Genetically engineered GSDME vesicle coated nanoparticle for immune activating and pyroptosis inducing synergetic therapy of fibrosarcoma

Abstract

Fibrosarcoma is a kind of highly malignant sarcoma with high recurrence rate after surgery and unsatisfying efficacy of radiotherapy and chemotherapy. Increasing the concentration of gasdermin in tumor cell to induce pyroptosis is a promising strategy to treat fibrosarcoma. However, the poor cytotoxic T lymphocytes (CTL) infiltration in tumor area limits the therapeutic effect with disabled positive feedback between pyroptosis and anti-tumor immunity. Induing tumor immunogenic cell death to activate anti-tumor immune and delivering gasdermin E (GSDME) to enhance tumor pyroptosis will rebuild the positive feedback. Hence, we proposed and constructed a genetically engineered GSDME vesicle coated nanoparticle (HV-FI) to achieve effective treatment of fibrosarcoma. First, the GSDME membrane-expressing tumor cells are constructed and the GSDME-displaying vesicles are extracted. Then the HV-FI is prepared by hybridizing red blood cell membranes and encapsulating ICG-loaded mesoporous Fe₃O₄ nanoparticles. The GSDME proteins on the vesicles can effectively induce fibrosarcoma cell pyroptosis with granzyme B. After NIR irradiation, fibrosarcoma cells released DAMPs and activated dendritic cells in vitro. In animal experiments, fibrosarcoma cells underwent pyroptosis and CTL infiltration was boosted in tumor microenvironment. That subsequently enhanced tumor cell pyroptosis with delivered GSDME, creating a synergetic effect of tumor cell pyroptosis and anti-tumor immune activation, ultimately leading to effective tumor clearance. Meanwhile, the immunological memory is established, preventing tumor recurrence after treatment. This work proposes and validates the applicability and effectiveness of nanotherapy based on enhanced tumor pyroptosis induction and immune activation, providing a new method for the treatment of unresectable fibrosarcoma.

Graphical abstract

1. Introduction

Fibrosarcoma is a type of malignant soft tissue sarcoma that originates from mesenchymal cells [1]. With insidious onset, typical cases such as retroperitoneal fibrosarcoma can grow to a large size, strongly compressing surrounding organs and leading to severe clinical consequences [2]. Currently, the preferred treatment for fibrosarcoma is surgery. However, the recurrence rate after surgery remains high while more patients are ineligible for surgical intervention. These patients with unresectable fibrosarcoma commonly undergo systemic treatments such as radiation and chemotherapy, which have limited efficacy and significant side effects [3].

Pyroptosis is a programmed inflammatory cell death pathway characterized by the formation of pores on the cell membrane, leading to cell swelling, rupture, and the release of cellular contents and a series of inflammatory factors such as IL-1β, IL-18, inducing inflammatory reactions [[4], [5], [6]]. Under stimuli such as toxic drugs, microbial infections, or endogenous danger signals, caspase proteins within the cells are activated, leading to the cleavage and activation of gasdermin proteins, producing abundant N-terminal fragments of gasdermin. Multiple N-terminal fragments of gasdermin proteins aggregate on the cell membrane to form pores, ultimately causing cell pyroptosis [[7], [8], [9], [10]]. Gasdermin E (GSDME) is one of the execution proteins of pyroptosis. It can be activated by Caspase-3 and granzyme B (GzmB) [[11], [12], [13]]. GzmB is one of the main components secreted by cytotoxic T lymphocytes (CTL) to exert anti-tumor function. It can activate both GSDME and caspase 3 protein [14]. Activated caspase-3 can also activate GSDME [15]. Therefore, this pathway shows high efficiency of activating GSDME. More importantly, activated GSDME (GSDME-NT) can mediate cell pyroptosis, releasing a series of damage-associated molecular patterns (DAMPs) to enhance infiltration of CTLs in the tumor microenvironment. The GzmB secreted by CTLs can further induce cell pyroptosis, leading to a positive feedback of cell pyroptosis and immune response in the tumor microenvironment [14,[16], [17], [18]].

Therefore, GSDME-based anti-tumor immunotherapy has attracted a lot of attention in tumor treatment area [[19], [20], [21]]. But the question remains whether the tumors response to pyroptosis induction. It is reported that gasdermin proteins are highly expressed in most normal cells but lowly expressed in most tumors, leading to tumor resistance to pyroptosis induction [22,23]. This is also one of the reasons why other treatment methods such as chemotherapy have poor therapeutic efficacy [24,25]. Inducing tumor expression of gasdermin artificially can effectively address this issue. It is reported that tumor cell pyroptosis can be enhanced by deliver gasdermin proteins to tumor or artificially overexpression of gasdermin proteins in tumor [[26], [27], [28]]. However, due to the immunosuppressive microenvironment created by the tumor, this positive feedback effect of cell pyroptosis and immune activation is greatly restricted [13]. Hence, it is necessary to employ other methods to trigger sufficient immune activation.

In recent years, biomimetic nanomaterials have been studied due to their high biocompatibility and functional diversity [29,30]. The red blood cell membrane, extracted from red blood cells, inherits the CD47 protein, which is a classic immune evasion-related molecule. By encapsulating nanoparticles with the red blood cell membrane, the circulation time in the bloodstream can be significantly extended, and this approach has been widely used in anti-tumor research [[31], [32], [33]]. Tumor cell membrane, which is extracted from tumor cells, are widely used in drug delivery because of their immune evasion and homologous targeting abilities [[34], [35], [36], [37]]. Meanwhile, genetic engineering membrane display technology is an effective method to modify target proteins onto cell membrane delivery vehicle and has been effectively applied in tumor treatment research [[38], [39], [40], [41]]. By using this technology to prepare tumor cell-derived nanovesicles displaying gasdermin, tumor-targeted delivery of gasdermin can be achieved.

Hence, we proposed and constructed a bio-engineering vesicle coated nanoparticle to achieve effective treatment of fibrosarcoma by enhanced pyroptosis induction and immune activation. The fibrosarcoma cell-derived nanovesicles expressing GSDME was prepared with protein membrane exhibition method [42]. Then hybrid vesicle was prepared by fusing red blood cell vesicle to enhance their immune evasion ability and prolong circulation time in vivo. Next, the mesoporous Fe₃O₄ nanoparticles was prepared and loaded with the photosensitizer ICG. Last, the nanoparticles were encapsulating with hybrid vesicle. After intravenous injection, the nanoparticles were aggregated in tumor areas with tumor homologous targeting ability. Under laser treatment, tumor cells undergo apoptosis and ferroptosis, thereby inducing preliminary anti-tumor immunity by releasing DAMPs. Then the GSDME protein on the tumor surface is cleaved and activated by cleaved-caspase3 released by dead cells and GzmB released by CTLs, inducing tumor cell pyroptosis. These pyroptosis cells further released inflammatory factors to activate and recruit more CTLs into the tumor microenvironment. More CTLs can release more GzmB to induce cell pyroptosis, enhancing the feedback loop to increase the tumor-killing effect, ultimately eradicating the tumor ( Scheme 1).

Scheme 1.

Scheme illustration of HV-FI for treatment of fibrosarcoma. a) The preparation of the HV-FI. b) Synergistic strategy of enhanced pyroptosis induction and immune activation. After intravenous injection, HV-FI is enriched to the tumor area and get into tumor cells. With NIR irradiation, the DAMPs released from dead tumor cells effectively activate the immune system. GzmB released by CTL activate the delivered GSDME, leading to pyroptosis and release of more DAMPs. The CTL infiltration is promoted and the positive feedback between pyroptosis and anti-tumor immunity was established.

2. Results and discussions

2.1. Construction of stable membrane-expressing GSDME tumor cells

To validate the project design, stable membrane-expressing full-length GSDME cells were constructed. The signal peptide and transmembrane region of EGFR protein were used to achieve membrane expression of GSDME. The full-length GSDME was located in extracellular region while the enhanced green fluorescent protein (EGFP) was placed in intracellular domain for subsequent screening. All sequences were acquired from GenBank and constructed into a lentiviral overexpression plasmid (Fig. S1).

Next, this plasmid and other lentiviral packaging plasmids were transfected into HEK 293T lymphocyte. After collecting the virus, further infection of WEHI 164 fibrosarcoma cells was carried out. After flow cytometry screening for EGFP-positive cells, cells were further selected in medium containing puromycin for 2 weeks until clear green fluorescence was observable in all cells under an inverted fluorescence microscope, indicating successful integration of the exogenous gene into the cell genome (Fig. S2).

To further validate the expression and cellular localization of GSDME-EGFP protein, we conducted immunofluorescence experiments using anti-GSDME antibodies and fluorescent secondary antibodies. Cells infected with lentivirus showed green fluorescence on the cell membrane, while cells without not infected with lentivirus infection did not exhibit green fluorescence. Cells infected with lentivirus and incubated with anti-GSDME antibodies and fluorescent secondary antibodies were labeled with red fluorescence, whereas cells not incubated with anti-GSDME antibodies did not bind to the fluorescent secondary antibodies and showed no red fluorescence, proving the expression of GSDME protein on the cell membrane surface in the experimental group (Fig. 1a). In the flow cytometry experiment, cells overexpressing GSDME-EGFP were labeled with red fluorescence after incubation with anti-GSDME antibodies and fluorescent secondary antibodies, while cells not incubated with anti-GSDME antibodies did not bind to the fluorescent secondary antibodies, similarly demonstrating the successful construction of stable membrane-expressing GSDME tumor cells (Fig. S3).

Fig. 1.

Preparation of HV. a) The membrane expression of fusion protein was verified by immunofluorescence. Blue: cell nucleus; Green: EGFP; Red: GSDME. Scale bar = 20 μm. b) Images of vesicles stained with membrane dye. Scale bar = 5 μm. c) SDS-PAGE showed the protein composition of vesicles. Red boxes show the hybrid vesicle inherited proteins from RV and GV. d) Pyroptosis occurred after tumor cells incubated with GSDME vesicle and GzmB. The arrows indicated pyroptosis. Scale bar = 50 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

2.2. Preparation of the hybrid vesicle

The tumor cell membrane vesicles expressing GSDME on the membrane surface (GV) were extracted and purified using the method previously reported. Western blot experiment was conducted to verify the presence of GSDME protein on GV. The results showed that vesicles extracted from tumor cells infected with lentivirus exhibited significant GSDME expression, while control group vesicles (CV) extracted from tumor cells not infected with lentivirus showed no apparent signal, demonstrating the presence of GSDME protein on the vesicles after extraction (Fig. S4).

Red blood cell vesicles (RV) were also extracted using a similar method, and after fusion of GV and RV using ultrasonication, hybrid membrane vesicles (HV) were obtained. Then validation of the fusion of hybrid membranes was carried out. Initially, GV and RV were labeled with two different cell membrane fluorescent dyes respectively. Then the fluorescence signals were observed after fusion under a fluorescence microscope. The results showed that the hybrid cell membrane exhibited both red and green fluorescence, confirming the presence of membrane phospholipid components from both RV and GV in the hybrid cell membrane (Fig. 1b).

Further analysis of the protein source on HV was conducted using SDS-PAGE gel and Coomassie brilliant blue staining, which revealed that the proteins on HV originated from both RV and GV (Fig. 1c). Additionally, WB experiments indicated the presence of GSDME and CD47 protein on HV after fusion procedure (Figs. S5–S6).

After the presence of GSDME protein on HV was validated, further verification of the activity of GSDME protein was carried out. By co-incubating WEHI 164 tumor cells with HV and GzmB, the cells exhibited significant vacuolization, a characteristic manifestation of cell pyroptosis. In contrast, the control group did not show any significant occurrence of cell pyroptosis, confirming that the GSDME on HV possessed the activity to induce cell pyroptosis. (Fig. 1d).

2.3. Preparation of HV-FI

Mesoporous Fe₃O₄ nanoparticles were synthesized and loaded with the photosensitizer indocyanine green (ICG) to obtain Fe-ICG nanoparticle (FI). Subsequently, FI encapsulated with HV (HV-FI) was prepared using ultrasonication. Then preliminary morphological characterization was conducted. The size and morphology of HV-FI were examined through transmission electron microscopy (TEM) and dynamic light scattering (DLS), revealing that the size of HV-FI was around 200 nm which was slightly larger than FI. The TEM images showed a fuzzy translucent layer at the edge of HV-FI nanoparticles while the edge of FI was clear, indicating the encapsulation of FI nanoparticles within HV (Fig. 2a and b). By measuring the zeta potential, it was found that mesoporous Fe₃O₄ nanoparticles exhibited a positive charge, while FI loaded with negatively charged ICG showed a larger negative charge. The cell membrane HV exhibited a weaker negative charge, and the charge of HV-FI fell between FI and HV (Fig. 2c). The ultraviolet absorption spectroscopy revealed that mesoporous Fe₃O₄ nanoparticles did not exhibit characteristic absorption peaks within this wavelength range, whereas ICG, FI, and HV-FI all displayed characteristic absorption peaks of ICG (Fig. 2d). These experimental results confirmed the presence of ICG in FI and HV-FI, as well as the effective encapsulation of FI by HV in HV-FI.

Fig. 2.

Preparation of HV-FI. a) TEM images of FI and HV-FI. Scale bar = 200 nm. b) The Size of HV-FI. c) The Zeta potential of HV-FI. d) Absorption spectra showed the FI nanoparticle was successfully encased in the HV. e) FC analysis of tumor cell after incubated with HV-FI. f) Fluorescent images of the tumor cell incubated with HV-FI. Blue: Cell nucleus; Green: vesicle fused to the cell membrane; Red: FI particle. Scale bar = 20 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

To verify the homologous targeting function of tumor vesicle, the cell membrane of RV-coated FI (RV-FI) and HV-FI were labeled with the DIO membrane dye. Subsequently, after co-incubation, the uptake of nanoparticles by tumor cells was observed using confocal microscopy. The results showed a significant green fluorescent signal within tumor cell in the HV-FI group, while the fluorescence in the RV-FI group was weaker. Additionally, the ICG fluorescence was stronger in the HV-FI group compared to the RV-FI group, with the FI group displaying the weakest ICG fluorescence (Fig. 2e). Meanwhile, FC assessment was performed and the data showed the ICG signal in HV-FI group was much higher than other control groups (Fig. 2f). These data demonstrated that the HV-FI group exhibited a good tumor-targeting ability.

2.4. Characterization of anti-tumor function of HV-FI in vitro

Following the confirmation of the preparation of HV-FI, further validation of the anti-tumor function of HV-FI was performed through in vitro experiments. First, the cytotoxic effect of HV-FI on tumor cells were evaluated in vitro. After incubating tumor cells with different concentrations of HV-FI and subjecting the cells to laser treatment, cell viability was assessed using the MTT assay. The results indicated that HV-FI effectively killed tumor cells under laser irradiation, with over 50 % tumor cell death observed at a concentration of 100 μg/ml of HV-FI (quantified based on Fe₃O₄ nanoparticles), while the cytotoxic effect of HV-FI without laser treatment was limited (Fig. S7).

Subsequently, the tumor-kill mechanism of HV-FI with laser was further demonstrated. To assess the ability of HV-FI with laser to induce apoptosis in tumor cells, tumor cells were incubated with HV-FI and subjected to laser treatment. Mitochondrial membrane potential detection probe JC-1 was added, revealing significant green fluorescence within the cells of the HV-FI Laser (+) group, indicating a significant decrease in mitochondrial membrane potential and early apoptosis in the cells (Fig. 3a and b).

Fig. 3.

Characterization of anti-tumor function of HV-FI in vitro. a) Fluorescent images of JC-1 probe indicated cell apoptosis after treated with HV-FI and laser. Scale bar = 20 μm. b) FC assessment of JC-1 probe indicated cell apoptosis after treated with HV-FI and laser. c) Fluorescent images of lipoperoxides distribution in tumor cells after treatment. Scale bar = 20 μm. d) Fluorescent images of ROS generated in tumor cells after treatment. Scale bar = 100 μm. e) FC assessment of ROS generated in tumor cells after treatment. f) Fluorescent images of Calcein&PI probe indicated cell membrane destruction and cell death after treated with HV-FI and laser. Scale bar = 40 μm. g-i) ICD occurrence reflected in cell membrane translocation of CRT, release of ATP and HMGB-1. I-III: PBS, HV-FI Laser (−), HV-FI Laser (+). N = 3, ∗∗P < 0.01, ∗∗∗P < 0.001, Student's t-test. j) DC maturity (CD80⁺CD86⁺) after incubated with tumor cell debris and releases. N = 3, ∗∗∗P < 0.001, Student's t-test.

Next, the functionality of HV-FI with laser in inducing ferroptosis in tumors was validated. By co-incubating tumor cells with HV-FI and subjecting them to laser treatment, ferroptosis-related indicators were evaluated using ROS detection probes, lipid peroxidation probes, and GSH detection reagents. The results showed that tumor cells in the HV-FI group exhibited noticeable green fluorescence on the cell membrane, indicating lipid peroxidation on the tumor cell membrane (Fig. 3c). Meanwhile, the data showed that HV-FI generated a significant amount of ROS within the cells under laser treatment (Fig. 3d and e), depleting intracellular reducing substances and progressively reducing the content of glutathione (Fig. S8). These results demonstrated that HV-FI with laser effectively induced oxidative damage related to ferroptosis in tumors.

Finally, tumor cells were treated using the same method and stained with calcein AM/propidium iodide to assess cell death. The results showed extensive cell death in the HV-FI Laser (+) group (Fig. 3f).

In the absence of laser irradiation, HV-FI didn't produce significant levels of ROS. It only exhibited the effects of iron overload. This iron overload alone could only cause slight lipid peroxidation of the cell membrane and was insufficient to induce effective cell killing. However, under laser irradiation, the photodynamic effects from ICG generated a substantial amount of ROS, significantly enhancing lipid peroxidation of the tumor cell membranes while also inducing apoptosis, as evidenced by a decrease in mitochondrial membrane potential.

Overall, the experiments demonstrated that incubation of HV-FI with tumor cells followed by laser treatment effectively induced apoptosis and ferroptosis in tumor cells, resulting in significant tumor-killing effect.

To investigate the release of DAMPs and activation of dendritic cells (DCs) in vitro, the tumor cells and supernatant from tumor cells treated with HV-FI + laser were collected. The ATP and HMGB-1 protein levels in the cell supernatant were measured, and the expression of cell surface calreticulin (CRT) was detected using flow cytometry. The results showed a significant increase in ATP and HMGB-1 protein levels in the cell supernatant of the HV-FI Laser (+) group, with HMGB-1 levels more than doubled compared to the control group and ATP levels more than quadrupled (Fig. 3g and h). Flow cytometry analysis revealed that the expression of CRT on the surface of tumor cells in the HV-FI Laser (+) group was significantly higher than that in the control group (Fig. 3i). Then the DCs were incubated with the treated tumor cell suspension. By analyzing the expression of DC surface markers CD80 and CD86 using flow cytometry and calculating the proportion of CD80⁺CD86⁺ DCs, it was found that the activation of DCs was approximately twice as high in the HV-FI Laser (+) group compared to the control group (Fig. 3j). These experiments demonstrated that tumor cells in HV-FI Laser (+) group underwent ICD and effectively activated DCs by releasing DAMPs.

Prior to conducting in vivo experiments, the biosafety of HV-FI was assessed in vitro. Hemolysis assays showed that HV-FI did not cause significant hemolysis within the therapeutic concentration range (Fig. S9). The cytotoxicity of HV-FI was tested using two normal cell lines and the data showed that the HV-FI had no significant cytotoxicity for non-homologous cells without laser treatment (Fig. S10). Besides, the size stability of HV-FI was tested and the date showed that the HV-FI remained stable within 3 weeks (Fig. S11). Further evaluation of HV-FI stability in plasma was performed and the ICG leakage was tested in data showed that ICG in supernatant of HV-FI group was much less than supernatant of FI, indicating a good stability of HV-FI in blood (Figure S12). Depending on the above experiments, it can be concluded that HV-FI exhibits good biocompatibility and is safe for in vivo application.

2.5. Application of HV-FI in vivo

First, the distribution of HV-FI in vivo was investigated. HV-FI was injected into mice via tail vein, and the distribution of HV-FI in the mouse body was monitored over time using live fluorescence imaging equipment. After reaching the final time point, the mice were euthanized, and major organs and tumors were isolated for fluorescence imaging. The results showed that HV-FI rapidly accumulated in the tumor area of the mice and remained there for an extended period, reaching optimal tumor-specific accumulation at 24 h with strong signal intensity (Fig. 4a–b). From excised organ imaging and statistical analysis, it was found that the tumor signal in the HV-FI group was significantly stronger than in the other control groups, with no excessive distribution in major organs (Fig. 4c and d). Besides, examination of tumor tissue under TEM revealed the distribution of the nanoparticles within the tumor (Fig. 4e).

Fig. 4.

Biodistribution test and treatment effect assessment in animal tumor models. a-b) Statistics and representative fluorescence images of living animal showed the in vivo accumulation characteristics of HV-FI. c) Representative fluorescence images of isolated organ and tumor. d) Statistics of fluorescence signal of isolated organ and tumor. N = 5, ∗∗P < 0.01, Student's t-test. e) TEM images of HV-FI remains in tumor tissue. Scale bar = 2 μm and 200 nm. f) Scheme of therapeutic schedule. g) Statistics of tumor volume change of mice in each group. N = 5, ∗∗∗P < 0.001, Student's t-test. h) Image of isolated tumors after finishing treatment. I-VII: PBS, FI, RV-FI, HV-FI, FI + laser, RV-FI + laser, HV-FI + laser.

Subsequently, the therapeutic efficacy of HV-FI was evaluated in tumor-bearing mice. HV-FI was injected via tail vein, followed by NIR laser irradiation of the tumor area after 24 h. Treatments were administered every three days for a total of three times (Fig. 4f). Tumor size changes were monitored during the treatment period. After the treatment ended, the mice were euthanized, and the tumors were isolated for observation. The results showed that the mice in the HV-FI + laser group exhibited effective tumor control and achieved a good therapeutic outcome (Fig. 4g and h).

Following the validation of the therapeutic efficacy in mouse tumors, tumor cell pyroptosis after treatment was further investigated. Observation through immunohistochemistry of mouse tumor sections revealed that in the three groups subjected to laser irradiation, FI + laser, RV-FI + laser, and HV-FI + laser, a significant presence of cleaved-Caspase3 protein was observed within the tumors. The distribution of cleaved-Caspase3 was more widespread in the RV-FI + laser and HV-FI + laser groups compared to the FI + laser group, with cleaved-Caspase3 being more concentrated intracellularly in the RV-FI + laser group and intercellularly in the HV-FI + laser group (Fig. 5a). This difference was attributed to cell pyroptosis occurring in the HV-FI + laser group, which released cleaved-Caspase3 into the extracellular space. Examination of HMGB-1 distribution in the tumors of each group showed that abundant HMGB-1 was released in intercellular substance after HV-FI treatment, indicating a significant ICD occurred and release of DAMPs (Fig. 5b). Further examination of GzmB distribution in the tumors of each group revealed that the HV-FI + laser group, due to the immunostimulatory effect, recruited more CTLs to the tumor area, leading to the release of GzmB (Fig. 5c). Subsequent Western blot detection of GSDME-NT in tumor tissues showed that the HV-FI and HV-FI + laser groups had more GSDME-NT distribution compared to other groups due to the introduction of exogenous GSDME. Furthermore, the HV-FI + laser group, with higher levels of cleaved-Caspase3 and GzmB in tumor tissue, exhibited the highest distribution of GSDME-NT, indicating significant GSDME-mediated cell pyroptosis (Fig. 5d, Fig. S13).

Fig. 5.

Evaluation of cell proptosis in tumor tissue and immune cytokines in serum. I-VII: PBS, FI, RV-FI, HV-FI, FI + laser, RV-FI + laser, HV-FI + laser. a) Representative IHC images of tumor tissue section detecting the cleaved-Caspase3. The cleaved-Caspase3 in HV-FI + laser group accumulated in the intercellular space while that in FI + laser group and RV-FI + laser group accumulated in cells. Scale bars = 50 μm and 20 μm. b) Representative IHC images of tumor tissue section detecting the HMGB-1. Scale bars = 50 μm. c) Representative IHC images of tumor tissue section detecting the GzmB. Scale bars = 50 μm d) Image of WB detecting the GSDME N-terminal in tumor tissue. e-h) ELISA experiment to detecting immune cytokines released in serum. N = 3, ∗P < 0.05, ∗∗P < 0.01, Student's t-test.

Then the immune activation and immune memory of tumor-bearing mice after treatment were evaluated. The mouse serum was collected, and relevant cytokines were detected using ELISA kits. The HV-FI + laser group showed significantly higher cytokine secretion level compared with other treatments, indicating high immunostimulatory efficiency (Fig. 5e–h).

Furthermore, cells from mouse tumors and lymph nodes were isolated for flow cytometric analysis. In tumor tissue, the proportion of CD45⁺ immune cell in the HV-FI + laser group was more than 40 % higher than that in other control groups. Meanwhile, the proportion of CD8⁺ T lymphocyte in the HV-FI + laser group was more than 30 % higher than that in other laser groups (FI + laser and RV-FI + laser) and more than twice as many as that in other no laser groups (PBS, FI, RV-FI, HV-FI). In lymph node, the proportion of mature DCs in the HV-FI + laser group was more than 40 % higher than that in other laser groups and more than twice as many as that in other no laser groups. The proportion of CD8⁺ T lymphocyte in the HV-FI + laser group was more than 25 % higher than that in HV-FI group, FI + laser group and RV-FI + laser group Additionally, the proportion of CD8⁺ T lymphocyte in the HV-FI + laser was more than 80 % higher than that in PBS group, FI group and RV-FI group. Meanwhile, the proportion of central memory CD8⁺ T lymphocyte (Tcm) in the HV-FI + laser group was more than 40 % higher than that in other laser groups and more than five times as many as that in no laser groups. (Figure S14, Fig. 6a–f). These experimental results indicated that the HV-FI + laser group significantly activated the anti-tumor immune response in tumor-bearing mice and established effective immune memory after treatment.

Fig. 6.

Evaluation of immune activation and immune memory establishment. a) Flow cytometry analysis of CD45⁺ immune cells in tumor, mature DCs in lymph node and CD8⁺ lymphocyte changing in tumor and lymph node. b-f) Statistics of immune cells changing: b-c) immune cells (CD45⁺) and CD8⁺ T lymphocytes (gated on CD45 and CD3) in tumor; d) Mature DCs in lymph node (gated on CD45 and CD11c). e-f) CD8⁺ T lymphocyte (gated on CD45 and CD3) and CD8⁺ T_cm (gated on CD45, CD3 and CD8) in lymph node. f) The tumor volume change in tumor rechallenge experiment.

Next, secondary tumor inoculation was performed on mice after HV-FI + laser treatment while untreated mice were used as control. The results showed that the immune system of treated mice effectively suppressed and cleared the tumor cells, demonstrating the effective establishment of immune memory (Fig. 6g).

Finally, the biosafety of HV-FI + laser application in vivo was assessed. First, H&E staining of major organs from mice treated with HV-FI + laser showed no significant abnormalities in organ morphology compared with control groups (Fig. S15). Then analysis of changes in mouse body weight during treatment and testing of serum liver and kidney function markers revealed no significant differences among the groups, indicating that HV-FI + laser application in vivo was safe (Figs. S16–S18).

3. Conclusion

We successfully prepared nanovesicles displaying GSDME on the membrane surface. By further hybridizing red blood cell membranes and encapsulating FI nanoparticles, we prepared tumor therapeutic biomimetic nanomaterials HV-FI. Through in vitro experiments, it was verified that HV-FI + laser could effectively induce apoptosis and ferroptosis in tumor cells and subsequently activate DCs with released DAMPs. Further validation through in vivo experiments demonstrated that HV-FI could efficiently accumulate in tumor areas by homologous targeting effect of tumor cell membrane, leading to effective tumor elimination under NIR laser irradiation. Subsequent series of experiments confirmed the presence of abundant GzmB, cleaved-caspase3, and GSDME-NT in the tumor area of the HV-FI + laser group, indicating significant cell pyroptosis and enhanced CTL infiltration in tumor area. Further effective activation of anti-tumor immunity and establishment of effective immune memory in the HV-FI + laser group was proved through the detection of serum cytokines and T lymphocytes, as well as a secondary tumor inoculation experiment in treated mice. Additionally, no major organ dysfunction was observed in mice during treatment, demonstrating the good biosafety of HV-FI + laser application. Hence, this study established an efficient synergy between enhanced cell pyroptosis induction and immune activation, providing a credible strategy for the effective treatment of fibrosarcoma.

4. Experimental section

4.1. Materials

Puromycin and Anti-rabbit second antibody (HRP) were purchased from Invitrogen. Lipo6000 transfection reagent, Calcein/PI Assay Kit, ROS assay kit and ATP detection assay kit were purchased from Beyotime China. Lentiviral plasmid, Hoechst and TNF-β ELISA kit were purchased from Sangon China. Anti-rabbit second antibody (Alexa647) was purchased from Affinity China. Anti-GSDME antibody and anti-caspase3 antibody were purchased from Proteintech. Anti-GSDME-NT antibody, anti-calreticulin antibody, Anti-sodium-potassium ATPase antibody, Anti-CD47 antibody, DID, DIO and firefly luciferase substrate were purchased from Abcam. Granzyme B and JC-1 were purchased from MCE. Glutathione test kit was purchased from Acmec China. IL-1 ELISA kit, IL-2 ELISA kit, HMGB-1 ELISA kit, TNF-α ELISA kit and IFN-γ ELISA kit were purchased from Abclonal China. Anti-CD45 antibody, Anti-CD80 antibody, Anti-CD86 antibody, Anti-CD11c antibody, Anti-CD3 antibody, Anti-CD44 antibody and Anti-CD62L antibody were purchased from Biolegend. PMSF and BODIPY-C11 were purchased from Sigma-Aldrich. WEHI-164 cell line (GDC0092) was purchased from China Centre for Type Culture Collection. LO2 and HEK293T cell line were acquired from Prof. Yu's laboratory.

4.2. Methods

4.2.1. Generation and characterization of GSDME expressing cell line

4.2.1.1. Generation of GSDME membrane-expressing cell line

Vector containing full-length GSDME coding sequence was constructed. EGFP was used to verify the cell transfection and protein expression. The signal peptide and transmembrane part of EGFR was used to achieve protein membrane expression. These coding sequences (from GenBank) were cloned into the lentiviral plasmid. Subsequently, the plasmids were transfected into HEK293T cells with packaging plasmids following instructions described in the manual. Lentivirus was collected and used to infect the WEHI164 cells. To establish the stably membrane expression cells, infected WEHI164 cells were further selected with flow cytometer and incubated in DMEM medium containing 10 % FBS and 2 μg/mL puromycin for 2 weeks.

4.2.1.2. Characterization of GSDME expressing cell line

The expression of GSDME on the cell membrane was analyzed by indirect immunofluorescence (IF) and flow cytometer (FC). In IF assay, the GSDME-overexpressing cells (5.0 × 10⁵ cells/well) were seeded in a coverglass bottom dish. The cells were incubated overnight and washed by cold PBS for three times. Then the cells were fixed with 4 % paraformaldehyde for 20 min. Subsequently, the cells were incubated with anti-GSDME antibody (1:100, 1 h at 37 °C) and secondary antibody conjugated to AlexaFluor647 (1:200, 1 h at 37 °C), cell nuclei was stained by Hoechst for 5 min. Between each step, the cells were washed for 3 times with PBS. The parental cells and infected cells without primary antibody incubation were used as controls. Fluorescence images were acquired by a confocal laser scanning microscope (Zeiss, LSM880). In FC assay, the GSDME-overexpressing cells were seeded in a 6 cm cell dish. The cells were incubated overnight. Then the cell suspension was prepared and washed by cold PBS for three times. Subsequently, the cells were incubated with anti-GSDME antibody (1:100, 30 min at 37 °C) and secondary antibody conjugated to AlexaFluor647 (1:200, 30 min at 37 °C). Between each step, the cells were washed for 3 times with PBS. The infected cells without primary antibody were used as a control. Fluorescence signals were acquired by a flow cytometer (Beckman, Cytoflex).

4.2.2. Generation of hybrid vesicle (HV)

The GSDME-overexpressing cells were washed by cold PBS twice to remove cell secretion and culture medium. Then the cells were suspended in PBS mixed with protease inhibitor and sonicated under low power (22.5 W, 1 min) in ice bath. Then GSDME nanovesicles (GV) were isolated by multi-steps ultracentrifugation. The blood cells were isolated from whole blood of mice. The red blood cell vesicles (RV) were generated with similar procedure. Then GV and RV was mixed in a 1:1 mass ratio. The hybrid vesicles (HV) were acquired by ultrasound and ultracentrifugation.

4.2.3. Characterization of hybrid vesicle (HV)

The protein component of HV was compared with RV and GV by coomassie brilliant blue stained (RT, 1h) SDS-PAGE gel. The membrane phospholipid component of HV was compared with RV (Labeled with DIO) and GV (Labeled with DID) by fluorescence imaging. Briefly, the HV was generated with RV (Labeled with DIO) and GV (Labeled with DID). Then the HV was absorbed to a slide and observed under inverted fluorescent microscope (Olympus, IX51). To verify the presence of GSDME on the HV, the western blot assay was performed with anti-GSDME antibody (1:1000, 4 °C, overnight) and HRP secondary antibody (1:10000, RT, 1h). The exposure condition was the default automatic mode. RV and GV were used as controls.

4.2.4. Preparation and of hybrid vesicle coated Fe-ICG nanoparticle (HV-FI)

4.2.4.1. Synthesis of mesoporous Fe₃O₄ nanoparticles

Hexahydrate ferric chloride and sodium citrate were weighted (mass ratio of 27:1) and dissolved in ethylene glycol. Sodium acetate was dissolved in ethylene glycol and slowly added it to solution above. The mixture was stirred for 30 min and then heated at 200 °C for 16 h. The product was separated with a magnet and washed with ultrapure water and anhydrous ethanol.

4.2.4.2. Preparation of the HV-FI

Indocyanine green powder was dissolved in an ethanol solution. Mesoporous Fe₃O₄ nanoparticles were resuspended in an ethanol solution. The two components were mixed in equal masses, thoroughly vortexed, and transferred to a test tube. The test tube was placed on a stirrer and stirred until the liquid was reduced to half to one-third of the original volume. The mixture was centrifuged at 12,000 g at room temperature for 15 min to remove the liquid. The sediment was washed twice with PBS to remove unbound indocyanine green. The iron-indocyanine green nanoparticles (FI) were resuspended in PBS, and 3 times the mass of hybrid vesicles was added and thoroughly vortexed. The mixture was sonicated in an ice bath for 1 min and placed on ice for 30 min. Then the mixture was centrifuged at 4 °C, 12,000 g for 15 min. The precipitate was collected, washed twice with PBS to remove blank vesicles.

4.2.5. Physics characterization of HV-FI

Particle size and Zeta potential of HV-FI was measured with DLS/Zeta instrument (Malvern, ZS 90). The morphology of HV-FI was observed with a transmission electron microscope (Thermo Scientific, G2 Spirit Biotwin). The Ultraviolet absorption spectrum of HV-FI was measured with a microplate reader (Thermo Scientific, Varioskan Flash). The size stability of HV-FI was tested by setting HV-FI in PBS at RT for 30 days and the DLS was used to assess the size change. The structure stability of HV-FI was tested by setting HV-FI in plasma at RT for 3 days and ICG in supernatant was detected using a microplate reader (OD780).

4.2.6. Functional characterization of HV-FI in vitro

4.2.6.1. Test of tumor cell targeting function of HV-FI

The RV and HV were labeled with green fluorescent membrane dye DIO. WEHI 164 cells were subcultured into confocal dishes and divided into three groups. FI, RV-FI and HV-FI were added to the three groups separately and then incubated at 37 °C for 4 h. Removed the supernatant and cells were washed three times with PBS. Then cells were observed with a confocal laser scanning microscope (CLSM, Zeiss, LSM880). For FC assessment, the WEHI 164 cells were subcultured into confocal dishes and divided into three groups. FI, RV-FI and HV-FI were added to the three groups separately and then incubated at 37 °C for 4 h. Removed the supernatant and washed cells for three times with PBS. Then cells were tested with a flow cytometry (FC, Cytoflex).

4.2.6.2. Cytotoxicity testing of HV-FI In vitro

WEHI 164 tumor cells were subcultured in a 96-well plate to a concentration of 10,000 cells per well with a liquid volume of 200 μl. After overnight incubation, different concentrations of HV-FI were added to the wells. After 4 h incubation, cells were treated with a laser (808 nm, 1W/cm²) for 3 min, then returned to the incubator. After 24 h of continued incubation, MTT was added to the wells. After 4 h incubation, the supernatant was carefully removed and the formazan was dissolved with DMSO. The absorbance at 490 nm was measured using a microplate reader (Thermo Scientific, Varioskan Flash).

LO2, 293T and NIH3T3 cells were subcultured in a 96-well plate to a concentration of 10,000 cells per well with a liquid volume of 200 μl. After overnight incubation, different concentrations of HV-FI were added to the wells. After 24 h of continued incubation, MTT was added to the wells. After 4 h incubation, the supernatant was carefully removed and the formazan was dissolved with DMSO. The absorbance at 490 nm was measured using a microplate reader (Thermo Scientific, Varioskan Flash).

4.2.6.3. Detection of ROS generation

WEHI 164 cells were subcultured to prepare cell slides. FI and HV-FI were respectively added. After 4 h incubation, the supernatant was removed and cells were washed three times with PBS. Then the ROS probe DCFH-DA (10 μM) was added. After 1 h incubation, the supernatant was removed, and cells were washed three times with PBS. Then cells were treated with a laser (808 nm, 1W/cm²) for 3 min. Observation and photography were conducted using a CLSM (Zeiss, LSM880) and FC (Cytoflex).

4.2.6.4. Detection of glutathione (GSH) depletion

WEHI 164 cells were subcultured in 6 cm cell culture dishes and grouped by time points. Then HV-FI was added. After 4 h incubation, the supernatant was removed and cells were washed three times with PBS. Then cells were treated with a laser (808 nm, 1W/cm²) for 3 min. Cells were returned to the incubator and collected at different time points. Collected cells were centrifuged and resuspended in reagent 1 of GSH detection kit. Then the GSH concentration was tested fallowing the instruction manual.

4.2.6.5. Detection of cell membrane peroxidation

Tumor cell slides was prepared. Then HV-FI was added and incubated for 4 h. Removed the cell culture supernatant and washed cells for 3 times with PBS. Cells were treated with a laser (808 nm, 1W/cm²) for 3 min and then incubated in incubator for 1 h. Lipid peroxidation probe solution was added, and cells were incubated at 37 °C for 30 min. Cells were observed using a CLSM (Zeiss, LSM880).

4.2.6.6. Mitochondrial membrane potential detection

Tumor cell slides was prepared and HV-FI was added. After 4 h incubation, the supernatant was removed and cells were washed for 3 times with PBS. Cells were treated with a laser (808 nm, 1W/cm²) for 3 min and then incubated in incubator for 1 h. Then mitochondrial membrane potential probe JC-1was added to a final concentration of 2 μM, and cells were incubated at 37 °C for 15 min. Next, the supernatant was removed, and the cells were washed for 3 times with PBS. Cells were assessed using a CLSM (Zeiss, LSM880) and FC (Cytoflex).

4.2.6.7. Cell Calcein/PI staining assay

Tumor cell slides was prepared and HV-FI was added. After 4 h incubation, the supernatant was removed and cells were washed for 3 times with PBS. Cells were treated with a laser (808 nm, 1W/cm²) for 3 min and then incubated in incubator for 1 h. Next, the cell culture medium was replaced with the calcein/PI working solution. After 30 min incubation, the supernatant was removed. Cells were washed for 3 times with PBS and tested with a CLSM (Zeiss, LSM880).

4.2.6.8. Validation of immunogenic cell death (ICD) of tumor cells

WEHI 164 cells were subcultured into 6 cm cell culture dishes and HV-FI was added. After 4 h incubation, the supernatant was removed and cells were washed for 3 times with PBS. Then cells were put into a 96-well plate, treated with the laser (808 nm, 1W/cm²) for 3 min and then incubated in incubator for 1 h. Part of the cell suspension was centrifuged to separate the cells and supernatant. First, the cells were used to detect the calreticulin on cell membrane using FC (Beckman, Cytoflex). Briefly, the cells were sequentially incubated with anti-calreticulin antibody (1:100, 37 °C for 30 min) and fluorescent secondary antibody (1:200, 37 °C for 30 min). Next, the concentration of high mobility group box 1 protein (HMGB-1) and ATP in supernatant were detected using experimental kits following the user manual. Last, another part of cell suspension (both cells and supernatants) was incubated with DC cells. After 24h, the DC cells were incubated with anti-CD80 fluorescent antibody and anti-CD86 fluorescent antibody synchronously (1:200, 37 °C for 30 min) and tested with a FC (Beckman, Cytoflex).

4.2.7. Functional characterization of HV-FI in vivo

4.2.7.1. Construction of tumor-bearing mouse model

All animal experiments were approved by the Management and Ethics of Laboratory Animals Committee of Xiamen University (NO. XMULAC20220245). All laboratory mice (Balb/c, 5 weeks old) were acquired from Xiamen university laboratory animal center (XMULAC) and kept in facility of XMULAC in conditions that meet animal welfare requirements. WEHI 164 cells suspension was prepared, counted, and adjusted to a concentration of 5 × 10⁶ cells per 150 μl. The hair of Balb/c mice on the left dorsal side was shaved off. Each mouse was subcutaneously injected with 150 μl of cell suspension on the left dorsal. The tumor size was observed and recorded regularly.

4.2.7.2. Imaging of tumor-bearing mice In vivo

Twelve Balb/c tumor-bearing mice were evenly divided into 4 groups, and the hair on the left dorsal side was shaved off. ICG, FI, RV-FI and HV-FI were injected separately via the tail vein of the mice. Infrared fluorescence images were captured at 0 h, 6 h, 12 h, 24 h, 36 h, 48 h, and 60 h post-injection using a small animal in vivo imaging equipment (Perkinelmer, IVIS Lumina III). Then the mice were sacrificed. The tumors, hearts, livers, spleens, lungs, and kidneys were dissected and infrared fluorescence imaging of them were captured.

4.2.7.3. Evaluation of therapeutic effect of HV-FI on tumor-bearing mice

35 Balb/c tumor-bearing mice were randomly divided into 7 groups. PBS, FI, RV-FI and HV-FI were injected into the mice (except for PBS, all others were divided into two groups: laser group and non-laser group). After 24 h of injection, the tumor area of the corresponding group of mice was treated with 808 nm laser (1W/m²) for 3 min. Treatment was performed every three days, three times in total. The tumor size and mouse weight were recorded every 3 days for a total of 15 days.

4.2.7.4. Detection of cell pyroptosis after tumor-bearing mouse treatment

After treatment, the mice were sacrificed and tumors were dissected. Part of tumor tissues were frozen in liquid nitrogen. Another part of tumor tissues was fixed, dehydrated and embedded in paraffin using conventional methods. Paraffin blocks were subjected to tissue chemical staining using conventional methods to detect GzmB and cleaved Caspase-3 proteins, with anti-GzmB and anti-Cleaved-Caspase3 antibodies diluted at a ratio of 1:200. The protein of frozen tumor tissues was extracted. Then the protein samples were prepared and the WB assay was performed to detect GSDME-NT in the tissue, with anti-GSDME-NT antibody (1:1000, 4 °C overnight) and HRP-conjugated secondary antibody (1:10000,1 h at room temperature).

4.2.7.5. Detection of immune activation status in tumor-bearing mice after treatment

The mice after treatment were sacrificed and the serum, tumor tissues, lymph nodes were sampled. The concentration of IFN-γ, TNF-α, IL-1, and IL-2 were detected using ELISA assays kits. Cells in tumor tissues, lymph nodes were separated. Then FC assay was performed using fluorescently labeled antibodies against CD45, CD11c, CD80, CD86, CD3, CD8, CD44 and CD62L (1:200, 30 min at 37 °C). Five tumor-bearing mice received aforementioned treatment using HV-FI. Then the mice were subcutaneously injected with tumor cells on another dorsal side of mice. Healthy mice subcutaneously injected with tumor cells on right dorsal side of mice were used as controls. The tumor size was observed and recorded regularly.

4.2.8. Biosafety evaluation

4.2.8.1. The hemolysis test in vitro

Mouse RBC suspension was extracted and diluted to an appropriate concentration with saline. Then different concentration of HV-FI were added respectively. Saline was used as negative control and deionized water (DW) was used as positive control. After incubated for 2 h at 37 °C, the mixture was centrifugated at 12000 rpm for 10 min at room temperature and the supernatant was collected. The absorbance of the supernatant at 575 nm was measured and the hemolysis rate was calculated as following formula:

hemolysis (%) = (Ab₅₇₅ HV-FI - Ab₅₇₅ saline) / (Ab₅₇₅ DW - Ab₅₇₅ saline) × 100 %

4.2.8.2. Organ toxicity evaluation

Whole blood was sampled and the serum was separated every 3 days after starting treatment. The concentration of ALT, AST and BUN was detected using an automatic biochemical analyzer (Mindray). The hearts, livers, spleens, lungs and kidneys were dissected. Fixation, dehydration, embedding, sectioning and H&E staining were performed according to routine method. Then the images of samples were observed and captured.

4.2.9. Data statistical methods

All data were presented as mean ± standard deviation (N ≥ 3). Statistical significance was determined by ordinary two-tailed Student's t-test using GraphPad Prism software. P value < 0.05 was considered as statistically significant. Statistical values were displayed in figures as following: ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.

CRediT authorship contribution statement

Fenglin Miao: Writing – review & editing, Writing – original draft, Visualization, Validation, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Zhao Wang: Writing – review & editing, Visualization, Validation, Methodology, Investigation, Conceptualization. Qing Wang: Writing – review & editing, Visualization, Validation, Methodology, Investigation, Data curation. Changsheng Zhou: Writing – review & editing, Visualization, Validation, Methodology, Investigation, Conceptualization. Xiao Zhou: Investigation, Methodology. Jialiang Zheng: Visualization, Methodology, Investigation, Conceptualization. Yuan Ma: Methodology, Investigation. Zhenhang Lin: Methodology, Investigation. Yilai Gao: Supervision, Resources, Project administration, Funding acquisition. Ting Wu: Supervision, Software, Resources, Project administration, Methodology, Formal analysis. Yong Zhang: Supervision, Software, Resources, Project administration, Funding acquisition, Formal analysis. Jing Gao: Methodology, Investigation. Wengang Li: Writing – review & editing, Supervision, Software, Resources, Project administration, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Ethics approval and consent to participate

All animal experiments were approved by the Management and Ethics of Laboratory Animals Committee of Xiamen University (NO. XMULAC20220245).

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82272935), Natural Science Foundation of Xinjiang Uygur Autonomous Region (No.2023D01A56), Joint laboratory of School of Medicine, Xiamen University-Shanghai Jiangxia Blood Technology Co. Ltd (No. XDHT2020010C), Joint research center of School of Medicine, Xiamen University-Jiangsu Charity Biotech Co. Ltd (No.20233160C0002).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.