Interaction with the DNA Binding Domain of MLL and a Small Molecular Drug Induces Proteolysis of MLL-Fusion Proteins and Ablates Leukemic Stem Cell Activity

Introduction

Leukemias harbouring 11q23 abnormalities causing mixed-lineage leukaemia gene (MLL) rearrangements are associated with poor clinical outcomes. Despite being an aggressive leukaemia, the MLL rearranged infant ALL has among the lowest mutation rates reported for any cancer. This means that to improve survival for patients with this aggressive leukaemia we need drugs that target the abnormal proteins produced by the MLL fusion gene or that interact with the abnormal MLL fusion protein to shut down the cellular machinery that drives these leukemias. Indeed, targeting MLL fusion dependent gene pathways has become a major focus.

Our previous studies have shown that inhibition of MLL-fusions, in a conditional mouse model of MLL-ENL driven acute myeloid leukaemia, resulted in a block in self-renewal of the leukemic cells and ablated the leukaemia in the mice. This led us to hypothesise that if, we could achieve pharmacological inactivation of the MLL fusion proteins, we could achieve improved clinical outcomes. To achieve this, we set out a drug screening programme in acute leukaemia with the aim to discover drugs that can inactivate MLL-fusion oncoproteins.

Results

Our drug discovery pipeline screened clinical approved drugs for their ability to inhibit the function of the MLL fusion protein. This lead to the discovery of a drug that interacts with the DNA binding domain within the MLL fusion protein. This interaction destabilises the MLL fusion protein so that the fusion protein gets degraded within 24 hours of addition of the drug. So far, we have shown that we can inhibit and induce the degradation of MLL-AF9, MLL-AF6 and MLL-AF4 (and WT MLL) in the human MLL rearranged cell lines (THP-1, SHI-I and MV4-11), in primary immortalized cells in which the MLL-AF9 is overexpressed from a lentiviral backbone and in patient derived leukemic samples.

Inactivation/degradation of the MLL fusion protein should shut down the cellular machinery that drives these leukemias. It is well established that MLL-fusions lead to abhorrent upregulation of its target genes HOXA9, MEIS1 and c-MYB. Treatment of MLL rearranged cells resulted in the downregulation of these MLL-fusion target genes within 24hrs of addition of the drug. Furthermore, Gene Set Enrichment Analysis of drug treated MLL-AF9 cells showed strong negative enrichment to various published MLL fusion target gene sets. Inactivation of MLL fusion protein should also result in block in self-renewal as we have previously shown in our conditional mouse model. Indeed, Gene Set Enrichment Analysis showed negative enrichment to published Leukemic Stem Cell gene set. To analyse the impact of drug treatment on self-renewal, we used a well-established self-renewal assay, whereby self-renewal is assessed by their ability to form colonies derived from single cells in methylcellulose. While treatment had no significant impact on the colony formation of CD34 positive cord blood progenitors, the drug was able to block the colony formation ability of MLL rearranged cell lines while only slowing a slight reduction in in the colony numbers of non MLL rearranged cell lines.

Conclusion

Overall, the data indicates that we may have discovered a new targeted treatment for MLL rearranged leukemia, which shows excellent clinical properties. We have successfully generated Patient Derived Xenografts (PDX) models and we are currently testing this drug to verify its effectiveness in the treatment on PDX. We will include this data in our presentation.