Dysregulated circARF3/miR-27b-3p/YAP1 contributes to impaired osteogenic differentiation of periodontal ligament stem cells in periodontitis

Abstract

BACKGROUND

Periodontitis is a prevalent chronic inflammatory disease whose progression can lead to the disrupted periodontal structures and subsequent tooth loss, severely impacting patients’ lives. Hence, exploring effective treatments for periodontitis is crucial. Periodontal ligament stem cells (PDLSCs) show promise for regenerating periodontal and alveolar bone tissues. Understanding the molecular mechanisms that control the osteogenic differentiation of PDLSCs is essential for developing regenerative therapies for periodontitis. Circular RNAs (circRNAs) are a novel type of noncoding RNA that form closed loops and are more stable than linear RNAs. Nevertheless, the relationship between circRNAs, periodontitis, and the osteogenic differentiation of PDLSCs has not been extensively studied.

AIM

To explore the roles of circRNA ADP-ribosylation factor 3 (circARF3) in the osteogenic differentiation of PDLSCs and periodontitis.

METHODS

Quantitative real-time polymerase chain reaction was utilized to compare RNA expression in periodontal ligament tissues from periodontitis patients and healthy individuals. PDLSCs were isolated, and flow cytometry was used to identify their surface markers. Alizarin red staining and alkaline phosphatase staining were performed to assess PDLSC osteogenic differentiation. Western blotting was employed to measure the expression of proteins related to osteogenic differentiation, including Runx2, Osterix, Osteocalcin, and Col1a1. The ENCORI website was applied to predict the interaction between miR-27b-3p and circARF3 and the 3’ untranslated region of Yes associated protein 1 (YAP1), and this interaction was validated using luciferase reporter gene assays, RNA immunoprecipitation, and RNA pulldown.

RESULTS

circARF3 expression was decreased in periodontal ligament tissues of patients with periodontitis compared with healthy individuals (P < 0.01). circARF3 knockdown resulted in decreased alkaline phosphatase activity, alizarin red staining, and osteogenic gene expression (P < 0.01). miR-27b-3p directly binds to circARF3 and the 3’ untranslated region of YAP1. Overexpression of miR-27b-3p mimicked, and depletion of YAP1 reversed, the effect of circARF3 on PDLSC osteogenic differentiation (P < 0.01).

CONCLUSION

Depletion of circARF3 suppresses PDLSC osteogenic differentiation in periodontitis via the miR-27b-3p/YAP1 axis.

Key Words: Periodontitis; Periodontal ligament stem cells; Osteogenic differentiation; Circular RNA; MicroRNAs; Yes associated protein 1

Core Tip: Periodontal ligament stem cells are promising candidate for the regenerative treatment of periodontal and alveolar bone tissues. Circular RNA ADP-ribosylation factor 3 was decreased in the periodontal ligament tissues of periodontitis patients. Circular RNA ADP-ribosylation factor 3 promotes the osteogenic differentiation of periodontal ligament stem cells in periodontitis by acting as a sponge for miR-27b-3p and increasing the expression of its target gene, Yes associated protein 1. This suggests it could be a potential therapeutic target for regenerative treatment of periodontitis.

INTRODUCTION

Periodontitis is a prevalent chronic inflammatory disease characterized by the breakdown of connective tissue surrounding the teeth and an inflammatory response triggered by microbes[1,2]. The progression of periodontitis can lead to disruption of the periodontal structures and subsequent tooth loss, significantly impacting patients’ quality of life[3]. Therefore, identifying effective treatments for periodontitis is crucial.

The periodontal ligament is a soft tissue that supports the teeth and contributes to orthodontic tooth movement, alveolar bone remodeling, and tooth nutrition[4]. Periodontal ligament stem cells (PDLSCs) are a type of mesenchymal stem cells (MSCs) derived from the periodontal ligament. They exhibit MSC characteristics, including self-renewal and the ability to differentiate into multiple cell types[5-7]. PDLSCs are readily accessible, possess strong bone-forming capabilities, and can be easily expanded, making them promising candidates for regenerating tooth-supporting bone[8]. A recent study revealed the potential of PDLSCs in regenerating supporting tissues in patients[9]. For example, in vivo studies demonstrated that PDLSCs can form cementum/PDL-like structures, periodontal ligaments, bone tissues, blood vessels, and peripheral nerves[10,11]. Notably, growing evidence suggests that inflammation impairs the regenerative capacity of human PDLSCs[12]. Consequently, understanding the molecular mechanisms that regulate the osteogenic differentiation of PDLSCs is critical for developing regenerative therapies for periodontitis.

Circular RNAs (circRNAs) are a novel type of noncoding RNA that form covalently closed continuous loops. They are more resistant to RNaseR degradation than linear RNAs[13]. Typically, circRNAs regulate gene function by directly competing with microRNAs (miRNAs) or by binding to RNA-binding proteins[14,15]. An increasing number of studies have demonstrated that circRNAs participate in various physiological and pathological processes, such as cellular activities, embryonic development, and multiple human diseases[16-18]. Nevertheless, the correlation of circRNAs with periodontitis and PDLSC osteogenic differentiation remains largely unexplored[19,20]. circRNA ADP-ribosylation factor 3 (circARF3) is a recently identified circRNA that modulates osteosarcoma cell growth and cell cycle[21]. Studies have demonstrated that circARF3 overexpression counteracts high glucose-induced renal mesangial cell fibrosis in a diabetic nephropathy model[22,23]. Other studies indicate that circARF3 also reduces dust mite-induced nasal mucosal cell apoptosis and inflammatory response[24]. Furthermore, circARF3 can improve neuroinflammation and obesity-associated metabolic disorders[25,26]. In this study, we explored the correlation of circARF3 expression with PDLSC osteogenic differentiation and periodontitis and the underlying molecular mechanisms. Here, we found that circARF3 is downregulated in periodontitis tissues compared with healthy controls. Ectopic circARF3 expression facilitated the osteogenic differentiation of PDLSCs by acting as a sponge for miR-27b-3p to upregulate Yes associated protein 1 (YAP1) expression. As a key component of the Hippo signaling pathway, YAP1 is essential for the osteogenic differentiation of MSCs, and its overexpression enhances their differentiation capacity[27]. We propose that circARF3 modulates osteogenic differentiation of PDLSCs and may be a promising therapy for periodontitis.

MATERIALS AND METHODS

Clinical tissues

Healthy donors (n = 17) and patients with periodontitis (n = 17) who underwent premolar extraction at Peking University School of Stomatology were recruited for this study. There were no statistically significant differences in age or sex between the two patient groups. Human periodontitis PDLSCs and healthy PDLSCs were isolated from periodontal ligament tissues. All participants provided written informed consent to allow the use of their tissues for this work. This study received approval from the Ethics Committee of China Medical University (No. 2024PS237K).

Cell culture and transfection

Periodontal ligament tissues were separated from the teeth roots, shredded, and digested with 1 mg/mL collagenase type I (Sigma, MO, United States) at 37 °C for 20 minutes. The cell suspension was filtered, and the obtained cells were cultured in low-glucose DMEM (Hyclone, UT, United States) containing 10% fetal bovine serum (Gibco, CA, United States) in a 37 °C incubator with 5% CO₂. Small hairpin RNAs against circARF3 (shcircARF3), YAP1 (shYAP1), the negative control (shNC), circARF3 overexpressing vector (ARF3 OE), miR-27b-3p mimics, and miR-27b-3p inhibitors were synthesized by RiboBio (China). Cell transfection was performed using Lipofectamine 3000 reagent (Invitrogen, CA, United States).

Osteoblastic induction

To perform osteoblastic induction, cells were treated with StemProTM Osteogenesis Differentiation Kit (Gibco, CA, United States). Briefly, PDLSCs were placed into 12-well plates and incubated in MSC growth medium for 5 days, followed by incubation with complete osteogenesis differentiation medium (Invitrogen, CA, United States).

Flow cytometry analysis

Cells were incubated with antibodies against PDLSC markers CD90, CD105, CD45, and CD34, and anti-IgG1 as a control. The samples were analyzed by Flow Cytometer (BD Biosciences, CA, United States).

Alkaline phosphatase staining

PDLSCs were placed into a 12-well plate, washed with phosphate buffered saline, and fixed with 4% paraformaldehyde for 30 minutes. The BCIP/NBT alkaline phosphatase (ALP) staining buffer (Beyotime, China) was added for 2 hours in the dark. ALP activity was quantified using a spectrophotometer at 562 nm.

Alizarin red staining

PDLSCs were washed with phosphate buffered saline, fixed with 4% paraformaldehyde for 30 minutes, followed by staining with 1% alizarin red S solution (Sigma, MO, United States). The bound dye was extracted with sodium phosphate containing 10% cetylpyridinium chloride and measured by a spectrophotometer at 405 nm.

Western blotting

Cells were lysed with RIPA buffer to obtain total proteins. An equal amount of 30 μg proteins was separated by sodium-dodecyl sulfate gel electrophoresis and blotted onto polyvinylidene difluoride membranes (Millipore, MA, United States). The blots were blocked with 5% skim milk (Sigma, MO, United States), incubated with primary antibodies against Runx2 (1:1000, 20700-1-AP, Proteintech, Wuhan, Hubei Province, China), Osterix (1:2000, 28694-1-AP, Proteintech, Wuhan, Hubei Province, China), Osteocalcin (1:1000, 20277-1-AP, Proteintech, Wuhan, Hubei Province, China), Colla1 (1:500, ab138492, Abcam, United Kingdom), or GAPDH (1:5000, 10494-1-AP, Proteintech, Wuhan, Hubei Province, China) overnight at 4 °C. The blots were visualized after incubation with corresponding secondary antibodies and ECL reagent (Millipore, MA, United States).

Quantitative real-time polymerase chain reaction

Total RNA was extracted from cells and tissues with TRIzol reagent (Thermo, MA, United States) following the manufacturer’s instructions. A total of 1 μg RNA was reverse-transcribed into cDNA using a cDNA Reverse Transcription Kit (Thermo, MA, United States). RNA levels were quantified using the SYBR Green system. GAPDH was used as an internal control.

Luciferase reporter gene assay

The potential binding sites of miR-27b-3p on circARF3 and YAP1 3’ untranslated region (3’UTR) were predicted using the ENCORI website. The sequences of circARF3 and YAP1 3’UTR with wild-type (WT) and mutated (MUT) putative binding sequence of miR-27b-3p were inserted into pmirGLO luciferase vectors (Promega, WI, United States). The plasmids and miR-27b-3p mimics were co-transfected into HEK-293T cells for 48 hours. The luciferase activity was measured using a Dual Luciferase Assay (Promega, WI, United States).

RNA immunoprecipitation assay

RNA immunoprecipitation assay was performed using an RNA Binding Protein Immunoprecipitation Kit (Millipore, MA, United States). Briefly, cells were lysed and incubated with Ago2-coated or IgG-coated Dynabeads at 4 °C overnight. The level of precipitated circARF3 was measured by quantitative real-time polymerase chain reaction (qRT-PCR).

RNA pulldown

Biotin-labeled WT and MUT miR-27b-3p were purchased from RiboBio (China) and transfected into cells. The cells were then lysed, sonicated, and incubated with magnetic beads (Invitrogen, CA, United States). The beads were washed and lysed in TRIzol reagent, and RNA level was quantified by qRT-PCR.

Statistical analysis

Data were presented as mean ± SD of at least three independent experiments and analyzed with SPSS software (version 16). Student’s t test and one-way analysis of variance (ANOVA) were conducted to analyze differences between two or more groups. P < 0.05 was considered statistically significant.

RESULTS

circARF3 expression is correlated with periodontitis

To evaluate the correlation between circARF3 expression and periodontitis, we isolated PDLSCs from healthy donors and patients with periodontitis and determined the levels of cell surface biomarkers by flow cytometry. As shown in Figure 1A and B, the isolated PDLSCs expressed high levels of CD90 and CD105 and low levels of CD45 and CD34. Subsequently, we identified decreased expression of circARF3 in periodontitis tissues compared to healthy tissues (Figure 1C). In addition, we found that circARF3 expression was upregulated during osteogenic induction, with levels peaking at day 7 (Figure 1D).

Figure 1 Circular RNA ADP-ribosylation factor 3 expression is correlated with periodontitis. A and B: Expression levels of periodontal ligament stem cell surface markers including CD90, CD105, CD45, and CD34 in periodontal ligament stem cell from periodontitis patients (A) and healthy donors (B) identified by flow cytometry; C: Circular RNA ADP-ribosylation factor 3 expression in periodontitis tissues and heathy samples. ^bP < 0.01, compared with healthy group; D: Level of circular RNA ADP-ribosylation factor 3 at 0 hour, 24 hours, 7 days, 14 days, and 21 days after osteogenic induction. ^aP < 0.05, compared with 0 hour; ^bP < 0.01, compared with 0 hour. pPDSLC: Periodontitis periodontal ligament stem cell; hPDSLC: Healthy periodontal ligament stem cell; ARF3: ADP-ribosylation factor 3.

circARF3 modulates osteogenic differentiation of PDLSCs

We next confirmed the function of circARF3 in osteogenic differentiation of PDLSCs by overexpressing or depleting circARF3. Transfection with circARF3 overexpressing plasmids or shcircARF3 successfully upregulated or downregulated the level of circARF3, respectively (Figure 2A and B). Results from alizarin red staining (ARS) and ALP staining indicated that circARF3 overexpression promoted the osteogenic activity of PDLSCs, while circARF3 depletion repressed the osteogenic activity (Figure 2B and C). Moreover, the levels of osteoblast makers Runx2, Osterix, Osteocalcin, and Col1a1, increased upon circARF3 overexpression and decreased upon circARF3 knockdown (Figure 2D).

Figure 2 Circular RNA ADP-ribosylation factor 3 modulates osteogenic differentiation of periodontal ligament stem cells. A: Level of circular RNA ADP-ribosylation factor 3 (circARF3) in periodontal ligament stem cells treated with circARF3 overexpressing plasmids and small hairpin RNAs that target circARF3. ^bP < 0.01, compared with shNC group or ovNC group; B and C: Alkaline phosphatase and alizarin red staining of periodontal ligament stem cells treated with circARF3-overexpressing plasmids and small hairpin RNAs that target circARF3. ^bP < 0.01, compared with ovNC group; ^cP < 0.01, compared with shNC group; D: The expression of osteoblast makers, Runx2, Osterix, Osteocalcin and Col1a1, was determined by western blotting assay. ^bP < 0.01, compared with ovNC group; ^cP < 0.01, compared with shNC group. ARF3: ADP-ribosylation factor 3; ARS: Alizarin red staining; ALP: Alkaline phosphatase.

circARF3 acts as sponge for miR-27b-3p

To investigate the regulatory mechanisms of circARF3 in the osteogenic differentiation of PDLSCs, we identified putative miRNA interactors of circARF3 through online analysis, and identified the potential binding between circARF3 and miR-27b-3p (Figure 3A). The transfection efficiency of miR-27b-3p mimics was confirmed by qRT-PCR assay (Figure 3B). Notably, miR-27b-3p repressed the luciferase activity of WT circARF3 but not the MUT circARF3 promoter vector (Figure 3C). Moreover, RNA pulldown and RNA immunoprecipitation assay demonstrated a direct interaction between circARF3 and miR-27b-3p (Figure 3D and E). Ectopic circARF3 expression resulted in decreased miR-27b-3p expression in PDLSCs (Figure 3F).

Figure 3 Circular RNA ADP-ribosylation factor 3 acts as sponge for miR-27b-3p. A: Potential binding site of miR-27b-3p on circular RNA ADP-ribosylation factor 3 (circARF3) predicted by ENCORI website; B: Level of miR-27b-3p determined by quantitative real-time polymerase chain reaction. ^bP < 0.01, compared with mimic NC group; C: Luciferase activity of wild-type (WT) and mutated (MUT) circARF3 promoter gene vectors. ^bP < 0.01, compared with mimic NC group; D: RNA pulldown assay to check interaction between biotin-labeled WT and MUT miR-27b-3p and circARF3. ^bP < 0.01, compared with Bio-NC group; ^cP < 0.01, compared with Bio-miR-27b-3pWT group; E: RNA immunoprecipitation assay to check interaction between miR-27b-3p and circARF3. ^bP < 0.01, compared with IgG group; F: Level of miR-27b-3p in periodontal ligament stem cells after transfection with circARF3-overexpressing vectors determined by quantitative real-time polymerase chain reaction. ^bP < 0.01, compared with ovNC group. ARF3: ADP-ribosylation factor 3; WT: Wild-type; MUT: Mutated.

circARF3 regulates PDLSC osteogenic differentiation via miR-27b-3p

We next evaluated whether miR-27b-3p participates in circARF3-meidated osteogenic differentiation. We transfected PDLSCs with miR-27b-3p mimics or inhibitors (Figure 4A). Similar to the effects of circARF3 overexpression, miR-27b-3p inhibition increased expression of osteogenic differentiation markers (Figure 4B and C) and osteogenic activity, confirmed by ALP and ARS staining (Figure 4D), while miR-27b-3p mimics impaired osteogenic differentiation. Importantly, miR-27b-3p levels were higher in periodontitis tissues compared with healthy tissues (Figure 4E).

Figure 4 Circular RNA ADP-ribosylation factor 3 regulates periodontal ligament stem cell osteogenic differentiation via miR-27b-3p. A: Level of miR-27b-3p in periodontal ligament stem cells transfected with miR-27b-3p mimics or inhibitors. ^bP < 0.01, compared with inhibitor NC group; ^cP < 0.01, compared with mimics NC group; B and C: The expression of osteoblast markers Runx2, Osterix, Osteocalcin and Col1a1, was determined by western blotting assay. ^bP < 0.01, compared with inhibitor NC group; ^cP < 0.01, compared with mimics NC group; D: Alkaline phosphatase and alizarin red staining of periodontal ligament stem cells treated with miR-27b-3p mimics or inhibitors. ^bP < 0.01, compared with inhibitor NC group; ^cP < 0.01, compared with mimics NC group; E: Level of miR-27b-3p in periodontitis tissues and heathy samples. ^bP < 0.01, compared with healthy group. ARS: Alizarin red staining; ALP: Alkaline phosphatase.

circARF3 affects PDLSC osteogenic differentiation via the miR-27b-3p/YAP1 axis

Prediction by an online website indicated the potential binding of miR-27b-3p to the 3’UTR region of the YAP1 promoter (Figure 5A). miR-27b-3p transfection suppressed WT YAP1 activity, but not MUT YAP1 promoter activity (Figure 5B), while simultaneously downregulating YAP1 mRNA levels (Figure 5C). Notably, YAP1 mRNA levels were significantly lower in periodontitis tissues compared with healthy tissues (Figure 5D). Moreover, YAP1 RNA and protein levels downregulated upon circARF3 depletion, which was rescued by co-transfection with miR-27b-3p inhibitors (Figure 5E and F). Evaluation of osteogenic differentiation suggested that administration of miR-27b-3p mimics and shYAP1 abolished the circARF3-elevated ARS and ALP staining (Figure 6A and B), as well as the expression of osteoblast makers Runx2, Osterix, Osteocalcin, and Col1a1 (Figure 6C and D).

Figure 5 miR-27b-3p targets the Yes associated protein 1 3’ untranslated region. A: Potential binding site of miR-27b-3p on Yes associated protein 1 (YAP1) 3’ untranslated region was predicted by ENCORI website; B: Luciferase activity of wild-type (WT) and mutated (MUT) YAP1 promoter gene vectors. ^bP < 0.01, compared with mimics NC group; C: Level of YAP1 in periodontal ligament stem cells after transfection with miR-27b-3p mimics was determined by quantitative real-time polymerase chain reaction. ^bP < 0.01, compared with mimics NC group; D: Level of YAP1 in periodontitis tissues and heathy samples. ^bP < 0.01, compared with healthy group; E and F: RNA and protein levels of YAP1 in periodontal ligament stem cells were determined by quantitative real-time polymerase chain reaction and western blotting assay, respectively. ^bP < 0.01, compared with shNC group; ^cP < 0.01, compared with shARF3 group. YAP1: Yes associated protein 1.

Figure 6 Circular RNA ADP-ribosylation factor 3 affects periodontal ligament stem cell osteogenic differentiation via miR-27b-3p/Yes associated protein 1 axis. A and B: Alkaline phosphatase and alizarin red staining of periodontal ligament stem cells; C and D: Expression of osteoblast makers, Runx2, Osterix, Osteocalcin and Col1a1 was determined by western blotting assay. ^bP < 0.01, compared with ovNC group; ^cP < 0.01, compared with ovARF3 group. ARF3: ADP-ribosylation factor 3; ARS: Alizarin red staining; ALP: Alkaline phosphatase.

DISCUSSION

Taken together, periodontal tissues from periodontitis have decreased circARF3 and YAP1 RNA levels and increased miR-27b-3p levels compared with healthy periodontal tissues. Molecular analysis demonstrated that miR-27b-3p could bind to circARF3 and the 3’UTR region of YAP1. circARF3 acted as a sponge for miR-27b-3p, increasing YAP1 expression and promoting osteogenic differentiation of PDLSCs. Recent studies have extensively investigated stem cell dysfunction in various diseases[28,29]. As MSCs, PDLSCs can self-renew, differentiate into multiple cell types, and contribute to the regeneration of periodontal and alveolar tissues[30].

Noncoding RNAs, including long noncoding RNA, miRNAs, and circRNAs, have crucial roles in numerous developmental processes, such as erythropoiesis, stem cell pluripotency regulation, apoptosis, and keratinocyte differentiation[14]. As the understanding of noncoding RNA functions grows, it becomes essential to explore their role in MSC commitment. Increasing evidence indicates that noncoding RNAs influence osteogenic differentiation[31,32]. For example, miR-24-3p inhibits osteogenic differentiation by regulating Smad5 expression[33]. Long noncoding RNA MSC-AS1 promotes osteogenic differentiation of bone marrow-derived MSCs and alleviates osteoporosis progression through the miR-140-5p/bone morphogenetic protein 2 regulatory axis[34]. Nevertheless, the study of circRNAs and PDLSC osteogenic differentiation is limited. circCDK8 overexpression represses PDLSC osteogenic differentiation by promoting mammalian target of rapamycin (mTOR)-regulated autophagy under hypoxic conditions[19]. circCDR1as acts as a sponge for miR-7, increasing growth differentiation factor 5 (Gdf5) levels and subsequently activating Smad1/5/8 and p38 mitogen-activated protein kinases phosphorylation to promote PDLSC osteogenic differentiation[20]. In this study, we demonstrated that circARF3 expression was decreased in periodontitis tissues and that increasing circARF3 levels facilitated PDLSC osteogenic differentiation.

To further explore how circARF3 regulates bone formation in PDLSCs, we evaluated potential miRNAs that interact with circARF3. Our data demonstrate that miR-26b-3p competes with circARF3 to modulate YAP1 expression and function in PDLSC osteogenic differentiation. miR-27b-3p levels were higher in periodontitis tissues compared with healthy samples, while YAP1 expression was lower. YAP1 is the critical downstream component of Hippo pathway, which regulates tissue homeostasis, organ size, and tumor development[35,36]. YAP1 phosphorylation affects its localization within the cell, regulates downstream gene expression, and modulates various cellular functions[37]. Moreover, crosstalk between Hippo-YAP signaling and other pathways affects osteogenic ability and host immune responses[38,39]. Tumor necrosis factor-α can inhibit PDLSC proliferation and osteogenic differentiation, but this can be reversed by the YAP1 gene through the Hippo signaling pathway[40]. Additionally, calcitonin gene-related peptide-alpha-transfected healthy PDLSCs showed elevated expression of osteogenic phenotypes, along with increased YAP expression[41]. We revealed that YAP1 mediates circARF3-modulated PDLSC osteogenic differentiation in periodontitis.

This study suggests that circARF3 may be a promising therapeutic target for regenerative therapy of periodontitis. Mechanistically, circARF3 promotes PDLSC osteogenic differentiation by sequestering miR-27b-3p and upregulating YAP1 expression. However, these conclusions were only verified through in vitro cell experiments. Future research will be conducted using animal models to confirm these findings in vivo.

CONCLUSION

In conclusion, circARF3 sponged miR-27b-3p to upregulate YAP1 expression, thereby promoting the osteogenic differentiation of PDLSCs in periodontitis.