Neutrophil membrane-encapsulated nanosonosensitizer with ultrasound-reinforced ferroptosis in Pseudomonas aeruginosa pneumonia

Abstract

Pneumonia caused by Pseudomonas aeruginosa (P. aeruginosa) infection remains a formidable clinical challenge due to persistent biofilm formation and intrinsic antibiotic resistance, exacerbated by bacterial iron homeostasis that stabilizes biofilm architecture and neutralizes oxidative stress. Herein, we present Fe/TNT@NM, a biomimetic nanosonosensitizer activated by ultrasound (US) to dismantle biofilms through dual extracellular-intracellular mechanisms. The nanosonosensitizer features an iron-doped titanate nanotube (Fe/TNT) core encapsulated within a neutrophil membrane (NM). Under US irradiation, Fe/TNT@NM generates sonodynamic reactive oxygen species (ROS) extracellularly and enhances Fe³⁺ release. These ions catalyze the Fenton reaction extracellularly to amplify chemodynamic effects and disrupt intracellular iron homeostasis, triggering bacterial ferroptosis. The NM coating enables immune evasion and biofilm-targeted delivery. This ultrasound-reinforced ferroptosis strategy synchronizes extracellular ROS storms with intracellular iron dyshomeostasis, achieving dual-action biofilm dismantling and eradication of drug-resistant P. aeruginosa. In a murine pneumonia model, Fe/TNT@NM suppresses biofilms and mitigates pulmonary injury. By converging biomimetic targeting, sonodynamic-chemodynamic cascades, and ultrasound-augmented ferroptosis, this nanosonosensitizer presents a paradigm-shifting approach to combat refractory biofilm infections and antibiotic resistance.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-03676-5.

Introduction

Bacterial pneumonia remains a formidable global health burden, exhibiting substantial morbidity and mortality rates worldwide [1]– [2]. Among causative pathogens, P. aeruginosa emerges as a leading etiological agent of nosocomial pneumonia, often progressing to severe complications such as pulmonary abscesses and necrotizing pneumonia, which critically threaten patient survival [3]. Current clinical management predominantly relies on antibiotic regimens targeting P. aeruginosa [4]– [5]. However, the pathogen’s remarkable capacity for biofilm formation significantly enhances bacterial persistence and accelerates antimicrobial resistance evolution, particularly fostering multidrug-resistant strains [6–9]. This adaptive mechanism complicates therapeutic efficacy, posing unprecedented challenges to modern antimicrobial stewardship [8, 10]. Consequently, the development of innovative therapeutic strategies capable of disrupting P. aeruginosa biofilms and eradicating resistant strains represents an urgent unmet need in combating this recalcitrant infection.

The pathogenic microenvironment of P. aeruginosa infections presents four critical barriers to effective treatment: hypoxia, acidic pH, H₂O₂ accumulation, and iron metabolic homeostasis, which collectively impair therapeutic outcomes in pneumonia [11]– [12]. To address these challenges, engineered nanomaterials capable of responding to endogenous/exogenous stimuli and converting intrinsic substrates into bactericidal agents represent a revolutionary strategy against biofilm-associated antibiotic-resistant infections. This approach enables multimodal therapeutic convergence through precise spatiotemporal control, particularly via external triggers such as microwaves, optical radiation, thermal energy, and US [13–16]. Among these, US demonstrates exceptional promise in biofilm eradication due to its non-invasive nature, deep tissue penetration, and precise spatiotemporal resolution [17]. Sonodynamic therapy (SDT), leveraging sonosensitizers like metal-doped titanium dioxide and metal-organic frameworks, achieves microbial clearance through US-activated ROS generation from water/oxygen molecules [18–20]. Titanate nanotubes (TNTs) further enhance SDT efficacy via their high surface area, hollow architecture, and efficient electron transport properties [21].

P. aeruginosa employs an intricately tuned iron-homeostasis network to sustain intracellular redox balance and effectively neutralize exogenous ROS, thereby establishing a fundamental basis for resistance [22–24]. While iron is indispensable for bacterial energy metabolism, respiration, and DNA synthesis [25–27], its excessive accumulation paradoxically triggers the Fenton reaction, inducing ferroptosis via lipid peroxidation cascade [28]– [29]. This dual role of iron has been exploited through iron-based nanomaterials (e.g., FeSO4, Fe3O4, FeCl3), which amplify labile iron pool (LIP) and ROS accumulation to induce ferroptosis in resistant pathogens [30–36]. These insights motivate the rational design of multifunctional nanocomposites that simultaneously target biofilm architecture and hijack bacterial iron metabolism, offering a synergistic strategy to combat P. aeruginosa pneumonia.

In this study, we engineered Fe/TNT@NM, a biomimetic nanotherapeutic platform that synergizes extrinsic (ultrasound) and intrinsic (microenvironmental) stimuli for targeted treatment of P. aeruginosa pneumonia (Scheme 1A). The nanocomposite comprises an iron-doped titanate nanotube (Fe/TNT) core enveloped within a neutrophil membrane (NM). Upon exposure to high-penetration US irradiation, Fe/TNT@NM initiates rapid extracellular ROS generation via sonodynamic activation, achieving a robust antibacterial response. Concurrently, US stimulation accelerates Fe ion release from the nanotube core. These liberated Fe ions exhibit dual functionality: (1) extracellularly catalyzing the Fenton reaction to amplify ROS-mediated chemodynamic effects, and (2) intracellularly disrupting bacterial iron homeostasis. Excessive intracellular Fe accumulation induces lipid peroxidation cascades, ultimately triggering ferroptosis. The NM coating confers dual advantages: evading immune clearance through biomimetic camouflage and promoting site-specific accumulation at infected loci via inflammation-targeting receptors. Precision-controlled US irradiation orchestrates a coordinated ROS surge across extracellular and intracellular compartments, synergistically disrupting P. aeruginosa biofilms and eliminating drug-resistant pathogens. Validated in a murine pneumonia model, Fe/TNT@NM demonstrated potent biofilm eradication and significant mitigation of pulmonary inflammation (Scheme 1B). This multimodal therapeutic paradigm—integrating sonodynamic-chemodynamic cascades with ferroptosis induction—establishes a durable antibacterial strategy against recalcitrant P. aeruginosa infections.

Scheme 1.

Construction and therapeutic mechanism of Fe/TNT@NM for combating Pseudomonas aeruginosa pneumonia. (A) Schematic illustration of Fe/TNT@NM nanostructure fabrication. (B) Dual-mode antibacterial mechanism of Fe/TNT@NM: combining an extracellular antibacterial strategy based on synergistic sonodynamic and chemodynamic effects with an intracellular antibacterial strategy mediated by ferroptosis, this system enables comprehensive biofilm eradication and mitigation of pulmonary inflammation

Results and discussion

Preparation and characterization of Fe/TNT@NM

To achieve precise spatiotemporal control over P. aeruginosa biofilm-associated pneumonia, we developed Fe/TNT@NM, a biomimetic nanocomposite that synergistically responds to endogenous (microenvironmental) and exogenous (ultrasound) triggers. The synthesis commenced with the preparation of hollow iron-doped titanate nanotubes (Fe/TNT) via a one-step hydrothermal method [37]. Transmission electron microscopy (TEM) revealed Fe/TNT’s uniform tubular morphology, with an average diameter of 20 nm and length of 220 nm. High-resolution TEM (HR-TEM) confirmed its crystalline structure, displaying lattice fringes with a spacing of 0.354 nm corresponding to the anatase TiO₂ (101) phase. Concurrently, neutrophil membranes (NM) were isolated from lipopolysaccharide (LPS)-activated murine peripheral blood [38] and purified (Scheme 1A). The Fe/TNT@NM nanocomposite was subsequently fabricated by encapsulating Fe/TNT with NM, as evidenced by TEM imaging showing intact membrane coating (Fig. 1A).

Fig. 1.

Structural and functional characterization of Fe/TNT@NM. (A) TEM and HR-TEM images of Fe/TNT nanotubes (left), purified NM (middle), and Fe/TNT@NM (right). (B) XRD patterns of Fe/TNT@NM and TNT@NM. (C) XPS spectra with Fe 2p of Fe/TNT@NM. (D) Element mapping of Ti, Fe, C, N, and O in Fe/TNT@NM. (E) UV-vis spectra of Fe/TNT@NM and TNT@NM. (F) The hydraulic diameter and zeta potential of Fe/TNT@NM. Flow cytometry analysis of LFA-1 expression in Fe/TNT, neutrophils, and Fe/TNT@NM (G), and quantitative analysis (H). (I) Expression of key receptor proteins in neutrophils, NM, and Fe/TNT@NM. (J) Total protein composition assessed by Coomassie Brilliant Blue staining of Fe/TNT, Fe/TNT@NM with different doping ratios, and purified NM. (K) Fluorescence images of mouse alveolar macrophages after incubation with Fe/TNT@NM. Red: Fe/TNT@NM (pre-labeled with Rhodamine B); Blue: cell nuclei; Green: cytoplasm. Data are represented as mean ± SD (n = 3). *** P < 0.001, n.s P > 0.05, n.s, no significance

X-ray diffraction (XRD) analysis validated the crystalline framework of Fe/TNT@NM, with characteristic peaks aligning with TiO₂ (JCPDS 21-1272) and H₂Ti₂O₅·H₂O (JCPDS 47-1024). A downward shift in the diffraction peak at 2θ = 25.3° confirmed successful iron doping on the titanate nanotube surface (Fig. 1B). Fourier-transform infrared (FTIR) spectroscopy further corroborated Fe incorporation, showing a characteristic peak shift toward higher wavenumbers (Figure S1A, Supporting Information). X-ray photoelectron spectroscopy (XPS) detected Fe³⁺, Ti⁴⁺, O, and C in Fe/TNT@NM, confirming elemental composition and valence states (Fig. 1C; Figure S1B-D, Supporting Information). Energy-dispersive X-ray spectroscopy (EDS) mapping verified homogeneous distribution of Fe, Ti, C, and O across the nanotube surface (Fig. 1D). UV-vis spectroscopy revealed a reduced bandgap (2.98 → 2.23 eV) due to Fe doping, which facilitates electron-hole separation and enhances sonodynamic activity under ultrasound irradiation (Fig. 1E; Figure S 1E, Supporting Information). Physicochemical characterization demonstrated that NM coating increased the hydrodynamic diameter of Fe/TNT@NM to 87 nm and reduced the zeta potential from − 13.3 mV (Fe/TNT) to −17.5 mV (Fig. 1F; Figure S2A, B, Supporting Information). Stability assays confirmed that Fe/TNT@NM maintained its structural integrity for 15 days in PBS and serum when the NM-to-Fe/TNT ratio was optimized to 2:1 (Figure S2C-E, Supporting Information).

To validate the successful encapsulation of NM onto the Fe/TNT core, we systematically analyzed the surface protein composition of Fe/TNT@NM. Energy-dispersive X-ray spectroscopy (EDS) elemental mapping confirmed the presence of C and N in Fe/TNT@NM, providing direct evidence of NM integration (Figure S3A, Supporting Information). Flow cytometry quantification revealed comparable lymphocyte function-associated antigen-1 (LFA-1) levels between Fe/TNT@NM and native neutrophils (Fig. 1G, H; Figure S3B-F, Supporting Information), confirming preservation of membrane protein integrity. Western blot analysis further demonstrated the enrichment of key inflammatory targeting receptors—TNF-α receptor (TNF-R), IL-1 receptor (IL-1R), and LFA-1—on the nanocomposite surface (Fig. 1I), which are critical for site-specific delivery to infected tissues [39, 40]. Coomassie Brilliant Blue staining revealed near-identical protein band patterns between Fe/TNT@NM and purified NM (Fig. 1J), verifying comprehensive retention of membrane-associated proteins during encapsulation. Crucially, NM coating reduced phagocytic clearance by mouse alveolar macrophages compared to uncoated Fe/TNT (Fig. 1K; Figure S3G, Supporting Information), demonstrating enhanced immune evasion capability.

Ultrasound enhanced sonodynamic activity and Fe³⁺ release of Fe/TNT@NM

As a TiO₂-based nanocomposite with iron dopants, Fe/TNT@NM leverages the inherent advantages of titanate nanotubes (TNTs), including high specific surface area and hollow architecture, which synergize with Fe-induced electronic modulation to amplify sonodynamic activity [41]. Specifically, Fe doping enhances electron-hole pair separation under ultrasound irradiation, thereby optimizing ROS generation efficiency [42]. To quantify ROS production, 1,3-diphenylisobenzofuran (DPBF)—a selective singlet oxygen (¹O₂) scavenger—was employed. UV-vis spectral analysis demonstrated a time-dependent decrease in DPBF absorption at 416 nm under US exposure (1.0 MHz, 1.5 W/cm²) (Fig. 2A), with Fe/TNT@NM exhibiting a 42.41% reduction in absorbance compared to water controls after 10 min of US (Fig. 2B; Figure S4A, Supporting Information). This decay correlated linearly with irradiation duration (0–10 min), confirming sustained ¹O₂ generation. Cyclic stability tests revealed robust performance over three consecutive US on-off cycles, with sonodynamic efficiency remaining above 85% of initial activity. However, prolonged US exposure (five cycles) induced structural degradation, reducing efficiency by 32.7% due to nanotube fragmentation (Fig. 2C). Electron spin resonance (ESR) spectroscopy further validated ¹O₂ production, displaying a characteristic 1:1:1 triplet signal intensity ratio under US activation (Fig. 2D). These findings collectively establish Fe/TNT@NM as a highly stable and efficient sonosensitizer, capable of generating cytotoxic ¹O₂ under spatiotemporal US control.

Fig. 2.

Sonodynamic activity and Fe^3⁺ release of Fe/TNT@NM. (A) UV-vis spectra of DPBF in Fe/TNT@NM under US (0–10 min). (B) UV absorption of DPBF in water and Fe/TNT@NM under US. (C) Sonodynamic efficiency of Fe/TNT@NM after 5 on-off US cycles. (D) ESR spectra of Fe/TNT@NM under US. (E) Fe^3⁺ release from Fe/TNT@NM under US in acidic pH. (F) Peroxidase activity of Fe/TNT@NM with various H₂O₂ concentrations. (G) Peroxidase activity of Fe/TNT@NM with H₂O₂ over time. (H) Steady-state kinetic of H₂O₂ catalysis by Fe/TNT@NM. Main plot: Michaelis-Menten plot; Inset: Lineweaver-Burk plot. (I) ESR spectrum of Fe/TNT@NM with H₂O₂. (J) Absorbance at 440 nm of Fe/TNT@NM, H₂O₂ and OPD under US. (K) Mechanisms of US-activated sonodynamic activity and Fe^3⁺ release of Fe/TNT@NM. Data are represented as mean ± SD (n = 3). ** P < 0.01, *** P < 0.001, n.s P > 0.05, n.s, no significance

Under the acidic microenvironment characteristic of P. aeruginosa infections, bacterial metabolic activity generates substantial hydrogen peroxide (H₂O₂), while host immune responses further amplify oxidative stress [43]– [44]. This acidic milieu, combined with external stimuli such as ultrasound, synergistically enhances the release of metal ions from the nanomaterials [45–47]. Kinetic analysis under simulated infection conditions revealed a time-dependent increase in Fe^3⁺ concentration, with US irradiation significantly accelerating Fe^3⁺ release compared to passive diffusion (Fig. 2E). The liberated Fe^3⁺ ions exploited the H₂O₂-rich microenvironment via pH-responsive Fenton reactions, catalytically converting H₂O₂ into cytotoxic hydroxyl radicals (·OH) [48]. UV-vis spectroscopy with a 3,3’,5,5’-tetramethylbenzidine (TMB) probe demonstrated concentration-dependent ·OH generation, evidenced by a linear increase in 652 nm absorption (Fig. 2F; Figure S4B, Supporting Information). Time-course experiments further validated sustained ·OH production, correlating with prolonged H₂O₂ exposure (Fig. 2G; Figure S4C, Supporting Information). Steady-state kinetic analysis revealed Michaelis-Menten saturation kinetics, with a maximum reaction velocity of 1.069 and a Michaelis constant of 10.85 mM (Fig. 2H). Lineweaver-Burk plots validated these parameters, highlighting Fe/TNT@NM’s catalytic efficiency. ESR spectroscopy corroborated ·OH generation, displaying a characteristic 1:2:2:1 intensity ratio under H₂O₂ incubation (Fig. 2I). Importantly, US irradiation synergistically enhanced the Fenton reaction, elevating ·OH levels by 2.3-fold compared to standalone chemodynamic therapy (Fig. 2J; Figure S4D, Supporting Information). Time-dependent US exposure further amplified ROS signals (Figure S4E, Supporting Information), demonstrating Fe/TNT@NM’s ability to harness both endogenous H₂O₂ and exogenous US energy for amplified oxidative stress. These findings demonstrate Fe/TNT@NM’s ability to integrate sonodynamic ROS generation, Fenton reaction-driven chemodynamic activity, and controlled Fe^3⁺ release into a unified therapeutic platform (Fig. 2K). The coordinated interplay of these mechanisms—extracellular oxidative stress induction and intracellular ferroptosis triggering—enables comprehensive biofilm eradication and antimicrobial efficacy against P. aeruginosa pneumonia.

Bactericidal efficacy of Fe/TNT@NM against P. aeruginosa

Given the dual sonodynamic-chemodynamic activities and Fe²⁺-release capability of Fe/TNT@NM, we systematically investigated the bactericidal efficacy of Fe/TNT@NM against P. aeruginosa, a leading pathogen in infectious pneumonia. Quantitative analysis revealed concentration-dependent antibacterial effects (Fig. 3A, Figure S5A, Supporting Information), with Fe/TNT@NM achieving 75.48% bacterial inhibition without US activation, which escalated to 90.63% under US irradiation (1.5 W/cm², 1.0 MHz) (Fig. 3B, C). Notably, US alone showed no significant antimicrobial activity compared to PBS controls. Live/dead fluorescent staining further corroborated these findings: while PBS-treated P. aeruginosa exhibited uniform green fluorescence (acridine orange, viable cells), Fe/TNT@NM-treated groups displayed prominent red fluorescence (ethidium bromide, dead cells), with near-complete bacterial lethality observed post-US activation (Fig. 3D). Scanning electron microscopy (SEM) confirmed the mechanism of action—untreated bacteria maintained intact membranes, whereas Fe/TNT@NM-exposed cells showed controllably induced membrane disruption, characterized by marked membrane invagination and cytoplasmic extravasation (Fig. 3E). These results demonstrate Fe/TNT@NM’s capacity to synergize exogenous US energy with intrinsic catalytic activity, achieving superior antibacterial efficacy through multimodal mechanisms.

Fig. 3.

Antibacterial activity of Fe/TNT@NM in vitro. (A) Inhibition of P. aeruginosa by Fe/TNT@NM at different concentrations under US. (B) Inhibition of P. aeruginosa by Fe/TNT@NM or PBS under US. (C) Representative diagram of P. aeruginosa inhibition by Fe/TNT@NM or PBS under US. (D) CLSM live/dead staining of P. aeruginosa under various treatments. (E) SEM images of P. aeruginosa after different treatments. (F) Remaining P. aeruginosa biofilm after Fe/TNT@NM treatment at different concentrations under US. (G) Remaining P. aeruginosa biofilm after Fe/TNT@NM treatment at different US times. (H) CLSM 3D images of P. aeruginosa biofilm under various treatments. (I) Remaining P. aeruginosa biofilm under various treatments. Data are represented as mean ± SD (n = 3). * P < 0.05, ** P < 0.01, *** P < 0.001, n.s P > 0.05, n.s, no significance

The disruption of P. aeruginosa biofilms by Fe/TNT@NM

The formidable antibiotic resistance and immune evasion capabilities of P. aeruginosa biofilms exacerbate infection severity and complicate clinical treatment [49]. To address this challenge, we systematically evaluated Fe/TNT@NM’s biofilm-disrupting efficacy using quantitative crystal violet staining. Fe/TNT@NM exhibited concentration-dependent biofilm inhibition, achieving 83% suppression at 250 µg/mL—a level statistically equivalent to higher concentrations (500 and 1000 µg/mL; 85% inhibition) (Fig. 3F; Figure S5B, Supporting Information). This established 250 µg/mL as the optimal therapeutic concentration. Time-course studies under US irradiation further demonstrated time-dependent biofilm inhibition, with maximal suppression (77%) observed at 8–10 min of US exposure (Fig. 3G; Figure S5C, Supporting Information). Subsequent experiments thus employed an 8-min US protocol. Structural analysis via FITC/EB dual staining revealed stark contrasts: PBS or US treated biofilms retained structural integrity and thickness, whereas Fe/TNT@NM treated biofilms—with or without US activation—showed marked disintegration and thinning (Fig. 3H). Quantitatively, Fe/TNT@NM alone achieved 45.48% biofilm clearance, escalating to 72.49% under US activation (Fig. 3I; Figure S5D, Supporting Information). These results confirm Fe/TNT@NM’s dual-action efficacy—directly eradicating planktonic bacteria through membrane disruption while destabilizing mature biofilms via oxidative and metabolic interference.

Extracellular oxidative stress storm elicited by Fe/TNT@NM

The extracellular oxidative stress mechanism underpinning Fe/TNT@NM’s antibacterial action was investigated through ROS generation analysis. As pivotal destructive agents in biofilm disruption and pathogen eradication [21, 50], ROS production in multidrug-resistant P. aeruginosa was quantified via flow cytometry and immunofluorescence. Flow cytometric profiles revealed baseline ROS levels in PBS- and US-only groups, whereas Fe/TNT@NM exposure induced significant ROS generation (1.8-fold increase vs. controls). US activation further amplified this response, evidenced by a rightward shift in fluorescence intensity peaks (510 nm channel) and a 2.3-fold ROS elevation compared to non-irradiated Fe/TNT@NM (Fig. 4A). Immunofluorescence imaging corroborated these findings: intense green fluorescence signals—indicative of ROS accumulation—were localized to Fe/TNT@NM-treated bacterial clusters, contrasting sharply with minimal signals in PBS/US controls (Fig. 4B). Collectively, these results confirm that US-activated Fe/TNT@NM orchestrates a precisely controlled oxidative storm at P. aeruginosa biofilm-microenvironment interfaces, destabilizing biofilm architecture while inducing lethal oxidative damage to embedded pathogens.

Fig. 4.

Extracellular and Intracellular antibacterial mechanisms of Fe/TNT@NM. (A) Flow cytometry analysis of ROS in P. aeruginosa treated with Fe/TNT@NM under US. Left, representative images, right, quantitative analysis. (B) Fluorescence microscope of ROS in P. aeruginosa treated with Fe/TNT@NM under US. (C) Cytoplasmic Fe²⁺ levels in P. aeruginosa after different treatments. (D) Lipid peroxidation in P. aeruginosa detected by flow cytometry with BODIPY-C11 after different treatments. (E) Fluorescence microscopy of lipid peroxidation in P. aeruginosa using the BODIPY-C11 after different treatments. (F) MDA levels in P. aeruginosa after different treatments. (G) GSH levels in P. aeruginosa after different treatments. (H) Fe/TNT@NM antimicrobial mechanism: extracellular sonodynamic and chemodynamic effects combined with intracellular ferroptosis. Data are represented as mean ± SD (n = 3). * P < 0.05, ** P < 0.01, *** P < 0.001, n.s P > 0.05, n.s, no significance

Fe/TNT@NM triggered intracellular bacterial ferroptosis

The ferroptosis-inducing capacity of Fe/TNT@NM in P. aeruginosa was mechanistically validated through iron metabolism dysregulation and lipid peroxidation assays. As demonstrated in Fig. 2E, US activation synergized with the acidic infection microenvironment to enhance Fe³⁺ release from Fe/TNT@NM. These Fe³⁺ ions were internalized via bacterial iron-siderophore transport systems and intracellularly reduced to Fe²⁺ [33], leading to a marked elevation in cytoplasmic Fe²⁺ levels compared to PBS controls (Fig. 4C). BODIPY-C11 probe was used to evaluate lipid peroxidation and ferroptosis in the P. aeruginosa. Both Fe/TNT@NM and US-activated Fe/TNT@NM induced significant lipid peroxidation in P. aeruginosa, as evidenced by a pronounced shift in fluorescence intensity at 510 nm. This oxidative cascade was effectively suppressed by the ferroptosis inhibitor Ferrostatin-1 (Fer-1) (Fig. 4D). Fluorescence microscopy further corroborated these findings: US-activated Fe/TNT@NM shifted emission from red (intact lipids) to green (oxidized lipids), while Fer-1 pretreatment reversed this phenomenon (Fig. 4E). Biochemical profiling confirmed ferroptotic mechanisms. US-activated Fe/TNT@NM elevated malondialdehyde (MDA)—a lipid peroxidation marker—while depleting glutathione (GSH), the primary cellular antioxidant. These alterations were significantly attenuated by Fer-1, confirming ferroptosis as the dominant cell death pathway (Fig. 4F, G). Collectively, these results demonstrate Fe/TNT@NM’s dual bactericidal action: Extracellular oxidative storm via US-triggered ROS generation (sonodynamic/chemodynamic effects); Intracellular ferroptosis driven by iron overload-induced lipid peroxidation. This bidirectional mechanism synergistically disrupts P. aeruginosa through membrane damage and metabolic collapse (Fig. 4H).

Biocompatibility evaluation of Fe/TNT@NM

The biocompatibility of Fe/TNT@NM was rigorously validated through multi-level safety assessments. Beyond material safety, we ensured ultrasound biosafety in pulmonary applications using optimized parameters (1.0 MHz, 1.5 W/cm², 50% duty cycle) [51–55]. In vitro hemolysis assays demonstrated exceptional blood compatibility, with Fe/TNT@NM inducing < 5% hemolysis at concentrations up to 500 µg/mL (Fig. 5A). Cellular viability assays revealed > 75% survival rates in human umbilical vein endothelial cells (HUVECs) and Bronchial epithelial cells-2 bronchial (BEAS-2B) even at maximum therapeutic doses (200 µg/mL), confirming minimal cytotoxicity (Fig. 5B, C). Flow cytometry analysis further showed > 90% viability in BEAS-2B cells across all treatment groups (Fe/TNT@NM, alongside US and Fer-1), with apoptosis/necrosis rates comparable to PBS controls (Fig. 5D-H). Crucially, MDA levels remained unchanged, excluding ferroptotic damage to mammalian cells (Fig. 5I). In vivo biosafety profiling in healthy mice revealed no hematological abnormalities: complete blood counts (RBC, HGB, WBC, HCT, MCV, MCH, MCHC, PLT) (Fig. 5J-L; Figure S6A-F, Supporting Information) and serum biochemistry (ALT, AST, ALP, CK, CREA) (Fig. 5M, N; Figure S6G-J, Supporting Information) all fell within physiological ranges. Histopathological examination of major organs (heart, liver, spleen, lungs, kidneys) showed intact tissue architecture without inflammatory infiltration or necrosis (Fig. 5O). These results collectively establish Fe/TNT@NM as a biosafe nanotherapeutic platform, combining targeted antimicrobial efficacy with systemic biocompatibility for pulmonary applications.

Fig. 5.

Biocompatibility and biosafety of Fe/TNT@NM. (A) Hemolytic activity of red blood cells exposed to Fe/TNT@NM at different concentration. (B) Viability of HUVECs treated with various concentrations of Fe/TNT@NM. (C) Viability of BEAS-2B cells treated with various concentrations of Fe/TNT@NM. (D) Flow cytometry analysis of apoptosis and necrosis in BEAS-2B cells treated with US, Fe/TNT@NM, or Fer-1. Proportion of viable (E), necrotic (F), late apoptotic (G), and early apoptotic (H) BEAS-2B cells among various treatment groups. (I) MDA levels in BEAS-2B cells treated with US, Fe/TNT@NM, or Fer-1. RBC (J), HGB (K), and WBC (L) levels in healthy mice 7 days after treatment with US, Fe/TNT@NM, or Fer-1. ALT (M) and AST (N) levels in serum of healthy mice 7 days after treatment with US, Fe/TNT@NM, or Fer-1. (O) Histopathological analysis of heart, liver, spleen, lung, and kidney in healthy mice 7 days after treatment with US, Fe/TNT@NM, or Fer-1. Data are represented as mean ± SD (n = 3)

Alleviation of P. aeruginosa pneumonia by Fe/TNT@NM

Encouraged by the robust in vitro biofilm-disrupting activity of Fe/TNT@NM, we established a murine model of P. aeruginosa pneumonia and further evaluated the in vivo therapeutic efficacy of Fe/TNT@NM in conjunction with US irradiation and Fer-1 treatment (Fig. 6A). To achieve effective penetration through pulmonary tissue, we employed ultrasound parameters (1.0 MHz, 50% duty cycle, 1.5 W/cm²) known to balance deep penetration (> 3–5 cm in vivo) with minimized thermal effects, ensuring coverage of the entire murine lung parenchyma (5–7 mm thick) [56]. Histopathological analysis of lung tissues via hematoxylin and eosin (H&E) staining revealed severe inflammatory infiltration and lung tissue damage in untreated infected mice. US-activated Fe/TNT@NM markedly attenuated these pathological changes, restoring near-normal lung architecture. However, treatment with Fer-1 reversed these therapeutic effects (Fig. 6B; Figure S7A, Supporting Information), confirming ferroptosis as a critical therapeutic mechanism. To assess the inflammatory response of P. aeruginosa-infected lung tissues after Fe/TNT@NM treatment, RT-qPCR analysis demonstrated significant upregulation of cytokines (IL-1β, TNF-α, IL-6) in infected lungs compared to healthy controls. While standalone US or Fe/TNT@NM showed limited impact, their combination synergistically suppressed cytokine expression to baseline levels. Notably, Fer-1 treatment partially reversed this suppression, indicating that bacterial ferroptosis triggered by Fe/TNT@NM-induced iron overload is central to resolving inflammation (Fig. 6C; Figure S7B, C, Supporting Information).

Fig. 6.

Therapeutic effect of Fe/TNT@NM of P. aeruginosa pneumonia in vivo. (A) Experimental procedure for treating P. aeruginosa pneumonia. (B) HE staining of lung tissue injury in mice under different treatments. (C) Relative expression of IL-1β, TNF-α, and IL-6 in lung tissue under different treatments. (D) Quantification of residual P. aeruginosa in mouse lung tissues under different treatments. (E) Representative images of residual P. aeruginosa by plate count method. (F) Analysis of residual P. aeruginosa in mouse lung tissue under different treatments. (G) Gram stains of residual P. aeruginosa in mouse lung tissue. (H) Flow cytometry of lipid peroxidation in residual P. aeruginosa in mouse lung tissue. (I) Quantification of lipid peroxidation in residual P. aeruginosa in mouse lung tissue. Data are represented as mean ± SD (n = 5). * P < 0.05, ** P < 0.01, *** P < 0.001, n.s P >0.05, n.s, no significance

The in vivo bactericidal efficacy of Fe/TNT@NM was systematically assessed in a murine P. aeruginosa pneumonia model. While standalone US irradiation or Fe/TNT@NM administration significantly reduced pulmonary bacterial burdens, their combined application achieved near-complete pathogen eradication, with residual bacterial loads reduced by 92% compared to PBS controls (Fig. 6D, E; Figure S7D, Supporting Information). Gram staining consistently corroborated the significant inhibitory impact of Fe/TNT@NM on P. aeruginosa proliferation, with a pronounced enhancement in eradication efficacy when combined with US treatment (Fig. 6F, G; Figure S7E, Supporting Information). Crucially, treatment with Fer-1 significantly weakened the bacteriostatic effect mediated by Fe/TNT@NM and US irradiation on P. aeruginosa. Mechanistic evaluation revealed that US-activated Fe/TNT@NM triggered robust lipid peroxidation in intrapulmonary P. aeruginosa, which was partially mitigated by Fer-1 treatment (Fig. 6H, I), confirming bacterial ferroptosis triggered by Fe/TNT@NM as the dominant bactericidal pathway. The collective findings underscored the potent antimicrobial efficacy of Fe/TNT@NM in an in vivo P. aeruginosa pneumonia model, significantly attributed to the triggering of intracellular ferroptosis in P. aeruginosa. The targeted bactericidal effect led to a significant reduction in the pulmonary infiltration of pro-inflammatory cytokines, thereby \substantially mitigating lung injury associated with P. aeruginosa infection.

Conclusion

In summary, we rationally developed Fe/TNT@NM, an ultrasound-responsive biomimetic nanotherapeutic platform that integrates exogenous (ultrasound) and endogenous (microenvironmental) activation mechanisms to achieve comprehensive biofilm eradication in Pseudomonas aeruginosa pneumonia. Our design leverages precision-controlled ultrasound irradiation to orchestrate a dual ROS assault: extracellular ROS generated through synergistic SDT and Fenton reaction-driven CDT, coupled with intracellular lipid peroxidation via iron overload-induced ferroptosis. Distinct from conventional ROS-generating systems, Fe/TNT@NM exhibits compartmentalized therapeutic action. Externally, SDT and CDT synergistically degrade biofilm integrity through oxidative damage, while internally, liberated Fe²⁺ ions hijack bacterial iron metabolism, accumulating to cytotoxic levels that trigger membrane disintegration via lipid ROS. The neutrophil membrane coating further enhances therapeutic efficacy by enabling immune evasion through biomimetic surface camouflage and inflammation-targeted delivery via adhesion receptor mediation. This dual functionality ensures rapid accumulation at infection sites while minimizing systemic clearance.

While our murine model of P. aeruginosa pneumonia demonstrated robust therapeutic efficacy, it is important to note that this system may not fully recapitulate the complexity of human pulmonary infections, particularly in immunocompromised hosts. Nevertheless, the bidirectional therapeutic mechanism of Fe/TNT@NM—simultaneously targeting extracellular biofilm architecture and intracellular bacterial viability—addresses a critical limitation of conventional monotherapies, which typically struggle to penetrate biofilm barriers or neutralize bacterial redox defense systems. By integrating extracellular oxidative stress induction with intracellular iron overload-triggered ferroptosis, this approach achieves synergistic biofilm eradication while circumventing resistance mechanisms. This dual-targeting strategy holds broad implications for advancing next-generation therapies against recalcitrant infections characterized by biofilm persistence and multidrug resistance.

Methods

Collection of mouse neutrophils. Female ICR mice (6–8 weeks old) were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. All experimental procedures were conducted in compliance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals, and approved by the Ethics Committees of Tianjin Medical University General Hospital (approval number: IRB2024-DWFL-055). For neutrophil isolation, mice received intraperitoneal injection of lipopolysaccharide (LPS; 1.5 mg/kg body weight) to induce systemic inflammation. 6 h post-injection, whole blood was collected via cardiac puncture under terminal anesthesia. Following erythrocyte lysis, leukocyte-rich fractions were separated through discontinuous Percoll density gradient centrifugation (52%, 69%, 78%) at 1,500 × g for 30 min (4 °C). Neutrophil populations were isolated from both the 69%/78% interface and the upper 78% layer. Cell characterization was performed using anti-LFA-1 antibody staining followed by flow cytometric analysis to verify activation status. Purified neutrophils were subsequently washed with PBS and cryopreserved at −80 °C.

Extraction of neutrophils membrane. Neutrophil membrane extraction was performed using ice-cold hypotonic lysis buffer (30 mM Tris-HCl, 225 mM mannitol, 75 mM sucrose, 0.2 mM EGTA) supplemented with protease/phosphatase inhibitor cocktail (1:100 v/v). Following cellular disruption through 30 strokes in a Dounce homogenizer, the cellular suspension was collected and centrifuged at 4 °C at 20,000 × g for 25 min. The supernatant was then centrifuged at 4 °C at 100,000 × g for 35 min to obtain the cell membrane fraction, which was subsequently stored at −80 °C.

Materials. Sodium hydroxide (NaOH) and iron (III) chloride were purchased from Tianjin Damao Chemical Reagent Factory. Titanium dioxide was purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. Hydrochloric acid was purchased from Tianjin Bohua Chemical Reagent Co., Ltd. Anhydrous ethanol was purchased from the Tianjin Hedong District Quanta Service Department. All reagents were of analytical grade and were used directly without further purification. Deionized water was used in this study.

Synthesis of Fe/TNT. Titanate nanotubes (TiO₂NTs) were synthesized hydrothermally according to our previous work [57]. 30 g NaOH, 1.6 g TiO₂ and 75 mL deionized water were mixed in a polytetrafluoroethylene reactor at 150℃ for 10 h. The product was centrifuged and washed until neutral. The pH was adjusted by adding concentrated hydrochloric acid. After centrifugal washing and neutralization, TiO₂NTs was prepared by drying at 80℃. Subsequently, 0.16 g TiO₂NTs were added to 30 mL of 1 mol/L ferric chloride solution and subjected to continuous stirring for 24 h to facilitate ion exchange. The Fe/TNT nanotubes were collected by centrifugation, and washed to remove the residue metal cations, and then dried overnight at 80 °C.

Synthesis of Fe/TNT@NM nanocomposite. Cell membrane-coated NPs were synthesized according to established methods [38]. Briefly, the neutrophil membrane was mixed with Fe/TNT in a weight ratio of 2:1 and subjected to bath sonication for 5 min to prepare Fe/TNT@NM. The structure and morphology of Fe/TNT@NM were observed using Transmission Electron Microscopy (TEM), High-Resolution Transmission Electron Microscopy (HR-TEM), and Scanning Electron Microscopy (SEM). The crystalline structure of Fe/TNT@NM was analyzed using X-ray Diffraction (XRD). The chemical composition of Fe/TNT@NM was determined by X-ray Photoelectron Spectroscopy (XPS) and Energy Dispersive X-ray Spectroscopy (EDS). The absorption spectrum was obtained in the range of 200–800 nm using Ultraviolet-Visible Diffuse Reflectance Spectroscopy (UV-vis, PE λ 750). Fourier Transform Infrared Spectroscopy (FTIR) was recorded using a Nicolet iS50. Dynamic Light Scattering (DLS) was employed to measure the hydrodynamic diameter, particle size, polydispersity index, and Zeta potential of Fe/TNT@NM.

Flow cytometric analysis. The abundance of LFA-1 in Fe/TNT@NM was assessed using flow cytometry. Fresh suspensions of Fe/TNT, Fe/TNT@NM, and neutrophils were prepared. Anti-mouse LFA-1 antibody was used for staining at 4 °C for 40 min. The fluorescence intensity of LFA-1 was measured using a Becton Dickinson FACSCanto-II flow cytometer and analyzed by FlowJo software.

Coomassie bright blue staining. The membrane protein composition of Fe/TNT@NM was analyzed using Coomassie Brilliant Blue staining. Firstly, we performed SDS-PAGE gel electrophoresis on both the labeled proteins and the sample proteins. Then, the gel was completely immersed in Coomassie Brilliant Blue staining solution and incubated at 4 °C overnight. The following day, the gel was rinsed with deionized water and incubated with a destaining solution at room temperature for 24 h. The destaining solution was replaced every 4 h until the blue background faded and distinct protein bands became visible. Lastly, the protein expression levels among different groups were compared and analyzed.

Western blotting. The samples from different groups were subjected to SDS-PAGE gel electrophoresis. The proteins were transferred from the gel onto a PVDF membrane. The membrane was then blocked with a 5% solution of non-fat dry milk. Following this, the membrane was incubated with the primary antibodies overnight at 4 °C, including antibodies against mouse tumor necrosis factor-alpha receptor (TNF-R), interleukin-1 receptor (IL-1R), and lymphocyte function-associated antigen-1 (LFA-1). The next day, the membrane was washed three times with TBST buffer and then incubated with the horseradish peroxidase (HRP)-conjugated secondary antibody for 1 h at room temperature. Finally, the expression levels of the target proteins were detected and analyzed using a ChemiDoc Imaging System.

Cell phagocytosis assay. Mouse alveolar macrophages (MH-S) were seeded in 12-well tissue culture plates at a confluence of 50% and cultured overnight. Fe/TNT@NM and Fe/TNT labeled with RB dye were added to the MH-S supernatant and incubated for 72 h. After washing three times with PBS, the cells were blocked with 0.5% BSA for 30 min. Tubulin antibodies were incubated overnight. The following day, an F488 fluorescent secondary antibody was incubated at room temperature for 1 h in the dark. The cell nuclei were stained using the nuclear dye DAPI. After mounting the coverslip, the cells were observed under a fluorescence microscope to assess the phagocytic effect of Fe/TNT@NM and Fe/TNT.

Detection of Singlet Oxygen (¹O₂) under US irradiation. The ¹O₂ generation was detected using the DPBF probe to evaluate the sonodynamic efficiency of Fe/TNT@NM. A 1 mL solution of Fe/TNT@NM (0.3 mg/mL) was mixed with 30 µL of DPBF (3 mg/mL). The absorbance change of DPBF at 416 nm was measured under ultrasonic conditions (1.0 MHz, 50% duty cycle, 1.5 W/cm²) at different times (0, 2, 4, 6, 8, and 10 min), indicating the production of ¹O₂. Water treated under the same ultrasonic conditions was used for comparison.

Detection of Hydroxyl Radical (•OH). The ·OH generation was detected using the TMB probe to evaluate the chemodynamic efficiency of Fe/TNT@NM. A 1 mL solution of Fe/TNT@NM (0.3 mg/mL) was incubated with different concentrations of H₂O₂ (0, 31.3, 62.5, 125, 250, 500, and 1000 µM). Then, 15 µL of TMB (10 mg/mL) was added to the mixture. After incubation with H₂O₂ at different times (0, 2, 5, 10, and 20 min), the absorbance change of TMB at 652 nm was measured, indicating the production of ·OH due to the Fenton reaction. Meanwhile, the Michaelis-Menten equation was used for nonlinear regression fitting, where V_max represents the maximum reaction rate, S represents the substrate concentration, and K_m is the Michaelis constant. K_m and V_max were calculated from a double reciprocal plot (or Lineweaver-Burk plot).

US enhanced •OH generation. The combined activity of sonodynamic and chemodynamic efficiency of Fe/TNT@NM was detected using the OPD probe. A 1 mL solution of Fe/TNT@NM (0.3 mg/mL) was mixed with H₂O₂ (1000 µM). Then, 15 µL of OPD (10 mg/mL) was added to the mixture. After exposure to US at different times (0, 2, 4, 6, 8, and 10 min), the absorbance changes of OPD at 440 nm were measured to reflect the production of ·OH by US and the Fenton reaction.

Evaluation of iron release. The release of Fe ions from Fe/TNT@NM was assessed under acidic conditions (pH 5.5) at different ultrasonication times. In brief, Fe/TNT@NM was incubated in water at pH 5.5 and treated with ultrasonication for varying durations: 0, 2, 4, 6, 8, and 10 min. The average concentration of free Fe ions in the supernatant was determined using Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES). All experiments were performed in triplicate.

The Plate Culture Counting. A bacterial suspension of multidrug-resistant P. aeruginosa strains was centrifuged at 5000 rpm for 5 min, after which the supernatant was discarded. The bacteria were then washed three times with PBS. The bacterial OD600 was adjusted to 1.023 using PBS. The suspension was divided into four groups: the PBS group, the Fe/TNT@NM group, the PBS + US group, and the Fe/TNT@NM + US group. Briefly, each bacterial suspension was incubated with PBS or Fe/TNT@NM at 37 °C for 2 h. Subsequently, the suspensions in the US groups were exposed to ultrasound (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 10 min. The bacterial suspensions were then diluted and spread onto solid agar plates, followed by incubation at 37 °C for 16–18 h. Colony counts were performed the following day. Each experimental group was repeated three times for statistical analysis.

The Dead/Live Bacterial fluorescent staining. The bacterial suspension of multidrug-resistant P. aeruginosa strains was collected by centrifugation. The OD600 of the bacterial suspension was adjusted to approximately 1.2 using PBS. The suspension was divided into four groups: the PBS group, the Fe/TNT@NM group, the PBS + US group, and the Fe/TNT@NM + US group. Briefly, each bacterial suspension was incubated with either PBS or Fe/TNT@NM at 37 °C for 2 h. Subsequently, the suspensions in the US groups were exposed to ultrasound (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 10 min. Each group was treated with AO and EB staining solutions and incubated at 37 °C in the dark for 15 min. The bacterial suspensions were washed with PBS and centrifuged to remove the supernatant. The bacteria were then observed using CLSM after drying overnight.

Crystal violet staining assay. The bacterial suspension of multidrug-resistant P. aeruginosa strains was collected by centrifugation. The OD600 of the bacterial suspension was adjusted to approximately 0.05 using PBS. The suspension was plated in a 96-well plate and incubated at 37 °C for 24 h to cultivate the biofilms. The following day, the biofilms were immersed in either PBS or Fe/TNT@NM at 37 °C for 4 h. Subsequently, the biofilms were exposed to ultrasound (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 5 min. Each group was then fixed with methanol for 40 min and stained with crystal violet at room temperature for 30 min. 33% acetic acid was added to each well and left at room temperature for 1 h. The optical density at 590 nm was measured using an enzyme-linked immunosorbent assay (ELISA) reader. Each experimental group was repeated three times for statistical analysis.

CLSM 3D Imaging Assay. A bacterial suspension of multidrug-resistant P. aeruginosa strains with an OD600 of 0.05 was plated in a 6-well plate and incubated at 37 °C for 48 h to cultivate a biofilm. On the following day, the biofilms were gently washed with PBS. Each group was then incubated with either PBS or Fe/TNT@NM at 37 °C for 4 h and exposed to ultrasound (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 8 min. Subsequently, the biofilms from each group were fixed with 4% paraformaldehyde for 4 h and incubated with FITC-conjugated concanavalin A (FITC-conA) and ethidium bromide (EB) staining solutions at 37 °C in the dark for 15 min. The bacterial biofilms were then observed using CLSM.

Ferroptosis in P. aeruginosa. The OD600 of the bacterial suspension of multidrug-resistant P. aeruginosa strains was adjusted to approximately 1.0. The suspension was divided into five groups: the PBS group, the Fe/TNT@NM group, the PBS + US group, the Fe/TNT@NM + US group, and the Fe/TNT@NM + US + Ferrostatin-1 (Fer-1) group. Briefly, each bacterial suspension was treated with either PBS or Fe/TNT@NM at 37 °C for 2 h. Subsequently, the suspensions in the US groups were exposed to ultrasound (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 10 min. The DCFH-DA probe was used to evaluate the ROS levels in P. aeruginosa. The level of lipid peroxidation in P. aeruginosa was detected using the BODIPY581/591-C11 fluorescent probe and an MDA detection kit. The levels of glutathione were detected using a GSH assay kit.

Hemolysis Analysis. Red blood cells were obtained from the whole blood of healthy mice. An erythrocyte suspension with a final concentration of 2% was prepared by diluting with PBS. Then, different gradient concentrations of Fe/TNT@NM (500, 250, 125, 62.5, 31.2, 15.6, 7.8, 3.9, and 2 µg/ml), PBS (as a negative control), and Triton-X-100 (as a positive control) were incubated with the same amount of 2% erythrocyte suspension at 37 °C for 1 h, respectively. Finally, the supernatant was collected, and the absorbance at 540 nm was measured using a Microplate Reader. The hemolysis rate (HR) of different groups was quantitatively analyzed according to the following equation: HR (%) = (Ab sample – Ab PBS)/(Ab Triton-X-100 – Ab PBS) × 100%. Each experimental group was repeated three times for statistical analysis.

Cell viability assay. Human umbilical vein endothelial cells (HUVECs) and human normal lung epithelial cells (BEAS-2B) were purchased from the BeNa Culture Collection (BNCC). HUVECs were cultured in a specialized medium containing fetal bovine serum, heparin, vascular endothelial growth factor, and F12. BEAS-2B cells were routinely cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with F12 and fetal bovine serum. The cells were grown in a humidified environment at 37 °C with 5% CO₂ and 95% air. Cell viability was assessed using the MTS assay. When HUVECs and BEAS-2Bs reached 90% confluence, the cells were trypsinized and then seeded in 96-well plates at a specified density for 24 h of culture. The next day, the cells were incubated with different concentrations of Fe/TNT@NM (500, 250, 125, 62.5, 31.2, 15.6, 7.8, 3.9, and 2 µg/mL) for 24 h. Finally, 20 µL of MTS reagent was added to each well and incubated at 37 °C for 2 h. The absorbance was measured at 545 nm using a microplate reader. Each experimental group was repeated three times for statistical analysis.

Apoptosis and necrosis assay. Cell apoptosis and necrosis were measured using a PE Annexin V apoptosis detection kit according to the manufacturer’s instructions. When BEAS-2B cells reached 90% confluence, they were trypsinized and seeded in 6-well plates. The next day, the cells were incubated with Fe/TNT@NM (200 µg/mL) and Fer-1 (2 µM) for 24 h and then exposed to US (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 10 min. PBS treatment served as a negative control. The cells were then rinsed with cold PBS, stained with Annexin V-PE and 7-AAD, and incubated in the dark at room temperature for 15 min. Finally, the proportions of apoptosis (early and late stages), necrosis, and viable cells were analyzed by flow cytometry. Each experimental group was repeated three times for statistical analysis.

MDA levels detection. Intracellular MDA levels were assessed using an MDA assay kit. BEAS-2B cells were collected and seeded in 6-well plates. The cells were treated with PBS, Fe/TNT@NM (200 µg/mL), and Fer-1 (2 µM) for 24 h, and then irradiated with US (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 10 min. The cells were fully lysed with extraction solution, followed by sonication in 3-second on and 10-second off cycles for 90 s. The supernatant was collected after centrifugation at 8000 g at 4 °C for 10 min. MDA detection solution was mixed into the supernatant and incubated in a 100 °C water bath for 60 min. The absorbance at 523 nm was measured using a microplate reader after the mixture had cooled to room temperature in an ice bath. Each experimental group was repeated three times for statistical analysis.

Biosafety and toxicity in vivo. Healthy ICR mice were treated with Fe/TNT@NM (200 µg/mL) and Fer-1 (0.8 mg/kg). The mice were anesthetized via intraperitoneal injection of 10% chloral hydrate. Anticoagulant whole blood was collected from the heart for routine blood tests, including assessments of red blood cells (RBC), hemoglobin (HGB), white blood cells (WBC), neutrophils, hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), and platelets (PLT). The blood was left at room temperature for 1 h and then centrifuged at 3500 rpm for 15 min. The supernatant serum was collected for biochemical analysis, including measurements of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), creatine kinase (CK), creatinine (CREA), and urea (UREA). Moreover, the heart, liver, spleen, lung, and kidney tissues were fixed for 24 h and embedded in paraffin. The pathological structures of each tissue were observed and analyzed following hematoxylin and eosin (H&E) staining.

Efficacy of Fe/TNT@NM in P. aeruginosa Pneumonia model. The murine pneumonia model infected with multidrug-resistant P. aeruginosa was performed as previously described [58]. Briefly, fresh bacterial suspensions of multidrug-resistant P. aeruginosa strains were cultured in LB medium at 37 °C overnight. The following day, P. aeruginosa was collected by centrifugation and washed three times with PBS. The concentration of the bacterial suspension was adjusted to 1 × 10⁷ CFU/mL. Mice were anesthetized with an intraperitoneal injection of 10% chloral hydrate and then infected intratracheally with 50 µL of fresh P. aeruginosa suspension for 24 h.

After infection, all animals were divided into the following groups: Normal, PBS, Fe/TNT@NM, Fe/TNT@NM + US, and Fe/TNT@NM + US + Fer-1. Mice in the PBS and Fe/TNT@NM groups received the same dose of PBS or Fe/TNT@NM (200 µg/mL) intranasally, the clinically preferred route for pulmonary infections because it delivers agents directly to the infected lung (Figure S8). The final group of mice received Fer-1 via tail vein injection at a dose of 0.8 mg/kg. Mice in the US treatment groups were exposed to US irradiation (1.0 MHz, 50% duty cycle, 1.5 W/cm²) for 6 min, with a 1-min pause every 2 min. The entire treatment period was maintained for a total of 5 days. At the end of the experiment, lung tissues from each group of mice were collected for histopathological analysis.

Residual P. aeruginosa in lung tissues. At the end of the experiment, mice were euthanized, and lung tissues were excised. The lung tissues were rinsed with PBS and then homogenized. The homogenate was centrifuged at 1500 rpm for 5 min to remove cells and debris. The supernatant was collected and centrifuged again at 5000 rpm for 5 min to harvest residual bacteria from the lung tissue. The bacteria were spread onto agar plates and incubated at 37 °C for 16–18 h. The bacterial CFU were counted the following morning. The experiment was repeated three times for statistical analysis.

P. aeruginosa ferroptosis in lung tissues. At the conclusion of the experiment, the mice were euthanized, and lung tissues were excised. The lung tissues were washed with PBS and then homogenized. The homogenate was centrifuged at 1500 rpm for 5 min to pellet cells and debris. The supernatant was collected, and the pellet was resuspended and then centrifuged again at 5000 rpm for 5 min to harvest residual bacteria from the lung tissue. The bacterial pellets were resuspended in PBS and stained with C11-BODIPY for 30 min at 4 °C. The level of ferroptosis in the bacteria was analyzed using fluorescence microscopy and flow cytometry.

Inflammatory cytokine levels. After different treatments, mouse lung tissues were isolated on ice and rinsed with PBS. The lung tissues were homogenized with an appropriate amount of Trizol on ice for 10 min. Total RNA was extracted from the lung tissues according to the manufacturer’s instructions, and reverse transcription was performed using a reverse transcription kit. Finally, the levels of inflammatory cytokines (IL-1β, TNF-α, and IL-6) in the lung tissues of mice from different groups were detected by real-time quantitative PCR.

Histopathological analysis. After different treatments, the lung tissues of mice were collected and fixed in 10% formalin for 24 h. Subsequently, the tissues were subjected to paraffin embedding and sectioning. Finally, the tissue sections were stained with hematoxylin and eosin, and the pathological changes in the lung tissues were observed under a microscope.

Ultrasound parameters. The selection of US parameters was rigorously optimized based on three critical factors: (1) deep tissue penetration requirements for pulmonary therapy, (2) minimization of bioeffects in air-filled tissues, and (3) synergy with nanosonosensitizer activation kinetics. At 1.0 MHz, ultrasound achieves 3–5 cm penetration depth in vivo, far exceeding murine lung thickness (5–7 mm) [56, 59]. This frequency minimizes acoustic scattering at air-tissue interfaces, ensuring energy delivery to deep pulmonary lesions [56, 60, 61]. Moreover, low-frequency US (0.8–1.5 MHz) is clinically established for transthoracic therapies (e.g., sonothrombolysis, drug delivery) due to its ability to bypass ribcage shadows and reach peripheral lung zones. Pulmonary ultrasound safety is governed by the thermal and mechanical indices: maintaining a thermal index ≤ 0.7, safely below the FDA limit of 1.0 for lung exposure, and a mechanical index < 0.4, well under the cavitation threshold of 0.7, ensures that exposure at 1.5 W/cm² lies entirely within the established safety envelope for murine lung tissue [51, 52]. This intensity falls within the 1.0–2.0 W/cm² range routinely employed in clinical pulmonary sonotherapy, such as nebulizer-enhanced antibiotic delivery, underscoring the translational relevance and safety margin of our chosen parameters. In summary, we selected 1.5 W/cm², 1 MHz and 50% duty cycle ultrasound, which penetrates the entire mouse pulmonary parenchyma without inducing thermal damage.

Statistical analysis. All quantitative data were first tested for normality using the Shapiro-Wilk test (α = 0.05). Results are presented as mean ± standard deviation from at least three independent experiments (n ≥ 3). Statistical comparisons among multiple groups were conducted with one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test, whereas two-group comparisons were performed using an unpaired two-tailed Student’s t-test. Statistical significance was considered when *P < 0.05, **P < 0.01, and ***P < 0.001.

Author contributions

Chunhong Du (Writing – original draft: Lead)Shuai Wang (Data curation: Lead; Methodology: Lead)Yijie Cheng (Methodology: Supporting)Jie Li (Software: Supporting)Yufei Zhang (Formal analysis: Supporting)Baolin Zhu (Formal analysis: Equal; Resources: Supporting)Zhongming Wu (Funding acquisition: Equal; Visualization: Supporting)Xinge Zhang (Funding acquisition: Lead; Writing – review & editing: Supporting)Lingyi Zhou (Data curation: Lead; Writing – original draft: Equal; Writing – review &editing: Lead).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.