Bmscs loaded exosome hydrogel promotes the repair of rotator cuff injury in rats in vivo

Abstract

To explore whether the hydrogels loaded with bone marrow mesenchymal stem cell-derived exosomes can facilitate the repair and regeneration of tendons in rats with acute rotator cuff injury through regulating transforming growth factor-beta1 (TGF-β1). After the extraction and identification of mesenchymal stem cell-derived exosomes, 60 male Sprague–Dawley rats were randomly divided into five groups (n = 12 per group): control (untreated), repair-alone, GelMA (hydrogel only), BMSC-Exos (hydrogel + BMSC-Exos), and BMSC-Exos + TGF-β1 inhibited (hydrogel + BMSC-Exos + TGF-β1 inhibitor P144). An acute rotator cuff injury repair model was established in the left shoulder joint of the rats, and different treatments were administered to the rats according to the groups. Six weeks later, the rats in each group were sacrificed, and HE staining, Masson staining, Sirius Red staining, biomechanical tests, and PCR detection were carried out. BMSC-Exos were successfully isolated and characterized. Biomechanical tests showed that the BMSC-Exos group exhibited significantly higher maximum failure load and stiffness compared with the repair-alone, GelMA, and BMSC-Exos + TGF-β1 inhibited groups (all P < 0.05), reaching levels similar to the normal control. Histological scoring revealed that the BMSC-Exos group had better collagen fiber continuity, parallelism, density, and fewer inflammatory cells and blood vessels at the tendon-bone interface. Gene expression analysis demonstrated that the BMSC-Exos group significantly upregulated the mRNA levels of ColⅠ, ColⅢ, Scx, and α-SMA compared with the other experimental groups (all P < 0.05). The hydrogels loaded with mesenchymal stem cell-derived exosomes can enhance the repair and regeneration of tendons in rat shoulder cuff injuries through TGF-β1.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-40392-y.

Introduction

Rotator cuff tears (RCT) constitute a prevalent musculoskeletal disorder in clinical practice, resulting in severe pain and disability¹. This injury typically occurs during sports and daily activities. Research has indicated that the estimated prevalence of shoulder cuff tear rupture among individuals over 60 years old is at least 10%, leading to the deterioration of shoulder joint function, accompanied by pain and restricted mobility². Statistics reveal that more than 250,000 operations are conducted annually in the United States to repair torn rotator cuff tendons, with associated medical expenses exceeding $1.5 billion³. Consequently, shoulder cuff tears and related issues have garnered increasing attention from clinicians^4,5.

The tendon-bone interface is the region where tendons and ligaments interface with bone and constitutes an important connection between soft tissue and bone. The tendon-bone interface is subjected to substantial stress and is prone to injury during movement. Tendon-bone healing demands the proliferation and differentiation of various cells. The quality of tendon-bone healing directly influences the possibility of rotator cuff retear after rotator cuff repair. The tendon-bone interface after rotator cuff repair is mainly filled with scar tissue, which fails to restore the normal tissue structure⁶. The tensile strength of Sharpey fiber scar tissue at its insertion point is significantly lower than that at the normal insertion point⁷.

Currently, an increasing number of methods are being used to treat rotator cuff injuries⁸.The preferred clinical treatment approach for shoulder cuff tears is surgical tendon repair^9,10. Nevertheless, some studies have demonstrated that the recurrence rate in cases of massive tear can reach up to 94%¹¹. At the same time, surgical outcomes are also influenced by a variety of factors^12–15.Complete tendon-bone healing depends on the strength of the bone-attachment site, and the regeneration of the fibrous cartilage area is critical for the strength of the attachment site. However, due to the scarcity of blood vessels and poor blood supply in this area, the healing process is slow. Fibrous cartilage tissue usually relies on scar tissue rich in type I collagen for healing after injury¹⁶. With the advancement of regenerative medicine, mesenchymal stem cells (MSCs) have drawn escalating attention as a type of multipotent stem cell with multi-directional differentiation potential and hold promising prospects in the field of regenerative medicine tissue^17,18. However, this study did not establish an animal model for rotator cuff repair surgery to verify the therapeutic effects and mechanism of action of BMSC-Exos¹⁹. Recent evidence demonstrates that mesenchymal stem cells (MSCs) exert therapeutic effects primarily through paracrine mechanisms, wherein MSC-derived exosomes (MSC-Exos) play a pivotal role²⁰.Exosomes are nanoscale extracellular vesicles with diameters ranging from 30 to 200 nm. These vesicles encapsulate a diverse array of biomolecules, including proteins, DNA, RNA, mRNA, and microRNAs, which facilitate significant intercellular communication. Recently, mesenchymal stem cell-derived exosomes (MSC-Exos) have emerged as a promising therapeutic approach for tissue regeneration and have found extensive application in conditions such as myocardial infarction, stroke, limb ischemia, liver and kidney injury, as well as osteochondral damage^21,22.

At present, certain studies have demonstrated that mesenchymal stem cells can facilitate the repair of rotator cuff injury. They can differentiate into osteoblasts, chondrocytes and tenocytes to promote osteogenesis, collagen fiber remodeling and maturation as well as extracellular matrix enrichment. BMSCs can promote meniscus regeneration and healing, and also repair tendon injury. BMSCs-derived exosomes not only possess some functions similar to BMSCs, but also have the advantages of low immunogenicity and no tumorigenicity, which renders them to have broad prospects in the field of regenerative medicine. It has been reported that the secreted proteome of BMSCs combined with tissue engineering therapy can improve chronic massive rotator cuff tears²³. He et al. found that BMSC-Exos could effectively promote cartilage repair and extracellular matrix synthesis, and alleviate knee pain in OA rats²⁴. Therefore, we propose that BMSCS-derived exosomes can accelerate the healing process of rotator cuff injury by acting on the injured rotator cuff tissue.

Transforming growth factor-beta 1 (TGF-β1) is a critical cytokine involved in cellular activities. Previous studies have demonstrated that TGF-β1 plays a significant role in regulating tendon tear healing and promoting tendon cell proliferation²⁵. Our research group has previously shown that BMSC-Exos enriched with TGF-β1 can enhance the expression of TGF-β1 in rotator cuff tenocytes in vitro, thereby promoting cell proliferation, migration, and fibrosis activity²⁶. To comprehensively evaluate the downstream therapeutic effects mediated by this TGF-β1 signaling, we focused on a panel of gene markers integral to tendon-bone healing. Collagen type I (Col I) and type III (Col III) are the primary structural components of the tendon ECM; their expression ratio is a crucial indicator of repair maturity, with a higher Col I/Col III ratio correlating with superior mechanical strength. Scleraxis (Scx) is a tendon-specific transcription factor essential for tenocyte differentiation and tendon development. α-smooth muscle actin (α-SMA) marks the presence of myofibroblasts, cells responsible for wound contraction and ECM remodeling during fibrosis and repair. Lastly, Tenascin-C (TnC) is a glycoprotein involved in cell adhesion and migration, dynamically expressed during tissue remodeling and mechanotransduction. Monitoring the expression of these genes provides a multi-faceted assessment of the healing process, from matrix synthesis (Col I/III) and cell fate (Scx) to contractile activity (α-SMA) and matrix reorganization (TnC).This study aims to establish an acute rotator cuff repair model in rats to confirm the therapeutic effects and the role of TGF-β1 in the repair of rotator cuff injuries by BMSC-Exos. It is anticipated that this will offer a new potential therapeutic strategy for treating rotator cuff injuries and enhancing postoperative recovery.

Materials and methods

The study was approved by the Animal Experiment Ethics Committee of Chengde Medical College’s Affiliated Hospital and followed the Laboratory Animal Center guidelines. Thirty male Sprague–Dawley (SD) rats, aged 3–4 weeks, were used for extraction and identification of BMSCs.Animals were euthanized per standard protocols, then surface sterilized in 75% alcohol for 10 min. Femurs were aseptically isolated in an ultra-clean bench and placed in PBS with 1% gentamicin. The bone marrow cavity was exposed by removing the epiphyses, then flushed six times with DMEM/F12 medium containing 1% penicillin–streptomycin until bones appeared clear. The suspension was collected in a 15 mL centrifuge tube. After discarding the supernatant, cells were resuspended and cultured at 37 °C with 5% CO₂. Media changes were partial after 24 h, complete after 48 h, and every three days thereafter. Cells from passages 3 to 6 were used for experiments.

Identification of bone marrow mesenchymal stem cells

MSCs are cells that express CD73, CD90, CD44H and CD105 but do not manifest CD14, CD11b, CD19, or CD79, CD34, CD45, and human leukocyte antigen D-related (HLA-DR). Initially, the phenotype of BMSCs was determined via flow cytometry. After the BMSCs were cultivated to the third generation, the cells were digested, enumerated, and re-suspended with an adequate amount of PBS to attain a cell concentration of 1 × 10⁷ cells/mL. The cells were uniformly blown until a single-cell suspension was formed for flow cytometry analysis.Three flow antibodies were utilized: monoclonal antibody against CD90,CD34 and CD44H monoclonal antibody(Thermo Fisher, USA).BMSCs exhibit the capacity to differentiate into diverse tissue lineages, such as bone, cartilage, and adipocytes. The third generation of bone marrow mesenchymal stem cells was chosen for three-line induction differentiation experiments. BMSCs were cultivated to the second passage and subsequently digested and enumerated. Cells were inoculated in 6-well plates at an initial cell density of 2 × 10⁵ cells/well, and 2 mL of rat BMSCS complete medium was added to each well. Once the cells were completely adhered to the wall, the induction differentiation medium was prepared in strict accordance with the instructions of the kit (rat mesenchymal stem cell three-line induction differentiation medium, Punosai, China) for the three-line induction differentiation experiments.

Extraction of exosomes

When the cell confluence attained 60–70%, BMSCs were cultivated in DMEM/F12 medium containing 10% FBS prepared with exosome-free fetal bovine serum (VivaCell, China) for 24 h. The medium supernatant was collected and centrifuged at 2000 g for 30 min. The supernatant was transferred to a fresh centrifuge tube and centrifuged at 12000 g for 30 min to eliminate dead cells, cell debris, and large particles (shedding vesicles and apoptotic bodies). Centrifugation was executed at 110,000 g for 70 min using a Type 70Ti rotor (Beckman Coulter, USA). The precipitate was washed in phosphate buffered saline (PBS) to eliminate contaminating proteins and centrifuged anew at 110,000 g for 70 min. PBS was removed and exosomes were freshly re-suspended in 100 μL PBS. All centrifugation steps were conducted at 4 °C.

Nanoparticle tracking analysis (NTA)

The exosomal solution was diluted to a vesicle suspension concentration ranging between 1 × 10⁷/ml and 1 × 10⁹ and examined by a Nanosight NS300 (NanoSight Ltd., Amesbury, UK) equipped with a 405 nm laser line for the determination of the size and number of isolated particles. Sixty seconds of video was captured at a frame rate of 30 frames per second and analyzed with NTA software.

Transmission electron microscopy (TEM)

20 to 40 μL of the exosome solution was dropped onto the copper mesh and was negatively stained with 2% phosphotungstic acid solution for 10 min. The samples were then dried under an incandescent lamp for 2 min. Transmission electron microscopy (JEOL, Japan) was utilized for observing exosome morphology.

Western blot analysis

A 5 × loading buffer was added to the exosome supernatant and subjected to denaturation through heating in a metal bath with rabbit polyclonal antibodies CD81 (Proteintech, USA), TSG101 (Proteintech, USA), and HSP70 (Proteintech, USA). Proteins were visualized on a protein visualizer.

Animal study design and rat RCT model

Sixty male SD rats (6–8 weeks) weighing 220–270 g were randomly allocated into 5 groups: the repair-alone group, the GelMA group (hydrogel), the BMSC-Exos group (hydrogel + BMSC-Exos), the control group, and the BMSC-Exos + TGF-β1 Inhibited group (hydrogel + BMSC-Exos + TGF-β1 inhibitor P144), with 12 rats in each group(n = 8 for biomechanical testing, and 4 rats for both histologic analysis and PCR test for each group)²⁷. Normal Control Group: Rats did not undergo any surgical procedure or injury, serving as intact healthy controls.Repair-alone Group (Injury + Repair): RCT model established and repaired surgically without hydrogel or exosome treatment.GelMA Group: RCT model repaired and treated with blank GelMA hydrogel.BMSC-Exos Group: RCT model repaired and treated with GelMA hydrogel loaded with BMSC-derived exosomes.BMSC-Exos + TGF-β1 Inhibited Group: RCT model repaired and treated with GelMA hydrogel loaded with both BMSC-Exos and the TGF-β1 inhibitor P144.A single postoperative time point of 6 weeks was selected for evaluation based on previous studies demonstrating significant tendon-bone healing at this stage in rat models, which allows for clear comparison of treatment efficacy under standardized conditions.

Anesthesia was implemented by continuous inhalation of 2.5% isoflurane. The anesthetized rats were splinted and maintained in the right lateral decubitus position, the left shoulder joint was exposed, the skin was prepared, and the surgical area was sterilized. The left limb of the rat was adducted and externally rotated to ensure adequate exposure of the surface marker of the humeral head. The skin in the surgical area was longitudinally incised to create a longitudinal incision, and the subcutaneous tissue was separated bluntly to fully expose the deltoid muscle. After the deltoid was cut, the subdeltoid tissue was likewise separated bluntly, and the humeral head and supraspinatus tendon were fully exposed. A 6–0 suture was utilized to pass under and lift up the supraspinatus tendon to expose the footprint area, and the supraspinatus tendon was separated from the humeral head using a scalpel knife. After confirming that the tendon was completely separated from the humeral head, the residual tendon and cartilage in the footprint area were removed to rejuvenate the wound. After the successful separation of the supraspinatus tendon from the humerus, a bone tunnel was formed by transverse puncture from both sides of the humeral head using a small towel forceps. A 4–0 non-absorbable suture was employed to pass through the bone tunnel, and the suture was passed through the left and right sides of the tendon. Saline or hydrogel 30μL was injected into the imprinted area above the nodules in accordance with the group. The hydrogel was solidified by blue light, and the suture was tightened to restore the tendon to its original insertion position. After the operation was completed, layer-by-layer suture was conducted, and the dressing was covered after skin disinfection (Fig. 1). After surgery, the animals were permitted free and unrestricted movement in the rearing cage, and their general health was inspected daily. After 6 weeks of treatment, rats were euthanized by isoflurane overdose inhalation.These doses are referenced from the experimental literature²⁷ and were approved in the IACUC protocol.

Fig. 1.

Rat RCT Model (From left to right, they are: exposure of the deltoid muscle; exposure of the supraspinatus tendon; tenotomy at the greater tuberosity; debridement and freshening of the footprint area.).

Preparation of hydrogels containing BMSC-Exos versus those without BMSC-Exos

Gelatin methacryloyl (GelMA) hydrogels are prevalently employed in diverse biomedical domains due to their appropriate biological and modulable physical properties. It can function as a drug delivery vehicle to effectively attain local and sustained release of exosomes. Gelatin methacryloyl (GelMA) hydrogels were prepared as described previously. Briefly, a 10% (w/v) GelMA solution was obtained by dissolving GelMA powder in phosphate-buffered saline containing 0.5% (w/v) photoinitiator (LAP), followed by filtration and sterilization. For BMSC-Exos loading, exosomes were mixed with the GelMA precursor solution to achieve a final concentration of 4 × 10¹⁰ particles/mL, a ratio optimized based on prior in vitro release and bioactivity studies (see reference 23 and our previous work). The release profile of exosomes from GelMA hydrogel was confirmed in vitro, showing sustained release over 7 days. For the inhibition group, P144 (MedChemExpress) was dissolved in DMSO and added to the BMSC-Exos-loaded GelMA solution to a final concentration of 300 µg/mL.

Biomechanical testing of the RCT repair

The XWW-10 Materials Testing System (MTS) was used to evaluate the biomechanical properties of rat humerus samples. After a 6-week treatment, rats were euthanized with an overdose of isoflurane. The supraspinatus tendon and peritendinous tissue were removed, leaving only the tendon-humerus complex. The humerus was fixed in the upper jig, and the supraspinatus tendon was clamped and secured in the lower jig, ensuring alignment with the humerus’s long axis. The specimen was preloaded to 1N and pulled at 10 mm/min until fracture. The maximum failure load and tendon stiffness were determined from the stress–strain curve.

Histological analysis

Following 6 weeks of treatment, the rats were euthanized, and the supraspinatus humerus samples were meticulously dissected. The specimens were fixed with 10% paraformaldehyde for 3 days and subsequently decalcified through immersion in a rapid decalcification solution (Zhuoding, China) for 48 h, with the solution being changed every 12 h. Subsequently, the samples were immobilized in a biological tissue dehydrator and embedded. The embedded tissue was sliced into 10 μm thick sagittal sections using a paraffin microtome (Leica, Germany). The histological alterations were analyzed via hematoxylin–eosin staining, Masson staining, and Sirius red staining. The continuity, parallel orientation, density, vascular distribution, and cellular composition of collagen fibers at the tendon-bone interface were subsequently evaluated. Histological scores were evaluated employing a semiquantitative scoring system (0–3 grades for each item). Regarding the continuity of collagen fibers and the parallel orientation of collagen fibers, the score was defined based on the percentage: 0 = a proportion ranging from 0 to 25%; 1 = 25–50%; 2 = 50% to 75% ratio and 3 = 75% to 100% ratio. For collagen fiber density, scores were defined by the percentage: 0 = very loose, 1 = loose, 2 = dense, and 3 = very dense. For vascular distribution and cellular composition, scores were defined by the percentages: 0 = severe or significant, 1 = moderate presence, 2 = mild presence, and 3 = absence or minimal presence. The total score amounts to 15, with higher scores indicating better recovery. Analysis was carried out using ImageJ software (National Institutes of Health). The samples were evaluated by 2 independent observers.

RNA extraction and qRT-PCR

The supraspinatus humerus samples were meticulously dissected and placed in a tissue grinding tube to which TRIzol reagent was added, followed by tissue homogenization. Post-homogenization, chloroform and isopropanol were added sequentially for the extraction of total RNA. The RNA was washed twice with 75% ethanol, and upon drying, the RNA was dissolved in enzyme-free water. Total RNA was quantified by using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific), and cDNA synthesis (RT-PCR) was conducted with FastKing RT Kit (with gDNase) (TIANGEN Biotech; China). With the SuperReal PreMix Plus (TIANGEN; Beijing; China), total RNA (2 μg) was reverse transcribed into complementary DNA. Three real-time PCRs were carried out as previously described, and Relative expression was calculated using the 2 − ΔΔCt method normalized to GAPDH (endogenous loading control). The primers employed to detect gene marker expression included: Col1, Col3, SCX, TNC, α-SMA.

Total RNA was extracted from tendon-bone interface tissues (n = 4 per group) using TRIzol. cDNA was synthesized and real-time PCR was performed with SYBR Green. Gene expression was normalized to GAPDH and analyzed via the 2⁻ΔΔCt method. The following primer sequences were used:

Gene	Forward Primer (5′ → 3′)	Reverse Primer (5′ → 3′)
Col1a1	5′-ATGCTGAATCGTCCCACCAG-3′	5′-AGAACAGTCTCTCCCACCCA-3′
Col3a1	5′-GGGACAACGAGGTCCTTCAG-3′	5′-GTCTTCGCCCTTAGGTCCTG-3′
Scx	5′-CAGATCTGCACCTTCTGCCT-3′	5′-TAGTGGCATCACCTCTTGGC-3′
α-SMA	5′-TGGGATTGCTGACAGGATGC-3′	5′-AGAGAAGGCCCTGTGACTTT-3′
TnC	5′-CCAGCCATTGCCACTGTAGA-3′	5′-CCAGACACTGTGCGGGTAAT-3′
GAPDH	5′-GGCTCATGACCACAGTCCAT-3′	5′-ACATTGGGGGTAGGAACACG-3′

To confirm the inhibitory effect of P144, TGF-β1 mRNA expression was also quantified in the BMSC-Exos + TGF-β1 inhibited group and compared with the BMSC-Exos group.

Statistical analysis

Data are presented as mean ± standard deviation. One-way analysis of variance (ANOVA) was used to compare differences among the five groups, followed by Tukey’s post-hoc test for multiple comparisons. A P-value < 0.05 was considered statistically significant. Sample sizes for biomechanical (n = 8), histological (n = 4), and PCR (n = 4) analyses were determined based on previous similar studies and pilot experiments; however, we acknowledge that a formal power analysis (e.g., using G*Power) would further strengthen the experimental design for future studies.

Results

1. Extraction and Identification of Bone Marrow Mesenchymal Stem Cells.

Rat BMSCs were isolated and purified in accordance with the aforementioned experimental methods. When the cells were passaged to the third generation, the morphology of the cells was regular, mostly fusiform and polygonal, which constitutes the typical morphology of BMSCs(Fig. 2A).

Fig. 2.

(A) First-passage BMSCs (left) and Third-passage BMSCs (right). Fig. (B) Bone marrow mesenchymal stem cell tri-lineage induction differentiation experiment results. The first row was captured at 100 × magnification, the second row was captured at 200 × magnification. (C) Flow Cytometry Characterization Results of Bone Marrow-Derived Mesenchymal Stem Cells.

The tri-lineage induction differentiation experiment was conducted to systematically evaluate the multi-differentiation potential of BMSC. As illustrated in Fig. 2B, in the osteogenic induction culture group, after a 3-week osteogenic induction period, the cells transitioned from spindle-shaped and polygonal forms into morphologies characteristic of mature osteoblasts. Microscopic examination revealed opaque regions, and Alizarin Red staining demonstrated dense orange-red nodular precipitates indicative of calcium deposition. In the adipogenic differentiation induction group, following a 3-week adipogenic induction period, cell morphology progressively became more rounded, resembling pre-adipocytes. Lipid droplets were clearly visible under the microscope, increasing in size as differentiation progressed and eventually fusing to form single or multiple prominent lipid droplets. Oil Red O staining highlighted these lipid droplets in red, confirming adipogenesis. In the chondrogenic differentiation induction group, after a 4-week differentiation induction period, cartilage nodules developed. Following embedding and sectioning, Alcian Blue staining revealed blue-green chondrocytes under microscopic observation, indicating glycosaminoglycan-rich extracellular matrix formation.The results of flow cytometry indicated that the rat bone marrow-derived cell population isolated in this study displayed classic characteristics of bone marrow mesenchymal stem cells, with high expression of CD44 and CD90 and no expression of CD34, which conformed to the identification criteria for mesenchymal stem cells (Fig. 2C).All experiments were performed in triplicate.

• 2.Phenotwwypic identification of exosomes derived from bone marrow mesenchymal stem cells.

TEM results demonstrated that BMSC-Exos exhibited typical spherical vesicles with a lipid bilayer surrounded structure (Fig. 3A). NanoSight reports that BMSC-Exos have vesicle sizes ranging from 90 to 200 nm with a peak at 141 nm, indicating that the exosomes in the sample possess typical size characteristics. The concentration of the resuspended exosome solution was calculated to be approximately 1.2 × 10¹¹ particles /mL (Fig. 3B).Western blot (WB) results confirmed the positive expression of the exosomal marker proteins CD81, TSG101, and HSP70 in the isolated vesicles (Fig. 3C). These results indicated that exosomes were successfully isolated from the BMSC culture supernatant.

Fig. 3.

Exosome characterization results: (A) Particle size concentration distribution graph. (B) Transmission Electron Microscopy image of exosomes. Morphology of BMSC-derived exosomes was observed using TEM. Scale bars, 500 nm (left), 100 nm (right). (C) Western blot results.

Biomechanical test

Tendon-bone complex rupture occurred mostly at the tendon-bone interface. At 6 weeks post-operation, the maximum failure load of the normal control group (27.59 ± 2.32 N, N = 8) was greater than that of the repair-alone group (16.67 ± 2.59 N, N = 8) (P < 0.05). The GelMA group (18.17 ± 2.04 N, N = 8) showed no statistically significant difference compared to the repair-alone group (P > 0.05). Similarly, there was no statistically significant difference between the BMSC-Exos + TGF-β1 Inhibited group (16.16 ± 2.36 N, N = 8) and the repair-alone group (P > 0.05). The maximum failure load of the BMSC-Exos group was 23.79 ± 4.26 N (N = 8), which was significantly higher than that of the repair-alone group, the GelMA group, and the BMSC-Exos + TGF-β1 Inhibited group (P < 0.05 for comparisons with each group), indicating that the exosome-loaded hydrogel provided superior repair enhancement compared to the hydrogel alone. The maximum failure load of the BMSC-Exos group was similar to that of the normal control group, with no statistically significant difference (P > 0.05). The stiffness results were consistent with the trend of maximum failure load. The stiffness of the repair-alone group was the lowest (5.68 ± 0.33 N/mm) and was significantly lower than that of the normal control group (13.57 ± 2.60 N/mm) (P < 0.05). The stiffness of the BMSC-Exos + TGF-β1 Inhibited group (8.85 ± 2.46 N/mm) and the GelMA group (8.81 ± 0.90 N/mm) was marginally higher than that of the repair-alone group, but with no statistically significant difference (P > 0.05). The stiffness of the BMSC-Exos group was 11.62 ± 3.99 N/mm, significantly higher than that of the repair-alone group and the GelMA group (P < 0.05), and showed no statistically significant difference from the normal control group (P > 0.05) (Fig. 4).

Fig. 4.

Biomechanical results. Figure (A) shows the supraspinatus tendon-humeral complex placed on the biomechanical tester. Figure (B) shows the maximum failure load results. Figure (C) shows the stiffness results. N = 8, * is P < 0.05; ** is P < 0.01; *** is P < 0.001.

Histological analysis

The assessments were performed by two pathologists blinded to the experimental groups. The results of Masson’s trichrome staining are presented in Fig. 5A. Compared with the normal control group, the repair-alone group had lower collagen fiber density and poorer alignment, with a statistically significant difference in histological score (P < 0.05). In contrast to the other experimental groups, the tendon-bone interface in the BMSC-Exos group exhibited denser collagen fibers, better continuity and parallelism, and fewer inflammatory cells and blood vessels, more closely resembling the natural interface. The histological scores of the BMSC-Exos group were significantly higher than those of the repair-alone group, the GelMA group, and the BMSC-Exos + TGF-β1 Inhibited group (P < 0.05). The histological scores of the GelMA group and the BMSC-Exos + TGF-β1 Inhibited group were higher than that of the repair-alone group, but the differences were not statistically significant (P > 0.05).

Fig. 5.

Histological analysis: (A) HE staining and Masson staining results and histologic scoring. Figure A shows the results of HE staining and Masson staining, scale bar = 50 μm. Figure B histologic scores, n = 4,* is P < 0.05; ** is P < 0.01; *** is P < 0.001. (B) Observation of Sirius Red staining under polarized light microscopy. Scale bar = 50 μm.

Sirius red staining results showed that under polarized light microscopy, the birefringence intensity of the repair-alone group was lower, with shorter collagen fibers and poor continuity. The birefringence intensity of the GelMA group and the BMSC-Exos + TGF-β1 Inhibited group was higher than that of the repair-alone group, but fiber continuity remained insufficient. The normal control group and the BMSC-Exos group exhibited high birefringence intensity and favorable fiber continuity. Semi-quantitative analysis of the stained areas using image analysis software revealed that the ratio of type I collagen (orange-red or bright yellow, appearing as thick fibers) to type III collagen (green or yellow-green, appearing as a fine mesh) in the BMSC-Exos group was comparable to that in the normal control group, and both were significantly better than the other experimental groups (P < 0.05). Compared with the normal control group, the collagen color in the BMSC-Exos group still displayed some heterogeneity (Fig. 5B).

Changes in mRNA levels

In the qRT-PCR assay, the gene expression levels in the GelMA group were similar to those in the repair-alone group, with no statistical significance (P > 0.05). Compared with the repair-alone group, the mRNA expression of ColⅠ, ColⅢ, Scx, α-SMA, and TnC in the BMSC-Exos group was significantly increased (all P < 0.05). The expression of Col I, Scx, α-SMA, and TnC genes in the BMSC-Exos group was also significantly higher than that in the GelMA group (P < 0.05). Compared with the BMSC-Exos + TGF-β1 Inhibited group, the mRNA expression of ColⅠ, ColⅢ, α-SMA, and TnC in the BMSC-Exos group was significantly elevated (P < 0.05). The mRNA expression of TnC in the BMSC-Exos group was lower than that in the normal control group (P < 0.05) but higher than that in the other three groups (P < 0.05). Furthermore, to verify the efficacy of the TGF-β1 inhibitor, we examined the expression of TGF-β1 and its downstream signaling molecules. The results showed that compared with the BMSC-Exos group, the BMSC-Exos + TGF-β1 Inhibited group exhibited significantly reduced TGF-β1 mRNA expression (P < 0.05), and the expression levels of downstream target genes (such as those indicative of Smad2/3 phosphorylation, inferred from related gene expression or subsequent protein assays) also showed a decreasing trend, indicating that P144 effectively inhibited the activity of this pathway (Fig. 6).

Fig. 6.

A mRNA expression levels. Figures A–E show the expression comparison of Col Ⅰ, Col Ⅲ, Scx, α-SMA, and TnC between groups. (n = 4,* is P < 0.05; ** is P < 0.01; *** is P < 0.001.)

Discussion

In this study, we assessed the therapeutic effect of BMSC-Exos on rotator cuff repair and discovered that BMSC-Exos could bring about significant improvement in tendon biomechanics during RCT treatment. Histological observation indicated that collagen fibers were well organized and the expression of tendon and tendon-bone relative markers was upregulated.

SD rats have been widely used as experimental models for the establishment of acute and chronic rotator cuff injury models, demonstrating good results^28,29. Rats generally possess strong tissue healing capabilities, and the repair and regeneration processes of rotator cuff tears (RCT) may take up to 8 weeks or longer³⁰. Moreover, similar to the study by Chunfeng Zhao et al., in our study, a modeling period of 6 weeks also showed favorable repair outcomes²⁷.

The role of TGF-β in physiological healing has been comprehensively examined. Studies have substantiated that transforming growth factor-β1 (TGF-β1) can stimulate the proliferation of fibroblasts and potentiate the synthesis of collagen and non-collagen proteins in fibroblasts³¹. When undertaking anterior cruciate ligament reconstruction, it has been ascertained that exogenous administration of TGF-β1 can significantly enhance the bonding strength between the graft and the tunnel wall at 3 weeks³². Additionally, Anaguchi et al. discovered that the application of TGF-β1 significantly augmented the tangential modulus and tensile strength of the regenerated fibrous tissue of the patellar tendon after removal of the central portion³².

In our previous in vitro study, we have discerned that TGF-β1 is highly expressed in BMSC-Exos, and BMSC-Exos containing TGF-β1 can promote the expression of TGF-β1 in tendon cells in vitro, thereby impacting the proliferation, migration and fibrosis activity of tendon cells²⁶. The effect of BMSC-Exos in tenocytes was significantly mitigated by the use of TGF-β1 receptor inhibitors³³. In this study, the results of biomechanical experiments manifested that the maximum tensile load and linear stiffness of rats in the BMSC-Exos group were significantly higher than those in the repair-alone group (P < 0.05) and showed a strong increasing trend compared to the GelMA group, although the comparison with the GelMA group alone did not reach statistical significance in all parameters,and were proximate to the normal control group. From the histological examination, it was discerned that the collagen fibers at the tendon-bone interface of rats in the BMSC-Exos group were denser, and the continuity and parallelism were superior to those in the repair-alone group and the GelMA group. Compared with the other experimental groups, the collagen fibers in the BMSC-Exos group were more continuous and had brighter color, and the histological pattern of collagen composition, as observed under polarized light after Sirius red staining, appeared more similar to that of the normal control group, with a balanced presence of both thick (likely Col I) and fine (likely Col III) fibers.These results buttressed our conclusion that BMSC-Exos could facilitate rotator cuff repair in vivo.

The gene expressions of ColⅠ, ColⅢ, Scx, α-SMA, and TnC were detected through PCR. The PCR results manifested that the expression of Col Ⅰ, Col Ⅲ, Scx, and α-SMA in the BMSC-Exos group was significantly higher than that in the other groups. The expression of the TnC gene in the BMSC-Exos group was higher than that in the repair-alone group, the GelMA group, and the BMSC-Exos + TGF-β1 Inhibited group, but lower than that in the normal control group. This phenomenon could potentially occur because TnC plays a complex and multi-faceted role in tendon injury, both as part of the injury response and as a factor restricting normal healing and functional recovery; the specific reasons for which remain to be explored.

In conclusion, an in vivo experiment was carried out to investigate the effect of BMSC-Exos on rotator cuff repair. The results disclosed that BMSC-Exos could promote the repair of rotator cuff injury in rats through TGF-β1. This study offers a new potential treatment for rotator cuff tear by conducting in vivo experiments.

Conclusions

In vivo experiments demonstrated that BMSC-Exos could promote the repair of rotator cuff injury through TGF-β1.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

RCI

Rotator cuff injury

Exos

Exosomes

BMSCs

Bone marrow mesenchymal stem cells

BMSC-Exos

Bone marrow mesenchymal stem cell-derived exosomes

TGF-β1

Transforming growth factor-β1

TEM

Transmission Electron Microscope

NTA

Nanoparticle Tracking Analysis

TSG101

Recombinant Tumor Susceptibility Gene 101

HSP70

Heat Shock Protein 70

Col Ⅰ

Collagen Type I

Col Ⅲ

Collagen Type III

α-SMA

α-Smooth muscle actin

Scx

Scleraxis

TnC

Troponin C

GelMA

Gelatin Methacryloyl

Author contributions

Peng, Kun:Conceptalization, Data Curation, Fommal AnalysisInvestigation, Methodology, Sofware,Visualization. Writing-Original DraftWriting-Review & Editing. Wang Shubin:Conceptalization, Data Curation, Fommal AnalysisInvestigation, Methodology, Sofware,Visualization. Writing-Original DraftWriting-Review & Editing. Xu Cong：Conceptalization, Funding Acquisition.Resources, Supervision, Validation, Writing-Review &Editing. Li, Jia:Resources. Zhang, Jian:Formalanalysis. Gao, Yaxian:Resources. Wang, Yongwei:Resources. Liu, Fei:Supervision. Guan, Chenwei:Formalanalysis. Wang Shubin and Peng Kun made equal contributions to this research.

Funding

The authors are grateful for the financia support provided by the Hebei Provincial Department of Science and Technology (Grant No.H2021406023),Health Commission of Hebei Province (Grant The “14th Five-Year Plan” medical key disciplines of Hebei province special funds).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

This study and included experimental procedureswere approved by the institutional animal care anduse committee of Chengde Medical University Affiliated Hospital. All animal housing and experimentswere conducted in strict accordance with the institutional guidelines for care and use oflaboratory animals.This study was approved by the ethics committee of Chengde Medical University Affiliated Hospital (approval No.CYFYLL2022136).We certify that the study wasperformed in accordance with the 1964declaration of HELSINKl and later amendments.Our manuscript confirming the study is reported in accordance with ARRIVE guidelines.