Standardization strategies for characterizing and manipulating the human bone marrow microenvironment

The bone marrow (BM) niche is a complex cellular, molecular, and physical microenvironment capable of homing and supporting hematopoiesis. It is characterized by its ability both maintain and to drive hematopoietic stem and progenitor cells (HSPCs) self-renewal and differentiation which ultimately generate all blood cell types. However, how the niche elements interact in a human context remains largely elusive due to the difficulty of engineering and exploiting relevant human-specific models. To address these limitations, this thesis explores the generation of new cellular tools and systems for the in vivo and in vitro bioengineering of human BM niches. The work is largely centered on the exploitation of human mesenchymal stromal cells (MSCs), the main regulatory elements of the BM niche.
This thesis opens on the possibility to harness MSCs’ ability to form humanized ossicles (hOss), a human BM organ hosting hematopoiesis in mice. While hOss provides an advanced in vivo model for human niche investigation, we report here that the lack of (1) a standardized protocol, (2) a stable MSCs source, and (3) functional characterization limits its exploitation. We thus aim at addressing these shortcomings.
We subsequently report the reproducible generation of hOss using the human MSCs line MSOD-B (Mesenchymal Sword of Damocles Bone morphogenetic protein type-2). Using this standardized tool, we demonstrate the robust hOss formation, offering superior engraftment of human healthy and leukemic blood cells in hOss compared to mouse bones. This was correlated to the MSOD-B capacity to reconstitute the mesenchymal elements of the human BM niche. These results prompted us to explore if our MSCs line could form hematopoietic niche in vitro.
The in vitro engineering of human BM niches is performed by combining scaffolding material and 3D perfusion bioreactor systems. Using such set-up, we describe the generation of osteoblastic niches through the functionalization of a collagen scaffold by directed human MSCs differentiation. HSPCs are then infused, leading to niche interactions reconstitution. Standardization is then achieved by combining MSOD and our open-source 3D printed perfusion bioreactor. We established the rapid design of human hematopoietic niche of custom sizes and further biologically validated our system for up to two weeks. We observed MSCs- HSPCs interactions throughout the niche leading to the phenotypic expansion of the blood stem cell populations.
The last part of the thesis focuses on the study of mitochondrial transfer between MSCs and HSPCs, a key aspect of niche communication. This recently proposed niche interaction route remains cryptic, and the exploitation of bioengineering systems would largely help uncovering this mechanism. To this end, a MSCs line bearing inducible mCherry mitochondrial tag was first generated, the iMSOD-mito. By exploiting iMSOD-mito in 2D and 3D in vitro culture systems, we evidenced significant mitochondrial transfer from mesenchymal to leukemic and healthy HSPCs. Most importantly a preferential transfer towards phenotypic CD34+/CD38- /CD45RA-/CD90+/EPCR+ stem cells was identified. We further associated this transfer with retained quiescence in single-cell divisional assays.
In summary, this thesis presents the development and exploitation of advanced standardized models of the human BM niche. We envision that this work will facilitate the understanding of the mesenchymal regulation of hematopoiesis in both healthy and malignant contexts.