EZH2, via an association with KDM2B, modulates osteogenic differentiation of root apical papillary stem cells

Abstract

BACKGROUND

Stem cells from apical papilla (SCAPs) represent promising candidates for bone regenerative therapies due to their osteogenic potential. However, enhancing their differentiation capacity remains a critical challenge. Enhancer of zeste homolog 2 (EZH2), a histone H3 lysine 27 methyltransferase, regulates osteogenesis through epigenetic mechanisms, but its role in SCAPs remains unclear. We hypothesized that EZH2 modulates SCAP osteogenic differentiation via interaction with lysine demethylase 2B (KDM2B), offering a target for therapeutic intervention.

AIM

To investigate the functional role and molecular mechanism of EZH2 in SCAP osteogenic differentiation.

METHODS

SCAPs were isolated from healthy human third molars (n = 6 donors). Osteogenic differentiation was assessed via Alizarin red staining and alkaline phosphatase assays. EZH2 overexpression/knockdown models were established using lentiviral vectors. Protein interactions were analyzed by co-immunoprecipitation, transcriptomic changes via microarray (Affymetrix platform), and chromatin binding by chromatin immunoprecipitation-quantitative polymerase chain reaction. In vivo bone formation was evaluated in immunodeficient mice (n = 8/group) transplanted with SCAPs-hydroxyapatite scaffolds. Data were analyzed using Student’s t-test and ANOVA.

RESULTS

EZH2 overexpression increased osteogenic markers and mineralized nodule formation. In vivo, EZH2-overexpressing SCAPs generated 10% more bone/dentin-like tissue. Co-immunoprecipitation confirmed EZH2-KDM2B interaction, and peptide-mediated disruption of this binding enhanced osteogenesis. Transcriptome analysis identified 1648 differentially expressed genes (971 upregulated; 677 downregulated), with pathway enrichment in Wnt/β-catenin signaling.

CONCLUSION

EZH2 promotes SCAP osteogenesis via antagonistic interaction with KDM2B, and targeted disruption of this axis offers a translatable strategy for bone regeneration.

Key Words: Bioactive peptides; Bone tissue engineering; Enhancer of zeste homolog 2; Osteogenic; Apical papillary stem cell

Core Tip: Enhancing the osteogenic potential of stem cells is crucial for bone regeneration. This study reveals that overexpressing enhancer of zeste homolog 2 (EZH2), a key methyltransferase, promotes osteogenic differentiation in root apical papillary stem cells. EZH2 enhances osteogenesis by regulating related protein and gene expression, both in vitro and in vivo. Furthermore, the interaction between EZH2 and lysine demethylase 2B inhibits this process, suggesting that disrupting this interaction with small-molecule peptides could improve apical papillary stem cell-based therapies for bone defects. This finding opens new avenues for optimizing stem cell-based treatments through targeted molecular modulation.

INTRODUCTION

Maxillofacial bone defects are a common clinical challenge resulting from various causes, including congenital malformations, high-energy trauma, open fractures, resection of bone tumors, and infections that require debridement. Although bone tissue possesses a remarkable ability to heal and repair itself, when the damage exceeds the regenerative capacity, bone grafting becomes necessary to support repair and reconstruction[1]. The main types of grafting materials include autografts, allografts, xenografts, synthetic grafts, and combinations of these. Autologous bone grafts are considered the gold standard due to their availability of osteoblasts, growth factors, and strong osteogenic potential[2]. However, their use is limited by the amount available from the donor site, and harvesting can cause pain and increase the risk of infection[3,4]. Allogeneic and xenogeneic grafts overcome these limitations but carry risks of immunogenicity and disease transmission. Furthermore, the loss of mineral content during processing weakens their mechanical properties and osteogenic capacity, making them less suitable for repairing large, complex bone defects. As a result, these grafts are primarily used as filler materials and are less effective for structural bone reconstruction[5].

Bone tissue engineering has evolved as a promising strategy for bone defect repair thanks to material science advancements, aiming to achieve functional bone regeneration with minimal complications. Osteoblasts that generate bone tissue matrix are a key component of bone tissue engineering. Mesenchymal stem cells (MSCs), possessing self-renewal and multilineage differentiation capacities[6], are particularly valuable due to their low immunogenicity and adaptability for autologous applications[7]. Among MSC sources, dental-derived populations - including from apical papilla (SCAPs), periodontal ligament stem cells, and dental pulp stem cells - demonstrate exceptional dentinogenic potential[8,9]. Compared to non-dental MSCs, dental-derived stem cells exhibit superior performance in oro-facial tissue regeneration[10] with SCAPs showing enhanced proliferative activity and osteo/dentinogenic capacity relative to periodontal ligament stem cells and dental pulp stem cells[11,12]. These characteristics position SCAPs as ideal candidates for regenerative therapies. To fully harness MSC potential, precise control over differentiation pathways is critical[13]. Preclinical studies demonstrate SCAP-mediated bone repair in periodontal disease models[14] and mineralized tissue formation in subcutaneous transplantation assays[15], though optimizing lineage-specific differentiation remains a key challenge.

Epigenetic regulation profoundly influences skeletal development and MSC osteogenic commitment, primarily through histone modifications, DNA methylation, and non-coding RNAs. Histone-modifying enzymes, including acetylases/deacetylases and methyltransferases/demethylases, are increasingly recognized as bone remodeling regulators[16-18]. The H3K36 trimethyltransferase WHSC1 (NSD2) functions through cooperative interactions with runt-related protein 2 and p300 to activate osteogenic gene transcription[19]. Conversely, lysine-specific demethylase 1 suppresses MSC osteoblastic differentiation by erasing H3K4/K9 mono-/bi-methylation modifications[20]. Enhancer of zeste homolog 2 (EZH2), a polycomb repressive complex 2 component mediating histone H3 lysine 27 (H3K27) trimethylation[21,22], exhibits context-dependent duality in skeletal biology: While pharmacological EZH2 inhibition enhances bone formation, its genetic ablation disrupts osteoblast maturation by altering cell cycle dynamics[21]. Previous studies have reported that pharmacological inhibition of EZH2 in wild-type mice was used to enhance osteogenesis and stimulate bone formation[23]. However, conditional knockout of EZH2 early in the mesenchymal lineage (i.e., through Prrx1 promoter-driven Cre expression) leads to skeletal abnormalities due to pattern defects. EZH2 defect in bone marrow-derived mesenchymal cells inhibits osteogenic differentiation and impedes cell cycle progression. Thus, knockdown of EZH2 in mouse preosteoblasts inhibits osteogenesis in part by inducing cell cycle changes[21]. This functional duality - promoting progenitor proliferation while restraining lineage commitment - underscores its complex regulatory nature. Emerging studies reveal non-canonical EZH2 mechanisms involving transcriptional complex formation and non-histone protein methylation[24-26], yet its role in human SCAP (hSCAP) osteogenesis remains incompletely characterized.

Studies on epigenetic mechanisms have suggested a functional synergy or antagonism between epigenetic factors[27]. By site prediction, we identified histone lysine demethylase 2B (KDM2B) as a possible interacting molecule of EZH2[28]. KDM2B mainly plays the role of removing methylation modifications at histone H3K4 and H3K36 sites, and plays an important role in physiological processes such as cell proliferation and senescence[29]. In addition, bioinformatics analysis revealed that KDM2B is closely related to the pathogenesis of osteoarthritis[30]. In addition, KDM2B^WT/ΔCxxC mice have a phenotype of abnormal skeletal development of the spinal bones[31]. These findings suggest that KDM2B may play an important function in the skeletal system. Therefore, it is necessary to further investigate the role of EZH2 interacting with KDM2B in regulating the osteogenic differentiation potential of MSCs.

In this study, hSCAPs were used to explore the roles of EZH2 and KDM2B in osteogenic differentiation. Our results suggest that overexpression of EZH2 in SCAPs has a promotive effect on their osteogenic differentiation capacity. Specifically, our results suggest that EZH2 binds to KDM2B and blocks the osteogenic differentiation potential of SCAPs. We found for the first time that there is a direct binding interaction between EZH2 and KDM2B, revealed that the key binding sites of EZH2 and KDM2B are located in the JmjC, CxxC, and PHD functional domains of KDM2B, and presented the detailed sequence information of the action sites for the first time. Based on these sequences, we obtained the functional biopeptides that can effectively intervene in the binding of EZH2 to KDM2B by our innovative synthesis.

MATERIALS AND METHODS

Cell culture

The hSCAP experiments involved in this study was approved by the Ethical Committee of Beijing Stomatological Hospital, Capital Medical University (Ethical Review No. CMUSH-IRB-KJ-PJ-2022-24). hSCAPs were extracted from the immature root tip tissue of wisdom teeth with appropriate patient consent. SCAP cultures at the third to fifth passages were utilized for conducting the subsequent experiments. Further details regarding the culture of SCAPs have been elucidated in our previous publication[32].

Plasmid construction and viral infection

The plasmids were prepared using established protocols. EZH2-specific short hairpin RNA (shRNA), control shRNA (LV3 shRNA, Consh), and EZH2 overexpression constructs were commercially sourced (Taihegene, Beijing). For ectopic EZH2 expression, the full-length human EZH2 sequence was cloned into the pQCXIN retroviral vector via BamH1/PacI restriction sites. Empty vector served as the negative control. Viral transduction followed our prior methodology[33], with puromycin selection (2 μg/mL) initiated 72 hours post-transfection to establish stable cell lines.

Alkaline phosphatase assay and Alizarin red staining

Following induction with osteogenic medium (100 μM ascorbic acid, 2 mmol/L β-glycerophosphate, 1.8 mmol/L KH₂PO₄, 10 nM dexamethasone), alkaline phosphatase (ALP) activity was quantified using a commercial kit (Sigma-Aldrich, MA, United States) at days 3 and 5. Mineralized matrix deposition was evaluated by Alizarin red staining (Sigma-Aldrich) after 14 day[32]. Triplicate wells were analyzed per experimental condition.

RNA isolation and real-time quantitative polymerase chain reaction

Total RNA was isolated from SCAPs using a commercial kit (Vazyme RC112-01). Reverse transcription of 1 μg RNA into cDNA was performed (Invitrogen kit, MA, United States), followed by quantitative polymerase chain reaction (qPCR) amplification as previously detailed[34]. GAPDH served as the endogenous control for mRNA quantification. All reactions were performed in triplicate. Primer sequences are provided in Table 1.

Table 1 Primer sequence.

Gene	Forward	Reverse
FGF2	CATCAAGCTACAACTTCAAGCA	CCGTAACACATTTAGAAGCCAG
EZH2	AAATCAGAGTACATGCGACTGA	GTATCCTTCGCTGTTTCCATTC
STC2	TCATCAAAGACGCCTTGAAATG	CAGCAAGTCCTTGAAATGGATC
TRIB3	CTACGTGGGACCTGAGATACTC	GAGTCCTGGAAGGGGTAGT
PCK2	CATCCCAACTCTCGATTTTGTG	TTCCCAGAAGTCCTTTGTGTTC
PSAT1	CCAAGTTTGGTGTGATTTTTGC	TGAGATTTGATGGAGCTAAGCT
GAPDH	GTATCGTGGAAGGACTCATGAC	ACCACCTTCTTGATGTCATCAT

FGF2: Fibroblast growth factor 2; EZH2: Enhancer of zeste homolog 2; STC2: Stanniocalcin 2; TRIB3: Tribbles pseudokinase 3; PCK2: Phosphoenolpyruvate carboxykinase 2; PSAT1: Phosphoserine aminotransferase 1.

Western blot analysis

Total proteins were resolved from SCAP, and SDS-polyacrylamide gel electrophoresis was performed as described previously[35]. The primary antibodies used in this study were those against dentin sialophosphoprotein (DSPP) (Cat. No. bs10316R, Bioss, China), bone sialoprotein (BSP) (Cat. No. bs-0026R, Bioss, China), and GAPDH (Cat. No. G8795, Sigma-Aldrich, MA, United States).

Co-immunoprecipitation assay

Cell lysates prepared with immunoprecipitation buffer containing protease inhibitors were incubated overnight at 4 °C with anti-EZH2 (CST #5246) or anti-KDM2B (CST #44570) antibodies. Protein A/G agarose beads (Sigma, MA, United States) were added for 2 hours, followed by three washes with chilled immunoprecipitation buffer. Immunoprecipitated complexes were subjected to immunoblotting.

Genome-wide H3K4me3 and H3K27me3 chromatin immunoprecipitation sequencing

After obtaining the genomic DNA fragments bound to the target proteins using the ChIP extraction kit, sequencing was performed to analyze the DNA binding sites of the target proteins in the whole genome. Following the process of ChIP library construction, data filtering, sequence comparison, and peak scanning, the data generated by Illumina sequencing were quality controlled and filtered, and compared with the reference genome using the comparison tool. Unique sequences were extracted and the results were stored in BED files, which were used for subsequent analysis, including reads analysis and peak scanning. Peak scanning was performed in the whole genome, and the associated genes of the scanned peaks were analyzed, including Gene Ontology and pathway enrichment analysis.

Microarray-based transcriptome profiling

Gene expression analysis was performed using the PrimeView^™ Human Gene Expression Array. Differentially expressed mRNAs were identified through Affymetrix GeneChip Operating Software, with statistical thresholds set at |fold change| ≥ 2 and P value ≤ 0.05.

Immunohybridization reaction of protein and peptide microarrays

The full-length protein sequence of human EZH2 was obtained by searching the Uniprot Protein Information website (https://www.uniprot.org), and based on the full-length sequence of 736 amino acids, we designed the second proposed synthetic peptide according to the principle of designing and synthesizing small molecule biologically active peptides by taking 15 amino acids as the observation window, starting from the first amino acid, moving back 10 amino acids, and then designing the second proposed synthetic peptide. Then, the immunohybridization reaction between KDM2B protein and EZH2 peptide microarray chip was carried out,the images were analyzed with TotalLab image analysis software for the optical density value of the chromogenic spots, and the “Spot Edge Average” algorithm was used to read out all the peptide sites. The highest optical density value of the chromogenic spots on the membrane was set as 100%, and using the background value around each spot as a reference, the optical density value of the remaining spots was calculated as a percentage of the optical density value of the spot. Bioinformatics comparison analysis was performed to obtain the possible direct binding sites and fragment sequences of EZH2 and KDM2B.

Design and synthesis of small molecule bioactive peptides

For the obtained positive binding site peptide sequence, according to the order from the carboxyl end to the amino end of the synthesis, the peptide crude product was purified using high-performance liquid chromatography to a purity of not less than 95%; the purified liquid was lyophilized to obtain high-purity peptide lyophilized powder.

Animal transplantation and histological analysis

An in vivo transplantation assay was performed using 16 immunodeficient mice (6 weeks old, male, n = 8/group: Vector-SCAP group and HA-EZH2-SCAP group). SCAPs (2 × 10⁶ cells) were combined with 20 mg hydroxyapatite/tricalcium phosphate (Engineering Research Center for Biomaterials, Sichuan University) and pre-incubated at 37 °C for 2 hours. The cell-scaffold complexes were subcutaneously implanted into 10-week-old female nude mice. After an 8-week incubation period, transplanted tissues were harvested, fixed in 10% neutral buffered formalin, and decalcified using 10% EDTA. Specimens were processed for hematoxylin-eosin (H&E) histology. Immunohistochemical analysis followed established protocols[33] using primary antibodies against osteocalcin (OCN) (Bioss bs-4917R) and DSPP (Bioss bs10316R). All procedures complied with institutional animal care guidelines (Ethics Committee of Beijing Stomatological Hospital, Approval No. KQYY-202110-003).

H&E staining

Harvested grafts underwent H&E staining using established protocols[36]. Quantification of newly formed mineralized tissues was performed. Tissue sections were randomly selected through a simple random sampling protocol to ensure unbiased analysis.

Statistical analysis

All experiments were independently replicated three times or more. Data analyses were conducted using GraphPad 7 (GraphPad Software, La Jolla, CA, United States) or SPSS 22.2 (IBM Corporation, New York, NY, United States). Appropriate statistical methods - Student’s t-test for pairwise comparisons, one-way ANOVA for multi-group analyses, or Kruskal-Wallis test for non-normally distributed data - were selected based on data characteristics. Statistical significance was defined as P ≤ 0.05.

RESULTS

EZH2 knockdown impairs SCAP osteogenic differentiation

To assess EZH2’s role in hSCAP osteogenesis, lentiviral vectors encoding EZH2-specific shRNA were transduced into hSCAPs. Puromycin selection (2 μg/mL, 3 days) ensured stable infection. EZH2 knockdown efficiency was confirmed by real-time qPCR (RT-qPCR) and Western blot (Figure 1A and B). ALP activity decreased in EZH2-suppressed hSCAPs after 7 days of mineralization induction (Figure 1C). Alizarin red staining and calcium quantification at 3 weeks revealed reduced mineral deposition compared to controls (Figure 1D and E). RT-qPCR demonstrated downregulated expression of osteogenic markers (BSP, DSPP, and OCN) at days 7 and 14 (Figure 1F-H).

Figure 1 Knockdown of enhancer of zeste homolog 2 inhibits steo/dentinogenic differentiation potential of human apical papillary stem cells. A: Quantitative polymerase chain reaction showed that the expression of enhancer of zeste homolog 2 (EZH2) was inhibited in human apical papillary stem cells (hSCAPs); B: Western blot analysis confirmed the knockdown of EZH2 in hSCAPs; C: Knockdown of EZH2 decreased alkaline phosphatase activity in hSCAPs; D and E: Alizarin red staining and quantitative calcium analysis demonstrated that knockdown of EZH2 inhibited mineralization in hSCAPs; F-H: Quantitative polymerase chain reaction showed that knockdown of EZH2 downregulated mRNA expression levels of bone sialoprotein (F), dentin sialophosphoprotein (G), and osteocalcin (H) in hSCAPs. GAPDH and ACTB was used as the internal controls. Data are presented as the mean ± SD (n = 3). Statistical analysis was performed using Student’s t-test. ^aP ≤ 0.05, ^bP ≤ 0.01, ^cP ≤ 0.001. EZH2: Enhancer of zeste homolog 2; BSP: Bone sialoprotein; DSPP: Dentin sialophosphoprotein; OCN: Osteocalcin.

EZH2 overexpression enhances SCAP osteogenic capacity

Lentiviral-mediated EZH2 overexpression in hSCAPs (puromycin-selected, 2 μg/mL, 3 days) was validated by RT-qPCR and Western blot (Figure 2A and B). Enhanced ALP activity was observed at 7 days post-induction (Figure 2C). After 3 weeks, Alizarin red staining and calcium quantification showed increased mineralization compared to controls (Figure 2D and E). RT-qPCR confirmed upregulated BSP, DSPP, and OCN expression at days 7 and 14 (Figure 2F-H). In vivo transplantation assays (8 weeks) revealed greater bone/dentin-like tissue formation in SCAP-EZH2 grafts vs controls (Figure 2I). Quantitative histomorphometry and immunohistochemistry confirmed elevated DSPP/BSP levels in EZH2-overexpressing transplants (Figure 2J).

Figure 2 Overexpression of enhancer of zeste homolog 2 enhances osteo/dentinogenic differentiation potential of human apical papillary stem cells. A: Quantitative polymerase chain reaction showed that enhancer of zeste homolog 2 (EZH2) was overexpressed in human apical papillary stem cells (hSCAPs); B: Western blot analysis confirmed overexpression of EZH2 in hSCAPs; C: Overexpression of EZH2 increased alkaline phosphatase activity in hSCAPs; D and E: Alizarin red staining and quantitative calcium analysis results demonstrated that overexpression of EZH2 enhanced mineralization in hSCAPs; F-H: Quantitative polymerase chain reaction showed that overexpression of EZH2 upregulated mRNA expression levels of bone sialoprotein (F), dentin sialophosphoprotein (G), and osteocalcin (H) in hSCAPs; I: Hematoxylin-eosin staining and quantitative measurement showed that overexpression of EZH2 promoted bone/dentin-like tissue formation. Scale bar = 100 μm (B: Bone/dentin-like tissues; HA: Hydroxyapatite tricalcium carrier; CT: Connective tissue); J: Immunohistochemical staining and quantitative analysis of dentin sialophosphoprotein and bone sialoprotein. GAPDH and ACTB were used as the internal controls. Data are presented as the mean ± SD (n = 3). Statistical analysis was performed using Student’s t-test. ^aP ≤ 0.05, ^bP ≤ 0.01, ^cP ≤ 0.001. EZH2: Enhancer of zeste homolog 2; BSP: Bone sialoprotein; DSPP: Dentin sialophosphoprotein; OCN: Osteocalcin.

EZH2 regulates expression of osteogenic differentiation-related genes independently of histone methyltransferase activity

After initially clarifying the role of EZH2 in the osteogenic differentiation of SCAP, in order to explore the potential regulatory mechanisms of EZH2 in the osteogenic differentiation of SCAPs, we performed mRNA transcriptome microarray analysis of SCAPs with EZH2 knockdown. Through in-depth bioinformatics comparative analysis, a total of 1648 significantly differentially expressed genes (DEGs) were identified, including 971 up-regulated and 677 down-regulated DEGs (Figure 3A). To validate the accuracy of the results of mRNA transcriptome microarray analyses, we randomly chose the mRNAs such as phosphoserine aminotransferase 1, stanniocalcin 2, tribbles pseudokinase 3, phosphoenolpyruvate carboxykinase 2, and fibroblast growth factor 2 genes, and detected the expression alterations of these five genes in SCAPs with EZH2 knockdown by RT-qPCR, and the RT-qPCR results were consistent with the alteration trend of mRNA transcriptome microarray analysis (Figure 3B). Next, we performed enrichment analysis of significantly different relevant signaling pathways and Gene Ontology entries in DEGs (Figure 3C and D), and the results indicated that knockdown of EZH2 mainly affected the ossification function of SCAPs. Then, we detected genome-wide H3K27me3 and H3K4me3 modification level changes in SCAPs with EZH2 knockdown with the help of genome-wide ChIP-seq technology (Supplementary Figures 1 and 2). It is well known that EZH2 is a methyltransferase that plays a role in silencing related genes by increasing H3K27me3 levels, so we performed an overlay analysis of genes enriched to the H3K27me3 peak and genes up-regulated by EZH2 silencing in SCAPs (Figure 3E), in which 73 genes with an intersection were mainly associated with rheumatoid arthritis, axonal guidance, thyroid hormone secretion, and other signaling pathways (Figure 3F). According to the finding of Kim et al[24], EZH2 can also play a role similar to that of a transcription factor independently of histone methyltransferase activity, directly regulating the transcription level of downstream genes. Therefore, we focused our study on how to enhance the promotional effect of EZH2 on the osteogenic differentiation potential of SCAPs.

Figure 3 Enhancer of zeste homolog 2 inhibits expression of osteogenic differentiation-related genes independently of histone methyltransferase effect. A: Heatmap illustrating differential gene expression in human apical papillary stem cells (SCAPs) following enhancer of zeste homolog 2 (EZH2) knockdown; B: Quantitative polymerase chain reaction analysis validated the expression changes of five randomly selected genes, consistent with the findings of the mRNA transcriptome microarray analysis; C: Histogram presenting significant results of Gene Ontology analysis for differentially expressed genes in SCAPs with EZH2 knockdown compared to the control group; D: Scatter plot depicting Kyoto Encyclopedia of Genes and Genomes pathway enrichment statistics for differentially expressed genes in SCAPs with EZH2 knockdown compared to the control group; E: Venn diagram showing the intersection between genes upregulated in human apical papillary stem cells following EZH2 knockdown and those associated with reduced trimethylation levels at the H3K27 locus; F: Histogram displaying significant Gene Ontology analysis results for differentially expressed genes in SCAPs with EZH2 knockdown relative to the control group. Data are presented as the mean ± SD (n = 3). Statistical analysis was performed using Student’s t-test. ^bP ≤ 0.01, ^cP ≤ 0.001. PSAT1: Phosphoserine aminotransferase 1; STC2: Stanniocalcin 2; PCK2: Phosphoenolpyruvate carboxykinase 2; TRIB3: Tribbles pseudokinase 3; FGF2: Fibroblast growth factor 2; EZH2: Enhancer of zeste homolog 2.

Reducing EZH2 binding to KDM2B promotes osteogenic differentiation of SCAPs

After initially identifying the role of EZH2 on the osteogenic differentiation of SCAPs in vitro and the ability of bone-like deposition in vivo, we further examined possible interactions between EZH2 and KDM2B during the osteogenic differentiation of SCAPs. By site prediction, we found that KDM2B is a possible interacting molecule of EZH2 (Figure 4A). qPCR results showed that the protein expression level of EZH2 gradually increased on days 3 and 7 of osteogenic differentiation induction, while the protein expression level of KDM2B gradually decreased on days 3 and 7 of osteogenic differentiation induction (Figure 4B). This suggests that a portion of EZH2 with increased expression may be involved in the physiological process of binding to KDM2B during osteogenic differentiation of SCAPs, depleting the total amount of EZH2, which positively contributes to osteogenic differentiation. So, we next tried to find a way to competitively substitute EZH2 for binding to KDM2B and inhibit the inhibitory effect of KDM2B on osteogenic differentiation, thus also allowing more EZH2 to better fulfill its facilitating role. Therefore, we next simulated the binding mode and affinity of EZH2 protein and KDM2B protein with the help of molecular docking technique (Figure 4C and D), and obtained the most likely and stable pose for the binding of the two.

Figure 4 Reduction of enhancer of zeste homolog 2 binding to lysine demethylase 2B promotes osteogenic differentiation of human apical papillary stem cells. A: The protein-protein interaction network analysis revealed potential direct or indirect interactions between enhancer of zeste homolog 2 (EZH2) and lysine demethylase 2B (KDM2B); B: Quantitative polymerase chain reaction confirmed that EZH2 expression progressively increased, while KDM2B expression gradually decreased during the induction of osteogenic differentiation in human apical papillary stem cells; C and D: Molecular docking simulations predicted the binding modes and affinities between EZH2 and KDM2B proteins. C illustrates the two-dimensional interactions, while D presents the three-dimensional binding conformations. Data are presented as the mean ± SD (n = 3). Statistical analysis was performed using Student’s t-test. ^aP ≤ 0.05, ^bP ≤ 0.01, ^cP ≤ 0.001. EZH2: Enhancer of zeste homolog 2; KDM2B: Lysine demethylase 2B.

EZH2 bioactive peptides can effectively block binding of KDM2B to EZH2 and promote osteogenic differentiation of SCAPs

Based on the above discovery of the interaction mechanism between EZH2 and KDM2B, and according to the amino acid sequence of the EZH2 protein, we designed and composed a peptide microarray based on the full-length protein sequence of EZH2. The KDM2B full-length recombinant epitope protein solution was subjected to immunohybridization reaction with the EZH2 peptide microarray chip, and then we collected and read the detailed information of the hybridization reaction for each peptide site (Figure 5A and B), and obtained the percentage of gray value of the colorimetric readout of each peptide site. The results showed that there was a total of 16 positive peptide sites whose percentage of gray value fell within the 80%-100% interval (Figure 5C). Comparison of these positive peptide sequences with the full-length amino acid sequence of EZH2 initially labeled the positional information of these positively reported binding sites with the functional structural domain of the EZH2 protein (Figure 5D). These results suggest that the KDM2B protein has significant, direct binding interactions with certain peptide segments on the EZH2 peptide microarray. Next, we selected four positive binding fragments with clear overlap with the functional structural domains of EZH2 proteins whose studies have been reported, and synthesized the corresponding bioactive peptides against the obtained peptide sequences. One negative binding site was also selected as a negative control and the peptides were similarly synthesized. To assess the function of these synthesized peptides, we pretreated SCAPs with these peptides for 24 hours. Co-immunoprecipitation assays showed that peptides PP7 and PP8 were particularly effective in inhibiting the interaction between EZH2 and KDM2B (Figure 5E). In addition, to explore the effects of EZH2 peptides on osteogenic differentiation, we added these peptides to the osteogenic differentiation induction medium of SCAPs and continued the induction of differentiation. The results showed that EZH2 peptides significantly enhanced the osteogenic differentiation of SCAPs (Figure 5F).

Figure 5 Enhancer of zeste homolog 2 bioactive peptides effectively block lysine demethylase 2B binding to enhancer of zeste homolog 2 and promote human apical papillary stem cell osteogenic differentiation. A and B: Peptide site immunohybridization results confirmed the presence of a direct binding site between enhancer of zeste homolog 2 (EZH2) and lysine demethylase 2B; C: Gray value analysis identified 16 positive loci, with gray values ranging between 80%-100%; D: Schematic illustration of different functional domains of EZH2 protein; E: Co-immunoprecipitation assays indicated that peptides PP7 and PP8 effectively inhibited interaction between EZH2 and lysine demethylase 2B; F: Alkaline phosphatase staining demonstrated that EZH2 peptides significantly promoted osteogenic differentiation of apical papillary stem cells. Data are presented as the mean ± SD (n = 3). Statistical analysis was performed using Student’s t-test. ^aP ≤ 0.05. EZH2: Enhancer of zeste homolog 2; KDM2B: Lysine demethylase 2B; ALP: Alkaline phosphatase.

DISCUSSION

With the advancement of tissue engineering, the use of MSCs for tissue regeneration has garnered significant attention. A key challenge in this field is how to effectively regulate the homing and directed differentiation of MSCs to promote efficient tissue regeneration. In this study, we identified EZH2 as a crucial regulator of the osteogenic and dentogenic differentiation of root SCAPs, both in vitro and in vivo. Although previous studies have suggested that EZH2 may negatively regulate osteogenesis, our results suggest a more complex role. For instance, it has been reported that during osteogenic differentiation of human dental follicle stem cells, both EZH2 expression and H3K27me3 levels decrease, indicating a potential inhibitory effect on osteogenesis[37]. It has also been shown that by conditionally knocking out the EZH2 gene in mouse myeloid cells, a significant increase in mature osteoblasts and bone formation biomarkers (P1NP and osteocalcin) was found in the bone marrow of such mice. This contradicts our results[38]. And interestingly, in Dudakovic et al’s study, it was found by initiating Cre at different time points that deletion of EZH2 in the pre-osteoblastic phase produced phenotypically normal mice[21]. Juvenile EZH2 deletion, on the other hand, resulted in mice with a hypoplastic bone trabecular phenotype. Thus, the effect of EZH2 on bone formation is stage-specific[21]. Indeed, deletion of EZH2 in bone marrow MSCs inhibits osteogenic differentiation and impedes cell cycle progression, which is reflected in reduced metabolic activity, decreased cell number, cell cycle distribution, and changes in cell cycle marker expression[39]. In conclusion, this suggests that EZH2 has a dual function in bone formation, and that EZH2 may play very different roles in cells at different levels of differentiation, which also suggests that we should strictly grasp the appropriate time point when applying EZH2’s regulation of bone formation to help repair bone defects.

By jointly analyzing the sequencing results of transcriptome microarrays and ChIP microarrays, we found that among the genes enriched to the H3K27me3 peak and the genes up-regulated by EZH2 silencing, the 73 genes intersected were mainly related to the signaling pathways of rheumatoid arthritis, axon guidance, and thyroid hormone secretion, but not the signaling pathways related to ossification. Based on previous studies, there is increasing evidence that EZH2 has many “unconventional” functions beyond methylating the histone H3K27 site[40]. First, EZH2 can methylate many non-histone proteins and thus regulate cellular processes in an H3K27me3-independent manner[41-43]. In addition, EZH2 relies on both methyltransferase-dependent and methyltransferase-independent mechanisms to regulate cellular gene expression programs and/or epigenomic patterns. Importantly, independently of polycomb repressive complex 2, EZH2 also forms physical interactions with many DNA-binding factors and transcriptional co-activators that context-dependently affect gene expression[44,45]. For example, EZH2 can enhance or repress the gene regulatory activities of transcription factors through methylation of their respective genes: GATA binding protein 4 (GATA4) is a key dosage-sensitive regulator of cardiac development[26]. Part of the role of this cardiac transcription factor is to promote gene transcription by recruiting the histone acetyltransferase p300 to specific chromatin sites. EZH2 directly interacts with lysine 299 of GATA4 and methylates GATA4, an event that reduces the GATA4-P300 interaction and thus attenuates GATA4 function[26]. In addition, EZH2 contains a hidden, partially disordered trans-activating structural domain that directly interacts with coactivators and acetyltransferase p300 to activate gene expression in a p300-dependent manner[46,47]. Thus, we hypothesize that EZH2 may not exert its regulatory effects on osteogenic genes in SCAPs through the classical methyltransferase-dependent pathway. However, the specific regulatory mechanisms are still being explored in our laboratory, and this study is more concerned with how to more effectively promote osteogenic differentiation of SCAPs and thus aid in tissue regeneration, so we shifted our research focus to how to improve the potential of EZH2 to enhance osteogenic differentiation of SCAPs.

CONCLUSION

Mechanistically, our findings reveal that EZH2 positively regulates the osteogenic or dentogenic differentiation capacity of SCAPs and that EZH2 binds to KDM2B and exerts its regulatory effects through mutual antagonism. Our findings provide new insights into the mechanisms that promote the osteogenic differentiation of dental MSCs and offer potential therapeutic targets and promising small-molecule drugs for promoting maxillofacial bone defect repair. Nevertheless, the small molecule peptide drugs developed in this study still have some limitations. At present, the effect of these bioactive peptides in promoting the osteogenic differentiation ability of SCAPs cells are still relatively limited, so the next step is that we will try to develop a more effective small molecule active peptide by combining the functions of the various structural domains of EZH2 with the remaining 9 sites obtained from the immunohybridization analysis of the protein and peptide microarrays. In the future, we will also try to modify and optimize existing bioactive peptide drugs to enhance their activity.