Conserved and Unique Regulatory Mechanisms of RUNX1 in Hematopoietic Stem/Progenitor Cell Subpopulations in Both Normal Hematopoiesis and FPD/AML Development

RUNX1 is a RUNT domain core-binding factor indispensable for the establishment of definitive hematopoiesis in vertebrate. In humans, heterozygous RUNX1 germline mutations are closely associated with Familial Platelet Disorder with propensity to develop AML (FPD/AML), a rare autosomal dominant disorder characterized by thrombocytopenia, platelet dysfunction, and markedly elevated risk for myelodysplastic syndrome and leukemia. RUNX1 regulates hematopoietic genes directly or indirectly, acting as a context-dependent activator or repressor of gene transcription. Here we have examined how RUNX1 regulates various subpopulations of hematopoietic stem/progenitor cells (HSPCs) to provide important insights into its physiologic and pathogenic roles and mechanisms. We used two complementary mouse models: a conditional RUNX1 knockout mouse, i,e, Mx-Cre;Runx1flox/flox, and a tetracycline-inducible FPD/AML patient derived RUNX1 S291fsX300 mutation knock-in crossed to a MLL-PTD knock-in mouse, i.e. MLL-PTD;RUNX1 S291fsX300. We monitored hematopoiesis and/or disease progression in both models by tracking CBCs in peripheral blood and analyzing bone marrow cells by flow cytometry. The bone marrow samples were sorted by flow cytometry to yield HSPC subpopulations: HSC (LSK CD34-Flt3-), GMP (LK FCRII/III+CD34+), and MEP (LK FCRII/III–CD34-). Cells were analyzed for transcriptome by deep RNA-seq, epigenetics by ATAC-seq, and gene targets by anti-RUNX1 CHIP-seq. Focusing on the HSC, GMP, and MEP subpopulations, we found that RUNX1 deletion in the hematopoietic system causes thrombocytopenia and HSC engraftment defects, and dramatic changes in HSPC subpopulations: RUNX1 deficiency resulted in an increase MegaK-primed HSCs via blocking their direct differentiation to megakaryocytes leading to thrombocytopenia, and caused a marked increase of GMPs and a significant decrease of MEPs. Induction of RUNX1 S291fsX300 mutant in MLL-PTD;RUNX1 S291fsX300 mice mimicked that of RUNX1 deletion, resulting similarly in thrombocytopenia, an expansion of Mega-HSCs, increased GMPs, and decreased MEPs. “Correction” of the RUNX1S291fsX300 mutation by doxycycline withdraw reversed the FPD phenotype as well as the cellular distribution changes in HSCs, GMPs and MEPs. RNA-seq analyses of DEGs found that RUNX1 acts both as an activator and a suppressor in gene transcription in HSPCs. In HSCs, among 1879 DEGs, 608 are upregulated and 1271 are downregulated in Runx1^-/- compared with WT cells. In GMPs, 2042 DEGs were seen, among which 908 are upregulated and 1134 are downregulated whereas in MEPs, 316 transcripts were upregulated and 1214 were downregulated amongst 1530 DEGs in Runx1^-/- cells. Combining the DEGs from RNA-seq with annotated ATAC-seq peaks, we discovered through gene ontology analyses that a set of systemic development genes are commonly regulated by RUNX1 via chromosome accessibility in all HSPC subpopulations, with potentially shared targets such as Jam2, Bcar1, Irf4, Insrr, Cxcl12, Tnfsf4, Rap1gap, Aldh1a1. In contrast, in HSCs, a unique set of hemostasis/platelet regulatory genes such as Itga2b, Gp5, Pdgfb, and Vwf were seen as potential targets, consistent with our recent findings (Blood Adv 7:2590). In GMPs, RUNX1 appears to specifically regulate inflammatory and innate immunity genes with potential targets such as Gfi1b, Ets1, Tnfrsf11a, while in MEP, RUNX1 specifically regulates cell motility and migration gene sets affecting targets such as Notch1, Ifitm1, Hoxa7, Cd300a, Cxcl10, Fut7, suggesting a role of RUNX1 in modulating MEP movement/location in the bone marrow. RUNX1 CHIP-seq analyses are under way to further clarify the potential targets in each HSPC subpopulation. Our studies indicate that RUNX1 regulates common as well as distinct transcriptional events in various HSPC subpopulations by context-dependent activation and suppression mechanisms. RUNX1 plays a universal role in the systemic developmental program of various HSPC subpopulations but appears to engage in unique lineage specific factors to deliver subpopulation-specific functions such as mega-K cell fate-priming of HSCs, inflammatory program in GMPs, and bone marrow location signals in MEPs. Ongoing analyses of the FPD/AML mouse genomic data in this context are expected to reveal contributions of patient RUNX1 mutations to the disease development.