Notch ligand-bearing thymic epithelial cells initiate and sustain Notch signaling in thymocytes independently of T cell receptor signaling.

Thymic epithelial cells are specialized to play essential roles at multiple stages of T cell development in the thymus, yet the molecular basis of this specialization is largely unknown. Recently, the Notch family of transmembrane proteins has been implicated in thymocyte development. Such proteins interact with cell surface proteins of the Delta-like and Jagged families. It is known that Notch ligands are expressed intrathymically, and that Notch signaling is regulated by Notch ligands expressed on either the same or third-party cells. However, functional analysis of Notch ligand expression, and elucidation of the mechanism of Notch ligand signaling in thymocyte development, are unclear. Here, we find that Notch ligand expression in the thymus is compartmentalized, with MHC class II(+) thymic epithelium, but not thymocytes nor dendritic cells, expressing Jagged-1, Jagged-2 and Delta-like-1. We also provide evidence that contact with Notch ligands on thymic epithelium is necessary to activate and sustain Notch signaling in thymocytes, and that this can occur independently of positive selection induction. Our data suggest that Notch ligand expression by thymic epithelium may partly explain the specialization of these cells in supporting thymocyte development, by regulating Notch activation via an inductive signaling mechanism independently of signaling leading to positive selection.

Specialisation of thymic epithelial cells for positive selection of CD4+8+ thymocytes.

Following their migration into the thymus, hemopoeitic stem cell precursors enter a complex developmental pathway involving proliferation, differentiation and alphabetaT-cell receptor (alphabetaTCR)-mediated selection procedures, in order to generate mature T-cell populations ready for export to the periphery. Thus, a critical stage during intrathymic T-cell development involves the generation of functionally mature CD4+8- and CD4-8+ cells from immature CD4+8- precursor thymocytes, a poorly understood process referred to as positive selection. While interactions between the alphabetaTCR and MHC-peptide complexes are known to be essential for the initiation of positive selection, additional unknown signals are also required. Using an in vitro reaggregate thymic organ culture system which allows comparison of the abilities of various cell types to induce maturation of CD4+8+ precursors, we provide evidence that both MHC-peptide complexes and specialised accessory molecules must be provided by thymic epithelium for efficient mediation of positive selection. Moreover, analysis of positive selection in the presence of thymic and non-thymic stromal cells expressing MHC class II molecules with the same limited peptide array suggests that this unique ability of thymic epithelium to mediate positive selection of CD4+8- cells is not solely due to presentation of a specialised peptide repertoire, but is dependent upon provision of specialised accessory interactions.

Lymphostromal interactions in thymic development and function.

The generation of a peripheral T-cell pool is essential for normal immune system function. CD4+ and CD8+ T cells are produced most efficiently in the thymus, which provides a complexity of discrete cellular microenvironments. Specialized stromal cells, that make up such microenvironments, influence each stage in the maturation programme of immature T-cell precursors. Progress has recently been made in elucidating events that regulate the development of intrathymic microenvironments, as well as mechanisms of thymocyte differentiation. It is becoming increasingly clear that the generation and maintenance of thymic environments that are capable of supporting efficient T-cell development, requires complex interplay between lymphoid and stromal compartments of the thymus.

Microenvironmental regulation of T cell development in the thymus.

T cell development in the thymus occurs through a series of events beginning with thymic colonization by migrant precursors and ending with the emigration of functionally competent CD4+ and CD8+ T cells to the periphery. It is well accepted that signals through the pre-T cell receptor (pre-TCR) and alpha-beta TCR (alphabetaTCR) complex play pivotal roles in the maturation of CD4-8- and CD4+8+ thymocytes, respectively. It is clear that stromal cells constituting the thymic microenvironment provide non-TCR-mediated interactions which are essential for several developmental events. Examples of such will be discussed here in relation to early and late events in T cell development.

The role of mesenchyme in thymus development.

We have reviewed the evidence that thymic mesenchymal cells and their progeny thymic fibroblasts play an important role in early T-cell development. Although it is possible that mesenchyme plays an inductive role in thymic epithelial morphogenesis, we have presented evidence to suggest that there is a direct effect of mesenchyme and fibroblasts on T-cell development. Moreover the association of these cell types with an ECM raises the possibility that the latter might be important in integrin and/or cytokine presentation especially during the CD4(-)8- phase of T-cell development.

RNA and protein expression of the murine autoimmune regulator gene (Aire) in normal, RelB-deficient and in NOD mouse.

Mutations in the putative transcription factor autoimmune regulator (AIRE) gene are responsible for autoimmune polyendocrinopathy-candidiosis-ectodermal dystrophy (APECED; OMIM#240300), a monogenic recessively inherited disease characterized by destructive autoimmune diseases of the endocrine organs, chronic candidiosis of mucous membranes and ectodermal dystrophies. In this study the expression of murine homolog for AIRE protein, Aire, was detected in a fraction of thymic medullary epithelial cells. Subcellularly, in thymus the protein appears as concentrated into nuclear dot-like structures, whereas in transfected cells the protein is also bound along a cytosolic fibrillar network. By RT-PCR Aire mRNA was detected in thymus, lymph node, spleen and testis although the second round PCR amplified Aire specific band from most mouse tissues analyzed. Furthermore, the Aire mRNA was detected in dendritic cell (DC) populations isolated from thymus and spleen, representing both myeloid- and lymphoid-related lineages of DC. We also demonstrate that the Aire protein is absent in the thymus of RelB-deficient mouse and in NOD thymus most of the Aire positive cells showed an abnormal morphology. These results suggest that the Aire protein is associated with the normal development and/or action of a subset of thymic medullary stromal cells involved in tolerance induction.

Studies on the role of IL-7 presentation by mesenchymal fibroblasts during early thymocyte development.

T cell precursor development depends upon poorly defined interactions between thymocyte precursors and stromal cells involving cell surface molecules, extracellular matrix (ECM) components and soluble growth factors. To determine whether presentation of soluble factors by ECM is involved in early T cell development, we analyzed expression of ECM components in individual thymic stromal subsets and investigated their ability to present ECM-associated IL-7, a factor known to be important during early thymocyte development. We show that MHC class II(+) thymic epithelium and fibroblasts - essential requirements for development of CD4(-)8(-)precursors - both show surface expression of ECM components such as fibronectin and heparan sulfate. Use of biotinylated IL-7 protein indicates that both cell types bind IL-7, while enzymatic disruption of specific ECM components indicates that IL-7 presentation by both cell types is dependent upon heparan sulfate. However, disruption of IL-7 presentation specifically on fibroblasts does not affect their ability to contribute to T cell development. Collectively, these data suggest that while ECM-mediated presentation of IL-7 may be a general function of thymic stromal cells during thymocyte development, heparan sulfate-mediated IL-7 presentation specifically by fibroblasts is not essential and that the specific requirement for fibroblasts in early development involves additional undefined interactions.

An essential role for the IL-7 receptor during intrathymic expansion of the positively selected neonatal T cell repertoire.

Intrathymic T cell development is a multistage process involving discrete phases of proliferation as well as differentiation. From studies on IL-7 or IL-7Ralpha-deficient mice, it is clear that the IL-7 receptor (IL-7R) plays a critical role during the initial stages of intrathymic CD4-8- precursor development. In contrast, the role of IL-7R in later stages of thymocyte development are unclear. Here, we have used various approaches to investigate directly the role of the IL-7R in thymocyte positive selection and the recently described phase of postselection proliferation. First, we show that positive selection involves selective up-regulation of IL-7Ralpha- and IL-7Rgamma-chains, with the majority of CD4+ and CD8+ cells being IL-7R+. Second, MHC class II+ thymic epithelium-which drives postselection proliferation-expresses IL-7 mRNA. Finally, analysis of positive selection and postselection proliferation in thymocytes from IL-7Ralpha-/- neonates shows that positive selection occurs normally, whereas postselection expansion is drastically reduced. Thus, our data provide the first evidence that, as well as playing a role during early phases of thymic development, IL-7R mediates intrathymic expansion of positively selected thymocytes, which may aid in establishment of the neonatal peripheral T cell pool.

An essential role for thymic mesenchyme in early T cell development.

We show that the mesenchymal cells that surround the 12-d mouse embryo thymus are necessary for T cell differentiation. Thus, epithelial lobes with attached mesenchyme generate all T cell populations in vitro, whereas lobes from which mesenchyme has been removed show poor lymphopoiesis with few cells progressing beyond the CD4(-)CD8(-) stage of development. Interestingly, thymic mesenchyme is derived from neural crest cells, and extirpation of the region of the neural crest involved results in impaired thymic development and craniofacial abnormalities similar to the group of clinical defects found in the DiGeorge syndrome. Previous studies have suggested an inductive effect of mesenchyme on thymic epithelial morphogenesis. However, we have found that mesenchyme-derived fibroblasts are still required for early T cell development in the presence of mature epithelial cells, and hence mesenchyme might have a direct role in lymphopoiesis. We provide an anatomical basis for the role of mesenchyme by showing that mesenchymal cells migrate into the epithelial thymus to establish a network of fibroblasts and associated extracellular matrix. We propose that the latter might be important for T cell development through integrin and/or cytokine interactions with immature thymocytes.

Specialized ability of thymic epithelial cells to mediate positive selection does not require expression of the steroidogenic enzyme p450scc.

Thymic epithelial cells are uniquely efficient in mediating positive selection, suggesting that in addition to providing peptide/MHC complexes for TCR ligation, they may also provide additional support for this process. Recent studies have shown that although engagement of either the TCR or glucocorticoid (GC) receptors can individually induce apoptosis in thymocytes, together these signals are mutually antagonistic. This had led to the suggestion that local GC production by thymic epithelial cells, by opposing TCR signaling for apoptosis, provides the basis of the ability of these cells to mediate thymocyte positive selection. In this paper we have examined this possibility directly and shown that highly purified cortical epithelial cells, which have the functional ability to mediate positive selection in reaggregate cultures, do not express mRNA for the key steroidogenic enzyme P405scc. Thus we conclude that the ability of thymic epithelial cells to support positive selection does not rely on their ability to produce GC. However, we find that P450scc mRNA is up-regulated in thymocytes on the initiation of positive selection, raising the possibility that any local protective effect of steroid production is mediated at the level of thymocytes themselves.

Positive selection of thymocytes: the long and winding road.

Positive selection is a crucial stage in T-cell development because it is here that CD4+CD8+ cells bearing T-cell receptors that interact with self-major histocompatibility complex molecules are rescued from cell death, resulting in the generation of mature T cells. Here, Graham Anderson and colleagues review recent studies indicating that positive selection is a multistage process involving interactions with thymic epithelium.

Cellular and matrix interactions during the development of T lymphocytes.

The thymus contains an extensive extracellular matrix. Although thymocytes express integrins capable of binding to matrix molecules, the functional significance of the matrix for T cell development is uncertain. We have shown that the matrix is associated with thymic fibroblasts which are required for the CD44+ CD25+ stage of double negative (CD4-8-) thymocyte development. The survival of cells at this stage is dependent on IL-7 and we propose that the role of fibroblasts is to present, via the matrix, IL-7 to developing T cells.

Factors regulating stem cell recruitment to the fetal thymus.

Colonization of the thymic rudiment during development is initiated before vascularization so that hemopoietic precursors must leave the pharyngeal vessels and migrate through the perithymic mesenchyme to reach the thymus, suggesting that they may be responding to a gradient of chemoattractant factors. We report that diffusible chemoattractants are produced by MHC class II+ epithelial cells of the fetal thymus, and that the response of precursors to these factors is mediated via a G protein-coupled receptor, consistent with factors being members of the chemokine family. Indeed, a number of chemokine receptors are expressed by thymic precursors, and several chemokines are also expressed by thymic epithelial cells. However, these chemokines are also expressed in a tissue that is unable to attract precursors, although the thymus expressed chemokine, TECK, is expressed at higher levels in thymic epithelial cells and we show that it has chemotactic activity for isolated thymic precursors. Neutralizing Ab to TECK, however, did not prevent thymus recolonization by T cell precursors, suggesting that other novel chemokines might be involved in this process. In addition, we provide evidence for the involvement of matrix metalloproteinases in chemoattractant-mediated T cell precursor recruitment to the thymus during embryogenesis.

CD69 expression discriminates MHC-dependent and -independent stages of thymocyte positive selection.

In the thymus, phenotypically and functionally mature single positive cells are generated from immature CD4+8+ precursors by a process known as positive selection. Although this event is known to involve alphabetaTCR ligation by peptide/MHC complexes expressed on thymic stromal cells, it is clear that positive selection is a multistage process involving transition through an intermediate CD4+8+69+ phase as well as subsequent postselection phases. By analyzing the development of preselection CD4+8+69- and intermediate CD4+8+69+ thymocytes in the presence of MHC class I-deficient, MHC class II-deficient, and MHC double-deficient thymic stromal cells, we investigated the role of MHC molecules at three distinct points during positive selection. Although the initiation of positive selection is critically dependent upon MHC interactions, we find the that later stages of maturation, involving the differentiation of CD4+8- and CD4-8+ cells from CD4+8+69+ thymocytes, occur in the absence of MHC molecules. Moreover, an analysis of the postselection proliferation of newly generated CD4+8- and CD4-8+ thymocytes shows that this also occurs independently of MHC molecules. Thus, our data provide direct evidence that, although positive selection is a multistage process initiated by TCR-MHC interactions, continuation of this process and subsequent postselection events are independent of ongoing engagement of the TCR.

A novel method of cell separation based on dual parameter immunomagnetic cell selection.

We describe a novel method of cell purification involving two stage immunomagnetic selection which permits isolation of cells based on a second cell surface marker without the need for removal of beads used in the first selection step. This approach takes advantage of the size differences between commercially available immunomagnetic beads and/or particles and their differing properties in terms of attraction to magnetic fields of various strengths. The first stage of separation involves positive selection of cells using the Miltenyi MiniMacs system, utilising 50 nm MicroBeads and a MiniMacs magnet. Cells obtained from this procedure--still rosetted with 50 nm MicroBeads--can then be subjected to further positive or negative selection using either streptavidin M280 or anti-rat M450 Dynabeads, without the need for prior bead removal, since the strength of the magnetic field of the Dynal separator is sufficient to attract the larger Dynabeads but not the MicroBeads. Here, we show that this system can be used to isolate a number of cell types including very rare target cell populations such as haemopoietic stem cells, using two different surface markers without perturbing subsequent functional capacity.

Studies on the phenotype of migrant thymic stem cells.

Stem cells first enter the thymus around the 11th to 12th days of gestation in BALB/c mouse embryos. The phenotype of these stem cells has been difficult to determine because their entry occurs when the thymic primordium is very small and involves too few stem cells to allow studies by flow cytometry. We have been able to microdissect the thymus from embryos during this stage and immunophenotype cells in sections using a sensitive tyramide amplification system. Our results show that migrant stem cells express CD45, c-kit, CD44, CD34 and alpha4 integrin, but other markers such as CD62L, CD25, Thy-1.2, CD3epsilon, alpha5 integrin and RAG-1 expression are detected only after stem cell entry. These results should help to improve the isolation and characterization of migrant thymic stem cells.

In vitro models of T cell development.

A number of in vitro systems are currently being used to study both thymocyte development and thymic stromal cell function. However, the usefulness of dispersed culture systems is limited since they often involve disruption of interactions within the normal three-dimensional architecture of the thymus in vivo which are critical for normal development to proceed. In contrast, Fetal Thymus Organ Culture (FTOC) provides an experimental system where such interactions are maintained, thereby allowing in vitro access to key aspects of thymocyte development. More recently, Reaggregate Thymus Organ Cultures (RTOCs) have allowed detailed analysis of thymic stromal cell function, while retroviral transfection of thymocyte subsets under FTOC conditions provides a rapid means to investigate thymocyte development at the molecular level. Current use of the FTOC approach is summarised here, and where appropriate is compared to the use of dispersed culture systems.

Differential effects of peptide diversity and stromal cell type in positive and negative selection in the thymus.

Thymocyte positive selection results in maturation to the single-positive stage, while negative selection results in death by apoptosis. Although kinetic analyses indicate only 3-5% of CD4+ 8+ cells reach the single-positive stage, the balance of positive and negative selection and the nature and quantity of cells mediating maximal negative selection are uncertain. Here, using a system where the number and type of stromal cells and thymocytes can be controlled, we investigated the maturation of CD4+ 8+ thymocytes in the presence or absence of thymic epithelium and dendritic cells (DC) from wild-type (wt) and H-2M(-/-) mice expressing different peptide arrays. We find that titration of wt DC into reaggregates of wt epithelium has a dramatic effect on the number of CD4+ cells generated, with 1% DC causing a maximal 80% reduction. Moreover, while addition of 1% wt DC into cultures of H-2M(-/-) epithelium causes a 90% reduction in CD4+ cells, no effect was observed when similar numbers of wt thymic epithelium were added. Collectively, these data provide the first accurate indication of the quantity and quality of stromal cells required for maximal negative selection in the thymus, demonstrate the importance of peptide diversity in T cell selection, and highlight a large degree of overlap between positive and negative selection events.

Use of explant technology in the study of in vitro immune responses.

The establishment of in vitro culture systems provides an accessible means to study events within the immune system. In contrast to either dispersed suspension or two-dimensional monolayer culture, the explantation of tissue fragments under organ culture conditions is, to date, the only method which allows essential three-dimensional cellular interactions to be maintained under conditions which permit controlled experimental manipulation in vitro. Recent modifications of explant technology, particularly within the area of fetal thymic organ culture, now allow the controlled reassociation of defined cellular subsets and manipulation of gene expression, under conditions where the functioning of both lymphoid and stromal cell types closely resembles that in vivo.