MESENCHYMAL STEM- AND STROMAL CELLS IN BONE MARROW AND OSTEOSARCOMA

Sammanfattning: Primary mesenchymal stem cells (MSC) play a central role in bone marrow (BM) and during haematopoiesis. Yet, the exact phenotype and spatial distribution of primary MSC in the human BM are unknown. Their cultured progeny are promising candidates for clinical applications. MSC cultures however, are heterogeneous and, while critical for clinical applications, their exact composition is not known. We have therefore analysed the phenotype of primary MSC in the BM and found that mesenchymal progenitors were highly enriched in the lin-/CD271+/CD45- fraction. Interestingly, CD146 expression pertained to the in-vivo localization of primary MSC in the human bone marrow (perivascular/endosteal) while CD146 expression in-vitro was oxygen level dependent. Cultured MSC were analysed after carboxyfluorescein succinimidyl ester (CFSE) staining for cell division tracking. Sorting for slowly dividing and rapidly dividing sub-populations and global gene expression analysis yielded 102 differentially expressed genes. Two of these genes translated into proteins that enabled for the prospective identification of a VCAM+/ FMOD+ sub-population, with low progenitor activity and a limited differentiation potential. On the other hand, MSC support tumour growth and metastasis and have even been suggested as osteosarcoma (OS) stem cells. We therefore analysed MSC in OS (OS-MSC) and compared them with BM-MSC. OS samples contained very high frequencies of mesenchymal progenitor cells. OS-derived MSC (OS-MSC) did not show chromosomal aberrations, had normal MSC morphology and expressed the typical MSC surface marker profile. A global gene expression analysis yielded a set of genes differentially expressed between OS- and BM-MSC. Of these, 3 genes responsible for membrane-associated proteins were analysed: CD142, LY6H, and OSS3' were 24.9- , 7.2- , and 66.4-fold higher expressed in OS-MSC. The OSS3-protein was expressed in all analysed primary OS samples and only OSS3 identified the majority of mesenchymal progenitor cells in uncultured tumour samples. Taken together, we propose here a marker combination for a highly enriched primary MSC population and show that CD146 expression relates to the perivascular (versus endosteal) localization of primary human MSC. This is of importance to further studies of the haematopoietic environment. We also show the presence of sub-populations within MSC cultures and propose markers for the isolation of a functionally impaired population. This is important for safe and efficient clinical application of cultured MSC. Furthermore we could demonstrate high numbers of colony forming mesenchymal progenitors in OS, suggesting that MSC are a major constituent of the OS stroma and consequently represent a target for therapy. Finally the above data show that a sub-population of OS cells expresses OSS3 and we furthermore demonstrate that OSS 3 identifies the majority of colony forming mesenchymal progenitors within OS. Presumably, these cells are the origin of a considerable part of the CAF population within the tumour and they should therefore be considered a target for therapy. Experiments investigating the use of OSS3 antibodies in an antibody-dependent cell-mediated cytotoxicity-based approach are currently underway. 'Due to a possible patent application, we were advised to substitute the original gene name with an alias.