Synthesis and characterization of CGP

To fabricate CGP, the initial step involved the guanidinylation of NH2-PEG-PCL using 1-Amidinopyrazole hydrochloride to form G-PEG-PCL [35]. This modified polymer serves as an effective carrier for the photosensitizer Ce6, forming CGP. Notably, G-PEG-PCL enhances the stability of Ce6 in physiological environments, boosts its association with bacteria, aids in biofilm penetration, and additionally functions as a nitric oxide donor. The successful guanidinylation was confirmed by the emergence of of 1H nuclear magnetic resonance peak at 3.41 ppm (–N–CH2–C–O–). Other characteristic peaks at 3.50 ppm (–CH2CH2O–) and 2.27 ppm (–O–CO–CH2–) were identified, corresponding to the PEG and PCL components of the compound, as depicted in Additional file 1: Fig. S1.

Figure 1a illustrates the preparation of CGP nanoparticles. The encapsulation efficiency of Ce6 in CGP was found to be 97.02%, with a loading capacity of 5.57% (Additional file 1: Table S1). As a control, CPP was simultaneously prepared using NH2-PEG-PCL loaded with Ce6. The particle sizes and transmission electron microscope images of CPP and CGP exhibited uniform spherical shapes, with sizes of 172.7 ± 5.2 nm and 166.6 ± 7.3 nm (Fig. 1b–e). In aqueous solution, CPP and CGP displayed a distinct green color, contrast to free Ce6, which exhibited poor water solubility, resulting in aggregation at the bottom of the Eppendorf tube (Fig. 1f). These results confirm the successful encapsulation of lipophilic Ce6 by NH2-PEG-PCL and G-PEG-PCL, enabling homogeneous dispersion in aqueous solution.

Fig. 1

Fabrication and characterization of CGP nanoparticles. a Schematic illustration for the preparation of CGP nanoparticles. b The particle size distribution and c transmission electron microscope images of CPP. d The particle size distribution and e transmission electron microscope images of CGP. f Solubility of Ce6 in different groups. g Zeta potential of each group. h Stability of CGP during 14 d. i Free Ce6, CPP, and CGP internalized by or bound tightly to E. faecalis and j the quantitative fluorescence intensity of Ce6 in E. faecalis. k Confocal laser scanning microscope images of E. faecalis biofilms treated by various groups. Data are presented as mean ± SEM, n = 3, *p < 0.05, **p ≤ 0.01, ****p ≤ 0.0001. Comparison between CGP versus other groups

The ζ-potential of PP was 7.04 mV, while that of GP was 8.72 mV (Fig. 1g). The increased ζ-potential in GP relative to PP may be due to the elevated pKa value of the guanidino group, as opposed to the amino group [35, 36]. Furthermore, CGP demonstrated good stability over 14 days at room temperature, as evidenced by the consistent aqueous solution color (Additional file 1: Fig. S2), particle size, and dispersibility index (Fig. 1h). Additionally, the ultraviolet absorption and fluorescence emission spectra of CGP were found to be in close alignment with those of free Ce6 (Additional file 1: Figs. S3 and S4), suggesting that the physical encapsulation of Ce6 did not alter its fundamental absorption and emission characteristics.

Bacterial association and biofilm permeation of CGP

Flow cytometry analysis was utilized to investigate the impact of guanidino groups on the association of CGP with bacteria. Employing Ce6 as a fluorescent indicator, it was observed that CGP exhibited a notably higher degree of internalization or adhesion to E. faecalis bacteria, in comparison to CPP (Fig. 1i). The quantitative fluorescence data further substantiated this observation, revealing that E. faecaliss treated with CGP displayed fluorescence intensities approximately 1.61-fold higher than those treated with CPP (Fig. 1j). This finding indicates that guanidino groups in CGP markedly augments its bacterial interaction capabilities [37]. Such an enhancement in bacterial association is pivotal, as it substantially bolsters the bactericidal efficacy of CGP, making it a more potent antimicrobial agent.

The efficacy of antimicrobial agents is significantly influenced by their ability to penetrate biofilms, which are known to hinder the therapeutic effectiveness of such agents [38]. The biofilm permeability of CGP was assessed using a CLSM, employing SYTO-9 as an indicator for biofilm staining. After 5 min of co-incubation with E. faecalis biofilm, CLSM analysis showed a faint red fluorescence emanating from Ce6 within the biofilm treated with free Ce6, suggesting limited penetration. This finding underscores the challenge posed by biofilms in obstructing the entry of antibacterial substances. In contrast, a modestly stronger fluorescence signal was observed in the CPP-treated biofilm, implying a slight improvement in biofilm permeability. Remarkably, the CGP-treated biofilm exhibited a significantly intensified fluorescence signal (Fig. 1k). These results suggest that while enhancing Ce6’s solubility and stability, as well as incorporating the amino group in CPP, may slightly improve biofilm permeability, the introduction of guanidinylation markedly enhances CGP's capacity to penetrate biofilms. This enhanced permeability is a key factor in boosting the antimicrobial efficacy of CGP against biofilm-protected bacteria.

ROS and NO generation driven by aPDT

After confirming increased bacterial association and improved biofilm penetration by guanidinylated CGP, we employed the Griess assay to evaluate its NO production capabilities in response to aPDT. As shown in Fig. 2a, PBS + Laser, CGP, Ce6 + Laser, and CPP + Laser displayed negligible NO production. In contrast, CGP + Laser exhibited a rapid production of 4.33 μM NO within 10 min. These findings suggest that the capability of CGP nanoparticles to produce NO is likely attributed to the presence of guanidino groups and their activation by aPDT. To further validate the hypothesis that CGP consumes H2O2 and subsequently produces NO during the aPDT process, we employed an H2O2 assay kit to detect the generation of H2O2. The assay operates through the oxidation of divalent iron ions by H2O2 to produce trivalent iron ions, which combine with xylenol orange to form a purple product [39]. As shown in Fig. 2b, CPP + Laser, and CGP + Laser generated 25.01 and 15.73 μM H2O2, following irradiation for 10 min. Notably, the H2O2 generation of CGP + Laser (synchronized with NO generation) was obviously lower than that of CPP + Laser. implying the consumption of H2O2 in the course of NO generation. This finding suggests that H2O2 is being consumed during the aPDT process in CGP + Laser group, to facilitate the generation of NO.

Fig. 2

NO and ROS generation profiles of CGP. a NO, b H2O2, and c 1O2 generation profiles of each group. d, e NO, f, g ROS, and h, i H2O2 generation and mean fluorescence intensity in E. faecalis bacteria after receiving various treatments. Data are presented as mean ± SEM, n = 3, **p ≤ 0.01, ****p ≤ 0.0001, ns: no significance. Comparison between CGP + Laser versus other groups

It has been reported that 1O2 also oxidizes the guanidino groups to produce NO [17]. Given that 1O2 is a pivotal factor in the bactericidal effectiveness of aPDT [40], we anticipated that its production would not be compromised. To investigate this, we employed SOSG to monitor the changes in 1O2 levels within CGP under aPDT. SOSG possesses faint blue fluorescence but reacts with 1O2 to produce SOSG endoperoxide, which emits strong green fluorescence [41]. As shown in Fig. 2c, the measured fluorescence intensities of SOSG endoperoxide for both CPP + Laser and CGP + Laser exhibited a similar increase, indicating that CGP can rapidly generate NO by consuming H2O2-produced during aPDT, without hindering the production of 1O2. This is crucial for the synergistic antimicrobial and biofilm eradication efficacy of aPDT and NO.

To corroborate the aforementioned results, we employed CLSM to observe the production of NO, ROS, and H2O2 by CGP in bacteria. The production of NO by CGP in bacteria was detected using DAF-FM DA, a compound that traverses the cell membrane and subsequently undergoes catalysis by intracellular esterase to form DAF-FM, which is incapable of crossing the cell membrane [42]. DAF-FM exhibits weak fluorescence, but it exhibits strong fluorescence upon reacting with NO. As shown in Fig. 2d and e, a notable green fluorescence signal indicative of NO was observed in E. faecalis bacteria treated with CGP + Laser, while only negligible fluorescence was noted in bacteria treated with Ce6 + Laser and CPP + Laser. To detect ROS produced by CGP in bacteria, we used DCFH-DA, a non-fluorescent molecule that and can freely traverse the cell membrane and, once inside, is converted into DCFH by intracellular esterase. ROS within the cell can oxidize nonfluorescent DCFH, transforming it into the fluorescent compound DCF [43]. As shown in Fig. 2f and g, the green fluorescence indicative of ROS in E. faecalis treated with CPP + Laser was slightly higher than in CGP + Laser treated bacteria, in line with the results obtained in the solution phase (Additional file 1: Fig. S5). H2O2 produced by CGP in bacteria was detected through the utilization of Amplex Red and horseradish peroxidase [44]. Under the action of horseradish peroxidase, H2O2 oxidizes Amplex red, resulting in the production of the red fluorescent [45]. Figure 2h and i demonstrate a diminished red fluorescence of H2O2 in CGP + Laser-treated bacteria in comparison to CPP + Laser. These results further support the possibility that the production of NO actually consumes a part of the ROS and H2O2 generated during aPDT.

Antimicrobial effect of CGP in vitro

After confirming that the generation of NO driven by aPDT did not impede the generation of 1O2, the synergistic antimicrobial effects of aPDT and NO were assessed through bacterial colony counting and bacterial live/dead staining. Figure 3a and b displays that the bacterial viability in CPP and CGP without irradiation was higher than 8 log units, which is consistent with that of PBS + Laser. However, a notable reduction in bacterial viability was observed among the remaining groups exposed to laser irradiation: Ce6 + Laser, CPP + Laser, and CGP + Laser, with viability counts of 3.84, 3.63, and 0.77 log units, respectively. This significant decrease in bacterial viability, especially in the CGP + Laser group, highlights the potent antimicrobial effect facilitated by the combination of aPDT and NO. To further elucidate the effects on bacterial cell integrity, we utilized the live/dead BacLight viability kit. Within this kit, the green fluorescent nucleic acid dye SYTO-9 penetrates and labels all bacteria, while PI can only penetrate compromised membranes and the insertion of PI causes a reduction of SYTO 9 fluorescence. As shown in Fig. 3c, PI fluorescence signals were hardly observed in the E. faecalis treated with PBS + Laser, CPP, and CGP, suggesting that the bacteria have intact membranes. CPP + Laser and Ce6 + Laser-treated E. faecalis exhibited relatively strong green and red fluorescence, indicating a moderate disruption of the bacterial membranes. In contrast, the CGP + Laser-treated bacteria exhibited strong red fluorescence and weak green fluorescence, indicating severe damage to the bacterial membranes. These results confirm that the antimicrobial effect of CGP + Laser is superior to that of CPP + Laser, thus indicating the synergistic antimicrobial effect of aPDT and NO. Furthermore, the study explored the antimicrobial mechanism of the nanoparticles through protein leakage, changes in alterations in ATP levels, and alterations in bacterial morphology.

Fig. 3

In vitro antibacterial activity of CGP. a Representative images of plate samples and b bacterial viability, c Live/Dead staining, and d scanning electron microscope images of E. faecalis after receiving various treatments. e The amount of total protein released from E. faecalis and f intracellular adenosine 5'-triphosphate production levels in E. faecalis after receiving various treatments. The red arrows point to the holes in cell membranes. The red arrows point to the damaged bacterial membranes. Data are presented as mean ± SEM. In (b) n = 3, while in (e, f) n = 5. **p ≤ 0.01, ****p ≤ 0.0001, ns: no significance. In (b) comparison between PBS versus other groups, while in (e, f) between the CGP + Laser versus other groups

Figure 3d shows that bacteria treated with PBS + Laser, CPP, and CGP were predominantly smooth and intact, indicating minimal damage. However, the bacteria treated with Ce6 + Laser and CPP + Laser displayed slightly crumpled membranes with small holes, suggesting moderate damage. Notably, bacteria in the CGP + Laser group displayed extensive and deep holes in their cell membranes, indicative of significant damage and disruption. The extent of bacterial membrane damage correlated with the degree of protein leakage observed. As quantified in Fig. 3e. This trend clearly demonstrates the enhanced membrane-disrupting capability of the CGP + Laser treatment. In addition, ATP levels, crucial for cell function, typically diminish in cells undergoing apoptotic, necrotic, or in a toxic state [46]. As shown in Fig. 3f, ATP levels were reduced by 80.61%, 82.05%, and 94.72% in the Ce6 + Laser, CPP + Laser, and CGP + Laser groups, respectively, compared to the PBS group. This substantial reduction in ATP levels, particularly in the CGP + Laser group, underscores the profound impact of the combined aPDT and NO treatment on bacterial vitality. These fundings suggest a synergistic antimicrobial mechanism of aPDT and NO that may be attuned to the following factors: (1) the ability of guanidino groups to permeate biofilms and associate with bacteria, enhancing the overall antimicrobial action; (2) the combined disruptive effect of aPDT-generated ROS and NO on bacterial membranes, leading to significant protein leakage; (3) the synergistic inhibition of mitochondrial ATP production by both ROS and NO, impeding essential bacterial functions; and (4) the triggering of lipid peroxidation, DNA damage, and protein dysfunction by the byproducts of ROS and NO, ultimately resulting in bacterial death. Moreover, unirradiated CPP and CGP were not included in the other experiments as controls due to their negligible antimicrobial activity.

After establishing the synergistic antimicrobial effect of aPDT and NO, we conducted further analysis on the correlation between various concentrations of nanoparticles and their corresponding antimicrobial effectiveness. And compared the antimicrobial efficacy of these nanoparticles with that of NaClO, a commonly used clinical root canal irrigant. As shown in Additional file 1: Figs. S6 and S7, the bactericidal effects of free Ce6 + Laser, CPP + Laser, and CGP + Laser against E. faecalis bacteria progressively intensified with increasing nanoparticle concentrations. Additionally, across all tested concentrations, CGP + Laser consistently demonstrated superior antibacterial activity compared to Ce6 + Laser and CPP + Laser. This observation reaffirms the enhanced antimicrobial impact attributed to the synergistic action of aPDT and NO. For instance, bacterial viabilities following treatments with PBS, Ce6 (0.5 µg/mL) + Laser, CPP (8 µg/mL) + Laser, and CGP (8 µg/mL) + Laser were quantified as 9.02, 6.64, 5.14, and 4.30 log units, respectively. Furthermore, NaClO exhibited strong antibacterial properties, with bacterial viability decreasing to 3.28 log units at a 0.5% concentration, and a complete absence of bacterial colonies at a concentration of 1%. Additionally, it is observed that 18 µg/mL of CGP is more effective than 0.5% NaClO, yet not as effective as 1% NaClO. Remarkably, at a concentration of 36 µg/mL, CGP achieved complete eradication of bacterial colonies. These findings indicate that CGP can serve as a formidable alternative to traditional antimicrobial agents in endodontic therapy, offering a novel approach in the fight against bacterial infections in dental treatments.

Biofilm eradication of CGP in vitro

Considering the remarkable abilities of CGP in terms of biofilm penetration and synergistic antimicrobial activity, we postulated that it could efficiently eradicate biofilms. Our initial investigation into this matter involved assessing the biofilm eradication property of CGP on E. faecalis biofilms, utilizing the live/dead bacterial staining method. As shown in Fig. 4a, the presence of green fluorescence from SYTO-9 in the E. faecalis biofilm treated with PBS + Laser was evident, whereas the red fluorescence from PI was scarcely observed. Comparatively strong signals of both green and red fluorescence were visualized in the biofilms treated with Ce6 + Laser and CPP + Laser, indicating that only a portion of the biofilms was disrupted. In contrast, intense PI signals were observed in the biofilms treated with CGP + Laser and 1% NaClO, implying that the biofilms were nearly eradicated. To corroborate these findings, we also utilized crystal violet staining to evaluate the biofilm eradication efficacy of CGP. As shown in Fig. 4b and c, the biofilms remained relatively intact following treatment with Ce6 + Laser and CPP + Laser. In stark contrast, CGP + Laser and 1% NaClO exhibited efficient eradication efficacy, leaving only 20.24% and 14.22% of biofilm residuals. On the one hand, these results further indicate that CGP exhibits superior synergistic biofilm eradication properties compared to aPDT alone. On the other hand, although NaClO, due to its corrosive nature, slightly surpasses CGP in biofilm eradication efficacy, its potential for severe side effects and well-known cytotoxicity must be acknowledged. Consequently, we believe that CGP still holds potential for application in root canal irrigation. In addition, it was evident that the antimicrobial and biofilm eradication efficacies of free Ce6 + Laser were relatively inferior compared to those of CPP + Laser and CGP + Laser. Therefore, in subsequent experiments, free Ce6 + Laser was excluded from being used as a control group, allowing for a more focused analysis on the potent biofilm-targeting properties of CGP.

Fig. 4

In vitro biofilm eradication efficacy of CGP. a Live/Dead staining, b crystalline violet staining, and c residual rates of E. faecalis biofilms after receiving various treatments. d Sample plate representative images, e bacterial viability, and f scanning electron microscope images of E. faecalis biofilms in human extracted teeth after receiving various irrigations. Data are presented as mean ± SEM, n = 3, *p ≤ 0.05, ****p ≤ 0.0001, ns: no significance. Comparison between 1% NaClO versus other groups

Encouraged by the favorable outcomes of CGP + Laser in live/dead bacterial staining and crystal violet staining, we proceeded to cultivate E. faecalis biofilms in the root canals of extracted human teeth to assess the biofilm eradication efficacy. As shown in Fig. 4d and e, the bacterial viability post CPP + Laser treatment was observed at 2.14 log units, whereas CGP + Laser treatment reduced it to 0.93 log units. Notably, the efficacy of CGP + Laser was nearly akin to that of 1% NaClO, which showcased a bacterial viability of just 0.77 log units. To gain further insights, the treated tooth was longitudinally dissected along its long axis, focusing on the apical portion of the root canals for detailed observation under a scanning electron microscope. As shown in Fig. 4f, abundant E. faecalis biofilm densely colonizing the root canal wall and obstructing the dentinal tubules in the control group treated with 0.9% NS. In contrast, only a few bacterial residues were evident in CPP + Laser treated canals, while the canals treated with CGP + Laser and 1% NaClO appeared virtually devoid of any bacterial presence. These results collectively suggest that CGP + Laser treatment is highly effective in eradicating E. faecalis biofilm within the root canals, demonstrating an efficacy nearly comparable to the commonly used endodontic irrigant, 1% NaClO. Given the temporal and spatial controllability, along with its similar antimicrobial properties and biofilm eradication efficacy to NaClO, CGP emerges as a potential candidate for clinical applications in endodontics.

Hemocompatibility and biocompatibility assays

Biosafety is a paramount consideration for in vivo applications. In this context, we used erythrocyte aggregation assay and hemolysis assay to demonstrate the blood compatibility of the nanoparticles. As shown in Fig. 5a and b, when tested with free Ce6 (2 μg/mL), CPP (36 μg/mL), and CGP (36 μg/mL), there were no noticeable enhancements in erythrocyte aggregation, nor were there any disruptions in erythrocyte morphology. Moreover, in the absence of laser irradiation, free Ce6, CPP, and CGP exhibited < 0.25% erythrocyte hemolysis (Fig. 5c). This result starkly contrasts with the outcome of erythrocytes treated with 0.25% NaClO, which lost their integrity, leaving only cellular fragments observable under the microscope. Due to the strong corrosive and oxidizing properties of NaClO [47], the blood was decolored after co-incubation with 0.25% NaClO for 1 min.

Fig. 5

Blood compatibility and biocompatibility of CGP. a Microscopic images of erythrocyte aggregation in different treatment groups. b, c Hemolytic effects in different treatment groups. d Cell Counting Kit-8 assay of MC3T3-E1 and e hPDLSCs cells after treated with various groups. f Live/dead cell staining in different groups of hPDLSCs. Data are presented as mean ± SEM. In (c) n = 3, while in (d, e) n = 5. *p < 0.05, ***p ≤ 0.001, ****p ≤ 0.0001, ns: no significance. In (c) comparison between PBS versus other groups, while in (d, e) between the control group versus other groups

Besides, the cell viability of CGP was consistently > 85% after co-incubation with MC3T3-E1 cells or hPDLSCs for 24 h (Fig. 5d and e). Notably, the morphology of hPDLSCs remained unaffected after this period, with no significant red fluorescence from PI staining observed (Fig. 5f). As predicted, NaClO exhibited pronounced cytotoxicity, with less than 2.5% cell survival after 1 min of co-incubation with hPDLSCs and MC3T3-E1 cells. The above experiments show that CGP has good hemocompatibility and biocompatibility, with great potential for clinical applications.

Osteogenic differentiation in vitro

In the treatment of AP, in addition to controlling the infection, we hope that the CGP will also promote periapical bone repair and regeneration. Bone regeneration is a complex process associated with inflammation, angiogenesis, and osteogenesis. NO may play an important role in ALP activity and mineralization in osteoblastic lineage [33]. Our cellular experiments employed H2O2 to simulate inflammatory environments, and drive CGP to generate NO. After 7 days of treating MC3T3-E1 cells with H2O2 and H2O2 + CPP under mineralized conditions (add β-glycerol and ascorbic acid in the medium), the degree of ALP staining was slightly lower than that of PBS, while H2O2 + CGP was darker than PBS (Fig. 6a). This was corroborated by quantitative ALP data, which revealed that the ALP activity in the H2O2 + CGP group was 19% higher compared to H2O2 alone (Fig. 6b). During osteoblast differentiation, calcium compounds are deposited on the cell surface to form calcium nodules. These calcium nodules can be stained orange by using alizarin red staining [48]. After 21 days of treatment, MC3T3-E1 cells treated with H2O2 and H2O2 + CPP showed fewer mineralized nodules than those treated with PBS, whereas they were more abundant in H2O2 + CGP (Fig. 6c and d). These findings suggest that while H2O2 and H2O2 + CPP substantially impeded the mineralization process of MC3T3-E1 cells, H2O2 + CGP alleviated the adverse impacts of H2O2 and reinstated the count of mineralized nodules to levels comparable to those of the PBS group.

Fig. 6

Mechanisms of CGP promoting repair of periapical bone defects. a, b Images and quantitative data of ALP staining and c, d alizarin red staining in MC3T3-E1 cells after receiving various treatments in mineralizing conditions. e The expression levels of OCN and f RUNX2 in MC3T3-E1 cells after receiving various treatments in mineralizing conditions. Data are presented as mean ± SEM, n = 3, *p < 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, ns: no significance. Comparison between H2O2 + CGP versus other groups

ELISA was utilized to analyze the secretion of osteogenesis-related markers, encompassing OCN and RUNX2, in MC3T3-E1 cells with various treatment. OCN is known to play a crucial role in bone metabolism, particularly in calcium homeostasis and bone mineralization processes [49]. RUNX2, on the other hand, functions in the growth and differentiation of osteoblasts by stimulating the expression of subsequent biomarkers [50, 51]. As shown in Fig. 6e, the concentration of OCN in the supernatants of MC3T3-E1 cells varied across treatment groups. The cells treated with PBS showed an OCN level of 98.6 ng/mL. In the presence of H2O2 and H2O2 + CPP, this level decreased slightly to 91.9 and 94.7 ng/mL, suggesting some disruption of osteogenic activity under inflammatory conditions. Remarkably, cells treated with H2O2 + CGP demonstrated a significant increase in OCN concentration, reaching 129.6 ng/mL. This suggests that the presence of CGP, not only mitigates the inhibitory effect of H2O2 but also enhances osteogenic activity. Similarly, the RUNX2 levels in MC3T3-E1 cell lysates exhibited a corresponding pattern. Cells treated with PBS had a RUNX2 level of 152.0 ng/mL. In the presence of H2O2 and H2O2 + CPP this level decreased slightly to 143.0 and 144.6 ng/mL. However, in the H2O2 + CGP group, there was a notable increase in RUNX2 concentration, reaching 172.6 ng/mL (Fig. 6f). This increase suggests that CGP enhances osteoblast differentiation and bone formation, even under conditions of inflammatory.

These results indicate a notable suppression of osteogenic differentiation in the presence of H2O2, a ROS that is commonly present in inflammatory conditions [52]. However, the introduction of CGP appears to counteract this inhibitory effect. The mechanism underlying this phenomenon involves the consumption of H2O2 by CGP, leading to the generation of NO. This release of NO from CGP is instrumental in stimulating osteogenic differentiation in MC3T3 cells, as evidenced by increased ALP activity at early stages of differentiation, initiation of calcium deposition at later stages, and the up-regulation of key osteogenic markers such as OCN and RUNX2. It is noteworthy that this assay does not accurately mirror the actual in vivo scenario, but H2O2 is a ROS that persistently exists in inflammatory sites in the long run, and ROS produced by periapical tissues obstruct osteoblast differentiation [53]. In therapeutic applications, residual H2O2, either from aPDT or periapical inflammation, could potentially oxidize CGP, thereby producing trace amounts of NO. This process is expected to promote the repair and regeneration of periapical bone defects, highlighting the potential of CGP as a therapeutic agent in conditions where inflammation and oxidative stress impede normal bone healing and regeneration processes.

In vivo treatment of apical periodontitis

CGP exhibited favorable biocompatibility, antimicrobial ability, eradication of biofilm, and promotion of mineralization properties in vitro experiments. This provides a foundation for carrying out in vivo therapeutic evaluation using a rat apical periapical model. Figure 7a presents the detailed experimental procedure for establishing an AP model for treatment and effect assessment. Briefly, the pulp cavity of the left maxillary first molar of the rat was opened, and a small cotton ball containing E. faecalis suspension was placed to infect the root canal, after which the crown was sealed with glass ions. Following 2 weeks of infection, the rats were treated with PBS, 1% NaClO, CPP + Laser, and CGP + Laser, respectively. After 3 weeks of treatment, the left maxillary first molar and the surrounding maxillary area were collected from the rats, and the extent of periapical bone defects was measured using micro-computed tomography. The results revealed notable reductions in bacterial viability across all treatment groups on the day of irrigation (Fig. 7b and c) and at day 21, indicating significant antimicrobial effects (Fig. 7d and e). Consistent with the results of root canal irrigation in extracted teeth, the antimicrobial effect was higher than 96% in all groups. Figure 7f shows the coronal and sagittal planes of the three-dimensional reconstructed images of the representative first molar for each group. The figure clearly illustrates large bone resorption cavity in the periapical tissues of the PBS group, indicating that E. faecalis infection of the root canal led to substantial periapical bone defect. In contrast, this defect obviously decreased in the 1% NaClO, CPP + Laser, and CGP + Laser groups, indicating gradual healing of the bone defect after successful infection control. Particularly noteworthy is the significantly smaller resorption cavity observed in the CGP + Laser group compared to the other groups, with its periapical condition closely resembling that of the healthy group, suggesting enhanced healing of periapical bone defects. And the quantitative data presented in Fig. 7g, the volume of the resorption cavity was 2.38, 1.12, 0.94, and 0.18 mm2 for PBS, 1% NaClO, CPP + Laser, and CGP + Laser, respectively. CPP + Laser exhibited a 7.5% reduction in bone defects in periapical inflammation treatment compared to the NaClO group (PBS group as 100%), likely due to the extrusion effect of NaClO, causing periapical tissue toxicity [54]. The conventional syringe irrigation method was employed in this study, and apical extrusion was inevitable except for negative pressure irrigation systems [55]. Remarkably, the resorption cavity of CGP + Laser was 32.2% lower than that of CPP + Laser. CGP + Laser showed excellent therapeutic effects with antimicrobial effects, and the production of NO during therapy aided in the healing of periapical bone defects. The synergistic application of aPDT with NO has been proven to treat AP both effectively and safely.

Fig. 7

In vivo therapeutic efficacy of CGP apical periodontitis model rats. a Experimental design of apical periodontitis model rats. b, c Sample plate representative images and bacterial viability of E. faecalis in rat teeth after receiving various irrigations on day 0 and d, e day 21. f Representative images and g the total resorption volume of the left maxillary first molar in each group. h Representative HE staining and i TRAP staining. The green arrows point to the neutrophils and the red arrows point to active osteoclasts. Data are presented as mean ± SEM, n = 6, ****p ≤ 0.0001. Comparison between PBS versus other groups

H&E and TRAP staining were performed to facilitate additional evaluation of the synergistic impact of aPDT and NO. As shown in Fig. 7h, the periapical area of the PBS group exhibited abundant neutrophil infiltration, whereas the 1% NaClO and CPP + Laser groups displayed minor inflammatory cell infiltration, indicates that inflammation in the periapical tissues gradually diminished after successful infection control. Conversely, the CGP + Laser group showed minimal neutrophil infiltration, suggesting a milder inflammatory response in this group. Additionally, the periapical tissues in this group exhibited morphological integrity without discernible defects, indicating the most favorable treatment outcome. As shown in Fig. 7i, TRAP staining showed numerous active osteoclasts in the periapical tissues of the PBS, while the number of osteoclasts was reduced in NaClO and CPP + Laser group. Notably, osteoclasts were virtually absent in CGP + Laser, while histological features and the periodontal ligament space were similar to those of healthy group. These findings may be attributed to the potent antimicrobial effect and the capacity to stimulate bone regeneration associated with CGP + Laser, demonstrating optimal therapeutic efficacy in endodontic treatment. In addition, no significant toxicity of CGP was observed during treatment, and no pathological abnormalities were observed in the major organs of the rats (Additional file 1: Fig. S8), which indicated that the safety of our multifunctional antibacterial nanoplatform [56, 57]. These findings substantiate the potential future application of CGP + Laser in endodontics, while also presenting a novel candidate treatment strategy for root canal irrigation.

Although this study has yielded some promising results, it is important to acknowledge certain limitations. Subsequent research should focus on conducting comprehensive analyses of the exact mechanisms and efficacy of CGP within multispecies bi