Human periodontal ligament stem cells (PDLSCs) were isolated from freshly extracted orthodontic premolars obtained from three healthy donors (two females and one male, aged 14–18 years) with written informed consent. The experimental protocol was approved by the Ethics Committee of the Ninth People’s Hospital (SH9H-2020-T36-2). PDLSCs were cultured and identified following previously reported protocols (Moshaverinia et al., 2014). Cells from different donors were not pooled; experiments were conducted independently for each donor to account for biological variability. Passages three to five were used for all experiments.

Zoledronate (ZOL) was obtained from Novartis Pharmaceuticals. PDLSCs were treated with ZOL at 0.01, 0.1, 0.5, 1, and 10 µM. The ZA concentrations used in this study were chosen based on human pharmacokinetic studies, ensuring that low-dose conditions (0.01–0.5 µM) approximate systemic and early bone surface exposures in osteoporosis, while higher-dose conditions (1–10 µM) model local skeletal accumulation observed in oncology patients receiving repeated high-dose therapy. In humans, a standard 4 mg IV infusion for osteoporosis yields peak plasma concentrations of 0.11–0.26 µM. Rapid skeletal uptake reduces plasma levels within 24 h, while local bone surface concentrations are modestly higher, supporting the use of 0.01–0.5 µM to mimic systemic and early bone exposures in osteoporotic patients (Skerjanec et al., 2003; Raccor et al., 2013; Cai et al., 2018). Oncology patients with bone metastases receive more frequent, higher-dose regimens (4 mg every 4 weeks), resulting in repeated high local skeletal concentrations (0.4–4.6 µM) due to both increased dosing and higher bone turnover (Ruggiero et al., 2014). Accordingly, 1–10 µM ZOL was employed to model high local exposure conditions associated with clinical MRONJ risk. Thus, the 0.01–10 µM range covers physiologically relevant plasma levels and potential local enrichment on bone surfaces.

PDLSCs (5 × 10³ cells/mL) were seeded into 96-well plate and treated with ZOL (0.01, 0.1, 0.5, 1, 10 μM) or vehicle (0.9% NaCl). Cell proliferation was assessed at 1, 3, 5 and 7 days using CCK-8 assay.

PDLSCs were seeded into 6-well plate and treated with ZOL for 72 h. Cell apoptosis was analyzed using FITC Annexin V Apoptosis Detection Kit and Cell Death Fluorescein Detection Kit.

PDLSCs were seeded in 12-well plate and treated with ZOL for 72 h. Morphology was observed by phase-contrast microscopy, and cytoskeleton stained with TRITC-phalloidin (YEASEN) was imaged under fluorescence microscopy.

PDLSCs (5 × 10³ cells/well) were seeded in 24-well plates and allowed to attach overnight. Cells were then continuously treated with ZOL at different concentrations for 72 h. After osteogenic induction for 7 days, ALP staining was performed with BICP/NBT Kit. Then cells were lysed with 1% Triton X-100 and subjected to three cycles of freezing and thawing. ALP activity was measured at 405 nm and normalized to protein content. ZOL was continuously present in the culture medium (total 10 days).

Cell treatment and processing steps were carried out following the same protocol as the ALP staining. PDLSCs were induced in the osteogenic media for 14 days, then cells were fixed in 4% paraformaldehyde and stained with 1% Alizarin Red. Alizarin Red dye was extracted with 400 µL of 10% cetylpyridinium chloride for 15 min and quantified at 562 nm.

PDLSCs were seed in 6-well plate and treated with ZOL for 72 h. Then total RNA was isolated using with Trizol, reverse transcribed by Superscript II. Real-time PCR was performed with Light Cycler^® 480 II. β-Actin was used to normalize gene expression. Primers used in this study are shown in Supplementary Table S1.

PDLSCs were treated with ZOL for 72 h, fixed, permeabilized, and blocked, then incubated with Runx2 and OCN antibodies (overnight, 4 °C), followed by secondary antibodies and DAPI staining. Fluorescence images were captured using fluorescence microscope.

Constructs were prepared by seeding 5 × 10⁷ PDLSCs onto β-tricalcium phosphate (β-TCP) cuboids and transplanted into aseptically created subcutaneous pockets in 6-week-old male nude mice. A total of 30 constructs were randomly assigned to divided into six groups: control (β-TCP + PDLSCs without ZOL), and ZOL-treated groups at 0.01, 0.1, 0.5, 1, and 10 µM (n = 5 per group, one construct per mouse). After 12 weeks, the implants were harvested and fixed in 4% paraformaldehyde for histological analysis. All animal experimental procedures were approved by the Experimental Animal Welfare and Ethics Branch of Shanghai Ninth People’s Hospital (SH9H-2020-A670-1).

Transplants were decalcified in 10% EDTA for 2 months, then washed, dehydrated, and embedded in paraffin. The sections were cut at 5 μm thickness and stained with H&E and Masson staining. The sections were also evaluated by immunofluorescence, as described previously (Limones et al., 2020).

Periodontal ligament stem cells (PDLSCs) from a single donor were treated with ZOL (0.5 µM) or vehicle control for 3 days (n = 3, biological replicates per group). Gene expression profiling was performed using Affymetrix GeneChips. Raw data were normalized by the RMA method in R, and differentially expressed genes were identified as those showing a ≥2-fold change with p < 0.05. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted using the clusterProfiler package (v3.16.0). P-values were adjusted using the Benjamini–Hochberg (BH) method, and pathways with an adjusted p ≤ 0.05 were considered significant.

Cells were lysed in RIPA buffer, and proteins were extracted, separated by 10% SDS-PAGE, and transferred to membranes. After blocking, membranes were incubated with primary antibodies (overnight, 4 °C) and secondary antibodies (1 h, RT). GAPDH served as a loading control. Signals were visualized with ChemiDoc XRS and quantified using ImageJ.

All the data were obtained from at least three independent experiments. The values were presented as mean ± standard deviation (SD). The Student’s t-test and one-way ANOVA were used to perform statistical analyses using GraphPad Prism 7.0 (GraphPad Software, United States). P < 0.05 was considered statistically significant.

Primary adherent cells migrated from the periodontal ligament (PDL) tissues within 3–5 days, proliferating progressively to 90% confluence within 3–12 days and exhibiting spindle-shaped, fibroblast-like morphology (Figure 1A). Cells at passages 3–5 maintained this morphology (Figures 1B,C) and were used for experiments. PDLSCs exhibited colony-forming ability (Figures 1D,E) and multipotency: osteogenic differentiation was confirmed by ALP staining and mineralized nodule formation (Figures 1F,G), adipogenic differentiation by lipid droplet accumulation (Figure 1H), and chondrogenic differentiation by Alcian blue staining (Figure 1I). Flow cytometry showed strong expression of MSC markers STRO-1, CD90, and CD105, with minimal hematopoietic marker CD34 and CD45 expression (Figure 1J).

High concentrations ZOL (1 and 10 μM) significantly suppressed proliferation at 72 h, with 10 μM showing the strongest effect, while ≤0.5 μM had minimal impact (Figure 2A). To further determine whether the observed the inhibition of proliferation was associated with ZOL-induced cell deaths, apoptotic activity was assessed using complementary approaches. Flow cytometry analysis using Annexin V-FITC/PI staining, revealed a significant increase in the proportion of early apoptotic cells (Annexin V⁺/PI⁻) in both 1 μM and 10 μM ZOL-treated groups compared with controls, while the late apoptotic/necrotic population (Annexin V⁺/PI⁺) showed no statistically significant change (Figures 2B,C). Consistently, TUNEL staining demonstrated a markedly higher percentage of TUNEL⁺ cells in the high-dose ZOL groups (Figures 2D,E), indicating increased DNA fragmentation.

To further distinguish apoptosis from pyroptosis, we examined key molecular markers of distinct cell death pathways. Immunofluorescence staining and Western blot analysis (Figures 2F–I) showed robust upregulation of cleaved caspase-3 and cleaved PARP in PDLSCs treated with high-dose ZOL, confirming activation of the canonical apoptotic cascade. In contrast, expression levels of caspase-1, a central mediator of pyroptosis, remained unchanged across treatment groups, indicating that inflammatory/pyroptotic cell death was not activated under these conditions. These molecular findings are concordant with the flow cytometry results, which showed selective induction of early apoptosis without a concomitant increase in late apoptotic/necrotic cells. Morphological assessment further supported these observations. Control and ≤0.5 μM ZOL-treated PDLSCs retained typical spindle-shaped fibroblastic morphology with well-organized F-actin cytoskeletons (Figures 2J,K). In contrast, cells exposed to 1 μM and 10 μM ZOL exhibited pronounced morphological alterations, including cell shrinkage, irregular contours, and disrupted actin filament organization, which are characteristic of apoptotic cytoskeletal remodeling.

Collectively, these results demonstrate that ZOL exerts a dose-dependent cytotoxic effect on PDLSCs. At concentrations ≥1 μM, ZOL suppresses proliferation and induces classic apoptosis—characterized by early apoptotic signaling, caspase-3/PARP activation, and cytoskeletal disruption—without engaging pyroptotic pathways. These findings establish apoptosis as the primary mode of ZOL-induced cell death in PDLSCs and provide a mechanistic basis for dose selection in subsequent experiments.

ALP staining on day 7 (Figure 3A) and alizarin red staining on day 14 (Figure 3B) revealed a biphasic response: low concentrations of ZOL (≤0.5 μM) enhanced osteogenic differentiation in a dose-dependent manner, whereas high concentrations (10 μM) markedly suppressed osteogenic activity. This dose-dependent dual effect was further supported by quantitative analysis of ALP activity (Figure 3C) and alizarin red content (Figure 3D), both of which peaked at 0.5 μM and declined sharply after 1 μM.

To further elucidate the molecular mechanisms underlying the biphasic effect, we assessed transcriptional levels of key osteogenic genes by RT-qPCR. ALP, RUNX2, and OCN peaked at 0.5 μM ZOL, whereas BMP2 peaked at 0.1 μM. High-dose ZOL (10 μM) significantly suppressed all four genes. In the 1 μM group, mRNA levels were lower than 0.5 μM but remained higher than control (Figure 3E). Consistently, immunofluorescence showed highest OCN and Runx2 at 0.5 μM, moderate reduction at 1 μM, and marked suppression at 10 μM (Figure 3F). Collectively, ZOL regulates PDLSC osteogenesis in a dose-dependent, biphasic manner: low concentrations (≤0.5 μM) enhance differentiation, high concentrations (≥10 μM) inhibit it, and 1 μM represents a transitional threshold. Thus, 0.5 μM was selected as the optimal concentration for subsequent assays.

Further evaluate the in vivo effects of ZOL on the osteogenesis of PDLSCs, β-TCP scaffolds seeded with PDLSCs were implanted subcutaneously into nude mice. After 12 weeks, these implants were harvested for histological analysis. H&E staining (Figure 4A) demonstrated a marked increase in new bone formation in the 0.1 μM and 0.5 μM ZOL groups compared with the control group. In contrast, new bone formation was profoundly diminished in the 1 μM and 10 μM groups, showing a statistically significant reduction compared with the control group (Figure 4D). Masson’s trichrome staining (Figure 4B) revealed abundant, well-organized collagen fibers in the 0.1 μM and 0.5 μM ZOL groups. In contrast, 1 μM and 10 μM groups displayed markedly reduced and sparse collagen deposition. Quantitative analysis of collagen fraction confirmed a significant increase in collagen content in the 0.1 μM and 0.5 μM groups compared to the control (Figure 4E). Immunofluorescence staining for OCN and Runx2 (Figure 4C) further validated these observations. Both the intensity and area of positive staining were markedly elevated in the 0.1 μM and 0.5 μM groups, moderate in the control group, and substantially diminished in the 1μM and 10 μM groups (Figures 4F,G).

These results demonstrate that ZOL at 0.1 μM and 0.5 μM significantly promotes ectopic bone formation of PDLSCs in vivo, whereas 0.01 μM shows negligible effect, and higher concentrations (1 μM and 10 μM) markedly suppress their osteogenesis.

To explore mechanisms of ZOL-induced osteogenesis in PDLSCs, we performed an Affymetrix Gene Expression Array. Differentially expressed genes exhibiting more than 2-fold up or downregulation were identified in the heat map (Figure 5A). GO and KEGG analyses identified the top 20 enriched biological processes and signaling pathways (Figures 5B,C). Among these, the MAPK and Wnt signaling pathways as prominent signaling cascades potentially mediating ZOL-induced osteogenic differentiation. Differentially expressed genes involved in these two pathways, showing greater than twofold changes, are listed in Figures 5D,E.

To verify the involvement of the Wnt/β-catenin and MAPK pathways in ZOL-promoted osteogenesis in PDLSCs, we performed Western blot analysis under both low-dose and high-dose ZOL conditions (Figures 5F–I). Under osteogenic-promoting conditions (0.5 µM ZOL), β-catenin protein expression was significantly increased as early as 12 h and remained consistently elevated from 12 to 72 h (Figure 5F). Concomitantly, phosphorylated GSK (p-GSK) levels were markedly decreased, suggesting inhibition of GSK kinase activity and consequent stabilization of β-catenin. Activation of MAPK signaling was evident after 24 h of ZOL exposure, as indicated by sustained and pronounced increases in phosphorylation of p38 and JNK from 24 to 72 h, with peak activation observed at 72 h. Quantitative densitometric analysis (Figure 5G) confirmed significant elevations in β-catenin, p-p38, and p-JNK levels, together with a robust reduction in p-GSK expression. In contrast, high-dose ZOL treatment (10 µM) failed to elicit sustained activation of either pathway (Figures 5H,I). At early time points (12–24 h), β-catenin abundance and phosphorylation levels of p38 and JNK remained largely unchanged compared with controls. However, prolonged exposure (48–72 h) resulted in a significant reduction of β-catenin protein, accompanied by increased p-GSK levels, indicative of enhanced β-catenin degradation and suppression of canonical Wnt signaling. In parallel, phosphorylation of p38 and JNK was markedly decreased at 48 and 72 h, while total protein levels remained unchanged, reflecting inhibition of MAPK pathway activation rather than altered protein expression. Quantitative analysis further substantiated these inhibitory effects. Together, these data demonstrate a dose-dependent divergence in signaling responses to ZOL: low-dose ZOL induces coordinated and sustained activation of Wnt/β-catenin and MAPK pathways, whereas high-dose ZOL suppresses both signaling axes. This comparative analysis establishes high-dose ZOL as an effective negative control for pathway activation and supports the conclusion that selective activation of Wnt/β-catenin–MAPK signaling underlies the osteogenic-promoting effects of low-dose ZOL in PDLSCs.

To determine whether the Wnt/β-catenin and MAPK signaling pathways mediate the osteogenic effects of ZOL on PDLSCs, we selectively inhibited β-catenin (LF3), p38 (SB203580), and JNK (SP600125). PDLSCs were pretreated with each inhibitor before ZOL stimulation. Western blot analysis revealed that LF3, SB203580, and SP600125 effectively reduced β-catenin accumulation, as well as the phosphorylation of p38, and JNK, respectively, after 24 h (Figures 6A–C). Quantitative densitometry confirmed significant suppression of pathway activation (Figures 6D–G). Functionally, inhibition of either the Wnt/β-catenin or MAPK pathways markedly reduced ALP activity (Figure 6H) and mineralized nodule formation (Figure 6I) in ZOL-treated PDLSCs. Consistently, real-time PCR analysis showed that blockade of these pathways significantly downregulated the expression of osteogenic marker genes (BMP2, RUNX2, ALP, OCN) (Figure 6L). Collectively, these findings demonstrate that activation of the Wnt/β-catenin and MAPK pathways plays a positive and functional role in mediating the ZOL-induced osteogenic differentiation of PDLSCs.