*3.5. Myogenic Di*ff*erentiation by MT-derived EVs*

MT-derived EVs express specific cell-adhesion molecules on their surfaces (ITGB1, CD9, CD81, CD44, Myoferlin) that are involved in the recognition and adhesion during the process of myoblast fusion [45–47].

Further, MT-derived EVs contain functionally active proteins of the G-protein family, which are involved in many cellular processes including myogenesis [48].

Wnt signaling and its modulation via GSK3 inhibitors, such as CHIR99021 (CHIR), is essential in the mesoderm induction and to obtain a reliable, reproducible and efficient myogenic differentiation protocol [18,19,49].

On the basis of these findings, we developed a method to induce a robust differentiation of iPSC toward skeletal muscle phenotype combining MT-derived EV cargo and the chemical modulator CHIR for paraxial mesoderm-like muscle progenitor commitment (Figure 4A).

PC- and FB-derived iPSCs, derived from three donors, were divided in four groups and treated as follows: (i) group 1, PC-/FB-derived iPSCs cultured in differentiation medium; (ii) group 2, PC-/FB-derived iPSCs cultured in differentiation medium augmented with EVs; (iii) group 3, PC-/FB-derived iPSCs cultured in differentiation medium enriched with CHIR; and (iv) group 4, PC-/FB-derived iPSCs cultured in differentiation medium supplemented with EVs and CHIR.

Cells were treated with 10 uM CHIR for 48 h to induce the expression of paraxial mesoderm genes. Consistent with other studies [18,19], iPSCs from both sources presented evident morphological changes losing the typical ES-like morphology after 24 h of treatment (Figure 4B). Following replacement of CHIR with FGF2, cells underwent proliferation and reached full confluence approximately at day 8–10.

**Figure 4.** Myogenic differentiation by MT-derived EVs. (**A**) Schematic diagram of the myogenic differentiation procedure: 30% confluent iPSCs were differentiated to early mesoderm using CHIR for 48 h, subsequently proliferation and expansion of muscular progenitors were stimulated utilizing EVs, FGF2, IGF, and the myotube-like cell maturation was induced, removing growth factors from the culture medium. (**B**) Morphological changes of PC-derived iPSC colonies during skeletal muscle differentiation. Scale bars represent 100 μm. (**C**) Representative qRT-PCR for the expression of early (Mesogenin, Pax3, Pax7, MyoD, MyoG) and late (MyH8, MCK, MHC) skeletal muscle genes. The analysis was performed at different time points (d0, d10, d20, d30) in the following conditions: FB- and PC-derived iPSCs not treated with EVs or CHIR (FB-iPSCs E6 and PC-iPSCs E6), exposed to CHIR or EVs only (FB-iPSC EVs, FB-iPSC CHIR, PC-iPSC EVs, PC-iPSC CHIR) and treated with both, EVs and CHIR (FB-iPSC EVs/CHIR and PC-iPSC EVs/CHIR) (n = 3 donors). Results were normalized to GAPDH.

Following 48 h of treatment, CHIR was removed and, in order to enhance the myogenic differentiation, the media were enriched with freshly MT-derived EVs (50 μg/mL) and replenished every 2 days with fresh vesicles until the formation of myoblast-like cells. Between days 15 and

#### *Cells* **2020**, *9*, 1527

25 paraxial mesoderm-like muscle progenitors can be either expanded or terminally differentiated upon withdrawal of all the growth factors from the medium. The outcome of myotube formation was greater when myoblast-like cells were passaged 2–3 times and consequently exposed longer to MT-derived EVs.

We have analyzed the expression of genes related to mesoderm induction and myogenic differentiation at different time points through qRT-PCR experiments (Figure 4C).

In PC- and FB-derived iPSCs treated with the CHIR/EV combination, pre-myogenic mesoderm genes, such as Mesogenin, Pax3, and Pax7, were upregulated between day 10 and day 20. The myogenic regulatory factor MyoD and MyoG started to be expressed around day 10 and increased up to day 30. Finally, the mature myocyte genes, Myh8 (myosin heavy chain 8), MCK (muscle creatine kinase) and MHC, were detected after day 20 and augmented until the end of the differentiation confirming the increase in differentiation and maturation of the iPSCs treated with both factors.

Cells cultured in differentiation medium were negative for the expression of myogenic genes, while iPSCs treated with EVs exhibited only pre-myogenic genes (Pax3 and Pax7), indicating that the MT-derived EVs are not sufficient, at least within 30 days, to induce myotube differentiation. Lastly, cells exposed to CHIR presented a myogenic inclination similar, but less efficient, to PC-/FB-derived iPSCs treated with the CHIR–EVs mishmash.

Our method proved that the combination CHIR-EVs induces a higher differentiation compared to CHIR treatment (Figure 5A).

To verify whether PC-derived iPSCs, generated from the three donors, possessed a greater propensity to differentiate into myotubes compared to FB-derived iPSCs, we matched the expression of Myh8, MCK, and MHC by qRT-PCR analysis at day 30 of differentiation (Figure 5A). The expression of mature myogenic genes resulted as being higher in differentiated PC-derived iPSCs.

To further determine the role of the EVs/CHIR combination we calculated the fusion index, which was significantly greater compared to the index obtained with the separate exposure of EVs or CHIR (Figure 5B). Consistently, also the cell number expressing MHC was also significantly greater in those generated using the CHIR–EVs cocktail compared to the other treatments (Figure 5C).

The ability of PC-derived iPSCs to generate differentiated cells presenting a characteristic mark of myogenesis, was further confirmed by FACS analysis. Myogenic regulatory factors, Pax3, Pax7, CD56, MYOD, MYOG, and MF20, were used to characterize the expression of myogenic proteins (Figure 5D). At day 30 roughly 80% of the cells were positive for the myocyte mature marker MF20.

Finally, after 25 days of CHIR–EV differentiation, PC-iPSC-derived muscular cells, were implanted in vivo in a limb girdle muscular dystrophy murine model, namely, SCID-Beige α–Sarcoglycan null mice (αSGKO/SCIDbg) [5] in order to evaluate their engraftment capabilities. Three different PC-derived iPSC clones were intramuscularly injected (5 <sup>×</sup> 105 cells/injection) into the anterior tibialis (TA) muscle of αSGKO/SCIDbg, showing a sufficient engraftment and integration into regenerating muscle fibers revealed by the labelling of human derived cells by immunofluorescence against human specific lamin A/C antibody (Figure 5E).

**Figure 5.** iPSC-derived MT characterization. (**A**) Representative qRT-PCR for the expression of Myh8, MCK and MHC after 30 days of differentiation. The combination CHIR-EVs induced a greater differentiation compared to CHIR treatment and PC-derived iPSCs possessed a greater propensity to differentiate into myotubes compared to FB-derived iPSCs. Results were normalized to GAPDH. (**B**) Fusion index was calculated as the percentage of nuclei within myosin heavy chain–positive myotubes (≥2 nuclei) divided by the total number of nuclei. A minimum of five random fields at 10X magnification were counted. (**C**) Immunofluorescence labeling for MHC (green), validating the differentiation capacity of PC-derived iPSCs toward skeletal muscle phenotype after 30 days of differentiation. The images represent sequentially: PC-derived iPSCs without treatment, exposed to EVs alone, to CHIR only and to the EV–CHIR cocktail. CHIR in combination with MT-derived EVs induced a higher expression of the skeletal marker (n = 3 donors). Nuclei were stained with DAPI. Scale bar represents 100 μm. (**D**) Representative histograms indicating the percentage of PAX3, PAX7, CD56, MYOD, MYOG, and MF20 (black peaks) determined by flow cytometry in PC-iPSC-derived MTs exposed to CHIR and MT-derived EVs after 30 days of differentiation. Matched isotypes were used as negative controls (grey peaks) (n = 3 donors). (**E**) Representative immunofluorescence against human lamin A/C (magenta) and laminin (yellow) on mouse anterior tibialis (TA) injected with human PC-derived iPSC exposed to CHIR-EVs (n = 3 donors). Arrows indicate center-nucleated human derived myofibers, arrowhead labelling integrated human nuclei into mature host muscle fiber. Nuclei were stained with DAPI. Scale bars represent 25 μm.

#### **4. Discussion**

In the last few years, there has been a growing interest in the use of iPSCs as a source of myogenic progenitors for cell-based treatment in muscle degenerative diseases. iPSCs overcome several of the limitations related to the use of adult myoblast therapy, such as, non-invasive biopsy for cell isolation, unlimited proliferative capacity in vitro, and a tool for the in vitro correction of genetically mutated somatic cells derived from patients affected by dystrophies [14–16,19,50–52]. Moreover, it is well known that iPSCs maintain the epigenetic memory of the original cell source, influencing the re-differentiation toward the same lineage [30,53]. Pericytes resident in adult skeletal muscle have shown remarkable angiogenic and myogenic differentiation capacities in vitro and in vivo [7]. Hence, in this study, exploiting the advantage of the "epigenetic retention", we proposed the isolation of human derived muscular pericytes to verify whether PC-derived iPSCs are more prone to differentiate into muscular cells compared to FB-derived iPSCs.

To date, the strategies used for iPSC myogenic differentiation are subdivided into two approaches: transgenic (by forced expression of Pax7 or MyoD) or non-transgenic (co-culture, EBs, small molecules) [15,16,51]. The former consists in the differentiation via overexpression of myogenic transcription factors, obtaining efficient iPSC differentiation into myogenic lineage in a relatively short amount of time [15,16,51]. However, forced expression of skeletal master genes drastically reduces the proliferative capacity of skeletal muscle progenitor cells, hindering the molecular mechanisms leading to myogenic differentiation (important for disease modeling in vitro), as well as their in vitro and in vivo self-renewal. Another point to be considered is the potential of insertional mutagenesis events due to the random integration, since these genes are commonly delivered using lentiviral vectors. Integrated virus genomes are frequently associated with chromosomal damage, rearrangements and deletions [54], making these approaches not suitable in terms of clinical applications. On the other hand, the non-transgenic methods use different sequential culture conditions, including cell sorting. These protocols are successful at producing myogenic progenitors capable of engrafting in vivo but lack reproducibility, besides being inefficient and time-consuming [55–57].

First-generation of transgene-free protocols employing different sequential culture conditions were not satisfactory and usually required post-differentiation selection to isolate expandable skeletal muscle progenitor cells [57]. In addition, in some described methods, iPSCs failed to express mature myogenic markers or form myotubes in vitro, underlying the poor efficiency of spontaneous myogenic differentiation method [55,56].

Improvements in muscle differentiation protocols were recently performed by employing a small molecule, CHIR99021, a GSK-3b inhibitor that activates Wnt signaling cascade, which plays a crucial role during early somite induction. These approaches have been proven to be efficient at boosting iPSC towards a myogenic pathway, leading to an ameliorated myotube differentiation in vitro [19,58,59].

Quite recently, considerable attention has been dedicated to extracellular vesicles (EVs) as important mediators of cell-to-cell communication. EVs can influence the behavior of recipient cells and are involved in many processes, including immune signaling, angiogenesis, stress response, proliferation, and cell differentiation [60–62]. Furthermore, EVs, through their paracrine signaling can be used in tissue engineering to modulate cell recruitment, differentiation, and proliferation [63]. Skeletal muscle cells secrete a large number of myokines and EVs that influence the growth, function and development of muscle tissue [26,27,41]. Transcriptome and proteome studies reported that EVs from C2C12 or human myoblasts are enriched in miRNAs and proteins that are implicated in myogenesis [22,27,40,41,64]. EVs isolated from differentiated C2C12 cells express specific cell-adhesion molecules on their surfaces (ITGB1, annexins CD56, CD9, CD81, CD44, Myoferlin), and other myokines (HGF, IGFI/II, FGF2, PDGF, TGFβ, myostatin) involved in myoblast fusion and muscle regeneration. [40,41,45,47,65]. In addition, differentially expressed miRNAs also contained in EVs during C2C12 myotube differentiation, such as miR-1, miR-133a, miR-133b, and miR-206, have been linked with muscle differentiation [64,66]. Interestingly, the Wnt signaling and Sirt 1 that are involved in muscle gene expression and differentiation, were predicted as the most significant targeted pathway by skeletal muscle EVs–miRNAs [64]. In the present study, we found that MT-derived EVs harbored myogenic factors that are crucial in enhancing the iPSC myogenic commitment. Despite several scientific

papers reporting transcriptome and proteome profiling of EVs from skeletal muscle, the mechanisms and the key factors playing a major role in promoting skeletal muscle differentiation remain unclear. Evidence suggests that instead of single factors, the diverse EVs myogenic factors such as miRNAs or proteins, synergistically trigger skeletal muscle differentiation in recipient cells [22,26,27,40,41,64]. Further studies are required to define the biological role/function of single myogenic factors in MT-derived EVs.

Here, we have developed an efficient method for iPSC skeletal muscle differentiation combining defined factors with myogenic elements carried by EVs. We were able to generate a consistent number of iPSC-derived positive MHC skeletal muscle cells, evidencing differences in epigenetic signatures between PC-derived iPSCs and their counterpart FB-derived iPSCs. Indeed, quantitative PCR analyses performed at different timepoints and at the end of the differentiation, showed that PC-derived iPSCs are more prone to differentiate in skeletal muscle cells when compared to FB-derived iPSCs. Importantly, the myoblasts obtained with this approach could be expanded, and cryopreserved at all steps during expansion, without losing fusion competence. The obtained MT-like cells were positive for MyoD, MYOG, and MHC, and were able to actively participate in myogenesis in vivo. Transplanted myogenic progenitors demonstrated their ability to successfully engraft and integrate into TA muscle of αSGKO/SCIDbg, identified by the detection of centrally located human lamin A/C positive nuclei, a characteristic perceived only in fusion-competent myoblasts.

In this paper, we explored for the first time EVs as "physiological liposomes" enriched with myogenic factors that were able to trigger skeletal myogenesis in iPSC, raising exciting possibilities for therapeutic use. The use of EVs represent novel tools to deliver signaling molecules for treating muscle-wasting diseases that are better tolerated by the immune system. Additional studies are now required to investigate the role of EVs' skeletal muscle content that may allow the development of better therapeutic approaches to promote skeletal muscle growth, differentiation and regeneration.

Our method may better recapitulate human developmental myogenesis providing inroads for patient-specific drug testing, disease modeling and regenerative approaches.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4409/9/6/1527/s1, Figure S1: Fibroblast characterization, Figure S2: Silencing of the exogenous factors, Table S1: List of human primers used for qRT-PCT analysis.

**Author Contributions:** Conceptualization, D.B., R.R., and C.B.; data curation, D.B.; formal analysis, D.B., M.C., F.M., S.G., and A.B.; funding acquisition, R.R., and C.B.; investigation, D.B., M.C., V.P., F.M., M.M., A.R., P.S., D.P., S.G., and A.B.; methodology, D.B., M.C., R.R., and C.B.; project administration, D.B., R.R., and C.B.; resources, S.C., C.L., C.G., R.R., and C.B.; supervision, S.C., C.G., R.R., and C.B.; validation, D.B., M.C., V.P., F.M., S.G., and A.B.; writing—original draft preparation, D.B.; writing—review and editing, C.L., C.G., R.R., and C.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Italian Ministry of Health—grant Giovani Ricercatori 2009 (GR-2009-1606636), by the Ministry of Education, University and Research—grant PRIN 2010-2011 (2010B5B2NL), by Regione Lazio, LAZIO INNOVA (85-2017-15095) and by the Italian Regenerative Medicine Infrastructure (IRMI), Cluster ALISEI (CTN01\_00177\_888744).

**Acknowledgments:** We acknowledge Gustavo Mostoslavsky from Boston University School of Medicine, for providing the STEMCCA plasmid. We thank Massimiliano Gubello for the English editing and proofreading. We want also to thank the reviewers for their insightful comments in previous version of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
