HPG, PLA-HPG and PLA-cy5 was synthesized as previously described (Deng et al., 2014; Lin et al., 2022). 100 mg PLA-HPG with or without 5 mg SKN were dissolved in 1.2 ml DMSO. The mixture was added to 40 ml deionized water using a syringe under vortex, in order to construct PLA-HPG nanoparticles (NNP). The NNP were filtered with Amico ultra-centrifuge filtration unit (100 kDa) and washed twice with deionized water. 1 vol NNP (resuspended at 25 mg/ml) was incubated with 1 vol of 0.1 M NaIO₄ for 2 min to obtain BNP, and added with 3 vol of 0.2 M ethylene glycol to quench the reaction. BNP were washed with deionized water and filtered four times with Amico ultra-centrifuge filtration unit (100 kDa), then resuspend at an appropriate concentration.

Suspensions of BNP or SKN/BNP were loaded on carbon-supported copper grids and air dried at room temperature. The samples were then stained by 2% phosphotungstic acid and processed to image using a transmission electron microscope [TEM, JEM-2010(Hr), JEOL]. The diameter and polydispersity index (PDI) of BNP and PTX/BNP were determined using a 90Plus PALS particle size analyzer.

For SKN release from SKN/BNP in vitro, 1.6 mg SKN in SKN/BNP were loaded in a dialysis bag (MWCO 3.5 kD) and immersed in 80 ml PBS (pH7.4) supplemented with 1% SDS at 37°C in a shaker. The medium was collected at a series of time points, and the amount of SKN released was quantified using an UV-Vis spectrophotometer.

SKN or SKN/BNP suspended in PBS (1.5 mg/ml) were added on a lysine-coated glass slide and incubated at 37°C for 30 min. After incubation, the slide was washed five times with PBS, and visualized. Ex vivo retention was evaluated using fresh pig skin with the same procedure.

The adhesion of BNP on bacteria biofilm was performed using PLA-cy5 loaded NNP or BNP. In brief, PLA-cy5/NNP and PLA-cy5/BNP were prepared as described (Lin et al., 2022). After biofilm formation, the PLA-cy5/NNP and PLA-cy5/BNP were added to the medium at a final concentration of 1 mg/ml, and incubated at 37°C for 1 h. Then the biofilm was washed three times with PBS and imaged using a fluorescent microscope.

For in vivo retention, SKN or SKN/BNP solution (1.5 mg/ml) were added to the wound of anesthetized C57BL/6 mice on day 3 of burn would modeling, when the wound was filled with abundant exudate. After incubation for 2 h, PBS-soaked cotton buds were used to remove non-adherent drugs. Retention of SKN or SKN/BNP on the humid wound were photographed.

Murine fibroblast L929 cells were obtained from Procell (Wuhan, China), and were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin. Cell proliferation was quantified with a Cell Counting Kit-8 (CCK-8) reagent. L929 cells were seeded into a 96-well plate at the density of 5 × 10³ cells per well and incubated for 24 h. The cells were then treated with different concentrations of SKN, BNP, SKN/BNP using DMSO as a solvent control for 24 h. After incubation, the supernatant was aspirated and premixed CCK-8 (10 μl) with medium were added to each well, and incubated for 1–3 h. Absorbance was determined at 450 nm.

Wound healing assay was performed using L929 cells and culture inserts (Ibidi) on a 24-well plate. Cells were suspended at 1.4 × 10⁶ cells/ml in DMEM supplemented with 10% FBS, and 70 μL of the cell suspension was transferred to each well of the culture-insert. After 12 h of incubation, the culture-inserts were removed, and the cells were washed thrice with PBS, and treated with SKN, BNP, SKN/BNP using DMSO as a solvent control in DMEM supplemented with 1% FBS. Photos were taken and the gap closure was measured to calculate the migration ability.

The Matrigel tube formation assay using human endothelial EA. hy926 cells was performed to assess in vitro angiogenesis (Li et al., 2022b). The EA. hy926 cells were obtained from Procell (Wuhan, China), and were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin. 250 μl BD Matrigel was pipetted into a well of a 24-well plate and incubated for 1 h at 37°C to coagulate. Then, serum-starved 7.5 × 10⁴ EA. hy926 cells were added to each well and incubated in serum-free medium with SKN, BNP, SKN/BNP using DMSO as a solvent control. After 10 h, capillary-like tubes in each well were photographed with an inverted microscope. The branch points and capillary length were analyzed by ImageJ software.

S. aureus MRSA (ATCC 43300) was purchased from the American Type Culture Collection (Manassas, VA, United States). It was cultured in 5 ml of Tryptic Soy Broth (TSB) media at 37°C overnight. The bacteria from overnight culture (2% inoculums) were transferred to 10 ml of fresh TSB, and treated with SKN, BNP, SKN/BNP respectively using DMSO as a solvent control. The mixture was incubated at 37°C with shaking at 220 rpm. To monitor the growth trend, samples were withdrawn aseptically at 3-h intervals and OD₆₀₀ of the bacteria was measured to determine bacterial growth kinetics.

The antibacterial activity was also evaluated by measuring the diameter of inhibition zones. The MRSA suspension at 10⁶ cfu/ml was inoculated onto the LB agar media and spread consistently on the plates. 8 mm diameter wells were created on the plates, and 60 µl PBS containing SKN, BNP or SKN/BNP was placed into the wells. After incubation at 37°C for 20 h, the antibacterial activity was determined by measuring the diameter of the inhibition zone.

Crystal violet was used to quantify biofilm biomass. The initial bacteria concentrations were set according to OD, which had been correlated to a standard curve of OD-cfu. The bacteria were seeded in a 96-well-plate (5 × 10⁶ cfu/well) in TSB medium supplemented with 2% glucose (w/v) and 2% sodium chloride (w/v), and treated with SKN, BNP, SKN/BNP using DMSO as a solvent control. The plate was incubated at 37°C for 24 h to establish biofilm. Then the medium with suspended bacteria were removed by pipetting, followed by washing with PBS twice. The remaining biofilm was stained with crystal violet staining solution for 5 min. Afterwards, the wells were washed twice with DI water, and 33% of acetic acid was added to fully dissolve the crystal violet. The absorbance was measured at 595 nm to determine the biofilm biomass.

For biofilm viability assay, the bacteria were seeded in a 24-well-plate (2 × 10⁶ cfu/well) in TSB medium supplemented 2% glucose and 2% sodium chloride, and treated with SKN, BNP, SKN/BNP as described above for 24 h. Afterwards, all wells were washed 3 times with PBS. The remaining biofilm was dispersed in 1 ml PBS by repeated pipetting and sonication for 30 min. Dispersed biofilm suspension was diluted in PBS, and 25 μl of each dilution was spread on LB agar plate, incubated at 37°C for 16 h. Colonies per plate were counted to determine the number of viable bacteria in the biofilm.

The animal study was approved by the Ethics Committee of Guangzhou University. The study was conducted in accordance with the local legislation and institutional requirements. 7-week-old female C57BL/6 mice were housed at room temperature (25°C) and in a light-dark cycle of 12 h. During the experiment, we used 1% pentobarbital sodium (50 mg/kg) to anesthetize the mice via intraperitoneal injection. After anesthetization and shave, a round cutaneous burn wound (φ 12 mm) was created on the back of each mouse, and the wounds were covered with a sterile plastic film. The mice were divided into 4 groups (n = 7 or 8 mice per group) and treated with PBS, SKN (30 μg/ml), SKN/BNP (containing 30 μg/ml SKN), BNP (equivalent dose of BNP, 0.83 mg/ml), respectively, on day 1, 5, 9 post-modeling. 1 × 10⁷ cfu of MRSA in PBS were inoculated on the wounds on day 2. After 3 days, obvious yellowish biofilm was seen on the wound infected with MRSA. The wound was dosed and photographed every 4 days. At the end of all procedures (15 days post-modeling), mice were euthanized using the cervical dislocation technique under anesthetization, and tissue samples were harvested. The wound skin was collected and fixed in 4% paraformaldehyde for 24 h, dehydrated, embedded in paraffin. For hematoxylin and eosin (H&E) and Masson’s trichrome staining, tissues embedded in paraffin were cut into 6 and 4 µm sections respectively, and stained following standard procedures. For immunohistochemical staining of CD31, the samples were cut into 4 μm slices. After deparaffinization, the sections underwent antigen retrieval in sodium citrate buffer (pH 6.0) using the high-pressure method. The slices were then incubated with 3% H₂O₂ for 15 min to remove the activity of endogenous peroxidase. After blocking with goat serum for 15 min, the slices were incubated with anti-CD31 primary antibody (Cell signaling technology, #77699T, 1:200) at 4°C overnight. The slices were then incubated with HRP-labeled secondary antibody at room temperature for 60 min, and visualized with DAB solution. The slices were counterstained with hematoxylin and analyzed under a microscope.

Data were shown as mean ± standard deviation. The statistical trend was analyzed by Student’s t-test and was considered as significant if p < 0.05.

As an amphiphilic copolymer, PLA-HPG self-assembled into nanoparticles (NNP) via a precipitation method in aqueous solution. Using the co-precipitation method, SKN could be incorporated into the hydrophobic core of the nanoparticles (SKN/NNP). The oxidative conversion of NNP to BNP was achieved by NaIO₄ treatment, which was reported to have no effect on the particle size and appearance of nanoparticles (Mai et al., 2022). Transmission electronic microscopy (TEM) confirmed the spherical shape of the BNP and SKN/BNP (Figure 1A). The diameters of BNP and SKN/BNP were about 29 and 33 nm, respectively, measured by dynamic light scattering (Figures 1B, C), with the average diameter of SK/BNP slightly larger than blank BNP. The polydispersity index of BNP and SKN/BNP was lower than 0.3, indicating the uniformity of the nanoparticles. The amount of SKN loaded in BNP was 3.6% ± 0.1% (g/g nanoparticles). As shown in Figure 1D, SKN displayed a steady and linear release profile. There was only 16% SKN released from SKN/BNP during the initial 24 h of incubation without burst release, and the rest of the encapsulated drug (∼85%) was slowly released over a period of 8 days in PBS supplemented with 1% SDS at 37°C. These results showed that BNP could be used as a drug delivery system for continuous and sustained drug release for SKN.

The aldehyde groups on the surface of BNP could form Schiff-base bonds with the amino groups on substances such as polylysine and proteins, resulting biological adhesion. To test whether BNP provides prolonged retention for SKN, polylysine-coated slides and pig skin were incubated with 1.5 mg/ml SKN or SKN/BNP at 32°C for 0.5 h. SKN/BNP showed rapid and robust adhesion compared with free SKN (Figures 2A, B). In addition, the adhesive effectiveness of SKN/BNP on burn wound with abundant exudate 3 days after modeling was also tested. Similarly, SKN/BNP at 1.5 mg/ml also adhered strongly to the burn wound after 2-h incubation at 32°C (Figure 2C), and even gentle wiping with PBS-soaked cotton buds can hardly remove the purple substance of SKN/BNP.

The healing of burns requires fibroblast proliferation, which serves a vital role in filling the wound and synthesizing collagen fibers involved in wound healing (Bi et al., 2020). As only ∼16% of SKN were released from SKN/BNP after 24 h of incubation (Figure 1D), SKN/BNP at higher dose of free SKN was used for comparative studies. SKN at 1 μM increased cell viability of L929 (p < 0.01, Figure 3A), however, SKN at 2 μM inhibited cell growth significantly (p < 0.001), demonstrating that SKN may induce cytotoxicity when applied at high concentration. Interestingly, BNP at the concentration of 80 and 160 μg/ml also promoted cell proliferation of L929 cells (p < 0.01), and with increasing concentrations up to 400 μg/ml, it had no influence on cell viability compared with the control group (Figure 3A). BNP at high concentrations (1 and 1.5 mg/ml) also had no toxicity in L929 cells (Supplementary Figure S1). Other studies also reported that unloaded BNP are non-toxic and did not show any toxicity at 1 mg/ml (Deng et al., 2016). SKN/BNP promoted cell proliferation in a dose-dependent manner, and its greatest effect at 20 μM was stronger than that of SKN or BNP alone (Figure 3A). In addition, 1 μM SKN, 20 μM SKN/BNP and equivalent dose of BNP (160 μg/ml) promoted the cell migration of L929 fibroblasts (p < 0.01, Figures 3B, C).

Angiogenesis is another crucial process involved in wound healing (Li et al., 2022b). To further determine their effect in angiogenesis, we investigated the tube-forming ability of EA.hy929 cells. Our results suggested both SKN (1 μM) and SKN/BNP (20 μM) possessed the ability to stimulate tube formation, as indicated by increased branch points and capillary length (p < 0.05, Figures 3D–F). We also noticed that the effect of SKN/BNP showed a slight lower trend than free SKN, which might be caused by less SKN released during shorter incubation time of 10 h.

The converting of HPG on the surface of NNP to an aldehyde-rich corona grants BNP an unprecedented bioadhesive property in the skin tissue. The adhesion of BNP on biofilm has not been investigated. 1 mg/ml PLA-cy5/NNP or PLA-cy5/BNP was incubated with mature MRSA biofilm for 1 h, and free nanoparticles were removed via washing with PBS. PLA-cy5/BNP showed substantially higher biofilm retention than PLA-cy5/NNP (Figure 4A), indicating BNP might be an ideal carrier for topical antimicrobial drug delivery in combating biofilm.

In general, SKN has been reported by to be a potent phytomedicine against gram-positive bacteria, including S. aureus, as well as against various species of lactic acid bacteria (Andújar et al., 2013). As shown in Figure 4B, SKN at 20 μM blocked the growth of MRSA over a 9-h period (p < 0.001), while SKN/BNP at 400 μM inhibited MRSA growth to a lesser extent (p < 0.001). BNP alone had little effect on the proliferation of MRSA (Figure 4B). Similar results were seen in terms of inhibition zone diameter (Figures 4C, D). BNP did not show any detectable inhibition zone, while 20 μM SKN had larger inhibition zone than 400 μM SKN/BNP (p < 0.001). This may be induced by the lower concentration of SKN released from the SKN/BNP during incubation. In addition, SKN/BNP at 400–1,200 μM showed progressive dose-dependent inhibitory activity against MRSA both in bacterial growth kinetics and diameter of the inhibitory zone (p < 0.001, Supplementary Figures S2A–C).

Crystal violet staining revealed that 20 μM SKN significantly inhibited the formation of MRSA biofilm (p < 0.001, Figures 4E, F), as reported previously (Lee et al., 2015). Surprisingly, BNP also had a significant effect on the biomass of residual MRSA biofilm (p < 0.001). In the experimental flow of biofilm staining using crystal violet, we noticed that the amount of biofilm remained almost the same before washing with PBS, however, it broke into pieces with a gentle shake and it could be easily removed from the plate. These phenomena indicated that BNP might affect the structural integrity of MRSA biofilm. SKN/BNP showed the best efficiency in reducing the biomass of MRSA biofilm (p < 0.001, Figures 4E, F). With BNP/SKN retention in the biofilm, it might exert a close-range effect on bacteria within the biofilm and avoid the shielding effect of the biofilm on exogenous free drugs. The cfu count of bacteria in remaining biofilm after different treatment also showed that SKN/BNP is significantly more potent in inhibiting MRSA biofilm than SKN or BNP individually (p < 0.001, Figures 4G, H).

`A burn wound with a 12 mm diameter was created on the back of C57BL/6 mice at Day 0. PBS, SKN (30 μg/ml), SKN/BNP (containing 30 μg/ml SKN), BNP (equivalent dose of BNP, 0.83 mg/ml) were applied to the wound on Day 1 post-modeling. Then precultured MRSA was inoculated to establish the mouse model with an infected burn wound (Figure 5A). As showed in Figure 5B, much exudate and obvious yellow biofilm existed on the wound on day 5, the biofilm turned much thicker on day 9. Treatment of SKN and SKN/BNP reduced the formation of MRSA biofilm effectively (p < 0.001), whereas BNP had little effect on the amount of biofilm in vivo (Figure 5B). Wound size in Figure 5C showed that the area of the burn wound began to reduce at least 5 days post-modeling. It also demonstrated that wound healing was faster in mice treated with SKN and SKN/BNP, with SKN/BNP exhibiting better therapeutic effect than SKN on Day 9 and Day 15 (p < 0.05). However, BNP also show little effect on wound size (Figure 5C). Consistently, histopathologic examination of the burn wound by H&E staining on Day 15 showed the existence of biofilm and profuse inflammatory infiltrates in the model group. With the treatment of SKN or SKN/BNP, the thickness of the biofilm significantly reduced, and reepithelialization almost completed in mice treated with SKN/BNP (Figure 5D). Masson staining showed almost none residual cutaneous appendages, severe collagen denaturation in mice of the model group. More collagenous fibers (stained in blue) were found after SKN and SKN/BNP treatment (Figure 5E). Notably, immunohistochemical examination demonstrated the expression of CD31, a specific endothelial marker to detect angiogenesis, were significantly elevated in mice treated with SKN and SKN/BNP, compared with the mice in the model group (p < 0.001, Figures 5F, G). The microvessel density was higher in mice treated with SKN/BNP than that treated with SKN (p < 0.05), indicating better angiogenesis with the adhesive and sustained release features of SKN from SKN/BNP (Figures 5F, G). Taken together, these results suggested that BNP based coating could enhance the wound healing ability and angiogenesis of SKN in MRSA-infected burn wound.

We assessed the in vivo toxicity of SKN, BNP and SKN/BNP using H&E staining. As showed in Supplementary Figure S3, no obvious structural abnormality was observed in the major organ of mice treated with SKN, BNP and SKN/BNP compared with those in the model group, indicating good compatibility of SKN/BNP in anti-infectious and burn wound therapies.