Therapeutic effects of bone marrow mesenchymal stem cells‐derived exosomes on osteoarthritis

Abstract Mesenchymal stem cells (MSCs) have shown chondroprotective effects in clinical models of osteoarthritis (OA). However, effects of MSC‐derived exosomes on OA remain unclear. The study aimed to investigate the therapeutic potential of exosomes from human bone marrow MSCs (BM‐MSCs) in alleviating OA. The anterior cruciate ligament transection (ACLT) and destabilization of the medial meniscus (DMM) surgery were performed on the knee joints of a rat OA model, followed by intra‐articular injection of BM‐MSCs or their exosomes. In addition, BM‐MSC‐derived exosomes were administrated to primary human chondrocytes to observe the functional and molecular alterations. Both of BM‐MSCs and BM‐MSC‐derived exosomes alleviated cartilage destruction and subchondral bone remodelling in OA rat model. Administration of BM‐MSCs and exosomes could reduce joint damage and restore the trabecular bone volume fraction, trabecular number and connectivity density of OA rats. In addition, in vitro assays showed that BM‐MSCs‐exosomes could maintain the chondrocyte phenotype by increasing collagen type II synthesis and inhibiting IL‐1β–induced senescence and apoptosis. Furthermore, exosomal lncRNA MEG‐3 also reduced the senescence and apoptosis of chondrocytes induced by IL‐1β, indicating that lncRNA MEG‐3 might partially account the anti‐OA effects of BM‐MSC exosomes. The exosomes from BM‐MSCs exerted beneficial therapeutic effects on OA by reducing the senescence and apoptosis of chondrocytes, suggesting that MSC‐derived exosomes might provide a candidate therapy for OA treatment.


| INTRODUC TI ON
Osteoarthritis (OA) is the most common type of chronic joint disease throughout the world, with the decline in living standards, functional disability and degenerative joint disorder. 1,2 Clinically, osteoarthritis (OA) is characterized by degradation of articular cartilage, inflammation of synovial and remodelling of subchondral bone. 3,4 OA patients suffer from severe pain, limitation of movement and a high risk of disability. Despite improvements in pharmacological, nonpharmacological and joint replacement surgery, 5 the current treatments for osteoarthritis are a great challenge due to unsatisfactory overall effects and substantial adverse events. 6 Knee arthroplasty removes the arthritic tissue but 10-20% of patients still report that pain remains after surgery. 7 Although joint replacement surgery and knee arthroplasty temporarily alleviate the symptoms, overall results are still disappointing with the progression of OA. 8 Therefore, it is of great significance to explore new therapeutic approaches against OA.
Mesenchymal stem/stromal cells (MSCs) are a population of multipotent stem cells, with the potential in adipogenesis, osteogenesis and chondrogenesis. Stem cell therapies exhibit great potential for the treatment of various diseases with its capability of tissue repair, immunomodulatory regulation and anti-inflammatory effects. 9 Interestingly, studies using cell tracking in cartilage repair show only limited cartilage formation by chondrogenic differentiation of the injected MSCs. 10,11 By contrast, rather than direct differentiation into the target tissue, accumulating evidence indicates that the biological and therapeutic effects of MSCs are mainly attributed to paracrine mechanisms mainly via secreting molecules, including growth factors, chemokines, cytokines and extracellular vesicles (EVs). 12 Among these paracrine molecules, EVs may be the most valuable therapeutic factor. 13 Extracellular vesicles, mainly made up of exosomes and microvesicles (MVs), are now recognized to participate in cell-to-cell communication by transporting various proteins, microRNAs (miRNAs) and mRNAs. 14 The therapeutic effects of exosomes for regenerative medicine have been thoroughly investigated.
Because parent cell selection enhancing cargo specificity, avoiding of adverse immune responses or rejection seen in cellular therapies, exosomes continue to improve in regenerative medicine. 15 Given the abundant contents and potential advantage as a carrier, the administration of MSCs-EVs has provided a promising option in treating a great deal of benign or malignant diseases.
Lung spheroid cell-secretome (LSC-Sec) and exosomes (LSC-Exo) exhibited great therapeutic potential for lung regeneration in idiopathic pulmonary fibrosis (IPF). 16 Better yet, exosomes derived from MSCs have facilitated immunosuppressive, pro-angiogenic, anti-apoptotic and anti-fibrotic features in various tissues, including the skin, 17 limbs, 18 heart 18 and lung. 15 Recent a targeting peptide, cardiac homing peptide (CHP) is used to target intravenously infused exosomes to the infarcted heart, enhancing the targeting of exosomes. 19 Furthermore, the role of exosomes as a regenerative medicine has potentially provided a novel nonsurgical, disease-modifying approach for OA. 20 Meanwhile, the contents in the exosomes could reduce MMP-13 production in the joint space, which lead to the decreased cartilage destruction and the balance of cartilage synthesis. There are currently no therapeutic interventions that can reverse the process of OA. However, the exact roles of MSC-derived exosomes in OA remain partially unclear. In this study, we aimed to explore the therapeutic effects and potential mechanism of MSCs-exosomes on the OA animal.

| Animal model
The rats were housed in groups of 5 per plastic cage on sawdust bedding in a 12:12 light-dark cycle (light-on period, 6:00 AM-6:00 PM) with controlled temperature. They were fed a standard diet and were provided access to filtered water ad libitum. The animals were acclimatized for 1 week before the experiments. Cartilage destruction was evaluated using the Osteoarthritis Research Society International (OARSI) scoring system according to the percentage of the vertical clefts/erosion to the calcified cartilage. 21 OA was induced unilaterally in the knee joint of donor animals by complete excision of the medial meniscus and resection of the anterior cruciate ligament. The knee joint of rats was injected with 100 µl exosome solution (100 µg) or MSCs (10 6 cell) every week post-surgery, which were sacrificed at the 8th week.

| Isolation of bone marrow MSCs
Primary bone marrow MSCs (BMSCs) were incubated in lowglucose Dulbecco's modified Eagle's medium (DMEM; Gibco, MD, USA) with 10% foetal bovine serum (FBS; Gibco) and 1% penicillin/ streptomycin (Gibco). The cells were cultured in a humid incubator with 5% CO 2 . The initial cell number is 10 6 -10 7 /ml. After that, change the medium with L-DMEM complete medium (10% FBS, L-DMEM, 1% penicillin/streptomycin) every 3 days. Use an inverted microscope to observe its growth status every day. When the cells are cultured to 70%-80% confluence of adherent cells, the cells are passaged. 22 Cells at passage 3 were characterized for the following experiments.

| Exosome isolation and procedures
This protocol of exosome isolation was based on previous exosome isolation method. 23 Exosomes were isolated from conditioned medium of BM-MSCs by ultracentrifuge, and almost 50ml conditional medium was collected. Initial spins consisted of

| Transmission electron microscopy of MSCs-Exosomes
The morphology of MSCs-exosomes was assessed by transmission electron microscopy (TEM). The exosome pellets were fixed in 3% (w/v) glutaraldehyde and 2% paraformaldehyde in cacodylate buffer and then loaded to copper grids coated with formvar. After washing, the grids were contrasted in 2% uranyl acetate. Following air drying, the samples were examined by TEM (Morgagni 268D, Philips, Holland).

| Flow cytometry
The apoptosis assay was conducted according to the manufacturer's instructions of apoptosis detection kit (BD Biosciences, San Jose, CA). Cells were collected by trypsin and rinsed with PBS two times.
Then, the samples were subsequently incubated with 5 μl Annexin V-FITC/ PI for 15min in the dark, and thereby detected by using FACSCalibur flow cytometer (BD Biosciences, San Jose, CA).

| Western blotting
The expression of protein was measured by Western blotting. Cells were lysed by radioimmunoprecipitation assay (RIPA) with 1 mM phenylmethanesulfonyl fluoride (PMSF, Beyotime, China). Bicinchoninic acid (BCA) protein assay reagent (Pierce, Rockford, USA) was used to determine the protein concentration of lysates. Equivalent amounts of protein were applied to SDSpolyacrylamide gels for electrophoresis and transferred to PVDF membranes as described previously. 24 Following blocking in 5% BSA solution, the membranes were incubated in primary antibody overnight at 4℃. After that, samples were treated with secondary antibody for 2 h at room temperature, followed by washing with PBS three times. At last, the samples were visualized using ECL (Millipore, USA).

| Quantitative real-time PCR
Total RNA was extracted from cells by using TRIzol (Invitrogen) according to the manufacturer's instructions. The first-strand cDNA was synthesized by using reverse transcriptase Kit (TAKARA, Dalian, China). Quantitative real-time PCR (RT-qPCR) was con-

| Anterior cruciate ligament (ACL) transection and medial meniscectomy (MM)
The left knee was prepared in a surgically sterile fashion. Through mini-arthrotomy, the anterior cruciate ligament (ACL) was transected with a scalpel, and MM was performed by excising the entire medial meniscus. The knee joint was irrigated, and the incision was closed. Ampicillin was administrated at the concentration of 50 mg/kg post-surgery. The surgical site and the activities of the animals were observed daily. Six male Sprague-Dawley rats (mean weight = 358 ± 5 g) underwent a unilateral sham operation as an additional control group.

| SAβ-Gal staining
Senescence-associated β-galactosidase (SAβ-gal) staining was conducted to evaluate the senescence induced by IL-1β according to the instruction. In brief, following exposure to IL-1β, cells were washed in phosphate-buffered saline (PBS) three times, followed by fixed in 4% paraformaldehyde for 15 min at room temperature. Then, the cells were incubated in SAβ-gal staining solution overnight at 37°C. The number of senescent cells was observed and counted under light microscope. The experiment was repeated independently three times.

| Histological assessment
At 8th week post-surgery, treatment-related cartilage and bone changes in the ACLT-joints were scored by two independent observers based on gross morphological appearance compared with the sham controls. Knee joints were fixed in 4% paraformaldehyde and decalcified in neutral 10% ethylenediaminetetraacetic acid (EDTA, PH 7.2) for 4 weeks at room temperature. Thereafter, the knees were embedded in paraffin blocks in the coronal position.

Sections of 6 μ m thickness were performed for Safranin-O and
Fast-Green staining. The articular cartilage was assessed using the Osteoarthritis Research Society International (OARSI) histological scoring system. 25 After behavioural tests, the rats were sacrificed, and the gross morphologic changes in their right femoral condyle cartilage lesions were scored as follows: 0 = surface appears normal; 1 = minimal fibrillation or slight yellowish discoloration of the surface; 2 = erosion extending to the superficial or middle layers; 3 = erosion extending to the deep layer; and 4 = erosion extending to the subchondral bone. 26 The right knee was routinely fixed with 4.0% paraformaldehyde at 4°C for 48 h and then decalcified with 10% EDTA at PH 7.4 for 4 weeks. After decalcification, each specimen was embedded in paraffin and sagittally sectioned through the medial femoral condyle at a thickness of 6 μm. The sections were then dewaxed in xylene and hydrated with graded ethanol series before being stained with haematoxylin and eosin (H&E).

| Histological staining
The complete articulation of the knees was maintained in poly formaldehyde (4%) and was fixed for twenty-four hours. Subsequent to traditional paraffin embedding, the articulation was cut into pieces

| Statistical analysis
Statistical analysis was performed using GraphPad Prism 5 (GraphPad Software, CA, USA). The data collected were expressed as mean ± standard deviation (SD). Statistical significance was assessed by Student's t test for comparing two groups and one-way analysis of variance (ANOVA) for comparing multiple groups. A value of p < 0.05 was considered to evaluate statistical significance. All experiments were performed independently at least three times.

| Characterization of BM-MSC-derived exosomes and the morphological effects on OA rats
The BM-MSC-derived exosomes were extracted from serum-free medium by gradient centrifugation. The TEM showed that ex- Osteophyte formation and synovial hyperplasia were not observed.
Compared with the OA group, the articular cartilage surface of the exosomes group was still intact without the formation of osteophytes, and the fluid in the articular cavity was significantly reduced.
The joint cavity condition of the MSCs group was also significantly better than that of the OA group. However, both of BM-MSCs and their exosomes improved the status of the articular cavity, cartilage and femur in OA rats ( Figure 1C). Semi-quantification showed that score of the OA group was higher than that of control group, while the exosomes and MSC administration significantly reduced the OA score ( Figure 1D).  Figure 1F). Furthermore, MSC and exosome treatments increased the intensity of toluidine blue staining in cartilage matrix of OA rats ( Figure 1G). Masson staining, mainly for collagen evaluation, demonstrated that MSC and exosome treatments increased positive area (red dye) and alleviate the destruction of articular cartilage surface ( Figure 1H). Additionally, immunohistochemical staining of cartilage tissues showed that the intensity of collagen II expression in OA group significantly decreased compared with the control group, while the administration of MSCs and exosomes obviously recovered the expression of collagen II ( Figure 1I). Mankin score was performed according to the structure, cell arrangement, matrix colouring, hydatid integrity and cartilage surface damage. Based on the staining above, MSCs and exosome groups had significantly lower Mankin score than that of the OA group (Fig. S1).

| Effects of exosomes on degeneration of trabecular microarchitecture of subchondral bone in OA rats
It is well-known that degenerative changes in the subchondral region are accompanied by the degeneration in the articular car-

| Effects of MSCs-exosomes on IL-1β-induced senescence and apoptosis in chondrocytes
Based on the results of the OA model, we further discovered the effects of exosomes derived from BM-MSCs on OA in vitro. Normal chondrocytes were treated with the IL-1β to mimic the OA condition.

| BM-MSCs-exosomes effected on chondrocytes through MEG-3
Various no-coding RNAs are proved to play key roles in exosomesmediated biological functions. Initially, we found that IL-1β treat-

| MEG-3 inhibited IL-1β-induced senescence and apoptosis in chondrocytes
Evidence above showed effects of BM-MSCs-exosomes on chondrocytes could reverse IL-1β-induced senescence and apoptosis.
Thus, we continued to explore the effects of its content MEG-3 on the chondrocytes. Overexpressing or knockdown MEG-3 inhibited or increased the β-gal staining intensity ( Figure 6A,B) and

| DISCUSS ION
Osteoarthritis is a degenerative disease involved in various inflammatory processes. Nonsteroidal anti-inflammatory drugs (NSAIDs) partially relieve OA symptoms. However, no obvious benefits are observed in the degeneration of cartilages and trabecular F I G U R E 5 Exosomal MEG3 from BM-MSCs induced molecular changes in chondrocytes. A, The expression of MEG-3 in chondrocytes with IL-1β or exosomes treatment detected by RT-qPCR. B, The melt curves of MEG3 were observed by RT-qPCR. (Figure 5B). C-E, The expression of MMP-13, collagen II and ADAMTS5 was detected by Western blotting. F-H, The relative expression of markers in C-E. *p < 0.05; **p < 0.01 microarchitecture of subchondral bone. In addition, long-term administration may lead to a range of serious side effects. With the capacity to alleviate the symptoms of osteoarthritis and cartilage damage, stem cell therapy is an emerging option for various diseases.
Recent studies indicated that MSCs had shown therapeutic efficacy in OA animal model and clinical trials. MSC-based tissue repair was first predicated on the hypothesis that these cells could differentiate into chondrocytes to replace the damaged tissue. However, it is now widely accepted that paracrine effects of MSC are a central mechanism of cell therapy promoting tissue regeneration, characterized by secretion of a broad spectrum of growth factors, chemokines, cytokines and EVs. 28,29 In the present study, we aimed to explore the therapeutic effects and underlying mechanisms of MSC-derived exosomes on OA.
Previous studies indicated that EVs from MSCs could alleviate osteoarthritis through balancing synthesis and degradation of cartilage extracellular matrix and promoting chondrocyte proliferation.
ACLT, a widely validated model, was performed in the current study to mimic OA features. 30 Notably, administration of MSCs-EXO significantly altered the key pathological features of OA, known as inflammation, catabolic matrix degradation and cellular apoptosis.
In particular, they improved the status of the articular cavity, cartilage and femur in OA rats and alleviated the destruction of articular cartilage. OA progression is often accompanied by increased subchondral bone remodelling that enables mechanical forces to dynamically modify its structure. Thus, we further assessed effects of treatments on histo-morphometric parameters of bone by micro-CT.
Bone formation is mainly caused by osteoblasts, and bone resorp- Actually, previous studies tried to discover the biological contents in MSCs-exosomes attributed to the therapeutic effects on OA. They found that MSC-derived exosomes suppressed IL-1β-induced apoptosis of chondrocytes and promoted cartilage repair in vivo by its content lncRNA-KLF3-AS1. 36 Exosomal KLF3-AS1 could act as a competitive endogenous RNA (ceRNA) by sponging miR-206 to facilitate GIT1 expression, which led to chondrocyte proliferation induction, apoptosis inhibition and subsequent cartilage repair. 37 Besides, exosomal Annexin A1 derived from Ad-MSCs could serve as a therapeutic factor for OA by chondroprotective effects through activation of NF-κB and activator protein-1. 38 Moreover, a recent study showed that exosomes from miR-92a-3p-overexpressing MSCs enhanced chondrogenesis and ameliorated cartilage degradation by directly targeting Wnt5a. Indeed, previous studies indicated that lncRNA MEG-3 regulated the proliferation, apoptosis and ECM degradation of chondrocytes through miR-93/TGFBR2 axis or miR-16/ SMAD7 axis. 39,40 In the current study, we explored whether the

CO N FLI C T O F I NTE R E S T
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The raw data supporting the conclusions of this article will be made available from the corresponding author on reasonable request.