Abstract
The high prevalence of malocclusion and dentofacial malformations means that the demand for orthodontic treatments has been increasing rapidly. As the biological basis of orthodontic treatment, the mechanism of mechanical force–induced alveolar bone remodeling during orthodontic tooth movement (OTM) has become the key scientific issue of orthodontics. It has been demonstrated that bone mesenchymal stem cells (BMSCs) are crucial for bone remodeling and exhibit mechanical sensing properties. Mechanical force can promote osteoblastic differentiation of BMSCs and osteogenesis, but the key factor that mediates mechanical force–induced osteogenesis during OTM remains unclear. In this study, by performing reverse-phase protein arrays on BMSCs exposed to mechanical force, we found that the expression level of forkhead box O3 (FOXO3) was significantly upregulated during the mechanical force–induced osteoblastic differentiation of BMSCs. The number of FOXO3-positive cells was consistently higher on the OTM side as compared with the control side and accompanied by the enhancement of osteogenesis. Remarkably, inhibiting FOXO3 with repaglinide delayed OTM by severely impairing mechanical force–induced bone formation in vivo. Moreover, knockdown of FOXO3 effectively inhibited the mechanical force–induced osteoblastic differentiation of BMSCs, whereas the overexpression of FOXO3 enhanced this effect. Mechanistically, we revealed a novel regulatory model in which FOXO3 promoted osteocalcin transcription by activating its promoter in cooperation with runt-related transcription factor 2 (RUNX2). We collectively obtained the first evidence that FOXO3 is critical for OTM, where it responds to mechanical force and directly regulates downstream osteoblastic differentiation in an efficient manner.
Keywords
Introduction
The prevalence of dental-maxillofacial malformations is as high as 70% to 90% (Gois et al. 2012). Dental-maxillofacial malformations are serious problems in contemporary society due to high morbidity and their harmful effects on appearance and oral functions. Thus, the demand for orthodontic treatment is increasing each year. Orthodontic treatment is based on the principle that tooth movement occurs when prolonged mechanical stimuli is applied to the teeth and the periodontal tissues, including alveolar bone remodel. In essence, orthodontic tooth movement (OTM) is primarily a process of alveolar bone remodeling activated by mechanical force (Krishnan and Davidovitch 2009). Hence, clarifying the underlying mechanisms associated with mechanical force–regulated alveolar bone remodeling is of great importance.
Alveolar bone remodeling involves the removal of mineralized bone via osteoclasts, followed by the formation of new bone matrix by osteoblasts. This continuous osteoclastogenesis-osteogenesis remodeling enables the bone to adapt to various external stimuli, such as mechanical stress (Hadjidakis and Androulakis 2006). It has been demonstrated that bone mesenchymal stem cells (BMSCs) are crucial for bone remodeling and that they exhibit mechanical sensing properties. BMSCs are not only the progenitor of osteoblasts, where they serve as one of the regulators of osteoclastic differentiation, but also the direct responder to external stress, since they can sense the mechanical force, convert it into intracellular biological signals, and initiate bone remodeling to meet the changing mechanical needs (Engler et al. 2006; Ikebuchi et al. 2018; Ransom et al. 2018; Wang et al. 2020). Thus, the mechanisms involved in osteoblastic differentiation of BMSCs under mechanical force are the key components of mechanical stress–regulated alveolar bone remodeling during OTM.
Previous studies have shown that multiple mechanisms are involved in the force-induced osteoblastic differentiation of BMSCs, including PI3K-AKT-mTOR and Wnt/β-catenin (Wu et al. 2012; Wang et al. 2017; Ouyang et al. 2018). As a downstream effector of PI3K-AKT signaling, forkhead box O3 (FOXO3) is a vital transcription factor that contributes to diverse cellular processes and participates in bone metabolism (Paik et al. 2007; Tsai et al. 2008; Martins et al. 2016). More important, evidence suggests that FOXO3 responds to mechanical stress in oocytes (Brunet et al. 2004; Nagamatsu et al. 2019; Shimamoto et al. 2019). However, the role of FOXO3 in mechanical force–regulated alveolar bone remodeling and its characteristics in OTM remain unclear.
The aim of this study is to reveal the key factors that mediate mechanical force–induced osteogenesis during OTM. In this work, we found that FOXO3 responded to force and effectively regulated downstream osteoblastic differentiation during OTM. Furthermore, we showed that FOXO3 could promote osteocalcin transcription by activating its promoter in cooperation with runt-related transcription factor 2 (RUNX2). These findings may facilitate the development of new solutions to clinical issues in orthodontic treatment.
Materials and Methods
Some methods are described in the Appendix.
Application of Tension
A Flexcell Strain Unit Fx-5000 (Flexcell Corp.) was used to apply mechanical tension in the cell incubator (37 °C, 5% CO2). BMSCs were seeded at a density of 2.0 × 105 onto BioFlex collagen I–coated plates (Flexcell Corp.) and subjected to cyclic tension (10% elongation, 0.5 Hz) in osteogenic induction medium for 8 h/d.
Reverse-Phase Protein Arrays
Reverse-phase protein arrays (RPPAs) were conducted at the MD Anderson Cancer Center, University of Texas, according to its protocols. BMSCs were exposed to mechanical force for 0, 4, 8, 12, and 24 h and analyzed. Array-Pro was used to analyze 272 unique antibodies, and the results were processed by SuperCurve Rx64 3.1.1.
OTM Model and Repaglinide Treatment
All animal experimental procedures were approved by the Animal Care Committee at Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, and performed according to institutional guidelines and the ARRIVE guidelines (Animal Research: Reporting In Vivo Experiments). The OTM model was performed with 6-wk-old male C57BL/6J mice as described previously (n = 6; Cao et al. 2014). In the FOXO3 inhibitor treatment, intraperitoneal injection of repaglinide was performed at a concentration of 25 mg/kg every day in the OTM model for 10 consecutive days.
Histologic Analysis
Alveolar bone tissues were harvested at OTM day 10, sectioned at 6 μm, and stained with rabbit monoclonal FOXO3 (2497, 1:50; Cell Signaling Technology), mouse monoclonal OCN (sc-390877, 1:200; Santa Cruz Biotechnology), or mouse monoclonal RANKL (12A668, 1:200; Abcam) primary antibody. TRAP staining (TRAP Kit 387A; Sigma-Aldrich) was performed as described previously (Zou et al. 2013). For calcein–alizarin red S double labeling, mice were injected intraperitoneally with 20 mg/kg of calcein and 40 mg/kg of alizarin red S during OTM on days 4 and 8, respectively. The region of interest is described in Appendix Figure 2B. Images were captured with a laser scanning confocal microscopy (LSM 880; Zeiss), and quantification was based on 3 fields per animal via ImageJ 1.47 software (National Institutes of Health).
Plasmids
The HA-RUNX2 plasmid was a gift from Dr. Gerard Karsenty’s laboratory (Wang et al. 2004). The FLAG-FOXO3 plasmid, FOXO3-shRNA plasmid, Ocn promoter and mutant Ocn promoter-driven pGL3-based luciferase reporters were synthesized.
Luciferase Reporter Assay
HEK 293T cells were transfected with polyetherimide with a luciferase reporter plasmid and Renilla luciferase (Promega) with various combinations of expression plasmids, as indicated. At 48 h posttransfection, cells were lysed and subjected to a dual-luciferase reporter assay (Promega).
Co-immunoprecipitation
Co-immunoprecipitation (co-IP) was performed according to a previously described method (Dai, Xie, et al. 2017). Flag-FOXO3 and HA-RUNX2 expression plasmids were transfected with polyetherimide and immunoprecipitated with rabbit polyclonal FLAG (SAB4301135, 1 μg; Sigma-Aldrich) or mouse monoclonal HA antibody (66006-2-Ig, 1 μg; Proteintech).
Chromatin Immunoprecipitation
Chromatin immunoprecipitation (ChIP) analysis was performed with an Enzymatic Chromatin Immunoprecipitation Kit (EZ ChIPTM 17-371; Merck-Millipore) according to the manufacturer’s instructions. Immunoprecipitation was performed with mouse monoclonal FOXO3 (66428-1-Ig, 1:50; Proteintech) in C3H10 t1/2 cells. The precipitated DNA was detected by quantitative reverse transcription polymerase chain reaction (qRT-PCR) with specific primers.
Statistical Analysis
All quantitative data were expressed as mean ± SD based on at least 3 independent samples. Student’s t test was used to compare differences between 2 groups. Statistical comparisons among >2 groups were conducted by 1-way analysis of variance. Multiple comparison of groups was performed with the S-N-K method. P < 0.05 indicated a statistically significant difference.
Results
Mechanical Force Promoted Expression of FOXO3 during OTM
To elucidate the mechanism associated with mechanical force–induced osteogenesis, an in vitro tension loading model was constructed (Zhang et al. 2015). Mandibular BMSCs (Appendix Fig. 1A–C) were seeded in osteogenic induction medium and subjected to mechanical tension of 10% at 0.5 Hz for 8 h/d. The tension promoted osteoblast differentiation of the BMSCs, as indicated by the increased alkaline phosphatase (ALP) activity (Fig. 1A) and upregulated expression levels of the early osteogenic marker genes Runx2 and Alpl and the late osteogenesis marker genes osteocalcin (Ocn) and bone sialoprotein (Ibsp; Fig. 1B; Appendix Fig. 1D, E). Next, to investigate the key factors that mediate mechanical stimuli–induced osteoblast differentiation, BMSCs exposed to tension for 0, 4, 8, 12, and 24 h were analyzed by RPPAs. We found that the FOXO3 protein level was significantly upregulated by mechanical loading, especially in the 8- and 12-h groups (Fig. 1C). Next, we confirmed that the protein expression level and mRNA level of FOXO3 were increased in BMSCs after tension loading, according to Western blotting and qRT-PCR analyses (Fig. 1D, E). In addition, immunofluorescence staining showed that the number of FOXO3-positive BMSCs was notably higher in the tension application group as compared with the control group (Fig. 1F).

Mechanical force promoted expression of FOXO3 during orthodontic tooth movement (OTM). (
Next, we assessed the in vivo reaction of FOXO3 to mechanical force during OTM. Significant tooth movement was observed in the OTM model after 10 d (Fig. 1G; Appendix Fig. 2A, B). Consistent with the in vitro tension-loading results, the number of FOXO3-positive cells was significantly higher on not only the tension and compression side but also the interradicular region (hereafter, OTM side) as compared with the control side (Fig. 1I; Appendix Fig. 2H, I). This finding accompanied enhancement of osteogenesis, as indicated by the increased mineral apposition rate, according to calcein and alizarin red S double labeling, and the increased number of OCN-positive osteoblasts (Fig. 1H; Appendix Fig. 2D, E). More TRAP-positive osteoclasts suggested activation of bone resorption activity as well (Appendix Fig. 2F, G). Overall, these results demonstrated that mechanical force effectively promoted the expression of FOXO3 during OTM.
FOXO3 Was Required for Bone Remodeling during OTM
To characterize the role of FOXO3 in OTM and mechanical force–induced bone formation, we performed intraperitoneal injection of repaglinide, a FOXO3 inhibitor, in the OTM model to silence the transcriptional activity of FOXO3 (Salcher et al. 2019). Remarkably, we found that the tooth movement was significantly impaired in the repaglinide-treated group, as the moving distance showed a 2.5-fold reduction as compared with the vehicle control according to micro–computed tomography analysis (Fig. 2A, B). The calcein and alizarin red S double-labeling assay showed that mineral apposition rate was lower in the repaglinide-treated group as compared with the vehicle control (Fig. 2C, Appendix Fig. 3A). Immunostaining of OCN also showed that the number of OCN-positive osteoblasts on the OTM side decreased in the repaglinide group (Fig. 2D), thereby indicating that FOXO3 inhibition could block bone formation in OTM. In addition, TRAP staining detected fewer TRAP-positive multinuclear osteoclasts surrounding the alveolar bone in the repaglinide-treated group (Fig. 2E, Appendix Fig. 3B). Furthermore, immunofluorescence staining showed that the number of RANKL-positive cells (receptor activator of nuclear factor kappa–Β ligand) was significantly decreased in the repaglinide-treated group as compared with the vehicle group (Fig. 2F, Appendix Fig. 3C). These results indicated that the bone-remodeling activity was impaired when FOXO3 was inhibited and that FOXO3 was essential for alveolar bone remodeling during OTM.

FOXO3 was required for bone remodeling during orthodontic tooth movement (OTM). (
FOXO3 Was Critical for Mechanical Force–Induced Osteoblast Differentiation of BMSCs
After demonstrating that FOXO3 contributed to alveolar bone remodeling during OTM, we explored the role of FOXO3 in the mechanical force–induced osteoblast differentiation of BMSCs. First, we knocked down the expression of FOXO3 via lentivirus-expressed shRNA interference (short hairpin RNA) in BMSCs. After confirming the inhibitory effect of shRNA on the RNA and protein levels for FOXO3 (Fig. 3A, C), we analyzed the influence of FOXO3 knockdown on the osteogenic differentiation of BMSCs under conditions with mechanical tension application or no stress loading. As shown in Figure 3B and C, the ALP activity and mRNA levels of osteogenesis-related markers such as Alpl, Ocn, Ibsp, and osteopontin (Opn) were upregulated by tension, but this effect was notably reduced in the FOXO3-shRNA group, thereby indicating that interference with FOXO3 gene expression could significantly inhibit the mechanical force–induced osteoblastic differentiation of BMSCs. Next, FOXO3 was overexpressed in BMSCs with lentivirus to determine the role of FOXO3 in osteogenesis. The overexpression of FOXO3 was verified by Western blotting and qRT-PCR (Fig. 3D, F). As shown in Figure 3E and F, the overexpression of FOXO3 increased the ALP activity and mRNA levels of downstream osteogenesis-related genes. These results indicated that FOXO3 acted as a vital intermediate molecule in the response to mechanical force because its expression level was increased by tension; it then played a significant role in regulating the downstream osteoblastic differentiation of BMSCs.

FOXO3 was critical for mechanical force–induced osteoblast differentiation of bone mesenchymal stem cells (BMSCs). (
FOXO3 Regulated Ocn Transcription by Activating Its Promoter in Cooperation with RUNX2
We sought to gain insights into the underlying molecular mechanism through which FOXO3 regulated downstream osteogenesis-related markers. The expression of Ocn was changed in an equidirectional manner with FOXO3 expression, where its mRNA level was downregulated by FOXO3 interference and upregulated by FOXO3 overexpression (Fig. 3C, F), so we hypothesized that FOXO3 might directly regulate Ocn expression. After matching the FOXO3 motif with 2 kb upstream of the Ocn transcription start sites and finding a potential FOXO3 binding site (Fig. 4A), we performed a ChIP assay. It showed that FOXO3 was enriched in the promoter of Ocn when the antibody against FOXO3 was used for immunoprecipitation as compared with the control immunoglobulin G (Fig. 4B), suggesting that FOXO3 bound directly to the Ocn promoter. Subsequently, we analyzed the effects of FOXO3 on the transcriptional activity of Ocn by utilizing its promoter-driven luciferase reporter. As shown in Figure 4C, the activity of the Ocn promoter was effectively increased by FOXO3 in a concentration-dependent manner. Moreover, we constructed a mutant Ocn promoter where the predicted FOXO3 motif was mutated from GATAAACA to AGCGGGTG (OG2-mu). The mutation abrogated the activity of FOXO3 on Ocn promoter, indicating that this FOXO3 binding site was essential for its transcription (Fig. 4D). Furthermore, the FOXO3 transcription inhibitor repaglinide blocked the increased transcription activity of the Ocn promoter, which confirmed that FOXO3 could regulate Ocn transcription (Fig. 4E). Thus, we found that FOXO3 could directly bind to the Ocn promoter and activate its transcriptional activity.

FOXO3 regulated Ocn transcription in cooperation with RUNX2. (
Previous studies have shown that RUNX2 is crucial for the transcription of Ocn by binding to its promoter (Ducy et al. 1997). RUNX2 expression was not affected by FOXO3 in the FOXO3 interference or FOXO3 overexpression conditions (Fig. 3C, F), so we wondered whether there was cooperation between these transcription factors in regulating osteogenesis-related genes. Co-IP indicated that a protein interaction occurred between FOXO3 and RUNX2 (Fig. 4F). The HA-RUNX2 and FLAG-FOXO3 plasmid cotransfection assay consistently showed that FOXO3 and RUNX2 colocalized in the nucleus, which indicated the formation of their physical complex (Fig. 4G). Moreover, FOXO3 and RUNX2 were cotransfected into HEK 293T cells to analyze their effects on the activity of the Ocn promoter. Transfection with RUNX2 or FOXO3 alone promoted the transcriptional activity of Ocn, but cotransfection with both had a synergistic effect because they further increased the promoter activity of Ocn (Fig. 4H). In addition, repaglinide blocked this enhanced effect of FOXO3 and RUNX2 on the Ocn promoter activity (Fig. 4I). These results indicated that FOXO3 cooperated with RUNX2 to drive Ocn transcription.
Discussion
The demand for orthodontic treatments is increasing rapidly each year due to the high prevalence of dental-maxillofacial malformations. Orthodontic treatment has become a more prominent part of dentistry in recent years, and this trend is expected to continue (Gudipaneni et al. 2018). Nevertheless, several clinical issues remain to be solved, such as root and alveolar bone resorption (Ikeda et al. 2004; Jiang et al. 2010; Richter et al. 2011). As the essence of orthodontic treatment is force-regulated alveolar bone remodeling, the key to solve those clinical problems is, to our knowledge, clarifying the underlying mechanisms of mechanical force–induced remodeling of alveolar bone. In our previous studies, we focused on the biomechanical responses of BMSCs to changes in mechanical force, but the key factors that mediate force-regulated bone remodeling during OTM remain unclear (Zhang et al. 2015; Ouyang et al. 2018). In the present study, we found that FOXO3 may play a critical role in OTM because its expression level was upregulated after force loading in OTM and that the inhibition of FOXO3 could greatly impair tooth movement.
To elucidate the mechanism associated with orthodontic force–induced osteogenesis, an in vitro force loading model was conducted. We have analyzed the effects of different parameters, such as amplitude, frequency, and duration, and found that the tension of 10% elongation, 0.5 Hz, 8 h/d, is suitable for the osteoblastic differentiation of BMSCs induced by force (Zhang et al. 2015; Ouyang et al. 2018), although the force applied in vitro is difficult to completely mimic that applied in vivo due to the complicated microenvironment.
Increasing evidence indicates that the alveolar bone–remodeling activity is one of the most important factors that affect the rate of OTM (Verna et al. 2000; Huang et al. 2014). Similarly, our previous study showed that force-induced increased osteogenesis enables accelerated tooth movement in ovariectomized rats due to enhanced alveolar bone-remodeling activity (Dai, Zhou, et al. 2017). Interestingly, in the present study, we found that FOXO3 inhibition significantly decreased osteogenesis and osteoclastogenesis, which suggests that inhibiting FOXO3 could delay OTM by decreasing the bone-remodeling activity.
The FOXO family comprises 4 members: FOXO1, FOXO3, FOXO4, and FOXO6. At the cellular level, the transcriptional targets of FOXOs are involved in regulating the stem cell cycle and oxidative stress resistance (Kitamura et al. 2007; Zhang et al. 2011). Among the FOXO family, the specific role of FOXO1 and FOXO3 in bone metabolism has been validated (Matsuzaki et al. 2018; Lee et al. 2020). However, whether FOXOs play a role in orthodontic force–induced alveolar bone remodeling remains unknown. In the present study, for the first time, we demonstrated that FOXO3 responded to orthodontic force and regulated the downstream osteoblastic differentiation of BMSCs. By performing RPPA on BMSCs exposed to tension, we found that the FOXO3 expression level was significantly upregulated in the force-induced osteoblastic differentiation. We also observed that β-catenin expression levels went down; this might be partially due to the suppression effect of FOXO3 on β-catenin expression (Iyer et al. 2013), though the relationship between FOXO3 and Wnt signaling under mechanical stress is worthy of further study. The number of FOXO3-positive BMSCs was notably and consistently higher in the force application group than the control group in vitro and in vivo. We found that FOXO3 was mostly detected in the nucleus of BMSCs while few of them detected in the cytoplasm; this may be due to the specific cell adhesive area of the culture plate for in vitro tension loading (McBeath et al. 2004; Dupont et al. 2011). Furthermore, we found that the knockdown of FOXO3 in BMSCs could effectively block force-induced osteoblast differentiation, whereas the overexpression of FOXO3 could enhance this effect. These findings indicate that FOXO3 itself acted as an inducer of osteogenic differentiation and that its upregulated expression level corresponded to force, resulting in enhancing osteoblastic differentiation of BMSCs.
Bone remodeling depends on the regulation of sophisticated transcription networks (Rosen 2008). As one of the most important markers of mature osteoblasts, the expression of Ocn changed in an equidirectional manner with FOXO3 expression. The activity of the luciferase reporter driven by the Ocn promoter and enhanced by FOXO3 expression as well as the ChIP assay indicated that FOXO3 could bind to the Ocn promoter and directly regulate its transcription. Additionally, FOXO1 can interact with RUNX2 to regulate the promoter of osteoblast marker genes, and RUNX2 is crucial for the transcription of Ocn (Ducy et al. 1997; Teixeira et al. 2010; Siqueira et al. 2011). Thus, we hypothesized that FOXO3 may directly interact with RUNX2 and cooperate in the transcriptional regulation of Ocn. Co-IP assays revealed that FOXO3 physically interacted with RUNX2. Moreover, cotransfection of FOXO3 and RUNX2 showed a synergistic effect as they could further increase the promoter activity of Ocn. Finally, the synergistic effects of FOXO3 and RUNX2 on Ocn promoter activity were dependent on FOXO3 transcription activity. Taken together, FOXO3 responded transcriptionally to mechanical stimuli and regulated osteogenesis by directly promoting Ocn transcription and cooperating with RUNX2, which may play a part in force-induced osteoblast differentiation (Fig. 4J).
In conclusion, for the first time, we demonstrated that FOXO3 responded to mechanical stress and orchestrated downstream osteoblastic differentiation during OTM. In addition, we showed that FOXO3 could promote Ocn transcription by activating its promoter in cooperation with RUNX2. These findings highlight the important role of FOXO3 in mechanical stimuli–regulated alveolar bone remodeling and may pave an avenue for developing FOXO3-based approaches to manipulate OTM for solving current clinical issues in orthodontic treatment, although additional research is still required.
Author Contributions
A. Jin, contributed to data acquisition, analysis and interpretation, drafted and critically revised the manuscript; Y. Hong, contributed to conception and design, drafted and critically revised the manuscript; Y. Yang, X. Gao, X. Gong, contributed to data acquisition and analysis, critically revised the manuscript; H. Xu, X. Huang, contributed to data analysis and interpretation, critically revised the manuscript; Q. Dai, L. Jiang, contributed to conception and design, critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of the work.
Supplemental Material
sj-pdf-1-jdr-10.1177_00220345211021534 – Supplemental material for FOXO3 Mediates Tooth Movement by Regulating Force-Induced Osteogenesis
Supplemental material, sj-pdf-1-jdr-10.1177_00220345211021534 for FOXO3 Mediates Tooth Movement by Regulating Force-Induced Osteogenesis by A. Jin, Y. Hong, Y. Yang, H. Xu, X. Huang, X. Gao, X. Gong, Q. Dai and L. Jiang in Journal of Dental Research
Footnotes
Acknowledgements
The authors thank Dr. Qian Bian from the Shanghai Institute of Precision Medicine, Shanghai Ninth People’s Hospital, for useful discussions. The authors also thank the Shanghai Institute of Precision Medicine for bioimaging support.
A supplemental appendix to this article is available online.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported in part by grants from the National Natural Science Foundation of China (82071083, 81870740, 81570950, 81800949); the Program of Shanghai Academic/Technology Research Leader (20XD1422300); Clinical Research Plan of SHDC (SHDC2020CR4084); Shanghai “Rising Stars of Medical Talent” Youth Development Program-Outstanding Youth Medical Talents (SHWJRS2019-72); the Cross-disciplinary Research Fund of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine (JYJC201902); the Project of Biobank of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine (YBKB201909); the Innovation Research Team of High-Level Local Universities in Shanghai (SSMU-ZLCX20180501); the National Science Foundation of Shanghai (21ZR1436900); the SHIPM-mu Fund from the Shanghai Institute of Precision Medicine, Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine (JC201809).
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
