2023 年 70 巻 2 号 p. 161-171
Humoral factors that are secreted from skeletal muscles can regulate bone metabolism and contribute to muscle-bone relationships. Although extracellular vesicles (EVs) play important roles in physiological and pathophysiological processes, the roles of EVs that are secreted from skeletal muscles in bone repair have remained unclear. In the present study, we investigated the effects of the local administration of muscle cell-derived EVs on bone repair in control and streptozotocin-treated diabetic female mice. Muscle cell-derived EVs (Myo-EVs) were isolated from the conditioned medium from mouse muscle C2C12 cells by ultracentrifugation, after which Myo-EVs and gelatin hydrogel sheets were transplanted on femoral bone defect sites. The local administration of Myo-EVs significantly improved delayed bone repair that was induced by the diabetic state in mice 9 days after surgery. Moreover, this administration significantly enhanced the ratio of bone volume to tissue volume at the damaged sites 9 days after surgery in the control mice. Moreover, the local administration of Myo-EVs significantly blunted the number of Osterix-positive cells that were suppressed by the diabetic state at the damage sites after bone injury in mice. Additionally, Myo-EVs significantly blunted the mRNA levels of Osterix and alkaline phosphatase (ALP), and ALP activity was suppressed by advanced glycation end product 3 in ST2 cells that were treated with bone morphogenetic protein-2. In conclusion, we have shown for the first time that the local administration of Myo-EVs improves delayed bone repair that is induced by the diabetic state through an enhancement of osteoblastic differentiation in female mice.
THE CLINICAL RELEVANCE of sarcopenia and osteoporosis suggests relationships between skeletal muscles and bone [1, 2]. Humoral factors secreted from skeletal muscles and myokines can regulate bone metabolism [3]. Furthermore, extracellular vesicles (EVs) are secreted from cells in various types of tissues and circulate in bodily fluids. EVs play important roles in physiological and pathophysiological processes by affecting remote organs through the delivery of EV membrane proteins or their contents, such as RNAs and proteins. However, the details of the roles of EVs in muscle-bone relationships remain unknown.
Guescini et al. isolated presumptive skeletal muscle-derived EVs from human plasma by using immunoaffinity capture beads with an antibody against α-sarcoglycan, which is a protein that is highly expressed in skeletal muscles, and they showed that EVs are highly enriched for skeletal muscle-specific miR-206 and represent approximately 1–5% of the total EV population [4]. Reports from human studies have shown that the amount of serum EVs is increased after exercise, and the expression of the same miRNAs is elevated in serum EVs and muscle tissues in response to exercise, thus suggesting that EVs that are derived from skeletal muscles may have systemic effects [5, 6]. Recently, Vechetti et al. reported that EVs derived from mechanical overload-induced skeletal muscles promote adipose tissue lipolysis through the delivery of miR-1 in mice [7]. However, the physiological actions of EVs from skeletal muscles in vivo have not been properly elucidated due to the limitations of the experimental techniques. To clarify the involvement of EVs in muscle-bone relationships, we investigated the effects of muscle-derived EVs (Myo-EVs) on bone metabolism in vitro. We have previously reported that Myo-EVs secreted from mouse muscle C2C12 cells suppress osteoclast formation and enhance osteoblastic differentiation of preosteoblasts in mouse cells [8, 9]. Moreover, we have demonstrated that several miRNAs are related to the suppressive effects of Myo-EVs on osteoclast formation. In addition, we have recently shown that fluid flow shear stress to mouse muscle cells enhances the suppressive effects of Myo-EVs on osteoclast formation [8, 10]. These findings suggest that skeletal muscle-derived EVs physiologically regulate bone metabolism.
It has been reported that the local administration of EVs secreted from human bone marrow-derived mesenchymal stem cells promotes femoral fracture healing and bone repair of calvarial defects in mice [11, 12], although the mechanisms of this bone repair by EVs are unclear. Ikebuchi et al. reported that the local administration of EVs secreted from mouse osteoclast-like Raw264.7 cells promotes the bone repair of calvarial defects via the receptor activation of nuclear factor-κB (RANK) expressed on the surface of EVs in mice [13]. Furthermore, Kang et al. recently reported that EVs secreted from mouse macrophage-like cells modulate bone repair [14]. The bone repair process is divided into inflammatory, restoration and remodelling phases, and macrophages, chondrogenesis and angiogenesis are involved in the restoration phase, which is partially related to inflammation [15, 16]. However, there have been no reports about the effects of skeletal muscle-derived EVs on bone repair.
It is well known that a diabetic state induces osteoporosis and delayed bone repair after fractures [17, 18]. The impairment of the mobilization of bone marrow stem cells, angiogenesis, the accumulation of macrophages, chondrogenesis and osteoblastic bone formation have been reported to be related to diabetic delayed bone repair [18, 19]. Moreover, we have previously reported that plasminogen activator inhibitor-1 (which is an adipocytokine) is involved in delayed bone repair induced by the diabetic state in female mice [19, 20]. Thus, in the present study, we examined the effects of the local administration of Myo-EVs on bone repair after femoral bone injury by using diabetic mice. We have previously reported that the diabetic state induced by streptozotocin (STZ) decreases bone mineral density and osteogenic gene expression in the tibia, which was more pronounced in female mice than in male mice [21]. Therefore, we used female mice in the present study.
Anti-Osterix antibodies were purchased from Abcam (Cambridge, UK). STZ was purchased from Sigma-Aldrich (St. Louis, MO). In addition, fetal bovine serum (FBS; Sigma-Aldrich) was heat inactivated at 56°C for 30 min. EVs contained in the FBS were depleted via ultracentrifugation at 130,000 × g for 16 hr at 4°C. Moreover, DAPI solution was purchased from Dojindo (Kumamoto, Japan). TRIzol reagent was purchased from Invitrogen (Carlsbad, CA).
Isolation of Myo-EVsWe collected Myo-EVs from the conditioned medium (CM) of mouse muscle C2C12 cells (American Type Culture Collection, Manassas, VA) by using ultracentrifugation, as previously described [9, 22]. Briefly, C2C12 cells were seeded on 10 cm culture plates coated with collagen (Cellmatrix type I-C, Nitta Gelatin, Inc., Osaka, Japan) and cultured in high-glucose Dulbecco’s modified Eagle’s medium (DMEM; FUJIFILM Wako, Osaka, Japan) supplemented with 10% FBS and 1% penicillin/streptomycin until confluent. Afterwards, C2C12 cells were cultured in fresh DMEM supplemented with 10% EV-depleted FBS for 48 hr at 37°C. The CM of C2C12 cells was centrifuged at 3,000 × g for 5 min to remove dead cells and cell debris, filtered with a Stericup filter unit with a 0.22-μm PVDF filter (Millipore, Bedford, MA, USA) and ultracentrifuged at 130,000 × g for 70 min at 4°C with a Himac CP80NX system (HITACHI, Tokyo, Japan) to isolate a pellet of Myo-EVs. Previously, we validated that the size of the isolated particles was almost less than 200 nm via a NanoSight LM10 V-HS system (Malvern Instruments, Malvern, UK), and the particles were positive for CD9 and CD81, which are well-defined EV markers, thus suggesting that particles rich in typical small EVs can be obtained by the previously mentioned method [9]. The pellet of Myo-EVs was suspended in fresh PBS and stored at –80°C until use. The protein amount of the Myo-EVs was measured with the use of a BCA Protein Assay Kit (Pierce, Rockford, IL).
AnimalsC57BL/6J female mice were purchased from CLEA Japan (Tokyo, Japan). Diabetes was randomly induced in 8- to 10-week-old mice by using intraperitoneal injections of the pancreatic β-cell toxin STZ (50 mg/kg body weight in saline), as previously described [20]. Control mice were injected with saline. Blood glucose levels of the mice were measured at 2 weeks after the first injection by using a blood glucose meter (Glutest Ace, Sanwa Kagaku Kenkyusyo, Nagoya, Japan). Mice with blood glucose levels higher than 300 mg/dL were determined to be diabetic. In addition, bone defect surgery was performed 2 weeks after the induction of diabetes. The mice were fed normal food and water, which were available freely to ingest. The mice were maintained in a specific pathogen-free environment on a 12-hour light-dark cycle at 23 ± 1°C and 55 ± 10% humidity (3 or 4 per cage). We assumed that the activities of the mice in each group were similar. All of the experiments were performed according to the guidelines of the National Institutes of Health and the institutional rules for the use and care of laboratory animals at Kindai University. The experiments were approved by the animal ethics committee of Kindai University (approval number: KAME-31-051).
Bone defect model and treatment with Myo-EVsA bone defect was induced in the right femurs of the mice according to a previously reported method (with some modifications) [19, 23]. Briefly, an incision (5 mm in length) was made in the anterior skin of the central femur of the right leg under anesthetization by using 2% isoflurane. After splitting the muscle, the surface of the femoral bone was exposed, and a hole was made in the femur by using a drill with a diameter of 0.8 mm. The hole was irrigated with saline to prevent thermal necrosis of the margins. For the local administration of Myo-EVs, a disc-shaped cationized gelatin (CG) hydrogel sheet (which is a sustained release carrier of Myo-EVs) with a diameter of 1.5 mm was infused with Myo-EVs (10 μg) or saline and placed into the femoral defect [24]. The incised skin was then sutured in a sterile manner, and anesthesia was discontinued. The ratio of bone volume to tissue volume (BV/TV) within the bone defect region was calculated to evaluate new bone formation after bone injury.
Quantitative computed tomography (qCT) analysisThe mice were anesthetized with 2% isoflurane for qCT analysis. The femurs of the mice were scanned by using Cosmo Scan GX II (Rigaku Corp, Tokyo, Japan) according to the manufacturer’s instructions (tube voltage: 90 kV; tube current: 88 μA and isotropic voxel size: 25 μm). The area of the bone defects in the femur was quantified by using Analyze 14.0 (Analyze Direct, Inc., Overland Park, KS).
Histological analysisThe femurs of the mice were removed under anesthesia with 2% isoflurane at 4 days after the femoral bone defect surgery. The removed femur was fixed in 4% paraformaldehyde, demineralized in 22.5% formic acid and 340 mM sodium citrate solution and embedded in paraffin for immunostaining. Four-micrometer-thick sections of the femur were incubated with the anti-Osterix antibody at a dilution of 1:200, followed by incubation with the secondary antibody conjugated with horseradish peroxidase. Positive signals were visualized by using a tyramide signal amplification system (PerkinElmer, Waltham, MS, USA), and the sections were counterstained with DAPI solution for nuclear staining. The section was photographed via a fluorescence microscope, and the number of Osterix-positive cells per 0.1 mm2 in the microscopic fields of the defective area of the femur was quantified.
Cell cultureST2 cells (RIKEN, Tsukuba, Japan), which represent a mouse mesenchymal cell line, were maintained in low-glucose Dulbecco’s modified Eagle’s medium (DMEM, FUJIFILM Wako, 5.5 mM glucose) with 10% FBS and 1% penicillin–streptomycin. ST2 cells were cultured until they were confluent and then treated with or without Myo-EVs, BMP-2 (FUJIFILM Wako), advanced glycation end product 3 (AGE3) and high glucose (25 mM). AGE3 was prepared as previously described [25]. Moreover, alkaline phosphatase (ALP) activity was measured with the Lab Assay ALP kit (FUJIFILM Wako) according to the manufacturer’s instructions.
Quantitative real-time PCRThe femurs of the mice were removed under anesthesia with 2% isoflurane at 4 days after femoral bone defect surgery and frozen in liquid nitrogen. Total RNA was isolated from a 5 mm piece of femur containing the defective site by using the TRIzol reagent. The total RNA of the ST2 cells was extracted with a Nucleo Spin® RNA Plus kit (Takara Bio, Shiga, Japan). A reverse transcription reaction of the extracted RNA from the femur or ST2 cells was performed by using the Prime Script RT reagent Kit with gDNA eraser (Takara Bio). The incorporation of SYBR Green into double-stranded DNA, which was performed by using an Applied Biosystems Step One PlusTM Real-Time PCR System (Applied Biosystems, Carlsbad, CA), was assessed via quantitative real-time PCR. The PCR primers are listed in Table 1. The specific mRNA amplification of the target was determined as the Ct value, followed by normalization to the mRNA level of β-Actin or glyceraldehyde-3-phosphate dehydrogenase (Gapdh).
| Gene | Primer sequence | |
|---|---|---|
| Runx2 | Forward | 5'-AAATGCCTCCGCTGTTATGAA-3' |
| Reverse | 5'-GCTCCGGCCCACAAATCT-3' | |
| Osterix | Forward | 5'-AGCGACCACTTGAGCAAACAT-3' |
| Reverse | 5'-GCGGCTGATTGGCTTCTTCT-3' | |
| ALP | Forward | 5'-ATCTTTGGTCTGGCTCCCATG-3' |
| Reverse | 5'-TTTCCCGTTCACCGTCCAC-3' | |
| Osteocalcin | Forward | 5'-CCTGAGTCTGACAAAGCCTTCA-3' |
| Reverse | 5'-GCCGGAGTCTGTTCACTACCTT-3' | |
| Col 1 | Forward | 5'-AACCCTGCCCGCACATG-3' |
| Reverse | 5'-CAGACGGCTGAGTAGGGAACA-3' | |
| IGF-1 | Forward | 5'-CAAGCCCACAGGCTATGGC-3' |
| Reverse | 5'-TCTGAGTCTTGGGCATGTCAG-3' | |
| FGF2 | Forward | 5'-GCGACCCACACGTCAAACTA-3' |
| Reverse | 5'-CCGTCCATCTTCCTTCATAGC-3' | |
| BMP-2 | Forward | 5'-GGGACCCGCTGTCTTCTAGT-3' |
| Reverse | 5'-TCAACTCAAATTCGCTGAGGAC-3' | |
| Aggrecan | Forward | 5'-CCTGCTACTTCATCGACCCC-3' |
| Reverse | 5'-AGATGCTGTTGACTCGAACCT-3' | |
| Col 2 | Forward | 5'-CCTCCGTCTACTGTCCACTGA-3' |
| Reverse | 5'-ATTGGAGCCCTGGATGAGCA-3' | |
| Col 10 | Forward | 5'-TGGGTAGGCCTGTATAAAGAACGG-3' |
| Reverse | 5'-CATGGGAGCCACTAGGAATCCTGAGA-3' | |
| PPARγ | Forward | 5'-GGAAAGACAACGGACAAATCAC-3' |
| Reverse | 5'-TACGGATCGAAACTGGCAC-3' | |
| aP-2 | Forward | 5'- ATCACCGCAGACGACAGGA -3' |
| Reverse | 5'- CTCATGCCCTTTCATAAACT -3' | |
| β-Actin | Forward | 5'-AATCGTGCGTGACATTAAG-3' |
| Reverse | 5'-GAAGGAAGGCTGGAAGAG-3' | |
| Gapdh | Forward | 5'-AGGTCGGTGTGAACGGATTTG-3' |
| Reverse | 5'-GGGGTCGTTGATGGCAACA-3' |
Runx2 = Runt-related transcription factor 2; ALP = alkaline phosphatase; Col 1 = type I collagen; IGF-1 = Insulin like growth factor-1; FGF2 = Fibroblast growth factor 2; BMP-2 = Bone morphogenetic proteins-2; Col 2 = type II collagen; Col 10 = type X collagen; PPARγ = peroxisome proliferator-activated receptor γ; aP-2 = adipocyte protein-2, Gapdh = glyceraldehyde-3-phosphate dehydrogenase.
All of the data are expressed as the mean ± the standard error of the mean (SEM). One-way ANOVA (followed by the Tukey–Kramer post hoc test) was used to perform multiple comparisons. The significance level was set at p < 0.05. GraphPad PRISM 6 software (La Jolla, CA) was used for all of the statistical analyses.
The body weights of the female mice were significantly decreased by STZ treatment at 2 weeks after the first injection of STZ (Fig. 1A), and the blood glucose levels of the mice were markedly elevated by STZ treatment before bone defects were elicited, thus indicating that STZ induced a diabetic state in the mice (Fig. 1B). The bone defect areas were analysed with qCT at 0, 7 and 9 days after surgery. As shown in Fig. 1C, the bone defects were not filled in any group at 7 days after surgery. Although the bone defect areas were almost filled in the control mice at 9 days after surgery, STZ treatment significantly suppressed bone repair after femoral bone injury, and the local administration of Myo-EVs significantly blunted delayed bone repair at 9 days after surgery in diabetic mice (Fig. 1C). In addition, the BV/TV of the new bone in the bone defect region was significantly increased by the local administration of Myo-EVs at 9 days after surgery in both control and diabetic mice (Fig. 1D and 1E).
Effects of the local administration of Myo-EVs on bone repair after femoral bone injury in female mice.
(A) Growth curve during experiments in control and streptozotocin (STZ)-treated mice with or without the local administration of Myo-EVs. ##: p < 0.01, versus control group with saline. **: p < 0.01 versus the control group treated with Myo-EVs. Cont, control. (B) Blood glucose levels in control and STZ-treated mice with or without the local administration of Myo-EVs at 0 and 9 days after surgery. (C) Quantification of the bone defect area at the damaged sites in control and STZ-treated mice with or without the local administration of Myo-EVs at 0, 7 and 9 days after surgery. The data represent the mean ± SEM of 8–10 mice per group. N.S., not significant. (D) Representative three-dimensional images at the damaged sites of femurs at 9 days after surgery, as assessed by the qCT images of each group. The scale bar indicates 1 mm. (E) BV/TV (%) of the mineralized bone formed in the hole region at 9 days after surgery was analysed by using qCT. Data represent the mean ± SEM of 8–10 mice per group.
Histological analyses of the damaged sites in the femurs were performed to clarify the mechanism by which the local administration of Myo-EVs blunts delayed bone repair in diabetic mice. The number of Osterix-positive cells at the damaged sites of the femurs was significantly decreased in diabetic mice at 4 days after surgery, and the local administration of Myo-EVs significantly blunted the number of Osterix-positive cells that were suppressed by the diabetic state (Fig. 2A and 2B). Subsequently, we examined the effects of the local administration of Myo-EVs on the mRNA levels of osteogenic genes (Runx2, Osterix, ALP, osteocalcin and type 1 collagen [Col 1]), insulin-like growth factor-1 (IGF-1), fibroblast growth factor 2 (FGF2) and bone morphogenetic protein-2 (BMP-2) at the damaged sites at 4 days after surgery in mice. The mRNA levels of Runx2, Osterix, ALP, osteocalcin and Col 1 were significantly reduced in diabetic mice without Myo-EV treatment at 4 days after surgery (Fig. 2C). The local administration of Myo-EVs significantly blunted osteocalcin mRNA levels that were suppressed by the diabetic state at 4 days after surgery, and it tended to blunt the mRNA levels of Runx2, Osterix, ALP and Col 1 that were suppressed by the diabetic state (without any statistically significant differences) (Fig. 2C). Moreover, the local administration of Myo-EVs significantly increased the mRNA levels of IGF-1 and FGF2 (but not BMP-2) in control mice at 4 days after surgery, and it tended to increase the mRNA levels of IGF-1 and FGF2 in diabetic mice (without statistically significant differences) (Fig. 2C).
Effects of the local administration of Myo-EVs on the expression of osteogenesis at damaged sites at 4 days after femoral bone injury.
(A) Representative microphotographs of Osterix-positive cells at the damaged sites at 4 days after surgery in female mice with or without STZ treatment and/or the local administration of Myo-EVs. Scale bars indicate 100 μm and 50 μm (enlarged images). The enlarged images show the square area in the low-magnification figure shown above. The dotted line indicates the boundary with the cortical bone. (B) Quantification of the number of Osterix-positive cells per 0.1 mm2 of the microscopic field at the damaged sites at 4 days after surgery. Data represent the mean ± SEM of 5–6 mice per group. (C) The mRNA levels of osteogenic genes (Runx2, Osterix, ALP, osteocalcin and Col 1), IGF-1, FGF2 and BMP-2 at the damaged sites at 4 days after surgery. Data are expressed as relative to β-Actin mRNA values and the mean ± SEM. N = 5–7 in each group. N.S., not significant. Ct, cortical bone.
We examined the effects of the local administration of Myo-EVs on the mRNA levels of chondrogenic genes (aggrecan, type 2 collagen [Col 2] and type 10 collagen [Col 10]), as well as adipogenic genes (peroxisome proliferator activated receptor γ [PPARγ] and adaptor protein 2 [aP2]), at the damaged sites at 4 days after surgery to assess the effects of Myo-EVs on chondrogenesis and adipogenesis during bone repair after a femoral bone defect induction in mice. The local administration of Myo-EVs significantly increased the mRNA levels of Col 2 and Col 10 in control mice at 4 days after surgery, although it did not affect the mRNA levels of chondrogenic genes in the diabetic mice (Fig. 3A). Furthermore, the local administration of Myo-EVs did not affect the mRNA levels of adipogenic genes in diabetic mice at 4 days after surgery, although it significantly suppressed PPARγ mRNA levels in the control mice (Fig. 3B).
Effects of the local administration of Myo-EVs on the expression of chondrogenic and adipogenic genes at damaged sites after femoral bone injury.
The mRNA levels of (A) chondrogenic genes (aggrecan, Col 2 and Col 10) and (B) adipogenic genes (PPARγ and aP2) at the damaged sites at 4 days after surgery in female mice with or without STZ treatment and/or the local administration of Myo-EVs. Data are expressed relative to β-Actin mRNA values and as the mean ± SEM. N = 5–10 in each group. N.S., not significant.
A previous study demonstrated that AGE levels in the serum, liver and kidney are significantly elevated in STZ-induced diabetic model rats [26]. Finally, we examined the effects of Myo-EVs on AGEs and high glucose-suppressed osteoblastic differentiation in mouse mesenchymal ST2 cells to investigate the mechanisms by which Myo-EVs blunt osteoblast differentiation that is suppressed by the diabetic state at damaged sites after femoral bone injury in mice. Myo-EVs significantly elevated Osterix and ALP mRNA levels that were enhanced by BMP-2 in ST2 cells (Fig. 4A). Moreover, Myo-EVs significantly blunted the mRNA levels of Osterix and ALP that were suppressed by AGE3 in the presence of BMP-2 treatment, and Myo-EVs significantly blunted Osterix mRNA levels that were suppressed by high-glucose conditions in the presence of BMP-2 (Fig. 4A). In addition, Myo-EVs significantly blunted AGE3-induced ALP activity suppression in the presence of BMP-2 (Fig. 4B).
Effect of Myo-EVs on osteoblastic differentiation of mouse mesenchymal cells.
ST2 cells were cultured with or without 10 μg/mL Myo-EVs, 200 ng/mL BMP-2, 200 μg/mL AGE3 or high-glucose conditions for 72 h. (A) The mRNA levels of Osterix, ALP, osteocalcin and Col 1 in ST2 cells. Data are expressed relative to glyceraldehyde-3-phosphate dehydrogenase (Gapdh) mRNA values and as the mean ± SEM of 4 experiments in each group. (B) ALP activity of ST2 cells. Data are expressed as the mean ± SEM of 4 experiments in each group. N.S., not significant. **p < 0.01 versus the ST2 cells treated with only BMP-2.
In the present study, we demonstrated that the local administration of Myo-EVs improves delayed bone repair induced by a diabetic state after a femoral bone defect in female mice. Moreover, the local administration of Myo-EVs significantly enhanced BV/TV at the damaged sites after a femoral bone defect in control mice. These findings indicate that the local administration of skeletal muscle cell-derived EVs may be an effective tool for the enhancement of bone repair and regeneration after bone injury, bone defects or fractures, which are compatible with the effectiveness of EVs that are derived from other types of cells, as has been previously reported [11-14]. Although the significance of skeletal muscle-derived EVs in circulation is still unknown at the present time, several studies have suggested the effectiveness of Myo-EVs on bone cells in vitro [9, 27]. Therefore, we speculated that skeletal muscle-derived EVs may partially participate in muscle/bone relationships and subsequently influence the bone repair process.
In our study, the local administration of Myo-EVs with CG hydrogel sheets improved delayed bone repair that was induced by the diabetic state after a femoral bone defect in female mice. Due to the fact that Myo-EVs did not affect serum glucose levels or body weight, the effects of Myo-EVs on bone repair may be due to their effects on the bone repair process, rather than due to changes in glucose metabolism and nutritional states. CG hydrogels are useful materials for the local sustained release of negatively charged components due to degradation [24]. The CG hydrogel was completely degraded at 9 days after transplantation, and Myo-EVs were considered to be continuously released in the bone defect area during the experiment and contributed to an improvement in bone repair in the present study.
Osteoblastic bone formation and the recruitment of osteoblastic cells are critical for the restoration phase during the bone repair process. In the present study, the local administration of Myo-EVs blunted the number of Osterix-positive cells and putative osteoblasts, which were decreased by the diabetic state at the damaged sites during the bone repair process after a femoral bone defect was elicited in female mice. Moreover, the local administration of Myo-EVs significantly blunted the decrease in osteocalcin expression that was induced by the diabetic state at the damaged sites, and it seemed to blunt the expression of other osteogenic genes, such as Runx2, Osterix, ALP and type I collagen, which were suppressed by the diabetic state at the damaged sites of femurs in diabetic mice (without statistical significance). These data suggest that Myo-EVs enhanced bone repair after femoral bone injury, which was partially achieved through an enhancement of osteoblastogenesis and osteoblastic differentiation at damaged sites during the bone repair process in mice. In our data, the local administration of Myo-EVs significantly enhanced the expression of IGF-1 and FGF2 at the damaged sites during the bone repair process after a femoral bone defect was induced in control mice, and it seemed to enhance the expression of IGF-1 and FGF2 at the damaged sites in diabetic mice (without statistical significance). Due to the fact that IGF-1 and FGF2 are important for osteoblastic bone formation [28, 29], Myo-EVs may partially enhance bone repair through local IGF-1 and FGF2 changes induced by bone injury.
AGEs are well known to be related to diabetic complications [30], and AGEs inhibit osteoblastic differentiation through receptors of AGEs [31]. High-glucose conditions can also suppress osteoblastic differentiation by regulating STAT3/SOCS3 signaling [32]. Thus, AGEs and high glucose may partially explain the mechanisms of delayed bone repair in the diabetic state. Therefore, we examined the effects of Myo-EVs on osteoblastic differentiation induced by BMP-2 in the presence of AGEs and high glucose by using ST2 cells. Our data showed that Myo-EVs blunted the expression of Osterix and ALP and that ALP activity was suppressed by AGE3 in the presence of BMP-2 in ST2 cells. Moreover, Myo-EVs blunted Osterix expression, which was suppressed by high glucose conditions in ST2 cells. When considering that the local administration of Myo-EVs improved delayed bone repair and osteoblastogenesis that were suppressed by the diabetic state at damaged sites after femoral bone injury in mice, Myo-EVs may contribute to the improvement of delayed bone repair in diabetic mice by cancelling the negative effects of AGEs and high glucose on osteoblastic differentiation at damaged sites in the restoration phase during the bone repair process.
It has been reported that the diabetic state influences chondrogenic and adipogenic differentiation [33, 34]. In the present study, the local administration of Myo-EVs did not affect the expression of chondrogenic genes (such as aggrecan, Col 2 and Col 10) at the damaged sites at 4 days after surgery in diabetic mice, although Myo-EVs increased the mRNA levels of Col 2 and Col 10 in control mice. Due to the fact that the mRNA levels of Col 2 and Col 10 were markedly suppressed in diabetic mice, the suppressive effect of diabetes may have been greater than the positive effect of Myo-EVs on chondrogenic differentiation. However, we cannot rule out the possibility that Myo-EVs may partially enhance bone repair through an enhancement of chondrogenesis in mice. Moreover, the local administration of Myo-EVs did not affect the expression of adipogenic genes (such as PPARγ and aP2) at the damaged sites at 4 days after surgery in diabetic mice. Taken together, these findings suggest that Myo-EVs improved delayed bone repair that is induced by the diabetic state, mainly through osteoblastic bone formation (but not through chondrogenesis and adipogenesis).
The types of Myo-EV components responsible for the effects of Myo-EVs on bone repair remained unknown in our study. Various miRNAs are included in EVs, and miRNAs can regulate gene expression via the translational inhibition of messenger RNAs. We previously analysed the miRNAs that are included in Myo-EVs and reported that the expression of some miRNAs in Myo-EVs is greater than that in mouse primary osteoblasts and bone marrow cells [8]. Among them, it has been reported that three miRNAs (miR143, miR196 and miR365) enhance osteogenic differentiation in mouse cells [35-39]. These miRNAs are abundantly included in Myo-EVs and may contribute to the effects of Myo-EVs on bone repair, although there is no specific evidence that these miRNAs in Myo-EVs are transferred to cells at the bone defect area. Further studies are necessary to clarify these issues.
We employed EVs from C2C12 myoblasts because a large number of EVs are necessary to apply Myo-EVs to mice in vivo, and C2C12 myoblasts are considered to be more efficient in isolating Myo-EVs in large-scale cell cultures than C2C12 myotubes, which are easily detached from the culture dish. Moreover, the secretion of EVs from C2C12 myotubes was less than that from C2C12 myoblasts in our previous study, although the suppression of EVs from C2C12 myoblasts and myotubes on osteoclast formation from mouse bone marrow cells was comparable in vitro [9]. Further studies will be necessary to determine the efficacy of Myo-EVs from myotubes on bone repair in vivo.
There are several limitations in this study. First, the STZ injection model might not represent type 2 diabetes. We used the STZ injection model in the present study, since the delayed bone repair after femoral bone injury by diabetic state is clear. It has been previously reported that db/db mice (leptin receptor-deficient mice) display compromised bone regeneration capacity [40]. Therefore, it might be useful to examine the effects of the local administration of Myo-EVs on bone repair after femoral bone injury using db/db mice, although the controversy exists about the usage of db/db mice for type 2 diabetic mouse model. Second, we showed the levels of osteogenic, chondrogenic or adipogenic genes in Figs. 2, 3 and 4 using only the mRNA level data, but not protein level.
In conclusion, our study showed that the local administration of Myo-EVs with CG hydrogel sheets improves delayed bone repair that was induced by the diabetic state, which was partially achieved through an enhancement of osteoblastic differentiation in female mice. These results suggest that EVs secreted from skeletal muscles may affect bone repair and bone metabolism in vivo.
We thank Mr. Katsumi Okumoto (Life Science Research Institute, Kindai University) for assistance with the histological analysis. The study was partially supported by grants from Takeda Science Foundation to Y.T., Tokyo Biochemical Research Foundation to Y.T., The Osaka Medical Research Foundation for Intractable Diseases to Y.T., the Cooperative Research Program (Joint usage/Research Center program) of the Institute for Frontier Life and Medical Sciences, Kyoto University, to Y.T., a JSPS KAKENHI Grant-in-Aid for Early Career Scientists (19K18480) and a Grant-in-Aid for Scientific Research (C: 21K09240) to Y.T., as well as a Grant-in-Aid for Scientific Research (C: 20K09514) to H.K. from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.
The authors declare no conflicts of interest.
