Thymus and T‐lymphocyte development: what is new in the 21st century?

It has now been 45 years since J. F. A. P. Miller published his ground-breaking discovery that the thymus has a crucial immunological function (1). It seems hard now to believe that for much of the 20th century, physicians and scientists believed that, at best, the thymus was an endocrine organ and that, at worst, it was an evolutionary remnant with no function at all (2). This view was derived, at least in part, from experiments demonstrating that adult thymectomy had no impact on immunological competence in mice. As is often the case with important scientific discoveries, Dr Miller did not set out to define the immunological function of the thymus. Rather, he was interested in elucidating the role of the thymus in the pathogenesis of murine leukemia viruses. He began by showing that adult thymectomy could prevent virus-induced leukemia. However, such leukemogenic viruses had to be inoculated at birth to cause leukemia. Therefore, to determine whether the viruses could multiply outside the thymus, he clearly required neonatal thymectomy. After many trials and tribulations, he eventually succeeded in devising a surgical method to accomplish this in a way that left the mice viable. To his great surprise, he soon realized that these neonatally thymectomized mice succumbed to a host of viral infections that the adult thymectomized control mice survived. Although it took some time for his findings and conclusions to be accepted, these studies formally demonstrated that thymus-derived (T) lymphocytes perform essential functions in adaptive immunity. During the ensuing decades, the myriad complexities of the thymus and thymocyte development have been intensively investigated. We now know that, during embryonic life, the thymus is seeded by progenitors from the fetal liver but that sustained T lymphopoiesis in postnatal life requires continued influx of thymus-seeding progenitors (TSPs) from the bone marrow. Signaling through the Notch1 receptor has been discovered to be crucial to initiate T-lineage specification in TSPs and their progeny. Surprisingly, however, Notch1 signaling is also required to suppress intrathymic B-cell development. Thus, the thymus proves to provide a suitable microenvironment for B-cell development when Notch1 signaling is impaired. Progression through the CD4/CD8 double-negative-1 (DN1) to DN3 stages of thymocyte development is then regulated by an interplay of Notch1 and interleukin-7 (IL-7) receptor signaling. During these phases, rearrangements of the T-cell receptor (TCR) γ, δ, and β loci occur. Successful (in-frame) TCRβ rearrangement allows DN3 thymocytes to assemble pre-TCR complexes to induce ‘β-selection’. This process allows DN3 thymocytes to survive, proliferate, and generate a large pool of CD4/CD8 double-positive (DP) thymocytes that begin TCRα rearrangement after exiting the cell cycle. However, DN2/DN3 thymocytes that make successful γ and δ TCR rearrangements typically become DN γδ T cells. In order for DP thymocytes to escape a default fate of programmed cell death, they must express αβTCR complexes that can recognize self-major histocompatibility complex (MHC) proteins complexed with self-peptides on cortical thymic epithelial cells. However, cells with high affinity αβTCRs will be eliminated by negative selection, usually in response to interaction with self-MHC/self-peptide complexes on medullary epithelium. This issue of Immunological Reviews highlights some of the most recent advances made in these areas over the last 5 years and also covers important work in several areas of thymus biology that have only recently become elucidated. These areas include thymus organogenesis, transcriptional regulation of intrathymic cell fate choices, and chemokine-regulated migration of T-cell precursors through different thymic microenvironments. The following paragraphs provide a brief preview of the topics covered in this volume. We begin with two articles describing thymic organogenesis and the contributions of non-hematopoietic stromal cells to many aspects of intrathymic T-cell development (3, 4). This contribution was made clear with the discovery that mutation of the FoxN1 gene in both mice and humans causes T-cell immune deficiency because of defective development of the thymic epithelium. G. Anderson and colleagues (3) review recent progress in defining mechanisms that regulate the development and functioning of the thymic epithelium, including the role of particular transcription factors and soluble mediators and the recent identification of thymic epithelial progenitor cells. They also discuss the bi-directional nature of lymphostromal interactions during thymic organogenesis. Thymic stroma clearly supports survival and proliferation of T-cell progenitors, but recent studies document that T-cell progenitors critically regulate development of thymic stroma. The latter interaction is particularly important for the development of medullary epithelial cells that can induce negative selection to a wide array of self-antigens not expressed in the thymus (5). Holländer and colleagues (4) also discuss the development of thymic stroma but focus more on thymic organogenesis during fetal life. They describe how early interactions between endodermal epithelium and neural crest-derived mesenchyme generate the thymic primordium, and they also delve into the functions of several transcription factors in regulating patterning and growth of thymic epithelium. During postnatal life, T lymphopoiesis must be sustained by continued importation of progenitors from the bone marrow, which enter the thymus from the blood stream by extravasating through postcapillary venules near the corticomeduallary junction of the thymus. Many aspects of this process remain poorly understood. Schwarz and Bhandoola (6) discuss signals and molecular interactions that may be involved in mobilizing T-cell progenitors to leave the bone marrow and enter the circulation, as well as interactions that may regulate egress of TSPs from the blood into the thymic parenchyma. Goldschneider (7) discusses recent work from his laboratory documenting that thymus seeding is periodic and is regulated by the periodic opening of microvascular gates to allow TSPs to occupy a limiting number of thymic niches. He describes the likely existence of a feedback loop between the thymus and the bone marrow to induce release of TSPs from the marrow when thymic niches become vacant. The requirement for Notch1 signaling in postnatal T-cell development was revealed by conditional ablation and ectopic activation of Notch1 signaling. Two families of structurally distinct Notch ligands, Delta-like and Jagged, are expressed by cortical thymic epithelial cells. However, it is not known which Notch ligands are necessary for intrathymic T-cell development. Lunatic Fringe is a glycosyltransferase that enhances sensitivity of Notch receptors to Delta-like ligands. By genetically manipulating thymocyte expression of Lunatic Fringe, my laboratory has documented that TSPs in postnatal mice retain potent B-cell potential and adopt the T-cell fate intrathymically in response to Notch activation (8). Ectopic expression of Lunatic Fringe in cortical thymocytes alters the thymic microenvironment to impair Notch1 activation in TSPs, allowing many to aberrantly adopt the B-cell fate, after they enter the thymus near the corticomedullary junction (see cover illustration). Thus, suppression of B-cell development is the first order of business for TSPs. Additional studies from my group have shown that thymocytes must continuously compete for Notch1 signals throughout the DN phases of intrathymic T-cell development and suggest that Lunatic Fringe-mediated competition for limiting Delta-like ligands is an important mechanism for regulating thymus size. It has long been assumed that the complex three-dimensional thymic architecture is indispensable for T-cell development, but two articles describe experiments demonstrating otherwise. Schmitt and Zúñiga-Pflücker (9) have shown that overexpression of Delta-like-1 in the OP9 bone marrow stromal cell line abrogates its ability to support T-cell development and instead promotes T-cell development in a simple monolayer cell culture system. Their article describes the utility of this system for studying many aspects of T lymphopoiesis in vitro. Perreault and colleagues (10) describe their studies documenting that lymph nodes can be made to support thymus-independent T-cell development in mice overexpressing oncostatin M. This pathway differs from the thymus-dependent pathway in several important respects. It begins with different early progenitors, utilizes hematopoeitic rather than epithelial cells to mediate positive selection, and generates mature T cells with quite unusual functional properties. This extrathymic pathway may provide a means of generating T cells in athymic individuals. Progression through the different stages of T-cell development requires migration of T-cell precursors through distinct thymic microenvironments. Förster and colleagues (11) have defined an important role for the CCR7 chemokine receptor in regulating this directed migration and describe these studies as well as the roles of other chemokine gradients in regulating progenitor localization in the thymus. Progression from the DN through the intermediate and late stages of T-cell development requires expression of T-cell antigen receptors generated by V(D)J recombination. Jackson and Krangel (12) describe mechanisms regulating accessibility and recombination of the TCRβ locus, and they discuss recent progress defining the molecular basis by which recombination of TCR genes is developmentally ordered and allelically excluded. V(D)J recombination induces DNA damage, albeit site specific. Mice and humans lacking the ataxia-telangiectasia mutated (ATM) DNA damage-response pathway are highly susceptible to T-cell leukemia and are T-cell deficient. Matei and colleagues (13) discuss ATM diverse functions in regulating DNA damage checkpoints and describe their recent findings suggesting that ATM regulates survival of T-cell progenitors undergoing TCR rearrangements. It has been clear for several years that expression of γδ versus αβ TCRs at the DN2/3 thymocyte stage profoundly influences whether T-cell progenitors will remain DN and become γδ T cells or whether they will progress to the DP stage γδ T cells. Several articles in this issue discuss the role of signals through the γδ TCR versus pre-TCR in regulating early thymocyte development. Aifantis and coworkers (14) describe molecular mediators that regulate thymocyte proliferation and survival downstream of pre-TCR signaling, and they also discuss the possibility that aberrant pre-TCR signaling may contribute to T-cell leukemogenesis. Articles from the groups of Love (15) and Wiest (16) discuss their recent data demonstrating that signaling through γδ TCR versus pre-TCR differs quantitatively rather than qualitatively. Strong signals induce progenitors to become γδ T cell, whereas weaker signaling allows progenitors to enter the αβ T-cell lineage as DP thymocytes. Because most γδ TCRs induce stronger signals than the pre-TCR, lineage choice is generally correlated with TCR subtype, but these authors describe studies demonstrating that cells can be induced to adopt the wrong cell fate when the strength of TCR signaling is altered. Clearly, an important goal of future studies will be to determine how signal strength is regulated and calibrated by the γδ TCR versus pre-TCR, as well as how TCR signaling is integrated with other pathways active at this stage, such as IL-7 and Notch signaling. Cytokine, TCR, and Notch signaling pathways control survival, proliferation, and lineage specification and commitment of early T-cell progenitors by driving changes in transcriptional networks that control gene expression. M. Anderson (17) describes how early phases of T-cell development are controlled by successive waves of regulated gene expression. Anderson discusses specific roles of several transcription factors and regulatory modules in controlling expression of particular early T-cell genes and developmental transitions. Interestingly, evidence from Anderson and others suggests that the transcription factor networks active in DN thymocytes can be influenced by the presence or absence of DP thymocytes, providing evidence that downstream T-cell progenitors can have a feedback regulation on earlier progenitors. Such feedback regulation may explain, at least in part, some of the differences in phenotype and function between fetal versus adult DN thymocytes, which develop in the absence versus presence of DP thymocytes, respectively. Rothenberg and colleagues (18) further explore transcriptional regulatory differences between fetal and adult T-cell development. Although many aspects of T-lineage specification are constant in fetal and postnatal life, there appears to be significant regulatory flexibility in these pathways. Examples include differential dependence on extracellular survival signals such as thymic stromal-derived lymphopoietin and IL-7, as well as differentially stringent requirements for certain transcription factors. Analyses of these differences led them to suggest that changes in transcriptional networks induced by β selection may differ in the fetal and adult thymus. The remaining articles in this volume are concerned with selection of the αβTCR repertoire during the transition of DP thymocytes to the mature CD4 or CD8 single-positive (SP) stages. During this transition, precursors capable of recognizing self-MHC are positively selected for survival, and αβTCR specificity for class I versus class II MHC dictates the choice between the CD4 helper and the CD8 killer cell lineages. However, precursors recognizing self-MHC with high affinity undergo negative selection by clonal deletion to minimize the contribution of overtly autoreactive cells to the mature T-cell repertoire. Kappes and colleagues (19) review their recent discovery of the Th-Pok transcription factor, which is necessary and sufficient for commitment to the CD4 lineage. They discuss the impact of this finding on the debate between instructional versus stochastic/selective mechanisms for lineage commitment and speculate that Th-Pok may be induced by TCR signaling in DP thymocytes. However, it seems likely from previous studies that TCR signals controlling early aspects of positive selection (such as survival) may be distinct from those controlling later aspects such as CD4/CD8 lineage commitment. Aliahmad and Kaye (20) provide a more general review of how changes in expression and/or function of particular transcription factors are linked to TCR signaling in DP thymocytes. Of particular interest is the TCR-induced TOX transcription factor, which they have shown can promote CD8 lineage commitment even in the absence of TCR signals. Although Th-Pok and TOX are clearly important mediators of the CD4/CD8 lineage choice, it is not yet clear how DP thymocytes with different TCR specificities choose distinct lineages, because both the factors are induced by TCR signaling. One possibility is that TCR signals inducing these transcription factors are of distinct strength or duration. B. J. Fowlkes and colleagues (21) discuss studies they and others have performed which strongly support this notion. Interestingly, their recent results suggest that Notch activation enhances TCR signaling in DP thymocytes; hence, it seems likely that integration of Notch and TCR signals will be important for the CD4/CD8 lineage decision. Soon after Miller demonstrated that the thymus is an immunological organ, several groups demonstrated that self-tolerance largely occurs in the thymus by negative selection. Although several decades have now passed, the mechanisms and cellular mediators of negative selection remain poorly understood. An important experimental tool to study negative selection was the generation of αβTCR transgenic mice in which all T cells express a single TCR of defined peptide/MHC specificity. However, as discussed by von Boehmer and Kisielow (22), TCR expression is usually abnormally high and abnormally early in these mice; hence, it has been difficult to determine whether the developmental stage specificity and mechanism of negative selection in these models accurately recapitulate the process in normal mice. These investigators more recently have generated mice that express transgenic TCR in a more ‘timely’ physiological manner, and their studies suggest negative selection that can induce clonal deletion (apoptosis) of self-reactive cells at the DP and SP stages. Gallegos and Bevan (5) discuss mechanisms by which negative selection can induce central tolerance to tissue-specific self-antigens (TSAs) expressed primarily outside the thymus. They review studies documenting the surprising discovery that medullary thymic epithelial cells are programed to express a variety of TSAs, allowing them to centrally induce tolerance to these antigens. However, they have shown that dendritic cells can induce tolerance to more ubiquitously expressed self-antigens and can also cross-present TSA derived from medullary epithelium to induce tolerance. Thus, central tolerance to non-thymic antigens is achieved by collaboration between different antigen-presenting cell types. The articles in this issue highlight many of the current controversies and questions that still need to be answered to provide a complete mechanistic understanding of the thymus in health and disease. However, it is often true that the more one knows about a topic, the more one realizes how much is still to be learned. I have no doubt that as our experimental toolbox becomes ever more sophisticated, many of the unresolved issues discussed in this volume will be elucidated. I am equally certain that these new discoveries will provide new sources of controversy and reveal layers of mechanistic complexity to thymus and T-cell development that we cannot currently envision. Thus, we can be confident that the fruits of this scientific generation's labors will provide plenty of fodder to carry thymus research into the next century.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.