Iron‐Hijacking Trojan Horse Nanoplatform Combats Implant‐Associated Biofilm Infections Through Immuno‐Fibrotic Remodeling

ABSTRACT

Implant‐associated biofilm infections persist due to a vicious cycle of biofilm resilience and fibrosis‐driven immune exclusion. To break this cycle, we engineered a photothermal nanoplatform (CHPB@H) consisting of curcumin‐loaded hollow Prussian blue (CHPB) nanoparticles embedded in an injectable thermosensitive chitosan hydrogel for near‐infrared (NIR)‐triggered local delivery. Upon NIR irradiation, the platform releases iron ions and curcumin that self‐assemble into Fe‐curcumin (Fe‐Cur) complexes. These complexes act as a molecular “Trojan horse” that hijacks bacterial iron‐acquisition pathways to induce ferroptosis. Concurrently, CHPB scavenges reactive oxygen species (ROS) via its intrinsic nanozyme activity and suppresses oxidative stress‐driven fibroblast activation, thereby dismantling the fibrotic barrier that excludes immune effector cells. This dual action eliminates biofilms and restores immune cell infiltration into the infection site. In preclinical animal models, the therapy clears established implant infections, remodels the immune‐fibrotic microenvironment, and promotes tissue regeneration. This work establishes a “clear‐and‐remodel” paradigm that integrates active targeted ferroptosis (via iron hijacking) with immuno‐fibrotic remodeling, offering a translatable strategy for complex implant‐associated biofilm infections.

Under NIR irradiation, CHPB@H releases iron ions and curcumin to form Fe–Cur complexes. These complexes act as a “Trojan horse” by hijacking the bacterial iron uptake system to induce ferroptosis, while the platform scavenges ROS and suppresses fibroblast activation, thereby dismantling the fibrotic barrier and restoring immune clearance. This dual action enables biofilm eradication and promotes tissue repair.

1. Introduction

Implant‐associated biofilm infections represent a serious clinical complication in orthopedic and reconstructive surgery, with high morbidity and recurrence rates [1, 2, 3]. Their refractoriness stems from classic bacterial defenses (biofilm formation and antibiotic resistance) and a dysregulated local microenvironment caused by sustained pathogen‐host interactions [4, 5, 6]. A key feature of this microenvironment is infection‐driven fibrosis exacerbated by reactive oxygen species (ROS) [7]. Persistent oxidative and inflammatory signaling drives aberrant fibroblast activation, forming dense fibrotic capsules around the infection site [8, 9]. These capsules create a physical barrier that restricts immune effector cell infiltration [10, 11, 12]. Consequently, even a robust systemic immune response is ineffective if effector cells cannot reach the pathogens. Yet current therapeutic strategies largely overlook this need to restore immune access [13, 14].

Nanozyme‐based antibacterial systems have advanced rapidly [15, 16]. Prussian blue platforms are widely studied for their photothermal and enzyme‐mimetic properties [17, 18], and ferroptosis‐based strategies have emerged as a promising bactericidal mechanism [19, 20, 21]. However, both face limitations in treating implant‐associated biofilm infections. Prussian blue systems mostly rely on photothermal or catalytic killing; the few that involve ferroptosis depend on passive iron release with limited active targeting [21]. Ferroptosis‐based studies, even those incorporating immune modulation, have rarely addressed the fibrotic immune‐exclusion barrier [22]. Thus, an ideal strategy should combine active targeted killing with fibrotic niche remodeling.

To address these challenges, we designed a multifunctional nanoplatform based on an “iron hijacking” concept. The design exploits polyphenolic–metal coordination chemistry [23, 24], enabling a “Trojan horse” strategy that leverages the bacterial iron uptake system for active targeting. Specifically, hollow Prussian blue (HPB) nanoparticles serve as an iron reservoir and a multifunctional nanozyme, releasing iron ions under near‐infrared (NIR) irradiation and scavenging ROS via their enzyme‐mimetic activities [25, 26]. Prussian blue is FDA‐approved for internal contamination with cesium or thallium, underscoring its clinical safety and translational potential [27]. Curcumin, acting as both the coordinating ligand and a direct anti‐fibrotic agent, was then incorporated to yield curcumin‐loaded HPB (CHPB) [28, 29]. Finally, we embedded CHPB nanoparticles in an injectable thermosensitive chitosan hydrogel (CHPB@H), enabling localized sustained release and on‐demand drug delivery [30, 31]. This platform operates through two synergistic mechanisms (Figure 1). Under the “Trojan horse” strategy, NIR triggers the release of iron ions and curcumin to form Fe–Cur complexes, which are actively internalized via the bacterial iron uptake system to induce ferroptosis [32]. Meanwhile, the platform scavenges excess ROS and suppresses fibroblast activation, thereby dismantling the fibrotic barrier and restoring immune cell infiltration.

FIGURE 1.

Dual mechanisms of the CHPB@H nanoplatform against biofilm infection. Under NIR irradiation, CHPB@H releases iron ions and curcumin to form Fe–Cur complexes. These complexes act as a “Trojan horse” by hijacking the bacterial iron uptake system to induce ferroptosis, while the platform scavenges ROS and suppresses fibroblast activation, thereby dismantling the fibrotic barrier and restoring immune clearance. This dual action enables biofilm eradication and promotes tissue repair. PB: Prussian blue; HPB: hollow Prussian blue; CHPB: curcumin‐loaded HPB.

Through systematic investigation, we demonstrate that this dual‐mode strategy disrupts the cycle of “oxidative stress–fibrosis–immune exclusion,” achieving effective biofilm clearance, immune infiltration, and bone regeneration in preclinical models. Relative to existing Prussian blue‐based and ferroptosis‐based systems, this work offers two innovations: active targeted delivery via the “iron hijacking” strategy (unlike passive iron release), and immune‐fibrotic remodeling that addresses the previously overlooked immune‐exclusion barrier. Thus, we establish an integrated paradigm that concurrently targets pathogens and the host‐derived barrier, offering a translatable therapeutic framework for complex implant‐associated infections.

2. Results and Discussion

2.1. Synthesis and Characterization of CHPB

HPB nanoparticles with a cubic morphology and mesoporous shell were prepared via a controlled hydrothermal etching method, exhibiting an average diameter of approximately 100 nm as observed by transmission electron microscopy (TEM) (Figure 2a,b) [33]. Curcumin was then loaded using an ethanol‐assisted adsorption process [34]. After loading, the cubic shape was well preserved, but the mesopores became less visible in TEM images, suggesting successful drug occupancy (Figure 2c). Quantification revealed a drug loading capacity of 8.0 ± 1.7% and an encapsulation efficiency of 69.5 ± 3.4%. X‐ray diffraction (XRD) patterns confirmed that the face‐centered cubic lattice of Prussian blue remained intact after drug loading, with no evidence of phase transition or amorphization (Figure 2d). Nitrogen physisorption measurements further substantiated pore filling: the BET surface area decreased from 224.76 m²/g (bare HPB) to 101.02 m²/g (CHPB), and the average pore diameter narrowed from 3.4 to 1.93 nm (Figure 2e and Figure S1). Energy‐dispersive X‐ray spectroscopy (EDS) elemental mapping revealed a homogeneous distribution of Fe, C, N, and O throughout the CHPB architecture (Figure 2f). X‐ray photoelectron spectroscopy (XPS) survey spectra confirmed the presence of Fe, C, N, and O (Figure S2). High‐resolution Fe 2p spectra showed that the characteristic Fe²⁺/Fe³⁺ redox states of the Prussian blue framework were preserved in CHPB (Figure 2g), and the C 1s region exhibited enhanced signals from oxygen‐bearing carbon species compared to bare HPB, directly evidencing curcumin incorporation (Figure 2h). Ultraviolet–visible (UV–vis) absorption spectroscopy provided complementary evidence: the CHPB spectrum displayed two distinct peaks at 426 nm (curcumin) and 720 nm (Prussian blue) (Figure 2i) [35, 36]. Dynamic light scattering (DLS) measurements gave a hydrodynamic diameter of 183 nm and a zeta potential of −28.3 mV (Figure 2j,k). Owing to the strongly negative surface charge, the nanoparticles exhibited excellent colloidal stability, with negligible changes in size or surface potential over 7 days in both phosphate‐buffered saline (PBS) and 10% serum‐containing medium (Figure S3) [37].

FIGURE 2.

Physicochemical properties of CHPB. (a) Schematic diagram of the CHPB synthesis process. (b) TEM image of HPB. Scale bar, 100 nm. (c) TEM image of CHPB. Scale bar, 100 nm. (d) XRD pattern of CHPB. (e) N₂ adsorption–desorption isotherm and pore size distribution of CHPB. (f) EDS elemental mapping of CHPB. Scale bar, 100 nm. (g) High‐resolution XPS Fe 2p spectra of HPB and CHPB. (h) High‐resolution XPS C 1s spectra of HPB and CHPB. (i) UV–vis spectrum of curcumin, HPB, and CHPB. (j) Hydrodynamic diameters of HPB and CHPB (n = 3). (k) Zeta potentials of HPB and CHPB (n = 3). (l) Kinetics of Fe release from CHPB with or without NIR irradiation. (m) UV–vis spectra of CHPB at different time points following NIR irradiation. (n) CAT‐mimicking, SOD‐mimicking, and POD‐mimicking activity of CHPB (n = 3). (o) XPS analysis of the antioxidant capability of CHPB. (p) Transformation between Fe²⁺ and Fe³⁺ states in CHPB by XPS. (q) Schematic illustration of the multi‐enzyme activity mechanism of CHPB. Data are presented as mean ± SD.

The release behavior of CHPB was evaluated at pH 5.4 in the presence of 10% serum. Without NIR irradiation, the release of both iron and curcumin was minimal, reaching only 3.2% and 13.1%, respectively, after 12 h. In contrast, exposure to NIR light (808 nm, 1.0 W/cm², 5 min) dramatically accelerated the release, yielding 13.2% iron release and 67.2% curcumin release at the same time point (Figure 2l; Figure S4). This NIR‐triggered release is attributed to the photothermal effect, which disrupts the nanoparticle structure. Time‐dependent UV–vis spectroscopy showed that upon NIR irradiation, the characteristic absorption peak of CHPB at 720 nm gradually diminished, indicating nanoparticle degradation (Figure 2m). Concurrently, a new absorption band emerged at 390 nm and increased in intensity over time; we ascribe this feature to a coordination complex formed between iron ions and curcumin (Fe–Cur complex) [38]. None of these spectral changes was observed in the absence of NIR irradiation.

2.2. Antioxidant and Photothermal Activity

Having validated the structural integrity and drug‐loading properties of CHPB, we next investigated its functional capabilities, specifically its potential to counteract oxidative stress and generate photothermal effects. Chronic infections are characterized by excessive production of ROS, which perpetuates tissue damage [39]. CHPB addresses this challenge through multiple enzyme‐mimetic activities (Figure 2n). It exhibits concentration‐dependent catalase (CAT)‐like activity, efficiently decomposing H₂O₂; superoxide dismutase (SOD)‐like activity, scavenging superoxide anions (·O₂ ⁻); and peroxidase (POD)‐like activity, enabling neutralization of various ROS. This multi‐enzymatic repertoire allows CHPB to intercept oxidative stress at multiple points within the ROS cascade [40]. The underlying mechanism relies on the redox‐active iron centers of the Prussian blue lattice [41]. High‐resolution XPS confirmed reversible Fe²⁺ and Fe³⁺ cycling (Figure 2o). Upon exposure to H₂O₂, hydroxyl radicals (·OH), and ·O₂ ⁻, a clear shift toward the Fe³⁺ state was observed, indicating effective electron transfer during catalytic reactions (Figure 2p). This redox versatility enables CHPB to act as an electron shuttle for sustained ROS scavenging (Figure 2q).

In addition to its antioxidant function, CHPB exhibits strong photothermal activity due to the intrinsic NIR absorption of Prussian blue [17, 42]. When a 50 µg/mL dispersion was irradiated with an 808 nm laser, the temperature reached 45.6°C within 5 min (Figure S5). Moreover, the heating efficiency remained stable over four consecutive on/off cycles (Figure S6), demonstrating excellent photostability for repeated therapeutic use.

2.3. In Vitro Antibacterial and Antibiofilm Efficacy

Leveraging the photothermal capacity of CHPB established above, we examined its antibacterial efficacy against Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli). Neither CHPB alone nor NIR irradiation alone reduced bacterial viability. When combined, however, CHPB + Laser decreased the colony‐forming units (CFU) of both strains by approximately 3 log₁₀ (Figure 3a–d). This effect was more pronounced than that of HPB + Laser, indicating that curcumin contributes synergistically to bacterial killing. TEM of treated bacteria revealed extensive ultrastructural damage: membrane disruption, cytoplasmic leakage, and formation of intracellular vacuoles (Figure 3e). Consistent with the morphological changes, we observed increased leakage of intracellular proteins (Figure S7) and β‐galactosidase (Figure S8), providing additional evidence of lost membrane integrity [43].

FIGURE 3.

In vitro antibacterial efficacy and mechanism of CHPB. (a, b) Representative photographs of bacterial colonies from S. aureus (a) and E. coli (b) after indicated treatments (serial dilutions 10⁻² to 10⁻⁵). (c,d) Quantitative analysis of viable bacteria (log₁₀ CFU) for S. aureus (c) and E. coli (d) (n = 5). (e) Representative TEM images of S. aureus and E. coli after Control or CHPB + Laser treatment. Scale bar, 500 nm. (f, g) 3D reconstructed confocal laser scanning microscope (CLSM) images of S. aureus (f) and E. coli (g) biofilms stained with SYTO 9 (live, green) and PI (dead, red). Scale bar, 50 µm. (h,i) Anti‐biofilm efficacy against S. aureus (h) and E. coli (i): total biomass (OD₅₉₅, crystal violet) and viable biofilm‐embedded bacteria (log₁₀ CFU) (n = 3). (j) Temperature changes and corresponding thermal images of bacterial suspensions subjected to different treatments (n = 3). (k) Siderophore secretion assay showing increased siderophore activity after CHPB + Laser treatments (n = 3). (l) Time‐lapse fluorescence microscopy images of S. aureus stained with an Fe²⁺‐specific probe after CHPB + Laser treatment. Scale bar, 10 µm. (m) Time‐course quantification of intracellular Fe²⁺ and Fe³⁺ concentrations in S. aureus following CHPB + Laser treatment (n = 3). (n) Flow cytometry analysis of intracellular ROS levels in S. aureus using the DCFH‐DA probe. (o) CLSM images of S. aureus stained with the lipid peroxidation probe C11‐BODIPY^581/591. Green fluorescence indicates lipid peroxidation. Scale bar, 10 µm. (p) Activities of respiratory chain complexes I–IV and ATP levels in S. aureus after indicated treatments. (q) Schematic illustration summarizing the multimodal antibacterial mechanism of the CHPB “Trojan Horse” strategy. Data are presented as mean ± SD.

We next asked whether CHPB + Laser could also dismantle established biofilms, a critical challenge in chronic infections [44, 45]. Live/dead staining provided direct visual evidence: untreated biofilms consisted almost entirely of live cells (green), whereas those exposed to CHPB + Laser showed pervasive red fluorescence, indicating widespread cell death (Figure 3f,g). Quantitative analysis confirmed these visual findings. The treatment reduced total biofilm biomass by 88.6% (S. aureus) and 89.8% (E. coli), and lowered the viable bacterial counts within the biofilms by 2.3 and 2.7 log₁₀ CFU, respectively (Figure 3h,i). Taken together, CHPB + Laser possesses a dual ability: it disrupts the biofilm matrix while eradicating the embedded bacteria.

To elucidate the mechanism behind biofilm disruption, we analyzed the extracellular matrix (ECM)—a protective scaffold of DNA, proteins, and polysaccharides [46]. Agarose gel electrophoresis showed that CHPB + Laser reduced the integrity of high‐molecular‐weight DNA by 96% (Figure S9), and SDS‐PAGE revealed an 85% reduction in total protein signal extracted from the biofilm matrix (Figure S10). Collectively, these results demonstrate that CHPB + Laser effectively degrades both the nucleic acid and protein components of the ECM, thereby facilitating penetration into the biofilm and enabling complete bacterial eradication [47].

2.4. Mechanism of Bacterial Ferroptosis Induction

The antibacterial action of CHPB + Laser proceeds through a cascade of events initiated by mild photothermal heating. Irradiation raised the temperature of the bacterial suspension to 47.9°C (Figure 3j). This thermal effect alone did not kill bacteria but was indispensable for efficacy: omitting the laser reduced antibacterial activity by approximately 80% (Figure S11). Heating is thought to facilitate the release of iron and curcumin from CHPB, accelerate Fenton reaction kinetics, and increase membrane permeability [48, 49, 50]. In response to the released iron, bacteria mounted an iron‐starvation response, evidenced by increased siderophore secretion (Figure 3k) and upregulation of iron‐transport genes (Figure S12). These observations led to the hypothesis that iron and curcumin form Fe–Cur complexes that enter bacteria via the siderophore‐mediated uptake pathway. To test this hypothesis, we used two inhibitors that block iron acquisition through distinct mechanisms: sodium orthovanadate (Na₃VO₄, an ABC transporter ATPase inhibitor) and gallium nitrate (Ga(NO₃)₃, an Fe³⁺ mimic that competes for siderophore binding) [5, 51]. At sub‐inhibitory concentrations, both inhibitors reduced intracellular iron and curcumin levels, and consequently attenuated bacterial killing, with CFU decreasing by approximately 2.2 and 1.8 log₁₀, respectively (Figure S13). Notably, the reductions in iron and curcumin content differed by less than 15%, confirming that Fe–Cur is internalized as an intact complex through the bacterial iron uptake system.

Once inside the cytoplasm, the Fe–Cur complex underwent a transformation. Fluorescent probes revealed substantial accumulation of Fe²⁺ in bacteria after CHPB + Laser treatment (Figure 3l). Time‐course analysis showed that over 60 min, both Fe³⁺ and Fe²⁺ levels increased, but Fe²⁺ rose much more steeply than Fe³⁺ (Figure 3m), indicating that within the reducing environment of the bacterial cytoplasm, a portion of Fe³⁺ is reduced to Fe²⁺. The resulting excess Fe²⁺ then drives Fenton reactions, generating large quantities of ROS (Figure 3n; Figure S14). The ensuing oxidative stress triggered lipid peroxidation—a hallmark of ferroptosis (Figure 3o). Two classical markers confirmed this: malondialdehyde (MDA) levels increased, and the GSH/GSSG ratio collapsed; both changes were largely reversed by the iron chelator deferoxamine (DFO) (Figure S15) [19, 52]. To determine whether the killing was specifically ferroptotic or merely a consequence of general oxidative damage, we compared two inhibitors. Fer‐1, a specific ferroptosis blocker, strongly protected bacteria, raising CFU by approximately 1.53 log₁₀. In contrast, the broad‐spectrum ROS scavenger N‐acetylcysteine (NAC) provided only weak and non‐significant protection (CFU increase of ∼0.47 log₁₀) (Figure S16). Thus, CHPB + Laser kills bacteria primarily through iron‐dependent lipid peroxidation (ferroptosis), rather than non‐specific oxidative stress.

Ferroptosis was not the only consequence. The treatment also severely impaired bacterial energy metabolism: it inhibited respiratory chain complexes I–IV, leading to a 79.6% depletion of intracellular ATP (Figure 3p) [53]. Moreover, it suppressed quorum‐sensing (QS) gene expression, reduced biofilm‐forming capacity (Figure S17), and decreased secretion of key virulence factors (Figure S18). When DFO was added, respiratory chain activity and ATP levels largely recovered, further underscoring the central role of iron‐mediated damage. Collectively, these findings establish a proof‐of‐concept that CHPB + Laser can deliberately trigger bacterial ferroptosis by hijacking the pathogen's own iron‐uptake machinery (Figure 3q). Unlike conventional therapies that rely on indiscriminate damage, this “Trojan horse” strategy achieves potent, iron‐dependent lipid peroxidation.

2.5. Cytoprotective and Anti‐Fibrotic Properties

Having established the potent antibacterial activity of CHPB, we next evaluated its biocompatibility and its ability to counteract oxidative stress and fibrosis. Before functional testing, we confirmed that CHPB is safe for cellular use. At concentrations up to 100 µg/mL, cell viability remained > 95% regardless of NIR irradiation (Figure 4a). Hemolysis was below 5% across all tested doses, and erythrocytes preserved their normal biconcave shape (Figure 4b,c).

FIGURE 4.

In vitro biocompatibility, antioxidant, and anti‐fibrotic properties of CHPB. (a) Viability of fibroblasts after 24 h incubation with increasing concentrations of CHPB, with or without NIR irradiation (n = 3). (b) Hemolysis rate of mouse erythrocytes after 2 h exposure to CHPB (n = 3); representative images are shown above. (c) Representative bright‐field images of erythrocyte morphology after treatment with water (positive control), saline (negative control), or CHPB. Scale bar, 20 µm. (d) Representative fluorescence images of intracellular ROS in fibroblasts stained with DCFH‐DA. Scale bar, 100 µm. (e) Activities of the antioxidant enzymes Gpx1, CAT, and SOD2 in fibroblasts under indicated treatments (n = 5). (f) Western blot analysis of Nrf2 (cytosolic and nuclear fractions) and HO‐1 expression; β‐actin and Lamin B1 served as loading controls. (g) Densitometric quantification of protein levels (n = 3). (h) Immunofluorescence images of α‐SMA (green) in fibroblasts; nuclei were stained with DAPI (blue). Scale bar, 100 µm. (i) Relative mRNA expression levels of fibrosis‐related genes (n = 5).(j) Western blot analysis of TGF‐β/Smad pathway. (k) Densitometric quantification of protein levels (n = 3). (l) Schematic illustration summarizing the proposed antioxidative and anti‐fibrotic mechanisms of CHPB. Data are presented as mean ± SD.

We next assessed whether CHPB could protect cells from oxidative damage. Exposure to H₂O₂ strongly elevated intracellular ROS, while CHPB not only suppressed this increase (Figure 4d) but also rescued the activity of three endogenous antioxidant enzymes that H₂O₂ had silenced. Glutathione peroxidase 1 (Gpx1), CAT, and SOD2 were elevated by 5.6‐, 6.9‐, and 6.4‐fold, respectively (Figure 4e). Mitochondrial protection was confirmed by JC‐1 staining, which showed preserved membrane potential (ΔΨm) (Figure S19) [54]. In addition, scratch assays revealed enhanced fibroblast migration, further supporting the cytoprotective effect of CHPB (Figure S20). Mechanistically, CHPB activated the Nrf2/HO‐1 axis, a master antioxidant pathway [55]. It drove nuclear translocation of nuclear factor erythroid 2‐related factor 2 (Nrf2) and boosted its target heme oxygenase‐1 (HO‐1) by 2.4‐fold (Figure 4f,g).

Because persistent oxidative stress drives fibrosis, we next evaluated whether CHPB could counteract pro‐fibrotic signals. H₂O₂ strongly induced the expression of myofibroblast marker α‐smooth muscle actin (α‐SMA) and fibrosis‐related genes (ACTA2, COL1A1, COL3A1, FN1); CHPB abolished these effects in a dose‐dependent manner (Figure 4h,i). Focusing on the central TGF‐β/Smad pathway [56, 57], we found that H₂O₂ elevated the p‐Smad2/Smad2 and p‐Smad3/Smad3 ratios, as well as collagen I (COL‐I) expression. CHPB brought all three parameters down to levels comparable to those achieved with a TGF‐β inhibitor. Critically, siRNA‐mediated knockdown of Nrf2 largely eliminated CHPB's protection, proving that the anti‐fibrotic action depends on Nrf2 (Figure 4j,k).

In summary, CHPB scavenges ROS and activates Nrf2, which in turn inhibits the TGF‐β/Smad signaling axis, reduces COL‐I expression, and attenuates fibroblast activation. For implant‐related infections, this combined activity is expected to dismantle the dense fibrotic capsule that would otherwise block immune cell infiltration, thereby promoting bacterial eradication (Figure 4l).

2.6. In Vivo Anti‐Biofilm Efficacy of CHPB@H

Building on the robust in vitro performance of CHPB, we engineered a localized delivery system to provide sustained therapeutic action at the infection site. CHPB was incorporated into a thermosensitive chitosan‐based hydrogel, forming the composite CHPB@H. This system demonstrated excellent injectability at room temperature (Figure S21) and underwent a rapid sol‐to‐gel at physiological temperature (37°C), enabling in situ depot formation (Figure 5a). Rheological characterization determined the gelation temperature to be 37.5°C (Figure S22). Morphological analysis via scanning electron microscopy (SEM) revealed a highly porous, interconnected three‐dimensional network (Figure 5b). The release kinetics of CHPB@H indicated prolonged drug release under simulated infection conditions (Figure 5c). Weight loss analysis demonstrated that the hydrogel underwent slow degradation in 10% serum, confirming its stability (Figure S23). In vivo, fluorescently labeled CHPB@H injected subcutaneously into the dorsal region of mice remained detectable for 14 days with gradual degradation, indicating retention at the implant site (Figure 5d; Figure S24). Thus, CHPB@H serves as a versatile local platform for sustained combination therapy.

FIGURE 5.

Immuno‐fibrotic remodeling therapy eradicates biofilm infection in a murine subcutaneous implant model. (a) Sol–gel transition of CHPB@H upon heating to physiological temperature (37°C). (b) SEM image of CHPB@H. Scale bar, 1 µm. (c) Release profiles of iron and curcumin from CHPB@H over 14 days. (d) Representative fluorescence images of the injection site at different time points after subcutaneous injection of fluorescently labeled CHPB@H. (e) Schematic illustration of the subcutaneous implant infection model and treatment regimen. (f) Representative in vivo thermal images of the dorsal region. (g) Quantitative analysis of the maximum local temperature at the implant site (n = 3). (h) Representative photographs taken on day 14 post‐infection: gross wound appearance (left), explanted titanium disc with surrounding tissue (middle), and disc surface (right). Red, green, and blue arrows indicate skin ulceration, fibrous capsules, and surface biofilms, respectively. (i) Quantitative bacterial burden on implant surfaces on day 14 (n = 5). (j) Representative Masson's trichrome staining of peri‐implant tissue sections on day 14. Scale bar, 500 µm (left) and 50 µm (right). (k) Thickness of the fibrous capsule surrounding the titanium implant (n = 3). (l) Multiplex immunofluorescence staining of peri‐implant sections for α‐SMA (myofibroblasts), CD3 (T cells), F4/80 (macrophages), and Ly6G (neutrophils), with quantitative analysis of fluorescence intensity. Scale bar, 50 µm (left) and 20 µm (right). (m) Schematic diagram summarizing the proposed in vivo mechanism of immune‐fibrotic remodeling therapy. (n) Flow cytometric analysis of peri‐implant cellular composition during the acute infection phase (day 3) (n = 3). (o) Heatmap of inflammatory cytokine levels measured in peri‐implant tissues on day 3. Data are presented as mean ± SD.

We next evaluated the therapeutic efficacy of CHPB@H in a clinically relevant murine model of subcutaneous implant infection using S. aureus‐coated titanium discs (Figure 5e). Upon NIR irradiation, in vivo thermal imaging confirmed localized photothermal heating in the CHPB@H + Laser group, with the implant site temperature reaching ∼44.9°C (Figure 5f,g). Macroscopic assessment on day 14 revealed that CHPB@H + Laser treatment mitigated infection‐induced tissue damage, exhibiting nearly intact skin, clean implant surfaces, and a 94.5% reduction in ulceration area (Figure 5h; Figure S25). SEM of explanted implants showed minimal adherent bacteria (Figure S26), and CFU enumeration confirmed an ∼1.8 log₁₀ reduction in viable bacteria recovered from implant surfaces (Figure 5i). Consistently, pro‐inflammatory cytokines in the peri‐implant tissues of the CHPB@H + Laser group were downregulated, confirming effective resolution of local inflammation (Figure S27).

At the histological level, Masson's trichrome staining showed an 84.5% reduction in fibrotic capsule thickness in the CHPB@H + Laser group (Figure 5j,k). Meanwhile, CHPB@H + Laser treatment reduced local ROS levels (Figure S28), enhanced antioxidant enzyme activities (Figure S29), and upregulated Nrf2 expression (Figure S30), confirming activation of the endogenous antioxidant pathway. Immunohistochemistry also revealed decreased COL‐I deposition (Figure S31), consistent with reduced fibrosis. To investigate the relationship between the fibrotic barrier and immune infiltration, we performed multiplex immunofluorescence staining. The results showed that CHPB@H + Laser reduced α‐SMA⁺ myofibroblasts near the implant surface and increased the infiltration of CD3⁺ T cells, F4/80⁺ macrophages, and Ly6G⁺ neutrophils (Figure 5l). We propose that relieving fibrosis enhances local immune infiltration and promotes infection clearance (Figure 5m).

We further performed flow cytometric analysis on peri‐implant tissues harvested on day 3 post‐treatment. The results showed a significant decrease in α‐SMA⁺ fibroblasts accompanied by a marked increase in CD45⁺ leukocyte infiltration. These immune cells were enriched in CD86⁺ M1‐like macrophages and CD8⁺ T cells (Figure 5n), indicating activation of both innate and adaptive immunity at the early stage. Macrophage polarization was further confirmed by immunofluorescence staining: CHPB@H + Laser significantly increased the expression of the M1 marker CD86, while the M2 marker CD206 slightly decreased (Figure S32). Given the role of M1 macrophages in antibacterial host defense, this polarization shift is expected to contribute to enhanced bacterial clearance [58]. In addition, cytokine profiling revealed broad elevation of pro‐inflammatory cytokines, reflecting effective activation of local innate immunity (Figure 5o) [59].

2.7. Transcriptomic Evidence of Host Microenvironment Reprogramming

To delineate the molecular mechanisms, we performed genome‐wide transcriptomic analysis on peri‐implant tissues harvested at day 3 post‐treatment (Figure 6a). RNA‐sequencing data revealed high reproducibility among biological replicates (Figure S33a). Comparative analysis between the CHPB@H + Laser group and the infected control identified extensive transcriptional reprogramming, with 596 genes significantly upregulated and 529 genes downregulated (Figure 6b; Figure S33b).

FIGURE 6.

Molecular mechanisms underlying host microenvironment reprogramming. (a) Schematic workflow for transcriptomic and molecular analysis of peri‐implant tissues harvested on day 3 post‐treatment. (b) Volcano plot of DEGs in peri‐implant tissues from the CHPB@H + Laser group compared to the Control. (c, d) GO Biological Process (c) and KEGG pathway (d) enrichment analyses of upregulated DEGs. (e–g) GSEA plots showing the normalized enrichment scores (NES) for the downregulated Wnt signaling pathway (e), and the upregulated Chemokine signaling (f) and Phagosome pathways (g). (h, i) qPCR validation of the relative mRNA expression levels of key Wnt pathway genes (h) and chemokine genes (i) in peri‐implant tissues (n = 5). (j) Schematic diagram of the in vitro Transwell co‐culture assay used to assess the link between fibrosis inhibition and immune cell migration. (k) Representative images of crystal violet‐stained migrated macrophages under the indicated conditions. (l) Quantitative analysis of macrophage migration efficiency (n = 3). Data are presented as mean ± SD.

Gene Ontology (GO) enrichment analysis revealed strong association with innate and adaptive immune activation (Figure 6c; Figure S33c–e). Biological processes such as “immune response”, “defense response to Gram‐positive bacterium”, “antigen processing and presentation”, and “inflammatory response” were significantly enriched. Kyoto Encyclopedia of Genes and Genomes (KEGG) further highlighted upregulation of critical immune‐related pathways, including “antigen processing and presentation”, “Th1/Th2/Th17 cell differentiation”, and “natural killer cell‐mediated cytotoxicity” (Figure 6d). These signatures indicate that, in addition to directly killing bacteria, CHPB@H + Laser therapy activates the host immune system.

In parallel, transcriptomic profiling demonstrated suppression of pro‐fibrotic signaling. Gene Set Enrichment Analysis (GSEA) showed significant downregulation of the Wnt/β‐catenin pathway (Figure 6e), a central regulator of fibroblast activation and ECM deposition [60]. Conversely, pathways essential for immune cell trafficking and function, such as “chemokine signaling” and “phagosome formation”, were upregulated (Figure 6f,g; Figure S34). This dual transcriptional profile was validated by qPCR, which confirmed suppression of key Wnt target genes (AXIN2, SP5, CCND1, C‐MYC) and the induction of chemokines involved in leukocyte recruitment (CXCL8, CXCL10, CCL2, CCL21) in treated tissues (Figure 6h,i) [61]. Together, these data provide a molecular rationale for the attenuation of fibrosis and enhanced immune infiltration observed in vivo.

To functionally link reduced fibrosis with improved immune access, we established an in vitro fibroblast‐macrophage co‐culture model simulating a fibrotic barrier (Figure 6j). Macrophage migration was severely impaired (by 72.3%) when co‐cultured with H₂O₂‐activated fibroblasts, mimicking the restrictive microenvironment of infection‐driven fibrosis. CHPB@H treatment reversed this impairment, restoring macrophage migratory capacity by 2.8‐fold (Figure 6k,l). These functional data demonstrate that CHPB@H‐mediated attenuation of fibroblast activation removes a physical barrier to immune cell movement, thereby enabling effective immune surveillance and pathogen clearance.

In summary, these transcriptomic and functional analyses reveal that CHPB@H + Laser therapy orchestrates a reprogramming of the infected host microenvironment, simultaneously enhancing immune‐activating pathways and suppressing pro‐fibrotic signaling. This dual‐pronged reshaping of the tissue niche dismantles the physical and molecular barriers that perpetuate infection, providing a mechanistic foundation for the synergy between pathogen clearance and host‐directed tissue repair.

2.8. Healing of Infected Bone Defects

To assess the translational potential of CHPB@H in regenerating infected bony tissue, we established a rat femoral bone defect model with infection (Figure 7a). Local NIR irradiation of CHPB@H induced rapid photothermal heating to 44.8°C (Figure 7b,c). Body weight and knee joint circumference were monitored over 8 weeks to assess systemic and local infection. All treatment groups showed progressive weight gain, with the CHPB@H + Laser group showing the largest increase (Figure S35). Knee joint swelling was observed in all groups during the first two weeks. However, the CHPB@H + Laser group consistently displayed the mildest swelling, which gradually resolved from week 4 and returned to baseline by week 8. In contrast, swelling persisted in the infected control group (Figure S36). These observations indicate that CHPB@H + Laser effectively controls local infection and improves overall host condition.

FIGURE 7.

Therapeutic evaluation in a rat model of infected bone defect. (a) Schematic illustration and surgical images showing the establishment of the infected femoral defect model and the treatment protocol. (b) Representative in vivo thermal images of the knee region following different treatments. (c) Quantitative analysis of the maximum local temperature at the defect site. (d) Representative photographs of bacterial culture plates from homogenized femoral tissues at weeks 4 and 8 post‐surgery. (e) Quantitative analysis of bacterial burden (log₁₀ CFU per femur) at the indicated time points (n = 3). (f) Representative micro‐CT images (coronal, sagittal, and 3D‐reconstructed views) of femurs at weeks 4 and 8. (g–j) Quantitative micro‐CT morphometric analysis of bone regeneration: BV/TV (g), Tb.N (h), Tb.Th (i), and Tb.Sp (j) (n = 5). (k) Representative histological images of the defect area (H&E and Masson's trichrome staining) at weeks 4 and 8. Scale bars, 500 µm. (l) Serum biochemical analysis at 8 weeks post‐surgery. Data are presented as mean ± SD.

As a clinically relevant comparison, we compared CHPB@H + Laser with local vancomycin powder. At day 7 post‐treatment, CHPB@H + Laser achieved a greater reduction in bacterial load than vancomycin (Figure S37), indicating faster bactericidal kinetics. This advantage may be attributed to the photothermal‐enhanced “Trojan horse” mechanism, which overcomes the limitations of conventional antibiotics in biofilm‐infected and poorly vascularized bone defects [62]. At weeks 4 and 8, quantitative cultures of femoral specimens showed that CHPB@H + Laser reduced bacterial loads by approximately 1.0 log₁₀ and 1.2 log₁₀ CFU per femur, respectively, outperforming all other groups (Figure 7d,e). Giemsa staining further confirmed reduced bacterial colonization in the treated group (Figure S38).

Micro‐CT analysis revealed that CHPB@H + Laser promoted bone regeneration. By week 8, the defect site exhibited bridging callus formation and near‐complete closure, contrasting sharply with the limited repair in control groups (Figure 7f). Morphometric quantification confirmed superior osteogenic outcomes: bone volume fraction (BV/TV) increased 1.96‐fold over the infected control, trabecular number (Tb.N) and thickness (Tb.Th) increased 2.59‐fold and 1.63‐fold, respectively, and trabecular separation (Tb.Sp) decreased 1.81‐fold (Figure 7g–j). These parameters indicate not merely increased bone mass but a definitive shift toward a denser, more mechanically competent trabecular architecture [63].

Histological evaluation confirmed these findings (Figure 7k). At week 4, the CHPB@H + Laser group showed early resolution of inflammation and active filling of the defect with reparative tissue. By week 8, the defect was largely bridged by new bone. Masson's trichrome staining revealed minimal fibrotic tissue in the treatment group, in contrast to the dense fibrous scar in controls. COL‐I immunohistochemistry confirmed reduced collagen deposition, consistent with the Masson staining (Figure S39). This anti‐fibrotic effect aligns with the in vitro observations of Nrf2/TGF‐β/Smad pathway inhibition. Thus, in addition to directly killing bacteria, CHPB@H remodels the host fibrotic barrier, creating a favorable microenvironment for bone healing.

2.9. In Vivo Biosafety and Translational Considerations

We further evaluated the in vivo biosafety profile of CHPB@H. Subcutaneous injection of CHPB@H into mice caused no obvious local inflammation, necrosis, or fibrosis at the injection site after 14 days (Figure S40a). Complete blood count remained within normal ranges, confirming the absence of systemic toxicity (Figure S40b). To assess iron metabolism and clearance, we measured iron content in major organs (heart, liver, spleen, lung, and kidney) and local bone tissue following CHPB@H injection in the rat femoral defect model [64]. Iron was primarily distributed in local bone, followed by liver and kidney, peaked on day 1, and gradually decreased over time; by day 21, iron levels in all organs dropped to very low levels (Figure S41). Iron excretion occurred via both urine (peak on day 1) and feces (peak on day 4) (Figure S42), demonstrating effective clearance without long‐term accumulation. At week 8 post‐injection, serum biochemical parameters (liver, kidney, and cardiac markers) were within normal physiological ranges (Figure 7l). Histopathological examination of major organs revealed no tissue damage or inflammatory infiltration (Figure S43). Collectively, these data confirm the favorable safety profile and biodegradability of CHPB@H, laying a foundation for its translational application.

Although the above safety data support the translational potential of CHPB@H, we acknowledge that the limited penetration depth of NIR light into deep tissues remains a practical challenge for photothermal therapy. Nevertheless, several strategies could mitigate this issue. Surgical debridement, already a clinical standard for deep implant infections, can expose the implant site to allow direct NIR irradiation. Minimally invasive techniques, such as arthroscopy or percutaneous optical fiber insertion, can deliver NIR light to deep foci with minimal tissue damage [65]. Emerging technologies, such as injectable mechanoluminescent nano‐transducers that generate light in deep tissues under focused ultrasound, also offer future possibilities for combining with our platform [66]. These considerations suggest that CHPB@H still holds potential for treating deeper infections through appropriate clinical or technological adaptations.

3. Conclusion

This study presents CHPB@H, an integrated nanoplatform that overcomes implant‐associated infections by concurrently disrupting resilient biofilms and remodeling the immunosuppressive fibrotic niche. Through a “Trojan Horse” mechanism, CHPB@H hijacks bacterial iron‐acquisition to deliver Fe‐cur complexes, inducing ferroptosis, while its antioxidant and anti‐fibrotic functions dismantle the physical barrier that blocks immune cell access. Validated in both a subcutaneous implant model and an infected bone defect model, this dual‐mode strategy achieves potent biofilm clearance, immune restoration, and functional bone regeneration. Our work establishes a “clear‐and‐remodel” paradigm that directly targets the pathological microenvironment, offering a translatable strategy for treating complex fibrotic‐barricaded infections.

4. Experimental Section

4.1. Animal Ethics and Husbandry

All experimental protocols rigorously complied with the NIH Guide for Laboratory Animal Care and Use alongside ARRIVE guidelines. The study protocol was approved by the Animal Ethics Committee of Zhengzhou University (Approval Number: ZZU‐LAC20241227[15]). Mice (BALB/c, male, 6–8 weeks old) and rats (Sprague Dawley, male, 250–300 g) were housed in a specific pathogen‐free (SPF) environment with free access to food and water.

4.2. Statistical Analysis

All experiments were independently repeated at least three times. Data were presented as mean ± standard deviation (SD). Statistical analysis was performed using GraphPad Prism 9.0 software. Comparisons between two groups were performed using an unpaired, two‐tailed Student's t‐test. Comparisons among multiple groups were performed using one‐way or two‐way analysis of variance (ANOVA), followed by Tukey's post hoc test. A p‐value < 0.05 was considered statistically significant (^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: adma73711‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding authors upon reasonable request.