3.1 Characterizations of the CSO@PM
To prepare the CSO@PM, CSO NPs were successfully synthesized by a hydrothermal method using SiO2 spheres as a sacrificial template [42] and then underwent encapsulation with PM by an ultrasound-assisted method. As shown in Fig. 2A, the CSO is synthesized as chrysanthemum-like particles with some agglomeration, which is consistent with a previous report [43]. This may be due to the change in the surface charge of silica caused by its hydrolysis under alkaline conditions [44]. Compared with bare CSO, CSO@PM has a clear core-shell structure (Fig. 2A, B, and Figure S3A, B), where the thickness of the outer shell is ~ 13 nm, indicating that the PM provides sufficient coverage for the NPs (Fig. 2B). The average zeta potential of CSO@PM is − 18.4 ± 1.7 mV, similar to that of PM (− 17.23 ± 2.5 mV) but significantly lower than that of bare CSO (6.4 ± 1.1 mV) (Fig. 2C), indicating the successful coating of the NPs with PM. The average hydrodynamic diameter of CSO@PM is 523.2 ± 14.9 nm (Fig. 2D), which is slightly higher than that of bare CSO (499.5 ± 8.9 nm). These results are consistent with a previous report [45].
The surface areas and pore volumes of CSO@PM were measured by N2 physical adsorption–desorption isotherm analysis. As shown in Figure S1A, the isotherms for CSO@PM present typical type-IV patterns, indicating the presence of mesopores in the materials. The Brunauer–Emmett–Teller surface area of CSO@PM is 138.5 m2 g− 1. The pore size distribution curve of CSO@PM derived from Barrett–Joyner–Halenda analysis revealed an average mesopore size of 4.01 nm (Figure S1B). The XRD pattern for CSO could be indexed to chrysocolla (Cu2 − XSi2O5(OH)3·XH2O, Figure S2) without any impurities. It is worth noting that the CSO has a bimodal pore-size distribution, which is due to the effects of its unique structure and hydrothermal treatment. A large pore size is suitable for capturing large biomolecules [46].
To verify whether the synthetic process influences the integrity of MPs, we carried out a proteomic analysis to categorize the quantities of proteins belonging to different cellular components based on Gene Ontology Annotation. Up to 69.1% of the membrane proteins were preserved, while 89.2% of the intracellular proteins were removed during the preparation of CSO@PM (Fig. 2E). To further confirm the retention of functional membrane proteins on the CSO@PM, Western blot analysis was performed. CSO@PM bears specific proteins of the key targeted bacteria (FPR1 and TLR4) and the inflammation mediators GPVI and CLEC-2 (Fig. 2F). The stability of CSO@PM was examined by observing the zeta potential changes over time, and CSO@PM showed negligible zeta potential change for 7 days (Figure S4). Furthermore, the NPs remain stable in PBS over 7 days at 4°C (Figure S5A), and the change in average hydrodynamic diameter for 7 days is also negligable (Figure S5B). Collectively, these results demonstrate the excellent stability of CSO@PM in biological environments.
Previous studies have demonstrated that CSO strongly adsorbs NIR light [47.48]. In our study, the UV-Vis-NIR spectra of CSO@PM in PBS showed that the CSO@PM exhibits strong absorbance at ∼808 nm (Figure S6), making it suitable for photothermal therapy with 808 nm laser irradiation. Clearly, the temperature change of the aqueous CSO@PM solution exhibits a concentration-dependent relationship, and the temperature of the aqueous CSO@PM (50 µg mL− 1) solution exceeds 65°C within 10 min at 1.5 W cm− 2, while the temperature of water without CSO@PM shows no increase (Fig. 2G, H). Furthermore, no noticeable changes in the temperature of CSO@PM are observed after undergoing six cycles of laser irradiation (1.5 W cm− 2, 10 min, Fig. 2I), indicating its excellent photothermal stability. These results indicate that CSO@PM is a promising candidate for photothermal sterilization.
3.2 In vitro targeting properties and antibacterial activity of CSO@PM
SEM was used to investigate the targeting properties of CSO@PM. As demonstrated in Fig. 3A, CSO and CSO@PM NPs are bound to the surface of P. aeruginosa, with CSO@PM group binding significantly more than CSO (Fig. 3A). The EDS element mapping technique analysis provides more supportive evidence for bacterium-targeting properties of CSO@PM (Figure S7). This may be attributed to PM expressing formyl peptide receptors, TLRs, and chemokine receptors to detect bacteria-related molecular patterns and target bacteria [31.32]. In order to further investigate the targeting properties of CSO@PM, we incubated CSO and CSO@PM with P. aeruginosa. Quantitative ICP-MS analysis showed that 23.5% of the CSO@PM adheres to the P. aeruginosa bacteria cells, while the corresponding value for CSO is only 7.69% (Fig. 3B). This demonstrates that CSO@PM has significant bacteria-targeting ability. Additionally, we also observed marked morphological changes of P. aeruginosa upon co-incubation with CSO@PM. The untreated bacteria remain fully active with typical club-like shapes and intact surfaces. However, the P. aeruginosa treated with CSO@PM under 808 NIR laser irradiation exhibit distorted morphology and wrinkled cellular walls and membranes with clear lesions and holes (Fig. 3A). The above results indicate that the PM imparts significant bacteria-targeting properties to CSO NPs. Furthermore, owing to the excellent photothermal properties of copper, CSO@PM can induce bacterial lysis under 808 nm NIR laser irradiation.
To further investigate the antibacterial activity of CSO@PM, in vitro antibacterial assays were performed. P. aeruginosa bacteria was chosen as a representative multi-drug-resistant gram-negative because it is resistant to many antibiotics, including carbapenems, which are the most commonly used antibiotics for multi-drug-resistant bacteria [49]. As expected, both CSO (50 µg mL− 1) and CSO@PM (50 µg mL− 1) exhibit very weak antibacterial activity against P. aeruginosa (Fig. 3C, D). However, when CSO and CSO@PM are exposed to 808 nm NIR laser irradiation, they exhibit strongly enhanced antibacterial effects (Fig. 3C, D). Furthermore, the CSO@PM + NIR treatment has a much higher antibacterial effect than CSO + NIR treatment (P < 0.001), where CSO@PM almost completely kills the bacteria at 50 µg mL− 1 and CSO only kills ~ 80.8% of the bacteria (Fig. 3C, D). This may be attributed to the bacterial targeting imparted by the PM, making the bactericidal efficiency of CSO@PM + NIR is much higher than that of CSO + NIR.
We also compared the antibacterial effects of CSO@PM and cephalosporins, which has been proven to have activity against an expanded spectrum of Gram-negative bacterial infections [50–52]. CFP is one of the most active of these cephalosporins and has a high activity against P. aeruginosa. Studies have demonstrated that ~ 80% of P. aeruginosa strains are killed by 128 µg mL− 1 CFP [53.54]. As shown in Fig. 3E, F, the bactericidal activity of CSO@PM plus laser irradiation against Gram-negative P. aeruginosa is significantly higher than that of CFP, and no bacterial regrowth is observed within 24 h (P < 0.001). The results clearly demonstrate that CSO@PM combined with NIR irradiation only requires a drug dose of 50 µg mL− 1, which is significantly lower than the effective antibacterial concentration of CFP (128 µg mL− 1) [54]. Overall, these results clearly indicate that CSO@PM has excellent bactericidal efficiency, which is beneficial for the treatment of drug-resistant infections.
3.3. In vitro anti-inflammatory activity of CSO@PM
In view of the benefits of platelet infusion treatment for LPS-induced sepsis [27.28] and the advantages of mesoporous materials with large specific surface areas and pore structure as adsorbents [46], we sought to determine whether CSO@PM can be used as a potential adsorption carrier for LPS.
At an LPS concentration of 1 ng mL− 1, adsorption is positively correlated to CSO@PM concentration (Fig. 4A). Furthermore, the adsorption capacity for the CSO@PM group is significantly higher than that of bare CSO (Fig. 4B). These results clearly demonstrate that CSO@PM can adsorb LPS, mainly due to the PM on its surface. This suggests that CSO@PM could adsorb bacteria-secreted LPS, thereby alleviating a series of inflammatory responses caused by LPS, and that CSO@PM shows promise as an anti-inflammatory material.
To further investigate the anti-inflammatory activity of CSO@PM, LPS-stimulated murine macrophages (RAW264.7) were used to mimic the inflammatory environment (Fig. 4C) [55]. As shown in Fig. 4D, LPS (100 ng mL− 1) induces inflammation in RAW264.7 cells, as evidenced by the upregulated mRNA expression of pro-inflammatory cytokines IL-1β and IL-6 in RAW264.7 cells upon LPS stimulation. Compared with the control group (treated with PBS), the mRNA expressions of IL-1β and IL-6 increase 5.07- and 5.22-fold upon LPS stimulation, respectively (P < 0.001). In contrast, treatment with CSO@PM (50 µg mL− 1) significantly inhibits the LPS-induced expression of IL-1β and IL-6 (P < 0.001).
Next, the effects of CSO@PM on the production of cytokines in LPS-stimulated RAW264.7 cells were investigated using ELISA. As shown in Fig. 4E, for the positive control treated with LPS (100 ng mL− 1), the concentrations of secreted IL-1β and IL-6 are 555.26 ± 23.46 and 400.18 ± 3.85 pg mL− 1, respectively, while the addition of CSO@PM (50 µg mL− 1) to the stimulated cultures suppresses IL-1β and IL-6 secretion to 263.65 ± 17.97 and 227.65 ± 17.40 pg mL− 1, respectively, which are significantly lower than the corresponding values for the LPS group (P < 0.001).
Secretion of both IL-1β and IL-6 in macrophages is used extensively as a biomarker of inflammation. They are pro-inflammatory cytokines with many functions, including those involved in chronic inflammatory reaction [56]. Thus, our results indicate that CSO@PM exerts anti-inflammatory activity by inhibiting the LPS-induced expression of pro-inflammatory cytokines IL-1β and IL-6 in RAW264.7.
3.4 Therapeutic effect of CSO@PM on P. aeruginosa-infected wounds
For in vivo studies, a multi-drug-resistant P. aeruginosa-infected murine skin wound model was utilized. P. aeruginosa is resistant to a great deal of antibiotics, including carbapenems [49]. Immediately upon the application of CSO@PM and NIR irradiation (1.5 W cm− 2, 10 min), the temperature of the wound rapidly increases to 56°C, while the temperature for the control group increases to only 38°C (Figure S8A,B). Thus, these data suggest that CSO@PM exhibit remarkable photothermal effects in vivo.
To further investigate the effects of CSO@PM in vivo. Infected wounds were treated with PBS (control), CFP (128 µg mL− 1, the effective inhibitory concentration), CSO (50 µg mL− 1), CSO@PM (50 µg mL− 1), CSO (50 µg mL− 1), or CSO@PM (50 µg mL− 1), sometimes combined with NIR irradiation (CSO + NIR, CSO@PM + NIR). The NIR irradiation (808 nm) was applied at an intensity of 1.5 W cm− 2 for 10 min. As shown in Fig. 5A-C, the appearance of wounds and quantitative closed wound areas in the control, CFP, CSO, CSO@PM, CSO + NIR and CSO@PM + NIR groups show that, in the 2 days after treatment, no obvious difference is observed in the closed wound area between all the groups (P > 0.05). However, continuous observation of the wounds revealed that the wound healing for the CSO@PM + NIR group is always significantly better than that of the control group on days 5, 7, and 9 post-surgery (P < 0.001) (Fig. 5A-C). On day 9, the CSO@PM + NIR mice exhibit a healing rate of ~ 90% (Fig. 5C).
To evaluate the actual bactericidal effect of CSO@PM in vivo, the wound tissues were harvested and homogenized to quantify the amount of residual bacteria. As shown in Fig. 5D,E, compared with the CFP and CSO + NIR groups, significantly fewer bacterial colonies are observed in the CSO@PM + NIR group (P < 0.001). These data confirm that CSO@PM combined with NIR irradiation has outstanding antibacterial activity, leading to rapid wound healing. The recovery of re-epithelialization and granulation tissue formation are key factors for evaluating wound healing [57]. Therefore, the wound tissue was collected and H&E stained to investigate the effect of CSO@PM on re-epithelialization and granulation tissue formation (Fig. 5F). As shown in Fig. 5G, H, the length of newly regenerated epidermis and the thickness of granulation tissue for the CSO@PM + NIR group are significantly higher than those for the corresponding control group (P < 0.001). In addition, compared with the corresponding control group, extensive collagen deposition is observed for the CSO@PM + NIR group (Fig. 5I).
These data show that CSO@PM in combination with NIR irradiation has significant bactericidal activity and can promote wound healing by accelerating the regenerative epithelialization, granulation thickening, and collagen deposition of infected wounds.
3.5. Therapeutic effect of CSO@PM on LPS-infected wounds
To verify the effect of CSO@PM on LPS-infected wounds, wound healing experiments were conducted in LPS-treated full-thickness skin wound model. The wounds were treated with PBS, CSO (50 µg mL− 1), PM, and CSO@PM (50 µg mL− 1). As shown in Fig. 6A, macroscopic analysis of the wound closures showed that, compared with the control group, wounds treated with CSO@PM show significantly faster closure. Continuous observation of wounds showed that on day 9, the CSO@PM-treated wound has undergone ~ 90% wound closure, while the corresponding control groups have closure rates of ~ 54–65% of the (Fig. 6A-C). The results clearly demonstrate that the healing rate for the CSO@PM group is significantly higher than that of the control (P < 0.001). Although previous studies have revealed that PM may adsorb toxins [34], the healing rate for the LPS-infected wounds treated with PM is significantly lower than that of the CSO@PM group (P < 0.001), as shown in Fig. 6C. This may be because CSO@PM not only adsorbs toxins, it also locks them away in its porous structure. This hypothesis is supported by previously reported results [34].
Again, the wound tissue was collected and H&E staining was performed to investigate the effect of CSO@PM on re-epithelialization and granulation tissue formation (Fig. 6D). As shown in Fig. 6E, F, compared with the control, the granulation tissues for the CSO@PM + NIR group are the thickest and the neo-epidermis lengths are the longest (P < 0.001). In addition, extensive collagen deposition is observed in the wounds treated with CSO@PM (Fig. 6I). These results indicate that CSO@PM has immense promise for treating LPS-infected wounds.
In order to confirm that CSO@PM adsorbs LPS to reduce the inflammatory response in LPS-infected wound to promote wound healing. RNA was extracted from the wound tissues of the control (not infected with LPS) and LPS-infected wounds treated with PBS, CSO (50 µg mL− 1), PM, and CSO@PM (50 µg mL− 1). Then, real-time fluorescent quantitative PCR was used to analyze the mRNA expression of pro-inflammatory mediators (IL-1β and IL-6). As shown in Fig. 6G, compared with the control group (treated with PBS), the expression of IL-1β and IL-6 mRNA for the LPS group is significantly increased (P < 0.001). However, CSO@PM treatment significantly inhibits the LPS-induced increase of IL-1β and IL-6 mRNA expression (Fig. 6G). In addition, we also analyzed the expression of cytokines in LPS-infected wounds by ELISA. As shown in Fig. 6H, the IL-1β and IL-6 levels for the LPS group are 704.21 ± 27.84 and 844.59 ± 15.68 pg mL− 1, respectively. However, the IL-1β and IL-6 levels are decreased to 380.13 ± 11.58 pg and 492.58 ± 14.16 pg mL− 1, respectively, by CSO@PM treatment. Thus, CSO@PM significantly inhibits the expression of inflammatory biological markers IL-1β and IL-6 [56], which may be a synergistic result of the anti-inflammatory properties of PM [27] and the porous structure of CSO [46].
In general, these data confirmed that CSO@PM inhibits the expression of IL-1β and IL-6 through LPS adsorption, thereby reducing the inflammatory response of the wound and ultimately promoting wound healing.
3.6 In vitro and in vivo biocompatibility
Biosafety is a crucial factor for an antibacterial agent. Accordingly, the in vivo and in vitro biosafety of CSO@PM were assessed. Firstly, to assess cytotoxicity in vitro, we chose NIH-3T3 fibroblasts, which are the main components of cutaneous tissues, as the cell model [58]. NIH-3T3 fibroblasts treated with CSO and CSO@PM show similar cell viabilities (> 90%) to that for the control group after 24 h, revealing that they exhibit no obvious cytotoxicity to 3T3 fibroblasts at the test concentrations (Fig. 7A). Additionally, the hemolysis rate of a nanomaterial must be less than 5% to ensure safety during intravenous administration [59.60]. As shown in Figure S9, the hemolysis rate for 500 µg mL− 1 CSO@PM (i.e., 10-times the concentration applied in vivo to treat infected wounds) is less than 5%. The results clearly show that the in vivo toxicity of CSO@PM at test concentrations is negligible.
Next, complete blood count (CBC) analysis was conducted to evaluate the hematological toxicity of CSO@PM in healthy mice. As shown in Fig. 7B-D, compared with the control group, there is no significant difference in blood routine index, white blood cell count, red blood cell count, and platelet count, which remain at normal levels (P > 0.05). To assess the effects of CSO@PM on visceral organs, serum enzyme level detection and histological assessment were performed. There are no significant changes in alanine transaminase (ALT), aspartate amino-transferase (AST), blood urea nitrogen (BUN), and creatinine (Cr) levels, which are indicators of liver and kidney function (Fig. 7E-I). No damage or appreciable abnormalities of the main organs (kidney, lung, spleen, liver, and heart) are observed 21 days after CSO and CSO@PM injection at 500 µg mL− 1. This indicates that CSO@PM will have no side effects in the process of bactericidal therapy.
Overall, the excellent biocompatibility of CSO@PM makes it a very promising antibacterial agent for biomedical applications.