A dual-ion/immune checkpoint nanoplatform-activated pyroptosis/cGAS-STING signaling transform iMWA into systemic immunotherapy against HCC recurrence

Abstract

Incomplete microwave ablation (iMWA) remains a major challenge in the clinical management of hepatocellular carcinoma (HCC), as residual tumors often foster an immunosuppressive microenvironment and upregulate PD-L1 expression, thereby promoting immune evasion and recurrence. This study systematically elucidates the mechanisms underlying HCC progression and metastasis following iMWA and describes the design of a dual-ion/immune checkpoint nanoplatform (aP@Mn/Ca) based on a metal-organic framework (MOF). Leveraging the excellent biocompatibility and tumor microenvironment-responsive degradability of MOFs, this platform enables efficient co-delivery of Ca²⁺, Mn²⁺, and an anti-PD-L1 (aPD-L1) antibody specifically to residual tumor tissues to overcome post-ablation immune barriers. Specifically, upon tumor accumulation, controlled Ca²⁺ release induces mitochondrial stress and inflammasome activation, triggering gasdermin-mediated pyroptosis and immunogenic cell death (ICD). Concurrently, Mn²⁺ activates innate immunity by potentiating the cGAS-STING signaling pathway. This synergistic induction of pyroptosis and STING activation promotes inflammatory cytokine production, antigen presentation, and cytotoxic T cell priming. Meanwhile, the local delivery of aPD-L1 mitigates adaptive immunosuppression and reduces systemic toxicity. Consequently, aP@Mn/Ca transforms iMWA into a systemic immunotherapeutic strategy, effectively suppressing HCC recurrence and metastasis both in vitro and in vivo. In summary, this work establishes an ion-based immunomodulatory paradigm integrating ICD, innate immune activation, and immune checkpoint blockade, offering a promising nanotherapeutic framework for preventing post-ablation HCC recurrence and enhancing immunotherapy in solid tumors.

Graphical abstract

Highlights

• •IMWA-induced PD-L1 upregulation and immune evasion in HCC are elucidated.
• •A TME-responsive MOF nanoplatform is designed for Mn/Ca and aPD-L1 co-delivery.
• •Ca²⁺-pyroptosis and Mn²⁺-STING activation synergistically trigger immune response.
• •The platform enhances antigen presentation and cytotoxic T cell infiltration.
• •Recurrence and metastasis are inhibited by remodeling the post-ablation TME.

1. Introduction

Microwave ablation (MWA) has established itself as a first-line, minimally invasive approach for treating hepatocellular carcinoma (HCC) [1,2]. However, due to the complexity and heterogeneity of tumors, precise control of ablation parameters is often hindered, thereby posing a risk of incomplete microwave ablation (iMWA), which fails to achieve complete tumor clearance [3,4]. Mechanistic studies reveal that iMWA can trigger a detrimental biological cascade, exacerbating immunosuppression and enabling immune evasion by increasing myeloid-derived suppressor cells (MDSCs), regulatory T cells (Tregs), and cytokine/chemokine levels in residual tumors, thereby accelerating HCC progression and metastasis [[5], [6], [7]]. Notably, iMWA also induces significant upregulation of the immune checkpoint molecule PD-L1 on surviving tumor cells, allowing tumors to escape immune surveillance [8,9]. Therefore, despite its technical success, local and distant tumor recurrence after iMWA remains a major clinical challenge [6,10]. Developing efficient methods to resolve the dilemma in iMWA is promising for improving the prognosis of HCC [11].

Accumulating evidence suggests that immunogenic cell death (ICD) plays a pivotal role in converting local tumor ablation into a systemic antitumor immune response [12,13]. Among various ICD modalities, pyroptosis, a gasdermin-mediated inflammatory form of programmed cell death, has attracted increasing attention due to its potent ability to release danger-associated molecular patterns (DAMPs), pro-inflammatory cytokines, and tumor antigens [14,15]. These signals can robustly activate innate immune sensing pathways, particularly the cyclic GMP-AMP synthase-stimulator of interferon genes (cGAS-STING) axis, thereby promoting dendritic cell maturation, type I interferon production, and cytotoxic T lymphocyte priming [16,17]. Nevertheless, pyroptosis is rarely achieved efficiently in residual HCC lesions after iMWA, limiting the conversion of local ablation into effective systemic immunity [18]. Intracellular ion homeostasis, especially Ca²⁺ and Mn²⁺ signaling, has emerged as a critical regulator of both inflammatory cell death and innate immune activation [[19], [20], [21], [22]]. Ca²⁺ overload is a well-established trigger for mitochondrial dysfunction, inflammasome activation, and gasdermin cleavage, all of which are essential steps in pyroptosis execution [23,24]. Meanwhile, Mn²⁺ has been identified as a potent endogenous enhancer of the cGAS-STING pathway by increasing cGAS sensitivity to cytosolic DNA and stabilizing cGAS-DNA interactions [[25], [26], [27]]. Despite their complementary immunostimulatory roles, the simultaneous and localized delivery of Ca²⁺ and Mn²⁺ to residual tumor tissues remains technically challenging, and their therapeutic potential in the context of post-ablation HCC recurrence is largely unexplored [28]. Metal-organic frameworks (MOFs), a class of hybrid porous materials assembled from metal ions and organic linkers, have gained significant attention in cancer immunotherapy due to their high drug loading capacity, excellent biocompatibility, and stimuli-responsive degradability [29,30]. These unique properties make MOFs particularly attractive as multifunctional delivery platforms capable of co-encapsulating and releasing diverse therapeutic ions and biologics in a controlled manner within the tumor microenvironment [31].

Meanwhile, elevated PD-L1 expression is a significant signal of incomplete ablation following MWA [29,30]. Immune checkpoint blockade, particularly targeting the programmed death ligand 1 (PD-L1), has shown promise in restoring T cell function in HCC. However, the efficacy of anti-PD-L1 (aPD-L1) monotherapy is often limited by poor tumor immunogenicity and insufficient innate immune activation [31]. Clinical and preclinical studies increasingly indicate that checkpoint blockade requires a pre-existing or therapy-induced inflammatory immune contexture to achieve durable responses [32,33]. Moreover, systemic administration of aPD-L1 antibodies is suboptimal due to PD-L1 expression in normal tissues [34]. This leads to reduced drug concentration at the tumor site, diminished efficacy, and potential off-target immune-related adverse events (irAEs) [35]. Therefore, strategies that can synchronize innate immune activation, ICD, and checkpoint inhibition are urgently needed to overcome post-ablation immune resistance and prevent HCC recurrence.

In this study, we first established cellular and animal models to systematically investigate the mechanisms driving HCC progression and metastasis following iMWA. Subsequently, we engineered a dual-ion/immune checkpoint nanoplatform aPD-L1@Mn/Ca-MOF (aP@Mn/Ca) based on a metal-organic framework (MOF) that co-delivers Mn²⁺/Ca²⁺ together with aPD-L1 to synergistically amplify post-iMWA antitumor immunity. Benefiting from the excellent biocompatibility, high drug loading capacity, and tumor microenvironment-responsive degradability of MOFs, this nanoplatform enables efficient and targeted co-delivery of therapeutic ions and antibodies specifically to residual tumor tissues. Specifically, this nanoplatform is rationally designed to accumulate in ablated tumor tissues, where controlled release of Ca²⁺ induces mitochondrial stress and inflammasome-mediated pyroptosis [36], while Mn²⁺ simultaneously potentiates cGAS-STING signaling in both tumor cells [37,38]. The synergy between pyroptosis and Mn²⁺-potentiated cGAS-STING activation results in a robust release of inflammatory factors, which enhances antigen availability and triggers a potent immune response. Concurrent delivery of aPD-L1 alleviates adaptive immune suppression, enabling robust T cell-mediated tumor clearance [39]. By integrating ion-mediated pyroptosis induction, STING pathway activation, and immune checkpoint blockade into a single nanosystem, this strategy effectively converts iMWA from a local cytoreductive treatment into a systemic immunotherapeutic intervention. Ultimately, it substantially augments the anti-tumor efficacy of iMWA both in vitro and in vivo. Collectively, aP@Mn/Ca establishes a previously underexplored ion-based immunomodulatory paradigm for preventing HCC recurrence after ablation and provides a versatile nanotherapeutic framework for coupling innate immune activation with checkpoint blockade in solid tumors (Scheme 1).

Scheme 1.

Schematic illustration of the mechanism by which aP@Mn/Ca reverse iMWA-induced tumor immunosuppression to enhance HCC treatment.

2. Results and discussion

Immunosuppression Drives Local Tumor Progression and Metastasis in Hepatocellular Carcinoma following iMWA: To enhance clinical relevance, we established an iMWA-treated HCC mouse model to evaluate post-treatment changes in local tumor growth and the tumor immune microenvironment (Fig. 1A). The results showed that residual tumors in iMWA-treated mice grew significantly faster than those in untreated controls (Fig. 1B). Moreover, at the designated endpoint, these tumors exhibited a significantly larger volume compared to the control group (Fig. 1C). Besides, the survival curves indicated that tumor-bearing mice in the iMWA group had a lower survival rate than untreated controls. Notably, while the untreated group maintained a high survival rate of 80% at day 30, the rate in the iMWA-treated group plummeted to 20% (Fig. 1D). This phenomenon may be attributed to the iMWA intervention in HCC reprogramming the tumor microenvironment (TME) and induce tolerance to immunotherapy, potentially explaining the high progression and metastasis rates observed clinically.

Fig. 1.

Effects of iMWA on HCC progression and immunity. A) Schematic diagram of the iMWA mechanism used in mice. B) Imaging images of residual tumors after different treatments in C57BL/6J mice. C) Growth curves of tumors after different treatments. D) Survival curves of mice after different treatments. Representative flow cytometric analysis of tumor-infiltrated lymphocytes after gating on E, F) Foxp3+CD4⁺ cells and G, H) Gr1+CD11b + cells. p-value in these figures were calculated by the two-tailed Student's t-test. *p < 0.05, ***p < 0.001. All the data are shown as the means ± SDs (n = 3). I) Histological analysis of tumor slices via immunofluorescence staining for Gr1, Foxp3 (scale bar = 50 μm). J) The relative amounts of PD-L1 exposure detected by flow cytometry. K) Brightfield images from wound healing assays at 0-, 24- and 48 h postheat stress pretreatment (Scale bar: 200 μm). L) Transwell assays assessing migration and invasion capabilities of Hepa1−6 cells postheat stress pretreatment (Scale bar: 100 μm) (n = 3). (G1: Untreaded; G2: iMWA). (P < 0.05 was considered statistically significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 was considered highly significant).

To further confirm our hypothesis of the immuno-editing effects of iMWA on the HCC TME, we collected residual tumor tissues on day 3 post-iMWA and analyzed changes in immunosuppressive cell populations by flow cytometry (FCM). The FCM data confirmed an enhanced infiltration of Tregs, a potent immunosuppressive T-cell subset (Fig. 1E and F), along with an increasing trend in MDSCs compared to untreated tumors (Fig. 1G and H). Besides, multiple recent studies have consistently demonstrated that iMWA treatment led to a decrease in CD4⁺ T cells, CD8⁺ T cells, and NK cells, along with an increase in monocytes and MDSCs [18]. These findings are consistent with our observations of Tregs d MDSCs following iMWA and provides a mechanistic basis for the observed therapy tolerance. Consistent with this, immunofluorescence staining for Gr1 and FOXP3 further confirmed a more immunosuppressive microenvironment in HCC following iMWA (Fig. 1I). Consequently, we assessed the expression of the immune checkpoint molecule programmed death-ligand 1 (PD-L1) by flow cytometry as an indicator of the local immune status. As shown in Fig. 1J, PD-L1 expression in tumor tissue increased by approximately 0.7-fold after iMWA treatment, suggesting the establishment of an immunosuppressive TME. To investigate the underlying mechanism, we modeled the localized thermal effect induced by iMWA under controlled in vitro conditions using Transwell migration and invasion assays. The results demonstrated that heat stress significantly enhanced the migratory and invasive capabilities of Hepa1-6 cells (Fig. 1K, L and Figs. S1 and S2), indicating that thermal stress acts as a potent trigger for activating an invasive tumor cell phenotype. Collectively, these findings suggest that iMWA in HCC may exacerbate the immunosuppressive landscape by upregulating PD-L1—a key driver of tumor progression and metastasis. This shift pushes the TME toward a more destructive state, ultimately leading to altered lymphocyte infiltration and unfavorable immunotherapy outcomes. Thus, our study provides a novel perspective on the mechanisms of immune evasion following iMWA.

2.1. Preparation and characterization of aP@Mn/Ca

To address the existing challenges faced by iMWA, we further developed a MOF-based nanoparticle formulation named aP@Mn/Ca.The Ca-Mn-MOF carrier was synthesized via a one-pot solvothermal method through the self-assembly of Ca/Mn bimetallic centers with carboxylic acid ligands.Subsequently, aPD-L1 antibody was stably conjugated onto the MOF surface via covalent bonds using EDC/NHS chemical coupling based on the carboxyl groups on the MOF surface. Transmission electron microscopy (TEM) imaging revealed the spheroidal morphology of the aP@Mn/Ca NPs (Fig. 2B). Elemental mapping (Fig. 2C) combined with energy-dispersive X-ray spectroscopy (EDS) confirmed the homogeneous distribution of both Ca²⁺ and Mn²⁺ throughout the nanoparticles (Fig. 2F). The chemical states of elements within aP@Mn/Ca were assessed using X-ray photoelectron spectroscopy (XPS). The results indicated that both Ca²⁺ and Mn²⁺ existed primarily in the +2 oxidation state (Fig. 2D and E). Furthermore, dynamic light scattering (DLS) measurements showed no significant changes in the hydrodynamic diameter (Fig. 2G) or zeta potential (Fig. 2H), indicating that drug loading had negligible impact on the nanoparticle size. We next characterized the acid-responsive release behavior of aP@Mn/Ca, a key feature for targeting the tumor microenvironment (Fig. 2I, J, K). The drug loading efficiency of aPD-L1 on the MOF was determined to be approximately 18.7%.(Fig. S3). At pH 5.4, the nanoparticles demonstrated rapid dissolution, releasing ∼83% of Ca²⁺, ∼81% of Mn²⁺, and ∼70% of aPD-L1 within 24 h—a release rate roughly three times greater than that observed at pH 7.4. This accelerated release under acidic conditions is attributed to the degradation of the calcium phosphate matrix and the detachment of the chelated payloads. Conversely, at pH 7.4, the sub-25% release of all components over 24 h confirms excellent colloidal stability and minimal off-target leakage, highlighting the formulation's potential for site-specific delivery. The pH-responsive release profile of aP@Mn/Ca is critical for its tumor-selective therapeutic function. The rapid release of Ca²⁺, Mn²⁺, and aPD-L1 under acidic conditions ensures that the therapeutic payload is preferentially delivered to the tumor site, while the minimal release at physiological pH (7.4) minimizes off-target toxicity. This TME-responsive release is the initiating event that enables the subsequent Ca²⁺ induced pyroptosis and Mn²⁺ potentiated STING activation.

Fig. 2.

Synthesis and characterization of the bimetallic Mn/Ca nanoagonist aP@Mn/Ca. A) Schematic representation of the synthesis process of aP@Mn/Ca. B) TEM images of the aP@Mn/Ca (Scale bar: 100 nm). C) Elemental mapping of C, O, Ca, Mn of aP@Mn/Ca (Scale bar:100 nm). D)XPS spectra of Ca2p, Mn2p, O1s, C1s of aP@Mn/Ca. E) XPS of aP@Mn/Ca. F) energy-dispersive X-ray spectroscopy (EDS) of aP@Mn/Ca. G) Particle sizes and H) zeta potential of Ca, aP@Ca, Mn/Ca and aP@Mn/Ca. I) Drug release profiles of Ca²⁺, J) Mn²⁺ and K) aPD-L1 at pH 5.4 and 7.4.

2.2. Evaluation of cytotoxicity and pyroptosis induction in vitro

Effective intracellular delivery is crucial for anti-tumor effect of aP@Mn/Ca NPs. Both confocal microscopy (CLSM) and flow cytometry (FCM) analyses confirmed the efficient uptake of DiD-labeled nanoparticles by Hepa1-6 cells within 8 h (Fig. 3A and B), demonstrating successful cellular internalization and thereby providing the basis for their cytotoxic efficacy. In parallel, we evaluated the proliferative and toxic effects of aP@Mn/Ca on HCC cells. At high concentrations, aP@Mn/Ca significantly inhibited the growth of HCC cells (Fig. 3C). While aP@Mn/Ca at elevated concentrations, did not exert overt toxicity on normal hepatocytes (Supporting Information Figure S4). This selective cytotoxicity is attributed to the weakly acidic microenvironment of the HCC and excellent biocompatibility. Notably, HCC cell proliferation was significantly suppressed at concentrations exceeding 200 μg/mL. Numerous studies utilizing calcium carbonate nanoparticles as drug carriers have demonstrated considerable promise and efficacy. This is primarily due to their ability to decompose in the weakly acidic tumor microenvironment (TME), releasing Ca²⁺. This Ca²⁺ release can induce mitochondrial damage via calcium overload and increased reactive oxygen species (ROS) production, thereby activating cytochrome c (CytC) and Caspase-3 [40]. This cascade ultimately leads to gasdermin E (GSDME) cleavage and the induction of pyroptosis [41]. Using a Ca²⁺-sensitive fluorescent probe, significant Ca²⁺ release (green fluorescence) was observed specifically in treatment groups containing the calcium carbonate component (G3-G6), aligning with the acid-responsive release profile. This Ca²⁺ release triggers a cascade that includes ROS generation and mitochondrial calcium overload, resulting in loss of mitochondrial membrane potential. In line with this, CLSM imaging confirmed concurrent mitochondrial Ca²⁺ accumulation, elevated ROS levels, and diminished membrane potential in aP@Mn/Ca-treated HCC cells (Fig. 3D–F and Fig. S5,6). To better mimic the in vivo tumor microenvironment, 3D spheroids were constructed using Hepa1-6 cells. After treatment with the various formulations, cell viability was assessed using calcein-AM/propidium iodide (PI) dual staining. Compared to other groups, the aP@Mn/Ca-treated spheroids exhibited a strong red fluorescence signal (Fig. 3G), indicative of extensive cell death. These results underscores the potent cytotoxicity of aP@Mn/Ca against cancer cells in a more physiologically relevant model. A study described a“dual-gating”strategy whereby ultrasound (US) selectively activates TRPV2 channels overexpressed on tumor cells [19], inducing a massive influx of extracellular Ca²⁺ that leads to intracellular calcium overload. This cascade subsequently triggers mitophagy and apoptosis, culminating in highly precise suppression of tumor growth, which is consistent with our findings. Concurrently, to investigate the impact on tumor cell invasion and migration, Transwell assays were performed. The aP@Mn/Ca group most effectively suppressed the migratory and invasive capacities of tumor cells, an effect likely attributable to the synergistic action of its components (Supplementary Fig. 7). This inhibitory effect on Hepa1-6 cell migration was further confirmed by a wound-healing assay. At both 24 and 48 h post-scratching, the aP@Mn/Ca group exhibited the lowest wound closure rate (Supplementary Fig. 8). Compared to the control group, Bio-TEM images of the aP@Mn/Ca-treated group revealed significant mitochondrial swelling (Fig. 3H). Furthermore, the released Mn²⁺ and Ca²⁺ triggered ROS accumulation and mitochondrial damage, which stimulated the Caspase-3-mediated pyroptotic pathway. This culminated in plasma membrane permeabilization of HCC cells accompanied by the release of pyroptotic vesicles. This process, in turn, enhanced the ICD profile and activated a systemic immune response. Western blot analysis confirmed the upregulation of key proteins in this pathway, including Cytochrome c (CytC), cleaved Caspase-3, and cleaved GSDME (Fig. 3J). To unequivocally establish pyroptosis as the dominant mechanism underlying aP@Mn/Ca-induced cytotoxicity, we utilized the selective Caspase-3 inhibitor Z-DEVD-FMK. As illustrated in Supplementary Fig. S9, pharmacological inhibition of Caspase-3 led to a substantial decrease in GSDME-N expression, which fell to levels closely resembling those observed in untreated controls. This result firmly corroborates that aP@Mn/Ca triggers pyroptosis via the Caspase-3/GSDME signaling cascade. Collectively, these results demonstrate that aP@Mn/Ca effectively suppresses tumor cell growth and viability while simultaneously reprogramming the immunosuppressive microenvironment in iMWA-residual tumors.

Fig. 3.

In vitro assessment of aP@Mn/Ca targeting and activation of oxidative stress. A) Comparison of phagocytosis fluorescence intensity of Hepa1-6 cells in response to DID-labeled aP@Mn/Ca incubated for different durations. B) Flow cytometry expression levels of Hepa1-6 cells incubated with DiD-labeled aP@Mn/Ca for different durations. C) Cell viability of Hepa1-6 cells after treatment with different concentrations of aP@Mn/Ca. D) Fluorescence images of Hepa1-6 cells co-stained with Fluo-4 AM after different treatments (Scale bar: 100 μm). E) CLSM images of ROS production in Hepa1-6 cells after different treatments (Scale bar: 100 μm). F) CLSM images of mitochondrial membrane dysfunction in Hepa1-6 cells assessed using JC-1 (aggregates in red, monomers in green) after different treatments (Scale bar: 50 μm). G) Fluorescence images of 3D Hepa1-6 tumor spheroids co-stained with Calcein AM/PI after various treatments (Scale bar: 500 μm). H) Biological transmission electron microscope (Bio-TEM) of Hepa1-6 cells in Control and aP@Mn/Ca groups. I) Schematic representation of the mechanism by which aP@Mn/Ca induces cell death through the Caspase3-mediated cellular pyroptosis pathway.J) WB analysis of cellular pyroptosis-related proteins. (G1: Control; G2: aPDL1; G3: Ca; G4: aP@Ca; G5: Mn/Ca; G6: aP@Mn/Ca). (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

2.3. Mechanisms of in vitro immune response and remodeling of the immunosuppressive microenvironment

A key strategy to enhance tumor immunogenicity involves inducing ICD, a process that prompts the release of damage-associated molecular patterns (DAMPs) such as calreticulin (CRT), high-mobility group box 1 (HMGB1), and adenosine triphosphate (ATP). The release of these DAMPs effectively promotes tumor antigen presentation and activates anti-tumor immune responses (Fig. 4A). As shown in Fig. 4B and C, fluorescence imaging revealed the most intense green signal in the aP@Mn/Ca group, indicating prominent CRT exposure. Measurement of extracellular ATP levels showed an approximately two-fold increase in the aP@Mn/Ca group compared to the control (Fig. 4D). Furthermore, the translocation of HMGB1 from the nucleus to the extracellular space was evidenced by a marked reduction in nuclear green fluorescence in the aP@Mn/Ca group, a finding corroborated by elevated HMGB1 levels in the cell culture supernatant (Fig. 4E–G). Collectively, these findings demonstrate the robust ICD-inducing capability of aP@Mn/Ca NPs. In addition to inducing pyroptosis, aP@Mn/Ca stimulated innate immunity and enhanced immunomodulatory capacity by activating the cGAS-STING pathway. Western blot analysis confirmed elevated phosphorylation levels of STING and IRF3 in both the Mn/Ca and aP@Mn/Ca groups (Fig. 4H). To distinguish whether the upstream mechanism of STING activation by Mn²⁺ involves direct enhancement of cGAS-DNA binding or indirect secondary activation triggered by ROS-induced DNA damage, we employed NAC and RU.521. NAC is a ROS scavenger, while RU.521 is a cGAS inhibitor that directly suppresses cGAS enzymatic activity and blocks its binding to DNA. As shown in Fig. S10, treatment with the cGAS inhibitor almost completely abolished aP@Mn/Ca-induced STING phosphorylation. In contrast, NAC (5 mM) treatment significantly reduced aP@Mn/Ca-induced ROS production, yet the level of pSTING remained partially elevated compared with untreated cells. These data indicate that even when the ROS-mediated DNA damage pathway is blocked, the majority of STING activation persists. This strongly demonstrates that Mn²⁺ exerts a direct, ROS-independent effect on cGAS-STING activation, whereas the Ca²⁺ dependent ROS burst and subsequent DNA damage constitute a secondary amplification loop that further potentiates the response. This activation can be attributed to the presence of bimetallic Mn/Ca in these formulations, which enhances intracellular ROS accumulation and DAMP release. These signals trigger cGAS activation and its downstream STING signaling, ultimately promoting a systemic immune response. To evaluate the adaptive immune response induced by aP@Mn/Ca, we established an in vitro Transwell co-culture model to study dendritic cell (DC) maturation, in which DC2.4 cells were co-cultured with Hepa1-6 cells subjected to different treatments (Fig. 4J). As shown in Fig. 4K, flow cytometry analysis revealed that the aP@Mn/Ca group exhibited the highest percentage of mature DCs (52.5%), significantly surpassing that of the control group (13.6%), indicating that aP@Mn/Ca acts as an effective stimulator for DC maturation. The upregulation of MHC-II molecules, a key maturation marker that enhances antigen-presenting capacity, was further examined. aP@Mn/Ca treatment significantly increased MHC-II expression on DCs by approximately 26% compared to the control (Fig. 4L). Additionally, flow cytometry using anti-mouse H-2Kb/H-2Db antibodies showed that aP@Mn/Ca also markedly elevated MHC-I expression on DCs (Fig. 4M). In contrast, the Ca group alone had a limited effect on enhancing MHC-I expression. These data collectively indicate that aP@Mn/Ca stimulates DC activation and adaptive immunity through a coordinated mechanism: Ca²⁺ overload induces pyroptosis and DAMP release, while Mn²⁺ overload potentiates DNA sensing by cGAS to activate the cGAS-STING pathway. This dual action upregulates the expression and trafficking of MHC-I, synergistically enhances co-stimulatory molecules like CD80/CD86, and ultimately drives a robust adaptive immune response.

Fig. 4.

In vitro immune response of aP@Mn/Ca. A) Mechanism of aP@Mn/Ca-induced upregulation of ICD and MHC-1 to enhance in vitro immune responses. B) Fluorescence imaging and C) quantitative statistics of CRT exposure on Hepa1-6 cells after different treatments was assessed by inverted fluorescence microscopy. Scale bar, 50 μm. D) ATP from tumor cells after different treatments. E) Fluorescence imaging and F) quantitative statistics of HMGB-1 release from Hepa1-6 cells after different treatments was assessed by inverted fluorescence microscopy. Scale bar, 50 μm. G) HMGB-1 from tumor cells after different treatments. H) The expression of the STING-related proteins in Hepa1-6 cells following different treatments measured by western blot assay. I) Schematic representation of the mechanism by which aP@Mn/Ca induces cell death through the STING-related pathway. J) Schematic diagram of transwell mode constructed for DCs mature evaluation. K) Flow cytometric analysis of CD80/CD86 expression to assess DCs maturation. L) MHC-Ⅱ expression in DCs after different treatments. M) MHC-I expression in tumor after different treatments. (G1: Control; G2: aPDL1; G3: Ca; G4: aP@Ca; G5: Mn/Ca; G6: aP@Mn/Ca). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

2.4. Evaluation of in vivo antitumor immunity and tumor microenvironment remodeling

Given that both the bimetallic ions (Mn²⁺/Ca²⁺) and aPD-L1 in aP@Mn/Ca can moderately enhance immune cell function, we sought to investigate their synergistic potential in activating systemic immunity and alleviating immunosuppression in vivo. Following a 3day treatment regimen, single-cell suspensions were prepared from tumor-draining lymph nodes (TDLNs), spleens, and primary tumors of mice in each group to evaluate the antitumor immune effects (Fig. 5A and B). Given the pivotal role of dendritic cells (DCs) in antigen presentation, we first assessed DC maturation in TDLNs. The aP@Mn/Ca group showed the highest DC maturation rate (59.6%) (Fig. 5C and D). These mature DCs are crucial for T cell activation. Consistently, aP@Mn/Ca treatment induced the highest level of CD4⁺ T cell activation—approximately 3.5-fold and 1.8-fold higher than the control and aPD-L1 monotherapy groups, respectively (Supplementary Fig. 11). Furthermore, the proportion of tumor-infiltrating CD8⁺ T cells in the aP@Mn/Ca group was about 12.6-fold and 2.5-fold greater than in the control and aPD-L1 groups (Fig. 5E and F). To determine whether these infiltrating lymphocytes were functional, we assessed their proliferation and effector status. aP@Mn/Ca treatment resulted in high frequencies of proliferating (CD8⁺Ki67⁺: 46.3%), IFN-γ-producing (CD8⁺IFN-γ⁺: 28.6%), and granzyme B-expressing (CD8⁺GZMB⁺: 26.3%) CD8⁺ T cells (Supplementary Fig. 12). These percentages were 4-fold, 3-fold, and 2.8-fold higher than the control, and 3-fold, 2.4-fold, and 2.1-fold higher than the aPD-L1 group, respectively, indicating potent tumor-killing capacity and a shift toward a pro-inflammatory TME. We next analyzed changes in immunosuppressive cell populations. aP@Mn/Ca significantly reduced the proportions of MDSCs (from 35.2% to 19.8%; Fig. 5G and H) and Tregs (from 22.7% to 8.91%; Fig. 5I and J). Concurrently, it increased the ratio of pro-inflammatory M1-like macrophages (from 6.05% to 11.6%; Fig. 5K and L) while decreasing that of M2-like macrophages (from 19.8% to 5.8%; Fig. 5M and N). Cytokine levels in mouse tumor tissues (Fig. 5O–R) further corroborated the flow cytometry findings. Specifically, the aP@Mn/Ca group exhibited significantly elevated levels of the pro-inflammatory cytokines IL-2 and IL-6, alongside corresponding reductions in the anti-inflammatory cytokines IL-10 and TGF-β. In summary, while aPD-L1 monotherapy moderately reduced Tregs and MDSCs and promoted immune cell infiltration, its effects were substantially inferior to those of aP@Mn/Ca. The Mn/Ca-based delivery of aPD-L1 more effectively enhanced T lymphocyte infiltration into tumors and reversed the iMWA-induced immunosuppressive microenvironment, demonstrating superior antitumor immunotherapeutic efficacy.

Fig. 5.

In vivo evaluation of the anti-tumor properties of aP@Mn/Ca. A) Schematic diagram of the experimental design to evaluate the antitumor properties of aP@Mn/Ca in vivo and to induce the immune response. B) Schematic diagram of the remodeling of the tumor immune microenvironment by induced cellular pyroptosis.C) Flow cytometry of DC, E) CD8⁺ T cells, G) MDSC, I) Tregs, K) M1, M) M2 cells in different treatment groups of mouse tumors, tissues and lymph nodes. D) Quantitative analysis of DC, F) CD8⁺ T cells, H)MDSC, J)Tregs, L) M1, N) M2 cells.O) ELISA quantitative analysis of IL-2, P) IL-6, Q) IL-10, R) TGF-β in different treatment groups. (G1: Control; G2: aPDL1; G3: Ca; G4: aP@Ca; G5: Mn/Ca; G6: aP@Mn/Ca). (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

2.5. In vivo validation of antitumor properties and underlying mechanisms of HCC growth suppression

Based on the ability of aP@Mn/Ca to induce a systemic antitumor immune response in iMWA-treated HCC, we further evaluated its therapeutic efficacy in an iMWA-treated Hepa1-6 mouse model. We first investigated the in vivo biodistribution of aP@Mn/Ca via fluorescence imaging. In contrast to the control group, the tumor fluorescence signal intensified within 24 h post-injection, peaked at 24 h, and gradually declined over the next 36 h (Fig. 6A and Supplementary Fig. 13), indicating efficient tumor accumulation via the bloodstream and retention through the enhanced permeability and retention (EPR) effect. Ex vivo imaging of resected organs (Fig. 6B) confirmed this, showing the strongest signal in tumors and weaker signals in the liver and kidneys. To assess antitumor efficacy, iMWA-treated mice bearing Luc-Hepa1-6 tumors were randomized into groups and treated as per the scheme in Fig. 6C. Tumor growth was monitored by bioluminescence imaging every three days and caliper measurements every two days, alongside body weight tracking. aP@Mn/Ca treatment significantly attenuated bioluminescence signals and suppressed tumor proliferation (Fig. 6H). While the control group exhibited the most rapid tumor growth, aPD-L1 monotherapy partially constrained tumor volume to 800–1000 mm³ (Fig. 6D and E), consistent with its immunomodulatory activity. However, its efficacy was limited by the iMWA-induced immunosuppressive microenvironment. In contrast, aP@Mn/Ca demonstrated superior antitumor activity, maintaining tumor volumes at approximately 200 mm³ by day 14. Body weights remained stable across all groups (Fig. 6F), and H&E staining of major organs revealed no significant toxicity (Supplementary Figs. 14 and 15). A 60-day survival analysis showed that aP@Mn/Ca-treated mice achieved the highest survival rate (Fig. 6G). Histological examination of end-point tumors via H&E staining revealed nuclear fragmentation and loss of normal architecture in the aP@Mn/Ca group, suggesting programmed cell death. Further evaluation by TUNEL and Ki67 staining confirmed extensive DNA fragmentation (increased TUNEL signal) and significant suppression of proliferation (decreased Ki67 expression) in aP@Mn/Ca-treated tumors (Fig. 6I). Concurrently, significant CRT exposure and HMGB1 translocation were detected (Supplementary Fig. 16), indicating activation of ICD. In summary, as discussed above, aP@Mn/Ca effectively improves the therapeutic outcome post-iMWA, likely by alleviating the immunosuppressive microenvironment and remodeling the TME.

Fig. 6.

In vivo validation of the anti-tumor properties of aP@Mn/Ca. A) In vivo biofluorescence imaging of DID-labeled aP@Mn/Ca mice. B) Biofluorescence imaging of tumors and major organs 36 h after DID-aP@Mn/Ca injection. C) Schematic diagram of the mouse iMWA model. D) Tumor growth curves of different treatment groups. E) Tumor growth ratios of the mice in each treatment group. The data are shown as the means ± SDs (n = 5). F) Body weights of the mice from each treatment group during the entire evaluation. The data are shown as the means ± SDs (n = 5). G) Survival rates of the mice in each treatment group throughout the entire evaluation period (days 0–60) (n = 5). H) Biofluorescence imaging of mouse tumor volume in different treatment groups (n = 5). I) H&E staining, Ki67 and TUNEL staining of tumor cells in different treatment groups (Scale bar: 100 μm). (G1: Control; G2: aPDL1; G3: Ca; G4: aP@Ca; G5: Mn/Ca; G6: aP@Mn/Ca). (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

2.6. The long-term in vivo tumor suppression and anti-metastatic effects of aP@Mn/Ca

To further evaluate the inhibitory effect of aP@Mn/Ca on HCC growth, an orthotopic liver tumor model was established (Fig. 7A). Hepa1-6-Luc tumor cells were orthotopically implanted into the livers of C57BL/6J mice. Treatment was initiated one week after tumor inoculation, allowing for stable engraftment. Tumor progression was monitored via bioluminescence imaging and survival analysis in vivo. As shown in Fig. 7B and C, the aP@Mn/Ca group demonstrated the most potent tumor growth suppression and the longest survival, confirming its robust anti-tumor efficacy in the orthotopic HCC model.Given that recurrent distant metastasis is a major clinical concern following iMWA, we further investigated the anti-metastatic efficacy of aP@Mn/Ca in a lung metastasis mouse model (Fig. 7D). Mice were inoculated with Hepa1-6 cells on day-7 to establish the metastasis model. By day 19, extensive metastatic nodules were detected in the lungs of control mice, whereas only a few nodules were observed in the aP@Mn/Ca group (Fig. 7E and F). The significant reduction in lung weight in the treatment group further supported the anti-metastatic effect (Fig. 7G).

Fig. 7.

The long-term in vivo tumor suppression and anti-metastatic effects of aP@Mn/Ca. 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. Analysis of the number of metastatic nodules in lung tissue from mice treated with different regimens (F) and Lung weight (G). Flow cytometry analysis of the frequencies of effector memory T cells (TEM) and central memory T cells (TCM) (H, I, J), Tregs (K, M), CD8⁺ T cells (L, N) in tumor tissue. O) The levels of cytokines, including TNF-α, IFN-γ, and IL-2, in mouse serum from each treatment group. (G1:Control; G2: aPDL1; G3: Ca; G4: aP@Ca; G5:Mn/Ca; G6:aP@Mn/Ca). P values in these figures were calculated by the two-tailed Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

A key advantage of immunotherapy is its potential to induce immunological memory, thereby preventing metastasis and recurrence. To assess this potential for aP@Mn/Ca, we analyzed memory T cell subsets in lung tissues by flow cytometry. Treatment with aP@Mn/Ca increased the frequencies of effector memory T cells (TEM) and central memory T cells (TCM) by 2.5-fold and 1.6-fold, respectively, compared to the control (Fig. 7H–J), confirming the activation of long-term immunological memory. Concurrently, the immunosuppressive Treg cell population decreased from 14% to 5.67% (Fig. 7K–M), while tumor-infiltrating CD8⁺ T cells significantly increased (Fig. 7L–N). As the anti-metastatic effect likely stems from a systemic antitumor immune response, we measured serum cytokine levels. Compared to the control, the aP@Mn/Ca group showed markedly elevated levels of TNF-α (∼2.6-fold), IFN-γ (∼1.2-fold), and IL-2 (∼2.5-fold), indicating a robust, systemically activated immune state (Fig. 7O–Q). Concurrently, we performed the same validation experiments using the H22 cell line (Supplementary Fig. 17A). Compared with the control group, Mice were inoculated with H22 cells on day-7 to establish the metastasis model. By day 19, extensive metastatic nodules were detected in the lungs of control mice, whereas only a few nodules were observed in the aP@Mn/Ca group (Supplementary Fig. 17B and C). The significant reduction in lung weight in the treatment group further supported the anti-metastatic effect (Supplementary Fig. 17D). aP@Mn/Ca treatment increased the frequencies of both TEM and TCM (Supplementary Fig. 17F–G), confirming the activation of long-term immunological memory. Meanwhile, the proportion of Tregs decreased from 17.3% to 7.89% (Supplementary Fig. 17H and J), whereas tumor-infiltrating CD8⁺ T cells increased by approximately 2-fold (Supplementary Fig. 17I and K). In addition, the aP@Mn/Ca group exhibited significantly elevated levels of TNF-α, IFN-γ, and IL-2, indicating a potent and systemically activated immune state (Supplementary Fig. 17L–N).

To demonstrate the existence of robust and durable adaptive immunity, we performed definitive tumor rechallenge experiments in cured mice (Supplementary Fig. 18A; G1: Surgery; G2: Cured). Notably, in mice that had been cured of their primary H22 tumors by aP@Mn/Ca, all rechallenged tumors were eventually completely cleared (Supplementary Fig. 18B). In fact, all five mice in this group survived for more than 110 days (Supplementary Fig. 18C), and the number of memory T cells was significantly increased in aP@Mn/Ca-treated mice (Supplementary Fig. 18D and E). These results indicate that aP@Mn/Ca-induced tumor regression elicits a rather potent immunological memory effect.

3. Conclusions

In summary, our study proposes a therapeutic strategy to remodel the iMWA-induced immunosuppressive microenvironment by dually and directly/indirectly mediating inflammatory necrosis and enriching aPD-L1 at the tumor site. Recent years have witnessed substantial progress in the research of MOF materials. Compared with previous studies [[42], [43], [44]], aP@Mn/Ca offers a more integrated approach: (i) it simultaneously delivers therapeutic ions (Ca²⁺ and Mn²⁺) and an immune checkpoint antibody (aPD-L1) in a single platform; (ii) Ca²⁺ and Mn²⁺ induces pyroptosis and activates the cGAS-STING pathway synergistically; and (iii) it combines local ICD induction with systemic immune checkpoint blockade. The aP@Mn/Ca, engineered as a MOF based nanoplatform, leverages the inherent advantages of MOFs including high loading capacity, excellent biocompatibility, and tumor microenvironment-responsive degradability to responsively release bimetallic ions (Ca²⁺/Mn²⁺) and aPD-L1 antibodies within the tumor tissue. The ions disrupt the oxidative stress balance in residual cancer cells via mitochondrial calcium deposition, enhancing cytosolic ROS generation, which leads to mitochondrial damage and pyroptosis. Concurrently, the combined action of the cGAS-STING pathway synergistically enhances inflammation and necrosis in HCC cells. In the presence of the aPD-L1 antibody, a specific antitumor immune response is potently activated, recruiting more tumor-killing T lymphocytes into the tumor. Both in vitro and in vivo experiments confirmed that aP@Mn/Ca not only effectively eliminates iMWA-treated subcutaneous tumors and enhances ICD but also triggers a robust antitumor immune response to reprogram the immunosuppressive HCC microenvironment, demonstrating great potential for improving the prognosis of advanced HCC patients treated with MWA.

CRediT authorship contribution statement

Mengqi Zhang: Conceptualization, Methodology, Writing – original draft. Chao Rong: Methodology, Writing – original draft. Hejia Qin: Methodology. Xiali Qin: Resources. Zelong Jiang: Methodology, Software. Guanhua Qiu: Methodology, Software. Tian Qin: Methodology. Guchun Qin: Software. Niqiang Zhou: Software. Weizhong Tang: Resources, Supervision. Danke Su: Resources, Supervision. Yunxi Huang: Resources, Supervision, Writing – review & editing. Chang Zhao: Conceptualization, Funding acquisition, Supervision, 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

No data was used for the research described in the article.