In-situ ²²³Ra-doped calcium-alginate composite microspheres: a high-LET and immunoactivating platform for α-particle radioembolization in hepatocellular carcinoma

Abstract

Transarterial radioembolization (TARE) with β-emitting radionuclides is widely used for hepatocellular carcinoma (HCC), but its clinical efficacy remains to be further improved. α-particle-emitting radionuclides possess high linear energy transfer (LET) and unique advantages in cancer therapy, motivating α-particle based composite platform. Accordingly, we engineer the first clinically mimetic α-TARE microsphere by in-situ ²²³Ra-doped calcium–alginate composite microsphere (²²³Ra/Ca-ALG MS) using a hydrogel matrix, in which alginate “egg-box” coordination captures Ra²⁺ to provide stable radiolabeling, delivered via selective hepatic arterial injection to HCC. The microspheres exhibited excellent radiolabeling stability (88% retention after 384 h) and potent, dose-dependent cytotoxicity against HCC cells under hypoxia. In an orthotopic rat HCC model, ²²³Ra/Ca-ALG MS-based TARE achieves precise intratumoral localization and sustained retention on SPECT/CT; ¹⁸F-FDG PET/CT and histopathology indicate a robust antitumor response, while serum biochemistry and histology support a favorable safety profile. Moreover, ²²³Ra/Ca-ALG MS provide powerful immune-activating capacity. Transcriptomics reveals activation of DNA-damage response, immunogenic cell death, and antigen-presentation pathways, flow cytometry and immunohistochemistry show increased dendritic-cell maturation and CD8⁺ T-cell infiltration. Collectively, ²²³Ra/Ca-ALG MS demonstrates hypoxia-tolerant cytotoxicity, immune-activating potential, offering new insights for the development of immune-based TARE strategies in HCC and showing promising prospects for clinical translation.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-03841-w.

Introduction

HCC remains the third leading cause of cancer-related mortality worldwide and is the predominant cause of death in patients with cirrhosis [1–3]. More than 60% of HCC cases are diagnosed at an unresectable stage, limiting curative treatment options [4]. In such cases, transarterial approaches are widely employed, leveraging the arterial hypervascularity of HCC, in contrast to the portal venous supply of healthy hepatic tissue. Transarterial chemoembolization (TACE) is currently the standard therapy for intermediate-stage disease [5, 6]. More recently, TARE — which delivers radioactive microspheres via selective intra-arterial catheterization—has gained clinical attention due to its targeted delivery, minimal systemic toxicity, and preservation of liver function [7]. Over the past five years, TARE adoption has increased by nearly 50%, supported by a consistently low incidence of treatment-related adverse events (< 6%) across multicenter studies.

TARE has demonstrated favorable safety and delivery profiles, yet it has not yielded a statistically significant overall survival benefit over conventional therapies [8]. The primary radionuclide used in TARE, yttrium-90 (⁹⁰Y), is a beta emitter with an average tissue penetration depth of approximately 2.5 mm [9]. As a low linear energy transfer (LET) radiation source, its cytotoxic effects are highly dependent on the oxygen enhancement effect, whereby molecular oxygen amplifies radiation-induced DNA damage through reactive oxygen species (ROS) generation [10]. However, the hypoxic microenvironment commonly found in solid tumors severely limits the efficacy of ⁹⁰Y-based radiotherapy by impairing ROS-mediated cytotoxicity, resulting in incomplete tumor eradication and frequent persistence of viable tumor cells post-treatment [11]. In contrast, alpha-emitting radionuclides maintain their cytotoxic potency even under hypoxic conditions, due to their markedly reduced oxygen enhancement ratios (OERs) [12]. This oxygen-independence makes alpha particles particularly well suited for treating radioresistant, poorly vascularized tumor regions where beta radiation is less effective. Additionally, alpha particles exhibit LET values that are 400–500 times higher than those of beta radiation, producing densely ionizing tracks over a short range. This leads to direct, irreparable double-strand DNA breaks and efficient induction of tumor cell apoptosis [13]. Based on our team’s prior research on alpha-emitting radionuclides [14], these unique radiobiological properties render alpha emitters a promising alternative to overcome the limitations of β-based TARE, particularly in addressing hypoxia-driven treatment resistance.

In addition to their well-established tumoricidal effects, alpha-emitting radionuclides have recently attracted attention for their capacity to stimulate antitumor immunity. Our previous study has demonstrated alpha-emitting radionuclides can trigger the release of damage-associated molecular patterns (DAMPs) to enhance tumor immunogenicity, providing therapeutic benefits beyond direct tumor cell killing [14]. Furthermore, alpha emitter–based intratumoral and intravenous injection have been shown to elicit immune-mediated clearance of both primary and distant tumor sites [15–18], underscoring the potential of localized alpha radiation to initiate systemic immune activation. Despite these promising findings, the immunological effects of alpha-emitting radionuclides-based microspheres following TARE remain poorly understood, particularly in HCC. As an immunologically “cold” tumor, HCC responds poorly to immune checkpoint blockade, highlighting the urgent need for strategies that can reshape its immune landscape [19]. Unlike conventional administration routes, α-TARE enables targeted embolization and sustained locoregional delivery of alpha radiation within hepatic tumors. Therefore, elucidating how TARE modulates tumor immunity, and leveraging this effect through rational immunotherapeutic combinations, may offer a promising strategy to overcome immune resistance in HCC and enhance the therapeutic impact of liver-directed alpha radiotherapy.

As the only globally FDA-approved alpha-emitter, ²²³Ra is particularly suitable for developing TARE. Its current clinical application has primarily been confined to bone metastasis treatment, research on TARE applications remains limited. The high energy of alpha particles poses technical challenges in radiopharmaceutical design. As an s-block element, ²²³Ra interacts with ligands via electrostatic forces. Being the largest stable + 2 ion in the periodic table, its low charge-to-radius ratio results in weaker metal-ligand bonds compared to smaller alkaline earth ions [20, 21]. Consequently, no suitable chelators have been identified for ²²³Ra in embolization microspheres. Alginate, a biocompatible and FDA-approved material for embolization, ensures the microspheres possess excellent injectability and embolization properties. The G-blocks in alginate microspheres can chelate with Ca²⁺ions to form stable ionic cross-links, resulting in a robust three-dimensional network structure. Leveraging the chemical similarity between radium and calcium (group 2 elements), Microspheres were first successfully synthesized via an in situ ²²³Ra loading method, thereby enabling the establishment of an α-TARE platform based on ²²³Ra.

This study designed novel ²²³Ra/Ca alginate microspheres for the treatment of HCC via TARE. These microspheres integrate embolization, alpha-particle radiotherapy, and immunostimulatory functions. Specifically, the ²²³Ra/Ca alginate microspheres are engineered to achieve both vascular embolization and localized alpha irradiation, providing direct cytotoxic effects while simultaneously promoting the release of tumor-associated antigens and proinflammatory cytokines. By coupling radiotherapy with immunomodulation, this strategy aims to enhance overall antitumor efficacy. To further potentiate immune activation, immunologic adjuvant CpG ODN 1826—a TLR9 agonist known to activate dendritic cells and amplify radiation-induced antitumor immunity—was co-administered [22]. This combinatorial approach is designed to overcome the immunosuppressive tumor microenvironment (TME) characteristic of HCC and augment the therapeutic outcomes of α-TARE. Importantly, the formulation exclusively utilizes clinically approved materials—²²³Ra, alginate, and CpG ODN 1826—to ensure translational relevance and facilitate future clinical implementation.

Methods

Materials

Sodium alginate (≥ 99%, S100128), isooctane (I113065), Span 80 (S110839), Tween 80 (T274282), calcium chloride (CaCl₂, ≥ 99%, C290953), and anhydrous ethanol (E130059) were purchased from Aladdin Reagent Co., Ltd. (Shanghai, China). ²²³RaCl₂ injection solution (Xofigo^®, Bayer, Germany) was used as the radioactive precursor, and ¹⁸F-FDG was obtained from Shanghai Atom Kexing Pharmaceutical Co., Ltd. (Shanghai, China). Ultrapure water was prepared using a Milli-Q system (Millipore, USA). Primary antibodies against γ-H2AX (AP1555), HSP70 (A23457), β-Actin (AC026), and HMGB1 (A2553), as well as secondary antibody Goat anti-Rabbit IgG (H + L, AS014), were purchased from ABclonal (China), and the CALR antibody (ab92516) was purchased from Abcam (UK). Cell Counting Kit-8 (CCK-8, Yeasen, 40203ES), Annexin V-FITC/PI Apoptosis Detection Kit (Yeasen, 40302ES60), enhanced ATP assay kit for intracellular ATP (Beyotime, S0027), extracellular ATP luminescence assay kit (Tongren Chemical, E299), and Live/Dead cell staining kit (Beyotime, C2015S) were used according to the manufacturers’ instructions. Western blot ECL chemiluminescence kit was purchased from Servicebio (G2014, China). All chemicals and reagents were of analytical grade and used as received.

Synthesis of ²²³Ra/Ca-ALG MS

To synthesize the ²²³Ra/Ca alginate microspheres, sodium alginate was first dissolved in ultrapure water, yielding a 2%–3% (w/v) solution (designated as solution A), which was left undisturbed for 2 h to remove air bubbles. In parallel, isooctane was mixed with surfactants Span 80 and Tween 80 in 3:1 ratio to create solution B. Solution B was then gradually added to solution A under continuous stirring, and the mixture was stirred for an additional 50 min at room temperature, forming a milky emulsion. During the subsequent crosslinking process, a 25% CaCl₂ solution was slowly introduced into the emulsion using a Pasteur pipette in a dropwise manner under constant stirring at approximately 600 rpm on a digital magnetic stirrer. After the CaCl₂ addition was completed, the system was allowed to equilibrate for 5 min, and then a calibrated activity of ²²³RaCl₂ injection solution (Xofigo^®, Bayer) was added dropwise using a 1 mL micropipette. Ra²⁺, being chemically similar to Ca²⁺, can effectively incorporate into the hydrogel matrix via this co-crosslinking and doping process. The mixture was stirred for an additional 2 h to facilitate co-precipitation, and then anhydrous ethanol was added to break the emulsion. The ²²³Ra/Ca alginate microspheres were collected by centrifugation, with the supernatant carefully removed to obtain the precipitated microspheres.

Characterization of ²²³Ra/Ca-ALG MS

The morphology and surface structure of the ²²³Ra/Ca alginate microspheres were observed by optical microscopy (Olympus BX53, Japan) and scanning electron microscopy (SEM, Hitachi SU8020, Japan). The particle size distribution was measured by laser diffraction (wet dispersion) using Mastersizer 2000 (Malvern Instruments, UK). Elemental composition and spatial distribution of C, O, Na, and Ca within the microspheres were analyzed by energy-dispersive X-ray spectroscopy (EDS, Oxford Instruments, UK) coupled to SEM. The crystallinity of the microspheres was determined by X-ray diffraction (XRD, Bruker D8 Advance, Germany). For evaluation of radiolabeling efficiency and in vitro radiostability, SPECT imaging was performed with a clinical gamma camera system (Siemens Symbia T16, Germany). Radiolabeling efficiency and retention were further quantified by incubating the microspheres in phosphate-buffered saline (PBS) and fetal bovine serum (FBS) at 37 °C, with radioactivity measured at multiple time points up to 384 h using a gamma counter (PerkinElmer 2480 Wizard2, USA).

Cell lines and cell culture

The human hepatocellular carcinoma (HCC) cell lines Huh7 cells (RRID: CVCL_0336) and Hep3B cells (RRID: CVCL_0326), as well as the rat hepatoma cell line N1S1, were all purchased from Shanghai Quicell Biotechnology Co., Ltd. (Shanghai, China). Since the N1S1 cell line is not currently listed in the RRID database, its authenticity was verified by short tandem repeat (STR) profiling and species identification (see Supporting Information). All cell lines were tested and confirmed to be free of mycoplasma contamination. Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Gibco, Thermo Fisher Scientific, USA) supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% penicillin–streptomycin (Gibco) at 37 °C in a humidified incubator with 5% CO₂. Cells were passaged every 2–3 days and used for experiments during the logarithmic growth phase. Cytotoxicity assays were performed under hypoxic conditions (1% O₂, 5% CO₂, balanced N₂) using a hypoxia incubator chamber.

Cell viability assessment

All in vitro assays were performed under hypoxic conditions (1% O₂) to mimic the ischemic and oxygen-deprived microenvironment that typically develops in tumors after transarterial radioembolization (TARE), where arterial embolization leads to sustained local hypoxia. Cell viability was evaluated using the Cell Counting Kit-8 (CCK-8, Beyotime Biotechnology, China) according to the manufacturer’s instructions. Huh7, Hep3B, and N1S1 cells were seeded into 96-well plates at a density of 5 × 10³ cells per well and allowed to adhere overnight. The cells were then treated with various concentrations of ²²³Ra/Ca-ALG MS under hypoxic conditions (1% O₂, 5% CO₂, 94% N₂) for 24 h. After treatment, 10 µL of CCK-8 reagent was added to each well, followed by incubation at 37 °C for 2 h. The absorbance at 450 nm was measured using a microplate reader (Thermo Fisher Scientific, USA). Cell viability was calculated as a percentage relative to untreated controls.

Colony formation assay

Huh7 and Hep3B cells were seeded in 6-well plates at a density of 2,000 cells per well and allowed to adhere overnight. ²²³Ra/Ca-ALG MS or control medium was then added, and the cells were continuously treated under hypoxic conditions (1%O₂, 5%CO₂, 94%N₂) for 7 days. During this period, the culture medium was replaced as needed. Each time the medium was changed, all microspheres were carefully collected from the supernatant by centrifugation and re-added to the wells together with fresh complete DMEM. After 7 days of treatment, the medium was replaced with fresh DMEM, and the cells were cultured for an additional 7 days to allow colony formation. Colonies were fixed with 4% paraformaldehyde for 20 min, stained with 0.1% crystal violet for 15 min, washed gently with PBS, and air-dried. Colonies containing more than 50 cells were counted manually under a microscope. All experiments were performed in triplicate.

Cell apoptosis assay

Apoptosis was assessed using the Annexin V-FITC/PI Apoptosis Detection Kit (Yeasen, 40302ES60) according to the manufacturer’s protocol. Huh7 and Hep3B cells were seeded in 6-well plates and incubated overnight. The cells were then treated with ²²³Ra/Ca-ALG MS or control medium under hypoxic conditions (1%O₂, 5%CO₂, 94%N₂) for the indicated times. At the end of treatment, both floating and adherent cells were harvested, washed twice with cold PBS, and resuspended in binding buffer. Annexin V-FITC and propidium iodide (PI) were added to the cell suspension and incubated for 15 min at room temperature in the dark. Apoptotic cells were analyzed by flow cytometry (Beckman Coulter, USA) within one hour. Data analysis was performed using FlowJo software. All experiments were performed in triplicate.

Live/Dead cell staining assay

Cell viability was further assessed using a Live/Dead cell staining kit (Beyotime, C2015S) according to the manufacturer’s instructions. Huh7 and Hep3B cells were seeded in 24-well plates and allowed to adhere overnight. The cells were then treated with ²²³Ra/Ca-ALG MS or control medium under hypoxic conditions (1%O₂, 5%CO₂, 94%N₂) for the indicated times. After treatment, cells were washed twice with PBS and stained with calcein-AM and propidium iodide (PI) working solution at 37 °C for 30 min in the dark. Stained cells were visualized and imaged using an inverted fluorescence microscope (Olympus IX73, Japan). Live cells emitted green fluorescence, while dead cells emitted red fluorescence. All experiments were performed in triplicate.

Immunofluorescence analysis of γ-H2AX assay

The formation of γ-H2AX foci was evaluated by immunofluorescence staining. Huh7 and Hep3B cells were seeded on glass coverslips in 24-well plates and allowed to adhere overnight. After treatment with ²²³Ra/Ca-ALG MS or control under hypoxic conditions (1%O₂, 5%CO₂, 94%N₂) for the indicated times, cells were washed with PBS (Servicebio, G4202) and fixed with 4% paraformaldehyde for 20 min at room temperature. Cells were permeabilized with 0.2% Triton X-100 in PBS for 10 min and blocked with 5% bovine serum albumin (BSA, Servicebio,) for 1 h. The cells were then incubated with primary antibody against γ-H2AX (AP1555, ABclonal) at 4 °C overnight, washed, and incubated with FITC-conjugated secondary antibody (Servicebio, GB22303) for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI (Servicebio, G1012) for 5 min. Coverslips were mounted with anti-fade mounting medium and observed under a fluorescence microscope (Olympus IX73, Japan). Representative images were acquired, and γ-H2AX foci per nucleus were quantified using ImageJ software. All experiments were performed in triplicate.

Western blot analysis of ICD biomarker

To analyze the expression of ICD biomarkers, Huh7 and Hep3B cells were treated with ²²³Ra/Ca-ALG MS or control medium under hypoxic conditions (1%O₂, 5%CO₂, 94%N₂) for the indicated times. After treatment, cells were harvested and lysed in RIPA buffer (Servicebio, G2002) containing protease inhibitor cocktail (Servicebio, G2006). Total protein concentrations were determined using a BCA Protein Assay Kit (Beyotime, P0010). Equal amounts of protein were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore, IPVH00010). Membranes were blocked with 5% non-fat milk in TBST (Servicebio, G0004) for 1 h at room temperature, then incubated overnight at 4 °C with primary antibodies against CALR (ab92516, Abcam), HSP70 (A23457, ABclonal), HMGB1 (A2553, ABclonal), and β-Actin (AC026, ABclonal) as the internal control. After washing, membranes were incubated with HRP-conjugated secondary antibody (AS014, ABclonal) for 1 h at room temperature. Bands were visualized using ECL chemiluminescence reagent (Servicebio, G2014) and detected using a ChemiDoc imaging system (Bio-Rad, USA). Band intensities were quantified using ImageJ software, with β-actin used for background subtraction and normalization. All experiments were performed in triplicate.

Immunofluorescence analysis of CALR assay

To evaluate CALR exposure on the cell surface, immunofluorescence staining was performed as follows. Huh7 and Hep3B cells were seeded onto confocal dishes and allowed to adhere overnight. After treatment with ²²³Ra/Ca-ALG MS or control under hypoxic conditions (1% O₂, 5% CO₂, 94% N₂) for the indicated times, cells were gently washed with PBS (Servicebio, G4202) and fixed with 4% paraformaldehyde for 20 min at room temperature. Cells were blocked with 5% bovine serum albumin (BSA, Servicebio, G5001) for 1 h at room temperature, and then incubated with primary antibody against CALR (ab92516, Abcam) at 4 °C overnight. After washing with PBS, the cells were incubated with 594-Goat Anti-Rabbit Recombinant Secondary Antibody (H + L) (Proteintech, RGAR004) for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI for 5 min. The samples were mounted using anti-fade mounting medium and imaged using a confocal laser scanning microscope (CLSM, Nikon Corporation, Japan). Representative images were acquired, and membrane-localized CALR fluorescence intensity was quantified using ImageJ software. All experiments were performed in triplicate.

Measurement of ATP level

Intracellular and extracellular ATP levels were measured using the Enhanced ATP Assay Kit (Beyotime, S0027) and the Extracellular ATP Assay Kit–Luminescence (Tongren Chemical, E299), respectively, following the manufacturers’ instructions. Huh7 and Hep3B cells were seeded in 6-well plates and treated with ²²³Ra/Ca-ALG MS or control under hypoxic conditions (1%O₂, 5%CO₂, 94%N₂) for 24 h. For intracellular ATP measurement, cells were harvested, washed with cold PBS, lysed using the kit lysis buffer, and the lysate was centrifuged to collect the supernatant. ATP content in the supernatant was quantified using a microplate reader (Thermo Fisher Scientific, USA) to detect relative light units (RLU). Protein concentration was determined with the BCA Protein Assay Kit (Beyotime, P0010), and ATP levels were normalized to total protein, expressed as nmol ATP per mg protein. For extracellular ATP detection, the culture supernatant was collected at 24 h and ATP levels were determined directly using the luminescence assay kit and the same microplate reader; results were reported as RLU. All experiments were performed in triplicate.

RNA extraction and sequencing (RNA-Seq) analysis

Total RNA was extracted from treated and control Huh7 cells using Trizol reagent (Invitrogen, USA) according to the manufacturer’s instructions. RNA concentration and integrity were assessed by NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). RNA quality control, library construction, high-throughput sequencing, and all subsequent bioinformatics analyses were performed by Hangzhou Lianchuan Biotechnology Co., Ltd. (Hangzhou, China). RNA-seq libraries were constructed using the standard Illumina mRNA library preparation protocol and sequenced on an Illumina NovaSeq 6000 platform. Clean reads were mapped to the human reference genome and gene expression levels were quantified as fragments per kilobase of transcript per million mapped reads (FPKM). Differentially expressed genes (DEGs) between ²²³Ra/Ca-ALG MS-treated and control groups were identified using the edgeR or DESeq2 package in R, with screening criteria of |fold change| >2 and p-value < 0.05. Principal component analysis (PCA) was conducted to assess sample clustering and group separation. Volcano plots and heatmaps were generated to visualize the distribution and clustering of DEGs. Functional enrichment analysis, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment, was performed to identify significantly associated biological processes and pathways. Gene set enrichment analysis (GSEA) was also conducted to further investigate global changes in relevant pathways. All RNA-Seq experiments were performed with three independent biological replicates per group.

Establishment of orthotopic N1S1 HCC model and treatment procedure

All animal experiments were approved by the Animal Research Committee of First Affiliated Hospital of Naval Medical University (CHEC (A.E)2025-001). The rat N1S1 model was established by orthotopic tumor implantation method. When the N1S1 cell density reached 70–80%, approximately 3 × 10⁶ cells were harvested for implantation per rat. N1S1 cells were mixed with matrigel and the mixture was injected into the liver middle lobe by syringe. One week after implantation, ¹⁸F-FDG PET/CT (Siemens Biograph TruePoint) imaging was performed as both a pre-treatment screening modality and a baseline imaging tool. The presence of intrahepatic hypodense nodules with increased ¹⁸F-FDG uptake on PET/CT confirmed successful establishment of the rat HCC model. ²²³Ra/Ca-ALG MS were administered via hepatic arterial catheterization. After performing an upper midline laparotomy, the gastroduodenal artery was isolated. ²²³Ra/Ca-ALG MS solution was injected via retrograde catheterization through the gastroduodenal artery into the PHA. The injected dose of ²²³Ra/Ca-ALG MS was 88 kBq, with a total injection volume of 0.2 mL. Penicillin was administered at a dose of 200,000 units per rat to prevent infection.

SPECT/CT and biodistribution

SPECT/CT imaging (Siemens Symbia T16 SPECT/CT) was performed at 1 h, 5 d, 10 d, and 20 d post-TARE to evaluate the biodistribution of ²²³Ra/Ca-ALG MS. Anesthesia was induced with isoflurane at a concentration of 5% and a flow rate of 1 L/min, followed by maintenance anesthesia at 2% isoflurane with a flow rate of 500 mL/min. A high-energy collimator was used. Imaging was collected using three characteristic energy peaks of ²²³Ra decay: 85 keV, 154 keV, and 270 keV. The matrix size was 128 × 128 with a magnification factor of 1.43. The CT acquisition was performed with a tube current of 40 mA, a pitch of 1.0, and a slice thickness of 0.75 mm. Animals were positioned in the supine position during scanning. Three-dimensional ROIs corresponding to distinct anatomical structures were delineated based on post-processing whole-body imaging data. Both the counts and cumulative volumes of these ROIs were measured (Fig. S4). Tissue distribution of the ²²³Ra/Ca-ALG MS was then calculated and presented as the percentage of the injected dose per milliliter (%ID/mL). The tumor-to-liver background ratio of radioactive counts was calculated.

Evaluation of tumor volume and metabolic activity using ¹⁸F-FDG PET/CT

¹⁸F-FDG PET/CT imaging was performed to assess N1S1 tumor volume and metabolic activity. Follow-up ¹⁸F-FDG PET/CT was performed at pre-treatment and on postoperative days 2, 7, and 14. The radiotracer ¹⁸F-FDG was provided by Shanghai Atom Kexing Pharmaceutical Co., Ltd., with a radiochemical purity exceeding 95%. Prior to imaging, animals were fasted for 12 h and deprived of water for 6 h. Body weights were measured and recorded for each rat. ¹⁸F-FDG was administered via tail vein injection at a dose of ​9 MBq​/kg. One hour after injection, animals were also anesthetized with isoflurane, positioned supine, and subjected to PET/CT scanning. PET data were acquired at 3 min per bed position and reconstructed using the ordered-subset expectation maximization algorithm. CT parameters were as follows: tube voltage, 100 kV; tube current, 150 mA; slice thickness, 1 mm; slice interval, 1 mm; and reconstruction with the medium-smooth kernel (H40s). The acquired imaging data were transferred to the MedEx image post-processing system for image fusion, three-dimensional multiplanar reconstruction. ROIs were delineated semi-automatically on the tumor at each time point to calculate metabolic parameters, including the SUVmax and MTV.

Bio-TEM

To observe ultrastructural and organelle-level changes induced by ²²³Ra/Ca-ALG MS-based TARE, tumor tissue samples were collected from each group at the indicated time after treatment. Fresh tumor tissue was cut into 1 mm³ cubes and immediately fixed in 2.5% glutaraldehyde (Servicebio, G1102) in 0.1 M phosphate buffer (pH 7.4) at 4 °C overnight. After washing with phosphate buffer, samples were post-fixed in 1% osmium tetroxide at 4 °C for 1–2 h, dehydrated through a graded ethanol series, and embedded in epoxy resin. Ultrathin Sects. (70–90 nm) were cut with an ultramicrotome, collected on copper grids, and double-stained with uranyl acetate and lead citrate. Ultrastructural features were examined using a transmission electron microscope (HT7800 RuliTEM). All sample processing and TEM analysis were performed by Wuhan Servicebio Technology Co., Ltd.

Tumor-infiltrating immune cell analysis

Tumor-infiltrating lymphocytes and dendritic cells were analyzed by flow cytometry in all groups at 7 days after TARE. Fresh tumor tissues were minced and digested in RPMI-1640 medium containing 0.1% collagenase IV (Sigma, C5138), 0.01% DNase I (Sigma, D5025), and 2% fetal bovine serum (FBS, Gibco) at 37 °C for 45 min with gentle agitation. The digested tissue was filtered through a 70 μm cell strainer (Corning), washed twice with cold PBS (Servicebio, G4202), red blood cells were lysed using RBC lysis buffer (BioLegend, 420301). Single-cell suspensions were incubated with anti-CD16/32 antibody (BioLegend, 101320) to block Fc receptors and then stained on ice with fluorochrome-conjugated antibodies for CD4 (BioLegend, 100406), CD8 (BioLegend, 100708), CD80 (BioLegend, 104712), and CD86 (BioLegend, 105012) for 30 min. After staining, cells were washed, resuspended in PBS with 2% FBS, and analyzed by flow cytometry (Beckman Coulter, USA). Data were analyzed using FlowJo software. CD8⁺ T cells were gated as CD8a⁺ cells, and activated dendritic cells (DCs) were identified as CD80⁺CD86⁺ double-positive cells. The proportions of each immune subset were calculated as a percentage of total viable cells. All groups consisted of four independent replicates (n = 4).

Biosafety evaluation

To comprehensively evaluate the biosafety of ²²³Ra/Ca-ALG MS-based TARE, rats in all treatment groups underwent serial serum biochemical analysis and histopathological examination. Blood samples were collected at 2, 7, 14, and 28 days after TARE for measurement of serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), albumin (ALB), and creatinine (CREA) using a fully automated biochemical analyzer (Hitachi 7600, Japan). All values were compared to reference physiological ranges. Major organs including the liver, heart, lung, kidney, and spleen were harvested at 28 days post-treatment, fixed in 4% paraformaldehyde, paraffin-embedded, sectioned, and stained with hematoxylin and eosin (H&E) for histopathological assessment. Pathology sectioning and H&E staining were performed by Shanghai YiLianBo Technology Co., Ltd. The architecture and morphology of each organ were examined for evidence of inflammatory infiltration, necrosis, or structural abnormalities. No significant pathological alterations were detected in any organs across all groups. All biosafety evaluations were performed in four independent replicates (n = 4).

Statistical analysis

Statistical analyses were performed using GraphPad Prism. Data are presented as mean ± SD unless otherwise indicated. Group differences were analyzed using one-way ANOVA or Student’s t-test, as appropriate. A significance threshold of p < 0.05 was considered statistically significant. RNA-seq data were analyzed using standard pipelines with Benjamini–Hochberg multiple-testing correction. Exact sample sizes and statistical tests are indicated in the figure legends.

Results and discussion

Synthesis and characterization of ²²³Ra/Ca alginate microspheres

Radium-223/Calcium alginate microspheres (²²³Ra/Ca-ALG MS) were synthesized via an oil-in-water emulsification synthesis method, as illustrated in Scheme 1a. Sodium alginate [(C₆H₇O₆Na)_n], a biopolymer obtained as a byproduct during iodine and mannitol extraction from marine macroalgae, is a linear polysaccharide composed of alternating β (1→4)-D-mannuronic acid (M) and α (1→4)-L-guluronic acid (G) units. This anionic copolymer forms hydrogels through cationotropic gelation, whereby divalent cations (Ca²⁺, Ba²⁺, Sr²⁺) replace sodium ions within the G-block regions to create a 3D network. Ca-based microspheres were selected for all subsequent experiments because calcium is an essential physiological element with a well-established history of clinical use and regulatory approval.

Scheme 1.

(a) Schematic illustration of the ²²³Ra/Ca-ALG MS synthesis. (b) The mechanism of the antitumor effect includes the direct cytotoxic effect and immune activation after ²²³Ra/Ca-ALG MS-based TARE in HCC

The ²²³Ra/Ca-ALG MS are homogeneously dispersed in water (Fig. 1a). The microspheres exhibit a smooth surface and well-defined boundaries under optical microscopy (Fig. 1b). Scanning electron microscopy (SEM) revealed spherical morphologies with coarse surfaces for both Ca-ALG MS and ²²³Ra/Ca-ALG MS (Fig. 1c&d). Energy-dispersive X-ray spectroscopy (EDS) mapping images indicate that elements C, O, Na, and Ca are homogeneously distributed over the surface of the microspheres (Fig. 1e&f) and calcium shows the highest content. Due to the extremely low concentration of ²²³Ra (kBq-level activity corresponding to sub-ppm elemental levels) and its weak X-ray emission beyond the typical detection range of conventional EDS, the characteristic Ra peak was not observed in the spectra, the Ra incorporation was further confirmed by the localized radioactive signal and excellent radiolabeling stability demonstrated in Fig. 1i&j. Based on our previous studies on transarterial embolization (TAE) in rat hepatic arteries, the terminal arterioles in rat livers typically range from 10 to 50 μm in diameter [23]. To achieve effective embolization of HCC in rats and prevent microspheres from ectopically distributing to non-target organs, the optimal microsphere size for rat HCC TAE is 40–100 μm, with diameters above 100 μm being ineffective for proper embolization [24]. The synthesized microspheres were analyzed for size distribution measured by laser diffraction (wet dispersion) size analyzer, the results indicated an average diameter of 67.58 ± 0.94 μm (Fig. 1g) demonstrating that the prepared microspheres met the size criteria for effective embolization in rat models. X-ray diffraction (XRD) analysis revealed predominantly amorphous peaks in the spectrum, indicating that the alginate microspheres maintained an amorphous polymer structure without detectable crystalline impurities (Fig. 1h). To confirm the successful synthesis and radiolabeling of ²²³Ra/Ca-ALG MS, visual comparison and SPECT imaging were performed (Fig. 1i). The ²²³Ra/Ca-ALG MS sample appeared as a distinct white pellet at the bottom of the tube, similar to blank Ca-ALG MS. Importantly, SPECT imaging revealed a stark contrast in radioactive signal distribution: the ²²³RaCl₂ solution exhibited diffuse radioactivity throughout the liquid phase, whereas the ²²³Ra/Ca-ALG MS displayed a sharply localized radioactive signal strictly confined to the microsphere pellet. This demonstrates that ²²³Ra was efficiently and stably incorporated into the microsphere matrix, rather than freely diffusing in solution. No radioactivity was detected in the blank Ca-ALG MS, confirming the specificity of the labeling. Radiostability assessment of ²²³Ra/Ca-ALG MS was further evaluated in vitro (Fig. 1j). When incubated in both PBS and FBS, the microspheres retained up to 88% of their initial radionuclide load after 384 h, indicating robust stability of the radiolabel under physiological and serum-containing conditions. These findings confirm that the developed ²²³Ra/Ca-ALG MS possess high labeling efficiency and sustained radiostability, making them suitable for subsequent biological and therapeutic studies.

Fig. 1.

Characteristics of microspheres. (a) Photograph of Ca-ALG MS (left) and ²²³Ra/Ca-ALG MS (right). (b) Optical microscopy image of ²²³Ra/Ca-ALG MS dispersed in deionized water. (c) SEM images of Ca-ALG MS. (d) SEM images of ²²³Ra/Ca-ALG MS. (e) EDS elemental mapping images of ²²³Ra/Ca-ALG MS. The spectra mainly reflect the major matrix elements (C, O, Na, Cl, and Ca), as the characteristic X-ray signal of ²²³Ra is below the detection limit of EDS due to its extremely low concentration and weak X-ray yield (sub-ppm level). (f) Quantitative elemental maps of C, O, Na, and Ca obtained by EDS analysis. (g) Size distribution of Ca-ALG MS and ²²³Ra/Ca-ALG MS measured by laser diffraction (wet dispersion, Mastersizer 2000). (h) XRD analysis of Ca-ALG MS and ²²³Ra/Ca-ALG MS. (i) Representative photographs and corresponding SPECT/CT images of ²²³RaCl₂, Ca-ALG MS, and ²²³Ra/Ca-ALG MS. (j) Radiostability of ²²³Ra/Ca-ALG MS in PBS and FBS

In vitro antitumor effects of ²²³Ra/Ca-ALG microspheres

TARE disrupts the arterial blood supply to tumors, resulting in acute and sustained hypoxia within the tumor region. Such a hypoxic microenvironment has been shown to significantly increase tumor resistance to conventional radiotherapy and to promote recurrence [25]. To authentically mimic the physiological conditions following TARE, all in vitro results reported in this study were obtained under hypoxic conditions (1% O₂), enabling us to rigorously evaluate the antitumor efficacy of ²²³Ra/Ca-ALG MS under these “radiation-unfavorable” circumstances.

We first focused on evaluating the acute inhibitory effect of ²²³Ra/Ca-ALG MS on cell viability under hypoxic conditions, while also designing a dose gradient to determine the IC₅₀ and to guide dosing for all subsequent experiments. To this end, we performed CCK-8 assays in which Huh7 and Hep3B cells were exposed to varying concentrations of ²²³Ra/Ca-ALG MS for 24 h under 1% O₂. The results showed a significant, dose-dependent reduction in cell viability; at the highest concentration tested (14.28 kBq/mL), the viability of Huh7 and Hep3B cells dropped to just 5% and 2%, respectively (Fig. 2a). The calculated IC₅₀ values were 1.322 kBq/mL for Huh7 cells and 1.040 kBq/mL for Hep3B cells (Fig. 2b), demonstrating that ²²³Ra/Ca-ALG MS maintained strong acute cytotoxicity even under hypoxic conditions. Based on these findings, and with reference to previous high-quality studies on the antitumor effects of ²²³Ra, we ultimately selected 3.7 kBq/mL—a dose that reduced cell viability to approximately 20%—as the standard working concentration for the subsequent series of in vitro experiments [14]. This choice ensured robust antitumor efficacy while allowing clear observation of relevant intracellular molecular biological effects. Live/dead cell staining assays were performed to evaluate the direct cytotoxic effects of ²²³Ra/Ca-ALG MS on HCC cells, both Huh7 and Hep3B cells treated with ²²³Ra/Ca-ALG MS exhibited a dramatic increase in cell death compared to the control and Ca-ALG MS groups (Fig. S1). Viable cells (stained green) were abundant in the control and Ca-ALG MS groups, while the majority of cells in the ²²³Ra/Ca-ALG MS group stained red, indicating loss of membrane integrity and cell viability. This effect was consistently observed in both Huh7 and Hep3B cell, confirming the potent cytotoxic activity of ²²³Ra/Ca-ALG MS in vitro.

Fig. 2.

Evaluation of the in vitro antitumor effects of ²²³Ra/Ca-ALG microspheres under hypoxic conditions. (a) Cell viability of Huh7(i) and Hep3B(ii) cells after 24 h incubation with escalating doses (0–14.28.28 kBq/mL) of ²²³Ra/Ca-ALG MS under 1% O₂, assessed by CCK-8 assay (n = 3). (b) Dose–response curves and IC₅₀ values of ²²³Ra/Ca-ALG MS for Huh7 and Hep3B cells, calculated by nonlinear regression. (c-i) Representative images of colony formation in Huh7 and Hep3B cells following 7 days of continuous exposure to 3.7 kBq/mL ²²³Ra/Ca-ALG MS under hypoxia; (c-ii) quantification of colony numbers in Huh7 cells; (c-iii) quantification of colony numbers in Hep3B (n = 3). (d-i) Representative flow cytometry dot plots of Annexin V-FITC/PI-stained Huh7 and Hep3B cells after 12 h and 24 h treatment with 3.7 kBq/mL ²²³Ra/Ca-ALG MS under 1% O₂; (d-ii) quantification of apoptotic rates in Huh7 cells; (d-iii) quantification of apoptotic rates in Hep3B cells (n = 3). (e-i) Immunofluorescence images of γ-H2AX (green) in Huh7 and Hep3B cells after 24 h exposure to 3.7 kBq/mL ²²³Ra/Ca-ALG MS under hypoxia, with DAPI (blue) nuclear counterstain; (e-ii) quantification of γ-H2AX fluorescence intensity in Huh7 cells; (e-iii) quantification of γ-H2AX fluorescence intensity in Hep3B (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Clonogenicity and the ability to proliferate indefinitely are among the most critical biological features of malignant tumors. Unlike the short-term cytotoxicity reflected by the CCK-8 assay, the colony formation assay focuses on the long-term proliferative capacity of tumor cells. This method also more accurately models the sustained suppressive effect of radioactive microsphere embolization on tumor clonality and proliferation after TARE. In this study, we continuously treated Huh7 and Hep3B cells with 3.7 kBq/mL ²²³Ra/Ca-ALG MS for 7 days. The results showed that only 410 colonies formed in the Huh7 cells and 360 in the Hep3B cells, representing a significant reduction compared to controls (Fig. 2c), indicating that ²²³Ra/Ca-ALG MS exerts a lasting inhibitory effect on tumor cell clonogenicity and proliferation. Notably, the Ca-ALG MS group also exhibited a slight, but statistically insignificant, reduction in colony numbers. We attribute this minor decrease mainly to the physical presence of microspheres occupying space on the culture plate, which may interfere with cell attachment and distribution, rather than any intrinsic cytotoxicity of the carrier itself [26].

We performed flow cytometric analysis to assess apoptosis in cells treated with ²²³Ra/Ca-ALG MS under hypoxic conditions. The results showed a significant, time-dependent increase in the proportion of apoptotic cells following exposure to 3.7 kBq/mL ²²³Ra/Ca-ALG MS: after 12 h, the Annexin V–positive fraction (Annexin V⁺/PI⁻ and Annexin V⁺/PI⁺)s in Huh7 and Hep3B cells reached 51.2% and 43.3%, respectively, and further increased to 80.1% and 58.5% at 24 h, while the control group consistently remained at a low baseline level (Fig. 2d). These findings indicate that ²²³Ra/Ca-ALG MS induces robust tumor cell death—including apoptosis and necrosis—even under hypoxia.

One key mechanism of radiation-induced tumor cell death is DNA damage. The degree and irreversibility of DNA damage are crucial determinants of treatment efficacy and the likelihood of tumor recurrence. To further elucidate the extent of DNA damage induced by ²²³Ra/Ca-ALG MS under hypoxic conditions, γ-H2AX immunofluorescence staining was performed after exposing Huh7 and Hep3B cells to 3.7 kBq/mL ²²³Ra/Ca-ALG MS for 24 h. γ-H2AX is the phosphorylated form of the histone variant H2AX and serves as a highly sensitive marker for DNA double-strand breaks [27]. The results revealed abundant clustering of γ-H2AX signals within the nuclei of treated cells (Fig. 2e), indicating widespread and extensive DNA double-strand breaks. This observation is consistent with the high LET characteristics of alpha particles, which can induce densely clustered and largely irreparable DNA lesions irrespective of oxygen concentration.

In summary, our systematic in vitro studies under hypoxic conditions—mimicking the clinical environment after TARE—demonstrated that ²²³Ra/Ca-ALG MS possesses robust and sustained antitumor activity against HCC cells. ²²³Ra/Ca-ALG MS not only caused potent, dose-dependent acute cytotoxicity, but also persistently inhibited tumor cell clonogenicity and proliferation. Mechanistically, these effects were accompanied by a marked, time-dependent induction of apoptosis and extensive DNA double-strand breaks, as evidenced by γ-H2AX foci formation. Importantly, all these antitumor effects were maintained even under severe hypoxia, highlighting the unique advantage of high LET alpha particles in overcoming the radioresistance typically associated with oxygen-deprived TME. Meanwhile, our results also demonstrated that Ca-ALG MS alone exhibited excellent biocompatibility and safety, with no significant cytotoxicity observed in any of the assays; pronounced antitumor effects were exclusively seen in the radiolabeled group. Collectively, these findings lay a strong experimental foundation for further exploration of ²²³Ra/Ca-ALG MS as a promising TARE platform to overcome hypoxia-induced therapeutic limitations in HCC.

Transcriptomic profiling reveals immunogenic cell death and immune activation induced by ²²³Ra/Ca-ALG MS in HCC cells

Principal component analysis (PCA, Fig. 3a) revealed a clear separation between the ²²³Ra/Ca-ALG MS-treated and control groups, with tight clustering observed within each group. This finding indicates that ²²³Ra/Ca-ALG MS treatment induces broad and consistent transcriptomic reprogramming in HCC cells. The substantial distance between the two groups along principal components demonstrates that α-emitter therapy exerts a significant and highly reproducible effect across biological replicates, establishing a solid foundation for subsequent identification of differentially expressed genes and pathway enrichment analysis.

Fig. 3.

Transcriptomic analysis of Huh7 cells following treatment with ²²³Ra/Ca-ALG MS. (a) PCA plot showing distinct transcriptional profiles between control and ²²³Ra/Ca-ALG MS groups. (b) Volcano plot illustrating differentially expressed genes between the two groups. Red dots represent significantly upregulated genes, and blue dots indicate downregulated genes. (c) Heatmap of differentially expressed genes associated with ICD and immune active displaying clear separation between control and ²²³Ra/Ca-ALG MS groups. (d) GO enrichment analysis revealed that the most significantly enriched biological processes in ²²³Ra/Ca-ALG MS groups. (e) KEGG pathway enrichment analysis of differentially expressed genes. (f) GSEA demonstrated significant enrichment of hallmark gene sets associated with apoptosis, immune response, and innate immune response in ²²³Ra-treated HCC cells

The volcano plot (Fig. 3b) and heatmap (Fig. 3c) further illustrated the distribution of differentially expressed genes following ²²³Ra/Ca-ALG MS treatment. A total of 2,174 genes were significantly upregulated and 588 genes were downregulated (Fig. S2). Notably, several classical Immunogenic cell death (ICD) and immune-activating genes—including Heat Shock Protein 70 (HSP70) family members (HSPA1A, HSPA1B), calreticulin (CALR), and IL-1β—were among the most highly upregulated genes. Other key inflammatory mediators such as NLRP3, TNF, and IL18 also showed prominent upregulation [28, 29]. This transcriptomic profile suggests that ²²³Ra/Ca-ALG MS not only triggers cell death but also activates multiple immune-stimulatory pathways at the molecular level.

Gene Ontology (GO) enrichment analysis (Fig. 3d) showed that upregulated genes were mainly enriched in biological processes such as “apoptotic process,” “immune response,” and “inflammatory response.” In addition, we observed enrichment in “ATP binding,” which further supports enhanced extracellular ATP signaling—a hallmark of ICD [30]. These findings provide mechanistic insights into the coordinated upregulation of cell death and immune activation pathways following ²²³Ra/Ca-ALG MS exposure.

KEGG pathway enrichment analysis (Fig. 3e) revealed significant activation of “cytokine-cytokine receptor interaction,” “TNF signaling pathway,” and “Toll-like receptor signaling pathway.” These pathways are known to regulate the release of inflammatory signals and the recruitment of immune cells, suggesting that ²²³Ra/Ca-ALG MS may “shatter” tumors via high LET, induce the release of multiple tumor antigens, promote the chemotaxis of antigen-presenting cells, and orchestrate the immunogenic reprogramming of HCC cells.

Gene set enrichment analysis (GSEA, Fig. 3f) further demonstrated significant enrichment of gene sets related to “apoptosis,” “immune response,” and “innate immune response.” This provides additional evidence that ²²³Ra/Ca-ALG MS treatment induces a comprehensive immune-activating phenotype at the transcriptomic level.

In the current era of cancer immunotherapy, against the backdrop of rapidly developing immunotherapeutic strategies and agents, we are encouraged to observe that these transcriptomic results demonstrate that ²²³Ra/Ca-ALG MS not only induces robust cytotoxicity but, more importantly, also drives the upregulation of multiple ICD markers and immune pathways. This establishes a direct mechanistic link between α-emitter TARE and immunogenic remodeling of HCC cells, suggesting that high-LET α-particle TARE could be a powerful strategy for overcoming tumor immune tolerance and enhancing antitumor immunity in HCC.

Induction of ICD by ²²³Ra/Ca-ALG MS

ICD is a regulated form of cell death that provokes a robust antitumor immune response through the release of DAMPs and the activation of antigen-presenting cells. Building upon our transcriptomic findings, we believe that α-particles, owing to their high LET, may more readily induce ICD by causing densely clustered and irreparable cellular damage in tumor cells.

To validate this ICD-associated trend, we treated Huh7 and Hep3B cells with 3.7 kBq/mL ²²³Ra/Ca-ALG MS under hypoxic conditions and collected protein samples at various time points to assess changes in key ICD markers. Western blot analysis confirmed a time-dependent upregulation of CALR and HSP70 expression following ²²³Ra/Ca-ALG MS treatment (Fig. 4a). CALR is a multifunctional chaperone that, upon translocation to the cell membrane, serves as a potent “eat-me” signal to facilitate dendritic cell phagocytosis of dying cells. HSP70 functions both as a molecular chaperone and a DAMP, enhancing immune recognition of dying tumor cells. Meanwhile, we observed a slight decrease in intracellular High Mobility Group Box 1 (HMGB1) levels, consistent with its translocation from the nucleus and passive release during ICD. As a nuclear protein, HMGB1 acts as a key DAMP when released into the extracellular space, promoting immune activation.

Fig. 4.

Evaluation of ICD induced by ²²³Ra/Ca-ALG MS treatment in HCC cells in vitro. (a-i) Western blot analysis of ICD-associated markers in Huh7 and Hep3B cell lines treated with ²²³Ra/Ca-ALG MS for 6 h, 12 h, and 24 h. (a-ii) Quantification graphs of CALR, HSP70, and HMGB1 expression in Huh7 and Hep3B cell lines treated with ²²³Ra/Ca-ALG MS for 6 h, 12 h, and 24 h. All protein levels were normalized to the β-actin (n = 3). (b-i) Representative confocal laser scanning microscopy images showing CALR exposure on the cell surface of Huh7 and Hep3B cells treated with ²²³Ra/Ca-ALG MS for 6 h, 12 h, and 24 h. CALR was visualized by immunofluorescence using a red-labeled antibody, and nuclei were counterstained with DAPI (blue). (b-ii) Quantification of CALR fluorescence intensity (n = 3) (c) Quantification of ATP concentration and extracellular ATP levels in Huh7 and Hep3B cells after treatment. Cells were treated with ²²³Ra/Ca-ALG MS, Ca-ALG MS, or left untreated as a control (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Upon observing CALR upregulation at the protein level, and considering its function as an antigenic “bait” anchored on the cell membrane, we further examined its subcellular localization. Confocal immunofluorescence imaging demonstrated that ²²³Ra/Ca-ALG MS significantly increased CALR exposure on the cell surface—a hallmark of ICD that promotes dendritic cell-mediated phagocytosis of dying tumor cells. Semi-quantitative fluorescence analysis also confirmed a marked, time-dependent increase in CALR membrane localization with sustained α-particle exposure (Fig. 4b).

Additionally, we evaluated ATP dynamics, as extracellular ATP release is a classical hallmark of ICD and efficiently attracts and activates antigen-presenting cells. After 24 h of ²²³Ra/Ca-ALG MS treatment, intracellular ATP levels were significantly reduced, while extracellular ATP release was markedly increased; no significant changes were observed in either the control or Ca-ALG MS groups (Fig. 4c).

In summary, our results demonstrate that ²²³Ra/Ca-ALG MS effectively induces ICD in HCC cells under hypoxic conditions, as evidenced by CALR exposure, upregulation of HSP70, HMGB1 release, and increased extracellular ATP levels. These findings suggest that high-LET α-particles not only exert potent direct cytotoxic effects on tumor cells but also create a favorable immunogenic microenvironment for the recruitment and activation of immune cells. This molecular mechanism lays the groundwork for further in vivo studies of its antitumor efficacy of ²²³Ra/Ca-ALG MS, particularly regarding its capacity to modulate the tumor immune microenvironment and synergize with immunotherapeutic strategies.

²²³Ra/Ca-ALG MS–based TARE and in vivo biodistribution monitoring

Recent studies have emphasized that rationally combining radioembolization with potent DC activators can help overcome the immunosuppressive TME and promote systemic antitumor immunity [31–33]. Following the confirmation of ²²³Ra/Ca-ALG MS induced immune activation, we conducted in vivo experiments combining TARE with immunotherapy using an immune adjuvant. The overall experimental protocol is illustrated in Fig. 5a. We strategically selected ODN 1826—a synthetic CpG oligodeoxynucleotide and TLR9 agonist—as the immunoadjuvant for combination with ²²³Ra/Ca-ALG MS-based TARE. The experimental groups included a sham embolization group, Ca-ALG MS group, ²²³Ra/Ca-ALG MS group, ²²³Ra/Ca-ALG MS + ODN 1826 group.

Fig. 5.

²²³Ra/Ca-ALG MS–based TARE: intra-arterial delivery and longitudinal SPECT/CT biodistribution monitoring with tumor-to-liver ratio tracking. (a) Schematic illustration of the TARE workflow and post-treatment evaluation. The diagram outlines the animal modeling strategy, group allocation for TARE administration, and subsequent imaging and cellular assessments. (b-i) ²²³Ra/Ca-ALG MS were delivered via the gastroduodenal artery (GDA) into the proper hepatic artery (PHA) supplying the tumor; the common hepatic artery (CHA) is indicated. (b-ii) H&E staining shows intrahepatic tumor formation after orthotopic N1S1 implantation. (b-iii) H&E section from the Ca-ALG MS group showing perivascular microsphere deposition with largely preserved liver lobular architecture, indicating minimal carrier-related toxicity. (b-iv) H&E section from the ²²³Ra/Ca-ALG MS group showing extensive coagulative necrosis in and around the tumor, with microspheres distributed within the necrotic zone. Scale bars: 200 μm (b-ii), 100 μm (b-iii& iv). (c) Fusion imaging of 3D-reconstructed CT images with SPECT imaging to visualize the spatial distribution of ²²³Ra/Ca-ALG MS in vivo. (d) Quantitative biodistribution of ²²³Ra/Ca-ALG MS in major organs, expressed as percentage of injected dose per milliliter (%ID/mL), at 20 d post-administration (n = 4). ****p < 0.0001. (e) Temporal evolution of the tumor-to-liver uptake ratio (T/L) at 1 h, 5 d, 10 d, and 20 d post-TARE, showing sustained tumor retention and increased selectivity in both groups (n = 4)

The rat N1S1 model was established via orthotopic tumor implantation method. H&E staining revealed typical tumor architecture and cell morphology, validating successful model establishment (Fig. 5b-ii). Subsequently, ²²³Ra/Ca-ALG MS-based TARE was performed in the orthotopic HCC model via hepatic arterial catheterization (Fig. 5b-i). Histological examination post-embolization showed that in the Ca-ALG MS group, the perivascular liver lobule structure remained largely intact, indicating minimal toxicity of the microsphere carrier itself (Fig. 5b-iii). In contrast, ²²³Ra/Ca-ALG MS embolization induced marked necrosis both around and within the tumor, confirming that α-emitting microspheres produce a cytotoxic effect far beyond that caused by mechanical vascular occlusion alone (Fig. 5b-iv).

Owing to the short path length of alpha particles (< 100 μm), direct in vivo quantification of alpha-emitting radiopharmaceutical distribution remains unfeasible. However, since alpha radionuclides or their decay products emit photons, γ-rays that can be efficiently detected by SPECT/CT, providing an effective means for noninvasive, visual, and quantitative analysis of the in vivo biodistribution of radiolabeled microspheres. In this study, SPECT/CT imaging was performed at 1 h, 5 days, 10 days, and 20 days after TARE (Fig. 5c). The imaging results confirmed the technical success of the embolization procedure, as the radioactive signal from the ²²³Ra-labeled microspheres was highly confined to the tumor region, indicating precise and effective embolization. Throughout the entire observation period, no detectable radioactivity was found in extrahepatic organs, further demonstrating the safety of the ²²³Ra/Ca-ALG MS-based TARE. Over time, as ²²³Ra undergo decay, the radioactive signal within the tumor region gradually decreased. For a more comprehensive assessment of spatial and temporal biodistribution, representative coronal and sagittal SPECT/CT images for each group at all time points are presented in Fig. S3, further confirming persistent tumor retention.

To systematically evaluate the in vivo tumor-targeting specificity and safety of ²²³Ra/Ca-ALG MS-based TARE, we performed quantitative biodistribution analysis across major organs and tissues (Fig. 5d). In nuclear medicine research, accurately assessing the radioactivity distribution in various organs is crucial—not only to confirm whether the radiolabeled microspheres are precisely embolized within the tumor vasculature, but also to directly evaluate procedural safety and postoperative complications. Our results showed that, in both the ²²³Ra/Ca-ALG MS group and the group receiving combined immunoadjuvant ODN 1826, the vast majority of radioactivity remained in the tumor, with minimal distribution in other major organs such as the heart, liver, lung, kidney, and bone. This high tumor selectivity was consistent with SPECT/CT imaging, highlighting the advantage of intra-arterial microsphere delivery for local tumor control. Additionally, we analyzed the tumor-to-liver uptake ratio (T/L ratio, Fig. 5e), which is a classical parameter in nuclear medicine imaging used to quantify the selective accumulation of radiotracers. In this context, the “tumor” (target) refers to the region of interest containing the tumor tissue, while the “liver” (background) represents the surrounding normal hepatic tissue. The T/L ratio directly reflects how effectively the therapeutic agent accumulates within the tumor relative to non-tumor liver tissue. A higher T/L ratio indicates superior tumor selectivity and reduced off-target exposure, which are critical for maximizing local efficacy and minimizing toxicity to normal liver. In our study, the T/L ratio in both groups steadily increased from 1 h to 20 days post-TARE, ultimately exceeding 15. This sustained elevation further confirms the excellent tumor-targeting efficiency and in vivo retention properties of our ²²³Ra/Ca-ALG MS-based TARE platform.

In summary, SPECT/CT imaging further validated the effectiveness of ²²³Ra/Ca-ALG MS embolization, with radioactive signals strictly confined to the tumor region. Quantitative biodistribution analysis across major organs demonstrated excellent biosafety, as well as a consistently high and increasing tumor-to-liver uptake ratio, underscoring the superior tumor selectivity and retention of the microspheres in vivo. Importantly, the addition of the immunoadjuvant ODN 1826 did not affect the radiolabeling stability of ²²³Ra on the microspheres or their embolization stability within the tumor microvasculature. These results further validate the safety, reliability, and translational potential of our ²²³Ra/Ca-ALG MS-based TARE combined immunotherapy strategy, offering a promising and safe locoregional treatment modality for HCC.

Antitumor efficacy of ²²³Ra/Ca-ALG MS–based TARE

We used 2-deoxy-2-[¹⁸F] fluoro-D-glucose (¹⁸F-FDG) positron emission tomography/X-ray computed tomography (PET/CT) to monitor tumor responses in HCC-bearing rats across different treatment groups. In vivo PET/CT scans were performed preoperatively and on days 2, 7, and 14 after treatment to longitudinally track both the ¹⁸F-FDG uptake intensity and the metabolic tumor volume (Fig. 6a-i). In the Ca-ALG MS group, a transient decrease in SUVmax was observed on day 2 after embolization, likely due to temporary metabolic suppression caused by acute vascular occlusion; however, both SUVmax and metabolic tumor volume (MTV) increased rapidly thereafter, indicating that embolization alone could not achieve durable tumor control. In the sham embolization (control) group, both tumor SUVmax and MTV continued to increase throughout the follow-up period. In contrast, both the ²²³Ra/Ca-ALG MS group and the ²²³Ra/Ca-ALG MS + ODN 1826 group exhibited sustained decreases in SUVmax and significant inhibition of tumor growth (Fig. 6a-ii&iii). To provide a more intuitive and comprehensive visualization, whole-body ¹⁸F-FDG PET/CT images—including maximum intensity projection (MIP), transverse, coronal, and sagittal views—are presented in Fig. S5, highlighting dynamic changes in intratumoral FDG uptake and overall tumor burden before and after TARE in each group.

Fig. 6.

Antitumor efficacy of ²²³Ra/Ca-ALG MS–based TARE. (a-i) Representative ¹⁸F-FDG PET/CT coronal images of orthotopic HCC-bearing rats from each group at pre-TARE, 2, 7, and 14 days post-treatment. White dashed circles indicate tumor regions. (a-ii) Quantitative analysis of tumor SUVmax at the indicated time points (n = 4). (a-iii) Tumor volume curves during the observation period for each group. (b) Representative immunohistochemical staining (and quantification) of Ki67 and TUNEL in tumor sections collected 14 days post-TARE (n = 4). (c) Representative bio-TEM images showing ultrastructural changes in vascular walls (c i–iv) and tumor cells (c v–viii) after ²²³Ra/Ca-ALG MS–based TARE. Red arrows indicate representative tissue and cellular alterations. (d) Gross images of excised tumor tissues from each group at the endpoint, demonstrating intergroup differences in tumor size and morphology. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

To further validate antitumor efficacy, TUNEL and Ki67 immunohistochemistry was performed on tumor sections collected on day 14 post-TARE (Fig. 6b). TUNEL staining demonstrated a significant increase in apoptotic cell populations in the ²²³Ra/Ca-ALG MS group compared to controls, with the combination therapy group exhibiting the strongest apoptotic signal. Correspondingly, Ki67 staining indicated that tumor cell proliferation was markedly suppressed in groups receiving ²²³Ra/Ca-ALG MS treatment, particularly with the addition of ODN 1826. Semi-quantitative analysis showed statistically significant differences in both TUNEL and Ki67 indices among groups, supporting a robust induction of apoptosis and potent inhibition of proliferation by ²²³Ra/Ca-ALG MS-based TARE and its immunomodulatory combination.

To directly visualize the ultrastructural and organelle-level damage effects of ²²³Ra/Ca-ALG MS on tumor microarchitecture, Tumor tissues were collected after treatment and examined by transmission electron microscopy (TEM). As shown in Fig. 6c, panel i illustrates ultrastructural alterations in multiple layers of the vascular wall, resulting from internal irradiation and cessation of blood flow induced by ²²³Ra-labeled microspheres embolized into the tumor-feeding arteries. The upper left arrow indicates the characteristic subendothelial layer with a continuous internal elastic membrane of the vessel; endothelial cell detachment and subsequent intravascular thrombus formation can be observed, with the central arrow pointing to fibrin within the thrombus. The lower right arrow highlights the consecutively aligned vascular smooth muscle cells. Panel ii shows damaged and fragmented mitochondria within vascular smooth muscle cells, while panel iii illustrates edematous intercellular stroma in the same cell type. Panel iv reveals extensive infiltration of inflammatory cells at the site formerly occupied by endothelial cells. The lower row of images displays various apoptotic features in tumor cells. Panel v presents the overall morphology of apoptotic tumor cells. Panel vi depicts chromatin condensation and nuclear fragmentation. Panel vii demonstrates swollen endoplasmic reticulum, and panel viii shows the residual mitochondrial membranes after cellular damage.

Consistent with the imaging findings, macroscopic examination of excised tumors at the study endpoint further confirmed the pronounced antitumor efficacy of ²²³Ra/Ca-ALG MS–based TARE. As shown in Fig. 6d, tumors from the ²²³Ra/Ca-ALG MS and combination groups were significantly smaller than those from the sham embolization and Ca-ALG MS groups. Notably, varying degrees of tumor necrosis were observed in post-embolization specimens, with the most pronounced effect seen in the combination therapy group.

Collectively, these results demonstrate that ²²³Ra/Ca-ALG MS-based TARE, either alone or in combination with the immunoadjuvant ODN 1826, delivers robust and sustained suppression of tumor progression in the orthotopic HCC model. The therapeutic benefit arises from the dual effects of direct cytotoxicity—manifested as extensive tumor cell and vascular damage at the ultrastructural level. Compared to β-emitters, α-emitting radionuclides can achieve comparable or superior antitumor efficacy at significantly lower doses, owing to their high LET, short path length, and enhanced biological effectiveness. In HCC, therapeutic efficacy has been reported with β-emitters such as ¹³¹I and ¹⁷⁷Lu at doses around 37 MBq [34, 35], whereas α-emitters, as used in our study, demonstrated significant effects at doses as low as 7.4 kBq. This stark contrast highlights the high potency and low-dose therapeutic advantage of α-emitting radionuclides in targeted radiotherapy.

Notably, the rational integration of α-particle TARE with TLR9 agonist-mediated immunostimulation further augments antitumor immunity, as evidenced by increased apoptosis, reduced tumor cell proliferation, and heightened immune activation. These findings highlight the translational potential of this strategy as a promising and comprehensive approach for advanced HCC therapy, offering both local tumor control and systemic immunological benefit.

In vivo immune activation analysis

As anticipated from the comparable tumor control observed by PET/CT imaging in both ²²³Ra/Ca-ALG MS-based TARE and combination therapy groups, subsequent evaluation of the TME revealed distinct differences in immune activation profiles across treatments (Fig. 7). Flow cytometric analysis of dissociated tumor tissues at day 7 post-TARE showed a progressive increase in CD8⁺ T cell infiltration from sham and Ca-ALG MS groups (21.4% and 23.0%, respectively) to ²²³Ra/Ca-ALG MS (27.5%), with the highest levels observed in the combination group (32.4%). A similar trend was noted for dendritic cells (DCs, CD80⁺CD86⁺), with percentages rising from 8.7% (sham) and 15.9% (Ca-ALG MS) to 20.2% (²²³Ra/Ca-ALG MS) and peaking at 29.3% in the combination group (Fig. 7a).

Fig. 7.

Immune activation by ²²³Ra/Ca-ALG MS–based TARE in vivo. (a) Flow cytometric analysis of dissociated tumor tissues at day 7 post-TARE (n = 4 ). (b) Representative immunohistochemical staining of CD8, GzmB, HSP70, CALR, and HMGB1 and quantitative analysis of the positive staining rates in tumor sections (n = 4 ). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

To further validate immune activation in vivo, immunohistochemical staining was performed for both immune effector and ICD-related markers (Fig. 7b). Quantitative analysis showed that the densities of intratumoral CD8⁺ T cells and GzmB positive cells were both markedly elevated in the ²²³Ra/Ca-ALG MS and combination groups, with the ²²³Ra/Ca-ALG MS + ODN 1826 group achieving the highest levels. This highlights a robust recruitment and activation of cytotoxic T lymphocytes, supporting the flow cytometry findings. For ICD markers, the expression of HSP70 and CALR was strongly upregulated, while HMGB1 nuclear retention was diminished, consistent with its release as a DAMP during ICD, with these effects being most pronounced in the combination group. These changes are consistent with the induction of ICD following α-emitter–mediated therapy, as both the exposure of HSP70/CALR and the release of HMGB1 are recognized hallmarks of ICD that promote antigen presentation and dendritic cell activation.

The selection of immunoadjuvant and its mechanistic synergy with the radioembolization platform remain key issues. Li et al. employed the TLR7/8 agonist R848 in combination with ¹³¹I-labeled microspheres for TARE in a rat HCC model, resulting in enhanced DC maturation and CD8⁺ T cell infiltration [36]. ODN 1826, a TLR9 agonist known to activate dendritic cells and enhance radiotherapy-induced antitumor immunity, was selected in our study. First, ODN 1826 has been extensively validated in rodent models, including rats, and is known for its high efficacy and safety. Second, compared with other CpG sequences such as ODN 1018, ODN 1826 is widely recognized as the standard sequence for robustly activating rat DCs and inducing Th1-type immune responses [37, 38]. This is particularly important in the context of ²²³Ra/Ca-ALG MS-based TARE, which induces ICD and tumor antigen release, necessitating efficient DC activation to initiate an effective antitumor immune response [39]. Furthermore, compared to R848, ODN 1826 offers a more targeted mode of immune activation, focusing on DC maturation and avoiding the risk of systemic inflammatory toxicity associated with broad-spectrum innate immune stimulators, which remains a major obstacle to their clinical application. Thus, ODN 1826 is particularly suitable for our rat ²²³Ra/Ca-ALG MS-based TARE model, achieving an optimal balance of efficacy, safety, and mechanistic synergy with radiation-induced tumor antigen release.

Specifically, based on our Western blot data, dying tumor cells gradually secrete DAMPs after TARE, reaching a peak at 24 h post-treatment (Fig. 4). Therefore, administration of the first dose of ODN 1826 at 12 h after TARE can coincide with the period of increasing antigen release, thereby enhancing the initiation of antitumor immunity. By 48 h after TARE, the immune response has entered the effector phase of T cell-mediated cytotoxicity; a second injection of ODN 1826 at this time helps maintain a Th1-polarized immune environment and supports continued recruitment of CD8⁺ T cells [40, 41]. This two-dose, time-staggered regimen is designed to fully leverage the ICD peak induced by ²²³Ra/Ca-ALG MS-based TARE, while avoiding the pitfalls of pre-irradiation CpG exposure, such as loss of activity or excessive systemic inflammation, thereby maximizing the activation and expansion of antitumor immune responses.

These comprehensive in vivo analyses provide direct mechanistic support for the notion that, although both ²²³Ra/Ca-ALG MS and its combination with ODN 1826 effectively suppress tumor growth, only the latter group achieves maximal activation of antitumor immunity within the TME. These results underscore the synergistic benefit of integrating rationally timed immunoadjuvant administration with α-emitter–mediated TARE, which not only augments the infiltration and activation of effector immune cells but also amplifies ICD-related signaling in HCC. Such immunological remodeling of the TME is expected to translate into improved long-term therapeutic efficacy and resistance to tumor recurrence.

In vivo biosafety evaluation of ²²³Ra/Ca-ALG MS-based TARE

To comprehensively assess the biosafety of ²²³Ra/Ca-ALG MS-based TARE, we performed longitudinal serum biochemistry and histopathological analysis in all treatment groups (Fig. S6). Serial measurements of serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), albumin (ALB), and creatinine (CREA) at 2, 7, 14 and 28 days post-TARE demonstrated that all values remained within the normal physiological range in each group. Mild transient elevations of AST and ALT were observed at day 2 post-TARE in both ²²³Ra/Ca-ALG MS-based TARE groups, likely reflecting acute hepatic microinjury related to embolization, but these values rapidly normalized by day 7 and remained stable thereafter. ALB and CREA levels were stable throughout the study, indicating preserved liver synthetic function and renal integrity (Fig. S6).

Histological examination of major organs (liver, heart, lung, kidney, and spleen) by H&E staining at 28 days post-treatment revealed intact architecture in all groups, with no evidence of inflammatory infiltration, necrosis, or structural damage. Except from the tumor area, the surrounding normal liver parenchyma remained preserved. Other organs exhibited no discernible pathological changes, further confirming the absence of off-target toxicity. Throughout the 42-day observation period, all groups exhibited gradual recovery or maintenance of body weight, with no significant long-term weight loss observed in the ²²³Ra/Ca-ALG MS or combination groups, indicating good systemic tolerance and acceptable safety profiles following treatment (Fig. S6).

Collectively, these results demonstrate that ²²³Ra/Ca-ALG MS-based TARE possesses a favorable safety profile, with minimal systemic toxicity and no detectable adverse effects on major organ function or structure, supporting its translational potential for clinical HCC therapy.

Conclusion

In this study, we innovatively developed ²²³Ra/Ca-ALG MS, fully harnessing the unique radiobiological advantages of high-LET α-particles. This strategy not only maintained potent cytotoxicity under severe hypoxia but also robustly induced ICD in tumor cells, thereby reshaping the tumor immune microenvironment. Leveraging this intrinsic immune activation, further combination with a TLR9 agonist significantly amplified intratumoral immune infiltration and activation, establishing a dual antitumor mechanism of “site-specific ablation plus immune-mediated clearance,” which synergistically enhanced overall therapeutic efficacy. In addition, the embolization process using these microspheres was highly precise and controllable, demonstrating stable and safe in vivo radionuclide distribution along with excellent biocompatibility. Nevertheless, this study is limited by the lack of long-term efficacy evaluation and detailed radiation dose study in large-animal models. Future work will focus on expanding the preclinical validation of α-TARE, and exploring synergistic combinations with immune checkpoint blockade to further enhance clinical translational potential.

Abbreviations

TARE

Transarterial radioembolization

HCC

Hepatocellular carcinoma

LET

Linear energy transfer

²²³Ra/Ca-ALG MS

Radium-223 doped calcium-alginate microspheres

SPECT/CT

Single-photon emission computed tomography / computed tomography

¹⁸F-FDG PET/CT

2-deoxy-2-[18F]fluoro-D-glucose positron emission tomography / computed tomography

DAMPs

Damage-associated molecular patterns

ICD

Immunogenic cell death

DNA

Deoxyribonucleic acid

ROS

Reactive oxygen species

OER

Oxygen enhancement ratio

TLR9

Toll-like receptor 9

CpG ODN 1826

Cytosine-phosphate-guanine oligodeoxynucleotide 1826

TME

Tumor microenvironment

Ca-ALG MS

Calcium-alginate microspheres

PBS

Phosphate-buffered saline

FBS

Fetal bovine serum

SEM

Scanning electron microscopy

EDS

Energy-dispersive X-ray spectroscopy

XRD

X-ray diffraction

PET/CT

Positron emission tomography / computed tomography

IC50

Half maximal inhibitory concentration

CCK-8

Cell Counting Kit-8

ATP

Adenosine triphosphate

PCA

Principal component analysis

DEGs

Differentially expressed genes

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

GSEA

Gene set enrichment analysis

CALR

Calreticulin

HSP70

Heat shock protein 70

HMGB1

High mobility group box 1

PI

Propidium iodide

FITC

Fluorescein isothiocyanate

DAPI

4’,6-diamidino-2-phenylindole

PVDF

Polyvinylidene difluoride

SDS-PAGE

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

FPKM

Fragments per kilobase of transcript per million mapped reads

SUVmax

Maximum standardized uptake value

MTV

Metabolic tumor volume

TUNEL

Terminal deoxynucleotidyl transferase dUTP nick-end labeling

TEM

Transmission electron microscopy

GzmB

Granzyme B

DCs

Dendritic cells

AST

Aspartate aminotransferase

ALT

Alanine aminotransferase

ALB

Albumin

CREA

Creatinine

T/L ratio

Tumor-to-liver ratio

Author contributions

Jinming Tian: Conceptualization, Methodology, Investigation, Formal analysis, Writing – original draft.Zeyu Zhang: Methodology, Investigation, Data curation, Writing – review & editing.Chao Cheng: Investigation, Resources, Visualization.Fei Yu: Formal analysis, Validation.Tao Wang: Resources, Supervision.Bin Cui: Supervision, Project administration.Ye Peng: Visualization, Data curation.Shuang Qiu: Validation.Fuming Wang: Investigation.Weijing Cheng: Resources, Supervision.Rong Luo: Methodology.Guorong Jia: Conceptualization, Supervision, Funding acquisition, Writing – review & editing. Changjing Zuo: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.All authors have read and approved the final manuscript.

Funding

This work was supported by the Major Research Plan of National Natural Science Foundation of China (Grant No. 92359204), National Natural Science Foundation of China (Grant No. 82302247 and 82272040) and Shanghai Pujiang Program (Grant No. 22PJD014) and Medical Research Project of Shanghai Hongkou District Health Commission (Grant No. 2303-23).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval

All animal experiments were performed in accordance with the guidelines of the Animal Experiment Ethics Committee of Changhai Hospital, Naval Medical University. The study protocol was reviewed and approved by the Committee (Approval No. CHEC (A.E)2025-001, Programme No. 2025-016).

Competing interests

The authors declare no competing interests.