Preparation and characterization of sPDMA@ICG NPs
A star-shaped polycation sPDMA was first synthesized via ATRP of DMA monomer from CD-Br initiator according to our previous works [23, 31], and the synthesis route is shown in Fig. S1. From the 1H NMR spectrum of CD-Br initiator in Fig. S2, the number of Br per β-CD molecule was calculated to be 7, indicating that 7 PDMA chains were grafted from one β-CD. After polymerization, the characteristic peaks representing protons in PDMA appeared at 4.1, 2.6 and 2.3 ppm (Fig. 1a), proving the success of polymerization. Since the content of β-CD in the polymer was relatively small, the proton signals of β-CD were not observed in the spectrum. According to the polymerization yield, the degree of polymerization (Dp) was estimated to be about 12.2 per chain and the molecular weight of sPDMA was approximately 15,600. The molecular weight distribution (Mw/Mn) was 1.24 characterized by gel permeation chromatography. To validate whether the polymer is in molecular brush regime, the gyration radius (Rg) and the distance (D) between the neighboring chains were estimated. According to the following equation [32], Rg = 0.5Dp0.5, Rg of PDMA chain was calculated to be about 1.75 nm. While D was estimated to be less than the outer diameter of β-CD (1.54 nm) [33]. Thus, D/Rg would be less than 0.88, meaning that the PDMA chains are in molecular brush regime. With such high grafting density, the PDMA chains will be forced into a stretched state due to the steric constraint, endowing sPDMA unique properties different from its random coil state.
sPDMA@ICG NPs were prepared from self-assembly of ICG and sPDMA in aqueous solution using the nanoprecipitation method described in the Supporting information. Fig. S3 shows the ultraviolet-visible-near infrared (UV-Vis-NIR) absorption spectra of ICG and sPDMA@ICG NPs. The characteristic absorption peak of ICG at 780 nm showed an obvious red shift after assembled with sPDMA, indicating that ICG had entered a less polar environment. Besides, the peak intensity of ICG monomer at 780 nm significantly decreased, while that of ICG dimer at 700 nm increased a little at the same time. This demonstrated that the distance between ICG molecules is closer in the NPs than in aqueous solution according to a previous report [34]. sPDMA@ICG NPs with different weight ratios of ICG/sPDMA were further prepared using the same method. In the UV-Vis-NIR spectra of these NPs, the similar peak shapes were observed clearly (Fig. 1b), meaning that ICG was loaded in similar environment with similar forms. When the ICG/sPDMA weight ratio was 4/14, sPDMA@ICG NPs displayed the smallest particle size and the narrowest polydispersity index (PDI) (Fig. 1c). Hence, sPDMA@ICG NPs used in the following investigation were prepared at this weight ratio. Under the observation of transmission electron microscope (TEM), sPDMA@ICG NPs had a regularly spherical shape (Fig. 1d). The diameter of sPDMA@ICG NPs determined using the dynamic light scattering method was 206 nm with a narrow size distribution, and their zeta potential was about + 18.4 mV (Fig. 1e). Within 5 days, the size and size distribution of sPDMA@ICG NPs changed little (Fig. 1f). In the UV-Vis-NIR absorption spectra, the peak intensities of sPDMA@ICG NPs were almost the same in this period (Fig. 1g), while free ICG showed obviously decreased absorption over time (Fig. 1h). These results demonstrate sPDMA@ICG NPs have better storage stability than free ICG.
Photothermal and photodynamic performances of sPDMA@ICG NPs
ICG is a photosensitizer with both photothermal and photodynamic performances. After assembled with sPDMA to form PDMA@ICG NPs, the photothermal and photodynamic performances of ICG were further investigated and the results are shown in Fig. 2. After irradiation with an 808 nm laser at 2 W/cm2, sPDMA@ICG NPs showed significant temperature elevations in an ICG concentration-dependent manner and the temperatures of their solutions rapidly increased within 2 min (Fig. 2a). Moreover, sPDMA@ICG NPs had slightly higher photothermal efficiency than that of free ICG at the ICG concentration of 35 µg/mL and their solution temperature increased from 22 to 55°C (Fig. S4A and B). It thus can be deduced that sPDMA@ICG NPs have good photothermal property. The photodynamic performance of sPDMA@ICG NPs was next evaluated by detecting the ROS generation using a green fluorescence probe, Singlet Oxygen Sensor Green (SOSG). The fluorescence intensity of SOSG at 525 nm for free ICG and sPDMA@ICG NPs (10 µg/mL ICG) both gradually increased with the increase of irradiation time, while it was not observed in phosphate buffered saline (PBS) or sPDMA (Fig. 2b). These results confirmed the significant photodynamic performances of ICG and sPDMA@ICG NPs. By comparison, sPDMA@ICG NPs showed an obviously weaker photodynamic performance, which may be owing to the change of light absorption property of ICG after self-assembly with sPDMA.
Bacterial surface adsorption and outer membrane penetration of sPDMA@ICG NPs
ICG is negatively charged and water soluble, hence it is difficult for it to pass through the bacterial cell membrane that is also negatively charged and characterized by lipid bimolecular structure. In this study, sPDMA@ICG NPs were prepared from self-assembly of ICG and sPDMA in aqueous solution, hoping to promote the adsorption and penetration of ICG in bacterial cells by taking advantage of the high positive charge density of sPDMA. The surface adsorption of sPDMA@ICG NPs towards Pg, a Gram-negative bacterium that is directly associated with periodontitis, was firstly evaluated through monitoring the change of bacterial surface charge property after 3-hour incubation. As shown in Fig. 3a, the negative charges on the surface of Pg incubated with sPDMA@ICG NPs were neutralized continuously, reflecting in the change of zeta potential from negative to positive and the further enhanced zeta potential value as the ICG concentration increasing. But in the meantime, the zeta potential of Pg incubated with free ICG maintained negative values. The above results demonstrate that sPDMA@ICG NPs can be adsorbed onto the bacterial surfaces via electrostatic interaction.
N-phenyl-1-naphthylamine (NPN) is often applied as a fluorescence probe to detect the penetration and damage of bacterial cell membrane [35], since it emits weak fluorescence in aqueous solution but strong fluorescence after spontaneously entering the hydrophobic region of cell membranes. Herein, NPN was used as a fluorescence indicator to evaluate the penetration of sPDMA@ICG NPs through the outer membrane of Pg, and the results are shown in Fig. 3b. NPN displayed very strong fluorescence signals in the bacteria after 3-hour incubation with sPDMA and sPDMA@ICG NPs. However, no visible increased fluorescence intensity of NPN was observed in the bacteria incubated with free ICG as compared to that incubated with PBS. It thus can be deduced that sPDMA has significant bacterial cell membrane-penetrating ability due to the superhigh density of positive charges on its brush layer. Moreover, upon 808 nm laser irradiation at 2 W/cm2 for 5 min, free ICG and sPDMA@ICG NPs further enhanced the permeability of outer membrane of Pg, demonstrating the damage of bacterial membrane caused by the photothermal and photodynamic performances of ICG.
The uptakes of free ICG and sPDMA@ICG NPs in Pg were also compared after 3-hour incubation through observing the intracellular fluorescence of ICG. Figure 3c shows the confocal microscopic images of the bacteria with staining of 4',6-diamidino-2-phenylindole dihydrochloride (DAPI). The intensive red fluorescence representing ICG was clearly observed in sPDMA@ICG NP-incubated Pg when the ICG concentration was 10 µg/mL, while almost no fluorescence was observed for free ICG with the same concentration. This indicated that sPDMA@ICG NPs were efficiently accumulated in the bacterial cells after penetrating the cell membranes, and thus would facilitate ICG to exert antibacterial effects through synergistic PTT and PDT.
In vitro antibacterial effects of sPDMA@ICG NPs with laser irradiation
Since sPDMA@ICG NPs can effectively deliver ICG into the bacteria cells and exhibit synergistic PTT and PDT performances, we further evaluated the antibacterial effects of sPDMA@ICG NPs. First, the PTT effect of sPDMA@ICG NPs was examined in Pg. After incubation with sample solutions for 3 h, the bacterial suspensions were centrifuged to remove the unbound samples and the precipitations were dispersed in PBS. These bacterial suspensions were irradiated with an 808 nm laser at 2 W/cm2 for 10 min, and their temperatures were further recorded using an infrared thermal camera within this period. sPDMA@ICG NPs maintained high photothermal conversion efficiency in the bacterial suspension, which was significantly stronger than free ICG (Fig. 4a and b). After laser irradiation for 5 min, the temperature raised and maintained in the range of 45–50°C, which was greatly higher than the body temperature. At this temperature range, the oral pathogenic bacteria can be damaged and inhibited effectively. This further proves that sPDMA@ICG NPs efficiently delivered ICG into the bacteria to exert the PTT efficacy against periodontitis.
PDT can induce the generation of ROS that possesses strong cytotoxicity via destroying bacterial lipids, proteins and genes. To evaluate the PDT performance of sPDMA@ICG NPs in Pg, 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA) was used as a fluorescent probe to detect the intracellular production of ROS. When a large amount of ROS is produced in the cells, non-fluorescent DCFH-DA can be oxidized into a fluorescent product 2’,7’-dichlorofluorescein (DCF) through a series of reactions. The fluorescence intensity of DCF is related directly to the intracellular level of ROS. Figure 4c compares the fluorescence intensities of DCF in the bacteria. Upon 10 minute-laser irradiation at 808 nm at 2 W/cm2, free ICG and sPDMA@ICG NPs both induced the notably enhanced fluorescence intensities of DCF, indicating that they exerted the PDT performances to trigger the generations of intracellular ROS. By comparison, sPDMA@ICG NPs showed a significantly more potent PDT performance, inflecting in a higher fluorescence intensity of DCF. This further proves sPDMA@ICG NPs efficiently delivered ICG into the bacteria to exert the PDT efficacy against periodontitis.
The antibacterial activity of sPDMA@ICG NP-mediated PTT and PDT was next evaluated in Pg using the CCK-8 assay. The influence of sPDMA on the bacterial viability was assessed firstly at different concentrations. Results in Fig. S5 shows that the bacterial viability decreased notably when the sPDMA concentration was larger than 50 µg/mL. Therefore, the concentration of sPDMA below 25 µg/mL was chosen in the following study. Without laser irradiation, the bacterial viabilities in the groups of free ICG and sPDMA@ICG NPs maintained at the high levels in the ICG concentration range of 1–10 µg/mL. After 808 nm laser irradiation at 2 W/cm2 for 5 min, free ICG and sPDMA@ICG NPs both inhibited the bacterial growth remarkably, but by comparison, sPDMA@ICG NPs displayed a much higher antibacterial effect (Fig. 4d). Under the TEM observation, the adsorption of sPDMA@ICG NPs was clearly visible on the surfaces of Pg, and after laser irradiation, the bacterial membrane was ruptured and the bacterial cells were disintegrated (Fig. 4e). Colony formation assay was further applied to evaluate the antibacterial activity visibly in Pg. As shown in Fig. 4f, sPDMA@ICG NPs with laser irradiation almost completely inhibited the bacterial growth at the ICG concentration of 10 µg/mL, while the bacteria in the other groups grew in different degrees. These results indicated that sPDMA@ICG NPs exerted synergistic PTT and PTT.
Destruction effect of sPDMA@ICG NPs with laser irradiation on the plaque biofilm
Dental plaque, which actually is the same as the bacterial biofilm, has been known to be the basis for the survival of microorganisms. The dental plaque can resist the penetration of antibacterial agents and thus reduce their antibacterial effects [36, 37]. This shows that the destruction of dental plaque will help to antibacterial treatment. Herein, the pathogenic bacteria were extracted from the periodontitis rats and incubated in the liquid culture medium for 3 d to form the plaque biofilm, which was further applied for evaluation of the destruction effect of sPDMA@ICG NP-mediated PTT and PDT. Only upon laser irradiation (808 nm, 2 W/cm2 and 5 min), free ICG and sPDMA@ICG NPs remarkably destroyed the plaque biofilms at the ICG concentration of 10 µg/mL (Fig. 5a). Not surprisingly, sPDMA@ICG NPs with laser irradiation exhibited a more potent destruction effect on the plaque biofilm than free ICG with laser irradiation, reflecting in a larger area (2.87 cm2) of damaged biofilm (Fig. 5b). These damaged plaque biofilms were further stained with the LIVE/DEAD (SYTO9/PI) BacLight Bacterial Viability Kit to evaluate the antibacterial activity. Under a fluorescence microscope, the live and dead bacteria emitted green and fluorescence separately. Similar to the above results, the bacteria in the group of sPDMA@ICG NPs with laser irradiation were almost completely killed (Fig. 5c). These results suggest that synergistic PTT and PDT mediated via sPDMA@ICG NPs is a promising antibacterial method for periodontitis treatment because of its destruction effect on the plaque biofilm and penetration ability through the bacterial membrane.
In vivo PTT and PDT performances of sPDMA@ICG NPs
A periodontitis animal model was constructed in Sprague Dawley (SD) rats according to our previous method [17]. After anesthetization, the rats were ligated with orthodontic steel wires at the gingival sulcus of the left maxillary second molar and then feed with 10% sucrose water for 4 weeks to induce periodontitis. These rats were given the treatments of sPDMA, free ICG and sPDMA@ICG NPs alone and their combination with laser irradiation. Here, 50 µL of sample solutions containing 20 µg/mL of ICG and/or 70 µg/mL of sPDMA were administrated via smear and laser irradiation was carried out at 808 nm at 2 W/cm2 for 5 min. Figure 6a shows the photos of periodontitis rat receiving the treatment of sPDMA@ICG NPs and laser irradiation.
DCFH-DA was next used as a fluorescence probe to detect the generation levels of ROS in periodontal tissues at treatment site. Figure 6b shows the confocal microscope images of frozen tissue sections with DCFH-DA staining. The green fluorescence of produced DCF was only observed in the periodontal tissues of the rats receiving the treatments of free ICG and sPDMA@ICG NPs with laser irradiation, demonstrating the generations of ROS induced by these treatments. Figure 6c shows the corresponding quantitative analysis of DCF fluorescence intensities. It was obvious that sPDMA@ICG NPs with laser irradiation induced a much higher level of ROS generation than free ICG with laser irradiation. These results indicated that sPDMA@ICG NPs had a potent PDT performance in periodontitis rats owing to their adhesion and penetration towards the plaque biofilms and bacterial membranes.
Within 5 min of laser irradiation, the temperature changes at periodontal tissues in the rats with administration of PBS (the control) and sPDMA@ICG NPs were also detected using a thermal imaging camera. As shown in Fig. 6d, sPDMA@ICG NPs increased the temperature of periodontal tissue rapidly to 43.1°C within 1 min of laser irradiation and 51.2°C within 5 min of laser irradiation. According to an investigation we previously reported [38], this temperature will not bring about the serious damages to the normal tissues. On the contrary, the temperature of periodontal tissue in the control rat did not increase significantly within this period of laser irradiation. These results indicated that sPDMA@ICG NPs also had a potent PTT performance in periodontitis rats.
In vivo anti-periodontitis effects of sPDMA@ICG NPs with laser irradiation
Periodontitis rats were used to evaluate the in vivo anti-periodontitis effects of synergistic PTT and PDT mediated by sPDMA@ICG NPs. The treatments including sPDMA, free ICG, sPDMA@ICG NPs alone and their combination with laser irradiation as described above were carried out once a week for consecutive 3 weeks. Afterwards, the resorption of alveolar bones in these treated rats was analyzed using the microcomputed tomography (micro-CT). The three-dimensional digital and tomographic images of alveolar bones were separately reconstructed with CTVox software and DataViewer software, shown in Fig. 7a and b. Compared to the negative (-) control rat, the alveolar bone resorption was very obvious in the positive (+) control rat, confirming the success construct of the periodontitis rat model. Upon laser irradiation, free ICG and sPDMA@ICG NPs both remarkably inhibited the alveolar bone resorption in periodontitis rats. By comparison, sPDMA@ICG NPs with laser irradiation exhibited a much stronger inhibitory effect on the alveolar bone resorption, and the treated rats had the alveolar bone heights similar to that of the control (-) rats. The distance between alveolar bone crest (ABC) and cement enamel junction (CEJ) was determined by the three points from the mesial to distal root surface of the second molars. Bone volume (BV), tissue volume (TV) and their ratios (BV/TV) around the ligated molars were further calculated by using CTAn software. Results in Fig. 7c and d showed that the rats receiving the treatment of sPDMA@ICG NPs with laser irradiation had the shortest distance of CEJ-ABC and the highest of BV/TV. These results were basically consistent with the in vitro antibacterial results, confirming that sPDMA@ICG NPs can efficiently prevent the alveolar bone resorption through exerting the PTT and PDT performances.
When all treatments were completed, the rats were euthanized and their periodontitis lesions (palatal gingiva and surrounding mucosa) were removed from the left maxillary second molar for histopathological examination. The microscopic images of tissue sections stained with hematoxylin and eosin (H&E) are shown in Fig. 8a. Different degrees of inflammatory responses (incomplete epithelium, epithelial hyperplasia, irregular arrangement of basal cells, and infiltration of inflammatory cells) were observed in the control(+) rats and the rats receiving the treatments of sPDMA with and without laser irradiation, free ICG and sPDMA@ICG NPs. However, these inflammatory responses were greatly ameliorated in the rats treated with free ICG and sPDMA@ICG NPs after laser irradiation. More importantly, sPDMA@ICG NPs with laser irradiation almost completely prevented the progress of periodontitis, which was manifested in the normal morphological and closely arranged epithelial cells as well as no obvious inflammatory changes in lamina propria.
As two of the inflammatory factors, tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β) play the important roles in the regulation of local inflammatory responses in periodontitis [39]. TNF-α can induce the inflammation, promote the secretion of matrix metalloproteins, activate the osteoclasts, and destroy the periodontal tissue [40, 41]. IL-1β can recruit and activate the neutrophile granulocytes, improve the release of inflammatory mediators, and decompose the connective tissue [42, 43]. These inflammatory factors will be convenient for pathogenic bacteria and their toxic metabolites to invade deep tissues and promote the occurrence and development of periodontitis [44]. Therefore, we further evaluated the protein levels of TNF-α and IL-1β in tissues by immumohistochemical staining. As can be seen in Fig. 8b and c, sPDMA@ICG NPs with laser irradiation evidently reduced the protein levels of these two inflammatory factors as compared to the control(+) and the other treatments. It thus can be deduced that synergistic PTT and PDT mediated by sPDMA@ICG NPs relieved the inflammatory responses in periodontitis rats.