Deletion of the Artery Specific Enhancer Promotes B Cell and Impairs T Cell Development from the Hemogenic Endothelium

HSCs and HSC-independent lymphoid progenitors in the embryo arise from hemogenic endothelium (HE) at multiple anatomical locations. Although activation of arterial program in HE has been shown to play an important role in HSC development, the impact of the arterial program on specification of HSC-independent lymphoid cell remains unclear. During vascular development, arterial fate is specified following induction of DLL4 expression by VEGF by signaling through a highly conserved VEGF-responsive artery-specific enhancer located within the third intron of DLL4 (DLL4-i3) which is controlled by ETS, SOXF and RBPJ factors. To investigate how the arterial program affects T and B cell lymphoid development, we generated DLL4-i3:Venus reporter and knockout (DLL4^Di3/Di3) human pluripotent stem cells (hPSCs). Using a 3-dimensional, serum- and feeder-free spheroid differentiation system, we demonstrated that DLL4: i3 enhancer activity recapitulates endogenous DLL4 expression and arterial endothelial cell specification in human hPSC cultures. Moreover, DLL4-i3: Venus⁺ HE and hematopoietic progenitors produced more myeloid CFCs and T cells as compared DLL4-i3: Venus^- hemogenic cells. Deletion of DLL4i3 enhancer impaired arterial specification from hPSCs including formation of arterial DLL4⁺CXCR4^+/- HE. In contrast to wild type (WT), hematopoietic progenitors (HPs) generated from DLL4^Di3/Di3 hPSCs showed impaired CFC potential with significantly reduced G-CFCs and markedly diminished T cell potential. However, DLL4^Di3/Di3 hPSCs showed efficient and highly reproducible B cell potential as compared to WT. To define the impact of artery-specific enhancer deletion on transcriptional program during endothelial-to-hematopoietic transition (EHT) and HP formation, we performed scRNAseq of all endothelial and hematopoietic cells from D5 and D8 differentiation cultures of DLL4-i3:Venus and DLL4^Di3/Di3 hPSCs. Through unsupervised clustering, 10 cell clusters (C0-C9) were distinguished and visualized using UMAP. Based on the expression panel of known pan-, arterial and hemogenic endothelial and hematopoietic markers we defined three clusters of HE with arterial features (AHE; high EFNB2, CDH5 and RUNX1 positive and SPN negative), two non-arterial HE clusters (low EFNB2, CDH5 and RUNX1 positive and SPN negative), two EHT clusters (CDH5 and SPN positive, high RUNX1, and GFI1b), and three HP clusters (SPN positive, high RUNX1 and CDH5 negative). GSEA of DLL4i3: Venus+ cells with active enhancer vs DLL4^Di3/Di3 cells revealed enrichment in ribosomes and oxidative phosphorylation pathways across all clusters and focal adhesion and tight junction pathways in HE and EHT clusters. DEG analysis demonstrated upregulation of EFNB2, MECOM, HES4 in DLL4-i3: Venus+ cells within HE and EHT clusters, while DLL4^Di3/Di3 cells demonstrated upregulation of NRP2 and CDH13 genes. Interestingly, in EHT and HP clusters, we noticed that DLL4-i3 knockout significantly upregulates the expression of GAS5 lncRNAs known to be involved in growth arrest of human peripheral blood T cells. Using siRNA, we demonstrated that knockdown of GAS5 in D5 cultures of DLL4^Di3/Di3 cells rescued the T cell potential of the hematopoietic progenitors suggesting that NOTCH-mediated GAS5 suppression is essential for the T cell specification and expansion following EHT. In contrast to T cells, GAS5 downregulation had no effect on B cell specification of DLL4^Di3/Di3 HE. Overall, we established that DLL4-i3 enhancer is required for arterial programming of HE with T cell potential while deletion of this enhancers promotes B cell development. In addition, we demonstrated an important role of GAS5 lncRNA in the regulation of T cell development from HE.