The differential role of PU.1 in normal and malignant haematopoiesis: from master regulator of differentiation to coordinator of leukaemia networks

The behaviour of cells is determined by the proteins that they contain. This protein composition greatly differs between cell types and even between cells within tissues and explains the different cell behaviours and functions that we observe. The levels of these proteins and why different proteins are present in specific cells are determined by tightly-regulated processes called transcription, where an intermediate product called a transcript is generated from a gene's DNA-template, and translation, where the transcript is then used to assemble the specific protein coded for by that gene. The process of transcription has been well studied in blood formation. Blood is a very complex tissue, containing different specialised cells that carry oxygen, cells that cause the blood to clot and cells that fight infection. However, every cell within the blood is generated from a single source: the haematopoietic stem cell (HSC), with the differing cells maturing and becoming more specialised through an ordered process called differentiation. In general, differentiation is orchestrated by specialised proteins called transcription factors that bind to and regulate the DNA of specific genes whose protein products are necessary for specialisation. The binding of the transcription factors then instruct the generation of transcripts from that gene, resulting in protein expression. The expression of these transcription factors is also tightly controlled and changes through differentiation and specialisation.

One of the cardinal features of leukaemias, and most cancers, is that this ordered process of differentiation is blocked and in effect, no specialised mature blood cells are formed. Acute leukaemias, such as Acute Myeloid Leukaemia (AML), are aggressive cancers with a dismal survival rate and are thus an unmet medical need. We have recently discovered that a transcription factor called PU.1 that is usually associated with normal blood differentiation is also absolutely required to maintain AML, a cellular state that lacks differentiation. This project will focus on answering the interesting question of why PU.1 drives differentiation in normal cells yet maintains AML. This seems at first-glance counterintuitive, however we speculate that: 1) differences in levels of the PU.1 protein, 2) differences in its binding to and regulation of different target genes and 3) differences in the proteins that it interacts with and alterations in its structure (called post-translational modifications) between the normal and leukaemia state explain this altered behaviour.

We will interrogate these avenues in this proposal, utilising state-of-the-art techniques in relevant model systems and multiple AML patient samples. Moreover, we will target PU.1 and its interaction partners with specific drugs and genetic perturbations to determine if it may be a relevant and safe therapeutic target in patients with AML. Therefore, the likely outcomes of this project are that we will significantly improve our understanding both of leukaemia biology and of the normal differentiation of blood cells, that we will identify PU.1 and its interacting proteins as potential therapeutic targets in AML and potentially that we may suggest an effective therapy in this aggressive disease.

In normal haematopoiesis, the ETS TF PU.1 is a master regulator of myeloid differentiation. However, we have recently demonstrated that in many AML subtypes, PU.1 is required for leukaemia maintenance. We propose that this relates to the involvement of PU.1 in regulating different genes and networks in each context, likely related to its presence within different gene regulatory complexes and/or possibly via differential post-translational modifications. Elucidation of these mechanistic differences will inform normal haematopoiesis and AML biology and is likely to have therapeutic implications for leukaemia.

We will utilise state-of-the-art genomic (CUT&RUN, HiChIP, bulk ATAC- and RNA-Seq, linked Multi-omics-single nucleus RNA-Seq and ATAC-Seq), proteomic (qPLEX-RIME, Phosphoproteomics) and functional genomic techniques (Perturb-Seq) in a relevant murine experimental system representative of the most common AML genotype (Npm1c/Flt3-ITD mutated), complemented by extensive corroboration across human patient samples and cell lines with a wide range of AML genotypes. The proposal is composed of 4 objectives:

1. Determine alterations of PU.1 binding pattern and 3D-genome coordination during the evolution of murine AML, with corroboration in human AML.

2. Link PU.1 activity to differential chromatin accessibility and downstream gene expression.

3. Determine the protein interactome and post translational modifications (PTM) of PU.1 in normal and malignant haematopoiesis.

4. Construction and perturbation of PU.1 networks in AML.

Taken together this project will greatly improve our understanding of both normal myeloid differentiation and AML biology, will likely confirm PU.1 as a poitential target in AML and may further identify specific interaction partners as therapeutic targets across AML.