Mapping of skeletal progenitors in embryonic limb cartilage Their application in bone tissue engineering

Large bone defects can be caused by major trauma, infection, prosthetic revision, bone tumour resection or non-healing fractures and in clinical practice, their healing remains a therapeutic challenge. Current treatments such as iliac crest autografts or cadaver allografts require multiple and repetitive interventions and are associated with various risks resulting in a high socio-economic burden. Several tissue engineering strategies have been developed to overcome these challenges and one of them is based on bone developmental engineering. This approach involves the in vitro manufacturing of a living cartilage tissue construct that upon implantation forms bone in vivo by mimicking the process of endochondral ossification as it takes place during embryonic development.

Briefly, during that process, Prrx1 expressing limb mesenchymal cells condense and differentiate into Sox9+ chondrocytes. These chondrocytes proliferate, organize in columns and enter hypertrophy under the control of an Ihh/PTHrP loop. After cell maturation into Runx2+ hypertrophic chondrocytes, a shift in matrix synthesis occurs from collagen type II to type X. This matrix calcifies and is replaced by bone by invading osteoblasts and transdifferentiating non-apoptotic hypertrophic chondrocytes, both characterized by Osterix expression and secretion of osteoid matrix.

The cell sources to engineer cartilage intermediates for in vivo bone healing can be diverse with the periosteum currently considered an excellent cell source. Lineage tracing experiments in mice have shown that during bone repair, osteoblasts and osteoclasts originate from the bone marrow, endosteum and periosteum, but that callus chondrocytes are primarily derived from the periosteum. More recently, it has been shown that human periosteal cells can be primed in vitro, by using conditioned medium and cell aggregation, to a cartilaginous intermediate tissue able to develop in vivo into bone ectopically and facilitate healing in an orthotopic long bone defect. However, these cells still generated excessive fibrous tissue. Enrichment for osteochondrogenic precursors is expected to result in an enhanced bone forming potential and improved purity of cell-based treatments.

Several studies in mice focused on the identification and contribution of skeletal stem cells and osteochondroprogenitors in bone development, homeostasis and fracture healing. These cells were found either in the zone underneath the growth plate, blood vessel niches or the periosteum. Several molecular markers have been associated with these cells such as Nestin, Gremlin1, Leptin Receptor, Gli1 and Periostin. In another study, “rainbow” adult mice displayed a high frequency of clonal regions in the growth plate, characterized by Alpha V Integrin (CD51) expression but negative for CD45 and TER119. This population was subsequently divided into eight subpopulations based on differential expression of CD105, CD90.2, CD200 and 6C3 cell surface markers. By combining this strategy with in vivo and in vitro approaches, they mapped bone, cartilage and stromal development from a postnatal mouse skeletal stem cell to its downstream progenitors in a hierarchical program similar to hematopoiesis.

In this thesis we hypothesized that enrichment for the skeletal progenitors leads to enhanced bone formation, and the possibility to reduce the cell dosage without losing potency of the cell-based construct. To test this hypothesis, we optimized the prospective isolation of stem and progenitor cell populations from the mouse embryonic hind limb cartilage 14.5 dpc.

We showed that primary mouse embryonic cartilage cells continued their developmental program and formed a bone organoid in an in vivo ectopic bone forming assay when encapsulated in collagen I hydrogel. Tracking experiments with eGFP+ embryonic cartilage cells revealed the contribution of donor cells to the osseous tissue. To purify the progenitors from the limb cartilage, we designed a polychromatic flow cytometry protocol, based on the previously described markers indicative of skeletal stem cells markers. We purified from the embryonic cartilage cells two cell populations, namely the mouse skeletal stem cell and a pre-progenitor, a direct descendent of the skeletal stem cell. Both populations were able to form bone in the collagen I hydrogel, and quantification of bone volume indicated more tissue formed in comparison to a pool of embryonic cartilage cells, even at lower cell density. In addition, the observed bone tissue was predominantly formed by endochondral ossification, as after one-week hypertrophic cartilage was observed. We noticed however that the potency of the progenitors was affected by the hydrogel encapsulating the cells, since a significant reduction in bone formation was observed when the cells were encapsulated in alginate hydrogel.

Next, we aimed to investigate the possibility to expand the primary embryonic cartilage cells, and simultaneously tried to enrich for progenitors during this expansion phase. For this, we culture expanded the embryonic cartilage cells in the presence of FGF2, a standard ligand used in stem cells expansion protocols. Gene expression analysis revealed no dedifferentiation of FGF2 expanded cells in comparison to cells expanded in foetal bovine serum containing growth medium as determined by the expression of ACTA2. Flow cytometry analysis of the CD marker set revealed an enrichment for stem cells and progenitors after two passages. However, when the stem cells and progenitors from expanded cells were implanted in collagen I, a major loss in in vivo bone formation was observed. Retrieved explants were 50% smaller, bone tissue volume was reduced, and the explants contained fibrotic tissue.

In this thesis we showed that purification of progenitors has a beneficial effect on bone formation upon in vivo implantation, and that it enabled reduction of cell dosage without negatively affecting the in vivo bone forming potency. However, prospective validation of CD markers identifying skeletal progenitors is required and may depend on the cell source. Our findings indicate that the relation between in vitro cell identity and in vivo tissue formation can be altered by in vitro manipulations.