Epigenomic Mechanisms of Action of Novel Mutant Isocitrate Dehydrogenase Inhibitors in Acute Myeloid Leukaemia

Acute myeloid leukaemia (AML) is a most common aggressive leukaemia in adults and is incurable in most patients. In AML, gene mutations cause immature cells in bone marrow to stop making mature cells (differentiation block and bone marrow failure) and to increase is numbers in patient bone marrow (expansion). Patients develop symptoms of anaemia, bleeding and infections, which lead to death is untreated. Most effective treatments for AML use chemotherapy drugs which kill leukaemic cells. These are highly toxic as they also harm other cells in the body. The majority of patients with AML are elderly and unable to tolerate these types of treatment. We therefore need to have more effective, less toxic treatments to restore blood and bone marrow function (disease remission) to both improve rates of cure, prolong patient survival, improve quality of life.

20% of AML patients have a mutation in genes coding for the enzymes isocitrate dehydrogenase 1 or 2 (IDH1/2). The mutant enzyme (mIDH) produces an abnormal chemical (or metabolite), dextro-2-hydroxyglutarate (d2HG). This metabolite is known to cause cancer (including AML and brain tumours), most probably by stopping cells from maturing properly.

New drugs which inhibit mIDH (mIDHi), is effective for ~40% of Isocitrate Dehydrogenase mutant (IDHm) AML patients. It works by reducing the number of immature AML cells by causing them to become useful mature blood cells. This benefits patients by reducing the need for blood transfusions and infection risk. Unfortunately, most patients who initially respond to mIDHi will develop resistance and relapse. We do not understand why this happens, or why some patients never respond to mIDHi. We also do not fully understand how mIDHi work, but it is likely to involve changes to the control mechanisms of genes

To address these fundamental questions, I want to understand what goes wrong in the control mechanism of genes which are usually expressed when blood cells mature, and how this causes AML. The aims of my proposal is to
1) Study how blood cells differentiate and mature in normal bone marrow and how this process of differentiation and maturation goes wrong in AML (i.e. differentiation block)
2) Investigate how drugs like mIDHi are able to re-programme AML cells to make them differentiate into functional mature cells
3) In patients where mIDHi are not effective, or when a patient's disease becomes resistant to mIDHi, find out why AML cells remain, or become blocked. If I can discover how we can target these blocked pathways, for example, by using novel drugs or by combining the effects of mIDHi with other drugs, then this could be useful for treating patients.

The methods I will use to study the behaviour of AML cells include using genetic sequencing techniques to look at how genes are expressed in cells, and also what the mechanisms are which control gene expression (epigenomics). This could provide a better characterisation and understanding of AML cells and be used to determine if AML patients are more or less likely to respond to a particular treatment. This can help clinicians to decide which treatments are best for an individual patient, and help develop combinations of treatments which are more likely to work for an individual.

Aims and objectives
1. Investigate how abnormal epigenetic gene regulatory mechanisms result in differentiation block and leukaemic transformation in IDHm AML.
2. Investigate the molecular mechanisms of action of mutant IDH inhibitors in inducing differentiation of AML, focusing on effects on transcription factors involved in normal granulocyte-monocyte (GM) differentiation.
Methodology
1. Compare gene expression, DNA methylation and chromatin accessibility, performed in parallel (multi-genomics) in normal bone marrow versus AML. Perform unbiased genome-wide bioinformatic analysis, and analysis focused on differentially expressed transcription factors required for normal GM differentiation (GM-TFs) but suppressed in AML.
2. Test the hypothesis that failure to express GM-TFs results in leukaemic transformation using CRISPR knockdown of candidate GM-TFs in IDH1-mutant preleukaemic mouse model.
3. Perform multi-genomic sequencing on mIDHi-treated AML LSC in vitro to investigate how mIDHi upregulates GM-TFs. Identify GM-TFs upregulated with mIDHi in responsive AML samples but which remain repressed in non-responsive samples. Test if re-expression restores response in resistant AML samples.
4. Correlate GM-TFs required for response (from point 3) with gene expression in IDHm AML patients who are resistant to mIDHi treatment in vivo. These may be important targets for novel therapies.

Applications and benefits
This work will improve our understanding of the molecular mechanisms of differentiation block and its contribution to pathogenesis of AML. It will help explain why differentiation therapies may not cure patients. The data has the potential to identify new genes/ pathways that should be targeted by novel therapies to improve clinical outcomes. It will provide the rationale for design of future clinical trials of combination therapy. Multi-genomic characterisation of patients may be useful as a prognostic tool, and to guide therapeutic decisions.

Work from this proposal has the potential to benefit several groups beyond those working in the field of Haemato-oncology.

The wider Scientific Community: The proposal is cross-disciplinary and brings together experts on AML biology, epigenomic gene regulation, DNA biochemistry, cellular metabolism and cancer genomics. Output of this research will not only be of interest to those working on AML, but also solid tumour cancer biologists and scientists interested in fundamental principles of gene regulation, lineage commitment and cell fate determination. I am proposing an innovative, integrated approach to study AML biology through combining functional experiments with multi-genomic sequencing techniques to study primary patient samples. Data generated by this proposal will be an important resource to colleagues and facilitate future studies in primary patient samples.

Clinicians: impact on improving health outcomes : Inhibitors of mutant IDH (mIDHi) are currently in phase 2/3 clinical trials for AML patients. They represent a new therapeutic paradigm for non-APL AML patients. However, we need to understand how best to deploy these agents to maximise benefit for patients. This proposal seeks to address why resistance to mIDHi occurs, to inform rational design of combination therapies and to identify new targets for differentiating agents of the future. Multi-genomic characterisation of patient disease could be developed as a better diagnostic and prognostic tool to guide therapeutic choices.

Industry: Pharmaceutical/ Biotechnology : The multi-disciplinary approach in this proposal illustrates how we can obtain maximum translational research data which is directly relevant to patient care from precious and limited primary patient samples. These data can inform design of new treatments and of future clinical trials of combination therapy.

Patient groups and members of the public : New therapies bring hope to patients but are often not fully effective. I am proposing to understand the mechanisms of action of mIDHi so that clinicians know are able to use these drugs more effectively. In the longer term, this work seeks to identify new therapeutic targets for patients whose disease is resistant, and those with other subtypes of AML. Findings from this proposal may also have implications for patients with other cancers.

Education and training : This proposal draws on a wide range of state-of-the-art technologies to address important questions about AML biology. It will provide high quality training for a post-doctoral research assistant, and through collaborations with groups from other fields will promote cross-disciplinary engagement and exchange of ideas.