Understanding the interplay between RUNX1 and Polycomb proteins in leukaemias with fusion proteins

RUNX1 is a critical regulator of the development of the haematopoietic system and a frequent target of genetic alteration in leukaemia. In addition, wildtype RUNX1 is required for maintenance of disease in leukaemias with fusion proteins, including MLL-AF4 and RUNX1-RUNX1T1 leukaemias. However, the exact mechanisms of how RUNX1 shapes the chromatin landscape and impacts gene regulation in normal and leukaemic cells remain elusive. In this thesis, it was investigated how RUNX1 binding affects the recruitment of gene-regulatory proteins and impacts the transcriptional regulation of its target genes, and how its binding is affected by the chromatin landscape. Using a refined intracellular tethering system, it was shown that RUNX1 alone is sufficient to recruit co-activators and the transcriptional machinery, and also co-repressors, including Polycomb group (PcG) proteins. In turn, it was shown that the chromatin environment affects the ability of RUNX1 to bind to its binding sites, as gains in the repressive histone marks H3K27me3 and H3K9me3 reduce RUNX1 binding. In addition, the existing chromatin landscape impacts RUNX1-mediated recruitment of co-factors. More specifically, at active sites RUNX1 promotes recruitment of co-activators, while at PcG-repressed genes RUNX1 binding stabilises PcG recruitment. By further investigating RUNX1-mediated recruitment of PcG proteins, it was shown that a key function of RUNX1 is the stabilisation of Polycomb Repressive Complex 1 (PRC1) in MLL-AF4 ALL and RUNX1-RUNX1T1 AML. PcG mutations are found in a subset of RUNX1-RUNX1T1 and RUNX1-mutant leukaemia patients and are often associated with a worse clinical outcome, but the mechanistic details of the functional RUNX1-PRC1 cooperation have not been investigated. In this thesis, it is shown that the RUNX1-RUNX1T1 fusion protein, and other truncated RUNX1 mutant proteins lacking the RUNX1 C-terminal region, cannot associate with PcG proteins. In these leukaemias, a key role of wildtype RUNX1 is to stabilise PcG binding at PcG-repressed genes, including a subset of genes upregulated in PcG-mutant RUNX1-RUNX1T1 patients. Overall, these results demonstrate that an important function of RUNX1 is to maintain PcG repression and that this activity is disrupted in RUNX1-mutant leukaemias, particularly those with PcG mutations, thus providing a mechanistic explanation for the positive selection of leukaemic clones with RUNX1/PcG mutations