Immunomodulatory and pro-mineralizing effects of an injectable baicalein-loaded methacrylated gelatin hydrogel for vital pulp therapy
Beatriz Ometto Sahadi, Igor Paulino Mendes Soares, Chloe Gifford, Caroline Anselmi, Pedro Henrique Chaves de Oliveira, Renan Dal-Fabbro, Maedeh Rahimnejad, Marcelo Giannini, Marco C. Bottino
TL;DR
A new injectable hydrogel with baicalein shows promise for pulp therapy by reducing inflammation and promoting mineralization in dental tissues.
Contribution
The development of a GelMA hydrogel incorporating BA-loaded MSNs for sustained drug delivery and pulp regeneration.
Findings
Baicalein enhanced mineralized nodule formation and reduced inflammation in dental pulp stem cells and macrophages.
The GelMA/MSNs-COOH-BA hydrogel showed improved mechanical strength and sustained drug release over 10 days.
In vivo tests showed no significant differences in biocompatibility across formulations over 28 days.
Abstract
This study first investigated the biological function of baicalein (BA) and then developed a photocrosslinkable methacrylated gelatin (GelMA) hydrogel incorporating BA-loaded, carboxylated mesoporous silica nanospheres (MSNs-COOH-BA) for vital pulp therapy. Initially, BA (1–20 μM) was tested for cytocompatibility, in vitro mineralized nodule formation as an early indicator of odontogenic potential, and antioxidant/anti-inflammatory functionality on dental pulp stem cells (DPSCs) and macrophages. Then, 15% (w/v) GelMA was formulated with MSNs-COOH-BA (10 or 20 mg/mL). Hydrogels were characterized by SEM/EDS for their microstructure morphology and chemical composition, as well as for compression, swelling, degradation, and BA release. Biological assessments included DPSC cytocompatibility and early mineralization responses under or without LPS stimulation, macrophage cytokine modulation,…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsEndodontics and Root Canal Treatments · Flavonoids in Medical Research · Hydrogels: synthesis, properties, applications
Introduction
Dental caries is among the most commonly treated conditions in dental practice and is frequently diagnosed at advanced stages, resulting in deep cavities [1]. During caries removal, the risk of iatrogenic pulp exposure rises substantially [1,2]. In these situations, direct pulp capping may be indicated to stimulate reparative dentin formation and create a “biological seal” between the restorative material and the pulp tissue [3,4]. This strategy is part of vital pulp therapy (VPT), a conservative approach aimed at preserving pulp vitality while supporting the dentin-pulp complex's intrinsic repair and development, which can be compromised by carious lesions or traumatic injuries [3-5].
The dentin-pulp complex responds to injury through a combination of inflammatory and mineralization processes, and the balance between pulpitis and repair is critical for preserving vitality [6,7]. Reparative dentinogenesis is regulated by bioactive cues, including growth factors released from the dentin matrix upon injury, as well as calcium and hydroxyl ions liberated from clinically available pulp capping biomaterials [7-9]. Accordingly, direct pulp-capping materials should not only protect the exposure site but also promote reparative dentin formation [2-5]. Although widely used, calcium hydroxide and mineral trioxide aggregate (MTA) have limitations in consistently controlling inflammation due to their caustic behavior, highlighting the need for improved strategies that actively support dentinogenesis by stimulating the pulp's intrinsic regenerative responses [10-12].
To meet these challenges, gelatin-based biomaterials, particularly gelatin methacryloyl (GelMA), have emerged as a promising platform for developing new VPT strategies [10-12]. GelMA is a soft, porous, highly hydrated hydrogel with tunable network formation, cytocompatibility, and biodegradability [13,14]. Incorporating mesoporous silica into GelMA can enhance its mechanical performance and osteogenic potential [15,16], while providing drug-reservoir capacity for controlled release of bioactive agents, features suitable for injectable, biologically active pulp-capping applications.
Building on GelMA's drug-reservoir capacity, natural compounds have gained attention as bioactive molecules for local delivery due to their anti-inflammatory and antibacterial properties [17-19]. Among them, the flavonoid baicalein (BA) exhibits antioxidant, anti-inflammatory, and anti-carcinogenic activities [20-22]. Although BA's dose-dependent toxicity requires further clarification, it has been shown to promote mineralization in mesenchymal stem cells, indicating potential for hard-tissue regeneration [18,23,24]. To optimize its therapeutic effects and minimize cytotoxicity, BA can be first encapsulated within mesoporous silica nanoparticles (MSNs), which protect the compound from premature degradation and allow controlled, localized release [25,26]. These BA-loaded MSNs can then be incorporated into a GelMA matrix, providing a biocompatible and injectable platform for VPT that offers sustained delivery and enhances anti-inflammatory activity in pulp cells compromised by an inflamed environment.
Guided by this rationale, we designed a two-phase study. First, BA's dose-response was quantified in vitro, assessing cytocompatibility, mineralization, and antioxidant/anti-inflammatory activity. Second, a photoactivated GelMA hydrogel embedding BA-encapsulated MSNs was engineered, characterized, and biologically evaluated as a potential pulp-capping material, characterizing drug release and biological effects on dental pulp stem cells. This strategy aimed to combine GelMA's tunable, biocompatible scaffold with BA's regenerative and anti-inflammatory properties to support a pro-regenerative response from pulp cells in VPT applications.
Materials and methods
This study was conducted in two phases. First, different concentrations of baicalein were evaluated through cytotoxicity, mineralization, and anti-inflammatory assays. Second, a photoactivated methacrylated gelatin (GelMA) hydrogel incorporating baicalein-loaded mesoporous silica nanospheres (MSNs) was engineered as a potential pulp-capping material, designed to combine GelMA's tunable mechanical properties and biocompatibility with baicalein's regenerative and anti-inflammatory effects to foster pulp tissue healing and protection, and further assessed in vivo through a subcutaneous implantation model.
Phase I
2.1.
Viability of dental pulp stem cells (DPSCs) in contact with BA
2.1.1.
Considering that baicalein is a hydrophobic compound, a primary stock was prepared at 100 mM in 100% dimethyl sulfoxide (DMSO; Sigma-Aldrich). This was then diluted to a working stock containing 1% DMSO to minimize solvent exposure and lower the baicalein concentration before use. Final working solutions were prepared by diluting the 1% DMSO stock into α-Minimum Essential Medium (α-MEM; Gibco, Thermo Fisher Scientific) to yield 1, 2.5, 5, 10, and 20 μM baicalein. The maximum DMSO content in the culture medium was 0.02% (v/v).
For all biological assays, dental pulp stem cells (DPSCs) at passages 4–6 were used. Cells were seeded at 3 × 10^3^ cells/well in 96-well plates and allowed to adhere for 24 h at 37 °C in a humidified atmosphere of 5% carbon dioxide (CO_2_) using complete α-MEM supplemented with 15% fetal bovine serum (FBS; Gibco) and 1% penicillin-streptomycin. After 24 h, cells were treated with baicalein at 1, 2.5, 5, 10, or 20 μM by adding 100 μL of each solution per well, and incubation continued for an additional 24 h under the same conditions. The culture medium was then maintained and refreshed every 2 days for all groups until day 7. Cell viability was assessed on days 1, 3, and 7 using a 10% Alamar Blue solution (Invitrogen, Carlsbad, CA, USA). For each time point, Alamar Blue was added at 100 μL/well and incubated for 3 h at 37 °C in 5% CO_2_. Fluorescence was measured at 555/595 nm (SpectraMax iD3, Molecular Devices, San Jose, CA, USA) against blank wells. Cells maintained in α-MEM without baicalein served as the negative control (NC), and a 0.3% phenol solution (Reagents, Charlotte, NC, USA) served as the positive control (PC). Fluorescence intensities were converted to percentages and normalized to the NC to determine viability at each time point [27].
In vitro mineralized nodule formation as an early indicator of odontogenic potential in DPSCs exposed to different BA concentrations
2.1.2.
Mineral deposition was evaluated at 14 and 21 days using osteogenic medium (complete α-MEM supplemented with 100 nM dexamethasone, 10 mM β-glycerol phosphate, and 50 μg/mL ascorbic acid). DPSCs were seeded at 1 × 10^4^ cells per well in 48-well plates and allowed to adhere for 24 h at 37 °C in a humidified 5% CO_2_ atmosphere [27]. Cultures were assigned to two conditions: osteogenic medium alone or osteogenic medium supplemented with lipopolysaccharide (LPS, 10 μg/mL) from Escherichia coli O111:B4 (Sigma-Aldrich) to model an inflammatory environment [28]. The osteogenic medium was replaced every other day for the first 7 days, after which baicalein solutions (5–20 μM) were added and maintained for 48 h. Cultures were then continued to the 14- and 21-day endpoints, with medium changes every 2 days.
Antioxidant and anti-inflammatory actions in RAW 264.7 macrophages by BA
2.1.3.
To evaluate the antioxidant potential of BA, we performed the Reactive Oxygen Species (ROS) assay. RAW 264.7 murine macrophages were seeded at 5 × 10^4^ cells per well in 96-well plates and cultured for 24 h at 37 °C in a humidified 5% CO_2_ atmosphere in Dulbecco's Modified Eagle's Medium (DMEM; Gibco) supplemented with 10% FBS and 1% penicillin-streptomycin. Cells were then assigned to two conditions: with or without LPS exposure. For the LPS group, DMEM containing 0.1 μg/mL LPS was applied for 6 h; control wells received refreshed DMEM without LPS. Media were subsequently replaced with fresh DMEM containing the indicated concentrations of baicalein and incubated for 1 h [27]. The fluorescent probe H_2_DCFDA (2′,7′-dichlorodihydro-fluorescein diacetate; Invitrogen, Carlsbad, USA) was then added for 10–15 min, and fluorescence was recorded at 490 nm excitation and 520 nm emission using a SpectraMax iD3 plate reader.
To assess the anti-inflammatory potential, cytokine quantification was performed using the Enzyme-Linked Immunosorbent Assay (ELISA). RAW 264.7 cells were seeded at 1 × 10^5^ cells per well in 24-well plates and cultured under the same conditions described above. Cells were exposed or not to 0.1 μg/mL LPS for 6 h, after which media were replaced with fresh DMEM containing the designated baicalein concentrations and incubated for an additional 24 h. Supernatants were collected and analyzed for TNF-α (tumor necrosis factor-α), IL-6 (interleukin-6), and IL-1α (interleukin-1α) using cytokine-specific ELISA assays (BioLegend, San Diego, CA, USA) [29]. Absorbance was measured at 450 nm using a SpectraMax iD3 plate reader, following the manufacturer's instructions.
Phase II
2.2.
Functionalization of mesoporous silica nanospheres (MSNs)
2.2.1.
The functionalization of mesoporous silica nanospheres (MSNs) was carried out following the standardized protocol described by Qu et al. [28]. Briefly, MSNs (approximately 200 nm in diameter, 4 nm pore size) were amine-functionalized (MSNs-NH_2_) by dispersing 200 mg of particles in 50 mL of ethanol and adding 400 μL of (3-aminopropyl)triethoxysilane (APTES). The suspension was stirred at 400 rpm for 24 h at 60 °C, after which the particles were collected by centrifugation, washed three times with alternating ethanol and deionized water, and vacuum-dried. For carboxylation, 150 mg of MSNs-NH_2_ were redispersed in 10 mL ethanol containing 75 mg succinic acid, 98 mg EDC, and 98 mg HOBt, stirred at room temperature for 5 h, and subsequently isolated by centrifugation and washed alternately with ethanol and deionized water to obtain MSNs-COOH. Successful surface modification was confirmed by ATR-FTIR (Nicolet iS50, Thermo Fisher Scientific, Waltham, MA, USA) collected in the 4000–600 cm^−1^ range at 4 cm^−1^ resolution.
Preparation of baicalein-loaded MSNs and encapsulation efficiency
2.2.2.
Carboxylated MSNs (20 mg) were incubated with 10 mL of a baicalein solution in alcohol (2 mg/mL) and stirred at room temperature for 24 h. The baicalein-loaded MSNs-COOH were recovered by centrifugation, washed three times with PBS, and dried under vacuum. To determine loading efficiency, the supernatant and all wash fractions were pooled, and the residual (unloaded) baicalein was quantified by UV–Vis spectroscopy at 270 nm using a SpectraMax iD3 microplate reader. Loading efficiency (LE%) was calculated as: LE% = [(BA1 − BA2) / BA1] × 100, where BA1 is the initial amount of baicalein and BA2 is the amount remaining in the combined supernatant and washes [30].
Release kinetics of baicalein from functionalized MSNs (MSNs-COOH-BA)
2.2.3.
To evaluate the cumulative release kinetics of baicalein, BA-loaded nanospheres (MSNs-COOH-BA; 1, 2, 5, or 10 mg) were each suspended in 1 mL PBS and incubated at 37 °C for 10 days. A control containing 10 mg of MSNs-COOH without BA in 1 mL PBS was run in parallel. At predetermined intervals, 100 μL of supernatant was withdrawn and immediately replaced with an equal volume of fresh PBS to maintain constant volume. The amount of baicalein released in each aliquot was quantified by UV–Vis absorbance at 270 nm using a SpectraMax iD3 microplate reader, and concentrations were calculated based on a previously established standard curve. Cumulative release profiles were obtained by summing the amounts of BA collected at each time point.
Incorporation of MSNs-COOH-BA into GelMA
2.2.4.
The synthesis of gelatin methacryloyl (GelMA) followed standardized protocols as reported by previous studies [13,31,32]. Briefly, type-A gelatin (10 g; 10% w/v in 100 mL PBS) was dissolved at 50 °C under continuous stirring. Methacrylic anhydride (8 mL) was then added dropwise at a rate of 0.5 mL/min while maintaining the temperature at 50 °C. After 2 h, the reaction was quenched by adding 100 mL of preheated PBS (50 °C). The solution was transferred into hydrated dialysis membranes (MWCO 12–14 kDa; ~30 cm in length, each containing ~50 mL solution) and dialyzed against deionized water at 50 °C for 7 days with twice-daily water changes. The dialyzed solution was filtered, aliquoted into 50 mL tubes, frozen at −80 °C overnight, and lyophilized (Labconco FreeZone 2.5 L, Labconco Corporation, Kansas City, USA) for 7 days. The resulting gelatin methacryloyl (GelMA) foam was stored at −20 °C until use.
Lyophilized GelMA (150 mg) was dissolved in 1 mL PBS (15% w/v) [11,33] at 60 °C with stirring at 300 rpm until homogeneous. Baicalein-loaded MSNs (MSNs-COOH-BA; 10 or 20 mg) were added and mixed under the same conditions to ensure uniform dispersion. A control formulation contained non-functionalized MSNs-COOH at 20 mg/mL (i.e., 20 mg per 1 mL GelMA). After 24 h of stirring, lithium phenyl-2,4,6-trimethylbenzoyl phosphinate (LAP) was added to each mixture to a final concentration of 0.05% w/v and stirred for an additional 30 min [28]. Aliquots (120 μL) were cast into silicone molds (6 × 2 mm or 8 × 3 mm, depending on the assay) and photocrosslinked using a dental curing light (Bluephase Style, Ivoclar, Amherst, USA) for 30 s per side [11].
Morphological and elemental characterization of MSNs-COOH-BA-GelMA hydrogels
2.2.5.
Hydrogel microstructure and elemental composition were examined by field-emission SEM with EDS (MIRA3 FEG-SEM, Tescan USA, Warrendale, USA). Samples (n = 3 per group) were prepared, frozen, and lyophilized for 24 h, then sectioned to expose internal cross-sections. Specimens were sputter-coated with gold for 90 s and analyzed to assess pore morphology by SEM, with EDS spectra/maps collected to determine elemental distribution. Pore size analysis was performed using ImageJ software (NIH, Bethesda, MD, USA), with calibration based on the scale bar of each SEM image to determine the average pore diameter.
Collagenase-mediated baicalein release from MSNs-COOH-BA-GelMA hydrogels
2.2.6.
Hydrogel discs (6 × 2 mm; n = 6) were placed in glass vials (VWR International, Radnor, USA) containing 5 mL PBS with collagenase type A (1 U/mL) and incubated at 37 °C. At predetermined intervals over 10 days, 100 μL of supernatant was withdrawn and replaced with an equal volume of fresh PBS. Baicalein in the collected aliquots was quantified by UV–Vis spectroscopy at 270 nm using a SpectraMax iD3 microplate reader, and concentrations were obtained from a previously established standard curve.
Compression, swelling, and degradation of MSNs-COOH-BA-GelMA hydrogels
2.2.7.
For the compression test, hydrogel cylinders (8 × 3 mm; n = 6) were equilibrated in PBS at 37 °C for 24 h, gently blot-dried to remove surface moisture, and tested in compression under hydrated conditions using a mechanical tester (MTESTQuattro, ADMET, Norwood, USA) equipped with a 250 kgf load cell. Samples were compressed at a rate of 1 mm/min to ~25–30% strain. Stress-strain curves were recorded (MTESTQuattro Controller v5.07.13), the compressive Young's modulus was calculated from the linear elastic region, and the ultimate compressive strength was taken as the maximum stress prior to failure [30].
Regarding swelling capacity, discs (6 × 2 mm; n = 6) were immersed in 5 mL PBS at 37 °C for 24 h, removed, gently blotted with tissue, and weighed to obtain the wet mass (W_w_). The same specimens were then lyophilized for 24 h and reweighed to determine the dry mass (Wd). The swelling capacity (%) was calculated using the following formula: . To assess degradation rate, discs (6 × 2 mm; n = 6) were placed individually in 5 mL PBS containing collagenase type A (1 U/mL) and incubated at 37 °C. The initial weight (W_0_) was recorded, then specimens were weighed daily from day 0 to day 3 and every two days thereafter. At each time point (W_t_), samples were removed, rinsed twice with deionized water, gently blot-dried, and reweighed; the PBS-collagenase solution was replaced at every measurement. The degradation ratio (%) was calculated using the following formula [34]:
Cytocompatibility, early mineralized matrix formation, and anti-inflammatory action of MSNs-COOH-BA-GelMA hydrogels
2.2.8.
For the cytotoxicity (extract) test, cylindrical hydrogels (6 × 2 mm; n = 8) were UV-sterilized for 30 min per side, then each was placed in a sterile glass vial containing 5 mL α-MEM supplemented with 15% FBS, l-glutamine, 1% penicillin-streptomycin, and collagenase type A (1 U/mL) and incubated at 37 °C. At days 1, 3, and 7, 500 μL of the incubation medium (extract) was withdrawn and replaced with an equal volume of fresh supplemented medium. DPSCs were seeded at 3 × 10^3^ cells per well in 96-well plates and allowed to adhere for 24 h at 37 °C in 5% CO_2_. Extracts were sterile-filtered (0.22 μm), after which the culture medium was replaced with 100 μL of the filtrate, and the cells were exposed for 24 h. Metabolic activity was quantified using Alamar Blue for 3 h, and fluorescence was read at 555/595 nm on a SpectraMax iD3 with blank subtraction. Cells maintained in α-MEM served as the negative control (NC), and 0.3% phenol as the positive control (PC). Fluorescence values were expressed as a percentage of NC to determine cell viability at each time point [35].
To assess whether MSNs-COOH-BA-GelMA hydrogels promote early mineralized matrix formation as an indicator of osteogenic/odontogenic commitment under inflammatory conditions, an indirect-contact Transwell assay was performed [13,28]. DPSCs (1.5 × 10^4^ cells/well) were seeded in 24-well plates and, after attachment, the medium was replaced with osteogenic induction medium with or without 10 μg/mL LPS. Transwell inserts were placed into the same wells, and hydrogel formulations were added to the upper chambers to allow diffusion of bioactive factors while preventing direct cell-material contact, thereby enabling evaluation of soluble-factor–mediated cellular responses rather than direct material-induced differentiation. Cultures were maintained for 14 or 21 days, then mineral deposition was evaluated by Alizarin Red S. For staining, cells (n = 6) were fixed in 70% ethanol at 4 °C for 1 h, rinsed with deionized water, stained with 40 mM Alizarin Red (pH 4.2) for 15 min, and washed five times with deionized water. Mineralized nodules, representing early-stage matrix mineralization, were imaged using an optical microscope (Echo Revolve; BICO Company) and quantified by dye extraction with 10% cetylpyridinium chloride followed by absorbance measurement at 570 nm on a SpectraMax iD3. DPSCs cultured in osteogenic medium with and without LPS served as positive and negative controls, respectively.
To evaluate anti-inflammatory activity, hydrogel extracts generated after direct contact for 1, 3, or 7 days were collected and stored at −80 °C until use. RAW 264.7 murine macrophages (passage 8) were seeded in 24-well plates at 1 × 10^5^ cells/well and cultured in DMEM supplemented with 10% FBS, 1% penicillin-streptomycin for 24 h at 37 °C in 5% CO_2_. An inflammatory response was then induced with LPS (0.1 μg/mL) for 6 h, after which the medium was replaced with the respective hydrogel extracts. Following incubation, supernatants were harvested and assayed for TNF-α, IL-6, and IL-1α using ELISA kits (BioLegend) according to the manufacturer's instructions, with absorbance read at 450 nm on a SpectraMax iD3.
In vivo subcutaneous biocompatibility of MSNs-COOH-BA-GelMA hydrogels
2.2.9.
Sixteen male Fischer 344 rats (10 weeks old; approximately 260 g) were purchased from Envigo RMS (Oxford, MI, USA) and randomly assigned to 7- or 28-day post-operative endpoints (n = 8 per time point) under an approved IACUC protocol (PRO00012062). Under isoflurane anesthesia, dorsal hair was clipped and the skin prepared with povidone-iodine. A 2 cm midline incision was made with a #15 blade, and blunt dissection created four subcutaneous pockets. Polythene tubes (10 mm length, 1 mm internal diameter, 2 mm external diameter) were implanted according to a randomized allocation: empty (control) or filled with 15% GelMA, 10 mg/mL MSNs-COOH-BA, 20 mg/mL MSNs-COOH-BA, or 20 mg/mL MSNs-COOH, yielding six tubes per group at each time point. Incisions were closed with coated Vicryl^®^ (polyglactin-910) sutures (Ethicon Endo-Surgery, Cincinnati, OH). Animals were monitored until full recovery and euthanized at the assigned time point by CO_2_ overdose [36].
Explanted specimens were fixed overnight in 10% neutral-buffered formalin, processed for paraffin embedding, and sectioned at 5 μm. Hematoxylin-eosin staining was used to assess inflammatory infiltration at the tube opening (material-tissue interface). Images were captured at 100× magnification using an ECHO Revolve microscope (BICO). For quantification, eight fields per sample were analyzed. Images were processed in ImageJ (NIH, Bethesda, MD, USA) using a consistent intensity threshold to generate binary masks. Cells were then counted using the Analyze Particles tool, which employed predefined size and circularity filters to include single cells and exclude debris, as described previously [29,37].
Statistical analysis
2.3.
All quantitative results are reported as mean ± standard deviation (SD). Experiments were performed in triplicate unless otherwise noted, and the number of independent samples or replicates (n) for each experiment is indicated in the corresponding figure legends. Statistical analyses were conducted using GraphPad Prism software version 10.6.1 (GraphPad Software, San Diego, CA, USA). Prior to statistical comparisons, data distribution was assessed for normality, and, when applicable, for homogeneity of variances. No data were excluded as outliers unless otherwise stated. Depending on the experimental design, statistical comparisons were performed using one-way or two-way analysis of variance (ANOVA), followed by appropriate post hoc tests (Tukey's or Sidak's) when indicated. A p-value of less than 0.05 was considered statistically significant.
Results
Viability of dental pulp stem cells (DPSCs) in contact with different concentrations of BA
3.1.
Across all time points, the positive control yielded the lowest cell viability. On day 1, treatment with baicalein at a concentration of 20 μM (BA 20 μM) resulted in a statistically significant reduction in cytocompatibility relative to both the negative control (p = 0.0028) and the 1 μM baicalein (BA 1 μM) group (p = 0.0261), corresponding to decreases of 21.08% and 16.14%, respectively. At the subsequent time points (days 3 and 7), no statistically significant differences were observed between any of the BA-concentrations groups and the negative control (p > 0.05), Fig. 1A.
In vitro mineralized nodule formation as an early indicator of odontogenic potential in DPSCs exposed to different BA concentrations
3.2.
In vitro mineralized nodule formation, used here as an early indicator of osteo/odontogenic commitment, was quantified in DPSCs after 14 and 21 days under non-inflammatory (−LPS) and inflammatory (+LPS) conditions. A similar overall pattern emerged at both time points. In the absence of LPS, baicalein (5–20 μM) did not change mineral deposition relative to the control (p > 0.05). Under LPS, mineralization in the control group was markedly reduced, whereas all baicalein doses increased mineral deposition (all p < 0.0001). At day 14, 20 μM yielded less mineral than 5 or 10 μM (p < 0.0001); by day 21, differences among baicalein concentrations were no longer significant (p > 0.05) (Fig. 1B). Representative micrographs align with these results: under −LPS, groups appear similar, while under +LPS, baicalein-treated cultures show visibly greater mineralized nodule formation than the control (Fig. 1C).
Antioxidant and anti-inflammatory actions in RAW 264.7 macrophages by BA
3.3.
Oxidative stress (ROS) decreased in a condition-dependent manner with the administration of baicalein. Without LPS, only 20 μM significantly lowered ROS compared to the control (p = 0.0484) and did not differ from the other baicalein doses (p > 0.05). Under LPS, all concentrations significantly reduced ROS compared with the control (all p < 0.0001), as shown in Fig. 1D. Pro-inflammatory cytokines showed a similar suppression under inflammatory conditions (Fig. 1E-G). For TNF-α and IL-6, every baicalein dose reduced levels relative to control (p < 0.0001), with 10 and 20 μM producing greater IL-6 reductions than 5 μM (p < 0.0001). For IL-1α, only 20 μM significantly decreased cytokine levels compared to the control (p = 0.0005).
Functionalization of mesoporous silica nanospheres (MSNs)
3.4.
FTIR confirmed successful surface functionalization of the MSNs. The Si–O–Si framework band appeared at ~1102 cm^−1^, with O─H features evident as a broad stretch near ~3455 cm^−1^ and an H–O–H bending band around ~1641 cm^−1^. Critically, a distinct C═O stretching band at ~1719 cm^−1^ verified the presence of surface carboxyl groups, consistent with complete carboxylation (Fig. 2A) [11]. Signals corresponding to baicalein were not clearly visible in the MSNs-COOH-BA spectra.
Encapsulation efficiency and release kinetics of baicalein from functionalized MSNs (MSNs-COOH-BA)
3.5.
Baicalein loading into MSNs-COOH was 16.57%, calculated by subtracting the residual drug quantified in the supernatant and wash fractions from the initial input using the standard curve. Accordingly, of the 20 mg of baicalein added, approximately 3.31 mg was incorporated into the nanoparticles. The release of baicalein was highest from the 5 and 10 mg/mL MSNs-COOH-BA formulations, each showing an early burst within the first 3 h. After this peak, concentrations declined and then plateaued, remaining relatively stable through 10 days (Fig. 2B).
Morphology and collagenase-mediated BA release from MSNs-COOH-BA-GelMA hydrogels
3.6.
SEM revealed a nanoporous network in 15% GelMA, with clear incorporation of MSNs-COOH within the matrix. In the MSN-COOH-BA groups, nanosphere clustering was evident (Fig. 3A). Pore size analysis revealed that the 15% GelMA control group exhibited an average pore diameter of 12.0 ± 3.5 μm, whereas the incorporation of MSNs, with or without BA, reduced pore size by approximately 33%, yielding an average of 8.0 ± 2.3 μm. EDS confirmed the expected elemental signals (Si, O, Na, Cl) (Fig. 3B). Hydrogels containing 10 or 20 mg/mL MSNs-COOH-BA released baicalein in distinct 10-day patterns when incubated in PBS and collagenase. Both formulations started releasing the compound within about 6 h, but by 24 h, the 20 mg/mL group had released around four times more than the 10 mg/mL group. The lower-load formulation maintained a relatively steady, sustained release, whereas the higher-load formulation exhibited a pronounced day-3 burst followed by a gradual decline (Fig. 4C).
Compression, swelling, and degradation of MSNs-COOH-BA-GelMA hydrogels
3.7.
Stress-strain responses showed concentration-dependent reinforcement by MSNs-COOH-BA: relative to 15% GelMA, 10 mg/mL increased stress across the strain range, and 20 mg/mL produced the highest response, especially beyond ~10% strain, whereas 20 mg/mL MSNs-COOH (no BA) yielded the lowest stresses (Fig. 3E). Consistent with these curves, Young's modulus increased with BA functionalization: the 20 mg/mL MSNs-COOH-BA group was stiffer than 15% GelMA (p = 0.0012), while 10 mg/mL did not differ from GelMA (p > 0.05); the 10 and 20 mg/mL BA groups were similar to each other (p = 0.9219) and both exceeded the unfunctionalized MSNs-COOH group (p < 0.05) (Fig. 3G). Ultimate compressive strength followed the same pattern: MSNs-COOH (no BA) was significantly weaker than the other groups (p < 0.05), whereas the BA-functionalized groups did not differ from each other (p = 0.2501) or from 15% GelMA (p > 0.05) (Fig. 3H).
Swelling behavior varied by formulation: 20 mg/mL MSNs-COOH (no BA) showed reduced equilibrium swelling versus 15% GelMA (p < 0.0001), and the two BA-functionalized groups also differed (p = 0.0078) (Fig. 3F). Degradation studies indicated that MSNs-containing hydrogels, with or without BA, degraded more gradually over 10 days, whereas 15% GelMA degraded more rapidly, reaching complete degradation within the study period (Fig. 3D).
Cytocompatibility, in vitro mineralized nodule formation as an early indicator of odontogenic potential, and anti-inflammatory activity of MSNs-COOH-BA–GelMA hydrogels
3.8.
Across all time points, the positive control showed the lowest cell viability. On day 1, hydrogels containing MSNs-COOH-BA at 10 and 20 mg/mL reduced viability relative to the negative control (−11.06%, p = 0.0482; −12.96%, p = 0.0318), with no difference between these doses (p > 0.999). By day 3, all GelMA groups were comparable to the negative control (p > 0.05). At day 7, only the 20 mg/mL MSNs-COOH-BA group remained significantly lower than the negative control (−20.86%, p = 0.0005), and it did not differ from the 10 mg/mL group (p = 0.6323) (Fig. 4A).
Mineralized nodule formation by DPSCs, used here as an early in vitro indicator of odontogenic potential, was quantified at 14 and 21 days under non-inflammatory (−LPS) and inflammatory (+LPS) conditions. At day 14 under −LPS, 15% GelMA did not differ from control (p = 0.0565), whereas all MSN-containing groups, with or without baicalein, showed greater mineral deposition than control (p < 0.0001). Under +LPS, every tested group surpassed control (p < 0.0001), with 20 mg/mL MSNs-COOH-BA yielding the highest values, though not different from 20 mg/mL MSNs-COOH (p = 0.6333) or 10 mg/mL MSNs-COOH-BA (p = 0.9383) (Fig. 4B). By day 21 in −LPS, no group differed from control (p > 0.05), but within the baicalein formulations a dose effect emerged: 20 mg/mL > 10 mg/mL (p = 0.0003). In +LPS, only 20 mg/mL MSNs-COOH-BA remained significantly higher than control (p < 0.0001) and all other groups (p < 0.05), indicating enhanced mineralization under inflammatory challenge (Fig. 4C). Representative micrographs taken at 14 and 21 days reflect the quantitative trends (Fig. 4D). Under non-inflammatory conditions, differences in mineralized nodule formation are modest across groups. With LPS stimulation, cultures treated with MSNs-COOH-BA exhibit a clear increase in nodule density compared to the control, consistent with the mineral quantification results reported.
RAW 264.7 macrophages exposed to extracts from hydrogels incubated in collagenase-containing medium for 1, 3, or 7 days were evaluated for TNF-α, IL-6, and IL-1α production (Fig. 4E-G). For TNF-α, levels rose over time in controls; on day 1 only 20 mg/mL MSNs-COOH-BA reduced TNF-α versus 15% GelMA (p = 0.0130), by day 3 both baicalein groups 10 and 20 mg/mL were lower than control (p = 0.0160 and p = 0.0084), and by day 7 only 20 mg/mL remained reduced (p = 0.0176, Fig. 4E). IL-6 showed a similar temporal pattern in controls; no group differed on day 1, whereas at days 3 and 7 both baicalein formulations lowered IL-6 compared with 15% GelMA and MSNs-COOH (p < 0.05, Fig. 4F). IL-1α remained relatively stable across time; a decrease appeared only on day 1 for 20 mg/mL MSNs-COOH-BA versus 15% GelMA (p = 0.0111), with no difference versus 20 mg/mL MSNs-COOH (p = 0.2336) or 10 mg/mL MSNs-COOH-BA (p = 0.5002), and no significant group differences at days 3 or 7 (p > 0.05, Fig. 4G).
In vivo subcutaneous biocompatibility of MSNs-COOH-BA-GelMA hydrogels
3.9.
In a rat subcutaneous implant model using polyethylene tubes, all groups exhibited an acute inflammatory infiltrate at 7 days, which diminished by 28 days (Fig. 5A-B). Mean ± SD inflammatory cell counts at 7 days were 314 ± 105 for empty tubes, 358 ± 130 for 15% GelMA, 461 ± 112 for 20 mg/mL MSNs-COOH, 388 ± 196 for 10 mg/mL MSNs-COOH-BA, and 470 ± 132 for 20 mg/mL MSNs-COOH-BA. By 28 days, counts declined to 217 ± 72, 266 ± 53, 316 ± 52, 238 ± 104, and 296 ± 86, respectively. Within each group, the decrease from 7 to 28 days was statistically significant, whereas no significant differences were detected among groups at a given time point (p > 0.05). The magnitude of reduction ranged from roughly 26% to 39%, consistent with progressive resolution of the inflammatory response across all formulations.
Discussion
This study developed and characterized a photocrosslinkable GelMA hydrogel incorporating baicalein-loaded silica nanospheres and evaluated its cytocompatibility, in vitro mineralization-related responses, and anti-inflammatory activity for vital pulp therapy. In cases of pulp exposure where direct pulp capping is indicated, drug-releasing hydrogels show strong potential to resolve inflammation and may delay or even avoid more invasive endodontic procedures [38,39]. Owing to their soft, porous structure, hydrogels act as depots that enable slow, controlled, and on-demand drug release [40]. Beyond the direct incorporation of bioactive agents, auxiliary carriers such as carboxylated mesoporous silica nanospheres (MSNs-COOH) offer a high surface area, facilitating functionalization and supporting the efficient loading and delivery of therapeutics [41,42].
Before evaluating the different GelMA formulations incorporated with MSNs-COOH-BA, BA was first tested in isolation to establish a reference for its direct cellular effects. Flavonoids such as BA exhibit multiple properties, including antioxidant and pro-differentiation activities, but can also display cytotoxicity even at relatively low concentrations [43,44]. Defining this non-cytotoxic bioactive window was essential not only to ensure experimental safety but also to provide a comparative basis for interpreting the subsequent results, since BA was later incorporated into two distinct systems (MSNs-COOH and GelMA). In such composite systems, physicochemical parameters, such as nanoparticle-matrix interactions, entrapment within the hydrogel, and potential photodegradation during crosslinking, may influence the bioavailability of BA, making it essential to separate the flavonoid's intrinsic activity from the influences of the delivery system.
Building on the dose-range assessment described above, we tested increasing concentrations of BA in DPSCs to define a non-cytotoxic, biologically active window. At 24 h, the highest dose (20 μM) reduced viability by 21% compared to the negative control; nonetheless, viability remained >70% across all groups, meeting the ISO 10993-5 criterion for non-cytotoxicity. By 3 and 7 days, no concentration caused a significant reduction in viability. To better approximate clinical conditions, we then modeled inflammation by challenging DPSCs with LPS. Under this stimulus, every BA dose enhanced mineralized nodule formation at 14 and 21 days, which is commonly interpreted as an early in vitro indicator of osteogenic/odontogenic differentiation potential. This promineralization effect is consistent with BA's reported modulation of differentiation pathways, including activation of Wnt/β-catenin signaling and upregulation of ALP, RUNX2, and OCN in mesenchymal and DPSCs [45]. Although these pathways have been reported in the literature, their specific involvement was not investigated in the present study, as our primary focus was on evaluating in vitro mineralization-related outcomes as early cellular responses. Thus, we next evaluated BA's antioxidant and anti-inflammatory activities. Across the tested concentrations, BA reduced intracellular ROS, attenuating a key driver of pulpal inflammation. BA also downregulated the pro-inflammatory cytokines TNF-α, IL-1α, and IL-6, with more pronounced effects at concentrations of 10–20 μM, indicating a dose-responsive modulation of inflammatory signaling in DPSCs. Mechanistically, these actions are consistent with BA-mediated inhibition of MAPK/NF-κB and JAK/STAT pathways and suppression of ROS-driven stress responses reported in macrophages and other cell types [46,47].
To translate BA's bioactivity into a clinically practical delivery format, we first encapsulated it in carboxylated mesoporous silica nanospheres (MSNs-COOH) to achieve controlled release within the pulp environment. Because 20 μM BA was found to be non-cytotoxic and bioactive (osteogenic and anti-inflammatory), we set a local release target of ≥20 μM. Among the loading conditions tested, only MSNs-COOH loaded at 10 mg/mL reached a concentration of ≥20 μM BA in PBS by day 3 and then sustained ~15 μM over 10 days. These measurements represent cumulative release over time, with periodic sampling and replacement of the medium to maintain constant volume, ensuring an accurate approximation of total BA availability from the nanoparticles. Anticipating that the GelMA network would further slow diffusion, we selected two MSN-COOH-BA loadings (10 and 20 mg/mL) for incorporation into the hydrogel to cover the desired therapeutic window. Considering the measured loading efficiency of 16.57%, 10 mg/mL of MSNs corresponds to ~1.657 mg/mL BA, equivalent to ~6135 μM, substantially higher than the observed cumulative release. This apparent discrepancy can be attributed to several complementary factors: (I) strong interactions between BA and the carboxylated MSNs, which limit diffusion from the particle surface; carboxylated MSNs have been reported to enhance drug retention through covalent bond linkages [36]. Supporting this, FTIR analysis of BA-loaded MSNs did not reveal distinct BA peaks compared to the spectra of unloaded MSNs (Fig. 2A), indicating that the flavonoid is effectively bound or embedded within the nanoparticle matrix. Additionally, partial entrapment of BA-loaded MSNs-COOH within the GelMA network may restrict mass transport; (II) aggregation of nanoparticles within hydrogel pores, particularly at higher loadings, creating local diffusion barriers; and (III) potential photodegradation of BA during GelMA photo-crosslinking, as the blue light (~420 nm) used overlaps with BA's absorption spectrum, potentially reducing the amount of bioactive compound available [48]. Interactions between BA-loaded MSNs and the GelMA matrix, which enhance crosslink density and mechanical integrity, may further retain BA within the reinforced polymer network. Collectively, these mechanisms explain why cumulative release from the hydrogel is modest despite the high nominal loading, while still providing therapeutically relevant concentrations for cellular bioactivity.
Building on the sustained but limited release of BA from MSNs-COOH, we next examined how incorporation into GelMA influenced hydrogel architecture and mechanical performance. SEM revealed that the overall microstructure of GelMA remained unchanged after the addition of MSNs-COOH, regardless of whether BA was present or not; all formulations retained the characteristic wide, interconnected porous network of GelMA. EDS detected sodium (Na) and chloride (Cl), consistent with successful modification of the MSNs via interactions with the crosslinking system [30]. Only one representative EDS spectrum is shown, as no differences were observed among the groups; all formulations, including different concentrations of MSN-COOH-BA, presented the same elemental characteristics. In the MSN-COOH-BA groups at both tested loadings, however, visible particle aggregation was apparent within the pores, even after 24 h of stirring, suggesting that higher nanoparticle content promoted interparticle interactions and clustering, and thus a less uniform dispersion. Finally, to determine whether these compositional changes compromised functional performance, we performed mechanical characterization to assess the structural integrity of the GelMA matrix, an essential property for an injectable pulp-capping material. As a soft, degradable hydrogel, GelMA is not intended to match the compressive strength of Ca(OH)2 or MTA; instead, the mechanical benchmark was set by the unmodified GelMA matrix, with MSNs and BA incorporated to enhance stiffness and structural integrity relative to the unmodified hydrogel while maintaining injectability and bioactive function.
Although localized aggregation of MSNs was observed within the GelMA pores, particularly at higher loadings (20 mg/mL MSN-COOH-BA), its impact on release behavior, mechanical performance, and cellular responses appears to be limited. In hydrogel-nanoparticle composites, drug release is primarily governed by diffusion through the polymeric network rather than by perfect nanoparticle dispersion, such that both isolated particles and aggregates can act as sustained-release reservoirs, reducing the likelihood of localized high- or low-concentration zones. Similarly, from a mechanical perspective, clustered inorganic fillers can function as reinforcing domains, enabling effective stress transfer even in the presence of local heterogeneity. Importantly, because MSNs were dispersed within the hydrogel matrix and evaluated indirectly, cells were exposed only to composite-mediated baicalein release rather than free nanoparticles, which likely accounts for the maintained cytocompatibility, mineralization, and anti-inflammatory effects. During material preparation, the hydrogel precursor containing MSNs was handled and injected prior to photo-crosslinking without any observable difficulties, indicating that the degree of aggregation observed by SEM did not interfere with handling. Together, these findings suggest that the degree of MSN aggregation present in the system did not compromise its functional or biological performance. Nevertheless, future optimization strategies may focus on improving nanoparticle dispersion by refining ultrasonication protocols, modulating mixing conditions, or tailoring MSN surface functionalization to enhance material homogeneity.
To determine whether these microstructural features resulted in measurable mechanical effects, we next analyzed the compressive and swelling behavior of the GelMA incorporated with MSNs-COOH. Even so, mechanical testing showed that this BA-loaded group achieved the highest Young's modulus and ultimate strength, outperforming the same MSN-COOH concentration without BA. A plausible explanation is that BA's multiple hydroxyl groups form hydrogen bonds and possibly covalent linkages with GelMA chains, effectively increasing crosslink density and reinforcing the network despite partial aggregation [49,50]. In contrast, MSNs-COOH without BA lack these stabilizing interactions and can disrupt the polymer architecture, weakening the hydrogel. Consistent with prior work, higher nanoparticle contents (≥ 1.5 mg/mL) in similar systems can decrease compressive modulus by diminishing MSN-matrix bonding [30], with reduced crosslink density proposed as an underlying mechanism [51]. This interpretation is further supported by pore size analysis, which revealed that MSNs-COOH incorporation, with or without BA, decreased the average pore diameter of the GelMA network. Smaller pores are generally associated with tighter crosslinking [52], providing a structural explanation for the improved stiffness and reduced swelling observed in the MSNs-COOH-BA groups. This lower crosslinking likely also accounts for the greater water uptake in the MSN-COOH group, which showed a significantly higher swelling ratio than the control. By comparison, BA-functionalized MSNs appeared to stabilize the matrix, limiting water absorption and yielding swelling comparable to the 15% GelMA control. Degradation data further support this interpretation: the MSN-COOH-BA formulations exhibited reduced mass loss over time relative to other groups, indicating enhanced structural integrity and a degradation profile conducive to maintaining mechanical support while sustaining therapeutic release during the early stages of pulp repair.
Building on the enhanced structural integrity and slower degradation of the MSN-COOH-BA hydrogels described above, we next asked whether they could sustain BA delivery while preserving its bioactivity. Because the material will come into contact with exposed pulp, we reevaluated cytocompatibility with DPSCs during release. In PBS, BA reached ~40 μM by 24 h and continued to increase (>60 μM at later sampling times). Despite this, eluates from both MSN-COOH-BA loadings (10 and 20 mg/mL) reduced viability by <30% at 24 h, within the ISO 10993-5 non-cytotoxic range, and this remained true for the higher loading at subsequent time points. Notably, these both concentrations measured likely overestimate in vivo exposure; in the pulp, fluid turnover and tissue diffusion would dilute the BA, so cells are unlikely to experience the cumulative concentration in full. Functionally, the 20 mg/mL MSN-COOH-BA hydrogel produced the greatest mineralized nodule formation under simulated inflammation, particularly at 21 days, consistent with sustained, therapeutically relevant dosing. LPS was used to simulate an inflammatory microenvironment typical of pulpitis, which impairs osteo/odontogenic differentiation and mineralization in dental pulp cells [28]. This model enables the assessment of BA-loaded hydrogels under inflammatory conditions while also providing a means to investigate pulp cell regenerative mechanisms and screen bioactive compounds [53]. In this context, the hydrogel's ability to preserve or enhance nodule formation underscores its potential to counteract inflammation-induced inhibition and support tissue regeneration.
To better contextualize these effects, it is important to relate the observed biological responses to the concentrations of baicalein released from the GelMA/MSNs-COOH system. Importantly, the biological responses observed for the BA-loaded hydrogels are consistent with the concentration range identified as bioactive in our direct BA exposure experiments, supporting a clear dose-response relationship between BA release and cellular effects. In isolated cultures, BA concentrations between 10 and 20 μM were sufficient to reduce oxidative stress, suppress pro-inflammatory cytokines, and enhance mineralized nodule formation under inflammatory conditions. When delivered via MSNs-COOH and GelMA, cumulative release profiles demonstrated that BA concentrations reached and remained within this effective window over time. Although release values measured in PBS likely overestimate in vivo exposure, the sustained suppression of inflammatory cytokines and enhancement of mineralization indicate that the released BA concentrations were sufficient to drive the observed biological effects.
Within the context of this study, all in vitro mineralization outcomes are interpreted as early cellular indicators of dentinogenic potential rather than as evidence of dentin regeneration. Taken together, the quantitative alignment between baicalein release and the observed cellular responses supports the biological relevance of the GelMA/MSN-COOH-BA system. However, while the enhanced mineralized nodule formation observed under inflammatory conditions supports the bioactivity of released baicalein, it is important to note that in vitro calcium deposition, as assessed by Alizarin Red S staining, represents an early indicator of osteogenic/odontogenic commitment rather than a direct surrogate for reparative dentin bridge formation in vivo. Dentinogenesis involves additional biological features, including odontoblast polarization, tubular dentin organization, and spatially regulated matrix deposition, which cannot be fully recapitulated in standard two-dimensional cell culture models [54]. Therefore, the present findings should be interpreted as supportive evidence of regenerative potential, warranting further validation in orthotopic pulp exposure models and through the evaluation of more specific odontogenic markers, such as dentin sialophosphoprotein (DSPP) [55].
After establishing sustained release and DPSC compatibility, we evaluated whether BA released from the 15% GelMA/MSN-COOH-BA hydrogels would attenuate macrophage inflammation. In an LPS-challenged RAW 264.7 model, both MSN-COOH-BA loadings reduced TNF-α and IL-1α. Because the hydrogel extracts were applied after LPS stimulation (post-treatment), absolute cytokine levels rose over time as expected for the LPS response. However, when expressed as percent inhibition relative to the LPS control at each time point, both concentrations produced consistent suppression, indicating that released BA retained anti-inflammatory activity during/after hydrogel degradation. Clinically, the 24-h effects are most noticeable, as early dampening of inflammation at the material-tissue interface is critical. To complement the in vitro anti-inflammatory findings, we evaluated in vivo biocompatibility in a rat subcutaneous implant model using polyethylene tubes. All groups exhibited an acute inflammatory infiltrate at 7 days, which subsided by 28 days. Within each formulation, the decrease from 7 to 28 days was statistically significant, whereas no differences were detected among groups at a given time point. The 26–39% reduction across groups indicates a progressive resolution of the tissue response and supports the notion that BA-functionalized MSNs in GelMA are at least as well tolerated as GelMA alone, aligning with the macrophage suppression observed in vitro.
The subcutaneous implantation model was intentionally selected at this stage as an initial in vivo screening platform to assess host tissue compatibility, inflammatory response, and material stability under well-controlled conditions. However, we acknowledge that this ectopic model does not recapitulate key features of the pulp microenvironment or the presence of specialized odontogenic cell populations. As a result, the localized anti-inflammatory effects of BA, which are expected to be most relevant in the early post-implantation phase of the pulp, may not be fully captured in this model, potentially explaining the similar tissue responses observed across groups. Consequently, outcomes central to vital pulp therapy, particularly dentin bridge formation, odontoblastlike cell organization, and functional pulp preservation, cannot be evaluated in this setting and should not be inferred from subcutaneous data alone.
Accordingly, the present findings should be interpreted as evidence of favorable biocompatibility and immunomodulatory behavior rather than definitive proof of regenerative efficacy in the pulp. Future studies should focus on optimizing nanoparticle dispersion within GelMA and fine-tuning BA dosing to enhance therapeutic efficacy. Importantly, orthotopic tooth-level pulp exposure models are required as the next translational step to validate dentin bridge formation, maintenance of pulp vitality, and restoration of functional tissue under clinically relevant conditions. Such models would also allow exploration of the cellular and molecular mechanisms underlying BA-mediated osteogenic and anti-inflammatory effects, providing a deeper understanding of the hydrogel's regenerative potential. Additionally, the development of alternative scaffold systems that avoid blue-light photo-crosslinking could reduce potential photodegradation of BA and improve its bioavailability, further enhancing the effectiveness of this delivery platform.
Conclusion
Overall, our results show that the GelMA hydrogel system loaded with MSNs-COOH-BA satisfies key preclinical criteria for a vital pulp-capping material. The system provided sustained BA release within a therapeutically relevant window while preserving bioactivity, attenuating ROS and pro-inflammatory cytokine production, and promoting in vitro mineralized nodule formation as an early indicator of odontogenic potential, even under LPS challenge. BA functionalization enhanced mechanical performance and mitigated swelling/degradation relative to MSNs without BA, indicating a more stable network suitable for injectable use. In vivo, the subcutaneous implantation model provided preliminary evidence of fundamental biosafety and tissue compatibility of the GelMA/MSNs-COOH-BA system, with tissue responses comparable to those elicited by GelMA alone. Importantly, this ectopic model was not intended to assess localized anti-inflammatory efficacy or regenerative outcomes within the dental pulp environment. Accordingly, the present in vivo findings should be interpreted as confirming material biocompatibility rather than as evidence of in vivo anti-inflammatory or pro-regenerative superiority. Collectively, these findings support the GelMA/MSNs-COOH-BA system as a multifunctional, drug-eluting hydrogel that integrates cytocompatibility and mechanical stability with in vitro anti-inflammatory activity and early pro-odontogenic cellular responses, while the validation of its anti-inflammatory and regenerative functions in vivo must be addressed in future orthotopic pulp exposure models to determine its capacity to support reparative dentin formation and functional pulp preservation under clinically relevant conditions.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Conrads G, About I, Pathophysiology of dental caries, Monogr. Oral Sci 27 (2018) 1–10, 10.1159/000487826.29794423 · doi ↗ · pubmed ↗
- 2Tong HJ, Seremidi K, Stratigaki E, Kloukos D, Duggal M, Gizani S, Deep dentine caries management of immature permanent posterior teeth with vital pulp: A systematic review and meta-analysis, J. Dent 124 (2022) 104214, 10.1016/j.jdent.2022.104214.35793760 · doi ↗ · pubmed ↗
- 3Duncan HF, Present status and future directions-vital pulp treatment and pulp preservation strategies, Int. Endod. J 55 Suppl 3(Suppl 3) (2022) 497–511, 10.1111/iej.13688.35080024 PMC 9306596 · doi ↗ · pubmed ↗
- 4Sriudomdech P, Santiwong B, Linsuwanont P, Outcomes of vital pulp treatment in permanent teeth with carious pulp exposure with signs and symptoms of irreversible pulpitis, Clin. Oral Investig 28 (10) (2024) 551, 10.1007/s 00784-024-05923-9.39320508 · doi ↗ · pubmed ↗
- 5European Society of Endodontology developed b, Duncan HF, Galler KM, Tomson PL, Simon S, El-Karim I, , European Society of Endodontology position statement: management of deep caries and the exposed pulp, Int. Endod. J 52 (7) (2019) 923–934, 10.1111/iej.13080.30664240 · doi ↗ · pubmed ↗
- 6Minic S, Florimond M, Sadoine J, Valot-Salengro A, Chaussain C, Renard E, , Evaluation of pulp repair after biodentine(TM) full pulpotomy in a rat molar model of pulpitis, Biomedicines 9 (7) (2021), 10.3390/biomedicines 9070784.PMC 830133134356848 · doi ↗ · pubmed ↗
- 7Cooper PR, Takahashi Y, Graham LW, Simon S, Imazato S, Smith AJ, Inflammation-regeneration interplay in the dentine-pulp complex, J. Dent 38 (9) (2010) 687–697, 10.1016/j.jdent.2010.05.016.20580768 · doi ↗ · pubmed ↗
- 8Smith AJ, Vitality of the dentin-pulp complex in health and disease: growth factors as key mediators, J. Dent. Educ 67 (6) (2003) 678–689, 10.1002/j.0022-0337.2003.67.6.tb 03668.x.12856968 · doi ↗ · pubmed ↗