SEARCH

SEARCH BY CITATION

Keywords:

  • Eph;
  • Ephrin;
  • T cell differentiation;
  • Thymic epithelial cell;
  • Thymus

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements

The Eph and ephrin families are involved in numerous developmental processes. Recently, an increasing body of evidence has related these families with some aspects of T cell development. In the present study, we show that the addition of either EphB2-Fc or ephrinB1-Fc fusion proteins to fetal thymus organ cultures established from 17-day-old fetal mice decreases the numbers of both double-positive (CD4+CD8+) and single-positive (both CD4+CD8 and CD4CD8+) thymocytes, in correlation with increased apoptosis. By using reaggregate thymus organ cultures formed by fetal thymic epithelial cells (TEC) and CD4+CD8+ thymocytes, we have also demonstrated that ephrinB1-Fc proteins are able to disorganize the three-dimensional epithelial network that in vivo supports the T cell maturation, and to alter the thymocyte interactions. In addition, in an in vitro model, Eph/ephrinB-Fc treatment also decreases the formation of cell conjugates by CD4+CD8+ thymocytes and TEC as well as the TCR-dependent signaling between both cell types. Finally, immobilized EphB2-Fc and ephrinB1-Fc modulate the anti-CD3 antibody-induced apoptosis of CD4+CD8+ thymocytes in a process dependent on concentration. These results therefore support a role for Eph/ephrinB in the processes of development and selection of thymocytes as well as in the establishment of the three-dimensional organization of TEC.

Abbreviations:
7-AAD:

7-amino actinomycin D

DN:

double negative

DP:

double positive

FTOC:

fetal thymus organ culture

RTOC:

reaggregate thymus organ culture

SP:

single positive

TEC:

thymic epithelial cell

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements

During thymocyte development, thymic stromal cells sequentially provide signals to T cell precursors that drive their differentiation to mature T cell 1. The process results in movement of the developing thymocytes through the different thymic compartments as well as events of attachment-detachment processes to/from the different stromal cells. Both chemokines and integrins have previously been pointed out to be involved in these cell-to-cell interactions in the thymus 2, 3, but attractive and repulsive phenomena are also concerned 3 and the nature of the molecules governing them remains to be conclusively known. These processes of cell positioning, cell movement and cell-to-cell interactions parallel to cell differentiation are regulated by signaling pathways, including the Eph family of tyrosine kinase receptors and their ligands, the ephrins, common or equivalent in different developmental systems. Ephs and ephrins constitute a complex system involved in regulating cell-to-cell and cell-to-extracellular matrix attachments-detachments and in chemokine-driven cell migration in different cell types 4, including thymocytes 5. Furthermore, Eph-ephrin activation, directly or by crosstalk with other surface receptors 6, can lead to cell division 79, programmed cell death 7, 10, 11 or cell differentiation 12. The fact that the Eph family constitutes the largest family of known tyrosine kinase receptors together with the promiscuity of Eph-ephrin binding, although presenting different ligand-receptor affinities, makes this family of proteins a plastic system for multiple possible interactions. Furthermore, although in many cases the system seems to be redundant, the different affinities and patterns of expression suggest a certain specialization and the possibility of regulating a wide spectrum of cell functions 13, 14.

Most Eph family members have been detected in the thymus, exhibiting an overlapping expression of receptors and ligands 15, 16. Moreover, the expression of the different members of the family seems to be regulated, and several Ephs and ephrins appear on the same thymic cell type and at the same thymocyte developmental stage 17, 18, indicating a wide plasticity that allows different combinations of signals. We have previously demonstrated a general role for the EphA subfamily in thymocyte development 16 and a more specific role for EphA4 in the thymic epithelial network organization which indirectly affected the T cell development 19. Recent evidence of regulation of anti-CD3 antibody-induced apoptosis has also been reported 20. The role of the EphB subfamily in T cell development and function is largely restricted to T cell costimulation (reviewed in 21), and both Eph families have also been reported to be involved in the regulation of lymphocyte chemotaxis 5, 22.

Here, we describe that EphB signaling contributes to T cell development through the regulation of both attachments and three-dimensional arrangements of thymocytes and thymic epithelial cells (TEC) and TCR-mediated responses. Therefore, Eph-ephrin signaling emerges as an important mechanism for governing cell-to-cell interactions within the thymus as well as some of the signals that are produced in such interactions.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements

EphrinB1-Fc and EphB2-Fc treatment affects thymocyte development and survival in fetal thymus organ culture

In order to determine whether ephrinB1 and its receptors could have a role in thymocyte development, and in particular in the double positive (DP; CD4+CD8+) to single positive (SP; CD4+CD8 or CD4CD8+) transition, 17-day-post-coitum fetal thymic lobes were cultured for 6 days in the presence of either soluble 10 µg/mL ephrinB1-Fc, EphB2-Fc, or human IgG-Fc as a control. At 17 days post coitum, only double-negative (DN; CD4CD8), immature CD3CD4CD8+ and DP thymocytes were present in the ex vivo thymus (data not shown). After 6 days in culture, lower numbers of thymocytes were recovered from both ephrinB1-Fc- and EphB2-Fc-treated lobes (Fig. 1A). Thymocyte suspensions from cultured lobes were analyzed for CD4, CD8 and TCRαβ expression. In ephrinB1-Fc- and EphB2-Fc-treated lobes, lower numbers of all thymocyte subsets were found (Fig. 1A). In both cases, the decrease in the absolute numbers of thymocytes mainly affected the DP thymocytes whose relative proportion also decreased (Table 1). Although, due to this decrease of the DP thymocyte proportion, the relative proportion of SP CD4+CD8 cells was higher than in the control lobes, lower numbers of total mature TCRαβ high-expressing thymocytes were recovered from the Eph-Fc- or ephrin-Fc-treated lobes (Fig. 1B). However, when the condition was analyzed in relation to the different T cell subpopulations, in the case of CD8+ TCRαβhigh cells, no statistically significant differences were observed. This result is probably due to the variability in cell recovery from fetal thymus organ cultures (FTOC) as well as to the fact that in a minor cell population the differences are difficult to appreciate. Both ephrinB1-Fc and EphB2-Fc treatment might therefore affect thymocyte survival and/or their differentiation.

thumbnail image

Figure 1. EphrinB1-Fc and EphB2-Fc treatments affect thymocyte development and survival in FTOC. At 17 days post coitum, fetal thymus lobes were cultured for 6 days in complete medium or supplemented with 10 µg/mL of either human IgG, ephrinB1-Fc or EphB2-Fc. Thymocyte suspensions were analyzed for CD4, CD8α and TCRαβ expression and for Annexin V staining. Absolute numbers represent the number of events counted by a flow cytometer when a fixed volume of the total cell suspension was analyzed and are therefore proportional to the total counts. Data represent the mean and standard deviation of at least three independent experiments. The significance of the Student's t-test probability is indicated: * p <0.05; ** p <0.01. (A) Numbers of recovered total thymocytes and of thymocyte subpopulations defined by CD4/CD8 expression from control and treated FTOC. (B) Numbers of recovered TCR high-expressing CD4+CD8+ thymocytes and mature TCRhigh SP (CD4+CD8 and CD4CD8+) cells from the same cultures. (C) Percentages of total and CD4/CD8 thymocyte subsets of Annexin V-positive cells within the total 7-AAD-negative population.

Download figure to PowerPoint

Table 1. EphrinB1-Fc and EphB2-Fc treatments affect the proportions of thymocyte subsets in FTOCa)
 DNDPCD4CD8
  1. a) Percentages of the four thymocyte subsets defined by the expression of CD4 and CD8 obtained as described in the Materials and methods section.

Medium12.30 ± 1.0160.29 ± 1.8720.01 ± 0.377.41 ± 0.49
hIgG11.31 ± 0.1860.04 ± 0.4520.28 ± 0.168.42 ± 0.25
EphB2-Fc10.41 ± 1.3552.85 ± 3.3529.09 ± 2.117.64 ± 0.11
EphrinB1-Fc10.51 ± 0.9548.04 ± 0.5933.03 ± 0.128.05 ± 0.48

We evaluated thymocyte apoptosis in both treated and control lobes by Annexin V staining of thymocyte suspensions, in combination with the expression of CD4 and CD8 markers. Dead cells were excluded by 7-amino actinomycin D (7-AAD) staining. A higher percentage of apoptotic thymocytes was found in ephrinB1-Fc- and EphB2-Fc-treated lobes than in control lobes. All thymocyte subsets suffered higher apoptosis when thymic lobes were cultured in the presence of either ephrinB1 or EphB2 (Fig. 1C). These results indicate that ephrinB1-EphB interaction affects thymocyte survival, directly or indirectly.

EphrinB1-Fc inhibits thymocyte-epithelium organization in reaggregate thymus organ cultures

The above-described results therefore indicated that the abrogation of EphB-ephrinB interactions compromised the DP-to-SP transition and thymocyte survival. In turn, DP cell survival and development are dependent on their interactions with stromal cells, mainly TEC. Under physiological conditions thymocyte-epithelium interactions are complex, including the regulated migration and the attachment-detachment processes of thymocytes along a well-structured epithelial network. In order to evaluate the possible involvement of Eph/ephrinB in these processes, we tested thymocyte-TEC interactions in a three-dimensional multicellular model. Reaggregate thymus organ cultures (RTOC) constituted by fetal TEC and isolated DP thymocytes in a 1 : 5 ratio were cultured in the presence of ephrinB1-Fc or human IgG-Fc as a control. After 24 h, the cell reaggregates were fixed and stained for cytokeratin and with Hoechst 33342, a nuclear staining. Control RTOC showed a compact structure in which thymocytes were packed by epithelial cell processes that formed a continuous epithelial meshwork. On the contrary, in the ephrinB1-Fc treated RTOC, the epithelial cells were rounded and unable to form a network housing the DP thymocytes. Furthermore, the keratin filaments, as revealed by using cytokeratin staining, were not organized within the TEC (Fig. 2). Thus, in the presence of ephrinB1-Fc, thymocytes and TEC are unable to organize the typical, three-dimensional epithelial network housing the developing thymocytes.

thumbnail image

Figure 2. EphrinB1-Fc inhibits thymocyte-epithelium organization in RTOC. Fetal stromal cells and DP thymocytes from MHC-deficient mice were mixed in a 1 : 5 ratio, pelleted, and cultured as RTOC in the presence of either human IgG as a control or ephrinB1-Fc. After 24 h, the cell reaggregates were fixed and stained for cytokeratin and with Hoechst 33342. Images were taken with a HCPl APO ×20/0.7 oil objective and digitally zoomed in a TCS SP2 Leica confocal microscope. A proper three-dimensional epithelial network housing thymocytes is formed in control RTOC (arrows). In ephrinB1-Fc-treated RTOC, such an organization is not found. Epithelial cells appear rounded without cell processes, and dotted instead of filamentous cytokeratin staining is observed (see detail).

Download figure to PowerPoint

EphrinB1-Fc inhibits DP thymocyte-TEC interactions and modifies thymocyte activation

The Eph receptor family role in determining both cell-to-cell adhesion and interactions is well known in other systems, and ephrinB1 has been reported to be involved in modulating TCR signaling 17, 23, 24. Thus, we evaluated whether EphB2-ephrinB1 signaling regulated both the interaction of DP thymocytes with stromal cells and/or the establishment of a functional TCR-dependent signaling. For this purpose, we used a model of thymocyte-fetal TEC conjugates described previously 25. First, the incidence of Eph/ephrin signaling on thymocyte-TEC conjugate formation was evaluated. In this respect, PKH26-stained DP thymocytes from MHC-deficient mice and CFSE-stained TEC were pre-incubated with either soluble ephrinB1-Fc, EphB2-Fc or human IgG and mixed (1 : 1) in a cell suspension. Conjugates were quantified as DP events by flow cytometry analysis at several time points of incubation in the presence of either ephrinB1-Fc, EphB2-Fc or human IgG.

Control human IgG-treated thymocyte-TEC conjugate formation describes a curve with a maximum reached at 30 min, after which the number of conjugates gradually decreases. Conjugates formed in the presence of ephrinB1-Fc presented a similar temporal behavior, but the frequency of them was lower along the studied period, being around 50% of control values at the maximum point (Fig. 3A). A similar effect could be observed when conjugates were formed in the presence of EphB2-Fc. Therefore, conjugate formation is significantly modulated by Eph-ephrin signaling.

thumbnail image

Figure 3. EphrinB1-Fc and EphB2-Fc inhibit DP thymocyte-TEC interactions and modify thymocyte activation. (A) PKH26-stained double-positive thymocytes from MHC-deficient mice and CFSE-stained TEC were pre-incubated with soluble ephrinB1-Fc, EphB2-Fc or human IgG and mixed (1 : 1) in a cell suspension. Data represent the meant percentage of formed conjugates from three independent experiments, at several time points of incubation in the presence of 10 µg/mL of ephrinB1-Fc, EphB2-Fc or human IgG. (B) Unstained DP-TEC conjugates were formed as above for 30 min and immunostained with anti-pTyr antibody. Conjugates with pTyr accumulation at the interaction surface (positive event) and without such accumulation (negative event) were manually counted under a confocal microscope. Bars represent the mean number of positive and negative events, from three independent experiments with 60 conjugates each, when conjugates were formed in the presence of 10 µg/mL human IgG (control) or ephrinB1-Fc. The significance of the Student's t-test probability is indicated: * p <0.05.

Download figure to PowerPoint

As previously suggested by other authors 26, we first confirmed the presence of both EphB2 and ephrinB1 in this complex. Conjugates were stained with either anti-ephrinB1 or anti-EphB2 antibodies. In the conjugates, both proteins were mainly located at the interaction surface of DP thymocytes and TEC (Fig. 4). Only occasionally they occurred in other locations (data not shown).

thumbnail image

Figure 4. EphrinB1 and EphB2 are located at the interacting surface in DP thymocyte-TEC conjugates. DP thymocyte-TEC conjugates formed for 30 min were immunostained for ephrinB1 or EphB2. Accumulation of both ephrinB1 and EphB2 (arrows) could be mainly found at the interacting surface.

Download figure to PowerPoint

DP thymocyte-TEC interaction forms a multimolecular signaling complex, known as the immunological synapse, which leads to a TCR-dependent activation of signaling pathways involving tyrosine phosphorylation 25. Thus, we evaluated whether ephrinB1-Fc treatment could affect the induced signaling response by analyzing tyrosine phosphorylation in DP thymocyte-TEC conjugates. We found that these conjugates showed significantly less phospho-tyrosine accumulation (45 ± 2.1/60 positive events) at the interaction surface between the two cell types than the control ones (56 ± 0.7/60 positive events) (Fig. 3B). Therefore, ephrinB1-Fc treatment prevented the thymocyte-TEC association and decreased the ability to induce a signaling response once the conjugate is formed.

Both ephrinB1-Fc and EphB2-Fc modulate anti-CD3 antibody-induced apoptosis

The above results indicate that EphB2 and ephrinB1, both expressed on thymocytes and TEC (17, 18 and our own not shown results), affect thymocyte apoptosis and are involved in both DP-TEC interactions and TCR-dependent signaling, which in turn affect also thymocyte survival. Also, one of the main factors governing thymocyte survival is the TCR selection, which is dependent on TCR signaling itself and has been described to be also modulated by EphB signaling 24. Accordingly, we analyzed anti-CD3 antibody-induced apoptosis of DP thymocytes after being costimulated or not with either ephrinB1-Fc or EphB2-Fc fusion proteins.

Total thymocytes from adult thymus were seeded on 5 or 10 µg/mL phase-immobilized anti-CD3 antibody with different concentrations of either co-immobilized ephrinB1-Fc or EphB2-Fc or on immobilized ephrin/Eph-Fc alone. After 24 h, thymocytes were recovered, stained for CD4, CD8α, Annexin V and 7-AAD and analyzed by flow cytometry. Results showed that Eph or ephrin stimulation alone, i.e. in the absence of anti-CD3 antibody stimulation, had little or no effect on the spontaneous apoptosis of thymocytes. When thymocytes were costimulated with anti-CD3 antibody and low concentrations of either ephrinB1-Fc or EphB2-Fc, anti-CD3 antibody-induced apoptosis of DP thymocytes was enhanced by the presence of the recombinant receptor or its ligand (Fig. 5). At higher concentrations, this costimulatory effect was reduced, and at the highest tested concentration of EphB2-Fc, apoptosis decreased below the level of anti-CD3 antibody stimulation alone. However, in the case of ephrinB1-Fc costimulation, these values remained higher than those found when DP cells were signaled only with anti-CD3 antibody without either EphB2-Fc or ephrinB1-Fc treatment. Therefore, both EphB2-Fc and ephrinB1-Fc modulated the anti-CD3 antibody-induced apoptosis in a dose-dependent manner.

thumbnail image

Figure 5. EphrinB1-Fc and EphB2-Fc modulate anti-CD3 antibody-induced apoptosis of thymocytes. Total thymocytes (500 000) from adult CD1 mouse thymus were seeded on 5 µg/mL (antiCD3=5) or 10 µg/mL (antiCD3=10) phase-immobilized anti-CD3 mAb with different concentrations of either co-immobilized ephrinB1-Fc or EphB2-Fc or on immobilized ephrin/Eph-Fc alone (antiCD3=0). After 24 h, thymocytes were recovered and stained for CD4, CD8α, Annexin V and 7-AAD. Data represent the meant percentage of Annexin-positive cells within the 7-AAD-negative DP thymocyte subset from two independent experiments conducted in replicate.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements

Different experimental approaches have suggested a role for Ephs and ephrins in T cell development 16, 17 as well as in the process of T cell selection which occurs within the thymus 17, 20. Here, we show that soluble ephrinB1-Fc or EphB2-Fc proteins added during 6 days to FTOC established from 17-day-old fetal thymus lobes cause a decrease in the yield of mature SP thymocytes, although their relative proportions increase, mainly due to the decrease in the percentage of DP thymocytes. These decreased numbers of thymocytes, as well as the altered proportions of different thymocyte subsets, correlate with the finding of increased apoptosis in the treated FTOC. This apoptosis is not restricted to the DP cell population, whose number is particularly decreased probably due to variations in cell cycle or increased apoptosis in earlier developmental stages. In this respect, in the CD8+ population, which includes SP TCRαβhigh CD8+CD4 thymocytes and also a proliferating immature cell population between the DN and DP compartments, the ephrinB1-Fc treatment induces the highest increase of apoptotic cells, but the decrease of the total numbers of CD8+ cells is relatively moderate, presumably due to increased cell proliferation. In agreement with these results, recently Yu and colleagues 17 observed that ephrinB1-Fc (but not ephrinB2-Fc or ephrinB3-Fc) treatment of FTOC results in decreased cellularity, which correlated with a significant increase in cell proliferation and therefore with significantly increased apoptosis. We have previously demonstrated that EphA-Fc and ephrinA1-Fc fusion proteins supplied to 16-day-old rat FTOC significantly reduced the yielded numbers of cells and increased the proportions of apoptotic cells 16. In other cell types, Eph/ephrin has been claimed to be important for regulating in vivo9, 10 and in vitro11, 27 cell survival. Therefore, our finding of increased apoptosis in ephrinB1/EphB2-Fc treated FTOC is not surprising.

Thymocyte development and survival is dependent on thymocyte-stromal cell interactions, mainly thymocyte-TEC. TEC provide thymocytes with different developmental signals as they move along the different stromal microenvironments. The definition of when and which cell-cell interactions must be given is a crucial aspect of thymocyte development. Our results demonstrate a role for EphB2/ephrinB1 in these processes. They indicate that ephrinB1-Fc proteins are capable of disorganizing the thymic three-dimensional epithelial meshwork that in vivo supports T cell maturation, and of altering the thymocyte distribution in RTOC formed by fetal thymic epithelial cells and isolated DP thymocytes. A similar thymic epithelial disorganization has been reported in mice deficient for different Ephs. EphA4-deficient mice that show a hypocellular thymus with a decreased proportion of DP thymocytes and increased numbers of apoptotic cells, as observed in our FTOC after EphB2-Fc or ephrinB1-Fc treatment, also exhibit an altered development of TEC that results in an extreme reduction of the thymic cortex and densely packed epithelial cells 19. Furthermore, mice deficient in different EphB show profound alterations in both the differentiation of TEC precursors and the organization of the epithelial network. Briefly, EphB-deficient thymi showed lower cellularity, disarrangement of the epithelial network, altered cortex-medulla distribution and thymocyte-epithelium disposition, as well as morphologically and phenotypically altered TEC (Fig. 6), (manuscript in preparation). In other epithelial tissues, Ephs (largely EphB2 and EphB3) regulate different aspects of their biology, including relative positioning of differentiating and proliferating cells, cell cycle progression of precursors, tumorigenesis, etc.14.

thumbnail image

Figure 6. EphB2-null, EphB3-null and EphB2/B3 double-mutant mice show altered thymic epithelium organization. Thymi from EphB2 mutants, EphB3 mutants and EphB2/B3 double mutants were processed for semi-thin sectioning and optical microscopy (OM), electron microscopy (EM) and immunostaining over cryosections (IS). Optical microscopy sections show general disarrangement of the thymic histology, with decreased numbers of thymocytes and abnormal thymocyte-epithelium disposition in mutant mice. By electron microscopy, the decreased cellularity was confirmed, as well as the presence of numerous degenerated TEC (stars). Keratin5 (green) and keratin8 (red) double immunostaining shows an altered epithelial network arrangement in mutant mice, with no-keratin-expressing areas (arrows), abnormal distribution of cortical and medullary areas and a higher proportion of double-stained (K5+K8+) TEC (yellow). Moreover, TEC are unable to organize a normal, continuous three-dimensional network.

Download figure to PowerPoint

As known, numerous data support a role for Eph receptors in the dynamics of cellular protrusions and cell migration by modifying the cytoskeletal organization 14. Thus, they regulate changes in cell shape in numerous tissues. During encounters between Eph- and ephrin-expressing non-neuronal cells, retractions have been observed in the receptor-expressing cells 28, 29. In our RTOC, the blockade of EphB signaling by soluble ephrinB1-Fc results in rounding of the TEC, disappearance of cell processes and disorganization of the cellular cytoskeleton network. In this same aspect, EphB forward signaling has been pointed out to be involved in the morphological changes that occur during dendritic spine morphogenesis in hippocampal neurons 30. In our model, it remains to be elucidated, however, whether the blockade of Eph-ephrinB interactions directly affects the behavior of TEC or whether the observed changes are indirectly mediated through the thymocytes, or whether both processes are involved. In any case, our experiments using RTOC are instructive on the presumptive role played by Ephs/ephrins in the organization of the three-dimensional epithelial meshwork of the thymus, essential for the normal maturation of T cell progenitors. Indeed, thymocyte-TEC interactions are important for many processes occurring within the thymus. Thus, migration of developing T cell progenitors through the different thymus compartments requires chemokine-dependent chemotaxis as well as attachments/detachments to/from the thymic epithelial meshwork 31. Different ephrins have been demonstrated for inhibiting T cell chemotaxis induced by SDF-1α 5, 32, and, as mentioned above, the forces regulating cell attraction/repulsion are largely governed by Ephs/ephrins 14.

Going further with the involvement of Ephs/ephrins in thymocyte-TEC interaction that allows thymocyte selection, we evaluated the effects of both ephrinB1-Fc and EphB2-Fc on the formation of DP thymocyte-TEC conjugates. This is a well-known experimental procedure dependent on TCR-peptide-MHC interaction and which implies the polymerization of the actin cytoskeleton at the point of epithelial cell contact and the formation of an immunological synapse 25. The disruption of the actin cytoskeleton in thymocytes abrogates the conjugate formation, indicating that the cytoskeleton dynamics, one of the major functions played by Ephs/ephrins in the cells, is important for the formation of stable conjugates 23, 26, 33. On the other hand, conjugate formation is dependent not only on actin cytoskeleton regulation within the DP thymocytes but also within the TEC 25. Another key process for the formation of immunological synapses is cell adhesion 34, whose regulation is remarkably mediated by Eph/ephrin signaling 4, 14. In this respect, our results demonstrate that ephrinB1-Fc and EphB2-Fc treatment prevents thymocyte-TEC association and alters TCR signaling. EphrinB1-EphB signaling is therefore important to stabilize the conjugates and for the induction of a phospho-tyrosine signaling cascade. Supporting this interpretation, both EphB2 and ephrinB1 mainly locate at the interaction surface between the two cells that form the conjugate. Other authors have also observed the migration of EphB to concentrate in lipid rafts in peripheral T lymphocytes when the TCR was strongly engaged 23, 33, 35, as well as the costimulatory properties of Eph signaling in TCR-dependent T cell activation (see review by Wu et al.21) and the involvement of Ephs in thymocyte selection 17, 20, 23.

Thus, Eph/ephrin signaling regulates DP thymocyte-TEC interaction and TCR-dependent signaling. Accordingly, we examined whether the effects of EphB2/ephrinB1-Fc fusion proteins on thymocyte apoptosis were the result of the lack of developmental and survival signals due to improper thymocyte-TEC interactions, a direct incidence of Eph/ephrin signaling directly on thymocyte survival or also reflected their involvement in thymocyte selection through the regulation of TCR signaling.

In agreement with other results 17, neither ephrinB1-Fc nor EphB2-Fc stimulation alone affected thymocyte survival, but both molecules were able to modulate anti-CD3 antibody-induced apoptosis of thymocytes. Recently, Freywald and colleagues 20 demonstrated that the stimulation of murine thymocytes with ephrinA1-Fc reduced the TCR-mediated induction of apoptosis in DP thymocytes. Previously, this same group had observed that cross-linking of EphB6 with anti-ephrinB1 antibody protected CD3-positive thymocytes from anti-CD3 antibody-induced apoptosis 24. Yu and colleagues 17 have also found that ephrinB1-Fc costimulation protected from anti-CD3 antibody-induced apoptosis of thymocytes. Our results demonstrate that the modulation of anti-CD3 antibody-induced apoptosis of DP thymocytes by EphB2-Fc or ephrinB1-Fc is dependent on the density of the molecules on the plate surface. Thus, growing concentrations of either immobilized EphB2-Fc or ephrinB1-Fc result in a gradual increase, faster in the case of EphB2-Fc than after ephrinB1-Fc treatment, of the proportions of anti-CD3 antibody-induced apoptotic DP thymocytes, to decrease later.

These apparently contradictory results could be explained by the capacity of Ephs/ephrins to modulate the strength of TCR signaling and thymocyte selection 21. In agreement with this hypothesis, Freywald and colleagues 20 recently pointed out that the levels of ephrinA expressed on thymic antigen-presenting cells could modulate the intensity of TCR engagement required for thymocyte selection, resulting in positive or negative selection. A dependence of cell responses upon Eph/ephrin stimulation on concentration, including the regulation of cell attachment/detachment, has also been demonstrated before 36.

These results suggest therefore that both EphB2-Fc and ephrinB1-Fc can modify the strength of TCR signaling and result in increased numbers of negatively selected DP thymocytes. Indirect evidence supports this hypothesis. It has been observed that ephrinB2, the main ligand of EphB6 and EphB4 18, but also EphB2 ligand, mainly distributes through the thymic cortex, which is occupied by DP thymocytes, SP cells in transition to the medulla and cortical TEC, and, to a lesser extent, through the cortico-medullary border 33. EphrinB1 is largely expressed on both DP thymocytes and CD8 cells (∼30%), but also on DN cells (13.5%) and finally CD4 cells (6.2%) 17. By means of RT-PCR, we have confirmed the presence of EphB2, EphB3, ephrinB1 and ephrinB2 in the main adult thymocyte populations (DN, DP, SP) isolated by flow cell sorting and in the TEC, as well as in total thymocytes and TEC from 15-day-post-coitum fetal thymus (data not shown). On the other hand, our results demonstrate that not only forward signaling, through EphB2, but also reverse signaling, mediated by ephrinB, can modify the strength of TCR signaling and result in altered proportions of apoptotic DP thymocytes. Eph/ephrin co-expression is a mechanism to quantitatively modulate the Eph/ephrin signaling balance that finally results in qualitatively different responses (reviewed in 37). It is therefore possible that different Eph-ephrin signaling combinations and/or Eph/ephrin/TCR signaling combinations could modulate the TCR response, even turning increased sensitivity to cell death into protection. On the other hand, Eph and ephrin signaling can regulate common or similar downstream pathways, despite their having very different cytoplasmic domains 6, 38. How they exert similar or even opposite responses in different cell types remains unresolved 6, 14.

Eph/ephrinB signaling could converge with TCR signaling in regulating intracellular pathways affecting both cytoskeleton dynamics and cell adhesion. Although molecular mechanisms involved in these processes are not totally known, it has been pointed out that Eph could interfere with TCR signaling through regulation of the activation of small GTPases 38, largely those of the Rho family, and then control MAPK activation, integrin and cytoskeleton organization 3941. Vav1, a guanine nucleotide exchange factor for members of the Rho family of GTPases, is necessary for positive selection 42 and regulates TCR-dependent activation of LFA-1 as well as some actin cytoskeleton-dependent events at the immunological synapse 43. In addition, negative regulation of MAPK pathway activation has been proposed as a mechanism by which EphA receptors could modulate TCR signaling and affect thymocyte selection 20. Published data about the effects of Ephs on MAPK activation are, however, controversial 5, 24, 26 and could be dependent on the cell type 21.

Our results imply that Eph/ephrin signaling conditions thymocyte-TEC interactions, which is not only relevant for thymocyte behavior but also for the normal biology of TEC and for the mutual responses induced through these interactions between the two cell types. Also, they can modulate some of the signals induced in these interactions, as is the TCR signaling.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements

Animals

CD1 (The Jackson Laboratory, Bar Harbor, ME) and MHC-deficient (Taconic, Germantown, NY) mice were bred and maintained at our animal facilities. Fetuses were obtained from CD1 timed mating. The day of appearance of a vaginal plug was designated as day 0 of gestation.

Fetal thymus organ culture

Fetal thymus lobes (17 days old) were cultured on polycarbonate membranes (Millipore Ibérica, Madrid, Spain) in RPMI 1640 (Invitrogen SA, Barcelona, Spain) cell culture medium 5% FCS containing either 10 µg/mL soluble EphB2-Fc or ephrinB1-Fc fusion proteins (R&D Systems, Oxon, UK) or purified human IgG (Sigma-Aldrich Quimica, Madrid, Spain). After 6 days, the lobes were processed for phenotypic analysis by flow cytometry.

Flow cytometry

Cell suspensions were stained for 20 min in PBS 1% FCS with specific mAb against either CD4, CD8α (Caltag Laboratories, Invitrogen) or TCRαβ (BD Biosciences, Erembodegem, Belgium) labeled with PE, FITC, Tricolor or APC, or with anti-Annexin V antibody (Annexin-V-FLUOS; Roche Diagnostics, Basel, Switzerland). After washing in PBS, stained cells were resuspended in PBS and analyzed in a FACSCalibur or LSR (BD Biosciences) at the Microscopy and Cytometry Center, Complutense University of Madrid.

Preparation of thymocytes and epithelial cells

Where indicated, DP CD4+CD8+ thymocytes at a preselection stage were obtained from MHC-deficient mice. TEC were obtained by disaggregating 15-day CD1 thymus lobes depleted of lymphoid cells by 2-deoxyguanosine treatment, as described 44. Typically, the resultant suspension mainly consisted of epithelial cells 45.

Reaggregate thymus organ cultures

Thymocytes and TEC were mixed together by centrifugation at a ratio of 5 : 1 and the pellet was transferred to a 0.8-µm filter in organ culture 44, 46. After 24 h, the reaggregates were fixed in 2% formaldehyde for 45 min, washed and stained with FITC-conjugated anti-pan cytokeratin mAb (clone c-11; Sigma-Aldrich) in PBS 0.1% Tween-20 for 30 min at RT, washed, counterstained with Hoechst 33342 (Sigma-Aldrich-Quimica) and analyzed in a TCS SP2 Leica confocal microscope at the Microscopy and Cytometry Center, Complutense University of Madrid.

Formation and analysis of thymocyte-epithelial cell conjugates

Thymocyte-epithelial cell conjugates were formed and evaluated by flow cytometry as described 25. In some experiments, unstained conjugates were formed and, after 10 min of fixation in 2% formaldehyde, washed and stained with anti-phospho-tyrosine (anti-pTyr) antibody (Santa Cruz Biotechnology, Santa Cruz, CA) or rat anti-mouse ephrinB1 mAb or goat anti-mouse EphB2 antibody (R&D Systems, Minneapolis, MN), in PBS 0.1% Tween-20 for 30 min. Staining was revealed with anti-mouse IgG-Alexa 488 (Molecular Probes, Invitrogen), anti-rat IgG-Texas Red or anti-goat IgG-Texas Red antibody (Jackson Laboratories), respectively, and when indicated a second incubation with FITC-conjugated anti-pan cytokeratin mAb (clone c-11; Sigma-Aldrich) was performed.. Immunostained conjugates were seeded on 6-well Teflon-printed slides (EMS, Washington, USA) and imaged under a TCS SP2 Leica confocal microscope at the Microscopy and Cytometry Center, Complutense University of Madrid.

Activation-induced cell death in thymocytes

Flat-bottom culture plates (BD Biosciences) were coated with 100 µL purified anti-CD3 mAb overnight at 4°C. For costimulation assays, the anti-CD3 mAb was co-immobilized with either ephrinB1-Fc or EphB2-Fc. After a first overnight incubation with anti-CD3 antibody, plates were rinsed with PBS three times and a second incubation with either ephrinB1-Fc or EphB2-Fc for 2 h at 37°C was performed.

Total thymocytes from adult CD1 mouse thymus were seeded on coated plates and incubated in RPMI 10% FCS at 37°C in a 5% CO2 humidified atmosphere. After 24 h, cells were washed and incubated with anti-CD4 and anti-CD8α antibodies before staining with Annexin V and 7-AAD.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements

We would like to thank Drs. Simona Rossi and William Jenkinson for their experimental help and advice and Javier Arias for his critical reading of the manuscript and his comments. We also thank the Microscopy and Cytometry Center of the Complutense University of Madrid for the use of their facilities and technical assistance. This work was supported by grants BCM2001–2025 and BFU2004–03132 from the Spanish Ministry of Education and Culture.

  • 1

    WILEY-VCH

  • 2

    WILEY-VCH

  • 3

    WILEY-VCH

  • 4

    WILEY-VCH

  • 5

    WILEY-VCH

  • 6

    WILEY-VCH