RUNX2 and SOX9 protein mobility correlates to osteogenic and chondrogenic differentiation of mesenchymal stem cells

Purpose: RUNX2 and SOX9 are the (master) transcription factors for bone and cartilage development respectively. In a subset of osteoarthritis (OA) patients, cartilage abnormalities are characterized by hypertrophic differentiation of primary articular chondrocytes and subsequent endochondral ossification. However, the switch between chondrocyte and hypertrophic differentiation is not yet fully understood. Mapping the molecular mechanism involved in this process will help us to understand disease mechanism and identify therapeutic targets to prevent hypertrophic differentiation. We have previously shown that Fluorescence Recovery After Photobleaching (FRAP) can be used to study the activity of SOX9 in chondrocytes. Here, we apply FRAP to study the dynamics of these factors from the progenitor level i.e., mesenchymal stem cells (hMSCs) by differentiating them to chondrogenic, osteogenic and adipogenic lineages, to understand their regulation in physiological conditions

Methods: hMSCs were isolated from bone marrow from patients with no known musculoskeletal diseases. hMSCs were grown on poly-l-lysine or fibronectin coated coverslips and differentiated to chondrogenic, adipogenic and osteogenic lineages. Steady state FRAP measurements were done on differentiating cells, which transiently expressed SOX9-mGFP or eGFP-RUNX2, at 2d, 8d, and 23d time points or on proliferating cells. FRAP was performed in at least 40 cells per condition. Results were analyzed using Matlab™ and statistics were performed using Mann-Whitney U tests. To confirm the differentiation, qPCR and histological staining for different lineages were performed.

Results: When compared to proliferating hMSC, RUNX2 transcriptional activity was significantly increased in osteogenic differentiating hMSCs, whereas the activity was decreased in chondrogenic differentiation. SOX9 transcriptional activity was slightly increased in chondrogenic differentiation, whereas it was decreased in osteogenic differentiation as compared to proliferating hMSCs. Both RUNX2 and SOX9 transcriptional activity were decreased in adipogenic differentiation. An increased fast diffusion population (A1) and an increased effective diffusion rate of RUNX2 and SOX9 was observed for all the three differentiation lineages and was correlated to a decreased transcriptional activity as observed by qPCR analysis of target genes and lineages specific protein production. For all the three lineages, and at all the time points tested, we observed at least two populations with respect to the protein mobility. Based on the similarity of the mobility to the mobility of RUNX2 in undifferentiated cells, we observe a differentiated and an undifferentiated cell population. This is evidenced by increased or decreased immobile fraction (IF) of eGFP-RUNX2 respectively.

Conclusions: We have identified that the mobility, and the correlated activity, of the transcription factors SOX9 and RUNX2 is related to the differentiation of cells toward either the chondrogenic, osteogenic or adipogenic lineage. In addition, we observe different cell populations with different transcription factor mobility. Differentiated and undifferentiated cells could be segregated based on the mobility of the transcription factors. The population differentiated cells increased in time. To our knowledge, this is the first report of differential transcriptional dynamics of RUNX2 and SOX9 during chondrogenic, osteogenic and adipogenic differentiation of hMSCs.