MicroRNA-27a transfected dental pulp stem cells undergo odonto/osteogenic differentiation via targeting DKK3 and SOSTDC1 in Wnt/BMP signaling in vitro and enhance bone formation in vivo

Abstract

Background

MicroRNAs (miRNAs) play a crucial role in cell differentiation through epigenetic regulation of gene expression. In human dental pulp cells, we have identified miRNA-27a being upregulated under inflammatory conditions. Here, we aimed to examine whether (i) overexpression of miRNA-27a in human dental pulp stem cells (hDPSCs) enhances their odonto/osteoblastic differentiation via Wnt and bone morphogenetic protein signaling; and (ii) hDPSCs overexpressing miRNA-27a promote new bone formation in vivo.

Methods

hDPSCs were cultured in osteogenic medium to promote differentiation. To examine the role of miRNA-27a, hDPSCs were transfected with either a miRNA-27a mimic to enhance or an inhibitor to suppress miRNA-27a expression. Odonto/osteoblastic differentiation was assessed by evaluating the expression of specific markers, Wnt and bone morphogenetic protein (BMP) signaling molecules, and mineralization capacity using RT-qPCR, western blotting, Alizarin Red S (ARS) staining, and alkaline phosphatase (ALP) activity. Potential miRNA-27a binding sites in the 3'UTRs of DKK3 and SOSTDC1 were identified via bioinformatics analysis and validated through the luciferase reporter assay. In vivo, miRNA-27a-overexpressing hDPSCs were seeded into collagen honeycomb scaffolds and implanted into mouse calvarial bone cavities to assess new bone formation.

Results

MiRNA-27a was highly upregulated in hDPSCs committed to odonto/osteoblastic differentiation. Overexpression of miRNA-27a led to increased expression of odonto/osteoblastic markers and enhanced mineralization capacity, while inhibition of miRNA-27a had the opposite effect. MiRNA-27a targeted DKK3, promoting β-catenin nuclear translocation and inhibiting SOSTDC1, which enhanced SMAD1/5 phosphorylation. Binding sites for miRNA-27a were identified in the 3'UTRs of DKK3 and SOSTDC1. In vivo, miRNA-27a-overexpressing hDPSCs promoted new bone formation in mouse calvaria bone cavities.

Conclusion

Transfection of miRNA-27a in hDPSCs enhanced their odonto/osteoblastic differentiation by targeting DKK3 and SOSTDC1, thereby promoting the Wnt and BMP signaling. Transplantation of miRNA-27a-overexpressing hDPSCs promoted new bone formation in vivo. These findings deepen our understanding of the effects of miRNA on Wnt and BMP pathways and suggest a potential clinical application for miRNA-27a in promoting hard tissue regeneration, offering a promising therapeutic target for dental and craniofacial tissue reconstruction.

Background

The dental pulp is typically shielded by a mineralized layer of dentin and enamel that ensures its protection from external threats [1]. However, advanced caries lesions or trauma allow bacteria or bacterial products to enter the pulp through the dentin tubules, causing pulpal inflammation. This condition is exaggerated by continuous bacterial invasion, and the necrosis or apoptosis of odontoblasts and the infiltration of inflammatory cells such as macrophages can be induced [2]. This pulpal inflammation is characterized by various mediators that are synthesized from immunocompetent cells and pulpal fibroblasts, which are stimulated by bacterial products [3]. Inflammatory mediators were traditionally associated only with tissue destruction; however, they are involved in the initiation of tissue regeneration and the process of inflammation regulation [4].

As dental caries progresses, bacteria invade the dentinal tubules, triggering the development of pulpitis. If untreated, the infection can lead to irreversible pulpitis and subsequent development of apical periodontitis, leading to a progressive bone resorption in the periapical area [5]. Once the infection is controlled, the bone defect may be repaired through new osteogenesis; however, regenerative approaches may enhance the healing of large bone defects, particularly those created after periapical surgery [6]. The demand for bone tissue engineering continues to grow, since autologous bone grafts face challenges such as scarcity of sources and immune rejection [7]. Bone tissue engineering requires the combination of osteogenic cells, 3D (three-dimensional) scaffolds, and bone-inductive chemicals to create mature bone structures [8]. Hence, optimizing stem cells, scaffolds and signaling molecules could provide a more effective alternative to autologous bone grafting for dental tissue regeneration [9].

Dental pulp stem cells (DPSCs) have garnered extensive interest as tissue engineering seed cells because of their abundant sources, ease of isolation and acquisition, low immunogenicity, the absence of ethical controversy, and low transplant rejection rate [10, 11]. Culturing human DPSCs (hDPSCs), a type of human mesenchymal stem cells, in an appropriate induction medium can effectively stimulate their differentiation process, promoting their transition into specific lineages such as osteoblasts [12]. Exosomes produced from DPSCs (DPSC-EXO) have been shown to prevent periodontitis in rats and to encourage the healing of alveolar bone [13]. The clinical regeneration of pulp-like tissue with many blood capillaries and dentin occurs by combining autologous DPSCs with leukocyte platelet-rich fibrin [14]. The application of deciduous DPSCs that contain hydroxyapatite–collagen to alveolar bone deficiencies during secondary dental eruption has been demonstrated to clinically promote bone repair [15]. DPSCs exhibit diverse differentiation potentials and have a strong inherent capacity for differentiation into hard tissue-forming cells. In dense culture conditions, activation of integrin signaling enhances the expression of hard tissue-forming cell markers in hDPSCs [16]. Furthermore, under hypoxic conditions, Wnt/β-catenin signaling is activated by HIF1α/BCL9, leading to the increased expression of hard tissue-forming cell markers [17]. Epigenetic processes play a crucial role in regulating stemness and the self-renewal of DPSCs and influencing their lineage commitments [18].

MicroRNAs (miRNAs) are small, non-coding RNAs that are essential transcriptome epigenetic regulators. MiRNAs typically bind to particular sites in the 3′ untranslated region (UTR) of target mRNAs, thereby suppressing targeted gene expression [19]. The capacity to regulate multiple transcripts allows miRNAs to change the transcriptome quickly, thoroughly, and reversibly in response to external stimuli such as inflammation. This capability suggests that miRNAs can be broadly considered inflammatory mediators [20]. The important regulatory role of miRNAs and their products in tissue regeneration is currently well-recognized [21]. However, the precise mechanism of the epigenetic modulation of miRNAs in the odonto/osteoblastic differentiation of DPSCs remains unclear.

The expression of miRNAs in hDPSCs stimulated by lipopolysaccharide (LPS) has been evaluated in our previous study [22], and miRNA-27a has been identified as one of the highly synthesized miRNAs. MiRNA-27a is a promising therapeutic target and an encouraging biomarker in a range of tumor types [23]. Additionally, miRNA-27a has been shown to reduce bone resorption and osteoclast formation by targeting peroxisome proliferator-activated receptor γ and adenomatous polyposis coli [24]. We have also demonstrated an anti-inflammatory function of miRNA-27a through suppressing the nuclear factor kappa B (NF-κB) signaling pathway [25]. By modulating the release of transforming growth factor（TGF）-β and other pro-inflammatory cytokines, NF-κB stimulates inflammation and fibrosis [26]. NF-κB activation also has been demonstrated to decrease osteoblast activity [27, 28] and to be a target for the treatment of inflammation-related bone disorders [29]. tumor necrosis factor (TNF)α, as a key inflammatory cytokine [30], inhibits bone morphogenetic protein (BMP) signaling during osteoblast development, which is closely linked to the NF-κB pathway [31]. Moreover, NF-κB directly blocks the binding of Wnt-induced β-catenin and BMP-induced runt-related transcription factor 2 (RUNX2) to components of a consensus response, decreasing the transcription of osteocalcin and bone sialoprotein [32]. These findings imply that miR-27a may be crucial for odonto/osteoblastic differentiation.

Based on the above insights, we hypothesized that miRNA-27a stimulates the odonto/osteoblastic differentiation of hDPSCs via Wnt and bone morphogenetic protein (BMP) signaling. In this study, we aimed to examine whether overexpression of miRNA-27a in hDPSCs enhances their odonto/osteoblastic differentiation via Wnt and bone morphogenetic protein signaling. We also aimed to demonstrate that hDPSCs overexpressing miRNA-27a promote new bone formation in vivo, providing fundamental data for a novel bone regeneration strategy.

Materials and methods

All experiments were authorized by the Tokyo Medical and Dental University Ethical Committee (#D2023-066) and conducted in compliance with the Ethical Guidelines for Clinical Studies. Each participant provided informed consent in accordance with the Ethical Guidelines for Clinical Studies.

Cell culture

Freshly extracted wisdom teeth (n = 7) were used to produce hDPSCs, which were grown in α-modified minimum essential medium (Fujifilm Wako Pure Chemical, Osaka, Japan) that was supplemented with an antibiotic and antifungal solution (Fujifilm Wako) and 10% fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA) under standard conditions (37 °C, 5% CO₂) that have been previously described [16, 17]. hDPSCs from passages two through six were used and the culture media were replaced every 3 days. The precise isolation methods are detailed in the supplementary materials and methods, while the properties of DPSCs are presented in supplementary Fig. 1. LiCl (20 mM), which inhibits GSK-3β [33], was used as a Wnt signaling activator.

miRNA-27a-5p mimic/inhibitor transfection

hDPSCs were transfected by the mirVana miRNA mimic hsa-miRNA-27a-5p (miRNA-27a mimic; Thermo Fisher Scientific), and the mirVana miRNA inhibitor hsa-miRNA-27a-5p (miRNA-27a inhibitor; Thermo Fisher Scientific) was transfected with Lipofectamine RNAiMAX transfection reagent (Thermo Fisher Scientific). Two separate controls were used: the mirVana miRNA mimic negative control #1 (Thermo Fisher Scientific) and the mirVana miRNA inhibitor negative control #1 (Thermo Fisher Scientific).

Scratch wound assay

To evaluate cell migration in vitro, the wound healing assay was performed. hDPSCs were seeded into a 24-well plate at a density sufficient to achieve 90–100% confluence within 24 h. A sterile 200 µL pipette tip was used to produce a scratch once the cells reached confluence. The detached cells were carefully removed by washing twice with phosphate-buffered saline (PBS). Then the cells were cultured in serum-free medium to eliminate the influence of growth factors in the added fetal bovine serum, allowing for migration to be the primary factor in closing the scratch. An inverted microscope was used to capture images of the scratch area, and the wound area was determined using ImageJ (Version 1.54i, https://imagej.net/ij/) by manually tracing the cell-free area in the images.

Odonto/osteoblastic differentiation

Odonto/osteoblastic differentiation was induced by culturing hDPSCs in an osteogenic medium containing 10 mmol/L β-glycerophosphate (Fujifilm Wako), 50 mg/L ascorbic acid (Fujifilm Wako), and 10 nmol/L dexamethasone (Fujifilm Wako). The culture medium was replaced every 2 days. Alkaline phosphatase (ALP) activity was measured using the LabAssay ALP kit (Fujifilm Wako). Alizarin Red S staining was performed to detect the mineralized nodule formation. Cells were fixed using methanol (Fujifilm Wako) and stained with Alizarin red S (1% solution, Fujifilm Wako, pH was adjusted at 6.4 by diluted ammonia solution). ALP staining was performed to detect the activity of alkaline phosphatase. Samples were fixed in 4% paraformaldehyde (Fujifilm Wako) for 5 min and then stained with an ALP staining solution containing 0.5% N, N-dimethylformamide (Fujifilm Wako), 0.1 mg/mL naphthol ASMX phosphate (Sigma-Aldrich, St. Louis, MO, USA), 0.6 mg/mL of fast blue BB salt (Sigma-Aldrich) and 2 mM MgCl₂ (Fujifilm Wako) in 0.1 mM Tris (pH 8.5) for 10 min. Further quantification was performed using ImageJ.

RNA extraction and quantitative real-time polymerase chain reaction

The reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) PrimeScript™ RT Master Mix (Takara Bio, Kusatsu, Japan) and the QuickGene RNA cultured cell kit S (Fujifilm Wako) were utilized for RNA extraction and cDNA synthesis, respectively. The CFX96 Real-Time qPCR System (Bio-Rad, Hercules, CA, USA) was utilized to conduct RT-qPCR using cDNA, particular primers (Supplementary Table 1), and GoTaq qPCR Master Mix (Promega, Madison, WI, USA). Internal control was achieved using actin beta.

For the evaluation of miRNA-27a expression, the mirVana miRNA isolation kit MicroRNA Reverse Transcription Kit (Thermo Fisher Scientific), and the specific reverse transcription primers for U6 (the internal control) and hsa-miRNA-27a-5p were used for RNA extraction and cDNA synthesis, respectively. The CFX96 Real-Time qPCR System (Bio-Rad) was utilized for RT-qPCR using the TaqMan Universal Master Mix II (Thermo Fisher Scientific).

Western blotting

A protease inhibition cocktail (Complete; Sigma-Aldrich) and a phosphatase inhibition cocktail (PhosSTOP; Sigma-Aldrich) were added to the radioimmunoprecipitation buffer. Nuclear and cytoplasmic fractions were extracted using a NE-PER nuclear and cytoplasmic extraction reagent (Thermo Fisher Scientific). Samples that were separated in the electrophoresis gels were transferred to a polyvinylidene fluoride membrane (Immobilon P; Merck Millipore, Burlington, MA, USA). Membranes were blocked with PVDF blocking reagent (Toyobo, Osaka, Japan) and probed with primary antibodies overnight at 4 °C: anti-β-catenin (1:1000, 15B8, #37447, monoclonal, mouse; Cell Signaling Technology, Danvers, MA, USA), anti-Lamin B1 (1:1000, GTX103292, polyclonal, rabbit; GeneTex, Irvine, CA, USA), anti-dickkopf (DKK) 3 (1:1000, #10365-1-AP, polyclonal, rabbit, Proteintech Group, Inc., Rosemont, IL, USA), anti-sclerostin domain containing (SOSTDC) 1 (1:1000, ab99340, polyclonal, rabbit; Abcam, Cambridge, UK),anti-lymphoid enhancer binding factor (LEF) 1 (1:1000, C12A5, #2230, monoclonal, rabbit; Cell Signaling Technology), anti-transcription factor (TCF)7 (1:1000, C63D9, #2203, monoclonal, rabbit; Cell Signaling Technology), anti-suppressor of mothers against decapentaplegic (SMAD)4 (1:1000, #9515, monoclonal, rabbit; Cell Signaling Technology), anti-phospho-SMAD1/5 (1:1000, 41D10, #9516, monoclonal, rabbit; Cell Signaling Technology), HRP-conjugated anti-tubulin (1:4000, PM054-7; Medical & Biological Laboratories, Nagoya, Japan), and horseradish peroxidase (HRP)-conjugated anti-Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH, 1:4000, PM053-7; Medical & Biological Laboratories, Nagoya, Japan). Following a wash with Tris-buffered saline containing Tween 20 (0.1% v/v, Fujifilm Wako), secondary antibodies were applied: HRP-conjugated anti-rabbit IgG (1:4000, W4011; Promega) and HRP-conjugated anti-mouse IgG (1:4000, W4021; Promega). Chemiluminescence was detected using Immobilon substrate (Merck Millipore) and the LAS-3000 mini-imaging system (Cytiva, Malborough, MA, USA) was used to capture images. Band intensity was quantified using ImageJ.

Immunofluorescence

Approximately 2.5 × 10⁴ hDPSCs were seeded per well in an 8-well chamber slide (IbidiFitchburg, Fitchburg, WI, USA), and fixed at 4 °C for 10 min with 4% paraformaldehyde. Samples were blocked for 30 min with 10% normal donkey serum (Abcam) then with 0.1% Triton X-100 (Fujifilm Wako) for permeabilization. Subsequently, primary antibodies including anti-β-catenin (1:500, 15B8, #37447, monoclonal, mouse; Cell Signaling Technology), anti-DKK3 (1:200, #10365-1-AP, polyclonal, rabbit, Proteintech Group, Inc.), anti-RUNX2 (1:1000, D1L7F, #12556, monoclonal, rabbit; Cell Signaling Technology), anti-Osteocalcin (OC, 1:500, E8B9X, #59757, monoclonal, rabbit; Cell Signaling Technology), anti-SOSTDC1 (1:500, ab99340, polyclonal, rabbit; Abcam), anti-SMAD4 (1:500, #9515, monoclonal, rabbit; Cell Signaling Technology), and anti-phospho-SMAD1/5 (1:500, 41D10, #9516, monoclonal, rabbit; Cell Signaling Technology) were added to the samples and incubated overnight. Alexa Fluor 568-conjugated anti-mouse IgG (1:500, donkey; ab175472, Abcam) and Alexa Fluor 488-conjugated anti-rabbit IgG (1:500, donkey; ab150073, Abcam) were used as secondary antibodies. As a nuclear counterstain, DAPI was applied (flouroshield mounting media with 4′,6-diamidino-2-phenylindole; Abcam, Cambridge, UK). LAS AF confocal software (Version 1.8.3; Leica Microsystem, Wetzlar, Germany) was used to examine images captured by a confocal laser scanning microscope (Leica TCS-SP8; Leica Microsystems) for histological evaluation.

Luciferase assay

The pGL4.49[luc2P/TCF-LEF/Hygro] vector (Promega) containing a transcription factor (TCF)–lymphoid enhancer-binding factor (LEF) response element was used to measure Wnt signaling activity. The pMIR-REPORT vectors (Thermo Fisher Scientific) were used for the 3' UTR assay of DKK3 and SOSTDC1. These vectors contained synthesized 400 bp fragments of the 3′-UTR of DKK3 or SOSTDC1, with or without mutations at the predicted miR-27a binding site, inserted into the XhoI and HindIII sites.

Animal experiments

The Animal Committee of Tokyo Medical and Dental University authorized all procedures of animal experiments (approval number: A2019-297A). Animal Research: Reporting of In Vivo Experiments (ARRIVE) 2.0 guidelines were adhered to in this animal study. Every attempt was made to lessen the suffering of animals and the quantity of animals used.

hDPSCs (1 × 10⁵ cells) transfected with the miRNA-27a mimic or the NC were seeded into the 3D Honeycomb Boosted collagen scaffold (Koken, Tokyo, Japan) and were transplanted into the bone cavities formed in the calvaria and back subcutaneous tissues of male Institute of Cancer Research (ICR) mice (n = 6, CLEA Japan, Inc., Tokyo, Japan). Ketamine (Daiichi Sankyo Company, Tokyo, Japan)/xylazine (Fujifilm Wako) (100/10 mg/kg body weight) was used for anesthesia. For cavity preparation in the calvaria, a low-speed handpiece equipped with a trephine bur was applied to create a 2.5 mm diameter circular defect on the left (miRNA mimic) and right (control) sides. The detailed information regarding the surgical procedures is provided in the supplementary materials and methods.

After mice were sacrificed by CO₂ inhalation, the entire calvaria and the back dermal tissues were extracted at 28 days and fixed at 4 °C for 24 h with 4% paraformaldehyde. Calvaria were demineralized in 0.5 M ethylenediamine tetra acetic acid (EDTA) at 4 °C for 3 weeks. Elastic Verhoeff-Van Gieson staining, and hematoxylin and eosin staining were applied on cryostat sections. ALP staining on sections was performed using the same method as for cultured DPSCs. Micro-computed tomography images (inspeXio SMX-100CT; Shimadzu, Kyoto, Japan) were taken on days 14 and 28 using the following parameters: a 0.05 mm voxel image, a 1.0 mm thick aluminum filter, a current of 40 μA, and a voltage of 85 kV. Subsequently, Amira 5.4.4 software (Visage Imaging, CA, USA) was utilized to examine the three-dimensional structure of newly formed bone at surgical locations.

Histological evaluation

The hematoxylin and eosin (H&E) staining was performed following standard protocols. The ALP staining on sections was conducted using the same method as described for the in vitro experiment. For immunohistochemical staining, sections were incubated with the following primary antibodies: anti-BCL9 (1:400, bs-12393R, polyclonal, rabbit; Bioss Antibodies), anti-β-catenin (1:500, 15B8, #37447, monoclonal, mouse; Cell Signaling Technology), anti-Osteocalcin (OC, 1:500, E8B9X, #59757, monoclonal, rabbit; Cell Signaling Technology) and anti-pSmad1/5 (1:500, 41D10, #9516, monoclonal, rabbit; Cell Signaling Technology). The primary antibodies were applied overnight at 4 °C. HRP-conjugated secondary antibodies, including anti-rabbit IgG (1:500, W4011; Promega) and anti-mouse IgG (1:500, W4021; Promega) were used as secondary antibodies. A DAB substrate kit (3,3′-diaminobenzidine; Abcam, ab64264) was used to detect immunoreactivity in accordance with the manufacturer's instructions. To prevent overstaining, the DAB chromogen was applied for 3 to 5 min at room temperature while the reaction was observed under a light microscope. Finally, slides were counterstained with methyl green for 10 min, followed by dehydration and mounting for observation.

Software and statistical analysis

Targetscan (http://www.targetscan.org/) was used to predict miRNA-27a targets. Statistical analysis was conducted using GraphPad Prism 9.0 (GraphPad Software, CA, USA). Data analysis techniques included the Mann–Whitney U test, one-way ANOVA followed by Tukey’s post hoc test, and the unpaired Student's t-test. p < 0.05 was considered statistically significant. At least three tests for each experiment were conducted.

Results

miRNA-27a-5p was highly elevated during hDPSC osteogenic differentiation and enhanced cell migration

Using bioinformatics methods, the binding sites of miR-27a-5p were predicted through the miRWalk, mirDIP, and TargetScan databases, identifying 233 genes as potential target genes of miR-27a-5p (Fig. 1A). Gene Ontology (GO) analysis revealed that these target genes are significantly associated with immune response, various cell differentiation processes, and tissue developments (Fig. 1B).

Fig. 1

miRNA-27a is involved in multiple cell differentiation and promoted cell migration. A miR-27a-5p target genes predicted by the TargetScan, mirDIP and miRWalk algorithms. B Thorough examination using GO analysis of 233 target genes. C Scratch wound assay. The area that remained uncovered by the cells represents the closing of the wound (upper and middle panels). miR-27a overexpressing hDPSCs showed significantly higher migration compared to negative control (mean ± SD, n = 3). D The expression of miRNA-27a was upregulated in hDPSCs that were cultured in an osteogenic medium on day 3 (mean ± SD, n = 3). Statistical analysis was performed using an unpaired two-tailed Student’s t-test. Significance level is indicated as follows: *p < 0.05. hDPSCs: human dental pulp stem cells; mimic NC: miRNA mimic negative control #1; miR-27a mimic: miRNA mimic for hsa-miRNA-27a-5p

miR-27a overexpressing hDPSCs demonstrated improved migration in the same timeframe in the scratch wound assay (Fig. 1C). To evaluate the expression of miRNA-27a during osteogenic differentiation, hDPSCs were treated with an osteogenic medium for 3 days and miRNA-27a was markedly elevated (Fig. 1D).

miRNA-27a-5p promoted the odonto/osteoblastic differentiation of hDPSCs in vitro

The expression of odonto/osteoblastic markers, such as osteopontin (OPN), osteocalcin (OC), BMP4, RUNX family transcription factor 2 (RUNX2), Sp7 transcription factor (SP7), dentin sialophosphoprotein (DSPP), dentin matrix protein 1 (DMP1), and ALP, was significantly upregulated in miRNA-27a mimic-transfected hDPSCs (Fig. 2A). Furthermore, ALP activity (Fig. 2B, C) and mineralized nodule formation (Fig. 2D) were upregulated in miRNA-27a mimic-transfected hDPSCs.

Fig. 2

miRNA-27a promoted the odonto/osteoblastic differentiation in hDPSCs. A The expression of osteoblastic marker mRNAs was upregulated in miRNA-27a mimic-transfected hDPSCs (mean ± SD, n = 3). B Alkaline phosphatase (ALP) activity, C ALP staining and D mineralized nodule formation stained by the alizarin red solution were upregulated in miRNA-27a mimic-transfected hDPSCs (mean ± SD, n = 3). Statistical analysis was performed using an unpaired two-tailed Student’s t-test. Significance levels are indicated as follows: *p < 0.05, **p < 0.01, and ***p < 0.001. hDPSCs: human dental pulp stem cells; mimic NC: miRNA mimic negative control #1; miR-27a mimic: miRNA mimic for hsa-miRNA-27a-5p; OPN: Osteopontin; OC: Osteocalcin; SP7: Transcription factor Sp7, also called osterix; DSPP: Dentin sialophosphoprotein; DMP1: Dentin matrix protein 1; RUNX2: Runt-related transcription factor 2; BMP4: Bone morphogenetic protein 4

miRNA-27a stimulated the Wnt pathway in hDPSCs

Enforced expression of miRNA-27a promoted Wnt signaling in hDPSCs with or without LiCl, a potent activator of the Wnt/β-catenin signaling pathway (Fig. 3A). In contrast, the miRNA-27a inhibitor downregulated Wnt signaling in hDPSCs with LiCl (Fig. 3A). The enforced expression of miRNA-27a downregulated the mRNA expression of axis inhibition protein 2 (AXIN2) and adenomatous polyposis coli (APC)—the negative regulators of Wnt signaling—and upregulated transcription activators transcription factor 7 (TCF7) and lymphoid enhancer binding factor 1 (LEF1) (Fig. 3B). Protein expression of TCF7 and LEF1 was also promoted in miRNA27a-transfected hDPSCs (Fig. 3B). Furthermore, western blotting and immunofluorescent staining were used to confirm the nuclear translocation of β-catenin (Fig. 3C, D).

Fig. 3

The miRNA-27a mimic upregulated the Wnt signaling pathway in hDPSCs. A Wnt reporter activity was promoted by the miRNA-27a mimic with or without LiCl, an agonist of canonical Wnt signaling. Promoted Wnt reporter activity by LiCl was suppressed by the miRNA-27a inhibitor (mean ± SD, n = 4). B The miRNA-27a mimic downregulated the mRNA expression of axis inhibition protein 2 (AXIN2) and adenomatous polyposis coli (APC) and upregulated the mRNA expression of transcription factor 7 (TCF7) and lymphoid enhancer binding factor 1 (LEF1) (mean ± SD, n = 4). C, D The miRNA-27a mimic promoted the expression of β-catenin in hDPSCs. In particular, nuclear translocation of β-catenin was detected using western blotting on nuclear fraction (mean ± SD, n = 3) (C) and immunofluorescence (D). Arrows indicate localized β-catenin in the nucleus. Statistical analysis was performed using a one-way ANOVA followed by Tukey’s post hoc test and an unpaired two-tailed Student’s t-test. Significance levels are indicated as follows: *p < 0.05, **p < 0.01, and ***p < 0.001. hDPSCs: human dental pulp stem cells; LiCl: lithium chloride, which enhances Wnt signaling by inhibiting GSK-3b, was used as Wnt signaling activator; mimic NC: miRNA mimic negative control #1; miR-27a mimic: miRNA mimic for hsa-miRNA-27a-5p; inhibitor NC: miRNA inhibitor negative control #1; miR-27a inhibitor: miRNA inhibitor for hsa-miRNA-27a-5p; RLU: relative light unit; DAPI: 4′,6-diamidino-2-phenylindole. Scale bars: 25 μm

DKK3 and SOSTDC1 were targets of miRNA-27a

DKK3 and SOSTDC1 were identified as potential targets of miRNA-27a, and binding sites were predicted using Targetscan. The overexpression of miRNA-27a significantly downregulated the expression of DKK3 and SOSTDC1 (both mRNA and protein; Fig. 4A, C, D); the downregulation was reversed upon inhibition of miRNA-27a expression (Fig. 4B, E). miRNA-27a downregulated the luciferase activity of wild-type DKK3 and SOSTDC1 3’ UTR-inserted pMIR-REPORT vectors; the downregulation was canceled by inserting the mutation in the predicted miRNA-27a binding site in DKK3 and SOSTDC1 (Fig. 4F).

Fig. 4

The miRNA-27a mimic downregulated DKK3 and SOSTDC1 in hDPSCs. The miRNA-27a mimic downregulated the mRNA (A) and protein (C) of DKK3 and SOSTDC1 in hDPSCs. Immunofluorescence also indicated the upregulation of DKK3 and SOSTDC1 in the miRNA-27a mimic-transfected hDPSCs (D) However, the miRNA-27a inhibitor upregulated the mRNA (B) and protein (E) of DKK3 and SOSTDC1 (mean ± SD, n ≥ 3). F The hsa-miRNA-27a-5p target sequences within DKK3 3′- untranslated region (UTR) and SOSTDC1 3′-UTR of wild-type (DKK3 wt; SOSTDC1 wt) and mutant (DKK3 mut; SOSTDC1 mut) were indicated. The 3’ UTR activity of DKK and SOSTDC1 was downregulated by the miRNA-27a mimic, which was attenuated by their mutation (mean ± SD, n = 4). Statistical analysis was performed using a one-way ANOVA followed by Tukey’s post hoc test and an unpaired two-tailed Student’s t-test. Significance levels are indicated as follows: *p < 0.05, **p < 0.01, and ***p < 0.001. hDPSCs: human dental pulp stem cells; mimic NC: miRNA mimic negative control #1; miR-27a mimic: miRNA mimic for hsa-miRNA-27a-5p; inhibitor NC: miRNA inhibitor negative control #1; miR-27a inhibitor: miRNA inhibitor for hsa-miRNA-27a-5p; RLU: relative light unit; DAPI: 4′,6-diamidino-2-phenylindole. Scale bars: 50 μm

miRNA-27a stimulated the BMP pathway in hDPSCs

miRNA-27a downregulated the mRNA expression of gremlin 1 (GREM1), a secreted antagonist of bone morphogenetic proteins [34] (Fig. 5A). The expression of phosphorylated small mothers against decapentaplegic 1/5 (pSMAD1/5) and SMAD4 was upregulated by miRNA-27a (Fig. 5B, C).

Fig. 5

The miRNA-27a mimic upregulated bone morphogenetic protein (BMP) signaling pathway in hDPSCs. A Transfection of the miRNA-27a mimic downregulated the expression of GREM1 (mean ± SD, n = 3). B Transfection of the miRNA-27a mimic upregulated the expression of phosphorylated small mothers against decapentaplegic 1/5 (pSMAD1/5) and SMAD4 using western blotting (mean ± SD, n = 3) and immunofluorescence (C). Statistical analysis was performed using an unpaired two-tailed Student’s t-test. Significance levels are indicated as follows: * p < 0.05, ***p < 0.001, and ****p < 0.0001. hDPSCs: human dental pulp stem cells; mimic NC: miRNA mimic negative control #1; miR-27a mimic: miRNA mimic for hsa-miRNA-27a-5p. Scale bars: 50 μm

Transplantation of miRNA-27a-transfected hDPSCs promoted hard tissue formation in vivo

The transplantation of miRNA-27a-transfected hDPSCs seeded in the 3D Honeycomb Boosted collagen scaffold (Figs. 6A, 7A) promoted collagen matrix deposition in the back subcutaneous tissue and hard tissue formation in cranial defects markedly superior to the transplantation of NC-transfected hDPSCs (Figs. 6B, C). Bone-like tissue was newly formed along the edge of artificial bone cavities in the miRNA-27a transfection group (Fig. 7B, C, D). However, the volume of such newly formed bone-like tissue was limited in the NC-transfected group (Fig. 7B, C, D).

Fig. 6

miRNA-27a promoted osteoblastic differentiation while downregulated DKK and SOSTDC1 in vivo. A The 3D honeycomb boosted scaffolds, in which hDPSCs transfected with or without miRNA-27a were seeded, were transplanted into the subcutaneous region of the dorsal aspect of mice. Sections of transplanted tissues were stained with hematoxylin and eosin stain B, Elastica van Gieson stain C and immunofluorescence stain of RUNX2, osteocalcin (D), SOSTDC1 and DKK3 (E). DKK3: dickkopf 3; SOSTDC1: sclerostin domain containing 1; hDPSCs: human dental pulp stem cells; NC: miRNA mimic negative control #1; miR-27a: miRNA mimic for hsa-miRNA-27a-5p. White scale bars: 10 μm. Black scale bars: 200 μm

Fig. 7

miRNA-27a promoted craniofacial bone regeneration and the expression of Wnt/BMP signaling activators in vivo. A The same 3D honeycomb boosted scaffolds were transplanted into the bone defects of mice calvaria. B, C Micro-computed tomography images showing enhanced mineralized tissue formation in miRNA-27a mimic-transplanted samples, miRNA-27a group significantly upregulated the bone formation compared to NC group on 14 days and 28 days (mean ± SEM, n = 6). Statistical analysis was performed using a Mann–Whitney U test. Significance levels are indicated as follows: *p < 0.05. D The bone mineral density was observed by micro-computed tomography images. E Immunohistochemical localization of signaling related molecules, BCL9, β-catenin and pSMAD1/5. F The formation of bone-like tissues was evaluated using alkaline phosphatase (ALP) and hematoxylin and eosin staining. The margins of the bone cavities are marked by red arrowheads. NB: Newly formed bone; Black arrowheads: Newly formed bone-like tissue, ALP positive area, OC positive cell; NC: miRNA mimic negative control #1; miR-27a: miRNA mimic for hsa-miRNA-27a-5p. Scale bars: 100 µm

Consistent with the in vitro experiments, in vivo findings also revealed increased expression of key odonto/osteoblastic factors RUNX2 and OC (Figs. 6D, 7F), and significant downregulation of the target genes DKK3 and SOSTDC1 in the miRNA-27a group (Fig. 6E). Additionally, Wnt pathway activation-related factors B-Cell Lymphoma 9 (BCL9) and β-catenin, as well as BMP pathway activation-related factor pSMAD1/5, were markedly upregulated in the newly formed bone-like tissue (Fig. 7E). Hematoxylin and eosin (HE) staining and alkaline phosphatase (ALP) staining further confirmed that the miRNA-27a group had a stronger positive effect on bone defect regeneration (Fig. 7F).

Discussion

Pulp infection-induced inflammation plays a key role in the development of apical periodontitis. LPS from infected root canals stimulates macrophages to secrete IL-1α and TNFα, which in turn regulate MMP-1 production, leading to inflammatory bone resorption in apical periodontitis [35]. The synergistic action of these cytokines may also promote RANKL, an NF-κB receptor activator ligand essential for bone resorption, potentially accelerating lesion growth [36]. Our previous research indicated that NF-κB signaling boosts miRNA-27a expression in LPS-stimulated hDPCs, which in turn inhibits NF-κB signaling pathway, creating negative feedback mechanisms that downregulates follistatin-like protein-1 (FSTL1) [25]. In FSTL1-overexpressing esophageal squamous cell carcinoma cells, the BMP pathway is silenced while the NF-κB pathway remains active, emphasizing their crosstalk [37]. Additionally, NF-κB activated by TNFα stimulation or IKK transfection directly inhibits bone matrix proteins production promoted by Wnt and BMP signaling [32]. This study highlights the crucial role of miRNA-27a in hard tissue regeneration after inflammation-induced injury by controlling the intricate network between the NF-κB, Wnt, and BMP pathways.

This study demonstrated that miRNA-27a transfection facilitates the odonto/osteoblastic differentiation of hDPSCs by targeting DKK3 and SOSTDC1 and activating Wnt and BMP signaling pathways. We are the first to report the comprehensive regulation of the NF-κB/Wnt/BMP signaling network by miRNA-27a, emphasizing its potential use in addressing inflammation-induced hard tissue defects.

hDPSCs overexpressing miRNA-27a exhibited noticeably increased migration capacity, which is crucial for disease progression, wound healing, and biomaterial-mediated regeneration [38]. This indicates that miRNA-27a may promote tissue repair and regeneration as well as support the directed migration of cells within biomaterial scaffolds. Significantly increased expression of miRNA-27a was observed in hDPSCs after being cultured in an osteogenesis-inducing medium for 3 days, and the expression of odonto/osteoblastic markers, and mineralized nodule formation significantly increased after transfection with miRNA-27a mimic. The formation of a mineralized matrix is the final stage of osteoblast differentiation, and DMP1, which was upregulated by the overexpression of miRNA-27a, is a regulatory protein involved in mineralization [39]. These findings suggest that miRNA-27a expression, which is promoted during the odonto/osteoblastic differentiation of hDPSCs, is involved in this differentiation process. A previous study reported a marked upregulation of miRNA-27 during the odontoblastic differentiation with miRNA-27a of a clonal odontoblast-like cell line called MDPC-23, significantly enhancing mineralized nodule formation [40]. Similarly, miRNA-27a expression increased when the well-established osteoblastic cell line MC3T3-E1 was cultured in an osteogenic induction medium. Transfection with miRNA-27a in these cells enhances osteoblast marker expression and promotes mineralized nodule formation [41]. Transfection of bone marrow stem cells with miRNA-27a promotes RUNX2 expression and calcified nodule formation [42]. Collectively, these findings suggest that miRNA-27a is intimately involved in the differentiation of hard tissue-forming cells.

The odonto/osteoblastic markers that are upregulated by miRNA-27a include RUNX2, which is a master gene of osteoblast differentiation [43]. Given that Wnt is the major upstream signal that induces RUNX2 expression [44], we speculated that miRNA-27a is involved in the induction and activation of Wnt signaling. Our findings confirmed that Wnt signaling activity was enhanced by miRNA-27a, and the expression of TCF7 and LEF1—the target genes of Wnt signaling—increased. Additionally, nuclear translocation of β-catenin, a typical phenomenon of active Wnt signaling, was observed. Therefore, we hypothesized that miRNA-27a induces the activation of Wnt signaling by suppressing the expression of negative regulators of Wnt signaling. Our study demonstrated that the expression of APC and AXIN2 was downregulated by miRNA-27a. APC plays a central role in suppressing the canonical Wnt signaling pathway [45], and miRNA-27a has been reported to downregulate the expression of APC [40]. In this study, we showed that AXIN2, another negative regulator and a target of Wnt/β-catenin signaling [46], was negatively regulated by miRNA-27a for the first time. We then focused on the negative regulators of Wnt signaling, and DKK3 and SOSTDC1 were screened as miRNA-27a targets. The expressions of DKK3 and SOSTDC1 were down- and upregulated in the overexpression and inhibition, respectively, of miRNA-27a in hDPSCs. Furthermore, a 3’UTR luciferase assay revealed previously unreported miRNA-27a binding sites located in the 3’ UTRs of DKK3 and SOSTDC1.

The Dickkopf genes consist of a single DKK3-related gene, DKKL1 (soggy), and a small gene family of four members (DKK1–4) that have been preserved across evolution [47]. The exact mechanism of the regulation of DKK3 on the Wnt signaling remains unclear. However, DKK3 inhibits Wnt signaling through a complicated and context-dependent interaction with low-density lipoprotein receptor-related protein (LRP) 5/6 [48]. DKK3 also interacts with β-transducin repeat-containing protein to inhibit the translocation of β-catenin into the nucleus, thereby attenuating the transcriptional activity and production of the β-catenin protein [49]. Compared with the regulatory mechanism of DKK3, that of SOSTDC1 on Wnt signaling is more clearly defined. SOSTDC1 inhibits Wnt signaling by reducing cell surface LRP6 [50]. SOSTDC1 deletion increases Wnt signaling, which in turn increases cortical bone [51]. Furthermore, SOSTDC1 directly binds to selected BMP proteins such as BMP2, -4, and -7, thus inhibiting BMP signaling [52]. Our study revealed that miRNA-27a upregulated the phosphorylation of SMAD1/5—key signaling molecules of the BMP pathway—and SMAD4, a downstream signaling molecule of SMAD1/5 called co-SMAD. Their expression in the miRNA-27a of over-expressed hDPSCs was also upregulated. Moreover, expression of GREM1, a ligand-sequestering antagonist for BMP2, -4, and -7 [34], was also depressed by miRNA-27a. Furthermore, miRNA-27a attenuates adipogenesis and promotes osteogenesis by targeting peroxisome proliferator-activated receptor γ and GREM1 [53]. These data suggest that miRNA-27 induced odonto/osteoblastic differentiation of hDPSCs by promoting Wnt and BMP signaling, which are critical for the differentiation of hard tissue-forming cells (Fig. 8).

Fig. 8

A schematic diagram showing the function of miRNA-27a-5p in hDPSCs. DKK3 inhibited Wnt signaling via the interference of β-catenin translocation, which negatively regulates TCF/LEF expression. SOSTDC1 is an antagonist of Wnt and BMPs and blocks both Wnt and BMP signaling. miRNA-27a-5p targeted DKK3 and SOSTDC1 and activated both Wnt and BMP signaling in our study. Furthermore, miRNA-27a-5p downregulated GREM1—one of the typical BMP antagonists—and led to the activation of BMP signaling

To investigate further, we examined whether miRNA-27a has the potential to induce new bone tissue formation in vivo. When miRNA-27a-overexpressing hDPSCs were seeded in collagen scaffolds and transplanted into cranial bone cavities, more bone-like hard tissue was formed compared to hDPSCs without forced miRNA-27a expression. These results indicated that miRNA-27a promotes bone-like hard tissue formation in hDPSCs. MiRNA-27a plays a crucial role in orchestrating osteoclast-mediated skeletal remodeling; depletion of miRNA-27a in mice resulted in heightened osteoclast proliferation and activity, which caused an increase in bone resorption [54]. It also said that miRNA-27a stimulates both osteogenesis and angiogenesis through its ability to reverse the inhibitory effects of TNFα on bone formation [55]. Our findings also confirm the involvement of miRNA-27a in enhancing odonto/osteoblastic differentiation, corroborating its importance in hard tissue formation.

MiRNA therapy has emerged as a prominent area of RNA-based therapeutics, and is regarded as a part of biopharmaceutical advancements, typically viewed as the “next generation” of medical interventions. The discovery of therapeutic miRNAs is considered one of the most significant and exciting breakthroughs in contemporary medicine. With their promising results, miRNA-based treatments are poised to emerge as the next wave of pharmaceuticals to treat a wide range of diseases [56].

MiRNA therapy can be combined with guided tissue regeneration (GTR) to regulate cellular behavior and signaling pathways, thereby enhancing tissue regeneration. Using miRNAs to stimulate osteogenic differentiation in cells for the regeneration of oral and craniofacial hard tissues has also emerged as a highly promising therapeutic approach, requiring a degree of enhanced biosafety, biocompatibility, and stability. In the future, regenerative engineering may address the lack of organs [57]. Although the role of scaffolds for regenerative engineering was not specifically examined in this research, the stiffness and porosity of scaffolds are critical elements influencing osteogenic differentiation in practical applications [58]. Recent studies have also demonstrated that three-dimensional cultured cells more closely resemble the natural tissue state than two-dimensional cultures, and that promoting interactions between cells and the surrounding matrix can enhance cell physiology [59]. Functional modifications, such as incorporating triphenylphosphine cations for mitochondrial targeting, hold potential for enhancing the accuracy and efficiency of delivery systems, and techniques for delivering particular targets are just as relevant to the advancement of miRNA supply [60]. In brief, identifying ways to induce the migration of stem cells to specific tissues, guide the differentiation of stem cells to participate in odonto/osteoblastic differentiation, and effectively use miRNAs in the clinic will be of great interest in future research.

As we have previously confirmed that miRNA-27a inhibits inflammatory mediator expression [25], our findings suggest that miRNA-27a controls inflammation and promotes hard tissue formation even in the presence of mild tissue inflammation. These findings guide the way for potential therapeutic applications of miRNA-27a in regenerative medicine, in which managing inflammation while enhancing tissue regeneration is crucial. This study utilized the transplantation of miRNA-27a overexpressing hDPSCs to achieve new bone formation; however, further studies are necessary to explore the devices and methods for the application of miRNA-27a.

Conclusion

Transfection of miRNA-27a in hDPSCs promoted their odonto/osteoblastic differentiation by targeting DKK3 and SOSTDC1 and activating Wnt and BMP signaling. Transplantation of miRNA-27a-overexpressing hDPSCs into calvarial bone cavities resulted in increased new bone formation in vivo. These findings emphasize the potential role of miRNA-27a in facilitating bone tissue regeneration, providing fundamental insights for a novel bone regeneration strategy.