INTRODUCTION

Mesenchymal stromal cells (MSCs) as universal regulators of tissue homeostasis are a promising material for regenerative medicine, in particular for the restoration of skeletal muscles after injury. Although skeletal muscle recovery is mainly provided by satellite cells (a resting tissue-specific population of cells), MSCs, both resident and those leaving the bone marrow and entering into the bloodstream, followed by migration to the damaged tissue under the influence of chemoattractants, also participate in the regeneration (Ramírez et al., 2006; Granero-Moltó et al., 2009; Liu et al., 2009; Chen et al., 2010; Fong et al., 2011). Experimental data obtained in recent years indicate that stimulation of muscle regeneration can be achieved by introducing MSCs into the area of injury (Natsu et al., 2004; Winkler et al., 2012; Andrade et  al., 2015).

Three possible mechanisms of MSC participation in skeletal muscle regeneration are currently under consideration: differentiation in the myogenic direction, fusion with myoblasts, and a paracrine effect on myogenesis. When cultured in the presence of certain inducers, in particular 5-azacytidine (Krupnick et al., 2004), steroid hormones (Gang et al., 2004), and galectin-1 (Chan et al., 2006), MSCs can acquire signs of differentiation into skeletal muscle, expressing the specific markers MyoD (Abcam, United Kingdom) and myogenin and merging into multinucleated myotubes containing desmin and myosin. However, the capability of MSCs for myogenic differentiation in response to various inducers is not confirmed by all authors and, apparently, is not the same for cells from different organs and at various stages of ontogenesis (Liu et al., 2003; Gang et al., 2004; Balana et al., 2006; Chan et al., 2006). Studies of the possible involvement of MSCs in regeneration by fusion with myogenic cells are also contradictory (Shi et al., 2004; Sassoli et al., 2012; Kulesza et al., 2016). At the same time, the positive effect of MSCs on the restoration of muscle tissue may be due to their paracrine function, which is the production of many biologically active factors with a wide spectrum of action (Sassoli et al., 2012).

The aim of this work is to study the possibility of direct differentiation of MSCs into myogenic cells, as well as their paracrine effect on myogenesis.

MATERIALS AND METHODS

We used adult male and female Wistar rats weighing 200–250 g, obtained from the Stolbovaya laboratory animal nursery (Moscow oblast), as well as 20‑day-old fetuses and 9- to 10-day-old pups.

Isolation and cultivation of MSCs and fetal fibroblasts. MSCs were obtained from the bone marrow of mature rats, which was isolated from the femur and tibia, washing the diaphysis with the α‑MEM medium (HyClone, United States). Fetal fibroblasts were isolated from the skin of 20-day-old fetuses by treating tissue fragments with a 0.075% type I collagenase solution (Sigma, United States) for 30 min at 37°С. Bone marrow and fetal skin cells were cultured in 75 cm2 plastic flasks (Greiner, Germany) at 37°С and 5% CO2 in the α‑MEM medium with 10% fetal calf serum (FCS) supplemented with L-glutamine, an antibiotic–antimycotic, penicillin, and streptomycin (HyClone). Passaging was carried out upon reaching a confluent monolayer, removing cells with a 0.25% trypsin solution with ethylene diamine tetraacetate (HyClone) and replanting them in a ratio of 1 : 2.

Induction of myogenic differentiation of MSCs. The differentiation of bone marrow MSCs was induced at the second passage. Due to the lack of a generally accepted protocol for the induction of myogenesis and the ambiguity of the literature data on the ability of MSCs to differentiate in this direction (Liu et al., 2003; Gang et al., 2004; Balana et al., 2006; Chan et al., 2006), the effectiveness of three different induction variants was evaluated. Cells were scattered in 12-well plates (Greiner) with a density of 1 × 104 cells/mL in the induction media (three variants).

Variant I. Medium: Dulbecco modified Eagle medium (DMEM) + 20% FCS + 10 μM 5-azacytidine (Sigma); control: DMEM + 20% FCS. After two days, the medium was replaced with DMEM with 10% horse serum (HyClone), after which cultivation was continued with a periodic change of this new medium.

Variant II. DMEM + 2% FCS + 3 μM 5-azacytidine + 10 ng/mL epidermal growth factor (EGF) + 10 ng/mL basic fibroblast growth factor (bFGF) + 1% ITS supplement + 1 + 50 μg/mL 2-phospho-L-ascorbic acid trisodium salt (Sigma); control: a medium of the same composition without 5-azacytidine. After four days in both the experimental and control cultures, the medium was replaced with the same fluid without 5-azacytidine and cultivation was continued with a periodic change of this new medium.

Variant III. DMEM + 10% FCS + 0.1 μM dexamethasone + 50 μM hydrocortisone (Sigma); control: DMEM with 10% FCS. After four days in the experimental culture, the medium was replaced by DMEM + 5% horse serum + 0.1 μM dexamethasone + 50 μM hydrocortisone, and in the control culture, DMEM with 5% horse serum. Further cultivation was carried out with a periodic change of these new media.

In all cases, cultivation was continued for 14–28 days with a change of the medium twice a week, after which the cultures were fixed for morphological and immunocytochemical studies.

Transfection of MSCs with the green fluorescent protein (GFP) gene. MSCs were labeled by adenovirus infection in the second passage after the formation of a cell monolayer. Gene-bearing viral particles GFP were designed at the laboratory of molecular immunology of the Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sceinces, based on the Ad.MAX system (SignaGen Laboratories, United States), which includes the shuttle vector DC311, carrying the gene GFP, and the genomic plasmid pBHGloxΔE13Cre. The 293 cell culture (human embryonic kidney epithelial cells) was infected twice with engineered viral particles to increase their number, after which they were frozen and thawed to lyse the cells. Viral particles were separated from the cell debris by centrifugation. After removal of the medium, 0.5 mL of a suspension of the virus and 1.5 mL of α‑MEM with 10% FCS were added to the MSC culture vials; after two hours, a fresh culture medium was added. The medium was changed after 1–3 days. The efficiency of transfection was evaluated by intravital observation of GFP fluorescence in an inverted microscope. In the experiments, cultures containing at least 50% fluorescent cells were used 4–5 days after transfection.

Co-cultivation of MSCs with muscle cells. Cells from the muscles of 9- to 10-day-old rat pups were isolated by treatment with a 0.2% type I collagenase solution for 30 min at 37°C. Freshly isolated muscle cells were co-cultured with bone marrow MSCs in 12-well plates in DMEM medium with 10% FCS. The following three variants were used for co-cultivation.

(1) To assess the capability of MSCs for myogenic differentiation under the influence of factors produced by myoblasts, MSCs were placed at the bottom of the plate after the first passage (1 × 104 cells per well), and muscle cells (1 × 105 cells per well) were put on the surface of membrane inserts with a pore size of 0.4 μm (Becton Dickinson, United States). MSCs cultured in the absence of muscle cells served as a control. The duration of cultivation was 21 days.

(2) To assess the paracrine effect of MSCs on myogenesis for in vitro muscle cells, muscle cells were placed on the bottom of the plate, precoated with fibronectin from human blood plasma (Payushina et al., 2010) (1 × 105 cells per well), on the surface of the membrane insert were MSCs, passing one passage (1 × 104 cells per well). Muscle cells cultured on a fibronectin coating in the absence of MSCs served as the control. The duration of cultivation was 10 days.

(3) To assess the capability of MSCs to fuse with myoblasts, 1 × 104 GFP-labeled bone marrow cells were placed in each well of a fibronectin-coated plate with 1 × 105 muscle cells and cultured for 13 days.

Subcutaneous muscle tissue transplantation. To study the effect of MSCs on myogenesis in vivo, we used an experimental model of ectopic transplantation of muscles developed by Studitskii (Studitskii, 1977). Muscle tissue of the hind limbs of a rat was crushed and mixed with a suspension of MSCs of the second passage (labeled with GFP or unlabeled). Two subcutaneous pockets were formed on the back of adult rats under chloral hydrate anesthesia. Muscle tissue mixed with MSCs was placed in one of them (7 × 105–1.6 × 106 cells for each recipient); in the other only muscle tissue was added. Stitches were applied to the skin incision. After five days, the animals were removed from the experiment. Five animals with the introduction of MSCs labeled with GFP and five animals with the administration of unlabeled MSCs were analyzed.

The introduction of cells and conditioned media into the area of muscle injury. Mechanical damage was done to mature rats under chloral hydrate anesthesia by a deep transverse incision of the cranial tibial muscle, followed by suturing the skin wound. To assess the regenerative potential of MSCs at different periods of the wound healing process, one group of animals was injected into the lesion area with 1 × 106 MSCs of the second passage in 100 μL of the DMEM medium immediately after surgery, and other groups were injected four days later. The control was the introduction of 100 μL of DMEM medium without cells or 1 × 106 fetal skin fibroblasts of the second passage in 100 μL of DMEM. Similarly, 100 μL of the medium conditioned by MSCs of the second passage were injected into the muscles immediately after their injury. As a control, the same volume of DMEM was injected into the muscles with 10% FCS or a medium conditioned with fetal skin fibroblasts of the second passage. Some experimental animals received four repeated injections of 100 μL of conditioned by MSCs or the control medium 3, 5, 7, and 10 days after the injury. In a number of experiments, the MSC-conditioned medium and the control medium (DMEM with 10% FCS) were concentrated by lyophilization before being introduced into the muscles, followed by dilution at 1/20 of the original volume. Animals were removed from the experiment 5 or 14 days after the injury. Each experimental group included five rats.

Analysis of the results. For morphological studies, cell cultures were fixed with 96% ethanol and stained with azure-eosin. Immunocytochemical staining was carried out on cultures fixed with 10% formalin, an indirect method using monoclonal antibodies to the muscle marker MyoD (Abcam) and second antibodies with fluorochrome Alexa Fluor 488 (Invitrogen, United States). Cell nuclei were stained with Hoechst 33342 (Sigma). The regenerating muscles and subcutaneous grafts were fixed with 10% formalin after 30 minutes in physiological saline, paraffin sections were prepared and stained with hematoxylin-eosin or the Mallory method according to the manufacturer’s protocol (Bio-Optica, Italy). The area of the regions of inflammation and fibrosis was measured on micrographs of the transverse sections of the studied muscle using the ImageJ program and expressed as a percentage of the total section area. The effectiveness of angiogenesis was assessed by counting blood vessels per 1 mm2 cutting area. Three sections from each damaged muscle were analyzed, and the average values of the estimated parameters were determined. Quantitative data were processed statistically using Strelkov tables (Strelkov, 1999); the significance of differences was evaluated using the Mann–Whitney test.

RESULTS AND DISCUSSION

The myogenic potencies of MSCs and their ability to fuse with myoblasts in vitro. During the cultivation of MSCs in the presence of substances known as inducers of myogenesis (dexamethasone and 5-azacytidine), the myogenic marker MyoD was not detected in them and the formation of myotubes did not occur (Figs. 1a, 1b). In some cases, differentiation of MSCs in the adipogenic direction was observed in the medium with 5-azacytidine (Fig. 1c). Our data are consistent with the results of other authors who failed to cause myogenic differentiation of MSCs with 5-azacytidine (Liu et al., 2003; Martin-Rendon et al., 2008). It is known that factors secreted into the environment by myoblasts can act as inducers of myogenesis (Salvatori et al., 1995; Wise et al., 1996), and in some cases, they cause myogenic differentiation of MSCs more efficiently than 5-azacytidine (Chan et al., 2006). However, we did not find morphological or immunocytochemical signs of differentiation of MSCs that are co-cultivated with muscle cells through a semipermeable membrane (Fig. 1d). The results obtained indicate the absence of expressed potentials for myogenesis in rat bone marrow MSCs.

Fig. 1.
figure 1

MSCs do not differentiate in vitro in the myogenic direction. The cultivation of MSCs in the presence of (a) 0.1 μM dexamethasone, (b) 10 μM 5-azacytidine, (c) 3 μM 5-azacytidine (dashed lines around the foci of adipogenesis), and (d) cells from muscles in the indirect co-cultivation system. (a, b, d) staining with hematoxylin-eosin, (c) staining with oil red O. The nuclei are stained with hematoxylin. Magnification: ob. ×20, eyepiece ×10.

Full size image

With direct co-cultivation of GFP-labeled MSCs with cells isolated from muscles, only in one case did we observe the formation of a labeled myotube as a result of the fusion of MSCs with myoblasts (Fig. 2). The absence or paucity of cases of the merger of MSCs from differentiating myogenic cells in vitro or in vivo have been reported by other authors (Natsu et al., 2004; Motohashi et al., 2008; Sassoli et al., 2012), although it was also reported that, when co-culturing GFP-labeled MSCs with C2C12 myoblasts in the presence of myogenesis inducers, the frequency of formation of hybrid myotubes was 38 cases per 1000 cells (Shi et al., 2004).

Fig. 2.
figure 2

Co-cultivation of GFP-labeled MSCs and myoblasts. (a) Culture of MSCs labeled with GFP; (b–d) joint culture of myoblasts and labeled MSCs (hybrid myotube is marked with an arrow). (a, c) Fluorescence microscopy, (b) phase contrast, (d) combination. Magnification: (a) ob. ×20, eyepiece ×10; (b–d) ob. ×10, eyepiece ×10.

Full size image

It can be assumed that the low frequency of transdifferentiation of MSCs or their fusion with recipient cells is characteristic of intact tissue. When skeletal muscle is damaged or when radiation causes the death of satellite cells, MSCs can occupy the niche of the latter and enter the path of myogenic differentiation (LaBarge and Blau, 2002; Dreyfus et al., 2004). There is evidence of the fusion of MSCs with myogenic cells in muscles that regenerate after chemical damage (Shi et al., 2004; de la Garza-Rodea et al., 2012). Thus, our data do not completely exclude the possibility of MSCs participating in the regeneration of muscle tissue by differentiation or fusion.

Regulatory effect of MSCs on in vitro myogenesis. In the indirect co-cultivation system, when MSCs were placed on the surface of the inserts and muscle cells turned out to be effector cells, activation of myogenesis was observed, which was manifested in an increase in the number of myotubes, as well as the ratio of the average number of nuclei in their composition to the total number of nuclei in the field of view (Figs. 3a–3c). A multiple increase in the number of forming myotubes was also noted during the cultivation of cells from muscle tissue in a medium conditioned by MSCs (Fig. 3d). Since muscle cells did not come into contact with MSCs in these experiments, it is obvious that the stimulation of myotube formation was due to secreted MSC factors, i.e., through the paracrine effect of MSCs on the differentiation and fusion of muscle cells.

Fig. 3.
figure 3

The effect of MSC secretory products on myoblast differentiation. (a) Muscle cell culture in the absence of MSCs; (b) muscle cells during indirect co-cultivation with MSCs. Myotubes are shown by arrows. Hematoxylin-eosin stain; magnification: ob. ×10, eyepiece ×10; (c) proportion of myotube nuclei in the total number of nuclei during indirect co-cultivation of MSCs with muscle cells (control, cultivation without MSC; experiment, co-cultivation with MSC); (d) number of myotubes per 1 cm2 during cultivation of muscle cells in a medium conditioned by MSCs (control, DMEM medium; experiment, conditioned medium).

Full size image

The effect of MSCs on in vivo myogenesis during subcutaneous muscle tissue transplantation. The effect of MSCs on the early stages of myogenesis was investigated on a model of subcutaneous administration of cells from damaged muscles in conjunction with a suspension of MSCs. During transplantation of ground muscle tissue and MSC labeled with GFP, five days after surgery, fluorescent cells were found in the transplant, located mainly near fragments of muscle fibers or adjacent to their surface, but not merging with myoblasts (Figs. 4a, 4b). The addition of MSCs to the transplanted tissue led to the activation of myogenesis, which was expressed in a larger number and size of newly formed myotubes in comparison with control samples (transplantation of crushed muscles without MSC injection) (Figs. 4c, 4d) and in their earlier formation (by two days) than in the control. The process of angiogenesis also intensified: the number of blood vessels with the introduction of MSCs and in the control was 6.3 ± 1.1 and 2.1 ± 0.2 per 1 mm2, respectively (differences are significant at p < 0.01). Thus, the addition of MSCs to transplanted myogenic cells stimulated the early stages of the recovery process.

Fig. 4.
figure 4

Transplantation of crushed muscles under skin of rats five days after surgery. (a) Labeled GFP MSCs in the graft; (b) the same field of view, cell nuclei stained with Hoechst 33342; (c) control (transplantation of crushed muscles); (d) experiment (transplantation of crushed muscles with the addition of MSCs). Myotubes are shown by arrows. (a, b) Fluorescence microscopy, (c, d) staining by the Mallory method. Magnification: (a, b) ob. ×20, eyepiece ×10; (c, d) ob. ×10, eyepiece ×10.

Full size image

The effect of MSCs on the regeneration of mechanically damaged muscles. To assess the effect of MSCs on skeletal muscle regeneration, we also used a model of mechanical damage to muscles through a deep incision, which allows us to analyze the state of the tissue not only in the early stages after damage (five days), but also later (14 days). MSCs were injected into the area of injury; the control was the introduction of fibroblasts of fetal skin or the culture medium. A histological analysis of the state of the muscles five days after the damage showed that MSCs introduced immediately after the injury did not affect the early stage of regeneration: in the control and in the experimental samples, extensive zones of muscle tissue destruction and severe inflammation were observed (Figs. 5a, 5b).

Fig. 5.
figure 5

Damaged muscles after the introduction of MSCs. (a) Control (DMEM), five days after injury; (b) MSCs, five days after injury; (c) control (DMEM), 14 days after injury; (d) MSC, 14 days after injury; (e) delayed (after four days) introduction of DMEM, 14 days after injury; (f) delayed (four days) introduction of MSCs, 14 days after injury. Fibrous tissue is indicated by arrows. Hematoxylin-eosin stain. Magnification: (a, b, d, e, f) ob. ×10, eyepiece ×10; (c) ob. ×20, eyepiece ×10.

Full size image

Fourteen days after the damage, regeneration both in the control (Fig. 5c) and in the experiment (Fig. 5d) was not yet completed. Moreover, if a connective tissue scar was formed in the muscle of the culture medium or skin fibroblasts at the site of damage in 80–100% of cases, then with the introduction of MSC, the frequency of its formation was ~50%. At the same time, the area of fibrous tissue after the introduction of MSC was comparable with that in the control (Table 1). Other authors who reported a significant improvement in the functional ability of damaged muscles after the introduction of MSCs, associated with an increase in the number and maturity of muscle fibers, did not find a decrease in signs of fibrogenesis (Natsu et al., 2004; Andrade et al., 2015). The introduction of MSCs in our experiments did not lead to a significant decrease in the area of the inflammation zone in comparison with the control (medium injection), although, as is known, MSCs are able to suppress proinflammatory M1 macrophages and activate anti-inflammatory M2 macrophages (Le Blanc and Mougiakakos, 2012; Abumaree et al., 2013). With the introduction of skin fibroblasts, the area of inflammation increased compared with the control muscles into which the medium was introduced, and the regeneration was delayed at the macrophage stage. It is possible that this was due to the secretion by fibroblasts of the pro-inflammatory cytokines interleukin-6 and interleukin-8 (Vistejnova et al., 2014). It should be noted that with the introduction of fibroblasts, most muscle fibers remained immature, which indicates a slowdown in their formation. The introduction of MSCs, on the contrary, facilitated the formation of more mature muscle fibers, and increased the number of vessels in the regenerating tissue (Fig. 6a).

Table 1.   The effect of MSCs on muscle regeneration after mechanical injury
Full size table
Fig. 6.
figure 6

Average number of vessels per 1 mm2 area of damaged muscle 14 days after introduction of (a) cells and (b) the conditioned medium. CMcont, DMEM with 10% FCS; CMexp, medium conditioned by MSC; CMlyoph. cont and CMlyoph. exp, respectively, control and conditioned MSC medium concentrated by lyophilization. Significance of differences: * p < 0.01, ** p < 0.05.

Full size image

In the case when MSCs were injected into the muscle four days after the damage, we observed a decrease in the area of fibrous tissue and the area of inflammation compared with the control (introduction of the medium) (Figs. 5e, 5f), as the injection of skin fibroblasts had no effect.

A single injection of the medium conditioned by MSC did not have a significant effect on muscle regeneration, like the introduction of the medium conditioned by skin fibroblasts. However, with its repeated administration, the area of the regions of inflammation decreased by approximately three times in comparison with the control (Table 2). Preliminary experiments on the introduction of a conditioned MSC medium, the concentration of which was increased by lyophilization, showed that it leads to more active vessel formation compared to the control (Fig. 6b). The results obtained indicate that the anti-inflammatory and angiogenic effects of MSCs during regeneration are associated with the regulatory molecules produced by them; in particular, stimulation of blood vessel formation may be due to the presence of vascular endothelial growth factor (VEGF) and angiopoietin-1 (Wu et al., 2007).

Table 2. The effect of MSC conditioned media on muscle regeneration after mechanical injury
Full size table

CONCLUSIONS

Judging by the results of these experiments, MSCs from rat bone marrow apparently do not have the potential for myogenic differentiation under the influence of inducers and have a weak ability to fuse with myoblasts. However, they stimulate the differentiation of muscle tissue cells through paracrine influence. When MSCs are introduced into the damaged muscle, they positively affect the course of the recovery process. They contribute to a decrease in the frequency of fibrosis (and sometimes its area), the formation of blood vessels, and the formation of mature muscle fibers and, in some cases, reduce the area of the inflammation zone. Obviously, the beneficial effect of MSCs on the regeneration of muscle tissue is associated with their paracrine function, since repeated administration of a medium conditioned by them or its introduction after lyophilization, which increases the concentration of the contents, has a positive effect.