Myelofibrosis (MF) is the most severe of the Philadelphia-negative myeloproliferative neoplasias (MPNs), with a median survival of 5-6 years. Whether diagnosed de novo (Primary Myelofibrosis, PMF) or secondary to another MPN, the features of MF are similar. A subpopulation of haematopoietic cells derived from the pathological clone releases pro-inflammatory cytokines and growth factors into the bone marrow microenvironment. In response, the bone marrow microenvironment undergoes remodelling, resulting in osteosclerosis and fibrosis of the mesenchymal stromal cells (MSCs) associated with loss of haematopoietic support. The 2016 WHO classification includes a premyelofibrosis state to facilitate early diagnosis of patients at increased risk of progression. However, although major progress has been made in understanding the pathogenesis of the disease, notably with the description of the so-called 'driver' mutations responsible for myeloproliferation (JAK2, CALR and MPL), apart from haematopoietic stem cell allotransplantation, which only concerns a minority of patients, current treatments are mainly symptomatic and have little influence on the natural history of MF.
Recently, we demonstrated that activation of the nuclear receptor PPARy (Peroxisome Proliferator-Activated Receptor-gamma) by its pharmacological ligands (Actos®) or (Pentaza®) reduced the development of osteosclerosis and reticulin fibrosis of the bone marrow (BM) and prevented anaemia resulting from bone marrow remodelling in three preclinical mouse models of MF (Lambert, Saliba et al. 2021). These results position PPARy agonists as interesting therapeutic candidates. However, before considering their therapeutic repositioning in the treatment of MF, it is imperative to characterise the status and function of PPARy within medullary MSCs both at the physiological stage and during the development of NMPs.
In this project, our initial results show that PPARy expression is decreased in murine and human MSCs at the MF stage. In contrast, no change in PPARy expression was observed in MSCs derived from other MPNs. Transcriptomic analyses also demonstrated that TGF-B, a major cytokine in the development of MF, is capable of negatively regulating PPARy expression in MSCs. In order to mimic this expression defect, we invalidated PPAR-y (KO) in two bone marrow MSC lines, the first murine (MS5), the second human (HS5, under characterisation). Under these conditions, basal expression of a panel of genes associated with MF is increased in MSC-KO to the level of wild-type lines stimulated by TGF-B. Expression of this panel was further increased in MSC-KO in the presence of TGF-B, indicating potentiation of the TGF-B-mediated signal in the absence of PPARy. This transcriptomic signature associated with KO-MSCs is found in murine MSCs from the thrombopoietin (TPOhigh) induced MF model as well as in human MSCs from patients with PMF. However, this expression profile was not found in MSCs from patients with another MPN, indicating that it is indeed a sign of a stage of MF.
Invalidation of PPARy does not affect the phenotypic signature of bone marrow MSCs, but their multipotent character is altered with a loss of adipocyte differentiation capacity associated with an increase in osteo-chondrocyte differentiation potential. These histological observations are corroborated by the decrease in the production of adipocyte factors by MSC-KO and an increase in the expression of the osteoblastic factor Runx-2. In addition, the supernatant of the KO line showed a marked increase in osteoprotegerin (OPG), a soluble molecule produced by osteoblasts that leads to apoptosis of osteoclasts. This deregulation of the osteoblast/osteoclast balance in KO conditions could explain the osteosclerosis observed in patients with MF. In addition, the production of CXCL12 (CXC motif Chemokine Ligand 12) and the bone marrow growth factor SCF (c-kit ligand) are greatly reduced in MSC-KO conditions, at both transcriptomic and protein levels. These data recapitulate the results described during transcriptomic analyses of MSC from patients with fibrosis. At the same time, the capacity of MSC-KO to support haematopoiesis, in both the short and long term, is significantly reduced, reflecting the cytopenias associated with MF.
In silico, RNA-Seq analyses were carried out on the MS5-WT and MS5-KO lines. Initial gene set enrichment analyses (GSEA) show that the pathways most significantly affected are inflammation, myogenesis (MSC to myofibroblast transition) and the cell cycle. Comprehensive analyses are now required to identify new therapeutic candidate genes and gain a better understanding of the development of bone marrow fibrosis.
These initial in vitro results support the key role of the PPARy receptor in the homeostasis of the bone marrow microenvironment and in the genesis of its remodelling during the development of myelofibrosis. However, in vitro approaches alone are unable to capture the full complexity of a disease involving multiple players including haematopoietic cells, immunological cells and all the cell types making up the bone marrow microenvironment. To integrate all these parameters, we have established a mouse model in which PPARy expression is reduced (haploinsufficiency) or invalidated (KO) in the medullary MSCs of animals. It is the study of this model that will form the core of the project. Initially, in vivo, it will be used to:
1) Characterise the role of PPARy in the homeostasis of the bone marrow microenvironment.
2) Assess the impact of reduced expression on the development of bone marrow fibrosis.
3) To validate the positioning of PPARy as a therapeutic target in the management of bone marrow fibrosis and to consider the repositioning of its pharmacological agonists (Actos®; Pentaza®) in this pathology.

The presence of medullary pre-fibrosis/fibrosis is a poor prognostic factor in MPN or acute myeloid leukaemia (AML). However, it is difficult to determine whether this condition is simply an indicator or whether it plays an active role in the development of haemopathies. The use of these models (Haplo-insufficient or KO for PPARy in MSCs) in association with preclinical mouse models of MPN (CML (BCR-ABL); PV (JAK2 V617F), ET (CALRDel52)) will allow, in a second phase, to determine whether:
1) The presence of a predisposition to bone marrow fibrosis influences the natural history of haemopathies.
2) In these diseases, which are purely haematopoietic in origin (mutation of the haematopoietic stem cell), it is appropriate to combine treatment targeting the malignant clone with treatment aimed at preventing the development of bone marrow fibrosis (activation of the PPARy receptor by its ligands in the haploinsufficiency condition).

This entire project is part of the Tomorrow's Biotechnologies (F) initiative, which aims to improve patient care through the development of personalised medicine.