Exosomes From Human Umbilical Cord-derived Mesenchymal Stem Cells Alleviate Osteoarthritis by Balancing the Synthesis and Degradation of Cartilage Extracellular Matrix and Regulating Macrophage Polarization

Background Osteoarthritis (OA) is one of the most common joint diseases and a major public health concern. Current therapies for OA can relieve symptoms but offer no potential for cartilage regeneration. Mesenchymal stem cells (MSCs) have been widely used for the treatment of OA owing to their paracrine secretion of trophic factors, a phenomenon in which exosomes may play a major role. Here, we investigated the potential of exosomes from human umbilical cord-derived MSCs (hUC-MSCs-Exos) at alleviating OA. incorporated EdU through a reaction between the alkyne group of EdU the uorescent azide copper-catalyzed azide-alkyne cycloaddition. Chondrocytes treated with 4 × 10 7 particles/mL hUC-MSCs-Exos an EdU-555 cell proliferation visualized an inverted uorescence microscope. EdU-free chondrocytes served a negative control. effect chondrocyte 4 × 10 7 particles/mL hUC-MSCs-Exos transwell assay. cells μL DMEM/F-12 h PFA min, stained with 0.5% crystal violet min, with upper upper cells not migrated to the lower chondrocytes with crystal violet lower ImageJ software. Chondrocyte apoptosis electron microscopy radioimmunoprecipitation assay phenylmethylsulfonyl polyvinylidene uoride (PVDF); cell counting kit-8 (CCK-8); specic pathogen-free (SPF); osteoarthritis research society international anterior cruciate ligament transection in combination with medial meniscectomy (ACLT+pMMx); analysis of variance (ANOVA)

and human studies (6)(7)(8)(9). Additionally, hUC-MSCs are considered a better choice compared to MSCs owing to the painless collection procedure, absence of ethical issues with respect to their use, and high proliferation potential, cell vitality, and paracrine potential (9)(10)(11). However, due to drawbacks such as the safety issues of MSCs, inconvenient storage and transportation, and differentiation for non-therapeutic purposes, there is limited research on their applications. Similar to all cell-based therapies, MSC-based therapy is associated with signi cant operational costs and challenges as cell-based medicine requires stringently monitored manufacturing and storage to ensure optimal viability and vitality of the cells at all stages (harvest, expansion, storage, and delivery to patients) (3,12,13). Moreover, many studies have shown that MSCs exert their therapeutic effects, i.e., reduction of cellular injury and enhanced repair through the secretion of reparative factors (12,14,15). Co-culture studies further demonstrated that MSCs secrete trophic factors to promote chondrocyte proliferation and matrix synthesis (16).
Exosomes have been identi ed as the principal agents mediating the therapeutic e cacy of MSCs in several diseases (14,17). Exosomes are extracellular vesicles with a size range of 30 to 200 nm (averagẽ 100 nm) that have an endosomal origin. Many exosome components are derived from their cells of origin; these include DNA, RNA, lipids, metabolites, as well as cytosolic and cell-surface proteins (17). Repair of osteochondral defects using MSC-derived exosomes was characterized by increased proliferation and in ltration of chondrocytes, and enhanced ECM synthesis (18,19). Wu et al. showed that exosomes derived from infrapatellar fat pad MSCs could protect the cartilage from damage and ameliorate gait abnormalities observed in OA mice (20). Wang et al con rmed that exosomes from embryonic MSCs exert a bene cial therapeutic effect on OA by balancing the synthesis and degradation of chondrocyte-derived ECM (21). Nevertheless, ameliorating in ammation constitutes an important component of strategies aimed at promoting cartilage repair (18,22). Zhang et al. found that exosomes derived from MSCs have an immunomodulatory effect and can induce M2 macrophages to in ltrate the defects and synovium of OA cartilage, reduce the in ltration of M1 macrophages, and downregulate the expression of IL-1β and TNF-α, thereby inhibiting the in ammatory response in OA. Similar immunoregulatory roles of exosomes derived from MSCs have been observed in in ammatory bowel disease (23), spinal cord injury (24), and diabetic peripheral neuropathy (25). The potential use of MSCderived exosomes as a novel cell-free therapy for the treatment of OA is worthy of recognition. Therefore, we investigated the regenerative and immunomodulatory effects of hUC-MSCs-Exos in a rat model of OA.

Methods
The umbilical cords were obtained from mothers of healthy full-term fetuses who had provided written informed consent at The Second Hospital of Jilin University. Consent for publication was also obtained from all donors. This study was performed in accordance with the Guidelines for Stem Cell Research and Clinical Translation (2016, ISSCR) and approved by the Ethics Committee of The Second Hospital of Jilin University (approval 2019-142).
Cell culture hUC-MSCs were isolated as described previously (10). Brie y, umbilical cords were washed with phosphate-buffered saline (PBS, Thermo Fisher Scienti c, USA) containing 5% penicillin/streptomycin solution (P/S, 100 IU/mL penicillin, 100 IU/mL streptomycin; Thermo Fisher Scienti c). Umbilical cord tissues were carefully dissected to remove vessels and then cut into pieces of approximately 1 mm 3 , followed by suspension in Dulbecco's Modi ed Eagle medium containing nutrient mixture F-12 (DMEM/F-12; Thermo Fisher Scienti c) supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scienti c), 1% P/S, and 10 ng/mL basic broblast growth factor (bFGF; PeproTech, UK). The suspension was used to coat the pre-incubated petri dishes. The explant-seeded plates were incubated for 1-2 h in an inverted position to ensure the proper attachment of the explants. Thereafter, 3 mL of fresh medium was gently added to the seeded plate, followed by incubation at 37 °C in a humidi ed 5% CO 2 incubator. The medium was changed every three days. An outgrowth of adherent cells from the tissue explants were observed after 1 week of seeding. After 10-14 days, the tissue explants were removed from the petri dishes, leaving behind the attached cells. When the hUC-MSCs had proliferated to approximately 80% con uence, cell harvesting was carried out using 0.25% EDTA-Trypsin (Thermo Fisher Scienti c), followed by centrifugation at 300 ×g for 5 min. Harvested cells were further grown to passage 4 for all the experiments. An inverted uorescence microscope (Olympus IX71, Japan) was used to monitor the phenotype and growth characteristics of the hUC-MSCs. These cells were sub-cultured and used as cell sources in subsequent experiments. Human chondrocytes (C-12710) for the knee and hip joint cartilage tissue were obtained from PromoCell's cell culture facility and cultured in a chondrocyte growth medium (PromoCell, Germany). At 24 hours after cell seeding, we changed the medium into that with 10 ng/mL interleukin 1 beta (IL-1β) (PeproTech, USA) according to the experimental design. The nest day, the medium with or without 4 × 10 7 particles/mL hUC-MSCs-Exos was added according to the experimental design, and the cells were collected and analyzed after 48 h. Normal cultured chondrocytes served as control. Human monocytic THP-1 cells (TIB-202, ATCC; USA) were cultured in RPMI-1640 medium (Thermo Fisher Scienti c) supplemented with 10% FBS, and 1% P/S. Based on a previous study (26), THP-1 cells were induced into M0 macrophages by incubation with 25 ng/mL phorbol 12-myristate 13-acetate (PMA; Sigma-Aldrich, USA) for 24 h. Once the cells became adherent, they were polarized to M1 macrophages by incubation with lipopolysaccharide (LPS; Sigma-Aldrich; 100 ng/mL) for 24 h. M0 and M1 macrophages (5 × 10 5 cells) were exposed to hUC-MSCs-Exos (4× 10 7 particles) in RPMI-1640 medium.
Immuno uorescence staining and ow cytometry assay  10 μg/mL; Thermo Fisher Scienti c) and visualized under an inverted uorescence microscope. For surface marker detection, a ow cytometry assay was performed. Cells were incubated with primary mouse anti-human monoclonal antibodies (1:100): anti-CD31, anti-CD45, anti-CD90, anti-CD105, anti-CD44, and anti-CD73. At the end of incubation, cells were washed twice with PBS and then incubated with goat anti-mouse secondary antibody (Alexa Fluor 488; 1:2000). After washing twice with PBS, target cells were analyzed on a FACS Calibur instrument (BD Biosciences, USA). Cells that were stained with only the secondary antibody served as a negative control.

Isolation and identi cation of exosomes
When hUC-MSCs reached 50-60% con uence, they were washed with PBS and cultured in conditioned medium containing DMEM/F-12, 10% non-exosomes-FBS (SBI, USA), 1% P/S, and 10 ng/mL bFGF for an additional 48 h at 37 °C in a humidi ed 5% CO 2 incubator. Exosome isolation was carried out using ultracentrifugation (28). Collected culture suspension was transferred to conical tubes for centrifugation at 300 ×g for 10 min at 4 °C to obtain the pellet. The supernatant was centrifuged a second time at 2,000 ×g for 10 mins at 4 °C to further remove cell debris, followed by a third time at 10,000 ×g for 30 min at 4°C to further remove apoptotic bodies and other organelles. Finally, the supernatant was ultracentrifuged at 120,000 ×g for 90 min at 4 °C in a SW32Ti rotor (Beckman Coulter, USA) to obtain the exosome pellet.
The pellet containing hUC-MSCs-Exos was resuspended with PBS and centrifuged at 120,000 ×g for 90 min at 4 °C. The supernatant was discarded, and the pellet containing hUC-MSCs-Exos was resuspended with 400 μL PBS and stored at -80 °C. The particle number of hUC-MSCs-Exos was quanti ed using an exosome ELISA complete kit (CD81 detection; System Biosciences, USA) following the manufacturer's instructions. Morphology of the exosomes was observed by using transmission electron microscopy (TEM). The size distribution of exosomes was measured using ZETASIZER Nano series-Nano-ZS (Malvern Instruments, UK), and was analyzed using Zetasizer software (Malvern Instrument). The exosomes were loaded onto copper grids and contrasted using 2% uranyl acetate (Sigma-Aldrich), dried, and observed using a TECNAI 12 TEM (FEI, USA). Antibodies against CD63 (1:1000; ab271286, Abcam, UK) and CD81 (1:1000; ab79559, Abcam) proteins were used to analyze the incorporation of each protein into exosomes using western ow cytometry and blotting. PKH26 (Sigma-Aldrich) was used to track the cellular entry of exosomes.

Western blotting
Cells or exosomes were lysed in a radioimmunoprecipitation assay (RIPA) and phenylmethylsulfonyl uoride (PMSF) buffer (ratio: RIPA:PMSF = 100:1; Beyotime, China) for 30 min on ice and centrifuged at 13,000 ×g for 20 min at 4 °C. The supernatant was collected, and the total protein concentration was determined using a BCA kit (Beyotime). An amount of 20 μg total protein was fractionated by using 10% SDS-PAGE and the separated proteins were blotted onto 0.22-μm polyvinylidene uoride (PVDF) membranes (Beyotime). Membranes were then blocked with 5% skim milk in TBST (10 mM Tri-HCL, 150 mM NaCl, 0.25% Tween-20, pH 7.5) at room temperature for 1 h, followed by overnight incubation with the following primary antibodies: anti-COL2A1, anti-SOX9, anti-ACAN, anti-MMP13, anti-COL1A2, and anti-ADAMTS5; anti-GAPDH (1:1000, Cell Signaling Technology) served as the protein-loading control. After washing with TBST, the membranes were incubated for 1 h at room temperature with the secondary antimouse or anti-rabbit IgG antibody (1:2000, Cell Signaling Technology). After three washes with TBST, signals were detected by chemiluminescence using the ECL-Plus detection system (TransGen, China).
The relative amount of proteins on the blots was determined using ImageJ (National Institutes of Health, USA).

qRT-PCR
Total RNA was extracted from cells using Trizol (Thermo Fisher Scienti c) according to the manufacturer's instructions and reverse-transcribed into cDNA using a reverse transcription kit (Roche, Switzerland), following the manufacturer's instructions. qRT-PCR was performed with FastStart Universal Chondrocyte proliferation, migration, and apoptosis assays The effect of hUC-MSCs-Exos on chondrocyte proliferation was evaluated using the cell counting kit-8 (CCK-8; Dojindo, Japan), as described previously (30). Cell proliferation curves were constructed by measuring the amount of formazan dye generated by the activity of cellular dehydrogenase using a microplate reader (TECAN, Austria) at a wavelength of 450 nm. The incorporated EdU was detected through a reaction between the alkyne group of EdU and the uorescent azide in a copper-catalyzed azide-alkyne cycloaddition. Chondrocytes treated with 4 × 10 7 particles/mL hUC-MSCs-Exos were measured using an EdU-555 cell proliferation kit (Ribobio, China) following the manufacturer's instructions and visualized under an inverted uorescence microscope. EdU-free chondrocytes served as a negative control. The effect of chondrocyte stimulation with 4 × 10 7 particles/mL hUC-MSCs-Exos was evaluated using a transwell assay. In brief, after digestion and counting, approximately 5 × 10 4 cells were seeded into the upper chamber of a 24-well 8-μm-pore-size transwell plate (Corning, USA). Subsequently, 600 μL of DMEM/F-12 medium containing exosomes was added into the lower chamber and 400 μL of DMEM/F-12 medium was added into the upper chamber before incubation for 24 h at 37 °C. The upper chamber was then xed with 4% PFA for 15 min, stained with 0.5% crystal violet for 10 min, and washed with PBS three times. The upper surface of the upper chamber was carefully wiped using a cotton swab to remove cells that had not migrated to the lower surface. Images were collected by using an inverted uorescence microscope. The chondrocytes stained with crystal violet that migrated to the lower chamber were counted by using ImageJ software. Chondrocyte apoptosis was detected using a FITC-annexin apoptosis detection kit (BD Biosciences) according to the manufacturer's instructions. After 24 h of passage 3 chondrocyte seeding, the DMEM/F-12 medium was replaced with medium containing 10 ng/mL IL-1β. The nest day, the medium with or without 4 × 10 7 particles/mL hUC-MSCs-Exos was added according to the experimental design, and the cells were collected and analyzed after 48 h. Normal cultured chondrocytes served as control. Chondrocyte apoptosis was detected by following the manufacturer's instructions. Samples were analyzed on a FACS Calibur instrument.

Animal studies
Male Sprague-Dawley rats (approximately 12 weeks old) weighing 300-350 g, housed in a speci c pathogen-free (SPF) animal laboratory with 12:12 h light/dark cycle, controlled temperature environment (20-26 °C), and steady humidity (40-70%), were used. All protocols involving animals were performed in accordance with the ethical guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the ethics committee of the Guangdong Medical Laboratory Animal Center Care and Use Committee (B202007-15). Based on a previous report, OA in our rat model was surgically induced by using the anterior cruciate ligament transection in combination with medial meniscectomy (ACLT+pMMx) method, without damaging the tibial surface (31). All rats were randomly divided into three groups: (1) Normal (without surgery; received articular cavity injection of 100 μL PBS on day 1 and 4 of every week from the 5th to 8th week; 6 knee joints from 3 rats, n = 6); (2) OA (rats underwent surgery and received articular cavity injection of 100 μL PBS on day 1 and 4 of every week from the 5th to 8th week after surgery; 6 knee joints from 3 rats, n = 6); (3) OA+ Exos (rats received articular cavity injection of 100 μL hUC-MSCs-Exos in PBS, 10 11 particles/mL of exosomes on day 1 and 4 of every week from the 5th to 8th week after surgery; 6 knee joints from 3 rats, n = 6). Eight weeks after surgery, rats were sacri ced, and the knee samples were harvested for the evaluation of disease progression.

Histology and immunohistochemical analysis
The tibiofemoral joints were removed, and the femoral condyles were xed in neutral-buffered formalin (containing 4% formaldehyde) for 24 h after the knee samples were harvested. The xed femoral condyles were decalci ed in EDTA (Leagene, China) for 28 days (refreshed every day) before dehydration using graded ethanol (Beijing Chemical Works, China), vitri cation using dimethyl benzene (Beijing Chemical Works), para n embedding, and tissue sectioning (5 μm), prior to staining with H&E for morphological evaluation. Tissue sections were stained using a safranin-O and fast green staining kit (Solarbio, China) and examined for matrix proteoglycan and overall joint morphology. Immunohistochemistry was performed to evaluate the function of articular chondrocytes. Rabbit anti-COL2A1, rabbit anti-MMP13, mouse anti-CD86, and rabbit anti-CD163 were used as primary antibodies (1:200; Abcam). After washing off excess primary antibodies, samples were incubated with secondary antibodies conjugated with HRP: HRP-labeled goat anti-mouse IgG (1:200; Beyotime) or goat anti-rabbit IgG antibody (1:200; Beyotime). Peroxidase-conjugated antibody (1:600; Beyotime) was diluted in 1% (w/v) BSA solution and incubated for 1 h. A DAB detection system (Beyotime) was used to visualize the sections. For quanti cation analysis, positively stained cells were counted in three randomly selected elds by using ImageJ software and the mean number of cells per high power eld (cells/HPF) was determined. Based on safranin-O and fast green staining, the osteoarthritis research society international (OARSI) score, which is a well-recognized histological scoring system (31), was used by three independent pathologists in a blinded manner to evaluate the OA progression in every sample of each group.

Statistical analyses
All experiments were performed with at least three biological replicates. Data are expressed as the mean ± standard deviation. Statistical analyses were performed using GraphPad Prism 7.0 software. Multiple comparisons were analyzed using analysis of variance (ANOVA) or Student's t test followed by Bonferroni correction. P < 0.05 was considered statistically signi cant.

Isolation and identi cation of hUC-MSCs
Cells migrated out from the UC pieces between days 7 and 14 and proliferated in the tissue culture plates (Fig. 1a). Isolated cells showed the formation of colonies with broblast morphology. Cells at passage 4 were identi ed and used in subsequent experiments. After reaching 80-90% con uence, cells adopted a spindle-like shape (Fig. 1b). The cells were collected, pooled, and subjected to phenotypic characterization and multipotency assays. When cultured in chondrogenic, osteogenic, or adipogenic medium, cells could be induced to differentiate along the chondrogenic, osteogenic, or adipogenic pathways, respectively. Chondrogenic potential was con rmed by sectioning beads and staining sulfated glycosaminoglycans using toluidine blue (Fig. 1c). Osteogenic potential was con rmed using calcium mineral deposits stained with Alizarin red (Fig. 1d). Adipogenic potential was evaluated by the observation of small cytoplasmic lipid droplets stained using Oil Red O (Fig. 1e). Following the criteria for identifying MSCs, we analyzed the surface markers of the cells. Analysis of surface antigen expression using immuno uorescence staining demonstrated that the cells were positive for CD44, CD73, CD90, and CD105, and negative for CD31 and CD45 (Fig. 1f). Flow cytometry con rmed the expression rate of the cell differentiation markers CD31 (2.3%), CD44 (95.2%), CD45 (3.2%), CD73 (96.7%), CD90 (94.3%), and CD105 (98.2%) (Fig. 1g). The spindle-shaped cells that expressed the MSC differentiation markers CD44, CD73, CD90, and CD105 and possessed multipotent differentiation potentials towards chondrocytes, osteoblasts, and adipocytes were classi ed as hUC-MSCs.

Isolation and identi cation of hUC-MSCs-Exos
TEM revealed that the vesicles from hUC-MSCs exhibited a round-shaped morphology with a diameter of 30-200 nm (Fig. 2a). Nanoparticle tracking analysis revealed that 82.5% of vesicles displayed a particle size distribution between 20 and 200 nm (Fig. 2b). Flow cytometry con rmed the expression of the vesicle differentiation markers CD63 (78.1%) and CD81 (79%) (Fig. 2c, d). Western blotting also detected the presence of exosome marker proteins, namely CD63 and CD81 (Fig. 2e). All these data indicated that hUC-MSCs-Exos were successfully isolated.

Exosomes promote chondrocyte proliferation, migration, and inhibit apoptosis
To con rm whether the chondrocytes could internalize hUC-MSCs-Exos, hUC-MSCs-Exos were labeled using red uorescent PKH26. Chondrocytes were then incubated for 24 h with PKH26-labeled exosomes. After washing with PBS, the PKH26-labeled exosomes were observed in the cytoplasm of the chondrocytes, con rming the internalization by chondrocytes. Almost all chondrocytes were positive for exosome internalization (Fig. 3a). After quantifying the exosomes using an exosome ELISA complete kit, proliferation was evaluated following the stimulation of chondrocytes with 0, 0.5, 1, 2, or 4 × 10 7 particles/mL of hUC-MSCs-Exos. Based on our concentration range, CCK-8 assay results con rmed that all concentrations of exosomes could promote the proliferation of chondrocytes in a dose-dependent manner (Fig. 3b). EdU assay results further showed that the proliferative ability of chondrocytes was notably enhanced by hUC-MSCs-Exos (Fig. 3c). The results of the transwell assay showed that the migratory ability of chondrocytes was markedly strengthened by hUC-MSCs-Exo treatment (Fig.  3d). Inhibition of apoptosis is an important factor in promoting cell proliferation. We used annexin V-FITC/PI staining and ow cytometry to detect the effect of exosomes on the inhibition of chondrocyte apoptosis (Fig. 3e). IL-1β was used to induce apoptosis (22.24%) and hUC-MSCs-Exo treatment decreased the apoptotic rate of chondrocytes (8.02%). These results clearly demonstrated the superior e cacy of hUC-MSCs-Exos in promoting chondrocyte proliferation and matrix synthesis and inhibiting chondrocyte apoptosis.

Exosomes reverse IL-1β-induced chondrocyte injury
We used a model of OA-like chondrocytes reported in a previous study (32). In our experiment, qPCR results showed that IL-1β could reduce the gene expression levels of COL2A1, SOX9, and ACAN and increase those of MMP13, ADAMTS5, and COL1A2. The OA-like chondrocyte model was successfully constructed. Interestingly, the supplementation of hUC-MSCs-Exos in chondrocytes dramatically reversed the effect of IL-1β on the gene expression of COL2A1, SOX9, ACAN, MMP13, ADAMTS5, and COL1A2 (Fig.  4a), consistent with the results obtained from the immuno uorescence staining assay (Fig. 4b). Western blotting further con rmed that IL-1β could reduce protein expression of COL2A1, SOX9, and ACAN in chondrocytes and increase the expression of MMP13, ADAMTS5, and COL1A2 at the protein level. However, more importantly, results from western blotting con rmed that hUC-MSCs-Exos could reverse the effect of IL-1β on the COL2A1, SOX9, ACAN, MMP13, ADAMTS5, and COL1A2 proteins (Fig. 4c, d). In conclusion, exosomes could reverse the injury of chondrocytes induced by IL-1β in a model of OA-like chondrocytes in vitro.

Role of exosomes in macrophage polarization
In order to evaluate the role of exosomes in macrophage activity, we performed an in vitro assay using a polarization protocol of the THP-1 monocyte cell line. THP1 was induced into M0 and M1 macrophages using PMA and LPS. When the M0 and M1 macrophages were treated using hUC-MSCs-Exos, qPCR results showed that the expression levels of anti-in ammatory factors IL-10 and ARG1 increased, whereas the expression levels of IL-1β and TNF-α decreased signi cantly (Fig. 5a). Immuno uorescence showed that compared to the control group, the number of CD163-positive cells was signi cantly increased, and the number of CD86-positive cells was signi cantly decreased in M0 and M1 macrophages treated with exosomes (Fig. 5b).

Exosomes attenuate OA progression
In the rat model, OA was surgically induced using the ACLT+pMMx method. Rats underwent surgery and received articular cavity injection of PBS or hUC-MSCs-Exos from the 5th to 8th week after surgery (days 1 and 4 of every week). Eight weeks after the surgery, knee samples were harvested for the evaluation of disease progression (Fig. 6a). No adverse events occurred in any of the experimental groups. We veri ed the potential of hUC-MSCs-Exos for OA prevention in an OA rat model (Fig. 6b). In the OA+PBS group, H&E and safranin-O and fast green staining showed that the cartilage showed moderate surface irregularity, super cial brillation, loss of proteoglycan, and loss of cartilage in the super cial zone, but none of these effects were observed when the cartilage was treated with exosomes. Based on safranin-O and fast green staining, the OARSI score was used to evaluate OA progression. The OARSI score of group OA+Exos was 1.12 ± 0.26 and that of group OA+PBS was 2.92 ± 0.42 (Fig. 6c). In the OA+PBS group, expression of COL2A1 in the cartilage decreased, whereas that of MMP13 was observed on the cartilage surface. In the OA+Exos group, joint wear and cartilage matrix loss were scarcely observed. Expression of COL2A1 and MMP13 in the cartilage was similar to that in the normal group, based on the cartilage surface (Fig. 6b, d, e). As indicated by CD86-positive cells, M1 macrophage numbers reduced in the cartilage of the OA+Exos group compared with that in the OA+PBS group (Fig. 6b, f). In contrast to the M1 macrophages, we observed an abundance of M2 macrophages, as indicated by CD163-positive cells in the cartilage of the OA+Exos group compared with those in the OA+PBS group (Fig. 6b, g). These data indicated that the exosomes attenuated OA progression and prevented severe damage to the knee articular cartilage in the rat OA model, which was caused by instability of the knee joint.

Discussion
OA is a prevalent chronic joint disease (1). Current therapies attempt to relieve the symptoms, but they cannot block or reverse the ongoing cartilage degeneration (2). Ideal treatment aiming toward achieving optimal OA joint repair should promote the regenerative properties of chondrocytes and eliminate the destructive effects of in ammation (33). Here, we report, for the rst time, the effect of exosomes derived from hUC-MSCs on the treatment of OA. Our data demonstrated that hUC-MSCs-Exos slowed the progression of OA and prevented severe damage to knee articular cartilage in the rat OA model. We also demonstrated that hUC-MSCs-Exos balance the synthesis and degradation of the cartilage extracellular matrix and regulate macrophages. In the treatment of OA, exosomes derived from hUC-MSC offer many advantages. First, many previous studies have validated the safety of hUC-MSCs-Exos, which is crucial for the design of future clinical trials(34). We did not observe any adverse events in our OA model rats administered hUC-MSCs-Exos. Second, hUC-MSCs-Exos have been reported to play a role in repairing tissue damage in vital organs, especially in the kidney (35), liver (36), lung (37), and heart (38,39). Our study demonstrated that direct administration of hUC-MSCs-Exos may effectively treat cutaneous wounding (28). Third, hUC-MSCs have been presented as the best source of exosomes because of the absence of ethical concerns, easy availability, non-invasive isolation with high yield, and greater differentiation and immunomodulatory potential (10,38). Fourth, the high abundance of miRNAs in hUC-MSCs-Exos plays an important role in maintaining cartilage homeostasis, promoting chondrogenesis, and regulating in ammation; these miRNAs include miRNA-21-5p (40,41), 146a-3p (42,43), 26a-5p (44), 100-5p (20,45), and let-7a (46). The high-throughput sequencing results of exosomes in our study showed that the abovementioned miRNAs were abundant (results not shown). We obtained the exosomes of umbilical cord-derived MSCs from three volunteers by using an ultracentrifugation isolation method. The results of TEM, ow cytometry, and particle size analysis showed that the exosomes of hUC-MSC were consistent with those reported in the literature (17,28). We used the method of ultracentrifugation, with varying speed and time, to gradually remove impurities in the sample and achieve the separation and enrichment of exosomes, which is convenient for up-scaling and minimizes residues and impurities.
Previous studies have demonstrated that exosomes derived from the human bone marrow (47) Unilateral surgical ACLT+pMMx in mature rats leads to the development of progressive cartilage degeneration. For all rat models of surgically induced OA, it is recommended that skeletally mature animals, 12 weeks of age or older, be used to best mimic the development of OA in humans. This model is also su ciently sensitive for the investigation of new therapeutic strategies that could prevent OA progression (31). In this model, exosomes by themselves were demonstrated to be able to prevent lesion progression and exert a similar regenerative effect as exosomes from other stem cells (31). In our animal experiment, both hind limb knee joints of rats were surgically intervened for the construction of an OA model. Compared to other studies using animal models(49, 51), we increased the number of exosome injections to twice a week for four weeks. Results showed that increasing the frequency of exosome injection was effective to enhance the therapeutic effect of exosomes on OA arthritis.
OA is considered a multifactorial cartilage lesion disease within a chronic in ammatory microenvironment (18). Macrophage-associated in ammation is a driver of OA structural damage and progression. Hikmat et al. suggested that targeting macrophages and macrophage-associated in ammatory pathways may be an effective means to treat OA(52). Anti-in ammatory intervention to ameliorate in ammation has been shown to promote chondrocyte survival and mitigate the risk of development to post-traumatic OA.

Conclusions
In this study, we con rmed the e cacy of exosomes derived from hUC-MSCs at treating OA. Exosomes derived from hUC-MSCs can not only promote the proliferation and migration of chondrocytes in vitro but also reverse IL-1β-associated damage. hUC-MSCs-Exos can inhibit the secretion of pro-in ammatory factors, promote the expression of anti-in ammatory factors, and regulate the polarization of macrophages. Exosomes derived from hUC-MSCs slowed the progression of early OA in vivo and prevented severe damage to the knee articular cartilage that arises as the result of the instability of the knee joint in the rat OA model (Fig. 7).

Consent for publication
All donors of umbilical cords provided their consent for publication.

Availability of data and materials
The datasets used and/or analyzed in the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.      images of elds showing red-colored nucleus of chondrocytes in each group. Hoechst 33342 staining was performed to detect nuclear localization (blue color). Scale bar: 100 μm. c, d Protein expression levels of COL2A1, ACAN, MMP-13, ADAMTS5, and COL1A2 in chondrocytes were detected using western blotting. This experiment was repeated three times. *P < 0.05, **P < 0.01, ***P < 0.001 compared to untreated chondrocytes.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.