Research Collection

Doctoral Thesis

Parameters influencing efficient T cell repertoire selection

Author(s): Martinic, Marianne M.A.

Publication Date: 2003

Permanent Link: https://doi.org/10.3929/ethz-a-004618228

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

Diss. ETH No.: 15096

PARAMETERS INFLUENCING EFFICIENT

T CELL REPERTOIRE SELECTION

A dissertation submitted to the

SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH

for the degree of

Doctor of Natural Sciences

presented by

MARIANNE M. A. MARTINIC

Dipl. Natw. ETH

born 29.11.1974

from France

Accepted on the recommendation of

Prof. Dr. H. Hengartner, examiner

Prof. Dr. R.M. Zinkernagel, co-examiner

2003

Meinen Eltern

Table of Contents

- 5 -

Table of Contents

1 Summary..................................................................................................... 7

2 Zusammenfassung .................................................................................... 9

3 Abbreviations ........................................................................................... 11

4 Introduction .............................................................................................. 13

4.1 Thymus architecture and development ................................................. 13

4.2 T cell development................................................................................... 15

4.2.1 Commitment to the ααααββββ or γγγγδδδδ T cell lineage .................................. 18

4.2.2 T cell repertoire selection............................................................. 19

4.2.3 CD4/CD8 T cell lineage commitment ........................................... 21

4.2.4 NK / γγγγδδδδ / CD8αααααααα+ / CD4+CD25+ T cells........................................... 21

4.3 H-Y-specific TCR transgenic mice.......................................................... 23

4.4 Central Question ...................................................................................... 26

5 Results Part I:

Efficient T cell repertoire selection in tetraparental chimeric mice

independent of thymic epithelial MHC ....................................................27

6 Results Part II:

Influence of MHC class I H-2Db density on selection and survival of

H-Y-specific TCR transgenic T cells....................................................... 43

7 Results Part III

Selection of the H-Y-specific transgenic TCR in an athymic versus

euthymic environment............................................................................. 57

8 General Discussion ................................................................................. 73

9 References................................................................................................ 83

10 Curriculum Vitae ...................................................................................... 93

11 Bibliography ............................................................................................. 95

12 Danke, Merci, Thank You, Gracias ......................................................... 97

Summary

- 7 -

1 Summary

During T cell maturation, T cell precursors migrate from the bone marrow via the

bloodstream into the thymus. In the thymus, maturing thymocytes first rearrange their T cell

receptor (TCR)-chain genes followed by a stringent selection process. During this selection

process, only thymocytes expressing productively rearranged TCR with weak to intermediate

overall avidity to self-MHC (major histocompatibility complex)/self-peptide complex (useful

TCR) receive a survival signal (positive selection) whereas thymocytes expressing TCR with

high overall avidity to self-MHC/self-peptide complex (potentially self-reactive TCR) die via

TCR-induced apoptosis (negative selection). This selection process ensures survival of

exclusively self-MHC restricted and self-tolerant thymocytes. During the last maturation step,

thymocytes are committed to the CD4 or CD8 T cell lineage followed by emigration of now

mature and functional T cells into the periphery. Survival and expansion of these T cells,

however, is only guaranteed if they remain in continuous interaction with self-MHC.

The aim of this study is to obtain a better understanding of the requirements needed for

efficient T cell repertoire selection and survival. The first part of the results section addressed

the role of thymic epithelial (TE) versus non-TE MHC in T cell repertoire selection. In the

second and third part of the results section this thesis focused on selection and survival of

the H-Y-specific transgenic TCR, which is specific for a male antigen-derived peptide (H-Y)

presented on MHC class I H-2Db molecules. The influence of TCR-restricting H-2Db density

on H-Y-specific transgenic TCR selection and survival was analysed in the second part of the

results section. Finally, the last part compared selection of the H-Y-specific transgenic TCR

in a euthymic (optimal) versus an athymic (sub-optimal) environment.

The data obtained here showed that 1) non-TE MHC was sufficient to efficiently select a

mature and functional T cell repertoire, 2) efficiency of selection and survival of low-affinity

TCR was dependent on optimal TCR-restricting MHC density and 3) although under sub-

optimal conditions, protection against self-reactivity was still guaranteed, the efficiency of

positive selection, however, was too low to provide protective immunocompetence under

physiological conditions (i.e. in a non-TCR transgenic situation).

Zusammenfassung

- 9 -

2 Zusammenfassung

Während der T-Zell-Reifung wandern Vorläufer-T-Zellen aus dem Knochenmark über die

Blutbahn in den Thymus. Im Thymus beginnen die Thymozyten zuerst mit der

Rearrangierung ihrer T-Zell-Rezeptor (TCR)-Kettengene gefolgt von einem strikten

Selektionsprozess. Während dieses Selektionsprozesses werden nur diejenigen

Thymozyten, die TCR mit einer geringen bis mittleren Gesamtbindungsstärke für Selbst-

MHC (Haupthistokompatibilitätsantigen-Komplex)/Selbst-Peptid-Komplex exprimieren

(nützliche TCR), ein Überlebenssignal erhalten (positive Selektion). Thymozyten, deren TCR

eine hohe Gesamtbindungsstärke für Selbst-MHC/Selbst-Peptid Komplex aufweisen

(potentiell selbst-reaktive TCR), sterben durch TCR-vermittelte Apoptose (negative

Selektion). Dieser Selektionsprozess stellt somit sicher, dass ausschliesslich Selbst-MHC-

restringierte und selbst-tolerante Thymozyten überleben. Während der letzten Reifungs-

Etappe erfolgt der Entscheid zur CD4- oder zur CD8-positiven T Zelle mit anschliessender

Emigration in die Peripherie als reife und funktionelle T Zelle. Das weitere Überleben sowie

die Expansion dieser T Zelle sind nur dann garantiert, wenn sie in ständigem Kontakt mit

Selbst-MHC bleibt.

Das Ziel dieser Arbeit ist es, ein besseres Verständnis über die verschiedenen Bedingungen

zu erhalten, die zum Erreichen einer effizienten T Zell Repertoire-Selektion notwendig sind.

Im ersten Resultate-Teil wurde die Funktion von Thymusepithel (TE) versus Nicht-TE MHC

bei der T Zell Repertoire-Selektion untersucht. In den darauffolgenden Resultate-Teilen

wurden die Selektion und das Überleben des H-Y-spezifischen, transgenen TCR analysiert,

welcher ein männliches Peptidantigen auf MHC Klasse I Molekül H-2Db spezifisch erkennt.

Der Einfluss der TCR-restringierenden H-2Db-Dichte auf die Selektion und das Überleben

des H-Y-spezifischen transgenen TCR wurde im zweiten Resultate-Teil verfolgt. Zum

Abschluss wurde im dritten Resultate-Teil die Selektion des H-Y-spezifischen transgenen

TCR in einer athymischen (sub-optimal) und in einer normalen Umgebung verglichen.

Die hier erhaltenen Daten konnten zeigen, 1) dass Nicht-TE MHC ausreichend war, um ein

reifes und funktionelles T Zell Repertoire zu selektionieren, 2) dass die Effizienz der

Selektion und des Überlebens niedrig affiner TCR von einer optimalen Dichte an TCR-

restringierenden MHC-Molekülen abhing und 3) dass unter sub-optimalen Bedingungen der

Schutz gegen Selbst-Reaktivität zwar gegeben, die Effizienz der positiven Selektion dagegen

aber viel zu niedrig war, um unter physiologischen Bedingungen (z. B. in einer nicht-TCR

transgenen Situation) noch genügend schützende Immunkompetenz zu vermitteln.

Abbreviations

- 11 -

3 Abbreviations

APC Antigen presenting cells

BM Bonne marrow

CLP Common lymphoid precursors

cTEC Cortical TEC

CTL Cytotoxic T lymphocytes

DAG Diacylglycerol

DC Dendritic cells

DEC Dendritic epidermal cells

DN Double negative

DP Double positive

ED Embryonic day

ELISA Enzyme-linked immunosorbent assay

end. Endogenous

ERK Extracellular-signal-regulated kinase

FL Fetal liver

FLT3+ HSC Fms-like tyrosine kinase 3+ HSC

Gads SH2 and SH3 domain-containing adaptor

proteins, bind to tyrosine-phosphatase Shc

GPI Glucose-6-phosphate-isomerase

Grb2 Growth-factor receptor-bound protein 2

HSA Heat stable antigen, CD24

HSC Haematopoietic stem cells

H-Y Male antigen-derived peptide

IEL Intestinal intraepithelial lymphocytes

Ig Immunoglobulin

IgH Immunoglobulin heavy chain

IgL Immunoglobulin light chain

IP3 Inositol-1,4,5-trisphosphate

ISP Immature single positive

ITAM Immuno-tyrosine based activation motif

Itk Tec-family tyrosine kinase

JNK c-jun N-terminal kinase

LAT Linker for activation of T cells

LCMV Lymphocytic choriomeningitis virus

Abbreviations

- 12 -

LCMV-GP LCMV-glycoprotein

LCMV-NP LCMV-nucleoprotein

MFI Mean fluorescence intensity

MHC Major histocompatibility complex

MLP Myeloid- and lymphoid precursors

mTEC Medullary TEC

NF-κB Nuclear factor κB

NF-AT Nuclear factor of activated T cells

NK Natural killer cells

p38 MAP kinase

pfu Plaque-forming units

PKC Protein kinase C

PLC-γ1 Phospholipase C-γ1

PNAr Peanut agglutinin receptor

pTα Invariant pre-TCRα chain

Rag-1,2 Recombination-activating genes-1,2

Ras Protein products regulate cellular growth and

differentiation (family of proto-oncogenes)

RasGRP Ras activator with a DAG-binding C1 domain

Scid Severe combined immunodeficiency

Slp76 Cellular adaptor protein

SP Single positive

TCR T cell receptor

TCRα-CPM TCRα-chain connecting peptide motif

TE Thymic epithelial

TEC Thymic epithelial cells

tg Transgenic

TN Triple negative

VSV Vesicular stomatitis virus

VSV-IND VSV serotype Indiana

WHN Winged-helix nude

ZAP-70 Zeta-Associated Protein-70 Src-family

tyrosine kinase

Introduction

- 13 -

4 Introduction

4.1 Thymus architecture and development

In mice as well as in humans, the thymus is the organ where progenitor T cells

differentiate to mature T cells expressing a self-MHC-restricted and self-tolerant T

cell receptor (TCR). The thymus lies in the upper thorax of vertebrate animals,

resting on the heart and extending into the base of the neck. It is bilaterally

symmetrical, composed of two lobes, which join at the midline. It consists primarily of

T cells at different developmental stages (thymocytes) and a heterogeneous group of

supporting cells - thymic epithelial cells, fibroblasts, macrophages, dendritic cells and

B cells - forming the thymic stroma. A capsule of connective tissue, which repeatedly

invaginates to form septae leading to numerous lobules, surrounds the thymic lobes,

each filled with thymocytes and stromal cells (Figure 1). Each thymus lobule is

organized into three different regions: a subcapsular, cortical and medullary zone.

Baskets of epithelial reticular cells characterize the subcapsular region. Fibroblasts,

macrophages and a network of spider-shaped and sheetlike epithelial cells are

located in the cortical region. The medullary region consists of numerous dendritic

cells, macrophages and a network of stubby epithelial cells. B cells are mostly

present in the corticomedullary junction. The different cell-type compositions in each

region lead to distinct MHC-staining patterns: a weak staining in the subcapsular

region, a reticular pattern in the cortex (due to epithelial network) and a confluent

pattern in the medulla (due mainly to dendritic cells and a subset of medullary

epithelial cells) (Benoist and Mathis, 1999).

In mice, the beginning of thymus development starts at fetal day 9 with the

invagination of the ectoderm of the third branchial cleft and the endoderm of the third

or fourth pharyngeal pouch (Anderson et al., 1996). With continuous invagination, the

ectoderm surrounds the endoderm. The developing organ becomes surrounded by

mesenchymal cells initially of neural crest and later of mesodermal origin, playing an

inductive as well as structural role during thymus development (Le Lievre and Le

Douarin, 1975). At fetal day 11, invaginations close to form a thymic rudiment or

anlage. Cells of ectodermal origin give rise to cortical thymic epithelial cells (cTEC);

cells of endodermal origin give rise to medullary TEC (mTEC). Blood vessels

associated with mesoderm-derived mesenchyme permeate the thymus anlage

Introduction

- 14 -

allowing cells of haematopoietic origin (from the yolk sac or fetal liver) to enter. From

this point on, all progenitor T cells as well as stromal cells are present in the

developing thymus. Thymus architecture, however, is not completed yet. There is no

distinct division into a subcapsular, cortical and medullary region. Only few medullary

islets are present in the dominating cortical region. Just before birth a distinct

medullary region appears in parallel with the emergence of mature αβT cells. The

right three-dimensional thymus architecture as well as development of mature T cells

in the thymus is only achieved in the presence of a proper cross talk between

developing T cells and thymic stromal cells (Anderson and Jenkinson, 2001).

Nude mice have a loss-of-function mutation in the WHN (winged-helix nude) gene

encoding for the transcription factor WHN. The thymus anlage can still develop in

absence of WHN but it will only contain primitive epithelial cells unable to differentiate

into mature thymic epithelial cells. The nude thymic rudiment will therefore never

establish a functional three-dimensional epithelial network nor have the ability to

promote the development of progenitor T cells into mature T cells.

Figure 1: Thymus organization. A thymus section through a single lobule is shown. The different forms of stromal

cells in each region are indicated: cortical and medullary epithelial cells, dendritic cells, macrophages, and

fibroblasts. The maturation of thymocytes from blasts in the subcapsular zone, to small resting cells in the cortex,

to larger cells in the medulla is also illustrated (Benoist and Mathis, 1999).

Introduction

- 15 -

Figure 2: Development of TCRαβ+ T cells in the thymus. The major different stages and sites of thymocyte-

development are shown from the most immature stage on the left (DN stage) to the most mature stage on the

right (SP stage). Time-span of different TCR-chain gene rearrangements, expression of pre-TCR and TCRαβ,

degree of cell division, time spent at the various stages, percentage of thymocytes belonging to the different

stages out ot total thymocytes, anatomic location in which thymocytes reside during the different stages, and total

thymocyte numbers passing the thymus per day are shown from top to bottom. Adapted from (Benoist and

Mathis, 1999).

4.2 T cell development

Haematopoietic stem cells (HSC) are pluripotent and show a self-renewal capacity.

In the murine fetal liver, HSC differentiate to lymphoid precursors with myeloid- and

lymphoid- reconstitution potential (FL-MLP) (Akashi et al., 2000; Cumano et al.,

1992; Sagara et al., 1997) and then further on to either cells committed to the B or to

the T cell lineage (Kawamoto et al., 2000; Kawamoto et al., 1997) (Figure 2).

Whereas during embryonic development T lineage commitment occurs prior to

thymic entry, in adults, T lineage commitment occurs in the thymus and is Notch-1

signalling-dependent (Ashton-Rickardt et al., 1994; Osborne and Miele, 1999). In the

adult bone marrow (BM), HSC differentiate first to Flt3+ HSC (fms-like tyrosine kinase

3) with transient myeloid and long-term lymphoid reconstitution potential and then to

common lymphoid precursors (CLP) (Figure 2). These cells show only limited self-

Introduction

- 16 -

renewal capacity and have a potential of producing lymphoid but not myeloid cell

types such as T, B, natural killer (NK) and dendritic cells (DC) (Borowski et al., 2002).

In absence of Notch-1 signalling, CLP develop to the Pro-B cell stage (Borowski et

al., 2002). In presence of Notch-1 signalling, CLP migrate via the bloodstream into

the thymus to further develop to the Pro-T cell stage. CLP first entering the thymus at

the corticomedullary junction are CD4-CD8-TCR- triple negative (TN) (Figure 2). This

stage can further be divided into four different stages depending on expression levels

of CD44 and CD25. The most immature thymocytes are CD44+CD25- (DN1) (Figure

2). These cells still can be committed to the NK, DC or B cell lineage. In presence of

Notch-1 signalling, DN1 thymocytes up-regulate CD25 to become CD44+CD25+

(DN2) (Figure 2). DN2 thymocytes migrate to the subcapsular region of the thymus

and show an IL-7-dependent proliferation (Rodewald et al., 1997). DN2 cells still can

develop to DC but not any more to NK or B cells (Benoist and Mathis, 1999) (Figure

2). After down-regulating CD44, CD44loCD25+ DN3 thymocytes stop proliferating.

From this developmental stage on, all cells are committed to the T cell lineage

(Benoist and Mathis, 1999). DN3 thymocytes start to rearrange β-, γ- and δ-TCR

chain genes (Figure 2). A productive γδ-chain gene rearrangement followed by

expression of a functional γδTCR leads to commitment to the γδ T cell lineage

(MacDonald et al., 2001) (Figure 2). A productive β-chain gene rearrangement

followed by successful pairing with the invariant pre-TCRα chain (pTα) forming the

pre-TCR commits cells to the αβ T cell lineage (MacDonald et al., 2001). Further

thymocyte survival and differentiation along the appropriate lineage is dependent on

signals from the correctly assembled CD3+ γδTCR or CD3+ αβTCR (Starr et al.,

2002).

pTα is a type 1 transmembrane protein with a single immunoglobulin-like

extracellular domain (Saint-Ruf et al., 1994). Pre-TCR signalling in contrast to αβTCR

or γδTCR signalling occurs in a cell-autonomous or constitutive manner (Saint-Ruf et

al., 2000). Productive pre-TCR signalling - might occur in the absence of any

extrathymic ligand (Irving et al., 1998) - leads to thymocyte survival and down-

regulation of CD25 (DN4 stage) (Figure 2). CD44-CD25-pre-TCRlo DN4 thymocytes

express already some CD3 subunits. DN4 thymocytes still can develop to γδ T cells

(Petrie et al., 1992). However, it is not known whether these cells down-regulated

CD25 before or after commitment to the γδ T cell lineage and whether they are

Introduction

- 17 -

exactly the same CD44-CD25- DN4 thymocytes that develop later on to αβ T cells

(Benoist and Mathis, 1999) (Figure 2). DN4 thymocytes show a massive increase in

cell size and proliferation, allelic exclusion at the TCRβ locus and CD4 and CD8 co-

receptors up-regulation (Benoist and Mathis, 1999; Starr et al., 2002). Some α-chain

gene rearrangements are observed (Benoist and Mathis, 1999; Starr et al., 2002)

(Figure 2). Between DN4 and double-positive (DP) stage, thymocytes express low

levels of CD8 representing immature single positive (ISP) thymocytes. This

developmental stage cannot be seen in vitro (Benoist and Mathis, 1999) (Figure 2).

CD4+CD8+pre-TCRlo DP thymocytes show strongly up-regulated Rag gene

expression resulting in α-chain gene rearrangements. Productive α-chain gene

rearrangement does not prevent further Vα-Jα rearrangement within the same cell,

there is no feedback inhibition as seen in productively rearranged TCRβ- ,

immunoglobulin (Ig) heavy (H)- and Ig light (L)-chains (Borgulya et al., 1992; von

Boehmer, 1990). Whereas during B cell development, most properly assembled H/L

chain pairs trigger allelic exclusion and further differentiation (Casellas et al., 2001;

Nemazee, 2000), during T cell development TCRαβ assembly itself is not enough to

promote further differentiation and Rag gene repression (Starr et al., 2002). TCRα-

chain gene rearrangement stops only then when the newly rearranged TCRαβ has

been positively selected (Borgulya et al., 1992; Borowski et al., 2002; Brandle et al.,

1992). One third of mature T cells harbour two productively rearranged TCRα alleles

(Casanova et al., 1991), however, less than 30% of mature T cells express actually

two different TCRαβ complexes due to strong competition for binding to TCRβ-chain

(Heath et al., 1995; Kisielow et al., 1988b).

The development of DN thymocytes to DP thymocytes takes about 14 days and

comprises around 3% of the whole thymocyte fraction (Benoist and Mathis, 1999)

(Figure 2). DP thymocytes are located in the cortical region of the thymus (Benoist

and Mathis, 1999). With continuous maturation, large and cycling DP thymocytes

(CD4+CD8+pre-TCRlo) stop cycling, return to small cell size, up-regulate Rag gene

expression and rearrange α-chain genes. Productively rearranged α-chain

assembles with the already expressed TCRβ -chain to form TCRαβ

(CD4+CD8+TCRαβlo) (Figure 2). The TCRαβ of these DP thymocytes is subjected to

a selection process to allow only survival of DP thymocytes, which express a self-

MHC-restricted and self-tolerant TCR. DP thymocytes expressing a TCRαβ not able

Introduction

- 18 -

to recognize self-MHC are eliminated by glucocorticoid-induced apoptosis (death by

neglect) (Surh and Sprent, 1994). DP thymocytes expressing a TCRαβ, which

recognizes self-peptide/MHC complexes with too high avidity and therefore be

potentially auto-reactive, get eliminated by TCR-induced apoptosis (death by

negative selection) (Benoist and Mathis, 1999). During this selection process over

95% of DP thymocytes get deleted (see section 4.2.2) (Figure 2). Only positively

selected DP thymocytes (CD4+CD8+TCRαβhi) (Figure 2) are able to produce a signal

that leads to Rag-gene repression and therefore stop of further α-chain gene

rearrangement, long-term survival, migration into medullary region of the thymus and

further differentiation into mature T cells (Wilkinson et al., 1995). Thymocytes remain

3-4 days in the DP stage and comprise about 80% of the whole thymocyte fraction

(Benoist and Mathis, 1999) (Figure 2).

The last step of thymocyte maturation involves commitment to the CD4- or CD8 T

cell lineage. Positively selected DP thymocytes down-regulate one of either co-

receptors to become unreactive CD4 single positive (CD4-SP) or CD8-SP T cells

(TCR/CD3intHSAhiPNArhi) (Figure 2). The exact mechanisms involved during this

maturation step are not fully known yet. The three current models for CD4/CD8

lineage commitment are discussed in a later section (see section 4.2.3). SP

thymocytes remain 7-14 days in the medullary region and comprise about 15% of the

total thymocyte fraction (Benoist and Mathis, 1999) (Figure 2). They are exported

from the thymus as fully mature and competent SP cells (TCR/CD3hiHSAloPNArlo) via

blood and lymph vessels at the corticomedullary junction (Chaffin and Perlmutter,

1991) (Figure 2).

4.2.1 Commitment to the ααααββββ or γγγγδδδδ T cell lineage

Parameters involved for commitment to either αβ or γδ T cell lineage are not

completely known yet. The current favoured model is the competitive model (Benoist

and Mathis, 1999). This model predicts that αβ and γδ T cells derive from a common

precursor. During a certain period of thymocyte differentiation (mostly between the

DN3 and DN4 stage) until now unknown factor(s) will favour commitment to either αβ

or γδ T cell lineage. β-, γ- and δ-chain gene rearrangement begins at the DN3 stage.

Thymocytes with a productive γ- and δ-chain gene rearrangement followed by

expression of a functional γδTCR will develop along the γδ T cell lineage. Thymocytes

Introduction

- 19 -

with a productive β-chain gene rearrangement and successful pairing with pTα

before productive γδ-chain gene rearrangement and γδTCR expression will develop

along the αβ T cell lineage. Definitive commitment to the αβ T cell lineage, however,

is only given after α-chain gene rearrangement on both alleles followed by excision

of the δ-chain locus. Commitment to the αβ T cell lineage is probably the most

frequent because it only implies one productive β-chain gene rearrangement versus

two productive γ- and δ-chain gene rearrangements for the γδ pathway.

4.2.2 T cell repertoire selection

DP thymocytes expressing low levels of the newly rearranged TCRαβ

(CD4+CD8+TCRαβlo) go through a very strict selection process to ensure survival of

exclusively self-MHC-restricted and self-tolerant T cells. Over 95% of these DP

thymocytes will die during this selection process. Positive selection is defined as the

rescue of self-MHC-restricted DP thymocytes from programmed cell death (Sebzda

et al., 1999); negative selection is defined as deletion of potentially auto-reactive DP

thymocytes by TCR-induced apoptosis (Benoist and Mathis, 1999); DP thymocytes

which do not recognize the self-peptide/MHC complex are neglected and die by

glucocorticoid-induced apoptosis (Benoist and Mathis, 1999). The current model to

explain T cell repertoire selection on a molecular level is the differential TCR

signalling model (Figure 3) (Starr et al., 2002). DP thymocytes with an overall weak

to intermediate avidity interaction between their TCRαβ and self-peptide/MHC

complexes will receive a survival signal whereas DP thymocytes with an overall

strong avidity interaction between their TCRαβ and self-peptide/MHC complexes will

receive a signal to undergo programmed cell death (Figure 3) (Starr et al., 2002). The

overall avidity interaction between TCR and self-peptide/MHC complexes includes

the intrinsic affinity of the TCR (and its co-receptor) for self-peptide/MHC complex,

expression levels of TCR, co-receptor, MHC molecules and adhesion molecules,

peptide concentration, and duration of the contact between the TCR and self-

peptide/MHC complex (Ashton-Rickardt et al., 1994; Sebzda et al., 1999).

With continuous differentiation, DP thymocytes increase their TCR expression levels.

Potentially auto-reactive thymocytes, which at the DP-TCRαβlo developmental stage

still get a survival signal after contact with self-peptide/MHC, will receive at the DP-

TCRαβhi stage a stronger signal and therefore get negatively selected. The increase

Introduction

- 20 -

in TCR-expression levels therefore increases the safety net that prevents mature T

cell auto-reactivity (Sebzda et al., 1999).

Figure 3: Differential TCR Signalling Model. (A) Positive Selection. A weak to intermediate overall avidity

interaction between TCR and self-peptide/MHC complex stimulates the TCR, which transmits a signal via the

TCRα-CPM, CD3γ-ITAM and CD3δ activating ZAP-70, which will partially phosphorylate LAT. Partial LAT

phosphorylation recruits PLC-γ1 and Gads, Slp7, Itk, resulting in activation of PLC-γ1 and production of DAG and

IP3. RasGRP is turned on by DAG and stimulates sustained, but low-level ERK-activation leading to thymocyte

survival. NF-AT and NF-κB-activation leads to thymocyte proliferation and differentiation. (B) Negative selection.

A high overall avidity interaction between TCR and self-peptide/MHC complex results in fully phosphorylated LAT,

Introduction

- 21 -

recruitment of Grb2/Sos1, p38- and JNK-activation, and transient, but strong ERK-activation leading to thymocyte

apoptosis. Adapted from (Starr et al., 2002).

4.2.3 CD4/CD8 T cell lineage commitment

During the last step of thymocyte development, positively selected DP thymocytes

(CD4+CD8+TCRαβhi) will down-regulate either one of their co-receptors to exit the

thymus as mature single-positive T cells. Three models – “instructive model”,

“stochastic model” and “co-receptor reversal” or “Singer instructive model” – are

currently present to explain CD4/CD8 T cell lineage commitment (Borowski et al.,

2002; Singer, 2002). The first model, “instructive model”, predicts that the co-receptor

instructs the final phenotype of mature T cells. Recognition of the appropriate self-

peptide/MHC complex by the TCR allows co-ligation of the appropriate co-receptor,

which induces a co-receptor signal instructing the DP thymocyte to terminate

expression of the inappropriate co-receptor. The second model, “stochastic model”,

predicts that the DP thymocyte stochastically terminates expression of one co-

receptor. Only thymocytes expressing the co-receptor capable of binding the MHC

molecule for which their TCR is restricted to, will receive a survival signal. In the

latest model, “co-receptor reversal” or “Singer-instructive model”, recognition of the

appropriate MHC molecule by the TCR is followed by a pre-programmed CD8 down-

regulation (Brugnera et al., 2000). Interaction between MHC I-restricted TCR and

MHC I molecule results in a weak TCR-signal. In presence of IL-7, co-receptor

reversal leads to CD4-silencing, CD8 up-regulation and final maturation to the CD8-

SP stage. Interaction between MHC II-restricted TCR and MHC II molecule results in

a strong TCR-signal. Cells are now refractory to co-receptor reversal and develop

therefore to CD4-SP cells (Brugnera et al., 2000).

4.2.4 NK / γγγγδδδδ / CD8αααααααα+ / CD4+CD25+ T cells

The first γδ T cells appear at embryonic day (ED) 14 (Benoist and Mathis, 1999).

Between ED14 and ED17 γδ T cells expressing Vγ5+Vδ1+ TCR predominate. They

are defined as dendritic epidermal cells (DEC) due to their dendritic morphology and

location in the epidermis. DEC seem to be selectively activated by a product of

“stressed” keratinocytes leading to secretion of Th1-like inflammatory cytokines as

well as lymphotactin resulting in recruitment of conventional lymphocytes (Benoist

and Mathis, 1999). In addition, they support the growth of epithelial tissues by

secreting epithelial growth factors, especially during wound healing (Benoist and

Introduction

- 22 -

Mathis, 1999). Vγ6+ T cells are the second wave of γδ T cells. They are produced

between ED17 and birth and migrate to the reproductive tract and tongue where they

reside as intramucosal γδ T cells (Benoist and Mathis, 1999). After birth, a more

heterogeneous population of γδ T cells is produced. These cells populate the thymus,

gut, spleen and other secondary lymphoid organs. Comparable to conventional αβ T

cells, γδ T cells show cytotoxicity, provide B cell help via CD40-CD40L interaction,

and activate macrophages via IFN-γ secretion (Benoist and Mathis, 1999). In

addition, they are able to respond to microbiologic infections (Hiromatsu et al., 1992;

Ladel et al., 1995; Mombaerts et al., 1993) and to recognize non-peptidic antigens

(Constant et al., 1994; Pfeffer et al., 1990; Tanaka et al., 1994). They can even

directly recognize antigen without processing and presentation by MHC-like

molecules (Schild et al., 1994; Weintraub et al., 1994).

Natural Killer (NK) T cells, CD8αα+ intestinal intraepithelial lymphocytes (IEL) and

CD4+CD25+ regulatory T cells are produced during T cell development in the thymus.

They all show an activated phenotype, seem to exert regulatory functions and to

require a high-affinity interaction with self-antigen for proper development (Bendelac

et al., 1996; Capone et al., 2001; Curnow et al., 1995; Hayday et al., 2001; Levelt et

al., 1999; Rocha et al., 1992; Starr et al., 2002; Sydora et al., 1993).

NK T cells have a thymic precursor (DN TCRαβ+ NK1.1+), express NK1.1, are

predominantly Vα14+ and selected on the non-classical MHC I molecule CD1

(Bendelac et al., 1997; Benlagha et al., 2002; Guy-Grand et al., 2003). The natural

ligand is unknown. They regulate conventional T cell responses through cytokine

secretion (Bendelac et al., 2001).

CD8αα+ TCRαβ+ IEL are predominantly found in the gut epithelium (Starr et al.,

2002). Just recently it was suggested that DN TCRαβ+ NK1.1- thymocytes are the

thymic precursors of CD8αα+ IEL (Guy-Grand et al., 2003). These precursors reach

the gut epithelium via mesenteric lymph nodes and the thoracic duct lymph where

they up-regulate CD8αα-expression (Guy-Grand et al., 2003). CD8αα+ TCRαβ+ IEL

express classical or non-classical MHC I-restricted TCR, use both ζ- and γFcεRI-

chains as CD3-associated signal transmitting module, express Ly49 NK receptors

and display NK cytotoxic abilities (Guy-Grand et al., 2003; Guy-Grand et al., 1996;

Guy-Grand et al., 1994; Park et al., 1995; Starr et al., 2002).

Introduction

- 23 -

CD4+CD25+ regulatory T cells are able to prevent the development of gastritis, colitis

and diabetes in vivo and to inhibit T cell proliferation in vitro (Read and Powrie, 2001;

Starr et al., 2002). Whereas development of CD4+CD25+ regulatory T cells is directed

by TEC, BM-derived antigen presenting cells (APC) in the thymus direct development

of CD4+CD25- regulatory T cells (Apostolou et al., 2002; Jordan et al., 2001). In

contrast to conventional αβ T cells, which get negatively selected (clonal deletion)

after a high overall avidity interaction between their TCR and self-peptide/MHC

complexes, regulatory T cells require a high-affinity interaction with self-antigen for

their development. Differences in avidity and the type of APC presenting the self-

peptide/MHC complex might decide if clonal deletion (central tolerance) and/or

selection of regulatory T cells occurs during thymocyte development (Apostolou et

al., 2002; Bensinger et al., 2001; Jordan et al., 2001; Jordan et al., 2000; Modigliani

et al., 1996; Read and Powrie, 2001). Regulatory T cells present in the periphery

could inhibit auto-reactive T cells having escaped clonal deletion in the thymus and

therefore increase even further the safety net to prevent mature T cell auto-reactivity.

4.3 H-Y-specific TCR transgenic mice

H-Y-specific T cell receptor (TCR) transgenic (tg) mice express a transgenic TCRαβ

(Vα3+Vβ8.2+) specific for a male antigen-derived peptide (H-Y) presented on MHC

class I H-2Db molecules (Kisielow et al., 1988a). The minor histocompatibility (H)

male specific (Y) antigen (H-Y) (KCSRNRQYL), which is expressed in all male

tissues, is derived from a protein encoded by the SMCY gene located on the Y

chromosome (Simpson et al., 1997). The transgenic TCRαβ can be followed by flow

cytometry analysis using transgenic TCRα- and TCRβ-specific antibodies (T3.70

and Vβ8.1/2, respectively) (Teh et al., 1989).

In males, presence of the self-peptide H-Y together with the TCR-restricting MHC

molecule H-2Db leads to negative selection in the thymus of H-Y-specific TCR

transgenic CD8+ T cells (Table 1 and Figure 4, first row) (Kisielow et al., 1988a).

CD4-CD8- tg TCRαβhi cells which were shown to be aberrant γδ-lineage T cells, and

cells showing a down-regulation of their co-receptor and transgenic TCR, resulting in

CD8αβlo tg TCRαβ lo-hi cells, are also present in TCR tg male mice (Table 1 and

Figure 4, first row) (Bruno et al., 1996). These double negative (DN) tg TCRαβhi and

CD8αβlo tg TCRαβlo-hi cells are not H-Y-reactive anymore, can therefore escape

Introduction

- 24 -

negative selection and migrate into the periphery as mature self-MHC-restricted and

self-tolerant T cells. These cells do not recognize H-Y/H-2Db complex, do not

proliferate after contact with male stimulator cells but are functional against other

antigens (Kisielow et al., 1988a; Teh et al., 1989). In females, absence of the peptide

H-Y together with the presence of the TCR-restricting MHC molecule H-2Db leads to

positive selection in the thymus of H-Y-specific CD8αβhi tg TCRαβhi cells (Table 1

and Figure 4, second row) (Kisielow et al., 1988a). Because the tg TCRαβ is already

expressed at the DN stage during thymocyte maturation, few DN tg TCRαβ+ cells are

also found in the periphery of TCR transgenic female mice representing the aberrant

γδ-lineage cells (data not shown) (Bruno et al., 1996). However, CD8αβlo tg TCRαβlo-

hi cells as seen in TCR transgenic males are not found in TCR transgenic females

(Figure 4, second row). TCR transgenic mice, which do not express the TCR-

restricting MHC molecule H-2Db do not show positive selection of H-Y-specific

CD8αβhi tg TCRαβhi cells (Table 1 and Figure 4, third row) (Teh et al., 1988; von

Boehmer, 1990). Peripheral DN tg TCRαβ+ and CD8αβlo tg TCRαβ lo-hi cells are

absent in TCR transgenic mice on a non-selecting background (Figure 4 and data not

shown). H-Y-specific CD8αβhi tg TCRαβhi cells escape death from neglect via

endogenous rearrangement of TCRα-chain genes (Table 1 and Figure 4, third row

and data not shown) (Teh et al., 1988). In summary, positive selection of H-Y-specific

TCR transgenic CD8+ cells requires presence of TCR-restricting MHC class I

molecule H-2Db and absence of self-peptide H-Y. Transgenic TCRαβ does not get

positively selected by the presence of MHC class II molecules of the selecting H-2b

haplotype nor by the presence of MHC class I molecules of H-2b, d or k haplotype

(Table 1 and Figure 4 and data not shown) (Kisielow et al., 1988b; Teh et al., 1988;

von Boehmer, 1990).

Introduction

- 25 -

Table 1: Selection of H-Y-specific TCR transgenic T cells.

Mouse T cell selection

Positive Selection

Phenotype of selected cells

CD8αβhi tgTCRαβ+ cells (Figure 4)

H-Y/H-2Db-

spec. responses

YES

Negative Selection No CD8αβhi tgTCRαβ+ cells (Figure 4)

Escape from Neg. Selection

CD4-CD8- tgTCRαβ+ cells (not shown)

NOCD8αβlo tgTCRαβ+ cells (Figure 4)

Gender

F2 (H-Y)

H-2dd

F0 (H-Y)

H-2bb

F0 (H-Y)

H-2bb

No CD8αβhi tgTCRαβ+ cells (Figure 4)Neglect

NOEscape from Neglect

(Figure 4 and not shown)

CD8αβhi TCRαend.+βtg

+ cells

Gated on live blood lymphocytes Gated on CD8αβ+ cells

50

101

103

80

101

103 91.2%

A)

H-2bb

F0 (H-Y)

73701

50

29

Co

un

ts

101

103 27.0%

Vβ8

.1/2

CD

8β

101

103

Co

un

ts80

485B)

H-2bb

F0 (H-Y)

CD8α CD8αCD8β T3.70 T3.70Vβ8.1/2101 103

80

050

101 103

-

101 103

772

101

103

101 103

0.6%

101

103

101 103

080

101 103

492C)

H-2dd

F2 (H-Y)

CD8lo CD8lo

CD8hi

CD8hiCD8hi

CD8hi

128MFI

61MFI

28MFI

150MFI Escape from Neg. Selection:

CD8αβlo tgTCRαβ+ (UR)

Positive Selection:

CD8αβhi tgTCRαβ+ (UR)

Escape from Neglect:

CD8αβhi TCRαend.+βtg

+ (UL)

Figure 4: Selection of H-Y-specific transgenic TCR in mice with selecting (H-2bb) or non-selecting (H-2dd) H-2

haplotype. Blood lymphocytes of H-Y-specific TCR transgenic males (F0 (H-Y), H-2bb, �) (A) and females (F0 (H-

Y), H-2bb, �) (B) with selecting (H-2 bb) or non-selecting (H-2dd) H-2 haplotype (F2 (H-Y), H-2dd, � �) (C) were

stained for surface expression of CD8α- (column 1) and CD8β-chain (column 2). Numbers in histogram plots

represent mean fluorescence intensity (MFI) of CD8α and CD8β, respectively. Circles in dot plots (column 3)

represent CD8αβlo and CD8αβhi populations. Percentage of CD8αα+ cells was always below 0.5%. Blood

lymphocytes of H-Y-specific TCR transgenic mice were gated on CD8αβ+ cells (CD8αβlo for H-Y-specific TCR

transgenic males (A), CD8αβhi for H-Y-specific TCR transgenic females (B) and H-Y-specific TCR transgenic

mice with a non-selecting H-2 haplotype (C)) and stained for surface expression of transgenic α- (T3.70) (column

4) and transgenic β-chain (Vβ8.1/2) (column 5) of H-Y-specific TCR. Numbers in histogram plots represent MFI of

transgenic TCRα and transgenic TCRβ-chain, respectively. Numbers in upper right (UR) quadrants represent

percentage of CD8αβ+ cells expressing transgenic TCRαβ (column 6).

Introduction

- 26 -

4.4 Central Question

The key question of this thesis is which parameters are involved in selection and

survival of a functional and mature T cell repertoire. The first part of the Results

section addressed the role of thymic epithelial (TE) versus non-TE MHC in T cell

repertoire selection (Results Part I). The second part analysed the influence of TCR-

restricting MHC density on selection and survival of the above already described low-

affinity H-Y-specific transgenic TCR (Results Part II). The last part of the Results

section compared intra- and extrathymic selection of the H-Y-specific transgenic TCR

(Results Part III). The aim of this study and all performed experiments was to obtain

at the end a better understanding of the different requirements for efficient T cell

repertoire selection so that in a near future we might be able to improve therapies

against autoimmune diseases and for successful transplantations.

Results Part I

- 27 -

5 Results Part I

Efficient T cell repertoire selection in tetraparental chimeric mice

independent of thymic epithelial MHC

Marianne M. Martinic*¶||, Thomas Rülicke†||, Alana Althage*,**, Bernhard

Odermatt‡, Matthias Höchli§, Alain Lamarre*††, Tilman Dumrese*, Daniel E.

Speiser*‡‡, Diego Kyburz*§§, Hans Hengartner* and Rolf M. Zinkernagel*¶

*Institute of Experimental Immunology, Department of Pathology, University Hospital Zurich,

Schmelzbergstrasse 12, CH-8091 Zurich, Switzerland†Institute of Laboratory Animal Science, Central Biological Laboratory, University Hospital

Zurich, Sternwartstrasse 6, CH-8091 Zurich, Switzerland‡Laboratory of Molecular Diagnostics, Department of Pathology, University Hospital Zurich,

Schmelzbergstrasse 12, CH-8091 Zurich, Switzerland§Laboratory of Electron Microscopy, University of Zurich, Gloriastrasse 30, CH-8028 Zurich,

Switzerland¶To whom reprint requests should be addressed. Phone: +41-1-255 29 89, Fax: +41-1-255

44 20, E-mail: marmar@pathol.unizh.ch, rolf.zinkernagel@pty.usz.ch.||These authors contributed equally to this work.**Present address: Department of Molecular and Experimental Medicine, The Scripps

Research Institute, 10550 North Torrey Pines Rd., 92037 La Jolla, CA††Present address: Institut National de la Recherche Scientifique-Institut Armand-Frappier,

531 Boul. des Prairies, H7V 1B7 Laval, Quebec, Canada‡‡Present address: Ludwig Institute for Cancer Research / CHUV, Division of Clinical Onco-

Immunology, Hôpital Orthopédique, Av. Pierre-Decker 4, CH-1005 Lausanne, Switzerland§§Present address: Rheumaklinik und Institut für Physikalische Medizin, University Hospital

Zurich, Gloriastrasse 25, CH-8091 Zurich, Switzerland

Abbreviations: LCMV, lymphocytic choriomeningitis virus; LCMV-GP, LCMV glycoprotein;

LCMV-NP, LCMV nucleoprotein; VSV, vesicular stomatitis virus; CTL, cytotoxic T

lymphocyte; pfu, plaque-forming units.

Results Part I

- 28 -

ABSTRACT

Non-thymic epithelial cells were compared to thymic epithelial cells for their

role in T cell repertoire selection. Tetraparental aggregation chimeras were

generated from T- and B cell-deficient mice (H-2d Scid or H-2b Rag-/-) and

thymus-deficient nude mice (H-2b or H-2d). These tetraparental mice showed

primary protective CD8+ T cell responses after lymphocytic choriomeningitis

virus (LCMV) infection, which were restricted to either thymic or non-thymic

epithelial MHC at comparable levels. These chimeras also mounted

neutralizing IgG responses dependent upon cognate CD4+ T helper cell activity

restricted to non-thymic epithelial MHC. Therefore, in contrast to earlier results

with irradiation or thymus chimeras, these relatively undisturbed tetraparental

mice reveal that the MHC of non-thymic epithelial cells efficiently selects a

functional T cell repertoire.

It is well established that the thymus is essential for T cell receptor rearrangement

and T cell maturation (Miller, 1961). It is also widely accepted that the MHC of radio-

resistant cells of the thymus – presumably thymic epithelial cells – selects the T cell

repertoire. This conclusion is based on a series of classical irradiation bone marrow

and thymus chimera experiments (reviewed in Moller, 1978; von Boehmer, 1990).

Several groups have shown that (H-2a x H-2b) F1-bone marrow cells reconstituting

lethally irradiated parental (H-2a)-mice generate H-2a-restricted but virtually no H-2b-

restricted virus-specific cytotoxic T lymphocytes (CTL) in a primary immune response

(Moller, 1978; von Boehmer, 1990). Although this view has since been accepted by

most immunologists and in textbooks, exceptions have been reported (Doherty and

Bennink, 1979; Longo and Schwartz, 1980; Matzinger and Mirkwood, 1978; Wagner

et al., 1980). However, examples of such non-thymic epithelial MHC-restricted T cells

have been rare and usually reflected weak responses that needed priming and

several rounds of in vitro restimulation before they could be measured. Surprisingly,

experiments with nude mice reconstituted with a completely allogeneic d14 fetal

thymus graft demonstrated that the T cell repertoire was almost exclusively specific

for the recipient MHC haplotype (Zinkernagel et al., 1980).

Therefore, to clarify the respective roles of the MHC of thymic epithelial versus non-

thymic epithelial cells in T cell repertoire selection, we have generated tetraparental

aggregation chimeras from T- and B-cell deficient mice (H-2d Scid or H-2b Rag-/-)

Results Part I

- 29 -

and thymus-deficient nude mice (H-2b or H-2d). In the resulting Scidd � nudeb

tetraparental mice, thymic epithelial cells are exclusively of Scid H-2d haplotype, and

T and B cells exclusively of nude H-2b haplotype. In the Rag-/-b ↔ nuded

tetraparental chimeras, thymic epithelial cells are exclusively of Rag-/- H-2b

haplotype, and T and B cells exclusively of nude H-2d haplotype. These relatively

undisturbed (i.e. non-irradiated and non-reconstituted) and well mixed tetraparental

adult chimeras allow a unique opportunity to study whether during a primary immune

response the T cell repertoire is restricted exclusively to the MHC of thymic epithelial

cells or to both parental haplotypes. Moreover, we analyzed whether these mice

could clear LCMV as efficiently as control mice and whether a protective CD4+ T

helper cell-dependent antibody response was present against LCMV or vesicular

stomatitis virus (VSV). Since B cells in these tetraparental mice express non-thymic

epithelial MHC, a protective CD4+ T helper cell-dependent antibody response would

be direct evidence for the presence of functional effector T cells restricted to non-

thymic epithelial MHC.

MATERIALS AND METHODS

Mice. C57BL/6 (H-2b), BALB/c (H-2d), C57BL/6-Rag1-/- or -Rag2-/- (Rag-/-b) were

obtained from the Institute of Laboratory Animal Science, University of Zurich,

Switzerland. C57BL/6-nude (nudeb) and BALB/c-nude (nuded) were purchased from

Biological Research Laboratories, Füllinsdorf, Switzerland. C.B-17-Scid (Scidd) were

purchased from IFFA CREDO, L'Arbresle, France. All mice were kept under specific

pathogen-free conditions.

Aggregation Chimeras. Mouse chimeras were generated by aggregation of 8-cell

embryos recovered from genetically immunodeficient mutants: Nuded and Scidd (both

albino, Gpi-1a, H-2d), nudeb and Rag-/-b (all black, Gpi-1b, H-2b). Embryos of each

strain were obtained by mating homozygous unreconstituted parents. Females were

super ovulated according to standard procedures with 5 iU PMSG/hCG and

additionally stimulated by the Whitten effect to improve their response. On day 3 of

gestation the embryos were flushed from the oviduct and their zona pellucida was

removed by brief incubation in pronase solution. After washing, embryos were

transferred in drops of M16 culture medium under liquid paraffin. Double embryo

aggregates of the following combinations were produced by gently pushing two

uncompacted morulae together: Scidd ↔ nudeb and Rag-/-b ↔ nuded. After 30 h of

Results Part I

- 30 -

incubation at 37°C and 5% CO2 in air, most aggregates formed early blastocysts.

Morphologically normal embryos were transferred into the uteri of pseudopregnant

histocompatible CB6F1 surrogate foster mothers under SPF conditions. Offspring

were born after 18 days of gestation and chimeras were recognized by the presence

of albino and pigmented skin patches a few days later. Additionally, several tissues

were tested for chimerism by GPI (glucose-6-phosphate isomerase)-isoenzyme gel

electrophoresis (Eppig et al., 1977). The contributions from both parental strains

were approximately equal, indicating that there was no strong strain-specific selective

advantage during embryonic development.

Cell Lines, ELISA, 51Cr-Release Assay, VSV-IND neutralization assay, and

Virus. EL-4 (H-2b) and P815 (H-2d) cells were obtained from the American Type

Culture Collection (Rockville, Maryland, USA). The LCMV nucleoprotein-specific

enzyme-linked immunosorbent assay (ELISA), the 51Cr-release assay, the VSV-IND

(VSV Indiana serotype) neutralization assay, LCMV-WE (LCMV strain WE) and VSV-

IND have been previously described (Bachmann et al., 1993; Battegay et al., 1993;

Charan and Zinkernagel, 1986; Kyburz et al., 1993; McCaren et al., 1959; Speiser et

al., 1992).

Immunohistology. Thymi were immersed in HBSS, snap-frozen, and 5-µm cryostat

sections were cut and fixed in acetone for 10 min. Sections were incubated with the

following antibodies: rat monoclonals against murine CD4 (YTS 191), CD8 (YTS

169), CD45R/B220 (RA3-6B2) and CD11b (M1/70), biotinylated mouse monoclonal

antibodies against murine MHC class I H-2Kb (AF6-88.5), H-2Kd (SF1-1.1), MHC

class II IAb (AF6-120.1) and IAd (AMS-32.1), followed by incubation with streptavidin-

alkaline phosphatase (all from Pharmingen). Primary antibodies were detected by

sequential incubation with goat antibodies against species-specific immunoglobulins,

followed by alkaline phosphatase labeled donkey anti-goat antibodies (Jackson

ImmunoResearch). Alkaline phosphatase was visualized using naphthol AS-BI (6-

bromo-2-hydroxy-3-naphtholic acid-2-methoxy anilide) phosphate and new fuchsin as

substrate, yielding a red color reaction product. Endogenous alkaline phosphatase

was blocked by levamisole. Sections were counterstained with hemalum.

Confocal Fluorescence Microscopy. Thymic epithelial cells were stained with a

polyclonal rabbit anti-cytokeratin antiserum (wide spectrum screening; dilution

1:1500; Dako). Primary rabbit antibodies were detected by sequential incubation with

affinity purified, rhodamine labeled goat anti-rabbit Ig antibodies followed by

Results Part I

- 31 -

rhodamine labeled donkey anti-goat Ig antibodies (dilutions 1:100 in TBS containing

5% normal mouse serum; Jackson ImmunoResearch). MHC class II antigens were

revealed with biotinylated mouse anti-IAb antibodies (clone AF6-120.1; dilution 1:200)

or biotinylated mouse anti-IAd antibodies (clone AMS-32.1; dilution 1:60; both from

Pharmingen) and fluorescein streptavidin (dilution 1:200; Dako). The appropriate

primary and secondary reagents were mixed and incubated in three steps of 30 min

each; anti-MHC class II antibodies were added twice. Slides were mounted with

Dako medium. Images were recorded with a confocal laser scanning system TCS-

SP2 (Leica laser technique, Mannheim, Germany) and processed using Openlab

software (Improvision).

Flow Cytometric Analysis. Peripheral blood or splenic cells were stained with the

following antibodies: anti-CD8alpha-APC (53-6.7), anti-CD8b.2-PE (53-5.8), anti-

B220-PE (RA3-6B2), anti-CD11b-PE (M1/70), anti-H-2Dd-FITC (34-2-12), anti-H-2Db-

PE (KH95), anti-H-2Kb-Biotin (AF6-88.5), Streptavidin-PerCP and Streptavidin-APC

(all from Pharmingen). For double tetramer stains, peripheral blood or splenic cells

(7.5x105) were stained with equal amounts of APC-labeled LCMV-WE GP33 tetramer

(GP33-Db) and PE-labeled LCMV-WE NP118 tetramer (NP118-Ld) and incubated for

20 min at 37°C. One microgram of anti-CD8alpha-FITC antibody (53-6.7) was added

to each sample and incubated for another 20 minutes at 4°C. All samples were

acquired on a FACSCalibur and analyzed using CellQuest software (Becton

Dickinson).

RESULTS

Tetraparental aggregation chimeras show a well-mixed chimerism.

Tetraparental aggregation chimeric mice (chimeras) of the following combinations

were generated: Scidd ↔ nudeb (Figure 5A) and Rag-/-b ↔ nuded. All chimeras had a

well-mixed chimerism, shown by the presence of both GPI-isoforms (GPI-1A, GPI-

1B) in sections of spleen, kidney, liver, lung, thymus and heart (Figure 5B and Figure

5C). Blood lymphocytes of Rag-/-b ↔ nuded chimeras were tested for H-2d and H-2b

expression by FACS-analysis. Whereas CD8+ and B220+ cells expressed only the

parental BALB/c-nude H-2d, CD11b+ macrophages were distributed into two

populations, expressing either the parental BALB/c-nude H-2d or the parental Rag-/-

H-2b (Figure 5D). Like CD8+ T cells, CD4+ T cells expressed only H-2d (data not

shown). Equivalent expression patterns were observed with Scidd ↔ nudeb chimeras

Results Part I

- 32 -

(data not shown). Non-mucosal CD8+ T cells in these chimeras were all CD8αβ as

shown by co-staining with CD8α- and CD8β-specific antibodies (data not shown).

GPI-1B

GPI-1A

GPI-1B

GPI-1A

H H HThSp LuLiKi

C57B

L/6

BA

LB

/c-n

ud

e

C57B

L/6

+

BA

LB

/c-n

ud

e

Scidd ↔ nudeb

Rag-/-b ↔ nuded

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

H H HHSp LuLiKi

B

C

DA

101

103

101

103

101

103

101 103101 103

101

103

101 103

(Rag-/-b ↔ nuded)- Chimera #8

(Rag-/-b ↔ nuded)- Chimera #7

BALB/c (H-2d)

C57BL/6 (H-2b)

Blood lymphocytes gated on

CD8+ cells B220+ cells CD11b+ cells

H-2Db H-2Kb

H-2

Dd

Figure 5: Phenotypic analysis of tetraparental chimeric mice. (A) Picture of a six-week-old Scidd ↔ nudeb

chimera. Distribution of GPI (Glucose-6-phosphate isomerase)-isoforms (GPI-1A, GPI-1B) in different tissues (H =

Heart, Sp = Spleen, Ki = Kidney, Li = Liver, Lu = Lung, Th = Thymus) of a Scidd ↔ nudeb chimera (B, lanes 3-7)

and a Rag-/-b ↔ nuded chimera (C, lanes 3-7). As controls, heart preparations of a C57BL/6 (GPI-1B) (B and C,

lane 1), BALB/c-nude (GPI-1A) (B and C, lane 2) and a 50:50-mixture of both (B and C, lane 8) were used. The

experiment was repeated three times with similar results. Blood-FACS analysis of six-to-eight-week-old chimeric

and control mice (D). Blood lymphocytes of naïve C57BL/6, BALB/c and Rag-/-b ↔ nuded chimeras #7 and #8

were gated on CD8+, B220+ or CD11b+ cells and stained with H-2Db-, H-2Kb- and H-2Dd- specific antibodies. The

experiment was repeated five times with similar results.

No rescue of nude thymic rudiments in tetraparental chimeras. Expression of

MHC class I and II, CD4, CD8, B220 and CD11b in chimeric and control thymi was

assessed by immunohistological analysis of frozen serial sections (Figure 6A, a-x).

Whereas analysis of MHC class I and II expression of the Rag-/- H-2b or Scid H-2d

haplotype (representing thymic epithelial haplotype) revealed a typical thymic

network (Figure 6A, i-j+q and Figure 6A, o-p+s), MHC class I and II expression

patterns of the nude haplotype revealed the presence of individual cells rather than

this thymic pattern (Figure 6A, k-l+r and Figure 6A, m-n+t). The presence of CD4+,

CD8+, B220+ and CD11b+ cells in Scidd ↔ nudeb chimeras was similar to control

mice (Figure 6A, u-x and data not shown). Some B220+ cells of C57BL/6-nude or

BALB/c-nude origin, respectively, were found in both the thymic medulla and in the

cortex (Figure 6A, w, arrows and data not shown).

In order to exclude eventual rescue of nude thymic rudiments (Blackburn et al., 1996;

Holub et al., 1975; Hsu et al., 1975; Kindred, 1979; Rodewald et al., 2001) and

Results Part I

- 33 -

confirm that thymic epithelial cells in these chimeras express exclusively the Rag-/-

or Scid- but not the nude haplotype, two-color thymus histology was performed

(Figure 6B, a-x). Chimeras and control mice were analyzed for expression of MHC

class II and cytokeratin (CK), the latter being a characteristic marker for epithelial

cells. Sections of chimeras and control mice revealed an intense yellow stain when

thymic epithelial MHC (Rag-/- or Scid-haplotype for chimeras) and cytokeratin stains

were superimposed, showing that both markers coincided on the thymic epithelial

network (Figure 6B, c+o and l+x). When nude MHC and cytokeratin stains were

superimposed, cytokeratin-negative, nude MHC class II-positive cells were found

(Figure 6B, r and u; green). Therefore, these cells must represent non-thymic

epithelial cells of haematopoietic nude origin, which have migrated into the thymus.

In summary, in all chimeras tested we found no evidence that nude thymic rudiments

were rescued as thymic epithelial cells were always of non-nude donor origin.

Results Part I

- 34 -

Figure 6: Immunohistological (A) and confocal immunofluorescence (B) analysis of the thymus from LCMV-

memory chimeras. (A) Frozen thymic sections of LCMV-memory C57BL/6, BALB/c, Rag-/-b ↔ nuded and Scidd ↔

nudeb were stained with monoclonal antibodies specific for MHC class I H-2Kb and H-2Kd and MHC class II IAb

and IAd (a-t). The thymus of (Scidd ↔ nudeb) chimera was additionally stained with monoclonal antibodies specific

for CD4, CD8, B220 and CD11b (u-x). Arrows in panel w indicate B220+ cells in the thymic cortex. The

experiment was repeated three times with similar results. (B) Two-color thymus histology from LCMV-memory

C57BL/6, BALB/c, Rag-/-b ↔ nuded and Scidd ↔ nudeb (a-x). Sections were stained with monoclonal antibodies

specific for MHC class II IAb (a, g, m, s; green) and cytokeratin (CK) (b, h, n, t; red) (overlay in c, i, o, u) or MHC

class II IAd (d, j, p, v; green) and CK (e, k, q, w; red) (overlay in f, l, r, x). Non-epithelial cells of haematopoietic

nude origin are CK-negative, MHC class II-positive (r, u; green). Cells double positive for MHC class II (green)

and CK (red) stain in yellow (c, l, o, x). The experiment was repeated three times with similar results.

Results Part I

- 35 -

Chimeras mount primary protective virus-specific CD8+ T cell responses

restricted to both thymic and non-thymic epithelial MHC. The effector function of

T cells of nude origin maturing in a thymic environment composed of epithelial cells

expressing non-matching MHC molecules was evaluated during an immune

response against LCMV. Chimeras aged 6-8 weeks showing comparable furry/nude

and pigmented/albino skin patches (Figure 5A) were infected intravenously with 200

plaque-forming units (pfu) of LCMV-WE. Eight days later, mice were

hemisplenectomized and cytotoxic CD8+ T cell activity was measured directly ex vivo

in a 5 hour 51Cr-release assay on H-2b (EL-4) and H-2d (P815) target cells prepulsed

with the immunodominant LCMV peptides LCMV-GP33-41 (H-2Db) and LCMV-NP118-

126 (H-2Ld) (Figure 7A) or the subdominant LCMV peptide LCMV-NP396-404 (H-2Db)

(data not shown). The chimeras exhibited strong primary CTL activity specific for all

three peptides tested. The CTL responses observed in chimeras were similar to

those of LCMV-infected C57BL/6 or BALB/c mice (Figure 7A). Chimeras were

efficiently protected against viral infection as indicated by the absence of detectable

virus in spleen and other organs eight days following infection (data not shown).

On day 36 after infection, lymph node cells from the hemisplenectomized mice were

restimulated in vitro for 5 days with peptide labeled H-2d or H-2b spleen cells (Fig.

3B). The strong CTL activity of chimeric lymph node cells was comparably restricted

to both thymic and non-thymic MHC (Figure 7B). As expected, no alloreactivity

against chimeric MHC was seen (Figure 7B). In contrast, in a standard mixed

lymphocyte culture assay, alloreactivity to third party H-2k was found for control and

chimeric effector cells (data not shown).

Results Part I

- 36 -

1 9 81

0

25

50

75

100

90 30 10 30

25

50

75

100

90 30 10 3

A

B

90 30 10 3 90 30 10 3

1 9 81 1 9 811 9 81

% S

pecif

ic L

ysis

% S

pecif

ic L

ysis

Effector : Target Ratio

Dilution of Standard Culture

Day 8 (directly ex vivo)

Day 36 (2° in vitro)

alloreactivity alloreactivity

Rag-/-b ↔ nuded

Scidd ↔ nudeb

BALB/c (H-2d)

C57BL/6 (H-2b)

EL-4 / ∅ EL-4 / GP33 P815 / ∅ P815 / NP118

EL-4 / ∅ EL-4 / GP33 P815 / ∅ P815 / NP118

(H-2b) (H-2

b) (H-2

d) (H-2

d)

(H-2b) (H-2

b) (H-2

d) (H-2

d)

naive C57BL/6

naive BALB/c

Figure 7: Primary ex vivo and secondary in vitro CTL response of LCMV-infected chimeras. Eight-week-old

C57BL/6 (�), BALB/c (�), Scidd ↔ nudeb (�) and Rag-/-b ↔ nuded (�) were infected intravenously with 200 pfu

of LCMV-WE. (A) On day 8 post infection, mice were hemisplenectomized and single cell suspensions were

tested directly ex vivo for 5h in a standard 51Cr-release assay on LCMV-GP33-loaded (EL-4 / GP33) or control

(EL-4 / ∅) EL-4 targets (H-2b) and on LCMV-NP118-loaded (P815 / NP118) or control (P815 / ∅) P815 targets

(H-2d). (B) On day 36 after infection, mice were sacrificed and pooled lymph node cells were restimulated in vitro

for 5 days with LCMV-GP33-loaded irradiated C57BL/6 splenocytes or LCMV-NP118-loaded irradiated BALB/c

splenocytes as stimulator cells. Cultures were tested for 5h in a standard 51Cr-release assay on LCMV-GP33-

loaded (EL-4 / GP33) or control (EL-4 / ∅) EL-4 targets and on LCMV-NP118-loaded (P815 / NP118) or control

(P815 / ∅) P815 targets. Similar results were obtained using restimulated splenocytes as effectors (data not

shown). Because of alloreactivity, C57BL/6 effectors lysed equally well peptide-loaded and control P815 targets.

The same was true for BALB/c effectors with peptide-loaded or control EL-4 targets. Therefore, results of

C57BL/6, or BALB/c effectors, with NP118-loaded P815 targets, or GP33-loaded EL-4 targets, are omitted for

clarity. The experiment was repeated six times with similar results.

To characterize the CD8+ T cell repertoire in more detail, day 8 and day 13 effector T

cells were also analyzed at the receptor level using tetramer staining (Figure 8 and

data not shown). Chimeras with strong primary CTL activity restricted to each of the

parental H-2 haplotypes always showed two distinct effector T cell populations, which

either bound LCMV-GP33 (H-2Db) tetramer or LCMV-NP118 (H-2Ld) tetramer, but

not both (Figure 8, 2nd-4th column and data not shown). While LCMV-NP118 (H-2Ld)

Results Part I

- 37 -

tetramer binding on day 8 was lower in chimeras than in BALB/c controls, by day 13,

LCMV-NP118 (H-2Ld) tetramer bindings were comparable (data not shown).

C57BL/6 (H-2b)

BALB/c (H-2d)

BALB/c-nude

(H-2d)

(Rag-/-b ↔ nuded)- Chimera #7

(Rag-/-b ↔ nuded)- Chimera #8

Gated on

CD8+ cells

NP

11

8-L

d

GP

33

-Db

Co

un

ts

NP

11

8-L

d

GP33-DbCD8

Gated on living splenocytes

Spleen (d8)

101

103

0.0

0.0

101

103

0.1

23

101

103

8.1

0.1

101

103

101 103

6.1

3.0

101

103 2.5

6.3

08

0

65.6

08

0

101 103

35.5

08

0

30.3

08

0

0.6

08

0

33.0

101

103 0.1

101

103 7.9

101

103 0.1

101

103

101 103

1.2

101

103 0.8

101

103 5.5

101

103 0.0

101

103 0.1

101

103 2.0

101

103

101 103

2.1

Figure 8: Tetramer analysis of LCMV-immune T cells of tetraparental chimeras. Eight-week-old C57BL/6, BALB/c,

BALB/c-nude and Rag-/-b ↔ nuded chimeras #7 and #8 were infected with 200 pfu of LCMV-WE intravenously.

Eight days after infection, mice were hemisplenectomized and 7.5x105 splenocytes were tested for binding to

LCMV-GP33 tetramer (GP33-Db) and LCMV-NP118 tetramer (NP118-Ld). Histogram plots show the percentage

of CD8+ cells amongst living splenocytes (1st column). Dot plots gated on living splenocytes show double staining

with a CD8-specific antibody and GP33-Db (2nd column) or double staining with a CD8-specific antibody and

NP118-Ld (3rd column). Numbers in upper right quadrants represent percentage of tetramer and CD8 double

positive splenocytes. Dot plots gated on CD8+ splenocytes show staining with equal amounts of GP33-Db and

NP118-Ld (4th column). Numbers in upper left (UL) and lower right (LR) quadrants represent percentage of CD8+

splenocytes binding to either NP118-Ld or to GP33-Db, respectively. The experiment was repeated three times

with similar results.

Non-thymic epithelial MHC-restricted CD4+ T cells co-operate efficiently with B

cells and mediate protective IgG responses. As B cells of tetraparental chimeras

express MHC class II of the non-thymic haplotype, co-operation is only possible with

CD4+ T cells being restricted to non-thymic MHC (Table 2). Therefore, to assess

whether CD4+ T cells restricted to non-thymic MHC are present and functional in

these chimeras, VSV- neutralizing IgG or LCMV-NP-specific IgG titers – which have

both been shown previously to be strictly dependent upon cognate MHC class II-

restricted CD4+ T helper cell activity (Leist et al., 1987; Ochsenbein et al., 2000) -

were monitored following VSV or LCMV infection, respectively. VSV-infected

chimeras generated protect ive neutral iz ing IgG responses (

Results Part I

- 38 -

Table 3) and LCMV-infected chimeras generated LCMV-NP-specific IgG responses

(Figure 9) comparable to control mice. These normal B cell responses confirm the

presence of functional CD4+ T cells restricted to non-thymic epithelial MHC.

Table 2: Productive T-B co-operation in tetraparental chimeras is only possible if CD4+ T cells express TCR

restricted to non-thymic MHC of B cells.

Scidd ↔ nudeb

Rag_/_b ↔ nuded

Tetraparental

ChimeraT / B cooperation

Thymic

epithelial

cells

H-2d

H-2b

T cells

B cells

H-2b

H-2d

H-2b-restricted TCR

Thb

TCR

Bb

Peptide

MHC II

H-2d-restricted TCR

Thb

TCR

Bb

Peptide

MHC II

H-2b-restricted TCR

ThThd

TCR

Bd

Peptide

MHC II

H-2d-restricted TCR

ThThd

TCR

Bd

Peptide

MHC II

Results Part I

- 39 -

Table 3: Non-thymic epithelial MHC-restricted CD4+ T cells mediate protective B cell IgG responses in

tetraparental chimeras following intravenous infection with 2x106 pfu VSV-IND.

Mouse*

VSV-IND neutralizing IgG titer** (serum dilution)

Rag-/- (H-2b)

Rag-/-b ↔ nuded

BALB/c (H-2d)

C57BL/6 (H-2b)

<1/40 <1/40

1/80'0001/80'000

1/40'0001/80'000

1/40'0001/40'000

Day 20Day 8

*Rag-/-b ↔ nuded tetraparental chimeras and control mice were infected intravenously with 2x106 pfu of VSV-IND.**VSV-IND-neutralizing IgG titers were monitored on day 8 and day 20 after infection. The highest serum dilution

that neutralizes 50% of input virus is expressed. At least four individual mice were tested in each group. Values

represent averages; variations were always less than one dilution step of two.

10

100

1000

10000

detection limit

An

ti-L

CM

V-N

P I

gG

[Re

cip

roc

al

of

Tit

er

Dil

uti

on

]

d1

4 C

57

BL

/6 (

H-2

b)

d1

4 B

AL

B/c

(H

-2d

)

d1

4 R

ag

-/-b

↔ n

ud

ed

d1

4 S

cid

d ↔

nu

de

b

na

ive

Scid

d ↔

nu

de

b

na

ive

C5

7B

L/6

(H

-2b

)

Figure 9: CD4+ T helper cells co-operate efficiently with B cells of non-thymic H-2 haplotype in LCMV-immune

chimeras. Eight-week-old C57BL/6, BALB/c, Scidd ↔ nudeb and Rag-/-b ↔ nuded were infected with 200 pfu of

LCMV-WE intravenously. On day 14 after infection, serum was prediluted 30-fold and tested for the presence of

LCMV-NP-specific IgG in a standard ELISA. Naïve Scidd ↔ nudeb and naïve C57BL/6 were used as negative

controls and signals were always below detection level. The same was true for non-infected Rag-/-b ↔ nuded-

chimeras or d14 LCMV-infected BALB/c-nude or for C.B-17-Scid or Rag-deficient mice (data not shown). The

experiment was repeated three times with similar results.

Results Part I

- 40 -

DISCUSSION

In summary, these chimeras showed protective virus-specific primary CD8+ T cell

responses restricted to both thymic and non-thymic MHC to comparable levels.

Virus-neutralizing IgG responses – strictly dependent on CD4+ T cell help restricted

to non-thymic epithelial MHC – were generated in these chimeras as efficiently as in

control mice. Taken together, these results demonstrate that cells other than thymic

epithelial cells are efficient in selecting a mature and functional T cell repertoire.

These findings challenge the widely accepted concept which postulates that MHC-

restriction is determined predominantly by the MHC of thymic epithelial cells (thymic

nurse cells) (Wekerle and Ketelsen, 1980) or the radio-resistant portion of the thymus

(reviewed in Moller, 1978; von Boehmer, 1990).

The present study was prompted by data obtained from experiments with nude mice

reconstituted with day 14 fetal thymus grafts from histoincompatible donors

(Zinkernagel and Althage, 1999; Zinkernagel et al., 1980). These nude thymus

chimeras only generated nude MHC- but not thymic MHC-restricted effector T cells.

The conclusion from these earlier studies was either that there was rescue of the

nude thymic rudiment or that cells other than thymic epithelia were efficient in and

essential for positive selection of MHC-restricted T cell specificities. More recently,

tetraparental chimeras between thymus-competent and nude donors of distinct MHC

haplotypes revealed that the thymic rudiment of the nude donor could not be rescued

anatomically in a tetraparental chimeric situation (Blackburn et al., 1996; Rodewald et

al., 2001), as we confirm here (Figure 6). In addition, histological analysis of 20

chimeric thymi showed no evidence of thymic epithelial cells, cysts or other

rudiments of nude origin in well-mixed chimeras.

The discrepancies between the present results and those obtained with F1(AxB) →

A-irradiation bone marrow chimeras and/or F1(AxB) nude grafted with an A-thymus

yielding virtually exclusively A-restricted but not B-restricted T cell responses are

particularly important (Matzinger, 1993; Moller, 1978; Singer, 1988). We believe they

are best explained as follows: it is possible that lethal or supralethal irradiation is not

capable of eliminating all host cells that contribute to T cell receptor interactions

resulting in survival of such T cells regardless of whether they are encountered only

in the thymus or also in the periphery. For example, radio-resistant follicular dendritic

cells in the spleen and lymph nodes, fibroblasts, or other mesenchymal cells would

fulfill such requirements. Also lymphohaemopoietically derived cells, including

Results Part I

- 41 -

macrophages and dendritic cells can probably not be eliminated completely and

current detection limits cannot exclude the presence of 0.5-2% of “contaminating”

cells. Therefore, precursor T cells migrating into the thymus of such F1(AxB) → A-

irradiation bone marrow chimeras will predominantly see A-expressing cells and

therefore will first and preferentially be restricted to A. Those A-restricted T cells will

strongly proliferate in the thymus and leave the thymus as A-restricted T cells. B-

expressing cells from F1(AxB) bone marrow first have to migrate into the thymus

leading to a time disadvantage compared to A-expressing cells already present in the

A-recipient. As the proliferation rate in the thymus is enormous, A-restricted T cells

will have a huge advantage over B-restricted T cells. For example, the numerical

advantage of A-restricted T cells in an A-thymus of an F1(AxB) → A-irradiation

chimera or in an F1(AxB) nude grafted with an A-thymus may readily reach factors of

10-30 (even after a subsequent second depletion of T cells) within 3-5 cell divisions

(corresponding to only 1-3 days!) (Lemischka et al., 1986; Longo and Schwartz,

1980; Zinkernagel, 1982).

In the case of nude mice reconstituted with a fully allogeneic fetal thymus, precursor

T cells migrating into the thymus predominantly encounter thymic epithelial MHC

(Zinkernagel et al., 1980). Only those T cells seeing non-thymic epithelial MHC on

cells from nude origin having migrated into the thymus will be restricted to non-thymic

MHC but will be at a numerical disadvantage compared to the former subset.

However, as soon as these hypothetical T cells would leave the thymus they will

contact cells expressing exclusively non-thymic MHC. Therefore, only T cells

restricted to non-thymic MHC will survive in these chimeras since it is now well

accepted that peripheral amplification and survival of mature T cells are strongly

MHC-dependent (Kirberg et al., 1997; Rocha and von Boehmer, 1991).

The advantage of the chimeras described here over F1(AxB) → A-irradiation bone

marrow chimeras or F1(AxB) nude grafted with an A-thymus is that cells expressing

either thymic epithelial or non-thymic epithelial MHC are present in the thymus and

the periphery in roughly equal numbers from the beginning, i.e. from the time-point

when precursor T cells migrate into the thymus. Therefore, selection and/or MHC-

dependent amplification and survival of a mature T cell repertoire restricted to either

thymic epithelial or non-thymic epithelial MHC is equivalent.

In conclusion, selection of a mature and functional T cell repertoire involves multiple

stages: Thymocyte maturation, followed by positive selection leading to self-MHC-

Results Part I

- 42 -

restricted T cells, and finally the continuous interaction between MHC-restricted

mature T cells and cells expressing the T cell restricting MHC molecules ensuring

their amplification and survival. Whereas thymic epithelium is crucial for T cell

maturation, the results presented here show that the MHC of thymic epithelial cells is

not required for the two latter stages of T cell repertoire selection.

We thank E. Horvath, K. Rappold, K. Osei-Boadum and N. Wey for technical help,

and Drs. A. Macpherson, K. McCoy and M. van den Broek for helpful comments on

the manuscript. This work was supported by the Swiss National Foundation for

Science and the Kanton of Zurich, Switzerland.

Results Part II

- 43 -

6 Results Part II

Influence of MHC class I H-2Db density on selection and survival of

H-Y-specific TCR transgenic T cells

Marianne M. Martinic, Hans Hengartner and Rolf M. Zinkernagel

Institute of Experimental Immunology, Department of Pathology, University Hospital

Zurich, Schmelzbergstrasse 12, CH-8091 Zurich, Switzerland

Results Part II

- 44 -

Abstract

H-Y-specific T cell receptor (TCR) transgenic mice express a transgenic TCR

specific for a male antigen-derived peptide (H-Y) presented on MHC class I H-

2Db molecules. The influence of different parameters on overall positive

selection of these transgenic T cells was analysed. Selection was

independent of age. However, positive selection of H-Y-specific TCR

transgenic T cells in females and escape from negative selection in males

were both strongly dependent on the density of H-2Db molecules on antigen

presenting cells. The higher the density of H-2Db molecules, the higher the

percentage of peripheral CD8+ T cells expressing transgenic TCR in females

and the stronger the down-regulation of the CD8 co-receptor in males. Thus

minor changes in the density of TCR-restricting H-2Db influence selection of

the H-Y-specific TCR.

Introduction

The minor histocompatibility (H) male specific (Y) antigen (H-Y) (KCSRNRQYL),

which is expressed in male tissues, is derived from a protein encoded by the SMCY

gene located on the Y chromosome (Simpson et al., 1997). The H-Y-specific

transgenic (tg) T cell receptor (TCR) binds specifically to the male antigen-derived

peptide (H-Y) presented on MHC class I molecules H-2Db (Kisielow et al., 1988a).

In males, presence of the self-peptide H-Y together with TCR-restricting MHC class

I molecules H-2Db leads to negative selection in the thymus of H-Y-specific CD8αβhi

tg TCRαβ+ cells (Table 4 and Figure 10, first row) (Kisielow et al., 1988a). CD4-

CD8- double negative (DN) tg TCRαβ+ cells which were shown to be aberrant γδ-

lineage T cells, and cells showing a down-regulation of their co-receptor and

transgenic TCR, resulting in CD8αβlo tg TCRαβ+ cells, are also present in TCR

transgenic males (Table 4 and Figure 10, first row and data not shown) (Bruno et

al., 1996). These DN tg TCRαβ+ and CD8αβlo tg TCRαβ+ cells are not measurably

H-Y-reactive anymore and are not negatively selected (Bruno et al., 1996). CD8αβlo

tg TCRαβ+ cells where binding to self-peptide/MHC complex probably is of low

avidity apparently have escaped negative selection in the thymus. They migrate into

the periphery as mature self-MHC-restricted, H-Y non-reactive T cells, do not

proliferate after contact with male stimulator cells but are functional against non H-Y

Results Part II

- 45 -

antigens (Kisielow et al., 1988a; Teh et al., 1989). In females, absence of the

peptide H-Y together with the presence of TCR-restricting MHC class I molecules

H-2Db leads to positive selection in the thymus of H-Y-specific CD8αβhi tg TCRαβ+

cells (Table 4 and Figure 10, second row) (Kisielow et al., 1988a). However, only

maximally 40% of peripheral CD8+ cells express the actual H-Y-specific transgenic

TCR (Figure 10, second row). The remaining CD8+ population have recombined the

endogenous TCRα loci and therefore have changed TCR specificity during

thymocyte selection (Figure 10, second row and data not shown) (Buch et al., 2002;

Huesmann et al., 1991; Merkenschlager et al., 1994). Because the transgenic

TCRαβ is already expressed at the DN stage during thymocyte maturation, few DN

tg TCRαβ+ cells are found also in the periphery of TCR transgenic females

representing the aberrant γδ-lineage cells (data not shown) (Bruno et al., 1996).

CD8αβlo tg TCRαβ+ cells as seen in TCR transgenic males are not found in TCR

transgenic females (Figure 10, second row). H-Y-specific TCR transgenic mice,

which do not express TCR-restricting MHC class I molecules H-2Db do not show

positive selection of H-Y-specific CD8αβhi tg TCRαβ+ cells (Table 4 and Figure 10,

third row) (Fink and McMahan, 2000; Teh et al., 1988; von Boehmer, 1990). These

cells, however, can escape death from neglect via endogenous rearrangement of

TCRα-chain genes (Table 4 and Figure 10, third row and data not shown) (Fink and

McMahan, 2000; Teh et al., 1988; von Boehmer, 1990). Peripheral DN tg TCRαβ+

and CD8αβlo tg TCRαβ+ cells are absent in H-Y-specific TCR transgenic mice with a

non-selecting H-2 haplotype (Figure 10, third row and data not shown). It also has

been shown that neither the presence of MHC class II molecules of the selecting H-

2b haplotype nor MHC class I molecules of the H-2d or H-2k haplotype were

sufficient to rescue H-Y-specific transgenic T cells from programmed cell death

(Table 4 and Figure 10 and data not shown) (Kisielow et al., 1988b; Teh et al.,

1988; von Boehmer, 1990). In summary, positive selection of H-Y-specific TCR

transgenic T cells seemed to be exclusively dependent on the presence of TCR-

restricting MHC class I molecules H-2Db and escape from negative selection was

modulated via CD8αβ co-receptor down-regulation. We analysed here the influence

of different H-2Db densities on selection of these TCR transgenic T cells and their

survival in the periphery.

Results Part II

- 46 -

Table 4: Selection of H-Y-specific TCR transgenic T cells.

Mouse T cell selection

Positive Selection

Phenotype of selected cells

CD8αβhi tgTCRαβ+ cells (Fig.10)

H-Y/H-2Db-

spec. responses

YES

Negative Selection No CD8αβhi tgTCRαβ+ cells (Fig.10)

Escape from Neg. Selection

CD4-CD8- tgTCRαβ+ cells (not shown)

NOCD8αβlo tgTCRαβ+ cells (Fig.10)

Gender

F2 (H-Y)

H-2dd

F0 (H-Y)

H-2bb

F0 (H-Y)

H-2bb

No CD8αβhi tgTCRαβ+ cells (Fig.10)Neglect

NOEscape from Neglect

(Fig.10 and not shown)

CD8αβhi TCRαend.+βtg

+ cells

Gated on live blood lymphocytes Gated on CD8αβ+ cells

50

101

103

80

101

103 91.2%

A)

H-2bb

F0 (H-Y)

73701

50

29

Co

un

ts

101

103 27.0%

Vβ8

.1/2

CD

8β

101

103

Co

un

ts80

485B)

H-2bb

F0 (H-Y)

CD8α CD8αCD8β T3.70 T3.70Vβ8.1/2101 103

80

050

101 103

-

101 103

772

101

103

101 103

0.6%

101

103

101 103

080

101 103

492C)

H-2dd

F2 (H-Y)

CD8lo CD8lo

CD8hi

CD8hiCD8hi

CD8hi

128MFI

61MFI

28MFI

150MFI Escape from Neg. Selection:

CD8αβlo tgTCRαβ+ (UR)

Positive Selection:

CD8αβhi tgTCRαβ+ (UR)

Escape from Neglect:

CD8αβhi TCRαend.+βtg

+ (UL)

Figure 10: Selection of H-Y-specific transgenic TCR in mice with selecting (H-2bb) or non-selecting (H-2dd) H-2

haplotype. Blood lymphocytes of H-Y-specific TCR transgenic males (F0 (H-Y), H-2bb, �) (A) and females (F0 (H-

Y), H-2bb, �) (B) with selecting (H-2 bb) or non-selecting (H-2dd) H-2 haplotype (F2 (H-Y), H-2dd, � �) (C) were

stained for surface expression of CD8α- (column 1) and CD8β-chain (column 2). Numbers in histogram plots

represent mean fluorescence intensity (MFI) of CD8α and CD8β, respectively. Circles in dot plots (column 3)

represent CD8αβlo and CD8αβhi populations. Percentage of CD8αα+ cells was always below 0.5%. Blood

lymphocytes of H-Y-specific TCR transgenic mice were gated on CD8αβ+ cells (CD8αβlo for H-Y-specific TCR

transgenic males, CD8αβhi for H-Y-specific TCR transgenic females and H-Y-specific TCR transgenic mice with a

non-selecting H-2 haplotype) and stained for surface expression of transgenic α- (T3.70) (column 4) and

transgenic β-chain (Vβ8.1/2) (column 5) of H-Y-specific TCR. Numbers in histogram plots represent MFI of

transgenic TCRα and transgenic TCRβ-chain, respectively. Numbers in upper right (UR) quadrants represent

percentage of CD8αβ+ cells expressing transgenic TCRαβ (column 6). Similar results were obtained with splenic

lymphocytes (data not shown). At least three individual mice were tested in each group. One out of twelve similar

experiments is shown.

Results Part II

- 47 -

Results

Selection of H-Y-specific transgenic TCR is independent of age.

The percentage of CD8+ cells expressing the H-Y-specific transgenic TCR in

females with the selecting H-2 haplotype (H-2bb) varied between 15-40%

independent of the age of these transgenic mice (Figure 10 and Figure 11). The

variability was probably due to the fact that these mice were housed under

conventional conditions and that they were not on a pure C57BL/6 but on a mixed

C57BL/6xSv129 background (Kisielow et al., 1988a). The percentage of CD8+ cells

expressing transgenic TCRαβ in females on a pure C57BL/6 background housed

under SPF-conditions was 25%±1% independent of age (Benedita Rocha, personal

communication). In H-Y-specific TCR transgenic males with the selecting H-2

haplotype (H-2bb), the percentage of CD8αβlo tg TCRαβ+ cells was always above

90% confirming earlier results (Figure 10 and Figure 11) (Kisielow et al., 1988a).

T3.70

Vβ8

.1/2

F0 (H-Y)

1 month

F0 (H-Y)

10 months

F0 (H-Y)

5 months

Gated on

CD8αβhi cells

Gated on

CD8αβlo cells

101

103 26%

101

103 97%

101

103 99%

101

103 17%

101

103

101 103

101

103

101 103

97%34%

Figure 11: Selection of H-Y-specific transgenic TCR is independent of age. Blood lymphocytes of H-Y-specific

TCR transgenic mice (F0 (H-Y)) of different ages were gated on CD8αβ+ cells and stained for surface expression

of transgenic TCRαβ (T3.70+Vβ8.1/2+). Numbers in UR quadrants represent percentage of CD8αβ+ cells

expressing transgenic TCRαβ. Similar results were obtained with splenic lymphocytes (data not shown). At least

three individual mice were tested in each group. One out of eight similar experiments is shown.

Results Part II

- 48 -

The higher the H-2Db density, the higher the percentage of CD8+ cells

expressing H-Y-specific transgenic TCR in females.

To analyse percentage of CD8+ cells expressing transgenic TCRαβ in mice with half

H-2Db density, we crossed H-Y-specific TCR transgenic mice (F0 (H-Y), H-2bb) with

either BALB/c (H-2dd), CBA (H-2kk) or C57B10.G (H-2qq) mice obtaining TCR

transgenic mice with a heterozygous H-2 haplotype (F1 (H-Y), H-2bd, H-2bk or H-2bq,

respectively) (Figure 12 and Figure 13). Additionally we crossed F1 (H-Y), H-2bd

and F1 (H-Y), H-2bk mice with BALB/c or CBA mice, respectively, to obtain F2 mice

with a non-selecting H-2 haplotype (F2 (H-Y), H-2dd or H-2kk) (Figure 12 and Figure

13 and data not shown). All F1 (H-Y) transgenic females exhibited a strong

reduction in their percentage of peripheral CD8+ cells expressing transgenic TCRαβ

resulting in only 5-10% (Figure 12). The remaining CD8+ population still expressed

the transgenic TCRβ-chain but paired with endogenous TCRα-chains (Figure 12

and data not shown). The drastic decrease in percentage of CD8+ cells expressing

transgenic TCRαβ was mainly observed in the periphery because analysis of F1 (H-

Y) versus F0 (H-Y) CD8-single positive (SP) thymocytes revealed only a slight

decrease in the percentage of CD8-SP thymocytes expressing transgenic TCRαβ

(80% CD8-SP thymocytes in F0 (H-Y) versus 60% in F1 (H-Y) TCR transgenic

females expressed transgenic TCRαβ, data not shown). A possible explanation for

the latter observation could be the excellent three-dimensional thymic

microenvironment offering cytokines and chemokines, which could overcome the

lack of sufficient H-2Db density for efficient positive selection of transgenic

thymocytes. In the periphery, however, these conditions are not available optimally

for further survival and expansion of all positively selected transgenic T cells

(Kirberg et al., 1997; Rocha and von Boehmer, 1991). The reduced H-2Db density in

the periphery of F1 (H-Y) transgenic mice seemed not sufficient for survival of all

positively selected T cells and therefore only about one fourth to half of the once

positively selected T cells survived and expanded in the periphery (Figure 12 and

Benedita Rocha, personal communication). CD8+ cells from F2 (H-Y) transgenic

females with a non-selecting H-2 haplotype (H-2dd or H-2kk) did not show positive

selection of H-Y-specific transgenic TCR but expressed high levels of the

transgenic TCRβ-chain paired with endogenous TCRα-chains confirming previous

observations (Figure 12 and data not shown) (Fink and McMahan, 2000; Teh et al.,

Results Part II

- 49 -

1988; von Boehmer, 1990). In summary, in H-Y-specific TCR transgenic females on

a mixed C57BL/6xSv129 background, kept under conventional conditions, a

reduction in H-2Db density by 50% (F1 (H-Y) females) resulted in a 2 to 4-fold

decrease in percentage of peripheral CD8+ cells expressing transgenic TCRαβ. In

H-Y-specific TCR transgenic females on a pure C57BL/6-background, kept under

SPF-conditions, reduction in H-2Db density by 50% resulted in a 50% decrease in

percentage of peripheral CD8+ cells expressing transgenic TCRαβ and almost no

decrease in percentage of thymic CD8+ cells expressing transgenic TCRαβ

(Benedita Rocha, personal communication). Overall, the efficiency of positive

selection and survival in the periphery for a low-affinity TCR such as the H-Y-

specific TCR seemed therefore to depend on high TCR-restricting H-2Db densities

and high CD8αβ-expression levels (Figure 12) (Cruz et al., 1998; Podd et al., 2001;

Zerrahn et al., 1999). The latter results were already observed by Cruz et al.

showing that only CD8αβ+ cells expressing transgenic TCRαβ were positively

selected in females whereas CD8αα+ cells expressing transgenic TCRαβ were of

too low overall avidity to get positively selected (Cruz et al., 1998). The observation

that H-Y-specific TCR transgenic females with a mixed C57BL/6xSv129

background, housed under conventional conditions, showed a different percentage

of CD8+ cells expressing transgenic TCRαβ than TCR transgenic females with a

pure C57BL/6 background, housed under SPF conditions, showed that not only H-

2Db density but also other parameters such as the background of these mice and

chronic infections could directly influence positive selection and peripheral survival

of TCR transgenic T cells.

Results Part II

- 50 -

H-2DbGated on

CD8αβhi cells

T3.70

Vβ8

.1/2

Fluorescence Intensity

Co

un

ts200

200

52

0200

101 103

200

56

200

47

F0 (H-Y)

H-2bb

F2 (H-Y)

H-2dd

F1 (H-Y)

H-2bq

F1 (H-Y)

H-2bk

F1 (H-Y)

H-2bd

0-1 %

15-40 %

5-10 %

101

103

101

103

101

103

101

103

101

103

101 103

0.3%

9%

7%

9%

26%

139

MFI

Figure 12: In females, selection of H-Y-specific transgenic TCR is dependent on the density of MHC class I H-2Db

molecules. Blood lymphocytes of H-Y-specific TCR transgenic females (F0, F1, F2 (H-Y)) with different H-2

haplotypes (H-2bb, H-2bd, H-2bk, H-2bq, H-2dd) were stained for surface expression of MHC class I molecules H-2Db

(numbers in histogram plots represent MFI of H-2Db). Blood lymphocytes of transgenic females were additionally

gated on CD8αβhi cells and stained for surface expression of transgenic TCRαβ (T3.70+Vβ8.1/2+). Numbers in UR

quadrants represent percentage of CD8αβhi cells expressing transgenic TCRαβ. Numbers beside dot plots

represent averages of percentages of CD8αβhi cells expressing transgenic TCRαβ in females with the selecting

(H-2bb), with a mixed (H-2bd, H-2bk, H-2bq) or with a non-selecting H-2 haplotype (H-2dd). At least 6 individual mice

were tested in each group. One out of three similar experiments is shown.

The higher the H-2Db density, the stronger the down-regulation of CD8ααααββββ-co-

receptor in H-Y-specific TCR transgenic males.

Analysis of the percentage of CD8αβlo tg TCRαβ+ cells in H-Y-specific TCR

transgenic males with an exclusively selecting (F0 (H-Y), H-2bb) or heterozygous

(F1 (H-Y), H-2bd or H-2bk or H-2bq) H-2 haplotype revealed no difference in the

thymus or in the periphery (Fig. 4 and data not shown). As already observed in H-Y-

specific TCR transgenic females with a non-selecting H-2 haplotype (F2 (H-Y), H-

2dd or H-2kk), male counterparts expressed exclusively transgenic TCRβ-chains

paired with endogenous TCRα-chains (Figure 10 and data not shown) (Fink and

McMahan, 2000; Teh et al., 1988; von Boehmer, 1990). Interestingly, although no

Results Part II

- 51 -

change could be observed in the percentage of CD8αβlo tg TCRαβ+ cells in F0 (H-

Y) versus F1 (H-Y) transgenic males, a drastic difference was seen in CD8α- and

CD8β-chain expression levels on both thymic and peripheral CD8+ cells (Figure 13

and data not shown). CD8+ cells from F0 (H-Y) transgenic males showed the

strongest down-regulation for both CD8α- and CD8β-chains (Figure 13). CD8+ cells

from F1 (H-Y) transgenic males had less down-regulated CD8α- and CD8β-chains

(Figure 13) and CD8+ cells from F2 (H-Y) transgenic males exhibited the same

CD8αβ-expression levels as their female counterparts (Figure 10 and Figure 13).

Taken together, these results showed that TCR transgenic T cells in males actively

modulated their CD8αβ-expression levels during development in order to escape

clonal deletion. High H-2Db density in H-Y-specific TCR transgenic males (F0 (H-Y),

H-2bb) resulted in a too high overall avidity interaction between transgenic TCR and

self-peptide/self-MHC (H-Y/H-2Db) complexes on APC leading to negative selection

of H-Y-specific CD8αβhi tg TCRαβ+ cells (Table 4 and Figure 10) (Kisielow et al.,

1988a). Through strong down-regulation of both CD8α- and CD8β-chains, the

overall avidity was decreased and therefore CD8αβlo tg TCRαβ+ cells could escape

negative selection and migrate into the periphery (Table 4 and Figure 10 and Figure

13) (von Boehmer, 1990). Although decreased H-2Db density in TCR transgenic

mice with a heterozygous H-2 haplotype (F1 (H-Y), H-2bd, H-2bk or H-2bq) lead to a

lower overall avidity interaction between transgenic TCR and H-Y/H-2Db complexes,

clonal deletion of all H-Y-specific CD8αβhi tg TCRαβ+ cells was still observed

(Figure 13 and data not shown). Escape from negative selection in these mice,

however, was already achieved by a less strong down-regulation of both CD8α- and

CD8β-chains than the one observed in F0 (H-Y) transgenic males (Figure 13).

This study shows that the density of TCR-restricting MHC class I molecules H-2Db

is an important parameter in determining the fate of H-Y-specific TCR transgenic T

cells. In females, positive selection and survival of peripheral H-Y-specific CD8αβhi

tg TCRαβ+ cells was most efficient at high H-2Db density (Figure 12). In males,

negative selection of H-Y-specific CD8αβhi tg TCRαβ+ cells was independent of H-

2Db density (Figure 10 and Figure 13). Escape from negative selection, however,

was dependent on H-2Db density; the higher the H-2Db density, the stronger the

down-regulation of CD8αβ co-receptor (Figure 13).

Results Part II

- 52 -

H-2Db CD8β

Fluorescence Intensity

Co

un

ts

F0 (H-Y)

H-2bb

F2 (H-Y)

H-2dd

F1 (H-Y)

H-2bq

F1 (H-Y)

H-2bk

F1 (H-Y)

H-2bd

> 90 %

0-1 %

> 90 %

CD8α

200

0200

101 103

200

53

200

44

200

47

080

101 103

491

80248

80

270

80

257

80

101 103

773

322

310

363

141MFI

151MFI

123MFI

Figure 13: Expression levels of CD8αβ co-receptor in H-Y-specific TCR transgenic males are dependent on the

density of MHC class I H-2Db molecules. Blood lymphocytes of H-Y-specific TCR transgenic males (F0, F1, F2

(H-Y)) with different H-2 haplotypes (H-2bb, H-2bd, H-2bk, H-2bq, H-2dd) were stained for surface expression of H-

2Db, CD8α - and CD8β-chains (numbers in histogram plots represent MFI of H-2Db, CD8α and CD8β ,

respectively). Numbers beside dot plots represent averages of percentages of CD8αβlo cells expressing

transgenic TCRαβ in males with the selecting (H-2bb), a heterozygous (H-2bd, H-2bk, H-2bq) or with a non-selecting

H-2 haplotype (H-2dd). At least 6 individual mice were tested in each group. One out of three similar experiments

is shown.

Discussion

The thymus is the organ where thymocyte maturation takes place. During this

maturation phase, double positive thymocytes (CD4+CD8+ TCRαβlo) undergo a

selection process by interacting with APC presenting self-peptide/MHC complexes

(reviewed in Starr et al., 2002). The outcome of this selection process is determined

by the overall avidity of the APC-thymocyte interaction. The overall avidity is the sum

of the intrinsic affinity of the TCR and its co-receptor for a self-peptide/MHC complex,

surface density of TCR and co-receptors on the thymocyte and of self-peptide/MHC

complexes on APC, surface density of certain adhesion molecules on both thymocyte

and APC, peptide concentrations, and the duration of the interaction between TCR

and self-peptide/MHC complexes (Ashton-Rickardt et al., 1994; Sebzda et al., 1999).

Results Part II

- 53 -

An overall weak to intermediate avidity interaction transmits a survival signal to the

maturing thymocyte (positive selection) whereas an overall strong avidity interaction

transmits a signal to the maturing thymocyte to undergo programmed cell death

(negative selection) (reviewed in Starr et al., 2002). This selection process ensures

therefore only survival of self-MHC-restricted (useful) and self-tolerant T cells, which

leave the thymus as mature T cells (reviewed in Ohashi, 2002; Starr et al., 2002).

Survival and further expansion of each T cell in the periphery, however, is only

achieved if the T cell remains in continuous interaction between its TCR and TCR-

restring MHC molecules on APC (Kirberg et al., 1997; Rocha and von Boehmer,

1991).

In this study we showed that positive selection and subsequent peripheral survival of

the low-affinity H-Y-specific transgenic TCR was less efficient with reduced TCR-

restricting MHC density (Figure 12 and Benedita Rocha, personal communication). In

contrast, this is not the case for high-affinity TCR such as the 318 or 2C transgenic

TCR (Pircher et al., 1989; Sha et al., 1988). In 318 TCR transgenic mice with

exclusively selecting (H-2bb) or heterozygous H-2 haplotype (H-2bd, H-2bk or H-2bq),

the percentage of peripheral CD8+ cells expressing transgenic TCR always remains

above 50% independent of reduced TCR-restricting H-2Db density (M. Martinic,

unpublished observation). Moreover, Zerrahn et al. show that only 2C TCR but not H-

Y-specific TCR transgenic cells are positively selected and survive in the periphery

when exclusively haematopoietic cells express TCR-restricting MHC molecules

(Zerrahn et al., 1999). Further, under sub-optimal conditions as given in Oncostatin M

(OM) transgenic mice, which show a profound thymic atrophy but massive

extrathymic T cell development (Boileau et al., 2000; Clegg et al., 1996), positive

selection and survival of the H-Y-specific but not of the 2C transgenic TCR is

reduced (Terra et al., 2002).

Taken together, these results show that as soon as environmental conditions are

sub-optimal – reduced TCR-restricting MHC density, different microenvironment,

different cytokine milieu – efficiency of positive selection and survival of low-affinity

but not of high-affinity TCR is decreased. A possible explanation could be that even

under sub-optimal conditions the intrinsic affinity of the high-affinity TCR but not of

the low-affinity TCR for self-peptide/MHC complexes is still high enough to provide

sufficient survival signal.

Results Part II

- 54 -

Negative selection of self-reactive T cells, however, is always complete. Even under

sub-optimal conditions, as already shown in H-Y/LckOM and 2C TCR/LckOM double

transgenic mice, all self-reactive T cells are absent (Terra et al., 2002). We confirmed

that observation, showing absence of all H-Y-specific cells (CD8αβhi tg TCRαβ+) in

TCR transgenic males with normal or with reduced H-2Db density (Figure 13 and

Benedita Rocha, personal communication). Additionally, even in a complete athymic

environment as is the case for nude H-Y-specific TCR transgenic males, we do not

detect any self-reactive T cells (Martinic et al., Manuscript in preparation-b).

Finally, escape from negative selection is probably only possible with low-affinity

TCR as seen with the H-Y-specific transgenic TCR (Figure 10 and Figure 13). There,

already a small modulation of the co-receptor was sufficient to fall under the overall

avidity threshold for negative selection. High-affinity TCR, in contrast, have under

physiological conditions most probably no chance to escape negative selection

because even by modulating co-receptors or other molecules on their cell-surface,

the intrinsic affinity of the high-affinity TCR itself remains too high.

In summary, selection and survival of a low-affinity TCR is most efficient under

optimal conditions, for example like in this study when restricting MHC antigens are

optimally present both in the thymus for initial positive selection and in the periphery

to ensure further survival. Furthermore, it is crucial that even under sub-optimal

conditions, negative selection remains tight to avoid potential auto-reactivity.

Results Part II

- 55 -

Materials and Methods

Mice

H-Y-specific TCR transgenic mice (F0 (H-Y), H-2bb) were a generous gift from Jörg

Kirberg. C57BL/6 (H-2bb), BALB/c (H-2dd), CBA (H-2kk) and C57B10.G (H-2qq) were

obtained from the Institute of Laboratory Animal Science, University of Zurich,

Switzerland. H-Y-specific TCR transgenic mice were bred with BALB/c, CBA or

C57B10.G mice yielding H-Y-specific TCR transgenic mice (F1 (H-Y)) with a

heterozygous H-2 haplotype (H-2bd, H-2bk or H-2b q, respectively). To obtain

transgenic mice with a non-selecting H-2 haplotype (H-2dd), H-Y-specific TCR

transgenic mice were bred twice with BALB/c (F2 (H-Y)) and tested with PCR and

FACS analysis for expression of transgenic TCR, presence of H-2dd and absence of

H-2bb MHC haplotype. All mice were kept under conventional conditions.

PCR

DNA was prepared from mouse-tail. 2 µl of mouse-tail-DNA were used for PCR

analysis. Primers used for amplification were specific for the H-Y-specific transgenic

TCR (Vβ8.2-fwd (5’ ACA AGG TGG CAG TAA CAG GA 3’) and Jβ2.3-rev (5’ ACA

GTC AGT CTG GTT CCT GA 3’) primers), for both H-2bb and H-2dd MHC haplotypes

(Ea5’-fwd (5’ AGT CTT CCC AGC CTT CAC ACT CAG AGG TAC 3’) and Ea3’-rev

(5’ CAT AGC CCC AAA TGT CTG ACC TCT GGA GAG 3’) primers) and for the H-

2dd MHC haplotype (K5’-fwd (5’ CAT GGG CAT AGA AAG GGC AGT CTT TGA ACT

3’) and Ea3’-rev primers). Expected bands were 302 bp for presence of H-Y-specific

transgenic TCR, 155 bp and 700 bp for presence of H-2b and H-2d MHC haplotype,

respectively, and 210 bp for presence of H-2d MHC haplotype.

Flow Cytometric Analysis

Peripheral blood cells were stained with the following antibodies: anti-CD8α-APC

(53-6.7), anti-CD8β.2-PE (53-5.8), T3.70-Biotin (specific for transgenic TCRα-chain

of H-Y-specific TCR) (generous gift from Benedita Rocha), anti-Vβ8.1/2-FITC

(specific for transgenic TCRβ-chain of H-Y-specific TCR) (MR5-2) and anti-H-2Db-

Biotin (KH95). Except for T3.70-Biotin, all antibodies were purchased from

Pharmingen. Streptavidin-PerCP and Streptavidin-Tricolor were purchased from

Results Part II

- 56 -

Pharmingen and Caltag Laboratories, respectively. All samples were acquired on a

FACSCalibur and analysed using CellQuest software (Becton Dickinson).

Acknowledgments

We thank Dr. Benedita Rocha for the generous gift of T3.70 mAb, and Drs. M. van

den Broek for helpful comments on the manuscript. This work was supported by the

Swiss National Foundation for Science and the Kanton of Zurich, Switzerland.

Results Part III

- 57 -

7 Results Part III

Selection of the H-Y-specific transgenic TCR in an athymic versus

euthymic environment

Marianne M. Martinic, Hans Hengartner and Rolf M. Zinkernagel

Institute of Experimental Immunology, Department of Pathology, University Hospital

Zurich, Schmelzbergstrasse 12, CH-8091 Zurich, Switzerland

Results Part III

- 58 -

Abstract

To compare intrathymic and extrathymic T cell selection, H-Y-specific T cell

receptor (TCR) transgenic mice (H-2bb) were crossed onto athymic C57BL/6- (H-

2bb) or BALB/c-nudes (H-2dd). Athymic and euthymic offspring were analysed

for selecting (H-2bb, H-2bd) or non-selecting (H-2dd) H-2 haplotype, and for

expression of transgenic TCR. Athymic mice had much lower absolute

numbers of CD8+ cells expressing transgenic TCR than euthymic controls. The

percentage of positively selected H-Y-specific transgenic CD8+ cells in females,

however, was comparable to euthymic controls. Negative selection of

transgenic TCR in athymic males was complete, as has been known for

euthymic males. Also escape from negative selection via down-regulation of

CD8ααααββββ co-receptor was observed in athymic males. In contrast, in athymic

mice escape from neglect via endogenous Vαααα-chain gene rearrangement was

less efficient than in euthymic mice. These results indicate that the thymic

microenvironment is 1) not essential for but much improves measurable

maturation and 2) is not necessary for negative selection of highly frequent

transgenic TCR expressing T cells.

Introduction

Our aim was to compare intrathymic and extrathymic T cell selection to analyse

whether the periphery could substitute a non-functional thymus by providing an

environment capable of selecting a functional self-MHC-restricted and self-tolerant T

cell repertoire. As it is well known that the thymus is absolutely required for T cell

receptor (TCR) rearrangement (Miller, 1961), we took advantage of a TCR transgenic

(tg) mouse system to compare selection of this specific transgenic TCR in an intra-

versus extrathymic environment. We analysed selection of the H-Y-specific

transgenic TCR, which is specific for a male antigen-derived peptide (H-Y) presented

on MHC class I H-2Db molecules (Kisielow et al., 1988a). The advantage of this

transgenic TCR is that without need of any external manipulation and by solely

comparing transgenic females with transgenic males, various parameters of T cell

selection can be directly assessed. In H-Y-specific TCR transgenic females (H-2bb),

presence of self-MHC (H-2Db) and absence of self-antigen (H-Y) lead to positive

selection of self-MHC-restricted H-Y-specific CD8αβhi tg TCRαβ+ cells (Kisielow et

al., 1988a). Maximally only 40% of CD8+ cells express the actual transgenic TCR, the

Results Part III

- 59 -

remaining CD8+ population recombined the endogenous TCRα loci and changed

TCR specificity during thymocyte selection (Buch et al., 2002; Huesmann et al.,

1991; Merkenschlager et al., 1994). In H-Y-specific TCR transgenic males (H-2bb),

presence of self-MHC (H-2Db) and of self-antigen (H-Y) clonally deletes all H-Y-

specific CD8αβhi tg TCRαβ+ cells (Kisielow et al., 1988a). Escape from negative

selection occurs through down-regulation of CD8αβ co-receptor resulting in H-Y-

unreactive CD8αβlo tg TCRαβ+ cells (Kisielow et al., 1988a; Teh et al., 1989). In H-Y-

specific TCR transgenic mice with non-selecting H-2 haplotype (H-2dd), H-Y-specific

CD8αβhi tg TCRαβ+ cells do not get positively selected and escape death from

neglect via endogenous rearrangement of TCRα-chain genes (Fink and McMahan,

2000; Teh et al., 1988; von Boehmer, 1990). To compare positive and negative

selection, escape from negative selection and escape from neglect in an intra- versus

extrathymic environment, euthymic H-Y-specific TCR transgenic mice (H-2bb) were

crossed twice with athymic C57BL/6- (H-2bb) or BALB/c-nudes (H-2dd) yielding

euthymic and athymic offspring with a selecting (H-2bb or H-2bd) or non-selecting H-2

haplotype (H-2dd).

Results

Athymic TCR transgenic mice show a strong reduction in absolute numbers of

transgenic cells compared to euthymic counterparts.

The biggest and striking difference between euthymic and athymic TCR transgenic

mice was the drastic reduction in absolute cell numbers (Table 5). Euthymic males

always had at least 40-fold more CD8αβlo tg TCRαβ+ splenocytes than athymic

males (Table 5). The difference between H-Y-specific CD8αβhi tg TCRαβ+ and

CD8αβhi tg TCRαend.+βtg

+ splenocytes in euthymic and athymic females was even

more striking, reaching a 120-fold decrease for athymic females (Table 5). Similar

reductions in absolute numbers were observed in liver and blood of athymic TCR

transgenic mice (data not shown). A possible explanation for the higher decrease in

absolute numbers in athymic females versus males may be that in athymic males

presence of self-MHC and self-antigen may lead to a primary survival signal for the

transgenic T cell via the TCR, allowing this T cell even under sub-optimal conditions

to move on to the next maturation step, which would be escape from negative

selection through CD8αβ co-receptor down-regulation. In presence of self-MHC but

Results Part III

- 60 -

absence of self-antigen - representing the athymic female environment – not all TCR

transgenic T cells may receive a sufficient high and long survival signal to move on to

the next maturation step, which would either be positive selection of H-Y-specific

CD8αβhi tg TCRαβ+ cells or internalisation of transgenic TCRα-chain followed by

endogenous Vα-chain gene rearrangement and expression of a TCR with new

specificity. A thymic environment, however, provides optimal conditions via the

perfect three-dimensional micro-architecture, its professional antigen presenting cells

and the specialized chemokine and cytokine milieu allowing immature thymocytes

not only to survive longer and therefore giving the thymocyte the chance to move on

to the next maturation step but also to increase their proliferation rate resulting in

higher absolute output cell numbers. Finally, thymocytes thanks to the thymus/blood

barrier are protected against peripheral chemokines such as TNF-α, which is

secreted during an immune response and is toxic for maturing cells, giving

thymocytes an additional survival advantage over extra-thymically maturing T cells

(Martin and Bevan, 1997; Orange et al., 1995; Wang et al., 1994).

Additionally, we analysed athymic TCR transgenic mice for presence of CD8αα+ tg

TCRαβ+ cells in the gut, blood, spleen and liver (data not shown). As observed

previously for euthymic TCR transgenic mice (Cruz et al., 1998), CD8αα+ tg TCRαβ+

cells were found exclusively in the gut of euthymic and athymic males but not at all in

euthymic or athymic females (data not shown).

Results Part III

- 61 -

Table 5: In athymic H-Y-specific TCR transgenic mice, absolute numbers of transgenic T cells are drastically

reduced when compared to euthymic transgenic mice.

TCR H-Y+/-

H-2bd

TCR H-Y+/-

H-2bd

nu/+

nu/+

nu/nu

nu/nu

MiceFold-

Difference

40

120

3.5E6

2.9E3

3.6E5

8.7E4

tg TCRαβ+ cells

# of CD8lo ( )

or CD8hi ( )Fold-

Difference

120

-

1.0E3

1.2E5

# of CD8hi

TCRαend.+βtg

+

cells

-

-

Euthymic (nu/+) and athymic (nu/nu) H-Y-specific TCR transgenic mice were sacrificed and total numbers of

CD8+ tgTCRαβ+ and of CD8+ TCRαend.(Vα8)+βtg

+ splenocytes assessed.

Self-MHC-restricted H-Y-specific TCR transgenic cells are positively selected in

athymic females.

The percentage of positively selected H-Y-specific CD8αβhi tg TCRαβ+ cells in

euthymic and athymic transgenic females was assessed via FACS-analysis (Figure

14). As already observed in euthymic females (Kisielow et al., 1988a), H-Y-specific

CD8+ cells in athymic females were only positively selected in presence of self-MHC

(Figure 14). Although absolute numbers of H-Y-specific TCR transgenic cells in

athymic females (H-2bd) were decreased 120-fold (Table 5), percentage of CD8+ cells

expressing transgenic TCR was comparable to euthymic females, being around 5-

10% (Figure 14) (Martinic et al., Manuscript in preparation-a). The remaining CD8+

population in euthymic (Buch et al., 2002; Huesmann et al., 1991; Merkenschlager et

al., 1994) and most of the remaining in athymic females performed endogenous Vα-

chain gene rearrangement expressing transgenic TCRβ-chain paired with

endogenous Vα-chain (Figure 14 and Figure 17A and data not shown). Again,

absolute numbers of T cells expressing tg TCRβ-chain paired with endogenous Vα8-

chain were 120-fold higher in euthymic compared to athymic transgenic females

(Table 5). Whether percentage of CD8+ cells in athymic females expressing

Results Part III

- 62 -

transgenic TCRαβ is also dependent on MHC-density levels as observed in euthymic

females remains to be proven (Martinic et al., Manuscript in preparation-a). H-Y-

specific CD8αβhi tg TCRαβ+ cells in euthymic females with a non-selecting H-2

haplotype (TCR H-Y+/-, H-2d d) escaped death from neglect via endogenous

rearrangement of TCRα-chain genes (Figure 14 and data not shown) (Fink and

McMahan, 2000; Teh et al., 1988). Athymic females with a non-selecting H-2

haplotype (TCR H-Y+/-, H-2dd), however, did not escape death from neglect via

endogenous Vα-chain gene rearrangement to detectable levels (Figure 14 and data

not shown). The latter observation may reflect again the disadvantage of a sub-

optimal athymic environment where absence of both self-MHC and self-antigen may

result in a too low and short survival signal for the maturing T cell.

101

103

101 103

101

103

101

103

101

103

Gated on CD8αβhi cells

nu/+ nu/nu

T3.70

Vβ8

.1/2

74%

90%

98%

18%

25.9%

9.4%

0.4%

0.0%

101

103

101

103

101 103

101

103 10% 0.7%

3% 0.0%

69% 7.4%

TCR H-Y+/-

H-2dd

TCR H-Y-/-

H-2bd

TCR H-Y+/-

H-2bb

TCR H-Y+/-

H-2bd

Figure 14: In presence of self-MHC (H-2b), positive selection of H-Y-specific transgenic TCR is seen in both

euthymic and athymic females. Blood lymphocytes of euthymic (nu/+) and splenocytes of athymic (nu/nu) H-Y-

specific TCR transgenic females (TCR H-Y+/-) expressing the selecting (H-2bb), a mixed (H-2bd) or non-selecting

H-2 haplotype (H-2dd) were gated on CD8αβhi cells and stained for surface expression of transgenic TCRαβ

(T3.70+Vβ8.1/2+). Transgenic negative littermates (TCR H-Y-/-, H-2bd) were used as negative controls. Numbers in

upper left (UL) and upper right (UR) quadrants represent percentage of CD8+ cells expressing exclusively

transgenic TCRβ-chain, and transgenic TCRαβ, respectively. Similar results were obtained with liver lymphocytes

of euthymic and athymic mice (data not shown). At least three individual mice were tested in each group. One out

of three similar experiments is shown.

Results Part III

- 63 -

Athymic TCR transgenic males show absence of all H-Y-specific transgenic

cells. Escape from negative selection occurs via down-regulation of CD8ααααββββ co-

receptor.

To compare negative selection in an athymic versus euthymic environment, H-Y-

specific TCR transgenic males were analysed for transgenic TCRαβ expression

(Figure 15 and Figure 16). Without exception, all males with selecting H-2 haplotype

(H-2bb, H-2bd) showed complete absence of H-Y-specific CD8αβhi tg TCRαβ+ cells

(Figure 15 and Figure 16 and data not shown). In addition, as already observed in

euthymic males, athymic males showed escape from negative selection via down-

regulation of CD8αβ co-receptor (Figure 16) leading to H-Y unreactive CD8αβlo tg

TCRαβ+ cells (Figure 15) (Kisielow et al., 1988a; Teh et al., 1989). The degree of

CD8αβ co-receptor down-regulation was strongly dependent on the density of MHC

class I H-2Db molecules; the higher the density of H-2Db, the stronger the down-

regulation of CD8αβ co-receptor in both euthymic and athymic males (Figure 16)

(Martinic et al., Manuscript in preparation-a). Escape from neglect via endogenous

Vα-chain gene rearrangement as observed in euthymic males with non-selecting H-2

haplotype (H-2dd) was also seen in athymic males (Figure 15 and Figure 17) (Fink

and McMahan, 2000; Teh et al., 1988). Interestingly, only athymic males but not

females with a non-selecting H-2 haplotype (H-2dd) were able to escape death from

neglect via endogenous Vα-chain gene rearrangement (Figure 14, Figure 15 and

Figure 17 and data not shown). One possible explanation may be that the H-Y-

specific transgenic TCR in athymic males with non-selecting H-2 haplotype (H-2dd)

cross-reacted with a male-specific antigen presented on H-2d leading to a high and

long enough survival signal to perform subsequent endogenous gene rearrangement

even in absence of a thymic environment. The resulting TCR (TCRαend.+βtg

+) with

new specificity and restriction to self-MHC H-2d would then enable the transgenic T

cell to survive and expand through continuous peripheral interaction with self-MHC

(Kirberg et al., 1997; Rocha and von Boehmer, 1991).

Results Part III

- 64 -

101

103

101

103

101

103

101 103

101

103

101

103

101 103

101

103

101

103

Gated on CD8αβlo cells

nu/+ nu/nu

T3.70

Vβ8

.1/2

95.9%

90.1%

1.4%

0.0%

1.2%

0.0%

92.4%

2.9%

92.9%

4.3%

17.4%

97.0%

7.7%

1.7%

TCR H-Y+/-

H-2dd

TCR H-Y-/-

H-2bd

TCR H-Y+/-

H-2bb

TCR H-Y+/-

H-2bd

Figure 15: In presence of self-MHC (H-2b) and self-peptide (H-Y), H-Y-specific transgenic T cells are absent in

both euthymic and athymic males. Blood lymphocytes of euthymic (nu/+) and athymic (nu/nu) H-Y-specific TCR

transgenic males (TCR H-Y+/-) expressing the selecting (H-2bb), a mixed (H-2bd) or non-selecting H-2 haplotype

(H-2dd) were gated on CD8αβlo cells and stained for surface expression of transgenic TCRαβ (T3.70+Vβ8.1/2+).

Transgenic negative littermates (TCR H-Y-/-, H-2bd) were used as negative controls. Numbers in UL and UR

corners represent percentage of CD8+ cells expressing exclusively transgenic TCRβ-chain, and transgenic

TCRαβ, respectively. Similar results were obtained with splenocytes and liver lymphocytes of euthymic and

athymic mice (data not shown). At least three individual mice were tested in each group. One out of three similar

experiments is shown.

Results Part III

- 65 -

040

101 103

4040

40

101 103

040

101 103

4040

40

101 103

nu/+ nu/nu

MFI133

268

291

144

MFI79

338

391

243

MFI106

MFI83

141

173

172220

245

176

Co

un

ts

CD8α CD8α CD8βCD8β

TCR H-Y+/-

H-2dd

TCR H-Y+/-

H-2bd

TCR H-Y+/-

H-2bb

TCR H-Y+/-

H-2bd

Figure 16: Escape from negative selection in euthymic and athymic H-Y-specific TCR transgenic males occurs via

down-regulation of CD8αβ co-receptor. Blood lymphocytes of euthymic (nu/+) and athymic (nu/nu) H-Y-specific

TCR transgenic males (TCR H-Y+/-) expressing the selecting (H-2bb), a mixed (H-2bd) or non-selecting H-2

haplotype (H-2dd) were stained for surface expression of CD8α- and CD8β-chain. Numbers in histogram plots

represent mean fluorescence intensity (MFI) of CD8α- and CD8β-chain, respectively. H-Y-specific TCR

transgenic females (TCR H-Y+/-, H-2bd) were used as positive controls for CD8αβ expression levels. At least 3

individual mice were tested in each group. One out of five similar experiments is shown.

Endogenous TCRαααα-chain gene rearrangement does not take place in athymic

TCR transgenic males.

To test whether endogenous Vα-chain gene rearrangement was suppressed in

athymic H-Y-specific TCR transgenic males with selecting H-2 haplotype (H-2bd) but

allowed in athymic females with selecting H-2 haplotype (H-2bd) as well as in athymic

mice with non-selecting H-2 haplotype (H-2dd), euthymic and athymic TCR transgenic

mice were gated on CD8+ cells and stained for surface expression of transgenic

TCRβ-chain and endogenous Vα8- (Figure 17A) or Vα2-chain (Figure 17B). Athymic

H-Y-specific TCR transgenic females but not males with selecting H-2 haplotype (H-

2bd) showed endogenous Vα-chain gene rearrangement (Figure 17), confirming

previous results obtained with euthymic mice [Buch, 2002 #646;Huesmann, 1991

#647;Merkenschlager, 1994 #648]. As already mentioned before, in athymic TCR

transgenic mice with non-selecting H-2 haplotype (H-2dd), endogenous Vα-chain

gene rearrangement could only be detected in males (Figure 14, Figure 15 and

Figure 17 and data not shown). This is in contrast to their euthymic counterparts

Results Part III

- 66 -

where both females and males performed endogenous Vα -chain gene

rearrangement (Figure 14, Figure 15 and Figure 17 and data not shown) (Fink and

McMahan, 2000; Teh et al., 1988).

101

103

101

103

101

103

101

103

101

103

101

103

101

103

101 103

101

103

101 103

nu/+ nu/nu

Vα8

Vβ8

.1/2

96%

97%

96%

15%

0.0%

1.9%

3.1%

0.8%

7.2% 1.0%

2.0% 0.0%

45% 8.4%

CD8αβhiCD8αβhi

CD8αβloCD8αβlo

78% 0.0%

TCR H-Y+/-

H-2bd

TCR H-Y+/-

H-2dd

TCR H-Y-/-

H-2bd

TCR H-Y+/-

H-2bd

TCR H-Y+/-

H-2dd

TCR H-Y+/-

H-2bd

Vα2

CD8αβhiCD8αβhi

CD8αβloCD8αβlo

Vβ8

.1/2

101

103

101 103

101

103

101

103

101 103

101

103 99%

92%

0.0%

4.8% 96% 1.6%

98% 0.0%

A

B

Figure 17: Endogenous Vα-chain gene rearrangement is less efficient in athymic H-Y-specific TCR transgenic

mice. Splenocytes of euthymic (nu/+) and athymic (nu/nu) H-Y-specific TCR transgenic mice (TCR H-Y+/-) with a

mixed (H-2bd) or non-selecting H-2 haplotype (H-2dd) were gated on CD8αβlo or CD8αβhi cells and stained for

surface expression of endogenous Vα8-chain (Vα8) and transgenic TCRβ-chain (Vβ8.1/2) (A) or endogenous

Vα2-chain (Vα2) and transgenic TCRβ-chain (Vβ8.1/2) (B). Transgenic negative females (TCR H-Y-/-, H-2bd) were

used as negative controls. Numbers in UL and UR quadrants represent percentage of CD8+ cells expressing

exclusively transgenic TCRβ-chain, and endogenous Vα- and transgenic TCRβ-chain, respectively. Similar

Results Part III

- 67 -

results were obtained with blood lymphocytes of euthymic and athymic mice (data not shown). At least three

individual mice were tested in each group. One out of three similar experiments is shown.

In summary, selection of the H-Y-specific transgenic TCR in an athymic versus

euthymic environment was mostly qualitatively (Figure 14, Figure 15, Figure 16,

Figure 17 and Table 6) but not quantitatively (Table 5) comparable. Athymic TCR

transgenic females with selecting H-2 haplotype (H-2bd) showed positive selection of

H-Y-specific CD8αβhi tg TCRαβ+ cells (Figure 14) whereas these cells were

completely absent in males (Figure 15 and Figure 16). Escape from negative

selection via CD8αβ co-receptor down-regulation was as efficient as in euthymic

males (Figure 16) (Kisielow et al., 1988a). The level of CD8αβ co-receptor down-

regulation was dependent on TCR-restricting H-2Db density confirming previous

observations with euthymic TCR transgenic males (Figure 16) (Martinic et al.,

Manuscript in preparation-a). Escape from neglect via endogenous Vα-chain gene

rearrangement in athymic TCR transgenic mice with non-selecting H-2 haplotype (H-

2dd) was only observed in males whereas euthymic controls showed escape from

neglect in both females and males (Figure 14, Figure 15 and Figure 17 and data not

shown) (Fink and McMahan, 2000; Teh et al., 1988). Finally, absolute numbers of H-

Y-specific CD8αβhi tg TCRαβ+ cells and of CD8αβhi TCRαend.+βtg

+ cells in athymic

females with selecting H-2 haplotype (H-2bd) were reduced 120-fold, absolute

numbers of CD8αβlo tg TCRαβ+ cells in athymic males having escaped negative

selection were reduced 40-fold when compared to their euthymic counterparts (Table

5).

Results Part III

- 68 -

Table 6: Selection of H-Y-specific transgenic TCR in an intra- versus extrathymic environment.

Intrathymic FigureExtrathymic

Escape from Negative Selection YES 15, 16YES

Negative Selection YES 15, 16YES

Positive Selection YES 14YES*

Endogenous Vα-chain gene rearrangement YES 17only in presence of

self-MHC or self-antigen**

Escape from Neglect YES 14, 15, 17only in presence of

self-MHC or self-antigen**

*Dependency on H-2Db density remained to be proven.

**Escape from neglect and endogenous Vα-chain gene rearrangement were not detected in TCR H-Y+/- females

with non-selecting H-2 haplotype (H-2dd) (absence of self-MHC and of self-antigen). TCR H-Y+/- females with

selecting H-2 haplotype (H-2bd) (presence of self-MHC) and TCR H-Y+/- males with non-selecting H-2 haplotype

(H-2dd) (H-Y-specific transgenic TCR might have cross-reacted with self-antigen presented on H-2d), however,

showed endogenous Vα-chain gene rearrangement.

Conclusion

Under these sub-optimal conditions (athymic environment), low frequencies of non-

transgenic precursor T cells will not show maturation of measurable numbers of T

cells except perhaps for allo responses, as seen in old nude mice, representing

probably thousands of specificities. Under optimal conditions (euthymic

environment), however, low frequencies of precursor T cells will mature in the thymus

to measurable numbers. Thus, the thymic microenvironment provides an excellent

three-dimensional architecture, chemokine milieu and specialized cells involved in

TCR rearrangement and assembly rendering the thymus therefore the optimal organ

for efficient T cell repertoire selection. TCR transgenic mice, however, with 103-105

times higher precursor frequencies seem to offer sufficient numbers for maturation in

the absence of a thymus. Negative selection of these T cells is as complete as

observed in euthymic mice. However, in absence of a TCR-rearrangement-favouring

thymic environment and/or because of the much lower frequencies, use of an

endogenous TCRVα is barely measurable in absence of a thymus. The periphery

seems therefore clearly not specialized for efficient T cell maturation and selection

Results Part III

- 69 -

but as already known is essential for further survival, expansion and differentiation

(Kirberg et al., 1997; Rocha and von Boehmer, 1991; Zinkernagel and Althage,

1999). Thus, T cell maturation is probably 103-104 times more efficient in a thymus

but can occur rarely in its absence. Biologically, however, this numbers game

separates immunocompetence from general immunodeficiency as experienced by

thymus-deficient nude mice or humans.

Results Part III

- 70 -

Materials and Methods

Mice

H-Y-specific TCR transgenic mice (TCR H-Y+/+, H-2bb) were a generous gift from Jörg

Kirberg. BALB/c- (H-2dd) and C57BL/6-nudes (H-2bb) were purchased from the

Institute of Laboratory Animal Science, University of Zurich, Switzerland, and from

RCC Biotech, Füllinsdorf, Switzerland, respectively. H-Y-specific TCR transgenic

mice were bred with C57BL/6- or BALB/c-nudes for two generations yielding

euthymic (nu/+) and athymic (nu/nu) H-Y-specific TCR transgenic mice expressing

different H-2 haplotypes (TCR H-Y+/-, H-2bb, H-2bd or H-2dd). The H-2 haplotype of

these mice and expression of transgenic TCR were tested using PCR- and FACS-

analysis.

PCR

DNA was prepared from mouse-tail. 2 µl of mouse-tail-DNA were used for PCR

analysis. Primers used for amplification were specific for the H-Y-specific transgenic

TCR (Vβ8.2-fwd (5’ ACA AGG TGG CAG TAA CAG GA 3’) and Jβ2.3-rev (5’ ACA

GTC AGT CTG GTT CCT GA 3’) primers), for both H-2bb and H-2dd MHC haplotypes

(Ea5’-fwd (5’ AGT CTT CCC AGC CTT CAC ACT CAG AGG TAC 3’) and Ea3’-rev

(5’ CAT AGC CCC AAA TGT CTG ACC TCT GGA GAG 3’) primers) and for the H-

2dd MHC haplotype (K5’-fwd (5’ CAT GGG CAT AGA AAG GGC AGT CTT TGA ACT

3’) and Ea3’-rev primers). Expected bands were 302 bp for presence of H-Y-specific

transgenic TCR, 155 bp and 700 bp for presence of H-2b and H-2d MHC haplotype,

respectively, and 210 bp for presence of H-2d MHC haplotype.

Flow Cytometric Analysis

Peripheral blood cells or splenocytes were stained with the following antibodies: anti-

CD8α-APC (53-6.7), anti-CD8α-Tricolor (53-6.7), anti-CD8β.2-PE (53-5.8), T3.70-

Biotin (specific for transgenic TCRα-chain of H-Y-specific TCR) (generous gift from

Benedita Rocha), anti-Vβ8.1/2-FITC (MR5-2) (specific for transgenic TCRβ-chain of

H-Y-specific TCR), anti-Vα8-Biotin (KT50) and anti-Vα2-PE (B20.1). Except for

T3.70-Biotin antibody, all antibodies were purchased from Pharmingen. Streptavidin-

PerCP was purchased from Pharmingen, Streptavidin-Tricolor from Caltag

Results Part III

- 71 -

Laboratories. All samples were acquired on a FACScan or FACSCalibur and

analysed using CellQuest software (Becton Dickinson).

Acknowledgments

We thank Dr. Benedita Rocha for the generous gift of T3.70 mAb. This work was

supported by the Swiss National Foundation for Science and the Kanton of Zurich,

Switzerland.

General Discussion

- 73 -

8 General Discussion

An optimal T-cell mediated immune response is achieved in the presence of a

mature, functional and diverse T cell repertoire, which consists of self-MHC-restricted

and self-tolerant T cells (reviewed in Sebzda et al., 1999; Stefanova et al., 2003; von

Boehmer et al., 2003). Based on the established data and results obtained here,

selection and survival of such an efficient T cell repertoire can be summarized as

follows with respect to what defines this T cell repertoire, why is it needed, which

cells are involved, when and where does it occur and which are the different

parameters influencing it (Figure 18).

T cell repertoire

selection and survival

WHAT?

WHERE?

HOW?

WHEN?WHO?

WHY?

Figure 18: On T cell repertoire selection and survival.

General Discussion

- 74 -

WHAT and WHY?

The immune system of humans and other vertebrates is an adaptive defence system

that has evolved to protect from invading pathogenic microorganisms (bacteria, virus,

fungi) and cancer (Goldsby et al., 2000). It consists of both non-specific (innate

immunity) and specific (adaptive immunity) components. Just after contact to a

pathogen, innate immunity provides the first line of defence (i.e. via phagocytic cells,

release of interferon, complement activation). If the invading pathogen evades the

innate immunity or is not cleared by it (as observed with many co-evolved

pathogens), the specific branch of the immune system gets triggered. The advantage

of the adaptive immunity is the ability of specific recognition and selective elimination

of foreign microorganisms and molecules resulting from a tremendous diversity in its

recognition molecules together with the capability to distinguish between self and

non-self (Goldsby et al., 2000). Failure of the latter results in auto-immunity and

destruction of the host. Therefore, with emergence of the adaptive immune system it

is an absolute prerequisite that the T cell repertoire is able to discriminate between

self and non-self (Stefanova et al., 2003; von Boehmer et al., 2003). To ensure this

self-non-self discrimination, thymocytes go through a strict selection process during

their maturation. Thymocytes expressing productively rearranged TCR with weak to

intermediate overall avidity to self-peptide/self-MHC complex receive a survival signal

(positive selection) whereas TCR with high overall avidity to self-peptide/self-MHC

complex are deleted via TCR-induced apoptosis (negative selection) (reviewed in

Benoist and Mathis, 1999). This selection process ensures that all T cells leaving the

thymus are self-MHC-restricted and self-tolerant. Finally, for further survival in the

periphery, it is essential that T cells remain in continuous interaction with self-MHC

(Kirberg et al., 1997; Rocha and von Boehmer, 1991). Furthermore, self-MHC-

restriction gives T cells a greater chance of being activated by non-self-peptide/self-

MHC complex than by non-self-peptide alone or together with non-self-MHC (Bevan,

1977; von Boehmer et al., 1978; Zinkernagel et al., 1978). These observations are

supported by recent data using new methods analysing TCR-MHC interactions and

TCR signalling events (Davis, 2002; Stefanova et al., 2002). Mark Davis et al. show

that MHC surface amino acids accessible for the TCR contribute significantly to the

on-rate of the TCR-MHC interaction whereas the peptide contributes to the off-rate

(Davis, 2002). Thus, TCR self-MHC interaction positions the TCR already in the right

position for future peptide binding and therefore increases the strength and duration

General Discussion

- 75 -

of the future TCR peptide/MHC interaction. Stefanova et al. show that the TCR self-

MHC interaction itself increases the sensitivity of the TCR towards foreign antigen by

constantly transmitting a signalling threshold through partial phosphorylation of the

TCRζ chain, therefore preparing the T cell to respond to a low density of foreign

peptide already early in an infection (Stefanova et al., 2002; Stefanova et al., 2003).

In summary, evolution of a T cell repertoire capable of distinguishing between self

and non-self made it possible to increase the diversity of the T cell repertoire and

therefore the quality of the immune response helping humans and other vertebrates

to eliminate more efficiently even highly evolved pathogens.

WHO?

During the past twenty to thirty years, many experiments have been performed to

analyse which cells are involved in T cell repertoire selection. Until recently the

predominant point of view was that the MHC of radio-resistant cells of the thymus

(presumably thymic epithelial (TE) cells) is responsible for efficient T cell repertoire

selection. This conclusion is based on a series of classical irradiation bone marrow

and thymus chimera experiments (Fink and Bevan, 1978; Kappler and Marrack,

1978; Kisielow et al., 1988b; von Boehmer et al., 1978; Zinkernagel et al., 1978) and

reviewed in Moller, 1978; von Boehmer et al., 2003). For example, reconstitution of

lethally irradiated A-recipients with F1 (AxB) bone marrow and/or F1 (AxB) nude

grafted with a fetal thymus A yield A-restricted but virtually no B-restricted virus-

specific cytotoxic T lymphocytes (CTL) in a primary immune response (Table 7) (Fink

and Bevan, 1978; Zinkernagel et al., 1978). In parallel to this view, some researchers

suggested that the MHC of non-TE cells could also be involved in T cell repertoire

selection (Bix and Raulet, 1992; Doherty and Bennink, 1979; Hugo et al., 1993;

Longo and Schwartz, 1980; Matzinger and Mirkwood, 1978; Wagner et al., 1980;

Zinkernagel et al., 1980). However, these earlier studies were often based on relative

complicated protocols - in vivo filtration experiments, several adoptive transfers into

lethally irradiated recipients (Doherty and Bennink, 1979) – or required several

rounds of in vitro restimulation before detecting any non-TE MHC-restricted

responses (Bix and Raulet, 1992; Hugo et al., 1992; Hugo et al., 1993; Longo and

Schwartz, 1980; Matzinger and Mirkwood, 1978; Pawlowski et al., 1993; Vukmanovic

et al., 1992; Wagner et al., 1980). In addition, the specificity of non-TE MHC-

restricted T cells was not defined at the peptide level, but rather involved activity

General Discussion

- 76 -

against minor histocompatibility antigens (Bix and Raulet, 1992; Matzinger and

Mirkwood, 1978), allo-antigens (Bix and Raulet, 1992; Longo and Schwartz, 1980;

Wagner et al., 1980) and/or synthetic polymers (Bix and Raulet, 1992; Hugo et al.,

1992; Hugo et al., 1993) detected via secondary CTL, mixed lymphocyte reaction

and/or proliferation assays, respectively. In another set of experiments, injection of

allogeneic MHC class II-positive A-fibroblasts or A-thymic epithelial cells into the

thymus of lethally irradiated B-mice reveals restriction to both donor and host MHC

haplotype (Hugo et al., 1992; Hugo et al., 1993; Pawlowski et al., 1993; Vukmanovic

et al., 1992). These results suggest that even fibroblasts not specialized to process

and present antigens in the context of MHC class II molecules can mediate positive

selection when transfected with the appropriate MHC molecule (Hugo et al., 1993).

However, it has to be kept in mind that before intrathymic injection of A-positive cells,

lethally irradiated B-mice were reconstituted with (AxB)-fetal liver cells and therefore

A-positive cells are present both in the thymus and in the periphery allowing positive

selection and further survival of A-restricted T cells. Surprisingly, Zinkernagel et al.

show that nude mice reconstituted with a completely allogeneic day-14 fetal thymus

graft develop virus-specific primary immune response restricted almost exclusively to

the nude- but not to the thymus graft-MHC haplotype (Table 7) (Zinkernagel et al.,

1980). These results were again not entirely accepted because of potential rescue of

the nude thymic rudiment although recent data using tetraparental chimeras between

euthymic and nude donors of distinct MHC haplotype reveals that the thymic

rudiment of the nude donor cannot be rescued anatomically to form mature thymic

epithelial cells (Blackburn et al., 1996; Rodewald et al., 2001). Taken together, the

above-discussed experiments were not able to define satisfactorily the role of non-TE

MHC in T cell repertoire selection and survival. Therefore, in order to clarify the role

of TE versus non-TE MHC in T cell repertoire selection and survival, tetraparental

aggregation chimeras from T and B cell-deficient mice (H-2d Scid or H-2b Rag-/-) and

thymus-deficient nude mice (H-2b or H-2d) were generated (Results Part I) (Martinic

et al., 2003). Analysis of these relatively undisturbed (non-irradiated and non-

reconstituted) tetraparental aggregation chimeras confirmed the absence of rescue of

nude thymic rudiment (Results Part I) (Martinic et al., 2003). After viral infection,

tetraparental aggregation chimeras revealed primary protective CD8+ T cell

responses that were peptide-specifically restricted to either TE or non-TE MHC at

comparable levels (Results Part I) (Martinic et al., 2003). Furthermore, protective

General Discussion

- 77 -

virus-neutralizing IgG responses dependent on cognate CD4+ T helper cells

restricted to non-TE MHC were also mounted (Results Part I) (Martinic et al., 2003).

Therefore, these results showed convincingly that non-TE MHC was sufficient to

select a functional T cell repertoire (Results Part I) (Martinic et al., 2003).

Table 7: On the role of thymic epithelial (TE) versus non-TE MHC in T cell repertoire selection.

WHERE and WHEN?

In addition to the now obtained knowledge with respect to which cells are involved in

T cell repertoire selection, the next step is to analyse whether there exists a specific

location and a time-frame for most efficient T cell repertoire selection. Unfortunately it

was not possible to design any experiment where the periphery could be observed

independently from the thymus, therefore the focus is set on both thymus and

periphery simultaneously. Based on previously published data and results obtained

here, T cell repertoire selection and survival of positively selected T cells is most

efficient when TCR-restricting MHC-expressing cells are present both in the thymus

and in the periphery at all times (Table 7). This is shown by the following examples:

Reconstitution of lethally irradiated A-mice with (AxB)-bone marrow leads to survival

General Discussion

- 78 -

of A- but not B-restricted cells (Table 7) (Fink and Bevan, 1978; Longo and Schwartz,

1980; Zinkernagel et al., 1978). In these mice, MHCA is present in both the thymus

and the periphery at all times, MHCB however is present initially exclusively in the

periphery and only after a few months BM-derived MHCB-positive cells will have

migrated into the thymus (Longo and Schwartz, 1980). By that time it is too late for

selection and survival of B-restricted T cells because A-restricted T cells have

already outnumbered B-restricted T cells and therefore only A-restricted T cells are

detected in the periphery (Lemischka et al., 1986; Longo and Schwartz, 1980;

Zinkernagel, 1982). Interestingly, this exclusive restriction to MHCA will change if

these mice are depleted of thymocytes and mature T cells eight months after BM-

reconstitution and allowed to replenish the T cell pool. Analysis of these mice three

months after T cell depletion reveals an A- and B-restricted T cell repertoire (Longo

and Schwartz, 1980) confirming the here developed hypothesis that survival of MHC-

restricted T cells is only guaranteed if TCR-restricting MHC-expressing cells are

present both in the thymus and in the periphery from the start. However, it remains to

say that Longo et al. do not use assays analysing defined antigen-specificity but

rather proliferation measurements or mixed lymphocyte reaction assays. In addition,

A- and B-restricted responses are never obtained ex vivo but only after in vitro

restimulation. Therefore, it might be interesting to repeat above-discussed

experiments but this time using defined antigens and more sensitive assays (i.e.

tetramer staining). Finally, with the here obtained results, it might be interesting to

readdress the question whether MHC class II positive haematopoietic cells can

mediate positive selection in MHC class II deficient mice (Markowitz et al., 1993).

Markowitz et al. show that lethally irradiated MHC class II-deficient mice reconstituted

with bone marrow from MHC class II-positive donors or nude mice grafted with

irradiated MHC class II-deficient thymus do not show positive selection and survival

of CD4+ cells. Positive selection and survival of CD4+ cells is only observed when TE

cells express MHC class II molecules (Markowitz et al., 1993). It would be interesting

to analyse what would happen if donor-MHC would be present both in the thymus

and in the periphery. This can be easily achieved if eight months after reconstitution

with bone marrow from MHC class II-positive donors, the peripheral T cell pool and

all thymocytes would be eliminated by cortisone and anti-thymocyte-serum treatment

(similar to the experiments performed by Longo et al. (Longo and Schwartz, 1980).

Three months after T cell repertoire depletion, newly emerging T cells would

General Discussion

- 79 -

encounter donor-APC (MHC class II-positive) already in the thymus, could therefore

be positively selected on donor-MHC, followed by commitment to the CD4 T cell

lineage and emigration into the periphery. There, further survival would be

guaranteed through continuous interaction with donor-MHC. Another way to test for

the ability of bone marrow-derived MHC class II-positive cells to positively select DP

thymocytes would be analysis of H-2b Rag-/- MHC class II-/- ↔ nuded tetraparental

aggregation chimeras. In these mice, all TE cells are MHC class II-deficient and only

professional APC from nuded donor are MHC class II-positive. In these mice but not

in the irradiation chimeras from Markowitz et al, MHC class II-positive cells are

present from the start both in the thymus and in the periphery. If after LCMV-infection

or VSV-infection, tetraparental aggregation chimeras are able to mount virus-specific

CD4-dependent IgG responses, haematopoietic cells are efficient in selecting CD4+ T

cells. If, however, these mice show a reduced CD4+ T cell response this might be

explained by the overall reduced density of MHC class II molecules in these

tetraparental aggregation chimeras as already observed by different labs for MHC

class I-deficient irradiation chimeras (Bix and Raulet, 1992; Terra et al., 2002;

Zerrahn et al., 1999).

Another example with nudeA mice grafted with fetal thymusB reveals survival of A-

restricted T cells (Table 7) (Zinkernagel et al., 1980). Again, MHCA-expressing cells

are the only cells present, both in the periphery and in the thymus, whereas MHCB-

expressing cells are only present in the thymus and therefore MHCB-restricted T cells

have no chance to survive further on in the periphery.

Finally, analysis of Rag-/-A ↔ nudeB and ScidB ↔ nudeA tetraparental aggregation

chimeras where both MHCA and MHCB-expressing cells are present from the start in

the thymus and in the periphery show survival of both A- and B-restricted T cells

confirming the above formulated hypothesis (Table 7) (Martinic et al., 2003).

Furthermore, these results are in agreement with data obtained from transplantation

studies. Mice with a high degree of haematopoietic chimerism showing host- and

donor-derived cells, both in the thymus and in the periphery, reveal complete allograft

tolerance and show both host- and donor-restricted T cell response after LCMV- or

VSV-infection (Adams et al., 2001; Williams et al., 2003). In another study, mice

receiving syngeneic thymus graft show presence of syngeneic MHC class II-

restricted CD4+ cells only if syngeneic MHC class II is found also in the periphery

(Rodriguez-Barbosa et al., 2002). Surprisingly, mice receiving xenogeneic porcine

General Discussion

- 80 -

thymus graft show presence of porcine MHC class II-restricted CD4+ cells even in

absence of porcine MHC class II in the periphery. However, this is only possible

because of the presence of murine MHC class II IAb in the periphery, which shows

highly conserved amino acid sequence with porcine MHC class II DQd probably

substituting therefore the need for porcine MHC class II (Rodriguez-Barbosa et al.,

2002).

HOW?

It is well known and accepted that the most efficient site for initial TCR

rearrangement and maturation is the thymus itself (Bach, 1979; Goldstein and White,

1971; Miller, 1961). The thymus provides an excellent three-dimensional

microenvironment including compartmentalization of maturing DP and SP

thymocytes and therefore optimal cytokine milieu for each individual compartment, a

physical barrier between the thymus and the periphery protecting maturing

thymocytes from the peripheral environment (optimised for efficient T cell activation

and differentiation but not for maturation) and preventing the re-entry of mature

thymocytes into the thymus (impeding therefore possible competition for certain

cytokines), and finally presence of professional APC with the unique capability of

promiscuous expression of peripheral self-antigens enabling negative selection of

thymocytes with high overall avidity against peripheral self-antigens (Anderson et al.,

2002). It was therefore interesting to analyse the efficiency of T cell repertoire

selection and survival under sub-optimal conditions, for example in the presence of

reduced TCR-restricting MHC density (Results Part II) or in the absence of a thymus

(Results Part III). Selection and survival of the low-affinity H-Y-specific transgenic

TCR but not of the high-affinity 2C transgenic TCR were less efficient under reduced

TCR-restricting MHC density (Results Part II) (Zerrahn et al., 1999). In an athymic

environment, positive selection and survival of the H-Y-specific transgenic TCR in

females were even more drastically reduced (Results Part III). Absence of self-

reactive H-Y-specific transgenic TCR in athymic males, however, was complete

(Results Part III). Other labs studying T cell repertoire selection and survival under

sub-optimal conditions have obtained similar results (Terra et al., 2002) and Perreault

et al., manuscript in preparation). Perreault et al, for example, show reduced positive

selection and survival of the H-Y-specific transgenic TCR in double transgenic H-

Y/LckOM females but unimpaired positive selection and survival of the 2C transgenic

General Discussion

- 81 -

TCR in 2C/LckOM double transgenic mice (Terra et al., 2002). However, the overall

TCR signalling strength in these mice represented by CD5 expression levels in DP

and SP mesenteric lymph node cells is reduced when compared to the ones in DP

and SP thymocytes of H-Y or 2C TCR single transgenic mice (Terra et al., 2002). To

compare the immune response of thymically derived with the one of extrathymically

derived T cells, C57BL/6 control mice and adult thymectomized lethally irradiated

Rag2-/- mice reconstituted with fetal liver of LckOM transgenic mice were infected

with LCMV and VSV (Perreault et al., manuscript in preparation). Only few of the

extrathymically derived T cells were able to show some cytotoxic activity against

LCMV-derived epitopes after in vitro restimulation and only few were able to produce

partial VSV-specific IgM and very low IgG titers. Nevertheless, none of the mice with

extrathymically derived T cells were able to clear LCMV or VSV-infection (Perreault

et al., manuscript in preparation). Taken together, these results show that although

protection against self-reactivity is still guaranteed under sub-optimal extrathymic

conditions, efficiency of positive selection is so reduced that it is not sufficient for full

immunocompetence. Just recently Guy-Grand et al. analysed extrathymic

lymphopoiesis in detail in athymic and euthymic mice (Guy-Grand et al., 2003). As

already seen in LckOM transgenic mice, mesenteric lymph nodes (mLN) are the

major site and Peyer’s patches (PP) a minor site of extrathymic lymphopoiesis (Guy-

Grand et al., 2003). However, in contradiction to previous observations (Poussier and

Julius, 1994; Rocha et al., 1991; Saito et al., 1998), the cryptopatches and the gut

epithelium prove rather inefficient (Guy-Grand et al., 2003). Guy-Grand also

concludes that the reduced efficiency of selection and survival of an extrathymically

derived T cell repertoire is due to the sub-optimal athymic environment. As an

example, Guy-Grand shows that the absence of CD8αα+ cells in athymic mice is due

to the lack of the right self-peptide/MHC complexes and the absence of the right

cytokine milieu (lack of IL-7 and IL-15) for survival and expansion of CD8αα+

precursors (DN TCRαβ+ NK1.1- cells which are present in the thymus of euthymic

mice) in the mLN (Guy-Grand et al., 2003). In euthymic mice, both specific

peptide/MHC complexes and the right cytokine milieu are provided in the thymus and

therefore CD8αα+ cells are present (Guy-Grand et al., 2003). In conclusion, in

healthy euthymic mice, extrathymic lymphopoiesis is absent. It can resume, however,

in conditions of severe lymphocyte depletion, absence of TCRαβ+ cells or marked

General Discussion

- 82 -

thymic atrophy to ensure at least some T cell repertoire selection and survival (Guy-

Grand et al., 2003).

In summary, optimal conditions (presence of a thymus and high TCR-restricting MHC

density present at all times throughout the whole body) lead to efficient selection and

survival of a self-MHC-restricted and self-tolerant T cell repertoire capable to

generate immune responses against all kinds of pathogens. Trying to understand

more in detail the exact mechanisms governing this selection and survival process,

might help us to envisage new strategies for improved therapies against autoimmune

diseases and for successful organ transplantations.

References

- 83 -

9 References

Adams, A. B., Durham, M. M., Kean, L., Shirasugi, N., Ha, J., Williams, M. A., Rees,P. A., Cheung, M. C., Mittelstaedt, S., Bingaman, A. W., et al. (2001). Costimulationblockade, busulfan, and bone marrow promote titratable macrochimerism, inducetransplantation tolerance, and correct genetic hemoglobinopathies with minimalmyelosuppression. J Immunol 167, 1103-1111.

Akashi, K., Reya, T., Dalma-Weiszhausz, D., and Weissman, I. L. (2000). Lymphoidprecursors. Curr Opin Immunol 12, 144-150.

Anderson, G., and Jenkinson, E. J. (2001). Lymphostromal interactions in thymicdevelopment and function. Nat Rev Immunol 1, 31-40.

Anderson, G., Moore, N. C., Owen, J. J., and Jenkinson, E. J. (1996). Cellularinteractions in thymocyte development. Annu Rev Immunol 14, 73-99.

Anderson, M. S., Venanzi, E. S., Klein, L., Chen, Z., Berzins, S. P., Turley, S. J., vonBoehmer, H., Bronson, R., Dierich, A., Benoist, C., and Mathis, D. (2002). Projectionof an immunological self shadow within the thymus by the aire protein. Science 298,1395-1401.

Apostolou, I., Sarukhan, A., Klein, L., and von Boehmer, H. (2002). Origin ofregulatory T cells with known specificity for antigen. Nat Immunol 3, 756-763.

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 aviditymodel of T cell selection in the thymus. Cell 76, 651-663.

Bach, J. F. (1979). Thymic hormones. J Immunopharmacol 1, 277-310.

Bachmann, M. F., Kundig, T. M., Kalberer, C. P., Hengartner, H., and Zinkernagel, R.M. (1993). Formalin inactivation of vesicular stomatitis virus impairs T-cell- but not T-help-independent B-cell responses. J Virol 67, 3917-3922.

Battegay, M., Moskophidis, D., Waldner, H., Brundler, M. A., Fung-Leung, W. P.,Mak, T. W., Hengartner, H., and Zinkernagel, R. M. (1993). Impairment and delay ofneutralizing antiviral antibody responses by virus-specific cytotoxic T cells [publishederratum appears in J Immunol 1994 Mar 1;152(5):1635]. J Immunol 151, 5408-5415.

Bendelac, A., Bonneville, M., and Kearney, J. F. (2001). Autoreactivity by design:innate B and T lymphocytes. Nat Rev Immunol 1, 177-186.

Bendelac, A., Hunziker, R. D., and Lantz, O. (1996). Increased interleukin 4 andimmunoglobulin E production in transgenic mice overexpressing NK1 T cells. J ExpMed 184, 1285-1293.

Bendelac, A., Rivera, M. N., Park, S. H., and Roark, J. H. (1997). Mouse CD1-specific NK1 T cells: development, specificity, and function. Annu Rev Immunol 15,535-562.

Benlagha, K., Kyin, T., Beavis, A., Teyton, L., and Bendelac, A. (2002). A thymicprecursor to the NK T cell lineage. Science 296, 553-555.

References

- 84 -

Benoist, C., and Mathis, D. (1999). T-lymphocyte differentiation and biology. InFundamental Immunology. W. E. Paul, ed., Fourth edn (Philadelphia, Lippincott-Raven Publishers).

Bensinger, S. J., Bandeira, A., Jordan, M. S., Caton, A. J., and Laufer, T. M. (2001).Major histocompatibility complex class II-positive cortical epithelium mediates theselection of CD4(+)25(+) immunoregulatory T cells. J Exp Med 194, 427-438.

Bevan, M. J. (1977). In a radiation chimaera, host H-2 antigens determine immuneresponsiveness of donor cytotoxic cells. Nature 269, 417-418.

Bix, M., and Raulet, D. (1992). Inefficient positive selection of T cells directed byhaematopoietic cells. Nature 359, 330-333.

Blackburn, C. C., Augustine, C. L., Li, R., Harvey, R. P., Malin, M. A., Boyd, R. L.,Miller, J. F., and Morahan, G. (1996). The nu gene acts cell-autonomously and isrequired for differentiation of thymic epithelial progenitors. Proc Natl Acad Sci U S A93, 5742-5746.

Boileau, C., Houde, M., Dulude, G., Clegg, C. H., and Perreault, C. (2000).Regulation of extrathymic T cell development and turnover by oncostatin M. JImmunol 164, 5713-5720.

Borgulya, P., Kishi, H., Uematsu, Y., and von Boehmer, H. (1992). Exclusion andinclusion of alpha and beta T cell receptor alleles. Cell 69, 529-537.

Borowski, C., Martin, C., Gounari, F., Haughn, L., Aifantis, I., Grassi, F., and vonBoehmer, H. (2002). On the brink of becoming a T cell. Curr Opin Immunol 14, 200-206.

Brandle, D., Muller, C., Rulicke, T., Hengartner, H., and Pircher, H. (1992).Engagement of the T-cell receptor during positive selection in the thymus down-regulates RAG-1 expression. Proc Natl Acad Sci U S A 89, 9529-9533.

Brugnera, E., Bhandoola, A., Cibotti, R., Yu, Q., Guinter, T. I., Yamashita, Y.,Sharrow, S. O., and Singer, A. (2000). Coreceptor reversal in the thymus: signaledCD4+8+ thymocytes initially terminate CD8 transcription even when differentiatinginto CD8+ T cells. Immunity 13, 59-71.

Bruno, L., Fehling, H. J., and von Boehmer, H. (1996). The alpha beta T cell receptorcan replace the gamma delta receptor in the development of gamma delta lineagecells. Immunity 5, 343-352.

Buch, T., Rieux-Laucat, F., Forster, I., and Rajewsky, K. (2002). Failure of HY-specific thymocytes to escape negative selection by receptor editing. Immunity 16,707-718.

Capone, M., Troesch, M., Eberl, G., Hausmann, B., Palmer, E., and MacDonald, H.R. (2001). A critical role for the T cell receptor alpha-chain connecting peptidedomain in positive selection of CD1-independent NKT cells. Eur J Immunol 31, 1867-1875.

Casanova, J. L., Romero, P., Widmann, C., Kourilsky, P., and Maryanski, J. L.(1991). T cell receptor genes in a series of class I major histocompatibility complex-restricted cytotoxic T lymphocyte clones specific for a Plasmodium bergheinonapeptide: implications for T cell allelic exclusion and antigen-specific repertoire. JExp Med 174, 1371-1383.

References

- 85 -

Casellas, R., Shih, T. A., Kleinewietfeld, M., Rakonjac, J., Nemazee, D., Rajewsky,K., and Nussenzweig, M. C. (2001). Contribution of receptor editing to the antibodyrepertoire. Science 291, 1541-1544.

Chaffin, K. E., and Perlmutter, R. M. (1991). A pertussis toxin-sensitive processcontrols thymocyte emigration. Eur J Immunol 21, 2565-2573.

Charan, S., and Zinkernagel, R. M. (1986). Antibody mediated suppression ofsecondary IgM response in nude mice against vesicular stomatitis virus. J Immunol136, 3057-3061.

Clegg, C. H., Rulffes, J. T., Wallace, P. M., and Haugen, H. S. (1996). Regulation ofan extrathymic T-cell development pathway by oncostatin M. Nature 384, 261-263.

Constant, P., Davodeau, F., Peyrat, M. A., Poquet, Y., Puzo, G., Bonneville, M., andFournie, J. J. (1994). Stimulation of human gamma delta T cells by nonpeptidicmycobacterial ligands. Science 264, 267-270.

Cruz, D., Sydora, B. C., Hetzel, K., Yakoub, G., Kronenberg, M., and Cheroutre, H.(1998). An opposite pattern of selection of a single T cell antigen receptor in thethymus and among intraepithelial lymphocytes. J Exp Med 188, 255-265.

Cumano, A., Paige, C. J., Iscove, N. N., and Brady, G. (1992). Bipotential precursorsof B cells and macrophages in murine fetal liver. Nature 356, 612-615.

Curnow, S. J., Boyer, C., Buferne, M., and Schmitt-Verhulst, A. M. (1995). TCR-associated zeta-Fc epsilon RI gamma heterodimers on CD4-CD8- NK1.1+ T cellsselected by specific class I MHC antigen. Immunity 3, 427-438.

Davis, M. (2002). Gene expression and signaling in the immunal system. Paperpresented at: Cold Spring Harbour Symposium (Cold Spring Harbour).

Doherty, P. C., and Bennink, J. C. (1979). Vaccinia-specific cytotoxic T-cellresponses in the context of H-2 antigens not encountered in thymus may reflectaberrant recognition of a virus-H-2 complex. J Exp Med 149, 150-157.

Eppig, J. J., Kozak, L. P., Eicher, E. M., and Stevens, L. C. (1977). Ovarianteratomas in mice are derived from oocytes that have completed the first meioticdivision. Nature 269, 517-518.

Fink, P. J., and Bevan, M. J. (1978). H-2 antigens of the thymus determinelymphocyte specificity. J Exp Med 148, 766-775.

Fink, P. J., and McMahan, C. J. (2000). Lymphocytes rearrange, edit and revise theirantigen receptors to be useful yet safe. Immunol Today 21, 561-566.

Goldsby, R. A., Kindt, T. J., and Osborne, B. A. (2000). Kuby immunology, Fourthedn (New York, W. H. Freeman and Company).

Goldstein, A. L., and White, A. (1971). Role of thymosin and other thymic factors inthe development, maturation, and functions of lymphoid tissue. Curr Top ExpEndocrinol 1, 121-149.

Guy-Grand, D., Azogui, O., Celli, S., Darche, S., Nussenzweig, M. C., Kourilsky, P.,and Vassalli, P. (2003). Extrathymic T cell lymphopoiesis: ontogeny and contributionto gut intraepithelial lymphocytes in athymic and euthymic mice. J Exp Med 197, 333-341.

References

- 86 -

Guy-Grand, D., Cuenod-Jabri, B., Malassis-Seris, M., Selz, F., and Vassalli, P.(1996). Complexity of the mouse gut T cell immune system: identification of twodistinct natural killer T cell intraepithelial lineages. Eur J Immunol 26, 2248-2256.

Guy-Grand, D., Rocha, B., Mintz, P., Malassis-Seris, M., Selz, F., Malissen, B., andVassalli, P. (1994). Different use of T cell receptor transducing modules in twopopulations of gut intraepithelial lymphocytes are related to distinct pathways of T celldifferentiation. J Exp Med 180, 673-679.

Hayday, A., Theodoridis, E., Ramsburg, E., and Shires, J. (2001). Intraepitheliallymphocytes: exploring the Third Way in immunology. Nat Immunol 2, 997-1003.

Heath, W. R., Carbone, F. R., Bertolino, P., Kelly, J., Cose, S., and Miller, J. F.(1995). Expression of two T cell receptor alpha chains on the surface of normalmurine T cells. Eur J Immunol 25, 1617-1623.

Hiromatsu, K., Yoshikai, Y., Matsuzaki, G., Ohga, S., Muramori, K., Matsumoto, K.,Bluestone, J. A., and Nomoto, K. (1992). A protective role of gamma/delta T cells inprimary infection with Listeria monocytogenes in mice. J Exp Med 175, 49-56.

Holub, M., Rossmann, P., Tlaskalova, H., and Vidmarova, H. (1975). Thymusrudiment of the athymic nude mouse. Nature 256, 491-493.

Hsu, C., Whitney, R. A., Jr., and Hansen, C. T. (1975). Thymus-like lymph node innude mice. Nature 257, 681-682.

Huesmann, M., Scott, B., Kisielow, P., and von Boehmer, H. (1991). Kinetics andefficacy of positive selection in the thymus of normal and T cell receptor transgenicmice. Cell 66, 533-540.

Hugo, P., Kappler, J. W., Godfrey, D. I., and Marrack, P. C. (1992). A cell line thatcan induce thymocyte positive selection. Nature 360, 679-682.

Hugo, P., Kappler, J. W., McCormack, J. E., and Marrack, P. (1993). Fibroblasts caninduce thymocyte positive selection in vivo. Proc Natl Acad Sci U S A 90, 10335-10339.

Irving, B. A., Alt, F. W., and Killeen, N. (1998). Thymocyte development in theabsence of pre-T cell receptor extracellular immunoglobulin domains. Science 280,905-908.

Jordan, M. S., Boesteanu, A., Reed, A. J., Petrone, A. L., Holenbeck, A. E., Lerman,M. A., Naji, A., and Caton, A. J. (2001). Thymic selection of CD4+CD25+ regulatory Tcells induced by an agonist self-peptide. Nat Immunol 2, 301-306.

Jordan, M. S., Riley, M. P., von Boehmer, H., and Caton, A. J. (2000). Anergy andsuppression regulate CD4(+) T cell responses to a self peptide. Eur J Immunol 30,136-144.

Kappler, J. W., and Marrack, P. (1978). The role of H-2 linked genes in helper T-cellfunction. IV. Importance of T-cell genotype and host environment in I-region and Irgene expression. J Exp Med 148, 1510-1522.

Kawamoto, H., Ikawa, T., Ohmura, K., Fujimoto, S., and Katsura, Y. (2000). T cellprogenitors emerge earlier than B cell progenitors in the murine fetal liver. Immunity12, 441-450.

References

- 87 -

Kawamoto, H., Ohmura, K., and Katsura, Y. (1997). Direct evidence for thecommitment of hematopoietic stem cells to T, B and myeloid lineages in murine fetalliver. Int Immunol 9, 1011-1019.

Kindred, B. (1979). Nude mice in immunology. Prog Allergy 26, 137-238.

Kirberg, J., Berns, A., and von Boehmer, H. (1997). Peripheral T cell survival requirescontinual ligation of the T cell receptor to major histocompatibility complex-encodedmolecules. J Exp Med 186, 1269-1275.

Kisielow, P., Bluthmann, H., Staerz, U. D., Steinmetz, M., and von Boehmer, H.(1988a). Tolerance in T-cell-receptor transgenic mice involves deletion of nonmatureCD4+8+ thymocytes. Nature 333, 742-746.

Kisielow, P., Teh, H. S., Bluthmann, H., and von Boehmer, H. (1988b). Positiveselection of antigen-specific T cells in thymus by restricting MHC molecules. Nature335, 730-733.

Kyburz, D., Aichele, P., Speiser, D. E., Hengartner, H., Zinkernagel, R. M., andPircher, H. (1993). T cell immunity after a viral infection versus T cell toleranceinduced by soluble viral peptides. Eur J Immunol 23, 1956-1962.

Ladel, C. H., Blum, C., Dreher, A., Reifenberg, K., and Kaufmann, S. H. (1995).Protective role of gamma/delta T cells and alpha/beta T cells in tuberculosis. Eur JImmunol 25, 2877-2881.

Le Lievre, C. S., and Le Douarin, N. M. (1975). Mesenchymal derivatives of theneural crest: analysis of chimaeric quail and chick embryos. J Embryol Exp Morphol34, 125-154.

Leist, T. P., Cobbold, S. P., Waldmann, H., Aguet, M., and Zinkernagel, R. M. (1987).Functional analysis of T lymphocyte subsets in antiviral host defense. J Immunol 138,2278-2281.

Lemischka, I. R., Raulet, D. H., and Mulligan, R. C. (1986). Developmental potentialand dynamic behavior of hematopoietic stem cells. Cell 45, 917-927.

Levelt, C. N., de Jong, Y. P., Mizoguchi, E., O'Farrelly, C., Bhan, A. K., Tonegawa,S., Terhorst, C., and Simpson, S. J. (1999). High- and low-affinity single-peptide/MHC ligands have distinct effects on the development of mucosalCD8alphaalpha and CD8alphabeta T lymphocytes. Proc Natl Acad Sci U S A 96,5628-5633.

Longo, D. L., and Schwartz, R. H. (1980). T-cell specificity for H-2 and Ir genephenotype correlates with the phenotype of thymic antigen-presenting cells. Nature287, 44-46.

MacDonald, H. R., Radtke, F., and Wilson, A. (2001). T cell fate specification andalphabeta/gammadelta lineage commitment. Curr Opin Immunol 13, 219-224.

Markowitz, J. S., Auchincloss, H., Jr., Grusby, M. J., and Glimcher, L. H. (1993).Class II-positive hematopoietic cells cannot mediate positive selection of CD4+ Tlymphocytes in class II-deficient mice. Proc Natl Acad Sci U S A 90, 2779-2783.

Martin, S., and Bevan, M. J. (1997). Antigen-specific and nonspecific deletion ofimmature cortical thymocytes caused by antigen injection. Eur J Immunol 27, 2726-2736.

References

- 88 -

Martinic, M. M., Hengartner, H., and Zinkernagel, R. M. (Manuscript in preparation-a).Influence of MHC class I H-2Db density on selection and survival of H-Y-specificTCR transgenic T cells.

Martinic, M. M., Hengartner, H., and Zinkernagel, R. M. (Manuscript in preparation-b).Selection of the H-Y-specific transgenic TCR in an athymic versus euthymicenvironment.

Martinic, M. M., Rulicke, T., Althage, A., Odermatt, B., Hochli, M., Lamarre, A.,Dumrese, T., Speiser, D. E., Kyburz, D., Hengartner, H., and Zinkernagel, R. M.(2003). Efficient T cell repertoire selection in tetraparental chimeric mice independentof thymic epithelial MHC. Proc Natl Acad Sci U S A 100, 1861-1866.

Matzinger, P. (1993). Why positive selection? Immunol Rev 135, 81-117.

Matzinger, P., and Mirkwood, G. (1978). In a fully H-2 incompatible chimera, T cellsof donor origin can respond to minor histocompatibility antigens in association witheither donor or host H-2 type. J Exp Med 148, 84-92.

McCaren, L. C., Holland, J. J., and Syverton, J. T. (1959). The mammalian cell-virusrelationship: I. Attachment of poliovirus to cultivated cells of primate and non-primateorigin. J Exp Med 109, 475-485.

Merkenschlager, M., Benoist, C., and Mathis, D. (1994). Evidence for a single-nichemodel of positive selection. Proc Natl Acad Sci U S A 91, 11694-11698.

Miller, J. (1961). Immunological function of the thymus. Lancet 2, 748-749.

Modigliani, Y., Bandeira, A., and Coutinho, A. (1996). A model for developmentallyacquired thymus-dependent tolerance to central and peripheral antigens. ImmunolRev 149, 155-120.

Moller, G. e. (1978). Acquisition of the T cell Repertoire. Immunol Rev 42, 3-270.

Mombaerts, P., Arnoldi, J., Russ, F., Tonegawa, S., and Kaufmann, S. H. (1993).Different roles of alpha beta and gamma delta T cells in immunity against anintracellular bacterial pathogen. Nature 365, 53-56.

Nemazee, D. (2000). Receptor selection in B and T lymphocytes. Annu Rev Immunol18, 19-51.

Ochsenbein, A. F., Pinschewer, D. D., Odermatt, B., Ciurea, A., Hengartner, H., andZinkernagel, R. M. (2000). Correlation of T cell independence of antibody responseswith antigen dose reaching secondary lymphoid organs: implications forsplenectomized patients and vaccine design. J Immunol 164, 6296-6302.

Ohashi, P. S. (2002). T-cell signalling and autoimmunity: molecular mechanisms ofdisease. Nat Rev Immunol 2, 427-438.

Orange, J. S., Salazar-Mather, T. P., Opal, S. M., Spencer, R. L., Miller, A. H.,McEwen, B. S., and Biron, C. A. (1995). Mechanism of interleukin 12-mediatedtoxicities during experimental viral infections: role of tumor necrosis factor andglucocorticoids. J Exp Med 181, 901-914.

Osborne, B., and Miele, L. (1999). Notch and the immune system. Immunity 11, 653-663.

References

- 89 -

Park, S. Y., Arase, H., Wakizaka, K., Hirayama, N., Masaki, S., Sato, S., Ravetch, J.V., and Saito, T. (1995). Differential contribution of the FcR gamma chain to thesurface expression of the T cell receptor among T cells localized in epithelia: analysisof FcR gamma-deficient mice. Eur J Immunol 25, 2107-2110.

Pawlowski, T., Elliott, J. D., Loh, D. Y., and Staerz, U. D. (1993). Positive selection ofT lymphocytes on fibroblasts. Nature 364, 642-645.

Petrie, H. T., Scollay, R., and Shortman, K. (1992). Commitment to the T cellreceptor-alpha beta or -gamma delta lineages can occur just prior to the onset ofCD4 and CD8 expression among immature thymocytes. Eur J Immunol 22, 2185-2188.

Pfeffer, K., Schoel, B., Gulle, H., Kaufmann, S. H., and Wagner, H. (1990). Primaryresponses of human T cells to mycobacteria: a frequent set of gamma/delta T cellsare stimulated by protease-resistant ligands. Eur J Immunol 20, 1175-1179.

Pircher, H., Burki, K., Lang, R., Hengartner, H., and Zinkernagel, R. M. (1989).Tolerance induction in double specific T-cell receptor transgenic mice varies withantigen. Nature 342, 559-561.

Podd, B. S., Aberg, C., Kudla, K. L., Keene, L., Tobias, E., and Camerini, V. (2001).MHC class I allele dosage alters CD8 expression by intestinal intraepitheliallymphocytes. J Immunol 167, 2561-2568.

Poussier, P., and Julius, M. (1994). Thymus independent T cell development andselection in the intestinal epithelium. Annu Rev Immunol 12, 521-553.

Read, S., and Powrie, F. (2001). CD4(+) regulatory T cells. Curr Opin Immunol 13,644-649.

Rocha, B., Vassalli, P., and Guy-Grand, D. (1991). The V beta repertoire of mousegut homodimeric alpha CD8+ intraepithelial T cell receptor alpha/beta + lymphocytesreveals a major extrathymic pathway of T cell differentiation. J Exp Med 173, 483-486.

Rocha, B., and von Boehmer, H. (1991). Peripheral selection of the T cell repertoire.Science 251, 1225-1228.

Rocha, B., von Boehmer, H., and Guy-Grand, D. (1992). Selection of intraepitheliallymphocytes with CD8 alpha/alpha co-receptors by self-antigen in the murine gut.Proc Natl Acad Sci U S A 89, 5336-5340.

Rodewald, H. R., Ogawa, M., Haller, C., Waskow, C., and DiSanto, J. P. (1997). Pro-thymocyte expansion by c-kit and the common cytokine receptor gamma chain isessential for repertoire formation. Immunity 6, 265-272.

Rodewald, H. R., Paul, S., Haller, C., Bluethmann, H., and Blum, C. (2001). Thymusmedulla consisting of epithelial islets each derived from a single progenitor. Nature414, 763-768.

Rodriguez-Barbosa, J. I., Zhao, Y., Zhao, G., Ezquerra, A., and Sykes, M. (2002).Murine CD4 T cells selected in a highly disparate xenogeneic porcine thymus graftdo not show rapid decay in the absence of selecting MHC in the periphery. JImmunol 169, 6697-6710.

References

- 90 -

Sagara, S., Sugaya, K., Tokoro, Y., Tanaka, S., Takano, H., Kodama, H., Nakauchi,H., and Takahama, Y. (1997). B220 expression by T lymphoid progenitor cells inmouse fetal liver. J Immunol 158, 666-676.

Saint-Ruf, C., Panigada, M., Azogui, O., Debey, P., von Boehmer, H., and Grassi, F.(2000). Different initiation of pre-TCR and gammadeltaTCR signalling. Nature 406,524-527.

Saint-Ruf, C., Ungewiss, K., Groettrup, M., Bruno, L., Fehling, H. J., and vonBoehmer, H. (1994). Analysis and expression of a cloned pre-T cell receptor gene.Science 266, 1208-1212.

Saito, H., Kanamori, Y., Takemori, T., Nariuchi, H., Kubota, E., Takahashi-Iwanaga,H., Iwanaga, T., and Ishikawa, H. (1998). Generation of intestinal T cells fromprogenitors residing in gut cryptopatches. Science 280, 275-278.

Schild, H., Mavaddat, N., Litzenberger, C., Ehrich, E. W., Davis, M. M., Bluestone, J.A., Matis, L., Draper, R. K., and Chien, Y. H. (1994). The nature of majorhistocompatibility complex recognition by gamma delta T cells. Cell 76, 29-37.

Sebzda, E., Mariathasan, S., Ohteki, T., Jones, R., Bachmann, M. F., and Ohashi, P.S. (1999). Selection of the T cell repertoire. Annu Rev Immunol 17, 829-874.

Sha, W. C., Nelson, C. A., Newberry, R. D., Kranz, D. M., Russell, J. H., and Loh, D.Y. (1988). Selective expression of an antigen receptor on CD8-bearing Tlymphocytes in transgenic mice. Nature 335, 271-274.

Simpson, E., Scott, D., and Chandler, P. (1997). The male-specific histocompatibilityantigen, H-Y: a history of transplantation, immune response genes, sexdetermination and expression cloning. Annu Rev Immunol 15, 39-61.

Singer, A. (1988). Experimentation and thymic selection. J Immunol 140, 2481-2483.

Singer, A. (2002). New perspectives on a developmental dilemma: the kineticsignaling model and the importance of signal duration for the CD4/CD8 lineagedecision. Curr Opin Immunol 14, 207-215.

Speiser, D. E., Stubi, U., and Zinkernagel, R. M. (1992). Extrathymic positiveselection of alpha beta T-cell precursors in nude mice. Nature 355, 170-172.

Starr, T. K., Jameson, S. C., and Hogquist, K. A. (2002). Positive and Negative T CellSelection. Annu Rev Immunol.

Stefanova, I., Dorfman, J. R., and Germain, R. N. (2002). Self-recognition promotesthe foreign antigen sensitivity of naive T lymphocytes. Nature 420, 429-434.

Stefanova, I., Dorfman, J. R., Tsukamoto, M., and Germain, R. N. (2003). On the roleof self-recognition in T cell responses to foreign antigen. Immunol Rev 191, 97-106.

Surh, C. D., and Sprent, J. (1994). T-cell apoptosis detected in situ during positiveand negative selection in the thymus. Nature 372, 100-103.

Sydora, B. C., Mixter, P. F., Holcombe, H. R., Eghtesady, P., Williams, K., Amaral, M.C., Nel, A., and Kronenberg, M. (1993). Intestinal intraepithelial lymphocytes areactivated and cytolytic but do not proliferate as well as other T cells in response tomitogenic signals. J Immunol 150, 2179-2191.

References

- 91 -

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 gammadelta T cells. Proc Natl Acad Sci U S A 91, 8175-8179.

Teh, H. S., Kishi, H., Scott, B., and Von Boehmer, H. (1989). Deletion of autospecificT cells in T cell receptor (TCR) transgenic mice spares cells with normal TCR levelsand low levels of CD8 molecules. J Exp Med 169, 795-806.

Teh, H. S., Kisielow, P., Scott, B., Kishi, H., Uematsu, Y., Bluthmann, H., and vonBoehmer, H. (1988). Thymic major histocompatibility complex antigens and the alphabeta T-cell receptor determine the CD4/CD8 phenotype of T cells. Nature 335, 229-233.

Terra, R., Labrecque, N., and Perreault, C. (2002). Thymic and extrathymic T celldevelopment pathways follow different rules. J Immunol 169, 684-692.

von Boehmer, H. (1990). Developmental biology of T cells in T cell-receptortransgenic mice. Annu Rev Immunol 8, 531-556.

von Boehmer, H., Aifantis, I., Gounari, F., Azogui, O., Haughn, L., Apostolou, I.,Jaeckel, E., Grassi, F., and Klein, L. (2003). Thymic selection revisited: how essentialis it? Immunol Rev 191, 62-78.

von Boehmer, H., Haas, W., and Jerne, N. K. (1978). Major histocompatibilitycomplex-linked immune-responsiveness is acquired by lymphocytes of low-responder mice differentiating in thymus of high-responder mice. Proc Natl Acad SciU S A 75, 2439-2442.

Vukmanovic, S., Grandea, A. G., 3rd, Faas, S. J., Knowles, B. B., and Bevan, M. J.(1992). Positive selection of T-lymphocytes induced by intrathymic injection of athymic epithelial cell line. Nature 359, 729-732.

Wagner, H., Rollinghoff, M., Rodt, H., and Thierfelder, S. (1980). T cell-mediatedcytotoxic immune responsiveness of chimeric mice bearing a thymus graft fullyallogeneic to the graft of lymphoid stem cells. Eur J Immunol 10, 521-525.

Wang, S. D., Huang, K. J., Lin, Y. S., and Lei, H. Y. (1994). Sepsis-inducedapoptosis of the thymocytes in mice. J Immunol 152, 5014-5021.

Weintraub, B. C., Jackson, M. R., and Hedrick, S. M. (1994). Gamma delta T cellscan recognize nonclassical MHC in the absence of conventional antigenic peptides. JImmunol 153, 3051-3058.

Wekerle, H., and Ketelsen, U. P. (1980). Thymic nurse cells--Ia-bearing epitheliuminvolved in T-lymphocyte differentiation? Nature 283, 402-404.

Wilkinson, R. W., Anderson, G., Owen, J. J., and Jenkinson, E. J. (1995). Positiveselection of thymocytes involves sustained interactions with the thymicmicroenvironment. J Immunol 155, 5234-5240.

Williams, M. A., Adams, A. B., Walsh, M. B., Shirasugi, N., Onami, T. M., Pearson, T.C., Ahmed, R., and Larsen, C. P. (2003). Primary and secondaryimmunocompetence in mixed allogeneic chimeras. J Immunol 170, 2382-2389.

Zerrahn, J., Volkmann, A., Coles, M. C., Held, W., Lemonnier, F. A., and Raulet, D.H. (1999). Class I MHC molecules on hematopoietic cells can support intrathymicpositive selection of T cell receptor transgenic T cells. Proc Natl Acad Sci U S A 96,11470-11475.

References

- 92 -

Zinkernagel, R. M. (1982). Selection of restriction specificities of virus-specificcytotoxic T cells in the thymus: no evidence for a crucial role of antigen-presentingcells. J Exp Med 156, 1842-1847.

Zinkernagel, R. M., and Althage, A. (1999). On the role of thymic epithelium vs. bonemarrow-derived cells in repertoire selection of T cells. Proc Natl Acad Sci U S A 96,8092-8097.

Zinkernagel, R. M., Althage, A., Waterfield, E., Kindred, B., Welsh, R. M., Callahan,G., and Pincetl, P. (1980). Restriction specificities, alloreactivity, and allotoleranceexpressed by T cells from nude mice reconstituted with H-2-compatible or -incompatible thymus grafts. J Exp Med 151, 376-399.

Zinkernagel, R. M., 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.

Curriculum Vitae

- 93 -

10 Curriculum Vitae

Name: Marianne M. A. Martinic

Work address: Institute of Experimental Immunology

Department of Pathology

University Hospital Zurich

Schmelzbergstrasse 12

8091 Zurich, Switzerland

Tel: +41 1 255 29 89

Fax: +41 1 255 44 20

Home address: Haldenbachstrasse 22

8006 Zurich, Switzerland

Tel: +41-79 407 20 83

Date of birth: 29.11.1974

Place of birth: La Paz, Bolivia

Nationality: Bolivian and French

Marital status: Single

Education:

1980-1982 Godehard Grundschule, Göttingen/Germany

1982-1992 Deutsche Schule “Mariscal Braun”, La Paz/Bolivia

1993-1995 Undergraduate study of biochemistry, Tübingen/Germany

1995-1998 Studies in biochemistry, immunology, genetics, neurobiology and physiology,

molecular biology and biophysics (Abt. XAe)

Swiss Federal Institute of Technology (ETH), Zurich/Switzerland

1996-1997 Diploma thesis at the Institute of Experimental Immunology.

Supervisors: Dr. M. van den Broek and Prof. H. Hengartner

Subject: “Establishment of a transgenic mouse model for inducible, tissue-

specific expression of neo-self antigens.”

1998 Start of PhD thesis at the Institute of Experimental Immunology.

Supervisors: Prof. H. Hengartner and Prof. R. M. Zinkernagel

Subject: “Parameters influencing efficient T cell repertoire selection.”

Bibliography

- 95 -

11 Bibliography

Martinic M. M., Rülicke T., Althage A., Odermatt B., Höchli M., Lamarre A., Dumrese

T., Speiser D. E., Kyburz D., Hengartner H., Zinkernagel R. M. (2003). Efficient T cell

repertoire selection in tetraparental chimeric mice independent of thymic epithelial

MHC. Proc Natl Acad Sci USA. 2003 Feb 18;100(4):1861-6.

Martinic M. M., Hengartner H., Zinkernagel R. M. (2003). Influence of MHC class I

H-2Db density on selection and survival of H-Y-specific TCR transgenic T cells.

Manuscript in preparation.

Martinic M. M., Hengartner H., Zinkernagel R. M. (2003). Selection of the H-Y-

specific transgenic TCR in an athymic versus euthymic environment. Manuscript in

preparation.

Macpherson A. J., Martinic M. M., Harris N. (2002). The functions of mucosal T cells

in containing the indigenous commensal flora of the intestine. Cell Mol Life Sci. 2002

Dec;59(12):2088-96. Review.

Hunziker L., Recher M., Ciurea A., Martinic M. M., Odermatt B., Hengartner H.,

Zinkernagel R. M. (2002). Antagonistic variant virus prevents wild-type virus-induced

lethal immunopathology. J Exp Med 2002 Oct 21;196(8):1039-46.

Medana I., Martinic M. M., Wekerle H., Neumann H. (2001). Transection of major

histocompatibility complex class I-induced neuritis by cytotoxic T lymphocytes. Am J

Pathol. 2001 Sep;159(3):809-15.

Ciurea A., Hunziker L., Martinic M. M., Oxenius A., Hengartner H., Zinkernagel R. M.

(2001). CD4+ T-cell-epitope escape mutant virus selected in vivo. Nat. Med. 2001

Jul;7(7):795-800.

Bibliography

- 96 -

Medana I. M., Gallimore A., Oxenius A., Martinic M. M., Wekerle H., Neumann H.

(2000). MHC class I-restricted killing of neurons by virus-specific CD8+ T

lymphocytes is effected through the Fas/FasL, but not the perforin pathway. Eur J

Immunol. 2000 Dec;30(12):3623-33.

Oxenius A., Martinic M. M., Hengartner H., Zinkernagel R. M. (1999). CpG-

containing oligonucleotides are efficient adjuvants for induction of protective antiviral

immune responses with T-cell peptide vaccines. J Virol. 1999 May;73(5):4120-6.

Acknowledgements

- 97 -

12 Danke, Merci, Thank You, Gracias

Zuallererst möchte ich mich ganz herzlich bei Hans Hengartner und Rolf Zinkernagel

für die Unterstützung und Anregungen während der gesamten Diss-Zeit bedanken.

Ich möchte mich insbesondere auch für die Freiheit bedanken, die ich bei den

einzelnen Projekten, beim Ausprobieren neuer Methoden und bei meiner eigenen

Meinungsbildung bezüglich bestimmter Themen jeweils hatte. Zuletzt möchte ich

mich nochmals speziell bei Rolf Zinkernagel für seine Geduld bei unseren ewigen

Diskussionen bedanken, die mich zwar manchmal zur Weissglut gebracht haben,

dafür aber doch immer sehr hilfreich und bereichernd waren.

J’aimerai aussi envoyer un très grand merci à Alain Lamarre, qui m’a toujours

soutenue et aidée pendant ces cinq dernières années - tu étais même d’accord de

discuter avec moi le sujet de T cell selection et ça en étant un 100% B cell type :-).

Ich möchte mich ganz fest bei Edit Horvath für ihre grosse Hilfe bei den

verschiedensten assays und ihrer immer anhaltenden guten Laune bedanken, sowie

für die schönen und sehr leckeren Scrabble-Abende zusammen mit Karin, Kathrin

und Co.

A big, big, big thank you to Kathy McCoy, Nathalie Oetiker, Veronika Pochanke,

Nicola Harris, Chiara Nembrini and Régine Dayer from the G43-lab for all their

support during the whole time inside and outside of the lab. I really enjoyed our

dinner-partys, snowboard-weekends, Christmas-celebrations and coffee- and lunch-

breaks talking about everything and nothing resulting in a perfect relaxing and

energy-re-loading atmosphere. In addition, I would like to thank Alana Althage and

Kathy McCoy for their constant support during the first year of my PhD-thesis, which

really helped me a lot.

Ganz besonders möchte ich mich beim gesamten Labor für die immer vorhandene

Unterstützung bedanken. Bei Elisabeth Hörhager, Simone Schmaderer und Daniela

Schuppisser möchte ich mich recht herzlich für die Beantwortung aller möglichen und

unmöglichen und vor allem dringenden Fragen sowie für die moralische

Acknowledgements

- 98 -

Unterstützung während der letzten paar Monate bedanken. Bei allen Laborantinnen

sowie Ljiljana Milojevic, Ioannis Chantziaras und Aytac Altuncevahir bedanke ich

mich für die perfekte Organisation unseres Labors, die es jedem ermöglicht, innert

kürzester Zeit die unmöglichsten Sachen wiederzufinden. Un gran agradecimiento va

al “feo, peludo y gordo” Lars Hangartner por su paciencia y gigante ayuda con todos

mis problemas de computación. Ein grosser Dank gilt allen Doktoranden für den

guten Zusammenhalt und für die chaotische, aber eben doch gemütliche Atmosphäre

im PhD-Raum. Zu guter letzt möchte ich mich noch bei Maries van den Broek für ihre

fortwährende Unterstützung und aufschlussreichen Diskussionen während der

gesamten Zeit hier bedanken.

Ein grosser Dank geht an Bernhard Odermatt und an sein gesamtes Histologie-Team

für die hervorragenden histologischen Schnitte. Ganz speziell möchte ich Bernhard

Odermatt für die sehr geduldige Einführung in die Thymus-Histologie und für das

Beantworten all meiner Fragen bedanken.

Ich möchte mich ganz herzlich bei Norbert Wey für seine ausserordentliche Geduld

und Hilfsbereitschaft beim Erklären und Vorführen aller möglichen Mikroskopier-

Computer-Techniken bedanken sowie für die enorme Hilfe beim Zusammenstellen

der Figuren.

Matthias Höchli vom Elektronischen Mikroskopie-Labor möchte ich danken für seine

grosse Unterstützung beim Erklären und Auswerten der Fluoreszenzmikroskopie-

Daten.

Ein grosser Dank gebührt auch Thomas Rülicke, dem ich mein ganzes Wissen über

die Herstellung der tetraparentalen Aggregationschimären zu verdanken habe.

Zum Schluss möchte ich meinen Freunden ausserhalb des Labors für ihre immer

anhaltende Unterstützung während der gesamten Diss-Zeit bedanken. Ganz speziell

möchte ich mich bei Familie Bock dafür bedanken, dass ich mich bei ihnen immer

daheim fühlen durfte und somit jedes Mal nach einem Wochenende in Tübingen

wieder völlig entspannt und voller frischer Energien nach Zürich zurückfahren konnte.

Hier in Zürich gilt mein ganz, ganz grosser Dank Sabine Bösch, die alle meine “Diss-

Acknowledgements

- 99 -

Launen“ miterleben durfte, und es doch jedes Mal wieder geschafft hat, mich zum

lachen zu bringen und das Leben in der Haldenbachstrasse wunderschön zu

gestalten.

Finalmente quisiera agradecer a mi hermana, Julie Martinic, por el lindo e

interesante tiempo que pasamos juntas en Zürich y por su constante apoyo como

también por sus consejos durante todo este tiempo.

Am allermeisten aber möchte ich mich ganz, ganz herzlich bei meinen Eltern

bedanken, dafür, dass sie immer, wirklich immer, für mich da waren und sind, für ihre

permanente Unterstützung während meiner gesamten Zeit hier in Europa und vor

allem dafür, dass es sie gibt.