2.1 Synthesis of perylene monoamide-glycoconjugates (PMI-3Gal and PMI-3Fuc)
As shown in Scheme S1, compounds PMI-3Gal and PMI-3Fuc were synthesized by a click reaction of the intermediate PMI-1 with the azide group modified D-galactose and L-fucoside, respectively, and followed by deprotection of the acetyl groups. The intermediates and the target molecules (PMI-3Gal and PMI-3Fuc) were fully characterized by 1H and 13C nuclear magnetic resonance (1H NMR and 13C NMR) as well as high resolution mass (HRMS) spectra. (Figure S1-S12, Supporting information).
2.2 Assembly properties of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc
Due to the strong π-π stacking interactions of the perylene backbones, PMI-3Gal and PMI-3Fuc (Fig. 2a) could form the self-assemblies and the co-assemblies in water. Firstly, solvent-dependent UV-vis spectra of PMI-3Gal and PMI-3Fuc were used to study the self-assembly and co-assembly behaviors. As shown in Fig. 2b and 2c, PMI-3Gal and PMI-3Fuc showed the maximum absorption band at 630 nm and 636 nm in DMSO solution, respectively, indicating that different types of the carbohydrate modification influenced the optical property of PMI. These optical properties of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc (1:1) in DMSO solution remained a non-aggregated state [37]. As shown in Fig. 2d, the maximum absorption band of the co-assembly PMI-3Gal@PMI-3Fuc was at 633 nm, which was lower than the maximum absorption of PMI-3Fuc and higher than the maximum absorption of PMI-3Gal. Upon increasing of the water ratio, the intensities of the maximum bands decreased, and the maximum bands underwent a hypsochromic shift to 566 nm, 569 nm and 566 nm for PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc, respectively, indicating the formation of H-aggregates due to strong intermolecular π-π stacking interactions in water [37]. Fluorescence spectra of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc showed a similar result. As shown in Figure S13, strong fluorescence emission bands at 736 nm, 736 nm and 738 nm of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc were found in DMSO solution, and decreased along with the increasing of water due to the enhanced π − π stacking interactions.
The self-assembly and co-assembly behaviors were further investigated by the dynamic light scattering (DLS) measurements. As shown in Figure S14a-c, the assemblies with the mean diameters of 259.85 nm, 259.33 nm and 261.98 nm for PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc were observed, and no obvious differences in the self-assemblies and the co-assemblies were found. Different morphologies were characterized by the SEM images, as shown in Figure S14 d-f, which showed silk ribbon for PMI-3Gal and PMI-3Gal@PMI-3Fuc, and flocculent aggregates for PMI-3Fuc.
In addition, the co-assembly model of PMI-3Gal and PMI-3Fuc was investigated by a turbidity assay (Fig. 2e). Two possible assembly models can be postulated, that is, the complexes (Model I, Fig. 2f) of the assemblies of PMI-3Gal and the assemblies of PMI-3Fuc or the co-assemblies (Model II, Fig. 2f) of PMI-3Gal and PMI-3Fuc with an interlaced mode. It is well known that lectin-carbohydrate interactions are specific and selective [39], and the self-assembled glycoclusters show enhanced binding interactions [40] with unique lectin through multivalent effect. As shown in Fig. 2e, the turbidity of the assemblies of PMI-3Gal showed a sharp increase upon addition of peanut agglutinin (PNA) lectin because PNA selectively bound to the terminal-galactosyl residues [40]. However, the turbidity of the mixture of the self-assemblies of PMI-3Fuc with PNA showed no obvious change, indicating no binding to PNA. When adding co-assemblies of PMI-3Gal and PMI-3Fuc, the turbidity increased, but was lower than that of the self-assemblies of PMI-3Gal. This result was due to that the insertion of non-recognizable carbohydrate in the assemblies weakened the specific and selective binding interactions of PNA with the terminal-galactosyl residues. These results indicated that the possible assembly model was the co-assemblies (Model II, Fig. 2e) of PMI-3Gal and PMI-3Fuc with an interlaced mode.
2.3 Photothermal properties of PMI-3Gal and PMI-3Fuc
Benefited from the strong π-π stacking interactions, the assemblies of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc showed very weak fluorescence in water, which was advantageous to the photothermal effect [41]. The photothermal effects of the assemblies and the co-assemblies of PMI-3Gal and PMI-3Fuc were studied under different laser powers. Under the concentration of 100 µM, as shown in Fig. 2g-i, the temperatures of PMI-3Gal at the laser powers of 0.2, 0.4 and 0.6 W/cm2 increased to 48.2°C, 65.6°C and 72.6°C, respectively, which were higher than that of PMI-3Fuc and lower than that of PMI-3Gal@PMI-3Fuc.
It is well known that photothermal therapy for tumor needs sufficient light and intense power to penetrate the tissues of the organism [36]. However, it is dangerous to the surrounding healthy tissue, especially for wound infections, which are not as deep as tumors [42]. Therefore, concentration-dependent photothermal effects under a power of 0.4 W/cm2 were studied. As shown in Fig. 2j-i, the temperatures of PMI-3Gal increased to 48.2°C,56.2°C, 61.0°C, and 65.6°C under the concentrations of 25 µM, 50 µM, 75 µM and 100 µM, respectively. The co-assemblies of PMI-3Gal@PMI-3Fuc also showed the best photothermal effect with maximum temperatures of 49.1°C, 56.4°C, 62.7°C, and 66.3°C. PMI-3Fuc showed the low temperatures of 46.2°C, 54.8°C, 59.5°C, and 64.2°C. This result indicated that different types of carbohydrate modification in PMI influenced the photothermal effect. As a control experiment, the temperature of the water solution only increased from 27.0°C to 29.0°C under laser irradiation. It is very interesting to note that the co-assemblies of PMI-3Gal@PMI-3Fuc exhibited the best photothermal effect.
Furthermore, the quantitative photothermal-conversion efficiency (𝜂) was calculated by warming/cooling curves with a Roper’s method [43]. As shown in Figure S15, the co-assemblies of PMI-3Gal@PMI-3Fuc showed a high photothermal conversion efficiency value of 63%, which was higher than that of PMI-3Gal (55%) and PMI-3Fuc (48%). Combined with the optical properties and the morphologies, we can conclude that different types of carbohydrate modification in PMI influence the optical properties, assembly behaviors, and the photothermal effect. In addition, the photostability of the self-assemblies of PMI-3Gal and PMI-3Fuc, and the co-assemblies of PMI-3Gal@PMI-3Fuc was evaluated upon five cycles under laser irradiation. A cycle was performed by irradiation of the sample for 10 min with laser, and then remove the laser, and the temperature reached a high point and then decreased to the room temperature. There were no obvious changes for the high temperature during five cycles of laser irradiation, indicating that PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc possessed high photostability.
2.4 Lectin-targeted antimicrobial activity of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc for P. aeruginosa
The tetravalent lectins LecA and LecB of P. aeruginosa exhibited selective recognition with D-galactose and L-fucose and played crucial roles in bacterial adhesion, biofilm formation, and virulence. Firstly, the adhesion actions of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc toward P. aeruginosa were performed.
As shown in Figure S16 and Table S1, P. aeruginosa with a working concentration of 9 × 108 colony forming units (CFU) mL− 1 was incubated for 6 h at 37°C with various concentrations (50 µM, 75 µM, 100 µM, 125 µM, 150 µM, and 200 µM) of the assemblies. The adhesion concentrations, determined by the Lambert-Beer law, were 16 µM, 29 µM, 56 µM, 92 µM, and 140 µM for PMI-3Gal, 12 µM, 24 µM, 45 µM, 68 µM, 79 µM, and 125 µM for PMI-3Fuc, and 17 µM, 30 µM, 56 µM, 80 µM, 93 µM, and 140 µM for PMI-3Gal@PMI-3Fuc. Compared with PMI-3Fuc, PMI-3Gal showed a better adhesion interaction for P. aeruginosa. Considering the result that PMI-3Fuc showed a weak adhesion interaction for P. aeruginosa, the co-assemblies of PMI-3Gal@PMI-3Fuc should exhibit weaker adhesion effect for P. aeruginosa than that of PMI-3Gal in theory. However, it is interestingly found that the co-assemblies of PMI-3Gal@PMI-3Fuc showed the best adhesion effect for P. aeruginosa, which might due to a hetero-multivalent effect [44] that enhances the affinity towards the target lectin by cooperative recognition interactions with different saccharides.
Benefited from the strong adhesion effect for P. aeruginosa and high photothermal effect of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc, their antibacterial activities were further studied. The antibacterial performances of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc with and without laser irradiation against P. aeruginosa in vitro were evaluated by observing the counts of colonies growing on the agar plate (Fig. 3a-c). As shown in Fig. 3d-f, there were no “dark” therapeutic effects of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc against P. aeruginosa with the viabilities of 99.75 ± 0.08%, 99.38 ± 0.41% and 99.95 ± 0.03% at high concentrations of 140 µM, 125 µM, and 140 µM, respectively. Under laser irradiation of 0.40 W/cm2 for 10 minutes, the remarkably photothermal antibacterial activities were observed. Along with increasing of the concentrations, PMI-3Fuc exhibited enhanced photothermal antibacterial activities for P. aeruginosa with the antibacterial ratios of 23.19 ± 0.90%, 34.57 ± 1.77%, 55.87 ± 1.06%, 66.23 ± 1.38%, 74.46 ± 0.69%, and 85.75 ± 0.43%, respectively. As expected, PMI-3Gal showed a better photothermal antibacterial activity than PMI-3Fuc due to stronger adhesion effect for P. aeruginosa and higher photothermal effect. The antibacterial ratios were 39.90 ± 0.21%, 57.47 ± 0.48%, 70.79 ± 0.33%, 77.15 ± 2.70%, 84.07 ± 0.34%, and 99.62 ± 0.12% under the concentration of 16 µM, 29 µM, 56 µM, 92 µM, and 140 µM, respectively. Moreover, the co-assemblies of PMI-3Gal@PMI-3Fuc demonstrated a significant cooperative therapeutic effect with simultaneous targeting to LecA and LecB lectins, resulting in concentration-dependent photothermal antibacterial activities of 39.07 ± 2.24%, 61.62 ± 1.64%, 80.46 ± 0.11%, 88.19 ± 0.10%, 98.14 ± 0.21%, and 100% under the concentration of 16 µM, 29 µM, 56 µM, 92 µM and 140 µM, respectively. Full eradication of P. aeruginosa was observed under the concentration of 140 µM for the assemblies of PMI-3Gal and PMI-3Gal@PMI-3Fuc, indicated that LecA lectin showed a main adhesion effect for P. aeruginosa. Under the same concentration of 93 (92) µM, the photothermal antibacterial activity of PMI-3Gal@PMI-3Fuc was 98.14%, larger than that of PMI-3Gal (84.07%), suggesting a cooperative therapeutic effect for targeting of LecB lectin.
Furthermore, the antibacterial performances of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc against multidrug-resistant P. aeruginosa isolated from mucus samples of Neurocritical care patient (WE3101, that is resistant to Imipenem and Meropenem) were evaluated (Figure S17). Along with increasing of the concentrations, as expected, PMI-3Gal showed a better photothermal antibacterial activities for multidrug-resistant P. aeruginosa than that of PMI-3Fuc.The bacterial survival rate was calculated as 90.15 ± 1.24%, 79.58 ± 1.51%, 65.26 ± 1.36%, 48.90 ± 3.72%, 33.98 ± 1.38% and 12.43 ± 1.15% under the concentration of 43 µM, 76 µM, 100µM, 126 µM, 158 µM and 165 µM (Fig. 3g). The antibacterial activities were observed for PMI-3Fuc, which exhibited the bacterial survival rate of 97.81 ± 1.92%, 92.43 ± 2.79%, 84.13 ± 2.08%, 73.77 ± 2.40%, 65.54 ± 1.71% and 50.25 ± 1.45% under the concentration of 35 µM, 65 µM, 83 µM, 120 µM, 145 µM and 155 µM (Fig. 3h), respectively. Moreover, the co-assemblies of PMI-3Gal@PMI-3Fuc became more significant after NIR irradiation. The bacterial survival rate decreased to 82.02 ± 3.05%, 65.46 ± 2.45%, 48.63 ± 0.93%, 35.90 ± 0.91%, 16.96 ± 1.02% and 2.09 ± 0.81% under the concentration of 52 µM, 88 µM, 112 µM, 149 µM, 164 µM and 172 µM (Fig. 3i), respectively.
Moreover, the PTT antibacterial efficacy of the co-assemblies of PMI-3Gal and PMI-3Fuc with different assembling ratios were investigated. As shown in Figure S18, the best antibacterial effect was observed for the 1:1 co-assembly of PMI-3Gal and PMI-3Fuc (PMI-3Gal@PMI-3Fuc), which exhibited nearly 95% inhibition ratio for P. aeruginosa after coincubation with the bacteria for 6 h under 635 nm laser irradiation for 10 min. When the assemble ratios of PMI-3Gal and PMI-3Fuc changed to 1:2 and 2:1, the PTT antibacterial inhibition ratios for P. aeruginosa were 83% and 86%, lower than that of PMI-3Gal@PMI-3Fuc. Moreover, the selectivity of the co-assemblies of PMI-3Gal@PMI-3Fuc was also studied against E. coli and S. aureus, which showed the antibacterial activities of 72% and 48%, as shown in Figure S18b-c, lower than the PTT antibacterial activity for P. aeruginosa. These results indicated that PMI-3Gal@PMI-3Fuc with 1:1 ratio demonstrated a high PTT antibacterial effect and exhibited selectively killing effect for P. aeruginosa.
In addition, the bacterial live/dead staining assay was further performed to investigate the photothermal antibacterial effect for P. aeruginosa with SYTO 9 and PI staining live and dead bacteria, respectively. As shown in Fig. 3j and S19, abundant green fluorescence labeled live bacteria (indicating no damage effect to the bacteria) for five groups: PBS groups regardless of laser irradiation, and the assemblies of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc at high concentrations of 140 µM (125 µM) without irradiation. Under 635nm (0.40 W/cm2) NIR irradiation for 10 min, live bacteria with green fluorescence decreased and dead bacteria with red fluorescence increased, both showed a concentration-dependent antibacterial activity. As shown in Fig. 3k-m, the antibacterial effects of PMI-3Fuc were calculated to be 7.81 ± 1.50%, 24.37 ± 2.11%, 44.97 ± 2.19%, and 77.48 ± 2.05% at 12, 45, 79, and 125 µM using Image J. Enhanced antibacterial activities were observed for PMI-3Gal, which exhibited antibacterial ratios of 25.57 ± 4.52%, 46.73 ± 2.59%, 72.26 ± 2.33%, and 90.96 ± 0.34% under the concentrations of 16, 56, 92, and 140 µM, respectively. As expected, the co-assemblies of PMI-3Gal@PMI-3Fuc showed the best antibacterial activities with the antibacterial ratios of 38.26 ± 0.03%, 55.63 ± 0.73%, 89.23 ± 1.45%, and 99.92 ± 0.03% at 17, 56, 93 and 140 µM, respectively. These results were consistent with colony growth on the LB agar plates.
The morphological changes of bacteria treated with PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc with and without laser irradiation were further explored. As shown in Fig. 3n, an integrated cell membrane with clear edges and smooth surface of bacteria were observed for the PBS group and the unirradiated PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc. Under laser treatment, the collapsed cell membranes for P. aeruginosa were found, and almost all the bacterial membranes were destroyed due to the PTT effects.
2.5 Lectin-targeted antibiofilm activity of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc for P. aeruginosa
Formation of bacterial biofilm is one of the most critical factors leading to drug resistance, which further prevents antimicrobial drugs from contacting bacteria, thereby delaying the healing of infected wounds. Fortunately, the formation of biofilm is related with lectins LecA and LecB presenting on the outer membrane of P. aeruginosa, which brings opportunity for the development of antibiotic-free antibacterial agents (Fig. 4a).
The antibiofilm performances of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc against P. aeruginosa biofilms were evaluated by crystal violet (CV) staining assay. Due to strong adhesion interactions of galactose with LecA and fuctose with LecB, the co-assemblies of PMI-3Gal@PMI-3Fuc showed the highest biofilm inhibition effects with the ratios of 25.08%, 34.96%, 45.94%, and 58.14% (Fig. 4b) under the concentrations of 10 µM, 20 µM, 40 µM, and 80 µM, respectively. PMI-3Gal displayed a better antibacterial effect than PMI-3Fuc. These results suggested that lectin-targeted agents showed a potential application for antibiotic-free antibiofilm effects for P. aeruginosa. It is well known that removal of mature biofilms is more difficult than the inhibition of biofilm formation due to the formation of a microbial community with polysaccharides, extracellular DNA (eDNA), lipids, and proteins. Upon increasing of the concentrations to 20 µM, 40 µM, 80 µM and 160 µM, the assemblies of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc also exhibited dispersion effects on the mature biofilms for P. aeruginosa, as shown in Fig. 4c, the trends were PMI-3Gal@PMI-3Fuc > PMI-3Gal > PMI-3Fuc. PMI-3Gal@PMI-3Fuc displayed the dispersion ratios of 28.83%, 33.06%, 39.97%, and 49.41%. PMI-3Gal with the dispersion ratios of 25.22%, 33.18%, 39.96%, and 46.80%; and PMI-3Fuc with the dispersion ratios of 24.69%, 33.05%, 38.49%, and 43.93% were observed. These results indicated that lectin-targeted agents showed the application for antibiotic-free antibiofilm effects both for mature biofilms and immature biofilms.
Moreover, antibiofilm effects were further examined with fluorescence imaging by live/dead bacterial staining. Intense green fluorescence for the control group was observed, indicated a very intact biofilm structure. When addition of 50 µM and 100 µM of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc, green fluorescence on biofilm was decreased, suggesting inhibition biofilm effect (Fig. 4d). In addition, removal effects of biofilm based on PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc with and without laser irradiation were also studied by a live/dead bacterial staining (Fig. 4e-g). With laser irradiation, the removal ratios of biofilm based on PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc increased from 38.53%, 27.14%, and 44.18–51.10%, 42.22%, and 55.83% at the concentration of 100 µM, respectively. Increasing the concentration to 200 µM, the removal ratios of biofilm based on PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc with laser irradiation were 89.73%, 75.97%, and 99.60%, which were larger than that (55.38%, 45.17% and 59.14%) of the group without laser irradiation. These results demonstrated that PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc could effectively inhibited and destroyed the formation of bacterial biofilms, showing the potential to eliminate bacteria.
2.6 In vitro cell migration and in vivo wound repair of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc
Cell migration and proliferation are critical processes for wound healing [45]. The biocompatibility of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc with L929 cells was studied by MTT assay under various concentrations. As shown in Figure S20 Near 100% cell viability was observed even at a high concentration of 100 µM, suggesting no toxicity to L929 cells. Moreover, the effects of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc on cell migration were studied by cell scratch assay at different time. The L929 cells were used to mimic wound infection in vitro. As illustrated in Figure S21, PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc all showed promoting effect on cell migration. PMI-3Gal and PMI-3Gal@PMI-3Fuc exhibited a relatively high ability to promote cell migration compared with the other groups, which showed the healing rates of 65.92 ± 0.32% and 65.69 ± 3.33% at 24 h through calculating the scratched area, both higher than that of PMI-3Fuc (51.02 ± 0.37%). Increasing of the cell incubation time to 48 h, the cell scratch coverages increased to 96.21 ± 1.06%, 82.03 ± 2.03% and 99.56 ± 0.16% for PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc, respectively. These results suggested that galactose residue was advantageous to promote cell migration, and the combination of galactose residue and fuctose residue exhibited a collaborative promotion effect.
Hemolysis is a key evaluation factor of the further application in vivo for multivalent glycoclusters. A hemolysis rate over 5% is adverse according to the standard of International Organization for Standardization (ISO) [46]. As shown in Figure S22, bright red supernatant of the Triton X-100 group was observed, indicated a serious hemolysis effect for the control group. However, the hemolysis rates of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc were lower than 3% even at a high concentration of 200 µM, suggesting good biocompatibility.
Furthermore, the wound healing effects of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc were investigated. Firstly, P. aeruginosa infected mouse whole skin wound models based on BALB/c mice were successfully established after P. aeruginosa infection, and the bacterial colonies was observed on Day 2. These mice were randomly divided into eight groups, including: PBS groups without (I) and with laser irradiation (II), PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc groups without (III-V) and with laser irradiation (VI-VIII). Treatments with PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc were implemented on the infected sites. Upon 635 nm laser irradiation, the temperatures of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc on wounds were recorded by an IR camera. As shown in Figure S23, the temperatures increased to 49.2℃, 44.7℃, 52.0℃ for PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc after irradiation for 10 min, which were higher than that of PBS group (37.9℃).
Dynamic wound healing process was photographed on dyes 2, 4 and 7. All wound sizes progressively reduced during the treatment period. As shown in Fig. 5a, it was obvious found that the wounds of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc with and without laser irradiation were smaller than those of the PBS groups with and without laser irradiation. The average unhealing areas (Fig. 5b) in PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc without laser irradiation were 46.72 ± 0.51%, 56.27 ± 2.54%, and 30.58 ± 1.25% on the 7th day, respectively, indicating that carbohydrate-lectin interactions can promote wound healing due to lectin-targeted antibacterial effects. With laser irradiation, the average unhealing areas in PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc increased to 27.79 ± 0.14%, 37.82 ± 1.84% and 16.79 ± 0.95% on the 7th day, respectively, suggesting a photothermal promoted killing bacteria activity. As the control groups, the average unhealing areas in PBS groups were 55.59 ± 0.89% (without laser irradiation) and 63.93 ± 1.15% (with laser irradiation), respectively. From the photographic images, the wound closure simulation plots (Fig. 5c) suggested that the skin regeneration of the assemblies with laser irradiation groups is obvious faster than the assemblies without laser irradiation groups, and which all faster than the control. The trend of the skin regeneration is PMI-3Gal@PMI-3Fuc > PMI-3Gal > PMI-3Fuc.
In addition, the antibacterial effects of PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc during infected wound healing process were investigated. As shown in Fig. 5d, tissue fluids were collected from the infected wound surface after treatment on dyes 2, 4 and 7. The extracted P. aeruginosa was incubated on agar plates, and the bacterial colonies were evaluated using Image J. After treated for 4 days, the bacteria colonies in PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc groups decreased to 67.68 ± 2.48%, 74.40 ± 1.71% and 57.30 ± 0.36% (Fig. 5e), respectively, which were lower than that (88.69 ± 1.09%) of PBS group. After treated for 7 days, the bacteria colonies continuously decreased to 57.03 ± 1.48%, 65.56 ± 0.17%% and 38.15 ± 0.71% (Fig. 5e) for PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc, respectively.
With laser irradiation, significantly bacteria killing effects were observed. The bacteria colonies decreased to 30.37 ± 1.14%, 37.02 ± 1.81%, and 1.68 ± 0.04% (Fig. 5e) for PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc after treatment of 7 days, respectively. These results demonstrated that synergistic therapy combined with inherent antimicrobial effect and photothermal killing activity (Fig. 5f) based on PMI-3Gal, PMI-3Fuc and PMI-3Gal@PMI-3Fuc in vivo showed potent wound healing in mice.
2.7 Histopathologic evaluations of collagen deposition, neovascularization, and inflammation microenvironment
Histopathologic evaluations of the regenerated skin after 7 d treatment provided insight into the wound healing effect. The skin sections of Hematoxylin & Eosin (H&E) staining (Fig. 6a) showed severe infection in tissues for the control groups. After 7 d treatment with the assemblies of the perylene-carbohydrate conjugates without laser irradiation, the infiltrated inflammatory cells decreased, suggesting an antibacterial effect through a mechanism of carbohydrate-lectin interactions. Moreover, an obvious reduction of the infiltrated inflammatory cells was found based on the assemblies of the perylene-carbohydrate conjugates with laser irradiation, indicating a remarkable antibacterial effect by a synergistic therapy combined with inherent antimicrobial effect and photothermal killing activity. In addition, an intact epidermis with a thick granulation tissue (Fig. 6b) was observed in the treatment groups of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc (with laser irradiation) with the thickness of 233.80 ± 18.16, 153.53 ± 7.79, and 252.05 ± 8.94 µm, respectively. The granulation tissues of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc (without laser irradiation) showed a thickness of 132.23 ± 7.63, 113.21 ± 10.22, and 164.23 ± 5.75 µm, lower than that of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc (with laser irradiation) groups. On the other hand, the thickness of granulation tissue for the control was only 90.11 ± 2.40 and 95.60 ± 3.86 µm for PBS groups with and without laser irradiation.
In addition, the organs in mice (heart, liver, spleen, lung, and kidney) were isolated and collected to investigate the biocompatibility. H&E staining images (Figure S24) showed no significant toxic side effects, and the body weight (Figure S25) of mice under the treatment process has no significant changes.
Furthermore, the collagen deposition and arrangement of the healing skin were employed by a Masson staining assay (Fig. 6c and 6d). As shown in Fig. 6c, abundant collagen with dense and organized structures were observed for the groups of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc (with laser irradiation), which were larger than that of the other groups. In them, the groups of PMI-3Gal@PMI-3Fuc exhibited a higher content of collagen in the dermis. Angiogenesis is an essential parameter for wound regeneration [47]. CD31 is a classical marker expressed in vascular endothelial cells [30]. The neovascularization for the regenerated tissue after 7 d treatment was investigated by an immunofluorescence staining method (Fig. 6e and 6f). As shown in Fig. 6e, marked red fluorescence was observed, which exhibited a relative capillary intensity of 31.22 ± 4.12 for PMI-3Gal@PMI-3Fuc group with laser irradiation. This result was higher than the groups of PMI-3Gal (16.27 ± 0.69) and PMI-3Fuc (14.26 ± 0.78) under laser irradiation. Without laser irradiation, the treatment groups of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc also showed a proangiogenic effect with the relative capillary intensities of 9.25 ± 0.44, 2.80 ± 0.10, and 10.24 ± 0.16, respectively, higher than the control group (1.00 ± 0.11). These results clearly demonstrated that the co-assemblies of PMI-3Gal@PMI-3Fuc with laser irradiation considerably promoted P. aeruginosa infected wound healing with an intact epidermis and regeneration of other appendages via abundant collagen deposition and prominent angiogenesis.
Moreover, the level of interleukin- 6 (IL-6) and tumor necrosis factor-α (TNF-α) at the wound site can reflect the level of tissue inflammation to some extent [48]. The cytokines of IL-6 and TNF-α were studied to evaluate the activation and termination of many cellular activities related to repair during wound healing. The IL-6 levels (Fig. 6g and 6h) of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc with laser irradiation were 0.05 ± 0.002, 0.19 ± 0.03 and 0.02 ± 0.002, lower than the levels (0.32 ± 0.02, 0.65 ± 0.02, and 0.23 ± 0.02) of PMI-3Gal, PMI-3Fuc, and PMI-3Gal@PMI-3Fuc without laser irradiation. These results all outperformed the control group. Similar result was also observed for the TNF-α expression. With laser irradiation, the TNF-α level (Fig. 6i and 6j) in PMI-3Gal@PMI-3Fuc group was 0.01 ± 0.002, lower than PMI-3Gal (0.05 ± 0.004), PMI-3Fuc (0.20 ± 0.10), and control (1.00 ± 0.07). These results indicated that the co-assemblies of PMI-3Gal@PMI-3Fuc with laser irradiation significantly showed a higher repair effect to promote wound healing by reducing the levels of IL-6 and TNF-α [49].