Distinct Reparative and Immunomodulatory Signatures of Extracellular Vesicles From Human Cartilage-Derived Chondroprogenitors,Chondrocytes,and Bone Marrow Mesenchymal Stem Cells in a Rat Osteoarthritis Model

Randomization and Blinding

Animals were assigned identification numbers by the institutional animal facility at the time of housing. Formal randomization and allocation concealment procedures were not employed. Personnel responsible for routine animal monitoring were unaware of treatment allocation. Following tissue harvest, samples were processed using coded identifiers, and histological sections were labelled with study-specific numbers prior to evaluation. Modified Mankin scoring was performed in a blinded manner, with investigators unaware of treatment group allocation during outcome assessment.

Isolation of BM-MSCs

Bone-marrow aspirate was harvested from the femur transection site and transferred into pre-heparinized vacutainers. Within 45 minutes of collection, the aspirate was diluted 1:1 with Ficoll-Paque PREMIUM (GE Healthcare, Cat. No. 17-5442-02) and gently mixed, or treated with RBC lysis buffer (1:5 dilution), followed by centrifugation. Samples treated with Ficoll were centrifuged at 400g for 45 minutes, while samples treated with RBC lysis buffer were centrifuged at 480g for 5 minutes. The mononuclear cell (MNC) layer was collected, washed twice with 1X phosphate-buffered saline (PBS; GIBCO, Thermo Fisher Scientific, Cat. No: 10010023), and resuspended in a basal medium. The resuspended cells were plated into T-25 or T-75 flasks and incubated under standard culture conditions (37⁰C, 5% CO₂, humidified atmosphere). After 24 hours, non-adherent cells were removed by washing with PBS. The culture medium composition included α-MEM (Minimum essential medium, Alpha modification; Lonza, Cat. No: BE02-002F) supplemented with 10% fetal bovine serum (FBS; GIBCO, Cat. No: 10270106), human fibroblast growth factor-2 (FGF-2) at 2 ng/mL (Abclonal, Cat. No: RP01042), and 1% antibiotic–antimycotic (GIBCO, Cat. no: 15240062). Adherent cells were expanded and passaged upon reaching near confluence. Cells were maintained until passage 2 or passage 3 for further experimental applications. For expansion at passage 3, BM-MSCs were seeded at a 5000 cells/cm² density. At sub-confluence, cells were harvested for downstream analyses.

Isolation of Chondrocytes

Cartilage slices devoid of visible calcification were rinsed thrice with PBS, finely minced to fragments measuring approximately 1–3 mm² and subjected to sequential enzymatic digestion. Initially, samples were incubated in a 37°C shaking water bath with 12 IU/mL pronase (Roche Diagnostics, Cat. No: 10165913103) in serum-free DMEM/F12 (Dulbecco’s Modified Eagle’s medium, Sigma, Cat. No: D8900-1L) for 3 hours. This was followed by overnight digestion (12–14 hours) with 100 IU/mL collagenase type II (Worthington Biochemical Corporation, Cat. No: LS004176) under identical conditions. The digested debris and the resulting cell suspension were filtered through a 40 μm nylon mesh (HiMedia, Cat. No: TCP024). Viable chondrocytes were quantified using the trypan blue dye exclusion assay (Gibco, Cat. No: 15250061), and cells were seeded at 5000 cells/cm² in standard expansion medium comprising DMEM/F12 supplemented with 10% FBS, 1% antibiotic-antimycotic solution. Cultures were maintained at 37°C with 5% CO₂, and medium was refreshed every 72 hours until cells reached ∼80% confluence for downstream applications.

Isolation of FAA-CPs and MCPs

To enrich FAA-CPs, freshly isolated chondrocytes were seeded on fibronectin-coated tissue culture plates (10 μg/mL in PBS; Sigma, Cat. No: F1141) at a seeding density of 4000 cells/cm². The fibronectin coating was prepared by overnight incubation at 4 °C. After 20 minutes of cell seeding at 37 °C, non-adherent cells were gently removed with PBS, and the adherent population was maintained in standard expansion medium comprising DMEM/F12 supplemented with 10% FBS, 5 ng/mL human FGF-2, and 1 ng/mL transforming growth factor beta-2 (TGF-β2; Abclonal Cat. No: RP00452). The medium was replaced every three days. Clonal expansion was monitored under an inverted microscope, and wells containing colonies with more than 32 cells (indicating at least five population doublings) were pooled. The enriched FAA-CPs were expanded to passage 2 in T-25 flasks for further analysis and experimental use. In MCP isolation, cartilage shavings measuring 8–10 mm² from non-weight-bearing regions were selected and washed with PBS. These were incubated in standard complete stromal medium comprising DMEM/F12 supplemented with 10% FBS, 10 mM GlutaMAX™ (Gibco, Cat. No: 35050061) and 1% antibiotic-antimycotic solution. The explants were initially equilibrated for 48 hours at 37 °C in a humidified incubator. Following equilibration, the explants were exposed to 0.1% collagenase type II (Worthington, Cat. No: LS004176) for 2 hours at 37 °C to facilitate loosening of the extracellular matrix. After thorough washing to eliminate released cells and enzymatic residues, the cartilage slices were placed into fresh stromal medium in 6-well culture plates (Corning, Cat. No: 3516) and maintained undisturbed to allow cellular outgrowth from the explants. Medium changes were performed every 72 hours. Upon reaching 85–90% confluence, migrated cells were monitored and harvested using 0.125% Trypsin-EDTA (Gibco, Cat. No: 25200-072, Gibco) and expanded to passage 2 under the same culture conditions.

Flow Cytometry Analysis

BM-MSCs, chondrocytes, FAA-CPs, and MCPs were phenotypically characterized for cell surface markers using FACS analysis. Once the cells reached confluency, they were trypsinized, and 0.5 x 10⁶ cells were resuspended in each FACS tube. Appropriate fluorochrome-conjugated antibodies were added at the recommended dilutions according to the manufacturer’s protocol, with unstained cells serving as controls. Positive MSC markers (CD105, CD73, CD106, CD90), negative MSC markers (CD34, CD45, CD14), integrin markers (CD49e, CD49b, CD29), immunogenic markers (HLA ABC, HLA DR, CD86, CD80) and potential markers of chondrogenesis (CD166, CD146, Podoplanin) were included (Table S1). Data acquisition was performed using the BD FACS CytoFlex, and results were analyzed using BD FACSDiva v8.0.1.1 and FlowJo software.

Cell Cycle Analysis

DNA content-based analysis was performed to evaluate cell cycle progression in the cell groups. Cultures at approximately 70% confluence were harvested using trypsin (0.125%), fixed in ice-cold 70% ethanol and incubated in 1 μg/mL DAPI(4′,6-diamidino-2-phenylindole, Thermo Scientific, Cat. No: 62248) solution containing 0.1% Triton X-100 (Merck, Cat. No: 9036-19-5) for 30 minutes at room temperature and subjected for cytometric acquisition (10,000 events/sample). Samples were analyzed using the BD FACS CELESTA flow cytometer (BD Biosciences, USA). Data analysis was carried out using FlowJo v10.8.1 software. Initial gating (P1) was applied on forward and side scatter plots to isolate single-cell events and exclude debris and aggregates. Within this P1 population, DNA content histograms were generated to evaluate distribution across G0/G1, S, and G2/M phases using the Dean-Jett model.

qRT-PCR

Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Cat. No: 74104), following the manufacturer’s protocol. RNA purity/concentration was assessed using a NanoDrop spectrophotometer, and 280 ng of RNA from each sample was reverse-transcribed into cDNA using the Takara PrimeScript RT Reagent Kit (Takara, Cat. No: 6110A). Quantitative PCR was performed using SYBR green master mix (Takyon, Cat. No: UF-RSMT-B0701) on a QuantStudio™ 12K Flex system (Applied Biosystems), with each 10 µL reaction containing 7 ng of cDNA and gene-specific primers. qRT-PCR was performed using RNA from n = [2] independent donors, each analysed in two technical replicate wells per gene. The cycling protocol included initial denaturation at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min, with melt curve analysis to confirm product specificity. Gene expression levels of SOX9, ACAN and COL2A1 as the chondrogenic markers, COL1A1, COL10A1, and RUNX2 as hypertrophic markers were quantified, normalized to GAPDH, and analyzed using the 2^−ΔΔCt method, relative to the FAA-CP group. A comprehensive compilation of intricate primer sequences, gene identifiers, accession codes, and base pair lengths is available within Table S2.

Multilineage Differentiation

Adipogenic, osteogenic, and chondrogenic differentiation were carried out using StemPro™ differentiation kits (Thermo Fisher, Cat no: A1007201, A1007001, and A1007101). For adipogenic differentiation, cells were plated at a density of 4000 cells/cm² in 24-well plates and cultured until they reached 60% confluence. The growth medium was then replaced with StemPro™ adipogenic differentiation medium. For chondrogenic differentiation, 0.5 million cells were placed into Eppendorf tubes, centrifuged at 400g for 12 minutes to form cell pellets, and incubated undisturbed for 48 hours. Subsequently, StemPro™ chondrogenic and osteogenic media were applied to the corresponding pellets. Media were refreshed every three days for adipogenic cultures (over 21 days) and every other day for osteogenic and chondrogenic cultures (over 28 days). Control adipogenic cultures were maintained in standard growth medium throughout the same period.

Osteogenic differentiation was validated using 1% Alizarin Red S, Von Kossa, and Picrosirius Red staining. Chondrogenic differentiation was confirmed through staining with 1% Alcian Blue, Safranin O Fast Green, Toluidine Blue, and Picrosirius Red.

Immunohistochemistry (IHC): Collagen Type II and Collagen X

Following chondrogenic differentiation, collagen type II IHC was performed to confirm differentiation. Pellets underwent antigen retrieval using pronase and hyaluronidase, followed by incubation with a mouse monoclonal antibody specific to collagen type II (5 μg/mL; DSHB IIII6B3). Subsequently, the samples were incubated with an HRP-conjugated goat anti-mouse polyclonal antibody (Pierce 31430) as the secondary antibody, and staining was visualized using 3,3′-diaminobenzidine (DAB) as the chromogen, with hematoxylin used as a counterstain. Controls were included to ensure antibody specificity: negative controls (without both primary and secondary antibodies), secondary controls (without the primary antibody), and positive controls (articular cartilage sections). Stained sections were analyzed using an Olympus BX43F microscope.

Osteogenic pellets were evaluated for matrix protein formation through IHC staining for collagen type X. Antigen retrieval was performed using chondroitinase ABC and pepsin, followed by incubation with a mouse monoclonal antibody specific to collagen type X (5 μg/mL; DSHB IIII6B3). The samples were then treated with an HRP-conjugated goat anti-mouse polyclonal antibody (Pierce 31430) as the secondary antibody. Staining was visualized using DAB as the chromogen, with hematoxylin serving as the counterstain.

Flow Cytometric Profiling and Confocal Imaging of EVs

Flow cytometric analysis was performed using the exosome-Human CD63 Isolation/Detection Kit (ThermoFisher Scientific, Cat. No: 10606D) following the manufacturer’s guidelines to assess the expression of canonical surface markers on EVs. Purified EV samples were diluted in PBS and incubated with CD63-specific magnetic Dynabeads provided in the kit, enabling immunocapture of vesicles expressing CD63 tetraspanin for 18-20 hours in 4°C. After incubation, the bead-bound EVs were washed using isolation buffer, followed by staining with APC-conjugated anti-human CD81 antibody and PE-conjugated anti-human CD63 antibody for 1 hour in dark at room temperature. Samples were analyzed using a CytoFLEX LX flow cytometer (Beckman Coulter, Brea, CA, USA) with 488 nm and 638 nm lasers. Acquisition and compensation were managed using CytExpert software (Version 2.5), and downstream data analysis was performed in FlowJo v10.8 (BD Life Sciences). Unlabelled controls, including beads incubated with unstained EVs, were included to establish baseline fluorescence (Table S1).

Following flow cytometric assessment, EV-bound magnetic beads were processed for imaging using a cytocentrifugation-based deposition technique. In brief, 100 µL of the stained EV-bead suspension was loaded into a cytofunnel mounted over a frosted glass slide backed by a filter card (Thermo Scientific Shandon, Cat. No: 5991022). Centrifugation was performed using a Cytospin 4 cytocentrifuge (Thermo Fisher Scientific, Waltham, MA, USA) at 600 rpm for 5 minutes. After centrifugation, the filter card was carefully detached without disturbing the deposition, and a glass coverslip was gently placed over the concentrated region and sealed. Slides were then imaged under an Olympus FV1000 confocal laser scanning microscope using consistent laser power, pinhole diameter, and gain settings across all samples.

Western Blot Characterization of EVs

Western blot analysis was performed to validate the presence of key exosomal markers. Protein lysates were prepared from isolated EVs using RIPA buffer supplemented with protease inhibitor (Sigma, Cat. No: P8340). Protein concentration quantification was conducted using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Cat. No: 23225). A total of 40 µg protein per sample was separated on gradient SDS-PAGE gels under reducing conditions. Following electrophoresis, protein was transferred onto polyvinylidene fluoride (PVDF) membranes (Thermo Fisher Scientific, Cat. No: 88520) using wet transfer at 90V; 200 mA for 2 hours in 4°C. Membranes were incubated with 4% BSA in Tris-buffered saline containing 0.05% Tween-20 (TBST; 150 mM; Thermo Fisher Scientific, Cat. No: 9005-64-5) for one hour at room temperature to minimize nonspecific binding. For primary antibody incubation, samples were treated overnight at 4°C with the following antibodies, diluted in TBST (150 mM) with 4% BSA: rabbit monoclonal anti-CD9 (Cell Signalling Technology, Cat. No: 13174; 1:1000), EpCAM (D13) (Cell Signalling Technology, Cat. No: 2626; 1:1000), HSP70 (D69) (Cell Signalling Technology, Cat. No: 4876; 1:1000), Flotilin-1 (D2V7J) (Cell Signalling Technology, Cat. No: 18634; 1:1000), and Annexin-V (Cell Signalling Technology, Cat. No: 8555; 1:1000), along with mouse monoclonal anti-Alix (3A9) (Cell Signalling Technology, Cat. No: 2171; 1:1000). After three 10-minute washes in TBST, membranes were incubated with appropriate secondary antibodies: HRP-conjugated goat anti-mouse IgG (Cell signalling technology, Cat. No: 7076; 1:1000) or goat anti-rabbit IgG (Cell signalling technology, Cat. No: 7074; 1:1000) for 1 hour at room temperature. Protein detection was achieved using Clarity™ Western ECL Substrate (Bio-Rad, Cat. No: 1705060), and chemiluminescent signals were imaged on a ChemiDoc Touch Imaging System (Bio-Rad, Hercules, CA, USA).

Zeta Sizer Analysis

The size distribution and hydrodynamic diameter of EVs in suspension were analyzed using a Zetasizer Ultra instrument (Malvern Panalytical, UK). EV samples were diluted 1:1000 in sterile distilled water and gently mixed to avoid aggregation. The diluted samples were loaded into the capillary tubes provided in the Zeta sizer kit, sealed with the scrapping wax, and placed into the instrument’s sample holder. Measurements were carried out at 250C using the default settings of the instrument. Each sample was analyzed in triplicate, and the average particle size in suspension (hydrodynamic diameter) values were recorded. Data were processed using ZetaView software, and all procedures were conducted according to the manufacturer’s guidelines to ensure consistency and reliability.

Nanoparticle Tracking Analysis

The concentration and size distribution of EVs were assessed using NTA on a NanoSight NS300 instrument (Malvern Panalytical, UK) equipped with a camera and controlled via NTA software version 3.2. Before analysis, EV samples were subjected to a stepwise dilution using sterile distilled water to reach the optimal particle concentration suitable for the linear detection range of the instrument. Initially, the sample was diluted 1:1000 in sterile distilled water, followed by two successive 1:10 dilutions, resulting in a final working dilution of 1:100,000 to achieve the final working dilution. The diluted EV suspensions were introduced into the sample chamber with the help of 0.2 μm syringe filter attached to the syringe. For each sample, three technical replicates were recorded, under consistent capture settings. Particle tracking was conducted in scatter mode using the built-in software with the detection threshold set to 5 and automatic settings enabled for blur and maximum jump distance parameters. Post-capture analysis included the determination of mean, mode, and standard deviation of particle sizes, as well as the total particle concentration per mL. To estimate EV yield, the raw particle count was corrected by the cumulative dilution factor.

Induction of OA Using the MIA Model

OA was induced using the MIA model. Prior to intra-articular injections, rats were anesthetized using an intraperitoneal injection of ketamine (50 mg/kg) combined with xylazine (5 mg/kg). After confirming adequate depth of anaesthesia, both knee joints were prepared under aseptic conditions. OA with 8% MIA was induced by bilateral intra-articular injection of 2 mg MIA reconstituted in 50 µL of sterile PBS into the tibiofemoral joints. Following injection, the joints were gently flexed and extended to ensure uniform distribution of the reagent. Animals were allowed a one-week induction period to establish early osteoarthritic changes.

Experimental Groups and EV Administration

Following OA induction, animals were randomly assigned to experimental groups. Sham control animals underwent unilateral plain needle puncture without injection (n = 2), while OA control animals received intra-articular PBS injections (50 µL; n = 2). Treatment groups received EVs derived from different human cell populations: BM-MSC-EVs (40 µg in 50 µL PBS; n = 6), chondrocyte-EVs (40 µg in 50 µL PBS; n = 6), FAA-CP-EVs (40 µg in 50 µL PBS; n = 6), and MCP-EVs (40 µg in 50 µL PBS; n = 6). As injections were performed bilaterally, each animal contributed two joints for analysis, resulting in a total of 12 joints per treatment group (n = 6 animals). EVs were administered via intra-articular injection once weekly from week 2 to week 7 post-MIA induction.

Tissue Harvesting and Histological Evaluation

At the end of the eighth week, animals were euthanized following approved protocols. Tibiofemoral joints and spleens were harvested for downstream analyses. Joint specimens were fixed, decalcified, processed, and sectioned for histological evaluation. Sections were stained with hematoxylin and eosin (H&E) for general morphology, Masson’s trichrome for collagen organization, and Safranin O–Fast Green for proteoglycan content. Cartilage repair was assessed using a modified Mankin scoring system evaluating cell morphology (0–4), matrix staining with Safranin O–Fast Green (0–3), cartilage thickness (0–2), and integration with adjacent host cartilage (0–2), resulting in a total score ranging from 0 to 11, with higher scores indicating greater cartilage degeneration. Blinded histological scoring was performed to assess cartilage repair. Immunohistochemical staining for cartilage-specific markers, including collagen type II and collagen type X, was carried out on joint sections to evaluate the quality of regenerated cartilage. Histological sections were assigned coded identifiers prior to evaluation, and scoring was performed independently by two investigators blinded to treatment allocation as per the Modified Mankin score ²⁷ (Table S3).

Quantitative Image Analysis of Collagen II and Collagen X Immunostaining

Collagen II- and Collagen X-immunostained sections were whole-slide imaged using the EVOS FL Auto Full Scan Imaging System (Life Technologies) at 10X magnification under standardised brightfield illumination and exposure settings. Regions of interest comprising the full articular cartilage thickness at anatomically comparable femorotibial sites were selected uniformly across the four EV-treatment groups following image calibration in FIJI (ImageJ) software. Staining intensity within each region of interest was quantified by colour deconvolution to isolate the DAB-specific channel, followed by measurement of mean grey value under identical thresholding and acquisition parameters across all samples. Group-wise comparisons (BM-MSC-, chondrocyte-, FAA-CP-, and MCP-EV) were performed by one-way ANOVA with Tukey’s HSD post hoc test.

Spleen Processing and IHC Analysis

Harvested spleens were fixed in 10% neutral buffered formalin (NBF) for 10 days to ensure preservation of tissue architecture. Following fixation, tissues were washed with PBS, processed, and embedded in paraffin. Sections were cut and stained with hematoxylin and eosin to assess overall splenic architecture. For IHC analysis, antigen retrieval was performed using 10 mM citrate buffer. Sections were incubated overnight at 4 °C with primary antibodies against interleukin-6 (IL-6; 1:250 dilution) and tumor necrosis factor-alpha (TNF-α; 1:100 dilution). The following day, sections were incubated with appropriate secondary antibodies for 30 minutes at room temperature, followed by visualization using DAB chromogen. Hematoxylin was used as a counterstain. These analyses were performed to assess systemic immunological responses following EV administration. Immunoreactivity for IL-6 and TNF-α was semi-quantitatively assessed, and a cut-off value of 30% was established based on the maximum baseline cytokine expression observed in sham control spleens. Values exceeding this threshold were considered indicative of a positive systemic inflammatory response.

Histological Assessment of Cartilage Repair

Eight weeks following MIA-induced OA, with intra-articular EV injections administered once weekly from week 2 through week 7, tibiofemoral joints were harvested and evaluated histologically. (Figure 5). Hematoxylin and eosin staining revealed cartilage erosion, surface fibrillation, and loss of structural integrity in OA control joints. In contrast, all EV-treated groups demonstrated improved cartilage morphology, with partial restoration of surface continuity and defect filling (Figure 5A).

OPEN IN VIEWER

Figure 5.

Histological evaluation of tibiofemoral joints following EV treatment (H&E and Masson’s trichrome). (A) Hematoxylin and eosin staining of sham, OA control, BM-MSC-EV, chondrocyte-EV, FAA-CP-EV, and MCP-EV groups at week 8. (B) Masson’s trichrome staining of corresponding sections demonstrating cartilage matrix distribution and collagen organization. Scale bars included. Hematoxylin stains nuclei blue-purple, eosin stains cytoplasm pink-orange; Trichrome stains cartilage matrix green, nuclei dark blue, and calcifying cartilage red

Among the EV-treated groups, cartilage-resident EVs, including chondrocyte-derived EVs, FAA-CP-derived EVs, and MCP-derived EVs, consistently exhibited improved structural organisation compared to BM-MSC-derived EVs as revealed by Masson’s Trichrome staining (Figure 5B).

Proteoglycan Content and Modified Mankin Scoring

Safranin O staining revealed marked proteoglycan depletion in OA control joints, confirming disease induction (Figure 6). All EV-treated groups showed increased Safranin O uptake relative to OA controls, indicating preservation or restoration of proteoglycan-rich cartilage matrix.

OPEN IN VIEWER

Figure 6.

Proteoglycan evaluation using Safranin O and modified Mankin scoring. Representative images of Safranin O–stained tibiofemoral sections from sham, OA control, BM-MSC-EV, chondrocyte-EV, FAA-CP-EV, and MCP-EV groups. FAA-CP-EV and MCP-EV groups showed greater proteoglycan retention. Orange-red staining indicates proteoglycan-rich cartilage. Modified Mankin scores comparing structural integrity and matrix composition across groups (n=12/group). Data were analysed using one-way ANOVA followed by Tukey’s HSD post-hoc test. BM-MSC-EV-treated joints showed significantly higher Modified Mankin scores than chondrocyte-EV (p = 0.006), FAA-CP-EV (p = 0.014), and MCP-EV (p = 0.010) groups; no significant differences were observed among the cartilage-resident EV groups (p > 0.05)

Notably, joints treated with cartilage-resident EVs demonstrated greater and more uniform Safranin O positivity compared to BM-MSC-EV-treated joints. Within the cartilage-resident groups, FAA-CP-EV and MCP-EV treatments showed staining patterns that were largely comparable to each other and modestly more pronounced than those observed with chondrocyte-derived EVs.

Quantitative assessment using modified Mankin scoring demonstrated marked improvement in histological scores across all EV-treated groups relative to OA controls (Figure 6). Cartilage-resident EV–treated joints exhibited significantly lower Mankin scores compared to BM-MSC-EV–treated joints (Tukey HSD: chondrocyte-EV vs BM-MSC-EV, p=.006; FAA-CP-EV vs BM-MSC-EV, p=.014; MCP-EV vs BM-MSC-EV, p=.010), reflecting improved cartilage architecture, surface integrity, and matrix composition. FAA-CP-EV and MCP-EV groups showed closely aligned scores, with both groups demonstrating comparable repair outcomes to chondrocyte-derived EVs (all pairwise comparisons among cartilage-resident groups, p>0.05, Tukey HSD).

IHC Evaluation of Cartilage Matrix Composition

IHC staining for Collagen type II demonstrated positive matrix deposition within regenerated cartilage regions in all EV-treated groups (Figure 7A). Collagen II staining was more continuous and evenly distributed in joints treated with cartilage-resident EVs compared to BM-MSC-EV–treated joints.

OPEN IN VIEWER

Figure 7.

Immunohistochemical evaluation of collagen II and collagen X. (A) Representative immunohistochemical staining and ImageJ-based quantification of collagen II expression, demonstrating hyaline cartilage matrix deposition in regenerated chondral regions across all groups. (B) Representative immunohistochemical staining and ImageJ-based quantification of collagen X expression, demonstrating hypertrophic chondrocyte distribution at the cartilage–bone interface across all groups. Positive immunoreactivity is indicated by brown staining. Scale bars as shown

Collagen type X staining was primarily localised to the cartilage–bone interface across all groups, consistent with endochondral and hypertrophic changes within the repair tissue (Figure 7B). No marked differences in the spatial distribution or intensity of Collagen X staining were observed between EV-treated groups.

Systemic Immunogenicity Assessment Following EV Administration

To assess systemic immune responses following xenogeneic EV administration, spleen tissues were evaluated at week 8 (Figure 8). Hematoxylin and eosin staining demonstrated preservation of overall splenic architecture across all EV-treated groups, with no gross disruption of splenic white or red pulp regions observed (Figure 8A).

OPEN IN VIEWER

Figure 8.

Immunogenicity assessment in spleen tissue following EV administration. (A) Hematoxylin and eosin staining of spleen sections from OA control, BM-MSC-EV, chondrocyte-EV, FAA-CP-EV, and MCP-EV groups at week 8. (B) IL-6 immunohistochemistry across seven groups: positive control, sham control, OA control, and EV-treated groups. BM-MSC-EVs exhibited the lowest IL-6 staining intensity. (C) Immunohistochemical quantification of IL-6 and TNF-α expression in spleen tissue following extracellular vesicle treatment. Data are expressed as percentage of splenic immunopositive cells (mean ± SEM). The dashed line indicates the sham-derived cut-off (30%). (D) TNF-α immunohistochemistry across corresponding groups demonstrating relative TNF-α expression. BM-MSC-EVs showed least TNF-α staining. Brown staining indicates inflammatory marker expression

IHC analysis revealed differential expression of IL-6 and TNF-α among treatment groups. BM-MSC-EV–treated animals exhibited the lowest expression levels of both IL-6 and TNF-α, consistent with the established immunomodulatory properties of MSC-derived vesicles. In contrast, cartilage-resident EV–treated groups demonstrated detectable IL-6 and TNF-α expression, with staining patterns resembling those observed following chondrocyte-derived EV administration (Figure 8B-D).