Role of PI3K in the bone resorption of apical periodontitis

This study was conducted in full accordance with the World Medical Association Declaration of Helsinki and was approved by the Ethics Committee of the Affiliated Stomatological Hospital of Dalian Medical University School of Stomatology (Approval number: 2021005). Participants were patients who were seen by a doctor at the Department of Endodontics or Maxillofacial Surgery of the Dental Hospital of Dalian Medical University between January 2021 and December 2021 due to chronic apical periodontitis. The clinical cases were all apical periodontitis caused by caries-derived infection, and the diagnosis was mainly based on X-ray, with a periapical index (PAI) score ≥ 3 [25, 26]. A total of 26 periapical tissue samples were collected from patients who needed apical surgery or tooth extraction due to apical periodontitis. Among these patients, 14 were males and 12 were females, aged 22–53 years. At the same time, healthy periodontal ligament tissues from 10 patients after extraction of premolars or third molars as part of orthodontic treatment during the same period were collected as the control group, comprising six males and four females, aged 19–54. The subjects were fully informed and provided signed consent to provide tissue samples for research. The chronic apical periodontitis patients were diagnosed based on clinical manifestations and X-rays and they were in good health with no systemic diseases or immune system diseases. Exclusion criteria were: ① the tooth was affected by periodontal disease or combined periodontal and pulpal disease; ② antibiotics or non-steroidal anti-inflammatory drugs had been taken in the 3 months before tooth extraction; ③ allergies; ④ refusal to participate in this study. Each clinically-obtained tissue sample was placed into an Eppendorf (EP) tube containing 0.5 mL Trizol and stored at − 80 °C for real-time quantitative polymerase chain reaction (RT-qPCR).

A total of 25 C57BL/6L female mice, 6–8-weeks old, were purchased from the Experimental Animal Center of Dalian Medical University. The animal experiments were approved by the Ethics Committee of Dalian Medical University (approval number: 2021006). An intraperitoneal injection of Ketamine (90–150 mg/kg) + Xylazine (7.5–16 mg/kg) was administered to anesthetize the mice, then apical periodontitis was established in the mandibular first molar by exposure of the pulp cavity [27]. The mice were fed on standard chow and allowed free access to food and water. At 0, 1, 2, 3, and 4 weeks after opening the pulp cavity, five mice were randomly selected for euthanasia. The mouse mandibular periapical tissue was prepared for hematoxylin–eosin (HE) staining, and the expression of relevant proteins was analyzed by immunohistochemical staining, while enzymatic histochemical staining was used for osteoclast detection. The methods of HE staining and immunohistochemical staining were as described in our previous study [16]. The PI3K polyclonal antibody (Bioworld, Nanjing, China) and AKT polyclonal antibody (Bioworld, Nanjing, China) were both used at a dilution of 1:100. Histochemical staining was carried out using a tartrate-resistant acid phosphatase (TRAP) staining kit (Sigma-Aldrich, Missouri, USA) according to the manufacturer’s instructions. The expression of PI3K and Akt, and the appearance of osteoclasts in mice with apical periodontitis was calculated using the average optical density value at 450 nm [28, 29].

Induced osteoclasts or osteoblasts were chosen and inoculated at a density of 2 × 10⁶ cells/well into a six-well plate. After allowing the cells to adhere for 24 h, the cell culture medium was replaced and fresh medium plus 100 ng/ml LPS (Sigma-Aldrich, Missouri, USA) was added to the experimental group [30], while the control group received vehicle alone. After 24 h of exposure, the cells were harvested and the gene expression levels of PI3K, Acp5 and NFATc1, and the osteoblast-related factors BMP-2 and Runx2, were analyzed by RT-qPCR, while western blotting was used to analyze the protein expression levels of these factors.

Osteoclasts or osteoblasts of the third or fourth generation were seeded at 2 × 10³ cells/well into 96-well plates. After the cells were allowed to adhere, the original culture medium was discarded, and inhibition of PI3K activity with the specific inhibitor LY294002 (Cell Signaling Technology, Boston, USA) at 0 or 20 μmol/L, was added to the cells. After a further 24 h of culture, CCK-8 (APExBIO, Houston, USA) reaction solution was added, and cells were incubated at 37 °C in a 5% CO₂ incubator for 1 h, then the OD value at 450 nm was measured using a microplate reader.

Osteoclasts or osteoblasts in good condition were selected to inoculate into a 6-well plate at a density of 2 × 10⁶ cells/well. After the cells were allowed to adhere for 24 h, the cell culture medium was replaced, and LY294002 was added at 0 or 20 μmol/L to the control or experimental groups, respectively. After a further 24 h, the cells were harvested for analysis. The effect of the PI3K inhibitor was analyzed by RT-qPCR to detect the gene expression of Akt, Acp5 and NFATc1 in osteoclasts, and Akt, BMP-2 and Runx2 in osteoblasts. Western blotting was used to analyze the protein expression of p-Akt, Akt, TRAP and NFATc1 in osteoclasts, p-Akt, Akt, BMP-2 and Runx2 in osteoblasts.

RNA was extracted from tissues or cells according to the instructions provided with the total RNA extraction kit (Takara, Shiga, Japan), and reverse transcribed to cDNA. Each cDNA sample was added to an eight-tube and centrifuged to mix. After mixing, the tube was placed in a qPCR reactor to carry out the PCR reaction. The RT-qPCR program consisted of a heating step at 95 °C for 30 s followed by 40 cycles of 95 °C for 5 s and 56 °C for 30 s. The primer sequences of human-related genes are shown in Table 1 and those of mouse-related genes in Table 2. The primer sequences were all designed and produced by Shanghai Bioengineering Co., Ltd. (Shanghai, China). The 2^−△△Ct method was used to calculate the relative expression of each target gene according to the Ct value of the internal control and target gene.

Total protein was extracted from experimental samples of clinical tissues or harvested cells. The protein concentration and protein denaturation were determined using a BCA assay kit (Sigma-Aldrich, Missouri, USA), according to the instructions supplied. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) was performed according to the electrophoresis gel preparation kit, then the separated proteins were transferred to polyvinylidene difluoride membranes, the blots were cut prior to hybridisation to the antibodies, and incubated with the primary antibody at 4 °C overnight. The antibodies used were as follows: rabbit monoclonal anti-GAPDH (1:1000 dilution; Bioworld, Nanjing, China), rabbit monoclonal anti-PI3K (1:500 dilution; Bioworld, Nanjing, China), rabbit monoclonal anti-Akt (1:5000 dilution; Cell Signaling Technology, Boston, USA), rabbit monoclonal anti-p-Akt (1:5000 dilution; Abcam, Cambridge, UK), rabbit monoclonal anti-TRAP (1:500 dilution; ABclonal, Wuhan, China), rabbit monoclonal anti-NFATc1 (1:500 dilution; Elabscience Biotechnology, Wuhan, China), rabbit monoclonal anti-BMP-2 (1:500 dilution; Bioworld, Nanjing, China), rabbit monoclonal anti-Runx2 (1:5000 dilution; Bioworld, Nanjing, China). After washing three times with TBST on a shaker, each time for 10 min, the membranes were incubated with the secondary antibody (1:1000 dilution; Bioworld, Nanjing, China). After further washes with TBST, membranes were incubated with ECL luminescent solution (Thermo Fisher Scientific, Waltham, USA), and a Bio-Rad gel imaging system was used to capture the western blot results. Image Lab software (Bio-Rad) was then used to analyze the results. After the protein bands were obtained in the experiment, the gray value of each band was obtained using Image Lab software. The internal reference protein GAPDH was used for normalization, and the ratio of the value of the target protein to GAPDH was the relative gray value of that protein. Each experiment was repeated three times, and the obtained relative gray values were used for statistical analysis.

In vivo experiments, HE staining, immunohistochemical staining and TRAP staining have 0, 1, 2, 3, 4 time points, 5 animals at each time-point, 5 non-consecutive sections per animal, a total of 25 data-points in each group for statistical analysis. For in vitro experiments, each experiment was undertaken at least three times unless stated otherwise. SPSS 17.0 software was used to analyze the data. Measurement data were expressed as mean ± SD, comparisons between two groups were performed using Tukey's test (α = 0.05), while comparisons between multiple groups were performed using one-way analysis of variance. The significance level for all analyses was set at 5% (α = 0.05). Bivariate correlation analysis of the relationship between PI3K, AKT and osteoclasts was performed.

As shown in Fig. 1, the expression of PI3K (P = 0.009, Fig. 1A), Acp5 (P < 0.001, Fig. 1B) and NFATc1 (P = 0.003, Fig. 1C) genes were all higher in the chronic apical periodontitis samples than in healthy periodontal ligament tissue, and the difference was statistically significant.

The results of HE staining and TRAP staining are shown in Fig. 2A, B. With the continuous progression of periapical inflammation, the exudation of periapical tissue and the extent of bone destruction gradually increased. Immunohistochemical staining (Fig. 2C, D) showed that cells positive for PI3K and Akt were yellow–brown, and their expression levels were highly consistent. Correlation analysis (Fig. 2E) showed that there was a high correlation between PI3K OD value and osteoclast OD value (r = 0.883, P < 0.001), and there was also a high correlation between AKT OD value and osteoclast OD value (r = 0.827, P < 0.001), while there was a moderate correlation between PI3K OD value and AKT OD value (r = 0.615, P < 0.001).

As shown in Fig. 3, after 5 days of induction of RAW264.7 cells, TRAP staining stained the cytoplasm wine-red with acid phosphatase granules, and most of the cells were giant cells with three or more nuclei. RT-qPCR analysis revealed that the mRNA expression of Acp5 was significantly higher after 5 days of induction of RAW264.7 cells compared with 0 days (P < 0.05). The above results indicate that RAW264.7 cells were successfully induced to form osteoclasts.

The results of ALP staining are shown in Fig. 4. Compared with 0-day MC3T3-E1 cells, MC3T3-E1 cells induced for 21 days contained a large number of flaky, dark red calcium nodules, and the mRNA expression of ALP was significantly increased after induction of MC3T3-E1 cells for 7 days compared with day 0 (P < 0.05). The above results indicate that MC3T3-E1 cells were successfully induced to form osteoblasts.

Gene and protein expression results of cells cultured in an LPS-mediated inflammatory microenvironment are shown in Fig. 5. In the experimental group of osteoclasts, the mRNA levels of PI3K and osteoclast-related factors Acp5 and NFATc1 were significantly increased (P < 0.05) (Fig. 5A), while the protein levels of PI3K, TRAP and NFATc1 (Fig. 5C, D) were consistent with the RT-qPCR results (P < 0.05).

In osteoblasts, gene expression analysis (Fig. 5B) showed that mRNA levels of PI3K and the osteogenic factors BMP-2 and Runx2 were significantly reduced (P < 0.05). The protein levels of PI3K, BMP-2 and Runx2, analyzed by western blotting, were consistent with RT-qPCR results (Fig. 5E, F).

The results of CCK-8 assay are shown in Fig. 6A, B. The RT-qPCR results are shown in Fig. 6C, D, The mRNA levels of Akt and the osteoclast differentiation markers Acp5 and NFATc1 decreased significantly in the osteoclast experimental group (P < 0.05), while mRNA levels of Akt, BMP-2 and Runx2 were similarly reduced in the osteoblast experimental group (P < 0.05). The ratio of p-Akt/Akt protein in the osteoclast experimental group decreased, and the protein expression levels of the osteoclast differentiation markers TRAP and NFATc1 corresponded to the RT-qPCR results, with the expressions of all being significantly decreased (P < 0.05) (Fig. 6E, F). As shown in Fig. 6G, H, the ratio of p-Akt/Akt protein in the osteoblast experimental group decreased and the protein levels of the bone formation-related factors BMP-2 and Runx2 also all decreased significantly, consistent with the results of RT-qPCR (P < 0.05).