A biomimetic magnetic MOF-based nanoplatform for H₂S-mediated thermal re-sensitization and immune reprogramming in multimodal hyperthermia therapy

Abstract

Magnetic hyperthermia therapy (MHT) has emerged as a promising anti-cancer strategy due to its precise spatial controllability and immune-activating effects. However, tumor cells can rapidly develop thermotolerance through the upregulation of heat shock proteins (HSPs), activation of the NF-κB signaling pathway, and recruitment of immunosuppressive cells, etc. Herein, we design a tumor cell membrane coated H₂S-releasing magnetic nanoplatform (CmMN@ADT) to overcome this resistance. This nanoplatform was synthesized by coordinating Fe₃O₄ nanoparticles with 1,3,5-benzenetricarboxylic acid (BTC) to form a magnetic metal organic framework (Fe₃O₄@MIL-100, MN), which was subsequently loaded with the H₂S donor ADT-OH and coated with tumor cell membranes for homotypic targeting. Upon exposure to an alternating magnetic field (AMF), the Fe₃O₄@MIL-100 core enables localized hyperthermia, while acidic tumor microenvironment triggers ADT-OH release for sustained H₂S generation. The released H₂S enhances tumor cell sensitivity to hyperthermia by inhibiting NF-κB activation and downregulating HSP expression. Suprisingly, H₂S can also augment the MN induced ferroptosis. In vitro and in vivo studies have demonstrated that CmMN@ADT effectively induces tumor ablation and elicits potent anti-tumor immune responses, ultimately achieving the inhibition of the growth of both primary and metastatic tumors. Collectively, this study presents a novel H₂S driven magnetic MOF nanoplatform that achieves dual mode synergy between H₂S Augmented MHT and ferroptosis, providing a mechanistically guided strategy to overcome tumor thermotolerance and achieve durable tumor suppression.

Graphical abstract

1. Introduction

The complexity of the tumor microenvironment (TME) presents a significant challenge to effective cancer therapy, characterized by hypoxia, acidosis, and an abundance of immunosuppressive components, etc [[1], [2], [3]]. To overcome these barriers, emerging strategies have focused on remodeling the tumor immune microenvironment (TIME) to enhance anti-tumor immunity [[4], [5], [6], [7]]. Among them, magnetic hyperthermia therapy (MHT) has attracted increasing attention due to its precise spatial controllability, potent thermal cytotoxicity, and ability to stimulate immune responses with minimal systemic toxicity [[8], [9], [10], [11]]. Nevertheless, in tumors that have established a dynamic equilibrium and engage in chaotic interactions with the host system, effectively treating malignant cells remains challenging. The inherent unpredictability and adaptive responses of these complex systems to external perturbations often contribute to therapeutic resistance. Especially, recent studies have revealed that tumor cells can develop thermotolerance after hyperthermia therapy through the upregulation of heat shock proteins (HSPs), activation of the nuclear transcription factor-κB (NF-κB) signaling pathway, and recruitment of myeloid-derived suppressor cells (MDSCs), which further significantly limit MHT efficacy [[12], [13], [14]]. Consequently, developing new and efficient methods to break thermal resistance represents a major challenge for improving the efficacy of MHT (see Scheme 1).

Scheme 1.

ADT-OH-loaded Fe₃O₄@MIL-100 nanoplatform (CmMN@ADT) for magneto-thermal ablation sensitization and NF-κB inhibition to amplify antitumor immunity.

Hydrogen sulfide (H₂S), as an endogenous gaseous signaling molecule, has been shown to enhance tumor cell sensitivity to hyperthermia, suppress activation of the NF-κB pathway, and further promote systemic anti-tumor immune responses, rendering it highly suitable for effectively overcoming the current limitations of MHT and enabling more durable tumor suppression [[15], [16], [17], [18]]. Nevertheless, the targeted delivery and controlled release of H₂S in tumor tissues remain formidable challenges, as conventional sulfide or polysulfide donors suffer from instability and poor biodegradability [19,20]. Notably, ADT-OH, a new organic H₂S donor, exhibits excellent stability, biocompatibility and safety, making it a promising candidate for therapeutic application [[21], [22], [23]]. Therefore, given the synergistic potential of H₂S-Augmented MHT, the development of nanomedicines capable of concurrently delivering H₂S and enabling MHT represents a significant advancement toward overcoming current therapeutic limitations.

Furthermore, iron oxide nanoparticles (IONs) have been extensively investigated for cancer MHT due to their unique superparamagnetic properties, nanoscale dimensions, favorable biocompatibility, surface functionalizability, and high magnetic responsiveness [[24], [25], [26]]. However, conventional ION-based MHT systems typically rely on solid magnetic nanoparticles with high saturation magnetization—such as γ-Fe₂O₃, Fe₃O₄, or α-Fe₂O₃—to achieve effective thermal ablation, which may hinder integration with H₂S-based therapeutic approaches [27,28]. In contrast, metal–organic frameworks (MOFs) offer high porosity, large specific surface areas, and tunable structural features, making them promising precursors for engineering porous magnetic nanocomposites [[29], [30], [31]]. Recently, magnetic MOF-based platforms have demonstrated considerable potential for synergistic cancer treatment strategies [32,33]. Among them, Fe₃O₄@MIL-100 integrates the advantageous properties of both IONs and MOFs, exhibiting superior performance in drug delivery applications for H₂S Augmented MHT, and synergistic treatment with MHT and ferroptosis. More importantly, Fe₃O₄@MIL-100 can be encapsulated by cell membranes to achieve targeted delivery [34].

Based on the aforementioned analysis, this study designed a multifunctional regulatory system, CmMN@ADT, which combines H₂S release capability with alternating magnetic field (AMF) responsive thermosensitivity. Specifically, the system was constructed through a coordination reaction between Fe₃O₄ nanoparticles and 1,3,5-benzenetricarboxylic acid (H₃BTC) under acidic conditions, forming a magnetic iron-based metal–organic framework designated as Fe₃O₄@MIL-100 (MN). The MN structure enables H₂S release via loading and delivery of ADT-OH, while the Fe₃O₄ component responds to AMF stimulation to induce effective MHT. Notably, the surface of CmMN@ADT is coated with tumor cell membranes, conferring homotypic targeting ability that facilitates specific recognition and internalization by tumor cells. In vitro and in vivo experimental results demonstrate that upon AMF exposure, the system generates localized heating, leading to thermal ablation of tumor cells. Concurrently, under the weakly acidic condition characteristic of the TME, the outer MOF structure undergoes degradation, releasing ADT-OH, which subsequently generates H₂S. This induces intracellular acidification and enhances chemodynamic therapy (CDT) efficacy. Furthermore, H₂S can overcome adaptive drug resistance in the tumor microenvironment (TME) following mild hyperthermia therapy (MHT) through multiple mechanisms, including downregulating the expression of heat shock proteins (HSPs), inhibiting the NF-κB signaling pathway, suppressing ATP synthesis, and reducing the infiltration of immunosuppressive cells. Collectively, this nanoplatform not only enables efficient MHT ablation through Fe₃O₄@MIL-100 but also functions as a robust H₂S gas delivery system, thereby offering a promising solution to the limitations of MHT and facilitating a dual mode synergy between MHT and ferroptosis. Ultimately breaking the chaotic balance of TME and improving the cure rate for tumors.

2. Results and discussion

2.1. Synthesis and characterization of CmMN@ADT

Based on previous studies, Fe₃O₄@MIL100 magnetic MOF nanoplatforms (MNs) were constructed by coating Fe₃O₄ nanoparticles with the metal-organic framework material MIL100 (Fe) [29,35]. Subsequently, cancer cell membrane (Cm) was modified onto the MN@ADT surface through co-extrusion to synthesize CmMN@ADT (Fig. 1A). Transmission electron microscopy (TEM) revealed that CmMN@ADT exhibits a spherical structure with an average diameter of approximately 140 nm (Fig. 1B), which is consistent with the results of dynamic light scattering (DLS) (Fig. 1C). The size of the nanoparticles increased with synthesis: 106.2 ± 5.07 nm (MN), 117.9 ± 2.22 nm (MN@ADT), and 144.4 ± 15.12 nm (CmMN@ADT). Additionally, the zeta potential of the nanoplatform changed during this process: −31.6 ± 0.8 mV (MN), −17.33 ± 1.3 mV (MN@ADT), and −21.13 ± 0.55 mV (CmMN@ADT) (Fig. 1C). These changes in size and zeta potential confirmed the successful completion of each synthesis step. Besides, elemental mapping analysis (Fig. 1E) and UV-visible absorption spectroscopy (Fig. 1F) further confirmed the successful loading of ADT-OH. X-ray diffraction (XRD) analysis also revealed similar diffraction peaks in Fe₃O₄, MIL-100, and MN samples, indicating successful deposition of Fe-MOF onto the Fe₃O₄ surface (Fig. S1). Subsequently, we tested the magnetic responsiveness of Fe₃O₄ and MN using a vibrating sample magnetometer (VSM). The results showed that both Fe₃O₄ and MN can respond to magnetic fields, thus being applicable for magnetic hyperthermia therapy (Fig. S2). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis demonstrated that CmMN@ADT retained the Cm protein (Fig. 1D), which not only effectively prevents CmMN@ADT from being cleared by the mononuclear phagocyte system (MPS), but also endows it with targeted functionality.

Fig. 1.

CmMN@ADT fabrication and characterization. (A) Synthesis of MN, preparation of CmMN@ADT. (B) TEM image of CmMN@ADT. (C) Zeta potential and DLS characterization of MN, MN@ADT and CmMN@ADT. (D) Obtain the CmMN@ADT protein profile by SDS-PAGE. (E) Element mapping of CmMN@ADT. (F) Ultraviolet-visible absorbance of MN, ADT-OH and CmMN@ADT. (G) The influence of acidic environment on the release of CmMN@ADT drugs. (H) CmMN@ADT hemolysis test. (I, J) Real-time temperature change and warming curves of CmMN@ADT under the action of an alternating magnetic field for 10 min. (K) Magnetic-thermal stability properties of CmMN@ADT after five heating cycles.

To simulate the release of ADT-OH from CmMN@ADT within tumor cells, CmMN@ADT was placed in dialysis bags in environments with pH values of 6.5 and 7.4, and samples were taken at different time points for ultraviolet-visible absorption spectroscopy detection. The results indicated that CmMN@ADT was more prone to release ADT-OH under acidic conditions, which is a key prerequisite for H₂S-Augmented MHT (Fig. 1G). To verify the degradability of CmMN@ADT under acidic and heated conditions and its potential for releasing H₂S in the acidic tumor microenvironment (TME), we measured the UV-visible spectra of the nanoparticles after incubation with methylene blue under different conditions. As shown in Fig. S3, CmMN@ADT effectively releases H₂S under acidic and thermal conditions, providing crucial evidence for subsequent experiments. Meanwhile, we also assessed the POD enzyme activity of the nanomaterials. As shown in Fig. S4, the nanocomposite platform exhibited excellent POD enzyme activity at temperatures suitable for mild hyperthermia.

We then evaluated the stability of CmMN@ADT by determining the polydispersity index (PDI) in double-distilled water (ddH₂O), phosphate-buffered saline (PBS), and Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS). The results demonstrated that CmMN@ADT maintained a relatively stable PDI value in both FBS-supplemented medium and PBS over a 7 day observation period (Fig. S5A), indicating favorable stability of the material under physiological conditions. Furthermore, we characterized the stability of the material following incubation in PBS at different pH values. As shown in Fig. S5B, the magnetic MOF retained excellent stability even after partial degradation of its surface structure under acidic conditions, which constitutes a critical prerequisite for its subsequent magnetothermal therapeutic application. Moreover, biosafety is the essential prerequisite for the effective functioning of CmMN@ADT. The hemolysis test further confirmed that no obvious hemolysis occurred even at high concentrations, demonstrating its excellent biosafety (Fig. 1H). Subsequently, the magnetic hyperthermia efficiency of CmMN@ADT in PBS solution was evaluated. Different concentrations (0, 0.1, 0.5, 1 mg/mL) of CmMN@ADT nanoparticles were dispersed in ddH₂O and tested under an alternating magnetic field (AMF) of 15 kA m⁻¹. As shown in Fig. 1I and J, compared with the control group, the experimental groups containing CmMN@ADT nanoparticles (NPs) exhibited significant temperature increases, with the 1 mg/mL group reaching 73 °C within 10 min, confirming the efficient magnetic hyperthermia effect of CmMN@ADT. Subsequently, we investigated the magnetothermal conversion efficiency of the magnetic MOFs treated with PBS at different pH values. As shown in Fig. S6, the nanomaterials exhibited enhanced magnetothermal heating performance after acidic treatment. This result confirms that the nanomaterials can still exert an effective magnetothermal therapeutic effect following the release of ADT-OH and ROS under acidic conditions. Furthermore, CmMN@ADT maintained excellent heating properties after five AMF on/off cycles, indicating stable magnetocaloric conversion performance (Fig. 1K). In summary, the above results demonstrate the successful construction of CmMN@ADT as a targeted magnetocaloric immunomodulator, which exhibits outstanding magnetocaloric properties and biocompatibility.

2.2. Cellular uptake and anti-tumor effects of CmMN@ADT In vitro

Based on the excellent physicochemical properties of CmMN@ADT, its cellular uptake efficiency, biosafety profile, and antitumor efficacy were further evaluated in this study. The effective uptake of nanomaterials by tumor cells is a critical prerequisite for the antitumor effect of CmMN@ADT. To monitor cellular internalization, we co-incubated Hepa1-6 tumor cells with Cy5.5-labeled CmMN@ADT and found that Cy5.5-CmMN@ADT could be efficiently internalized by tumor cells (Fig. 2A). Furthermore, upon incubation at various time points, the cells exhibited fluorescent signals of increasing intensity by the flow cytometry assay (Fig. S7), suggesting that the internalization process of CmMN@ADT in Hepa1-6 cells is significantly time-dependent. Meanwhile, given the specific recognition of Cm, CmMN@ADT did not exhibit significant cytotoxicity in normal hepatocytes (Thle-2) (Fig. S8), which further confirms its excellent biocompatibility and tumor targeting selectivity.

Fig. 2.

Evaluation of in vitro therapeutic effects. (A) Endocytosis of Hepa1-6 cells by Cy5.5-CmMN@ADT confocal images. (B) Hepa1-6 cell viability after co-incubation with different concentrations of CmMN@ADT. (C) CmMN@ADT induced cancer cell death by chemodynamic therapy. (D) Annexin V/PI staining and (E) Calcein AM/PI staining evaluated the effect of in vitro treatment were used. (F) LPO staining images of cells after different treatments. (G) Flow cytometry detection of H₂S released by Wsp-5 probe, (H) statistical analysis. Intracellular glucose content (I), lactate concentration (J) and H₂O₂ concentration (K) under different treatment conditions. (L) Detection of released ROS by flow cytometry with DCFH-DA probe, (M) statistical analysis. (N) Intracellular GSH levels after different treatments. n = 3, one-way ANOVA with Tukey’s test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

The antitumor efficacy of CmMN@ADT arises from the synergistic interplay of its multiple mechanisms (Fig. 2C). To evaluate the influence of the magneto-thermal effect to tumor cell responses following CmMN@ADT treatment, the viability of Hepa1-6 cells was detected using the CCK-8 assay. The results showed that CmMN@ADT combined with MHT exhibited a more significant cytotoxic effect compared to CmMN@ADT monotherapy (Fig. 2B). Subsequently, flow cytometry (FCM) was performed to analyse cell apoptosis under different treatment conditions. As shown in Fig. 2D, the CmMN@ADT combined with AMF treatment group induced the most pronounced tumor cell death. The results of Calcein-AM/PI double staining assay also showed a consistent trend (Fig. 2E). Additionally, relative to the control group, the CmMN@ADT + AMF treatment group exhibited significantly downregulated expression of the antiapoptotic B-cell lymphoma 2 (Bcl-2) protein, accompanied by marked upregulation of the proapoptotic Bcl-2-associated X protein (Bax) (Fig. S9). Meanwhile, we further detected the activation status of caspase-3, a classical apoptosis marker. The results showed that the expression of cleaved Caspase-3 was significantly elevated in CmMN@ADT + AMF treated Hepa1-6 cells, and its ratio to total Caspase-3 was also significantly increased (Fig. S9). The above results further confirmed the antitumor efficacy of CmMN@ADT in the synergistic treatment with MHT and ferroptosis.

Additionally, it has been reported that H₂S promotes glucose uptake and metabolism in tumor cells, leading to an increase in lactate production, which in turn induces intracellular acidification and creates favorable microenvironmental conditions for chemodynamic therapy, thereby facilitating the generation of large quantities of reactive oxygen species (ROS) that induces oxidative stress death in tumor cells [36,37]. Therefore, we further evaluated the capacity of CmMN@ADT to release H₂S at the tumor sites. Intracellular H₂S levels were detected by loading Hepa1-6 cells with Wsp-1, a green fluorescent hydrogen sulfide probe. FCM analysis revealed that CmMN@ADT was able to release H₂S efficiently (Fig. 2G–H, and S10). Notably, H₂S release mediated by CmMN@ADT was further enhanced following application of an AMF. Consistent with H₂S levels, the CmMN@ADT + AMF treatment group significantly promoted glucose consumption (Fig. 2I) and increased intracellular lactate content (Fig. 2J). Additionally, this treatment group resulted in a significant increase in intracellular H₂O₂ levels (Fig. 2K). Given that CmMN@ADT elevated intracellular H₂O₂ concentration and induced acidosis, we further detected intracellular total ROS levels using DCFH-DA probe. As shown in Fig. 2L–M, and S10, the CmMN@ADT + AMF treated group exhibited the strongest DCF green fluorescent signal, indicating the most significant ROS accumulation. Meanwhile, Fe³⁺ generated by the MN through the fenton reaction could further deplete intracellular glutathione (GSH) (Fig. 2N) and disrupt cellular redox homeostasis, thus positively promoting the accumulation of ROS [38,39]. Notably, as GSH levels decrease under CmMN@ADT treatment, GSH is oxidized to glutathione disulfide (GSSG), subsequently inhibiting GPX4 and SLC7A11 expression. This process is clearly validated by Western blot analysis in Fig. S11. High levels of ROS can trigger lipid peroxidation (LPO) of cell membranes, which ultimately triggers ferroptosis [40,41]. Thus, we utilized C11-BODIPY 581/591 probe to assess the degree of lipid peroxidation in Hepa1-6 cells under different treatment conditions. As shown in Fig. 2F, the green fluorescence signal of lipid peroxidation was significantly enhanced in the CmMN@ADT + AMF group. The MDA content in each group of cells also showed the same trend (Fig. S12), suggesting marked activation of ferroptosis.

In addition, it has been reported that excessive ROS production can disrupt mitochondrial respiratory chain complex IV (COX IV), impairing mitochondrial structure and function, and ultimately decreasing intracellular ATP levels [42]. Biotem imaging revealed abnormal mitochondrial morphology in Hepa1-6 cancer cells after CmMN@ADT + AMF treatment (Fig. S13A). FCM analysis further demonstrated that CmMN@ADT + AMF significantly decreased mitochondrial membrane potential (Fig. S13B–C) and reduced intracellular ATP content (Fig. S14). Collectively, these findings demonstrate that CmMN@ADT can generate substantial amount of ROS, trigger oxidative stress in tumor cells, and effectively induce ferroptosis, thereby significantly inhibiting tumor growth.

2.3. Immunomodulation and thermal Re-sensitization of CmMN@ADT

More importantly, effective immune activation is critical for achieving long-term antitumor effects. MHT has been reported to induce immunogenic cell death (ICD), thereby activating dendritic cells (DCs) and subsequently triggering CD8⁺ T cell-mediated systemic immune responses [43,44]. Additonally, the tumor immunosuppressive microenvironment is a key inducement for tumors to escape immune surveillance. This immunosuppressive state not only involves T cell exhaustion and immune checkpoint overexpression, but also includes the accumulation of immunosuppressive cells such as regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs). Given these findings and the demonstrated cytotoxicity, we further evaluated the immune microenvironment reprogramming capacity of CmMN@ADT. As shown in Fig. 3A, the levels of high mobility group box 1 protein (HMGB1) in Hepa1-6 cells treated with CmMN@ADT + AMF exhibited significantly lower levels of HMGB1 but markedly higher levels of calreticulin (CRT) and secreted ATP compared to other treatment groups, which was consistent with the quantitative results (Fig. 3B–Fig. S15). These damage-associated molecular patterns (DAMPs) “heat up” TME and further activate the antitumor immune response [[45], [46], [47]]. As shown in Fig. 3C and D, CmMN@ADT + AMF treated Hepa1-6 cancer cells effectively promoted DCs maturation (45.1 ± 2.1%) compared to the PBS group (12.4 ± 0.2%). Notably, we used flow cytometry to detect the expression levels of major histocompatibility complex class I (MHC-I) in each group of DCs. MHC-I, as an antigen presenting molecule, was upregulated on mature DCs to stimulate antigen-specific T cells [48]. The CmMN@ADT + AMF group showed the highest level of MHC-I expression compared to other groups (p < 0.05) (Fig. S16). Moreover, the levels of antitumor cytokines TNF-α and IL-6 were significantly elevated in the CmMN@ADT + AMF group, reaching 3.22-fold and 2.55-fold higher than those in the PBS group, respectively (Fig. 3E and F), which aligned with our ICD findings.

Fig. 3.

Evaluation of CmMN@ADT mediated immune response and NF-κB inhibition. (A) CRT expression and intracellular HMGB1 levels in Hepa1-6 cancer cells under different conditions, and the level of released ATP (B). Proportion of CD80⁺ and CD86⁺ dendritic cells under different stimulation conditions (C,D), and concentrations of TNF-α (E) and IL-6 (F) in supernatants after co-culture. (G) Western blot analysis of metastasis-associated proteins in Hepa1-6 cells after different stimuli. (H) Wound-healing, migration and invasion assays and quantitative analysis (I, J) of Hepa1-6 cancer cells under different treatment conditions. (K) Schematic representation of the therapeutic effect of CmMN@ADT on cancer cells. n = 3, one-way ANOVA with Tukey’s test, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Although CmMN@ADT + AMF has demonstrated substantial advantages in antitumor therapy and significantly induced apoptosis in tumor cells. However, with prolonged treatment, the local hyperthermia induced by magnetothermal therapy may drive the development of “lose-sensitivity” in tumor cells, resulting in treatment failure. This “lose-sensitivity” manifests as reduced tumor cell sensitivity to oxidative stress and potential upregulation of NF-κB expression—a key mediator that enhances tumor resistance to therapy, ultimately facilitating tumor cell survival and metastasis [[49], [50], [51]]. Heat shock proteins are ATP-dependent molecular chaperones a that are major defence molecules against heat stress [52]. It has been shown that H₂S, a gaseous signaling molecule, inhibits the expression of HSP70 and blocks the NF-κB signaling pathway, thereby breaking tumor adaptive resistance and generating antitumor activity [17]. Ultimately, this process in tumor cells regaining sensitivity and once again responding to MHT treatment. Therefore, we further evaluated the ability of CmMN@ADT re-sensitization in MHT and inhibit tumor metastasis. As shown in Fig. S17, the expression of COX IV and HSP70 proteins was effectively downregulated by CmMN@ADT + AMF treatment compared to the MN + AMF treatment group without H2S production. In addition, during AMF treatment, tumor cells suffered lethal injury, which in turn activated the NF-κB signaling pathway in order to generate a resistance response to treatment. Consequently, the NF-κB signaling pathway activity in Hepa1-6 cells was significantly enhanced in the MN + AMF-treated group (Fig. 3G). Notably, due to the inhibitory effect of H₂S on the NF-κB pathway, the activation of this signaling pathway was significantly inhibited in the CmMN@ADT + AMF treated group than in the MN + AMF treated group. As an important signaling pathway mediating tumor migration capacity, differences in NF-κB pathway activity affected metastasis related protein expression [53]. As shown in Fig. 3G–Fig. S18, the expression of both matrix metalloproteinase (MMP-2 and MMP-9) was significantly reduced in the CmMN@ADT + AMF group. The results of the wound-healing assay and transwell assay showed that the migration ability of cancer cells was significantly reduced after treatment with CmMN@ADT + AMF (Fig. 3H–J). Meanwhile, E-Ca expression was elevated while N-Ca and Snail expression was decreased in Hepa1-6 cells after CmMN@ADT + AMF treatment (Fig. S19), indicating that the Epithelial-Mesenchymal Transition (EMT) process was significantly inhibited. Cumulatively, these findings demonstrate that CmMN@ADT mediated combinatorial magneto-thermal therapy and chemodynamic therapy exerts multiple antitumor effects in vitro not only directly kills tumor cells, but also promotes cellular immune response by inducing ICD effect, which in turn activates systemic antitumor immune response. In addition, this therapeutic strategy can effectively inhibit the adaptive resistance of the tumor to the magnetothermal effect and downregulate the activity of the HSP70 and NF-κB signaling pathways, thus significantly inhibiting the invasion and migration ability of the tumor (Fig. 3K).

2.4. Therapeutic effect of CmMN@ADT in vivo

Inspired by the remarkable and long-term antitumor effects in in vitro experiments, we further evaluated the in vivo therapeutic efficacy of CmMN@ADT in a hepatocellular carcinoma (HCC) mice model. Effective accumulation of nanomedicines at the tumor site is a critical prerequisite for exerting their antitumor activity. Therefore, we assessed the distribution of CmMN@ADT in vivo by the Hepa1-6 tumor-bearing mice model. As shown in Fig. S20A, gradual accumulation of CmMN@ADT at the tumor site was detected at 4 h after injection of CmMN@ADT into the tail vein of C57BL/6J mice. Ex vivo organ imaging results further confirmed that CmMN@ADT was predominantly localized in tumor tissues (Fig. S20B and C), indicating favorable tumor-targeting capability and retention properties. Furthermore, we quantified the elemental content in mouse tissues at different time points using inductively coupled plasma-mass spectrometry (ICP-MS) technology. As shown in Fig. S21, CmMN@ADT continuously entered and accumulated in tumors within 24 h post-administration. Meanwhile, the retention levels of CmMN@ADT in key organs such as the heart, spleen, and lungs gradually decreased over time. This result indicates that CmMN@ADT can be efficiently cleared through in vivo metabolic pathways, with almost no non-specific accumulation in vital organs. To further evaluate the magneto-thermal conversion performance of the magnetic component in vivo, we applied AMF to tumor-bearing mice 24 h after injection of CmMN@ADT. As shown in Fig. S22, under AMF irradiation of 15 kA⋅m-¹, CmMN@ADT could rapidly increase the local temperature from 25 °C to 45 °C within 10 min, demonstrating that it still possesses an excellent magneto-thermal conversion efficiency in vivo.

Subsequently, when the subcutaneous xenograft tumor volume reached approximately 100 mm³, mice in each group were injected with different nanoparticles (NPs) in the tail vein and received AMF treatment (Fig. 4A). As shown in Fig. 4B and Fig. S23A, CmMN@ADT alone only slightly inhibited the growth of Hepa1-6 xenograft tumors compared to the PBS group and the single AMF-treated group, whereas CmMN@ADT combined with AMF treatment significantly enhanced the antitumor effect. Survival analysis showed a consistent trend: the tumor progressed rapidly in the PBS group and the AMF alone group, and all mice died within 38 days, whereas the CmMN@ADT + AMF treatment group not only significantly inhibited the tumor growth, but also significantly enhanced the survival rate (Fig. 4C). To further evaluate the therapeutic effect, the tumor tissues were stripped and weighed at the end of the experiment. As shown in Fig. S23B–C, the CmMN@ADT + AMF group exhibited the smallest tumor volume and the lightest average tumor weight. Meanwhile, tumor tissues were stained with hematoxylin-eosin (H&E) staining and Ki67 immunofluorescence staining to carry out pathological analysis. As shown in Fig. S23D, areas of mild to moderate necrosis were observed in the CmMN@ADT and MN + AMF groups, whereas extensive and severe tissue necrosis with the lowest number of proliferative tumor cells was observed in the CmMN@ADT + AMF group. In summary, consistent with the results of in vitro experiments, CmMN@ADT combined with AMF effectively inhibited the proliferation and promoted the apoptosis of Hepa1-6 tumors in vivo.

Fig. 4.

Immune response of CmMN@ADT in Hepa1-6 tumor mice. (A) Schematic diagram of Hepa1-6 tumor inoculation and different treatments. Tumor growth curves (B) and survival analysis (C) of Hepa1-6 tumor mice (n = 6) after different treatments. Western blot analysis of tumor tissue after different treatments (D), IL-6 (E), and TNF-α (F) levels. Representative profiles (left) and relative frequencies (right) showing the expression of CD80⁺CD86⁺ DCs (G) and CD8⁺ T (H) cells in tumor-draining lymph nodes after different treatments. Representative profiles (left) and relative frequencies (right) showing the expression of Gra B⁺CD8⁺ (I) and IFNγ⁺CD8⁺ (J) T cells in mice treated with different therapies. Representative profiles (left) and relative frequencies (right) showing Foxp3⁺CD25⁺ (K) and Gr-1⁺CD11b⁺ (L) cells infiltrating tumors after different treatments. n = 3, one-way ANOVA with Tukey’s test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

2.5. Immune response to CmMN@ADT in vivo

As an important strategy for combined tumor therapy, MHT not only directly eliminates tumor cells via thermal effects, but also disrupts the TIME by inducing robust ICD, thus activating effective antitumor immune responses [54]. To further elucidate the antitumor mechanism of CmMN@ADT, we assessed the ICD effect and the alterations in relevant immune cell subsets within tumor tissues across different treatment groups. As shown in Fig. 4D, the results showed that the expression of CRT was significantly upregulated and the extracellular release of HMGB1 was markedly increased in the CmMN@ADT + AMF treated group. These findings indicate that CmMN@ADT mediated magneto-thermal therapy synergistically enhanced the ICD effect in the tumor cells with chemodynamic therapy. Serum enzyme-linked immunosorbent assay (ELISA) results further demonstrated that CmMN@ADT + AMF treatment elevated the serum levels of antitumor cytokines interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) to 4.04-fold and 2.21-fold those of the control group, respectively (Fig. 4E and F). Meanwhile, this treatment significantly promoted the maturation and differentiation of DCs (Fig. 4G), which facilitated the initiation of T cell-mediated adaptive immune responses. Flow cytometric analysis of tumor-infiltrating lymphocytes (TILs) revealed that CmMN@ADT + AMF effectively induced the expansion of effector T cells. Specifically, the proportion of CD8⁺ T cells reached (38.4 ± 2.5%), which was significantly higher than the other groups (13.6% ∼ 36.9%) (Fig. 4H). Consistent with this, the CmMN@ADT + AMF group exhibited the highest proportion of Granzyme B⁺ CD8⁺ T cells (32.1 ± 0.4%, Fig. 4I) and peak levels of IFN-γ⁺ CD8⁺ T cells (21.1 ± 0.4%, Fig. 4J), confirming the functional activation of CD8⁺ T cells. In addition, FCM analysis of tumor tissue also revealed that CmMN@ADT + AMF treatment significantly reduced the proportion of immunosuppressive cell populations, including regulatory T cells (Tregs, 6.89 ± 0.2%, Fig. 4K) and myeloid-derived suppressor cells (MDSCs, 4.23 ± 0.7%, Fig. 4L). Collectively, these results suggest that CmMN@ADT + AMF not only effectively induces antitumor immune responses, but also significantly alleviates the immunosuppressive state in the tumor microenvironment, thereby strongly inhibiting tumor proliferation.

2.6. Long-term tumor-suppressive effects of CmMN@ADT

To further evaluate the inhibitory effect of this magnetic nanoplatform on the growth of HCC, we established an orthotopic liver tumor model (Fig. 5A). Hepa1-6-Luc tumor cells were implanted into the liver tissues of C57BL/6J mice, and treatment was initiated one week after stable tumor growth. Tumor progression was monitored via in vivo bioluminescence imaging along with survival analysis. As shown in Fig. 5B and C, the CmMN@ADT + AMF group exhibited the most pronounced inhibition of tumor growth, with the longest survival duration, which confirmed the robust antitumor activity of CmMN@ADT in the orthotopic HCC model. To further assess the efficacy of CmMN@ADT + AMF in inhibiting tumor metastasis and preventing distant dissemination, we established a lung metastasis model via tail vein injection of Hepa1-6 cells (Fig. 5D) and compared the inhibitory effects of different treatment groups on lung metastatic nodules. As shown in Fig. 5E and F, multiple larger lung metastatic nodules were seen in the PBS group and the AMF alone group. In contrast, the number of lung metastatic nodules in both the CmMN@ADT group and the MN + AMF group decreased. Notably, the CmMN@ADT + AMF group exhibited minimal metastasis, indicating that this combination therapy was effective in suppresses tumor metastasis. The results of survival analysis also showed a consistent trend, with the CmMN@ADT + AMF treated group of mice having the longest median survival (Fig. 5G). To elucidate the potential mechanisms underlying long-term antitumor immunity, we further analyzed the proportions of effector memory T cells (Tems, CD62L⁻CD44⁺) and central memory T cells (Tcms, CD62L⁺CD44⁺) ratios to assess their ability to induce immune memory for long-term control of distant metastases. The results showed that CmMN@ADT + AMF treatment significantly increased the ratios of both Tcms and Tems (Fig. 5H and I). Additionally, CmMN@ADT + AMF treatment significantly increased the proportion of CD3⁺CD8⁺ T cells in the spleen (Fig. 5J and K) and promoted the proliferative activity of CD8⁺ T cells (Fig. 5L and M). Collectively, these findings demonstrate that CmMN@ADT + AMF not only effectively inhibits orthotopic tumor growth and distant metastasis, but also elicits a robust immune memory response. This multi-faceted therapeutic efficacy highlights the considerable translational potential of CmMN@ADT for improving survival outcomes in HCC models.

Fig. 5.

Long-term antitumor effects of CmMN@ADT. (A) Schematic diagram of the construction and treatment of an orthotopic tumor model. Bioluminescence imaging (B) and survival rate (C) analysis of mice after orthotopic implantation of Luc-Hepa1-6 cells in the liver. (D) Schematic diagram of lung tumor metastasis induced by intravenous injection of Hepa1-6 cancer cells and different treatment regimens. (E) Representative photographs of lung tissue from mice treated with different regimens (upper) and H&E staining (lower). Analysis of the number of metastatic nodules in lung tissue from mice treated with different regimens (F) and overall survival rate (G). Flow cytometry analysis of the frequency of memory CD8⁺ T cells in the spleen (H, I) and the infiltration of CD8⁺ T cells (J, K) and Ki67⁺ CD8⁺ T cells (L, M) in tumor tissue. n = 3, one-way ANOVA with Tukey’s test, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

2.7. Biosafety assessment

Finally, we performed a systematic in vivo biocompatibility assessment of the tumor-bearing mice across all treatment groups. H&E staining pathology of major organs, including heart, liver, spleen, lungs and kidneys, revealed no obvious tissue damage or abnormal pathological changes in any treatment group compared to the control group (Fig. S24). In addition, all mice showed stable weight gain throughout the treatment period, with no significant fluctuations observed (Fig. S25A). To further assess systemic safety, blood samples were collected from the retro-orbital venous plexus of mice in the CmMN@ADT + AMF group and control group for routine hematological analysis (Fig. S25B) and blood biochemical testing (Fig. S25C). The results showed that the hematological parameters such as leukocytes, erythrocytes and platelets, as well as biochemical indices related to liver and kidney functions were within the normal reference range without significant differences between the two groups. Collectively, these results indicate that CmMN@ADT exhibits good biocompatibility and safety in vivo.

3. Conclusions

In summary, our study successfully developed a multifunctional biomimetic magnetic nanoplatform, CmMN@ADT, which integrates H₂S-Augmented MHT and ferroptosis, reverse the TIME, ultimately effectively and persistently inhibit tumor growth. Specifically, by combining the AMF-responsive heating of Fe₃O₄ with the controlled release of H₂S from ADT-OH via a degradable MOF structure, the platform induces potent tumor cell ablation while simultaneously overcoming thermal resistance through mechanisms including HSP downregulation and NF-κB pathway inhibition. Furthermore, tumor cell membrane coating enhances targeted delivery, improving therapeutic specificity. Both in vitro and in vivo studies confirmed that, upon exposure to an AMF, CmMN@ADT disrupts the chaotic equilibrium of the TME, enabling synergistic H₂S Augmented MHT and ferroptosis, while concurrently enhancing ICD, facilitating CD8⁺ T cell infiltration, and reversing the immunosuppressive TIME, highlighting its significant potential for sustained tumor suppression. Compared to privious studies, our research not only improves the therapeutic efficacy of ferroptosis through the use of H₂S but also employs H₂S to overcome tumor adaptive resistance during MHT, thereby offering a novel paradigm for tumor treatment.

4. Experimental section

Materials: Anethole trithione was sourced from Shanghai Bide Pharmatech Co., Ltd. Wsp-1 probe, DCFH-DA, C11-BODIPY, Membrane and Cytoplasmic Protein Extraction Kit, JC-1 Assay Kit, Calcein/PI Live/Dead Assay Kit, ATP Assay Kit, and GSH Content Assay Kit were all obtained from Beyotime Biotechnology Co., Ltd. (Shanghai). Annexin V-FITC Apoptosis Detection Kit was acquired from Shanghai Bestbio Biotechnology Co., Ltd. A Cell Counting Kit-8 (CCK-8) was obtained from Nanjing NuoWei Biotechnology Co., Ltd. A H₂O₂ Content Assay Kit was obtained from Sangon Biotech (Shanghai). A L-Lactic Acid Assay Kit was purchased from Beijing Solarbio Technology Co., Ltd. A Cellular Ferrous Colorimetric Assay Kit was purchased from Elabscience Biotechnology Co., Ltd. (Wuhan). All reagents were used directly as received without further purification.

Cancer cell membrane (Cm) Collection: Cm were obtained according to the membrane and cytoplasmic protein extraction kit instructions. Briefly, Hepa 1-6 cells are resuspended with extraction reagent containing PMSF, left in an ice bath for 30 min, followed by cell disruption and centrifugation at 4 °C to obtain Cm.

Preparation of CmMN@ADT: Fe₃O₄ NPs were synthesized by the classical chemical coprecipitation method. Firstly, FeCl₃⋅6H₂O (0.70 mg) and FeCl₂⋅4H₂O (0.35 mg) were dissolved in 20 ml of water at 80 °C under nitrogen protection and stirred vigorously. Then, 15 ml of NH₄OH solution (0.11 g/mL) was added until the pH reached 10-12. The mixture was kept at 80 °C and reacted for 2 h. Subsequently, 6 mL of polyacrylic acid (PAA) solution (0.164 g mL⁻¹) was added to the dark reaction solution and stirred at room temperature for 8 h. The mixture was separated by a magnet to collect the precipitate, which was washed three times with deionized water (25 mL). After centrifugation at 12,000 rpm for 10 min, the supernatant was discarded and the precipitate was collected to obtain PAA-modified Fe₃O₄ NPs (Fe₃O₄-PAA).

Fe₃O₄-PAA NPs (200 mg) were ultrasonically dispersed in 20 mL of FeCl₃ solution (10 mM) and mechanically stirred at 70 °C for 0.5 h. Then, 10 mg of 1,3,5-benzenetricarboxylic acid (H₃BTC) was added and the mixture was mechanically stirred at 70 °C for 6 h. The Fe₃O₄@MIL100 was collected by a magnet.

10 mg of ADT-OH was dissolved in 10 mL of methanol and ultrasonically treated at room temperature for 10 min. Then, 2 mg of Fe₃O₄@MIL-100 was added and the mixture was mechanically stirred at room temperature for 24 h. After stirring, the precipitate was collected by a magnet and washed three times with ethanol to remove unbound ADT-OH, obtaining MN@ADT.

The obtained MN@ADT was sonicated with Cm solution (1 mg ml⁻¹) and then passed sequentially through a membrane with a pore size of 400 and 200 μm to obtain CmMN@ADT. Detection of membrane protein loading in CmMN@ADT using Coomassie brilliant blue method.

Characterization of CMmMN@ADT: The morphology of CmMN@ADT was characterized by transmission electron microscopy (TEM, FEI, Tecnai G2 F30, USA). The particle size distribution and zeta potential of MNs, Cm and CmMN@ADT were determined by dynamic light scattering (DLS) (Nano ZS90, UK). To evaluate the stability of CmMN@ADT, the CmMN@ADT was dispersed in PBS buffer with pH 7.4 and DMEM with 10% FBS, and their particle size and polydispersion index (PDI) at different times were determined by DLS. The absorption spectra of MN, ADT-OH and CmMN@ADT were determined by UV-vis spectrophotometer (UV2600, Shimadzu, Japan). In addition, to study the magneto-thermal conversion efficiency of CmMN@ADT, different concentrations of CmMN@ADT solutions were placed in alternating magnetic field coils (15 kA m⁻¹), and the real-time temperature was recorded using an infrared thermal imaging thermometer (Hikvision, H21).

Evaluation of ADT-OH Releasing from CmMN@ADT: 3 mg of CmMN@ADT was placed in a 10 kDa dialysis bag and immersed in 30 mL of PBS buffer with a pH of 6.5 or 7.4, respectively, and shaken at 37 °C at 300 rpm. Subsequently, 1 mL of liquid was collected at 0, 1, 2, 4, 8, 16, 24, and 48 h (with an equal volume of buffer solution added), and the amount of ADT released from CmMN@ADT was determined using a UV-vis spectrophotometer.

Cell lines and Animals: Hepatocellular carcinoma cell lines (Hepa1-6 cells) and human hepatocytes (Thle-2 cells) were procured from Zhejiang Matson Cell Technology Co., Ltd. (Zhejiang, China) and grown in DMEM media (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin at 37 °C in a 5% CO₂ atmosphere. Four-week-old female C57BL/6j mice were procured from the Animal Experiment Center of Guangxi Medical University and maintained at the same facility, with experiments conducted in accordance with the guidelines established by the Ethics Committee of the Affiliated College of Cancer Medicine of Guangxi Medical University.

Cell Uptake: CLSM and FCM were used to evaluate the uptake CmMN@ADT of cells. After co-incubation with Cy5.5-labeled CmMN@ADT for 0.5, 1, 2, 4, and 6 h, Hepa 1-6 cells were digested, resuspended, PBS washed, and recorded by FCM (cytoflex, Beckman). In addition, Hepa 1-6 cells were seeded in 24-well plates having pre-positioned coverslips, and then replaced with medium containing 10 μg mL⁻¹ free Cy5.5 or CmMN@ADT for 6 h, respectively. The cells of the coverslips were then fixed, permeabilized, Hoechst 33,342 stained, and finally images were acquired with CLSM (Zeiss LSM 980, Germany).

Cell viability assay: Hepa 1-6 cells and Thle-2 cells were plated in 96-well plates at a density of 5 × 10³ cells per well and treated with different concentrations of CmMN@ADT for 24 h. Subsequently, CCK-8 was added and incubated for 2 h, and cell viability was assessed at 450 nm by multimode microplate reader (Synergy H1, USA).

Apoptosis assay: Hepa 1-6 cells were seeded in 12-well plates with 1 × 10⁵ cells per well, and the cells were treated with PBS, MN, and CmMN@ADT for 24 h. Cells were collected, lightly washed 3 times with PBS and resuspended, followed by staining with Annexin V-FITC/PI for 15 min. Finally, apoptotic cells were analyzed by FCM (cytoflex, Beckman).

ROS and H₂S Measurement: Hepa 1-6 cells were cultured in 12-well plates (1 × 10⁵ per well) for 12 h, after which the media was substituted with one containing various drugs and incubated for an additional 6 h. Following the loading of the DCFH-DA and WSP-1 probe (Beyotime, Shanghai) in accordance with the provided instructions, and incubation in darkness or AMF for 30 min, the cells were collected for the study of ROS and H₂S levels via flow cytometry (Cytoflex, Beckman).

Calcein-AM/PI staining: Hepa 1-6 cells were cultivated in 12-well plates (1 × 10⁵ per well). Following 24 h of treatment with various drugs, Calcein-AM and propidium iodide (PI) were added for staining. Incubate for 30 min in the dark and acquire images by fluorescence microscopy.

GSH Measurement: Hepa 1-6 cells were cultivated in 6-well plates (4 × 10⁵ per well) for 12 h. The cells were incubated with various drugs for 6 h and subsequently exposed to an alternating magnetic field for 10 min. The cells were collected and lysed. Finally, the intracellular GSH level was measured according to the kit instructions (Beyotime, Shanghai).

H₂O₂ measurement: Utilize the kit to assess the amount of H₂O₂ in cells in accordance with the manufacturer’s guidelines. Briefly, Hepa 1-6 cells were inoculated at 4 × 10⁵ cells per well and cultured overnight in 6-well plates. After the different treatments, the cells were collected for disruption and centrifugation to isolate the supernatant. Finally, the characteristic absorbance value of the titanium peroxide composite was determined at 415 nm.

C11-BODIPY Staining: C11-BODIPY staining was conducted in accordance with the manufacturer’s guidelines. Briefly, inoculated 5 × 10⁴ Hepa 1-6 cells into a 24-well plate with pre-placed coverslips and cultured for 12 h. After different treatments, cells were washed 3 times with PBS, stained with C11-BODIPY working solution for 30 min in the darkness, and then stained with Hoechst 33,342. Ultimately, images were acquired using the CLSM.

Western blot: Hepa 1-6 Cells were lysed using RIPA lysate buffer containing phosphatase and protease inhibitors, and then the protein concentration of the samples was standardized. The proteins are then isolated using gel electrophoresis and deposited to PVDF membranes. After blocking with 5% BSA for 30 min, the primary antibody was incubated overnight at 4 °C in accordance with the manufacturer’s guidelines, followed by incubation with the fluorescent secondary antibody for 1 h at room temperature. Finally, images were acquired by a gel imaging system and the grayscale of the bands was counted with Image J.

ATP and Lactate Assays: Measured the levels of ATP and lactate in the intracellular and supernatant in accordance with the manufacturer’s guidelines. Briefly, after different treatments of cells in 6-well plates, the cells and supernatant were collected, followed by lysis and filtration, respectively. Finally, the RLU value corresponding to the ATP content and the lactate reactant value at 540 nm was measured by multimode microplate reader.

Mitochondrial Status Assessment: Hepa 1-6 cells were seeded into confocal dishes and cultured overnight. After different treatments, the mitochondrial membrane potential changes were assessed by JC-1 staining using the kit according to the manufacturer’s instructions, and finally images were observed and acquired using CLSM. In addition, after PBS and CmMN@ADT + AMF treatment, Hepa 1-6 cells were collected and fixed in 2.5% glutaraldehyde. Finally, the changes in mitochondrial morphology and ultrastructure were observed by electron microscopy (JEM-1200EX).

Immunofluorescence: Hepa 1-6 cells were seeded in 24-well plates with pre-positioned coverslips overnight for various treatments. Cells were fixed using 4% paraformaldehyde, subjected to 0.1% Triton X-100 treatment for 15 min, and subsequently blocked with 5% BSA for 30 min. Then, primary antibodies were added and incubated overnight at 4 °C, followed by 1 h with fluorescein-conjugated secondary antibodies, and Hoechst 33,342 stained for 15 min. Finally, images were observed and acquired by CLSM.

BMDCs extraction and culture: Tibiae and femurs were aseptically harvested from 8 week C57BL/6j mice. Bone marrow cells were flushed from the medullary cavities using sterile PBS through a syringe needle. The isolated cells were cultivated in 12-well plates with RPMI 1640 complete medium supplemented with recombinant murine GM-CSF (20 ng/mL) and IL-4 (10 ng/mL), denoted as Day 0. On days 2 and 4, half of the medium was replaced with fresh cytokine-containing medium. Immature dendritic cells (imDCs) were collected on day 6 for subsequent experiments. To validate the ICD effect prompted by nanodrugs treatment, Hepa1-6 cells were co-cultured with immature DC cells for 24 h, after which the supernatant was harvested for cytokine analysis and DC cells were collected for flow cytometry assessment of their activation.

Wound healing assay: Hepa 1-6 cells were cultivated in 6-well plates (3 × 10⁵ cells) and cultured overnight. A pipette tip of 200 μL to form a straight line with the cells in the well plate and wash off the detached cells with PBS. Different treatments were subsequently performed, and images were observed and taken with an inverted microscope at 0 h and 24 h. The wound healing rate was determined using the formula: wound healing rate = (wound distance at 0 h - wound distance at 24 h)/wound distance at 0 h × 100%.

Cell migration and invasion assay: 200 μL of pre-treated Hepa 1-6 cells (3 × 10⁴ cells) were inoculated into the upper chamber (or containing Matrigel) of Transwell inserts (24-well plates) and then placed in DMEM with 10% FBS for 24 h. The chamber was then fixed with 4% paraformaldehyde for 20 min, and stained with crystal violet dye for 30 min. Observations and imaging were conducted using an inverted microscope. Migration/invasion rate = number of cells in the sample group divided by the number of cells in the control group × 100%.

Animal model establishment: All animal experiments have been approved by the Medical Ethics Committee of Guangxi Medical University Cancer Hospital with an approval number: 2025-E0620. A subcutaneous hepatocellular carcinoma model wasestablished by subcutaneous injection of 100 μL of Hepa 1-6 tumor cell suspension (3 × 10⁶ cells per mouse) into the right lower extremity of mice. For orthotopic HCC models, Luc-Hepa 1-6 cells were injected into the left lobe of the liver of female C57BL/6j mice (1 × 10⁶ cells per mouse), after which the formation and progression of HCC in situ was assessed by bioluminescence signals using an IVIS spectral imaging system (PerkinElmer). To construct a model of lung metastasis after liver cancer treatment, Hepa 1-6 tumor cells were injected subcutaneously into C57BL/6j mice (3 × 10⁶ cells per mouse). After different treatments, Hepa 1-6 tumor cells (1 × 10⁶ cells per mouse) were injected into the tail vein. Body weight of the mouse and tumor volume were measured and documented throughout the treatment. The tumor volume was determined using the formula: V = (L × W²)/2, where L and W represent the length and width of the tumor, respectively.

Therapy effects of CmMN@ADT: The immune response and therapeutic effect of CmMN@ADT have been studied in a subcutaneous tumor-bearing Hepa 1-6 animal model. Different drugs (5 mg kg⁻¹ CmMN@ADT) were injected subcutaneously. After the last vaccination, tumor tissue from mice was extracted for analysis using hematoxylin and eosin (H&E) staining and Ki-67 immunofluorescence staining. At the same time, another batch of mice with the same treatment is monitored for survival analysis. We also collected TDLNs and spleens from various groups of mice for FCM analysis to study the activation, phenotype, and function of immune cells.

Biological safety: C57BL/6j mice bearing Hepa 1-6 tumors were injected with various drugs in the tail vein (n = 3 in each group). At 48 h, the mice were sacrificed and blood was collected for hemolysis test, serum biochemistry and blood count analysis. At the same time, principal organs like the heart, liver, spleen, lungs, and kidneys were harvested for H&E staining to assess organ damage.

Statistical Analysis: One-way ANOVA was used to evaluate the data, which were shown as mean ± standard deviation (SD). The survival curve was analyzed by log-rank test. The statistically significant expressions were as follows: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.001.

CRediT authorship contribution statement

Zelong Jiang: Conceptualization, Data curation, Writing – original draft. Hong Huang: Formal analysis, Investigation, Methodology. Mengqi Zhang: Validation. Feng Lin: Data curation, Methodology. Guchun Qin: Validation. Niqiang Zhou: Investigation. Guanhua Qiu: Data curation. Jie Chen: Supervision. Duo Wang: Conceptualization, Visualization. Yunxi Huang: Validation, Writing – review & editing. Chang Zhao: Supervision, Validation, Writing – original draft, Writing – review & editing.

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.