A total of 24 male Sprague-Dawley rats (12 weeks old; 250–300 g; Liaoning Changsheng Biotechnology Co., Ltd.) were used in the present study. All rats were housed under standard laboratory conditions (12-h light/dark cycle, 24–25°C and 50–55% humidity) and provided free access to food and water. The experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (8th edition) (17) and approved by the Institutional Animal Care and Use Committee of the Affiliated Hospital of Shandong University of Traditional Chinese Medicine (approval no. 2019-58).

All rats were randomized into control, SANFH, or SANFH+TFRD groups (n=8 each). TFRD was purchased from Beijing Qihuang Pharmaceutical Manufacturing Co., Ltd. (trade name: Qiang gu capsuled, drug approval number: Z20030007). Based on the instructions of Qiang gu capsule, clinical usage (0.75 g·day for adults) and the Meeh-Rubner equation (18) of dose conversion, 20 mg/kg dosage was chosen for administration to rats by intraperitoneal injection daily. A previous study reported that rats were administered with Rhizoma drynariae extract at dose of 20 mg/kg body weight once daily (19). The SANFH model was induced by intramuscular injection of lipopolysaccharides (LPS) and methylprednisolone (MPS). Briefly, rats in the SANFH and SANFH+TFRD groups were given two doses of 20 µg/kg LPS (Escherichia coli 055: B5; Sigma-Aldrich; Merck KGaA) on days 0 and 1 (twice at 24 h intervals) and administered 40 mg/kg MPS (Pfizer, Inc.) on days 3, 4 and 5 (three times after a 24-h interval). Rats in the control group were injected with normal saline. Additionally, rats in the SANFH+TFRD group were intraperitoneally injected with 20 mg/kg/day TFRD every day, while the remainder of the rats were injected with normal saline. The time interval for all injections was 24 h. All rats were weighed weekly. Subsequently, four weeks after the final MPS injection, all rats were sacrificed by intraperitoneal overdose anaesthesia with pentobarbital sodium (240 mg/kg, Sigma-Aldrich; Merck KGaA) and femoral heads were harvested for subsequent tests.

The femur specimens were fixed in 4% paraformaldehyde for 48 h at 4°C and then decalcified with 10% ethylenediaminetetraacetic acid for 4–6 weeks at room temperature. After complete decalcification, the specimens were dehydrated in graded ethanol, embedded in paraffin and cut into 4-µm sections. Thereafter, the sections were dewaxed with xylene and then gradually dehydrated via gradient ethanol. The sections were stained with haematoxylin solution for 5 min at 37°C followed by eosin solution for 2 min at 37°C. Finally, the sections were dehydrated with gradient ethanol, washed three times using xylene and then mounted in neutral gum. The histopathological changes were examined under a light microscope (Leica Microsystems GmbH) by two independent authors. Three sections were obtained from each femoral head and the percentage of empty lacunae was assessed at ×100 magnification in five randomly selected fields per section. The diagnosis of ONFH was established on the basis of the pathological features of empty bone lacuna numbers and surrounding necrotic bone marrow tissues.

TUNEL assay was used to detect the apoptosis rate of osteoblasts using a Situ Cell Death Detection kit (Nanjing Jiancheng Bioengineering Institute) according to the manufacturer's protocol. Following routine dewaxing and dehydration, the sections were digested with pepsin (~0.25%-0.5% hydrochloric acid solution) for 25 min at room temperature, incubated with 50 µl TUNEL reaction mixture for 1 h at 37°C and then incubated with 50 µl of peroxidase for 30 min at 37°C. Thereafter, the sections were dyed with 3,3-diaminobenzidine reagent and stained with hematoxylin for 2 min at room temperature. Cells with brown nuclei were considered as TUNEL-positive cells and were further assessed under a fluorescence microscope (Olympus Corporation). The apoptosis rate of osteoblasts was defined as the proportion of the number of TUNEL-positive cells to the total number of cells.

Total protein was extracted from the osteoblasts and tissues of the femoral head using radioimmunoprecipitation assay buffer containing protease and phosphatase inhibitors (Abcam). A bicinchoninic acid (BCA) kit (Thermo Scientific, Inc.) was used to measure the total protein concentration. Equal amounts of protein (30 µg/lane) were loaded on 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis and then transferred onto polyvinylidene fluoride membranes (Pall Life Sciences). Thereafter, the membranes were blocked with 5% non-fat milk (Sigma-Aldrich; Merck KGaA) in Tris-buffer saline containing 0.1% Tween-20 for 1 h at room temperature and then incubated with the following primary antibodies overnight at 4°C: Anti-caspase-3 (cat. no. 14220; 1:1,000), anti-Bax (cat. no. 14796; 1:1,000), anti-RUNX2 (cat. no. 12556; 1:1,000), anti-phosphorylated (p-)PI3K (cat. no. 17366; 1:1,000), anti-p-AKT (cat. no. 4060; 1:1,000), anti-AKT (cat. no. 4691; 1:1,000) and anti-glyceraldehyde 3-phosphate dehydrogenase (internal parameter; cat. no. 5174; 1:1,000 all from Cell Signaling Technology, Inc.; anti-Bcl-2 (ab196495; 1:1,000) and anti-OCN (ab13418; 1:1,000), from Abcam; and anti-VEGF (sc-7269; 1:1,000), anti-OPG (sc-390518; 1:1,000), anti-RANKL (sc-377079; 1:1,000) and anti-PI3K (sc-1637; 1:1,000) from Santa Cruz Biotechnology, Inc. The membranes were then incubated with corresponding horseradish peroxidase-conjugated secondary antibodies (1:1,000, Wuhan Boster Biological Technology Co., Ltd.) for 1 h at room temperature. Finally, the labelled proteins were treated with enhanced chemiluminescence (ECL) solution (Pierce; Thermo Fisher Scientific, Inc.) and the protein bands were analysed using ImageJ software, version 1.8.0 (National Institutes of Health).

Osteoblasts were isolated from calvariae of neonatal rats as previously described (20). Briefly, heads of 2-day-old neonatal rats were cut off, soaked in phosphate-buffered saline (PBS; HyClone; Cytiva) and the calvariae were then dissected out. Thereafter, the calvariae were cut into pieces, incubated with 1% trypsin (Gibco; Thermo Fisher Scientific, Inc.) for 15 min at 37°C and then incubated with PBS containing 0.25% collagenase (Type I; Sigma-Aldrich; Merck KGaA) for 60 min at 37°C. Following centrifugation at 300 × g for 10 min at 37°C, the osteoblasts were obtained and cultured in α-minimal essential medium containing 12% foetal bovine serum and antibiotics (Gibco; Thermo Fisher Scientific, Inc.) at 37°C with 5% CO₂.

Osteoblasts were treated with different concentrations of dexamethasone (Dex; 0, 10⁻⁶, 10⁻⁷ and 10⁻⁸ M; Sigma-Aldrich; Merck KGaA) for 0, 24, 48 and 72 h at 37°C or TFRD (0, 0.1, 0.5, 2.5, 12.5 and 62.5 µg/ml) for 48 h at 37°C to estimate the dosage effects on cell proliferation. Osteoblasts were divided into control (without treatment), Dex (osteoblasts treated with 10⁻⁶ M Dex for 48 h at 37°C), Dex+TFRD [osteoblasts pre-treated with 12.5 µg/ml TFRD (the maximum concentration of TFRD without cytotoxicity on osteoblasts) for 2 h at 37°C and then incubated with 10⁻⁶ M Dex for 48 h at 37°C], Dex+TFRD+LY294002 group (osteoblasts pre-treated with 12.5 µg/ml TFRD for 2 h at 37°C and 10 µM LY294002 [an inhibitor of PI3K/AKT pathway; Beyotime Institute of Biotechnology] for 1 h at 37°C and then incubated with 10⁻⁶ M Dex for 48 h at 37°C). Following treatment, osteoblasts were collected for further experiments.

Osteoblast proliferation was assessed using the MTT assay. Briefly, osteoblasts (2.0×10⁴ cells/ml) were seeded onto 96-well plates and exposed to TFRD, Dex and LY294002 as detailed in the previous experiment, respectively. Thereafter, osteoblasts were incubated with 20 µl MTT solution (5 mg/ml, Beyotime Institute of Biotechnology) at 37°C for 4 h and then incubated with 150 µl dimethyl sulfoxide. Finally, optical density (OD) was measured at 490 nm with a spectrophotometric plate reader (Bio-Rad Laboratories, Inc.). The experiments were repeated three times.

The Annexin V-fluorescein isothiocyanate (FITC) early apoptosis detection kit (Cell Signaling Technology, Inc.) was used to analyse osteoblast apoptosis. Briefly, osteoblasts were incubated with Annexin V-FITC (10 µl) and propidium iodide (5 µl) for 30 min without light at room temperature. The apoptosis rate of osteoblasts (including early and late apoptotic cells) was measured using a BD FACSCanto II flow cytometer (Becton, Dickinson and Company).

The ROS level in osteoblasts was measured using a dichlorofluorescein-diacetate (DCFH-DA) dye (Sigma-Aldrich; Merck KGaA). Briefly, osteoblasts (3×10³ cells per well) were seeded into 24-well plates and exposed to different treatments as detailed in the previous experiments. The osteoblasts were then collected, washed with PBS and incubated with 10 µM DCFH-DA for 30 min at 37°C without light. Fluorescence intensity was measured via flow cytometry.

All data were presented as the mean ± standard deviation and analysed using GraphPad Prism 7.0 software (GraphPad Software, Inc.). Significant differences among three or more groups were assessed by one-way analysis of variance (ANOVA) with Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

In the present study, LPS and MPS were used to establish a rat model of SANFH. To investigate the influence of TFRD on SANFH rats, rats were intraperitoneally injected with TFRD. During the experimental period, all rats survived and there were no significant differences in the mean body weights among the control, SANFH and SANFH+TFRD groups (P>0.05; Fig. 1A). Haematoxylin and eosin staining was used to assess the pathological changes in the femoral head. As shown in Fig. 1B, bone structure was intact in the control group, while there were many empty bone lacunae and broken trabeculae in the SANFH group. In addition, the percentage of empty lacuna in the SANFH+TFRD group was lower compared with the SANFH group (P<0.05). TUNEL assay was conducted to detect apoptosis of the bone cells of the femoral head. As presented in Fig. 1C, the apoptosis rate in the SANFH and SANFH+TFRD groups was higher compared with the control group (P<0.01), however treatment with TFRD reduced the apoptosis rate compared with the SANFH group (P<0.05). In addition, the expression of apoptosis-related proteins (caspase-3, Bax and Bcl-2) was detected via western blotting. The results showed that the expression of caspase-3 and Bax in the SANFH and SANFH+TFRD groups was increased whereas the expression of Bcl-2 was decreased in comparison with the control group (Fig. 1D-G; P<0.01), while treatment with TFRD increased the expression of Bcl-2 and decreased the expression of Bax and caspase-3 compared with that in the SANFH group (Fig. 1D-G; P<0.01).

Protein expression in the femoral head of rats was detected via western blotting. As shown in Fig. 2, Compared with the control group, the expression of VEGF, RUNX2, OPG and OCN in the SANFH and SANFH+TFRD groups was decreased (P<0.01), whereas the expression of RANKL was increased (P<0.01). Compared with the SANFH group, treatment with TFRD increased the expression of VEGF, RUNX2, OPG and OCN and reduced the expression of RANKL (P<0.05).

To explore the protective mechanism of TFRD in SANFH rats, the PI3K/AKT pathway-related proteins were detected by western blotting. The results showed that the SANFH and SANFH+TFRD groups had lower expression of p-P13K/P13K and p-AKT/AKT than that in the control group (P<0.01; Fig. 3), while treatment with TFRD increased the protein expression of p-P13K/P13K and p-AKT/AKT compared with the SANFH group (P<0.01; Fig. 3).

The effect of different concentrations of Dex at different time points on osteoblast proliferation was investigated using MTT assay. The results showed that Dex inhibited osteoblast proliferation in a dose- and time-dependent manner and that 10⁻⁶ M Dex inhibited osteoblast proliferation (P<0.05; Fig. 4A). Additionally, MTT assay was used to evaluate the cytotoxic effect of TFRD on osteoblasts. As shown in Fig. 4B, when the concentration of TFRD was less than 12.5 µg/ml, TFRD had no cytotoxic effect on osteoblasts. Therefore, 10⁻⁶ M Dex and 12.5 µg/ml TFRD were selected for subsequent experiments. As can be seen in Fig. 4C-E, treatment with Dex reduced the OD₄₉₀ value, increased apoptotic rate and ROS production in osteoblasts compared with the control group (P<0.01), while the opposite effects of TFRD in osteoblasts were observed. Compared with the control group, Dex increased the expression of caspase-3 and Bax and decreased the expression of Bcl-2, p-P13K/P13K and p-AKT/AKT (P<0.01; Fig. 4F), but this effect was weakened by TFRD.

In order to elucidate the mechanism of TFRD-induced protection of osteoblasts from Dex-induced damage, the PI3K/AKT pathway inhibitor (LY294002) was used to treat the osteoblasts in the Dex+TFRD group. As shown in Fig. 5A, the expression of p-P13K/P13K and p-AKT/AKT in the Dex+TFRD+LY294002 group was reduced compared with that in the Dex+TFRD group (P<0.01), indicating that the activation of the PI3K/AKT pathway induced by TFRD was blocked via LY294002. Subsequently, the effects of LY294002 on proliferation, apoptosis and ROS production was detected in osteoblasts. As presented in Fig. 5B-D, following LY294002 treatment, the OD₄₉₀ value was decreased, while the apoptotic rate and ROS production were increased in the osteoblasts (P<0.05).