Loss of BCOR Protein in Development of Acute Myeloid Leukaemia

Acute myeloid leukaemia (AML) is an aggressive haematological malignancy with a dismal clinical outcome and represents an unmet medical need. However, therapeutic progress will not be made without a greater understanding of the varied molecular pathogenesis of this heterogeneous disease. Recently, the majority of cases of the largest subgroup of AML, AML with a normal karyotype (AML-NK), were found to harbour recurrent loss-of-function mutations in epigenetic regulators.

Germline mutations in the BCL6 Co-Repressor (BCOR) gene, situated at Xq27-28, are associated with two overlapping X-linked dominant disorders, the Oculo-facio-cardio-dental (OFCD) syndrome and syndromic micropthalmia. In addition, somatic mutations in BCOR were described within the last few years in adults and children with AML, particularly in cases arising in the context of aplastic anaemia and in patients with antecedent myelodysplastic syndromes.

BCOR encodes a ubiquitously expressed protein which is involved in BCL6-mediated transcriptional repression. In contrast to other BCL6 partners, NCOR and SMRT, it was recently demonstrated to occupy the majority of its DNA binding sites independently from BCL6, reflecting its multiple interaction partners. BCOR also forms a Polycomb repressor complex 1 (PRC1)-like complex with PCGF1, KDM2B, RING1, SKP1, RYBP, and RNF2, and can bind to both class I and II histone deacetylases. Its spectrum of protein-binding partners and studies of mutant forms suggest that BCOR is an epigenetic modulator. Therefore, we predict that loss of BCOR in AML results in loss of its complex repressive function and leads to an activation of an important leukaemogenic transcriptional programme.

To test this hypothesis, we generated a mouse model of leukaemia using CRISPR-Cas9 genome editing. Using a common endonuclease assay, we validated 6 different gRNA constructs against Bcor targeting the region of known human mutations described in AML. The three most efficient constructs were subsequently used to create a pool of lentiviral vectors. In order to increase the efficiency of targeting the haematopoietic progenitor cells, we harvested lineage depleted bone marrow cells from mice expressing the Cas9 cDNA from the Rosa locus (Rosa-Cas9). In parallel, we performed the same experiments with Rosa-Cas9 mice harbouring a Flt3 internal tandem duplication (Flt3-ITD), a mutation that can co-occur with BCOR mutations in individual AML patients.

First, we transduced bone marrow lineage negative cells with lentiviruses carrying the three BCOR sgRNAs at 60-70% efficiency and performed a serial re-plating assay to investigate the clonogenic potential of targeted cells. We observed a significant increase in the colony numbers in both Cas9 only and Cas9 Flt3-ITD targeted cells in comparison to the empty lentiviral vector controls from the second re-plating onwards. Using a standard endonuclease assay we have confirmed the presence of Cas9-induced mutations within the target Bcor exons, indicative of clonal selection for Bcor targeted cells and an increased self-renewal potential of hematopoietic precursors that lack BCOR. Transduced cells were also injected into lethally irradiated recipient mice in the presence of helper cells to investigate development of leukaemia in vivo. We have been able to confirm the presence of the Bcor mutated cells in the blood of tested mice 12 weeks post injection and these mice continue to be monitored for evidence of leukemia.

We are currently conducting integrated genomic analysis with ChIP-Seq, RNA-Seq and methylation analysis to link the BCOR mutations and their altered cellular phenotype to aberrant epigenetic control and pre/leukaemic gene expression programmes. Finally, we are also going to assess co-operativity of Flt3-ITD and BCOR mutations in AML leukaemogenesis. We anticipate that our work will inform the biology of AML-NK associated with BCOR, identify potential therapeutic targets and provide platforms for the assessment of novel therapeutics.