MPC characterization
MPC isolated from skeletal muscles of TdTomato and Luciferase positive mice were analyzed for their myogenic identity and skeletal muscle differentiation potential in vitro. Desmin, a general myogenic cell marker and intermediate type filament necessary for muscle contraction [20], stained positive in 94.67 ± 9.24% (mean ± SD) of MPC compared to non-myogenic cells (NMC) (Fig. 1A). Furthermore, MPC cultivated under differentiating conditions, were found to be fusion-competent and formed multinucleated myofibers (Fig. 1B). Moreover, MPC demonstrated significantly higher AChE activity compared to NMC (Fig. 1C), suggesting high skeletal myogenic differentiation potential of MPC. CD9, a tetraspanin surface marker required for normal fusion of myotubes and muscle regeneration [21], as well as CD98, a surface marker identifying activated muscle stem cells [22], were found highly positive at a mean ± SD of 79.75 ± 8.45 (CD9) % and 86.71 ± 7.54 (CD98) % in MPC, respectively (Fig. 1D).
Cell implantation: dose definition and cell distribution
In order to study the fate of MPC following intramuscular injection, an in vivo model of MPC implantation and tracking was established. MPC from Luciferase or TdTomato mice were implanted into the tibialis anterior or gastrocnemius muscle of immunodeficient SHO-PrkdcscidHrhr mice. It has previously been shown that intramuscularly injected myogenic cells do not migrate well [23]. Thus, to distributed MPC over the whole muscle, a multiple needle applicator with 4 needles (each 30G) placed at a distance of 1 mm from each other was designed (Supplementary Fig. 3). Signals of Luciferase MPC injected by single or multi-needles were visible at the site of injection over the whole study course of 75 days (Figs. 2A + B), suggesting long-term engraftment of cells in both cases. Significantly higher luciferase signals were found when cells were injected by a 4-needle multi-needle applicator compared to those injected by a single needle applicator, suggesting that a 4-needle distribution is favorable to reach high cell engraftment in the subsequent ischemia approach (Fig. 2B). Immunohistological analysis of muscle specimen injected with TdTomato cells revealed TdTomato positive myofibers on POD 70 (Fig. 2C), suggesting fusion of MPC with existing myofibers and/or formation of new myofibers. Successful engraftment of injected cells is necessary for myofiber formation and thus hypothesized to be essential for regenerative effects of MPC. Therefore, an effort was made to increase MPC engraftment. Optimal cell dose per needle for an injection depth of 0.25 mm was calculated to be 2.5*105 cells per needle according to Skuk et al. (1*106 total for a 4-needle applicator) [24]. Comparison of luciferase signals emitted by either 1*106 or 1*105 cells injected per muscle over time, revealed significantly lower signals in muscles injected with fewer cells (Fig. 2D), thus confirming dose-signal relation and cell dose definition. Again, luciferase signals significantly increased over time until POD 75, thus suggesting ongoing proliferation of injected MPC (Fig. 2D).
Hindlimb ischemia/reperfusion model
To establish a murine model of muscle damage due to warm ischemia and reperfusion in extremities, WIT ranging from 30 min to 3 h were tested with the goal to induce muscle damage without major muscle necrosis. Short WIT (group A1, 30 min WIT) did not lead to macroscopic and only minimal histopathologic changes (Fig. 3A), including few internal nuclei with prominent nucleoli. Macroscopic signs of IRI including progressive swelling and erythema of the ischemically injured leg were observed in animals subjected to prolonged (≥1 h) WIT (data not shown). Histopathologic evaluation of muscle biopsies in groups A2 (1 h WIT) and A3 (2 h WIT) demonstrated mild and moderate to severe leukocyte infiltrations on POD 3 and 7, respectively (Fig. 3A). In addition, signs of muscle regeneration were seen in biopsies by POD 14 (Fig. 3A). Especially muscle tissue in group A3 displayed characteristic cytoplasmatic basophilia on POD 3 and internal rows of vesiculated nuclei with prominent nucleoli on POD 14 (Fig. 3A, bottom row). While animal survival was excellent in groups A1 to A3, a sharp decline in animal survival to 25% was observed after increasing WIT to 3 h in group A4 (Fig. 3B). As animals in group A3 displayed signs of moderate to severe ischemic injury and excellent postoperative survival rates, 2 h of WIT were identified as most suitable to investigate the regenerative properties of myogenic progenitor cells.
MPC engraftment and persistence following IRI in vivo
In order to study the engraftment and persistence of implanted MPC following IRI, luciferase reporter expressing MPC were injected into the tibialis anterior muscle of C57BL/6 mice without WIT (group D1) or after 2 h WIT (group D2). Bioluminescence of injected cells was visible throughout the entire study period (POD 75) in all mice of groups D1 and D2 (Fig. 4A). The signal persisted at the area of injection without further distribution. Quantification of the bioluminescence signals demonstrated increasing signals over time in both study groups (p < 0.001) and consistently higher signals in the 2 h WIT (D2) animals (Fig. 4B) (p = 0.0155). These findings suggest that MPC engraft and persist at the injection site, with a higher number of cells present after IRI.
Interplay of MPC engraftment, tissue damage and regeneration following IRI in vivo
Biopsies of the tibialis anterior muscle from animals subjected to 2 h WIT taken on POD 2 (group B) displayed moderate to severe leukocyte infiltration without signs of major muscle necrosis. No difference in the extent of leukocyte infiltration or muscle cell damage was observed between animals receiving sham (B1, Fig. 5A) or MPC injections (B2, Fig. 5B). In contrast to animals in group B1, accumulating cell infiltrates were seen around the co-injected beads in all animals of group B2 (Fig. 5B, bottom row). TdTomato and desmin + cells were detected at these sites (Fig. 6A), suggesting that injected cells remained at the injection site co-located with fluorescent beads.
On POD 14, pronounced muscle regeneration as evident from the presence of myotubes with multiple internal vesiculated nuclei and prominent nucleoli, known to occur following fusion of single nucleated muscle progenitor cells with each other [25], was observed in all groups challenged by WIT (Fig. 7A and 7B, black circles). Animals with sham injection (C1) and with MPC injection (C2) both displayed a median muscle regeneration score (Table 2) of 3 (range, C1: 1–3; C2: 3–3; p > 0.9) (Fig. 8A). On POD 14, only few MPC were left aggregated around beads in C2 mice (Fig. 7B, middle row). As assessed through fluorescence imaging, the TdTomato signal, originally expressed by injected MPC, was located in newly formed (central nuclei containing) desmin expressing myofibers, indicating that injected MPC have contributed to myofiber regeneration (Fig. 6B). Accordingly, a low degree of fibrosis and tissue damage was observed in muscle biopsies of both groups (Fig. 7A and 7B, bottom row).
75 days after MPC injection only very few cells were still present in close proximity to the co-injected beads in ischemically injured animals of group D1 (Fig. 9B). This was in stark contrast to histopathologic results from animals in group D2 (MPC injection in sham operated animals, no WIT), where large aggregates of cells were still visible around the co-injected beads (Fig. 9A), suggesting that IRI increased the cellular turnover in infiltrates at the injection site. In group D1, a high degree of muscle regeneration was still seen on POD 75 reflecting in a median regeneration score of 3 (range, 3–3). In contrast, animals in group D2 displayed little to no muscle tissue regeneration and thus a median score of 0 (range, 0–1; p = 0.029; Fig. 9B), suggesting that IRI led to an increase in muscle regeneration. Similarly to earlier time-points, no relevant amounts of fibrosis and scaring were seen 75 days after MPC injection.