Queralt Vallmajó Martín
The bone marrow (BM) microenvironment, or niche, significantly affects behaviors of its resident stem cell populations. Disruptions in the BM niche contribute to a number of severe clinical pathologies. Discovery of novel therapeutic strategies for BM-related diseases necessitates development of superior preclinical models to study parameters affecting BM homeostasis. To address this, a poly(ethylene glycol) hydrogel featuring a particular crosslinking chemistry based on the enzyme transglutaminase factor XIII, dubbed TG-PEG, was utilized to engineer humanized bone marrow models. Its modular nature was exploited throughout this work to systematically evaluate contributions of biophysical, cellular, and biochemical BM niche components to stem cell biology.
Stiffness of TG-PEG hydrogels was optimized for each cell type and individual application explored. For bone formation by bone marrow stromal cells (BMSCs), an intermediate stiffness was optimal and significantly improved total bone volume over softer TG-PEG gels or natural materials. Results indicated that stability of the scaffold was more important than their chemical and biological properties. Alternatively, for hematopoietic stem and progenitor cells (HSPCs), softer gels were optimal for preserving multipotency while promoting cell proliferation.
A number of BM-specific stem and progenitor cell types with richly diverse functions were explored. Skeletal stem cells (SSCs) encapsulated in TG-PEG hydrogels demonstrated their ability to form de novo bone in healing critically sized calvaria defects in mice. Next, subcutaneous implantation of BMSC-laden gels revealed their significant contribution to formation of ossicles. For some BMSC donors, ossicles could be formed in absence of bone morphogenetic protein (BMP)-2 offering a novel tool to evaluate their intrinsic capacity. Later, titration of BMSC seeding densities and BMP-2 doses yielded optimal conditions for robust BM formation, regardless of donor, enabling reproducibility for follow-on studies involving BM models. In humanized mice, homing of HSPCs to BMSC-laden TG-PEG gels was observed. Finally, humanized xenograft models were employed to determine vulnerability of the BM niche to subversion by cancer cells simulating osteotropism for both leukemia and breast cancer metastasis.
Molecular factors known to trigger particular cell reactions within the niche were also investigated. Including RGD cell-adhesion ligands, as well as matrix metalloproteinase (MMP)-susceptible crosslinks to render the gels degradable, were critical for facilitating migration of recruited BM cells from murine hosts in xenograft studies. Hyaluronic acid (HA), when added to the hydrogel backbone, conferred superior engraftment to transplanted BMSCs and afforded reduced immunogenicity. Incorporation of the Notch-signaling ligands Jagged1 or DLL-4 into soft TG-PEG gels during in vitro HSPC expansion substantially stimulated proliferation, but recovered cells could no longer ensure long-term reconstitution.
Aging, trauma, or cancer can cause disruptions of the tightly controlled balance between cells and their microenvironment in the BM niche. This body of work presents a number of bioengineered tools optimized for studying how niche components are affected by or contribute to such pathologies. Further work with these or similar models will no doubt highlight key clinically relevant targets and inform novel therapeutic approaches.
EPFL2019
