Introduction
Orthodontically induced root resorption (OIRR) is a frequent yet poorly understood complication of orthodontic treatment. Emerging evidence links oxidative stress to mechanical loading. However, the regulation of redox homeostasis in periodontal tissues under varying force magnitudes remains unclear. Nuclear factor erythroid 2–related factor 2 (Nrf2), a master regulator of antioxidant defense, modulates inflammation and bone remodeling, but its force-dependent role in OIRR is undefined.
Methods
Periodontal ligament fibroblasts were subjected to graded compressive forces (0-2 g/cm 2) in vitro to assess Nrf2/kelch-like ECH–associated protein 1 (Keap1)/sequestosome 1 (p62) pathway activation and downstream inflammatory and osteoclastic responses. Genetic and pharmacologic modulation of Nrf2 signaling was performed. In vivo, a murine orthodontic tooth movement model applying light (10 g) and heavy (40 g) forces was used to evaluate Nrf2 function in periodontal remodeling and root resorption.
Results
Nrf2 displayed threshold-dependent regulation. Moderate force (≤1.5 g/cm 2) activated the Nrf2/HO-1 pathway, preserving redox balance and limiting inflammation, whereas excessive force (2 g/cm 2) led to Nrf2 saturation, resulting in ROS accumulation, amplified inflammation, and enhanced osteoclastogenesis. Keap1 knockdown restored antioxidant capacity and reduced inflammation, whereas p62 knockdown impaired Nrf2 activation and aggravated tissue injury. In vivo, heavy force induced sustained interleukin-1β expression and severe root resorption, intensified by Nrf2 inhibition.
Conclusions
Force-dependent saturation of the Nrf2/Keap1/p62 pathway acts as a molecular switch linking oxidative stress and inflammation in OIRR. Nrf2 serves as a mechanosensitive regulator of periodontal homeostasis and a potential therapeutic target to prevent root resorption.
Graphical abstract
Highlights
-
•
Compressive force regulates Nrf2 activity in a force threshold-dependent manner.
-
•
Nrf2/Keap1/p62 pathway plays a key role in linking oxidative stress and inflammation in orthodontically induced root resorption.
-
•
Activation of the Nrf2/Keap1/p62 pathway offers a therapeutic strategy to prevent root resorption.
Orthodontically induced root resorption (OIRR) is a common iatrogenic consequence of orthodontic treatment, characterized by the irreversible loss of cementum and dentin. Epidemiological studies indicate that 48%-66% of orthodontically treated teeth experience up to 2 mm of root shortening, and 1%-5% lose more than 4 mm. The extent of OIRR varies with multiple clinical and biological factors, including the magnitude and duration of mechanical loading, tooth anatomy, periodontal status, etc. Although the application of light forces is known to mitigate resorption, precise force thresholds remain undefined, and the underlying biological mechanisms are incompletely understood, leaving OIRR prevention largely empirical.
Among periodontal tissue, periodontal ligament fibroblasts (PDLFs) serve as primary mechanosensors and effectors of alveolar bone remodeling. Under orthodontic force, the periodontal ligament (PDL) transduces mechanical stimuli through complex mechanotransduction cascades that orchestrate bone resorption and formation. Excessive compression, however, disrupts PDL microarchitecture and impairs local perfusion, giving rise to hypoxia and excessive accumulation of reactive oxygen species (ROS). The resulting oxidative stress, a state in which ROS and reactive nitrogen species overwhelm intrinsic antioxidant defenses, triggers redox-sensitive signaling pathways such as nuclear factor kappa–light–chain–enhancer of activated B cells pathway, thereby promoting proinflammatory cytokine release (interleukin- [IL-] 1β, IL-6, and tumor necrosis factor-α [TNF-α]) and disturbing the receptor activator of nuclear factor kappa–B ligand (RANKL)/osteoprotegerin (OPG) balance to favor osteoclastogenesis and root resorption. Although the interplay between mechanical stimulation and ROS signaling has been demonstrated in vascular endothelial cells, chondrocytes, and nucleus pulposus cells, its direct role in orthodontic tooth movement (OTM) remains insufficiently defined. In particular, how PDLFs regulate ROS and inflammation via endogenous antioxidant systems, and whether these processes vary with force magnitude, remains unclear.
Nuclear factor erythroid 2–related factor 2 (Nrf2) is a key transcriptional regulator of cellular redox homeostasis. Under physiological conditions, Nrf2 is maintained at low intracellular levels through ubiquitin-mediated degradation via its interaction with the kelch-like ECH–associated protein 1 (Keap1)–cullin 3 complex. Elevated ROS levels oxidize cysteine residues on Keap1, releasing Nrf2 to translocate into the nucleus, in which it activates the transcription of downstream antioxidant genes, such as heme oxygenase-1 (HO-1) and glutathione synthetase. The selective autophagy adaptor sequestosome 1 (p62 or SQSTM1) further modulates this process by competitively binding Keap1 through its Keap1-interacting region, thereby preventing Nrf2 degradation and forming a positive feedback loop—the Nrf2/Keap1/p62 pathway. This pathway exerts cytoprotective effects across diverse inflammatory and metabolic diseases. For instance, Nrf2 suppression enhances NOD-like receptor family pyrin domain containing 3 inflammasome activation in osteoarthritis, and impaired Nrf2 function aggravates ROS-driven bone loss in apical periodontitis. In orthodontics, Nrf2 activation has been shown to attenuate both OTM and relapse. Moreover, recent studies have elucidated the interplay between mechanical stimulation, oxidative stress, and periodontal remodeling. Cyclic stretch elevates intracellular ROS and promotes Nrf2-dependent osteogenic differentiation in PDLFs, whereas nicotinamide adenine dinucleotide (NAD⁺)/Nrf2 signaling maintains redox balance and drives osteogenic commitment in bone marrow mesenchymal stem cells under cyclic stretch.
Despite these insights, direct evidence linking oxidative stress with OIRR and elucidating how Nrf2 activity is modulated by mechanical force intensity remains limited. The molecular mechanisms by which PDLFs coordinate oxidative stress, inflammatory signaling, and osteoclastogenesis under different loading conditions are poorly defined. Moreover, it remains unknown whether Nrf2 exhibits force-dependent activation thresholds or saturation effects under excessive stress.
Here, we propose that mechanical force regulates oxidative stress, inflammation, and root resorption in a force-intensity–dependent manner via the Nrf2/Keap1/p62 pathway. Specifically, we hypothesize that moderate compressive force activates Nrf2 to enhance antioxidant defenses, whereas excessive forces surpass the adaptive capacity of this pathway, leading to impaired redox homeostasis, uncontrolled ROS accumulation, elevated osteoclastogenesis, and aggravated root resorption. To test this, we employed a graded in vitro compressive force model and an in vivo OTM model, integrating pharmacologic and genetic manipulation of Nrf2 signaling. This study aims to elucidate the mechanistic role of Nrf2 in force-mediated periodontal remodeling and to establish a theoretical framework for antioxidant-based therapeutic strategies in orthodontics.
Material and methods
Eight-week–old male C57BL/6 mice were purchased from Dashuo Biology Technology (Chengdu, China) and acclimated under a 12-hour light and dark cycle at 25°C-28°C for 1 week. All animal procedures were approved by the Animal Care Committee of Sichuan University (ethical number: WCHSIRB-D-2022-157).
To establish the OTM model, a nickel-titanium coil spring was ligated between the maxillary right first molar and maxillary incisors. Nrf2 inhibitor ML385 (HY-124305, MedChemExpress, Shanghai, China) was intraperitoneally injected 1 hour before force application at a dosage of 30 mg/kg once a day until euthanization. Control mice received an equal volume of saline. Mice were euthanized at 1, 3, 7, and 14 days after force application. Maxillae, including the maxillary right first molar, were harvested and fixed in 4% paraformaldehyde overnight at 4°C.
Samples were scanned using a micro-computed tomography (CT) 45 system (Scanco Medical, Bruttisellen, Switzerland) at 10 μm resolution. Three-dimensional (3D) reconstructions of the maxillary right first molar were generated with Mimics (version 21.0; Materialise, Belgium) after established protocols. Root resorption lacunae were identified and outlined on gray scale images, with intact and resorbed regions marked in red and yellow, respectively. The remaining root and resorbed volumes were quantified using Mimics, and the root resorption volume ratio (percentage) was calculated as resorbed volume/remaining root volume.
After micro-CT scanning, the maxillae were decalcified, dehydrated, and paraffin-embedded. Serial 4 μm sections were cut in the sagittal plane. Sections were stained with hematoxylin and eosin (HE; Solarbio, Beijing, China), Sirius red (Solarbio), and tartrate-resistant acid phosphatase (TRAP; Wako, Saitama, Japan) according to the manufacturers’ instructions. The PDL width was determined on HE-stained sections by drawing a perpendicular line from the midpoint of the mesial surface of the distal root to the root surface. TRAP-positive multinucleated cells on the surface of the compression side were counted.
For immunohistochemistry (IHC), staining was performed using the VECTASTAIN Elite ABC-HRP kit (PK-6101, Vector Laboratories, Calif). After deparaffinization, hydration, antigen retrieval, and blocking, sections were incubated with primary antibodies, including anti-IL1β (1:200, ab205924, abcam, Waltham, Mass), anti-HO1 (1:1000, ab13243, abcam), and anti-Nrf2 (1:200, AF0639, Affinity, Shanghai, China). Images from the apical one third section of the mesial compression side were captured by a light microscope (Eclipse 80i; Nikon, Tokyo, Japan). Quantitative analyses were performed using ImageJ (Media Cybernetics, Bethesda, Md), and the mean optical density was calculated as mean optical density (MOD) = Integrated optical density/Area.
As for in vitro experiments, human PDLFs (hPDLFs) were isolated from premolars extracted for orthodontic treatment purposes as described previously. All procedures were approved by the Research Ethics Committee of the State Key Laboratory of Oral Diseases, West China Hospital of Stomatology, Sichuan University (ethical number: WCHSIRB-D-2021-539). Cells were seeded in 6-well plates (5 × 10 4 cells per well) and subjected to graded compressive forces after adhesion. For Nrf2 inhibition, cells were treated with 5 μM ML385 (HY-124305, MedChemExpress) for 24 hours according to the previous research.
For small interfering (si)–RNA transfection, hPDLFs were transfected with siRNA (100 nM, Hanbio, Shanghai, China) using Lipofectamine RNAiMAX (Invitrogen, Waltham, Mass). The siRNA sequences were as follows: siKeap1 (forward 5’-GGAGCGCUACGAUGUGGAATT-3’ and reverse 5’-UUCCACAUCGU AGCGCUCCTT-3’) and sip62 (forward 5’-CCUGCAGACCAAGAACUAUTT-3’ and reverse 5’- AUAGUUCUUGGUCUGCAGGTT-3’). Transfection efficiency was confirmed by real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) and western blot analysis.
For RNA-seq analysis, total RNA was extracted by Trizol reagent (Invitrogen). mRNA sequencing was performed by NewCore Biotech (Shanghai, China) as described previously. Reverse transcription and RT-qPCR were performed by PrimeScript FAST RT reagent kit with gDNA Eraser (Takara, Shiga, Japan) and TB Green Premix ExTaq II FAST qPCR kit (Takara). Primer sequences are provided in the Supplementary Table . Relative gene expression levels were normalized to glyceraldehyde 3–phosphate dehydrogenase (GAPDH) and calculated via the 2 -△△Ct method.
For protein analysis, total protein was extracted by the total protein extraction kit (Signalway Antibody, Greenbelt, Md). The protein was subjected to a 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride (PVDF) membrane. After blocking, membranes were incubated with the following primary antibodies: anti-GAPDH (1:5000, ER1706-83, Huabio, Zhejiang, China), anti-HO1 (1:2000, ab13243, abcam), anti-p62 (1:1000, 5114, Cell Signaling Technology, Danvers, Mass), anti-Keap1 (1:1000, 4678, Cell Signaling Technology), and anti-Nrf2 (1:1000, AF0639, Affinity). Bands were visualized by the ChemiDoc touch imaging system (Bio-Rad, Hercules, Calif) and quantified using ImageJ after normalization to GAPDH.
Intracellular ROS were measured using the ROS assay kit (S0033S, Beyotime, Shanghai, China). Cells were incubated with DCFH-DA (1:1000) and Hoechst 33342 (C1027, Beyotime) at 37°C for 30 minutes and imaged under the fluorescence microscope (Leica DMi8, Wetzlar, Germany).
For immunofluorescence (IF) staining, hPDLFs were fixed, permeabilized, and blocked, then incubated with primary antibody anti-Nrf2 (1:200, 12721, Cell Signaling Technology). Nuclei and cytoskeleton were stained with 4,6-diamino-2-phenyl indole (DAPI; D9542, Sigma-Aldrich, Saint Louis, Mo) and phalloidine (6 mmol/L, Invitrogen). Images were captured with confocal laser microscopy (CLSM; FV-3000, Olympus, Tokyo, Japan), and fluorescence intensity was quantified with ImageJ.
To assess the ability of PDLFs to induce osteoclast differentiation of Tohoku hospital pediatrics–1 (THP-1) cells, supernatants from treated PDLFs were collected. THP-1 cells (2×10 5 cells per well) were prestimulated with PMA (100 ng/mL, APExBIO, Houston, Tex) for 48 hours. Afterward, cells were cultured in PDLF-conditioned medium supplemented with M-CSF (40 ng/mL, AMK9001, Amizona, China) and RANKL (40 ng/mL, AMK9001, Amizona), with medium replacement every 3 days for 12 days. Afterward, TRAP (Wako), nuclei, and cytoskeleton staining were performed, and both bright-field and fluorescence images were acquired under identical fields. TRAP + multinuclear cells (nuclei ≥ 3) were counted in 5 randomly selected fields per well.
Statistical analyses
GraphPad Prism (version 9; San Diego, Calif) was used for statistical analysis. Data were expressed as the mean ± standard deviation from at least 3 independent experiments. The statistical difference among groups was analyzed by one-way analysis of variance with Tukey’s honest significant difference test. P <0.05 was considered statistically significant.
Results
The OTM model was established to evaluate periodontal tissue remodeling under 10 g (light) force and 40 g (heavy) force. HE and Sirius Red staining revealed the extensive hyalinization in the 40 g group during days 1-3 ( Supplementary Fig 1 , A and B ). PDL width decreased significantly on days 1 and 3 under both forces, with a more pronounced reduction in the 40 g force group. By day 7, the PDL width in the 40 g force group returned to baseline, whereas the 10 g force group exhibited gradual widening, indicating early tissue remodeling ( Supplementary Fig 1 , C ). TRAP staining demonstrated several osteoclasts along the root surface in the 40 g force group, showing 2.3-fold and 3.3-fold increases relative to the 10 g force group at days 7 and 14, respectively ( Fig 1 , A and B ; Supplementary Fig 1 , D ). Consistently, micro-CT analysis showed that root resorption lacunae in the 40 g force group were 60% and 52% larger than the 10 g force group at days 7 and 14, respectively ( Fig 1 , C and D ; Supplementary Fig 1 , D ). IHC staining of IL-1β showed significant elevation in both groups at days 1 and 3. The expression normalized by day 7 under light force but persisted through day 14 under heavy force ( Fig 1 , E and F ). Collectively, these findings indicate that light orthodontic force elicits adaptive alveolar bone remodeling, whereas excessive force induces prolonged inflammation, augmented osteoclastogenesis, and root resorption.
Excessive mechanical force exacerbates periodontal inflammation and osteoclastogenesis in vivo and in vitro: A and B, Representative images of TRAP staining of PDL on the compression side of the distal root and quantitative analysis of the number of TRAP-positive cells on the root surface. Scale bar = 50 μm; C and D, Representative 3D micro-CT images of the distal root and quantitative analysis of the resorption volume ratio; E and F, Representative immunohistochemical images and semiquantitative analysis of IL-1β at the mesial surface of distal roots. Scale bar = 20 μm. A , alveolar bone; P , PDL; and R , root. n = 5 for each group; G and H, RT-qPCR analysis of proinflammatory factors (IL-1β, IL-6, IL-8, and TNF-α), RANKL, OPG mRNA levels, and RANKL/OPG ratio after graded compressive forces; I and J, Representative images of TRAP staining and quantification of the number of TRAP + multinuclear cells after treatment by supernatant from treated PDLFs ( red , TRAP + multinuclear cells [nuclei ≥ 3]); K , GO enrichment analysis of differentially involved signaling pathways between the 2 g/cm 2 force group and the control; L , Heat map showing differentially expressed mRNAs; M , GO chord plot showing possible crosstalk between enriched signaling pathways. Data are shown as the mean ± standard deviation. n = 3 for each group. ∗ P <0.05, ∗∗ P <0.01, ∗∗∗ P <0.001, and ∗∗∗∗ P <0.0001.
To dissect the cellular mechanisms underlying force-induced responses, hPDLFs were subjected to graded compressive forces (0, 1, 1.5, and 2 g/cm 2) for 3 hours. RT-qPCR revealed upregulation of inflammatory cytokines (IL-1β, IL-6, IL-8, and TNF-α), the RANKL/OPG mRNA ratio, and osteoclast-related genes (Cathepsin K, matrix metalloproteinase–3 [MMP3], and nuclear factor of activated T-cells cytoplasmic 1 [NFATc1]) under 1 and 1.5 g/cm 2 of forces, with comparable expression levels between these groups, suggesting similar cellular responses to moderate stress levels. In contrast, 2 g/cm 2 force markedly enhanced the expression of all these genes ( Fig 1 , G and H ; Supplementary Fig 1 , E ). Moreover, conditioned medium from the 2 g/cm 2 force group further promoted osteoclast-like differentiation in THP-1 cells ( Fig 1 , I and J ). Given the clinical association between excessive force and OIRR in clinical practice, RNA sequencing was performed on the 2 g/cm 2 force group and control to further elucidate the underlying mechanisms. A total of 482 differentially expressed genes were identified, including 120 upregulated and 362 downregulated genes ( Supplementary Fig 1 , F ). Gene ontology (GO) enrichment analysis highlighted biological processes related to cellular hypoxia response, oxidative stress response, inflammation, and osteoclast differentiation ( Fig 1 , K ). Heatmap visualization revealed significant upregulation of oxidative stress regulators, including Nrf2 (NFE2L2), HO-1 (HMOX1), and p62 (SQSTM1), alongside inflammation and osteoclastogenesis-related genes ( Fig 1 , L ). GO chord diagram confirmed strong connectivity between Nrf2, HO-1, p62, and redox–inflammatory pathways ( Fig 1 , M ). These data implicate the Nrf2/Keap1/p62 axis as a key mediator of force-induced oxidative and inflammatory signaling.
To further explore Nrf2/Keap1/p62 pathway activation, hPDLFs were exposed to 2 g/cm 2 compressive force for 0, 1, 2, 3, 6, 12, and 24 hours. Nrf2 expression peaked at 3 hours, followed by a decline, whereas HO-1 expression reached its maximum at 6 hours ( Fig 2 , A ), suggesting HO-1 responds later than Nrf2 under compressive force. A 3-hour loading duration was therefore selected for subsequent experiments. Cellular ROS levels in the 2 g/cm 2 force group increased 1.39-fold and 1.28-fold relative to the 1 and 1.5 g/cm 2 force groups, respectively ( Fig 2 , B ). Western blot analysis revealed a force-dependent increase in Nrf2 expression and a decrease in Keap1, with an upregulation of HO-1 and p62 at moderate forces. However, at 2 g/cm 2 of force, Nrf2 expression plateaued and neither HO-1 nor p62 levels showed further elevation, suggesting antioxidant defense saturation ( Fig 2 , C ). IF analysis confirmed nuclear translocation of Nrf2 at moderate stress levels, but no further increase at 2 g/cm 2 force ( Fig 2 , D and E ). Consistently, in vivo IHC staining of PDL tissues from the 40 g force group showed significantly increased expression of Nrf2 and HO-1 on days 1 and 3 compared with the 10 g force group ( Fig 2 , F – I ).
