Human myogenic reserve cells are quiescent stem cells that contribute to muscle regeneration after intramuscular transplantation in immunodeficient mice

Satellite cells, localized within muscles in vivo, are Pax7+ muscle stem cells supporting skeletal muscle growth and regeneration. Unfortunately, their amplification in vitro, required for their therapeutic use, is associated with reduced regenerative potential. In the present study, we investigated if human myogenic reserve cells (MRC) obtained in vitro, represented a reliable cell source for muscle repair. For this purpose, primary human myoblasts were freshly isolated and expanded. After 2 days of differentiation, 62 ± 2.9% of the nuclei were localized in myotubes and 38 ± 2.9% in the mononucleated non-fusing MRC. Eighty percent of freshly isolated human MRC expressed a phenotype similar to human quiescent satellite cells (CD56+/Pax7+/MyoD−/Ki67− cells). Fourteen days and 21 days after cell transplantation in immunodeficient mice, live human cells were significantly more numerous and the percentage of Pax7+/human lamin A/C+ cells was 2 fold higher in muscles of animals injected with MRC compared to those injected with human myoblasts, despite that percentage of spectrin+ and lamin A/C+ human fibers in both groups MRC were similar. Taken together, these data provide evidence that MRC generated in vitro represent a promising source of cells for improving regeneration of injured skeletal muscles.

In this study, we generated and characterized human MRC and assessed their potential as a source of myogenic stem cells able to improve muscle regeneration in vivo. For this purpose, we generated high amounts of human MRC from freshly isolated primary human myoblast cultures and transplanted them in immunodeficient mice. We demonstrated that 80% of human MRC were quiescent Pax7 + /MyoD − cells, and importantly, that human MRC exhibited an enhanced survival and an enhanced potency to generate Pax7 + cells after transplantation in immunodeficient mice compared to human myoblasts. These data highlight a potential role of human MRC in improving muscle healing and in the maintenance of a pool of SC in vivo.
Myoblasts were replated, grown to confluence and cultured in differentiation medium (DM) for 48 hours. A majority of the cells formed myotubes in vitro (MF20 + /MEF2 + cells) with fusion index values of 62.0 ± 2.9% and 38.0 ± 2.9% of cells were human myogenic reserve cells (MRC), escaping the terminal differentiation process (Fig. 1c).
The expression pattern of myogenic factors (Pax7, MyoD, MyHC and Myf5) was also analyzed in myoblasts, myotubes and MRC by Western blot (n = 4). Proliferating myoblasts (in GM) expressed high level of MyoD and low level of Pax7. In MRC, compared to myoblasts, Pax7 expression was enhanced by 4 fold whereas MyoD expression was massively down regulated to reach very low levels (P < 0.005, n = 4, Fig. 2b). Myf5 expression was similar in myoblasts and in MRC but down regulated in myotubes. Purity of our cell preparation was confirmed by myosin heavy chain (MyHC) that is expressed only in myotubes extracts.
Finally, the myogenic potential of MRC generated in vitro was tested. Myotubes were removed from the culture dishes and exposed the remaining adherent mononuclear MRC to GM. This induced cell proliferation until
Improved survival of MRC as compared to human myoblasts after injection in lacerated murine muscles. Human quiescent Rluc + and proliferating myoblasts Rluc + were injected in lacerated Gastrocnemius muscles of immunodeficient mice and cell survival was quantified by bioluminescence imaging (BLI) at various time point. The percentage of cell survival at day 4, 7 14 and 21 always refer to the 100% survival obtained by measuring BLI 3 h after cell transplantation for each cell injection. No significant difference in the percentage of human live cells remaining in mice was observed at day 4 and day 7 between the 2 groups. By contrast, differences in human cell survival were observed 14 days after cell injection (52.3 ± 4.1% vs. 35.9 ± 5.2% for MRC and myoblast respectively) and 21 days after cell injection (54.1 ± 5.2% vs. 31.5 ± 4.8% for MRC and myoblast respectively). These differences were statically significant (P < 0.05, n = 12, Fig. 5a,b), suggesting that MRC survived better after xenotransplantation than that of myoblasts.
MRC and myoblasts contribute equally to lacerated muscle regeneration. Skeletal muscle fibers of human origin were defined as fibers containing two human-specific markers, i.e., human spectrin and a human lamin A/C. We also used an antibody against mouse/human dystrophin to reveal muscle fibers of both human and murine origin. Quantification of human fibers was defined as the ratio of spectrin positive fibers to   . The percentage of surviving cells was normalized respective to the signal measured 3 h after cell injection (day 0), and considered as 100% in each group. Data are represented as mean ± SEM (n = 12; *p < 0.05; **p < 0.001, using unpaired Student's t test). total number of Lamin A/C positive nuclei to compensate for variation in the number of human cells present in an injected muscle section. Three weeks after cell injection, transplanted myoblasts and MRC contributed equally to muscle regeneration as indicated by the presence of a similar percentage of spectrin + human fibers (related to the total number of lamin A/C + cells) in both groups (23.1 ± 7% and 27.1 ± 4.6% respectively for myoblasts and MRC; P = 0.6, n = 6, Fig. 6).
Pax7 + cells of human origin are more numerous after MRC injection than myoblast injection. To identify SC of human origin in grafted mice, muscle sections from animals injected with either myoblasts or MRC, were stained with antibodies against human lamin A/C and the SC marker Pax7 (Fig. 7). This showed that the percentage of human lamin A/C + /Pax7 + cells (indicated by white arrows) in representative muscle sections was significantly increased after MRC injection compared to myoblast injection (23.8 ± 1.9% vs. 10.1 ± 1.7% of lamin A/C + Pax7 + cell respectively (n = 6, Fig. 7). We further observed that injected human myoblasts and human MRC express MyoD at 21 days after injection (Fig. 8a). In vivo localization of these cells was also investigated using an antibody directed against laminin, a basal lamina protein that cover the SC niche. Double positive cells for Pax7 and human Lamin A/C were localized surrounded by laminin suggesting that some human cells adopt a SC position after intramuscular injection (Fig. 8b).

Discussion
Myogenic precursor cell transplantation is one option for the treatment of Duchenne patients and/or for the treatment of skeletal muscle injuries 21,27 . Satellite cells (SC) and their progeny, myoblasts, are essential for muscle regeneration to occur 2 and as such, may represent an "ideal" cell source to restore skeletal muscle integrity. Unfortunately, the in vitro amplification process required to obtain therapeutic doses of myoblasts, reduces their regenerative capabilities 23,28 . Myogenic reserve cells (MRC) have been identified 20 years ago in vitro, as cells able to escape spontaneously terminal differentiation but which remain myogenic if further stimulated 24 . In the present study, we demonstrated that human MRC are similar to quiescent human SC, participate to muscle regeneration with an enhanced potency compared to myoblasts to generate Pax7 + cells in vivo, and apparently fulfill all the criteria expected from a therapeutic tool. SC fate is governed by the expression of various transcription factors including Pax7 and MyoD. Pax7 is expressed in quiescent and activated satellite cells, whereas MyoD is expressed in activated satellite cells and during differentiation 29,30 . In our in vitro model, we generated a heterogeneous population of MRC after differentiation initiation. A majority of these cells (80%) were Pax7 + /MyoD − /Ki67 − cells, similar to quiescent SC. Down-regulation of MyoD, which is known to be required for generation of MRC 26,31 , was confirmed by western blot analysis. A number of properties of MRC described in vitro are similar to those of quiescent muscle SC. Likewise, 80% of human MRC are quiescent mononucleated Pax7 + /MyoD − cells, and proliferate in GM while retaining a high differentiation potential in DM. Two markers, CD82 and CD114 (G-CSF receptor), recently identified in myogenic precursor cell populations 32,33 were also expressed at similar level by human MPC and human MRC. Compared to earlier studies of damaged muscles regeneration by xenogenic myoblasts [34][35][36][37][38] , we obtained in this study a mean survival of human myoblasts 7 days after implantation which was 4 fold superior compared to that of grafted clonal human myoblasts described in another study 38 . This could be explained in part by the use of severe immunodeficient NOD/Shi-scid, IL-2R γ null (NOG) mice 39 , that lack mature lymphocytes (B, T), natural killer cells, have dysfunctional macrophages and dendritic cells, and a reduced complement activity. Such altered dynamics of the early innate response and immune cellular reactions may favor early myoblasts survival post injection 40 . Moreover, the in vitro cell expansion protocol we used which involved polyclonal myoblasts rather than clonal myoblasts decreases the number of cell doublings prior to injection and may favor survival 23,41 . Finally, the state of the muscle at the time of injection may also affect its regeneration 42,43 , such as the laceration of the muscle prior to injection may favor xenogenic cell survival.
An important finding of the present study is the difference in cell cycle status between MRC and myoblasts at the end of the in vitro culture immediately prior to transplantation, and the significant difference in survival/ engraftment in the host between these cells observed 14 and 21 days after transplantation. The increased engraftment seems not to be due to a differential expression of myogenic markers such as CD56, CD82 or CD146 or the myogenic regulatory factor Myf5, since these were expressed at a similar level on both cell types. By contrast MyoD expression may impact on graft take since it has been reported that myogenic cells lacking or with reduced MyoD expression, implant more efficiently than cells expressing high level of MyoD 44,45 . It may thus be expected that myoblasts which express MyoD would exhibit a reduced ability to engraft in vivo 16,23,46 , while MRC whose 80% do not express MyoD by the time of grafting (if undertaken after 48 hours of exposure to DM, see our data) may engraft with a higher efficiency than myoblasts early after transplantation. Nevertheless, as already demonstrated for human myoblasts that were shown to differentiate rapidly after implantation in vivo 47 , we also observed that myoblasts and MRC express MyoD at 21 days after injection.
The difference in cell cycle observed between the 2 cell types at the time of injection may also impact grafting efficiency. Quiescent MRC may adopt a metabolic profile favorable to their survival and increase their autophagy as demonstrated recently for quiescent preconditioned mesenchymal stem cell 48 . As such, initial quiescence may favor myogenic stem cell survival after transplantation in an ischemic environment. This in accordance with a recent study showing that satellite cells maintained in an undifferentiated stage by leukemia inhibitory factor engrafted more efficiently 49 . The mechanisms involved in the regulation of muscle stem cell quiescence are not fully identified yet, but mechanisms involving Sprouty1 50, 51 , autophagy 52 , notch signaling 53 , miRNAs 54 , Kruppel-like factor 7 55 and epigenetic events 56 could favor quiescent MRC emergence.
Another important property of human myogenic cells is their ability to regenerate injured muscle. MRC participated to muscle regeneration, as revealed by the presence of human nuclei in host myofibers positive for human spectrin. The ratio of Spectrin + fibers to total number of Lamin A/C + nuclei was rather low but similar to results obtained after human myoblast transplantation. Since these results could not explain the increased survival observed with MRC as compared to myoblasts, we hypothesized that MRC should better survive as non-fusing cells.
Using a Pax7 antibody, we demonstrated that human MRC gave rise to significantly more progenitor myogenic cells, as compared to human myoblasts. Thus, the fate of human myoblasts versus human MRC is different after intramuscular injections in a xenogenic host. A subpopulation (25%) of the injected quiescent MRC did not form myotubes although expressing Pax7. Some of these double positive Pax7/Lamin A/C cells adopt a satellite cell position. Their ability to respond to future muscle damage was not tested in our model but one may expect some of these cells to be at least in part functional, as observed after injection of myogenic CD133 + cells 16 or human satellite cells 57 . In summary, we have demonstrated that human MRC are quiescent Pax7 + /MyoD − cells, with enhanced survival and enhanced potency to form Pax7 + cells in vivo as compared to that of human myoblasts. Compared to other potential myogenic stem cells 16,19,58 , MRC hold the advantage to be generated in vitro in number compatible with future therapeutic applications. It will be important to examine and compare the expression of genes involved in MRC and myoblasts self-renewal, to identify which pathway is differently activated between these two cell types. Their respective metabolic profiles should also be examined to decipher the possible metabolic shift existing between MRC and myoblasts. The use of artificial niche that maintain stem cell quiescence may also improved the therapeutic potential of MRC 59 .

Methods
Human muscle biopsies. Human muscle biopsies were collected during orthopedic surgery of 22 healthy patients (25.6 ± 1 years old). All methods relating to the human study were performed in accordance with the guidelines and regulations of the Swiss Regulatory Health Authorities and approved by the Comission Cantonale d'Ethique de la Recherche from the Geneva Cantonal Authorities, Switzerland (protocol CER n° 12-259). Informed and written consents were obtained from all subjects involved in the study.  Primary human myoblasts cell culture. Human skeletal muscles were minced and subjected to enzymatic dissociation (Trypsin EDTA 0,05%, Gibco) for 60 min under agitation at 37 °C. The process was stopped with 10% FCS. The cell suspension was then filtered through a 70 μm and a 40 μm nylon filter. Cell suspension was expanded in growth medium (GM) consisting of Ham's F10 (Life Technologies) supplemented with 15% FCS (Life Technologies), bovine serum albumin (Sigma-Aldrich; 0.5 mg/ml), fetuin (Sigma-Aldrich; 0.5 mg/ ml), epidermal growth factor (Life Technologies; 10 ng/ml), dexamethasone (Sigma-Aldrich; 0.39 μg/ml), insulin (Sigma-Aldrich; 0.04 mg/ml), creatine (Sigma-Aldrich; 1 mM), pyruvate (Sigma-Aldrich; 100 μg/ml), uridine (Sigma-Aldrich; 50 μg/ml), gentamycin (Life Technologies; 5 μg/ml) for 5 to 7 days.
For cell cycle analysis, cells were permeabilized by incubation on ice for 20 minutes with Fix/Perm buffer (BD Biosciences). Cells were washed twice in Perm/Wash buffer (BD Biosciences) and incubated with a mouse anti human Ki67 AlexaFluor ® 647 or an isotype control AlexaFluor ® 647 mouse IgG1k for 30 minutes on ice (BD Biosciences). After washing with Perm/Wash buffer, cells were incubated for 15 minutes at 37 °C with 5 μl of Hoechst 33342 (1 mg/ml, Invitrogen). The fluorescence was measured using a LSR Fortessa (BD Biosciences) and data were analyzed using FlowJo 10.2 (FlowJo LLC, USA).

Generation of human myogenic reserve cells. Human myoblasts were cultured in GM until confluence
and then transferred to differentiation medium (DM). DM is a DMEM-based medium (Life Technologies) supplemented with bovine serum albumin (Sigma-Aldrich; 0.5 mg/ml), epidermal growth factor (Life Technologies; 10 ng/ml), insulin (Sigma-Aldrich; 0.01 mg/ml), creatine (Sigma-Aldrich; 1 mM), pyruvate (Sigma-Aldrich; 100 μg/ml), uridine (Sigma-Aldrich; 50 μg/ml), gentamycin (Life Technologies; 10 μg/ml). After 2 days in DM, both human myotubes and non-fused cells, defined as human MRC were observed. For transplantation experiments and flow cytometry analysis, human MRC were specifically isolated using a short trypsinization (30 seconds) that specifically removed all myotubes and left only quiescent undifferentiated cells adherent. To further eliminate small myotubes, trypsinized MRC were filtered using 20 μm pre-separation filters (Miltenyi Biotec, Bergisch Gladbach, Germany) before flow cytometry analysis or cell transplantation experiments. All in vitro and in vivo experiments were carried out with cells that had divided less than 20 folds. Western blot. Total protein extract was obtained by harvesting human myogenic cells (myoblasts, myotubes, MRC) after trypsinization in lysis buffer (50 mM Tris (pH 7.4), 250 mM NaCl, 5 mM EDTA, 50 mM NaF, 1 mM Na3VO4 et 1% Nonidet P40) containing protease inhibitors (Complete Tablet, Roche) and phosphatases inhibitors (PhosSTOP, Roche) on ice for 5 minutes. Cell lysates were centrifugated 5 min at 13'000 rpm and supernatant protein content was dosed using a Bradford assay.
Renilla luciferase lentiviral vector. Virus was produced as described previously 38 . A three-plasmid expression system was used to generate second-generation lentiviral vectors by transient transfection of 293T cells. Infections of human myoblasts (10 5 adherent cells) were performed at a multiplicity of infection (MOI) of 1, in the presence of polybrene (8 μg/ml) in GM. After overnight incubation, cells were washed and incubated in GM for 3 days. Transgene expression was confirmed by bioluminescence imaging before transplantation.
Muscle laceration and intramuscular transplantation. Isoflurane (Abbott, Baar, Switzerland), supplemented with oxygen through a semi-closed circuit inhalation system, was used to anesthetize mice. Two days before cell transplantation, Gastrocnemius muscles were lacerated using a reliable method as previously described 60,61 . Briefly, Gastrocnemius muscles were exposed and cut at 60% of the length from their distal insertion, through 75% of their width and 50% of their thickness. Muscles were then sutured using a polypropylen 6.0 (Ethicon, Somerville, NJ, USA). Thereafter, either human myoblasts or human MRC (5 × 10 5 cells in 15 μl PBS) were injected into one location of the injured muscle, using a 25 μl syringe with a 29-gauge needle (Hamilton Company, Bonaduz, Switzerland). Mice received intraperitoneal injection of Temgesic (0.05 mg/kg, Essex Chemie, Luzern, Switzerland) for 3 days after muscle injury.
Bioluminescence analysis. Imaging was performed using an IVIS 200 optical imaging system (PerkinElmer, Schwerzenbach, Switzerland) and data were processed using Living Image ® Software (PerkinElmer, Schwerzenbach, Switzerland).
For live cell culture analysis, bioluminescence imaging (BLI) was performed immediately upon addition of the substrate (coelenterazine native; 1 μg/ml; Biotium, Hayward, CA, USA), diluted in cold PBS ++ (containing Ca 2+ and Mg 2+ ). Acquisition time for in vitro experiments were 1 second. For in vivo experiments, bioluminescence was monitored immediately after Coelenterazine i.v injection in mice through the retro-orbital route (1 mg/kg diluted in 100 μl PBS ++ ) and continuously for 12 minutes (4 consecutives acquisition of 3 min) A region of interest (ROI) was manually selected over the signal intensity. The area of the ROI was kept constant and signal recorded as maximum [photon/sec]. For technical reasons, the 3h-point post injection was defined as the reference 100% survival (day 0). BLI values measured at day 4, 7, 14 and 21 were related to the bioluminescence data obtained at day 0.
Differentiation state was quantified by assessing nuclear expression of the myogenic transcription factor MEF2 and by assessing expression of the myosin heavy chain protein. For this purpose, human cells were incubated with mouse anti-myosin heavy chain antibody (1:1000; DSHB, Iowa, USA) and with rabbit anti-MEF2 antibody (1:300, Santa Cruz Biotechnology Inc., Heidelberg, Germany). Fusion indexes were defined as the ratio of MEF2 positive nuclei count inside MF20 positive myotubes and divided by the number of total DAPI positive nuclei.