Mesenchymal Stromal Cells (MSC) Isolated from Human Osteosarcomas Show a High Progenitor Cell Frequency, Typical MSC Morphology, Surface Marker Profile, and Differentiation Capacity, and They Are Considerably Affected by Tyrosine Kinase Inhibitors in Vitro.

Bone-marrow derived mesenchymal stromal cells (MSC, also referred to as mesenchymal stem cells) are multipotent cells with intriguing properties (proliferation and differentiation capacity, stroma function, immune modulation) making them promising candidates for clinical use. Recently, it has been reported that culture-derived MSC can acquire chromosomal abnormalities, and abnormal culture-derived murine MSC have been found to give rise to osteosarcomas in transplantation experiments. Furthermore, MSC have been implicated in osteogenic sarcoma, and we therefore aimed to isolate and characterize MSC from primary human osteosarcomas (OS). Single cell suspensions were prepared from OS specimens and cells could be clearly identified morphologically as tumor cells. Stromal progenitor content was assessed using the CFU-F assay. CFU-F frequencies in the OS specimens were 980 ± 450 colonies per 1 × 10⁵ cells (n=6), which is considerably higher than in normal bone marrow (1.3 ± 0.2 colonies per 1 × 10⁵ cells, n=8, bone marrow from healthy volunteer donors). Culture-derived MSC could be generated from every OS sample tested using standard tissue culture flasks and serum-containing MSC medium. OS-derived MSC were similar to bone marrow (BM) derived MSC showing typical MSC morphology and typical MSC surface marker profile, i.e. cells were positive for CD105, CD73, CD90, CD44, HLA-class I, CD166, and negative for CD45, CD34, CD14, CD19, HLA-DR, and CD31. Furthermore, three of three tested OS-derived MSC samples could be differentiated into the osteoblastic and chondroblastic lineages, and all but one showed adipocytic differentiation capacity. Karyotyping of OS-derived MSC showed that the majority of MSC samples were normal, as were the results of FISH analyses for chromosomes 7, 10, 13, and 17. MSC derived from one OS specimen contained cells that showed a complex abnormal karyotype, which, however, was different from the karyotype of the primary tumor. BM- and OS-derived MSC cultures exposed to the first and second generation tyrosine-kinase inhibitors (TKI) imatinib (IM) and nilotinib (NI) (0 – 10 μM) showed a significant inhibition of MSC growth at every time point studied (0 – 4 weeks). After 4 weeks, BM-derived MSC were reduced by IM to 31.4 ± 12.0% (0.625 μM), 24.6 ± 9.1% (1.25 μM), 23.8 ± 12.3% (2.5 μM), 10.6 ± 0.9% (5.0 μM), and 0.9 ± 0.1% (10.0 μM) compared to controls. The lowest numbers of OS-MSC were observed after 2 weeks exposure to IM (0.625 μM, 6.9 ± 4.6%; 1.25 μM, 8.5 ± 4.5%;2.5 μM, 10.3 ± 6.4%; 5.0 μM, 9.2 ± 4.9%; 10 μM, 0.6 ± 0.1%). BM-MSC and OS-MSC growth was also considerably inhibited by NI, but the effects were less pronounced for most doses and time points studied; minimum numbers of BM-MSC (0.8 ± 0.2% of control) and OS-MSC (0.6 ± 0.1% of control) were observed after 3 weeks culture with 10 μM NI. Taken together, our results show that osteosarcoma samples contain high numbers of mesenchymal progenitor cells and that MSC can be successfully generated from OS. OS-MSC are likely to represent stromal cells, i.e. cancer-associated fibroblasts, rather than the tumor cell population. Interestingly, strong similarities were observed between OSMSC and BM-MSC, implicating that these two cell types are closely related.