Chemoradiotherapy‐Integrated Tumor Cell‐Derived Microparticles Mediate Tumor Eradication in Malignant Pleural Effusion

ABSTRACT

Malignant pleural effusion (MPE) portends poor prognosis in advanced cancer. Strategies that integrate potent tumor killing with immunosuppressive microenvironment reprogramming are crucial for the treatment of MPE. Studies show that irradiated tumor cell‐derived microparticles (RT‐MPs) possess natural tumor‐targeting cytotoxicity and innate immune activation properties. To further boost the tumoricidal effects of RT‐MPs, we developed innovative chemoradiotherapy‐integrated tumor cell‐derived microparticles (CR‐MPs) by loading RT‐MPs with chemotherapeutic agents, including methotrexate (MTX), monomethyl auristatin E (MMAE), or doxorubicin (DOX). CR‐MPs exhibited superior tumoricidal activity over both RT‐MPs and free drugs against a range of tumors. Specifically, MTX‐loaded CR‐MPs (CR‐MPs@MTX) triggered mitochondrial oxidative stress and immunogenic ferroptosis in tumor cells, while directly reprogramming macrophages toward the M1 phenotype and stimulating dendritic cells via cGAS‐STING/NF‐κB pathway activation. In murine MPE models, CR‐MPs effectively suppressed tumor progression, extended survival, and demonstrated favorable biosafety. When combined with immunotherapy, this approach achieved a cure rate of up to 70%, induced durable immunological memory, and retained efficacy against chemotherapy‐resistant tumors. This study establishes CR‐MPs as a novel platform with robust therapeutic efficacy against MPE, highlighting their translational potential as a precision concurrent chemoradiotherapy strategy for MPE management in clinical settings.

Highlights:

• •CR‐MPs functionally simulate concurrent chemoradiotherapy.
• •CR‐MPs trigger immunogenic ferroptosis and activate innate immunity.
• •Synergy with immunotherapy drives complete tumor eradication.

1. Introduction

Malignant pleural effusion (MPE), a common clinical manifestation of primary pleural tumors or metastatic invasion from other malignancies, is associated with a poor prognosis and diminished quality of life (Clive et al. 2014; Gonnelli et al. 2024). Lung cancer is a leading cause of metastatic MPE (Tian et al. 2021). Approximately 50% of patients with non‐small cell lung cancer (NSCLC) eventually develop pleural metastasis, with a median survival of less than seven months (Taghizadeh et al. 2017). Metastatic pleural disease lacks effective definitive treatment, with standard approaches such as thoracentesis, pleural catheter drainage, and surgical pleurodesis providing only palliative benefit (Bibby et al. 2018; Feller‐Kopman et al. 2018). Intrapleural chemotherapy represents an etiologic strategy for MPE management by directly killing tumor cells and stimulating chemical pleurodesis. However, its therapeutic benefits are often compromised by several factors: (1) inadequate drug penetration and accumulation within tumor tissues, (2) resistance to chemotherapeutic agents, (3) chemotherapy‐induced adverse effects, (4) tumor microenvironment (TME) rich in both suppressive myeloid immune cells and inhibitory cytokines, among other factors (Huang et al. 2021; Murthy et al. 2019; Sakaguchi et al. 2017). Therefore, strategies that co‐engage both tumor cells and suppressive immune cells, ensuring tumor‐specific targeting and cytotoxicity while simultaneously reversing immunosuppression are urgently needed to improve current MPE management.

Radiotherapy (RT) is a cornerstone of cancer treatment, administered to approximately 50% of cancer patients (Chandra et al. 2021). Despite its direct cytotoxic effects, emerging evidence suggests its potent immunostimulatory effects that are often interpreted as an “in situ tumor vaccine” (Wang et al. 2024). Irradiation induces immunogenic cell death (ICD) of tumor cells, characterized by the expression of damage‐associated molecular patterns (DAMPs), release of inflammatory cytokines, and activation of antigen‐presenting cells (APCs) to cross‐prime CD8⁺ T cells (S. Guo et al. 2023a; Rodriguez‐Ruiz et al. 2020; Zhu et al. 2021). This cascade of immune responses facilitates systemic tumor control, highlighting its therapeutic potential in metastatic disease. However, RT is not a viable option for a subset of conditions, notably MPE (Feller‐Kopman et al. 2018). Localized radiotherapy fails to effectively control diffusely distributed metastatic lesions within the pleural cavity. Moreover, escalation of RT dose poses substantial risks to adjacent organs such as the heart and lungs. These limitations underscore the necessity for alternative approaches to harness radiotherapeutic effects.

Cellular microparticles (MPs), a subclass of extracellular vesicles ranging from 100 to 1000 nm in diameter, are produced by the direct shedding of the plasma membrane upon various stimuli (van Niel et al. 2018). They shuttle bioactive cargo, including nucleic acids, proteins, and lipids, between cells to mediate intercellular communication in physiological and pathological processes (Marar et al. 2021). Our prior research identified irradiated tumor cell‐derived microparticles (RT‐MPs) as key mediators of the radiation‐induced bystander effect (RIBE) (Wan et al. 2020). These microparticles recapitulate the immunostimulatory effects inherent to RT by inducing ICD and reprogramming the immunosuppressive tumor‐associated macrophages into a pro‐inflammatory state (Wan et al. 2020). In our previous study, administration of RT‐MPs extended the survival of MPE mice, yet the benefit was limited in duration. Thus, further enhancement of the anti‐tumor capacity of RT‐MPs is imperative. Originating from cancerous cells, RT‐MPs inherit the surface composition of their parent cells, conferring them with natural cancer affinity and prolonged tumor retention, which qualifies them as an optimal drug delivery system (DDS) (Luk and Zhang 2015). To this end, we exploited the inherent advantages of RT‐MPs by loading them with chemotherapeutic drugs, formulating a novel combination therapy that simulates concurrent chemoradiotherapy to surmount resistance in MPE.

Here, we developed a novel platform based on chemoradiotherapy‐integrated tumor cell‐derived microparticles (CR‐MPs) by encapsulating multiple chemotherapeutic agents within RT‐MPs. CR‐MPs exhibit enhanced tumoricidal activity and comparable immunostimulatory effects compared with RT‐MPs, showing favorable therapeutic efficacy and safety profiles in preclinical MPE models. Notably, when combined with immune checkpoint inhibitors, the combination therapy achieved about 70% complete remission and long‐term immune memory in MPE‐bearing mice. By delivering the combined effects of chemoradiotherapy, CR‐MPs provide a practical paradigm for the integration of “radio‐chemo‐immunotherapy”, holding promise for MPE clinical practice.

2. Materials and Methods

2.1. Chemical Reagents

MTX (MB1156‐1) and DOX (MB1087) were purchased from Meilunbio. MMAE (HY‐15162), Mito‐tempo (HY‐112879) and Azalamellarin N (HY‐162150) were purchased from MCE. Ferrostatin‐1 (S7243), Emricasan (S7775), and Necrostatin‐1 (S8037) were purchased from Selleck. Recombinant Murine GM‐CSF (713704) and M‐CSF (576402) were obtained from Biolegend. Recombinant anti‐mouse PD‐1 mAb (D265A) were obtained from STARTER.

2.2. Cell Lines and Cell Culture

All cell lines were obtained from the American Type Culture Collection (ATCC) and maintained under standard conditions. NCI‐H1299 and NCI‐H460 were propagated in RPMI‐1640 medium (Gibco), while LLC, Hela, T98G, PANC‐1, and KPC cells were cultured in Dulbecco's Modified Eagle Medium (DMEM; Gibco). Both media were supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% penicillin/streptomycin (BL505A, Biosharp). Cells were incubated at 37°C in a humidified 5% CO₂ atmosphere and routinely screened for mycoplasma contamination.

For the LLC‐Luc stable cell line, parental LLC cells were transduced with a lentiviral vector encoding firefly luciferase and selected with puromycin (2 µg/mL, BL528A, Biosharp) for 7 days. Luciferase expression was validated by bioluminescence imaging (Bruker) after incubation with D‐luciferin (abs42075819, Absin).

2.3. Generation and Isolation of RT‐MPs and CR‐MPs

Tumor cells were seeded in 10‐cm culture dishes and grown to 70% confluency. Cells were subsequently irradiated with 10 Gy using 6‐MV X‐rays (600 MU/min, Trilogy System Linear Accelerator, Varian Medical Systems) and replaced with 25 mL of complete medium. After 48 h, the conditioned medium was collected and sequentially centrifuged at 1,000 × g for 10 min (to remove cells debris) and 14,000 × g for 2 min (to pellet apoptosis body), both at 4°C. The supernatant was further centrifuged at 14,000 × g for 1 h to isolate RT‐MPs, which were then washed and resuspended in PBS.

For CR‐MPs generation, tumor cells were irradiated as described above, and the medium was replaced with 25 mL of fresh complete medium. After 24 h, chemotherapeutic drugs (MTX, MMAE, and DOX) were added at specified concentrations, and those irradiated cells were incubated for an additional 24 h. The conditioned medium was collected and centrifuged as described above (sequential centrifugation at 1000 × g, 14,000 × g, and final isolation at 14,000 × g). CR‐MPs were washed to remove free drugs and resuspended in PBS. CR‐MPs for subsequent experiments were prepared by treating irradiated cells with 1 mg/mL MTX, 500 µg/mL MMAE, or 250 µg/mL DOX. RT‐MPs and CR‐MPs were quantified identically by protein content via BCA assay (G2026‐200T, Servicebio).

2.4. Quantification of Drug Payload

The quantification of MTX within RT‐MPs was conducted via high‐performance liquid chromatography (HPLC). The MTX standard was sourced from meilunbio. Approximately 200 µg of MTX‐loaded RT‐MPs were reconstituted in 200 µL of a lysis buffer (methanol:acetonitrile = 2:1, v/v), followed by centrifugation at 12,000 × g for 10 min to precipitate proteins, with the protein‐free supernatant analyzed using a 1260 Infinity II HPLC system (Agilent) equipped with a C18 column (Nucleosil, 250 × 4.6 mm, Alltech). The eluent consisted of 90% PBS and 10% acetonitrile (pH 7.4), delivered at a flow rate of 1 mL/min. Separation was performed under isothermal conditions at 40°C, and MTX was quantified by monitoring UV absorbance at 306 nm. Chromatographic data were acquired and integrated using OpenLAB CDS software (Agilent).

The quantification of MMAE in RT‐MPs was carried out using Liquid Chromatography‐Tandem Mass Spectrometry (LC‐MS/MS). The MMAE standard was sourced from MCE. Sample pretreatment involves lysis and protein precipitation, as mentioned above. Separation was achieved on an Ultimate 3000 UHPLC system (Thermo Fisher) with a C18 column (Hypersil GOLD, 50 × 2.1 mm, Thermo Fisher). The column was maintained at 40°C. The mobile phases consisted of 0.1% formic acid in water (A) and acetonitrile (B). A gradient elution program was applied as follows: 10% B at 0 min, 30% B at 3 min, 95% B at 5 min, returning to 10% B by 9.1 min and maintaining until 12 min. The flow rate was set to 0.3 mL·min^{− 1}, and 5 µL of each sample was injected. Mass spectrometric detection was conducted on a Triple Quadrupole instrument (Thermo Fisher) operated in positive ionization mode with the following settings: heater temperature, 350°C; sheath gas flow, 35; auxiliary gas, 10; sweep gas flow, 0; and spray voltage, 3.5 kV. Data acquisition and processing were performed using Xcalibur software (v3.0, Thermo Fisher), with compound quantification based on integrated peak areas from the extracted ion chromatograms.

The concentration of DOX in RT‐MPs was measured by microplate reader (Agilent). The DOX standard was sourced from meilunbio. Sample pretreatment involves lysis and protein precipitation, as mentioned above. The fluorescence of the pretreated samples and DOX standard was measured using excitation and emission wavelengths of 480 nm and 560 nm, respectively.

2.5. Transmission Electron Microscopy and Dynamic Light Scattering (DLS)

Both RT‐MPs and CR‐MPs were negatively stained with 2% phosphotungstic acid for 5 min before deposition on copper grids. The morphology of RT‐MPs and CR‐MPs was examined using TEM (HT7800/HT7700, Hitachi). The size distributions of both RT‐MPs and CR‐MPs were characterized using DLS (ZetaSizer Nano‐ZS, Malvern Instruments).

2.6. Western Blot and Elisa

Protein samples were resolved by SDS‐PAGE and then transferred to PVDF membranes. After blocking with 5% skim milk in TBST (0.1% Tween‐20) for 1 h at room temperature, membranes were probed with primary antibodies overnight at 4°C. Following secondary antibody incubation, protein bands were visualized using ECL Ultra (MA0186, meilunbio). Mouse HMGB1 (EM0382, Finetest) and human HMGB1 (E‐EL‐H1554, Elabscience) produced in the supernatants were evaluated by ELISA assays according to the manufacturer's instructions.

2.7. In Vitro Cytotoxicity Assays

For evaluating cell viability, cells were seeded in 96‐well plates at 5000 cells/well and cultured for 24 h prior to treatment. Following exposure to varying concentrations of RT‐MPs, CR‐MPs, or chemical agents for 48 h, viability was assessed using a CCK‐8 kit (BS350B, Biosharp).

For analyzing cell death, cells (5×10⁴ cells/well) were plated into 12‐well plates with 1 mL complete medium and cultured for 24 h. Following a 48‐h treatment with CR‐MPs, RT‐MPs, or chemical agents, cells were harvested, stained with Annexin V/7‐AAD (640920/640934, Biolegend) according to the manufacturer's protocol, and incubated at 4°C for 30 min before flow cytometric analysis.

2.8. RNA‐Sequencing

Total RNA was isolated using TRIzol Reagent (Invitrogen), followed by DNase I digestion to remove genomic DNA. RNA quality and concentration were assessed using a Qubit 3.0 Fluorometer with the Qubit RNA Broad Range Assay kit (Life Technologies). For library preparation, 2 µg of total RNA was processed using the KC Stranded mRNA Library Prep Kit for Illumina (Wuhan Seqhealth Co., Ltd., China) according to the manufacturer's protocol. Raw sequencing reads were quality‐filtered using Trimmomatic (v0.36) to remove low‐quality sequences.

Differential gene expression analysis was conducted using edgeR (v3.12.1), with significance thresholds set at a |log2 fold‐change| ≥ 1 and adjusted p‐value < 0.05. Functional enrichment analysis, including GO and KEGG pathway analysis, was performed using KOBAS (v2.1.1) with a significance cutoff of p < 0.05.

2.9. Evaluation of ROS

Cells (5×10⁴ cells/well) were seeded in 12‐well plates and treated with CR‐MPs, RT‐MPs, or chemical reagents for 24 h. Following treatment, cells were incubated with H2DCFDA (general ROS, S0033S, Beyotime), MitoSOX (mitochondrial superoxide, 720827ES, Yeasen), or C11‐BODIPY (lipid peroxidation, D3861, Thermo Fisher) probes in serum‐free medium for 30 min at 37°C. After washing three times with serum‐free medium, cells were resuspended in PBS and analyzed by flow cytometry.

2.10. Immunofluorescence

Cells were cultured on glass coverslips in 24‐well plates. Following treatment, samples were fixed with 4% paraformaldehyde for 10 min, permeabilized with 0.2% Triton X‐100 for 15 min, and blocked with 5% BSA for 1 h. Primary antibody incubation was performed overnight at 4°C, followed by fluorescence‐conjugated secondary antibody staining at 37°C for 1 h. Nuclei were stained with Hoechst (G1012, Servicebio,), and images were acquired using a laser scanning confocal microscope.

2.11. Generation of BMDMs and BMDCs

Bone marrow‐derived macrophages (BMDMs) and dendritic cells (BMDCs) were isolated from femurs of C57BL/6 mice. Following red blood cell lysis (BL503B, Biosharp), cells were cultured in RPMI 1640 medium containing 10% FBS and either 20 ng/mL M‐CSF (for BMDMs) or 20 ng/mL GM‐CSF (for BMDCs). Medium was refreshed every 48 h. Both BMDM and BMDCs were harvested on day 7 for subsequent assays.

2.12. In Vitro Uptake Assay

For detecting colocalization of cell and CR‐MPs, cells were seeded in glass coverslips in 24‐well plates and incubated with PKH‐26–labeled (PKH26GL‐1KT, Sigma‐Aldrich) CR‐MPs for 4, 12, and 24 h. After three times wash by PBS, cells were stained with carboxyfluorescein diacetate succinimidyl ester (CFSE, Y229292‐10 mg, Beyotime) for 5 min, washed with serum, fixed with 4% paraformaldehyde for 15 min and washed by PBS again. Colocalization was assessed by laser scanning confocal microscopy.

For quantification of cellular uptake, cells in 12‐well plates were treated identically, washed with PBS for three times, stained with CFSE, fixed, washed, and resuspended in PBS for flow cytometric analysis.

2.13. In Vivo Cellular Uptake Assay

To assess the cellular uptake of CR‐MPs@MTX within the MPE microenvironment, MPE‐bearing mice received a single intrapleural injection of PKH26‐labeled CR‐MPs@MTX (approximately 100 µg of MP protein) on day 8 post‐tumor inoculation. 24 h post‐injection, mice were euthanized. Pleural effusion was collected, and pleural tumor nodules were dissected. Cells from the effusion were pelleted by centrifugation and resuspended. Tumor nodules were mechanically dissociated to prepare single‐cell suspensions. Cells from both sources were filtered through a 70‐µm strainer, subjected to red blood cell lysis, and stained for viability using the Zombie Violet Fixable Viability Kit. Subsequently, cells were stained with fluorochrome‐conjugated antibodies against specific surface markers. The uptake of PKH26‐labeled CR‐MPs was quantified by flow cytometry as the percentage of PKH26‐positive cells within defined populations: CD45⁻ tumor cells, CD45⁺CD11b⁺F4/80⁺ macrophages, CD45⁺CD11b⁺CD11c⁺ dendritic cells, CD45⁺CD11b⁺LY6G⁺ neutrophils, and CD45⁺CD3⁺ T cells.

2.14. Real‐Time Quantitative PCR

Total RNA was isolated from samples using TRIzol Reagent (X2372, Invitrogen), with concentration measured by NanoDrop ND‐1000 (Thermo Fisher). RNA was reverse‐transcribed to cDNA using HiScript III RT SuperMix (#R323, Vazyme). Quantitative PCR was performed on a StepOne system with ChamQ SYBR qPCR Master Mix (Q331‐02, Vazyme). Gene expression was analyzed via the comparative Ct method. Primer sequences are listed in Table S1 .

2.15. Animal Studies and Evaluation of Therapeutic Effect

C57BL/6J mice (7–8 weeks old) were obtained from SLB Laboratory Animal Co. (Hunan, China) and maintained in individually ventilated cages. All procedures were performed in compliance with protocols approved by the Hubei Provincial Animal Care and Use Committee and followed the ethical guidelines of the Animal Experimentation Ethics Committee of Huazhong University of Science and Technology (HUST, Wuhan, China, Approval No. 4758). MPE murine models were generated by intrapleural injection of 2×10⁵ LLC‐Luc cells in 50 µL PBS through the 10th/11th intercostal space at the midaxillary line under 1% pentobarbital sodium anesthesia. Following randomization, mice were allocated to receive one of the following treatments: Control animals received intrapleural injections of 100 µL 1× PBS. In the RT‐MPs and CR‐MPs treatment groups, the respective microparticles were suspended in 100 µL 1× PBS at a concentration of 2 µg/µL for intrapleural administration. For MTX treatment, two dosing regimens were employed: an equivalent dose (10 µg/mL matching the drug content in CR‐MPs) and a standard experimental dose (500 µg/mL as recommended for animal studies), both prepared in 100 µL 1× PBS for intrapleural delivery. For PD‐1 treatment, mice received anti‐PD‐1 monoclonal antibody at 10 mg/kg via intraperitoneal injection. Survival condition of mice from each group was documented during observation, and tumor burden was evaluated through in vivo bioluminescence imaging.

2.16. In Vivo Bioluminescence Imaging

For tumor burden assessment, MPE‐bearing mice were anesthetized with 1% pentobarbital sodium and administered 150 mg/kg firefly luciferin via intraperitoneal injection. Imaging was performed using the Bruker in Vivo MS FX PRO Imager following a 10‐min incubation period. Three distinct imaging modalities were acquired: white light radiographs (0.175 s exposure), X‐ray photographs (30 s exposure), and luminescence images (3 min exposure) to comprehensively evaluate tumor progression. To evaluate the retention of CR‐MPs, we intrapleurally injected DiR‐labeled (HY‐D1048, MCE) microparticles, and detected fluorescence using the Bruker In Vivo MS FX PRO Imager (750 nm excitation and 780 nm emission). Imaging timepoints were established at 2 h and 24 h post‐injection.

2.17. In Vivo Flowmetry Analysis

For comprehensive immune cell profiling in the TME, pleural lavage fluid and tumor nodules were harvested from MPE‐bearing mice. Single‐cell suspensions were prepared by mechanical dissociation followed by centrifugation (500 g, 5 min). After erythrocyte lysis, cells were washed with PBS and resuspended in PBS. Viability assessment was performed using Zombie Violet (423114) or Zombie NIR (423106) Fixable Viability Kits prior to surface marker staining. For surface marker staining, cells were stained with CD45 (103114), CD3 (100306), CD4 (100408), CD8a (100752), F4/80 (111703), CD11c (117305), Gr‐1 (108412), CD80 (104706), and CD86 (159216). Antibodies were used at manufacturer's recommended concentrations with 30 min incubation at 4°C. For intracellular marker staining, cells were stimulated for 5 h at 37°C/5% CO2 with monensin sodium salt (1.5 µg/mL), ionomycin calcium salt (1 µg/mL), and PMA (100 ng/mL) before fixation/permeabilization. This enabled detection of effector molecules IFN‐γ (505808) and granzyme B (396414) in T lymphocyte populations.

For analyzing changes in memory T lymphocytes, the spleen and lymph nodes of MPE mice were collected, mechanically dissociated, and lysed to remove erythrocytes. Single cell suspensions prepared from the spleen and lymph nodes were stained with a panel of flow cytometry antibodies, including anti‐mouse Zombie NIR Fixable Viability Kit (423106) for live/dead discrimination, followed by surface markers for immune cell characterization: CD45 (103114), CD3ε (100306), CD4 (100408) and CD8a (100752), as well as CD44 (103012) and CD62L (161204) to assess memory T cell phenotypes. All antibodies were purchased from Biolegend, used at manufacturer's recommended concentrations and incubated at 4°C for 30 min.

2.18. Statistical Analysis

Data statistical analyses were performed with GraphPad Prism 10.1.2. Quantitative results are expressed as mean ± s.d. for in vitro experiments, or as mean ± s.e.m. for in vivo experiments. For comparisons between two or more groups, the following tests were applied: two‐tailed Student's t test, one‐way ANOVA with Tukey's multiple comparisons, and two‐way ANOVA with Tukey's post hoc test. Survival analysis was conducted using the log‐rank test. A P value of less than 0.05 was considered statistically significant.

3. Results

3.1. Generation and Characterization of CR‐MPs

Tumor‐derived microparticles (T‐MPs) naturally contain tumor‐associated antigens and immunostimulatory molecules that are capable of activating APCs, conferring a unique advantage for T‐MP‐based DDS (Sun et al. 2023). However, different induction methods alter the compositions of T‐MPs, resulting in variable APCs activation and divergent immune responses. To identify the optimal DDS, we first evaluated the immunogenicity of various T‐MPs. Microparticles were derived from normally cultured tumor cells (NC‐MPs), ultraviolet‐treated tumor cells (UV‐MPs), or irradiated tumor cells (RT‐MPs) Figure S1A–B. Gene Set Enrichment Analysis (GSEA) of protein profiles revealed that RT‐MPs exhibited significant positive enrichment in immune activation‐associated pathways, compared with NC‐MPs and UV‐MPs Figure S1C–F. Relative to NC‐MPs and UV‐MPs, RT‐MPs exhibited higher levels of multiple immune‐related signaling molecules Figure S1G–H. Western blot analysis confirmed the elevated expression of calreticulin (CRT), a canonical ICD marker identified in the proteomic data Figure S1I. To further compare their immunostimulatory function, we incubated NC‐MPs, UV‐MPs, and RT‐MPs with two classical APCs, bone marrow‐derived dendritic cells and macrophages (BMDCs and BMDMs). RT‐MPs most significantly promoted the activation of both APCs Figure 1A–B. Collectively, these results indicate that RT‐MPs, with the strongest immunogenicity, represent an ideal T‐MP‐based DDS.

FIGURE 1.

Generation and characterization of CR‐MPs.

(A–B) Mean fluorescence intensity (MFI) of CD80 on BMDCs (A) and BMDMs (B) treated with NC‐MPs, UV‐MPs, and RT‐MPs. (C) Schematic diagram of therapeutic intervention for intrapleural RT‐MPs administration at different timepoints (initiated at day 5 or day 9, 200 µg RT‐MPs per treatment) in murine MPE models. (D) Survival analysis of MPE‐bearing mice treated with RT‐MPs at varied intervention timing indicated by (C) (n = 8 per group). (E) In vivo bioluminescence images of pleural tumor burden before and after RT‐MPs treatment. (F) Schematic diagram of the production pipeline for CR‐MPs. Tumor cells were first irradiated by a single dose of 10 Gy X‐ray and incubated with high concentrations of chemotherapeutic drugs, followed by gradient centrifugation for isolation. (G–H) Quantification of MTX payload in LLC cell‐derived RT‐MPs. (G) Representative HPLC chromatograms and (H) quantified concentration of MTX loaded after incubation with the indicated drug concentrations. (I–J) Quantification of MMAE payload in LLC cell‐derived RT‐MPs. (I) Representative LC/MS chromatograms and (J) quantified concentration of MMAE loaded after incubation with the indicated drug concentrations. (K–M) Representative TEM images of (K) RT‐MPs, (L) CR‐MPs@MTX, and (M) CR‐MPs@MMAE. Scale bar = 100 nm. (N) Size distribution of RT‐MPs, CR‐MPs@MTX, and CR‐MPs@MMAE. (O) Statistical analysis of particle diameters for the RT‐MPs, CR‐MPs@MTX, and CR‐MPs@MMAE. (P) Zeta potential of RT‐MPs, CR‐MPs@MTX, and CR‐MPs@MMAE. (Q) Western blot analysis of surface markers (CD63, TSG101) and negative marker (Calnexin) of RT‐MPs, CR‐MPs@MTX, and CR‐MPs@MMAE. All data are presented as mean ± s.d. One‐way ANOVA with Tukey's multiple comparisons test was performed for (A, B, H, J, O, P). Log‐rank test was performed for (D). *P < 0.05, **P < 0.01, and ***P < 0.001; ns, no significance.

Our previous findings demonstrated that RT‐MPs prolonged the survival of MPE‐bearing mice (Wan et al. 2020). To explore whether RT‐MPs remain effective under conditions of higher tumor burden, we implemented a delayed treatment regimen Figure 1C. However, when intrapleural administration of RT‐MPs was postponed by 4 days, corresponding to a stage of increased tumor burden, the therapeutic efficacy of RT‐MPs was reduced Figure 1D–E and Figure S2A. To further enhance tumor control against MPE, we developed a strategy to integrate chemoradiotherapy through the incorporation of chemotherapeutic drugs into RT‐MPs. CR‐MPs were generated by incubating irradiated tumor cells with high‐concentration chemo‐agents, followed by isolation via gradient centrifugation Figure 1F. RT‐MPs could efficiently encapsulate multiple cytotoxic agents, including MTX, DOX, and MMAE, the cytotoxic payload of antibody‐drug conjugate (ADC) drugs. Microparticles were quantified by protein contents and various drug payloads were determined through standard curves. The drug loading capacity positively correlated with the concentration of drugs during tumor cell incubation. RT‐MPs derived from murine lung cancer cells LLC were able to carry up to 5 µg MTX (upon incubation with 1 mg/ml MTX) or 3 µg MMAE (upon incubation with 0.5 mg/ml MMAE) per milligram of microparticle protein Figure 1G–J and Figure S2B–E. Human lung cancer cell H1299‐derived RT‐MPs had an MTX loading capacity of 2‐2.5 µg (with 1 mg/ml MTX) and a DOX loading capacity of 25 µg (with 0.25 mg/ml DOX) per 100 µg of microparticle protein Figure S2F–H. Given the above loading capacities, CR‐MPs generated under the corresponding drug incubation conditions were utilized for all subsequent studies.

We next characterized CR‐MPs in terms of morphology, structure, physicochemical properties, and protein content. Transmission electron microscopy (TEM) verified that drug loading did not compromise the structural integrity of RT‐MPs membranes Figure 1K–M and Figure S2I–K. Upon drug loading, RT‐MPs exhibited an increase in particle size to varying extents, as revealed by nanoparticle tracking analysis (NTA) Figure 1N–O and Figure S2L–M. Zeta potential analysis indicated that CR‐MPs retained their native negative surface charge Figure 1P and Figure S2N. The presence of CD63 and TSG101 further confirmed the extracellular vesicle identity of drug‐loaded RT‐MPs Figure 1Q and Figure S2O. These results collectively demonstrate the potential of RT‐MPs as effective carriers for various therapeutic molecules.

3.2. CR‐MPs Enhance Tumor Cytotoxicity by Promoting Ferroptosis

To explore the potential antitumor activity of CR‐MPs, we first confirmed that tumor cells could efficiently take up CR‐MPs upon direct contact. Flow cytometry revealed time‐dependent uptake of CR‐MPs@MTX by homologous tumor cells Figure 2A. Immunofluorescence staining further demonstrated the accumulation of internalized CR‐MPs@MTX within H1299 cells Figure 2B. Moreover, efficient internalization was observed across heterologous tumor cell lines Figure S3A–B.

FIGURE 2.

CR‐MPs enhance tumor cytotoxicity by promoting ferroptosis.

(A) Uptake of PKH26‐labeled, H1299‐derived CR‐MPs@MTX by H1299 cells was assessed by flow cytometry at 4, 12, and 24 h. (B) Representative immunofluorescence images of H1299 cells (green) after 24‐hour uptake of PKH26‐labeled CR‐MPs@MTX (red). Scale bar = 100 µm. (C) CCK‐8 viability assay evaluated cytotoxic effects of LLC‐derived CR‐MPs@MTX versus RT‐MPs against LLC cells. (D) CCK‐8 viability assay evaluated cytotoxic effects of LLC‐derived CR‐MPs@MTX versus RT‐MPs against Hela, H460, and T98G cells. (E) Heatmap of cytotoxicity profiles of LLC‐derived CR‐MPs@MTX against LLC, H1299, Panc, and KPC cells (CCK‐8 assay). (F) 7‐AAD‐based cytotoxicity assessment of LLC‐derived CR‐MPs@MTX, RT‐MPs, and equivalent free MTX against LLC cells. (G) Annexin V/7‐AAD‐based cytotoxicity assessment of LLC‐derived CR‐MPs@MTX, RT‐MPs, and equivalent free MTX against H460 cells. (H) Dose‐response curves showed sensitivity to MTX in LLC‐MTXR versus parental LLC cells. (I) CCK‐8 viability assay of LLC and LLC‐MTXR cells treated with LLC‐derived CR‐MPs@MTX versus RT‐MPs. (J) KEGG pathway enrichment analysis of differentially expressed genes (DEGs) in LLC cells 24‐hour post CR‐MPs@MTX treatment. (K) Relative expression of ferroptosis driver genes in H1299 and LLC cells treated with CR‐MPs@MTX based on RNA‐sequencing. (L) Representative C11‐BODIPY immunofluorescence images in H1299 cells after 24‐hour treatment with CR‐MPs@MTX, RT‐MPs, or equivalent free MTX (Non‐oxidated state, red; oxidated state, green). Scale bar = 20 µm. (M) Flow cytometric quantification of lipid ROS (C11‐BODIPY) in LLC cells treated with LLC‐derived CR‐MPs@MTX, RT‐MPs, or equivalent free MTX. (N) Flow cytometric quantification of lipid ROS (C11‐BODIPY) in H1299 cells treated with H1299‐derived CR‐MPs@MTX, RT‐MPs, or equivalent free MTX. (O) Partial rescue of CR‐MPs@MTX cytotoxicity by ferroptosis inhibitor Ferrostatin‐1 (Fer‐1, 10 µM). All data are presented as mean ± s.d. Two‐way ANOVA with Tukey's multiple comparisons test was performed for (C, D, I, O). One‐way ANOVA with Tukey's multiple comparisons test was performed for (F, G, M, N). *P < 0.05, **P < 0.01, and ***P < 0.001; ns, no significance.

We next investigated the tumoricidal effects of CR‐MPs on tumor cells. CR‐MPs@MTX showed significant concentration‐dependent killing effects on both homologous and heterologous tumor cells, outperforming equivalent RT‐MPs carrier, as revealed by CCK‐8 assays Figure 2C–D and Figure S4A–B. The cytotoxicity extended to other types of tumor cells, including H1299, Panc, and KPC cells, with comparable dose‐response effects Figure 2E. CR‐MPs@MMAE and CR‐MPs@DOX similarly exhibited significantly improved cytotoxic effects compared with RT‐MPs across all doses Figure S4C–D. The concentration resulting in approximately 50% cell death was utilized in subsequent functional assays. Moreover, CR‐MPs@MTX surpassed free MTX at equivalent doses in killing efficiency on both homologous and heterologous tumor cells Figure 2F–G and Figure S4E–F. Consistent results were observed with CR‐MPs@MMAE and CR‐MPs@DOX Figure S4G–H. A critical challenge in cancer treatment is the development of chemoresistance. We established MTX‐resistant LLC cells (LLC‐MTXR) through progressive dose escalation, and confirmed their resistance by an IC50 shift in dose‐response curves Figure 2H. notably, these cells remained sensitive to CR‐MPs treatment. Although CR‐MPs@MTX was less effective against these resistant cells than against parental cells, it was still markedly superior to RT‐MPs. Figure 2I. These findings indicate the robust broad‐spectrum tumoricidal efficacy of the CR‐MPs therapeutic platform Figure S4I.

To elucidate the underlying mechanism of CR‐MPs‐induced cell death, transcriptomic profiling was performed on LLC and H1299 cells treated with CR‐MPs@MTX. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis revealed significant enrichment of ferroptosis‐related pathways and key ferroptosis‐driving genes Figure 2J–K. Flow cytometry and immunofluorescence with C11‐BODIPY staining showed that CR‐MPs@MTX treatment markedly increased lipid reactive oxygen species (ROS), the key executors of ferroptosis, in tumor cells compared with either RT‐MPs or MTX alone Figure 2L–N. A similar elevation of lipid ROS was observed upon treatment with CR‐MPs@MMAE or CR‐MPs@DOX Figure S5A–B. Furthermore, CR‐MPs@MTX‐induced cytotoxicity was attenuated by the ferroptosis inhibitor ferrostatin‐1 (Fer‐1), but not by inhibitors of pyroptosis (Azalamellarin‐N), necroptosis (Necrostain‐1), or apoptosis (Emricasan) Figure 2O and Figure S5C.

To explore the mechanism underlying CR‐MPs‐mediated ferroptosis, we traced the subcellular distribution of CR‐MPs. PKH26‐labeled CR‐MPs@MTX were observed to predominantly accumulate in mitochondria and lysosomes, while showing minimal colocalization with the endoplasmic reticulum or Golgi apparatus Figure S6A. This suggested that CR‐MPs may be transported via the endolysosomal pathway and then reside in mitochondria to act there. To test this hypothesis, we further investigated mitochondrial damage upon CR‐MPs treatment. Both total cellular ROS and mitochondrial ROS (mtROS) were observed to be markedly elevated following CR‐MPs@MTX treatment compared with RT‐MPs or MTX treatments Figure S6B–D. Concurrently, mitochondrial membrane potential depolarization was shown by a notable increase in JC‐10 monomers via flow cytometry and immunofluorescence assays Figure S6E–G. The mtROS scavenger Mito‐tempo further mitigated CR‐MPs@MTX induced cytotoxicity Figure S6H. Collectively, these results suggest that CR‐MPs target mitochondrial vulnerabilities to amplify oxidative stress damage, which initiates ferroptosis in tumor cells.

3.3. CR‐MPs Promote APCs Activation in Vitro

Ferroptosis is a form of ICD, characterized by the release of tumor‐associated antigens and DAMPs, which promotes phagocytosis of dying cells and activates APCs (Catanzaro et al. 2024). To determine whether, beyond direct tumor cells killing, CR‐MPs could also initiate APCs activation by triggering ICD, we first evaluated the release of classical DAMPs. CR‐MPs@MTX treatment potently induced the release of high mobility group box 1 (HMGB1), CRT, and ATP, demonstrating superior immunogenicity over RT‐MPs or MTX alone Figure 3A–D and Figure S7A–C. Transcriptomic profiling of tumor cells revealed an enrichment of multiple pro‐inflammatory signaling pathways and elevated inflammatory signatures following CR‐MPs@MTX treatment Figure S7D–E. Consistent with these results, the upregulation of various cytokines and chemokines genes such as Tnfa, Il‐12a, and Ccl7 further supported that CR‐MPs@MTX treated tumor cells have the potential to recruit and activate APCs Figure S7F. Accordingly, we co‐cultured tumor cells subjected to different treatments with the two APC subsets. Tumor cells treated with CR‐MPs@MTX significantly reprogrammed BMDMs toward a pro‐inflammatory M1 phenotype, as evidenced by elevated CD80 and reduced CD206 expression Figure 3E–F. Moreover, macrophages exhibited significantly enhanced phagocytic activity against CR‐MPs@MTX treated tumor cells Figure 3G. In parallel, CR‐MPs@MTX treated tumor cells also induced maturation of BMDCs, characterized by increased surface expression of MHC‐II and CD86 Figure S7G–H. Collectively, these findings support that CR‐MPs activate APCs by inducing ICD in tumor cells Figure S7I.

FIGURE 3.

CR‐MPs promote APCs activation in vitro.

(A) MFI of CRT on control, MTX treated, RT‐MPs treated and CR‐MPs@MTX treated LLC cells. (B, C) Relative (B) HMGB1 level and (C) ATP level in culture supernatant of control, MTX treated, RT‐MPs treated and CR‐MPs@MTX treated LLC cells. (D) Representative immunofluorescence images of CRT (red) on control, MTX treated, RT‐MPs treated and CR‐MPs@MTX treated H1299 cells. Scale bar = 20 µm. (E, F) MFI of (E) CD80 and (F) CD206 on BMDMs co‐cultured with control, MTX treated, RT‐MPs treated and CR‐MPs@MTX treated LLC cells. (G) Phagocytosis ratio of BMDMs to control, MTX treated, RT‐MPs treated and CR‐MPs@MTX treated LLC cells. (H) Uptake of PKH26‐labeled, LLC‐derived CR‐MPs@MTX by BMDCs was assessed by flow cytometry at 4, 12, and 24 h. (I, J) MFI of (I) CD80 and (J) CD86 on control BMDCs and BMDCs treated with MTX, RT‐MPs and CR‐MPs@MTX. (K) KEGG pathway enrichment analysis in CR‐MPs@MTX‐treated BMDCs versus control BMDCs. (L) GSEA of the interferon‐alpha (IFNα) response pathway in CR‐MPs@MTX‐treated BMDCs versus control BMDCs. (M) Western blot analysis of phospho‐TBK1, phospho‐p65, phospho‐IRF3, and actin in CR‐MPs@MTX treated or MTX treated BMDCs at 5‐, 10‐, 30‐, and 240‐minute post‐treatment. (N) RT‐qPCR evaluated expression of indicated genes in BMDCs after 24‐hour treatment with MTX, RT‐MPs, or CR‐MPs@MTX, compared with untreated control. All data are presented as mean ± s.d. One‐way ANOVA with Tukey's multiple comparisons test was performed. *P < 0.05, **P < 0.01, and ***P < 0.001; ns, no significance.

Given that RT‐MPs showed potent immunostimulatory effects on activation of BMDCs and BMDMs (Hu et al. 2024; Sun et al. 2023; Wan et al. 2020), we next investigated whether CR‐MPs could directly activate the two classical APC subsets in a similar manner. We first assessed the phagocytic capacity of APCs for CR‐MPs@MTX. Both APC subsets could efficiently internalize these microparticles. Figure 3H and Figure S7J. Treated BMDMs exhibited a marked increase in CD80 and a decrease in the M2 marker CD206, indicating a shift toward an M1 phenotype Figure S5K–L. Concurrently, CR‐MPs@MTX upregulated pro‐inflammatory genes (e.g., Il1a, Tnfa, Cxcl9, Cxcl10, Ccl2, Ccl5) and downregulated immunosuppressive markers (e.g., Tgfb, Chi) Figure S7M. We next sought to determine the effects of CR‐MPs@MTX on DCs. Following efficient internalization, CR‐MPs@MTX induced activation in BMDCs, evidenced by elevated expression of CD80 and CD86 Figure 3I–J. Subsequent transcriptomic analysis further confirmed the activation of BMDCs, characterized by the upregulation of immunostimulatory genes and downregulation of immunosuppressive markers Figure S8A–C. Multiple pro‐inflammatory pathways were significantly enriched. KEGG pathway analysis identified significant enrichment of TNF‐α signaling, NF‐κB activation, and cytosolic DNA sensing pathways, aligning with the enrichment of type I interferon alpha revealed by GSEA Figure 3K–L. Western blot analysis of BMDCs verified that CR‐MPs@MTX treatment rapidly induced phosphorylation of key molecules in the cGAS‐STING/NF‐κB pathways (TBK1, p65, and IRF3) Figure 3M. Consistently, Gene Ontology (GO) pathway analysis indicated positive regulation of cytokine production Figure S8D, aligning with the upregulation of multiple cytokines and chemokines genes regulated by IRF3 and NF‐κB Figure 3N. These results suggested that CR‐MPs triggered inflammatory activation in DCs and ultimately activated downstream transcription factors NF‐κB and IRF3. Taken together, CR‐MPs exert immunostimulatory effects by inducing ICD of tumor cell and directly activating APCs.

3.4. CR‐MPs and Combination with Immunotherapy Prolong Survival in MPE Murine Model

Having elucidated the tumoricidal role of CR‐MPs in vitro, we next evaluated the antitumor effects of CR‐MPs in vivo. We established an MPE murine model by intrapleural inoculation of LLC‐Luc cells and monitored tumor progression via bioluminescence imaging. In vivo imaging following intrapleural injection of DiR‐labeled CR‐MPs@MTX indicated intense fluorescence signal accumulation in the thoracic cavity, confirming pleural retention of the microparticles Figure S9A. We next analyzed cellular uptake within the MPE microenvironment. PKH26‐labeled CR‐MPs@MTX was administered intrapleurally to MPE‐bearing mice. Cells from pleural effusions and tumor nodules were then collected and analyzed by flow cytometry Figure S9B. PKH26‐labeled CR‐MPs@MTX was efficiently internalized by multiple cell populations. Notably, 41% of tumor cells (CD45⁻) were positive for PKH26 in MPE and 31% in tumor nodules. CR‐MPs were also largely taken up by key APCs, including macrophages (CD11b⁺F4/80⁺) and dendritic cells (CD11b⁺CD11c⁺) in both MPE and tumor nodules, whereas minimal uptake was observed in CD3⁺ T cells Figure 4A. To evaluate the therapeutic efficacy of CR‐MPs@MTX, three intrapleural doses were administered at two‐day intervals, conducting on day 9 post‐inoculation Figure 4B. A significant improvement in survival was observed in CR‐MPs@MTX treatment, superior to that of equivalent RT‐MPs, MTX and saline control, as verified by in vivo bioluminescence imaging Figure 4C–D and Figure S9C.

FIGURE 4.

CR‐MPs and combination with immunotherapy prolong survival in MPE mouse model.

(A) Quantification of the accumulation of PKH26‐labeled CR‐MPs@MTX in various cell types within tumor nodules and pleural effusion in MPE‐bearing mice. (B) Schematic diagram of therapeutic intervention for MPE. Saline, MTX (10 µg/mL matching the drug content in CR‐MPs), RT‐MPs (200 µg per treatment), and CR‐MPs (200 µg per treatment) were intrapleurally injected on day 9, 11, and 13 post‐inoculation. Mouse survival was monitored. (C) Survival analysis of indicated treatments (relative to B) in MPE‐bearing mice (n = 12 per group). (D) In vivo bioluminescence images of pleural tumor burden in MPE‐bearing mice after indicated treatments (day 25 post‐inoculation). (E) Schematic diagram of therapeutic intervention of CR‐MPs@MTX combined anti‐PD‐1 mAb for MPE. Saline and CR‐MPs (200 µg per treatment) were intrapleurally injected on day 4, 6, 8, and 10 post‐inoculation, and anti‐PD‐1 mAb (10 mg/kg) was intraperitoneally injected on day 4, 6, and 8 post‐inoculation. Mouse survival was monitored. (F) Survival analysis of indicated treatments (relative to E) in MPE‐bearing mice (n = 10 per group). (G) In vivo bioluminescence images of pleural tumor burden in MPE‐bearing mice after indicated treatments (day 12 post‐inoculation). (H) Quantification of total photon flux (photons/second) of pleural tumor burden in mice after indicated treatment. (I) Representative immunofluorescence images and quantitation of CRT in MPE‐bearing mice with the indicated treatments. Scale bar = 100 µm. (J) Schematic diagram of therapeutic intervention for MTX‐resistant MPE model. Saline, MTX (2.5 mg/kg), and CR‐MPs (200 µg per treatment) were intrapleurally injected on day 4, 6, 8, and 10 post‐inoculation, and anti‐PD‐1 mAb (10 mg/kg) was intraperitoneally injected on day 4, 6, and 8 post‐inoculation. Mouse survival was monitored. (K) Survival analysis of indicated treatments (relative to J) in mice with chemotherapy‐resistant MPE following intrapleural inoculation with LLC‐MTXR cells (n = 11 per group). (L) Representative in vivo bioluminescence images and thoracic cavity anatomical diagrams of indicated treatments in chemotherapy‐resistant MPE model. All data are presented as mean ± s.e.m. One‐way ANOVA with Tukey's multiple comparisons test was performed for (A, H, I). Log‐rank test was performed for (C, F, K). *P < 0.05, **P < 0.01, and ***P < 0.001; ns, no significance.

As noted above, CR‐MPs@MTX facilitated immunogenic death, activated two APC subsets, and concurrently upregulated PD‐L1 expression Figure S9D–E. These findings prompted us to investigate a combination strategy with anti‐PD‐1 monoclonal antibody (mAb) in subsequent experiments Figure 4E. The combination therapy significantly extended survival and achieved complete regression in 7 of 10 mice without recurrence, outperforming both anti‐PD‐1 and CR‐MPs@MTX monotherapies, as confirmed by in vivo bioluminescence imaging Figure 4F–H. Consistently, immunofluorescence images showed that the ICD marker CRT was significantly elevated after CR‐MPs@MTX monotherapy and further potentiated by combination immunotherapy Figure 4I.

As CR‐MPs@MTX alone showed cytotoxicity against MTX‐resistant cells in vitro, we investigated the synergistic potential of CR‐MPs@MTX and anti‐PD‐1 combination therapy in an MTX‐resistant MPE model. We established an MTX‐resistant MPE model by inoculating LLC‐MTXR cells into the pleural cavity Figure 4J. MTXR cells exhibited poor sensitivity to MTX therapy (2.5 mg/kg, i.p.) in vivo, and notably, the combination of CR‐MPs@MTX and anti‐PD‐1 mAb significantly prolonged the survival of mice bearing MTX‐resistant MPE, achieving complete remission in 60% (7 of 11) of treated mice, as revealed by both diminished bioluminescence intensity and eradication of pleural tumor nodules Figure 4K–L and Figure S9F. Collectively, these results from preclinical models provide a compelling rationale for advancing CR‐MPs toward clinical development.

3.5. CR‐MPs Reverse Immunosuppression in MPE through Activating APCs and T Cells

In light of our previous findings that combination immunotherapy potentiated CR‐MPs@MTX monotherapy efficacy in MPE, we next clarified the immune microenvironment changes in vivo. We performed flow cytometric analysis of immune cell populations among hemorrhagic pleural effusion and pleural tumor nodules within 48 h after indicated treatment Figure S10A–B. CR‐MPs@MTX monotherapy significantly augmented CD45⁺ immune cells frequencies, with combination therapy further enhancing CD45⁺ leukocyte infiltration compared with either monotherapy or anti‐PD‐1 alone Figure 5A. Combination therapy promoted the expression of CD80 on DCs over monotherapy or anti‐PD‐1 Figure 5B. Similarly, it enhanced macrophage activation, as indicated by elevated levels of both CD80 and CD86 Figure 5C–D. These results aligned with the immunostimulatory effects of CR‐MPs in vitro, indicating APCs activation. Concurrently, the immunosuppressive myeloid‐derived suppressor cells (MDSCs) population was decreased in both CR‐MPs@MTX monotherapy and combination treatment groups Figure 5E.

FIGURE 5.

CR‐MPs reverse immunosuppression in MPE through activating APCs and T cells.

(A) Percentage of CD45⁺ cells within the live cells population. (B) Percentage of CD80⁺ dendritic cells within the CD45⁺ cells population. (C) Percentage of CD80⁺ macrophages within the CD45⁺ cells population. (D) Percentage of CD86⁺ macrophages within the CD45⁺ cells population. (E) Percentage of MDSCs (Gr‐1⁺) within the CD45⁺ cells population. (F) Percentage of CD3⁺ T cells within the CD45⁺ cells population. (G) Percentage of CD4⁺ T cells within the CD45⁺ cells population. (H) Percentage of CD8⁺ T cells within the CD45⁺ cells population. (I) Percentage of CD8⁺Gzmb⁺ T cells within the CD45⁺ cells population. (J) Percentage of CD4⁺Gzmb⁺ T cells within the CD45⁺ cells population. (K) Ratio of Treg cells (CD4⁺Foxp3⁺) to Th1 cells (CD4⁺Gzmb⁺). (L) Representative immunofluorescence images and quantitation of activated TAMs (F4/80⁺CD86⁺) in MPE‐bearing mice with the indicated treatments. Scale bar = 100 µm. (M) Representative immunofluorescence images and quantitation of CTLs (Gzmb⁺CD8⁺) in MPE‐bearing mice with the indicated treatments. Scale bar = 100 µm. (N–P) Flow cytometry analysis of T cell subsets in the spleen of MPE‐bearing mice with the indicated treatments. (N) Percentage of CD4⁺Gzmb⁺ T cells within the CD4⁺ T cells population. (O) Percentage of CD4⁺IFNg⁺ T cells within the CD4⁺ T cells population. (P) Percentage of CD8⁺Gzmb⁺ T cells within the CD8⁺ T cells population. All data are presented as mean ± s.e.m. One‐way ANOVA with Tukey's multiple comparisons test was performed. *P < 0.05, **P < 0.01, and ***P < 0.001; ns, no significance.

The activation of APCs facilitated enhanced T cell priming, leading to increased infiltration of CD3⁺ T cells, CD4⁺ T cells, and CD8⁺ T cells in CR‐MPs@MTX and combination treatment groups Figure 5F–H. Augmented T cells exhibited marked anti‐tumor activity, as revealed by increased frequencies of cytotoxic T lymphocytes (CTLs; CD8⁺Gzmb⁺) and T helper type 1 cells (Th1; CD4⁺Gzmb⁺) Figure 5I–J. Notably, CD4⁺ Th1 cells showed a significant increase in the combination therapy group compared with monotherapy. In addition, the ratio of regulatory T cells (Treg; CD4⁺ Foxp3⁺) to Th1 cells was reduced in CR‐MPs@MTX and combination therapy groups Figure 5K. Multiplex immunohistochemistry showed greater infiltration of CD86⁺ macrophages and Gzmb⁺CD8⁺ T cells with combination therapy than with monotherapy Figure 5L–M and Figure S11A. To determine whether systemic immune activation also occurred, we next examined effector T cell responses in peripheral immune organs. The combination therapy significantly increased CTLs (CD8⁺Gzmb⁺) and Th1 cells (CD4⁺IFNg⁺) in the spleen compared with monotherapy or anti‐PD‐1 alone Figure 5N–P. In the lymph nodes, CTLs were elevated in both monotherapy and combination groups, while Th1 cells were significantly increased by the combination therapy over monotherapy Figure S11B–C. Collectively, these data demonstrate that CR‐MPs@MTX remodel the immunosuppressive MPE microenvironment by enhancing APCs activation and suppressing MDSCs infiltration, thereby promoting adaptive immunity. Integrating immunotherapy with monotherapy further potentiates a coordinated antitumor immune response.

3.6. CR‐MPs Combined with Immunotherapy Induce Long‐Term Immunological Memory

As demonstrated by prior observations that CR‐MPs@MTX combined with immunotherapy cured 70% of MPE‐bearing mice, we next examined whether these cured mice developed immunological memory. Tumor rechallenge experiment was conducted in mice cured for at least 30 days and age‐matched control mice Figure S12A. Cured mice exhibited significantly prolonged survival compared with naïve mice upon tumor re‐inoculation Figure 6A. In vivo imaging verified the absence of pleural effusion in cured mice during the observation period Figure 6B. We next evaluated the presence of memory T cells Figure S12B. Flow cytometric analysis showed a substantial increase in the proportions of effector memory T cells (T_EM) and central memory T cells (T_CM) in the spleen Figure 6C–F and the lymph nodes Figure 6G–J from cured mice compared with control mice. Taken together, these results indicate the establishment of long‐term immunological memory Figure 6K.

FIGURE 6.

CR‐MPs combined with immunotherapy induce long‐term immunological memory.

(A) Survival analysis comparing healthy control mice (Ctrl) and cured mice (Cured) following intrapleural inoculation with LLC‐luc cells. (B) Representative in vivo bioluminescence images of control and cured mice at day 7 and 14 post‐inoculation. (C–F) Flow cytometric analysis of memory T lymphocyte populations in the spleen from control and cured mice. (G–J) Flow cytometric analysis of memory T lymphocyte populations in the lymph nodes from control and cured mice. (K) Schematic illustration of the immunological memory generation mechanism mediated by combined CR‐MPs@MTX and anti‐PD‐1 therapy. (L–O) Hematological and biochemical analysis of blood samples from treated mice, including quantification of (L) red blood cells (RBCs), (M) white blood cells (WBCs), (N) alanine transaminase (ALT), and (O) aspartate transaminase (AST) levels. All data are presented as mean ± s.e.m. Log‐rank test was performed for (A). Two‐tailed Student's t test was performed for (C–J). One‐way ANOVA with Tukey's multiple comparisons test was performed for (L‐O). *P < 0.05, **P < 0.01, and ***P < 0.001; ns, no significance.

To evaluate the biocompatibility of CR‐MPs@MTX monotherapy and combination therapy, we documented weight changes during intervention, conducted hematological and biochemical analyses, and performed histopathological examinations of major organs (heart, liver, spleen, lung, and kidney). Stable body weights were maintained throughout CR‐MPs@MTX monotherapy and combination therapy intervention Figure S12C. Peripheral blood analysis revealed that counts of red blood cells (RBC) and white blood cells (WBC) maintained within physiological ranges across all treatments Figure 6L–M. Hepatic and renal function markers of mice receiving treatments, including alanine transaminase (ALT), aspartate transaminase (AST) Figure 6N–O, blood urea nitrogen (BUN), and creatinine (CR) Figure S12D‐E were also within normal ranges. Moreover, HE staining of heart, liver, spleen, lung, and kidney sections confirmed the absence of pathological changes following CR‐MPs@MTX monotherapy or combination therapy Figure S12F. These findings collectively demonstrate the excellent biosafety profile of CR‐MPs@MTX‐based therapies.

4. Discussion

MPE remains largely refractory to conventional chemotherapy and is ineligible for radiotherapy, leaving patients with limited therapeutic options (Gayen 2022). In this study, we demonstrate that CR‐MPs represent a novel therapeutic modality that functionally mimics an indirect form of “concurrent chemoradiotherapy”. CR‐MPs significantly suppressed tumor progression and improved survival across varied tumor burdens, while maintaining favorable tolerability. When combined with PD‐1 blockade, they conferred complete tumor remission in a subset of MPE mice, while inducing durable immunological memory. Notably, efficacy was maintained in chemotherapy‐resistant tumors, underscoring the exceptional translational potential of CR‐MPs for refractory MPE.

Conventional intrapleural chemotherapy demonstrates limited efficacy against MPE and is frequently associated with tumor recurrence. The limited efficacy stems from the non‐targeted nature of conventional chemotherapeutic agents, thereby compromising efficient delivery and accumulation in tumor tissues (Elsharkasy et al. 2020). Furthermore, free chemo‐agents readily diffuse into the systemic circulation, which contributes to their dose‐limiting systemic toxicity (Vader et al. 2016). To overcome these challenges, T‐MPs have emerged as promising drug carriers due to their inherent biocompatibility, active tumor‐targeting properties, and enhanced permeability and retention effects (M. Guo et al. 2019; Liang et al. 2019; Tang et al. 2012). Their autologous origin preserves the surface adhesion molecules of their parent tumor cells, enabling homotypic binding to recipient cancer cells, which is further complemented by EPR effect afforded by their nano‐scale size, leading to improved tumor site retention (Fang et al. 2020; Luk and Zhang 2015). Moreover, the encapsulation of chemotherapeutic drugs within T‐MPs effectively mitigates off‐target toxicity, enhancing the therapeutic window (M. Guo et al. 2019). Nevertheless, the application of T‐MPs as a therapeutic intervention for MPE remains underexplored. Specifically, UV‐MPs loaded with MTX (UV‐MPs@MTX) have demonstrated superior tumor‐targeting and cytotoxic capability in both cellular and animal experiments compared with free drug (M. Guo et al. 2019; Tang et al. 2012). Although UV‐MPs@MTX treatment could prolong survival in MPE‐bearing mice, complete tumor eradication was not achieved (M. Guo et al. 2019). These findings suggest that enhanced drug delivery alone may be insufficient for MPE control, stressing the need for the reprograming of TME and establishment of long‐term immune surveillance.

RT represents a potentially curative treatment modality renowned for its capacity to activate anti‐tumor immunity (Guo et al. 2023a). Moreover, the combinatorial strategy of RT and immunotherapy exhibits synergistic efficacy, significantly improving therapeutic responses (Ngwa et al. 2018). Our previous findings reveal that RT‐MPs, key mediators of the RIBE, can effectively inhibit tumor progression in both primary and metastatic models (Deng et al. 2024; Hu et al. 2024; Li et al. 2024; Sun et al. 2023; Wan et al. 2020). In our study, proteomic profiling and western blot analysis revealed that RT‐MPs were significantly enriched in immunostimulatory markers (e.g., CRT) compared with NC‐MPs or UV‐MPs. We further confirmed RT‐MPs possessed the most potent capacity to reprogram APCs. These finding is consistent with their previously reported superior capacity to prevent tumor growth (Sun et al. 2023; Wan et al. 2020), solidifying RT‐MPs as a highly promising immunotherapeutic candidate among the evaluated T‐MPs. However, RT‐MPs exhibit limited intrinsic tumoricidal activity, as evidenced by our observations that they are often insufficient to induce timely tumor regression under high tumor burden Figure 1D. Building on this rationale, we combined the advantages of RT‐MPs and chemotherapy, constructing CR‐MPs designed to simultaneously deliver cytotoxic payloads while activating localized anti‐tumor immunity. We employed an active drug‐loading approach, wherein irradiated tumor cells were directly incubated with chemotherapeutic agents. This process led to a moderate increase in particle size, attributable to the combined effects of drug‐induced cellular stress and the physicochemical properties of the payload. Nevertheless, the average diameter of CR‐MPs@MTX was around 313 nm, which is optimal for EPR‐mediated tumor accumulation. Supporting this, our in vivo imaging and flow cytometry analyses confirmed that following intrapleural administration, CR‐MPs were effectively retained within the thoracic cavity for at least 24 h and were subsequently internalized by both malignant cells and APCs Figure 4A and Figure S9A. We verified the superior broad‐spectrum cytotoxic activity of CR‐MPs over RT‐MPs carrier across homologous, heterologous, and even chemotherapy‐resistant tumor cells. Mechanistically, CR‐MPs increased lipid ROS levels, induced ferroptosis‐mediated immunogenic death and triggered DAMPs release, thereby amplifying innate antitumor immunity. At the same time, CR‐MPs preserve the APC‐activating properties of RT‐MPs, enabling dual targeting of malignant cells and pleural APCs. This synergistic mechanism sustained efficacy in preclinical MPE model Figure 4C. Furthermore, combining CR‐MPs with PD‐1 blockade potently reversed the immunosuppressive TME, as evidenced by increased APCs reprogramming and T cell priming. Consequently, this combination therapy eradicated established chemotherapy‐sensitive and ‐resistant MPE while inducing durable immunological memory Figure 4F, providing a compelling rationale for ‘radio‐chemo‐immuno’ triad strategies in clinical practice. Beyond mechanistic efficacy, CR‐MPs exhibit high biosafety in preclinical models. This profile, combined with the established clinical safety of autologous T‐MPs packaging MTX, underscores the feasibility of generating patient‐specific CR‐MPs from resected tumors, paving the way for personalized precision therapies (Dong et al. 2022; M. Guo et al. 2019).

In future research, we will investigate the potential mechanisms underlying CR‐MPs induced ferroptosis and APCs reprograming. We will also further investigate the therapeutic effects of CR‐MPs at different stages of MPE progression, and evaluate the therapeutic potential and safety of CR‐MPs in clinical studies. Additionally, we will develop standardized manufacturing processes, further enhancing batch‐to‐batch consistency in CR‐MPs characteristics, including size, composition, and drug encapsulation efficiency. Finally, we will establish customizable CR‐MPs platforms through genetic engineering (e.g., immune‐stimulating factors), antibody conjugation, and versatile drug loading strategies.

In conclusion, we have established an effective platform for synchronous radio‐chemotherapy delivery by CR‐MPs, and demonstrated its therapeutic potential against MPE both as monotherapy and in combination with immunotherapy. This approach represents a transformative strategy particularly suited for MPE management, presenting new clinical solutions for these challenging conditions.

Author Contributions

Minqi Zhou: writing – original draft, investigation, conceptualization. Lingyi Kong: investigation, writing – original draft. Weidong Zhong: investigation, writing – original draft. Yijun Wang: investigation. Wenqian Yuan: investigation. Xixi Liu: investigation. Jiacheng Wang: investigation. Mengyao Su: investigation. Yuze Zheng: investigation. Zishan Feng: writing – review and editing. Zhiyuan Zhou: writing – review and editing. Yue Deng: writing – review and editing. Wenwen Wei: writing – review and editing. Xiao Yang: writing – review and editing. Yan Hu: writing – review and editing. Chao Wan: writing – review and editing. Yuhan Sheng: conceptualization, writing – review and editing. Yajie Sun: conceptualization, writing – review and editing. Kunyu Yang: conceptualization, writing – review and editing.

Funding

This study was conducted with the support by Noncommunicable Chronic Diseases‐National Science and Technology Major Project (Grant No. 2023ZD0503003), Key R&D Program of Hubei Province (Grant No. 2024BCB051), National Natural Science Foundation of China (Grant No. 82330085), Chinese Society of Clinical Oncology Foundation (Grant No. Y‐MSDZD2022‐0476), Natural Science Foundation of Wuhan (Grant No. 2024040801020347), Natural Science Foundation of Hubei Province (Grant No. 2025AFB502), and National Natural Science Foundation of China (Grant No. 82102843).

Ethics Statement

All experimental procedures involving animals were approved by the Animal Experimentation Ethics Committee of Huazhong University of Science and Technology (HUST, Wuhan, China; Approval No. 4758). These procedures were performed in strict adherence to the committee's guidelines, which align with the policies of the Hubei Provincial Animal Care and Use Committee.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information: jev270277‐sup‐0001‐figureS1‐S12.docx

Supporting Information:jev270277‐sup‐0002‐TableS1.docx

Data Availability Statement

All data supporting the findings of this study are provided within the main paper and its Supplementary Information files. Any additional data required are available from the corresponding author upon reasonable request.