Skip to main content
Journal of Pharmaceutical Analysis logoLink to Journal of Pharmaceutical Analysis
. 2026 Apr 13;16(10):101633. doi: 10.1016/j.jpha.2026.101633

Orchestrating enhanced starvation and ROS storm therapy with MnO2 nanozyme-driven cascade nanoreactors

Jiahui Cai a, Zutong Cui a, Wenming Liang a, Jianming Sun a, Zihan Zhang a, Feifei Sun a, Feng Yang a, Zhiwei Liu a, Liming Gao b, Jidong Wang a, Jian Li a,⁎
PMCID: PMC13742849  PMID: 42997822

Abstract

The therapeutic efficacy of glucose oxidase (GOx)-mediated starvation therapy and β-Lapachone (LPC)-induced oxidative stress is significantly constrained by tumor hypoxia and elevated antioxidant defenses in the tumor microenvironment (TME). To address these limitations, a multifunctional nanoplatform, LPC/GOx@ZIF-8@MnO2 (LGZM), was developed through the integration of LPC and GOx within a zeolitic imidazolate framework-8 (ZIF-8) carrier followed by surface modification with manganese dioxide (MnO2), and its antitumor efficacy and mechanistic basis were systematically evaluated using HeLa cells in vitro and U14 tumor-bearing mouse models in vivo. Comprehensive in vitro and in vivo studies demonstrated significant tumor growth inhibition with excellent biosafety, while mechanistic investigations revealed that this nanocomposite achieves synergistic antitumor effects through four coordinated mechanisms: MnO2-catalyzed conversion of endogenous hydrogen peroxide to oxygen alleviates hypoxia while sustaining GOx-mediated glucose depletion; a self-amplifying cascade reaction enhances LPC-induced reactive oxygen species (ROS) generation; photothermally activated upregulation of nicotinamide adenine dinucleotide (phosphate) (NAD(P)H):quinone oxidoreductase 1 (NQO1) via the heat shock protein 70 (HSP70)/NQO1 axis enhances LPC efficacy; and depletion of glutathione (GSH) disrupts cellular antioxidant defenses. Furthermore, it was also revealed that LGZM acts on caspase-3, high mobility group box 1 (HMGB1), and matrix metalloproteinase-1 (MMP-1) to coordinate apoptosis induction and metastasis suppression, with molecular docking further clarifying the key binding sites of LPC with these three targets. Collectively, this work presents an innovative strategy to overcome therapeutic limitations imposed by tumor hypoxia, provides molecular-level insights into multimodal synergistic therapy, and offers new ideas for precision tumor therapy.

Keywords: Starvation therapy, ROS storm, NQO1, MnO2 nanozyme, Tumor hypoxia alleviation

Graphical abstract

Image 1

Highlights

  • •

    Pioneered a hypoxia-modulating nanoplatform enabling self-amplifying cascade therapy.

  • •

    Quadruple synergy mechanism overcomes tumor antioxidant defenses for oxidative stress therapy.

  • •

    MnO2 dual-function enables novel photothermal-chemodynamic synergy.

  • •

    Multi-omics reveals β-lapachone's multi-target mechanisms.

1. Introduction

The rapid proliferation of tumor cells is critically dependent on endogenous glucose uptake through the glycolytic pathway [1]. Targeting this metabolic vulnerability by disrupting glucose supply to cancer cells has emerged as a promising therapeutic strategy for inhibiting tumor growth. In particular, glucose oxidase (GOx)-mediated starvation therapy has demonstrated significant potential due to its ability to deplete intratumoral glucose [[2], [3], [4]]. Under aerobic conditions, GOx catalyzes the oxidation of glucose, generating cytotoxic hydrogen peroxide (H2O2) and gluconic acid as byproducts [5,6]. However, the hypoxic nature of the tumor microenvironment (TME) poses a major limitation to the efficacy of GOx-mediated therapy [7,8]. This challenge is further exacerbated by the oxygen consumption inherent to the GOx-catalyzed reaction, creating a detrimental feedback loop that intensifies hypoxia and impairs therapeutic outcomes [9,10].

β-Lapachone (LPC) represents another class of TME-dependent anticancer agents, which exerts its therapeutic effect through a unique nicotinamide adenine dinucleotide (phosphate) (NAD(P)H):quinone oxidoreductase-1 (NQO1)-mediated redox cycling process [11]. NQO1 is overexpressed in many types of malignant cells, with levels up to 100-fold higher than those in normal cells, endowing LPC with remarkable tumor selectivity [12]. NQO1 reduces LPC to its hydroquinone form, which subsequently undergoes auto-oxidation back to the parent quinone in an oxygen-dependent manner, generating substantial amounts of superoxide anion (O2−) and H2O2 in the process [13,14]. However, this efficient reactive oxygen species (ROS)-generating process of LPC is also strictly oxygen-dependent. Consequently, hypoxia in the TME severely restricts its therapeutic efficacy. It is thus evident that hypoxia represents a common and critical bottleneck for both GOx-based starvation therapy and LPC-induced oxidative stress therapy. Overcoming this limitation is essential for achieving effective combination therapy.

Fortunately, the TME itself provides a clue for addressing this challenge: it exhibits significantly elevated levels of H2O2 compared to normal tissues [15]. This observation has motivated researchers to exploit the high reactivity of manganese dioxide (MnO2) with H2O2, catalyzing its decomposition into oxygen as an effective strategy to alleviate tumor hypoxia [[16], [17], [18], [19]]. Building on this principle, various nanosystems have been developed to ameliorate hypoxia. For instance, cascade systems integrating GOx and catalase (CAT) have been constructed [[20], [21], [22]], where CAT decomposes the H2O2 produced by GOx to generate oxygen in situ. However, such enzymatic systems often face challenges including poor in vivo stability, therapeutic efficacy overly reliant on endogenous H2O2 levels, and a relatively monotherapeutic mode. On the other hand, various functional materials have been extensively explored. Conventional MnO2 nanomaterials, while effective in generating oxygen or depleting glutathione (GSH), typically exhibit limited drug-loading capacity and functional singularity [16]. In contrast, nanocarriers like zeolitic imidazolate framework-8 (ZIF-8) possess excellent drug-loading capacity and pH-responsive properties but inherently lack the ability to modulate the TME [23].

To address these limitations, we constructed a multifunctional nanoplatform, designated as LPC/GOx@ZIF-8@MnO2 (LGZM) (Scheme 1A), aiming to achieve synergistically enhanced therapeutic outcomes. The innovation of this design is reflected in several key aspects: Firstly, leveraging the high drug-loading capacity and pH-responsive properties of ZIF-8 [[23], [24], [25]], we co-loaded GOx and LPC, ensuring the stability of the active components in the physiological environment and their targeted release at the tumor site. Secondly, by coating the surface with an MnO2 shell [26], we established a self-sustaining oxygen supply cycle: the H2O2 produced by the GOx-catalyzed consumption of glucose and oxygen is decomposed by MnO2 to generate oxygen, which in turn continuously supports the catalytic activity of GOx and the redox cycling of LPC. More importantly, the MnO2 component serves multiple functions herein, acting not only as an oxygen generator but also exhibiting GOx-mimetic activity [16,27], GSH-scavenging capability [28], and photothermal conversion properties [[29], [30], [31]]. This integration enables the deep fusion of starvation therapy, oxidative stress therapy, and photothermal therapy (PTT).

Scheme 1.

Scheme 1

Schematic illustration of the synthesis of LPC/GOx@ZIF-8@MnO2 (LGZM) and utilizing manganese dioxide (MnO2) to enhance starvation therapy and oxidative stress therapy. (A) Schematic illustration of the preparation route of LGZM. (B) Schematic illustration of the therapeutic mechanism of LGZM for enhanced starvation therapy and oxidative stress therapy. LPC: β-Lapachone; GOx: glucose oxidase; ZIF-8: zeolitic imidazolate framework-8; NIR: near-infrared; ROS: reactive oxygen species; HIF-1α: hypoxia-inducible factor 1-alpha; NQO1: NAD(P)H:quinone oxidoreductase 1; GSH: reduced glutathione; GSSG: oxidized glutathione.

To validate the design concept, we investigated the in vivo behavior and therapeutic mechanisms of the platform. Upon intravenous administration, the nanoplatform, owing to its optimized nanosize, preferentially accumulated at tumor sites via the enhanced permeability and retention (EPR) effect [4,32]. Subsequently, the acidic TME triggered the rapid degradation of the ZIF-8 framework, leading to the release of the encapsulated GOx and LPC. GOx catalyzes the oxidation of endogenous glucose, producing gluconic acid and H2O2. The latter is subsequently decomposed by the outer MnO2 shell into oxygen, thereby realizing the aforementioned self-sustaining oxygen supply cycle and continuously enhancing GOx-mediated glucose deprivation. It is noteworthy that MnO2 itself also exhibits GOx-like activity and can directly participate in catalyzing glucose oxidation [16,27], further intensifying the starvation therapy. On the basis of restored oxygen supply, the released LPC is bioactivated by the highly expressed NQO1 in tumor cells, undergoing an oxygen-dependent redox cycling process that triggers a massive burst of ROS and depletion of ATP [12]. This subsequently modulates key signaling pathways including P-glycoprotein (P-gp), nuclear factor kappa-B (NF-κB), hypoxia-inducible factor 1-alpha (HIF-1α), and caspase activation ultimately inducing tumor cell apoptosis [12,33]. Therefore, the incorporation of MnO2 not only fundamentally alleviates the oxygen-dependent limitations of GOx and LPC but also establishes a multi-layered, self-amplifying synergistic therapeutic system (Scheme 1B). Furthermore, the MnO2 component possesses significant photothermal conversion capability, enabling localized PTT under near-infrared (NIR) laser irradiation [[29], [30], [31]]. The induced thermal effect not only directly kills tumor cells but also upregulates the expression of heat shock proteins (HSPs), particularly HSP70 [34,35]. HSP70, in turn, further stabilizes and enhances the protein level of NQO1 [36,37], forming a positive feedback loop that strengthens LPC redox cycling and substantially improves ROS generation efficiency [38,39]. Concurrently, the continuous consumption of GSH by MnO2 effectively weakens the antioxidant defense capacity of tumor cells, further augmenting ROS-mediated cytotoxicity [16,40].

In summary, the core innovation of this study lies in the rational design of the core-shell structured nanoplatform LGZM. This platform not only establishes a self-sustaining chemical cascade to enable continuous supply of oxygen and ROS but, more critically, it creates a mutually enhancing synergy between two originally independent modalities: the photothermal effect and biochemical drug efficacy. Specifically, the photothermal effect upregulates NQO1 expression via the HSP70/NQO1 signaling axis, thereby proactively amplifying LPC's redox cycling and the subsequent ROS burst. This design achieves deep integration and synergistic amplification of oxidative stress therapy, starvation therapy, and PTT at the mechanistic level, transcending the conventional approach of simple functional combination and resulting in highly effective synergistic antitumor activity. In vivo results demonstrate that this platform exhibits outstanding tumor growth inhibition and favorable biosafety over a 15-day treatment period. Further mechanistic investigations at the molecular level elucidate its critical action pathways, including modulation of the HSP70/NQO1 axis, depletion of GSH, and activation of the caspase cascade. These findings provide a new paradigm and a solid theoretical and experimental foundation for the development of highly effective and low-toxicity precision cancer therapeutics.

2. Materials and methods

2.1. Materials

2-Methylimidazole (M915172), β-Lapachone (L872641), and glucose-oxidase (G810485) were purchased from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China). Zn(NO3)2·6H2O (Z111706) and KMnO4 (P399447) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). All chemical reagents utilized were analytical grade.

The cell experiment kits were provided by Beyotime Biotechnology Co., Ltd. (Shanghai, China). Glucose assay kits were supplied by Sigma-Aldrich (St. Louis, MI, USA), and H2O2 assay kits were sourced from Sangon Biotech Co., Ltd. (Shanghai, China). Assay kits for GSH, superoxide dismutase (SOD), malondialdehyde (MDA), oxidized glutathione (GSSG), alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and creatinine (CRE) were obtained from Nanjing Jiancheng Biological Engineering Institute (Nanjing, China). Recombinant NADH dehydrogenase, quinone 1 (NQO1) was supplied by Shanghai Zeye Biotechnology Co., Ltd. (Shanghai, China).

Antibodies targeting specific proteins were acquired as follows: HIF-1α (sc-13515), NQO1 (sc-393736), HSP70 (sc-32239), caspase-3 (sc-65497), matrix metalloproteinase-1 (MMP-1; sc-21731), high mobility group box 1 (HMGB1; sc-56698), and β-actin (sc-47778) from Santa Cruz Biotechnology (Shanghai) Co., Ltd. (Shanghai, China); phosphorylated extracellular signal-regulated kinase 1/2 (p-ERK1/2; #4370S) and total extracellular signal-regulated kinase 1/2 (t-ERK1/2; #9102S) from Cell Signaling Technology, Inc. (Danvers, MA, USA).

2.2. Cell lines and mice

HeLa and U14 cells were sourced from the National Collection of Authenticated Cell Cultures (Shanghai, China). Female Kunming mice (4 weeks old, 20 ± 2 g, specific pathogen-free (SPF) grade) used for animal experiments were supplied by Beijing Charles River Laboratory Animal Technology Co. Ltd. (Beijing, China).

2.3. Synthesis of LGZM

2.3.1. Synthesis of ZIF-8

A solution of zinc nitrate hexahydrate (Zn(NO3)2·6H2O, 0.6 g) in methanol (10 mL) and a solution of 2-methylimidazole (1.6 g) in methanol (10 mL) were prepared separately. These solutions were then rapidly combined and vigorously stirred to initiate the reaction. The mixture was maintained at 25 °C under continuous stirring for 1 h (1000 rpm). Upon reaction completion, the resulting white precipitate was isolated by centrifugation (10,000 rpm, 10 min). The solid product was subsequently washed three times with methanol to remove residual reactants and soluble by-products. Finally, the purified material was vacuum-dried at 60 °C to yield ZIF-8 as a white powder.

2.3.2. Loading of LPC and GOx

A dispersion was prepared by adding 10 mg of ZIF-8 powder to 10 mL of a 0.3 mg/mL LPC solution. This mixture was subjected to rotary incubation at 25 °C for 24 h (500 rpm). Following incubation, LPC@ZIF-8 nanoparticles were separated by centrifugation (12,000 rpm, 10 min). The supernatant was collected for subsequent quantification of unencapsulated LPC, enabling calculation of the encapsulation efficiency (EE%) and drug loading capacity (DL%) using standard (1), (2).

Following LPC encapsulation, 10 mg of GOx was introduced into the LPC@ZIF-8 suspension and the mixture underwent continuous stirring for 24 h at ambient temperature (500 rpm). The resultant GOx-conjugated nanoparticles (LPC/GOx@ZIF-8 (LGZ)) were isolated via centrifugation (12,000 rpm, 10 min). The encapsulation efficiency and loading capacity of GOx were subsequently determined by quantifying the concentration of unencapsulated GOx in the supernatant using the Bradford protein assay.

EE%=(mi−mu)mi×100% (1)
DL%=(mi−mu)mt×100% (2)

The initial input mass of LPC (or GOx) is defined as mi, the mass of unencapsulated LPC (or GOx) in the supernatant as mu, and the total mass of the drug delivery system as mt.

The presence of successfully loaded GOx was demonstrated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the nanoparticle fraction after thermal denaturation in loading buffer. Briefly, the nanoparticles were denatured in 4× loading buffer containing 5% β-mercaptoethanol at 95 °C for 5 min to dissociate the carrier structure and release the encapsulated proteins prior to electrophoresis. For comparison, standard GOx (positive control) and the empty ZIF-8@MnO2 carrier (negative control) were processed in parallel under identical conditions.

2.3.3. Modification of MnO2 on LGZ

After the successful preparation of LGZ, a total of 20 mg of KMnO4 was introduced into the LGZ suspension under vigorous stirring (1000 rpm). The mixture was agitated continuously until a characteristic yellowish-brown hue developed, indicating MnO2 formation. The resulting composite nanoparticles LGZM were isolated by centrifugation (14,000 rpm, 15 min), washed thoroughly with deionized water (3 ×), and vacuum-dried to yield the final product.

In selected syntheses, the nano-drug delivery system was prepared with Cy5-labeled LPC for drug tracking, denoted LPC(Cy5)/GOx@ZIF-8@MnO2 (Cy5/LGZM).

2.4. Characterization of LGZM

The nanoparticles’ morphology was characterized by Hitachi HT-7700 transmission electron microscopy (TEM; Hitachi, Tokyo, Japan) and Hitachi SU8230 scanning electron microscopy (SEM; Hitachi, Tokyo, Japan). Hydrodynamic diameter and zeta potential measurements were performed using a Malvern Zetasizer Nano ZS instrument (Malvern, Worcestershire, UK). Ultraviolet-visible (UV-Vis) absorption spectra of aqueous dispersions were acquired with a Shimadzu UV-2550 spectrophotometer (Shimadzu, Kyoto, Japan). The chemical composition and crystal structure of the synthesized nanomaterials were characterized using Thermo Scientific ESCALAB QXi2 X-ray photoelectron spectroscopy (XPS; Thermo Scientific, Waltham, MA, USA) and Rigaku SmartLab X-ray diffraction (XRD; Rigaku Corporation, Tokyo, Japan).

2.5. In vitro performance measurement

The stability, photothermal effect, ROS generation, drug release performance, enzymatic activity, anti-serum protein adsorption capacity, and hemocompatibility of LGZM nanoparticles were evaluated in vitro. Detailed experimental procedures are provided in the Supplementary data 1.1.

2.6. Cell assays

The cellular uptake, cytotoxicity, intracellular ROS levels, mitochondrial membrane potential (ΔΨm), intracellular oxygen production capacity, intracellular H2O2 levels, cell thermal imaging, colony formation, invasion, migration, GSH, MDA, and GSSG levels, as well as tumor penetration ability of LGZM nanoparticles were systematically evaluated using HeLa cells. Detailed experimental procedures are provided in the Supporting data 1.2.

2.7. Animal assays

The in vivo photothermal conversion performance, tumor hypoxia relief performance, pharmacokinetic profiles, biodistribution characteristics, tumor growth and metastasis inhibition efficacy, and biosafety of LGZM were evaluated using a U14 tumor-bearing mouse model. To further investigate the underlying antitumor mechanisms, tumor tissues were fixed in 4% paraformaldehyde and subsequently subjected to hematoxylin and eosin (H&E) staining, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, and Ki67 expression analysis. Additionally, immunohistochemical analysis and Western blotting were performed to assess the expression of relevant proteins in tumor tissues. Detailed experimental procedures are provided in the Supporting data 1.3.

All animal experiments were conducted at the Animal Research Center of Yanshan University (Qinhuangdao, China). The experimental protocols were approved by the Institutional Animal Care and Use Committee of Yanshan University (Approval No.: YD202417), and all procedures adhered to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health.

2.8. Molecular docking

β-Lapachone from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP) were obtained by the mol2 file of, then its structure was optimized using Chem3D 22.0.0. Simultaneously, the 3D PDB structures of HMGB1, MMP-1, and caspase-3 were retrieved from the RCSB PDB database, ensuring they are X-ray crystallography structures with a resolution better than 2.5 Å. PyMOL was used to remove water molecules and small ligands, and subsequently the proteins were processed in AutoDock Tools 1.5.7 by adding hydrogens, assigning charges, and merging nonpolar hydrogens. Finally, molecular docking was performed using AutoDock Vina 1.1.2, then the results were evaluated by sorting receptor-ligand pairs based on binding free energy (affinity, kcal/mol). A binding energy lower than −5 kcal/mol indicated good binding, while values below −7 kcal/mol suggested strong activity. Finally, the docking poses were visualized in PyMOL.

2.9. Bioinformatics analysis

Mouse cervical cancer (U14 Cell Line) is a mouse-derived squamous cell carcinoma cell line of the uterine cervix. We employed an integrated multi-database strategy to systematically screen potential targets of LPC against squamous cell carcinoma. Initially, the Simplified Molecular Input Line Entry System (SMILES) structure of LPC was retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). Drug-like components were then screened using SwissADME (http://www.swissadme.ch/), followed by target prediction through Swiss Target Prediction (http://www.swisstargetprediction.ch/). Concurrently, squamous cell carcinoma-related genes were obtained from GeneCards and OMIM databases, with common targets identified through Venn diagram analysis using Venny 2.1.0.

Subsequently, a protein-protein interaction (PPI) network was constructed using the STRING database (https://cn.string-db.org/), followed by topological analysis with Cytoscape 3.9.1. Based on the degrees, betweenness, and closeness centrality measures, we identified the top 7 core targets. Further functional enrichment analysis was performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID; https://david.ncifcrf.gov/) to identify Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (P < 0.05). The results were visualized using the ggplot2 package in R software.

2.10. Statistical analysis

The quantitative data are presented as mean ± standard deviation (mean ± SD). Statistical comparisons between two groups were performed using two-tailed unpaired Student's t-tests, while multiple group comparisons were analyzed by one-way analysis of variance (ANOVA) followed by Tukey's post-hoc test using GraphPad Prism (GraphPad Prism 9.0 Software). Statistical significance for specific longitudinal datasets, including LPC release profiles, murine tumor photothermal heating curves, and tumor growth kinetics, was determined by comparing area under the curve (AUC) values across groups using one-way ANOVA with Tukey's post-hoc test. The significance levels were defined as follows: ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001; ns, not significant.

3. Results

3.1. Synthesis and characterization of LGZM

The fabrication process of the LGZM nanoplatform was illustrated in Scheme 1A. Initially, ZIF-8 nanoparticles were synthesized via a co-precipitation method. Subsequently, a MnO2 nanolayer was deposited on the ZIF-8 surface through the decomposition of KMnO4. Morphological characterization by TEM and SEM revealed that the pristine ZIF-8 particles exhibited a characteristic rhombic dodecahedral morphology (Figs. 1A and B), while the MnO2-coated ZIF-8 (ZIF-8@MnO2) maintained a spherical structure (Figs. 1C and D). Furthermore, elemental mapping analysis via SEM confirmed the co-existence and homogeneous distribution of both Zn and Mn elements within the LGZM structure (Fig. S1).

Fig. 1.

Fig. 1

Characterizations of LPC/GOx@ZIF-8@MnO2 (LGZM). (A, C) Transmission electron microscopy (TEM) images and hydrodynamic diameter distribution of zeolitic imidazolate framework-8 (ZIF-8) (A) and ZIF-8@MnO2 (C), respectively. (B, D) Scanning electron microscopy (SEM) images of ZIF-8 (B) and ZIF-8@MnO2 (D), respectively. (E) High-resolution X-ray photoelectron spectroscopy (XPS) survey spectra of ZIF-8 and ZIF-8@MnO2. (F) Magnified view of the Mn 2p region (635–660 eV) in ZIF-8@MnO2. (G) High-resolution Mn 2p spectrum of ZIF-8@MnO2, showing the Mn 2p1/2 and Mn 2p3/2 peaks. (H) X-ray diffraction (XRD) patterns of ZIF-8 and ZIF-8@MnO2. (I, J) Hydrodynamic sizes (I), and zeta potentials (J) of ZIF-8, LPC@ZIF-8 (LZ), LPC/GOx@ZIF-8 (LGZ), and LGZM nanoparticles. (K) Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis confirming the successful loading of glucose oxidase (GOx) in the LGZM nanoplatform. (L) Ultraviolet-visible (UV-Vis) absorption spectra of the synthesized nanoparticles. Sample concentrations: ZIF-8 (1 mg/mL), GOx (0.59 mg/mL), ZIF-8@MnO2 (2.35 mg/mL), β-Lapachone (LPC; 0.25 mg/mL), LGZM (3.20 mg/mL). (M) Hydrodynamic size and polydispersity index (PDI) changes over 5 days for LGZM. Data presented as mean ± standard deviation (SD), n = 3 in I, J, and M.

High-resolution XPS analysis provided key evidence for the successful coating of MnO2. Compared with pure ZIF-8, the XPS survey scan of ZIF-8@MnO2 revealed distinct new peaks in the 635–660 eV region, which are ascribed to the Mn 2p orbitals (Figs. 1E and F). High-resolution spectra of the Mn 2p region showed a characteristic doublet with binding energies of 641.8 eV for Mn 2p3/2 and 654.7 eV for Mn 2p1/2 (Fig. 1G), confirming the presence of Mn4+ [41]. This result verifies the successful modification of ZIF-8 with manganese dioxide, demonstrating the successful formation of the ZIF-8@MnO2 composite. To further elucidate the crystal structure of the ZIF-8@MnO2 nanoparticles, XRD was performed. As shown in Fig. 1H, the composite material displays a series of diffraction peaks at 12.7°, 18.2°, 28.1°, 37.6°, 49.9°, and 60.2°, which align well with the characteristic reflections of the (110), (200), (310), (211), (411), and (521) crystal planes of α-phase MnO2 (standard reference ICDD#04-005-4884) [42], indicating a well-defined crystalline structure in the outer MnO2 shell. It is worth noting that the diffraction peaks of ZIF-8@MnO2 are significantly broadened compared to typical sharp reflections, suggesting nanoscale crystallite dimensions and low crystallinity, likely resulting from the nano-sized effects of the MnO2 shell and the epitaxial growth on the ZIF-8 surface. These results confirm the successful coating of MnO2 on ZIF-8, and the unique crystal structure combined with nano-effects are expected to enhance the surface activity and reactivity of the composite material.

The successful stepwise assembly of the nanoplatform was confirmed by a combination of dynamic light scattering (DLS) and TEM. Bare ZIF-8 nanoparticles exhibited a hydrodynamic diameter of 68.67 ± 4.24 nm by DLS (Fig. 1I), and appeared as uniform, well-defined rhombic dodecahedra as revealed by TEM (Fig. 1A). Loading with LPC (LPC@ZIF-8) increased the diameter to 80.43 ± 3.89 nm (Fig. 1I), and TEM imaging revealed the appearance of light-contrast patches on the particle surface (Fig. S2A), indicating successful drug encapsulation. Subsequent encapsulation of GOx (LGZ) further expanded the size to 91.20 ± 4.12 nm (Fig. 1I), accompanied by a morphological transition to a more spherical shape with a fragmented, thin surface layer under TEM (Fig. S2B). Finally, coating with MnO2 (LGZM) resulted in the largest particles at 128.80 ± 6.47 nm (Fig. 1I), which corresponded to the spherical particles with a rough surface texture visible in TEM images (Fig. S2C). This stepwise size growth was corroborated by corresponding changes in surface charge, measured via ζ-potential. The positively charged ZIF-8 (+29.03 ± 3.36 mV) exhibited a slight decrease upon LPC loading (+25.53 ± 2.53 mV). A more significant reduction occurred after GOx encapsulation (+16.17 ± 1.56 mV), and crucially, the final MnO2 coating reversed the surface charge, yielding a negative ζ-potential of −11.27 ± 1.99 mV (Fig. 1J). The systematic changes in nanoparticle size, surface potential, and the distinct morphological evolution collectively provide conclusive evidence for the successful stepwise assembly and surface modifications of the nanoplatform.

The successful loading of both GOx and LPC was further verified through SDS-PAGE assay and UV-Vis spectroscopy, respectively. SDS-PAGE analysis confirmed the presence of GOx in the LGZM nanoparticles, as evidenced by a characteristic band at ∼80 kDa that aligned with the free GOx standard. (Fig. 1K). UV-Vis spectral analysis demonstrated characteristic absorption peaks in the LGZM solution that matched the LPC standard spectrum, confirming effective drug incorporation (Fig. 1L). The drug loading (DL%) and encapsulation efficiency (EE%) of the two active components were quantitatively determined. For LPC, the DL% was 7.86% ± 1.24% with an EE% of 83.69% ± 4.84%. For GOx, the loading capacity was 18.5% ± 2.63% with an EE% of 59.09% ± 5.52%. These comprehensive characterization results collectively confirm the successful synthesis of the LGZM nanoplatform.

A 5-day stability assessment showed that the hydrodynamic diameter and polydispersity index (PDI) of LGZM remained stable in phosphate buffered saline (PBS), with all PDI values below 0.3 (Fig. 1M), preliminarily demonstrating its good dispersibility and stability. To better simulate physiological conditions, the samples were incubated in PBS (37 °C), 10% fetal bovine serum (FBS)/PBS (room temperature), and 10% FBS/PBS (37 °C) for 5 days, with daily monitoring of their particle size and PDI. The results indicated that LGZM maintained good dispersion and stability under all tested conditions, with no significant changes in hydrodynamic size or PDI throughout the observation period (Fig. S3). This confirms that the nanocomposite exhibits excellent stability in physiologically relevant media containing serum, supporting its suitability for subsequent biological studies.

3.2. Property evaluation of LGZM

Previous studies have demonstrated that modified MnO2 nanostructures exhibited enhanced photothermal performance [16,43]. Building upon this foundation, and considering the critical role of photothermal effects in modulating drug release kinetics and therapeutic synergy, we systematically evaluated the photothermal characteristics of both MnO2 and LGZM in vitro. Under 808 nm laser irradiation, LPC/GOx@ZIF-8@MnO2 demonstrated concentration- and power-dependent temperature elevations (Figs. 2A and S4). The temperature rise curves and thermal images showed that a 2 mg/mL nanoparticle solution achieved a temperature increase from 24.7 °C to 49.4 °C within 10 min at 2.5 W/cm2, surpassing the threshold for thermal ablation of tumor cells (43 °C). Control experiments with MnO2 alone also showed photothermal activity, reaching 46.1 °C under identical conditions (Figs. 2B and S5). The photothermal conversion efficiency of LGZM was calculated to be 37.09% (Fig. 2C), confirming the significant contribution of the MnO2 coating. Furthermore, the composite exhibited excellent photostability, with minimal temperature variation (ΔT < 1.5 °C) across four laser “on/off” cycles (Fig. 2D). These results collectively establish LGZM as a promising candidate for combinational PTT applications.

Fig. 2.

Fig. 2

In vitro performance evaluation of LPC/GOx@ZIF-8@MnO2 (LGZM). (A) Photothermal heating curve of LGZM solution at different concentrations after exposure to 808 nm near-infrared (NIR) laser (2.5 W/cm2). (B) Photothermal heating curve of PBS, MnO2, and LGZM after exposure to 808 nm NIR laser (2.5 W/cm2). (C) Photothermal conversion efficiency (η) of LGZM under 808 nm NIR irradiation (2.5 W/cm2). (D) The LGZM maintained consistent temperature profiles throughout four successive heating-cooling cycles. (E) Singlet oxygen generation of different nanoparticles evaluated by 1,3-diphenylisobenzofuran (DPBF). (F) Ultraviolet-visible (UV-Vis) absorption spectra of methylene blue (MB) + H2O2 after different treatments. (G–J) Glucose decomposition induced by different nanoparticles: effects on glucose consumption (G), O2 content (H), H2O2 generation (I), and pH (J). (K, L) In vitro β-Lapachone (LPC) release rate of LGZM under pH 7.4, 5.0, and 5.0 + Laser without (K) or with glucose (L). Data presented as mean ± standard deviation (SD), n = 3 in B, E, G–L. ∗P < 0.05, and ∗∗∗∗P < 0.0001. ZIF-8: zeolitic imidazolate framework-8; NQO1: NAD(P)H:quinone oxidoreductase 1; LGZ: LPC/GOx@ZIF-8; Glu: glucose.

The ROS production capability of LGZM was quantitatively assessed using 1,3-diphenylisobenzofuran (DPBF). In the presence of recombinant NQO1, LPC underwent redox cycling to generate ROS, evidenced by a 47.5% decrease in DPBF fluorescence. The MnO2 shell, leveraging its peroxidase-like activity, catalyzed the decomposition of H2O2 into oxygen, which significantly amplified the cascade generation of ROS. This catalytic role was confirmed by a marked enhancement in the DPBF quenching rate to 67.5% after MnO2 modification. Furthermore, under NIR laser irradiation, the composite's prominent photothermal effect enhanced reaction kinetics and molecular diffusion, elevating the DPBF quenching rate further to 72.4% (Fig. 2E).

To further delineate the nature of the generated ROS, we employed a methylene blue (MB) degradation assay, which is highly indicative of hydroxyl radical (•OH) formation. The assay confirmed that the NQO1-mediated “futile cycle” of LPC, which generates substantial fluxes of H2O2 and O2−, creates a highly oxidizing environment conducive to •OH formation. This was evidenced by a 52.58% decrease in the characteristic MB absorption at 665 nm upon NQO1-driven activation of LPC. Furthermore, we demonstrated that the MnO2-induced ROS cascade significantly contributes to •OH production (Fig. 2F).

The enzymatic performance of LGZM was systematically evaluated through dynamic monitoring of glucose consumption, H2O2 generation, oxygen release, and pH variation during glucose oxidation. Comparative analysis revealed significantly enhanced glucose oxidation capacity in LGZM compared to LGZ (Fig. 2G). This enhancement is attributed to the multifunctional role of the MnO2 shell: it exhibits intrinsic GOx-like activity [16,27], directly catalyzing glucose oxidation to gluconic acid while generating H2O2 (Fig. S6); simultaneously, it decomposes both endogenous and newly generated H2O2 into oxygen and water. This oxygen supply effectively alleviates tumor hypoxia and provides the essential substrate for the embedded natural GOx. Thereby, it creates a sustainable catalytic microenvironment that synergistically potentiates starvation therapy via intensified glucose consumption.

The photothermal effect of MnO2 further accelerated enzymatic kinetics, promoting glucose catabolism. Dissolved oxygen profiles (Fig. 2H) demonstrated efficient oxygen generation through MnO2-mediated H2O2 decomposition, with photothermal enhancement substantially accelerating oxygen release kinetics. This catalytic cascade significantly elevated oxidative metabolites (H2O2 and gluconic acid) as quantified in Figs. 2I and J, confirming the system's efficient glucose depletion through enzymatic catalysis.

The release kinetics of LPC from the nanoplatform were evaluated under physiologically relevant conditions using dynamic membrane dialysis. Under neutral conditions (pH 7.4 PBS), the system demonstrated minimal drug release (6.17% ± 1.31% over 24 h), confirming excellent stability in normal tissues. In contrast, acidic conditions (pH 5.0) triggered significant LPC release (67.91% ± 2.56% over 24 h), attributable to pH-dependent structural collapse of the ZIF-8@MnO2 framework. NIR irradiation (808 nm, 2.5 w/cm2) increased LPC release to 74.61% ± 2.37% within 24 h (Fig. 2K), demonstrating effective photothermal control. In glucose-containing buffer, GOx-mediated acidification synergized with photothermal effects to achieve 83.83% ± 4.05% release (Fig. 2L). These results collectively establish the system's capacity for TME-specific drug delivery through combined pH responsiveness and photothermal activation.

3.3. In vitro cytotoxic effects and mechanistic insights of LGZM

The systematic evaluation of ZIF-8 and LGZM demonstrated excellent biocompatibility, with ZIF-8 maintaining over 90% cell viability across various concentrations (Fig. 3A) and the nanoplatform showing exhibited minimal cytotoxicity to normal 293T cells (80.51% ± 3.67% viability after 24 h incubation). In marked contrast, the system exhibited potent selective cytotoxicity against tumor cells, achieving approximately 50% inhibition in 4T1, MCF-7, and A549 cells while reducing HeLa cell viability to 40.48% ± 3.16% (Fig. 3B). The tumor-selective cytotoxicity mechanism arises from the synergistic combination of pH-responsive nanoplatform disintegration in the acidic TME and NQO1 enzyme overexpression in malignant cells, which together enable targeted drug release and selective bioactivation. This dual-targeting approach, combined with the synergistic therapeutic effects of the loaded components, ensures precise anticancer efficacy through both physical and biological discrimination mechanisms while maintaining excellent biosafety in normal cells that lack these characteristic tumor features.

Fig. 3.

Fig. 3

Comprehensive in vitro evaluation of the LPC/GOx@ZIF-8@MnO2 (LGZM): cytotoxicity, cellular uptake, and therapeutic mechanisms. (A) The viability of HeLa cells treated with zeolitic imidazolate framework-8 (ZIF-8) or LGZM + 808 nm near-infrared (NIR) laser at various concentrations. (B) The viability of 293T cells, HeLa cells, A549 cells, MCF-7 cells, and 4T1 cells treated with different formulations. (C) Colocalization observation by confocal laser scanning microscope of the Cy5/LGZM (red) and LysoTracker (green) in HeLa cells. (D) Evolution with time of the Pearson's correlation coefficients between the signals from the Cy5/LGZM and LysoTracker. (E) The viability of HeLa cells treated with the different formulations. (F–K) Representative fluorescence images and corresponding quantitative analysis of cell status in HeLa cells treated with different nanoparticles: Live/dead staining (F, G), intracellular reactive oxygen species (ROS) accumulation (H, I), and mitochondrial membrane potential (J, K). Data presented as mean ± standard deviation (SD), n = 5 in A, B, E, n = 3 in D, G, I, K. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001. ns: not significant; LPC: β-Lapachone; GOx: glucose oxidase; LGZ: LPC/GOx@ZIF-8; LGZM + L: LPC/GOx@ZIF-8@MnO2 with near-infrared laser irradiation; calcein-AM: calcein acetoxymethyl ester; PI: propidium iodide; DCFH-DA: 2′,7′-dichlorodihydrofluorescein diacetate; ROS FI: reactive oxygen species fluorescence intensity.

Lysosomal escape is critical for maintaining the pharmacological activity of endocytosed nanodrugs. We investigated the lysosomal escape behavior of Cy5/LGZM nanoparticles using fluorescence co-localization imaging. The nanoparticles (red) and lysosomal tracker (green) were visualized by confocal laser scanning microscopy (CLSM), and their co-localization was quantified by Pearson's correlation analysis. CLSM images showed that after 2 h of incubation, the nanoparticles began to enter cells and were endocytosed into lysosomes, with a Pearson's coefficient of 0.663 ± 0.040 between the Cy5 signal and lysosomal tracker. When the incubation time was extended to 4 h, the Cy5/LGZM nanoparticles predominantly remained trapped in lysosomes, as indicated by strong co-localization and a significantly increased Pearson's coefficient of 0.840 ± 0.026 compared to the 2 h group, likely due to more complete internalization. By 6 h, the Cy5 signal had diffused throughout the cytoplasm, accompanied by a decrease in the Pearson's coefficient to 0.533 ± 0.045, indicating lysosomal escape and drug release. Notably, clear nuclear localization was observed, confirming successful nuclear translocation of the drug (Figs. 3C and D). This sequential localization pattern—from lysosomal entrapment to cytoplasmic release and finally nuclear accumulation—supports efficient cellular internalization via the endolysosomal pathway and subsequent intracellular drug delivery. Further evaluation using a 3D tumor spheroid model showed remarkable penetration capacity, with fluorescent nanoparticles reaching depths of 80–100 μm within the tumor mass after 6 h (Fig. S7), indicating the system's ability to overcome dense tumor barriers. These findings collectively demonstrated both effective cellular uptake and deep tissue penetration capabilities are achieved by LGZM, which are critical prerequisites for successful antitumor therapy.

Building upon its favorable biosafety and cellular uptake profile, we further evaluated the antitumor efficacy of the LGZM nanoplatform. The therapeutic efficacy evaluation revealed a HeLa cell viability of 96.25 ± 1.80% in the PBS control group. Treatment with LPC or GOx alone resulted in cell viability rates of 66.14% ± 3.51% and 75.09% ± 2.99%, respectively, while the LGZ co-delivery system showed enhanced cytotoxicity with cell viability decreasing to 63.04% ± 5.01%, confirming the advantage of combined drug delivery. Notably, the MnO2-modified LGZM system exhibited significantly stronger antitumor effects, further reducing cell viability to 50.50% ± 3.01%. This cytotoxicity was substantially enhanced by PTT, which decreased HeLa cell viability to 39.44% ± 5.36% (Fig. 3E). This enhancement originated from the efficient photothermal conversion and rapid temperature increase, as confirmed by thermal imaging (Fig. S8), which promoted drug release and directly induced tumor cell ablation. Consistent with the Cell Counting Kit-8 (CCK-8) assay, live/dead staining images demonstrated the highest rate of cell death in the final treatment group (Figs. 3F and G). These results collectively establish that the trimodal combination of PTT, oxidative stress therapy, and starvation therapy yields significantly superior antitumor effects compared to monotherapies, highlighting the clinical potential of this nanosystem.

Assessment of intracellular molecular oxygen dynamics using the [Ru(dpp)3]Cl2 probe provided further mechanistic insights. Compared to the PBS control, bright [Ru(dpp)3]Cl2 fluorescence was observed in cells treated with LPC or GOx alone, indicating oxygen consumption via LPC's redox cycling and GOx-mediated glucose oxidation, respectively. In contrast, a significant fluorescence quenching of the probe was detected in the LGZM group, revealing that the MnO2 coating substantially alleviated hypoxia and elevated intracellular oxygen levels. This oxygenating effect was further augmented under laser irradiation, where localized heating promoted oxygen release and diffusion, leading to enhanced tumor cell oxygenation (Fig. S9). Furthermore, while the individual actions of LPC and GOx consumed oxygen and concurrently elevated intracellular H2O2 levels, no significant net change in H2O2 was observed in the LGZM group (Fig. S10). It is plausible that the absence of a significant dynamic change was due to the consumption of H2O2 in the MnO2-catalyzed cascade reaction.

2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) fluorescence imaging further revealed weak green fluorescence of ROS in the PBS control group and relatively stronger signals in the LGZ group, while formulations containing MnO2 exhibited markedly enhanced fluorescence intensity (Figs. 3H and I). These results demonstrate that the oxygen generated by MnO2 further promoted LPC redox cycling, significantly elevating intracellular ROS levels. It is worth noting that the GZM group also showed relatively high ROS levels, consistent with the intrinsic properties of MnO2, such as its potential Fenton-like activity. However, the ROS intensity triggered by the LGZM group was significantly higher than that of the GZM group and markedly exceeded that of any single-component treatment group (Fig. S11). This comparison strongly suggests that the dramatic ROS burst observed in the LGZM group primarily stems from the synergistic amplification of LPC redox cycling facilitated by MnO2-mediated hypoxia alleviation, rather than from a simple additive effect of individual ROS-generating capabilities. In summary, this hypoxia-relieving effect promoted an LPC-induced ROS storm, thereby achieving more potent tumor cell killing.

Mitochondrial membrane potential analysis demonstrated that the LGZM nanosystem induced significant mitochondrial depolarization under 808 nm laser irradiation, as evidenced by the marked decrease in JC-1 red/green fluorescence ratio (Figs. 3J and K). These results confirm that the nanoplatform promotes early tumor cell apoptosis through mitochondrial dysfunction, establishing the crucial involvement of the intrinsic apoptotic pathway in its antitumor mechanism.

Given that over 90% of cancer-related mortality results from metastatic dissemination [44,45], we systematically evaluated the anti-metastatic potential of LGZM through three complementary assays. Transwell invasion assays demonstrated 62.80 ± 1.40% inhibition of fetal bovine serum-induced chemotaxis (Figs. 4A and S12), while wound healing assays revealed an 18.33% ± 1.49% migration rate compared to 36.19% ± 1.49% in PBS controls (Figs. 4A and S13). Clonogenic assays showed a 51.19% ± 6.79% reduction in colony formation versus controls (Figs. 4A and S14), confirming durable growth inhibition at the population level. Notably, these anti-metastatic effects were significantly enhanced under 808 nm laser irradiation, demonstrating the nanosystem's multimodal capacity to simultaneously suppress tumor proliferation, migration, and colony formation [46].

Fig. 4.

Fig. 4

LPC/GOx@ZIF-8@MnO2 (LGZM) suppresses metastatic behavior and induces oxidative damage in tumor cells. (A) Representative invasion, wound-healing, and colony formation images of HeLa cells treated with different formulations. (B–D) Changes in malondialdehyde (MDA) content (B), glutathione (GSH) level (C), and superoxide dismutase (SOD) activity (D) in HeLa cells treated with different formulations. (E) Representative Western blot analysis of hypoxia-inducible factor 1-alpha (HIF-1α), NAD(P)H:quinone oxidoreductase 1 (NQO1), heat shock protein 70 (HSP70), caspase-3, high mobility group box 1 (HMGB1), matrix metalloproteinase-1 (MMP-1), and extracellular signal-regulated kinase 1/2 (ERK) activation in HeLa cell lysates. (F) Quantitative analysis of HIF-1α, NQO1, and HSP70 expression levels normalized to β-actin. (G) Quantitative analysis of caspase-3, HMGB1, and MMP-1 expression levels normalized to β-actin, and ERK activation expressed as the ratio of phosphorylated extracellular signal-regulated kinase 1/2 (p-ERK) to total extracellular signal-regulated kinase 1/2 (t-ERK). (H) The molecular mechanism of LGZM promoting tumor cell death and suppressing tumor metastasis. Data presented as mean ± standard deviation (SD), n = 3 in B–D, F, G. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. ns: not significant; LPC: β-Lapachone; GOx: glucose oxidase; ZIF-8: zeolitic imidazolate framework-8; LGZ: LPC/GOx@ZIF-8; LGZM + L: LPC/GOx@ZIF-8@MnO2 with near-infrared laser irradiation; ROS: reactive oxygen species; MAPK: mitogen-activated protein kinase; GSSG: oxidized glutathione.

To elucidate the molecular mechanisms underlying the anti-proliferative and anti-migratory effects of LGZM, we systematically analyzed oxidative stress markers and protein expression profiles. Our findings demonstrated: i) significant elevation of MDA levels (Fig. 4B) coupled with depletion of GSH and SOD (Figs. 4C and D), accompanied by MnO2-mediated GSH oxidation to GSSG (Figs. S15 and S16), confirming severe oxidative stress induction and antioxidant system impairment; ii) marked reduction of HIF-1α expression (Figs. 4E and F), demonstrating effective hypoxia alleviation by MnO2; iii) molecular signatures of apoptosis activation (upregulated caspase-3) and suppressed tumor invasiveness (downregulated HMGB1, MMP-1, and p-ERK1/2) (Figs. 4E and G). Notably, NIR irradiation induced photothermal stress responses, evidenced by 2.4-fold and 1.9-fold increases in HSP70 and NQO1 expression respectively (Figs. 4E and F), with expression levels exhibiting power density-dependence on NIR irradiation (Fig. S17). These molecular alterations collectively demonstrate that the nanosystem exerts its anti-tumor effects through coordinated modulation of oxidative stress, hypoxia relief, apoptosis induction, and photothermal stress response pathways.

To further validate the functional necessity of the photothermally activated HSP70/NQO1 axis, we performed pharmacological inhibition assays using the HSP70 inhibitor pifithrin-μ and the NQO1 inhibitor dicoumarol. As shown in Fig. S18, while NIR irradiation significantly upregulated both HSP70 and NQO1 expression in LGZM-treated cells, inhibition of either HSP70 or NQO1 effectively suppressed this activation. More importantly, flow cytometry analysis revealed that the enhanced ROS generation induced by photothermal activation was markedly attenuated upon inhibition of the HSP70/NQO1 axis. These results confirm that the photothermally activated HSP70/NQO1 axis plays an indispensable role in amplifying the LPC-mediated ROS storm, providing direct evidence for its functional significance in the nanosystem's therapeutic mechanism.

Building on these mechanistic insights, we propose that the LGZM nanoplatform exerts synergistic antitumor effects through three interconnected mechanisms (Fig. 4H): i) Hypoxia alleviation: MnO2 catalytically converts endogenous H2O2 to oxygen, which ameliorates tumor hypoxia and downregulates HIF-1α, thereby enhancing therapeutic sensitivity. ii) Oxidative stress amplification: Concurrent depletion of GSH and SOD compromises cellular antioxidant defenses, while NIR-activated photothermal effects upregulate the HSP70/NQO1 axis, establishing a positive feedback loop for sustained ROS generation through continuous LPC redox cycling. iii) Apoptotic induction: Elevated MDA levels and caspase-3 activation confirm oxidative damage-induced apoptosis, complemented by reduced tumor invasiveness (downregulated p-ERK1/2, HMGB1 and MMP-1). This coordinated mechanism sequentially disrupts tumor antioxidant defenses through hypoxia modulation, photothermal sensitization, and bioreductive amplification, ultimately generating a lethal oxidative storm that effectively induces tumor cell apoptosis.

3.4. In vivo antitumor efficacy of LGZM

Following intravenous administration in U14 tumor-bearing mice (tumor volume ≧200 mm3), the photothermal properties of LGZM were evaluated under 808 nm laser irradiation (2.5 W/cm2). Infrared thermography revealed rapid temperature elevation in tumors treated with LGZM, reaching 47.2 °C within 6 min and stabilizing at 50.2 °C (Figs. 5A and B). This pronounced photothermal effect, absent in PBS controls, demonstrates both efficient tumor accumulation of the nanoplatform and the exceptional photothermal conversion capability of the loaded MnO2 in LGZM. The achieved temperature range is appropriate for inducing tumor-specific heat shock responses while preventing nonspecific thermal damage to surrounding tissues.

Fig. 5.

Fig. 5

Thermographic images, biodistribution, and pharmacokinetics of LPC/GOx@ZIF-8@MnO2 (LGZM) in U14 tumor-bearing mice. (A, B) Photothermal imaging (A) and photothermal heating curve (B) of U14 tumor-bearing mice after intravenous injection with saline and LGZM respectively for 6 h and then treated with 808 nm near-infrared (NIR) laser. (C) Representative fluorescence images of U14 tumor-bearing mice at 0, 2, 4, 6, 8, 12, and 24 h post-injection of free Cy5 or Cy5/LGZM. (D) Ex vivo fluorescence images of excised tumors and major organs from U14 tumor-bearing mice. (E) Quantified fluorescence intensity of the excised tumors and organs. (F) Concentration-time profiles of β-Lapachone (LPC) in mice plasma after intravenous injection of free LPC or LGZM. Data presented as mean ± standard deviation (SD), n = 3 in B, E, and F. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗∗P < 0.0001. ns: not significant; LGZM + L: LPC/GOx@ZIF-8@MnO2 with near-infrared laser irradiation; con: concentration.

In vivo targeting capability of Cy5/LGZM was quantitatively assessed using real-time fluorescence imaging in tumor-bearing mice. Biodistribution analysis revealed rapid tumor accumulation of Cy5-labeled nanoparticles, with detectable targeting observed as early as 2 h post-injection and progressive accumulation over 24 h (Fig. 5C), demonstrating the combined effects of EPR and TME-responsive accumulation. Comparative analysis showed higher tumor fluorescence intensity in the Cy5/LGZM nanoparticle group versus free Cy5 controls, confirming improved circulation time and targeting efficiency. Ex vivo organ imaging at 24 h post-injection revealed preferential tumor accumulation (Fig. 5D), demonstrating a significantly higher tumor-to-liver ratio (1.61 ± 0.12) compared to free Cy5-treated group (0.51 ± 0.04), with minimal off-target distribution in reticuloendothelial organs (Figs. 5E and S19). These quantitative results establish the system's capacity for tumor-specific delivery.

Pharmacokinetic evaluation revealed superior performance of the LGZM nanocomplex compared to free LPC, demonstrating a 1.84-fold prolongation of plasma half-life (4.621 ± 0.374 h vs. 2.514 ± 0.114 h, Fig. 5F). This enhanced circulation stability is supported by the significantly lower protein adsorption of the nanocomplex (<4% bovine serum albumin binding) compared to free LPC, confirming its effective “stealth” properties (Fig. S20), along with excellent colloidal stability in biological media, as evidenced by only a 7.9 ± 4.3 nm size increase after 48 h in 10% FBS/PBS (Fig. S3) without aggregation. These favorable properties, including structural stability and a low protein adsorption rate, not only prolonged the systemic circulation time but also established a solid pharmacokinetic foundation for tumor-selective accumulation via the EPR effect.

To achieve early intervention by targeting the initial TME establishment, different formulations were administered via tail vein injection 24 h after tumor inoculation. Six hours post-injection, mice in the corresponding group received photothermal treatment with an 808 nm NIR laser (2.5 W/cm2, 6 min). Treatments were performed every other day for a total of seven sessions (Fig. 6A). LGZM nanoparticles demonstrated remarkable antitumor efficacy in the U14 tumor-bearing mouse model, exhibiting a 61.5% reduction in tumor mass (1.03 ± 0.15 g vs. saline control 2.66 ± 0.26 g) that significantly outperformed free LPC (33.1% inhibition), free GOx (38.7% inhibition), and LGZ (52.7% inhibition) (Fig. 6B). The therapeutic superiority was attributed to MnO2-mediated catalytic conversion of endogenous H2O2 to oxygen, which simultaneously enhanced GOx-mediated starvation therapy by disrupting tumor energy metabolism and promoted LPC redox cycling for amplified ROS generation. Photothermal combination therapy further elevated the inhibition rate to 76.2% (vs. 61.4% without irradiation), with tumor volume kinetics and excised tumor morphology consistently validating these therapeutic outcomes (Figs. 6C and D).

Fig. 6.

Fig. 6

Therapeutic evaluation of LPC/GOx@ZIF-8@MnO2 (LGZM) in vivo. (A) Experimental schedule of U14 tumor-bearing mice treated with different protocols. (B) Isolated tumor weights, (C) tumor volume changes, and (D) isolated tumor images of U14 tumor-bearing mice with different formulation treatments. (E) Representative images of hematoxylin and eosin (H&E), terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL), and Ki67 staining in tumor tissues. (F–H) Corresponding quantitative analysis of the nuclear-cytoplasmic ratio (F), apoptotic area (G), and Ki67 proliferation index (H). (I) Representative photographs of lung surfaces and corresponding H&E-stained sections from mice receiving different treatments. (J) Quantification of lung metastatic nodules in different treatment groups. Data presented as mean ± standard deviation (SD), n = 5 in B, C, F–H, J. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.0001, and ∗∗∗∗P < 0.0001. ns: not significant; NIR: near-infrared; NS: normal saline; ZIF-8: zeolitic imidazolate framework-8; LPC: β-Lapachone; GOx: glucose oxidase; LGZ: LPC/GOx@ZIF-8; LGZM + L: LPC/GOx@ZIF-8@MnO2 with near-infrared laser irradiation.

Histopathological examination provided mechanistic insights into the therapeutic effects, with H&E staining revealing significantly reduced nuclear-to-cytoplasmic ratios (0.57 ± 0.03 vs. 0.65 ± 0.04 in saline controls, Figs. 6E and F) and characteristic therapeutic damage patterns in the LGZM + Laser group. TUNEL assays demonstrated 2.46-fold increased apoptotic areas compared to controls (Figs. 6E and G), while Ki67 immunohistochemistry showed markedly decreased proliferation indices (Figs. 6E and H), collectively confirming the treatment's dual capacity to induce apoptosis and suppress proliferation.

The anti-metastatic efficacy was further validated using a lung nodule formation model, which demonstrated the significant capacity of LGZM to inhibit tumor dissemination. Compared to the saline control group, mice treated with LGZM exhibited a remarkable reduction in the number of surface metastatic nodules on the lungs. H&E staining and histopathological analysis of lung tissue sections revealed extensive tumor cell foci and severe destruction of alveolar structures in the control group. In contrast, only sporadic micro-metastatic foci were observed in the LGZM treatment group, with the majority of lung tissues maintaining intact architecture and showing no significant tumor infiltration (Figs. 6I and J). These findings collectively demonstrate that LGZM effectively suppresses tumor metastasis by reducing metastatic colonization and preserving normal pulmonary architecture.

3.5. Therapeutic mechanism of LGZM

Having established the potent antitumor efficacy of LGZM, we next elucidated the mechanistic basis for its therapeutic effects through multiple coordinated pathways:

Hypoxia amelioration, evidenced by significant downregulation of HIF-1α in tumors treated with LGZM (Figs. 7A and B), confirming MnO2 mediated improvement of the TME. This finding was further substantiated by direct measurement of intratumoral oxygen partial pressure (pO2). As shown in Fig. S21A, compared to the saline control group (16.55 ± 1.41 mmHg), LGZ slightly reduced tumor pO2 to 13.30 ± 1.18 mmHg due to oxygen consumption. In contrast, MnO2 modification markedly alleviated hypoxia, elevating pO2 to 24.45 ± 1.43 mmHg in the LGZM group. Under laser irradiation, local heating further enhanced oxygen release and diffusion, increasing tumor pO2 to 28.78 ± 2.05 mmHg. Spatial mapping of intratumoral oxygen distribution consistently demonstrated that MnO2 decoration combined with laser irradiation effectively improved overall tumor oxygenation (Fig. S21B).

Fig. 7.

Fig. 7

The mechanism of LPC/GOx@ZIF-8@MnO2 (LGZM) in inhibiting tumor growth and metastasis. (A) Representative Western blots and quantitative analysis of hypoxia-inducible factor 1-alpha (HIF-1α), NAD(P)H:quinone oxidoreductase 1 (NQO1), heat shock protein 70 (HSP70), caspase-3, and high mobility group box 1 (HMGB1) expression and extracellular signal-regulated kinase 1/2 (ERK) activation in tumor tissue homogenates. (B) Representative immunohistochemistry (IHC) images of HIF-1α, NQO1, caspase-3, HMGB1, and matrix metalloproteinase-1 (MMP-1) in tumor tissue sections, and the semi-quantification of these IHC images. (C–E) glutathione (GSH) level (C), malondialdehyde (MDA) content (D) and superoxide dismutase (SOD) activity (E) in tumor tissues of U14 tumor-bearing mice after different treatments. Data presented as mean ± standard deviation (SD), n = 3 in A, n = 5 in B–E. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001. ns: not significant; p-ERK: phosphorylated extracellular signal-regulated kinase 1/2; t-ERK: extracellular signal-regulated kinase 1/2; NS: normal saline; ZIF-8: zeolitic imidazolate framework-8; LPC: β-Lapachone; GOx: glucose oxidase; LGZ: LPC/GOx@ZIF-8; LGZM + L: LPC/GOx@ZIF-8@MnO2 with near-infrared laser irradiation.

Synergistic oxidative stress potentiation and photothermal enhancement were systematically validated through multi-level analysis. In tumor tissues, this combined treatment strategy triggered a significant oxidative stress response, characterized by elevated MDA levels and depletion of GSH and SOD (Figs. 7C–E). This redox imbalance worked in concert with the marked upregulation of the apoptosis executioner protein caspase-3 (Figs. 7A and B), collectively promoting tumor cell apoptosis. Systemic oxidative stress monitoring further revealed the whole-body effects of the treatment. Serum analysis showed a significant increase in MDA levels, while circulating SOD and GSH levels exhibited a compensatory rise (Fig. S22), indicating a systemic antioxidant response to the MnO2-mediated disruption of local redox homeostasis in the tumor.

The enhancement mechanism of PTT operates through a dual pathway: the direct thermal ablation effect of NIR laser irradiation combined with the photothermally activated HSP70/NQO1 signaling axis (Fig. 7A and B), which significantly augments the LPC-induced oxidative stress therapy. This synergy-induced intense oxidative burst was directly confirmed at the tissue level: dihydroethidium (DHE) staining and quantitative analysis demonstrated a significant increase in O2− generation in the LGZM treatment group. Furthermore, NIR laser irradiation further enhanced both the intensity and distribution range of the fluorescence signal, clearly indicating that the photothermal effect effectively triggered a burst of ROS generation (Figs. S23A and B). Flow cytometric analysis independently verified these findings. The data revealed a significant increase in total ROS levels in the LGZ group, while the LGZM group exhibited even higher ROS levels. After laser irradiation, the ROS level in this group peaked among all experimental groups (Fig. S23C). These quantitative results were highly consistent with the DHE imaging conclusions, collectively confirming the capability of this nanoplatform to induce intense oxidative stress under photothermal regulation.

Metastasis suppression was demonstrated by reduced expression of HMGB1 (involved in tumor cell migration), MMP-1 (critical for extracellular matrix degradation), and p-ERK1/2 (key signaling mediator of invasion pathways) (Figs. 7A and B), which indicates the disruption of molecular pathways essential for tumor invasion and migration.

Collectively, these data demonstrate that LGZM achieves potent antitumor efficacy through multimodal mechanisms encompassing hypoxia relief, oxidative damage induction, metabolic interference, apoptotic activation, and metastasis inhibition.

3.6. Biosafety and biocompatibility assessment of LGZM

Comprehensive biosafety assessment demonstrated excellent biocompatibility of the LGZM platform. In vitro hemolysis tests revealed minimal erythrocyte damage (hemolysis rate <5% at 0–1000 μg/mL concentrations, Fig. 8A), satisfying ISO 10993-4 standards [47] for biomedical materials. Longitudinal monitoring showed no significant differences in body weight trajectories between treatment and control groups (Fig. 8B), with all animals maintaining normal growth patterns. Post-treatment analysis indicated physiological organ indices (heart, liver, spleen, lung, kidney, thymus) within normal limits, with no pathologic enlargement or atrophy signs (Fig. 8C). Serum biochemistry confirmed preserved hepatic and renal function: ALT (11.69 ± 0.62 U/L), AST (78.21 ± 5.47 U/L), CRE (24.91 ± 2.13 μmol/L), and BUN (25.25 ± 1.47 mmol/L) levels all remained within normal ranges (Figs. 8D–G). Histopathological examination of major organs revealed well-preserved tissue architecture with no significant signs of pathological alterations, such as inflammation, necrosis, or fibrosis (Fig. 8H). These findings indicate that LGZM did not induce obvious toxic side effects in these organs under the experimental conditions [48].

Fig. 8.

Fig. 8

Biosafety assessment of LPC/GOx@ZIF-8@MnO2 (LGZM). (A) The hemolysis effect of erythrocyte after incubating with LGZM at different concentrations. (B, C) Body weight changes (B), and organ indices (C) of U14 tumor-bearing mice after different treatments. (D–G) Alanine aminotransferase (ALT) (D), aspartate aminotransferase (AST) (E), creatinine (CRE) (F), and blood urea nitrogen (BUN) (G) levels in the serum of U14 tumor-bearing mice following different treatments. (H) Representative images of the major organ sections after hematoxylin and eosin (H&E) staining. Data presented as mean ± standard deviation (SD), n = 5 in A–G. NC: Normal control; NS: normal saline; ZIF-8: zeolitic imidazolate framework-8; LPC: β-Lapachone; GOx: glucose oxidase; LGZ: LPC/GOx@ZIF-8; LGZM + L: LPC/GOx@ZIF-8@MnO2 with near-infrared laser irradiation.

These multimodal safety data, encompassing hemocompatibility, systemic toxicity, and organ function assessments, establish a favorable short-term biosafety profile for the LGZM nanosystem and confirm its potent antitumor efficacy, thereby supporting its preliminary translational potential for clinical oncology applications. While these short-term biosafety results are promising, the long-term in vivo fate, potential accumulation-related toxicity, and immunogenicity of the nanoplatform warrant further thorough evaluation. The ZIF-8 framework is expected to undergo gradual biodegradation under acidic conditions into Zn2+ ions and imidazole linkers, which can be metabolized and cleared renally. Similarly, the MnO2 shell can be reduced to Mn2+ by intracellular GSH and other reducing agents for subsequent excretion. Nevertheless, it is acknowledged that a portion of the nanoparticles may be sequestered by the reticuloendothelial system (RES), primarily in the liver and spleen. Future studies should therefore focus on the long-term biodistribution, RES accumulation kinetics, and potential immunogenicity of this platform. Strategies such as surface PEGylation or the application of biocompatible coating materials could be explored to minimize RES uptake and enhance overall biosafety, representing critical next steps toward advancing this nanoplatform along the clinical translation pathway.

3.7. Polypharmacological mechanism and therapeutic potential of LPC

To explore the multi-target mechanism of LPC at the molecular and systems levels, and to assess its potential as the primary active agent for future nanomedicine design, we conducted an integrated computational study combining molecular docking and network pharmacology.

Molecular docking was performed to predict potential interactions between LPC and key proteins (caspase-3, HMGB1, MMP-1) implicated in our experimental findings. The results revealed strong binding affinities of LPC with caspase-3 (binding energy = −7.7 kcal/mol) via hydrogen bonds with TYR-202 and ARG-167 in its active center, with HMGB1 (binding energy = −7.1 kcal/mol) involving residues ARG-104 and PRO-102, and with MMP-1 (binding energy = −8.5 kcal/mol) at the catalytic domain residues ASN-180 and HIS-183 (Fig. S24). These structural insights suggest a plausible model for LPC's multi-target activity: promoting apoptosis through caspase-3 engagement, disrupting HMGB1-mediated signaling, and inhibiting MMP-1-driven invasion.

To further elucidate the broader polypharmacological landscape of LPC, a network pharmacology approach was employed. Systematic screening identified 43 overlapping targets of LPC and squamous cell carcinoma (Figs. 9A and B). Topology analysis of the protein-protein interaction network highlighted seven core targets: proto-oncogene tyrosine-protein kinase Src (SRC), signal transducer and activator of transcription 3 (STAT3), epidermal growth factor receptor (EGFR), matrix metalloproteinase 2 (MMP2), androgen receptor (AR), mitogen-activated protein kinase 8 (MAPK8), and poly (ADP-ribose) polymerase 1 (PARP1) (Figs. 9C and D). KEGG pathway enrichment analysis demonstrated their significant involvement in critical oncogenic processes, including apoptosis, necroptosis, EGFR tyrosine kinase inhibitor resistance, the programmed cell death ligand 1/programmed cell death protein 1 (PD-L1/PD-1) checkpoint pathway, Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling, extracellular matrix organization, and metabolic reprogramming. (Fig. 9E). GO enrichment analysis further indicated that these targets regulate key molecular functions such as protein kinase activity and ATP binding, and are involved in essential biological processes including apoptotic regulation and signal transduction (Fig. 9F).

Fig. 9.

Fig. 9

Bioinformatics analysis uncovers β-Lapachone's pleiotropic anti-tumor mechanisms in squamous cell carcinoma. (A) The potential targets of β-Lapachone intersected with squamous cell carcinoma disease targets. (B) Subnetwork of candidate core targets screened based on topological parameters (Degree >9.95, Betweenness >44.65, Closeness >0.12). (C) Subnetwork of core targets obtained after secondary topological screening. (D) Interaction network of the final screened hub targets. (E) Integrated Kyoto Encyclopedia of Genes and Genomes (KEGG) and (F) Gene Ontology (GO) pathway enrichment analysis of common targets. SRC: proto-oncogene tyrosine-protein kinase Src; STAT3: signal transducer and activator of transcription 3; EGFR: epidermal growth factor receptor; MMP2: matrix metalloproteinase 2; AR: androgen receptor; MAPK8: mitogen-activated protein kinase 8; PARP1: poly (ADP-ribose) polymerase 1.

Collectively, this computational work positions LPC as a promising antitumor drug candidate capable of coordinately regulating a network of targets spanning apoptosis induction, immune microenvironment modulation, and metabolic reprogramming. These predictions not only offer a deeper mechanistic rationale for the observed efficacy of the LGZM nanosystem but, more importantly, deliver valuable clues and a strategic direction for the independent development of LPC and the rational design of next-generation combination nanotherapies.

4. Discussion

In recent years, significant progress has been made in nanoplatforms designed to modulate the TME for enhanced synergistic therapy, providing an important reference and comparative context for this study. A comprehensive analysis of recent representative works reveals that the mainstream design paradigm primarily revolves around constructing efficient chemical catalytic cycles. For example, by integrating GOx with MnO2 or other metal-based nanozymes, a cascade reaction is established from glucose consumption and H2O2 generation to ROS production, thereby achieving synergy between starvation therapy and chemodynamic therapy [26,[49], [50], [51]]. Some advanced designs have further incorporated iron-based Fenton reagents or photosensitizers to expand the pathways for ROS generation or to combine photodynamic therapy, demonstrating excellent direct killing efficacy [41,52]. Notably, the latest research has begun exploring the potential of these nanosystems in inducing immunogenic cell death and regulating the tumor immune microenvironment, opening new dimensions for treatment [41,53].

Compared to the aforementioned cutting-edge works, the LGZM platform constructed in this study proposes a distinct paradigm in terms of integration strategy and mechanism of action. The core of this work is not merely the optimization of catalytic cycles or the introduction of new materials, but rather the pursuit of deep mechanistic coupling between “bioactive molecule-driven” processes and “inorganic nanocatalytic systems”. Specifically, this study innovatively integrates the unique redox cycle of LPC, which depends on the tumor-overexpressed enzyme NQO1, as a core therapeutic module into the nanoplatform. This design not only introduces a powerful and tumor-selective endogenous ROS burst source but also, through its continuous depletion of NAD(P)H and ATP, forms a mechanism complementary to and reinforcing starvation therapy at the fundamental level of energy metabolism [12].

More importantly, this study reveals and validates a synergistic mechanism that transcends mere physical mixing. Experimental results indicate that the photothermal effect of the MnO2 component in the platform can upregulate the protein expression level of NQO1 by activating the HSP70/NQO1 signaling axis, thereby positively amplifying the biocatalytic efficacy of LPC and forming a positive feedback loop of “physical stimulation enhancing bioactivity”. This synergistic mode, which strategically interconnects external physical energy with internal bio-enzyme activity at the signaling pathway level, is a significant characteristic of our platform compared to existing systems. Therefore, the contribution of this work lies in providing a novel design philosophy centered on “bioactivity-guided deep mechanistic coupling”. By integrating the unique mechanism of action of the chemotherapeutic agent LPC with nanocatalytic medicine and PTT, it offers a new paradigm and experimental foundation for overcoming TME limitations and achieving highly efficient synergistic therapy.

While this proof-of-concept study has made clear progress in mechanistic exploration and efficacy validation, its translation toward clinical application faces several key challenges, which also chart the course for future research. First, regarding the optimization of the treatment regimen, this study employed a relatively intensive dosing and irradiation schedule to maximize synergistic signals. Future translational research must systematically explore the potential for reducing treatment frequency, such as achieving efficacy with a single dose or only a few sessions. Our pharmacokinetic data demonstrate that the nanoplatform possesses favorable tumor targeting and retention properties, providing a foundation for investigating “reduced-dose with enhanced effect” strategies based on higher single doses or optimized laser parameters. Second, in terms of preparation process, although we employed mature standard methods and demonstrated good batch-to-batch consistency, the multi-step synthesis protocol requires further simplification and standardization for large-scale production. Future efforts will focus on developing more stable and scalable manufacturing processes. Finally, concerning the dimension of efficacy comparison, this study focused on internal mechanism validation and structural necessity. To comprehensively evaluate its competitive advantage, follow-up research requires direct, quantitative head-to-head comparisons with representative advanced platforms, particularly regarding key translational metrics such as tumor targeting efficiency, long-term biodistribution, and immune activation effects. Furthermore, confirming the causality of the HSP70/NQO1 axis through gene-editing techniques, along with systematically assessing long-term in vivo safety and potential immunogenicity, are indispensable steps toward advancing this platform toward clinical application.

In summary, this study provides a novel synergistic therapeutic paradigm for overcoming TME limitations, and addressing the aforementioned translational challenges will be key to realizing its clinical potential.

5. Conclusion

In this study, we successfully constructed a ZIF-8-based nanosystem, LGZM, which achieved highly efficient synergistic antitumor effects by alleviating the hypoxic TME and combining starvation therapy, oxidative stress therapy, and PTT. The nanosystem utilizes MnO2 to catalyze the overexpressed H2O2 in the TME to generate oxygen, thereby alleviating hypoxia while enhancing GOx-mediated glucose depletion and triggering a self-reinforcing cascade reaction that persistently disrupts tumor energy metabolism. Furthermore, the photothermal effect of the MnO2 component further upregulates NQO1 expression by activating the HSP70/NQO1 signaling axis, enhancing the LPC-mediated ROS burst. Simultaneously, its ability to deplete GSH effectively weakens the antioxidant defense capacity of tumor cells. In vivo experimental results demonstrated that the nanosystem exhibited significant tumor growth inhibition and favorable biosafety over a 15-day treatment period. To further elucidate the multi-target mechanism of LPC, we integrated molecular docking and network pharmacology analysis to evaluate its potential as a primary active component in future nanomedicine design. In summary, this study not only proposes an innovative synergistic strategy to overcome tumor hypoxic limitations but also preliminarily reveals the multi-target mechanism of LPC, providing an important theoretical and experimental foundation for the development of precise and efficient antitumor nanomedicines.

CRediT authorship contribution statement

Jiahui Cai: Writing – original draft, Methodology, Formal analysis, Conceptualization. Zutong Cui: Validation, Data curation. Wenming Liang: Validation, Data curation. Jianming Sun: Investigation. Zihan Zhang: Investigation. Feifei Sun: Visualization, Software. Feng Yang: Visualization, Software. Zhiwei Liu: Resources. Liming Gao: Supervision. Jidong Wang: Supervision. Jian Li: Writing – review & editing, Project administration, Funding acquisition.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files. Additional datasets are available from the corresponding author upon reasonable request.

Declaration on the use of AI assisted technology

During the research and manuscript preparation for this study, the authors utilized DeepSeek for assistance. This included language polishing and grammar checking of certain sections. The authors have carefully reviewed, verified, and revised all AI-generated content and assume full responsibility for the integrity, accuracy, and academic integrity of the work.

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.

Acknowledgments

The authors would like to greatly appreciate the Hebei Natural Science Foundation (Grant No.: H2023203003). The flowcharting was supported by Figdraw (https://www.figdraw.com/#/).

Footnotes

Peer review under responsibility of Xi'an Jiaotong University.

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jpha.2026.101633.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Multimedia component 1
mmc1.docx (3.1MB, docx)

References

  • 1.Xu Q., Lan X., Lin H., et al. Tumor microenvironment-regulating nanomedicine design to fight multi-drug resistant tumors. WIREs Nanomed. Nanobiotechnol. 2023;15 doi: 10.1002/wnan.1842. [DOI] [PubMed] [Google Scholar]
  • 2.Lei S., Zhang J., Blum N.T., et al. In vivo three-dimensional multispectral photoacoustic imaging of dual enzyme-driven cyclic cascade reaction for tumor catalytic therapy. Nat. Commun. 2022;13:1298. doi: 10.1038/s41467-022-29082-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Hu Y., Cheng H., Zhao X., et al. Surface-enhanced Raman scattering active gold nanoparticles with enzyme-mimicking activities for measuring glucose and lactate in living tissues. ACS Nano. 2017;11:5558–5566. doi: 10.1021/acsnano.7b00905. [DOI] [PubMed] [Google Scholar]
  • 4.Zhang W., Zhou Z., Lv Q., et al. O2-generation-enhanced responsive starvation/photothermal synergistic tumor therapy based on the AuNRs@MnO2@SiO2 nanocarrier and thermosensitive biomimetic camouflaging. ACS Appl. Bio Mater. 2023;6:4775–4790. doi: 10.1021/acsabm.3c00544. [DOI] [PubMed] [Google Scholar]
  • 5.Liang X., Chen W., Wang C., et al. A mesoporous theranostic platform for ultrasound and photoacoustic dual imaging-guided photothermal and enhanced starvation therapy for cancer. Acta Biomater. 2024;183:264–277. doi: 10.1016/j.actbio.2024.05.040. [DOI] [PubMed] [Google Scholar]
  • 6.He T., Xu H., Zhang Y., et al. Glucose oxidase-instructed traceable self-oxygenation/hyperthermia dually enhanced cancer starvation therapy. Theranostics. 2020;10:1544–1554. doi: 10.7150/thno.40439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Wang M., Wang D., Chen Q., et al. Recent advances in glucose-oxidase-based nanocomposites for tumor therapy. Small. 2019;15 doi: 10.1002/smll.201903895. [DOI] [PubMed] [Google Scholar]
  • 8.Pan Y., Liu L., Mou X., et al. Nanomedicine strategies in conquering and utilizing the cancer hypoxia environment. ACS Nano. 2023;17:20875–20924. doi: 10.1021/acsnano.3c07763. [DOI] [PubMed] [Google Scholar]
  • 9.Gong P., Cui H., Li C., et al. Self-stablized monodispersing nano-MOFs for controlled enzyme delivery. Chem. Eng. J. 2024;489 [Google Scholar]
  • 10.Gong P., Zhao K., Liu X., et al. Fluorescent COFs with a highly conjugated structure for combined starvation and gas therapy. ACS Appl. Mater. Interfaces. 2022;14:46201–46211. doi: 10.1021/acsami.2c11423. [DOI] [PubMed] [Google Scholar]
  • 11.Wu Y., Wang X., Chang S., et al. β-Lapachone induces NAD(P)H: quinone oxidoreductase-1-and oxidative stress-dependent heat shock protein 90 cleavage and inhibits tumor growth and angiogenesis. J. Pharmacol. Exp. Therapeut. 2016;357:466–475. doi: 10.1124/jpet.116.232694. [DOI] [PubMed] [Google Scholar]
  • 12.Ye M., Han Y., Tang J., et al. A tumor-specific cascade amplification drug release nanoparticle for overcoming multidrug resistance in cancers. Adv. Mater. 2017;29 doi: 10.1002/adma.201702342. [DOI] [PubMed] [Google Scholar]
  • 13.Wang J., Zhang H., Lv J., et al. A tumor-specific ROS self-supply enhanced cascade-responsive prodrug activation nanosystem for amplified chemotherapy against multidrug-resistant tumors. Acta Biomater. 2023;164:522–537. doi: 10.1016/j.actbio.2023.04.014. [DOI] [PubMed] [Google Scholar]
  • 14.Dai L., Li X., Duan X., et al. A pH/ROS cascade-responsive charge-reversal nanosystem with self-amplified drug release for synergistic oxidation-chemotherapy. Adv. Sci. 2019;6 doi: 10.1002/advs.201801807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.He L., Wang J., Zhu P., et al. Intelligent manganese dioxide nanocomposites induce tumor immunogenic cell death and remould tumor microenvironment. Chem. Eng. J. 2023;461 [Google Scholar]
  • 16.Yang G., Ji J., Liu Z. Multifunctional MnO2 nanoparticles for tumor microenvironment modulation and cancer therapy. WIREs Nanomed. Nanobiotechnol. 2021;13 doi: 10.1002/wnan.1720. [DOI] [PubMed] [Google Scholar]
  • 17.Zhang L., Yang R., Yu H., et al. MnO2-capped silk fibroin (SF) nanoparticles with chlorin e6 (Ce6) encapsulation for augmented photo-driven therapy by modulating the tumor microenvironment. J. Mater. Chem. B. 2021;9:3677–3688. doi: 10.1039/d1tb00296a. [DOI] [PubMed] [Google Scholar]
  • 18.Zhang W., Chen J., Guo Q., et al. Reversal of doxorubicin-resistance of MCF-7/Adr cells via multiple regulations by glucose oxidase loaded AuNRs@MnO2@SiO2 nanocarriers. Colloids Surf. B Biointerfaces. 2025;253 doi: 10.1016/j.colsurfb.2025.114748. [DOI] [PubMed] [Google Scholar]
  • 19.Zheng W., Zhou Z., Lv Q., et al. Oxygen-generated hierarchical-structured AuNRs@MnO2@SiO2 nanocarrier for enhanced NIR- and H2O2-responsive mild-hyperthermia photodynamic/photothermal combined tumor therapy. Adv. Ther. 2022;5 [Google Scholar]
  • 20.Cui C., Fang Y., Chen B., et al. Glucose oxidation performance is improved by the use of a supramolecular self-assembly of glucose oxidase and catalase. Catal. Sci. Technol. 2019;9:477–482. [Google Scholar]
  • 21.Zhao M., Zhu A., Zheng X., et al. Multistage-responsive dual-enzyme nanocascades for synergistic radiosensitization-starvation cancer therapy. Adv. Healthc. Mater. 2023;12 doi: 10.1002/adhm.202300118. [DOI] [PubMed] [Google Scholar]
  • 22.Chen Z., Li Y., Xiang Q., et al. Metallic copper-based dual-enzyme biomimetic nanoplatform for mild photothermal enhancement of anticancer catalytic activity. Biomater. Res. 2024;28:34. doi: 10.34133/bmr.0034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Wang Q., Sun Y., Li S., et al. Synthesis and modification of ZIF-8 and its application in drug delivery and tumor therapy. RSC Adv. 2020;10:37600–37620. doi: 10.1039/d0ra07950b. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Yu Y., Zhang J., Huang W., et al. Spatiotemporally controlled AuNPs@ZIF-8 based nanocarriers for synergistic photothermal/chemodynamic therapy. Acta Biomater. 2024;175:317–328. doi: 10.1016/j.actbio.2023.12.030. [DOI] [PubMed] [Google Scholar]
  • 25.Park K.S., Ni Z., Côté A.P., et al. Exceptional chemical and thermal stability of zeolitic imidazolate frameworks. Proc. Natl. Acad. Sci. U. S. A. 2006;103:10186–10191. doi: 10.1073/pnas.0602439103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Wang J., Yao L., Hu E., et al. MnO2 decorated ZIF-8@GOx for synergistic chemodynamic and starvation therapy of cancer. J. Solid State Chem. 2021;298 [Google Scholar]
  • 27.Tang W., Fan W., Zhang W., et al. Wet/sono-chemical synthesis of enzymatic two-dimensional MnO2 nanosheets for synergistic catalysis-enhanced phototheranostics. Adv. Mater. 2019;31 doi: 10.1002/adma.201900401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Peng J., Gong P., Song S., et al. Biomineralized synthesis of a smart O2-regenerating nanoreactor for highly efficient starvation/gas therapy. Mater. Sci. Eng. C. 2021;126 doi: 10.1016/j.msec.2021.112132. [DOI] [PubMed] [Google Scholar]
  • 29.Liu Z., Zhang S., Lin H., et al. Theranostic 2D ultrathin MnO2 nanosheets with fast responsibility to endogenous tumor microenvironment and exogenous NIR irradiation. Biomaterials. 2018;155:54–63. doi: 10.1016/j.biomaterials.2017.11.015. [DOI] [PubMed] [Google Scholar]
  • 30.Wang L., Guan S., Weng Y., et al. Highly efficient vacancy-driven photothermal therapy mediated by ultrathin MnO2 nanosheets. ACS Appl. Mater. Interfaces. 2019;11:6267–6275. doi: 10.1021/acsami.8b20639. [DOI] [PubMed] [Google Scholar]
  • 31.Chen W., Yan Y., Han R., et al. Biodegradable flower-like manganese for synergistic photothermal and photodynamic therapy applications. Photochem. Photobiol. Sci. 2021;20:153–160. doi: 10.1007/s43630-020-00010-w. [DOI] [PubMed] [Google Scholar]
  • 32.Peng J., Gong P., Li S., et al. A smart bioresponsive nanosystem with dual-modal imaging for drug visual loading and targeted delivery. Chem. Eng. J. 2020;391 [Google Scholar]
  • 33.Chang N., Zhao Y., Ge N., et al. A pH/ROS cascade-responsive and self-accelerating drug release nanosystem for the targeted treatment of multi-drug-resistant colon cancer. Drug Deliv. 2020;27:1073–1086. doi: 10.1080/10717544.2020.1797238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Wang Y., Li S., Zhang P., et al. Photothermal-responsive conjugated polymer nanoparticles for remote control of gene expression in living cells. Adv. Mater. 2018;30 doi: 10.1002/adma.201705418. [DOI] [PubMed] [Google Scholar]
  • 35.Do Cong T., Wang Z., Hu M., et al. Extraspecific manifestation of nanoheater’s position effect on distinctive cellular photothermal responses. ACS Nano. 2020;14:5836–5844. doi: 10.1021/acsnano.0c00951. [DOI] [PubMed] [Google Scholar]
  • 36.Dong G., Youn H., Park M.T., et al. Heat shock increases expression of NAD(P)H: quinone oxidoreductase (NQO1), mediator of β-Lapachone cytotoxicity, by increasing NQO1 gene activity and via Hsp70-mediated stabilisation of NQO1 protein. Int. J. Hyperther. 2009;25:477–487. doi: 10.1080/02656730903049836. [DOI] [PubMed] [Google Scholar]
  • 37.Xue C., Li M., Liu C., et al. NIR-actuated remote activation of ferroptosis in target tumor cells through a photothermally responsive iron-chelated biopolymer nanoplatform. Angew. Chem. Int. Ed. 2021;60:8938–8947. doi: 10.1002/anie.202016872. [DOI] [PubMed] [Google Scholar]
  • 38.Yin W., Ke W., Chen W., et al. Integrated block copolymer prodrug nanoparticles for combination of tumor oxidative stress amplification and ROS-responsive drug release. Biomaterials. 2019;195:63–74. doi: 10.1016/j.biomaterials.2018.12.032. [DOI] [PubMed] [Google Scholar]
  • 39.Son S., Won M., Green O., et al. Chemiluminescent probe for the in vitro and in vivo imaging of cancers over-expressing NQO1. Angew. Chem. Int. Ed. 2019;58:1739–1743. doi: 10.1002/anie.201813032. [DOI] [PubMed] [Google Scholar]
  • 40.Zhu J., Xiao T., Zhang J., et al. Surface-charge-switchable nanoclusters for magnetic resonance imaging-guided and glutathione depletion-enhanced photodynamic therapy. ACS Nano. 2020;14:11225–11237. doi: 10.1021/acsnano.0c03080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Zhang Y., Yang Y., Shi J., et al. A multimodal strategy of Fe3O4@ZIF-8/GOx@MnO2 hybrid nanozyme via TME modulation for tumor therapy. Nanoscale. 2021;13:16571–16588. doi: 10.1039/d1nr04196g. [DOI] [PubMed] [Google Scholar]
  • 42.Chen S., Sun P., Humphreys J., et al. Acetate-based ‘oversaturated gel electrolyte’ enabling highly stable aqueous Zn-MnO2 battery. Energy Storage Mater. 2021;42:240–251. [Google Scholar]
  • 43.Chen Z., Zhang Q., Huang Q., et al. Photothermal MnO2 nanoparticles boost chemo-photothermal therapy-induced immunogenic cell death in tumor immunotherapy. Int. J. Pharm. 2022;617 doi: 10.1016/j.ijpharm.2022.121578. [DOI] [PubMed] [Google Scholar]
  • 44.Fidler I.J. The pathogenesis of cancer metastasis: the ‘seed and soil’ hypothesis revisited. Nat. Rev. Cancer. 2003;3:453–458. doi: 10.1038/nrc1098. [DOI] [PubMed] [Google Scholar]
  • 45.Yang H., Wang B., Wang T., et al. Toll-like receptor 4 prompts human breast cancer cells invasiveness via lipopolysaccharide stimulation and is overexpressed in patients with lymph node metastasis. PLoS One. 2014;9 doi: 10.1371/journal.pone.0109980. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Song S., Wang D., Zhao K., et al. Donor-acceptor structured photothermal COFs for enhanced starvation therapy. Chem. Eng. J. 2022;442 [Google Scholar]
  • 47.Liu C., Lin Z., Qiao C., et al. Hemocompatibility assay of a micro-catheter using hydrophilic coating biomaterials. Bio Med. Mater. Eng. 2019;30:1–9. doi: 10.3233/BME-181028. [DOI] [PubMed] [Google Scholar]
  • 48.Gong P., Li C., Wang D., et al. Enzyme coordination conferring stable monodispersity of diverse metal–organic frameworks for photothermal/starvation therapy. J. Colloid Interface Sci. 2023;642:612–622. doi: 10.1016/j.jcis.2023.03.178. [DOI] [PubMed] [Google Scholar]
  • 49.Zhang L., Yang Z., He W., et al. One-pot synthesis of a self-reinforcing cascade bioreactor for combined photodynamic/chemodynamic/starvation therapy. J. Colloid Interface Sci. 2021;599:543–555. doi: 10.1016/j.jcis.2021.03.173. [DOI] [PubMed] [Google Scholar]
  • 50.Shao L., Liu D., Liu X., et al. Glucose oxidase and MnO2 functionalized liposome for catalytic radiosensitization enhanced synergistic breast cancer therapy. Biomed. Pharmacother. 2024;179 doi: 10.1016/j.biopha.2024.117402. [DOI] [PubMed] [Google Scholar]
  • 51.Jiang R., Hang L., Li W., et al. Tri-stimulus-responsive hollow mesoporous MnO2 nanocarriers for magnetic-resonance-imaging-guided synergistic starvation/photodynamic therapy of breast cancer. ACS Appl. Nano Mater. 2024;7:1450–1461. [Google Scholar]
  • 52.Hinolan D.S., Cheng P., Kaveevivitchai W., et al. Core@Shell MnO2@MOF nanozymes with accelerated oxygen generation for synergistic chemodynamic and starvation therapy. ACS Appl. Nano Mater. 2025;8:16267–16278. [Google Scholar]
  • 53.Ren H., Bai Y., Liu Z., et al. A multifunctional cascade gas-nanoreactor with MnO2 as a gatekeeper to enhance starvation therapy and provoke antitumor immune response. Acta Biomater. 2024;190:501–517. doi: 10.1016/j.actbio.2024.11.004. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.docx (3.1MB, docx)

Data Availability Statement

All data generated or analyzed during this study are included in this published article and its supplementary information files. Additional datasets are available from the corresponding author upon reasonable request.


Articles from Journal of Pharmaceutical Analysis are provided here courtesy of Xi'an Jiaotong University

RESOURCES