Resveratrol-loaded human dental pulp stem cell-derived exosomes enhance the efficacy of blue LED-mediated antimicrobial photodynamic therapy against mature cariogenic biofilms

Abstract

Antimicrobial photodynamic therapy (aPDT) is a promising adjunctive approach for biofilm-associated oral diseases, though the poor stability and bioavailability of natural photosensitizers remain limiting factors. This study evaluated the anti-biofilm efficacy of resveratrol-loaded human dental pulp stem cell-derived exosomes (Res@hDPSC-Exos) activated by blue LED irradiation against mature cariogenic biofilms. hDPSC-Exos were isolated, characterized, and loaded with resveratrol; physicochemical properties, encapsulation efficiency, and exosomal integrity were assessed by TEM, DLS, and CD63 expression. Mature Streptococcus mutans and Lactobacillus acidophilus biofilms were established on human enamel specimens. The minimum biofilm reduction concentration (MBRC) of Res@hDPSC-Exos and minimum biofilm reduction dose (MBRD) of blue LED were determined, and biofilms were treated according to six groups: untreated control, free resveratrol, empty hDPSC-Exos, blue LED alone, and Res@hDPSC-Exos-mediated aPDT at 1/2× and 1/4×MBRC. Biofilm biomass, metabolic activity, and gtfB/slpA virulence gene expression were assessed by crystal violet staining, XTT assay, and qRT-PCR, respectively. Resveratrol was successfully encapsulated into hDPSC-Exos (encapsulation efficiency ≈ 63%) without compromising vesicle morphology or membrane integrity. The MBRC of Res@hDPSC-Exos was 250 µg/mL for S. mutans and 125 µg/mL for L. acidophilus, and the MBRD of blue LED was 300 s for both species. Free resveratrol, empty hDPSC-Exos, and blue LED alone produced only modest, non-significant effects (≈ 5–15%, P > 0.05). In contrast, Res@hDPSC-Exos-mediated aPDT at 1/2×MBRC with sub-MBRD irradiation (240 s) significantly reduced biofilm biomass (84.7% and 75.2%) and metabolic activity (88.9% and 85.0%) in S. mutans and L. acidophilus, respectively (P < 0.001), and downregulated gtfB and slpA expression to 0.15-fold and 0.19-fold of controls (P < 0.001). Res@hDPSC-Exos-mediated aPDT effectively suppressed biofilm growth, metabolic activity, and virulence gene expression in mature cariogenic biofilms under conservative irradiation conditions. Given the minimal activity of the individual components relative to the combined treatment, exosome-assisted delivery may contribute to enhancing the photodynamic performance of resveratrol; however, as this study did not directly compare free resveratrol- and exosome-mediated aPDT under identical conditions, this interpretation should be regarded as preliminary. Further in vivo, clinical, and comparative studies are warranted to confirm its translational potential.

Introduction

Dental caries remains one of the most prevalent chronic oral diseases worldwide and continues to pose a significant public health burden despite advances in preventive and restorative dentistry. It is currently recognized as a biofilm-mediated and diet-modulated disease resulting from dysbiosis of the oral microbiome rather than infection by a single pathogen [1, 2]. Dental plaque is a structurally organized biofilm in which microbial communities are embedded within an extracellular polymeric substance (EPS) matrix that promotes adhesion, protection, and survival. Compared with planktonic microorganisms, biofilm-associated cells exhibit increased tolerance to antimicrobial agents and host defense mechanisms, thereby contributing to the initiation and progression of carious lesions [3–5].

Among cariogenic microorganisms, Streptococcus mutans is considered a major etiological agent owing to its acidogenicity, aciduricity, and biofilm-forming capacity [6]. A key virulence determinant of S. mutans is the gtfB gene, which encodes glucosyltransferases responsible for the synthesis of insoluble glucans that facilitate bacterial adherence and biofilm accumulation on tooth surfaces [7]. In addition, Lactobacillus acidophilus contributes to the progression of established carious lesions through sustained acid production and survival under acidic conditions. The surface-layer protein A (SlpA) has been identified as an important factor involved in adhesion, biofilm maturation, and maintenance of biofilm architecture in this species [8, 9].

Current caries management primarily relies on mechanical disruption of biofilms supplemented by antimicrobial agents such as chlorhexidine (CHX). Although CHX is widely used as a chemical plaque control agent, its long-term application is associated with several adverse effects, including tooth staining, taste alteration, and disturbances in oral microbial ecology [10, 11]. Consequently, the development of alternative antimicrobial approaches capable of effectively controlling cariogenic biofilms while minimizing side effects remains highly desirable.

Antimicrobial photodynamic therapy (aPDT) has emerged as a promising non-invasive strategy for the management of oral infections. This approach involves activation of a photosensitizer by light of an appropriate wavelength in the presence of oxygen, resulting in the generation of reactive oxygen species (ROS) that induce oxidative damage to microbial cells [12]. Owing to its broad-spectrum antimicrobial activity and low potential for inducing microbial resistance, aPDT has gained increasing attention as an adjunctive approach for controlling cariogenic biofilms [13]. However, the effectiveness of aPDT largely depends on the physicochemical properties, stability, and delivery efficiency of the photosensitizer employed.

Resveratrol (3,5,4′-trihydroxystilbene), a naturally occurring polyphenolic compound, possesses antimicrobial, antioxidant, anti-inflammatory, and anti-biofilm activities and has recently been proposed as a potential photosensitizer for aPDT [14–16]. Nevertheless, its clinical application is limited by poor aqueous solubility, low bioavailability, rapid metabolism, and susceptibility to degradation [17].

To overcome these limitations, an appropriate delivery strategy is required to improve the physicochemical stability and local availability of resveratrol during photodynamic treatment. Previous studies have emphasized that the poor aqueous solubility, limited chemical and photostability, and rapid metabolism of resveratrol can substantially restrict its effective delivery and biological performance, thereby supporting the development of encapsulation and nanocarrier-based approaches [17–19]. In the context of aPDT, this issue is particularly relevant because the photodynamic response depends on the effective availability of the photosensitizer at the microbial target during light activation [13, 16]. Thus, improving the formulation and delivery of resveratrol represents a rational strategy for enhancing its photodynamic performance rather than simply increasing the concentration of free compound.

Among the available delivery platforms, extracellular vesicles, particularly exosomes, have attracted considerable interest as naturally derived nanocarriers. Exosomes are nanoscale, membrane-enclosed extracellular vesicles generated through the endosomal pathway and released following fusion of multivesicular bodies with the plasma membrane. They are typically reported within a size range of approximately 30–150 nm and are surrounded by a lipid bilayer composed primarily of phospholipids and cholesterol, with membrane-associated proteins that contribute to vesicle stability, cellular interaction, and uptake [20–22]. Their molecular cargo is heterogeneous and may include membrane and cytosolic proteins, lipids, messenger RNAs, microRNAs, other non-coding RNAs, and, depending on the parental cell and culture conditions, additional bioactive molecules [20–22]. Importantly, exosomes do not have a single defined molecular weight because they are heterogeneous biological particles whose mass depends on their size, membrane composition, and molecular cargo; therefore, their physicochemical characterization is more appropriately described in terms of particle size, morphology, surface markers, and molecular composition [23, 24]. Their lipid-bilayer structure enables the incorporation and transport of a broad range of bioactive molecules, including hydrophobic compounds, while their endogenous biological origin may provide favorable biocompatibility and low immunogenicity compared with some conventional synthetic delivery systems [20–22]. Importantly, exosome-based delivery can potentially protect encapsulated cargo from premature degradation and modify its local availability and release characteristics. These properties provide a mechanistic rationale for using exosomes to overcome formulation-related limitations of resveratrol and to facilitate its availability during photoactivation.

The selection of human dental pulp stem cell-derived exosomes (hDPSC-Exos) was based on their specific biological relevance to the oral environment rather than on their use simply as a generic nanocarrier. Dental pulp stem cells are an accessible population of dental mesenchymal stem cells, and their extracellular vesicles have demonstrated regenerative, immunomodulatory, and tissue-supportive activities in oral and dental applications [25–29]. Their dental tissue origin therefore provides a biologically relevant platform for developing cell-free therapeutic approaches intended for dental applications. In addition, hDPSC-derived extracellular vesicles have been investigated as potential vehicles for bioactive molecules and as therapeutic mediators in oral tissue repair, supporting their suitability as a dental-context delivery platform [28, 29].

Importantly, the rationale for the present strategy is not that hDPSC-Exos are intrinsically required for antimicrobial activity or that they are superior to all other nanocarriers. Rather, they were selected as a biologically derived and dental tissue-relevant carrier with the potential to improve the delivery characteristics of resveratrol while maintaining compatibility with oral therapeutic applications. This distinction is important because the use of hDPSC-Exos for photodynamic delivery against cariogenic biofilms represents a relatively unexplored application. To the best of our knowledge, the use of hDPSC-Exos specifically for resveratrol delivery during aPDT against mature cariogenic biofilms has not previously been investigated. The present study therefore addresses a specific knowledge gap by integrating a dental stem cell-derived extracellular vesicle platform with a natural photosensitizer and blue LED-mediated aPDT.

Accordingly, the present study aimed to evaluate the anti-biofilm efficacy of aPDT based on resveratrol-loaded hDPSC-Exos (Res@hDPSC-Exos) against S. mutans and L. acidophilus in an ex vivo cariogenic biofilm model. Furthermore, the effects of this treatment on the expression of the biofilm-associated virulence genes gtfB and slpA were investigated. We hypothesized that exosome-mediated delivery of resveratrol would enhance the photodynamic efficacy of resveratrol against cariogenic biofilms, resulting in greater reductions in biofilm biomass and bacterial metabolic activity, together with downregulation of virulence-associated genes.

Materials and methods

Isolation and characterization of hDPSC-Exos

Human dental pulp stem cells (hDPSCs; IBRC C11306, National Center for Genetic and Biological Resources of Iran) were cultured in low-glucose Dulbecco’s Modified Eagle Medium (DMEM; Gibco, USA) supplemented with GlutaMAX and 10% exosome-depleted fetal bovine serum (FBS; Gibco, USA). After 3–5 days of culture, the conditioned medium was collected, filtered through a 0.22-µm syringe filters (Jet Biofil, Guangzhou, China), and subjected to exosome isolation using the EXOCIB™ Exosome Isolation Kit (SibBio, Iran) according to the manufacturer’s protocol [30]. Briefly, the conditioned medium was incubated with the precipitation reagent overnight at 4 °C, followed by centrifugation at 3000 rpm for 45 min. The resulting exosome pellet was resuspended in sterile phosphate-buffered saline (PBS). The protein concentration of the isolated exosome preparations was quantified using a colorimetric protein assay. For morphological characterization, exosome suspensions were prepared for transmission electron microscopy (TEM) and examined for vesicle morphology and membrane integrity using a TEM instrument (Zeiss EM10C, Germany) operated at an accelerating voltage of 100 kV. Particle size and size distribution were determined by dynamic light scattering (DLS) using a DLS instrument (Malvern Instruments Ltd., UK). Flow cytometry was performed to assess the expression of the exosomal surface marker CD63. Exosome preparations were stained with a fluorochrome-conjugated anti-CD63 antibody and analyzed using a flow cytometer with appropriate controls. Following characterization, purified exosome suspensions were aliquoted and stored at − 70 °C until further experimental use. A schematic diagram of the experimental steps is shown in Fig. 1.

Fig. 1.

A schematic diagram of the experimental steps

Preparation and characterization of Res@hDPSC-Exos

Resveratrol (purity ≥ 99%; Sigma-Aldrich, Germany) was incorporated into hDPSC-Exos at a weight-to-volume ratio of 1:5. The exosome–resveratrol suspension was incubated in the dark at 4 °C for 6–8 h to facilitate resveratrol association with the vesicles. To enhance loading efficiency, the mixture was subjected to a sonication-assisted loading procedure. Briefly, samples were placed in an ice bath and exposed to ultrasonic waves using a probe sonicator (SONOPULS, Bandelin, Germany; 30 kHz, 9 W, continuous mode for 10 s), thereby promoting transient permeabilization of the exosomal membrane and facilitating resveratrol encapsulation. Following loading, unencapsulated resveratrol was removed by dialysis using a membrane with a molecular weight cut-off of 12–14 kDa. The physicochemical properties of the resulting Res@hDPSC-Exos were subsequently evaluated using the same DLS, TEM, and flow cytometric procedures described in Sect. Isolation and characterization of hDPSC-Exos to assess particle size, vesicle morphology, and CD63 expression, respectively.

In vitro release profile of resveratrol from Res@hDPSC-Exos

The release behavior of resveratrol from Res@hDPSC-Exos was assessed using a dialysis-based method as previously described with minor modifications [31, 32]. Briefly, 4 mL of the exosome formulation was transferred into a dialysis membrane (12–14 kDa MWCO) and immersed in 10 mL PBS maintained at a constant temperature in a water bath. At predetermined time intervals, aliquots of the release medium were collected and replaced with an equal volume of fresh PBS to maintain sink conditions. The concentration of released resveratrol was quantified by UV–visible spectrophotometry. Absorbance measurements were recorded at 306 nm, corresponding to the characteristic absorption maximum of resveratrol and showing negligible interference from exosomal components. The cumulative release percentage was calculated according to the following equation:

Bacterial strains and growth conditions

S. mutans ATCC 35,668 and L. acidophilus ATCC 4356 (National Center for Genetic and Biological Resources of Iran) were used in this study. S. mutans was cultured on mitis salivarius agar (Merck, Germany), whereas L. acidophilus was grown on de Man–Rogosa–Sharpe (MRS) agar (Merck, Germany). All cultures were incubated for 24 h at 37 °C under 5% CO₂.

Determination of minimum biofilm reduction concentration (MBRC) of Res@hDPSC-Exos

The anti-biofilm efficacy of Res@hDPSC-Exos was evaluated using a microtiter plate-based assay. Briefly, 200 µL of each bacterial suspension separately (1 × 10⁸ CFU/mL) were inoculated into sterile 96-well plate and incubated for 24 h at 37 °C in 5% CO₂ to allow mature biofilm formation. Non-adherent cells were subsequently removed by washing the wells with PBS. Preformed biofilms were then exposed to serial concentrations of Res@hDPSC-Exos (31.2, 62.5, 125, 250, and 500 µg/mL) and incubated for an additional 24 h under the same conditions. Following treatment, the supernatant was carefully discarded, and the wells were gently rinsed twice with PBS. Biofilm cells were detached by resuspension in 100 µL PBS with vigorous pipetting. Serial dilutions were prepared, and 10 µL aliquots were plated onto brain heart infusion (BHI) agar (Merck, Germany). After 24 h of incubation at 37 °C in 5% CO₂, colony-forming units (CFUs) were counted. The MBRC was defined as the lowest concentration of Res@hDPSC-Exos that resulted in a statistically significant reduction in viable biofilm-associated bacteria compared with the untreated control [33, 34].

Determination of minimum biofilm reduction dose (MBRD) of blue LED light

To evaluate the light effect, mature biofilms were prepared as described above and washed with PBS to remove planktonic and loosely attached cells. The biofilms were then exposed to a blue LED light source (DY400-4, Denjoy Dental Co., Ltd., Shenzhen, China) with a wavelength of 450 ± 20 nm. Irradiation was performed at an output power density (irradiance) of 1 W/cm² for exposure times of 60, 120, 180, 240, and 300 s, corresponding to light fluences of 60, 120, 180, 240, and 300 J/cm², respectively. During irradiation, the light probe was maintained at a distance of 1 mm above the surface of each well, while a black paper sheet was placed underneath the microplate to minimize light reflection. To avoid potential cross-illumination between adjacent wells, empty wells were used as separators whenever feasible. Following irradiation, biofilm disruption, serial dilution, plating on BHI agar, and CFU enumeration were conducted as described above. The MBRD was defined as the minimum irradiation time required to achieve a statistically significant reduction in viable biofilm cells compared with the non-irradiated control group [35].

Formation of a microbial biofilm on the enamel surface

Sample selection

Human teeth extracted for therapeutic purposes, including orthodontic and periodontal indications, were included in this study. Only teeth free from caries, cracks, or fractures were selected. All specimens were used exclusively for research purposes and were not subjected to any procedures beyond the study protocol. Prior to analysis, all samples were fully anonymized to ensure that no donor-identifiable information could be traced.

Experimental groups

The experimental design incorporated appropriate component-specific and treatment controls to distinguish the individual and combined effects of resveratrol, hDPSC-Exos, and blue LED irradiation. The following groups were included:

• Group 1. Untreated control (without resveratrol, exosomes, or light irradiation)

• Group 2. Free resveratrol without exosomes and light irradiation

• Group 3. Empty hDPSC-Exos without resveratrol and light irradiation

• Group 4. Blue LED irradiation alone

• Group 5. aPDT using Res@hDPSC-Exos (1/2×MBRC) and light irradiation

• Group 6. aPDT using Res@hDPSC-Exos (1/4×MBRC) and light irradiation

Sample size calculation

The required sample size was determined using PASS software (version 21) based on a one-way ANOVA power calculation. Assuming α = 0.05, 95% statistical power (1 − β = 0.95), and an anticipated effect size of 1.7, the minimum required sample size was calculated as five independent specimens per group. Accordingly, with six experimental groups, 30 specimens were required for each bacterial species. Because S. mutans and L. acidophilus biofilms were established and evaluated separately, a total of 60 independent specimens were included when both bacterial models were considered. The specimens were randomly allocated to the experimental groups to minimize allocation bias.

Development of an in vitro enamel biofilm model

Enamel slabs (3 × 3 × 1 mm) were prepared from human tooth crowns containing sound, flat enamel surfaces [36]. To simulate acquired pellicle formation, the specimens were immersed in filtered human saliva (passed through a 0.22-µm syringe filter) for 30 min, promoting surface conditioning and enhancing bacterial adhesion. Following pellicle formation, the samples were gently rinsed with PBS and placed into wells of a sterile 48-well plates containing 1 mL of tryptone–yeast extract broth (Merck, Germany) supplemented with 1% sucrose (Sigma-Aldrich, Germany). A bacterial suspension of either S. mutans or L. acidophilus, each at a concentration of 1.5 × 10⁶ CFU/mL, was separately added to the wells to initiate biofilm formation and the plates were then incubated for 8 h at 37 °C under 5% CO₂. Subsequently, the specimens were transferred to fresh medium containing 1 mM glucose (Sigma-Aldrich, Germany) and further incubated for 24 h. During the subsequent maturation phase, biofilms were exposed to 10% sucrose solution eight times daily, each exposure lasting 3 min, for a total period of 96 h to simulate cariogenic dietary conditions [36].

Evaluation of biofilm biomass by crystal violet assay

Following treatment of mature biofilms formed on the enamel specimens with the treatment groups, samples were gently rinsed with PBS to remove non-adherent biomass. Biofilm mass was quantified using a crystal violet staining assay [37]. Briefly, specimens were incubated with 0.1% crystal violet solution for 15 min at room temperature, followed by thorough washing with PBS to eliminate unbound dye. The retained stain was solubilized using 80% ethanol, and absorbance was measured at 595 nm using a microplate spectrophotometer. Biofilm biomass was expressed as a percentage relative to untreated controls.

Evaluation of metabolic activity using XTT reduction assay

Following treatment of mature biofilms formed on the enamel specimens with the study groups, the metabolic activity was assessed using the XTT reduction assay [38]. Following disruption of biofilms by sonication (30 kHz, 9 W, continuous mode for 10 s), samples were incubated with XTT reagent at 37 °C in the dark for 3 h. Enzymatic reduction of XTT to water-soluble formazan was quantified spectrophotometrically at 492 nm, and metabolic activity was expressed as a percentage relative to the control group.

Evaluation of biofilm-associated gene expression by quantitative real-time polymerase chain reaction (qRT-PCR)

The impact of treatment groups on bacterial virulence was evaluated by quantifying the expression levels of gtfB in S. mutans and slpA in L. acidophilus. Following treatment under sub-MBRC (1/2×MBRC and 1/4×MBRC) and sub-MBRD conditions, total RNAs were extracted using a commercial RNA isolation kit (AnaCell, Iran) according to the manufacturer’s protocol. The concentration and purity of the isolated RNAs were assessed using a NanoDrop^® ND-1000 spectrophotometer (Thermo Fisher Scientific, USA). Complementary DNA (cDNA) was synthesized from the purified RNA using a reverse transcription kit (DENAzist Asia Biotechnology, Iran) following the manufacturer’s instructions.

Specific primers were designed using Primer3Plus software (v.4.1.0; http://bioinfo.ut.ee/primer3/), and their sequences are provided in Table 1. qRT-PCR reactions were carried out in a final volume of 20 µL, comprising 1 µL of cDNA, 0.5 µM of each primer, 10 µL of SYBR™ Green PCR Master Mix (Zist Virayesh, Iran), and nuclease-free water. Thermal cycling was performed with an initial denaturation at 95 °C for 5 min, followed by 35 amplification cycles consisting of denaturation at 95 °C for 10 s, annealing at 60 °C for 15 s, and extension at 72 °C for 15 s. Gene expression changes were quantified using the comparative Ct (2^−ΔΔCt) method [39], whereby relative expression levels of target genes were normalized to the 16S rRNA housekeeping gene and expressed as fold changes compared with the untreated control group.

Table 1.

Primers sequence for studied genes

Gene	Primers	Sequence (5’-3’)	Amplicon size (base pair)
gtfB	Forward	TGTTGTTACTGCTAATGAAGAA	103
	Reverse	GCTACTGATTGTCGTTACTG
slpA	Forward	ACTGCTAAC AAC ACTCCAGC	153
	Reverse	TTC AGC AGG TTTAACGGC AG
16S rRNA	Forward	CCTACG GGAGGC AGC AGTAG	121
	Reverse	CAACAG AGC TTTACG ATCCGAAA

Statistical analysis

All experiments were performed in triplicate, and the resulting data were analyzed using SPSS software (version 25.0; SPSS Inc., Chicago, IL, USA). Differences among groups were assessed using one-way analysis of variance (ANOVA). A P value < 0.05 was considered statistically significant.

Results

Characterization of hDPSC-Exos and Res@hDPSC-Exos

The successful isolation of hDPSC-Exos was confirmed by TEM, DLS, and flow cytometric analyses (Fig. 2). TEM images revealed that hDPSC-Exos exhibited the characteristic spherical morphology with intact membrane structures (Fig. 2a). Following resveratrol loading, Res@hDPSC-Exos retained their vesicular architecture without noticeable aggregation or structural damage, indicating that the loading procedure preserved exosomal integrity (Fig. 2b).

Fig. 2.

Characterization of hDPSC-derived exosomes (hDPSC-Exos) and resveratrol-loaded exosomes (Res@hDPSC-Exos): (a) TEM image of hDPSC-Exos, (b) TEM image of Res@hDPSC-Exos, (c) Flow cytometric analysis of CD63 expression in hDPSC-Exos. (d) Flow cytometric analysis of CD63 expression in Res@hDPSC-Exos, (e) DLS profile of hDPSC-Exos, and (f) DLS profile of Res@hDPSC-Exos

Flow cytometric analysis demonstrated positive expression of the exosomal marker CD63 in both formulations. The percentage of CD63-positive vesicles was 71.5% for hDPSC-Exos and 52.3% for Res@hDPSC-Exos (Fig. 2c and d), confirming the preservation of exosomal surface characteristics after resveratrol encapsulation.

DLS measurements revealed that hDPSC-Exos possessed a mean hydrodynamic diameter of approximately 145 nm (Fig. 2e). After resveratrol encapsulation, the average particle size increased moderately to approximately 160 nm (Fig. 2f), reflecting successful drug loading within the exosomal nanocarriers. Importantly, both formulations maintained a narrow size distribution within the nanoscale range, indicating preservation of colloidal stability following the loading process. These findings confirmed the successful isolation of hDPSC-Exos and demonstrated that resveratrol loading did not markedly affect exosomal morphology, size distribution, or CD63 expression.

Release profile of resveratrol from Res@hDPSC-Exos

The concentration of free resveratrol in the sample was determined using a calibration curve established from spectrophotometric analysis (y = 0.0014x + 0.1866, R² = 0.9617) (Fig. 3). Based on the measured OD value of 0.63 for the unknown sample, the concentration of free resveratrol was calculated to be 316.7 µg/mL. Considering an initial total resveratrol concentration of 500 µg/mL in the system, the encapsulation (or loading) efficiency was determined to be 63.3%. These findings indicate a moderate encapsulation performance of the developed exosomal system under the applied experimental conditions.

Fig. 3.

Standard calibration curve of resveratrol concentration measured by UV-Vis spectrophotometry, with signal intensity representing absorbance at 306 nm wavelength

MBRC of Res@hDPSC-Exos

The anti-biofilm activity of Res@hDPSC-Exos against mature biofilms of S. mutans and L. acidophilus is presented in Table 2. For S. mutans, treatment with 31.2, 62.5, and 125 µg/mL Res@hDPSC-Exos resulted in reductions of 6.5%, 12.3%, and 23.1% in viable biofilm-associated bacteria, respectively, with no statistically significant differences compared with the untreated control (P > 0.05). A significant reduction was first observed at 250 µg/mL, which decreased the viable bacterial count by 69.6% (P = 0.002), while treatment with 500 µg/mL further reduced bacterial viability by 91.8% (P < 0.001). Accordingly, the MBRC of Res@hDPSC-Exos against S. mutans was determined to be 250 µg/mL. In contrast, L. acidophilus exhibited greater susceptibility to Res@hDPSC-Exos. Although no significant reductions were detected at 31.2 and 62.5 µg/mL (P > 0.05), treatment with 125 µg/mL significantly reduced viable biofilm-associated bacteria by 53.5% compared with the untreated control (P = 0.004). Increasing the concentration to 250 and 500 µg/mL resulted in further reductions of 71.7% and 88.2%, respectively (P < 0.001). Therefore, the MBRC of Res@hDPSC-Exos against L. acidophilus was established at 125 µg/mL.

Table 2.

Anti-biofilm activity of Res@hDPSC-Exos against mature bacterial biofilms

Microorganism	Treatment	Minimum	Maximum	Mean ± SD (CFU/mL)	Reduction (%)	P value
S. mutans	Control	9.41 × 10⁵	1.02 × 10⁶	9.77 ± 0.40 × 10⁵	—	—
	31.2 µg/mL	8.86 × 10⁵	9.34 × 10⁵	9.13 ± 0.24 × 10⁵	6.5	0.446
	62.5 µg/mL	8.26 × 10⁵	8.89 × 10⁵	8.57 ± 0.32 × 10⁵	12.3	0.173
	125 µg/mL	7.18 × 10⁵	7.74 × 10⁵	7.51 ± 0.29 × 10⁵	23.1	0.061
	250 µg/mL	2.74 × 10⁵	3.25 × 10⁵	2.97 ± 0.26 × 10⁵	69.6	0.002
	500 µg/mL	6.70 × 10⁴	9.20 × 10⁴	8.00 ± 1.26 × 10⁴	91.8	< 0.001
L. acidophilus	Control	9.26 × 10⁵	9.74 × 10⁵	9.50 ± 0.24 × 10⁵	—	—
	31.2 µg/mL	8.67 × 10⁵	9.02 × 10⁵	8.86 ± 0.18 × 10⁵	6.7	0.398
	62.5 µg/mL	8.18 × 10⁵	8.56 × 10⁵	8.39 ± 0.19 × 10⁵	11.7	0.146
	125 µg/mL	4.26 × 10⁵	4.61 × 10⁵	4.42 ± 0.18 × 10⁵	53.5	0.004
	250 µg/mL	2.46 × 10⁵	2.89 × 10⁵	2.69 ± 0.22 × 10⁵	71.7	< 0.001
	500 µg/mL	9.70 × 10⁴	1.26 × 10⁵	1.12 ± 0.15 × 10⁵	88.2	< 0.001

CFU colony-forming units, SD standard deviation

MBRD of blue LED light

The anti-biofilm activity of blue LED irradiation against mature biofilms of S. mutans and L. acidophilus is summarized in Table 3. For S. mutans, irradiation for 60, 120, 180, and 240 s resulted in gradual reductions in viable biofilm-associated bacteria by 6.7%, 12.2%, 20.8%, and 31.2%, respectively, with no statistically significant differences compared with the non-irradiated control (P > 0.05). A significant reduction was first achieved after 300 s of irradiation, reducing the viable bacterial count by 52.3% relative to the control (P = 0.002). Therefore, the MBRD for S. mutans was determined to be 300 s.

Table 3.

Anti-biofilm activity of blue LED irradiation against mature bacterial biofilms

Microorganism	Irradiation time (s)	Minimum	Maximum	Mean ± SD (CFU/mL)	Reduction (%)	P value
S. mutans	Control	9.39 × 10⁵	1.01 × 10⁶	9.74 ± 0.36 × 10⁵	—	—
	60	8.95 × 10⁵	9.22 × 10⁵	9.09 ± 0.14 × 10⁵	6.7	0.481
	120	8.36 × 10⁵	8.71 × 10⁵	8.55 ± 0.18 × 10⁵	12.2	0.196
	180	7.48 × 10⁵	7.96 × 10⁵	7.71 ± 0.24 × 10⁵	20.8	0.082
	240	6.52 × 10⁵	6.86 × 10⁵	6.70 ± 0.17 × 10⁵	31.2	0.058
	300	4.48 × 10⁵	4.86 × 10⁵	4.65 ± 0.19 × 10⁵	52.3	0.002
L. acidophilus	Control	9.43 × 10⁵	9.91 × 10⁵	9.67 ± 0.24 × 10⁵	—	—
	60	8.97 × 10⁵	9.28 × 10⁵	9.13 ± 0.16 × 10⁵	5.6	0.501
	120	8.56 × 10⁵	8.84 × 10⁵	8.70 ± 0.14 × 10⁵	10.0	0.223
	180	7.79 × 10⁵	8.26 × 10⁵	8.04 ± 0.24 × 10⁵	16.9	0.097
	240	6.88 × 10⁵	7.29 × 10⁵	7.08 ± 0.21 × 10⁵	26.8	0.061
	300	4.97 × 10⁵	5.22 × 10⁵	5.07 ± 0.14 × 10⁵	47.6	0.003

CFU colony-forming units, SD standard deviation

A similar time-dependent trend was observed for L. acidophilus. Irradiation for up to 240 s produced progressive but non-significant reductions in viable biofilm cells (5.6–26.8%; P > 0.05). In contrast, irradiation for 300 s significantly decreased the viable bacterial count by 47.6% compared with the untreated control (P = 0.003), establishing 300 s as the MBRD for L. acidophilus.

Anti-biofilm effect of aPDT

The effect of treatment groups on biofilm biomass was evaluated by the crystal violet assay (Fig. 4a). For S. mutans, free resveratrol, empty hDPSC-Exos, blue LED alone produced only modest, non-significant reductions in biofilm biomass (approximately 15%, 6%, and 7%, respectively). In contrast, 62.5 µg/mL Res@hDPSC-Exos (1/4×MBRC) combined with 240 s blue LED irradiation significantly reduced biofilm biomass by 62.3% (P = 0.002), while 125 µg/mL (1/2×MBRC) achieved an 84.7% reduction (P < 0.001).

Fig. 4.

Effect of treatment groups on: (a) Biofilm biomass determined by crystal violet assay, (b) Biofilm metabolic activity determined by the XTT reduction assay, (c) Relative expression levels of the gtfB and slpA genes in the treated groups compared with the untreated control group. The biofilm biomass, biofilm metabolic activity, and the expression levels in the control group were normalized to 1.0. The treatment resulted in significant reductions compared to the untreated control group (*P < 0.05)

Similarly, in L. acidophilus, the individual treatments produced limited, non-significant reductions of approximately 14%, 5%, and 7%, respectively. The combination of 31.2 µg/mL Res@hDPSC-Exos (1/4×MBRC) with 240 s irradiation reduced biofilm biomass by 58.6% (P = 0.002), whereas 62.5 µg/mL (1/2×MBRC) resulted in a 75.2% reduction (P < 0.001). Overall, increasing Res@hDPSC-Exos from 1/4×MBRC to 1/2×MBRC significantly enhanced the anti-biofilm efficacy of aPDT in both species, with the greatest reductions observed at 1/2×MBRC + 240 s (84.7% for S. mutans and 75.2% for L. acidophilus).

Anti-metabolic activity of aPDT

The metabolic activity of mature biofilms following treatment is presented in Fig. 4b. In S. mutans, free resveratrol, empty hDPSC-Exos, blue LED alone produced only modest reductions in metabolic activity (approximately 15%, 7%, and 8%, respectively), with no statistically significant differences from the untreated control (P > 0.05). In contrast, aPDT with 62.5 µg/mL Res@hDPSC-Exos (1/4×MBRC) plus 240 s blue LED irradiation significantly reduced metabolic activity by 68.7% (P = 0.001). Increasing the concentration to 125 µg/mL (1/2×MBRC) further enhanced the effect, resulting in an 88.9% reduction (P < 0.001).

A similar trend was observed for L. acidophilus. Free resveratrol, empty hDPSC-Exos, blue LED alone produced minor reductions of approximately 13%, 5%, and 7%, respectively, without significant differences from the control (P > 0.05). Treatment with 31.2 µg/mL Res@hDPSC-Exos (1/4×MBRC) plus 240 s irradiation reduced metabolic activity by 67.0% (P = 0.001), while 62.5 µg/mL (1/2×MBRC) achieved an 85.0% reduction (P < 0.001).

One-way ANOVA demonstrated significant differences among groups for both S. mutans and L. acidophilus (P < 0.001). Overall, the limited effects of the individual components compared with the pronounced reductions following Res@hDPSC-Exos-mediated aPDT support the enhanced anti-biofilm efficacy of the combined treatment, with the 1/2×MBRC concentration showing the greatest suppression of metabolic activity in both species.

Anti-virulence effect of aPDT

The effect of aPDT on biofilm-associated virulence gene transcription was assessed by qRT-PCR using the comparative 2^−ΔΔCt method (Fig. 4c). Free resveratrol, empty hDPSC-Exos, blue LED alone (240 s) did not significantly affect gtfB expression in S. mutans or slpA expression in L. acidophilus compared with the untreated control (P > 0.05).

In contrast, aPDT using sub-MBRC concentrations of Res@hDPSC-Exos combined with sub-MBRD blue LED irradiation significantly suppressed both virulence genes. In S. mutans, 62.5 µg/mL (1/4×MBRC) and 125 µg/mL (1/2×MBRC) Res@hDPSC-Exos reduced gtfB expression to 0.42-fold (P = 0.003) and 0.15-fold (P < 0.001), respectively. Similarly, in L. acidophilus, 31.2 µg/mL (1/4×MBRC) and 62.5 µg/mL (1/2×MBRC) reduced slpA expression to 0.47-fold (P = 0.004) and 0.19-fold (P < 0.001), respectively.

Thus, while the individual components had negligible transcriptional effects, their combination in aPDT produced a pronounced, concentration-dependent downregulation of biofilm-associated virulence genes, consistent with the observed reductions in biofilm biomass and metabolic activity.

Discussion

The present study demonstrated that aPDT mediated by Res@hDPSC-Exos combined with blue LED irradiation effectively inhibited mature cariogenic biofilms formed by S. mutans and L. acidophilus. The combined treatment significantly reduced biofilm biomass, metabolic activity, and the expression of the major virulence genes gtfB and slpA. Importantly, these effects were achieved using sub-MBRC concentrations of Res@hDPSC-Exos together with sub-MBRD irradiation, indicating that exosome-mediated delivery enhanced the photodynamic efficacy of resveratrol while reducing the required photosensitizer concentration and light dose [40–44].

To the best of our knowledge, this is among the first studies to evaluate hDPSC-Exos as biological nanocarriers for resveratrol delivery during aPDT against cariogenic biofilms. Unlike synthetic nanocarriers, exosomes exhibit excellent biocompatibility, low immunogenicity, and efficient transport of hydrophobic compounds, making them attractive delivery vehicles for oral photodynamic applications [21, 40, 45]. The specific selection of hDPSC-Exos also warrants consideration. The purpose of using dental pulp stem cell-derived vesicles was not merely to provide a nanoscale carrier, but to employ a biologically derived platform with direct relevance to the oral and dental environment. Dental stem cell-derived extracellular vesicles have been increasingly investigated as cell-free therapeutic mediators because they can participate in tissue repair, immunomodulation, angiogenesis, and odontogenic processes [25–29]. This biological context distinguishes hDPSC-Exos from a purely synthetic delivery vehicle and provides a conceptual basis for their investigation in dental applications. In the present study, this dental relevance is particularly attractive because an anti-caries intervention should ideally combine effective biofilm control with a delivery platform that is biologically compatible with oral tissues. Nevertheless, our findings should not be interpreted as demonstrating that hDPSC-Exos are universally superior to other nanocarriers. Rather, they provide proof-of-concept that a dental stem cell-derived extracellular vesicle platform can be used to formulate resveratrol and support its application in blue LED-mediated aPDT against mature cariogenic biofilms. Physicochemical characterization further confirmed successful encapsulation while preserving exosomal integrity. Their nanoscale size and acceptable encapsulation efficiency may facilitate diffusion within the biofilm matrix, improve local stability of resveratrol, and enhance its availability during photoactivation [14, 21, 22, 46, 47].

Importantly, the limited effects observed with free resveratrol, empty hDPSC-Exos, and blue LED irradiation alone further support the enhanced efficacy of the combined treatment. Free resveratrol produced only modest, non-significant reductions in biofilm biomass and metabolic activity, although previous studies have reported concentration-dependent anti-biofilm activity of resveratrol against S. mutans [48]. Similarly, empty hDPSC-Exos showed minimal effects, suggesting that their primary contribution was likely related to resveratrol delivery rather than direct antibacterial activity. Blue LED irradiation alone also produced only minor reductions, consistent with previous evidence that the antimicrobial activity of blue light depends on wavelength, fluence, exposure time, and biofilm characteristics [49–51]. Collectively, these findings indicate that the pronounced anti-biofilm activity observed with Res@hDPSC-Exos-mediated aPDT was mainly associated with the combined treatment rather than the individual components.

Consistent with previous studies, L. acidophilus exhibited greater susceptibility than S. mutans, probably because the glucan-rich EPS produced by S. mutans limits penetration of antimicrobial agents and contributes to greater biofilm tolerance [41, 43, 52]. Nevertheless, Res@hDPSC-Exos-mediated aPDT significantly reduced both biofilm biomass and metabolic activity, indicating disruption of mature biofilm architecture and bacterial physiological activity. These findings agree with previous reports showing that aPDT effectively impairs oral biofilms through oxidative damage to bacterial cells and the surrounding biofilm matrix [3, 4, 40–42, 52–54].

One of the most important findings of this study was the significant downregulation of gtfB in S. mutans and slpA in L. acidophilus. Since these genes are closely associated with glucan synthesis, adhesion, and biofilm persistence, their suppression suggests attenuation of bacterial virulence in addition to reduced bacterial viability [6–9]. Similar modulation of virulence-associated genes following aPDT has been reported previously [21, 40, 53–56], supporting the potential of exosome-assisted resveratrol delivery to interfere with key pathogenic mechanisms involved in cariogenic biofilm development.

The pronounced antimicrobial efficacy observed when blue LED irradiation was combined with Res@hDPSC-Exos, despite the limited effect of blue LED alone, is consistent with the established principles of aPDT, whereby light activation of a photosensitizing agent promotes the generation of ROS, thereby enhancing microbial killing and contributing to the disruption of biofilm structure and viability [40–42]. Although ROS generation was not directly quantified in the present study, the marked reductions in biofilm biomass, metabolic activity, and virulence gene expression are consistent with the proposed photodynamic mechanism. Nevertheless, the lack of direct ROS measurement limits definitive attribution of the observed effects to ROS-mediated photodynamic activity, as intrinsic antimicrobial effects of resveratrol and/or exosomes may also contribute. Future studies should directly quantify ROS generation to further confirm the proposed mechanism. Blue LED (~ 450 nm) represents a suitable light source for dental aPDT because of its compatibility with natural photosensitizers, low thermal effects, and widespread clinical availability. In addition, blue wavelengths may activate endogenous bacterial chromophores, potentially enhancing the overall photodynamic response [52, 53, 57–59].

The improved therapeutic performance of Res@hDPSC-Exos may be related to the ability of extracellular vesicles to provide a protective lipid-bilayer environment for hydrophobic cargo and potentially modify its local availability during photoactivation [20–22]. In the present study, successful resveratrol loading and preservation of vesicular characteristics were demonstrated, while the marked anti-biofilm effects were observed under sub-MBRC and sub-MBRD conditions. These findings are consistent with the interpretation that exosome-assisted formulation improved the functional availability of resveratrol during aPDT. However, because intracellular ROS production, direct biofilm penetration of the vesicles, and comparative performance against alternative nanocarriers were not directly evaluated, these mechanisms should be considered plausible explanations rather than experimentally confirmed pathways. Further studies incorporating fluorescent tracking, ROS measurements, and head-to-head comparisons with synthetic nanocarriers will be required to establish the precise contribution of exosome-mediated delivery.

From a clinical perspective, achieving significant anti-biofilm activity using sub-MBRC concentrations of the photosensitizer together with sub-MBRD irradiation is particularly encouraging. Reducing both photosensitizer concentration and irradiation dose may shorten treatment time while minimizing unnecessary light exposure, supporting the potential application of this approach as a minimally invasive adjunctive strategy for caries management [12, 40–42, 52].

The present study also has several strengths. Unlike conventional planktonic models, antimicrobial efficacy was evaluated in mature biofilms formed on human enamel specimens, providing greater biological relevance. Moreover, structural, metabolic, and molecular outcomes were assessed simultaneously, allowing a more comprehensive evaluation of treatment efficacy. Nevertheless, several limitations should be acknowledged. The study was limited to a dual-species ex vivo biofilm model, only a single irradiation protocol was investigated, and in vivo validation was not performed. Future studies should therefore investigate multispecies biofilms, optimize photodynamic parameters, quantify ROS generation, and confirm therapeutic efficacy in animal models and clinical trials.

Conclusion

The combination of Res@hDPSC-Exos with blue LED irradiation produced a synergistic anti-biofilm effect against mature S. mutans and L. acidophilus biofilms that clearly exceeded the sum of its individual components: free resveratrol, empty hDPSC-Exos, and light exposure alone each failed to produce meaningful reductions in biofilm biomass, metabolic activity, or virulence gene transcription, whereas their combination as Res@hDPSC-Exos-mediated aPDT achieved substantial suppression of gtfB and slpA expression alongside marked declines in biofilm viability, even at sub-inhibitory photosensitizer concentrations and conservative light doses.

This pattern suggests a functional contribution of the exosomal carrier to the photodynamic process, most plausibly by protecting resveratrol from premature degradation and improving its local availability at the biofilm–light interface. However, because free resveratrol was tested only in the absence of light rather than as a light-activated comparator, the present design cannot isolate the exosome’s specific contribution from that of resveratrol itself, and this distinction remains an open question for future work rather than a settled conclusion.

From a translational standpoint, achieving pronounced anti-biofilm and antivirulence effects at reduced photosensitizer concentrations and shorter irradiation times supports the feasibility of this platform as a conservative, tissue-compatible adjunct to caries control. Confirming this potential will require studies directly comparing free resveratrol- and exosome-mediated aPDT under matched conditions, quantifying reactive oxygen species generation, and extending evaluation to multispecies and in vivo models.

Author contributions

MP: Conceptualized the study, designed the methodology, conducted the laboratory experiments, collected and analyzed data, drafted the manuscript, and critically reviewed the manuscript.AB: Conceptualized the study, designed the methodology, assisted in experiment execution, and critically reviewed the manuscript.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.