MKpoiesis is a very original differentiation process leading to the generation of platelets  (1011 platelets daily in human adult). Multipotent HSCs of the bone marrow commit into megakaryocytic progenitors, which differentiate into promegakaryoblast (PMKB) precursors.  At this step the classical system of mitosis switches to endomitosis, which is a unique process that allows to increase DNA content from 2N to 4N, 8N, 16N, …, without cell division. This gives rise to polyploid megakaryoblasts (MKBs) that differentiate into megakaryocytes (MKs). At the end of endomitosis, the cytoplasm and thus MK size expand due to increased protein synthesis and organelle development (including alpha and dense granules). At the end of differentiation, the platelets are generated by fragmentation of MK cytoplasmic membrane in the blood. MKpoiesis is mainly regulated by thrombopoietin (TPO) and the TPO receptor (MPL)/JAK2 signaling pathway. MPL is a homodimeric type I receptor without intrinsic tyrosine kinase activity. The soluble Janus tyrosine kinase JAK2 is essential in mediating MPL phosphorylation leading to the activation of downstream signaling pathways (STAT, ERK, PI3K/AKT). All the steps of MKpoiesis are finely regulated by several transcription factors (TFs) and epigenetic regulators.

1) Normal MKpoiesis: Role of TFs and epigenetic regulators in MKpoiesis
(Hana Raslova & William Vainchenker)
We focus on RUNX1, FLI1 and SRF TFs regulating early and/or late stages of MKpoiesis (RNAi approach, ChIP-seq) to better understand the mechanisms by which the genetic alterations in these TFs, particularly those in RUNX1 and FLI1, induce the thrombocytopenia.
As mutations in some epigenetic regulators induce abnormalities of the MKpoiesis (EZH2, ASXL1, ASXL2) we are studying their roles in the regulation of MKpoiesis by using RNAi and pharmacological inhibition in primary human cells or using KO mouse models. 

2) Pathological MKpoiesis: Thrombopenia

Genetic alterations of the MK lineage can lead to decrease in platelet number (thrombocytopenia) and platelet functions (thrombopathy).

• Mechanisms implicated in inherited thrombocytopenia (Hana Raslova)

We model different inherited thrombocytopenias through induced pluripotent stem cells (iPSCs) that are differentiated towards MKs in 2D cultures as well as in 3D-silk based scaffold. We developed a consortium SilkFusion (https://silkfusion.eu/) aiming not only to develop a new technology to produce platelets at large scale for transfusion but also to decipher the mechanisms implicated in these thrombocytopenias allowing to identify druggable targets with the goal to improve patient management (collaborations Alessandra Balduini, Pavia, Italy; Cédric Ghevaert, Cambridge, UK). In term of personalized medicine, the platelet production from primary progenitors- or iPSC-derived MKs in miniaturized 3D model could predict the response to identified treatments. 

• Understanding the pathogenesis of a new type of acquired thrombocytopenia, previously classified in immune thrombocytopenia (ITP) (William Vainchenker) 

This type of thrombocytopenia is due to an intrinsic abnormality of MK differentiation. Our work hypothesis is that this disorder is a peculiar form of clonal hematopoiesis. This is done in collaboration with the team of Chloé James (Bordeaux).

• Identifying new platelet antigens (human platelet antigens, HPA) responsible of fetal and neonatal alloimmune thrombocytopenia (William Vainchenker) 

For this part, we are using a molecular approach. This work is performed in collaboration with the department of platelet Immunology (Rachel Peterman) at the INTS (Paris) 

3) Pathological MKpoiesis: Myeloproliferative neoplasms (MPNs)

Pathologies of the MK lineage can also lead to increased platelet number (thrombocytosis) and myelofibrosis (MF) which is due to an excessive production of fibrotic tissue by the bone marrow stromal cells in response to inflammatory cytokines released by the MKs.

Those conditions are found in MPNs, which are studied by our lab since many years. MPNs are a group of clonal hematological malignancies (polycythemia vera (PV), essential thrombocytemia (ET) and primary myelofibrosis (PMF)) characterized by increased proliferation of myeloid lineages, leading to abnormally high number of mature blood cells. The major genetic alterations have been identified in 90% of cases with mutations in 3 genes: JAK2V617F, MPL and calreticulin (CALR). These MPN driver mutations all cause deregulated activation of the JAK signaling pathways.

• Determining MPN initiation by identifying germline predisposition and by characterizing their functions  (Isabelle Plo, Christine Bellanné-Chantelot & Antonio di Stefano)

Familial forms of MPNs resemble MPN sporadic cases in terms of clinical and molecular features. In particular, they present JAK2V617F, MPL and CALR mutations. Since MPNs display a low incidence of development, the fact that several cases aggregate in families suggests that germline variants contribute to the initiation and/or development of the disease in these families. We identified several germline variants after pangenomic analyses.  We described a germline 700-kb duplication of the chromosome 14 including ATG2B and GSKIP genes, that strongly predisposes to MPN in families from French West Indies. We study the function of these predisposition factors using mouse and iPSC modeling as well as primary cell from patients. 

• Understanding the driver functions in MPN physiopathology (Isabelle Plo & Caroline Marty) 

In the laboratory we have identified the JAK2V617F mutation in MPN, and have decipher the function of JAK2, MPL and more recently CALR mutations on hematopoiesis. The two most frequent CALR mutations found in ET and PMF patients are a deletion of 52 bp (del52) or type 1 and a 5 bp insertion (ins5) or type 2. We have shown that CALR mutants specifically bind and activate MPL and the JAK2/STAT pathway that controls MK differentiation, thus explaining the tropism of these mutations. We are now deciphering the role of these mutations in human and mouse hematopoiesis.

      - Role of CALRdel52 (Type 1) and ins5 (Type 2) mutations in MPN in mouse (Caroline Marty)

We have developed 2 conditional KI mouse models that can express CALRdel52 or ins5 at a heterozygous or homozygous endogenous level. Using these models, our main objective is to understand the mechanisms involved in their phenotypic differences (thrombocytosis and MF), especially at the HSC level using techniques such as competitive engraftments, mass cytometry and single cell RNAseq.

      - Role of CALRdel52 (Type 1) and ins5 (Type 2) mutations in MPN in human
(Isabelle Plo, William Vainchenker, Christophe Marzac & Florence Pasquier)

We investigate the role of CALR mutations in human hematopoiesis particularly on stem cells as well as their mechanism of action using primary cells from patients (cell culture, PDX models, iPSC models).

• Deciphering the mechanisms of MF (Hana Raslova)

It takes several months to current MPN mouse models to develop MF. We recently developed a model of accelerated MF by crossing p19 KO with Jak2V617F mice. These mice develop a MF in only 2 weeks after JAK2V617F induction. We will use this model to determine the mechanisms implicated in MF, by studying dysmegakaryopoiesis and cytokine secretion.

Our ultimate goal is to improve MPN and AML treatments though a better understanding of the mechanisms implicated in pathogenesis.
 
1) Drug screening for MF (Hana Raslova)
The new p19 KO/ Jak2V617F KI mouse model will be used to establish the efficacy of several selected drugs in MF. Their combination with fedratinib or ruxolitinib both inhibiting JAK2V617F will be tested. 

2) Effect of IFN treatment in MPN patients
(Isabelle Plo, Jean-Luc Villeval, Christophe Marzac, Florence Pasquier)
Most therapies used in MPN have limited effect on the malignant cells but IFNα has demonstrated some efficacy in inducing molecular remission. Therapy improvement relies on i) better stratifying the patients. We follow the dynamic of mutated cells during IFN treatment in patients at the level of progenitors and mature cells and infer the dynamic at the level of disease initiating stem cells depending on the type of mutations, the disease, IFN doses and associated mutations using a mathematical model; ii) determining its mechanism of action. We have shown that PML, an IFNα downstream effector whose targeting by arsenic (AS) was implicated in acute promyelocytic leukemia eradication, is a therapeutic vulnerability in MPN. In a Jak2V617F mouse model, the combination strongly improved hematological and molecular responses at the level of early progenitors leading to disease initiating cell eradication. These very interesting results now open the way to clinical trials between Gustave Roussy and Saint Louis Hospital.

3) Evolution of clonal hematopoiesis upon solid cancer treatment
(Jean-Baptiste Micol)
Clonal hematopoiesis of indeterminate potential (CHIP) describes the identification, in individuals without hematologic disease, of one or more somatic mutations in hematopoietic cells. These mutations, especially in genes as TP53 and PPM1D, can represent a preleukemic state. We are particularly interested in the evolution of these clones upon treatment with chemo/radiotherapy in cancer patients and how they can contribute to leukemogenesis in therapy-related myeloid neoplasms.

4) Drug screening for AML (Iléana Antony-Debré)
We are currently testing new drugs in order to improve AML treatment, by using mouse models and primary patients cells (collaboration Stéphane de Botton, Gustave Roussy).