The Lineage Differentiation and Dynamic Heterogeneity of Thymic Epithelial Cells During Thymus Organogenesis

Abstract

Although much progress has been made recently in revealing the heterogeneity of the thymic stromal components, the molecular programs of cell lineage divergency and temporal dynamics of thymic epithelial cell (TEC) development are largely elusive. Here, we constructed a single-cell transcriptional landscape of non-hematopoietic cells from mouse thymus spanning embryonic to adult stages, producing transcriptomes of 30,959 TECs. We resolved the transcriptional heterogeneity of developing TECs and highlighted the molecular nature of early TEC lineage determination and cortico-medullary thymic epithelial cell lineage divergency. We further characterized the differentiation dynamics of TECs by clarification of molecularly distinct cell states in the thymus developing trajectory. We also identified a population of Bpifa1+ Plet1+ mTECs that was preserved during thymus organogenesis and highly expressed tissue-resident adult stem cell markers. Finally, we highlighted the expression of Aire-dependent tissue-restricted antigens mainly in Aire+ Csn2+ mTECs and Spink5+ Dmkn+ mTECs in postnatal thymus. Overall, our data provided a comprehensive characterization of cell lineage differentiation, maturation, and temporal dynamics of thymic epithelial cells during thymus organogenesis.

Key Factors for Thymic Function and Development

“…Though both of these approaches have shown the identity of the main TEC populations, many unknown TEC subtypes, in particular, thymus tuft cells, were identified by scRNA-seq in both mice ( 62 , 92 ) and humans ( 63 ). In the recent study authors additionally identified a population of Bpifa1 + Plet1 + mTECs that was preserved during thymus organogenesis in mice, and these mTECs highly expressed tissue-resident adult stem cell markers ( 93 ). Depending on the levels of MHCII and CD80, mTECs are broadly subdivided into mTECs lo ( Cldn4 , lower levels of HLA class II) and mTECs hi ( Spib, Aire, Fezf2 , higher levels of HLA class II).…”

Sex biased human thymic architecture guides T cell development through spatially defined niches

“…Recent single cell sequencing resources demonstrating the diversity of human thymus tissue are incongruous with our current framework of thymus structure and organization [7][8][9][10][11][12][13][14][15][16][17][18][19] which describes a general migratory path a thymocyte takes through the cortex and medulla during conventional αβT cell development. Recent spatial transcriptomic sequencing of human fetal and postnatal thymus has demonstrated that a deeper granularity of thymic niches is present within the tissue and that these niches evolve during fetal development to support different waves of non-conventional T cells 19,20 .…”

Thymic epithelial organoids mediate T cell development

“…Unsupervised clustering identified 7 main clusters of epithelial cells (Fig. S3A), which we annotated according to in vivo datasets (Baran-Gale et al, 2020; Bautista et al, 2021; Bornstein et al, 2018; Gao et al, 2022; Nusser et al, 2022; Park et al, 2020): ‘early cTECs’, ‘cTECs’, ‘early mTECs’, ‘pre-Aire mTECs’, ‘Aire and Spink5 mTECs’, ‘tuft-like mTECs’, and ‘adult bipotent progenitor-like’ (Fig. S3B).…”

“…Although thymic epithelial organoids recapitulate many key organoid features such as cell population diversity and possibility to be expanded and passaged, maintaining their functionality in the long term remains challenging. This is probably linked to some niche factors missing in the current relatively minimal culture conditions, which over time either generally prevent functionality to be maintained or enrich for specific subsets that might lack functionality (Gao et al, 2022). Future work including single-cell transcriptomic analysis of the organoids will most likely help identify yet unexplored but necessary niche factors.…”

“…Each cell type was then subset and thresholded based on EpCAM , Ptprc and Pdgfra expression. Epithelial clusters were identified using 18 dimensions and a resolution of 0.4, leading to 7 clusters that were named based on markers from previous datasets (Baran-Gale et al, 2020; Bautista et al, 2021; Bornstein et al, 2018; Gao et al, 2022; Kernfeld et al, 2018; Park et al, 2020). Immune clusters were identified using 18 dimensions and a resolution of 3.…”