Platelet-Derived Factors Allow Human Mesenchymal Stem Cells to Spontaneously Undergo Endochondral Bone Differentiation and Provide Bone Marrow Support in a Xenogenic In Vivo Model

Mesenchymal Stem/Progenitor Cells (MSPC) are regarded as the universal skeletal progenitor, theoretically capable of differentiating into cartilage, bone, tendon and muscle. Their functions as pericytes as well as key bone marrow stromal cells are also well documented. Tremendous benefits could therefore be achieved by cell therapy with MSPC. Unfortunately, the very successful research aimed at isolating, expanding and differentiating these cells in vitro has so far failed to translate into significant clinical advances. In particular, bone regeneration studies are disappointing, because application of MSPC in vivo requires their osteogenic pre-differentiation in vitro and/or their co-implantation with bone chips, and because results are highly donor-dependant. In this context, recent studies indicate that MSPC expanded from human Bone Marrow (BM) in media supplemented by Human Platelet Lysate (HPL) in lieu of Fetal Bovine Serum (FBS) have a higher rate of bone differentiation. Here, we use a new, simplified mouse model of ectopic bone formation requiring neither pre-differentiation nor bone matrix. We show that human BM-MSPC expanded in HPL-supplemented medium from all donors spontaneously form bone through an endochondral mechanism. Importantly, the ossicles generated from MSPC from almost half of the donors become the site of ectopic bone marrow development. Further experiments suggest that maintenance of MSPC stemness by platelet-derived factors during cell expansion is paramount to this effect.

MSPC are expanded from human bone marrow following standard protocols, with culture media supplemented either with 10% FBS or with 10% HPL. Cultured MSPC are resuspended in a non-mineral collagen/laminin matrix (Matrigel^®) and injected subcutaneously into immune-deficient NSG mice. Bone formation is monitored non-invasively by osteosensitive near-infrared imaging and/or by histology on paraffin-embedded ossicles. HPL-derived MSPC from 13 out of 13 donors form bone in vivo compared to only 2 out of 10 FBS-derived MSPC. Extensive cartilage formation is observed as early as one week after implantation, while signs of ossification appear from the third week onwards. In addition, ossicles generated by HPL-MSPC from 6 out of 13 donors become colonised by mouse bone marrow, indicating that platelet-derived factors maintain the capacity of MSPC to reconstitute a functional bone marrow niche. Bone formation in vivo is considerably delayed, but not blocked, by pre-treatment of the cells with Cholera Toxin, indicating that a Gα protein-coupled receptor (GPCR-α) ligand is partly responsible for the biological effect of HPL. In vitro, HPL and FBS-derived cells display remarkably little phenotypic differences, with the notable exception of the stemness-associated surface marker SSEA-4 which is consistently more expressed on HPL- than on FBS-derived MSPC. Interestingly, serial passage in FBS medium of cells first expanded in HPL medium leads to the rapid loss of surface SSEA-4 expression, paralleled by a loss of osteogenic and bone marrow support potential in vivo. Loss of SSEA-4 expression is also observed in cells serially passaged in HPL medium containing the PDGF receptor-β inhibitor Imatinib, and ossicles derived from these cells are no longer able to attract mouse bone marrow in vivo. These data suggest that human platelet-derived factors allow MSPC to retain their stem cell potential in culture as well as in the ectopic bone microenvironment. This hypothesis is further supported by the observation that human MSPC can be re-isolated and re-expanded from bone marrow-infiltrated ossicles and, remarkably, build bone again when re-injected into NSG mice.

Our data suggest that platelet-derived factors, including PDGF and an (as yet unidentified) GPCR-α ligand, contribute in vitro to maintain BM-MSPC stemness. In vivo, they efficiently drive BM-MSPC to differentiate along the chondrogenic and osteogenic lineages, while preserving MSPC bone marrow-support function. We conclude that therapeutic approaches using MSPC for skeletal regeneration should preferentially use early passage BM-MSPC expanded in HPL-supplemented medium. Furthermore, the observed correlation between surface SSEA-4 expression and multipotency in BM-MSPC can be exploited to monitor the quality of the cell preparations.

No relevant conflicts of interest to declare.