Hypoxic preconditioned bone marrow-derived mesenchymal stromal/stem cells enhance myoblast fusion and skeletal muscle regeneration

Background: The skeletal muscle reconstruction occurs thanks to unipotent stem cells, i.e., satellite cells. The satellite cells remain quiescent and localized between myofiber sarcolemma and basal lamina. They are activated in response to muscle injury, proliferate, differentiate into myoblasts, and recreate myofibers. Many stem and progenitor cells support skeletal muscle regeneration, which could be disturbed by extensive damage, sarcopenia, cachexia, or genetic diseases like dystrophy. Many lines of evidence showed that the level of oxygen regulates the course of cell proliferation and differentiation. Methods: In the present study, we analyzed hypoxic’s impact on human and pig bone marrow-derived mesenchymal stromal cell (MSC) and mouse myoblast proliferation, differentiation, and fusion. Moreover, the influence of the transplantation of human bone marrow-derived MSCs cultured under hypoxic conditions on skeletal muscle regeneration was studied. Results: We showed that bone marrow-derived MSCs increased VEGF expression and improved myogenesis under hypoxic conditions in vitro . Transplantation of hypoxic preconditioned bone marrow-derived MSCs into injured muscles resulted in the improved cell engraftment and formation of new vessels. Conclusions: We suggested that SDF-1 and VEGF secreted by hypoxic preconditioned bone marrow-derived MSCs played an essential role in cell engraftment and angiogenesis. Importantly, hypoxic preconditioned bone marrow-derived MSCs more efficiently engrafted injured muscles, however, they did not undergo myogenic differentiation.


Introduction
Skeletal muscle regeneration is a complex process that allows restoration of skeletal muscle homeostasis lost due to the injury, such as intensive exercise, surgical procedures, and diseases. Skeletal muscle regeneration covers two distinct phases. The first one includes tissue degeneration, accompanied by inflammation, necrosis of damaged myofibers, and their phagocytosis by immune cells.
The second one is regeneration, leading to new myofiber formation followed by their maturation, tissue reinnervation, and finally skeletal muscle functional recovery (1).

Muscle necrosis occurs when myofibers' integrity is severely disrupted what involves increased
sarcolemma permeability, organelle dysfunction, and loss of myofiber architecture. Necrotic cell death and loss of plasmalemma integrity lead to the release of intracellular components that act as damageassociated molecular patterns (DAMPs) and trigger an inflammatory response (2,3). Necrotic myofibers release many cytokines, growth factors, and chemoattractants. These signals activate tissue-resident and circulating inflammatory cells (4,5). Neutrophils are the first to infiltrate the site of injury. These cells phagocytize damaged myofibers and release numerous factors which induce migration of local monocytes and their differentiation into macrophages (6)(7)(8). Two days after injury, macrophages become the predominant cell population present within damaged tissue (5,9,10). They can be divided into two distinct subpopulations -M1, also considered as pro-inflammatory macrophages, characterized by the presence of CD68, and responsible for phagocytosis of necrotic tissue and releasing pro-inflammatory factors, such as tumor necrosis factor α (TNF-α), interleukin-1β (IL-1β), IL6, IL12, as well as nitric oxide (NO) and reactive oxygen species (ROS) (11)(12)(13). The second described population is M2, also called anti-inflammatory macrophages, characterized by the presence of CD163, releasing factors, like tumor growth factor β (TGF-β), IL4, IL10 or IL13, and for supporting myoblast differentiation, angiogenesis, and extracellular matrix (ECM) deposition (14,15). The next phase of skeletal muscle repair covers myofiber regeneration which is possible due to satellite cells (SCs) -skeletal musclespecific stem cells, characterized by a PAX7 transcription factor. These cells are tightly connected to the myofibers and located between basal lamina and sarcolemma. In healthy muscles, SCs remain quiescent, but after injury, they become activated, re-enter the cell cycle, start to proliferate, differentiate into myoblasts which further fuse to form myotubes. Finally, myotubes' maturation leads to new functional myofibers' formation (7,(16)(17)(18).
Quiescent SCs can quickly respond to changes in their niche and specific signals present in their microenvironment. Among the crucial factors causing SC activation and proliferation are mechanical disruptions of sarcolemma and action of growth factors released by inflammatory cells, endothelial cells, interstitial cells, such as fibroblasts, or released from the ECM by specialized proteases -metalloproteinases (MMPs) (19,20). Fibroblast growth factor (FGF) is one such factor. Its action via transient receptor potential channels (TRPC) leads to the translocation of nuclear factor of activated Tcells (NFATc) into the nucleus resulting in SC activation (21,22). FGF action is also known to activate the p38 mitogen-activated protein kinase (p38 MAPK) pathway, which acts as a molecular switch during SC activation (23). Another ECM-derived factor crucial for SC activation is hepatocyte growth factor (HGF), which binding to hepatocyte growth factor receptor (HGFR/c-Met) promotes SC re-entry into the cell cycle (24). One of the signals produced by such cells as fibroblasts or myofibers is insulin-like growth factor 1 (IGF-1), which is well known to stimulate the PI3K/Akt/mTOR pathway and to downregulate the activity of transcription factor Forkhead box O (FOXO), what in consequence leads to downregulation of p27 cell cycle inhibitor and activation of SC cell cycle (25). Some of the previously mentioned factors are known to directly or indirectly regulate SC activation and proliferation. Among them are, for example, TNF-α activating nuclear factor kappa B (NF-κB) pathway, which leads to silencing Notch1, NO which stimulates MMP expression and ECM remodeling or IL6, which stimulates SC proliferation in signal transducer and activator of transcription 3 (STAT3)-dependent manner (19,(26)(27)(28). Finally, activated and proliferating SCs start to differentiate.
The myogenic differentiation of SCs is regulated by sequentially expressed transcription factors, called myogenic regulatory factors (MRFs). MRF family consists of MYF5, MYOD, myogenin, and MRF4 (29,30). Quiescent SCs are characterized by the presence of paired box transcription factor 7 (PAX7).
PAX7 and MYF5 are present in proliferating SCs and myoblasts. PAX7 regulates the expression of MYF5 and MYOD, while MYF5 presence enhances the expression of MYOD. MYOD is a critical factor of myogenic differentiation. It facilitates the transition from myoblast proliferation to the myocyte differentiation stage by inducing the myogenin and p21 and p57 cell cycle inhibitor expression (31,32).
Further, MYOD and myogenin trigger the expression of other genes essential for muscle cell function, such as MRF4, myosin heavy and light chains, muscle creatine kinase, or troponin (33). The expression of myogenin and MRF4 is accompanied by the downregulation of PAX7, MYF5, and MYOD. Some cells do not undergo differentiation but remain PAX7 positive, downregulate MYOD, and restore the SC population necessary for the next rounds of muscle regeneration (34)(35)(36). Those that differentiated fuse to each other to result in the formation of multinucleated myotubes and then myofibers (37,38).
Alternatively, differentiated myocytes can fuse with already existing myofibers during the regeneration of slightly damaged skeletal muscles (39). Finally, newly formed myotubes and myofibers undergo maturation to become fully functional. During maturation, myofibers grow, myofibers' proper contractility is restored, and neuro-muscular junctions are formed (40, 41).
In skeletal muscle diseases, sarcopenia, or cachexia, skeletal muscle regeneration is disturbed.
Many populations of stem and progenitor cells are studied for potential therapeutic use. Two main strategies to support skeletal muscle regeneration are considered. First, the transplanted cells could participate in myofiber reconstruction; second, transplanted cells' secreted factors could support regeneration. One of the studied cells are bone marrow-derived stromal cells, also known as bone marrow-derived "mesenchymal" stem/stromal cells (bone marrow-derived MSCs). However, it should be noted that these cells do not present naïve myogenic potential (42). Bone marrow-derived MSCs are a heterogeneous population (43), typically isolated from bone marrow based on their ability to adhere to the culture plate's surface. It was proven that bone marrow-derived MSCs contain a population of cells that fulfill the rigorous criteria of stem cells (44). This subpopulation of bone marrow-derived MSCs present long-term expansion without phenotypic change, self-renewal probed during in vivo serial transplantations, and multipotency examined by in vivo differentiation assay at the single-cell level (44)(45)(46)(47)(48). CD146 appeared to be a handy marker to select and isolate stem cell subpopulations from bone marrow-derived MSCs (45). Human CD146+ bone marrow-derived MSCs were shown to be able to selfrenew, differentiate into bone and bone marrow, a support organization of endothelial cells into functional blood vessels, and differentiate into chondrocytes and adipocytes (45,49).
In the current study, we focused on the hypoxic effect on bone marrow-derived MSC and myoblast co-cultures. We also followed if cultured under hypoxic condition bone marrow-derived MSCs could more efficiently support skeletal muscle regeneration. The level of oxygen is an essential factor regulating gene transcription and cell fate. The level of O2 during in vitro culture under hypoxic conditions (1-3%) is much more similar to the level present in the physiological bone marrow-derived MSC niche in the bone marrow (2-7%) than that observed under standard in vitro culture conditions. Accordingly, it was previously shown that bone marrow-derived MSCs cultured under hypoxic conditions induced their proliferation, migration, elevated colony-forming unit capabilities, increased ECM deposition, osteogenic and adipogenic potential, and angiogenic factors expression (65)(66)(67)(68)(69)(70)(71).
Moreover, preconditioning of bone marrow-derived MSC with hypoxic increased their ability to engraft injured tissues after transplantation. In the subacute murine limb ischemia model, hypoxic preconditioned bone marrow-derived MSCs injected into skeletal muscles engrafted this tissue more efficiently, induced neoangiogenesis, and improved blood flow (69). Similar results were observed after transplantation of hypoxic preconditioned bone marrow-derived MSCs to other ischemic tissues, including heart, brain, lung, and liver (70)(71)(72)(73)(74). We hypothesized that hypoxic preconditioning impacts the human bone marrow-derived MSC secretome. As a result, these cells could more efficiently engraft injured skeletal muscle, support myoblast fusion, and skeletal muscle regeneration. To follow this problem, we choose to investigate human and pig bone marrow-derived MSCs. We selected cells of two species as we previously showed that as far as MSCs are concerned, the cells' origin may determine their reaction to the same factors (75). Moreover, pig serves as a valuable model in preclinical research.

Migration assay -scratch assay
Migration of pMSCs or hMSCs cultured either under hypoxic or normoxic was analyzed using scratch wound healing assay (78). Briefly, cells were cultured to obtain 90-100% confluence. Next, the cells were scratched from the plate using a plastic tip to create the "wound". The wound healing manifested by the ability of the cells to refill the created gap was observed. After 3.5h, 8h and 24h cells were fixed with cold methanol and stained using Giemsa-May Grünwald method. The pictures were taken, and the area of the scratch was calculated using GIMP 2.

Myoblast and bone marrow-derived MSC co-culture, fusion index, and hybrid myotube analysis
Co-cultures were obtained by seeding mPM in a 1:1 ratio with either pMSCs or hMSCs. Cells were cultured under normoxic or hypoxic conditions in MSCmed or PMmed for 5-7 days. C2C12 myoblasts (3x10 4 or 6x10 4 ) were cultured in the absence of hMSCs or pMSCs or co-cultured with hMSCs or pMSCs in 3:2.5; 3:5; 3:7.5 ratio. Cells were cultured under normoxic or hypoxic conditions in C2C12med for 5-7 days. Further, cells were fixed, and fusion index or proportion of hybrid myotubes were estimated. Fusion index of C2C12 or mPM cultured alone or in co-cultures either with pMSCs or hMSCs was calculated. Briefly, differentiated cells were fixed in cold methanol and stained according to the Giemsa-May Grünwald method. Images from 4 fields of view were collected, and nuclei number was counted.
Fusion index was calculated as a percentage of nuclei present in myotubes compared to all visible cell nuclei.
Myotubes formed by either C2C12 or PM co-cultured either with pMSCs or hMSCs were visualized using skeletal myosin's immunolocalization. The participation of hMSCs in myotube formation was evaluated by visualization of human nuclei. pMSC contribution in hybrid myotubes formation was verified by the presence of GFP within myotubes. In contrast, the contralateral leg was injected with 0.9% NaCl solution (saline-treated muscles served as a control). After 14 days of regeneration, mice were sacrificed, muscles were isolated, and analyzed.

Immunocytochemistry and immunohistochemistry
Selected antigens were immunolocalized in in vitro cultured cells as well as in muscle crosssections. In vitro cell cultures were fixed with 3% PFA, washed with PBS, and stored in 4°C. Dissected skeletal muscles were frozen in isopentane, cooled down with liquid nitrogen, transferred to -80°C, and cut into 10 μm sections using cryomicrotome (Microm HM, Thermo Fisher Scientific). Cryosections were fixed with 3% paraformaldehyde, washed with PBS, and stored in 4°C. Further fixed cells or cryosections were permeabilized with 0.1% Triton X-100 (Sigma-Aldrich) in PBS and incubated with 0.25% glycine AlexaFluor 488 (A11008, ThermoFisher Scientific), or donkey anti-rabbit conjugated with AlexaFluor 594 (A21207, ThermoFisher Scientific).

Muscle histology
Dissected skeletal muscles were frozen in isopentane, cooled down with liquid nitrogen, transferred to -80°C, and cut into 10 μm sections using cryomicrotome (Microm HM, Thermo Fisher Scientific). Cryosections were fixed with 3% PFA, washed with PBS, and stored in 4°C. Samples were hydrated in PBS, incubated in Harris hematoxylin solution (Sigma-Aldrich), and washed in distilled water.
Then, fixed sections were incubated in eosin Y solution (Sigma-Aldrich) and washed in distilled water.
Specimens were mounted with UltraMount (Dako Cytomation) and analyzed using inverted light microscope Eclipse TE200 (Nikon) and ImageJ software (NIH).

Gene expression analysis
Total RNA was isolated from muscles, C2C12, mPM, pMSCs, and hMSCs cultured alone or in cocultures, using High Pure Isolation Kit (Roche) and from dissected muscles using mirVana™ miRNA Isolation Kit (Thermo Fischer Scientific) and purified with Turbo DNA-free Kit (Thermo Fischer Scientific), according to the manufacturers' protocols. cDNA was obtained in reverse transcription reaction performed using RevertAid First Strand cDNA Synthesis Kit (ThermoFisher Scientific) according to manufacturer's protocol. The conditions of reverse transcription were as follows: 25°C for 10 min, 42°C for 60 min, 85°C for 5 min. Next, mRNA levels were examined using quantitative real-time PCR analysis (qPCR) with TaqMan assays (ThermoFisher Scientific) for the following genes: human: CD9  Figure 4A). Also, the level of transcripts encoding human CD9 and CDH15 significantly increased in cells cultured under hypoxic conditions. Hypoxic and myoblasts' presence did not alter the MYOG expression in pMSCs, but the high level of DES (desmin) was noticed in cells cultured in MSCmed ( Figure 4B).

The transplantation of human bone marrow-derived MSCs cultured under normoxic and hypoxic conditions into mouse injured skeletal muscles
As we demonstrated, human and pig bone marrow-derived MSCs influenced myoblasts' proliferation and fusion when cultured under hypoxic conditions. Since any significant differences between human and pig cells' impact on mouse myoblasts were found, we decided to inject human Finally, hMSC cultured under normoxic conditions and hypoxic preconditioned hMSCs were transplanted into ctx injured skeletal muscle of SCID mice ( Figure 6). The muscle mass, nerve area, and a number of newly formed myofibers did not differ between muscles that received hMSCs cultured under either normoxic or hypoxic conditions ( Figure 6A, B). Importantly, the area of blood vessels was higher after hypoxic preconditioned hMSC transplantation. Moreover, a higher number of human cells was detected in mouse muscles injected with hypoxic preconditioned hMSCs ( Figure 6A, C). These cells were found between myofibers ( Figure 6C). However, the higer area of connective tissue in muscles injected with hMSCs cultured under hypoxic conditions compering to control muscles was noticed ( Figure 6A). Then, we analyzed the level of mouse and human transcripts after cell transplantation. The level of mouse Vwf was lower in injured muscles than in intact muscle; however, it did not differ between muscles transplanted with hMSC cultured under either standard or hypoxic conditions ( Figure   6D). Notably, only in mouse muscles injected with hypoxic preconditioned hMSCs the human transcripts such as laminin, VCAM, MCAM, PDGFRb, CSPG4 (NG2), FAP, CXCL12 (SDF-1), and VEGF were found ( Figure 6E). Besides, the WNT, MYH3, MYF5, MYOD1, and MYOG transcripts were not detected. We concluded that hypoxic preconditioned hMSCs efficiently engrafted injured muscle but did not follow myogenic differentiation based on obtained results.
In our study, we noticed an increase of primary myoblast proliferation when cultured under hypoxic conditions, which stays in agreement with other studies showing the higher proliferation of rat, human, and mouse primary myoblasts cultured in hypoxic (2-6% O2) (95)(96)(97)(98). The C2C12 proliferation also increased in co-cultures with mouse bone marrow-derived MSCs in VEGF dependent manner (99).
Importantly, under hypoxic conditions, we observed the increased VEGF expression in human bone marrow-derived MSCs. However, neither human nor pig bone marrow-derived MSC proliferation or migration changed when comparing the cells cultured under normoxic or hypoxic conditions. Although no significant influence of hypoxic on rat bone marrow-derived MSC migration was found, it was shown that the oxygen concentration affected bone marrow-derived MSC response to chemokines, inflammatory cytokines, and growth factors (100). On the other hand, mouse bone marrow-derived MSCs migrated more efficiently, in transmembrane migration assay, in response to conditioned medium under hypoxic conditions (68). Human bone marrow-derived MSCs cultured under hypoxic conditions increased their migration rates and HGF receptor expression, i.e., c-Met (101). Thus, the influence of hypoxic on bone marrow-derived MSC migration abilities depends on many variables.
In the current study, we documented that primary myoblasts and C2C12 myoblasts fused less efficiently under hypoxic conditions. Other studies showed that the impaired fusion of mouse C2C12 myoblasts cultured under hypoxic conditions was also connected to myotube atrophy and a lower number of nuclei per myotube (102). Also, mouse primary and H-2K myoblast differentiation were less efficient under such conditions (103). Moreover, most of the studies described the inhibition of myogenic differentiation under hypoxic (91). The response of mouse satellite cells to oxygen level changes is regulated by HIF2A (89). Quiescent satellite cells residing within the niche are hypoxic and express HIF2A, which maintains their quiescence and self-renewal and blocks differentiation (89). We was also documented that mouse bone marrow-derived MSCs overexpressed Vegf and secreted a higher VEGF level under hypoxic conditions (67,68). Moreover, both mouse primary myoblasts and C2C12 myoblasts expressed VEGF and its receptors (107). VEGF enhanced C2C12 myoblast migration and prevented apoptosis (107) and myogenic differentiation, which resulted in the promotion of myotube hypertrophy of C2C12 cells, increased mitogenic activity, migration, and proliferation (99,108). Thus, we concluded that an increase of VEGF secretion by bone marrow-derived MSCs, observed under hypoxic conditions, could account for improved myoblast proliferation and differentiation.
Finally, we transplanted hypoxic preconditioned human bone marrow-derived MSCs to injured skeletal muscles of SCID mice to study their influence on tissue regeneration. Bone marrow-derived MSCs cultured under hypoxic more efficiently engrafted the muscle. However, they were found only between myofibers. We were able to detect human lamina, VCAM, NG2, CD146, PDGFR, and FAP in muscles after hypoxic preconditioned bone marrow-derived MSC transplantation with more effective engraftment. However, we did not notice the presence of myofibers formed by human bone marrowderived MSCs, and the expression of human MYF5, MYOD, MYOG, and MYH3 was not detected. Thus, we concluded that hypoxic preconditioned bone marrow-derived MSCs more efficiently engrafted injured muscles but did not follow myogenic differentiation. Notably, the human SDF-1 and VEGF transcripts were present in mouse muscle after hypoxic preconditioned bone marrow-derived MSC transplantation, and these factors could impact skeletal muscle regeneration. We suggested that VEGF, which is upregulated in hypoxic preconditioned bone marrow-derived MSCs, could play a vital role in cell engraftment after transplantation. Verma and coworkers showed that tissue-resident satellite cells expressed VEGF, which recruited endothelial cells (109). In this way, satellite cells induced capillary formation in their niche. We suggested that improved expression of VEGF in hypoxic preconditioned bone marrow-derived MSCs could induce vessel formation and support cell engraftment. A higher number of vessels was found in muscles after hypoxic preconditioned bone marrow-derived MSC transplantation. Reconstruction of vessel network is essential for muscle reconstruction, and the reduction of skeletal muscle network was described in dystrophic, ALS, or denervated muscles (110).
Acute muscle damage led to disruption in the microvasculature, hypoxic, and activation of HIF-1α signaling -the main factor of hypoxic response (111). One of the HIF-1 target genes is VEGF, i.e., a welldescribed factor triggering angiogenesis also in skeletal muscles (111). VEGF and angiogenesis improved skeletal muscle regeneration and chronic skeletal muscle diseases (108,(111)(112)(113). Similarly, SDF-1 expression increased in injured muscles, presented a proangiogenic effect, and mobilized stem cells to injured muscles (114)(115)(116)(117)(118). The restoration of blood flow and vascular formation was also enhanced after intra-arterial injection of hypoxic preconditioned mouse bone marrow-derived MSCs to mice with hind limb ischemia (101). Hypoxic preconditioned mouse bone marrow-derived MSC transplantation increased WNT4 expression in skeletal muscles, and WNT4 was shown to induce bone marrow-derived MSC proliferation and migration well as endothelial cell migration and myoblast differentiation (101).
Based on all the abovementioned results, we concluded that SDF-1 and VEGF secreted by hypoxic preconditioned bone marrow-derived MSCs increased new vessel formation during skeletal muscle reconstruction.

Conclusions
The hypoxic induced proliferation of myoblasts but delayed their differentiation, decreased transcripts encoding CD9, and increased PAX7 and MRFs transcripts expression.