NIR Light‐Driven Photocatalytic Antibacterial Hydrogels for Synergistic MRSA Biofilm Eradication and Wound Regeneration
Dong Mo, Yujia Wei, Meng Pan, Wen Chen, Yun Yang, Kang Li, Xicheng Li, Jianan Li, Qingya Liu, Hanzi Deng, Mei Zhu, Zhenpeng Zhang, Zhaolin Xiao, Zhiyong Qian
TL;DR
A new hydrogel uses light to kill drug-resistant bacteria and promote wound healing, offering a promising treatment for infected wounds.
Contribution
The first synthesis of Fe-Bi2O2-S nanoflowers for NIR light-driven antibacterial and wound regeneration applications.
Findings
Fe-BOS@C/H Gel eliminates 97% of MRSA biofilms under NIR light.
The hydrogel promotes cell proliferation and wound healing without light.
Transcriptomic analysis reveals multiple antibacterial mechanisms of the hydrogel.
Abstract
The treatment of drug‐resistant bacterial biofilm infections remains a significant challenge in clinical practice. To address this challenge, 3D oxygen vacancy (OV)‐rich iron‐doped Bi2O2S (Fe‐Bi2O2‐XS, Fe‐BOS) nanoflowers (NFs) are synthesized for the first time via an ion‐exchange method. The resulting material exhibits a small band gap, abundant OVs, and favorable charge‐transfer properties. It also shows robust photothermal performance and strong photocatalytic reactive oxygen species (ROS)‐generation ability. Fe‐BOS@C/H Gel is subsequently prepared by crosslinking hydrazide‐modified chondroitin sulfate, the Fe‐BOS NFs, and oxidized hyaluronic acid via a dynamic Schiff reaction. Fe‐BOS@C/H Gel not only shows good hemostasis and injectability, but also achieves 97% methicillin‐resistant Staphylococcus aureus (MRSA) biofilm elimination. Transcriptomic analyses reveal that Fe‐BOS@C/H…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
SCHEME 1
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7
FIGURE 8
FIGURE 9
FIGURE 10- —National Natural Science Foundation of China10.13039/501100001809
- —Natural Science Foundation of Sichuan Province10.13039/501100018542
- —Natural Science Foundation of Chongqing
- —West China Hospital, Sichuan University10.13039/501100013365
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNanoplatforms for cancer theranostics · Advanced Photocatalysis Techniques · Advanced Nanomaterials in Catalysis
Introduction
1
The human skin is the first line of defense against external microbes and pathogens [1]. However, wounds severely damage the integrity of the skin barrier, significantly increasing the risk of infection by pathogenic bacteria such as Staphylococcus aureus, Escherichia coli (E. coli), and methicillin‐resistant Staphylococcus aureus (MRSA) [2]. The rapid colonization of numerous bacteria in wounds and the formation of mature biofilms [3] can lead to a series of complications, including sepsis, tissue necrosis, organ failure, and even death in severe cases. The robust extracellular polymeric substance barrier [4, 5], which consists of polysaccharides, proteins, extracellular deoxyribonucleic acid, etc. [6], not only prevents antibiotics from penetrating biofilms, leading to antibiotic resistance, but also provides significant resistance to the host immune defense, allowing internalized bacteria to easily evade immune attacks [7]. Current clinical treatments for infected wounds include cleaning, debridement, antibiotic therapy, and dry dressings. Unfortunately, these traditional treatments are ineffective in removing drug‐resistant bacteria and biofilms and lack the ability to promote tissue regeneration [8]. Therefore, novel therapeutic strategies with antibiotic‐independent bactericidal capabilities and the ability to promote wound regeneration are urgently required to address the current treatment dilemma for wounds infected with drug‐resistant bacterial biofilms.
Hydrogels [9], which are similar in composition to the extracellular matrix (ECM) [10], have attracted significant attention because of their high water content and 3D spatial structures [11]. Hydrogel systems provide mechanical support for cell migration and proliferation, and maintain a moist environment that promotes skin wound healing [12]. Among the different types of hydrogels available, natural polymer‐based hydrogels with injectable properties stand out not only because they have excellent properties such as easy preparation, good hydrogel formation, and biocompatibility, but also because they possess a unique ability to adapt to irregularly shaped defects and closely fit wound surfaces [13, 14]. Hyaluronic acid (HA) [15], a natural macromolecule, is an important component of the ECM and has excellent properties, particularly in terms of rheology, biocompatibility, biodegradability, and nontoxicity [16]. HA promotes cell proliferation and migration, accelerates skin regeneration, and exhibits promising potential for application in wound healing [17, 18]. As a multifunctional signal regulator in human tissue, chondroitin sulfate (CS) [19, 20] can greatly increase the concentration of cell growth factors in wounds, thereby promoting wound healing [21]. However, the bioactivity and antibacterial properties of these unmodified natural hydrogels are insufficient to meet the therapeutic requirements for bacterially infected wounds. Given that hydrogel systems can also serve as carriers for therapeutic drugs, bio‐macromolecules, and nanomaterials [22], the development of novel, antibiotic‐independent, and multifunctional nanocomposite hydrogel dressings is crucial for preventing bacterial infections and promoting wound healing and skin tissue regeneration.
Antibacterial photocatalytic therapy (APCT) is an antibiotic‐independent strategy for infection control that presents numerous advantages, including broad‐spectrum sterilization, spatiotemporal controllability, and induction of bacterial resistance [23, 24]. Under light irradiation, photoexcited electrons or holes on the surface of a photocatalyst can react with nearby O_2_ or water molecules to generate different types of reactive oxygen species (ROS). These ROS can undergo peroxidation reactions with the bacterial membrane, resulting in the leakage of cytoplasmic components and, ultimately, antibacterial effects [25, 26]. Generally, most photocatalysts with wide band gaps utilize only ultraviolet‐visible (UV–Vis) light, which accounts for less than 50% of the full solar spectrum [27]. Near‐infrared (NIR) light‐driven photocatalytic reactions that generate ROS for antibacterial therapies are severely limited. Various 2D layered materials, including BiOX (X = Cl, Br, I) [28], MoS_2_ [29, 30], WS_2_ [31], and Bi_2_O_2_S [32], can be excited by NIR light owing to their narrow band gap [33, 34]. Bi_2_O_2_S [35], in particular, possesses excellent properties such as outstanding stability and photoelectric response, efficient charge separation, and a narrow band gap of ∼1.5 eV [36, 37, 38]. The narrow band gap of Bi_2_O_2_S can facilitate the capture of a wide range of the solar spectrum, thereby increasing the utilization of solar energy. However, Bi_2_O_2_S presents poor APCT capability owing to some inherent limitations of layered materials, including poor charge‐transfer and separation efficiencies.
As one of the most effective methods for regulating the electronic structure of photocatalysts, the metal doping strategy [39] is widely used to modify the properties of photocatalytic materials [40]. The principle behind this strategy involves the induction of energy‐level hybridization by introducing metal ions into the original structure of a photocatalyst, thereby achieving charge redistribution without altering its crystal structure. Chen et al. reported that the Cu‐doped BiOCl nanocomposites [41] demonstrated significantly improved light absorption, enhanced electron–hole separation, and excellent O_2_ adsorption. Therefore, Cu‐BiOCl can produce large amounts of active substances such as ROS and holes, and exhibits good photocatalytic antibacterial efficiency. Defect engineering is another effective strategy for regulating the electronic structure and chemical properties of photocatalysts. Studies have reported that introducing low‐charge elements such as Co^2+^ and Fe^3+^ into BiOCl crystals can help form vacancies in the oxygen sublattice. These surface vacancies in BiOCl act as electron traps, accelerating charge carrier separation and effectively suppressing electron‐hole recombination [42]. Given this background, we hypothesized that composite materials based on Fe‐doped Bi_2_O_2‐X_S (Fe‐BOS) and hydrogels will exhibit both strong APCT ability and antibacterial activity, potentially offering a solution to antibiotic‐resistant bacterial‐infected wounds. To the best of our knowledge, however, no such material has yet been reported in this literature.
In this study, we synthesized a photocatalytic antibacterial and pro‐regeneration hybrid hydrogel dressing to treat MRSA biofilm‐infected wounds. The hydrogel was prepared using iron‐doped Bi_2_O_2_S (Fe‐Bi_2_O_2‐X_S, Fe‐BOS) as an antibacterial guest molecule and an adipic acid dihydrazide‐modified CS/oxidized HA (CS‐ADH/OHA, C/H) natural polymer hydrogel as a host carrier to form Fe‐BOS@C/H Gel with a host–guest structure. First, novel 3D oxygen vacancy (OV)‐rich Fe‐BOS nanoflowers (NFs) were obtained by introducing Fe during the synthesis of Bi_2_O_2_S (BOS) nanosheets (NSs) (Scheme 1a). The intrinsic mechanism of Fe doping in regulating the band structure and carrier dynamics of BOS was examined to elucidate the photothermal performance and light‐driven ROS generation activity of the Fe‐BOS photocatalysts. Fe‐BOS@C/H Gel was subsequently prepared by Schiff base crosslinking between the CS‐ADH‐containing Fe‐BOS photocatalysts and OHA (Scheme 1b). The antibacterial performance of Fe‐BOS@C/H Gel toward E. coli and MRSA, along with its MRSA biofilm elimination efficiency and cell regeneration‐promoting properties, was systematically evaluated. Finally, a mouse model of MRSA biofilm‐infected wounds was established to verify the light‐driven antibacterial ability of Fe‐BOS@C/H Gel and explore the internal mechanism behind the nanocomposite hydrogel‐accelerated healing of infectious wounds (Scheme 1c).
Preparation process of (a) the Fe‐BOS photocatalysts and (b) Fe‐BOS@C/H Gel. (c) Treatment process of MRSA biofilm‐infected wounds using Fe‐BOS@C/H Gel.
Results and Discussion
2
Synthesis and Characterization
2.1
The morphology of the synthesized photocatalytic materials was characterized using transmission electron microscopy (TEM). Figure 1a shows that BOS has a 2D ultrathin sheet‐like morphology, measuring approximately 80 nm in length and width. Fe‐BOS was successfully prepared by doping the BOS NSs with Fe during their synthesis. Fe‐BOS exhibited a unique flower‐like morphology with a hierarchical structure and measured approximately 250 nm in size (Figure 1b). Changes in the size and morphology of Fe‐BOS were attributed to the successful integration of Fe into the BOS lattice, the nucleation of novel domains, and/or the promotion of the growth of pre‐existing domains. High‐resolution TEM of the Fe‐BOS NFs revealed interplanar lattice spacings of 0.266, 0.191, and 0.158 nm, corresponding to the (111), (002), and (132) planes of the orthorhombic BOS crystal, respectively (Figure 1c,d). This finding indicates that Fe incorporation does not compromise the original crystal structure of BOS. High‐angle annular dark field scanning TEM (HAADF‐STEM) and elemental mapping were performed to further confirm the existence of Fe atoms in Fe‐BOS. Fe was homogeneously distributed in the Fe‐BOS crystals despite its low content, indicating that it was successfully incorporated into the Fe‐BOS structure (Figure 1e).
TEM images of the (a) BOS NSs and (b) Fe‐BOS NFs, (c) HRTEM images, and (d) SAED pattern of Fe‐BOS crystal. (e) HAADF‐STEM image of Fe‐BOS NFs and the corresponding elemental mappings of Fe, Bi, O, and S; (f) XRD pattern of Fe‐BOS; (g) XPS survey of Fe‐BOS; (i) ESR pattern of BOS and Fe‐BOS; High‐magnification XPS spectra of (h) Fe 2p, (j) Bi 4f, and (k) O 1s in the Fe‐BOS NFs.
Time‐resolved experiments were conducted to fully understand the formation process of the Fe‐BOS NFs. The morphological evolution of the materials over time is shown in Figure S1a. Small Fe‐BOS crystals emerge and rapidly aggregate after 1 min of reaction (Figure S1b). After 10 min (Figure S1c), numerous ultrathin Fe‐BOS NSs emerge, interconnecting and gradually assembling into a flower‐like structure. The increased quantity of Fe‐BOS crystals and their self‐assembly into flower‐like structures are especially evident after 20 min of reaction (Figure S1d). When the reaction time reaches 30 min (Figure S1e), the Fe‐BOS NFs reveal a distinct morphology and continue to grow, crosslinking along the entire edges of the BOS NSs and remaining relatively dispersed. TEM further indicated that the morphology of the Fe‐BOS crystals remained unchanged as the Fe doping ratio increased, consistently exhibiting a flower‐like structure; however, the size of the crystals progressively increased (Figure S2). These findings suggest that the formation mechanism of Fe‐BOS is influenced by doping with transition metals, which alters the thermodynamics of the resulting crystals and induces an Ostwald ripening effect [43]. The continuous deposition of smaller Fe‐BOS crystals on larger Fe‐BOS NFs causes them to grow and induces a significant enhancement in the atomic mobility of the NSs, which allows them to coalesce and agglomerate more effectively. These phenomena accelerate the secondary self‐assembly of the NSs and form a 3D NF structure with a low energy state [44].
Powder X‐ray diffraction (XRD) was performed to analyze the crystalline structures of the Fe‐BOS photocatalysts. The XRD patterns (Figure 1f) revealed 11 distinct peaks at 14.87°, 24.23°, 27.42°, 29.96°, 33.66°, 45.04°, 47.35°, 53.75°, 55.26°, 58.40°, and 69.17°, corresponding to the (020), (110), (120), (040), (111), (141), (002), (112), (221), (132), and (212) phases of orthorhombic BOS NS, respectively (JCPDS No. 34–1493). The XRD pattern of Fe‐BOS exhibited diffraction peaks that corresponded precisely to those of orthorhombic BOS crystals, with no discernible new peaks. This result indicates that Fe doping does not alter the original crystalline structure of BOS. The chemical compositions of BOS and Fe‐BOS were analyzed using X‐ray photoelectron spectroscopy (XPS), which revealed the presence of Bi, O, and S (Figure 1g). In addition, the characteristic peaks of Fe were observed only in Fe‐BOS (Figure 1h), confirming the effective incorporation of Fe into BOS. Electron spin resonance (ESR) was used to identify OVs in BOS and Fe‐BOS. Figure 1i shows analogous signals centered at g = 2.002 for both photocatalysts, indicating electrons trapped by OVs. A stronger signal at g = 2.002 was observed for Fe‐BOS than for BOS, indicating more OVs in the latter. High‐resolution XPS further confirmed the higher concentration of OVs in Fe‐BOS than in BOS. The Bi 4f spectrum exhibited two prominent peaks at 158.51 and 163.80 eV, corresponding to Bi 4f_7/2_ and Bi 4f_5/2_, respectively (Figure 1j). The binding energies of Bi 4f in Fe‐BOS shifted toward higher values compared with those of pristine BOS, likely because of the reduced distance between Bi and O owing to Fe doping. The O 1s spectrum for Fe‐BOS (Figure 1k) exhibited three distinct peaks at 529.68, 531.14, and 534.78 eV, which were assigned to lattice oxygen, OVs, and surface hydroxyl groups, respectively. In addition, the ratio of the characteristic peaks corresponding to OVs significantly increased, further indicating that Fe doping promotes the formation of OVs in the crystal structure of Fe‐BOS.
Optical Properties and Band Structure
2.2
Diffuse reflectance spectroscopy (DRS; Figure 2a) was employed to assess the optical absorption characteristics of BOS and Fe‐BOS. Both materials exhibited good absorption in the visible range, with the introduction of Fe significantly improving the light absorption wavelength of the original BOS and the light absorption wavelength of Fe‐BOS extending to the far‐infrared region. The band‐gap energies (Eg) of the materials were estimated using the Kubelka–Munk function, which is defined as (𝛼hν)1/m = A(hν – Eg). The Eg values of BOS and Fe‐BOS were calculated from Tauc plots and found to gradually decrease from ≈0.96 to ≈0.75 eV (Figure 2b). Mott–Schottky (M‐S) measurements are an effective method for confirming the type and flat‐band potential (*E_fb_ *) of semiconductors (Figure 2c). The M–S plots of BOS and Fe‐BOS have a positive slope, indicating that both photocatalysts are typical n‐type semiconductors. The estimated *E_fb_
- values of BOS and Fe‐BOS were confirmed by extrapolating the lines to 1/C_2_ = 0 and found to be –0.19 and –0.40 V versus Ag/AgCl, respectively. The conduction band (CB) potential (*E_CB_ *) was generally approximately −0.2 V more negative than the *E_fb_
- of n‐type semiconductors. Thus, the *E_CB_
- values of BOS and Fe‐BOS were –0.19 and –0.40 V versus a normal hydrogen electrode (NHE, *E_NHE_ * = *E_Ag/AgCl_
-
- 0.197 V) [45]. The valence band (VB) potentials (*E_VB_ *) of BOS and Fe‐BOS were calculated using the equation *E_VB_ * = *E_CB_
-
- Eg (Figure 2d) [46] by combining the M‐S plots and Eg of BOS (0.96 eV) and Fe‐BOS (0.75 eV), and found to be 0.77 and 0.35 eV, respectively, versus NHE. These results demonstrate that Fe‐BOS is an effective NIR photocatalyst with a robust capacity for reducing CB electrons, rendering it capable of proton reduction.
Energy level structures of BOS and Fe‐BOS: (a) Diffuse reflectance spectra, (b) Tauc plots, (c) M–S plots, and (d) energy band structure diagrams. DFT calculations of the (e,f) geometric structures, (g,h) charge density distributions, and (i,j) electron localization function patterns of BOS and Fe‐BOS. Band structures and projected DOS of (k) BOS and (l) Fe‐BOS. Optoelectronic properties of BOS and Fe‐BOS: (m) TRPL spectra, (n) photoelectrochemical responses, and (o) EIS plots. (p) LSV plots of BOS and Fe‐BOS under dark (D) and light (L) conditions.
Photocatalytic Mechanism
2.3
The energy levels of photocatalysts are significantly affected by their electronic structures, which, in turn, are affected by their photocatalytic activity. Theoretical calculations enhance our understanding of the correlation between electronic structure and photocatalytic activity [47]. Density functional theory (DFT) calculations were conducted to investigate the geometries, optimized band structures, and densities of state (DOS) of the BOS and Fe‐BOS crystals. The optimized geometries are shown in Figure 2e,f. Substitution of the six‐coordinate Bi atom on the pristine orthorhombic BOS (111) surface with an Fe atom leads to a local distortion. Compared with the Bi−O bond length of 2.452 Å and the Bi−S bond length of 3.209 Å in the ac plane in the pristine BOS system, the corresponding Fe−O and Fe−S bond lengths (blue arrow in Figure 2f) decreased to 2.367 and 3.171 Å, respectively. The four in‐plane Fe−O and Fe−S bond lengths decreased owing to the ionic‐radius mismatch between the Bi and Fe atoms. Figure 2g–j reveals the charge density distribution near Fe atoms and OVs in Fe‐BOS. The numerous electrons surrounding the Fe atom suggest that the doping energy level (Figure 2d) is due to the integration of Fe into the [Bi_2_O_2_]^2+^ layers. DFT calculations were performed to elucidate the effects of Fe doping and OVs on the electronic structure of Fe‐BOS. Figure 2k,l illustrates the proposed band topologies and DOS of the two photocatalysts. Fe doping and OVs alter both the CB and VB of Fe‐BOS, resulting in the appearance of new energy levels between the CB maximum and Fermi levels. DFT calculations further indicated that the band gap decreased from approximately 1.06 eV (BOS) to approximately 0.78 eV (Fe‐BOS). The theoretical calculation results were nearly consistent with the experimental results. The underestimated band gaps of BOS and Fe‐BOS may be attributed primarily to the application of the Perdew–Burke–Ernzerhof functional [48].
Photochemical Properties
2.4
The photochemical properties of BOS and Fe‐BOS were analyzed using steady‐state and transient fluorescence spectrometry to investigate differences in their carrier dynamics. Photoluminescence (PL) spectroscopy (Figure S3) indicated that the Fe‐BOS photocatalysts have lower PL intensities than the BOS photocatalysts in the same wavelength range, indicating significant suppression of carrier recombination. Time‐resolved photoluminescence (TRPL) decay spectroscopy (Figure 2o) revealed that the PL decay profiles could be accurately modeled using a tri‐exponential equation, implying the existence of three distinct relaxation paths for the photocarriers. The average lifetimes of the BOS and Fe‐BOS photocatalysts are summarized in Table S2. The Fe‐BOS photocatalysts have a longer lifetime (318.92 ns) than the BOS photocatalysts (315.22 ns), indicating improvements in carrier‐separation efficiency in Fe‐BOS.
Transient photocurrent response curves, electrochemical impedance spectroscopy (EIS), and current–voltage (I–V) measurements were used to investigate the role of charge‐carrier transport and recombination dynamics in the photocatalytic oxidation of the Fe‐BOS photocatalysts. The transient photocurrent response curves (Figure 2m) demonstrate that the photocurrent response of the Fe‐BOS photocatalysts is much stronger than that of the BOS photocatalysts, indicating that the photoinduced electrons and holes in Fe‐BOS are effectively segregated and undergo faster charge transfer than those in BOS. EIS (Figure 2n) revealed a reduction in the arc radius of Fe‐BOS, indicating a limitation of its impedance during the charge‐transfer process. These results are consistent with the photocurrent results. I–V measurements obtained via linear sweep voltammetry (LSV, Figure 2p) were performed under dark and light conditions to further illustrate the enhanced carrier‐separation efficiency of Fe‐BOS. Whereas the I–V curves of the BOS film exhibited linear behavior, those of the Fe‐BOS film exhibited pronounced nonlinear (rectifying) behavior under light irradiation, suggesting that light irradiation significantly improves the electron‐transfer and carrier‐separation efficiencies of Fe‐BOS.
Photothermal Properties In Vitro
2.5
Previous studies have indicated that metal doping can facilitate the nonradiative recombination of photogenerated carriers in the form of thermal energy by enhancing the inharmonicity of phonon vibrations (i.e., enhancing phonon–phonon scattering) and significantly improving photothermal conversion efficiency [49, 50]. Temperature elevation curves and photothermal images (Figure 3a) were obtained to investigate the photothermal properties of the prepared photocatalysts. The temperature elevation curve of BOS was similar to that of pure water, demonstrating that pure BOS photocatalysts have negligible photothermal properties because of their low absorbance at 808 nm and large band gap. The temperature of the Fe‐BOS photocatalysts significantly increased compared with that of the BOS photocatalysts, revealing that Fe doping significantly improves the photothermal performance of Fe‐BOS. Further studies showed that the photothermal performance of Fe‐BOS strongly depends on its concentration (Figure 3b) and the laser power density (Figure 3c). The excellent photothermal stability of Fe‐BOS was demonstrated using photothermal cycling experiments (Figure 3d). Similar to other photothermal agents, the absorption wavelength of Fe‐BOS ranged from the visible to the NIR region. The absorption intensity of Fe‐BOS was linearly correlated with its concentration (Figure S4a). The mass extinction coefficient (ε) of Fe‐BOS at 808 nm was calculated from the absorption spectra and found to be 3.22 L·g^−1^·cm^−1^ (Figure S4b). The photothermal conversion efficiency (η) of the Fe‐BOS photocatalyst was calculated to be 41.44% (Figure 3e, f), which is remarkably higher than that of previously reported Bi‐semiconductor photocatalyst materials [51, 52].
(a) Temperature elevation curves and photothermal images of BOS and Fe‐BOS at a concentration of 200 µg·mL−1 after irradiation with an 808 nm laser. Temperature‐change profiles of Fe‐BOS suspensions under different (b) concentrations and (c) laser powers. (d) Photothermal periodic curve of the Fe‐BOS suspensions. (e) Photothermal curve of a 200 µg·mL−1 Fe‐BOS solution exposed to NIR light irradiation for 10 min, followed by the cessation of light exposure. (f) Linearity between the cooling time and –ln(θ) obtained from the cooling period in (e). ROS generation by the BOS and Fe‐BOS photocatalysts under NIR light irradiation: (g) Fluorescence spectra of DCFH to detect total ROS generation, (h) absorption spectra of NBT to detect •O2 − production, (i) degradation curves of DPBF to detect 1O2 production, and (j) degradation curves of MB to detect •OH− production. (k) Schematic of ROS generation by the Fe‐BOS photocatalyst under NIR light irradiation. Time‐dependent ESR signals of Fe‐BOS photocatalysts with (l) TEMP–·O2 − (m) TEMP−1O2 and (n) BMPO−•OH− as free‐radical trapping probes under NIR light irradiation.
In vitro NIR Light‐Driven ROS Generation by the Fe‐BOS Photocatalysts
2.6
Considering the excellent NIR absorption capacity and good photothermal conversion properties of Fe‐BOS, we investigated its photocatalytic activity for generating ROS. 2′,7′‐Dichlorodihydrofluorescein (DCFH) was used to measure the amount of total ROS produced during the photocatalytic reaction (Figure 3g). The results showed that the fluorescence intensity in the Fe‐BOS group was 2.1 times that in the BOS group, revealing that Fe doping significantly improves the photocatalytic activity of Fe‐BOS to generate ROS. The generation of different ROS was also examined. Nitro blue tetrazolium (NBT) can be reduced into blue formazan with maximum absorbance at 520 nm by superoxide radicals (•O_2_ ^−^) (Figure 3h). While the absorbance of both groups increased after NIR light irradiation, that of the Fe‐BOS group was 1.8 times that of the BOS group. Next, 1,3‐diphenylisobenzofuran (DPBF) and methylene blue (MB) were selected as fluorescent probes to detect singlet oxygen (^1^O_2_) and hydroxyl radicals (•OH^−^), respectively (Figure 3i,j). The absorbance of DPBF and MB in the BOS group slightly decreased, whereas that in the Fe‐BOS group significantly decreased after NIR light irradiation. Moreover, the photocatalytic activities of ^1^O_2_ and •OH^−^ in Fe‐BOS were 1.92 and 1.88 times higher than those in BOS, respectively. These results demonstrate that Fe doping dramatically improves the photocatalytic activity of BOS for generating ROS (Figure 3k). Finally, the photocatalytic activity of Fe‐BOS for generating •O_2_ ^−^, ^1^O_2_, and •OH^−^ was tested by ESR spectroscopy (Figure 3l, m, and n). The ESR signal intensities of •O_2_ ^−^, ^1^O_2,_ and •OH^−^ in the Fe‐BOS group increased with prolonged illumination, indicating that the photocatalytic ROS‐generating activity of Fe‐BOS is time‐dependent.
Synthesis and Characterization of OHA, CS‐ADH, and Fe‐BOS@C/H Gel
2.7
Fe‐BOS@C/H Gel (Figure 4a) was prepared from OHA and CS‐ADH polymers according to the synthesis routes shown in Figure S5 and S6. The chemical structures of the CS‐ADH and OHA polymers were determined by proton nuclear magnetic resonance hydrogen (^1^H NMR) and Fourier transform infrared (FTIR) spectroscopy. The ^1^H NMR spectrum of OHA indicated that, compared with that in HA, the −CHO group in the OHA polymer exhibited three novel chemical shifts at 4.82, 4.98, and 5.08 ppm owing to the influence of −OH (Figure 4b). The OHA peak at 2.0 ppm was attributed to methyl groups. By comparing the specific signal peak regions of the aldehyde and methyl groups, we determined the degree of oxidation of HA to be 50.5% (oxidation degree = integral of aldehyde groups/integral of methyl groups). The amount of ADH in the conjugate was quantitatively evaluated by integrating the characteristic signals of the methylene group of ADH (1.64 ppm) and the N‐acetyl group of CS (1.97 ppm). According to the ^1^H NMR results (Figure 4c), the degree of carboxymethyl substitution in CS‐ADH is approximately 49.8%. The FTIR spectrum of OHA exhibited an absorption peak at 1730 cm^−1^, corresponding to the symmetric vibrations of the aldehyde group (Figure 4d). The absorption peaks in the spectrum of CS‐ADH at 1644 and 1550 cm^−1^ corresponded to amide II, indicating that ADH was successfully coupled with CS (Figure 4e).
(a) Synthesis diagram of Fe‐BOS@C/H Gel. Physicochemical and structural characterization and multifunctional properties of Fe‐BOS@C/H Gel: 1H NMR spectra of (b) OHA and (c) CS‐ADH and FTIR spectra of (d) OHA and (e) CS‐ADH. (f) Optical photograph of Fe‐BOS@C/H Gel. (g) Typical SEM images of C/H Gel and Fe‐BOS@C/H Gel. (h) Variations in the G′ and G′′ of Fe‐BOS@C/H Gel as the oscillation strain varied from 1% to 300%. G′ and G′′ of (i) C/H Gel and (j) Fe‐BOS@C/H Gel samples over three high (300%)/low (1%)‐strain cycles. (k) Self‐healing performance of Fe‐BOS@C/H Gel. (l) Injectability of Fe‐BOS@C/H Gel. (m) Electrical conductivity of Fe‐BOS@C/H Gel. (n) Temperature change curves of different hydrogels after NIR light irradiation. ROS generation of different samples under NIR light irradiation: (o) UV–vis absorption spectra of NBT to detect the generation of •O2 −, (p) MB degradation curves to detect the generation of •OH–, and (q) DPBF degradation curves to detect the generation of 1O2.
The states of C/H Gel samples with different concentrations of OHA and CS‐ADH were examined. Dynamic chemical bonds can be formed through a Schiff base reaction between the aldehyde groups of OHA and the amino groups of CS‐ADH, leading to the construction of C/H Gel. Eight C/H gel samples with different concentrations of OHA and CS‐ADH were prepared. When the concentration of CS‐ADH was 8%, C/H Gel demonstrated gelatinous characteristics as the concentration of OHA increased to 4% (Figure S7a). Conversely, when the concentration of OHA was 4%, C/H Gel appeared gel‐like as the concentration of CS‐ADH increased to 4% (Figure S7b). Therefore, C/H Gel and Fe‐BOS@C/H Gel samples containing 4% CS‐ADH and 4% OHA were synthesized to form a 3D network structure for further experiments (Figure 4f). The morphology of the hydrogels was observed by scanning electron microscopy (SEM). As shown in Figure 4g, both samples exhibit highly interconnected porous structures, which can facilitate the exchange of gases and nutrients near wound sites and the absorption of blood or wound exudates. Energy dispersive spectrometry (EDS) revealed that Fe‐BOS NFs were distributed within Fe‐BOS@C/H Gel (Figure S8).
Evaluation of Rheological and Self‐Healing Properties, Injectability, and Electrical Conductivity
2.8
The mechanical, self‐healing, and viscoelastic behaviors of C/H Gel and Fe‐BOS@C/H Gel were characterized using rheological measurements. A strain amplitude sweep test (Figure 4h) showed that the intersections of the storage (G′) and loss (G″) moduli of C/H Gel and Fe‐BOS@C/H Gel were 140% and 193%, respectively, indicating that the hydrogels are at the critical point between the gel and sol states. Alternating high/low‐strain scanning tests were conducted to assess the self‐healing properties of Fe‐BOS@C/H Gel (Figure 4i,j). When the strains exerted on C/H Gel and Fe‐BOS@C/H Gel exceeded a critical point (250% and 300%, respectively), the G′ values of both gels sharply decreased and fell below their G″ values. When the strain was returned to 1%, G′ and G″ immediately recovered to their initial values. These findings indicate that the crosslinked structure of Fe‐BOS@C/H Gel can be rapidly reformed owing to the presence of dynamically reversible hydrogen and imine bonds. C/H Gel and Fe‐BOS@C/H Gel were cut into two semicircles to characterize their self‐healing capability (Figure 4k). When the two semicircular hydrogels were placed in contact with each other for 10 min, they successfully reformed into complete hydrogels. Hydrogels with injectable properties can increase the convenience of administration and help fully seal wounds. The excellent injectability of Fe‐BOS@C/H Gel was clearly demonstrated in a hydrogel injection experiment (Figure 4l). Viscosity tests (Figure S9) revealed that the viscosity of Fe‐BOS@C/H Gel decreases with increasing shear rate. This shear‐thinning property further demonstrates the injectability of Fe‐BOS@C/H Gel. Conductive hydrogel dressings produce electrical stimulation, create a microenvironment similar to a human physiological electric field for wounds, guide the migration of cells to the wound site, accelerate cell proliferation, and promote the formation of new tissues [53]. An analog circuit experiment was conducted to investigate the electrical conductivity of Fe‐BOS@C/H Gel. Compared with the pure C/H Gel group, the Fe‐BOS@C/H Gel group showed brighter light under a small lamp, with a conductivity of 2.009 S·m^−1^ because Fe‐BOS has good conductivity as a semiconductor material (Figures 4m and S10). Finally, we investigated the degradation performance of Fe‐BOS@C/H Gel. The hydrogel degradation curve showed 46% degradation within 1 d and approximately 80% degradation after 21 d (Figure S11). The release curve of the Fe‐BOS NFs in Fe‐BOS@C/H Gel (Figure S12) further indicated that the Fe‐BOS NFs were gradually released as the hydrogel degraded.
The photothermal performance of Fe‐BOS@C/H Gel was investigated. Photothermal images and temperature elevation curves are shown in Figures 4n and S13. The temperature of Fe‐BOS@C/H Gel significantly increased to 48.4°C under NIR light irradiation. By contrast, the temperatures of C/H Gel and BOS@C/H Gel barely changed. This result indicates that Fe‐BOS@C/H Gel has good photothermal properties. The ability of Fe‐BOS@C/H Gel to generate total ROS, •O_2_ ^−^, ^1^O_2_, and •OH^−^ upon photoexcitation was monitored using DCFH, NBT, DPBF, and MB as probes (Figures 4o–q and S14), with the results revealing the strong photocatalytic ability of Fe‐BOS@C/H Gel to generate ROS. The hemostatic properties of Fe‐BOS@C/H Gel were investigated using a mouse liver hemorrhage model (Figures S15,S16). The control and commercial fibrin glue groups exhibited larger bleeding areas on filter paper, with total blood losses of approximately 900 and 940 mg, respectively. When Fe‐BOS@C/H Gel was adhered to the bleeding site, bleeding decreased significantly, resulting in a total blood loss of only 100 mg. These results demonstrate that the hemostatic effect of Fe‐BOS@C/H Gel is better than that of commercial fibrin glue.
Antibacterial Ability in Vitro
2.9
An ideal dressing exhibits not only hemostatic properties, but also excellent antimicrobial properties to reduce pathogen counts and inflammation, thereby promoting wound healing. Motivated by the outstanding photothermal performance and ROS‐generation capacity of the hydrogels, we evaluated the antibacterial properties of Fe‐BOS@C/H Gel against gram‐negative E. coli and gram‐positive MRSA. Spread‐plate images of the bacteria after inoculation with the control treatment, C/H Gel, BOS@C/H Gel, and Fe‐BOS@C/H Gel in the presence (+) or absence (–) of NIR light irradiation are displayed in Figure 5a,b. Notably, the control (+) and C/H Gel (+) groups, as well as all treatment groups without NIR light irradiation, showed no antibacterial activity. By contrast, the Fe‐BOS@C/H Gel (+) group exhibited pronounced antibacterial activity against E. coli and MRSA, likely owing to the excellent ROS‐generation capacity and photothermal properties of Fe‐BOS. The antibacterial efficacy of Fe‐BOS@C/H Gel was calculated by measuring its optical density at 600 nm (OD_600_) (Figure S17). The relative antibacterial rates of the Fe‐BOS@C/H Gel (+) group were 98.50% ± 12.43% for E. coli and 97.84% ± 9.35% for MRSA. This result indicates that Fe‐BOS@C/H Gel has the best antibacterial ability among the treatments investigated. The morphology and membrane integrity of the bacteria were visualized by SEM (Figure 5c,d). The bacteria in all treatment groups without NIR light irradiation displayed a rod‐shaped (E. coli) or typical and intact spherical (MRSA) morphology. By contrast, the morphology of the group exposed to BOS@C/H Gel (+) treatment exhibited subtle alterations, characterized by slight deformations or undulations of the bacterial membrane. After Fe‐BOS@C/H Gel (+) treatment, local hyperthermia and ROS storms generated by the photothermal and photocatalytic processes effectively altered the permeability of the bacterial membranes, eventually rupturing or completely dissolving them. These results demonstrate that Fe‐BOS@C/H Gel has robust sterilization capabilities against different types of bacteria.
*Antibacterial properties of Fe‐BOS@C/H Gel in vitro. Spread‐plate images of (a) E. coli and (b) MRSA colonies after exposure to the control treatment, C/H Gel, BOS@C/H Gel, and Fe‐BOS@C/H Gel. SEM images of (c) E. coli and (d) MRSA under different treatments. (e) CLSM images and (f) dead/live bacteria ratio of treated MRSA biofilms. Dead bacteria are stained red, while live bacteria are stained green. (g) Photographs and (h) OD595 values of CV‐stained MRSA biofilms subjected to various treatments. Data are presented as mean value ± SD. (*p < 0.05; **p < 0.01; ***p < 0.001; ***p < 0.0001; NS, not significant).
Evaluation of Antibacterial Ability Against Biofilms in vitro
2.10
Bacteria in infected chronic wounds usually exist in the form of biofilms. Therefore, the removal of bacterial biofilms is crucial to promote the healing of chronic wounds. Bacterial biofilms were analyzed using the live/dead staining method and absorbance measurements at OD_600_. Confocal laser scanning microscopy (CLSM) 3D images (Figure 5e) revealed dense biofilm structures in the control (+) and C/H Gel (+) groups, as well as all treatment groups without NIR light irradiation. By contrast, the area of dead bacteria in the Fe‐BOS@C/H Gel (+) group was approximately 97% (Figure 5f). Crystal violet (CV) staining (Figure 5g) and UV–Vis analysis (Figure 5h) showed that the MRSA biofilms in the control (+) and C/H Gel (+) groups were unaffected by the applied treatments. The biofilms in the BOS@C/H Gel (+) group were only partially removed (elimination rate, ∼20%), whereas those in the Fe‐BOS@C/H Gel (+) group were nearly completely eliminated (elimination rate, > 90%). These results indicate that Fe‐BOS@C/H Gel effectively disrupts the MRSA biofilm structure through synergistic APCT and photothermal effects.
Analysis of Differentially Expressed Genes
2.11
The antibacterial mechanism of the Fe‐BOS@C/H hydrogel was investigated by RNA sequencing and gene transcription analysis of MRSA. All MRSA isolates were characterized using polymerase chain reaction (PCR), which indicated that the samples were of high quality (Figure 6a). 16S rRibosomal RNA (rRNA) sequence analysis showed that the bacteria isolated from all samples were MRSA. Generally, small RNAs can interact with messenger RNA or various functional proteins and play a key role in regulating downstream gene expression [54]. The expression and other biological processes of MRSA intracellular proteins were affected by Fe‐BOS@C/H Gel (+) treatment. A gene expression quantity box plot of log10 (TPM+1) was analyzed to determine the gene distribution and examine the overall differences in mRNA expression between the control and Fe‐BOS@C/H groups. The TPM+1 quantity box plot showed changes in gene expression dispersion after Fe‐BOS@C/H Gel (+) treatment (Figure 6b). Principal component analysis (PCA) and a heatmap of the correlation analysis (Figure 6c,d) revealed a significant difference in expression between the two groups and low variability within each group, demonstrating the reliability of the RNA sequence analysis. A total of 2758 variables were identified, of which 453 were significantly upregulated, and 423 were downregulated in the Fe‐BOS@C/H Gel (+) group compared with the control group (Figure 6e). The expression levels of differentially expressed genes (DEGs) in the Fe‐BOS@C/H Gel and control groups are presented as a heat map in Figure S18. The biological functions of the DEGs were subsequently analyzed using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). GO enrichment analysis (Figure 6f) revealed several energy‐metabolism‐related functions, such as transferase activity, regulation of metabolic process, regulation of biological process, cytochrome‐c oxidase activity, electron transfer activity, active transmembrane transporter activity, and phosphotransferase activity (alcohol group as acceptor). KEGG pathway enrichment (Figure S19) and KEGG chord diagram (Figure 6g) analyses revealed significant effects on pathways such as oxidative phosphorylation, the phosphotransferase system (PTS), ribosomes, arginine biosynthesis, quorum sensing (QS), S. aureus infection, adenosine 5′‐triphosphate (ATP)‐binding cassette transporters (ABC transporters), and necroptosis. The effects of Fe‐BOS@C/H Gel on the environmental information processing, cellular processes, genetic information processing, and metabolism of MRSA were further analyzed using heat maps. Genes associated with oxidative phosphorylation and PTS and ABC transporters were significantly downregulated, suggesting disruption of transmembrane proteins coupled to ATP hydrolysis [55]. The transcriptome results indicated that Fe‐BOS@C/H Gel (+) treatment significantly downregulates genes related to the tricarboxylic acid cycle and ATP synthesis, which aligns with mechanisms in eukaryotic cells. Therefore, we plotted a cluster heatmap of the DEGs related to oxidative phosphorylation and PTS and ABC transporters (Figure 6h,i). Compared with those in the control group, the phosphate transporter‐related genes qoxB, qoxA, mtlA, and pstC, and the oligopeptide transporter‐related gene opuBD in the Fe‐BOS@C/H Gel group were downregulated. Therefore, we speculated that the membrane transport function and osmotic pressure of the bacteria are severely disrupted after treatment with Fe‐BOS@C/H Gel. Genes (such as nikA and mprF) associated with QS and S. aureus infection were plotted as heat maps (Figure S20,S21), which showed significant reductions in their expression levels. Hence, Fe‐BOS@C/H Gel may disrupt QS, thereby weakening bacterial virulence and infectivity. Finally, we designed primers (Table S3) corresponding to the DEGs related to oxidative phosphorylation, PTS, QS, S. aureus infection, and ABC transporter regulation using quantitative PCR and validated their expression levels (Figure S22). Based on the above results, the effects of Fe‐BOS@C/H Gel appear to occur via a synergistic photothermal and photocatalytic antibacterial mechanism (Figure 6j). Fe‐BOS with photothermal properties generates localized high temperatures that affect membrane protein function, destroy bacterial membranes and DNA structures, and impart antibacterial properties. Furthermore, the large amount of ROS generated by Fe‐BOS with photocatalytic activity destroys the antioxidant defense system of bacteria and disrupts bacterial metabolism, ultimately leading to bacterial death. In summary, owing to its photocatalytic ROS‐generation capacity and photothermal properties, Fe‐BOS exhibits strong antibacterial activity and inhibits the formation of MRSA biofilms.
RNA sequence analysis of MRSA after Fe‐BOS@C/H Gel (+) treatment. (a) Identification of bacteria by agarose gel electrophoresis. (b) Box plot of the expression of DEGs in the control and Fe‐BOS@C/H Gel groups. (c) PCA analysis, (d) volcano plot, and (e) heatmap of total DEGs between the control and Fe‐BOS@C/H Gel groups. (f) GO enrichment analysis of DEGs in the Fe‐BOS@C/H Gel group. (g) KEGG chord diagram of the “Ribosome”, “Galactose metabolism”, “Arginine biosynthesis”, “ABC transporters”, “Quorum sensing (QS)” and “Oxidative phosphorylation and phosphotransferase system (PTS)” pathways. Cluster heat map of genes associated with (h) “oxidative phosphorylation and PTS” and (i) “ABC transporters”. (j) Illustration of the antibacterial mechanism of Fe‐BOS@C/H Gel with NIR light irradiation.
In vitro Cyto‐Compatibility and Hemocompatibility
2.12
We tested the cytotoxicity of Fe‐BOS@C/H Gel using L929 cells and human umbilical vein endothelial cells (HUVECs). No significant inhibitory effect on cell proliferation was observed after co‐culturing with different concentrations of Fe‐BOS@C/H Gel for 24 or 48 h (Figure 7a,b). Live/dead staining (Figures 7c, d) showed that cell proliferation was not inhibited when HUVECs were co‐incubated with the different hydrogels for 1 d. However, when the co‐incubation time was extended to 3 d, HUVECs in the different hydrogel‐treated groups showed more obvious cell proliferation than the control group. These results indicate that none of the hydrogel treatments affected cell growth and that the degradation products of C/H Gel, including HA and CS, effectively promote cell proliferation. A scratch assay was performed to evaluate the migration ability of HUVECs in the presence of C/H Gel, BOS@C/H Gel, or Fe‐BOS@C/H Gel (Figure 7e). All hydrogel‐treated groups exhibited higher migration ratios than the control group after 24 h of co‐incubation. Moreover, migration areas reached nearly 100% (Figure S23). These findings indicate that the degradation molecules of C/H Gel continuously stimulate cell migration and further verify that CS and HA can significantly promote the regeneration of fibroblasts and endothelial cells.
*Cytotoxicity and cytology evaluations of Fe‐BOS@C/H Gel. Viability of (a) L929 cells and (b) HUVECs after treatment with different hydrogels at different concentrations for 24 or 48 h (n = 4). Fluorescence images of live/dead HUVECs following calcein‐AM/PI assay. The cells were co‐cultured with the media extracts of different hydrogels and incubated for (c) 1 or (d) 3 d. (e) Scratch assay of HUVECs treated with C/H Gel, BOS@C/H Gel, and Fe‐BOS@C/H Gel. Data are presented as mean ± SD (*p < 0.05; **p < 0.01; ***p < 0.001; ***p < 0.0001; NS, not significant).
In vivo Evaluation of Fe‐BOS@C/H Gel for the Treatment of Infected Wounds Caused by MRSA Biofilms
2.13
A mouse model of wounds infected with MRSA biofilms was established to study the in vivo therapeutic effect of Fe‐BOS@C/H Gel. The entire treatment process is illustrated in Figure 8a. Initially, the mice were randomly divided into four groups and received different treatments with or without NIR light: control (0.9% NaCl), C/H Gel, BOS@C/H Gel, and Fe‐BOS@C/H Gel. After NIR light irradiation, the temperature at the wound site was recorded using a thermal imaging camera. The temperature of wound sites in the control (+), C/H Gel (+), and BOS@C/H Gel (+) groups did not change significantly, reaching only 34.3, 34.5, and 37.8°C, respectively (Figure 8b). By contrast, the temperature of wounds in the Fe‐BOS@C/H Gel (+) group reached 48.6°C after 10 min. This result confirms that Fe‐BOS@C/H Gel exhibits a highly efficient photothermal therapeutic effect. Images of the mouse wound sites were captured on days −3, 0, 2, 4, 6, and 8 to record changes in wound area in each group. In the absence of NIR light irradiation, no significant differences in wound size were observed between the control and other treatment groups. However, when all treatment groups were irradiated with NIR light, the BOS@C/H Gel (+) and Fe‐BOS@C/H Gel (+) groups showed potently accelerated wound healing on days 2, 4, 6, and 8 (Figure 8c). The healing rate of the BOS@C/H Gel (+) group was significantly higher than that of the C/H Gel (+) group on days 2, 4, 6, and 8. The Fe‐BOS@C/H Gel (+) group exhibited the best wound‐closure performance at every time point considered, with final wound‐closure rates of 97.29% ± 0.31% on day 8 (control (+) group, 66.82% ± 2.57%; control (−) group, 65.40% ± 2.36%) (Figure 8d). Body weight measurements showed that, with the exception of those in the Fe‐BOS@C/H Gel (+) group, mice in all other treatment groups lost weight (Figure 8e). These results suggest that, compared with C/H Gel and BOS@C/H Gel, Fe‐BOS@C/H Gel significantly promotes the healing process of MRSA biofilm‐infected wounds under NIR light irradiation.
*Wound healing and regeneration in a mouse model of MRSA biofilm–infected wounds. (a) Schematic illustration showing the experimental timeline. (b) Thermal photographs of different treatment groups under NIR light irradiation. (c) Photographs and (d) relative wound areas of MRSA biofilm‐infected wounds under different treatments (n = 4). (e) Statistics of mouse body weight throughout the treatment period. (f) Photographs and (g) quantification of the number of residual MRSA adhering to peripheral soft tissues on postoperative day 3 (n = 5). Data are presented as mean ± SD (*p < 0.05; **p < 0.01; ***p < 0.001; ***p < 0.0001; NS, not significant).
The in vivo antibacterial properties of Fe‐BOS@C/H Gel were evaluated by analyzing residual bacteria. Bacterial plate counting showed notable reductions in bacterial colonies in the Fe‐BOS@C/H Gel (+) group, with a corresponding antibacterial rate of 94.81% ± 1.32%, indicating significant bactericidal capacity (Figure 8f,g). Histological staining was performed to examine the extent of bacterial infection at the wound site. Giemsa staining (Figure 9a) revealed a large amount of MRSA (black arrows) remaining in all treatment groups without NIR light irradiation, indicating the continued presence of severe bacterial infection at the wound site. By comparison, the number of bacteria at the wound site was slightly lower in the control (+) and C/H Gel (+) groups. MRSA was barely observed in wound tissues in the Fe‐BOS@C/H Gel (+) group, further demonstrating the excellent in vivo antibacterial performance of Fe‐BOS@C/H Gel.
Histological analysis of MRSA biofilm‐infected wounds in different treatment groups. (a) Giemsa staining on day 3 (the black arrows point to bacteria) (b) H&E and (c) Masson staining on day 8. Scale bars: (a) photographs = 1 mm, enlarged image = 150 µm; (b,c) photographs = 2 mm, enlarged image = 150 µm.
Histological experiments were performed to examine the regenerative effects of Fe‐BOS@C/H Gel on skin tissue. Hematoxylin and eosin (H&E) staining (Figure S24) revealed obvious tissue infection, an acute inflammatory response, and neutrophil infiltration in the skin wound tissues of all groups after 3 d of treatment; however, the dermal gap in the Fe‐BOS@C/H Gel (+) group was significantly reduced. After 8 d of treatment (Figure 9b), the control (+), C/H Gel (+), and all treatment groups without NIR light irradiation showed obvious acute inflammation and a large dermal gap. By contrast, the BOS@C/H Gel (+) and Fe‐BOS@C/H Gel (+) groups showed gradually reduced inflammation and a narrow dermal gap. Notably, the Fe‐BOS@C/H Gel (+) group exhibited the lowest levels of inflammation and the smallest dermal gaps. Masson staining of skin tissues on days 3 and 8 showed significantly increased collagen contents in all hydrogel‐treated groups compared with those in the control group (Figures 9c; S25). This result verifies that the degradation products of C/H Gel, such as CS, promote collagen regeneration in the skin tissue [56]. Among the treatments investigated, the Fe‐BOS@C/H Gel (+) group showed the highest level of collagen deposition. These results reveal that Fe‐BOS@C/H Gel has good antibacterial and anti‐inflammatory properties and optimal collagen regeneration‐promoting activity. Thus, Fe‐BOS@C/H Gel has good application prospects for wound healing.
Staining with cytokine indices such as CD31, Ki67, and collagen I (COL‐I) was conducted to investigate the effect of Fe‐BOS@C/H Gel on accelerating skin wound healing. CD31, a specific marker of vascular endothelial cells, is a key indicator of neovascularization in wounds [57]. After 3 d of treatment, the expression of CD31 in wound tissues obtained from the Fe‐BOS@C/H Gel (+) group was significantly higher than that in wound tissues obtained from other treatment groups; more capillaries were also observed (Figure 10a). These results indicate that Fe‐BOS@C/H Gel can significantly promote angiogenesis under NIR light irradiation. During the proliferative phase of wound healing, cell proliferation and epithelial regeneration, in addition to angiogenesis, are observed. Ki67, an indicator of cell proliferation, was used to assess the tissue‐regenerative capacity of the hydrogels [58]. Immunofluorescence images of the Ki67‐stained tissues are shown in Figure 10b; in these images, the Ki67 protein, which is located in the nucleus, is stained red. The number of Ki67‐positive cells in the Fe‐BOS@C/H Gel (+) group increased significantly on day 8, indicating that the rapid proliferation of granulation tissue promotes faster wound healing. Fibroblasts achieve epidermal remodeling by regulating COL‐I expression. During wound healing, fibroblasts gradually secrete more COL I, providing suitable conditions for fibroblast and epidermal cell proliferation [59]. According to the immunofluorescence staining images shown in Figure 10c, cells from the Fe‐BOS@C/H Gel (+) group show brighter green fluorescence than those from the control and other hydrogel‐treated groups, indicating that Fe‐BOS@C/H Gel significantly promotes the rapid proliferation of fibroblasts and epidermal cells in the wound, thereby accelerating epidermal remodeling.
Tissue immunofluorescence staining of wound tissues in different hydrogel groups after 8 d of treatment. (a) CD31 staining on day 8 (the black arrows point to bacteria). Images of immunofluorescence staining for (b) Ki67 (red) and (c) COL‐Ι (green) on day 8. Scale bars: (a) photographs = 100 µm; (b,c) photographs = 70 µm.
Biosafety
2.14
Biosafety is an important property of the antibacterial agents used in biomedical applications. Blood chemistry and H&E histological analyses of five organs and tissues (liver, heart, kidney, spleen, and lung) were performed to investigate the in vivo biosafety of Fe‐BOS@C/H Gel. No significant differences in blood chemistry parameters (Figure S26) were observed between the control and hydrogel‐treated mice. H&E staining analysis of major organs after 3 d of treatment with different hydrogel dressings did not show an inflammatory response (Figure S27). These results demonstrate that Fe‐BOS@C/H Gel dressings do not cause long‐term tissue damage or toxic effects, thereby revealing their potential for future biomedical applications owing to their good biosafety.
Conclusion
3
We successfully developed a nanocomposite hydrogel dressing (Fe‐BOS@C/H Gel) to effectively eliminate MRSA biofilms and promote wound regeneration and healing. First, we synthesized novel 3D Fe‐BOS NFs. Fe doping decreased the bandgap of the Fe‐BOS photocatalysts, increased the number of OVs, inhibited the recombination of photogenerated electron–hole pairs, and effectively promoted charge transfer. Fe‐BOS exhibited excellent photothermal performance and strong photocatalytic activity for ROS generation under NIR light irradiation. We subsequently prepared Fe‐BOS@C/H Gel in situ through a dynamic Schiff reaction by crosslinking the CS‐ADH‐containing Fe‐BOS photocatalysts and OHA. Fe‐BOS@C/H Gel exhibited good hemostasis, injectability, and antibacterial and pro‐proliferative properties, and can fill wounds infected with MRSA biofilms in situ. Fe‐BOS@C/H Gel effectively eliminated MRSA biofilms owing to its excellent photothermal ability and photocatalytic ROS‐generation capacity. The CS and HA in Fe‐BOS@C/H Gel promoted wound collagen deposition, angiogenesis, and wound healing. This study provides an innovative solution that utilizes the synergistic strategy of light‐driven antibacterial performance and pro‐regeneration to treat drug‐resistant bacterial biofilm‐infected wounds, thereby promoting the application of photocatalytic materials and nanocomposite hydrogels in the biomedical field.
Experimental Section
4
Synthesis of Fe‐BOS NFs
4.1
Fe‐BOS NFs were synthesized using a facile one‐step liquid‐phase ion‐exchange strategy [60]. In a typical synthesis, different molar ratios of Bi(NO_3_)3⋅5H_2_O and Fe(NO_3_)3⋅9H_2_O (Table S1) were dissolved in 40 mL of deionized water to form a reaction solution. Next, 2 mL of SC(NH_2_)2 (19 mg·mL^−1^) was added to the reaction solution, followed by continuous stirring for 10 min. Then, 18 mL of EDTA⋅2Na (82.72 mg·mL^−1^) was added to the mixture. After sufficient stirring for 10 min, the mixed solution was dropped into 20 mL of KOH (30 mg·mL^−1^) and NaOH (80 mg·mL^−1^). The mixed solution was vigorously stirred for 30 min and allowed to stand for 12 h. Finally, the precipitates were collected by centrifugation, washed twice with deionized water, and surface‐modified with polyvinylpyrrolidone to obtain Fe‐BOS NFs as the final product. Bi_2_O_2_S (BOS) NSs were synthesized following the same procedure without the addition of Fe(NO_3_)3·9H_2_O.
Photothermal Effects of Fe‐BOS NFs
4.2
Different samples were irradiated with NIR light, and their photoactivated heating curves were monitored using a thermal infrared camera (E6xt, FLIR, USA) and a digital thermometer (Checktemp1‐H198509, HANNA, Italy) in phosphate‐buffered saline (PBS). The influence of Fe‐BOS concentration and laser power intensity on the photothermal property of the NFs was investigated by exposing 200 µg·mL^−1^ of different samples (PBS, BOS, and Fe‐BOS) to NIR light (808 nm) irradiation with a laser power of 1.0 W·cm^−2^ for 10 min; 200 µg·mL^−1^ Fe‐BOS samples to NIR light (808 nm) irradiation with laser powers of 0.5, 1.0, or 1.5 W·cm^−2^ for 10 min; and Fe‐BOS samples with different concentrations (50, 100, 200, 400 µg·mL^−1^) to NIR light (808 nm) with a laser power of 1.0 W·cm^−2^ for 10 min. Photothermal stability was assessed using five laser on/off cycles. All temperature changes were recorded every 30 s using a thermal infrared camera.
Generation of Total ROS by Fe‐BOS NFs
4.3
DCFH was used as a fluorescent probe to detect total ROS levels [61]. First, 0.1 mg of each sample was dispersed in 0.95 mL of PBS and then added with 0.05 mL of the DCFH solution (0.02 mg·mL^−1^). Subsequently, the mixed solution was irradiated with an 808 nm laser with a laser power density of 1 W·cm^−2^ for 10 min. Finally, 0.9 mL of the supernatant was centrifuged, and fluorescence spectra were detected from 300 to 700 nm using a fluorescence spectrophotometer.
Generation of •O2
− of Fe‐BOS NFs
4.4
Nitroblue tetrazolium (NBT, DMSO solution) was used as a chromogenic agent for the detection of superoxide radicals (•O_2_ ^−^) [62] 0.1 mL of the samples (1 mg·mL^−1^) was mixed with 0.9 mL of NBT (25 µm) solution, and the control group was 0.1 mL of PBS buffer (pH = 7, 0.2 m) without samples. Therefore, the relative amount of •O_2_ ^−^ generated by the photocatalyst after 10 min of near‐infrared light (808 nm, 1.0 W·cm^−2^) irradiation was calculated based on the absorbance of the mixed solution. Finally, 0.9 mL of the supernatant was centrifuged, and the absorbance of the as‐generated formazan (FM) was then quantified by UV–Vis absorption spectroscopy. For further verifying the production of •O_2_ ^−^, 5‐tert‐butoxycarbonyl 5‐methyl‐1‐pyrroline N‐oxide (BMPO), as a spin trapping agent for •O_2_ ^−^ , and Fe‐BOS (0.1 mg·mL^−1^) was mixed. Then, ESR was employed to test and record the •O_2_ ^−^ signal under the irradiation of 808 nm NIR light every 10 min.
Generation of •OH− by Fe‐BOS NFs
4.5
The generation of •OH^−^ in the samples was determined using the MB method [63]. Briefly, a sample suspension (0.3 mg·mL^−1^, 0.985 mL) was mixed with MB solution (0.09 mg·mL^−1^, 0.015 mL). Thereafter, the mixed solution was subjected to 808 nm NIR light irradiation (1.5 W·cm^−2^) for 15 min. The absorbance of the mixed solution at 560 nm was quantified using UV–Vis absorption spectroscopy. The production of •OH^−^ was verified by mixing 5,5‐dimethyl‐1‐pyrroline N‐oxide, as a spin trapping agent for •OH^−^, and Fe‐BOS (0.1 mg·mL^−1^). ESR was subsequently conducted to record •OH^−^ signals under 808 nm NIR light irradiation at 10 min intervals.
Generation of 1O2 by Fe‐BOS NFs
4.6
DPBF, a highly specific fluorescent probe, was reacted with ^1^O_2_ to form an internal peroxide, which was subsequently decomposed into 1,2‐dibenzoylbenzene without the characteristic UV–vis absorption peak at 440 nm [64]. First, 10.0 mg of DPBF powder was dissolved in 20 mL of DMSO to obtain a DPBF solution. Then, 0.1 mg of different samples was dispersed in 0.97 mL of DMSO as the test solution and added with 0.03 mL of the DPBF solution. The mixed solution was irradiated with an 808 nm laser with a laser power density of 1.0 W·cm^−2^ for 10 min. Finally, 0.8 mL of the supernatant and 1 mL of PBS were mixed and transferred to a quartz dish. The UV–Vis absorption spectra of DPBF in the range of 250–700 nm were measured. The production of ^1^O_2_ was verified by mixing 2,2,6,6‐tetramethylpiperidine‐1‐oxyl (TEMP), as a spin trapping agent for ^1^O_2_, and Fe‐BOS (0.1 mg·mL^−1^). ESR was employed to record ^1^O_2_ signals under 808 nm NIR light irradiation at 10 min intervals.
Preparation of Fe‐BOS@C/H Gel
4.7
The synthesis and characterization of OHA and CS‐ADH are discussed in the Supporting Information. CS‐ADH was dissolved in PBS (pH = 7.2) to obtain a 16% (w/v) CS‐ADH solution, and OHA was dissolved in PBS (pH = 7.2) to form an 8% (w/v) OHA solution. Subsequently, 0.05 mL of Fe‐BOS solution (1.0 mg·mL^−1^) and 0.05 mL of CS‐ADH solution (16%, w/v) were added to 0.1 mL of OHA (8%, w/v). After intense shaking, a Fe‐BOS@CS‐ADH/OHA nanocomposite hydrogel (Fe‐BOS@C/H Gel) was obtained. CS‐ADH/OHA hydrogel (C/H Gel) was prepared in the same manner using 0.05 mL of PBS instead of 0.05 mL of Fe‐BOS solution.
Antibacterial Performance of Fe‐BOS@C/H Gel
4.8
The antibacterial performance of the hydrogel was evaluated using the spread plate method and scanning electron microscopy. Detailed experimental procedures are provided in the supporting information.
The Clearing Capacity of Fe‐BOS@C/H Gel Toward Bacterial Biofilm
4.9
CV staining, UV–Vis measurement, and live/dead bacterial staining experiments were conducted to evaluate the ability of the hydrogel to remove bacterial biofilms. Detailed experimental procedures are provided in the supplementary information.
Cytotoxicity Assay In Vitro
4.10
L929 cells and HUVECs were cultured in medium supplemented with 10% fetal bovine serum (JYK‐FBS‐301, Inner Mongolia Jinyuankang Biotechnology Co., Ltd.) and 1% penicillin–streptomycin solution (HyClone) in a humidified atmosphere of 5% CO_2_ at 37°C. The medium was changed every 2 d. Cytotoxicity tests were performed using an indirect contact method. Cells were seeded into the lower chambers of a 24‐well Transwell chamber (NEST Biotechnology). After 1 d of culture, the upper chamber containing different hydrogels was placed in the lower chambers, and further culture was conducted for 24 or 48 h. Then, 0.01 mL of CCK‐8 (Life‐iLab, China) was added to each well, followed by incubation for another 2 h. The absorbance of each well was measured at a wavelength of 450 nm using a microplate reader. Cell viability (X) was calculated using Equation (1), in accordance with ISO 19003–5.
where OD 1 is the mean absorbance of the experimental sample and negative control groups; and OD 2 is the mean absorbance of the positive control group.
Live/Dead Staining Assay
4.11
Live/dead staining assay was used to visually assess cell viability. Briefly, HUVECs were seeded in the lower chambers of a 24‐well Transwell chamber at a density of 5 × 10^3^ cells·well^−1^ and cultured in a humidified 5% CO_2_ incubator at 37°C for 24 h. Then, the upper chamber containing different hydrogels was added to the lower chambers. After incubation for 24 h, the upper chamber and medium were removed. The cells were subsequently stained using a Hoechst 33342/PI Double Stain Kit (Solarbio, Cat#CA1120) for 15 min and imaged using an inverted fluorescence microscope.
Cell Scratch Experiment
4.12
HUVECs were cultured in 24‐well plates (NEST Biotechnology) at a density of 10^5^ cells·well^−1^ and allowed to form a monolayer. Then, 200 µL pipette tips were used to scratch two parallel lines in each well. The wells were rinsed gently with PBS to remove cell debris. The cells were grown in serum‐free medium, treated with various hydrogels, cultured with an Incucyte S3 live cell analyzer, and photographed once every 12 h. Migration area ratios (%) were analyzed using ImageJ software according to Equation (2).
where M 0 and M t represent the initial and healed scratch areas, respectively.
Evaluation of the Mouse Model of Infected Wound caused by MRSA Biofilms in Vivo
4.13
All animal experiment protocols were approved by the Ethics Committee of the Animal Experimental Center of the State Key Laboratory of Biotherapy of Sichuan University (SKLB20240629004), and all procedures were complied with the Laboratory Animal Requirements of Environment and Housing Facilities. BALB/c mice (famale, 6–8 weeks) were randomly assorted into two different treatment series included treated with NIR light series (Control (+), C/H Gel (+), BOS@C/H Gel (+), Fe‐BOS@C/H Gel (+)) and treated without NIR light series (Control (−), C/H Gel (−), BOS@C/H Gel (−), Fe‐BOS@C/H Gel (−)). To construct a mouse model with an MRSA biofilm‐infected wound, the mice were first anesthetized with 5% chloral hydrate (1.5 mL·kg^−1^) in the abdominal cavity. Then, the upper back of the mice was shaved, sterilized, and a rounded skin wound (10 mm in diameter) was made. Subsequently, each wound was inoculated with 0.05 mL of MRSA suspension (1 × 10^8^ CFU·mL^−1^). For bacterial reproduction, the wounds were covered with a bare PU membrane (Tegaderm Film, 3 m, USA) for 3 days. Then, various groups of nanomaterials were injected into the infectious areas every other day. For better immobilization of nanoparticles, the wounds were covered with 3 m Liquid bandage (Cavilon No Sting Barrier Film, 3 m, USA). Treated with NIR light series were irradiated by NIR light at a power density of 0.4 W·cm^−2^ for 10 min. The temperatures of the back tissues were recorded during NIR light irradiation. To view the wound healing process, firstly, the mouse body weight was recorded on days 0, 2, 4, 6, and 8. Subsequently, the infected wounds were recorded the infected wounds were recorded and photographed on days 0, 2, 4, 6, and 8. Wound healing rates were calculated according to Equation (3):
where Wound area_0_ represents the initial wound area on day 0, and t represents the day, like 0, 2, 4, 6, and 8.
Histological
4.14
On the third and eighth day of the wound healing experiment, the skin tissue surrounding the wound site in each group was collected and fixed in 4% formaldehyde for 24 h. Then, they were embedded in paraffin and sectioned to 5 µm thick for Giemsa, H&E, Masson's staining, and immunofluorescence staining (Ki67 and COL‐I).
Statistical Analysis
4.15
The result was presented as mean ± standard deviation (SD) (n ≥ 3). The statistical analysis was performed with Origin 9.0 software through Student's t‐test and one‐way analysis of variance. The statistical significance was indicated by *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; and NS, not significant, respectively.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Supporting File: advs73614‐sup‐0001‐SuppMat.docx.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1T. A. Harris‐Tryon and E. A. Grice , “Microbiota and Maintenance of Skin Barrier Function,” Science 376 (2022): 940–945, 10.1126/science.abo 0693.35617415 · doi ↗ · pubmed ↗
- 2X. Hou , H. Wang , X. Yao , Q. Zhou , and X. Niu , “Pt‐Induced Sublattice Distortion Facilitates Enzyme Cascade Reactions for Eradicating Intracellularly Methicillin‐Resistant Staphylococcus aureus and Enhancing Diabetic Wound Healing,” ACS Nano 19 (2025): 17709–17727, 10.1021/acsnano.5c 01894.40307061 · doi ↗ · pubmed ↗
- 3Y. Li , Y. Dong , Z. Zhang , Z. T. Lin , C. Liang , and M. X. Wu , “Efficient Photolysis of Multidrug‐Resistant Polymicrobial Biofilms,” Advanced Science 12 (2025): 2407898, 10.1002/advs.202407898.39708333 PMC 11809414 · doi ↗ · pubmed ↗
- 4S. Zhang , W. He , J. Dong , Y. K. Chan , S. Lai , and Y. Deng , “Tailoring Versatile Nanoheterojunction‐Incorporated Hydrogel Dressing for Wound Bacterial Biofilm Infection Theranostics,” ACS Nano 19 (2025): 10922–10942, 10.1021/acsnano.4c 15743.40071724 · doi ↗ · pubmed ↗
- 5S. Lai , B. Cao , X. Ouyang , et al., “Fluorescent Microneedle‐Based Theranostic Patch for Naked‐Eye Monitoring and On‐Demand Photo‐Therapy of Bacterial Biofilm Infections,” Advanced Functional Materials 35 (2025): 2415559, 10.1002/adfm.202415559. · doi ↗
- 6H. Fan , Q. Sun , K. Dukenbayev , et al., “Carbon Nanoparticles Induce DNA Repair and PARP Inhibitor Resistance Associated With Nanozyme Activity in Cancer Cells,” Cancer Nanotechnology 13 (2022): 39, 10.1186/s 12645-022-00144-9. · doi ↗
- 7M. Chen , Y. Sun , B. Xu , et al., “Photo‐Responsive Nanozyme Disrupts Bacterial Electron Transport Chain for Enhanced Anti‐Biofilm Therapy,” Advanced Functional Materials 35 (2025): 2417354, 10.1002/adfm.202417354. · doi ↗
- 8M. Mirhaj , S. Labbaf , M. Tavakoli , and A. M. Seifalian , “Emerging Treatment Strategies in Wound Care,” International Wound Journal 19 (2022): 1934–1954, 10.1111/iwj.13786.35297170 PMC 9615294 · doi ↗ · pubmed ↗