The preTCR-dependent DN3 to DP transition requires Notch signaling, is improved by CXCL12 signaling...

Post on 07-May-2023

4 views 0 download

Transcript of The preTCR-dependent DN3 to DP transition requires Notch signaling, is improved by CXCL12 signaling...

The preTCR-dependent DN3 to DP transition requiresNotch signaling, is improved by CXCL12 signalingand is inhibited by IL-7 signaling

Roxane Tussiwand1, Corinne Engdahl1, Nadine Gehre1, Nabil Bosco2,

Rod Ceredig3 and Antonius G. Rolink1

1 Developmental and Molecular Immunology, Department of Biomedicine, University of Basel,

Basel, Switzerland2 Nestle Research Department, Lausanne, Switzerland3 REMEDI, University of Galway, Galway, Ireland

The requirement for Notch signaling during T-cell development has been extensively

studied. Nevertheless, the developmental stage at which it is required and whether

additional signaling pathways are needed are still poorly understood. By using a stromal-

cell-free culture system, we show that sorted double-negative 3 (DN3) thymocytes only

require a Delta-like-4-induced Notch signal to differentiate into double-positive (DP) cells.

This differentiation process is preTCR-a dependent. DN3 cells undergo 4–5 proliferation

cycles, and the addition of the chemokine CXCL12 improves proliferation. IL-7 blocks the

differentiation of DN3 cells to DP cells but not the Notch-induced proliferation of cultured

DN3 cells. The impaired differentiation correlates with an inhibition of Rag-2 up-

regulation. Overall, the in vitro stromal-cell-free culture system presented here also

provides a powerful and unique tool for studying the mechanisms involved in the positive

and negative selection of T cells.

Key words: Delta-like 4 and preTCR . DN3 cells . Double-positive cells . Notch

Introduction

To maintain T lymphopoiesis, progenitor cells from the BM

constantly seed the thymus. In transplantation settings, it has

been shown that different progenitors can home to the thymus

and generate T cells [1–9]. However, under physiological

conditions the role of each of these progenitors is still unclear.

It is generally believed that the so-called thymic seeding

precursor (TSP) still has the capability of generating B cells, NK

cells, DCs and other myeloid cells [2, 5, 6, 10, 11]. Upon

receiving the Notch signal, the differentiation potential towards

the B-cell lineage of the TSP is rapidly lost [2, 10, 11], whereas

other lineage options are still maintained [2, 5, 6, 10, 11].

However, a recent study using an IL-7Ra reporter mouse

indicated that the non-T-cell lineage differentiation of these

progenitors is not, or only very rarely, occurring under

physiologic conditions [12, 13].

Within the thymus, the earliest thymocytes are characterized

by the absence of CD4 and CD8 expression and are therefore

called double-negative (DN) cells. Based on the expression of

CD25, CD44 and CD117, DN cells can be subdivided into four

subpopulations. DN1 cells express high cell surface levels of CD44

and CD117, and are negative for CD25, whereas their DN2

progeny express all three markers [14–16]. In vitro, DN1 and

DN2 cells still possess the capacity to differentiate into NK cells,

DCs and other myeloid cells [5, 6, 11, 17] suggesting that they

are not yet restricted to the T-cell lineage. T-cell commitment is

achieved at the DN3 stage. However, the mechanisms operatingCorrespondence: Prof. Antonius G. Rolinke-mail: antonius.rolink@unibas.ch

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380 DOI 10.1002/eji.201141824 Leukocyte signaling 3371

during the transition from DN2 to DN3 cells that guide this

commitment are still poorly understood. DN3 cells express high

levels of CD25 and low levels of both CD44 and CD117 [14–16].

It is at this stage of development that the rearrangement of the

TCR-b chain locus is completed, and the cells are selected

for a productive TCR-b chain rearrangement, also known as the

b-selection checkpoint [18–20]. DN3 cells that have rearranged

their TCR-b chain locus unproductively can either differentiate

into g/d T cells or will otherwise undergo apoptosis. DN3

cells carrying a productive TCR-b chain will express it at the

cell surface as a complex together with the preTCR-a (preTa)

and CD3 thereby forming the so-called preTCR [18–20].

Most preTCRs expressing DN3 will differentiate into the a/bT-cell lineage. However, a small proportion might still give

rise to g/d T cells as suggested by the finding that about 20% of

g/d T cells contain a TCR-b chain capable of forming a preTCR

[21–24].

The vast majority of DN3 cells that express a TCR-b poly-

peptide will undergo a preTCR-mediated proliferative expansion

[18–20] and will differentiate into DN4 cells. DN4 cells have lost

the expression of CD25, CD44 and CD117 and are the direct

precursors of the immature single-positive (ISP) CD81 cells

[14–16]. ISPs will then differentiate into CD41 and CD81-

expressing double-positive (DP) cells. At this stage of T-cell

development, the TCR-a chain locus will be rearranged and those

cells that are able to express a functional ab TCR will undergo

positive and negative selection.

Several signaling pathways playing a crucial role in this

complex T-cell developmental program have been identified, and

of these, Notch signaling is possibly the most crucial. Thus, it was

shown that conditional inactivation of Notch 1 [25] or one of its

ligands, Delta-like 4 (DL4) [26, 27], resulted in a complete

abrogation of T-cell development in the thymus. Moreover,

IL-7–IL-7R [28, 29] as well as the stem cell factor (SCF)–c-kit

[30–33] signaling axes have been shown to be crucial for efficient

T-cell development. WNT and Sonic hedgehog have also been

implicated in the development of T cells [[34, 35] and references

herein] and additional signaling pathways might also be neces-

sary for efficient T-cell generation. Recently, it was suggested that

the transition from a Notch-dependent to Notch-independent

stage of T-cell development, mediated by either Notch ligand

Delta-like 1 (DL1) or DL4, occurred at the CD81 ISP stage [36].

However, the stages during T-cell development at which stage

these signaling pathways are required, and how they synergize,

are still not clear.

Several years ago Zuniga-Pflucker and colleagues

generated an OP9 stromal cell line expressing DL1 [36, 37]. This

stromal cell line is able to promote differentiation into the

T-cell lineage of early hematopoietic precursors [36, 37].

This culture system has been of great utility in our understanding

of the early stages of thymocyte development. However,

in such cultures, the density of DL1 distribution and the kinetics

of Notch signal delivery are hard to control and additional

signals, either cell surface or soluble delivered by OP9 cells,

remain unknown.

Therefore, to identify the minimal requirements necessary for

T-cell commitment and differentiation, we developed a stromal-

cell-free culture system.

Results

WT DN3 cells only require DL4-Notch signaling fortheir transition into DP thymocytes

Using the OP9-DL1 stromal cell culture system, we previously

showed that wild-type DN3 differentiated into DP thymocytes

without the necessity of adding other growth or differentiation

factors [4]. However, this study did not rule out the possibility

that other signaling components provided by the OP9 stromal

cells also contributed to this differentiation. In fact, differentia-

tion of DN3 to DP cells on OP9-DL1 stromal cells occurs via CD41

and not as under physiological conditions through a CD81-ISP

[14–16].

Moreover, it was recently shown that DL4 and not DL1 is the

physiologic thymic Notch ligand responsible for T-cell develop-

ment [26, 27]. Therefore, to address the minimal requirements

for the transition from DN3 to DP, we established a stromal cell-

free Notch-signaling culture system. DL4-human IgG1-Fc (DL4-

Fc) fusion protein (see the Materials and methods) was coupled to

a tissue culture plate pre-coated with 10 mg/mL monoclonal

mouse anti-human IgG1-Fc antibody (Huf 5.4).

In a first set of experiments, sorted DN3 cells were added to

plates pre-coated with 10 mg/mL anti-human-Fc and then coupled

with various amounts of DL4-Fc, as indicated (Fig. 1A). On

day 5, the number of viable cells for the various conditions was

determined by Trypan Blue exclusion. No viable cells could be

recovered in the absence of DL4-Fc (Fig. 1A). However,

viable cells were readily detectable in all the wells coupled with

DL4-Fc and their numbers correlated with the concentration of

DL4-Fc protein (Fig. 1A), peaking at about 1000 ng/mL DL4-Fc.

The maximal number of viable cells recovered at this dose was

three times the input. FACS analysis revealed that practically all

cells expressed CD4 and CD8 irrespective of DL4-Fc concentra-

tion, as shown for 10 ng and 1mg (Fig. 1B and C). Thus, the

differentiation of DN3 cells into DP cells only requires Notch

signaling and its efficiency, in terms of cell recovery, is DL4 dose

dependent.

To determine the kinetics as well as the phenotype of inter-

mediate stages in this DN3 to DP transition, sorted DN3 cells were

plated in the presence of 2 mg/mL coupled DL4-Fc. Expression of

CD4 and CD8b was determined from day 2 to day 5, while TCR-

g/d and a/b on day 5. As shown in Fig. 2A (upper left) about 25%

cells were CD8b ISP and about 35% DP after 2 days of culture. By

day 3, the percentage of CD8b ISP decreased to about 12% while

the DP population increased to 70% (Fig. 2A, upper right). By

days 4 and 5, the vast majority (over 85%) of cells were found to

be DP (Fig. 2A). Mirroring in vivo thymic T-cell development,

initial differentiation of DN3 cells into DPs on recombinant DL4

occurred via a CD8 single-positive (ISP) stage.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380Roxane Tussiwand et al.3372

Around 15% of the DN cells recovered on day 5 expressed a

TCR-g/d and none expressed a TCR-a/b (Fig. 2B, upper left

panel). Around 30% of DPs on day 5 showed a decrease in CD8

expression (Fig. 2A, lower right). However, all CD4high/CD8high

and all CD4high/CD8low cells expressed identical, homogenous

levels of TCR-a/b (Fig. 2B, upper middle and right panels)

comparable to those found on ex vivo isolated DPs (Fig. 2B, lower

middle panel) and far less then expressed by ex vivo isolated

single-positive CD4 cells (Fig. 2B, lower right panel). These

findings suggest that CD4high/CD8low cells had down-modulated

CD8 expression and had not been positively selected. The finding

that both CD4high/CD8high and CD4high/CD8low cells could not be

activated by cell-surface CD3 cross-linking (data not shown)

supports this conclusion.

The DN3 to DN transition is preTCR dependent and isimproved by CXCL12

The fact that the yield of DP cells on day 5 of culture was about

three times the input of DN3 cells (Fig. 1) indicates that at least

some of the cells must have proliferated. To determine the extent

of proliferation and the percentage of cells that underwent

division, sorted DN3 cells were CFSE labeled. On day 5 of culture,

CFSE signal of viable cells was 15- to 30-fold reduced compared

with the starting population (Fig. 3A). This finding indicates that

the original DN3 cells that survived the 5-day culture and

differentiated into DPs had divided 4–5 times. Considering the

recovery, this experiment suggests that about 10% of the original

DN3 cells proliferated and differentiated into DPs.

That few cells expressing the initial level of CFSE were seen

after 5 days (Fig. 3A) presumably means that non-dividing DN3

cells failed to survive in these cultures. To determine the number

of non-dividing DN3 cells more precisely, we sorted DN3 from

Bcl2 transgenic mice, which should be protected from apoptosis,

irrespective of their differentiation and/or proliferation. On day 5

of culture, practically all cells from Bcl2 transgenic mice were

alive indicating that the Bcl2 transgene indeed protects them

from undergoing apoptosis. The number of cells that could be

recovered was about 4.5 times the input. The CFSE profile

revealed that around 35% of the cells had divided once and the

Figure 1. Proliferation and differentiation of DN3 cells into DP cells isDL4-Fc dose dependent. (A) 2� 105 DN3 cells were cultured with theindicated concentrations of DL4-Fc. The number of live cells wasdetermined on day 5 of culture. The plotted numbers represent themean values and standard deviation as percentage of input cells oftriplicate cultures. (B and C) Flow cytometric analysis of CD4 andCD8 expression from DN3 cells cultured for 5 days in the presence of(B) 10 ng/mL and (C) 1 mg/mL DL4-Fc respectively. The data arerepresentative of three individual experiments.

Figure 2. Kinetics of DN3 to DP transition upon Notch signaling.(A) Sorted DN3 cells were cultured in the presence of 2mg/mL DL4-Fc.CD8b and CD4 expression was determined by flow cytometry on days 2,3, 4 and 5. (B) Analysis of TCR-g/d (bold) and TCR-ab (broken line) ondouble-negative (DN), CD41/CD8high (DP) and CD4high/CD8low (CD4)cells obtained on day 5 of culture is shown in the upper panel. Analysisof TCR-g/d (bold line) and TCR-a/b (broken line) on double-negative(DN), CD41/CD81 DPs and CD41 single-positive cells ex vivo isolatedfrom the thymus is shown in the lower panel. The data arerepresentative of three individual experiments.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380 Leukocyte signaling 3373

remainder between two and maximally five times (Fig. 3B

and C). The small shift to the left of the undivided cell peak in

Fig. 3B is commonly ascribed to a leakage of CFSE upon cultur-

ing. Overall, these findings largely confirmed the findings of WT

cells in that about 10% of them divide 4–5 times upon culture on

plate-bound DL4. As previously shown [38], some DN cells

acquire CD4 and CD8 expression without division (Fig. 3C,

middle panel). Moreover, CD4 and TCR-b expression analysis

revealed that only those cells that divided 4–5 times carried

high levels of CD4 and low levels of TCR-ab on their surface

(Fig. 3C).

During the transition from DN3 to DP cells, the preTCR

complex comprising preTa and TCR-b protein plays a crucial role

in selecting and amplifying by proliferation of those DN cells that

have undergone a complete productive TCR-b rearrangement

[18–20]. Cytoplasmic TCR-b expression can first be detected in

DN3 cells and has been associated with higher CD27 levels [39].

To test whether CD27high and CD27low DN3 cells differ in their

capacity to proliferate and differentiate into DPs on plate-bound

DL4, we sorted and cultured them for 5 days. Cell recovery

from cultures initially seeded with sorted CD27high cells was

10- to 15-fold the input, whereas it was only 25% the input in

those seeded with CD27low cells (Fig. 4A). Moreover, the analysis

of CD4 and CD8 expression revealed that most (90%) CD27high

cells had differentiated into DPs, whereas only 40% of the ini-

tially CD27low cells expressed CD4 and CD8 (Fig. 4B and C).

Taken together, these results indicate that mainly those DN3 cells

that expressed a TCR-b protein proliferated and differentiated

into DPs.

Based on the above results and the recently published data

[20], it was reasonable to assume that this differentiation and

proliferation are mediated by preTCR in the presence of Notch

signaling. To directly test this hypothesis, DN3 cells were sorted

from preTa-deficient mice and cultured for 5 days on plate-bound

DL4. At the end of the culture, practically no live cells (o10% of

input) could be recovered. Moreover, FACS analysis revealed that

very few surviving cells expressed CD4 and TCR-b on their cell

surface (Fig. 5A). In marked contrast, about 35% recovered cells

Figure 3. Proliferation analysis of DN3 cells upon Notch signaling. (A) Sorted WT and (B) Bcl2 transgenic DN3 cells were CFSE labeled and culturedfor 5 days in the presence of 2 mg/mL DL4-Fc. CFSE dilution was determined on day 0 (thin line) and day 5 (bold line) by flow cytometry.(C) Expression of CFSE versus CD4 (middle panel) and CFSE versus TCR-ab (right panel) on DN3 cells from Bcl2 transgenic mice cultured for 5 dayson plate-bound DL4-Fc. The data are representative of three individual experiments.

Figure 4. CD27high but not CD27low cells efficiently proliferate anddifferentiate into DP cells upon Notch signaling. (A) 2� 105 CD27high

(white bar) and CD27low (black bar) DN3 cells were sorted and culturedin the presence of 2 mg/mL DL4-Fc. On day 5 of culture, the number oflive cells was determined. Indicated are mean values and standarddeviation of triplicate cultures. (B and C) CD4 and CD8 expression by(B) CD27high and (C) CD27low cultured DN3 cells on day 5 in the presenceof 2 mg/mL DL4-Fc. The data are representative of three individualexperiments.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380Roxane Tussiwand et al.3374

expressed TCR-gd (Fig. 5B). Thus, proliferation and differentia-

tion of DN3 cells into DPs on plate-bound DL4 are preTCR

dependent. The finding that sorted DN3 cells from Rag-2-defi-

cient mice did not even survive for 3 days in the present culture

system further supports this conclusion.

It was recently suggested that PI3 kinase signaling mediated

by CXCR4 is also involved in the DN3 to DP transition [40–42].

To evaluate the contribution of CXCL12 during this differentia-

tion step in the present culture system, we sorted DN3 cells and

cultured them in the presence of 10 nM CXCL12. On day 5 of

culture, over 90% of the cells that could be recovered were DP

irrespective of the absence (Fig. 6A) or presence (Fig. 6B) of

CXCL12. However, the number of cells from CXCL12-containing

cultures was about four times higher (Fig. 6C). To distinguish

between the increases in the number of divisions undertaken by

surviving cells from an increased proportion of initially plated

cells proliferating, DN3 cells were sorted, labeled with CFSE and

analyzed on day 5 of culture. As shown in Fig. 6D, in the presence

of CXCL12 the original DN3 cells had divided two times more

indicating that CXCL12 improves the preTCR-dependent Notch-

signaling-induced proliferation and differentiation.

IL-7 blocks the Notch-induced DN3 differentiation butnot their proliferation

It is well established that IL-7 plays a crucial role in early DN

T-cell development [28, 29]. Moreover, it was recently shown

that IL-7 plays a redundant role in post-selected thymocyte

maturation and emigration [43]. However, its contribution

during DN3 to DP transition has so far not been investigated.

In fact, we and others have suggested that IL-7 might even inhibit

the DN3 to DP transition [4, 44]. To validate our hypothesis, we

decided to test the effect of IL-7 on the DN3 to DP transition in

this culture system.

Thus, sorted DN3 cells were cultured on plate-bound DL4 in

the presence or absence of 100 U/mL IL-7. FACS analysis

performed on day 5 revealed that in the presence of IL-7 53%

remained DN, 33% expressed low levels of CD8 and only 11%

became DP (Fig. 7A, left dot plot). This phenotype contrasts with

those cultured in the absence of IL-7 (Fig. 7A, right dot plot)

where 77% of the cells were DPs. Moreover, in the presence of

IL-7, about 30% of DN cells expressed a TCR-gd (Fig. 7B, left

histogram overlay) while all DP cells expressed low levels of

TCR-ab (Fig. 7B, right histogram overlay). The CD8low cells were

TCR-g/d and a/b negative indicating that they represented ISP

cells (Fig. 7B, middle histogram overlay). Thus, IL-7 blocks the

Notch-induced differentiation of DN3 cells to DP cells. To test

whether IL-7 also interferes with Notch-induced proliferation,

sorted DN3 cells were labeled with CFSE. The CFSE signal

obtained on day 5 of culture revealed that proliferation in the

presence or absence of IL-7 was similar (Fig. 7C versus Fig. 2A).

Moreover, it was found that cells expressing a TCR-gd had made

one division less than those expressing a TCR-ab (Fig. 7D).

Our experiments with CD27high and CD27low DN3 cells indi-

cated that only TCR-b protein-expressing CD27high cells prolif-

erated and differentiated in this culture system (Fig. 4). To test

whether cells that had undergone Notch-induced proliferation in

the presence of IL-7 were also enriched for TCR-b protein

expression, cells obtained after day 5 of culture were analyzed for

cell-surface TCR-g/d and cytoplasmic TCR-b expression. Of the

TCR-g/d-negative cells most (over 90%, 78.3/85.7) had TCR-b

Figure 5. DN3 cells from preTa-deficient mice demonstrate poordevelopment to DP cells but differentiate into gd T cells upon Notchsignaling. Flow cytometric analysis of (A) CD41 versus TCR-ab and(B) CD41 versus TCR-g/d expression on DN3 cells from preTa-deficientmice cultured for 5 days in the presence of 2mg/mL DL4-Fc. The dataare representative of three individual experiments.

Figure 6. CXCL12 improves proliferation and differentiation uponNotch signaling. Flow cytometric analysis of CD4 and CD8 expressionby DN3 cells cultured for 5 days on 2 mg/mL plate-bound DL4-Fc(A) in the absence or (B) in the presence of 10 nM CXCL12. (C) Thenumber of live cells recovered from 2� 105 seeded DN3 cells culturedfor 5 days on plates pre-coated with 2mg/mL DL4-Fc in the absence(black bar) or presence (white bar) of 10 nM CXCL12. Numbersrepresent mean values and standard deviation of triplicate cultures.(D) Flow cytometric analysis of CFSE dilution of DN3 cells grown onplates pre-coated with 2 mg/mL of DL4-Fc for 5 days in the absence (thinline) or presence of 10 nM CXCL12 (bold line). The filled histogramrepresents the CFSE signal on day 0. The data are representative ofthree individual experiments.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380 Leukocyte signaling 3375

protein in their cytoplasm, whereas only about 30% (4.2/14.3)

TCR-g/d cells expressed cytoplasmic TCR-b protein (Fig. 7E).

Thus, in either the presence or absence of IL-7, the Notch-induced

proliferation of DN3 cells seems to be mainly restricted to

those DN3 cells that had undergone a productive TCR-brearrangement.

A hallmark of the DP stage of thymocyte development is the

rearrangement and subsequent expression of the TCR-a gene.

Therefore, it could be envisaged that the block in differentiation

observed in the presence of IL-7 might be due to inefficient TCR-arearrangement. To test, at least in part, this hypothesis we

determined Rag-2 expression in differentiating DN3 cells in the

presence or absence of IL-7 by qPCR. At 48 and 96 h after

initiation of cultures, Rag-2 levels were 4–5 times lower in DN3

cells cultured on plate-bound DL4 in the presence versus absence

of IL-7 (Table 1). This finding suggests that the partial block in

Rag2 up-regulation by IL-7 might result in an inhibition of

TCR-a rearrangements which in turn might account for the fact

that DN3 cells cultured on plate-bound DL4 with IL-7 inefficiently

differentiate into DP cells.

Discussion

In this study, we have analyzed some of the parameters involved

in the differentiation of mouse thymocytes from the DN3 to DP

stage in a stromal-cell-free culture system. In this culture system,

soluble DL4-human IgG1-Fc fusion protein is immobilized on

plastic through a pre-coated mouse anti-human Fc monoclonal

antibody.

Figure 7. IL-7 blocks the Notch-induced DN3 to DP transition. (A) Flow cytometric analysis of CD4 and CD8 expression on sorted DN3 cells culturedfor 5 days on plates pre-coated with 2mg/mL DL4-Fc in the presence of 100 U/mL IL-7 (left) or in the absence of IL-7 (right). (B) Flow cytometricanalysis of TCR-g/d (bold line) and TCR-ab (broken line) expression on double-negative cells (DN), CD81 cells (CD8) and DP cells obtained fromsorted DN3 cells cultured for 5 days on plates pre-coated with 2 mg/mL DL4-Fc in the presence of 100 U/mL IL-7. (C) CFSE dilution analysis of sortedDN3 cells cultured for 5 days on plates pre-coated with 2 mg/mL DL4-Fc in the presence of 100 U/mL IL-7 (bold line). Thin line histogram representsthe CFSE signal on day 0. (D) CFSE dilution analysis of TCR-g/d (bold line) and TCR-ab (broken line) expressing cells obtained from DN3 cellscultured for 5 days on plates pre-coated with 2 mg/mL DL4-Fc in the presence of 100 U/mL IL-7. Thin line histogram represents the CFSE signalon day 0. (E) Flow cytometric analysis of cell-surface TCR-g/d and cytoplasmic TCR-b expression of sorted DN3 cells cultured for 5 days on platespre-coated with 2 mg/mL DL4-Fc in the presence of 100 U/mL IL-7. The data are representative of three individual experiments.

Table 1. Relative Rag-2 expression

IL-7 Time in culture

48 h 96 h

Absent 5.0 44.0

Present 1.1 11.1

Ct values were normalized against HGPRT and fold increase wascalculated as 2DDCt . Results are compared with the arbitrary unit of 1.0assigned to ex vivo sorted DN3 cells. Mean values of duplicates of twoindependent experiments.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380Roxane Tussiwand et al.3376

The results obtained show that signaling via the Notch ligand

DL4 alone is sufficient to drive the differentiation of DN3 to DP

cells, defining the minimal requirement for this differentiation

step during T-cell commitment. This process involves multiple

rounds of cell divisions and requires the expression of the preTacomplex. The addition of IL-7 prevents differentiation and is

characterized by diminished Rag-2 expression, presumably

reflecting reduced TCR-a rearrangements. Signaling through the

PI3-kinase CXCR4 via CXCL12 is not essential for this differ-

entiation step, but its addition results in an increased number of

cell divisions, leading to a four-fold increase of cell recovery.

Taken together, we were able to show in a highly controlled in

vitro system the contribution of the different signaling pathways

involved during the transition from DN to DP. Moreover, the

identified intermediates reflect, by means of surface markers, the

physiologic counterparts found in the thymus. In addition, this

novel culture system will also provide a unique tool for studying

the mechanisms involved in positive and negative selection of

T cells by adding back various thymic epithelial and/or DCs.

In the mouse thymus, differentiation of T cells is a carefully

orchestrated process involving multiple signaling molecules

produced and expressed by non-lymphoid thymic elements as

well as the corresponding receptors on developing thymocytes.

Clearly, Notch, IL-7, SCF, CXCR4 and preTa signaling all interact

with one another. As recently reviewed by Janas and Turner [42]

also chemotaxis is involved in this DN3 to DP differentiation

process with an initial outward migration towards the cortex

followed by an inward migration to the medulla as post-DP

selected SP cells. Simultaneously, differentiation of TCR-gd cells

takes place with cells distributed throughout the cortex. In vitro

culture systems trying to recapitulate this complex differentiation

process were of great utility but have shown some limitations. In

particular, the OP9 stromal cells, incapable of supporting

myelopoiesis and made to express the Notch ligands DL1 or DL4,

were clearly instrumental in revolutionizing the ability to study

T-cell differentiation in vitro [36, 37]. However, such culture

systems are not ideal since the density of Notch ligand expression

cannot be controlled and other signals may be delivered to the

co-cultured cells. For these and other reasons, we decided to

develop the stromal-cell-free culture system described herein.

Initial experiments using the soluble DL4-Fc fusion protein

alone were universally unsuccessful despite the use of different

grades of protein-fixing plastic. This lack of success can now be

explained by the fact that addition of soluble Notch ligand to

cultured cells inhibits T-cell differentiation (data not shown).

Nevertheless, once generated, the mouse anti-human Fc mono-

clonal antibody overcame this problem showing a DL4 dose

dependency of T-cell differentiation (Fig. 1). The initial transition

from DN to DP in vitro, like the in vivo counterpart, occurred via

a CD4low/CD8high (CD8 ISP) intermediate. At later time points,

following the appearance of DP cells some CD4high/CD8low cells

that did not proliferate upon TCR engagement could be recov-

ered. These latter cells most likely represent the post-DP ‘CD4

wannabes’ described by Jameson and Bevan [45] or CD4high

/CD8low ‘intermediate’ cells reviewed by Singer et al. [46] that

have not been TCR selected but have nevertheless down-modu-

lated CD8 expression.

TCR-gd cells appeared as well in these cultures. Their

generation was clearly Notch-dependent and their proportion,

but not number, increased in the absence of preTa signaling. This

result perhaps favors the notion that the expression of preTadiverts cells from a (default) TCR-gd to a TCR-ab fate and that in

preTa KO mice, all T cells develop as TCR-gd cells. The appear-

ance of TCR-gd cells again supports on one hand a close reflection

of this in vitro culture system to physiological thymic develop-

ment and on the other hand shows its flexibility as a tool for

further studies on differentiation and selection.

In the literature, there has been some confusion as to the

relative roles and contribution of preTa, Notch and IL-7 in the

DN3 to DP transition. Moreover, increased survival could rescue

the IL-7R defect [47, 48], suggesting that a delicate equilibrium

exists between all the signaling pathways.

Cell recovery from in vitro cultures is the difference of cell

proliferation and cell death. Using CFSE labeling, we could show

that most recovered cells had proliferated; yet total cell recovery

could not be accounted for proliferation alone. A large proportion

of DN3 cells failed to survive in these in vitro conditions. The use

of DN3 cells from Bcl2 transgenic mice revealed that the loss of

DN3 cells must have taken place in cultures of WT DN3 cells.

The single contribution of each signaling pathway: preTa,

Notch, IL-7 and CXCR4 to the DN3 to DP commitment was

assessed by our stromal-free culture system. The dependence on

preTa expression was tested in two ways, first by sorting DN3

cells into CD27high (TCR-b-positive) or CD27low (TCR-b-negative)

(Fig. 4) and furthermore by using DN3 cells from preTa-deficient

mice (Fig. 5). In both situations, it could be shown that

in the absence of preTa signaling, differentiation from DN3

to DP was severely compromised. Recovery of CD27low cells was

considerably reduced, again confirming that survival (Fig. 3) and

consequent differentiation of DN3 cells are dependent upon a

productive TCR-b rearrangement.

While in vivo, CXCR4 was shown to be required for the

progression of thymocyte differentiation beyond the DN1 stage

[49], a recent report suggested that it also contributed to the

b-selection-dependent transition from DN3 to DP [41, 42]. We

could show that the dependence on the CXCR4 ligand CXCL12

(also known as SDF-1a) was not absolute, but its addition did

result in a four-fold increase in cell recovery.

The role of IL-7 in thymocyte differentiation is complex. At

early (DN1 and DN2) stages, IL-7 clearly plays an essential role in

cell survival and proliferation. IL-7 is also involved in regulating

rearrangement of the TCR-g locus and from such experiments it

was perhaps assumed that IL-7 favored rearrangements of all

TCR gene loci. However, the role of IL-7 in the DN3 to DP tran-

sition was still controversial. We have previously shown that as

during in vitro B-cell development, the presence of IL-7 inhibits

differentiation, by preventing IgL chain rearrangements. Mature

naıve BCR-expressing B cells emerge in culture only upon with-

drawal of IL-7 [50]. Similarly, in stromal-cell-based cultures of

T cells, we have previously shown that the presence of IL-7

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380 Leukocyte signaling 3377

reduced the development of DP cells. This is also seen in fetal

thymus organ cultures where the addition of IL-7 prevents

TCR-ab and favors TCR-gd differentiation. We could show that in

the absence of stromal cells, the presence of IL-7 is neither

altering proliferation (Fig. 7C) nor the proportion of cells

expressing intracellular TCR-b (Fig. 7E), but that fewer DP cells

(Fig. 7A) and more TCR-gd cells (Fig. 7B) are generated. Taken

together with the analysis of Rag-2 transcripts, the simplest

interpretation of these results would be that like for B-cell

development, IL-7 inhibits T-cell differentiation in vitro. Although

the processes of proliferation and differentiation are intimately

linked, there is no universal rule dictating whether proliferation is

necessary for differentiation. Regarding RAG proteins, their level

of phosphorylation, which influences their stability and degra-

dation, is dependent on the proliferative status of the cell and

thereby plays an important role in gene rearrangement.

Normally, in vivo, lymphocytes have to exit cell cycle to

successfully rearrange antigen-specific receptor genes and maybe

the presence of IL-7 prevents cells entering this state.

Collectively, we could specifically address the contribution of

each signaling pathway known to play a role during this crucial

transition from DN3 to DP cells. Moreover, we introduce herein

our stromal-cell-free in vitro culture system, which offers

numerous advantages over stromal-cell-based systems for study-

ing T-cell commitment. In particular, we could show and want to

emphasize that dose, strength and duration of Notch signaling

delivered to cultured cells can be easily controlled. Furthermore,

the absence of a stromal layer allows dissection of the role of

known and potential signaling pathways involved in T-cell

development, in the absence of any other perturbation. Finally,

since there is no evidence of antigen receptor selection, this

culture system offers the possibility of specifically address

fundamental questions on selection mechanisms by adding back

to those cultures factors whether cellular and/or soluble.

Materials and methods

Mice

Female C57BL/6, C57BL/6-PreTa-deficient, C57BL/6-Rag-2-defi-

cient and C57BL/6 mice expressing BCL-2 as transgene under the

control of the H-2K promoter of 5–8 wk of age were used. All

mice were bred and maintained in our animal facility under

specific pathogen-free conditions. All animal experiments were

carried out within institutional guidelines (authorization number

1888).

Recombinant DL4-Fc

The extra-cellular part of mouse DL4 (amino acids 1–521) was

cloned as a fusion protein 50 of the Fc portion of human IgG1

followed by an IRES-GFP. The modified pCR 3 vector, containing

the Fc portion of human IgG1 followed by IRES-GFP, was kindly

provided by Dr. Pascal Schneider (Department of Biochemistry,

University of Lausanne, Switzerland). Upon transfection in

Chinese hamster ovary cells, the cells expressing high GFP levels

were sorted and used for DL4-Fc production. DL4-Fc fusion

protein was purified by affinity chromatography by passing the

supernatant of transfected cells over a protein A-Sepharose

column.

Flow cytometry and cell sorting

FITC-, PE-, APC- or biotin-labeled mAb specific for CD3, CD4,

CD8a, CD8b, CD25, CD27, CD117, TCR-b and TCR-g/d were

purchased from BD Biosciences or e-Biosciences. Staining of the

cells was performed as described [4]. Flow cytometry was

performed using a FACS Calibur (BD Biosciences) and data were

analyzed using the Cell Quest Pro Software (BD Biosciences). For

cell sorting, the FACS Aria (BD Biosciences) was used. Re-analysis

of the sorted cells indicated that in all instances they were over

98% pure. DN thymocytes were prepared as previously described

[16], washed and dead cells removed by centrifugation over

Ficoll-Paque Plus (Amersham Biosciences). From the total DN

thymocytes, DN3 cells were sorted as CD251, CD44– and CD3–.

In some of the experiments thus sorted, DN3 cells were further

subdivided into CD271 and CD27– cells. CFSE labeling was

performed according to standard procedures; briefly, sorted cells

were labeled in the dark at 371C for 5 min.

Cell cultures

In all experiments, 48- or 24-well tissue culture plates from Nunc

(Nunc A/S, Roskilde, Denmark) were used. Wells were pre-

coated overnight or longer with 10 mg/mL anti-human IgG-Fc

(hybridoma Huf-5.4 generated in house) in PBS (0.25 mL per 48

wells and 0.5 mL per 24 wells) at 41C. Thereafter, wells were

washed twice with IMDM supplemented with 5�10�5 M

b-mercaptoethanol, 1 mM glutamine, 0.03% w/v Primatone

(Quest, Naarden, The Netherlands), 100 U/mL penicillin,

100mg/mL streptomycin and 5% heat-inactivated fetal bovine

serum (complete medium) and then incubated with complete

medium containing various amounts of DL4-Fc overnight

(0.25 mL per 48 wells and 0.5 mL per 24 wells) at 41C. Before

culturing, the cells plates were washed twice with complete

medium and 5� 104–2.5� 105/mL sorted DN3 cells were

cultured in complete medium (0.5 mL per 48 wells and 1 mL

per 24 wells). When indicated 10 nM CXCL12 or 100 U/mL IL-7

was added to the culture medium.

Real-time Rag-2 PCR

Total RNA was extracted from sorted ex vivo or cultured cells at

the indicated time points and conditions (as described in cell

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380Roxane Tussiwand et al.3378

culture) using TRI Reagents (MRC, Cincinnati, OH, USA), and

first-strand synthesis was performed with Superscripts RT kit

(Roche) according to manufacturer’s guidelines.

For RAG2 expression, the following primers were used:

50-TGCCAAAATAAGAAAGAGTATTTCAC-30 and 50-GGGACATTTT

TGATTGTGAATAGG-30. Quantification of the transcript was

performed by real-time PCR on an ABI Prism 7000 light cycler

(Applied Biosystems, Zug, Switzerland) using SYBR green PCR

MasterMix (Fermentas). Ct values were normalized against

hypoxanthine guanine phosphoribosyl transferase (HGPRT), and

fold increase was calculated as 2DDCt . Results are compared with

ex vivo sorted DN3 cells.

Acknowledgements: Antonius Rolink is the holder of the chair in

Immunology endowed by F. Hoffman-La Roche Ltd., Basel.

This work was supported by grants from the Swiss National

Science Foundation to Antonius Rolink. Rod Ceredig is

supported by Science Foundation Ireland under grant number

SFI 09/SRC/B1794 and by a Science Foundation Ireland Stoke’s

Professorship.

Conflict of interest: The authors declare no financial or

commercial conflict of interest.

References

1 Kondo, M., Weissman, I. L. and Akashi, K., Identification of clonogenic

common lymphoid progenitors in mouse bone marrow. Cell 1997. 91:

661–672.

2 Benz, C. and Bleul, C. C., A multipotent precursor in the thymus maps to

the branching point of the T versus B lineage decision. J. Exp. Med. 2005.

202: 21–31.

3 Martin, C. H., Aifantis, I., Scimone, M. L., von Andrian, U. H., Reizis, B.,

von Boehmer, H. and Gounari, F., Efficient thymic immigration of B2201

lymphoid-restricted bone marrow cells with T precursor potential. Nat.

Immunol. 2003. 4: 866–873.

4 Balciunaite, G., Ceredig, R., Massa, S. and Rolink, A. G., A B2201CD1171

CD19– hematopoietic progenitor with potent lymphoid and myeloid

developmental potential. Eur. J. Immunol. 2005. 35: 2019–2030.

5 Bell, J. J. and Bhandoola, A., The earliest thymic progenitors for T cells

possess myeloid lineage potential. Nature 2008. 452: 764–767.

6 Wada, H., Masuda, K., Satoh, R., Kakugawa, K., Ikawa, T., Katsura, Y. and

Kawamoto, H., Adult T-cell progenitors retain myeloid potential. Nature

2008. 452: 768–772.

7 Rodewald, H. R., Kretzschmar, K., Takeda, S., Hohl, C. and Dessing, M.,

Identification of pro-thymocytes in murine fetal blood: T lineage

commitment can precede thymus colonization. EMBO J. 1994. 13:

4229–4240.

8 Serwold, T., Ehrlich, L. I. and Weissman, I. L., Reductive isolation from

bone marrow and blood implicates common lymphoid progenitors as the

major source of thymopoiesis. Blood 2009. 113: 807–815.

9 Umland, O., Mwangi, W. N., Anderson, B. M., Walker, J. C. and Petrie,

H. T., The blood contains multiple distinct progenitor populations

with clonogenic B and T lineage potential. J. Immunol. 2007. 178:

4147–4152.

10 Ceredig, R., Bosco, N. and Rolink, A. G., The B lineage potential of thymus

settling progenitors is critically dependent on mouse age. Eur. J. Immunol.

2007. 37: 830–837.

11 Balciunaite, G., Ceredig, R. and Rolink, A. G., The earliest subpopulation

of mouse thymocytes contains potent T, significant macrophage, and

natural killer cell but no B-lymphocyte potential. Blood 2005. 105:

1930–1936.

12 Schlenner, S. M., Madan, V., Busch, K., Tietz, A., Laufle, C., Costa, C.,

Blum,C. et al., Fate mapping reveals separate origins of T cells and

myeloid lineages in the thymus. Immunity 2010. 32: 426–436.

13 Schlenner, S. M. and Rodewald, H. R., Early T cell development and the

pitfalls of potential. Trends Immunol. 2010. 31: 303–310.

14 Godfrey, D. I., Kennedy, J., Mombaerts, P., Tonegawa, S. and Zlotnik, A.,

Onset of TCR-beta gene rearrangement and role of TCR-beta expression

during CD3–CD4–CD8– thymocyte differentiation. J. Immunol. 1994. 152:

4783–4792.

15 Godfrey, D. I., Zlotnik, A. and Suda, T., Phenotypic and functional

characterization of c-kit expression during intrathymic T cell develop-

ment. J. Immunol. 1992. 149: 2281–2285.

16 Ceredig, R. and Rolink, T., A positive look at double-negative thymocytes.

Nat. Rev. Immunol. 2002. 2: 888–897.

17 Rolink, A. G., Balciunaite, G., Demoliere, C. and Ceredig, R., The potential

involvement of Notch signaling in NK cell development. Immunol. Lett.

2006. 107: 50–57.

18 Fehling, H. J., Krotkova, A., Saint-Ruf, C. and von Boehmer, H., Crucial

role of the pre-T-cell receptor alpha gene in development of alpha beta

but not gamma delta T cells. Nature 1995. 375: 795–798.

19 von Boehmer, H., Aifantis, I., Azogui, O., Feinberg, J., Saint-Ruf, C., Zober,

C., Garcia, C. et al., Crucial function of the pre-T-cell receptor

(TCR) in TCR beta selection, TCR beta allelic exclusion and alpha beta

versus gamma delta lineage commitment. Immunol. Rev. 1998. 165:

111–119.

20 Pang, S. S., Berry, R., Chen, Z., Kjer-Nielsen, L., Perugini, M. A., King, G. F.,

Wang, C. et al., The structural basis for autonomous dimerization of the

pre-T-cell antigen receptor. Nature 2010. 467: 844–848.

21 Burtrum, D. B., Kim, S., Dudley, E. C., Hayday, A. C. and Petrie, H. T., TCR

gene recombination and alpha beta-gamma delta lineage divergence:

productive TCR-beta rearrangement is neither exclusive nor preclusive of

gamma delta cell development. J. Immunol. 1996. 157: 4293–4296.

22 Dudley, E. C., Girardi, M., Owen, M. J. and Hayday, A. C., Alpha beta and

gamma delta T cells can share a late common precursor. Curr. Biol. 1995.

5: 659–669.

23 Wilson, A. and MacDonald, H. R., A limited role for beta-selection during

gamma delta T cell development. J. Immunol. 1998. 161: 5851–5854.

24 Bosco, N., Engdahl, C., Benard, A., Rolink, J., Ceredig, R. and Rolink, A. G.,

TCR-beta chains derived from peripheral gammadelta T cells

can take part in alphabeta T-cell development. Eur. J. Immunol. 2008. 38:

3520–3529.

25 Radtke, F., Wilson, A., Stark, G., Bauer, M., van Meerwijk, J., MacDonald,

H. R. and Aguet, M., Deficient T cell fate specification in mice with an

induced inactivation of Notch1. Immunity 1999. 10: 547–558.

26 Hozumi, K., Mailhos, C., Negishi, N., Hirano, K., Yahata, T., Ando, K.,

Zuklys, S. et al., Delta-like 4 is indispensable in thymic environment

specific for T cell development. J. Exp. Med. 2008. 205: 2507–2513.

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380 Leukocyte signaling 3379

27 Koch, U., Fiorini, E., Benedito, R., Besseyrias, V., Schuster-Gossler, K.,

Pierres, M., Manley, F. et al., Delta-like 4 is the essential, nonredundant

ligand for Notch1 during thymic T cell lineage commitment. J. Exp. Med.

2008. 205: 2515–2523.

28 Peschon, J. J., Morrissey, P. J., Grabstein, K. H., Ramsdell, F. J.,

Maraskovsky, E., Gliniak, B. C., Park, L. S. et al., Early lymphocyte

expansion is severely impaired in interleukin 7 receptor-deficient mice.

J. Exp. Med. 1994. 180: 1955–1960.

29 von Freeden-Jeffry, U., Vieira, P., Lucian, L. A., McNeil, T., Burdach, S. E.

and Murray, R., Lymphopenia in interleukin (IL)-7 gene-deleted mice

identifies IL-7 as a nonredundant cytokine. J. Exp. Med. 1995. 181:

1519–1526.

30 Waskow, C., Terszowski, G., Costa, C., Gassmann, M. and Rodewald,

H. R., Rescue of lethal c-KitW/W mice by erythropoietin. Blood 2004. 104:

1688–1695.

31 Waskow, C., Paul, S., Haller, C., Gassmann, M. and Rodewald, H. R.,

Viable c-Kit(W/W) mutants reveal pivotal role for c-kit in the main-

tenance of lymphopoiesis. Immunity 2002. 17: 277–288.

32 Agosti, V., Corbacioglu, S., Ehlers, I., Waskow, C., Sommer, G., Berrozpe,

G., Kissel, H. et al., Critical role for kit-mediated Src kinase but not PI

3-kinase signaling in pro T and pro B cell development. J. Exp. Med. 2004.

199: 867–878.

33 Massa, S., Balciunaite, G., Ceredig, R. and Rolink, A. G., Critical role for

c-kit (CD117) in T cell lineage commitment and early thymocyte

development in vitro. Eur. J. Immunol. 2006. 36: 526–532.

34 Staal, F. J. and Clevers, H. C., WNT signalling and haematopoiesis: a WNT-

WNT situation. Nat. Rev. Immunol. 2005. 5: 21–30.

35 Crompton, T., Outram, S. V. and Hager-Theodorides, A. L., Sonic

hedgehog signalling in T-cell development and activation. Nat. Rev.

Immunol. 2007. 7: 726–735.

36 Schmitt, T. M. and Zuniga-Pflucker, J. C., Induction of T cell development

from hematopoietic progenitor cells by delta-like-1 in vitro. Immunity

2002. 17: 749–756.

37 Schmitt, T. M., de Pooter, R. F., Gronski, M. A., Cho, S. K., Ohashi, P. S. and

Zuniga-Pflucker, J. C., Induction of T cell development and establishment

of T cell competence from embryonic stem cells differentiated in vitro.

Nat. Immunol. 2004. 5: 410–417.

38 Ceredig, R., Sekaly, R. P. and MacDonald, H. R., Differentiation in vitro of

Lyt 21 thymocytes from embryonic Lyt 2– precursors. Nature 1983. 303:

248–250.

39 Taghon, T., Yui, M. A., Pant, R., Diamond, R. A. and Rothenberg, E. V.,

Developmental and molecular characterization of emerging beta- and

gammadelta-selected pre-T cells in the adult mouse thymus. Immunity

2006. 24: 53–64.

40 Ara, T., Itoi, M., Kawabata, K., Egawa, T., Tokoyoda, K., Sugiyama, T.,

Fujii, N. et al., A role of CXC chemokine ligand 12/stromal cell-derived

factor-1/pre-B cell growth stimulating factor and its receptor CXCR4 in

fetal and adult T cell development in vivo. J. Immunol. 2003. 170:

4649–4655.

41 Janas, M. L., Varano, G., Gudmundsson, K., Noda, M., Nagasawa, T. and

Turner, M., Thymic development beyond beta-selection requires phos-

phatidylinositol 3-kinase activation by CXCR4. J. Exp. Med. 2010. 207:

247–261.

42 Janas, M. L. and Turner, M., Stromal cell-derived factor 1alpha and

CXCR4: newly defined requirements for efficient thymic beta-selection.

Trends Immunol. 2010. 31: 370–376.

43 Weinreich, M. A., Jameson, S. C. and Hogquist, K. A., Postselection

thymocyte maturation and emigration are independent of IL-7 and ERK5.

J. Immunol. 2011. 186: 1343–1347.

44 Ikawa, T., Hirose, S., Masuda, K., Kakugawa, K., Satoh, R., Shibano-

Satoh, A., Kominami, et al., An essential developmental checkpoint for

production of the T cell lineage. Science 2010. 329: 93–96.

45 Jameson, S. C. and Bevan, M. J., T-cell selection. Curr. Opin. Immunol. 1998.

10: 214–219.

46 Singer, A., Adoro, S. and Park, J. H., Lineage fate and intense debate:

myths, models and mechanisms of CD4- versus CD8-lineage choice. Nat.

Rev. Immunol. 2008. 8: 788–801.

47 Maraskovsky, E., O’Reilly, L. A., Teepe, M., Corcoran, L. M., Peschon, J. J.

and Strasser, A., Bcl-2 can rescue T lymphocyte development in

interleukin-7 receptor-deficient mice but not in mutant rag-1–/– mice.

Cell 1997. 89: 1011–1019.

48 Akashi, K., Kondo, M., von Freeden-Jeffry, U., Murray, R. and Weissman,

I. L., Bcl-2 rescues T lymphopoiesis in interleukin-7 receptor-deficient

mice. Cell 1997. 89: 1033–1041.

49 Plotkin, J., Prockop, S. E., Lepique, A. and Petrie, H. T., Critical role for

CXCR4 signaling in progenitor localization and T cell differentiation in

the postnatal thymus. J. Immunol. 2003. 171: 4521–4527.

50 Rolink, A., Grawunder, U., Haasner, D., Strasser, A. and Melchers, F.,

Immature surface Ig1B cells can continue to rearrange kappa and

lambda L chain gene loci. J. Exp. Med. 1993. 178: 1263–1270.

Abbreviations: DL1: Delta-like 1 � DL4: Delta-like 4 � DN: double-

negative � DP: double-positive � ISP: immature single-positive � preTa:

preTCR-a

Full correspondence: Prof. Antonius G. Rolink, Developmental and

Molecular Immunology, Department of Biomedicine, University of

Basel, Mattenstrasse 28, 4058-Basel, Switzerland

e-mail: antonius.rolink@unibas.ch

Received: 3/6/2011

Revised: 11/7/2011

Accepted: 9/8/2011

Accepted article online: 23/8/2011

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji-journal.eu

Eur. J. Immunol. 2011. 41: 3371–3380Roxane Tussiwand et al.3380