Development of double-positive thymocytes at single-cell resolution

Abstract

Background T cells generated from thymopoiesis are essential for the immune system, and recent single-cell studies have contributed to our understanding of the development of thymocytes at the genetic and epigenetic levels. However, the development of double-positive (DP) T cells, which comprise the majority of thymocytes, has not been well investigated. Methods We applied single-cell sequencing to mouse thymocytes and analyzed the transcriptome data using Seurat. By applying unsupervised clustering, we defined thymocyte subtypes and validated DP cell subtypes by flow cytometry. We classified the cell cycle phases of each cell according to expression of cell cycle phase-specific genes. For immune synapse detection, we used immunofluorescent staining and ImageStream-based flow cytometry. We studied and integrated human thymocyte data to verify the conservation of our findings and also performed cross-species comparisons to examine species-specific gene regulation. Results We classified blast, rearrangement, and selection subtypes of DP thymocytes and used the surface markers CD2 and Ly6d to identify these subtypes by flow cytometry. Based on this new classification, we found that the proliferation of blast DP cells is quite different from that of double-positive cells and other cell types, which tend to exit the cell cycle after a single round. At the DP cell selection stage, we observed that CD8-associated immune synapses formed between thymocytes, indicating that CD8sp selection occurred among thymocytes themselves. Moreover, cross-species comparison revealed species-specific transcription factors (TFs) that contribute to the transcriptional differences of thymocytes from humans and mice. Conclusions Our study classified DP thymocyte subtypes of different developmental stages and provided new insight into the development of DP thymocytes at single-cell resolution, furthering our knowledge of the fundamental immunological process of thymopoiesis.

From pseudotime to true dynamics: reconstructing a real-time axis for T cells differentiation

“…A similar but delayed pattern is seen in expression of cleaved-PARP1 (cPARP), which is produced during both apoptosis and necrosis(Elmore, 2007) (Figure 6C). Proliferation, identified by IDU incorporation, also exhibits a bi-wave profile, where cells divide more extensively around day 5 during the third phase of the DN state (DN3)(Laurent et al, 2004; Paiva et al, 2021), and then again at the DP state(Ciofani and Zúñiga-Pflücker, 2007; Li et al, 2021) between days 15 and 20 (Figure 6B).…”

“…Proliferation, identified by IDU incorporation, also exhibits a bi-wave profile, where cells divide more extensively around day 5 during the third phase of the DN state (DN3) (Laurent et al, 2004;Paiva et al, 2021), and then again at the DP state (Ciofani and Zúñiga-Pflücker, 2007;Li et al, 2021) between days 15 and 20 (Figure 6B).…”

The transcriptional regulator Sin3A balances IL-17A and Foxp3 expression in primary CD4 T cells

“…However, it should be noted that while DN thymocytes undergo a rapid and extensive proliferative phase (6-8 cell cycle division) to ensure the generation of a large pool of progenitors, DP cells enter a resting phase to allow rearrangement of the TCR alpha chain (Taghon et al , 2006; Ciofani & Zúñiga-Pflücker, 2007). Indeed, by performing single cell RNA-seq analyses Li et al found that up to 40% of intermediate SP were in the M/G1 phase, while this was not the case for DP blasts (Li et al , 2021). Thus, it could be speculated that deletion of Sin3A at the DN stage (Cowley et al , 2005) results in the selection of a cell subset capable to survive positive thymic selection, and yet lacking proper responsiveness at the mature stage.…”

Calcium Signaling Is Impaired in PTEN-Deficient T Cell Acute Lymphoblastic Leukemia

“…The final cell clusters were identified using Seurat 'FindNeighbors' and 'FindClusters' (with Louvain Algorithm option) functions with the first 20 PCs and coarse grain resolution of 0.7. Each computed cell clusters were assigned to thymopoiesis stages by manual curation, using thymopoiesis stage markers previously identified (19)(20)(21)(22) and Immgen Database (23). To infer pathway activity, genes sets of pathways of interest (Supplementary Table 2) were retrieved from Molecular Signature Database (MSigDB) (24) and were used to score pathway activity for each cell with AUCell version 1.12.0 (25).…”

“…After sequencing data preprocessing, which includes the removal of poor quality cells, of celldoublets and non-T cells (see Materials and Methods, Single-Cell RNA Sequencing Data Analysis), we projected the data using the UMAP non-linear dimensionality reduction method that enabled us to visualize cell transcriptome heterogeneity (17) (Figure 1B). We identified 8 clusters and assigned them to cell type according to established markers of thymopoiesis (19)(20)(21)(22)(23) (Figures 1C, D; see also Supplementary Figure 2 for unsupervised cluster gene markers). Cluster 1 encompasses DN, ISP and DP blast cells, cluster 2 corresponds to DP small , cluster 4 to DP CD69+ cells, and cluster 5 to naive CD4 + cells.…”