Micro- and Nano-Topographies to Enhance Non-viral Transfection and Non-viral Neuronal Transdifferentiation of hMSCs

Abstract

Transfection and direct conversion, which is referred to as “transdifferentiation” in this thesis, play significant roles in regenerative medicine research including gene therapy, vaccines, modeling of diseases and disorders, drug testing, cell replacement therapy, and tissue engineering. Most of this research has depended on the use of viral vectors, which can be associated with adverse conditions such as genotoxic integration of viral payloads and immunogenicity. Non-viral vectors have the potential to eliminate these problems and have added benefits such as ease of production and targetability, however, they come at the cost of efficiency. Substrates topographies are known to be able to modulate cell behaviors including efficiency of non-viral transfection and non-viral transdifferentiation. However, current research is limited to only a few combinations of cell type, substrate material, and topographical geometry. In this thesis, we specifically investigated the hypotheses that nano- or micro-topographical PDMS substrates would be able to increase non-viral transfection efficiency and/or enhance non-viral neuronal transdifferentiation of human mesenchymal stem cells (hMSCs). We screened the effects of 16 different nano- and micro-scale topographies, with different geometries, on both processes. Results from our transfection study showed that five of the topographical patterns, which included nano- and micro-gratings, concave micro-lenses, micro-holes, and nano-pillars, increased the efficiency of Lipofectamine-mediated transfection. However, these results were not statistically significant, so further rigorous study of these selected topographies is needed. Convex and concave micro-lenses interestingly showed opposite effects on transfection efficiency, and are suggested for further study to investigate the significance of the relationship between direction of lens curvature and transfection efficiency. Our studies on transdifferentiation were neither able to confirm or disprove our hypothesis. No neuronal morphology was seen in samples that went through the transdifferentiation procedure. Immunofluorescent staining for the neuronal lineage marker, microtubule-associated protein 2 (MAP2), showed weak fluorescence in all samples including negative controls. This could have been due to non-specific staining or weak MAP2 expression previously shown to exist in hMSCs without any induction. We found that the hMSCs became over 95% confluent during the transdifferentiation procedure which may have interfered with the transfections or topography-cell interactions during the transdifferentiation procedure. We also found that the poly(amido amine) transfection reagent used during the transdifferentiation procedure was more toxic to hMSCs than, as reported elsewhere, to mouse embryonic fibroblasts or COS-7 cells. Together these observations indicate that parameters affecting cell confluence (such as seeding density or length of transfection phase) and the specific non-viral transfection reagent used should be reconsidered for future work on neuronal transdifferentiation of hMSCs. Alternatively, since the neuronal transdifferentiation procedure has previously been shown to work on fibroblasts, investigations of topographical influence on neuronal transdifferentiation with fibroblasts may lead to more informative results.