Ethics approval information

This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of Sichuan University. Animal procedures were performed according to protocols approved by the Animal Ethics Committee of Sichuan University (WCHSIRB-D-2020-476). All animal experiments followed the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. For collecting patient medical images, the study was approved by the institutional review board of West China Hospital of Stomatology, Sichuan University (WCHSIRB-D-2020-116), and all patients signed the informed consent. Patients or the public were not involved in the design, conduct, reporting, or dissemination plans of our research.

Power calculation

The number of mice per experimental group was based on a power calculation using data from our preliminary experiments. We found that a sample size of 5 mice per group would provide greater than 80% power to detect at least a 50% difference between ADD and Sham groups in RFP+ cell numbers in Gli1-CreERT+; Tmfl/− mice. To account for the possibility of a loss of mice due to death before study completion, we used an N of 5–8 for RFP+ cell number counting.

Animals

C57BL/6 wild-type male mice were purchased from Dashuo Experimental Animal Laboratories (Chengdu). Gli1tm3(creERT2)Alj/J mice (JAX#007913, abbreviated Gli1-CreER+), Tg(Sox9-cre/ERT2)1Msan/J mice (JAX#018829, abbreviated Sox9-CreER+) and B6.Cg-Gt(ROSA)26Sortm14(CAG-tdTomato)Haze/J mice (JAX#07908, abbreviated Tmfl/fl) were obtained from the Jackson Laboratory. Offspring were genotyped at 2 weeks old and conducted experiments at 4 weeks old. To generate tdTomato-conditionally activated mice, Tmfl/fl mice were crossed with Gli1-CreER+ mice and Sox9-CreER+ mice. Animals were housed in the Experimental Animal Center of the West China Hospital with free access to water and food at 25 °C with 40% humidity under a 12 h light/dark cycle. For ADD animal model and DR animal model generations described below, confounders between different groups were not controlled.

Generation of a mouse model with surgically induced ADD

The surgical method of ADD mouse model was modified based on previous studies described for rats and rabbits.38,39,40 Mice were anesthetized using i.p. injection with tribromoethanol at 0.12 μL·g−1 body weight (T903147, Macklin, China). The right preauricular region, as the surgical region, was shaved and sterilized with iodine and 75% ethanol and followed by local anesthesia with subcutaneous injection of 0.1 ml lidocaine hydrochloride (2%). A curved incision at approximately 1 cm in diameter was made parallel to the ears, extending from the temporalis muscle to the inferior edge of the masseter muscle. The overlying parotid gland was carefully exposed and dissected, and the superficial temporal vein was electro-coagulated. The zygomatic arch and TMJ capsule were exposed. The periosteum of the zygomatic arch was separated forward to the lower orbital margin and backward to expose the zygo-temporal suture. The lateral joint capsule was opened by a tweezer, the medial attachment was retained, and the articular disc and the lower condyle were exposed. A hole at 1 mm diameter was carefully drilled anterior to the front junction of the zygomatic arch using a 0.3 mm round bur. A round needle with a 6-0 nylon suture attached crossed the hole and was tied, then penetrated through the anterior band of the disc. The disc was pulled forward by the nylon suture. After confirming the disc was at the front of the condylar head, the suture was double-tied to the prepared hole in the zygomatic arch. After disc displacement, the wound was thoroughly irrigated with saline, and the anterior position of the disc was verified by about 1 mm deviation of the lower midline of teeth to the non-operative side. Finally, the wound was sutured in layers. Mice in the Sham group went through the same surgical procedure, including exposing the joint capsule, except drilling the hole in the zygomatic arch and pulling the disc forward. After ADD/Sham surgery, soft food was provided, and normal chow was provided one week later.

DR in ADD mouse model

ADD mice were randomly selected for disc repositioning surgery 5 days after the initial ADD surgery. After anesthesia by i.p. injection with tribromoethanol at 0.12 μL·g−1 body weight (T903147, Macklin, China), another incision was made along the previous incision, the TMJ capsule was exposed. The 6-0 nylon suture was carefully moved with the intact disc preserved. Then the displaced disc was released from adhesion and was pulled back on the top of the condyle. The muscle and skin were sutured. The re-Sham mice went through the second operation that exposed the TMJ capsule again without changing the disc position. The same postoperative care as initial surgery was provided.

Specimen preparation

Mice in ADD and Sham groups were randomly chosen and were sacrificed at 1, 2, 4, and 8 weeks after model generation. Mice in DR and re-Sham group were sacrificed at 2, 4, and 8 weeks after ADD surgery. After anesthetizing with isoflurane, 25 mL ice-cold heparin PBS (10 U·mL−1 heparin sodium in 0.01 mol·L−1 PBS) was injected transcardially. Next, 25 mL 4% PFA (4% paraformaldehyde in 0.01 mol·L−1 PBS, pH 7.4) was infused transcardially for tissue fixation. The intact TMJs in different groups were dissected and fixed at 4 °C for 24 h. For decalcification, Specimens were decalcified with 15% EDTA (15% ethylenediaminetetraacetic acid disodium salt solution, pH 7.5) for 7–21 days at 4 °C. Then the TMJ specimens were paraffin-embedded sagittally and were consecutively cut at 5 μm. For Gli1-CreER+ and Tmfl/fl mouse, the decalcified samples were embedded and frozen in optimum cutting temperature compound (OCT) after dehydrating by 30% sucrose solution overnight and cut into 6 μm thick sections.

Micro CT

Samples at 2, 4, and 8 weeks were selected for Micro CT scanning (Scanco, Switzerland) at 70 kVp, 200 µA with a spatial resolution of 7 µm. Scanco Evaluation software was used for bone analysis and three-dimensional reconstruction. Three cubes in size of 210 × 210 × 210 µm3 were selected in the anterior, middle, and posterior subchondral bone regions. 3D reconstruction of the mandible was performed to measure CL, width, and RH.41 The RH is measured by the vertical distance from the sigmoid notch to the midpoint of the condyle.

Histology and modified Mankin score

Hematoxylin and eosin staining (H&E, Biosharp, China) and Safranin O/ fast green (SO) staining (SO&FG, Solarbio, China) were performed according to the manufacturer’s instructions. The number of chondrocytes and thickness of cartilage were counted using H&E staining sections. The thickness of condylar cartilage was assessed three times at quarter points of the anterior, middle, and posterior cartilage in each region. The average thickness of cartilage was calculated as the mean value of nine data from each TMJ sample and ultimately used for statistical analysis (Fig. S8). The Safranin O+ area was calculated using ImageJ 1.51. Modified Markins score was used to assess the severity of cartilage arthritis, based on the number and morphology of chondrocytes, the safranin O positive region of cartilage matrix, and cartilage integrity, normal cartilage is scored 0-1 point. Totally, 2–4 points are assigned to mild OA. The Image 6.0 image analysis system (Media Cybernetics, Rockville, MD) was used for semiquantitative analysis.

Immunofluorescent staining

For immunofluorescent staining, the following antibodies were used: Aggrecan (1:100, 969D4011, Invitrogen), Collagen II (1:50, SC-52658, Santa Curz), SOX9 (1:200, ab185966, Abcam), RUNX2 (1:100, ET1612-47, Huaan). The secondary antibodies included goat anti-rabbit Alexa Fluor 488/ 568/ 647, anti-mouse Alexa 488/ 568 (1:500, Invitrogen). Apoptotic cells were detected by TUNEL staining (In Situ Cell Death Detection Kit, 11684795910, Roche’s) according to the recommended protocol. Slices were detected by laser scanning confocal microscopy (FV3000, Olympus).

Aggrecan+ and Collagen II+ area was calculated using ImageJ software. Three square frames (0.12 mm × 0.12 mm) in cartilage or subchondral bone were randomly selected to calculate the numbers of SOX9+ cells, RUNX2+ cells, and TUNEL+ cells.

Tartrate-resistant acid phosphatase (TRAP) staining

TRAP staining was conducted to detect osteoclasts based on the TRAP staining kit instruction (NO387, Sigma). Three regions of interest (0.15 mm × 0.15 mm) were selected from the subchondral bone in the anterior, middle, and posterior regions of the condyle. Purplish red cells containing more than one nucleus were counted into osteoclasts.

RNA sequencing analysis

The mRNA levels were detected by RNA sequencing. Fresh cartilage in condyles of 1w Sham group and ADD group mice were carefully removed and cleaned with 4°enzyme-free PBS. The cartilage tissues were removed completely and sent for analysis as far as possible. After Total RNA extraction by Trizol reagent (Thermofisher, 15596018), RNA was purified and reverse-transcribed to create the cDNA library that was sequenced and run with Illumina NovaseqATM 6000 sequence platform. Genes differential expression analysis was performed by DESeq2 software between two different groups and by edgeR between two samples. The genes with the parameter of false discovery rate (FDR) below 0.05 and absolute fold change ≥ 1.5 were considered differentially expressed genes.

Tamoxifen administration

Tamoxifen (T5648, Sigma) was suspended in corn oil (S50856, Yuanye) at 10 mg·mL−1. In the ADD and DR model generation experiment, Gli1-CreER+; Tmfl/- mice were injected intraperitoneally with tamoxifen 2 days before model generation (75 mg·kg−1 body weight per day, for 2 consecutive days). Sox9-CreER+; Tmfl/− mice were injected intraperitoneally with tamoxifen 7 days before model generation (75 mg·kg−1 body weight per day, for 3 consecutive days).

EdU assay

EdU (30 mg·kg−1) (A10044, Invitrogen) was injected i.p. 24 hours prior to sacrifice. EdU+ cells were detected with the Cell-LightTM EdU Apollo 488 In Vitro Kit (C10310-3, Riobobio) according to the recommended protocol.

Patients and study design

A retrospective study included 12 condyles (6 normal and 6 with ADDwoR) from adolescent patients with unilateral ADDwoR treated by disc repositioning surgery at West China Hospital of Stomatology, Sichuan University, was conducted from April 2021 to August 2022.

Inclusion criteria were (1) ≤18 years old; (2) diagnosed with unilateral ADD without reduction; (3) treated by disc repositioning surgery; (4) with complete preoperative and postoperative Cone-beam CT (CBCT) data of both TMJ.

Patients with syndromes, cleft palates, orthodontics treatment, and orthognathic surgery history were excluded. ADDwoR patients were diagnosed by experienced doctors according to the diagnostic criteria for temporomandibular disorders.

Imaging

CBCT data were collected 2–4 weeks before and 6–12 months after surgery. The scanning speed of the CBCT was 26.3 s, with 120 kV and 20 mA, and each layer thickness was set as 0.16 mm. Patients were in a seated position with the orbital-ear plane parallel to the horizon and asked to keep the eyes forward, intercuspation, and tongue position during the examination.

3D image measurements

The scan data were stored in the format of “.dicom”, which was imported into and constructed virtual 3D surface models of the anatomic condylar structures of interest with Mimics 21 software (Materialize, NV, Belgium). The 3D measurement started from condylar image segmentation and reconstruction based on previous studies.42,43 The superior contour of the condyle was determined as the first radiopaque point found in the upper articular area. The lowest point of the sigmoid notch is the marker point of the bottom of the condyle. The 3D condylar image was reconstructed after the isolation, and the condylar volume was measured. Condylar height was measured as the vertical distance from the apex of the condyle to the largest cross-section of the reconstructed condyle in the sagittal view, and CL was analyzed as the distance between the most anterior and posterior points of the condylar head.

Statistical analysis

The mean and standard deviation of data were analyzed by GraphPad Prism 9 analysis software. Two-way analysis of variance (ANOVA) was used to compare the data of the two groups at multiple time points, a P value < 0.05 was considered to be statistically significant. The staining results were analyzed by ImageJ 1.51 (Leeds Precision Instruments, USA) by an individual analyzer using blind sections.