The thymus: a house dividing

Comment on Gray et al, page 3777

The relative radioresistance of thymic epithelium has led to the assumption that this population is static and exhibits little turnover during postnatal life.

Recent studies demonstrating significant proliferation by thymic epithelial cells in the postnatal thymus have challenged the notion of a static population.^1,2  In this issue of Blood, Gray and colleagues present data that underscore the dynamic nature of thymic epithelium in the postnatal thymus and that provide new insight into several fundamental questions of thymic epithelial cell (TEC) biology. Here, they demonstrate that the rate of thymic epithelial turnover approaches that of the epidermis, with a replacement time of 10 to 14 days. This means that the composition and organization of thymic epithelial compartment represents a “snapshot” of the dynamic relationship between progenitor cells, transit-amplifying cells, and loss of cells from terminally differentiated epithelial compartments. From this perspective, thymic involution becomes less mysterious and can now be viewed as a failure of progenitor and/or transit-amplifying TECs to maintain the thymic epithelial compartment at steady-state levels. The reversal of thymic involution following castration can be considered as another perturbation of the dynamics of thymic epithelial (TE) differentiation, where the populations of progenitor and/or transit-amplifying epithelial cells are expanded in response to castration-induced alterations of the hormonal milieu.FIG1 

The mitotic activity of TE subsets defined in this report helps to model the developmental relationship between subsets of thymic epithelium. Gray and colleagues have demonstrated that the MHC II^hi (CD80^hi) medullary TEC (mTEC) subset is heterogeneous and exhibits greater mitotic activity than the MHC II^lo subset, and have suggested that the former population may represent a transit-amplifying population. Furthermore, the temporal relationship observed between the appearance of MHC II^hi and MHC^lo mTECs may reflect a precursor-product relationship. The developmental relationship suggested by these data is presented in the figure.

As levels of tissue-restricted antigen (TRA) expression are highest in the MHC II^hi (CD80^hi) subset of mTECs,³  this model of TE differentiation associates these properties with immature mTECs, rather than terminally differentiated mTECs as has been proposed.³  However, the program of mTEC differentiation suggested by these data is consistent with the notion that TRA expression reflects the activity of immature mTECs.^1,4  It follows that the death of thymic epithelial cells that would accompany this dynamic epithelial turnover may facilitate the cross presentation of mTEC-derived antigens by bone marrow–derived antigen-presenting cells within the medullary compartment⁵  and thus enhance the efficiency of this negative selection process. This scenario would reconcile the efficient negative selection of TRA-specific thymocytes that has been previously observed^6,7  with the established rarity and spatial distribution of the epithelial cells that express them.^1,7,8 

This report also presents provocative data regarding the antigen-presenting capacity of different thymic epithelial subsets. The prevalent view of cortical TEC (cTEC) and mTEC function proposes a division of labor, where cTECs mediate positive selection and function poorly as conventional antigen-presenting cells, while mTECs can serve as antigen-presenting cells and mediate negative selection but do not mediate positive selection.⁹  In demonstrating that isolated cTECs can effectively present peptide antigens to TCR transgenic T cells, Gray and colleagues challenge this widely held view of cTEC properties. If validated, this observation will require a re-evaluation of the role of cortical epithelial cells in shaping the T-cell antigen receptor specificity repertoire. However, it will be important to demonstrate that the level of dendritic cell contamination present in the cTEC populations does not contribute to the T-cell responses observed. In sum, this study contributes significantly to a dynamic view of thymic epithelium that will influence the design of new models and experimental approaches to understand the differentiation and function of thymic epithelium.

The authors declare no competing financial interests. ▪