Timosaponin B- II Enhances Osteogenic Differentiation of Human Periodontal Ligament Stem Cells via PI3K/AKT/GSK3β Signaling Pathway

Abstract

Background

This investigation aims to elucidate the effects of Timosaponin B-II (TB-II) on the proliferation and osteogenic differentiation of human periodontal ligament stem cells (hPDLSCs) through both in vitro experiments and an in vivo orthodontic tooth movement model utilizing rats. The primary objective is to clarify the mechanisms by which TB-II influences the remodeling of periodontal tissue under biomechanical stress, thereby providing insights into its potential role in reducing relapses after orthodontic tooth movement.

Methods

hPDLSCs were isolated and characterized via flow cytometry and multilineage differentiation assays (osteogenic and adipogenic induction). The impact of TB-II on the expression levels of osteogenic genes and proteins, including runt-related transcription factor-2 (RUNX-2), alkaline phosphatase (ALP), and collagen type 1 (COL-1), was evaluated through quantitative real-time PCR (qRT-PCR) and Western blotting. Alizarin Red Staining (ARS) was utilized to assess the formation of mineralized nodules. Additionally, the involvement of the phosphatidylinositol - 3 – kinase (PI3K)/ protein kinase B(AKT)/ glycogen synthase kinase - 3β(GSK3β) signaling pathway in TB-II-mediated osteogenesis was explored using pharmacological inhibitors (LY294002 for PI3K/AKT and CHIR-99021 for GSK3β). Western blot analysis identified key osteogenic markers (GSK3β, p-GSK3β, AKT, p-AKT) in treated cells. For in vivo validation, eighteen male Wistar rats were randomly divided into TB-II-treated and saline-control groups. Micro-computed tomography (micro-CT) evaluated tooth movement and alveolar bone structural changes. Histological assessment included hematoxylin–eosin (HE) staining, Masson trichrome staining, and tartaric-resistant acid phosphatase (TRAP) staining to analyze periodontal tissue morphology. Immunohistochemical (IHC) analysis assessed osteogenic markers (RUNX-2, ALP, COL-1) and the osteoclastogenic regulator RANKL to evaluate tissue remodeling. All statistical analyses were performed using GraphPad Prism 8. Comparisons between groups were conducted via one-way/two-way ANOVA with Tukey’s post-hoc test. Values of p < 0.05 were regarded as statistically significant.

Results

In vitro studies revealed that TB-II at 20 μM significantly enhanced the proliferation, ALP activity, and mineralized nodule formation of hPDLSCs, accompanied by markedly elevated expression of RUNX-2, ALP, COL-1, p-AKT and p-GSK3β. Pharmacological inhibition of the PI3K/AKT pathway via LY294002 abolished TB-II’s osteogenic effects, while treatment with CHIR99021 indicated that GSK3β activity was downstream and regulated by the PI3K/AKT signaling axis. In vivo, TB-II administration in a rat orthodontic tooth movement (OTM) model upregulated RUNX-2, ALP, and COL-1 expression on the tension side of tooth roots, while simultaneously reducing TRAP + osteoclast numbers and inhibiting RANKL expression.

Conclusion

TB-II stimulates the proliferation and osteogenic maturation of hPDLSCs in vitro by activation of the PI3K/AKT/GSK3β signaling axis. In vivo investigations using OTM model further demonstrate that TB-II enhances periodontal tissue regeneration. Collectively, these results highlight the therapeutic potential of TB-II in preventing relapses following OTM, positioning it as a viable candidate for clinical strategies aimed at stabilizing orthodontic outcomes.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s12015-025-10962-0.

Background

Periodontitis, a prevalent inflammatory disorder characterized by progressive periodontal tissue destruction, remains a leading cause of tooth loss [1]. Healthy periodontal architecture is critical for sustaining masticatory forces and enabling successful orthodontic tooth movement (OTM) [2]. Following the completion of orthodontic treatment, repositioned teeth and dental arches often tend to revert to their pre-treatment positions, a phenomenon known as relapse [3]. This tendency is more pronounced in patients with periodontitis [4]. Consequently, accelerating periodontal tissue regeneration and enhancing the remodeling rate of alveolar bone has become a key focus in research. Furthermore, it is acknowledged that the biological mechanisms underlying relapse following OTM involve periodontal tissue remodeling processes that resemble those occurring during OTM itself [5]. To achieve optimal therapeutic outcomes and minimize relapses, an in-depth study of periodontal tissue regeneration and remodeling is essential, encompassing the etiology, pathophysiology, and treatment approaches of periodontal disease [6, 7].

Currently, several approaches for periodontal tissue regeneration exist, such as tissue engineering, gene therapy, biomaterials, and pharmacotherapy [8, 9]. Each of these methods possesses unique advantages and limitations, necessitating further investigation into their applicability and clinical efficacy. At present, tissue engineering is the primary method for periodontal tissue regeneration [8]. This approach constructs three-dimensional graft complexes in vitro by utilizing seed cells, bioactive molecules, and biological scaffolds, thereby aiming to restore the health and function of periodontal tissues [10]. Seed cells are one of the core and indispensable components in tissue engineering [11, 12]. Among them, human periodontal ligament stem cells (hPDLSCs) are adult stem cells with biological properties: they exhibit self-renewal capacity, clonal proliferative capability, and multilineage differentiation potential toward osteogenic, adipogenic, chondrogenic, and odontogenic lineages. Additionally, hPDLSCs possess immune-modulatory activity that suppresses inflammatory responses while promoting tissue repair, making them ideal for regenerative applications in inflamed periodontal environments [13, 14]. Notably, hPDLSCs secrete paracrine factors such as VEGF, BMP-2, and TGF-β, which enhance angiogenesis, osteogenesis, and extracellular matrix remodeling. Since hPDLSCs can be isolated from periodontal tissues, they are among the most suitable seed cells for periodontal regeneration [14]. These properties collectively enable hPDLSCs to actively participate in maintaining periodontal homeostasis and facilitate regeneration of periodontal ligament, alveolar bone, and cementum in vivo [14, 15]. Bioactive molecules mainly include drugs and growth factors that control inflammation, promote cell proliferation, and induce differentiation [16].

Anemarrhena asphodeloides (Zhi Mu), a Liliaceae plant-derived herb in traditional Chinese medicine, has garnered attention for its multitarget efficacy in neurodegenerative diseases, osteoporosis, and inflammatory disorders [17]. Modern pharmacological studies have demonstrated that Zhi Mu can alleviate cognitive impairment, osteoporosis, and Alzheimer's disease, and it exhibits anti-inflammatory, antidepressant, hypoglycemic, and antitumor properties [18, 19]. Qin et al. [20] reported that TB-II can promote osteoblast proliferation and enhance serum alkaline phosphatase activity. Wang et al.[21] established an osteoporosis model in mice through diabetic model and found that TB-II intervention can controvert the decrease of trabecular bones and their bone mineral content caused by the diabetes in the model.

The PI3K/AKT signaling pathway is integral to various physiological and pathological processes, playing a pivotal role in biological functions such as differentiation regulation, cell proliferation influence, and modulation of metabolism and survival [22, 23]. It is one of the most crucial pathways for controlling osteogenic differentiation and bone growth. Prior research has indicated that the activation of the PI3K/AKT pathway greatly enhances the expression of osteogenic genes [24], whereas the use of the pathway inhibitor LY294002 produces the opposite effect [25]. A key downstream effector of this pathway is glycogen synthase kinase 3β (GSK3β), which modulates osteogenic gene expression through β-catenin signaling [26]. Pharmacological tools such as CHIR99021 are widely used to study the pathway’s role in bone formation and remodeling. Activation of PI3K/AKT inhibits GSK3β, promoting osteogenic differentiation [27].

Based on these observations, we propose that TB-II stimulates proliferation and osteogenic maturation of hPDLSCs in vitro through activation of the PI3K/AKT/GSK3β signaling pathway. This mechanism likely drives periodontal tissue remodeling, thereby attenuating OTM relapse.

Materials and Methods

Isolation and Characteristics of hPDLSC

This research received approval from the Ethics Committee of the School of Stomatology at Shandong University (Approval No.20230703). Legal guardians were provided with a detailed explanation of the study's purpose and gave consent for their children to participate. Premolars from 14- to 22-year-old orthodontic patients, free of caries and periodontal disease, were collected following informed consent. Immediately post-extraction, the premolars were placed in pre-chilled α-minimum essential medium (α-MEM; Yuanpei, Shanghai, China), supplemented with L-glutamine, nucleosides, deoxynucleotides, and 1% antibiotic–antimycotic solution (PSG, 100 ×; Biosharp, Guangzhou, China). The teeth were transported on ice to the laboratory, rinsed with PSG-containing phosphate-buffered saline (PBS; Solarbio, Beijing, China), and processed under sterile conditions. Periodontal ligament tissue from the mid-root region was dissected and plated onto culture flasks. Cultures were maintained in α-MEM supplemented with 20% fetal bovine serum (FBS; Yuanpei, Shanghai, China) and 1% PSG at 37 °C with 5% CO₂. After 7 days, spindle-shaped hPDLSCs emerged from tissue fragments. Upon reaching 70–80% confluence, cells were detached using trypsin (Yeasen, Shanghai, China), neutralized with FBS-supplemented α-MEM, and observed under microscopy to confirm cell viability. Single-cell clones were isolated via limiting dilution, and cells between passages 3 (P3) and 5 (P5) were selected for subsequent experiments.

Flow Cytometry

A suspension of hPDLSCs in PBS was prepared, adjusting the cell concentration to 1 × 10⁷ cells/mL. Subsequently, 100 μL of this suspension was allocated into five distinct 1.5 mL microcentrifuge tubes. One tube served as the control, receiving an equivalent volume of PBS, while the remaining four tubes were separately treated with antibodies targeting CD34, CD44, CD45, and CD105 (Elabscience, Houston, TX, USA). To safeguard the samples from light exposure, all tubes underwent incubation in dark conditions. Following a 20-min incubation period on ice, the cells were washed thoroughly. Flow cytometry was utilized to analyze the samples, with data processed and evaluated using FlowJo software (version 10, BD Biosciences, USA).

CCK-8 Assay

hPDLSCs were seeded at a density of 2 × 10³ cells per well in 96-well plates. After confirming cell adhesion, α-MEM supplemented with TB-II (purity ≥ 98%; Yuanye, Guangzhou, China) at concentrations of 0, 1, 5, 20, 80, and 100 μM was added to each well, with quintuplicate replicates for each condition. Cell proliferation was evaluated at 1, 3, and 5 days using the CCK-8 assay (Biosharp, Guangzhou, China). For the assay, 100 μL of culture medium containing CCK-8 reagent (10α-MEM: 1CCK-8 ratio) was added to each well, followed by incubation at 37 °C for 45 min. Optical density (OD) values were measured at 450 nm using a microplate reader (BMG Labtech, Ortenberg, Germany).

ARS and Calcium Quantification

hPDLSCs were seeded 1 × 10⁵ cells per well in 6-well plates. Following attachment, osteogenic differentiation was induced using a specialized medium containing α-MEM supplemented with 10% FBS, 50 μg/mL ascorbic acid, 10 mM β-glycerophosphate, and 0.01 μM dexamethasone. After 28 days of differentiation, cells were washed thrice with PBS and fixed in 95% ethanol. Mineralized nodules were stained with Alizarin Red S (ARS, Beyotime Biotechnology, Shanghai, China) and imaged using an Olympus microscope (Olympus Corporation, Japan). Quantification of mineralization was achieved by extracting the dye with 10% (w/v) cetylpyridinium chloride (CPC), followed by incubation for 15 min. Absorbance of the eluate was measured at 562 nm to assess calcium deposition.

Oil Red O Staining

The same cell culture method as previously described was used in this experiment. Adipogenic differentiation was induced by treating the cells with a medium containing α-MEM supplemented with 10% FBS, 0.5 μM hydrocortisone, 500 μM isobutyl methylxanthine, 60 μM indomethacin, and 10 μM insulin. After 21 days of induction, the cells were rinsed with PBS and fixed in 4% formaldehyde. Oil Red O staining solution (Solarbio, Beijing, China) was then applied to the cells. Stained cells were observed and captured under a microscope.

ALP Staining and ALP Activity Assay

Cells were seeded at a density of 5 × 10⁴ per well in a 12-well plate. After cell attachment, the medium was replaced with osteogenic induction medium. Following 14 days of incubation, the cells were washed with PBS and fixed in 4% formaldehyde. BCIP/NBT Alkaline Phosphatase Colorimetric Assay Kit (Beyotime, Shanghai, China) was applied, where a deeper blue color indicated higher ALP activity.

After the 14-day induction period, hPDLSCs were washed with ice-cold PBS and lysed using RIPA buffer (Beyotime, Shanghai, China) supplemented with protease and phosphatase inhibitors. The lysates were incubated on ice for 15 min, followed by centrifugation to collect the supernatant. ALP activity in the supernatant was evaluated using the Alkaline Phosphatase Assay Kit (Solarbio, Beijing, China), while protein concentration was determined with the BCA Protein Assay Kit (Solarbio, Beijing, China). For further analysis, the protein samples were mixed with 5 × SDS-PAGE loading buffer (Beyotime, Shanghai, China) and heated at 100 °C for 5 min to ensure protein denaturation.

RNA Extraction and qRT-PCR Analysis

Following a 14-day osteogenic induction period, total RNA was isolated from the cells using the Trizol method. The obtained RNA was converted into complementary DNA (cDNA) utilizing a reverse transcription kit (Yeasen, Shanghai, China). Each sample was processed in triplicate for accuracy. The primer sequences used in this study were carefully designed and analyzed as Supplementary Table 1.

Western Blot Analysis

Protein samples (20 µg/lane) were separated by SDS-PAGE and transferred to PVDF membranes (Millipore, Merck, USA). Membranes were blocked for 15 min at room temperature using rapid blocking solution (Servicebio, Wuhan, China). Primary antibodies—sourced from Proteintech (Wuhan, China) and Abways (Shanghai, China)—were selected based on target protein molecular weights and included COL-1, ALP, RUNX-2, and GAPDH (Proteintech); GSK3β, AKT, and phosphorylated-AKT (p-AKT) (Abways). All primary antibodies were incubated overnight at 4 °C.

Following incubation, membranes were washed three times with TBST (Tris-buffered saline with 0.1% Tween-20; Servicebio) for 10 min each. Secondary antibodies were matched to the species of primary antibodies. Proteins were detected using a chemiluminescent substrate (Yeasen, Shanghai, China) and visualized on an Amersham Imager 600 (GE Healthcare, Little Chalfont, UK). Experiments were performed in quintuplicate to ensure reproducibility.

Immunofluorescence Assay

Cells were fixed at 20–25 °C for 30 min using 4% paraformaldehyde, followed by permeabilization with 0.1% Triton X-100 and blocking with 5% bovine serum albumin (BSA; both reagents from Solarbio, Beijing, China) for 12 min and 1 h, respectively. Primary antibody targeting COL-1 (1:300 dilution, Immunoway, Suzhou, China) was applied overnight at 4 °C. Following removal of the primary antibody, cells were incubated with FITC-conjugated goat anti-rabbit IgG secondary antibody (1:300 dilution, Proteintech, Wuhan, China) in the dark for 1 h at room temperature. Nuclei were counterstained with DAPI (Solarbio), and fluorescent images were captured using a Leica fluorescence microscope (Wetzlar, Germany). Fluorescence intensity quantification was performed via ImageJ software (National Institutes of Health, Bethesda, MD, USA).

LY294002 and CHIR-99021 Treatment

LY294002, a non-specific inhibitor of the PI3K/AKT signaling pathway, and CHIR-99021, a selective GSK3β inhibitor, were used in combination with TB-II to investigate their effects on osteogenic differentiation. Four experimental groups were established: (1) Control group—cells cultured in osteogenic medium alone; (2) TB-II group—cells treated with osteogenic medium supplemented with 20 μM TB-II (Yuanye, Guangzhou, China); (3) TB-II + LY294002 group—cells exposed to osteogenic medium containing 20 μM TB-II and 10 μM LY294002 (Solarbio, Beijing, China); and (4) TB-II + CHIR-99021 group—cells cultured in osteogenic medium with 20 μM TB-II and 5 μM CHIR-99021 (Solarbio, Beijing, China). After 14 days of osteogenic induction, alkaline phosphatase (ALP) staining and Western blotting were performed to evaluate protein expression levels.

Establishment of the OTM Model

Eighteen male Sprague–Dawley rats (7 weeks old, 250 ± 5 g) were sourced from Beijing Charles River Laboratories and maintained under specific pathogen-free (SPF) conditions with controlled temperature (20–25 °C), humidity (65–70%), and a 12-h light–dark cycle. Experimental sample size was determined via the resource equation method (see Supplementary Data 1). All procedures adhered to the NIH Guidelines for Laboratory Animal Care and were approved by Shandong University’s IACUC (Approval No. 20230702).

Animals were anesthetized with inhaled isoflurane (Yipin Pharmaceutical Co., Hebei, China). The orthodontic appliance comprised a 0.20 mm stainless steel archwire (Aosu, Hangzhou, China), a nickel-titanium tension spring, and polymer anchorage material (Charisma; Heraeus Kulzer, Germany). The spring applied a calibrated 50 g force (measured via an orthodontic dynamometer, XIHUBIOM, Hangzhou, China) to displace the left maxillary first molar toward the midline by connecting it to the central incisor. Rats were randomly divided into two groups (n = 9 per group): the control group received daily intraperitoneal injections of saline, while the TB-II group was administered TB-II (80 mg/kg/day) over the experimental period.

Micro-CT

Following anesthesia, rats underwent transcardial perfusion via the left ventricle with 4% paraformaldehyde to fix the maxillary bone, with a focus on the right maxillary first molar. Post-perfusion, specimens were immersed in 4% paraformaldehyde at 4 °C for 24 h, followed by dehydration and mounting on a specimen platform. Micro-CT scanning (Quantum GX2, PerkinElmer, USA) was performed under standardized parameters (50 kV, 88 μA, 36 μm slice thickness) to capture the right maxillary region, encompassing the three molars and adjacent alveolar bone (scan duration: 14 min per sample). Exported DICOM datasets were processed using RadiAnt Viewer (Medixant, Poland) for 3D reconstruction and axial alignment. OTM distance was quantified as the shortest linear separation between the distal surface of the first molar and the mesial surface of the second molar on the right maxilla. Bone microstructural parameters—including bone volume fraction (BV/TV), trabecular thickness (Tb.Th), and trabecular separation (Tb.Sp)—were analyzed in the region of interest (ROI) using CTAn software (Skyscan NV).

HE and TRAP Staining

The specimens were decalcified over 12 weeks in a 10% EDTA solution (pH 7.2). Once decalcification was complete, the samples were dehydrated gradually using ethanol and embedded in paraffin wax. Sections were prepared by slicing the specimens in a proximal-to-distal orientation, with each slice measuring 4 μm in thickness.

To assess periodontal tissue morphology, hematoxylin and eosin (HE) staining was performed on the slides. Additionally, a TRAP staining kit (Solarbio, Beijing, China) was utilized to identify TRAP-positive (TRAP +) cells, which indicate osteoclast activity. The TRAP + cells were quantified using ImageJ software for further analysis.

Masson’s Trichrome Staining

For Masson trichrome staining, 5–7 µm paraffin sections are dewaxed to water, followed by nuclear staining with Weigert’s iron hematoxylin solution (Servicebio, Wuhan, China) for 10–15 min. Excess dye is removed via 10-s differentiation in 1% HCl ethanol, after which sections are rinsed and incubated in acidic Biebrich scarlet solution (pH 1.5–2.0, Servicebio, Wuhan, China) for 60 min in the dark to stain collagen fibers red. Collagen-specific staining is enhanced by treating sections with phosphomolybdic acid differentiator for 1–2 min to eliminate non-collagen red pigments. Sections are then immersed in Aniline Blue solution (pH 2.5, Servicebio, Wuhan, China) for 10–15 min to deepen collagen blue staining, followed by a brief 0.5% acetic acid wash. Subsequent counterstaining includes 3–5 min in Orange G solution (Servicebio, Wuhan, China) for muscle fibers and 1–2 min in weak acid fuchsin for cytoplasm. Finally, sections are dehydrated through ethanol gradients, cleared in xylene, and mounted with neutral resin. The final staining reveals collagen fibers in vivid blue, muscle/cytoplasm in pink, nuclei in black, and erythrocytes/matrix in pale yellow or colorless, enabling morphological analysis of collagen and connective tissues.

IHC Staining

The slide was submerged in citrate buffer solution for antigen retrieval via high-temperature pretreatment. Non-specific binding was blocked by incubation with 5% goat serum for a 40-min period. The specimen was then immersed in a primary antibody cocktail comprising polyclonal anti-RUNX-2 (rabbit origin, 1:300; Proteintech, Wuhan, China), COL-1 (1:300; ImmunoWay, Plano, TX, USA), and RANKL (1:300; Servicebio, Wuhan, China) antibodies, followed by overnight incubation at refrigerated temperature (4 °C). After thorough washing, the sample underwent secondary antibody conjugation using a biotinylated anti-mouse/rabbit IgG antibody (minimum 30-min incubation). The staining protocol concluded with SP reagent kit (Servicebio, Wuhan, China) processing for signal amplification after final wash cycles. Nuclei were counterstained with hematoxylin, and microscopic visualization was conducted to assess immunoreactivity patterns. Quantitative analysis of immunostaining intensity was performed via ImageJ software to calculate the arithmetic mean of optical density (AOD) values across analyzed fields.

Data Analyze

Statistical analyses were conducted using SPSS 16.0 software (SPSS Inc., Chicago, IL, USA). Data followed a normal distribution and were expressed as mean ± standard deviation. Each experiment was independently repeated a minimum of three times. Comparative analyses were performed using the Student's t-test or one-way ANOVA. A significance threshold of P < 0.05 was applied for all statistical evaluations.

Results

Isolation and Characteristics of hPDLSCs

Within 1 to 2 weeks, primary cells began to migrate from the edges of the periodontal tissue block (Fig. 1A). These cells were harvested, enzymatically digested, and cultured, with passages 3 to 5 utilized for subsequent experiments (Fig. 1B). Following 21 days of osteogenic induction and 28 days of adipogenic induction, ALP, Alizarin Red and Oil Red O staining confirmed the cells’ multilineage differentiation potential (Fig. 1D, E and F). Flow cytometry revealed low expression levels of CD34 and CD45, alongside high expression levels of CD44 and CD105, confirming the stem cell characteristics of the isolated population (Fig. 1G, H, I, J).

Fig. 1.

Cultivation and characterization of hPDLSCs. (A) Spherical cells migrated from the edges of the tissue blocks. Scale bar: 500 μm. (B) Around two weeks later, hPDLSCs displayed a spindle-like morphology and organized into a spiral arrangement. Scale bar: 500 μm. (C) Chemical structure of Timosaponin B-II (TB-II; C45H76O19). (D) After osteogenic differentiation, Alizarin Red staining revealed dark brown mineralized nodules in hPDLSCs. Scale bar: 500 μm. (E) After osteogenic differentiation, alkaline phosphatase staining revealed blue nodules in hPDLSCs. Scale bar: 500 μm. (F) Following adipogenic differentiation, Oil Red O staining highlighted lipid droplets of varying sizes within the hPDLSCs. Scale bar: 500 μm. (G–J) Flow cytometry analysis of hPDLSC surface markers. CD34 (G) and CD45 (I) showed negligible expression, while CD44 (H) and CD105 (J) were strongly expressed

Effect of TB-II on Cell Proliferation

The CCK-8 assay was performed to assess the proliferation of hPDLSCs treated with TB-II for 1, 3, and 5 days. The results indicated that TB-II at concentrations of 20 μM slightly enhanced cell proliferation on days 1 and 3, though the effect was minimal (Fig. 2A and B). By day 5, significant promotion of cell proliferation was observed at 5 μM and 20 μM TB-II, while 1 μM and 80 μM TB-II showed no notable impact on proliferation. 100 μM TB-II showed a decrease in proliferation. According to the above results, 5, 20, and 80 μM CAT were selected as low, medium, and high concentrations for follow-up experiments.

Fig. 2.

Investigation into TB-II’s effects on hPDLSC proliferation. (A and B) The proliferative capacity of hPDLSCs exposed to TB-II at concentrations of 0, 1, 5, 20, 80, and 100 μM was assessed via CCK-8 assays at 1, 3, and 5 days. Two-way ANOVA analysis revealed that 20 μM TB-II induced the most significant proliferative response on day 5 compared to control (***P < 0.001). Error bars represent mean ± SD

Effect of TB-II on ALP Activity and Calcified Nodules

ALP activity assay revealed a dose-dependent increase in ALP activity in TB-II treated group compared to the control group, with significant elevation observed at concentrations of 20 μM (P < 0.0001) (Fig. 3B). ALP staining further demonstrated enhanced osteogenic differentiation, as evidenced by intense dark-blue staining in the 20 μM TB-II group (Fig. 3A). ARS staining qualitatively confirmed this trend, with denser and more mineralized nodules observed in TB-II treated group, particularly in the 20 μM group (Fig. 3C). Calcium quantification corroborated these findings, showing a significant increase in calcium deposition in TB-II treated group compared to control group (P < 0.0001) (Fig. 3D). Collectively, these results indicate that TB-II promotes osteogenic maturation in hPDLSCs in a concentration-dependent manner, with optimal effects observed at 20 μM.

Fig. 3.

Investigation of TB-II's influence on the osteogenic differentiation of hPDLSCs. (A) ALP staining was performed on hPDLSCs treated with TB-II (0, 1, 10, or 100 μM) during 14 days of osteogenic induction. The most pronounced staining (blue) was discovered at 20 μM TB-II (Scale bar: 500 μm). (B) Quantification of ALP activity revealed significant increases at 5 μM (*P < 0.05) and 20 μM (**P < 0.01) TB-II compared to control (one-way ANOVA). (C) After 28 days of differentiation, ARS staining demonstrated enhanced mineralization (intense red deposits) and larger calcified nodules at 20 μM TB-II, with minimal effects at 80 μM (Scale bar: 500 μm). (D) 20 μM TB-II induced the strongest mineralization compared to control in the quantitative ARS analysis (****P < 0.0001, one-way ANOVA). All data are presented as mean ± SD

Effect of TB-II on the Expression of RUNX-2, ALP, COL-1, p-AKT and GSK3β

qRT-PCR demonstrated that treatment with 20 and 50 μM TB-II significantly upregulated osteogenesis-related genes, including RUNX2 (Fig. 4E, P < 0.001), ALP (Fig. 4F, P < 0.001), and COL1A1 (Fig. 4G, P < 0.0001). No statistically significant difference was observed between the control group and 5 μM TB-II. Western blot analysis (Fig. 4A and H) confirmed these findings, showing that TB-II at 5, 20, and 50 μM concentration-dependently increased protein expression of osteogenesis markers: RUNX-2 (Fig. 4B), ALP (Fig. 4C), COL-1 (Fig. 4D), p-GSK3β (Fig. 4I), and p-AKT (Fig. 4J) (P < 0.0001). The most pronounced effects were observed at 20 μM TB-II, consistent with the qRT-PCR results. Based on these experiments, 20 μM TB-II was selected as the optimal concentration for following experiments.

Fig. 4.

TB-II enhanced the expression of markers associated with osteogenic differentiation in hPDLSCs. (A) Western blot analysis of RUNX-2, ALP, and COL-1 protein levels in hPDLSCs treated with TB-II (0, 5, 20, 80 μM) after 14 days of osteogenic induction. (B-D) Quantification of protein expression for RUNX-2 (B), ALP (C), and COL-1 (D). One-way ANOVA revealed significant increases at 20 μM TB-II (****P < 0.0001) compared to control. Error bars represent mean ± SD. (E–G) qRT-PCR analysis of RUNX-2 (E), ALP (F), and COL-1 (G) mRNA levels. One-way ANOVA showed significant upregulation at 20 μM TB-II (***P < 0.001 for E and G; ****P < 0.0001 for F). Error bars represent mean ± SD. (H) Western blot analysis of p-GSK3β and p-AKT in hPDLSCs exposed to TB-II (0, 5, 20, 80 μM) after 14 days of osteogenic induction. (I and J) Quantification of p-GSK3β (I) and p-AKT (J) protein levels. One-way ANOVA demonstrated significant increases at 20 μM TB-II (****P < 0.0001). Error bars represent mean ± SD

TB-II Stimulates Osteogenic Differentiation by Activating the PI3K/AKT/GSK3β Pathway

To elucidate the molecular mechanisms underlying TB-II induced osteogenic differentiation of hPDLSCs, cells were treated with 20 μM TB-II in combination with PI3K inhibitor LY294002 or Wnt pathway activator CHIR99021. Co-administration with LY294002 significantly suppressed osteogenic maturation: ALP activity decreased (P < 0.0001), staining intensity diminished (Fig. 5A and B), calcium nodule formation reduced (P < 0.0001, Fig. 5D), and ARS staining became lighter (Fig. 5C). Conversely, CHIR99021 co-treatment enhanced osteogenesis, evidenced by elevated ALP activity (P < 0.0001), intensified ALP staining and ARS staining (Fig. 5A and C), and increased mineralized nodule deposition (P < 0.01, Fig. 5D). Western blot analysis (Fig. 6A and E) revealed that combined TB-II/LY294002 treatment downregulated key osteogenic markers RUNX-2 (P < 0.0001, Fig. 6B), ALP (P < 0.0001, Fig. 6C), and COL-1 (P < 0.0001, Fig. 6D), alongside reduced phosphorylation of GSK3β (P < 0.0001, Fig. 6F) and AKT (P < 0.0001, Fig. 6G) compared to TB-II alone. This indicated concurrent inhibition of both the PI3K/AKT and Wnt/β-catenin pathways, resulting in suppressed osteogenic differentiation. Conversely, TB-II/CHIR99021 co-treatment upregulated osteogenic marker expression (RUNX-2: P < 0.01, ALP: P < 0.001, COL-1: P < 0.001) and enhanced phosphorylation of GSK3β (P < 0.001) and AKT (P < 0.05), demonstrating dual activation of the Wnt/β-catenin and PI3K/AKT pathways. Notably, CHIR99021’s ability to potentiate TB-II’s osteogenic effects, coupled with the observation that AKT phosphorylation preceded Wnt pathway activation, provides experimental evidence that the Wnt/β-catenin pathway functions downstream of the PI3K/AKT signaling axis in mediating TB-II’s osteogenic effects. Immunofluorescence analysis (Fig. 7A) demonstrated that treatment with 20 μM TB-II combined with CHIR99021 induced the highest COL-1 expression and exhibited the strongest fluorescence intensity, whereas the same TB-II concentration paired with LY294002 resulted in the lowest intensity (Fig. 7B). These findings align with our previous observations.

Fig. 5.

TB-II activated the PI3K/AKT/GSK3β signaling pathway to promote osteogenic differentiation in hPDLSCs. (A) ALP staining images were captured after 14 days of osteogenic induction in control group, TB-II-treated group (20 μM), and co-treated groups with TB-II plus either LY294002 or CHIR99021. (B) ALP activity was quantified via one-way ANOVA, revealing significant differences among control, TB-II, TB-II + LY294002, and TB-II + CHIR99021 groups (****P < 0.0001; mean ± SD). (C) Representative ARS staining images displayed mineralization patterns after 28 days of osteogenic induction across control, TB-II, TB-II + LY294002, and TB-II + CHIR99021 groups. (D) Quantitative ARS analysis via one-way ANOVA demonstrated marked distinctions between experimental conditions (****P < 0.0001; mean ± SD), with all TB-II-containing groups maintained at 20 μM concentration under identical osteogenic protocols

Fig. 6.

TB-II activated the PI3K/AKT/GSK3β signaling pathway to promote osteogenic differentiation in hPDLSCs. (A and E) Western blot analysis assessed protein levels of RUNX2, ALP, COL1, p-GSK3β, and p-AKT after 14 days of osteogenic induction in control hPDLSCs, TB-II-treated cells (20 μM), and co-treated groups with TB-II combined with either LY294002 or CHIR99021. (B–D, F–K) Quantitative analysis via one-way ANOVA revealed statistically significant differences in relative protein expression of RUNX2 (F), ALP (G), COL1 (H), p-GSK3β (J), and p-AKT (K) among control, TB-II, TB-II + LY294002, and TB-II + CHIR99021 groups. Error bars represent mean ± SD, with significance denoted as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. All TB-II-containing groups were maintained at 20 μM concentration under identical experimental conditions

Fig. 7.

TB-II activated the PI3K/AKT/GSK3β signaling pathway to promote osteogenic differentiation in hPDLSCs. (A) Immunofluorescence detection of the expression of COL-1 for the control group, TB-II group (20 μM), TB-II (20 μM) with LY294002 group, TB-II (20 μM) with CHIR99021 group. TB-II (20 μM) with CHIR99021 group was observed with the strongest immunofluorescence intensity. Scale bar: 50 μm. (B) Quantitative analysis confirmed significantly elevated COL-1 expression in TB-II-treated cells (**P < 0.01 vs. control), which was further amplified by CHIR99021 (**P < 0.001 vs. TB-II alone) and suppressed by LY294002 (***P < 0.001 vs. TB-II alone). All data are presented as mean ± SD, with statistical comparisons performed via one-way ANOVA

OTM Model and Micro-CT Analysis

The OTM model was established in Wistar rats, as shown in Fig. 8A. Throughout the study, all animals remained in good health, and the TB-II treated group exhibited no significant body weight changes compared to the control group (Fig. 8B; Supplementary Table 2). The left maxillary first molar was mesially translated via 50 g force application over 14 days, resulting in reduced tooth displacement (Fig. 8C and D; Supplementary Table 3) in TB-II treated group (80 mg/kg/day) versus control group (saline) (P < 0.01). Concomitantly, TB-II administration slowed the rate of OTM progression during this period. Micro-CT analysis of the first molar’s distal buccal root revealed significant bone remodeling on the tension side. Specifically, TB-II treatment increased bone volume fraction (BV/TV) and trabecular thickness (Tb.Th) while decreasing trabecular spacing (Tb.Sp) compared to controls (P < 0.01 for BV/TV and Tb.Th; P < 0.05 for Tb.Sp) (Fig. 8E–H). On the pressure side, similar trends were observed for these parameters but did not reach statistical significance.

Fig. 8.

TB-II Inhibits Orthodontic Tooth Movement via Modulating Alveolar Bone Remodeling in a Rat Model. (A) A schematic diagram illustrates the experimental setup for induced mesial movement of the right maxillary first molar in rats. (B) Bodyweight measurements over 14 days showed no significant differences between control and TB-II-treated groups (10 mg/kg/day, *P > 0.05; one-way ANOVA), indicating no systemic toxicity (mean ± SD, n = 9). (C) Representative anatomical landmarks for mesial displacement quantification are depicted (1st molar = 1 st; 2nd molar = 2nd). (D) Quantitative analysis revealed that TB-II treatment significantly reduced mesial movement distance compared to controls (**P < 0.01; one-way ANOVA). (E) Regions of interest (ROIs) for bone analysis were defined in vertical and horizontal micro-CT views (red boxes) near the first molar. (F–H) Trabecular bone parameters on day 14 demonstrated that TB-II treatment altered alveolar remodeling: compared to controls, the TB-II group exhibited reduced bone volume/total volume (BV/TV) and trabecular thickness (Tb.Th) on the compression side (**P < 0.01), while increasing trabecular separation (Tb.Sp) on the tension side (*P < 0.05). All data are presented as mean ± SD (n = 9)

HE, Masson and TRAP Staining Observation

Histological analysis of maxillary first molar PDL architecture revealed distinct biomechanical adaptations under orthodontic force. HE staining demonstrated PDL narrowing on the compression side and widening on the tension side, with elongated spindle-shaped cells (Fig. 9A). Notably, TB-II-treated group exhibited superior PDL organization compared to the control group, displaying continuous and densely packed cellular arrangements. Masson's trichrome staining further highlighted enhanced extracellular matrix remodeling in TB-II group, showing more densely packed and orderly collagen fibers on both tension and compression sides relative to the control group (Fig. 9B). TRAP staining identified multinucleated osteoclasts with characteristic resorption pits on root and alveolar bone surfaces (Fig. 9C). Quantitative analysis revealed a statistically significant reduction (P < 0.05) in osteoclast numbers on the compressive root surface in TB-II-treated group compared to the control group (Fig. 9D), suggesting suppressed bone resorption activity.

Fig. 9.

Histological and Molecular Profiling of Orthodontic Tooth Movement in TB-II-Treated Rats. (A) HE staining revealed morphological changes in periodontal tissues on compression (control: a; TB-II: b) and tension (control: c; TB-II: d) sides after 14 days of orthodontic loading. (B) Masson’s trichrome staining highlighted collagen distribution in compression-side (a, b) and tension-side (c, d) tissues between control and TB-II-treated groups. (C) TRAP staining identified osteoclast-like cells (black arrows) on compression-side tissues, with reduced TRAP-positive cell numbers in the TB-II group compared to controls. (D) Quantification confirmed TB-II treatment significantly decreased TRAP-positive cell counts (mean ± SD, n = 9) (**P < 0.05, one-way ANOVA). (E–K) Immunohistochemistry demonstrated spatially distinct protein expression patterns: RUNX-2 (E and F), alkaline phosphatase (ALP) (G and H), and collagen-1 (COL-1) (I and J) were enriched on the tension side, while receptor activator of nuclear factor-κB ligand (RANKL) (K and L) predominated on the compression side. TB-II treatment suppressed RUNX-2 (**P < 0.01), ALP (**P < 0.01), and COL-1 (**P < 0.01) expression on the tension side, while RANKL expression was reduced on the compression side (*P < 0.05). Scale bars: 20 μm for all histological images. Statistical analyses used one-way ANOVA with mean ± SD (n = 9)

IHC staining Analysis

Immunohistochemical analysis (Fig. 9E–L) evaluated protein expression profiles of key osteogenic (RUNX-2, ALP, COL-1) and osteoclastogenic (RANKL) markers in the PDL. In TB-II group, tension-side PDL exhibited significantly elevated expression of RUNX-2 (P < 0.05), ALP (P < 0.01), and COL-1 (P < 0.01) compared with the control group, while compression-side RANKL expression was markedly suppressed (P < 0.01).

Discussion

During orthodontic treatment or after its completion, some patients may experience varying degrees of alveolar bone resorption. In severe cases, this can lead to periodontal attachment loss, tooth mobility, and even tooth loss, posing significant threats to oral health. Under specific inductive signaling conditions, stem cells can differentiate into various connective tissues, including bone, dentin, cementum, and adipose tissues, making them ideal seed cells for tissue and organ damage repair, particularly in the reconstruction of oral and maxillofacial tissues [28–30].

Although stem cell-based therapies hold great promise, they have not yet met clinical standards due to a range of significant challenges [31]. One of the primary hurdles is the limited ability of stem cells to promote bone formation in the local microenvironment, often caused by a lack of adequate mechanical and biochemical signals. When osteogenic differentiation at the transplant site is insufficient, it leads to poor bone formation, which negatively impacts the overall success of tissue regeneration. A crucial factor in achieving effective bone formation is the proper regulation of key signaling pathways, such as PI3K/AKT, Wnt, and RUNX-2 [32–34]. These pathways play a vital role in controlling essential biological processes, including cell proliferation, differentiation, and matrix mineralization. Therefore, activating these signaling mechanisms can significantly improve osteogenic differentiation and enhance therapeutic outcomes.

hPDLSCs are mesenchymal stem cells located in the periodontal ligament with self-renewal and multidirectional differentiation potential. They can differentiate into osteoblasts and cementoblasts to form bone and cementum-like structures, as well as into fibroblasts to generate structures resembling natural periodontal ligament [35, 36]. In this study, hPDLSCs were successfully isolated and cultured. Flow cytometry confirmed the absence of contamination by other cells, and in vitro induction demonstrated their adipogenic and osteogenic differentiation capabilities. Previous studies have shown that protein crotonylation can promote the osteogenic differentiation of hPDLSCs via the PI3K/AKT signaling pathway [37]. Research on regeneration and reconstruction has revealed that hPDLSCs exhibit high alkaline phosphatase activity and robust mineralization ability, making them an ideal type of seed cell within periodontal tissues [38].

Anemarrhena asphodeloides, a commonly used herb in traditional Chinese medicine, has recently garnered attention in this field. Among its key components, TB-II stands out as particularly significant. Previous studies have demonstrated that TB-II exhibits beneficial effects in treating conditions such as Alzheimer’s disease, diabetes, and osteoporosis [37–39]. Due to these promising properties, we selected TB-II as the focus of our investigation. Our study aimed to enhance the osteogenic potential of hPDLSCs by identifying an effective approach. We observed a significant upregulation of RUNX-2, ALP, and COL-1 expressions after 14 days of osteogenic induction under the use of TB-II. RUNX-2, a crucial regulator of osteogenesis, activates genes involved in osteoblast differentiation [40]. ALP, an essential early-stage marker of calcification, plays a vital role in osteoblast maturation [41]. Among the tested concentrations, 20 μM TB-II exhibited the highest ALP activity and the most intense blue ALP staining. COL-1, a key component of bone matrix and collagen production, is closely linked to osteogenesis and, when mutated, contributes to osteogenesis imperfecta (OI) [42]. Immunofluorescence analysis revealed a substantial increase in COL-1 expression in the 20 μM TB-II group. RUNX-2 regulates ALP and COL-1 by interacting with cis-acting elements such as OSE2 in the OC promoter region, highlighting its importance in skeletal development [43]. Additionally, calcium nodule formation, a key indicator of osteogenic differentiation and bone maturation, was significantly enhanced. These findings suggest that TB-II, particularly at a concentration of 20 μM, effectively promotes the osteogenic differentiation of hPDLSCs.

The PI3K/AKT pathway plays a crucial role in osteogenesis by promoting the proliferation, differentiation, and survival of osteoblasts [23]. Activation of PI3K/AKT signaling enhances osteogenic differentiation by upregulating key transcription factors like RUNX-2 and BMP-2, which drive bone formation [44, 45]. Additionally, the pathway interacts with other signaling cascades, such as Wnt and BMP, to synergistically enhance bone formation [46]. Dysregulation of PI3K/AKT signaling is associated with impaired osteogenesis and bone-related disorders, highlighting its importance in skeletal health. LY294002 is a potent, cell-permeable inhibitor of the PI3K/AKT pathway. It specifically targets the ATP-binding site of PI3K, blocking its activity and downstream signaling, including AKT activation [47]. LY294002 helps elucidate the roles of PI3K in cellular processes such as proliferation, survival, metabolism, and differentiation. It has been instrumental in studying osteogenesis [48].

CHIR-99021, a selective glycogen synthase kinase-3 (GSK-3) inhibitor, enhances osteogenesis through the Wnt/β-catenin signaling pathway [49]. GSK3β is one of the two isozymes of the GSK-3 family. Normally, GSK3β regulates osteoblast differentiation and bone formation by suppressing this pathway [50]. By inhibiting GSK3β, CHIR-99021 stabilizes β-catenin, upregulating osteogenic genes (RUNX-2, OSX) and promoting collagen synthesis and mineralization. In mesenchymal stem cells (MSCs), it favors osteoblast differentiation over adipogenesis or chondrogenesis. Preclinical studies in vitro and animal models show increased bone density and trabecular thickness. This compound highlights GSK3β inhibition as a viable strategy to augment bone regeneration.

In our study, 20 μM TB-II activated the PI3K/AKT pathway, as evidenced by elevated p-AKT levels, which was reversed by the PI3K inhibitor LY294002, confirming pathway dependency. Co-treatment with CHIR-99021 further enhanced p-GSK3β, indicating Wnt pathway activation via GSK3β inhibition, which was supported by upregulated p-GSK3β. The accumulated evidence underscores that TB-II promotes osteogenic maturation in human periodontal ligament stem cells by modulating the PI3K/AKT/GSK3β signaling continuum, thereby positioning this compound as a dual-action regulator capable of synchronizing critical signaling networks to enhance osseous regeneration outcomes.

We investigated the effects of TB-II on the adaptation of periodontal tissue in living systems through in vivo experiments. Our micro-CT analysis revealed that the tension side exhibited a higher bone volume fraction (BV/TV) and trabecular thickness (Tb.Th), along with a lower trabecular spacing (Tb.Sp) compared to the control groups, indicating an accelerated bone production process. Additionally, immunohistochemical analysis of COL-1 and RUNX-2 supported TB-II's role in promoting bone formation on the tension side. While we did not find significant overall statistical differences between the two groups, we did observe notable increases in BV/TV and Tb.Th, along with a reduction in Tb.Sp on the compression side. HE and Masson staining both showed a better cellular arrangement in the TB-II group. TRAP staining showed a decrease in osteoclast numbers on the compression side treated with TB-II, suggesting a reduction in bone resorption. Furthermore, TB-II was found to lower RANKL expression in the compressed root areas, indicating that its effects may be linked to the inhibition of osteoclast development through the suppression of the RANKL pathway and related gene activity. These findings suggest that TB-II enhances bone formation on the tension sides while simultaneously limiting resorption on the compression sides under mechanical strain, resulting in smaller tooth movements. This balanced dual action on bone metabolism provides potential strategies to prevent orthodontic relapses and improve methods for maintaining corrected dental alignment.

Our study has several limitations that warrant acknowledgment. First, the relatively short experimental duration for hPDLSC culture precludes insights into the long-term biological impacts and potential adverse effects of TB-II. Additionally, the reliance on an animal model limits our ability to extrapolate findings to human applications, necessitating further investigations to establish optimal dosing regimens and clinical safety profiles. Future research directions will prioritize addressing these gaps through extended longitudinal studies and focus on elucidating TB-II’s mechanistic influence on osteoclast genesis. All the things above are crucial for ensuring its safe and effective integration into clinical therapies targeting bone remodeling disorders.

Conclusion

Our study demonstrates that TB-II stimulates proliferation and osteogenic maturation of hPDLSCs in vitro through activation of the PI3K/AKT/GSK3β signaling pathway. Furthermore, in vivo experiments using OTM model revealed that TB-II enhances periodontal tissue remodeling, thereby reducing post-OTM relapse. These findings establish TB-II as a promising candidate for mitigating relapses following orthodontic treatment.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

TB-II

Timosaponin B- II

ANOVA

Analysis of variance

BV/TV

Bone volume/total volume

CPC

Cetylpyridinium chloride

DAPI

4,6-Diamidino-2-phenylindole

EDTA

Ethylenediaminetetraacetic acid

FBS

Fetal bovine serum

GAPDH

Glyceraldehyde-3-phosphate dehydrogenase

HE

Hematoxylin and eosin

IHC

Immunohistochemical

micro-CT

Micro-computed tomography

OTM

Orthodontic tooth movement

PDL

Periodontal ligament

PI3K

Phosphatidylinositol-3-kinase

RUNX-2

Runx family transcription factor 2

SPF

Specific pathogen-free

TBST

Tris-buffered saline containing 0.1% Tween 20

TRAP

Tartaric-resistant acid phosphatase

α-MEM

α-Minimum essential medium

AOD

Average optical density

ALP

Alkaline phosphatase

AKT

Protein kinase B

p-AKT

Phosphoprotein kinase B

GSK3β

Glycogen synthase kinase - 3β

p-GSK3β

Phosphorylated Glycogen Synthase Kinase 3β

qRT-PCR

Quantitative reverse transcription polymerase chain reaction

COL-1

Collagen type 1

CCK-8

Cell-counting Kit-8

ARS

Alizarin Red Staining

RANKL

Nuclear factor-κB (NF-κB) ligand

ROI

Region of interest

Tb.Th

Trabecular thickness

Tb.Sp

Trabecular separation/spacing

OD

Optical density

BCA

Bicinchoninic acid

BSA

Bovine serum albumin

hPDLSCs

Human periodontal ligament stem cells

PBS

Phosphate-buffered saline

Author Contributions

DWY conducted the experiments and wrote the manuscript. JZ conceived and designed the manuscript, and approved the final manuscript. XBC conceived and drafted the manuscript. RD coordinated the execution of the research, participated in the experimental work and revised the manuscript. ZYW responsible for collection and assembly of data and performed compound preparation. YXZ participated in the experiments. PY analyzed the data. All authors read and approved the final manuscript.

Funding

This work was supported by the Province Natural Science Foundation of Shandong Province, grant number ZR2021QH340.

Data Availability

The data generated in this study are available from the corresponding author upon request.

Declarations

Ethics Approval and Consent to Participate

A series of procedures for animal experiments approved by Shandong University’s Institutional Animal Care and Use Committee (Approval No. 20230702).

Consent to Publication

Not applicable.

Competing interests

The authors declare no competing interests.