Biomimetic Cu_2−xSe nanoplatforms for efficient glioblastoma treatment: overcoming the blood-brain barrier and boosting Immunogenetic cell death

Abstract

Glioblastoma (GBM) is an aggressive and highly heterogeneous brain tumor that continues to pose a significant clinical challenge. Current therapeutic strategies, including surgical resection, radiotherapy, and chemotherapy, are hindered by the tumor’s invasive behavior, resistance to treatment, and the difficulty of selectively targeting tumor cells. Emerging modalities, such as immunotherapy and photodynamic therapy, hold considerable promise; however, their efficacy in treating GBM is limited by critical barriers, including poor penetration of the blood-brain barrier (BBB), tumor heterogeneity, and insufficient accumulation of therapeutic agents at the tumor site. In this study, innovative biomimetic copper selenide nanoparticles (CS@CM) are developed for targeted photothermal therapy of GBM. These nanoparticles are functionalized with glioma cell membranes (CM), and this biomimetic design leverages the homing capability of the membranes to achieve efficient BBB penetration and enhanced targeting of GBM tissues. CS@CM act as potent photothermal agents upon light activation, which can amplify reactive oxygen species-induced oxidative stress to damage glioma cells. Such combination therapy effectively triggers immunogenic cell death to achieve splendid antitumor efficacy, offering a promising therapeutic strategy for GBM. Collectively, this approach addresses the limitations of conventional treatments, paving the way for improved clinical outcomes in managing this formidable malignancy.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-03696-1.

Introduction

Glioblastoma (GBM) is the most common and aggressive primary brain tumor, classified as a high-grade malignant glioma [1–3]. Its pronounced heterogeneity, rapid proliferation, and extensive invasiveness present significant challenges for effective clinical management [4, 5]. While surgical resection remains the cornerstone of treatment, complete removal is nearly unattainable due to the diffuse and infiltrative nature of the tumor, which obscures clear boundaries with surrounding brain tissue [6, 7]. Radiotherapy and chemotherapy can transiently slow tumor progression; however, their efficacy is limited and often accompanied by damage to healthy brain tissue [8, 9]. Despite advancements in targeted therapies and immunotherapies, clinical outcomes for GBM remain dismal. These approaches frequently fail to account for the diverse subpopulations of tumor cells, resulting in incomplete eradication and eventual tumor recurrence [10, 11]. Addressing these challenges requires a deeper understanding of GBM’s molecular mechanisms, optimization of existing treatment modalities, and the development of innovative therapeutic strategies to improve patient outcomes [12].

Reactive oxygen species (ROS) are highly reactive molecules containing oxygen, including free radicals like superoxide (·O₂⁻), hydroxyl radical (·OH), and non-radicals like hydrogen peroxide (H₂O₂), which are generated during cellular metabolism, and play a pivotal role in cellular signaling, gene expression, and immune defense [13, 14]. On the other hand, when produced in excess, ROS can induce oxidative stress and leads to cellular damage [15]. Traditional modalities for GBM treatment, including radiotherapy and chemotherapy, can generate ROS to induce DNA damage and cell death [16–18]. In additional to these conventional approaches, ROS generation by nanocatalysts offers a promising alternative strategy for GBM therapy [19, 20]. By mimicking natural enzyme activities, nanocatalysts can catalyze ROS production in targeted ways by reacting with the high-level substrates in tumor microenvironment (TME), which can induce severe tumor-specific oxidative stress damage and overcome tumor resistance to realize the full potential of ROS-mediated therapy [21, 22]. Nevertheless, it is still requisite to combine such nanocatalytic therapy with other modalities to enhance the treatment efficacy and minimize off-target effects. In another aspect, photothermal therapy (PTT) is a promising non-invasive cancer treatment modality that employs photothermal agents (often nanoparticles) to convert absorbed light energy into heat, thereby inducing localized tumor destruction and minimizing the damage to surrounding healthy tissue [23, 24]. Notably, photo-induced hyperthermia may also improve the enzymatic activity of nanocatalysts via enhanced diffusion of reactants and improved catalytic kinetics [25, 26]. To this end, the combination of oxidative stress damage and hyperthermia-induced effects may produce synergistic antitumor outcomes, making it a promising approach for GBM therapy. This strategy offers advantages such as high targeting precision, the ability to overcome tumor resistance, and minimized side effects.

Among various photothermal agents, copper chalcogenides (Cu_2−xE, where E = S, Se, Te) with surface defects have demonstrated significant potential in photothermal therapy due to their excellent surface plasmon resonance absorption properties [27–29]. These materials are particularly effective in the second near-infrared (NIR-II) region, which allows for deep tissue penetration and localized heating upon light irradiation. Their ability to convert light into heat efficiently makes them ideal candidates for enhancing tumor treatment through hyperthermia-induced damage. Additionally, the tunable composition of copper chalcogenides nanomaterials allows for optimization of their optical properties, further improving their effectiveness in photothermal therapy [30]. On the other hand, Cu⁺/Cu²⁺, as metastable metal ions, catalyze reactions similar to Fenton chemistry across a wide pH range, promoting the generation of ROS and alleviating the hypoxia in TME [31, 32]. Compared to other metal ions, Cu⁺/Cu²⁺ ions exhibit higher reactivity in acidic conditions, which makes them particularly effective in the acidic TME [33, 34]. These properties make copper chalcogenides effective for synergistic GBM treatment strategies through potentiating the combinatorial effects of PTT and oxidative-induced damage. Selenium, as a key chalcogen, offers significant advantages in tumor therapy, including its ability to generate ROS, low toxicity, enhanced accumulation in the hypoxic TME, and activation of immunogenic cell death (ICD) [35–37]. In this context, Cu_2−xSe (CS) holds significant potential for targeted GBM therapy through the synergistic action of multiple mechanisms.

The blood-brain barrier (BBB) is a highly selective and protective membrane that presents a significant challenge in delivering therapeutic agents to the brain [38]. Due to its tight junctions, enzymatic barriers, and active efflux mechanisms, nanoagents often struggle to cross the BBB. Overcoming this barrier remains one of the major obstacles in the development of effective therapies for GBM [39–41]. To address this challenge, membrane-camouflaged nanoparticles have emerged as a promising strategy to enhance the delivery of therapeutic agents across the BBB. These nanoparticles can exploit the natural cell membranes’ properties to improve BBB penetration and enable more efficient targeted treatment of GBM [42, 43]. Cancer membrane-coated nanoparticles carry certain surface markers or ligands that interact with specific receptors on BBB endothelial cells, potentially facilitating nanoparticle transport across the barrier. Herein, glioma cell membrane-camouflaged CS nanoparticles (CS@CM) were developed for efficient GBM therapy via simple intravenous injection (Scheme 1). During blood circulation, this camouflage effectively facilitated CS@CM to penetrate BBB and reach glioma tissue. After tumorous enrichment, intense hyperthermia and amplified ROS generation can be simultaneously achieved under NIR-II laser irradiation, without causing significant adverse effects. Such combinatorial PTT and oxidative stress damage not only extensively ablated the tumor tissue, but also triggered dramatic ICD. The resultant release of damage-associated molecular patterns (DAMPs) may enhance the immune response and achieve more durable GBM treatments [44, 45]. This multimodal therapy represents a significant advantage in the treatment of GBM, addressing critical limitations in existing therapies and providing a promising strategy for overcoming the resistance and recurrence associated with this aggressive cancer. The structural features and physiochemical properties of CS@CM were investigated through multiple characteristic techniques. Cell uptake behavior and cytotoxicity of CS@CM on glioma were studied at the cellular level. Lastly, the performance of CS@CM on GBM treatment was validated by using tumor-bearing mouse models. The results aim to bring new opportunities and provide valuable insights for future GBM treatment.

Scheme 1.

Diagrammatic representation of the synthetic procedure of CS@CM and its therapeutic mechanism toward orthotopic GBM

Results and discussion

Structural characterizations

CS nanoparticles were synthesized using a straightforward chemical reduction method under a nitrogen atmosphere. The selenium precursor was prepared by reducing selenium powder with sodium borohydride, followed by coordination with Cu²⁺, resulting in the formation of ultra-small CS, which exhibited a characteristic brown to black color [46]. The CS has a regular spherical morphology and good aqueous dispersion, exhibiting an extremely tiny particle size, from observing the transmission electron microscopy (Fig. 1a and S1). The hydrated particle size of the CS is 5.5 nm, determined through dynamic light scattering (DLS). High resolution TEM image reveals a lattice spacing of 0.28 nm, which corresponds to the (220) crystal plane of CS (Fig. 1b). To synthesize cytomembrane-camouflaged CS, GL261 Cell membrane was first obtained through ultrasonic fragmentation and purification. Afterwards, biomimetic CS@CM was prepared by a physical extrusion method. TEM image shows a typical membrane-coated nanoparticle of CS@CM, presenting the similar crystalline form with CS (Fig. 1c and S2). Notably, a significant number of small CS particles are present within each CS@CM, resulting in An average size of approximately 133 nm, which falls within the effective range for enhanced permeability and retention (EPR) taking action. The hydrodynamic diameter of CS@CM, measured by DLS, was found to be 182.0 nm, which is comparable to that of membrane vesicles, measured at 147.7 nm (Fig. 1d and e). Zeta potentials of CS and CS@CM were recorded to be −13.72 and − 11.16 mV, respectively (Fig. 1f). The moderate surface charge change is attributed to the cell membrane coating.

Fig. 1.

Structural characterizations. (a) TEM image and (b) HR-TEM image of CS. (c) TEM image of CS@CM. (d) AFM image of CS. (e) Hydrodynamic dimensions of CS and CS@CM as determined by DLS. (f) Zeta potentials of CM, CS, and CS@CM. (g) XRD pattern of CS. (h) Core-level XPS spectra of Cu 2p. (i) FTIR spectra of PVP, CS, and CS@CM. (j) UV-visible spectra of CS at different concentrations (0, 1.25, 2.5, And 5 mg mL⁻¹). (k) SDS-PAGE analysis of CM, CS, and CS@CM

The crystal structure of as-synthesized CS was analyzed using powder X-ray diffraction (XRD). The XRD pattern of CS shows weak diffraction peaks, which are attributed to the standard berzelianite phase of selenium (JCPDS card No. 06-0680) (Fig. 1g). Surface defects may increase the amorphous content, leading to a reduction in the intensity of the diffraction peaks. Structural defects can introduce localized states and enhance light absorption, thereby preserving or even improving the photothermal conversion efficiency [30]. In addition, the presence of surface defects may facilitate charge separation and electron transfer, which are beneficial for catalytic processes, including ROS generation via redox cycling between Cu(I) and Cu(II) [31]. X-ray photoelectron spectroscopy (XPS) Analysis of the Cu 2p region confirmed the presence of both Cu⁺ and Cu²⁺ species, further validating the copper-deficient nature of the CS nanoparticles (Fig. 1h). Fourier transform infrared (FTIR) spectroscopy was used to examine the functional groups on the surface of the nanoparticles. As shown in Fig. 1i, the characteristic peaks of PVP at 2828–2982 cm⁻¹ (C-H stretching), 1705–1725 cm⁻¹ (C = O stretching), And 1020–1360 cm⁻¹ (C-N stretching) were significantly attenuated after the cell membrane was coated onto the CS. The optical property of CS was characterized using UV-vis-NIR spectroscopy. Strong localized surface plasmon resonance (LSPR) of CS was observed in the NIR region (600–1000 nm), which is associated to the high vacancy density of copper in the structure (Fig. 1j). This LSPR feature is of particular interest for potential biomedical applications in photoacoustic imaging and photothermal therapy in the NIR region. SDS-PAGE was performed on the CM, CS, and CS@CM samples to verify the successful loading of the cell membrane. Coomassie Brilliant Blue staining displayed clear bands for both CM and CS@CM, confirming the effective incorporation of the cell membrane onto the CS surface (Fig. 1k and S3). EDS mapping analysis confirmed the presence of Cu, Se, O, S, N, and P in CS@CM, further validating the successful formulation of the final product (Figure S4).

Physiochemical properties

The photothermal properties of CS and CS@CM were comprehensively evaluated under 1064 nm laser irradiation (1 W/cm²). After 10 min of NIR irradiation, CS reached a temperature of 53 °C, indicating a strong photothermal effect for potential tumor cell ablation (Fig. 2a). In comparison, CS@CM reached a peak temperature of 48.7°C, likely due to the cell membrane coating, which partially inhibits the photothermal conversion. When exposed to four consecutive ‘on’ and ‘off’ laser cycles, CS@CM exhibited stable temperature increases, highlighting their capacity to sustain repetitive thermogenesis (Fig. 2b). The photothermal conversion efficiency of CS@CM was calculated to be 37.68%, which is exceptionally high for PTT in the NIR-II biological window (Figure S5). It is important to note that the hyperthermia generated by CS@CM is closely linked to both the agent concentration and laser irradiation period, making it conducive to controllable photothermal therapy (PTT) (Fig. 2c).

Fig. 2.

Photothermal property and ROS generation capability of CS@CM. (a) Heating curve of the aqueous solution containing CS or CS@CM (2 mg/mL) during irradiation with a 1064 nm laser for 30 min. (b) Temperature change of CS@CM dispersion (2 mg/mL) under four cycles of cyclic NIR irradiation (1064 nm, laser on for 7 min per cycle). (c) Thermal mapping of different concentrations of CS@CM under 1064 nm laser irradiation for different periods. (d) Absorption spectra of OPD (2 mg/mL) after reacting with various agents (100 µg/mL) in the presence of H₂O₂ (100 µM) for 30 min. (e) Absorption spectra of OPD (2 mg/mL) after reacting with CS@CM at different concentrations (0–100 µg/mL) in the presence of H₂O₂ for 30 min. (f) Absorption spectra of MB (0.5 mg/mL) after reacting with various agents (100 µg/mL) in the presence of H₂O₂ (100 µM) for 30 min. Absorption spectra of (g) MB (0.5 mg/mL), (h) TMB (0.5 mM) and (i) ABTS (0.5 mM) after reacting with CS@CM at different concentrations (0–100 µg/mL) in the presence of H₂O₂ (100 µM) for 30 min. (j) Fluorescence emission spectra of TA (6 mM) after reacting with CS@CM (100 µg/mL) in the presence of H₂O₂ (100 µM) for 30 min. (k) ESR spectrum of CS@CM (100 µg/mL) in the H₂O₂ solution (100 µM) containing the spin trap of DMPO (100 mM). NIR-II laser (1064 nm) irradiation was carried out for 5 min in the applicable groups. (l) Dissolved oxygen level in H₂O₂ solution (100 µM) containing CS@CM (100 µg/mL) during incubation for 9 min

Multiple colorimetric assays were employed to investigate the ROS generation mediated by CS@CM. Fluorescence assays to detect 2,3-diaminophenazine (OPDox) were conducted, with ROS production monitored at 566 nm (Fig. 2d and e). ‘CS + L’ group produces the highest amount of ROS, followed by the ‘CS@CM + L’ group. The reduced ROS generation in the ‘CS@CM + L’ group suggests that the cell membrane coating partially shields ROS production compared to the uncoated CS. Similar results were obtained by using the colorimetric probe of MB, further confirming that CS@CM can effectively generate ROS for potentially aggravating cellular oxidative stress (Fig. 2f). In addition, based on the assays of OPD, MB, TMB and ABTS, ROS generation by CS@CM is strongly dependent on the concentration of the agent (Fig. 2g-i). TA was employed to specifically probe for hydroxyl radicals (·OH) in order to characterize the types of ROS generated. The CS@CM group exhibited intense fluorescence at 425 nm, indicating the yield of highly reactive ·OH (Fig. 2j). Electron spin resonance (ESR) spectroscopy double confirmed the formation of ·OH, as evidenced by the typical characteristic quartet peaks with the ratio of 1:2:2:1 using DMPO as the spin trap (Fig. 2k). These findings highlight the significant ROS generation capacity of CS@CM, particularly the production of ·OH, which may contribute to their therapeutic potential in PTT and ROS-induced oxidative damage. In another aspect, the catalase-like activity of CS@CM was accidentally discovered by monitoring the dissolved oxygen levels in H₂O₂ solution (Fig. 2l). In the presence of CS@CM, a rapid increase in O₂ concentration was observed within the first 4 min. Such localized oxygenation holds promise for alleviating hypoxia in the TME, potentially leading to tumor regression.

Cellular uptake and cytotoxicity in vitro

Cell membrane camouflage is an effective strategy for enhancing the in vivo stability and tumor-targeting efficiency of nanoparticles by preventing nonspecific clearance through the mononuclear phagocyte system (MPS). Additionally, such camouflage promotes homologous targeting of nanoparticles by retaining tumor-specific antigens on the cytomembrane surface. Therefore, the homologous targeting and cellular uptake of CS@CM were assessed using confocal laser scanning microscopy. Both CS and CS@CM were effectively internalized by GL261 glioma cells, as evidenced by the red fluorescence of Cy5.5 in cytosolic region (Fig. 3a and b). It is noteworthy that CS@CM exhibited significantly higher fluorescence intensity compared to the CS group at specific time points (0, 0.5, 1, 2, 4, And 6 h) (Fig. 3c and d). CS@CM are coated with glioma cell membranes, which are rich in tumor-specific surface proteins and receptors. These membrane proteins enable CS@CM to specifically recognize and bind to glioma cells, as the cell membrane has an affinity for homologous tumor cell surface markers. This homotypic targeting allows the nanoparticles to be taken up by glioma cells more efficiently, as the cancer cell membranes on the nanoparticles interact with the same molecules on the surface of glioma cells, thereby enhancing cellular uptake [47, 48]. To further verify the specific targeting capability of CS@CM, cellular uptake assay using non-glioma 4T1 murine cells was employed as a heterologous control. As shown in Figure S6, there was no obvious difference in the Cy5.5 fluorescence signal between CS and CS@CM in 4T1 cells, suggesting that the GL261 cell membrane coating does not enhance nanoparticle uptake in unrelated cell types. These findings confirm that the cell membrane camouflage facilitates selective uptake by glioma cells while minimizing nonspecific internalization by other cell types, thereby improving the targeting precision of CS@CM.

Fig. 3.

Cellular uptake and cytotoxicity of CS@CM in vitro. Confocal microscopic images of GL261 glioma cells after incubation with Cy5.5-labeled (a) CS and (b) CS@CM for different periods (scale bar: 200 μm). Cell nucleus and cytoskeleton are fluorescently labeled with DAPI and phalloidin-FITC, respectively. Quantitative analysis of the MFI of Cy5.5-labeled (c) CS and (d) CS@CM. (e) Confocal microscopic images of GL261 cells after various treatments and stained with DCFH-DA fluorescent probe (scale bar: 200 μm). (f) MFI of DCF in different groups corresponding to panel (e). Cell viability of GL261 cell after treatment with CS@CM at the concentration of (g) 25 µg/mL, (h) 50 µg/mL, (i) 100 µg/mL, and (j) 200 µg/mL, with or without 1064 nm laser irradiation. (k) Flow cytometry dot plots of GL261 cells subject various treatments and stained with Annexin V-FITC/PI. (l) Histogram to quantify cell number in the stages of non-apoptosis, early apoptosis, late apoptosis and necrosis stages corresponding to panel (k). Groups are assigned to be (1) Blank, (2) Laser, (3) CS, (4) CS + laser, (5) CS@CM, (6) CS@CM + laser. Data are displayed as mean ± SD (n = 4). ^*p < 0.05, ^**p < 0.01, ^***p < 0.001

DCFH-DA is a non-fluorescent compound that can freely penetrate cell membranes. Once inside the cell, it is hydrolyzed by intracellular esterases to form DCF. Thus, ROS level in GL261 cells were fluorescently analyzed by using DCFH-DA probe (Fig. 3e). The administration of CS and CS@CM can effectively elevate the intracellular ROS level, attributed to the Cu(I)/Cu(II)-mediated Fenton-like reaction (Fig. 3f). Furthermore, the amount of ROS increased significantly upon NIR laser irradiation, which is associated with the enhanced activity of nanocatalysts under hyperthermic conditions. Compared to CS, CS@CM resulted in a tremendously higher content of cytosolic ROS, resulting from the improved cellular uptake of the biomimetic nanoparticles. ROS can trigger oxidative stress, leading to cellular damage through multiple pathways, including DNA strand breaks and base modifications, lipid peroxidation that compromises membrane integrity, and protein oxidation that impairs essential enzymatic and structural functions. These cumulative damages activate apoptosis or necrosis, ultimately resulting in tumor cell death [49, 50].

The biocompatibility of CS@CM was evaluated by assessing its toxicity on human umbilical vein endothelial cells (HUVECs). Cell viability remained above 90% even after incubation with CS@CM at concentrations up to 200 µg/mL, demonstrating excellent biocompatibility (Figure S7). Tumor-specific cytotoxicity of CS@CM was investigated through both 3-(4,5-dimethylthiazol-2-yl)−2,5-diphenyltetrazolium bromide assay (MTT) assay and Annexin V apoptosis assay. Both CS and CS@CM exhibited dose-dependent cytotoxicity toward GL261 glioma cells under 1064 nm laser irradiation (1 W/cm² for 5 min), thanks to the PTT effect and ROS-induced oxidative damage. CS@CM demonstrated significantly stronger phototoxicity compared to CS, likely due to the enhanced cellular uptake and tumor-targeting capabilities provided by the cell membrane coating (Fig. 3g-j). On this basis, the apoptotic status of glioma cells was further uncovered after various treatments. Flow cytometry results show that the highest percentage of apoptosis/necrosis, 54.38%, was achieved with the ‘CS@CM + laser’ treatment, attributed to the most severe photoinduced damage caused by hyperthermia and oxidative stress (Fig. 3k and l). To further evaluate the pro-apoptotic effect of CS@CM, a TUNEL assay was conducted on GL261 cells after different treatments. As shown in Figure S8, the ‘CS@CM + laser’ group displayed the strongest TUNEL fluorescence, indicating the highest apoptotic level among all groups. This result is consistent with our in vitro cytotoxicity assays and testifies that the combination of CS@CM with NIR irradiation can effectively trigger apoptosis in glioma cells, thereby contributing to its potent antitumor efficacy.

To assess the ability of CS@CM to penetrate the BBB, a Transwell assay was conducted using a bEnd.3 cell layer in the upper chamber to mimic the BBB. The results clearly demonstrate that CS@CM exhibits efficient BBB penetration, as evidenced by the significantly stronger Cy5.5 fluorescence observed in GL261 cells in the lower chamber compared to the control (Figure S9). This enhanced BBB penetration can be attributed to the glioma-specific receptors present on the BBB, which facilitate receptor-mediated transcytosis and endocytosis of CS@CM [51]. Additionally, the surface characteristics of the nanoparticles, such as size and charge, further enhance their permeability, while the glioma cell membrane camouflage helps to reduce immune clearance, allowing for prolonged circulation and efficient tumor targeting [52]. This multi-faceted mechanism effectively facilitates CS@CM’s ability to cross the BBB and deliver the therapeutic agents directly to glioma cells, enhancing its potential for GBM treatment.

DAMPs release in vitro

ICD is a specific form of cell death that not only eliminates cells but also activates the immune system, particularly the antitumor immune response. DAMPs play a central role in ICD by acting as signals that alert the immune system to the presence of dying or stressed cells. These molecular signals are released or exposed during ICD and help to activate the immune response, leading to the recognition and elimination of tumor cells. To assess the release of DAMPs from tumor cells during CS@CM-mediated therapy, an optimized immunostaining protocol was applied after treating the glioma cells with various treatments. Calreticulin (CRT) plays a pivotal role in ICD by acting as a key “eat-me” signal for the immune system. It is a calcium-binding chaperone that is typically located in the endoplasmic reticulum (ER) of cells, assisting in protein folding and quality control. However, during ICD, CRT undergoes a translocation from the inside of the cell to the cytomembrane, which marks the cell for recognition and phagocytosis by dendritic cells (DCs) and other immune cells. Confocal microscopy revealed the highest translocation of CRT induced by ‘CS@CM + laser’, as evidenced by the most intense red fluorescence on the membrane of glioma cells (Fig. 4a-c). These results demonstrate that CS@CM effectively triggers ecto-CRT translocation through pronounced ER stress. As Another critical DAMPs, high mobility group box 1 (HMGB1) is a nuclear protein that typically functions in DNA stabilization and the regulation of gene expression, but it can also be released extracellularly during cellular stress or death. When released, HMGB1 acts as a potent alarmin, signaling to the immune system for initiating and modulating immune responses. In contrast to its normal localization in the nuclear region of untreated cells, ‘CS@CM + laser’ treatment induces a remarkable leakage of HMGB1 into the cytoplasm and extracellular space. This is demonstrated by the significant reduction of green fluorescence in the cytosolic region, indicating the leakage of HMGB1 (Fig. 4d and e). Adenosine triphosphate (ATP) plays a multifaceted role in ICD, acting both as an energy currency of the cell and as an important danger signal in the context of immune activation. ATP is released from dying tumor cells and acts on immune cells through purinergic receptors, especially P2 × 7, to activate the innate immune system, recruit immune cells, and promote dendritic cell maturation. As observed, ‘CS@CM + laser’ induces significant ATP release from the cytosolic region, likely due to irreversible damage to the plasma membrane (Fig. 4f). The remaining intracellular ATP closely mirrored the trend of released ATP (Fig. 4g). All of the above results demonstrate that CS@CM administration effectively induced the release of DAMPs, and the confirmed ICD is crucial for stimulating antitumor immunity in interventional immunotherapy (Fig. 4h).

Fig. 4.

DAMPs release from GL261 cells in vitro. (a) Confocal microscopy images of GL261 cells after various treatments and stained with CRT (scale bar: 50 μm). (b) MFI of CRT corresponding to panel (a). (c) Flow cytometric dot plots to analysis the fluorescence intensity of CRT. (d) Confocal microscopy images of GL261 cells after diverse administrations and stained with HMGB1 (scale bar: 50 μm). (e) MFI of HMGB1 corresponding to panel (d). (f) The amount of ATP released from GL261 cells subject to different treatments. (g) The remnant ATP in cytoplastic region in terms of various regiments. (h) Schematic diagram to illustrate the DAMPs release for ICD stimulation. Groups are assigned to be (1) Blank, (2) Laser, (3) CS, (4) CS + laser, (5) CS@CM, (6) CS@CM + laser. Data are displayed as mean ± SD (n = 4). ^*p < 0.05, ^**p < 0.01, ^***p < 0.001

Biodistribution of CS@CM in vivo

To further evaluate the lesion targeting capability and accumulation of CS@CM in glioma, orthotopic GBM models were established on C57BL/6 mice. Fluorescence imaging was employed to monitor the biodistribution of Cy5.5-labeled CS and CS@CM upon their intravenous injection via the tail vein (Fig. 5a-d). Mice treated with CS@CM exhibited significantly higher fluorescence signals in the brain compared to those treated with CS, implying that the GL261 cell membrane coating enhances the targeting ability of the CS@CM (Fig. 5a). The fluorescence signal gradually increased over time, reaching its peak at 12 h post-injection (Fig. 5b). It is noteworthy that CS@CM selectively accumulates in GBM tissue while showing minimal uptake in the heart, demonstrating excellent targeting efficacy (Fig. 5c and d). The glioma-specific enrichment of CS@CM was further confirmed by immunofluorescence staining of the brain histological sections (Fig. 5e). The glioma cell membrane proteins on CS@CM may interact with endothelial cells of the BBB, facilitating the transport of nanoparticles across the BBB into the glioma regions. Moreover, this homologous targeting allows the CS@CM to recognize and preferentially bind to glioma cells over normal brain cells, enhancing their uptake by the tumor and reducing off-target effects. Although the circulation half-life was not directly measured in this study, membrane camouflage has been demonstrated to reduce immune clearance, which likely prolongs the circulation time of CS@CM, allowing for more efficient tumor targeting. These findings suggest that the glioma membrane coating improves both targeting specificity and circulation half-life, thereby enhancing therapeutic efficacy [53, 54].

Fig. 5.

Biodistribution of Cy5.5-labeled CS@CM in brain region after intravenous injection. (a) In vivo fluorescence imaging of the brain cerebral site in GL261 glioma-bearing C57BL/6 mice at different time points (4, 6, 8, 12, 24, And 36 h) following intravenous injection of Cy5.5-labeled CS or CS@CM (5 mg/kg). (b) MFI in brain region corresponding to panel (a). (c) Ex vivo fluorescence images of the Heart And brain tissue at 12 h post-injection of different agents in GBM-bearing mice. (d) MFI in different tissues corresponding to panel (c). (e) Representative fluorescence images of brain sections at 12 h post-injection of CS or CS@CM (scale bar: 250 μm). (f) H&E staining of the histological sections of major organs on day 28 in terms of different treatment groups (scale bar: 200 μm). Data are displayed as mean ± SD (n = 4). ^**p < 0.01, ^***p < 0.001

GBM suppression by CS@CM in vivo

Inspired by the impressive tumor-killing performance of CS@CM in vitro and its favorable biodistribution in vivo, its therapeutic efficacy against solid tumors was evaluated in GBM-bearing C57BL/6 mice. The safety of the treatment and its tumor-suppressing effects were both evaluated following a detailed treatment protocol (Figs. 5f and 6a). All the mice were assigned into the following five groups at random, including (1) Blank, (2) CS, (3) CS + laser, (4) CS@CM, (5) CS@CM + laser. Tumor-bearing mice were intravenously injected with various agents on day 0, 2 And 4. One day post-injection, the tumor sites were irradiated with a pulsed 1064 nm laser (1 W/cm², 5 min). On day 1, infrared thermal imaging revealed a rapid temperature increase to 43.4 °C at the tumor region after 5 min of 1064 nm laser irradiation upon CS@CM administration, whereas in the CS group, the temperature increased to 41.6 °C (Fig. 6b and c). The outstanding thermal response of CS@CM aligns with its enhanced tumor-homing properties, which may enable more effective PTT and increase oxidative stress-induced damage with high specificity.

Fig. 6.

GBM inhibition in vivo. (a) Timing program of the treatment procedures for animal experiments. (b) NIR thermographic images in orthotopic GBM-bearing C57BL/6 mice under 1064 nm laser irradiation (1 W/cm²) one day after intravenous injection with CS or CS@CM (5 mg/kg). (c) Temperature elevation curves at tumor site corresponding to panel (b). (d) Bioluminescence images of GBM-bearing C57BL/6 mice on day 0, 14 And 28. (e) MFI of the tumor site corresponding to panel (d). (f) Tumor inhibition rate in terms of different administrations. (g) Kaplan-Meier survival curves of mouse models after receiving diverse regimens. (h) H&E staining of the cerebral tissue sections. The area enclosed by the dashed line represents the GBM region. Groups are assigned to be (1) Blank, (2) CS, (3) CS + laser, (4) CS@CM, (5) CS@CM + laser. Data are displayed as mean ± SD (n = 4). ^***p < 0.001

The therapeutic efficacy of CS@CM was further assessed using bioluminescence imaging (Fig. 6d). Compared to the blank control, glioma growth in the ‘CS@CM + laser’ group was significantly inhibited by the end of treatment, as evidenced by the minimal Luc + GL261 fluorescence in cerebral region on day 28 (Fig. 6e). The tumor inhibition rates were calculated to be 9.07%, 47.68%, 16.51% And 88.64%, in terms of the treatment groups of CS, CS + laser, CS@CM, and CS@CM + laser, respectively (Fig. 6f). Kaplan-Meier survival curves showed a clear survival benefit in the treated groups (Fig. 6g). Non-treated mice had a median survival of 26.5 days, while those treated with ‘CS@CM + laser’ exhibited the longest median survival of 49 days. These results demonstrate the potential of CS@CM in inhibiting GBM growth or progression and extending survival. The changes in glioma size after various treatments were further analyzed on histological sections. After hematoxylin & eosin (H&E) staining, the minimum area of tumor region was visualized in the group of ‘CS@CM + laser’, which was in line with the findings in the fluorescence imaging in vivo (Fig. 6h). To evaluate the immune response triggered by CS@CM treatment, flow cytometry analysis was performed to assess the activation of T cells and DCs at the end of the animal study. The ‘CS@CM + laser’ group exhibited the highest percentage of CD4⁺/CD8⁺ T cells and CD80⁺/CD86⁺ DCs, indicating robust activation of both T cells and DCs (Figure S10). This result suggests that CS@CM treatment effectively stimulates a strong immune response, involving both T cell activation and DC maturation, which enhances immune surveillance and promotes efficient tumor elimination [55, 56].

The therapeutic safety of nanoparticles is crucial for tumor therapy due to their potential to cause both beneficial and harmful effects. To this end, pathological analysis of major organs (heart, liver, spleen, lung, and kidney) was conducted using H&E staining on day 28 (Fig. 5f). No notable histopathological alterations were detected in any of the treatment groups, suggesting that the systemic damage to healthy tissues was minimal. Biochemical assays were further carried out to assess the functional status of the liver and kidneys by measuring specific biomarkers in blood. Major liver and kidney function indicators showed no significant changes following treatment with 'CS@CM + laser', indicating that the administration of CS@CM minimally contributed to the development of chronic disorders in these organs (Figure S11). These findings collectively demonstrate the excellent biosafety profile of CS@CM, supporting its potential for future clinical applications in oncology therapy toward GBM.

Conclusions

In summary, we have successfully engineered novel biomimetic CS@CM nanoparticles for oxidative stress-induce damage of glioma cells, enhanced by photo-induced hyperthermia. Upon intravenous administration, glioma cell membrane on CS@CM can facilitate the transport of nanoparticles across the BBB into the GBM regions And this homologous targeting allows agents to preferentially recognize And bind to glioma cells, thereby enhancing tumor-specific uptake while minimizing off-target effects. Under 1064 nm laser irradiation, thermal generation in the GBM region not only induces tumor ablation but also promotes catalytic reactions, leading to enhanced oxidative damage. ICD induction by the synergistic therapy may effectively stimulate antitumor immunity, thereby enhancing therapeutic efficacy. The drastic suppression of GBM was achieved in orthotopic tumor-bearing mice without inducing significant adverse side effects. Taken together, this approach offers new perspectives on the design of biomimetic nanocatalysts for effective GBM therapy, with a focus on overcoming the BBB and optimizing the therapeutic potential of the TME.

Experimental section

Synthesis of CS: A one-step method was employed in the aqueous phase. 50 mL of ultrapure water was added to a 100 mL three-neck round-bottom flask, followed by nitrogen purging for 10 min to remove oxygen. Selenium powder (39.48 mg, 0.5 mmol) and NaBH₄ (56.7 mg, 1.5 mmol) were then introduced, And the mixture was stirred for 2 h under a nitrogen atmosphere until the solution became colorless. The resulting solution was labeled Solution (A) Then, 25 mL of ultrapure water was added in a separate 100 mL three-neck flask, followed by CuCl₂·2H₂O (85 mg, 0.5 mmol) and PVP (200 mg). The mixture was stirred for 20 min under nitrogen to ensure complete coordination of PVP with copper ions, yielding Solution (B) Solution A (25 mL) was then added to Solution B using a syringe, resulting in the solution to turn brown. The mixture was stirred for An additional 2 h at room temperature. The resulting solution was centrifuged at 10,000 rpm for 10 min, And the supernatant was collected And dialyzed using a 14 kDa dialysis bag to remove excess PVP. The purified products were placed in 50 mL centrifuge tubes, freeze-dried for 24 h, And stored at 4 °C for subsequent experiments.

Synthesis of CS@CM: Mouse Luc + GL261 cells were cultured at 37 °C in a 5% CO₂ incubator. Adherent cells were detached And resuspended to prepare a single-cell suspension. A total of 5× 10⁷ cells were cryopreserved in Liquid nitrogen And subjected to three rapid freeze-thaw cycles. Cell membrane disruption was performed using An ultrasonic disruptor for 10 min while maintaining the suspension in an ice-water bath to prevent excessive Heating. Following membrane disruption, the mixture was centrifuged at 600 rpm for 10 min to remove large organelles. The supernatant was collected And further centrifuged at 2,700 rpm for 10 min, followed by ultra-high-speed centrifugation at 11,480 rpm for 30 min at 4 °C to isolate the final supernatant. The supernatant was then lyophilized to obtain the cell membrane. The cell membrane was extruded through a 400 nm polycarbonate membrane using a cell membrane extruder for 7 cycles to produce small vesicles. These vesicles were then mixed with CS And extruded through a 100 nm polycarbonate membrane for An additional 7 cycles. The final products, composed of the CS encapsulated within the Luc + GL261 cell membrane vesicles, was obtained and designated as CS@CM.

Photothermal property of CS@CM: To investigate the photothermal effects of CS and CS@CM, thermographic images were taken and temperature changes were recorded at various sample concentrations (0, 0.05, 0.1, 0.5, 1, And 2 mg/mL) under 1064 nm laser irradiation (1 W/cm²) for 5 min. Each sample was measured three times independently, with photos And temperature readings taken every 30 s. The photothermal cycling performance of CS@CM (2 mg/mL) was also evaluated. In this experiment, the material was first irradiated for 5 min under 1064 nm laser irradiation (1 W/cm²) for 7 min (‘on’ state), then allowed to cool naturally down to room temperature (‘off’ state), constituting one cycle. Totally five ‘on-off’ cycles were performed, with temperature recorded every 30 s.

ROS generation by CS@CM: ROS generation by CS@CM under different endogenous and exogenous stimuli was investigated using colorimetric assays. UV absorption was initially measured under different treatments, including CS, CS@CM, and CS@CM + L. The ROS yield can be quantified with multiple colorimetric probes, including O-phenylenediamine (OPD), methylene blue (MB), 3,3’,5,5’-tetramethylbenzidine (TMB) And 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS). Specifically, CS@CM dispersions at concentrations of 0, 12.5, 25, 50, And 100 µg/mL were incubated with OPD (2 mg/mL), MB (500 µg/mL), TMB (0.5 mM) and ABTS (0.5 mM) for 30 min. After incubation, the supernatants were collected for spectrophotometric Analysis, with absorbance measured at 450 nm (OPD), 655 nm (MB), 652 nm (TMB), And 734 nm (ABTS). Alternatively, ROS generation was also detected by using terephthalic acid (TA), with fluorescence emission at 425 nm. The yielded free radical was further categorized via electron paramagnetic resonance spectroscopy (ESR) using the spin trap of 5,5-dimethyl-1-pyrroline N-oxide (DMPO). Briefly, 100 µg/mL CS@CM solution was mixed with DMPO (100 mM) in the presence of H₂O₂ (100 µM). ESR signal was finally recorded after reaction for 30 min.

Cellular uptake behavior of CS@CM: GL261-Luc cells were seeded in confocal Petri dishes (4 × 10⁴ cells/dish) And cultured for 24 h. The cells were then incubated with Cy5.5-labeled CS and CS@CM for various time points (0, 0.5, 1, 2, 4, And 6 h). After incubation, the cells were washed three times with PBS, fixed with 4% paraformaldehyde, and stained with DAPI for subsequent examination. Cellular uptake was assessed using a confocal laser scanning microscope (LSM710, Carl Zeiss, Germany) and concurrently analyzed by a flow cytometer (BD FACSVantage SE, USA).

Intracellular ROS generation: 2,7-Dichlorofluorescin diacetate (DCFH-DA) fluorescence probe was used to detect total intracellular ROS production. Specifically, GL261 cells were seeded on glass-bottom dishes at a density of 8 × 10⁴ cells per well And cultured for 24 h. After incubation with CS or CS@CM (100 µg/mL) for 6 h, the GL261 cells were washed twice with PBS And further incubated for 12 h. DCFH-DA was then added to the cells And incubated at 37 °C under 5% CO₂ for 30 min. The cells were either irradiated or left unirradiated with a pulsed 1064 nm laser (1 W/cm²) for 5 min, followed by incubation for Another 2 h under the same culture conditions (37 °C, 5% CO₂). Afterward, the culture medium was removed, and the GL261 cells were stained with Hoechst 33,342 for 15 min. Intracellular ROS production was then characterized and analyzed by using both confocal microscopy and flow cytometry.

Transwell assay: bEnd.3 cells (6 × 10⁴ cells per well) were seeded onto the upper chamber of a 12-well Transwell plate And cultured for 7 days. The transendothelial electrical resistance (TEER) was measured periodically to monitor the integrity of the endothelial monolayer, And the culture was continued until the TEER value exceeded 150 Ω·cm². GL261 cells (1 × 10⁵ cells per well) were then seeded into the lower chamber And cultured for 24 h. Subsequently, Cy5.5-labeled CS and CS@CM were added to the upper chamber And incubated for 8 h. After incubation, the chambers were washed twice, fixed using DAPI staining, and fluorescence was observed under confocal microscopy to assess the ability of the nanoparticles to cross the BBB.

Apoptosis induced by CS@CM: The apoptosis of cells upon various treatments was assessed using flow cytometry. GL261 cells were seeded in 6-well plates (1 × 10⁵ cells per well) And incubated for 24 h. Subsequently, CS and CS@CM were added to the wells And co-cultured with the cells for An additional 6 h. The cells were then washed, cultured in drug-free medium, And irradiated with or without a 1064 nm laser (1 W/cm², 5 min). After irradiation, the cells were incubated for 24 h. Following this incubation, cells were harvested using trypsin, washed with PBS, and stained with Annexin V-FITC and propidium iodide (PI) according to the manufacturer’s protocol. At last, apoptosis was analyzed by flow cytometry.

DAMPs release by CS@CM: Immunofluorescence imaging was used to assess the exposure of CRT on the plasma membrane surface. GL261 cells (1 × 10⁵ cells per well) were seeded in 6-well plates And cultured for 24 h. Subsequently, CS and CS@CM (100 µg/mL) were added to the wells. After 6 h of incubation, the cells were maintained in fresh medium And exposed to or without the 1064 nm laser (1 W/cm², 5 min). After An additional 24 h of incubation, the cells were harvested after trypsinization, washed with PBS, And stained with PE-conjugated Anti-mouse CRT antibody, and incubated at 4°C for 20 min. Finally, the cells were collected and analyzed by both confocal microscopy and flow cytometry. To investigate the release of HMGB1, similarly, GL261 cells were plated into confocal Petri dishes (5 × 10⁴ cells/dish) And cultured for 12 h. Once the cells adhered to the dish surface, CS and CS@CM were added into the wells. After 6 h of phagocytosis, unadhered nanoparticles were aspirated And discarded. The cells were then irradiated with the 1064 nm laser (1 W/cm², 5 min). After 24 h of additional incubation, the cells were fixed and permeabilized to stain with Alexa Fluor^® 647-conjugated anti-HMGB1 antibody, followed by examination with confocal microscopy. The extracellular release of ATP was measured using an ATP detection kit. Briefly, GL261 cells were cultured as described above And exposed to or without the 1064 nm laser (1 W/cm², 5 min). The released ATP in the supernatant was then detected following the manufacturer’s protocol.

Orthotopic GBM establishment in C57BL/6 mice: All experimental protocols in this study were conducted in accordance with the guidelines approved by the Ethics Committee of Chongqing Medical University and the Animal Ethics Committee of the Second Affiliated Hospital of Chongqing Medical University (Approval number: IACUC-CQMU-2023-0096). Male C57BL/6 mice, aged 6–8 weeks, were purchased from Chongqing Enswell Biotechnology Co., Ltd. (China) And bred in SPF-class facilities in the laboratory. For orthotopic GBM implantation, the mice were Anesthetized with 1% pentobarbital sodium and fixed in a stereotactic apparatus. A suspension containing GL261 cells (3 × 10⁵ cells in 5 µL of PBS) was then injected into the striatum of the mice (coordinates: bregma 1.0 mm, right lateral 2.0 mm, depth 3.5 mm) for tumor inoculation. In situ tumor growth was monitored using an IVIS Imaging System. After seven days, the GBM-bearing mice were ready for the following experiments.

Biodistribution of CS@CM: in vivo After the establishment of the orthotopic GBM model, 200 µL Cy5.5-labeled CS or CS@CM (5 mg/kg) were intravenously injected into GBM-bearing C57BL/6 mice. Biodistribution of the nanoagents in vivo was monitored at various time points (4, 6, 8, 12, 24, And 36 h) by fluorescence imaging. At 12 h post-injection, major organs (liver, spleen, lung, kidney, heart, and brain) were harvested, rinsed with PBS, And imaged using the IVIS Imaging System. To further assess nanoagent distribution within the tumor tissue, the brains of the mice were fixed, embedded in OCT compound, And sectioned at 10 μm thickness. The cell nuclei were stained with DAPI, and the distribution of the nanoagents in the tissue was visualized from confocal microscopy. In another aspect, GBM tissues from mice treated with CS or CS@CM were collected at 12 h post-injection, followed by fixed in 2.5% glutaraldehyde for further ultrastructural analysis.

GBM-specific hyperthermia generation in vivo: The laser beam spot was aligned with the tumor implantation site, and the laser-induced temperature increase at the tumor site was recorded using an infrared thermal imaging camera. To evaluate the laser-induced hyperthermia generation in orthotopic GBM in vivo, tumor-bearing mice were injected via the tail vein with CS or CS@CM at the injection dose of 5 mg/kg. One day post-injection, laser irradiation was applied through a small skull window using a 1064 nm laser (1 W/cm², 5 min). NIR thermal imaging was utilized to monitor the local temperature of tumor during the irradiation.

GBM inhibition in vivo: Luc + GL261-bearing C57BL/6 mice were randomly divided into five groups, each consisting of six mice. The groups were allocated as follows: (1) Blank, (2) CS, (3) CS + laser, (4) CS@CM, and (5) CS@CM + laser. Tumor-bearing mice were injected via the tail vein with various agents at the injection dose of 5 mg/kg on day 0, 2 And 4. One day post-injection, the tumor sites were irradiated with a pulsed 1064 nm laser (1 W/cm², 5 min). Therapeutic efficacy was monitored at various time points using the IVIS imaging System. On day 28, GBM tumors from all the groups were harvested, fixed, and stained with H&E. Tumor tissues were finally examined using confocal microscopy to assess the antitumor efficacy.

Statistical analysis: Data are presented as mean ± standard deviation (SD, n = 4, unless stated otherwise). The error bars represent the SD of independent sample measurements. Statistical Analysis was conducted using GraphPad Prism 8.0 (GraphPad Software, USA). Differences between two groups were evaluated using an unpaired two-tailed Student’s t-test. Significance levels are indicated as ^*p < 0.05, ^**p < 0.01, and ^***p < 0.001.

Author contributions

S. Lin, H. Xing and Y. Zeng contributed equally to the study. S. Lin and H. Xing performed the majority of the experimental work, including in vitro and in vivo studies, data analysis, and interpretation. Y. Zeng prepared and edited most figures. S. Liu, H. Xing and Y. Zeng wrote the main manuscript. E. Galimova and A. Chernov provided critical feedback on the manuscript. G. Liu and P. Xue supervised the study and designed the research framework. P. Xue revised the manuscript with input from all authors.

Funding

This study is supported by Key Research and Development Project of Sichuan Provincial Science and Technology Plan (2024YFFK0249), Open Research Project from Anhui Provincial Key Laboratory of Tumor Evolution and Intelligent Diagnosis and Treatment (KFKT202405), Chongqing Innovative Medical Device Application Demonstration Project (CQEIC2024MDAD-035), Science and Technology Innovation Key R&D Program of Chongqing (CSTB2025TIAD-STX0010), and Chongqing Graduate Joint Training Base (lpjd202409).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.