Twenty male BALB/c mice (6-week-old; 23–25 g) were purchased from the Animal Center of Chinese Academy of Sciences (Shanghai, China) and housed in a specific-pathogen free laminar flow room under constant temperature (25–27°C), a 12-h light/dark cycle and humidity (40–50%) with access to food and water ad libitum. All experiments and animal care procedures were approved by the Animal Center of Sichuan University (Sichuan, China). All experimental procedures were approved by the Experimental Animal Committee of the State Key Laboratory of Oral Diseases, West China College of Stomatology, Sichuan University (Chengdu, China).

The PDL tissues were obtained from the premolar teeth of three donors, which were extracted for orthodontic purposes at the West China Hospital of Stomatology of Sichuan University. The protocols regarding the use and manipulation of PDL tissues were approved by the Institutional Review Board of West China Hospital of Stomatology, Sichuan University (Chengdu, China) and written informed consent was obtained from each donor. The extracted teeth were rinsed and placed in Dulbecco's modified Eagle's medium (DMEM; GE Healthcare Life Sciences, Logan, UT, USA) supplemented with 1,000 U/ml penicillin and 1,000 µg/ml streptomycin (Hyclone; GE Healthcare Life Sciences, Logan, UT, USA). The remaining procedures were performed, as described by Arnold et al (25). PDLs attached to the middle third of the root were removed with a curette to avoid contamination with the gingival and apical tissues. The PDL tissues were cut into ~1 mm² pieces and placed in 25 mm² culture flasks for cell culture in DMEM medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml of penicillin and 100 µg/ml of streptomycin at 37°C, under 5% CO₂ and 95% humidity. After reaching confluence of ~75% (in ~7 days), the cells were treated with 0.25% trypsin-0.1% EDTA (Hyclone; GE Healthcare Life Sciences) for cell passage. PDLCs at the 5th or 6th passage were used for the present studies.

The periodontal pathogens Porphyromonas gingivalis (P.g; ATTC33277), Prevotella intermedia (P.i; ATTC25611) and Fusobacterium nucleatum (F.n; ATTC25586) were obtained from State Key Laboratory of Oral Diseases (Chengdu, China) and cultured in brain heart infusion broth (Oxoid Ltd, Basingstoke, UK) under anaerobic conditions at 37°C for 48 h. The supernatants of the P.g culture were collected by centrifugation for 20 min at 9,600 × g, and stored at −80°C. Blood samples (50 ml) were taken from six healthy donors and the PBMCs were isolated using human lymphocyte separation tubes (DAKEWE, China), according to the manufacturer's protocols. The supernatants of the PBMCs were collected by centrifugation at 1,000 × g for 15 min and stored at −80°C. The full details of the procedure were as described by Lu et al (23). The protocols regarding the use and manipulation of PBMCs were approved by the Institutional Review Board of West China Hospital of Stomatology, Sichuan University (approval no. WCHSIRB-D-2013-039) and written informed consent was obtained from the donors.

The PDLCs were seeded in a 24-well plate at a density of 8×10⁴ cells/well and were incubated overnight at 37°C, under 5% CO₂ and 95% humidity. The cells were then stimulated with 10 ng/ml interleukin (IL)-1, IL-6, TNF-α or interferon (IFN)-γ, 50 ng/ml lipopolysaccharide (LPS), a combination of IL-1, IL-6, TNF-α and IFN-γ mixed at the ratio of 1:1:1:1 to a final concentration of 10 ng/ml of cytokines, or periodontal pathogenic bacteria at a PDLC:bacteria ratio of 1:50. The surface expression of PD-L1 on the stimulated PDLCs was measured using flow cytometric analysis. Recombinant human IFN-γ, IL-1, IL-2, IL-6 and TNF-α were purchased from R&D Systems (Minneapolis, MN, USA).

The pretreated PDLCs were harvested, washed twice with FCM buffer, comprising phosphate-buffered saline (PBS; ZSGB-BIO, Beijing, China) with 5% FBS (Hyclone; GE Healthcare Life Sciences) and 0.1% NaN₃ (Sigma-Aldrich), and incubated with either phycoerythrin (PE) mouse anti-human PD-L1 monoclonal antibody (3 µl per sample; dilution, 1:40; BioLegend, Inc., San Diego, CA, USA; cat. no. 329706) or isotype control antibody (3 µl per sample; dilution, 1:40; BioLegend, Inc.; cat. no. 400320) for 30 min at 4°C. The stained cells were washed twice with FCM buffer and analyzed using flow cytometry (Beckman Coulter FC500; Beckman Coulter, Miami, FL, USA) with Submit 5.2 software (Beckman Coulter).

The PDLCs were pretreated with 10 ng/ml of TNF-α or IFN-γ for 48 h to induce the surface expression of PD-L1. The isolated PBMCs were seeded at a density of 1×10⁷ cells/well in a six-well plate and incubated with 10 µg/ml phytohemagglutinin (PHA; Roche Diagnostics GmbH, Mannheim, Germany) for 72 h at 37°C to activate T lymphocytes (26). The surviving TNF-α or IFN-γ pre-treated, PDLCs were collected and then co-cultured with PHA-activated PBMCs at a ratio of 1:50 for 48 h for 37°C; untreated PDLCs were used as a negative control. To inhibit the binding of PD-L1 to PD-1, 10 µg/ml of purified mouse anti-human monoclonal CD274 (PD-L1, B7-H1) antibody (dilution, 1:40; eBioscience, San Diego, CA, USA; cat. no. 14-5983-82) was also added to the TNF-α-pretreated PDLC and PHA-activated PBMC co-culture. Following 18 h of incubation at 37°C, the cells were collected and labeled with propidium iodide (PI)/Annexin V-fluorescein isothiocyanate (FITC; Annexin V-FITC Apoptosis Detection kit; Nanjing KeyGen Biotech Co., Ltd., Nanjing, China)/APC-mouse anti-human CD4 monoclonal antibody (dilution, 1:40; BioLegend, Inc.; cat. no. 317416) or PI/annexin V-FITC/APC-mouse anti-human monoclonal CD8 antibody (dilution, 1:40; BioLegend, Inc.; cat. no. 300911) at 4°C for 30 min. The labeled cells were subjected to flow cytometric analysis, and the resulting APC⁺/FITC⁺ cells were gated to identify apoptotic T lymphocyte populations.

The PDLCs (1×10⁶) pretreated with 10 ng/ml TNF-α were stained with carboxyfluorescein diacetate succinimidyl ester (CFSE; Invitrogen; Thermo Fisher Scientific, Inc.) to label the live cells. Briefly, these PDLCs were incubated with CFSE (final concentration 10 µM) at 37°C for 10 min with gentle agitation. The cells were then washed twice with DMEM supplemented with 10% FBS and resuspended in DMEM. The CFSE-labeled cells were co-cultured with PHA-treated PBMCs at a ratio of 1:50 for 48 h. PBMCs and PDLCs alone were cultured as CFSE-negative and CFSE-positive controls, respectively. To inhibit the binding of PD-1 to PD-L1, 10 µg/ml of purified anti-human CD274 (PD-L1, B7-H1) antibody was added to the co-culture system at the time of mixing of the two cell types. Following 18 h of incubation, the cells were collected and stained with PI (final concentration 10 µg/ml) for 10 min to label the dead cells. The cells were then subjected to flow cytometric analysis, and signals gated as CFSE⁺/PI⁻ were considered live cells.

To establish an experimental model or peridontitis, 10 of the mice were randomly selected and were injected with P.g (P.g-injected group); and another 10 mice were injected with PBS as healthy controls (27). All surgery was performed under 200 mg/kg chloral hydrate [Meryer (Shanghai) Chemical Technology Co., Ltd., Shanghai, China] anesthesia and efforts were made to minimize suffering. For analysis of the outcomes, the severity of periodontitis in the model was categorized into two case types: Case type I was termed mild periodontitis and was defined by the presence of bleeding on probing, furcation cul-de-sac involvement, and facial-lingual tooth movement, without movements in a vertical or mesial direction. Case type II was termed severe periodontitis and was defined by the presence of bleeding on probing, furcation through-and-through involvement, and tooth movement in facial-lingual, vertical and mesial directions.

The mice were sacrificed by cervical dislocation following anesthesia at 10 weeks following the injection, and the inflammatory tissues of the mice were removed. Half of the tissues were embedded in paraffin (Hualin Kangfu, Shanghai, China) for sectioning, followed by immunohistochemical staining of the sections to visualize the expression levels of PD-L1 in the tissues. Single periodontal tissue cells of the inflammatory periodontium were isolated from the other half of the tissues by enzymatic digestion, and the expression levels of PD-L1 and PD-1 on the surface of the cells were analyzed using flow cytometry. Briefly, the tissues were minced with scalpels into sizes <1 mm³. The minced tissues were then digested with 2 mg/ml type I collagenase (Sigma-Aldrich) and 100 mg/ml DNase (Sigma-Aldrich) in PBS at 4°C overnight. Finally, the digested tissues were passed through a 200-mesh filter (Yangin Biological, Shanghai, China) to obtain single cells. The spleens of the mice were also removed and passed through the 200-mesh filter to obtain single cells. The cells were stained with rat PE-anti-mouse PD-L1 (3 µl per sample; dilution, 1;40; BioLegend, Inc.; cat. no. 124304) or APC-rat anti-mouse PD-1 antibodies (3 µl per sample; dilution, 1:40; BioLegend, Inc.; cat. no. 135209), and the expression levels of PD-L1 and PD-1 were determined using flow cytometric analysis.

Immunohistochemical analyses were performed, as described by Karlsson et al (28). The paraffin-embedded specimens were sliced into 4–5 µm-thick sections. The tissue sections were deparaffinized in xylene (Shanghai Macklin Biochemical Co., Ltd., Shanghai, China) for 3–10 min and dehydrated through a series of graded alcohols (100, 100, 95 and 80%) to displace the water. All sections were treated with 100% methanol containing 0.3% H₂O₂ for 15 min to block any endogenous peroxidase activity. The tissue sections were immersed in 50 ml 10 mM citrate buffer (pH 6.0) and placed in a microwave oven. Antigens were retrieved by microwaving in citrate buffer (pH 6.0) for three 6-min cycles. Bovine serum albumin (ZSGB-BIO) was used to block nonspecific IgG bindings. The sections were incubated with rabbit anti-mouse PD-L1 polyclonal antibodies (dilution, 1:200; Abcam, Cambridge, MA, USA; cat. no. ab58810) overnight at 4°C, followed by incubation with biotinylated goat anti-rabbit secondary antibodies (1:5,000; ZSGB-BIO, Beijing, China) for 30 min at room temperature, and with streptavidin-peroxidase complex solution for another 30 min at room temperature. The slides were stained with 3, 3′-diaminobenzidine (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). Finally, the sections were counterstained with hematoxylin and observed under an optical microscope (BX51TR; Olympus Corporation, Tokyo, Japan). Image-Pro Plus version 6.0 (Media Cybernetics, Inc., USA) was used for image analysis.

Statistical analysis was performed using SPSS 17.0 software (SPSS, Inc., Chicago, USA). All values are presented as the mean ± standard error of the mean. Data were analyzed by one-way analysis of variance, followed by Bonferroni's test. P<0.05 was considered to indicate a statistically significant difference.

In the present study, PDLCs were stimulated with five inflammatory cytokines that are present in the inflammatory microenvironment of the body, including IL-1, IL-6, TNF-α, IFN-γ and LPS, to examine the inducible expression of PD-L1. As shown in Fig. 1, the expression of PD-L1 was upregulated by IL-1, IL-6, TNF-α, IFN-γ and LPS, and by the combination of IL-1, IL-6, TNF-α and IFN-γ, compared with the unstimulated control.

The inflammatory cytokines secreted by immune and non-immune cells are known to be involved in the destruction of periodontal tissues, and are inducible upon bacterial infection (29). Therefore, the present study examined whether infection with periodontal bacteria induces the expression of PD-L1 on PDLCs. For this assessment, three strains of extensively investigated periodontal pathogens, including P.g, F.n and P.I, were selected for co-culture with the PDLCs for 18 h, followed by flow cytometric analyses of the surface expression of PD-L1. As shown in Fig. 2, the expression of PD-L1 was upregulated by P.g, F.n, and P.i, compared with the unstimulated control.

The present study further investigated the molecular function of PD-L1 on PDLCs by co-culturing PHA-activated PBMCs with PDLCs pretreated with TNF-α or IFN-γ, followed by two-color flow cytometric analysis. The TNF-α- and IFN-γ-induced expression of PD-L1 were shown to cause significant apoptosis of the activated CD4⁺ and CD8⁺ T lymphocytes. Pretreatment of the PDLCs with TNF-α led to increases in the percentages of apoptotic CD4⁺ and apoptotic CD8⁺ T cells, and the addition of IFN-γ resulted in increases in the percentages of apoptotic CD4⁺ and CD8⁺ T cells (Fig. 3A–D). The PI⁻ cells were gated to exclude necrotic cells (Fig. 3E). In addition, the TNF-α-pretreated PDLCs induced higher levels of apoptosis of the CD4⁺ and CD8⁺ T lymphocytes, compared with the IFN-γ-pretreated PDLCs. This was consistent with the more marked inducibility of TNF-α, compared with IFN-γ on the expression of PD-L1, as shown in Fig. 1A. To clarify the cause of the apoptosis, anti-PD-L1 antibodies were added to the cell co-culture at the time of mixing of the PHA-activated PBMCs with TNF-α-pretreated PDLCs. As shown in Fig. 3C and D, the percentages of apoptosis of the CD4⁺ and CD8⁺ T lymphocytes were reduced significantly, suggesting that the apoptosis of lymphocytes was correlated with the induced expression of PD-L1 on the PDLCs, and that PD-L1 may have negatively regulated the inflammatory responses. The increasing apoptosis of lymphocytes resulting from the upregulation of PD-L1 on the PDLCs inhibited the progression of excessive inflammatory immune responses, and thus reduced the destruction of the PDLCs.

To validate these initial observations of the present study, the cellular effect of the induced expression of PD-L1 on PDLCs was examined by tracking the viability of the PDLCs with CFSE and PI staining. The CFSE-stained PDLCs were co-cultured with PHA-activated PBMCs for 48 h, followed by staining with PI, resulting in a decrease in viable (CFSE⁺/PI⁻) PDLCs (Fig. 4A–C). However, the percentage of CFSE⁺PDLCs increased by 20%, compared with the untreated control when the PDLCs were pretreated with TNF-α to induce the expression of PD-L1 prior to co-culturing with the PHA-activated PBMCs (Fig. 4D),. In addition, anti-PD-L1 antibodies were added to the TNF-α-pretreated PDLCs and activated PBMCs co-culture at the time of mixing of the two cell cultures, to confirm the cause of cell survival. The percentage of viable PDLCs was reduced to 33.9%, which was not statistically different with that of the untreated control (Figs. 4E and F). These findings further confirmed the protective role of PD-L1 on PDLCs against inflammatory damage.

A mouse model of experimental periodontitis, exhibiting alveolar bone loss at the maxillary first molar was established in the present study. The expression of PD-L1 in the inflamed periodontium of the mice was analyzed using flow cytometry and immunohistochemistry. As shown in Fig. 5A, the mice with mild periodontitis expressed ~19% more PD-L1, compared with those with severe periodontitis, indicating that the presence of overexpressed PD-L1 may have ameliorated inflammation in the periodontal tissues, which resulted in periodontitis with less alveolar bone loss. The immunohistochemical staining of the corresponding tissue sections also indicated a considerable increase in the expression of PD-L1 in the periodontal tissues of the mice with mild periodontitis only (Fig. 5B and C). The animal experiments revealed a negative correlation between the expression of PD-L1 and the severity of destruction of the periodontal tissues (Fig. 5D). This is in accordance with the suggested protective role of PD-L1 in immunological damage.

As the PD-L1/PD-1 pathway is known to inhibit T cell-mediated immune responses (30), the present study also examined the expression levels of PD-1 in the periodontal tissues, peripheral blood and spleen of the periodontitis mouse model. Notably, no significant correlation was found between the expression levels of PD-L1 and PD-1 in the periodontal tissues (Fig. 5E and F) and other tissues (data not shown). This suggested the involvement of an alternative mechanism.