Enhancement of myogenic and muscle repair capacities of human adipose-derived stem cells with forced expression of MyoD

Abstract

Muscle disorders such as Duchenne muscular dystrophy (DMD) still need effective treatments, and mesenchymal stem cells (MSCs) may constitute an attractive cell therapy alternative because they are multipotent and accessible in adult tissues. We have previously shown that human multipotent adipose-derived stem (hMADS) cells were able to restore dystrophin expression in the mdx mouse. The goal of this work was to improve the myogenic potential of hMADS cells and assess the impact on muscle repair. Forced expression of MyoD in vitro strongly induced myogenic differentiation while the adipogenic differentiation was inhibited. Moreover, MyoD-expressing hMADS cells had the capacity to fuse with DMD myoblasts and to restore dystrophin expression. Importantly, transplantation of these modified hMADS cells into injured muscles of immunodepressed Rag2(-/-)gammaC(-/-) mice resulted in a substantial increase in the number of hMADS cell-derived fibers. Our approach combined the easy access of MSCs from adipose tissue, the highly efficient lentiviral transduction of these cells, and the specific improvement of myogenic differentiation through the forced expression of MyoD. Altogether our results highlight the capacity of modified hMADS cells to contribute to muscle repair and their potential to deliver a repairing gene to dystrophic muscles.

Figure 1

Characterization of MyoD-hMADS cells. (a–c) Transduction of hMADS cells with the lentiviral vector HIV PGK-MyoD. (a) hMADS cells were labeled with an anti-MyoD antibody revealed with peroxydase staining 2 days after transduction. Note the positive nuclear staining of almost all cells. (b) Effect of MyoD on hMADS clonogenicity expressed as the percentage of clones compared to that obtained with WT-hMADS cells (WT). MOI, multiplicity of infection; n = 7 independent experiments and *P = 0.001. (c) Flow cytometry analysis comparing the percentage of CD56⁺ cells in WT- and MyoD-hMADS cells. In this representative experiment, cells were used one passage after viral transduction and MyoD expression enhanced the percentage of CD56⁺ cells from 2% in WT-hMADS cells to 33% in MyoD-hMADS cells. (d–f) Myogenic differentiation of MyoD-hMADS cells. (d) The microscope phase-contrast view shows a typical field of MyoD-hMADS cells 10 days after the switch in myogenic differentiation medium. Multinucleated myotube-like structures are clearly visible on a background of mononucleated cells. (e) Abundance of the cellular elongated syncitia varied according to the MOI levels. The histogram indicates the fusion index of three independent experiments and *P = 0.05. The fusion index was calculated as the ratio of the number of nuclei inside myotubes to the number of total nuclei × 100 at day 8 of myogenic differentiation. The number of nuclei was estimated by the average of nuclei counted in 20 independent and randomly chosen microscope fields. A myotube was defined by the presence of at least three nuclei within a continuous cell membrane. (f) Comparison by western blot of MyoD phosphorylation status between WT- and MyoD-hMADS cells after 5 and 13 days in myogenic differentiation medium. Total cellular extracts were stained with an anti-MyoD antibody (1/200) for phosphorylated (47 kd) and dephosphorylated (45 kd) MyoD protein. A clear shift toward dephosphorylated MyoD was reproducibly found. Lanes were normalized according to protein bands labeled with an anti-tubulin antibody (not shown). hMADS, human multipotent adipose-derived stem.

Figure 2

Muscle differentiation markers expressed in MyoD-hMADS cells. (a) Comparative expression by RT-PCR between WT- and MyoD-hMADS cells after 3 and 7 days in myogenic differentiation medium. Many markers could be amplified at day 7 for MyoD-hMADS cells. PCR was run for 30 cycles, which gave sufficient amplification signal for MyoD-hMADS cells at day 7 without any amplification for other cDNAs under comparison. All the primers were human sequence specific. (b) Immunofluorescence analysis of MyoD-hMADS cells in myogenic differentiation medium. The muscle proteins desmin, nestin, skeletal myosin, and dystrophin were detected using specific antibodies revealed with Rhodamin fluorochrome (red). Nuclei were counterstained with Hoechst 34580 (blue) to detect multinucleated myotubes. Note the presence of mononucleated hMADS cells also expressing desmin beside the desmin positive myotubes. Skeletal myosin is detected with an anti-myosin heavy chain. Dystrophin could be observed in some myotubes at its expected membrane localization. Bar = 20 µm. cDNA, complementary DNA; RT-PCR, reverse transcriptase-PCR; WT-hMADS, human multipotent adipose-derived stem.

Figure 3

hMADS and DMD cell cocultures. Dystrophin-deficient DMD myoblasts were transduced with an eGFP lentiviral vector (>90% efficiency measured 2 days after transduction) and cocultured with MyoD-hMADS cells (MOI 30). Cells were cultured on glass to allow microscopic examination and were stained at the end of the longest survival time, ~18 days under such conditions. Immunostaining of dystrophin was observed with an Alexa Fluor 594-conjugated secondary antibody. The figure shows two views of myotubes positive for both eGFP (green) and dystrophin (red) which are hybrid myotubes formed from DMD and MyoD-hMADS cells. Nuclei were labeled in blue with Hoechst-34580 and show the multinucleated structure of myotubes. Myotubes were often found isolated in the cultures because of the plating on glass, which was required for the immunostaining. Bar = 20 µm. DMD, Duchenne muscular dystrophy; eGFP, enhanced green fluorescent protein; hMADS, human multipotent adipose-derived stem; MOI, multiplicity of infection.

Figure 4

Differentiation potential of MyoD-hMADS cells. (a) The adipogenic differentiation potential was compared between WT- and MyoD-hMADS cells. The phase-contrast microscope views show typical fields of WT- and MyoD-hMADS cells (MOI 30) subjected to adipogenic culture conditions for 2 weeks. Adipocytes are characterized by the presence of tiny lipid droplets in their cytoplasm. They cover the entire culture dish with WT-hMADS cells, while they are very rare with MyoD-hMADS cells (red arrow). The MyoD-hMADS cells could also be found as multinucleated myotubes (black arrow), in accordance with the very close adipogenic and myogenic culture conditions. The histogram visualizes a representative experiment comparing the GPDH adipocyte marker activity between WT- and MyoD-hMADS cells at the indicated MOI. (b) A single clone in adipogenic differentiation medium is shown at low and high magnification after Oil Red O staining for triglycerides. The histogram compares the percentage of adipogenic clones in WT- and MyoD-hMADS cells according to MOI; n = 5; *P = 0.05 and **P = 0.001. (c) For comparison, the same clonal analysis as in b is shown in myogenic differentiation medium. The low magnification picture is a phase-contrast view of a single myogenic clone and the high magnification picture shows nuclei labeled with an anti-MyoD antibody revealed by peroxydase staining. As opposed to the adipogenic clones, the histogram of this representative experiment shows that the percentage of myogenic clones drastically increases with the MOI level. hMADS, human multipotent adipose-derived stem; MOI, multiplicity of infection; GPDH; glycerol-3-phosphate dehydrogenase; WT, wild type.

Figure 5

Detection of hMADS cells–derived nuclei in mouse muscles. The presence of hMADS cells–derived nuclei was assessed 1 month after injection of hMADS cells in the tibialis anterior muscles of Rag2^−/− γC^−/− mice. hMADS cells–derived nuclei were labeled on transverse cryostat sections with a human specific anti-lamin A/C antibody conjugated to Alexa Fluor 488 (green). The basal lamina was labeled with an antibody crossreacting with both human and mouse laminin and detected with Alexa Fluor 405 (blue). Human positive nuclei were found as central nuclei in regenerated fibers (arrow) or in cells located in interstitial spaces (arrowhead). Nuclei derived from MyoD-hMADS cells are presented. The same immunofluorescent pattern was observed for the different types of hMADS cells, except that MyoD-hMADS cell–derived nuclei were more often observed integrated in the muscle fibers than nuclei derived from LacZ- or WT-hMADS cells. Bar = 10 µm. hMADS, human multipotent adipose-derived stem; WT, wild type.

Figure 6

Detection of hMADS cells–derived muscle fibers in mouse muscles. The presence of hMADS cells–derived muscle fibers was assessed 1 month after injection of hMADS cells in the tibialis anterior muscles of Rag2^−/− γC^−/− mice. (a) Assembled representative views of transverse cryostat sections from muscles transplanted with MyoD-and WT-hMADS cells. The sections were immunolabeled with a human specific anti-spectrin β-chain antibody conjugated with Alexa Fluor 594. Note the variations in muscle fiber calibers. The histogram shows the quantification of human spectrin-positive fibers per section found for WT- and MyoD-hMADS cells at MOI 10 and 30; n = 18, 3, and 6 muscles, respectively, and *P = 0. 05. (b) Double staining of muscle fibers derived from hMADS cells. Sections found positive for anti-spectrin antibody were colabeled with human specific anti-dystrophin or anti-δ-sarcoglycan antibodies to ascertain whether the hMADS cells–derived muscle fibers were expressing characteristic human muscle proteins. All antibodies gave the expected membrane labeling of muscle fibers. Spectrin was labeled with Alexa Fluor 594 fluorochrome (red), dystrophin and δ-sarcoglycan with FITC fluorochrome (green), and nuclei with Hoechst-34580 (blue). (c) Expression of human muscle markers in mouse muscles analyzed by RT-PCR. RNA was extracted from 6 to 8 pools of 50 cryostat sections cut along the tibialis anterior muscles. All the amplifications were performed at 30 cycles with human specific primers. PCR amplifications are illustrated with cDNAs from muscles transplanted with MyoD-hMADS cells. Human cDNAs obtained from noninjected muscles were used for negative control (no cell). Equilibrium of cDNA quantities was verified with mouse β-actin. Bar = 20 µm. cDNA, complementary DNA; hMADS; human multipotent adipose-derived stem; MOI, multiplicity of infection; RT-PCR, reverse transcriptase-PCR; WT, wild type.