Low-level laser therapy prevents medication-related osteonecrosis of the jaw-like lesions via IL-1RA-mediated primary gingival wound healing

Five MRONJ patients, diagnosed according to AAMOS criteria [1], and five healthy controls were included in this study. MRONJ gingival samples and healthy gingival samples were obtained during the MRONJ-related surgeries or orthopedic surgeries at Peking University School and the Hospital of Stomatology. The collection process was standardized. Gingiva samples around the tooth were collected from patients in both groups. Specifically, the gingival samples of the MRONJ group were collected around the MRONJ-involved tooth, and the control gingival samples were collected around the healthy tooth during orthopedic surgeries. The harvested tissues were fixed in 4% paraformaldehyde and subjected to further analyses.

The present study was approved by the Institutional Review Board (IRB) of the Peking University Hospital of Stomatology (PKUSSIRB-202170184). All enrolled individuals provided written informed consent. The clinical features of healthy controls and MRONJ patients of the study population are shown in Additional file 1: Table S2.

Zoledronate (Zol) was purchased from Sigma. An anti-OCN antibody was purchased from Abclonal. Anti-IL-1β, anti-TNF-α, and anti-IL-1RA antibodies were acquired from Abcam. Alexa Fluor 488 secondary antibody was purchased from Abcam. IHC-secondary antibodies were bought from ZSGB-BIO. IL-1 RA neutralizing and negative control antibodies were purchased from R&D Systems. IL-1β ELISA kit was purchased from Biolegend. Masson’s trichrome stain kit was obtained from Solarbio. Tartrate-resistant acid phosphatase (TRAP) stain kit was purchased from Sigma-Aldrich. Cell counting kit-8 was purchased from Dojindo.

A 6–8-week-old male C57BL/6N mice were acquired from the Beijing Vital River Laboratory Animal Technology Co., Ltd. All the animals were housed in a specific pathogen-free facility with a relative humidity of 50 ± 1%, a temperature of 22 ± 1 °C, and a light/dark cycle of 12/12 h and were given standard chow and water. All animal experiments were approved by the Ethics Committee of the Peking University Health Science Center (LA2018265) and carried out in compliance with ARRIVE guidelines.

In order to reflect the similar clinical disease state of MRONJ and further explore its mechanism, we built the MRONJ mouse model and compared the mouse gingival samples with human gingival tissues. Mice were randomly divided into the following groups (n = 5/group): (1) Ctrl group, which received a vehicle and no local treatment on the extraction site; (2) Zol group, which received a zoledronate and no local treatment on the extraction site; (3) Zol + LLLT group, which received a zoledronate and LLLT on the extraction site.

To build an MRONJ-like mouse model, mice were administered with zoledronate (500 μg /kg, SML0223, Sigma, United States) intraperitoneally twice a week for four weeks according to our preliminary animal experiments and previous studies [21]. After two weeks of drug treatment, tooth extraction of bilateral maxillary second molars was performed under general anesthesia using pentobarbital (50 mg/kg, P3761, Sigma, USA). The ctrl group or the Zol group received no local treatment for tooth sockets after tooth extraction. For local treatment, LLLT was locally employed to the tooth sockets immediately post extraction. In the subsequent rescue animal experiments, mice were randomly divided into four following groups (n = 5/group): (1) Ctrl group, which received a vehicle and negative control antibodies on the extraction site; (2) Zol group, which received a zoledronate and negative control antibodies on the extraction site; (3) Zol + LLLT group, which received a zoledronate, LLLT, and negative control antibodies on the extraction site; (4) Zol + LLLT + IL-1RA NAb group, which received a zoledronate, LLLT, and interleukin-1 receptor antagonist neutralizing antibody (IL-1RA NAb) on the extraction site. After tooth extraction and/or LLLT treatment, IL-1RA NAb (10 μg per mouse, AF-480-NA, R&D, USA) was topically injected into the extraction sockets using a micro sampler (91234910, Solarbio, China) in the Zol + LLLT + IL-1RA NAb group. Negative control antibodies (10 μg per mouse, AB-108-C, R&D, USA) were topically injected into the extraction sockets using a micro sampler for the Ctrl group, Zol group, and Zol + LLLT group. The local treatment was repeated 2 and 4 days after the tooth extraction. At two weeks post-extraction, all mice were euthanized by cervical dislocation. Untreated healthy mice were euthanized as natural healing control. The maxillae were collected and fixed in 4% paraformaldehyde, after which the fixed maxillae were subjected to further analyses.

We used a low-level laser (Velure S9 980 – Lasering) with the main wavelength of 980 nm to treat the tooth extraction. Three LLLT sessions were performed at the tooth extraction site at 0, 2, and 4 postoperative days. The irradiation parameters were as follows: 0.5 W power; spot size of 0.007829 cm²; continuous operation mode; power intensity of 63.87 W/ cm². Irradiation was employed at a single point with the laser tip 5 mm above dental extraction sites for 45 s.

The fixed maxillae were scanned using micro-computed tomography (microCT) (60 kV, 2 mA, J. Morita Corp., Kyoto, Japan) to assess hard tissue repair in the tooth extraction sockets. The area of interest was chosen, and bone mineral density (BMD) and bone volume/total volume (BV/TV) were quantified by three-dimensional (3D) bone morphometric software (SIEMENS, Munich, Germany). Data were analyzed using Image J.

The fixed maxillae, including the extraction sockets, were decalcified in 10% EDTA, paraffin-embedded, and sectioned using a microtome (4 μm thick slices). Also, the fixed human gingival samples were sectioned using a microtome (4 μm thick slices). Hematoxylin and eosin (H&E) staining was then conducted for histological observations of the gingiva and tooth extraction sites (Solarbio, China). Masson’s trichrome staining was conducted to observe the collagen fibers following the instructions (Solarbio, China). TRAP staining was conducted to indicate osteoclasts following the instructions (Sigma-Aldrich, USA).

Immunochemistry-paraffin (IHC-P) staining and immunofluorescent Staining (IF) were conducted to study LLLT-induced primary gingival wound healing employing the following primary antibodies: anti-OCN (A6205, Abclonal); anti-IL-1β (ab9722, Abcam); anti-TNF-α antibodies (ab1793, Abcam), anti-IL-1RA (ab124962, Abcam). For IHC-P staining, sections were deparaffinized, rehydrated, and inactivated endogenous peroxidase activity with 3% H₂O₂ in dark for 20 min. Then, sections were incubated in 0.01 M sodium citrate buffer solution at 98 °C for 20 min. After cooling to room temperature, sections were incubated with anti-OCN, or IL-1β or anti-TNF-α antibodies overnight at 4 °C. Then, sections were stained with secondary antibodies (PV-9001, ZSGB-BIO or PV-9002, ZSGB-BIO) and a DAB kit (ZLI-9018, ZSGB-BIO).

For IF staining, sections were handled as before and incubated with anti-IL-1RA antibodies overnight at 4 °C, after which they were stained with Alexa Fluor 488 secondary antibodies (1:200, Abcam) for 1 h and counterstained with DAPI (ZLI-9557, ZSGB-BIO). Images were acquired with an Olympus microscope. Images were analyzed using Image J.

A high-throughput tissue crusher was used to lyse gingival tissue into a paste. RNAs were extracted from mice gingival tissues with TriZol Reagent (15,596,026, Invitrogen, Thermo Fisher Scientific, USA) and reverse-transcribed (Takara, Japan) into complementary DNA (cDNA). The resultant cDNAs were used for the following experiments. qRT-PCR was applied to detect the gene expressions with the ABI Prism 7500. The relative expression level of targeted genes was quantified with β-actin (used as the internal control) and calculated with the 2^–ΔΔCT method. Additional file 1: Table S1 shows all of the primer sequences.

A high-throughput tissue crusher was used to lyse gingival tissue into a paste. Tissue supernatant of mice gingival samples were collected. IL-1β in gingival tissues was detected with ELISA kits (432604, Biolegend) under the manufacturer’s instructions. Then, a microplate reader measured the absorbance at 450 nm (Elx808; BioTek).

HaCaT cells, obtained from the central laboratory of Peking University Hospital of Stomatology, were cultured in DMEM (Dulbecco’s Modified Eagle’s Medium, Sigma-Aldrich, USA) supplemented with 1% penicillin − streptomycin (Gibco, USA) and 10% fetal bovine serum (Gibco, USA) in a humidified atmosphere at 37^◦C in 5% CO₂. The effect of Zol on HaCaT cells was examined using the cell counting kit-8 (CCK-8) assay (Dojindo, Japan). HaCaT cells were seeded in 96-well plates at a density of 2 × 10³ cells per well and incubated with 200 μL of the complete medium overnight to allow them to attach. Then, the medium was replaced with a fresh medium containing 1, 5, 10, 25, 50 μM Zol. After 24 h, 10 µL CCK-8 solution was added to each well and incubated for 2 h, and the absorbance of each well was measured at 450 nm absorbance. The effect of LLLT on the migration of HaCaT cells treated with Zol was assessed using the migration assay. In six-well plates, HaCaT cells were seeded and divided into the following groups: (1) Ctrl group: no drugs treatment; (2) Zol group: cells were incubated with 10 μM Zol for 24 h; (3) Zol + LLLT group: cells were incubated with 10 μM Zol for 24 h and then received LLLT (Velure S9 980 – Lasering, 980 nm, 0.5 W delivering energy doses at 0.5 J/cm² for 10 s). After incubation with 10 μM Zol for 24 h, a sterile 10-μl pipet tip was used to scratch three separate wounds through the cells. The cells were gently rinsed in PBS to remove floating cells and incubated in the medium at 37 °C, 5% CO₂/95% air environment. Images of the scratches were taken using an inverted microscope (Olympus, Lake Success, NY) at 0, 24, 48, and 72 h of incubation. The wound closure percentage was quantified using Image J.

Experiments were performed according to the schedule (Additional file 1: Fig. 1A). At 2 weeks post-extraction, stereoscope observations revealed that the tooth socket wounds were completely healed; consecutive mucosal covering was observed in both Ctrl and Zol + LLLT groups. At the same time, incomplete mucosal healing, lack of epithelial coverage, and exposed bone at the tooth socket wounds were seen in the Zol group (Additional file 1: Fig. 1B). MicroCT results further showed no obvious new bone regeneration at tooth-extraction sockets in the Zol group, while the Ctrl group and the Zol + LLLT group both showed ample newly formed bone (Additional file 1: Fig. 1C). The quantified data of microCT results indicated that BV/TV and BMD in alveolar sockets were significantly higher in the Ctrl group and Zol + LLLT group than in the Zol group (P < 0.05, Fig. 1D,E).

Histologically, the HE staining results showed intact epithelial coverage and adequate bone regeneration in sockets in the Ctrl and LLLT groups. In the Zol group, defective epithelial lining and necrotic bone with empty lacunae were observed (Fig. 1F and Additional file 1: Fig. S1A). In addition, Masson staining results showed more collagen deposition in the extraction sockets in the Zol + LLLT group compared to the Zol group (Fig. 1H). According to the results of TRAP staining, the Zol group had significantly fewer osteoclasts per bone marrow area (#/mm²), while LLLT application increased osteoclasts formation (Additional file 1: Fig. S1B and S1C). Additionally, immunohistochemical staining indicated significantly increased osteocalcin (OCN) expression in the Zol + LLLT group than in the Zol group (P < 0.05, Additional file 1: Fig. S1D and S1E).

Moreover, MRONJ patients had more inflammatory cell infiltration in the gingival tissue compared with healthy controls (Fig. 1G). Masson staining results showed that collagen arrangement was more disordered in MRONJ patients compared with healthy controls (Fig. 1I).

To investigate the mechanism of how LLLT promoted gingival wound healing, mice gingiva was harvested at 11 days post-extraction. The qPCR results showed that IL-1β and IL-6 expression in the gingival tissues were downregulated, while IL-1RA expression was more upregulated in the Zol + LLLT group than in the Zol group (Fig. 2A–C). The results of ELISA further showed that LLLT decreased IL-1β expression in the gingival tissue compared with the Zol group (Fig. 2D). Immunohistochemistry indicated that IL-1β and TNF-α in the gingival tissue were upregulated in the Zol group compared with the Ctrl group. However, these factors were downregulated in the LLLT group (Fig. 2E). In addition, patients’ gingival samples of MRONJ lesions showed higher expression of IL-1β and TNF-α than the healthy controls (Fig. 2F). Immunofluorescence also indicated that IL-1 RA was highly expressed in the LLLT group compared with the Zol group (Fig. 2G).

A previous study has suggested that IL-1RA is important for wound healing [22]. Herein, we found increased IL-1RA expression in MRONJ lesions after LLLT. Then, we examined whether IL-1RA has a major part in LLLT promoting gingival wound healing. LLLT was performed on the tooth extraction sockets with or without IL-1RA NAb. The in vivo results showed that IL-1RA NAb abolishes the ability of LLLT to promote wound healing. Experiments were performed following the schedule (Fig. 3A). Stereoscope observations showed incomplete mucosal healing in the Zol + LLLT + IL-1RA NAb group compared with the Zol + LLLT group (Fig. 3B). MicroCT analysis further showed decreased bone formation, BMD, and BV/TV in the Zol + LLLT + IL-1RA NAb group compared with the Zol + LLLT group (P < 0.05, Fig. 3C–E). Histological analysis revealed that wound healing was delayed, and bone formation at the extraction site was reduced by using IL-1RA NAb (Fig. 3F and Additional file 1: Fig. S2A). The IL-1β and TNF-α expression increased again in the mouse gingiva in the Zol + LLLT + IL-1RA NAb group compared with the Zol + LLLT group, confirmed by immunohistochemistry (Fig. 3G, H). The above results suggest IL-1RA is involved in LLLT-induced primary gingival wound healing and MRONJ prevention.

A previous study has shown that BPs impede soft tissue wound healing by inhibiting epithelial cell proliferation and migration [23]. To detect the effects of Zol on HaCaT cells, we treated HaCaT cells with different concentrations of zoledronate, 1 μM, 5 μM, 10 μM, 25 μM, 50 μM, and found that 10 μM was the lowest concentration, at which HaCaT cells proliferation was inhibited at 24 h (P < 0.05, Fig. 4A). Thus, we chose 10 μM Zol for the migration assay. The migration assay demonstrated that Zol slows down the HaCaT cells’ migration rate, whereas LLLT significantly accelerates the migration rate at 24 h, 48 h, and 72 h (P < 0.05, Fig. 4B and 4C).