Traumatic temporomandibular joint bony ankylosis in growing rats

Sprague–Dawley rats were provided by the Laboratory Animal Center of the Fourth Military Medical University (Xi'an, China) and all experiments were approved by the animal welfare ethics committee of the School of Stomatology [36] (Approval ID 2020–0950). The calculated sample size was 24 (PASS 11.0, Test for one correlation power analysis: Power = 0.9, α = 0.05, R = 0.74). Twenty-four male Sprague–Dawley rats at 3 weeks old (weight 60–80 g) were obtained from the Animal Center of the Fourth Military Medical University.

Each Sprague–Dawley rat underwent general anesthesia with 2% isoflurane and oxygen. The TMJ region was isolated with sterile drapes. The temporal and preauricular regions were shaved and disinfected. In the operation group, the right TMJ underwent a sham operation (sham side) whereby a curved preauricular incision was made and the TMJ complex was exposed, then the wound was closed in layers. For the left TMJ of the operation group, the bony ankylosis-induced side (experimental side), a 1 cm long curved preauricular incision was made and the flap was lifted to expose the TMJ complex (Fig. 1A). The capsule was exposed by blunt dissection (Fig. 1B). The condyle was isolated with a periosteal elevator, and the articular disc was exposed by the separation of the lateral attachment of the disc through the joint space (Fig. 1C). The anterior and posterior attachments of the disc were then cut off and the articular disc was removed (Fig. 1D). The joint space was exposed by horizontal blunt dissection (Fig. 1E), the fibrocartilage on the condyle was removed using a Gracey scaler (Fig. 1F) and grid grooves were carved using an apex elevator on the surface of the glenoid fossa (Fig. 1G). The joint space was then narrowed (Fig. 1H), the capsule was sutured, and the wound was closed in layers (Fig. 1I). After surgery, each rat was administered an antibiotic (penicillin, 2 mg/100 g; X–Y Biotechnology) and an analgesic (pentazocine, 0.1 mg/100 g; X–Y Biotechnology) for 5 days. The control group did not undergo any operations.

The vertical passive maximum mouth opening (PMMO) and mandibular lateral movement were measured and recorded at 1, 4, and 8 weeks postoperatively according to our previous study [3]. Vertical PMMO was measured under general anesthesia as the distance between the maxillary and mandibular incisors upon the perpendicular application of 1.0 g of force to the jaw by a ruler scale (Fig. 1J, K). The force of 1.0 g was chosen based on pilot experiments which indicated that this amount of force was sufficient for the Sprague–Dawley rat jaws to reach the end-feel of ‘capsular feel’ or ‘springy-block’ [3]. Mandibular lateral movement was also measured and recorded under general anesthesia according to our previous study [18] (Fig. 1M, N).

All rats were euthanized with a lethal dose of pentobarbitone sodium by intraperitoneal injection at 1, 4, and 8 weeks postoperatively. All TMJ complexes were fixed in 4% paraformaldehyde. Then, TMJ complex samples were rinsed for 6 h with water and then decalcified in 17% EDTA for 6 weeks before embedding in paraffin. 4-μm sections were cut along the mesiodistal direction of the dentition.

All Sprague–Dawley rats underwent general anesthesia at 1, 4, and 8 weeks postoperatively. Their skulls were scanned by Micro-CT (Siemens AG) at a voltage of 80 kV, a current of 500 μA, and a resolution of 10 μm. Subsequently, three-dimensional (3D) images of TMJ complexes were reconstructed with the Inveon Research Workplace (Siemens AG). In the operation group, the region of interest (ROI) included three cubes (each 0.5 × 0.5 × 0.5 mm), which were selected along an imaginary line between the concave point of the condylar lateral neck and the upper edge of the zygomatic arch base. The morphological parameters of trabecular bone microarchitecture, including bone volume fraction (BV/TV), trabecular thickness (Tb.Th), bone specific surface (BS/BV), trabecular separation (Tb.Sp) and trabecular number (Tb.N), were measured.

The paraffin sections were stained with Hematoxylin and Eosin (H&E) and modified Safranine O-Fast Green FCF Cartilage Stain (Solarbio, G1371), according to the manufacturer's instructions. Slides were preincubated with 3% H₂O₂ for 10 min. Goat serum was used to block nonspecific binding, and then sections were incubated overnight at 4 °C with Coll II (1:1000, Abcam), MMP13 (1:250, Proteintech), RUNX2 (1:500, SANTA). Subsequently, sections were incubated with the corresponding secondary antibodies at room temperature for 1 h.

Total RNA extracted from TMJ complexes was used for RNA sequencing by Beijing Genomics Institute (ShenZhen, China). The samples were sequenced at BGI (Cambridge, MA) with paired-end 100 bp in the BGISeq-500 platform. Each group had three replicates, and all sequence data were deposited in the Gene Expression Omnibus database. Pathway analysis was performed using the DAVID functional annotation bioinformatic tool.

Data are presented as mean ± SD and were analyzed with SPSS 18.0 software (SPSS Inc, Chicago, IL, USA). The values of the control and experiment were compared at each time point. Micro-CT parametric data were compared by paired t-test. The vertical PMMO and the mandibular lateral movement between the control and operation groups were compared by independent t-test. P < 0.05 was considered statistically significant.

All Sprague–Dawley rats could tolerate the compound trauma in the operation group, and no postoperative wound infection occurred. The vertical PMMO of the operation group was significantly lower than that of the control group at 4 and 8 weeks postoperatively (Fig. 1J, K). There was statistical significance in the vertical PMMO between the control and operation groups at 8 weeks postoperatively (Fig. 1L, P < 0.05). At 4 and 8 weeks postoperatively, the mandibular lateral movement of the operation group was significantly greater than that of the control group (Fig. 1M, N), and the difference was statistically significant between the two groups (Fig. 1O; Post-OP 4 W, P < 0.01; Post-OP 8 W, P < 0.001). At 8 weeks postoperatively, a clear limitation of vertical PMMO was observed in rats of the operation group.

On the experimental side, gross observations were made of the TMJ external appearance with a stereoscopic microscope at 8 weeks postoperatively (Fig. 2A-C). TMJ bony ankylosis and irregular new bone formation on the TMJ complex surfaces were observed (Fig. 2C). Moreover, bony adhesions were observed in the joint space of the experimental side (Fig. 2B, C; white dotted line).

On the sham side, bony ankylosis was absent (Fig. 2D-F). The glenoid fossa and articular cartilage did not exhibit fusion.

The microarchitectural data revealed that the image of TMJ bony ankylosis was observed on the experimental side but not on the sham side.

On the experimental side, some new cartilage and bone were observed at 1 week postoperatively. There was moderate new cartilage and bone on the experimental side at 4 weeks postoperatively and joint space was narrower at 4 weeks than that at 1 week postoperatively. At 8 weeks postoperatively, TMJ bony ankylosis was found on the experimental side, whereby the joint space had nearly disappeared and was filled with abundant new cartilage and bone (Fig. 3A; white arrow) and the degree of calcification increased over time. The morphological parameters of the trabecular bone microarchitecture were calculated (Fig. 3B-F). The volume of TMJ complexes was 51.54 ± 3.584 mm³ at 8 weeks postoperatively (Fig. 3G).

On the sham side, TMJ regular upper and lower articular surfaces were observed, however TMJ bony ankylosis was absent. There was no new cartilage or bone in the joint space of the sham side at 1, 4, and 8 weeks postoperatively. The image of the joint space showed that it was slightly narrowed. Similarly, the morphological parameters of the trabecular bone microarchitecture were calculated. (Fig. 4B-F). The volume of TMJ complexes was 9.827 ± 0.716 mm³ at 8 weeks postoperatively (Fig. 4G).

BV/TV (P = 0.014) and Tb.Th (P = 0.172) of the experimental side TMJ complexes were significantly increased than that on the sham side at 8 weeks postoperatively (Fig. 3B, E). BS/BV (P = 0.033) and Tb.Sp (P = 0.258) of the experimental side TMJ complexes increased more at 1 week postoperatively than they did on the sham side (Fig. 3C, F). Tb.N (P = 0.811) of the experimental side was lower than that on the sham side at 1 week postoperatively (Fig. 3D). The volume difference of TMJ complexes was statistically significant between the experimental and sham sides at 4 and 8 weeks postoperatively (Fig. 3G) (Post-OP 4 W, P = 0.004; Post-OP 8 W, P = 0.000), There was no significant difference between the other parameters between the two sides at other times.

All TMJ complexes on the experimental and sham sides were verified histologically (Fig. 4A-D).

On the experimental side, H&E staining showed a small amount of new cartilaginous matrix and cartilage cells, and a large amount of fibrous tissue in the joint space at 1 and 4 weeks postoperatively (Fig. 4A). Bony ankylosis was observed in the joint space, which was abundantly filled with new cartilaginous matrix, cartilage cells, endochondral ossification, and a small amount of fibrous tissue at 8 weeks postoperatively (Fig. 4A, B). Safranin O staining was used to evaluate the amount of red glycosaminoglycan (GAG) in the cartilage of TMJ complexes (Fig. 4C). A small amount of GAG was observed in the joint space at 1 and 4 weeks postoperatively (Fig. 4 C; black arrow), however GAG was abundant and distributed irregularly in the joint space at 8 weeks postoperatively, when endochondral ossification and cartilage margin ossification proceeded simultaneously (Fig. 4C; white arrow). IHC staining for Col II showed that the Col II-positive cells contributed 16.5%, 34.2%, and 44.3% of all Col II cells at 1, 4, and 8 weeks after the operation, respectively.

On the sham side, bony ankylosis and endochondral ossification were absent in the joint space, there was no statistical significance in GAG and Col II expression at 1, 4, and 8 weeks postoperatively.

RNA-sequencing was performed for dissected TMJ complexes on both the experimental side and sham side. Whole TMJ-complex RNA-Seq analysis identified transcripts corresponding to 37,246 genes. A volcano plot identified both upregulated and downregulated genes, while the biological coefficient of variation was between 0.4 and 0.5, within an acceptable range variation (Fig. 5A). The x-axis is the log2 scale of the fold change of gene expression between the two sides (log2 (fold change)). Negative/positive values indicate downregulation/upregulation. The y-axis is the minus log10 scale of the adjusted p values (-log10), which indicates the significant differences in the level of expression. The red dots represent significantly upregulated genes with at least a twofold change, while the blue dots represent significantly downregulated genes with at least a twofold change (Fig. 5A).

At 8 weeks postoperatively, differentially expressed genes (DEGs) were enriched in 101 terms, including 56 biological process (BP) terms, 16 cellular component (CC)terms and 29 molecular function (MF) terms. DEGs with endochondral ossification primarily included MMP13 and RUNX2. MMP13-related terms were as followed: in the BP domain, the meaningful enriched GO terms focused on extracellular matrix organization (GO:0,030,198), ossification (GO:0001503), proteolysis (GO:0006508), endochondral ossification (GO:0001958), and osteoblast differentiation (GO:0001649) (Fig. 5B). The most enriched CC terms were correlated with the extracellular space (GO:0,005,615). The MF terms were closely related to the extracellular matrix structural constituent (GO:0005201), and platelet-derived growth factor binding (GO:0048407). RUNX2-related terms were as followed: in the BP domain, the meaningful enriched GO terms focused on endochondral ossification (GO:0001958), ossification (GO:0001503), osteoblast differentiation (GO:0001649), protein processing (GO:0016485), and skeletal system development (GO:0001501) (Fig. 5B). RUNX2 was not enriched in the CC or MF terms. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of the 101 DEGs identified 20 pathways, which are shown a bubble diagram. The KEGG pathways with endochondral ossification that were enriched at 8 weeks after the operation primarily included Parathyroid hormone synthesis, secretion and action, the Relaxin signaling pathway, and the IL-17 signaling pathway (Fig. 5C). We used a heatmap to show the distribution of DEGs between the experimental and sham sides. Red and blue stripes represent genes with high or low expressions, respectively (Fig. 5D). MMP13 and RUNX2 showed a significant difference between the experimental and sham sides (Fig. 5D; black arrow).