Runx-mediated gene regulation during thymic T cell development and stage specific genomic interactions of Runx3

Runx transcription factors play pivotal roles in the development of hematopoietic cells, including T cells. However, the Runx-target genes in early stages of thymic T cell development have been only partially elucidated. Moreover, the relationships of different Runx paralogs, whether they compete or collaborate with each other, are unknown. We have performed CRISPR-Cas9 mediated acute deletion of Runx1 and Runx3, alone and in combination, to define the functions of Runx factors and their target genes. We focused on two distinct stages of early T cell development; before lineage commitment (Phase1) and after commitment (Phase2). Our data suggest that Runx1 and Runx3 are functionally redundant in Phase1, hence the absence of one factor was compensated by the other factor. In Phase2, Runx1 becomes a major contributor, while low levels of Runx3 still strengthened the Runx1-mediated target gene regulation and showed an additive effect. Of importance, we found that Runx1 and Runx3 upregulate T cell programs and inhibit alternative lineage genes. In order to capture the developmental status of Runx-perturbed cells, we performed supervised principal component analysis (PCA) using a fixed frame of PC loadings from single cell RNA-seq analysis of normal in vivo-derived thymocytes. Unlike the reference in vitro and in vivo pro-T cell population or sgControl introduced cells, Runx knockout cells either failed to progress fully (Phase1) or deflected from the normal developmental trajectory (Phase2), suggesting essential roles of Runx factors in early stages of T development. In addition, we analyzed the correlation between stage-specific genomic occupancy of Runx1 and Runx3 (using previously published data for Runx1, GSE103953) and Runx-mediated gene regulation responses at each stage. Interestingly, genes with dynamic, stage-sensitive expression showed a very significant association with phase-specific genomic interactions of Runx factors, and an especially strong correlation with Runx activated target genes. In summary, our data suggest a pivotal role for Runx transcription factors in thymic T cell development by imposing T lineage specification in Phase1 and instructing the lineage-committed progenitor cells to progress through a normal T developmental trajectory. ChIP-seq: Bone-marrow progenitor cells from C57BL/6J wild-type animals were cultured on OP9-DLL1 stromal cells with IL-7 and Flt3-ligand to promote T-lineage development. The following populations were FACS sorted for ChIP-seq: Phase1 cells: Lineage marker (Lin)- CD45+ cKithigh CD25- cells on day5 Phase2 cells: Lin- CD45+ cKitlowCD25+ cells on day10. Approximately 5-10 million cells were fixed at RT in 1 mg/ml DSG followed by addition of 1% formaldehyde and subjected to ChIP-seq. RNA-seq after Phase1 and Phase2 Runx-perturbation: Bone-marrow progenitor cells from B6.CRISPR-Cas9/Bcl2 transgene expressing mice were cultured on OP9-DLL1 stromal cells with IL-7 and Flt3-ligand for 2 days (for Phase1 perturbation) or 10 days (for Phase2 perturbation). Three sgRNAs against Runx1 and/or three sgRNAs against Runx3 were retrovirally introduced either to target a single gene or to target both in combination. After 3 days post infection, Lineage marker (Lin)- infection marker+ cKit high (for Phase1 perturbation) or Lin - infection marker+ CD25+ (for Phase2 perturbation) cells were FACS sorted and subjected to bulk RNA-seq. RNA-seq for in vitro ETP, Bcl11b- DN2a, and Bcl11b+ DN2b pro-T cells: Bone-marrow progenitor cells from B6.Bcl11b-YFP reporter animals were cultured on OP9-DLL1 stromal cells with IL-7 and Flt3-ligand for 4 or 6 days. The following gating strategy was applied for FACS sorting of the indicated populations: ETP: Lin- cKit high CD44 high CD25- Bcl11b(YFP)- Bcl11b- DN2a: Lin- cKit high CD44 high CD25+ Bcl11b(YFP)- Bcl11b+ DN2a: Lin- cKit high CD44 high CD25+ Bcl11b(YFP)+ Sorted cells were subjected to bulk RNA-seq analysis.