| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Microscopy and Cytometry Center, Complutense University, Madrid, Spain;
Department of Cell Biology, Faculty of Biology, Complutense University, Madrid, Spain; and
Department of Cell Biology, Faculty of Medicine, Complutense University, Madrid, Spain
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Eph/ephrins are regulators of the morphogenesis, cell positioning, and cell migration through several mechanisms that specifically regulate the restriction of cell movement guiding the cells to their definitive position, the establishment of tissue domains, and/or boundary formation (5, 6). In summary, these molecules define when and where a certain cell type must move, attach, or detach (4, 7, 8) In general, these roles are accomplished by means of regulating cytoskeleton dynamics, cell adhesion, and integrin activity (4, 9), but there is growing evidence that Eph-ephrin activation can modulate gene expression through the cross-activation of other receptors (10) or by the regulation of distinct intracellular pathways (11). There are examples of their participation in the regulation of cell proliferation (12), cell differentiation (13), or cell survival (14, 15).
As in other morphogenetic and developmental processes, thymus development during fetal life takes place through a combination of positional information and cell fate definition of the different types of participating cells. Moreover, the developmental process continues after birth because T cell development implies the entrance of bone marrow cell precursors into the thymus that differentiate in a mature and compartmented thymic stroma. The process involves cell movement through different thymic compartments as well as processes of attachment and detachment while developmental and cell fate signals are given to the developing thymocytes by the thymic stromal cells (16, 17).
Eph/ephrins are thus candidate molecules to regulate some of the processes that occur in the thymus as they do in other developing systems. In fact, the presence of almost all Eph receptors and ephrins have been found in thymus in a compartmented pattern and differential expression in distinct thymic cell subsets (18, 19). We have also previously demonstrated a role for Eph/ephrins A in the in vitro T cell development (18), and increasing evidence confirms the involvement of these molecules in the functioning of immune system, as in other tissues (20). Ephrin B2 is recruited to patched rafts after TCR cross-linking, and enhances the T cell proliferation induced by suboptimal anti-CD3 stimulation and MAPK pathway activation (21). EphB6 cross-linking induces apoptosis in Jurkat cells (22), modulates TCR-mediated responses, and inhibits JNK activation (23). The cell chemotaxis in response to different chemokines is also modulated by costimulation with ephrins A or B (24). In contrast, human dendritic cells respond to an ephrinA3-Fc fusion protein by adhering to fibronectin via integrins (25).
Because many Eph/ephrins are expressed in thymus, it is difficult to assign specific roles for different Eph receptors and ephrins. In fact, a role for Eph/ephrins in the in vivo thymus organogenesis and/or T cell development has not been conclusively defined. In other systems, analysis of different ephrin gene-disrupted mice has provided important information (26), but very limited data are available on the immune system. An EphB6 knockout mouse does not reveal substantive modifications in its proportions of thymocyte subpopulations (27), although it shows defects in T cell-mediated responses (28). Preliminary results (29) indicate that EphB6-transgenic mice show decreased numbers of thymocytes and altered proportions of different T cell subsets. However, these same authors were unable to find important alterations in the thymus phenotype of EphB2 knockout mice (29).
In the present study, we accomplish the role of EphA4 in in vivo thymus development by analyzing the phenotype of an EphA4-deficient mouse. EphA4 is particularly interesting because it is able to bind both ephrins A and B. There are several examples of the role of EphA4 in defining cell positioning of muscle precursors (30), in the morphogenetic processes that allow epithelialization of somites (31) and in the migration of neural crest cells (32), which is directly associated with thymus development (33, 34). In muscle cells, EphA4 activates the Jak/STAT pathway (11) which is common to many signaling events occurring in the thymus (35). Our current results demonstrate that EphA4 has a key role in the generation of a proper epithelial network and, as a consequence, in the T cell differentiation. This is, therefore, the first report demonstrating an in vivo role for Eph/Ephrins in thymopoiesis, confirming our previous in vitro findings and shedding new insight into the mechanisms governing both the establishment and the maintenance of the different thymic microenvironments as well as in the thymocyte-stromal cell relationships.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
EphA4 knockout (KO)3 mice in a C57BL/6-DBA-2-mixed background were provided by Dr. P. Charnay (l’Institut National de la Sante et de la Recherche Medicale, Ecole Normale Superievre, Paris, France). These animals and EphA4 KO in the C57BL/6-DBA-2-CD1-mixed background were maintained in pathogen-free conditions and descendents from heterozygous parents were used for analysis in all cases.
Flow cytometry
Thymus cell suspensions were stained for 20 min in PBS 1% FCS with specific mAbs against either CD4, CD8
(Caltag Laboratories), TCR
, TCR
, CD69, CD25, CD44, CD19, CD3, CD117 (c-Kit) (BD Biosciences) labeled with either PE, FITC, Tricolor, or allophycocyanin and/or annexin V (Annexin VFUOS; Roche Diagnostics/Molecular Biochemicals), washed, and resuspended for analysis. For cell cycle analysis, after surface labeling, cells were fixed in Cellfix (BD Biosciences) overnight and stained with Hoechst 33342 (Sigma-Aldrich-Química) in EtOH 30% in PBS 1% BSA for 1 h at room temperature. At least 20,000 cells/sample were analyzed in a FACSCalibur or LSR (BD Biosciences) and CellQuest software at the Centro de Microscopia y Citometria (Complutense University, Madrid, Spain).
Microscopy
For immunofluorescence, cryosections from thymi were fixed in acetone for 10 min and air dried. Afterward, slides were incubated with the corresponding primary Ab(s): anti-ephrin A rabbit antiserum (Santa Cruz Biotechnology), biotin-conjugated anti-Cl II mAb, unlabeled MTS10 mAb (BD Biosciences), that specifically recognizes mouse medullary thymic epithelial cells (TECs), anti-mouse keratin 5 rabbit antiserum (Covance) or anti-keratin 8 mAb Troma-1 (Developmental Studies Hybridoma Bank) (at 5–10 µg/ml in PBS 1% FCS). After washing, staining was revealed with Texas Red/FITC-conjugated donkey anti-rabbit IgG Ab, avidin-Texas Red, multiadsorbed Texas Red-conjugated donkey anti-rat IgM or biotin-conjugated donkey anti-rat IgG and avidin AMCA (Jackson ImmunoResearch Laboratories), respectively, at 5 µg/ml. For MTS10-cytokeratin double staining, an additional incubation with FITC anti pan-cytokeratin (Sigma-Aldrich) (2 µg/ml) was performed. Each incubation step was conducted for 30 min at 4°C followed by 3 x 5 min washes in PBS.
Sections were photographed with a Spot 2 digital camera on a Zeiss Axioplan microscope and Metamorph software at the Centro de Microscopia y Citometria.
For electron microscopy, thymi were fixed in 2.5% glutaraldehyde/0.1 M sodium cacodylate at 4°C for 3–5 h and embedded in PB 812 (PolySciences) as previously described (36). Semithin sections stained with toluidine blue were used for identifying the most interesting areas under a light microscope. Ultrathin sections were double stained with uranyl acetate and lead citrate and examined in a JEOL 10.10 electron microscope at the Centro de Microscopia y Citometria.
Reconstitution assays
Bone marrow cell suspensions were labeled with FITC-conjugated anti-CD3, -CD4, -CD8, -DX5, and -CD19 mAbs, and negatively isolated with anti-FITC MACs beads (Miltenyi Biotec) in an Automacs (Miltenyi Biotec) according to the supplier’s instructions. Typically, >95% of the obtained cell suspension was negative for these markers. A total of 2 x 106 cells from this negative cell fraction were i.v. injected in SCID mice. For epithelium functionality assays, fetal thymic lobes from 15-day-postcoitum fetuses were treated with 1.35 mM 2-deoxiguanosine for 5 days and grafted under the renal capsule of SCID mice. Afterward, grafted mice were reconstituted with bone marrow precursors as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
EphA4-deficient mice present decreased cellularity and altered thymocyte subsets
The thymus of EphA4–/– mice was smaller in size compared with their littermate wild-type (wt) controls. This thymic hypoplasia became more evident as mice grew up. Thus, at around 4 wk, the thymus size of all EphA4–/– mutants was severely reduced, with respect to their body size, and thymocyte numbers were up to 100 times lower than those of control thymi. As compared with EphA4+/+ wt mice, EphA4+/– heterozygous animals showed no significant variations in the cellularity of lymphoid organs or only slight differences. The thymic hypocellularity found in the EphA4–/– thymus, also resulted in decreased numbers of peripheral T lymphocytes, since early age (Fig. 1A).
|
molecules showed an abnormal T cell development in 
70% (25 of 35) of the 4-wk-old EphA4–/– mice. In these mutants, the absolute numbers of all lymphoid cell subsets were importantly reduced, the double-positive (DP) (CD4+CD8+) cell compartment being the most affected, and consequently, an extreme reduction in its percentage was observed. In the most severe phenotypes, this thymic cell population was almost absent. On the contrary, heterozygous EphA4+/– mice showed normal proportions of the different thymocyte subsets (Fig. 1B). In younger animals, the frequency of EphA4–/– mice exhibiting reduced proportions of DP thymocytes was lower. Thus, only 3 of the 27 mice examined during the first 2 wk of postnatal life showed a reduced percentage of DP cells, whereas 13 of 38 3-wk-old deficient mice exhibited this reduction (Fig. 1C). This variation in the frequency of mice that exhibited the described thymocyte phenotype was dependent on the mixed background of used KO mice. When the C57BL6/DBA-mixed background colony was backcrossed for four generations into a C57BL6 background the penetrancy of the observed phenotype, evaluated as the decreased percentage of DP thymocytes, was reduced. On the contrary, when they were backcrossed for three generations into a DBA background or when introduced the CD1 (outbred) background, the penetrancy remained unmodified (data not shown). The above-described data and the following reported results correspond to mixed background colonies.
EphA4–/– mice with normal or slightly decreased percentages of DP cells showed other alterations of T cell development. Thus, 8 of 10 animals without reduced percentages of DP thymocytes showed a slight increase in the percentage of double-negative (DN) thymocytes and decreased proportions of mature TCRhigh cells. In correlation with the decreased proportions of total TCRhigh cells in these animals, we also found decreased numbers of both DP TCRhigh cells and DP CD69+ cells (Fig. 2).
|
high and DP CD69+ cells suggest that the mentioned DP to mature single-positive transition could be made difficult by an inefficient TCR selection. To determine possible alterations at the earliest stages of intrathymic T cell development that could explain both the accumulation of DN cells and the reduced proportions of DP cells in most affected animals, we performed a flow cytometric analysis of the more immature thymocyte subsets defined by the expression of CD117 (c-kit), CD44, and CD25 cell markers. To restrict our study to the T cell lineage, a Lin– (CD19–CD3–CD4–CD8–) cell population was gated in a CD44 vs CD117 (c-kit) analysis excluding the CD44+CD117– population which contains non-T cell precursors (38) (NK cells, macrophages, and dendritic cells). This gated cell population was further analyzed for CD44 and CD25 expression (Fig. 3A).
|
Both increased apoptosis and reduced numbers of cycling cells occurred in EphA4 mutant mice
To examine other factors that could contribute to the reported thymus hypocellularity, we evaluated the proportions of both apoptotic cells and cycling cells occurring in the different T cell subsets of either EphA4–/– or EphA4+/+ mice. Apoptosis was measured by staining cell suspensions with annexin V in combination with the expression of CD4 and CD8 T cell markers and analyzed by flow cytometry. This analysis revealed a higher percentage of apoptotic cells in the EphA4 mutant mice. This increase in the proportion of annexin V-positive cells were observable from the younger animals, although, in our analysis, changes were statistically significative in animals older than 15 days, affecting also those which did not show a dramatic reduction in the percentage of DP cells. The increased apoptosis affected mainly to the DP cell population (Fig. 4, A and B), suggesting that this would be an additional factor to account for both the reduced thymic cell numbers and the decreased percentage of DP cells observed in EphA4–/– mice. The analysis of DNA content showed that the cell cycle was also reduced in the EphA4 mutants. In this case, however, the proportions of proliferating cells were reduced in all the thymocyte subsets (Fig. 4, C and D). Again, this decrease in the proportion of cycling cells was observable in younger animals, affecting mainly DN and DP cell subpopulations (Fig. 4D).
|
EphA4-deficient mice showed an abnormal histological organization of the thymic epithelial network
Light and electron microscopy examination of the thymus of EphA4–/– mice showed important histological alterations which could explain the defective development of T cell precursors found in these animals. As compared with the wt thymus, the EphA4–/– thymus exhibited the already mentioned reduction of cell content as well as a decrease in the area occupied by the thymic cortex. The electron microscopy analysis confirmed the extreme reduction of the thymic cortex in the EphA4–/– mice which consisted of a few layers of developing thymocytes, largely medium and large cells, in a meshwork of TECs arranged in parallel to the thymic capsule. On the contrary, the wt thymus contained a cortex full of small and medium thymocytes closely packed among the narrow cell processes of TECs arranged perpendicularly to the thymus surface (Fig. 5A).
|
EphA4–/– bone marrow precursors develop normally in SCID mice, but the T cell development is altered in an EphA4-deficient thymic stroma
Because both the thymic epithelial network and thymocyte development were altered in EphA4-deficient mice, we analyzed whether this phenotype could be due to a direct cell autonomous role of EphA4 on thymocyte development or the consequence of T cell progenitors developing in an altered thymic microenvironment.
To test whether the observed phenotype was a consequence of the lack of EphA4 signaling in thymocytes, we first analyzed the development of EphA4–/– bone marrow cell precursors in the wt thymic environment of SCID mice. A total of 2 x 106 EphA4–/– or EphA4+/+ bone marrow cells depleted of mature lymphoid cells were i.v. injected in SCID mice. EphA4–/– bone marrow were collected from animals showing severe reduction of DP thymocytes. After 6 wk, both EphA4–/– cell precursors and EphA4+/+ control progenitors had completely reconstituted the lymphoid organs of the SCID mice. In both cases, the successfully reconstituted thymus showed normal numbers of cells (88.5 x 106 ± 19.53 recovered from EphA4–/– bone marrow-reconstituted mice vs 94.72 x 106 ± 23.82 from controls) and normal distribution of the developing thymocyte subsets (Fig. 6A).
|
These results excluded the need for EphA4 signaling on T cell precursors for their normal development and suggested a defective environment as being the cause for the altered T cell development observed in the EphA4 mutant mice. To confirm this possibility, 15-day-old fetal thymic lobes from either EphA4-deficient mice or wt control animals, depleted of lymphoid cells by d-Guo treatment, were grafted under the renal capsule of SCID mice. Afterward, grafted SCID mice were injected with wt bone marrow cell precursors depleted of mature lymphoid cells as in the previous experiments. After 6 wk, both SCID thymus and grafted thymus were reconstituted but whereas newly SCID thymus was normally reconstituted with the wt precursors (data not shown), the thymi developed from the EphA4-deficient grafted thymic lobes presented lower numbers of cells (0.65 ± 0.8 x 106 thymocytes recovered from EphA4–/– mice vs 2.48 ± 0.3 x 106 recovered from wt controls) and impaired T cell development compared with those found in the control thymi developed from the grafted wt thymic lobes (Fig. 6, B and C). When d-Guo-treated lobes were transplanted into SCID mice which were not reconstituted with bone marrow progenitors just DN thymocytes could be recovered. No developing DP thymocytes were found in either wt or mutant grafts (data not shown). Thus, the altered T cell development found in EphA4-deficient mice would largely be the consequence of the development of normal T cell precursors in an altered thymic stroma.
Alterations of the thymic epithelial network maturation appear earlier than those of T cell development
To further correlate defective thymic microenvironment and altered development of T cell precursors, we studied the thymus histology at earlier developmental stages. In neonatal EphA4–/– thymi, densely packed epithelial cells appeared already in some thymic areas whereas in others, no keratin-positive cells were present. A higher proportion of K5+K8+ presumptive immature TECs than in control wt mice were also found (Fig. 7A). No important variations were found in the development of other elements of thymus stroma, such as blood vessels or connective tissue (data not shown). On the contrary, as above mentioned (Fig. 1C) at neonatal stages, although the thymic cellularity was reduced, all thymocyte CD4/CD8 subsets were present with normal or almost normal proportions. Similarly altered histological TEC disposition could be already observed in 17-day-postcoitum EphA4–/– fetal thymi (Fig. 7B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
In the current study, we demonstrate that the lack of EphA4 results in important alterations in the development of the epithelial network which in turn results in a defective T cell differentiation. Previous reports had failed to find thymus alterations in Eph B6-deficient mice (27). Only recent preliminary data demonstrate decreased numbers of thymocytes and slightly altered T cell development in mice which carry an Eph B6 transgene (29). These studies require further confirmation and, in any case, the thymus phenotype of Eph B6-transgenic mice is unrelated to that reported here for EphA4–/– mice. EphA4 is a member of this large family of tyrosine kinase receptors related to the migration of neural crest cells which are required for pharyngeal arch development involved in thymus organogenesis (32). We had previously demonstrated EphA4 expression in the rat thymus and how the addition of EphA-Fc fusion proteins to rat fetal thymus organ cultures significantly decreased the total number of isolated cells from the lobe and particularly that of the DP (CD4+CD8+) cell compartment (18).
The EphA4-deficient mice present thymus hypoplasia that becomes more evident as the mice grow up. Both fetal and neonatal thymus contain already lower cell numbers than that of EphA4+/+ mice, but the thymus of 4-wk-old EphA4–/– mice shows a severe reduction of the cell numbers that can be as low as 1/100 of the control cellularity. In addition, most EphA4–/– animals present both a low proportion of DP (CD4+CD8+) thymocytes and altered proportions of the different subpopulations of TN (CD3–CD4–CD8–) cells defined by the expression of CD117 (c-Kit), CD44 and CD25 cell markers. Most Eph A4 mutants exhibit increased proportions of the CD44–CD25+ TN cell subset and in some of them the increase even affected the early CD117+CD44+CD25– TN cell population. In these last cases, our cytometrical study demonstrated that the increase affected the T cell lineage (c-kit+ cells) rather than B lymphocytes (CD19+ cells) or CD44+CD117– cells (NK cells, macrophages, dendritic cells). In contrast, EphA4–/– mice with less severe phenotype show also a partial blockade of the transition from DP to SP cell compartment, with decreased proportions of both DP TCRhigh thymocytes and DP CD69+ cells. Taken together, these results suggest that the lack of EphA4 could be affecting both TCR and TCR selection. Remarkably, Krox20/Erg2B, which is a direct regulator of EphA4 expression in other cell types, is concerned with the DN to DP thymocyte transition (43) and with thymic selection (44).
In addition to this less efficient developmental transition of thymocytes, other factors contribute to the phenotype of EphA4–/– mice. Thus, a higher proportion of apoptotic cells occurs mainly in the DP cell compartment of EphA4–/– thymi than in wt control ones and there is also a reduced percentage of cycling cells in all the thymocyte subsets of mutant mice.
Also, the thymic epithelial stroma shows extensive alterations in EphA4-deficient mice. In adult mutants, our electron microscopy study confirms the alterations of the cortical epithelium observed by immunofluorescence. Although the thymic epithelium retains the expression of important functional molecules such as MHC class II Ags and of ephrins A1 and A3, the higher affinity ligands of EphA4, the meshwork is profoundly collapsed. The TECs appear arranged in parallel to the thymic capsule rather than forming perpendicular lines of TECs which extend from the connective tissue capsule to penetrate deeply in the thymic parenchyma.
Remarkably, these alterations of the thymic epithelium appeared early during the life of EphA4-deficient mice. In EphA4–/– neonatal thymi, there is already a packed cortical epithelium as well as increased numbers of K5+K8+ epithelial cells and TEC disposition defects can be already observed in 17-day-postcoitum fetal thymi. The question arises as to which is the origin of these defects observed in the thymus of EphA4-deficient mice.
EphA4–/– mice are known to have defects in other organ systems that could be supporting disturbing stress that could explain the observed hypocellularity and increased apoptosis. However, the phenotype of EphA4–/– thymi is different of that reported for a stressed mouse, in which no such alterations of the thymic epithelial network are observed (45, 46, 47). Accordingly, mutant thymic stroma grafted in SCID mice is not functional and neonatal EphA4–/– thymus exhibits an altered epithelial microenvironment but normal or almost normal proportion of thymocytes.
In contrast, the EphA4 receptor has been reported to be involved in the cell migration in other biological systems (30, 32, 42, 48, 49). Some results (32, 48, 50) point out a role for EphA4 in the organization of pharyngeal arches, through the targeted migration of branchial neural crest cells, and, therefore, in the formation of thymic primordium. However, defects in neither the migration of neural crest cells or pharyngeal arch patterning have been previously described in EphA4-deficient mice. In our study, EphA4 mutants thymic primordium is formed and colonized by normal numbers of c-kit+ fetal thymocyte precursors (data not shown) which are partially able to differentiate and a correctly positioned thymus is developed. Although thymic hypoplasia in neonatal thymus is observed, as in other neural crest defects, in EphA4–/– mice, it becomes severe in postnatal life.
EphA4-deficient mice phenotype might be explained by a direct role of EphA4 on thymocyte precursor development. However, our results on bone marrow chimeras, discard this possibility, EphA4-deficient bone marrow precursors colonize the thymus and develop properly intrathymically. Notably, bone marrow precursors used for these experiments were isolated from mice with a severely reduced number of thymocytes and percentage of DP cells. Thus, EphA4 signaling is not directly necessary for thymocyte precursor development.
On the contrary, wt bone marrow precursors undergo altered development, with decreased proportions of DP (CD4+CD8+) thymocytes, in an alymphoid EphA4-deficient thymic stroma grafted under the renal capsule of SCID mice suggesting that the observed alterations in EphA4–/– thymi are due to the observed defects in the patterning and/or function of the thymic stromal cell components, largely epithelial cells. In fact, our previous results, and in the current study, had demonstrated that EphA4 higher expression was associated with the TECs, mainly in the fetal thymus, rather than with the thymocytes (18). Other authors have detected EphA4 transcript in human fetal thymic stroma (51). However, Vergara-Silva et al. (19) reported that EphA4 is weakly expressed in the mouse thymus and is associated with SP thymocytes and medullary epithelial cells. These disagreeing results could be reflecting the different techniques used to detect Eph expression.
On this direction, the lack of EphA4 in TECs could affect directly the T cell development through the reverse signals mediated by ephrin A expressed on developing thymocytes (18, 19). Thus, EphA4 receptors could provide ephrin-positive thymocytes with the necessary signals for their correct migration, intrathymic positioning, and/or development. However, this is difficult because ephrins A, the highest affinity ligands of EphA4, although it is able to bind weakly ephrins B, are molecules bound to the cell surface via GPI. It has been reported that GPI-deficient thymocytes develop correctly (52). Furthermore, ephrins expressed on developing thymocytes could be bound by other members of Eph family A normally expressed on TECs. On the contrary, our results indicate that the lack of EphA4 results in an altered development of the thymic epithelial compartment, rather than other elements of the thymic stroma, most probably due to the deprivation of forward signals in the EphA4–/– TEC. An indirect role for these nonepithelial stromal cell components cannot be, however, completely discarded.
Actually, a role for EphA4 in epithelial cell differentiation has already been described. During somite morphogenesis, mesenchymal cells in the paraxial mesoderm differentiate into epithelial cells that delineate the boundaries between forming somites. EphA4-ephrin interactions cause the cells at the boundary interface to detach from each other. The cells at the two sides of the forming boundary acquire columnar morphology, become polarized and relocalize the -catenin to the apical junctions, the centrosome, and the nucleus. EphA4 forward signaling plays a direct role in this process in the expressing cells but also a cell-nonautonomous effect in the ephrin-expressing cells (31). Similarly, EphA4 activation in early Xenopus embryos disrupted cadherin-dependent cell adhesion. Cells acquire a rounded morphology, lose apical microvilli, and the apical/basolateral boundary is also lost, indicating that cells lost their apical/basolateral polarity (53). The molecular bases of these changes are unknown. EphA4 activity has been associated with Rho family of GTPases and cytoskeleton dynamics (54, 55) and is known to activate the Jak/STAT pathway (11). In addition, cross-talk with the fibroblast growth factor receptor (FGFR) family has also been reported (56, 57). Mice deficient for FGFR2-IIIb exhibit hypoplastic thymus and a block of the maturation of the epithelium with increased K5-positive cells and decreased numbers of proliferating stromal cells (58). EphA4 could, therefore, act in correlation with other morphogenetic agents, such as FGF, defining repulsion vs adhesion/attraction by means of regulating cell-to-cell and cell-to-extracellular matrix adhesion molecules and cytoskeleton dynamics. This general role assumed for Ephs regulates cell movements and cell position, but also, cell shaping and cell polarity. Other examples relate Eph family with epithelium development. Thus, in the intestinal epithelium EphB2/B3, whose expression is regulated by TCF and -catenin, regulate the relative position of proliferative and differentiated cell populations (59). Finally, Eph family is also important in defining three-dimensional epithelial structures. Activation of EphA receptors that are expressed in cultured kidney epithelial cells inhibits branching morphogenesis induced by hepatocyte growth factor in collagen gels (12). Also, targeted expression of a dominant-negative FGF receptor blocks branching morphogenesis and epithelial differentiation of the mouse lung (60).
Apart from this role in the epithelial cell differentiation, Eph family could be involved in the maintenance of differentiated epithelial cells. Its expression is altered in many epithelial tumors and has been related with metastatic behavior (61). A role in maintaining epithelial identity and avoiding lost of cell adhesion and epithelial-mesenchyme transition has been proposed (62). Lost of cell adhesion and epithelial characteristics and/or certain mesenchymal transition could explain the appearance of keratin-negative cells which still maintain certain epithelial characteristics in EphA4 mutants.
It has been previously reported that TECs from mice with early arrested T cell development present a two-dimensional structure and are unable to organize the three-dimensional structure proper of the normal thymic epithelium (63). These alterations could also explain the strong packing observed in the epithelial network of EphA4–/– mice. In this case, the presence of all thymocyte subsets, although in altered proportions, does not allow the proper organization of TECs. Other defective mice, such as those expressing a truncated form of the transcription factor Foxn1, also retain certain T cell differentiation, but the lack of signaling to epithelial cells results in the impediment of their development (64).
The lack of a well-structured stroma could affect the T cell development throughout changes in the cell-to-cell interactions, known to be a requisite for the correct development of thymocytes. The absence of Eph A4 signaling in the TECs of mutants could also affect the expression of epithelial genes involved in governing the proper T cell maturation. On this respect, a similar adult histological organization has been observed in an epithelium conditioned STAT3 mutant mouse. In agreement with our findings on the thymus of EphA4–/– mice, the thymus of epithelium conditioned Stat3 mutants exhibits an altered thymic epithelial meshwork which results in an age-dependent reduction of total cell content and the percentage of DP cells, accumulation of CD44+CD25– TN cell precursors and increased proportions of apoptotic cells (65). Further analysis should help to determine the molecular and cellular basis of these physiological alterations on thymic development and function resultant from the lack of EphA4.
Therefore, EphA4 is a new molecule to incorporate to the list of relevant factors for the regulation of the proper development of the thymic epithelial network. Its lack provokes profound alterations in the thymic epithelium that indirectly results in both thymic hypocellularity and blockade of T cell differentiation.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by Grants BCM2001-2025, BFU2004-03132, and BCM2003-01901 from the Spanish Ministry of Education and Culture and CAM08.1/0038.1/2003 and GR/SAL/0582/2004 from the Comunidad Autónoma de Madrid. 
2 Address correspondence and reprint requests to Dr. Agustín Zapata, Department of Cell Biology, Faculty of Biology, Complutense University of Madrid, Antonio Novais SN 28040, Madrid, Spain. E-mail address: zapata{at}bio.ucm.es
3 Abbreviations used in this paper: KO, knockout; DP, double positive; DN, double negative; TN, triple negative; wt, wild type; FGF, fibroblast growth factor.
Received for publication December 29, 2005. Accepted for publication April 21, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-Catenin and TCF mediate cell positioning in the intestinal epithelium by controlling the expression of EphB/ephrinB. Cell 111: 251-263. [Medline]
| ||||||||||||||||||||||||||||||||||||||
