PRE-TCR AND TCRCXP SIGNALING DURIlVG T CELL DEVELOPMENT Timothy C. Groves A thesis subrnitted in conformity with the requirements for the degree of Ph.D. Graduate Department of Irnmunology University of Toronto a Copyright by Timothy C. Groves ( 1997)
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
PRE-TCR AND TCRCXP SIGNALING DURIlVG T CELL DEVELOPMENT
Timothy C. Groves
A thesis subrnitted in conformity with the requirements for the degree of Ph.D.
Graduate Department of Irnmunology University of Toronto
a Copyright by Timothy C. Groves ( 1997)
National Library I * m of Canada Bibliothèque nationale du Canada
Acquisitions and Acquisitions et Bibliographie Services services bibliographiques
395 Wellington Street 395. rue Wellington Ottawa ON K1A ON4 Ottawa ON K1 A ON4 Canada Canada
The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microfom, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/nlm, de
reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts 60m it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.
Pre-TCR and TCRap Signding During T ce11 Development
Ph.D. 1997
Timothy C. Groves
Graduate Department of Imrnunology
University of Toronto
Abstract:
During thymic development. the TCRp-containing pre-TCR cornpiex mediütes signaling at
the DN to DP transition, whereas TCRap-mediated MHC recognition triggers maturation at the
DP to SP transition. Various studies have examined the role of the Src PTKs Fyn and Lck in
pre-TCR/TCRap signaling. Although thymic development is normal in l~n -/- mice. the
jeneration of DP çells as well as the development of TCRap and TCRyG cells is partially
comprornised in Ick -/- mice. Since Fyn and Lck expression levels are comparable during T
ce11 development, Fyn may play a role in pre-TCR or TCR signaling in lck -/- mice. This study
assessed thyrnic development in lck -/-jjn -1- rnice and demonstrated thüt i t is completely
impaired with virtually al1 thymocytes arrested at the immature CD15+ DN stage. and no
detectable peripherd TCRaP+ and TCR@ cells. Additionally, expression of the activatedjjm
(TFF ) transgene cornpietrly restores the production of DP thymocytes and peripheral TCRyS
cells in lck -1- rnice. The TFF transgene also piirtially improves developrnent of transitionaYSP
thymocytes but fails to increüse the nurnber of peripheral TCRuP cells in k k 4- mice.
Collectively. these results suggest that Fyn rnediates development of DP thyrnocytes in lck -/-
mice, and transgenic expression of constitutively activatedfw can almost completely replace
Lck during the DP to SP transition. Another aspect of this thesis examined the TCRaP
signding requirements in the maturation of DP thymocytes. Using an in vitro mode1 to m e s s
the response of purified TCRI0 DP cells to immobilized TCR specific antibodies, the results
demonstrated that DP cells undergo multiple changes associated with positive selection in vivo.
Following TCR stimulation. DP thymocytes undergo increased expression of CD5 and Bcl-2.
a reduction in RAGl and pre-Ta gene expression. and a switch in Ick promoter usage.
Signais provided by CD4 and CD28 synergize with TCR signüls to mediate various aspects of
positive seleciion. However. TCR-rnediated signrils fail to trigger al1 hallmarks of DP ceIl
maturation becüuse clonal deletion. CD4lCD8 lineage cornmitment. and other aspects
associated with positive selection are not observed.
Acknowledgements:
I gratefully ackncwledge my supervisor Dr. Cynthia Guidos for providing me the
opportunity to work in her laboratory. As well, I thank Dr. Guidos for her guidance and
support during my graduate training. My supervisory cornmittee rnembers. Drs Iayne Danska,
Michael Julius, and Tania Watts, dso deserve praise for their support, advice and suggestions.
During my tïve years working in Dr. Guidos' laboratory. 1 had the opportunity to work
with m m y great people. I thank both past and present members (Trang Duong. Betty-JO
Edgell. and Dianne Holland) of the Guidos lab. In addition, 1 appreciated the support and
kindness of the Dünska Iüb. 1 ülso çxtend my gratitude to Gordon Cheng and Dr. Patty Smiley
for their tremendous support and friendship. As rvell. i greatly appreciated Gisrle Knowles
not only for her assistance in tlow cytometry and ceEl soning but also for her enthusiasrn.
Finally, I thank my family. Mom. Dad, Mike. Pam and Natalie. for their stradfast love.
support, and understanding. 1 am also extremety gratefùl to my family for their time and
patience.
Table of Contents:
Title page .................................................................................................... i
.................................................................................................... Abstract u
...................................................................................... Acknowledgements iv
......................................................................................... Table of Contents v ... ............................................................................................ List of Tables viu
List of Figures ............................................................................................ ix
.................................................................................... List of Abbreviations xi
.............................................................................................. Publications xiv
................................................................................... Chapter 1: Introduction 1
7 ........................................................................................... I . Overview - 2 . General Aspects of T Cell Development ........................................................ 3
........................................................................ A . Thymus architecture - 3
......................................................... B . Pre-thymic and thymic precursors 4
..................................................... C . T cell precursor-product relationships 5
.............. . D Development of TCR$i+ cells .. ............................................ 6
3 . The DN to DP Transition ......................................................................... 8
................................................................................ A . Role of TCRP 8
..................................................... B . Identification and function of pre-Ta 9
...................................................................... C . Role of CD3 proteins - 1 1
D . Role of other molecules in pre-TCR signaling ........................................... 13
............... 4 . Role of Src family Tyrosine Kinases in TCR Signaling and Development 13
.................................................................................... A . Overview -13
................................................................................. B . Role of Lck -15
................................................................................. C . Role of Fyn -18
................................................. 5 . Positive Selection at the DP to SP Transition 19
........................................................................ A . Historical overview -19
............................................................ Chapter IV: Discussion ............. .... 106
1 . The DN to DP Transition ...................................................................... 107
...................... A . Lck and Fyn in pre-TCR signaling at the DN to DP transition 107
B . Role of Lck and Fyn in extrathymic development ................................... 110
...................................................................... 2 . The DP to SP Transition 111
................................ A . The response of DP thymocytes to TCR engagement 111
B . The significance of thymic stroma1 cells in positive selection ........................ 113
.......................................................... . C CD4/CD8 lineage cornmitment 114
................... . D Positive selection: a multi-step process of thymocyte maturation 116
. E Modulation of TCR-mediated signaling in DP thymocytes .......................... 118
............................ . F Significance of thymic strornal cells in negative selection 120
.................................................................. i ) B7ICD28 interactions 121
.............................................................. ii) gp39lCD40 interactions 122
..................................... i i i) LFA- MCAM- 1 and CDDLFA-3 interactions 122
.................................................................... iv) CD30 interactions 133
3 . Conclusions ..................................................................................... 123
................................................................................. Chapter V: References 125
vii
List of Tables:
Chapter II:
Table II- 1 Failure of TCR engagement to induce denth of DP thymocytes ................. -49
Table 11-2 Anti-TCRP induces death of CD4 and CD8 SP thymocytes irt viirtl ............. 52
............................... Table 11-3 Xnti-CD3e induces death of DP thymocytes iri v i iw 53
Table I I4 Surface phrnotype of DP thy mocytes after TCR engagement in ipitro ........... 60
Chapter III:
Table III- I Lymph node T ceil subsets in mice lacking Lck, Fyn or both .................... 82
Table 111-2 Effect of TFF trrinsgene on peripheral T ce11 number in lck -1- mice ........... 94
List of Figures:
Chapter II:
Figure II- 1
Figure II-ZA
Figure II-2B
Figure II-2C
Figure 11-3
Figure II--CA
Figure II4B
Figure II-5A
Figure II-5B
Figure II-jC
Figure 11-6
Chapter III:
Figure III- 1 A
Figure III- 1 B
DP thyrnocytes decreasr expression of both CD4 and CD8 in response to TCR
................................................................................. ligation -48
TCR-induced down-regulation of coreceptor expression by DP thymocytes is
................................... not correlated with the induction of üpoptosis.. .55
TCR-induced down-regulation of coeceptor expression by DP thymocytes is
................................... not correlated with the induction of apoptosis.. -57
........................... Developmentd potential of TCR-stimulated DP blasts -58
Purified blast and PM DP thymocytes express CD5 and CD69 in response to
........................................................................... TCR ligation .6 1
CDJ-mediüted signals enhÿnce TCR-induccd expression of CD5 or Bcl-7 in DP
............................................................................ thy rnoc ytes ..62
CDZ8-mediated signals enhance TCR-induced expression of CD5 in DP
........................................................................... thymocytes.. -64
DP thy mocytes terminate expression of M G 1 and pre-Ta in response to TCR
................................................................................. ligütion -66
DP thymoc ytes terminate expression of RAGl and pre-Tu in response to TCR
................................................................................. ligation -66
... DP thymocytes terminate expression or pre-Tu in response to TCR ligation 67
TCR ligation alters the ratio of lck type I and type I I transcripts in DP
.......................................................................... t hymoc ytes.. ..68
Abrogation of T ceIl development in lck - l -~vn -/- mice. Absence of TCR@+
..................... and TCR/S+T cells in the periphery of lck z 4- mice 83
Abrogation of T ce11 developrnent in lck - / - f y i 4- mice. Early block in
................................... thvmocvte develooment in lck -1- fyn -1- mice .84
Figure 111-2 Western blot analysis of Fyn protein levels in thymocytes . ... . . .. . ... . . . . .. . . .. -87
Figure III-3A Effect of the TFF transgene on T ce11 development in RAGl -/- rnice.. . . . . . . . -88
Figure 111-38 Effect of the TFF transgene on T ce11 development in RAGl 4- and lck -1-
mice.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -89
Figure III-JA Effect of the TFF transgene on the DP to SP transition in ick -1- mice.. . . . . . .9 1
Figure III-4B Effect of the TFF trünsgene on the DP to SP transition in lck -1- mice.. . . . . . .92
Figure 111-5 Effect of TFF transgene on the frequency of TCRp+ and CD4+ splenic cells.. . .
....... .. . ...... ...................... . . . . . . . . . . ............... . . 95
Figure IIMA Effect of Lck drficiency on TCR-mrdiated signal transduction in DP
t hy rnocytes. TCFUCD4-induced protein tyrosine phosphoryliit ion in Ick +/+
and l c k -1- DP thymocytes ............................................................ 96
Figure IIMB Effect of Lck deficirncy on KR-mediated signal transduction in DP
thymocytes. RT-PCR ünalysis of M G - I and /3-<lctin transcripts in lck +If and
lck -/- DP thymocytes ................................................................. 98
Figure III-6C Effect of Lck deficiency on TCR-mediatrd signal transduction in DP
thymocytes. Basal and TCR-induced CD5 and CD69 expression by normal vs
l c k -'- DP thymocytes ................................................................. 99
List of Abbreviations:
P2m: beta 2 micro_slobulin
B6: C57BU6
C: constant
CGy: centigray
Cs: cesium
D: diversity
DN: double negritive
DNA: deoxyribonucleic acid
dNTP: dinucleotide triphosphate
DP: double positive
DTT: dithiothreitol
ECL: enhmced cherniluminescence
EDTA: ethylenediaminetetrri-acetic acid
FCS: fetal calf serum
FITC: tluorescein isothiocy rinate
HBSS: Hank's but'fered salt solution
HRP: horseradish peroxidase
HPRT: hypoxanthine phosphoribosyltransferüse
HSA: heat stable antigen
Ig : immunoglobulin
ITAM: intracellular tyrosine activation motif
J: junctional
kDa: kiloddton
LCMV: lymphocytic choriomeningitis virus
LFA: lymphocyte f~inction-associatrd antigen
MAPK: mitogen-activated protein kinase
MF?: mean lluorescence intensity
MHC: major histocompatability complex
Mls: minor lymphocyte-stimulating antigen
OVA: ovdburnin
PAGE: polyücrylümide gel electrophoresis
PBS: potassium buffered saline
PCD: prograrnmed ce11 decith
PE: phycoerythrin
PI: propidium iodide
PM: post-mitotic
pre-Tu: pre-TCRa chain
PTK: protein tyrosine kinase
RAG: recombinrt.le-activating gene
RNA: ribonucleic acid
&VASE: ribonuclerise
RT-PCR: reverse transcription polymerase chüin reriction
scid : severe combined irnmunodetïcienc
SDS: sodium docedy l sulfate
SP: sin@ positive
SH: src homoiogy
TAP: peptide transporter associated with antigen processing
TCR: T ce11 receptor
Tdt: terminal deoxynucleotide transferase
TFF: constitutively active jjri trrinsgene
TUNEL: terminal Jeoxynucleotidyl transferase-mediated d-UTP-biotin nick end labeling
V(D)J: variable (diversity) junctional
V: variable
xii
ZAP-70: zeta-associated tyrosine kinase
. *.
X l l l
Publications:
Chapter II:
Groves T.. M. Parsons. N.G. Miyamoto. and C.I. Guidos. 1997. TCR engagement of C D ~ '
CD8' thymocytes in vitro induces rarly aspects of positive selection. but not apoptosis. J.
Irnrnurzoi. 158: 65-75.
Chapter III:
Groves T.. P. Smiley. 1M.P. Cooke. K. Forbush. R.M. Perlmutrer. C.J. Guidos. 1996. Fyn
c m partidly substitute for Lck in T lymphocyte development. It>inrlo~i~ 5 : 4 17-428.
X I V
CHAPTER 1 tNTRODUCTION
1. Overview
T lymphocytes play a major role in protecting the host agÿinst f'oreign antigens by virtue of
their ability to recognize antigen through diverse. clonally distributed surface T ce11 receptors
(TCR). Most T cells express TCRaP heterodimers but it minor class expresses TCRyG
heterodimers. TCRap cells recognize smdl antigenic peptides bound to cfass I or class II major
histocornpatability complex (MHC) molecules. Two TCRuP T ce11 lineages can be
distinguished by phenotype. function. and specificity: CD4f helper and CD8+ cytotoxic T
cells. The rnajority of helper T cells express the MHC class II coreceptor. CD4. and TCR
specific for peptidr/MHC class II ligands, whereas rnost cytotoxic T cells express the MHC
class I coreceptor. CD& and TCR spccitic for peptide/MHC c las I ligands. The generation of
functionally mature CD4+ and CD8+ T cells o c c m in two phases which depend upon TCRP
and TCRap expression, respectively. During the tirs< phase. CD4/CD8 double negittive or
(DN) thymocytes that have functionally rearranged TCRp are selected to undergo clonal
expansion and differentiation into CD4/CD8 double positive (DP) thymocytes. This selection
process is mediated by the pre-TCR complex composed of pre-Ta. TCRP and CD3 proteins.
Subsequently. DP thymocytes begin to rearrange TCRn genes. and TCRup+ celis then
undergo a selection process biised on TCRap-rnediated inreractions with thyrnic MHC
molecules. This second selection process. re ferred to as positive selection. ge nerates self-
iMHC restrictrd CD4 or CD8 sinsir positive (SP) T cclls that preferentially recognize foreign
peptides associiiied with self-MHC molecules. During the DP to SP transition. autoreactive T
cells that respond to self peptides plus self-MHC proteins are negatively selected by clona1
deletion or tùnctional inactivation. Consequentiy, positive and negative selection events in the
thymus result in the genention of highly diverse but self-toleriint. self-MHC-restricted TCRaP
cells. One focus of my thesis examines the importance of the Src türnily protein tyrosine
kinases (PTKs) Fyn and Lck in pre-TCR-mediated thymocyte selection iii vivo . The second
aspect of my thesis assesses the capacity of DP thymocytes to undergo TCRap-mediated
positive and negative selection in response to in virm TCR engagement. Before describing my
results, 1 will first review important tindings in the ÿrea of T ce11 development.
2. General Aspects of T Ce11 Development
A. Thymus architecture
The thymus contains a frimewcrk of non-lymphoid stroma1 cells interspersed with
immature and mature T cells. Histologically, the thymus is composed of two major regions,
the cortex and the medulla. and is encapsulated by connective tissue (reviewed in Boyd et al.,
1993). T ce11 precursors are tliought to enter the thymus either at the subcapsule during fetal
Me or the cortico-medullary junction of the adult thymus (Ceredig and Schreyer, 1984;
Miyasaka et al., 1990). Following entry, the precursors rapidly proceed to the cortical
subcapsular region and then rnigrate frorn the cortex to the medulla as they differentiate into
mature cells (Kyrwski. 1087: Penit and Vasseur, 1988). Thus. the thymic cortex largely
contains immature DN and DP thyrnocytes whereas the medulla is a site of mature SP
thymocytes.
The thymic stroma includes epithelial and hrmatopoictically-derived cells. Thymic
epithelium is suggested to derive from the third and/or founh pharyngeal pouch endoderm and
ectoderm from the third brachial cleft and neural crest (reviewed in Boyd et al.. 1993). The
thymic epithelial ceIl cornpartment is heterogeneous. and there are numerous differences
between cortical and medullary epithelial cells. Cortical epithelial cells express high levels of
MHC class II molecules and low amounts of MHC class 1 molecules, whereas medullary
epithelial cells express iMHC class I but have variable expression of MHC class 11 molecules
(Wekerle et al., 1980; Surh et al., 1992). Furthermore. non-polymorphic MHC clriss II 1-0
molecules are highly expressed on bone-marrow-derived cells and on medullary but not on
cortical epithelium (Surh et al.. 1992). Thymic stroma1 crlls of hematopoietic origin include
macrophages and dendritic cells. Thymic macrophages are rarely found in the medulla but are
prominent in the cortex and at the cortico-medullary junction and mostly lack expression of
MHC class II molecules (van Ewijk et al., 1980). In contrast. dendritic cells express high
levels of MHC class 11 molecules and are evident throughout the thymus, especially at the
cortico-medullary junction (van Ewijk et al.. 1980). Thus. the thymic stroma consists of both
epithelial and bone rnarrow-derived cells. which interact with T ce11 precursors to promoie their
development into CD4+ helper and CD8+ cytotoxic T cells.
B. Pre-thymic and thymic precursors
T ce11 progenitors arise in the fetd livcr or adult bone marrow and migrüte to the thymus. It
is not clear whether cornmitment to the T ceil linerige occurs in the fetal liver or rtdult bone
marrow. or occurs as a consequence of thyrnic colonization (reviewed in Rodewald, 1995).
Recent results suggest the former possibility because the ktal blood contain precursors that are
T [inerige committed but lack progenitor potential for the B cell. myeloid and erthyroid lineages
(Rodewald et ai.. 1994). These cells have ü Thy- l + c-kitIo CD3- phenotype. and they
reconstitute the T ce11 lineage after both intrüthymic and intrüvenous trans fers. However, the
Thy- l + c-kitIo CD3- cells fail to colonize the bone marrow and lack prccursor activity for other
hematopoietic lineüges. These rcsults thus strongly suggest that the fetal blood contains
precursors that are committed to the T ce11 lineage and thus T ceil cornmitment can precede
thymic colonization. In contrast. the identification of lymphoid-restrictrd precursors in the
adult bone rnarrow remains to be resolved. A precursor population was identified in adult
thymus that can give rise to T cells. B cells. NK cells. and dendritic çells (Wu et al.. 1991;
Matsuzaki et al.. 1993). These cells frature low expression of CD4 and high expression of c-
kit (Wu et al., 199 l : Matsuzaki et d., 1993). A corresponding population is also present in the
fetal thymus. but they have a poor capacity to reconstitute the T ce11 lineage (Antica et al..
1993). Thus. both extrathymic and intrithymic precursors for T crll development have been
c haracterized.
C. T cell precursor-product relationships
Both TCR@ and TCR@ cells can develop from immature DN thymocytes. DN
thymocytes contain the eürliest precursor T cells. as lirst dcmonstrated by intravenous and
intrathymic transfer studies (Fowlkes et al.. 1985: Crispe et al.. 1087: Shimonbevitz et al..
1987: Guidos et al.. 1989a). DN thymocytes are developmentally and phenotypically
heterogenous, and can be subdivided into distinct maturational stages based upon CD44 and
CD25 expression. Analysis of the reconstitution potential of the DN subsets following
intrathy rnic trans fer or in lymphocyte-depleted fetal thy rnic organ cultures ( FTOCs; Crispe et
al.. 1987; Shimonkevitz et al.. 1987: Godfrey et al.. 1993) indicütes that development of DN
thymocytes proceeds via an ordered srquence of maturational stages: CD44+CD25-;
CD44+CD25+ ; C D 4 4 CD25f; and CD44- C D W . Commitrnent to either the TCR./6 or
TCRup Lineage is thought to occur after the C D 4 4 CD25+ stage since these crlls were the most
differentiated DN subset to reconstitute both lineiiges in lymphocyte-depleted FTOCs (Godfrey
et al.. 1993). Subsequent analysis from the intrathymic trünsfer system and in vitro studies
indicatrd that TCR- DN thyrnocytes mature into TCRI<' DP cells through a TCRTD4- CD8+
intermediate stage (Nikolic-Zugic and Bevan. 1988; Guidos et al.. 1989b; Nikolic-Zugic et al..
1989).
DP thymocytes can bc subdividcd into cycling blasts. rvhich represent approximütely 10-
15% of total DP cells. and small post-mitotic (PM) cells (Shortman et al., 1990). Using the
inirathymic transfer system. only large DP thyrnocytes demonstrated precursor activity for
mature thymocytes (Guidos ct al.. 1989b). Later studies demonstrated that DP PM thymocytes
also have precursor potential for mature SP cells but were less efficient than DP blasts in
generating mature progeny (Lundberg and Shortman. 1994). Further studies showed that
TCRIO DP blasts could develop into ~CRmed CDJf CD810 or TCRmcd ~ ~ 4 1 0 CD8+ transitional
intermediates and then T C R ~ ~ CD4 SP or T C R ~ ~ CD8 SP cells. respectively (Guidos et al.,
I W O ; Guidos and Weissman. 1993). Thus, T C R ~ ~ ~ CD4+ C ~ 8 1 0 cells were suggested to be
transitional intermediates between TCRI" DP blasts and T C R ~ ~ CD4 SP cells. Conversely,
maturation of DP cells into TCRhi CD8 SP cells was proposed to procred via T C R ~ C ~
CD8f transitional intermediates. However. more recent studies suggest that ~ C ~ m e d CD4+
C D ~ ~ O transitional cells have precursor üctivity not only for T C R ~ ~ CD4 SP cells but also for
T C R ~ ~ CD8 SP cells (Lundberg et al.. 1995: Suzuki et al.. 1995). Thus, the seneration of
mature S P thy moc ytes may proceed via several intermediate stages from DN thy moc ytes.
During fetal ontogeny. thyrnic development is accompanied by TCR rearrangements and
expression of the variable (V) . junctional (J). diversity (D), and constant (C) segments rit the
TCRa, P. 6 and y genes. The transition from the CDU+CD?Sf to CD41-CD25+ stage is
associated with rearrüngernents at the TCRP. y, and 8 loci. While the TCRp and S loci undergo
DJ rearrangrments at day 1.1 followed by V(D)J rearrangemrnts at day 15- 16 of gestation. the
TCRy locus undergoes VJy rearrangements (Fowlkes and Pardoll. 1988). The onset in TCR
rearrangernents is accompanied by loss of CD44 expression and high level expression of the
recombinase-iictivatinj genes. RAG 1 and RAG2 (Godfrey et al.. 1993: Wilson et ai.. 1994).
As a result. the proportion of TCRy6f cells is maximal at day 17 and then declines until birth.
During the transition from the CD4-V CD25+ to the DP stage. the TCRU locus undergoes VJa
rearrangements at day 17 of gestation and this correlates with high RAGl and RAG2
expression (Wilson et al.. 1994). Consequently, the tiequency of TCRap+ cells initially
begins at day 1 7 and is maximal at binh. In summary, TCRap and TCRy6 rearrangement and
expression demonstrate different kinetics diiring fetal thymic development.
D. Development of TCRy6+ cells
During fetal ontogeny. successive waves of TCR-{8+ cell subsrts are generated in the
thymus prior to development of mature TCRap+cells (Pardoil et al.. 1987). Analysis of TCR
rearrangements in TCR./Gf cells indicates selective usage of particular V, J. D and C segments
at the TCR 7 and 6 locus. Distinct subsets of K R $ + cells that differ in V gene usage appear
in a series of overlapping waves (Havnin and Allison. 1988). TCRy6+ cells that appear in the
first and second waves express Vy3 and Vy4 chains and have invariant junctions at the third
complementary drtermining region. The Vy3 and V-fl chains associate with ü V61 chah also
with invariant junctions. By day 18 of drvelopment. the Vy3+ and Vy4+ thymocytes
disappear and are replaced by cells expressing Vy2 or Vy1 paired with diverse VS chains. The
various TCRySf subsets migrate to panicular peripheral sites. In normal mice. Vy3+ cells are
prominent in the skin. V-fi+ cells in the tongue and fernale reproductive tract. and Vylf and
V$+ cells in the lymphoid tissues and the blood. Thus. the first T crlls to be produced in the
thymus express TCRyS and migrate to epidermal and mucosd epithelia.
Developmen t of TCRySf cells has been extensively examincd in both TCRyS transgenic
and non-trrinsgenic models. Initially. drvelopment of TC%{&+ crlls was suggested to be
iMHC-dependent. based upon studies involving two different TCRy6 transgenic strains (G8
and KN6) which generate cells with specificity for a ligand encodrd by the non-classical MHC
class Ib gene. T12 (Ito et al.. 1990: Wells et al.. 1991: Haas et al.. 1993; Wells et al.. 1993).
However. another study demonstrated thüt development of non-transgenic TCR$+ cells was
normal in MHC citrss II - and B3 microgloblin (p2m)-det'icient mice (Correa et al., 1993).
Additionally. a recent study found that functionally mature G8 TCR-{S+ crlls could develop in
p2m -detïcient mice (Schweighoffer and Fowlkes. 1996). suggesting that P3m-containing
MHC class I molecules are not reqiiired for the developrnent of TCR@+ cells. Thus. MHC
molecules are not necessay for TCRyS ceIl development.
Multiple studies have iissessed the requirernents for antigen recognition by ?'CR@+ cells.
Results from previous reports suggest that antigen recognition by MHC class I I and class Ib
specific TCR./6+ cells does not require antigen processing and presentation of peptides bound
to MHC molecules (Schild et al.. 1994; Weintraub et al., 1994). The non-MHC-encoded CD 1
molecules, which consist of monomorphic p2m-associated glycoproteins. c m serve as tÿrgets
for TCRy6+ cells (Faure et al.. 1990). Additionally, other studies demonstrüted that TCR~G+
cells can ülso recognize various antigens in the absence of rintigen presenting cells, including
mycobacterial peptides. heat shock proteins (Born et al.. 1990; Rajasekar et al.. 1 W O ) , and
small non-peptide ligands (Tanaka et al.. 1994). Thus. these results suggest that antigen
recognition by TCRyS+ cells is not restriçted by peptiddMHC complexes.
3. The DN to DP Transition
A. Role of TCRp
An important role for TCRB chain in the DN to DP transition was reveded by studies of
mice bearing targeted mutations in the RAG and TCR loci. These studies involved mice that
were dekctive in V(D)J recornbination, such as those with mutations in RAGI or RAG2
(Mombrierts et al.. 1993b; Shinkai et al., 1992). or mice with the severe combined
immunodeficient (scid ) mutation (Bosrna et al.. 1983). In these mice. thymic development is
mested at the C D W CD25f DN stage, and thymic celluliirity is diminished by almost 100-
fold. However. introduction of a productively rearranged TCRp trrinsgene in these mice
induces the developrnent of DP cells and restores thymic cellularity. suggesting that TCRp
rearrangement and expression are necessary for the di fferent iation and expansion of DP
thymocytes (Kishi et al.. 199 1: Mombaens et 31.. 1993: Shinkai et al.. 1993). In çontrast to
R4Gl or R4GZ -deficient ( M G -1-) or scid mice. TCRp-detïcient (TCRP-1-) mice have thyrni
with ripproximatrly IO-fold fewer cells than normal mice. and they contain a smnli number of
DP cells (~Mombaerts et al., 1991a). However, in mice bearing nul1 mutations in both the
TCRp and 6 loci. DP thymocytes are not generated. resulting in a phenotype comparable to
M G -/- and 3rd mice (Itahiiro et al., 1993). One potential explanütion for these results is that
productive reürrangements at the 6 locus may infrequently or inefficiently prornote the
maturation of DN thymocytes. In contmt to f c ~ p - 1 - mice. both the thymic crllularity and the
number of DP cells is normal in ~ c ~ n - 1 - mice (Mornbüerts et al.. 1992a: Philpott et al., 1992).
Thus. although most DP cells express a complete TCRap. the TCRa chain is not essential for
their developrnent.
The role of the different domains of the TCRP chain in T cell development have been
addressed in TCR traiisgenic mice expressing mutant versions of the TCRp chain. In TCR
transgenic mice bearing a TCRP chah lacking the variable domain. the generation of DP
thymocytes is not perturbed. suggesting that the variable domain is not necessary for the DN to
DP transition (Ossendorp et al., 1992). More recently. a study demonstrated that the
extracellular constant (C) and trmsrnembrane domains of the TCRB chain rire both necessary in
regulating the number of DP thymocytes (Jacobs et al.. 1996). Collectively. these results
suggesi that the TCRP chain regulates the nurnber of DP thymocytes but is not essential for
their generation.
Examination of TCRp expression during thymic development also confimis the importance
of productive TCRP chain remmgements in mediatins the developrnent of DP thymocytes. In
DP thyrnocytes derived frorn KI3a-I- micr. 7 5 8 of V(D)J rearrangernents are in frame at the
TCRp locus. a rate that is close CO the 7 1.4% cxpected if selection occurs versus the 33%
predicted if no selection takes place (Mallick et al.. 1993). In a sirnilar andysis of thymocytes
from normal mice. the lrss mature CDU- C D 3 + DN subset contains rearanged TCRP genes
that are out-of-frünie. whereas approximütely 7 5 6 of TCRP rearrangements in CD@- CD25-
DN cells are productive (Dudley et al.. 1994). These results suggest that selection for
productive TCRP rearrangements regulates dttveloprnrnt at the CD44 C D 3 f DN to DP
transition. Productive TCRP gene rearrangements also nrgütively regulate TCRP
rearrangements by a mechanisrn known as allelic exclusion. Expression of a functionally
rearranged TCRlj iransgenr suppresses rrlirrangemrnt of endogenous TCRP (üematsu et al.,
1988), thereby excludin_j a ce11 frorn expressing two different TCRp çhains. In surnmary.
these results suggest that expression of a functional TCRp chah promotes clonal expansion.
TCRP allelic exclusion. and maturation of DN thymocytes to the DP stage.
B. Identification and function of pre-Tcx
The above genetic studies suggested that ii TCRP-contüining pre-TCR complex controls the
DN to DP transition. Biochernical evidence for such a complex was first derived from a scid
thymic ce11 line (SCB.29) rransfected with a rearranged TCRP gene (Groettrup et al.. 1992).
Initially, the TCRp chain was suspected to form TCRP homodimers at the ce11 surface. but
funher analysis found that TCRP forms a heterodirner with a 33 kDa protein. referred to as
gp33 or pre-TCRa (pre-Tu; Groettmp et al.. 1993). The pre-Ta gene encodes a type I
transmembrane protein (Saint-Ruf et al., 1994) with moderate homology ( 2 0 - 3 % ) to the C
domains of the imm~inoglobulin ( Ig) siipergene hmi l y. The cxtracrllular dornain of pre-Ta
pairs with the cxtrücellulx C dornain of the TCRP chain. possibly ienving the TCRp V-domain
available for pairing with another protein. In pre-B cells. the pre-B ce11 receptor consists of an
Ig heavy chain (p) associated with 15 and VpreB proteins. which pair with the C and V
domüins of Ig p, respectively (von Boehmer et al.. 1993). Thus. the pre-Ta chain in the pre-
TCR receptor müy be analogous to the À5 protein in the pre-B ce11 receptor. However, unlike
h5. the pre-Ta chain has a transmembrane and a cytopIasmic domain, featuring two potential
phosphorylation sites for protein kinase C and a possible SH3 binding region (Saint-Ruf et al.,
1994). The role of the pre-Tucytoplasmic domüin in signal transduction remains to be
identifîed.
Expression o l the pre-Tu gene differs signi ficüntly from the TCRP gene. The pre-Ta chah
is poorly expressed in CD.IJ+CDlj- DN cells but undergoes a progressive increase in
expression in the more diffenntilited DN subsrts. particularly the C D 4 4 C D B f and
CD44+CD25+ cells (Saint-Ruf et al.. 1994). Expression of pre-Tu is weak in DP cells and
not drtectable in mature SP thymocytes. Another feüture of the pre-Ta grne is that. unlike the
TCRP gene. it does not undergo rearrangrment because pre-Tu transcripts were observed in
remangement-defective R4GZ -'- mice (Saint-Ruf et al.. 1994). Thus, expression of the pre-
Ta gene does not require rearrangement and is predorninüntly expressed in immature DN
thy mocytes.
Analysis of mice lacking the pre-Ta gene indicates that it plays a role in eürly T ce11
development (Fehling et al.. 1995). Pre-Ta -1- mice develop 30-fold fewer than normal DP
thymocytes. and very few mature thyrnocytes and peripheral T cells. The pre-Ta chah thus
regulates thymic cellularity but is not essential for the deveiopment of DP thymocytes.
However. development of TCR@ crlls is normal in pre-Ta -/- mice. suggrsting that pTa
gene expression is only important in TCRaP. not TCR./S. ce11 development.
C. Role of CD3 proteins
In mature T crlls. the CD3 (y, S. E ) and TCRC proteins regulrite assernbly and transport of
the TCRaP complex (reviewed in Ashwell and Klüusner. 1990). Another function for CD3
and TCRS proteins is to transduce intracellular signals from the CD3/TCR complex in response
to stimulation. During signal transduction. the CD3 and TCRC subunits al1 undergo
phosphorylation on two tyrosine residues present in the immune receptor tyrosine-based
activation motif (ITAM; Reth, 1989; Weiss. 1993). This motif consists of ri pair of tyrosine-
X-X-lrucine/isoleucines sequencrs (where X corresponds to any variable residue) separated by
seven to eight variable residues. This motif is present as threr copies in the TCRi subunit and
as one copy in the othrr CD3 chains. When chimeric molecules made from the extracellular
domains of CD4 or CD8 and the intracellular domain of TCR j are crosslinked. a full spectrum
of TCR signaling rvrnts are induced. suggesting that ITAlM motifs are sufficient for mediating
these rvents (Irving and Weiss. 199 1 ; Romeo and Seed. 1 9 1 ). Howrver. TCRT chimeras
bearing only one of the three ITAM motifs results in reduced activation versus those expressing
al1 the motifs (Rorneo et al.. 1992; Wegener et al.. 1993). Furthermore. mutation of tyrosine
residues in the iT.4M motif of CD25lCDk and CD8ITCRT chimeras abolishes the signaling
properties of rhr chimeraï ( Letourneur and Klausner. 1992: Romro r i al.. 1992). Thus. the
ITAii motif in the CD3 subunits hüs an important function in TCRKD3-mediated signding.
Expression of the CD3 proteins is developmentally rrgulated. The CD3 subunits are first
detected by intracellular staining in prethymic precursors in the human fetal liver (Lanier et al.,
1992) and murine tëtal thymocytss as early as days 13 to 14 of gestation (Levelt et al.. 1993b).
and TCRC are expressed intracellularly at days 13 and 14. respectively. prior to
intrncellular expression of TCRP at day 15 of gestation. Since the antibody used to detect
c D 3 ~ can only detect CD& in the presence of CD31 or CD36. this suggests that CD3d6 or
CD3d-t dimers are rxpressed iit this devrlopmrntal stage (Levelt et al., 1 9 9 3 ~ ) .
Irnrnunoprecipitation studies showed that CD3 proteins cün also associate with TCRP in the
absence of TCRa (Shinkai et al., 1993; Iacobs et al.. 1994). These results suggest that all the
CD3 subunits are physically associated with the pre-TCR cornplex.
In vitro studies have demonstrated a rolr for C D ~ E in pre-TCR signaling in DN
thyrnocytes. Administration of C D ~ E specific antibodies to organ cultures of normal fetal
thymi at d;iy 14, which is the timepoint before completion of TCRp rearrangrment. induces the
maturation of DP thymocytes (Levelt et al.. 1993a). Funherrnore. treatrnent of M G 2 mice
(Jacobs et al., 1994; Shinkai and Alt. 1994) and organ cultures ofrither TCR-p-1- or of RAGI
-le fetal thymi (Levelt et al.. 1993c) with C D ~ E specific antibodirs promotrs the drvelopment
and expansion of DP thymocytes. These results suggest that the signaling function of C D 3 on
the surface of DI\; thymocytes exists prior to and independent of TCRP expression.
Results from iti i ~ i v o and in vitm studies supporr the idea that and TCRS participate
in early thymocytc maturation. In TCR ;-1- mice, thyrnic cellularity is reduced 10- to 20-fold
with impaired development of DP thymocytes. implicating a role for TCRL in the expansion
during the DN to DP transition (Liu et al., 1993; Love et al., 1993: iMalissen et al.. 1993; Ohno
et al.. 1993). However. the developmental defects in TCR je'- mice rnay not reîlect a lack of
TCR signa@. rather. dekctive TCR expression (Shores et al., 1994). Consistent with this.
in viw üdniinistratlon of C D ~ E specific üntibodies induces DN thymocytes in £?.AG2 TCR 5
-/- mice to mature to the DP ceil stage. sugesting tliat TCRL is not cssential for signaling at the
DN to DP transition (Levelt et al.. 1995). In contrast to f~~6-l- mice. the DN to DP transition
is virtuülly arrested in CD3y. CD3& and C D ~ P deficient mice (Malissen et al.. 1995),
suggesting an essential rolr for CD3~16 and/or CD3dy dimers in early T çrll maturation.
However. results obtained from transgenic mice expressing CD25/CD3& or CD25/TCR<
chimeric genes suggested that signaling via rither CD% or TCRC is sufficient for thymocyte
maturation to the DP stage (Shinkai et al.. 1995). Collectively. these results suggest that C D ~ E
is necessary for pre-TCR signaling, whereas multiple TCRG ITAMs likely amplify signaling
via the pre-TCR complex.
D. RoIe of other rnolecufes in pre-TCR signaling
Various studies have rxamined the role of additional molrcules in pre-TCR signaling at the
DN to DP transition. A recent study examined the importance of CD27KD70 interactions in
pre-TCR signaling (Gravestein et al.. 1906) and suggested that these interactions may
cooperate in pre-TCR signaling. Additionally. results of another study indirectly suggested
that the thymic stroma1 antigen CD81 may be a potential ligand for the pre-TCR complex
(Boismenu et al.. 1996). CD8 1, also known as TAPA-1. is a member of the tetrapanin, or
transrnembrane 4 integral membrane protein hmily (Oren et al.. 1990). Administration of
CD8 1 specific antibody to FïOCs signitïcantly inhibited the generation of DP thymocytes but
had no effects on TCRy6 T ce11 developrnent. Furthermore. day 14 DN fetal thymocytes
differentiate into DP thyrnocytes in reaggreption cultures only in the presence of CD81-
transfected tibroblasts. These findings suggcst that the pre-TCR complex and CD8 1 may both
be involved in the DN to DP transition and participate in the developrnent of TCR@ cells. and
not TCRyS+ cells. However. a recrnt report foound thüt T ceIl maturation is normal in CD81 -1-
mice. suggesting thrit CD8 1 mity not be essentiril during the DN to DP transition (Maecker and
Lrvy, 1997). T\Jeverthrless. these results indicare that CD27 and CD81 may have a potential
roie in pre-TCR signaling.
1. Role of Src Family Tyrosine Kinases in TCR Signaling and Development
A. Overview
The earliest detectable signal transduction event initiated through the TCR is the activation
of protein tyrosine kinases (PTKs), resulting in the phosphorylrition of various proteins. in T
cells, two Src fümily PTKs. Lck and Fyn. mediate TCRKD3 signaling. Fyn has two
isoforms, Fyn (B) in the brain and Fyn (T) in hernatopoietic cells, that result from mutually
exclusive splicing offvn exons 7A and 7B (Coolie and Perlmutter. 1989). Following TCR
engagement. Lck and Fyn initiate phosphorylation of ITAM motifs in the cytoplasmic domains
of CD3ITCRi (Sarosi et al.. 1992; Straus and Weiss. 1992). This results in the recruitment of
the ZAP-70ISyk farnily of PTKs to CD3TTCRL ITAM motifs. an event mediated by two
tandemly arranged SH? domains of ZAP-70 or Syk interacting with two phosphotyrosines in
the ITAM motif of the CD3lTCRC chain (Chan et al., 1992: Wange et al.. 1992; Wange et al.,
1993; Chan et al.. L994b). Thus, TCR and coreceptor CO-engagement leads to the activation of
non-receptor Src PTKs. Lck and Fyn, which results in the recruitment of another PTK. ZAP-
70.
Lck and Fyn are cytoplasmic non-receptor PTKs with an unique riniino (N) terminus, Src
homology (SH) regions 7 and 3 (SH2 and SH3). and a carboxyl terminal kinase domain. At
the amino terminus. myristoylation at glycine position 2 prrmits Lck and Fyn to interact with
the inner Iratlet of the plasma membrane. Fyn also associates with CD3 proteins. iilbeit with
very low stoichiometry as measured by immunoprecipitation studics (Sarosi et al.. 1997). In
contrat. the amino terminus of Lck contains cysteinrs 20 and 23 that eftïcirntly interact with
CD4 and CD8 iit much higher stoichiornetries (Shaw et al.. 1989: Turner et al.. 1990).
Additionally. Lck interacts with other surface proteins such as CD45 (Schraven et al., 1991:
Rothstein et al.. 1993). interleukin-2 receptor P (Hritakeyama et al., 199 l ) , CD2 (Danielian et
al., 199 1: Danielian et al.. 1992). and the glycosylphosphotidylinositol-linked proteins, Thy-1
and Ly-6 (Stefanovü et al.. 199 1 ). Activation of Lck and Fyn Irads to autophosphorylation at
tyrosine 394 and tyrosine 117. respectively. At the carboxyl terniinus. Lck and Fyn are
nrgatively regulated by Csk-rnediatrd phosphorylation iit tyrosine 505 and tyrosine 528.
respectively (Bergman et al.. 1992) (Chow et al.. 1993). These C-terminal tyrosines are
dephosphory lated by the receptor tyrosine phosphatase CD45 (Hurlry et al., 1989: Mustelin et
al., 1989; Ostergaard et al.. 1989: Shiroo et al., 1992: McFarland et al.. 1993: Sieh et al.,
1993). Thus. TCR-mediated signal transduction is regulated by Lck and Fyn kinase activity,
which in tum retlects a balance between Csk and CD45 activities in T cells.
B. Role of Lck
In mature T çells. Lck plays an essential role in trmsducing CD4CD8 signals dunng T ce11
activation (Weiss and Littman, 1994). However. Lck can also mediate TCR signaling
independent of its association with CD4 (Abraham et al., 199 1c: Straus and Weiss, 1992).
Thus, Lck kinase activity is not strictly dependent upon its association with CD4
In the thymus, Lck expression begins in immature DN thymocytes and continues
throughorit developrnent to the mature SP stage (iMrirth et al., 1985: Perlmutter et al.. 1988;
Reynolds et al.. 1990: Wildin et al., 199 1 ). During thymocyte maturation. two promoters
regulate transcription of the lck gene (Reynolds et al.. 1990: Wildin et al.. 1991). The
proximal promoter. located adjacent to the Ick coding sequence. and the distal proximal
promoter. positioned at Ieaït 9 kb upstreüm from the proximal prornoter in mice. are both active
in immature thymocytes ( D N and DP) while the distal promoter preferentially funciions in
mature thymocytrs and T cells (Reynolds et al.. 1990: Wildin et al., 199 1 ). Usage of the
proximal promoter results in expression of type 1 [ck mRNA while the distal promoter directs
transcription of type II Ick mRNA. Type 1 and type II Ick mRNAs. which differ only in their
5' untranslated regions. are present at similar levels in immature thymocytes. but 5- to 9-fold
more type I I than type 1 ick mRNA is present in mature thymocytes and T cells. The
difference in expression of thrse transcripts in mature T cells is suggested to reflect inactivation
of transcription at the proximal promoter.
In mature T cclls and immature DP thymocytes, Lck prclerentially associates with CD4
relative to CD8. In immature DP thymocytes, 2550% of surfxe CD4 molecules versus 2 9 of
su r fxe CD8 associates with intracellular Lçk (Wiest et al., 1993). This reHects intrinsic
differences in the ability of the cytoplasmic tails of CD4 versus CD8 to interüct with Lck. In
addition, DP thymocytes express approximately equal amounts of CD8a and tailles CD8a
( ~ ~ 8 a ' ; Parnes. 1989). the latter which fails to associate with Lck (Zarnoyska et al., 1989).
The amount of Lck associated with CD4 is decreased by intrathymic CD4 and MHC class II
interactions (Wiest et al.. 1993). This reduction in üvailüble Lck-associated CD4 appears to
impair TCR signaling in DP thy mocytes. unless TCRs are coaggregated w ith coreceptors,
presumably due to insufficient lsvels of CM-itssociated Lck for phosphorylation and activation
of ZAP-70 (associated with tyrosine-phosphoryllited TCRC) and the initiation of subsequent
TCR signaling events (Wiest et al.. 1996).
A role for Lck in thymocyte development was elucidated in rrünsgeniç rnice braring either
wild-type or a constitutively active Ick (containing a phenylalanine for tyrosine substitution at
position 503) under the Ick proximal promoter (Abraham et al.. 199 la). leading to expression
in DN and DP thymocytes (Reynolds et al., 1990: Wildin et al.. 1991). In transgenic mice
expressing either wild-type or activated Ick (lckF505 ). anirnals with high levels of the
triinsgene demonstrated thymic CD3- DN tumors with irnpaired V(D)JP rearriingrments at the
TCRP Iocus (Abraham et al.. I99 16; Anderson et al.. 1992). Mice with Iower transgene copy
numbers had thvmi consisting primarily of DP ceils that Iück TCR surface expression with
ïeduced lrvels of V(D)JP rearrüngements at the TCRP locus. These results suggest li potential
role for Lck in mediating rvtnts normally associated with TCRP expression. such as alielic
exclusion and maturation of DP thy mocytes.
Definitive rvidence of a role for Lck in mediating allelic exclusion and the DN to DP
transition was confirmed in Ick trünsgenic mice. Trünsgrnic mice were grnerated expressing a
dominant nrgative lck (contüining an arginine for lysine substitution at position 173 which
inhibits phosphate transfer) under control of the lck proximal promoter (Lrvin et al.. 1993).
The results demonstrated that the lckR273 transgne causes a dose-dependent inhibition of the
DN to DP transition md TCRP remangement (Levin et al.. 1993). Thymic development in the
most severely affected line. expressing 12-fold more Lck R173 protein over endogenous Lck,
is arrested iit the CDISf DN ce11 stage with thymic cellularity at 1% of normal levels. In
lckR273 mice. overexpression of the lckR273 transgene prevents a TCRP transgene from
mediating allelic exclusion of endogenous V(D)JP rearrangements. suggesting that the Lck
R273 protein competes with wild-type Lck in mediating TCRP signaling (Anderson et al.,
1993). A role for Lck in pre-TCR signaling is also supported by results from RAGI '1- mice
expressing the IckF505 transgene. since thyrnic cellularity as well as DP cell development is
restored in kkFj05 IRAGI -Id mice (Mornbaerts et al.. 1994). The thyrnic cellularity is
restored in FTOCs of M G -/- thyrni treated with anti-CD3 (Levelt et al.. 1993~) and in RAG
-1- rnice following either expression of a TCRP transgene (Mombaerts et al.. 199%: Shinkai et
al., 1993) or iri vivo anti-CD~E treatrnent (Jacobs et al., 1994; Wu et al., 1996). Furtherrnore,
expression of a TCRP transgene (Mombaerts et al., 1994) or in vivo administration of anti-
CD% (Levelt et al.. 1995: Wu et al.. 1996) fails to improve thyrnic developrnent in RAG-I -'- k k -1- mice. Collrctively. these results suggest that TCRPICD3-stimulation of the DN to DP
transition and TCRp allelic exclusion occurs in m Lck-dependent mrinner.
The role of Lck in thymocyte development has been further assessrd in another loss of
tiinction model. ln mice bearing a targeted disruption in the lck genr. rhere is a 10-fold
reduction in the absolute number of DP çells rind a severe decrease in the number of mature SP
crlls (~Molina et al.. 1993). Peripheral T cells are both grearly reduced in numbers and
functionally impaired (Wen et al.. 1995). The absence of Lck also impairs the development of
VS3 dendritic rpidermal T crlls (Kawai et al.. 1995) and Vylo 1 l TCR trünsgenic T cells
(Penninger et al.. 1993). However, allelic exclusion at the TCRp locus is virtually normal in
lck -1- mice (Wallace et al.. 1995). Various reüsons may explain the observed differences in
thyrnic development between lck and lckR273 mutant mice (Molina et al.. 1992; Levin et
al.. 1993). Since the Ick grne is disrupted in the most 3' coding exon. a srna11 amount of
functional kinase may be generated from mutant fck transcripts. Second. i t is possible that
other PTKs prornote thyrnic developrnent in lck -'- mice via an alternative pathway. In lck
R273 mice, the IWold higher expression levels of LckR273 protein relative to endogenous
Lck müy compete not only with endogenous Lck but other PTKs and inhibit pre-TCR
signaling. A potential candidate is the Src hrnily non-receptor PTK Fyn.
C. Role of Fyn
Various studies have assessed the role of Fyn in TCR signaling and T ce11 development. In
transgenic rnice overexpressing wild-typelvn under the control of the ick proximal prornoter,
thymocytes demonstrate enhanced responsiveness to TCR crosslinking (Cooke et al., 199 1).
These results are specific to Fyn because thymocytes frorn transgenic mice overexpressing the
closely-related Src fümily PTKs. Lck and Hck. fiil to exhibit hyperproliferative responses
(Cooke et al.. 199 1 ). Expression of constitutively activütedfvn (T ) orj jn (BI (containing a
phenylalanine for tyrosine substitution at position 528) or both Fyn isoforms in antigen-
specific T ceIl hybridoma ceils enhances the responsiveness of these cells to TCR stimulation
(Davidson et al.. 1992). Conversely, thymocytes derived from transgenic rnice bearing
catalytically inactive fin (B) (containing a glutamine for lysine substitution at position 296) fail
to undergo TCR-rnediated stimulation (Cooke et al.. 199 1 ). Finüily. mature thy mocytes and T
cells from micr drficient forjjm ( T ) (Appleby et al.. 1993: Swan et al.. 1995) or bothhn (7')
andfin ( B ) (Stein et al.. 1992) dernonstrate impaired TCR signaling. These results suggest
that Fyn has a prominent role in KR-mediated signaling in thymocytes 2nd mature T cells.
The level of Fyn expression increases by approximately IO-fold as DP thymocytes mature
to SP thymocytes (Cooke et ai.. 1991). The requirement for Fyn in thymic development has
been tiirther rxarnined in mice lackingfjw (7) (Appleby et al.. 1991) or both Fyn isoforms
(Stein et al.. 1992). Howevrr. normal numbers of DP and SP thymocytes are generated infin
-1- rnice dernonstrating that Fyn does not play an essential role in T crll development. In
addition. positive selection of TCR trrinsgenic thymocytes proceeds normlilly in the absence of
Fyn (T) (Swan et al.. 1995). Collectively. the results suggest that Lck plays a critical role in
thymocyte development. whereiis Fyn is not essential.
5. Positive Selection at the DP to SP Transition
A. Historical overview
The vast majority of DP thymocytes do not mature further and undergo programmed ce11
death (PCD) within 3 to 4 days. However. a subset of DP crlls beiiring TCR that can
recognize self MHClpeptidr ligand with sufficient at'finity are positively selected and
differentiate into CD4+ and CD8+ T cells. The concept of thymocyte positive selection based
upon TCR specificity was first demonstrated by studies of bone marrow chimeras. In bone
marrow chimeras established by reconstituting irradiated MHC A type mice with MHC
heterozygous ( A x B)F 1 bone marrow cells. cytotoxic T cells were generated that only lysed
tÿrgets of the MHC A type (Bevan. 1977: Zinkemagel et al.. i 978). Thus. the MHC genotype
of the imdiated host. rüther than that of bone marrow precursors. determines the speci tïcity of
the MHC restriction. Treatmcnt of mice with antibodies specific for MHC class II or 1
molecules frorn birth onward results in ri specific reduction in mütiire CD4+ or CDS+ T cells,
respectively (Kruisbeek et al.. 1985; Marusic-Golesic et al.. 1988). Similar studies (Ramsdell
and Fowlkes. 1989; Zunigü-Ptlucker et al., 1989: Zuniga-Pflucker et al.. IWO) were
conducted with CD4 and CD8 specific antibodies. and these treatrnrnts also inhibited the
generation of the respective mature CD4+ or CD8+ T celi subpopulations. Positive selection
was also confirmed by studies using antibodies specific for the Vp 1 7af subset. demonstrating
the influence of MHC hiiplotype on positive selection of specific TCR-VP segments (reviewed
in Blackman et al.. 1990). Vp 17ü+ CD4+ cells are present rit a greater frequency in H-2s
( 14%) than in H-2b (4%) mice (Kappler et al., 1989). The high frequency of VP 17a+ CD4+
cells is also dominant in H-2 heterozygous (b x q ) FI mice. thus retlecting positive selection by
H-2s. rather t han negative selection by H-zb (Blackrnan et al.. 1990). Further iinalysis of bone
marrow and thymic chimeras demonstrated that the frequency of VP17a+ CD4+ cells is
determined by MHC molecules expressrd by thymic epithelial cells (Blückman et al., 1990).
Thus, multiple approaches have demonstrated evidence of positive selection in T ceil
development.
B. TCR transgenic rnodels
Positive selection has also been studied by generating mice transgenic for rearranged TCR
a and p chains from T cell clones with a detïned specificity for MHC and peptide (Berg et al..
1989; Kaye et al., 1989; Sha et al., 1988; Teh et al.. 1988). In the original K R transgenic
model (Teh et al., 1988). mice were created that express TCRaP transgenes specific for the
male HY antigen presented in association with k I - 7 ~ ~ . T ce11 developrnent is arrested at the DP
stage. unless the restricting ~ - 2 ~ b molecule is expressed in the thymus (Kisielow et al.. 1988;
Teh et ai., 1988; Scott et al.. 1989). Thus, the anti-HY TCR is predominantly expressed on
DP and CD8+ thymocytes and mature T cells, but not on CD4+ thymocytes in H-2b fernale
mice. Similady. TCR transgenic CD8+ thymocytes are only generated in irradiated femaie H-
2 ~ b mice reconstituted with stem cells from anti-HY TCR triinsgenic mice, indicating that
positive selection is dependent upon thyrnic expression of selecting MHC molecules (Kisielow
et al., 1988). In non-H-2b anti-HY TCR transgenic mice. the TCR transgene is expressed on
DN and DP cells but not on CD4+ and CD8+ thymocytes (Scott et al., 1989). Thus, the
specificity of the transgenic anti-HY TCR for MHC class 1 rnolecules results in the selection of
CD8+ cells expressing high levels of the anti-HY TCR. However. some CD4f T cells
expressing transgenic TCRuP are also generated because the TCRa transgene incompletely
suppresses rearrangrments of endogenous TCRa genes (Bluthmann et al.. 1988). Thus,
CD4+ T cells generated in these mice express a TCR composed of a transgenic TCRp chain
paired with an endogenous TCRa chain that is selected by MHC class II molecules.
Consequently, CD4+ T cells are not apparent in anti-HY TCR transgenic rnice homozygous for
the scid mutation because of a defective remangement mechanism in these mice (Scott et ai.,
1989). Thus. the anti-HY TCR transgenic animal model has been important in understanding
positive selection.
Results from the anti-HY TCR transgenic model have now been extended to other MHC
class 1-specific TCRs (Sha et al., 1988b; Sha et al., 1988a) as well as to MHC class II-specific
TCRs (Berg et al., l989b) transgenic animal models. In the latter case. transgenic TCRap
were derived from the CD4+ T ce11 clone 2B4 specific for 1 - E ~ associated with a pigeon
cytochrome c peptide (Berg et al., 1989b). In contrÿst to MHC clüss 1-specific TCR transgenic
rnodels. there is a bias towards the generation of CD4+ peripheral T cells in 281 transgenic
mice and other MHC class II-specific TCR transgenic rnodels (reviewed in Robey and
Fowlkes, 1994). In surnmary, the results indicrite that the LMHC specitïcity of the TCR
determines whether DP thymocytes develop into the CD4 or the CD8 lineage.
It is clear that TCR transgenic micc have been extremely useful tools in studying positive
seleciion. However, there are some problems in studying these mice (von Boehmer and
Kisielow, 1990). For example. TCRup trünsgenes are often prematurely expressed during T
ce11 development in TCR transgenic mice (Teh et al., l9F)O). and this rnay result in cells
prematurely responding to signais that induce thymic maturation. Additionally, the TCR
transgenes are expressed at abnormally high levels in immature thymocytes, which may
interfere with norrnal T ce11 maturation.
C. Role of the thymic microenvironment
In the thymus, positive selection is thought to be primarily mediated by rüdioresistant
thymic epithelial ceils based upon studies of borh bone marrow (Bevan. 1977: Zinkernagel et
al.. 1978) and thymic chirneras (Zinkernagel et al.. 1978; Lo and Sprent. 1986). Multiple
studies suggest thnt the thymic cortex is the p n m q site of positive selection. Trünsgenic mice
were generated that expressed 1-E cc genes cxrying specific deletions (AX and AY) in the 5'
regdatory regions resulting in 1-E expression in specific thymic cell types (van Ewijk et al.,
1988). The AX rnice feature 1-E expression in the thymic medulla, but not in the thymic
cortex, whereas I-E was reciprocally expressed in AY mice. Development of CD4f TCR+
cells (Berg et al., 1989b) and CD4+VP6+ T cells (Benoist and Mathis. 1989; Bill and Palmer,
1989). both of which depend upon LE moiecutes, was irnpaired in AX mice but not in AY
mice, suggesting a requircrnent for MHC expression in the thymic cortex (Benoist and Mathis.
1989; Berg et ai., 1989b; Bill and Palmer, 1989; Cosgrove et al., 1992). Additionally, a recent
study demonstnted that expression of I-A molecules rxclusively in the thymic cortex also leads
to positive srlection of CD4+ T cells (Laufer et al.. 1996). In surnmüry. positive selection of
the rnajority of CD4+T cells is dependent upon interictions with thymic epithelial cells in the
cortex.
D. Peptides in positive selection
Recent studies have demonstrated the importance of peptides in thymic selection. Initial
studies of MHC mutant mice showed a role for peptides in thymocyte development (Nikolic-
Zugic and Bevan, 1990; Sha et al.. 1990). In these mice. mutations in MHC c l a s I molecuIes
affected residues involved in peptide binding withour altering residues that contact the TCR.
The MHC mutant mice demonstrated irnpüired devrlopment of CDS+ cells bearing a class I-
specific TCR transgene. suggesting a role for self peptides in thymic selrction. More recently,
an iri stifro mode1 system using ktal thymi from rnice defective in MHC chss i surface
expression has bern developed to address the rolr of peptides in restoring drvelopment of
CD8+ cells (Ashton-Rickardt et al.. 1993: Hogqiiist et al.. 1993). Stable MHC class 1
expression is dependent upon the MHC çlass 1 heavy chain associating with p2m and cytosolic
peptides. which are translocated into the endoplasmic reticulum by a peptide pump derived
from products of the peptide trcotsporter cissocitrted \rvifli mri,qeei proce.s.siir,q-l c i r d -2 (TAP-I
and TAP-2 ) genes. Since rnice lacking either &?m or TAP-I express unfolded MHC cIass 1
heavy chain. addition of peptides and PZm to fetal thymi derived from these MHC class I
mutant mice restores MHC class I surface expression and developrnent of mature CD8+ cells
(Ashton-Rickardt et al.. 1993; Hogquist et al.. 1993). However, only some of the peptides
capable of rescuing MHC class I expression induce positive selection of CD8+ cells,
suggesting that peptides do not simply siabilize MHC class I molecules but intluence the
specificity of the TCWMHC interaction during positive selection (Ashton-Rickardt et al..
1993).
Subsequent studies used FTOCs from MHC class 1-restricted TCR tninsgenic, MHC class
1-mutant mice to directly address the role of peptide speciticity in positive selection. Bevan's
group assessed positive selection in piri FTOCs expressing a transgenic TCR specific for
the ovalbumin (OVA) peptide associated with H - X b (Hogquist et al.. 1994). while other
groups examined thymic selection in TAP-1 -1- and Ptti -1- FTOCs expressing a transgenic
TCR specific for the mouse lymphocytic choriorneningitis virus (LCMV) peptide bound to H-
2 ~ b ( Ashton-Rickardt et al.. 1994: Sebzda et al.. 1991). These studies employed variants of
antigenic peptides, some of which have been identified as TCR agonists and antagonists
depending upon their activity of mature T cells. Agonist peptides lire strongly antigenic for
mature T cells. whereas mtagonist peptides are classifird as ligands that inhibit the response of
antigenic-specific mature T cells to suboptimd amounts of antigenic peptide. The Bevan study
showcd that positive selsction of anti-OVA TCR transgenic CD8+ cefls in p2tti -1- ETOCs is
induced by a subset of antagonist peptides and not by control peptides (Hogquist et al.. 1994).
Similarly. others found that low concentrations of agonist LCMV peptides prornote positive
selrction of anti-LCMV transgenic CD8+ cells in either TAP-l -1- or b2m -/- FTOCs, whereas
control peptides do not ( Ashton-Rickxdt et al.. 1994: Srbzda et al.. 1994). In surnmary, these
results suggest that peptides have a specific role in positive selection via interaction with the
TCR and support the affinityhvidity mode1 of positive selection (see Section 7 8 ).
E. Consequences of positive selection
Positive selection of DP thymocytes results in a complrx series of drveloprnentd changes
leading to survival and maturation of thyrnocytes. While some of thesr changes are evident at
the DP stage, other changes are apprirent during or after the DP to SP transition. In addition to
changes in CD4 and CD8 expression (see below), there are other features of positive selection.
During the DP to SP transition, pre-Ta expression is terrninüted (Saint-Ruf et al., 1994).
Additionally, there is an upregulation of TCRap expression, with a 10- to 30-fold increase in
TCR density during the DP to SP transition (Havran et al.. 1987: Guidos et al.. 1990). This
c m be accounted for by the elevation of TCRa RNA levels and protein synthesis. resuiting in
increased assembly of TCRap complexes (Kearse et al.. 1995). Unselected TCR+ D P
thymocytes are suggested to undergo multiple TCRa rearrangements until they undergo
positive selection (Petrie et al.. 19934. In vivo studies dcmonstrated that positive selection
prevents further TCRa rearrangements by terrninating RAGI and RAGZ transcription
(Borgulya et al.. 1997: Briindle et al.. 1992: Kouskoff et al.. 1995). TCR engagement of DP
thyrnocytes in r g i r r r , similürly induces a decrease in RAGI and RAGZ mRNA expression
(Turka et al.. 199 1 b). Thymocytes thus maximize their chances for positive selection by
continually making new receptor combinations until RAG expression is shut off due to
positive selection or ce11 death.
Other changes t hat accornpany positive selection include expression of activation and
adhesion molecules. Expression of CD69. an activation molecule that is rapidly induced in
mature T cells following TCR stimulation ( H u a et al.. 1986; Cosulich et al.. 1987: Testi et al..
1989). is expressed on thymocytes that are undergoing thymic selection (Swat et al.. 1993;
Yamashita et al.. 1993). Other changes that accompany positive selection include upregulation
in surface expression of CD5. MHC class 1. CD45RA. and down-regulation of Thy- 1. heat-
stable antigen (HSA). and the novel thymic difkrentiation üntigen. F3 (Fowlkes and Pardoll,
1988; Barthlott et al.. 1996). Similarly. the functions of the intrgrins VLA-4 and VLA-5 are
upregulated during the DP to SP transition leading to a loss of firm adhesion, possibly
allowing positively selected çells to undrrgo migration from the cortex to the medulla (Crisa et
al.. 1996).
Positive selection is also iiccompanied by enhanced ce11 survival. Thymocyte survival
during T ce11 development is regulated by the Bcl-2 family of proteins. including Bcl-2. Ba,
B ~ 1 - x ~ . and Bcl-x,. The promoters of cell survival. Bcl-2 and Bcl-x,. whose functions are
modulated by interactions with other members of the Bcl-2 family, such as Büx (Oltvai et al.,
1993). Bad (Yang et al., 1995). and Bag-1 (Takayama et al.. 1995). are reciprocally expressed
during T ce11 development. Bcl-2 is predorninantly expressed in immature DN and mature SP
thyrnocytes and is low in immature DP thymocytes (Veis et al.. 1993). However. it is
upregulated during positive selection of DP thymocytes (Linette et al.. 1994). In contrast.
expression of B ~ 1 - x ~ . the major isoform of Bcl-x. increases from the DN to the DP stage. and
is down-regulated in the DP to SP transition and is absent in mature T cells (Ma et ai.. 1995).
Transgenic studies demonstrated that Bcl-2 enhances the survival of DP thymocytes in culture
but hils to inhibit thymocytes from undergoing negative selection (Sentman et al.. 1991:
Strasser et al., 199 1; Veis et al., 1993; Tao et al., 1994). in the absence of Bcl-2. thymocyte
maturation is normal. but shonly after binh both immature and mature thyrnocytes demonstrate
extensive apoptosis. leading to a depletion of T cells in the thymus and periphery (Nakayama et
al., 1993~). Although Bcl-2 and Bcl-xL are expressed at distinct stages of T cd1 developrnent,
trmsgenic studies drmonstrated that B ~ 1 - x ~ can rescue mature T cells in hcl-2 -1- rnice (Chao et
1 , 9 9 ) In the absence of Bcl-xL. DP thyrnocytes demonstrate rnhanced susceptibility to
apoptosis with no effect upon the survivül of mature thymocytes and mature T cslls (Motoyama
et al., 1995). Thus. Bcl-xL appeürs to regulate the lifespün of DP thymocytes prior to
selection. whereas Bcl-2 maintains the survival of positively selected cells.
F. I\IIodels of CD4/CD8 lineage cornmitment
Various rnodels have bren put fonviird to rxplain how DP cells brcome comrnitted to the
CD4 or CD8 lineage (reviewed in von Boehmer. 1996). The instructional mode1 posits that DP
cells bearing an MHC class 1-restricted TCR are instructed to become mature CD8+ SP cells by
coengagement of TCR and CD8 with MHC class Vpeptide complexes. whereas DP cells
bearing an MHC class II-restricted TCR differentiate into mature CD4+ SP cells upon
coengagement of TCR and CD4 with MHC class IVpeptide complexes (von Boehmer, 1986;
Robey et al.. 1991). Thus, this model predicts thiit different signals are delivered to DP
thymocytes by TCR coengagement with CD4 or CD8. Howrvrr. the instructional model has
been challenged by the results of several studies (von Borhmcr. 1986; Chan et al., 1993;
Crurnp et al.. 1993: Baron et al.. 1994; Itano et al., 1994: Robey et al.. 1994: van Meerwijk et
al., 1995). For exarnple. several studies suggest that the development of TCRmcd ~ ~ 4 1 0 CD8+
cells is not dependent upon TCWMHC class 1 interactions. and the generiition of K ~ m e d
C D ~ + C D ~ ~ O cells does not require TCWMHC clriss II interactions (Chan et al., 1993; Cmmp
et al., 1993; van Meerwijk et al., 1995). If thymocytes with fhese transitional phenotypes
indeed represent lineage-committed intermediates. these results challenge the instructional
model for CD4/CD8 lineage commitment and lend support to a stochastic/selective model as a
rnechanism for lineage commitment (Robey et al., 199 1 ). In this model. DP cells randomly
extinguish expression of either CD4 or CD8 following TCR and coreceptor coengagement. In
the second step. those cells braring a coreceptor rnatched to the same MHC specificity of the
TCR receive a selective signal rescuing them from ce11 death following TCR and coreceptor
coengagement with MHC clliss 1 or class 11 molecules. These committed cells then complete
differentiation in response to a coreceptor-dependent TCR signding event.
The selective/stochastic model hüs also been challenged by results from in vitro and in vivo
studies indicriting that sorted ~ C ~ m e d CD4+ ~ ~ 8 1 ~ cells contained precursors not only for
T C R ~ ~ CD4 SP but also for T C R ~ ~ CD8 SP cells (Lundberg et al.. 1995; Suzuki et ai.. 1995).
In conirüst, soned TCRmcd CD& CD8+cells contained precursors exclusively for T C R ~ ~ CD8
SP crlls. These lindings led the authors to suggest a delàult/instructive mechanism that
proposed asyrnrnetric signaling requirements for CD4 versus CD8 lineage cornmitment. CD4
lineage commitment is postulated to occur in the absence of MHC-dependent interactions
whereas cornmitment to the CD8 lineage is dependent upon an instructive signal involving TCR
and CD8 CO-engagement of MHC class Vpeptide complexes. However. interpretûtion of these
studies is clouded by potential problems such as the purity of the sorted celis used in these
studies. The precursor potential differences between T C R ~ ~ ~ C D ~ + C D ~ I ~ cells and T C R ~ ~ ~
~ ~ 4 1 0 CD8+ cells may not redect differences in lineage committment. rather. differences in
survivai.
Recently, yet another model for CD4KD8 linesge commitment has emerged. In this
revised two-step instructivelselective model. it is proposed thai MHC class I and class II
molecules, in response to coengagement of TCR and coreceptors. deliver quantitatively
different signais such that stronger signals induce cornmitment to the CD4 lineage and weaker
signals promote CD8 lineage commitment (Itano et al.. 1996: Mütschük et al.. 1996). It is
suggested that this reflects a greater ability of Lck to bind to the cytoplasmic domain of CD4
than to CDS. Following coreceptor down-regulation, ttioss cells with TCR matched to the
same MHC specificity as the coreceptor are selected to cornplete differentiation.
In surnrnary, there are currently four models to üccount for CD4/CD8 lineage cornmitment.
However. i t rernains to be determined which mechanism best explains commitment to the CD4
or CD8 lineage. Despite the differences betwern the models. there is generül agreement that
positive selection and lineage cornmitment does not result from a single TCWMHC interaction
but requires multiple TCR engagements with the thyrnic microenvironment.
G. Positive selection as a multi-step process
The DP to SP transition is currently viewed as a process thüt is dependent upon multiple
interactions with thymic MHC molecules. For example. a rrcent study dernonstrated that
positive1 y seiected TCRf DP çells undergo sr veral TCR engagements with positive selecting
MHC molrcules until they cornplete differentiation (Kisielow and Maizek. 1995). Imponantly.
only those cells that successfully completed the whole maturation program were rescued from
apoptosis. In another study. an itr vitro reüggregate culture system was established in which
purified precursor T cells are mixed with purified thymic epithelium (Wilkinson et al.. 1995).
Puritkd CD69+ DP thymocytes, which are predominantly T C ~ m e d ' h i cells that have had at
least one TCR engagement. required the presence of thymic epithelium to efticiently cornplete
differentiation into SP thymocytes. However. others found that the thymic stroma is not
necessary for this process because purified T C R ~ C J ~ ~ ~ DP thymocytes differentiate into T C R ~ ~
CD8 SP cells in the absence of thymic stroma1 cells (Petrie et al.. 1993b: Kydd et al.. 1995).
The generation of TCRhi CD8 SP cells is more efficient when TCRmed'hi DP thymocytes,
derived from mice transgenic for the ce11 survivül regulator Bcl-2. are cultured without stroma
in vitro (Kydd et al.. 1995). Thus. the improved production of T C R ~ ~ CD8 SP thymocytes in
the presence of thyrnic epithrlium may reflect the capacity of the thymic stroma to deliver
survival signals to differentiating thymocytes. In summary. positive selection does not involve
a single maturation event but appears to requise sustained TCR/MHC interactions with the
thyrnic microenvironment.
6. Negative Selection at the DP to SP Transition
A. Role of thymic microenvironment
During T ce11 development. negative selection eliminütes potrntially autoreactive
thymocytes. Analysis of bone marrow chimeras and thymectomized. irradiated. bone marrow-
reconstituted mice dernonstrüted that clonal deletion of autoreüctive T cells is largely rnediated
by hematopoietic ctills (Ramsdell et al.. 1989; Roberts et al.. 1989; Speisrr et al., 1989;
Speiser et al.. 1990). Studies in chimeras and MHC transgenic mice dernonstrüte that thymic
epithelial cells can also mediate negative selection via rither clonal anergy (van Ewijk et d.,
1988: Ramsdell et al.. 1989; Roberts et al.. 1989: Gao et al.. L990; Schonrich et al.. 1992;
Bonomo and Matzinger, 1993: Oukka et al., 1996) or clonal deletion (Gao et al.. 1990: Speiser
et ai.. 1992: Kosaka and Sprent, 1993: Hoffmann et al.. 1995). Thus. both hernatopoietic and
non-hematopoietic cells can tolerize self-reactive thymocytes.
B. Models of negative selection
The mechanisms by which self-tolerance is estabiished have been investigated in normal
mice. Analysis of thymocytes bearing TCR-VP 17a. whicli a r t specific for L E MHC class II
rnolecules. indicates that they represent approximately 1-20 95 of peripheral T cells in mice
lacking I-E (Küppler et al.. 1987). In contrast. only O. 1 Cic of the peripheral T cells are TCR-
VP 17a+ in mice expressing I-E MHC class 11 rnolecules. since most are clonally deleted before
T ce11 development is completed (Kappler et al.. 1987). Similar findings can be extended to
other subsets of thyrnocytes bearing a puticulÿr TCR-VP specific for retrovirus-encoded minor
lymphocyte stimulatory (Mls) üntigens (Kappler et al.. 1988; MacDonald et al.. 1988: Pullen et
al.. 1988).
Trmsgenic models have dso proved useful in understanding negative selection. In fernale
mice expressing a transgenic TCR specific for the male HY antigen in association with H-ZD~,
most transgenic TCR thyrnocytes and T cells generated are biased towards the CD8 lineage.
By contrast, significantly fewer TCR transgenic T cells rire present in malt: mice due to clonal
deleiion, and those remaining are functionally inactivated (Kisielow et al.. 1988). Similarly, in
anti-H-2~d TCR transgenic mice, thymic expression of H-2Ld leüds to either clona1 deletion or
functional inactivation of autoreactive TCR transgenic cells (Sha et al.. I988a). Additionally,
in mice bearing the anti-LCiMV transgenic TCR. LCMV infection at binh induces clonal
deletion and functional inactivation of TCR transgenic T cells (Pircher et al.. 1989). Thus,
thymic tolerance occurs by both clonal deletion and fùnctiond inactivation.
C. Developmental stage susceptibility to negative selection
It il; not well understood which developmental stages are susceptible to clonal deletion or
functional inactivation. ResuIts from i i i ititro FTOC studies. and in ~*ii:o studies,
demonstrated that treatment with rinti-CD3 diminsites most DP cells (Smith et :il.. 1989; Shi et
al., 1991). In contrast, others found that FTOCs treated with anti-TCRp induced cIonal
deletion in only a subsrt of DP cells suggesting that the TCR and CD3 are functionally
uncoupled at the DP stage (Finkel et al.. 1989). Results from TCR transgenic mice studies
have also not cleÿrly rstablished which developmental stage is susceptible to negative selection.
In the anti-HY TCR transgenic model. DP thymocytes are deletrd (Kisielow et al.. 1988) or
prevented from developing (Takahama et al.. 1992) in the negütively selecting H-ZDb male
mice. In anti-LCMV TCR transgenic mice. which contain T cells bearing TCR with dud
specificity for LCMV and MIS". LCMV mediates deletion at the DP stage whereas Mlsa-
induced deletion is not evident until the SP thymocyte stage (Pircher et al., 1989). Thus, the
ligand (MHClpeptide or MHCIMls") may determine which stage is susceptible to clona1
deletion, likrly a result of differential ligand expression in the cortex versus the medulla. In a
non-TCR transgenic model. clonal deletion of cells bearing Mls-specific TCR-VP's primarily
occurs after the TCRIO DP stage (Guidos et al.. 1990). Thesr results suggest that the nature of
the ligand may govern which thyrnic stage is susceptible to negative xlection.
D. Role of accessory molecules
Interactions other than those involving TCR and MHClpeptide interactions rnay play a role
during negative selection. In vitro studies suggest that B7- 1 and ICAM- 1 together mediate
negative selection ( Kishirnoto et al.. 19%). while in vivo studies indicate potential roles for
CD40lgp39 (Foy et al.. 1995) and CD30lCD30 ligand (Amakawa et al.. 1996) interactions in
negative selection. Although FaslFas ligand interactions participate in apoptosis of
thymocytes. they have no requisite role in negative selection (reviewed in Nagata and Suda.
1995). The cos tirnulatory molecule CD28 is reported to increrise the sensitivi ty of thymocytes
to üpoptosis in rcsponse to anti-CD~E treatrnent (Punt et al.. 1994). However. others found
that CD28 doçs not have a critical role in negative selection of either TCR transgenic or Mls-
reactive V ~ 3 f T cells (Walunas et al.. 1996). Thus. the role of CD18 and othsr costirnulatory
molecules in negative st-lection requires further andysis.
7. Models to explain the paradox of positive and negative selection
A. Microenvironment model
Various rnodels have been proposed to address the püradigm of how both positive and
negative selection c m resulr from identical TCR and sel f-iMHClpepiide complex interactions.
One model suggests that positive and negative selection are rnediated by thymic epithelial cells
and hematopoietic cells. respectively (Sprent et al.. 1988). However, data from various
studies challenge this model. For ex;imple. positive selection is rnediated not only by thymic
epitheliül cells but also by fibroblasts (Hugo et al.. 1993; Priwlowski et al.. 1993). In addition.
bone-marrow derived ceils positively select class I iMHC-restricted NK cells (Bix and Raulet,
1992). Thus, positive and negative selection are not strictly mediated by two distinct subsets
of thymic stromal cells.
B. Affinity/Avidity model
A second model. known as the affinity model. suggests that the TCR affinity t'or its ligmd
determines the developrnental fate of a ceIl (Droge. 198 1). For a given self MHUpeptide
ligand, cells with high affinity undego clonal deletion or anergy. those with low TCR affinity
are positively selected. while thymocytes with TCRs bearing no affinity for self MHC are
neglected and die. Since interactions between thymocytes and antigen presenting cells are not
based upon TCR aftïnity alone but the overall avidity. the affinity hypothesis was modified to
becorne the avidity hypothesis (Janeway et al.. 1992). In addition to TCR affinity. factors
such as the density of the TCR andor its ligand (Berg et al.. 1990). as well as CD4 (Davis and
Littman, 1995) and CDS (Robey et al.. L992; Sherman et al.. 1992; Smith et al.. 1996) also
contribute to the overall avidity of these interactions. To test the üvidity hypothesis. multiple
studies üssessed development of TCR transgenic CD8+ SP thymocytes in MHC class 1-mutant
fetal thymi cultured in the presence of exogenous plm and varying amounts of peptide
(Ashton-Rickardt et al.. 1994: Sebzda et al., 19941. Two studies demonstrated that low
concentrations of agonist peptides prornote positive selection of anti-LCMV TCR transgenic
CD8+ cells. while highrr doses induce negative selection. In a different mode1 system
assessing thymic selection of anti-OVA TCR transgrnic CD8+ cctlls. an OVA antagonist
peptide mediates the generation of CD8f thy mocytes in P-7111 -1- thyrni and partial negative
selection in PZni 4- thyrni (Hogquist et al.. 1994). Thus. the OVA antagonist peptide induces
positive selection when expressed at low ligand density on PZ,,r -/* cells and partial negative
selection when expressed at high ligand density on 4- cells. Results from the LCMV and
OVA studies suggest that the overdl üvidity of the TCRMHC-peptide interaction determines
whether the outcome is either positive or negütive selection. However, results from a recent
study disagree with this strict definition of the differentid avidity model of thymic selection as
low quantities of an agonist peptide induce the generation of non-tunctional OVA-specific
T C R ~ ~ C D ~ ~ O cells in PZni 4- lobes (Hogquist et al.. 1995). These findings conflict with a
recent report that described a novel peptide agonist capable of producing functionally mature
T C R ~ ~ CD8+ thy rnocytes over a Iarge rmge of doses (Se bzda et al.. 1996). In addition, this
peptide agonist modifies the response of the cells suçh that they are rendered tolerünt against it
but can respond to a related peptide rigonist (Sebzda et al., 1996). In sumrnary. these studies
suggest that both peptide antagonists and peptide agonists can mediate positive selection.
C. Developrnental stage model
Another mode1 suggests that thymocytes are developrnentally programmed to undergo
positive and negativr selection üt different stages of maturation. This is supported by the
identification of two developmentally distinct subsets. TCRI" DP blasts and T C R ~ ~ ~
trmsitional cells (CD4+8I0 and CD410 8+ ). as Iikely targets of positive and negative selection.
respectively (Guidos et al., 1990). It is suggested that maturation of TCRIO DP blasts into
~ ~ ~ m e d cells occurs as a result of positive selection. based on the high frequency of T C R ~ ~ ~
cells beüring a particular TCR-VP in strains with the required positive selrcting element. In
contrast. a reduced frequrncy of cells bearing a TCR-VP specific for :i particular negative
selecting element is not observrd until the T C R ~ ~ mature stage. suggesting negative selection
occurs at or after the T C R ~ C ~ ceIl stage. Thus. positive selection precedes negarive selection in
this model. However, one caveat of this model is that the developmental fate of thymocytes
was assessed in response to MHClMls ligands and thus a different oiitcome rnay be obtained
using conventional MHClpeptide ligands.
The three models described above each potentially explain the paradox of positive and
negative selection in the thymus, but these models rnay not be rnutually exclusive. In the
thymus, both the cortex and the medulla are very heterogeneous compartments, and it is
possible that differences between thymic stroma1 cells in mediating positive and negative
selection may reflect differential expression of accessory molecules. Interactions involving
accessory molecules between thymocytes and thymic stroma1 cells c m complement TCWMHC
interactions. which in turn can potentiate the overall avidity of these interactions. In addition,
the sensitivity of thymocytes to TCR ligands rnay b r developmentally prograrnmed. resulting in
distinct TCR signaling thresholds during the DP to SP transition. At a particular stage, low
avidity interactions may be critical to drive positive selection. but higher avidity interactions
may be necessary to induce negative selection at the next devrlopmental stage. Consequentiy,
expression of accessory molecules in the thymic conex may be criticül in allowing interactions
mediated by cortical stromÿl crlls to reach a threshold necessary to trigger positive selection of
DP thymocytes. In contrast. the thymic medulla may express a different array of accessory
molecules that are incapable of promoting positive selection of DP thymocytes but instead
induce negative selection at a subsequent stage of differentiation. Thus. al! three mechanisms
may play a role in specifying positive versus negative selection.
8. TCRap Signaling in Positive and Negative Selection
The role of numerous signaling molecules in positive and negative selection has been
assesscd using mutant mice deficient for these proteins. Studies o f fw micc demonstrated
that neithrr positive and negative selection are affected by the absence of Fyn (Stein et al..
1993; Swan et al.. 1995). In contrast. multiple studies suggest that Lck has a role in positive
selection of DP thymocytcs (Teh et al., 199 1 ; Carrera et al.. l991a; Carrera et al.. 199%; van
Oers et al.. 1993: Ericsson and Teh. 1995) and lineage commitrnent (Itüno et al.. 1996).
Although Lck is demonstrüted to participate in TCRap signaling in mature T ceils (Abraham et
al., 1991~; Straus and Weiss, 1992). the role of Lck in TCRolp-mediated signaling during
positive selection of DP thyrnocytes remains unclear. The requirement for Lck in thymic
selection was recently assessed using trmsgenic mice expressing the same catalytically inactive
lck that completely blocks T ce11 deveiopment at the DN stage (Levin et al., 1993). under the
control of the k k dista1 promoter such that it is highly expressed in DP thyrnocytes, mature SP
thyrnocytes and T cells (Hashimoio et al., 1996). The expression of catalytically inactive Lck.
levels 2-fold above endogenous Lck. partially inhibits positive selection but fails to alter
negative selection of anti-HY TCR transgenic thymocytes. Thus, Lck may play a critical role
in both pre-TCR and TCRaP signaling during T ce11 development.
Recent studies have identitïed the role for ZAP-70 in pre-TCR and TCRap signaling.
Three different reports characterized a selective T ce11 immunodeficiency in humans that results
from mutations in the ZA P-70 gene ( A p i a et al.. 1994; Chan et al., 1994a; Elder et al.,
1994). An interesting fiature is that only CD4+ T cells were present in the periphery,
suggesting a ditferential requirement for ZAP-70 in positive selection of CD4+ versus CD8+ T
cells ( Arpriia et al.. 1994; Chan et al., 1994a; Elder et al.. 1994). However, results from ZAP-
70 -1- mice demonstratcd that positive and negative selection of both lineages are dependent
upon ZAP-70 (Negishi et al.. 1995). In contrast. normal numbers of DP cells are generated in
Z4P-70 -/- rnice. indicaring that this kinase is not essential for pre-TCR signaling. However.
a recent study demonstrated that pre-TCR signalin- results in the activation of ZAP-70,
suggesting that ZAP-70 may play a role signaling via the pre-TCR complex (van Oers et al.,
1995). In summary. ZAP-70 is critical in the DP to SP transition. but hirther studies are
required to determine the role of ZAP-70 during the DN to DP transition.
In addition to Fyn. Lck and SAP-70. the role of other molecules that rday signals
following TCR and coreceptor coengagement have also been addressed. The significance of
the mitogcn-activatrd protein kinase ( MAPK) cascade in thy mic developrnent was assessed in
genetic studies in anti-HY TCR transgenic mice expressing either dominant-nrgative Ras or
Mrk-1 transgenes (Alberolü-Ila et al.. 1995: Swan et al., 1995). The results suggest thnt
signaling via MAPK is necessary in positive selection. but not for negative selection. Sirnilar
results were observed in anii-HY TCR transgenic mice expressing the dominant-negative Ras
and Mrk-1 transgenrs simultaneously (Alberola-lia et al.. 1996). These results suggest
differential efkcts of the MAP kinase cascade on positive and negative selection.
Differential effects on positive and negütive selection have also been reported for the
transmembrane molecule CDS. During positive selection. thyrnocytes undergo increased CD5
expression (Fowlkes and Pardoll. 1988). Results in CD5 -1- rnice demonstrated that thymic
maturation is normal without any effects on negative selection (Tarakhovsky et al.. 1995).
However. CD3 -'- thymocytes were hyperresponsive to TCR-mediated signding in vitro.
suggesting that CD5 is a negative regulator of TCR signaling. Nevenheless. the lack of CD5
had variable rffects on thymic selection of thymocytes in three differrnt TCR tninsgenic
models. The lack of CDS improved positive selection of one TCR. increaïed negative selection
of a second TCR. and inhibited positive selrction of the third TCR. Büsed upon these results.
CD5 is suggested to be a negative regulutor of TCR signaling and the outcome of thyrnic
selection may retlect the overall avidity of the TCRMHC interaction.
9. Objectives
In the deveioprnental transition from the DN to DP stage. abundant evidence bas implicated
Lck in pre-TCR signaling. In the absence of Lck. thymocyte development is impaired.
resulting in a 10-fold reduction in the thymic cellularity and the numbcr of DP thymocytes.
Furthermore. there is a sevrre reduction in the number of mature SP thymocytes and peripheral
T cells. th es^ rrwits thus suggest that Lçk is important but not essential for thymocyte
developmrnt. Nrïertheless. the results raisc a numbcr of questions. Sincr the lack of Lck
incompletely inhibits thymocyte development at the DN to DP transition. do other PTKs
substitute for Lck in this rvent'? One potential candidate is Fyn because i t is cxpressed at
similar lrvels in DN and DP cells and associates with the CD3lTCR complex. Alrhough Fyn
does not play a critical rolr in thyrnic development, it may play a role during development when
Lck is absent. If so. does the hnction of Fyn depend upon a TCR expression'? To address the
role of Fyn in thymic developrnent when Lck is absent. I have generated lck -/- rnice that are
iilso deficient forLw. Additionally. 1 have genrrated lck -1- mice that express a constitutively
active mutant transgene. To determine whether the function of the JYII transgene is
dependent upon pre-TCR expressionJ~n transgenic mice were generated that lack both Lck
and RAGl. Using these two genetic approaches, thymic development was analysed by
changes in thymus cellularity and as well as phenotypic changes by tlow cytometry.
In the DP to SP phase of developrnent, various studies indicate a requirement of
thyrnocytes to undergo continuous TCR engagements in order to complete differentiation.
However. the ability of DP thymocytes to undergo maturation inro SP thymocytes in response
to TCR engagement alone has not been iissessed. Additionally. müturational changes
associated with positive selsction needs to be determined in DP thymocytes following TCR
stimulation. To address this. purified DP thyrnocytes were cultured i ~ z ritro in the presence of
imrnobilized TCR speci fîc üntibodies. TCR-stimulated DP thy rnocytes were t hen assessed for
phenotypic and molecular changes thnt accompany positive selection in vivo.
Chapter II
TCR Engagement of CD4+CD8+ Thymocytes In Vitro Induces Early Aspects of Positive Selection, But not Apoptosis
Tim Grovesl, Michael ~arsonsz, Neil G. Miyarnotoz, Cynthia J.
Guidosl
Division of Immunology & Cancer, Hospital for Sick Children Research
lnstitute and Department of Immunology, University of Toronto, Ontario
Cancer Institute, Princess Margaret Hospital, Toronto, Ontario,. Canada MSG
1 x 8
The contents of this Chapter were published in Journal of Imrnunology, 158,
65-75 (1997). O Copyright 1997. The American Association of Imrnunologists.
The majority of the work was performed bg Tim Groves except for Figure II-
6, which was done in cooperation with Michael Parsons.
II. 1. Introduction
During T ceIl development in the thymus, complex positive and negative selection
processes ensure that the peripheral T ceIl pool consists primarily of cells bearing highly
diverse. but self MHC-restricted and self-tolerant TCRaP. These selection processes begin
after DN precursors generate large numbers of immature TCRapl" DP precursors in response
to signals smaniiting from a pre-TcclTCRP (pre-TCR) complex (Saint-Ruf et al., 1994;
Anderson and Perlmutter. 1995). DP ihymocytes express a highly diverse array of
clonotypically distinct up TCR. the specificity of which determines the developmental fate of
each cell. Most DP thymocytes express TCR thar cannot recognize self MHC. and they
undtrgo programmed cell death within 3-4 days. In contrast. DP precursors bearing
potentially usehl. self MHC-restricted TCR are positively selrcted to survive and develop into
functionally mature CD4 or CD8 SP T cells (von Boehmer. 1991). Two models have been
proposcd to explain how most DP precursors bearing MHC class II-specific TCR will retain
expression of CDI . the MHC clüss I I coreceptor. whereas those bearing MHC class I-specific
TCR will retain expression of CD8. the MHC class 1 coreceptor. as they develop into mature T
4 1 s (Robry and Fowlkes. 1994). The stochastic/selectivr mode1 argues that lineage
commitment ocçurs randomly. perhaps in response to TCR signals (Chan et al.. 1993; van
Meenvijk and Germain. 1993). but without regard to TCR specificity. thus senzrating some
cells wi th mismatched TCR and coreceptor speciticities. Subsequently. coengagement of TCR
and CD4 or CD8 by the same MHClpeptide ligand would be required to positively select
lineage-committed cells bearing mütched TCR and coreceptor specificities. Thus. in this
model. lineage cornmitment and positive selection are temporally separated and mechanistically
distinct. In contrast. the instructional model posits lineage commitment is induceci by
coengagement of TCRKD4 or TCWCD8 on DP thymocytes. CD4 and CD8 are thought to
deliver distinct signals io DP thymocytes, causing them to selectively extinguish expression of
the inappropriate coreceptor (Robey and Fowlkes. 1994).
Positive selection also triggers a series of developmental changes that ultimately lead to
survival and matur~tion of lineage-committed thymocytes. Thymocyte sunival is regulated by
members of the bcl-2 gene hmily. Bcl-.r acts as a surviviil gene in DP thymocytes. but the
related bcl-2 gene is required for long-terrn survival of positively selrcted SP T cells
(Nakayama et al.. 1993a; Ma et al.. 1995; Motoyama et al.. 1995). Maturÿtion encompasses a
cornplex array of developmental changes that result in cessation of pre-Ta expression (Saint-
Ruf et al., 1994) and allelic exclusion of TCRa chains by repressing transcription of RAGI
and RAGZ (Borgulya et al.. 1997; Briindle et al.. 1992: Petrie et al.. 1993a). Additionally,
maturation involves quantitative changes in expression of ce11 surfice proteins that regulate
mature T ce11 functions. such as CD5. TCRap. MHC cliiss 1. heat stable nntigen (HSA). and
Thy- 1 (Fowlkes and Pardoll. 1988). Recent studies suggest thüt maturation is a complex.
multi-step process which requires multiple TCR engagements (Chan et al.. 1993: Davis et al..
1993; Petrie et al.. 1993b: van Mernvijk and Germain. 1993: Kisielow and Maizek. 1995; van
Meenvijk et al.. 1995). and that i t begins before DP thymocytes begin fo lose CD4 or CD8
(Hugo et al.. 1990: Ohashi et ai.. IWO; Shortman et al.. 199 1 : Luciis et al.. 1993: Lucas et al..
1994).
Multiple lines of evidence suggest that DP thymocytes can also undergo clonal deletior.
(negative selection). in response to TCR-ligand interactions (Robey and Fowlkes. 1994). For
example. DP thy mocy tes expressing transgenic TCRuP of de fined speci ficity are deleted in the
presence of the appropriate MHCIpeptide complex. but are positively selected when self MHC
is expressed in the absence of the cognate peptide (Kisielow et al.. 1988; Sha et al.. 1988a: Teh
et al.. 1988; Murphy et al.. 1990). In addition. apoptosis of non-TCR transgenic DP
thymocytes crin be rapidly induced in vivo by treatment with antibodies specitïc for the
TCRKD3 complex (Smith et al., 1989; Shi et al.. 1991). However, precursors of DP
thymocytes prernüturely express high levels of transgenic TCRaP (Teh et al.. 1990; Takahama
et al.. 1992). and in non-TCR transgenic mice. low levels of surface C D ~ E and TCRP, perhaps
cornplexed to pre-Ta. are expressed on precursors of DP thymocytes (Leveh et al.. 1993~ ;
Saint-Ruf et al., 1994). Ligation of C D ~ E or TCRp on TCR transgenic or non-transgenic
precursor cells generates signüls which inhibit their developmrnt into DP thymocytes
(Takahama et al., 1992). Therefore. it is not clear that DP thyrnocytes are the only or even the
major targets of deletional signüls in these in vivo experirnents. Other studies suggest that
clonal deletion can also take place later in thymocyte development. during the DP to SP
transition (Berg et al.. 1989a: Pircher et al.. 1989; Cuidos et al.. 1990).
Several important questions remüin unanswered by thesr i j r vivo studies of positive and
negative selection. First. given the emerging consensus that positive seltsction is rt multi-step
process involving a series of TCR engagements. which outcomes are immedinte and direct
consequences of TCR signtiling in DP thyrnocytes? Second, do purified DP thymocytes
undergo ripoptosis in response to engagement of TCRKD3 in the absence of other signals frorn
the microenvironment'? We have used an in vitro approach to address these issues and to
begin to resolve individual steps in the cornplex developmcntal process that governs the DP to
SP transition. Our results suggest that only a subset of developmenttil outcomes induced by
TCR-ligand interactions during the DP to SP transition in vivo are directly linked to TCR-
mediated signaling pathways in DP thyrnocytes. We also show that C D 4 or CDWmediated
signals can potentiate TCR signüls to induce these outcomes. which include some aspects of
the maturation and survival programs, but not CD-I/CDS lineagr cornmitment or clonal
dele tion.
II. 2. Materials and ~Methods
Antibodies
The sources of antibodies specific for CD4 (GK1.5). CD8a (53-6.7). CD5 (53-7.3). B220
(6B2). and Thy-l (53-2.1 ) have been described (Danska et al.. 1994). Antibodies specific for
HSA ( M 1/69), iL-ZRu (7D4). CD8a (YTS-169.4), TCRp (H57-597). and C D ~ E ( 145-?Cl 1)
were also used. These antibodies were puntied from tissue culture supematants contÿining 5%
NuSerum by protein G- or protein A-Sepharose chromatography . and conjugated using
standard techniques to tluroscein isothiocyanate (FITC). biotin. or phycoerythrin (PE), dl
purchased from Molecular Probes (Eugene, OR). The rernaining antibodies were purchased
from Pharmingen (San Diego, CA): H 1.2F3 (CD69), 14.8 (CDLCS-RA), MB23G2 (CD45-
RB). AF6-88.5 (H-Xb). 2D7 (LFA-1). and 37.5 1 (CD%). Antibodies specific for murine
Bcl-2 (3F 1 1 ) was kindly provided by Dr. S. Korsmeyer. The secondary reagents used include
avidin-PE (Criltag. San Francisco. CA) and a tandem conjugatr: of avidin-Cychrome-5/PE that
was prepared by conjugating PE to Cychrome-5 (Biological Detrction Systerns Inc..
Pittsburgh. PA) according to manuîixturer's instructions. Cychrome-SlPE was then coupled
ro avidin (iLIoleculiir Probes) using N-gamma-rnaleimidobutyry1oxys~~ccinimide (GMBS;
Calbiochrm. San Diego. C A ) and 2-iminothiolanr HCl ( X T . Pierce. Rockford. IL) as
described for PE çoiipling. Heteroconjugates of H57-597 and GK 1.5 (anti-CD4) or YTS-
169.4 (anti-CD8u) were prepared using the GMBW-IT procedure (Lcdbetter et al.. 1989).
Al1 antibodies were used at saturaring concentrütions. Rat IgG2a. rat IgGlb. and hamster IgG
isotype control antibodies conjugated to FITC. biotin. or PE were purchased from Pharmingen
and used at 1-3 yg/ml.
Mice
C57BLf6 (B6) and C.B- 17 scid mice were bred and maintüined at the Hospital for Sick
Children. Anti-H-Y TCR transgenic mice were kindly provided by Dr. Philippe Poussier,
Wellesley Hospital Research Institu te. Toronto. Ont.. wi th the permission of Dr. Harald von
Boehrner, Basel Institute for Immunology, Basel. Switzerland. MHC class lJ- x class 11-1- 86
mice were provided by Drs. Michad Gnisby and Laune Glimcher.
Purification of Thymocyte Subsets
Single-ce11 suspensions of thymocytes or lymph node cells were prepared in HBSS
containing 2% calf serum and 10 mM Hepes. Smal! post-mitotic (PM) and large blast DP
thymocytes were purified from unseparated cells by a combination of centrifuga1 elutnation,
which separates srnall PM from large cells, and fluorescence-üctivated ce11 sorting using a
FACStar Plus (Becton Dickinson. Mountain View, CA). Fractions enriched for PM cells or
large cells were stained with anti-CD&-PE (53-6.7) plus anti-CD5-FITC and then sorted for
high expression of CD8. low expression of CD5 (which correlates with TCR expression) and
low (PM) or high (blasts) fonvürd li_oht scatter. To assess the purity. sorted cells were stained
with allophycoerythrin (APC)-GKl.5 or APC-H57-597. Rrsults indicated that sorted cells
were 99% T C R ~ D ~ + C D ~ + .
CD4 SP thymocytes and lymph node cells were purified as follows. Cells were stained
with biotinylatrd anti-CD8u (and anti-B720 to depletr B crlls from lymph node cells). washed,
and then resuspended in 1 ml staining medium plus 20 ml avidin-conjugated paramagnetic
beads (Advanced Magnetics. Inc., Cambridge. MA). After 30 min at - 1 T with constant
mixing, bead-coarrd cells were removed using a Bio-MAG magnetic separator (Advanced
Magnetics) as described (Guidos et al.. 1990). CDS-depletrd thymocytes were then stained
with iividin-Cychrome-5lPE. FITC-53-7.3 and PE-GK i .5 and CD4+ SP cells were sorted
(98-998 purity ). CD8 and BZO-depleted lymph node cells wrre 92-95% CDM+ and were not
puri fied further.
Cell Stimulation
Tissue culture wells were pre-coated with PBS. protein A-purified 857-597, or
heteroconjugates of H57-597lGK 1.5 or H57-597lYTS- 169.4 for 2 hr at 3 7 T . or ovemight at
4 O C and washed 3 times with PBS before the addition of the soncd DP subsets. Ionomcyin
(Calbiochern) was used at 500 ng/rnl. Dexarnethasone was purchased from Sigma. Cells were
cultured in RPMI 1640 containing 10% FCS, 5 x 1 0 - ~ M 2-ME. 10 mM glutamine. 10 rnM
Hepes, penicillin ( 100 Ulml), and streptomycin ( 100 pglml) at 2- 10 x lo5 Iwell in 48-well
tissue culture plates. After 1 or 1 days of culture. çells were harvested and
immunofluorescence staining was used to quantitate surface or intracellular expression of
various markers.
Immunofluorescence and Flow Cytometry
Detaiis ot' the staining procedure for surface molecules have been previously described
(Groves et al.. 1995). Cultured DP thymocytes were stained for CDS, CD69, CD4, or CD8a
using either FITC-conjugated antibodirs or biotinylated antibodies and avidin-Cychrome-YPE.
Intracellular BcI-3 expression was evaluated by staining celis in HBSSI2% cclf serum with
Cychrome-YPE. which is excluded by live cells. followed by biotinylated 3Fl 1 and avidin-PE
in HBSSI7lc calf serum containing 0.1 8 saponin (Danska et al.. 1994). Fluorescence was
analyzed on ii FACScan flow cytometer (Becton Dickinson) with LYSIS II software. Dead
celIs and debris were excluded from the analysis on the basis of Iow tonvard scatter and high
propidium iodide (PI) tluorrscence.
RNA Analyses
Expression of prr-Ta RNA in VL3-3MZ cells was asscssed by Northern analysis as
described (Groves et al.. 1995). RAGl and pre-Ta mRNA levels in DP thymocytes were
evaluated by semi-quantitative RT-PCR. RNA was extracted from unstimulated and stimulated
cells and reverse-transcribed into cDNA as previously described (Danska et al., 1994). Five-
fold serial dilutions of the cDNA were then used as a template for PCR amplications using
primers specific for RAGI. pre-Ta mRNA, or P-crctiiz . The primers used were: RAGl F3 and
R6 (Chun et al.. 199 1 ), fl -actif1 sense (Danska et al., 1994). -crc~iri anti-sense 5'-TGA GGT
AGT CTG TCA GGT CCC; pre-Ta sense (CTG CAA CTG CGT CAT GCT TC); pre-Ta
antisense (TCA GAG GGG TGG GTA AGA TC). Primers for both p -~[ctal and RAGI were
included in the same PCR reüctions ( S O C annealing, 37 cycles) and yielded 56 1 bp and 429
bp products, respectively. PCR reiictions for pre-Ta (59W iinnealing, 37 cycles, 660 bp
product) and P - m r i n were run separately. After amplitïcation, the PCR products were
electrophoresed on a 1.5% agarose gel, transferred ont0 nylon membranes and subjected to
Southem analysis. The membranes were UV cross-linked. prehybridized at 650C in 6X SSC,
0.5% SDS, 0.5X Denhardt's solution and 50 @ml denatured salrnon sperm DNA, and then
hybridized overnight in the same solution with 32~-dCTP-labelled probes for murine pre-Ta,
M G - I , or P -cictiri cDNA. Al1 probes have been previously described (Groves et al., 1993,
except for pre-Tu, which was kindly provided by Dr. Jayne Danska. Blots were washed three
times (final wash in 0.2X SSC and 0.1% SDS at 65OC) and then exposed to the MoIecular
Dynamics Phosphorimager (Sunnyvale. CA) system and then analyzrd using Imagequant 3.0
software (iMolecular Dy namics ).
For quüntitating relative amounts of Ick type 1 and type II transcripts by RT-PCR. total
RNA was rxtracted and reverse-transcribed into cDNA as follows: RNA (0.1-2 pg) was added
to a solution contiiining 3 pl of deionized formamide. 1 u1 of IOX S 1 hybridizütion buffer (4 M
NaCI. 400 rnM Pipes. 10 miM EDTA. pH 6.4) and 10 ng of the reverse primer Mlck EX?,
made iip to a total volume of 10 pl with H20. The RNA solution wüs then heated to 85OC for 5
min and allowed to cool slowly to 3PC. Single stranded cDNA was made through reverse
transcription of the RNNhybrid by adding it to a mixture containing 50 rnM TRIS-HCl pH
8.3. 8 miM MgCl?. 10 mM DTT, 0.125 mM dNTPs, 7U RNASE inhibitor (Pharmacia. Baie
d'Urfe. Que.) and 12.5U AMV reverse transcriptase (Boehringer Mannheim, Laval. Que.). in
ri total volume of 100 pl H20. This mixture was incubated at 42°C for 1 hr and inactivated at
85OC for 10 min. PCR was performed using rnurine 1cK sense primers locatrd in exon la
(Mlck II: AGC ATC ATG TGA ATA GGC CAG AAG GCT CCC) and exon 1 b (Mlck 1; AGA
ACA GGG CTC TAG GAT GTC TGA TGT TGG) together with an anti-sense primer located
in exon 2 (Mlck EX?; ATA GGT GAC CAG TGG GTC CCG CAC TTC AGA). Human P
-clcrin sense (CGC CCC GCG AGC ACA GAG CCT CGC CTT TCC C ) and anti-sense
primers (GCG CCC CAC GAT GGA GGG GAA GAC GGC CCG) were used as controls.
Ten microliters of the reverse transcription rnix was then used in standard PCR reactions (620C
annealing) that included 0.5- 1 pl of a-32P dATP 3000 Ci/mmol. PCR products were resolved
on a 7.58 acrylarnide gel. The gel was dried. exposed to a phosphor screen. and quantitation
of the PCR products was performed as described above.
Intrathymic Injections
The details of this procedure have been described rlsewhere (Guidos et al.. 1990). Briefly.
Io hi 1 x 10h control or TCR-stimulated CD5 8 blasts in 10 pl PBS were injected into e x h thyrnic
lobe of anesthetized. unirradiated 4-6 wk-old Thy- 1 congenic host mice. Mice injected with 10
pl PBS per thymic lobe senred as negütive controls. Three days Iüter. thymocyte suspensions
made from individuai thymic lobes were stained with biotinylated 53-2.1. which is specific for
the donor Thy- 1.2 cillele. Cells were washed and resuspended in 1 ml striining medium plus 20
pl avidin-conjugatrd paramagnetic beads ( Advanced magne t ics. Lnc. ). After 30 min at 40C
with constant mixing, cell-bead complexes were removed from free cells using a Bio-MAG
magnetic separator. Enriched ce11 srrspensions were then stained with avidin-Cychrome-5PE
and PE-GK 1.5 and FITC-53-6.7. The results indicated that the frequency of Thy- 1 .I+ cells in
PBS-injected lobes was less than 3%. whereas the frequency of Thy-l.2+ cells in lobes
injected with cells was 7-24%. Data was collected tiom Thy- 1 . 2 ~ gatrd cells for analysis of
CD4 and CD8 expression.
Apoptosis Assays
Cells were stained with PI and DNA content was quantitated on a per ceIl basis using flow
cytometry. as described (Groves et al.. 1995). Apoptotic cells were identifiable aï those with a
subdiploid DNA content. Alternatively, the Apoptag kit (Oncor, Gaithersberg. MD) was used
according to manufacturer's instructions to detect cells in the earfy stages of üpoptosis using the
TUNEL assay (Kishimoto et al.. 1995). For qualitative mdysis of apoptosis. genomic DNA
was analyzed for the presence of nucleosornal-sized fragments of DNA that give the
characteristic Iaddrr appearance. as previously described (Groves et al.. 1995).
II. 3. Results
Purification of DP Thymocyte Subsets
We and othrrs have previously demonstrated thiit cyciinz DP blasts are much more efficient
than post-rnitotic (PM) DP cells at generating SP thymocytes lifter intrathymic transfer (Guidos
et al.. 1989b: Lundbers and Shortman, 1994). Therefore. we evaluated both cycling and PM
DP thymocytes for their responses to TCR engagement in i i t ro throughout this study. We
used centrifuga1 elutriation to separate large cycling blasts from small PM thymocytes and then
isolated DP thymocytes from each elutriated fraction by ce11 sorting. This procedure yielded
99% DP thymocytes that were either 99% PM Gl cells or 8 5 9 5 % SIGZIM blasts upon re-
analysis. To avoid potential artefacts generated by antibody staining and sorting of DP
thymocytes. in several experirnents we used unstained. unfractionated thymocytes MHC ckc~s I
-1- II -1- B6 mice. which are 9045% DP (Crump et al.. 1993: Figure II- 1 ).
TCR Engagement of DP Thymocytes In Vitro Fnils to Induce Apoptosis.
Since numerous studies have shown that il? vivo TCR engagement using either C D ~ E or
TCRD specific antibody. or MHCIpeptide ligand causes most DP thymocytes to rapidly
undergo apoptosis (Smith et al.. 1989: Murphy et al.. 1990: Shi et al.. 1991). we decided to
first assess whrther TCR engagement could induce apoptosis in purified DP thymocytes in
vitro. Highly pure (98-99%) blast or PM DP thymocytes were cultured for 1 or 7 days alone.
or in wells that had been pre-coated with purified TCRB or CD3e specific antibody as described
(Groves et al., 1995). DP blasts increased slightly in number, as rxpected of a cycling
population. and very few apoptotic cells were detected (Table 11-1). In contrast. 30-504 of
PM DP thymocytes dicd when cultured overnight in simple media (Table 11-1). Thus,
susceptibility of DP thymocytes to spontaneous apoptosis. which likely removes unselected DP
thyrnocytes in vivo. appears to be quite low when they tirst arise üs blasts and increüses with
tiriie after they cease dividing. This difference may account for the higher efficiency of DP
blasts at generating mature progeny following inirathymic transfer (Guidos et al.. 1989b:
Figure 11-1. DP thymocytes decrease expression of both CD4 and CD8 in response to TCR
ligation. Thymocytes from MHC class I - / - I I 4- rnice were cultured in duplicate for 16 hr
with 10 @ml of immobilized a n t i - C D ~ E (2C l l ) or anti-TCRp ( M i r ) , or 10-6 M
dexamethasone (Dex). Cultures were then split into two aliquots and stained by TUNEL, to
identiQ apoptotic cells. or with anti-CD4-PE and anti-CD8-RTC. Numbers in the left panel
indicate the percentage of apoptotic cells in each culture. Fifty to seventy percent of DP
thymocytes became CD4lo~D8lo in response to anti-CD~E or anti-TCRP treatment. (ND, not
determined).
Table 11-1: Failure of TCR engagement to induce death of DP Thymocytes
Ir, Viable Ce11 Recovery (% Apoptotic CeIls)'
Day 1 Day 2
DP Blsts Media
TCRp
CD&
TCRpICD4
TCRP+LFA- 1
TCRp+CD28
Ionornycin
Ionornycin
M C 1 Media
TCRp
C D ~ E
lL The indicated subsets of soned DP thymocytes or unfractionated, unstained thymocytes from
MHC ckiss I -1-11 -1- mice were cultured in media alone. or in wells that had been coated with
the indicated antibodies.
b TCRPICD4 refers to a chernical heteroconjugate of H57-597 and GK 1.5 which was used at
10 g i rn l . For the dual antibody combinations. each antibody was used at 10 @ml.
After 1 day ( 15-20 hours) or 2 days of culture. cells were counted and compÿred to the input
cell number (5% viable recovery). In addition. cells were analyzed by PI staining and flow
cytometry to determine the frequency of apoptotic cells in each culture. Note that the frequency
of apoptotic cells is inversely proportional to the % viable recoveiy.
d ND, not determined.
Lundberg and Shortman, 1994).
TCRP or C D ~ E engagement of DP thymocytes did not decrease the recovery of viable cells.
nor incresise the tiequency of üpoptotic cells after 1 day of culture (Table II- 1). We dso
exarnined cells after 3 days of culture, but did not observe TCRP- or CD3e-induced death of
DP thymocytes (Table II- 1). No TCRP- or CD3einduced DP thymocyte death was observed
after 1 or 2 days culture of unfractionated MHC clr.~s I -/-II -I- thymocytes, ruling out the
possibility thiit the antibodies used for soning DP thymocytes interfered with TCR-rnediated
death in virro. In addition, it is notewonhy that the rate of spontaneous ceIl death seen with
unmanipulated DP thy mocytrs was similar to that seen with sorted DP thymocy tes. Thus, the
high rate of DP ceIl death NI virro was not an artefact of the isolation procedure. Interestingly.
when we carried out similar experiments with unstained. unfractionated thyrnocytes from
normal 86 mice. we obsrrved a preferential TCR-induced loss of CD4 and CD8 SP
thymocytes (Table 11-2). Thus. in contrast to immature DP thymocytes. mature SP thymocytes
are susceptible to TCR-induced deüth in virro.
The same antibody prrparations which failed to induce death of DP thymocytes in vitro
induced substiintial loss of DP thymocytes and apoptosis 10 hr after intraperitoneal injection
into mice (Table 11-3). This striking difference bctween Nt v i vo versus itl rritro TCR
engagement çould indicate that clonal deletion in vivo ülso involves the participation of non-
TCWligand interactions. such as LFA- 1-mediüted adhesion or CD28-medinted CO-stimulation
(Carlow et al.. 1992: Page et al.. 1993; Punt et al.. 1994). Howrver. we found that
simultaneous engagement of TCRP and CD28 or TCRP and LFA- 1 did not increase üpoptosis
in purified DP blasts or PM thymocytes relative to the spontaneous levels in control cultures
with media (Table II- 1 ). Sirnilar results were obtained when TCRP and CD4 were coengaged
by an anti-TCRplanti-CD4 heteroconjugate (Table 11- 1 ). Thus. engaging the TCR ülone or in
combination with other surface receptors did not increase the frequency of DP rhymocytes
dying in culture.
Table 11-2: Anti-TCRp Induces Death of CD4 and CD8 SP Thymocytes In Vitro
SubsetLf No. of Crlis ~ecoveredl~
(X 10-6)
Mediri Anti-TCRB
Unfiactionrited B6 thymocytes were cultured overnight in media alone, or in wells that had
been coated with rinti-TCRP.
b Cells werc counted. stained with anti-CD4-PE and anti-CD8-FITC and ünülyzed by flow
cytornetry. The psrcentage of each thymocyte subset was multipiitrd by the total number of
thymocytes in each culture to obtain the ribsolute number of thymocyres in erich subset.
Table 11-3: Anti-CD& lnduces Death of DP Thymocytes In Vivo
No. DP ~ h ~ r n o c ~ t e s ~
Treatmen tc' Dose (millions)
Hamster IgG 500 pg 142, 147
y Irradiation 500 cGy 12, 9
l l Six wrek-old B6 mice were irradiated with 500 centigray (cGy) frorn ü ' 3 ' ~ s source, or
injected intraperitoneally with the indicated antibodies. The lower e ffectiveness of anti-ICRP
relative to anti-CD3e at inducing deletion N i vivo has been previously noted (Shi et al.. 199 1).
Twenty hours after treatment of mice with irradiation or antibodies. thymocytes were
counted. stained with anti-CD4-PE and ünti-CD8-FITC. and anüly zed by tlow cytometry. The
percentage of DP thyrnocytes wüs multipiied by the total nurnber of thyrnocytes in each sample
to obtain the absolute number of DP thymocytes in each thymus. Two micelgroup were
analyzed and duplicate values are shown. Agarose gel electrophoresis of DNA isolated from
each sample showed that nucleosomai-sized DNA fragments were prominent only in the 500
cGy and anti-CD3e (500 pg) treated groups (data not shown), indicüting the presence of
apoptotic cells.
In Vitro TCR Engagement of DP Thymocytes Induces the Development of
Non- Apoptotic ~ ~ 4 1 0 ~ ~ 8 1 0 Cells
ln agreement with previous studies (Groves et al.. 1995: Keÿrse et al., 1995). we found
that overnight TCRP stimulation reduced surface expression of both CD4 and CD8 by 5- 10
fold on 50-7096 of the cells (Figure II- 1). C D 4 1 0 ~ ~ 8 l o thymocytes which are generated by
overnight culture of DP thymocytes in simple media have been shown to contain apoptotic cells
(Swat et al.. 199 1). Therefore. we quantitated apoptotic cells from these same cultures using
the highly sensitive TUNEL technique, which has been shown to be 4-10 times more sensitive
than PI staining at detecting apoptotic thymocytes after 16 hours of culture (Kishimoto et al.,
1995). However, the frequency of apoptotic cells detected by TUNEL was not increased by
TCRp or CD& engagement (Figure 11-1 ). The frequency of C D ~ ~ ~ C D ~ ~ ~ ceils increased to 80-
90% in TCRKD3-stimulated cultures after 2 days but was only 5- [O% in unstimulated cultures
(data not shown). Despite this dramatic TCR-induced difference in the frequency of
~ ~ 4 1 ~ ~ ~ 8 1 0 cells. similar numbers of viable cells were recovered fiom control and stimulated
cultures even at this later tirnepoint (Table II- 1 ). Collectively. these data demonstrate that the
C D ~ ~ ~ C D S I O thymocytes which were generated in response to TCR engagement did not
represent DP thymocytes undergoing TCR-induced clonal deletion in vitro. To address this
possibility further, we cmied out several additional experiments.
First. we evaluated purified C D ~ ~ T D ~ ] ~ thymocytes quditatively and quantitatively for the
presence of apoptotic cells. Consistent with the experiments reported by Swat et al. (Swat et
al., 1991). we found that ~ ~ 4 1 0 ~ ~ 8 1 0 cells arose spontaneously after overnight culture of
normal or anri-H-Y TCR transgenic B6 thymocytes (Figure II-2A. population 3) . However,
these cells were readily stained by PI (Figure II-2A). indicating that they were non-viable and
did not have intact membranes at the time of staining with CD4 and CD8 antibodies. Thus.
they cannot be classified as ~ ~ 4 1 0 ~ ~ 8 l o with any degree of confidence, since non-viable cells
lacking intact membranes are known to bind üntibodies non-specifically. In agreement with
Swat et al.. we found thüt the spontaneously arising ~ ~ 4 l o ~ ~ 8 l o population is clearly enriched
Control 0'1)
Figure II-2A. TCR-induced down-regulation of coreceptor expression by DP thymocytes is
not correlated with the induction of apoptosis. Freshly isolüted thymocytes from fernale H-2b
(B6) and female H-zb anti-H-Y TCR transgenic (HY) mice were cultured overnight alone or
with immobilized anti-TCRP. Cells were then stained with antibodies specific for CD4 (PE-
GKl.5) and CD8u (FITC-53-6.7). In the CD4 vs CD8 plots (A). boxes indicate gates used to
quantitate and sort ~ ~ 4 ~ ~ 8 ~ ~ and ~ ~ 4 ~ ~ 8 " populations. The frequency of cells in each
population is shown for each plot. Note that in the absence of PI. dead cells lacking intact
membranes stain nonspecifically with low levels of both antibodies and thus 14- 17% of control
(not stirnulated) thyrnocytes are classed as ~ ~ 4 ~ ~ 8 " gate (top. population 2). However.
inclusion of PI in the staining media shifts most of these cells out of this gate (middle,
population 4)- since PI fluorescence is brighter than PE fluorescence. Thus, the high
fluorescence signals in the PE channel of the CD4 vs CD8 plot represent PI-stained cells. The
higher frequency of PI-stained cells in the TCR-stimulated cultures reflects TCR-induced death
of CD4 SP and CD8 SP cells, but not death of DP thymocytes, as shown in Table II.
in apoptotic cells, as demonstrated by the abundance of nucleosornal DNA fragments and a
hjgh percentüge of cells (50-708) with subdiploid DNA content (Figure II-2B). In addition to
the non-viable CD41°CD810 thymocytes that arose spontaneously, we also observed a large
number of viable, PI-excluding CD410CD81° thymocytes thüt ÿrose in response to ovemight
TCR engagement of B6 or ami-H-Y TCR transgenic thymocytes (Figure [MA, population 6) .
In contrast to the spontaneously uising C D ~ ~ O C D ~ ' ~ thymocytes, the viable, TCR-induced
~ ~ 4 i o C D 8 1 0 population, as well as the CD4hiCD8hi cells (population 5). did not contain
visible nucleosomai DNA fragments and very few cells had a subdiploid DNA content (Figure
11-23). These results show that CD41°~~810 thymocytes which devrlop in response to TCR
engagement are viable. non-apoptotic cells.
Although TCR-induced ~ ~ 4 l ~ C D 8 l ~ thymocytes did not display apoptotic chûracteristics, it
remained possible that these cells required a second signal before succumbinp to apoptosis
(Page et al., 1993). To address this possibility. we used the intrathyrnic adoptive transfer
assay to assess the developmental potential of C D ~ ~ ~ C D ~ ~ ~ thymocytes generated by ovemight
TCR engagement of purified DP thymocytes. Three days after intrathymic transfer of control
DP blasts, 3.1 + 0.9% (n=3) the nurnber of input donor cells were recovered, as compared to
2.6'1 0.3% (n=3) for TCR-stirnulated DP blasts. In addition, TCR-stimulated DP blasts
generated sirnilar numbers of CD4 and CD8 SP as control DP blasts after intrathymic tram fer
(Figure II-2C). Interestingly. C D ~ ~ ~ C D ~ ~ Q thymocytes were observed among the progeny of
both control and TCR-stimulated DP thymocytes. although ü t much higher frequency in the
latter case. These observations indicate that C D ~ ~ ~ C D ~ ~ U thyrnocytes generated in response to
in virro TCR engagement can survive for at least 2-3 days when returned to the thymic
microenvironment, and at least some of these cells c m develop into CD4 or CD8 SP.
Strain: Pop.:
Figure II-ZB. TCR-induced down-regulation of coreceptor expression by DP thymocytes is
not correlated with the induction of apoptosis. Agarose gel electrophoresis of DNA was
perforrned using total DNA isolated from 6x 1 O' BO s r anti-H-Y TCR transgrnic (HY) cefls of
gated populations 1, 7 . 5. and 6 from the CD4 vs CD8 plots (L=DNA ladder: F=fresh cells;
lo=ionomycin-treated cells). The percentage of apoptotic cells was determined using a flow
cyrometrk assay to quantitate cells with subdiploid DNA content.
CDS - Figure II-2C. Developmental potential of TCR-stirnulated DP blasts. Purified DP blasts
were cultured for 1 day alone or with 10 @ml of immobilized anti-TCRp. Cells were
harvested and injected intrathyrnically into unirradiated Thy- 1 congenic rnice ( I x lo6/lobe). The
CD4 vs CD8 profiles are shown for donor-denved (Thy- 1 .I+) cells in 2 different thymic lobes
frorn conuol (upper) or TCR-stirnulated DP blasts (bottom) 3 days after intrathymic injection.
Percentages in each quadrant are indicated.
TCR Engagement of DP Thymocytes Induces Developmental Progression In
Vitro
We next sought to determine whether aspects of the maturation and survival programs
triggered by positive selection in vivo could be induced by in vitro TCR engagement of DP
thymocytes in short-terrn suspension cultures. After culturing blast or PM DP thymocytes for
1 or 2 days alone. or in wells that had been pre-coated with purified TCRP sprcitïc antibody,
Row cytometry was iised to quantitate surface expression of T cell maturation markers known
to increase ( C D 5 MHC class 1. CD45-RB) or decrease (HSA. Thy- 1 ) as a result of positive
selection itt iivu. We did not observe TCR-stirnulated changes in expression of HSA. Thy- 1,
MHC class I . or CD45 (Table 11-4). However. K R engagement induced CD5 expression in
approximately 50% and 7 5 6 of PM DP cells after one day or two days of culture. respectively
(Figure 11-3). This observation is similar to a previous study of unfractionated DP thymocytes
(Kearse et al.. 1995). which contain 85-958 PM cells. A hallmark of KR-mediated signal
transduction in mature T cells is rapid expression of the CD69 activation molecule. In contrast
to another report (Swat et al.. 1993). we found that PM DP thyrnocytes also respond to TCR
cross-linking by rapidly inducing expression of CD69 (Figure 11-3. bottom). The blast subset
also increased CD5 and induced CD69 expression in response to TCR engagement, but with
slower kinetics thün PM DP thymocytes (Figure 11-3). This mliy indicate that DP thyrnocytes
must complete DNA synthesis and ce11 division before becoming responsive to TCR signais.
Finally, we found that TCR engagement of DP thyrnocytes increased expression of the Bcl-2
survival protein by 2-4 fold, especially after 2 days of culture (Figure [MA, top). These data
suggest that iti vitro TCR engagement of DP thymocytes induces selective developmentd
changes associüted with maturation and survival.
Table 11-4: Surface Phenotype of DP Thymocytes After TCR Engagement in
Vitro
Stimulation Mean Fluorescence Intensity oka
Thv- 1 HSA CD45-RA CD45-RB H-2K CD5
Media 933 213 3 L 63 3 62
TCRB 708 390 3 164 3 247
TCRB/CD4 ND^ ND 4 195 5 1014
a After stimulation of PM DP thymocytes for 2 days with media. anti-TCRP. or anti-TCRpIanti-
CD4 heteroconjugate. expression of the indicated ce11 surface proteins was evaluated by flow
cytometry.
b ND. not deterrnined.
BLASTS PM
1 Day
2 Days
1 Day
2 Days
Figure 11-3. Purified blast and PM DP thymocytes express CD5 and CD69 in response to
TCR ligation. The indicated subsets were cultured for I or 2 days alone or with 10 pglml of
immobilized anti-TCRp. and then stained with antibodies specific for CD5 and CD69. For
each plot. staining of unstimulated (shaded histograrns) and stimulated (clear histograms) cells
is compared. Negative controls (unstained cells) has mean t'uorescence intensities (MFI) of 3-
5 (not shown). Values in brackets indicate MF1 of TCR-stimulated cells with induced CD5 or
CD69 expression.
IO IOO 1808 10 108 1080 IO 100 1~ IO 109 1080
CDS - Bcl-2 - Figure 11-4 A. CD4-mediated signals enhance TCR-induced expression of CD5 or Bcl-2 in
DP thymocytes. Sorted PM DP thymocytes were cultured for 1 or 2 days alone, or with 10
@ml of imrnobilized anti-TCRP or a heteroconjugate of anti-TCRphnti-CD4. Cells were then
analyzed for surface CD5 expression or intracellulx Bcl-7 expression. Staining of control
(shaded histograms) and stimulated (dark outlinrd histogrürns) cells is compared to staining
with isotypr-matched control antibodies (clear histograms). Values in brackets indicate MF1 of
TCR-stimulated cells with induced CD5 or Bcl-2 expression.
CD4 or CD28-Mediated Signals Enhance TCR-Induced Maturation of DP
Thymocytes In Vitro
Positive selection in vivo likely involves the participation of several receptor/ligand
interactions, in addition to TCR engagement by MHUpeptidr. In pürticuliir. previous studies
have shown that coengagement of the CD4 or CD8 coreceptors with the TCR substantially
improves early TCR signaling events in DP thymocytes (relative to TCR signaling alone) as
measured by protein tyrosine phosphorylation and mobilization of intmcellular calcium stores
(Gilliland et al.. 199 1: Turka et al., 199 la). This improvement is ihought to reflect recruitmeni
of the coreceptor-associated Lck protein tyrosine kinase to the TCR complex. Additionally, DP
thymocytes express high levels of CDX. which recognizes members of the 87 sene family
and activates ü KR-independent signal transduction pathway that provides important
costimulatory si_ennls to mature T cells (Gross et al., 1992). Therefore. we assessed whether
CD3 or CD28 signals could synergize in vitro with TCR signals to increase CD5 or Bcl-2
expression by DP thymocytes.
Coengagement of TCRICD4 induced more cells to express CD5, and induced higher levels
of CDS per cell. than did TCR engagement alone (Figure I I 4 A and Table 11-3). Similady.
TCRICD4 coengügrment significantly enhanced induction of Bcl-2 protein. relative to TCR
engagement atone (Figure II-?A). Engagement of CD4 or CD8 alone hiled to induce CD5
expression in DP çells (data not shown). The costimulatory potential of the CD28 signaling
pathway in DP thymocytes wüs evaluated by quantitating CDS expression after culture in wells
coated with either anti-CD28, a sub-optimal dose of anti-TCRp (0.25 p g ) . or with a mixture of
sub-optimal anti-TCRP and anti-CD28. The results showed that anti-CD28 could synergize
with ü sub-optimal concentration of mti-TCRP to induce CD5 expression in both blast and PM
DP thymocytes (Figure W B ) .
BLASTS Stimulation:
Figure II4B. CD28-mediüted signals enhance TCR-induced expression of CD5 in DP
thymocytes. Purified DP subsets were cultured for 1 day in media, or with 2.5 pg or 0.25 pg
of immobilized anti-TCRP, 12.5 pg anti-CD28, or 0.25 pg anti-TCRp plus 12.5 pg anti-CD28.
Cells were harvested and then stained with antibodies specific for CD5 or CD69. Negative
controls (unstained cells) have mean fluorescence intensities of 3-5 (not shown).
In Vitro TCR Engagement of DP Thymocytes Induces Additional Outcornes
Associated with Positive Selection In Vivo
Nonhern blot analyses show that M G 1 expression is rapidly diminished in response to
TCR engagement of DP thyrnocytes and clonal DP ce11 lines in vitro (Turka et al.. 1991b;
Groves et al.. 1995; Kearse et al., 1995). To facilitate analyses of small ceIl numbers, we used
a semi-quantitative RT-PCR strategy to determine whether changes in gene expression that
occur as a result of positive selection in vivo are inducible by short-term TCR engagement of
purified DP thymocytes in suspension culture. This RT-PCR technique readily demonstrated a
dramatic and specific decrease in RAGl expression following ovemight TCR engagement of
DP thyrnocytes (Figure [MA) . We then extended our analysis to expression of the pre-Ta
gene, which continues to be expressed in TCR@ DP thymocytes. though it is thought to
function primarily in immature DN precursors of DP thymocytes (Saint-Ruf et al.. 1994;
Fehling et al., 1995). The RT-PCR assay showed that pre-Tu expression in DP thymocytes is
barely detectable üfter TCR engagement (Figure 1133). Northern b1ot analyses of the VL3-
3M2 immature DP ceIl line showed a similar TCR-induced reduction in pre-Tu expression
(Figure II-5C). Thus, in vitro TCR engagement of DP thymocytes and an immonalized DP
ce11 line simultaneously decreases R A G and pre-Ta expression, and increases TCRa
expression (Groves et al., 1995; Kearse et al., 1995).
Another hallmark of T ce11 development is the switch in lck promoter usage. In immature
DN and DP thyrnocytes, the lck gene is transcribed using both the proximal and distal
promoters, whereas in mature peripheral T cells, the distal promoter is preferentially used
(Reynolds et al.. 1990; Wildin et al., 199 1 ). We used RT-PCR to evaluate whether TCR
engagement of DP thymocytes altered the relative abundance of proximal (type 1) and distal
(type 11) transcripts. In accordance with other studies using RNAse protection (Reynolds et
al.. 1990). the RT-PCR technique demonstrates that the ratio of type 1 relative to type 11
transcripts is low (0.2-0.4) in mature CD4 SP thymocytes and lymph node T cells (Figure 11-6,
lanes 2 and 3), whereas this ratio is high (2.1-2.6) in unstimulated DP thymocytes (lanes 4 and
A
cDNA dilution:
RAG-1
p-actin
Control TCR
Thymocytes
Stimulation: None TCR- P None -
cDNA dilution:
Figure II-SA and B. DP thymocytes terminate expression of RAGl and pre-Ta in
response to TCR ligation. DP blasts were cultured overnight alone (control) or with 10 pg/ml
of immobilized anti-TCM. Total RNA was extracted, reverse-transcnbed into cDNA, and 5
fdd serial dilutions of the cDNA were subjected to PCR amplifications with RAG1 and #?-actin
-specific primers in the same reaction (A), or with pre-Ta and $-actin -speci fic primers in
separate reactions (B). Total scid thymocytes were used as a positive control for pre-Ta
mRNA. PCR products were resolved on 1.5% agarose gels, transferred to nylon membranes,
and probed with 32-~-labelled fragments of RAGI, pre-Ta, or pactin DNA.
Stimulation: - + -
pre-Ta
HPRT
Figure II-SC. Northem analysis was performed to anaIyze pre-Ta expression in total
thymocytes and in VU-3M2 ceils culnired overnight alone (-), or with 10 pglml of
immobilized anti-TCRg (+). The blot was stripped and re-probed with HPRT to venfy equal
loading of RNA.
BLASTS C S
&ilIl 2.6 0.3
kiJ
Figure 11-6. TCR ligation drcrs [tic r;iiio k i t ' 1c.k r>*pc 1 mit tupi: I I triinscript.. in DP cclls.
RSX ~ i h c.'irsactcci l'ro~ii wi-rcd DP hliisth ;inci D P Ph1 cclls c~iiti~rcd i i~ r I da'* ( P M or 2 days
i h l ~ h 1 alonc ( C 1. or \ t iiiiiilatcd wirli IO ug/iiil O t. anii-TCRp ( S 1. R N A \vii> a l w ohtaincd Inmi
i i n l r ~ i c t i n t c thy nioc!.ie\. \oricd CD-!+ SP ttiyiiioq.ies. and CD-!+ l y t i p h nodi: T cclls. RNA
i t a i-c\.crsc-tr;in\cribc~i into CDS.\ ;incl r~ ih jcc id t o PCR an;il>.si\ u.itti Ii.k wnse prirners
1 7 i r i r i i ~ c i i : c l \';iluc\ iiidic;itc itic ratio 01' tlic ri:l;iiivc iiniiii inih ol' i 1 . p ~ I ;inil iypc I I
iranxripts a\ ~lctcriiiincii witti ;i h~lolcciilar D>.naiiiicr plio\plioriiii;igcr.
6). However. overnight TCR engagement of DP thymocytes caused the ratio of type Ytype 11
transcripts to FaIl to 0.3-0.8 (lanes 5 and 7). which is similar to the ratio for CD4 SP
thymocytes (lane 3 ).
II. 4. Discussion
The earliest observable outcomes of positive selection in vivo include upregulation of
surface TCR, CD5. and CD69 (Guidos et al., 1990; Hugo et al.. 1990; Ohashi et al., 1990;
Shortman et al., 199 1 ; Bendelac et al., 1992; Guidos and Weissman, 1993). decreased
expression of RAGl and ferminal dm-yuicieotide trclnsfrrcrsr ( TdT ; Borgulya et ai., 1992;
Brandie et al., 1992; Briindle et al., 1994; Kouskoff et al., 1995), and increased Bcl-2
expression (Linette et al., 1994). Expenments reported here and previously show that in vitro
TCR engagement decreases RAG and TdT expression and increases CD5 and TCRa
expression in DP thyrnocytes and ce11 lines (Turka et al.. 199 1 b; Groves et al.. 1995; Kearse et
al.. 1995). We have extended these observations to show that TCR ligation of DP thyrnocytes
also increases Bcl-2 expression (Figure 11-4). terminates pre-Ta gene expression (Figure II-
5B). and causes a switch in lck prornoter usage (Figure 11-6). In addition, we showed ihat
CD4- or CD28-mediated signais enhance TCR-induced maturation and Bcl-2 expression in DP
thymocytes (Figure 11-4; Tables II- 1 and 11-4). Coreceptor engagement also enhances early
outcomes of positive selection in vivo (Chan et al.. 1994c: Dutz et al.. 1995). Interestingly,
our results indicütç thüt the response of DP thyrnocytes to TCR engagagrnent iri vitro is
selective. sincr changes in expression of Thy- 1. HSA, MHC class 1, and CD45-RB were not
induced (Table 11-4). One possible explanation for this selective response is that TCR
engagement may be sufficient to induce some outcomes, but others may require additional.
non-TCR-mediated signds trom the thyrnic microenvironment. We cannot rule out this
possibility, but it is interesting to note that the outcornes which are not TCR-inducible in vitro
are known to occur well after DP thymocytes have upregulated TCR, CD5 and CD69 in vivo
(Lucas et al., 1993; Lucas et al.. 1994). Thus, an alternative explanation is that the later
outcomes become TCR-inducible only after the earIy maturational events have occurred.
CoIlectively, these observations are consistent with the notion that TCR engagement of DP
thymocytes, in the absence of other signals from the thymic microenvironment, is sufficient to
induce many early outcomes of positive selection in vitro, wi tliout inducing apoptosis.
Despite multiple lines of evidence from in vivo studies suggesting that ligation of the
TCRKD3 cornplex triggers apoptosis of DP thymocytes, we found that TCR or TCWCD4
engagement significantly increased expression of the Bcl-2 suwival protein. and did not induce
apoptosis in blast or PM DP thymocytes (Table II- 1). Despite the increased Bcl-2 expression.
survival of DP thymocytes was not improved (relative to unstimulated cells) by these
treatments. This may indicate that the arnount of Bcl-2 protein was insufficient. or that the
expression of other survival and death genes was not appropriately modulated by these
experimental manipulations. Singer's group has also reported that TCR engagement fails to
induce apoptosis of DP thymocytes in virro (Punt et al., 1994; Keürse et al., 1995). but in
contrast to these investigators (Punt et al.. 1994). we did not find that CD28 costirnulation
revealrd a TCR-inducible death pathway in DP thymocytes (Table 11-2). Although the reason
for this discrepancy is not clear, other studies have shown that CD28/B7 interactions are not
required for clona1 deletion mediüted by MHClpeptide on antigen-presenting cells in vivo or in
vitro (Tan et al., 1992; Jones et ai., 1993; Page et al., 1993; Shahinian et al.. 1993; Wdunas
et al.. 1996). Page et al. have provided evidence that a CD28/B7-independent costirnulatory
signal is required to observe TCR-induced death of DP thymocytes in vitro (Page et al.,
1993), but the nature of this signal remains obscure.
The inability of antibodies to the TCRKD3 cornplex to induce apoptosis of DP thymocytes
in vitro contrasts sharply with the rapid and massive death induced by these antibodies in vivo
(Table 11-3; Smith et al., 1989: Shi et al.. 1991). Removal of DP thymocytes from the thyrnic
microenvironment is known to alter the expression and function of the K R on DP thymocytes
(Nakayarna et al., 1989; Nakayama et al.. 1990; Nakayama et ai., 199 1 ), which could explain
our failure to observe TCR-induced DP ce11 death in vitro. However, i t is not c l ex that DP
thyrnocyte death in vivo is a direct consequence of TCR-mediated signaling in DP thymocytes.
Rather than inducing death of DP thyrnocytes, in vivo administration of TCR/CD3 specific
antibodies may inrerfere with the generation of DP thymocytes (Takahama et al., 1992). since
this process is known to depend on TCRICD3 signals (Levelt and Eichmann, 1995). Under
some circumstances, addition of TCR or CD3 specific antibodies to fetai thymus organ cultures
fails to induce deletion (Finkel et al., 1992; Levelt et al., 1993a). and may even induce positive
selection (Takahama et al., 1994). Thus, death of DP thymocytes is not an inevitable
consequence of treatmsnt with antibodies to the TCRKD3 complex in vivo.
The observation that DP thymocytes decrease expression of both coreceptors in response to
TCR engagement (Figure II- 1 : Swat et al.. 199 1; Kelirse et al., 1995) is not easily reconciled
with either stochastic or instructional models of CD4/CD8 lineage commitment (Robey and
Fowlkes. 1994). Others have shown that coengagernent of TCR and CD8 by presentation of
appropriate MHC class Upeptide complexes on an tigen-presen ting cells induces TCR
transgenic DP thymocytes to down-regulate both CD4 and CD8 (Swat et al., 199 1; Curnow et
al., 1994). The ~ ~ 4 1 ~ ~ ~ 8 1 0 thymocytes generated in response to TCR specific antibody
(Figures 11-1 and 11-2) or MHC/peptide on antigen-presenting cells (Cumow et al., 1994) are
not apoptotic, and the former can survive and differentiate in vivo as well as control DP
thymocytes pre-cultured in media alone (Figure II-2C). Thus, decreased expression of CD4
and CD8 does not universally signify a commitment to die. The signitïcance of coreceptor
downregulation in response to TCR engagement remains to be estabiished. but a recent study
suggested that ~ ~ 4 1 o C D 8 1 0 thymocytes may represent a previously unrecognized
developrnental intermediate stage during the DP to SP transition (Suzuki et ai., 1995).
Our results suggest tliat lineage cornmitment. as signified by selective shut-off of CD4 or
CDS, is not an immediate consequence of TCR or TCRKD4 signaling in DP thymocytes
(Figure 11-1). This may indicate that TCR signals are necessary, but not sufficient, to induce
CD4/CD8 iineüge cornmitment. This notion is consistent with studies dernonstrating that, in
addition to presentation of MHClpeptide ligands, positive selection in vitro requires non-TCR-
mediated signals provided by thymic epithelial cells (Anderson et al., 19941; Poirier et al.,
1994). However. it is not clear whether the unique signais delivered by thymic epithelium are
essential for lineage commitment itself, for survivai of lineage-committed precursors, or for
both. An alternative possibility is that TCR signaling plays no requisite role in the selective
loss of CD4 or CD8 from DP thymocytes. This idea conflicts with the recent suggestion that in
vivo TCR engagement induces DP thymocytes to stochastically shut-off CD4 or CD8 and
develop into coreceptor transitional ( c D ~ [ o ~ + or ~ ~ 4 ~ 2 3 ~ 0 ) thymocytes (Chan et al.. 1993; van
Meenvijk and Germain. 1993). However, mice deficient in both MHC cltss I and clnss II,
which lack any potential TCRaP ligands. also develop CD4108+ or CD4+8I0 transitional
thymocytes. but these cells are largely T C R ~ ~ / C D ~ ~ - (Crump et al.. 1993). In addition,
Fischer et al. have recently showed that the absence of the Vüv signaling molecule severely
impairs TCR signaling and early maturational events. such as upregulation of TCR and CD5,
in DP thymocytes. but does not prevent the developrnent of CD4+ 8'0 transitionai thymocytes in
vivo (Fischer et al.. 1995). Collectively these results suggest that TCR signaling is essential
to induce early maturational events in vivo. but is dispensible for development of a coreceptor
transitionai phenotype. Alrhough increased expression of TCR m d CD5. as well as decreased
expression of RAG genes. occur prior to or coincidently with downregulation of CD4 or CD8
in vivo, we suggest that these events are elicited by distinct mechanisms. The TCR-induced
events may be essential for the early maturation and subsequent positive selection of lineage-
committed thymocytes that arise stochastically. Elevüted CD5 expression rnay be particularfy
important in this regard, since this molecule wüs recently shown to influence TCR signaling
during positive selection (Tarakhovsky et al., 1995).
Our observations support conclusions from in vivo systems suggesting that multiple,
temporally separated TCR engagements are required to effect the entire spectrum of
developrnental changes associated with positive selection. Furthermore. our results suggest
that the multiple outcornes of this developmental process may be independently regulated. The
in vitro approach WC describe here offers a means of resolving individual steps and signals in
the cornplex developrnental process that govems the DP to SP transition in vivo.
Chapter III
Fyn Can Partially Substitute For Lck in T Lymphocyte
Development
Tim Groves 1, Patricia Smiley 196, Michael P. Cooke 2.7,
Katherine Forbush 375, Roger M. Perlmutter 2,33495, and Cynthia J.
Guidos1
l Division of Imrnunology and Cancer, Hospital for Sick Children and
Research Institute, 555 University Avenue, Toronto, Ontario, Canada MSG
1x8
and Department of Irnmunology, University of Toronto, Toronto, Ontario,
Canada M5S l A 8
9 - Department of Biochemistry, 3 Department of Immunology, 4 Department of
Medicine, Howard Hughes Medical Institute, University of Washington,
Seattle, Washington 98195-7650
Present Address: Westrninister College, Salt Lake City, Utah 84105
7 Present Address: Systemix, 3155 Porter Drive, Pa10 Alto, California 94304
The contents of this Chapter were published in Immunity 5, 417-428. (1996)
III. 1. Introduction
Two structurally distinct TCRP-containing receptor complexes regulate T lymphocyte
development. The pre-TCR complex, thought to consist of pre-Ta. TCRP. and CD3 proteins.
is expressed on CDX+ DN thyrnocytes (Levelt and Eichmann, 1995). In mutant mice lacking
RAGI or RAGZ (Mombaerts et ai., 1992b; Shinkai et al.. 1992), TCRP (Mombaerts et al.,
1992a). pre-Ta (Fehling et al., 1995), C D ~ E (Malissen et al., 1995). or TC& (Liu et al..
1993: Love et al.. 1993; Malissen et al., 1993; Ohno et al., 1993). development of DP
thymocytes from CD25+ DN progenitors is diminished or completely abrogated, reducing
thymic cellularity 10- 100 fold. Thus, expression of the pre-TCR complex regulates a critical
checkpoint in thymocyte development. Consistent with this notion, productively r e m n g e d
TCRP transgenes (Kishi et al., 199 1; Mombaerts et al.. 1992b: Shinkai et al.. 1993; Shores et
al.. 1993) or anti-CD3e treatment (Levelt et al., 1993~; Jacobs et al., 1994; Shinkai and Ah,
1994) restore thymic cellularity and the developrnent of DP thymocytes in scid or RAGI - or
RAGZ -deficient mice. Collectively, these results demonstrate that pre-TCR signals stimulate
TCRp+ DN progenitors to proliferate. down-regulate C D 5 and to express the CD4 and CD8
MHC coreceptors. At this stage. TCRa expression begins and TCRnp-mediated recognition of
peptide/MHC ligands expressrd on thymic stroma1 cells selects some DP thymocytes to mature
into CD4+ helper or CD8+ cytotoxic T cells (Guidos. 1996).
Fyn and Lck. two lymphocyte-restricted members of the Src tyrosine kinase farnily. have
both been shown to play important roles in TCRap/CD3-mediated signal transduction in mature
T cells (Anderson et al., 1994b; Weiss and Littman, L994). However. Fyn plays no essential
role in pre-TCR or TCRap-induced maturation. since normal nurnbers of DP and SP
thymocytes are generated i n h n -1- mice (Appleby et al.. 1992; Stein et al.. 1992; Swan et al.,
1995). and expression of a dominant negativebn allele does not impair T ce11 development
(Cooke et al., 199 1) . In contrast. multiple lines ofevidence point to an important role for Lck
at both stages of T ce11 development. First. the ability of TCRP transpnes or injection of anti-
CD3e to restore normal DP thymocyte developrnent to RAGI -'- or RAGZ -/- rnice requires
Lck activity (Mombaerts et al.. 1994; Levelt et al.. 1995; Wu et al., 1996). Second. over-
expression of a constitutively active Ick transgene promotes the development of DP
thymocytes in the absence of TCRp expression (Mombaerts et al., 1994). suggesting that Lck
functions downstream of the pre-TCR complex in regulating the DN to DP transition. In
addition, Lck signais regulate allelic exclusion of TCRp gene reamngemeni in DN thyrnocytes
(Anderson et al., 1992; Anderson et al., 1993). Lck has also been irnplicated in TCRaP-
mediated signaling (McCarthy et al.. 1988; Nakayama et al.. 1989; Nakayama et al., 1990;
Nakayama et al., 1991; Wiest et ai., 1993; van Oers et al.. 1996a; Wiest et al.. 1996), positive
selection (Teh et al., 199 1; Carrera et al., 1992a; Carrera et al., 1992b; van Oers et al., 1992;
Hashimoto et al., 1996), and CD4/CD8 lineage cornmitment (Itano et al.. 1996) of DP
thymocytes. Finally, Lck is required for the development of some subsets of TCRyS+ T cells
(Penninger et al.. 1993; Kawai et al., 1995).
Although these observations define an important role for Lck-mediated signaling events
during development of TCRy6+ and TCRap+ T cells. it is not ciear whether Lck is essentiai for
these processes. Analyses of mice over-expressing a dominant negative I d transgene suggest
an obligate role for Lck in pre-TCR signaling. since development of DP thymocytes and allelic
exclusion of TCRP are both completely abrogated in these mice (Anderson et al.. 1993; Levin
et al., L 993). By contrast, some accumulaiion of DP thymocytes does occur in Ick -deficient
mice created by targeted disrupion of Ick (Molina et al., 1992). Moreover. Ick -1- mice exhibit
efficient allelic exclusion of TCRP (Wallace et al., 1995). Similarly, RAGl -/- Ick 4- mice still
develop small numbers of DP thymocytes in response to expression of a TCRP transgene
(Mombaerts et al., 1994) or anti-CD3~treatrnent (Levelt et al.. 1995; Wu et a[.. 1996).
Finally, srnall numbers of peripheral SP T cells develop in Ick -/- mice (Molina et al., 1992).
arguing that in the absence of Lck, some other signaling rnolecule can subserve its functions in
positive selection of DP thymocytes into the SP lineages.
One obvious candidate for such a signaling element is Fyn. since i t is expressed at
comparable levels to Lck in DN and DP thymocytes (Olszowy et al.. 1995) and is known to
associate with CD3 proteins (Gauen et al., 1992). Although Fyn does not play an essential
role in T ce11 development. we hypothesized that Fyn might play a redundant role in pre-TCR
and/or TCRapKD3 signaling. To address this possiblity, we generated lck -1- rnice harbonng
either recessive loss-of-function or dominant gain-of-functionfvn alleles. Thymic cellularity
in lck -1-fyn -1- mice was reduced 5-10 fold relative to fck -/' mice. and DP thymocytes,
peripheral TCRap+ T cells and TCR$if T cells were virtually undetectable. These data
indicate that Fyn rnediates development of DP thymocytes in ick mice. This notion was
funher supported by Our finding that a constitutively active mutantfvn (7) transgene (TFF )
could completely restore DP thymocyte development in Ick -1- mice. However. in contrast to
the analogous gain-of-function ick transgene (Mornbaens et al.. 1994). the TFF transgene
could not promote efficient generation of DP thymocytes in RAGl '/- mice. Finally, we
provide evidence that Lck is important for signal transduction during the DP to SP transition,
and that this Lck function cm be partidly subserved by thefj~z TFF transgene.
III. 2. Materials and Methods
Mice
Al1 strains were bred and maintained at the Hospital for Sick Children. Mice deficient for
either lck (Molina et al., 1992) orfin (Appleby et al., 1992) were intercrossed to generate
double-deficient animals. Peripherd blood from F2 offspring was iinalyzed by tlow cytometry
and Ick -1- mice were identified based upon a reduction of TCRPf T cells relative to C57BU6
mice. The f jn genotype of these animals was determined by a PCR-based assay using
oenomic tail DNA. A sense primer (S'fin : CAG GTC TCT GCT GCC GCC TAG) from 3
withinfiti (T) exon 7A and an anti-sense primer (3'f,vn : CGA GTC ACG TGC AAC TTC
CCA) from exon 7B amplified a 600 bp fragment of the wild-type fj*n allele. In a second
reaction, the S f i n primer and a 3' neotnycin primer SDL33 (Appleby et al.. 1992) was used
to arnplify a 1.6 kb fragment of the disruptedfvn (T) allele. which contains the nromycin
gene inserted into exon 7B. The amplification cycle ( 1 min at 94OC. 90s at 55OC. and 75s at
72OC) was repeated 37 times in a Perkin-Elmer 480 thermalcycler. PCR products were
resolved on 2% agarose gels containing 0.5 pdml ethidiurn bromide and photographed under
UV light.
A B6-TFF transgenic line (5525) wüs established using standard procedures (Cooke et ai.,
199 1 ) and then crossed to RAGI -1- lck 4- rnice (Wu et al.. 1996). TFF + F 1 progeny were
identified by slot-blot screening of tail DNA with a hurnan growth hormone probe as
previously described (Cooke et al., 199 1 ), and then backcrossed to RAGI -/- Ick -'- mice.
Backcross progeny were then typed for the presence of the TFF trünsgene. and for their
RAGI and Zck genotype as previously described (Wu et al.. 1996).
Antibodies and Flow Cytometry
Fluorochrome-conjugated antibodies specific for the following surface molecules were
prepared using standard techniques: CD4 (GKlS) , CD8a (53-6.7 or YTS 169.4). CD5 (53-
7.3), TCRp (H57-597), C D ~ E (YCD3- 1; 145-2C 1 l ) , and CD44 (IM 178). The remaining
antibodies were purchased: CL3 (TCRfi) and 7D4 (CD25; Pharmingen. San Diego, CA);
1G 10 (phosphotyrosine; Upstate Biotechnology Incorporated, Lake Placid, NY); horseradish
peroxidase (HRP)-conjugüted protein A and HM-conjugüted goat anti-rnouse IgG (Amersharn
Corp. Arlington Heights. IL). Fyn specific polyclonal rabbit antisera were generously
provided by Dr. A. Veillette (McGill University, Montreal. Que).
For flow cytometric analyses. single ce11 suspensions from thymus, lymph nodes, and
erythrocyte-depleted periphenl blood and spleen were prepüred in HBSS containing 2% calf
semm and 10 m M Hepes and stained for surface markers as described previously (Groves et
al., 1995; Guidos et al., 1995). Al1 antibodies were used at saturating concentrations. The
secondary reagents used included avidin-PE (Caltag, San Francisco, CA) and a tandem
conjugate of avidin-Cychrome-5lPE that was prepared by conjugating PE (Molecular Probes,
OR) to Cychrome-5 (Biological Detection Systems Inc.. Pittsburgh. PA) according to
manuhcturer's instructions. Rat [gGZa, rat IgG2b. and hamster IgG isotype control
antibodies conjugated to FITC, biotin, or PE were purchased from Pharmingen and used at 1-2
pg./rnl. Two- or three-color immunotluorescence was anülyzed on a FACScan flow cytometer
with LYSIS [I software (Becton Dickinson & Co., Mountain View, CA) as previously
described (Guidos et al., 1990; Fischer et al., 1995). Dead ceils and debris were excluded
from the analysis on the basis of low forward scatter andlor high propidium iodide
fluorescence.
Purification of DP Thyrnocytes
DP thymocytes from lck 4 - and lck -'- mice were purified using a dual laser FACSTAR
PLUS ce11 sorter (Becton Dickinson). Cells were first stained with saturating concentrations of
FITC-conjugated anti-CDS (53-7.3) and PB-conjugated anti-CD8 (53-6.7). followed by cell
sorting of the C D S ~ O C D ~ ~ ~ population. The sorted cells were 99% DP thymocytes. as assessed
by staining an aliquot of purified cells with saturating concentrations of biotinylated anti-CD4
(GK1.5) followed by avidin-CyYPE. This procedure allowed purification of DP thymocytes
without engaging either the TCR or CD4 molecules. Studies of TCR-induced changes in CDS,
CD69, and RAGI expression were ciirried out as previously described (Fischer et al.. 1995;
Groves et al., 1997).
Western Blot Analyses
Anti-Fyn imrnunoblotting was performed on cells lysed at 5 x 1 0 ~ cellslmi 50 rnM Tris, pH
8.0. 20 rnM EDTA. pH 8.0.20 @l Na3V04,50 miMNaF, 1% NP-40, 20 pg/ml leupeptin and
20 pg/rnl aprotinin. Insoluble material was removed by microcentrifugation at 16,000 rpm for
15 min ai 40C. Post-nuclear supernatants frorn the indicated nurnbers of ce11 equivdents were
resolved on a 8% non-reducing SDS-PAGE gel, transferred to nitrocellulose, probed with anti-
Fyn serum and horseradish peroxidase-conjugated protein A, and then developed using the
Enhanced Cherniluminescence (ECL) detection system (Amersham Corp). For studies of
TCR-induced prorein tyrosine phosphorylation, DP thymocytes purified as described above
were stained at 5x lo7/ml with biotinylated YCD3- 1. £457. GK 1.5. or H57 plus GK1.5 for 20
min at 40C. After washing, cells were resuspended at 1.5x107/ml and pre-warmed to 37OC for
3 min. Biotinylated antibodies were then crosslinked by the additiori of IOpg/ml avidin
(Molecular Probes, Eugene, OR) for 1 min at 370C. The reaction was stopped by addition of 1
ml cold PBS containing 400 pM Na3V04. Cells were lysed as described above for Fyn
immunoblotting, and post-nuclear supernatants were resolved on a 8% reducing SDS-PAGE
gel, transferred to nitrocellulose, and probed with 4G 10 followed by horseradish peroxidase-
conjugated anti-mouse IgG.
III. 3. Results
Profound Failure of T Cell Development in Ick -/-fyn 4- Mice
To determine whether the srndl numbers of DP thymocytes found in lck -/- rnice develop in
a Fyn-dependent manner, we bred lck -'-lvn -/- and lck -/-/yn 4- mice. Consistent with
previous reports (Molina et al., 1992). flow cytometnc analysis of lymph node cells from 3-6
week-old k k -1- mice dernonstrated a 10-20 fold reduction in the frequency and absolute
number of CD4+ and CD8f T cells (Figure III- 1 A and Table III- 1). In contrast,
TCRap+ T cells were reduced by less than 2-fold infin -/- mice (Table III- 1). The nurnber of
lyrnph node TCRysC C D ~ E + cells was similar in k k -/- andlvn -/- mice, and was reduced 2-
fold or less relative to normal 86 mice. However, TCRap+ or TCRy6+ cells were virtually
undetectable in the lymph nodes (Figure III-1A; Table III-1) of the double mutant rnice,
demonstrating that development of both T ce11 lineages requires the participation of either Lck
or Fyn.
Analysis of thyrnocyte number and phenotype suggested that the lack of peripheral
TCRap+ celk in mice deficient for both Ick andfyn derived from a profound and early block
in thyrnocyte maturation. In lck - / ~ w -1- mice. thymic cellularity was 1-56 of normal, and
thyrnocytes were predorninantly large DN cells, with orily 1-38 progressing to the DP stage
(Figure III- I B ). B y contrat, thymic cellularity in k k -/-~5.n +/+ mice was 10-20% of normal,
and JO-60% of the thymocytes were small DP cells (Molina et al., 1992; Figure III- 1 B).
Thyrnocyte nurnber and phenotype were very comparable in I d -/-fyn fi+ and lck -/-fin 4-
mice, revealing thatf i gene dosage had no effect on thymocyte development in the absence
of Lck. The severe paucity of DP thymocytes in the double-mutant mice was accompanied by
an accumulation of immature CD44-/lo CD25+ thyrnocytes. at the expense of their C D 4 4
CD25- progeny (Figure III-IB). Thus, abrogation of Lck and Fyn expression rnediates a
developrnental arrest virtuaiiy identical to that in pre-TCR-deftcient RAGln -/- and scid rnice.
or in rnice expressing a high level of catalytically inactive ick (reviewed in Anderson and
Perlrnutter, 1995; Levelt and Eichmann, 1995).
Table 111-1: Lymph Node T Cell Subsets in Mice Lacking Lck andor Fyn
Genotype No. of T CeIls
lck Lw N TCRap+ (x IOe6) TCRyGf (x 10-5)
+/+ +/+ 3 14.5 t 3.5 3.6 t 0.3
+/+ -1- 4 8.9 + 3.6 1.7 t 0.7
4- +/+ 3 0.9 t O. 1 1.6 2 0.8
-/- -1- 3 ND ND
N. number of individuals per group; ND, not detectable. Peripheral lymph nodes cells from
individual mice of each genotype were counted and stained with antibodies specific for TCRM
(H57) or TCR$ (GL3). The frequency of each subset was rnultiplied by the total ce11 count to
determine the absolute number present in each animal. Values shown represent the mean + SD
for the number of individud mice andyzed in each group.
Figure III-1A. Abrozation of T ce11 developrnent in fck -I-jjn -1- mice.
Absence of TCRup+ and TCkt6+ T cells in the periphery of ick -l-j:\.,z -/- mice. The plots
display two-color analyses of CD3e vs TCRyS. TCRP vs CD4 and TCRP vs CD8 expression
in lymph node çells from 3-6-week old mice of each genotype. Numbers on each plot indicate
the prrcentage of cells in that quadrant. Quadrant markers were set based on staining with
isotype-rnatched control antibodies. Frequencies of <O.% were considered nezative. baed
on cornparison with isotype-matched control stains.
- FSC -
Figure III-1B. Abrogation of T cell development in lck -/-LW --/ mice.
Early block in thymocyte development in lck - '- l)vt -/- mice. Thymocytes from 3-6 week old
rnice of each genotype were stained with the indicated pairs of antibodies and analyzed by two-
coior tlow cytometry. The numbers below each genotype dcsigniition refer to the average
nurnber of thymocytes (+ SD) for that group. Numbers on each plot indicate the percentage of
cells in that quadrant. Quadrant markers were set based on stüining wiih isotype-matched
control antibodies. except for CD4 vs CD8 and CD44 vs CD25 stains, in which the quadrants
dernarcate the prominent DP or CD25+ CD44Io subsets. respectively.
Low levels of TCRP (Figure III- 1B) were detected on the surface of 1575% of lck -/-fin
-1- thymocytes. arguing that the developmental arrest was not due to failure to express a
T C R ~ I C D ~ E - C O ~ taining pre-TCR complex. TCRP+ thymocytes in the double-mutant mice
were predominantly CD25f whereas TCRP+ thyrnocytes in 86 mice were virtually al1 CD25-
(Figure III- L B), suggesting that loss of CD25 from pre-TCRf thy mocytes was substantially
impaired in the absence of Lck and Fyn. CD25- TCRP+ thymocytes developed in lck -/-fin
+/- mice, but this process appeared to be inefficient or slower than normal. since up to 20% of
TCRpf thymocytes were CD25+ (Figure III- 1B). Collectively, these results sugpst that pre-
TCR signaling, which regulates the loss of CD25, acquisition of CD4 and CD8. and clonal
expansion of TCRP+ thyrnocytes, are significantly impaired in the absence of Lck. but are
completely compromised in the absence of both Lck and Fyn. Thus, Fyn can partially
compensate for the absence of Lck in the pre-TCR-rnediated DN to DP trmsition.
Expression of a Gain-of-Function fyn Transgene Restores DP Thymocyte
Development in lck -DePicient Mice
Src family tyrosine kinases contüin a tyrosine residue near the C-terminus which negatively
regulates catalytic activity . Phosphory lation of ty mine 528 negatively regulates Fyn kinase
function. and substitution of this residue with phenylalanine generates a more active kinase
(Davidson et ai.. 1992). Expression of the analogous gain-of-function lck transgene directs
the maturation of DP thymocytes in the absence of TCRP rearrangement (Anderson et al.,
1992; Mombaerts et al., 1994). and this effect is not dependent on over-expression of mutant
relative to endogenous Lck. Thus, Lck functions genetically downstream of the pre-TCR
complex to promote the DN to DP transition. These observations prompted us to examine
whether a gain-of-function f)n transgene could function similarly in RAGl -/- mice, and
whether it could restore normal thymocyte development in Ick -/- mice. A m u r i n e b (Ti
cDNA, referred to as TFF, was engineered to harbor the tyrosine to phenylalanine gain-of-
function mutation at position 528 and was cloned into the pl017 transgene expression cassette.
which contains the vanscriptional control elernents of the Ick proximal promoter (Cooke et al.,
1991). A Bo-TFF transgenic line was established and then crossed to R4Gl -1- lck -1- mice
(Wu et al.. 1996) to generate TFF -expressing mice that lacked either fck or R A G I gene
function.
We first evaiuated expression of Fyn protein in TFF transgenic thymocytes. Transcription
from the Ick proximal promoter is downregulated in SP relative to DP thymocytes (Reynolds
et al.. 1990; Wildin et al., 1991) consistent with the expression patterns of several other
transgenes expressed under control of the proximal lck regulatory elements (Anderson et al.,
1994b). Therefore. we compared Fyn protein levels in TFF transgenic thymocytes to those in
Ick -1- thymocytes (a mixture of DN and DP cells) and MHC ciiss 1 -1-11 -1- (prirnarily DP)
thymocytes. Fyn protein was expressed at 2-4 fold higher levels in TFF transgenic relative to
these non-transgenic strains (Figure 111-2). Because the signal in TFF transgenic thymocytes
derives from both transgenic and endogenous Fyn. these results suggest that mutant Fyn
protein is expressed at 1-2 times that of endogenous wild-type Fyn. Figure III-3A shows that
the TFF transgene c m promote the development of a k w (5 5%) DP thymocytes in RAGl -/-
lck f/- mice, but thymic cellularity was not improved. Thus, in contrast to constitutively
active Lck, constitutively active Fyn does nor efficiently promote the DN to DP transition in the
absence of a TCRP-containing pre-TCR cornplex.
In contrast to the marginal effect of the TFF transgene on thymocyte development in
M G 1 -1- mice. the data in Figure III-3B demonstrate that it has a profound effect in RAGl
ick 4- rnice. In ick -/- mice, thyrnic cellularity is 5 -108 of normal. due to a 10-20 fold
reduction in the production of DP thymocytes (Molina et al.. 1992; Wallace et al.. 1995), but
the TFF transgene restored thymic cellularity and development of normal numbers of DP
thyrnocytes to fck -I- rnice (Figure III-3B). As noted previously (Molina et al.. 1992). we
found that TCRPICD3e expression on fck -I- DP thymocytes was 3-4 fold higher than on wild-
type DP cells (Figure III-4A), though a prominent T C R ~ ~ ~ ~ ~ ~ subset was absent (Figure III-
3B). We d s o noted that expression of CDS. which is normally upregulated in parallel wiih
TFF Transgene: - + - + Genotype: MW-/- kUGI+/- RAGl+/+ RAGI+/-
MHCII -/- Ick +/- k k -1- lck -1- No. Thymocytes
Relative Densi ty :
Fyn +
Figure 111-2. Western blot analysis of Fyn protein levels in thymocytes. Post-nuclear
supematants from the indicated number of cells were separated by SDS-PAGE (non-reducing),
transferred to ni trocellulose, and blotted wi th Fyn-speci fic anti-sera. Thymocytes from MHC
clms I -1-11 -1- mice are primaril y DP cells, and Ick -1- thymocytes consist of approximatel y
equal numbers of DN and DP cells. Densitometnc analysis was performed, and the results
were nonnalized to the signal from MHC -/- thymocytes, which was arbi trady designated as
1.0.
" -QI .rr *-QllC
TFF - MG1 lck - - / +/-
3.4+ 1.4 x 10 6
Figure III-3A. Effect of the TFF transgene on T ce11 development in RAGl -/- mice.
Thyrnocytes from 3-6-week old TFF f and TFF ' R4GI -1- mice were counted and analyzed
for expression of CD4 vs CDS. CD25. and TCRP as described for Figure III- [B. Since al1
TFF transgenic micr wrrr drrived from TFF + x TFF - matings. they express only 1 copy of
the transgene. Staining wirh isotype-controi matched antibodies (shaded histograrns) and
antibodies specific for C D 3 or TCRP (clear histograms) is compared.
TFF RAGI lck
Figure III-3B. Effect of the TFF transgene on T ceIl developrnent in RAGI -1- and Ick -'- mice. Thymocytes frorn 3-6-week old TFF + and TFF - RAGI -1- mice were counted and
analyzed for expression of CD4 vs CD& CD25, and TCRP as described for Figure III-3A.
Since al1 TFF transgenic mice were denved from TFF + x TFF - rnatings. they express only 1
copy of the transgene. The histogram rnarkers denote the mature ~ c ~ m e d ' h i subset in each
strain. Note that this population is not detected in TFF -1ck -/- mice. The numbers in brackets
refer to mean fluorescence intensity of cells falling into the TC~medhi gate.
TCR during thymocyte development (Bendelac et al., 1992: Guidos and Weissman. 1993) was
5-6 fold lower than normal on lck '/- relative to normal DP thymocytes (Figure III-4A).
Interestingly, the TFF transgene ülso restored normal levels of thymocyte TCR and CD5
expression to DP thymocytes in RAGl +/- lck -/- mice (Figure III-4B). Thus, the TFF
transgene can replace Lck Function to prornote the expansion and phenotypic development of
DP thymocytes.
The TFF Transgene Improves the DP to SP Transition in lck -1- Thyrnocytes
Lck-deficiency also reduces, but does not abrogate, the development of CD4 and CD8 SP
T cells (Molina et al.. 1992). This reduction couid reflect the reduced pool size of DP
precursors in lck 4- mice. and/or a specific defect in positive selection of lck -1- D P
thymocytes. In the former case. DP and SP thymocytes should both be reduced in number but
present at normal frequencies. as has been shown for interleukin-7 -1- mics (von Freeden-
Jeffry et al.. 1995). However, very few SP thymocytes are detectable in Ick -1- mice (Molina
et al., 1992). and most thyrnocytes falling into the SP gates have CD4Wo or CD4io8+
transitional phenotypes (Figure III-3B). Furthermore. few transitionaVCD4 SP thymocytes in
lck -1- rnice expressed mature levels of TCR. CD5, and CD69 (Figure II14A). suggesting that
positive selection of DP thymocytes into the CD4 lineage is defective.
Expression of the TFF transgene enhanced the production of transitional/SP and
TC ~ m e d l h i thymocytes in lck -/- mice (Figure III-3B). suggesting that activated f yn
significantly improves the DP to SP transition in the absence of Lck. This notion was further
supported by the observation that transitional/SP thyrnocytes from TFF + RAGl 4- lck -/-
mice expressed nearly normal levels of TCR, CD5, and CD69 (Figure III-4B). in contrast to
those from lck -1- mice (Figure III4A). These data suggest that the TFF transgene can
partially replace Lck function in promoting the DP to SP transition, but several findings
suggested that positive selection of lck -'- DP thyrnocytes was not cornpletely norrnalized by
the TFF transgene. First, expression of TCR and CD5 on transitiond/SP thymocytes from
Figure III-4A. Effect of the TFF trinsgene on the DP to SP transition in lck -'- mice.
Thyrnocytes trom 3-6-week old 86 vs lck -1- rnice were analyzed by three-color flow
cytometry for CD4 and CD8 expression vs TCR. CD5. or CD69. For each genotype. the RI
gate defines DP thymocytes and the R2 gate defines transitionallCD4 SP thymocytes. The
histogram overiays compare the expression of TCR. CD5. and CD69 on R I or R?-gated cells
from 86 (shaded) vs lck -/- (clear) mice.
- TCR- - CD5 - Figure III-JB. Effcct of the TFF transgene on the DP to SP transition i n ick -/- mice.
Thyrnocytes ti-om 3-6-werk old 86 vs TFF + lck -/- micc were malyzed by three-color flow
cytometry for CD4 and CD8 expression vs TCR. CD5. o r CD69 For cach genotype. the R1
gatr detïnes DP thymocytes and the R? gaie drfinrs transitional/CD? SP thymocytes. The
histogram ovcrlays compare the expression of TCR. CD5. and CD69 on RI or R2-gated cells
frorn 86 (shaded) vs TFF + lck 4- (clear) mice.
TFF + mice. although improved relative to non-iransgenic lck -/- thymocytes, was still slightly
lower than normal (Figure III-4B). Furthermore, although TFF + lck -/- mice had slightly
higher numbers of splenic TCRap+ T cells than TFF -Zck -1- mice. wild-type nurnbers of
TCR@ T cells were not observed (Table III-2), and K R and CD4 levels rernained
abnormally low (Figure 111-5). Notably . however. the TFF transgene restored normal
numbers of' splenic TCR@ T cells to Ick -/- mice (Table 111-2). The TFF transgene had
relatively little impact on thyrnocyte development in RAGl +" Ick +/- or RAGl +/+ Zck +/+
rnice, although marginally increased numbers of SP thymocytes expressing slightly reduced
TCR levels were sometimes observed (Figure III-3B). However, penpheral T cells frorn these
mice had a normal phenotype (Figure 111-5).
TCRapKD3 and CD4 Signaling in lck 4- DP Thyrnocytes
The above results demonstrate that maturation of lck -Id DP thymocytes into the SP lineages
is defective, and that this c m be amelionted by expression of constitutively activebn . These
observations suggest that Lc k-deficiency impairs signal transduction even ts required for
efficient positive seiection of DP thymocytes in vivo. in accord with a recent study (Hashimoto
et al., 1996). Thrrefore. we assessed TCR- and CD4-mediated signal transduction in normal
and lck -/- DP thymocytes. As previously noted by several investigators (Gilliland et al., 199 1;
Turka et al., 199 la: Wiest et al.. 1996), we found that aggregation of TCR or CD3 on fresh ex
vivo DP thymocytes frorn normal mice produces only marginal increases in protein tyrosine
phosphorylation. but this response can be markedly improved by coaggregation of the TCR
with CD4 (Figure III-6A). However. CD4-induced tyrosine phosphorylation. most notably of
the 1201130 kDa protein substratr, was abrogated by Lck deficiency (Figure III-6A).
Nonetheless. Zck -1- DP thymocytes exhibited robust TCR-induced protein tyrosine
phosphorylation without CD4 coaggregation, arguing that proximal TCR signaling pathways
are intact, and perhaps improved, in the absence of Lck. Distal K R signaling pathways were
also activated, since TCR stimulation increased CD5 and CD69 expression and decreased
Table 111-2: Effect of TFF Transgene on Peripheral T Ce11 Number in lck -1-
Mice
Geno type No. of T Cells
lck RAGZ TFF N TCRap+ (x TCR$5+ (x 10-5)
+/+ +/+ - 5 47 14 1 6 2 3
-1- +/+ - 5 6 + 4 4 + 2
-1- +/- + 3 18 k 5 19+9
N. number of individuals per group. Erythrocyte-depleted spleen cells from individual mice of
each genotype were counted and the absolute number of l'CR@+ and TCRyS+ T cells in each
sarnple was determined as described for Table ILI- 1.
Figure 111-5. Effect of TFF transgene on the frequency of TCRp+ and CD-lf splenic T
cells. Erythrocyte-depleted spleen cells from individual 3-6-week oid mice of each genotype
were analyzed for TCRP and CD4 expression as described for Figure III- 1A. Two to five rnice
of each genotype were analyzed and a representative animal from each group is shown.
Stiiining with isotype-control matched üntibodies (shaded histograms) and antibodies specific
for TCRP or CD4 (clear histograms) is compared.
DP Thymocytes Ick +f+ lck "-
Figure III-6A. Effect of Lck-deficiency on TCR-mediated signal transduction in DP
thymocytes. TCRlCD4-induced protein tyrosine phosphorylation in lck +/+ vs Ick - f - DP
thyrnocytes. Punfied DP thymocytes from kk +/+ and kk -1- mice were cultured for 1 min at
370 C without stimulation or after antibody-mediated cross-linking of the indicated surface
molecules. Cellular proteins from equal ce11 numbers were separated by SDS-PAGE (non-
reducing) , transfemd to nitrocellulose, and blotted w ith a monoclonal anti-phosphotyrosine
antibody (4G 10). Arrows indicate several proteins that undergo TCWCWinducible tyrosine
phosphorylation. Numbers on the left indicate the migration of MW standards.
RAGI expression in lck -1- DP thymocytes, although these responses were slightly l e s
efficient in the absence of Lck (Figures III-6B and III-6C). Thus, the TCR c m tsansduce
signals leading to activation/matuntion of lck -/- DP thymocytes iri vitro. yet these cells fail to
mature normally Nt ~ ~ i v o .
lck +/+ Eck -1-
Stimulation: - TCR - TCR ---- Vi rn in YI
m C " * M P ? - = - 0 . . cDNA Dilution: , , , ,
Figure III-6B. Effect of Lck-deficiency on TCR-mediated signal transduction in DP
thymoctyes. Reverse transcri ption-pol ymerase chain reaction anal y sis of RAGl and 6-actin
transcripts in DP thymocytes culturecl ovemight done or with anti-TCR@ The indicated cDNA
dilutions were PCR ampified with primers sWc for M G 1 or pactin and the products were
fractionated on agarose gels, blotîed ont0 nylon membrane, probed with 32~-labelled RAGl or
pactin cDNA fragments, and exposed to a Phosphorimager screen. Note that both normal and
kk -1- DP thymocytes show TCR-induced down-regulation of RAGl transcripts, whereas #l-
aciin tranxnpt abundance is similar in al1 samples.
CDS - Figure III-6C. Effect of Lck-deficiency on TCR-mediated signal transduction in DP
thymoctyes. Büsül and TCR-induced CD5 and CD69 expression by normal vs. lck -'- DP thymocytes. DP thymocytes from each genotype were purified by ce11 sorting and then
cultured ovemight alone or with immobilized TCRP specific antibody. Histograms show CD5
and CD69 expression on unstimulated (shaded histograms) or anti-TCRP-stimulated (clear
histograms) DP thyrnocytes.
III. 4. Discussion
This study has revealed that Fyn and Lck can serve redundant functions dunng T ce11
development. in accord with the documented ability of Src farnily kinases to substitute for one
another in regulating growth and differentiation of other cd1 types (Lowe11 and Sonano, 1996).
This redundancy was evident in the development of different T ce11 lineages (TCRaP and
TCRyG), as well as during pre-TCR and TCR-rnediated signaling events required for TCRaP
ce11 development. However. the degree of redundancy was not the same in al1 of these
processes. For exarnple, development of lymph node TCRyGf T cells was comprornised to a
similar but minor extent in the absence of Lck or Fyn, but was virtually abolished in the
absence of both molecules, suggesting that there is a high degree of functional overlap between
these two kinases. In contrat, developmental transitions mediated by the pre-TCR and TCR
are severeiy compromised in the absence of Lck, but remain largely intact in the absence of
Fyn, suggesting a greater reliance on Lck in these developmental processes. Nonetheless, we
show here that transgenic expression of constitutively active Fyn cm dmos t completely replace
Lck function in al1 of these aspects of T ce11 developrnent.
Functions of Fyn and Lck in TCRyS Cell Development
We observed small nurnbers of TCR@ thymocytes in fck -1- j jn -1- anirnals, suggesting
that the absence of these cells in spleen and lymph nodes (Figure III- IA; Table III- 1) may
reflect defective selection, export from the thymus, ancüor peripheral expansion of these cells.
Although Vy 2+ T cells are positively selected on MHC class 1. developrnent of Vy 3+ skin T
cells is class 1-independent (Wells et al.. 1991; Correa et al., 1992). In addition. available
evidence suggests that selection of Vy 3+ is not dependent on TCR-ligand interactions
(Asarnow et al., 1993). Thus, i t remains unclear whether the TCRyS or other ce11 surface
receptors interact with Fyn and Lck during TCRy6 ce11 developrnent.
Notably, the TFF transgene restored splenic TCRy6+ ce11 numbers to normal in fck
-deficient mice (Table III-2), arguing that it can effectively compensate for Lck in this regard.
However, development of Vy 7-transgene-bearing T cells appean to be strictly Lck-de pendent
(Penninger et al.. 1993), suggesting that Fyn can replace Lck function for the production of
most, but not d l . thy micnlly-derived TCRys+ cells. Interestingly . the development of intra-
epithelial TCRyS+ cells in the gut, thought to occur extrathymically (Poussier and Julius,
1995), was previously shown to be Lck-independent (Penninger et al., 1993). Studies are in
progress to determine whether Fyn plays an essential or a functionally redundant role in the
development of this TCRy6+ ce11 subset.
Functional Redundancy of Lck and Fyn in Pre-TCR-Mediated Positive
Selection
Although Fyn plays no essential role in the DN to DP transition (Appleby et al., 1992;
Stein et al., 1992). experiments described here show that development of DP thymocytes in kk
-1- mice is Fyn-dependent (Figure III- 1 ). These findings likely explain why RAG -1- lck -1-
mice can still develop small numbers of DP thymocytes after treatment with ionizing radiation
or CD3 specific antibodies (Levelt et al.. 1995: Wu et al.. 19%). In addition, our observations
suggest an expianation for the different thymocyte phenotypes of k k -'- rnice versus mice
expressing a dominant-negarive mutant ïck transgene. DP thymocyte developrnent and TCRP
allelic exclusion (Anderson et al., 1993; Levin et al., 1993) are completely abrogated in the
latter mice (Anderson et al., 1993; Levin et al., 1993). whereas these events are only partially
compromised in lck -/- mice (Molina et al., 1992; Wallace et al., 1995). We suggest that the
dominant negative lck transgene, which was expressed at 17-fold higher levels than
endogenous Ick , did not permit adventitious signaling by Fyn, which clearly c m participate in
DP thymocyte development in lck -/- mice.
In mice facking both Lck and Fyn, we observed an accumulation of TCRP+ CD25+ DN
thymocytes (Figure III-IB). suggesting that pre-TCR-expressing precursors could not
transduce signais required for developmental progression and proliferation. In accord with this
idea, expression of a constitutively a c t i v e m transgene completely obviated the need for Lck
function at this stage of T cell developrnent (Figure III-3B). suggesting that the two kinases
could in principle serve overlapping functions in the DN to DP transition. However, several
observations argue that Fyn and Lck do not function identically in this regard. First, DP
thyrnocyte developrnent is severely impaired in lck -/- mice, but is normal infyn -1- mice
(Appleby et al., 1992; Molina et al., 1992; Stein et al.. 1992). This is not likely due to
quantitative differences, since i t has been reported that Fyn and Lck are cxpressed at sirnilar
levels in DN and DP thymocytes (Olszowy et al., 1995). Second, transgenic expression of
constitutively active lck both restores normal DP thymocyte development to RAGI -1- mice
(Anderson et al., 1992; Anderson et al.. 1993; Mombaerts et al., 1994). but the analogous TFF
transgene does not (Figure III-3A). Again, this difference does not correlate with different
levels of transgene expression, since the effect of the mutant Lck was observed when it was
expressed at endogenous Lck levels. yet the TFF transgene had only marginal effects despite
being over-expressed at least 2-fold relative to endogenous Fyn (Figure 111-2). Finally, over-
expression of a dominant negatjve Ick transgene cornpletely abrogates DP thymocyte
development, whereas a dominant negativefin transgene does not affect T ce11 development
(Cooke et al., L99 1).
Collectively, these data suggest that Lck and Fyn may interact with an overlapping but not
identical set of substrates to mediate pre-TCR signals. either because they possess inherently
different substrate specificities or due to distinct intracellular distributions (Ley et al., 1994).
In mature T cells, Fyn and Lck are thought to be activated by aggregation of distinct ce11
surface receptors, since Fyn is found nssociated with the cytoplasrnic tails of CD3 chains and
other surface receptors, whereas at least sorne cellular Lck is associated. via its unique N-
terminal domain, with cysteine residues in the cytoplasmic tails of CD4 and CD8a (Weiss and
Littman, 1994). However. once activated, both Fyn and Lck are thought to phosphorylate
tyrosine-based activation motifs in TCRC, C D ~ E , CD3y. and CD36, and subsequently, ZAP-
70 and/or Syk, mernbers of a distinct tyrosine kinase family (Anderson et al., 1994b; Weiss
and Littman, 1994). By contrast, neither the intracellular distributions nor the protein
substrates of Lck and Fyn in pre-TCR-expressing immature thymocytes have been identified.
In the simplest rnodel. Fyn and Lck would both be associated with the cytoplasmic tails of pre-
TCR components, allowing their direct activation by aggregation of pre-TCR complexes, but
this has not yet been demonstrated biochemically, and the existence of pre-TCR ligands
remains speculative. Moreover, Fyn and Lck also interact with cytokine receptors and other
surface molecules. such as CD2. known to be expressed in immature thymocytes (Anderson et
al., 1994b; Seckinger and Fougereau, 1994; von Freeden-Jeffry et al.. 1995). Thus. Lck and
Fyn need not be directly associated with the pre-TCR. Substrates phosphorylated by Lck and
Fyn in pre-T cells are also unidentified, although ZAP-70 was shown to activated by CD3
crosslinking of a pre-TCR-expressing immature T ce11 line (van Oers et al.. 1995). However.
the absence of ZAP-70 or Syk does not compromise development of DP thymocytes (Cheng et
al., 1995; Negishi et al.. 1995; Turner et al.. 1995). Thus, ZAP-70 and Syk may no< be
involved in pre-TCR signding, or they may also play redundant roles in this process.
Role of Lck and Fyn in Positive Selection of DP Thymocytes
Fyn-rnediated signal transduction is dispensible for positive selection of DP thymocytes
into the CD4 or CD8 lineage (Appleby et al., 1992: Stein et al., 1993; Swan et al.. 1995). In
contrast. very few SP thymocytes and T cells develop in k k -'* mice (Molina et al., 1992).
We show here that the DP to SP transition is significantly comprornised in the absence of Lck,
suggesting that Lck-mediated signals are important for positive selection of most DP
thymocytes. In accord with this idea. a recent study dernonstrated that expression of
catalytically inactive k k specifically in DP cells using the distal lck promoter effects an
unarnbiguous block in positive selection (Hashimoto et al.. 1996). The role of Lck in positive
selection could involve transducing signals from CD4/CD8 and/or the TCR complex. We
show here that defective positive selection of lck '/" DP thymocytes in vivo does not appear
to correlate with defective TCR signaling il1 vitro (Figure 111-6). B y contrast, we found that
CD4-dependent signaling was abrogated in lck -1- DP thymocytes (Figure III-6A). This defect
could impinge on positive selection in vivo, since a recent study (Wiest et al., 1996) showed
that activation of ZAP-70, an essential event in positive selection (Negishi et al., 1995),
requires coaggregation of CD4 with the TCR in DP thymocytes. However, other defects in lck
-/- DP thymocytes might also account for their defective maturation into SP T cells in vivo.
For example. Lck-deficiency caused decreased CD5 and increased TCR expression in DP
thymocytes, and both defects were normalized by expression of the TFF transgene (Figure III-
4). Interestingly, these results funher suggest that DP thymocytes modulate CD5 leveis in
response to Lck- and Fyn-dependent signal transduction Ni vivo. CD5 has been reported to
negatively regulate TCR signaling in thymocytes and to play an important role in positive
selection (Tarakhovsky et al., 1995), suggesting that abnormal CD5 expression couId impair
positive selection of lck -1- DP thymocytes.
Although endogenous Fyn appears to cornpensate poorly for Lck in development of
transitionaVSP thymocytes and peripherai T cells, the TFF transgene could partially restore
this developrnental transition (Figure 111-4, Table III-?), demonstrating that Fyn c m transduce
signals important for positive selection of DP thymocytes. The differential effectiveness of
endogenous wild-type Fyn and transgenic TFF might be explained by the slightly higher
expression of the latter in DP thymocytes (Figure III-?), but it is important to note that
endogenous Fyn expression increases during the DP to SP transition (Cooke et al., 1991),
whereas TFF transgene is expected to decrease across this transition, due to its expression
under control of the Ick proximal promoter. Downregulation of TFF transgene expression
likely begins during the earliest phases of positive selection, since we have recently shown that
ovemight TCR engagement of DP thymocytes in vitro significantly decreases transcription
from the lck proximal promoter (Groves et al., 1997). Positive selection is thought to be a
multi-step process requiring several TCR engagement events (Guidos, 1996). so
downregulation of the TFF transgene during the late stages of positive selection could account
for its failure to completely restore the development of SP T cells in lck -/- mice. An equally
plausible explanation for the differential effectiveness of endogenous Fyn and transgenic TFF
in this regard is that endogenous Fyn may not be activated efficiently by TCR aggregation Ni
vivo, perhaps because positive selection of DP thymocytes usually occurs in response to low
affinity TCR-ligand interactions (Jarneson et al., 1995). Finally. the tyrosine to phenylaianine
substitution at position 528 could alter the substrate specificity of TFF relative to wild-type
Fy n.
In surnrnary, Our studies have identifed overlapping functions of Fyn and Lck at multiple
stages of T ceIl development. funher emphasizing the pivotal role that Src family kinases play
in this regard.
Chapter IV Discussion
The major objective of this thesis was to examine the signaling requirements at two major
checkpoints of T ce11 development. The first principal step occurs at the DN to DP transition,
imd is dependent upon expression of the pre-TCR complex. which is composed of TCRP chain
in association with pTa chain and CD3 proteins. The second important checkpoint occurs at
the DP to SP transition. and is regulated via TCRup signaling, which mediate positive and
negative selection during this transition. Below I will discuss my results in the context of these
two transitions.
1. The DN to DP Transition
A. Lck and Fyn in pre-TCR signaling at the DN to DP transition
In vivo studies (Molina et al., 1992; Levin et al.. 1993) demonstrate that Lck functions in
signding at the DN to DP transition. Fyn is expressed at comparable levels to Lck in immature
thymocytes (Olszowy et al.. 1995) but does not play an essential role in T ceIl development
(Appleby et al.. 1992: Stein et al., 1992). The present study (Chapter III) addressed the
developmentül importance of Fyn when Lck is absent. My results demonstrated that, in the
absence of both Src tarnily PTKs. TCRap ceil development was severely impaired. Thyrnic
cellularity in the double-mutant mice was approximately 5-fold less than in Ick -/- thymi (Table
III- 1 ). Furthermore, virtually al1 thyrnocytes frorn Ick dyr.ii 4- thymi were arrested at the
C D X f DN stage (Figure III-lB), resulting in the lack of detectable mature thymocytes and T
cells. Similar findings were reported in another study (van Oers et al., 1996b). Additionally,
van Oers et al. found that the development of Vy3+ dendritic epidermal cells was more severely
affected in Ick -/-fin mice versus Ick 4- mice, whereas the functional activity of NK cells
was unperturbed in lck -/-jjn -1- mice. These results thus reveal a role for Fyn in T cell
development when Lck is absent.
The role of Fyn at the DN to DP transition was further assessed by generating Ick -1- mice
that expressed a constitutively active fyn (TFF ) transgene under the lck proximal promoter.
leading to TFF expression in immature DN and DP thymocytes. Expression of the TFF
transgene increased the thymic cellularity of lck -1- mice by IO-fold, back to wild-type nurnbers
(Figure III-3B). The transgene d s o enhanced the generation of transitional and SP mature
thymocytes, which expressed nearly normal levels of TCR, CDS, and CD69 (Figure II14B).
However, it failed to mediate the generation of wild-type numbers of mature TCR@ cells in
the periphery (Figure 111-5). Thus, the TFF transgene completely restores the DN to DP
transition in Ick -1- rnice, but only partially restores the DP to SP transition (see below). In
summary, results frorn the Fyn Ioss-of-function ( I d -l-fvn -/' mice) and gain-of-function
(TFF transgene) studies suggest that Fyn can partially substitute for Lck in mediating thyrnic
development.
The function of Fyn in substituting for Lck at the DN to DP transition was further
examined by generating RAGI ' h c k 4- mice expressing the TFF transgene. Previous work
demonstrated that expression of an activated Ick (lck 505 ) transgene rescues the development
of DP thymocytes in RAG -1- mice. indicating thiit an activated Lck can bypass the need for
pre-TCR expression (Mombaerts et al., 1994). In contrast, expression of the TFF transgene
failed to significantly promote the generation of DP thymocytes in R A G I 4' fck -l- mice,
suggesting that Fyn cannot substitute for Lck in rescuing DP thymocyte development in the
absence of pre-TCR expression (Figure III-3A). Recent studies found that Lck is necessary in
permitting DP thymocyte maturation to be restored in RAG -1- rnice following treatment with
anti-CD3e or ionizing radiation (Levelt e t al.. 1995: Wu et al.. 1996). However, it is not clear
whether Fyn clin substitute for Lck in mediating T ce11 developrnent in R A G -1- mice in
response to these treatments. If Fyn substitutes for Lck in these responses, then the generation
of DP thyrnocytes will occur in TFF + Ick -/- RAGI -/- rnice following treatment with either
ionizing radiation or anti-CD~E. Recent expenments suggested that treatment of TFF + lck -'- RAGl -1- mice with anti-CD~E promotes the generation of DP thymocytes (B.J. Edgell, C.J.
Guidos, unpublished observations). This finding argues that the TFF transgene requires pre-
TCR or CD3 aggregation to mediate the DN to DP transition. whereas this is not the case for
the lck 505 transgene.
Differences in the abilities of the activated Ick versus thefjri transgenes to relieve the
developrnental West at the DN stage in RAG -'- mice may reflect distinct properties of Fyn
and Lck. A study assessing Fyn and Lck expression in human T cells found that the two
PTKs have distinct intracellular localization pattems (Ley et al.. 1994). Lck is predorninantly
found at the plasma membrane but is also obsewed in pericentrosomal vesicles. In contrast.
Fyn does not CO-localize with Lck but rather is associated with the centrosorne and microtubule
bundles arising from the centrosome. In addition to differences in intracellulx localization
patterns, Lck and Fyn may regulate overlapping as well as distinct substrates. Lck has been
shown to primarily mediate tyrosine phosphorylation of K R 6 and ZAP-70 following TCR
stimulation (van Ocrs et al., 1996a). In contrast, a recent study demonstrated that Fyn
specifically induces tyrosine phosphorylation of Pyk2. a member of the focal adhesion kinase
farnily PTK, during TCR signaling (Qian et al., 1997). Thus. differences in intracellular
expression of Lck and Fyn in T cells may lead to regulation of distinct substrates involved in
downstrem pathways of pre-TCR signaling.
Recent studies have examined the MAPK pathway in pre-TCR signaling at the DN to DP
transition. Following TCRaP stimulation. Ras is activated leading to successive
phosphorylation and activation of kinases of the MAPK cascade (reviewed in Alberola-Ila et
al.. 1997). In a recrnt study assrssing the role of the MAPK cascade in pre-TCR signaling, an
activated R m transgene was demonstrated to permit RAG -/- DN thymocytes to undergo
differentiütion into DP cells as well as cellular expansion to wild-type numbers (Swat et al.,
1996). In onother report. retrovirus-mediated gene transfer of mutant MAPK kinase
(MAPKK) into FTOCs was empioyed to assess the role of the MAPK signaling pathway in the
regulation of thymocyte development (Crompton et al., 1996). The results showed that
catalytically inactive MAPKK inhibits the generation of DP thymocytes in RAG '/- FTOCs
following anti-CD3e treatment. Similarly. development at the DN to DP transition is impaired
in lobes of TCR a-/- FTOCs overexpressing catalytically inactive MAPKKI. These results
suggested that the MAPK signaling pathway is important during the DN to DP transition. In
contrast. Perlmutter's group found that the MAPK pathway may not be criticnl in pre-TCR
signaling, since transgenic expression of either dominant-negative MAPKKI and/or dominant
negative Ras tninsgenes did not perturb the DN to DP transition (Alberola-Ila et al.. 1995;
Swan et al.. 1995: Alberola-Ih et al., 1996). One possible explanation for the differences
between these two studies is that the dominant negative trrinsgenes may induce different effects
in fetal versus adult thymocytes. Alternatively. in the Perlmutter study, the transgenes rnay not
have been expressed at high enough levels to interfere with the iictivities of the endogenous
gene products. In summary, the role of the MAPK pathway as well as other downstrearn
signaling events in pre-TCR signaling needs to be further explored.
B. Role of Lck and Fyn in extrathymic T ce11 development
The results of this thesis (Chapter IiI) demonstrated that the absence of both Lck and Fyn
profoundly affected the developrnent of TCRaPf cells and TCRyGf cells in the thymus, but 1
did not assess the effects of both Lck and Fyn deficiency on extrathyrnic T ce11 developrnent.
The intestinal epitheliurn is a thymus-independent site of T ce11 development that results in the
ocneration of both TCRaP+ and TCR-,6+ intra-epithelial lymphocytes ( IELs: Poussier and b
Julius, 1994). A subset of TCR~P'IELS and the majority of TCR*pS+ IELs express CD8aa
homodimers. whereas thyrnically-derived CD8+ T celis express CD8uP heterodimers. The
TCRaP- CD8aa+ IELs are thought to be gut-derived. w hereas TCRap+ CD8up+ IELs are
considered to originate from the thymus. In contrat to development of thyrnically-derived T
cells, the maturation of most [EL subsets is relatively normal in fck -/- mice (Penninger et al..
1993; Page et al.. 1997). Nonetheless, the number of TCRap+CD8aa+ LELs is reduced by
two-fold, but the number of K R y * C D 8 a d cells is similar to normal rnice (Penninger et al..
1993: Page et al., 1997). Similar to findings for thymocyte development in jyn -1- mice
(Appleby et al.. 1992. Soriano et al.. 1992). a recent study by Pullen's group found that
extrathyrnic developrnent of IELs is normal i n h l 4- mice (Page et al., 1997). This study also
found that TCRapf CD8aa+ IELs are absent in lc-k -/-fin -/- mice (Page et al.. 1997). Thus,
similar to results from my thesis (Chapter III) for thymocyte development in lck -1-fyn -1-
mice. Fyn plays a role in IEL development when Lck is absent. Interestingly, TCR~B'IELS
are vinually absent in lck - / ~ n +/- mice, indicating that T C R ~ ~ ' E L development requires
two copies 0fij.n when Lck is absent (Page et al.. 1997). In contrast. data from my thesis
(Chapter III; Figure III-1B) demonstrated that thymocyte development in lck -1- mice was
identical in tck -/-fin 4 - rnice and lck - / - f i n rnice. Thus. jjw gene dosage has no
affect on thyrnocyte development but is criticai for E L development in lck -/-mice. In conirast
to TCRap+ CD8aa+ IELs. some TCR@ CD8aa+ IELs, albeit in small numbers, were
generated in lck -1-f~ 4- rnice. indicating that the development of these cells can occur
independently of either Lck or Fyn (Page et al.. 1997). Similady, results (not shown) from
rny thesis (Chapter III) demonstrated that TCR@+ thymocytes are evident in ick -/-fin -1-
mice. indicating that the generition of both gut- and thymic-derived TCR@ cells are not
critically dependent upon Lck or Fyn. In summary, the development of both gut- and thymic-
derived TCRao+ and TCR-,6+cells have similar as well as distinct requirements for Lck and
Fy n.
2. The DP to SP Transition
A. The response of DP thymocytes to TCR engagement
The first component of this thesis (Chapter II) examined the in vitro capacity of DP
thymocytes to undergo maturation in response to TCR engagement alone. To assess this,
purified DP thymocytes were cultured in vitro in the presence of immobilized TCR specific
antibody. The results demonstrated that both DP PM and DP blasts were functionally capable
of undergoing some aspects of maturation associated with positive selection. In response to
TCR stimulation. borh DP subsets demonstrated increased surface CD5 and CD69 expression
(Figure II-3), as well as decreased RAGI and pTa chain mRNA expression (Figures II-5A-
C). Other changes observed included a switch in lck promoter usage from the proximal to the
distal promoter (Figure 11-6). and increased intracelIular expression of the ceil survival
regulator BcI-2 (Figure II-4A). However, other phenotypic changes that accornpany positive
selection were not observed. including CD4/CD8 lineage cornmitment. as well as increased
Thy-1 and MHC class 1 expression. and decreased HSA expression (Table 11-4). In addition.
DP thyrnocytes did not undergo TCR-induced apoptosis (Table 11- 1 ). Thus. TCR stimulation
of DP thymocytes induces aspects of thymic maturation but fails to mediate complete T ce11
development.
Various explanations may üccount for the iàilure of DP thymocytes to complete T ce11
development in response to TCR engagement in virrn . One possibility is chat stimulation via
TCR specific antibodies rnay not retlect the physiologiciil conditions in vivo. Thus, K R
engagement may induce di fferent signaling pathway s in DP thy mocytes versus those induced
via iMHC/peptide ligand. Consistent with this. a recent study by Germain's group
demonstrated that TCR and CD4 coengagement during TCR engagement of MHC/peptide
ligands delivers a different TCR-mediüted signaling response in T cells versus TCR
engagement done (Madrenas et al.. 1997). Stimulation of T cells with heterobivaient anti-CD4
and anti-CD3e was dernonstrated to induce a response pattern similar to agonist peptide/MHC
class I I ligands. In contrast. bivalent anti-CD3e induces a response resembiing partial
agonists/antagonists in T cells. In a sirnilar way. I generated heteroconjugates of anti-CD4 and
anti-TCRP to more closely mimic a MHCIpeptide ligand. and used thern to stimulate DP
thymocytes. However. these heteroconjiigates induced simiiar phenotypic changes in DP
thymocytes as those induced by anti-TCRP, suggesting that stimulation via heteroconjugates is
not significantly different from stimulation via anti-TCRP. Since these heteroconju, aates were
multimers. they rnay not be inducing the appropriate signals in DP thymocytes as those
delivered by antibody coligation of CD4 and CD3 via bivalent crosslinking as described in the
Germain study. Another explanation for the failure of DP thymocytes to complete T ce11
development in response to TCR engagement in vitro rnay refiect the absence of necessary
non-TCR derived signals provided by the thymic microenvironment (Page et al., 1993). A
third possibility is that maturation of DP thymocytes may require prolonged and continuous
engagements. These two latter explanations rnay not be mutually exclusive such that
maturation of DP thymocytes rnay undergo prolongeed or multiple engagements involving TCR
and non-TCR signals.
B. The significance of thymic stromal cells in positive selection
One explanation for the Iÿilure of TCR-rnediated signals to mediate the complete maturation
of DP thymocytes is that this in vilru mode1 system Iücked thymic stromal cells, which may
provide additional signals n s c e s s q to mediiite the DP to SP transition. The rolr of stromal
cells in mediating thymic selrction has bern assrsscd by various studies with contlicting results
(reviewed in Anderson et al., 19%). Multiple studies suggest that thymic spithelial cells
primarily rnediate positive selection (reviewed in Jamrson et al., 1995). Consistent with this,
intrathymic transfer of MHC class II+ bone rnarrow cells into MHC c1cr.s.s II -deficient rnice
fails to mediate positive selection of MHC class II-restricted CD4+ cells (Markowitz et al.,
1993). However. çvidence from other intrathymic trünsfer and bone marrow chimera studies
suggest that non-epitheliai cells such as fibroblasts (Hugo et al., 1993: Pawlowski et al., 1993)
can mediate positive selection of either MHC clüss I or class II-restricted T cells.
The role of thyrnic strornai cells in mediating thymic selection has also been addressed
using the in vitro model system whereby DP thymocytes and individual stroma! ce11 types are
mixed to form reaggregate organ cultures (Anderson et al.. I994a). Anaiysis of various MHC
class II+ stromal cell types demonstrated that MHC class II+ thymic cortical rpitheliiil cells are
unique in mediating the development of DP thymocytes into fi~nctionally mature SP cells. since
MHC class II+ saliva- and gut epithelial cells failed to support maturation (Anderson et al..
1994a). These results suggest that, in addition to TCIUMHC interactions. thymic cortical
epithelial cells mediate positive selection by providing unique accessory signals to DP
thymocytes. Results frorn the in v i fro model systern using chemically-fixed thymic cortical
epithelial cells funher dernonstrated that these signals are rnediated by interactions with surface
ligands rather than soluble factors (Anderson et al., 1994~). Consistent with these findings.
maturation of a DP ce11 line (DPK) derived from a TCR transgenic mouse. which expresses a
TCR specific for pigeon cytochrome c peptide associated with 1-E" was shown to occur in the
presence of I-E~-bearing thymic epithelial cells in reaggregation cultures (Poirier et al., 1994).
Further analysis dernonstrated that I -~~-bear ing thymic medullary epithelial cells failed to
mediate the differentiation of the DPK cells in vitro ~ind il1 rirw , consistent with the role for
thymic cortical epithelial cells in rnediating positive selection (Poirier et al.. 1994).
Additionally, purified MHC class II' cortical epithelial cells. unlike MHC çlass 11' bone
rnarrow cells. were shown to be unique in mediating positive seleciion of DP cells. further
suggesting the importance of interactions involving MHC as well as non-MHC l ipnds by
stromal cells in this regard (Ernst et al.. 1996).
In addition to IMHC/TCR interactions, recent studies have examined the role of CD28/B7
interactions. which are necessary in the costirnulatory pathway of T ce11 activation (reviewed in
June et al.. 1994). in positive selection of thymocytes. In CD28 -/' mice. positive selection of
TCRaP transgenic thymocytes is not impaired (Walunas et al.. 1996). but this result may
reflect the übility of other molecules to substitute for CD28. Nevcrtheirss. another study
dernonstrated that the addition of B7-specific antibodies to reiiggregatr cultures of MHC class
IIf thymic epithelial cells and DP thymocytes did not impair the DP to SP transition (Jenkinson
et al.. 1994). Howevrr. in this thrsis (Chüpter I I ) . results from the i ~ i i-irm mode1
demonstrated that CD28 synergized with TCR signals to incretlse CD5 expression by DP
thymocytes. Thus. CD28 c m costirnulate TCR-dependent maturation of DP thymocytes. The
physiological significance of these results is not clear, but the role of CD28 in positive selection
could Vary depending upon the overdl avidity of TCRMHC interactions.
C. CD4/CD8 lineage cornmitment
Another feature of thymic maturation that ensues in DP thymocytes following interactions
with thymic stroma1 cells is cornmitment to either the CD4 or CD8 Iineage. This process
involves TCRlMHC interactions between thymocytes and thymic stromal cells such that
thymocytes with MHC class 1-restricted TCRs develop as CD8+ cells, whereas thymocytes
bearing MHC class II-restricted TCRs develop as CD4+ cells (reviewed in von Boehmer,
1996). In this thesis (Chapter IL), results from the in vitro model failed to demonstrate
evidence of lineage commitment (Figure II- 1 ), but this may retlect the absence of accessory
signals provided by thymic stroma1 cells. Recent rvidence suggests that lineagr commitment
rnay not only be int'luenced by TCRIMHC interactions but also by celiular interüctions
involving the Notch receptor (Robey et d.. 1996). Since Notch h a been shown to play a role
in determining cell fate decisions in inverrebrates (reviewed in Artavanis-Tsakonas et ai.,
1995), and Notchl is expressed in developing thymocytes (Weinmaster et al., 1991;
Weinmaster et al., 1992; Hasserjian et al., 1996), the effect OC expressing rictivated Notchl in
immature thymocytes on CD4/CD8 lineage comrnitment was examined in MHC mutant mice
(Robey et al.. 1996). The results dernonstrated that expression of activated ~Vorclrl biüses DP
thymocyte development towards the CD8 lineage. A model was proposed suggesting that
MHC clüss 1 engagement by DP thymocytes permits those cells that receive Notch signals to
develop as CD8+ cells (Robey et al.. 1996). In contrast. MHC class II recognition might
prevent DP cells from responding to Notch signals. biüsing them towards the CD4 lineage. In
the presence of activated ~Vorcii. DP thymocytrs undergo development to the CD8 lineage in
either MHC cZoss I - or ciriss II -deficient mice. but development of CD8+ cells is very
inefficient in M H C ci<iss I and ciass II double-mutant mice. These results suggest that in the
presence of activated Norcli, recognition of either MHC clüss 1 or class I I molecules pennits
DP thyrnocytes to develop into cD8' cells.
In spite of the novel findings of the Robey et al. study. there is one potential problem with
this report. Previous work hns suggested that activared Nordi can function as an oncogene
because overexpression of activated Notchl in murine bone marrow cells lrads to T ce11
lymphomas (Pear et al., 1996). Consequently, expression of activated Notcll in immature
thymocytes rnay lead to proliferation ancilor enhanced cellular survival of thymocytes. Thus.
expression of activated Norcli may selectively promote the development of thymocytes that
otherwise would not survive the selection process by extending their lifespan. Consistent with
this interpretation, a previous study examined the effect of transgenic expression of the proto-
oncogene bcl-2 in immature thymocytes of various MHC mutant mice (Linettc et al., 1994).
Overexpressed Bcl-2 enhanced the number of celts dong the CD8 lineage in both MHC clms I
4- rnice as well as in MHC class II-restricted TCR transgenic mice. but rhese cells were
immature due to their H ~ ~ m e d - h i phenotype. In contrast, Bci-2 had no effect on development
of CD4 lineage cells in iMHC cluss II -1- rnice and failed to promote development of
thymocytes beyond the DP stage in MHC class i and clmx Ii double-mutant mice. These
results suggest that transgenic bcl-2 expression selecti vely rescues CD8 lineage cells that
would otherwise die in the thymus. Similarly. the rtbility of activated Nord1 to skew
developrnent towards CD8 lineage cells in MHC cirss I -1- or MHC cims II 4- rnice but not in
M H C clrss I and clms II double-mu tant mice may reflect selective survival of CD8 lineage
ceils rather than a specific mechanism in cornmitment towards the CD8 linrage (Robey et al..
1996). In summary. funher analysis is required to discriminate the ability of activated Notclz
to mediate lineage cornmitment versus enhancing the lifesprin of thymocytes.
D. Positive selection: a multi-step process of thymocyte maturation
A major feature of this thesis (Chapter II) wüs that TCR engagement of DP thyrnocytes in
vitro did not result in the çornplete differentiation of DP cells to SP thymocytes. Interestingly,
I observed thüt K R stimulation of DP thymocytes NI iitro down-reguinted expression of both
CD4 and CD8 and simultaneously induces CD69 expression. giving rise to
C D ~ ~ ~ C D ~ ~ ~ C D ~ ~ ~ cells (Figure II- 1). Funhermore. these coreceptor du11 cells were not only
non-apoptotic but hüd precursor activity for mature SP thymocytes. suggesting that they are not
dead-end cells but may be involved in thymic differentiation (Figures II-IA-C). A recent study
by Germain's group dernonstrated that CD410CD81oCD69+ cells represent a transitional stage
in T ce11 maturation and commitment (Lucas and Germain. 1996). These authors proposed a
new mode1 for lineage commitment in thymic development, in which TCR engagement
prevents DP thymocytes from dying by neglect and induces their differrntiation to the
C D ~ ~ O C D ~ ' ~ C D ~ ~ + ceIl stage. Thymoc ytes then mature to ~ C R m c d CD4+ C D ~ ~ U cells, w hich
in turn undergo MHC class-mediated termination of either CD4 or CD8 expression perhaps via
instructional signals. Those cells that undergo TCR-mediated interactions with MHC class 1
molecules are permitted to sustain CD8 expression while turning off CD4 synthesis. giving rise
to mature T C R ~ ~ CD8+ cells via the TCRmrd CD 410 C Dg+ stase. In contrüst. T c R ~ ' ~
C D ~ + C D ~ ] O cells bewing TCR that in~eract with MHC class II molecules are prrmitted to
sustain CD4 expression while terminating CD8 synthesis, giving rise to mature T c R ~ ~ CD4+
cells. In conclusion. results from this thesis (Figure 11-1) demonstrating the presence of
C D J ~ D ~ ~ ~ C D ~ ~ + cells from in vitro cultures of TCR-stimulated DP thymocytes are
consistent with the notion, as recently shown in a revised mode1 of lineüge cornmitment (Lucas
and Germain. 1996). that ~ ~ 4 1 " ~ ~ 8 ~ 0 ~ ~ 6 9 + cells are transitional cells during thymic
maturation.
In the Germain study. C D ~ I ~ C D ~ ~ ~ C D ~ ~ ~ crils isolated directly from the thymus could
spontaneously develop into T C R ~ ~ ~ CD@ CD8lo cells hi rirro ( Lucas and Germain. 19%).
However. 1 generated C D ~ ~ U C D ~ ~ < ) C D ~ ~ + cells iri itirro tollowing TCR engagement of DP
cells. but they did not progress beyond this stage (Chapter II). This observation suggests that
non-TCR signals, perhaps in conjunction with sustüined or additional TCR engagements. may
be required to mcdiiite this next developrnrntal transition. Similar conclusions have been made
by other investigators. Using the in vitro reaggregate culture system. one study showed that
the complete maturation of DP CD69+cells. which have begun a TCR-dependent process.
requires additional interactions with thyrnic epithelial cells to cornplete the DP to SP transition
(Wilkinson et al., 1995). Similarly, another study demonstrated that positively selected T C R ~ ~
DP thymocytes from TCRlhcl-2 double transgenic mice fail to generate mature SP cells
following intrathymic transfer into recipient mice unless they express posiiively selecting MHC
class 1 molecules (Kisielow and Maizek, 1995). Thus. both the survival and maturation of
T C R ~ ~ DP thymocytes into functionally mature T cells may be dependent upon sustained TCR
engagements and thymic stroma1 ce11 signÿls. This is further supponed by results from this
thesis (Chapter III). Expression of the TFF trmsgene in lck -/* mice significantly improved
the DP to SP transition but friiled to restore wild-type nurnbers of TCRap+ T cells in the
periphery (Figure III-4B; Table III-?). The impaired generation of peripheral TCRc@+ T cells
in TFF + I d -/- mice müy reflect down-regulated TFF trünsgene expression dunng the latter
stages of positive selection. This is a plausible explanation because TFF transgene expression
is under control of the lck proximal promoter. which has reduced transcriptional activity in DP
thymocytes following TCR stimulation (Figure 11-6). Consequently, the requirement for TFF
transgene expression in mature thymocytes may allow them to complete thymic differentiation
in fck -1- mice. The inability of SP thymocytes to complete their maturrition program in TFF
+lck-/- mice rnay explain the Iow yield of peripheral TCRapf T cells due to either irnpaired
migration to the periphery andor failure to survive in the periphery. Collectively. these data
argue that positive selection may involve ü series of TCR signaling events ro induce the
cornplete maturation of DP to SP cells. as well as rescue tiom PCD those cells that have
successfully completed differentiation.
E. Modulation of TCR-mediated signaling in DP thymocytes
Multiple in virro and irt i ~ i w studies indicate that various signal transduction pathways
directly regulate TCR-mediated signaling. Similar to mature T cells, approximately 50% of
surface CD4 molecules are associated with Lck in DP thyrnocytes. whereüs only 2% of surface
CD8 molecules are iissociated with Lçk (Wiest et cl.. 1993). Thymocytes from lck -/- mice
express increased surface TCR levels relative to normal thyrnocytes (Figure III-3B; Molina et
al., 1997). This upregulation in TCR expression by DP thymocytes is reported to reflect the
regulatory role of Lck in TCR expression (Wiest et al., 1993; Ericsson and Teh, 1995). Lck
appears to negatively regulate TCR expression in DP thymocytes iri vivo via M H C class
IUCD4-associated Lck interactions, and this is relieved by culturing DP thymocytes in vitro at
37°C (Nakayama et al., 1990). Thus. disruption of in vivo MHC class IVCD4-associated Lck
interactions prevents Lck-mediüted down-modulation of TCR expression in DP thymocytes as
dernonstrated in MHC ckiss II '/- mice (Wiest et al.. 1993) and anti-CD4 treated mice
(McCarthy et al.. 1988). Similxly. Lck-mediated regulation of TCR expression is decreased
in CD4 trmsgenic mice overexpressing CD4 molecules (Nakayama et al., 1993b). An increase
in CD4 molecules that are not associüted with Lck is suggested to reduce the iimount of Lck
aggregrated by C W crosslinking and thus decreased Lck activation. Thus. thex results clearly
demonstrate a role for Lck in reguiüting TCR expression. Moreover. TCR-mediated signaling
in DP thymocytes fiiils to activate ZAP-70 and thus does not induce downstream TCR
signaling events because intrüthymic iMHC class IVCD4 interactions reduce the number of Lck-
associated CD4 molecules available to rtctivate ZAP-70 ( Wiest et al., 1996). However,
coaggregation of CD4 and K R in DP thymocytes permits DP thymocytes to undergo ZAP-70
activation and productive TCR signüling events. Another major negative regulator of TCR
signaling is the transrnembrane molecule CD5 (Türakhovsky et al.. 1995). In CD5 -/- mice.
DP thymocytes are hypersensitive to TCR-mediated signriling in virro and demonstrate
abnormal positive selrction. Thus. both CD4-associated Lck and CD5 play major roles in
modulating TCR-mediated signal ~ansduction during thymic developrnent.
In addition to negatively replat ing surhce TCR expression and signaling. Lc k also
appears to positively regulate signaling events induced by TCR and CD4 coaggregation.
Consistent with previous reports (Gilliland et al.. 199 1 : T urka et al.. 199 1 a: Wiest et al.,
1996). lck +/+ DP thymocytes undergo a weak increase in TCRlCD3-mediatrd protein tyrosine
phosphorylaticn. but coaggregütion of the TCR with CD4 permits the cells to undergo strong
protein tyrosine phosphorylation (Figure III-6A). Interestingly. TCR stimulation induces Ick
-I- DP thyrnocytes to undergo strong protein tyrosine phosphorylation suggesting that TCR
signaling is improved in the absence of Lck. In contrast to Ick +If DP thyrnocytes, CD4-
induced protein tyrosine phosphorylation of various proteins. most prominently the 120/ 130
kDa protein substrate. is impaired in lck 4- DP thymocytes (Figure III-6A). These results
contrast the findings of a recent study which demonstrated that CD3-mediatcd protein tyrosine
phosphorylation by lck -/- thymocytes is reduced compared to lck +It thymocytes (van Oers et
al.. 1996a). The major difference between thrse studies is thrit unfractionated k k +if and lck
4- thymocytes were used in the van Oers et al. study. Results of the van Oers et ai. study
likely reflects CD3-mrdiated protein tyrosine phosphorylation by mature lck +/+ SP
thymocytrs as these cells express superior TCR signaling activity compüred to DP thymocytes
and are presrnt in significantly greater numbers in lck +/+ versus lck -1- mice. Cotisequently,
in the van Oers et al. report. differences between lck +/+ and Zck -1- thymocytes in CD3-
induced protein tyrosine phosphorylation are likely attributed to CD3-rnediated signaling events
by mature SP thymocytes rather than by immature DP thymocytes. Based upon proiein
tyrosine phosphorylation ünalysis. results from this thrsis suggest that Lck is not critical for
TCRlCD3-signaling but is important for CD4-induced signaling in DP thymocytes.
The role of Lck in distal signding events was also examined in this study (Chapter III). in
lck -1- DP thyrnocytes. TCR stimulation rnhanced CD5 and CD69 expression (Figure III-6C)
and decreased RAGl expression (Figure III-6B). albeit thcse responses were slightly less
prominent than in fck +If DP thymocytes. Similürly. the van Oers et al. study observed that
TCR engagement induced CD69 expression in Ick -1- DP thymocytes, although the response
was less efficient compared to [CA- +/+ DP thymocytes (van Oers et al.. 1 W6a). Thus. these
results suggest thÿt both early and distal TCR-mediated signaling rvents c m occur. though
perhaps not optimdly. in the absence of Lck.
F. Significance of thymic stromal cells in negative selection
In this thesis, the hilure to observe TCR-induced ce11 derith of DP thymocytes irr vitro rnay
reflect the intrinsic resistance of these cells to undergo PCD or may relate to the absence of a
cornplete signal provided by thy mic stromal cells. Vürious studies have examined the
importance of thyrnic stromal cells in negative selection. In the rraggrrgate organ culture
system. Jenkinson et al. showed that bone marrow-derived dendritic cells. unlike MHC class
II+ thymic epithelial cells. efficiently delete SEB-reactive VP8+ DP thymocytes (Jenkinson et
al . 199) . LTsing this system. Merkrnshlager et al. demonstrated thüt deletion of autoreactive
anti-HY TCR transgenic thymocytes by male thymic dendritic cells (Merkenshlager et al..
1994) requires fewer stromal cells per thymocyte versus positive selection. which is dependent
upon a 1: 1 ratio of stromal cells and thyrnocytes. Thus. hemütopoietic-denved cells are the
major stromal cells that rnediate self-tolerance, and they likely mediate negative selection by
forming multiple contacts with thymocytes. The interplay between TCR signals and TCR-
independent signals is likely ro be important in mediating thymocyte and stromal ce11
interactions. Accessory molecules expressed by thymic stromal cells müy provide the TCR-
independent signüls necessary to mediate negative selection. Described below are potential
candidate accessory rnolecules that have been implicated in negative selection.
i ) B7lCD28 interactions
The costirnulatory molecule CD28 is expressed by DP thymocytes (Gross et al., 1992).
and expression of its ligands B7-1 and B7-2 is apparent in the thymus (Nelson et al.. 1993;
Degermann et al.. 1994). The role of CD28/B7 interactions in negative selection has been
assessed. but there have been conflicting results. In this thesis. TCR and CD28 engagement
failed to increue apoptosis in both DP suhsets. In contrast, another study drmonstrated that
TCR and CD28 costimulation enhances üpoptosis in DP thymocytes (Punt et al., 1994). The
basis for these different results is not clear, but others found that irz ~ i v o administration of the
CD28/B7 antügonist. CTLA4Ig. does not inhibi t deletion of autoreactive TCR transgenic
thymocytes or Mls-reüctive cells in negatively selecting mice (Tan et al.. 1992: Jones et al.,
1993). Funhermore, negative selection of both autoreactive TCR transgenic thymocytes. as
well as Mls-reactive cells, is unaltered in CD28 -1- mice (Walunas et al.. 1996). but other
costimulütory molecules may substitute for CD28 in self-tolerance. To elirninate this problem,
a recent study exarnined the role of CD28lB7 interactions by studying deletion of TCR
transgenic DP thymocytes cultured in vitro with specific peptide presented by MHC class I-
transfected Drospllilu cells (Kishimoto et al., 1996). 87 expression by MHC I+ Drosphilu
cells correlated with induction of apoptosis of TCR transgenic thy mocytes. Collectively. these
results indicate that the role of CDWB7 interrictions in negative selection requires further
analysis.
i i ) gp39KDJO interactions
Since pp39KD.IO interactions are essential in T cell-dependent B ceIl responses and CD40
is expressed by thy mic cpithelial cells and drndritic cells. the importance of thrsc interactions
was recently assessed in ü comprehensive study using various models of negative selection
(Foy et al.. 1995). Analysis of mice treated with gp39 specific antibodies and in ,13739 -l- mice
demonstrated that delet ion of Mls-reactive thymocytes as weli as autoreactive TCR transgenic
cells is impaired in negatively selecting rnice when gp39KD40 interactions are prevented (Foy
et al.. 1995). Thus. these results suggest that gp39KD40 interrictions and/or costimulatory
molecules induced by CD40 signaling have an important role in negative selection.
i i i ) LFA-l/ICAM-1 and CD2/LFA-3 interactions
Since LFA- I is rxpresssd by DP thymocytes whereas its ligand. ICAM- 1. is expressed
throughout the thymus, the contribution of LFA- IIICAM- 1 interactions to negative selection
hüs been üssessed by other studies. An in ~litro mode1 assessed the fate of anti-HY TCR
transgenic thymocytes in response to male dendritic crlls (Ciirlow et al.. 1992) and
demonstrüted that Ni iitro drletion depends upon LFA- IIICAM-1 interactions. Others showed
that the combined expression of B7- 1 and ICAM- 1 is required Tor irz virro clonal deletion of
TCR transgenic thymocytes cultured in the presencr of specific peptides and MHC class I-
transfected Drosphilo cells ( Kishimoto et al., 1996). Thus. LFA- MCAM- 1 interactions
significantly contribute in rnediating clonai deletion of thymocytes. In contrast. the role of CD2
interactions do not have the same importance in negative selection (Killeen et al.. 1992) since
clonal deletion of thymocytes bearing either MHC class I and class II-restricted transgenic
TCRs is unaffrcted in CD2 '1- rnice. Thus. CD2 likely does not play a critical role in negative
selection, but another molecule may substitute for CD2 in this event.
iv) CD30 interactions
Others have suggested that CD30. a member of the tumor nrcrosis hctor receptor farnily of
proteins involved in triggering ce11 drath. may play a role in negative selection ( Arnakawa et
al.. 1996). In CD30 -1- mice bearing various TCR transgenes. the results suggested that CD30
has an effect in the clonal deletion of either anti-HY TCRaP transgenic ûr ünti-Thb TCRyG
transgenic thymocytes in negatively selecting mice ( Amakawa et al.. 1996). In contrat. CD30
does not appear to be required for clonai deletion of thymocytes reactive for the endogenous
superantigen MM? The authors of the Amakawa et al. study ülso concluded that CD3-
induced death of DP thymocytes is irnpaired in CD30 rnice. but this interpretation is
questionable. The study assessed apoptosis by culturing lin frac tionated t hy mocytes from
CD30 +/- and CD30 -1- mice for 24 hr in the presence of immobilized anti-CD3 and then
determining ce11 viability as well as DNA fragmentation by agarose gel electrophoresis. üsing
a similar assay to assess apoptosis, results from this thesis (Tables 11-1 and 11-7) demonstrated
that mature SP thymocytes. rather than immature DP thyrnocytes. are sensitive to TCRfCD3-
induced apoptosis iri vitro. Hrnce, the Arnükawa et al. study should compare the response of
purified DP cells from CD30 4- and CD30 -1- mice to CD3-induced apoptosis in order to
assess whether CD30 -1- DP thymocytes are l e s susceptible to CD3 specific antibody-mediated
apoptosis versus CD30 fi- DP thymocytes. Nrvertheless. results from this study rnay suggest
a potential role for CD30 in thymic tolerance.
3. Conclusions
This thesis documents a number of noveI findings. First, Fyn was demonstrated to
mediate development of DP thymocytes in Ick -1- mice. because the DN to DP transition was
partially irnpaired in lck -1- mice but was virtually arrested in Ick -1- mice. In addition,
the TFF trmsgene complrtely restored the generation of DP thymocytes in Ick -1- mice,
suggesting that Fyn c m mediate pre-TCR signnling in Ick -'- mice. Second, the TFF
transgene. unlike the activated fck transgene. failed to overcome the DN to DP developmental
arrest in RAG 4- micr, suggesting that activated Fyn cannot bypass the need for a pre-TCR
cornplex. One possible explmation for this result may retlect different intracellular localization
patterns of Fyn and Lck in T cells, leading to regulation of distinct substrates invoived in pre-
TCR signaling. Third. Lck was shown to be imponant in the DP to SP transition, and this Lck
function was panially replacrd by the TFF transgene. Finally. TCR engagement of DP
thymocy tes 0 1 i+tm induçed a number of phenotypic changes that accompany positive
selection. These include increüsed expression of CD5 CD69. and Bcl-3. a reduction in RAGI
and pre-Ta expression. and a switch in lck promoter usage. However, TCR-stimulated DP
thyrnocytes hiled to undergo al1 aspects of t hymic maturat ion because c lonal deletion,
CD4/CD8 lineiige cornmitment, and changes in Thy-1. HSA. MHC class 1. and CD45RB,
were not observed. Thus. TCR engagement alone failed to promote the complete maturation
prograrn in DP thymocytes. Consequently. these results may retlect the jack of thymic stromal
cells in the in ivitro mode1 system since these celIs have been shown to express various
accessory molecules thnt are important for both positive and negative selection. Thus, future
studies should üssess the response of DP thymocytes to both immobilized TCR specific
antibodies, as well as to thymic stromal cells.
Chapter V References
Abraham, K. M., Levin, S . D., iMarth, J. D., Forbush, K. A., and Perlmutter, R. M. ( 199 1 a). Delayed thymocyte development induced by augmentrd expression of p561ck. J. Exp. Med. 173, 142 1 - 1432.
Abraham. K. M., Levin, S. D., Marth, J. D., Forbush, K. A., and Perlmutter, R. M. (199 1 b). Thymic tumorigenesis inducrd by overexpression of p56l'k Proc. Natl. Acad. Sci. USA 88, 3977-398 1.
Abraham, N., Miceli, M. C.. Parnes, 3. R., and Veillette, A. ( 199 Ic). Enhancement of T ce11 responsiveness by the lymphocyte-specific tyrosine protein kinase, p56lck. Nature 350, 62-66.
Alberola-Ila. J.. Forbush. K. A.. Seger, R., Krebs. E. G., and Perlmutter, R. M. (1995). Selective requirement for MAP kinase activation in thymocyte differentiütion. Nature 373,620- 623.
Alberola-Ila, J., Hogquist. K. A.. Swan, K. A., Bevan, M. J.. and Perlrnutter. R. M. (1996). Positive and negative selection invoke distinct signaling pathways. J. Exp. Med. 184. 9- 18.
Alberola-Ila, J., Takaki. S.. Kerner, J. D., and Perlmutter, R. M. (1997). Differential signaling by lymphocyte antigen receptors. Annu. Rev. Immunol. 15, 125- 154.
Arnakawa, R., Hakem, A.. Kundig, T. M., 'Matsuyama. T., Simard, J. J. L.. Timms, E., Wakeharn, A., Mittruecker, H. W.. Griesser, H., Takirnoto, H.. Schmits. R.. Penninger, J. M., and Mak. T. W. ( 1996). Irnpüired negative selection of T cells in Hodgkin's disease antigen CD30-deîïcient rnice. Ce11 84,55 1-562.
Anderson. G.. Owen. J. J. T., Moore. N. C., and Jenkinson, E. J. ( 1994a). Thymic epithelial cells provide unique signals for positive selection of CD4+CDgf thymocytes in vitro. J. Exp. Med. 179, 2037-203 1.
Anderson, G., Moore, N. C., Owen, J. J. T., and Jenkinson. E. J. ( 1996). Celluiar interactions in thymocyte developrnent. Ann. Rev. Immunol. 1-1.73-99.
Anderson, S. J.. and Perlmutter, R. M. (1995). A signaling piithway governing early thymocyte maturation. Immunol. Todiiy 16, 99- 105.
Anderson, S. J.. Abraham, K. M.. Nakayama. T., Singer. A.. and Perlrnutter. R. M. (1992). Inhibition of T-ce11 receptor P chain remangement by overexpression of the non-receptor protein tyrosine kinase pj61ck. EMBO J. 11,1877-4886.
Anderson, S. J.. Levin, S. D., and Perlmutter, R. M. ( 1993). Protein tyrosine kinase p561ck controls allelic exclusion of T-ce11 receptor P-chain genes. Nature 365.532-554.
Anderson, S. J., Levin, S. D., and Perlmutter. R. M. (1994b). Involvement of the protein tyrosine kinase p561ck in T cell signaling and thymocyte development. Adv. Immunol. 56, 15 1-178.
Antica. M., Wu, L.. Shortman, K., and Scollay. R. ( 1993). Intrathymic l ymphoid precursor cells during fetal thymus development. J. Immunol. 151, 5887-5895.
Appleby, M. W., Gross, J. A.. Cooke, M. P., Levin, S. D.. Qian. X., and Perlmutter. R. M. ( 1992). Defective T ce11 receptor signaling in mice lacking the thymic isoforrn of p59f? Ce11 70, 75 1 -763.
Arpaia. E.. Shahar, M., Dadi. H.. Cohen, A., and Roifman, C. ( 1994). Defective T ce11 receptor signding and CD8+ thymic selection in humans lacking ZAP-70 kinase. Ce11 76.947- 958.
Artavanis-Tsakonas, S.. Matsuno, K., and Fortini. M. E. ( 1995). Notch signaling. Science 268, 225-232.
Asarnow. D. M.. Cado. D.. and Raulet. D. H. (1993). Selrction is not required to produce invariant T-ce11 receptor gamma-gene junctional sequences. Nature 362. 158-60.
Ashton-Rickardt. P. G., Kaer. L. V.. Schumacher. T. N. M.. Ploegh. H. L., and Tonegawa, S. ( 1993). Peptide contributes to the speciticity of positive selection of CDSç T cells in the thymus. Cell 73, IO4 1 - IO49.
Ashton-Rickardt, P. G.. Bandeira, A., Delaney, J. R.. Van Kaer. L., Pircher, H.-P., Zinkernagel. R. M., and Tonegawa, S. ( 1994). Evidence for a differential avidity mode1 of T ce11 selection in the thymus. Ce11 76. 65 1-663.
Ashwell, J. D., and Klausner, R. D. ( 1990). Genetic and mutational anülysis OF the T ceil üntigen receptor. Annu. Rev. Immunol. 8, 139- 167.
Baron. A., Hakn. K., and von Boehmer, H. ( 1994). A human CD4 transgene rescues CD4- CD8+ cells in 02-microglobulin-detïcient mice. Eur. J. Immunol. 24. 1933- 1936.
Barthlott, T., Potocnik, A. J . . Kohler, H., Carsetti, R.. Pircher, W., Fowlkes, B. J., and Eichmann. K. ( 1996). A novel mouse thymocyte antigen (F3Ag): down-regulation during the CD4+CD8+ double-positive stage indicates positive srlection. Int. Immunol. 8. 10 1 - 1 13.
Bendelac, A.. Matzinger, P.. Srder. R. A.. Paul. W. E.. and Schwartz. R. H. (1992). Activation events during thyrnic selrction. J. Exp. Med. 175. 73 1-742.
Benoist, C., and Mathis. D. ( 1989). Positive selection of the T ce11 repertoire: Where and when does it occur? Ce11 58, 102% 1033.
Berg, L. J., Füzekas de St. Groth. B.. Pullcn, A. LM.. iind Davis. M. M. (1989a). Phenotypic differences between ap versus f3 T-ce11 receptor transgenic mice undergoing negative selection. Nature 340. 559-562.
Berg, L. J., Pullen. A. M.. Fazekas de St. Groth. B., Mathis, D.. Benoist. C.. and Davis, M. M. ( 1989b). AntigenlMHC-specific T cells are preferentially exported frorn the thymus in the presence of their MHC ligand. Ce11 58. 1035-lû46.
Berg. L. J.. Frank, G. D.. and Davis. M. M. (1990). The effects of MHC gene dosage and allelic variation on T ce11 receptor selection. Ce11 60, 1043- 1053. . Bergman. M., Mustelin. T.. Ortken, C.. Partanen. J.. Flint. N. A., Amrien. K. E., Autero, M., Burn, P.. and Alitalo, K. (1992). The human p50'sk tyrosine kinase phosphorylates p56[~k at tyr-505 and down-replates its catalytic activity. EMBO J. 11,29 19-2924.
Bevan, J. (1977). in a radiation chimaera. host H-2 antigens drtcrmine immune responsiveness of donor cytotoxic cells. Nature 269.4 1 7-4 19.
Bill, J.. and Palmer. E. ( 1989). Positive selection of CD4+ T crlls mediated by MHC class U- bexing svornal ce11 in the thymic cortex. Nature 341,649-63 1.
Bix, M.. and Raulet. D. (1992). Inefficient positive selection of T cells directed by haematopoietic cells. Nature 359. 330-333.
Blackman, M.. Kappler. I.. and Marrack, P. ( 1990). The role of the T ce11 receptor in positive and negative sclection of developing T cells. Science 248. 1335- 134 1.
Bluthmann, H.. Kisielow. P.. Uematsu, Y., Malissen. M.. Krimpenfort, P.. Berns, A., von Boehrner. H.. and S teinrnetz, M. ( 1988). T cell-specific deletion of T-ce11 receptor transgenes allows functional remangement of endogenous a- and P- genes. Nature 334. 156- 159.
Boismenu, R., Rhein, M., Fischer. W. H., and Havran, W. L. ( 1996). A role t'or CD81 in early T ceIl development. Science 171. 198-200.
Bonomo, A., and iMatzinger. P. ( 1993). Thymus epitheliurn induces tissur-specific tolerance. J. Exp. Med. 177. 1 133- 1 164.
Borgulya. P.. Kishi. H.. Uematsu. Y.. and von Boehmer. H. ( 1992). Exclusion and inclusion of a and p T ce11 receptor alleles. Ce11 69. 529-537.
Born. W.. Happ, M. P.. Dalias. A., Reardon, C.. Kubo. R.. Shinnick. T.. Brennan. P.. and O'Brien. R. ( 1990). Recognition of heat shock proteins and $5 ce11 function. Irnmunol. Today 11 , 40-43.
Bosma. G. C.. Custer, R. P.. and Bosma, LM. J . (1983). A severe combined imrnunodrticiency mutation in the mouse. Nature 301. 527-530.
Boyd. R. L.. Tucek. C. L.. Godfrey. D. 1.. Izon, D. J.. Wilson. T. J., Davidson, N. J. , Bean. A. G. D.. Ladyrnan, H. M.. Ritter, M. A.. and Hugo. P. (1993). The thymic microenvironment. Immunology Today 14,445-459.
Briindle. D.. ~Müller. C.. Rülicke, T.. Hrngartner. H.. and Pircher, H. ( 1992). Engagement of the T-ce11 receptor durins positive selection in the thymus down-regulates RAG- 1 expression. Proc. Natl. Acad. Sci. LISA 89, 9529-9533.
Brandie. D.. Müller. S., Müller. C. . Hengartner, H., and Pircher. H. ( 1994). Regulation of RAG-1 and CD69 expression in the thymus during positive and negative selection. Eur. J. Immunol. 23. 145- 15 1.
Cürlow, D. A.. van Oers. N. S. C . . Teh, S.-J., and Teh, H.-S. (1997). Deletion of antigen- specific immature thymocytes by dendritic cells requires LFA- IIICAM interactions. J. Immunol. 138, 1595- 1603.
Carrera, A. C.. Baker, C.. Roberts. T. M., and Pardoll. D. M. ( 1992a). Tyrosine kinase triggering in thymocytes undergoing positive selection. Eur. J. Irnmunol. 22.1289-2294.
Carrera. A. C., Wee, S.-B.. Roberts. T. M., and Pardoll, D. M. ( 1992b). Selective association between MHC class I-restricted T ceIl receptors, CD8. and activated tyrosine kinases on thymocytes undegoing positive selection. J. Immunol. 149, 3 112-3 149.
Ceredig, R., and Schreyer. M. ( 1984). Immunohistochernical localization of host and donor- derived cells in the regenerating thymus of radiation bone rnarrow chimeras. Thymus 6. 15-26.
Chan. A. C. . Iwashima. M., Turck. C . W.. and Weiss. A. (1991). ZAP-70: A 70 kd protein- tyrosine kinase that associates with the TCR 5 chüin. CeII 71.649-662.
Chan. A.. Kadlecek. T.. Elder. M.. Filipovich. A.. Kuo-W.-L.. Iwashirna. M.. Parslow, M., and Weiss, A. ( 1994a). ZAP-70 deficiency in an autosornai recessive form of severe combined irnrnunodeficiency. Science 264, 1599- 160 1.
Chan, A. C.. van Oers, N. S. C.. l'ran. A.. Turka, L.. Law, C L . . Ryan, J. C., Clark, E. C., and Weiss. A. ( 1994b). Differential expression of ZAP-70 and S y k protein tyrosine kinases, and the rolr of this farnily of protein tyrosine kinases in T ceIl iintigen receptor signaling. J. Immunol. f 52, -17584766.
Chan, S. H.. Cosgrove, D., Waltzinger. C.. Benoist. C.. and Müthis. D. (1993). Another view of the selective rnodel of thymocytr selection. CeIl 73. 735-736.
Chan. S. H.. Waltzinger. C.. Baron. A.. Benoist. C., and Mathis. D. ( 1994~) . Role of coreceptors in positive selection and lineage cornmitment. EMBO J. 13,4482489.
Chao, D. T.. Linette. G. P.. Boise. L. K.. White. L. S.. Thornpson, C. B.. and Korsmeyer, S. J. (1995). Bcl-xL and Bcl-2 repress a common püthway of ceil death. j. Exp. Med. 182, 82 1-828.
Cheng. A. LM., Rowiev. B.. Pno. W.. Hayday. A.. Bolen. J. B.. and Püwson. T. (1995). Syk tyrosine kinase requiréd for rnouse viability and B-ce11 devrlopment. Nature 378. 303-306.
Chow. L. M. L.. Fournel, M.. Davidson. D.. and Vrillette. A. ( 1993). Negative regularion of T-ce11 receptor signüiling by tyrosine protein kinase p 5 0 ~ . ~ ~ . Nature 3665. 156- 160.
Chun. J. J. M.. Schatz. D. G.. Oettinger. M. A.. Jaenisch. R.. and Baltimore. D. ( 1991). The recombination activating gene-l (RAG- 1 ) transcript is present in the murine central nervous system. Ce11 64. 189-700.
Cooke, M. P., and Perlmutter, R. M. ( 1989). Expression of ii novel form of thejyrz proto- oncogene in hematopoittic cells. The New Biologist 1. 66-74.
Cooke, M. P., Abraham, K. 1V.. Forbush, K. A., and Perlmutter. R. LM. (1991). Regulation of T ce11 receptor signaling by a src hmily protein-tyrosine kinase p5qi.11. CeIl 65. 28 1-29 1.
Coma, I., Bix. M.. Liao. N. S., Zijlstra, M.. Jaenisch, R., and Raulet. D. ( 1991). Most y6 T cells develop normally in beta 2-microglobulin-deficient rnice. Proc. Natl. Acad. Sci. USA 89, 653-657.
Cosgrove. D.. Chan. S. H.. Waltzinger, C.. Benoit. C.. and Mathis, L. D. (1992). The thymic cornpartment responsible for positive selection of CD4f T cells. Int. Immunol. 4, 707- 7 10.
Cosulich, M. E., Rubartelli. A., Risso, A., Cozzolino. F.. and Bargellesi. A. (1987). Functional characterization of an antigen involved in an early step on T-ce11 activation. Proc. Natl. Acad. Sci. USA 8442054209.
Crisa, L., Cirulli, V., Eilisman, M. H., Ishii, 1. K., Elices, M. J., and Salomon, D. R. (1996). Ceil adhesion and migration are regulated at distinct stages of thymic T ce11 development: the roles of fibronectin , VLA-5, and VLA-5. J. Exp. Med. IM. 2 15-228.
Crispe, 1. N., Moore, M. W.. Husmann, L. A.. Smith. L.. Bevan. M. J., and Shimonkevitz ( 1987). Differentiation potentiül of subsets of CD4-CD8- thymocytes. Nature 329. 336-339.
Crompton, T.. Gilmour, K. C.. and Owen, M. 1. (1996). The MAP kinase pathwüy controls differentiation from double-negative to double-positive thymocyte. CeIl 86. 243-25 i.
Crurnp, A. L., Grusby, M. J., Glimcher, L. H.. and Cantor. H. ( 1993). Thymocyte development in major histocornpatibility cornplex-defiçient mice: Evidence for stochastic cornmitment to the CD4 and CD8 lineages. Proc. Natl. Acad. Sci. USA 90. 10739- 10743.
Curnow, S. J., Barad, M., Brun-Roubereau, N., and Schmitt-Verhulst, A.-M. ( 1994). Flow- cytometric ünalysis of apoptotic and nonapoptotic T-ceIl receptor-transgenic thymocytes following in vitro presentation of antigen. Cytometry 16.4 1-18.
Danielian. S., Fagard, R., Alcover, A., Acuto, O.. and Fischer, S. (1991). The tyrosine kinase activity of p56" is increüsed in human T cells activated via CD?. Eur. J. Immunol. 21, 1697- 1970.
Danielian, S., Alcover, A.. Polissard, L., Stefanescu, M., Acuto, O., Fischer, S., and Fagard. R. ( 1992). Both T ce11 receptor (TcR)-CD3 complex and CD2 increüse the tyrosine kinase activity of ~ 5 6 ~ ' ~ : CD2 can mediate TcR-CD3-inde pendent and CDG-dependent activation of p56lCk. Eur. J. Imrnunol. 72. 19 15-292 1.
Danska, J. S., Ptlumio, F., Williams. C., Huner. O., Dick, J. E., and Guidos. C. J. ( 1994). Rescue of T cell-specific V(D)J recombination in SCID mice by DNA-damaging agents. Science 266, 450-455.
Davidson. D., Chow. L. M. L., Fournel. M., and Veillcttr. A. ( 1997). Differential regulation of T ce11 üntigen responsiveness by isotorrns of the src -related tyrosine protein kinase p59% J. Exp. Med. 175. 1.183- M E .
Davis, C. B., Killeen, N., Crooks, M. E. C., Raulet. D.. and Littman, D. R. (1993). Evidence for a s tochastic mechanism in the di fferentiation of mature subsets of T lymphocytes. Ce11 73, 237-247.
Davis, C. B.. and Littman. D. R. ( 1995). Disrupted development of thymocytes expressing a transgenic TCR upon CD4 overexpression. Int. Immunol. 7, 1977- 1986.
Degermann, S., Surh, C. D., Glimcher, L. H., Sprent. J., and Lo, D. (1994). B7 expression on thymic rnedullary epithelium correlates with epithelium-mediated deletion of VP5+ thymocytes. J . Immunol. 152, 3254-3263.
Droge, W. ( 198 1). Theoretical considerations and probability rnodels for the somatic development of the T ce11 repertoire. Ce11 Immunol. 57, 25 1-264.
Dudley, E. C., Petrie, H. T., Shah. L., Owen, M. J . , and Hayday, A. C . ( 1994). T ce11 receptor B chain gene rearrangement and srlection during thymocyte development in adult mice. Immunity 1, 83-93.
Dutz, J. P.. Ong. C. J.. Marth, J., and Teh, H.-S. ( 1995). Distinct differentiütive stages of CD4+CD8+ thymocyte development defined by the lac k of coreceptor binding in positive selection. J. Immunof. 154, 388-2599.
Elder. M., Lin. D., Clever, J.. Chan, A., Hope, T., Weiss. A.. and Parslow. T. ( 1994). Human severe combined immunodeficiency due to a defect in ZAP-70, a T ce11 tyrosine kinase. Science 264. 1596- 1599.
Ericsson, P.-E.. and Teh. H.-S. ( 1995). The protein tyrosine kinase p561'k regulates K R expression and T ceIl selection. Int. Imrnunol. 7,6 17-621.
Ernst, B. B.. Surh. C. D., and Sprent. J. (1996). Bone marrow-derived cells tail to induce positive selection in thymus reaggregation cultures. J. Exp. Med. 183. 1235- 1210.
Faure, F.. Jitsuküwa. S., Misossec. C., and Hercend, T. (1990). CDlc as a target recognition structure for human T lymphocytes: analysis with peripheral blood $3 ceils. Eur. .J. Imrnunol. 30, 703-706.
Fehling, K. J., Krotkovü, A.. Saint-Ruf. CC.. and von Boehmcr. H. ( 1993). Cr~icial role of the pre-T-ce11 receptor a gene in development of up but not y6 T cells. Nature 375. 795-798.
Finkel, T. H., Crimbier, J . C., Kubo, R. T., Born. W. K., Marrack. P., and Kappler, J. W. (1989). The thymus has two fi~nctionülly distinct populations of immature up+ T cells: one population is deleted by ligation of uPTCR. Ce11 58. 1047- 1054
Finkel, T. H.. Kappler. J. W., and Marrack. P. C. ( 1997). Immature thy mocytes are protected from deletion early in ontogeny. Proc. Natl. Acad. Sci. USA 89. 3372-3374.
Fischer, K.-D.. Zmuidzinas, A., Gardner, S., Barbacid. M., Bernstein. A.. and Guidos, C. ( 1995). Defective T-ceIl receptor signaling and positive selection of Vav-detïcirnt CDJf CD8+ thymocytes. Nature 374.374477.
Fowlkes, B. J.. and Pardoll. D. LM. (1988). Molecular and cellular events of T ce11 development. Adv. Imrnunol. 44, 207-264.
Fowlkes, B. J., Edison, L., Mathieson, B. J., and Chused. T. M. ( 1985). Early T lymphocytes. Di fferentiation in vivo of adult intrathymic precursor cells. J. Exp. Med. 162, 803-823.
Foy, T. M., Page, D. M.. Waldschmidt, T. J.. Schoneveld. A.. Laman, J. D.. Masters, S. R., Tygrett, L.. Ledbetter. J. A., Aruffo, A.. Claassen. E.. Xu, J. C., Flüvell. R. A.. Oehen, S., Hedrick, S. M.. and Noeile. R. J . (1995). An essential role for gp39. the ligand for CD40, in thymic selection. J. Exp. Med. 282. 1377- 1388.
Gao, E X . , Lo, D.. and J.. S. ( 1990). Strong T ce11 tolerance in parent to FI bone marrow chimeras prepared with supralethal irradiation. J. Exp. Med. 171. 1 10 1 - I 12 1.
Gauen. L. K. T., Kong, A.-N. T., Samelson. L. E.. and Shaw. A. S. (1992). p59/?rl tyrosine kinase associates with multiple T-ce11 receptor subunits through its unique amino-terminal domain. Molec. Cell. Biol. 12. 5438-5346.
Gilliland, L. K., Teh, H.-S., Uckun, F. M., Norris, N. A., Teh, S.-J., Schieven, G. L., and Ledbetter. J. A. (1991). CD4 and CD8 are positive regulators of T ce11 receptor signal transduction in early T ce11 di fferentiation. J. Immuool. 146. 1 759- 1 765.
Godfrey. D. I.. Kennedy. J.. Suda. T., and Zlotnik, A. ( 1993). A developmental pathway involving four phenotypically and funciionally distinct subsets of CD3-CD4-CD% triple- negative adult mouse thymocytes detïned by CD44 and CD15 J. Immunol. 150.42444253.
Gravestein, L. A., van Ewijk. W.. Ossendorp, F.. and Borst. J. ( 1996). CD27 cooperates with the pre-T ce11 receptor in the replation of murine T ce11 development. J. Exp. iMed. 184, 675-685.
Groettrup, M.. Baron. A., Griffiths. G.. Palacios. R., and von Boehmer. H. (1992). T ceIl receptor (TCR) P chain homodimers on the surface of immature but noi mature a, y. and 6 chain deficient T ce11 lines. EMBO J. 11,2735-2746.
Groettrup. M.. Ungeweiss. K.. Azogui. O.. Palacios. R., Owen. M.. Haydliy. A.. and von Boehrner, H. ( 1993). A novel disulfide-linked heterodimer on pre-T cells consists of the T ceil receptor P chüin and a 33 kd glycoprotein. Ce11 75. 333-294.
Gross. J., Callas, E., and Allison, J. P. (1992). Identification and distribution of the costirnulatory receptor CD28 in the mouse. J. Immunol. 119. 380-388.
Groves, T., Katis, P.. Madden, 2.. Mrinickam, K., Ramsden, D., Wu. G., and Guidos, C. J. ( 1995). In vitro maturation of clonal CD4+CD8f ce11 lines in response to TCR engagement. J. Immunol. 154. 50 1 1-5022.
Groves, T.. Parsons. LM.. ~Miyamoto. N., and Guidos, C. 1. (1997). TCR engagement of CD4+CD8+ thymocytçs in vitro induces early aspects of positive selection. but not apoptosis. J. Immunol. 158, 65-75.
Guidos. C. J. ( 1996). Positive selrction of CDJf and CD8+ T cells. Curr. Opin. Immunol. 8. 225-232.
Guidos. C. J.. and Weissman. 1. L. ( 1993). Sequential occurrence of positive and negative selection during T lymphocyte maturation. In Molrcular Mechanisms of Immunological Self- Recognition. F. Alt and H. J. Vogel. eds. (San Diego. CA: Aciidemic Press. Inc.). pp. 137- 147.
Guidos, C. J.. Weissman, I. L.. and Adkins. B. ( 198%). Developmental potential of CD4-8- thymocytes: Peripheral progeny include mature CD4-8- T-cells bearing np TCR. J. Immunol. 143, 3773-3780.
Guidos, C. J., Weissman, 1. L., and Adkins, B. (1989b). Intrathyrnic maturation of murine T lymphocytes from CD8+ precursors. Proc. Natl. Acad. Sci. USA 86. 7542-7546.
Guidos, C. J.. Danska. J. S.. Fathrnan, C. G.. and Weissman, 1. L. ( 1990). T ce11 receptor- mediated negative selection of autoreactive T lymphocyte precursors occurs after cornmitment to the CD4 or CD8 lineages. J. Exp. Med. 172. 835-845.
Guidos. C. J.. Williams. C. J . , Wu. G. E.. Paige. C. J.. and Danska. J . S. (1995). Developmen; of CD4+CDS+ thymocytes in RAG-deiicient mice through a T ce11 receptor P chain-independent pathway. J. Exp. Med. 181. 1 187-1 195.
Haas, W., Pereira. P.. and Tonegawri, S. ( 1993). Garnrnaldeltn cells. Annu. Rev. Immunol. 11. 637-685.
Hara. T., Jung. L. K. L.. Bjorndahl, J. LM.. and Fu, S. LM. (1986). Humün T ceIl activation III. Rapid induction of a phosphorylated 28W32kDa disulfide-linked eürly activation antigen (EA 1 ) by 12-o-tetradecanoyl phorbol- 13-acetüte. mitogens. and antigens. J. Exp. Med. 161, 1988-2005.
Hashimoto. K.. Sohn. S. J.. Levin. S. D., Tado, T.. Perlmutter. R. M.. and Nakayüma. T. ( 1996). Requirernent for p56rcP tyrosine kinase activation in T ce11 receptor-mediüted thyrnic selection. J. Exp. iMed. 183. 93 1-943.
Hasserjian, R.. Aser. J., Davi. D.. Weinberg, D.. and Sklür, J. ( 1996). Modulated expression of Notch I during thymocyte development. Blood 88.970-976.
Hatakeyama, M.. Kono, T.. Kobayashi. N.. Kawahara. A.. Levin. S. D., Perlmutter. R. M., and Taniguchi, T. ( 199 1 ). Interaction of the IL-2 receptor with the src family kinase p561ck Identification of a novel intermolecular association. Science 352. 1523- 1528.
Havran. W. L.. and Allison. J. P. ( 1988). Developmentdly ordered appearrince of thymocytes cxpressing different T-ce11 antigen receptors. Nature 335,44345.
Havran, W. L., Poenie, LM.. Kirnura. J., Tsien, R.. Weiss. A., and Allison. J . P. (1987). Expression and funct ion of the CD3-ant igen receptor on murine CD4+8+ thy mocytes. Nature 330, 170- 1 73.
Hoffmann, M. W., Heath, W. R., Ruschrneyer, D., and Miller, J. F. A. P. (1995). Deletion of high avidity T çells by thymic epithelium. Proc. Natl. Acad. Sci. USA 92.985 1-9855.
Hogquist. K. A., Gavin, M. A., and Bevan, M. J. (1993). Positive selection of CD8+ T cells induced by major histocompatability complex binding peptides in fetal thymic organ cultures. J. Exp. Mrd. 1993. 1469-1473.
Hopquist. K. A.. Jameson, S. C.. Heath, W. R.. Howard. J. L.. Bevan. M. J.. and Carbone, F. R. ( 1994). T ce11 receptor antagonisi peptides induce positive selection. Cell 76, 17-27.
Hogquist. K. A., Jameson, S. C.. and Bevan. M. J . (1995). Strong agonist ligands for the T cell receptor do not rnediate positive selection of functional CD8+ T cells. Immunity 3,79-86.
Hugo. P.. Boyd. R. L.. Waanders, G. A.. Petrie. H. T.. and Scollay, R. (1990). Timing of deletion of autoreactive Vp6+ cells and down-modulation of either CD4 or CD8 on phenotypically distinct CD4+8+ subsets of thymocytes cxpressing intermediate or high levels of T ce11 receptor. Int. Imm~inol. 3, 265-272.
Hugo, P.. Kappler, J. W., Godfrey, D. I., and Marrück, P. ( 1993). Fibroblasts can induce thymocyte positive selection in vivo. Proc. Nütl. Acad. Sci. USA 90. 10335- 10339.
Hurley, T. R., Luo, K.. and Sefton, B. M. (1989). Differential effects of expresion of the CD45 tyrosine protein phosphatase on the tyrosine phosphorylation of the ick. fyn and c-src tyrosine protein kinases. Mol. Cell. Biol. 13. 165 1- 1656.
Irving, B. A.. and Weiss. A. ( 199 1). The cytoplasrnic domain of the T ce11 receptor 6 chain is sufficient to couple to receptor-associated signal transduction pathways. Ce11 64. 89 1-90 1.
Itaharo. S., ;Mombaerts, P.. Lahille. J.. lacomini. J., Nelson, A., Clarke, A. R.. Hooper. M. L., Farr, A., and Tonegawa, S. (1993). T ce11 receptor 8 gene mutant mice: independent generation of ab T cells and programmed rearrangements of y6 TCR genes. Ce11 72, 337-348.
Itano, A., Kioussis. D., and Robey, E. ( 1994). Stochastic component to development of class I major histocornpatibility cornplex-specitic T cells. Proc. Natl. Acrid. Sci. USA 91, 220-224.
Itano, A., Salmon. P.. Kioussis, D., Tolaini, iM., Corbella. P.. and Robey. E. ( 1996). The cytoplasmic domriin of CD4 promotes the development of CD4 lineage T cells. J. Exp. Med. 183, 73 1-74 1.
Ito, K., van Kaer. L.. Bonneville, M., Hsu, S., Murphy, D. B., and Tonegüwa. S. (1990). Recognition of the product of a novet MHC TL region gene (27b) mouse T ce11 receptor. Ce11 62, 549-56 1.
Jacobs, H.. Vandeputte, D.. Tolkamp, L.. de Vries, E.. Borst. J.. and Bems. A. ( 1994). CD3 components at the surface of pro-T çells can mediate pre-T ce11 development il1 vivo. Eur. J. Immunol. 24, 934-939.
Jacobs. H., lacomini. J.. and van de Ven, M. ( 1996). Domains of the TCRp-chah required for early thymocyte development. J. Exp. Med. 184. 1833-1843.
Jarneson, S. C . . Hogquist, K. A.. and Annu. Rev. Immunol. 13. 93-126.
Janeway, C., Rudensky, A.. Rath, S., needle's eye. Curr. Biol. 2 , 26-28.
Bevan. M. J. ( 1995). Positive srlrction of thyrnocytes.
iMuiphy, O. ( 1992). It is easier for a carne1 to pass the
Jenkinson, E. J., Anderson. G., and Owen, J . J. T. ( 1992). Studies on T ce11 maturation on defined thymic stromril ce11 populations in vitro. J. Exp. Med. 176. 845-853.
Jenkinson, E. J., Anderson. G., Moore, N. C., Smith, C. A. S., and Owen, J . J . T. (1994). Positive selection by MHC clriss II+ thymic epithelial cells by B7 are not involved. Dev. lmmunol. 3. 265-27 1.
Jones. L. A., Izon, D. J., Niedlrind, J. D., Linsley. P. S.. and Kr~iisbeek, A. M. ( 1993). CD28-87 interactions are not required for intrrithyrnic clonal deletion. Int. immunol. 5. 503- 5 12.
June, C. H., Bluestone, J. A., Nadler, L. M., and Thompson, C. B. (1994). The B7 and CD28 receptor families. Immunol. Today 15. 32 1-33 1.
Kappler. J. W.. Roehm. N., and Marrack. P. ( 1987). T-cell tolerance by clonal elimination in the thymus. Ce11 49, 273-280.
Kappler, J. W., Staerz, U.. White, J., and Marriick. P. C. ( 1988). Self-toleriince eliminates T cells sprcific for Mls-modified products of the major histocompatibility complex. Nature 332, 35-40.
Kappler, J . W., Kushnir. E., and Murack, P. ( 1989). Analysis of VP17a expression in new mouse strains bearing the Vpa haplotype. I. Exp. Med. 169. 1533- 154 1.
Kawai, K.. Kishihara, K.. ~Molina, T. J., Wallace, V. A., Mak, T. W., and Ohashi, P. S. ( 1995). Impaired drvelopment of Vy3 dendritic epidermal T cells in p56lC' protein tyrosine kinase-detïcient and CD45 protein tyrosine phosphatase-deficient mice. J. Exp. Med. 181. 345-349.
Kearse, K.. Takahama, Y.. Punt. J., Sharrow, S., and Singer, A. (1995). Early molecular events induced by TCR signaling in immature CD4+CD8+ thymocytes. J. Exp. Med. 181, 193- 202.
Kitleen. N.. Stuart. S. G., and Littman. D. R. ( 1992). Development and function of T cells in mice with a dismpted CD2 gene. EMBO J. 11.43294336.
Kishi. H., Borguiya. P.. Scott. B., Karjalainen. K., Traunecker, A.. Kaufman. 1.. and Von Boehmer. H. ( 199 1 ). Surface expression of the p T cell receptor ( K R ) chain in the absence of other TCR or CD3 proteins on immature T cells. EMBO J 10.93- 100.
Kishimoto. H., Surh. C . D.. and Sprent, J. ( 1995). Upregulation of surface markers on dying thymocytes. J. Exp. iMed. I81. 649-655.
Kishimoto. H., Cai, Z., Brunmark, A.. Jackson, M. R.. Peterson, P. A.. and Sprent. J. ( 1996). Diflering roles for B7 iind intercelliilar adhrsion molecule- 1 in negative selection of thymocytes. J. Exp. Mcd. 184. 53 1-537.
Kisielow, P., Bluthmann. H., Sttierz, U. D., Steinmetz, M., and von Boehrner, H. ( 1988). Tolerance in T-ceil receptor transgenic mice involves de letion of nonmature CD4+8+ thymocytes. Nature 333, 743-746.
Kisielow. P.. and Maizek. A. ( 1995). Positive selrction of T cells: Rescue from programmed ce11 death iind differentiation require continua1 engagement of the T ce11 receptor. J. Exp. Med. I S I . 1975-1984.
Kosaka, H., and Sprent. J. (1993). Tolerance of CD8+ T cells developing in Parent to FI chimeras prepared with supralethal irradiation: step-wise induction of tolerance in the intrathymic and extrathymic environments. J. Exp. Med. 177. 367-378.
Kouskoff, V.. Vonesh, J.-L., Benoist, C., and Mathis, D. ( 1995). The influence of positive selection on RAG expression in thymocytes. Eur. J. Immunol. 25. 54-58.
Kruisbeek, A. M., iMond, J. I.. FowIkes. B. J., Carmen, J. A., Bridges, S., and Longo, D. L. ( 1985). Absence of the LytZ-, L3T4+ lineage oPT cells in mice treated neonatally with anti- 1-A correlates with the absence of intrithyrnic L A antigen-presenting ce11 function. J. Exp. Med. 161, 1029-1047.
Kydd. R., Lundberg. K.. Vremec. D.. Harris. A. W.. and Shortman. K. (1995). Intermediate steps in thymic positive selection. Generation of CD4-8+ T crlls in culture from CD4+ 8+ , cD~~"%+. and ~ ~ 1 + 8 i n t thymocytes with up-regulüted levels of TCR-CD3. J. Irnmunol. 155, 3806-35 14.
Kyewski. B. A. ( 1987). Seeding of thymic microenvironment defined by distinct thyrnocyte- stroma1 ce11 interactions is developmentally controlled. J. Exp. Med. 166. 570-538.
Lanier. L. L.. Chang. C. . Spits. H.. and Phillips, I. H. (1991). The developmental relationship between NK cells and T cells. J. Immunol. 119. 1876- 1880.
Laufer. T. M.. DcKoning, J.. Markowitz, J. S.. Lo. D.. and Glimcher, L. H. (1996). Unopposed positive selection and autoreüctivity in mice expressing class II MHC only on thymic cortex. Nature 383, 8 1-85.
Ledbetter, J. A., Norris, N. A., Grossrnann, A., Grosmaire, L. S.. June. C. H.. Uckun, F. M., Cosand, W. L.. and Rabinovitch. P. S. (1989). Enhanced trünsmembrane signalling activity of monoclonal antibody heteroconjugütes suggests interactions between receptors at the T ceIl surface. Mol. Immunol. 26. 137- 145.
Letourneur. F.. and Klausner. R. D. (1992). Activation of T cells by a tyrosine kinase activation domüin in the cytoplasmic tail of CD&. Science 255. 79-82.
Levelt. C. N., and Eichmann. K. ( 1995). Receptors and signals in early thymic selection. Immunity 3.667-672.
Levelt. C.. Ehrfeld. A.. and Eichmann. K. ( 1993a). Replation of thymocyte development through CD3. 1. Tirnepoint of ligation of C D ~ E determines clona1 drletion or induction of developmental prosram. J. Exp. Med. 177. 707-7 16.
Levelt. C. N.. Carsetti. R.. and Eichmann. K. ( 1993b). Regulation of thymocyte development through CD3. II. Expression of T ce11 receptor P CD3e and maturation to the CD4+8+ stage are highly correlated in individual thyrnocytes. J. Exp. Med. 178. 1867- 1875.
Levelt, C. N.. Mombaerts. P.. Iglesias. A.. Tonegawa. S.. and Eichmann. K. ( 1993~) . Restoration of rarly thymocyte differentiation in T-ce11 reccptor p-chain-deficient mutant mice by trmsmembrane signalling through C D ~ E . Proc. Natl. Acad. Sci. USA 90. 1 1401- 1 1405.
Levelt. C. N.. Mombaerts, P.. Wang. B., Kohler. K.. Tonegawü. S.. Eichmann. K., and Terhorst. C. (1995). Regulütion of thymocyte developnient through CD3: Functional dissociation between p561ck and CD35 in early thymic selection. Irnrnunity 3. 115-222.
Levin. S., Anderson, S.. Forbush. K., and Perlrnutter, R. ( 1993). A dominant-negative transgene defines a role for p56fck in thymopoiesis. EMBO J. 12, 167 1- 1680.
Ley, S. C.. Marsh. M., Bebbington, C. R.. Proudfoot. K.. and Jordan, P. ( 1994). Distinct intricellular localization of Lck and Fyn protein tyrosine kinases in human T lymphocytes. J. Cell Biol. 125. 639-649.
Linette. G. P.. Grusby. M. J.. Hedrick. S. M.. Hansen. T. H.. Glimcher. L. H.. and Korsmeyer. S . J. (1994). Bcl-2 is upregulated at the CDJ+CD8+ stage during positive
selection and promotes thymocyte differentiation 205.
Liu, C.-P., Ueda. R., She, J., Sancho, J., Wang, Dudley, E. C.. , A., Terhorst, C., and Huang, M. (1993).-~bnormal T ce11 developrnent in mutant mice and identification of a novrl T ce11 population in the intestine. EMBO J. 12,4863.
137
at several control points. Imrnunity 1, 197-
B., Weddell. G., Loring, J.. Kurahara, E.,
Lo, D., and Sprent, J. ( 1986). ldentity of cells thiit imprint H-2-restricted T-ce11 specificity in the thymus. Nature 31 9. 677-675.
Love, P. E., Shores, E. W., Johnson, M. D., Tremblay, M. L.. Lee. E. J., Grinberg, A., Huang. S. P., Singer, A.. and Wrstphal. H. (1993). T ceIl developrnent in rnice that lack the 6 chah of the T ce11 antigen receptor complex. Science 261.9 18-92 1.
Lowell, C. A.. and Soriano, P. ( 1996). Knockouts of Src-hmily kinases: stiff bones, wimpy T cells, and bad mernories. Genes Dev. 10, 1845- 1857.
Lucas, B.. and Germain, R. N. ( 1996). Unexpectrdly complex regulation of CD4/CD8 coreceptor expression supports a revised mode1 of CDJ+CD8+ thy rnocy te di fferentiation. Irnrnunity 5. 48 1-477.
Lucas. B.. Vasseur. F.. and Pcnit, C. ( 1993). Normal sequence of phenotypic trrinsitions in one cohort of 5-bromo-2'-deoxy uridine-pulse-labelrd thy mocytes. J. Immunol. 151, 4574- 4583.
Lucas, B., Vasseur. F., and Penit. C. (1994). Production. selection. and maturation of thymocytes with high surtace density of TCR. J. Immunol. 153.53-62.
Lundberg, K.. and Shortman, K. ( 1994). Srnall cortical thyrnocytes are subjcct to positive selection. J. Exp. Med. 179. 1475- 1483.
Lundberg, K.. Heath. W., Kontgen, F.. Carbone, F. R.. and Shortman. K. (1995). intermediate steps in positive selection: Differentiation of C D ~ + ~ ~ " T C R ~ ~ ' [hy mocy tes into C D ~ W T C R ~ ~ thymocytes. 1. Exp. bled. 181. 1643- 165 1.
Ma, A.. Pena. J. C., Chang, B., Margosian, E., Davidson. L.. Alt. F. W.. and Thompson. C. B. ( 1995). BcIx regulates the survival of double-positive thymocytes. Proc. Natl. Acad. Sci. USA 92. 4763-4767.
MacDonald. H. R., Schneider, R.. Lees, R. K., Howe, R. C., Plcha-Orbea, H.. Festenstein, H., Zinkernagel, R. iM., and Henpartner, H. ( 1988). T-ce11 receptor VP use predicts reactivity and tolerance to MW-encoded antigens. Nature 322,4045.
Madrenris, J., Chau, L. A., Smith, J., Bluestone. J. A., and Germain, R. N. (1997). The efficiency of CD4 recruitment to ligand-engüged K R controls the agonisVpartia1 agonist properties of peptide-MHC rnolecule ligands. J. Exp. Med. 185.2 19-229.
Maecker. H. T.. and Lrvy. S. ( 1997). Normal lymphocyte developrnent but delayed humoral immune response in CD8 1 -nul1 mice. J. Exp. Med. 185, 1505- 15 10.
Malissen, M., Gillet. A., Rocha. B., Trucy. J.. Vivier, E., Boyer. C.. Kontgen. F.. Brun, N., Mazza, G., Spanopoulou, E., Guy-Grand, D.. and Malissen. B. (1993). T ce11 development in rnice laking the CD3-Zjq gene. EMBO J. 12.4347.
Malissen, M., Gillet. A.. Ardouin. L., Bouvier, G.. Trucy, J.. Ferrier, P.. Vivier, E., and Malissen, B. ( 1995). Altered T ce11 development in mice with a targeted mutation of the CD3- epsilon gene. EMBO J. 14,46411653.
Mallick, C.. Dudley. E.. Viney, J., Owen. M., and Hayday. A. ( 1993). Rearrangement and diversity of T ce11 receptor P chain genes in thymocytes: a critical role for the p chain in development. Ce11 73. 5 13-5 19.
Markowitz, J. S.. Auchincloss, H.. Grusby, M. J., and Glimcher. L. H. ( 1993). Class IIC hernopoietic-cells cannot mediate positive selection of CD4f T-lymphocytes in class-II deficient mice. Proc. Natl. Acad. Sci. USA 90. 3779-3783.
Marth. J. D.. Peet, R.. Krebs. E. G., and Perlmutter. R. .M. ( 1985). A lymphocyte-specific protein tyosine kinase gene is rearranged and overexpressed in the murine T ce11 lymphoma LSTRA. Ce11 43. 393-404.
Marusic-Galesic. S.. Stephany. D. A.. Longo. D. C.. and Kruisbeek, A. LM. (1988). Deveiopment of CD4-8+ cytotoxic T-cells requires interactions with class 1 MHC determinants. Nature 333. 180- 183.
Matechak. E. O., Killeen, N., Hedrick, S. M., and Fowlkes, B. J. ( 1996). .MHC class 11- specific T cells can develop in the CD8 lineage when CD4 is absent. Immunity 4.337-347.
Matsuzaki. Y.. Gyotoku, J.4.. Ogawa. M.. Nishikawa. S.4.. Katsura. Y.. Gachelin. G., and Nakauchi. H. ( 1993). Characterization of c-kit positive intrathy mic stem cells that are restricted to lymphoid differentiation. J. Exp. Med. 178. 1283- 1797.
McCarthy. S. A.. Kruisbeek, A. M.. K.. U. 1.. Sharrow. S. 0.. and Singer. A. (1988). Engagement of the CD4 nioleculr intluences ce11 surface expression of the T-ce11 receptor on thymocytes. Nature 336. 76-79.
McFarland, E.. Hurley, T. R., Pingel, J . T.. SeRon. B.. Shaw, A.. and Thomas, M. L. ( 1993). Correlation between Src fdmily member regulation by the protein-tyrosine-phosphatase CD45 and transmembrane signalins through the T-ce11 receptor. Proc. Natl. Acüd. Sci. USA 90, 1402- 1406.
Merkenshlager. M.. Benoit. C.. and Mathis, D. ( 1994). Evidence for a single niche mode1 of positive selection. Proc. Natl. Acad. Sci. USA 91. 1 1694- 1 1698.
Miyasaka, T., Pabst, R., Dudler. L., Cooper, M., and Yamaguchi. K. (1990). Characterization of lymphatic and venous emigrünts from the thymus. Thymus 16, 29-43.
Molina, T. J., Kishihara. K., Siderovski, D. P., van Ewijk, W.. Narendran. A.. Timms, E., Wakeharn, A., Paige, C. J., Hartmann. K.-U., Veillette. A.. Davidson, D., and Mak, T. W. ( 1997). Profound block in thymocyte development in mice lacking p561c.k. Nature 357, 16 1- 164.
Mombaerts. P., Clarke, A., Rudnicki, M., Iacomini, J. . Itohara, S., LahiHe, J., Wang, L., Ichikawa. Y., Jaenisch, R., Hooper, M., and Tonegawa, S. (1993a). Mutations in T-ce11
antigen receptor genes cc and P block thymocyte development at dit'ferent stages. Nature 360. 225-33 1.
Mombaens. P.. Iacomini. J.. Johnson. R.. Herrup. K.. Tonegawa. S.. and Papaioannou, V. ( 1992b). RAG- 1 -deficient mice have no mature B and T lymphocytes. Cell68. 869-877.
Mombaerts. P.. Anderson. S. J.. Perlrnutter. R. M.. Mak, T. W.. and Tonegawa, S. (1994). An activated lck trrinsgene promotes thymocyte development in RAG-1 mutant mice. Immunity 1, 26 1-267.
~Motoyama, N.. Wang. F.. Roth, K. A.. Sawa. H.. Nakayama, KA., Nakayama, K.. Negishi, 1.. Senju. S.. Zhang. Q., Fuji. S.. and Loh. D. Y. ( 1995). Massive ce11 death of immature hematopoietic crlls and neurons in Bcl-+ -deficient mice. Science 267. 1506- 15 10.
Murphy. K. M., Heimbrrger. A. B.. and Loh. D. Y. (1990). Induction by antigen of intrathymic apoptosis of CD4+CD8+ TCRIO thymocytes in vivo. Science 250. 1710- 1723.
Mustelin. T.. Coggeshall. K. M., and Altman, A. ( L989). Rapid activation of the T-ce11 tyrosine protein kinase pp561ck by the CD45 phosphotyrosine phosphatase. Proc. Natl. Acad. Sci. LISA 86, 6303-6326.
Nagata, S.. and Suda, T. ( 1995). Fas and Fas ligand: Ipr and gld mutations. Immunol. Today 16, 39-43.
Nakayarna. T., Singer. A.. Hsi. E. D.. and Samelson. L. E. (1989). Intrathymic signalling in immature CD4+CD8+ thymocytes results in tyrosine phosphorylrition of the T-ce11 receptor zeta chain. Nature 34I. 65 1-654.
Nakayama. T.. June. C. H.. iMunitz. T. 1.. Sheard. LM.. iMcCiirthy. S. A.. Sharrow. S. O.. Samelson. L. E.. and Singer, A. ( 1990). Inhibition of T ce11 receptor expression and function in immature CD4+ CD8+ cells by CD4 Science 239. 1558- 156 1.
Nakayama. T.. Sümelson. L. E.. Nakayama, Y.. Munits. T. 1.. Shrard. M.. June. C. H.. and Singer. A. ( 199 1 ). Ligand-stimulated signaling events in immature CD4+ CD8+ thymocytes expressing competent T-ce11 receptor complexes. Proc. Natl. Acad. Sci. USA 88. 9949-9953.
Nakayama, K.-1.. Nakayama. K.. Negishi, 1.. K~iida. K.. Shinkai. Y.. Louie. M.. Fields, 1. E., Lucas, P. J.. Stewart. V.. Alt. F. W.. and Loh. D. Y. (1993a). Disappearance of the lyrnphoid system in Bcl-2 homozygous mutant chimrric mice. Science 261. 1584- 1588.
Nakayama. T.. Wiest. D. L.. Abraham. K. M.. Munitz. T. 1.. Perlrnutter. R. M.. and Singer. A. ( 1993b). Decreüsed signaling cornpetence as a result of receptor overexpression: overexpression of CD4 reduces its ability to activate p56/ck tyrosine kinase and to regulate T- ce11 antigen receptor expression in immature CD4+CD8+ thymocytes. Proc. Natl. Aciid. Sci. USA 90. 10534-10538.
Negishi, 1.. Motoyama. N.. Nakayarna. KA.. Nakayama. K.. Senju. S., Hatakeyama, S.. Zhang, Q., Chan. A. C. . and Loh. D. Y. (1995). Essential role for ZAP-70 in both positive and negative selection of thymocytes. Nature 376.4351138.
Nelson. A. J.. Hosier. S., Brady, W.. Linstey, P. S., and Farr, A. G. (1993). Medullary thyrnic epithelium expresses a ligand for CTLA4 in situ and in vitro. J. Irnrnunol. 151. 2453- 246 1 .
Nikolic-Zugic, J., and Bevan, M. J. ( 1988). Thymocytes expressing CD8 differentiate into CD4+ cells following intrathymic injection. Proc. Nati. Acad. Sci. USA 8.5. 8633-8637.
Nikolic-Zugic. J.. and Bevan. M. J. ( 1990). Role of self-peptides in positively selecting the T- ce11 repertoire. Nature 341.65-67.
Nikolic-Zugic, 3 . . Moore. M. W., and Bevan. M. J. (1989). Characterization of the subset of immature thymocytes which can undergo rapid in vitro differentiation. Eur. J. Immunol. 19, 649-653.
Ohashi, P. S.. Pircher, H.. Bürki. K.. Zinkernagel, R. AM., and Hengartner. H. (1990). Distinct sequence of negative or positive selection implied by thymocyte T-ce11 receptor densities. Nature 346. 86 1-863.
Ohno. H.. Aoe, T., Taki. S.. Kitamura. D.. Ishida. Y., Rajewsky, K.. and Saito. T. (1993). Developmental and functional impairment of T cells in micc lacking CD35 chains. EMBO J. 12, 4357-3366.
Olszowy. M. W.. Leuchtmann. P. L., Veillette, A.. and Shaw, A. S. (1095). Cornparison of p561ck and p5&" protein expression in thyrnocyte subsets. peripherül T cells. NK cells. and lymphoid ce11 lines. J. Immunol. Ij5.423612-10.
Oltvai. 2. N.. iMilliman. C. L.. and Korsmeyer. S. J . (1993). Bcl-2 heterodimerizes in vivo with a conserved homolog, B u . that accelerates prograrned ce11 death. Ce11 74.609-6 19.
Oren. R.. Takahashi. S., Doss, C.. Levy, R., and Levy. S. ( 1990). TAPA- 1. the target of an antiproliferative antibody. defines a new frimily of transmembrüne of proteins. Mol. Cell. Biol. 10, 4007-40 1 5.
Ossendorp, F., Jacobs. H., van der Horst, G.. de Vries, E., Berns. A.. and Borst. J. (1992). T ce11 receptor-ap lacking the p-chain V dornain can be expressed at the ce11 surface but prohibits T ce11 maturation. J. Immunol. 148, 37 14-3722.
Ostergaard. H. L.. Shackelford. D. A.. Hurley, T. R., Johnson. P.. Hymim. R.. Sefton, B. M.. and Trowbridge. I. S. ( 1989). Expression of CD45 alters phosphorylation of the lck- encoded tyrosine protein kinase in murine lymphoma T-ce11 lines. Proc. Natl. Acad. Sci. USA 86. 8959-8963.
Oukka. M.. Colucci-Guyon. E.. Lan Tran, P.. Cohen-Tannoudji. M.. Babinet. C., Lotteau, V., and Kosmatopoulos. K. ( 1996). CD4 T ce11 tolerancr to nuclear proteins induced by medullary thymic epithelium. Immunity 4,545-553.
Page, D. M.. Küne. L. P., Allison. J. P.. and Hedrick. S. M. (1993). Two signals are required for negative selection of CD4+CD8+ thyrnocytes. I. Immunol. 1.71, 1868- 1880.
Page, S. T., van Oers. N. S. C.. Perlmutter. R. M., Weiss. A.. and Pullen. A. (1997). Differential contribution of Lck and Fyn protein tyrosine kinases to intraepithelial lymphocyte development. Eur. J. Immunol. 27. 554-562.
Pardoll, D. M.. Fowlkes, B. J.. Bluestone. J. A.. Kruisbeek. A.. Maloy, W. L.. Coligan, J. E., and Schwartz. R. H. ( 1987). Differential expression of two distinct T-ce11 receptors during thymocyte development. Nature 326. 79-8 1.
Panes. J. R. ( 1989). Molecular biology and function of CD4 and CD8. Adv. Immunol. 44, 365-3 1 1.
Pawtowski, T., Elliott, J. D., Loh, D. Y., and Strierz. U. D. (1993). Positive selection of T lymphocytes on tibroblasts. Nature 364. 643-645.
Pear, W. S., Aster, J. C., Scott, M. L.. Hasserjian, R. P., Softèr, B.. Sklar. I.. and Baltimore, D. ( 1996). Exclusive development of T-ce11 neoplasrns in mice trrinsplünted with bone rnarrow expressing activated Notch alleles. I. Exp. Med. 183. 2283-239 1.
Penit. C., and Vasseur, F. ( 1988). Sequential events in thymocyte dilferentation and thymus regeneration revealed by a combination of bromodeoxyuridine DNA labelling and antimitotic drug treatment. J. Immunol. I JO, 33 15-3323.
Penninger. J., Kishihara, K.. Molina. T.. Wallace, V. A.. Timms. E.. Hedrick. S. M., and Mak. T. W. (1993). Requirement for tyrosine kinase p56l~k for thymic development of transgenic y3 T cells. Science 260, 358-36 1.
Perlrnuttrr. R. M.. Marth. 1. D., Ziegler, S. F.. Garvin. A. iM.. Pawar. S.. Cooke, M. P.. and Abraham, K. M. ( 1988). Specidized protein tyrosine kinase proto-oncogenes in hematopoietic cells. Biochim. B iophys. Acta. 948. 345-762.
Petrie. H. T.. Livak, F.. Schatz. D. Ci.. Strasser. A.. Crispe, 1. N.. and Shortman, K. ( 1993a). Multiple rearrangements in TCRa chain genes rnaximize the production of useful thymocytes. J. Exp. iMed. 178, 6 15-622.
Petrie. H. T.. Strasser. A.. Harris. A. W.. Hugo. P.. and Shortman. K. (1993b). CD4+8- and CD4-8+ mature thymocytes rrquire different post-srlection processing for final development. J. Immunol. 151. 1273- 1279.
Philpott, K.. Viney. J.. Kay. G.. Rüstan. S.. Gardiner. E.. Char. S.. Hayday. A.. and Owen, LM. ( 1992). Lymphoid deveioprnent in mice congenit;illy lacking T ce11 rcçeptor ap-expressing cells. Science 256. 1448- 1452.
Pircher. H.. Bürki. K.. Lang, R.. Hengartner. H., and Zinkernagel, R. M. (1989). Tolerance induction in double speci tic T-cell receptor transgenic micr varies w ith antigen. Nature 342, 559-56 1 .
Poirier. G.. Lo. D.. Reilly. C. R., and Kaye. 1. ( 1994). Discrimination between thymic rpithelial cells and peripheral antigen-presenting cells in the induction of immature T ce11 differentiation. Immunity 1 . 385-39 1.
Poussier. P.. and Julius. M. ( 1994). Thymus independent T ce11 development and selection in the intestinal epithelium. Ann. Rev. Immunol. 12, 52 1-53.
Poussier. P., and Julius, M. ( 1995). T-ce11 development and selection in the intestinal epithelium. Sem. Immunol. 7, 32 1-31.
Pullen. A. M., Marrack. P.. and Kappler, J. W. (1988). The T-ce11 repertoire is heavily influenced by tolerance to polyrnorphic self-antigens. Nature 335. 796-80 1.
Punt, J. A.. Osborne, B. A.. Takahama, Y., Sharrow. S . O., and Singer. A. ( 1994). Negative selection of CD4fCD8+ thymocytes by T ce11 recrptor-induced apoptosis requires a costirnulatory signal that c m be provided by CD28. J. Exp. bled. 179. 709-7 13.
Qian, D.. Lw. S.. van Oers. N. S. C., Dikic. 1.. Schlessinger. J.. and Weiss. A. (1997). Tyrosine phosphorylation of PykZ is selectively regulated by Fyn during TCR signaling. J. Exp. Med. 185, 1353- 1259.
Rajüsekar. R., Sim, G. K., and Augustin. A. ( 1990). Self heat shock and y3 T-ce11 reactivity. Proc. Natl. Acad. Sci. USA 87, 1767- 177 1.
Ramsdell, F., and Fowlkes, B. 3. ( 1989). Engagement of CD4 and CD8 accessory rnoiecuIes is required for T ce11 maturation. J. Immunol. 143. 1467- 147 1.
Rarnsdell, F., Lantz, T., and Fowlkes, B. J. ( 1989). A nondeletional mechanism of thymic self tolerance. Science 246, 1038- 104 1.
Reth, M. ( 1989). Antigen receptor tail clue. Nature 338, 383-384.
Reynolds, P. J.. Lesley. J.. Trotter, J.. Schulte, R.. Hyman. R., and Sefton. B. M. (1990). Changes in the relative abundance of type 1 and type II lck mRNA transcripts suggest differential promoter usage during T-ce11 development. Mol. Cell. Biol. IO, 42664270.
Roberts. J. L.. Sharrow. S. O., and Singer. A. S. ( 1989). Clonal deletion and clonai anergy in the thymus inducrd by cellular elements with different radiation sensitivities. J. Exp. Med. 171. 935-940.
Robey. E.. Chang. D.. Ltano. A.. Cado. D.. Alexander, H., Lans. D.. Wrinmaster, G.. and Salmon, P. ( 1996). An activated form of Notch influences the choice between CD4 and CD8 T ce11 lineages. Ce11 87,483-492.
Robey. E.. and Fowlkes, B. J . ( 1994). Selrctive events in T ce11 development. Annu. Rev. Immunol. 12. 675-705.
Robey. E. A.. Fowlkes. B. J., Gordon. J. W.. Kioussis. D.. von Boehmer. H.. Ramsdell. F.. and Axrl. R. (1991). Thymic selection in CD8 transgenic mice supports an instructive mode1 for commitrnent to a CD4 or CD8 lineage. Ce11 64.99- 107.
Robey. E. A.. Ramsdell. F.. Kioussis. D.. S1:a. W.. Loh. D.. Axel. R.. and Fowlkes, B. J. ( 1992). The level of CD8 expression can determine the outcome of thymic selection. Ce11 69, 1089- 1096.
Robey, E., Itano. A.. Fanslow, W.. and Fowlkes, B. J. (L994). Constitutive CD8 expression allows inefficient maturation of CD4+ helper T cells in class II major histocompatibility complex mutant mice. J. Exp. Med. 179. 1997-2001.
Rodewald. H.-R. ( 1995). Püthways from hernatopoietic stem cells to thymocytes. Curr. Opin. Immunol. 7, 176- 187.
RodewaId, H.-R., Kretzschmar, K., Takeda, S., Hohl, C., and Dessing, M. (1994). Identification of prothymocytes in murine fetal blood: T lineage commitment c m precede thymus colonization. EMBO J. 13.4229-4240.
Romeo. C.. and Seed. B. ( 199 1). Cellular immunity to HIV activated by CD4 fused to T ce11 or Fc receptor polypeptides. Cc11 64, 1037- 1046.
Romeo. C.. Amiot. M., and Seed. B. (1992). Sequence requirements for induction of cytolysis by the T ce11 antigen/Fc receptor < chain. Ce11 68. 1037- 10.16.
Rothstein, D. M., daSilva. A.. Sugita. K., Yamamoto. LM.. Praad. K. V.. Morimoto. C.,
Schlossman. S. F.. and Rudd, C . E. (1993). Human C D ~ / C D ~ ~ R A + and CDUCD45RA- T ceIl subsets express ~ ~ i l - p j 6 l ~ ~ complexe. CD4-associüred lipid kinases. T C R / C D ~ - ~ S ~ ~ ! ~ ~ complexes. and share sirnilar tyrosine kinase substrates. Int. Immunol. 5.4094 18.
Saint-Ruf. C.. Ungewiss. K.. Groettrup. M.. Bruno. L.. Fehling. H. J.. and von Boehmer. H. ( 1994). Analysis and expression of a cloned pre-T ce11 recrptor gene. Science 266. 1208- 1212.
Sarosi. G. P.. Thomas. P. M.. Egerton. M.. Phillips. A. F.. Kim. K. W.. Bonvini. E., and Samelson, L. E. ( 1992). Chüracterization of the T ce11 antigen receptor-p6Ofjfl protein tyrosine kinase association by chemical crosslinking. Int. Immunol. 4. 12 1 1- 12 17.
Schild. H., Mübaddat. N.. Litzenberger. C., Ehrich. E. W.. and Davis. M. M. (1994). The nature of major histocompatability çomplex recognition by y6 T cells. Ce11 76. 29-37.
Schonrich. G.. Momburg: F.. Hammerling. G. J.. and Arnold. B. ( 1992). Anergy induced by thyrnic medullüry spithelium. Eur. J. Immunol. 22. 1687- 169 1.
Schraven. B., Kirchgessner. H.. Gaber, B., Sarnstag, Y., and Meuer. S. (199 1 ). A functional complex is formed in human T lymphocytes between the protein tyrosine phosphatase CD45, the protein tyrosine kinase p561t.k and pp32, ii possible common substrate. Eur. J . Immunol. 2 1, 2469-2477.
Schweighoffcr. E.. and Fowlkes. B. J. ( 1996). Positive srlrction is not required for thymic maturation of transgenic -{S T cells. J. Exp. ~Med. 183. 2033-104 1.
Scott. B.. Bluthmann. H.. Teh. H. S.. and von Borhmer. H. (1989). The seneration of mature T cells requires interaction of the ap T-ce11 receptor with major histocompatibility antigens. Nature 338. 59 1-593.
Srbzdü. E.. Wallace. V. A.. Mayer. J.. Yeung. R.. Mak. T. W.. and Ohüshi. P. (1994). Positive and negative thymocyte selection induced by different concentrations of a single peptide. Science 263. 16 15- 16 18.
Sebzda. E., Kundig, T.M.. Thomson, C.T.. Aoki. K., Mak. S-Y.. Mayer. J.P.. Zamborelli, T., Nathenson, S.G.. Ohashi. P.S. ( 1996). Mature T ce11 reactivity altered by peptide agonist that induces positive selection. J. Exp. Med. 183, 1093-1 104.
Seckinger, P.. and Fougereau. M. ( 1994). Activation of src family kinases in human pre-B cells by IL-7. J. Immunol. 153. 97-109.
Sentrnan. C. L.. Shutter, J. R.. Hockenbery, D.. Kanagawa. O., and Korsmeyer, S. J. ( 199 1 ). bcl-2 inhibits multiple forrns of apoptosis but not negative selection in thymocytes. Ce11 67. 879-888.
Sha, W. C., Nelson, C. A., Newberry, R. D., Krrinz. D. M., Russell, J. H.. and Loh. D. Y. ( 1988a). Positive and negative selection of an antigen receptor on T cells in tmnsgenic mice. Nature 336. 73-76.
Sha. W. C., Nelson. C. A., Newberry. R. D., Kranz, D. M., Russell, J. H., and Loh, D. Y. (1988b). Selective expression of an antigen receptor on CD8-bearing T lymphocytes in transgenic mice. Nature 335, 27 1 -274.
Sha, W. C.. Nelson, C. A.. Newberrv, R. D.. Pullen. J. K.. Pease. L. R.. Russell. J. H.. and Loh. D. Y. ( 1990). Positive selrction of transgenic receptor-bearing thymocytes by K~ antigen is altered by Kb mutations that involve peptide binding. Proc. Natl. Acad. Sci. USA 87, 61864190.
Shahinian, A., Pfeffer. K., Lee, K. P.. Kündig. T., Kishihara, K.. Wiikeham. A.. Kawai. K., Ohashi, P. S.. Thompson. C. B.. and Mak. T. W. (1993). Differential T ceIl costirnulatory requirernents in CD28-deficient mice. Science 261. 609-6 12.
Shaw, A. S.. Amrein, K. E.. Hammond, C., Stern, D. F., Sefton. B. M.. and Rose, J. K. ( 1989). The lck tyrosine protein kinase interacts with the cytoplasmic tail of the CD4 glycoprotein through its unique amino-terminal domain. Ce11 59. 617-636.
Sherman, L. A., Hesse, S. V., Irwin. LM. J., La Face. D.. and Peterson, P. ( 1992). Selecting T ce11 receptors with high affinity for self-MHC by decreasing the contribution of CD8. Science 358, 8 L 5-8 18.
Shi, Y.. Bissonnette. R. P., Parfrey. N., Szalay, M.. Kubo, R. T.. and Green. D. R. (1991). In vivo administration of monoclonal üntibodies to the C D 3 n ce11 receptor complex induces ce11 death (apoptosis) in immature thyrnocytes. J. Immunol. 146,3340-3346.
Shirnonkevitz. R. P.. Husmann, L. A., Bevan, M. J.. and Crispe. 1. N. (1987). Transient expression of IL-? receptor precedes the differentiation of immature thyrnocytes. Nature 329, 157- 159.
Shinkai, Y.. and Mt. F. W. ( 1994). CD~E-mediated signüls rescue the development of CD4+CD8+ thyrnocytes in RAG-Z-I- rnice in the absence of TCRP chain expression. [nt. Immunol. 6, 995- 100 1.
Shinkai, Y.. Rathbun, G., Lam. K.-P., Oltz, E., Stewart. V., Mendelsohn. M.. Charron, J., Datta. M.. Young, F.. Stall. A.. and Alt, F. (1992). RAG-Z-deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cr11 68,855-567.
Shinkai. Y.. Koyasu, S., Nakayama, KA.. Murphy, K.. Loh, D., Reinhrrz. E., and Alt, F. (1993). Restoration of T ce11 development in RAG-2-deficient mice by functional TCR transgenes. Science 259, 812-825.
Shinkai, Y., Ma, A., Cheng, H.-L., and Alt, F. W. (1995). CD% and CD35 cytoplasmic dornains can independently generate signüls for T ce11 development and function. Immunity 2. 401-41 1.
Shiroo, M., Goff, L.. Biffen, M., Shivnan, E., and Alexander. D. ( 1992). CD45 tyrosine phophatase-activated p5gv"" couples the T ce11 iintigen receptor to pathways of diacylglycerol production, protein kinase C activation and calcium flux. EMBO J. 11,48874897.
Shores, E., Nakayama, T., Wiest, D., Takahama. Y., Sharrow. S., and Singer. A. ( 1993). Structurdly distinct T ce11 receptor complexes on developmentally distinct T ce11 populations in severe combined immunodeficiency mice expressing a TCRP transgene. J. Immunol. ISO, 1263- 1275.
Shores, E. W.. Huang, K.. Tran. T., Lee. E., Grinberg. A.. and Love. P. E. (1994). Role of TCRC chain in T ce11 development and selection. Science 266. 1047- 1050.
Shortman, K., Egerton, M., Spangrude, G . J.. and Scollay, R. ( 1990). The generation and h te of thymocytes. In Seminars in Immunology, J. Sprent. ed. (London: W. B. Saunders), pp. 3-12.
Shortman, K.. Vrernec. D.. and Egerton. M. (1991). The kinetics of T ce11 antigen receptor expression by su bgroups of CD4+8+ thymocytes: Delineütion of ~ ~ 4 + 8 + 3 ~ + thy mocytes as post-selection intermediates leading to mature T cells. J. Exp. Med. 173. 313-332.
Sieh, M., Bolen, J. B., and Weiss. A. ( 1993). CD45 specificülly modulates binding of Lck to a phosphopeptide encornpassing the negative regulütory tyrosine of Lck. EMBO J . 12, 3 15- 13 1
Smith. C. A., Williams. G. T.. Kingston, R., Jenkinson. E. J.. and Owen. J. J. T. (1989). Antibodies to the CD3TT-ce11 receptor complex induce death by apoptosis in immature T cells in thymic cultures. Nature 337, 18 1 - 184.
Smith, D. H., Lee, N.. and Lacy. E. ( 1996). Thymocyte ce11 fate alterrition by high levels of CD8a is independent of CDB-Id association. J. Irnmunol. 157. 108% 1095.
Speiser, D. E.. Lees. R. K.. Hengiirtner. H.. Zinkernagel, R. M., and MacDonald. H. R. ( 1989). Positive and negative selection of T ce11 receptor Vp domains controlled by distinct ce11 populations iin the thymus. J. Exp. Med. 170, 2 165-2170.
Speiser, D. E.. Chvatchko. Y.. Zinkernagel. R. M.. and MacDonald. H. R. ( 1990). Distinct fates of sel f-speci fic T cells develping in irradiation bone miirrow chimeras: Clonal deletion, clonal anergy. or in vitro responsiveness to self-Mls- 1" controlled by hernopoietic cells in the thymus. J. Exp. iMed. 172. 1305- 13 14.
Speiser. D. E.. Pircher. H.. Ohashi. P. S.. Kybiirz, D.. Hengartner. H.. and Zinkernagel, R. M. (1992). Clonai deletion induced by either radioresistant thymic host cells or lymphohemopoietic donor cells at different stages of class 1-restricted T ce11 ontogeny. J. Exp. Med. 17.5, 1277- 1283.
Sprent. J.. Lo. D.. Gao. E.-K.. and Ron, Y. (1988). T-ce11 selection in the thymus. Irnmunol. Rev. 101, 173- NO.
Stehnova. I . , Horejsi. V.. Ansotevi. 1. J.. Knüpp. W.. and Stockinger. H. ( 199 1). GPI- anchored cell-surface molecules çomplexed to protein tyrosine kinases. Science 254, 10 16- 1019.
Stein. P. L.. Lee. J.-iM.. Rich, S.. and Soriano. P. (1997). ppj*\'', mutant mice display differential signaling in thymocytes and peripheral T cells. Ccll 70.74 1-750.
Strasser, A., Harris. A. W.. and Cory, S. (1991). 6cl -2 transgene inhibits T ce11 death and penurbs thymic ce11 self-censorship. Ce11 6 7.889-899.
Straus. D. B.. and Weiss, A. ( 1992). Genetic evidence for the involvement of the lck tyrosine kinase in signai transduction through the T ce11 antigen receptor. Ce11 70.585-593.
Surh. C. D.. Gao. E.-K.. Kosaka. H.. Lo. D.. Ahn. C.. Murphy. D. B.. Karlsson. L., Peterson. P.. and Sprent, J. ( 1992). Two subsets of epithelial cells in the thymic medulla. J. Exp. Med. 176, 495-505.
Suzuki. H.. Punt. J.. Granger. L., and Singer. A. ( 1995). Asymrnetric signaling requirements for thymocyte commitrnent to the C M f versus CD8+ T ce11 lineages: A new perspective on thymic comniitrnent and selection. lmmunity 2.4 13-425.
Swan, K. A., Alberola-Ila. J.. Gross. J. A., Appleby, M. W.. Forbush. K. A.. Thomas, J. F.. and Perlmutter. R. M. ( 1995). Involvement of p2 lruS distinguishes positive and negative selection in thymocytes. EMBO J, 13.376-285.
Swat, W.. Ignatowicz. L.. von Boehmer. H.. and Kisielow. P. ( 1991). Clonal deletion of immature CD4+8+ thymocytes in suspension culture by rxtrathyrnic antigen-presenting cells. Nature 351. 150-153.
Swat. W.. Dessing, M.. von Boehrner, H., and Kisielow. P. ( 1993). CD69 expression during selection and maturation of CD4+8+ thymocytes. Eur. J. Immunol. 23. 739-746.
Swat. W.. Shinkai. Y.. Cheng. H.-L., Davidson. L.. and Alt. F. W. ( 1996). Activated Ras signals differentiation and expansion of CD4+8+ thymocytes. Proc. Nat 1. Acad. Sci. USA 93, 4683-4687.
Takahama. Y.. Shores. E. W.. and Singer. A. (1992). Negative selection of precursor thymocytes before their differentiation into CD4+CD8+ cells. Science 258.653-656.
Takahama. Y.. Suzuki. H.. Katz. K.. Grusby. M. J.. and Singer, A. ( 1994). Positive selection of CD4+ T cells by TCR ligation without aggregation even in the absence of MHC. Nature 371, 67-70.
Takayama. S.. Sato. T.. Krajewski. S.. Kochel. K.. Irie, S.. Millan. J. A.. and Reed, J. C. ( 1995). Cloning and functional iuialysis of BAG- 1 : a novel BcI-2-binding protein with anti-ce11 death activity. Ce11 80. 279-284.
Tan. R.. Teh. S.-J.. Lrdbctter. J. A.. Linslry. P. S.. and Teh. HA. ( 1992). B7 costirnulates proliferation of CD4-8+ T lymphocytes but is not required for the drletion of immature CD4+8+ thymocytes. j. Immunol. 139. 32 17-3224.
Tanaka. Y.. Sano, S., Nieves. E., De Libero, G., Rosa. D., Modlin. R. L., Brenner, M. B.. Bloom. B. R.. and Morita. C. T. (1994). Nonpeptide ligands for human T cells. Proc. Natl. Acad. Sci. USA 91, 8 175-8 179.
Tao. W.. Teh. S.-J.. Melhado. 1.. Jirik. F.. Konerneyer. S. J.. and Teh. Hm-S. (1994). The T ce11 receptor repertoire of CD4-8+ thyrnocytes is altered by overexpression of the BCL-2 protooncogene in the thymus. J. Exp. Med. 179, L45- 153.
Twakhovsky. A.. Kanner. S. B.. Hornbach, J.. Ledbetter. J. A.. Müller. W.. Killeen, N., and Rajewsky. K. ( 1995). A role for CD5 in KR-mediated signal transduction and thymocyte selection. Science 269, 535-537.
Teh, H. S., Kisielow, P., Scott, B., Kishi, H., Uernatsu, Y., Bluthrnann, ii., and von Boehrner, H. ( 1988). Thyrnic major histocornpatibility cornplex antigens and the ap T-ceIl receptor determine the CD4fCD8 phenotype of T-cells. Nature 335,129-233.
Teh. H. S., Kishi, H.. Scott, B., Borgulya, P., von Boehmer, H.. and Kisielow, P. ( 1990). Early deletion and late positive selection of T cells expressing a maie-specific receptor in T ce11 receptor trrinsgenic mice. Devel. irnmunol. 1, 1 - 10.
Teh, H.-S., Carvin, A. M., Forbush, K. A.. Carlow, D. A.. Davis. C . B., Littman, D. R., and Perlmutter. R. M. ( 199 1 ). Participation o l CD4 coreceptor molecules in T ceIl repertoire selection. Nature 349, 24 1-243.
Testi. R.. Phillips. J. H.. and Lanier, L. L. (1989). Leu 23 induction as an eürly marker of func tional CD31T ceIl an tigen receptor triggering. Requirement for receptor cross-linking, prolonged elevation of intracelluiür [Ca++] and stimulation of protein kinase C. J. Immunol. 132, 1854-1860.
Turka. L. A.. Linsley. P. A.. Paine, R.. Schieven, G. L.. Thompson. C . B.. and Ledbetter, J. A. ( 1 9 9 1 ~ ) . Signal transduction via C D 4 CD8. and CD28 in mature and immature thymocytes. Implications for thymic selection. J. Immunol. 116. 142% 1136.
Turka. L. A.. Schritz, D. G.. Oettinger, M. A., Chun, J., Gorka, C.. Lee, K.. McCorrnack, W. T.. and Thompson. C. B. (1991b). Thymocyte expression of RAG-1 and RAG-2: Termination by T ceIl receptor cross-linking. Science 253, 778-78 1.
Turner. J. LM., Brodsky, M. H.. Irving, B. A., Levin, S. D., Perlmutter. R. M.. and Littman, D. R. (1990). Interaction of the unique N-terminal region of tyrosine kinase p56lck with cytopiasmic domains of CD4 and CD8 is mediated by cysteine motifs. CeIl 60. 755-765.
Turner, LM.. Mee, P. J.. Costello, P. S.. Williams. O.. Price. A. A.. Duddy. L. P.. Furlong, iM. T.. Geahfen. R. L.. and Tybuiewicz, V. L. 3. ( 1995). Perinatal lethrility and blocked B-ce11 development in mice lacking the tyrosine kinase Syk. Nature 378, 298-301.
Uematsu. Y.. Ryser. S., Dembic, 2.. Borgulya, P.. Krimpenfort. P., Brrns. A.. von Boehmer, H.. and Steinrnetz. M. ( 1988). In transgenic mice the introduced function T ce11 receptor p gene prevents expression of endogenous P genes. CeIl 52,83 1-84 1.
van Ewijk, W.. Rouse. R. V.. Jones, P. P., and Weissmün, I. L. (1980). Distribution of H-2 microenvironments in the mouse thymus. Immunoelectronmicroscopic identification of 1-A and H-2K bearing cells. J. Histochem. Cytochem. 28. 1089- 1099.
van Ewijk, W., Ron, Y.. Monaco, J., Kappler, J., tMarrack, P.. Le Meur. M.. Gerlinger, P., Durand. B.. Benoist, C.. and Mathis, D. ( 1988). Compartmentalizütion of MHC class II gene expression in transgenic rnice. Cell 53, 357-370.
van Meerwijk, J . P. M., and Germain, R. N. (1993). Development of mature CD8+ thymocytes: selection rather than instruction'? Science 26 1.9 1 1-9 15.
van Meerwijk. J. P. M.. O'Connell, E. M.. and Germain, R. N. (1995). Evidence for lineage commitment and initiation of positive selection by thymocytes with intermediate surface phenotypes. J. Immunol. 153, 63 14-6323.
van Oers, N. S . C., Garvin, A. M., Davis, C. B., Forbush, K. A., Carlow, D. A., Littman, D. R., Perlmutter, R. M.. and Teh, H.-S. ( 1992). Disruption of CDS-dependent negative and positive selection of thymocytes is correlated with a decreased association between CD8 and the protein tyrosine kinase, p56/~.~. Eur. J. Immunol. 22. 735-743.
van Oers, N. S. C.. von Boehmer, H.. and Weiss, A. (1995). The pre-T çell receptor (TCR) complex is functiondly coupled to the TCR-6 subunit. J. Exp. Med. 182, 1585- 1590.
van Oers. Pi. S. C., Killeen, N., and Weiss, A. (1996ü). Lck regulates the tyrosine phosphorylation of the T ce11 receptor subunits and ZAP-70 in murine thymocytes. J. Exp. Med. 183, 10% 1062.
van Oers. N. S. C., Lowin-Kropf, B.. Finlay. D., Connolly. K.. and Weiss. A. ( 1996b). ap T ce11 development is abolished in mice lacking both Lck and Fyn protein tyrosine kinases. Immunity 5, 429-436.
Veis, D. J., Sentman, C. L., Bach. E. A., and Korsmeyer, S. J. ( 1993). Expression of the Bci-2 piotein in murine and human thymocytes and in peripheral T lymphocytes. J. Immunol. 1.51, 2546-2554.
von Boehmer, H. ( 1986). The selection of the ap heterodimeric T ce11 receptor for antigen. Immunol. Today 7. 333-336.
von Boehmer, H. ( L 994). Positive selection of lymphocytes. CeIl 76.2 lS)-IX
von Boehrner, H. ( 1996). CD4/CD8 lineage cornmitment: back to instruction? J. Exp. Med. 153, 7 13-7 15.
von Boehmer, H.. and Kisielow, P. ( 1990). Self-nonself discrimination by T c e k Science 348, 1369- 1373.
von Boehmer, H., Swat, W.. and Kisielow, P. (1993). Positive selcction of immature ocp T cells. Irnmunol. Revicws 135, 67-79.
von Freeden-Jeffry, U., Vieim, P.. Lucilin. L. A., McNeii. T., Burdach. S: E., and Murray, R. (1995). Lymphopenia in interleukin (IL)-7 gene-deletrd mice identifies IL-7 as a nonredundant cytokine. J. Exp. iMed. 181. 15 19- 1526.
Wallace, V. A.. Kslwai, K., Levelt, C. N., Kishihara, K.. iMolina. T.. Tirnms. E.. Pircher, H.. Penninger, J., Ohashi, P., Eichmann, K.. and Mak. T. W. (1995). T lymphocyte development in p561ck deficient mice: allelic exclusion is incornpiete but tliymocyle development is not restored by TCRP or TCRn iranspnes. Eur. J. Immunol. 25, 13 12- 13 18.
Walunas, T. L.. Sperling, A. 1.. Khattri, R., Thompson, C. B., and Bluestone. J . A. (1996). CD28 expression is not essential for positive and negative selection of thymocytes or peripheral T ce11 tolerance. J. Immunol. 156, 1006- 10 13.
Wange, R. L., Kong. A.-N. T., and Sarnelson, L. E. (1992). A tyrosine-phosphorylated 70- kDa protein binds a photoaffinity analogue of ATP and associates with both the < chain and CD3 components of the üctivüted T ce11 mtigen receptor. J. Biol. Chem. 267. 1 1685- 1 1688.
Wange, R. L.. Malek, S. N., Desiderio, S.. and Siimelson, L. E. (1993). Tandem SH2 domains of ZAP-70 bind to T cell antigen receptor < and C D ~ E frorn activated 3urkrit T cells. J. Biol. Chern. ,368. 19797- 1980 1.
Wegener, A.-M. K., Letourneur, F.. Hoeveler, A.. Brocker, T.. Luton, F.. and Malissen, B. ( 1992). The T ce11 receptorKD3 complex is composed of at l e s t two autonomous transduction modules. CeIl 68. 83-95.
Weinmaster, G., Robens, V. J., and Lemke, G. ( 199 1 ). A homologue of Drosophila Notch expressed during miirnrnalian development. Development I I 3. l9WO5.
Weinmaster, G., Robens. V. J., and Lemke, G. ( 1992). NotchZ: A second mammrilian Notch gene. Development 116.93 1-94 1.
Weintraub, B. C.. Jackson. M. R., and Hedrick. S. M. (1994). -/6 T cells c m recognize nonclassical MHC in the absence of conventionai antigenic peptides. J. Immunol. 153, 305 1- 3058.
Weiss, A. ( 1993). T cell antigen receptor signal transduction: A talc of tails and cytoplasmic protein-tyrosine kinases. CeIl 73, 209-7 12.
Weiss, A., and Littman. D. R. ( 1994). Signal transduction by lymphocyte antigen receptors. Ce11 76. 263-274.
Wekerle, H.. Ketelsen, Lr.-P., and Ernst, M. ( 1980). Thymic nurse cells. Lymphoepithelial cell complexes in murine thymuses: morphological and serological characterization. J. Exp. Med. 15 1, 925-944.
Wells, F. B., Gahm, S . I., Hedrick, S . M., Bluestone. J. A.. Dent, A., and Matis, L. A. ( 199 1 ). Requirement for positive selrction of gamma delta receptor-bearing T cells. Science 253, 003-5.
Wells, F. B.. Tatsumi, Y., Bluestone, J. A.. Hedrick, S. LM., Allison, J. P.. and Mritis, L. A. (1993). Phenotypic and functional analysis of positive selrction in the gammddelta T ce11 iineage. J. Exp. Med. 1 77. L O6 1-70.
Wen, T., Zhang, L., Kung, S. K. P.. ~Molina, T. J.. ~Miller, R. G., and Mrik. T. W. (1995). Allo-skin graft rejection. tumor rejection and natural M e r activity in mice lacking p56"k. Eur. J. Immunol. 35, 3 155-3 159.
Wiest, D. L., Yuan, L.. Jefferson, J., Benveniste, P., Tsokos, M.. Klausner, R. D., Glirncher. L. H.. Samelson, E.. and Singer, A. (1993). Regulation of T ceIl receptor expression in immature CDI+C D8+ thymocytes by p56Ick tyrosine kinase: Basis for differential signaling by CD4 and CD8 in immature thymocytes expressing both coreceptor molecules. J. Exp. Med. 178, 170 1- 17 12.
Wiest, D. L.. Ashe, I. M., Abe, R.. Bolen, J. B.. and Singer. A. (1996). TCR activation of ZAP-70 is impaired in CDJ+CD8+ thymocytes as a consequence of intrathymic interactions that diminish available p56@ Immunity 4,495-504.
Wildin, R. S., Garvin. A. M., Pawar, S., Lewis. D. B., Abraham, K. M., Forbush, K. A., Ziegler, S. F.. Allen. J. M., and Perlrnutter, R. M. (199 1). Developmental regulation of Ick gene expression in T lymphocytes. J. Exp. Med. 173,383-393-
Wilkinson, R. W., Anderson, G.. Owen, J. J . T., and Jenkinson. E. J. ( 1995). Positive selection of thymocytes involves sustained interxtions with the thymic microenvironment. J. imrnunol. 155,5234-5240.
Wilson, A., Held, W., and MacDonald, H. R. (1994). Two waves of recombinase gene expression in developing thymocytes. J. Exp. Med. 179. 1355- 1360.
Wu, L., Antica, M.. Johnson, G. R.. Scollay. R.. and Shortman, K. ( 199 1). Development potential of the eürliest precursor cells from the adult mouse thymus. J. Exp. Med. 174, 16 17- 1627.
Wu, G.. Danska. J. S., and Guidos, C. J. (1996). Lck -dependence of signaling pathways activated by -{-irradiation and C D ~ E engagement of immature RAG-de ficient thy mocytes. Int. Immunol. S. 1 159- 1 164.
Yamashira. 1.. Nagata. T.. Tada. T.. and Nakayama, T. ( 1993). CD69 cri1 surface expression identities developing thymocytes which audition for T d l recrptor-mediated positive selection. [nt. Immunol. 5, 1 139- 1 150.
Yang. E.. Zha. J., Jockel. J.. Boise. L. H.. Thompson. C. B., and Korsrneyer. S. J. (1995). Bad, a heterodimeric partner for Bcl-XI and Bcl-3, displaces Bax and promotes ce11 death. Ce11 80, 385-29 1 .
Zamoyska. R.. Derharn. P.. Gorman. S. D.. von Hoegen. P.. Bolen, J. B.. Veillette. A., and Parnes. J. R. ( 1989). Inability of CD8a' polypeptides to associate with p361'k correlates with irnpaired tünction in virro and lack of expression in viiw. Nature 342.178-28 1 .
Zinkernagel. R. Ivl.. Callahan, G. N.. Althage, A., Cooper, S., Klein. P. A.. and Klein, J. ( 1978). On the thymus in the differentiation of "H-2 self-recognition" by T-cells: Evidence for dual recognition'? J. Exp. Med. 147. 882-896.
Zuniga-Pflucker. J. C., McCürthy, S. A., Weston, M.. Longo. D. L.. Singer, A.. and Kruisbeek, A. M. ( 1989). Role o f CD4 in thymocyte selection and maturation. J. Exp. Med. 169, 2085-2096.
Zuniga-Pflucker. J. C., Jones. L. A.. Longo, D. L.. and Kruisbeek. A. M. (1990). CD8 is required during positive srlrction of CD4-/CD8+ T crlls. J. Exp. Mrd. 171. 427-437.
IMAGE NALUATION TEST TARGET (QA-3)
APPLIED 4 IMAGE. lnc 1653 East Main Street
Related Documents
