Anti-Triggering Receptor Expressed on Myeloid Cells 2‑Conjugated Nanovesicles Loaded Vadimezan Reprogram Tumor-Associated Macrophages to Combat Recurrent Lung Cancer

Abstract

Postoperative lung recurrent cancer exhibited characteristics of an immunosuppressive tumor microenvironment (TME) and low immunogenicity, hindering the therapeutic efficacy of monotherapy, which requires a combination of several treatment modules. Strategies that activate the cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) pathway and repolarize tumor-associated macrophages (TAMs) toward the antitumoral M1-like phenotype to reverse the TME are rarely reported. The triggering receptor expressed on myeloid cells 2 (TREM2) is a promising therapeutic target, owing to its critical role in enhancing tumor immunogenicity within the TME. This work describes the design of an anti-TREM2-modified FePt-based biomimetic nanovesicle (FP/Vad@CC-aT2) for the delivery of STING agonist Vadimezan (Vad), which increases tumor immunogenicity to sensitize recurrent lung tumors to immunotherapy. FePt not only acted as a photoacoustic/magnetic resonance imaging contrast agent but also enhanced ferroptosis by catalyzing a Fenton reaction with reactive oxygen species production under X-rays. Simultaneously, anti-TREM2 effectively repolarized TAMs into M1-type macrophages, thereby reversing immunosuppressive TME together with a Vad-activated STING pathway, which promoted the maturation of dendritic cells and enhanced the infiltration of cytotoxic T lymphocytes. Therefore, this study highlighted the FP/Vad@CC-aT2-mediated cascade immune response for suppressing lung cancer recurrence that involves ferroptosis potentiation, TAM repolarization, and STING pathway activation.

Non-small cell lung cancer (NSCLC) is one of the most prevalent and lethal malignancies globally, exerting a substantial burden on public health. The prognosis for advanced-stage NSCLC remains poor, with high rates of postoperative recurrence and metastasis despite the advances in treatments, including surgery, chemotherapy, and radiotherapy. , Immunotherapy, particularly the one using immune checkpoint inhibitors (ICIs), has emerged as a first-line treatment for advanced NSCLC, revolutionizing the therapeutic paradigm because it harnesses the body’s own immune system against cancer cells. , However, many patients with postoperative recurrent NSCLC show low response rates despite the treatment with ICIs, suggesting the development of an immunosuppressive tumor microenvironment (TME) after surgery. ,, The recurrence of the tumor post surgery is significantly associated with the exhaustion of cytotoxic T lymphocytes (CTLs) and the increase of anti-inflammatory M2 phenotype macrophages, which leads to the immunosuppressive characteristics of TME by secreting anti-inflammatory cytokines, including interleukin-10 (IL-10). Consequently, nanotechnology-based immunotherapy, which either activates the immune response or inhibits the immunosuppressive activity, is a promising strategy to inhibit metastasis and recurrence after NSCLC treatments.

Recent research indicated that activating the cyclic GMP-AMP synthase-stimulator of interferon genes (cGAS-STING) pathway is a key component involved in cancer immunity, and its activation enhances tumor immunogenicity. − This activation promotes cytokine secretion and improves antigen presentation, as well as T cell cytotoxicity. − Moreover, growing evidence has demonstrated the role of the STING pathway in reinvigorating the immunogenic TME and increasing the sensitivity to ICI therapy. ,,, Small molecule Vadimezan and natural cyclic dinucleotide − are typical STING agonists that should be encapsulated as a nanoplatform to overcome their metabolic instability and low bioavailability. However, effective immune responses against tumors through the STING pathway are achieved with the involvement of multiple complex pathways, resulting in strategies that target multiple pathways simultaneously to maximize the therapeutic efficacy.

Ferroptosis is an iron-dependent cell death mechanism driven by the Fenton reaction that accelerates the generation of reactive oxygen species (ROS) and the accumulation of lipid peroxidation (LPO) products on the cell membranes, − thus resulting in a strong killing effect on cancer cells, inducing immunogenic cell death (ICD), promoting dendritic cell (DC) maturation, and triggering adaptive T cell responses. Moreover, ferroptosis generates cytosolic dsDNA from damaged mitochondria, thus bidirectionally activating the cGAS-STING pathway. − Therefore, it is crucial to develop therapeutic agents that work together to bridge the gap between STING pathway activation and the initiation of ferroptosis to achieve an effective tumor therapy. ,, More strategies have been developed to effectively promote LPO accumulation to amplify ferroptosis through the increase in the level of Fe²⁺/H₂O₂ and the consumption of GSH in large amounts in the TME. , Additionally, the combination of ferroptosis-based treatments with other therapies such as PD-1/PD-L1 antibodies, X-ray dynamic therapy (XDT), , and photothermal therapy (PTT) has been evaluated, but that with the STING agonist has rarely been explored.

The integration of immunostimulatory properties with therapeutic functions in nanoplatforms has received considerable attention in recent years, despite the extensive development of different therapeutic modalities in nanomedicine. Tumor-associated macrophages (TAMs) are related to the occurrence, development, and metastasis of tumors, suggesting that their targeting might be an antitumor therapeutic strategy. TREM2 is a promising therapeutic target due to its widespread expression on TAMs and its potent immunosuppressive effects in the TME. − Moreover, some studies showed that the role of TREM2 in cancer is tumor-type dependent. Zhang et al. demonstrated that TREM2⁺ TAMs are predictive and prognostic markers for NSCLC in patients under ICI treatment. Binnewies et al. also showed that the treatment with anti-TREM2 alters the abundance and phenotype of TAMs in the TME under anti-PD-1 immune therapy in an orthotopic ovarian cancer model. Both Wang et al. and Wei et al. demonstrated that targeted TREM2 inhibition in the TME reprograms immunosuppressive networks, eliciting robust antitumor immunity and potentiating immune activation in mouse colorectal cancer models. , However, despite the promising therapeutic effect of TREM2 inhibition, current strategies targeting it showed suboptimal efficacy in clinical settings, requiring further optimization and refinement. Consequently, novel strategies are required to precisely modulate the TAM function and reprogram the immunosuppressive TME into an immunostimulatory state, thereby enabling effective ICI therapy.

In this work, we first report a biomimetic therapeutic integration platform for repolarizing M2-type macrophages by targeting tumor cells via homotypic recognition capability, inducing ferroptosis, and promoting STING pathways. By coating the core of Vad and FePt nanoparticles (NPs) with cancer cell membranes (CCM) and conjugating anti-TREM2 antibodies on the surface, the construct was specifically targeted to the tumor microenvironment rich in M2-type macrophages through homing effects and TREM2 recognition. XDT synergistically promoted ferroptosis and activated the cGAS-STING pathway; these effects potentiated anti-TREM2 therapy by DC maturation and enhanced CTL infiltration, ultimately amplifying tumor cell killing. Therefore, this work proposes a new approach based on biomimetic dual-modality imaging-guided radioimmunotherapy using FP/Vad@CC-aT2 nanovesicles and X-ray irradiation, with the aim of increasing the cascade of different crucial steps of immunotherapy. The designed nanovesicles are first engineered by the extrusion of the STING agonist Vad-encapsulated liposome (Vad@LP) and CCM-decorated radiosensitizer FePt (FP@CC) to obtain the CCM-decorated nanovesicle (FP/Vad@CC). Moreover, anti-TREM2 was incubated with the biofunctional linker Sulfo-SMCC to obtain maleimide-activated anti-TREM2 (aT2-Mal) for the conjugation of FP/Vad@CC through the reaction between maleimide and thiols (Figure a). The resulting tumor-targeting nanovesicle (FP/Vad@CC-aT2) possesses the homotypic binding ability and immune escape properties derived from the natural properties of CCM for greater tumor accumulation. This nanovesicle introduces a new dual-action strategy that uses X-rays to trigger and amplify ferroptosis, potentially bridging STING pathway-mediated immunity and radioimmunotherapy (Figure b). Once internalized, Fe²⁺ ions from FP/Vad@CC-aT2 increase ROS levels under X-ray irradiation, starting the release of mtDNA into the cytosol, enhancing cGAS’s sensitivity to mtDNA, boosting cGAMP synthesis, and increasing its affinity to STING. Thus, the cGAS-STING pathway is efficiently amplified in combination with the Vad-activated STING pathway, which enhances type I interferon (IFN-I) production and recruits CD8⁺ T cells. Overall, the combination of ferroptosis potentiation and STING pathway activation with the ICI therapy strategy induced by FP/Vad@CC-aT2 and radiotherapy promotes the maturation of DCs and the infiltration of CTLs, reverses the immunosuppressive TME, transforms the tumor from “cold” to “hot”, and induces a prolonged antitumor immunity against postoperative recurrent tumors (Figure c).

1.

Schematic illustration of the construction of the FP/Vad@CC-aT2 nanovesicle and its potential ability to improve the antitumor radioimmunotherapy. (a) Synthesis of FP/Vad@CC-aT2. (b) Synergistic therapy guided by PA/MR imaging, in which ferroptosis activation and the combined effect of STING-related innate immunity and radioimmunotherapy are involved. (c) Surgical resection creates an immunosuppressive TME characterized by low immune infiltration. FP/Vad@CC-aT2 increases the ROS levels after X-ray irradiation through the Fenton reaction, which further facilitates ferroptosis and the amplification of the cGAS-STING pathway in combination with the Vad-activated STING pathway. These synergistic processes increase tumor immunogenicity and boost the antitumor immune response in postoperative recurrent cancer.

Results and Discussion

Fabrication and Characterization of FP@CC-aT2 and FP/Vad@CC-aT2 NPs

The syntheses of these NPs are shown in Figure a. PEGylated FePt NPs (FP-PEG) were prepared according to our previous report. ,, CCM was extracted and extruded from Lewis lung cancer (LLC) cells as previously reported, , which was used to coat FP-PEG. CCM-coated FePt (FP@CC) NPs were obtained by mixing all components together with a procedure including the extrusion through a porous polycarbonate membrane. The homotypic targeting adhesion molecules, including Galectin-3, CD44, and E-cadherin, were enriched in purified LLC CMV and FP@CC, indicating that three representative homotypic cell adhesion proteins were successfully transferred to the shell of FP@CC (Figure S1a, Supporting Information). Furthermore, the STING agonist Vad-encapsulated liposome (Vad@LP) was prepared by using the film hydration technique. The transmission electron microscopy (TEM) image indicated the uniform spheroid morphology of the FP@CC nanovesicle (Figure S1b, Supporting Information). FP/Vad@CC was prepared by the fusion of FP@CC with Vad@LP according to a previous study, with minor revision. , The anti-TREM2 (aT2) antibody was conjugated to cancer cell membranes (CCM) via Sulfo-SMCC, a heterobifunctional cross-linker enabling stable covalent amide bond formation. FP@CC-aT2 possessed a uniform spheroid shape as revealed by the HAADF-STEM image (Figure b). The presence of Fe and Pt in the FP@CC-aT2 was confirmed by element mapping analysis (Figure S2, Supporting Information). FP@CC and FP/Vad@CC possessed ζ-potentials of −28.6 and −24.2 mV, while FP@CC-aT2 and FP/Vad@CC-aT2 possessed ζ-potentials of −5.1 and −3.5 mV, respectively, which confirmed successful decoration of aT2 antibody on the surface of NPs (Figure c). Furthermore, the drug Vad loading capacity was approximately 9.2 ± 0.4% determined by a triple quadrupole mass spectrometer. The in vitro Vad release profile showed that the released drug was approximately 40% at 4 h and 78% at 24 h in a buffer solution at pH 6.5, which was faster and more than that in a buffer solution at pH 7.4 (Figure e). According to our design, FePt was gradually disintegrated in an acidic condition to release Fe²⁺, which reacted with the high endogenous amount of H₂O₂ to produce ^•OH through the Fenton reaction. To verify it, the EPR measurements clearly revealed that the incubation of H₂O₂ and 3,4-dihydro-2,3-dimethyl-2H-pyrrole 1-oxide (DMPO) with FP/Vad@CC-aT2 resulted in a strong four-line EPR signal with a 1:2:2:1 peak-to-peak intensity pattern (Figure f), which was similar to a previous report.

2.

Synthesis and comprehensive characterization of nanovesicles. (a) Schematic illustration of the synthesis of biomimetic nanovesicles. (b) Representative HAADF-STEM images and corresponding elemental mapping of FP@CC-aT2. Scale bar: 200 nm. (c) Ζ-potentials of FP@CC, FP/Vad@CC, FP@CC-aT2, and FP/Vad@CC-aT2. (d) Time-dependent stability of FP@CC, FP/Vad@CC, FP@CC-aT2, and FP/Vad@CC-aT2 in PBS at pH 7.4. (e) Vad release profile of FP/Vad@CC in pH 6.5 and 7.4 buffer solutions. (f) EPR spectrum of formed DMPO-HO^• after coincubation of FP/Vad@CC-aT2 with H₂O₂. (g) T1 and (h) T2 relaxation rates of FP/Vad@CC-aT2 incubated at different concentrations. (i) PA signal intensity of FP/Vad@CC-aT2 versus a function of concentration in pH 6.5 and 7.4 buffer solutions. Results are expressed as the mean ± standard deviation. Figure (a) was created by Figdraw.

Previous studies revealed the use of FePt-based nanosystems for photoacoustic imaging (PAI) and magnetic resonance imaging (MRI) tumor imaging. ,, Thus, the potential T1 and T2-weighted MR imaging ability of FP/Vad@CC-aT2 at different concentrations was investigated. The concentration-dependent brightening effect of FP/Vad@CC-aT2 was monitored in T1- and T2-MR images at pH 6.5, and the corresponding signal-to-background ratio (SBR) was changed, as shown in Figure g,h. In addition, the concentration-dependent PA signal changes of FP/Vad@CC-aT2 were observed in buffers at pH 6.5 and pH 7.4, and the in vitro PA signal strength of FP/Vad@CC-aT2 showed a positive correlation with its concentration (Figure i). The hemolytic toxicity test was performed before starting the in vitro study, revealing that the FP@CC-aT2 nanovesicle did not significantly change the level of biotoxicity even up to 40 μg/mL (based on Fe concentration; Figure S3, Supporting Information), which was appropriate for the in vitro and in vivo studies.

In Vitro FP@CC-aT2-Potentiated Ferroptosis

The cellular uptake of Rho-labeled NPs (FP@CC_Rho and FP@CC_Rho-aT2) was observed by CLSM in vitro to evaluate the targeting effect of FP@CC-aT2. FP@CC_Rho and FP@CC_Rho-aT2 emitted a stronger red fluorescence signal than the control group in LLC cells after 3 h of incubation (Figure S4, Supporting Information), revealing that CCM decoration did significantly enhance tumor cell uptake.

According to the design, FP@CC-aT2 with X-ray irradiation and a H₂O₂ supply should remarkably improve the Fe²⁺-mediated Fenton reaction and consequently amplify ferroptosis in LLC. The 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) staining assay was used for the investigation of the intracellular ROS level (Figure a). The greatest green fluorescence signals were seen in LLC cells treated with FP@CC-aT2 + X-ray + H₂O₂, indicating that this therapy was most effective in producing ^•OH within tumor cells to enhance ferroptosis. Moreover, the ferroptosis inhibitor ferrostatin-1 (Fer-1) was employed to shed additional light on the mechanism underlying the FP@CC-aT2-induced cell death. Ferroptosis appears to be primarily engaged in the FP@CC-aT2-mediated cell death pathway, as indicated by the significantly reduced green fluorescence signal observed in cells incubated with FP@CC-aT2 + Fer-1 compared to that in other groups. Besides, ROS levels within LLC cells were analyzed using the DCFH-DA probe to compare the different treatment groups, which is similar to the result of fluorescence intensity, indicating ROS levels in different treatment groups. ROS levels within LLC cells were analyzed using flow cytometry by measuring fluorescence intensity to compare the FP@CC-aT2 + X-ray to other control groups. Fluorescence intensity was significantly increased in FP@CC-aT2 + X-ray compared to other treated groups (Figure S5, Supporting Information), confirming that ROS induced by FP@CC-aT2 and XDT was due to ferroptosis.

3.

In vitro study of the FP@CC-aT2-based nanoplatform. (a) CLSM images of LLC cells stained with DCFH-DA (green) to detect the intracellular-generated ROS level after various treatments (scale bar: 100 μm). (b) Representative confocal images of LLC cells stained with C11-BODIPY (green) to detect the LPO level (Scale bar: 100 μm). (c) Western blot analysis of GPX-4 expression in LLC cells after different treatment groups. I, PBS; II, FP@CC; III, FP@CC-aT2; IV, FP@CC-aT2 + X-ray. (d) CLSM images of LLC cells stained with Calcein-AM (green, live cells) and PI (red, dead cells) after different treatments (scale bar: 200 μm). (e) Western blot analysis of apoptosis-related proteins in LLC cells after various treatments. I, PBS; II, FP@CC; III, FP@CC-aT2; IV, FP@CC-aT2 + X-ray. (f) CLSM images of RAW264.7 cells stained with CFDA-SE (green) in 3D tumor spheroids after various treatments (scale bar: 500 μm).

The intracellular generation of LPO , was also assessed in LLC after different treatments using C11-BODIPY (Figure b). LLC cells treated with FP@CC-aT2 + X-ray + H₂O₂ exhibited the highest LPO generation, as revealed by the green fluorescence in the cell membrane. However, the addition of the ferroptosis inhibitor Fer-1 significantly reduced the fluorescence intensity. The expression of glutathione peroxidase 4 (GPX-4), which is the main regulator of ferroptosis, was also assessed to further study the mechanism of FP@CC-aT2-induced cell death. , GPX-4 protein expression was significantly increased in untreated LLC compared to that in other treated groups (Figure c), confirming that ferroptosis induced by FP@CC-aT2 and XDT was due to the inhibition of GPX-4 expression. Since the excessive ROS production in cells leads to apoptosis, the cytotoxic effect of FP@CC-aT2 with or without X-ray irradiation was assessed in LLC tumor cells after 24 h of treatment using the standard CCK-8 assay (Figure S6a, Supporting Information). All nanovesicles induced a dose-dependent inhibitory effect on LLC cell growth. Moreover, more than 80% cells were dead in the FP@CC-aT2 + X-ray group when the Fe concentration was 40 μg/mL, which was far higher than those of FP@CC-aT2 and FP@CC-aT2 + Fer-1 + X-ray groups, indicating the synergistic effect of ferroptosis and XDT. The therapeutic effect of FP@CC-aT2 combined with X-ray treatment was assessed using dead-live staining and CLSM images (Figure d). Many dead cells (red) in the FP@CC-aT2 + X-ray group were the highest due to its cooperative X-ray irradiation and anti-TREM2 that amplified ferroptosis. In contrast, neither the control nor FP@CC-aT2 + Fer-1 was able to effectively kill cancer cells. Apart from the regulation of ferroptosis, apoptotic death was further confirmed by measuring PARP1, Bcl-2, and caspase-3 protein expression (Figure e). PARP1 and caspase-3 were upregulated, while Bcl-2 was downregulated after treatment with FP@CC-aT2 + X-ray.

FP/Vad@CC-aT2 Induced Cell Death In Vitro and Related Mechanism

The results of the cytotoxicity effect of FP/Vad@CC-aT2 with or without X-ray irradiation toward LLC cells after 24 h of treatment revealed a dose-dependent inhibitory effect (Figure S6b, Supporting Information). The IC₅₀ of the FP/Vad@CC-aT2 + X-ray, FP/Vad@CC-aT2, and FP@CC-aT2 + X-ray was 15.16, 86.98, and 116.8 μg/mL, respectively, indicating the synergistic therapy, since the loaded Fe²⁺ and Pt could enhance the X-ray dynamic effect and subsequently trigger ferroptosis. The apoptotic rates of LLC cells treated with FP/Vad@CC-aT2 + X-ray and FP@CC-aT2 + X-ray were 14.8% and 8.51%, respectively, which were higher than that of the control group (3.12%), suggesting that the addition of Vad increased programmed cell death (Figure S7, Supporting Information).

FP/Vad@CC-aT2 Promoted Macrophage Phagocytosis

To comprehensively assess FP/Vad@CC-aT2 effects on macrophage activity, we established a 3D coculture system of LLC tumor spheroids and RAW264.7 macrophages. Confocal imaging of CFSE-labeled macrophages revealed significantly enhanced migration and infiltration into tumor spheroids after FP@CC-aT2 treatment with or without Vad versus FP@CC and CCM controls (p < 0.01), indicating improved recruitment/activation (Figure f). Flow cytometric phagocytosis assays (CFDA-SE⁺ tumor cells + CellTrace-red⁺macrophages) demonstrated markedly higher engulfment rates in FP/Vad@CC-aT2-treated groups than antibody-free controls (p < 0.001; Figure S8a, Supporting Information), with correlative CLSM confirming increased macrophage migration and tumor cell phagocytosis (Figure S8b, Supporting Information). Collectively, these data establish that FP/Vad@CC-aT2 promotes macrophage recruitment/infiltration into tumor spheroids and enhances phagocytic clearance of tumor cells, validating the nanovesicle’s dual immunomodulatory function. Notably, sustained nanoparticle internalization by tumor cells (Figure S4, Supporting Information) creates a self-amplifying therapeutic loop.

In Vivo Dual-Modality PA/MR Imaging

A subcutaneous LLC tumor-bearing mouse was created to further determine the feasibility of nanovesicles for dual-modality imaging in vivo and monitor their accumulation into the tumor. The T1-weighted and T2-weighted MRI images and the related PA signals at the tumor site were captured at different time intervals (0, 4, 8, 24, and 48 h) after the injection of FP/Vad@CC-aT2, FP@CC, and FP (5 mg/kg, 200 μL; Figure a). Notably, the enhanced T1-weighted (bright) and negatively enhanced T2-weighted (dark) MR images after the treatment with FP/Vad@CC-aT2 and FP@CC were clearly observed at 4 h postinjection (Figure b,c), indicating the release of more Fe²⁺ from FP/Vad@CC-aT2 and FP@CC when located in the acidic TME. A similar tendency was observed in the PA images at the tumor site (Figure d). The PA signal of FP/Vad@CC-aT2 and FP@CC was progressively increased, reaching a peak at 4 h postinjection, and gradually decreased thereafter, exhibiting a similar change trend. Furthermore, FP/Vad@CC-aT2 showed a clearly stronger PA signal than FP, indicating the adhering molecules and surface antigens present in the CCM placed on the surface of these nanovesicles. Moreover, the PA signals from nanovesicles at 4 h postinjection were distinct from endogenous signals (OxH and deOxH) as observed by PA multispectral scanning. These results confirmed the high biocompatibility of FP/Vad@CC-aT2 and its outstanding tumor-specific PA/MR imaging abilities. Moreover, the pharmacokinetics and excretion of FP/Vad@CC-aT2 showed that the circulation half-life of FP/Vad@CC-aT2 in vivo was calculated to be 4.5 h (Figure S9, Supporting Information), which is consistent with the maximum uptake time point for PA/MR imaging. For visual observation of the biodistribution, ICG-labeled FP/ICG@CC-aT2 was intravenously injected for in vivo and ex vivo imaging (Figure S10, Supporting Information). FP/ICG@CC-aT2 exhibited peak tumor accumulation at 4 h postinjection, followed by predominant hepatic clearance, demonstrating efficient tumor-targeting capability.

4.

In vivo PA and MR images of LLC tumor-bearing mice at different time points (0, 4, 8, 24, and 48 h) postinjection of FP/Vad@CC-aT2, FP@CC, and FP. (a) In vivo T1WI and T2WI of tumors at different time points after intravenous injection of FP/Vad@CC-aT2, FP@CC, and FP nanovesicles. (b, c) The corresponding T1 SBR (b) and T2 SBR (c) of the tumor area. (d) PA-US overlay images of tumors after the injection of different nanovesicles (blue). Multiplexed PA images of nanovesicles (green), oxygenated hemoglobin (OxH, red), and deoxygenated hemoglobin (deOxH, blue) in the tumors are superimposed on the ultrasound image (gray color). (e–g) Average PA intensity increment at 700 nm in the tumor after the injection of FP/Vad@CC-aT2 (e), FP@CC (f), and FP (g) (n = 3). Results are expressed as mean ± standard deviation. *P < 0.05, ***P <0.001.

In Vivo Antitumor Performance and Potential Immunologic Effects of FP@CC-aT2

Encouraged by the superior imaging performance of FP@CC-aT2 in vivo, its effect in a subcutaneous lung cancer model was assessed (Figure a). FP@CC-aT2 mediated ferroptosis and XDT induced direct tumor cell death and remodeling of the immunosuppressive TME, amplifying ferroptosis and enhancing ICD effects (Figure b). Kaplan–Meier survival curves showed a significantly better survival in the FP@CC-aT2 and FP@CC-aT2 + X-ray groups (Figure c). The FP@CC-alone group exhibited a minimal effect on tumor growth inhibition, suggesting the limited efficacy of FP-induced ferroptosis therapy. However, tumor growth was remarkably suppressed after the treatment with FP@CC-aT2 and FP@CC-aT2 + X-ray in comparison to the control group (Figure d). Anti-TREM2, ferroptosis, and XDT enhanced the antitumor efficacy. Additionally, the mouse body weight did not change during treatments, suggesting the biosafety of FP@CC-aT2 (Figure e). Neither toxicity was observed in the main organs (heart, liver, spleen, lung, and kidney) in different treated groups, nor significant changes in blood biochemistry (e.g., ALT, TBIL, DBIL, γ-GT, TBA, UREA, CREA, UA, and CK-MB) after the treatment with FP@CC-aT2 + X-ray compared to the control group (Figures S11 and S12, Supporting Information), demonstrating the biosafety and biocompatibility of the FP@CC-aT2 nanoplatform. Typical histopathological damage was observed in the combination therapy group (Figure f). However, no significant damage was found in the control group. In addition, evaluation of K67 expression revealed its weak expression in the combination therapy group, indicating an inhibitory effect on tumor proliferation. In addition, the percentage of apoptotic cells was the highest in the combined group, while no evident apoptosis was found in the control group. Tumor cell death mechanism induced by FP@CC-aT2 revealed the involvement of GPX-4, whose expression was downregulated, and 4-hydroxynonenal (4-HNE), whose expression was increased in the FP@CC-aT2 + X-ray group, revealing that ferroptosis was mediated by FP@CC-aT2 + X-ray (Figure g,h).

5.

Synergized anticancer effect and mechanism of action of FP@CC-aT2 and X-ray in the inhibition of LLC primary tumor growth. (a) Protocol for the evaluation of tumor implantation and treatments. (b) Schematic illustration of the mechanism of action of FP@CC-aT2-mediated ferroptosis and XDT in TME. (c–e) Kaplan–Meier survival curves (c), average tumor growth curves (d), and body weight (e) of tumor-bearing mice after different treatments (n = 5). (f) Representative images of tumors stained with hematoxylin and eosin (H&E), Ki67, and TUNEL (scale bar: 100 μm). (g, h) Representative immunofluorescence images of 4-HNE and GPX-4 expression in tumor slices. Scale bar: 100 μm. (i) Representative images of CRT and HMGB1 immunohistochemical staining (scale bar: 100 μm). (j, k) Quantification of CRT and HMGB1 expression in excised tumors (n = 5, mean ± SD, one-way ANOVA, n.s.= not significant. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). Figure (a,b) was created by Figdraw.

The expression of calreticulin (CRT) on the cell membrane and the release of the typical immunogenic substance high-mobility group box 1 (HMGB1) in the combination group and FP@CC-aT2 further confirmed that FP@CC-aT2 + X-ray induced an ICD effect after treatment (Figure i–k). These results suggested that FP@CC-aT2 + X-ray nanotherapy triggered ferroptosis through the cooperation of GXP4-CRT-HMGB1 and the Fe²⁺-ROS-Fenton pathway.

Transcriptomic Analysis of the Antitumor Mechanism

The immunologic phenotype changes in the residual lung tumors collected after surgery were assessed using immunofluorescence (IF) staining to explore the cellular mechanisms behind the immune changes in the TME triggered by surgery. The results showed a significant increase in the infiltration of CD206⁺ anti-inflammatory M2 phenotype macrophages, mostly TREM2⁺, in surgically treated tumors compared to that in normal tissue (Figure a). This result revealed that an increased number of M2 macrophages, along with decreased CTL infiltration, resulted in a markedly more immunosuppressive and exhausting TME in surgically resected tumors.

6.

FP@CC-aT2-mediated antitumor immunotherapy mechanism was determined by transcriptomic analysis. (a) Representative polychromatic immunofluorescent staining images of tumor and normal tissue from human surgical samples showing the infiltration of TREM2+ (red) and CD206⁺ (yellow) cells. Scale bar: 200 μm. (b) Principal component analysis of transcriptome expression of FP@CC-aT2, FP@CC, and control groups. (c) Heatmap of differentially expressed genes (DEGs) associated with tumor progression and immunosuppression in FP@CC-aT2, FP@CC, and control groups. (d) Gene Ontology (GO) enrichment analysis and (e) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway of the DEGs for the FP@CC-aT2 group. (f) Gene set enrichment analysis (GSEA) of the immune response signature for immunotherapy-related genes induced by FP@CC-aT2. (g) Assessment of immune cell infiltration in tumor tissues after treatment with FP@CC-aT2 and FP@CC groups.

Samples of LLC tumors were collected for transcriptomic analysis on day 18 after treatment with FP@CC-aT2 and FP to investigate the effect of FP@CC-aT2 on TME, while tumors from untreated tumor-bearing mice were used as a control group. The pathways and mechanisms of action of FP@CC-aT2 in vivo were assessed using RNA sequences (RNA-seq) after treatment with FP@CC-aT2, FP@CC, or PBS (Figure b). A total of 15,701 genes were analyzed, and among them, 394 differentially expressed genes (DEGs) were upregulated and 169 were downregulated (Figures S13a,b and S14, Supporting Information). The 563 DEGs were selected from the total genes by fold change ≥ 2 plus two standard deviations (Figure c), log2 |fold change| ≥ 1, and p < 0.05. aT2 Ab promoted the polarization of M1-like macrophages, which further enhanced immune cell receptor recognition, immune cell differentiation, cytotoxic effects, and tumor-killing response.

Gene Ontology (GO) analysis indicated that DEGs were significantly enriched in the molecular function (MF) category, particularly in receptor ligand activity and immunoglobulin receptor binding, then in biological processes (BP) related to immune functions and pathways, including differentiation of lymphocytes, mononuclear cells, and T cells, adaptive immune response based on recombination of immune receptors built, and second messenger-mediated signaling (Figure d). Some studies revealed that the core of anti-TREM2 immunotherapy is the activation of immune cells and their induction to kill tumor cells. FP@CC-aT2 treatment downregulated the PI3K-AKT signaling and upregulated macrophage differentiation and natural killer cell-mediated cytotoxicity signaling, as revealed by the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. KEGG analysis showed that FP@CC-aT2 treatment upregulated macrophage as well as Th1 and Th2 cell differentiation signals in the immune cell differentiation process-related content (Figure e). Gene Set Enrichment Analysis (GSEA) also indicated that the immune response signature for immunotherapy-related genes induced by FP@CC-aT2 was related to natural killer cell-mediated cytotoxicity, lymphocyte trans-endothelial migration, T cell, and NF-kappa B receptor signaling pathway (Figures f and S15, Supporting Information). Pathway analysis in T and B cell receptor signaling demonstrated that FP@CC-aT2 treatment induced an effective immune response and mobilized immune cells. Immune cell infiltration in the tumor after FP@CC-aT2 treatment was assessed to further evaluate the change in the TME. Activation of CD8⁺ T cells by the STING pathway is a vital part of the synergistic increase in the immune response.

Our results showed that multiple antitumor immune cell infiltrates, especially NK cells and activated effector CD8⁺ T cells, significantly increased in the FP@CC-aT2 group (Figure g). Collectively, the transcriptome analysis of the tumor further demonstrated that FP@CC-aT2 effectively reshaped the immunosuppressive TME to an immunogenic phenotype to improve the antitumor immune response.

Antitumor Effect of FP/Vad@CC-aT2 and XDT in the Recurrent Lung Tumor

TREM2 and STING acted as core proteins, forming a tight interaction network with immune cell activation proteins, as revealed by the protein–protein interaction (PPI; Figure a). The size and color of the rings in the PPI network indicated the importance of different genes. The immune cell activation proteins CD8A, CD4, CD68, CD80, and CD86 were activated to form mature CD8⁺ T cells, CD4⁺ T cells, DC cells, and macrophages, respectively. An orthotopic lung cancer model was first established, and lungs were harvested at the endpoint for H&E staining (Figure b). The H & E-stained sections of the orthotopic Matrigel model revealed invasive growth patterns characteristic of clinical lung cancer, thereby validating the model’s physiological relevance. The FP/Vad@CC-aT2 + X-ray group showed significant tumor regression (≤10% residual viable tumor cells) and no evidence of recurrence in all mice, compared to extensive residual disease in controls. In addition, the FP/Vad@CC-aT2 + X-ray group showed a higher apoptosis rate than the single treatment group, indicating the promotion of tumor cell apoptosis and necrosis. The result is consistent with the WB result of obvious apoptosis induced for the FP/Vad@CC-aT2 + X-ray group (Figure e). These data validate FP/Vad@CC-aT2 plus X-ray efficacy against orthotopic lung cancer, demonstrating the translational value of the Matrigel-based model for metastasis recapitulation and recurrence inhibition. Subsequently, a postoperative recurrent LLC tumor model was established to evaluate the synergistic effects of FP/Vad@CC-aT2 and XDT. FP/Vad@CC-aT2 + X-ray significantly prolonged the survival of recurrent tumor-bearing mice compared to the other groups (Figure c), indicating a therapeutic effect on postoperative recurrent tumor. FP/Vad@CC-aT2 + X-ray effectively suppressed tumor growth compared to the control groups (Figures d,f and S16, Supporting Information). The CC-aT2-alone group showed little effect on suppression of the tumor growth compared to the control group (saline). It is worth noting that the treatment of both FP@CC + X-ray and Vad@CC presented some growth inhibition, indicating certain antitumor activity of themselves alone (Figure S16b, Supporting Information). Moreover, FP/Vad@CC-aT2 + X-ray notably presented significantly stronger antitumor efficacy compared to FP@CC-aT2 + X-ray and Vad@CC-aT2 + X-ray, reflecting the synergistic effect of radioimmunotherapy (Figure d,f). The photographs of excised subcutaneous tumors from different treatment groups in the recurrent lung tumor model also demonstrated the antitumor effect of various treatment groups (Figure S17, Supporting Information). Furthermore, the body weight of treated mice was slightly increased with no significant differences compared to the control group (Figure e). More necrosis and apoptosis, as well as less proliferation, were observed in the FP/Vad@CC-aT2 + X-ray group compared with other groups, as shown by H&E, TUNEL, and Ki67 staining results (Figure S18, Supporting Information). Tumor samples were collected for Western blot analysis at the end of the experiment, and the results revealed that the cGAS-STING pathway was significantly activated in the tumor region in the FP/Vad@CC-aT2 + X-ray group, showing an enhanced phosphorylation level of STING, TBK1 and IRF3 proteins (Figure g,h), which was attributed to the release of the STING agonist Vad, the activated ferroptosis induced by Fe²⁺, and the amplified dsDNA induced by radiotherapy. The differential gene clustering analysis of the transcriptome further demonstrated that the FP/Vad@CC-aT2 + X-ray group activated the antigen presentation signaling pathway and the interferon signaling pathway, and the high expression of related genes, such as IFN-γ, activated the macrophages with proinflammatory and tumor-killing phenotypes (Figure i). It was shown that the interferon pathway promotes T, NK, DC, and macrophage activity, facilitates antigen presentation, and plays a role in anticancer immunity. The FP/Vad@CC-aT2 + X-ray group showed a substantial infiltration of M1 macrophages, as revealed by the prominent green fluorescence (Figure j). Collectively, these results indicated that FP/Vad@CC-aT2 + X-ray enhanced the STING pathway and subsequently strengthened the powerful long-term immune responses against postoperative lung recurrent tumor. The content of the proinflammatory M1 macrophage marker CD86 (green fluorescence) and the anti-inflammatory M2 macrophage markers TREM2 (red fluorescence) and CD206 (yellow fluorescence) was measured to assess the effect of FP/Vad@CC-aT2 on the TME of a recurrent tumor. This result was due to the TREM2 antibody promoting macrophage phenotypic change, which in turn caused a significant buildup of M1-type macrophages. On the other hand, M2-type macrophages were more prevalent, and M1-type macrophages were less prevalent in the control group.

7.

Synergized anticancer efficacy and mechanism of FP/Vad@CC-aT2 with radiotherapy for the inhibition of postsurgical LLC recurrent tumor. (a) PPI network analysis of immune cell activation proteins. (b) Representative H&E images of the orthotopic lung model treated with PBS, Vad@CC + X-ray, FP@CC-aT2 + X-ray, Vad@CC-aT2 + X-ray, and FP/Vad@CC-aT2 + X-ray. (c) Kaplan–Meier survival curves, (d) average tumor volume, and (e) body weight changes of tumor-bearing mice after different treatments (n = 6). (f) Individual tumor growth curves of recurrent tumors. (g) Western blotting images and (h) corresponding analysis of the cGAS-STING pathway-related proteins expression of pSTING. I, PBS; II, FP@CC-aT2; III, FP/Vad@CC-aT2, IV, FP/Vad@CC-aT2 + X-ray. (i) Heatmap visualization of hierarchical clustering showing the differential content profile of antigen presentation and interferon signaling pathway-related genes. I, PBS; II, FP/Vad@CC-aT2; III, FP/Vad@CC-aT2 + X-ray. (j) CLSM examination of FP/Vad@CC-aT2 treatment in the tumor (blue: DAPI; green: CD86; yellow: CD206; red: TREM2; scale bar = 200 μm). Results are expressed as mean ± SD; n.s.= not significant. *P < 0.05 and **P < 0.01 and ***P< 0.001 by one-way ANOVA with Tukey’s multiple comparisons test.

In Vivo Immune Response and Reinvigoration of the Immunosuppressive TME through FP/Vad@CC-aT2 and XDT

Inspired by the proven results that FP/Vad@CC-aT2 induced the amplified cGAS-STING protein secretion after XDT, as well as its capacity to inhibit the invasion of cancer cells in vitro, we predicted that FP/Vad@CC-aT2 may also inhibit LLC tumor recurrence in vivo, relieving the defects of radiotherapy. To further explore the reconstruction of the tumor immune microenvironment caused by this treatment, we conducted further analysis of the obtained tumor tissue components and tumor-draining lymph nodes (Figure S19, Supporting Information). The maturation of DCs inside the lymph nodes was moderately increased when the mice were treated with Vad@CC, FP@CC-aT2, and Vad@CC-aT2, reaching 22.7, 25.4, and 36.8%, respectively. In contrast, the highest percentage of maturation of DCs (58.3%) was obtained in the FP/Vad@CC-aT2 + X-ray group, indicating that the most significantly increased maturation of DCs was achieved after FP/Vad@CC-aT2 + X-ray treatment and not with other treatments (Figure a,b). The tumor showed a significant increase in the number of proinflammatory M1-like macrophages (CD86⁺, Figure c,d) and a decrease in the number of M2-like macrophages (CD206⁺; Figure S20, Supporting Information) after treatment with FP/Vad@CC-aT2 + X-ray compared with with other treatment groups. Additionally, the ratio of M1/M2, which is a crucial marker of the immunostimulatory environment, was significantly increased in the FP/Vad@CC-aT2 + X-ray group (Figure g), indicating that this treatment reversed the polarization of TAMs from the anti-inflammatory M2 phenotype to the proinflammatory M1 phenotype, which transformed the immune microenvironment from the immune suppression form to the immune activation form. The significant DC maturation and TAM-M1 polarization demonstrated the cascade enhancement of immune effects after the induction of ferroptosis, anti-TREM2, and activation of the STING pathway in the FP/Vad@CC-aT2 + X-ray group.

8.

FP/Vad@CC-aT2 synergized with radiotherapy reverses the immunosuppressive TME. (a–f) Flow cytometry analysis and quantification of mature DCs (CD80⁺CD86⁺CD11c⁺CD45⁺) (a, b), TAMs-M1 (CD80⁺F4/80⁺CD11b⁺CD45⁺) (c, d), and CTLs (CD8⁺CD3⁺CD45⁺) (e, f) from recurrent tumors after different treatments. (g) Quantification of M1/M2 ratios measured by flow cytometry. TNF-α (h) and IFN-γ (i) expression in the tumor after different treatments measured by the ELISA assay. G1, PBS; G2, Vad@CC; G3, FP@CC-aT2; G4, Vad@CC-aT2; and G5, FP/Vad@CC-aT2. (j, k) Quantification and images of IHC staining of CD4⁺ and CD8⁺ T cell infiltration in the tumor after different treatments. Scale bar: 100 μm. (l) IF staining of the infiltration of CD80+ and CD86+ cells in the tumor after the treatment with the FP/Vad@CC-aT2 + X-ray group. Scale bar: 200 μm. (m) Heatmap visualization of hierarchical clustering showing the differential content profile of M1–M2 polarization and T cell signaling pathway-related genes. I, PBS; II, FP@CC-aT2; III, FP/Vad@CC-aT2 + X-ray. Results are expressed as mean ± SD; n = 3; n.s. = not significant. *P < 0.05 and **P < 0.01, ***P < 0.001,****P < 0.0001 by one-way ANOVA with Tukey’s multiple comparisons test.

The group treated with FP/Vad@CC-aT2 + X-ray showed a more significant infiltration of CTLs (CD8⁺ CD3⁺CD45⁺, Figure e,f) in the lung recurrent tumor compared with other groups due to the greater maturation of DCs and polarization of TAMs-M1. Moreover, the number of tumor-promoting regulatory T cells (Treg, Foxp3⁺) slightly decreased in the mice treated with FP/Vad@CC-aT2 + X-ray (Figure S21, Supporting Information), further confirming that postoperative recurrent tumor-driven immunosuppression was reduced. FP/Vad@CC-aT2 + X-ray treatment significantly increased the expression of tumor necrosis factor-α (TNF-α, Figure h) and interferon γ (IFN-γ, Figure i), thanks to the excellent tumor penetration of the nanovesicle due to CCM and the improved entry of Vad and anti-TREM2. However, treatment with Vad@CC-aT2 and X-ray alone promoted the secretion of inflammatory factors to a certain extent, presumably due to the moderate effect of Vad and anti-TREM2, while treatment with FP@CC-aT2 or Vad@CC alone weakly promoted the expression of inflammatory cytokines. The significantly increased levels of TNF-α and IFN-γ in the tumors treated with FP/Vad@CC-aT2 + X-ray indicated that the activation of the STING pathway facilitated IFN secretion, enhancing the efficacy of tumor immunotherapy.

The inadequate infiltration into the tumor itself may be related to the relatively poor antitumor effects and requires further investigation for other treated groups, such as Vad@CC and FP@CC-aT2. Then, we also used IHC and IF staining to investigate the composition and distribution of immune cells in the tumor. Similarly, IHC staining revealed a significant infiltration of CD4⁺ and CD8⁺ T cells in the tumor region of the FP/Vad@CC-aT2 group, scattered throughout the tumor tissue since the induction of ferroptosis with Fe²⁺-mediated activation of the STING pathway (Figure j,k). IF staining demonstrated a significant increase in the infiltration of CD80⁺ and CD86⁺ cells in the tumor of the FP/Vad@CC-aT2 group compared to the control group (Figure l), further corroborating the flow cytometry data. In addition, the heatmap visualization data demonstrated that the interferon pathway and macrophage phenotype switching play a key role in tumor immunotherapy (Figure m). FP/Vad@CC-aT2-mediated cancer immunotherapy repolarizes M2 TAMs to M1 TAMs, activates the STING pathway in M1 TAMs to promote intratumoral IFN-I secretion, and matures DCs to recruit CTLs compared to the PBS and FP@CC-aT2 group. Overall, the treatment with FP/Vad@CC-aT2 + X-ray represented a synergistic strategy, inducing a cascade enhancement effect represented by the induction of ferroptosis, antagonism of TREM2, and activation of the STING pathway.

Conclusions

A new anti-TREM2-decorated FePt-based nanovesicle was constructed to enhance tumor immunogenicity through the potentiation of ferroptosis, activation of the STING pathway, and repolarization of TAMs, resolving the challenges of the immunosuppressive TME and insufficient TILs encountered by conventional postoperative lung immunotherapy. FP/Vad@CC-aT2 more efficiently catalyzed the production of •OH from endogenous H₂O₂ in TME after X-ray irradiation, which amplified oxidative stress that in turn triggered ferroptosis to boost antitumor immunity. Moreover, STING pathway activation by Vad and TREM2 blockade by anti-TREM2 promoted a more efficient transformation of M2-like macrophages into the antitumor M1-like macrophages, which efficiently reversed the immunosuppressive TME, leading to a cascade of self-enhanced immune responses to postsurgical recurrent tumors. Furthermore, FP/Vad@CC-aT2 could be used as a contrast agent for dual-modality PA/MR imaging-guided tumor treatment. These results demonstrated that the introduction of anti-TREM2 and STING agonist reversed the immunosuppressive TME in recurrent tumors, and the combination with ferroptosis and XDT further improved the synergistic therapeutic effect on recurrent tumors.

Materials and Methods

Materials

Pt­(acac)₂, Fe­(acac)₃, 1,2-hexadecandiol, oleic acid, oleylamine dioctyl ether, DSPE-PEG2000, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), and cholesterol were purchased from Sigma. Sulfo-SMCC was purchased from Thermo Scientific. Vadimezan (DMXAA) and 5-carboxyfluorescein diacetate N-succinimidyl ester (CFDA-SE) was purchased from MedChemExpress (China). A ferroptosis inhibitor (Fer-1) was purchased from Selleck. H₂O₂ and 3,4-dihydro-2,3-dimethyl-2H-pyrrole 1-oxide (DMPO) were obtained from Sigma. Rhodamine-NHS was from Merck Chemicals. C11-BODIPY581/591 was purchased from DOJINDO. 2,7-Dichlorodihydrofluorescein diacetate (DCFA-DA) and AM/PI were obtained from Solarbio. Phalloidin was obtained from Thermo Fisher. DAPI, Annexin V-FITC, and propidium iodide (PI) were obtained from DOJINDO. Antibodies, including recombinant snti-TREM2 antibody (ab245227), MedChemExpress, HY-P80920, STING (ab288157), IRF3 (ab68481), caspase-3 (ab32351), cleaved caspase-3 (ab32042), glutathione peroxidase 4 (ab125066), and GADPH (ab8245), were purchased from Abcam. pSTING (19781), TBK1 (38066), P-TBK1 (5483), and p-IRF3 (79945) were purchased from Cell Signaling Technology. The Cell Counting Kit-8 (CCK-8) reagent was obtained from Biosharp (China). Other chemicals and reagents were of analytical grade and used as received.

Synthesis CCM-Modified FePt (FP@CC)

LLC cancer cell membrane (CCM) fragments was prepared according to our previous report. The FePt and PEG-modified FePt (FP-PEG) nanoparticles (NPs) were synthesized via a thermo-reduction based on our previous report. ,, Subsequently, FP-PEG was surface-functionalized using the CCM fragments via filtering the latter through 400 nm polycarbonate porous membranes (Avanti) 10 times to obtain FP@CC.

Synthesis of Vad@LP, FP/Vad@CC,and FP/Vad@CC-aT2

The STING agonist Vadimezan (Vad)-encapsulated liposome (Vad@LP) was synthesized by a thin-film hydration method, followed by extrusion based on a previous report. 30.3 mg of soy lecithin, 1.6 mg of cholesterol, and 14 mg of DSPE-PEG₂₀₀₀ were dissolved in 40 mL of chloroform, and 120 μL of Vad solution (7 mg/mL in THF) was added. The obtained solution was subjected to vacuum rotary evaporation to completely remove the solvent. Afterward, the formed thin film was hydrated with 5 mL of water under sonication. The dispersion was extruded against a 200 nm polycarbonate membrane 10 times. The solution was then transferred to a dialysis bag with a 100 kDa cutoff and dialyzed against DI water for 48 h. Afterward, Vad@LP and FP@CC were mixed and filtered through a 200 nm polycarbonate porous membrane 10 times on a mini extruder (Avanti Polar Lipids) to obtain biomimetic FP/Vad@CC. , For the control group, the CCM-modified Vad@LP (Vad@CC) was also prepared using a similar protocol.

To prepare anti-TREM2 antibody-modified FP/Vad@CC (FP/Vad@CC-aT2), bifunctional linker Sulfo-SMCC (0.1 mg, 5 mg/mL) was first incubated with the anti-TREM2 antibody (10 μg/mL) in a 1:5 ratio for 60 min, and then mixed with 5 mL of FP/Vad@CC in a 10:1 ratio. The mixture was incubated at 4 °C for 2 h to obtain the biomimetic nanovesicles FP/Vad@CC-aT2. A similar method was applied to prepare CC-aT2 and Vad@CC-aT2.

To investigate the biodistribution of FP/Vad@CC-aT2, ICG-labeled FP/ICG@CC-aT2 (ICG concentration: 0.5 mg/mL) was prepared using the above similar steps, except that ICG was encapsulated into a liposome.

To prepare Rho-labeled nanovesicles, NHS-rhodamine B was selected to mix with FP@CC or FP@CC-aT2 for 2 h, followed by dialysis for 48 h to obtain FP@CC_Rho and FP@CC_Rho-aT2.

Characterization of FP/Vad@CC-aT2

The morphology of FP@CC and FP/Vad@CC was characterized using a high-angle toroidal dark-field phase scanning transmission electron microscope (HAADF-STEM, JEOL JEM-2010). Element mapping was performed using a Themis Z aberration-corrected scanning transmission electron microscope (ac-STEM, FEI, USA). Energy-dispersive X-ray spectroscopy (EDX) spectra was acquired on an EX-250 EDX analyzer (HORIBA, Japan). The DMPO-trapped EPR spectrum was carried out by electron paramagnetic resonance spectroscopy (Bruker EMXnano, Germany). Mass chromatography was carried out at a triple quadrupole mass spectrometer (UPLC-QqQ-MS). The hydrodynamic size, ζ-potential, and stability of nanovesicles were determined using a dynamic light scattering instrument (NanoSight NS500, Malvern Instruments Ltd., UK), and the detailed method of drug loading capacity and the hemolysis assay of FP@CC-aT2 are illustrated in the Supporting Information.

In Vitro Cytotoxicity and Apoptosis Assay

To observe the cell death, LLC cells were seeded in confocal dishes and were subjected to various treatments (PBS, FP@CC-aT2 + Fer-1, FP@CC-aT2, and FP@CC-aT2 + X-ray ([Fe] = 100 μg/mL; [Fer-1] = 5 μg/mL; [X-ray] = 2 Gy)) for 24 h. Then, Calcein-AM (green) and propidium iodide (PI, red) costaining (Beyotime, China) was employed to stain live and dead cells, which was observed under a fluorescence microscope (Leica Thunder Imager 3D Live Cell).

Cell Imaging

Cellular uptake and distribution: LLC cells were seeded into confocal dishes (1 × 10⁵ cells per well) and incubated for 24 h. Then, the cells were coincubated with FP@CC_Rho or Rho/FP@CC_Rho-aT2 for 3 h. Then, cell nuclei and the cytoskeleton were stained with DAPI and Alexa Fluor 488 Phalloidin, respectively. The cells were washed with PBS three times before observation by Confocal laser scanning microscopy (CLSM) (ZEISS LSM 780, Germany).

Intracellular ROS: LLC cells were seeded into confocal dishes (1 × 10⁵ cells per well) and then treated for 6 h with different groups: control, FP@CC-aT2, FP@CC-aT2 + Fer-1, and FP@CC-aT2 + H₂O₂ and FP@CC-aT2 + X-ray + H₂O₂ ([Fe] = 100 μg/mL; [Fer-1] = 5 μg/mL; [H₂O₂] = 10 mM; [X-ray] = 2 Gy). For Fer-1 and X-ray groups, the cells received X-ray or Fer-1 treatment after 5 h of coincubation. The cells were stained with DCFH-DA (Beyotime, China) for 30 min for CLSM observation.

Intracellular lipid peroxide (LPO) generation: The intracellular LPO level was assessed using C11-BODIPY_581/591, and the oxidized products displayed green fluorescence (λ_ex: 485 nm, λ_em: 520 nm). LLC cells were seeded into the confocal dish and subjected to various treatments: control, FP@CC-aT2, FP@CC-aT2 + Fer-1, and FP@CC-aT2 + H₂O₂ and FP@CC-aT2 + X-ray + H₂O₂ ([Fe] = 100 μg/mL; [Fer-1] = 5 μg/mL; [H₂O₂] = 10 mM; [X-ray] = 2 Gy) for 6 h. For the X-ray group, the cells received X-ray irradiation after 5 h of coincubation. Afterward, the cells were stained with C11-BODIPY_581/591 (5 μM) for 20 min. The cells were observed by CLSM after being washed with PBS.

3D Cell Spheroid Formation

The Nunclon Sphera surface minimizes variability and supports the consistent formation of cancer spheroids that simulate the 3D structures of tumor growth and trace the migration and proliferation of macrophages. LLC cells were cultured in complete medium containing 0.75% methylcellulose as a thickening agent using 96-well ultralow attachment plates. After 48 h of incubation, RAW264.7 cells labeled with green emissive CFDA-SE (1 μM) were seeded in a 96-well plate containing LLC spheroids and further cultured for 24 h. Then, the spheroids were treated with various nanoformations (CCM, CCM-aT2, Vad@CCM, FP@CC, FP/Vad@CC, FP@CC-aT2, and FP/Vad@CC-aT2) for 6 h. Finally, the cells were fixed with 4% paraformaldehyde and stained with DAPI before CLSM observation (ZEISS LSM 780, Germany), and the cell labeling method is illustrated in the Supporting Information.

Animal Models and LLC Tumor Models

C57BL/6 mice (5 weeks old; weighing 18–20 g) were purchased from Guangdong Medical Laboratory Animal Center. All animal procedures were conducted according to the guidelines approved by the Care and Use of Laboratory Animals of the Guangdong Medical Laboratory Animal Center. The license number is KY2023–110–01.

For the subcutaneous LLC tumor model, 100 μL of PBS solution containing 5 × 10⁶ suspended LLC cells was subcutaneously injected into the inguinal region of the left hind limb of C57BL/6 mice. When the average tumor size reached 100 mm³, the mice were then randomly separated into five groups, including control, X-ray, FP@CC, FP@CC-aT2, and FP@CC-aT2 + X-ray.

For the postoperative recurrent tumor model, the established subcutaneous tumor-bearing mice were used for the surgical procedure (∼300 mm³). The mice were induced with 2% isoflurane anesthesia with an inhalation anesthesia machine, followed by maintenance at 1.5% to sustain anesthesia. The surgical area was prepared by shaving the fur and disinfecting the surgical site with an iodophor. An approximately 1 cm incision was made on the skin surface over the tumor area, corresponding to the size of the tumor. Tumor resection was performed by using standard blunt dissection. Due to the growth characteristics of subcutaneous tumors, approximately 5–10% of the tumor burden was intentionally left at the surgical margin to simulate a positive margin. , Skin closure was achieved using sterile 6–0 silk sutures. During the surgery and for 2 h postoperatively, buprenorphine was administered at a dose of 0.2 mg/kg as an analgesic medication. When the average tumor size reached 100 mm³, the mice were randomly divided into five groups, including control, Vad@CC, FP@CC-aT2 + X-ray, Vad@CC-aT2 + X-ray, and FP/Vad@CC-aT2 + X-ray.

In Vitro and In Vivo PA/MR Imaging

A 7.0T/310 MRI scanner was used to determine in vitro and in vivo MR imaging performance of nanovesicles (NOVA, Germany). The T1- and T2- relaxation rates of FP@CC-aT2 nanovesicle dispersed in pH6.5 and 7.4 buffer solution with varying Fe concentrations (0, 5, 10, 25, 50, and 100 μg/mL) were examined. For in vivo MR imaging evaluation, 100 μL of FP, FP@CC, or FP/Vad@CC-aT2 ([Fe] = 100 μg/mL) was injected into the LLC cell tumor-bearing mice via the tail vein, and the imaging was detected at different time intervals after injection. The imaging process was conducted continuously over 2 days, with images acquired at 2 h intervals.

The in vitro PA performance of FP@CC-aT2 with various concentrations (6.25, 12.5, 25, 50, 100, and 200 μg/mL) was tested using a Vevo 3100 PAI system (FUJIFILM VisualSonics, Japan). For in vivo PA imaging evaluation, 100 μL of FP, FP@CC, and FP/Vad@CC-aT2 (5.0 mg/kg FePt) was injected into the LLC cell tumor-bearing mice via the tail vein, and the imaging was detected at different time intervals after injection. Wavelengths of 700, 750, 780, 800, 850, and 900 nm were chosen for the calculation of the unmixing in vivo PA experiment. For each ROI, approximately 50 frames were evaluated.

Pharmacokinetics Study

The pharmacokinetics of FP/Vad@CC-aT2 was evaluated by collecting blood samples 5, 10, 20, 60, 120, and 240 min after intravenous injections. After centrifugation, 100 μL of serum was collected and dissolved in 0.4 mL of aqua regia. The samples were heated at 80 °C for 4 h. After digestion, the samples were diluted to 10 mL with 1% HNO₃ and added as an internal standard (10 ppb). Pt concentrations were measured using inductively coupled plasma-mass spectrometry (ICP-MS), and the data were analyzed by a traditional two-compartment pharmacokinetic model to estimate the half-life.

In Vivo Tumor Inhibition of FP@CC-aT2 in the LLC Subcutaneous Lung Tumor

The LLC tumor-bearing C57BL/6 mice (∼100 mm³) were randomly divided into five groups (n = 5) and intravenously injected with PBS, X-ray, FP@CC, FP@CC-aT2, and FP@CC-aT2 + X-ray (5.0 mg/kg FePt). At the first injections, tumor regions were irradiated by X-ray radiation (2 Gy, 1 min) after 3 h of injection of samples. Tumor volumes were calculated following a well-established formula V = (a × b ²)/2, where a and b stand for the longest and shortest diameters of the tumor, respectively. On day 18, mice were sacrificed, and the tumors were harvested and weighed. Heart, liver, spleen, lung, and kidney samples of the mice were collected and weighed.

Additionally, tumors were excised for histological examination after the mice were randomly sacrificed following the treatments. Hematoxylin and eosin (H&E), TdT-mediated dUTP nick end labeling (TUNEL) staining, and Ki67 staining were included in the examination. Following treatment, cleaved caspase-1, glutathione peroxidase 4 (GPX-4), 4-hydroxynonenal (4-HNE), CRT, and HMGB1 were also stained. The serum samples from each group of mice were collected after tumor inoculation on day 18, and the quantities of ALT, TBIL, DBIL, γ-GT, TBA, UREA, CREA, UA, and CK-MB were determined.

In Vivo Tumor Inhibition of FP/Vad@CC-aT2 in the Orthotopic Lung Cancer Model

Lung cancer cells (LCC) were suspended in Matrigel (Corning) and orthotopically implanted into the left lung of immunocompromised mice (n = 3) via surgical exposure, ensuring precise tumor localization. Matrigel was used to mimic the extracellular matrix (ECM) and enhance tumor microenvironment fidelity. Mice were randomized into 5 groups: (1) control, (2) Vad@CC + X-ray, (3) FP@CC-aT2 + X-ray, (4) Vad@CC-aT2 + X-ray, and (5) FP/Vad@CC-aT2 + X-ray (5.0 mg/kg FePt, 3.0 mg/kg Vad; X-ray: 2 Gy, 1 min). Treatment began 7 days after implantation, with treatment monitored for 4 weeks post-therapy. Lungs were harvested at the endpoint for H&E staining to assess tumor morphology, necrosis, and residual/recurrent lesions.

In Vivo Tumor Inhibition of FP/Vad@CC-aT2 in the Postoperative Recurrent Tumor

When the tumor burden of the postoperative recurrent tumor model reaches 100 mm³ again, the mice will be randomly divided into 5 groups (n = 5) and will receive intravenous injections via the tail vein of PBS, Vad@CC, FP@CC-aT2, FP/Vad@CC, and FP/Vad@CC-aT2 (5.0 mg/kg FePt, 3.0 mg/kg Vad). The sample will be administered every 3 days for a total of 3 doses, with X-ray irradiation (2 Gy, 1 min) after 4 h of injection, coinciding with the first doses. The survival and weight changes of the mice in each group will be recorded according to standard methods. The tumor volume will be measured, and the volume changes will be charted every 3 days according to the same standard method. On day 18, tumors were excised for histological examination after the mice were randomly sacrificed following the treatments. Hematoxylin and eosin (H&E), TdT-mediated dUTP nick end labeling (TUNEL) staining, and Ki67 staining were included in the examination, and detailed methods are illustrated in the Supporting Information.

Statistical Analysis

The data presented are expressed as mean ± SEM from five independent experiments, each performed with three replicates. Statistical analyses were conducted using GraphPad Prism 9.3.0 software (GraphPad Software Inc.) and FlowJo 10.0 (FlowJo Software Inc.). Differences between groups were assessed using two-tailed Student’s t tests and one-way analysis of variance. Statistical significance was denoted as *P < 0.05 **P < 0.01 ***P < 0.001 and ****P < 0.0001, indicating a significant difference between groups.

Glossary

Abbreviations

TREM2

triggering receptor expressed on myeloid cells 2

cGAS-STING

cyclic GMP-AMP synthase-stimulator of interferon genes

TME

tumor microenvironment

TAMs

tumor-associated macrophages

Vad

Vadimezan

NSCLC

non-small cell lung cancer

ICIs

immune checkpoint inhibitors

CTLs

cytotoxic T lymphocytes

IL-10

interleukin 10

ROS

reactive oxygen species

LPO

lipid peroxidation

ICD

immunogenic cell death

DC

dendritic cell

XDT

X-ray dynamic therapy

PTT

photothermal therapy

CCM

cancer cell membrane

IFN-I

type I interferon

PAI

photoacoustic imaging

MRI

magnetic resonance imaging

Fer-1

ferrostatin-1

GPX-4

glutathione peroxidase 4

CLSM

confocal laser scanning microscopy

4-HNE

4-hydroxynonenal

H&E

hematoxylin and eosin

CRT

calreticulin

HMGB1

high-mobility group box 1

DEGs

differentially expressed genes

GO

gene ontology

MF

molecular function

BP

biological processes

KEGG

Kyoto encyclopedia of genes and genomes

GSEA

gene set enrichment analysis

PPI

protein–protein Interaction

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsnano.5c10375.

• Detailed methods of drug loading capacity, hemolysis assay, cell culture, Western blotting, cell labeling, phagocytosis quantification, flow cytometry to evaluate immune cells, transcriptomic mRNA sequencing analysis, and ELISA analysis to evaluate cytokines; nanovesicle characterization (TEM, membrane protein by Western blot, elemental mapping of Fe/Pt) (Figures S1 and S2); cytotoxicity measured by the hemolysis assay (Figure S3); in vitro evaluation of cell uptake efficacy against LLC cells after different treatment groups (Figure S4); ROS levels in different treatment groups of LLC cells measured by flow cytometry (Figure S5); in vitro evaluation of therapeutic efficacy against LLC cells after different treatment groups (Figure S6); flow cytometric analysis of apoptosis in various treatment groups (Figure S7); macrophage–tumor cell phagocytosis imaging and quantification (confocal microscopy, flow cytometry) (Figure S8); in vivo fluorescence imaging and biodistribution (ICG labeling, Pt in serum) (Figures S9 and S10); histology and biosafety (H&E of organs; cardiac, hepatic, renal function) (Figures S11 and S12); transcriptomics (RNA-seq, DEGs, GO/KEGG/GSEA, PPI network, heatmaps) (Figures S13–S15); tumor growth inhibition in vivo (body weight, tumor volume, excised tumor images) (Figures S16–S18); immune profiling by flow cytometry (DCs, macrophages, T cell subsets; gating strategies) (Figures S19 and S20); IHC analysis of Foxp3 expression in recurrent tumor tissues from control and FP/Vad@CC-aT2 + X-ray mice and reference (Figure S21) (PDF)

#.

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. B.X., H.Q., and H.W. contributed equally.

The authors declare no competing financial interest.