Figure 17

Time evolution of the hMSC differentiation process. We projected the data over the first component vector (with explained variance of 85.64%) of the second mode (populations and time) obtained for Tucker3 analysis. At one end we have undifferentiated hMSC (Tissue Culture Plastic, or TCP) and at the other end we have the target state (fully differentiated hOST). In between we plot the data for each stimulus and time point (e.g., NSD2 = no stretch, day 2; SD5 = stretch, day 5). Because the "SD" points lie closer to the target (hOST) than their corresponding "NSD" conditions on days 2 and 5, we conclude that the stimulus "stretch" accelerates osteogenic differentiation when compared to the same stimulus without stretch.