Reciprocal communication between FAPs and muscle cells via distinct extracellular vesicle miRNAs in muscle regeneration

Significance

This study reports a pathway of reciprocal communication between fibroadipogenic progenitors (FAPs) [the major source of intramuscular adipose tissue (IMAT)] and muscle cells, which occurs after muscle injury. This pathway involves secretion of extracellular vesicles (EVs) from FAPs containing miR-127-3p, which is taken up by muscle stem cell and facilitates myogenesis, while muscle cells release EVs that contain miR-206-3p and miR-27a/b-3p, which are taken up by FAPs and repress adipogenesis. Together, this reduces IMAT accumulation and promotes healthy repair of injured muscle. This pathway may also provide a potential therapeutic for muscle injury and play a role in muscle-fat crosstalk in obesity.

Abstract

Muscle regeneration is a complex process relying on precise teamwork between multiple cell types, including muscle stem cells (MuSCs) and fibroadipogenic progenitors (FAPs). FAPs are also the main source of intramuscular adipose tissue (IMAT). Muscles without FAPs exhibit decreased IMAT infiltration but also deficient muscle regeneration, indicating the importance of FAPs in the repair process. Here, we demonstrate the presence of bidirectional crosstalk between FAPs and MuSCs via their secretion of extracellular vesicles (EVs) containing distinct clusters of miRNAs that is crucial for normal muscle regeneration. Thus, after acute muscle injury, there is activation of FAPs leading to a transient rise in IMAT. These FAPs also release EVs enriched with a selected group of miRNAs, a number of which come from an imprinted region on chromosome 12. The most abundant of these is miR-127-3p, which targets the sphingosine-1-phosphate receptor S1pr3 and activates myogenesis. Indeed, intramuscular injection of EVs from immortalized FAPs speeds regeneration of injured muscle. In late stages of muscle repair, in a feedback loop, MuSCs and their derived myoblasts/myotubes secrete EVs enriched in miR-206-3p and miR-27a/b-3p. The miRNAs repress FAP adipogenesis, allowing full muscle regeneration. Together, the reciprocal communication between FAPs and muscle cells via miRNAs in their secreted EVs plays a critical role in limiting IMAT infiltration while stimulating muscle regeneration, hence providing an important mechanism for skeletal muscle repair and homeostasis.

Skeletal muscle plays crucial roles in locomotion, basic physiological functions like respiration, and whole-body metabolism. Even in adults, skeletal muscle maintains its ability to regenerate due to the continued presence of muscle stem cells (MuSCs), also called satellite cells (1, 2). Following muscle injury, quiescent MuSCs are activated, proliferate, differentiate, and fuse into myotubes to repair the damaged muscle fibers (3–5). This process involves many other cell types, including vascular cells, immune cells, fibroblasts, and fibroadipogenic progenitors (FAPs) (6–8), which collaboratively provide a critical niche for MuSC differentiation and effective muscle regeneration.

FAPs are a class of interstitial mesenchymal stem cells in muscle that have the potential to differentiate into both adipocytes and fibroblasts (9, 10). FAPs are activated after skeletal muscle injury and play roles in facilitating clearance of necrotic debris and promoting myogenesis of MuSCs (7, 10, 11). Following severe muscle injury, FAP proliferation and differentiation also leads to accumulation of intramuscular adipose tissue (IMAT). In healthy mice, this IMAT is cleared as muscle is repaired (12). In MuSC-deficient mice, infiltration of IMAT is greatly increased, suggesting negative feedback between the activated MuSCs and regulation on adipogenic differentiation of FAPs during muscle regeneration (12). How MuSCs communicate with FAPs remains largely unknown.

Extracellular vesicles (EVs), including exosomes, can serve as paracrine mediators in cell-to-cell communication (13, 14), as well as indicators of physiological and pathological processes (15, 16). The small EVs/exosomes are 50- to 200-nm vesicles surrounded by a lipid bilayer and contain a wide range of constituents, including many RNA species, proteins, lipids, and metabolites (14). Various components of this cargo including the miRNAs can be transferred from the EVs into recipient cells, where they can act to modulate the recipient’s biological response. Indeed, multiple studies have shown that miRNAs in circulating EVs not only serve as biomarkers, but through their uptake in other cells also help regulate metabolism, affect the progression of a wide range of diseases, and modify tissue homeostasis, including skeletal muscle function (15, 17–20). Given their important role in cellular communication, EVs/exosomes are also being considered as a means to deliver therapeutic miRNAs.

In the present study, we demonstrate that after acute muscle injury, there is activation of FAPs, which in addition to giving rise to IMAT, release EVs that are enriched in several miRNAs encoded by an imprinted cluster of miRNAs on chromosome 12. The most abundant of these miRNAs is miR-127-3p, which can target S1pr3 in MuSCs and help promote myogenesis and muscle regeneration. In return, MuSCs and their derived myoblasts and myotubes release EVs that are enriched in miR-27a-3p, miR-27b-3p, and miR-206-3p that can inhibit adipogenic differentiation of FAPs by repressing Pparg (a master regulator of adipogenesis) and other regulators of adipogenesis, including C-Met and Runx1, thereby reducing the amount of IMAT infiltration. Thus, FAPs and muscle cells form a unique niche in which these two important cell types reciprocally communicate through their specific EV-miRNA secretion, and this plays an essential role in limiting accumulation of IMAT while promoting the repair of injured muscle. Importantly, mice given intramuscular injections of EVs from immortalized FAP (iFAP) show improved ability to repair damaged muscle than mice receiving vehicle control, suggesting that this pathway may provide a potential therapeutic in muscle regeneration.

Results

Depletion of Platelet-Derived Growth Factor Receptor Alpha (Pdgfrα⁺) Cells Reduces IMAT Infiltration but Delays Muscle Regeneration.

To explore the interaction between FAPs and MuSCs in muscle regeneration, BaCl₂ was injected into the tibialis anterior (TA) muscle of 8- to 10-wk-old wild-type (WT) male C57Bl/6J mice to produce acute skeletal muscle injury, after which we followed the repair process for 28 d. During the first 3 days post-injury (Dpi), hematoxylin and eosin (H & E) staining showed the injured muscle along with immune cell infiltration and some clearance of the damaged muscle fibers (Fig. 1A). By Dpi5, some newly formed muscle fibers began to appear. At this stage, there was also appearance of adipocytes between the injured muscle fibers, and this peaked around Dpi7, as shown by perilipin1 (Plin1) immunostaining and quantitation of fat area in TA muscle cross-sections (Fig. 1 A and B). During this process, paired box 7 (Pax7, a MuSC marker gene), myogenic differentiation 1 (Myod1), myogenin (Myog) [two transcription factors regulating myogenesis] and Pdgfrα, which partly suggests FAP action, were all up-regulated, peaking on Dpi3 and indicating the rapid response of both MuSCs and FAPs to muscle injury (Fig. 1 C–F). The mRNA levels of Plin1, a marker of mature adipocytes and infiltration of IMAT, demonstrated a more gradual increase during this early stage of muscle regeneration (Fig. 1G). However, by day 28, as the muscle regeneration progressed, the IMAT gradually disappeared, as indicated by disappearance of Plin1 immunostaining between the fibers of the TA muscle (Fig. 1H). This dynamic appearance and disappearance of IMAT during the recovery of the injured muscle suggests some form of communication coordinating this repair process.

Fig. 1.

Depletion of Pdgfra⁺ cells reduces IMAT infiltration but delays muscle regeneration. (A) H & E staining and Plin1 immunofluorescence staining on TA muscle cross sections from 8- to 10-wk-old WT mice on days 0, 1, 3, 5, and 7 post injury. Dpi: days post injury. (Scale bar = 50 μm.) (B) Quantitation of adipocyte area per whole TA muscle cross-sections on Dpi 0, 1, 3, 5, and 7, measured by ImageJ. (C–G) mRNA expression of myogenic genes (Pax7, Myod1, and Myog) and adipogenic genes (Pdgfra and Plin1) in TA muscle from mice on 0, 1, 3, 5, and 7 Dpi, measured by q-PCR. (H) H & E staining and Plin1 immunofluorescence staining of TA muscle cross-sections from WT mice on 28 Dpi. (Scale bar = 50 μm.) (I) H & E staining of TA muscle cross-sections from Ctrl and Pdgfrα⁺ cell^null mice generated as shown in Fig. S1A. (Scale bar = 50 μm.) (J) Plin1 staining on TA cross-sections from Ctrl and Pdgfrα⁺ cell^null mice. (Scale bar = 50 μm.) (K) Quantification of IMAT area in whole TA muscle cross-sections from Ctrl and Pdgfrα⁺ cell^null mice. (L) The ratio of regenerating myofibers containing two or more centralized nuclei per field of TA cross-section shown in panel I. (M) Distribution of myofiber size in Ctrl and Pdgfrα⁺ cell^null mice TA muscles as shown in I. More than 120 fibers were quantified per mouse. N ≥ 3, data represent mean ± SEM. Student’s t test, *P < 0.05, **P < 0.01 ***P < 0.005.

FAPs are a specialized form of multipotent Pdgfrα⁺ stem cells that reside between the muscle fibers (9). To determine the function of FAPs in the skeletal muscle regeneration process, we injected diphtheria toxin (DT) into mice carrying a Pdgfrα promoter-driven Cre^+/− inducible DTR (iDTR^+/−) (21), such that Pdgfrα-promoter driven Cre mediated excision of a STOP cassette renders the FAPs sensitive to DT. For these experiments, intramuscular DT injection was performed 15 d prior to creation of TA muscle injury, and Pdgfrα-Cre negative iDTR^+/− mice, also injected with DT, were used as control (SI Appendix, Fig. S1A). TA muscles were then collected on Dpi7, a time point at which there is normally significant IMAT infiltration.

In the control mice, there was clear immunostaining of Pdgfrα⁺ cells localized between muscle fibers, and this was reduced by more than 80% in the Pdgfrα-Cre^+/−; iDTR^+/− mice, demonstrating a successful depletion of most Pdgfrα⁺ cells (SI Appendix, Fig. S1 B and C). Consistent with the loss of Pdgfrα⁺ cells, IMAT in TA muscle of the Pdgfrα-Cre^+/−; iDTR^+/− mice was markedly reduced 7 Dpi (Fig. 1 I and J). Indeed, Plin1⁺ IMAT cell number and area in each TA cross section was reduced by >90% in Pdgfrα-Cre^+/−; iDTR^+/− mice compared to controls, demonstrating that Pdgfrα⁺ cells are the major source of IMAT observed during muscle regeneration (Fig. 1K and SI Appendix, Fig. S1D). H & E staining of TA muscle on Dpi7 also revealed that the proportion of muscle fibers with more than two centralized nuclei was significantly lower in TA muscle from Pdgfrα-Cre^+/−; iDTR^+/− mice compared to controls indicating that, in these mice, there were reduced fusion of differentiated satellite cells and formation of new muscle fibers resulting in delayed repair of the damaged muscle (Fig. 1 I and L). Consistent with this, there was an increased percentage of small diameter muscle fibers (300 to 600 μm²) and a reduced percentage of larger muscle fibers (900 to 1,800 μm²) in the Pdgfrα-Cre^+/−; iDTR^+/− mice compared to controls (Fig. 1M). Indeed, average muscle fiber size was reduced by ~50% in TA from Pdgfrα-Cre^+/−; iDTR^+/− mice (SI Appendix, Fig. S1E), indicating the significant delay in muscle regeneration due to Pdgfrα⁺ cell deletion. Thus, loss of FAPs not only reduces IMAT infiltration it also affects function and ability of MuSCs to contribute to skeletal muscle regeneration.

Reciprocal Communication between FAPs and MuSCs.

To more directly explore the communication between MuSCs and FAPs, we isolated these two cells from both WT and tdTomato mice using flow cytometry (10, 22) (SI Appendix, Fig. S2A). As expected, red fluorescence was expressed in the FAPs and MuSCs from tdTomato mice (tdFAPs and tdMuSCs) but not WT mice (SI Appendix, Fig. S2B). When cultured in vitro, the isolated MuSCs and FAPs differentiated into myotubes and adipocytes, respectively, as reflected by positive staining in myosin heavy chain (MyHC) for the former and by oil red O staining for the latter (SI Appendix, Fig. S2 C and D). TdMuSCs were then cultured alone or mixed co-culture in the same wells with an equal number of MuSCs or FAPs from WT mice. After 7 d in growth media, the cells were stained for MyHC. Among the three culture conditions, tdMuSCs co-cultured with FAPs formed the largest myotubes (Fig. 2A). This was also associated with a 2.5-fold increase in MyHC⁺ cells and a ~fourfold increase in the MyHC⁺ area (Fig. 2 A and B and SI Appendix, Fig. S3A). On the other hand, mixed co-culture of FAPs with an equal number of tdMuSCs impaired the ability of the FAPs to differentiate into adipocytes in growth media, as reflected by a dramatically lower proportion of Pparg⁺ cells in the co-culture with tdMuSCs compared with other two groups (Fig. 2 C and D). By contrast, co-culture of nonlabeled FAPs with tdTomato-labeled FAPs significantly increased the ability of the FAPs to undergo adipogenesis, as indicated by a threefold increase in percentage of Pparg⁺ cells compared to controls, although differentiation may, in part, be affected by the FAPs density (Fig. 2 C and D). Likewise, co-culture of unlabeled FAPs from WT mice with tdFAPs in adipogenic induction media resulted in an increased number of FAPs forming adipocytes, while co-culture of FAPs with tdMuSCs resulted in decreased adipogenesis of the FAPs (SI Appendix, Fig. S3 B and C).

Fig. 2.

Reciprocal communication between FAPs and MuSCs. (A) tdTomato red fluorescence, MyHC (Green), and DAPI (Blue) staining in MuSCs from tdTomato mice, either cultured alone (tdMuSC), or co-cultured with an equal number of MuSCs from WT mice (tdMuSC-MuSC), or co-cultured with an equal number of FAPs from WT mice (tdMuSC-FAP) in growth medium for 7 d. (Scale bar = 100 μm.) (B) Quantification of MyHC⁺ cells as a percentage of total cells from views shown in A, one image per individual repeat. MyHC⁺ area analysis by ImageJ is shown in SI Appendix, Fig. S3A. (C) tdTomato red fluorescence, Pparg (Green), and DAPI (Blue) staining in FAPs from WT mice, either cultured alone (FAP), or co-cultured with an equal number of FAPs from tdTomato mice (FAP-tdFAP), or co-cultured with an equal number of MuSCs from tdTomato mice (FAP-tdMuSC) in growth medium for 7 d. (Scale bar = 50 μm.) See SI Appendix, Fig. S3 B and C for Pparg staining and quantification of FAPs in different groups after 2 d of adipogenic inducement. (D) Quantification of the percentage of Pparg⁺ cells per view shown in C, two images per individual repeat. (E) Schematic figures of transwell culture system, separating the top and bottom wells with a 0.4 μm high-pore-density membrane. (F) MyHC (Green) and DAPI (Blue) staining of MuSCs cultured in the bottom well, with MuSCs (MuSC-musc) or FAPs (MuSC-fap) in the top well for 7 d. (Scale bar = 100 μm.) (G) Quantification of the percentage of MyHC⁺ cells per view showed in F, one image per individual repeat. (H) Pparg (Green) and DAPI (Blue) staining of FAPs cultured in the bottom well, with FAPs (FAP-fap) or MuSCs (FAP-musc) in the top well for 7 d. (Scale bar = 100 μm.) (I) Quantification of the percentage of Pparg⁺ cells per view showed in H, one to two images per individual repeat. N ≥ 3, data represent mean ± SEM. Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.005.

To determine whether cross talk between FAPs and MuSCs was through a secreted factor and to avoid any effects of cell density on adipogenesis and myogenesis, we performed co-culture experiments using a transwell system in which secreted factors that can pass through a 0.4-µ filter might influence cells in the paired transwell (Fig. 2E). We found that presence of FAPs, compared with MuSCs, in the top (donor) well resulted in increased MyHC⁺ cell ratio of MuSCs cultured in the bottom (recipient) well, indicating that the FAPs released factors that promote myogenesis (Fig. 2 F and G). On the other hand, when FAPs were cultured as the recipients in the bottom well and FAPs or MuSCs as donors in the top well, we found that, compared with FAP donors, MuSCs released factors that inhibited adipogenesis of FAPs, as reflected by a 50% reduction in the proportion of Pparg⁺ FAPs in the bottom well (Fig. 2 H and I). Taken together, these results demonstrate reciprocal communication between FAPs and MuSCs, which occurs via secreted factors and which positively affects the differentiation of the MuSCs and negatively affects the differentiation of the FAPs.

EVs Released from FAPs and Muscle Cells Are Involved in Their Crosstalk.

Recent studies have shown that, in addition to soluble factors, exosomes and EVs can serve important mediators of intercellular communication in maintenance of muscle function and homeostasis (18, 20, 23). To determine whether EVs could mediate the reciprocal communication between FAPs and MuSCs, we isolated the EVs from culture media of FAPs and muscle cells (MCs: including MuSCs and their derived myoblasts and myotubes), using sequential ultracentrifugation (hereafter referred to as EV pellets: EVPs) or size exclusion chromatography (hereafter referred to an EVs) (see Materials and Methods for details). Nanoparticle tracing analysis and electron micrograph of CD63 immunostaining revealed that the EVs/exosomes of MCs had an average diameter 198 nm and those of FAPs had an average diameter of 184 nm (Fig. 3 A–C). The EVPs and EVs were both positive for the exosomal markers CD63, HSP70, and TSG101 and negative for cytoskeleton marker such as β-ACTIN and secretory pathway-related protein CALNEXIN (Fig. 3D and SI Appendix, Fig. S4A). To facilitate having sufficient quantities of FAPs for in vitro EV transfer studies, primary isolated FAPs were immortalized with SV40 retrovirus. These immortalized FAPs (iFAPs) maintained their ability to differentiate into adipocytes after treatment with adipogenic induction cocktail (SI Appendix, Fig. S4B) and released EVs similar to those from primary FAPs in size, morphology and marker expression, as confirmed by nanotracker analysis, electron microscopy and staining with anti-CD63 antibody (SI Appendix, Fig. S4 C and D). To follow their transfer, the iFAP-derived EVs (iFAP-EVs) and the C2C12-derived EVs were labeled with PKH26, a tracking dye with red fluorescence (24). Then PKH26 labeled EVs, as well as the PKH26 dye itself, were purified using size exclusion to remove the non-bound or aggregated dye. When the labeled iFAP-EVs were added to a ~40% confluent culture of C2C12 myoblasts, within 1 d, red fluorescence could be observed in the C2C12 cells incubated with the labeled EVs, but not in C2C12 cells incubated with only the PKH26 dye, indicating uptake of iFAP-EVs by the muscle cells (Fig. 3E). Likewise, when EVs from C2C12 myoblasts were labeled with PKH26 and incubated with iFAPs for 24 h, red fluorescence was apparent in the iFAPs, but not in the iFAPs incubated with PKH26 only, indicating uptake of C2C12-derived EVs by the iFAPs (Fig. 3E).

Fig. 3.

EVs/Exosomes released from FAPs and muscle cells are involved in their cross-talk. (A and B) Size distribution of EVs isolated from muscle cells and FAPs measured by nanoparticle tracking analysis. (C) Electron microscopy of EVs secreted from MCs and FAPs. (Scale bar = 200 nm.) (D) Western blot of EV-related protein markers HSP70 and CD63 in EVPs/exosome from MCs and FAPs. (E) EVs/exosomes secreted by C2C12 muscle cell line or iFAPs were labeled with PKH26, and then were re-purified using size exclusion and used to treat iFAPs or C2C12 for 24 h, respectively. PKH26 fluorescence in iFAPs and C2C12s showing EVs uptake. EVs: cell = ~500:1. (Scale bar = 20 μm.) (F) MyHC staining of MuSCs treated with PBS, FAP-EVPs, or MC-EVPs for 5 d, EVs: cell = ~1,000:1. (Scale bar = 100 μm.) (G) Quantification of MyHC⁺ cells in F, one image per individual repeat. (H) Pparg staining in FAPs treated with PBS, FAP-EVPs, or MC-EVPs for 5 d. EVs: cell = ~1,000:1. (Scale bar = 100 μm.) (I) Quantification of Pparg⁺ cells in H, three to four images per individual repeat. N ≥ 3, data represent mean ± SEM. Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.005.

More importantly, we found that there was a strong effect to promote myogenesis of the MuSCs as demonstrated by an increase of the proportion and area of MyHC⁺ cells following treatment with FAP-EVPs, compared to either MuSCs incubated with EVPs from MCs or PBS as a control (Fig. 3 F and G and SI Appendix, Fig. S4E). This occurred without a change in the total number of MuSCs (SI Appendix, Fig. S4F). By contrast, adding MC-EVPs to FAPs in culture resulted in an obvious decrease in the percentage of Pparg⁺ cells when compared to controls treated with PBS or EVPs from FAPs, with no significant difference in total FAP number (Fig. 3 H and I and SI Appendix, Fig. S4G). Taken together, these data indicate that EVs released from FAPs crosstalk and contribute to the regulation of differentiation of recipient muscle cells, while EVs from muscle cells contribute to the regulation of adipogenic differentiation of FAPs.

Distinct miRNAs Enriched in FAPs, Muscle Cells, and Their EVPs.

MiRNAs have been shown to be packaged into EVs/exosomes, and, in this way, allow a secreting cell to regulate gene expression in a recipient cell (18, 19, 25, 26). To determine the miRNAs in the EVs secreted by FAPs and muscle cells that might contribute to their reciprocal communication, we performed small RNA sequencing using FAP- and MC-EV pellets. A total 639 of miRNAs were detected in the EVPs, of which 114 were significantly enriched (P < 0.05) in FAP-EVPs and 78 were enriched in MC-EVPs (Fig. 4A). The top 20 miRNAs distinctly enriched in the FAP- or MC- EVPs ranked by their relative abundance are shown in Fig. 4B. To better define the potential functional effects of miRNAs secreted in EVPs from FAPs and muscle cells, we determined the potential targets of these top differentially enriched miRNAs using DIANA-miRPath v3.0 (27) (Fig. 4 C and D). Among the top regulated pathways targeted by miRNAs enriched in FAP-EVPs were Hippo signaling, amino acid degradation, and adherens junction-related signaling (the cell junction whose cytoplasmic face is linked to the actin cytoskeleton), whereas miRNAs enriched in MC-EVPs were predicted to target genes involved in TGF-beta signaling, axon guidance, and transcriptional mis-regulation.

Fig. 4.

Distinct miRNAs in FAPs and muscle cells and their EVPs. (A) Venn diagram of miRNAs identified in FAP- and MC- EVPs. (B) Log 2 fold change of miRNAs enriched in FAP-EVPs compared to MC-EVPs is shown in heatmap, which is arranged by the log 2 of relative reads of miRNAs in FAP-EVPs (red bars) or in MC-EVPs (green bars). MiRNAs indicated by the red stars are the candidates contributing the cross-talk between the two cell types. (C and D) Pathway analysis of top 20 miRNAs enriched in FAP-EVPs and MC-EVPs targeted genes. (E) Chromosome location of top 20 miRNAs enriched in FAP-EVPs. Different fill patterns represent different miRNA clusters. (F) Q-PCR analysis of miR-127-3p in FAPs and muscle cells. (G) Q-PCR analysis of miR-206-3p, miR-27b-3p, and miR-27a-3p in FAPs and muscle cells. N ≥ 3, data represent mean ± SEM. Student’s t test, *P < 0.05, ***P < 0.005.

Interestingly, of the top 20 miRNAs enriched in FAP-EVPs, 14 miRNAs (70%) were encoded in 4 miRNA clusters located on mouse chromosome (Chr) 12, all of which are within the imprinted Dlk1-Rtl1-Dio3 region, suggesting activation of this Chr12 region in the activated FAPs (Fig. 4E, shown in red bars, filling pattern represent different miRNA clusters, SI Appendix, Table S2). As FAPs underwent adipogenic differentiation, most of these miRNAs were significantly decreased, suggesting inactivation of this region (SI Appendix, Fig. S5A). For example, miR-127-3p, the top enriched miRNA in FAP-EVPs, was 2.5-fold more highly expressed in FAPs than in muscle cells, and following FAP adipogenesis, decreased to around 10% of its initial level (Fig. 4F and SI Appendix, Fig. S5B). Thus, this cluster of miRNAs, including miR-127-3p, which is highly expressed in FAPs and enriched in FAP-EVPs appears to be a potential indicator of FAP identity and its multiple mesenchymal precursor fate.

The top 20 miRNAs abundant in MC-EVPs (Bottom half of Fig. 4B) were more dispersed throughout the genome, although several were clustered on Chr1, 2, 8, 13, 18, and X, with many near genes involved in myogenesis and muscle function (28, 29) (SI Appendix, Fig. S5C, shown in green, SI Appendix, Table S3). Most of the miRNAs enriched in muscle cell EVPs are known to be highly expressed in muscle cells, including miR-206-3p, miR-1a-3p, miR-1b-5p, miR-133a-3p, and miR-486a-5p/3p. The most enriched miRNA in EVPs from muscle cells, miR-206-3p, was more than 200-fold higher in MCs than FAPs and more than eightfold higher in MC-EVPs than in FAP EVPs (Fig. 4 B and G). Likewise, miR-27a-3p and miR-27b-3p were significantly higher in muscle cells and MC-EVPs than in FAPs and their EVPs (Fig. 4 B and G). Thus, these miRNAs were enriched in both muscle cells and their EVPs and could serve as markers of muscle cells.

MiR-127-3p Enriched in the FAP and iFAP EVs Facilitates Muscle Cell Myogenesis by Repressing S1pr3.

MiR-127-3p, the top enriched miRNA in FAP-EVPs, has been shown to promote myogenesis by repressing expression of G-protein coupled receptor S1pr3 (SI Appendix, Table S1A) (30). miR-127-3p could be confirmed to be inside the EVs, since treatment of EVs from iFAP with RNase A alone did not degrade the miRNA, while treatment of the EVs with Triton X-100 followed by RNase A allowed its degradation, proving that miR-127-3p is protected by the lipid bilayer of EVs (Fig. 5A). Treatment of primary muscle cells with FAP-EVs for 24 h resulted in a twofold increase of mature miR-127-3p as compared to the PBS-treated controls; this occurred with no increase in levels of precursor miR-127, consistent with transfer of the mature miRNA from the FAP-EVs to the muscle cells, rather than up-regulation of gene expression (Fig. 5B). Importantly, this was associated with a 20% decrease in mRNA expression of the miRNA-127-3p target gene S1pr3 (Fig. 5C).

Fig. 5.

MiR-127-3p enriched in the FAP and iFAP EVs facilitates muscle cell myogenesis by repressing S1pr3. (A) Q-PCR analysis of miR-127-3p in equal aliquots iFAP-EVs without any treatment or with treatment of RNase A or TritonX-100 plus RNase A. (B and C) Fold change of precursor and mature miR-127-3p as well as its target gene S1pr3 expression in MCs 24 h after FAP-EVPs treatment compared to PBS treated Ctrl. EVs: cell = ~500:1. (D) Fold change of miR-127-3p in iFAPs cultured in 6-well plate transfected with scramble ctrl or miR-127-3p inhibitor, three repeats per group. (E) Workflow of PBS ctrl, iFAP-EVs (Scr-EVs), and EVs from miR-127-3p knockdown iFAPs (KD-EVs) treatment on C2C12 cultured in a 12-well plate. The red arrow means PBS or EVs treatment. EVs: cell = ~500:1. (F) Fold change of miR-127-3p in recipient C2C12 treated with PBS, Scr-EVs, and KD-EVs. (G and H) miR-127-3p target gene S1pr3 expression at both mRNA and protein level in the recipient C2C12 as workflow showed. (I) Q-PCR analysis of MyoD and MyoG in same recipient C2C12 as E. N ≥ 3, data represent mean ± SEM. Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.001.

To explore the role of miR-127-3p from iFAP-EVs in myogenic regulation, we transfected iFAPs with a miR-127-3p inhibitor, i.e., an anti-miRNA-127-3p single-stranded oligonucleotide that can bind to pri-miRNA to inhibit Drosha activity or pre-miRNA to inhibit Dicer cleavage, as well as bind to and inhibit the function of the mature miRNA (31). This resulted in a >98% decrease in endogenous miR-127-3p in the iFAPs (Fig. 5D). C2C12 cells were then treated with PBS (Ctrl), EVs from iFAP transfected with scrambled miRNA (Scr-EVs), or EVs from iFAP transfected with the miR-127-3p inhibitor (KD-EVs) (Fig. 5E). In the recipient C2C12 cells, miR-127-3p was increased ~threefold in the Scr-EVs treated group as compared to the PBS or KD-EVs treated groups. As a result, S1pr3, the target gene of miR-127-3p, was reduced ~60% at the mRNA level and ~40% at the protein level in C2C12 cells treated with Scr-EVs, whereas there was no significant change S1PR3 protein and ~40% decrease of S1pr3 mRNA in C2C12 cells incubated with KD-EVs (Fig. 5 G and H). Moreover, the myogenesis-related genes MyoD and MyoG were significantly higher in the C2C12 cells treated with iFAP EVs as compared to the other two groups. Thus, loss of miR-127-3p in iFAP EVs reduces the positive effect of these EVs on myogenic regulation (Fig. 5I).

miR-27a/b-3p Inhibits FAP Adipogenesis by Targeting Pparg.

Among the miRNAs most enriched in the muscle cell EVPs were miR-206-3p, miR-27a-3p, and miR-27b-3p (Fig. 4B). MiR-206-3p has been shown to suppress adipogenesis of 3T3L1 cells and FAPs by targeting c-Met and/or Runx1 (32, 33), whereas miR-27a-3p and miR-27b-3p (miRs-27a/b-3p) are predicted to target the 3′UTR of Pparg (SI Appendix, Table S1C). Supporting the potential of transfer of these miRNAs from muscle to FAP, we found that FAPs treated with MC-EVPs showed increased amounts of mature miR-206-3p, miR-27b-3p, and miR-27a-3p with no change in the amounts of their precursor miRNAs (Fig. 5A and SI Appendix, Fig. S6 A and B). Again, the RNase protection assay proved that miR-206-3p was largely inside the EVs, i.e., protected from degradation in the absence of Triton treatment. miR-27a/b-3p were also partly, but not completely, protected from RNase A degradation, suggesting that some of these miRNAs were on the surface or in the membrane of the EVs, or that the membrane of C2C12-EVs is not that intact, and therefore the RNase A can enter the EVs and degrade some of the miRNAs. (SI Appendix, Fig. S6C).

To determine the potential role of miR-27a/b-3p enriched MC-EVs to their target gene Pparg in FAPs, we characterized the differentiation of the FAPs in vitro using a standard adipogenic induction cocktail without and with treatment of miR-27a-3p or miR-27b-3p mimics (chemically modified double-stranded RNA molecules designed to mimic endogenous miR-27a-3p or miR-27b-3p). In control FAPs, at days 2 to 6 post induction, there was a decrease in expression of endogenous miR-27a/b-3p, and this was accompanied by a time-dependent 30-fold increase in expression of Pparg mRNA and a fourfold increase in the level of PPARg protein by days 4 to 6 (SI Appendix, Fig. S6D and Fig. 6 F and G). This was followed by a time-dependent 400-fold increase of perilipin 1 (Plin1) at the mRNA level and 45-fold increase of PLIN1 at the protein level, reflecting adipocyte differentiation (SI Appendix, Fig. S6 E–G). Transfected FAPs with mimics of miR-27a-3p or miR-27b-3p prior to induction of adipogenic differentiation resulted 30- to 34-fold increase levels of miR-27a-3p and miR-27b-3p on day 4 of induction as compared to cells transfected with scrambled control (SI Appendix, Fig. S6H). Transfection with miR-27a/b mimics also resulted in a suppression of Pparg expression by around 35 to 45% at the mRNA level and by 60 to 69% at the protein level when compared to cells transfected with control sequences (Fig. 6 B–D). FAPs transfected with these two miRNA mimics also showed a reduction in adipogenesis with an ~50% decrease in the level of PLIN1 protein (Fig. 6 E and F) and a 30% reduction in lipid accumulation compared to controls as determined by oil red O staining 4 to 6 d after induction of differentiation (Fig. 6 G and H and SI Appendix, Fig. S6I). Taken together, these results indicate that the miR-27a/b-3p, which are enriched in MC-EVs, can be taken up and inhibit FAP adipogenesis by targeting Pparg.

Fig. 6.

MiR-27a/b-3p inhibit FAP adipogenesis by targeting Pparg. (A) Fold change of mature miR-27a-3p and miR-27b-3p expression in recipient FAPs 24 h after MC-EVPs treatment compared to PBS treated Ctrl. EVs: cell = ~500:1. (B and C) Q-PCR analysis of Pparg mRNA expression and western-blot of PPARg protein in FAPs transfected with scrambled Ctrl (Scr) or miR-27a/miR-27b-3p mimics. Cells were transfected on day −2, differentiated on day 0, and harvested on day 2. (D) Quantification of PPARg protein levels in C by ImageJ. (E) Western-blot showing PLIN1 protein level in FAPs transfected with Scr, miR-27a-3p, or miR-27b-3p mimics. Cells were transfected on day −2, differentiated on day 0, and harvested on day 4. (F) Quantification of PLIN1 protein level in E using Image J. (G) Oil red O staining of FAPs transfected with Scr and miR-27a/miR-27b-3p mimics after adipogenic differentiation for 6 d. (H) Quantification of oil red O extracted from differentiated FAPs. N ≥ 3, data represent mean ± SEM. Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.005.

miRNAs and EVs from iFAPs Promote Muscle Repair, while miRNAs in MC-EVs Serve as Negative Regulators of IMAT Infiltration during Muscle Regeneration.

To further explore the role of EVs from FAPs and muscle cells and their miRNAs on muscle regeneration, we treated injured TA muscle with EVs from iFAPs and assessed miRNAs enriched in MCs, as well as their target gene expression pattern during muscle regeneration. To this end, we injected the iFAP-EVs into injured muscles on days 1, 3, and 5 post-injury and assessed expression of Pax7 (a marker of MuSCs), Myod and Myog (markers of myogenesis), and eMyHC (embryonic MyHC), a marker of newly formed muscle, by western blotting (Fig. 7 A and B). This revealed ~twofold increases of both PAX7 and eMYHC in the iFAP-EV injected group compared to the vehicle controls at day 7 post-injury, with mild down-regulation of MYOD and trend up-regulation of MYOG protein levels, suggesting an increase in new muscle fibers developing from the MuSCs (Fig. 7 B–D and SI Appendix, Fig. S7 A and B). Using eMyHC and Laminin co-immunostaining of TA muscle on Dpi7, we also found a 1.5-fold increase in the average size of eMyHC⁺ muscle fibers in EV-treated groups compared to controls, indicating an increase in formation of new muscle fibers (Fig. 7 E and F). Thus, iFAP-EVs can promote muscle repair after injury.

Fig. 7.

EVs from iFAPs promote muscle repair, while miRNAs enriched in MC-EVs serve as negative regulators of IMAT infiltration during muscle regeneration. (A) Workflow of EV-treatment on injured TA muscles. Five mice per group. (B) Total protein level of PAX7, eMYHC, MYOD, and MYOG were measured by western-blot from TA muscles without or with iFAP-EVs treatment. (C and D) Quantification of Pax7 and eMyHC protein expression in B by ImageJ. (E) eMyHC and Laminin co-immunostaining of TA muscles from vehicle and iFAP-EVs injected groups on Dpi7. (Scale bar = 50 μm.) (F) Quantification of average size of eMyHC⁺ muscle fibers shown in E, more than 30 newly formatted muscle fibers were measured per mouse. (G and H) Q-PCR analysis of miR-27a/b-3p and their target gene Pparg expression in uninjured TA muscles and in TA muscles harvested on Dpi 3, 5, and 7. (I) Q-PCR analysis of miR-206-3p and its target gene c-Met expression in TA muscles as same time point as in G and H. N ≥ 3, data represent mean ± SEM. Student’s t test, *P < 0.05.

To explore the relationship between miR-27a/b-3p and miR-206-3p and their target genes in vivo during muscle regeneration, we conducted Q-PCR analysis on TA muscles collected under uninjured conditions or on days 3, 5, and 7 post-injury. As expected in the whole TA muscle, which is a mixed tissue, expression of Pparg mRNA reached a peak on day 3 post-injury, consistent with stimulation of IMAT formation. At this time, the expression of miRNA-27a/b-3p reached their nadir at about 30% of control levels (Fig. 7 G and H). However, by Dpi3, miR-206-3p levels begin to increase, reaching more than 30-fold increase at Dpi 5-7. The latter was associated with a decrease in the miR-206-3p target gene C-Met (Fig. 7I). Thus, as the regeneration progressed, MuSCs proliferated and differentiated into myoblasts and myotubes, leading to increased levels of muscle cells enriched miRNAs [miR-27a/b-3p and miR-206-3p] (Fig. 4B), which could be secreted in MC-EVs and taken up by FAPs to prevent IMAT over-accumulation, thus creating the cycle of repair.

Discussion

Precise spatiotemporal crosstalk between cells is essential for skeletal muscle homeostasis. MuSCs are indispensable for skeletal muscle regeneration and present throughout life (12). In this study, we show that in the process of repair of muscle injury, there is reciprocal communication between FAPs and muscle cells, and that this occurs via miRNAs in EVs/exosomes and plays a crucial role in control of the repair process (summarized in Fig. 8). Thus, shortly after injury, at an early stage of muscle regeneration, there is activation of FAPs, which is associated with an increase in release of EVs from FAPs. These EVs are enriched in miR-127-3p, which helps promote MuSCs differentiation and muscle regeneration by repressing S1pr3. As repair progresses, the MuSCs and their derived muscle cells also release EVs. These are enriched in several miRNAs, including miR-206-3p and miRs-27a/b-3p, which repress adipogenesis of FAPs by targeting c-Met, Runx1 (both effects of miR-206-3p) and Pparg (the target of miR-27a/b-3p), thus reducing IMAT infiltration and allowing full muscle regeneration.

Fig. 8.

Reciprocal crosstalk via distinct EV-miRNAs promotes muscle healing and reduces intramuscular fat. Following injury, FAP-derived intramuscular adipocytes infiltrate injured muscles. Activated FAPs secrete EVs containing miR-127-3p, which facilitate MuSC myogenesis. In return muscle cells secrete EV- miR-206-3p and miR-27a/b-3p repressing FAP adipogenesis.

Many recent studies have shown release of EVs/exosomes containing miRNAs is an important mechanism in control of metabolism (19), as well as in intramuscular regulation and cellular crosstalk (18, 20). Though it is not completely understood how EV/exosome-miRNAs are taken up and their cargo is dispersed in recipient cells, both clathrin-dependent and clathrin-independent endocytosis pathways, including caveolin-mediated uptake, macropinocytosis, phagocytosis, and lipid raft–mediated internalization have been suggested to play a role (34–36). Here we show that, whatever the mechanism, at each step of muscle regeneration, this is a highly regulated process both at the cellular and EV levels. Thus, we find miR-127-3p is twofold higher in FAPs than muscle cells, it is even more differentially enriched in FAP-EVs. When these FAP-EVs are incubated with muscle cells, there is uptake and a significant increase in levels of miR-127-3p in the muscle cells, consistent with transfer of this miRNA from the EVs to the MuSCs. This transfer then leads to a decrease in expression of S1pr3 mRNA and protein. Decreases in S1pr3 have been shown to promote myogenesis by releasing MuSCs from the quiescent stage and regulating cell cycle of MuSCs (30, 37). When production of miR-127-3p in FAPs is blocked by using an inhibitor, there is reduction of promoting myogenesis of FAP-EVs, suggesting FAP-EVs facilitate myogenic differentiation, at least partly, through miR-127-3p. This process of miRNA regulation may also be disrupted in diseases other than injury. For example, the myogenic capacity of MuSCs and muscle function are known to be reduced with aging and in patients with Duchenne muscular dystrophy (DMD), and in both cases, miR-127-3p has been shown to be involved (38). Likewise, in the mdx mouse model of DMD, overexpression of miR-127-3p ameliorate the dystrophic phenotype (30, 37). Maintaining skeletal muscle homeostasis and ability of regeneration is a complex process and involves not only MuSCs and FAPs, but also vascular cells and inflammatory cells. Here we show that EVs released from iFAPs speed up muscle regeneration by effects of the miRNAs they contain to promote myogenic differentiation. These findings along with our in vitro and in vivo study suggest that transferring of miR-127-3p via EVs/exosomes from FAPs or iFAPs into muscle of aged or DMD individuals could have therapeutic potential by improving satellite cell differentiation and muscle function.

Importantly, the cell crosstalk between FAPs and muscle cells is bidirectional. Thus, while FAP-EVs transfer important regulatory miRNAs to muscle, muscle cells release EVs that contain miRNAs that act to repress FAP adipogenesis. The existence of this EV crosstalk is clear by our demonstration that treatment of FAPs with MC-EVs represses expression of Pparg and adipose differentiation. Small RNA sequencing reveals that MC-EVs are significantly enriched with miR-206-3p, miR-27a-3p, and miR-27b-3p, all of which can have repressive effects on adipogenesis. Indeed, miR-206-3p is almost 250-fold and miR-27a/7b-3p are 4- to 5-fold more highly expressed in muscle cells than in FAPs. Following muscle injury, expression of miR-27a/b-3p is reduced, but as MuSCs proliferate, differentiate, and muscle repair continues, expression of miR-27a/b-3p gradually return. Likewise, miR-206-3p increases about 25-fold at 5 Dpi, and as expected, expression of the target genes of this miRNA shows an opposite pattern. This explains the early infiltration of IMAT and the reversal of this process later during muscle regeneration, when miRNAs in EVs released from MCs contribute to repression of IMAT infiltration. In addition to muscle injury, IMAT accumulation is increased in obesity, aging, type 2 diabetes, Duchene muscular dystrophy, and diseases associated with reduced muscle activity, and this likely creates a vicious cycle by contributing to muscle dysfunction and insulin resistance (39–42). In these disorders, natural or synthetic EVs containing miR-206-3p and miR-27a/b-3p may serve as a potential tool to reducing IMAT infiltration and improving muscle homeostasis.

Previous studies have shown that EVs from different cell types have unique protein, lipid, and miRNA cargo, reflecting differences in cellular content of each of these species, differences in EV formation and sorting of miRNAs for EV secretion (26, 43, 44). All of these have potential for transfer and involvement in regulation of cells at a distance. In the case of muscle cells and FAPs, the differential enrichment of miRNAs is striking and could serve as markers for FAP versus muscle cell activity in different physiological and disease states. For both cell types, the EV-enriched miRNAs appear to reflect both cellular function and selective activation. For the muscle cells, several of the miRNAs enriched in the EVs have been linked to myogenesis (45, 46), and while these are dispersed throughout the genome in terms of chromosomal location, 30% were in two regions on chromosome 8, and an additional 40% were in clusters on chromosomes 1 and 18. The miRNAs enriched in the FAP-EVs showed an even more interesting distribution, with 70% of the EV-enriched miRNAs coming an imprinted region on chr12 that includes the paternally expressed protein-coding genes Dlk1, Rtl1/Mart1, and Dio3 and the maternally expressed nonprotein coding genes Gtl2 (MEG3 in human), Rian (MEG8 in human), anti-Rtl1, and Mirg (47, 48). This region also contains the largest clusters of miRNAs in the human and mouse genome, with 54 and 61 miRNAs, respectively (49, 50). Previous studies have shown that this locus is highly regulated by methylation and involved in stem cell function, embryonic development, tissue growth, and differentiation (50, 51). In previous studies, we have shown the expression of these genes and miRNAs are markedly down-regulated in cells lacking both the insulin and IGF-1 receptors (52). In this study, we find in that as FAPs undergo adipogenic differentiation, most of the miRNAs enriched in FAP-EVs significantly decrease based on published miRNA analysis (33), suggesting inactivation of these same regions as the FAP transition from the multipotent mesenchymal stem cell state to a differentiated adipose cell state. Further understanding regulation of this miRNA cluster may not only provide insights into the function of EV/exosomal miRNAs in muscle repair but also to developing approaches to reducing IMAT infiltration and improving muscle homeostasis.

Two limitations of our study should be kept in mind. First, as noted above, there are multiple types EVs released by cells, and while we have focused on small EVs with the characteristics of exosomes, it is possible that we are dealing with a mixed population of vesicles. Second, while FAPs express high levels of Pdgfrα, making this a useful marker of this cell type (6), single cell transcriptional and translational analyses show that Pdgfrα can also be detected in other cell types, such as endothelial cells (6, 8). Thus, when using Pdgfrα promotor-driven Cre with iDTR mice to induce cell death, while FAPs are the largest population of Pdgfrα⁺ cells, we cannot rule out that some of the effects ascribed to FAPs on IMAT infiltration and muscle regeneration may come in part from other Pdgfrα⁺ cell types.

In summary, our in vitro plus in vivo studies demonstrate a highly coordinated and reciprocal crosstalk between FAPs and muscle cells following muscle injury. In this process, FAPs release EVs that contain miRNAs which, when taken up by MuSCs, promoting myogenesis and muscle regeneration. The MuSCs and the differentiating myoblasts and myotubes then release EVs that contain miRNAs which limit conversion of FAP to IMAT and thereby promote return of normal muscle morphology and function. Understanding this process of crosstalk involving miRNAs in EVs may help identify therapeutic targets with the potential of modifying or mimicking this process that ultimately improve muscle function and reduce IMAT infiltration not only in muscle injury, but in other diseases, including DMD, aging, and obesity.

Materials and Methods

Animal Care and Use.

All the animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee at both the China Agricultural University (Approval number: SKLAB-2011-04-03) and the Joslin Diabetes Center (IACUC: 2020-02). Mouse information is listed in SI Appendix. All the genotyping primers are listed in SI Appendix, Table S4.

FAP Deletion and Muscle Injury.

Single dose of 1 ng/g DT (Sigma-Aldrich: D0564) was injected into the TA muscle of Ctrl mice and Pdgfrα-Cre; iDTR mice to induce the FAP deletion. To induce injury, 50 μL of 1.2% BaCl₂ (Sigma: 10361-37-2) was injected into the TA muscle.

H & E, Immunofluorescence (IF), and Imaging on Fixed Samples.

Tissues or cells were fixed in 4% paraformaldehyde (PFA). Paraffin cross sections of the TA muscle were processed for H & E (Sigma) staining and IF staining. The primary antibodies are listed in SI Appendix. The primary antibody was detected using the corresponding IgG (H+L) antibodies conjugated to Alexa Fluor 594 or 488 (Invitrogen, 1:500).

MuSC and FAP Isolation by Fluorescence-Activated Cell Sorting.

The skeletal muscles of both hindlimbs were dissected and gently shredded. These were digested by collagenase type 2 (2.5 U/mL, Sigma: C6885/9001-12-1) and dispase II (2.4 U/mL, Sigma: Roche 49420780012). Blood cells were removed by ACK (Ammonium-Chloride-Potassium) lysis buffer (Thermo-Fisher: A1049201). The cells were incubated with antibodies for 30 to 60 min. Antibodies are listed in SI Appendix. Cell sorting was conducted by either the Beijing University or Joslin Diabetes Center Flow Cytometry Cores. MuSCs were identified as CD31⁻, CD45⁻, Sca-1⁻, α7-integrin⁺ cells. FAPs were identified as CD31⁻, CD45⁻, Sca-1⁺, α7-integrin⁻ cells. Procedure details are attached in SI Appendix.

FAPs Immortalization.

FAPs under 70 to 80% confluent were infected by retrovirus SV40 T-antigen followed by the selection with 2.5 µg/mL puromycin dihydrochloride (Thermo Fisher Scientific, A1113803) for 4 d.

FAPs/iFAPs Transfection.

When FAPs or iFAPs were 70 to 80% confluent, scramble (Dharmacon: CN-001000-01-05), miR-27a-3p mimics (Dharmacon: CTM-547903), and miR-27b-3p mimics (Dharmacon: CTM-547902) or miR-127-3p inhibitors (Horizondiscovery: IH-310397-08-0002) were transfected by Lipofectamine 3000 Reagent (Thermo Fisher Scientific: L3000-008), according to the dose and protocol provided.

Oil Red O Staining and Quantification.

Differentiated FAPs were fixed with 4% PFA and stained by the 0.5% oil red O (Sigma, 0-0625) in isopropanol for 1.5 h. For quantification of differentiated cells, stained Oil Red O was dissolved in isopropanol and its absorbance was measured at 518 nm.

EV/Exosome Purification.

EVPs were isolated by ultracentrifugation (Beckman: L7-65) as previously described. The procedure is attached in SI Appendix. EVs were collected using size exclusion via a 70-nm Gen2 Column (Izon Science, qEVoriginal, ICO-70).

RNase Sensitivity Assay.

EVs were isolated using size exclusion from iFAP or C2C12 culture media and divided into nine equal portions of 250 µL each. Three of these were used as controls without any treatment, three for RNase A (0.2 µL, 20 mg/mL, BioLabs: T3018-2) only treatment group, and three were treated with 1% Triton X-100 (Sigma: T9284) followed by 0.2 µL of RNase A. 1 µL of RNase inhibitor (PROMEGA: n2518 40 U/µL,) was added to block the function of RNase A, after which we conducted normal RNA extraction and Q-PCR for miR-127-3p in iFAP-EVs, and miR-206-3p and miR-27a/b-3p in C2C12-EVs.

EVs Intramuscular Injection.

EVs were collected using a 70-nm Gen2 Column (Izon Science, qEVoriginal, ICO-70) and quantitated using a NanoSight (NS300). For in vivo experiments, around 1.2 × 10⁹ EVs were separately intramuscular injected into the TA muscle of each mouse on days 1, 3, and 5 post-injury.

SmallRNA Sequencing and Pathway Analysis.

Total RNAs were isolated by the EV RNA extraction kit (Beijing microread: E1520R), and then, small RNA fractionations were identified and reverse transcription by NEXTflexTM Small RNA-Seq Kit V3 (Illumina Compatible). Sequencing was conducted on Agient2000, and the reads were mapped against miRBase to identify known miRNAs. Analysis details are attached in SI Appendix. The top 20 miRNAs enriched in the muscle cells and FAPs target genes pathway analysis was determined by DIANA-miRPath v3.0 (27). The identification of miRNAs clusters was based on data from miRbase.

RNA, Protein Extraction and Q-PCR, Western Blotting Analysis Are Attached in SI Appendix.

Quantification and statistical analysis.

All quantification results in the current study were conducted by ImageJ software. All data were presented as mean ± SEM. The Student t test was performed for comparison of two groups.

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLSX)

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information. All significant differential expressed miRNAs are listed in Dataset S1, miRNAs data source has been deposited in Zenodo, which is accessible with link: https://zenodo.org/records/10685908 (53).