Electroporation-primed cascade nanocatalysis boosts STING-driven chemoimmunotherapy in hepatocellular carcinoma

Abstract

Nanosecond pulsed electric field (nsPEF) ablation enables non-thermal tumor destruction with favorable tissue selectivity, but its long-term efficacy is often limited by postoperative recurrence. Here, we developed an electroporation-primed, tumor microenvironment-responsive core-shell cascade nanoparticle, ZnO₂-NiFc@MSCU (ZNFM), to potentiate nsPEF-mediated hepatocellular carcinoma (HCC) therapy and induce durable antitumor immunity. ZNFM contains a degradable ZnO₂ core that releases Zn²⁺ and generates H₂O₂, together with an acid-responsive NiFc shell that catalyzes H₂O₂ into cytotoxic reactive oxygen species, thereby enhancing local tumor eradication. Meanwhile, the incorporated prodrug MSCU, generated by conjugating the STING agonist MSA-2 with curcumin, undergoes esterase-triggered hydrolysis in tumor cells to co-release both agents. Mechanistically, Zn²⁺ potentiates STING activation, whereas curcumin promotes immunogenic cell death and alleviates immunosuppression by inhibiting the IDO-kynurenine axis. Importantly, nsPEF pretreatment serves as a delivery primer that markedly enhances intratumoral penetration and cellular uptake of ZNFM, thereby strengthening local therapeutic activation and promoting systemic antitumor immunity. In HCC models, nsPEF + ZNFM induces marked tumor regression, promotes dendritic cell maturation, elicits robust cytotoxic T-cell responses, and suppresses tumor rechallenge. These findings establish an electroporation-enabled cascade nanotherapeutic platform for durable HCC chemoimmunotherapy.

Graphical abstract

Highlights

• •An electroporation-primed core-shell cascade nanocatalytic platform was developed for hepatocellular carcinoma chemoimmunotherapy.
• •NsPEF acted as a delivery primer to enhance intratumoral deposition and cellular uptake of ZNFM.
• •ZNFM amplified oxidative stress, activated STING signaling, and suppressed the IDO-kynurenine axis in tumors.
• •The nsPEF + ZNFM strategy induced durable systemic antitumor immunity and protected against tumor rechallenge.

1. Introduction

Hepatocellular carcinoma (HCC) remains a major global health burden, with both incidence and mortality continuing to rise worldwide [1,2]. Among emerging local treatment modalities, nanosecond pulsed electric field (nsPEF) ablation has attracted considerable interest as a non-thermal strategy for solid tumors [3]. By delivering ultrashort high-voltage pulses, nsPEF induces irreversible nanopores in cellular membranes and ultimately triggers apoptotic cell death, while largely preserving adjacent blood vessels and surrounding normal structures [4,5]. This unique feature makes it particularly attractive for tumors located near critical anatomical sites. In addition to its direct cytotoxicity, nsPEF can promote the release of tumor-associated antigens and damage-associated molecular patterns, thereby initiating local antitumor immune responses [6,7]. Nevertheless, nsPEF monotherapy still has intrinsic limitations in solid tumors. Nonuniform electric-field distribution, tumor heterogeneity, and signal attenuation at the tumor margin may result in sublethal injury and incomplete tumor eradication, allowing residual cells to survive and drive local recurrence [8,9]. Moreover, although nsPEF can induce immunogenic cell death (ICD) and release tumor antigens and Damage-Associated Molecular Patterns (DAMPs), this endogenous immune stimulation is often insufficient to generate durable systemic antitumor immunity, especially within immunosuppressive solid-tumor microenvironments [10,11]. Recent studies indicate that nsPEF-induced antitumor immunity depends on dendritic cell activation, CD8⁺ T-cell responses, and cGAS-STING signaling, supporting the need for combination strategies that further amplify and sustain these immune effects [12]. Therefore, integrating nsPEF with immunomodulatory nanomedicine may provide a rational approach to improve local ablation efficacy while promoting durable systemic immune protection.

To address these limitations, increasing efforts have focused on integrating nsPEF with advanced drug-delivery platforms and immunomodulatory strategies. Among these approaches, the design of tumor microenvironment (TME) responsive prodrugs is particularly compelling [13,14]. Such prodrugs remain stable during systemic circulation but are selectively activated by tumor-associated biochemical cues, such as aberrantly expressed enzymes and acidic pH, thereby releasing the active drug [15]. For example, the elevated expression of esterases in tumor cells has emerged as an ideal trigger for the design of ester bond-linked prodrugs [16,17]. However, the therapeutic performance of such prodrugs is still frequently restricted by poor tumor accumulation and inefficient intracellular delivery, which together hinder their translational potential [[18], [19], [20]].

To address these challenges, the development of nanocarrier-based delivery systems has garnered significant attention. These systems enhance prodrug accumulation and promote intracellular release within tumors [21,22]. Advances in nanomedicine open new avenues for improving nsPEF efficacy. Nanoparticles, with their tunable size, surface properties, and drug-loading capacity, have become highly efficient carriers for tumor-targeted drug delivery [[23], [24], [25]]. Importantly, nsPEF pretreatment has been shown to transiently increase membrane permeability and loosen stromal barriers within tumors [6,[26], [27], [28]]. These effects create a more permissive intratumoral environment for nanoparticle penetration and cellular uptake. Accordingly, integrating nsPEF with nanomedicine may provide a rational route to achieve enhanced therapeutic efficacy through complementary mechanisms.

From an immunological perspective, the stimulator of interferon genes (STING) pathway plays a pivotal role in coupling tumor damage to innate immune activation [29]. In addition to pharmacological STING agonism, endogenous danger signals released from damaged tumor cells may also contribute to cGAS-STING activation. Among them, mitochondrial DNA (mtDNA) has been recognized as an important damage-associated molecular pattern. Upon mitochondrial injury, mtDNA can leak into the cytosol and be sensed by cyclic GMP-AMP synthase (cGAS), thereby activating the downstream STING–TBK1–IRF3 axis and type I interferon signaling [30]. In HCC, recent evidence further supports that cGAS-STING activation can enhance CD8⁺ T-cell recruitment and dendritic cell-dependent antitumor immunity [31,32]. Since nsPEF induces profound membrane perturbation, mitochondrial stress, and ICD, mtDNA leakage may represent an endogenous signal linking nsPEF-induced tumor damage to cGAS-STING activation. NsPEF triggers ICD within the TME, activating dendritic cells (DCs) [7,33], suggesting that pharmacological reinforcement of STING signaling may further amplify this process. Indeed, combining nsPEF with a STING agonist is expected to promote antigen presentation, strengthen T-cell priming, and increase the infiltration of CD8⁺ and CD4⁺ T cells into tumors [34]. Moreover, STING agonists can remodel tumor vasculature to support lymphocyte infiltration, further amplifying antitumor immunity [35]. On the pharmacological front, small-molecule STING agonists such as MSA-2 represent promising immunomodulators [36], while the natural compound curcumin not only induces ICD but also potently inhibits indoleamine 2,3-dioxygenase (IDO)—a key mediator of tumor immune escape [37,38]. However, both agents face pharmacokinetic challenges, including poor solubility, rapid clearance, and lack of tumor selectivity, which limits their therapeutic efficacy [39,40]. Therefore, advanced delivery systems are needed to overcome these limitations and improve drug targeting and stability.

In addition, recent studies have revealed that zinc ions (Zn²⁺) can enhance cGAS activity and stability, enabling sustained activation of the STING pathway [41]. In parallel, transition metal-based nanomaterials, such as nickel or iron, have shown promise in tumor therapy due to their ability to catalyze the conversion of tumor-overexpressed H₂O₂ into highly toxic reactive oxygen species (ROS) through a Fenton-like reaction under the mildly acidic tumor microenvironment [[42], [43], [44]]. This process can effectively induce tumor cell death and compensate for the limitations of nsPEF alone.

Motivated by the limited and short-lived immunity after nsPEF ablation and the delivery barriers imposed by dense HCC stroma, we engineer a cascade-responsive core-shell nanoplatform that converts electroporation-enabled deep deposition into amplified chemoimmunotherapy (Fig. 1). Specifically, we construct ZnO₂-NiFc@prodrug nanoparticles (ZNFM) consisting of (i) a degradable ZnO₂ core that provides a sustained Zn²⁺ flux while generating H₂O₂ as an endogenous substrate, (ii) a pH-labile NiFc hybrid shell that liberates catalytic species to drive Fenton-like ROS production in the mildly acidic TME, and (iii) an esterase-activatable bifunctional prodrug (MSCU) incorporated into the core-shell to co-deliver the STING agonist MSA-2 and curcumin. NsPEF pretreatment is leveraged as a “delivery primer” to transiently enhance membrane permeability, enabling ZNFM to penetrate deeply and be internalized efficiently within tumors. Once activated in situ, Zn²⁺ and MSA-2 cooperatively strengthen STING signaling to drive innate immune activation and dendritic cell maturation, whereas curcumin contributes cytotoxicity, reinforces immunogenic cell death, and relieves immune suppression via downregulating the IDO-kynurenine pathway. By integrating electroporation-assisted delivery with cascade chemical amplification and dual-payload prodrug activation, this work presents a functional materials strategy to potentiate nsPEF-based chemoimmunotherapy. This strategy enables effective control of primary HCC and promotes durable systemic antitumor immunity against recurrence and distant lesions.

Fig. 1.

Schematic illustration of the synergistic anti-tumor mechanism of nsPEF combined with the ZNFM nanoplatform.

2. Materials and methods

Synthesis of ZnO₂ and ZnO₂/Ni-Ferrocene (ZnO₂/NiFc) Nanoparticles: ZnO₂ nanoparticles were synthesized according to a previously reported method. In brief, zinc acetate (1.2 g) and PVP (1.2 g, Mw = 10,000) were dissolved in 60 mL of deionized water. Subsequently, 6 mL of 30 wt% H₂O₂ aqueous solution was rapidly added to the mixture under vigorous stirring. After 24 h of reaction, the resulting ZnO₂ nanoparticles were collected by centrifugation (11,800 rpm, 10 min), followed by washing three times with deionized water and ethanol, respectively. ZnO₂/NiFc nanoparticles were prepared via in situ assembly. Briefly, 10 mg of ZnO₂ and 30 mg of PVP-K30 were dispersed in a mixed solvent of 10 mL DMF and 10 mL water. Then, 0.35 mL of NiCl₂ solution (0.1 M) and 0.35 mL of Fc(COOH)₂ solution (0.1 M) were added dropwise under stirring. The mixture was stirred at 75°C for 2 h to form a NiFc shell containing coordinated Ni species and ferrocene-derived Fe-containing moieties on the ZnO₂ surface.

Synthesis of MSCU: Curcumin (200 mg, 1.0 equiv) and MSA-2 (500 mg, 3.0 equiv) were dissolved in anhydrous dichloromethane (DCM, 50 mL) at room temperature. EDCI (544 mg, 5.0 equiv) and 4-(dimethylamino) pyridine (DMAP, 200 mg, 3.0 equiv) were then added sequentially, and the reaction mixture was stirred at room temperature for 48 h. The reaction was quenched with saturated brine, and the aqueous phase was extracted with DCM. The combined organic layers were dried over anhydrous Na₂SO₄, filtered, and concentrated under reduced pressure. The crude product was purified by silica gel chromatography using DCM/ethyl acetate (20:1, v/v) as the eluent to afford MSCU as an orange-red powder (190 mg, 38% yield).

Synthesis of ZNFM: ZnO₂/NiFc nanoparticles (5 mg) and MSCU (5 mg) were dispersed in 15 mL of ethanol. After stirring for 12 h, the resulting ZNFM nanoparticles were collected by centrifugation (11,800 rpm, 10 min) and washed three times with ethanol. For clarity, the nomenclature of the nanoplatform was unified throughout the manuscript: ZnO₂ denotes the zinc peroxide core nanoparticles; ZNF denotes the ZnO₂-core/NiFc-shell carrier; and ZNFM denotes MSCU-loaded ZNF.

Live/Dead Cell Staining: Huh7 cells were seeded in 6-well plates and cultured for 24 h. Cells were then divided into six groups: (i) pH 6.5 DMEM supplemented with 10% fetal bovine serum (10% DMEM, control), (ii) MSCU (10 μM), (iii) ZNF (30 μg/mL), (iv) ZNFM (30 μg/mL), (v) nsPEF, and (vi) nsPEF + ZNFM (30 μg/mL). The nsPEF parameters were as described in the “nsPEF treatment” section. After the indicated treatments for 24 h, cells were incubated with calcein-AM (4 μM) and propidium iodide (PI, 8 μM) for 20 min at 37°C. Cells were washed with PBS to remove excess probes, and fluorescence images were acquired using an inverted fluorescence microscope.

Molecular docking: The 3D structure of Curcumin was retrieved from PubChem and subsequently imported into ChemBio3D Ultra 14.0 to undergo energy minimization. The crystal structure of the IDO was obtained from the Protein Data Bank under the accession number 5ETW. Following this, the structure was imported into PyMOL 3.1.1, where the original ligand was excised. Thereafter, the protein structure was transferred to AutoDockTools for additional analysis. Finally, the molecular docking simulation between Curcumin and IDO was executed with AutoDock Vina 1.5.7.

Intracellular Uptake Tracking of ZNFM: Huh7 cells were seeded in 6-well plates and incubated for 24 h. Cells were subjected to nsPEF and then incubated with rhodamine B–loaded ZNF (ZNF@RhB; equivalent carrier concentration: 30 μg/mL). At 1 and 6 h post-incubation, cells were washed with PBS to remove excess nanoparticles and imaged using an inverted fluorescence microscope to visualize intracellular rhodamine B fluorescence.

Intracellular ROS Detection: Huh7 cells were seeded in 6-well plates and incubated for 24 h. After 8 h of the indicated treatments, the culture medium was replaced with serum-free medium containing DCFH-DA (20 μM). Following incubation for 30 min at 37°C, the cells were washed with PBS and imaged using an inverted fluorescence microscope to record the ROS-associated fluorescence signal.

BMDCs Activation by Tumor-Conditioned Media: Hepa1-6 cells were treated with MSCU (9.21 μg/mL), ZNF (30 μg/mL), or ZNFM (30 μg/mL) for 24 h. The supernatants were collected and clarified by centrifugation (4°C, 10000 g, 10 min) to obtain Hepa1-6 conditioned media. Immature BMDCs were counted and seeded in 24-well plates, and then cultured in a 1:1 (v/v) mixture of Hepa1-6 conditioned media and fresh complete medium. After 24 h, BMDCs were collected for flow cytometric assessment of maturation markers (Fig. S30).

Biodistribution and Biosafety: For biodistribution analysis, tumor-bearing mice were randomly assigned to three groups and treated with (i) saline, (ii) ZNFM (20 mg/kg, i. v.), or (iii) ZNFM (20 mg/kg, i. v.) combined with nsPEF (ZNFM + nsPEF). At 24 h post-injection (n = 3), tumors and major organs were excised, weighed, and digested in acid. The contents of Zn and Fe in each tissue were quantified by ICP-OES. For biosafety evaluation, healthy mice were randomly divided into two groups and intravenously injected with saline or ZNFM (20 mg/kg) on day 0. Blood samples were collected on days 3 and 7 (n = 3) for routine hematology and serum biochemistry analyses. At the same time points (days 3 and 7), major organs were harvested for H&E staining and histopathological assessment.

In Vivo Antitumor Efficacy: Female C57BL/6 mice (6 weeks) were used for in vivo experiments. All animal procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals of Zhejiang University and were approved by the Animal Ethics Committee of Zhejiang University (Approval number: ZJU20250769).

To establish an orthotopic HCC model, Hepa1-6 cells (5 × 10⁶) were resuspended in a Matrigel/DMEM mixture (70%/30%, v/v; 20 μL) and injected under the liver capsule. Approximately 10 days post-implantation, mice with palpable tumors were assigned to receive nsPEF treatment or sham surgery. For the sham group, mice underwent the same surgical procedure and electrode placement without pulse delivery. Three days after treatment, mice were euthanized and orthotopic tumors from the nsPEF and sham groups were collected for RNA sequencing.

To establish the subcutaneous HCC tumor model, Hepa1-6 cells (1 × 10⁶) were suspended in PBS and injected subcutaneously into the right flank of each mouse. Tumor length and width were measured using a vernier caliper, and tumor volume was calculated as V = (width² × length)/2. When tumors reached approximately 200 mm³, mice were randomized into six groups and treated on days 0, 2, and 4 as follows: (i) saline (i.v.), (ii) MSCU (2.3 mg/kg, i. v.), (iii) ZNF (20 mg/kg, i. v.), (iv) ZNFM (20 mg/kg, i. v.), (v) nsPEF, and (vi) nsPEF + ZNFM (20 mg/kg, i. v.). For the nsPEF + ZNFM group, mice received nsPEF on day 0, immediately followed by tail-vein injection of ZNFM.

Body weights and tumor volumes were recorded throughout the treatment period. On day 21, 6 mice in each group were reserved for survival analysis, while the remaining mice were sacrificed. Tumors were excised, weighed, and fixed in 4% paraformaldehyde for histological analyses. Paraffin-embedded tumor sections were prepared for H&E staining. In parallel, frozen tumor sections were used for Ki67 and TUNEL immunofluorescence staining to evaluate therapeutic efficacy.

Immune Activation Analysis: An additional cohort of tumor-bearing mice (6 groups) was treated using the same dosing schedule described above. During the treatment period, mice were partially euthanized (n = 3) and tumors were collected for flow-cytometric quantification of cytotoxic T cells (CD3⁺CD8⁺) and helper T cells (CD3⁺CD4⁺) (Figs. S31–32). Cells were stained with anti-CD45-APC/Cy7 (BioLegend, 103,116), anti-CD3-PerCP/Cy5.5 (BioLegend, 100,217), anti-CD8a-PE/Cy7 (BioLegend, 100,721), and anti-CD4-APC (BioLegend, 100,411) prior to acquisition on a flow cytometer. In parallel, serum IL-6 and IFN-β levels were measured in mice receiving intravenous ZNFM (n = 3). Tumor-draining lymph nodes were harvested to analyze the maturation of dendritic cells by flow cytometry.

To evaluate systemic antitumor immunity, after completion of the 14-day treatment course, mice were rechallenged by subcutaneous inoculation of 2 × 10⁶ Hepa1-6 cells into the contralateral flank, and the growth of distant tumors was monitored for 14 days. On day 15 after rechallenge, tumors were excised, photographed, and measured for volume and weight. Tumor tissues were further processed for flow-cytometric analysis of CD3⁺CD8⁺ cytotoxic T cells and CD3⁺CD4⁺ helper T cells.

Statistical Analysis: All data values are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism and ImageJ. Student's t-test was used for comparisons between two groups, and one-way ANOVA followed by Tukey's multiple-comparisons test was applied for multiple-group comparisons. Differences were considered significant at p < 0.05 (*), p < 0.01 (**).

3. Results and discussion

3.1. Design rationale and validation of the esterase-activatable prodrug MSCU

To establish a molecularly informed strategy for augmenting nsPEF-enabled HCC immunotherapy, we first profiled the post-ablation tumor transcriptome (Fig. 2a). RNA-seq analysis revealed a large set of differentially expressed genes (DEGs) between nsPEF-treated and untreated tumors, indicating that nsPEF profoundly reshaped the tumor transcriptional landscape (Fig. 2b). Functional enrichment analysis further demonstrated that these DEGs were highly enriched in immune-associated pathways (Fig. 2c). Notably, enrichment of the “negative regulation of immune system process” suggested that nsPEF may also engage immunosuppressive programs within the TME, thereby constraining the initiation and amplification of effective antitumor immunity. This immunosuppressive signature implies that physical ablation alone may be insufficient for durable tumor control, and highlights the need for rational combination strategies that actively promote immune activation while maintaining cytotoxic tumor elimination.

Fig. 2.

NsPEF engages immunosuppressive programs in HCC and motivates the design and validation of MSCU. (a) Schematic illustration of transcriptomic profiling of tumor tissues after nsPEF ablation by RNA sequencing. (b) Volcano plot showing DEGs in nsPEF-treated tumors relative to untreated controls (n = 3). (c) GO and KEGG enrichment analysis of DEGs. (d) MS analysis of the synthesized MSCU molecule. (e) MS analysis of MSCU degradation products under esterase conditions. (f) ¹H NMR spectrum of MSCU. (g) CCK-8 assay of Huh7 cells after incubation with curcumin, MSA-2, or MSCU at the indicated concentrations for 24 h (n = 3). (h) Representative wound-healing images of Huh7 cells treated as indicated for 0 and 24 h. (i) Quantification of healing percent in different treatment groups (n = 3, **p < 0.01). (j) Live/dead staining evaluating the cytotoxic effects of curcumin, MSA-2, and MSCU on Huh7 cells. (k) Schematic diagram of the mechanism of action of MSCU. (l) Western blot analysis of STING pathway-related proteins and phosphorylation levels. (m) Schematic illustration of key markers of ICD. (n) Immunofluorescence detection of CRT exposure and HMGB1 release. (o) Representative flow cytometric analysis of BMDC maturation, as indicated by CD80 and CD86 expression after co-culture with differently treated tumor cells. Scale bars, 50 μm.

Motivated by this observation, we next designed a dual-functional prodrug aimed at integrating chemotherapy-driven tumor killing with immune stimulation. Since the STING pathway represents a central hub for cytosolic DNA sensing and can potentiate antitumor immunity, we selected the small-molecule STING agonist MSA-2 as the immunostimulatory module [45]. However, MSA-2 lacks direct cytotoxicity toward tumor cells, limiting its efficacy in reducing tumor burden. Furthermore, MSA-2 treatment leads to higher expression of IDO in both early and late-stage tumors, which negatively regulates the STING pathway [46,47]. To overcome these limitations, we chemically linked MSA-2 with curcumin, a chemotherapeutic agent that may inhibit IDO, via an esterase-cleavable ester bond to create MSCU (Fig. S1). This dual-functional prodrug, MSCU, effectively integrates chemotherapy-induced tumor cytotoxicity with immune activation, offering a promising approach to overcome the immunosuppressive characteristics of the TME and improve therapeutic outcomes in HCC. Mass spectrometry (MS) confirmed the expected molecular weight of MSCU, supporting its successful synthesis (Fig. 2d). Importantly, upon incubation with esterase, degradation products corresponding to both MSA-2 and curcumin were readily detected by MS, indicating efficient enzymatic cleavage of MSCU and release of both active payloads (Fig. 2e). This esterase-triggered conversion was further supported by UV-vis spectral changes (Fig. S2), and structural correctness was validated by ¹H NMR analysis (Fig. 2f). Together, these results demonstrate that MSCU is a structurally well-defined, esterase responsive prodrug capable of “prodrug-to-active drug” conversion under enzymatic conditions, making it suitable for biological applications in the TME.

We next evaluated the cytotoxic functions of MSCU by performing CCK-8 assays at different concentrations of curcumin, MSA-2, and MSCU over 24 h (Fig. 2g). The results revealed that 10 μM MSCU induced approximately 50% cell death in Huh7 hepatoma cells after 24 h, which was used for subsequent experiments. We then assessed the time-dependent cytotoxicity of MSCU at this concentration. As shown in Fig. S3, MSCU-induced cell death increased progressively over time, with cell viability decreasing to approximately 50% at 24 h, 24% at 48 h, and 14% at 72 h. MSCU also inhibited tumor cell migration, as demonstrated by the wound-healing assays (Fig. 2h–i). Specifically, the migration rate of Huh7 cells in the MSCU-treated group was reduced to approximately 21%, compared to 56% in the control group. Live/dead staining of Huh7 cells treated with 10 μM MSCU confirmed that the majority of the cytotoxicity observed was attributed to curcumin release (Fig. 2j). Mechanistically, MSCU was designed to release MSA-2 upon esterase cleavage, thereby activating the STING pathway and triggering downstream TBK1/IRF3 signaling (Fig. 2k). Consistent with this design principle, western blot analysis showed that both MSCU and its esterase-degraded products robustly increased phosphorylation levels of STING, TBK1, and IRF3, with activation magnitude comparable to free MSA-2 (Fig. 2l, Fig. S4a–c). These data demonstrate that MSCU preserves STING agonistic activity following enzymatic conversion, validating the feasibility of using an esterase-responsive linkage to enable immune activation in situ.

STING activation and chemotherapy-induced cellular stress can converge to promote ICD, thereby enhancing antigen presentation and facilitating T-cell priming. ICD is characterized by hallmark danger signals including calreticulin (CRT) exposure, ATP release, and High Mobility Group Protein B1 (HMGB1) release [48] (Fig. 2m). Immunofluorescence staining revealed that MSCU markedly increased CRT exposure and promoted HMGB1 release (Fig. 2n), indicating robust ICD induction. In line with this, MSCU-treated tumor cells showed elevated extracellular HMGB1 and ATP levels (Figs. S5–6), further supporting the activation of ICD-associated signaling. Importantly, these immunogenic cues also translated into functional immune activation, as MSCU significantly promoted bone marrow derived dendritic cells (BMDCs) maturation (Fig. 2o). Collectively, these findings establish MSCU as an esterase-activatable dual-payload prodrug that enables coordinated release of MSA-2 and curcumin, thereby integrating STING pathway activation with ICD induction. This prodrug design provides a mechanistic basis for reinforcing the transient immune activation induced by nsPEF while counteracting the accompanying immunosuppressive response.

3.2. Construction of ZNFM and TME-programmed cascade degradation and outputs

Building on MSCU, we next fabricated a ZnO₂-core/NiFc-shell carrier, designated as ZNF, and further incorporated MSCU to obtain the final nanoplatform, designated as ZNFM. Here, ZNF refers to the ZnO₂/NiFc core-shell carrier, whereas ZNFM refers to MSCU-loaded ZNF. This design aimed to translate TME acidity and esterase activity into sequential elemental Zn and H₂O₂ supply, catalytic ROS generation, and on-demand release of MSA-2 and curcumin.

ZnO₂ nanoparticles were synthesized by reacting zinc acetate with H₂O₂ in the presence of polyvinylpyrrolidone (PVP) in deionized water. For the assembly of ZNF, a ZnO₂-core/NiFc-shell structure was constructed via in situ coating of ZnO₂ using NiCl₂ and ferrocene dicarboxylic acid (Fc (COOH)₂) as shell precursors in a DMF/water mixture. As shown in Fig. 3a, scanning electron microscopy (SEM) images revealed that the particles were relatively uniform in morphology, while transmission electron microscopy (TEM) images clearly demonstrated a distinct core-shell structure. X-ray diffraction (XRD) analysis was subsequently performed to examine crystallinity. In Fig. 3b, the diffraction peaks of ZnO₂ matched well with the standard pattern (PDF#13-0311), whereas ZNF exhibited additional diffraction features attributable to the NiFc-derived shell, confirming successful shell deposition while preserving the crystalline characteristics of the ZnO₂ core. The elemental composition of the nanoplatform was first examined by X-ray photoelectron spectroscopy (XPS). The survey spectrum revealed the presence of Zn, Ni, Fe, and O (Fig. 3c), demonstrating successful incorporation of each element. High-resolution XPS spectra further confirmed the valence states of Zn²⁺, Fe²⁺, Ni²⁺, and peroxide O₂²⁻ (Figs. S7–10), indicating that the ZnO₂ core retained its peroxide feature and that the NiFc shell preserved coordinated Ni species together with ferrocene-derived Fe centers as redox-active components. To further confirm drug loading and enable subsequent quantification, the UV-vis absorbance characteristics of MSCU were first established. As shown in Fig. S11, MSCU exhibited concentration-dependent absorption with good linearity over the tested range, and the corresponding standard curve was used for quantification. On this basis, the UV-vis spectrum of drug-loaded nanoparticles showed the characteristic absorption signature of MSCU, further confirming successful loading of MSCU into the nanocarrier (Fig. 3d). TEM-coupled elemental mapping further demonstrated that Zn was mainly distributed throughout the nanoparticle core, whereas Ni and Fe were enriched at the particle periphery, consistent with the intended ZnO₂-core/NiFc-shell architecture (Fig. 3e). Notably, sulfur (S) was also detected after MSCU loading, providing additional evidence for successful drug incorporation because S originates from the MSA-2.

Fig. 3.

Fabrication and Stimuli-Responsive Performance of ZNFM. (a) SEM and TEM images of the synthesized nanoparticles. (b) XRD patterns of NF simulation, ZnO₂, and ZNF. (c) XPS survey spectrum of ZNF. (d) UV-vis spectra of MSCU-loaded nanoparticles, confirming successful drug loading. (e) TEM image and corresponding elemental mapping of ZNFM, showing the distribution of O, Zn, Ni, Fe, and S. (f) DLS size distribution of ZnO₂, ZNF, and ZNFM (n = 3). (g) Zeta potentials of ZnO₂, ZNF, and ZNFM (n = 3). (h) Schematic illustration of the proposed TME-responsive cascade mechanism of ZNFM. (i) Cumulative Zn²⁺ release from ZNFM at pH 5.8 and pH 7.4 (n = 3). (j) Cumulative Fe release from ZNFM at pH 5.8 and pH 7.4 (n = 3). (k) Cumulative MSA-2 release from ZNFM under different pH conditions in the presence or absence of esterase (n = 3). (l) H₂O₂ generation by ZNFM under acidic and neutral conditions (n = 3). (m) ESR spectra of DMPO-·OH adducts generated by ZNFM at pH 5.8 and pH 7.4. (n) TEM images showing the structural evolution of ZNFM after incubation at pH 7.4 or pH 5.8 for the indicated times.

The physicochemical properties of the nanoparticles were then evaluated. DLS analysis indicated narrow size distributions for ZnO₂, ZNF, and ZNFM, while loading of NiFc and MSCU led to an increase in hydrodynamic diameter (Fig. 3f). ZNFM showed an average hydrodynamic size of 172.1 nm (PDI = 0.1407), a size considered favorable for tumor accumulation. Meanwhile, zeta-potential measurement further supported NiFc shell modification and MSCU incorporation (Fig. 3g). The nanoparticles exhibited a negative surface charge mainly due to the carboxyl groups (-COOH) of Fc [49], and the surface potential became more negative after MSCU loading (∼-35 mV). This change may arise from enhanced association between MSCU and Fc-containing domains, potentially increasing the exposure of anionic groups and improving colloidal stability. Stability under aqueous and physiological-like conditions is essential for biomedical applications [50]. To evaluate the colloidal stability and possible agglomeration of ZNFM under storage and physiological-like conditions, ZNFM was dispersed in DI water, saline, and serum-containing culture medium, and its hydrodynamic size distributions were monitored on days 0, 3, 7, and 14. As shown in Fig. S12, ZNFM displayed narrow and similar size distribution profiles in all tested media throughout the 14-day observation period. No obvious peak broadening or shift toward larger particle sizes was observed over time, suggesting that ZNFM maintained good colloidal stability without evident agglomeration under these conditions. These results further support the suitability of ZNFM for subsequent biological applications.

We next investigated the TME-responsive degradation and functional activation of ZNFM. As illustrated in Fig. 3h, following cellular internalization, the acidic microenvironment promotes degradation of the NiFc shell, facilitating MSCU release. In parallel, MSCU can be hydrolyzed by abundant tumor-associated esterases into MSA-2 and curcumin, exerting immunostimulatory and chemotherapeutic effects, respectively. Meanwhile, ZnO₂ undergoes acid-triggered decomposition to release Zn²⁺ and generate H₂O₂. The generated H₂O₂ can further react with shell-associated redox-active species, including ferrocene-derived Fe centers and coordinated Ni species, through Fenton-like processes to produce hydroxyl radicals (·OH) and amplify intracellular ROS levels [51]. Consistent with this design, Zn²⁺ release from ZNFM was markedly accelerated under acidic conditions. As shown in Fig. 3i, cumulative Zn²⁺ release at pH 5.8 was substantially higher than that at pH 7.4, indicating acid-responsive decomposition of the ZnO₂ core. A similar pH-dependent trend was observed for Fe release (Fig. 3j), suggesting concurrent destabilization of the NiFc shell in an acidic environment. Since MSCU hydrolysis produces both MSA-2 and curcumin, MSA-2 was selected as a representative marker for prodrug activation. A standard UV-vis calibration curve of MSA-2 was first established for quantitative analysis (Fig. S13). As shown in Fig. 3k, the cumulative release of MSA-2 was significantly enhanced at pH 5.8 and was further promoted in the presence of esterase, whereas only limited release was observed at pH 7.4 regardless of esterase treatment. These results demonstrate that ZNFM possesses dual responsiveness to acidity and esterase activity, thereby enabling tumor-selective activation of the loaded prodrug.

Finally, the H₂O₂-generating capability and Fenton catalytic activity of ZNFM were evaluated. The UV-vis spectra and standard curve for H₂O₂ detection were first established (Fig. S14a–b). Considering that 30 μg/mL ZNFM was used as a representative working concentration in subsequent cellular experiments, H₂O₂ generation was directly assessed at this concentration under neutral and acidic conditions. As shown in Fig. 3l, ZNFM generated markedly higher levels of H₂O₂ under acidic conditions than under neutral conditions, confirming its acidity-responsive H₂O₂-generating capability at the biologically relevant concentration used in vitro. This result supports that ZNFM can provide an acid-triggered H₂O₂ source for subsequent catalytic ROS generation in the cellular treatment setting. The Fenton catalytic activity of ZNFM was further evaluated using a TMB chromogenic assay (Fig. S15a–c). Under acidic conditions, increasing the concentration of ZNFM led to a progressive increase in the absorbance of ox-TMB at around 652 nm, indicating enhanced ·OH generation. In addition, electron spin resonance (ESR) spectroscopy further confirmed the production of ·OH, as evidenced by the characteristic 1:2:2:1 quartet signal of the DMPO-·OH adduct (Fig. 3m). The pH-responsive structural evolution of ZNFM was directly visualized by TEM. As shown in Fig. 3n, the nanoparticles remained morphologically intact after incubation at pH 7.4, indicating good stability under physiological conditions. In contrast, under acidic conditions (pH 5.8), ZNFM gradually underwent collapse and fragmentation over time, with obvious structural disruption observed at 4 h and near-complete disassembly at 24 h. Together, these results demonstrate the pH-responsive H₂O₂-generating and Fenton catalytic properties of ZNFM, as well as its acid-triggered structural disassembly, supporting its potential for enhanced tumor-selective therapeutic activity.

3.3. NsPEF-ZNFM synergy is driven by electroporation-primed delivery and cascade immunochemical amplification

To clarify how electrical priming and chemical immunomodulation jointly contribute to the enhanced therapeutic effect, we dissected the nsPEF-ZNFM synergy into a three-step mechanistic sequence involving nsPEF-mediated delivery enhancement, ZNFM-driven intracellular cascade immunochemical amplification, and subsequent apoptosis, ICD induction, and immune remodeling. We first established an in vitro nsPEF treatment model to determine whether electroporation could facilitate nanoparticle internalization (Fig. 4a and Fig. S16). Cellular uptake was evaluated using Rhodamine-labeled ZNF. Compared with cells without nsPEF treatment, nsPEF-treated Huh7 cells exhibited stronger intracellular red fluorescence after incubation with ZNF, especially at 6 h (Fig. 4b). Quantitative analysis further confirmed that nsPEF significantly increased the intracellular fluorescence intensity of Rhodamine-labeled ZNF (Fig. S17), indicating that nsPEF acts as a physical delivery primer to enhance nanoparticle uptake by tumor cells. To further determine whether this delivery-promoting effect also occurred in vivo, ICP-OES analysis was performed to quantify the biodistribution of ZNFM-associated elements in major organs and tumors after intravenous administration. As shown in Fig. 4c and Fig. S18, the intratumoral levels of Zn and Fe in the nsPEF group reached approximately 40 μg/g and 195 μg/g, respectively, both of which were higher than those in the control and no nsPEF groups. By contrast, no evident abnormal retention of Zn or Fe was detected in the heart, liver, spleen, lung, or kidney. These results demonstrate that nsPEF not only promotes intracellular nanoparticle uptake in vitro but also enhances tumor-localized deposition of ZNFM in vivo.

Fig. 4.

Mechanistic basis of nsPEF-ZNFM synergy through electroporation-primed delivery and cascade immunochemical amplification. (a) Schematic illustration of the in vitro nsPEF treatment procedure. (b) Cellular uptake of Rhodamine-labeled ZNF in Huh7 cells with or without nsPEF at 1 and 6 h. (c) ICP-OES quantification of Zn in the heart, liver, spleen, lung, kidney, and tumor (n = 3). (d) Viability of AML12 cells after incubation with different concentrations of ZNFM for 24 and 48 h (n = 3). (e) Viability of Huh7 cells after incubation with different concentrations of ZNFM under neutral (pH 7.4) and mildly acidic (pH 6.5) conditions (n = 3). (f) Intracellular ROS generation in Huh7 cells after the indicated treatments. (g) Volcano plot of differentially expressed genes after treatment. (h) GO and KEGG enrichment analyses of differentially expressed genes. (i) Western blot analysis of STING pathway activation in Huh7 cells after the indicated treatments. (j) Molecular docking analysis of curcumin with IDO. (k) Western blot analysis of IDO expression in Huh7 cells after the indicated treatments. (l) ELISA quantification of KYN levels in culture supernatants after the indicated treatments (n = 3). (m) Representative Annexin V-FITC/PI flow cytometry plots and quantitative analysis of apoptosis in Huh7 cells after the indicated treatments (n = 3). (n) Immunofluorescence images of CRT exposure in Huh7 cells after the indicated treatments, with corresponding quantitative analysis of CRT fluorescence intensity (n = 3). Scale bars, 50 μm **P < 0.01.

Enhanced intracellular accumulation should theoretically increase the extent of intracellular cascade activation. We therefore next examined whether ZNFM could convert electroporation-enhanced delivery into amplified chemical and immunological stress. In normal mouse hepatocytes AML12, ZNFM caused negligible cytotoxicity over the tested concentration range of 0–40 μg/mL at both 24 and 48 h, suggesting favorable cytocompatibility toward non-malignant hepatocytes (Fig. 4d). In contrast, ZNFM markedly reduced the viability of Huh7 cells in a dose-dependent manner, and this cytotoxic effect was more pronounced under mildly acidic conditions than under neutral conditions (Fig. 4e). At a representative dose of 30 μg/mL, ZNFM induced pronounced tumor cell killing after 24 h, with approximately 55–60% cell death under mildly acidic conditions (Fig. 4e and Fig. S19). Moreover, time-course CCK-8 assays showed that ZNFM-induced Huh7 cell death progressively increased with prolonged incubation (Fig. S20), consistent with stimulus-responsive activation and cumulative intracellular stress. These results indicate that ZNFM possesses preferential cytotoxicity toward tumor cells while maintaining good compatibility with normal hepatocytes. We then examined whether ZNFM-induced tumor cell damage was associated with intracellular oxidative stress. DCFH-DA staining showed that both ZNF and ZNFM markedly increased intracellular ROS fluorescence compared with the control and MSCU groups (Fig. 4f). The strong ROS signal in nanoparticle-treated cells supports the contribution of the ZnO₂/NiFc cascade to oxidative stress amplification after cellular internalization. Together with the pH-dependent cytotoxicity results, these findings suggest that ZNFM can translate tumor microenvironment-responsive chemical activation into ROS-mediated tumoricidal stress.

To further explore the global molecular alterations induced by ZNFM and to guide subsequent mechanistic validation, RNA-sequencing analysis was performed in Huh7 cells after treatment with ZNFM or control medium. The volcano plot revealed a broad transcriptional remodeling profile in the ZNFM-treated group, with numerous differentially expressed genes identified between ZNFM and control cells (Fig. 4g). Notably, IDO1 was significantly downregulated after ZNFM treatment, suggesting that ZNFM may interfere with the IDO-associated immunosuppressive pathway. Functional enrichment analysis of the differentially expressed genes further showed that ZNFM treatment significantly affected multiple immune- and stress-related biological processes, including the TNF signaling pathway, cytokine-mediated signaling pathway, response to metal ion, leukocyte migration, NF-κB signaling pathway, cellular response to oxidative stress, regulation of innate immune response, response to reactive oxygen species, and antigen processing and presentation (Fig. 4h). These transcriptomic findings indicate that ZNFM not only induces oxidative stress-associated tumor cell injury but also remodels tumor cells toward an immunologically active state. Importantly, the enrichment of innate immune response, cytokine signaling, oxidative stress, and antigen presentation-related pathways, together with IDO1 downregulation, provided a rationale for the subsequent validation of STING pathway activation, IDO-KYN axis inhibition, and immunogenic cell death induction.

Guided by the transcriptomic results and the designed co-delivery function of the MSCU prodrug module, we next investigated whether ZNFM could activate STING signaling and modulate the IDO-KYN immunosuppressive axis in tumor cells. Western blot analysis showed that MSCU and ZNFM markedly increased the phosphorylation levels of STING and its downstream mediators TBK1 and IRF3, whereas ZNF alone induced only limited pathway activation (Fig. 4i and Fig. S21a–c). To further determine whether Zn²⁺ itself contributes to STING pathway activation, we performed an additional in vitro experiment using ZnCl₂ as a soluble Zn²⁺ control. Compared with the control group, ZnCl₂ alone evidently increased the phosphorylation levels of STING, TBK1, and IRF3 (Fig. S22), indicating that Zn²⁺ can independently contribute to STING-TBK1-IRF3 pathway activation. Notably, ZNFM induced stronger phosphorylation of STING, TBK1, and IRF3 than ZnCl₂ alone, suggesting that the STING-activating effect of ZNFM is not solely attributable to Zn²⁺, but may result from the combined contribution of Zn²⁺ release, MSCU-mediated STING agonism, and cascade oxidative stress amplification. Consistently, ELISA analysis showed significantly increased secretion of IFN-β and IL-6 in the MSCU and ZNFM groups (IFN-β to ∼120 pg/mL; IL-6 to ∼28 pg/mL) (Fig. S23a–b). These data indicate that the MSCU module within ZNFM preserves its STING-agonistic activity and effectively activates the canonical STING-TBK1-IRF3 signaling cascade. Since IDO is the key rate-limiting enzyme for kynurenine (KYN) biosynthesis and KYN accumulation is associated with immunosuppressive signaling that can indirectly constrain STING activity [52,53], we further examined whether curcumin released from MSCU could suppress the IDO-KYN axis. Molecular docking suggested a favorable interaction between curcumin and IDO, with a binding energy of −9.0 kcal/mol (Fig. 4j). Western blot analysis showed that MSCU and ZNFM reduced IDO expression in Huh7 cells (Fig. 4k and Fig. S24). Functionally, KYN levels were decreased from approximately 9000 pg/mL in the control group to approximately 7200 pg/mL in the ZNFM group (Fig. 4l). These results suggest that ZNFM not only activates STING signaling through MSA-2 release but also weakens the IDO-KYN immunosuppressive axis through curcumin-mediated modulation, thereby favoring a more immunostimulatory tumor-cell state.

Finally, we investigated whether the physical delivery enhancement mediated by nsPEF and the chemical cascade activation of ZNFM could be translated into amplified biological antitumor effects. Annexin V-FITC/PI flow cytometry showed that the apoptotic fraction increased from approximately 3% in the control group to approximately 30% in the nsPEF + ZNFM group, which was substantially higher than that induced by ZNFM or nsPEF alone (Fig. 4m). This result indicates that nsPEF-enhanced intracellular delivery and ZNFM-mediated cascade chemical activation cooperatively promote tumor cell apoptosis. In addition to direct tumor cell killing, ICD was further evaluated by detecting damage-associated molecular patterns. Immunofluorescence staining demonstrated that nsPEF + ZNFM markedly enhanced CRT exposure compared with each monotherapy group (Fig. 4n). In parallel, HMGB1 release was also increased after the combination treatment (Fig. S25), confirming strengthened ICD-associated danger signaling. These results demonstrate that the physical–chemical synergy between nsPEF and ZNFM is ultimately converted into biological functional outputs, including enhanced tumor cell apoptosis and ICD induction. Collectively, these findings support a three-step mechanistic model underlying nsPEF-ZNFM synergy. First, nsPEF acts as a physical electroporation primer that enhances intratumoral delivery and intracellular accumulation of ZNFM. Second, intracellularly activated ZNFM converts this delivery advantage into amplified ROS stress, STING-TBK1-IRF3 activation, and suppression of the IDO-KYN immunosuppressive axis. Third, these convergent immunochemical signals promote ICD-associated danger signaling and immune remodeling, thereby transforming transient nsPEF-induced immune perturbation into sustained antitumor immunity.

3.4. NsPEF-assisted ZNFM therapy potentiates STING-driven ICD and antitumor immunity in vivo

To evaluate the in vivo therapeutic efficacy of the combination strategy, we established a subcutaneous primary tumor model together with a secondary tumor model and applied a treatment regimen integrating local nsPEF stimulation with repeated intravenous administration of ZNFM (Fig. 5a). The in vivo nsPEF device and treatment interface are shown in Fig. 5b and Fig. S26, confirming the feasibility of localized pulse delivery during animal experiments. Throughout the treatment period, body weight remained stable in all groups, indicating good systemic tolerability of the treatment protocol (Fig. 5c). We next assessed the therapeutic efficacy against the primary subcutaneous tumors. At the endpoint, ex vivo tumor images showed that the nsPEF + ZNFM group exhibited the smallest tumor burden among all groups (Fig. 5d). Consistently, longitudinal monitoring of tumor growth demonstrated that the combination treatment achieved the strongest and most durable suppression of tumor progression (Fig. 5e). Notably, tumors in the nsPEF + ZNFM group remained very small at the endpoint (terminal volume ∼80 mm³), whereas the nsPEF-only group showed an initial inhibitory effect followed by tumor regrowth during the later stage of treatment. Tumor-weight analysis further confirmed the superior therapeutic effect of the combined treatment, with the nsPEF + ZNFM group showing the lowest terminal tumor mass (terminal tumor masses decreased to ∼70 mg) (Fig. 5f). When compared to the nsPEF-only group, the nsPEF + ZNFM group exhibited a tumor inhibition rate of approximately 79%, highlighting the enhanced antitumor efficacy of the combined treatment. Together, these data demonstrate that nsPEF markedly potentiates the in vivo antitumor efficacy of ZNFM against primary tumors.

Fig. 5.

NsPEF-assisted ZNFM therapy induces potent tumor suppression and promotes antitumor immune activation in vivo. (a) Schematic illustration of the treatment schedule in the subcutaneous primary tumor model and secondary tumor model. (b) Schematic of the nsPEF procedure applied to tumor-bearing mice. (c) Body weight changes of mice during treatment (n = 4). (d) Representative photographs of excised primary tumors collected from different treatment groups at the experimental endpoint. (e) Growth curves of primary tumors in different groups during treatment (n = 4). (f) Final tumor weights of primary tumors harvested from different groups (n = 4). (g) Western blot analysis of STING pathway-related proteins, including p-STING, STING, p-TBK1, and p-IRF3, in tumor tissues after the indicated treatments. (h) Western blot analysis of CRT expression in tumor tissues after the indicated treatments. (i) Representative H&E, Ki67, and TUNEL staining images of tumor sections collected from different groups. (j) Immunofluorescence staining of tumor sections showing CD3⁺CD8⁺ and CD3⁺CD4⁺ T-cell infiltration in the nsPEF and nsPEF + ZNFM groups. (k) Representative flow cytometry plots and quantitative analysis of mature DCs (CD80⁺CD86⁺) in tumor-draining lymph nodes. (l) Representative flow cytometry plots and quantitative analysis of CD8⁺ T cells in tumor. (m) Representative flow cytometry plots and quantitative analysis of CD4⁺ T cells in tumor. Scale bars, 100 μm *P < 0.05, **P < 0.01.

To elucidate the mechanism underlying this enhanced tumor suppression, we first examined STING pathway activation in tumor tissues. Western blot analysis showed that nsPEF + ZNFM markedly increased p-STING levels as well as the phosphorylation of downstream TBK1 and IRF3 compared with the other treatment groups (Fig. 5g and Fig. S27a–c). Because STING activation is closely associated with ICD, we next examined CRT exposure in tumor tissues. Western blot analysis showed elevated CRT expression after treatment, with the highest level observed in the nsPEF + ZNFM group (Fig. 5h and Fig. S28). Consistently, immunofluorescence staining further confirmed enhanced CRT exposure in the nsPEF + ZNFM group (Fig. S29). Histological examination further supported the strong therapeutic effect of the combined treatment. H&E staining revealed extensive structural destruction in tumor tissues, Ki67 staining showed markedly reduced proliferative activity, and TUNEL staining demonstrated increased apoptotic cell death in the nsPEF + ZNFM group (Fig. 5i). In addition, immunofluorescence staining of tumor tissues showed enhanced infiltration of CD3⁺CD8⁺ and CD3⁺CD4⁺ T cells after nsPEF+ZNFM treatment (Fig. 5j).

To investigate the immune activation effect of nsPEF + ZNFM, tumor-draining lymph nodes were collected from different treatment groups during the therapy period, and DC cell flow cytometry analysis was performed. nsPEF + ZNFM markedly promoted DC maturation, with CD80⁺CD86⁺ DCs reaching ∼36% (Fig. 5k), compared to 6% in the control group and 16% in the nsPEF group, suggesting enhanced antigen presentation capacity. Interestingly, intravenous administration of ZNFM facilitated DC activation in peripheral immune organs, such as lymph nodes, which then initiated T-cell responses. Accordingly, T-cell analysis revealed a substantial increase in CD8⁺ T cells to ∼30% (Fig. 5l), accompanied by an upward trend in CD4⁺ T cells to ∼9% (Fig. 5m). In contrast, CD8⁺ T cell activation in the control and nsPEF groups was approximately 5%, while CD4⁺ T cell activation was 4% and 5%, respectively. Notably, while nsPEF alone induced an increase in DC maturation, the immune response within the tumor region was partially suppressed, possibly due to nsPEF-associated immunosuppressive programs within the tumor microenvironment, which may limit sustained systemic immune activation and potentially contribute to tumor relapse. In contrast, the nsPEF + ZNFM group effectively overcame this issue, leading to enhanced tumor suppression and preventing relapse through strong systemic immunity. These in vivo results demonstrate that nsPEF-primed delivery effectively converts the cascade chemistry of ZNFM into heightened STING-centered immunogenic remodeling and superior primary tumor control with a favorable safety profile.

3.5. Contralateral rechallenge verifies systemic and durable antitumor immunity

Finally, to determine whether local nsPEF + ZNFM treatment could provide systemic protection beyond primary tumor control, we established a contralateral tumor rechallenge model to evaluate distant tumor inhibition and long-term immune efficacy (Fig. 6a). After resection of the primary tumor, Hepa1-6 cells were implanted on the opposite flank, and the growth of the rechallenged tumors was continuously monitored.

Fig. 6.

NsPEF-assisted ZNFM therapy elicits durable systemic antitumor immunity against tumor rechallenge. (a) Schematic illustration of the contralateral tumor rechallenge model following primary tumor resection. (b) Growth curves of rechallenged tumors in different treatment groups (n = 4). (c) Final weights of rechallenged tumors harvested from different groups (n = 4). (d) Representative photographs of excised rechallenged tumors from each group. (e) Kaplan-Meier survival curves of mice after tumor rechallenge (n = 6). (f, g) Serum cytokine levels of IFN-β (f) and IL-6 (g) measured after treatment (n = 3). (h, i) Representative flow cytometry plots and quantitative analysis of CD3⁺CD8⁺ T cells and CD3⁺CD4⁺ T cells in rechallenged tumors from different groups (n = 3). *P < 0.05, **P < 0.01.

As shown in Fig. 6b, tumors in the control group exhibited rapid progression, reaching an average volume of approximately 240 mm³ by day 14 after rechallenge. NsPEF monotherapy conferred only modest protection, with rechallenged tumors still growing to approximately 190 mm³, indicating that local ablation alone was insufficient to induce durable systemic immunity. In contrast, mice treated with nsPEF + ZNFM displayed pronounced suppression of distant tumor growth, with tumor volumes maintained at approximately 20 mm³ and remaining largely stable throughout the observation period, demonstrating effective inhibition of tumor outgrowth at the secondary site. Consistent with the growth-curve results, final tumor-weight analysis showed that the nsPEF + ZNFM group had the lowest tumor mass among all groups (Fig. 6c), and ex vivo photographs further confirmed the markedly reduced burden of rechallenged tumors in this group (Fig. 6d). These results demonstrate that nsPEF-assisted ZNFM therapy effectively inhibits tumor outgrowth at the secondary site.

This enhanced distant tumor control translated into a clear survival benefit. Mice receiving nsPEF + ZNFM showed the best survival outcome, with only 1 death observed up to day 60, whereas the nsPEF group exhibited 4 deaths during the same observation window, highlighting robust combination-induced systemic protection. Kaplan-Meier survival analysis further demonstrated a significant survival benefit in the nsPEF + ZNFM group (Fig. 6e).

To mechanistically link protection against rechallenge with immune activation, circulating inflammatory cytokines were quantified. Serum analysis revealed significantly elevated levels of IFN-β and IL-6 following nsPEF + ZNFM treatment (Fig. 6f and g), reaching approximately 85 pg/mL and 30 pg/mL, respectively. The robust induction of IFN-β indicates sustained innate immune signaling and is consistent with the reinforced STING activation observed in primary tumors. These findings support the notion that local nsPEF-assisted ZNFM therapy can drive systemic immunostimulatory signaling. We next analyzed immune-cell infiltration in the rechallenged tumors. Flow cytometric profiling showed that the proportion of CD8⁺ T cells was markedly increased in the nsPEF+ZNFM group (approximately 33% of total T cells) compared with the other treatment groups (Fig. 6h). Likewise, the percentage of CD4⁺ T cells was also elevated (approximately 12%) after combination treatment (Fig. 6i). These data indicate that combination therapy promotes durable T-cell mediated immune surveillance at distant tumor sites. The pronounced suppression of rechallenged tumors and prolonged survival indicate that nsPEF-assisted ZNFM therapy induces durable systemic antitumor immunity consistent with protective immunological memory.

3.6. In vivo hemocompatibility and systemic biosafety of ZNFM

To systematically evaluate the in vivo biosafety of ZNFM after intravenous administration, healthy mice were intravenously injected with ZNFM, and hemocompatibility, routine hematological parameters, serum biochemical indexes, and histopathological changes in major organs were examined after treatment (Fig. 7a).

Fig. 7.

In vivo hemocompatibility and systemic biosafety evaluation of ZNFM after intravenous administration. (a) Schematic illustration of intravenous ZNFM administration and subsequent systemic biosafety evaluation in mice. (b) Representative images and quantitative hemolysis rates of red blood cells incubated with ZNFM at different concentrations. (c–f) Routine hematological analysis, including red blood cell count (RBC, c), white blood cell count (WBC, d), platelet count (PLT, e), and hemoglobin level (HGB, f), in mice after ZNFM treatment for 3 or 7 days (n = 3). (g–l) Serum biochemistry analysis, including alanine aminotransferase (ALT, g), aspartate aminotransferase (AST, h), blood urea nitrogen (BUN, i), creatinine (CREA, j), cholinesterase (CHE, k), and albumin (ALB, l), in mice after ZNFM treatment for 3 or 7 days (n = 3). (m) Representative H&E-stained images of major organs, including the heart, liver, spleen, lung, and kidney, collected from control mice and ZNFM-treated mice at days 3 and 7. Scale bars, 100 μm ns, not significant; **P < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

The hemocompatibility of ZNFM was then evaluated using a hemolysis assay. As shown in Fig. 7b, ZNFM caused negligible hemolysis over the tested concentration range of 30–800 μg/mL, with hemolysis rates remaining far below the positive control group. This result demonstrates that ZNFM possesses favorable blood compatibility and is suitable for intravenous administration. The systemic biosafety of ZNFM was further examined by routine hematological analysis. As shown in Fig. 7c–f, no significant changes were observed in red blood cells, white blood cells, platelets, or hemoglobin at either day 3 or day 7 after ZNFM administration compared with the control group, suggesting negligible hematological toxicity. Consistently, serum biochemical analysis revealed no significant alterations in alanine aminotransferase, aspartate aminotransferase, blood urea nitrogen, creatinine, cholinesterase, or albumin among the groups (Fig. 7g–l), indicating that ZNFM treatment did not induce appreciable hepatic or renal dysfunction under the tested conditions. Histopathological examination further supported the favorable biosafety profile of the formulation. H&E staining of the heart, liver, spleen, lung, and kidney showed no obvious inflammatory infiltration, tissue necrosis, structural disruption, or pathological lesions at either day 3 or day 7 after ZNFM administration (Fig. 7m). Together with the hematological and biochemical data, these observations demonstrate that systemic administration of ZNFM is well tolerated in vivo.

4. Conclusion

In summary, we engineered a tumor microenvironment-responsive core-shell cascade nanoplatform, ZNFM, to transform nsPEF-mediated electroporation into amplified STING-driven chemoimmunotherapy against HCC. The therapeutic synergy between nsPEF and ZNFM follows a sequential physical–chemical–immunological mechanism. First, nsPEF functions as a physical delivery primer that improves intratumoral deposition and intracellular accumulation of ZNFM. Second, the internalized nanoplatform undergoes tumor-specific cascade activation: the degradable ZnO₂ core releases Zn²⁺ and supplies H₂O₂, while the pH-labile NiFc shell catalyzes ROS generation, thereby reinforcing tumor cell killing and ICD induction. Third, the esterase-activatable MSCU prodrug co-releases MSA-2 and curcumin, enabling Zn²⁺/MSA-2-mediated activation of the STING-TBK1-IRF3 axis and curcumin-mediated suppression of the IDO-KYN immunosuppressive pathway. These integrated material functions promote dendritic cell maturation, enhance antitumor T-cell responses, and establish systemic immune protection against tumor rechallenge. Overall, this study presents an electroporation-enabled cascade biomaterial strategy that couples local physical ablation with programmable nanocatalytic and immunomodulatory outputs, offering a promising approach for durable HCC chemoimmunotherapy.

Abbreviations

HCC	Hepatocellular carcinoma
nsPEF	Nanosecond pulsed electric field
TME	Tumor microenvironment
ZNFM	ZnO2-NiFc@MSCU
ZNF	ZnO2/NiFc
MSCU	MSA-2-curcumin prodrug
ROS	Reactive oxygen species
ICD	Immunogenic cell death
DCs	Dendritic cells
BMDCs	Bone marrow-derived dendritic cells
IDO	Indoleamine 2,3-dioxygenase
KYN	Kynurenine
DEGs	Differentially expressed genes
GO	Gene Ontology
KEGG	Kyoto Encyclopedia of Genes and Genomes
H&E	Hematoxylin and eosin
ANOVA	Analysis of variance

Ethics approval statement

All animal procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals of Zhejiang University and were approved by the Animal Ethics Committee of Zhejiang University (Approval number: ZJU20250769).

CRediT authorship contribution statement

Jiaxin Zhao: Data curation, Formal analysis, Investigation, Validation, Visualization, Writing – original draft, Writing – review & editing. Zheyu Fang: Data curation, Formal analysis, Investigation, Validation, Visualization. Ziang Luo: Data curation, Formal analysis, Investigation, Visualization. Jingye Yang: Validation, Visualization, Writing – original draft. Sakarie Mustafe Hidig: Formal analysis. Jie Zhou: Formal analysis. Fangshen Xu: Formal analysis. Huilin Zheng: Formal analysis. Lei Zhang: Formal analysis. Shengyong Yin: Conceptualization, Methodology, Supervision, Visualization, Writing – original draft, Writing – review & editing. Xiang Li: Conceptualization, Methodology, Supervision, Visualization, Writing – original draft, Writing – review & editing. Zhenhua Hu: Conceptualization, Funding acquisition, Methodology, Supervision, Visualization, Writing – original draft, Writing – review & editing.

Declaration of competing interests

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.