Genetically modified bone marrow derived mesenchymal stem cells (MSCs) have proven to be efficient cell carriers for local or systemic delivery of therapeutics for the treatment of diseases as well as for growth factors to augment tissue formation. Moreover, the effective transfection of MSCs inside tissue engineering constructs would transcend as an ideal approach to guide tissue formation in vitro and in vivo. However, the inability to efficiently transfect MSCs has limited these potential therapeutic applications. Most of the studies involving the enhancement of gene transfer have typically centered on improving the delivery vector systems, while the cellular microenvironment in which the cell resides and the cell itself are typically neglected. In the present study, the effect of structural extracellular matrix proteins, fibronectin (FN), laminin (LM), vitronectin (VT), collagen I (CL-I), and matrigel (MG), on non-viral gene transfer to cloned mouse MSCs (D1) is studied. In our gene transfer protocol cells are seeded on the immobilized structural proteins and DNA/PEI polyplexes are added post cell binding and spreading. Gene transfer was found to be significantly affected by the protein to which the cells were bound, with cells plated in MG having a statistically higher transgene expression than any other protein. In contrast, cells plated on collagen I resulted in statistically lower gene transfer efficiency than any other protein coating.

To determine the reason for MG enhancing and CL-I reducing the level of gene transfer we analyzed the morphology and proliferation for cells seeded on the ECM proteins. Although D1 cells were able to attach on all ECM proteins, cell spreading and the resultant morphology was different. For example cells plated on FN and LM resulted in well spread star-shaped cells with visually more actin stress fibers than cells seeded on CL-I, VT, and MG, while cells seeded on MG resulted in very elongated spindle cell morphology. Cells seeded on BSA coated wells (control) did not spread significantly. Surprisingly, although there were differences in cell morphology and spreading, cell proliferation was similar for all proteins except MG. Based on the above results it appears that gene transfer to D1 cells is affected by the cellular microenvironment were the cells reside, with ECM proteins that induce actin stress fibers and elongated spindle like cells resulting in more efficient gene transfer than ECMs that result in moderately spread cells. Proliferation rate was not able to explain the differences in gene transfer efficiency, with cells having the same proliferation rate resulting in drastically different gene transfer efficiency (e.g. FN vs CL-1) and cells with slower proliferation rate resulting in similar gene transfer rate as cells that had a faster proliferation rate (e.g. MG vs FN). We believe that the cellular microenvironment can be manipulated and engineered to achieve optimal non-viral gene transfer. The investigation of the effects of the cellular microenvironment on non-viral gene transfer will result into a comprehensive understanding of the limitations of gene transfer, which could further improve the current methods and vectors used for non-viral gene delivery.