 Human mesenchymal stem cells are important in poor tissue engineering because they are easy to isolate, can be expanded in culture and also undergo osteogenic differentiation under stimulation. The common approach using this cell is to combine the cell with three-dimensional scalpel in a perfusion reactor system to enhance their osteogenic potential. The primary role of a perfusion reactor is to mitigate the mass transfer indication and also provide mechanical stimuli to drive the cell down osteogenic lineage. There are two different flow configurations. In the first flow configuration in the perfusion reactor, the flow primarily occurs at the interface outside the scaffold and the shear stress only occurs at the interface between the scaffold and the bulk media. Alternatively, you can drive the media transversely across the scaffold. In that case, all the cells in the interior of the scaffold will be exposed to media flow and shear stress. An important property of human mesenchymal stem cell is their ability to secrete ECM and growth factors. How these two different flow configurations impact the accumulation, secretion and accumulation of ECM and growth factors in three-dimensional scaffold is largely unknown, and that's the motivation for this study. We used our in-house perfusion reactor system as a very simple design, had three scaffold tucked in the center of the chamber separating the chamber to upper and lower channels. Each channel has one in-light and one out-light, so when we pump the media through two in-lights and coming out to two out-lights, the media will perfuse on the surface of the scaffold, parallel to the surface of the scaffold. We call it parallel flow. Alternatively, we can close one in-light and one out-light, and in this case, the media will be driven transversely across the scaffold and generating so-called transverse flow in the system. In the current system, we put four identical chambers in a single system and used the dynamic seating with the same passage and same donor, so initial condition in each chamber was identical. After the seating, we run two chambers with parallel flow and rest of chamber with the transverse flow over 14 days. After seven days of culture, we took out one chamber from each flow configuration and rest of them at day 14, then tested the functional assay. The results are very interesting. The ECM accumulation in the parallel flow is considerably higher than that in the transverse flow, and what is the most striking is the accumulation of FGF2, an important growth factor for MIC in two flow configurations. By day seven, FGF2 concentration in parallel flow is over four times higher than what is in the transverse flow. This leads to the difference in cell proliferation and the CFU f-forming ability. Cells in the parallel flow maintain much better proliferation potential and higher CFU f-forming ability. In contrast, cells in the transverse flow are much further down osteogenic lineage as analyzed by PCR comparing to the cells in parallel flow. So together from these results, we think that one important function of media flow in the reactor system is to change the cellular microenvironment, the formation of such microenvironment, and through it direct the cell fate. It should be noted that shear stress we use in the two flow configuration are considerably lower than what is considered osteoinductive. So the primary reason for the observed difference in cell phenotype change in two flow conditions are mainly from the differences in their microenvironment. So in future studies, the changes of microenvironment should be carefully considered in the perfusion reactor system because they have a significant impact on cell fate in tissue formation process. And our venue script has detailed description of the experiment. And thank you and happy reading.