Bipotency of thymic epithelial progenitors comes in sequence


Full correspondence: Dr. Pärt Peterson, Professor of Molecular Immunology, Molecular Pathology, Institute of Biomedicine, Ravila 19, Biomedicum, University of Tartu, Tartu, 50411, Estonia

Fax: +372 737 4212


See accompanying Commentary by Baik et al.


In the thymus, in order to become MHC-restricted self-tolerant T cells, developing thymocytes need to interact with cortical and medullary thymic epithelial cells (TECs). Although the presence of a common bipotent progenitor for these functionally and structurally distinct epithelial subsets has been clearly established, the initial developmental stages of these bipotent cells have not been well characterized. In this issue of the European Journal of Immunology, Baik et al. [Eur. J. Immunol. 2013.43: 589–594] focus on the phenotypical changes of the early bipotent populations and show how the cortical and medullary markers are sequentially acquired during TEC development. These findings argue against a binary model in which both cortical and medullary lineages diverge simultaneously from lineage-negative TEC progenitors and highlight an unexpected overlap in the phenotypic properties of these bipotent TECs with their lineage-restricted counterparts.

The essential function of the thymus is to generate and select new T cells with functional and self-tolerant TCRs for proper adaptive immune responses. During embryogenesis, the thymus, together with parathyroid glands, originates from the third embryonic pharyngeal pouch, and in the mouse, starts to form around embryonic day E10 and E11 of gestation. After this stage, the interaction between the endoderm-derived thymic epithelium, neural crest-derived mesenchymal cells and immigrated hematopoietic precursors leads to the thymic expansion and compartmentalization into the outer cortical and inner medullary regions encompassing cortical and medullary thymic epithelial cells (cTECs and mTECs), respectively. The cTECs are primarily responsible for the generation and survival of the positively selected CD4+ CD8+ immature T-cell pool with an immunocompetent TCR repertoire, whereas the main function of mTECs and medullary DCs is to secure the negative selection of self-reactive T cells. The two epithelial cell types are morphologically and functionally distinct, nevertheless, the evidence for their common bipotent progenitor cells has started to accumulate during recent years. A paper by Baik et al. published in this issue of the European Journal of Immunology [1] adds new evidence and perspectives to our understanding of the bipotent thymic epithelial progenitor cell (TEPC) differentiation and lineage marker expression.

The early differentiation of TEPC depends on a transcriptional program activated by the transcription factor FoxN1; in mice with Foxn1 mutations TECs do not develop and thymopoiesis is blocked [2]. The transcriptional regulation of the later dichotomy of cTECs and mTECs has remained thus far unknown. What is known is that the separation between cTECs and mTECs is associated with changes in their keratin expression patterns. Though not absolutely, keratin K8+ K5 cells are predominantly cTECs and K8K5+ cells are mTECs, whereas K8+K5+ cells, as well as K14+ cells, are often considered as epithelial precursor cells at fetal stages [3, 4]. In the adult thymus, K8+K5+ cells are present at the cortico–medullary junction but their potency as progenitor cells is unknown. Other epithelial markers have proven to be informative tools in the identification of epithelial cell phenotypes. For example, cTECs express proteosomal subunit beta-5t (encoded by Pmsb11), Ly-51/CD249 (Enpep), delta-like ligand 4 (Dll4), serine protease 16 (Prss16) and CD205 (DEC-205, Ly75) with the earliest cTEC-specific markers detectable at E12. In contrast, the markers associated with mTECs are tight junction proteins claudin-3 and -4 (Cldn3 and 4) and lectin UEA1 with commitment to mTEC lineage at E13. The differentiation and full maturation of mTECs critically depends on RANK signaling that stimulates the expression of CD80, MHC class II, CD40 and Aire, all needed to promote tolerance towards self-antigens (reviewed in [5, 6]).

The presence of a large pool of thymic epithelial cells in the early thymus expressing cTEC and mTEC markers has been considered as an indication that both epithelial cell types share a common bipotent progenitor cell [7]. The clonal progenitor activity was initially described for the mTEC lineage using chimeric mice [8]. The existence of bipotent TEPCs was first indirectly addressed by the transplantation of bulk reaggregated thymic organ cultures under the kidney capsule [9-11], the direct evidence came from using a clonal assay with single thymic epithelial cells expressing yellow fluorescent protein (YFP) [12]. The YFP-expressing single epithelial cells, when microinjected into non-fluorescent E12 host thymuses and transplanted into recipient mice, gave rise to both cTEC and mTEC lineages (but not to a single cTEC or a single mTEC lineage) in the fully developed thymuses [12]. The pool of bipotent embryonic progenitors seems to be restricted, possibly due to the limited capacity for proliferation or lack of suitable stem cell niches [13], and is not compensated for in later developmental stages if depleted in during embryogenesis [14]. The evidence for bipotent and unipotent epithelial progenitors in the postnatal thymus was found using lineage-tracing based on the human K14 promoter driving Cre-recombinase and a reporter mouse that activates YFP only after Cre-mediated genomic rearrangement [15]. The rare activation of Cre-recombinase in epithelial progenitors, and hence the labeling of these cells with YFP in postnatal mice, created epithelial cell clusters containing only mTECs, only cTECs or both mTECs and cTECs. The capability of a single TEPC to generate a functional postnatal thymic microenvironment was further shown by reverting dormant single cells in a FoxN1-deficient thymus to cells expressing FoxN1 [15].

The paper by Baik et al. [1] now provides novel information on the sequential marker acquisition at the early stages of TEPC development. Using reporter mice with the green fluorescent protein driven by Foxn1 promoter (Foxn1:eGFP), the authors were able to monitor GFP expression from E11 onwards, thus covering the early transition into the TEPC phenotype (Fig. 1). Baik et al. [1] show that at the E11–E12 days of development, a distinct population of progenitors acquires CD205, a marker specific for mature cTECs. The changes in TEPC phenotype continue at E13, when the TEPC population starts to express CD40 and are accordingly positive for both the cTEC and mTEC markers. At E14, the TEPC population downregulates the expression of CD205 and remains positive for CD40, thus resembling the surface expression pattern of mTECs. To further show that the CD205+CD40 progenitor cells can give rise to mTECs, Baik et al. [1] examined the responsiveness of these CD205+CD40 progenitor cells to RANK signaling using agonistic antibody. Indeed, the cells responded to RANK stimulation with enhanced expression of CD40 and MHC class II as seen in mTEC differentiation. Most importantly, CD205+CD40 cells were able to form a functionally organized thymus microenvironment in transplantation experiments, with the expression of beta-5t and CD205 in cortical and CD80 and Aire in medullary epithelium. Collectively, these results demonstrate the plasticity of the thymic epithelium and establish CD205 as a marker for bipotent embryonic TEPCs.

Figure 1.

Sequential marker acquisition at the early stages of TEC development based on the data by Baik et al. [1]. The expression pattern of the cTEC marker CD205 and the mTEC marker CD40 is shown for the embryonic TEPC from E11 through E14 (blue cells). Bipotency was demonstrated for the CD205+CD40 population at E12, developing into both mature mTECs (turquoise) and cTECs (purple) (the bipotency of later progenitor populations remains to be studied).

What is the importance of this and earlier studies on TEPCs? As pointed out by the authors, the identification of progenitor markers and their consecutive expression pattern is an important step in the analysis of the prospective potential of TEPCs, as well as for establishing strategies to regenerate a functional thymus with full cortical and medullary compartments. Thymus transplantation is a promising therapy for the treatment of DiGeorge syndrome-associated immunodeficiency [16], and a recent report, using postnatal allograft transplantation, hinted at the role of K14+ and human cTEC-marker CDR2-positive epithelial cells in the reconstitution of the thymus allograft [17]. Certainly, the next step would be the identification of the progenitor markers in the adult thymus as this would have practical implications for human thymus transplantation and for the restoration of T-cell immunocompetence. Despite the fact that the thymus starts involution soon after birth and becomes atrophic with age [18], the adult thymic epithelium is constantly regenerated from a pool of adult progenitor cells, albeit with decreasing efficiency [7]. Thus, the capacity for renewed thymopoiesis is not lost with aging and could be restored therapeutically [19]. Different treatment strategies with growth factors (growth hormone, IGF-1, and FGF-7), IL-7 or sex steroids have been already applied in diverse experimental systems to improve age-related loss of thymic function (reviewed in [20]).

The differentiation of thymic epithelium shares features and markers with other epithelial tissues, including skin or mammary epithelial cells [21-23]. In this respect, lineage-tracing analysis of progenitor cells from mammary epithelium with cytokeratin promoters, has revealed the existence of a K14+ multi-potent progenitor at an early embryonic stage, whereas postnatal and adult development are ensured by K14/K5+ and K8/K18+ unipotent stem cells that differentiate into myoepithelial and luminal lineages, respectively, and are no longer maintained by rare multi-potent progenitors [24]. The shift from bipotent stem cell prevalence at embryonic stage to unipotent or compartment-specific progenitors at postnatal and adult tissues may well take place in thymus too—the rapid turnover and the capacity to regenerate after the selective ablation indicate the potency of cTEC and mTEC lineage-specific progenitors in the postnatal and adult thymus [25, 26].

The study by Baik et al. [1] raises unanswered questions, namely the persistence of embryonic bipotent TEPCs and the relation of these TEPCs to the bi- or unipotent progenitors in the adult thymus. The cTEC/mTEC marker pattern, identified here, should be useful for further isolation and then characterization of the progenitors. Finally, the bipotent TEPC (and possible cTEC lineage progenitor) specificity for CD205, an endocytic C-type lectin-like molecule with a role in the recognition of apoptotic cells for antigen uptake and processing [27] warrants further characterization.


The authors thank the European Regional Fund/Archimedes Foundation and the Estonian Research Council funding IUT2–2 for their support.

Conflict of interest

The authors declare no financial or commercial conflict of interest.


thymic epithelial cell


thymic epithelial progenitor cell


yellow fluorescent protein