Human DPSCs were provided by S-Evans Biosciences. The Ethics Committee of S-Evans Biosciences approved the collection of the human samples (no. 2020-01). Briefly, regular human third molars were extracted from adult donors (age, 18-28 years) during orthodontic treatment or due to impaction after obtaining written informed consent. Subsequently, the pulp tissue was lightly harvested and then digested in a collagenase type I (2 mg/ml) and dispase (4 mg/ml) solution for 1 h at 37°C. After passing through a 70-µm strainer (Falcon, BD Biosciences), the cell suspension was seeded into tissue culture plates at a density of 1×10⁵/well with the α modification of Eagle's medium (Gibco; Thermo Fisher Scientific, Inc.), supplemented with 15% fetal bovine serum (FBS, Gibco; Thermo Fisher Scientific, Inc.), 2 mM l-glutamine, 100 units/ml penicillin and 100 µg/ml streptomycin. The plates were then incubated at 37°C in 5% CO₂ using a humidified incubator and passaged by 0.25% trypsin-ethylenediaminetetraacetic acid (EDTA, Gibco; Thermo Fisher Scientific, Inc.) digestion at a 1:5 ratio.

Osteogenic, adipogenic and chondrogenic inductions were conducted in vitro, as previously described, to evaluate the multilineage differentiation capacity of the DPSCs (40). To evaluate mineralization, the cultures were stained with Alizarin Red S according to the manufacturer' instructions (Cyagen Biosciences, Inc.). Briefly, cells were fixed with 4% paraformaldehyde (Beyotime Institute of Biotechnology) for 30 min and then stained with Alizarin Red S for 5 min at room temperature. Lipid droplets formed following adipogenesis induction were visualized using Oil Red O staining according to the manufacturer' instructions (Cyagen Biosciences, Inc.). Briefly, cells were fixed with 4% paraformaldehyde for 30 min followed by Oil Red O staining for 30 min at room temperature. Successful chondrogenic differentiation was tested with the use of Alcian blue staining according to the manufacturer' instructions (Cyagen Biosciences, Inc.). Briefly, cells were fixed with 4% paraformaldehyde for 30 min followed by Alcian blue staining for 30 min at room temperature.

The phenotype of the DPSCs was examined at the third passage by flow cytometry (BD FACSVerse) using antibodies, including anti-CD14 (BD Biosciences, cat. no. 555397, at a dilution of 1:5 and incubation at room temperature for 15 min), anti-CD34 (BD Biosciences, cat. no. 555822, at a dilution of 1:5 and incubation at room temperature for 15 min), anti-CD73 (BioLegend, cat. no. 344004, at a dilution of 1:20 and incubation at room temperature for 15 min), anti-CD105 (BioGems, cat. no. 17111-60, at a dilution of 1:20 and incubation at room temperature for 15 min), anti-HLA-DR (BD Biosciences, Cat. no. 555811, at a dilution of 1:5 and incubation at room temperature for 15 min) and anti-CD90 (BioLegend, cat. no. 328110, at a dilution of 1:5 and incubation at room temperature for 15 min) according to the manufacturer' instructions. The suitable isotype-matched antibody (PE; BD Biosciences, cat. no. 555749; FITC, BD Biosciences, cat. no. 555573, at dilution of 1:5 and incubation at room temperature for 15 min) was utilized as a negative control. The data were analyzed using BD FACSuite software (1.0.5.3841).

Exosome isolation was performed using ultracentrifugation as previously described (41). In brief, culture supernatants of DPSCs were centrifuged at 3,000 × g for 30 min and then at 10,000 × g for a further 30 min at room temperature. Finally, exosomes were pelleted following centrifugation at 64,000 × g for 110 min at 4°C using an SW28 rotor (Beckman Coulter, Inc.) and washed once with 0.32 M sucrose. Nanosight 2000 analysis and transmission electron microscopy (FEI) were utilized for the identification of exosomes that were resuspended in phosphate-buffered saline.

DPSCs at the third passage were transfected with miR-140-5p mimic (20 µg/ml) or negative control (20 µg/ml) (Guangzhou RiboBio Co., Ltd.). The Amaxa Nucleofection system was adopted according to the instructions of the manufacturer of the Human Mesenchymal Stem Cell Nucleofector TM kit (Lonza Group, Ltd.). Briefly, DNA and cells were mixed in the Amaxa cuvette (Lonza Group, Ltd.) and placed in the nucleofector device. Nucleofection was conducted using the U-23 program (Lonza Group, Ltd.), which was optimized and performed as recommended by the manufacturer. Subsequently, the cuvette was rinsed with culture medium and the cells were transfer into the dish for further culture.

Total RNA was extracted using the Total Exosome RNA and Protein Isolation kit (Invitrogen; Thermo Fisher Scientific, Inc., cat. no. 4478545) according to the manufacturer's protocol, and the amount of RNA isolated was quantified using a NanoDrop spectrophotometer (NANO-100; Hangzhou Allsheng Instruments Co.,Ltd.). RT reactions were performed using 100 ng total RNA with the TaqMan miRNA RT kit (Applied Biosystems; Thermo Fisher Scientific, Inc., cat. no. 4366596) according to the manufacturer's instructions. qPCR reactions were conducted on a 7900 System (Applied Biosystems; Thermo Fisher Scientific, Inc.) using the Taq-Man miR-140 probe and gene-specific primers (Thermo Fisher Scientific, Inc., Assay ID 001187). Homo sapiens snRNA U6 qPCR Primer (Thermo Fisher Scientific, Inc., Assay ID 001973) were used as an internal control. qPCR was performed under the following conditions: 10 min at 95°C, followed by 40 cycles of 95°C for 10 sec, 20 sec at 60°C and 10 sec at 72°C.

Cartilage cells subjected to IL-1β (10 ng/ml; R&D Systems, Inc.) and DPSC-derived exosome (5×10⁸ particles/ml) treatment for 48 h were collected and total RNA was isolated using TRIzol reagent (Invitrogen, cat. no. 15596026) and reversed transcribed into first-strand cDNA using reverse transcriptase (Takara Bio, Inc.; cat. no. H2640A). RT-qPCR was performed using a SYBR-Green Master Mix (Takara Bio, Inc.; cat. no. RR820A) under the following conditions: 95°C for 4 min, 40 cycles of 95°C for 30 sec, 58°C for 30 sec, 72°C for 30 sec, followed by melting curve analysis of 95°C for 30 sec, 72°C for 30 sec, and 95°C for 30 sec. Relative gene expression was calculated using the 2^−ΔΔCq (42) method and the data were normalized against GAPDH. The primers used are listed in Table I. Each PCR trial was performed in triplicate and repeated at least 3 times.

Specific pathogen-free (SPF) male Sprague-Dawley (SD) rats (approximately 8 weeks old, weighing 200±20 g) were procured from the Shanghai SLAC Laboratory Animal Co., Ltd. [Certificate no. SCXK (Shanghai) 2017-0005] and maintained in the SPF laboratory of a local animal facility. The experimental protocol was approved by the local Medical Animal Experiment Ethics Committee of Zhejiang Provincial People's Hospital (no. 2019-034).

A total of 24 SD rats were randomly divided into the blank, OA model, exosome treatment and improved-exosome treatment groups, with 6 animals per group. Subsequently, 50 µl of 25 mg/ml iodoacetic acid (Sigma-Aldrich; Merck KGaA) was administered to the double knee joints to induce OA. The same amount of saline was administered into the knee joint cavities of the rats in the blank group. After 1 week, the exosome treatment group and improved-exosome treatment group were injected with 50 µl of exosomes (5×10¹⁰ particles/ml, isolated from the culture supernatants of DPSCs as described above) and miR-140-enriched exosomes (5×10¹⁰ particles/ml, isolated from the culture supernatants of miR140-transfacted DPSCs as described above), respectively, in the knee joint cavity of both hind limbs once a week; a total of 4 such injections were administered. The rats in the blank group and OA model group were administered the same amount of saline. The rats were sacrificed by an overdose of pentobarbital (150 mg/kg) intraperitoneally at 1 week after the end of treatment in accordance with a previous study (43).

At 1 week after the end of treatment, 3 rats were randomly selected from each group and anesthetized with 3% pentobarbital sodium (60 mg/kg) by intraperitoneal injection. An X-ray examination of the knee joint structure was then performed, and the grade of knee joint degradation of digital radiography films (PerkinElmer, Inc., CLS136341/F) was scored using a modified Kellgren-Lawrence scoring system for rats as previously described (44,45).

The rats were sacrificed, and their knee joints were isolated and fixed in 10% neutral-buffered formalin (Sangon Biotech) for 3 days and then decalcified for 2 weeks in buffered 12.5% EDTA (as previously described with some modifications) (46,47) and formalin solution. The tissues were then embedded in paraffin and were sliced along the longitudinal axis of the lower extremity at a thickness of 5 µm. They were subsequently stained with Safranin O-Fast Green (Sangon Biotech) and hematoxylin and eosin (H&E; Sangon Biotech). For Safranin O-Fast Green staining, sections were deparaffinized and stained with Safranin O for 5 min followed by Fast Green staining for 5 min at room temperature. For H&E staining, sections were deparaffinized and stained with hematoxylin staining solution for 10 min and then counterstained in eosin staining solution for 1 min at room temperature. The OARSI score was used to evaluate histological changes as previously described (48).

Normal human cartilage tissues were obtained from donation with written informed consent, and the protocol was approved by the Ethics Committee of Zhejiang Provincial People's Hospital in Hangzhou, China (2018 KY 012). Samples were obtained aseptically and digested using a standard protocol. Briefly, the tissues were cut into <1 mm³ sections; they were then transferred to a culture flask, mixed with 5-fold volumes of 0.25% trypsin, and placed in a 37°C, 5% CO₂ incubator for 30 min. The trypsin digestion solution was then discarded and 5-fold volumes of 0.05% type II collagenase (Sigma-Aldrich; Merck KGaA) were added to the tissue fragments. Subsequently, cells were collected every 4 h until the tissue block was digested. These chondrocytes were then cultured in DMEM supplemented with 20% FBS in a 37°C, 5% CO₂ incubator.

Normal human chondrocytes were treated with IL-1β to induce the inflammatory, in vitro model of OA as previously described with some modifications (49). The cells were divided into 4 different groups as follows: i) Normal culture; ii) IL-1β (10 ng/ml, R&D Systems); iii) IL-1β (10 ng/ml) + DPSC-derived exosomes (5×10⁸ particles/ml); and iv) IL-1β (10 ng/ml) + miR-140 enriched exosomes (5×10⁸ particles/ml). Cells were collected for use in western blot analysis and apoptosis analysis at different time points.

Western blot analysis was performed as previously described (50). Briefly, exosomes or chondrocytes were added to cold lysis buffer (Beyotime Institute of Biotechnology, cat. no. P0013B) and incubated for 30 min on ice. The lysates were then centrifuged at 13,000 × g for 10 min at 4°C. Subsequently, the supernatants were collected and quantified using the BCA protein assay kit (Beyotime Institute of Biotechnology, cat. no. P0011), and samples (20 µg) were separated on 10% polyacrylamide gels and transferred to Hybond-P membranes. 5% BSA (Beyotime Institute of Biotechnology, cat. no. ST025) was used for blocking at room temperature for 45 min prior to incubation with antibodies. The following antibodies were used: Rabbit anti-human CD63 (ProteinTech Group, Inc., cat. no. 25682-1-AP, at a dilution of 1:500), CD9 (ProteinTech Group, Inc., cat. no. 20597-1-AP at a dilution of 1:500) and Bcl-2 (ProteinTech Group, Inc., cat. no. 12789-1-AP at a dilution of 1:1,000), as well as rabbit anti-human Bax (Beyotime Institute of Biotechnology, cat. no. AF0057, at a dilution of 1:500) and Bad (Beyotime Institute of Biotechnology, cat. no. AF1009, at a dilution of 1:500). The primary antibodies were incubated at room temperature for 45 min. GAPDH (Beyotime Institute of Biotechnology, cat. no. AF1186, at a dilution of 1:2,000) served as the loading control. The blots were then incubated with HRP-conjugated goat anti-rabbit secondary antibody (Beyotime Institute of Biotechnology, cat. no. A0208, at a dilution of 1:1,000) for 45 min at room temperature and then processed for analysis using an ECL chemiluminescence kit (Beyotime Institute of Biotechnology). ImageJ (National Institutes of Health, v1.8.0) was further employed to analyze the grayscale values obtained by western blotting.

Apoptosis was determined using the Annexin V-propidium iodide (PI) binding assay as previously reported (51). The cells were harvested and labeled with Annexin V-FITC and PI for 15 min using a FITC Annexin V apoptosis detection kit I (BD Pharmingen). Apoptosis was evaluated using flow cytometry (Navios, Beckman Coulter, Inc.) within 30 min of staining, and the data were analyzed using FlowJo software v11.0. The extent of apoptosis was quantified as the percentage of Annexin V⁺ cells. All experiments were performed in triplicate.

Data are presented as the means ± standard deviations of the results of at least 3 independent studies. Student's t-tests were utilized to identify differences among groups, as suitable. One-way ANOVA with the Bonferroni test was performed for multiple group comparisons. A P-value <0.05 was considered to indicate a statistical significant difference.

DPSCs were isolated from the pulp tissue of extracted human third molar teeth without caries, periodontal diseases, or infections. These cells are closely associated with MSCs found in the stromal compartment of various tissues, including the bone marrow. As shown in Fig. 1A, the cells exhibited a typical spindle-shaped morphology. DPSCs followed the criteria that defined multipotent MSCs, as recommended by the International Society for Cellular Therapy guidelines in 2006 (52). Following culture in a specific differentiation medium for 2-3 weeks, the isolated DPSCs successfully differentiated into mesenchymal derivatives, including osteoblasts (identified by Alizarin Red labeling), adipocytes (identified by Oil Red O labeling) and chondrocytes (identified by Alcian blue) (Fig. 1A). Furthermore, as revealed by flow cytometric assay, these cells were positively stained for the common MSC-associated markers, CD73, CD90 and CD105, and did not exhibit an expression of the hematopoietic markers, CD14, CD34 and HLA-DR (Fig. 1B).

To examine the effects of miR-140-5p-transfected DPSCs on articular cartilage injuries, miR-140-5p mimic or negative control oligonucleotide were transfected into third-passage DPSCs via electroporation. The cells were collected 48 h later for total RNA extraction. RT-qPCR revealed that the cells transfected with the mimic exhibited a 10-fold increase in the miR-140-5p level compared with the negative control cells (Fig. 2A). Exosomes were collected from both types of DPSCs by centrifugation. Nanosight analysis indicated that the diameter of the majority of the DPSC-derived exosomes was approximately 134±29 nm (Fig. 2B). As shown in Fig. 2C, western blot analysis revealed that the DPSC-derived exosomes expressed exosomal markers, including CD9 and CD63. No significant differences in the expression levels of CD9 and CD63 between exosomes from the DPSCs with and without miR-140-5p transfection were observed (P>0.05). Finally, the results of RT-qPCR revealed a 4-fold increase in the miR-140-5p levels in exosomes derived from the miR-140-5p-overexpressing DPSCs than those from the control exosomes.

Subsequently, normal human chondrocytes were treated with IL-1β (10 ng/ml) with or without DPSC-derived exosomes. As shown in Fig. 3, IL-1β stimulation substantially decreased the chondrocyte ACAN, Col2α1 and Sox9 mRNA levels. However, the use of DPSC-derived exosomes elevated the corresponding mRNA levels, particularly the miR-140-5p-enriched exosomes.

Chondrocytes from normal humans were treated with IL-1β (10 ng/ml) with or without DPSC-derived exosomes. The cells were trypsinized for apoptosis analysis at various time points. As shown in Fig. 4A and B, when exposed to IL-1β for 48 h, the DPSC-derived exosome groups (DPSC-derived exosomes and miR-140-5p-enriched exosomes) exhibited a decreased apoptosis compared with the control group (16.29±0.1% and 12.32±0.2% vs. 21.91±0.3% Annexin V⁺ cells at 24 h; 24.91±0.1% and 18.14± 0.1% vs. 34.58±0.5% Annexin V⁺ cells at 48 h; all P<0.01). The miR-140-5p-enriched exosome group exhibited a more significant reduction in cell apoptosis than the DPSC-derived exosome group (P<0.01) (Fig. 4A and B). Likewise, western blot analysis indicated a significant inhibition of the expression of apoptosis-related proteins in chondrocytes treated with miR-140-5p-enriched exosomes (Fig. 4C and D).

Subsequently, the rat model of OA was treated with the 2 above-mentioned types of DPSC-derived exosomes and then assessed using X-rays to determine the effects of the treatments (Fig. 5). In comparison with the sham-operated group, the OA model group exhibited evident stenosis, with visible osteophytes and osteosclerosis. This was in line with previous findings (53,54), indicating the generation of the rat model of OA was successfully established. Both DPSC-derived exosomes reduced the grade and severity of knee joint damage in rats with OA, as determined by X-ray score, improved the joint cavity structure, and reduced the formation of osteophytes, articular surface irregularity and osteosclerosis. Moreover, these effects were substantially enhanced following miR-140-5p-enriched exosome treatment.

To further assess the effects of exosome treatment on the knee condition in rats with OA, the histological sections of rat cartilage in all 4 groups were examined. As observed following Safranin O-Fast Green and H&E staining (Figs. 6A and S2A), the knee joint cartilage was uneven in the rat model OA group, and the thickness of the cartilage was reduced with visible apoptotic bodies. Furthermore, the lower edge of the cartilage was uneven, and pathological changes, including inflammatory cell infiltration, could be observed. The multiple administration of DPSC-derived exosomes improved the pathology of the rat knee joints to varying degrees, and the miR-140-5p-enriched exosomes substantially reduced the total pathological OARSI grading score (Figs. 6B and S2B).