In the mouse thymus, newly formed single positive (SP) cells spend an average of 14 days in the thymic medulla. During this time, phenotypic and functional maturation occurs with down-regulation of CD69 and heat stable antigen (HSA), and up-regulation of Qa-2. Very little is known about the final steps that allow or direct these T cells to emigrate and join the recirculating peripheral T cell pool. Currently available data suggest that not all recent thymic emigrants (RTE) complete this maturational sequence in the medulla and that emigration may occur at any time during the medullary maturation stage. In this study, we have compared adhesion and activation marker expression on SP thymocytes, RTE and peripheral T cells to determine more precisely which SP medullary thymocytes are exported. Although RTE were heterogeneous for HSA and Qa-2 expression, they were quite uniform with regard to the expression of other molecules. In contrast to medullary SP thymocytes, most RTE were L-selectin(high) and CD69-. In addition, CD4+ CD8- and CD4- CD8+ RTE were phenotypically distinct from each other in that the former were beta7 integrin(-/low), CD45RB(intermediate) and CD45RC-, while the latter were beta7 integrin(high), CD45RB(high) and CD45RC(low). These phenotypes were comparable to only a minor (as little as 6%) subpopulation of medullary SP thymocytes. Overall, the data indicate that export of cells from the medullary pool of SP thymocytes is not random, but that a series of maturational events within the SP stage are necessary before export can occur.

The peripheral T cell pool is maintained both by export of naive T cells from the thymus and by post-thymic expansion of activated/memory T cells. However, it is not known whether the thymus can alter its output following peripheral T cell depletion. Using intrathymic injection of fluorescein isothiocyanate to detect recent thymic emigrants (RTE), we directly tested whether the thymus is able to alter the number of RTE or the CD4:CD8 ratio of RTE emigrating to the periphery in response to in vivo depletion of total peripheral T cells or CD4 T cells, respectively. Depletion of peripheral T cells was achieved with anti-Thy-1 or anti-CD4, at doses that did not affect thymocyte numbers. Depletion of greater than 70% of peripheral T cells by treatment with anti-Thy-1 in vivo did not alter the number or cell cycle status of RTE trafficking to lymph nodes or spleen during the peripheral reconstitution phase (6, 9, 12 days). Similarly, depletion of the majority of CD4 T cells, which significantly reduced the peripheral CD4:CD8 T cell ratio, did not alter the total number or the proportion of CD4+ CD8- RTE in peripheral lymphoid organs. These data clearly indicate that thymic output is not influenced by downstream alterations in peripheral T cell pool size or CD4:CD8 ratio. Rather we contend that thymic T cell export is internally regulated by as yet undefined mechanisms.

TNF, lymphotoxin (LT) and their receptors are expressed constitutively in the thymus. It remains unclear whether these cytokines play a role in normal thymic structure or function. We have investigated thymocyte differentiation, selection and thymic organogenesis in gene targeted mice lacking LTα, TNF, or both (TNF/LTα-/-). The thymus was normal in TNF/LTα-/- mice with regard to cell yields and stromal architecture. Detailed analysis of αβ and γδ T cell-lineage thymocyte subsets revealed no abnormalities, implying that neither TNF nor LT play an essential role in T cell differentiation or positive selection. The number and distribution of thymic CD11c+ dendritic cells was also normal in the absence of both TNF and LTα. A three-fold increase in B cell numbers was observed consistently in the TNF/LTα-/- thymus. This phenotype was due entirely to the LTα deficiency and associated with changes in the hemopoietic compartment, rather than the thymic stromal compartment of LTα-/- mice. Finally, specific Vβ8+ T cell deletion within the thymus following intrathymic injection of staphylococcal enterotoxin B (SEB) was TNF/LT independent. Thus, despite the presence of these cytokines and their receptors in the normal thymus, there appears no essential role for either TNF or LT in development of organ structure or for those processes associated with T cell repertoire selection.

Lymphotoxin (LT)‐α, a member of the TNF family, is recognized as an important mediator in different aspects of lymphoid organ development. Targeted disruption of this molecule resulted in a substantial reduction in the proportion of αEβ7‐integrinhigh CD8+ T cells detectable in peripheral lymphoid organs. This defect, however, was not observed on mature CD4–CD8+ thymocytes. To determine whether this was due to downregulation of β7‐integrin expression by peripheral CD8+ T cells or a failure of thymic emigration of CD8+ β7‐integrinhigh T cells, β7‐integrin was examined on recent thymic emigrants (RTE). When analysed within 16 h after leaving the thymus CD4–CD8+ RTE in both LT‐α–/– and wild type (wt) mice remained β7‐integrinhigh and were indistinguishable. However, within 3–5 days, emigration loss of β7‐integrin became evident in LT‐α–/– mice. Despite this loss, the proportion of thymically derived αβTCR+ T‐cell populations in the intestinal epithelium, an important target tissue of CD8+ αEβ7‐integrinhigh T cells, was increased in the absence of LT‐α. In contrast, B cells were detectable only rarely in the intestinal tissue of LT‐α–/– mice. The expression of E‐Cadherin remained unchanged. These results indicate that a LT‐α‐dependent process maintains a high level of αEβ7‐integrin expression by peripheral CD8+ T cells, and with this control mechanism LT‐α may help to regulate CD8+ T‐cell numbers in the tissues.