1. ¹Center for Gene Therapy, Tulane University Health Sciences Center, New Orleans, LA, USA
2. ²Department of Tumor Biology, Institute for Cancer Research, Norwegian Radium Hospital, Montebello Oslo, Norway

Correspondence: DJ Prockop, Tulane University New Orleans, Center for Gene Therapy, Tulane University Health Sciences Center, SL 99, 1430 Tulane Avenue, New Orleans, LA 70112, USA. E-mail: dprocko@tulane.edu

Received 21 February 2003; Accepted 8 July 2003.

Abstract

Human marrow stromal cells (hMSCs) are an attractive source of adult stem cells for autologous cell and gene therapy. To transfect hMSCs without the use of viruses, we developed improved conditions for stable transfection of the cells by electroporation. hMSCs were isolated by adherence to plastic, and were electroporated at 600 V and 100 s in a 2-mm gap cuvette with a plasmid containing enhanced green fluorescence protein (EGFP) and neomycin phosphotransferase gene (neo^r). After electroporation of 10⁶ cells with 10 g of the linearized plasmid DNA, hMSCs with stable DNA integration were selected by culturing with 200 g/ml G418. The transfected hMSCs were expanded another 300-fold in 14 days to obtain 89 million cells, of which 98% expressed EGFP. Chloroquine increased the number of hMSCs transiently expressing EGFP from 12% to over 50%, but decreased stable integration. Stable integration of plasmid DNA into rat MSCs by electroporation was also successful. The transfected MSCs retained their capacity to differentiate into both adipocytes and osteoblasts. Thus, MSCs were stably transfected with plasmid DNA and retained their differentiation capacity after expansion.

Marrow stromal cells (MSCs) are an attractive cell line for the development of cell and gene therapy because they are readily available from patients, and because they can differentiate into osteoblasts, adipocytes, chondrocytes, cartilage, skeletal muscle, cardiac myocytes, and early progenitors of neural cells.^1,2,3,4,5 MSCs may be useful in treating diseases through the use of autologous stem cell transplantation, but for the treatment of genetic disorders, the stem cells must be modified to express the exogenous gene.

For the experiments reported here, we considered several methods of introducing genes into MSCs. In preliminary experiments, we found that use of cationic lipids (Gene Porter and Gene Porter 2; Gene Therapy Systems, San Diego, CA, USA) or a proprietary electroporation system (Nucleofector; amaxa biosystems, Gmblt, Cologne, Germany) yielded transient transfectants of MSCs (14–16%), but inadequate yields of stable transfectants (not shown). We elected not to employ viral vectors because of their tendency to be inactivated through methylation, induce immune reactions, and impose a limit on the size of the exogenous gene that can be introduced.^6,7,8 Instead, we selected to develop improved conditions for transduction of MSCs by electroporation. Keating et al⁹ previously used electroporation to create stable cell lines of human MSCs (hMSCs) by expressing the immortalizing SV40 large T-antigen.^9,10 Our aim was to obtain stable transfectants without immortalizing the cells.

hMSCs were isolated from 2 ml aspirates of bone marrow from the iliac crest, as previously described.^11,12 Briefly, the aspirate was separated with a density gradient and the mononuclear cells were plated in complete medium at 37°C and 5% humidified CO₂. After 24 h, the nonadherent cells are removed and the cells are cultured for 5–7 days. The hMSCs were expanded through one passage by plating at 50 cells/cm² and incubating the cultures for 12 days. Rat MSCs (rMSCs) were isolated from the long bones of adult male Lewis rats as previously described, and expanded at 50 cells/cm² for 10 days prior to electroporation.¹³ Electroporation was performed using the Eppendorf Multiporator, which provides a constant square pulse wave and ease of optimization. Initially, the pulse length and strength had to be optimized to limit cell death, while being intense enough to allow the DNA to cross the cell membrane.

For electroporation (Figure 1), the cells were counted and resuspended in hypo-osmolar buffer. We tested 1, 5, 10, 15, and 20 g of plasmid DNA with 10⁶ hMSCs. A measure of 10 g of linearized plasmid vector was found to provide the best transfection efficiency and cell recovery (not shown). The voltage was varied from 300 to 700 V in 100 V increments. Additionally, the length of the pulse was varied from 50 to 125 s in 25 s increments. Immediately after electroporation, the cells were removed from the cuvette and plated on three 10-cm diameter tissue culture dishes in complete medium. After 24 h, the plates were washed twice with PBS and re-fed with complete medium.

Figure 1.

Flow chart for electroporation of hMSCs and maps of plasmids. (a) Schematic diagram of the electroporation protocol. hMSCs were seeded into a cell factory (surface area 6320 cm²; Nunc) at 50 cells/cm² and grown in complete culture medium (MEM (Cellgro, Herndon, VA, USA) supplemented with 18% FCS (lot selected for rapid growth, Atlantic Biologicals, Norcross, GA, USA), 100 units/ml penicillin/100 g/ml streptomycin (Invitrogen, Carlsbad, CA, USA), and 1 mM L-glutamine (Invitrogen, Carlsbad, CA, USA)) for 10 days. The cells were isolated by incubation in 0.25% trypsin and 1 mM EDTA (Invitrogen) at 37°C for 5 min. For electroporation, the cells were counted and resuspended in hypo-osmolar buffer (Eppendorf) at a concentration of 2.5 10⁶ cells/ml. A volume of 400 l of cell suspension was mixed with 10 g of linear plasmid DNA (pEGFP-IRESneo Vector for EGFP expression; Clontech, Palo Alto, CA, USA) in a sterile electroporation cuvette (2-mm gap width, Eppendorf). The voltage was varied from 300 to 700 V in 100 V increments. Additionally, the length of the pulse was varied from 50 to 125 s in 25 s increments. Immediately after electroporation, the cells were removed from the cuvette and plated on three 10-cm diameter tissue culture dishes (Nunc, Rochester, NY, USA) in complete medium. After 24 h, the plates were washed twice with PBS and re-fed with complete medium. To obtain stably transfected MSCs, the samples were incubated in complete medium for 72 h and then the complete medium was supplemented with 200 g/ml G418 sulfate (Invitrogen). The plates were washed with PBS and fresh selection medium was added every 2–3 days. After 14 days of selection, the stable expression of EGFP was determined by FACS analysis (FACSVantageSE; Becton-Dickinson, San Jose, CA, USA). To visualize the colonies, the plates were stained with 3% Crystal Violet in methanol for 10 min, and then washed with water three times. (b) The pEGFP-IRESneo vector was used for all experiments to allow the expression of both EGFP and the neo^r gene on one mRNA linked by an IRES sequence EGFP was inserted into the multiple cloning site (MCS). Electroporation with the control plasmid, pIRESneo, produced cells stably expressing the neo^r gene, but not EGFP.

Full figure and legend (59K)

To assay transient expression, the hMSCs on one of the three plates were isolated and assayed by flow cytometry for EGFP-expressing cells. After electroporation with pEGFP-IRESneo at 600 V and 125 s, transient expression of EGFP was about 12% (Figure 3h), with background fluorescence gated at less than 0.5% (Figure 3g). EGFP-positive hMSCs were readily detected by epifluorescence microscopy after 72 h (Figure 3d).

Figure 3.

Transient and stable expression of EGFP: hMSCs expressing enhanced green fluorescent protein (pIRES-eGFP, Clontech). (a, b) Control hMSCs expressing neomycin resistance gene after 2 weeks of selection in G418. (c, d) Transient expression of EGFP from hMSCs 3 days after electroporation. (e, f) Stable expression of EGFP from a clone of hMSCs after 2 weeks of selection in G418 for neomycin resistance. All photographs are at 200 magnification. FITC photographs (b, d and f) were exposed for 2 s. (g) FACS analysis of stable control electroporation with the pIRESneo control plasmid containing neo^r, but not EGFP. (h) Transient expression of EGFP after 24 h. (i) Stable expression of EGFP after selection; greater than 97% of the MSCs were positive for EGFP expression. (j) Stable expression of EFGP after 300 000 expansion of MSCs. Pooled colonies were plated at 50 cells/cm² in complete medium. After 2 weeks, the cells were harvested and analyzed by FACS.

Full figure and legend (414K)

To obtain stably transfected hMSCs, the samples were incubated first in complete medium for 72 h and then in complete medium supplemented with 200 g/ml G418 sulfate. After selection, cells electroporated at 600 V and 100 s were 97% EGFP positive when assayed by flow cytometry (Figure 3i). Colonies of hMSCs were detectable on the plates, and showed bright EGFP expression (Figure 3f).

The hMSCs stably expressing EGFP exhibited a phenotype characteristic of untransfected hMSCs. Up to 85 clones were obtained (Figure 2c). The pooled clones were expanded 300-fold to 89 million cells in 14 days, characteristic of the rapid proliferation of MSCs. About 98% of the hMSCs expressing EGFP as assayed by flow cytometry (Figure 3j). The cells continued to express high levels of EGFP after expansion by passage at low plating densities of 50 cells/cm² every 2 weeks over a 6-month period. The hMSCs were also multipotential, as evidenced by their ability to differentiate into adipocytes and osteoblasts after selection and expansion.

Figure 2.

Optimization of electroporation for transient and stable expression of EGFP in hMSCs and rMSCs. (a) Transient expression of EGFP of hMSCs, as determined by FACS analysis. The voltage was varied from 100 to 900 V, in 100-V increments. The length of the pulse was varied from 25 to 125 s, in 25-s increments. Transient and stable expression of EGFP was determined by FACS analysis after 72 h and 3 weeks, respectively. (b) Transient expression of EGFP was increased after treatment with chloroquine. The hMSCs were electroporated with 600 V and 100 s, and then immediately transferred to complete medium containing 50 or 100 M chloroquine. After incubating the hMSCs with chloroquine for 0.5, 2 and 4 h at 37°C, the cells were recovered by centrifugation, re-suspended in complete medium, and plated on three 10-cm diameter plates. Transient EGFP expression was determined 72 h after electroporation. (c) Number of clones produced after selection as a function of the voltage and the length of time of the pulse. After the cells were selected with G418 for 14 days, the plates were stained with 3% Crystal Violet in methanol for 10 min. The plates were washed three times with distilled water and the colonies counted. (d) Optimization of the electroporation of rMSCs. The rMSCs were electroporated as above, with 1 10⁶ cells mixed with 10 g of linear plasmid DNA. The pulse strength was varied from 400 to 800 V for 100 s. After selection for stable integration of the plasmid DNA, the clones were visualized by staining with 3% Crystal Violet as above. N=3 for all conditions.

Full figure and legend (120K)

For differentiation, the hMSCs were plated at high density and the medium was changed from complete medium to differentiation medium 24 h later. After 21 days in osteogenic medium, Alizarin Red-positive mineral deposits were visible all through the culture (Figure 4g). After 21 days in adipogenic medium, the lipid vacuoles were easily identified as bright red inclusions after Oil Red O staining within the cells (Figure 4h). The EGFP expression was still evident after differentiation to adipocytes and osteoblasts (Figure 4c–f).

Figure 4.

Differentiation of hMSCs stably expressing EGFP: differentiation of hMSCs to bone and fat. After selection for stable clones expressing GFP and neo^r, the hMSCs were plated at high density and exposed to the differentiation medium. Control hMSCs (a, b) were incubated in control medium. Bone differentiation is seen in (c, d), and stained with Alizarin Red for mineral deposition (g). Osteogenic differentiation medium contained MEM supplemented with 10% FBS, 1% penicillin, 1% streptomycin, 1 mM L-glutamine, 1 10^-8 M dexamethasone (Sigma), 10 M -glycerol phosphate (Sigma), and 50 M ascorbic acid-2 phosphate (Sigma). The medium was changed every 3–4 days. After 21 days, the plates are fixed in 10% neutral-buffered formalin for 20 min and stained with 0.5% Alizarin Red (pH 4.1; Sigma). The plates were washed 3 with water and the mineral deposits visualized by a dark red stain. Adipocyte differentiation is seen in (e, f), and stained for lipid vacuoles in (h) with Oil Red O. For adipogenic differentiation, the MSCs were grown in MEM supplemented with 10% FBS, 100 units/ml penicillin, 100 g/ml streptomycin, 1 mM L-glutamine, 0.5 M hydrocortisone (Sigma, St Louis, MO, USA), 0.5 M IBMX (Sigma), and 60 M indomethane (Sigma) for 21 days. The plates were then fixed in 10% neutral-buffered formalin for 20 min and stained with Oil Red O (stock: 0.5% in isopropanol, mixed three parts stock to two parts water and filtered through a 0.2 m filter; Sigma) for 30 min. After thorough washing of the plates with PBS, the lipid vacuoles were easily identified as bright red inclusions within the cells.

Full figure and legend (700K)

Electroporation of rMSCs was also optimized using similar conditions. After selection for stable integration of the plasmid DNA, the clones were visualized with Crystal Violet. After electroporation at 700 V, the rMSCs produced up to 340 clones per million cells electroporated (see Figure 2d).

In further experiments, hMSCs immediately after electroporation were treated with chloroquine to transiently inhibit the lysosomes and limit plasmid degradation.^14,15,16 FACS analysis for EGFP expression in these cells indicated that exposure to chloroquine increased the transient expression from 12 to 50% (Figure 2b). However, after selection with G418 for stably expressing hMSCs, only one colony was visualized on one plate. On the 17 other tissue culture plates, no colonies were visualized (data not shown). The results suggest that the chloroquine, although it can greatly increase the number of MSCs transiently expressing the transgene, inhibited the integration of the plasmid into the MSCs chromosomal DNA, thereby limiting stable expression of the transgene.

Electroporation was shown here to be an efficient method for stably expressing a transgene in hMSCs and rMSCs. The transfected MSCs can easily be expanded to produce 50 million cells within 2 weeks.