Recent Progress 2015

Understanding how blood cells are generated has clear implications for the treatment of blood diseases.  Such knowledge could potentially lead to defining new conditions to amplify haematopoietic stem cells (HSCs) or could translate into new methods to produce HSCs, or other types of blood cells, from human embryonic stem (ES) cells or induced pluripotent stem (iPS) cells.  Additionally, as most key transcription factors (TFs) regulating early haematopoietic development have also been implicated in various types of leukaemias, elucidating their function during normal development could result in a better understanding of their roles during abnormal haematopoiesis in leukaemia. In particular, the genes encoding the AML1/RUNX1 transcription factor and its cofactor CBFβ are master regulators of blood development and are frequently rearranged or mutated in human leukaemias such as acute myeloid leukaemia (AML) and acute lymphoblastic leukaemia (ALL). Similarly, the transcriptional co-activator MOZ that is involved in three independent myeloid chromosomal translocations fusing MOZ to the partner genes CBP, P300 or TIF2 in human leukaemia is a critical regulator of blood development. Our group studies, in particular, the function of RUNX1 and MOZ in haematopoietic development and maintenance in order to better understand how alterations of these functions lead to leukaemogenesis.

RUNX1 regulates a cell adhesion and migration prior to blood emergence
The earliest site of blood cell development in the mouse embryo is the yolk sac where blood islands, consisting of haematopoietic cells surrounded by a layer of angioblasts, develop at approximately day 7.5 of gestation.  The parallel development of these two lineages in close association provided the basis for the hypothesis that they arise from a common precursor, a cell called the haemangioblast. A conflicting theory however associates the first haematopoietic cells to a phenotypically differentiated endothelial cell with haematopoietic potential, i.e. a haemogenic endothelium.  Support for the haemangioblast concept was initially provided by the identification during embryonic stem (ES) cells differentiation of a clonal precursor, the blast colony-forming cell (BL-CFC), which gives rise (after 4 days of culture) to blast colonies with both endothelial, smooth muscle and haematopoietic potential.  We recently established a new model of haematopoietic development. We demonstrated that haematopoietic cells are generated from the haemangioblast through the formation of a haemogenic endothelium intermediate. During this process, the haemogenic endothelial cells first cluster and then lose their endothelial identity, altering their flat, adherent appearance into the characteristic round shape of mobile haematopoietic precursor cells.


Figure 1: Identification of Genes bound by RUNX1 and Identification of Genes differentially expressed in the absence of RUNX1 in Haemogenic Endothelium. (A) Schematic Representation of RUNX1b DamID system. Top. The RUNX1b::Dam fusion binds RUNX1 binding sites and the Dam methylates adenines within nearby GATC sequences. The methylated fragments are isolated using the methylation specific enzyme DpnI, amplified by PCR, subjected to high throughput sequencing and the sequences analysed with the DamID Peak Identification Pipeline. Non-specific binding is controlled by sequencing of untethered Dam samples. (B) Validation of detection of RUNX1 binding at some known transcriptional targets. Raw sequencing read data from RUNX1b:Dam and control Dam haemogenic endothelium samples. Red bars indicate the RUNX1b:Dam peaks as identified by the DamID-PIP pipeline. RUNX1b::Dam binds to Sox17 and AI467606 promoter region and the +35Kb enhancer element of Gfi1, which are all known RUNX1 targets. (C and D) Genes Differentially Expressed between Runx1 WT and Runx1 KO haemogenic endothelium. (C) Heat maps and (D) GSEA analyses.

The identification of the earliest transcriptional programme regulated by RUNX1 would be key to understanding the onset of haematopoiesis. Previous studies aiming to reveal the initial RUNX1 programme have however been largely hampered by the fact that the haemogenic endothelium represents a rare, transient subset of endothelial cells with low endogenous RUNX1 expression. In addition, the RUNX1-dependent transition to haematopoiesis takes place rapidly, making it difficult to distinguish between immediate effects of RUNX1 in haemogenic endothelium cells and later direct or indirect effects in committed blood progenitors. To overcome these caveats, we coupled DamID (DNA adenine methyltransferase identification) with high-throughput sequencing, to map RUNX1 binding sites in haemogenic endothelium. This alternative to ChIP relies on the deposition of "methylation tags" around the binding sites of RUNX1 by the E.coli DNA adenine methyltransferase (Dam) (Figure 1). The stability of the methylation marks and the alleviation of the need for antibodies, make this technique ideal for binding site analysis of low expressed DNA-binding proteins in small populations. Compared to ChIP-seq, DamID-seq generated more control peaks (untethered Dam vs ChIP IgG or input control) and wider peaks overall, rendering existing ChIP-seq analysis packages unsuitable.  We therefore developed a new TF specific DamID Peak Identification Pipeline (DamID-PIP) in collaboration with the group of Dr Crispin Miller (CRUK Manchester Institute). Integration of the haemogenic endothelium specific RUNX1-DamID binding profile with matching transcriptome data (Figure 2) revealed that RUNX1 binds to and up-regulates the expression of genes involved in cell adhesion and migration, including components of the integrin signaling pathway. This suggests that at this early stage of haematopoietic development, RUNX1 organises the formation of haemogenic endothelium clusters required for the release of blood progenitors. Overall, this study provides the first comprehensive genome-wide RUNX1 target profiling in the early haemogenic endothelium and demonstrates that RUNX1 acts in a stage-specific fashion by activating adhesion and migration associated genes prior to the emergence of haematopoietic cells and the down-regulation of the endothelial programme. Outside of the haematopoietic context, this endothelial-epithelial RUNX1 signature in haemogenic endothelium might also reflect the emerging role of RUNX1 in epithelial-based tumour formation and progression. In particular the RUNX1 targets that associate with cell migration in haemogenic endothelium may represent important regulators of the potential metastatic role of RUNX in solid tumours.

Figure 2: RUNX1b Binds to and Positively Regulates Genes associated with Cell Adhesion, Cell Migration and ECM Interaction. (A) Venn diagram showing the overlap between RUNX1-bound genes in haemogenic endothelium (HE), as determined by DamID-seq, and RUNX1b-dependent differentially expressed genes in haemogenic endothelium. Heatmap depiction of the 235 genes that are both bound and regulated by RUNX1b in haemogenic endothelium. RUNX1b- bound and positively correlated genes are depicted in red and RUNX1b-bound and negatively correlated genes are depicted in blue.  All genes are ordered from higher to lower expression. (B) IPA on RUNX1b-bound and positively correlated genes (top panel) and on RUNX1b-bound and negatively correlated genes (bottom panel).