Synthesis and characterization of E-Au NPs

For synthesis of E-Au NPs, EGCG first self-assembled under alkaline conditions, followed by rapid EGCG-Au coordination with HAuCl4, completing the synthesis process almost instantaneously (Fig. 1A). Transmission electron microscopy (TEM) images revealed that the E-Au NPs were uniform and spherical, with an average size of 5.38 ± 0.86 nm (Fig. 1B and Fig. S2). Dynamic light scattering (DLS) analysis showed consistent size distribution of E-Au NPs in DI water, phosphate buffer solution (PBS), and Luria-Bertani (LB), indicating E-Au NPs had its good dispersibility. The particle sizes of E-Au NPs in various solutions were around 40 nm, more significant than the size observed via TEM, possibly due to the strong adhesion of NPs to the hydration surface (Fig. S3). Fourier transform infrared spectroscopy (FT-IR) revealed a blue-shift of the -OH absorption peak from 3432 cm-1 to 3420 cm-1 and a red-shift of the carbonyl (C = O) absorption peak from 1625 cm-1 to 1628 cm-1 in E-Au NPs comparison with EGCG (Fig. 1C). These changes suggest that carboxyl and hydroxyl groups likely played a role in the synthesis and stabilization of E-Au NPs. X-ray photoelectron spectroscopy (XPS) showed E-Au NPs contained characteristic peaks of O 1s, C 1s and Au 4f (Fig. S4). Precisely, in the Au 4f spectrum, peaks at 83.4 eV and 87.1 eV corresponded to the two states 4f7/2 and 4f5/2 of Au, indicating that the Au element in E-Au NPs contained both Au3+ and Au0 (Fig. 1D). This observation confirmed the formation of an Au-O bond between Au3+ and EGCG, which served as a critical aspect in the synthesis of E-Au NPs. The UV-vis spectra of E-Au NPs displayed a characteristic MPN ligand-to-metal charge-transfer band at approximately 530 nm, indicating a prevalent bis-state metal-phenolic coordination (Fig. 1E). The successful synthesis of E-Au NPs was further verified through X-ray Diffraction (XRD), revealing four additional broad peaks at 38.3°, 44.4°, 64.7°, and 77.6°, corresponding to the (111), (200), (220), and (311) planes of the Au lattice (JCPDS Card NO. 04–0784) (Fig. 1F).

Fig. 1

Synthesis and characterization of E-Au nanoparticles. (A) Schematic illustration of synthetic process of E-Au NPs. (B) TEM image of E-Au NPs, scale bar = 2 nm. (C) FT-IR spectra of EGCG and E-Au NPs. (D) High-resolution XPS spectra of E-Au NPs. (E) UV–vis absorption spectra of EGCG and E-Au NPs. (F) XRD patterns of E-Au NPs. Thermal images (G) and corresponding temperature variation (H) of E-Au NPs with various concentrations (0–200 µg/mL) under 808 nm laser irradiation (1.0 W/cm2) for 5 min. (I) Five cycles of heating/cooling photothermal curves of the E-Au NPs.

The photothermal properties of E-Au NPs was then investigated by continually monitoring temperature fluctuations exposed to 808 nm NIR laser irradiation. As shown in Fig. 1G, the temperature change of E-Au NPs under 808 nm infrared laser irradiation for 5 min was recorded using an infrared camera. It was noted that E-Au NPs exhibited a concentration-dependent temperature increase when subjected to NIR irradiation at 1.0 W/cm2. The temperature difference (ΔT) of E-Au NPs was precisely quantified as 36.0 °C, whereas it only slightly increased 3.3 °C for water (Fig. 1H). Additionally, the temperature change presented a power-density-dependent photothermal performance (Fig. S5). A comparison was also made between the temperature increase of E-Au NPs and EGCG-Ag NPs irradiated for 5 min at 1.0 W/cm2 laser. The ΔT of EGCG-Ag NPs was measured at only 5.0 °C, whereas E-Au NPs exhibited a temperature increase of 25.1 °C. Furthermore, E-Au NPs displayed stable and repeatable heating capacity over five cycles of NIR irradiation on/off, indicating their potential for long-term photothermal applications (Fig. 1I).

In vitro antibacterial and antibiofilm activity of E-Au NPs

S. aureus, E. coli, and MRSA were selected to assess the antibacterial activity of E-Au NPs. First, the bacterial growth curves were monitored using the optical density at 600 nm (OD600) at various concentrations of E-Au NPs. All these three bacterial strains, including both Gram-positive and Gram-negative bacteria, were inhibited in a dose-dependent manner while E-Au NPs administration could not completely inhibit bacterial reproduction (Fig. 2A-C, Fig. S6). The antibacterial activity of E-Au NPs under NIR irradiation was evaluated by colony-forming units (CFUs) on LB-agar plates. The antibacterial effects for S. aureus and E. coli were significantly enhanced under NIR irradiation with the increase of E-Au NPs concentration or laser irradiation time (Figs. S7-S10). After treatment with 200 µg/mL E-Au NPs plus laser irradiation (808 nm, 1.0 W/cm2, 2.5 min), both Gram-positive and Gram-negative bacteria were almost eliminated and the survival rates were only 0.06% for S. aureus, 3.38% for MRSA, and 0.02% for E. coli, respectively (Fig. 2D-G, Fig. S8-S10). Notably, the E-Au NPs under NIR irradiation antibacterial strategy exhibited excellent antibacterial ability to MRSA, which was multidrug-resistant and highly pathogenic. These results suggested that E-Au NPs provided antibacterial efficacy to both Gram-positive and Gram-negative bacteria which could be further synergetic strengthened by laser irradiation. Besides, EGCG and Au3+ also exhibited certain antibacterial activity which could not completely eradicate the MRSA (Fig. S11).

Fig. 2

In vitro antibacterial and antibiofilm activities of E-Au NPs. Growth curves of S. aureus (A), E. coli (B) and MRSA (C) upon incubation with E-Au NPs at different concentrations. The photographs (D) and survival rate (E) of MRSA treated with E-Au NPs plus laser at varied time (1.0 W/cm2, 200 µg/mL). The photographs (F) and survival rates (G) of MRSA treated with E-Au NPs (at varied concentrations of 0, 25, 50, 100, 150 and 200 µg/mL, 1.0 W/cm2, 2.5 min). 3D confocal images (H) and thickness (I) of MRSA biofilms stained by SYTO 9 after different treatments, scale bar = 50 μm. (J) Images of MRSA biofilms stained by crystal violet after different treatments. (Data are presented as the mean ± s. d., *p < 0.05, *** p < 0.001)

Biofilm plays a crucial role by providing a protective shield for bacteria during infections, so the anti-biofilm effect is essential to the practical application of antibacterial agents [38]. The effectiveness of E-Au NPs against MRSA biofilm was first verified using morphological confocal images. In Fig. 2H and I, the control and NIR laser alone groups exhibited intact and dense biofilms with strong green fluorescence signals, while E-Au NPs partially hindered biofilm formation, resulting in a fragmented structure. In E-Au NPs plus laser group, the antibiofilm effect was further enhanced, and the bacteria could barely form a steady biofilm with only sporadic microcolonies existed. Besides, the thickness of MRSA biofilm was 3.539 ± 1.415 μm with E-Au NPs treatment and the thickness significantly decreased after treatment with E-Au NPs plus laser (1.068 ± 0.771 μm) compared with the control group (8.325 ± 0.928 μm) and laser alone group (8.128 ± 1.364 μm). Considering that biofilm biomass is an important factor in biofilm formation, its reduction was evident in the E-Au NPs plus laser group, as confirmed by crystal violet staining, aligning with the findings from the morphological confocal images (Fig. 2J, Fig. S12). Moreover, the efficacy against matured biofilm, which was more important to treat the infection, was further investigated (Fig. S13). As shown in micro and macro images, E-Au NPs plus laser treatment exhibited notable destructive effect to mature biofilms with the sparsest biofilms and the destruction rate was more than 90%, indicating the excellent antibiofilm ability of E-Au NPs at both biofilm formation stages.

Potential antibacterial mechanism of E-Au NPs

Based on its excellent antibacterial properties, the possible mechanisms of the E-Au NPs were investigated. The phenolic groups EGCG in E-Au NPs were prone to undergo oxidation to form electron-deficient semiquinones in the presence of Au3+. These highly reactive semiquinones could then bind with sulfhydryl residues of proteins, resulting in the formation of quinoproteins (quinone–protein conjugates) (Fig. 3A). Although these quinoproteins have been reported to participate in several physiological processes, such as apoptosis, their potential antibacterial effects had not been further investigated. To verify and quantify the amount of quinoproteins, a redox–cycling staining assay was employed. This assay involved that quinoproteins convert colorless nitroblue tetrazolium into insoluble purple formazan in the presence of glycine. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a marked protein to chosen to detect the production of quinoprotein. Following incubation of E-Au NPs with GAPDH, the purple bands on the sodium dodecyl-sulfate polyacrylamide in agarose gel electrophoresis revealed concentration-dependent formation of quinoproteins (Fig. 3B). Subsequently, the production of quinoproteins in bacteria was further evaluated by MRSA with E-Au NPs at varying concentrations and recording the optical density at 530 nm. The results showed an increase in quinoprotein levels with higher concentrations and longer incubation time, with minimal quinoprotein production in the control and Au3+ alone group. This highlighted the crucial role of phenolic ligands in the formation of quinoproteins (Fig. 3C, D).

Fig. 3

Elucidation of the antibacterial modes of E-Au NPs in vitro. (A) Schematic illustration of the formation of quinoproteins owing to the polyphenols in E-Au NPs. (B) Qualitative analysis of quinoprotein formation after treatment with E-Au NPs at different concentrations. (C) Quantitative analysis of quinoprotein formation in MRSA during treatment with E-Au NPs. (D) Maximum rate of quinone protein production at different concentrations. (E) Bright-field and DCFH-DA staining images of ROS after different treatments, scale bar = 200 μm. (F) SEM and TEM images of MRSA with different treatments, scale bar = 400 nm. Leakage of protein (G), nucleic acids (H) and K+ (I) from MRSA with different treatments. (Data are presented as the mean ± s. d., ** < 0.01, ***p < 0.001)

In addition, the NIR laser-induced phototherapy, including both photodynamic and photothermal effects, efficiently boosted the antibacterial efficacy. As depicted in Fig. 3E and Fig. S14, there was no ROS fluorescence signal in the control group and only a few signals in the E-Au group. However, the ROS fluorescence signal was significantly enhanced with the NIR laser due to the photodynamic effect. Increased ROS levels in cells can disrupt cellular physiological processes, attack key macromolecules, and damage the cellular structure of bacterial cells. Morphological examination of MRSA treated with E-Au NPs and laser irradiation was further conducted by scanning electron microscopy (SEM) and TEM. As shown in Fig. 3F, SEM displayed intact bacteria with smooth surfaces in the control group, whereas MRSA appeared shrunken and irregular after E-Au NPs treatment, and cells were fragmented with disrupted cell structures after treatment with E-Au NPs plus laser irradiation. Similarly, TEM images revealed slightly nonuniform MRSA cytoplasm after E-Au NPs treatment, with further disruption and leakage of cytoplasm upon treatment with E-Au NPs plus laser irradiation. The incomplete cell wall indicated intracellular component leakage, such as proteins, DNA, and K+, which was corroborated through various techniques (e.g., bicinchoninic acid protein assay; Fig. 3G-I).

Next, the antibacterial mechanisms were further explored through RNA sequencing. A number of 1433 differentially expressed genes (DEGs) including 763 upregulated genes and 642 down-regulated genes compared with the control were analyzed by the volcano plot analyses (Fig. 4A). As shown in Fig. 4B, partially significantly changed genes of DEGs were listed in the heatmap. Accessory gene regulator (Agr) plays an important role in intercellular signaling quorum sensing which was decreased after E-Au NPs plus laser irradiation treatment. The down-regulation of virulence factor (Spa, Hla) and antibacterial resistance-related gene (ermC) indicated the attenuation of pathogenicity and invasiveness. Besides, the expression of genes related to maintenance (icaA, icaD) and adhesion (Clfa, Clfb) hinted at interference with biofilm formation. In contrast to control group, oxidative stress-related genes were upregulated, leading to ROS production. Further analysis of RNA sequencing through gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis, revealed enrichment in cellular processes, catalytic activities, and membrane-related pathways (Fig. 4C-F). As for KEGG analysis, DEGs were enriched in pathways of metabolic pathway, membrane transport, and cell growth and death pathways, focusing on biosynthesis metabolites enrichment analysis.

Fig. 4

Gene transcriptome assay. (A) Volcano plot analyses of DEGs treated with E-Au NPs plus laser. Down-regulated genes are in blue. Up-regulated genes are in red. Gray dots represent genes are less than 2-fold differential expression. (B) Heatmap of several differentially expressed genes associated with quorum sensing, virulence, antibacterial resistance, biofilm, adhesion and oxidative stress. GO analysis (C), GO enrichment (D), KEGG pathways analysis (E) and KEGG enrichment (F) of DEGs between the control group and E-Au group

To sum up, the results revealed that the potential antibacterial mechanism of E-Au NPs involved quinoproteins production, ROS generation from photothermal therapy, cellular structure disruption, and regulation of genes associated with biofilm formation, pathogenicity, invasiveness, and quorum sensing.

In vivo treatment effect of MRSA-infected skin wounds

The photothermal effect and in vivo therapeutic performance of E-Au NPs based treatment were initially evaluated using the MRSA biofilm infected skin wounds murine model (Fig. 5A). After the establishment of the MRSA biofilm infected skin wounds model, mice were administered with E-Au NPs or control agents on Day 1 and 3, with PBS serving as the control group for wound healing. With NIR irradiation, infrared thermographic images showed that the temperature of infected skin wounds was improved both in EGCG and E-Au groups while the improvement was more significant in E-Au group indicating the potential ability of PTT therapy (Fig. 5B). As shown in Fig. 5C, D, wounds without MRSA infection, defined as the Un-inf group, healed naturally, and its wound area recovered to 22.91% compared to 37.92% for the PBS group on day 8. Wounds treated with EGCG showed slight improvement at 30.77% but this effect was not enhanced by NIR laser irradiation. Conversely, the wound area recovered to 28.98% following E-Au NPs treatment and achieved the best therapeutic effect at 17.21% after laser irradiation, surpassing even the healing rate of the Un-inf group.

Fig. 5

In vivo therapeutic effect of E-Au NPs for MRSA-infected skin wounds model. (A) Schematic illustration of the treatment of E-Au NPs with laser in MRSA-infected skin wounds model. (B) Infrared thermographic images disposed of with NIR laser on mice dorsal cutaneous wounds. (C) Photographs of wound in MRSA-infected mice after different treatment for 7 days. The Un-inf group represents uninfected skin wounds. (D) Quantification analysis of infected wound area. Evaluation of bacterial colonies inside infected skin after different treatments (E) and relative bacteria counts (F). (G) Evaluation of infected skin wound tissue by hematoxylin and eosin (H&E) staining. Red circles indicate the formation of blood vessels, scale bar = 1 mm/250 µm. (H) Immunohistochemical staining analysis of IL-6 in the skin. (Data are presented as the mean ± s. d., *p < 0.05, **p < 0.01, ***p < 0.001)

On day 8, infected wound tissues were collected for LB-agar plate CFU counting to quantify MRSA levels. The EGCG and EGCG plus laser groups showed numerous MRSA colonies compared to E-Au NPs group. On the contrary, mice treated with E-Au NPs had significantly lower MRSA, especially with NIR laser irradiation, confirming the in vivo antibacterial potential of E-Au NPs (Fig. 5E, F). Histopathological analysis was performed to evaluate the epithelial gap and anti-inflammatory treatment effect of E-Au NPs on the MRSA-infected skin wound tissues. H&E staining images revealed that the epithelial gap was shortened and wound healing was promoted after treatment with E-Au NPs and NIR laser irradiation (Fig. 5G, Fig. S15 and Fig. S16). Besides, immunohistochemistry (IHC) was performed to assess the amount of proinflammatory, such as cytokines interleukin-6 (IL-6). As shown in Fig. 5H and Fig. S17, IL-6 was highly expressed in PBS, EGCG, and EGCG plus NIR groups while notably decreased after the treatment of E-Au NPs, indicating the significant inhibition of the inflammatory reaction of infection.

In vivo treatment effect of MRSA-infected keratitis

The in vivo treatment effect performance of E-Au NPs was further evaluated in the MRSA-infected keratitis model (Fig. 6A). After 48 h incubation of MRSA suspension on the eyes, the typical symptoms of bacterial keratitis, such as cornea edema, conjunctival congestion, opacification, and periocular purulent exudation appeared, indicating the successful establishment of the MRSA-infected keratitis model. Mice were treated with E-Au NPs twice on day 2 and day 4, with PBS and cefazolin sodium (CS) used as controls for keratitis healing. During the 16 day-long experiments, the murine cornea of the control and CS group gradually deteriorated and showed severe MRSA biofilm infection and inflammation symptoms as well as abundant corneal neovascularization, with the clinical grading scale 14.33 ± 1.16 for the control group and 13.67 ± 0.58 for the CS group on day 16 (Fig. 6B, C). When mice treated with E-Au NPs, the cornea inflammatory symptoms deteriorated with edema and neovascularization in the first 8 days while were gradually relieved in the later 8 days (clinical grading scale: 7.67 ± 1.00). Upon laser irradiation, the symptoms of keratitis further decreased and achieved the best therapeutic effect. The cornea became transparent with no neovascularization, and infection of the cornea and surrounding tissue was effectively controlled (clinical grading scale: 2.67 ± 0.58). Besides, the corneal epithelium defect was evaluated through sodium fluorescein staining. In control and CS groups, the epithelium defect of the cornea was stained for a large area while the area was smaller after E-Au treatment. There was no fluorescein staining in the E-Au plus laser group, indicating superior corneal recovery and surface regularity. Keratitis can also reduce tear secretion, which was assessed using the phenol red thread assay. Indeed, tear secretion in the control group was impaired, with a wetting length of the phenol red of less than 2 mm. After treatment with E-Au NPs plus laser, tear secretion almost fully recovered to norma, with a wetting length of approximate 5.53 mm (Fig. 6D).

Fig. 6

In vivo therapeutic effect of E-Au NPs for MRSA-infected keratitis. (A) Schematic illustration of the treatment of E-Au NPs with laser in MRSA-infected keratitis model. (B) Slit lamp images of different treatments on day 0, 4, 8, 12, 16 and sodium fluorescein staining of corneas. (C) The clinical grading scale of different treatments. (D) Quantification analysis of phenol red thread test. H&E staining of the cornea (E) and quantification analysis of the corneal thickness (F). The parts framed with the black dotted line were enlarged, scale bar = 100 μm/20 µm. Immunohistochemical staining analysis of IL-6 in the cornea (G) and quantification analysis (H), scale bar = 100 μm. (I) H&E staining of the retina, scale bar = 100 μm. (J) Measurement of IOP. (K) Cytotoxicity of HCEC incubated with E-Au NPs for 24 h with or without laser irradiation. (Data are presented as the mean ± s. d., *p < 0.05, ***p < 0.001)

Histopathological analysis was conducted to assess the impact of E-Au NPs treatment on MRSA-infected cornea tissues in terms of antibacterial and anti-inflammatory effects. As shown in Fig. 6E, F, severe inflammation with inflammatory cell infiltration and corneal edema with irregular epithelium and stromal matrix were observed in control and CS groups through H&E staining. However, following treatment with E-Au NPs plus laser, a significant decrease in inflammatory cells and disappearance of edema were noted, along with restoration of normal corneal structure. Additionally, the expression of proinflammatory cytokines IL-6 was notably higher in the control and CS groups compared to the treatment group, indicating effective inhibition of the inflammatory response during the late stage of infection (Fig. 6G, H).

On the basis of the excellent therapeutic effect of MRSA biofilm-infected skin wounds and keratitis, the local and systemic biosafety of E-Au NPs was evaluated for the potential clinical translation. Sodium fluorescein staining revealed no significant corneal epithelial defect after E-Au NPs administration compared to the control group (Fig. S18). As depicted in Fig. 6I, pathological analysis of retinal biosafety showed no apparent structural changes post-treatment. Intraocular pressure (IOP) returned to normal levels after treatment, indicating no disruption to aqueous humor circulation (Fig. 6J). On the other hand, cell viability of HCEC was 81.57% and 87.58% after E-Au NPs treatment even at the concentration of 400 µg/mL with or without laser irradiation (Fig. 6K). No significant systemic or metabolic side effects were observed after E-Au NPs treatment, as indicated by routine blood examinations and blood biochemical levels remaining within normal ranges (Fig. S19). Furthermore, H&E staining of major organs confirmed no histological abnormalities or signs of toxicity (Fig. S20). These findings collectively suggest that E-Au NPs treatment is effective for MRSA biofilm-infected keratitis with high biocompatibility, marking it promising for potential clinical applications.