Evidence for TESC/TEPC in the postnatal thymus

Read Also

The first evidence for a common stem/progenitor cell activity in the postnatal thymus was provided by analysis of a subset of human thymic epithelial tumors that were found to contain cells that could generate both cortical and medullary sub-populations, suggesting that these arose from epithelial stem cells (92). These data, and subsequent data based mainly on shared marker expression in cultured cells have been extensively reviewed elsewhere (93) and will not be revisited here.

Further data regarding the phenotype of thymic epithelial progenitors came from analysis of mice with a secondary block in thymus development resulting from a primary T cell differentiation defect, which suggested that epithelial cells that co-express K5 and K8 have cTEC progenitor activity (33). This conclusion was based on the observation that in mice with complete, early block in thymocyte differentiation, the postnatal thymus is characterized by the predominance of K8⁺K5⁺ TECs. Normal cTEC development and architecture developed upon restoration of T cell differentiation, which lead to the conclusion that a precursor:progeny relationship existed between K8⁺K5⁺ TECs and K8⁺K5^- cTECs. It is unclear whether the TEC phenotype observed in these mouse mutants corresponds to an authentic differentiation arrest of a normally occurring TEC population, or reflects an abnormal state induced by the thymocyte differentiation block; also, the differentiation of cTEC from a minor subpopulation of undetermined phenotype in the mutant thymi cannot be excluded. Nevertheless, these data in conjunction with data indicating that all TEC in the developing thymic primordium at E11.5 share expression of these markers are consistent with the proposal that the K8⁺K5⁺ population of the adult thymus contains a TEPC activity. As K8 and K5 are co-expressed by only a population of cells at the cortico-medullary junction and scattered in the cortex in the postnatal thymus, their distribution pattern is also consistent with this hypothesis. Furthermore, a population of Plet1⁺ TEC also exists in the postnatal thymus as a minor subpopulation of mTEC, and overlaps partially with the medullary population of K8⁺K5⁺ TECs. Based on extrapolation from the characteristics of fetal Plet1⁺ TEC, it is tempting to speculate that these Plet1⁺K8⁺K5⁺ cells may be postnatal TEPC/TESC, however at present no data directly address this possibility.

The notion that the postnatal thymus is maintained principally by a common TEPC/TESC that generates both cortical and medullary TEC must be treated with caution, however, since the chimera study discussed above revealed no evidence for this type of activity (89). Furthermore, a lineage analysis based on low frequency epithelial-specific Cre recombination in the postnatal thymus demonstrated the existence of apparently clonal proliferative units that were limited to either medullary or cortical TEC, or spanned both cortical, cortico-medullary junction and medullary regions (87). These studies are often taken to provide support for sub-lineage restricted stem/progenitor cells in the postnatal thymus. However, neither demonstrates this definitively, as in both cases the data would also be consistent with proliferation of terminally differentiated epithelial cells - or a mixture of these activities. Also, the relatively early time of recombination (2 weeks postnatal) in the latter study leaves open the possibility that some or all of the activities detected are from residual fetal cells that are not maintained in the adult steady-state thymus. Similarly, while some recent data could be interpreted to support the existence of TE stem cells in the postnatal thymus, in each case the data are also consistent with proliferation of committed progenitor or terminally differentiated epithelial cells. For example, the changes seen in thymi upon castration demonstrate that regenerative capacity persists in the involuted thymus. While this would be consistent with persisting stem/progenitor cell activity, some mature cTEC and mTEC cells are still present in the involuted thymus, leaving open the possibility that regeneration could be due to their proliferation. In this regard, evidence that Fgf7 is mitogenic for all adult TECs provides a possible mechanism for proliferation-based expansion of differentiated cells during rebound (94). It is also possible that a facultative stem cell activity could be induced under regenerating conditions, similar to the oval cell activity induced by certain types of liver damage and the Ngn3⁺ ductal cell activity induced by partial duct ligation in the pancreas (95). Thus, at the present time, the question of whether the postnatal thymus is maintained during homeostasis by a stem cell or progenitor cell based mechanism, or by proliferation of terminally differentiated cells, remains unresolved.   

Several recent publications have begun to shed light on the molecular requirements for postnatal TEC maintenance and the potency of postnatal TEC. Analysis of the p63 knock out phenotype revealed the requirement for this protein, a critical regulator of the stratified epithelial cell program, in postnatal TEC. p63^-/- mice develop thymus hypoplasia that was suggested to result from loss of a TESC compartment (96). This study suggested an important connection between the mechanisms underpinning maintenance of the thymus and those that maintain other stratified epithelial cells. Recently, this issue has been explored further through functional experiments that demonstrated that postnatal rat TEC can be cultured clonally and indefinitely under the same culture conditions as skin/hair follicle keratinocytes (97). These cultured TEC, which as a population express Plet1 and a variety of stem cell-associated markers including p63, were shown to contribute to the thymic microenvironment in the presence of carrier fetal thymic epithelial cells. This assay was short-term, and therefore did not address whether the cultured cells had TESC activity. Furthermore, the contribution to these thymus reaggregates was largely limited to mTEC. However, the cultured rat TECs were able to contribute to both epidermal and hair follicle lineages in a skin morphogenetic assay, and could maintain this contribution for the long term. These cells therefore functioned as classical skin stem cells once reprogrammed to adopt the skin/hair follicle molecular program - notably, this reprogramming was induced by their microenvironment rather than by genetic intervention. While it is unclear at present how this study relates to the presence of epithelial stem cells in the postnatal thymus, the findings are of great interest and merit further investigation. In addition to these studies, a recent study has addressed whether a Foxn1-negative cell type may exist at the base of the postnatal TEC hierarchy (98), and concludes that postnatal TEPC express Foxn1, based on use of three different approaches to ablating Foxn1-expressing postnatal TEC. However, although technically elegant, these approaches do not exclude all scenarios for Foxn1-negative progenitors, including the possibility that Foxn1-positive niches are required to indirectly to maintain Foxn1-negative TESC.