Development and Exploitation of a Fully Human and Modular Organotypic Bone Marrow Niche Model to Study the Role of Stroma-Produced Factors in Human MDS

Background: Myelodysplastic syndromes (MDS) are a heterogenous group of stem cell driven disorders primarily affecting the elderly and characterized by inefficient production of mature blood cells and a high risk (30%) of evolution to secondary acute myeloid leukemia. Despite tremendous progress in the past decade, treatment options for MDS patients remain limited, and primarily address disease symptoms, rather than altering disease course. This points to the urgent need to better understand the pathogenesis of this heterogenous group of syndromes to develop new therapies that address disease vulnerabilities. However, this effort has been largely hampered by the limited availability of model systems that allow the exploration of MDS biology in a fully humanized setting. In recent years, studies from our lab and others, have highlighted the crucial role niche cells play in human MDS, hence reinforcing the notion that MDS is a disease of a tissue rather than hematopoietic cells alone. Therefore, exploration of MDS biology requires the further development of fully human MDS models in which both constituents of the disease, namely hematopoietic and niche cells, are present.

Methods: To address this issue we successfully isolated endothelial cells (ECs) and mesenchymal stromal cells (MSC) from bone marrow biopsies obtained from MDS patients or healthy age matched controls, and subsequently utilized them to develop fully human 2D and 3D organotypic niche models, which were successfully used to support normal and MDS HSPCs expansion ex-vivo. The 3D system makes use of a collagen scaffold, as this protein makes up for 90% of the matrix proteins in the bone. Importantly, MSC and EC cultures could be successfully established from several independent donors and immortalized to generate primary cell lines that can be used to reproducibly establish these ex-vivo systems in a robust manner. Moreover, we could show that these niche cells were easily amenable to genetic editing using CRISPR-Cas9 technology as well as modified to carry fluorescent reporter proteins for tracking cellular interactions using live cell imaging and confocal microscopy.

Results: In this work, we successfully isolated human mesenchymal and endothelial cells, from primary bone marrow biopsies (MDS and healthy) and established fully human 2D and 3D organotypic co-cultures ex-vivo. Of note, although bone marrow ECs represent an essential component of the hematopoietic niche, they have so far been omitted in previously described human bone marrow niche models, owing to the notorious difficulties in isolating and expanding this cell type from primary bone marrow biopsies. Therefore, we established immortalized EC lines (iECs) that faithfully recapitulate the morphological, phenotypic and functional features of primary bone marrow ECs. When cultured at defined ratios and under defined conditions, MSCs instructed ECs and iECs to form of vessel-like structures that mimic the meshwork observed in vivo and are typically escheated by aSMA positive cells that stabilize the structures. Genetic manipulation of the cellular components of the niche also allowed to explore the functional relevance of a specific ECM protein, which we previously identified to be significantly upregulated in MSCs isolated from MDS patients, namely the Secreted Protein Acidic and Rich in Cysteine (SPARC). SPARC ablation triggered enhanced proliferation of MDS derived HSPCs and sensitized them treatment with 5-Azacytidine, a standard of care hypomethylating agent used for the treatment of MDS patients. Additional studies are underway to further understand the underlying molecular mechanisms and define a potential druggable target that could sensitize MDS cells to standard of care treatment. Besides gene targeting studies, these organotypic models are also being used to evaluate the relative fitness of MDS and healthy stem/progenitor cells in healthy versus patient derived niches, to explore the contribution of niche components to the establishment of the progressive clonal dominance observed in MDS.