Immunomodulatory and dentinogenic potential of surface-engineered hesperetin-functionalized composite scaffolds for pulp-dentin regeneration

Abstract

Maintaining dental pulp vitality and promoting dentin regeneration in inflamed environments requires biomaterials that not only support cell differentiation but also actively regulate immune signaling. Here, we present a bioinstructive, immunomodulatory nanofibrous platform composed of polycaprolactone and nano-hydroxyapatite (PCL/nHA), surface-engineered through alkaline hydrolysis (H-) to enhance hydrophilicity and loaded with the flavonoid hesperetin (+HT) for sustained bioactive release. Physicochemical analyses confirmed improved surface wettability, calcium ion release, and diffusion-driven hesperetin delivery, all without compromising the scaffold's structure. In vitro, the scaffolds facilitated the migration, proliferation, and odontogenic differentiation of human dental pulp cells, leading to a significant increase in mineralized matrix formation. Using lipopolysaccharide-stimulated macrophages and a 3D inflammatory pulp capping model, the scaffolds modulated immune responses by downregulating pro-inflammatory cytokines and upregulating pro-resolving mediators, while also promoting dentinogenic and angiogenic gene expression. In vivo, subcutaneous implantation demonstrated favorable host integration, reduced inflammatory infiltration, early M2-skewed macrophage polarization, and organized extracellular matrix formation. Overall, this study introduces a surface-engineered, flavonoid-releasing scaffold as a bioinstructive platform that combines immunomodulation and odontoblastic signaling, providing mechanistic insight for designing next-generation materials aimed at regenerating the dentin-pulp interface under inflammatory conditions.

Graphical abstract

1. Introduction

Maintaining the vitality of exposed pulp and promoting dentin regeneration in teeth affected by caries remains a significant challenge. The most influential factors are the intraoperative assessment of pulp status, the material used, and the patient's age [[1], [2], [3]]. In cases of pulp exposure, low levels of inflammatory mediators are essential for supporting regenerative processes, stimulating reactionary dentinogenesis by primary odontoblasts and reparative dentinogenesis by progenitor cells; however, sustained bacterial infection can disrupt this balance, shifting the response from regeneration to degeneration, leading to cytotoxicity, necrosis, and impaired mineralization [[4], [5], [6], [7]]. Instead of encouraging a pro-resolving inflammatory response, materials currently applied to exposed pulp often release high concentrations of hydroxyl ions, which increase oxidative stress, exacerbate inflammation, and induce a superficial necrotic zone prone to secondary inflammation and bacterial recontamination [[8], [9], [10]]. Therefore, new materials must be able to orchestrate controlled immune-cellular interactions that favor resolution of inflammation, thereby reducing the risk of failure of vital pulp therapies (VPT), including direct pulp capping and pulpotomy [[11], [12], [13]].

Recent advances have increasingly focused on integrating bioactive agents for guiding cellular responses toward regeneration. Among them, flavonoids have emerged as a versatile class of natural molecules with multifunctional biological effects [[14], [15], [16], [17]]. Their potential in dental tissue engineering has been notably emphasized due to their ability to modulate key cellular responses for pulp healing under chronic exposure to inflammatory stimuli [14,15,18]. However, depending on their intrinsic properties and environmental conditions (e.g., pH and temperature), flavonoids can undergo extensive degradation and cellular metabolism in the body [19,20]. Therefore, a local delivery system for these molecules would be preferable.

Among these flavonoids, hesperetin (HT) stands out for its effective bioactivity at low concentrations (10-15 μM), as it modulates inflammation and promotes biomineralization [14,[21], [22], [23], [24], [25]]. HT has shown remarkable stability, with over 70% remaining intact in an aqueous medium after 15 h at 37 °C [26]. Even its metabolites exhibit anti-inflammatory properties when incubated with macrophages [26]. Due to its hydrophobic nature, recent strategies have focused on delivery systems that enhance the solubility of flavonoids while maintaining or amplifying their bioactivity. Controlled-release nanostructured platforms have shown promise by enabling the gradual delivery of flavonoids, improving stability and cytocompatibility, and promoting osteogenic outcomes [1,15,21,27]. However, the ability of these strategies to actively shape the inflammatory milieu and influence the immune-regenerative crosstalk in the dentin-pulp interface remains unexplored.

In the context of VPT, the design and composition of biomaterials are critical determinants of clinical success, as they directly influence the cellular phenotype and local immune response, which are central to tissue regeneration [11]. Fibrous scaffolds represent a particularly advantageous platform due to their high surface area-to-volume ratio and tunable microarchitecture, whose interconnected porous network enables simultaneous functions, including: (1) acting as a physical barrier to overlying restorative materials and bacterial infiltration; (2) serving as a temporary extracellular matrix that allows tissue-resident progenitor cells to proliferate, migrate, and differentiate to form new tissue; and (3) functioning as a reservoir for the controlled delivery of bioactive molecules [12,15,[28], [29], [30]]. Therefore, these biomaterials are promising candidates for stimulating the regeneration of the dentin-pulp interface. Recent studies have shown that polycaprolactone (PCL)-based scaffolds containing nano-hydroxyapatite (nHA) or calcium hydroxide (CH) support pulp cell proliferation, enhance adhesion and spreading, upregulate genes associated with dentinogenesis, and promote significant mineralized matrix deposition [15,30,31]. Building on these findings, functionalizing fibrous scaffolds for HT release may provide a cost-effective, commercially viable, and biologically multifunctional approach capable of not only enhancing dentinogenic differentiation but also modulating the inflammatory microenvironment in compromised pulp tissues.

In this study, we hypothesize that HT-functionalized PCL/nHA fibrous scaffolds can modulate inflammation and dentinogenic signaling in pulp cells. We investigated hesperetin (HT) adsorption onto composite PCL/nHA scaffolds and assessed how alkaline hydrolysis modifies the surface chemistry. This process would increase the scaffolds' hydrophilicity while modulating flavonoid release profile and cellular interactions [32,33]. Rather than comparing current pulp-capping procedures, our approach aims to elucidate the mechanistic interplay between surface chemistry, HT release, and immune-odontoblastic signaling, providing a conceptual framework for immunomodulatory scaffolds targeting dentin-pulp complex regeneration under inflammatory conditions. Our goal is to effectively deliver HT to enhance the regenerative and pro-resolving responses of dental pulp cells in inflamed microenvironments. This strategy advances the development of next-generation immuno-informed biomaterials for dentin tissue engineering-based VPT by integrating surface modification with controlled delivery of bioactive molecules.

2. Experimental section

2.1. Scaffold development and experimental formulations

PCL/nHA composite fibrous scaffolds were prepared as previously described [30]. Briefly, a chloroform/dimethylformamide solution (4:1) containing 10% poly(ε-caprolactone) (PCL) (80,000 g/mol, Sigma-Aldrich, St. Louis, MO, USA) and 20% (m/m) nano-hydroxyapatite (nHA) (<200 nm, Sigma-Aldrich) was stirred for 24 h. Then, the solution was electrospun using an 18-gauge industrial dispensing needle at a flow rate of 1 mL/h, an applied voltage of 12-13 kV, and a distance of 15 cm for 1 h. After evaporating the residual solvent in a desiccator for 72 h, round samples (13 mm diameter) were prepared for physicochemical and biological analyses.

All scaffolds were sterilized in UV light for 1 h each side before HT adsorption. Then, they were rinsed with 70% ethanol for 30 min and sterile PBS (phosphate-buffered saline, pH 7.4). Two conditions were set for HT adsorption: hydrolyzed and non-hydrolyzed scaffolds. For the hydrolyzed samples, the scaffolds were soaked in 500 μL of 5 M NaOH solution at room temperature for 30 min [33]. The samples were then washed three times with PBS to completely remove the NaOH. This was confirmed by measuring the pH, which returned to a neutral range. Meanwhile, the non-hydrolyzed scaffolds were kept in PBS and rinsed as well. HT adsorption was performed by immersing a sample in a 200 μM HT solution (in PBS), prepared from a 10 mM HT stock solution (in DMSO), for 15 h at 37 °C, protected from light. The HT used was hesperetin, ≥95% (W431300, LOT#SHBN0365, Sigma-Aldrich). This protocol was determined in a previous pilot study on drug release, aiming to provide a bioactive release of HT at approximately 10 μM within 24 h [14] (Fig. S1). Meanwhile, non-hydrolyzed and/or HT-free scaffolds were maintained under the conditions in pure PBS. After the adsorption protocol, the scaffolds were washed with PBS, and the samples were immediately used for experiments. Non-hydrolyzed scaffolds without HT adsorption were used as a control. The scaffold functionalization protocol is illustrated in Fig. 1, and Table 1 lists the formulated groups.

Fig. 1.

Schematic representation of the scaffold functionalization protocol. Polycaprolactone/nano-hydroxyapatite (PCL/nHA) scaffolds were subjected (or not) to alkaline hydrolysis (H) for 30 min before hesperetin (HT) adsorption.

Table 1.

Experimental groups according to scaffold formulation.

Experimental group	Hydrolysis (H-)	Hesperetin Adsorption (+HT)
PCL/nHA (Control)	Non-hydrolyzed	Non-adsorbed
PCL/nHA + HT	Non-hydrolyzed	Adsorbed with hesperetin
H-PCL/nHA	NaOH-Hydrolyzed	Non-adsorbed
H-PCL/nHA + HT	NaOH-Hydrolyzed	Adsorbed with hesperetin

2.2. Physicochemical characterization

2.2.1. Morphology, composition, and water contact angle

The morphology of the scaffolds was evaluated using scanning electron microscopy (SEM), and their composition was analyzed using energy-dispersive spectroscopy (EDS). Scanning electron microscopy (SEM) analyses were performed at 2000 × and 20,000 × magnification. After the samples were placed in stubs, dried, and sputter-coated with Au-Pd (80/20), the analyses were conducted using a field-emission scanning electron microscope (MIRA3, Tescan, Brno-Kohoutovice, Czech Republic). The diameter of 150 fibers was measured from three SEM images per formulation. The composition of the scaffold was also evaluated using Fourier Transform Infrared Spectroscopy (FTIR) coupled with a diamond attenuated total reflectance (ATR) accessory (VERTEX 70, Bruker, Germany). Spectra were collected from 64 scans, with a resolution of 4 cm⁻¹, in the wavelength range of 4000 to 400 cm⁻¹. To measure the hydrophilicity of the formulations, the contact angle between the fibrous scaffold surfaces and 1 μL of deionized water was analyzed after 2 s using a goniometer (SCA20, DataPhysics Instruments GmbH, Filderstadt, BW, Germany).

2.2.2. Hesperetin adsorption efficiency and release profile

The hesperetin adsorption efficiency and in vitro release profile were measured using UV/Vis spectroscopy and a calibration curve at 286 nm (Synergy H1, BioTek, Winooski, VT, USA) with known concentrations. After following the adsorption protocol, the absorbance of the PBS solution, which initially contained 200 μM HT, was measured. Adsorption efficiency was determined by the ratio of the actual to initial HT concentration in the solution. Then, the samples were immersed in 1 mL of fresh PBS and stored at 37 °C for up to seven days. At various time points during storage, 100 μL of the solution in contact with the samples was removed to measure the absorbance and calculate the cumulative release (μM). Then, another 100 μL of PBS was added to maintain a constant volume. To investigate release kinetics, the data were fitted to several mathematical models, and model quality was assessed by the coefficient of determination and residual analysis.

2.2.3. Calcium release and pH modulation

The scaffolds were stored in 300 μL of PBS, and the solution was changed daily for up to 50 days. At various time points, 100 μL samples were collected and subjected to an endpoint reaction with o-cresolphthalein complexone (Calcium Liquiform, Labtest Diagnostics, Lagoa Santa, MG, Brazil). The absorbance was then measured at 570 nm using a Synergy H1 [30]. Cumulative soluble calcium release was quantified using a calibration curve with known calcium concentrations. Additionally, pH was monitored at specific time points using a Hanna HI 2221 (Hanna Instruments, Woonsocket, RI, USA).

2.2.4. In vitro degradation

The scaffolds were dried at 37 °C for 24 h. The dry mass of each sample was measured using a high-precision scale (XS105 Dual Range, Mettler Toledo, Toledo, OH, USA) to determine the initial mass (M₀). After storing the samples in PBS, they were reweighed (M_x) as before for up to 120 days. The percentage change in mass during each storage period was then determined using the following equation: % dry mass change = 100 − (M_x × 100/M₀).

2.3. Scaffolds and dental pulp cells interaction

2.3.1. Human dental pulp cell culture

An established primary culture of human dental pulp cells (hDPCs) was isolated from healthy third molars (n = 4) by enzymatic digestion and used for the experiments (Research Ethics Committee approval #55269822.7.0000.5416). The hDPCs expressed positive mesenchymal stem cell markers (a minimum of 76.2% of the cell population) [4]. Detailed methodology for isolation and characterization is described elsewhere [4]. Cells were cultured in α-MEM containing 100 IU/mL penicillin, 100 μg/mL streptomycin, 2 mmol/L glutamine, and 0.25 μg/mL amphotericin B (all from GIBCO, Invitrogen, Carlsbad, CA, USA). The complete medium was also supplemented with 10% fetal bovine serum (FBS; GIBCO). Cell culture was carried out at 37 °C and 5% CO₂ in a humidified environment. Cells that achieved 80% confluence between passages 3 and 6 were used in all subsequent experiments.

2.3.2. Dental pulp cells migration and fibronectin expression

The analysis of cell migration and fibronectin expression was conducted through indirect contact with the scaffolds. In summary, hDPCs (3 × 10⁴) were seeded inside inserts with a fluorescence-blocking porous membrane (8 μm pore diameter, FluoroBlok, Corning Corporation, Corning, NY, USA). The insert-cells sets were positioned in wells of a 24-well plate containing the scaffolds, which were held at the bottom by a stainless-steel ring in complete α-MEM for 24 h. Following this, the cells were fixed, permeabilized, and blocked as previously described [30]. The cells were then incubated with an anti-fibronectin primary monoclonal antibody (1:100 dilution, Santa Cruz Biotechnology, Santa Cruz, CA, USA) at 4 °C for 12 h, followed by an additional hour-long incubation with FTIC-conjugated secondary antibody (1:100 dilution, Jackson Immunoresearch Laboratories, West Grove, Pennsylvania, PA, USA). Subsequently, the cells were incubated with a red fluorescent probe for actin (1:20, ActinRed 555 ReadyProbes reagent, Invitrogen), and cell nuclei were stained with Hoescht 33342 (1:5000, Invitrogen). The bottom surface of the inserts was analyzed under a fluorescence microscope at 10 × magnification. Images from multiple fields were acquired to quantify the number of migrated cells, the percentage of positive areas for actin and fibronectin, and the relative fibronectin expression, calculated as the fibronectin/actin ratio (QuPath software, version 0.5.1 for MacOS, [34]). Pores and artifacts were excluded from the analyses.

2.3.3. Direct interaction between scaffolds and human dental pulp cells

The seeding of hDPCs (5 × 10⁴) was conducted in a single drop (20 μL) onto a standardized area of the scaffolds, which were placed in 24-well plates [30]. Following 30 min of initial adhesion, 1 mL of complete α-MEM was added to each sample, and this solution was renewed every 2 days. Scaffold-cell sets were incubated at 37 °C and 5% CO₂ in a humidified environment until the following analyses were performed. The complete α-MEM was supplemented with 5 mM β-glycerophosphate and 50 μg/mL ascorbic acid for odontogenic differentiation experiments [30].

2.3.4. Cell viability, adhesion, and proliferation

These analyses were conducted at 1, 7, and 14 days. Cell viability was analyzed by cell metabolism using the alamarBlue reagent protocol (Invitrogen), considering paired samples [30]. The mean fluorescence obtained for cells cultured on the PCL/nHA scaffold (control) on day 1 was designated as 100%. Cell viability and their adhesion ability were also evaluated using the Live/Dead staining protocol [30]. At the same time points (1, 7, and 14 days), scaffold-cell sets were evaluated under direct fluorescence at 10 × magnification to verify the presence of viable cells (stained green) or cells in the process of death (stained red). Cell proliferation was inferred by comparing cell viability and population on the scaffolds over time.

2.3.5. Gene expression of odontogenic differentiation markers

Cells were cultivated on scaffolds for 14 days. The total RNA was extracted, purified, and treated with DNase using the RNAqueous Micro Total RNA Isolation Kit (Ambion, Life Technologies, Carlsbad, CA, USA). The reverse transcription was performed with 1 μg of RNA using random primers, and quantitative polymerase chain reactions (qPCR) were performed using custom assays with primers and hydrolysis probes (TaqMan Gene Expression Assays; Applied Biosystems, Life Technologies) for amplification of COL1A1 (ID: Hs01076756_g1), ALPL (ID: Hs01029144_m1), DSPP (ID: Hs00171962_m1), DMP1 (ID: Hs01009391_g1) genes. The comparative Cq method was used to calculate relative gene expression, using GAPDH (ID: Hs02786624_g1) and ACTB (ID: Hs01060665_g1) as reference genes, and the PCL/nHA scaffold as the control [30].

2.3.6. Dentinogenic proteins expression

Cells were cultured on scaffolds for 21 days. After this period, they were fixed with 4% formaldehyde, permeabilized with 0.1% Triton X-100 (Sigma-Aldrich), and blocked with 5% bovine serum albumin (BSA) for 30 min. Subsequently, the samples were incubated at 4 °C for 12 h with Alexa Fluor 488-conjugated anti-DSPP antibody (1:100, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and Alexa Fluor 546-conjugated anti-DMP-1 antibody (1:100, Santa Cruz Biotechnology). After two washes in PBS, nuclei were counterstained with Hoechst 33342 (1:5000, Invitrogen). Images from multiple fields were acquired, and the percentage of positive areas for DSPP and DMP-1 was quantified using QuPath software (version 0.5.1 for MacOS; [34]). The positivity threshold was established based on negative control groups, with meticulous exclusion of imaging artifacts from the analysis.

2.3.7. Mineralized matrix formation and soluble calcium content

Mineralized matrix formation and soluble calcium deposition were assessed after 21 days of culture. To assess mineralized matrix formation, scaffold-cell constructs were fixed in 70% ethanol and stained with Alizarin red S solution, as previously described [30]. Five washes with deionized water removed excess dye, and mineral deposition was visualized under a stereomicroscope. The retained stain was solubilized with cetylpyridinium chloride solution to quantify the calcium-bound matrix, and the absorbance was measured at 570 nm (Synergy H1). For soluble calcium quantification, parallel samples were incubated with 0.6 N HCl for 4 h at 4 °C to extract intracellular or loosely bound calcium. The supernatants were then subjected to colorimetric analysis using an endpoint reaction with o-cresolphthalein complexone, as previously described. Cell-free scaffolds were processed in parallel and used to subtract baseline background levels for each group. The PCL/nHA group (control) was considered the baseline reference (100%) for mineralized matrix formation, and calcium content was quantified using a standard curve.

2.4. Indirect interaction between scaffolds and inflammatory cells

RAW 264.7 murine macrophages (TIB-71, ATCC, Rockville, MD, USA) were cultured in high-glucose Dulbecco's modified Eagle's medium (DMEM; GIBCO) supplemented with 10% FBS (GIBCO) and 1% penicillin/streptomycin (GIBCO) at 37 °C in 5% CO₂. Cells that had reached 80% confluence between passages 17 and 18 were used for these experiments. Cells were seeded in 24-well plates (1 × 10⁵) for 24 h. Thereafter, cells were stimulated (positive control) or not (negative control) with lipopolysaccharides (LPS) from Escherichia coli (E. coli; 100 ng/mL) to modulate an inflammatory response. After 3 h, the supernatant/extracts (1 mL) from the scaffolds, which had been kept in FBS-free DMEM for 24 h, were collected and applied to the macrophages for another 3 h. The immediate indirect immunomodulatory effect of the scaffold component release was analyzed by gene expression (RT-qPCR) of Tnf (ID: Mm00443258_m1), Il1b (ID: Mm00434228_m1), and Nos2 (ID: Mm00440502_m1), as previously described, using Gapdh (ID: Mm99999915_g1) as a reference. The fold change of gene expression was calculated based on the negative control group, in which the macrophages were maintained in DMEM without LPS.

After 24 h of exposure to the extracts or controls, cell viability was assessed using alamarBlue as previously described. The culture supernatants were collected to measure the synthesis of Tnf-α and Il-1β using ELISA, following the manufacturer's protocol (Cat. #432604 and #430904, Biolegend, San Diego, CA, USA). In addition, reactive nitrogen species production (measured by nitrite levels) was determined by mixing 50 μL of culture supernatant with a Griess reagent system (Promega, Madison, WI, USA). The resulting reaction was read at 540 nm (Synergy H1), and the nitrite concentration was determined using a calibration curve.

2.5. In vitro pulp capping under a simulated inflammatory microenvironment

An established three-dimensional model of inflamed dental pulp tissue was used to evaluate the modulatory effects of scaffolds on the phenotype of hDPCs in a simulated direct pulp capping scenario [15]. Dentin discs (1 mm thick, 8 mm diameter) were obtained from healthy third molars (Research Ethics Committee approval #55269822.7.0000.5416) and centrally perforated (2 mm) to simulate pulp exposure. Discs were mounted in artificial pulp chambers (APCs) and treated with 0.5 M EDTA to remove the smear layer before sterilization with ethylene oxide.

To simulate dental pulp tissue in vitro, three-dimensional cultures of hDPCs were used. Briefly, a collagen type I solution (from rat tail, 4.39 mg/mL, Corning, Discovery Labware Inc., Bedford, MA, USA) was mixed with 10x α-MEM (5:1; GIBCO, Invitrogen), followed by pH neutralization. Then, aliquots of 200 μL were transferred to 48-well plates, followed by polymerization at 37 °C for 30 min. The 3D matrices were then seeded with 1 × 10⁵ hDPCs (passage #5) and incubated for 48 h at 37 °C, facilitating the uniform distribution of cells. To create a pre-existing inflammatory microenvironment, cells were seeded in complete α-MEM supplemented with 10 μg/mL E. coli LPS (Sigma-Aldrich) [4,15]. During the 48-h incubation period, the LPS-containing medium was refreshed after 24 and 48 h, for a total of three exposures [4,15]. Cells and 3D cultures cultivated in LPS-free α-MEM were used as negative controls (NC).

The 3D cultures were placed in the pulp side of the APCs, and the scaffolds (3 mm diameter) were placed in the central perforation (in contact with the 3D culture) to mimic a clinical situation of direct pulp capping. The cells were then cultured in 1 mL of LPS-free complete α-MEM. Following 3 h and 14 days of direct contact between the scaffolds and the 3D cultures, an evaluation of inflammatory and dentin-pulp regeneration gene expression was conducted, as previously described. For the inflammatory response, the following sequences were amplified: IL8 (ID: Hs00174103_m1), IL6 (ID: Hs00174131_m1), IL1B (ID: Hs01555410_m1), TNF (ID: Hs00174128_m1), COX2 (ID: Hs00153133_m1), MMP9 (ID: Hs00957562_m1), IL10 (ID: Hs00961622_m1), and HMOX1 (ID: Hs01110250_m1). For the dentin-pulp regeneration response, the following sequences were amplified: COL1A1 (ID: Hs01076756_g1), ALPL (ID: Hs01029144_m1), DSPP (ID: Hs00171962_m1), DMP1 (ID: Hs01009391_g1), OPN (ID: Hs00959010_m1), OCN (ID: Hs01587814_g1), TGFB1 (ID: Hs00998133_m1), and VEGFA (ID: Hs00900055_m1). The H-PCL/nHA + HT formulation was compared to the control scaffold (PCL/nHA). Untreated and LPS-treated non-capped cultures served as negative control (NC) and positive control (PC), respectively. The relative gene expression was calculated using the comparative Cq method, with GAPDH (ID: Hs02786624_g1) and ACTB (ID: Hs01060665_g1) as reference genes, and normalized to the negative control.

2.6. In vivo subcutaneous biocompatibility, immunomodulation, and mineralization potential

Under an approved IACUC protocol (PRO00012062), twelve male Fischer 344 rats (12 weeks old; 300 g; Envigo, Oxford, MI, USA) were randomized to retrieval at 1, 4, or 8 weeks (n = 4 per time point). Anesthesia was induced and maintained with isoflurane. After dorsal hair clipping and skin preparation with povidone-iodine, a 2 cm midline dorsal incision was made with a No. 15 blade, and blunt dissection created four subcutaneous pockets. Sterile 10 × 10 mm scaffolds, PCL/nHA (control) and H-PCL/nHA + HT, were implanted, two from each group per animal in separate pockets (four total implants per animal), yielding eight scaffolds per group at each time point. Wounds were closed with coated polyglactin-910 sutures (Vicryl®, Ethicon Endo-Surgery, Cincinnati, OH). Animals were observed until fully ambulatory and monitored thereafter. At the designated endpoints, rats were humanely euthanized by CO₂ inhalation and implants retrieved for analysis [35,36].

Then, samples were fixed overnight in 10% neutral-buffered formalin, processed for paraffin embedding, and sectioned at a thickness of 5 μm. Hematoxylin-eosin (H&E) staining was used to assess the inflammatory cell response to the scaffold. To quantify inflammatory cells, eight fields per sample were analyzed. Images were processed in ImageJ (U.S. National Institutes of Health, Bethesda, Maryland, USA) by applying a consistent intensity threshold to generate binary masks. Cells were then counted using the Analyze Particles function, with predefined size and circularity filters applied to include single cells and exclude debris, as previously described [37,38].

For immunofluorescence, sections were incubated overnight at 4 °C with primary antibodies diluted 1:100: anti-iNOS (Abcam, ab283655) together with anti-CD163 (Abcam, ab182422) to examine scaffold-driven macrophage polarization at 1 and 4 weeks, and anti-osteocalcin (Proteintech, #23418-1-AP) and anti-osteopontin (Abcam, ab216402) to evaluate mineralization potential at 8 weeks. Detection was performed using the secondary antibody goat anti-rabbit IgG-Alexa Fluor 488 (Abcam, ab150077; 1:100, 2 h, room temperature), and nuclei were counterstained with VECTASHIELD mounting medium containing DAPI. Images were captured at 10 × on an ECHO Revolve microscope (BICO). Quantification was performed in ImageJ by analyzing eight randomly selected, non-overlapping fields per group, splitting the fluorescence channels, and reporting the percentage of immunolabeling area that was positive for the target antibody [39,40].

2.7. Statistical analysis

The number of replicates for each protocol is indicated in the corresponding figure legends in the Results section and was determined to ensure a statistical power of at least 80%, based on two or more independent experiments. Data were assessed for normality and homoscedasticity to inform the selection of the most appropriate statistical analysis. Depending on the dataset, the following statistical tests were applied: Kruskal-Wallis followed by Dunn's test, one-way ANOVA followed by Tukey's test, Welch's ANOVA followed by the Games-Howell test, two-way ANOVA followed by Sidak's test, or unpaired t-test. The data on the release of hesperetin, calcium, pH modulation, and in vitro degradation were analyzed using 95% confidence intervals. All statistical analyses were performed using GraphPad Prism version 10.3.1 for Mac (GraphPad Software, San Diego, CA, USA), adopting a significance level of 5%.

3. Results

3.1. Morphological and physicochemical characterization of scaffolds

SEM analysis (Fig. 2A) revealed that all groups exhibited fibers with random orientation. The adsorption of HT did not alter fiber morphology. The higher frequency of fiber diameters around 500 to 600 nm for non-hydrolyzed scaffolds, compared to 400 nm after alkaline hydrolysis (Fig. 2B), confirms this nanoscale morphological change. No significant alterations were observed following HT adsorption, irrespective of hydrolysis treatment. The hydrolysis protocol preserved fiber integrity, resulting in slightly rougher surfaces, indicating surface etching followed by enhanced precipitation of PBS salts, as confirmed by elemental analysis. EDS spectra also confirmed the presence of calcium (Ca) and phosphorus (P) from nHA and PBS salts, along with chlorine (Cl) and sodium (Na) from the latter, with no major compositional differences observed after treatments (Fig. 2C). Fiber diameter measurements (Fig. 2D) demonstrated a significant decrease in diameter following hydrolysis (H-PCL/nHA), suggesting an etching effect on the fiber surface.

Fig. 2.

Morphological and chemical characterization of the experimental scaffolds. (A) Scanning electron microscopy (SEM) images of the scaffold surfaces at 2000 × and 20,000 × magnification. (B) Fiber diameter distribution plots of the experimental scaffolds (n = 150). (C) Corresponding energy-dispersive X-ray spectroscopy (EDS) plots indicate elemental composition and highlight the presence of Ca and P. (D) Fiber diameter of the experimental scaffolds (n = 150). Statistical analysis was performed using Kruskal–Wallis followed by Dunn's post hoc test (p < 0.05). (E) Fourier-transform infrared spectroscopy (FTIR) spectra of neat PCL, nHA, HT, and the experimental scaffolds. Full spectra (4000 to 400 cm⁻¹) and insert from 4000 to 3000 cm⁻¹ wavenumbers.

The FTIR spectra (Fig. 2E) confirmed the characteristic bands of PCL observed at ∼2865 cm⁻¹ and ∼2945 cm⁻¹ (CH₂ stretching), ∼1720 cm⁻¹ (C=O stretching), and ∼1180–1045 cm⁻¹ (C–O–C stretching) [41]. These bands remained visible in all formulations, indicating chemical stability of the polymer structure after both hydrolysis and HT adsorption. The presence of phosphate-related (PO₄³⁻) bands from nHA was confirmed at ∼1040 cm⁻¹, ∼600 cm^−1, and ∼560 cm⁻¹ [42], validating the presence and structural integrity of the inorganic phase. HT-containing scaffolds exhibited weak bands near 1580 cm⁻¹ and 1500 cm⁻¹ (C=C from aromatic stretching) and ∼1260 cm⁻¹ and ∼1040 cm⁻¹ (C–O stretching from phenolic groups), consistent with HT adsorption. A slight reduction in the intensity of the C=O band was observed after hydrolysis. The broad –OH stretching band (∼3600–3200 cm⁻¹) exhibited an increase in intensity and breadth in modified formulations, particularly in the H-PCL/nHA + HT group, suggesting the introduction of hydroxyl groups through alkaline hydrolysis.

Physicochemical characterization revealed a significant increase in surface wettability after alkaline hydrolysis. Contact angles decreased from ∼129° (PCL/nHA) and ∼122° (PCL/nHA + HT) in non-hydrolyzed scaffolds to ∼50° (H-PCL/nHA) and ∼16° (H-PCL/nHA + HT), indicating a shift from hydrophobic to hydrophilic behavior (Fig. 3A). Hesperetin adsorption (Fig. 3B) increased to ∼36% in H-PCL/nHA + HT (∼72 μM HT) compared to ∼20% in PCL/nHA + HT (∼40 μM HT), confirming enhanced drug binding after surface hydrolysis (p < 0.0001). The released fractions corresponded to ∼25% and ∼50% of the adsorbed amount at 7 days for H-PCL/nHA + HT and PCL/nHA + HT, respectively. Hesperetin release profiles (Fig. 3C) demonstrated a sustained release pattern, with hydrolyzed scaffolds reaching ∼13 μM cumulative release at 24 h, compared to ∼15 μM in the non-hydrolyzed scaffolds. A plateau in cumulative release was observed by 168 h (7 days), with H-PCL/nHA + HT and PCL/nHA + HT reaching ∼18 μM and ∼20 μM, respectively. The Weibull and Korsmeyer-Peppas models were the best fits for the release data of both scaffolds, suggesting a diffusion-controlled release mechanism (Fig. 4).

Fig. 3.

Physicochemical characterization of the experimental scaffolds. (A) Contact angle analysis of the experimental scaffolds. Data are mean ± SD. Statistical analysis was performed using Welch's ANOVA followed by the Games-Howell post hoc test (p < 0.05). The horizontal dashed line indicates the threshold between the hydrophobic (>90°) and hydrophilic (<90°) properties. Representative images of water droplets on the scaffold surfaces are shown. (B) Hesperetin adsorption (%) onto non-hydrolyzed and hydrolyzed PCL/nHA scaffolds. Data are mean ± SD. Statistical analysis was performed using an unpaired t-test. (C) pH values of the PBS immersion medium over 28 days. (D) Hesperetin release profile from HT-loaded scaffolds over 168 h. (E) Calcium release from the scaffolds over 49 days. (f) Remaining scaffold mass (%) over a 120-day degradation period. For (C-F), data are mean ± 95% confidence intervals. n = 8 for all experiments.

Fig. 4.

Cumulative release profiles of hesperetin from PCL/nHA + HT and H-PCL/nHA + HT scaffolds were fitted to six mathematical models: Zero-order, First-order, Higuchi, Korsmeyer–Peppas, Weibull, and Peppas–Sahlin. The best fits were observed for the Weibull (R² = 0.996–0.999) and Peppas–Sahlin (R² = 0.997–0.998) models. For the Korsmeyer–Peppas model, the release exponent n ranged from 0.21 to 0.22, suggesting Fickian diffusion from a cylindrical matrix. The Weibull shape parameter β was ∼0.40–0.45, also indicating a diffusion-dominated mechanism. In the Peppas–Sahlin model, the diffusion coefficient (k₁) was greater than the relaxation coefficient (k₂), confirming the predominance of Fickian's transport. Notably, hydrolyzed fibers (H-PCL/nHA + HT) exhibited slightly higher release constants (k) across models, indicating enhanced release kinetics following alkaline treatment.

Calcium release increased gradually, with a burst release up to 3 days, ranging from ∼1.8 to ∼2 mM for PCL/nHA + HT and H-PCL/nHA + HT, respectively (Fig. 3D). Overall, hydrolyzed scaffolds exhibited elevated calcium release over 49 days (equivalent to 7 weeks), with H-PCL/nHA + HT reaching ∼4.3 mM, compared to ∼2.8 mM in the control (PCL/nHA). pH measurements remained stable over 28 days for all groups, with values ranging from ∼6.3 to 7.8, and a slight increase was observed at day 5 (Fig. 3E). Scaffold degradation was slow over 120 days (4 months), with all groups retaining over 95% of their original mass, thereby confirming the long-term structural integrity of the PCL-based scaffolds even after alkaline hydrolysis (Fig. 3F).

3.2. Dental pulp cells migration and fibronectin expression

Using a Transwell® migration assay, we evaluated the 24-h hDPC migration toward bioactive cues released by the experimental scaffolds (Fig. 5A). The quantification of migrated cells demonstrated increased migration in groups exposed to hydrolyzed scaffolds and HT-containing formulations (Fig. 5C). The H-PCL/nHA + HT group exhibited the highest number of migrated cells. Actin staining revealed the cytoskeletal organization of migrated cells (Fig. 5D). Cells exposed to H-PCL/nHA + HT exhibited enhanced actin area, indicating increased spreading and reorganization linked to migratory activity. Immunostaining for fibronectin (Fig. 5B) revealed larger fibronectin-positive areas in cells exposed to hydrolyzed and HT-adsorbed scaffolds, with the most intense labeling observed for H-PCL/nHA + HT (Fig. 5E). The fibronectin/actin ratio was quantified (Fig. 5F), revealing elevated values for all modified scaffolds compared to the PCL/nHA scaffold. The H-PCL/nHA + HT scaffold exhibited the highest ratio, thereby confirming that migrated cells synthesized greater amounts of fibronectin.

Fig. 5.

Dental pulp cells migration and fibronectin expression. (A) Schematic of the experimental setup for evaluating cell migration. Human dental pulp cells (hDPCs) were cultured inside porous inserts, and scaffolds were placed at the bottom of the well, supported by stainless steel rings. Cells that migrated to the lower surface were analyzed after 24 h. (B) The upper row shows representative fluorescence images of actin filaments (red) and nuclei (blue) in migrated cells. Scale bars = 250 μm. The lower row shows fibronectin expression (green) in migrated cells, overlaid with actin (red) and nuclei (blue). Scale bars = 100 μm. (C) Quantification of migrated cell number per scaffold group. (D) Quantification of actin-positive area (%). (E) Quantification of fibronectin-positive area (%). (F) Fibronectin/actin ratio. All data are presented as mean ± SD (n = 12 areas from 4 independent samples). Statistical analyses were performed using one-way ANOVA followed by Tukey's post hoc test (p < 0.05).

3.3. Direct interaction between scaffolds and dental pulp cells

The direct interaction between the experimental scaffolds and human dental pulp cells (hDPCs) is shown in Fig. 6. A time-dependent increase in cell metabolism was observed for all groups at all time points (p < 0.0001), indicating that all formulations supported proper cell proliferation (Fig. 6A). The Live/Dead images confirmed the cytocompatibility and proliferation of hDPCs on the scaffolds, revealing dense and increasing populations of viable (green) cells compared to dead (red) cells over time (Fig. 6B). On the seventh day, a divergence in cell metabolism was observed among the formulations. A significant increase in metabolic activity was detected in response to HT adsorption, hydrolysis, and their combination, which exhibited the highest metabolic activity (Fig. 6A). Furthermore, alkaline hydrolysis enhanced cell spreading on day 1, as evidenced by a discernible decrease in cell clustering in surface-modified samples (Fig. 6B).

Fig. 6.

Direct interaction between scaffolds and human dental pulp cells. (A) Cell metabolism (alamarBlue assay) at 1, 7, and 14 days. Statistical analysis was performed using two-way ANOVA followed by Sidak's post hoc test (p < 0.05). (B) Representative Live/Dead staining images at 1, 7, and 14 days show viable (green) and membrane-damaged/dead cells (red). Scale bars = 100 μm. (C) Relative gene expression of COL1A1, ALPL, DSPP, and DMP1 at day 14. Statistical analysis was performed using one-way ANOVA followed by Tukey's test (p < 0.05). (D) Immunofluorescence analysis of DSPP and DMP-1 protein expression at day 21. Scale bars = 250 μm. Quantification of the positive area was analyzed using one-way ANOVA, followed by Tukey's test. (E) Representative images of mineralized matrix formation on scaffolds stained with Alizarin Red at day 21. Scale bars = 1 mm. Darker red–brown regions indicate mineral deposition, whereas bright-red coloration reflects intrinsic scaffold background staining (see scaffold-only – no cells – control). Quantification reflects whole-scaffold extraction. Alizarin Red staining (%) and soluble calcium content (mM) were analyzed using one-way ANOVA followed by Tukey's test (p < 0.05). Scale bars = 1 mm. All data are presented as mean ± SD (n = 8).

Relative gene expression on day 14 showed upregulation of ALPL and DSPP for HT adsorption, hydrolysis, and their combination. However, only the H-PCL/nHA + HT scaffold demonstrated a significant upregulation of DMP1 (Fig. 6C). Conversely, both hydrolyzed scaffolds downregulated COL1A1 expression (Fig. 6C). Immunofluorescence staining at day 21 revealed that the DSPP-positive area reached ∼18% for H-PCL/nHA + HT, the only group significantly higher than the control (∼10%) (Fig. 6D). However, DMP-1 immunoexpression was comparable across the formulations (p = 0.0726).

Alizarin Red staining revealed that alkaline hydrolysis enhanced matrix mineralization regardless of HT adsorption. The most intense mineralization activity was observed for H-PCL/nHA + HT (∼170%), which was nearly 1.7 times higher than that of PCL/nHA. Soluble calcium quantification corroborated these findings, showing significantly lower values in H-PCL/nHA + HT (∼1.4 mM) compared to PCL/nHA (∼1.7 mM), confirming the enhanced bioactivity of H-PCL/nHA + HT scaffolds due to higher calcium uptake and deposition by cells into the mineralized matrix (Fig. 6E).

3.4. Indirect interaction between scaffolds and inflammatory cells

The effects of scaffold extracts on RAW 264.7 murine macrophages are depicted in Fig. 7. Following stimulation with LPS, the metabolic activity of macrophages was above 70% viability. Moreover, formulations incorporating HT demonstrated enhanced cell metabolism in comparison to PC (Fig. 7A). Gene expression analysis showed no significant modulation of Tnf or Nos2 in macrophages treated with extracts from all formulations after 3 h relative to PC (p ≥ 0.247) (Fig. 7B and D). In contrast, PCL/nHA + HT significantly downregulated Il1b (Fig. 7C). After 24 h of exposure, HT-containing formulations significantly reduced the levels of secreted TNF-α, IL-1β, and nitrite compared to PC (Fig. 7E–G).

Fig. 7.

Indirect interaction between scaffolds and inflammatory cells. RAW 264.7 murine macrophages were previously stimulated (positive control, PC) or not (negative control, NC) with LPS (100 ng/mL), followed by treatment with 24-h extracts from each scaffold. (A) Cell metabolism (alamarBlue assay) of macrophages after 24-h exposure to scaffold extracts. Statistical analysis was performed using one-way ANOVA followed by Tukey's post hoc test (p < 0.05). (B-D) Relative gene expression of Tnf, Il1b, and Nos2 after 3-h exposure to the extracts. Statistical analysis was performed using Welch's ANOVA followed by the Games-Howell post hoc test (p < 0.05). (E-G) Secreted TNF-α, IL-1β, and nitrite levels in macrophage culture supernatants after 24-h exposure to the extracts. Statistical analysis was performed using Welch's ANOVA followed by the Games-Howell post hoc test (p < 0.05). All data are presented as mean ± SD (n = 8). The asterisk indicates no detectable protein in the NC group extract.

3.5. In vitro pulp capping under an inflammatory microenvironment

The inflammatory and regenerative responses evaluated through a 3D in vitro pulp capping model using hDPCs are shown in Fig. 8A. The Live/Dead assay revealed high levels of cell viability in 3D cultures, with minimal membrane-damaged or dead cells in both non-challenged (−LPS) and LPS-challenged (+LPS) conditions (Fig. 8B). Furthermore, the detection of reactive oxygen species (ROS) revealed increased oxidative stress in +LPS groups compared to −LPS controls (Fig. 8C). These analyses validated the model. After 3 h of treatment, H-PCL/nHA + HT scaffolds significantly downregulated IL8, IL6, IL1B, TNF, and MMP9 inflammation-related gene expression in LPS-stimulated hDPCs compared to PC (Fig. 8D). H-PCL/nHA + HT increased IL10 expression compared to PC, whereas HMOX1 remained unaffected by any condition (Fig. 8D). After 14 days, regenerative markers were differentially expressed, with H-PCL/nHA + HT significantly increasing COL1A1, ALPL, DSPP, OPN, and OCN; TGFB1 remaining unchanged; and PCL/nHA significantly increasing VEGFA (Fig. 8E).

Fig. 8.

Inflammatory and regenerative responses in a 3D in vitro pulp capping model. (A) Schematic of the in vitro pulp capping model used in this study. Human dental pulp cells (hDPCs) were cultured in 3D collagen matrices in the absence (–LPS) or presence (+LPS) of chronic inflammatory stimulus. The 3D cultures were placed inside artificial pulp chambers set with dentin discs. An exposure site was created in the dentin to expose the 3D culture, mimicking a pulp capping procedure with the experimental scaffolds. (B) Representative Live/Dead staining images of cells spread in the 3D matrix after culturing cells in the absence (–LPS) or presence (+LPS) of chronic inflammatory stimulus. Viable cells are shown in green, and membrane-damaged/dead cells in red. Brightfield filter contrasts with the filters. Scale bars = 1 mm and 250 μm. (C) Detection of reactive oxygen species (ROS, green) in the 3D matrix after culturing cells in the absence (–LPS) or presence (+LPS) of chronic inflammatory stimulus. Brightfield filter contrasts with the filter. Scale bars = 1 mm and 250 μm. (D) Relative gene expression of inflammation-related markers (IL8, COX2, IL6, MMP9, IL1B, IL10, TNF, and HMOX1) in LPS-stimulated hDPCs after 3-h exposure to the scaffolds. (E) Relative gene expression of regeneration-related markers (COL1A1, OPN, ALPL, OCN, DSPP, TGFB1, DMP1, and VEGFA) in 3D-cultured cells after 14 days. All gene expression data were analyzed using Welch's ANOVA followed by the Games-Howell post hoc test (p < 0.05) and are shown as mean ± SD (n = 6).

3.6. In vivo subcutaneous biocompatibility, immunomodulation, and mineralization potential

Regarding biocompatibility, all scaffolds integrated well with the surrounding tissues, and no necrosis was observed in any of the animals (Fig. 9A and B). Histologically, at 1 week, the H-PCL/nHA + HT scaffold elicited a lower inflammatory response (1531 ± 478 cells per field) than PCL/nHA (2051 ± 1027; p = 0.0213). At 4 weeks, both scaffolds exhibited similar inflammatory cell counts (approximately 1700 cells per field), with no significant difference. By 8 weeks, H-PCL/nHA + HT again demonstrated better biocompatibility, with fewer inflammatory cells (1791 ± 485) than the PCL/nHA control (2131 ± 472; p = 0.0001).

Fig. 9.

In vivo subcutaneous biocompatibility. (A) H&E-stained sections of PCL/nHA and H-PCL/nHA + HT scaffolds explanted at 1-, 4-, and 8-weeks post-implantation (100 × ; scale bar, 100 μm). Yellow arrows indicate collagen fiber deposition, red arrows indicate inflammatory cells, and dark blue brackets with arrows delineate the fibrous capsule surrounding the scaffold. (B) Inflammatory cell counts (cells/field) were quantified with ImageJ and reported as mean ± SD. Two-way ANOVA with Tukey's post hoc test; n = 8; α = 0.05.

Over time, cells progressively infiltrated between the scaffold fibers, accompanied by increased collagen fibril deposition, with the H-PCL/nHA + HT scaffolds tending to exhibit a more organized deposition pattern (Fig. 9A). Qualitatively, H-PCL/nHA + HT appeared to elicit a milder and more organized host response, with a relatively thinner capsule, fewer inflammatory cells, and more seamless integration at 8 weeks (Fig. 9A).

To evaluate the immunomodulatory effect of H-PCL/nHA + HT, we quantified the areas of iNOS (M1) and CD163 (M2) immunopositivity (Fig. 10A). At 1 week, H-PCL/nHA + HT exhibited a markedly lower iNOS ⁺ area than PCL/nHA (0.542 ± 0.117% vs. 1.450 ± 0.695%; p < 0.0001), corresponding to a 2.7-fold decrease, while CD163⁺ was higher with H-PCL/nHA + HT (0.766 ± 0.095% vs. 0.478 ± 0.061%; p < 0.0001), a 1.6-fold increase (Fig. 10C). Therefore, the M2:M1 ratio at 1 week was significantly higher with H-PCL/nHA + HT than with PCL/nHA, indicating a pronounced early shift toward a pro-resolving macrophage phenotype. By 4 weeks, iNOS⁺ levels in the PCL/nHA decreased to match those of H-PCL/nHA + HT, indicating resolution of inflammation over time, and CD163⁺ values were similar between groups at this later time point. Collectively, these data indicate that H-PCL/nHA + HT drives the early suppression of pro-inflammatory (M1) signaling and enhances anti-inflammatory (M2) polarization, with group differences diminishing as the response matures. To assess the osteogenic potential of the H-PCL/nHA + HT scaffold at 8 weeks, we examined late-stage markers osteopontin (OPN) and osteocalcin (OCN) by immunostaining. No samples in either group exhibited detectable OPN or OCN signals at this time point (Fig. 10B).

Fig. 10.

Immunofluorescent assessment of macrophage polarization and osteogenic markers after subcutaneous implantation. (a) iNOS (M1) and CD163 (M2) around PCL/nHA and H-PCL/nHA + HT at 1- and 4-weeks post-implantation. (b) Osteopontin (OPN) and osteocalcin (OCN) at 8 weeks post-implantation. All panels were labeled with an Alexa Fluor 488-conjugated secondary antibody (green) and imaged at 200 × ; scale bar, 50 μm. (c) Areas of positive labeling were quantified in ImageJ and reported as mean ± SD. Two-way ANOVA with Tukey's post hoc test; n = 8; α = 0.05.

4. Discussion

Unlike conventional scaffolds designed solely for cytocompatibility, this study developed and characterized an innovative fibrous scaffold functionalized with the flavonoid hesperetin (HT) for actively instructing immune and progenitor cell behavior. We proposed a surface-modified platform based on previously established polycaprolactone/nano-hydroxyapatite (PCL/nHA) fibers [30], which was modified through alkaline hydrolysis to enhance hydrophilicity, cell affinity, and ion release, while also modulating HT adsorption and release. This design integrates surface chemistry engineering with drug delivery, aiming to recapitulate key immunobiological mechanisms underlying dentin-pulp regeneration. This multifunctional strategy demonstrated effective immunomodulatory activity, enhanced the proliferation and migration of cells, and stimulated odontogenic gene expression and mineralized matrix formation by pulp cells, even after chronic LPS challenge. In vivo, it reduced inflammation and induced an early pro-resolutive response, facilitating the deposition of ECM. Importantly, the dentinogenic potential is supported by in vitro and 3D pulp model findings, while in vivo focused primarily on immunomodulation. Collectively, these results reinforce the concept that scaffold surface modification and targeted flavonoid release can act synergistically to coordinate inflammatory and regenerative events, advancing immunomodulatory biomaterials for VPT.

Alkaline hydrolysis effectively modified the surface of PCL/nHA scaffolds, significantly improving wettability from ∼129° (control) to below 50°, helping to overcome the intrinsic hydrophobicity of PCL and promote biological interactions [32,33]. FTIR analysis revealed spectral changes that are consistent with surface chemical modification. The observed decrease in the C=O peak supports the hypothesis of partial surface scission of ester bonds in the PCL matrix [32], while the persistence of sharp C–O–C bands indicates that the backbone structure of PCL was preserved. Furthermore, the observation of a broadening of the hydroxyl stretching band was predominantly evident in samples that underwent hydrolysis, particularly for the H-PCL/nHA + HT group. This finding indicates an enhancement in hydrogen bonding between HT and newly introduced polar groups. The enhanced HT adsorption efficiency in this group corroborates the role of hydrolysis in enabling more specific interactions (likely hydrogen bonding between HT and newly introduced polar groups), which extend beyond nonpolar forces such as hydrophobicity or van der Waals interactions. Release data confirmed a controlled, diffusion-driven release mechanism. Morphological analysis showed that hydrolysis reduced the average fiber diameter (from ∼600 nm to ∼400 nm) without compromising the uniform interconnected fibrous structure. Degradation behavior further confirmed that scaffold stability was not compromised, and the pH remained within physiological levels (6-8) across all formulations, which was compatible for retaining HT stability and driving dentin-pulp complex regeneration [8,19,20,26,30]. Overall, these physicochemical findings validate alkaline hydrolysis as a simple and effective strategy to enhance hydrophilicity, drug-matrix interactions, and molecular retention, while maintaining scaffold integrity.

Effective regeneration in VPT requires stem cell migration and extracellular matrix (ECM) remodeling. Therefore, our first biological experiments utilized a Transwell assay to indirectly expose hDPCs to the formulations. Hydrolyzed scaffolds enhanced cell migration, with H-PCL/nHA + HT showing the greatest chemotactic effect. Calcium ions are key regulators of many cellular processes, and their effects on cell migration are mediated by the calcium-sensing receptor (CaSR) in response to changes in extracellular calcium [43,44]. Moreover, MAP kinases are central to migration, and calcium ions have been shown to contribute to their activation [44]. In our study, scaffold calcium release profiles demonstrated a gradual increase in calcium ions over time, with hydrolyzed scaffolds releasing more calcium than non-hydrolyzed ones. These data suggest that hydrolysis-induced increases in calcium availability, combined with local HT release, created a bioactive ionic and molecular gradient capable of enhancing chemotaxis and promoting reparative cell recruitment.

The increased calcium availability, especially for the H-PCL/nHA + HT formulation, may have activated CaSR signaling, stimulating pathways that promote actin polymerization and cytoskeletal reorganization, driving cell migration [45,46]. Thus, the calcium release from the scaffolds supports cell adhesion and proliferation and enhances the active migration of hDPCs, an essential step for successful pulp regeneration. In addition, elevated extracellular calcium has been shown to upregulate fibronectin synthesis in pulp cells [47]; in our work, hydrolyzed scaffolds increased fibronectin deposition, as evidenced by a higher fibronectin-to-actin ratio. Highly expressed in the dental basement membrane [48], fibronectin is an ECM component that mediates odontoblast elongation, polarization, and terminal differentiation from undifferentiated pulp cells [49], which are essential for dental pulp regeneration. Hence, the hydrolyzed HT-laden scaffolds appear to generate an instructive interface that combines biochemical (HT, Ca²⁺) and topographical cues to guide pulp cell behavior toward a regenerative phenotype.

Once hDPCs reach the biomaterial, they must proliferate and differentiate into odontoblast-like cells. To assess these processes, we evaluated the behavior of hDPCs cultured directly on the scaffolds, which reflects mechanisms critical for successful vital pulp therapy. All scaffold groups supported cell viability and proliferation, as evidenced by time-dependent increases in metabolic activity and the presence of dense populations of viable cells, confirming overall cytocompatibility. Alkaline hydrolysis enhanced early cell attachment and proliferation, likely by increasing surface hydrophilicity and topography, which promote protein adsorption and integrin-mediated adhesion [33]. The metabolic increase observed for H-PCL/nHA + HT at day 7 can be attributed to the synergistic effect of calcium ions and HT release, which can both activate MAP kinases signaling [23,24,44]. The activation of BMP signaling and upregulation of RUNX2 expression have also been associated with the effects of HT on osteogenic and odontogenic differentiation [18]. In addition, this microenvironment supported early odontogenic differentiation, as evidenced by the upregulation of ALPL and DSPP across HT- and hydrolysis-modified groups. The exclusive upregulation of DMP1 by H-PCL/nHA + HT suggests that the synergistic effect of surface functionalization and HT delivery more effectively activates differentiation pathways. The immunostaining confirmed enhanced odontoblastic activity, as indicated by a higher DSPP-positive area for H-PCL/nHA + HT, which reflects the presence of both the DSPP precursor and its cleaved product, dentin phosphoprotein (DPP), representing approximately 50% of the dentine non-collagenous proteins and involved in dentin matrix formation and mineralization [50]. Together, these results demonstrate that the integration of hydrolysis-induced surface activation and HT-mediated signaling orchestrates a favorable microenvironment for dentinogenesis.

Calcium ions are known to effectively drive the odontogenic differentiation of pulp cells by activating the mTOR and BMP2-mediated Smad1/5/8 and Erk1/2 signaling pathways [51,52]. Importantly, hesperetin has also been shown to induce the osteo/odontoblastic differentiation of mesenchymal stem cells via the ERK and Smad signaling pathways [24], also supporting the upregulation of odontogenic markers observed in this study. Additionally, the Wnt/β-catenin signaling pathway has also been related to the odontogenic differentiation of dental pulp cells by naringenin, a flavanone presenting a similar chemical structure [53]. Although phosphorylation-specific assays were not included in the present work, these pathways should be interpreted as biologically coherent, literature-supported hypotheses aligned with our multi-level functional outcomes. Therefore, the combined release of these two molecules may have synergistically mediated the odontogenic potential in our study. The downregulation of COL1A1 in modified scaffolds indicates that the scaffold structure can properly mimic the native ECM structure. This reflects a rapid shift from early matrix deposition toward biomineralization, which was also potentiated by hydrolysis, as shown by stronger mineralized matrix deposition. The contrast between Alizarin Red staining and soluble calcium levels highlights a key mechanism: higher calcium deposition into the mineralized matrix rather than accumulation in the soluble fraction confirms active mineralization rather than passive precipitation. These results reinforce the dual role of the system: biochemical induction through HT and ionic signaling through calcium, establishing a biomimetic environment for the regeneration of the dentin–pulp complex. Future studies will include targeted phosphorylation assays to directly confirm these mechanisms under orthotopic models.

Modulating the inflammatory response without completely suppressing it is essential to support regeneration. Macrophages stimulated with LPS were used to assess the scaffold's pro-resolving potential, mimicking the early immune microenvironment of an inflamed, caries-exposed pulp tissue. The use of 24-h scaffold extracts was based on the HT release profile, which demonstrated a burst release within the first day, ranging to previously reported immunoresolutive concentrations (10 to 20 μM) [14,22]. This early window is also crucial for immune cell crosstalk and fate decisions at the site of exposure [6]. Release data reinforced that the early burst and sustained HT release are related to the diffusion mechanism from the scaffold matrix. Although no changes in gene expression were observed at 3 h, the effects on cytokine secretion after 24 h support a post-transcriptional regulatory mechanism. Hesperetin exhibits immunoresolutive activity even at concentrations as low as 1 μM, whether in its free form or delivered via nanocrystal systems [18], highlighting its potent bioactivity and suitability for low-dose therapeutic strategies. Mechanistically, HT has been shown to modulate the NF-κB pathway and NLRP3 inflammasome activation, which explains the reduced secretion of TNF-α and IL-1β [14,22,54]. Importantly, previous work from our group demonstrated that hesperetin inhibits NF-κB activation and reduces oxidative stress in LPS-challenged cells [14,55,56], reinforcing the biological plausibility of the mechanisms inferred here. Although the absolute cytokine levels were within a moderate range, this finding is consistent with the known requirement of a secondary stimulus to promote full inflammasome activation in macrophages, particularly for IL-1β maturation before secretion [57]. The reduced nitrite levels also suggest a reduction in nitric oxide production, a diffusible molecule often upregulated in pulpitis, which contributes to oxidative stress and extracellular matrix breakdown [5,7]. These molecules are key mediators of pulp tissue degradation and nociceptive signaling. Therefore, the ability of HT-laden scaffolds to modulate, rather than block, the inflammatory response may contribute to a more favorable environment for pulp regeneration. This selective modulation is particularly relevant, given clinical data demonstrating elevated IL-1β levels in teeth diagnosed with irreversible pulpitis [58], which reinforces its role as a key mediator of disease progression.

To closely mimic the clinical environment of the dental pulp under inflammatory conditions, we used a 3D in vitro pulp capping model. This model simulates the dentin-pulp interface and supports pulp cell differentiation by incorporating artificial pulp chambers (APCs) and dentin discs, able to release growth factors [15]. The exposure to LPS for 3 days aims to mimic a pulpitis-like microenvironment, while the 3-h and 14-day time points were selected to capture the immediate transcriptional response to the early burst release of hesperetin and its ability to modulate an established LPS-induced inflammatory phenotype and longer-term regenerative processes [4,15]. The H-PCL/nHA + HT scaffolds exhibited an initial burst release profile of HT, followed by sustained release over 7 days. This controlled multimodal delivery effectively modulated chronic inflammation, likely through downregulation of genes encoding pro-inflammatory cytokines, such as IL-8, TNF-α, and IL-6, and restored the regenerative potential of dental pulp cells by promoting odontogenic signaling, including the upregulation of DSPP and DMP1. This dual-phase mechanism (early immune modulation followed by regenerative activation) illustrates the temporal coordination needed for successful dentin–pulp regeneration. These findings align with previous studies demonstrating that sustained drug delivery systems can effectively address persistent inflammation while supporting stem cell-driven mineralization in VPT [15,[59], [60], [61], [62]].

As noted above, successful VPT depends on modulating, rather than ending, the inflammatory response [11]. Although our evaluation utilized an ectopic subcutaneous model, this platform is a well-established first step in vivo for probing biomaterial behavior. In this context, both H-PCL/nHA + HT and control PCL/nHA scaffolds demonstrated high biocompatibility in vivo, integrating well with host tissue and exhibiting no signs of necrosis or adverse reactions. Notably, the hesperetin-releasing scaffold elicited a markedly tempered inflammatory response relative to the control, evidenced by fewer infiltrating inflammatory cells and immunohistochemical markers of a pro-healing macrophage phenotype. Compared with PCL/nHA, it showed reduced M1 and elevated M2 macrophages. This polarization pattern indicates an early immune shift from a pro-inflammatory to a pro-resolutive microenvironment, a hallmark of biomaterials capable of actively engaging host immunity rather than merely evading it [63,64].

This early skewing toward an M2-dominant response aligns with HT's known immunoresolutive activity, as citrus flavanones can inhibit key inflammatory pathways and downregulate the expression of pro-inflammatory cytokines in injured tissues [65,66]. Although by week 4, the inflammatory cell counts in the two groups had converged, indicating resolution of the acute response in both, by week 8, the H-PCL/nHA + HT group again showed lower chronic inflammation than the control. This sustained immunomodulatory profile suggests that the local release of HT maintained macrophage homeostasis and prevented the persistence of chronic inflammation, contributing to a more favorable healing trajectory. Histologically, the hesperetin-loaded implants were surrounded by a thinner, more organized collagenous capsule at 8 weeks, in contrast to the thicker, less organized fibrous capsule around plain PCL/nHA. A thin, well-organized capsule is indicative of a gentle foreign body response, suggesting that local hesperetin release continued to modulate the immune microenvironment favorably over the long term [63,64]. Such findings reinforce the emerging view that the success of bioactive scaffolds lies not only in their osteo/odontogenic potential but also in their capacity to direct host immune adaptation, ensuring long-term tolerance and integration.

Conversely, the in vivo subcutaneous (ectopic) model did not demonstrate the scaffold's osteogenic activity under the conditions tested. By 8 weeks post-implantation, neither group showed detectable OPN or OCN, markers of new mineralized matrix, indicating no ectopic bone formation. This likely reflects the limitations of subcutaneous implantation for assessing osteogenesis, as it lacks the native odontogenic and osteoinductive cues of bone sites. The absence of mineralized tissue formation in this setting highlights the model's intrinsic constraint rather than a deficiency of the material itself [67]. Ectopic environments typically support only minimal bone formation without strong inductive stimuli [67]. Accordingly, although the H-PCL/nHA + HT scaffold clearly promoted an immunoresolutive milieu in vivo, its hypothesized dentinogenic potential could not be confirmed in this setting. Nonetheless, the immunomodulatory behavior observed here provides a mechanistic basis for future orthotopic validation, where the interplay between macrophage polarization, pulp cell recruitment, and dentin bridge formation can be more accurately captured. This result does not rule out HT's potential for regeneration but rather points to a limitation of the study and highlights the need for organotypic or orthotopic dental/pulp models, such as those described in previous studies [55,56], to assess whether the observed immunomodulation can lead to dentin bridge formation and will be pursued in future work.

While encouraging, the present findings should be interpreted with consideration of certain limitations. The subcutaneous implantation model, although not representative of the dentin-pulp environment, was strategically adopted as an initial translational step to dissect immunomodulatory mechanisms before orthotopic validation, preventing the interference of additional tissue- and material-specific variables [68]. Although prioritizing and exhausting mechanistic and alternative in vitro models as recommended [68], this approach does not fully capture the complex interplay of inflammatory events, cellular responses, and mineralized tissue formation that occur in the pulp exposure, especially when combined with the lining and restorative materials. Another limitation lies in the stability of the bioactive molecule and the reproducibility of its release profile within the pulp capping scenario, which are critical to ensuring consistent therapeutic outcomes.

Nevertheless, these limitations are not unique to experimental systems; even clinically established materials exhibit intrinsic drawbacks, such as uncontrolled alkalinity, transient ion release, and limited immunoresolutive capacity, contrasting with the tunable, sustained bioactivity achieved in the present system. Moving forward, more elaborate orthotopic in vivo studies considering chronic caries-driven exposures are required to validate these results in clinically relevant settings, allowing direct comparisons with established VPT materials and glass ionomer cements. Such studies will help determine whether this strategy can contribute to post-operative pain control, ensuring long-term safety and dentin bridge formation, further consolidating the link between immunomodulation and functional regeneration. Compared to current clinical gold standards (e.g., Mineral Trioxide Aggregate – MTA), which rely on an initial necrotic layer and strong alkaline stimulation, the present scaffold acts through controlled Ca²⁺ release and delivery of HT. This combination may promote early inflammatory resolution and subsequent odontogenic signaling. Furthermore, the fibrous architecture offers barrier-like behavior distinct from particulate or putty-like bioceramics, addressing handling and integration considerations noted clinically [69], and potentially optimizing long-term outcomes in VPT through controlled molecular and immune signaling.

5. Conclusion

This study introduces a multifunctional fibrous scaffold for controlled HT release and surface modification through alkaline hydrolysis, improving interaction with hDPCs. By increasing hydrophilicity, maintaining drug release, and supplying calcium, the scaffold enhanced hDPC migration, proliferation, and odontogenic differentiation, while also modulating inflammation and promoting regenerative gene expression in both traditional and 3D models. In vivo, it decreased inflammation, supported organized ECM deposition, and triggered an early M2-skewed response. These findings highlight the translational potential of multifunctional scaffolds as biologically rational alternatives to conventional capping materials, with future research needed to confirm their regenerative potential in clinically relevant environments for more predictable vital pulp therapy outcomes.

CRediT authorship contribution statement

Igor Paulino Mendes Soares: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Visualization, Writing – original draft. Caroline Anselmi: Conceptualization, Investigation, Methodology, Validation, Writing – review & editing. Renan Dal-Fabbro: Formal analysis, Investigation, Methodology, Validation, Writing – review & editing. Lídia de Oliveira Fernandes: Data curation, Investigation, Methodology. Gabriel Cardoso Pinto: Formal analysis, Investigation, Methodology, Writing – review & editing. Rodolfo Debone Piazza: Formal analysis, Investigation, Methodology, Validation, Writing – review & editing. Carlos Alberto de Souza Costa: Conceptualization, Methodology, Resources, Validation, Writing – review & editing. Josimeri Hebling: Conceptualization, Funding acquisition, Methodology, Supervision, Validation, Writing – review & editing. Marco C. Bottino: Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Validation, Writing – review & editing.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Marco C. Bottino reports financial support was provided by National Institute of Dental and Craniofacial Research. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.