Oleuropein attenuates steroid-induced osteonecrosis of the femoral head by inhibiting osteoblast apoptosis via activation of the PI3K-AKT-Bcl2 pathway Authors: Jianlong Lin, Rui Zhang, Xingyu Lin, Xiangjun Luo, Haokun Lin, Jie Xu, Shishui Lin Published: 2026 ABSTRACT Abstractteroid-induced osteonecrosis of the femoral head (SONFH) is a severe bone disorder caused by long-term glucocorticoid administration and is characterized by osteoblast apoptosis. Oleuropein (OLP), a natural compound with anti-inflammatory and antioxidant properties, has demonstrated anti-apoptotic potential in bone-related diseases. However, its therapeutic role in SONFH has not yet been elucidated. This study aimed to investigate the therapeutic effects of OLP on SONFH and elucidate its underlying molecular mechanisms. In vitro, MC3T3-E1 osteoblasts treated with methylprednisolone (MPS) were co-incubated with OLP. Cell viability was assessed using a CCK-8 assay and live/dead cell staining. In vivo, a rat SONFH model was established with lipopolysaccharide and MPS, followed by OLP treatment. Bone microstructure was analyzed by micro-computed tomography and histopathological staining (H&E, Masson, Goldner). Network pharmacology and proteomics analyses were used to identify key targets and pathways related to the effects of OLP on SONFH. Apoptosis was examined with flow cytometry, TUNEL staining and ELISA. Protein and mRNA expression levels of relevant targets and pathways were examined with western blotting and quantitative real-time polymerase chain reaction. OLP significantly reversed MPS-induced osteoblast apoptosis and enhanced cell viability. In SONFH rats, OLP reduced empty lacunae, restored trabecular bone structure, and improved collagen organization. Molecular docking confirmed OLP’s binding to AKT and Bcl2. Network pharmacology and proteomics highlighted apoptosis regulation, PI3K-AKT signaling, and cell cycle control as key mechanisms. Furthermore, OLP activated the PI3K-AKT-Bcl2 pathway, increasing p-PI3K, p-AKT, and Bcl2 levels while decreasing Caspase3 and Caspase9. OLP also promoted osteogenesis (upregulated ALP and Runx2) and angiogenesis (increased vWF and CD31). OLP alleviates SONFH by inhibiting osteoblast apoptosis and promoting osteogenesis and angiogenesis through activation of the PI3K-AKT-Bcl2 pathway. These findings support OLP as a promising natural compound for early intervention in SONFH.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-48653-6.Keywords: Steroid-induced osteonecrosis of the femoral head, Oleuropein, Apoptosis, Molecular mechanismsSubject terms: Cell biology, Diseases, Drug discovery, Medical research, Molecular biology Introduction Glucocorticoids (GCs) are widely applied in the clinical management of autoimmune diseases, allergic reactions, and diverse inflammatory disorders because of their potent anti-inflammatory and immunosuppressive effects1,2. However, prolonged or high-dose administration of GCs is strongly associated with the onset of steroid-induced osteonecrosis of the femoral head (SONFH), a debilitating condition characterized by progressive degeneration of the hip joint. In recent years, the incidence of SONFH has steadily increased, making it the leading cause of non-traumatic osteonecrosis of the femoral head. Without timely and effective intervention, more than 80% of patients with SONFH develop femoral head collapse within three years, resulting in severe disability and posing a major challenge in orthopedic practice3,4. Therefore, early prevention and timely intervention are essential to reduce long-term functional impairment. The pathological features of SONFH include structural deterioration of the femoral head, collapse of the articular surface, and progressive joint dysfunction5. Its pathogenesis is multifactorial, involving dysregulated lipid metabolism, impaired microcirculation, and enhanced osteoblast apoptosis, the latter being recognized as a key contributor to disease progression6. Current clinical management primarily focuses on symptomatic relief and late-stage surgical approaches, such as total hip arthroplasty. However, these treatments fail to address the underlying pathological mechanisms and are associated with substantial risks of perioperative complications7. These limitations highlight the urgent need for early interventions and targeted therapies that can modulate the local microenvironment and specific molecular pathways involved in SONFH. In recent years, natural compounds have attracted increasing attention in the prevention and treatment of orthopedic diseases due to their multitarget mechanisms and favorable biocompatibility8. Evidence suggests that several plant-derived bioactive compounds, such as echinacoside and gastrodin, exert protective effects on damaged cells by modulating apoptosis-related signaling pathways, offering promising therapeutic strategies for steroid-induced bone injury9,10. Oleuropein (OLP), a naturally occurring phenylethanoid glycoside abundant in olive leaves and extra virgin olive oil, is well known for its anti-inflammatory, antioxidant, and anticancer properties11. Recent studies have reported that OLP exhibits therapeutic potential in animal models of osteoporosis and may protect osteoblasts from apoptosis12. In addition to its anti-osteoporotic effects, OLP has also been shown to reduce alveolar bone loss in periodontal disease by suppressing osteoclast activity and inflammation while enhancing osteoblast function, thereby further supporting its potential in maintaining bone homeostasis13. Preliminary evidence also indicates that OLP can modulate the PI3K-AKT signaling pathway in the context of cancer and cardiovascular diseases14. Nevertheless, the specific mechanisms by which OLP mitigates GC-induced osteoblast apoptosis and prevents SONFH remain to be elucidated. In this study, we first evaluated the cytocompatibility of OLP and its protective effects against methylprednisolone (MPS)-induced cytotoxicity in MC3T3-E1 osteoblasts. After confirming its favorable biocompatibility and ability to restore cellular injury in vitro, we established a SONFH rat model followed by OLP treatment. Bone microstructure was analyzed using radiological and histopathological assessments. Micro-CT, H&E, Masson, and Goldner staining demonstrated that OLP reduced empty lacunae, restored trabecular structure, and improved collagen organization. As a result, OLP exhibited significant therapeutic effects in SONFH rats along with favorable in vivo biocompatibility. Furthermore, network pharmacology and proteomics analyses were employed to identify potential targets and pathways of OLP in SONFH, and molecular docking confirmed its binding affinity with AKT and Bcl2. These analyses highlighted apoptosis regulation, PI3K-AKT signaling, and cell cycle control as key mechanisms. Consistently, flow cytometry and TUNEL staining showed that OLP reduced MPS-induced osteoblast apoptosis, while western blotting and qPCR further revealed that OLP attenuates GC-induced apoptosis by activating the PI3K-AKT-Bcl2 signaling pathway (Scheme 1). Collectively, these findings provide a solid theoretical foundation for the development of natural compound-based targeted therapies for early intervention and treatment of SONFH. Scheme 1 Schematic illustration of therapeutic mechanism of OLP against SONFH. Materials and methods Cell culture and cell viability assay Mouse osteoblastic MC3T3-E1 cell lines were obtained from the National Collection of Authenticated Cell Cultures (China) and cultured in α-MEM medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS). Cells were maintained at 37 °C in a humidified incubator with 5% CO₂. Cell viability following methylprednisolone (MPS; MedChemExpress, MCE, USA) treatment was assessed using the Cell Counting Kit-8 (CCK-8; Beyotime, China). MC3T3-E1 cells were seeded in 96-well plates at a density of 5.0 × 10³ cells per well and incubated with different concentrations of MPS (1, 10, 100, 200, 300, 400, 500, 1000 µM) for 24 h before performing the CCK-8 assay. Similarly, cells were exposed to increasing concentrations of OLP (0, 1, 10, 20, 30, 40, 50, 100 µg/mL) for 24 h. To evaluate the protective effect of OLP against MPS, cells were co-treated with 400 µM MPS and different concentrations of OLP (1, 10, 20, 30, 40, 50, 100 µg/mL) for 24 h. Live/dead cell staining The effect of OLP on osteoblast viability under MPS-induced cytotoxicity was assessed by live/dead cell staining. After co-incubation with 400 µM MPS and different concentrations of OLP (1, 10, 20, 30, 40, 50, 100 µg/mL) for 24 h, the culture medium was aspirated, and cells were washed once with phosphate-buffered saline (PBS). Staining was performed according to the manufacturer’s protocol using the Calcein/PI Cell Viability and Cytotoxicity Assay Kit (Beyotime, China). Stained cells were observed and imaged with a Leica DMIRB inverted fluorescence microscope (Leica, USA). Detection of apoptosis levels Annexin V-FITC assay by flow cytometry MC3T3-E1 cells were seeded in 6-well plates at a density of 2.0 × 10⁵ cells per well and incubated for 24 h. Cells were then divided into four groups: (a) control group (Ctrl), cultured in α-MEM with 10% FBS; (b) MPS group, treated with 400 µM MPS to induce glucocorticoid-mediated apoptosis; (c) low-dose OLP group (OLPL), co-incubated with 400 µM MPS and 10 µg/mL OLP; and (d) high-dose OLP group (OLPH), co-incubated with 400 µM MPS and 40 µg/mL OLP. After 24 h, cells were harvested, washed twice with PBS, and stained using the Annexin V-FITC Apoptosis Detection Kit (Beyotime, China). Apoptosis was quantified by flow cytometry. TUNEL assay MC3T3-E1 cells were seeded in 96-well plates at a density of 5.0 × 10³ cells per well and incubated for 24 h. Cells were then treated with 400 µM MPS for 24 h to induce apoptosis, followed by incubation with α-MEM containing OLP (10–40 µg/mL) for another 24 h. After washing twice with PBS, cells were stained using the One-Step TUNEL Apoptosis Assay Kit (Elabscience, China). Images were acquired using an inverted fluorescence microscope, and the percentage of TUNEL-positive cells was quantified with ImageJ software (NIH, USA). Enzyme-linked immunosorbent assay (ELISA) MC3T3-E1 cells were seeded in 6-well plates and cultured under standard conditions. Following exposure to 400 µM MPS for 24 h, cells were treated with OLP (10–40 µg/mL) for another 24 h. Culture supernatants were collected and centrifuged at 1000 × g for 10 min at 4 °C to remove debris. The concentrations of Bcl-2, Bax, Caspase-3, and Caspase-9 were measured using commercial ELISA kits (Elabscience, China), according to the manufacturer’s instructions. Absorbance was read at 450 nm with a microplate reader (Thermo Fisher Scientific, USA), and protein levels were calculated from standard curves. Western blot analysis Cells were washed 2–3 times with PBS, lysed with RIPA buffer (Santa Cruz Biotech, USA) on ice for 30 min, and centrifuged at 12,000 rpm for 10 min at 4 °C. Supernatants were collected, mixed with 5× loading buffer (4:1), and denatured by boiling for 15 min. Proteins were separated by SDS-PAGE and transferred to PVDF membranes (0.45 μm). After blocking with 5% skim milk in TBST for 30 min, membranes were incubated overnight at 4 °C with primary antibodies (1:1000, per manufacturer’s instructions). After washing, membranes were incubated with goat secondary antibodies (1:3000 in TBST) for 30 min at room temperature. Protein bands were visualized using ECL reagents (Thermo Fisher Scientific, USA) and imaged with a chemiluminescence system. Images were analyzed using AIWBwell software. Antibodies against PI3K, AKT, p-PI3K, p-AKT, Bcl-2, ALP, Runx2, vWF, CD31, Caspase-3, and Caspase-9 were purchased from Servicebio (China). Quantitative real-time PCR (qPCR) Total RNA was extracted using RNA lysis buffer (Servicebio, China), mixed with a chloroform substitute, and centrifuged at 12,000 rpm for 10 min at 4 °C. The aqueous phase was collected, precipitated with isopropanol at − 20 °C for 15 min, washed twice with 75% ethanol, air-dried, and dissolved in nuclease-free water. RNA concentration and purity were assessed using a Nanodrop 2000 spectrophotometer (Servicebio, China). cDNA was synthesized from 200 ng total RNA in a 20 µL reaction following the manufacturer’s protocol. qPCR was performed, and gene expression was analyzed using the 2^−ΔΔCT method. Primer sequences for PI3K, AKT, Bcl-2, Bax, Caspase-3, Caspase-9, vWF, CD31, ALP, and Runx2 are listed in Table S1. Network pharmacology analysis Potential OLP targets were predicted using SwissTargetPrediction, BATMAN-TCM, and SEA (SuPred) databases. Disease-related targets for “steroid-induced osteonecrosis of the femoral head” were retrieved from GeneCards, DisGeNET, and OMIM databases. Overlapping targets were identified using Venny 2.1.0. A protein–protein interaction (PPI) network was generated with STRING and analyzed using Cytoscape (v3.9.0). A “drug–core targets–disease” network was constructed. Gene Ontology (GO) and KEGG pathway enrichment analyses were conducted using DAVID, with visualization on www.bioinformatics.com.cn. Molecular docking was performed with AutoDock Vina, and results were visualized using PyMOL. Proteomics study Sample preparation MC3T3-E1 cells from the SONFH model group (400 µM MPS, 24 h) and OLP treatment group (400 µM MPS + 40 µg/mL OLP, 24 h) were ground in liquid nitrogen. Protein extraction, trypsin digestion, and peptide purification were performed using standard protocols. Peptides were labeled with Tandem Mass Tag (TMT) reagents, fractionated by reverse-phase HPLC (Agilent 300 Extend C18, 4.6 × 250 mm, 5 μm), vacuum-concentrated, and freeze-dried for analysis. Data analysis Protein identification quality was verified, and differential protein expression analysis was performed, including (1) global expression analysis, (2) identification of differentially expressed proteins (DEPs), and (3) clustering of DEP expression patterns. Functional analyses included GO enrichment, subcellular localization prediction, Clusters of Orthologous Groups classification, and KEGG pathway enrichment15–17. Animal experiments All procedures were approved by the Animal Ethics Committee of Fujian Medical University (IACUC FJMU 2024 − 0194). All methods were carried out in accordance with Helsinki and ARRIVE guidelines and regulations. Twenty-four male Sprague–Dawley (SD) rats were housed under controlled conditions (25 ± 1 °C, 55 ± 5% humidity, 12 h light/dark cycle). Rats were randomly divided into four groups (n = 6 each): (a) control (0.9% saline, intraperitoneal injection for 1 week); (b) MPS group (LPS 10 µg/kg + MPS 20 mg/kg, intraperitoneal and intramuscular injections for 6 weeks); (c) low-dose OLP group (MPS+OLPL, 20 mg/kg OLP via intraperitoneal and gluteal intramuscular injection); and (d) high-dose OLP group (MPS+OLPH, 80 mg/kg OLP via the same route). After treatment, rats were anesthetized with 2% sodium pentobarbital (40 mg/kg, intraperitoneal) and euthanized by inhalation of excessive carbon dioxide, and femoral heads were collected. H&E, masson, and goldner staining Rat femoral heads were fixed in formaldehyde for 48 h, decalcified in EDTA for 4 weeks, dehydrated, paraffin-embedded, and sectioned (5 μm). Hematoxylin and eosin (H&E) staining was performed to assess trabecular bone and empty lacunae. Masson’s trichrome staining evaluated collagen fibers and fibrosis, while Goldner’s trichrome differentiated mineralized bone (green) from osteoid (red). Immunohistochemical Staining Paraffin sections were rehydrated, subjected to pepsin antigen retrieval, and permeabilized with 0.5% Triton X-100. After blocking with 5% bovine serum albumin (BSA), sections were incubated overnight at 4 °C with antibodies against p-PI3K, p-AKT, Bcl2, Caspase-3, Caspase-9, ALP, Runx2, vWF, and CD31. Following washing, fluorescently labeled secondary antibodies were applied, and sections were mounted and visualized by fluorescence microscopy. Micro-computed tomography (micro-CT) Femoral heads were harvested and scanned using micro-CT for the evaluation of bone microarchitecture. For quantitative analysis, ROI was strictly defined within the epiphyseal and subchondral trabecular bone of the femoral head, which represents the primary pathological site of SONFH. The ROI was selected by identifying the subchondral bone plate as an anatomical landmark and including the underlying trabecular bone within the epiphysis, while excluding the metaphyseal region to avoid confounding effects from non-lesion areas. Bone microarchitectural parameters, including bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), and trabecular number (Tb.N), were quantified using the system software. Statistical analysis and software All experiments were independently repeated at least three times. Data are expressed as mean ± standard deviation (SD). Statistical analyses were performed using SPSS software (IBM Corp., Armonk, NY, USA; version 18.0; https://www.ibm.com/products/spss-statistics). Intergroup differences were evaluated by one-way ANOVA, with p < 0.05 considered statistically significant. Image analysis was performed using ImageJ (National Institutes of Health, Bethesda, MD, USA; version 1.53; https://imagej.net/software/imagej/). Western blot band quantification was conducted using AIWBwell software (AIWBwell Inc., China; version 1.0). Network pharmacology analysis was performed using Cytoscape (version 3.9.0; https://cytoscape.org/). Results Cytocompatibility and preliminary therapeutic evaluation in vitro The CCK-8 assay revealed that the cell viability of MC3T3-E1 was significantly decreased following the increasing doses of methylprednisolone (MPS), with an estimated half maximal inhibitory concentration (IC50) of approximately 400 µM (P < 0.05) (Fig. 1A). Based on this result, we selected 400 µM concentration of MPS as the in vitro model-establishing method for subsequent experiments. Meanwhile, there was no apparent cytotoxicity when MC3T3-E1 cells were treated with series concentrations of OLP for 24 h, indicating favorable biocompatibility of OLP in vitro (Fig. 1B). Cytocompatibility of OLP and its protective effects against MPS-induced cytotoxicity in MC3T3-E1 cells. (A) Cell viability of MC3T3-E1 cells exposed to increasing concentrations of MPS (0–1000 µM) for 24 h (mean ± SD, n = 6). (B) Cell viability of MC3T3-E1 cells treated with various concentrations of OLP (0–100 µg/mL) for 24 h (mean ± SD, n = 6). (C) Effects of OLP (10–100 µg/mL) on cell viability in MC3T3-E1 cells exposed to 400 µM MPS for 24 h (mean ± SD, n = 6). (D) Representative live/dead staining images of MC3T3-E1 cells treated with OLP (10–100 µg/mL) under normal conditions and MPS (400 µM)-induced apoptosis for 24 h (scale bar = 200 μm). (E) Quantitative analysis of live/dead staining results (mean ± SD, n = 3). Statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Notably, OLP effectively reversed MPS-induced cytotoxicity in a concentration-dependent manner, suggesting that OLP may exert a protective effect through specific underlying mechanisms (Fig. 1C). Among the tested concentrations, 40 µg/mL OLP demonstrated maximal efficacy in restoring cell viability, with no significant additional benefit compared to the higher concentrations (50–100 µg/mL). Therefore, 10 µg/mL and 40 µg/mL concentrations of OLP were selected as the low-dose and high-dose groups for subsequent investigations, respectively. To further assess the protective effects of OLP against MPS-induced cellular damage, we performed cell live/dead staining. When MC3T3-E1 cells were exposed to MPS, a significant increase of dead cell proportion was observed. However, OLP treatment significantly reduced the proportion of dead cells in a concentration-dependent manner, while the percentage of live cells correspondingly increased (Fig. 1D-E). These findings were consistent with the results of CCK-8, confirming that OLP not only demonstrates preferable biocompatibility in vitro but also effectively attenuates MPS-induced osteoblast injury. Evaluation of the therapeutic effects of OLP on SONFH animal models Then, we investigated the potential therapeutic effects of OLP through SONFH rat model. Micro-CT was used to assess trabecular microarchitectural changes in the femoral head. Compared to the control group (Ctrl), rats in the model group receiving MPS showed significant reductions in bone mineral density (BMD), bone volume fraction (BV/TV), and trabecular number (Tb.N) (p < 0.05), alongside a significant increase in trabecular separation (Tb.Sp) (p < 0.05) (Fig. 2A-F). These findings confirm successful induction of glucocorticoid-induced trabecular structural deterioration. In contrast, OLP-treated rats (OLPL and OLPH) demonstrated significant restoration of BMD, BV/TV, and Tb.N (p < 0.05), concurrently with a marked reduction in Tb.Sp (p < 0.05), compared to the MPS group (Fig. 2A-F). These micro-CT quantified results demonstrate that OLP effectively counteracted trabecular degeneration and maintained bone microarchitecture. Therapeutic effects of OLP on bone microarchitecture and histopathology in SONFH rat models. (A) Representative micro-CT images of femoral heads in coronal and transverse planes of the epiphyseal region. (scale bar = 1 mm, n = 6). (B) BV/TV. (C) Tb.N. (D) Tb.Sp. (E) Tb.Th. (F) BMD. (G) Representative H&E staining of femoral heads. (scale bar = 200 μm n = 6). (H) Representative Goldner’s trichrome staining of femoral heads (scale bar = 200 μm n = 6). (I) Representative Masson’s trichrome staining of femoral heads (scale bar = 200 μm n = 6). Histopathological analysis using hematoxylin and eosin (H&E) staining further confirmed the above findings in the epiphyseal region of the femoral head. In the Ctrl group, trabeculae were densely arranged, continuous, and structurally intact, with well-defined bone architecture and evenly distributed marrow cavities, and no obvious pathological alterations were observed. In contrast, the MPS group exhibited typical pathological features, including a marked increase in empty osteocyte lacunae within the trabeculae, extensive necrosis of marrow tissue, fractured and disorganized trabeculae, and reduced bone mass, indicating that glucocorticoid intervention compromised the structural integrity of bone tissue. Notably, OLP treatment markedly alleviated these degenerative changes, as evidenced by a reduction in empty lacunae, restoration of trabecular continuity and organization, and increased bone mass (Fig. 2G). Goldner’s trichrome staining provided complementary evidence supporting these findings in the epiphyseal region of the femoral head. Femoral head samples from OLP-treated rats exhibited improved restoration of mineralized bone continuity, markedly reduced matrix fragmentation, and an increased proportion of mineralized bone surface compared to the MPS group. Collectively, these results suggest that OLP not only preserves bone structure but also facilitates matrix mineralization and promotes reparative remodeling in glucocorticoid-induced osteonecrotic lesions (Fig. 2H). Masson staining of the epiphyseal region revealed extensive disruption, fragmentation, and disorganization of collagen fibers in the MPS group, indicative of severe matrix degradation within necrotic regions. Administration of OLP significantly ameliorated collagen fiber organization, reduced fragmentation, and effectively restored fiber continuity, demonstrating enhanced regeneration of the collagen matrix and improved structural integrity (Fig. 2I). Further, we evaluated the biocompatibility of OLP in vivo. The pathological sections stained by H&E showed that the tissue structure of the main organ in the various intervention groups were not significantly different from that in the Ctrl group (Fig. S1). Overall, these results demonstrated that OLP exhibited favorable biocompatibility as well as markedly mitigated glucocorticoid-induced structure deterioration, reducing empty osteocyte lacunae, restoring collagen fiber continuity, and improving matrix mineralization in vivo. Network pharmacology analysis As shown in the Fig. 3A-B, the “drug-target-disease” network illustrates that 123 potential protein targets of OLP are associated with SONFH. PPI analysis was conducted on these targets, and the visualization revealed 82 nodes and 594 edges, indicating the complex interactions between OLP and its associated proteins. In this network, each node represents a gene-encoded protein, the edges represent the interactions between the connected proteins. The top five core targets identified based on degree ranking were TNF, AKT1, EGFR, CASP3, and BCL2, suggesting that these may serve as potential therapeutic targets of OLP against SONFH (Fig. 3C). Network pharmacology and molecular docking analysis of OLP against SONFH. (A) Venn diagram showing the intersection of OLP- and SONFH-associated targets. (B) Drug–target–disease network diagram illustrating the interactions between OLP and SONFH targets. (C) Protein–protein interaction (PPI) network and identification of core targets. (D) Gene Ontology (GO) enrichment analysis of the identified targets (bubble plot). (E) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the identified targets (bubble plot). (F) Molecular docking of OLP with the five core target proteins (TNF, AKT1, EGFR, CASP3, and BCL2), showing the most stable conformations with the strongest binding affinities. Subsequently, the 123 key targets were imported into the DAVID database for Gene Ontology (GO) enrichment analysis. As shown in the Fig. 3D, OLP’s protective effect against SONFH primarily involves biological processes (BP) such as negative regulation of apoptotic processes, inflammatory response, response to external stimulation, and positive regulation of smooth muscle cell proliferation. In terms of cellular components (CC), the targets are enriched in the plasma membrane, extracellular exosomes, receptor complexes, neuronal cell bodies, and endoplasmic reticulum membrane. The molecular functions (MF) include enzyme binding, G protein-coupled adenosine receptor activity, endopeptidase activity, protein binding, zinc ion binding, protein homodimerization activity, metalloendopeptidase activity, and nuclear receptor activity. Further Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed that the main pathways involved in OLP’s therapeutic effects including apoptosis, pathways in cancer, lipid and atherosclerosis, and the neurotrophin signaling pathway, indicating a multi-targeted and multi-pathway mechanism of action (Fig. 3E). Finally, molecular docking analysis was performed using AutoDock Vina, in which OLP was used as the ligand and the five core proteins (TNF, AKT1, EGFR, CASP3, and BCL2) were treated as receptors. The docking results were visualized using Pymol, which demonstrated that OLP could form stable interactions with all five core target proteins, supporting its potential for further in-depth mechanistic investigation (Fig. 3F). Proteomics data analysis Proteomic screening identified 471 differentially expressed proteins (DEPs) between OLP treatment group and SONFH model group across six biological replicates, with the OLP group exhibiting 145 upregulated and 326 downregulated proteins relative to controls (Fig. 4A). Notably, OLP intervention upregulated anti-apoptotic, anti-inflammatory, and antioxidant genes (Gpnmb, Tsc22d3, Gclc, Npc2) while downregulating Nr3c1 and Ptgs2. Mechanistic analyses revealed that Gpnmb activates intracellular pro-survival signaling pathways to mediate anti-apoptotic effects. Tsc22d3 (glucocorticoid-induced leucine zipper) modulates cell cycle progression, AKT signaling, and NF-κB pathways to inhibit apoptosis. Gclc (glutamate-cysteine ligase catalytic subunit) enhances glutathione synthesis for reactive oxygen species (ROS) scavenging while suppressing inflammatory related signaling pathways (NF-κB, Nrf2), thereby conferring anti-inflammatory and antioxidant protection. Npc2 maintains cellular homeostasis through regulation of cholesterol transport, apoptosis, and inflammatory responses. Notably, reduced Nr3c1 expression correlates with PI3K-AKT-mTOR pathway activation in colon cancer. Collectively, these DEPs alterations demonstrate that OLP mitigates glucocorticoid-induced osteoblast apoptosis through coordinated modulation of cell cycle regulation, apoptotic signaling, inflammatory responses, and oxidative stress pathways, with subcellular localization analysis revealing predominant intervention sites within extracellular, cytoplasmic, and nuclear compartments (Fig. 4B). Proteomic analysis and functional enrichment of differentially expressed proteins (DEPs) following OLP intervention in SONFH rats. (A) Volcano plot of DEPs between OLP treatment and model groups. (B) Heatmap of DEPs across biological replicates. (C) Gene Ontology (GO) enrichment analysis of DEPs, with the top 10 significantly enriched terms for biological processes (BP), cellular components (CC), and molecular functions (MF) shown as bar charts. (D) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of DEPs, highlighting significantly enriched pathways related to cell cycle regulation, PI3K–AKT signaling, and apoptosis. We further performed GO enrichment analysis. The top 10 significantly enriched terms from each category—CC, MF, and BP—were visualized in bar charts (Fig. 4C). GO analysis revealed that DEPs were primarily localized to extracellular regions and nuclei, participating in cellular proliferation, cell cycle regulation, and cellular growth. KEGG pathway enrichment analysis quantified enrichment significance using Rich Factor, p-value (range: 0–1), and gene counts per pathway. Bubble plot visualization revealed pronounced enrichment of OLP-modulated DEPs in cell cycle regulation, PI3K-AKT signaling, and DNA replication pathways, with notable representation in cellular senescence processes. Complementary network pharmacology findings collectively demonstrate that OLP exerts therapeutic effects against SONFH primarily through apoptotic signaling, PI3K-AKT pathway, and cell cycle control mechanisms (Fig. 4D). In Vitro evaluation of the anti-apoptotic effects of OLP To investigate the anti-apoptotic effects of OLP, apoptosis levels in MC3T3-E1 cells were assessed via flow cytometry. MPS treatment significantly elevated the apoptosis rate to 56.6 ± 0.88%, compared to 33.23 ± 1.01% in the control group (p < 0.05), demonstrating its potent pro-apoptotic effect. Notably, OLP intervention dose-dependently reduced the apoptosis rate, decreasing to 41.02 ± 0.64% and 37.44 ± 0.47% in the low- dose and high-dose OLP groups, respectively (p < 0.05) (Fig. 5A-B). The in vivo anti-apoptotic effect of oleuropein (A, B) Flow cytometry analysis of apoptosis and quantification of apoptotic rates under OLP treatment at different concentrations (n = 3). (C, D) TUNEL staining of apoptosis and quantitative analysis of TUNEL-positive cells (scale bar = 200 μm, n = 3). (E–H) Protein expression levels of Bcl-2, Bax, Caspase-3, and Caspase-9 assessed by ELISA (n = 3). (I–L) mRNA expression levels of Bcl-2, Bax, Caspase-3, and Caspase-9 assessed by RT-qPCR (n = 3). Statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. To further confirm these findings, Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) fluorescence staining was employed to evaluating apoptosis level in different groups. MPS intervention induced a marked increase in red fluorescence intensity compared to the control group, reflecting a significantly higher apoptosis level. Conversely, OLP treatment dose-dependently attenuated red fluorescence intensity, corresponding to a reduction in TUNEL-positive cells. Notably, in the high-dose OLP group, the number of TUNEL-positive cells was comparable to that of the control group, demonstrating that high-dose OLP effectively restored apoptosis levels to baseline (Fig. 5C-D). To further confirm the anti-apoptotic effects of OLP, ELISA and qPCR were performed to analyze key apoptosis-related markers, including Bcl2, Bax, Caspase-3, and Caspase-9. OLP treatment dose-dependently increased the anti-apoptotic protein Bcl2 expression, suppressed the expression of pro-apoptotic protein Bax, and downregulated the expression of Caspase-3 and Caspase-9 protein (p < 0.05) (Fig. 5E-H). Consistent with these protein-level findings, qPCR analysis revealed same trends in the mRNA expression of these markers (Fig. 5I-L). All these results demonstrated that OLP effectively mitigates MPS-induced osteoblast apoptosis. Mechanistic validation via western blotting, qPCR, and In Vivo experiments To elucidate the molecular mechanisms underlying the protective effects of OLP, we performed Western blot analysis on key proteins encompassing the PI3K-AKT signaling pathway (phosphorylated PI3K [p-PI3K], total PI3K, phosphorylated AKT [p-AKT], total AKT), anti-apoptotic signaling (Bcl2), pro-apoptotic markers (cleaved Caspase-3, cleaved Caspase-9), osteogenic markers (alkaline phosphatase [ALP], runt-related transcription factor 2 [Runx2]), and angiogenic indicators (von Willebrand factor [vWF], platelet endothelial cell adhesion molecule-1 [CD31]). Compared to the control group, MPS intervention significantly suppressed PI3K-AKT pathway activation in MC3T3-E1, evidenced by markedly reduced phosphorylation levels of PI3K and AKT (p < 0.05). Concurrently, MPS downregulated Bcl2, ALP, and Runx2 protein expression (p < 0.05), while upregulating cleaved Caspase-3 and cleaved Caspase-9 (p < 0.05), indicating enhanced apoptotic signaling and impaired osteogenesis (Fig. 6A-D). Mechanistic validation of OLP effects via the PI3K–AKT signaling pathway. (A) Western blot analysis of phosphorylated PI3K (p-PI3K), total PI3K, phosphorylated AKT (p-AKT), and total AKT under interventions with OLP, PI3K inhibitor (LY294002), and PI3K activator (740 Y-P). (B) Western blot analysis of Bcl-2, Caspase-3, and Caspase-9 under the same conditions. (C) Western blot analysis of angiogenic proteins CD31 and vWF. (D) Western blot analysis of osteogenic proteins Runx2 and ALP. (E–M) RT-qPCR analysis of PI3K, AKT, Bcl-2, Caspase-3, Caspase-9, CD31, vWF, ALP, and Runx2 mRNA expression under the indicated interventions (n = 3). OLP treatment concentration-dependently reversed these MPS-induced alterations. Specifically, OLP enhanced PI3K and AKT phosphorylation (p < 0.05), increased Bcl2, ALP, and Runx2 expression (p < 0.05), and decreased cleaved Caspase-3 and cleaved Caspase-9 levels (p < 0.05). Furthermore, OLP upregulated the expression of the angiogenic proteins vWF and CD31 (p < 0.05). These results demonstrate that OLP exerts its protective effects by activating the PI3K-AKT pathway, thereby promoting anti-apoptotic and osteogenic activity, and may also stimulate angiogenesis under glucocorticoid-impaired conditions (Fig. S2). To further validate the role of the PI3K-AKT pathway in mediating OLP’s effects, LY294002 (PI3K inhibitor) and 740 Y-P (PI3K activator) were introduced. Western blot showed that co-treatment with 740 Y-P and OLP (40 µg/mL) further enhanced PI3K and AKT phosphorylation, significantly increased Bcl2, ALP, and Runx2 expression, and suppressed pro-apoptotic markers. In contrast, LY294002 attenuated PI3K and AKT phosphorylation, reduced Bcl2, ALP, and Runx2 expression, and led to increased Caspase-3 and Caspase-9 levels despite OLP co-treatment. The same trends were further confirmed by qPCR analysis (Fig. 6E-M). In vivo immunohistochemical analysis of the SONFH rat model further supported these findings. Compared to the control group, the MPS group exhibited significantly reduced expression of p-PI3K and p-AKT (by approximately 22.81% and 20.45%, respectively, p < 0.05), as well as decreased Bcl2, ALP, and Runx2 expression levels. Meanwhile, the expression level of Caspase-3 and Caspase-9 increased 57.54% and 34.91%, respectively (p < 0.05), indicating severe apoptotic and osteogenic impairment. Notably, rats in the MPS + OLPH treatment group showed significant restoration of p-PI3K, p-AKT, ALP, and Runx2 expression (p < 0.05), with Bcl-2 levels returning to near-baseline. The expression level of Caspase-3 and Caspase-9 was markedly reduced (p < 0.05), consistent with inhibition of apoptosis. Furthermore, the OLPH-treated group exhibited elevated expression of vWF and CD31, implying a pro-angiogenic effect (Figs. 7A-F and 8A-E). Immunohistochemical staining and quantitative analysis of PI3K–AKT signaling, apoptosis markers in SONFH rat femoral heads. (A) Representative immunohistochemical staining of p-PI3K, p-AKT, Bcl-2, Caspase-3, Caspase-9 in femoral head tissues from each group (10X scale bar = 400 μm, 40X scale bar = 100 μm). (B–F) Quantification of p-PI3K, p-AKT, Bcl-2, Caspase-3, Caspase-9 positive expression in femoral head tissues (n = 6). Statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Immunohistochemical staining and quantitative analysis of osteogenesis and angiogenesis markers in SONFH rat femoral heads. (A) Representative immunohistochemical staining of ALP, Runx2, CD31, and vWF in femoral head tissues from each group (10X scale bar = 400 μm, 40X scale bar = 100 μm). Quantification of Runx2, ALP, vWF and CD31 positive expression in femoral head tissues (n = 6). Statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Collectively, these findings provide strong evidence that OLP alleviates steroid-induced osteoblast apoptosis and promotes osteogenesis and angiogenesis through activation of the PI3K-AKT-Bcl2 signaling pathway. This mechanistic insight reinforces the therapeutic potential of OLP in delaying or reversing the progression of SONFH. Discussion SONFH remains a challenging clinical condition with limited effective pharmacological interventions18,19. From a pathophysiological perspective, increasing evidence suggests that osteoblast and osteocyte apoptosis within the epiphyseal and subchondral region of the femoral head represents an early pathological event in SONFH progression, preceding irreversible structural collapse and therefore constituting a critical therapeutic target20. Against this background, the present study provides compelling evidence that OLP, a natural phenolic compound derived from olive leaves, exerts significant therapeutic effects against SONFH by mitigating osteoblast apoptosis, promoting osteogenesis and angiogenesis, and activating the PI3K-AKT-Bcl2 signaling pathway. Our findings not only elucidate the molecular mechanisms underlying OLP’s protective actions but also highlight its potential as a novel natural agent for early intervention in SONFH. The pathogenesis of SONFH is multifactorial, involving dysregulated lipid metabolism, impaired vascular perfusion, and—critically—accelerated osteoblast apoptosis21,22. GCs such as MPS are known to induce oxidative stress and mitochondrial dysfunction in osteoblasts, leading to caspase activation and programmed cell death23–25. In line with previous reports, our in vitro experiments confirmed that MPS significantly reduced MC3T3-E1 cell viability and increased apoptosis in a dose-dependent manner. Notably, co-treatment with OLP effectively reversed these cytotoxic effects, restoring cell viability and reducing apoptosis in a concentration-dependent fashion. These results are consistent with earlier studies demonstrating OLP’s anti-apoptotic properties in other cell types26,27. But to our knowledge, this is the first report to establish its efficacy in a SONFH context. Beyond cellular models, we established a rat model of SONFH using LPS and MPS—a well-validated protocol that recapitulates key features of human disease, including trabecular thinning, increased bone porosity, and empty osteocyte lacunae28,29. Micro-CT analysis revealed that OLP treatment significantly improved bone microarchitecture parameters such as BV/TV, Tb.Th, Tb.N, and Tb.Sp. Histopathological assessments further confirmed that OLP reduced empty lacunae, restored trabecular continuity, and enhanced collagen organization. These structural improvements were accompanied by enhanced mineralization, as evidenced by Goldner’s trichrome staining, and reduced fibrosis by Masson’s trichrome results. Importantly, OLP showed no signs of systemic toxicity in major organs, supporting its favorable biocompatibility profile. This LPS- and glucocorticoid-based model has been widely applied to reproduce key pathological characteristics of steroid-associated osteonecrosis, particularly trabecular deterioration and osteocyte loss within the femoral head, thereby supporting the translational relevance of the in vivo findings30. To unravel the mechanistic basis of OLP’s effects, we employed an integrated approach combining network pharmacology, proteomics, molecular docking, and functional validation. Network pharmacology analysis identified 123 overlapping targets between OLP and SONFH, with TNF, AKT1, EGFR, CASP3, and BCL2 emerging as core targets. KEGG enrichment analysis highlighted apoptosis, PI3K-AKT signaling, and cell cycle regulation as central pathways. These in silico predictions were corroborated by proteomic profiling, which revealed 471 differentially expressed proteins (DEPs) after OLP treatment. Among these, anti-apoptotic (e.g., Gpnmb, Tsc22d3), antioxidant (Gclc), and cholesterol regulatory (Npc2) proteins were upregulated, while pro-apoptotic and inflammatory mediators (e.g., Nr3c1, Ptgs2) were downregulated. These findings suggest that OLP exerts pleiotropic effects across multiple cellular processes, ultimately converging on enhanced cell survival and reduced inflammation. Importantly, by integrating network pharmacology–based target prediction with proteomic profiling and subsequent functional validation, our study delineates a coherent mechanistic framework linking OLP treatment to the regulation of osteoblast apoptosis and survival. In line with this integrative evidence, previous studies have likewise implicated the PI3K–AKT signaling axis as a central regulator of osteoblast survival and glucocorticoid-induced bone injury, supporting the biological relevance of this pathway in the pathogenesis of SONFH31. Molecular docking simulations further supported OLP’s high-affinity binding to key targets such as AKT and BCL2, providing a structural basis for its pharmacological activity. Functional experiments using Western blot and qPCR confirmed that OLP activates the PI3K-AKT pathway, as evidenced by increased phosphorylation of PI3K and AKT, upregulation of the anti-apoptotic protein Bcl2, and downregulation of cleaved Caspase-3 and Caspase-9. The use of pathway-specific modulators—LY294002 (PI3K inhibitor) and 740 Y-P (PI3K activator)—further validated that OLP’s anti-apoptotic and osteogenic effects are PI3K-AKT-dependent. Moreover, we demonstrated that OLP promotes osteogenic differentiation, as indicated by increased expression of ALP and Runx2, and may exert pro-angiogenic effects, as reflected by the upregulation of vWF and CD31. These findings are particularly relevant given the critical role of vascular insufficiency in SONFH progression. Although direct perfusion-based vascular imaging was not performed in the present study, angiogenesis-related markers such as CD31 and vWF are commonly used to indicate vascular involvement in bone remodeling, and accumulating evidence suggests a close coupling between angiogenesis and osteogenesis in skeletal homeostasis32,33. By simultaneously targeting apoptosis, osteogenesis, and angiogenesis, OLP may influence multiple pathological facets of SONFH, potentially distinguishing it from single-target therapies. Our results align with emerging research on natural compounds in bone protection. For instance, asiatic acid and proanthocyanidins have been shown to attenuate GC-induced osteonecrosis via similar pathways34,35. However, OLP offers a unique combination of anti-inflammatory, antioxidant, and anti-apoptotic properties, which may underlie its superior efficacy in our model. Furthermore, the ability of OLP to modulate cholesterol homeostasis via Npc2 upregulation may indirectly protect against lipid metabolism dysregulation—a key contributor to SONFH36, which should be investigated in future work. Despite these promising results, several limitations must be acknowledged. First, while our rat model recapitulates many features of human SONFH, species-specific differences may affect translational relevance. Second, vascular changes were primarily evaluated using immunohistochemical detection of CD31 and vWF, which provide indirect evidence of angiogenesis. More direct assessments, such as Microfil perfusion combined with micro-CT–based vascular imaging, would allow a more comprehensive evaluation of intraosseous blood supply and vascular reconstruction, which should be addressed in future investigations. Third, the precise bioavailability and tissue distribution of OLP in bone remain unclear. Future studies should employ pharmacokinetic profiling and advanced delivery systems (e.g., nanoparticles) to optimize OLP dosing and targeting. Fourth, although we focused on the PI3K-AKT-Bcl2 axis, other pathways (e.g., NF-κB, Nrf2) may also contribute to OLP’s effects and warrant further investigation37. Conclusions This study establishes OLP as a promising therapeutic agent for SONFH, with demonstrated efficacy in reducing osteoblast apoptosis, improving bone microstructure, and activating pro-survival signaling. Its multi-targeted mechanisms and natural origin offer significant advantages over conventional treatments. While challenges like bioavailability require attention, OLP’s potential for early intervention and bone repair warrants further clinical exploration. Future studies should focus on translational validation and long-term outcomes to advance OLP toward clinical application. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Author contributions Jianlong Lin: Writing – original draft, Visualization, Methodology, Conceptualization. Rui Zhang: Writing – review & editing, Writing – original draft, Visualization, Methodology. Xingyu Lin: Investigation, Data curation. Xiangjun Luo: Methodology, Formal analysis. Haokun Lin: Software, Data curation. Jie Xu: Supervision, Project administration. Shishui Lin: Supervision, Project administration. Funding This work was financially supported by the Joint Funds for the Innovation of Science and Technology, Fujian Province (grant number: 2024Y9035), the Natural Science Foundation of Fujian Province (grant number: 2025J01078), Fujian Provincial Health Technology Project (grant number: 2024CXA004), and the Special Funds of Fujian Provincial Department of Finance (grant number: 2024 − 881). Data availability The spectrometry proteomics data generated in this study have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD076592. The dataset is currently available for access (https://www.ebi.ac.uk/pride/archive/projects/PXD076592). Declarations Competing interests The authors declare no competing interests. Ethics approval and consent to participate This study was authorized by the Institutional Animal Ethics Committee of Fujian Medical University (IACUC FJMU 2024 − 0194). Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally to this work: Jianlong Lin, Rui Zhang, Jie Xu and Shishui Lin. Contributor Information Jie Xu, Email: 457884651@qq.com. Shishui Lin, Email: linshishui1@fjmu.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The spectrometry proteomics data generated in this study have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD076592. The dataset is currently available for access (https://www.ebi.ac.uk/pride/archive/projects/PXD076592).