PDLSCs were cultured using methods described in a previous publication (Lan et al., 2022). In brief, the cryopreserved PDLSCs were resuscitated and subsequently resuspended in a mesenchymal stem cell complete medium (Cyagen, China). After inoculation, the cells were cultured in an incubator at 37°C. When the cultured cells reached a confluency of 80%, they were passaged at a ratio of 1:2.

For the preparation of Cur-PDLSC-EV, the PDLSCs were cultured in an exosome-free medium containing 10 μg/mL curcumin for 24 h. The supernatants were collected and centrifuged at 300 g for 5 min at 4°C to remove the dead cells. The collected supernatants were then centrifuged at 10,000 g for 30 min at 4°C to remove cellular debris. The supernatants were transferred to a 38.5 mL ultracentrifugation tube (Beckman Coulter, United States), and centrifuged at 100,000 g for 70 min. Subsequently, the pellets were resuspended and added to a 5.5 mL ultracentrifugation tube (Beckman Coulter, United States). At the bottom of the ultracentrifugation tube, 500 µL Exojuice (WeinaBio, China) was added, followed by adding PBS to fill up. After centrifuging at 100,000 g for 70 min, 200 µL liquid was removed at the bottom, and then 200 µL liquid was pipetted, which contained extracellular vesicles of high purity, and stored at −80°C before use.

PDLSC-EV were isolated using methods previously described in the literature (Lan et al., 2022). Briefly, after culturing the PDLSCs in the exosome-free medium for 24 h, the cell supernatants were collected. Extracellular vesicles were isolated using the ultracentrifugation method, and purified using the Exojuice–Exosome Purification and Concentration Kit (WeinaBio, China).

The particle size and electric potential of extracellular vesicles were analyzed using a laser particle size analyzer. The morphology of extracellular vesicles was observed using a transmission electron microscope. Western blotting was used to detect the exosome-related specific proteins (CD9, CD63, CD81, and TSG101).

To confirm the absorption of Cur-PDLSC-EV and PDLSC-EV by cells, the extracellular vesicles were labeled with a red-fluorescent dye (DiI, Solarbio, China) as described previously in the literature (Lan et al., 2022). Briefly, EV were treated with 5 µl of 10 µM DiI-labeling dye in 1 ml serum-free medium and incubated for 30 min at 37°C. EV suspension was ultracentrifuged at 100,000 g for 70 min at 4°C. They were then incubated with the PDLSCs at 37°C for 3 h, and fixed in 4% paraformaldehyde for 15 min. FITC- phalloidin (Yeasen, China) was used to stain the cytoskeleton for 45 min, and DAPI (Solarbio, China) staining was used for the nuclei. The red signals in the cells were detected using a fluorescence microscope (Mshot, China).

PDLSCs were inoculated in a 96-well plate at a density of 5,000 cells per well, with five parallel sets designed for each group. After growing for 24 h, the original culture medium was removed, and 50 μg/mL Cur-PDLSC-EV or PDLSC-EV were added. After an additional 48 h of culture, the culture medium was removed, and 100 µL fresh culture medium containing 10% CCK-8 reagent (APExBIO, United States) was added. After culturing in a 37°C incubator for 1.5 h, the absorbance at 450 nm was measured using a microplate reader (Thermo, United States), and the data were recorded.

The cells were resuspended after digestion, and the suspension was diluted to 5 × 10⁵ cells/mL. Approximately 100 µL of the suspension was added to the upper chambers of the Transwell (Corning, United States), and 50 μg/mL PDLSC-EV or Cur-PDLSC-EV was added to the lower chambers. After 24 h of culture, the culture media in the upper and lower compartments were removed. After washing with PBS three times, 1 mL of 4% paraformaldehyde was added to the upper and lower compartments to fix for 30 min, washed again with PBS three times, and then 1 mL crystal violet was added for 30 min to stain. The excess dyes were subsequently removed, followed by washing with PBS three times. Subsequently, the cells in the upper chambers were gently wiped with a cotton swab, and the inserts were then properly air-dried. The migration results were observed and recorded under a microscope.

After digestion, the PDLSCs were seeded at a density of 2 × 10⁴ cells/well into a 24-well plate, and osteogenic differentiation medium (ODM, Cyagen, China) was added to induce osteogenesis. After 24 h, 50 μg/mL PDLSC-EV, Cur-PDLSC-EV or Cur-PDLSC-EV+(ICG-001 (10 μM) was added. ARS and AKP staining were performed after 7–10 days of culture-induced differentiation. Before staining, 4% paraformaldehyde (Beyotime, China) was added for fixing at room temperature for 30 min, followed by washing with PBS three times at 5 min each time. Subsequently, the ARS staining solution (Beyotime, China) and AKP staining solution (Beyotime, China) were added and stained for 5 and 30 min, respectively. The results were observed, pictures were taken, and recorded under a microscope.

RNA was extracted by using Trizol reagent and chloroform, precipitated by isopropanol, rinsed in 75% ethanol, and then resuspended in nuclease-free water. After measuring the concentration of the derived RNA, cDNA was synthesized by HiScript^® III RT SuperMix for qPCR (+gDNA wiper) (Vazyme, China). Quantitative PCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, China) to detect the mRNA levels of osteogenesis-related genes COL1/ALP/RUNX2. β-Actin was used for the normalization of mRNA expression. All gene sequences were obtained from GenBank, and the primers were designed (Primer Premier 5) and synthesized (Sangon Biotech, China). Primer sequences are shown in Table 1.

BCA working solution and serially diluted BSA standards were prepared according to the reagent kit operation manual. A volume of 10 µL of the BSA standards and samples to be tested were added to 100 µL of the BCA working solution, mixed well, and then reacted at 37°C for 30 min. The sample concentration was measured using an ultra-micro spectrophotometer. Subsequently, 10 µL protein loading buffer was added to an aliquot of 50 µL exosome extract for protein denaturing at 95°C for 10 min. A quantity of 50 μg of total protein was separated through 10% SDS-PAGE, whose electrophoresis conditions were 90 V for 30 min and 120 V for 60 min. The proteins were then transferred to a PVDF membrane, with a regulated voltage of 60 V and a transfer time of 2 h. The membrane was blocked using 5% nonfat milk for 1 h. GAPDH/GSK3β/β-catenin antibodies (1:1000 dilution) were added and incubated overnight at 4°C. The membrane was then washed with 1 × TBST four times at 5 min each time. The corresponding horseradish peroxidase (HRP)-labeled secondary antibodies (a dilution ratio of 1:5000) were added and incubated for 1 h. The membrane was washed with TBST four times at 5 min each time. The protein expression was detected using the ECL chemiluminescent reagents, and the signals were developed in a chemiluminescence imaging system.

According to the laser particle size analyzer, PDLSC-EV had an average diameter of 117 nm and an average electric potential of −26.8 mV; Cur-PDLSC-EV exhibited an average diameter of 126 nm and an average electric potential of −22.7 mV (Figure 1A). The transmission electron microscopy revealed that both PDLSC-EV and Cur-PDLSC-EV displayed cup-shaped vesicles (Figure 1B). The Western blotting data showed that both PDLSC-EV and Cur-PDLSC-EV expressed CD81, CD63, CD9, and TSG101 proteins, and the CD9 and CD81 expression of Cur-PDLSC-EV were higher than those of PDLSC-EV (Figure 1C; Supplementary Material). After co-culture of DiI-labeled PDLSC-EV or Cur-PDLSC-EV with PDLSCs for 3 h, it could be observed under a fluorescence microscope that the extracellular vesicles had already been successfully internalized by PDLSCs (Figure 1D).

The PDLSCs were treated with 50 μg/mL Cur-PDLSC-EV or PDLSC-EV, and after 48 h CCK-8 was added to measure the absorbance value at 450 nm. The OD₄₅₀ and cellular viability results (Figures 2A, B) showed that PDLSC-EV and Cur-PDLSC-EV had promoting effects on the proliferation of PDLSCs, and they did not exhibit a difference in the enhanced activities between each other.

The transwell assay (Figure 2C) data showed that compared with the PBS control group, PDLSC-EV and Cur-PDLSC-EV possessed facilitative effects on the migration of PDLSCs, and there was no difference in the enhancement between the two.

After osteogenic induction, the PDLSCs began to change from fibrous shape to polygonal, scale-like morphology. As the induction time increased, calcium nodules gradually formed, which after ARS staining could be visible as stained red mineral deposits. Based on the ARS and AKP staining results (Figures 3A–C), it was shown that either PDLSC-EV or Cur-PDLSC-EV promoted the osteogenic differentiation of PDLSCs, and Cur-PDLSC-EV had a better enhancement effect. Additionally, after adding ICG001, the pro-osteogenic ability of Cur-PDLSC-EV was reversed in PDLSCs (Figures 4A, B).

The RT-qPCR data revealed that PDLSC-EV had no effect on the expressions of osteogenesis-related genes COL1 and Runx2, while upregulating ALP gene expression (p < 0.05). Cur-PDLSC-EV had stimulating effects on the gene expression of all the osteogenesis-related genes COL1, Runx2, and ALP (p < 0.01). Comparing the PDLSC-EV and Cur-PDLSC-EV groups, there were extremely significant differences in COL1 and ALP gene expression (p < 0.01) (Figure 3D).

According to the Western blotting data, it was shown that Cur-PDLSC-EV upregulated the protein expression of p-GSK3β and β-catenin in PDLSCs. To further confirm that the activation of Wnt/β-catenin pathway is responsible for the progression of PDLSCs by Cur-PDLSC-EV treatment, PDLSCs were treated with Cur-PDLSC-EV in the presence of the inhibitor (ICG001) of Wnt/β-catenin pathway. Our results suggested that the Cur-PDLSC-EV-mediated activation of Wnt/β-catenin signaling could be blocked by ICG001, and the osteogenic differentiation was attenuated (Figure 4C; Supplementary Material). In addition, a luciferase assay based on a TOPflash/FOPflash reporter plasmid system was used to evaluate β-catenin-dependent transcriptional activity. The results demonstrated that Wnt/β-catenin signaling was activated following Cur-PDLSC-EV treatment (Figure 4D), accompanied by a significant increase in the mRNA levels of the downstream target gene CCND1 (Figure 4E). These results indicate that Cur-PDLSC-EV promoted osteogenesis by activating Wnt/β-catenin signaling pathway.