Innovative strategy for microRNA delivery in human mesenchymal stem cells via magnetic nanoparticles

Abstract

Bone marrow derived human mesenchymal stem cells (hMSCs) show promising potential in regeneration of defective tissue. Recently, gene silencing strategies using microRNAs (miR) emerged with the aim to expand the therapeutic potential of hMSCs. However, researchers are still searching for effective miR delivery methods for clinical applications. Therefore, we aimed to develop a technique to efficiently deliver miR into hMSCs with the help of a magnetic non-viral vector based on cationic polymer polyethylenimine (PEI) bound to iron oxide magnetic nanoparticles (MNP). We tested different magnetic complex compositions and determined uptake efficiency and cytotoxicity by flow cytometry. Additionally, we monitored the release, processing and functionality of delivered miR-335 with confocal laser scanning microscopy, real-time PCR and live cell imaging, respectively. On this basis, we established parameters for construction of magnetic non-viral vectors with optimized uptake efficiency (~75%) and moderate cytotoxicity in hMSCs. Furthermore, we observed a better transfection performance of magnetic complexes compared to PEI complexes 72 h after transfection. We conclude that MNP-mediated transfection provides a long term effect beneficial for successful genetic modification of stem cells. Hence, our findings may become of great importance for future in vivo applications.

Figure 1

Schematic representation of magnetic transfection complexes Magnetic transfection complexes consist of streptavidin coated paramagnetic iron oxide nanoparticles in the core and miR/PEI polyplexes bound on them via streptavidin-biotin connections.

Figure 2

Characterization of hMSC (A–C) Differentiation capacity of hMSCs was shown by immunostaining of aggrecan (green) for chondrocytes (A), osteocalcin (red) for osteocytes (B) and FABP-4 (green) for adipocytes (C); Nuclei were stained with DAPI (blue); (D,E) Immunophenotyping of hMSCs was performed by flow cytometry after staining for specific CD surface markers. The bright gray areas indicate CD marker isotope controls (D); Surface marker expression values are in percentage of positive cells and represented as mean ± standard error, n = 5 (E).

Figure 3

Transfection optimization with magnetic polyplexes in hMSCs. hMSCs were transfected with Cy™3 labeled miR/PEI or miR/PEI/MNP complexes and the uptake efficiency (A,C) and cytotoxicity (B,D) were determined by flow cytometry 5 h after transfection. (A,B) miR/PEI or miR/PEI/MNP complexes with various miR amounts (2.5, 5, 15 pmol/cm² miR) at NP ratio 10 with MNP concentrations ranging from 0.5 to 2 μg/mL iron (MNP 0.5 to MNP 2); (C,D) Different NP ratios (NP 2.5, 10, 33) with varied MNP amount ranging from 0.5 to 6 μg/mL iron (MNP 0.5 to MNP 6). miR amount was kept constant (5 pmol/cm² miR). Cells treated with naked miR were used as control. Values are presented as mean ± standard error, n = 3, *p ≤ 0.05 versus miR.

Figure 4

Characterization of transfection complexes. (A) Condensation of miR by PEI was examined by gel electrophoresis. Polyplexes with NP ratios from 0.1 to 33 and 20 pmol miR were investigated. miR alone was used as positive control. At NP ratio 0.5 the miR signal in the gel disappeared. Due to the bigger size of PEI complexes compared to miR alone, complexes do not run into the gel, but remain in the slots. This indicates a complete binding of miR to PEI; (B,C) Surface charge (B) and particle size (C) of MNP and transfection complexes were determined by DLS and PALS. Magnetic polyplexes with NP ratio 10, 20 pmol miR and MNP concentrations ranging from 0.5 to 6 μg/mL iron (MNP 0.5 to MNP 6) were used. Zeta potential data are presented as mean ± standard error, n = 10. Particle size data are presented as mean ± standard deviation, n = 10.

Figure 5

Processing of transfected precursor-miR. (A) hMSCs were transfected with precursor-miR-335 using miR/PEI or miR/PEI/MNP complexes and level of a mature miR-335 strand was detected by real time PCR 5, 24 and 72 h after transfection. Cells treated with miR only were used as a control. Dashed line indicates miR-335 expression in untransfected cells. Right plot shows a linear scale of miR-335 expression 72 h after transfection. Values were normalized to RNU6B expression and represented as mean ± standard error. The data are representative of 5 independent biological experiments (n = 5), each of which was measured in qPCR-triplicates. **p ≤ 0.001 versus miR, ^##p ≤ 0.001 versus miR/PEI-mediated transfection; (B,C) Labeled miR/PEI (B) and miR/PEI/MNP complexes (C) were visualized by confocal laser scanning microscopy 72 h after transfection in hMSCs. miR-335 was labeled with Cy™5 dye (cyan), PEI was labeled with Oregon Green^® 488 (yellow) and MNPs were labeled with Atto 565 (red). Nuclei were counterstained with DAPI (gray). The arrow indicates condensed miR/PEI complexes inside the nucleus. Scale bar = 5 μm.

Figure 6

Efficient knockdown of miR-335 target genes. (A,B) hMSCs were transfected with miR/PEI or miR/PEI/MNP complexes and relative gene expression of TNC (A) and RUNX2 (B) was measured by real-time PCR 5, 24 and 72 h after transfection. Cells treated with miR only were used as control. Dashed line indicates gene expression in untransfected cells. Values were normalized to GAPDH expression and are represented as mean ± standard error, n = 5, *p ≤ 0.05 versus miR, **p ≤ 0.001 versus miR, ^##p ≤ 0.001 versus miR/PEI - mediated transfection.

Figure 7

Migration activity of hMSCs after miR-335 transfection. (A) hMSCs were transfected with miR/PEI/MNP complexes and migration activity was tested 24 h after transfection. The overgrown surface area of transfected cells was measured before and 12 h after scratching. Untransfected cells were used as control. Data are presented as mean ± standard error, n = 5, **p ≤ 0.001; (B,C) Representative images after transfection with magnetic polyplexes containing either scrambled miR (B,B′) or miR-335 (C,C′). Images were taken immediately after (B,C) and 12 h after scratching (B′,C′). Values represent the non-overgrown surface area. Scale bar = 200 μm.