Characterisation of stemness and multipotency of ovine muscle‐derived stem cells from various muscle sources

Abstract Muscle stem cells (MSCs) are a promising tool for cell‐based therapy and tissue regeneration in veterinary medicine. Evaluation of MSCs from muscles of different origins improves our understanding of their regenerative potential. The present study compared the stemness, cell proliferation, migration potential, myogenic differentiation (MD), and multipotency of MSCs for four developmentally different muscles of ovine origin. MSCs were isolated from the hind limb (HL), diaphragm (DI), extraocular (EO), and masseter (MS) muscles. Cell proliferation, migration, and stemness were examined using sulforhodamine B, and colony formation assays. Evaluation of multipotency was examined using histological and morphometric analyses, alkaline phosphatase (ALP) activity, and the expression of myogenic, adipogenic, and osteogenic markers using RT‐qPCR. Data were statistically analysed using analysis of variance. The results revealed that all experimental groups expressed stem cell markers paired box transcription factor Pax7, α7‐integrin, CD90, and platelet‐derived growth factor receptor alpha. DI and HL muscle cells displayed higher proliferation, migration, and colony formation capacities compared to the EO and MS muscle cells. HL and DI muscle cells showed increased MD, as indicated by myotube formation and relative expression of MyoD at day 7 and Myogenin at day 14. Although MS and EO muscle cells displayed impaired MD, these cells were more prone to adipogenic differentiation, as indicated by Oil Red O staining and upregulated fatty acid‐binding protein 4 and peroxisome proliferator‐activated receptor gamma expression. DI muscle cells demonstrated a higher osteogenic differentiation capability, as shown by the upregulation of osteopontin expression and an elevated ALP activity. Our data indicate that ovine HL and DI MSCs have a higher regenerative and multipotent potential than the EO and MS muscle cells. These results could be valuable for regional muscle biopsies and cell‐based therapies.


| INTRODUC TI ON
Skeletal muscle stem cells (MSCs) are a valuable tool for investigating myogenic differentiation (MD) and muscle regeneration. It is well known that stem cells and muscles develop concomitantly. During embryonic development, a group of cells express paired box transcription factor 3 (Pax3), delaminate from the dermomyotome, and migrate to the mesenchyme of the limb bud to differentiate into limb muscles (He et al., 2003;Zammit et al., 2004). Although the limb bud cells expressing Pax3/Pax7 undergo proliferation and differentiation, another group of cells withdraws from the cell cycle to become MSCs (Zammit et al., 2004). In 1961, adult MSCs were identified between the basal lamina and sarcolemma of muscle fibres (Mauro, 1961). Adult quiescent MSCs express the paired box transcription factor Pax7 (Shioi et al., 1995), which is essential for their specification and survival (Kuang et al., 2006). In contrast, Pax3 is expressed only in a few muscle groups, including the diaphragm (DI) (Redshaw & Loughna, 2012). Unlike the limb and trunk muscles, head muscles, including the masseter (MS) and extraocular (EO) muscles, develop from the cranial paraxial mesoderm (Harel et al., 2009). In addition, the progenitor cells of the head muscles, excluding the tongue, do not express Pax3 (Zammit et al., 2004).
Adult skeletal muscles show remarkable regenerative potential under physiological (Parise et al., 2008) and pathological conditions (Lepper et al., 2011). Upon muscle injury, MSCs undergo both asymmetric and symmetric cell division (Kuang et al., 2007). In asymmetric cell division, Pax7 + /myogenic regulatory factor 5 (Myf5 -) parental cells give rise to two dissimilar daughter stem cells: one is a Pax7 + / Myf5cell, whereas the other is a Pax7 + /Myf5 + cell committed to MD. Proliferating MSCs express MyoD and are referred to as myoblasts. MD is accompanied by downregulation of Pax7 and upregulation of Myogenin expression, which, together with MyoD, is able to activate muscle-specific contractile protein production (Zammit et al., 2004(Zammit et al., , 2006. MSC multipotency has been previously reported in cells with intact muscle fibres that give rise to adipocytes and osteocytes (Asakura et al., 2001;Elashry et al., 2017). MSCs differ in their multipotency. In this context, it has been reported that CD56 + /CD34cells isolated from porcine skeletal muscle display myogenic potency, whereas CD56 + /CD34 + cells show both adipogenic differentiation (AD) and MD capacities (Perruchot et al., 2013). Similarly, CD56 + / CD44 + /CD45 -MSCs were able to differentiate into myotubes, adipocytes, and osteocytes in human beings (Coppi et al., 2006). Furthermore, studies have also revealed that only platelet-derived growth factor receptor alpha (PDGFRα)-positive cells showed efficient AD in vitro and in vivo (Oishi et al., 2013;Watt et al., 1987).
The limited use of MSCs in regenerative medicine, owing to their poor survival rate and cell population heterogeneity, necessitates the search for alternative sources. Although several studies have characterised MSCs isolated from the paraxial mesoderm, less attention has been given to MSCs derived from the head mesoderm.
Ovine MSCs are very similar to human beings MSCs in terms of their life span, metabolism, proliferation, and differentiation capacity (Yan et al., 2013). Moreover, ovine skeletal muscle contains similar myosin heavy chain isoforms as human beings, including type I, IIA, and IIX (Andruchov et al., 2004;Hemmings et al., 2009;Lefaucheur et al., 1998;Wieczorek et al., 1985). Although it has been reported that MSCs of ovine limb muscles can undergo MD, AD, and osteogenic differentiation (OD) (Yan et al., 2013), ovine MSC populations derived from developmentally different muscle groups have not been fully elucidated. Thus, the present study characterised the stemness and multipotency of MSCs isolated from developmentally different sources of ovine skeletal muscles. We hypothesised that MSCs derived from the paraxial mesoderm would display greater regenerative potency and multipotent properties than those derived from the head mesoderm, including the MS and EO muscles. To that end, we examined MSC stemness, cell viability, cell proliferation, and migration capability using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetraz olium bromide (MTT), Sulforhodamine B (SRB), wound healing, and colony-forming unit ability (CFU) assays. MD of various cell populations, as well as their multipotency regarding the adipogenic and osteogenic fate, were examined using immunohistochemistry, morphometric measurements, and RT-qPCR. The data demonstrate the effect of muscle development on the properties of their own stem cells. We show that MSCs of hind limb (HL) and DI muscles have enhanced proliferation, colony formation, and migration potential in comparison to MS and EO muscle cells. We report that cells of the trunk mesoderm, including HL and DI, demonstrate enhanced myogenic potential compared to the cells of the head mesoderm.
Although MS cells have the properties of impaired stem cells, these cells display a higher adipogenic potency. These results are important considering regional muscle biopsies for cell-based therapy and tissue engineering.

| MSC isolation and expansion
MSCs were obtained from 4-week-old male lambs (n = 6). MSCs were isolated from the gastrocnemius of the HL and DI, MS, and EO muscles. Muscle tissue was obtained from the Institute of Veterinary Pathology, Justus-Liebig University, Giessen, following standard and ethical regulations. Briefly, 1 mm 2 muscle pieces were transferred in cold phosphate-buffered saline (PBS, Gibco) with 1% penicillin/streptomycin (P/S, AppliChem). On a clean bench, fat, connective tissue, and blood vessel fragments were removed, and then, the muscle tissue was cut into small fragments using a sterile scalpel blade, placed into 50 ml falcon tubes with PBS, and centrifuged at 670 g for 5 min. The supernatant was discarded, and the pellets were digested with 0.2% collagenase IV diluted in 4.5 g/L Dulbecco's Modified Eagle Medium (DMEM, Gibco) on a shaker at 37°C for 40 min. Muscle samples were triturated mechanically using an 18-gauge needle to obtain single-cell suspension. An equal volume of 2% foetal calf serum (FCS, Biocell) in DMEM was added, and the mixtures were filtered through a 70μm cell strainer, followed by centrifugation at 240 g for 5 min.
Cell pellets were mixed with 1 ml of growth medium consisting of DMEM, 10% FCS, and 1% P/S. To select stem cell populations from other cell types, we used several validation criteria, including pre-plating, PCR, immunohistochemistry, and flow cytometry, to track stem cell and myogenic marker expression. Briefly, the cells were pre-plated in 20% FCS in Ham's F10 medium (Gibco) with 1% P/S in plastic culture dishes for 48 h. This method not only preserves the myogenic population in favour of other cell types, but also removes the fast adherent cells, particularly fibroblasts, as previously reported (Chirieleison et al., 2012;Elashry et al., 2017).
Non-adherent cells were cultivated in a fresh medium for expansion. In parallel, cells from each group were examined for stem cell markers, including CD90, MyoD, Myogenin, and PDGFRα, using PCR, as well as immunohistochemistry for Pax7, Myogenin, CD44, CD90, and α7-integrin. Cell populations showing less than 80% positivity for stem cell markers were discarded.

| Flow cytometry analysis
Briefly, 1 × 10 6 cells/ml of fresh medium was used. A volume of 100 μl of cell suspension per well was transferred to a 96-round-bottomed-well plate. After centrifugation at 400 g for 3 min at room temperature (RT, 15-25°C), cell pellets were washed in a buffer, comprising 1% bovine serum albumin, 0.01% NaN 3 , 0.5% goat serum, and 10% horse serum in PBS, and centrifuged at 400 g for 3 min at RT. The pellets were incubated with 50 μl of the primary antibodies, including mouse CD44 (1:100 antibody-online ABIN94121), CD45 (1:100, antibody-online ABIN319753), and major histocompatibility class II (MHCII, 1:200, Bio-Rad), for 20 min at RT and then centrifuged at 400 g for 3 min. The supernatant was discarded, and the pellets were washed twice in the washing buffer for 3 min, followed by incubation with 50 μl of anti-mouse-PE (1:200, BD Biosciences) secondary antibody for 20 min in the dark. After two washing steps, the pellets were suspended in PBS for flow cytometry analysis using an Accuri C6 ® (BD Biosciences) equipped with Accuri C6 software (BD Biosciences).
Selected cell populations were expanded in the growth medium in T-75 culture flasks at 37°C and 5% CO 2 . The cells were detached using trypsin (TrypLE, Gibco) at 37°C and 5% CO 2 for 6-7 min, and then centrifuged at 240 g for 5 min. Cell pellets were resuspended in fresh medium and counted. Subsequently, the cells were either passaged or cryopreserved for further experiments. For cryopreservation, 1 × 10 6 cells/ml were resuspended in freezing medium, containing DMEM, 1% P/S, 30% FCS, and 5% dimethyl sulfoxide (DMSO, Roth), and stored in cryotubes at −160°C for further experiments. Cells from passages P1-P4 were used for all experiments.

| MTT assay
The MTT assay (Sigma-Aldrich) was used to assess the degree of tetrazolium reduction by nicotinamide adenine dinucleotide phosphate (NADPH)-dependent cellular oxidoreductase, indicating cell viability. Cells of the HL, DI, MS, and EO muscles were seeded into 24-well plates at a seeding density of 5 × 10 3 cells/well for 3, 5, and 7 days. Briefly, the medium was replaced with 300 μl of growth medium, having 5 mg/ml MTT diluted in PBS, per well. The cells were incubated for 3 h at 37°C and 5% CO 2 . The solution was removed, and the cells from each muscle group was incubated with 200 μl of DMSO per well in triplicate. The plates were incubated in a shaker for 10 min at RT. Samples (200 μl) for each condition were pipetted in triplicate into a 96-well plate. Absorbance was measured at 570 nm using a microplate reader equipped with Magellan TM Data Analysis Software (Tecan).

| SRB assay
The SRB assay measures the amount of cellular protein indicative of cell number, as previously reported by Vichai and Kirtikara (2006).
MSCs were seeded in 24-well plates at a density of 5 × 10 3 cells/well in growth medium for 3, 5, and 7 days. Briefly, following the cultivation time points, cells were fixed in 4% paraformaldehyde (PFA, Roth) for 10 min. After three washes in PBS for 5 min each, the cells were incubated with 500 μl of SRB solution, containing 0.4% SRB sodium salt (Sigma-Aldrich) diluted in 1% acetic acid (Merck) for 10 min at RT.
A volume of 200 μl from the mixture was pipetted in triplicate per muscle for each animal (n = 6) into 96-well plates. Absorbance was measured at 565 nm using a microplate reader (Tecan).

| Wound-healing assay
Sterile silicon inserts were tightly placed in the middle of each well (1 insert/well) in a 24-well plate. MSCs of the HL, DI, MS, and EO muscles were seeded in growth medium at a density of 5 × 10 3 cells/well in triplicate. MSCs were cultivated at 37°C and 5% CO 2 until they reached 80% confluence. After 24 h, the inserts were carefully removed and fresh medium was added. The cells were examined under an inverted microscope (Axioobserver, Zeiss), and images were captured for up to 48 h. At the removal of culture inserts, the time was recorded as zero (0). Images were taken at 10× magnification using a Leica MC170 microscope (Leica Microsystems) equipped with Leica Application suite (LASV4.4) imaging software (Switzerland). MSC migration to close the intercellular gap was measured after 6, 12, 24, and 48 h. Image analysis and measurements of gap closure were performed using Adobe Photoshop CS6.

| CFU assay
MSCs were seeded at a density of 1 × 10 2 , 5 × 10 2 , and 1 × 10 3 cells/25 cm 2 in culture flasks with growth medium for 7 days. The medium was changed every 2 days. The cells were washed in PBS and fixed in 4% PFA for 10 min. Crystal violet (5 mg/ml, Roth), diluted in 2% ethanol, was added to each flask for 8 min. The extra stain was washed with distilled water three times for 3 min each.
The flasks were left to dry overnight at RT. Cell colonies stained in violet were identified and photographed using a Leica Microscope.
The total number of colonies were counted per flask at 4× objective. Then, each flask was divided into four equal squares, and the colonies inside each square were counted. The cell colonies were categorised into three groups: small (10-20 cells), medium (21-50 cells), and large (>50 cells).

| Induction of MD
The cells were seeded into 24-well plates at a density of 1 × 10 4 cells/well in growth medium. Upon reaching 70% confluency, the growth medium was switched to MD medium containing DMEM, 2% horse serum (Biocell), 20 ng/ml fibroblast growth factor (Thermo Fisher Scientific), and 1% P/S for up to 14 days.

| Induction of AD and OD
To examine multipotency, MSCs were induced into adipogenic and osteogenic fates using a specific induction medium. MSCs were cultivated in a growth medium for 48 h. Next, the medium was removed and the cells were incubated in AD and OD media for 14 and 21 days, respectively. The AD medium comprised 4.5 g/L glucose DMEM, 5% FCS, 1% P/S, 0.1 µM dexamethasone (Sigma-Aldrich), 5 µg/ml insulin-transferrin-selenium, and 5 µM rosiglitazone (Sigma-Aldrich).
Cells were detached using a cell scraper, and the lysates were centrifuged at 28,400 × g for 2 min. Lysates were incubated with paranitrophenylphosphate (NPP, 2 mg/ml, Roth) substrate, dissolved in a buffer solution, containing 1 M TRIS and 5 mM MgCl 2 (pH 9.0), in triplicate at 37°C for 2 h. The mixture was transferred into 96well plates, and the ability of alkaline phosphatase (ALP) to metabolise NPP into p-nitrophenol (PNP) was measured as previously reported (Bessey et al., 1946). The absorbance of PNP was measured at 405 nm using a microplate reader (Tecan) and Magellan TM Data Analysis Software.

| RT-qPCR
Throughout the study, RNA samples were isolated from primary MSCs and from cells induced to MD and AD at days 7 and 14, and OD at days 7, 14, and 21. Non-induced cells in the BM were processed in parallel, and served as negative controls for all experimental setups. Approximately 1 µg RNA samples were extracted from all the groups using an RNA purification kit (Sigma-Aldrich). RNA samples were treated with a recombinant DNAse I (Roche) and RNase inhibitor (Thermo Fisher Scientific), and then reverse-transcribed into cDNA using a Multiscribe TM Reverse Transcriptase (Thermo Fisher Scientific).
Parallel preparation without reverse transcriptase was used as the negative control. The following cycling conditions were applied: 21°C for 8 min, 42°C for 15 min, 99°C for 5 min, and 5°C for 5 min, followed by cooling to 4°C. The obtained cDNA was mixed with a master mix containing Biotherm Taq Polymerase, ultrapure water, buffer, dNTP, and the respective forward and reverse primers. Random hexamers (Microsynth AG) were used as primers as listed in Table 1. PCR was performed for 35 cycles using the following cycling conditions: 5 min at 95°C, 30 s at 94°C, 30 s at 60°C, 30 s at 72°C, and 1 min at 72°C. Gel electrophoresis was performed to evaluate the PCR products. Relative expression of MD (MyoD and Myogenin), AD (peroxisome proliferator activated receptor gamma, PPARγ and fatty acid-binding protein, FABP4), and OD (osteopontin, OPN) markers was evaluated in triplicate via RT-qPCR using the GoTaq qPCR Mix (Promega). A thermal cycler was used for 2 min at 95°C, 15 s at 95°C, 30 s at 60°C, 5 s at 60°C, and 5 s at 95°C for 40 cycles using Bio-Rad CFX Manager 2.1 software (Bio-Rad Laboratories GmbH). 18S1 was used as an endogenous reference, and the relative gene expression was quantified using the 2-∆∆CT method, as described previously (Schmittgen & Livak, 2008).

| Oil Red O staining
The cells were fixed in 4% PFA at days 7 and 14 post-AD, and then washed twice in distilled water for 3 min. Briefly, the cells were washed with 60% isopropanol for 5 min, and incubated with freshly

| Alizarin Red S staining
The cells previously induced for OD were fixed in 4% PFA at days 7, 14, and 21 post-induction. The cells were washed twice for 3 min with distilled water, and then incubated with 2% Alizarin Red S (Roth) in distilled water for 15 min at RT. After three washing steps in distilled water, the matrix mineralised by calcium deposits was recognised as having a red-orange colour, and was photographed under an inverted light microscope.

| Immunocytochemistry
MSCs were seeded on sterile glass coverslips in 24-well plates at a seeding density of 1 × 10 4 cells/well. The cells were fixed in 4% PFA

| Statistical analysis
The data were collected from MTT, SRB, CFU, and wound-healing assays, as well as ALP activity and Oil Red O semi-quantification in triplicate (n = 6 per experimental group). Two-way ANOVA was performed to evaluate the effect of cell source (variable 1), i.e. HL, DI, MS, and EO muscles, and the time required for the gap closure (variable 2), indicative of efficient regeneration. Two-way ANOVA was also carried out to analyse the effect of cell source (variable 1) at different time points (variable 2), including 3, 5, and 7 days, on both cell viability and cell protein content. The effect of muscle source and seeding densities (100, 500, and 1000 cells/flask) on the colony formation ability was also assessed using two-way ANOVA.
Non-parametric one-way ANOVA was performed to evaluate the number of Pax7-positive cells per microscopic field (n = 10) for all muscle groups. Effect of MD on the number, width, and length of myotubes; number of myonuclei per tube; and nuclear cytoplasmic ratio of the myotubes for all muscle groups was also examined using one-way ANOVA. To analyse the effect of differentiation medium (variable 1), including AD and OD versus BM on ALP activity, Oil Red O semi-quantification, and the quantification of relative expression of markers at various time points (variable 2), two-way ANOVA was performed. Multiple comparisons were performed and variable interactions were assessed using Tukey's and Sidak's post-hoc tests.
The output is presented as mean ± SEM. Statistical significance was set at p < 0.05. Statistical analyses were performed using GraphPad Prism 7.0 software.

| Evaluation of proliferation and migration of MSCs from various sources
Cell viability was measured using the MTT colorimetric assay. Our data Conversely, the number of MS and EO MSCs showed no significant difference on days 3 and 5 (Figure 1b).
MSC migration capacity was assessed using wound-healing assay. Cell migration to close the interface gap between the separated cell populations was determined by using culture inserts.  (Figure 1c-m).

| Characterisation of stemness for MSCs isolated from various muscle groups
Expression of stemness and myogenic markers was evaluated using

| DISCUSS ION
Skeletal MSCs provide an interesting tool not only for understanding MD, but also for investigating tissue regeneration from a therapeutic perspective. We hypothesised that muscle origin regulates their myogenic fate and reflects their regeneration capacity. Thus, we investigated the stemness, MD potential, and multipotency of ovine MSCs from various embryonic sources. All experimental groups revealed expression of MSCs and myogenic markers, as shown using immunohistochemistry, PCR, and flow cytometry. Although Pax7 antigen was clearly observed and quantified using immunostaining, the mRNA expression of Pax7 was undetectable even in early passages (P2-P4). A possible explanation is that Pax7 was rapidly downregulated to an undetectable level or that mRNA expression was rapidly turned over in the activated muscle precursors. It was found that Pax7 is the most reliable marker for MSC identification (Kuang & Rudnicki, 2008;McCullagh & Perlingeiro, 2015). Some studies have reported that Pax7 is a marker of quiescent satellite cells (Kuang & Rudnicki, 2008;Shioi et al., 1995), and its expression decreases shortly after MSC isolation and passaging (Ding et al., 2017). Similarly, it has been documented that after muscle injury, MSCs become activated and express MyoD and Myogenin upon commitment to MD (Zammit et al., 2004(Zammit et al., , 2006. The data revealed strong expression of PDGFRα in MSCs, which is in agreement with a previous report suggesting its expression in skeletal muscles by fibro-adipogenic progenitors (Joe et al., 2010). Furthermore, the latter might play a positive role in muscle regeneration by producing a scaffold for efficient myogenesis as previously reported (Watt et al., 1987).
The adherence and colony-forming ability of MSCs have been considered as one of the characteristic features of stem cells. It has been reported that stem cells that efficiently form colonies are more proliferative and possess a greater capacity to differentiate (Pochampally, 2008). The data revealed that DI and HL muscle cells had higher colony-forming capacity, viability, and proliferation rate compared to EO and MS muscle cells. Thus, our data suggest an enhanced regenerative ability for the trunk muscles compared to head muscles, which probably requires a higher cell dose to perform in a similar manner when considering cell transplantation. The woundhealing assay, which is an indicative tool for cell migration, demonstrated that the HL and DI muscle cells have a superior potential for cell migration compared to cells of the head mesoderm, including the EO and MS muscles. Previous studies have shown that MSCs possess extensive cell mobility between myofibers and adjacent muscles across the extracellular matrix during muscle development and regeneration (Hughes & Blau, 1990;Westerblad et al., 2010). By considering the difference in the embryonic origin of the muscles from which the cells were isolated as HL and DI muscles originate from somites, whereas, MS and EO muscle cells are derived from the head mesoderm (Harel et al., 2009;Zammit et al., 2004), suggesting that somite cells have conserved stem cell properties compared to the cells of the head mesoderm. Moreover, the superior regenerative capacity of cells from trunk muscles suggests that head muscles utilise different myogenic programmes to achieve muscle repair. It has been reported that head muscles follow a specialised programme to produce satellite cells for initiating MD and post-natal muscle regeneration (Nogueira et al., 2015). Similarly, a study showed that eye muscles are not affected in some muscular dystrophies causing weakness and degeneration of body musculature (Emery, 2002).
In the present study, HL and DI muscle cells showed efficient MD, as indicated by upregulated expression of MyoD at day 7 and Myogenin at day 14 and abundant myotube formation. In agreement with our results, a study demonstrated that DI MSCs exhibit a significant rate of myogenesis compared to cells from the semimembranosus muscle of the HL (Redshaw & Loughna, 2012;Redshaw et al., 2010). Interestingly, the DI muscle cells showed wider myotubes compared to the EO muscle cells, which might be due to the future differences between both muscles, as fine and rapidly contractile fibres are required for eye movement (Zhou et al., 2011). The data highlighted that impaired MD in MS cells might be related to either a poor proliferation rate or a different myogenic programme orchestrating the development of the head muscles. In this context, it has been reported that the regulatory pathways that control skeletal myogenesis and cell fate are different in trunk muscles compared to the muscles of the head region (Harel et al., 2009). Similarly, it has been found that several genes are involved in the specification of pre-myogenic progenitors in the trunk, including Pax3 and ladybird homeobox1 (Lbx1); however, other molecular cues control head muscle specification. For example, the bicoid-related homeobox transcription factor (Pitx2) is required for head myogenesis, because depletion of Pitx2 causes complete absence of EO and masticatory muscle development (Shih et al., 2007(Shih et al., , 2008. Interestingly, HL and DI muscle cells displayed moderate adipogenic potential and were more prone to MD. Our study did not reveal any difference in lipid accumulation between HL and DI muscle cells.

The cells of MS and EO
In contrast, another study showed that semimembranosus muscle cells had greater lipid accumulation compared to the DI (Redshaw & Loughna, 2012). It was reported that FABP4 is normally expressed in skeletal muscle fibres, and its expression is increased during exercise (Fischer et al., 2006). The latter could be due to enhanced metabolic activity and oxidative capacity during exercise rather than a greater adipogenic potential.
Evaluation of OD revealed enhanced ALP activity, a common biochemical marker for osteoblast activity , calcium ions deposition at day 21, and upregulated OPN expression in the HL and DI muscle cells. Upregulated OPN expression is an indicative marker for osteogenic commitment; OPN expression was observed in the course of OD of mesenchymal stem cells, which suggests its important role in osteogenesis (Zohar et al., 1998). OPN is not only expressed in osteoblasts, but also biosynthesised by myoblasts (Uezumi et al., 2010). OPN has been reported to play a role in skeletal muscle inflammation and fibrosis following injury (Pagel et al., 2014). A similar report revealed that OPN is produced in muscle cells F I G U R E 2 Characterisation of stemness for MSCs isolated from various muscle groups. (a) The expression profile of mesenchymal stem cells markers including CD90, PDGFRα, MyoD and Myogenin using PCR. Lysates of the HL, DI, EO and MS muscle cells were processed for RNA extraction and cDNA synthesis using a reverse transcription kit. (b-d) Colony forming assay evaluates the ability of stem cells to form colonies. HL, DI, EO and MS muscle cells were seeded at various seeding densities (SD), SD 1 × 10 2 (b), 5 × 10 2 (c) and 1 × 10 3 /culture flask.
The number of small (10-20 cells), medium (21-50 cells) and large (50> cells) colonies were counted using crystal violet staining after 48 h.  (Seale et al., 2000). Similar studies have indicated that EO and MS muscles express more developmental 18S1 house-keeping genes were used as endogenous references. (g, h, I, j) Representative images for myotubes (white arrows) stained with phalloidin (green) show increased myogenic differentiation ability of HL and DI compared to EO and MS muscle cells. Cell nuclei were counterstained with DAPI (blue). All data presented as mean ±SEM. *p < 0.05, **p < 0.01 and ***p < 0.001. Scale bare =100 μm [Colour figure can be viewed at wileyonlinelibrary.com] myosin isoforms that disappear in other adult muscles (D'Albis et al., 1986;Will et al., 2015). Several other factors considered to modulate the pattern of MSCs, such as breed, gender, post-natal age, muscle fibre origin, and phenotype, have an effect on MSC number, proliferation rate, and transcription factor expression in vitro (Harding et al., 2015;Manzano et al., 2011;Torrente et al., 2004).
In the same context, the number of MSCs varies between different muscles and within the same muscle between different fibre types (Ding et al., 2017;Gibson & Schultz, 1982;Keefe et al., 2015;Zammit et al., 2004). For example, DI muscles displayed more MSCs per unit Images of DI cells stained with Alzarin Red S reveal calcium ions deposition (red) following OD induction at day 21 compared to cells in BM and day 7 of OD. All data are presented as mean ± SEM. **p < 0.01 and ***p < 0.001. Scale bare =200 μm [Colour figure can be viewed at wileyonlinelibrary.com] volume than the limb muscles (Keefe et al., 2015). Furthermore, maternal nutrition during gestation may influence the properties of muscle cells. A study demonstrated that restricted maternal nutrition during gestation alters the expression of myogenic regulatory factors, changes the proliferation rate, and impairs MD in MSCs of lambs (Raja et al., 2016). Moreover, the number of foetuses during gestation may be a determinant factor for post-natal muscle performance. A study has revealed that twin foetuses/neonates showed less muscle weight compared to singles due to reduced myofiber cross-sectional area and lower number of myogenic progenitors. The latter might play a role in myofiber hypertrophy in late gestation as well as early post-natal period (McCoard et al., 2001). Alternatively, maternal obesity in sheep may enhance intramuscular adipogenesis during foetal development (Yin et al., 2013) and possibly predispose MSCs to adipogenic induction.

| CON CLUS ION
MSCs offer a promising tool for regenerative medicine and cellbased therapy. Evaluation of stem cell populations would improve our understanding of the regenerative ability of each muscle. We show that ovine MSCs isolated from different muscle groups express specific stem cell and myogenic markers. By comparing the colony formation potential, viability, proliferation, and migration capacity, we concluded that the cells of the HL and DI muscles possess a greater regenerative potential compared to those of EO and MS muscles. Head muscle cells had a higher adipogenic capacity compared to the trunk muscle cells. In contrast, cells of trunk muscles displayed enhanced osteogenic potential, as shown by increased ALP activity and upregulated OPN expression. Collectively, these results suggest that MSCs derived from the paraxial mesoderm and head mesoderm followed a different myogenic programme. Additionally, we could assume that the DI and HL muscle cells are more suitable sources for biopsies and stem cell isolation for tissue regeneration.

ACK N OWLED G EM ENTS
We thank the Education, Audiovisual and Culture Executive Agency of the European Commission for partially funding the project in the framework of the Erasmus Mundus Master Program EUCOMOR.
Further thanks given to Manuela Heimann, Eva Kammer, Kathrin Wolf-Hofmann and Susanne Schubert-Porth for technical support.

CO N FLI C T O F I NTE R E S T
The authors have declared no conflict of interest.

AUTH O R CO NTR I B UTI O N S
MIE, KG, AE, MCK collected the raw data set and analyzed the results. MIE wrote the original draft of the manuscript and interpreted the results, SW and SA revised and finalized the submitted version of the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data collected and the analyses performed to generate the manuscript results are available from the corresponding author on reasonable request.

S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the Supporting Information section.