Biomimetic PD-1-nanovesicles inducing tri-modal sonodynamic-chemo-immunotherapy for breast cancer therapy and lung metastasis inhibition

Abstract

Background

Current cancer therapeutic approaches demonstrate constrained efficacy and fail to effectively address tumor metastasis, underscoring the critical need for innovative treatment modalities.

Results

In this study, we developed an innovative PD-1⁺ cell membrane-coated nanoplatform (PD-1-Au@DOX) that integrates ultrasound (US)-responsive gold nanoparticles (AuNPs), doxorubicin (DOX) delivery, and PD-L1 blockade into a synergistic therapeutic system. In vitro studies demonstrate that PD-1-Au@DOX enables targeted delivery of DOX and AuNPs to PD-L1-expressing breast cancer cells (4T1 and MCF-7) via surface-bound PD-1, while concurrently blocking PD-L1 to initiate an immune response. Under US irradiation, AuNPs generate substantial reactive oxygen species (ROS), deplete intracellular glutathione, and increase DOX release. The synergistic interplay of these three pathways culminates in pronounced apoptosis in both 4T1 and MCF-7 cell lines. In murine breast cancer models, PD-1-Au@DOX + US demonstrated significant efficacy in primary tumor suppression, as evidenced by robust growth inhibition and prolonged survival. The coordinated activation of immune responses through PD-1/PD-L1 pathway modulation and US-mediated ferroptosis effectively suppressed lung metastasis.

Conclusions

These findings establish PD-1-Au@DOX + US as a tri-modal approach for breast cancer treatment and metastasis prevention, providing an effective new clinical pathway.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-04013-6.

Introduction

Breast cancer remains one of the most prevalent cancers worldwide [1]. The current standard of care for breast cancer, which includes surgical resection, radiotherapy, chemotherapy, and immunotherapy, continues to yield suboptimal outcomes. This limited efficacy is largely attributable to the constraints of individual treatment modalities and the complex, heterogeneous nature of the tumor [2–4]. ‌Furthermore, the inability of current therapies to effectively address breast cancer metastasis significantly contributes to treatment failure [5, 6]. Current paradigms of cancer therapy face challenges stemming from drug resistance and significant treatment-related adverse effects [7], necessitating the development of effective alternative therapeutic strategies.

Sonodynamic therapy (SDT), photodynamic therapy (PDT), photothermal therapy (PTT), and chemodynamic therapy (CDT) all enable precise treatment at tumor sites. However, compared with other modalities, SDT offers significant advantages, including exceptional tissue penetration depth, a non-invasive and non-ionizing nature, and selective cytotoxicity toward malignant tissues [8], thereby conferring distinct benefits over photodynamic therapy and chemodynamic therapy [9]. Previous study indicates that gold nanoparticles (AuNPs) can generate. Reactive oxygen species (ROS) under sonodynamic activation [10]. Elevated intracellular ROS directly trigger DNA strand breaks, lipid peroxidation, and mitochondrial dysfunction, ultimately culminating in cell death [11]. However, conventional sonodynamic nanomedicines typically rely on a single ROS generation mechanism and face limitations in overcoming intra-tumoral barriers, thereby hindering their therapeutic efficacy against cancer [12].

Immunotherapy has made significant advances in cancer treatment and has emerged as a promising therapeutic strategy for patients with advanced cancers [13–15]. Immune checkpoint inhibitors (ICIs), particularly anti-PD-1 antibodies, have become the cornerstone of cancer immunotherapy, revolutionizing the treatment landscape for numerous malignancies and gaining FDA approval for clinical use across multiple cancer types [16, 17]. However, anti-PD-L1 monoclonal antibodies demonstrate low efficacy in patients [18], potentially attributable to insufficient tumor-associated antigen (TAA) release and the highly immunosuppressive nature of the tumor microenvironment (TME) [19]. Targeting tumor cells via tumor-specific molecular targets represents a crucial strategy for enhancing the precision of cancer therapy and minimizing off-target toxicities. Studies have reported that spermine can be exploited to functionalize nanocarriers for tumor cell targeting through the polyamine transport system, thereby enabling precise drug delivery [20]. Similarly, nanoparticles functionalized with anti–PD-L1 antibodies have been shown to specifically recognize PD-L-1 expressed on the surface of tumor cells, facilitating tumor-targeted delivery [21]. Consequently, PD-1–based nanoparticle systems not only block PD-L1 on tumor cells to restore antitumor immune responses but also enhance drug delivery efficiency while reducing systemic toxicity and off-target adverse effects.

Biomimetic drug delivery systems offer a promising approach to overcoming these challenges through improved tumor targeting [22]. Cell membrane-coated nanoparticles integrate the drug-loading capabilities of nanoparticles with the biological functionalities inherent to cell membranes. This combination enables prolonged circulation, homotypic targeting, and immune evasion, ultimately leading to enhanced tumor accumulation [23]. Furthermore, engineered cells can express specific targeting proteins on their surface, thus enabling targeted delivery of nanocarriers coated with cell membranes [24]. Controlled drug release can effectively minimize the premature release of nanocarriers in non-tumoral tissues, thereby reducing off-target toxicities. As previously reported, prodrugs engineered with disulfide bonds enable redox-responsive drug release that mitigates systemic side effects [25, 26]. Additionally, ultrasound-triggered controlled release from membrane-based nanocarriers allows spatiotemporally precise payload delivery at tumor sites [27]. Together, these features position biomimetic membrane-based nanodelivery systems as a promising strategy for achieving high therapeutic efficacy with low toxicity in cancer treatment.

In this study, we developed a biomimetic nanoplatform (PD-1-Au@DOX) by coating nanoparticles with PD-1-engineered cell membranes and co-encapsulating doxorubicin (DOX) and ultrasound-responsive gold nanoparticles (AuNPs). This system integrates tumor-targeted immune checkpoint blockade, chemotherapy, and sonodynamic therapy (SDT) within a single construct. Unlike systemic anti–PD-1/PD-L1 antibodies that cause immune-related toxicities, our approach exploits surface-displayed PD-1 to achieve localized PD-L1 blockade in the tumor microenvironment. This not only enhances tumor accumulation through PD-1/PD-L1-mediated recognition but also initiates T cell–dependent antitumor immunity without systemic immune activation. Critically, the engineered membrane functions as an active immunomodulatory interface—where PD-1 serves dual roles as both a targeting ligand and a checkpoint inhibitor. Upon ultrasound irradiation, the co-delivered AuNPs generate reactive oxygen species (ROS), while DOX is released in a spatiotemporally controlled manner. Notably, both DOX and ROS are established inducers of ferroptosis—a form of immunogenic cell death—thereby enabling synergistic tumor killing and immune stimulation. In vitro, PD-1-Au@DOX selectively bound PD-L1-expressing 4T1 and MCF-7 cells and efficiently delivered its cargo, resulting in potent cytotoxicity. In murine breast cancer models, the platform significantly suppressed tumor growth, prolonged survival, and reduced metastasis. Together, this work presents a rationally integrated strategy that merges precision checkpoint modulation with chemo–sonodynamic therapy, highlighting its promise for clinical translation in breast cancer treatment.

Methods

Cell culture

The murine breast cancer cell line 4T1, and the human breast cancer cell line MCF-7 were purchased from Xiamen Immocell Biotechnology Co.,Ltd, and were cultured in RPMI 1640 medium containing 10% FBS and 1% penicillin/streptomycin. The PD-1⁺ macrophages based on immortalized murine bone marrow-derived macrophages (iBMDMs) were custom established by Cyagen Biosciences. In brief, PD-1 gene were transfected into iBMDMs via lentivirus. The cells were cultured in DMEM medium supplemented with 10% FBS and 1% penicillin/streptomycin. All the cells were maintained at 37 °C in a humidified incubator with 5% CO₂.

Synthesis of AuNPs nanoparticles

The synthesis of spherical AuNPs was performed through a sequential CTAB-assisted method. Initially, 250 μL of 0.01 M HAuCl4 aqueous solution was added to 10 mL of 0.1 M CTAB solution at 25–30 °C, followed by rapid injection of 600 μL ice-cold 0.01 M NaBH4 to produce a brownish-yellow seed solution that was aged for 2–5 h. For nanoparticle growth, 2.2 mL of 0.01 M HAuCl4 was combined with 50 mL DI water and 1.5 g CTAB at 27 °C, then 6.5 mL aliquots were treated with 70 μL fresh 0.1 M ascorbic acid before adding 1.3 mL seed solution, with vigorous mixing for 10 min yielding a wine-red dispersion [28].

Peroxidase-like activity of AuNPs

The peroxidase-like activity of AuNPs was assessed by incubating a mixture of TMB (0.8 mM, as substrate), H₂O₂ (1 mM), and AuNPs (100 μg mL⁻¹) in deionized water, and the US group was treated under ultrasound irradiation (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min). The development of a blue color indicated TMB oxidation, and the absorbance was quantified using a microplate reader.

Preparation of PD-1-Au@DOX

The PD-1⁺ macrophages membrane was isolated as previously described with modification [29]. In brief, the PD-1⁺ macrophages were adjusted to 2 × 10⁶ cells, which were then lysed in ice-cold tris-magnesium buffer (TM-buffer, pH 7.4, containing 0.01 M Tris and 0.001 M MgCl₂) for 1 h. The lysate was first centrifuged at 2,000 g for 10 min at 4 °C to remove cellular debris, followed by a second centrifugation at 18,000 g for 15 min at 4 °C to pellet the cell membrane fraction. The obtained membranes were homogenized by ultrasonication for 15 min in an ice-water bath using a cell disruptor.

For nanoparticle synthesis, 5 mg of AuNPs and 5 mg of DOX were co-incubated with 5 mL of a cell membrane suspension derived from 2 × 10⁸ cells. The mixture was vortexed for 1 min followed by ultrasonication in an ice-water bath for 20 min to promote efficient membrane coating. The resulting nanoparticles were harvested by centrifugation at 15,000 × g for 15 min at 4 °C, resuspended in buffer, and briefly sonicated (3 min in an ice-water bath) to ensure homogeneity. To achieve uniform particle size, the suspension was sequentially extruded through polycarbonate membranes with pore sizes of 400 nm, 200 nm, and 100 nm, yielding monodisperse nanoparticles with a hydrodynamic diameter of approximately 100 nm. The final product was isolated by centrifugation at 18,000 × g, and the AuNPs content was quantified spectrophotometrically by measuring absorbance at 530 nm. The DOX content was quantified spectrophotometrically by measuring absorbance at 485 nm.

And the encapsulation efficiency (EE, %) was calculated as:

Characterization of PD-1-Au@DOX

The hydrodynamic diameter of the synthesized AuNPs and PD-1-Au@DOX were determined by dynamic light scattering (DLS) using a Malvern Zetasizer HS III instrument (Malvern Panalytical, UK). Prior to measurement, AuNP samples were diluted in ultrapure water to an appropriate concentration to avoid multiple scattering effects, ensuring an attenuator index within the optimal range recommended by the manufacturer. Each sample was subjected to three independent runs, with each run comprising 10–15 sub-runs automatically optimized by the instrument software. The morphology of PD-1-Au@DOX was observed by TEM (HT7700, Hitachi, Japan). The expression of PD-1 on Au@DOX and PD-1-Au@DOX were determined with PD-1 Elisa kit (Ruixin biotech).

In vitro drug release of DOX and AuNPs

The release of DOX from PD-1-Au@DOX were conducted in PBS (pH 7.4 and pH 5.5) at 37 ℃. PD-1-Au@DOX was dissolved in 10 mL PBS at 37 ℃ and shaken at 100 rpm. The samples were collected at different time points (0.5, 1, 4, 12, 24, 48 and 72 h), and the samples were centrifugated at 15,000 g for 30 min. To evaluate the ultrasound-triggered release profile, PD-1-Au@DOX was dissolved in 10 mL PBS at 37 ℃, exposed to ultrasound irradiation (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min) and shaken at 100 rpm for 30 min. The samples were subjected to centrifugation at 15,000 g for 30 min. The accumulate release of DOX in supernatant was calculated by absorbance measurement at 490 nm from the established standard curve using TECAN Infinite F200 Microplate Reader.

The release of AuNPs from PD-1-Au@DOX was determined. Briefly, PD-1-Au@DOX nanoparticles were exposed to ultrasound (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min), followed by centrifugation of the suspension at 15,000 × g for 30 min. The release of AuNPs in supernatant was calculated by absorbance measurement at 530 nm from the established standard curve using TECAN Infinite F200 Microplate Reader.

In vitro cytotoxicity study

The cytotoxicity of Au@DOX and PD-1-Au@DOX were evaluated using MTT assays. 4T1 and MCF-7 breast cancer cells were seeded in 96-well plates at 8 × 10³ cells/well in RPMI 1640 medium supplemented with 10% FBS and incubated at 37 °C with 5% CO₂ for 24 h. Cells were then treated with free DOX, Au@DOX and PD-1-Au@DOX (2 μg/mL), then cells treated with/without ultrasound irradiation (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min), and with untreated cells as control. After 24 h treatment, MTT solution (1 mg/mL, 50 μL) was added and incubated for 4 h. The resulting formazan crystals were dissolved in DMSO (150 μL/well), and absorbance at 570 nm was measured using TECAN Infinite F200 Microplate Reader.

In vitro cellular uptake

To measure the cellular uptake efficiency of 4T1 and MCF-7 cells, cells were seeded into 6-well plates at a density of 2 × 10⁵ cells/well and incubated with Au@DOX and PD-1-Au@DOX (2 μg/mL) and for 4 h. To investigate the functional role of PD-1 in the PD-1-Au@DOX, tumor cells were pretreated with an anti-PD-L1 antibody (10 μg/mL) for 1 h at 37 °C to block surface PD-L1. Subsequently, PD-1-Au@DOX was incubated with the antibody-blocked tumor cells for 4 h at 37 °C. Post-incubation, cellular interactions and internalization were quantitatively analyzed using flow cytometry and confocal laser scanning microscopy.

Intracellular GSH detection

4T1 and MCF-7 cells were incubated in 6-well plates at 37 ℃ with 5% CO₂. Afterward, they were treated with Au@DOX, PD-1-Au@DOX (2 μg/mL) followed by ultrasound irradiation (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min), and co-incubation for 0.5, 1, 2, 4, 12 h, post-treatment, the culture medium was taken out, and cells were washed with PBS for three times. The intracellular GSH was determined with GSH ELISA kit (COIBO BIO, China). The absorption was determined utilizing TECAN Infinite F200 Microplate Reader at 450 nm.

Intracellular ROS detection

The intracellular ROS levels were quantified using 2′,7′-dichlorofluorescin diacetate (DCFH-DA) kit. Briefly, 4T1 and MCF-7 breast cancer cells were seeded in 12-well plates at a density of 5 × 10⁴ cells/well and cultured for 12 h (37 °C, 5% CO₂). Cells were then treated with DOX, Au@DOX, PD-1-Au@DOX nanoparticles (2 μg/mL) for 4 h, followed by ultrasound irradiation (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min). DOX treated alone was also determined. After an additional 2 h incubation, dual staining was performed: nuclei were labeled with DAPI, while ROS was detected with DCFH-DA (10 μM, 20 min). Fluorescence signals were captured using a confocal microscope (Leica TCS SP8) and flow cytometry.

Blocking the PD-L1 of tumor cells.

To evaluate the PD-L1 blocking capacity of PD-1-Au@DOX, 4T1 and MCF-7 breast cancer cells were seeded in 6-well plates at 1 × 10⁵ cells/well and cultured for 24 h, followed by 1-h incubation with Au@DOX, PD-1-Au@DOX, or non-fluorescent anti-PD-L1 antibody. After PBS washing, cells were stained with PE-conjugated anti-PD-L1 antibody for 1 h, and PD-L1 expression levels were subsequently analyzed by flow cytometry and confocal microscopy.

Cell apoptosis analysis

To evaluate the combinatorial cytotoxicity of PD-1-Au@DOX against breast cancer cells, 4T1 and MCF-7 cells were seeded in 6-well plates (1 × 10⁵ cells/well) and cultured in complete RPMI-1640 medium at 37 °C with 5% CO₂. Cells were divided into four treatment groups: (1) Au@DOX (2 μg/mL), (2) PD-1-Au@DOX (2 μg/mL), (3) anti-PD-L1 antibody (10 μg/mL, positive control), and (4) untreated control. The cells were incubated at 37 °C with 5% CO₂ for 6 h. Then the US irradiation group were subjected to ultrasound irradiation (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min) using a therapeutic ultrasound system. After 18 h incubation, purified CD8⁺ T cells (isolated from C57BL/6 mice) were added at a 1:1 effector-to-target ratio. Following an additional 48 h co-culture, apoptotic activity was quantified through intracellular staining of cleaved caspase-3 (Cell Signaling Technology, #9661) and flow cytometric analysis.

In vivo ultrasound irradiation protocol

For in vivo sonication, the ultrasound probe was placed in direct contact with the skin overlying the tumor. A uniform layer of ultrasound gel (Dandelion Clear Ultrasound Gel) was applied as the coupling medium to ensure efficient acoustic transmission, maintaining a consistent probe-to-skin distance of 0–2 mm. During sonication (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min), real-time temperature at the tumor site was monitored.

Anti-tumor efficacy of In situ breast tumor models

All animal experiments were conducted following the approved animal protocol procedures by the Institutional Animal Care and Use Committee (IACUC) of Fujian Medical University. To establish a breast cancer model, female BALB/c mice (n = 5 per group) were subcutaneously inoculated with 1 × 10⁶ 4T1 cells suspended in 100 μL PBS. When the tumors reached an average volume of approximately 100 mm³, the mice were randomized into 7 treatment groups, and i.v. injected with: 1. Control (PBS, 100 μL); 2. Control + US (PBS, 100 μL); 3. Au@DOX (10 mg/kg); 4. Au@DOX + US (10 mg/kg); 5. PD-1-Au@DOX (10 mg/kg); 6.PD-1-Au@DOX + US (10 mg/kg); 7. Free DOX (10 mg/kg) every 4 days for 5 times. And US irradiation was performed 6 h later (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min). Body weight was recorded simultaneously with tumor volume measurements. Mice were euthanized when tumors exceeded 1,500 mm³ or at the day 21. The tumors were analysized with H&E staining. TUNEL expression was detected via immunofluorescence assay, while immunohistochemistry was employed to evaluate Ki67, GPX4 expression. Finally, H&E staining was performed on critical organs (heart, liver, spleen, lung, and kidney) to assess systemic toxicity. TheTNF-α, IFN-γ, IL-6, and IL-12 (ELISA, Invitrogen) of the tumors were respectively measured using an ELISA kit. To evaluate survival rates, animals were observed daily until they reached either spontaneous death or a moribund state.

In vivo anti-metastasis efficacy.

The orthotopic breast cancer model was established by injecting 1 × 10⁶ 4T1 cells into the mammary fat pad of BALB/c mice. One week post-inoculation, mice were randomly allocated into four experimental groups (n = 4 per group): (1) PBS (100 μL), (2) PD-1-Au@DOX (10 mg/kg), (3) PD-1-Au@DOX + ultrasound (US), and (4) free DOX (10 mg/kg). The PD-1-Au@DOX + US group received intravenous administration of PD-1-Au@DOX (10 mg/kg) on days 0, 4, 8, 12 and 16, followed by US irradiation (1.0 MHz, 0.5 W/cm², 20% duty cycle, 5 min) 6 h post-injection. The treatment regimen was conducted over a 21-day period. At the experimental endpoint, mice were euthanized, and lung tissues were collected for subsequent analysis. Pulmonary metastatic nodules were quantified, and their presence was histologically confirmed through H&E staining.

Statistical analyses

Statistical analyses were performed using GraphPad Prism 8.0.2. Comparisons between two groups were evaluated using two-sided Student’s t-tests, while comparisons involving more than two groups were assessed via one-way or two-way analysis of variance (ANOVA). The survival curves were created using log-rank (Mantel–Cox) tests. A p-value < 0.05 was considered statistically significant (Fig. 1).

Fig. 1.

A schematic diagram illustrates the synergistic mechanism of ‌PD-1-Au@DOX‌ in breast cancer therapy, assisted by ultrasound. AuNP induce ‌ferroptosis‌ in tumor cells, while DOX‌ delivers chemotherapy. The ‌PD-1‌ component on PD-1-Au@DOX further enhances cytotoxicity by blocking the ‌PD-1/PD-L1‌, thereby amplifying ‌T cell-mediated immune responses‌ and reducing lung metastasis through a multi-modal therapeutic approach

Results and discussion

Synthesis and characterization of PD-1-Au@DOX

‌The use of natural cell membranes as nanoparticle coatings provides a strategy to simultaneously achieve low immunogenicity, high biocompatibility, and prolonged circulation in vivo [30–33]. Moreover, nanoparticles coated with PD-1-expressing cell membranes can specifically target and block tumor cell surface PD-L1 [34]. To synthesize PD-1⁺ membrane-coated nanoparticles, cell membrane derived from engineered murine PD-1-positive macrophages were coated with DOX and ultrasound-responsive AuNPs. Subsequent extrusion, performed as we previously described [35], yielded uniformly sized nanoparticles within the desired range (Fig. 2A). The size distribution of AuNPs and PD-1-Au@DOX were determined with DLS, with hydrodynamic sizes of ~ 2.3 nm and ~ 91.3 nm (Fig. 2B-C). Transmission electron microscopy (TEM) demonstrated that both AuNPs and PD-1-Au@DOX exhibited a spherical morphology with high homogeneity (Fig. 2D-E), which is consistent with previous literature [36]. The TEM-measured diameter of PD-1 Au@DOX (75.53 ± 4.98 nm) was smaller than the DLS-derived hydrodynamic diameter, as DLS includes the solvation layer and thus typically overestimates particle size compared to TEM.

Fig. 2.

Synthesis and characterization of ‌PD-1-Au@DOX. (A), Schematic illustration of the synthesis of PD-1-Au@DOX. B-C, The size distribution of AuNPs (B) and PD-1-Au@DOX (C) were measured with DLS. D-E, The TEM images of AuNP (D) and PD-1-Au@DOX (E). (F), The cumulative release of DOX from PD-1-Au@DOX was measured at pH 5.5 and pH 7.4 over different time points. (G), The cumulative release amount of DOX from PD-1-Au@DOX under ultrasonic and non-ultrasonic conditions at the 20-min time point under pH 5.5 and pH 7.4 conditions. (H), The UV–vis absorption spectra of AuNP. I, The expression of PD-1 on PD-1-Au@DOX and control (Au@DOX). (J), Schematic diagram to illustrate ROS generation under US. K, Catalytic decomposition of H2O2 by AuNP with/without US. The data are shown as the mean ± SD, as calculated using two-sided Student’s t-tests (G, I). Significant differences are indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. ns, not significant.‌

The cell membrane structure enables a drug encapsulation efficiency of 23.28 ± 3.40% for DOX and encapsulation efficiency of 20.06 ± 0.14% for AuNPs. The ultrasound-triggered release of AuNPs from PD-1-Au@DOX was also evaluated. The results showed a released AuNPs concentration of 0.11 ± 0.01 mg/mL, corresponding to a release efficiency of 55% relative to the initial loading capacity of 0.20 ± 0.01 mg/mL.

Given the acidic tumor microenvironment, we assessed the release profile of PD-1-Au@DOX at pH 5.5 and pH 7.4. The results demonstrate that under acidic conditions (pH 5.5), PD-1-Au@DOX exhibits rapid drug release, achieving a cumulative release of 86.57% after 72 h, compared to only 68.77% at pH 7.4 (Fig. 2F). This pH-responsive release behavior endows PD-1-Au@DOX with tumor-specific drug delivery capability. Furthermore, we investigated the effect of ultrasound on the release of PD-1-Au@DOX. Under acidic conditions (pH 5.5), ultrasound irradiation significantly enhanced drug release, with 18.34% of DOX released within 20 min compared to 16.83% in the absence of ultrasound (Fig. 2G). This demonstrates that ultrasound can effectively promote the drug release from PD-1-Au@DOX.

The expressions of PD-1 on PD-1-Au@DOX and Au@DOX were determined with PD-1 Elisa Kit. These results demonstrate that PD-1-Au@DOX exhibits significantly higher expression of PD-1 compared to Au@DOX (Fig. 2I), which provides the possibility for subsequent targeted delivery of AuNPs and DOX by PD-1-Au@DOX, as well as blocking of PD-L1 on the surface of tumor cells.

Leveraging the unique properties of gold nanoparticles, we demonstrated their capacity to generate ROS under ultrasonic stimulation (Fig. 2J). The exploitation of elevated H₂O₂ levels in tumor microenvironments to drive its conversion into cytotoxic ROS constitutes a promising therapeutic strategy [37]. The capacity of AuNPs to generate ROS was assessed by monitoring the colorimetric oxidation of 3,3’,5,5’-tetramethylbenzidine (TMB). In the presence of H₂O₂, AuNPs catalyze the formation of ·OH, which oxidize colorless TMB to its blue-colored oxidized form. As shown in Fig. 2K, AuNPs demonstrated ROS generation both with and without ultrasound exposure. However, AuNPs exhibited significant ultrasound responsiveness, with substantially higher ROS production observed under ultrasound irradiation compared to the non-irradiated control group.

‌These results demonstrate that the PD-1-nanovesicles we engineered exhibit uniform morphology, optimal size, and significant surface PD-1 expression. Furthermore, they possess the capability to enhance ROS generation and DOX release upon ultrasound activation, collectively contributing to precise and efficient therapeutic efficacy at the tumor site.

In vitro cellular uptake and cytotoxicity

The cellular uptake efficiency of nanocarriers critically determines their antitumor efficacy. Our designed PD-1⁺ cell membrane-coated nanoparticles have excellent biocompatibility, while surface-exposed PD-1 enables specific binding to tumor cell surface PD-L1, thereby facilitating efficient targeted delivery (Fig. 3A). Therefore, we assessed the uptake capacity of these nanoparticles in 4T1 and MCF-7 cells. Flow cytometry analysis revealed significantly enhanced cellular uptake of PD-1-Au@DOX compared to Au@DOX in 4T1 cells (Fig. 3B-C), likely attributable to the PD-1-mediated specific targeting of surface PD-L1 on tumor cells‌. Notably, pre-incubation of 4T1 cells with anti-PD-L1 antibodies prior to exposure to PD-1-Au@DOX resulted in a marked reduction in nanoparticle uptake‌. This dose-dependent inhibition further corroborates that the enhanced uptake is driven by specific PD-1/PD-L1 molecular recognition‌. Consistent results were also observed in MCF-7 cells (Fig. 3D-E), further validating our findings.

Fig. 3.

PD-1-Au@DOX specifically targets tumor cells. (A), Schematic illustration of PD-1-Au@DOX achieving tumor cell-specific delivery via PD-1 targeting of PD-L1. (B-E), Flow cytometric analysis and the quantification of Au@DOX, PD-1-Au@DOX, PD-1-Au@DOX (tumor cells were pre-blocked with PD-L1 antibody) and free DOX uptake by 4T1(B-C) and MCF-7 (D-E) cells after co-incubation for 4 h. (F-I), Confocal microscopy ‌shows‌ the intracellular uptake of Au@DOX, PD-1-Au@DOX, PD-1-Au@DOX (tumor cells were pre-blocked with PD-L1 antibody) and free DOX by 4T1(F-G) and MCF-7 (H-I) cells after co-incubation for 4 h. The data are shown as the mean ± SD, as calculated using one-way ANOVA (C, E, G, I). Significant differences are indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. ns, not significant

To further validate these findings, we employed confocal laser scanning microscopy to quantitatively assess cellular uptake of Au@DOX, PD-1-Au@DOX, and free DOX in both 4T1 and MCF-7 cell lines. Additionally, we systematically evaluated how anti-PD-L1 blockade of surface PD-L1 on these cells affected PD-1-Au@DOX uptake (Fig. 3F-I). The results demonstrate significantly enhanced cellular accumulation of PD-1-Au@DOX compared to Au@DOX controls. Importantly, this increased uptake was substantially diminished following PD-L1 blockade in both cell types, providing compelling evidence that surface-exposed PD-1 enables specific recognition of tumor cells through PD-1/PD-L1 interaction and facilitates precise targeted delivery.

Collectively, these findings demonstrate that PD-1-Au@DOX achieves targeted delivery to PD-L1-expressing tumor cells through PD-1-mediated recognition, enabling tumor-specific drug accumulation while minimizing uptake in normal cells. This selective delivery mechanism effectively reduces off-target toxicity toward healthy tissues.

We evaluated the in vitro cytotoxicity of Au@DOX, PD-1-Au@DOX, and free DOX. The results demonstrated that Au@DOX, PD-1-Au@DOX, and Free DOX all exhibited concentration-dependent cytotoxicity against 4T1 cells. However, ultrasound irradiation significantly enhanced the cytotoxicity of PD-1-Au@DOX while showing minimal enhancement for Au@DOX and Free DOX (SFig. 1A). This differential effect stems from PD-1-Au@DOX’s targeted delivery to 4T1 cells via PD-1/PD-L1 interaction, which promotes greater AuNPs uptake and consequently improves US-triggered ROS generation. In contrast, Au@DOX showed limited US-triggered ROS generation while Free DOX demonstrated no US effect. Similar trends were observed in MCF-7 cells (SFig. 1B), confirming the generalizability of this mechanism. ‌These results demonstrate that the PD-1 displayed on the surface of PD-1-Au@DOX enables targeted drug delivery by binding to PD-L1 on breast cancer cells, thereby reducing toxicity toward normal cells.

In vitro ROS production and GSH depletion

Our results demonstrates that AuNPs exhibit efficient ROS generation under ultrasound irradiation (Fig. 2J). Nevertheless, the robust cellular antioxidant defense mechanisms, particularly those involving glutathione (GSH), substantially limit this strategy through efficient ROS neutralization [38]. Therefore, depletion of GSH in malignant cells can potentiate ROS accumulation, thereby augmenting the efficacy of ROS-inducing therapeutic interventions. To assess this, we quantified the capacity of PD-1-Au@DOX to generate ROS and deplete GSH in 4T1 and MCF-7 cells under ultrasound stimulation (Fig. 4A). As shown in Fig. 4B-E, both Au@DOX and PD-1-Au@DOX exhibited time-dependent GSH depletion in 4T1 cells, with PD-1-Au@DOX demonstrating significantly higher GSH degradation capacity than Au@DOX. Notably, ultrasound irradiation further enhanced the GSH-depleting effect of PD-1-Au@DOX (Fig. 4B-C). Similar results were observed in MCF-7 cells (Fig. 4D-E). These findings indicate that PD-1-Au@DOX possesses ultrasound-responsive properties, and its superior GSH degradation compared to Au@DOX may stem from its enhanced cellular uptake due to stronger targeted delivery to both 4T1 and MCF-7 cells, ultimately leading to more efficient GSH depletion.

Fig. 4.

In vitro ROS generation. (A), ‌Schematic illustration of ROS generation and GSH depletion in cells induced by PD-1-Au@DOX under ultrasound irradiation. (B-D), The intracellular GSH level of 4T1 (B-C) and MCF-7 (D-E) cells after treated with Au@DOX or PD-1-Au@DOX (with/without ultrasound) at different time points. (F–H), Confocal microscopy ‌shows‌ the intracellular ROS levels in 4T1 cells (F) and their quantification (G), as well as MCF-7 cells (F) and their quantification (H) (scale bar = 50 μm). (I-L), Flow cytometric ‌shows‌ the intracellular ROS levels in 4T1 cells (I) and their quantification (J), as well as MCF-7 cells (K) and their quantification (L). The data are shown as the mean ± SD, as calculated using one-way ANOVA (G, H, J, L). Significant differences are indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. ns, not significant

To validate the ultrasound-induced ROS generation by PD-1-Au@DOX, we employed DCFH as a fluorescent probe for ROS detection via confocal microscopy and flow cytometric analysis. The results demonstrated that PD-1-Au@DOX generated significantly higher levels of ROS compared to Au@DOX in 4T1 cells under ultrasound irradiation (Fig. 4F-G). A similar trend was observed in MCF-7 cells (Fig. 4F, 4H). To further validate the ultrasound-triggered ROS generation capability of PD-1-Au@DOX, we quantitatively assessed this using flow cytometry. The results revealed that upon ultrasound irradiation, PD-1-Au@DOX significantly enhanced ROS production in both 4T1 and MCF-7 cells compared to all control groups (Fig. 4I-L). This enhanced ROS production may be attributed to the targeted delivery of PD-1-Au@DOX, leading to increased cellular uptake by tumor cells and consequently improved ROS generation efficiency. These findings are consistent with our previous observations (Fig. 3). Interestingly, DOX, used here as a positive control, also induced a detectable ROS response. This may be attributed to DOX’s ability to trigger ferroptosis, which is consistent with previous reports [39]. Notably, the DOX-only group (without DCFH-DA) exhibited weak green fluorescence, likely due to partial spectral overlap between DOX’s emission profile and the detection window of the DCFH-DA probe. However, this background fluorescence from DOX was minimal and did not interfere with the overall trend of DCFH-DA–derived fluorescence signals from cells treated with DOX (SFig. 2). Although the DCFH-DA assay confirmed a significant increase in intracellular ROS levels following ultrasound irradiation, DCFH-DA is broad spectrum probe. Therefore, the exact identity of the predominant ROS generated in the PD-1-Au@DOX/US system remains to be determined. Future studies employing selective chemical probes or scavenger-based assays will be necessary to delineate the specific ROS species responsible for the observed therapeutic effects.

These findings demonstrate that PD-1-Au@DOX exhibits robust ultrasound responsiveness, generating substantially higher ROS levels than the ultrasound-alone group, Au@DOX group, Au@DOX + US group, or PD-1-Au@DOX group without US across multiple breast cancer cell lines (4T1 and MCF-7 cells). Furthermore, PD-1-Au@DOX displayed pronounced GSH depletion under ultrasound irradiation in breast cancer cells, thereby exerting potent cytotoxic effects.

In vitro immunotherapy

PD-1/PD-L1 immune checkpoint inhibitors reactivate anti-tumor immunity by disrupting the PD-1/PD-L1 signaling axis between tumor cells and T cells, thereby alleviating immune suppression and restoring cytotoxic T-cell function [40]. To this end, we employed PD-1-Au@DOX to competitively block tumor cell surface PD-L1 via its surface-bound PD-1 moieties, thereby reactivating T-cell-mediated anti-tumor immunity (Fig. 5A). Following co-incubation of 4T1 cells with Au@DOX, PD-1-Au@DOX, or anti-PD-L1 antibody, flow cytometric analysis revealed that surface PD-L1 expression on 4T1 cells was significantly downregulated after treatment with PD-1-Au@DOX or anti-PD-L1 antibody, whereas no significant reduction was observed in the Au@DOX group (Fig. 5B-C). Consistent findings were replicated in MCF-7 cells (Fig. 5D-E).

Fig. 5.

In vitro immune checkpoint inhibition by PD-1-Au@DOX. (A), ‌Schematic illustration of T cell-mediated immunotherapy induced by PD-1-Au@DOX. (B-E), Flow cytometry shows the PD-L1 expression in 4T1 cells (B) and the quantification (C), as well as MCF-7 cells (D) and the quantification (E). (F–H), Confocal microscopy ‌shows‌ the PD-L1 expression in 4T1 cells (F) and their quantification (G), as well as MCF-7 cells (F) and their quantification (H) (scale bar = 50 μm). I-J, 4T1 (l), MCF-7 (J) cells were co-cultured with Au@DOX or PD-1-Au@DOX (with/without US) for 24 h, followed by 48 h co-culture with CD8⁺ T cells. Anti–PD-L1 antibody served as a positive control. Tumor cell apoptosis was assessed via flow cytometry for intracellular cleaved caspase-3 (left), with quantification (right). The data are shown as the mean ± SD, as calculated using one-way ANOVA (C, E, G, H, I, J). Significant differences are indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. ns, not significant

Consistent with our previous findings, confocal microscopy analysis confirmed that surface PD-L1 expression was significantly downregulated on both 4T1 and MCF-7 cells after treatment with PD-1-Au@DOX or anti-PD-L1 antibody, whereas no significant reduction was observed in cells treated with Au@DOX alone (Fig. 5F-H). This effect stems from the ability of surface-bound PD-1 on PD-1-Au@DOX to competitively engage and block tumor cell surface PD-L1, mirroring the mechanism of anti-PD-L1 antibodies.

To assess the cytotoxic capacity of effector T cells against tumor cells treated with PD-1-Au@DOX, we co-cultured 4T1 cells pre-incubated with Au@DOX, PD-1-Au@DOX, or anti-PD-L1 antibody (with or without ultrasound treatment) with effector T cells for 48 h. T-cell-mediated cytotoxicity was subsequently quantified by flow cytometry. As shown in Fig. 5I, ultrasound treatment alone (C + US) did not induce significant apoptosis in 4T1 cells (~ 71 vs ~ 81). Treatment with CD8⁺ T cells alone (C + CD8) resulted in a modest increase in tumor cell apoptosis compared to the control (~ 140 vs ~ 81). Meanwhile, incubation with PD-1-Au@DOX alone showed a relatively higher level of apoptosis (~ 486 vs ~ 216), likely due to the precise delivery capability of PD-1-Au@DOX, which facilitated greater uptake by 4T1 cells and induced a concentration-dependent chemotherapeutic effect. Following ultrasound irradiation, the apoptosis in the PD-1-Au@DOX + US group was notably higher than that in the Au@DOX + US group (~ 1422 vs ~ 354), attributable to increased cellular internalization of DOX and AuNPs in the PD-1-Au@DOX group, combined with ROS generation induced by ultrasound that synergistically enhanced the chemotherapeutic efficacy. When pre-incubated with Au@DOX, PD-1-Au@DOX, or anti-PD-L1 antibody, followed by co-culture with CD8⁺ T cells, the PD-1-Au@DOX + CD8 group exhibited the most pronounced apoptosis (~ 1873), likely due to dual mechanisms: PD-1-Au@DOX delivered DOX for chemotherapy, and its surface PD-1 blocked PD-L1 on 4T1 cells, thereby enhancing CD8⁺ T cell-mediated cytotoxicity. Furthermore, in the combined treatment involving pre-incubation of 4T1 cells with Au@DOX or PD-1-Au@DOX, US irradiation, and subsequent co-culture with CD8⁺ T cells, the PD-1-Au@DOX group achieved the highest apoptosis level (~ 3326), attributable to targeted DOX delivery, US-triggered ROS production, and PD-L1 blockade on 4T1 cells, which collectively elicited a strong cytotoxic response from CD8⁺ T cells. Similar results were observed in MCF-7 cells (Fig. 5J), confirming the robustness of this multimodal therapeutic approach.

Collectively, these data establish that PD-1-Au@DOX mediates potent breast cancer therapy through a triple-pronged mechanism: (1) PD-1-guided targeted chemotherapy via DOX delivery, (2) immune checkpoint blockade through competitive PD-L1 engagement restoring T-cell cytotoxicity, and (3) ultrasound-triggered ROS generation inducing ferroptosis-like death. The confluence of these therapeutic modalities culminates in synergistic tumor eradication, as validated in both 4T1 and MCF-7 models.

In vivo antitumor effect

Given the significant US-induced ROS generation, PD-1-mediated targeted delivery of DOX, and PD-1/PD-L1 immune checkpoint inhibition demonstrated by PD-1-Au@DOX at the cellular level—along with its unique therapeutic efficacy in vitro—we proceeded to evaluate its in vivo anticancer performance in a breast cancer-bearing animal model. Timing chart of the treatment procedures for animal experiments is illustrated in the Fig. 6A.

Fig. 6.

Tumor inhibition in vivo. (A), Timing chart of the treatment procedures for animal experiments. B, Mouse body weight during the treatment. (C), Tumor growth curve by dynamically recording the tumor volume during the therapeutic course. (D), Tumor volume change in terms of different groups. (E), Tumor weight in terms of different groups on day 21. F, Survival rate in different treatment groups. (G-J), H&E (G), Ki67 (H), GPX4 (I) immunohistochemistry staining, and TUNEL (J) immunofluorescence of excised tumors after treatment ended. Data are presented as mean ± s.d. and analyzed using one-way ANOVA (E) or log-rank (Mantel–Cox) tests (F). *, P < 0.05; **, P < 0.01; ***, P < 0.001, and ****, P < 0.0001. ns, not significant

Treatment with PBS (as negative control), US alone, Au@DOX (with/without US), PD-1-Au@DOX (with/without US), Free DOX (as postive control) did not induce significant body weight reduction in the animals (Fig. 6B). Tumor volume and tumor weight were monitored as the primary endpoint to assess the sustained therapeutic efficacy of PD-1-Au@DOX-enhanced US-triggered multimodal therapy over 21 days. The results showed that Control, Control + US, Au@DOX, Au@DOX + US, and PD-1-Au@DOX groups showed rapid tumor progression, while the combination of PD-1-Au@DOX with US irradiation achieved significant tumor suppression (Fig. 6C-E), the tumor growth inhibition reach 83.14% (217.40 vs 1288.89 mm³) demonstrating clear therapeutic benefits.

In addition, survival analysis revealed complete mortality in both C and C + US groups by day 28. While Au@DOX, Au@DOX + US, and PD-1-Au@DOX groups showed modest survival extension compared to controls, the positive control (Free DOX) achieved prolonged survival up to day 55. Notably, 4 mice in the PD-1-Au@DOX + US group remained alive at this timepoint, demonstrating statistically significant survival enhancement in breast cancer-bearing mice (Fig. 6F). Furthermore, H&E staining of tumor tissue sections was conducted to assess the extent of cellular damage and necrosis (Fig. 6G). Ki67 immunohistochemistry was employed to analyze tumor progression across groups. As anticipated, the PD-1-Au@DOX + US group exhibited the lowest Ki67 expression, indicating minimal tumor proliferation (Fig. 6H). Notably, GPX4 expression was significantly downregulated in this group (Fig. 6I), confirming the induction of intratumoral ferroptosis. Ferroptosis plays a critical role in cancer therapy, and notably, reduced GPX4 expression potently enhances ferroptotic cell death, thereby eliciting a more robust antitumor response within the tumor microenvironment [39]. Immunofluorescence analysis revealed markedly enhanced TUNEL expression in the PD-1-Au@DOX + US cohort compared to other groups (Fig. 6J), demonstrating its superior antitumor efficacy. These findings demonstrate that the PD-1-Au@DOX + US regimen synergistically combines three therapeutic mechanisms: (1) US-induced ferroptosis, (2) PD-1-mediated targeted DOX delivery, and (3) PD-1/PD-L1 pathway-mediated immune checkpoint blockade. This multimodal approach enables significant tumor growth inhibition and prolonged survival in breast cancer-bearing mice.

Furthermore, to assess intratumoral immune infiltration, we quantified CD8⁺ T cell populations within tumors using confocal microscopy post-treatment. As shown in SFig. 3A-B, PD-1-Au@DOX + US-treated tumors exhibited significantly higher green fluorescence intensity compared to other groups, indicating substantial CD8⁺ T cell accumulation. Furthermore, pro-inflammatory cytokine levels (TNF-α, IFN-γ, IL-6, and IL-12) were markedly elevated in the PD-1-Au@DOX + US group (SFig. 2C-F). This enhanced immune infiltration likely results from dual mechanisms: (1) PD-L1 blockade on tumor cells by PD-1-Au@DOX, and (2) ultrasound-mediated immune activation. Extensive plasma membrane disruption robustly triggers immunogenic oncolytic pyroptosis, thereby enhancing tumor immunogenicity while concurrently remodeling the immunosuppressive TME. This coordinated effect culminates in potent and long-lasting systemic antitumor immune responses [41]. Both of which are consistent with prior reports in the literature [42].

Lung metastasis inhibition

Cancer metastasis remains the leading cause of cancer-related mortality and is a major contributor to therapeutic failure across diverse treatment modalities [43]. Consistent evidence indicates that the lungs serve as the predominant site for distant metastasis in breast cancer [44]. ‌Given that both ROS generation and PD-1/PD-L1 pathway activation have been reported to suppress tumor metastasis—coupled with our system’s ability to induce ferroptosis via PD-1-Au@DOX + US while concurrently blocking the PD-1/PD-L1 axis^19,41—we sought to determine whether this combinatorial strategy could effectively inhibit breast cancer pulmonary metastasis.

The treatment schedule is depicted in Fig. 7A. Quantitative analysis demonstrated a significant reduction in pulmonary metastatic nodules in PD-1-Au@DOX + US-treated mice compared to controls (70.0%, Fig. 7B-C). While Free DOX exhibited comparable efficacy in primary tumor growth inhibition (72.6%, Fig. 6C-D), its anti-metastatic effect was substantially inferior to that of PD-1-Au@DOX + US (22.7% vs 70.0%). This distinction likely arises from the fundamentally different mechanisms of action: Free DOX primarily exerts its effects through conventional chemotherapy-mediated growth suppression of primary tumors, whereas PD-1-Au@DOX + US exerts synergistic antitumor activity through tri-modal mechanisms—(1) targeted delivery of DOX for chemotherapy, (2) ultrasound-induced ROS generation triggering ferroptosis, and (3) PD-1/PD-L1 blockade-mediated immune activation—this approach not only enhances tumor immunogenicity but also concurrently remodels the immunosuppressive TME, thereby achieving simultaneous breast cancer suppression and metastasis inhibition. In addition, HE staining revealed prominent lung nodules in the control, PD-1-Au@DOX, and Free DOX groups, suggesting that PD-1-Au@DOX + US treatment exhibited best efficacy in suppressing tumor metastasis (Fig. 7D).

Fig. 7.

Metastasis inhibition effect of PD-1-Au@DOX. A, Experimental scheme for PD-1-Au@DOX-induced in vivo anti-metastasis treatment. B, Number of tumor nodules in different groups. C, Representative photos of resected lung tissue in different treatment groups (n = 4). D, Images of H&E-stained lung tissues treated with the control, PD-1-Au@DOX, PD-1-Au@DOX + US groups and Free DOX. Data are presented as mean ± s.d. and analyzed using one-way ANOVA (B). *, P < 0.05; **, P < 0.01; ***, P < 0.001, and ****, P < 0.0001. ns, not significant

Conclusion

In summary, we have developed an innovative PD-1⁺ cell membrane-coated nanoplatform (PD-1-Au@DOX + US) that synergistically combines US-responsive AuNPs with DOX -targeted-delivery and PD-L1 blocking. This system operates through tri-modal mechanisms: (1) PD-1-guided targeted chemotherapy via DOX delivery and US-triggered controlled release, (2) the generation of ROS by AuNPs under sonodynamic action, combined with ROS enhancement by DOX, cooperatively induces ferroptosis, and (3) localized immune checkpoint blockade via membrane-displayed PD-1 acting as a competitive decoy for tumor PD-L1, thereby reinvigorating CD8⁺ T cell function within the tumor microenvironment. Unlike conventional combination approaches that administer chemotherapeutics, SDT agents, and immune modulators separately, our platform ensures synchronized action of all components at the tumor site. Characterization confirmed that PD-1-Au@DOX possesses an appropriate nanoscale size and displays PD-1 on its surface. Ultrasound irradiation effectively promotes the release of DOX from PD-1-Au@DOX. In vitro cellular experiments demonstrated that PD-1-Au@DOX achieves targeted delivery to 4T1 and MCF-7 cells via PD-1–mediated recognition of PD-L1 expressed on the cell surface. Under ultrasound exposure, PD-1-Au@DOX significantly depletes intracellular GSH levels and enhances ferroptosis (co-induced by DOX and ROS) in both 4T1 and MCF-7 cells. Furthermore, PD-1-Au@DOX effectively blocks PD-L1 on the surface of these cancer cells, thereby enabling a synergistic triple-combination therapy—comprising sonodynamic therapy, chemotherapy, and immunotherapy—to induce apoptosis in 4T1 and MCF-7 cells in vitro. In murine models of aggressive breast cancer, PD-1-Au@DOX combined with ultrasound irradiation achieved potent suppression of primary tumor growth, significantly prolonged survival, and markedly reduced lung metastasis. Critically, the synergy between ferroptotic cell death and restored T cell cytotoxicity created a pro-inflammatory tumor milieu that contributed to both local control and systemic antitumor immunity. This work advances the design paradigm of biomimetic nanomedicine by reprogramming the cell membrane from a passive stealth coating into an active immunomodulatory interface, while simultaneously leveraging chemotherapy and sonodynamic therapy to cooperatively induce ferroptosis and enhance antitumor efficacy. The tri-modal integration of tumor-targeted drug delivery, ultrasound-triggered sonodynamic therapy, and localized PD-1/PD-L1 blockade positions PD-1-Au@DOX as a promising candidate for clinical translation in metastatic breast cancer. Nevertheless, we acknowledge certain limitations in the current study. First, while our data demonstrate significant suppression of lung metastasis in vivo, the underlying molecular and cellular mechanisms by which PD-1-Au@DOX + US inhibits tumor metastasis were not fully elucidated. Second, the study lacks control formulations such as physically mixed combinations of PD-1@DOX + free Au, PD-1-Au + free DOX, or non-integrated PD-1 + Au + DOX, which would further validate the necessity of structural integration within a single nanoparticle. These limitations have been noted herein, and more in-depth mechanistic investigations into antimetastatic pathways, alongside refined control groups, will be prioritized in our future work.

Abbreviations

DOX

Doxorubicin

GSH

Glutathione

AuNPs

Gold nanoparticles

ICIs

Immune checkpoint inhibitors

ROS

Reactive oxygen species

SDT

Sonodynamic therapy

TAA

Tumor-associated antigen

TME

Tumor microenvironment

Author contributions

Xiujuan Zhang: Conceptualization, Methodology, Editing, Funding acquisition, Supervision. Leilei Liu: Conceptualization, Editing, Investigation, Supervision. Wenjin Lin: Investigation, Data acquisition and analysis. Zhenhu Lin: Investigation, Data acquisition and analysis. Zhiyong Li: Investigation. Fen Fu: Data acquisition and analysis. Haidong Dong: Conceptualization, Methodology. Xiaodong Xie: Methodology, Writing, Editing, Supervision.

Funding

This research was supported by Joint Funds for the innovation of science and Technology, Fujian province (2024Y9307;2024Y9802;2023Y9202), and Fujian Provincial Department of Science and Technology Guiding Science and Technology Planning Project (2024Y0022).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

All animal experiments were conducted following the approved animal protocol procedures by the Institutional Animal Care and Use Committee (IACUC) of Fujian Medical University, which are consistent with the International Council for Laboratory Animal Science (ICLAS) guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.