circ_0002456/FUS interaction inhibits NF-κB signaling to attenuate DNA damage and inflammatory responses in hDPSCs

Abstract

Objectives

circular RNAs (circRNAs) are emerging regulators of inflammatory diseases, but their role in pulpitis remains unclear. This study aimed to investigate the role of circRNA in the occurrence and development of pulpitis.

Methods

qRT-PCR and Western Blot were used to detect γ-H2AX, a marker of DNA double-strand breaks (DSBs), and inflammatory factors in pulp tissues. Bioinformatics identified dysregulated circRNAs. Loss- and gain-of-function experiments were conducted to explore the function of circ_0002456 in lipopolysaccharide (LPS)-induced DNA damage and inflammation in human dental pulp stem cells (hDPSCs). The interaction between circ_0002456 and fused in sarcoma (FUS) protein was validated by RNA fluorescence in situ hybridization (FISH), RNA pull-down and nucleoplasmic separation experimen. NF-κB pathway activation was assessed after circ_0002456/FUS siRNA transfection in hDPSCs, and NF-κB inhibitors were used to confirm its regulatory role in DNA damage and inflammation.

Results

DNA damage was positively correlated with inflammation in pulpitis. In vitro, circ_0002456 downregulation exacerbated LPS-induced DNA damage and inflammation, while overexpression alleviated these effects. Mechanistically, circ_0002456 bound FUS, restricting its nuclear export and suppressing NF-κB activation, thereby mitigating DNA damage and inflammation.

Conclusion

circ_0002456 interacts with FUS to inhibit NF-κB signaling, attenuating DNA damage and inflammation in hDPSCs, revealing a novel regulatory axis in pulpitis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13287-025-04391-6.

Background

Pulpitis is a common oral disease, the main cause of which is bacterial infection after pulp exposure [1]. Recent research on pulpitis has focused on controlling pulpitis and pulp regeneration [2]. Dental pulp stem cells (DPSCs), a type of mesenchymal stem cell (MSC) derived from dental pulp tissue, exhibit multilineage differentiation potential and can be directionally differentiated into odontoblasts [3], osteoblasts [4], adipocytes [5], neural cells [6], making them a key focus in pulpitis-related research. Under inflammatory microenvironments, the odonto/osteogenic differentiation capacity of DPSCs is significantly suppressed, while resolving inflammation restores their reparative functions [7, 8]. However, there are relatively few studies on the regulation of pulpitis by circRNA. Inflammation, a fundamental defense mechanism against pathological stimuli, involves reactive oxygen species (ROS) production to eliminate pathogens. However, excessive ROS under pathological conditions induces macromolecular damage, including DNA lesions [9, 10]. After damage occurs, the body activates the DNA damage response (DDR) to recruit repair proteins and mitigate damage [11]. If the damage is not repaired in time, it may trigger cell senescence, apoptosis or cancer [12]. DSBs is an extremely serious form of DNA damage. Intriguingly, human pulpitis tissues exhibit elevated expression of γ-H2AX, a DSB marker, alongside inflammatory mediators [13]. Regulation of the repair protein Ku70 in the nonhomologous end junction (NHEJ) pathway can reduce DSBs and attenuate pulp inflammation [13], establishing a direct link between DNA damage and pulpitis progression.

Circular RNAs (circRNAs) are closed circular RNA formed by reverse splicing that regulate gene expression at the transcriptional and post-transcriptional levels. As a form of epigenetic modification, circRNA has been implicated in a variety of human inflammatory diseases, including periodontitis, atherosclerosis, osteoarthritis, and gastrointestinal diseases [14–16]. Beyond inflammation, circRNAs influence disease pathogenesis, including chronic kidney disease [17], neurodegeneration [18], and cancer [19–21], through DNA damage response (DDR) regulation. However, whether circRNA modulate pulpitis progression via DDR mechanisms remains unexplored, and this area needs more attention and in-depth exploration. Here, we found that circ_0002456 is down-regulated in pulpitis tissues and demonstrated that circ_0002456 binds to the FUS protein, suppressing NF-κB pathway activation. This interaction attenuates DNA damage and inflammatory responses, revealing a novel circRNA-mediated mechanism in pulpitis pathogenesis. Our findings advance the understanding of circRNA functions in oral diseases and propose circ_0002456 as a potential therapeutic target for pulpitis management.

Methods

Patient dental pulp samples

Human pulp tissue were extracted from orthodontic extraction, impacted teeth, and teeth requiring root canal treatment in clinical practice. All teeth were collected from the Affiliated Stomatology Hospital of Guangzhou Medical University, after informed consent from each patient and permission from the Affiliated Stomatology Hospital of Guangzhou Medical University (No.JCYJ2023005). Twenty-six dental pulp tissues were collected. The obtained pulp tissues were preserved in liquid nitrogen for RNA and protein extraction. A complete pulp tissue is divided into two parts using ophthalmic scissors to extract RNA and protein. This study employed a staged experimental design. In the first phase, the association between γ-H2AX and IL-6 was investigated utilizing 4 healthy and 8 inflamed samples. Based on preliminary findings, the second phase expanded the research scope by supplementing with an additional 5 healthy and 9 inflamed samples to systematically analyze the interrelationships among γ-H2AX, IL-8, IL-1β, and TNF-α.

Cell culture

hDPSCs were derived from the premolars and third molars of healthy donors aged 13 to 21, who were treated at the Affiliated Stomatology Hospital of Guangzhou Medical University. Pulp tissue was separated from the pulp chamber and digested with 3 mg/mL collagenase at 37 ℃ for 40 min. Subsequently, cells were cultured in a medium containing 10% fetal bovine serum and 1% double antibiotics at 37 ℃ and 5% carbon dioxide. Cells between the second and sixth generations were used for subsequent experiments. The hDPSCs obtained by this method have been analysed by flow cytometry and have several differentiation potentials [3]. All experiments were approved by the Ethics Committee of the Affiliated Stomatology Hospital of Guangzhou Medical University (No.JCYJ2023005).

Cell transfection

The siRNA targeting circ_0002456 and FUS, as well as the negative control siRNA (si-NC), were synthesized by Sangon Biotech (China). Additionally, the overexpression plasmids for circ_0002456 and FUS, along with the pcDNA3.1 vector, were also synthesized by Sangon Biotech (China). When hDPSCs reached 70–80% confluence, transfection of small interfering RNA and overexpressed plasmids were carried out using Lipofectamine 3000 (Invitrogen, USA) following the manufacturer’s instructions. To verify the transfection efficiency, Quantitative real-time polymerase chain reaction (qRT-PCR) was used to detect the expression level of circ_0002456, and the expression level of FUS was detected by Western blot. After transfection for 48 h, hDPSCs was stimulated by LPS with a concentration of 1 µg/mL.

RNA Preparation and qRT-PCR

Total RNA from pulp tissue and hDPSCs was extracted via TRIzol reagent (Invitrogen, USA) and reverse transcribed via the Evo M-MLV reverse transcription reagent (Accurate Biology, China). The primer sequences used were synthesized by Sangon Biotec. Amplification of the cDNA was performed via SYBR Green Pro Taq HS premixed qPCR kit (Accurate Biology, China) according to the manufacturer’s instructions. The expression levels of the target genes were caculated via the 2-∆∆Ct method. The expression levels of the circRNAs and mRNAs were normalized to GAPDH. Primer sequences for qRT-PCR and PCR detection in this study were shown in Supplemental Table S1.

RNase R treatment

1000 ng of total RNA isolated from hDPSCs was incubated with RNase R (K3061, Apexbio, USA) for 5 min. The degradation of circ_0002456, DOCK1 and GAPDH mediated by RNase R was subsequently assessed via qRT-PCR (n = 3).

Western blot analysis

Cells were lysed in RIPA buffer (Beyotime, China) combined with a cocktail of protease inhibitors (Thermo Scientific, USA) and Phosphatase Inhibitor Cocktai (CW2383S, CWBIO, China). Equal amounts of proteins (20–30 µg) of different groups were separated by 12.5% SDS-PAGE and transferred to 0.22/0.45 μm PVDF membranes (Millipore, USA). The membrane was first blocked with 5% BSA for 1 h at room temperature and incubated at 4 °C overnight with primary antibodies: GAPDH (1:2500, ab9485, Abcam, UK), γ-H2AX (1:1000, ab81299, Abcam, UK), FUS (1:10000, 68262-1-Ig, Proteintech, China), NF-κB P65 (1:1000, #3033, Cell Signaling Technology, USA), Phospho-NF-κB p65 (1:1000, #4764, Cell Signaling Technology, USA). After washed with Tris-buffer saline containing 0.05% Tween 20 (TBST) for three times and 10 min each, the membranes were incubated with goat anti-rabbit (1:1000, A0208, Beyotime, China) or goat anti-mouse (SA00001-1, 1:5000, Proteintech, China) labeled with horseradish peroxidase at room temperature for 1 h. Visualization of the blotting using the Omni-ECL™ Femto photochemiluminescence kit (SQ202, Epizyme Biomedical, China) (n = 3).

Bioinformatics method

To predict miRNAs potentially interacting with circ_0002456, we utilized three databases: circBank, CircInteractome, and circAtlas. Among the candidate miRNAs, hsa-miR-1265 and hsa-miR-326 were identified as overlapping hits across all three databases. Functional and pathway enrichment analyses were subsequently performed on the downstream target genes of these two miRNAs to preliminarily screen the biological roles of circ_0002456. To investigate the protein-binding mechanism underlying circ_0002456’s function, potential RNA-binding proteins (RBPs) were predicted using Interactome, CSCD, RBPBD, and RBPmap. The protein FUS emerged as a common candidate in CSCD, RBPBD, and RBPmap. Functional enrichment analysis was further conducted on proteins predicted to interact with FUS.

RNA fluorescence in situ hybridization (FISH)

hDPSCs were seeded on cell climbing slides and subjected to FISH. Blue fluorescence (DAPI, C1005, Beyotime, China) and red fluorescence (circ_0002456, Cy3, RiboBio, China) were used to indicate the cell nuclei and circ_0002456, respectively. The samples were fixed with 4% paraformaldehyde for 10 min, and then washed with PBS three times for 5 min each time. Permeabilization was performed using 0.5% Triton X-100 at 4 °C for 5 min. Subsequent reagent preparation and procedures were conducted according to the instructions of the RiboTM Fluorescent In Situ Hybridization Kit (C10910, RiboBio, China). Imaging was performed using a confocal microscope (TCS SP5 II, Leica, Germany) (n = 3).

Immunofluorescence staining (IF)

hDPSCs were fixed with 4% paraformaldehyde, permeated with 0.5% Triton X-100 for 10 min at room temperature, then incubated in 5% BSA for 1 h at room temperature. Samples were then incubated with primary antibodies overnight at 4 °C. The following antibodies were used: γ-H2AX (1:250, ab81299, Abcam, UK), FUS (1:10000, 68262-1-Ig, Proteintech, China). The following day, samples were incubated with Coralite488-conjugated Goat Anti-Mouse antibody (SA00013-1, Proteintech, China). Nuclei were stained with DAPI (C1005, Beyotime, China). Immunofluorescence images were captured by confocal microscopy (TCS SP5 II, Leica, Germany) (n = 3).

RNA Pull-Down

The PierceTM Magnetic RNA-Protein Pull Down Kit (Pierce Biotechnology, Rockford, USA) was used to capture proteins bound to circ_0002456. The circ_0002456 probe was designed and synthesized by Sangon Biotech (China). Briefly, biotin-labeled RNA was added to 50 µL magnetic beads and incubated at room temperature for 30 min. Then the protein lysate was added and incubated for 1 h. Next, samples were incubated with the elution buffer at 37 °C for 30 min. The supernatants were subjected to Western blot analysis.

Mouse pulpitis model

All animal experiments were strictly conducted in accordance with the REACH 2.0 guidelines and approved by the Animal Ethics Committee of Guangdong Huawei Testing Co., Ltd. (No.202306006). A total of 55 C57BL/6 mice (8–10 weeks old, 20–30 g body weight, with balanced gender distribution) were selected, including 40 mice for the pulp exposure time-course study and 15 mice for the circ_0002456 overexpression experiment. Mice were purchased from Zhuhai Best Biotechnology Co., Ltd., quality-certified by the Guangdong Provincial Experimental Animal Monitoring Center, and housed under specific pathogen-free (SPF) conditions. They were randomly divided into four experimental groups based on pulp exposure durations: 0 h (control group), 1 h, 3 h, and 5 h, with 10 mice per group. Anesthesia was administered via intraperitoneal injection of tribromoethanol (0.2 mL/10 g, Nanjing Aibei Biotechnology Co., Ltd.). Under 30× magnification using a Leica EZ4W surgical microscope, a 1/4 round bur (diameter: 0.5 mm Henan Deruibang Medical Device Co., Ltd., China) was used to create an occlusal pulp exposure cavity in the maxillary first molar until the pulp was visualized. No signs of pain or distress were observed during or after the procedure, and all animals were included in subsequent experiments. At predefined time points, euthanasia was performed via cervical dislocation. Maxillae were harvested for structural analysis by Micro-CT scanning. A total of 80 maxillary first molars were collected, with 2 teeth per group used for RNA extraction, 3 teeth for protein isolation, and one maxillary side (1 tooth) for immunohistochemistry (IHC), with three biological replicates. Experimenters were blinded to group assignments. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was used to detect circ_0002456 and associated inflammatory factors (n = 3). Western blotting and immunohistochemistry (IHC) were employed to evaluate protein damage (n = 3).

To investigate the role of circ_0002456 in vivo, 0.5 µL of Entranster™ in vivo transfection reagent complex containing either a circ_0002456 overexpression plasmid (2 µg/µL) or an empty vector (control) was injected into the pulp chamber using a microsyringe (needle tip diameter: 0.26 mm, Shanghai Baolige Biotechnology Co., Ltd., China) following 5 h of pulp exposure. A sterile cotton ball saturated with the transfection mixture was placed into the pulp chamber and sealed with glass ionomer cement (GIC). Forty-eight hours post-transfection, mice were euthanized by cervical dislocation. Bilateral maxillary first molars (30 teeth total) were collected, with 2 teeth per group used for RNA extraction, 3 teeth for protein isolation, and three biological replicates. circ_0002456 transfection efficiency and inflammatory factor expression were analyzed via qRT-PCR (n = 3). Western blotting was performed to assess γ-H2AX expression levels across groups (n = 3).

Micro-CT analysis

The maxillary bone along with its dentition was secured on a specially designed fixture and subjected to micro-computed tomography analysis using a SkyScan 1172 system (Kontich, Belgium) with acquisition parameters set at 60 kV and 100 mA, employing a 0.5-mm aluminum filter. During the imaging process, the mounted specimen completed a 180° rotational movement about its vertical axis, with incremental rotation steps of 0.4° between projections. Subsequent image reconstruction was performed using NRecon software (version 1.6.3, Bruker-microCT) to generate serial axial sections revealing the internal anatomical features of each specimen (n = 3). After the scanning is completed, the sample is used for IHC.

Immunohistochemistry (IHC)

Mouse maxilla were fixed with 4% paraformaldehyde for 24 h. The fixed tissues were then decalcified and dehydrated, embedded in paraffin wax, and sectione using a microtome. The paraffin sections were deparaffinized, rehydrated and subjected to antigen retrieval by EDTA thermal repair method. The sections were incubated with 3% hydrogen peroxide in methanol for 30 min to block endogenous peroxidase activity, then incubated in 5% BSA for 1 h at room temperature. The sections were incubated at 4 °C overnight with the following primary antibody: γ-H2AX (1:00; Affinity Biosciences Inc, USA). After incubation, the specific binding was visualized by using diaminobenzidine (DAB, ZSGB-BIO, China). The nucleus was counterstained with haematoxylin. The tissue sections of the maxillary first molar of mice were observed. The light yellow and brown staining were γ-H2AX positive immunostaining (n = 3).

Statistical analysis

Data analysis was performed with GraphPad Prism 8 (GraphPad, La Jolla). The Student’s t-test (two-tailed) was employed to examine the statistical significance of the differences between the two groups. One-way analysis of variance (ANOVA) was utilized to test the differences among multiple independent groups, and linear analysis was conducted to detect the correlations between the indicators. All data are shown as the means ± SD from three independent experiment. Statistical significance was defined as P < 0.05.

Results

Positive correlation between DNA damage and inflammatory factor expression in clinical dental pulp tissues

In clinical dental pulp tissues, Western blot confirmed the high expression of the DNA damage protein marker γ-H2AX in inflamed pulp tissues (Fig. 1A, Fig.S1A), and qRT-PCR results revealed a strong positive correlation with statistical significance between γ-H2AX levels and inflammatory factors IL-8, TNF-α, and IL-6 (Fig. 1B–C, Fig.S1B). While IL-1β demonstrated a weak positive correlation with γ-H2AX expression (r = 0.4817), this association lacked statistical significance (P = 0.0811, Fig. 1D). C57BL/6 mouse pulpitis model has high stability and repeatability. We constructed a mouse pulpitis model to further evaluate the relationship between DNA damage and inflammation of dental pulp tissue (Fig. 1E). Micro-CT three-dimensional reconstruction confirmed that the pulp exposure site was located on the occlusal surface of the maxillary first molar, with no crown fracture. Sagittal and coronal images revealed no low-density shadows in the pulp-exposed area or periapical region, consistent with radiographic features of pulpitis (Fig. 1F). Western blot and immunohistochemistry analyses detected γ-H2AX protein levels in mouse pulp tissues at 0, 1, 3, and 5 h post-exposure. Results showed a gradual increase in γ-H2AX, which paralleled the upregulation of IL-6 and IL-1β mRNA levels quantified by qRT-PCR (Fig. 1G–J). Although the relationship between DNA damage and IL-1β may require validation in larger samples, these findings collectively indicate a significant positive correlation between DNA damage response and inflammatory activation. However, further experimental validation is required to elucidate its precise mechanistic role.

Fig. 1.

Correlation analysis between pulpitis and DNA damage. A Expression levels of γ-H2AX protein in clinical dental pulp tissues. B–D Expression of γ-H2AX protein in clinical dental pulp tissues and its correlation with IL-8, TNF-α and IL-1β. E The procedure of pulp exposure for maxillary molars in mice. F MicroCT shows the sagittal and coronal view after pulp exposure in mice. G The immunohistochemical results of γ-H2AX in the pulp tissues of mice. H Western blot was used to detect the expression level of γ-H2AX protein in mouse pulp tissue. I–J The expression levels of IL-6, IL-1β in the pulp. *p < 0.05, **p < 0.01, ***p < 0.001(Full-length blots were presented in Supplementary Figure S4A–B)

circ_0002456 was downregulated in hDPSCs inflammatory model

A cellular inflammation model of hDPSCs was established with 1 µg/ml LPS stimulation for 5 h. The expression profiles of circRNAs in the cells before and after stimulation were analysed by high-throughput sequencing (Fig. 2A). We screened out 13 human-mouse homologous circRNAs with down-regulated expression, and verified them by qRT-PCR. Among the significantly downregulated circular RNAs, circ_0002456 has the greatest degree of down-regulation (Fig. 2B). Subsequently, the miRNAs and downstream target genes that circ_0002456 might bind were subjected to GO analysis and Wiki Pathways analysis (Fig. 2C–D), and it was found that circ_0002456 might be related to DNA damage and repair. circ_0002456 is a circRNA formed by reverse splicing from exons 26 and 27 of DOCK1, 224 base pairs long, located on human chromosome 10. The reverse splicing site of circ_0002456 was identified by Sanger sequencing (Fig. 2E). Agarose gel electrophoresis was performed with the amplification products of Convergent primers and Divergent primers, and found that only Divergent primers could amplify the bands (Fig. 2F). The above results confirm that circ_0002456 has a ring structure formed by head-tail splicing. Subsequently, the stability of circ_0002456 was tested, and the results of qRT-PCR proved that circ_0002456 could resist the digestion of RNase R and had stronger stability (Fig. 2G). FISH and RNA nucleoplasmic isolation experiments were performed. The FISH showed that circ_0002456 was distributed in both the nucleus and cytoplasm of hDPSCs, and the fluorescence signal in the nucleus was stronger (Fig. 2H). RNA nucleoplasmic isolation experiments further revealed that in hDPSCs, about 62% of circ_0002456 was distributed in the nucleus and about 38% in the cytoplasm (Fig. 2I). Based on the above results, we concluded that circ_0002456 has a stable ring structure and is distributed in both nucleus and cytoplasm, which is related to inflammation and DNA damage.

Fig. 2.

circ_0002456 was downregulated in hDPSCs inflammatory model. A Volcanic map: Red dots represent up-regulated circRNA, blue dots represent down-regulated circRNA, and gray dots represent no significant difference (|log₂ FC| > 2, p-value < 0.001). B The expression levels of the top 13 down-regulated human and mouse homologous circRNAs by qRT-PCR. C–D GO analysis and Wiki Pathways were used to analyze the function of circ_0002456. E Genomic location of circ_0002456 and Sanger detection of the reverse splicing sites of circ_0002456. F Agarose gel electrophoresis showed that the divergent primers could not amplify circ_0002456 from genomic DNA (gDNA). G Expression changes of circ_0002456 and its parental gene DOCK1 after RNase R treatment. H FISH assay for the localization of circ_0002456 in cells; The circ_0002456 probe was labeled with Cy3 (red), and the cell nucleus was labeled with DAPI (blue). I Detection of the localization of circ_0002456 in cells by qRT - PCR. *p < 0.05, **p < 0.01, ***p < 0.001

circ_0002456 inhibited DNA damage and inflammatory expression

In order to further investigate the effect of circ_0002456 on LPS-induced DNA damage and inflammatory response, we designed and synthesized two small interfering RNAs (siRNA1 and siRNA2) and an overexpressed plasmid targeting circ_0002456. When siRNAs and plasmid were transfected into hDPSCs, the expression level of circ_0002456 was effectively regulated without affecting the expression of its parental gene DOCK1 (Fig. S2A–D). After treating hDPSCs with knockdown and overexpression of circ_0002456, we performed stimulation using 1 µg/ml LPS for 5 h to probe the function of circ_0002456. Western blot results showed that compared with the control group, the expression level of DNA double-strand break marker γ-H2AX protein was significantly up-regulated in the circ_0002456 knockdown cell group (Fig. 3A). IF results were similar, showing that the fluorescence intensity of γ-H2AX was significantly enhanced in the cell group knocked down circ_0002456 (Fig. 3B). Meanwhile, the qRT-PCR results showed that the related inflammatory factors IL-6, IL-8, IL-1β and TNF-α were also significantly upregulated compared with the control group (Fig. 3C). These results indicate that downregulation of circ_0002456 expression in hDPSCs exacerbates LPS-induced DNA damage and inflammatory responses. Conversely, transfection of the circ_0002456 overexpression plasmid demonstrated the opposite effect, where elevated circ_0002456 expression suppressed DNA damage and inflammation in LPS-stimulated hDPSCs. (Fig. 3D–F).

Fig. 3.

circ_0002456 inhibits DNA damage and inflammatory expression. A–B, D–E Western blot and Immunofluorescence were used to detect the expression levels of γ-H2AX proteins after LPS treatment and circ_0002456 silencing or overexpression. C, F The expression levels of inflammatory factors after LPS treatment and circ_0002456 silencing or overexpression. *p < 0.05, **p < 0.01, ***p < 0.001 (Full-length blots were presented in Supplementary Figure S4C–D)

circ_0002456 binds to FUS protein to regulate DNA damage and inflammatory expression

CircRNA can regulate protein function by binding to proteins. Using databases such as CircInteractome, CSCD, RBPBD, and RBPmap, we predicted potential circRNA binding proteins and performed KEGG enrichment analysis (Fig. 4A–B). Notably, the RNA binding protein FUS was identified as a candidate in all three databases: CSCD, RBPBD, and RBPmap. KEGG pathway analysis of downstream candidate binding proteins revealed that FUS is associated with nucleocytoplasmic transport and DNA damage. Importantly, previous studies have reported that FUS can participate in DNA damage repair. IF experiments showed that FUS was mainly located in the nucleus and had a spatial overlap with circ_0002456 (Fig. 4C). Subsequently, in order to further prove whether circ_0002456 could bind to the FUS binding protein, we conducted an RNA pull-down assay and found that circ_0002456 could directly bind to FUS through Western blot (Fig. 4D). After knocking down and overexpressing circ_0002456, hDPSCs were stimulated with LPS for 5 h, and the protein expression level of FUS was detected by Western blot. The results indicated that, compared with the control group, the protein expression level of FUS in the circ_0002456 knockdown group decreased significantly (Fig. 4E). On the contrary, the circ_0002456 overexpression group showed significantly up-regulated FUS protein expression (Fig. 4F). Subsequently, to investigate the functional role of the FUS binding protein in LPS-induced DNA damage and inflammatory responses in hDPSCs, we designed and synthesized two FUS-specific siRNAs (siRNA1 and siRNA2) and an overexpression plasmid, which were transfected into hDPSCs. Western blot results confirmed that both siRNAs and the overexpression plasmid significantly altered FUS protein expression levels in hDPSCs (Fig. S2E–F). After 5 h of LPS stimulation, Western blot analysis revealed a significant increase in the expression of γ-H2AX in the FUS-knockdown group compared to the control (Fig. 5A). Consistent with this, IF assays showed a marked upregulation of γ-H2AX fluorescence intensity following FUS knockdown (Fig. 5B). qRT-PCR results further demonstrated significant upregulation of inflammatory cytokines, including IL-6, IL-8, IL-1β, and TNF-α, in the FUS-knockdown group (Fig. 5C). These findings indicate that reduced FUS expression exacerbates LPS-induced DNA damage and inflammatory responses in hDPSCs. Conversely, transfection of the FUS overexpression plasmid produced opposite effects (Fig. 5D–F), demonstrating that FUS overexpression suppresses LPS-induced DNA damage and inflammatory responses.

Fig. 4.

circ_0002456 binds to FUS protein. A circ_0002456 downstream target proteins were predicted by CircInteractome, CSCD, RBPBD and RBPmap databases. B Perform KEGG analysis on the predicted target proteins. C Immunofluorescence of the spatial localization correlation between circ_0002456 and FUS. D The binding of circ_0002456 to FUS was detected by RNA assay. E–F Western blot analysis was used to detect the expression levels of FUS after circ_0002456 silencing and overexpression. *p < 0.05, **p < 0.01, ***p < 0.001 (Full-length blots were presented in Supplementary Figure S4E–G)

Fig. 5.

FUS inhibits DNA damage and inflammatory expression. A–B, D–E Western blot and Immunofluorescence were used to detect the expression levels of γ-H2AX proteins after LPS treatment and FUS silencing or overexpression. C–F qRT-PCR was used to detect the expression levels of inflammatory factors after LPS treatment and FUS silencing or overexpression. *p < 0.05, **p < 0.01, ***p < 0.001 (Full-length blots were presented in Supplementary Figure S4H–I)

circ_0002456 regulates DNA damage and inflammation of hDPSCs through FUS

To investigate the synergistic regulatory role of circ_0002456 and the FUS binding protein in LPS-induced DNA damage and inflammatory responses in hDPSCs, we conducted rescue experiments. Western blot analysis showed that transfection of FUS-specific siRNA significantly upregulated the protein expression of γ-H2AX, and immunofluorescence assays further confirmed enhanced γ-H2AX fluorescence intensity. Notably, when hDPSCs were co-transfected with the circ_0002456 overexpression plasmid and FUS siRNA, both Western blot and IF results demonstrated that circ_0002456 overexpression effectively reversed the FUS knockdown-induced upregulation of γ-H2AX protein levels and fluorescence intensity (Fig. 6A–B). Similarly, qRT-PCR analysis revealed that FUS silencing markedly increased the mRNA expression of inflammatory cytokines IL-6 and IL-8. Importantly, co-transfection of the circ_0002456 overexpression plasmid with FUS siRNA also reversed the elevated IL-6 and IL-8 mRNA levels caused by FUS knockdown (Fig. 6C – D). These results strongly suggest that circ_0002456 modulates LPS-induced DNA damage and inflammatory responses in hDPSCs through its interaction with the FUS protein.

Fig. 6.

circ_0002456 regulates LPS-induced DNA damage and inflammation through FUS. A–B Western blot and immunofluorescence were used to detect the expression levels of γ-H2AX in cells after overexpressing circ_0002456 and knocking down FUS. C–D qRT-PCR was used to detect the expression levels of inflammatory factors after overexpressing circ_0002456 and knocking down FUS. *p < 0.05, **p < 0.01,***p < 0.001 (Full-length blots were presented in Supplementary Figure S4J)

circ_0002456 inhibits FUS nuclear output

Western blot and immunofluorescence assay were used to investigate whether DNA damage induced by LPS stimulation of hDPSCs would be accompanied by nucleoplasmic shift of FUS. Western blot results showed that the expression of FUS protein in hDPSCs decreased after LPS stimulation (Fig. 7A). Analysis of FUS protein content in the nucleus and cytoplasmic components of hDPSCs showed that the expression level of FUS protein in the nucleus decreased and the expression of FUS protein in the cytoplasm increased after LPS stimulation of hDPSCs (Fig. 7B). The results of IF experiments were similar, and the fluorescence intensity of FUS in the cytoplasm significantly increased after LPS stimulation (Fig. 7C). These results suggest that LPS stimulation of hDPSCs causes FUS protein translocation into the cytoplasm. In order to investigate whether circ_0002456 binding to FUS protein affects the nucleoplasmic shift of FUS protein, we performed immunofluorescence experiments. The results showed that the fluorescence intensity of circ_0002456 in the nucleus was decreased after LPS stimulation of hDPSCs (Fig. 7D). The results of RNA karyoplasmic separation showed that circ_0002456 occupied less in nucleus and more in cytoplasm (Fig. 7E). These results suggest that circ_0002456 may be involved in the nucleoplasmic shift of FUS proteins. Subsequently, we transfected circ_0002456 overexpression vector to investigate the effect of circ_0002456 expression on the nucleoplasmic shift of FUS protein. Western blot results showed that after circ_0002456 overexpression, the expression of FUS protein increased in the nucleus and decreased in the cytoplasm, indicating that circ_0002456 could inhibit the nuclear output of FUS (Fig. 7F). In summary, circ_0002456 can directly bind to FUS protein. When stimulated by LPS, the reduction of circ_0002456 promotes the movement of FUS to the cytoplasm, resulting in the reduction of FUS proteins in the nucleus, which aggravates DNA damage and inflammation.

Fig. 7.

circ_0002456 inhibits the nuclear export of FUS. A–B Western blot analysis was performed to detect FUS expression levels and changes in nuclear and cytoplasm expression in LPS-treated cells. C Immunofluorescence was used to detect the nuclear and cytoplasmic distribution of FUS in LPS-induced hDPSCs. D FISH was used to detect the nuclear and cytoplasmic distribution of circ_0002456 in LPS-induced hDPSCs. E Nuclear and cytoplasmic distribution of circ_0002456 in hDPSCs induced by LPS were examined by qRT-PCR. F Western blot analysis was performed to detect FUS expression levels in the nucleus and cytoplasm after LPS treatment and circ_0002456 overexpression. *p < 0.05, **p < 0.01,***p < 0.001 (Full-length blots were presented in Supplementary Figure S4K–M)

circ_0002456 and FUS regulate DNA damage in inflammatory responses through the NF-κB pathway

Studies have shown that DNA damage activates the NF-κB pathway, and when the NF-κB pathway is activated, it stimulates the release of inflammatory factors and maintains the long-term existence of inflammation in the body [22, 23]. To determine if circ_0002456 or FUS overexpression inhibits NF-κB activation, we transfected cells with their respective overexpression plasmids. After circ_0002456 and FUS were overexpressed, LPS was used to stimulate hDPSCs for 5 h, and Western blot was performed to detect NF-κB pathway-associated proteins. Western blot results showed that, compared with the control group, the expression of p65 and phosphorylated p65 (p-p65) was downregulated in the circ_0002456 and FUS overexpression group (Fig. 8A-C, F-H), indicating that circ_0002456 and FUS may inhibit the NF-κB pathway by affecting the expression and phosphorylation of p65. Meanwhile, qRT-PCR results showed that the expressions of related inflammatory factors IL-6, IL-8, IL-1β, and TNF-α were decreased (Fig. 8E, J). However, when we overexpressed circ_0002456 and FUS and added an NF-κB pathway activator, the expression levels of DSBs marker γ-H2AX and related inflammatory factors increased (Fig. 8D–E, I–J). When we knocked down circ_0002456 and FUS and added NF-κB pathway inhibitors, the results were opposite (Fig. S3A–J). These results indicate that circ_0002456 and FUS can regulate DNA damage in inflammatory response through NF-κB pathway.

Fig. 8.

circ_0002456 and FUS regulate DNA damage in the inflammatory response through the NF-κB pathway. A–D Western blot analysis was performed to detect the protein expression levels of p-p65, p65 and γ-H2AX in cells after circ_0002456 overexpression alone or in combination with an NF-κB pathway activator. E After circ_0002456 overexpression alone or in combination with an NF-κB pathway activator, the expression levels of inflammatory factors were analyzed by qRT-PCR. F–I Western blot analysis was performed to detect the protein expression levels of p-p65, p65 and γ-H2AX in cells after FUS overexpression alone or in combination with an NF-κB pathway activator. J After FUS overexpression alone or in combination with an NF-κB pathway activator, the expression levels of inflammatory factors were analyzed by qRT-PCR. *p < 0.05, **p < 0.01,***p < 0.001 (Full-length blots were presented in Supplementary Figure S4N)

circ_0002456 inhibits DNA damage and inflammatory responses in pulpitis

To investigate the role of circ_0002456 in regulating DNA damage and inflammatory responses during pulpitis in vivo, its expression was evaluated in human dental pulp tissues. circ_0002456 expression exhibited a significant decrease correlated with progressive DNA damage and worsening inflammatory responses. (Fig. 9A–E). Following 5 h pulp exposure, circ_0002456 was markedly downregulated in mice, further supporting its role in inflammatory regulation. Based on the murine direct pulp capping model [24], dental pulp tissues were exposed for 5 h, followed by local injection of a circ_0002456 overexpression plasmid. The exposed sites were then sealed with glass ionomer cement (GIC) (Fig. 9F–G). After 48 h, the functional effects of circ_0002456 were assessed in vivo. At 48 h post-treatment, circ_0002456 was highly expressed in dental pulp tissues (Fig. 9H). Western blot analysis revealed a significant downregulation of γ-H2AX in the circ_0002456 overexpression group compared with the control group (Fig. 9I). Simultaneously, qRT-PCR results demonstrated decreased expression levels of IL-6 and IL-1β in the circ_0002456 overexpression group (Fig. 9J-K). These findings suggest that circ_0002456 exerts inhibitory effects on DNA damage and inflammatory responses in pulpitis.

Fig. 9.

circ_0002456 inhibits DNA damage and inflammatory responses in pulpitis. A Correlation between circ_0002456 and γ-H2AX in human dental pulp tissue. B–D Correlation between circ_0002456 and IL-8, IL-1β and TNF-α in human dental pulp tissue. E The expression levels of IL-6,IL-1β and circ_0002456 in mouse pulp tissue. F The process of local plasmid injection. G The sagittal and coronal view after GIC filling. H The transfection efficiency in vivo was detected by qRT-PCR. I The expression of γ-H2AX in dental pulp tissue after overexpression of circ_0002456 in vivo was detected by Western blot. J–K The expressions of IL-6 and IL-1β in dental pulp tissue after overexpression of circ_0002456 in vivo were detected by qRT-PCR. *p < 0.05, **p < 0.01,***p < 0.001 (Full-length blots were presented in Supplementary Figure S4O)

Discussion

Pulpitis is a common inflammatory disease in the oral disease. During disease progression, the balance between Gram-negative and Gram-positive bacterial populations shifts dynamically. While Gram-positive bacteria dominate during the carious stage, the infection transitions to a mixed pattern with Gram-negative bacterial predominance as pulpitis develops [25, 26]. LPS, a key component of the outer membrane in Gram-negative bacterial cell walls, has been demonstrated to induce DNA damage in hDPSCs [27]. Based on this pathogenic mechanism, our study employed LPS stimulation to establish an in vitro cell model of pulpitis. We found that the down-regulation of circ_0002456 and DNA damage were related to the occurrence and development of pulpitis. Our research results contribute to a further understanding of the genetic-epigenetic interactions in the development of pulpitis.

DNA is an indispensable biomacromolecule in living organisms. DNA damage may occur under internal factors or external environmental factors, resulting in different degrees of harm. More and more studies have observed the presence of DNA damage in inflammatory diseases. Recently, elevated expression of the DSBs protein marker γ-H2AX has been found in pulpitis tissues [13, 28]. Consistent with previous studies, we found that DNA damage exists in clinically inflamed dental pulp tissue, and the degree of DNA damage is positively correlated with the expression level of inflammatory factors [29]. It is notable that when the pathway agonists of NF-κB were added, the expression of inflammatory factors increased, and the function of circ_0002456/FUS in inhibiting DNA damage was weakened. This result suggests that there seems to be a bidirectional regulatory relationship between DNA damage and the inflammatory response. Mechanistically, DSBs is one of the most potent DNA damage signals for NF-κB activation. In DNA damage signaling, ataxic telangiectasia mutant (ATM) kinase promotes the activation of NF-κB in response to external stress through phosphorylation and ubiquitination of the key regulator of NF-κB, NEMO (IKγ). Concurrently, DNA damage leads to the accumulation of cytoplasmic free DNA, which is recognized by cGAS to synthesize cGAMP to activate the STING-TBK1 signaling axis, thereby triggering NF-κB and IRF3 activation [30]. This cascade drives the excessive secretion of pro-inflammatory mediators including, IL-6, IL-1β, IFN, and TNF-α. In addition, after the activation of NF-κB, an inflammatory microenvironment is formed [31]. Chronic inflammation leads to tissue hypoxia, upregulating HIF-1α [32] and iNOS to sustain ROS/RNS production. It can lead to oxidative stress, induce DNA damage, and exacerbate local tissue damage [33, 34]. ROS and inflammatory mediators can also participate in NF-κB and MAPK pathway activation, establishing a self-amplifying signaling loop between DNA damage and inflammatory responses [35], forming a self-amplifying signaling loop between DNA damage and inflammation. The DNA damage response (DDR) senses DNA damage sites and recruits DNA repair proteins to repair the damage. As an epigenetic regulator, circRNA plays a critical role in DDR signaling pathways [36] where it either exacerbates or suppresses DNA damage processes through distinct molecular mechanisms, thereby modulating disease pathogenesis and progression [21, 37, 38]. Additionally, in bacterial-induced inflammatory oral diseases, the functional responses of dental tissue-derived stem cells (DTSCs) are governed by epigenetic regulation [39]. Study have reported that circ_0138960 regulates the NF-κB signaling pathway, protecting hDPSC from LPS-induced inflammation [54]. In this study, we performed high-throughput sequencing of hDPSCs cells before and after LPS stimulation, and found that circRNA expression was different and related to DNA damage. In a murine pulpitis model, we demonstrated that local administration of circ_0002456 significantly suppressed both DNA damage and inflammatory responses. These findings confirm that circ_0002456 can regulate DNA damage and inflammatory processes in pulpitis. The future development of sustained-release nanoparticles containing circ_0002456 may provide a promising therapeutic strategy to restore the self-renewal and differentiation capacity of DPSCs under inflammatory conditions, ultimately promoting regeneration of the dentin-pulp complex.

Mechanistically, circRNAs exert their functional roles primarily through either miRNA sequestration via competitive endogenous RNA (ceRNA) networks or dynamic associations with RNA-binding proteins (RBPs), thereby orchestrating post-transcriptional regulation and modulating protein activity [40–42]. FUS is a multifunctional DNA/RNA binding protein. Studies demonstrate that it not only regulates cellular functions through interactions with circRNA [37, 43], but also directly participates in the DNA damage repair process [38, 44]. FUS, predominantly localized within the nucleus, serves as a core regulatory factor in the early DDR. In this study, overexpression of circ_0002456/FUS was shown to mitigate DNA damage, with circ_0002456 attenuating cytoplasmic accumulation of FUS. Mechanistically, FUS achieves rapid recruitment to DSBs sites via its C-terminal RGG2 domain (C-terminal 468–526) [45], which mediates specific binding to PAR polymers. Furthermore, FUS orchestrates DNA repair through liquid-liquid phase separation (LLPS), forming membraneless condensates that recruit critical repair factors (e.g., XRCC1/DNA ligase IIIα) from both NHEJ and HR pathways, thereby facilitating PARP1-dependent DNA ligation [46–48]. Notably, nuclear depletion of FUS reduces NHEJ/HR repair efficiency by approximately 30% [48, 50]. DSBs induced activation of DNA-dependent protein kinase (DNA-PK) triggers phosphorylation of FUS at N-terminal residues, promoting its nuclear-to-cytoplasmic translocation and impairing damage-induced transcriptional regulation [50]. In amyotrophic lateral sclerosis (ALS), FUS-NLS mutations cause cytoplasmic mislocalization, resulting in PARP-dependent DNA repair defects and neuroinflammatory toxicity [51]. Additionally, under pathological conditions, FUS may undergo functional prioritization between distinct molecular pathways. While its C-terminal region (C-terminal 274–525) acts as a co-activator for NF-κB p65 to amplify inflammatory signaling [52], our findings paradoxically demonstrate that FUS overexpression suppresses both DNA damage and NF-κB pathway activation. This apparent contradiction may arise from functional competition between structural domains: the PAR-binding RGG2 domain (C-terminal 468–526) spatially overlaps with the p65-interacting region (C-terminal 274–525). Notably, human dental pulp stem cells (hDPSCs), as mesenchymal stem cells, exhibit superior DSBs repair capacity compared to differentiated cells [53]. Under LPS stimulation in hDPSCs, FUS appears to prioritize PAR-mediated DNA repair over inflammatory signaling. The observed suppression of NF-κB activation may thus reflect a mechanistic hierarchy wherein circ_0002456/FUS-mediated DNA damage resolution attenuates secondary inflammatory cascades.

This study established an in vitro pulpitis model using LPS stimulation. However, clinical pulp infections are predominantly driven by polymicrobial communities, where Gram-positive components like lipoteichoic acid (LTA) activate inflammatory pathways through Toll-like receptor 2 (TLR2) via mechanisms distinct from LPS-mediated TLR4 signaling. Future investigations should develop mixed-infection models simulating clinical polymicrobial conditions to comprehensively evaluate the pathogenic role of circ_0002456 in pulpitis development.

Conclusion

In summary, circ_0002456 can directly bind the FUS protein and restrict the nuclear output of the FUS to inhibit DNA damage, and finally reduce the LPS-induced inflammatory response of hDPSCs by inhibiting the activation of NF-κB pathway (Fig. 10). These findings reveal the important regulatory role of circ_0002456 and FUS in DNA damage and inflammation in pulpitis, and provide new strategies for future treatment of pulpitis. However, further studies are needed to confirm the efficacy and safety of circ_0002456 and its related target proteins as potential in vivo therapeutics.

Fig. 10.

circ_0002456 functional mechanism diagram: Under the stimulation of LPS, the expression of circ_0002456 and FUS decreased, and the low expression of circ_0002456 promoted the movement of FUS to cytoplasm and aggravated DNA damage. After DNA damage occurs, the NF-κB pathway is activated to promote the expression of inflammatory factors.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

QZJ conceived the study. LCL, HXD and FL designed the research and performed the experiments. LCL, FL and XHL performed in vivo studies, FL is an experimenter and has no knowledge of animal grouping. JKZ, BTW and LL analyzed the data. KHH and YXZ performed the statistical analysis and visualization processing. LCL and HXD wrote the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (no. 82270966), Guangdong Basic and Applied Basic Research Foundation of Natural Science Foundation project (no.2024A1515012741 and no.2022A1515111150).

Data availability

The data of this study are available within the article and its supplementary materials. The original sequencing data reported in this paper has been added to the supplement file.

Declarations

Ethics approval and consent to participate

All experiments using cells derived from human patients were reviewed and approved by the Institutional Review Board of the Hospital of Stomatology of Guangzhou Medical University [No.JCYJ2023005; Project title: Function and mechanism of circ_dsir(circ_dental stem cell inflammatory related) regulating DNA damage response in pulpitis; Date of approval: August 20, 2023]. All animal experiments are reviewed, approved and supervised by the Animal Ethics Committee of Guangdong Huawei Testing Company (No.202306006; Project title: Functions and mechanisms of circ_0002456 regulating the DNA damage response in the development of pulpitis; Date of approval: June 5, 2023).

Consent for publication

Written informed consent for publication was obtained from all participants.

Artificial intelligence

There are no works generated using artificial intelligence in this article.

Competing interests

The authors declare that they have no conflict of interest.