Introduction

Osteoarthritis (OA) is the most prevalent form of degenerative joint disease, associated with symptoms such as pain, swelling, stiffness, and restricted joint mobility. Current treatment strategies focus on symptom management and disease progression control through multifaceted approaches including patient education, physical exercise, pharmacological interventions, and regenerative therapies [1, 2]. Notably, mesenchymal stem cells (MSCs) have demonstrated significant potential for OA treatment by alleviating pain, enhancing joint function, and promoting cartilage regeneration [3,4,5].

The Hoffa’s fat pad (HFP), also known as infrapatellar fat pad, plays a significant role in joint health and function. It is an intracapsular and extrasynovial adipose tissue on a fibrous scaffold located within the knee joint [6]. Located strategically, the HFP functions as a shock absorber, reducing mechanical stress on adjacent knee structures and contributing to the maintenance of joint stability during motion [7]. Moreover, HFP contains MSCs which have potential applications in cartilage regeneration and repair. Due to its intra-articular localization, HFP-MSCs possess better chondrogenic differentiation capacity than MSCs from subcutaneous adipose tissue or bone marrow [8, 9]. It has been shown in clinical trials that HFP-MSCs, either alone or combined with blood products, have disease-modifying effects in knee OA [10, 11]. However, MSCs exert their main therapeutic effects through paracrine signaling, releasing a diverse array of biologically active molecules, many of which are encapsulated within extracellular vesicles (EVs).

EVs are small, nano-sized particles released by cells into the extracellular environment. They serve as key mediators of cell-to-cell communication by transporting bioactive molecules, including proteins, lipids, and nucleic acids, which influence diverse physiological and pathological processes. The application of MSC-EVs represents a promising cell-free therapeutic strategy, offering a potential solution to the challenges associated with viable cells administration [12,13,14]. MSC-EVs have demonstrated the ability to enhance cartilage regeneration and suppress inflammation [15,16,17]. Their dual role in modulating inflammatory responses and facilitating cartilage matrix reconstruction is crucial for effective OA management [18].

OA is a complex disease involving multiple cell types, with chondrocytes, synovial fibroblasts, macrophages, and other immune cells playing significant roles in its pathogenesis. The primary aim of this study was to evaluate the influence of HFP-MSC-EVs on chondrocytes, synovial fibroblasts, and macrophages. The study by Kouroupis et al. on HFP-MSC-EV cargo analysis revealed a strong presence of factors that play crucial roles in regulating inflammatory and pain-related pathways [19]. Given that osteoarthritis (OA) is a disease characterized by inflammation and pain, and findings that HFP-MSCs possess superior chondrogenic capacity than MSCs from other fat tissues or bone marrow, we hypothesize that HFP may serve as a valuable source of MSC-EVs for cell-free therapy in OA [20, 21]. Various studies have demonstrated that efficacy of EVs can be influenced by priming cells of origin with specific molecules (e.g. cytokines, lipopolysaccharide) [22, 23]. Blood-derived products are increasingly being explored for their potential to improve outcomes in OA [24,25,26]. In addition, human blood-derived supplements, such as platelet lysate and plasma, are being explored as xeno-free alternatives to fetal bovine serum (FBS) for MSC expansion, addressing safety and regulatory concerns in clinical applications [27, 28]. Our previous studies revealed enhanced viability, metabolic activity and chondrogenic potential of HFP-MSCs with the use of blood products; platelet-rich plasma (PRP) and hyperacute serum (hypACT) [29, 30]. Therefore, we examined whether priming HFP-MSCs with blood-derived products, PRP and hypACT, could alter the therapeutic properties of the resulting EVs. The therapeutic efficacy of EVs could be further enhanced by improving their uptake by target cells [31, 32]. Thus, we explored whether polybrene (hexadimethrine bromide), a polycationic polymer originally used to reverse heparin-mediated anticoagulation, enhances EV internalization by target cells to improve treatment efficacy [33, 34].

Methods and materials

The workflow used in this research is summarized in Fig. 1.

Fig. 1
figure 1

Schematics of the experimental workflow (Created in BioRender. Gulová, S. (2025) https://BioRender.com/kkqzqx6)

Full size image

Tissue harvest

Cartilage (n = 3) and HFP tissue (n = 3) were obtained from different osteoarthritic patients undergoing total joint replacement surgery after approval by the ethics committee of Karl Landsteiner Private University (EK 1020/2020) and after obtained patient informed consent.

Cell isolation and expansion

Chondrocytes were isolated following a previously established protocol [35]. In brief, cartilage was dissected from femoral condyles, cut into 2–3 mm³ fragments, and digested with 0.2 WU/mL Liberase (Roche, Vienna, Austria, #05401119001) in DMEM-F12 GlutaMAX medium (GIBCO, Waltham, MA, USA, #31331-028) containing 200 U/mL penicillin, 0.2 mg/mL streptomycin, and 0.5 µg/mL amphotericin B (Sigma-Aldrich, St. Louis, MI, USA). The resulting cell suspension was filtered through a 40-µm strainer (BD Biosciences, Franklin Lakes, NJ, USA, #352340) to eliminate undigested material. Cells were then seeded at 10⁴ cells/cm² in the same medium, supplemented with 10% FCS. HFP-MSCs were isolated as described previously (Neubauer et al., 2019). Briefly, HFP tissue was minced into 5-mm pieces using a scalpel and digested with 9000 units of collagenase I (Sigma-Aldrich, #C0130) per 10 g of tissue in 15 ml DMEM (GIBCO, #21969-035) for 2 h at 37 °C on a roller in a 50-ml tube. The suspension was filtered through a cell strainer (Corning, #431,750) to remove undigested material, washed once with PBS, and resuspended in growth medium composed of high-glucose DMEM with GlutaMAX™ and pyruvate (GIBCO, #31966-047), 2% penicillin/streptomycin (Sigma-Aldrich, #P4333), 1% amphotericin B (Sigma-Aldrich, #A2942), 10% heat-inactivated fetal calf serum (GIBCO, #11550-356), 1% nonessential amino acids (GIBCO, #11140-050), and 1 ng/ml bFGF (Sigma-Aldrich, #F0291). Cells were seeded at 10,000 cells/cm² in T175 flasks and cultured at 37 °C with 5% CO₂ for expansion. Synovial fibroblasts were thawed (courtesy of Associated Tissue Bank), seeded in T75 flasks and cultured at 37 °C with 5% CO₂ for expansion.

Preparation of blood products

Whole blood samples were collected in-house from 5 to 7 volunteers, after obtaining informed consent. Donors were required to be between 25 and 45 years old and in good health on the day of donation, while those with conditions such as pregnancy, underweight, or diabetes were excluded based on a screening form. Platelet-rich plasma (PRP) was prepared using citrate-coated vacutainer tubes (VACUETTE 9NC trisodium citrate 3.2%, Greiner Bio-One, #455,322) and processed according to established protocols [36]. PRP samples were pooled and stored at -20 °C until use. For hypACT preparation, freeze-dried serum powder (courtesy of Orthosera GmbH) was reconstituted with 2 ml of sterile double-distilled water.

Extracellular vesicles enrichment and characterisation

Extracellular vesicles production in bioreactor

To produce EVs, HFP-MSCs were seeded into vertical bioreactor culture chambers (PBSMini, PBSBiotech) as described previously [37]. Briefly, on average 3,5 × 106 cells at maximum passage number 5, were mixed with 1.6 g Synthemax II microcarrier beads (Corning, #3781), corresponding to a seeding density of ~ 6000 cells/cm². The cell-bead suspension, in 30 ml growth medium (composed of high-glucose DMEM with GlutaMAX™ and pyruvate (GIBCO, #31966-047), 2% penicillin/streptomycin (Sigma-Aldrich, #P4333), 1% amphotericin B (Sigma-Aldrich, #A2942), 10% heat-inactivated fetal calf serum (GIBCO, #11550-356), 1% nonessential amino acids (GIBCO, #11140-050), and 1 ng/ml bFGF (Sigma-Aldrich, #F0291)), was incubated on a roller at 37 °C for 1 h to allow cell attachment. The suspension was then transferred to a PBSMini vertical wheel bioreactor (PBSBiotech, USA) with an additional 50 ml growth medium, and the rotation speed was adjusted from 18 to 25 rpm to maintain bead suspension. 20 ml of growth medium was refreshed daily during the expansion phase. Afterwards, medium was switched to serum-free medium with 5% FCS, CPRP, or hypACT, and 2 U/ml heparin for 24 h, following washing with serum free medium, to get rid of any FCS residuals. The cell-bead suspension was after priming transferred to 50 ml tubes and washed once with serum-free medium by centrifugation at 200 × g for 5 min to get rid of any blood residuals. After washing, the cells were resuspended in serum-free medium (composed of high-glucose DMEM with GlutaMAX™ and pyruvate (GIBCO, #31966-047), 2% penicillin/streptomycin (Sigma-Aldrich, #P4333), 1% amphotericin B (Sigma-Aldrich, #A2942)), returned to the bioreactor chamber, and cultured in a total volume of 80 ml for 48 h to allow for EV collection. Conditioned supernatant was harvested after gravitational settling, centrifuged (4000 × g, 15 min) to remove debris, and stored at -80 °C for further use.

Monitoring metabolic activity of MSCs in bioreactor culture via XTT assay

To assess metabolic activity of cells in the bioreactor culture, 100 µl of the cell suspension was sampled, and 20 µl of this was mixed with 80 µl of growth medium per well in 96-well plates (triplicates). Subsequently, 50 µl of XTT reagent, prepared according to the manufacturer’s instructions (Roche, #11,465,015,001), was added to each well. Following a 4-hour incubation at 37 °C and 5% CO₂, absorbance was recorded at 492 nm, with 690 nm as a reference wavelength, using a Synergy 2 plate reader. Blank wells containing only medium and XTT reagent served as controls.

Conditioned media ultrafiltration

Ultrafiltration of conditioned media was carried out following previously described method [37]. Briefly, conditioned media stored at -80 °C were thawed at 37 °C in a water bath. To prepare 100 kD Amicon ultrafiltration units (Merck, #UFC9100), the filters were prerun with PBS, 70% ethanol, PBS again, and ultrafiltrated 1% BSA in PBS to ensure sterility and minimize EV attachment. Media were processed in 15 ml aliquots by centrifugation (4,000 × g, 15 min, 4 °C) using the same filter unit per sample to accumulate retentates. After filtering, EVs were washed with 15 ml PBS and centrifuged again. The primary retentate (“EV”) was recovered in low-attachment tubes (Biozym, #710,176). Residual EVs were detached by adding 1 ml PBS, vortexing at 2,000 rpm for 20 s, recovering in low-attachment tubes and repeating the process two times for quantitative EV recovery.

Protein quantification

Total protein content was measured using the Micro BCA™ Protein Assay Kit (Thermo Scientific, #23235) in a microplate format, following manufacturer’s instructions. Briefly, 50 µl of either undiluted or appropriately diluted samples were used. To release EV protein content, samples were lysed with 10 µl RIPA buffer for 10 min at 4 °C. A standard curve (0 µg/ml to 200 µg/ml) was prepared using bovine serum albumin. Following a 2 h incubation at 37 °C, absorbance was measured at 562 nm using a plate reader.

Western blot analysis

EVs were lysed in RIPA buffer (Sigma, #R0278) supplemented with phosphatase and protease inhibitor cocktail (Thermo Scientific, #78440) by vigorous pipetting, followed by a 15-minute incubation at 4 °C. After protein quantification, 20 µg of total protein were separated on 10% or 4–12% gradient SDS-PAGE pre-cast gels (Invitrogen, #NP0301 and #NP0321). Samples were reduced with 1 µl beta-mercaptoethanol when detecting ApoB100, calnexin and syntenin. Protein samples were mixed with LDS sample buffer (Invitrogen, #NP0007) and heated at 95 °C for 10 min before loading. Proteins were transferred to PVDF membranes using semidry blotting. CD63 (BioLegend, #353,005) and CD9 (System Biosciences, #EXOAB-CD9A-1), syntenin (Cell Signaling, #27964), calnexin (Santa Cruz, #sc-46669) and ApoB100 (MYBioSource, #MBS2044849) antibodies, all diluted 1:1000 in 1% BSA in 1X PBS, were applied for detection. Enhanced chemiluminescence (ECL) using WesternBright ECL substrate (Advansta, #K-12,045-D20) was used for visualization. Blot images were processed with automatic white correction in GIMP software (version 2.10).

Nanoparticle tracking analysis

The size and concentration of particles were determined using nanoparticle tracking analysis (NTA) in scatter mode, while the presence of CD9 and CD73 markers was assessed in fluorescence mode on a ZetaView® TWIN device. Samples were diluted 1:1000 in PBS. In scatter mode, video acquisition was conducted at 11 positions, recording at 30 frames per second for 2 s per position, with camera sensitivity set to 80 and shutter speed at 100. Next, for detecting CD9 and CD73 in EV samples, antibodies were pre-diluted 1:10 in PBS. A mixture of 8 µl EV suspension and 1 µl of each prediluted antibody was incubated in the dark for at least 1 h. The stained EV suspensions were then diluted 1:333 in PBS and analyzed at a camera sensitivity of 90 to determine the presence of CD9- and CD73-positive particles. Video acquisition and analysis were conducted using Zeta navigator 1.0.3.8.6 and ParticleExplorer 4.1.0.16. To visualize the data from scatter and fluorescence mode in a graph, a correction factor was applied, calculated from the correlation curve between camera sensitivity and particle concentration in a given sample.

Cryo-electron microscopy

Different aliquots (2–4 µl) of EVs suspended in PBS were deposited on 1 μm mesh copper grids. Afterwards, samples were blotted for 1–2 s and plunge frozen in liquid ethane. A Glacios cryo-electron microscope was used to visualize EVs.

Confocal fluorescence microscopy for imaging cellular uptake of CFSE-labeled EVs

HFP-MSC-EVs were labelled using CFSE ( Invitrogen™, #C34570). Briefly, approximately 1.44E + 10 of HFP-MSC-EVs were mixed with 9 µl of 500 µM solution of CFSE in PBS resulting in a working concentration of 15 µM and incubated for 2 h at 37 °C in dark. Next, stained EVs were purified by Ultrafiltration in two Vivaspin 500 Ultrafiltration tubes (Sartorius, #VS0141) with a 100,000 MWCO cutoff. EVs were washed three times with PBS and the retentate was diluted with PBS following two additional washes to quantitatively transfer EVs. The particle concentration and presence of stained EVs was confirmed via NTA on a ZetaView® TWIN device. Video acquisition and analysis were conducted using Zeta navigator 1.4.7.6 and ParticleExplorer 4.3.4.4 softwares.

Chondrocytes and synovial fibroblasts were seeded into four-well Nunc Lab-Tek chamber slides (Thermo Fisher) at a density of 1 × 104 cells/well and cultured in DMEM-F12 GlutaMAX medium (GIBCO, Waltham, MA, USA, #11524436) containing 200 U/mL penicillin, 0.2 mg/mL streptomycin, and 0.5 µg/mL amphotericin B (Sigma-Aldrich, St. Louis, MO, USA). After cell attachment, the culture medium was changed to serum-free medium, either supplemented with 2 µg/mL polybrene (plb+) (Sigma-Aldrich, #H9268) or left unsupplemented (plb−, control). Labeled EVs were added to each group and incubated for 18 h in the dark. To detect internalization, we added 2 × 104 EVs per cell to media as higher doses provide higher spot count, intensity and maximum pixel [38]. After incubation, the cells were washed and fixed with 4% paraformaldehyde (ROTH) for 15 min at room temperature. The cells were then washed three times with PBS before mounting medium containing DAPI (Santa Cruz Biotechnology) was added. A coverslip was placed on top, and the slides were stored in the dark for 2 h to allow proper staining of the nuclei. Uptake was detected via confocal microscopy (Leica TCS SP8/ LAS X, Version 3.5.7.23225) at 63× magnification. To analyze the particle number, images were processed in ImageJ (2.14.0/1.54f) in several steps. First, images were converted into 8-bit format, and the threshold was set to a lower threshold level of 99 and an upper threshold level of 255. Next, images were converted into binary format, and the watershed function was applied to separate merging particles. Finally, particles were analyzed in the region of interest with a size range of 2–23 pixels.

HFP-MSCs tri-lineage differentiation

Tri-lineage differentiation was conducted on three donors. Cells were cultured in either growth medium or differentiation media, supplemented with 10% CPRP, 10% hypACT, or 10% FCS as a control. Since PRP required the addition of 2 U/ml heparin (Gilvasan Pharma GmbH, #3909969), heparin was added uniformly across all experimental conditions. The differentiation experiments lasted 3 weeks, and media were changed twice weekly.

Adipogenic differentiation

The adipogenic differentiation potential of HFP-MSCs was evaluated via Oil Red O staining, following the method described by Kuten et al. [36]. Briefly, cells were seeded in 6-well plates at a density of 2 × 10⁵ cells and cultured in normal growth medium until reaching 100% confluency. Differentiation was then induced using serum-free growth medium supplemented with StemXVivo® Adipogenic Supplement (Bio-Techne Ltd., Abingdon, UK) and either 10% CPRP, 10% hypACT, or 10% FCS (control). After three weeks, cells were washed with PBS, fixed in 10% formalin for 30 min, and stained with Oil Red O (Sigma, #01391) following the manufacturer’s protocol. Fixed cells were then washed with distilled water, treated with 60% isopropanol for 2 min, and incubated with Oil Red O working reagent (prepared in a 3:2 ratio of stock solution to distilled water) for 5 min. Excess dye was removed by rinsing with distilled water until no residual staining remained. Stained cells were imaged using a phase-contrast microscope. To quantify lipid accumulation, the incorporated dye was extracted with 300 µl of 100% isopropanol, centrifuged at 14,000 × g for 5 min, and the absorbance of 80 µl of supernatant was measured in triplicate at 490 nm using a Synergy 2 plate reader.

Osteogenic differentiation

The osteogenic differentiation potential was assessed using Alizarin Red S staining, as described by Neubauer et al. [29]. Briefly, MSCs were seeded at a density of 2 × 10⁵ per well in 6-well plates and maintained in normal growth medium until reaching full confluency (3–5 days), at which point differentiation was induced. The differentiation medium consisted of serum-free growth medium supplemented with 100 nM dexamethasone (Sigma-Aldrich, #D4902), 50 µg/ml ascorbic acid (Sigma-Aldrich, #A4403), and 10 mM β-glycerol phosphate (Sigma-Aldrich, #G9422), along with 10% CPRP, 10% hypACT, or 10% FCS (control). After 3 weeks, cells were washed with PBS, fixed in 10% formalin for 30 min, and stained with 2% aqueous Alizarin Red S solution (pH 4.3, Sigma-Aldrich, #A5533) for 30 min at room temperature. Stained cells were then rinsed three times with distilled water, and mineralized areas were visualized under a light microscope. For quantification, stained cells were incubated with 500 µl of 10% acetic acid for 30 min, collected using a cell scraper, vortexed, and heated at 85 °C for 10 min. The mixture was then cooled, centrifuged at 20,000 × g for 15 min, and the supernatant was neutralized with 500 µl of 10% ammonium hydroxide. The absorbance of 100 µl of the neutralized solution was measured in triplicate at 405 nm using a Synergy 2 plate reader.

Chondrogenic differentiation

Chondrogenic differentiation was assessed via chondropellet formation as previously described [35]. In brief, 2.5 × 10⁵ MSCs were suspended in chondrogenic medium containing DMEM high glucose (GIBCO, #10596010), 1% ITS liquid media supplement (Sigma-Aldrich, #I3146), 100 nM dexamethasone (Sigma-Aldrich, #D4902), 50 µg/ml ascorbic acid (Sigma.Aldrich, #A4403), 1% non-essential amino acids (GIBCO, #11140050), 5 ng/ml TGFβ-3 (PeproTech, #AF-100-36E), 4% methylcellulose (Sigma-Aldrich, #M7027), and either 10% CPRP, 10% hypACT, or 10% FCS (control). Cells were centrifuged at 4164 × g for 10 min and cultured at 37 °C, 5% CO2 in 15 ml tubes with loose cap. After three weeks, pellets were recovered, and excess liquid was removed. Pellets were washed with 1x PBS and fixed in 10% formalin overnight at 4 °C. They were then placed in base molds (Fisher Scientific, #22-363-554) pre-filled with a frozen layer of tissue matrix (Tissue-Tek® O.C.T.™, Sakura Finetek, #4583) and stored at − 80 °C for at least 24 h. Cryosectioning was performed using a CryoStar NX70 cryostat (Thermo Fisher Scientific). 6 μm-thick sections were mounted onto adhesive glass slides (Thermo Scientific™ SuperFrost Plus™, Thermo Fisher Scientific, #10,149,870), dried, and fixed in cold acetone (− 20 °C) for 10 min. Histological staining was conducted using Alcian Blue 8GX (Sigma-Aldrich, #A3157), prepared from 1 g Alcian Blue 8GX, 97 ml distilled water, and 3 ml 96% acetic acid (Sigma-Aldrich, #A9967). Sections were stained for 30 min, rinsed under running tap water for 1 min, and sequentially dehydrated in 95% ethanol and two changes of absolute ethanol (3 min each). Finally, slides were mounted with xylene, followed by cover glass placement. Sulfated glycosaminoglycans appeared blue and were visualized under a light microscope.

Differentiation of THP-1 into M1 macrophages

1 × 105 THP-1 monocytic cell line, (ATCC, #TIB-202 ™) were seeded into ThinCert cell culture inserts for 6-well plates (Greiner Bio-One, Kremsmunster, Austria) and cultured in growth medium (DMEM/F12 GlutaMAX I media supplemented with antibiotics (200 U/mL penicillin, 0.2 mg/mL streptomycin, and 2.5 µg/mL amphotericin B, (Sigma-Aldrich Chemie GmbH, Steinheim, Germany)) and 100 nM PMA (Sigma Aldrich, #P1585-1MG) for 3 days following 5 days in normal growth media to obtain resting M0 macrophages (rM0). Afterwards, rM0 were activated to the M1 phenotype by adding 20 ng/mL IFN-γ (Sigma-Aldrich Chemie GmbH) and 500 ng/mL LPS (Sigma-Aldrich Chemie GmbH) to fresh culture media for 2 days.

Co-cultivation of chondrocytes/synovial fibroblasts and M1 macrophages

To establish the co-culture, chondrocytes or synovial fibroblasts were seeded in 6-well plates (1 × 105 cells/well) and cultured in 2 mL growth medium for 48 h. ThinCerts bearing differentiated M1 macrophages were transfered to the chondrocyte/synovial fibroblast culture plates after washing macrophages and chondrocytes/synovial fibroblasts once with serum-free growth medium. Then 9 × 109 (90 000 EVs/cell) either FCS-, PRP- or HypACT-primed HFP-MSC-EVs (n = 3) were added to the co-culture in serum free media either supplemented with 2 µg/mL polybrene (plb+) (Sigma-Aldrich, #H9268) or left unsupplemented (plb−). After 48 h, supernatants were stored at − 80 °C and RNA was extracted immediately from cells. Chondrocytes (n = 3) and synovial fibroblasts (n = 3) used in co-cultures were isolated from different donors and treated with MSC-EVs from individual donors in subsequent experiments.

RT-qPCR analysis

Total RNA was extracted from chondrocytes, synovial fibroblasts and macrophages cultured in the inflammation model using a High Pure RNA Isolation Kit (Roche, #11,828,665,001). For cDNA synthesis a Transcriptor cDNA Synth Kit (Roche, #04897030001) was used according to the manufacturer’s protocol. For RT-qPCR, FastStart Essential DNA Probes Master (Roche, #06402682001) was mixed with 1 µL cDNA and 900 nM of primer (supplementary Table 1). CT values were obtained on a LightCycler 96 device (Roche, #05815916001). Data were normalized to GAPDH, and fold changes were calculated via the 2−ΔΔCt method.

Enzyme-linked immunosorbent assay (ELISA)

The concentrations of TNF-α, IL-4 and IL-10 in the co-culture supernatant were quantified using ABTS-based ELISA assays (PeproTech, London, UK; #900-K21K, #900-K14K, #900-K25K). Briefly, supernatants were analyzed in accordance with the manufacturer’s instructions, and absorbance was recorded at 405 nm with a wavelength correction at 650 nm using a BioTek Synergy 2 plate reader. Cytokine concentrations were calculated via BioTek Gen5 software (version 1.11.5).

Statistics

All statistical analyses were performed using GraphPad Prism 9.5, with a significance threshold of p < 0.05. Data distribution was assessed via the Shapiro-Wilk test. Depending on the experimental design, data were analyzed using either one-way ANOVA followed by Dunnett´s or Tukey’s multiple comparisons post-hoc test or two-way ANOVA followed by Tukey’s post-hoc test, as detailed in the figure legends. Data in graphs are represented as mean ± SD.

Results

Isolation and characterization of HFP-MSCs

HFP-MSCs were isolated from Hoffa’s fat pad tissues obtained from three donors (see Table 1 for donor details).

Table 1 Donor details. F-female, M-male, BMI- body mass index
Full size table

Once confluency was achieved, the HFP-MSCs were seeded onto microcarriers for 3D cultivation. Metabolic activity of HFP-MSCs was routinely assessed using the XTT assay to monitor the proliferation of HFP-MSCs cultured in the vertical wheel bioreactor and to determine the ideal time to start media harvest for EV isolation. As shown in Fig. 2A, the metabolic activity of HFP-MSCs in the bioreactor progressively increased, reaching a plateau after one week. Conversely, transitioning to serum-free media following blood priming led to a decline in metabolic activity (Fig. 2B). Our previous study provides a detailed description of the effect of 3D cultivation on the physiological state of cells, including senescence and apoptosis, as well as its impact on the yield and biological activity of EVs isolated from 3D culture [37].

Fig. 2
figure 2

Monitoring of HFP-MSCs metabolic activity over time in bioreactor culture, A from cell seeding to plateau phase, B from blood priming to supernatant harvest. Each graph reports data collected from 3 biological replicates and expressed as means ± SD. BP-blood products. SFM-serum free media

Full size image

To demonstrate the multilineage differentiation capacity of HFP-MSCs cultured in 3D conditions, cells were induced to differentiate into adipocytes, osteocytes, and chondrocytes (supplementary Fig. 1). Osteogenic differentiation was confirmed by Alizarin Red S staining, which revealed the deposition of calcified material (supplementary Fig. 1A). Significant differences in mineralization were observed depending on the type of media used (supplementary Fig. 1D), with PRP- and HypACT-supplemented media significantly enhancing osteogenic differentiation compared to undifferentiated cells or those in FCS-supplemented media. Adipogenic differentiation was evidenced by the presence of lipid droplets stained with Oil Red in all differentiated groups (supplementary Fig. 1B). Cells cultured in PRP-supplemented media showed significantly higher lipid dye retention, compared to those in FCS- or HypACT-supplemented media (supplementary Fig. 1E). Chondrogenic differentiation was evaluated through pellet formation and Alcian blue staining (supplementary Fig. 1C). Blue staining indicates deposition of glycosaminoglycans.

HFP-MSC-EVs enrichment and characterization

EVs were enriched from the culture supernatants of HFP-MSCs cultured under 3D conditions using ultrafiltration. The HFP-MSC-EVs samples were prepared from three pooled fractions; a primary retentate followed by two clean outs using PBS. HFP-MSC-EVs characterization was carried out according to the MISEV recommendations [39]. EVs were identified and characterized by assessing total protein content, particle number and size, cryo-electron microscopy and the presence of typical EV surface markers. The highest concentration (Mean ± SD) of particle/ml was measured in HypACT- EVs samples ( 5,12E + 10 ± 2,25E + 10), compared to PRP- ( 3,96E + 10 ± 1,40E + 10) or FCS-EVs ( 3,75E + 10 ± 1,99E + 10). The lowest protein concentration (Mean ± SD) was measured in FCS-EVs ( 453,6 ± 272,8 µg/ml) samples compared to PRP- ( 657,8 ± 79,83 µg/ml) or HypACT- ( 658,6 ± 69,73 µg/ml) EVs samples (Fig. 3A). To confirm the presence of EVs in our samples, surface markers were examined using Western Blot (Fig. 3B) and NTA in fluorescence mode (Fig. 3C). As shown in Fig. 3B, the typical EV marker proteins CD9 and CD63 and syntenin were detected in our samples. To confirm the absence of EV contaminants, calnexin (a marker of cell lysate contamination) and ApoB100 (a marker of blood plasma contamination) were analyzed. Furthermore, fluorescence NTA confirmed the presence of CD9 and CD73 on individual EVs. On average, 43.91 ± 23.49% of EVs expressed CD9, while 47.34 ± 24.99% of EVs displayed the CD73 marker on their surface. CD9⁺ EVs were smaller, with a mean mode size of 82.5 nm, compared to CD73⁺ EVs, which had a mean mode size of 161 nm (Fig. 3D). To analyze the morphology of isolated HFP-MSC-EVs, cryo-electron microscopy was performed (Fig. 3E). Particles of varying sizes and structures were observed, including single and double vesicles with lipid bilayers, as well as small electron-dense particles.

Fig. 3
figure 3

HFP-MSC-EVs characterization. A Analysis of particle concentration via NTA (left) and protein concentration via BCA assay (right) in FCS-, PRP- and HypACT primed HFP-MSC-EVs samples. Data are given as mean ± SD. *** p < 0.001. B Representative cropped blots of EV markers (CD63,CD9 and syntenin) and cell/plasma contaminant markers (Apo B100, calnexin), obtained from Western Blot. Full-length blots are presented in Supplementary Figs. 2, 3, 4 and 5 A and B. C Particle size distribution (black line) in scatter mode and fluorescent mode (pink and green line) obtained by NTA. D Comparison of mode size of CD9+ and CD73+ single positive EVs. Data are given as mean ± SD. *** p < 0.001. E Cryo-electron microscopy images showing structural differences of HFP-MSC-EVs samples. Single vesicles with lipid bilayer (white arrows), double vesicles with lipid bilayer (yellow arrows) and electron dense particle (blue arrow). Scale bar: 50 nm

Full size image

Internalization of HFP-MSC-EVs by chondrocytes and synovial fibroblasts increases with polybrene

To analyze whether polybrene enhances the internalization of EVs by synovial fibroblasts and chondrocytes, fluorescence imaging was performed. First, EVs were stained with CFSE dye, and NTA analysis revealed that more than 45% of EVs were CFSE-positive (Fig. 4E). CFSE-labeled EVs were then incubated with synovial fibroblasts (Fig. 4A, B) or chondrocytes (ratio of 2 × 104 EVs per cell ) (Fig. 4C, D) in the presence or absence of polybrene. To evaluate the efficacy of polybrene in promoting EV uptake, image processing was performed. As shown in Fig. 4F, the presence of polybrene significantly increased EV internalization by synovial fibroblasts, with an average of 49.67 EVs/cell, compared to 13.38 EVs/cell without polybrene. Similarly, polybrene enhanced EV uptake by chondrocytes, with 14.14 EVs/cell compared to 8.86 EVs/cell in the absence of polybrene.

Fig. 4
figure 4

HFP-MSC-EVs internalization by synovial fibroblasts (SF) and chondrocytes (CHRT). HFP-MSC-EVs were labeled with CFSE (green) and HFP-MSC-EVs uptake by chondrocytes and synovial fibroblasts was observed by confocal microscope; Scale bar: 20 μm. (A, B- synovial fibroblasts, C, D- chondrocytes). E) Particle size distribution in scatter mode (black line) and fluorescent mode (green line) obtained by NTA. F) Results from particle count analysis after image processing

Full size image

Induction of an inflammatory environment by M1 macrophages on synovial fibroblasts and chondrocytes in vitro

To establish an in vitro inflammatory osteoarthritis model, chondrocytes or synovial fibroblasts were co-cultured with differentiated M1 macrophages in the presence or absence of polybrene (plb). RT-qPCR analysis demonstrated that proinflammatory macrophages influenced cellular behavior significantly. In synovial fibroblasts, the presence of M1 macrophages led to heightened expression of inflammatory mediators, including cytokine IL-6, chemokines; IL8, CCL2/MCP-1, CCL5/RANTES, and the enzyme COX-2, relative to controls without macrophages (Fig. 5A). Chondrocytes co-cultured with M1 macrophages exhibited a downregulation of chondrogenic markers (COL1, COL2A) while catabolic enzyme genes (MMP3, MMP13) were upregulated compared to controls cultured without macrophages. Expression levels of ACAN and SOX-9 were slightly elevated (Fig. 5B).

Fig. 5
figure 5

The effect of M1 macrophages on gene expression in A synovial fibroblasts, B chondrocytes. Data were normalized to GAPDH. Normalized gene expression is shown as −ΔΔCt values (log₂ fold change is equivalent to -∆∆CT; positive values = upregulation, negative = downregulation), relative to the control (untreated cells). Each graph reports data collected from 3 biological replicates and expressed as means ± SD. Data were compared by 2way ANOVA, followed by Tukey´s post hoc test, with ****p < 0.0001, **p < 0.01, *p < 0.05

Full size image

Hoffa´s fat pad MSC-EVs treatment effects on chondrocytes/macrophages co-cultures

To evaluate the impact of HFP-MSC-EVs on chondrocyte gene expression, RT-qPCR analysis and measurement of pro-inflammatory and anti-inflammatory protein levels were performed after 48 h of treatment. Chondrocytes (n = 3) used in co-cultures were isolated from different donors and treated with HFP-MSC-EVs derived from individual donors in separate experiments. No donor matching or pooling was performed. HFP-MSC-EVs were found to alter chondrocyte gene expression. Specifically, we observed trends concerning upregulation of collagen genes (COL2A, COL1) and downregulation of catabolic genes (MMP3, MMP13), however, largely lacking significance due to substantial donor variability (Fig. 6). Absence or presence of polybrene (plb) did not affect gene expression changes.

Fig. 6
figure 6

HFP-MSC-EVs effects on expression of chondrogenic and matrix-degrading enzyme genes in chondrocytes within an inflammation model. Data were normalized to GAPDH. Each graph reports data collected from 3 biological replicates and expressed as means ± SD. Data were compared by 1way ANOVA, followed by Dunnett´s post hoc test, with *p < 0.05

Full size image

For example, the observed gene expression changes varied significantly with the age of the EV donors (Fig. 7), when comparing results from two separate experiments treated with EVs from different donors (Donor 1 and Donor 2) Notably, a statistically significant upregulation of the COL2A gene was observed in chondrocytes treated with EVs derived from the youngest donor compared to untreated cells, both in the presence (F = 133.5, p < 0.0001) and absence (F = 39.84, p < 0.0001) of polybrene. Conversely, EVs from the oldest donor resulted in a significant reduction in COL2A expression (F = 87.02, p < 0.0001) relative to untreated cells. Interestingly, the presence of polybrene significantly upregulated COL2A expression (F = 86.79, p < 0.0001) in chondrocytes treated with EVs from the oldest donor. Concerning blood product priming (FCS, PRP, HypACT), we did not observe significant differences, suggesting that priming cells for 24 h with different blood products prior to EV isolation did not affect their efficacy in vitro.

Fig. 7
figure 7

The effect of HFP-MSC-EVs from young/old donor on gene expression of COL2A. Data were normalized to GAPDH. Each graph reports data from 3 technical replicates of one chondrocyte donor and expressed as means ± SD. Data were compared by 1way ANOVA, followed by Dunnett´s post hoc test, with *** p < 0.001, **** p < 0.0001, ns - not significant

Full size image

To assess the influence of HFP-MSC-EVs on cytokine release, IL-4, IL-10, and TNF-α concentrations were measured in chondrocyte/macrophage co-culture supernatants using ELISA. After 48 h, EV treatment significantly modulated the release of anti-inflammatory cytokines IL-10 (F = 19.03, p < 0.0001) and IL-4 (F = 6.997, p < 0.0001), as well as the pro-inflammatory cytokine TNF-α (F = 3.705, p = 0.0013) (Fig. 8). The presence of polybrene affected the release of anti-inflammatory cytokines. Notably, IL-10 levels were consistently higher in co-culture supernatants without polybrene compared to their counterparts with polybrene across all EV groups.

Fig. 8
figure 8

HFP-MSC-EVs effects on pro/anti-inflammatory protein release. Each graph reports data collected from 3 biological replicates and expressed as means ± SD. Data were compared by ANOVA, followed by Tukey’s post hoc test, with *p < 0.05, **p < 0.01, ****p < 0.0001

Full size image

Hoffa´s fat pad MSC-EVs increased IL-10 gene expression in macrophages

To determine whether HFP-MSC-EVs influence macrophage polarization, we examined the gene expression of M1-associated (iNOS) and M2-associated (MRC1/CD206, IL-10) markers in macrophages from chondrocyte/macrophage co-cultures. Our findings revealed a significant increase in IL-10 expression in response to PRP-primed HFP-MSC-EVs (Fig. 9).

Fig. 9
figure 9

HFP-MSC-EVs effects on gene expression in macrophages from co-cultures. Data collected from 3 biological replicates were normalized to GAPDH and expressed as means ± SD. Data were compared by 2way ANOVA, followed by Tukey´s post hoc test, with *p < 0.05, **p < 0.01

Full size image

Hoffa´s fat pad MSC-EVs treatment on synovial fibroblasts/macrophages co-cultures

The impact of HFP-MSC-EVs on inflammation-related genes (IL-6, IL-8, CCL2/MCP-1, and CCL5/RANTES) and catabolic genes (ADAMTS4, ADAMTS5) was assessed in synovial fibroblasts using RT-qPCR. Levels of IL-4, IL-10, and TNF-α released into co-culture supernatants were also measured after 48 h of treatment. Synovial fibroblasts (n = 3) used in co-cultures were isolated from different donors and treated with HFP-MSC-EVs derived from individual donors in separate experiments. No donor matching or pooling was performed. HFP-MSC-EVs significantly downregulated the expression of CCL5/RANTES (F = 4.515, p = 0.0392) and COX2 (F = 7.203, p = 0.0116) compared to untreated controls. However, no significant downregulation of any inflammation-related gene was observed in groups treated in the presence of polybrene (Fig. 10).

Fig. 10
figure 10

HFP-MSC-EVs effects on expression of inflammatory genes in synovial fibroblasts. Data were normalized to GAPDH. Each graph reports data collected from 3 biological replicates and expressed as means ± SD. Data were compared by 1way ANOVA, followed by Dunnett´s post hoc test, with *p < 0.05, **p < 0.01

Full size image

The synovial fibroblasts/macrophage co-culture supernatants were analyzed to evaluate the influence of HFP-MSC-EVs on release IL-4, IL-10, and TNF-α cytokine levels. Following treatment with HFP-MSC-EVs, the levels of the anti-inflammatory cytokine IL-10 (F = 3.815, p = 0.0015) and the pro-inflammatory cytokine TNF-α (F = 2.676, p = 0.0130) were significantly or substantially altered, respectively (Fig. 11). IL-10 levels were particularly elevated in the co-culture supernatants, especially in synovial fibroblasts treated with HFP-MSC-EVs primed with FCS (with or without polybrene) and PRP compared to the control. Changes in TNF-α levels were not significant. No IL-4 were detected in any analyzed supernatants.

Fig. 11
figure 11

HFP-MSC-EVs effects on pro/anti-inflammatory protein release. Each graph reports data collected from 3 biological replicates and expressed as means ± SD. Data were compared by ANOVA, followed by Tukey’s post hoc test, with *p < 0.05, **p < 0.01

Full size image

Discussion

This study investigated the impact of HFP-MSC-EVs on the phenotype of cells critical to OA progression, namely chondrocytes, synovial fibroblasts, and macrophages. HFP-MSCs were cultured under 3D conditions and primed with blood products, including PRP and HypACT, which are routinely used in orthopedic practice, to assess their influence on EV efficacy. To enhance HFP-MSC-EV internalization by target cells, polybrene was investigated. Additionally, the effect of polybrene on the therapeutic potential of EV treatment was analyzed.

HFP-MSCs were cultured under 3D conditions in bioreactors on microcarriers following isolation to enhance their overall yield. A previous study reported an increased yield of extracellular vesicles (EVs) and higher EV secretion per cell in 3D cultures compared to 2D cultures [37]. Since achieving a sufficient quantity of EVs is essential for clinical application, maximizing their yield is a critical factor. Equally important is the selection of an appropriate MSC source for EV isolation and subsequent clinical use, and its relevance in the treatment of osteoarthritis (OA) [40]. Given that the Hoffa’s fat pad (HFP) is anatomically located within the knee joint, MSCs derived from this tissue retain a strong chondrogenic potential, even in elderly patients with OA [41]. Following 3D cultivation and the collection of culture media for EV isolation, the differentiation capacity of MSCs into adipocytes, chondrocytes, and osteoblasts was analyzed (Fig. 2).

The trilineage differentiation potential of the isolated cells was confirmed. However, the differentiation capacity of MSCs varied in response to different blood products, with PRP treatment yielding the most favorable results. PRP is often used in OA therapy and has been shown to provide significant improvements in pain and function scores [42]. Furthermore, human blood products, particularly those from plasma are under active investigation as cell culture supplements, offering a viable alternative to conventional animal-based options such as fetal bovine serum (FBS) [43].

HFP-MSC-EVs were isolated from conditioned media by ultrafiltration, as it is a gentler, simpler, faster, and more scalable method compared to ultracentrifugation [44, 45]. Using the NTA method, we measured higher particle concentrations per milliliter in EV samples isolated from MSCs primed with human blood products (PRP or HypACT) (Fig. 3A). Similarly, these samples exhibited higher protein levels compared to EVs derived from FCS-primed MSCs. Fluorescent NTA confirmed the presence of the EV-specific marker CD9 and the MSC marker CD73 in EV samples isolated from MSCs primed with FCS, PRP, or HypACT. The mean mode size of CD9⁺ particles was significantly smaller than that of CD73⁺ particles, which may be attributed to differences in their biogenesis pathways [37].

The presence of lipid bilayer vesicles in our samples was further confirmed by cryo-electron microscopy (Fig. 3E). The isolated vesicles varied in size and structure, with single lipid bilayer vesicles being predominantly observed. In the PRP-primed sample, small electron-dense particles were also detected. Western blot analysis further confirmed the presence of extracellular vesicles by the detection of EV markers CD9, CD63, and syntenin (Fig. 3B). The absence of cell contaminants in the EV samples was confirmed by the lack of the calnexin marker, in contrast to the cell lysate. The ApoB100 marker, was highly expressed in the plasma sample but significantly reduced in the PRP-primed samples. Based on the electron microscopy findings (presence of small dense particles) and Western blot analysis, the PRP-primed EV samples may contain small PRP residues.

To enhance the internalization of EVs by target cells and potentially improve their efficiency, we supplemented the culture media with polybrene. Surface charge plays a critical role in the cellular uptake of nanoparticles, as positively charged nanoparticles are more readily internalized by cells [46]. The surface charge of EVs is predominantly negative; however, it can be influenced by various factors, including the surrounding medium [47, 48]. Polybrene, a cationic polymer, can modify the surface charge of both cells and nanoparticles, thereby enhancing nanoparticle internalization. Cationic polymers adsorb onto negatively charged surfaces, leading to a reduction or reversal of surface charge due to electrostatic attraction between the positively charged polymer and the negatively charged particle surface [49, 50]. In our study, we observed a higher internalization of CFSE-positive particles by chondrocytes in the presence of polybrene (14.14 EVs/cell) compared to chondrocytes cultured without polybrene (8.86 EVs/cell). Similarly, synovial fibroblasts exhibited increased EV uptake in the presence of polybrene (49.67 EVs/cell) compared to those in polybrene-free media (13.38 EVs/cell) (Fig. 4). Although polybrene significantly enhanced EV internalization, we further investigated whether it also improves the therapeutic efficacy of EVs.

To determine the therapeutic efficacy of HFP-MSC-EVs, we analyzed their effect on the gene expression of inflammatory, catabolic, and chondrogenic markers, as well as inflammatory protein secretion, in cells critical for OA progression in vitro. To stimulate an inflammatory environment, chondrocytes or synovial fibroblasts were co-cultured with M1 macrophages. RT-qPCR analysis revealed that M1 macrophages induced the upregulation of inflammatory genes (IL6, IL8, CCL2, CCL5, COX2) in synovial fibroblasts and catabolic genes (MMP3, MMP13) in chondrocytes (Fig. 5). Macrophages play a vital role in OA progression by interacting with joint cells and modulating inflammation. The presence of activated macrophages in OA joints is associated with pain and disease severity [51]. Their depletion significantly reduces the production of pro-inflammatory factors, (e.g. IL1, TNF, IL6, IL8) in synovial tissue, reflecting their role in synovitis and inflammation. Macrophages removal is also associated with the downregulated of matrix degradation enzymes, such as metalloproteinases and aggrecanases, which are responsible for cartilage degradation [52,53,54].

Subsequently, HFP-MSC-EVs were added to the co-culture at a ratio of 9 × 10⁴ EVs per cell. As there is no universally accepted particle-to-cell ratio for treatment, we selected a relatively high EV dose to provide a robust stimulus under the experimental conditions. In chondrocytes, HFP-MSC-EVs treatment significantly downregulated the gene expression of catabolic enzymes MMP3 and MMP13 responsible for degradation of cartilage extracellular matrix, as well as the ACAN gene in the PRP-EV group. Furthermore, HFP-MSC-EVs upregulated COL1 and COL2A gene expression (Fig. 6). Interestingly, the upregulation of COL2A was dependent on the age of the EVs donor (Fig. 7). A statistically significant upregulation of the COL2A gene was observed in chondrocytes treated with EVs derived from a young donor compared to untreated cells, whereas EVs from an older donor resulted in a significant reduction in COL2A expression. To confirm these preliminary findings, it will be necessary to perform repeated analyses using samples from multiple donors. However, in a study by Cheng et al., it was shown that MSCs from older donors produced fewer small EVs and exhibited reduced immunopotency compared to those from younger donors [55]. This suggests that for potential future clinical applications, it is not only important to select a suitable tissue source for MSC-EVs, but also to consider the age and physiological condition of the donor. It should be noted that HFP samples used in this study were obtained from OA patients.

HFP-MSC-EVs also significantly increased the release of the anti-inflammatory cytokine IL-10 by chondrocytes compared to the control or chondrocytes co-cultured with M1 macrophages. Interestingly, the presence of polybrene significantly decreased IL-10 release across all EV groups compared to their polybrene-free counterparts. Additionally, IL10 gene expression was upregulated in M1 macrophages from chondrocyte/macrophage co-cultures (Fig. 9). Since IL-10 is secreted by M2 macrophages, our results suggest that M1 macrophages may have undergone polarization toward an anti-inflammatory M2 phenotype.

Similarly in synovial fibroblasts, HFP-MSC-EVs treatment increased the release of IL-10 compared to the control (Fig. 11), while it decreased CCL5 and COX2 gene expression (Fig. 10). CCL5 and COX2 are both associated with synovial inflammation, playing significant roles in the inflammatory processes within synovial tissues. CCL5, also known as RANTES, is a chemokine responsible for recruiting and activating immune cells at the sites of inflammation [56]. COX2 expression increases in response to proinflammatory cytokines (e.g., IL1β or TNF-α), and catalyzes the production of prostaglandins, contributing to disease progression. COX2 inhibitors are commonly used in OA therapy to reduce inflammation and pain [57,58,59]. Interestingly, we observed elevated levels of IL-10 in untreated synovial fibroblasts co-cultured with M1 macrophages, compared to the control. In inflammatory environments such as those present in osteoarthritis (OA), the synovium or cartilage may produce anti-inflammatory cytokines, including IL-10, as a counter-regulatory response to inflammation [60].

The results revealed that HFP-MSC-EVs could have a disease-modifying effect due to their anti-inflammatory, anti-catabolic, and pro-chondrogenic properties. A limitation of the study is the use of MSCs from OA patients with high donor variability, which may have influenced the significance of our results. As our findings suggest, it could be worthwhile to further investigate the effect of MSC-EVs from young versus older donors. Another limitation is the low number of donors, which limits the strength of our proof-of-concept that HFP-MSC-EVs from OA patients possess therapeutic effects independent of donor-specific characteristics. On the other hand, although we used MSCs from non-healthy sources, clinical studies have shown that MSCs isolated from the HFP of OA patients exhibit disease-modifying effects, such as reducing pain and improving joint function in individuals with knee OA [10, 11].

Furthermore, we did not observe significant differences in EV efficacy depending on MSC priming with blood products. It could also be worthwhile to culture MSCs in media supplemented with human blood products for longer period to determine whether PRP or HypACT could enhance their therapeutic efficacy. Priming MSCs with PRP appears promising due to its effects on multilineage MSC differentiation (Fig. 2) and the superior effect of EVs from PRP-primed MSCs on macrophage gene expression (Fig. 9). To support the hypothesis, that PRP-primed EVs have superior effect over the other priming conditions, further studies are required. Our results also revealed the possible presence of small residuals in PRP- primed samples, observed as small electron-dense particles in cryo-electron microscopy images.

Additionally, we observed increased EV internalization by target cells in the presence of polybrene (Fig. 4), which was not linked to improved treatment outcomes. To optimize EV internalization for potential clinical applications, further research is needed to explore its impact on EV treatment efficacy and translation to clinical practice.

Conclusion

In summary, HFP-MSC-EV treatment triggered cellular responses in chondrocytes and synovial fibroblasts. HFP-MSC-EVs significantly downregulated the expression of catabolic enzymes MMP3 and MMP13, responsible for cartilage matrix degradation, as well as CCL5 and COX2, which are associated with synovial inflammation. Although polybrene increased the uptake of HFP-MSC-EVs by target cells, it did not lead to improved therapeutic outcomes. Furthermore, we observed elevated anti-inflammatory IL-10 levels after treatment and its upregulation in M1 macrophages, suggesting a possible polarization toward an anti-inflammatory M2 phenotype. Interestingly, we also observed a significant upregulation of COL2A in chondrocytes in response to HFP-MSC-EVs from a young compared to an old HFP donor.