coculture, and compare to OP9-DL
cells as the standard, 1.5 × 105 cells per well, seeded overnight,
provide an almost confluent well in a 6-well plate for both
OP9-DL and mFibro-DL. For OP9-DL cells, coculture
medium is used and for mFibro-DL fibroblast medium. Once
the right number of cells is placed in 6-well plates, gently rock
the plate back and forth for even cell distribution.
3.3 Isolation of HSCs All experimental procedures related to the use of mice require
from Mouse d15 Fetal approval and are performed according to the guidelines specified
Liver by local institution’s animal user protocol guidelines.
1. CD1 (or desired mouse strain) timed pregnant mice are com-
mercially available from Charles River Laboratories.
Alternatively, timed paired matings may be set up in-house
in local animal facility.
2. At day 14–15 p.c., fetal mice are obtained and the livers dis-
sected using sterilized scissors and forceps.
3. Using the sterile rubber end of a 3 or 5 mL syringe plunger,
day 14–15 fetal livers are then crushed against a 40 μm cell
strainer that fits into a 50 mL conical tube. The strainer is
washed several times using coculture medium.
4. The cells are centrifuged at 480 × g for 5 min at 4 °C. The super-
natant is discarded and the cells in the pellet are resuspended in
4 mL of coculture medium per up to ten fetal livers.
5. To enrich for HSCs, mature cells are lysed using anti-CD24

complement lysis by adding the following to the cell suspen-
sion: (a) 1 mL of reconstituted rabbit complement and (b)
either 5 mL of filtered culture supernatant from J11D.2
hybridoma clone containing anti-CD24 mAb or 5 mL of
coculture media containing 1 ng of purified anti-CD24 mAb
per 106 cells. Incubate for 30 min at 37 °C.
6. 7 mL of Lympholyte-M density cell separation medium is then
carefully laid under the cell suspension as to prevent mixing.
The 2-phase solution is centrifuged at 1000 × g for 10 min at
room temperature.
7. At the end of centrifugation, the lysed cells will be in the pel-
let along with the red blood cells and the mononuclear cells
(mostly stem/progenitors) will form an opaque band at the
interphase between Lympholyte M and the coculture media.
The interphase layer is carefully removed using a serological
pipette and transferred to a 50-mL tube. Note that there will
be some mixing of the two phases while the cells are being
taken up. The volume of this mixture is measured and at least
twice at much coculture medium is added. The mixture is then
centrifuged at 700 × g for 5 min at 4 °C to obtain the HSC-
enriched cells in the pellet.
8. At this point, the cell population will be about 85 % CD117+
(c-kit+). The cells may be cryopreserved by resuspending them
in freezing medium at about 4 embryos/fetal livers per mL of
freezing medium per vial. At a later date, the cells may be
thawed by placing the cryovial containing the fetal liver cells at
37 °C for approximately 1 min. As soon as the cells are thawed,
they are gently place in a 15 mL conical tube containing 9 mL
of Hank’s Buffered Saline Solution and mixed before centrifu-
gation and staining for HSCs.
9. HSCs (CD117+ Sca1+ Lin/CD24−), which should account for
about 5 % of the cells, may be obtained using flow cytometric
fluorescent-activated cell sorting. The cells are stained with
antibodies against CD117 and Sca1 conjugated to fluoro-
chromes of choice—CD117-APC and Sca1-PE are recom-
mended, and sorted for CD117+ Sca1+ cells. After washing, the
sorted cells are ready to be placed on OP9-DL4 or mFibro-
DL4 to initiate the coculture.
3.4 Initiation Cells may be seeded in 6-well plates or 10 cm plates or other ves-
and Maintenance sels. Here we describe coculture in 6-well plates.
of Coculture
1. Seed 1 × 103 HSCs in 3 mL of coculture medium per 1 well of
~90 % confluent OP9-DL or mFibro-DL cells, and add 5 ng/
mL Flt-3L and 1 ng/mL IL-7 and 25 ng/mL SCF (Note, the
addition of SCF is not necessary when setting up OP9-DL1
cocultures). Place the culture plates in the incubator.
2. Day 5: Prepare fresh plates of OP9-DL and mFibro-DL (as

described above) for transfer on Day 6.
3. Day 6: Small Clusters of round, shiny cells will be present, with
some cells detached in suspension. Cells are disaggregated by
vigorously pipetting the medium up and down using a serologi-
cal pipette with pipette-filler. The stromal monolayer may be
disrupted during this process and detached from the plate. Once
detached, the stromal cells can form clumps that are filtered out
by passing the resuspended cells in the medium through 70 μm
nylon mesh filter into 15 mL conical tube. The wells may be
washed with 3 mL of PBS and passed through the same filter.
4. At this point, the cells are split into two tubes with one-sixth to
be transferred to a fresh plate of stromal cells. 1 mL of the
mixture is transferred to a 5 mL polypropylene round bottom
tube and centrifuged at 480 × g for 5 min at 4 °C before dis-
carding the supernatant and resuspension in 3 mL of coculture
media containing cytokines. The cells are then transferred to a
fresh plate of stromal cells. The remaining 5 mL of cells may be
used for cell counts and flow cytometry analysis comparing
OP9-DL and mFibro-DL cocultures (see Fig. 1a for compari-
son of OP9-DL1, mFibro-GFP (control), and mFibro-DL1).
5. Day 9 and onward. Follow the instructions outlined above in
Subheading 3.4, step 2 approximately once every 3–5 days. The
ratio of the cells transferred may be less or more depending on
the number and the frequency of the cells that one is interested
in. In addition, the culture can be expanded into larger plates.
4 Notes
1. We routinely test new lots of FBS serum against a standard lot

of FBS previously shown to support T lymphopoiesis in cul-
ture, such as fetal thymic organ cultures.
2. To preserve early passage stocks of mFibro-DL4 stromal cells,
once the plate’s confluency reaches 80–90 % of the 10 cm dish,
cells are split into four 10-cm dishes. Passaging procedures are
continued until the desired number of plates are 80 % conflu-
ent. Freeze one plate per cryovial in freezing media and place
in liquid nitrogen.
3. To ensure Dll1-HA or Dll4-HA is expressed, a sample of
GP+E-DL and mFibro-DL cells may be lysed to perform a
Western immunoblot assay using an Anti-HA antibody.
4. We illustrate here that both the fibroblasts and the hematopoi-
etic component are derived from the same strain, but not the
same individual mouse; nevertheless, the MHC of the mice
from which the stroma and the hematopoietic cells are derived
are identical, as shown in Fig. 1.
5. For QRT-PCR of the hematopoietic cells in coculture, it is not

enough to filter out the clumps of stromal cells through the
nylon mesh filter. There will be some single stromal cells that
will go through the filter and may change the values obtained.
We recommend placing the filtered resuspended cells onto a
tissue culture plate and place in the CO2 incubator. Both OP9
and fibroblast cells are quite adherent and within 15 min will
attach to the surface. The non-adherent hematopoietic com-
ponent will stay in suspension and can be harvested.
References
1. Jenkinson EJ, Owen JJ (1990) T-cell differen- levels shows differential lymphomyeloid lin-
tiation in thymus organ cultures. Semin eage commitment outcomes. J Immunol
Immunol 2:51 185:867
2. Takahama Y (2000) Differentiation of mouse 9. Zuniga-Pflucker JC (2004) T-cell development
thymocytes in fetal thymus organ culture. made simple. Nat Rev Immunol 4:67
Methods Mol Biol 134:37 10. Beaudette-Zlatanova BC et al (2011) A human
3. Deftos ML, Bevan MJ (2000) Notch signaling thymic epithelial cell culture system for the
in T cell development. Curr Opin Immunol promotion of lymphopoiesis from hematopoi-
12:166 etic stem cells. Exp Hematol 39:570
4. Schmitt TM, Zuniga-Pflucker JC (2002) 11. Mohtashami M, Zuniga-Pflucker JC (2006)
Induction of T cell development from hemato- Three-dimensional architecture of the thymus
poietic progenitor cells by delta-like-1 in vitro. is required to maintain delta-like expression
Immunity 17:749 necessary for inducing T cell development.
5. Dervovic DD, Ciofani M, Kianizad K, Zuniga- J Immunol 176:730
Pflucker JC (2012) Comparative and func- 12. Mohtashami M, Shah DK, Kianizad K, Awong
tional evaluation of in vitro generated to G, Zuniga-Pflucker JC (2013) Induction of
ex vivo CD8 T cells. J Immunol 189:3411 T-cell development by Delta-like 4-expressing
6. Dervovic DD et al (2013) Cellular and molec- fibroblasts. Int Immunol 25:601
ular requirements for the selection of in vitro- 13. Smeets MF et al (2013) Removal of myeloid
generated CD8 T cells reveal a role for Notch. cytokines from the cellular environment
J Immunol 191:1704 enhances T-cell development in vitro. Int
7. Holmes R, Zuniga-Pflucker JC (2009) The Immunol 25:589
OP9-DL1 system: generation of T-lymphocytes 14. Schmitt TM, Zuniga-Pflucker JC (2006) T-cell
from embryonic or hematopoietic stem cells development, doing it in a dish. Immunol Rev
in vitro. Cold Spring Harb Protoc 2009:pdb. 209:95
prot5156 15. Takashima A (2001) Establishment of fibro-
8. Mohtashami M et al (2010) Direct comparison blast cultures. Curr Protoc Cell Biol Chapter
of Dll1- and Dll4-mediated Notch activation 2:Unit 2.1
Chapter 15
In Vitro Analysis of Thymocyte Signaling

Abstract
From the moment a developing thymocyte expresses a TCR, it is subjected to numerous interactions with
self-peptide/MHC complexes that determine its ultimate fate. These include death by neglect, negative
selection (apoptosis and lineage deviation), positive selection, and lineage commitment. The identification
of signals that govern these unique cell fates requires the ability to assess the activity, level of expression,
subcellular location, and the molecular associations of numerous proteins within the developing T cell.
Thus, this chapter describes methods designed to analyze thymocyte signaling under various types of
peptide-based stimulation in vitro.
Key words T cell development, Positive selection, Negative selection, TCR signaling, Death by neglect
1 Introduction
The selection of functional T cells is mediated by interactions

between the T cell antigen receptor and self-peptide/major histo-
compatibility complex expressed on thymic epithelium. These
interactions either lead to survival and development, or death. The
affinity model proposes that the selection outcome is determined
by the affinity of the TCR for a peptide–MHC complex. Weak
TCR/peptide–MHC interactions do not support thymocyte sur-
vival (death by neglect); strong interactions lead to thymocyte
apoptosis, lineage deviation or receptor editing (collectively called
negative selection); and interactions between these extremes lead
to the development of mature T cells (positive selection) [1]. More
recently, a number of “third” signals have been given a role in the
selection process [2]. A longstanding issue remains as to how the
TCR reads the parameters of ligand engagement, along with these
third signals, to direct these distinct cell fates.
We use several standard biochemical techniques to assess signals
in our system. Immunoblot techniques are used to determine the
expression levels and activation states of proteins in large synchronous
populations of thymocytes: for these assays we stimulate, lyse,
169
SDS-PAGE, transfer and immunoblot (see below for stimulation

and lysis protocols). Flow cytometry is used to perform similar assays
on small/rare populations of cells that can be distinguished by a
specific phenotypic marker (also works for large populations): for
these assays we stimulate, stain surface markers and perform intracel-
lular staining. In general we use standard kits (e.g. BD Phosflow or
Cytofix/Cytoperm kits) and the methods are not described here.
However, we have used the fixation and permeabilization protocol
reagents below in single cell suspension to assess expression levels of
a number of different proteins with some success. Confocal micros-
copy provides activation and location information that has added to
our understanding of the diversity in the regulation of selection sig-
nals [3, 4]. These assays are described below.
Historically, immunoprecipitation followed by western blot-
ting (IP-Western) has been invaluable for elucidating the protein–
protein interactions that occur during thymocyte positive and
negative selection in the thymus. We have used these assays exten-
sively to map out differences in signaling in large synchronous
populations of T cells available from thymi of mice on a non-
selecting background [3]. However, IP-Western can have signifi-
cant drawbacks when targeting small/rare populations of T cells,
proteins that are expressed at low levels, or proteins whose avail-
able reagents are inefficient at IP. Thus, IP followed by flow cytom-
etry (IP-FCM) offers the benefit of being able to precipitate low
abundance proteins to generate data with a high signal to noise
ratio [5–7]. Additionally, this protocol offers the benefit of being
able to use significantly fewer cells for analysis (~1 × 105 for IP-FCM
vs. 100 × 106 for IP-Western). Others have even successfully per-
formed these types of analyses on <1 × 103 cells [7]. These assays
are described below.
2 Materials
Prepare all solutions using autoclave sterilized MilliQ (or equiva-

lent) water and molecular biology grade reagents. All reagents are
stored at room temperature unless otherwise noted.
2.1 Tissue 1. Complete media: RPMI1640 supplemented with 1 mM non-

Preparation essential amino acids (NEAA), 1 mM sodium pyruvate, 50 μM
β-Mercaptoethanol, 100 U/mL Penicillin 100 μg/mL
Streptomycin 0.292 mg/mL Glutamine, and 10 % FBS.
2. TCR transgenic mouse on a non-selecting background. For
CD8 T cell selection TCR transgenic Rag−/− β2m−/− mouse
preferably <6 weeks of age are ideal (see Note 1).
3. Alternatively, to assess cohorts of T cells at multiple stages of
development, thymi from TCR transgenic and polyclonal mice
on a selecting background can also be used (see Note 2).
2.2 Stimulation 1. Antigen-presenting cells: examples: bulk splenocytes, T cell-

with Peptide-Pulsed depleted spleens, bone marrow-derived dendritic cells or other
APC (See Note 3) cell lines that express the appropriate MHC.
2. Peptide solution, concentration previously optimized.
2.3 Stimulation 1. Peptide/MHC Tetramers (2 mg/mL).

with Peptide/MHC 2. Complete media (see Subheading 2.1).
I Tetramers
2.4 Confocal 1. 13 mm round glass coverslips (no. 1: 0.13 mm thick).

Microscopy 2. Microscope slides (pre-cleaned) 25 × 75 × 1.0 mm.
3. Humidity chamber: 6-well plates with a coverslip on diameter
filter paper, parafilm and ~1 mL water (Fig. 1) (see Note 4).
4. Complete media (see Subheading 2.1).
5. 0.2 % Triton solution.
6. Fresh 1.5 and 3.0 % (w/v) BSA/PBS solution.
7. Staining conditions for surface and/or cytosolic proteins: pre-
pare a stock solution of 4 % formaldehyde fix (1 part 16 %
formaldehyde plus 3 parts PBS). Make and keep at RT in
chemical hood.
8. Staining conditions for cytosolic and/or nuclear proteins: pre-
pare a stock solution of 8 % formaldehyde fix (1 part 16 %
formaldehyde plus 1 part PBS). Make and keep at RT in
chemical hood.
9. Blocking buffer: 25 mL 3 % (w/v) BSA/PBS. Keep on ice
until use.
10. OPTIONAL: DNA/nuclear staining reagent such as DAPI or
DRAQ-5AT if nuclear localization is desirable.
Fig. 1 Schematic of a humidity chamber for confocal microscopy. In general we

prepare a 6-well plate with a filter paper, covered by piece of parafilm with
approximately 1 mL of water. The filter paper and parafilm are circles that we
mark and cut using the diameter of a 50 mL conical tube as a template. The
cover of the plate is labeled with the conditions (stimulation, stain, time, etc.).
These are easily reused from experiment to experiment
2.5 IP-FCM 1. Lysis Buffer: 10 mM Tris pH 7.4, 1 % Triton X-100, 0.5 %

NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA. Add
20 mM NaF, 0.2 NaVO4, and 0.2 mM PMSF fresh every time
before use.
2. Carboxylate-modified polystyrene latex (CML) beads ~6 μM
diameter (Store at 4 °C).
3. Monoclonal antibodies for IP in PBS (≥0.2 mg/mL, must be
devoid of BSA).
4. Antibodies for detection (Must be directly conjugated to a
fluorochrome or a species other than IP mAb to avoid
cross-reactivity).
5. MES Coupling Buffer: 50 mM 2-(N-morpholino)ethanesul-
fonic acid (MES) pH 6.0, 1 mM EDTA.
6. EDAC-MES: 50 mg/mL 1-ethyl-3-(3-dimethylaminopropyl)
carbodiimide hydrochloride (EDAC).
7. Quenching/Blocking/Storage (QBS) Buffer: 1 % BSA + 0.02 %
Sodium azide in 1× PBS (Store at 4 °C).
8. FCM Staining Buffer: 1 % FBS + 0.2 % Sodium azide in 1×
PBS (Store at 4 °C).
3 Methods
3.1 Harvesting 1. Prepare sterile tools and small Petri dishes with 1 mL of com-
Thymi (See Note 5) plete media and 1 piece of sterile nylon mesh.
2. Place Petri dish on ice.
3. Harvest thymi from mice that have been euthanized according
to your institutional guidelines. Place in the Petri dish on ice
(one thymus per Petri dish).
4. Place another piece of mesh over the thymus and press with a
plunger from a 3 mL syringe to create a single cell
suspension.
5. Pass cells to a clean 15 mL conical tube that contains 3 mL of
complete media.
6. Rinse the mesh with 1 mL of fresh complete media and trans-
fer to the conical tube.
7. Spin cells at 483 × g (450–500 × g) for 5 min in centrifuge (4 °C).
8. Resuspend cells in fresh complete media (1–2 mL/thymus).
9. Transfer the cells to a new conical (50 mL) through cell
strainer (or nylon mesh) to remove large clumps of cells and
thymic debris.
10. Count with a hemocytometer (or electronic counter) and
dilute cells to a concentration 10 × 106/mL. Proceed to stimu-
lation of your choice.
3.2 Stimulation 1. Prepare media that contains the peptide at the desired concen-
with APC tration (see Note 6).
2. Resuspend APC at 20 × 106/mL in the peptide solution (from
3.2.1 Peptide-
step 1).
Loaded APC
3. Incubate at 37 °C in the culture incubator from a minimum of
30 min.
4. Wash cells with complete media, spin at 483 × g (450–500 × g)
for 5 min and discard supernatant.
5. Resuspend cells at 10 × 106 cells/mL of complete media.
3.2.2 Stimulation 1. Set the thermomixer or water bath at 37 °C.

with Peptide-Loaded APC 2. Mix T cells and antigen-loaded APCs at a ratio 1:1. Ideally, for
confocal microscopy or ICS-flow cytometry, mix 50 μL APCs
(10 × 106/mL) with 50 μL T cells (10 × 106/mL) in a 1.5 mL
microfuge tube.
3. Centrifuge in bench top mini-centrifuge. Use pulse/quick
spin mode for 7–10 s.
4. Tap the pellet.
5. Place in thermomixer at 37 °C for the planned time course.
6. Place the tubes on ice to stop stimulation.
3.2.3 Stimulation 1. Add approximately 1 × 106/staining condition to 1.5 mL

with Peptide/MHC microfuge tube. Example: for an experiment requiring three
I Tetramers staining conditions for each time point, one would add 3 × 106
thymocytes (300 μL of 10 × 106/mL suspension of thymo-
cytes from above).
2. Place cells in thermomixer (37 °C, 200–500 rpm mixing).
3. Add desired concentration of peptide/MHC tetramers and
incubate for planned time course (see Note 7).
4. Wash 2× with ice-cold media (to remove excess tetramer) and
place the tubes on ice to stop stimulation.
3.3 Confocal 1. Wash cells from stimulation (above) and resuspend pellet
Microscopy (10 × 106 cells/mL) by gently tapping the microfuge tube (see
Note 8).
3.3.1 Mounting Cells
2. Place 100 μL of stimulated cells on the appropriate coverslip.
TAKE NOTE OF PLATE LAYOUT.
3. To allow the cells to settle onto the coverslip, carefully place
covered 6-well plates in refrigerator for minimum of 10 min.
During this time prepare the staining buffers. Fixation:
(see Note 9).
4. Move plates to chemical hood and add:

(a) For surface and/or cytosolic stains: 100 μL of 4 % fix.
Incubate at RT for 45 min.
(b) For cytosolic and/or nuclear stains: 100 μL of 8 % fix.
Incubate at RT for 30 min.
5. Carefully aspirate liquid off each coverslip and wash 2× with
200 μL PBS. (You may stop the procedure at this step. To do
this, add 200 μL PBS and place plates at 4 °C). For cytosolic
and nuclear stains proceed to step 10.
3.3.2 Permeablization/ 1. Gently, but quickly add 100 μL of ice cold 0.2 % triton drop-
Blocking wise and incubate for 2 min. (For the best results, work with
six wells at a time. Start the timer after adding the triton to last
coverslip.)
2. Aspirate triton and add ice cold 3 % BSA/PBS.
3. Repeat steps 5 and 6 until all coverslips have been treated.
4. Place plates at 4 °C overnight. Alternatively, block for a mini-
mum of 1 h then proceed to staining procedure.
3.3.3 Staining 1. Prepare staining cocktails.

(a) For primary surface and cytosolic stains, add the desired
antibodies to 1.5 % BSA/PBS (w/v).
(b) For primary cytosolic and nuclear stains, add desired anti-
bodies to 1.5 % BSA/0.2%Triton solution.
(c) Prepare secondary antibodies by adding labeled antibod-
ies to 1.5 % BSA/PBS (w/v).
2. Aspirate coverslip and wash 2× with RT PBS. For cytosolic and
nuclear stains skip ahead to step 16.
3. Add primary antibody staining cocktail. TAKE NOTE OF
PLATE LAYOUT!!!
4. Incubate at RT in the dark for 1 h. During the staining incu-
bations make labels for slides.
5. Wash 2× with PBS and add secondary staining cocktail.
6. Incubate at RT in the dark for 1 h. For surface and cytosolic
stains proceed to step 17.
7. Add 50 μL of the primary stain cocktail in PBS/BSA 1.5 %/
Triton 0.2 % for 30 min at RT
8. Wash three times in PBS/BSA1.5 %
9. Add secondary staining cocktail and incubate for 1 h.
10. OPTIONAL: If DNA/Nuclear staining is desired, dilute
DRAQ-5 at 1/5000 (or other DNA stain) in PBS/formalde-
hyde 1 %. Add 50 μL to the coverslip and incubate for 5 min
at RT.
3.3.4 Mounting 1. Using a forceps to pick up coverslip, dab the edge of the cov-
the Coverslip erslip with a lint-free tissue to remove excess stain. Then dip
the coverslip in PBS and dab the edge of the coverslip with
tissue three times, then dip in water and dab once. Place the
coverslip CELL SIDE DOWN, onto a microscope slide that
has one drop of mounting media (see Note 10).
2. Cover the mounted coverslip/slides and let dry overnight at
RT, then paint the edges of each coverslip with clear nail polish
(to make a seal) and place in slide box and place at 4 °C. NOTE:
The slides are then ready for analysis by confocal microscopy.
While in theory the stains remain stable for up to a year, we
suggest they be analyzed within 3 months of mounting.
3.4 IP-FCM 1. Determine the concentration of CML beads by diluting beads

in PBS and enumerating using a hemacytometer under a
3.4.1 Generation
microscope.
of Ab-Coupled CML Beads
2. Gently vortex CML beads to suspend them evenly and transfer
20 × 106 beads to a 1.5 mL centrifuge tube.
3. Wash the beads three times in 0.5 mL MES Coupling Buffer,
centrifuging at ~16,000 × g for 5 min at 25 °C.
4. Following last wash, aspirate buffer leaving beads undisturbed.
Resuspend beads in 50 μL MES Coupling Buffer.
5. Add 20 μL EDAC-MES (made fresh) and mix by pipetting up
and down for 15 min to prevent beads from settling.
6. Wash the beads three times in 0.5 mL MES Coupling Buffer,
centrifuging at ~16,000 × g for 5 min at 25 °C.
7. Aspirate buffer leaving beads undisturbed and resuspend beads
in 50 μL PBS.
8. Add 50 μL mAb of choice.
9. Mix for 3–4 h in a thermomixer at ~1100 rpm at 25 °C, being
sure that no beads are settling at the bottom of the tube.
10. Wash beads three times in 1 mL PBS, centrifuging at ~16,000 × g
for 5 min at 25 °C being sure to discard as much supernatant
as possible without disturbing the CML bead pellet.
11. Following last wash, aspirate all PBS and resuspend beads in
100 μL QBS Buffer and store overnight at 4 °C.
12. Count beads by diluting in PBS and enumerating using a
hemacytometer under a microscope.
3.4.2 Sample Lysis 1. Lyse 10 × 106 thymocytes in 100 μL Lysis Buffer (ice-cold) in
a 1.5 mL microcentrifuge tube for 20 min on ice.
2. Following lysis spin crude lysate for 50 min at ~16,000 × g at
4 °C.
3. Pass supernatant to a fresh 1.5 mL microcentrifuge tube
(see Note 11).
4. Add 2.5 × 105 IP beads to the lysate (see Note12).

5. Place tubes in a thermomixer at ~1100 rpm overnight (place
mixer itself in a 4 °C fridge), making sure that beads do not
settle.
3.4.3 Staining of Beads 1. Wash CML beads that have been incubated with lysate three
for FCM times in ice-cold Lysis Buffer being sure to discard as much
buffer as possible without disturbing the CML bead pellet.
2. Resuspend CML beads in FCM staining buffer and pass beads
to 96-well round-bottom plate.
(a) Split the CML beads into enough wells to assess the pro-
teins of interest plus isotype control(s).
3. Add fluorochrome-conjugated mAbs (or primary Abs) to the
samples and incubate for 30 min on ice in darkness. α-POSH,
clone M-290, α-JIP-1 clone M-300 and α-JNK1 clone C-17
used for demonstration (Fig. 2) [5].
4. Wash beads two times with FCM Buffer, spinning at ~500 × g
for 5 min at 25 °C.
5. If unlabeled primary Abs are used, add fluorochrome-
conjugated secondary Ab and incubate for 30 min on ice in
darkness. If directly conjugated Abs are used, skip this step.
6. Resuspend IP beads in 200 μL FCM Staining Buffer and pro-
ceed to analysis on the flow cytometer.
Fig. 2 IP-FCM analysis of protein interactions in pre-selection DP thymocytes. (a) Gating on singlet beads
based on FSC and SSC properties. (b) Proper calibration of fluorescence voltage such that the isotype control
falls between 100 and 101 while a “bright” positive control falls on scale. (c) Representative experiment where
α-POSH CML beads are used on lysates of pre-selection DP thymocytes. Histograms represent the presence
of JIP-1 and JNK1 bound to POSH. Red histogram represents the “loading control” to quantify how much POSH
was precipitated
3.4.4 Acquisition of Data 1. Increase the FSC and SSC of the cytometer such that bead
by FCM singlets are clearly visible (see Note 13).
2. Draw a gate around the population of bead singlets (Fig. 2a).
3. Apply this gate and visualize the fluorescence data on a log
scale for each of the fluorescence channels being used.
4. Pass the isotype control sample and confirm that the generated
histogram falls between 100 and 101 (see Note 14).
5. Pass the control for a very bright sample (see Note 15).
6. Pass experimental samples (Fig. 2c).
7. Acquire >1000 bead events. Analyze as described [5, 7].
4 Notes
1. Older mice will work. However, they have significantly fewer

pre-selected DP.
2. Due to the heterogeneity of the T cell populations from these
thymi, confocal and flow cytometry-based assays are more
suitable to assess signaling in these T cells.
3. Due to potential contamination with APC-derived proteins,
this method is not appropriate for immunoblotting, IP or
IP-FCM.
4. Carefully plan out the stains and time course in advance. Label
the cover of the 6-well plate with the stimulation, stain and
timepoint for each coverslip. Keep in mind how you want to
collect the data at the confocal. One slide holds three cover-
slips max. An intelligent and logical layout greatly simplifies
the collection and analysis of the data. In addition, it greatly
reduces the probability of mislabeled coverslips.
5. For short-term assays, those that are performed in less than
24 h, this step can be performed at the bench. Longer time
points require FTOC (described in Chapter 12) and should be
performed in a sterile hood.
6. When using peptides of different strengths, always start by
handling the weakest peptide and move step by step to the
strongest peptide.
7. You may substitute anti-CD3 with or without cross-linking.
When using cross-linking, pre-incubate thymocytes with anti-
CD3 on ice for 15 min, wash, and add pre-warmed media with
appropriate concentration of cross-linking antibody, start timing.
8. For analyses involving confocal examination of thymocyte/
APC conjugates do not mix cells by pipetting.
9. If your cell surface target is sensitive to fixation, stain on ice
after stimulation and prior to fixation).
10. To avoid bubbles, hold coverslip cell side down at 45 % angle

and slide the coverslip across slide until it touches the drop of
mounting media then gently lay the coverslip down onto slide.
11. Alternatively, one can proceed to standard SDS-PAGE/immu-
noblot (described elsewhere [3]) procedures at this point.
12. The number of thymocytes used, the volume of lysis buffer,
and the number of CML beads used must be determined
empirically based on the proteins being analyzed.
13. The CML beads being used are smaller than thymocytes, as
such the FSC and SSC gain will have to be increased to make
them visible to the cytometer.
14. If fluorescence is too bright lower the channel voltage, if the
histogram does not show, increase the channel voltage.
α-Rabbit-488 used for demonstration purposes (see Fig. 2b).
15. A good control for a very bright sample is anti-mouse Ig (if
your Ab coupled to the CML bead is mouse) as it is highly
abundant on the CML bead. α-Mouse 488 was used for dem-
onstration purposes (see Fig. 2b).
Acknowledgements
The authors thank the members of the Daniels and Teixeiro labs
for assistance in compiling these protocols. Funding provided by
grants from the Missouri Mission Enhancement Fund and
University of Missouri Research Board is greatly acknowledged.
References
1. Starr TK, Jameson SC, Hogquist KA (2003) 5. Cunningham CA, Knudson KM, Peng BJ,
Positive and negative selection of T cells. Annu Teixeiro E, Daniels MA (2013) The POSH/
Rev Immunol 21:139–176 JIP-1 scaffold network regulates TCR-mediated
2. Stritesky GL, Jameson SC, Hogquist KA JNK1 signals and effector function in CD8 T
(2012) Selection of self-reactive T cells in the cells. Eur J Immunol 43(12):3361–3371
thymus. Annual Rev Immunol 30:95–114 6. Gil D, Schrum A, Alarcon B, Palmer E (2005)
3. Daniels MA, Teixeiro E, Gill J, Hausmann B, T cell receptor engagement by peptide – MHC
Roubaty D, Holmberg K, Werlen G, Holländer ligands induces a conformational change in the
GA, Gascoigne NR, Palmer E (2006) Thymic CD3 complex of thymocytes. J Exp Med
selection threshold defined by compartmental- 201(4):517–522
ization of Ras/MAPK signalling. Nature 7. Schrum AG, Gil D, Dopfer EP, Wiest DL,
444(7120):724–729 Turka LA, Schamel WW, Palmer E (2007)
4. Daniels MA, Teixeiro E (2010) The persistence High-sensitivity detection and quantitative
of T cell memory. Cell Mol Life Sci 67(17): analysis of native protein-protein interactions
2863–2878 and multiprotein complexes by flow cytometry.
Sci STKE 2007(389):pl2
Chapter 16
Molecular Analysis of Mouse T Cell Receptor α and β Gene

Rearrangements
Levi J. Rupp, Liang Chen, Michael S. Krangel, and Craig H. Bassing
Abstract
PCR on genomic DNA isolated from lymphocyte populations is an invaluable technique to analyze T cell
receptor (TCR) α and β gene rearrangements. Although this approach is powerful, it also has limitations
that must be accounted for in experimental design and data interpretation. Here, we provide background
required for understanding these limitations, and then outline standard PCR methods that can be used for
analysis of TCRα and β gene rearrangements in mice.
Key words VDJ recombination, T cell receptor (TCR) α and β genes, Variable (V), diversity (D), and
joining (J) gene segments, Thymocytes, αβ T cells
1 Introduction
1.1 Biochemistry Developing αβ T lymphocytes assemble TCRα and β genes from

of TCRα and β Gene germline variable (V), diversity (D), and joining (J) gene segments.
Rearrangements This recombination process is catalyzed by the lymphocyte-specific
RAG1/RAG2 (RAG) endonuclease and ubiquitously expressed
non-homologous end-joining (NHEJ) DNA repair proteins.
Recombination signal sequences (RSSs) comprised of conserved
heptamer and nonamer sequences separated by 12 or 23 nucleo-
tides (12-RSSs or 23-RSSs) flank TCRα and β V, D, and J gene
segments. RAG binds a single RSS, captures a second RSS to form
a synaptic complex, and then cleaves DNA between each RSS and
its flanking gene segment, forming blunt signal ends and hairpin-
sealed coding ends [1]. RAG only mediates synapsis and cleavage
of gene segments flanked by RSSs of distinct spacer lengths, a level
of regulation known as the “12/23” rule. In a subsequent reaction
phase, the RAG and NHEJ proteins cooperate to ligate signal ends
together to form signal joins and to open, process, and then ligate
coding ends together to form V(D)J coding joins [2]. Signal joins
are precise, however coding joins are imprecise due to heterogene-
ity in nucleotide position of hairpin-opening, deletion of coding
179
180 Levi J. Rupp et al.
end sequences, and addition of non-template nucleotides. Coding

joins form the second exons of TCRα and β genes and are tran-
scribed with first exons and downstream constant (Cα or Cβ)
region exons. In a population of αβ T cells, the combination of
possible joining events and inherent imprecision in V(D)J coding
join formation cooperate to generate TCRα and β gene diversity.
While combinatorial and junctional diversity are each important
for adaptive immunity, they pose substantial obstacles for analysis
of TCRα and β gene rearrangements.
1.2 Structure The genomic structure of TCRβ loci is conserved across commonly
and Recombination used lab mouse strains. The TCRβ locus spans 685 kb on chromo-
of TCRβ Loci some 6. It is comprised of 23 functional Vβ segments and two
Dβ-Jβ-Cβ clusters; each Dβ-Jβ-Cβ cluster spans ~6 kb and contains
a single Dβ segment (Dβ1 or Dβ2), six functional Jβ segments
(Jβ1.1-1.7 or Jβ2.1-2.7 where Jβ1.7 and Jβ2.6 are not functional),
and two sets of Cβ region exons (Cβ1 or Cβ2) (Fig. 1). All but two
of the 23 functional Vβs reside within a 234 kb Vβ cluster that ends
250 kb upstream of Dβ1. One of the other two Vβs resides 156 kb
upstream of this Vβ cluster and the second lies 10 kb downstream
of Cβ2. In addition to these 23 functional Vβs, the Vβ cluster con-
tains 12 non-functional Vβs that are pseudogenes or lack compe-
tent RSSs. Except for the Vβ downstream of Cβ2, all Vβs reside in
the same transcriptional orientation as the Dβ-Jβ-Cβ clusters.
Historically, Vβ segments were named in order numerically as
they were identified. Presently, the IMGT (www.imgt.org) and
NCBI name Vβ segments based on their genomic location with the
Vβ furthest upstream of Dβ1 called TRBV1 and the other func-
tional and non-functional Vβs designated by number in ascending
order moving along the locus (with the exception of the TRBV12
and TRBV13 families that are interspersed). Dβ and Jβ segments
Fig. 1 Schematic of TCRβ locus. Gene segments are listed below each region, from distal to proximal, and are
identified by NCBI and IMGT nomenclature. Gray lettering indicates that segments are non-functional. Locus is
not to scale, and double hash marks indicate large intergenic distances. Arrow denotes that TRBV31 is in
opposite transcriptional orientation to the rest of the locus
Molecular Analysis of Mouse T Cell Receptor α and β Gene Rearrangements 181
are also numbered in ascending order and labeled TRBD and TRBJ
in the current system. Peptides encoded by Vβ segments, however,
are still referred to by their historical names. For example, the
TRBV1 segment was historically called Vβ2, and TCRβ proteins
that contain this Vβ are still referred to as Vβ2+ TCRβ chains. In
Table 1, we summarize the differences between historical and
IMGT/NCBI nomenclature. For this chapter, we use the IMGT/
NCBI nomenclature for Vβ gene segments and encoded peptides.
Table 1
Current NCBI and historical nomenclature for TCRβ locus variable (V)
region segments
NCBI Historical Join PCR

nomenclature nomenclature primer [19]
TRBV1 Vβ2 TCCTGGGGACAAAGAGGTCAAATC
TRBV2 Vβ4 AGCTATCAAAAACTTATGGACAATCAG
TRBV3 Vβ16 GATTTTAGGACAGCAGATGGAGTTTC
TRBV4 Vβ10 GCGCTTCTCACCTCAGTCTTCAG
TRBV5 Vβ1 AAATGAGACGGTGCCCAGTCGTT
TRBV6 Vβ26 Non-functional
TRBV12-1 Vβ5.2 CAGCAGATTCTCAGTCCAACAGTTTa
TRBV13-1 Vβ8.3
TRBV12-2 Vβ5.1 CAGCAGATTCTCAGTCCAACAGTTTa
TRBV13-2 Vβ8.2
TRBV12-3 Vβ5.3 Non-functional
TRBV13-3 Vβ8.1 TATATGTACTGGTATCGGCAGGACA
TRBV14 Vβ13 CTGCTGTGAGGCCTAAAGGAACTAA
TRBV15 Vβ12 AGCTGAGATGCTAAATTCATCCTTC
TRBV16 Vβ11 TTCTCAGCTCAGATGCCCAATCAG
TRBV17 Vβ9 TTCCAATCCAGTCGGCCTAACAAT
(continued)
Table 1
(continued)

nomenclature nomenclature primer [19]
TRBV19 Vβ6 AAGGCGATCTATCTGAAGGCTATGA
TRBV20 Vβ15 CCCATCAGTCATCCCAACTTATCC
TRBV21 Vβ19 CTACAAGAAACCGGGAGAAGAACTC
TRBV23 Vβ20 CTGGTATCAACAAAAGCAGAGCAAA
TRBV24 Vβ17 TCGAAATGAAGAAATTATGGAACAAAC
TRBV26 Vβ3 GAAAAACGATTCTCTGCTGAGTGTCC
TRBV29 Vβ7 AGCTGATTTATATCTCATACGATGTTG
TRBV30 Vβ18 CCGGCCAAACCTAACATTCTCAAC
TRBV31 Vβ14 AGAGTCGGTGGTGCAACTGAACCT
Segments are listed starting with the most DβJβ distal and moving along the locus
toward the DβJβ cluster. Also contains Vβ primer sequences for join PCRs
a
This primer binds both TRBV12-1 and TRBV12-2
Recombination of TCRβ loci is dictated by RSSs flanking Vβ,

Dβ, and Jβ gene segments [3]. Relative to their transcriptional ori-
entation, Vβs are flanked by downstream 23-RSSs, Jβs are flanked
by upstream 12-RSSs, and Dβs are flanked by upstream 12-RSSs
and downstream 23-RSSs. By the 12/23 rule, recombination
between Vβ and Jβ segments should be possible. However, beyond
12/23 compatibility RSS joining restrictions prevent recombina-
tion between Vβ and Jβ segments.
Recombination of TCRβ loci is regulated along and between
alleles [4]. On each allele, TCRβ recombination is ordered with
Dβ-to-Jβ rearrangements preceding rearrangements of Vβ seg-
ments to DJβ complexes. The TRBD1 segment can rearrange to
any of the six functional TRBJ1 or TRBJ2 segments and possibly
to the TRBD2 segment, while the TRBD2 segment can only rear-
range to any of the six functional TRBJ2 segments. Each Vβ can
rearrange to whatever TRBD1TRBJ1, TRBD1(TRBD2)TRBJ2,
or TRBD2TRBJ2 complexes are available. In this regard, Vβ seg-
ments can recombine with TRBD2TRBJ2 complexes on alleles
that have previously rearranged a Vβ segment to a TRBD1TRBJ1
complex. All TCRβ recombination events occur through deletion
of intervening sequences, except for TRBV31 rearrangements

to DβJβ complexes that occur through inversion of interven-
ing sequences. While TCRβ alleles can assemble both a
TCRBV31TRBD2TRBJ2 rearrangement and a Vβ−TRBD1TRBJ1
rearrangement involving another Vβ segment, there is no evidence
that both of these would be expressed as part of functional TCRβ
genes. The Vβ-to-DβJβ recombination step occurs on one allele at
a time because of asynchronous initiation of Vβ recombination
between alleles and the ability of TCRβ proteins expressed from
one allele to signal feedback inhibition of Vβ recombination on the
other allele. Although Vβ segments only recombine to assembled
DβJβ complexes, it has not been determined whether Dβ-to-Jβ
rearrangements occur on both alleles prior to initiation of Vβ-to-
DβJβ recombination on a single allele.
1.3 Structure The TCRα locus is larger than the TCRβ locus and contains many
and Recombination more gene segments. Moreover, it is polymorphic and structurally
of TCRα Loci distinct in different strains of inbred mice (Fig. 2). Five different
Vα haplotypes were described based on restriction fragment length
polymorphisms [5]. Detailed sequence analysis is available for
strains 129 (which has the same haplotype as BALB/c, AKR,
CBA/J, C3H and A/J) and C57BL/6. The 129 TCRα locus spans
about 1.6 Mb on chromosome 14, whereas the C57BL/6 locus
spans about 1.8 Mb, with the major structural difference mapping
to the array of V gene segments. The 129 locus contains 104 V
gene segments, of which 88 are likely functional, whereas the
C57BL/6 locus contains 138 V gene segments, of which 117 are
likely functional. A complicating feature of the locus is the repeti-
tive nature of the V gene array. In strain 129 mice, there is a 400 kb
duplication in the central portion of the array, with nearly identical
V gene paralogues in each repeat. C57BL/6 mice have copies of
these repeats plus an additional repeat sandwiched between them,
creating even greater complexity. Both strains have unique V gene
segments in non-duplicated regions that flank the repeat region.
The repeat structure makes locus analysis rather challenging.
Downstream of the V gene segments are Dδ, Jδ and Cδ gene
segments that are used to assemble TCRδ genes. Downstream of
Cδ is a single V gene segment in opposite transcriptional orienta-
tion to other gene segments in the locus; this is followed by an
array of 60 Jα gene segments (of which 43 are likely functional as
defined by their ability to undergo recombination and splicing to
create mature TCRα transcripts) and Cα. TCRα rearrangement is
Vα-to-Jα, whereas TCRδ rearrangement is Vδ-to-Dδ-to-Jδ.
TCRα locus nomenclature has been particularly challenging
because there have been several versions; for example, Vα19 was
called AV19 by Arden [6] and is now referred to as TRAV1 by
IMGT and NCBI, which name the Vα gene segment families in
linear order of their first occurrence, from distal to proximal.
Fig. 2 Schematic of TCRα locus. The structures of the 129 and C57BL/6 TCRα loci are presented. The repeat
structure of the V array is noted and the portions shared by 129 and C57BL/6 as well as the portion unique to
C57BL/6 are indicated. Gene segments are listed below each region, from distal to proximal, and are identified
by NCBI and IMGT nomenclature. Gray lettering indicates that genes are non-functional based on the presence
of a stop codon or their inability to rearrange or undergo appropriate mRNA splicing. These designations
diverge somewhat from those in IMGT based on Jα repertoire studies conducted in our laboratory [15–18] as
well as deep sequencing of the peripheral Vα-Jα repertoire [14]. Those V gene segments that are recognized
by IMGT to serve as both a Vα and a Vδ have a dual designation (e.g., 15-1/DV6-1, short for TRAV15-1/
TRDV6-1); those that are recognized to serve exclusively as a Vδ have a single designation (e.g., DV1, short for
TRDV1). Data on TRAV haplotypes were previously reported [5]
The current NCBI nomenclature is compared to the classical Vα

nomenclature in Table 3. Similar comparisons of Jα nomenclatures
are presented in Table 4. Notably, fifteen of the functional V gene
segments in the locus have been classified as Vδ gene segments
(TRDV); some of these serve as both Vα and Vδ gene segments
(e.g., TRAV15-1/DV6-1) whereas others (e.g., TRDV1, TRDV5)
serve exclusively as Vδ gene segments (Fig. 2). All TCRα and
TCRδ gene recombination events occur by deletion except for
those involving the inverted TRDV5 gene segment, which under-

goes recombination with Dδ gene segments by inversion of the
Dδ-Jδ-Cδ region.
1.4 Developmental Mature αβ T cells develop through a differentiation program that

Control of TCRα and β links expression of TCRα and β proteins from productive V(D)J
Recombination rearrangements to further developmental progression (Fig. 3). In
the thymus, CD4−CD8− (double-negative or DN) thymocytes initi-
ate accessibility of TCRβ loci to promote TCRβ recombination [4].
Assembly and expression of functional VβDβJβ rearrangements
leads to TCRβ-mediated intracellular signals that inhibit further Vβ
recombination and promote survival, proliferation, and differentia-
tion of DN thymocytes into CD4+CD8+ (double-positive or DP)
thymocytes. One of the first marks of such TCRβ-selected cells is
up-regulated expression of CD28 (from CD28low DN3a cells to
CD28high DN3b cells) on DN thymocytes prior to their prolifera-
tion and further differentiation [7]. Similarly, up-regulated expres-
sion of CD27 is a mark of TCRβ-selected DN thymocytes [8]. Due
to inherent imprecision in coding join formation, only ~1/3 of
VβDβJβ rearrangements are assembled in-frame and capable of
expressing a functional TCRβ protein. In the ~2/3 of DN cells that
assemble out-of-frame VβDβJβ rearrangements on the first allele,
Vβ recombination can be re-initiated on the other allele in another
attempt to assemble an in-frame VβDβJβ rearrangement. In addi-
tion, on alleles containing non-functional rearrangements to the
Dβ1 cluster, secondary Vβ rearrangements can occur to Dβ2Jβ2
complexes until all possible Vβ-to-DβJβ recombination events are
exhausted. As a result, ~60 % of mature αβ T cells contain VβDβJβ
rearrangements on one allele and DβJβ rearrangements on the
other allele, while ~40 % contain VβDβJβ rearrangements on both
alleles where one is out-of-frame in most cells. Although the
TRBD1TRBJ1 cluster is rearranged on both alleles in all αβ T
Fig. 3 Diagram of αβ T cell development with markers relevant for cell sorting. Critical checkpoints in αβ T cell
development are shown, including surface markers that will differentiate pre- and post-selection DN thymo-
cytes, and pre- and post-selection DP thymocytes
lymphocytes, the TRBD2TRBJ2 cluster remains un-rearranged on

one or both alleles in a significant fraction of cells [9].
Thymic DP cells initiate accessibility of TCRα loci to promote
TCRα recombination [4]. TCRα recombination generally occurs
on both alleles and in multiple rounds on each allele. Initial rear-
rangements are targeted to the most Cα-distal J gene segments.
TRAJ61, although classified as a pseudogene, is the most frequent
target of these primary rearrangements, although Jα segments
immediately downstream are likely targets as well. These primary
rearrangements tend to involve relatively proximal Vα gene seg-
ments. However, it should be noted that newly generated DP thy-
mocytes will have already undergone complete or incomplete
TCRδ gene rearrangement on both chromosomes. Depending on
whether these rearrangements used relatively proximal or distal Vδ
gene segments, DP thymocytes initiating TCRα rearrangements
will have complete, nearly complete, or only partial Vα arrays to
work with. Primary TCRα rearrangements are typically followed
by several additional cycles of Vα-to-Jα rearrangement on both
alleles, using progressively more proximal Jα and more distal Vα
gene segments. The process is terminated either by positive selec-
tion based on the specificity of the assembled TCRαβ, or by DP
thymocyte apoptosis. The average lifespan of DP thymocytes is
thought to be about 3 days. In contrast to the TCRβ locus, TCRα
rearrangement is not subject to allelic exclusion. In keeping with
this, 25–30 % of TCRαβ T cells have productive Vα-to-Jα recom-
bination on both alleles, although a much smaller fraction actually
express two receptors on the cell surface [10–12].
1.5 Analysis PCR is the most commonly used technique for analysis of antigen
of TCRα and β Gene receptor gene rearrangements because it generates large amounts
Rearrangements of information in a short amount of time. Both conventional (see
Subheading 3.1) and real-time (see Subheading 3.2) PCR applica-
tions may be applied to characterize TCRα and TCRβ rearrange-
ments. Additionally, PCR may be coupled with a Southern blot
and probe approach (see Subheading 3.3) to increase sensitivity;
this application is particularly relevant for analysis of TCRα rear-
rangements but can be utilized at either locus.
The general PCR strategy is to amplify TCRα and β gene
rearrangements from genomic DNA using a forward primer that
anneals in a V segment or upstream (5′) of a D segment and a
reverse primer that anneals in or downstream (3′) of a J segment. In
most cases, these primers anneal to genomic sequences too far apart
to amplify products from germline (un-rearranged) alleles, the
exceptions being PCR reactions with 5′Dβ and 3′Jβ primers since
Dβ segments reside ~700 bp upstream of the 5′ Jβ segment in each
cluster. Although PCR reactions on serial dilutions of input DNA
are commonly used to determine the linear range of amplification
and provide semi-quantitative analyses of V(D)J recombination,
real-time PCR approaches can also be used. For essentially compre-

hensive analysis of TCRβ gene rearrangements through multiple
conventional PCR reactions we employ 43 separate PCR reactions
that each use a unique pair of primers. Due to the short distances
between the 5′Jβ and 3′Jβ segment in each cluster, one PCR reac-
tion that pairs a Dβ primer with a single primer downstream of
one of the Jβ clusters can amplify TRBD1 rearrangements to all
six functional TRBJ1 segments, TRBD1 rearrangements to all six
functional TRBJ2 segments or TRBD2 rearrangements to all six
functional TRBJ2 segments. Similar pairing of a primer specific for
any of the 23 individual Vβ segments (or all TRBV12 or TRBV13
family members) with a Jβ1 cluster primer can amplify TRBV-to-
TRBD1TRBJ1 rearrangements, whereas pairing of a Vβ primer
with a Jβ2 cluster primer can amplify TRBV-to-TRBD1TRBJ2,
TRBV-to-TRBD2TRBJ2, or TRBV-to-TRBD1(TRBD2)TRBJ2
rearrangements.
Any attempt at comprehensive analysis of TCRα rearrange-
ments is infinitely more challenging due to the much larger num-
bers of Vα and Jα gene segments and the large expanse of the Jα
array. There are too many Vα-Jα combinations to amplify, and any
individual combination would be present quite infrequently and
would be that much harder to detect. PCR analyses have generally
paired a primer specific for a Vα segment or all members of Vα seg-
ment family with a Jα primer. Individual Jα segments are typically
spaced by 0.5–2.0 kb, making it difficult to amplify many Jα rear-
rangements with a single primer. Nevertheless, by using long PCR
extension times, a combination of 9 different Jα primers has been
used to visualize rearrangements across the entire Jα array using a
blot and probe approach [13]. It is more typical to choose short
PCR extension times to limit the detected rearrangements to pri-
marily a single Jα gene segment. Due to the low abundance of indi-
vidual rearrangements, conventional PCR with direct visualization
by ethidium bromide staining can be challenging. Rather, it is best
to follow conventional PCR by Southern blotting and hybridization
to a Vα- or Jα-specific probe, or to analyze by a more quantitative
real-time PCR. In designing such experiments it is also important to
consider the developmental progression of Vα-to-Jα recombination
events that constrains the TCRα repertoire. As borne out by many
studies [13, 14], the unique proximal Vα gene segments rearrange
most frequently to the most distal Jα gene segments and the unique
distal Vα gene segments rearrange most frequently to Jα gene seg-
ments in the center and proximal half of the Jα array. Multi-member
V gene families that are scattered across the repeat regions tend, as
a group, to rearrange broadly across the Jα array. Thus a TRAV19
primer might logically be paired with a TRAJ58 primer, a TRAV1
primer might logically be paired with a TRAJ30 primer, and a
TRAV12 family primer could be paired with almost any Jα primer.
Common primers that we have used to amplify TCRβ and α
gene rearrangements are listed in Tables 1 and 2 or Tables 3 and 4,
188
Levi J. Rupp et al.
Table 2
Primers for TCRβ join PCRs including oligos 5′ of Dβ and 3′ of Jβ segments
Primer name Sequence Location Southern blot probe

Dβ1 GAGGAGCAGCTTATCTGGTGGTTT Upstream TRBD1
Dβ2 GTAGGCACCTGTGGGGAAGAAACT Upstream TRBD2
Jβ1.2 CCTGACTTCCACCCGAGGTT Downstream TRBJ1.2 CATCCTTCCTCTGATTAC [20]
Jβ1.7 AAGGGACGACTCTGTCTT Downstream TRBJ1.7 GATAAGACAGAATCCTTAGG
Jβ2.2 CTCCAACCCTGACTCAGATCCCCACC Downstream TRBJ2.2 GGTGGGGATCTGAGTCAGGGTTG
Jβ2.7 TGAGAGCTGTCTCCTACTATCGATT Downstream TRBJ2.7 TATGAACAGTACTTCGGTCCC [19]
Table 3
Current NCBI and historical nomenclature for representative TCRα locus
variable (V) region segments

nomenclature nomenclature primers
TRAV1 Vα19 CTATCTCTTCCTGATGGTGG (C)
TRAV2 Vα12
TRAV3 Vα5
TRAV4 Vα11
TRAV5 Vα13
TRAV6 Vα4 GAAGCAGCAGAGGGTTTGAAG (R)
TRAV7 Vα1
TRAV8 Vα18
TRAV9 Vα3
TRAV10 Vα15
TRAV11 Vα14 GTCCTCAGTCCCTGGTTGTC (C)
TRAV12 Vα8 GCAGCAGCTCCTTCCATC (R)
CAGACAGAAGGCCTGGTCAC (C)
TRAV13 Vα10 AAGAACGTCGCAGCTCTTTG (R)
TRAV14 Vα2 TGGAGACTCAGCCACCTACT (R)
TRAV15 Vα7 TCCATCAGCCTTGTCATTTCA (C,R)
TRAV16 Vα17 TGCAACAGTGGGTCATTATTCT (R)
TRAV17 Vα9 TGGAGCGACTCAGCCAAGTA (R)
TRAV19 Vα20 CCAGCCTGAAGACACAGCAG (R)
TRAV21 Vα6 GTATGGCTTTCCTGGCTATTGC (R)
Also contains Vα primer sequences for join PCRs. (C) and (R) identify primers that have
been used in our laboratory for conventional and real-time PCR, respectively. In gen-
eral, V gene segment primers for conventional PCR are chosen to create amplicons of
approximately 300 bp, whereas V gene segment primers for real-time PCR are chosen
to create amplicons of approximately 100 bp. Primers for multi-gene families were cho-
sen so that they amplify all members of a given family
respectively. The two major applications of PCR analysis of TCRα

and β gene rearrangements are to quantify defects in V(D)J recom-
bination (rates, efficiency, levels) and changes in repertoire. Below
we outline certain limitations of PCR analysis of TCRα and β gene
rearrangements that must be accounted for in experimental design
and data interpretation.
Table 4
190
Current NCBI and historical nomenclature for TCRα locus joining (J) region segments
NCBI Historical
nomenclature nomenclature Join PCR primers Southern blot probe
TRAJ61 Jα61 ATGAGTCTTCCAGTCATGGC (R)
TRAJ58 Jα58 GACTCACTGTGAGCTTTGCC (R,C) AGCTTAGACCCAGTGCCTTG
(R,C)
Levi J. Rupp et al.
TRAJ57 Jα57 AGCTCACTGTCAGCTTTGTCC AAAGATGAGCTTCGCAGACC

(R,C)
TRAJ56 Jα56 ACTCAGAACGGTTCCTTGACC AGCTGACTTTTGGTCAAGGAACCG
(R,C)
TRAJ53 Jα53 GGAGTCACAGTTAAGAGAGTTCC AATTGCTGCCTCCACTGTTC
(C)
TRAJ52 Jα52 TTGGATGGACCCTAAGGATG TGGAGCTAACACTGGAAAGCTC
(C)
TRAJ50 Jα50 ACGACTGATAAGGATGTCCC TTCAGCAAGCTGGTGTTTGG
(R,C)
TRAJ49 Jα49 GGAATGACAGTCAAACTTGTTCC AGAAGTTCTGGTAACCCCGTG
TRAJ48 Jα48
TRAJ47 Jα47
TRAJ45 Jα45
TRAJ44 Jα44
TRAJ43 Jα43
TRAJ42 Jα42 AGAGTTTAGTGCCTTTCCCG (C) TCTGGAGGAAGCAATGCAAAG
TRAJ40 Jα40 TGGTACCTGCTCCAAAGACG (R)
TRAJ39 Jα39
TRAJ38 Jα38
TRAJ37 Jα37 AAATGAGCATAAGCGACAG (R)
TRAJ35 Jα35
TRAJ34 Jα34
TRAJ33 Jα33
TRAJ32 Jα32
TRAJ31 Jα31 GCGTCCCATCACCAAAGAAG (R)
TRAJ30 Jα30 GGGAGAACATGAAGATGTGTCC (R)
TRAJ28 Jα28
TRAJ27 Jα27
TRAJ26 Jα26
TRAJ24 Jα24
TRAJ23 Jα23
TRAJ22 Jα22 TGTCAGTTGGGTTCCAGATCC (C) TCCAAAGATGAGTTGCCAGC
TRAJ21 Jα21
TRAJ18 Jα18 TCCCTGGAGTAGAAAGAAACCTACTCAC (C) CAGGTATGACAATCAGCTGAGTCCC
(R)
TRAJ17 Jα17 TGATGGCTAGGCTCCTTTTC
TRAJ16 Jα16
TRAJ15 Jα15
TRAJ13 Jα13
TRAJ12 Jα12
TRAJ11 Jα11
TRAJ9 Jα9
TRAJ7 Jα7
TRAJ6 Jα6
TRAJ5 Jα5
TRAJ4 Jα4
Molecular Analysis of Mouse T Cell Receptor α and β Gene Rearrangements
TRAJ2 Jα2 TACCGGGTTGCAAATGGTG (R)

Also contains reverse primers for join PCRs and oligonucleotide sequences for Southern blot probes
191
The fact that a TCRβ and a TCRα gene rearrangement are

both required for αβ T cell development must be accounted for in
PCR analyses to assay for and quantify defects in V(D)J recombina-
tion. In αβ T cells that develop from normal V(D)J recombination,
~60 % have a VβDβJβ rearrangement on one allele and a DβJβ rear-
rangement on the other, ~40 % have VβDβJβ rearrangements on
both alleles, and nearly 100 % have VαJα rearrangements on both
alleles. For TCRβ, this translates to VβDβJβ rearrangements on
70 % of alleles and DβJβ rearrangements on 30 % of alleles in the
entire population. While a complete block in V(D)J recombination
prevents αβ T cell development, substantially impaired V(D)J
recombination could lead to the selection of αβ T cells with reduced
frequencies of TCRα and β gene rearrangements. However, there
must be at least one fully rearranged allele at each locus. Thus,
PCR analysis (even if highly quantitative) would detect only small
differences in the amount of TCRβ and TCRα recombination
(maximally 30 and 50 %, respectively) in selected αβ T cells. Such
defects may be substantially greater when measured in non-
selected, immature thymocyte populations that are actually engaged
in V(D)J recombination. Defects in TCRα recombination may also
be detected as diminished usage of more Cα-proximal as compared
to Cα-distal Jα gene segments, since the former require more
rounds of Vα-to-Jα recombination than the latter. Additionally, the
distribution of Vβ and Jα segments used may also be skewed by
selection. Therefore, PCR analysis of TCRβ and TCRα gene rear-
rangements to assay for and quantify potential defects in V(D)J
recombination must be conducted on non-selected DN and DP
thymocytes isolated by flow cytometry-based cell sorting with
appropriate cell surface markers. Common cell surface markers
used to isolate non-selected DN and DP thymocytes are indicated
in Table 5 and Fig. 3. Of course, if the goal is to analyze perturba-
tions that affect selection of the TCR repertoire in the thymus, or
expansion of particular T cell subsets in the periphery, then PCR
analyses should be conducted on either selected single positive
(SP) thymocytes or mature peripheral αβ T cells, as appropriate.
Table 5
Cell surface markers for isolation of pre- and post-selection DN or DP
thymocytes
Cell population Gating scheme

DN3a CD4−CD8−CD44−CD25+CD28low
DN3b CD4−CD8−CD44−CD25+CD28high
Pre-selection DP CD4+CD8+TCRβloCD24+CD69−
Post-selection DP CD4+CD8+TCRβloCD24+CD69+
2 Materials
2.1 Conventional All PCRs should use molecular biology grade H2O, while standard
PCR tap water is acceptable for running buffers.
1. diH2O—molecular biology grade.
2. Cells of interest (see Note 1).
3. Rapid Lysis Buffer: 100 mM Tris-base pH 7.4, 0.2 % sodium
dodecyl sulfate, 5 mM EDTA, 200 mM NaCl.
4. Proteinase K. Re-suspended at 20 mg/mL in diH2O.
5. Isopropanol.
6. TE buffer pH 8.0: 10 mM Tris-base pH 8.0, 1 mM EDTA in
diH2O.
7. 10× PCR reaction buffer (as supplied by DNA polymerase
manufacturer).
8. DNA polymerase (see Note 2).
9. 10 mM dNTP mixture (2.5 mM each dATP, dTTP, dCTP,
dGTP), available commercially.
10. Primers of interest (see Tables 1, 2, 3, and 4). Re-suspend
primers at 5 pmol/μL using diH2O.
11. CD14 control primers. Sense primer: 5′-GCTCAAACTTTCA
GAATCTAC-3′; Antisense primer: 5′-AGTCAGTTCGTGGA
GGCCGGAAATC-3′; if following with Southern blot
(Subheading 3.3), probe: 5′-AGCAGATCTGGGGCAGTT
CACTGA-3′.
12. Standard thermocycler.
13. 1× TBE buffer: 89 mM Tris-base, 89 mM boric acid, 2 mM
EDTA in H2O. To make a 5× stock solution: Dissolve 1080 g
Tris-base, 550 g boric acid, and 400 mL 0.5 M EDTA in 5 L
H2O. Adjust final volume to 20 L, and dilute to 1× in H2O
prior to use. Adjustment to precise pH is not necessary at any
point.
14. Agarose.
15. Ethidium bromide (10 mg/mL stock solution).
16. 6× Gel Loading Dye.
17. 1 kb DNA ladder.
2.2 Quantitative 1. diH2O—molecular biology grade.

Real-Time PCR 2. Primers of interest for join detection (see Tables 3 and 4).
Prepare 10 mM working stock in diH2O.
3. 2× SYBR Green PCR Master Mix.
4. Genomic DNA of interest (50 ng/μL).
5. Real-time PCR instrument.
2.3 Southern Blot Molecular biology grade diH2O should be utilized for steps involv-
ing DNA while standard diH2O is suitable for running buffer,
transfer buffer, hybridization solution, and washing buffers.
1. Join PCR product from Subheading 3.1.
2. 1× TAE buffer: 40 mM Tris-base, 0.11 % glacial acetic acid
(v/v), 1 mM EDTA in H2O. To make a 50× TAE stock solu-
tion: Dissolve 968 g Tris-base in 2 L H2O. Add 228.4 mL gla-
cial acetic acid and 400 mL 0.5 M EDTA solution. Adjust final
volume to 4 L in H2O. Dilute to 1× in H2O prior to use.
Adjustment to precise pH is not necessary at any point.
3. Agarose.
4. Ethidium bromide (10 mg/mL stock solution).
5. DNA ladder (see Note 3).
6. UV source (gel box or wand).
7. Denaturing solution: 0.5 M NaOH, 0.6 M NaCl in diH2O.
8. Tupperware (see Note 4).
9. Orbital shaker such as Stovall Belly Dancer™.
10. Sponges (see Note 5).
11. Blotting paper (see Note 6).
12. 10× SSC: 1.5 M NaCl, 150 mM trisodium citrate in diH2O. To
make a 20× SSC stock solution: Dissolve 3506 g NaCl and
1764 g sodium citrate in a final volume of 20 L H2O. Dilute
as necessary in H2O prior to use. Adjustment to precise pH is
not necessary at any point.
13. Nylon membrane (see Note 7).
14. Paper towels.
15. Parafilm.
16. Disposable pipette.
17. Gel box and electrophoresis power source.
18. Drying oven.
19. 19× SSCPE: 2.85 M NaCl, 285 mM trisodium citrate,
247 mM KH2PO4, and 19 mM EDTA in H2O. Adjustment to
precise pH is not necessary.
20. Oligo pre-hybridization solution: 6× SSCPE, 5× Denhardt’s
solution, 1 % SDS in H2O. Prepare 100 mL by combining the
following: 31.5 mL 19× SSCPE, 10 mL 50× Denhardt’s solu-
tion, 10 mL 10 % SDS solution, and 48.5 mL diH2O. Adjustment
to precise pH is not necessary.
21. Oligo hybridization solution: 6× SSCPE, 3× Denhardt’s solu-
tion, 1 % SDS, 100 μg/mL yeast tRNA in H2O. Prepare
100 mL by combining the following: 31.5 mL 19× SSCPE,
6 mL 50× Denhardt’s solution, 10 mL 10 % SDS solution,
0.4 mL yeast tRNA (25 mg/mL stock, Invitrogen Catalog

15401-011), and 52.5 mL diH2O. Adjustment to precise pH
is not necessary.
22. Hybridization oven set to 42 °C.
23. Probe labeling reagents.
(a) 50–100 pmol oligo DNA (<6 μL total volume). See
Table 2 or 4 for oligo sequences, including CD14 loading
control.
(b) T4 polynucleotide kinase (10 U/μL).
(c) γ32P ATP, 6000 Ci/mmol, 150 mCi/mL [e.g. Perkin
Elmer Catalog: NEG-035C].
(d) Heating block set to 37 °C.
(e) TE buffer, pH 8.0 (see Conventional PCR, Subheading 2.1,
Item 6).
24. Wash buffer: 2× SSC, 0.1 % SDS in H2O. To prepare 4 L Wash
buffer, combine 400 mL 20× SSC with 40 mL 10 % SDS and
adjust volume to 4 L with H2O. Adjustment to precise pH is
not necessary.
25. Heated water bath shaker (see Note 8).
26. Funnel.
27. Glass roller bottle.
28. Saran wrap.
29. Film (see Note 9).
30. Autoradiography cassette.
31. Film developer.
3 Methods
3.1 Conventional 1. Isolate genomic DNA by method of choice. We utilize the lysis
PCR and isopropanol precipitation protocol described in steps 2–5.
2. Pellet 104–107 cells of interest in 1.7 mL eppendorf tube.
Re-suspend cells in 500 μL rapid lysis buffer (see Note 10)
with 5 μL Proteinase K (20 mg/mL stock). Incubate over-
night at 56 °C.
3. Add 800 μL isopropanol to Eppendorf tube containing lysed
cells. Mix thoroughly by shaking.
4. Centrifuge 10 min at maximum speed in bench-top centrifuge
to pellet DNA.
5. Aspirate supernatant. Re-suspend DNA in appropriate volume
TE buffer or diH2O (see Note 11) and place at 56 °C for 2 h.
6. Determine DNA concentration by method of choice. Dilute
an aliquot to 50 ng/μL in diH2O.
7. Prepare 25 μL PCR reaction as follows, using desired primers

from Tables 1, 2, 3, and 4. When performing multiple reac-
tions, preparing a master mix of all components except
genomic DNA is recommended:
2.5 μL 10× PCR buffer

0.5 μL dNTPs (10 mM)
1 μL Forward primer (5 pmol/μL)
1 μL Reverse primer (5 pmol/μL)
0.2 μL DNA polymerase (5 U/μL)
9.8 μL diH2O
10 μL Genomic DNA (50 ng/μL)
8. Include a control reaction for a non-rearranging region, such

as CD14.
9. Perform PCR amplification using a standard thermocycler and
the appropriate program for the locus being analyzed:
TCRβ
(a) 94° for 3 min.
(b) 94° for 45 s.
(c) 60° for 1 min 30 s.
(d) 72° for 2 min 30 s.
(e) 72° for 10 min.
TCRα
(a) 95° for 3 min.
(b) 95° for 30 s.
(c) 63° for 30 s.
(d) 72° for 30 s.
(e) 72° for 3 min.
Repeat steps b–d for 32–40 cycles total (see Note 12).
10. Analyze PCR products by standard agarose gel electrophoresis

with either 1 % (TCRβ) or 1.2 % (TCRα) agarose gel in TBE.
3.2 Quantitative 1. Isolate genomic DNA by the protocol described in

Real-Time PCR Subheading 3.1, steps 2–5.
2. Prepare 12 μL PCR reactions in triplicate as follows, using
desired primers from Tables 3 and 4. Include control wells for
a non-rearranging standard (such as CD14).
6 μL 2× SYBR Green PCR Master Mix

0.3 μL Forward primer (10 mM)
0.3 μL Reverse primer (10 mM)
3.4 μL diH2O
2 μL Genomic DNA (50 ng/μL)
3. Perform PCR amplification on real-time PCR instrument.

(a) 95 °C for 5 min.
(b) 95 °C for 10 s.
(c) 65 °C for 30 s.
Repeat steps b and c for 45 cycles.
4. Determine relative quantification of Vα-Jα rearrangements in

different samples by normalization to CD14 signal or non-
rearranging genetic region of your choice (see Note 13).
3.3 Southern Blot Unless otherwise described, all procedures may be carried out at
room temperature. All solutions may be stored at room tempera-
ture, but we store the oligo pre-hybridization solution, oligo
hybridization solution, and wash buffer at 42 °C for ease of use. All
steps involving radiation should be performed behind a radiation
shield and according to relevant institutional guidelines. We utilize
a dedicated sink, 37 °C heating block, and heated shaker for radia-
tion purposes.
1. Prepare running gel by dissolving 4 g agarose in 500 mL 1×
TAE with ethidium bromide (5 μL of stock 10 mg/mL solu-
tion). Combs should be sized to permit 50 μL loading volume
(see Note 14).
2. Load samples on running gel, leaving the outer two lanes on
each side empty, as these will be lost in transfer. Load one lane
of DNA ladder (see Note 15).
3. Run gel overnight (18–24 h) at 23 V.
4. Remove gel from gel box and examine by UV to verify sam-
ples have amplified (see Note 16).
5. Activate the gel by treating for 30 s with short-wave UV irra-
diation (~254 nm) using wand or gel box, with appropriate
shielding.
6. Fill a Tupperware container with 1 L of denaturing solution.
Place gel in Tupperware. Shake gently on orbital shaker for
1 h at room temperature (see Note 17).
7. Prepare for transfer. Place a series of sponges (large enough to
fully support the gel) side-by-side in large Tupperware con-
tainer. Lay one sheet of blotting paper over the sponges,
extending blotting paper over the edge of sponges. Add 10×

SSC to the Tupperware, filling until buffer level is ½″ below
top of sponges.
8. Activate membrane. Cut an appropriately sized piece of nylon
membrane and wet in Tupperware containing diH2O. Remove
membrane from water, allowing excess water to drain, and
place membrane in a separate Tupperware containing 10×
SSC for 1 h at room temperature.
9. Cut two pieces of blotting paper sized to fit the gel.
10. Set-up transfer as follows. Carefully remove gel from
Tupperware containing denaturing solution and place on top
of sponges/blotting paper prepared in step 7 (see Notes 18
and 19). Place activated nylon membrane on top of gel and
roll out bubbles using a disposable pipette. Place parafilm
around the edge of the membrane in a “framing” pattern.
Parafilm should cover the outer ¼″ of the membrane on all
sides. Wet one piece blotting paper in 10× SSC and place on
top of parafilm layer. Roll out bubbles as before. Place a dry
piece of blotting paper on top of the wet first piece. Place
paper towels on top of this blotting paper. Place a weight (such
as an eppendorf rack) on top of paper towels. The final order,
from bottom to top, is: sponge, blotting paper, gel, nylon
membrane, parafilm, wet blotting paper, dry blotting paper,
paper towels. Transfer overnight (~20 h), even though trans-
fer rate is ~1 kb per hour.
11. Disassemble transfer apparatus. Place membrane between
folds of larger piece of blotting paper and dry in vacuum oven,
1–1.5 h at ~110 °C (see Note 20).
12. Label the membrane and mark the ladder using an industrial
style pen to prevent label washing off during subsequent
washes. Ladder can be visualized by examining membrane
under UV light.
13. Perform pre-hybridization by placing rolled membrane into
glass roller bottle. Add 12 mL pre-warmed oligo pre-
hybridization solution to the tube, place in hybridization oven
and rotate for minimum of 1 h at 42 °C.
14. Prepare radiolabeled oligo probe per steps 15–17.
15. To 1.7 mL eppendorf tube, add 50–100 pmol oligo (total
volume ≤6 μL), 1 μL 10× kinase buffer, 1 μL T4 polynucleo-
tide kinase (10 U/μL), and 2 μL γ32P ATP. Bring total volume
to 10 μL with diH2O.
16. Incubate for 30 min at 37 °C.
17. Add 90 μL buffer TE pH 8.0.
18. After minimum 1 h incubation, discard pre-hybridization
solution from roller bottle of step 13. In a 15 mL conical, add
probe solution from step 17 to 13 mL hybridization solution

pre-warmed to 42 °C. Add immediately to roller bottle con-
taining membrane that has been pre-hybridized. Hybridize
overnight at 42 °C.
19. Adjust heated shaker temperature to 42 °C and allow tem-
perature to stabilize.
20. Using a funnel, empty roller bottle contents by pouring radio-
active probe into 15 mL conical tube. Cap tube containing
radioactive probe, then wash and dry before returning to
radiation storage lockbox if planning to re-use probe.
Otherwise, discard per appropriate protocol. Dispose of all
radioactive materials according to institutional guidelines.
21. Add pre-warmed wash buffer to roller bottle containing mem-
brane and swirl several times. Open roller bottle over sink and
remove membrane using forceps. Place in Tupperware con-
taining sufficient wash buffer (pre-warmed to 42 °C) to cover
membrane. Seal tupperware and place in heated shaker at
42 °C. Wash with shaking for 15 min. Rinse out roller bottles
in sink and decontaminate per standard protocol.
22. After 15 min wash, discard wash buffer in sink. Add fresh wash
buffer and repeat, washing three times total (see Note 21).
23. Remove membrane from last wash and dry by touching the
corner to a paper towel. Place on saran wrap and fold saran
wrap over blot, completely enclosing. Fix blot in autoradiog-
raphy cassette using tape. In dark room, place film in autoradi-
ography cassette. Expose per manufacturer’s protocol.
24. Develop film per standard protocol.
25. Optional: stripping blots. Often it is desirable to probe a single
blot with multiple different probes, for example one probe
detecting a rearrangement of interest and the second probe a
loading control such as CD14. To strip a blot, boil 1 L of H2O
in a beaker. Place blot in Tupperware container and pour boil-
ing water over blot, swirling for 30 s to 1 min. Discard water
and rinse blot until background radiation levels are detected
on the Geiger counter.
4 Notes
1. Possible populations include bulk thymocytes, pre- and post-

selection thymocytes isolated by flow cytometry, or mature T
cells.
2. We have had success with Qiagen HotStarTaq Polymerase but
others may be suitable.
3. We use TrackIt λ DNA/HindIII Fragments (Invitrogen).
4. A range of Tupperware/plastic containers are useful for this

protocol and should be sized to fit the anticipated gel size.
5. We utilize kitchen sponges ~1 in. thick. These may be cut to
size as desired.
6. We recommend VWR Grade 703.
7. We have had success with Zeta Probe GT Genomic Tested
Blotting Membranes (Bio-Rad).
8. Analogous to Bellco Catalog 7746-22110.
9. For example, HyBlot CL®, available from Denville Scientific
or other suppliers.
10. If using more cells, the lysis buffer volume may be increased
accordingly. Isopropanol in the following step should be scaled
to maintain a similar ratio of lysis buffer:isopropanol.
11. For low cell numbers (104–105 cells), re-suspend in
20–50 μL. Larger cell numbers can be re-suspended in 200 μL
or sufficient volume to fully re-suspend pellet and obtain accu-
rate DNA quantification. It is better to start with a smaller
volume and dilute as necessary.
12. For visualizing products by ethidium bromide staining, high
cycle numbers (e.g. 40 cycles) should be used. However, the
large amount of input DNA and high cycle number mean that
the reaction is not quantitative. To obtain a rough quantitative
measure of rearrangement levels one can perform serial two-
or fivefold dilutions with low cycle number (e.g. 32 cycles),
followed by Southern blotting because low cycle number
often makes products difficult to visualize by ethidium bro-
mide staining.
13. The efficiency of primer pairs should be validated for your buf-
fer system/instrument. To do this, perform qPCR on serial
dilutions of bulk thymic DNA or plasmid containing the rear-
rangement of interest. Incorporate this efficiency value into
relative quantification (ΔΔCt) calculations.
14. The two outermost lanes on each side will be lost in transfer,
so plan gel loading accordingly.
15. Often times it is necessary to examine multiple Jα probes to
obtain a complete picture of recombination. Rather than strip-
ping and re-probing a blot multiple times, the same series of
samples can be loaded two or more times on a single gel and
the membrane cut after transfer. In this case, leave several
empty wells between each sample, and be sure to load a ladder
with each set.
16. In some cases the control PCR reaction will fail to run. In this
case it is not worth proceeding, and the PCR conditions
should be optimized before repeating the protocol.
17. Multiple gels may be shaken in a single container, simply place

on top of each other.
18. Gels will be very brittle after denaturing. Take care to handle
gently. If the gel does crack or split, it is sometimes possible to
piece it together and continue with the transfer.
19. Denaturing solution may be re-used three times, after which it
should be discarded.
20. After drying the membrane is stable and may be stored at
room temperature indefinitely.
21. If the blot is particularly “hot,” additional washes can be per-
formed prior to exposure. Typically, if the blot is washed until
the Geiger counter gives a reading of 2–3000 cpm, an over-
night exposure will be sufficient. If the blot is significantly
“hotter,” additional washes are typically useful. Note, how-
ever, that the appropriate intensity of radiation and duration of
exposure must be determined empirically for each probe.
Acknowledgements
This work was supported by NIH grants T32 HG000046 (LJR),

R37 GM41052 and R01 AI49934 (LC and MSK), R01 CA125195
and R01 CA136470 (CHB), and a Leukemia and Lymphoma
Society Scholar Award (CHB).
References
1. Schatz DG, Swanson PC (2011) V(D)J recom- 7. Williams JA, Hathcock KS, Klug D et al (2005)
bination: mechanisms of initiation. Annu Rev Regulated costimulation in the thymus is criti-
Genet 45:167–202 cal for T cell development: dysregulated CD28
2. Bassing CH, Swat W, Alt FW (2002) The costimulation can bypass the pre-TCR check-
mechanism and regulation of chromosomal point. J Immunol 175:4199–4207
V(D)J recombination. Cell 109:S44–S55 8. Taghon T, Yui MA, Pant R et al (2006)
3. Sleckman BP, Gorman JR, Alt FW (1996) Developmental and molecular characterization
Accessibility control of antigen-receptor of emerging β- and γδ-selected pre-T cells in
variable-region gene assembly: role of cis-acting the adult mouse thymus. Immunity 24:53–64
elements. Annu Rev Immunol 14:459–481 9. Khor B, Sleckman BP (2002) Allelic exclusion
4. Krangel MS (2009) Mechanics of T cell recep- at the TCRβ locus. Curr Opin Immunol
tor gene rearrangement. Curr Opin Immunol 14:230–234
21:133–139 10. Malissen M, Trucy J, Jouvin-Marche E et al
5. Jouvin-Marche E, Morgado MG, Trede N (1992) Regulation of TCRα and β gene allelic
et al (1989) Complexity, polymorphism, and exclusion during T-cell development. Immunol
recombination of mouse T-cell receptor α gene Today 13:315–322
families. Immunogenetics 30:99–104 11. Gascoigne NR, Alam SM (1999) Allelic exclu-
6. Arden B, Clark SP, Kabelitz D, Mak TW (1995) sion of the T cell receptor α-chain: develop-
Mouse T-cell receptor variable gene segment mental regulation of a post-translational event.
families. Immunogenetics 42:501–530 Semin Immunol 11:337–347
12. Niederberger N, Holmberg K, Alam SM et al 16. Hawwari A, Bock C, Krangel MS (2005)

(2003) Allelic exclusion of the TCR α-chain Regulation of T cell receptor α gene assembly
is an active process requiring TCR-mediated by a complex hierarchy of germline Jα promot-
signaling and c-Cbl. J Immunol 170: ers. Nat Immunol 6:481–489
4557–4563 17. Abarrategui I, Krangel MS (2007) Noncoding
13. Pasqual N, Gallagher M, Aude-Garcia C et al transcription controls downstream promoters
(2002) Quantitative and qualitative changes in to regulate T-cell receptor α recombination.
V-Jα rearrangements during mouse thymo- EMBO J 26:4380–4390
cytes differentiation: implication for a limited 18. Hawwari A, Krangel MS (2007) Role for rear-
T cell receptor α chain repertoire. J Exp Med ranged variable gene segments in directing sec-
196:1163–1173 ondary T cell receptor α recombination. Proc
14. Genolet R, Stevenson BJ, Farinelli L et al Natl Acad Sci U S A 104:903–907
(2012) Highly diverse TCRα chain repertoire 19. Gärtner F, Alt FW, Monroe R et al (1999)
of pre-immune CD8(+) T cells reveals new Immature thymocytes employ distinct signaling
insights in gene recombination. EMBO J 31: pathways for allelic exclusion versus differentia-
1666–1678 tion and expansion. Immunity 10:537–546
15. Guo J, Hawwari A, Li H et al (2002) 20. Bassing CH, Alt FW, Hughes MM et al (2000)
Regulation of the TCRα repertoire by the sur- Recombination signal sequences restrict chro-
vival window of CD4 + CD8+ thymocytes. Nat mosomal V(D)J recombination beyond the
Immunol 3:469–476 12/23 rule. Nature 405:583–586
Chapter 17
Intrathymic Injection
Sugata Manna and Avinash Bhandoola
Abstract
Intrathymic injection is used in several T cell-associated immunological studies to deliver cells or other
substances directly into the thymus. Here, we describe the intrathymic injection procedure involving surgi-
cal incision of the mouse with or without a thoracotomy. Though this procedure can result in poor recov-
ery, postsurgical complications, and distress to the animal, it is actually a simple procedure that can be
carried out relatively easily and quickly with experience.
Key words Thymus, Mouse, Intrathymic injection, Thoracotomy, Surgical incision
1 Introduction
Thymus is the essential organ for differentiation and selection of

immunocompetent T lymphocytes. It comprises two distinct, sym-
metrical pyramidal lobes, which are in close contact and connected
by an isthmus of connective tissue. This specialized primary lym-
phoid organ is present in all jawed vertebrates, and is located ante-
rior to the heart and posterior to the sternum in the midline of the
thoracic cavity (in the anterior superior mediastinum). Thymus is
the first lymphoid organ to develop and in mice, its size is maximal
just before puberty followed by a gradual decrease of size and func-
tion along with age (age-associated involution) [1].
Intrathymic injection is used to study questions of tolerance
induction [2–4], tumor induction [5], cell transplantation [6],
T cell-specific gene therapy [7, 8], and T cell development [9–12]
in mice. This process can be accomplished by several means includ-
ing surgical incision of the sternum for direct visualization, acces-
sibility, and injection into the thymus [6–16]; a blind approach
following a small skin incision [17, 18] or directly into the thoracic
cavity [19]; or an ultrasound-guided approach [20, 21]. As intra-
thymic injection involving thoracic surgery may cause pain and
distress and result in postoperative complications, the latter less
203
204 Sugata Manna and Avinash Bhandoola
invasive techniques, not requiring a thoracotomy, have been devel-

oped. While success rates can be low by the blind injection meth-
ods, ultrasound-mediated guidance can alleviate these issues,
allowing visualization of both the needle and the injected material
as it enters the thymus. Nevertheless, many researchers may not
have access to ultrasound equipment for the implementation of
this procedure. Here, we describe the techniques involving thora-
cotomy as well as the blind-injection approach involving a small
skin incision.
2 Materials
1. Sterile drape.
2. Sterile latex surgical gloves.
3. Sterile gauge sponge or swab.
4. 1 cc insulin syringe, with 29 1/2 G ultrafine needle.
5. Betadine.
6. Surgical board.
7. Rubber bands.
8. Eye moisture salve.
9. Forceps, scissors appropriate for incision.
10. Wound clip.
11. 10 μL syringe (Hamilton Co; Reno, NV), attached with a
27 G needle (no dead volume).
12. Anesthetics: Ketamine and xylazine (see Notes 1 and 2).
Ketamine and xylazine should be diluted in 1× PBS. Ketamine/
xylazine anesthetic should be administered at a dose of
0.1 mg/g mouse weight and 0.02 mg/g mouse weight for
ketamine and xylazine, respectively.
3 Procedures
3.1 Intrathymic 1. Anesthetize the mice with ketamine and xylazine adminis-
Injection with tration at a dose/volume of 0.01 mL/g mouse weight
Thoracotomy (~200 μL for a 20 g mouse) via intraperitoneal injection
(use 1 cc insulin syringe, with 29 1/2 G ultrafine needle)
(see Notes 3 and 4).
2. Rinse the work area with betadine and lay down a sterile drape.
3. Place the anesthetized mouse on its back on a surgical board
and immobilize it by strapping its feet with rubber bands
(Fig. 1a) (see Note 5).
Intrathymic Injection 205
Fig. 1 (a) Anesthetized mouse placed and immobilized with rubber bands on surgical board. (b) Midline longi-
tudinal incision, with mouse’s rib cage cut for visualization of thymus. (c) Intrathymic injection, mouse chest
held open with forceps and needle with bevel up, is positioned and inserted into the thymus
4. Loosely stretch an additional rubber band across the mouth to

hold back the head and gently pull out the tongue with forceps
so that the mouse does not asphyxiate.
5. Place the mouse or board so that the head of the animal is
toward you.
6. Apply eye moisture salve to each eye to prevent the cornea

from drying out.
7. Swab the chest and neck area with betadine.
8. Pinch skin at upper thoracic region with forceps and make a
small longitudinal midline cut with a fine delicate scissors.
Continue incision through the skin with scissors; make one
continuous incision up to the xiphoid process visible as white
“V” under the skin (from the maxillary to the middle of the
rib cage).
9. Separate skin along fascial plane on either side of incision by
gently inserting forceps jaw underneath the skin, and spread
the cut skin outward on each side with forceps, creating two
“flaps” to expose the sternum.
10. Carefully lift the salivary gland, lying between the larynx and
sternal notch, with forceps, make a single cut in the connective
tissue attaching it to the rib cage (at the end pointing away
from you), and pull the gland superiorly in your direction to
visualize the top of the rib cage and trachea, taking care not to
tear them (see Note 6).
11. Using a clean high-quality fine forceps, very carefully pinch
thin muscle lying on top of the trachea. Once pinched, do not
let it go, and pull muscle inferiorly as far as possible (it should
eventually tear) to reveal a small invagination at the top of the
rib cage.
12. By introducing a scissors into the invagination make a verti-
cally oriented 3-mm incision (with an upward movement)
down through the sternum to bisect the upper sternum at the
centerline (slightly to the right, at the level of first two ribs).
Gently spread the opened ribcage sideways using the tip of
blunt, curved forceps to expose/reveal the thymus, the milky
white-translucent-colored organ pulsing through the opening
(Fig. 1b) (see Notes 7 and 8).
13. Fill 10 μL Hamilton syringe (with a 27 G needle) with 10 μL
cell suspension and remove air bubbles.
14. Maintain the split open by pushing it to the side with forceps
and with the other hand insert the needle, bevel up, into the
parenchyma of the thymus, 2–3 mm under the thymic capsule.
Inject 10 μL of cell suspension or solution in the lobe and
withdraw the needle carefully to minimize the backflow. If
needed repeat the injection for the other lobe (Fig. 1c) (see
Notes 9–11).
15. Put back down the salivary gland in place, to block the rib
cage opening. Release the mouse of rubber bands. Hold the
two flaps of skin together with forceps and close the wound by
stapling the skin gently with two to three wound clips.
16. Return mouse to the cage; check to ensure that tongue is still
out. Place subsequent mice next to previous ones to conserve
heat until they recover from anesthesia (may take up to 2 h for
complete recovery) (see Notes 12 and 13).
3.2 Intrathymic Intrathymic injection can be performed following intercostal

Injection Without injection, a minimally invasive technique, without thoracic cavity
Thoracotomy exposure.
1. Anesthetize and immobilize the mice as described above
(steps 1–7).
2. Make a small (4–5 mm) midline incision in the skin above the
sternum. Alternatively, you can make a transverse incision in
the skin near the second intercostal space, perpendicular to the
sternum (Fig. 2a).
3. Spread the cut skin outward with a forceps.
4. Position the needle between the third and fourth ribs, at
~30°–40° angle relative to the sternum. With needle vertex
pointing toward the manubrium, inject the thymus lobe
through the thoracic wall with 10 μL of cells or solution (if the
needle does not pass through the thymus, the injection should
be successful) (Fig. 2b).
5. The estimated injection depth is approximately 2–3 mm and
both thymic lobes can be accessed through the intercostal
space on either side of sternum.
6. Release the mouse of rubber bands. Hold the two flaps of skin
together with forceps and close the wound by stapling the skin
gently with two to three wound clips.
7. Return mouse to the cage for recovery (see Note 14).
Fig. 2 (a) Small incision made perpendicular to the sternum and across the midline of the upper thoracic
region. (b) Intercostal intrathymic injection, the operator inserts the needle between third and fourth ribs
4 Notes
1. Use sterile vials and aseptic techniques to prepare ketamine/

xylazine cocktail for surgical anesthesia.
2. Ketamine and xylazine are controlled substances and therefore
should be stored in a locked drawer, at room temperature.
3. It is easier to inject the thymus of young mice (6-week-old
recipients). It can be difficult to inject aged mice due to age-
related thymic involution.
4. Dosing of the anesthetics is crucial and should be evaluated
relative to weight, sex, and strain of the mice. Ketamine/xyla-
zine anesthetic should be administered at 100/20 mg/kg
mouse weight, ~200 μL for a 20 gm mouse.
5. Anesthesia takes effect in few minutes and lasts for 20–30 min
depending on the mouse strain used. Before starting the pro-
cedure, pinch the toe of the mice to assess the anesthetic
depth. You can inject a small amount of diluted ketamine solu-
tion again if the toe reflex is still active 5 min after ketamine/
xylazine anesthetic injection. Do not administer additional
ketamine/xylazine anesthetic because the mice will be asleep
for a long time and the risk of anesthetic death goes up.
6. If you are working with more than one mouse, inject next
mouse with ketamine/xylazine anesthetic at this point.
7. At the time of thoracotomy, keep the tip of the scissors away
from the heart and lungs. Bleeding may obscure the opening.
If necessary soak up the blood with a clean absorbent pad.
8. Cut only one-third of the sternum. If the tip of the thymus is
not visible, you may need to extend the sternum incision.
9. After administration of cell suspension or solution into the thy-
mus, allow the needle to remain in the lobe for few seconds so
that the internal pressure decreases. This gives the injected solu-
tion time to redistribute, minimizing leakage after withdrawal.
10. The injected volume should not exceed 10 μL per lobe (try to
inject close to the middle of the lobe ~2–3 mm deep). Due to
enormous size variation of the mouse thymus with age, it is
difficult to specify the injection depth.
11. Immediately after injection, flush the syringe three to four times
with sterile PBS. Take care not to bend the wire plunger.
12. To prevent hypothermia after surgery, place animals under a
heating lamp while still under anesthesia.
13. The efficiency of the technique can be improved by rapid sur-
gical incision and injection.
14. The blind injection technique can be practiced by injecting a
dye (e.g., India ink) followed by removal and sectioning of the
thymus to observe the site of injection.
Acknowledgments
We thank N. Taylor and V. Zimmerman for reading and critical

comments on this chapter. A special thanks to P. Sarkar for
photographs.
References
1. Rezzani R, Nardo L, Favero G et al (2014) cellular traffic from the thymus to the periph-
Thymus and aging: morphological, radiologi- ery in mice. Eur J Immunol 10:210–218
cal, and functional overview. Age 36:313–351 12. Weber BN, Chi AW, Chavez A et al (2011) A
2. Chen W, Issazadeh S, Sayegh MH et al (1997) critical role for TCF-1 in T-lineage specifica-
In vivo mechanisms of acquired thymic toler- tion and differentiation. Nature 476:63–68
ance. Cell Immunol 179:165–173 13. Guidos CJ, Weissman IL, Adkins B (1989)
3. Chen W, Sayegh MH, Khoury SJ (1998) Intrathymic maturation of murine T lympho-
Mechanisms of acquired thymic tolerance cytes from CD8+ precursors. Proc Natl Acad
in vivo: intrathymic injection of antigen Sci U S A 86:7542–7546
induces apoptosis of thymocytes and periph- 14. Jerabek L, Weissman IL (2002) Intrathymic
eral T cell anergy. J Immunol 160:1504–1508 injection for analysis of T-cell progenitor
4. DeMatteo RP, Chu G, Ahn M et al (1997) activity. In: Klug CA, Jordan CT (eds)
Long-lasting adenovirus transgene expression Hematopoietic stem cell protocols, vol 63,
in mice through neonatal intrathymic toler- Methods Mol Med. Humana Press, Totowa,
ance induction without the use of immuno- NJ, pp 161–165
suppression. J Virol 71:5330–5335 15. Hardy RR, Shinton SA (2004) Characterization
5. Muto M, Kubo E, Sado T (1987) Development of B lymphopoiesis in mouse bone marrow
of prelymphoma cells committed to thymic and spleen. In: Gu H, Rajewsky K (eds) B cell
lymphomas during radiation-induced thymic protocols, Methods in molecular biology.
lymphomagenesis in B10 mice. Cancer Res Humana Press, Totowa, NJ, pp 1–24
47:3469–3472 16. Rooke R, Benoist C, Mathis D (2000)
6. Goldschneider I, Komschlies KL, Greiner DL Intrathymic delivery of MHC genes using
(1986) Studies of thymocytopoiesis in rats and recombinant Adenoviruses. In: Kearse KP (ed)
mice. I. Kinetics of appearance of thymocytes T cell protocols: development and activation,
using a direct intrathymic adoptive transfer Methods in molecular biology. Humana Press,
assay for thymocyte precursors. J Exp Med Totowa, NJ, pp 69–79
163:1–17 17. Liu LL, Du XM, Wang Z et al (2012) A sim-
7. Adjali O, Marodon G, Steinberg M et al plified intrathymic injection technique for
(2005) In vivo correction of ZAP-70 immu- mice. Biotech Histochem 87:140–147
nodeficiency by intrathymic gene transfer. 18. de la Cueva T, Naranjo A, de la Cueva E et al
J Clin Invest 115:2287–2295 (2007) Refinement of intrathymic injection in
8. Seggewiss R, Dunbar CE (2005) A new direc- mice. Lab Anim (NY) 36:27–32
tion for gene therapy: intrathymic T cell- 19. Adjali O, Montel-Hagen A, Swainson L et al
specific lentiviral gene transfer. J Clin Invest (2009) In vivo and ex vivo gene transfer in
115:2064–2067 thymocytes and thymocyte precursors. In:
9. De Obaldia ME, Bell JJ, Bhandoola A (2013) Baum C (ed) Genetic modification of hemato-
Early T-cell progenitors are the major granulo- poietic stem cells, vol 506, Methods Mol Biol.,
cyte precursors in the adult mouse thymus. pp 171–190
Blood 121:64–71 20. Handon RB, Mueller K, Miller SH (2010) An
10. De Obaldia ME, Bell JJ, Wang X et al (2013) alternative method for intrathymic injections
T cell development requires constraint of the in mice. Lab Anim (NY) 39:248–252
myeloid regulator C/EBP-α by the Notch tar- 21. Tuckett AZ, Thornton RH, Shono Y et al
get and transcriptional repressor Hes1. Nat (2014) Image-guided intrathymic injection of
Immunol 14:1277–1284 multipotent stem cells supports lifelong T-cell
11. Scollay RG, Butcher EC, Weissman IL (1980) immunity and facilitates targeted immunother-
Thymus cell migration: quantitative aspects of apy. Blood 123:2797–2805
Chapter 18
Analysis of Cell Proliferation and Homeostasis

Using EdU Labeling
Abstract
Determination of cellular proliferation and population turnover is an important tool for research on lym-
phoid cell function. Historically this has been done using radiolabeled nucleotides or nucleoside analogs,
such as BrdU (5-bromo-2-deoxyuridine), that are incorporated into nascent DNA during S-phase.
Recently, a new procedure was developed to label nascent DNA using EdU (5-Ethynyl-2-deoxyuridine).
This new method overcomes limitations imposed by the procedure used to detect BrdU because EdU
detection is based on an easily performed chemical reaction that does not require DNA denaturation, is
quick and reproducible, and has a superior signal-to-noise ratio. This technique offers a wide range of
opportunities to analyze cellular proliferation, population homeostasis, and cell marking procedures.
Key words EdU (5-ethynyl-2′-deoxyuridine), Click it chemistry, Proliferation, S-phase, DNA

replication
1 Introduction
Measurement of the proliferative capacity and life span of defined

cell types is fundamental for understanding population dynamics
and homeostasis. One of the best approaches to accomplish this
goal is to directly measure DNA synthesis. Early studies employed
incorporation of radiolabeled thymidine followed by autoradiogra-
phy and quantification in populations of cells in situ or isolated by
a variety of different procedures [1]. However these procedures
were laborious and it was difficult to analyze large numbers of cells.
Later, methods using BrdU, a halogenated thymidine analog, were
introduced to directly measure de novo DNA synthesis based on
its incorporation during DNA synthesis and flow-cytometric detec-
tion using anti-BrdU antibodies were developed [2]. The great
advance of this approach was that it utilized flow cytometry to
simultaneously measure phenotypic markers, DNA content, and
211
BrdU incorporation of a large number of individual cells. With

some modifications, this overall approach has been used for the last
30 years to study proliferation dynamics and population homeo-
stasis of defined cell subsets [1–3]. The method and timing of
labeling depends on the experimental goals. For example, short
term labeling identifies the proportion of cells actively replicating
DNA during that period. When this approach is coupled with
DNA content analysis, it provides a rich source of information
about cellular proliferation. For example, one can calculate the dis-
tribution of cells actively dividing according to their distribution in
the G1-, S-, and G2/M-phases of the cell cycle. The number of
resting cells can also be easily calculated. In addition, information
about the duration of DNA synthesis, doubling time, and cell cycle
dynamics is available. Due to the fact that EdU is not reutilized,
“pulse-chase” studies can be done; dividing cells are labeled for a
defined period followed by sampling and analysis at various time-
points. This technique is ideal for tracing step-wise cellular differ-
entiation and population dynamics of post mitotic labeled cells
during the chase period, and has been frequently used in studies
on neural development and peripheral lymphocyte differentiation,
some lasting almost 3 months [4]. Another application, using con-
tinuous labeling, allows calculation of proliferation rates of defined
cell populations from the rate at which they incorporate labeled
nucleotide analogs. The lifetime of any phenotypically defined
lymphocyte population can be determined using continuous label-
ing experiments. The calculation is based on determining how
long it takes for 100 % of the cells in a population to become
labeled due to complete replacement of that population from pre-
cursor cells [4–6]. Specific applications of these procedures have
been applied in a vast array of in vivo and in vitro experimental
systems in biological models ranging from C. elegans and drosoph-
ila to humans [7].
In spite of the widespread use of BrdU to measure DNA syn-
thesis, this procedure includes relatively harsh fixation procedures
and requires opening of the DNA using heat, acid or DNase to
expose the BrdU epitope and allow access for anti-BrdU antibod-
ies. These factors limit phenotypic analysis using antibodies directed
to specific proteins due to destruction of epitopes. Additionally,
inconsistency in achieving exposure of the incorporated BrdU
results in variable signal-to-noise ratios. Recently a new flow-
cytometric method to label and detect nascent DNA using EdU
has been developed and made commercially available by Life
Technologies [8]. EdU, is a nucleoside analog of thymidine that is
incorporated into DNA during S-phase just like BrdU and is not
reactive in biological systems [9]. The EdU detection procedure
technique uses a copper (I) catalyzed click reaction chemistry to
covalently couple an azide modified fluorescent dye to incorporated
Analysis of Cell Proliferation and Homeostasis Using EdU Labeling 213
EdU to form a stable triazole ring [8]. Because of the small size of
the click detection reagents, no harsh DNA denaturation steps are
required. As a result, EdU detection uses a simple protocol that
takes less than 30 min with more reproducible results and greater
signal-to-noise ratio. A variety of fluorochromes for EdU detection
facilitate analysis of specific populations defined antibody based
phenotypic profiling by specific proteins. It is likely that the advent
of new and improved techniques using EdU to label nascent DNA
will usher in a new wave of creative scientific exploration and excit-
ing findings.
2 Materials
The components for EdU labeling and detection are included in

kits purchased from Life Technologies that utilize various fluoro-
chromes: Alexa Fluor 488, Alexa Fluor 647, or Pacific Blue.
2.1 In Vitro EdU 1. EdU (5-ethynyl-2′-deoxyuridine, Component A) (Life

Labeling Technologies) (see Note 1).
2. Media appropriate for type of cell to be cultured.
3. Phosphate buffered saline (PBS).
4. DMSO.
2.2 In Vivo EdU 1. EdU (5-ethynyl-2′-deoxyuridine, Component A) (Life

Labeling Technologies) (see Note 1).
2. Phosphate buffered saline (PBS).
3. Syringes with 25 G needles.
4. Sterile drinking water.
2.3 Stain Cell 1. FACS buffer (1× PBS, 0.5–1.0 % BSA, 0.1 % (W/V) sodium
Surface Markers azide) (see Note 2).
2. Tubes compatible with flow cytometer to be used.
3. Fluorescently labeled antibodies.
4. Anti CD16 (2.4G2) antibody use at 5 μg/ml.
2.4 Fix 1. Click-iT fixative (Component D).

and Permeabilize Cells 2. Click-iT saponin-based permeabilization and wash reagent
(Component E).
3. De-ionized water.
2.5 EdU Detection 1. CuSO4 (Component F).

2. Click-iT EdU buffer additive (Component G).
2.6 Optional 1. Cells that have been fixed, permeabilized, and stained for EdU
Additional Antibody with the Click-it reaction.
Staining for 2. Fluorescently labeled antibodies.
Intracellular
3. 1× Click-iT saponin-based permeabilization and wash buffer.
Antigens or with RPE,
PE-Tandem, or Qdot
Antibody Conjugates
2.7 Cell Cycle Stain 1. DAPI (4′,6-diamidino-2-phenylindole) solution 5 mg/ml in

diH2O.
2. FACS buffer.
2.8 Flow Cytometry 1. Flow cytometer and software to analyze data.

Analysis
3 Methods
3.1 In Vitro EdU 1. Establish culture conditions where cells are proliferating.
Labeling 2. Add 4 ml of DMSO or PBS to vial containing 10 mg EdU to
make a 10 mM solution (see Note 1).
3. Add EdU to culture at 2–10 μM. EdU can be dissolved in
DMSO or in aqueous solutions. EdU stock solution is 10 mM
(see Note 3). Include cultures treated with vehicle alone for
controls.
4. Culture for period to label nascent DNA (see Note 4). Harvest
cells, count, and wash with PBS. Resuspend in FACS buffer at
1 × 107 cells/ml and keep on ice.
3.2 In Vivo EdU 1. Prepare EdU at 1 mg/ml in sterile phosphate buffered saline
Labeling (PBS).
2. For short term or pulse labeling of adult mice, perform an
intraperitoneal (IP) injection of 100–200 μl dissolved EdU
(see Note 5). Harvest organs/cells at desired timepoints
(see Note 6).
3. For long term labeling, perform IP injection with 100 μl of
1 mg/ml EdU in PBS. Mice are then provided with drinking
water ad libitum containing 0.3 mg/ml EdU (see Note 7).
Replace drinking water with freshly prepared EdU every 2–3
days (see Notes 1 and 8).
4. Harvest lymphoid organs of interest and process into single
cell suspensions. Count cells and wash with PBS. Resuspend in
FACS buffer at 1 × 107 cells/ml and keep on ice.
5. If cells will not be stained with antibodies, place 100 μl of
the cell suspension (1 × 106 cells) into FACS tubes, add 3 ml
FACS buffer, pellet cells, remove supernatant and proceed to
Subheading 3.4.
3.3 Stain Cell 1. Prepare a chart for the staining strategy listing fluorochromes,
Surface Markers EdU detection and DNA stain. Antibodies using PE, PE-
tandem, or Qdot(R) conjugates should not be used before the
EdU detection step as their signal is reduced by the Click it reac-
tion conditions. Currently, Life Technologies offers azides with
Alexa Fluor 488, Alexa Fluor 647, and Pacific Blue. Also plan
for additional samples for controls including unstained cells and
single colors of each fluorochrome for compensation. When
analyzing rare populations, it is useful to use additional control
samples containing “all stains except one” to set the gates for
the population identified by the missing fluorochromes.
2. Place 100 μl of the cell suspension (1 × 106 cells) into FACS
tubes. Add 0.5 μg anti CD16/CD32 antibody in tubes to be
stained with antibodies to block antibody binding by the Fc
receptor and incubate on ice 10 min.
3. Set up single stain controls for each antibody. Reserve tubes
with cells for EdU detection reagents and DNA stains to be
added later. Incubate tubes on ice 15 min, add 3 ml FACS
buffer, pellet cells and remove supernatant. Proceed to
Subheading 3.4.
4. While control cells are staining, make a mixture with predeter-
mined amounts of all desired antibodies and add an aliquot to
samples to be stained. This procedure facilitates dispensing the
antibodies and reduces tube-to-tube variability. Incubate tubes
on ice 15–30 min, add 3 ml FACS buffer, pellet cells and
remove supernatant. Proceed to Subheading 3.4.
3.4 Fix and 1. All samples and controls should be fixed and permeabilized to
Permeabilize Cells ensure uniform characteristics when run on the flow cytome-
ter. Prepare needed amount of 1× Click-iT saponin-based per-
meabilization and wash reagent by diluting the provided 10×
stock (Component E) with water (see Note 9). A total of
3.6 ml per sample is needed for the standard reaction. If intra-
cellular staining will be done, a total of 6.7 ml per sample is
needed.
2. Gently drag tubes across a rack or use other method to break
up the cell pellets from Subheadings 3.2, step 5 and 3.3, steps
3 and 4.
3. Add 100 μl of Click-iT fixative (Component D) to each tube
and mix well. Incubate the cells for 15 min at room tempera-
ture in the dark.
4. Add 3 ml of FACS buffer, pellet cells, and remove the super-
natant (see Note 10).
5. Gently break up the cell pellet again and add 100 μl of 1×
Click-iT saponin-based permeabilization and wash reagent.
Mix well and proceed directly to Subheading 3.5 (see Note 11).
3.5 EdU Detection 1. Allow kit components to come to room temperature before
opening.
2. Prepare a working solution of fluorescent azide for EdU detec-
tion by adding 130 μl of DMSO to Component B of the kit
(Alexa Fluor 488, Alexa Fluor 647, or Pacific Blue) and mix
well. Any remaining working solution will be stable for a year
if stored at ≤−20 °C.
3. Add 2 ml of deionized water to the vial containing the Click-iT
EdU buffer additive (Component G) to make a 10× stock
solution and gently mix until fully dissolved. Any remaining
stock solution should be dispensed into single use aliquots and
is stable for a year when stored at ≤−20 °C.
4. Prepare a master mix of the components required for the
Click-it reaction. For each tube mix together 438 μl PBS or
TBS, 10 μl CuSO4 (Component F), 2.5 μl fluorescent azide
(prepared in Subheading 3.5, step 1), 50 μl 10× reaction buf-
fer additive (prepared in Subheading 3.5, step 2). Scale up the
mixture for the number of samples to be treated and add
500 μl to each tube. It is important to use the cocktail within
15 min of preparation. It is good practice to include a control
sample of cells not exposed to EdU. In addition, these cells are
needed for single staining compensation controls for intracel-
lular antigens or antigens stained with RPE, PE-tandem, or
Qdot antibody conjugates.
5. Incubate the mixture for 30 min at room temperature in the
dark.
6. Add 3 ml of the Click-iT permeabilization and wash buffer
(prepared in Subheading 3.4, step 1), pellet the cells and
remove the supernatant.
7. Gently dislodge the cell pellet. If intracellular antibody stain-
ing is desired, add 100 μl Click-iT permeabilization and wash
buffer and proceed to Subheading 3.6. Otherwise add 500 μl
of the Click-iT permeabilization and wash buffer and go to
Subheading 3.7 for DNA staining or Subheading 3.8 for flow
cytometry.
3.6 Optional Staining 1. Add predetermined amounts of antibodies to the cells and mix
for Intracellular well. Incubate on ice for 30 min protected from light.
Antigens or with PE, 2. Add 3 ml of the Click-iT permeabilization and wash buffer,
PE-Tandem, or Qdot pellet the cells, and remove the supernatant.
Antibody Conjugates 3. Gently dislodge the cell pellet and add 500 μl of the Click-iT
permeabilization and wash buffer and go to Subheading 3.7
for DNA staining.
3.7 Cell Cycle Stain 1. Dilute 5 mg/ml DAPI stock 1:100 with diH2O and add
12.5 μl to each tube (3 μM final concentration) and incubate
for 15 min at room temperature (see Note 12). Ensure that a
tube of fixed and permeabilized cells is stained with only DAPI
for compensation control. After incubation proceed to
Subheading 3.8.
3.8 Flow Cytometry 1. Use unlabeled and single color controls to set up compensa-
Analysis tion on cytometer and run samples. Use “all but one” controls
to set gates if needed (see Note 13).
2. Gate on the cell population of interest. Collect the fluorescent
signal from DAPI or other DNA content dyes using linear
amplification; all other fluorescent signals should be collected
with logarithmic amplification. When measuring cellular DNA
content on most flow cytometers, use a low flow rate (<500
events per second) during data collection. Use the same col-
lection rate for all samples in the experiment (see Note 14).
3. Cell clumping is often observed when processing samples for
flow cytometry. When analyzing DNA content, it is important
to distinguish between G1 doublets from a G2-M single event.
To discriminate the doublet from the singlet, plot the width
(W) versus area (A) for the channel used in a dot plot graph.
A population of single cells will form a diagonal. W increases
with the diameter of the clump, while the A of a G1 doublet
and the G2/M single cell is the same. Therefore, discrimina-
tion of G1 doublets and clumped cells from G2/M single cells
can be made by gating out events that deviate with greater
width from the diagonal as shown in Fig. 1a [10].
Fig. 1 Human epithelial cells were cultured with EdU for 2 h then stained with Pacific Blue azide and 7-aad. (a)
7-aad fluorescence was collected using the FL3 channel and plotted area (FL3-A) versus width (FL3-W) to
allow discrimination between clumped doublets and single cells shown within the rectangular gate. (b) Cells
were gated based on size and the percentage of EdU labeled cells is shown as a histogram. (c) Bivariate analy-
sis of EdU versus 7-aad staining is shown for cells gated as in (a). Numbers shown denote percentages.
Clockwise from the top are gates that show cells in S-phase (59 %), G2/M (13 %) and G0/G1 (20 %)
4. Using the desired phenotypic gate, the percent cells that syn-
thesized DNA and incorporated EdU during the pulse period
can be determined with a simple histogram (Fig. 1b). Plotting
EdU versus DAPI provides more information about the cell
cycle (Fig. 1c). In this example, human epithelial cells were
pulsed with EdU for 2 h and stained with 7-aad and Pacific
Blue azide. The bivariate analysis shows that 59 % of these cells
synthesized DNA during the pulse and represent cells in
S-phase (Fig. 1c). Twenty percent of the cells were quiescent
and remained in G1 phase while 13 % did not synthesize DNA
and were in G2-M phase (Fig. 1c). A population of stable qui-
escent cells can be identified by lack of EdU incorporation
over a longer labeling period, while a population of cycling
cells will become uniformly labeled.
4 Notes
1. After being dissolved, remaining EdU stock solution is stable

for up to 1 year when stored at −20 °C. EdU solution should
be portioned into single use aliquots to avoid freeze thawing.
EdU is incorporated into DNA and is a potential mutagen.
EdU has been identified as a teratogen in laboratory animals.
Proper protective clothing should be used when handling
EdU. Also proper procedures should be implemented to mini-
mize contamination and dispose of waste according to institu-
tional guidelines. Waste, including stock solutions, used
media; animal cage litter, feces, urine, and water containing
EdU should be considered as hazardous.
2. Fetal bovine serum can also be used at 3–10 % instead of
BSA. Sodium azide is used as a preservative; its addition to the
buffer is optional. Follow proper precautions when disposing
of sodium azide to avoid accumulation of potentially explosive
deposits in plumbing.
3. Preliminary experiments should be done to determine optimal
concentration for labeling and assess potential toxicity of
DMSO or EdU on the cells of interest. In general 10 μM EdU
has no detectable toxicity for a variety of different cell types. If
toxicity is noted, reduction of EdU concentration or short-
ened labeling times may be indicated.
4. Preliminary experiments should be done to optimize labeling
periods. In vitro EdU labeling can be detected in as little as
3 min. In most cases a 1–4 h labeling is sufficient. In general,
labeling periods used for BrdU studies can be used as a good
guideline for EdU labeling period, however EdU labeling can
often be detected with shorter labeling periods than
BrdU. Incorporated EdU is very stable and cells can be
washed, fixed and stored for months without loss of signal [8].
5. Alternatively mice can be injected with EdU twice 2–4 h apart

to initiate the pulse labeling. Start timing the chase after the
second injection. Newborn to 20-day-old mice can be injected
with 50 μl 1 mg/ml EdU solution IP or SC.
6. Give enough time for EdU to diffuse and label proliferating
cells. We have readily detected EdU incorporation in thymo-
cytes after 60–90 min after injection.
7. EdU can also be administered subcutaneously (SC) every
3–4 h for up to 5 days if needed. Successful long term EdU
labeling (in brain) has been achieved by injecting EdU every
3–4 h during a 12 h daytime period followed by a 12-h over-
night period with no injections for a 5-day period [11].
Sequential injections are indicated for mice under 20 days old
since they have not been weaned and do not drink much water.
However the extra handling and disturbance may affect the
experimental outcome so additional mice injected with vehicle
alone on the same injection schedule should be included as
controls. Long labeling periods may be toxic to some popula-
tions. Signs of distress, ruffled hair, lethargy and decreased
thymus size compared to controls [12] suggest possible toxic
effects. Labeling periods up to 5 weeks have been done with
BrdU to study lymphocyte turnover. We have not observed
toxicity with EdU labeling periods up to 7 days.
8. Unlike BrdU, EdU is not light sensitive so water bottles do
not need to be covered in foil.
9. The Click-iT permeabilization reagent maintains the morpho-
logical light scatter characteristics of leukocytes. It can be used
with whole blood or cell suspensions containing red blood
cells and will lyse red blood cells. The diluted 1× solution is
stable for 6 months when stored at 2–6 °C while the 10× stock
is stable for a year stored at −20 °C. In some cases, it may be
more convenient to make up 500 ml 1× buffer. Note that this
buffer contains sodium azide.
10. If red blood cell debris or hemoglobin is present in the sample,
repeat the wash step before proceeding.
11. Cells can be held at this stage for up to 30 min if needed.
12. DAPI is a potential mutagen so use proper precautions when
handling and disposing. DAPI stain may not compatible with
Pacific Blue depending on whether a violet laser is used. Other
DNA dyes include propidium iodide (FL2 channel) or 7-aad
(7-Aminoactinomycin D) (FL3 channel). Life Technologies
also sells several DNA dyes. RNase treatment is required if the
dye binds RNA to ensure accurate DNA content profile and
facilitate analysis.
13. Use this chart to determine the correct parameters to detect

EdU with the different azides available from Life Technologies.
Azide Excitation (mn) Emission filter (or similar)

Alexa Fluor 488 488 Green (530/30 nm)
Alexa Fluor 647 633/635 Red (660/20 nm)
Pacific Blue 405 Violet (450/50 nm)
14. If additional dilution of cells is needed, ensure that wash buf-

fer containing 3 μM DAPI is used to prevent reduction of
fluorescent signal.
Acknowledgement
The Intramural Research Program of the National Cancer Institute

at the National Institutes of Health supports the authors. The
authors thank Kevin Chua for comments on the paper. The authors
have no conflicts of interest to disclose.
References
1. Darzynkiewicz Z, Traganos F, Zhao H, mature lymphocytes. Eur J Immunol
Halicka HD, Li J (2011) Cytometry of DNA 20:1697–1708
replication and RNA synthesis: Historical per- 7. Cavanagh BL, Walker T, Norazit A, Meedeniya
spective and recent advances based on “click AC (2011) Thymidine analogues for tracking
chemistry”. Cytometry A 79:328–337 DNA synthesis. Molecules 16:7980–7993
2. Dolbeare F, Gratzner H, Pallavicini MG, Gray 8. Salic A, Mitchison TJ (2008) A chemical
JW (1983) Flow cytometric measurement of method for fast and sensitive detection of
total DNA content and incorporated bromo- DNA synthesis in vivo. Proc Natl Acad Sci
deoxyuridine. Proc Natl Acad Sci USA 80: USA 105:2415–2420
5573–5577 9. Buck SB, Bradford J, Gee KR, Agnew BJ,
3. Gray JW, Dolbeare F, Pallavicini MG, Beisker Clarke ST, Salic A (2008) Detection of S-phase
W, Waldman F (1986) Cell cycle analysis using cell cycle progression using 5-ethynyl-2’-de-
flow cytometry. Int J Radiat Biol Relat Stud oxyuridine incorporation with click chemistry,
Phys Chem Med 49:237–255 an alternative to using 5-bromo-2’-deoxyuri-
4. Tough DF, Sprent J (1994) Turnover of naive- dine antibodies. Biotechniques 44:
and memory-phenotype T cells. J Exp Med 927–929
179:1127–1135 10. Nunez R (2001) DNA measurement and cell
5. Parretta E, Cassese G, Santoni A, Guardiola J, cycle analysis by flow cytometry. Curr Issues
Vecchio A, Di Rosa F (2008) Kinetics of in Mol Biol 3:67–70
vivo proliferation and death of memory and 11. Clarke LE, Young KM, Hamilton NB, Li H,
naive CD8 T cells: parameter estimation based Richardson WD, Attwell D (2012) Properties
on 5-bromo-2’-deoxyuridine incorporation in and fate of oligodendrocyte progenitor cells in
spleen, lymph nodes, and bone marrow. the corpus callosum, motor cortex, and piri-
J Immunol 180:7230–7239. form cortex of the mouse. J Neurosci 32:
6. Rocha B, Penit C, Baron C, Vasseur F, 8173–8185
Dautigny N, Freitas AA (1990) Accumulation 12. Tough DF, Sprent J, Stephens GL (2007)
of bromodeoxyuridine-labeled cells in central Measurement of T and B cell turnover with
and peripheral lymphoid organs: minimal esti- bromodeoxyuridine. Curr Protoc Immunol
mates of production and turnover rates of Chapter 4:Unit 4 7
Chapter 19
Characterization and Isolation of Human T Cell Progenitors

Inge Van de Walle, Karin Davids, and Tom Taghon
Abstract
During their development, human T cells undergo similar genomic changes and pass through the same
developmental checkpoints as developing thymocytes in the mouse. The difference between both species,
however, is that some of these developmental stages are characterized by different phenotypic markers and
as a result, evidence emerges that the molecular regulation of human T cell development subtly differs
from the mouse [1–4]. In this chapter, we describe in detail how the different stages of human T cell
development can be characterized and isolated using specific surface markers.
Key words Human, T cell development, Thymocytes, Isolation of T cell progenitors, Fluorescent
activated cell sorting, Magnetic activated cell sorting
1 Introduction
In humans, the hematopoietic progenitor cells that seed the thy-

mus are still poorly defined and therefore remain controversial.
The group of Canque identified a CD34hiCD45RAhiCD7low pro-
thymocyte population in fetal bone marrow, cord blood, and thy-
mus that display preferential T- and NK-cell lineage potential
compared to their CD7− counterpart. Given their selective poten-
tial to colonize the thymus in a chimeric in vitro setting, they pro-
pose these cells to be the direct precursors of uncommitted
CD34+CD1− thymocytes (Fig. 1) [2]. In contrast, the group of
Crooks identified, besides this CD34+CD10+CD1−CD7low uncom-
mitted pro-thymocyte subset, also a CD7− population with clear
myeloid, lymphoid and possibly also erythroid differentiation
potential, suggesting that this is a more primitive subset of pro-
genitors that colonizes the human thymus [3, 5]. While these
CD7− precursors may not be as efficient to colonize the thymus, it
still remains possible that the CD7− and CD7low subsets both con-
tribute to human T cell development since it is anticipated that
very few cells effectively seed the thymus. Within the CD34+CD1−
uncommitted thymocyte population, a CD7high subset can be
221
222 Inge Van de Walle et al.
CD56 +
NKT CD3 +
TCR-αβ+
Bone Thymus
marrow
Foxp3 +
CD25 +
Treg
CD4 +
Commitment β-selection +/- selection CD3 +
TCR-αβ+
CD1a+ CD1a-
CD4 + CD8 CD3 +
CD34 + CD34 + CD34 +/- CD1a+ CD1a+ TCR-αβ+
CD8αβ +/- SP
CD10 + CD7 + CD1a+ CD4 + CD4 + CD3 + CD45RA +
CD7 + CD1a + CD4+ CD8αα+/- CD8αβ+/- TCR-αβ+ CD45RO -
CD1a- CD4 - CD8αα+/- CD28+ CD3 - CD69 -
CD1a-
CD4
HPC T-S T-C ISP4 β DP αβ CD3 +
SP TCR-αβ+
CD45RA +
CD45RO -
CD45RA +
CD38 - CD3 +
CD34 + TCR-γδ+
HPC γδ γδ
CD10 + CD1 -
CD7 -/low
CD34 + CD3 +
CD10 + TCR-γδ+
CD7 -/low TCR rearrangements CD1 +
δ α
γ
β
Fig. 1 Schematic overview of the different developmental stages that characterize human T cell development.
HSC hematopoietic stem cell, T-S T-lineage specified progenitor, T-C T cell committed progenitor, ISP4 CD4
immature single positive, β rearranged T cell receptor-β chain, DP double positive, αβ T cell receptor-αβ
positive cell, NKT natural killer T cell, Treg regulatory T cell, CD8 SP CD8 single positive, CD4 SP CD4 single
positive, γδ T cell receptor-γδ positive cell, CD cluster of differentiation
identified that contains more differentiated T-lineage specified

progenitor cells [3]. In each case, it is important to note that the
markers CD44 and CD25, used to characterize immature so-called
double negative thymocytes in the mouse, are not used to discrimi-
nate different development stages of early developing thymocytes
in human. While it is anticipated that the most immature human
thymocytes express c-kit, analogous to murine ETPs, this marker
has not been used thus far to isolate human ETPs, which may be
due to the lack of appropriate antibodies.
Irreversible commitment to the T cell pathway is completed
when the cells upregulate the human immature T cell marker
CD1a [6]. During these specification and commitment processes,
T cell receptor (TCR) rearrangements at the TCR-δ, TCR-γ, and
TCR-β loci are initiated, in that respective order [7], and peak dur-
ing the next developmental stage, which is characterized by the
gradual loss of the progenitor marker CD34 and the induction of
CD4 expression. This is an important difference between mouse
and human since these so-called immature single positive
thymocytes, characterized as CD34−/+CD4+CD3−CD28−, have not
Characterization and Isolation of Human T Cell Progenitors 223
yet passed the β-selection checkpoint, in contrast to mouse ISP

thymocytes that are the immediate precursors of DP thymocytes
and posses a functional rearranged TCR-β chain. In human, even
some CD4+CD8α+ thymocytes have not passed the β-selection
checkpoint, illustrating the importance of using appropriate sur-
face markers to correctly identify the discrete stages of human T
cell development [4].
The result of the various recombination processes directly
influences the developmental lineage outcome of the T cell precur-
sors. In frame rearrangements of TCR-γ and TCR-δ chains mainly
result in the generation of TCRγδ+CD3+ T cells that further
mature, while a functional TCR-β chain will pair with the surro-
gate pre-T-cell receptor α chain (pTα) that allows pre-TCR signal-
ing to induce β-selection [4, 8, 9]. Passage through this
developmental checkpoint results in the upregulation of CD28
[10], induces strong proliferation as also illustrated by the upregu-
lation of CD71 and finally results in the rapid differentiation into
CD4+CD8αβ+ double positive thymocytes that will initiate TCR-α
chain rearrangements, illustrating the importance of CD8β as a
marker of true β-selected DP thymocytes [9].
In frame TCR-α chain rearrangements result in the expression
of a full TCRαβ-CD3 complex in these DP progenitors that subse-
quently mature into naïve CD1−CD45RA+CD4+ T helper cells or
CD8+ cytotoxic T cells through a process of ligand-dependent posi-
tive and negative selection. We refer to Plum et al. [11] for details
on these processes. In addition to these subsets, also NKT and reg-
ulatory T cells develop intrathymically from DP thymocytes, but
these lineages are not discussed further in this chapter. An overview
of the human T cell development is presented in Fig. 1.
2 Materials
2.1 Extraction 1. 140 mm petri dish.

of Human Thymocytes 2. 10 ml syringe.
3. 50 ml conical tube.
4. 40 μM cell strainer.
5. Iscove’s Modified Dulbecco’s Medium (IMDM).
6. Penicillin/streptomycin (Pen/Strep).
7. L-glutamine.
8. Fetal bovine serum (FBS, specific serum testing is needed, see
Note 1); heat-inactivated at 56 °C for 1 h.
9. Dimethyl sulfoxide (DMSO).
10. Turk’s solution: 10.0 g/l CH3COOH.
11. Hemocytometer counting chamber.
2.2 Lymphoprep 1. Phosphate buffered saline (PBS).

Density Gradient 2. Dulbecco’s phosphate buffered saline (DPBS).
for Isolation of Viable
3. Lymphoprep.
Mononuclear Cells
4. MACS buffer (see Subheading 2.3).
5. Turk’s solution (see Subheading 2.1).
2.3 MACS Direct 1. MACS buffer: 2 % FBS (heat-inactivated at 56 °C for 1 h),

CD34 Progenitor Cell 2 mM EDTA in PBS. Degas buffer before use (prepared 1 day
Isolation Kit in advance) as air bubbles could block the column. Store at
2–8 °C up to 1 month.
2. Human wash buffer for analytical staining of human hematopoi-
etic cells (see Note 2):PBS containing 1 % BSA and 0.1 % NaN3.
Correct pH to 7.2–7.3 and adjust osmolarity to 275–285 mOsm.
Filter through 0.22 μM filter. Store buffer at 2–8 °C.
3. 15 ml polypropylene conical tube.
5. Trypan blue 0.4 %.
6. MACS Direct CD34 Progenitor Cell Isolation Kit, human,
Miltenyi, containing FcR blocking reagent and CD34 micro-
beads (Miltenyi; Bergisch Gladbach, Germany).
7. Magnetic cell separator and LS separation column (Miltenyi;
Bergisch Gladbach, Germany).
8. 30 μm pre-separation filter.
2.4 Flow-Cytometric 1. MACS buffer (see Subheading 2.3).

Cell Sorting 2. 5 ml polypropylene round-bottom tubes.
3. Iscove’s Modified Dulbecco’s Medium (IMDM).
4. Penicillin/streptomycin (Pen/Strep).
5. L-glutamine.
6. Fetal bovine serum (FBS); heat-inactivated at 56 °C for 1 h.
7. Monoclonal antibodies suitable for flow-cytometric analysis
(see Table 1 for an overview of monoclonal antibodies used in
the described protocols). These can be obtained from differ-
ent suppliers.
2.5 CD3+and CD8+ 1. Mouse anti-human monoclonal antibodies (MAbs): glycopho-

Depletion with rin-A, CD8 unlabeled, CD3 unlabeled.
Dynabeads 2. Sheep anti-mouse immunoglobulin-coated Dynabeads
(Invitrogen by Life Technologies; AS, Norway).
Table 1
Overview of monoclonal antibodies to sort different T cell subsets
Start population of interest Mouse anti-human monoclonal antibodies

Extrathymic progenitors from bone marrow or cord blood CD34, CD3, CD14, CD56, CD19, CD38
Double positive thymocytes CD4, CD8, CD3
+
CD34 thymocytes CD34, CD4, CD1
+
CD4 ISP thymocytes CD4, CD34, CD8, CD3, CD28
TCR-γδ T cells from thymus TCR-γδ, CD3, CD1
3. DynaMag™ Magnet (Invitrogen by Life Technologies; AS,

Norway).
4. Dulbecco’s Phosphate Buffered Saline (DPBS), 2 % FBS.
5. Fetal Bovine Serum (FBS); heat-inactivated at 56 °C for 1 h.
2.6 Anti TCR-γδ 1. MACS buffer (see Subheading 2.3).

Microbead Kit 2. Anti TCR-γδ microbead kit, human, Miltenyi, containing anti-
TCR γ/δ Hapten-Antibody (see Note 3) and Anti-Hapten
MicroBeads-FITC (Miltenyi; Bergisch Gladbach, Germany).
3. Magnetic cell separator and LS separation column (Miltenyi;
Bergisch Gladbach, Germany).
4. 30 μm pre-separation filter.
7. Alternative method can be performed as well (see Note 4).
3 Methods
3.1 Human 1. Transfer thymus tissue, obtained from cardiac surgeries (see
Thymocyte Cell Note 5), to a 140 mm petri dish with 25 ml IMDM supple-
Suspension mented with penicillin (100 U/ml), streptomycin (100 μg/
ml), L-glutamine (2 mM) and 10 % heat-inactivated FBS. Mince
thymus tissue into 4–5 mm pieces with a sterile scalpel.
2. Gently smash thymus fragments with the end of a plunger
from a 10 ml syringe to release the thymocytes. The medium
will become turbid.
3. Transfer cell suspension to a 50 ml polypropylene conical tube

with a 40 μm cell strainer to remove dispersed cells from tissue
fragments.
4. Rinse petri dish with 25 ml IMDM supplemented with peni-
cillin (100 U/ml), streptomycin (100 μg/ml), L-glutamine
(2 mM), and 10 % heat-inactivated FBS and repeat steps 2
and 3 until all tissue is processed.
5. Determine cell number with Turk’s solution. Thymus suspen-
sion can be used immediately for further purification, stored
overnight at 4 °C (at a concentration of 30–50 × 106 cells/ml)
or frozen for later use (see Note 6 and Chapter 20 for details).
During overnight incubation at 4 °C and after freezing and
thawing of thymus suspension, CD4+CD8+ double positive
thymocytes (>80 % of thymocytes) predominantly die, result-
ing in an enrichment of more immature thymocytes that makes
their purification easier and cheaper since less reagents are
required. It is strongly recommended to remove the dead thy-
mocytes using a Lymphoprep density gradient (see Note 7)
prior to further purifications as they will interfere with this
process (see Subheading 3.2).
3.2 Purification 1. Dilute cord blood 1:3 with PBS. Dilute bone marrow 1:5 with
of Viable Cells by PBS (see Note 8).
Lymphoprep Density 2. Transfer maximum 30 ml of diluted cord blood or thymus
Gradient suspension in a 50 ml polypropylene conical tube. Gently pipet
15 ml Lymphoprep™ below diluted cord blood or thymus
suspension. For bone marrow, 15 ml Lymphoprep™ is trans-
ferred in a 50 ml polypropylene conical tube. Gently pipet
30 ml of diluted bone marrow on top of the Lymphoprep™.
3. Centrifuge at 940 × g for 20 min at room temperature (accel-
eration and brake at 30 % of maximum).
4. Aspirate the interface (Fig. 7), containing mononuclear cells
from cord blood sample, bone marrow (see Note 9) or thymus
suspension, using a Pasteur pipette (see Note 10).
5. Pool all interface samples from cord blood, bone marrow or
thymus suspension in a 50 ml polypropylene conical tube and
wash twice in 50 ml cold PBS for cord blood and bone mar-
row, or cold DPBS for thymus suspension (centrifuge at
400 × g for 10 min at 4 °C).
6. Resuspend cell pellet in 50 ml MACS buffer and determine
cell number using Turk’s solution.
supernatant.
8. The mononuclear cell fraction is now ready for further purifi-
cation of specific subsets. Optional: cells can be frozen and
stored in liquid nitrogen at this stage for later use (see Note 6).
3.3 Isolation Extrathymic CD34 positive progenitor cells can be isolated from
of Extrathymic CD34 the mononuclear fraction obtained after Lymphoprep density gra-
Positive Cells Using dient of bone marrow or cord blood.
Magnetic Activated 1. Determine cell number (see Note 11), centrifuge cell suspen-
Cell Sorting (MACS) sion at 500 × g for 10 min, aspirate supernatant completely.
2. Resuspend cell pellet in 300 μl MACS buffer for up to 108
total cells, add 100 μl of FcR blocking Reagent, add 100 μl of
CD34 MicroBeads (see Note 12), mix well and incubate for
30 min in the refrigerator (2–8 °C, see Note 13).
3. Wash labeled cells twice by adding 15 ml MACS buffer and
centrifuge at 500 × g for 10 min, aspirate supernatant.
4. Resuspend cell pellet in 3 ml MACS buffer and proceed to
magnetic separation.
5. Place the LS separation column in the magnetic field of the
MACS Separator, prepare LS column by rinsing three times
with 3 ml MACS buffer.
6. Apply cell suspension onto the column (see Note 14), discard
flow-through containing unlabeled cells. Wash the column
three times with 3 ml MACS buffer; perform washing steps by
adding buffer aliquots only when the column reservoir is
empty.
7. Remove the column from the separator and place it in a 15 ml
polypropylene conical tube. Pipette 3 ml MACS buffer onto
the column and immediately flush out the magnetically labeled
cells by firmly pushing the supplied plunger into the column.
Optional: Cells can be passed on a second LS column if a
higher purity is desired without the need for cell sorting as
described below. Repeat steps 4–7.
8. Count the cells and check CD34+ purity on 10,000 cells by
adding a fluorescent labeled anti-human CD34 monoclonal
antibody.
9. Incubate for 30 min at 2–8 °C.
10. Wash cells by adding 2 ml of human wash buffer, centrifuge at
500 × g for 6 min, discard supernatant by decanting 5 ml tube.
11. Run sample on a flow cytometer (Fig. 2).
12. Purity of CD34+ cells obtained from cord blood is between 60
and 80 %.
Optional: the CD34+ enriched cell fraction can be frozen and
stored in liquid nitrogen (see Note 6).
3.4 Purification of Enriched CD34+ cells, either freshly isolated or frozen (see Chapter
CD34+ Precursors by 20 for thawing cells), are preferentially further purified to >99 %
Fluorescence-Activated purity by cell sorting to avoid outgrowth of contaminating mature
Cell Sorting (FACS) T cells.
a b c
Fig. 2 Gating strategy to sort extrathymic CD34+Lin− progenitor cells derived from cord blood. Within the lym-
phocyte gate (a), a viable (b) population of CD34+ lin− progenitor cells can be gated (c). These cells are negative
for the monocyte marker CD14, the B cell marker CD19, the NK cell marker CD56 and the T cell marker CD3
1. Resuspend the known number of cells in 100 μl MACS buffer

and stain with appropriate antibodies for 30–45 min at
2–8 °C. The antibodies that are required depend on the par-
ticular scientific question. For most experiments, it is sufficient
to start with CD34+ cells that are negative for lineage markers
(Lin− = CD3−CD56−CD14−CD19−). However, one can also
enrich for more primitive precursors by adding CD38 to the
panel and sort CD34+CD38−/lowLin− cells.
2. Following incubation with monoclonal antibodies, wash the
labeled cells once in 10 ml MACS buffer at 500 × g for 6 min
and resuspend the cells in an appropriate volume MACS buf-
fer (see Note 15) for cell sorting.
3. Sorted fractions are collected in 5 ml polypropylene round-
bottom tubes filled with 1.5 ml IMDM supplemented with
penicillin (100 U/ml), streptomycin (100 μg/ml), L-gluta-
mine (2 mM), and 50 % heat-inactivated FBS.
4. Transfer the collected cells to a 15 ml polypropylene conical
tube, dilute the volume to 10 ml with MACS buffer and cen-
trifuge at 500 × g for 10 min.
5. Resuspend the cells in 1 ml IMDM supplemented with peni-
cillin (100 U/ml), streptomycin (100 μg/ml), L-glutamine
(2 mM), and 10 % heat-inactivated FBS
6. Count the cells (see Note 16) and take out 10–20 μl (see Note
17) to check the purity of the sorted population. If the purity
is >99 %, cells can be used for downstream T cell development
assays (Chapter 20). If the purity is insufficient, cells can be
resorted if the remaining cell number is high enough.
Optional: the sorted cell fraction can be frozen and stored in
liquid nitrogen (see Note 6).
In the following sections, we describe the most effective pro-
cedures to isolate specific human thymocytes subsets.
3.5 Isolation 1. Thymus suspension is used as starting material, either fresh,

of Intrathymic CD34 following overnight incubation or after freezing–thawing
Positive Cells (see Subheading 3.1).
Using MACS 2. For purification of the most immature CD34+ thymocytes,
MACS CD34+ microbeads are used, similar as for extrathymic
cord blood or bone marrow precursors (see Subheading 3.3,
steps 1–7) (see Note 18).
3. Purity of CD34+ purified thymocytes should be around 95 %
following one column and 99 % if cells are sequentially passed
onto two columns.
Optional: the CD34+ enriched cell fraction can be frozen and
stored in liquid nitrogen (see Note 6).
4. Enriched CD34+ cells, either freshly isolated or frozen (see
Chapter 20 for thawing cells), can be further purified into
uncommitted CD34+CD4−CD1a− thymocytes and the earliest
committed CD34+CD4−CD1a+ thymocytes, depending on
the purpose of the experiment.
and stain with anti-human CD34, CD1 and CD4 antibodies
for 30–45 min at 2–8 °C.
6. Follow steps 2–6 from Subheading 3.4.
7. See Fig. 3 for representative pre-sort FACS plots.
3.6 Isolation 1. Thymus suspension is used as starting material, either fresh,

of Immature Single following overnight incubation or after freezing–thawing (see
Positive Cells Subheading 3.1).
(CD4+CD8−CD3−) 2. Determine the desired cell number (see Note 19), centrifuge
these cells at 500 × g for 6 min and aspirate the supernatant.
Resuspend 1 × 107 to 4 × 108 cells in 200 μl MACS buffer.
a b
Fig. 3 Gating strategy to sort CD34+ thymocytes. CD4 expression is used to dis-
criminate between the CD34+CD4− cells and the more differentiated CD34+CD4+
that immediately precedes the immature single positive cells (a). In the
CD34+CD4− population two populations can be discriminated based on the
expression of CD1 (b). Upregulation of CD1 correlates with commitment to the
T-cell lineage in postnatal human thymocytes
3. Label the cells first with an unlabeled anti-glycophorin-A

monoclonal antibody, and subsequently with unlabeled anti-
CD8 and -CD3 monoclonal antibodies for depletion of resid-
ual red blood cells and CD8+ and CD3+ cells (see Note 20).
4. Incubate for 30 min at 2–8 °C.
5. In the meantime, resuspend sheep anti-mouse immunoglobu-
lin coated Dynabeads in the vial by vortexing 30 s, or tilt
and rotate for 5 min. Ensure that all magnetic beads are
resuspended.
6. Transfer the desired amount of Dynabeads to 5 ml tubes (max
1.25 ml/tube) and dilute the beads to 4 ml in DPBS 2 % FBS
and resuspend. Use four beads for each cell that is being
labeled, so the starting volume would be 1 ml of a 4 × 108
beads suspension for 100 × 106 cells.
7. Place the tubes in the DynaMag™ Magnet for 2–5 min.
8. Aspirate the supernatant.
9. Take the tubes out of the magnetic field to resuspend the
beads in 4 ml DPBS 2 % FBS and place back in the DynaMag™
Magnet for 2–5 min.
10. Aspirate the supernatant.
11. Repeat washing steps 8 and 9 two more times.
12. Pool all beads and resuspend in the original starting volume
(see step 6) using DPBS 2 % FBS.
13. After 30 min incubation, wash the antibody labeled cells from
step 3 twice in 10 ml DPBS 2 % FBS and centrifuge at 500 × g,
4 °C for 6 min.
14. Resuspend the labeled cells in DPBS 2 % FBS at a final concen-
tration of 150 × 106 cells/ml and add 2:3 volume of the pre-
washed Dynabeads in a round bottom 15 ml conical tube.
15. Incubate on ice for 30 min, while gently shaking every 5 min
to keep the beads in suspension.
16. Dilute the cell suspension 4.5 times with DPBS 2 % FBS and
distribute the suspension in 5 ml tubes (3 ml/tube).
17. Place the tube(s) in the DynaMag™ Magnet for 5 min.
18. Aspirate the supernatant from the tubes and pool if multiple
tubes are used.
19. Centrifuge at 500 × g at 4 °C for 6 min.
20. Resuspend the cells in DPBS 2 % FBS in a volume equal to the
volume of the remaining beads and add the remaining
Dynabeads (step 14) in a round bottom 15 ml conical tube.
22. Use 5 μl of the supernatant in a hemocytometer to verify that
all beads are drawn to the magnet. If magnetic beads are still
a b
Fig. 4 Gating strategy to sort CD4 immature single positive thymocytes (4ISP)
after CD3 and CD8 depletion of total thymocytes. Two populations can be dis-
criminated in the CD4+CD3−CD8−CD34− cell fraction based on the expression of
CD28. The differential expression of CD28 discriminates between pre- and post
(CD28+) β-selected cells
present, elongate the incubation time in the DynaMag™

Magnet for another 5 min. If no beads are present, aspirate the
unlabeled cells from the tube, count the cells and centrifuge at
500 × g at 4 °C for 6 min.
23. The resulting CD3 and CD8 depleted thymocyte fraction can
subsequently be further purified into CD4+CD3−CD8−CD28−
pre- and CD4+CD3−CD8−CD28+ post-β selection subsets
using a cell sorter.
and stain with anti-human CD4, CD3, CD34, CD8, and
CD28 antibodies for 30–45 min at 2–8 °C.
26. See Fig. 4 for representative FACS plots.
3.7 Isolation 1. Due to their limited life span, CD4+CD8+ DP cells need to be
of Double Positive (DP) isolated directly from freshly prepared thymus suspension.
and Single Positive Both CD3− and CD3+ thymocytes can be isolated. Using the
(SP) Thymocytes same labeling, mature SP CD4+CD3+ and CD8+CD3+ thymo-
cytes can be sorted at the same time, if the available cell sorter
has the capacity to sort 4 populations simultaneously.
2. Put the desired cell number (see Note 21) into a 15 ml poly-
propylene conical tube, centrifuge at 500 × g for 6 min.
3. Resuspend in 100 μl MACS buffer and stain cells for CD4,
CD8, and CD3 for 30–45 min at 2–8 °C. Use the desired
antibodies in appropriate concentrations as recommended by
the manufacturers (see Note 20).
4. Follow steps 14–18 from Subheading 3.4 but pay attention
to Note 22.
5. See Fig. 5 for representative FACS plots.
a b
c d
Fig. 5 Gating strategy to sort DP and SP thymocytes. CD4+CD8+ double positive cells as shown in (a) can be
separated in CD3− and CD3+ subsets as shown in panel (b). Single positive cells CD4+CD8− and CD8+CD4−
thymocytes (a) express high levels of CD3 (c, d) that can be incorporated as a marker to sort for CD4 and CD8
SP cells
3.8 Isolation of TCR 1. Freshly prepared thymocytes (see Subheading 3.1) are used as
γδ Cells starting material.
2. Determine the cell number (see Note 23), centrifuge the cells
at 500 × g for 6 min and aspirate the supernatant.
3. Resuspend the cell pellet in 20 μl MACS buffer per 107 total
cells (see Note 24).
4. Label thymocytes with anti-TCR γδ haptenated (see Note 3)
antibodies (10 μl per 107 cells). The amount of antibody added
is based on 1:4 of the total cell number (see Note 25).
5. Mix well and incubate for 10 min at 4–8 °C.
6. Add 15 μl MACS buffer per 107 total cells and add MACS
Anti-Hapten MicroBeads-FITC (20 μl per 107 cells) based on
1:4 of the total cell number.
7. Mix well and incubate for 15 min at 4–8 °C.
8. Proceed with MACS protocol (see Subheading 3.3, steps 3–7).
a b
Fig. 6 Gating strategy to sort immature and mature TCR-γδ thymocytes. Two
populations can be obtained in the TCR-γδ population based on the expression
of CD1. Downregulation of CD1 correlates with a more mature phenotype
9. Count the cells and centrifuge at 500 × g at 4 °C for 6 min.

10. To purify and separate immature and mature TCR-γδ positive,
stain the MACS enriched cells with CD3 and CD1 monoclo-
nal antibodies for 30–45 min at 2–8 °C (see Note 26).
12. See Fig. 6 for representative FACS plots to sort immature
TCRγδ+CD3+CD1a+ and mature TCRγδ+CD3+CD1a− TCRγδ
T cells.
4 Notes
1. Technical disadvantages to using serum include the undefined

nature of serum, batch-to-batch variability in composition,
and the risk of contamination. Therefore, it is recommended
to perform serum testing. To test for T cell development, we
culture fetal thymic lobes (fetal day 14) of B6 mice in fetal
thymus organ culture for 14 days and check for efficiency to
support T cell development by cell counting and flow-
cytometric analysis of T cell differentiation.
2. This buffer is not recommended for staining cells that will be
cultured due to the presence of NaN3.
3. The anti-TCRγ/δ MicroBead Kit consists of a hapten-
conjugated anti-TCRγ/δ antibody and FITC-conjugated
anti-hapten MicroBeads.
4. An example of an alternative method, instead of anti-TCRγ/δ
MicroBead kit, is to first stain thymocytes with a TCRγ/δ-PE
conjugated antibody, followed by magnetically labeling of the
cells with anti-PE Microbeads. Subsequently, cell suspension is
loaded on MACS Column that is placed in the magnetic field

of a MACS separator. The magnetically labeled cells are
retained in the column.
5. Thymus tissue is obtained from children undergoing cardiac
surgery, with ages ranging from a couple of days old to mostly
3–4 years. We do not recommend to use the thymus from
children that are only a couple of days old since we experience
that fewer thymocytes are obtained compared to older chil-
dren and the thymocytes are less viable. The reason for this
difference is unclear. Please be aware that you must request
and obtain permission of the Ethical Commission of your
research Institute to obtain and use human cells and tissues.
6. Human hematopoietic progenitor cells can easily be frozen
and stored in liquid nitrogen for a prolonged period until
needed to start a culture experiment. However, it is highly
recommended to freeze–thaw hematopoietic progenitor cells
and especially thymocytes only once. Repeated freezing–thaw-
ing will significantly affect cell survival. Details on freezing are
provided in Chapter 20. Freezing or overnight incubation at
4 °C is not recommended for RNA analysis.
7. As an alternative to remove dead cells, the Dead Cell Removal
kit from Miltenyi Biotec can also be used.
8. After dilution, the bone marrow is filtered with a cell strainer
to remove aggregates.
9. Following centrifugation of bone marrow, a fat layer is formed
on top of the mononuclear cells. Aspirate this fat layer before
aspiration of the mononuclear cells.
10. Lymphoprep™ is used for the isolation of mononuclear cells
from cord blood and bone morrow, or for the isolation of via-
ble thymocytes after overnight incubation or after thawing of
thymus suspension. Lymphoprep™ is a density gradient
medium (1.077 g/ml). Granulocytes and erythrocytes from
cord blood/bone marrow samples have a higher density than
mononuclear cells and therefore sediment through the
Lymphoprep™ layer, while mononuclear cells are retained at
the interface (Fig. 7). For thymus suspension, dead cells sedi-
ment through the Lymphoprep™ layer and viable mononuclear
cells are retained at the interface due to their lower density.
11. The cell number is dependent on the amount of cells needed
for further analysis or experiments. Bone marrow mononu-
clear cells contain around 5 % CD34+ cells of which up to half
are CD34+CD19+. Cord blood mononuclear cells contain
1–3 % CD34+ cells.
12. When working with fewer than 108 cells, use the same volumes
as indicated. When working with higher cell numbers, scale up
all reagent volumes and total volumes accordingly.
/
dead cell fraction
Fig. 7 Lymphoprep density gradient. The different fractions are shown following
centrifugation
13. It is not recommended to incubate the cells with microbeads

on ice since this results in a lower temperature that interferes
with the binding efficiency.
14. For optimal performance, it is important to obtain a single cell
suspension before applying the cells onto the column. If
needed, pass the cell suspension through a 30 μm nylon mesh
to remove cell clumps which may clog the column. Moisten
filter with MACS buffer before use.
15. MACS buffer volume depends on the amount of cells present
before the sort. As a guideline, use at least 200 μl of MACS
buffer as lower volumes will lead to loss of a substantial frac-
tion of the cells during setup of the sort gates.
16. Be sure to count the cells following sorting as the numbers
shown by the cell sorter are never accurate and depend on the
sorting efficiency which may differ from sort to sort.
17. This volume depends on the amount of cells that were effec-
tively sorted. A minimum of 103 cells is desired to have a rep-
resentative image of the purity of the sort.
for further analysis or experiments. Only 1 % or less of total
thymocytes are CD34+ cells. This frequency increases follow-
ing overnight incubation at 4 °C or freezing and thawing, fol-
lowed by Lymphoprep density gradient.
for further analysis or experiments. CD4+ ISP thymocytes
compose around 1–3 % of total thymocytes.
20. Antibodies should be titrated to determine the optimal
amount. In general, 1 μg of monoclonal antibody per 106 cells
can be used as rule of thumb, but for thymocytes depletions,
this can be reduced to 1:4.
Fig. 8 Gating strategy to exclude doublets. The population of interest (singlets)

have an average width signal of 75,000. The formed doublets can be discrimi-
nated based on the width signal strength, that is double (150,000) in comparison
with the singlets

for further analysis or experiments. CD4+CD8+ DP thymo-
cytes constitute 70–85 % of total thymocytes, mature
CD4+CD3+ T cells around 8–14 % and mature CD8+CD3+ T
cells around 4–8 %.
22. Doublets or aggregates can be formed by two particles that
stuck together and can be seen by the sorter as a single cell.
Problems with purity can be caused by the inclusion of dou-
blets that appear as single particles, especially when sorting DP
thymocytes. The doublets can be identified and excluded by
appropriate sort gates as depicted in Fig. 8. By using signal
area (FSC-A) and width (FSC-W), it is possible to discriminate
between real double positive cells and doublets because of the
larger FSC-W signal.
for further analysis or experiments. Roughly 1–3 % of thymo-
cytes are TCR-γδ positive cells.
24. When working with fewer than 107 cells, use the same volumes
as indicated. When working with higher cell numbers, scale up
all reagent volumes and total volumes accordingly.
25. Cells labeled with the anti-TCR γδ Microbead Kit cannot be
additionally stained with another fluorochrome-conjugated
anti-TCR γδ antibody against the same epitope, since the anti-
TCR γδ Hapten antibody occupies most of the epitopes.
26. TCR-γδ antibody is not required because of the FITC-labeled
hapten-antibody in the MACS procedure.
References
1. Taghon T, Waegemans E, Van de Walle I 7. Dik WA, Pike-Overzet K, Weerkamp F et al
(2012) Notch signaling during human T cell (2005) New insights on human T cell develop-
development. Curr Top Micobiol Immunol ment by quantitative T cell receptor gene rear-
360:75–97 rangement studies and gene expression
2. Haddad R, Guimiot F, Sic E et al (2006) profiling. J Exp Med 201:1715–1723
Dynamics of thymus-colonizing cells dur- 8. Carrasco YR, Trigueros C, Ramiro AR et al
ing human development. Immunity 24: (1999) Beta-selection is associated with the onset
217–230 of CD8beta chain expression on CD4(+)
3. Hao QL, George AA, Zhu J et al (2008) CD8alphaalpha(+) pre-T cells during human
Human intrathymic lineage commitment is intrathymic development. Blood 94:3491–3498
marked by differential CD7 expression: identi- 9. Joachims ML, Chain JL, Hooker SW et al
fication of CD7- lympho-myeloid thymic pro- (2006) Human alpha beta and gamma delta
genitors. Blood 111:1318–1326 thymocyte development: TCR gene rearrange-
4. Taghon T, Rothenberg EV (2008) Molecular ments, intracellular TCR beta expression, and
mechanisms that control mouse and human gamma delta developmental potential–differ-
TCR-αβ and TCR-γδ T cell development. ences between men and mice. J Immunol
Semin Immunopathol 30:383–398 176:1543–1552
5. Weerkamp F, Baert MR, Brugman MH et al 10. Taghon T, Van de Walle I, De Smet G et al
(2006) Human thymus contains multipotent (2009) Notch signaling is required for prolifera-
progenitors with T/B lymphoid, myeloid, tion but not for differentiation at a well-defined
and erythroid lineage potential. Blood 107: beta-selection checkpoint during human T-cell
3131–3137 development. Blood 113:3254–3263
6. Ratajczak MZ (2008) Phenotypic and func- 11. Plum J, De Smedt M, Leclercq G et al (2008)
tional characterization of hematopoietic stem Human intrathymic development: a selective
cells. Curr Opin Hematol 15:293–300 approach. Semin Immunopathol 30:411–423
Chapter 20
Approaches to Study Human T Cell Development

Anne-Catherine Dolens, Inge Van de Walle, and Tom Taghon
Abstract
Not only is human T cell development characterized by unique changes in surface marker expression, but
it also requires specific growth factors and conditions to mimic and study T cell development in vitro. In
this chapter, we provide an overview of the specific aspects that need attention when performing T cell
differentiation cultures with human progenitors.
Key words Human, T cell development, OP9 coculture, Viral transduction, Gene perturbation,
In vitro, Hematopoietic stem cell, Lineage differentiation
1 Introduction
The ability to study human T cell development is of critical impor-

tance from a clinical perspective. Not only is there accumulating
evidence that there are fundamental differences between human
and mouse T cell development, there is also an increasing need to
be able to perform functional genetic experiments in a human set-
ting since the accumulating high-throughput sequencing data
reveal an endless amount of genetic variation that awaits functional
analysis. While some genetic abnormalities reveal a clear functional
defect, other variations might result in more subtle phenotypes [1].
As described in Chapter 19, the distinct developmental stages of
human T cell development can be identified using specific cell sur-
face markers that differ from mouse. However, there is increasing
evidence that also the molecular mechanisms that control this devel-
opmental process slightly differ between both species [2]. While the
molecules involved seem to be similar, the kinetics and their activity
sometimes differ between both species as exemplified by the Notch
signaling requirements [2, 3]. This may help to explain the differ-
ences in surface marker expression that also display different kinetics
with respect to developmental progression as for instance, illus-
trated for CD4 as described in Chapter 19. In addition, execution
of the full T cell developmental program starting from human
239
240 Anne-Catherine Dolens et al.
hematopoietic progenitor cells demands more time compared to

mouse precursors and this is an important aspect to consider when
studying human T cell development. In this chapter, we describe
how the OP9 coculture system can be employed to achieve efficient
human T cell differentiation, specify the growth factors and condi-
tions which are required and highlight specific aspects that require
special attention in comparison to when studying mouse T cell
development. Fetal thymus organ culture also remains an invaluable
tool to study human T cell development in vitro since this offers a
more physiological setting for developing thymocytes and even
allows the use of specific gene-deleted thymic microenvironments
[3]. However, this protocol has been described in great detail previ-
ously [4] and is therefore not described in this chapter.
2 Materials
2.1 Freezing 1. 15 ml polypropylene conical tube.

and Thawing 2. Sterile cryogenic vials.
of Human T Cell
3. Trypan blue stain 0.4 %.
Progenitors
5. Fetal bovine serum (FBS) heat-inactivated at 56 °C for 1 h
(see Note 1).
6. FBS with 20 % dimethyl sulfoxide (DMSO).
7. IMDM complete: Iscove’s Modified Dulbecco’s Medium
(IMDM), penicillin (100 U/ml), streptomycin (100 μg/ml),
L-glutamine (2 mM), 10 % heat-inactivated FBS.
2.2 Viral 1. 96-well round-bottom tissue culture plate.

Transduction 2. 96-well flat-bottom non-tissue culture plate.
of Human T Cell
3. Humidified cell culture incubator at 37 °C, 7 % CO2.
Progenitors
4. Fetal bovine serum (FBS) heat-inactivated at 56 °C for 1 h.
5. IMDM complete: Iscove’s Modified Dulbecco’s Medium
(IMDM), penicillin (100 U/ml), streptomycin (100 μg/ml),
L-glutamine (2 mM), 10 % heat-inactivated FBS.
6. Cytokines: human interleukin-7 (IL-7), human Flt-3 ligand

(Flt-3 L), human stem cell factor (SCF), human thrombopoi-
etin (TPO).
7. RetroNectin: 24 μg/ml in sterile PBS, filtered through a
0.22 μm filter. Unused solution can be stored at −20 °C.
8. Phosphate buffered saline (PBS) containing 2 % BSA, passed
through a 0.22 μm filter.
9. Recombinant retrovirus or lentivirus (homemade or commer-
cially available).
10. PBS without Ca and Mg.
In vitro T Cell Development 241
2.3 Preparation 1. Humidified cell culture incubator at 37 °C, 7 % CO2.

of OP9 Feeder Layers 2. OP9 stromal cells expressing human DLL4 (OP9-hDLL4) at
different levels (OP9-hDLL4low and OP9-hDLL4high) or
OP9 cells expressing no or other Notch ligands as described [5].
4. Medium to maintain OP9 cell lines (OP9 medium): Liquid
Minimum Essential Medium Alpha, penicillin (100 U/ml),
streptomycin (100 μg/ml), L-glutamine (2 mM), 20 % heat-
inactivated FBS.
6. 0.25 % trypsin–EDTA.
9. 24-well flat bottom tissue culture plate.
10. 100 mm × 20 mm cell culture dish.
2.4 Initiation 1. Humidified cell culture incubator at 37 °C, 7 % CO2.

and Maintenance 2. 24-well tissue culture plate containing confluent layer of OP9
of OP9 Coculture cells (see Subheading 3.3, step 4).
Experiments
4. Homemade Minimum Essential Medium Alpha (MEMα):
950 ml sterile water (see Note 2), 29.3 ml NaHCO3 (7.5 %),
10 g of alpha-MEM powder, stirred for 5–10 min. pH cor-
rected to 7.1–7.2 and osmolarity corrected to 260–285 mOsm.
Filter solution using a 0.22 μM bottle top filter.
5. OP9 coculture medium:
Homemade MEMα media, penicillin (100 U/ml), strepto-
mycin (100 μg/ml), L-glutamine (2 mM), 20 % heat-
inactivated FBS.
6. Cytokines: human interleukin-7 (IL-7), human Flt-3 ligand
(Flt-3L), human stem cell factor (SCF), human interleukin-15
(IL-15), human thrombopoietin (TPO), human granulocyte
colony-stimulating factor (GCSF), and human granulocyte-
macrophage colony-stimulating factor (GM-CSF).
2.5 Analysis of OP9 1. Flow cytometer capable of multi-color analysis (e.g., FACS
Coculture Experiments LSRII).
2. Anti-human FcR Blocking Reagent.
3. Anti-mouse FcRγII/III (clone 2.4.G2).

4. Human Wash Buffer: PBS without Ca and Mg, 1 % BSA, 0.1 %
NaN3, pH corrected to 7.2–7.3, osmolarity corrected to 275–
285 mOsm and passed through a 0.22 μm filter. Store buffer
at 2–8 °C.
5. Mouse anti-human monoclonal antibodies obtained from dif-
ferent suppliers.
6. 5 ml polystyrene round-bottom tubes.
3 Methods
3.1 Freezing 1. Human CD34+ T cell progenitors, isolated as described in

and Thawing Chapter 19, can easily be stored in liquid nitrogen and pre-
of Human T Cell served for initiating differentiation cultures when desired.
Progenitors Transfer cells into a polypropylene conical tube and determine
cell number with Trypan Blue. Cord blood progenitors and
early thymocytes can be frozen at a density of maximum
150 × 106 cells and 250 × 106 cells per vial respectively.
2. Centrifuge cells at 500 × g for 6 min at 4 °C and remove
medium.
3. Resuspend cell pellet in 1:2 volume of pure FBS.
4. Add 1:2 volume of FBS supplemented with 20 % DMSO
(see Note 3). This should be done dropwise, while mixing the
cell suspension by swirling the conical tube.
5. Aliquot cells per 1 ml into sterile cryogenic vials and put
immediately on ice.
6. Transfer cryogenic vials containing cells to a prechilled freez-
ing box at −80 °C.
7. After 24 h, transfer vials to liquid nitrogen (see Note 4).
8. To thaw the cells, remove cryogenic vials containing frozen
cells from liquid nitrogen and place in 37 °C water bath until
80 % is thawed.
9. Transfer cells to a 15 ml polypropylene conical tube on ice.
Gradually increase volume by adding IMDM complete drop-
wise while continuously shaking the tube (1 ml–2 ml–4 ml)
and put on ice for 1 min after addition of each volume.
supernatant.
11. Resuspend cells in IMDM complete and determine cell num-
ber and survival with Trypan Blue. For cord blood CD34+
precursors, recovery can be over 90 % while for CD34+ thymo-
cytes the survival rate is generally less.
3.2 Culture Prior to the transduction of progenitor cells, cytokine-stimulated

Progenitor Cells culture is essential for optimal proliferation of the cells, necessary
and Transduction for high transduction efficiencies.
1. Resuspend progenitor cells (CD34+Lin− cord blood precur-
sors or CD34+CD4−CD1−/1+ thymocyte precursors, see
Chapter 19) in IMDM complete medium supplemented with
combinations of cytokines and seed cells in 96-well round-
bottom tissue culture plate in 200 μl medium per well. Cord
blood progenitors are seeded at a density of 500 × 103 cells/ml
in medium containing 100 ng/ml SCF, 100 ng/ml Flt3-L,
and 20 ng/ml TPO, thymocyte precursors at 1 × 106 cells/ml
in medium containing 10 ng/ml SCF and 10 ng/ml IL-7.
2. Culture thymocyte precursors or cord blood progenitor cells
at 37 °C, 7 % CO2 for 24 h or 48 h respectively.
3. Transduce cells using RetroNectin™-coated 96-well flat-
bottom plates (see Note 5).
(a) Add 80 μl RetroNectin (at 24 μg/ml) per well and incu-
bate for 2 h at RT.
(b) Aspirate RetroNectin, transfer to a sterile vial, and store at
−20 °C (RetroNectin can be reused at least once).
(c) Add 80 μl PBS containing 2 % BSA and incubate for
30 min at RT.
(d) Remove PBS containing 2 % BSA and rinse the well by
adding 200 μl IMDM complete and remove
immediately.
(e) Add 100 μl viral supernatant to the well, followed by the
desired cell number in 100 μl IMDM complete with
appropriate cytokines (see Note 6).
4. 48 h after transduction, collect the cells by vigorous pipetting
and ensure that all cells are removed from the well.
5. A fraction of the cells can be used to determine transduction
efficiencies using flow cytometry (see Note 7). Remaining cells
can be sorted based on the expression of EGFP (optional,
see Note 8) and cultured on OP9 feeders or used in FTOC.
3.3 Preparation 1. Thaw OP9 cells expressing human DLL4 (OP9-hDLL4)

of OP9 Feeder Layers (see Note 9) 1 week prior to initiating coculture experiments.
(a) Prepare a 15 ml polypropylene conical tube with 10 ml
OP9 medium (RT).
(b) Thaw a vial with frozen OP9-hDLL4 cells in 37 °C water
bath.
(c) Quickly add cells to 15 ml polypropylene conical tube,
spin 500 × g for 6 min, discard supernatant.
(d) Flick pellet to help resuspension of the cells. Gently resus-

pend pellet in 10 ml OP9 medium (RT), transfer to
100 mm × 20 mm cell culture dish, and place in a humidi-
fied incubator at 37 °C and 7 % CO2.
(e) Change medium the next day.
Optional: Besides their T cell developmental capacity,
T cell progenitors, depending on their developmental
progression, can also be assayed for differentiation towards
other hematopoietic lineages. This may require different
OP9 stromal cells. We refer to ref. [5] for more details.
2. Passage the OP9 stromal cells that are not being used for
cocultures each time when the confluence reaches about
80–95 %. Do not let them reach confluency! To split OP9 cells:
(a) Carefully aspirate the medium from the cell culture dish
and wash the monolayer with 5 ml PBS to wash out the
serum.
(b) Add 2 ml 0.25 % Trypsin–EDTA, incubate for 5 min at
37 °C.
(c) Verify that all OP9 stromal cells have detached. If not,
incubate for another 5 min at 37 °C.
(d) Neutralize hydrolysis reaction by adding 4 ml of OP9
medium and transfer the cells from the culture dish into a
15 ml polypropylene conical tube.
(e) Spin at 500 × g for 6 min at 4 °C and aspirate
supernatant.
(f) Resuspend cells in 1 ml OP9 medium and determine cell
number with Trypan Blue.
3. Maintain OP9 stromal cells by seeding 200 × 103–250 × 103
cells per 100 mm × 20 mm cell culture dish in a total volume
of 10 ml OP9 medium. Split OP9 cells every 2–3 days.
4. For coculture experiments, prepare a single layer of OP9 stro-
mal cells 24 or 48 h before initiating the coculture. Therefore,
plate 20 × 103 or 10 × 103 cells, respectively, in a total volume
of 500 μl OP9 medium in a 24-well flat bottom tissue culture
plate (see Note 10).
3.4 Initiation 1. Verify using a microscope that the OP9 cells have generated a
and Maintenance confluent layer of cells and remove medium from the wells.
of OP9 Coculture 2. Resuspend the purified progenitor cells in the appropriate vol-
Experiments ume of “OP9 coculture medium” (see Note 11), supplemented
with the required cytokines. The medium volume depends on
the number of progenitor cells that will be used to initiate the
culture (see Table 1). Supplemented cytokines depend on the
hematopoietic lineage being studied (see Table 2).
Table 1
Maximum number of cells from different progenitor populations for
initiation of cocultures onto different well types
Plate Medium Number of cord Number of pediatric

format volume (μl) blood progenitors thymus progenitors
96-well 150 1–300 1–500
48-well 300 300–5 × 103 500–7 × 103
24-well 500 5 × 103–10 × 103 7 × 103–20 × 103
Table 2
Different cytokine cocktails for the generation of different cell types
T cell mix NK cell mix B cell mix Myeloid cell mix

5 ng/ml SCF 5 ng/ml SCF 20 ng/ml SCF 20 ng/ml SCF
5 ng/ml Flt3-L 5 ng/ml Flt3-L 20 ng/ml Flt3-L 20 ng/ml Flt3-L
5 ng/ml IL-7 5 ng/ml IL-7 20 ng/ml TPO 20 ng/ml TPO
10 ng/ml IL-15 10 ng/ml GM-CSF
10 ng/ml GCSF
3. Seed the resuspended progenitor cells onto the confluent OP9

cells.
4. Place the coculture plate in an incubator at 37 °C, 7 % CO2.
5. Check cocultures every 3 days. If cell density is to high due to
expansion of the progenitor cells or if OP9 cells start to detach,
split coculture progenitor cells by transferring them to new
OP9 feeder cells (see Note 12).
6. Harvest cocultures at different time points to split cocultures
or to perform flow-cytometric analysis.
(a) Disaggregate the cells by forceful pipetting (see Note 13)
and transfer them to a 15 ml polypropylene conical tube.
If necessary, rinse well with sterile PBS to remove any
remaining cells and transfer to the same 15 ml polypropyl-
ene conical tube.
(b) Centrifuge at 500 × g for 6 min at 4 °C.
(c) Carefully remove the supernatant and resuspend the cell
pellet in 500 μl pre-warmed OP9 coculture medium with
appropriate cytokines.
(d) Determine cell number with Trypan Blue.
(e) Transfer 1:2 or maximum 100 × 103 cells to new wells of a

24-well tissue culture plate (if sufficient cells, otherwise
use 96- or 48-wells) that contain a fresh confluent layer of
OP9 cells for further coculture.
(f) The rest of the cells can be used for flow-cytometric analy-
sis (see Subheading 3.5).
7. Repeat steps 5 and 6 until the cells have sufficiently differenti-
ated into the T cell lineage pathway as evident from the flow-
cytometric analysis (see Subheading 3.5). Given that human T
cell development progresses with slower kinetics compared to
in mouse, it is not trivial to maintain the cultures for a pro-
longed time. Timing of analysis depends on the starting popu-
lation. When cocultures are initiated with CD34+CD4−CD1−/+
thymocytes, it is recommended to analyze the cells at day 6,
12, 19, and 25 after initiating coculture. At this last time point,
cells will no longer expand and most will have successfully dif-
ferentiated into either TCR-γδ or TCR-αβ T cells. When
cocultures are initiated with cord blood or bone marrow
derived CD34+Lin− precursors, analysis is performed at day
12, 19, 26, 33, 40, 40, 55 of coculture. A representative image
of the kinetics of human T cell development form CD34+Lin−
cord blood precursors is given in Fig. 1.
3.5 Analysis of OP9 1. Before labeling, centrifuge the cells for 6 min at 500 × g.
Coculture Experiments 2. Discard the supernatant.
3. Resuspend the cells in 100 μl human wash buffer and pre-
incubate on ice for 5 min with 1 μl anti-human FcR Blocking
Reagent and 2 μl anti-mouse FcRγII/III 20 μg/ml (clone
2.4.G2) to avoid nonspecific binding.
Fig. 1 Kinetic analysis of human T cell development from CD34+Lin− cord blood progenitors on OP9-DLL4low
vs. OP9-DLL4high cells. Dot plots also show effects of low (OP9-DLL4low) versus high (OP9-DLL4high) levels
of Notch activation on human T cell development as discussed in more detail elsewhere [2]. Early specification
towards the T cell lineage is favored by high Notch activity and can be tracked by the upregulation of CD7 while
CD34+ expression is maintained. Further differentiation is characterized by the upregulation of CD5 and down-
regulation of CD34. Addition of HLA-DR to the panel is important to discriminate T cell progenitors from mono-
cytic/dendritic cells. While CD34+ cells coexpress HLA-DR, also following upregulation of CD7 (green cells),
CD7+CD5+ T cell progenitors will downregulate CD34+ as well as HLA-DR (blue cells), while CD5+ cells also
contain HLA-DR+CD4+ monocytic/dendritic cells (red cells). Thus, CD5 or CD7 are insufficient as sole markers
for early T cell progenitors. Further differentiation into CD4+CD8β+ double positive thymocytes and TCR-αβ+
CD3+ T cells is more efficient when Notch activity is reduced (OP9-DLL4low) but requires over 50 days
of coculture to develop. TCR-γδ T cells preferentially develop in conditions of high Notch signaling activity
(OP9-DLL4high) (color figure online)
CD7+CD5- CD7+CD5+ CD7-CD5+

Table 3
Antibody panels to study different early stage of human T cell development in vitro (see Note 14)
Early T cells Double positive cells TCRαβ/TCRγδ T cells Myeloid cells B cells
PI PI PI PI PI
CD45 CD45 CD45 CD45 CD45
(HI30) (HI30) (HI30) (HI30) (HI30)
CD34 CD8b TCRαβ CD4 CD19
(AC136) (2ST8.5H7) (BW242/412) (RPA-T4) (LT19)
CD7 CD8 TCRγδ HLA-DR HLA-DR
(M-T701) (SK1) (11F2) (LN3) (LN3)
CD5 CD4 CD3 CD14
(UCHT2) (RPA-T4) (UCHT1) (TUK4)
CD1a CD3 CD4 CD15
(HI149) (UCHT1) (RPA-T4) (VIMC6)
CD4 HLA-DR CD8b CD123
(RPA-T4) (LN3) (2ST8.5H7) (6H6)
HLA-DR CD1a CD303
(LN3) (HI149) (AC144)
Antibody clones are indicated in italicized parenthesis
4. Transfer 10–100 × 103 cells into 5 ml polystyrene round-

bottom FACS tube for each staining.
5. Cells are stained with panels of anti-human monoclonal anti-
bodies according to the guidelines of the manufacturer.
Antibody panels to study different early stages of T cell devel-
opment are represented in Table 3.
6. Incubate for 30–45 min at 4–6 °C, in the dark.
7. Wash the cells by adding 2 ml prechilled human wash buffer
per sample. Centrifuge at 500 × g for 6 min at 4 °C and remove
the supernatant. Keep the cells on ice and covered from day-
light until flow-cytometric analysis.
8. A representative image of the kinetics of human T cell devel-
opment form CD34+Lin− cord blood precursors is given
in Fig. 1.
4 Notes
1. Technical disadvantages to using serum include the undefined

nature of serum, batch-to-batch variability in composition,
and the risk of contamination. Therefore it is recommended to
perform serum testing. To test for T cell development, we
culture fetal thymic lobes (fetal day 14) of B6 mice in fetal

thymus organ culture for 14 days and check for efficiency to
support T cell development by cell counting and flow-
cytometric analysis of T cell differentiation.
2. Use a 1-L glass graduated cylinder for the preparation of
homemade MEMα. Do not use the cylinder for any other pur-
pose and do not clean with any detergent! After each use, rinse
the cylinder with milliQ H2O and then with 100 % EtOH,
autoclave occasionally. Store homemade MEMα at 4 °C and
use in cocultures for maximum 4 weeks after preparation!
3. For the preparation of FBS supplemented with 20 % DMSO,
DMSO should be added dropwise to FBS while keeping the
solution on ice to minimize the exothermic reaction.
4. DMSO is a cryoprotective agent, which slows down the cool-
ing rate, reducing the risk of ice crystal formation. The gradual
decrease in temperature, by temporarily storing the cells at
−80 °C, improves amorphous freezing. However, DMSO is
toxic to cells and should be added dropwise to minimize the
exothermic reaction that occurs when DMSO is added to water.
5. Large number of cells can be transduced using 24-well flat-
bottom RetroNectin™-coated plates. Therefore, volumes
need to be adjusted using the following protocol:
Add 320 μl RetroNectin 24 μg/ml per well and incubate for
2 h at RT.
Aspirate RetroNectin, transfer to a sterile vial and store at
−20 °C (can be used once again).
Add 320 μl PBS containing 2 % BSA and incubate for 30 min
at RT.
Remove PBS containing 2 % BSA and rinse well using 600 μl
IMDM complete.
Add 400 μl viral supernatant and 400 μl of precultured cells,
correct cytokine concentration as viral supernatant do not
contain cytokines.
6. Be sure to add sufficient cytokines for compensating the viral
supernatant that does not contain cytokines. Cytokine con-
centrations from the preculture should remain stable during
the complete transduction procedure.
7. It is recommended to use EGFP (or another fluorescent
marker gene) expressing retroviruses or lentiviruses (such as
LZRS or MSCV) to monitor transduced cells during the sub-
sequent T cell developmental assays. Transduction efficiency
peaks around 48 h after adding the virus to the cells.
8. To initiate cocultures with transduced cells, it is recommended
to start with a 100 % sorted EGFP+ transduced cell population
as this is the most accurate way to calculate cell numbers.

While a mixture of EGFP− and EGFP+ cells has the advantage
of containing an internal control for evaluating the efficiency
of the cultures (EGFP− cells should behave similarly for each
virus being tested), it is more difficult to track the fate of the
transduced cells since gene perturbations that lead to a nega-
tive effect on cell growth and/or differentiation may lead to a
transition of EGFP+ to EGFP− cells, making it difficult to make
accurate calculations.
9. DLL4 is the physiological Notch ligand that induces T cell
development in vivo and is therefore recommended although
we do not see major differences between DLL4 and DLL1.
We also don’t see any major differences in effects on human T
cell development when using OP9 cells that express a human
or mouse Notch ligand. Effects of other Notch ligands have
been described [3, 5] and can be used depending on the par-
ticular experiment.
10. Coculture experiments can be performed with 96-well, 48-well
or 24-well plates. It is important that OP9 stromal cells reach
100 % confluency before adding lymphoid progenitor cells to
ensure close interaction between the lymphoid progenitor
cells and the Notch ligand expressing OP9 cells.
11. It is essential to use OP9-coculture medium that is prepared
with fresh, homemade MEMα medium since MEMα that is
purchased as a liquid seems to have lost an unknown compo-
nent that is important for efficient human T cell development.
Therefore, the freshly prepared medium cannot be used for
more than 4 weeks to ensure reproducible results.
12. The growth rate of human hematopoietic progenitor cells is
highly variable due to the donor variety. As a result, cultures need
to be monitored continuously to avoid that cultures become too
dense as this will result in cell death. Cord blood or bone marrow
derived CD34+ progenitor cells usually display limited expansion
during the first week of OP9-hDLL4 coculture.
13. Forceful pipetting of OP9 cocultures will result in the detach-
ment of both OP9 and human hematopoietic progenitor cells.
There is currently no efficient protocol that allows the selec-
tive isolation of the hematopoietic cells without OP9 contami-
nation. This, however, does not affect downstream analysis or
further culture of the cells when replated on a fresh monolayer
of OP9 cells.
14. An anti-human CD45 monoclonal antibody is added to each
tube to distinguish human hematopoietic cells from OP9 stro-
mal cells. Alternatively, anti-mouse CD51 can also be used to
stain OP9 stromal cells. Propidium Iodide (PI) is added to
exclude dead cells.
References
1. Weinacht KG, Brauer PM, Felgentreff K et al control human TCR-αβ/γδ development by
(2012) The role of induced pluripotent inducing differential Notch signal strength. J Exp
stem cells in research and therapy of primary Med 210:683–697
immunodeficiencies. Curr Opin Immunol 4. Verhasselt B, Naessens E, Stove V (2003)
24:617–624 Generation of transgenic T cells from human
2. Taghon T, Waegemans E, Van de Walle I CD34+ cord blood cells. Methods Mol Biol
(2012) Notch signaling during human T cell 215:327–339
development. Curr Top Micobiol Immunol 5. Van de Walle I, De Smet G, Gärtner M et al
360:75–97 (2011) Jagged2 acts as a Delta-like Notch
3. Van de Walle I, Waegemans E, De Medts J et al ligand during early hematopoietic cell fate deci-
(2013) Specific Notch receptor-ligand interactions sions. Blood 117:4449–4459
Chapter 21
Humanized Mice to Study Human T Cell Development

Sarah Bonte, Sylvia Snauwaert, Stijn Vanhee, Anne-Catherine Dolens,
Tom Taghon, Bart Vandekerckhove, and Tessa Kerre
Abstract
While in vitro models exist to study human T cell development, they still lack the precise environmental
stimuli, such as the exact combination and levels of cytokines and chemokines, that are present in vivo.
Moreover, studying the homing of hematopoietic stem (HSC) and progenitor (HPC) cells to the thymus
can only be done using in vivo models. Although species-specific differences exist, “humanized” models
are generated to circumvent these issues.
In this chapter, we focus on the humanized mouse models that can be used to study early T cell devel-
opment. Models that study solely mature T cells, such as the SCID-PBL (Tary-Lehmann et al., Immunol
Today 16:529–533) are therefore not discussed here, but have recently been reviewed (Shultz et al., Nat
Rev Immunol 12:786–798).
Key words Immunodeficient mice, Humanized mice, In vivo mouse model, T cell development, BLT
mouse model, SCID-Hu mouse model
1 Introduction
1.1 Mouse Strains Engraftment of human hematopoietic cells and fetal hematopoietic
Used to Study In Vivo tissues was initially described in severe combined immunodeficient
T Cell Development (SCID) mice [1–3]. Due to the Prkdcscid (protein kinase, DNA
activated, catalytic polypeptide) mutation, these mice lack func-
tional T- and B-lymphocytes, allowing higher and more diverse
engraftment than was described earlier in nude and beige-nude-
Xid (bnx) mice, which support T cell but not B cell development
[4–6]. However, engraftment of human cells in SCID mice was
still suboptimal, owing to high levels of host natural killer (NK)
cell activity and innate immunity. Another way to delete T and B
cells in the murine host, is to target one of the recombination acti-
vating gene 1 and 2 (Rag1 and Rag2) genes, crucial for rearrange-
ments of both T and B cell receptor genes during T and B cell
development [7, 8]. However, these mice also retain high NK cell
253
254 Sarah Bonte et al.
activity that leads to low human cell engraftment. The scid mutation,
as compared to the Rag mutation, also results in defective DNA
repair (as the Prkdc gene encodes for a polypeptide important in
repair of double-stranded DNA breaks) and, consequently, higher
radiosensitivity [9].
The development of nonobese diabetic (NOD)-scid mice, by
crossing the scid mutation onto the NOD background, led to the
next breakthrough in the development of immunodeficient mouse
models. Additional defects in innate immunity in NOD-scid mice,
lacking hemolytic complement and showing functional defects
in both macrophages and NK cells, led to higher engraftment
of human hematopoietic cells, compared to scid mice [10–12].
However, the residual activity of both NK cells and macrophages
in NOD-scid mice still limited human hematopoietic engraftment,
and especially T cell development was very inefficient. To address
this problem, existing models were adapted, e.g., by injecting mice
with TM-β1, an IL-2 receptor beta chain (IL2Rβ) blocking anti-
body [13–15], further decreasing NK cell activity in these mice,
allowing the engraftment of T cell precursors in the murine thymus
with in vivo T cell development in about 50 % of the cases. The
major drawback of this model was that large amounts of antibody
and several intraperitoneal injections of the mice were necessary.
Later, new mouse strains lacking a functional IL-2 receptor gamma
chain (Il2rgnull) were developed such as NOD/Shi-scid Il2rgnull
(NOG) [16], NOD/LtSz-scid Il2rgnull (NSG) [17], BALB/c--
Rag2null Il2rgnull (BRG) [18], NOD/Rag1null Il2rgnull (NRG) [19].
Since the IL2Rγ chain is an important component of the receptors
for IL-2, IL-4, IL-7, IL-9, IL-15, and IL-21, these mice com-
pletely lack NK-cell development, as well as functional T- and
B-lymphocytes, features that lead to superior human hematopoi-
etic engraftment and allow for T cell development in the murine
thymus in a near to 100 % frequency. An important disadvantage
of the NOD-scid strain is the relatively short life span of these mice
due to thymoma development [20], thereby not allowing evalua-
tion of the injected and/or in vivo generated cells for an extended
period of time. Il2rgnull mice, however, have a longer median life
span (more than 90 weeks for NSG mice compared with 37 weeks
for NOD-scid mice) and thymoma development is very rare to
nonexistent [21]. Also in mice lacking beta2-microglobulin
(B2m−/−) (NOD/SCID/B2m−/−) [22] B, T, and NK cell develop-
ment and functions are disrupted, as β2m is necessary for major
histocompatibility complex (MHC) class I-mediated immunity.
These and other immunodeficient mouse strains are extensively
reviewed in ref. 21.
The thymic microenvironment and especially the thymic epi-
thelial cells support positive and negative selection processes through
interaction of MHC molecules on the epithelial cells with the TCR
Humanized Mice to Study Human T Cell Development 255
of the developing T cell. Therefore, in the above-mentioned models,

human T cells are selected on murine MHC molecules. To over-
come this hurdle, humanized mouse models have been developed
in which a human thymus is transplanted in an immunodeficient
mouse (see Subheading 1.2 SCID-hu and BLT models).
Currently, new immunodeficient mouse strains are being devel-
oped that allow differentiation of fully functional T-lymphocytes
from injected human HSC after selection on human MHC mole-
cules, without the need for transplantation of human organs. Such
strains include mice that lack MHC class I or MHC class II mole-
cules and that are engineered to express human MHC class I and/
or II molecules: HLA-A2 (NSG-A2, [23]), HLA-DR (NOG/
HLA-DR4, [24]), and both HLA-A2 and HLA-DR (NSG-A2/
DR1, [25]). In this way, xenogeneic graft-versus-host disease
(GVHD) is reduced and the development of human MHC-
restricted T-cells, able to recognize antigens presented by human
MHC-expressing antigen-presenting cells (APCs) in the periphery,
in the absence of mouse MHC-restricted T-cells can be achieved
using the Hu-HSC model [26] (see Subheading 1.2). The draw-
back here is that only T cells expressing a HLA-A2 and/or HLA-DR
restricted TCR are being selected, and there is a need for mouse
strains expressing other human MHC molecules, both class I and
class II.
Also, transgenic mice engineered to express human cytokines
are able to enhance engraftment and/or proliferation of human
hematopoietic cells. Engraftment of especially HSC and T lympho-
cytes is improved in stem cell factor (SCF)-expressing mice [27],
whereas human interleukin 7 (IL-7) (after retro-orbital injection of
a hIL-7 lentiviral supernatant into BRG mice) enhances homeo-
static proliferation of human T cells [28].
An overview of the available mouse strains to study T cell
development in a humanized mouse model is given in Table 1.
T cell development and HSC engraftment scores are based on
published data but a side by side comparison of the different strains
has not yet been performed, and the scores should be treated as such.
1.2 Immunodeficient Such immunodeficient mice can be reconstituted with human

Mouse Models progenitors and organs to generate humanized mice, according
to several outlines. The Hu-HSC model, where human HSC are
adoptively transferred in immunodeficient mice, was primarily
developed to study and characterize human HSC, as it is the only
model in which all properties of HSC (multilineage development,
self-renewal, high proliferative capacity and ability to reconstitute)
can be evaluated [12]. This model revealed the SCID repopulating
cell or SRC, as the ultimate definition of the HSC. After transfer of
HSC in the Hu-HSC model, these cells undergo multilineage dif-
ferentiation, including both T and B cell development. Multipotent
progenitors will home to the thymus, where they are educated on
Table 1
Currently available mouse strains optimally supporting human T cell development
(a) Basic
T cell HSC
Mouse strain Alternative name Abbreviation Background Genetic defect development engraftment Refs
NODShi.Cg-PrkdcscidIl2rgtm1Sug NOD/Shi-scid Il2rgnull NOG NOD/SCID IL-2Rγ chain lacks +++ ++ [16]
IC domain
NOD.Cg-PrkdcscidIl2rgtm1Wjl NOD/LtSz-scid Il2rgnull NSG NOD/SCID Null mutation of +++ +++ [17]
IL-2Rγ
NOD.Cg-Rag1tmMomIl2rgtmWjl NOD/Rag1null Il2rgnull NRG NOD Null mutation of +++ +++ [19]
Rag1 and IL-2Rγ
C.Cg-Rag2tmFwaIl2rgtmSug BALB/c-Rag2nullIl2rgnull BRG Balb/c Null mutation of ++ ++ [18]
Rag2 and IL-2Rγ
chain lacks IC
domain
t
C.Cg-Rag1tm1MomIl2rg mWjl BALB/c-Rag1nullIl2rgnull BRG Balb/c Null mutation of ++ ++ [19]
Rag1 and IL-2Rγ
chain lacks IC
domain
C;129S4-Rag2tmFlvIl2rgtm1Flv/J / BRG or Mixed (Balb/c x (Null mutation of ++ ++ [29]
C;129RG 129Rag2null) Rag2) Null
mutation of
IL-2Rγ chain
(b) Supplementally engineered mice to enhance T cell development
Supplemental
Background Additional genetic effects on T cell
Mouse strain Abbreviations (see ref. 30) defect development Refs
NOD.Cg-PrkdcscidIl2rgtm1WjlTg(HLA-A/H2-D/B2M)1Dvs NSG Tg(HLA-A2/ NSG Transgene HLA-A2 Supports [31]
B2M) and human development of [32]
NSG-A2/B2M β2-microglobulin human HLA-A2
restricted T cells
NOD.Cg-PrkdcscidIl2rgtm1WjlTg(HLA-A2.1)1Enge NSG Tg(HLA-A2) NSG Transgene HLA-A2 Supports [23]
NSG-A2 development of
human HLA-A2
restricted T cells
NOD.Cg-Rag1tm1MomIL2rgtmWjlTg(HLA-DRB1)31Dmz NSG Tg NSG Transgene Supports [33]
(HLA-DR4) HLA-DR4 development of
NSG-DR4 human HLA-DR
restricted T cells,
enhanced
immune function
NOD.Cg-PrkdcscidH2-Ab1tm1DolIl2rgtm1SugTg(HLA-DRA, NOG NOG Transgene Supports [24]
HLA-DRB1*0405)1Jic/Jic Tg(HLA-DR4) HLA-DR4 development of
NOG-DR4 human HLA-DR
restricted T cells,
enhanced
immune function
NOD.Cg-PrkdcscidIl2rgtm1Wjl/Sz Tg(HLA-A2*0201/ NSG Tg(HLA-A2/ NSG Transgene HLA-A2 Supports [25]
HLA-DRA*0101, HLA-DRB1*0101) DR1) and HLA-DR1 development of
NSG-A2/DR1 human HLA-A2
and HLA-DR
restricted T cells,
enhanced
immune function
Abbreviation: IC intracytoplasmic; nc not compared
Adapted from [26]
murine epithelial cells, and therefore will be tolerant to the mouse

host, resulting in a low frequency of GVHD [21]. However, the
generated T cells will be mouse MHC-restricted instead of human
MHC-restricted [26]. As mentioned above, newer mouse strains
with deleted mouse MHC molecules and transgenic expression of
human MHC molecules (see Table 1b) have answered this issue.
For the Hu-HSC model, HSC from several sources have been
used, from fetal liver [34], umbilical cord blood [16, 18, 35], bone
marrow [34], or granulocyte-colony stimulating factor (G-CSF)-
mobilized peripheral blood [17], all leading to acceptable levels of
engraftment. Originally, HSC were injected intravenously (i.v.)
into adult hosts, but engraftment was improved when the cells
were delivered intrahepatically (i.h.) into newborn mice (probably
due to a more tolerant environment) or after intrafemoral (i.f.)
injection into adult recipients. Intrafemoral injection circumvents
the need for homing of HSC to the bone marrow, but therefore
does not allow evaluation of homing characteristics.
To enhance the selection process of human thymocytes, the
SCID-Hu mouse model, wherein human fetal liver and thymus are
co-implanted under the renal capsule of SCID mice (SCID-Hu
thy/liv) [36], or human bone marrow is implanted subcutaneously
in SCID mice (SCID-Hu bone) [37], can be employed. Several
reports have demonstrated that this is a useful and relevant in vivo
system to study the developmental potential of transplantable
human hematopoietic progenitor and stem cells [38–40]. The
advantage of this assay is the presence of a human microenviron-
ment; the obvious disadvantage is the necessity for fetal material
and the lack of quantitative data. Moreover, although it is suitable
for studying T cell development, only low levels of hematopoietic
cell lineages other than thymocytes can be generated and the over-
all human immune system function is poor.
To overcome these hurdles, the bone marrow/liver/thymus
(BLT) model was developed. This model is in origin the same as
the SCID-Hu model, but, in addition, human HSC isolated from
the same fetal liver are injected i.v. and engraft the bone marrow,
resulting in a complete human immune system arising in the
immunodeficient mouse [41, 42].
An overview of the different humanized mouse models
supporting T cell development is given in Fig. 1 and Table 2.
Depending on their research goal, investigators need to choose the
appropriate humanized mouse model for their studies. All above-
mentioned strains and models, including advantages and disadvan-
tages, are reviewed in refs. 26 and 21. In terms of establishment of
engraftment of HSC and resulting T cell development, NSG mice
have been proven to be superior compared to other mouse strains,
especially at limiting HSC doses [43]. Moreover, female NSG mice
show higher engraftment levels than male NSG mice [44].
BLT SCID-Hu
fetal liver fetal liver

HSC +
and thymus and thymus
under kidney
intravenous
capsule
femur liver
kidney
intravenous intrafemoral intrahepatic

adult newborn
HSC
Hu-HSC
Fig. 1 Schematic overview of humanized mouse models
Human peripheral blood lymphocytes (PBL) can also be

injected when the function of mature T cells is being studied.
As this is beyond the scope of this chapter, which focuses on T cell
development, we refer to other reviews for the Hu-PBL model
[21, 30].
1.3 Methods Immunodeficient mice are usually preconditioned using total-body

to Enable Engraftment sublethal γ-radiation before injection of HSC to enable optimal
in Mice engraftment in these mice. The purpose of this irradiation is to
eliminate residual immune cells, to induce the production of cyto-
kines (improving engraftment and differentiation) and to create
space for the introduced HSC. The level of the employed radiation
differs, depending on the age of the mice (newborn or adult) and
the strain used (scid or Rag mutation). Newborn mice are more
sensitive than adult mice to irradiation, and as the scid mutation
leads to heightened radiosensitivity, a lower dose of irradiation
given to NSG or NOG mice causes the same effect as a higher dose
in BRG mice. Typically, mice are irradiated using a 137Cs gamma
irradiator, but any radiation source can be used when applying set-
tings specifically developed for mouse irradiation. Newborn mice
harboring the scid mutation (NSG, NOG) are typically irradiated
with 100 cGy [45], while newborn BRG mice can be irradiated
with up to 400 cGy [18]. Adult NSG or NOG mice tolerate radiation
Table 2
Humanized mouse models supporting T cell development in vivo
Humanized Organs showing

mouse model Thymic T cell
(abbreviated) Brief description microenvironment reconstitution Advantages Limitations Refs.
Hu-HSC Injection of human − Murine (selected Murine thymus, − Multilineage development (B/T cells, − T cells educated [12]
HSC (from FL, CB, on mouse MHC) blood, spleen, myeloid cells, NK cells) on murine TEC [16]
BM or mPB) i.v. or − Murine with bone marrow, − Higher T cell engraftment in newborn and murine [17]
i.f. in adult and i.h. in HLA expression lymph nodes mice MHC restricted [18]
newborn irradiated (class I and/or (seldom) − Low engraftment [34]
immunodeficient mice class II) only in levels of some [35]
HLA-tg mice hematopoietic
lineages in blood
SCID-Hu Co-implantation of Human Transplanted − T cells educated on human autologous − Specialized [36]
human FL and FT thymic organ TEC and human MHC restricted transfer method
(of the same donor) − Availability of large amounts of human (surgery)
under the kidney T cells in thymus − Human fetal
capsule material required
− Low engraftment
levels of
hematopoietic
lineages other
than thymocytes
BLT After establishment of Human (and Transplanted − Complete human immune system − Specialized [41]
SCID-Hu murine) thymic organ, − T cells educated on human autologous transfer method
murine thymus, TEC and human MHC restricted (surgery)
blood, spleen, − Availability of large amounts of human − Human fetal
bone marrow, T cells in thymus material required
lymph nodes − Higher levels of total human HSC
(seldom) engraftment than in Hu-HSC
Abbreviations: Hu human; SCID severe combined immunodeficiency; BLT bone marrow/liver/thymus; FL fetal liver; CB cord blood; BM bone marrow; mPB mobilized peripheral
blood; i.v. intravenous; i.f. intrafemoral; i.h. intrahepatic; MHC major histocompatibility complex; HLA human leukocyte antigen; TEC thymic epithelial cells
Adapted from [26]
doses up to 400 cGy [17] but mostly 200–300 cGy is used [46–49].
For adult BRG mice, radiation doses of 500–600 cGy are used
[50]. The only mouse strain that seems to overcome the need for
irradiation, and therefore might provide a more physiological envi-
ronment, is the newborn NSG mouse expressing membrane-bound
SCF [27] and the adult Rag2−/−Il2rg−/−KitWv/Wv mouse [51], in
which the mutant Kit receptor opens up the stem cell niches across
species barriers [52]. As human lymphoid development in the
Rag2−/−Il2rg−/−KitWv/Wv mouse has not yet been published, this
mouse strain is not discussed above.
An alternative method that substitutes irradiation is injection
of the mice with the chemotherapeutic agent Busulfan. Choi et al.
[53] showed that this leads to a higher survival rate and enhanced
T cell development.
Other engraftment-enabling methods that have previously
been used are injection of the mice with TM-β1, before injection
of human cells [13], a strategy which has been made redundant by
the newer strains, lacking IL2Rγ or β2m, and in vivo injection of
various recombinant human cytokines, such as IL-15, IL-7, and
SCF, after injection of human cells [54–56].
Also, intrafemoral injection of HSC can be considered as a
method to enable engraftment, as injection of HSC directly into
the bone marrow circumvents the need for homing of these cells to
the bone marrow and therefore allows for immediate engraftment.
2 Materials
2.1 Common 1. Phosphate-buffered saline (PBS), sterile.

Materials 2. 137
Cs gamma irradiator (see Note 1).
3. Sterile, autoclaved housing device for mice during irradiation
(see Note 2).
4. Sterile alcohol swab or 70 % ethanol spray.
2.2 CD34+ Isolation 1. Hematopoietic stem cell (HSC) source (umbilical cord blood,
mobilized peripheral blood, bone marrow, or fetal liver)
(see Note 3).
2. Lymphoprep or Ficoll, stored at room temperature.
3. Method to isolate CD34+ cells (see Note 4).
2.3 Intrahepatic 1. Immunodeficient mice breeding pairs (see Note 5).

Injection 2. 1-cc tuberculin syringes with 25-G needle or 1-cc insulin
syringe with 29-G needle.
3. Optional: topical nasal decongestant.
4. Ice bucket with ice.

5. Sterile gauze.
6. Sterile 100 mm petri dishes.
7. Heat lamp or heating device.
2.4 Intravenous 1. Immunodeficient mice: male or female, 6–10 weeks old

Injection (see Note 5).
2. 1-cc tuberculin syringes with 25-G needle.
3. Heat lamp or heating device.
2.5 Intrafemoral 1. Immunodeficient mice: male or female, 6–10 weeks old

Injection (see Note 5).
2. Anesthetic and/or anesthesia equipment (according to institu-
tional guidelines).
3. Surgical instruments: curved forceps, scalpel, wound closure
staple and clip or suture, needle holder, suture scissors
(see Note 6).
4. Hamilton syringe (Hamilton, Bonaduz, Switzerland).
5. Hair clipper.
6. Bone wax (Braun, Diegem, Belgium).
7. 25-G needle.
2.6 Generation 1. 18-G hypodermic needle.

of a Transplantation 2. 18-G spinal needle.
Needle
3. Sandpaper, very fine.
2.7 Subcapsular 1. SCID mice (preferably male, 6–8 weeks old) or NOD.
Kidney Transplant Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NOD/LtSz-scid Il2rg−/−) mice
(male or female, 6–10 weeks old) (see Note 5).
2. Human fetal liver and fetal thymus of 14–23 weeks of gesta-
tional age (preferably HLA-typed, see Note 7) (see Note 8).
3. Sterile PBS supplemented with 50 μg/ml gentamicin (Gibco,
Invitrogen, Merelbeke, Belgium).
4. Sterile RPMI-1640 medium (Gibco, Invitrogen, Merelbeke,
Belgium) supplemented with 50 μg/ml gentamicin.
5. Anesthetic and/or anesthesia equipment (according to institu-
tional guidelines).
6. Special injection device/transplantation needle (see Note 9).
7. Surgical instruments: scalpel, microdissecting scissors, Dumont
tweezers, microdissecting forceps, microdissecting scissors,
wound closure staple and clip and/or absorbable suture, nee-
dle holder and suture scissors, scissors and toothed tweezers
(see Note 6).
8. Sterile 100 mm petri dishes.

9. Hair clipper.
10. 10 ml syringe with 18-G needle.
11. Heat lamp or heat bags.
12. Optional (BLT model): sterile RPMI-1640 supplemented with
1 mg/ml collagenase/dispase (Roche, Mannheim, Germany)
and 0.5 U/ml DNase I (Roche).
13. Optional (BLT model): 70 μm cell strainer, sterile.
14. Optional (BLT model): 1-cc tuberculin syringes with 25-G
needle.
3 Methods
3.1 CD34+ Isolation 1. Isolate mononuclear cells from HSC source using Lymphoprep
or Ficoll density gradient centrifugation (see Chapter 19 by
K. Davids et al. for details).
2. From the obtained mononuclear cells, isolate CD34+ cells
(see Note 4 and Chapter 19 by K. Davids et al.).
3. Resuspend the isolated CD34+ cells in sterile PBS at the desired
cell density (volumes and cell numbers typically used are discussed
in Note 10) and keep on ice until transplanted (see Note 11).
3.2 Mice Irradiation 1. Place mice in sterile autoclaved housing device for irradiation
(see Notes 12 and 13 for additional information on handling
newborn mice).
2. Irradiate mice with the appropriate amount of total-body irra-
diation (see Notes 1 and 14).
3.3 Intrahepatic 1. Monitor breeding pairs for the birth of new litters (this can be
Injection in Newborn estimated using timed matings). The procedure should be per-
Mice formed on 1- (not earlier to avoid rejection by the mother) to
4- (not later as engraftment efficiency will decrease) day old
newborn pups.
2. Isolate CD34+ as described in Subheading 3.1.
3. Irradiate pups as described in Subheading 3.2. Pups are placed
back with their mother afterwards.
4. Wait for at least 2 h and not more than 24 h before starting
with injection procedure.
5. Place irradiated pups on a sterile gauze pad in a sterile petri
dish on ice for 5–10 min until anesthetized (when gross move-
ment ceases).
6. Load the CD34+ cell suspension into a 1-cc tuberculin syringe
with 25-G needle or a 1-cc insulin syringe with 29-G needle.
7. Swab abdomen of mice with sterile alcohol swab.
8. Deliver 50 μl of cell suspension directly into the liver (see Notes

15 and 16).
9. Place the pups under a heat lamp for 1–2 min (see Note 17).
10. Optional: before putting the pups back with the mother, apply
a small amount of topical nasal decongestant to the snout of
the mother (see Note 18).
11. Pups can be weaned from the mother at 3–4 weeks of age and
human cells can be detected in the blood from 4 weeks after
HSC injection on. Full evaluation is optimally done at 10–12
weeks of age by sacrificing mice and collecting organs (bone
marrow, thymus, liver, spleen, and blood by cardiac puncture).
A representative flow cytometric analysis of T cell reconstitu-
tion after intrahepatic injection of human CD34+ cells from cord
blood in newborn NSG and Rag−/− mice at 12 weeks after injec-
tion is shown in Fig. 2.
3.4 Intravenous 1. Isolate CD34+ as described in Subheading 3.1.

Injection in Adult Mice 2. Irradiate mice as described in Subheading 3.2.
3. Wait for at least 4 h and not more than 24 h before injecting
mice with HSC.
4. Put mice with their tail under a heat lamp or heating device to
cause vasodilatation (see Note 17). Mice can be immobilized
using restraining devices.
5. Swab tail vein with sterile alcohol swab.
6. Inject 200 μl of cell suspension into the lateral tail vein using a
1-cc syringe with 25-G needle.
7. Depending on the research question, different analyzing time
points can be set. Evaluation of homing of HSC can be done
immediately after injection of the cells. T cell development can
be analyzed from 3 to 4 weeks after injection on. Full evalua-
tion is optimally done at 10–12 weeks after HSC injection by
sacrificing mice and collecting organs (bone marrow, thymus,
liver, spleen, and blood by cardiac puncture).
3.5 Intrafemoral 1. Isolate CD34+ as described in Subheading 3.1.

Injection in Adult Mice 2. Irradiate mice as described in Subheading 3.2.
3. Wait for at least 4 h and not more than 24 h before injecting
mice with HSC.
4. Anesthetize mice according to institutional protocols.
5. Shave the fur broadly over the knee joint area of the mice using
a hair clipper.
6. Locally spray 70 % ethanol and let it dry.
7. After flexing the knee to 90°, make an incision of 2–3 mm in
the skin just above the patella using the scalpel.
NSG mice
RAG mice
Fig. 2 T cell reconstitution after intrahepatic injection of cord blood CD34+ cells in newborn NSG and Rag −/−
mice at 12 weeks after injection
8. Move the patella to the side.

9. Place a 25-G needle between the condyles at the end of the
femur and twist gently until resistance is lost to make an intra-
femoral tunnel in the bone marrow cavity. Remove the needle.
10. Fill the Hamilton syringe with the CD34+ cell suspension and
insert into the hole made in the bone marrow cavity. Inject the
cell suspension into the bone marrow cavity.
11. Seal the needle hole with bone wax and reposition the patella.
12. Close the skin incision with wound staples or absorbable suture
(see Note 19).
13. Depending on the research question, different analyzing time
points can be set. T cell development can be analyzed from
3 to 4 weeks after injection on. Full evaluation is optimally
done at 10–12 weeks after HSC injection by sacrificing mice
and collecting organs (bone marrow, thymus, liver, spleen, and
blood by cardiac puncture) (see Note 20).
3.6 Fetal Thymus 1. Place the fetal thymus in a petri dish and rinse twice with PBS/
Preparation gentamicin.
for Subcapsular 2. Transfer the thymus to a petri dish filled with cold RPMI
Kidney Transplant medium.
3. Remove the connective tissue from the thymus.
4. Cut the thymus into small pieces (1 × 1 × 1 mm) using micro-
dissecting scissors (see Note 21). Keep on ice.
3.7 Fetal Liver 1. Place the fetal liver in a petri dish and rinse twice with PBS/
Preparation gentamicin.
for Subcapsular 2. Transfer the liver to a petri dish filled with cold RPMI medium.
Kidney Transplant
3. Cut the liver into small pieces (1 × 1 × 1 mm) using microdis-
secting scissors (see Note 21). Keep on ice.
4. Optional (BLT model): Digest small tissue pieces using RPMI-
1640 supplemented with collagenase and DNase I at 37 °C for
1 h. Gently disrupt the tissues by mixing every 15 min. Filter
the cell suspension through a sterile 70 μm cell strainer. Isolate
CD34+ cells as stated above. If for the experiment cells arising
from injected FL CD34+ and transplanted FL pieces need to be
discriminated, CD34+ cells can be labeled (see Note 22).
3.8 Generation 1. A transplantation needle can be generated if necessary

of a Transplantation (see Note 9).
Needle 2. Smooth the tip of an 18-G hypodermic needle with sandpaper.
3. Cut the internal piece of an 18-G spinal needle to a length
2 mm longer than the hypodermic needle. Also smooth the
surface of the internal needle with sandpaper.
4. Place the internal piece of the spinal needle inside the hypoder-
mic needle.
3.9 Subcapsular 1. If necessary (see Note 9): generate a transplantation needle as

Kidney Transplant described in Subheading 3.8.
2. Prepare fetal thymus and fetal liver as described in
Subheadings 3.6 and 3.7, respectively.
3. Optional (BLT model): irradiate mice as described in
Subheading 3.2. Wait for at least 4 h and not more than 24 h
before starting transplantation procedure.
4. Anesthetize mice according to institutional protocols.
5. Shave the left side of the mouse from hip joint to shoulder
joint using a hair clipper.
6. Load transplantation needle with fetal tissues: using Dumont
tweezers, place a piece of fetal liver on the tip of the hypodermic
needle, and aspirate it inside by drawing back the internal nee-
dle. Similarly, load a piece of thymus and a second piece of liver
on the same needle, making a liver–thymus–liver sandwich.
7. Swab the left side of the mouse using alcohol swabs.

8. Make a longitudinal skin incision from approximately the last
rib to the hip joint (2 cm).
9. Loosen connective tissue under the skin using the blunt side
of scissors, until you see the kidney bulge beneath the
musculature.
10. Pull the musculature above the kidney upwards with tweezers
and make a small incision perpendicular to the axis of the kid-
ney in the abdominal wall.
11. Using thumb and index finger of both hands, squeeze the
kidney out of the abdominal cavity. Direct the pressure to keep
spleen and gut in place while the kidney pops out. When the
incision in the abdominal wall is small, the kidney will now
remain in place outside the abdominal wall (as a button in a
button hole).
12. Keep the kidney wet by regularly dropping a small amount of
sterile PBS onto it using a 10 ml syringe with an 18-G needle.
13. Using Dumont tweezers, gently tear the kidney capsule near
the lower pole, making a hole. Be careful not to damage the
parenchyma of the kidney as to prevent bleeding.
14. Insert the transplantation needle (loaded with human fetal
liver and thymus) through the hole and move the needle
upwards towards the upper pole of the kidney, between the
renal capsule and the kidney parenchyma. Make sure the kidney
remains wet (by dropping sterile PBS onto it).
15. Push the internal needle to drive out the human fetal liver and
thymus pieces. Make sure that the three pieces stay together
while retracting the needle.
16. Gently pull two edges of the muscular wall around the kidney
upwards with toothed tweezers, allowing the kidney to reenter
the abdomen.
17. Close the abdominal wall using absorbable sutures.
18. Close the skin incision with wound staples or absorbable suture
(see Note 19).
19. After surgery, make sure the mice maintain their body
temperature by placing the cage on heat bags or by using a
heat lamp.
20. Optional (BLT model): tissue implanted mice can be intrave-
nously injected with CD34+ cell suspensions from the same
fetal liver on the day of tissue transplantation (a few hours
after surgery). CD34+ isolation is performed as described in
Subheading 3.1.
21. Engraftment can be analyzed 10–12 weeks after transplanta-
tion (see Note 20).
4 Notes
1. Typically, a 137Cs gamma irradiator is used but in theory any

radiation device can be used (as described in Subheading 1.3).
2. Housing devices during irradiation may be Hepa filter top
cages or special irradiation devices.
3. The use of human HSC requires approval by the ethical
committee of your institution. Informed consent should be
obtained. When using human organs, be sure to follow the
appropriate (inter)national guidelines.
4. CD34+ cells can be isolated using anti-CD34 magnetic activated
cell sorting (MACS) beads (Miltenyi, Leiden, The Netherlands)
or by fluorescence activating cell sorting (FACS).
5. Immunodeficient mice should be housed in specific pathogen
free (SPF) conditions, in microisolator cages. When perform-
ing animal studies, be sure to follow the appropriate (inter)
national guidelines.
6. Surgical instruments need to be sterilized by autoclaving before
use or disposable instruments should be used.
7. HLA typing of fetal liver and fetal thymus can be done to
enable matching or discrimination, depending on the experi-
mental setup.
8. Typically, fetal thymus is of 14–23 weeks of gestational age and
fetal liver 18–23 weeks [57].
9. A transplantation needle for injecting fetal thymus and liver
can be constructed using 18-G hypodermic and spinal needles
[58] or an 18-G trocar can be used [57].
10. Total volumes typically used for injection are: 50 μl for
intrahepatic injection in newborn mice, 200 μl for intravenous
injection in adult mice and 10 μl for intrafemoral injection in
adult mice. Cell numbers typically used are: 5–50 × 104 for
intrahepatic injection, 1 × 105–5 × 106 cells for intravenous
injection (when using more than 107 cells, the high viscosity
will result in difficulties and mice may die soon after injection)
and 5–10 × 104 cells for intrafemoral injection. Depending on
the experimental setup, cell numbers may vary.
11. Cells need to be injected as soon as possible when resuspended
in PBS, to avoid cell death.
12. All investigators should wear sterile gloves and rub clean
bedding between their hands before handling newborn mice,
to mask any foreign odors on pups and improve the chance
that the mother will accept her pups back into the cage after
completion of the procedure.
13. Pups should have milk in their stomach (seen as a white spot
under the relatively transparent skin) before irradiation proce-
dures can be started.
14. Depending on the immunodeficient mice strain used, a differ-
ent amount of irradiation should be given (see Subheading 1.3).
Newborn NSG or NOG mice are typically irradiated with
100 cGy, newborn BRG mice with 400 cGy.
15. The liver can be seen as a dark spot under the relatively trans-
parent skin of the newborn pup and is situated above the white
stomach (filled with milk).
16. Mice are best restrained by one investigator with one hand at
their scruff and another holding hind legs and tale, and injected
by another investigator. When performing the injection with-
out a second investigator, mice can be held with two fingers
(thumb and forefinger) at the scruff and little finger at the tail
and injecting can be performed using the other hand.
17. Make sure not to overheat mice, the temperature should not
exceed 30–32 °C (85–90 °F).
18. The topical nasal decongestant applied to the snout of the mother
masks the foreign odors left by the investigators and improves
the chance that the mother will accept her pups back into the
cage after completion of the procedure (see also Note 12).
19. Staples are preferred as they are less easily removed by the
mice.
20. Mice should be monitored regularly for wound healing and
possible signs of distress. Wound staples should be removed
after 7–10 days.
21. The diameter of the fetal thymus at 14–23 weeks of gestational
age is typically 1–1.5 cm. The length of the fetal liver at 14–23
weeks of gestational age is typically 15–35 mm.
22. CD34+ cells can be labeled with an intracellular dye (e.g.,
CFSE, PKH, or cell trace) or by transduction with a fluores-
cent marker (e.g., EGFP), depending on the experiment.
References
1. Lapidot T, Pflumio F, Doedens M et al (1992) 3. Vormoor J, Lapidot T, Pflumio F et al (1994)
Cytokine stimulation of multilineage hemato- SCID mice as an in vivo model of human cord
poiesis from immature human cells engrafted blood hematopoiesis. Blood Cells 20:316–320
in SCID mice. Science 255:1137–1141 4. Kamel-Reid S, Dick JE (1988) Engraftment of
2. Vormoor J, Lapidot T, Pflumio F et al (1994) immune-deficient mice with human hemato-
Immature human cord blood progenitors poietic stem cells. Science 242:1706–1709
engraft and proliferate to high levels in severe 5. Dao MA, Nolta JA (1998) Use of the bnx/hu
combined immunodeficient mice. Blood xenograft model of human hematopoiesis to
83:2489–2497 optimize methods for retroviral-mediated
stem cell transduction. Int J Mol Med 93: engrafted with mobilized human hematopoietic
2414–2419 stem cells. J Immunol 174:6477–6489
6. Nolta JA, Dao MA, Wells S et al (1996) 18. Traggiai E, Chicha L, Mazzucchelli L et al
Transduction of pluripotent human hemato- (2004) Development of a human adaptive
poietic stem cells demonstrated by clonal anal- immune system in cord blood cell-transplanted
ysis after engraftment in immune-deficient mice. Science 304:104–107
mice. PNAS 93:2414–2419 19. Brehm MA, Cuthbert A, Yang C et al (2010)
7. Mombaerts P, Iacomini J, Johnson RS et al Parameters for establishing humanized mouse
(1992) RAG-1-deficient mice have no mature models to study human immunity: analysis of
B and T lymphocytes. Cell 68:869–877 human hematopoietic stem cell engraftment in
8. Shinkai Y, Rathbun G, Lam KP et al (1992) three immunodeficient strains of mice bearing
RAG-2-deficient mice lack mature lympho- the IL2rgamma(null) mutation. Clin Immunol
cytes owing to inability to initiate V(D)J rear- 135:84–98
rangement. Cell 68:855–867 20. Cashman JD, Eaves CJ (1999) Human growth
9. Fulop GM, Philips RA (1990) The scid muta- factor-enhanced regeneration of transplantable
tion in mice causes a general defect in DNA human hematopoietic stem cells in nonobese
repair. Nature 347:479–482 diabetic/severe combined immunodeficient
10. Conneally E, Cashman J, Petzer A et al (1997) mice. Blood 93:481–487
Expansion in vitro of transplantable human 21. Shultz LD, Ishikawa F, Greiner DL (2007)
cord blood stem cells demonstrated using Humanized mice in translational biomedical
a quantitative assay of their lympho-myeloid research. Nat Rev Immunol 7:118–130
repopulating activity in nonobese diabetic- 22. Christianson SW, Greiner DL, Hesselton RA
scid/scid mice. PNAS 94:9836–9841 et al (1997) Enhanced human CD4+ cell
11. Larochelle A, Vormoor J, Hanenberg H et al engraftment in beta2-microglobulin-deficient
(1996) Identification of primitive human NOD-scid mice. J Immunol 158:3578–3586
hematopoietic cells capable of repopulating 23. Strowig T, Gurer C, Ploss A et al (2009)
NOD/SCID Mouse bone marrow: implica- Priming of protective T cell responses against
tions for gene therapy. Nat Med 2:1329–1337 virus-induced tumors in mice with human
12. Hogan CJ, Shpall EJ, McNulty O et al (1997) immune system components. J Exp Med 206:
Engraftment and development of human 1423–1434
CD34(+)-enriched cells from umbilical cord 24. Suzuki M, Takahashi T, Katano I et al (2012)
blood in NOD/LtSz-scid/scid mice. Blood Induction of human humoral immune res-
90:85–96 ponses in a novel HLA-DR-expressing trans-
13. Kerre TC, De Smet G, De Smedt M et al genic NOD/Shi-scid/gammacnull mouse. Int
(2002) Adapted NOD/SCID model supports Immunol 24:243–252
development of phenotypically and function- 25. Billerbeck E, Horwitz JA, Labitt RN et al
ally mature T cells from human umbilical cord (2013) Characterization of human antiviral
blood CD34(+) cells. Blood 99:1620–1626 adaptive immune responses during hepato-
14. McKenzie JL, Gan OI, Doedens M et al (2005) tropic virus infection in HLA-transgenic
Human short-term repopulating stem cells are human immune system mice. J Immunol
efficiently detected following intrafemoral 191:1753–1764
transplantation into NOD/SCID recipients 26. Shultz LD, Brehm MA, Garcia-Martinez JV
depleted of CD122+ cells. Blood et al (2012) Humanized mice for immune
106:1259–1261 system investigation: progress, promise and
15. Shultz LD, Banuelos SJ, Leif J et al (2003) challenges. Nat Rev Immunol 12:786–798
Regulation of human short-term repopulating 27. Brehm MA, Racki WJ, Leif J et al (2012)
cell (STRC) engraftment in NOD/SCID mice by Engraftment of human HSCs in nonirradiated
host CD122+ cells. Exp Hematol 31:551–558 newborn NOD-scid IL2rgamma null mice is
16. Ito M, Hiramatsu H, Kobayashi K et al (2002) enhanced by transgenic expression of membrane-
NOD/SCID/gamma(c)(null) mouse: an excel- bound human SCF. Blood 119:2778–2788
lent recipient mouse model for engraftment of 28. O’Connell RM, Balazs AB, Rao DS et al
human cells. Blood 100:3175–3182 (2010) Lentiviral vector delivery of human
17. Shultz LD, Lyons BL, Burzenski LM et al (2005) interleukin-7 (hIL-7) to human immune sys-
Human lymphoid and myeloid cell development tem (HIS) mice expands T lymphocyte popula-
in NOD/LtSz-scid IL2R gamma null mice tions. PLoS One 5, e12009
29. Song J, Willinger T, Rongvaux A et al (2010) tive and innate immune responses to EBV and
A mouse model for the human pathogen TSST-1. Nat Med 12:1316–1322
Salmonella typhi. Cell Host Microbe 8: 42. Denton PW, Garcia JV (2009) Novel human-
369–376 ized murine models for HIV research. Curr
30. Tary-Lehmann M, Saxon A, Lehmann PV HIV/AIDS Rep 6:13–19
(1995) The human immune system in hu-PBL- 43. McDermott SP, Eppert K, Lechman ER et al
SCID mice. Immunol Today 16:529–533 (2010) Comparison of human cord blood
31. Shultz LD, Saito Y, Najima Y et al (2010) engraftment between immunocompromised
Generation of functional human T-cell subsets mouse strains. Blood 116:193–200
with HLA-restricted immune responses in 44. Notta F, Doulatov S, Dick JE (2010)
HLA class I expressing NOD/SCID/IL2r Engraftment of human hematopoietic stem
gamma(null) humanized mice. PNAS 107: cells is more efficient in female NOD/
13022–13027 SCID/IL-2Rgc-null recipients. Blood 115:
32. Jaiswal S, Pazoles P, Woda M et al (2012) 3704–3707
Enhanced humoral and HLA-A2-restricted 45. Pearson T, Greiner DL, Shultz LD (2008)
dengue virus-specific T-cell responses in Creation of “humanized” mice to study human
humanized BLT NSG mice. Immunology immunity. Curr Protoc Immunol Chapter 15,
136:334–343 unit 15.21
33. Danner R, Chaudhari SN, Rosenberger J et al 46. King MA, Covassin L, Brehm MA et al (2009)
(2011) Expression of HLA class II molecules Human peripheral blood leucocyte non-obese
in humanized NOD.Rag1KO.IL2RgcKO mice diabetic-severe combined immunodeficiency
is critical for development and function of interleukin-2 receptor gamma chain gene
human T and B cells. PLoS One 6, e19826 mouse model of xenogeneic graft-versus-host-
34. Holyoake TL, Nicolini FE, Eaves CJ (1999) like disease and the role of host major histo-
Functional differences between transplantable compatibility complex. Clin Exp Immunol
human hematopoietic stem cells from fetal 157:104–118
liver, cord blood, and adult bone marrow. Exp 47. Bruck F, Belle L, Lechanteur C et al (2013)
Hematol 27:1418–1427 Impact of bone marrow-derived mesenchymal
35. Ishikawa F, Yasukawa M, Lyons B et al (2005) stromal cells on experimental xenogeneic graft-
Development of functional human cord blood versus-host disease. Cytotherapy 15:267–279
and immune systems in NOD/SCID/IL2 48. Ali N, Flutter B, Sanchez RR et al (2012)
receptor {gamma} chain(null) mice. Blood Xenogeneic graft-versus-host-disease in NOD-
106:1565–1573 scid IL-2Rgammanull mice display a T-effector
36. McCune JM, Namikawa R, Kaneshima H et al memory phenotype. PLoS One 7, e44219
(1988) The SCID-hu mouse: murine model 49. Cai S, Wang H, Bailey B et al (2011)
for the analysis of human hematolymphoid Differential secondary reconstitution of
differentiation and function. Science 241: in vivo-selected human SCID-repopulating
1632–1639 cells in NOD/SCID versus NOD/SCID/
37. Kyoizumi S, Baum CM, Kaneshima H et al gamma chain mice. Bone Marrow Res
(1992) Implantation and maintenance of func- 2011:252953
tional human bone marrow in SCID-hu mice. 50. Kwant-Mitchell A, Pek EA, Rosenthal KL et al
Blood 79:1704–1711 (2009) Development of functional human NK
38. Baum CM, Weissman IL, Tsukamoto AS et al cells in an immunodeficient mouse model with
(1992) Isolation of a candidate human hema- the ability to provide protection against tumor
topoietic stem-cell population. PNAS 89: challenge. PLoS One 4, e8379
2804–2808 51. Waskow C, Madan V, Bartels S et al (2009)
39. Chen BP, Galy A, Kyoizumi S et al (1994) Hematopoietic stem cell transplantation with-
Engraftment of human hematopoietic precur- out irradiation. Nat Methods 6:267–269
sor cells with secondary transfer potential in 52. Waskow C, Cosgun N, Rahmig S (2013) Stable
SCID-hu mice. Blood 84:2497–2505 human hematopoietic stem cell engraftment
40. Shih CC, Hu MC, Hu J et al (1999) Long- supports continuous de novo generation of
term ex vivo maintenance and expansion of mature human blood cells in mice. Exp
transplantable human hematopoietic stem cells. Hematol 41:S11
Blood 94:1623–1636 53. Choi B, Chun E, Kim M et al (2011) Human
41. Melkus MW, Estes JD, Padgett-Thomas A et al B cell development and antibody production in
(2006) Humanized mice mount specific adap- humanized NOD/SCID/IL-2Rgamma(null)
(NSG) mice conditioned by busulfan. J Clin required to initiate engraftment of purified

Immunol 31:253–264 primitive human hematopoietic cells trans-
54. Biancotti JC, Town T (2013) Increasing hema- planted at limiting doses into NOD/SCID
topoietic stem cell yield to develop mice with mice. Bone Marrow Transplant 23:
human immune systems. Biomed Res Int 203–209
2013:740892 57. Namikawa R, Weilbaecher KN, Kaneshima H
55. Sun A, Wei H, Sun R et al (2006) Human et al (1990) Long-term human hematopoiesis
interleukin-15 improves engraftment of human in the SCID-hu mouse. J Exp Med 172:
T cells in NOD-SCID mice. Clin Vaccine 1055–1063
Immunol 13:227–234 58. Roncarolo MG, Carballido JM (2001) Cons-
56. Bonnet D, Bhatia M, Wang JC et al (1999) truction of human-SCID chimeric mice. Curr
Cytokine treatment or accessory cells are Protoc Immunol Chapter 4, Unit 4.8
Chapter 22
Using the Zebrafish Model to Study T Cell Development

Yong Zhang and David L. Wiest
Abstract
While zebrafish have for some time been regarded as a powerful model organism with which to study early
events in hematopoiesis, recent evidence suggests that it also ideal for unraveling the molecular requi-
rements for T cell development in the thymus. Like mammals, zebrafish possess an adaptive immune
system, comprising B lymphocytes as well as both the γδ and αβ lineages of T cells, which develop in the
thymus. Moreover, the molecular processes underlying T cell development in zebrafish appear to be
remarkably conserved. Thus, findings in the zebrafish model will be of high relevance to the equivalent
processes in mammals. Finally, molecular processes can be interrogated in zebrafish far more rapidly than
is possible in mammals because the zebrafish possesses many unique advantages. These unique attributes,
and the methods by which they can be exploited to investigate the role of novel genes in T cell develop-
ment, are described here.
Key words Zebrafish, T cell development, Morpholino, Functional rescue, In situ hybridization
1 Introduction
Zebrafish are becoming a powerful vertebrate model for in vivo

genetic studies of human development [1–3]. Zebrafish and mam-
mals share similar blood cells and use common molecular pathways
to regulate the production of blood cells including thymocytes
[4–6]. In zebrafish, the first definitive HSCs derive from the ventral
wall of the dorsal aorta at about 28–30 h post fertilization (hpf),
then migrate to the caudal hematopoietic tissue, and finally home
to the thymus and the kidney, where T cell development and
adult hematopoiesis occur, respectively [7]. T cell progenitors
marked by ikaros expression appear in the thymus by 3 days post
fertilization (dpf) [7]. As in mammals, zebrafish T cell development
gives rise to two T lineages, γδ and αβ, and is accompanied by the
rearrangement of four T-cell receptor (TCR) loci (TCRα, TCRβ,
TCRγ, and TCRδ) [8]. As in mammals, TCR genes in zebrafish
are assembled by a V(D)J recombination process that is depen-
dent upon the recombination activating genes, rag1 and rag2 [9].
273
274 Yong Zhang and David L. Wiest
Table 1
Transgenic lines for studying thymocyte development in zebrafish
Reporter line Gene promoter Cells marked References

Tg(itga2b:EGFP)la2 cd41 Lymphoid stem cells [13, 14]
Tg(ikzf1:GFP)fr101 ikaros Lymphoid stem/progenitor cells [15]
zdf8
Tg(rag2:EGFP) rag2 Lymphoid progenitors [16]
cz1
Tg(lck:EGFP) lck T progenitors/T cells [17]
The T cell specific nonreceptor tyrosine kinase lck is also found in

the lymphoid precursors and maturing T lymphocytes in the bilat-
eral thymic lobes of zebrafish [3, 10]. A recent review summarizes
additional genes that play an important role in zebrafish and serve
as markers of T cell development (c-myb, ikaros, rag1, rag2, lck,
TCRα, TCRβ, TCRγ, TCRδ, IL7R, jak3, ccr9a, ccr9b, zap70, gata3,
runx1, foxn1, etc.) [11].
Although the mouse remains the gold standard for immuno-
logical research, the zebrafish model provides a number of distinct
advantages that complement the use of mice. Specifically, zebrafish
females lay a large number of externally fertilized eggs that develop
ex utero and this allows easy kinetic analysis of the effect of muta-
tions that are lethal at early embryonic stages and would thus be
very difficult to analyze during mouse development. Most zebraf-
ish genes have mammalian orthologs, and the roles they play in
hematopoiesis in general, and T cell development in particular, are
highly conserved. Accordingly, insights gained into processes
underlying T cell development in zebrafish will be applicable to
mammalian biology [12]. Another significant advantage of using
the zebrafish model is that embryos are transparent for the first 3
weeks of life. The optical clarity of zebrafish embryos allows visual-
ization of development using fluorescent lineage tracers. These
tracers enable real time imaging to study migration, colonization,
and cell-fate determination of T progenitors without the need for
sophisticated intravital two-photon microscopy [13, 14]. Trans-
genic fluorescent lineage tracers available for studying T cell devel-
opment are shown in Table 1. For gene expression analysis, both
the expression level and cell type distribution of transcripts can be
easily assessed using whole mount in situ hybridization with anti-
sense probes. Finally, gene function can be rapidly assessed using
gain- and loss-of-function approaches. Indeed, the requirement of
a gene can be readily examined by loss-of-function using antisense
oligonucleotides called “morpholinos,” which interfere with either
splicing or translation of mRNA [18]. Morpholinos can be obtained
by providing Gene Tools with the target sequence, following
Studying T Cell Development in Zebrafish 275
which they will design sequence-specific oligos to block either

mRNA translation or splicing (https://oligodesign.gene-tools.
com/request/). Conversely, gain-of-function, overexpression experi-
ments can be done either by simply injecting the egg with mRNA
encoding the protein of interest or by using heat-inducible plas-
mids that allow manipulation of both the timing or level of induc-
tion. These advantages, which make zebrafish such a powerful
model for the dissection of T cell development, are described here.
Expected experimental outcomes are illustrated using our recent
analysis of the role of the ribosomal protein Rpl22 in T cell
development.
2 Materials
2.1 Zebrafish 1. AHAB zebrafish housing systems (Aquatic Habitats, Aquatic

Facility, Husbandry, Eco-Systems, Inc.): The fish room is maintained on a 14-h
and Culture day/10-h night cycle. We maintain and breed Zebrafish at
of Embryos 28.5 °C under the standard aquaculture conditions published
in The Zebrafish Book (http://zfin.org/zf_info/zfbook/
zfbk.html). Embryos were staged as described previously [19].
2. Egg water stock (60×): pH 7.0, 300 mM NaCl, 10.2 mM KCl,
19.8 mM CaCl2, 19.8 mM MgSO4. Prepare 60× Stock by add-
ing 17.5 g of NaCl, 0.76 g of KCl, 2.8 g of CaCl2, and 4.9 g
of MgSO4 to 1 l Millipore Q water. Adjust pH to 7.0, and
sterilize by autoclaving. Make a 1× working solution by dilut-
ing 60× stock into distilled water. Add methylene blue to a
final concentration of 0.01 %. The solution can be stored for
several weeks at room temperature (RT) (see Note 1).
3. Phenylthiourea (PTU): Prepare the 60× stock solution by dis-
solving 1.8 g powder in 1 L Millipore Q water and heating to
60 °C for several hours under agitation. Store at 4 °C. Make up
the working solution by dilution to 1× with egg water. The use
of PTU can prevent the embryos from developing pigment,
thereby prolonging the developmental window during which
they embryos are transparent.
4. Pronase: Dissolve 1 g powder in 100 ml egg water to make
10 mg/ml stock solution. Aliquot into 15 ml tubes and store
at −20 °C.
5. Petri dishes (100 × 20 mm).
6. Transfer pipettes
7. Incubator for zebrafish embryos
2.2 Zebrafish Lines Wild-type and transgenic zebrafish lines are available from Zebrafish
International Resource Center (ZIRC, Eugene, OR). For examples
described in this chapter, we employed the following fish lines: (1)
AB wild-type fish; and (2) Tg(lck:EGFP)cz1 [17].
2.3 Morpholino 1. Morpholinos (MO) (Gene Tools, Philomath, OR): We use

Design morpholinos to knock down gene expression. The 25-base
and Preparation morpholinos are complementary with their target RNAs and
can either inhibit mRNA translation or mRNA splicing.
Translation-blocking morpholinos often decrease protein exp-
ression to levels undetectable by Western blot and begin to act
immediately in fertilized embryos. Splice-blocking morpholi-
nos act after the switch from maternal to zygotic transcription
at 3.5dpf and allow selective deletion of particular protein
domains.
2. Danieau buffer: pH 7.6, 5.0 mM HEPES, 58 mM NaCl,
0.7 mM KCl, 0.4 mM MgSO4, 0.6 mM Ca(NO3)2, dissolved
in nuclease-free water.
3. The morpholino stock is dissolved in either Danieau buffer or
nuclease-free water at a concentration of 1–3 mM in 5 μl ali-
quots. Stocks should be stored frozen at −80 °C or in airtight
microtubes at RT to prevent evaporation (see Note 2). These
are sample sequences for MOs targeting the zebrafish rpl22
start codon (5′-CCGACAGTTTTGGCAGAAAGCCAGT-3′,
designated “Rpl22 MO”) and as well as a 5-base mismatch
control (5′-CCCACACTTTTCGCACAAACCCAGT-3′, des-
ignated “Rpl22 MM”).
2.4 Microinjection 1. Needle puller (Flaming/brown micropipette puller, Model

P-97, Sutter Instrument Co.).
2. Borosilicate filamented glass capillaries (World Precision instru-
ments, Inc.).
3. Microinjector (PM1000 cell microinjector, Microdata Instru-
ment, Inc.). This microinjector requires a separate compressed
air pump (Senco, Model: PC1010).
4. Micromanipulator holder (GJ-8 magnetic stand, Narishige
Scientific Instruments).
5. Micromanipulator arm: M-152 (Narishige Scientific Instruments).
6. Dissecting stereomicroscope.
7. Plastic Mold for making injection plates: See Fig. 1a for a
description of the plastic mold to make injection chambers.
This can also be found in the Zebrafish Book (http://zfin.
org/zf_info/zfbook/chapt5/5.1.html).
8. Stage micrometers.
9. Straight probes for micromanipulating zebrafish embryos
(Fine Science Tools, Cat#: 10140-01).
10. Petri dishes (100 × 20 mm).
Fig. 1 Microinjection and dechorionation for zebrafish embryos. (a) Injection mold. (b) Injection chamber/plate.
(c) Schematic of morpholino or plasmid injection. For morpholino knockdown, we inject the liquid (red spot) to
the interface between cell and yolk. For plasmid overexpression, we directly inject the solution to the cyto-
plasm (green spot) at 1-cell stage. (d) Dechorionation process. The arrows indicate the chorions around the
24hpf embryos. (e) Alignment of the one-cell stage embryos in the trenches of the injection plate. (f) Schematic
of the vertical section of the injection plate. (g) Lateral view of a 5dpf Tg(lck:EGFP)cz1 embryo. Arrow indicates
the EGFP+ T cells in the thymus
2.5 Transient Rescue 1. Plasmid DNA preparation kit for highly purified DNA.
Experiments 2. Water bath.
3. Construct: For heat shock mediated inducible expression, clone
cDNA encoding the gene of interest into the heat-inducible
pSGH2 vector [6, 20]. The pSGH2 vector contains two I-SceI
restriction sites.
4. I-SceI meganuclease: (New England Biolabs) Aliquot 2 μl each
upon arrival and store at −80 °C) along with I-SceI buffer
(New England Biolabs), supplied with the NEBuffer I-SceI
pack (see Note 3).
2.6 Western Blot [21] 1. Deyolking buffer: 55 mM NaCl, 1.8 mM KCl, 1.25 mM
NaHCO3.
2. Wash buffer: 10 mM Tris–HCl pH 8.5, 110 mM NaCl,
3.5 mM KCl, 2.7 mM CaCl2.
3. 200 μl Precision tips.
4. Protease inhibitor cocktail (Roche).
5. PMSF.
6. SDS lysis buffer: 100 mM Tris pH 6.8, 4 % SDS, 200 mM

DTT, and 20 % glycerol, supplemented with complete mini
protein inhibitor cocktail (Roche).
7. MOPS running buffer.
8. Transfer buffer: 20 mM Tris, 150 mM glycine, 20 % methanol,
and 0.038 % SDS.
9. 20× TBST solution: 500 mM Tris pH 7.4, 60 mM KCl, 2.8 M
NaCl, and 1.0 % Tween 20 in high purity H2O. Make 1× TBST
by diluting 20× TBST solution with distilled H2O.
10. Blocking buffer: 5 % nonfat milk powder in TBST.
11. Antibodies as needed. The following antibodies were used in
the experiments described in this chapter: Anti-human Rpl22
serum (detects N-terminus of human Rpl22, Transduction
Labs, San Jose, CA [22]), dilution: 1:1000; Anti-Actin (Sigma,
AC-40), dilution: 1:2000. Secondary anti-rabbit or mouse
IgG HRP-conjugated Antibodies (Cell Signaling), dilution:
1:1000.
12. 0.45 μm PVDF transfer membrane.
13. ECL substrate and chemiluminescent reagents.
14. Protein gel running and transfer equipment (e.g., Novex XCell
SureLock® Mini-Cell and XCell II™ Blot Module, Life tech-
nologies or other similar).
15. Autoradiography films.
2.7 Whole-Mount 1. In Situ Cell Death Detection Kit, TMR red (Roche).
TUNEL Assays 2. Acetone.
3. Permeabilization solution (0.1 % v/v Triton X-100 and 0.1 %
w/v sodium citrate in PBS): 100 μl Triton X-100 and 1 ml
10 % sodium citrate in 98 ml of pH 7.4 PBS, freshly
prepared.
2.8 Whole-Mount 1. Fixative: 4 % paraformaldehyde in PBS. Dissolve 40 g of pow-

In Situ Hybridization der in 1 L of PBS by heating to 60 °C, then adjust the pH to
7.4. Fixative can be aliquotted into 50-ml tubes and stored at
−20 °C (up to ~6 months).
2. Nuclease-free water.
3. PBS (10× Stock solution, pH 7.4).
4. Tween 20.
5. PBST: 1× PBS plus 0.1 % Tween 20.
6. Methanol.
7. Proteinase K stock: 10 mg/ml of proteinase K (Roche
Diagnostics) in PBST.
8. RNA polymerase (T7, Sp6, or T3) and transcription buffer

(Ambion).
9. DIG RNA labeling mix (Roche Diagnostics).
10. RNAse inhibitor (Ambion).
11. NucAway™ spin columns (Ambion,).
12. UltraPure formamide
13. 20× SSC stock solution
14. Prehybridization buffer (Hyb−): 50 % formamide, 5× SSC,
0.1 % Tween 20. Store at −20 °C.
15. Hybridization buffer (Hyb+): Hyb (−) plus 5 mg/ml RNA
from Bakers Yeast (Sigma) and 50 μg/mL heparin (Sigma,).
Store at −20 °C.
16. 2× SSCT: 2× SSC, 0.1 % Tween 20.
17. 2× SSCT/50 % formamide: 2× SSCT, 50 % formamide.
18. 0.2× SSCT: 20× SSC diluted to 0.2× with nuclease-free water,
0.1 % Tween 20.
19. Blocking reagent for nucleic acid hybridization.
20. Heat-inactivated FBS: Incubate FBS at 60 °C for 1 h, filter-
sterilize, then store at −20 °C.
21. MAB buffer: 10 mM Tris pH 9.5, 100 mM maleic acid,
150 mM NaCl.
22. MABT: MAB plus 0.1 % Tween 20.
23. 10 % BMB Block: 10 % blocking reagent in MAB solution.
24. Complete Block: MABT plus 10 % heat-inactivated FBS, 2 %
BMB Block.
25. Anti-digoxigenin–alkaline phosphatase antibody (Roche
Diagnostics).
26. Staining buffer: 100 mM Tris–HCl pH 9.5, 50 mM MgCl2,
100 mM NaCl, 0.1 % Tween 20, 1 mM levamisole (add fresh
from a 0.5 M stock).
27. Vector BCIP/NBT Staining solution: We use BCIP/NBT AP
substrate kit IV (Vector Laboratories). Add two drops of kit
solution per 5 mL of staining solution, mix; two drops solution
2, mix; and two drops solution 3, mix well.
28. 15 mm Netwell™ Insert with 74 μm Mesh Size Polyester
Membrane (Corning,).
29. Orbital shaker.
30. Tecan NanoQuant machine for RNA/DNA quantitation
(Tecan, Infinite M200).
31. Hybridization oven.
2.9 Mounting 1. Fluorescence stereomicroscope equipped with green fluorescent

Embryos protein (GFP) and DsRed filters, DS-Fi1 digital camera, and
for Observation imaging software.
2. Microscope depression slides (Science Enthusiast).
3. Tricaine: Prepare a 25× solution (4 mg/ml) by dissolving 1.2 g
in 300 ml of water. Store at 4 °C.
4. Methylcellulose: Prepare a 3 % solution in egg water. The living
or stained embryos are mounted in 3 % methylcellulose and
photographed.
2.10 Fluorescence- 1. 1× trypsin–EDTA.

Activated Cell Sorting 2. TRIzol reagent.
(FACS)
3. 35 mm culture dish.
4. 40 μm nylon mesh filter.
5. FACSAria II (BD) or comparable cell sorter.
3 Methods
3.1 Identifying Many mammalian genes have zebrafish orthologs. We use both
Zebrafish Ortholog homology and synteny-based approaches to identify the zebrafish
orthologs of mammalian genes. Conserved synteny (i.e., contigu-
ous genes or ESTs with conserved map order on the chromosomes
of different species) is a prominent feature of vertebrate genomes.
Analysis of zebrafish genomic mapping data has revealed conserved
rearrangements and homology segments between zebrafish and
human genomes [5]. Thus, the homology based syntenic analysis
can be used to effectively predict and find orthologous genes
between zebrafish and human.
1. Go to http://www.ensembl.org/Homo_sapiens/Info/index.
html and enter the name of the human gene (e.g., RPL22) in
the SEARCH section.
2. Click on ContigView, then the gene name, to get the amino
acid sequence as well as genomic position.
3. Go to http://www.ncbi.nlm.nih.gov/, select tBLASTn and
paste the amino acid sequence in the dialog box. Select
“Nucleotide selection (nr/nt)” in the “Choose search set-
Database” window, and “zebrafish” in the “Organism” win-
dow, and click on BLAST, followed by clicking on “Formatting”
in a new window to get the search summary.
4. Record the zebrafish gene name (e.g., zgc:123327, rpl22), then
go to http://www.genome.ucsc.edu/ to identify the genomic
position of this gene.
5. Repeat these steps for other genes located near the gene
of interest until it is clear that there is a good syntenic
relationship.
6. Next align the amino acid sequences of the human, mouse and
zebrafish orthologs of your gene of interest. The Clustal W2
algorithm is an effective tool for this task (http://www.ebi.
ac.uk/Tools/msa/clustalw2/).
3.2 Preparation 1. Prepare a 1.5 % agarose solution with 1× egg water.

of Microinjection 2. Pour enough molten agarose into a 100 cc petri dish to cover
Plates/Needles the bottom with a thin layer, then allow to solidify at room
temperature. When the agarose is solid, pour another layer of
molten agarose on top and place the plastic injection mold
(Fig. 1a) into the agarose solution. Make sure no bubbles are
trapped under the mold. Cover and allow it to cool.
3. Overlay the injection plate with ddH2O and store at 4 °C
(see Fig. 1b).
4. Pull microinjection needles using the P-97 micropipette puller.
Load the glass capillary tubes into the heating chamber with
the middle part of the capillary surrounded by a platinum heat-
ing film. The platinum heating film electrically heats up the
glass capillary, enabling a horizontal linear force to pull the
heated glass apart to produce two separate needles. Store nee-
dles into a petri dish on a row of plasticine to prevent the
needles from breaking.
5. Under a stereomicroscope, use a blade to cut the tip off of
the needle. A good microinjection needle should be long and
sharp but not flexible. This will minimize damage to the
embryo and help to pierce the chorions.
3.3 Preparation 1. The procedures to follow to set up matings are described in

of Zebrafish Embryos detail in the zebrafish book (http://zfin.org/zf_info/zfbook/
zfbk.html). Briefly, set up well-fed fish (6–18 months old) the
night before in pairs (two males and two females) in the mat-
ing chamber with the divider in place. The next morning,
remove the divider to allow the fish to spawn.
2. Collect embryos approximately 20 min after female and male
zebrafish begin to mate. Visually inspect the embryos to ensure
that the clutch is uniformly at the early single cell stage. For
gene knockdown analysis, as well as for most other manipulations
of zebrafish embryos, it is best to inject into the cytoplasm of
one-cell stage embryos, which produces the most consistent
results.
3.4 Dechorionation The chorion is the membrane around the developing zebrafish
embryo (see Fig. 1d, arrows). Embryos should be removed from
their protective chorions to facilitate further observation, fixation,
or other manipulations. Removal of the chorion, or dechorion-
ation, is accomplished by pronase digestion of embryos older than
18hpf. We generally dechorionate at 24hpf.
1. Dilute 10 mg/ml pronase stock with egg water to 2 mg/ml
working solution.
2. Transfer the embryos to a petri dish, removing as much egg
water as possible. Add 2 ml of the diluted pronase solution to
the dish and swirl the embryos. Incubate at room temperature
for 10 min.
3. Use a Pasteur pipette to gently transfer the embryos to a fresh
petri dish full of egg water. Most of chorions will have been
removed at this time (see Fig. 1d).
4. Rinse the embryos thoroughly (at least three times) with egg
water to remove the pronase.
3.5 FACS Sorting 1. Transgenic zebrafish embryos produced as above from appro-
of Lineage Marked priate lines [Table 1; e.g., Tg(lck:EGFP)cz1] are dechorionated
T Cell Progenitors by pronase treatment. For FACS sorting of fluorescently
labeled progenitors, we use about 100 transgenic embryos
(5dpf, see Fig. 1e) that have been grown in the 28.5 °C
incubator.
2. To anesthetize the embryos, transfer them to a petri dish con-
taining Tricaine in egg water at 0.2 mg/ml and incubate until
they are no longer swimming.
3. Transfer anesthetized embryos to deyolking buffer (keep on
ice) and pass them through a 200 μl pipette tip several times to
remove their yolk.
4. Dissect and macerate deyolked whole embryos in ice-cold
0.9 × PBS plus 5 % FBS using a scalpel blade to facilitate
digestion.
5. Transfer embryos to a 35 mm culture dish containing 1× trypsin–
EDTA solution and incubate for 30–60 min at 32 °C. During
incubation, gently agitate by pipetting every 10 min.
6. Terminate the digestion by adding CaCl2 to a final concentra-
tion of 1 mM and fetal calf serum to 10 %.
7. Pellet the cells for 5 min at 400 × g, then wash twice with
0.9 × PBS at 400 × g.
8. Filter the suspension through 40 μm nylon mesh to eliminate
debris (Keep cold).
9. Pellet and resuspend the cells in 0.9 × PBS/5 % FBS. Stain with
1 μg/ml propidium iodide (PI) 30 min on ice to identify dead
cells and debris.
10. Perform FACS on the resulting single cell suspensions at room

temperature. Use non-transgenic AB strain zebrafish cells to
set the gate control, then sort the progenitors of interest (e.g.,
lck-GFP+ T cell progenitors) and GFP− cells directly in TRIzol,
if you are preparing RNA.
11. Isolate total RNA from the purified cells by TRIzol extraction
according to manufacturer’s instructions. Quantitative or con-
ventional PCR can then be used to assess mRNA expression of
the gene of interest.
3.6 Morpholino 1. Before injection, heat the morpholino (MO; antisense oligo-
Microinjection nucleotide) in a thermocycler for 10 min at 65 °C to dissolve
any precipitates that can clog the microinjection needle.
2. Assemble the needle to the microinjection arm.
3. Turn on air the compressor pump and PM1000 microinjector.
4. Press the “Timer” and “BALN” buttons to activate balance
pressure. Press the “VENT” button.
5. Use the “Fill” button to load the needle with the MO target-
ing your gene of interest.
6. Use the “Inject” knob on the injector panel to adjust the injec-
tion air pressure and set the injection time. For the PM1000
cell microinjector, we use 60 ms injection time and 10-psi
injection pressure as the starting point. Use the “Balance”
knob to adjust the balance pressure between 0 and 0.4 psi to
prevent both liquid from leaking out of the needle and medium
from flowing back into the needle.
7. Place a drop of mineral oil on the microscope stage microme-
ter. Submerge the tip of the injection needle into the mineral
oil and press the foot pedal to finish the injection. The released
droplet will form a perfect sphere in the oil. Measure the diam-
eter of the droplet and calculate the volume according
to 4/3πR3. Adjust the injection pressure and injection time to
produce the desired injection volume and dose of morpholino
(see Table 2).
8. Transfer embryos into the trenches of the injection plate using
a Pasteur pipette (Fig. 1e).
9. Use the fine probe to align the embryos, with the animal pole
facing you (see Fig. 1e, f).
10. Manipulate the injection arm to advance the needle, gently
piercing the chorion of the embryos. Inject the morpholino
into the interface between cytoplasm and yolk of one-cell stage
embryos (see Fig. 1c, lower panel and Note 4).
11. Move the injection plate and repeat until all of the embryos on
the plate have been injected, then transfer the embryos to
a clean dish containing egg water, and place in 28.5 °C
incubator.
Table 2
Morpholino (MO) dose per microinjection
Diameter of Radius of Volume MO dose (ng) of MO dose (ng) of

drop (μm) drop (μm) (pL) 4/3πR3 0.5 mM = 4.15 ng/nl 2 mM = 16.6 ng/nl
1 0.5 0.52 0.00 0.01
2 1 4.19 0.02 0.07
3 1.5 14.13 0.06 0.23
4 2 33.49 0.14 0.56
5 2.5 65.42 0.27 1.09
6 3 113.04 0.47 1.88
7 3.5 179.50 0.74 2.98
8 4 267.95 1.11 4.45
9 4.5 381.51 1.58 6.33
10 5 523.33 2.17 8.69
11 5.5 696.56 2.89 11.56
12 6 904.32 3.75 15.01
13 6.5 1149.76 4.77 19.09
14 7 1436.03 5.96 23.84
15 7.5 1766.25 7.33 29.32
3.7 Western Blot 1. To examine protein expression, dechorionate embryos using

Analysis pronase as above, then wash three times in egg water to elimi-
nate residual pronase.
2. Transfer both control and morpholino injected embryos to
1.5 ml microcentrifuge tubes. We found that approximately
five embryo equivalents were required to detect Rpl22, but
this will vary with protein abundance and antibody quality.
3. Centrifuge at 300 × g for 5 min and remove all residual egg
water. Add 200 μl deyolking buffer and pipet up and down
with a 200 μl tip until the yolk is disrupted. 200 μl deyolking
buffer is enough for up to 100 embryos and the volume should
be adjusted in proportion as the number of embryos is varied.
4. Agitate the tubes on the orbital shaker for 5 min at 12.85 × g
to dissolve the yolk.
5. Centrifuge at 300 × g for 30 s and discard the supernatant.
Wash two more times, then add 0.5 ml wash buffer to the pel-
lets, agitate for 2 min at 12.85 × g, and pellet as above.
6. Solubilize the pelleted embryos in 2 μl 2× SDS-sample buffer
per embryo and heat for 5 min at 95 °C. No homogenization
is necessary, as the cells should dissolve rapidly in the buffer.
7. Remove insoluble material by centrifugation at 4 °C for 10 min

at 12,000 × g. After collection, the supernatant can be imme-
diately run on an SDS-PAGE gel or stored at −80 °C in
aliquots.
8. Approximately five embryo equivalents are loaded per lane,
resolved by SDS-PAGE, transferred to PVDF membrane, and
immunoblotted. The sample immunoblot employed mouse
anti-actin monoclonal antibody and rabbit anti-Rpl22 serum,
followed by incubation with appropriate secondary HRP-
conjugated-Abs. Bound antibody was visualized by enhanced
chemiluminescence (Fig. 2b) (see Note 5).
3.8 Fluorescent 1. T cell development in Tg(lck:EGFP)cz1 embryos is clearly

Microscopy evident by 5dpf. To examine the effect of gene knockdown on
T cell development in this model, the embryos must be immo-
bilized. To accomplish this, place 5dpf morphant embryos
(i.e., morpholino-injected) in a petri dish containing Tricaine
in egg water at 0.2 mg/ml and incubate until the embryos are
no longer swimming. This usually requires about 5 min.
2. 3 % methylcellulose works well for mounting 5dpf embryos for
regular microscopic observation and photography. To mount
embryos, add a drop of 3 % methylcellulose to the microscope
depression slide.
3. Transfer the embryo using a Pasteur pipette and position it in
the methylcellulose.
4. Use the steel probe to precisely orient the embryo for lateral or
dorsal view.
5. Photograph using a fluorescence stereomicroscope (see Fig. 2c).
3.9 Whole-Mount To determine if thymocytes are undergoing apoptosis, we employ

TUNEL Assays TUNEL staining.
1. Fix Embryos in 4 % PFA overnight at 4 °C.
2. Dehydrate in graded methanol/PBS (50, 70, 95, 100 %),
incubating for 5 min at room temperature at each step and
then storing at −20 °C overnight.
3. Incubate for 10 min in acetone at −20 °C.
4. Wash embryos twice in PBS for at least 5 min per wash.
5. Permeabilize the embryos for 15 min at RT in 0.1 % Triton
X-100 plus 0.1 % sodium citrate/PBS (freshly prepared).
Longer incubation times may be necessary when working on
older embryos.
6. Wash twice in PBS, 10 min per wash.
Fig. 2 Using zebrafish model to investigate the role of Rpl22 in T cell development. (a and b) Targeted knock-
down of Rpl22 in zebrafish embryos. (a) For Rpl22, the MO is designed to specifically block the initiation of
translation at the start codon in exon 1. The short black line and blue arrow indicate the positions of sequences
targeted. (b) The effect of 1 ng Rpl22-MO knockdown on expression of endogenous Rpl22 is assessed by
immunoblotting on 5dpf embryos. (c) Rpl22-MO injected (1 ng), but not MM (5-base pair mismatch) control
injected, Tg(lck:EGFP) fish exhibit a loss of EGFP-marked T cells at 5 dpf (lateral view, red circles; dorsal view,
yellow rectangles). Numbers refer to fraction of morphants with the depicted phenotypes. (d) Injection of 1 ng
Rpl22-MO causes apoptosis in the thymus at 3.5 dpf, while injection of the same quantity of 5-base pair mis-
match control morpholino (Rpl22-MM) caused no change in the thymus (green dashed rectangles). The white
dashed line delimits the eye. Numbers refer to fraction of morphants with the depicted phenotypes. (e–h) The
use of a heat-shock inducible, bidirectional expression construct to restore T cell development in rpl22 mor-
phants. (e) Upper panels depict schematics of the heat-inducible expression plasmid pSGH2-mRpl22. (f) The
heat shock-inducible expression plasmid associated with I-sceI mediated transient overexpression is injected
into one-cell stage embryos. At 3dpf, the embryos are heat shocked at 37 °C for 1 h and then effects on T cell
development are analyzed at 5 dpf using WISH for lck, a marker of thymic progenitors. (g) After heat-shock,
GFP+ embryos are selected for subsequent WISH analysis (pink arrowheads). (h) Heat shock induction of
mouse Rpl22 expression rescues T cell development in rpl22 morphants (5dpf lck WISH staining, red rectan-
gles). Numbers refer to fraction of morphants with the depicted phenotypes
7. Add 50 μl enzyme solution (containing terminal deoxynucleotidyl

transferase) to 450 μl label solution (containing nucleotide
mixture in reaction buffer) to obtain a total of 500 μl
TUNEL reaction mixture, then mix well (Roche TMR Kit
instructions).
8. Incubate embryos for 1 h at 37 °C in the dark in TUNEL reac-
tion mixture in a 1.5 ml, round-bottom microcentrifuge tube.
100–150 μl should be enough for up to 50 embryos.
9. In the dark, wash for 2 h (4 × 30′ in PBST), then photograph
(Fig. 2d).
3.10 Whole-Mount Whole mount in situ hybridization (WISH) employs an antisense

In Situ Hybridization ribonucleotide probe and can be used for a number of purposes
including determining the tissue distribution and gross expression
level of an mRNA species and/or the presence of a particular cell
population marked by a lineage-restricted mRNA.
1. To prepare DNA for making riboprobes, linearize the plasmid
encoding your probe and purify by two rounds of phenol–
chloroform extraction. Check 1 μl of digested DNA on a 1 %
agarose gel to ensure the DNA template is linear. Quantify the
DNA concentration of the template (e.g., Tecan NanoQuant).
2. To transcribe the riboprobes, add the following to the tran-
scription reaction: 1 μg of linearized DNA template, 2 μl of
10× transcription buffer, 2 μl of DIG-labeling RNA mix, 1 μl
of RNase inhibitor (40 U/μL), 2 μl of RNA polymerase (T7,
T3, or Sp6), and enough nuclease-free water to bring the reac-
tion to 20 μl. Incubate the reaction at 37 °C for 2 h, then
purify the riboprobe using a Ambion NucAway spin column.
Assess the quality and yield of the probe by running 1 μl of the
probe on an agarose gel and quantifying the concentration.
Probes can be stored at −80 °C for up to 2 years.
3. Dechorionate and fix 24 h embryos using pronase. After
dechorionation, grow the embryos in egg water containing
0.003 % PTU to prevent pigmentation, which starts to develop
at around 24 hpf. Fix embryos at 5dpf (or another developmental
stage appropriate for your experiment) with 4 % fix solution
and rock overnight at 4 °C (see Note 6). Thymocyte seeding
can be observed by 3.5–4dpf, but the optimal time to assess
whether knockdown of a gene blocks development is 5dpf.
Wash embryos in PBST three times, 5 min each. Wash with
100 % methanol, incubate for 5 min, then transfer the embryos
to fresh methanol and store at −20 °C.
4. Rehydrate embryos for 5–10 min each in 75 % Methanol/
PBST, 50 % Methanol/PBST, and 25 % Methanol/PBST. Wash
embryos in PBST once quickly, then twice more for 5 min each.
5. Digest embryos with 10 μg/ml Proteinase K in PBST for

30 min at room temperature with gentle rocking. These condi-
tions work well for zebrafish embryos from 3 to 5 dpf.
Proteinase K is made from a 10 mg/ml stock stored at −20 °C.
6. Remove the proteinase K with three quick PBST washes.
7. Re-fix embryos for 1 h at room temperature. This step is neces-
sary to eliminate residual proteinase K.
8. Remove the fixative with three quick PBST washes.
9. Hybridization: Preheat in situ hybridization buffer Hyb(−) and
Hyb(+) solutions to 65 °C. The temperature should be consis-
tent throughout the hybridization procedure. Add Hyb(−) to
the 1.5 ml tubes containing embryos and shake gently for
15 min at 65 °C. Replace the Hyb(−) with Hyb(+) and shake
gently for 1 h at 65 °C. Add probes (typical range is 0.3–
1.0 ng/μl) and rock gently overnight at 65 °C.
10. Wash embryos in 2× SSCT/50 % formamide twice for 30 min
at 65 °C.
11. Wash embryos in 2× SSCT twice for 15 min at 65 °C.
12. Wash embryos in 0.2× SSCT twice for 30 min at 65 °C.
13. Wash embryos in MABT three times at room temperature,
10 min each.
14. Transfer the embryos from MABT to complete blocking solu-
tion in 12-well plates containing 15 mm Netwell™ inserts with
polyester membrane and incubate with gentle shaking for at
least 2 h at room temperature (see Note 7).
15. Remove complete blocking solution and replace it with anti-
digoxygenin-AP antibody at 1:5000 dilution in complete block-
ing solution and rock overnight at 4 °C.
16. Wash embryos in complete blocking solution, 1 h at room
temperature.
17. Wash embryos in MABT solution, 30 min at room temperature.
18. Wash embryos in staining buffer at room temperature,
3 × 5 min, with gentle rocking.
19. Wash in 0.1 M Tris pH 9.5 for 5 min at RT with gentle
rocking.
20. Add Vector BCIP/NBP staining solution. Wrap the plate in
aluminum foil and incubate in the dark at room temperature
for 20–60 min monitoring the development of staining every
20 min. Alternatively, the embryos can be stained overnight
at 4 °C.
21. Stop the staining reaction by washing in PBST with gentle
rocking 3× for 5 min at room temperature.
22. Replace PBST with 50 % glycerol/PBS and wash for 15 min at

room temperature. Glycerol is the mild clearing reagent that
makes the stained embryos more transparent for photography.
Transfer embryos to a six-well plate containing 100 % glycerol,
then store at 4 °C.
23. Photograph. Effects of gene knockdown on development of
the T lineages can be assessed using the lck probe (see Fig. 2h).
3.11 Heat Inducible In zebrafish, it is critical to perform rescue experiments to ensure

Gene Expression that phenotypes resulting from MO-mediated gene knockdown
to Rescue are truly due to the absence of the intended target, rather than due
Developmental Arrest to off-target effects. This can be conveniently accomplished by
Caused by Gene injecting MO treated embryos with MO-resistant forms of the
Knockdown mRNA encoding the intended target; however, because injected
mRNA lasts only a few days, mRNA overexpression is only effective
for phenotypes at early embryonic stages. For later developmental
stages more than 3dpf, heat-shock inducible DNA constructs can
be used to temporally overexpress exogenous genes to study T cell
development [6, 20]. The I-SceI meganuclease was originally uti-
lized in Medaka to induce stronger promoter activity in the F0
founder. Coinjection of I-SceI significantly improves the specificity
of transient expression and decreases the mosaicism in zebrafish [23].
We have employed this approach to ensure that ectopic expression
of Rpl22 can restore the block in T cell development observed in
rpl22 morphants at 5dpf. In doing so, we heat induced expression
of mouse Rpl22 at 3dpf and we identified phenotypic rescue of
thymocyte development by performing WISH with an lck probe at
5dpf (Fig. 2e–h).
1. To enable heat-inducible expression of a gene of interest, clone
the cDNA of the gene in question into pSGH2 [20]. pSGH2
enable simultaneous expression of the gene of interest as well
as GFP to mark embryos that have productively incorporated
the vector (see Fig. 2e).
2. Upon cloning, prepare purified pSGH2 expressing the gene of
interest, using a high purity plasmid preparation kit.
3. Collect the one-cell stage embryos and align them on the
microinjection plate (see Note 8).
4. Mix pSGH2 plasmid with 0.5× I-SceI buffer and 0.5 U/μl
I-SceI meganuclease (New England Biolabs) and microinject
(100 pg) directly into the cytoplasm of one-cell stage embryos.
5. At 24hpf, check the quality of the injected embryos and remove
the dead embryos. Dechorionate the embryos by pronase
treatment and then culture at 28.5 °C to 3dpf.
6. At 3dpf, transfer the injected embryos into the 1.5 ml tubes

filled with 1 ml egg water, heat shock at 37 °C water bath for
1 h, then resume culturing the embryos at 28.5 °C (see Note 9).
7. pSGH2 vector has the heat shock-inducible bidirectional pro-
moter consisting of multimerized heat shock elements (HSE).
Thus, this heat shock-inducible expression construct allows
temporal control of ectopic expression of the gene of inter-
est, coupled with EGFP marking of expressing embryos
(see Fig. 2e–g). After heat-shock, GFP+ embryos are selected
for analysis (Fig. 2g). In Fig. 2h, WISH with an lck probe was
performed to identify developing T cells. This revealed that
the developmental arrest caused by Rpl22 knockdown can be
rescued by a morpholino resistant form of Rpl22 (pSGH2-
Rpl22) in a heat-inducible manner.
4 Notes
1. Methylene blue is used to prevent fungal growth in egg water;

however, it can also induce autofluorescence, especially for
embryos after 3dpf. Therefore, for observation under fluores-
cent microscopy, we recommend that embryos should be
cultured in egg water without methylene blue.
2. For morpholino storage, GeneTools recommends to store ali-
quots at RT. We use airtight microtubes to prevent evapora-
tion. If morpholino aliquots are stored for a long time,
GeneTools recommends freeze-drying the oligo. MOs should
be heated up at 65 °C for 10 min before microinjection to
dissolve possible precipitates.
3. Due to the low stability of I-sceI meganuclease, aliquots of
enzyme should be prepared (e.g., 2 μl) upon arrival and stored
at −80 °C. The microinjection mix should be prepared imme-
diately before injection and kept on ice.
4. We inject the morpholino at one-cell stage to get more consis-
tent results for T cell phenotypes at 5dpf.
5. We have an excellent antibody for testing the efficiency of
Rpl22 translation-blocking MO. Figure 2b shows a sample
immunoblot of Rpl22 expression in controls and embryos
treated with 1 ng of a morpholino that blocks translation
(Fig. 2a, b). However, sometimes antibodies are not available
for detection of zebrafish proteins. Thus, an alternative way
of evaluating the efficacy of start-site morpholinos that do not
alter mRNA size, is to produce a chimeric mRNA comprising
the target sequence surrounding the target genes’ ATG fused
in frame with GFP-coding sequence. Co-injection of this
fusion mRNA with MO will reveal whether the morpholino
blocks GFP synthesis, which serves as a surrogate marker for

expression of the endogenous target protein. Splice-blocking
morpholinos can also be used to knockdown target genes, and
the efficiency checked by assessing the predicted size shift by
RT-PCR.
6. For embryos older than 18 hpf, the chorion should be removed
before fixation to avoid a curved tail after fixation.
7. We use inserts with a polyester membrane to conveniently
transfer the embryos between wells during WISH washing and
staining steps.
8. The heat-inducible plasmid should be directly injected into
the cytoplasm of the cell. We strictly use only one-cell stage
embryos to ensure that the transient overexpression has consis-
tent effects on phenotypic rescue.
9. Temperature or incubation time of heat-shock treatment may
be varied for different ectopically expressed genes. Increasing
the temp/incubation time may increase the extent of overex-
pression; however, excessive heating may also kill embryos.
Different conditions should be tested to identify the best
compromise between overexpression and survival.
Acknowledgments
We are grateful to the support and help from Dr. Jennifer Rhodes,
Allison Ulrich and Alison N. Bilbee for the zebrafish work.
We gratefully acknowledge the assistance of the following core
facilities of the Fox Chase Cancer Center: Flow Cytometry, DNA
Sequencing, Imaging and Laboratory Animal/Zebrafish. This work
was supported by NIH grants AI081814, AI073920, NIH core
grant P01CA06927, Center grant P30-DK-50306. Y.Z. is a W.J.
Avery Postdoctoral Fellow of Fox Chase Cancer Center.
References
1. Amatruda JF, Zon LI (1999) Dissecting established by the Notch-Runx pathway. Genes
hematopoiesis and disease using the zebrafish. Dev 19(19):2331–2342
Dev Biol 216(1):1–15 5. Rhodes J, Hagen A, Hsu K, Deng M, Liu TX,
2. Lieschke GJ, Currie PD (2007) Animal models Look AT, Kanki JP (2005) Interplay of pu.1
of human disease: zebrafish swim into view. and gata1 determines myelo-erythroid progen-
Nat Rev Genet 8(5):353–367 itor cell fate in zebrafish. Dev Cell 8(1):
3. Trede NS, Langenau DM, Traver D, Look AT, 97–108
Zon LI (2004) The use of zebrafish to under- 6. Bajoghli B, Aghaallaei N, Hess I, Rode I,
stand immunity. Immunity 20(4):367–379 Netuschil N, Tay BH, Venkatesh B, Yu JK,
4. Burns CE, Traver D, Mayhall E, Shepard JL, Kaltenbach SL, Holland ND, Diekhoff D,
Zon LI (2005) Hematopoietic stem cell fate is Happe C, Schorpp M, Boehm T (2009)
Evolution of genetic networks underlying the 15. Hess I, Boehm T (2012) Intravital imaging
emergence of thymopoiesis in vertebrates. Cell of thymopoiesis reveals dynamic lympho-
138(1):186–197 epithelial interactions. Immunity 36(2):
7. Murayama E, Kissa K, Zapata A, Mordelet E, 298–309
Briolat V, Lin HF, Handin RI, Herbomel P 16. Langenau DM, Feng H, Berghmans S, Kanki
(2006) Tracing hematopoietic precursor mig- JP, Kutok JL, Look AT (2005) Cre/lox-
ration to successive hematopoietic organs regulated transgenic zebrafish model with con-
during zebrafish development. Immunity 25(6): ditional myc-induced T cell acute lymphoblastic
963–975 leukemia. Proc Natl Acad Sci U S A 102(17):
8. Schorpp M, Bialecki M, Diekhoff D, Walderich 6068–6073
B, Odenthal J, Maischein HM, Zapata AG, 17. Langenau DM, Ferrando AA, Traver D, Kutok
Boehm T (2006) Conserved functions of Ikaros JL, Hezel JP, Kanki JP, Zon LI, Look AT,
in vertebrate lymphocyte development: genetic Trede NS (2004) In vivo tracking of T cell
evidence for distinct larval and adult phases of development, ablation, and engraftment in
T cell development and two lineages of B cells transgenic zebrafish. Proc Natl Acad Sci U S A
in zebrafish. J Immunol 177(4):2463–2476 101(19):7369–7374
9. Willett CE, Cherry JJ, Steiner LA (1997) 18. Nasevicius A, Ekker SC (2000) Effective tar-
Characterization and expression of the recom- geted gene ‘knockdown’ in zebrafish. Nat
bination activating genes (rag1 and rag2) of Genet 26(2):216–220
zebrafish. Immunogenetics 45(6):394–404 19. Kimmel CB, Ballard WW, Kimmel SR, Ullmann
10. Langenau DM, Zon LI (2005) The zebrafish: B, Schilling TF (1995) Stages of embryonic
a new model of T-cell and thymic develop- development of the zebrafish. Dev Dyn 203(3):
ment. Nat Rev Immunol 5(4):307–317 253–310
11. Ma D, Wei Y, Liu F (2013) Regulatory mecha- 20. Bajoghli B, Aghaallaei N, Heimbucher T,
nisms of thymus and T cell development. Dev Czerny T (2004) An artificial promoter con-
Comp Immunol 39(1-2):91–102 struct for heat-inducible misexpression during
12. Zhang Y, Duc AC, Rao S, Sun XL, Bilbee AN, fish embryogenesis. Dev Biol 271(2):416–430
Rhodes M, Li Q, Kappes DJ, Rhodes J, Wiest DL 21. Link V, Shevchenko A, Heisenberg CP (2006)
(2013) Control of hematopoietic stem cell emer- Proteomics of early zebrafish embryos. BMC
gence by antagonistic functions of ribosomal pro- Dev Biol 6:1
tein paralogs. Dev Cell 24(4):411–425 22. Anderson SJ, Lauritsen JP, Hartman MG,
13. Kissa K, Murayama E, Zapata A, Cortes A, Foushee AM, Lefebvre JM, Shinton SA,
Perret E, Machu C, Herbomel P (2008) Live Gerhardt B, Hardy RR, Oravecz T, Wiest DL
imaging of emerging hematopoietic stem cells (2007) Ablation of ribosomal protein L22
and early thymus colonization. Blood selectively impairs alphabeta T cell develop-
111(3):1147–1156 ment by activation of a p53-dependent check-
14. Bertrand JY, Kim AD, Teng S, Traver D (2008) point. Immunity 26(6):759–772
CD41+ cmyb+ precursors colonize the zebraf- 23. Gupta S, Zhu H, Zon LI, Evans T (2006)
ish pronephros by a novel migration route to BMP signaling restricts hemato-vascular devel-
initiate adult hematopoiesis. Development opment from lateral mesoderm during somito-
135(10):1853–1862 genesis. Development 133(11):2177–2187
INDEX
A CD249............................................................................... 71
CD326............................................................. 66, 68–70, 72
Aire................................................. 11, 14, 27, 42, 65, 66, 70 CD1a ....................................................... 222, 229, 233, 248
Allelic exclusion......................................................... 39, 186 CD16/CD32 antibody ............................ 111, 113, 213, 215
7-Aminoactinomycin D (7-AAD) ........ 55, 59–62, 217–219 CD1d tetramers............................................ 6, 51, 55, 56, 60
alternatives ........................................................... 62, 219 CD51, mouse .................................................................. 250
Antibiotics ....................................................... 110, 111, 113 CD4 SP ................................ 14, 15, 48–51, 53–57, 222, 232
Antigen-presenting cells (APCs) ........................... 5, 14, 69, CD8 SP .......................... 15, 16, 37, 48–51, 53–56, 222, 232
117, 118, 120, 122–127, 142, 171, 173, 177, 255 Cell cycle .................................... 62, 101, 212, 214, 217, 218
Apoptosis.......................... 141, 142, 148, 169, 186, 285, 286 Central memory ................................................................ 51
agonist induced .......................................................... 142 Changes in repertoire ...................................................... 189
Chimeras
B
bone marrow ........................... 36, 41, 101, 104, 109–115
Bcl2 ......................................................................... 8, 10, 12 fetal liver ............................................................ 109–115
Bim .................................................................................... 12 Click reaction .................................................................. 212
BLAST............................................................................ 280 Collagenase ................................ 66, 67, 69, 71, 72, 263, 266
Bone marrow............................................... 6, 7, 14, 36, 41, 47, 75, Collection tubes, polystyrene ....................................... 95, 97
100, 101, 103–106, 109–115, 159, 171, 221, 225–227, Complement
229, 234, 246, 250, 258, 260, 261, 264, 265 guinea pig .................................................. 110, 112, 114
T cell depletion .......................................... 112–114, 171 rabbit ................................................. 110, 114, 161, 165
Bone marrow growth medium (BMGM)................ 102–104 Conditioning mix ............................................ 118, 122–126
Bone marrow transplantation (BMT) ..................... 109, 113 Confocal microscopy ................................. 82, 143, 170–175
Bone wax ................................................................. 262, 265 Continuous labelling ....................................................... 212
5-Bromo-2-deoxyuridine (BrdU) ............ 211, 212, 218, 219 Corticomedullary junction (CMJ) ..................................... 75
Busulfan........................................................................... 261 Cre recombinase ................................................................ 36
Cytotoxic T (Tc) ...................... 118–120, 122, 123, 125, 223
C
D
CCR7 ........................................................... 7, 10, 14, 17, 65
CD3 .................... 4, 8, 50, 118, 222, 224–225, 228–233, 248 Danieau buffer ................................................................. 276
CD7 ........................................................ 221, 222, 246–248 Death by neglect .............................................. 142, 143, 169
CD8 .............................................. 4–6, 9, 10, 12, 13, 15–18, Dechorionation ................................ 277, 282, 284, 287, 289
25–27, 37–39, 41, 47–56, 75, 76, 88, 91, 92, 96, 118, Defects in V(D)J recombination.............................. 189, 192
120, 125, 127, 144, 147, 160, 161, 170, 185, 192, Deletion reporters.............................................................. 36
222–226, 229–232, 236, 248 Dendritic epidermal T cells (DETC) .................... 25–28, 30
CD25 ........................... 15, 16, 49–51, 57, 62, 161, 192, 222 2′-Deoxyguanosine (dGuo) .............. 151, 152, 154, 155, 157
CD27 ......................................................... 9, 15, 17, 26, 185 Dermal fibroblast..................................................... 160, 162
CD28 .............................................. 9, 15, 17, 117, 118, 125, Developmental arrest ............................................... 289–290
185, 222, 223, 225, 231 4′,6-Diamidino-2-phenylindole
CD34+..................................... 221, 222, 225, 227–229, 234, (DAPI) ................................................... 62, 78, 85,
235, 242, 243, 246, 248, 250, 261, 263–269 97, 161, 171,214, 217–220
CD38 ...................................................................... 225, 228 Dimethyl sulfoxide (DMSO) ......................... 115, 162, 213,
CD44 .......................... 15, 16, 49, 50, 55, 120, 161, 192, 222 214, 216, 218, 223, 240, 242, 249
CD73 .......................................................................... 27, 29 Dispase ...................................................................... 71, 263
CD80 ................................................................................ 71 DN1 ...................................................................... 15, 16, 49
CD117 (cKit) ...................................................... 15, 16, 165 DN3 ................................................... 15–17, 37, 49, 50, 142
CD205................................................................... 66, 69–71 DNA content analysis ..................................................... 212
vol. 1323, DOI 10.1007/978-1-4939-2809-5, © Springer Science+Business Media New York 2016
293
T-CELL DEVELOPMENT: METHODS AND PROTOCOLS
294 Index
DNA dyes........................................................................ 219 Gene inactivation .............................................................. 36

Double negative (DN) ........................................ 5, 8, 14, 15, Gene knockdown............................................. 281, 285, 289
38, 47–52, 56, 60, 185, 192, 222 Glycophorin-A ................................................................ 224
Doublet ............................................................. 97, 217, 236 Graft-versus-host disease (GVHD)........................ 112, 113,
255, 258
E Granulocyte-colony stimulating factor
Early thymic precursors (ETP)........................................ 119 (G-CSF) ...................................................... 241, 258
Early T progenitors (ETP) ........................................ 15, 222 Granulocyte-macrophage colony-stimulating factor
E-box binding proteins................................................ 10, 13 (GM-CSF), human ............................................. 241
Effector memory ....................................... 51, 119, 120, 127
H
Egr3............................................................................. 29, 30
EpCAM ............................................................................ 71 Hamilton syringe ............................................. 206, 262, 265
Erk..................................................................................... 12 Hanging-drop cultures ............................................ 155–156
Ethidium bromide ............................ 187, 193, 194, 197, 200 Hapten .................................................... 225, 232, 233, 236
5-Ethynyl-2′-deoxyuridine (EdU) .......................... 211–220 HEK 293T ...................................................... 101, 161, 163
Ets1 ................................................................................... 13 Hematopoetic stem cell (HSC) .......................... 37, 38, 100,
Expected cell numbers 101, 105, 106, 109, 110, 112–114, 159, 161, 222, 255,
LN ......................................................................... 90–92 256, 258–261, 263–265, 268
splenocytes .................................. 120, 122–124, 127, 171 Hematopoiesis ................................................. 114, 273, 274
thymocytes................................................................... 59 Hematopoietic progenitor cells (HPCs) ................. 105, 151,
Extrathymic CD34 .................................................. 227, 228 221, 234, 240, 250, 258
Hemocytometer................................. 53, 110, 172, 223–225,
F 230, 240, 241
Feeder cells .............................................................. 118, 245 HLA-A2 ................................................................. 255, 257
Fetal bovine serum (FBS) ................................ 89, 92, 93, 95, HLA-DR ................................................ 246, 248, 255, 257
96, 101, 102, 114, 133, 152, 154–156, 160, 162, 166, HLA typing............................................................. 262, 268
170, 172, 218, 223–225, 230, 240, 242, 249, 282 Host-versus-graft disease (HVGD) ................................. 113
heat-inactivated .................................. 89, 160, 223–226, HoxA3............................................................................... 14
228, 240, 241, 279 Humidity chamber .................................................... 83, 171
Fetal liver (FL) ............................................ 24, 47, 100, 103,
109–115, 161, 164–165, 258, 260–262, 266–269 I
Fetal thymus ............... 24, 151, 154, 155, 262, 266, 268, 269 ikaros ........................................................................ 273, 274
Fetal thymus organ culture (FTOC) ........................ 41, 131, IL2Rγ chain ............................................................ 254, 261
141–148, 151, 177, 233, 240, 243, 249 Il2rgnull ..................................................................... 254, 256
Fibroblast media .............................................. 160, 162–164 Immunoblot.................................................... 143, 166, 169,
Fixation time ............................................................... 83, 84 170, 177, 178, 285, 286, 290
Flow cytometric sorting........................... 88, 89, 93–95, 127 Immunofluorescence antibody (IFA) staining ................... 76
Flow cytometry........................ 15, 17, 36, 47–63, 65–72, 89, Immunoprecipitation followed by western blotting
91–94, 103, 104, 114, 132, 133, 137–138, 143, 147, (IP-Western)................................................ 143, 170
156, 161, 164–166, 170, 177, 192, 199, 211, 212, 214, Inducible Treg (iTreg)..................................... 118–120, 122,
216–218, 224, 233, 243, 245, 246, 248, 249, 264 123, 125, 127, 222
Flt-3 ligand, human (Flt-3 L) ................................. 240, 241 Injection
Fluorescence minus one (FMO)........................................ 63 intrafemoral ........................ 258, 261, 262, 264–265, 268
Fluorescent lineage tracers............................................... 274 intrahepatic ................................................ 261–265, 268
5-Fluorouracil (5-FU) ............................................. 103, 105 intraperitoneal (IP) ............................ 115, 204, 214, 254
FoxP3 .................................................. 11, 13, 52, 55, 60, 62, intrathymic ................................................ 101, 203–208
63, 119, 125, 127 retro-orbital ....................................................... 115, 255
Freezing medium ............................................. 162, 165, 166 In situ hybridization ................................................ 274, 288
Interleukin-9 (IL-9), human ........................................... 254
G
Interleukin-15 (IL-15), human ............... 241, 245, 254, 261
Gamma irradiator ............................................ 259, 261, 268 Intracellular dye ............................................................... 269
Gata3 .......................................................... 7, 9, 13, 119, 274 Intracellular staining ....................................... 52, 54–57, 60,
Gelatin sponges ....................................................... 153, 157 62, 63, 66–70, 143, 170, 185, 215
Gel-foam sponges............................................ 144, 145, 147 Intraepithelial lymphocytes (IELs)...................................... 6
Index
295
IP followed by flow cytometry (IP-FCM) ............. 143, 170, Mucosal-associated invariant T cells (MAIT) ..................... 6
172, 175–177 Murine dermal fibroblasts expressing Dll4
IRF1 .................................................................................. 13 (mFibro-DL4) ..................................... 160, 164–166
Irradiation....................................................... 109, 111, 112, Murine stem cell virus (MSCV) ............... 99, 100, 102, 249
115, 197, 259, 261, 263, 268, 269 Mutagen .................................................................. 218, 219
I-sceI meganuclease ......................................... 277, 289, 290
ISP.................................................. 48–51, 53, 223, 225, 235 N
Natural killer NK cells .................................... 221, 228, 245,
K
253, 254, 260
Ketamine/xylazine ................................................... 204, 208 Natural killer T (NKT)..................................... 6, 51, 54, 56,
222, 223, 253
L Negative selection ................................ 11–12, 14, 27, 40–42,
lck...................................... 4, 37, 38, 274, 275, 286, 289, 290 50, 65, 87–89, 131, 132, 141–148, 169, 170, 223, 254
Lentiviral vectors ..................................................... 100, 104 Nomenclature
Lentivirus ......................................... 100–102, 105, 240, 249 historical ............................................ 181–182, 189–191
Lineage commitment ..................................... 7, 9, 15, 27, 28 IMGT/NCBI .................................... 180, 181, 183, 184
instructional ................................................................. 28 TCRα .......................................................... 25, 184, 189
stochastic ..................................................................... 28 Non-homologous end-joining (NHEJ) ........................... 179
Lineage deviation .................................................... 148, 169 Notch ........................................................ 7–9, 14, 159, 160,
Lineage markers ...................................... 101, 228, 282–283 206, 239, 241, 246, 250
Lineage mix ........................................................... 50, 58, 60 Notch ligand Delta-like....................................... 7, 159–167
Lipofectamine ......................................................... 102–104 Nozzle, 100mm ..................................................... 72, 94, 95
Live gate ...................................................................... 61, 62 Nucleoside analogs .......................................................... 212
Lymph nodes (LN) Nude mouse....................................................................... 13
mesenteric.................................................................... 96 Nur77 ................................................................................ 12
peripheral............................................................... 57, 96 Nylon mesh ......................................... 54, 57, 58, 66–68, 70,
133, 147, 153, 155, 162, 166, 167, 172, 235, 280, 282
M
O
Macrophages ............................................... 14, 54, 241, 254
Magnetic beads, antibody coated .......................... 87, 90–93, OP9 co-culture system .................... 240–242, 244–248, 250
114, 120, 121, 230 OP9-DL.......................................... 159–161, 164–166, 246
Major histocompatibility complex (MHC) .............. 5, 7–13, OP9 stromal cells expressing human DLL4
169, 254, 255, 258, 260 (OP9-hDLL4)..................................... 241, 243, 250
MAP kinases, selection...................................................... 12 Ortholog .......................................................... 274, 280–281
Methylene blue ........................................................ 275, 290
P
MHC Class I..................................................... 24, 254, 255
MHC Class II ................................................................. 255 Paraffin embedding ................................... 76–78, 80–81, 84
MHC, non-classical .................................................... 24, 30 Paraffin permeabilization................................................... 84
β2 Microglobulin ............................................... 41, 254, 257 Paraformaldehyde (PFA) ...................................... 63, 66, 68,
Microinjection .................................................. 38, 276–277, 76–80, 82, 84, 278, 285
281, 283–284, 289, 290 Plat-E cells .............................................................. 156, 158
mobilized peripheral blood .............................. 258, 260, 261 PLZF ................................................................................ 29
Morpholinos (MO) .......... 274, 276, 277, 283–286, 290, 291 Polybrene .......................................... 102–104, 106, 153, 156
Mouse model Polymerase chain reaction (PCR)
BALB/c-Rag2null (BRG) ................... 115, 183, 254–256, conventional ....................... 187, 189, 193, 195–196, 283
259, 261, 269 quantitative real-time......................... 167, 187, 193–197
BLT ............................................ 255, 258, 260, 263, 267 Positive selection.......................................... 9, 10, 12, 17, 40,
Hu-HSC ................................................... 255, 258, 260 50, 65, 87, 89, 131, 132, 142, 143, 147, 169, 186
NOD/LtSz-scid Il2rgnull (NSG)................. 254, 256, 262 Pre-T cell receptor (pTα)........................................... 27, 223
NOD/Rag1null IL2rgnull (NRG).......................... 254, 256 Primary mouse fibroblast ......................................... 162–163
NOD-scid ................................................................. 254 Prkdc ................................................. 253, 254, 256, 257, 262
NOD/Shi-scid Il2rgnull (NOG) .......................... 254, 256 Propidium iodide (PI) ..................................... 62, 66, 68, 69,
SCID HU.................................................. 255, 258, 260 219, 248, 250, 282
SCID PBL ........................................................ 258–261 Pseudotyping .................................... 100, 106, 111, 180, 186
296 Index
pSGH2 vector ......................................................... 277, 290 T cell depletion, complement ......................... 110, 112–114,
Pulse labelling.......................................................... 214, 219 120, 123–124, 127, 171
Puma ................................................................................. 12 T cells
CD4+ ..................................... 5, 9, 12, 13, 16, 25, 27, 37,
R 38, 41, 47–49, 76, 88, 96, 120, 127, 185, 192, 222,
RANK ............................................................................... 14 223, 225, 226, 229–232, 235, 236, 246
Recombination CD8+ ........................................... 5, 9, 10, 12, 13, 16, 25,
TCRα ............................................ 8, 179, 183–187, 189, 27, 37, 38, 41, 47–49, 76, 88, 96, 118, 120, 125, 127,
192, 200, 223, 273 185, 192, 236
TCRβ ............................. 8, 180–183, 185, 187, 192, 273 γδ .............................. 4, 6–8, 15, 23–30, 52, 58, 222, 273
Recombination activating gene 1 (Rag1)................ 8, 10, 37, naïve ................................................ 7, 29, 37, 38, 51, 55,
41, 179, 253, 254, 256, 257, 273, 274 117, 119, 120, 122
Recombination activating gene 2 (Rag2)................ 8, 10, 41, regulatory............................................. 11, 13, 51, 57, 62,
115, 179, 253, 254, 256, 261, 273, 274 65, 118, 141, 222, 223
Recombination signal sequences (RSSs).......... 179, 180, 182 TCR-α .......................................... 4, 5, 7–10, 12, 38–42, 50,
Reconstitution ........................................... 41, 101, 106, 109, 179–180, 183–192, 196, 223, 273, 274
110, 113, 115, 151–158, 162, 165, 255, 260, 264, 265 TCR-β............................................. 4, 5, 7–9, 15–17, 39–42,
Restriction fragment length polymorphisms ................... 183 49–54, 180–183, 185–188, 192, 196, 273, 274
Retinoic acid .................................................... 119, 125, 127 TCR-δ............................................... 4, 8, 52, 183, 184, 186,
Retronectin ............................... 102, 104–106, 240, 243, 249 222, 223, 273, 274
Retroviral transduction TCR-γ ....................................................... 4, 8, 9, 16, 39, 52,
hematopoietic cells ......................... 41, 99, 103–104, 161 221, 233, 273, 274
lymphoid cell lines ............................. 102, 104, 106, 109 TCR signal, nuclear targets of .......................... 4, 10–12, 16,
primary thymocytes ................................................... 104 28–30, 118, 223
Retrovirus Terasaki plates ......................................................... 153, 155
host range ............................................................ 99, 100 Teratogen......................................................................... 218
vectors, FIV-based ..................................................... 100 Th0 .......................................................... 118, 122, 123, 125
vectors, HIV-1 based ................................................. 100 Th1 ................................................... 118–120, 122, 123, 125
vectors, MSCV .................................... 99, 100, 102, 249 Th2 ................................................... 118–120, 122, 123, 125
RORγt ...................................................... 9, 29, 30, 101, 119 Th17 ......................................... 118–120, 122, 123, 125, 126
Runx ...................................................................... 9, 13, 274 Th17, Iscove's modified Dulbecco's medium
(IMDM) ............................................120, 125, 126,
S 223–226, 228, 240, 242, 243, 249
ThN......................................................................... 118, 125
Schnurri-2 ......................................................................... 12
Thoracotomy ........................................................... 204–208
SCID mice ...................................................... 253–260, 262
Thpok .......................................................................... 13, 38
Self-inactivating viruses (SIN) ................................ 100, 105
Thrombopoietin (TPO), human ............................. 240, 241
Serum testing, bovine ...................................... 223, 233, 249
Thymic epithelial cell (TEC) ............................... 6, 7, 9, 13,
Silanized slides .................................................................. 82
14, 27, 65–72, 75–77, 109, 142, 143, 169, 254, 260
Skint1 .......................................................................... 28, 30
cortical (cTEC) ................................... 65, 66, 69–72, 76
Southern blot............................ 186–188, 190, 193, 197–200
medullary (mTEC)......................... 27, 65, 66, 69–72, 76
Sphingosine 1 phosphate (S1P)......................................... 13
Thymic epithelium
Splice-blocking morpholinos................................... 276, 291
cortex ........................................ 6, 9, 10, 14, 65, 138, 142
Stat5 .................................................................................. 13
medulla ................................... 10–12, 14, 27, 42, 65, 142
Stem cell factor, human .................................. 102, 162, 165,
Thymic nurse cells (TNC) ................................................ 72
240, 241, 243, 245, 255, 261
Thymic organ culture
Stromal cells, thymic ...................................... 66–68, 71, 75,
fetal ...................................................... 41, 131, 142, 166
76, 131, 151, 152, 159–167, 241, 244, 250
reaggregate................................................................. 131
Surgery, cardiac........................................ 225, 234, 260, 267
reconstituted ................................................ 41, 151–158
T Thymic slices ........................................................... 131–139
Thymocytes
β5t ......................................................................... 65, 66, 70 average lifespan of DP ............................................... 186
TAP transporters ............................................................... 41 ISP............................................ 37, 48–51, 223, 225, 235
Index
297
Thymoproteasome, β5t ...................................................... 65 U

Timed matings ........................................................ 145, 263
Timed pregnancies .................................. 110, 114, 153, 164 Ulex europaeus agglutinin-1 (UEA-1) .............. 66, 68, 69, 71
Tissue morphology, fixation .............................................. 77 Ultrasound-guided .................................................. 203, 204
Tissue preparation Umbilical cord blood ............................................... 258, 261
fixed frozen ............................................................ 77–81 V
fresh frozen ............................................................ 77–79
paraffin embedded ................................................. 78, 80 Vector plasmid ................................................................. 102
Tox............................................................................. 13, 110 Viral packaging cell ......................................................... 163
TRAIL .............................................................................. 14
W
Transfection, transient ............................................. 101–102
Transgenic TCR ............................................. 12, 28, 38–42, Western blot ............................... 36, 170, 276–278, 284–285
118, 144, 170 Whole mount in situ hybridization
Transplantation needle .................................... 262, 266–268 (WISH) ............................... 274, 278–279, 286–291
Transplant, subcapsular kidney ................ 262–263, 266–267 Whole-mount, TUNEL .................................. 278, 285–287

(Methods in Molecular Biology 1323) Rémy Bosselut, Melanie S. Vacchio (Eds.) - T-Cell Development - Methods and Protocols (2016, Humana Press) PDF

Uploaded by

Copyright:

Available Formats

(Methods in Molecular Biology 1323) Rémy Bosselut, Melanie S. Vacchio (Eds.) - T-Cell Development - Methods and Protocols (2016, Humana Press) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Methods in Molecular Biology 1323) Rémy Bosselut, Melanie S. Vacchio (Eds.) - T-Cell Development - Methods and Protocols (2016, Humana Press) PDF

Uploaded by

Copyright:

Available Formats

What are some of the critical concepts that studies of T cell development have brought insight into?

What are some of the critical concepts that studies of T cell development have brought insight into?

What are the main cell types involved in T cell development within the thymus?

What are the main cell types involved in T cell development within the thymus?

Methods in

Molecular Biology 1323

For further volumes:

Methods and Protocols

Rémy Bosselut and Melanie S. Vacchio

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Library of Congress Control Number: 201506151

Springer New York Heidelberg Dordrecht London

Printed on acid-free paper

Humana Press is a brand of Springer

Bethesda, MD, USA Rémy Bosselut

PART I BACKGROUND INFORMATION

PART II CORE APPROACHES AND STRATEGIES

PART III SPECIFIC TECHNIQUES

13 Reconstituted Thymus Organ Culture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

JOSÉ ALBEROLA-ILA • Immunobiology and Cancer Program, Oklahoma Medical

FRANCIS A. FLOMERFELT • Experimental Immunology and Transplantation Branch,

LEVI J. RUPP • Division of Cancer Pathobiology, Department of Pathology and Laboratory

200 Million Thymocytes and I: A Beginner’s Survival

1 T Cells and T Cell Antigen Recognition

T lymphocytes (T cells) are critical components of the adaptive

transcription factors refer to mouse T cell development. Issues

CD3 CD3 P Lck

Fig. 1 T cell receptor structure and signaling. (a) Schematic representation of

and intracellular signaling to substrates not physically linked to

2 An Overview of T Cell Development

2.1 A Brief The “objective” of intrathymic T cell development is twofold.

Thymus Peripheral lymphoid organs

Hematopoietic Committed Repertoire Mature Mature Effector

T lineage Antigen receptor Antigen

2.2 Early T Cell The earliest steps of intrathymic development, thought to be

In addition to Notch1 signals, early T cell development requires

2.3 Emergence Upon successful completion of TCRβ gene rearrangement, the

the question arises how developing thymocytes distinguish between

2.4 Selection of αβ T The selection and differentiation of DP thymocytes into mature

polymorphism of MHC molecules (i.e. the large number of allelic

MHC-peptide complexes that engage the TCR with high avid-

2.8 Differentiation The selection of conventional TCRαβ thymocytes is accompanied

or helper differentiation [62, 63]. Recent work has identified

Foxn1 that intrinsically disrupts thymic epithelial cell develop-

thymic function in the elderly or in patients whose immune system

3 Tracking Developing Thymocytes

Investigating T cell development typically involves a variety of

Myeloid, NK, TCRβ, TCRγδ

HSC DN1a DN1b DN2a DN2b DN3a

3.2 αβ T Cell Surface markers useful to analyze the development of conventional

β−selection TCRα rearrangement

Marker Pre-selection DP Signaled DP Selected SP Mature SP Comments

CD8 (as mentioned, the TCRhi CD69hi CD4+CD8int subset

This brief overview highlights the tremendous progress made over

Development of γδ T Cells, the Special-Force Soldiers

with the generation of memory [4]. As such, γδ T cells combine attri-

2 Antigen Recognition by γδ T Cells

While αβ T cells recognize and respond to proteolytically derived

3 Links Between Vγ Usage and Behavior