2.1 Root repair model protocol and CTGF local administration
The establishment of a root repair model was consistent with our previous study[29]. 36 6-week-old male SD rats around 200 ± 10g weight were selected in our study. The experimental design was ratified by the Ethics Committee of West China Stomatological Hospital (WCHSIRB-D-2021-601). Under anesthesia, an orthodontic nickel-titanium spring was applied between the left maxillary first molar and incisor with 100 g force (Supplemental Fig.S1a), and OIRR was successfully established after two weeks with heavy force loading (Supplemental Fig.S1 b). After two weeks, the nickel-titanium tension spring was replaced by a ligature wire to achieve the retention effect. Detailed protocols are shown in Supplemental Fig.S1c. To calculate the volume of root resorption after 2 weeks, 6 rats in each group were sacrificed at this instant as initial controls (Day 0). Without orthodontic force, the root which impaired by heavy orthodontic forces initiates the biological activity of cementum regeneration. The rest of the rats were stochastically selected into the control group and CTGF-treated group. The CTGF-treated rats received 0.5ug of CTGF, which was injected every two days locally at the buccal gingiva of the left maxillary first molar, whereas the control group only received with a 0.1% acetic acid solution injected locally. Rats were sacrificed on the 14th (Day 14) and 28th (Day 28) days (n = 6), respectively, and the left maxilla was isolated for the following experiments.
2.2 Microcomputed tomography analysis
The samples were fixed in 4% paraformaldehyde (PFA) for 24 h, and then examined by the high-resolution (10 µm) Micro-CT50 system (Scanco Medical, Wangen-Bruttisellen, Switzerland). The detailed analysis of root resorption was consistent with our previous study.[29] The 3D model of the left maxillary first molar was reconstructed using Mimics 21.0 software to observe the absorption craters (Supplemental Fig. S1d, e). The blue-marked area represents the volume of the existing root, while the pink pseudo-color indicates the resorption root area which was manually delineated in multiple 2D images by the Mimics software according to the morphology of the resorption root (Supplemental Fig. S1f). The red area of the supplemental Fig.S1g indicates the area of cementum resorption. For test subjects, we divided the volume of the absorption craters by their root length to obtain resorption volume per millimeter of root length.
2.3 Histological and immunohistochemical (IHC) analysis
After Micro-CT, the left maxillary jaw was decalcified with 14% EDTA (pH 7.4) for 1.5 months, and then dehydrated in ascending graded alcohol and embedded in paraffin, and then cut into 5µm serial sections. Root resorption morphology was observed by H&E (G1340; Solarbio, Beijing, China). Immunohistochemistry was taken after deparaffinization in xylene and then rehydration in descending graded alcohol. Then the specimens adopted heat-induced epitope retrieval. After incubation with 5% BSA, sections were incubated overnight with anti-Col1(1:200,501352, Zen-Bio, China, anti-Osterix (1:200, Huabio, ER1914-47, China), anti-Runx2 (1:200, 860139, Zen-Bio, China), anti-CAP (1:200, sc-53947, Santa, America), anti- Cx43 (1:200, 340279, Zen-Bio, China), anti-active β-catenin (1:200, 19807, Cell Signaling Technology, America), following the incubation of the corresponding IgG secondary antibodies. The binding of the avidin-biotin-peroxidase complex to the DAB substrate Kit (ZLI-9017; Zhongshan BioTech, China) visualizes the positive immune signal.
2.4 Isolation, culture, and Identification of hPDLSCs
The detailed cell isolation protocol has been described in the previous study.[30] After obtaining the informed consent of the subjects, the periodontal ligament tissues of the premolars or third molars of 6 patients were collected. hPDLSCs were cultured as previously described, and passages of 3rd-5th periodontal ligament cells were used for subsequent experiments.[30] Alkaline phosphatase, alizarin red and crystal violet staining were used to detect the multi-directional differentiation ability. Flow cytometry was performed using primary antibodies stro1, CD146, CD34 and CD45 (Proteintech Group, USA) by the BD Accuri™ Cflow cytometer (BD Biosciences, Milan, Italy).
2.5 RNA Interference of Cx43 and CTGF
The RNA interference protocol was consistent with the previous study.[25] hPDLSCs were seeded in 24-well, 48-well and 6-well culture dishes, then transfected with corresponding siRNA (hanbio, China) and transiently transfected with Lipofectamine RNAiMAX (Invitrogen, Burlington, ON, Canada). Specific siRNA sequences are shown belows: Si-Cx43 (forward 5'- GCGACAGAAACAAUUCUUC-3' and reverse 5'- GAAGAAUUGUUUCUGUCGC-3'); Si-CTGF (forward 5'- GCACCAGUGUGAAGACAUATT − 3' as well as reverse 5'- UAUGUCUUCACACUGGUGCTT − 3'). Western bolt was performed to confirm transfection efficiency.
2.5 Alkaline Phosphatase (ALP) Staining
hPDLSCs were seeded in 48-well culture dishes, and cultivated for 7 days in mineralization medium (ascorbic acid (0.05M), dexamethasone(100nM), β-glycerophosphate (10mM)) with CTGF in an ascending concentration (0,10, 20, 50ng/ml). After fixation with 4% PFA, hPDLSCs were applied with BCIP/NBT detection kit (C3206, Beyotime, China) and fully stained for 15 minutes, then rinsed with water and stored for observation under the light microscope. To verify whether Cx43 is involved in CTGF-mediated cementogenic differentiation, we applied CTGF and Si-Cx43 together to hPDLSCs and fixed them for ALP staining seven days later. hPDLSCs were treated with the Cx43 agonist All-Trans-Retinoic Acid (ATRA) (R2625, Sigma, America) and Si-CTGF Immediately afterward, and then fixed for ALP staining seven days later respectively.
2.6 Alizarin Red Staining (ARS)
The detailed hPDLSCs treatment protocols were in accordance with the ALP staining experiments above. hPDLSCs were seeded in 24-well culture dishes, and then cultured in mineralization medium for 14 days with the similar treatment protocol described above. The hPDLSCs were dyed with 1% ARS (G1452, Solabio, China) for 30 min to evaluate the degree of mineralization, and then bathed with double distilled water for 3 min. The images were observed under a stereoscopic imaging microscope.
2.7 Western blotting (WB)
The detailed hPDLSCs treatment protocols were in accordance with the ALP staining experiments above. Total proteins were extracted by protein lysis buffer (BL504A, Biosharp, China) containing phosphatase inhibitors following the manufacturer's instructions. After denaturation, the proteins were separated by 10% SDS-polyacrylamide gel electrophoresis and transferred to PVDF membranes. Sealed with 5% skim for 1 hour, proteins were incubated with the corresponding primary antibody (1:1000) overnight: Rabbit anti-Col1; Rabbit anti-Osterix; Rabbit anti-Runx2; Mouse anti-Cap, Rabbit anti-Cx43; Rabbit anti-CTGF (R24001, Zen-Bio, China); Mouse anti-β-catenin (ab237983, Abcam, UK); Rabbit anti-active-β-catenin, following the incubation with the corresponding secondary antibody and reaction with the Western Blotting Reagent Kit (P90719, Millipore). Image-J software was applied to evaluate the intensity of protein bands (Media Cybernetics, Bethesda, MD, United States).
2.8 Quantitative RT –PCR
The detailed hPDLSCs treatment protocols were in accordance with ALP staining experiments above. RNA was collected in accordance with the instructions of the RNA Isolation Kit (Bioteke, Beijing, China.), and then 1 µg RNA was transformed into cDNA with reverse transcriptase (Thermo Fisher Scientific, MA, USA) and subjected to quantitative RT-PCR reaction (Bio-Rad, CA, USA). The relative expression of targeted genes (Table. S1) was calculated applying the ΔΔCt method as previously described.[30]
2.9 Immunofluorescence and Confocal Laser Scanning Microscope (CLSM)
The detailed immunofluorescence procedure was described formerly[25]. The hPDLSCs were fixed in 4% PFA for 15 min. Subsequently, hPDLSCs or tissue sections were permeated with 0.25% Triton X-100 for 10 min. The hPDLSCs or tissue sections were reacted with anti-Cx43 and anti-β-catenin antibodies overnight at 4℃ after being sealed with 5% BSA for 1 h. Anti-rabbit antibody AlexaFluor647 (ab150079, Abcam, UK) or anti-mouse antibody AlexaFluor488 (ab150113, Abcam, UK) was incubated at room temperature for 2h. DAPI (D9542, Sigma, America) staining was performed for periodontal ligament cells nuclei. The cytoskeletons were dyed with phalloidine (6µmol/L, Invitrogen, CA). Fluorescence staining images were visualized by CLSM.
2.10 Co-immunoprecipitation (Co-IP)
The detailed operation is performed according to the manufacturer's instructions (YJ201, Epizyme Biomedical Technology, China). After ice-cold protein lysis of hPDLSCs for 10 min, samples were interacted overnight with the 2.5ug Rabbit antibody β-catenin (R22820, Zen-Bio, China). Then the antibody-protein complexes were mixed with Protein A/G Plus magnetic beads overnight at 4℃. After immunoprecipitation, the immune complexes were resuspended in SDS-loading buffer after centrifugation and denatured as mentioned above.
2.11 Statistical Analysis
The data are representative of three independent experiments. The data have been described as mean ± standard deviation (SD). One-way ANOVA and Student's t-test were evaluated by GraphPad Prism version 8 software (GraphPad, San Diego,CA, United States) to calculate the statistical difference.