Direct head-to-head comparison of cationic liposome-mediated gene delivery to mesenchymal stem/stromal cells of different human sources: a comprehensive study

Abstract

Nonviral gene delivery to human mesenchymal stem/stromal cells (MSC) can be considered a very promising strategy to improve their intrinsic features, amplifying the therapeutic potential of these cells for clinical applications. In this work, we performed a comprehensive comparison of liposome-mediated gene transfer efficiencies to MSC derived from different human sources-bone marrow (BM MSC), adipose tissue-derived cells (ASC), and umbilical cord matrix (UCM MSC). The results obtained using a green fluorescent protein (GFP)-encoding plasmid indicated that MSC isolated from BM and UCM are more amenable to genetic modification when compared to ASC as they exhibited superior levels of viable, GFP(+) cells 48 hr post-transfection, 58 ± 7.1% and 54 ± 3.8%, respectively, versus 33 ± 4.7%. For all cell sources, high cell recoveries (≈50%) and viabilities (>85%) were achieved, and the transgene expression was maintained for 10 days. Levels of plasmid DNA uptake, as well as kinetics of transgene expression and cellular division, were also determined. Importantly, modified cells were found to retain their characteristic immunophenotypic profile and multilineage differentiation capacity. By using the lipofection protocol optimized herein, we were able to maximize transfection efficiencies to human MSC (maximum of 74% total GFP(+) cells) and show that lipofection is a promising transfection strategy for MSC genetic modification, especially when a transient expression of a therapeutic gene is required. Importantly, we also clearly demonstrated that intrinsic features of MSC from different sources should be taken into consideration when developing and optimizing strategies for MSC engineering with a therapeutic gene.

FIG. 1.

Transgene delivery into mesenchymal stem cells (MSC) derived from BM, AT, and UCM. Quantification and long-term monitoring of the levels of total GFP⁺ (gray squares) and GFP⁺PI⁻ (black bars) (percentage) were performed upon lipofection for each cell source. Comparison of transfection for cells plated at 1000 cells/cm² (a) and 3000 cells/cm² (b). Each bar and point represent the mean±standard error of mean (SEM), n=4, *p<0.05. BM, bone marrow; AT, adipose tissue; UCM, umbilical cord matrix; GFP, green fluorescent, PI, propidium iodide.

FIG. 2.

Yield of transfection and cell recovery after lipofection of MSC from the different sources. Comparison of cells plated at 1000 cells/cm² (a) and 3000 cells/cm² (b). Yield of transfection is represented as solid bars, whereas gray squares correspond to cell recovery. Each bar and point represents the mean±standard error of mean (SEM), n=4, *p<0.05.

FIG. 3.

Intracellular plasmid copy number present into BM- (a), AT- (b), and UCM-MSC (c) after lipofection. Two initial cell densities, 1000 cells/cm² (black bars) and 3000 cells/cm² (white bars) were tested. Each bar represents the mean±standard error of mean (SEM), n=4, *p<0.05.

FIG. 4.

Impact of liposome-mediated gene delivery on the proliferative status of MSC from the different human sources. The number of viable cells was determined for both transfected cells (solid lines) and nontransfected cells (dashed lines) (a). Representative microscopy images of UCM-, BM-, and AT-derived MSC plated at 3000 cells/cm² on days 1, 2, and 3 after transfection (b), where cells were still subconfluent. Fluorescence microscopy images were merged with bright field optical microscopy images. Each point represent the mean±standard error of mean (SEM), n=4, *p<0.05. Color images available online at www.liebertpub.com/hgtb

FIG. 5.

Cell division kinetics of MSC from different human sources after lipofection. Cells were plated at the initial cell density of 3000 cells/cm². Throughout culture GFP⁺ (black bars) and non-transfected MSC (white bars) were analyzed by flow cytometry and for each time point the percentage of cell population in each generation was determined using the Proliferation Wizard of the ModFit Software (a). GFP⁺ and GFP⁻ populations were discriminated using distinct FL1/FL3 gates as shown in the representative flow cytometry dot plots of MSC expressing GFP and labeled with PKH 26 dye (b). Each bar represent the mean±standard error of mean (SEM), n=3.

FIG. 6.

Immunophenotype, multilineage diferentiative potential, and clonogenic potential of MSC from different human sources after transfection. Flow cytometric analysis of surface antigen expression showed that lipofection had no impact on MSC immunophenotype profile (a). After gene delivery, MSC were induced to differentiate and stained for osteogenesis (Von Kossa and alkaline phosphatase), adipogenesis (Oil Red-O), and chondrogenesis (Alcian blue). No major differences were noticed among MSC sources. Representative fluorescence microscopy images of transfected UCM-MSC are merged with bright field optical microscopy images (b). Clonogenic potential of genetically modified MSC isolated from BM, AT, and UCM (black bars) compared to nontransfected cells (white bars) (c). Each bar represents the mean±standard error of mean (SEM), n=2. Color images available online at www.liebertpub.com/hgtb