Tumor cell membrane‑coated MOF‑199 nanoparticles co‑loaded with auranofin and celecoxib enhance single‑dose radiotherapy and anti‑PD‑L1 therapy in cervical cancer

Abstract

Cervical cancer (CC) remains one of the leading causes of cancer-related death among women worldwide. Radiotherapy (RT) is an integral component of multimodal treatment for CC. Beyond its direct cytotoxic effects mediated by radiation-induced DNA damage, RT can also elicit antitumor immunity by promoting the generation of reactive oxygen species (ROS) and inducing immunogenic cell death (ICD). This immunomodulatory property provides a strong rationale for combining RT with immune checkpoint inhibitors (ICIs), a combination strategy that holds promise for enhancing therapeutic efficacy in CC. However, CC cells can exploit their antioxidant defense systems to counteract the oxidative stress induced by RT. The glutathione (GSH) and thioredoxin 1 (Trx1) antioxidant systems constitute the major intracellular antioxidant barriers of CC. They can attenuate ICD by scavenging ROS to limit RT-induced antitumor immune activation, thereby diminishing the efficacy of RT and ICIs combination treatment. Additionally, RT can also activate cyclooxygenase-2 (COX-2) to promote the production and release of prostaglandin E2 (PGE2), in turn suppressing DC maturation, consequently weakening ICD-mediated activation of anti-tumor immunity. Hence, we herein designed a tumor cell membrane-coated copper-based metal-organic framework nanomaterial (MOF-199) co-delivering auranofin and celecoxib (MAC@M) for co-targeting the antioxidant system and COX-2/PGE2 pathway to augment radioimmunotherapy efficacy in CC. In vitro experiments demonstrated that MAC@M could significantly enhance RT to increase CRT exposure, HMGB1 release, and ATP secretion, as well as promote DC activation. In animal models, MAC@M combined with RT elicited robust CD8⁺ T cell and NK cell-mediated antitumor immune responses, which largely contributed to local tumor control and a pronounced systemic antitumor effect, ultimately improving ICI efficacy in CC. Collectively, our study may provide a new combinatorial treatment strategy against CC.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12951-026-05116-4.

Introduction

Cervical cancer (CC) remains one of the most common malignancies among women worldwide and continues to pose a major threat to women’s health. The World Health Organization reports that annually, approximately 660,000 new cases and 350,000 deaths are recorded globally [1]. Radiotherapy (RT) is an integral component of CC treatment. In addition to directly killing tumor cells through radiation-induced DNA damage, RT can promote immunogenic cell death (ICD) through radiation-induced oxidative stress and DNA damage [2, 3]. ICD is a specialized form of cell death capable of eliciting antitumor immune responses. During ICD, tumor cells release tumor-associated antigens (TAAs), adenosine triphosphate (ATP), and high-mobility group box 1 (HMGB1), accompanied by the exposure of calreticulin (CRT) on the cell surface [4]. These immunostimulatory signals are recognized by dendritic cells (DCs), promoting their maturation and activation, enhancing TAA uptake and cross-presentation, and subsequently initiating T cell-mediated immune responses, thereby facilitating the conversion of an immunosuppressive tumor microenvironment (TME) into an immunologically active state [5, 6]. Accordingly, RT not only provides local tumor control but also has the potential to remodel the tumor immune microenvironment and enhance antitumor immunity.

Immune checkpoint inhibitors (ICIs), which reinvigorate antitumor T-cell responses by blocking inhibitory checkpoint signaling, have emerged as an important therapeutic strategy for CC [7]. However, their efficacy is often constrained by an immunosuppressive TME characterized by low tumor immunogenicity, limited CD8⁺ T-cell infiltration, and the accumulation of immunosuppressive cell populations [8–10]. By increasing tumor immunogenicity and promoting immune activation, RT may help overcome these barriers and enhance the responsiveness of CC to ICI-based therapy [11, 12]. However, RT exerts a double-edged effect on tumor immunity, as it can stimulate antitumor immune activation while simultaneously inducing immunosuppressive events. For example, RT can activate cyclooxygenase-2 (COX-2), promoting the synthesis and release of prostaglandin E2 (PGE2), which in turn impairs DC maturation and attenuates ICD-mediated antitumor immune responses [13, 14]. Moreover, when RT induces oxidative stress in CC cells, tumor cells can exploit their endogenous glutathione (GSH)–thioredoxin 1 (Trx1) antioxidant system to counteract radiation-induced damage by scavenging reactive oxygen species (ROS), thereby undermining the immunogenic potential of RT [15, 16]. Collectively, these two mechanisms, namely insufficient generation of immunogenic tumor damage due to the intrinsic antioxidant barrier of tumor cells and the immunosuppressive feedback pathways induced by RT itself, act in concert to constitute a core obstacle that restricts RT-driven antitumor immune responses, ultimately compromising the synergistic therapeutic efficacy of RT combined with ICIs. Therefore, simultaneously relieving tumor-intrinsic antioxidant defenses and RT-induced immunosuppressive feedback may represent a critical strategy for further potentiating RT-mediated antitumor immunity and improving the therapeutic efficacy of RT in combination with ICIs.

Celecoxib, a kind of selective COX-2 inhibitor, is a commonly used nonsteroidal anti-inflammatory drug (NSAID) in clinical practice. It can suppress COX-2 activity to decrease PGE2 production in tumor cells, ultimately alleviating the immunosuppression within the TME [17]. Auranofin is a small-molecule antirheumatic agent that irreversibly disrupts the Trx1 antioxidant system to increase intracellular ROS accumulation in tumor cells, subsequently amplifying RT-induced DNA damage and ICD-elicited anti-tumor immune response [18]. Additionally, evidence has shown that the divalent Cu²⁺ can react with GSH to generate oxidized glutathione disulfide (GSSG), simultaneously being reduced to Cu⁺, which can participate in Fenton-like reactions to amplify oxidative stress and enhance ICD of tumor cells [19]. The metal-organic framework MOF-199 is a porous nanocrystalline structure self-assembled from Cu^2⁺ and the organic ligand benzene-1,3,5-tricarboxylic acid (H₃BTC) through coordination chemistry [20]. In the GSH-rich tumor microenvironment, MOF-199 undergoes responsive degradation and continuously releases Cu²⁺ to induce and augment ICD of tumor cells [21]. Of note, MOF-199 nanoparticles exhibit a certain degree of tumor-targeting capability, and coating it with tumor cell membranes can further enhance its tumor accumulation [22]. Moreover, owing to its high specific surface area, porous architecture, and favorable chemical stability, MOF-199 also serves as an ideal drug delivery carrier [23]. Based on these clues, in this study we focus on developing a CC cell membrane-coated MOF-199 nanoparticle platform co-delivering celecoxib and auranofin, and combining it with RT to enhance the therapeutic efficacy of ICIs in CC (Fig. 1).

Fig. 1.

MOF-199 nanoparticles enveloped with membranes from U14 cells and loaded with auranofin and celecoxib. These nanoparticles are designed to cooperate with RT to enhance the efficacy of PD-L1 in cervical cancer

Materials and methods

Preparation of MOF-199, A@M, C@M, and AC@M

MOF-199 was prepared as previously described [24]. Copper nitrate trihydrate (2.58 mmol) and trimesic acid (1.25 mmol) were dissolved in an ethanol/deionized water mixed solvent (1:1, 10 mL). After thorough mixing, triethylamine (250 µL) was added dropwise, and the mixture was stirred at room temperature for 30 min. The reaction solution was centrifuged, and the precipitate was washed three times with ethanol. The obtained product was dispersed in anhydrous ethanol and stored at room temperature. MOF-199 (1 mg) was ultrasonically dispersed in ethanol (1 mL), followed by the addition of auranofin (1 mg), celecoxib (1 mg), or their mixture. After stirring in the dark at 25 °C for 24 h, the products were collected by centrifugation, washed with PBS, and lyophilized to obtain A@M, C@M, and AC@M, respectively.

Preparation and characterization of M@M, MA@M, MC@M, and MAC@M

U14 cells at log phase were lysed in hypotonic buffer, frozen, thawed, and homogenized. Tumor cell membranes were obtained and stored. MOF-199, A@M, C@M, and AC@M were mixed with membranes, sonicated, and extruded. Final products (M@M, MA@M, MC@M, MAC@M) were obtained. We utilized a transmission electron microscope (TEM, Hitachi HT7800) and EDS mapping to scrutinize the morphology and elemental composition of MOF-199, AC@M, and MAC@M. For an in-depth look into the surface chemical makeup of MAC@M, we turned to X-ray photoelectron spectroscopy (XPS, Thermo Scientific ESCALAB 250Xi). To determine the crystal lattice of MOF-199, AC@M, and MAC@M, we conducted XRD analysis using an X-ray diffractometer (XRD, Bruker D8 Advance). The hydrodynamic diameters and zeta potentials of these materials were then measured with precision via dynamic light scattering (DLS, Malvern Zetasizer Nano ZS90). Cu⁺ generation was quantified via the neocuproine assay using absorbance at 457 nm. GSH depletion was determined using the DTNB assay by recording absorbance at 412 nm following incubation with GSH. Hydroxyl radical (•OH) formation was detected with MB as a probe, and absorbance at 665 nm was measured. To assess colloidal stability, 0.2 mg/mL MAC@M was dispersed in PBS and maintained at 37 °C, followed by DLS measurement of hydrodynamic diameter. The encapsulation efficiency and loading efficiency of celecoxib were determined using high-performance liquid chromatography (HPLC). Briefly, a calibration curve was constructed using a range of celecoxib concentrations from 0 to 50 mg/mL, and the amount of celecoxib loaded in the nanoparticles was then calculated accordingly. The loading efficiency and encapsulation efficiency of auranofin were assessed by inductively coupled plasma mass spectrometry (ICP-MS). Specifically, a calibration curve for Au content was established, and the Au content was measured to indirectly calculate the amount of auranofin loaded in the nanoparticles. The loading efficiency and encapsulation efficiency were calculated according to the following formulas: Loading efficiency (%) = (Amount of target drug loaded in nanoparticles / total amount of drug‑loaded nanoparticles) × 100%. Encapsulation efficiency (%) = (Amount of target drug loaded in nanoparticles / total amount of target drug initially added during preparation) × 100%.

SDS-PAGE analysis of cell membrane coating

The U14 cell membranes, AC@M, and MAC@M samples were individually combined with loading buffer before being subjected to heat treatment at 100 °C for a duration of 10 min. Following this preparation, 10 µL aliquots from each sample were carefully applied to a 10% precast Tris-glycine gel. Electrophoresis was then carried out at a constant voltage of 110 V for exactly one hour. To visualize the protein bands, the gel underwent staining with rapid Coomassie Brilliant Blue solution and subsequently captured for imaging.

Evaluation of tumor targeting and biocompatibility

U14 cells were seeded into confocal dishes and cultured overnight. FITC‑tagged AC@M or MAC@M were then added at a concentration of 50 µg/mL and incubated at 37 °C for 4 h. This was followed by a thorough rinsing with three doses of PBS. The cells were then treated with 4% paraformaldehyde for about a quarter of an hour and subsequently dyed with DAPI for ten minutes to set the stage. Once mounted with antifade solution, the specimens were put under the confocal lens (DAPI was set at 405/420 to 470 nm, and FITC was set at 488/500 to 550 nm). For the flow cytometry analysis, we collected the treated cells and noted the MFI in the FITC channel. To examine how much the cells were taking up over time, we allowed them to soak in FITC-MAC@M for periods ranging from zero to eight hours, then conducted the analysis as previously described.

As the tumor bulk neared 200 cubic millimeters, the rodents were given an intravenous shot of AC@M or MAC@M (200 micrograms per critter). Researchers carried out real-time imaging using fluorescence at the 0, 2, 4, 8, 12, 24, and 48-hour marks post-injection, with the light spectrum set at 745/800 nanometers. By the 48-hour mark, the mice were humanely euthanized, and key organs along with the tumors were harvested for further in vitro analysis. The fluorescence signals were then measured utilizing the Living Image platform.

The hemocompatibility test involved incubating 2% of mouse red blood cells with MAC@M in concentrations ranging from 20 to 100 micrograms per milliliter, at body temperature (37 °C) for two hours. Post centrifugation, the supernatant’s absorbance was quantified at a wavelength of 540 nm. To serve as references, phosphate buffered saline acted as the negative control while deionized water stood in as the positive one. To gauge the in vivo biocompatibility, we dosed BALB/c mice intravenously with MAC@M at a dose of 10 milligrams per kilogram of body weight. On the seventh day, blood samples were procured for both standard bloodwork and serum biochemical analyses, while major organs were then excised for hematoxylin and eosin (H&E) staining.

Cell culture and treatment

The U14 cell line was maintained in complete DMEM medium within a humidified incubator set at 37 °C and 5% CO2 atmosphere. All experimental procedures utilized cells that were in the exponential growth phase, ensuring optimal metabolic activity. To prepare for treatment, nanoparticles were introduced into the culture at a final concentration of 50 µg/mL and allowed to incubate for 6 h prior to exposure to X-ray radiation. The RT was administered using an RS-2000 Pro X-ray irradiator (Rad Source Technologies) operating at 160 kV and 25 mA, delivering radiation at a rate of 1 Gy/min. For irradiation, the samples were exposed to single doses of either 2–8 Gy, corresponding to their respective experimental groups, with a 0.3-mm copper filter employed during the procedure to modify the beam characteristics.

Detection of oxidative stress and DNA damage

Intracellular ROS concentrations were evaluated utilizing the DCFH-DA fluorescent probe. Mean fluorescence intensity was recorded in the FITC channel via flow cytometric analysis, while confocal microscopy was employed to capture fluorescence imagery. To determine DNA double-strand breakage, γ-H2AX immunofluorescence staining was conducted. Following DAPI nuclear counterstaining, cells were examined using confocal microscopy, and the percentage of cells exhibiting γ-H2AX positivity was determined through ImageJ software analysis.

Cell proliferation assay

U14 cell reproductive longevity was tested using colony formation assays. Cells from each treatment were counted and plated in triplicate. Cultures lasted 10–14 days, with half the medium refreshed every 3 days. Colonies were counted (≥ 50 cells) and formation rate calculated relative to controls.

Detection of immunogenic cell death markers

In this study, we gauged PGE2 and HMGB1 concentrations through ELISA procedures. To collect the cell culture supernatant, we spun the samples at 3,000 rpm for a solid 10 min at 4 degrees. The resultant absorbance readings at 450 nm were recorded. For the ATP secretion analysis, we employed a luminescence test with the CellTiter-Glo^® reagent. The surface CRT was scrutinized using immunofluorescence techniques. The cells were treated with 4% paraformaldehyde for fixation, then blocked with 5% BSA. Subsequently, they were exposed to anti-CRT primary antibodies at a 1:200 dilution, along with Alexa Fluor 555-labeled secondary antibodies, also at a 1:200 ratio, all in a fridge at 4 °C. The nuclei were stained with DAPI for clarity, and confocal microscopy was used to capture the images. The CRT-associated MFI values were then calculated with the help of ImageJ software.

Flow cytometry analysis

BALB/c mouse bone marrow cells differentiated into DCs in GM-CSF and IL-4 medium. Day 6 immature DCs were co-cultured with U14 cells. Cells were stained for CD11c, CD86, PD-L1, IL-12, and IL-10 and analyzed by flow cytometry. Tumor tissue was treated with collagenase IV and DNase I, then separated by Percoll gradient. Cells were stained for DC markers and CD8⁺ T and NK cells were identified. Samples were analyzed on a flow cytometer using FlowJo.

Animals and tumor models

The research utilized 18–22 gram female BALB/c mice, aged 6–8 weeks, sourced from Beijing Vital River. These subjects were maintained under controlled conditions with a temperature of 22 °C, 55% humidity, and a 12-hour light/dark cycle. All experimental procedures were approved by the Ethics Committee of The Second Xiangya Hospital, Central South University (approval no. 20260761). For the tumor model, U14 cells were administered by subcutaneous injection into the right armpit region. Seven days after inoculation, mice were randomly assigned to different treatment groups using a computer-generated randomization sequence. Tumor volume measurements, imaging data analyses, and histological assessments were performed by investigators blinded to the treatment assignments throughout the entire experimental period. To investigate the abscopal effect, a second set of cells was implanted in the left armpit five days later, with radiation therapy targeting exclusively the right-side tumor.

Evaluation of in vivo antitumor efficacy

Tumor-bearing mice (n = 5) were randomized into treatment groups: Control, 2 Gy, 2 Gy + PBS, 2 Gy + various nanoparticles (M@M, MA@M, MC@M, MAC@M), and 8 Gy. MAC@M (10 mg/kg) or controls were administered intravenously 6 h pre-irradiation. Tumors were measured biweekly (V = L×W²/2), with body weights recorded. Mice were sacrificed on day 17 for tumor analysis (weight, photography, collection). Survival was monitored until humane endpoints. For immune depletion, anti-CD8α/anti-asialo GM1 antibodies were administered ip 2d pre-treatment and every 3d thereafter. PD-L1 blockade used anti-PD-L1 (200 µg/mouse ip q3d ×4) combined with radiation and/or MAC@M.

Detection of apoptosis and proliferation in mouse tumor tissues

At the endpoint, tumors were fixed, dehydrated, cleared, paraffin-embedded, and sectioned. Apoptosis was assessed by TUNEL, and Ki-67 expression was measured by immunohistochemistry. Sections were imaged by confocal microscopy, and positive areas and cell fraction were quantified using ImageJ.

Statistical analysis

The results were presented as mean values ± standard deviations. All in vitro experiments were performed in triplicate to ensure reproducibility, and each in vivo cohort comprised five mice. For multi-group comparisons at a single time point, a one-way analysis of variance (ANOVA) was applied. Tukey’s multiple comparisons test was used when pairwise contrasts were conducted among all experimental groups, whereas Fisher’s least significant difference (LSD) post hoc test was employed only for pre-specified pairwise comparisons based on specific experimental hypotheses. Tumor development was analyzed using a two-way repeated measures ANOVA. Survival curves were generated with the Kaplan-Meier method and compared using the log-rank test. All statistical analyses were performed using GraphPad Prism 8.0, with statistical significance set at p < 0.05.

Results and discussion

Preparation and characterization of tumor cell membrane-coated and auranofin/celecoxib-loaded nanoparticles

MOF-199 was first prepared by coordination-driven self-assembly, followed by loading with auranofin and celecoxib. MAC@M nanoparticles were subsequently constructed by coating the drug-loaded nanoparticles with tumor cell membranes. TEM observation showed that MOF-199 exhibited a regular polyhedral structure [25]. After drug loading, the overall morphology of AC@M showed no obvious change, indicating that drug incorporation did not disrupt the basic structure of MOF-199. After further coating with tumor cell membranes, a low-electron-density membrane-like layer was observed around MAC@M, forming a typical core–shell morphology, suggesting the coating of U14 cell membranes (Fig. 2A). SDS-PAGE analysis further confirmed that the protein band profile of MAC@M was highly similar to that of the isolated tumor cell membranes, whereas AC@M showed no obvious characteristic membrane protein bands (Fig. 2B). These results indicated that cell membrane-associated proteins were effectively retained during the coating process, thereby endowing MAC@M with biomimetic properties. EDS elemental mapping showed that key elements, including Cu, Au, F, N, and S, were uniformly distributed in MAC@M (Fig. 2C). Among them, Cu was mainly derived from MOF-199, Au from auranofin, and F from celecoxib, whereas N and S were associated with cell membrane components and membrane proteins. These findings further confirmed the successful integration of the MOF-199 framework, dual-drug loading, and cell membrane coating at the elemental composition level. XPS analysis also revealed characteristic signals of Cu and Au in MAC@M (Fig. 2D–F), further supporting the effective incorporation of auranofin into the nanosystem. Quantitative analysis by ICP‑MS revealed that the loading efficiency and encapsulation efficiency of auranofin were 11.95 ± 0.76% and 19.58 ± 0.61%, respectively, whereas HPLC analysis showed that the corresponding values for celecoxib were 16.30 ± 0.45% and 25.07 ± 0.58%, respectively (Table S1). XRD analysis showed that MAC@M retained the main characteristic diffraction peaks of MOF-199 (Fig. 2G), indicating that drug loading and cell membrane coating did not markedly disrupt the crystalline structure of MOF-199. DLS showed that the hydrodynamic diameters of MOF-199, AC@M, and MAC@M increased sequentially (Fig. 2H). This change in particle size was consistent with the processes of drug loading and cell membrane coating. Meanwhile, the zeta potential gradually decreased from MOF-199 to MAC@M (Fig. 2I), suggesting the successful introduction of negatively charged components from the tumor cell membrane onto the nanoparticle surface. Notably, MAC@M demonstrated no discernible alterations in particle dimensions throughout 7-day incubation in PBS, mouse plasma, or 10% FBS medium (Figure S1), indicating good colloidal stability in different physiologically relevant media and providing a basis for subsequent in vivo applications。.

Fig. 2.

Preparation and characterization of MAC@M. (A) TEM images of MOF-199, AC@M, and MAC@M (scale bar = 100 nm). (B) SDS-PAGE of the membrane protein profiles of U14 cells, AC@M, and MAC@M. (C) EDS images of MAC@M (scale bar = 100 nm). (D–F) High-resolution XPS spectra of MAC@M (D), Cu 2p of MAC@M (E), and Au 4f of MAC@M (F). (G) XRD patterns of MOF-199, AC@M, and MAC@M. (H-I) Hydrodynamic size (H) and zeta potential (I) of MOF-199, AC@M, and MAC@M measured by DLS. (J) Representative photographs and TEM of MAC@M incubated in PBS or GSH (scale bar = 50 nm). (K) Detection of Cu⁺ generation in different groups. (L) GSH depletion by different concentrations of MAC@M. (M) Detection of hydroxyl radical generation induced by different concentrations of MAC@M

In addition to serving as a drug delivery carrier, MOF-199 itself also exhibits redox-regulatory activity [21]. In the GSH-rich tumor microenvironment, MOF-199 can undergo responsive disassembly and release Cu²⁺. Subsequently, Cu²⁺ can be reduced to Cu⁺ by GSH, accompanied by enhanced GSH depletion [26]. The generated Cu⁺ can further react with H₂O₂ via a Fenton-like reaction to induce •OH production, thereby amplifying intracellular oxidative stress in tumor cells [27]. To mimic the GSH-rich tumor microenvironment, MAC@M was co-incubated with 10 mM GSH. Compared with the PBS condition, MAC@M showed marked structural disruption after incubation with GSH for 24 h, as evidenced by collapse or disappearance of the original polyhedral morphology under TEM. The solution color also gradually changed to colorless (Fig. 2J), suggesting that GSH could trigger the disassembly of the MOF-199. Thereafter, we further investigated the in vitro release behaviors of celecoxib and auranofin from MAC@M under physiological conditions (pH 7.4) and in a simulated TME (high GSH concentration). As shown in Figure S2, both drugs exhibited low release levels at pH 7.4, whereas the presence of GSH significantly accelerated their release, demonstrating that MAC@M possesses notable GSH-responsive release properties. Kinetic fitting results further indicated that the release profiles of both drugs were best described by the first-order kinetic model, indicating a synchronized and sustained release pattern. Given that GSH can reduce Cu²⁺ to Cu⁺, Cu⁺ generation was further detected using the neocuproine assay. The results showed that the absorbance at 457 nm was markedly increased after incubation of MAC@M with GSH (Fig. 2K), indicating that Cu²⁺ in the system could be reduced to Cu⁺ by GSH. The GSH-depleting capacity of MAC@M was then evaluated using the DTNB assay. Higher MAC@M concentrations correlated with reduced absorbance at 412 nm (Fig. 2L), suggesting a reduction in the residual GSH level and indicating that MAC@M consumed GSH in a concentration-dependent manner. The Fenton-like activity of MAC@M was further evaluated using methylene blue (MB) as a •OH probe. The absorbance of MB at 665 nm gradually decreased with increasing MAC@M concentrations (Fig. 2M), suggesting that MAC@M promoted •OH generation in a concentration-dependent manner. Collectively, these results demonstrated that MAC@M nanoparticles were successfully constructed and exhibited good structural and colloidal stability, as well as GSH-responsive degradation and Fenton-like catalytic activity.

Tumor-targeting capacity and biocompatibility of MAC@M nanoparticles

To determine whether tumor cell membrane coating could enhance the uptake of nanoparticles by tumor cells, we first evaluated the cellular internalization capacity of MAC@M in vitro. Confocal microscopy showed that, compared with FITC-labeled AC@M, MAC@M-treated U14 cells exhibited markedly stronger green fluorescence signals, which were mainly distributed in the cytoplasmic region (Fig. 3A), suggesting that tumor cell membrane coating promoted nanoparticle entry into tumor cells. Flow cytometry further confirmed that, the intracellular FITC fluorescence peak in the MAC@M group was markedly shifted to the right, accompanied by a significant increase in mean fluorescence intensity (MFI) (Fig. 3B, C), indicating that tumor cell membrane coating substantially enhanced the cellular internalization efficiency of the nanoparticles. In addition, the fluorescence signal of MAC@M in U14 cells gradually increased with prolonged incubation time and reached a peak at 8 h (Fig. 3D, E).

Fig. 3.

In vitro and in vivo tumor-targeting capacity and biocompatibility of MAC@M. (A) Confocal micrographs of U14 cells following incubation with FITC-labeled AC@M or MAC@M (scale bar = 20 μm). Nuclei (DAPI) are blue, and FITC-labeled nanoparticles are green. (B, C) Flow cytometric analysis of intracellular FITC fluorescence signals in different groups (B) and the quantification of MFI. (C). (D, E) Flow cytometric analysis of intracellular FITC fluorescence signals after co-incubation of MAC@M with U14 cells for different time periods (D) and the quantification of MFI (E). (F) Fluorescent imaging of tumor-bearing mice at intervals from 0 to 48 h following intravenous administration of AC@M or MAC@M, complemented by ex vivo imaging of key organs and tumors at the 48-hour mark. (G) Hemolysis rate of MAC@M at concentrations ranging from 20 to 100 µg/mL. (H) H&E staining images in different groups

We further evaluated the in vivo tumor-targeting and accumulation of MAC@M. After tail vein injection, the fluorescence signal of MAC@M progressively accumulated at the tumor site and remained elevated over 24–48 h (Fig. 3F). Quantitative analysis showed that the tumor-associated fluorescence intensity of MAC@M was approximately 1.65-fold higher than that of AC@M at 24 h post-injection. Notably, MAC@M maintained a higher level of tumor accumulation even at 48 h, with a fluorescence intensity approximately 1.91-fold that of AC@M (Figure S3). Following a 48-hour ex vivo imaging procedure, the fluorescence intensity of MAC@M in tumor tissue turned out to be markedly higher than what was observed with AC@M. However, this contrast was not as noticeable in non-target organs, where the distribution of the fluorescent signal was lower (Fig. 3F). These results indicated that tumor cell membrane coating not only enhanced the in vitro cellular uptake of MAC@M, but also improved its in vivo tumor accumulation and retention. The enhanced tumor-targeting capability may primarily be attributed to functional proteins expressed on the tumor cell membrane, which confer homotypic recognition and adhesion properties to the nanoparticles and facilitate preferential interactions with tumor cells of the same origin. In addition, the membrane coating may reduce nonspecific interactions between the nanoparticles and normal tissues, potentially prolonging their systemic circulation [28, 29].

On this basis, we further evaluated the biocompatibility of MAC@M to determine its suitability for intravenous administration. The hemolysis assay showed that, the hemolysis rate induced by MAC@M remained below 5% and was markedly lower than that in the deionized water positive control group (Fig. 3G), indicating good hemocompatibility. H&E staining revealed no significant pathological changes in major organs across all mouse groups (Fig. 3H). In addition, hematological parameters and serum biochemical indices showed no obvious abnormalities (Figure S4), suggesting that MAC@M did not induce significant hematological toxicity or liver and kidney dysfunction in vivo. Collectively, these results demonstrated that MAC@M nanoparticles exhibited favorable tumor-targeting capacity and in vivo safety, providing a reliable basis for subsequent in vitro and in vivo antitumor studies in combination with RT.

MAC@M nanoparticles enhance RT-mediated cytotoxicity against CC cells in vitro

Multiple studies have verified that RT with a single dose of 2 Gy can not only inhibit tumor cell proliferation and induce tumor cell death to a certain extent, but also provide better tissue-protective advantages by reducing damage to surrounding normal tissues caused by high-dose RT [30]. Furthermore, single dose 2 Gy irradiation can also participate in the remodeling of the tumor immune microenvironment by inducing ROS generation, promoting DNA damage, and enhancing tumor antigen release, thereby providing the possibility for combination immunotherapy [31, 32]. However, when used alone, RT induces only limited oxidative stress, which is often insufficient to produce sustained and potent tumor-killing effects. The primary reason for this restriction can be attributed to the potent antioxidant defense system of tumor cells, with the GSH and Trx1 mechanisms acting as pivotal barriers in preserving cellular redox balance [15]. The aforementioned results suggest that MOF-199 may enhance oxidative stress by promoting GSH-mediated degradation and similar Fenton reactions. In addition, auranofin can inhibit thioredoxin reductase (TrxR) activity, disrupt the Trx1 antioxidant system, and further reduce the ability of tumor cells to scavenge ROS [33]. Therefore, MAC@M may theoretically interfere with both the GSH and Trx1 antioxidant pathways, weaken the oxidative stress defense of tumor cells, and further enhance RT-induced ROS accumulation, DNA damage, and tumor cell killing.

To verify this hypothesis, we first used the DCFH-DA fluorescent probe to detect ROS levels. The 2 Gy group exhibited an increase in intracellular green fluorescence, indicating that a single 2 Gy dose of irradiation could induce a certain degree of ROS accumulation. Compared with the 2 Gy group, intracellular ROS levels were elevated to varying extents in the 2 Gy + M@M, 2 Gy + MA@M, and 2 Gy + MC@M groups, with the 2 Gy + MAC@M group showing the strongest fluorescence signal and the highest MFI value (Fig. 4A, D). These results suggested that MAC@M could effectively enhance intracellular oxidative stress induced by a single 2 Gy dose of irradiation. The accumulation of ROS can trigger more oxidative damage to DNA, and one of the primary mechanisms by which RT eliminates cells is the induction of DNA double-strand breaks [34]. γ-H2AX is a well-known indicator of DNA double-strand fractures [35]. Based on this, we further evaluated the level of DNA damage by γ-H2AX immunofluorescence staining. The results showed that γ-H2AX-positive signals were weak in the Control group, whereas the 2 Gy group showed increased γ-H2AX signals, indicating that a single 2 Gy dose of irradiation could induce a certain degree of DNA damage. In contrast, the 2 Gy + MAC@M group exhibited the strongest nuclear γ-H2AX fluorescence signal and the highest proportion of γ-H2AX-positive cells (Fig. 4B, E). These results indicated that MAC@M not only enhanced ROS accumulation induced by a single 2 Gy dose of irradiation, but also further aggravated DNA double-strand breaks, thereby increasing the sensitivity of CC cells to a single 2 Gy dose of irradiation.

Fig. 4.

MAC@M nanoparticles enhance RT-induced oxidative stress and DNA damage in vitro and inhibit cervical cancer cell proliferation. (A, D) Images of DCFH-DA fluorescence staining (A) and the quantification of MFI (D) in cells from different groups, used to assess intracellular ROS levels. (B, E) Images of γ-H2AX immunofluorescence staining (B) and the quantification of cells (E) from different groups. Nuclei (DAPI) are blue, and γ-H2AX signals are red. (C, F) Images of colony formation assays (C) and the quantification of colony numbers (F) in cells from different groups.^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

To assess the long-term proliferative potential and survival abilities of cells under various experimental conditions, we conducted a colony formation assay. The number of colonies was reduced in the 2 Gy group, suggesting that a single 2 Gy dose of irradiation alone inhibited cell proliferation to some extent. Compared with the 2 Gy group, the colony-forming ability was further decreased in the 2 Gy + M@M, 2 Gy + MA@M, and 2 Gy + MC@M groups, whereas the 2 Gy + MAC@M group exhibited the lowest number of colonies and the most pronounced inhibitory effect (Fig. 4C, F). These results indicated that MAC@M combined with a single 2 Gy dose of irradiation significantly impaired the long-term proliferative potential of U14 cells. Collectively, these findings demonstrated that MAC@M combined with a single 2 Gy dose of irradiation significantly increased intracellular ROS levels in CC cells, aggravated DNA damage, and further inhibited colony formation, thereby exhibiting a clear radio-sensitizing effect.

MAC@M nanoparticles augmented RT-induced ICD of CC cells in vitro

ICD is a unique type of cell death that can trigger an immune response against tumors. It is marked by the release or exposure of DAMPs from dying or stressed tumor cells, primarily involving the release of CRT, ATP, and HMGB1 [4]. These DAMPs can be recognized by DCs, thereby promoting DC maturation, antigen uptake, and antigen presentation, and subsequently initiating antitumor immune responses [36]. The above findings showed that MAC@M combined with a single 2 Gy dose of irradiation significantly enhanced ROS accumulation and DNA damage in CC cells, both of which are important bases for ICD induction [2]. Interestingly, RT exerts bidirectional immunomodulatory effects. Studies have shown that RT can upregulate COX-2 expression and promote PGE2 release by inducing oxidative stress, DNA damage, and inflammatory responses in tumor cells, thereby forming an immunosuppressive feedback loop [37]. As an important immunosuppressive mediator downstream of the COX-2 pathway, PGE2 can inhibit DC maturation and antigen presentation, thereby weakening ICD-mediated antitumor immune responses [38]. Therefore, suppressing PGE2 release while enhancing tumor cell damage induced by a single 2 Gy dose of irradiation may be more conducive to amplifying ICD effects and promoting DC activation.

We first measured PGE2 levels in the cell culture supernatants from different treatment groups. PGE2 secretion was increased in the 2 Gy group, suggesting that a single 2 Gy dose of irradiation may activate the COX-2/PGE2 pathway while inducing tumor cell damage. In contrast, PGE2 levels were markedly reduced in the celecoxib-containing treatment groups, with the greatest decrease observed in the 2 Gy + MAC@M group, where the PGE2 level was approximately 0.51-fold that of the 2 Gy group (Fig. 5A). These results indicated that MAC@M could effectively suppress PGE2 release by delivering celecoxib, thereby alleviating PGE2-mediated immunosuppression and creating more favorable conditions for the subsequent conversion of ICD signals into DC activation. We next examined the release and exposure of DAMPs. ATP release, HMGB1 release, and CRT exposure represent the major “find-me” signal, inflammatory danger signal, and “eat-me” signal during ICD, respectively [39–41]. The results showed that, compared with the 2 Gy group, the 2 Gy + MAC@M group exhibited markedly increased ATP release and further elevated HMGB1 release in the cell culture supernatant (Fig. 5B, C). Meanwhile, immunofluorescence staining of CRT showed that the 2 Gy + MAC@M group had the strongest CRT fluorescence signal on the cell surface, and quantitative analysis likewise demonstrated the highest level of CRT exposure (Fig. 5D and Figure S5). These results indicated that MAC@M markedly enhanced DAMP release and exposure induced by a single 2 Gy dose of irradiation, suggesting that the combined treatment more effectively induced ICD in U14 cells.

Fig. 5.

MAC@M nanoparticles enhance RT-induced immunogenic cell death in cervical cancer cells in vitro by promoting DAMP exposure/release, thereby further facilitating DC activation. (A, C) ELISA was used to measure PGE2 (A) and HMGB1 (C) of different groups. (B) A luminescence assay kit was used to measure ATP of different groups. (D) Representative confocal fluorescence images of CRT exposure on the cell surface in different groups. DAPI-stained nuclei are blue, and CRT signals are red. (E–F) Flow cytometric of CD86 (E) and IL-12 (F) expression in BMDCs after stimulation with tumor cells of different groups. ns, no significant difference, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

To further determine whether the enhanced ICD signals could be translated into DC activation, U14 cells subjected to different treatments were co-cultured with BMDCs, and DC maturation and functional status were analyzed. BMDCs stimulated with tumor cells treated with 2 Gy + MAC@M exhibited the highest level of CD86 expression (Fig. 5E), indicating markedly enhanced DC maturation. Meanwhile, IL-12 expression was also significantly increased in the 2 Gy + MAC@M group (Fig. 5F). IL-12 is an important pro-inflammatory cytokine that promotes T-cell activation and antitumor immune responses, and its increase indicated that DCs not only acquired a more mature phenotype, but also exhibited further enhanced immune-activating function [42]. In addition, after stimulation with tumor cells treated with 2 Gy + MAC@M, the expression levels of PD-L1 and IL-10 in BMDCs were both markedly decreased (Figure S6). Among them, PD-L1 is an important immune checkpoint molecule, and its high expression can inhibit T-cell activation through the PD-1/PD-L1 axis [43]. IL-10, in contrast, is a typical immunosuppressive cytokine that can impair the antigen-presenting capacity of DCs and their secretion of pro-inflammatory factors [44]. Therefore, the decreased expression of PD-L1 and IL-10 suggested that MAC@M combined with a single 2 Gy dose of irradiation not only promoted DC maturation and activation, but also attenuated the immunosuppressive phenotype of DCs, thereby making them more favorable for subsequent T-cell activation and the initiation of antitumor immune responses. Collectively, these results demonstrated that MAC@M combined with a single 2 Gy dose of irradiation could enhance oxidative damage and ICD in CC cells, while suppressing PGE2-associated immunosuppressive signaling and further promoting DC maturation and activation, thereby more effectively converting tumor cell damage induced by a single 2 Gy dose of irradiation into antitumor immune activation.

MAC@M nanoparticles sensitized RT to resist CC in vivo

The above in vitro experiments showed that MAC@M enhanced ROS accumulation, DNA damage, and ICD induced by a single 2 Gy dose of irradiation, and promoted DC activation. Given its favorable in vivo tumor accumulation and biocompatibility, we further evaluated the in vivo antitumor efficacy of MAC@M combined with a single 2 Gy dose of irradiation in a U14 tumor-bearing mouse model.

After treatment, marked differences in excised tumor size were observed among the different groups. Tumors in the Control group were relatively large, indicating continuous tumor progression. Treatment with 2 Gy irradiation alone or nanoparticles alone only delayed tumor growth to a certain extent. In contrast, the 2 Gy + MAC@M group showed the smallest excised tumors, suggesting that the combination of MAC@M and a single 2 Gy dose of irradiation produced a stronger local tumor-inhibitory effect (Fig. 6A). Further analysis of the tumor growth curves showed that the tumor volume in the control group increased rapidly over time, whereas tumor growth in the 2 Gy + MAC@M group was significantly suppressed (Fig. 6B). Tumor volume analysis on day 17 similarly demonstrated that the final tumor volume in the 2 Gy + MAC@M group was significantly lower than that in the control group, corresponding to a tumor inhibition rate of 83.03% (Figure S7). These results indicated that MAC@M effectively enhanced the tumor-control effect of a single 2 Gy dose of irradiation in vivo. Survival analysis further supported these findings. Kaplan–Meier curves showed that, mice in the 2 Gy + MAC@M group had markedly prolonged survival, suggesting that the combined treatment not only delayed tumor progression but also improved the overall survival outcome of tumor-bearing mice (Fig. 6C). Combined with the above in vivo fluorescence imaging results, the effective accumulation and retention of MAC@M at the tumor site may provide an important basis for its enhancement of the in vivo therapeutic efficacy of a single 2 Gy dose of irradiation.

Fig. 6.

MAC@M nanoparticles sensitize RT against cervical cancer in vivo. (A) Representative photographs of excised tumors (n = 5). (B, C) Tumor growth curves during the 0–17 day observation period (B), and survival curves in different groups (C) (n = 5). (D) Representative TUNEL staining images of tumor tissues in different groups, used to assess the level of apoptosis. DAPI-stained nuclei are blue, and TUNEL-positive signals are pink. (E) Representative Ki-67 immunohistochemical staining images of tumor tissues

To further validate the in vivo antitumor effect of MAC@M combined with a single 2 Gy dose of irradiation at the histological level, TUNEL staining and Ki-67 immunohistochemistry were performed on tumor tissues. TUNEL staining showed that apoptotic signals were weak in the Control group, whereas a certain increase in TUNEL-positive signals was observed in the monotherapy groups. Notably, the 2 Gy + MAC@M group exhibited the strongest TUNEL-positive signal (Fig. 6D and Figure S8A), indicating that the combined treatment more effectively induced tumor cell death. Ki-67 immunohistochemistry showed that Ki-67 expression was high in the Control group, indicating active tumor cell proliferation. In contrast, the proportion of Ki-67-positive cells was markedly reduced in the 2 Gy + MAC@M group (Fig. 6E and Figure S8B), suggesting that the combined treatment significantly inhibited tumor cell proliferative activity. Taken together, these in vivo findings demonstrate that MAC@M combined with a single 2 Gy dose of irradiation effectively suppresses tumor progression, prolongs survival, and induces apoptosis while inhibiting proliferation. Despite the remarkable antitumor efficacy of MAC@M combined with RT observed in this study, several challenges remain to be addressed prior to clinical translation. Given its excellent catalytic activity and TME-responsive degradation properties, the long-term biosafety of MAC@M nonetheless requires systematic evaluation, particularly with respect to chronic toxicity following repeated administration, pharmacokinetic profiles, potential damage to normal tissues, and the risk of immune-related adverse effects. In addition, further optimization of scalable manufacturing processes, quality control standards, pharmacokinetic behavior, and tumor-targeting delivery efficiency is essential to support its clinical development. Overall, MAC@M represents a promising radiosensitizing nanoplatform for immunoradiotherapy, yet its clinical implementation will necessitate rigorous and systematic preclinical validation.

The immune mechanisms for the enhanced effect of MAC@M nanoparticles and RT in resisting CC in vivo

The above in vitro results showed that MAC@M combined with a single 2 Gy dose of irradiation enhanced ICD in CC cells and alleviated immunosuppression by inhibiting PGE2 release. To further determine whether the combined treatment could remodel the tumor immune microenvironment in vivo, we first examined ICD-related signals and DC maturation in tumor tissues. Confocal microscopy showed that, compared with the Control group, either 2 Gy irradiation alone or nanoparticle treatment enhanced CRT fluorescence signals in tumor tissues to varying degrees. Among all groups, the 2 Gy + MAC@M group exhibited the strongest CRT signal, with a mean fluorescence intensity significantly higher (Fig. 7A, B). These results suggested that MAC@M combined with a single 2 Gy dose of irradiation could also enhance ICD-associated “eat-me” signals in vivo, thereby providing stronger stimulation for antigen-presenting cells to recognize and engulf damaged tumor cells. We then measured PGE2 levels in tumor tissues. PGE2 is an important immunosuppressive mediator downstream of the COX-2 pathway and can inhibit DC maturation and antigen presentation, thereby weakening ICD-mediated immune activation [45]. The results showed that PGE2 were increased in the 2 Gy group, suggesting that tumor damage induced by a single 2 Gy dose of irradiation may be accompanied by PGE2‑related immunosuppressive feedback. In contrast, PGE2 levels were markedly reduced in the celecoxib-containing treatment groups, with the most pronounced decrease observed in the 2 Gy + MAC@M group (Fig. 7C). These results suggested that MAC@M may suppress local PGE2 elevation by delivering celecoxib, thereby alleviating PGE2-mediated immunosuppression and creating a more favorable tumor microenvironment for subsequent DC maturation and effector immune cell activation.

Fig. 7.

MAC@M nanoparticles combined with RT induce immunogenic cell death and promote DC maturation in vivo. (A–B) Representative confocal fluorescence images of CRT exposure in tumor tissues from mice in different groups (A) and quantitative analysis of the corresponding MFI (B, n = 5). DAPI-stained nuclei are blue, and CRT immunofluorescence signals are red. (C) ELISA was used to measure PGE2 in tumor tissues of different groups (n = 5). (D–E) Flow cytometry was used to analyze CD80 and CD86 on tumor-infiltrating dendritic cells in different groups (D) and the quantification of CD80⁺CD86⁺ DCs (E, n = 5). ns, no significant difference, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

DC maturation is a key process linking tumor cell ICD to the subsequent initiation of antitumor immunity. DAMPs released or exposed during ICD can promote the uptake of tumor antigens by DCs and induce the upregulation of co-stimulatory molecules, thereby enhancing antigen-presenting capacity [46]. Therefore, we further analyzed the maturation status of tumor-infiltrating DCs by flow cytometry. The results showed that the proportion of CD80⁺CD86⁺ DCs in tumor tissues was significantly upregulated in the 2 Gy + MAC@M group, with an increase of approximately 40.77% relative to that in the control group (Fig. 7D, E and Figure S9A). These findings indicated that MAC@M combined with a single 2 Gy dose of irradiation effectively promoted DC maturation in vivo, laying the foundation for the subsequent activation of adaptive and innate immune effector cells. The above findings indicated that ICD could promote the uptake of tumor antigens by DCs and induce DC maturation. Mature DCs can further activate CD8⁺ T cells through antigen cross-presentation and co-stimulatory signaling, driving their differentiation into effector T cells. CD8⁺ T cells are key effector cells in adaptive antitumor immunity and, once activated, can secrete IFN-γ and directly mediate tumor cell killing [47]. Meanwhile, PGE2 can inhibit T-cell activation, proliferation, and effector function, thereby weakening CD8⁺ T cell-mediated antitumor immune responses [48]. Therefore, we further examined the functional status of tumor-infiltrating CD8⁺ T cells. The results showed that the 2 Gy + MAC@M group exhibited the highest IFN‑γ expression level in these cells (Fig. 8A and Figure S9B), and the percentage of IFN‑γ⁺CD8⁺ T cells was increased by approximately 22.7% compared with the control group (Fig. 8B). Additionally, further examination revealed that the CD8⁺ T cell infiltration level in tumor tissues was increased by approximately 29.2% in the 2 Gy + MAC@M group relative to the control group (Figure S10A, B). These results indicated that MAC@M synergized with a single 2 Gy dose of irradiation to promote both tumor infiltration and effector activation of CD8⁺ T cells.

Fig. 8.

MAC@M nanoparticles combined with RT enhance antitumor immune responses in vivo, thereby inhibiting tumor progression. (A–B) Flow cytometric was used to analyze IFN-γ in CD8⁺ T cells (A) and the quantification of IFN-γ⁺CD8⁺ T cells (B) in different groups (n = 5). (C–D) Flow cytometric was used to analyze GZMB expression in NK cells (C) and the quantification of GZMB⁺NK1.1⁺ cells (D) in different groups (n = 5). (E) Representative photographs of excised tumors (n = 5). (F, G) Tumor growth curves (F), and Kaplan–Meier survival curves (G) in different groups (n = 5). ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

In addition to CD8⁺ T cells, NK cells are also important effector cells in antitumor immunity. NK cells can directly kill tumor cells without antigen-specific recognition. In cervical cancer, NK cells are important innate immune effector cells involved in tumor immune surveillance and cytotoxic killing, and their infiltration level and functional status are closely associated with antitumor immune efficacy [49]. Studies have shown that mature DCs can promote NK cell activation by secreting cytokines such as IL-12, thereby enhancing their cytotoxic function [50]. Meanwhile, PGE2 can also inhibit NK cell recruitment, activation, and granzyme release, thereby weakening NK cell-mediated tumor killing [51]. Therefore, we further analyzed the activation status of NK cells. The results showed that GZMB expression in NK cells was strongest in the 2 Gy + MAC@M group (Fig. 8C and Figure S9C), and the proportion of GZMB⁺NK1.1⁺ cells was also the highest (Fig. 8D). Meanwhile, further analysis revealed that NK1.1⁺ NK cell infiltration in the 2 Gy + MAC@M group was increased by approximately 21.4% relative to the control group (Figure S9D and Figure S10C, D). These results indicated that MAC@M synergized with a single 2 Gy dose of irradiation to significantly promote NK cell infiltration into tumors and enhance GZMB-mediated cytotoxic function. To further verify whether CD8⁺ T cells and NK cells were involved in the antitumor effect of MAC@M combined with a single 2 Gy dose of irradiation, in vivo immune cell depletion experiments were performed using anti-CD8 and anti-NK1.1 antibodies on the basis of 2 Gy + MAC@M treatment. Compared with the 2 Gy + MAC@M group, depletion of either CD8⁺ T cells or NK cells led to markedly increased tumor volume and reduced survival benefit. Moreover, simultaneous depletion of both CD8⁺ T cells and NK cells resulted in the most pronounced tumor progression and further shortened survival (Fig. 8E–G and Figure S11). These results demonstrated that both CD8⁺ T cells and NK cells contributed to the in vivo antitumor effect mediated by MAC@M combined with a single 2 Gy dose of irradiation.

RT and MAC@M nanoparticles cooperate to elicit abscopal anti-tumor effect and enhance immunotherapy efficacy

Whether local RT can further induce systemic antitumor immune responses is an important aspect for evaluating the translational potential of radioimmunotherapy combination strategies. Based on the above results, MAC@M combined with a single 2 Gy dose of irradiation enhanced ICD signals, promoted DC maturation, and activated the effector functions of CD8⁺ T cells and NK cells. Therefore, we further established a bilateral tumor model to investigate whether the combination of MAC@M and single-dose 2 Gy irradiation could elicit systemic antitumor activity. Specifically, U14 cells were first subcutaneously inoculated into the right flank of mice to establish primary tumors, and the same number of U14 cells was subcutaneously inoculated into the left flank 5 days later to establish secondary tumors. During treatment, only the right-sided primary tumors received local 2 Gy irradiation, whereas the left-sided secondary tumors did not receive direct irradiation, and non-target areas were shielded with a lead plate to minimize the effect of scattered radiation dose on secondary tumors. This model can be used to evaluate systemic immune responses induced by local treatment and their inhibitory effects on distant lesions [52].

The results showed that, in the right-sided primary tumors, tumors in the Control group grew rapidly and continuously, whereas a single 2 Gy dose of irradiation alone or MAC@M alone produced only limited inhibitory effects. In contrast, primary tumor growth was markedly suppressed in the 2 Gy + MAC@M group, and both the individual tumor growth curves and the mean tumor volume curves showed a stronger local tumor-control effect (Fig. 9A, C). More importantly, in the left-sided secondary tumors that did not receive direct irradiation, tumor growth was also markedly delayed in the 2 Gy + MAC@M group, with a greater inhibitory effect than that observed in the monotherapy groups (Fig. 9B, D). Kaplan–Meier survival analysis further showed that mice in the 2 Gy + MAC@M group had significantly prolonged survival, with greater survival benefit than the other treatment groups (Fig. 9E). These results indicated that MAC@M combined with a single 2 Gy dose of irradiation not only enhanced local control of irradiated tumors, but may also synergistically suppress the progression of unirradiated distant lesions through both systemic antitumor immune responses and direct effects resulting from nanoparticle accumulation in distant tumors. Moreover, in the established lung metastasis model, the combination regimen similarly significantly reduced the number of pulmonary metastatic foci, further confirming its inhibitory effect on metastatic lesions (Figure S12). Collectively, the combination of MAC@M and single-dose 2 Gy irradiation not only achieves favorable local therapeutic efficacy against primary tumors, but also suppresses the growth and metastasis of non-irradiated distant lesions, suggesting systemic antitumor activity and prolonging overall survival in bilaterally tumor‑bearing mice.

Fig. 9.

RT and MAC@M nanoparticles cooperate to elicit potent abscopal antitumor effects and enhance immunotherapy efficacy. (A–B) Individual tumor growth curves of primary tumors (A) and secondary tumors (B) (n = 5). (C–D) Mean tumor growth curves of primary tumors (C) and secondary tumors (D) (n = 5). (E) Kaplan–Meier survival curves from different groups (n = 5). (F) Representative photographs of excised tumors (n = 5). (G, H) Tumor growth curves (G), and Kaplan–Meier survival curves(H) in different groups (n = 5). ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

Having confirmed that the combination of MAC@M and a single 2 Gy dose of irradiation could trigger an abscopal effect, we next investigated whether this combination could augment the efficacy of checkpoint blockade immunotherapy. MAC@M combined with a single 2 Gy dose of irradiation promoted ICD, DC maturation, and effector immune cell activation, suggesting that it might convert the tumor microenvironment into a more immunologically activated state suitable for anti-PD-L1 therapy. Based on this, anti-PD-L1 treatment was further combined with MAC@M and a single 2 Gy dose of irradiation in this study. The results showed that anti-PD-L1 monotherapy produced a certain degree of tumor inhibition, while tumor growth was further delayed in the 2 Gy + anti-PD-L1 and MAC@M + anti-PD-L1 groups compared with the anti-PD-L1 monotherapy group. In contrast, the 2 Gy + MAC@M + anti-PD-L1 group exhibited the smallest excised tumors, the slowest tumor growth rate, and the lowest endpoint tumor volume, with an inhibition rate of approximately 93.12% (Fig. 9F, G and Figure S13). Kaplan–Meier survival analysis also showed that mice in the 2 Gy + MAC@M + anti-PD-L1 group achieved the most pronounced survival benefit (Fig. 9H). These results indicated that MAC@M combined with a single 2 Gy dose of irradiation could further enhance the efficacy of anti-PD-L1 and improve the survival outcome of tumor-bearing mice. The above findings demonstrate that MAC@M combined with a single 2-Gy irradiation not only elicits systemic antitumor immune responses through local treatment, but also sensitizes tumors to anti-PD-L1 immunotherapy. Mechanistically, MAC@M dismantles the intrinsic antioxidant defense barrier of tumor cells and the immunosuppressive feedback loop induced by RT, thereby establishing a mechanistically complementary dual-pathway immunopotentiation modality that effectively overcomes the insufficient immune activation associated with current radioimmunosensitization strategies. Collectively, the combination of MAC@M and RT exhibits marked antitumor efficacy and immune enhancement in a subcutaneous tumor-bearing model. However, to facilitate clinical translation, further studies are warranted in orthotopic CC models using a more clinically relevant fractionated RT regimen, in conjunction with ICB, to systematically evaluate and optimize this combinatorial therapeutic strategy.

Conclusion

In summary, this study developed a tumor cell membrane-coated copper-based metal-organic framework nanomaterial based on MOF-199 for the co-delivery of auranofin and celecoxib (MAC@M). In combination with RT, MAC@M could enhance oxidative stress, ICD-associated immune activation, and anti-PD-L1 therapeutic efficacy in a subcutaneous CC tumor model. Overall, this combination strategy provides a promising new approach for the treatment of CC. Nevertheless, further studies are warranted to perform direct target validation, evaluate long‑term biosafety, investigate drug release behavior, and employ more clinically relevant animal models, in order to further elucidate the immunomodulatory mechanisms and assess the feasibility of clinical translation.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

C.F. conceived the study, designed the MAC@M nanoplatform, performed in vitro experiments, and wrote the manuscript. J.W. synthesized and characterized the nanomaterial, conducted in vivo animal studies, and prepared all figures. Y.Z. carried out immune response analyses, including T cell and NK cell assays. Y.F. supervised the project, secured funding, provided conceptual guidance, and is the corresponding author. All authors reviewed and approved the manuscript.

Funding

This work was supported by Hunan Provincial Key R&D Program (2026AQ2002), Hunan Provincial Natural Science Foundation, Industry-Joint Fund (2026JJ81334), Hunan Provincial Natural Science Foundation, General Program (2022JJ30868), Beijing Science and Technology Innovation Medical Development Foundation (KC2023-JX-0186-RQ059), Beijing Life Oasis Public Welfare Service Center (cphcf-2023-035), Leling Linghang Program Research Fund [18], Beijing zhongkanglian Public Welfare Foundation (zk1030220250402822-KY-0001), Wu jieping Medical Foundation (320.6750.2024-13-108), and Tianshu Linghang Fund for Frontier Exploration in Tumor Immunotherapy and Safety (CETSDCORP311-09).

Data availability

Data will be made available on request.

Declarations

Competing interests

The authors declare no competing interests.