Modelling ineffective erythropoiesis in myelodysplastic syndromes with ring sideroblasts

Abstract

Myelodysplastic syndromes with ring sideroblast (MDS-RS) is a clonal hematological malignancy characterized by accumulation of iron filled erythroblasts called ring sideroblasts in the bone marrow, and recurrent somatic mutations in the splicing factor gene SF3B1. The disease mainly affects elderly individuals, causing severe anemia in patients for which no curative treatment exists. An important requisite to develop new treatments for MDS-RS is to identify the cell population that propagates and sustains the disease clone, as well as its functional downstream effects on erythropoiesis. Modelling the ineffective erythropoiesis of MDS-RS has been problematic both in vitro and in vivo, in particular the generation of ring sideroblasts, which has hampered functional studies of the disease. The focus of this thesis was to determine at which stage of the hematopoietic hierarchy SF3B1 mutations originate in MDS-RS patients, and to provide experimental models recapitulating the disease phenotype, allowing for functional studies and testing of new therapeutic options.

In study I we demonstrated that SF3B1 mutations in MDS-RS patients originate in the hematopoietic stem cell (HSC) compartment, before division into the myeloid and lymphoid lineages. We found clonal involvement in B-cell progenitors resulting in a negative effect on lymphoid development. Furthermore, we found that only HSCs and no other investigated progenitor populations isolated from MDS-RS patients could propagate the SF3B1 mutated clone both in vitro and in vivo. Transplantation of HSCs from MDS-RS patients into immunodeficient mice resulted in ring sideroblast formation, providing a novel in vivo model to study the disease. In study II we established a three-dimensional (3D) culture model capable of recapitulating healthy and aberrant terminal erythropoiesis. Suspension cultures of CD34+ progenitor cells from MDS-RS patients had thus far failed to generate mature erythroid cells, including ring sideroblasts. We therefore decided to compare long term cultures of CD34+ cells and mononuclear cells (MNCs) from healthy individuals and MDS-RS patients either in suspension (2D) or in 3D scaffolds that mimic the structure of the bone marrow. We found that the scaffolds provided the CD34+ cells with proliferative advantage and enabled them to preserve their selfrenewal potential. By comparison, the same cells did not survive beyond three weeks in 2D cultures. Additionally, the CD34+ 3D cultures predominantly facilitated erythropoiesis, including enucleation and erythroid island generation. MNC cultures maintained stable proliferation for the four-week culture period, supporting multi-lineage hematopoietic differentiation and cytokine secretion relevant to erythropoiesis and MDS-RS. The CD34+ 3D, MNC 3D and MNC 2D cultures maintained the SF3B1 mutated clone and generated ring sideroblasts de novo from the second week of culture, providing a novel in vitro model to assess therapeutic compounds aiming to alleviate the anemia in MDS-RS patients.

In study III we treated primary cells from healthy individuals and MDS-RS patients with luspatercept, a relatively new treatment option for MDS-RS, in the 3D model established in study II. Luspatercept is a transforming growth factor beta family ligand trap that has been shown to alleviate anemia in MDS-RS patients although its mechanism has not been elucidated. We found that luspatercept enhances proliferation and erythroid output of CD34+ cells and MNCs from healthy individuals in vitro, demonstrating that it can have a direct effect on hematopoietic progenitor cells. By contrast, luspatercept had no direct effect on hemopoiesis in the MDS-RS cultures, nor did it inhibit the SF3B1 mutated clone or ring sideroblast generation. This indicates that the drug may not directly target the disease clone although this will have to be confirmed in a larger population of responding patients. Interestingly, we found that luspatercept completely inhibited IL-6 secretion of CD34+ cells from healthy individuals in vitro, indicating that the drug can affect cytokine secretion. Since IL-6 can have a negative effect on erythropoiesis and is upregulated in a proportion of MDS patients it is worth exploring if it is upregulated in MDS-RS patients that respond to the drug.