Reducing Connexin43 Boosts Cell Cleanup to Help Tooth Formation in Inflamed Dental Tissue

Healthy and intact premolars were extracted for orthodontic treatment purposes. All patients provided informed consent. hDPCs were isolated and prepared as described previously (Huang et al. 2006). Cells were maintained in α‐minimal essential medium (α‐MEM; Thermo Fisher Scientific Inc., USA) with 10% fetal bovine serum (FBS; Thermo Fisher) and 1% penicillin/streptomycin (Thermo Fisher) at 37°C in a humidified atmosphere of 95% air and 5% CO₂. Cells from passages 3 through 10 were used for experiments. hDPCs were cultured with odontogenic induction medium (OI) consisting of α‐MEM containing 10% FBS, 50 mg/mL ascorbic acid, 10 mM sodium β‐glycerophosphate, and 10 nM dexamethasone (Sigma‐Aldrich, USA) (Pei et al. 2016), with or without 0.1/5 μg/mL Escherichia coli‐derived LPS (Cat. No. L2630, Sigma) for 7, 14, and 21 days, replacing the medium every 3 days. The cells were divided into three experimental groups: control group (OI medium without LPS stimulation), 0.1 μg/mL LPS‐treated group, and 5 μg/mL LPS‐treated group.

Human third molars were collected from patients aged 20–40 years under strict exclusion criteria (incomplete root development, cracked teeth). Based on clinical and radiological examinations, they were categorised into four groups: (1) intact healthy teeth (control), (2) superficial caries, (3) intermediate/deep caries, and (4) irreversible pulpitis (n = 3 per group). Informed consent was obtained from all participants. The teeth were fixed in 4% paraformaldehyde solution for 48 h, followed by decalcification in 17% ethylenediaminetetraacetic acid (EDTA) (pH 7.4) at room temperature for 16 months. For light microscope analyses, tissues were embedded in paraffin. Thin sections (5 μm) were stained with haematoxylin and eosin or processed for immunofluorescence analysis.

To establish conditional knockout (cKO) mice with Cx43 deletion specifically in dental pulp cells, we generated dentine sialophosphoprotein (DSPP) Promoter‐Cre; Cx43^flox+/+ mice using the Cre/LoxP recombination system. Cx43^flox+/− heterozygous mice and DSPP pro‐CreERT mice were obtained from Cyagen Biosciences (USA). Briefly, male Cx43^flox+/− heterozygous mice were crossed with wild‐type C57BL/6 (B6) mice to generate Cx43^flox+/+ littermates, which served as controls. Subsequently, male Cx43^flox+/− heterozygous mice were crossed with CreERT⁺ mice to produce CreERT⁺; Cx43^flox+/− offspring. These mice were intercrossed to obtain CreERT⁺; Cx43^flox+/+ progeny. To induce Cx43 deletion in mouse dental pulp cells, 4‐week‐old CreERT⁺; Cx43^flox+/+ mice received intraperitoneal tamoxifen (Cat. No. T5648; Sigma) administration (75 mg/kg) for five consecutive days. Detailed results of cKO mouse generation and genotypic validation were provided in the Supporting Information.

Dentine injury was induced in the first maxillary molars of 4‐week‐old Cx43^flox+/+ (control) and cKO mice, following a standardised mechanical drilling protocol as previously described (Zhao et al. 2018). Mice were anaesthetised via intraperitoneal injection of ketamine (80 mg/kg) and xylazine (8 mg/kg). A 0.3‐mm‐diameter round bur (Cat. No. E0123; Dentsply Maillefer, USA) mounted on a high‐speed dental handpiece was used to create standardised cavities on the occlusal surface of the first maxillary molars. To minimise direct pulp exposure, drilling was primarily directed toward the dental fossa. However, due to the anatomical complexity of mouse molars (including multiple cusps, limited tooth size, and bur dimensions), minor enamel removal near or at the cusp tips was unavoidable. Mice were euthanised at postoperative days 1, 7, and 14. Maxillae were dissected and fixed in 4% paraformaldehyde for 48 h at 4°C, followed by decalcification in 17% EDTA (pH 7.4) for 3 months with solution replacement every 3 days. Decalcified specimens were dehydrated through a graded ethanol series, embedded in paraffin, and sectioned at 5‐μm thickness for histological analysis.

To inhibit Cx43 expression in cells, hDPCs were transfected with lentiviral particles carrying short hairpin RNA targeting Cx43 (shRNA‐Cx43) or a non‐targeting control (shRNA‐mock) at a multiplicity of infection (MOI) of 40, as previously described (Long et al. 2024).

Total RNA was extracted from cultured cells and first‐strand cDNA was synthesised from 0.5 μg of RNA using an Advantage RT‐for‐PCR Kit (Takara Bio Inc., JP). Gene expression was analysed by real‐time quantitative polymerase chain reaction (RT‐qPCR) with an Applied Biosystems (Thermo Fisher) 7500 Real‐time PCR System in 20 μL of Titanium One‐Step RT‐PCR Kit (Takara) containing 10 mmol/L each of the specific primers (Table 1). The results were standardised to the housekeeping gene GAPDH and expressed as relative mRNA levels.

Total protein was isolated in RIPA buffer (Sigma), separated by SDS‐PAGE, and transferred to PVDF membranes (EMD Millipore/Merck KGaA, USA). Subsequently, the membranes were incubated with a blocking buffer containing 5% non‐fat dried milk in TBS Tween‐20 buffer and cultured at 4°C overnight with antibodies against LC3 (Cat. No. ab192890, Abcam, UK; 1:2000), Beclin‐1 (Cat. No. 11306‐1‐AP, Proteintech, USA; 1:5000), sequestosome‐1/SQSTM1 (p62; Cat. No. 18420‐1‐AP, Proteintech; 1:2000), DSPP (Cat. No. bs‐8557R, Bioss, CN; 1:1000), dentine matrix protein‐1 (DMP‐1; Cat. No. bs‐25502R, Bioss; 1:1000), Osterix (Osx; Cat. No. ab209484, Abcam; 1:1000), runt‐related transcription factor 2 (Runx2; Cat. No. 20700‐1‐AP, Proteintech; 1:500), Cx43 (Cat. No. ab11370, Abcam; 1:2000), and GAPDH (Cat. No. 10494‐1‐AP, Proteintech; 1:10 000), followed by incubation with secondary antibodies at 37°C for 2 h. Blots were developed with a Supersignal West Pico chemiluminescent substrate kit (Thermo Fisher) with GAPDH serving as a loading control. Quantification of protein bands was performed using ImageJ software (NIH, USA).

Deparaffinised sections were rehydrated through graded ethanol and stained with haematoxylin (Beyotime Biotechnology Co. Ltd., CN) for 5 min, followed by differentiation in acid alcohol and bluing in tap water. Sections were counterstained with eosin (Beyotime) for 1 min, dehydrated in graded ethanol, cleared in xylene, and mounted with neutral resin. Tissue sections were digitally scanned using a KF‐PRO‐020‐HI high‐resolution slide scanner (KFBIO, CN) for microscopic observation and analysis.

Tissue slices or cell samples were permeabilized with 0.1% Triton X‐100 for 15 min and blocked with 1% bovine serum albumin (BSA) in phosphate buffered saline (PBS) for 1 h. Then, the sections were incubated with Cx43 (1:50), DSPP (1:200), LC3 (1:500), and p62 (1:200) primary antibodies at 4°C overnight. The next morning, the sections were washed three times with tris buffered saline (TBS) followed by incubation with ALEXAFluor 488/647‐conjugated secondary antibodies (Cat. No. ab150077/ab150115, Abcam; 1:100). The nuclei were counterstained with DAPI (Sigma). The sections were then analysed by fluorescence microscopy using a LEICA DM4000B microscope equipped with a Photometrics CoolSnap1 camera and software (Leica, DE).

Freshly harvested cell specimens were fixed in 2.5% glutaraldehyde (in 0.1 M phosphate buffer, pH 7.4) at 4°C for 12 h, followed by post‐fixation in 1% osmium tetroxide for 2 h. Samples were then dehydrated through a graded ethanol series (50%, 70%, 90%, and 100% ethanol, 15 min each) and embedded in Epon 812 resin. Polymerisation was performed at 60°C for 48 h. Ultrathin sections (80 nm) were cut using a diamond knife on a Leica Ultracut UCT ultramicrotome and mounted on 200‐mesh copper grids. Sections were stained with 2% uranyl acetate in methanol for 20 min (protected from light). Imaging was performed using a Hitachi HT7800 transmission electron microscope operated at 80 kV. Images were captured at magnifications of ×5000 to ×50 000 using a Gatan OneView CMOS camera.

The mRFP‐GFP‐LC3 plasmid was obtained from GeneChem Co. Ltd. (CN). hDPCs were seeded in 12‐well plates and transfected with the mRFP‐GFP‐LC3 plasmid for 8 h. Subsequently, the medium was replaced with complete culture medium. Cells were fixed with 4% paraformaldehyde for 30 min and counterstained with DAPI for 10 min. Autophagic flux was analysed by fluorescence microscopy, monitoring the distribution and alteration of mRFP‐GFP‐LC3B fluorescent signals. Autophagosomes (characterised by yellow puncta, GFP⁺/RFP⁺) and autolysosomes (characterised by red puncta, GFP⁻/RFP⁺) were quantified in individual cells with ImageJ software (NIH).

The alkaline phosphatase (ALPase) activity of cells was determined using an alkaline phosphatase assay kit (Beyotime) according to the manufacturer's protocol. hDPCs cultured for 7 days were fixed with 4% paraformaldehyde at room temperature for 30 min. Subsequently, a 5‐bromo‐4‐chloro‐3‐indolyl phosphate (BCIP)/nitro blue tetrazolium (NBT) solution was added to each well and incubated for 5 min at room temperature for the chromogenic reaction. After washing, stained cells were observed under a stereomicroscope (Model: SMZ800N; Nikon, JP). For quantitative analysis of BCIP/NBT staining, images were analysed with ImageJ software (NIH).

Cells cultured for 21 days were fixed with 4% paraformaldehyde at room temperature for 30 min. Fixed cells were then stained with 40 mM Alizarin Red (Cyagen, USA; pH 4.2) for 10 min at room temperature. Following thorough washing, mineralisation nodules were visualised and imaged using a stereomicroscope (Nikon). For quantitative analysis, stained areas were measured using ImageJ software (NIH).

All data were presented as the mean ± standard deviation (SD). Prior to statistical analysis, normality was assessed using the Shapiro–Wilk test, and homogeneity of variances was evaluated via Levene's test. All statistical analyses were performed using GraphPad Prism v9.0 (GraphPad Software; GraphPad Software Inc., USA), including Student's t‐test (unpaired) and two‐way ANOVA with Tukey's post hoc test. All experiments were independently repeated in triplicate. p < 0.05 was regarded as statistically significant.

LC3, a key autophagy‐related gene implicated in autophagosome formation (Mizushima et al. 2010; Castillo et al. 2013), and the reduction of p62, a phenomenon associated with enhanced autophagic flux (Yoshii and Mizushima 2017; Castillo et al. 2013), constitute two mechanistically distinct yet interconnected autophagy markers. As illustrated in Figure 2a, haematoxylin and eosin staining demonstrated marked inflammatory cell infiltration (black arrows) subjacent to the lesion, which intensified with infection severity. IF analysis revealed minimal co‐localization of LC3 and p62 immunoreactivity within the odontoblast layer of intact, healthy dental specimens (control group). Notably, progressive microbial infiltration induced spatial expansion of both biomarkers' expression patterns, exhibiting extensive distribution across the odontoblast layer, sub‐odontoblastic areas, and certain regions within the pulp proper.

Quantitative analysis using ImageJ revealed a progressive increase in both the distribution area and expression levels of LC3 and p62 with disease severity (Figure 2b), indicating a positive correlation between infection severity and immunoexpression intensity of these markers. These findings collectively suggest that: (1) autophagic machinery becomes activated in pulpal tissues during infectious and inflammatory processes; (2) escalated pathogenic challenge promotes autophagosome biogenesis as evidenced by LC3 accumulation; while (3) paradoxical p62 accumulation despite enhanced autophagosome formation implies potential impairment of autophagic flux completion, possibly indicating lysosomal dysfunction or defective autophagosome–lysosome fusion mechanisms under sustained or highly inflammatory conditions.

Beclin‐1 serves as a critical regulator of autophagosome formation (Mizushima et al. 2010; Castillo et al. 2013). To investigate the dynamic alterations in autophagy and autophagic flux under inflammatory conditions in hDPCs, we initially analysed the protein expression profiles of LC3, Beclin‐1, and p62 via WB. The results demonstrated that 0.1 μg/mL LPS treatment significantly upregulated LC3‐II/I ratios and Beclin‐1 protein levels while downregulating p62 expression across multiple time points compared to the control group (Figure 3a). These observations suggest that low‐grade inflammation enhances autophagic initiation and potentially promotes autophagic flux in hDPCs. In contrast, 5 μg/mL LPS exposure similarly elevated LC3‐II/I ratios and Beclin‐1 expression but paradoxically increased p62 accumulation relative to both control and low‐concentration LPS groups (Figure 3a), suggesting a potential impairment of autophagic flux despite persistent autophagic activation under high‐grade inflammatory stimulation.

To further validate these findings, we employed GFP‐RFP‐LC3 dual fluorescence plasmid transfection to monitor autophagic progression. Autophagosomes (yellow puncta) and autolysosomes (red puncta) were quantified through fluorescent microscopy. Compared to controls, 0.1 μg/mL LPS treatment significantly increased red puncta density (white arrows, Figure 3b) in hDPCs, confirming enhanced autophagosome–lysosome fusion and autophagic flux completion. Conversely, 5 μg/mL LPS exposure elevated yellow puncta count (yellow arrows) while reducing red puncta formation compared to the low concentration group (Figure 3b), demonstrating impaired lysosomal fusion capacity under high‐grade inflammation.

Complementary ultrastructural analysis via TEM after 7‐day LPS exposure in hDPCs revealed distinct autophagic morphologies. The 0.1 μg/mL LPS group exhibited increased autolysosome formation (indicated by red arrows), whereas 5 μg/mL LPS promoted autophagosome accumulation (blue arrows) with suppressed autophagosome‐lysosome fusion (Figure 3c), consistent with immunofluorescence observations. Collectively, these multimodal analyses establish a concentration‐dependent dual regulatory effect of inflammatory stimulation: low‐intensity inflammation facilitates complete autophagic flux through coordinated autophagosome formation and lysosomal degradation, while high‐intensity inflammation induces autophagic initiation but disrupts autophagic flux completion through impaired lysosomal fusion.

LPS, a predominant pathogen‐associated molecular pattern in gram‐negative bacterial cell walls, is clinically associated with progressive carious lesions and pulpitis pathogenesis, triggering innate immune responses in dental pulp tissues (Wang et al. 2015). Previous studies have demonstrated that LPS at the appropriate concentration may enhance the proliferation capacity and osteo/odontogenic differentiation potential of dental stem cells (Huang et al. 2015; Liu et al. 2019). To systematically evaluate the dose‐dependent effects of inflammatory stimulation on differentiation processes, we quantified the expression levels of odontogenic biomarkers—DSPP and DMP‐1 (Hanks et al. 1998), along with key transcription factors Runx2 and Osx (Chen et al. 2009), in hDPCs exposed to graded LPS concentrations (0.1 vs. 5 μg/mL). RT‐qPCR analysis revealed that 0.1 μg/mL LPS significantly elevated the mRNA levels of DSPP, DMP‐1, Runx2, and Osx in hDPCs across multiple time points compared to the control (Figure 4a), a trend corroborated by corresponding protein elevation (Figure 4b). In contrast, 5 μg/mL LPS not only abolished these pro‐differentiation effects but also suppressed both mRNA and protein expression below baseline levels (Figure 4a,b).

Functional assays further demonstrated that 0.1 μg/mL LPS enhanced ALP activity, whereas 5 μg/mL inhibited it (Figure 4c). Alizarin Red S staining confirmed that low‐grade inflammation promoted mineralised nodule formation, while high‐grade inflammation suppressed mineralisation (Figure 4c). These data establish that sustained high‐intensity inflammatory signalling disrupts the reparative odontogenic programme in hDPCs, suggesting a mechanistic link between uncontrolled pulp inflammation and impaired dentine regeneration.

Emerging evidence underscores the regulatory role of autophagy in odontogenic differentiation under physiological and pathological conditions, with autophagic flux serving as a pivotal mediator of this process (Pei et al. 2016; Zhan et al. 2020). However, our experimental data revealed that high‐concentration LPS significantly suppressed both autophagic flux and odontogenic differentiation in hDPCs. To delineate the mechanistic interplay between autophagy and odontogenic differentiation under inflammatory stress, we pharmacologically induced autophagy using rapamycin (Fu, Zhang, et al. 2024)—a mechanistic target of rapamycin (mTOR) inhibitor that initiates autophagosome formation and enhances lysosomal fusion—prior to 5 μg/mL LPS exposure.

The optimal rapamycin concentration (100 nM, 24‐h treatment) was determined via Cell Counting Kit‐8 (CCK‐8; Beyotime) viability assays (Figure 5a). Immunoblot analysis demonstrated that rapamycin pretreatment in LPS‐stimulated hDPCs markedly upregulated the LC3II/I ratio and Beclin‐1 protein level while downregulating p62 expression (Figure 5b), confirming enhanced autophagic flux. Corroborating this, GFP‐RFP‐LC3 dual fluorescence imaging revealed a pronounced increase in autolysosome formation (quantified by red puncta, white arrows) in rapamycin‐treated cells compared to the LPS‐only group (Figure 5c). These findings collectively indicate that rapamycin not only activates autophagy but restores functional autophagic flux in hDPCs under high‐inflammatory conditions.

Notably, rapamycin‐mediated autophagic flux potentiated the odontogenic differentiation of hDPCs, as evidenced by elevated mRNA and protein expression levels of key markers (DSPP, DMP‐1, Osx, and Runx2) in LPS‐stimulated cells (Figure 5d,e). Extended mineralisation assays further demonstrated that rapamycin significantly augmented ALP activity at day 7 and accelerated extracellular matrix mineralisation at day 21 in LPS‐challenged hDPCs (Figure 5f).

To validate the causal role of autophagic flux, we inhibited autophagy using 3‐methyladenine (3‐MA; Liu et al. 2021; 5 mM, 24‐h pretreatment) prior to low‐dose LPS stimulation (0.1 μg/mL). CCK‐8 assays confirmed the non‐cytotoxic 3‐MA concentration (Figure 6a). WB analysis revealed that autophagy inhibition by 3‐MA attenuated LPS‐induced LC3II/I ratio and Beclin‐1 protein level, while elevating p62 accumulation (Figure 6b). These biochemical changes were corroborated by GFP‐RFP‐LC3 dual fluorescence assay showing reduced autophagosome‐lysosome fusion (red puncta indicated by white arrows, Figure 6c), collectively demonstrating impaired autophagic flux. Crucially, 3‐MA suppressed the upregulation of DSPP, DMP‐1, Osx, and Runx2 mRNA and protein expression levels triggered by 0.1 μg/mL LPS (Figure 6d,e), concomitant with diminished mineralisation capacity (Figure 6f).

These data conclusively establish that functional autophagic flux is indispensable for odontogenic differentiation and mineralisation in hDPCs under inflammatory stress. Enhancing autophagy with rapamycin or inhibiting it with 3‐MA directly affects the odontogenic capacity of cells. This demonstrates that autophagy is a viable therapeutic target to counteract inflammation‐related damage in pulp repair.

Emerging evidence indicates that Cx43 exerts dual regulatory control over autophagy through: (1) direct protein–protein interactions with core autophagy machinery (e.g., LC3) to govern autophagosome formation (Iyyathurai et al. 2016; Bejarano et al. 2014), and (2) GJ channel activity‐dependent modulation of autophagic flux (Zou et al. 2020). Our previous studies have independently demonstrated two context‐dependent functions of Cx43 in dental pulp biology: (1) Hemichannel blockade attenuates LPS‐induced inflammatory responses by limiting DAMPs efflux (Hu et al. 2024), while (2) GJ channel potentiation enhances Ca²⁺‐mediated the odontogenic differentiation through Erk1/2 signalling (Li et al. 2015). Thus, we hypothesised that Cx43‐mediated autophagy regulation might orchestrate odontogenic differentiation. In the current investigation, inhibition of Cx43 in LPS‐challenged (5 μg/mL) hDPCs significantly elevated LC3II/I ratios and Beclin‐1 expression levels while suppressing p62 accumulation (Figure 7a). In line with these biochemical alterations, Cx43 suppression augmented autophagosome‐lysosome fusion as evidenced by GFP‐RFP‐LC3 reporter‐based immunofluorescence (red puncta indicated by white arrows, Figure 7b), while TEM structurally validated this enhancement via quantification of autophagic vacuoles with lysosomal characteristics (red arrows, Figure 7c). The coordinated elevation in fusion frequency across both imaging modalities established Cx43 inhibition as a regulator of autophagic flux progression in hDPCs.

Furthermore, Cx43 inhibition upregulated both mRNA and protein expression levels of odontogenic markers in 5 μg/mL LPS‐challenged hDPCs (Figure 8a,b). Functional validation revealed concomitant enhancement of ALP activity and mineralised nodule formation (Figure 8c), suggesting that Cx43 modulation critically regulates the inflammatory microenvironment‐compromised odontogenic differentiation. These coordinated findings in vitro position autophagic flux as a pivotal mechanistic link between Cx43 activity and differentiation potential under high‐inflammatory conditions.

To establish clinical relevance, we utilised dental pulp‐specific Cx43‐cKO mice (driven by the DSPP promoter‐Cre) in a maxillary first molar dentine injury model. Notably, cKO mice exhibited accelerated tertiary dentine formation compared to the Cx43^flox+/+ littermates, with mineralised matrix deposition (black boxes) increasing at days 7 and 14 post‐injury, respectively (Figure 9a). Spatiotemporal analysis revealed: (1) IF‐based LC3 signal intensification subjacent to injury sites at day 7 and 14 compared to the control (Figure 9b); (2) Concomitant p62 downregulation (Figure 9c); (3) Upregulated DSPP expression paralleling repair progression (Figure 9d). This multimodal in vivo evidence not only recapitulates our findings in vitro but also establishes the Cx43‐autophagy axis as a conserved regulator of pulp‐dentine complex regeneration under pro‐inflammatory conditions. Collectively, our integrated experimental approach spanning molecular profiling, functional assays, and transgenic models reveals a previously unappreciated Cx43 gatekeeper function: through autophagic flux modulation, it dynamically calibrates odontogenic differentiation capacity in inflammatory milieus, ultimately dictating dental pulp repair outcomes.