Breaking the durability barrier with d band optimized Cu Fe twin crystal nanozymes for long term suppression of Botrytis cinerea

Abstract

Nanozymes that emulate the catalytic function of natural enzymes often suffer from self-depletion, limiting their long-term efficacy for practical applications. Here, we report an antifungal nanozyme composed of copper-iron nanocrystals (CuFe NCs) engineered through a Twinned Nanozyme Protection Strategy (TNPS). Leveraging the continuous electron-donating role of Cu²⁺ ions, these CuFe NCs sustain enzymatic activity for over a year, achieving prolonged antifungal effects. Computational chemistry analyses reveal that in the Cu-Fe twinned crystal nanozyme system, the asymmetric four-coordination of Fe atoms shifts the Fe 3 d orbital closer to the Fermi level, facilitating electron transfer from Cu to Fe via oxygen atoms. This mechanism enhances the nanozyme’s catalytic activity, ensuring lasting antifungal efficacy. Remarkably, CuFe NCs demonstrate potent antifungal activity by generating endogenous H₂O₂ under oxidative stress within the fungus. Additionally, a multifunctional film (CuFe NCs/sodium alginate/gelatin (Cu-Fe/SL)) is developed, combining water retention with antifungal properties, providing over 19 days of protection for perishable foods and maintaining moisture for more than a week. This antifungal film reduces production costs by 80% compared to conventional antimicrobial films. The durable activity of CuFe NCs offers a promising approach to advancing post-harvest food preservation strategies.

Nanozymes often suffer from self-depletion, limiting their long-term efficacy for practical applications. Here, the authors report Cu-Fe twincrystal nanozymes that leverage the continuous electron-donating role of Cu²⁺ ions to achieve sustain enzymatic activity and prolonged antifungal effects.

Introduction

Food waste represents an escalating global challenge, with the Food Waste Index Report 2024 from the United Nations Environment Programme indicating that each person wastes an average of 132 kilograms of food annually, which accounts for nearly one-fifth of all food produced¹. This problem is especially severe in perishable products like fruits and vegetables, which account for about 40% of total food waste, enough to meet the yearly nutritional needs of 190 million people². Plant pathogens contribute substantially to global food waste by diminishing crop yields^3,4, causing spoilage throughout the supply chain^5,6, and degrading food quality^7,8. Due to the open oxidative nature of food preservation, fungal contamination is a persistent risk at all stages from harvest to processing, particularly from the notorious pathogen Botrytis cinerea (B. cinerea), which adversely affects flavor, aroma, color, and texture while posing food safety risks^9–12. In addition, moisture loss in fresh foods contributes significantly to postharvest waste, as it impacts consumer acceptance and directly affects the sustainability of the food system^13,14. Thus, achieving both durable antifungal efficacy and effective moisture retention has become essential in advancing food preservation technologies to reduce food waste and promote sustainability. Traditional storage methods, such as waxing, refrigeration, and preservative use^15,16, help extend shelf life but are often costly, time-intensive, and can foster resistance, affecting consumer satisfaction^17,18. Additionally, addressing moisture loss caused by packaging material swelling remains a persistent challenge for these techniques, underscoring the urgent need for innovative solutions.

In recent years, attention has shifted to nanozymes, which blur the line between inorganic and biological materials by mimicking the catalytic properties of natural enzymes^19–21. Nanozymes offer an advantage over conventional preservation methods by circumventing specific targeting, thereby reducing the risk of resistance^22–24. Since the pioneering discovery of Fe₃O₄ nanoparticles (NPs) in 2007, recognized for their peroxidase (POD)-like activity, nanozymes have initiated a transformative era with applications spanning pharmaceuticals^25–27, biosensing, wastewater treatment, and biomedicine^26,28. Unlike natural enzymes, Fe₃O₄ nanozymes exhibit superior antimicrobial properties and stability under a wide range of temperatures and pH conditions^29–31. Moreover, their excellent hydrophilicity helps mitigate water loss associated with fungal infections, enhancing preservation effects and prolonging shelf life. These attributes underscore Fe₃O₄ nanozymes’ transformative potential in advancing food preservation, paving the way for broader adoption.

The POD-like activity of Fe₃O₄ NPs is generally attributed to Fenton-like reactions involving surface Fe²⁺ ions under acidic conditions³². However, studies have shown that limited electron transfer from Fe²⁺ to the surface can initiate a phase transition to γ-Fe₂O₃³³, causing rapid oxidation and a concurrent reduction in catalytic activity. In the open oxidative environment of food preservation, this leads to accelerated activity decay, challenging Fe₃O₄ nanozymes’ effectiveness for long-lasting antifungal action and moisture retention. Researchers have pursued modifications to improve Fe₃O₄’s structural stability and longevity. For example, composites like Fe₃O₄@MoS₂-Ag²⁹ and SnS₂@Fe₃O₄ have shown enhanced POD-like properties and antibacterial efficacy³⁴; however, low electron transfer efficiency within these heterogeneous structures remains an obstacle.

In this study, we introduce a Twinned Nanozyme Protection Strategy (TNPS) to enhance POD performance by harnessing the unique advantages of twinned crystal structures to establish an efficient electron transfer channel. We designed copper-iron nanocrystals (CuFe NCs) with a twinned Fe₃O₄-CuO crystal structure, which greatly improved electron transfer efficiency and accelerated Fe³⁺ to Fe²⁺ conversion. This innovation effectively preserves POD-like nanozyme activity, ensuring long-term stability under high humidity and oxidative conditions, with CuFe NCs retaining enzyme activity for over a year under standard storage. Notably, CuFe NCs generate endogenous H₂O₂ within fungal cells, triggering hydroxyl radical (·OH) production through a Fenton-like reaction without requiring external H₂O₂. This process effectively disrupts fungal cell membranes while sparing healthy tissues. Density functional theory (DFT) calculations reveal that Fe atoms in CuFe NCs, serving as catalytic centers, exhibit higher charge density induced by neighboring Cu atoms compared to those in Fe₃O₄. Transcriptomic analyses further illuminate CuFe NCs’ impact on oxidative stress and ergosterol biosynthesis in B. cinerea. To complement this nanozyme, we developed a multifunctional film composed of CuFe NCs, sodium alginate, and gelatin (Cu-Fe/SL) that combines moisture retention with antifungal properties. This film maintains moisture for over 7 d and provides antifungal protection for up to 19 d, effectively preserving food quality. By achieving both antifungal and moisture control functions, this approach marks a significant advance in food preservation technology. Furthermore, this strategy offers a sustainable and cost-effective framework for nanozyme applications across fields such as cancer therapy, antimicrobial dressings, and vaccine delivery, highlighting its broad potential.

Results and discussion

Synthesis and characterization of CuFe NCs

In our previous work, we successfully facilitated efficient heat transfer within the lattice by designing a CuS/MnS twinned crystal, paving a promising path in the realm of nanozymatic technology³⁵. Building on this concept, we have developed an approach for synthesizing CuFe NCs with a twinned crystal structure, enhancing electron transfer efficiency. The synthesis protocol for CuFe NCs is illustrated in Fig. 1a and is based on the TNPS. CuFe NCs were synthesized via a one-pot reaction, with optimal conditions determined by varying the concentrations of Iron (II) chloride tetrahydrate (FeCl₂·4H₂O), copper(II) chloride dihydrate (CuCl₂·2H₂O), and polyvinylpyrrolidone (PVP) to evaluate their effects on inhibiting B. cinerea. The inhibition rates against B. cinerea were recorded at 60.57%, 67.66%, 51.06%, and 9.08% for concentrations of 0.02, 0.04, 0.06, and 0.08 M of CuCl₂·2H₂O and FeCl₂·4H₂O, respectively (Supplementary Fig. 1a). We observed that lower concentrations of CuCl₂·2H₂O and FeCl₂·4H₂O resulted in smaller aggregates due to the formation of fewer nanocrystals. Conversely, a reduction in PVP concentration led to larger aggregates and diminished inhibitory effects, likely due to lower surface energy and less effective aggregation prevention (Supplementary Fig. 1b). Optimal fungal inhibition was achieved at a concentration of 0.04 M for both CuCl₂·2H₂O and FeCl₂·4H₂O, with 50 mg mL^–1 of PVP.

Fig. 1. Schematic illustration and TEM images of Cu_xFe_y NCs.

a The preparation process of Cu_xFe_y NCs. b–i TEM images of Cu NCs, Cu₉Fe₁ NCs, Cu₈Fe₂ NCs and Cu₅Fe₅ NCs, respectively. j–m HRTEM images of Cu NCs, Cu₉Fe₁ NCs, Cu₈Fe₂ NCs and Cu₅Fe₅ NCs, respectively. n EDS mapping images of Cu₈Fe₂ NCs. All experiments were independently repeated three times with similar results.

A series of CuFe NCs was synthesized by adjusting the ratios of Cu to Fe, resulting in the representative Cu_xFe_y NCs (where x = 9, 8, 5; y = 1, 2, 5), which were analyzed for their twinned crystal structures using transmission electron microscopy (TEM). The TEM images revealed uniform rod-like structures across all Cu_xFe_y NCs, with an average diameter of approximately 100 nm (Fig. 1b–i). High-resolution transmission electron microscopy (HRTEM) confirmed the presence of single nanocrystal domains, with the HRTEM image of Cu NCs displaying distinct lattice fringes corresponding to the (002) and (111) planes of CuO NPs (Fig. 1j). Similarly, magnified HRTEM images of Cu₉Fe₁, Cu₈Fe₂, and Cu₅Fe₅ NCs displayed lattice fringes corresponding to both CuO and Fe₃O₄, aligning with multiple characteristic planes of these NPs (Fig. 1k–m). Notably, while the size of Fe₃O₄ NPs in Cu₈Fe₂ NCs did not significantly differ from that in Cu₉Fe₁ NCs, the increased quantity of Fe₃O₄ was critical for enhancing electron transfer efficiency. This enhancement accelerated the Fenton-like reaction catalyzed by Fe₃O₄ nanozymes, positively influencing catalytic efficiency and advancing the reaction process. Energy-dispersive X-ray spectroscopy (EDX) mapping confirmed the homogeneous distribution of Cu, Fe, and O elements within the Cu₈Fe₂ NCs, indicating uniform compound formation (Fig. 1n). Dynamic light scattering (DLS) measurements revealed that Cu₉Fe₁, Cu₈Fe₂, and Cu₅Fe₅ NCs exhibited similar hydrodynamic diameters, averaging 127.8 nm, suggesting consistent physical properties across different compositions (Supplementary Fig. 2). To further confirm the dimensional uniformity of the nanocrystals, we precisely determined the polydispersity index (PDI) for each sample (Supplementary Table 1). All PDI values were consistently below 0.5, demonstrating a high degree of size uniformity in the prepared nanocrystals. However, it is worth noting that both Cu₉Fe₁ NCs and Cu₅Fe₅ NCs exhibit PDI values greater than 0.3 but less than 0.5, indicating their heterogeneous nature. This characteristic may partially account for their relatively weak biological activity.

X-ray diffraction (XRD) analysis confirmed the presence of multiple twin crystals within the same structure, as illustrated in Fig. 2a. All XRD peaks corresponding to the Cu NCs aligned with the CuO phase (JCPDS 04-005-4872), indicating the successful synthesis of pure CuO without Fe₃O₄ contamination. The introduction of Fe resulted in the appearance of distinct Fe₃O₄ peaks (JCPDS 97-025-2267), confirming its effective incorporation into the crystal matrix. The Cu_xFe_y NCs exhibited an intermediate diffraction pattern between those of Cu and Fe NCs, with no extraneous peaks observed, signifying the formation of twin crystals comprising CuO and Fe₃O₄, which was corroborated by HRTEM results. Fourier-transform infrared spectroscopy (FT-IR spectroscopy) revealed a strong absorption peak at 1652 cm^–1, characteristic of the C = O group in PVP, thus confirming its presence. Additional peaks at 1284 cm^–1, 1168 cm^–1, and 879 cm^–1 were attributed to C–N, C–O, and O–O bonds, respectively, suggesting that O–O Bonds may enhance antifungal efficacy through a potential H₂O₂ supply mechanism (Supplementary Fig. 3). The electron transfer properties of CuFe NCs were analyzed using X-ray photoelectron spectroscopy (XPS). This analysis confirmed the elemental composition on the surface of Cu_xFe_y NCs, including carbon, nitrogen, oxygen, Cu and Fe (Fig. 2b). Variations in the chemical states of Cu and Fe at different Cu-Fe ratios were detailed in the XPS spectra (Fig. 2c–d, Supplementary Fig. 4). Shifts in the Fe 2p spectrum indicated changes in the electronic environment due to the integration of CuO. These shifts suggested an increase in electron density around the Fe species, potentially enhancing the Fenton-like reactions. Remarkably, Cu₈Fe₂ NCs displayed the most pronounced binding energy offset, indicating that this specific ratio demonstrates exceptional electron transfer efficiency, with a notable increase in electron density around the Fe element. This characteristic significantly contributes to the sustained performance of the Fe₃O₄-like POD enzyme activity, perfectly aligning with optimal catalytic efficiency in redox cycling.

Fig. 2. The characterization of Cu_xFe_y NCs.

a XRD patterns of Cu_xFe_y NCs. b–d XPS analysis of Cu_xFe_y NCs. e Fe²⁺/ Fe³⁺ ratio of Cu_xFe_y NCs. f Cu⁺/ Cu²⁺ ratio of Cu_xFe_y NCs. g Quantitative distribution of OV in Cu_xFe_y NCs. h UV-vis absorption spectra of TMB + H₂O₂ + Cu_xFe_y NCs. i UV-vis spectrum of MB solution after reaction with Cu_xFe_y NCs.

Our findings suggest that the accelerated electron transfer processes in Cu_xFe_y NCs facilitate the establishment of a catalytic cycle between CuO and Fe₃O₄. This synergy not only promotes the interconversion of metal ions but also enhances POD-like enzyme activity and Fenton-like reactions (Eqs. 1–5). Detailed analysis of the Fe²⁺/Fe³⁺ and Cu⁺/Cu²⁺ ratios revealed that Cu₈Fe₂ NCs exhibit the highest reduction state ratios, 3.7 and 7.9 times greater than those of Cu and Fe NCs, respectively. This significant enhancement can be primarily attributed to the abundant CuO present within the twin structure of Cu₈Fe₂ NCs, which plays a supportive role and facilitates robust redox cycling of Fe₃O₄ through enhanced electron transfer capabilities (Fig. 2e, f). These comprehensive findings underscore the unique properties of CuFe NCs designed based on the TNPS, promoting enhanced catalytic reactions through an electron transfer mechanism, thereby advancing the field of nanozyme research.

The reactions involved are represented as follows:

$${\mathrm{Fe}}^{2+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{Fe}}^{3+}+\cdot \mathrm{\cdot OH}+{\mathrm{OH}}^{-}{k}_1=570{\mathrm{M}}^{-1}{\mathrm{S}}^{-1}$$ 1

$${\mathrm{Fe}}^{3+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{Fe}}^{2+}+\cdot \mathrm{\cdot }{\mathrm{O}}_2\mathrm{H}+{\mathrm{H}}^{+}{k}_2=0.0026{\mathrm{M}}^{-1}{\mathrm{S}}^{-1}$$ 2

$${\mathrm{Cu}}^{+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{Cu}}^{2+}+\cdot \mathrm{\cdot OH}+{\mathrm{OH}}^{-}{k}_1<100{\mathrm{M}}^{-1}{\mathrm{S}}^{-1}$$ 3

$${\mathrm{Cu}}^{2+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{Cu}}^{+}+\cdot \mathrm{\cdot }{\mathrm{O}}_2\mathrm{H}+{\mathrm{H}}^{+}{k}_2=470{\mathrm{M}}^{-1}{\mathrm{S}}^{-1}$$ 4

$${\mathrm{Cu}}^{+}+{\mathrm{Fe}}^{3+}\to {\mathrm{Fe}}^{2+}+{\mathrm{Cu}}^{2+}{\mathrm{\Delta }E}^0=0.6\mathrm{V}$$ 5

The XPS analysis further identified an increase in oxygen vacancies (OV) within CuFe NCs. The O 1 s peak was deconvoluted into three distinct peaks corresponding to C = O, OV, and O–O, located at binding energies of 529.6 eV, 531.2 eV, and 532.5 eV, respectively (Supplementary Fig. 5). The presence of the O–O peak corroborates the FT-IR findings (Supplementary Fig. 3). These combined factors, including H₂O₂ adsorption, OV, and nanoscale size effects, likely account for the presence of O–O bonds in CuFe NCs. Notably, these bonds were consistently observed across various Cu/Fe ratios in Cu_xFe_y NCs, regardless of the precise elemental ratio. The OV content in Cu₈Fe₂ NCs exhibited a significant increase to 24.3% compared to other groups (Fig. 2g). This alteration indicates that defects resulting from electron transfer enhance the substrate’s affinity for H₂O₂ and facilitate the critical redox reactions necessary for catalytic activity. Influenced by both electrons transfer and OV, the proportion of reduced-state Cu and Fe in the Cu₈Fe₂ NCs increased, suggesting enhanced mass transfer efficiency and redox reactions at the nanocrystal surface. Integrating these analyses, the optimal Cu-Fe ratio and the abundance of CuO within the twin structure of Cu₈Fe₂ NCs provide robust support for sustained electron transfer compared to other ratios. Moreover, the increase in OV further promotes the occurrence of Fenton-like reactions. This meticulously designed structural configuration, based on TNPS, significantly enhances the efficiency and effectiveness of the catalytic process.

The POD-like activity of CuFe NCs was demonstrated through the catalytic oxidation of 3,3′,5,5′-tetramethylbenzidine (TMB) in the presence of H₂O₂, resulting in a visible color change (Fig. 2h). Steady-state kinetic analysis yielded kinetic constants, including the Michaelis constant (K_m) and maximum reaction rate (V_m), which were notably superior in Cu₈Fe₂ NCs compared to other groups. The K_m was determined to be 0.54 mM, while the V_m reached 19.20 × 10^–8mM s^–1 (Supplementary Table 2). These values indicate a lower K_m and a higher V_m compared to those reported for other bimetallic nanozymes^36–39, underscoring the high substrate affinity and catalytic efficiency of CuFe NCs. Notably, the Cu₈Fe₂ NCs exhibited a significantly enhanced affinity for substrates, attributed to the increased relative proportion of the active Fe²⁺ center and elevated charge density (Fig. 2c, e). This heightened activity arises from the optimal ratio and spatial arrangement of Fe₃O₄ and CuO within the twin crystal structure, facilitating rapid electron transfer and efficient conversion from Fe³⁺ to Fe²⁺, thereby sustaining robust POD-like activity (Fig. 2h). Further exploration of the relationship between the Cu-Fe ratio and catalytic activity revealed an intriguing phenomenon: despite Cu₉Fe₁ NCs possessing a greater electron supply compared to Cu₈Fe₂ NCs, the overexposure of CuO resulted in a relative reduction of active sites on Fe₃O₄, consequently diminishing its catalytic efficiency, which did not surpass that of Cu₈Fe₂ NCs. Conversely, as the proportion of Cu decreases and that of Fe increases, the diminished electron supply fails to support continuous electron transfer, resulting in further oxidation of Fe₃O₄ and suboptimal catalytic efficiency.

Subsequent investigations focused on ·OH generation during Fenton-like reactions catalyzed by CuFe NCs. Consistent with our expectations, the electron transfer between CuO and Fe₃O₄ within Cu₈Fe₂ NCs significantly amplified Fenton-like catalysis, leading to the highest ·OH generation among the tested groups (Fig. 2i, Supplementary Figs. 6 and 7). This observation supports the hypothesis that CuO provides a stable framework that enhances the functionality of Fe₃O₄ under the TNPS guidance, allowing for sustained catalytic activity. The excellent POD-like activity of CuFe NCs in tandem with their unique structural features positions them as promising candidates for applications in wound healing and other biotechnological fields. Through the optimization of the Cu-Fe ratio and the exploitation of electron transfer mechanisms, CuFe NCs hold significant potential for advanced therapeutic interventions.

To elucidate the catalytic micro-reactions and electron migration dynamics on the surfaces of CuFe NCs, as well as those of Fe₃O₄ and CuO, we employed DFT to model these interactions. Our focus was directed toward Cu₂Fe₈ NCs and Cu₈Fe₂ NCs due to their distinct catalytic behaviors observed in preliminary experiments. These models, alongside those for Fe₃O₄ and CuO, were examined in the (010) and (001) crystallographic orientations (Fig. 3a and Supplementary Figs. 8 and 9). In these systems, H₂O₂ transformation proceeds through two key intermediates, *H₂O₂-H₂O₂ and *OOH-OH (Supplementary Figs. 10–13), which ultimately yield ·OH and ·OOH radicals. Notably, both Cu₂Fe₈ and Cu₈Fe₂ NCs exhibited significantly reduced potential barriers for catalytic reactions at the Fe-Fe active centers within the (010) orientation (Supplementary Figs. 14 and 15). This suggests that the (010) planes represent the predominant catalytic surfaces in CuFe NCs, with Fe atoms serving as the principal catalytic sites. The orientation-specific catalytic enhancement underscores the critical role of crystallographic orientation in optimizing nanomaterial catalytic performance. We mapped the micro-reaction pathways and computed the corresponding potential energy barriers (Fig. 3b–d). Our analysis shows that while the surfaces of Fe₃O₄ and CuO exhibit high reaction barriers exceeding 3.0 eV, the barriers for Cu₂Fe₈ NCs and particularly for Cu₈Fe₂ NCs are significantly lower. The latter displayed barriers below 0.2 eV, indicative of significantly enhanced catalytic reactivity, corroborating the experimental observations of increased catalytic efficiency in Cu₈Fe₂ NCs.

Fig. 3. Theoretical insights into catalytic mechanisms of CuFe NCs.

a The stable structures of Cu₈Fe₂ NCs and Cu₂Fe₈ NCs. b Free energy diagram for catalytic reaction over Cu₈Fe₂ NCs, Cu₂Fe₈ NCs, CuO and Fe₃O₄. The highlights indicate the rate-determining step with the values of the limiting energy barrier labeled. c Volcano plot of the overpotential η vs. the difference between the adsorption free energy of *OOH-OH and *H₂O₂-H₂O₂ for four models. d Adsorption free energy of *H₂O₂-H₂O₂ vs. the difference between the adsorption free energy of *H₂O₂-H₂O₂ and *OOH-OH for four models. e The charge density differences of Cu₈Fe₂ NCs and Cu₂Fe₈ NCs models. f The projected DOS for Cu₈Fe₂ NCs, Cu₂Fe₈ NCs, CuO and Fe₃O₄ models. g Electron localization function of Fe₃O₄ and Cu₈Fe₂ NCs. h The computed partial density of states of the d orbital of Fe atoms for Fe₃O₄ and Cu₈Fe₂ NCs.

Further insight was gained through electron density difference maps, which demonstrated a marked increase in electron density around the active Fe sites in Cu₈Fe₂ NCs compared to Cu₂Fe₈ NCs, indicating more effective electron mediation during catalytic reactions (Fig. 3e). The projected density of states (DOS) diagrams provided additional support: while Fe₃O₄ and CuO exhibited minimal overlap between the conduction and valence bands at the Fermi level, suggesting limited electron transfer efficiency, Cu₈Fe₂ NCs showed well-aligned conduction and valence bands that facilitate efficient charge carrier transfer and enhance the catalytic process (Fig. 3f).

Furthermore, electron localization function (ELF) analysis (Fig. 3g) reveals a distinct coordination environment in Fe₃O₄ and Cu₈Fe₂ NCs. In Fe₃O₄ NCs, Fe ions exhibit both tetrahedral and octahedral coordination with oxygen atoms, with electron density predominantly localized around oxygen. In contrast, Cu₈Fe₂ NCs display an asymmetric bonding arrangement where Fe ions adopt a distorted tetrahedral coordination and Cu ions an asymmetric octahedral coordination with oxygen. This asymmetry induces a Jahn-Teller distortion around the Fe coordination, shifting electron density closer to the Fe centers. Complementing these findings, DOS analysis demonstrates that the Fe 3d energy level (E_d) in Cu₈Fe₂ NCs (−2.23 eV) is positioned nearer the Fermi level compared to Fe₃O₄ NCs (−2.77 eV) (Fig. 3h). This elevation in E_d contributes to increased POD-like activity, thereby boosting ·OH generation via Fenton-like reactions. These theoretical insights underscore the enhanced catalytic performance of Cu₈Fe₂ NCs, facilitated by efficient electron transfer from Cu to O to Fe and reduced activation barriers, highlighting their suitability for industrial and environmental applications. The electronic state alignment in Cu₈Fe₂ NCs, optimized via TNPS, not only supports robust redox cycling but also elucidates the mechanistic foundation for their exceptional Fenton-like activity, rendering them especially promising for nanozyme applications.

Biological activity of Cu₈Fe₂ NCs against B. cinerea

This study explores the antifungal efficacy of CuFe NCs with varying Cu-Fe ratios against B. cinerea, focusing on key indicators such as mycelial growth inhibition and spore germination. The median effect concentration (EC₅₀) values for inhibiting mycelial growth ranged between 0.67 and 11.52 mg mL^–1, while those for spore germination spanned from 0.55 to 2.01 mg mL^–1 (Supplementary Fig. 16). Among the NCs, Cu₈Fe₂ exhibited strong antifungal potency, significantly surpassing both Cu and Fe NCs at similar concentrations. Notably, the EC₅₀ values for Cu₈Fe₂ NCs were markedly lower than those of the single-metal NCs, establishing their enhanced efficacy (Supplementary Figs. 17 and 18). After 7 d of incubation, we observed a concentration-dependent reduction in the mycelial colony area with increasing Cu₈Fe₂ NCs concentration (Supplementary Fig. 17a). The corresponding growth area decreased proportionally (Supplementary Fig. 17b). At 2 mg mL^–1, the colony size matched the original inoculation disk, and no visible hyphal growth was observed. The inhibition rate of mycelial growth reached 100% at ≥2 mg mL^–1 Cu₈Fe₂ NCs (Supplementary Fig. 17c). To further validate the reliability of our results, we additionally measured mycelial biomass using the wet weight method. The findings also confirmed that Cu₈Fe₂ NCs at ≥1.5 mg mL^–1 achieved 100% inhibition of hyphal growth (Supplementary Fig. 17d). Similarly, for spore germination, the number of germinated spores decreased with increasing Cu₈Fe₂ NCs concentration (Supplementary Fig. 18a), and no spore germination was observed at 1.5 mg mL^–1 (Supplementary Fig. 18b). The inhibition rate reached 100% at ≥1.5 mg mL^–1 (Supplementary Fig. 18c). However, it is important to emphasize that these results reflect inhibition, not elimination. Even though no growth was observed under high concentrations, the fungal cells were still present in the system and were not physically removed or lysed. In other words, the degree of inhibition does not directly correlate with the number of residual fungal cells, and 100% inhibition does not indicate complete eradication. The fungi were growth-inhibited, not eradicated. Similar phenomena are frequently reported in previous antifungal studies^17,40. Microscopic analysis revealed that elevated Cu₈Fe₂ NCs concentrations led to pronounced mycelial suppression and aggregation, indicative of potent fungicidal activity. At concentrations of 1.5 and 2 mg mL^–1, spores displayed severe structural deformation, underscoring the NCs’ capability to inflict extensive cellular damage (Supplementary Fig. 19). This concentration-dependent inhibition was further validated through flow cytometry, which showed dose-responsive damage to fungal cell integrity (Supplementary Fig. 20).

TEM illustrated significant structural disruptions within fungal cells post-treatment with Cu₈Fe₂ NCs, including cytoplasmic leakage and vacuolation, signs of advanced cellular stress (Fig. 4a). High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) elemental mapping confirmed disruptions in metal homeostasis, highlighting Cu and Fe infiltration into fungal cells and its subsequent toxicological effects. Propidium iodide (PI) and 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) staining further elucidated the antifungal mechanism of the NCs, indicating elevated reactive oxygen species (ROS) production and substantial membrane damage, particularly at higher NCs concentrations (Fig. 4b). The resultant oxidative stress led to mitochondrial dysfunction and increased permeability, ultimately driving fungal cell death. These observations demonstrate the NCs’ ability to trigger oxidative stress through disruption of cellular and metal equilibrium. Zeta potential measurements provided insights into the electrostatic interactions between the cationic surfaces of Cu₈Fe₂ NCs and the negatively charged fungal cells (Fig. 4c), which facilitates NCs uptake, exacerbating oxidative damage and resulting in the observed antifungal effects.

Fig. 4. Inhibitory activity and mechanism of Cu₈Fe₂ NCs on the growth of B. cinerea.

a Representative TEM and HAADF-STEM elemental mapping images of B. cinerea before and after treatment with Cu₈Fe₂ NCs at a concentration of 1.5 mg mL^–1. b The effect of Cu₈Fe₂ NCs on the cell membrane and ROS accumulation in B. cinerea. c Changes in the zeta potential of B. cinerea before and after treatment with Cu₈Fe₂ NCs. The raw output results from the instrument can be found in the Supplementary Data 2. d Volcano plot of differentially expressed genes between Cu₈Fe₂ NCs-treated and control groups. e–g Enrichment analysis of the GO metabolic pathway showed the most significant enrichment pathway. The column diagrams showed the average values for the assays in triplicate, and error bars represent the standard deviation. *** indicates p ≤ 0.001 and two-tailed Student’s test.

However, we fully acknowledge that potato dextrose broth (PDB) is not a conventional medium for zeta potential measurement. Its unique composition significantly affects zeta potential due to the presence of various electrolytes. As a result, although the zeta potential measured in PDB was relatively low (<10 mV), this does not necessarily indicate poor dispersibility of the nanomaterials themselves. Indeed, as shown in Supplementary Table 1 and Supplementary Fig. 2, the Cu₈Fe₂ NCs exhibited a low polydispersity index (PDI < 0.3) and a uniform particle size distribution, supporting their colloidal stability. In addition, we re-evaluated the zeta potential of Cu₈Fe₂ NCs in water under standard conditions. As shown in Supplementary Fig. 21, the zeta potential increased significantly to over +30 mV, confirming that Cu₈Fe₂ NCs possess excellent colloidal stability in aqueous systems. This result clearly demonstrates that the low zeta potential observed previously was due to the specific properties of the PDB medium, rather than an intrinsic limitation of the nanomaterials themselves.

To further validate this explanation, we conducted an additional experiment by gradually increasing the concentration of PDB in aqueous dispersions of Cu₈Fe₂ NCs. As shown in Supplementary Fig. 21 a concentration-dependent decrease in zeta potential was observed. The absolute zeta potential declined significantly as PDB concentration increased, indicating a strong negative correlation between PDB concentration and colloidal stability. These findings reinforce our conclusion that it is the measurement environment, rather than the nanomaterial, that accounts for the observed decrease in zeta potential under physiological conditions.

In summary, Cu₈Fe₂ NCs exhibit strong antifungal properties by inducing oxidative stress and destabilizing cellular integrity in B. cinerea. Their unique ability to disrupt metal homeostasis and elevate ROS production within fungal cells highlights their potential as innovative antifungal agents.

Transcriptomic analysis of B. cinerea upon treatment of Cu₈Fe₂ NCs

To uncover the molecular mechanisms underpinning the antifungal efficacy of Cu₈Fe₂ NCs, transcriptomic analyses were conducted. This investigation identified 470 significant differentially expressed genes (DEGs), including 224 upregulated and 246 downregulated genes, demonstrating a substantial transcriptional response to NCs exposure (|log2(fold change)| ≥1; p < 0.05), as illustrated in the volcano plots (Fig. 4d and Supplementary Fig. 22). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses further delineated the impact of Cu₈Fe₂ NCs on B. cinerea. GO analysis underscored significant changes in pathways associated with oxidative stress, underscoring ROS’s critical role in the NCs’ antifungal mode of action (Fig. 4e–g). Notably, expression of Bcmdm12, a key gene in mitochondrial autophagy, was significantly reduced, as was the expression of BCIN_05g03230 in the pentose phosphate pathway. Additionally, a pronounced downregulation in Bcgst22, involved in glutathione biosynthesis, pointed to a disruption in antioxidative pathways and a consequent increase in ROS, particularly H₂O₂, likely due to superoxide dismutase activity. The heightened intracellular ROS, especially H₂O₂, appears to drive the formation of highly toxic ·OH via Fenton-like reactions, which are amplified by the presence of Cu and Fe ions introduced by the Cu₈Fe₂ NCs. This oxidative stress, rather than direct toxicity from H₂O₂ alone, is likely responsible for the NCs’ antifungal effects (Supplementary Fig. 23). Moreover, transcriptomic data showed significant upregulation of genes associated with Cu and Fe ion binding and transport (Bclcc5, Bclcc8, Bclcc9, Bcfre7, and Bcthi4), reinforcing the role of these metal ions in mediating oxidative damage through ·OH generation.

Ergosterol, a fundamental component of fungal cell membranes, emerged as a primary target. Genes integral to ergosterol biosynthesis, including Bcerg1 and Bcerg3, were markedly downregulated, indicating that Cu₈Fe₂ NCs treatment disrupted fungal cell membrane integrity and function (Supplementary Fig. 24a). High-performance liquid chromatography (HPLC) further confirmed a substantial reduction in ergosterol content following Cu₈Fe₂ NCs exposure, aligning with the cell membrane damage patterns observed microscopically (Supplementary Fig. 24b). Compared to the control group, the ergosterol production in the cells treated with nanozyme materials decreases, although the extent of the reduction varies significantly. Specifically, the decrease is less pronounced with Cu NPs and Fe NPs, while the reduction is more substantial with the doped Cu_xFe_y NCs, particularly for the Cu₈Fe₂ NCs with nearly a 100-fold reduction in ergosterol expression (peak at 19 min), which exhibit the best nanozyme activity. Additionally, mass spectrometry (MS) analysis corroborated the HPLC findings, revealing a significant decrease in ergosterol levels following Cu₈Fe₂ treatment (Supplementary Fig. 24c–e). Reverse transcription quantitative polymerase chain reaction (RT-qPCR) corroborated these transcriptomic findings, demonstrating high consistency between the two datasets, thereby validating the transcriptional outcomes (Supplementary Table 3 and Supplementary Fig. 25). This comprehensive analysis reveals that Cu₈Fe₂ NCs exert antifungal effects primarily by inducing oxidative stress in B. cinerea. The NCs promote intracellular H₂O₂ accumulation, leading to robust ·OH production via Fenton-like reactions. These reactive hydroxyl radicals further disrupt ergosterol biosynthesis, compromise cell membrane integrity, and ultimately drive B. cinerea cell death.

Long-term antifungal activity of Cu₈Fe₂ NCs

In addition to their potent antifungal activity, the stability and longevity of Cu₈Fe₂ NCs are critical for practical applications, such as preventing fungal infections in perishable goods like fruits and vegetables. To evaluate the enduring antifungal performance of Cu₈Fe₂ NCs, we compared them with conventionally prepared Fe₃O₄ NPs using B. cinerea as a test organism. Initial assessments demonstrated that Cu₈Fe₂ NCs significantly outperformed Fe₃O₄ NPs in antifungal efficacy (Fig. 5a). Over time, Fe₃O₄ NPs exhibited a rapid decline in antifungal properties due to instability and susceptibility to oxidation, with a 50% reduction in efficacy by the 2nd week and a complete loss by the 3rd week. In contrast, Cu₈Fe₂ NCs maintained robust antifungal activity that showed no significant decrease over time, retaining high efficacy for up to one year (Supplementary Fig. 26). Fe₃O₄ nanozymes, despite their potential, face challenges due to phase transitions during oxidation that convert Fe²⁺ to Fe³⁺, leading to a decline in their POD-like activity. This remains a persistent issue in materials science^27,33. To address this, we designed CuFe NCs with a twin crystal structure based on thermodynamically neutral phase structures, where the close interaction between Fe₃O₄ and CuO domains enhances electron transfer and accelerates Fe³⁺ to Fe²⁺ conversion, maintaining a high Fe²⁺ proportion. This innovative approach effectively tackles the industry challenge of stabilizing POD-like nanoenzyme activity over extended periods. Consequently, we used the Fe²⁺/Fe³⁺ ratio as a primary indicator of efficacy and investigated the mechanism in detail.

Fig. 5. Long-lasting antifungal activity of Cu₈Fe₂ NCs against B. cinerea.

a XPS spectrum of Fe 2p and antifungal activity (insert) of Cu₈Fe₂ NCs and Fe₃O₄ NPs at different times. b Comparison of Fe²⁺/Fe³⁺ ratio of Cu₈Fe₂ NCs and Fe₃O₄ NPs at different times. c Comparison of ·OH generating capacity of Cu₈Fe₂ NCs and Fe₃O₄ NPs at different times. d Comparison of antifungal effects of Cu₈Fe₂ NCs and Fe₃O₄ NPs at different times. The column diagrams showed the average values for the assays in triplicate, and error bars represent the standard deviation. *** indicates p ≤ 0.001 and two-tailed Student’s test.

Analytical results revealed a steady decline in the Fe²⁺/Fe³⁺ ratio in Fe₃O₄ NPs over time. Specifically, this ratio dropped from an initial 3.30 to 0.46 by the 2nd week, reflecting an 86% reduction in Fe²⁺ content. By the 4th week, the ratio further decreased to 0.26, ultimately resulting in a phase transition to Fe₂O₃ (Supplementary Fig. 27). This decline in the Fe²⁺/Fe³⁺ ratio directly impacted the POD-like activity of Fe₃O₄ NPs, with complete activity loss by the 4th week, leading to reduced ·OH generation and, consequently, loss of antifungal function. In stark contrast, the TNPS-based twin crystal structure of Cu₈Fe₂ NCs facilitated continuous Fe₃O₄ functionality by enabling efficient electron transport within the twin structure. This design significantly increased the initial Fe²⁺/Fe³⁺ ratio and maintained it over time due to enhanced electron transfer mechanisms. Even after a year, the Fe²⁺/Fe³⁺ ratio remained as high as 4.56, nearly identical to the initial ratio of 4.64. This sustained high Fe²⁺/Fe³⁺ ratio was essential for preserving POD-like enzymatic activity and antifungal efficacy in Cu₈Fe₂ NCs. After one year, Cu₈Fe₂ NCs continued to exhibit high POD activity without signs of decline, ensuring a lasting antifungal effect (Fig. 5b–d and Supplementary Fig. 28). The twin crystal configuration of Fe₃O₄ and CuO within Cu₈Fe₂ NCs thus enabled persistent electron transfer facilitated by TNPS, effectively preserving the catalytic activity required for long-term antifungal action. These findings not only confirm the superior stability and prolonged antifungal properties of Cu₈Fe₂ NCs but also underscore their potential for broader applications, particularly in extending the freshness of fruits and vegetables by slowing down pathological water loss. The Cu₈Fe₂ NCs’ capacity to maintain effective antifungal activity over extended periods without degradation marks a substantial advancement in antifungal treatments, promising enhanced protection against fungal pathogens in agricultural storage settings.

Biosafety evaluation of Cu₈Fe₂ NCs

The potential toxicity of nanomaterials, especially for applications in food preservation, mandates rigorous biosafety assessments to ensure their safety for non-target cells while retaining robust antifungal properties. Thus, we conducted comprehensive in vitro and in vivo evaluations to determine the safety profile of Cu₈Fe₂ NCs. In vitro analysis employed A549 cells, a human alveolar epithelial cell model, exposed to various concentrations of Cu₈Fe₂ NCs. Cytotoxicity was assessed using Lactate Dehydrogenase (LDH) release assays, which revealed no significant increase in LDH levels compared to untreated controls, indicating that Cu₈Fe₂ NCs did not induce cellular damage. Fluorescence microscopy corroborated these findings, as treated cells exhibited no observable morphological changes compared to controls (Supplementary Fig. 29a, b).

In vivo studies involved oral administration of Cu₈Fe₂ NCs to mice via gavage at doses ranging from 0 to 100 mg kg^–1, which is up to 60 times the minimum inhibitory concentration and significantly higher than the amounts typically used in food packaging applications. Throughout the observation period, no mortality or signs of distress were noted; mice across all dose groups displayed normal behavior, alertness, and healthy fur. Furthermore, analyses of body weight and organ coefficients showed no statistically significant differences between treated and control groups, suggesting an absence of adverse physiological effects (Supplementary Fig. 29c, d). Histopathological examination provided deeper insight into potential systemic toxicity. Hematoxylin and eosin (H&E) staining of key organs (heart, liver, spleen, lung, and kidney) revealed no inflammation or histopathological abnormalities, confirming that Cu₈Fe₂ NCs did not elicit toxic effects on these organ systems at the administered concentrations (Supplementary Fig. 29e). This rigorous biosafety evaluation demonstrates that Cu₈Fe₂ NCs are non-toxic to healthy mammalian cells and tissues, supporting their potential as safe, effective antifungal agents in food preservation applications. These findings underscore the promise of Cu₈Fe₂ NCs as a solution for extending the shelf life of perishable goods without compromising biosafety.

Preparation and characterization of antifungal films

Antifungal films serve as semi-permeable barriers, reducing food exposure to air and thus diminishing microbial activity, effectively extending the shelf life of fruits and vegetables. In this study, we engineered a composite film composed of sodium alginate (SA) and gelatin (GL) embedded with Cu₈Fe₂ NCs, termed Cu-Fe/SL film, to enhance enzymatic activity while mitigating pathological and physical water loss in food items. GL, a biodegradable animal protein with excellent gas barrier properties, forms the foundational matrix of our film; however, its inherent brittleness and limited tensile strength are addressed by incorporating polysaccharides like SA⁴¹. While SA is frequently employed in biodegradable films, it typically lacks the robust mechanical and antimicrobial properties required for advanced applications, which we enhanced through the incorporation of Cu₈Fe₂ NCs. The long-lasting antifungal action of Cu₈Fe₂ NCs allows this modified film to effectively curb pathological water loss caused by fungal infections, thereby playing a significant role in food preservation. Interaction between the hydroxyl groups in SA and the reactive groups in GL bolsters the stability and mechanical strength of the composite film⁴². A 1:1 ratio of SA to GL was found optimal, maximizing tensile strength (TS) through effective polymer interactions (Supplementary Fig. 30)^43–45. Initial investigations into varying CuFe NCs concentrations within the SA/GL film matrix revealed that lower concentrations did not sustain antifungal efficacy, while higher concentrations offered negligible additional benefits. Ultimately, an optimal concentration of 1.5 mg mL^–1 CuFe NCs was determined, balancing cost-effectiveness with antifungal performance. Incorporating 1.5 mg mL^–1 of Cu₈Fe₂ NCs not only augmented the antifungal capacity of the film but also improved its mechanical properties. All films exhibited high transparency and uniformity, crucial attributes for food packaging applications, allowing clear visual inspection of contents, which is essential in commercial settings (Supplementary Fig. 31).

Scanning electron microscopy (SEM) analysis showed that while pure SA and GL films displayed smooth surfaces and compact cross-sections, SA/GL and Cu-Fe/SL films exhibited rougher surfaces with enhanced microporosity, analogous to the modifications seen with Ag NPs additions⁴² (Supplementary Fig. 32a). This increase in surface roughness upon embedding Cu₈Fe₂ NCs is likely due to hydrogen bond formation among film components, which helps restrict excessive swelling and maintain effective oxygen barrier properties in high-humidity environments. This effect is reminiscent of the lotus effect seen in plant stomata, where the structured surface minimizes water adhesion and preserves the transpiration barrier function, essential for retaining structural integrity⁴⁶. EDX mapping confirmed a uniform distribution of C, S, O, Cu, and Fe throughout the Cu-Fe/SL film, critical for consistent antifungal performance (Supplementary Fig. 32b). The hydrogen bonding interactions appear to account for observed changes in surface characteristics. The Cu-Fe/SL film demonstrated significantly enhanced break strain compared to other formulations, surpassing the mechanical performance benchmarks required for food packaging applications (Supplementary Fig. 32c). FT-IR spectra showed characteristic peaks for O–H, C=O, and C–O groups integral to the film’s structural matrix, with no new peaks appearing upon embedding Cu₈Fe₂ NCs, indicating no new chemical bonds formed within the composite (Supplementary Fig. 32d).

The film’s water vapor permeability (WVP) was optimized to balance moisture transfer, which is critical for extending food freshness. The ideal humidity range for strawberries is typically 85-95%⁴⁷. The Cu-Fe/SL film demonstrated superior water barrier properties compared to other films (Supplementary Fig. 32e). This improvement stems from the extensive hydrogen bonding network induced by Cu₈Fe₂ NCs, which restricts water vapor permeability, even in high-humidity environments. Following the “tortuous path” concept, the inclusion of Cu₈Fe₂ NCs increases the path length for water diffusion, thereby reducing the diffusion coefficient and enhancing the film’s water barrier capacity. Furthermore, water contact angle measurements indicated an increase in the hydrophobicity of the Cu-Fe/SL surface (Supplementary Fig. 32f). With its robust mechanical properties and potent antifungal action, the Cu-Fe/SL film presents a promising candidate for food preservation applications (Supplementary Fig. 33). Its capacity to minimize water loss and maintain structural integrity in humid conditions highlights its potential for broader use in the food packaging industry, offering an effective solution for prolonging the freshness and quality of perishable goods.

Enhanced longevity and efficiency of Cu-Fe/SL film in fruit preservation

In the commercial sector of fruit and vegetable preservation, maintaining freshness and robust antifungal properties is paramount to preventing fungal contamination effectively. This study introduces Cu-Fe/SL films composed of Cu₈Fe₂ NCs embedded in a matrix of SA and GL, designed to enhance the preservation performance of conventional packaging films. Unlike conventional preservatives that release antifungal agents in a single dose, our Cu-Fe/SL films exhibit prolonged antifungal activity, a critical advancement that bolsters their economic viability for post-harvest storage (Fig. 6a). Strawberries wrapped in Cu-Fe/SL films, isolated to prevent cross-contamination, showed no fungal growth or significant dehydration over extended periods, whereas untreated strawberries developed visible fungal decay within 3 d (Supplementary Fig. 34a). Similarly, strawberries stored with standard antifungal agents or in aseptic chambers displayed rapid dehydration or spoilage within 5 d, while those wrapped in Cu-Fe/SL films retained market quality for over 19 d at room temperature (Fig. 6a). These findings underscore the strong preservation capabilities of Cu-Fe/SL films, which outperform other nanoscale preservative films (Supplementary Table 4). The cost-efficiency of Cu-Fe/SL films was also analyzed. Conventional nanomaterial-based packaging solutions range from $2.78 to $4.45 per square meter, yet the Cu-Fe/SL films were produced at just $0.87 per square meter, making advanced nanomaterials more accessible for daily food packaging applications.

Fig. 6. Effects of Cu-Fe/SL film on the prevention of B. cinerea infection of strawberries.

a Long-lasting freshness diagram of Cu-Fe/SL film. b Sensory evaluation diagram. c Changes in physicochemical properties of strawberries during storage after Cu-Fe/SL film treatment.

The superior moisture retention properties of Cu-Fe/SL films are attributed primarily to their low WVP and robust antifungal performance. The Cu₈Fe₂ NCs promote extensive hydrogen bonding within the matrix, resulting in a reduced WVP that effectively blocks water vapor diffusion. Following Nielsen’s tortuous path theory, the NCs-enhanced matrix creates a complex pathway for diffusing molecules, significantly reducing permeability and minimizing water loss. Additionally, the inherent antifungal activity prevents pathological water loss caused by microbial infestations, preserving both nutritional and sensory qualities in strawberries (Supplementary Fig. 34b).

To evaluate Cu₈Fe₂ NCs in real food systems, we tested their antifungal efficacy against B. cinerea in strawberries, a common and perishable fruit. Untreated strawberries, as well as those wrapped solely in SA or GL films, exhibited B. cinerea growth, while those treated with Cu-Fe/SL films showed no visible fungal infestation even after 8 d (Supplementary Fig. 35). This effect is attributed directly to Cu₈Fe₂ NCs, as SA/GL films alone did not inhibit fungal growth. Sensory quality was also markedly better in strawberries wrapped in Cu-Fe/SL films, which maintained firmness, exhibited minimal wrinkling, and had reduced odor compared to control groups, further demonstrating the film’s effectiveness in preserving visual and textural qualities (Fig. 6b). Quantitative analysis of key physicochemical parameters essential to strawberry quality during storage, including weight loss, total soluble solids (TSS), pH, firmness, color stability, and ascorbic acid retention, confirmed the superior performance of Cu-Fe/SL films (Fig. 6c, Supplementary Fig. 36). Weight loss was significantly lower in Cu-Fe/SL films compared to other groups, likely due to the films’ low WVP, which reduces moisture transfer (Supplementary Fig. 36a). TSS levels, indicative of sugar content, were better retained, suggesting that the film slows metabolic processes associated with ripening (Supplementary Fig. 36b). Moreover, the film helped control pH shifts typically associated with spoilage and maintained the structural integrity of strawberries by mitigating softening (Supplementary Fig. 36c, d). Notably, these films effectively extended the shelf life of strawberries while preserving sensory and nutritional quality. The microenvironment created by Cu-Fe/SL films aligns with the optimal storage conditions for strawberries, namely 0 to 2 °C⁴⁸ and 90–95% relative humidity, helping to suppress microbial growth and reduce respiration rates⁴⁷. This controlled environment minimizes decay caused by excessive moisture and prevents water loss-induced shrinkage, preserving attributes like color and ascorbic acid content, both crucial for market appeal and health benefits (Supplementary Fig. 36e–g). Although electronic sensors excel in precise measurements and data processing, they still cannot match the human sensory system. The human sensory system, particularly the sense of smell, taste, and touch, is highly complex and sensitive, capable of detecting subtle changes and intricate sensory experiences. While electronic sensors can detect specific chemical substances or physical signals, they lack the ability to integrate across senses and are prone to biases when processing complex or ambiguous sensory data. The limitation of electronic sensors lies in their inability to fully understand and replicate the nuanced responses of the human sensory system to multisensory stimuli, which makes them unsuitable for replacing human perception in certain applications. Therefore, in future research, we will aim to substitute electronic sensors with the human sensory system whenever possible.

Safety assessments of Cu-Fe/SL films for food applications included testing metal migration levels in strawberries. Even after 8 d, Cu and Fe levels were well below international safety limits, supporting the films’ non-toxic nature in food packaging (Supplementary Table 5). Strawberries treated with Cu-Fe/SL films retained higher market value, exhibiting minimal physical degradation and superior color and nutritional profiles compared to controls. The films’ comprehensive preservation capabilities address critical challenges in the fresh produce logistics chain, enhancing the commercial value of perishable goods. Further correlation analysis of Cu₈Fe₂ NCs concentrations and quality metrics revealed strong positive relationships between higher NCs content and improved preservation attributes (Supplementary Fig. 37). Notably, weight loss reduction showed a correlation coefficient of 0.97 with NCs concentration, underscoring Cu₈Fe₂ NCs as a critical factor in maintaining food integrity. This correlation highlights the potential for Cu₈Fe₂ NCs in preserving both antifungal efficacy and sensory quality.

In this study, we synthesized CuFe NCs by integrating twin crystals of Fe₃O₄ and CuO using the TNPS method, achieving stable and enhanced POD-like enzyme activity facilitated by efficient electron transfer. These NCs leverage endogenous H₂O₂ produced during fungal respiration to initiate Fe²⁺-mediated Fenton-like reactions, generating ·OH without external H₂O₂, thus enhancing their practicality for food preservation. Our results reveal that CuFe NCs maintain consistent enzyme activity for up to a year under standard storage conditions, highlighting their durability. These nanocrystals exhibit selective antifungal action against B. cinerea, effectively controlling fungal infections while minimizing collateral damage to healthy tissue. This targeted mechanism significantly reduces safety concerns typically associated with nanomaterial applications in the food industry. Furthermore, Cu-Fe/SL films extended strawberry freshness by providing sustained antifungal protection for over 19 d and preserving moisture for over a week under real-world conditions. Notably, the production costs of these films are 80% lower than those of conventional antimicrobial films, offering substantial economic advantages. The development of Cu-Fe/SL films represents a scalable, cost-effective advancement in post-harvest preservation, introducing a class of antifungal agents poised to transform food safety practices. Additionally, this study highlights the potential of CuFe NCs for broader applications, such as cancer therapy, antimicrobial dressings, and vaccine delivery.

Methods

Materials

FeCl₂·4H₂O (purity ≥99%), CuCl₂·2H₂O (purity ≥98%) and PVP (average molecular weight 40,000, purity ≥99%) were sourced from Merck (Darmstadt, Germany). PDB and potato dextrose agar (PDA) were obtained from Shanghai AoBoXing Bio-Tech Co., Ltd. The chemical reagents, including TMB (purity ≥98%), phosphate buffer saline (PBS, pH 7.4), and H₂O₂ (30% w/w aqueous solution)_, were purchased from Macklin Chemical Reagent Co., Ltd. (Shanghai, China). Strawberries of the Dandong “999” were acquired from a local market and used immediately in the experiments. All chemicals and solvents were used as received without further purification.

Fungal strains and culture conditions

B. cinerea (Bio-81601) was sourced from the Query Network for Microbial Species of China. The fungal spores were cultured on PDB medium at 28 °C for 7 d. To harvest the spores, deionized water was added to the plates, and the mycelium was gently scraped using a spreader to suspend the spores in the solution.

Use of cells and animals

Human alveolar epithelial cells (A549) were acquired from the College of Veterinary Medicine, Jilin University. The experimental mice, specific pathogen-free (SPF) 6-week-old BALB/c mice (Mus musculus, outbred strain), were provided by Liaoning Changsheng Biotechnology Co. Mice were housed in standard cages (n = 5 mice per group) under controlled conditions: 12-h light/12-h dark cycle (lights on from 07:00 to 19:00), ambient temperature of 22 ± 2 °C, and relative humidity of 50 ± 10%. Food and water were available ad libitum.

Synthesis of CuFe NCs

For the synthesis of CuFe NCs, CuCl₂·2H₂O (1.6366 g, 9.6 mmol) and FeCl₂·4H₂O (0.4771 g, 2.4 mmol) were dissolved in an aqueous PVP solution (300 mL, 5 mg/mL) and stirred briefly under magnetic stirring. Subsequently, sodium hydroxide (NaOH, 3 mL, 4 M) followed by H₂O₂ (12 mL, 9.8 M) were added sequentially to the above mixture under continuous magnetic stirring. The mixture was then vigorously stirred at room temperature for 30 min to ensure complete reaction. After the reaction was complete, the resulting precipitate was isolated via centrifugation (8944 × g for 10 min), rinsed thoroughly with deionized water for three times to remove residual impurities, and subsequently freeze-dried under vacuum to yield the final CuFe NCs product.

Characterization of CuFe NCs

The morphology of CuFe NCs was analyzed using TEM (2000 FX, JEOL, Japan) and HRTEM (JEM-2010, JEOL, Japan). EDX mapping was performed using a TEM (JEM-2100F, JEOL, Japan) to determine the elemental composition of the nanocrystals. Particle size distribution and zeta potential were measured using DLS and a zeta potential analyzer (Malvern Zetasizer Nano ZS90, Malvern Instruments Limited, England), respectively. The crystal structure of the nanocrystals was investigated by XRD (Ultima + , Rigaku, Japan) using Cu Kα radiation (λ = 1.5406 Å) with a scan rate of 2° per minute over a range of 5–80° in 2θ. XPS (K-Alpha, Thermo Fisher Scientific, USA) was employed to analyze the valences of O, Cu, and Fe, as well as the electronic valence band (EVB) of CuFe NCs. The chemical composition was further characterized by FT-IR spectroscopy (Nicolet iS20, Thermo Fisher Scientific, USA).

POD-like activity and kinetic assay

The POD-like activity of CuFe NCs was assessed using TMB as the substrate in the presence of H₂O₂. Enzyme kinetics were then determined to evaluate the activity parameters of CuFe NCs⁵⁰. The POD-like activity of CuFe NCs was assessed using TMB as the substrate in the presence of H₂O₂. The reaction system (total volume 200 μL) contained 50 mM PBS (pH 4.0), 0.1–2.0 mM TMB (dissolved in Dimethyl sulfoxide (DMSO), final DMSO concentration ≤5%), 1 mM H₂O₂, and 0.1 mg mL^–1 CuFe NCs. Reactions were initiated by adding CuFe NCs and incubated at 37 °C. The oxidation of TMB (to a blue product) was monitored by measuring absorbance at 652 nm using a microplate reader at 30 s intervals for 5 min.

For enzyme kinetic analysis, initial reaction rates (v) were calculated from the linear phase of absorbance changes (first 3 min). To determine Michaelis-Menten constants (K_m) and maximum reaction rates (V_m), the concentration of one substrate (S) was fixed while the other was varied: (1) H₂O₂ concentration fixed at 1 mM, TMB concentration varied (0.1–2.0 mM); (2) TMB concentration fixed at 0.5 mM, H₂O₂ concentration varied (0.1–5.0 mM). Kinetic parameters were derived by fitting the data to the Michaelis-Menten equation (v = V_m[S]/(K_m + [S])) using non-linear regression in GraphPad Prism 9. Each concentration gradient was tested in triplicate, and results are presented as mean ± standard deviation.

Hydroxyl radical production assay

Hydroxyl radical assay kits (A018, Nanjing Jiancheng Bioengineering Institute, China) were used to evaluate the ·OH production efficacy of the Fenton-like reaction at various Cu/Fe ratios. CuFe NCs with different Cu/Fe ratios were substituted for the conventional Fenton reagents. A volume of 0.2 mL of each Cu_xFe_y NCs variant, representing a range of Cu/Fe ratios, was mixed with 0.2 mL of substrate solution. The reaction was quenched by adding a color developer quickly. The absorbance was then measured at 550 nm using a spectrophotometer to determine the ·OH production capacity for each variant.

Additionally, methylene blue (MB) degradation was assessed as an indicator of ·OH production in the Fenton-like reaction. Solutions of CuFe NCs with varying Cu-Fe ratios were mixed with 1 mL of MB solution (500 μg mL^–1). After a 4-h co-incubation period, the mixture was centrifuged (11,180 × g for 10 min) to separate the supernatant. The extent of MB degradation was then quantified by measuring the absorbance across the spectrum from 400 to 800 nm using a UV-Vis spectrophotometer.

Inhibitory activity of Cu₈Fe₂ NCs on B. cinerea

Analysis of mycelial growth effects

The effects on mycelial growth were evaluated using EC₅₀ values, inhibition rates, and microscopic observations⁵¹. After incubation for 3 d at 28 °C, the morphological changes of the mycelia were examined using an inverted microscope (IX71-A21PH, OLYMPUS, Japan) and a TEM (JEM-F200, JEOL, Japan). PBS was used as a negative control. The experiments were conducted in triplicate to ensure reproducibility.

Observation of metals inside mycelium

B. cinerea was incubated with Cu₈Fe₂ NCs at a concentration of 1.5 mg mL^–1 for 3 d at 28 °C, after which the mycelium was washed with PBS. The mycelium was fixed in 2.5% glutaraldehyde overnight, dehydrated through a graded ethanol series, and then prepared into ultrathin sections using an ultramicrotome^52,53. The sections were stained using a standard protocol, and HAADF-STEM (FEI Talos F200X, Thermo Scientific, USA) along with EDS mapping were performed to analyze the distribution of Cu and Fe within the fungal cells.

Effect on cell membrane integrity

The extent of fungal cell membrane damage was evaluated using a modified method from Zheng et al.⁵⁴. Various volumes of Cu₈Fe₂ NCs solution were added to 1 mL of PDB medium containing 0.3 g of mycelium, with PBS as the negative control. After incubation for 8 h at 28 °C, 200 μL of PI (100 μg mL^–1) fluorescent stain was added to the mycelial pellets and further incubated for 10 min. The excess dye was washed off, and the mycelial samples were examined using a fluorescence microscope (BX53, OLYMPUS, Japan). Additionally, flow cytometry assays were performed with a FACSCalibur system (BD Biosciences).

Effect on ROS

Following the pretreatment method described in Section Effect on cell membrane integrity, each sample was treated with 20 μL of 10 mM DCFH-DA, a fluorescent dye for ROS, and incubated for 10 min. After removing the excess dye, the mycelial samples were visualized using a fluorescence microscope.

Zeta potential analysis of fungal surfaces

The surface charge of the fungal samples was evaluated by measuring zeta potential values using a zeta-potential analyzer. B. cinerea spore suspensions were treated with Cu₈Fe₂ NCs and incubated overnight. PBS was used as a negative control. The zeta potentials of the treated fungal samples, untreated controls, and samples with Cu₈Fe₂ NCs only were then analyzed⁵⁵. Each measurement was conducted in triplicate, and the results are reported as mean values.

A 200 μg mL^–1 suspension of Cu₈Fe₂ NCs was first prepared in aqueous solution. The mixture was sonicated using a cell disruptor for 30 min, followed by an additional 30 min of ultrasonication in a water bath. After sonication, particle size was measured using a laser particle size analyzer.

Separately, PDB was prepared at a concentration of 25 mg mL^–1 as a fungal culture medium. This stock solution was then diluted with water to final concentrations of 0.5 mg mL^–1, 1.5 mg mL^–1, and 2.5 mg mL^–1. For each concentration, 200 μg of Cu₈Fe₂ NCs was added per milliliter of medium. The mixtures were subjected to 30 min of sonication using a cell disruptor, followed by another 30 min of bath sonication. Particle size was then determined using a laser particle size analyzer (Malvern Zetasizer Nano ZS90, Malvern Instruments Limited, England).

Transcriptome analysis

Transcriptomic analysis was conducted to identify changes in gene expression, following the protocols outlined in previous studies⁵⁶. B. cinerea and Cu₈Fe₂ NCs were co-cultured at 28 °C for 10 h, after which the samples were immediately flash frozen in liquid nitrogen for 30 min to preserve RNA. PBS served as the negative control. Detailed methods are available in the Supplementary Information.

Ergosterol inhibition assay

The inhibitory effects of CuFe NCs with varying Cu/Fe ratios on ergosterol synthesis were assessed using HPLC (LC-20AB, Shimadzu, Japan). CuFe NCs and B. cinerea were co-cultured at 28 °C for 10 h, then washed and lysed with PBS. Unsaponifiable lipids (NSLs) were isolated using a method adapted from Liu et al.⁵⁷. Briefly, 6 mL of 10% potassium hydroxide (KOH) in methanol was added to each sample for saponification at 80 °C for 100 min in a water bath. The mixture was cooled, and NSLs were extracted thrice with petroleum ether, followed by evaporation to dryness using a rotary evaporator (RE-52A, Yarong, China). The residues were reconstituted and separated on a Shimadzu Wondasil C18-WR column (250 mm × 4.6 mm, 5 μm) with the column temperature maintained at 30 °C. The mobile phase consisted of methanol (A) and water (B) at a ratio of 95:5 (v/v) with a flow rate set at 1.0 mL min^–1. The injection volume was 10 μL, and the detector wavelength was set at 282 nm (the maximum absorption wavelength of ergosterol). Ergosterol levels were quantified using the LC-20AB system, with peak areas measured under specified chromatographic conditions to calculate inhibition rates. Additionally, ergosterol was determined using MS (quadrupole-electrostatic field orbitrap combined ultra-high resolution mass spectrometer, Q Exactive) with the mobile phase consisting of phase A (acetonitrile + 0.1% formic acid) and phase B (0.1% formic acid in water) for further verification.

$$\mathrm{I}\mathrm{nhibition}\mathrm{rates}\left(\%\right)=\frac{\mathbf{peak}\mathbf{area}\mathbf{of}\mathbf{BK}-\mathbf{peak}\mathbf{area}\mathbf{of}\mathbf{CuFe}\mathbf{NCs}}{\mathbf{peak}\mathbf{area}\mathbf{of}\mathbf{BK}}\times 100$$ 6

In vitro safety evaluation of Cu₈Fe₂ NCs

The cytotoxicity of Cu₈Fe₂ NCs was assessed using a modified method adapted from Etienne et al.⁵⁸. A549 cells were cultured for 48 h at 37 °C in a 5% CO₂ atmosphere with saturated humidity, and then detached using 0.25% trypsin. The cells were seeded in 96-well plates at a density of 2.5 × 10⁵ cells mL^–1 and allowed to adhere overnight. Subsequently, Cu₈Fe₂ NCs at concentrations of 0, 1, 2, 4, 6, and 8 mg mL^–1 were added and incubated for 5 h. Cytotoxicity was measured using an LDH Cytotoxicity Assay Kit (C0016, Beyotime, China).

In a parallel experiment, A549 cells were inoculated into 96-well plates at a density of 2 × 10⁴ cells per well and cultured overnight. Different concentrations of Cu₈Fe₂ NCs were then introduced and incubated at 37 °C for 6 h. After the incubation, cell viability and cytotoxicity were assessed using a Calcein/PI Cell Activity and Cytotoxicity Assay Kit (C2015M, Beyotime, China) for 30 min. The metabolic activity of the cells was determined using a laser scanning confocal microscope (FV3000, Olympus, Japan).

In vivo safety evaluation of Cu₈Fe₂ NCs

The animal studies followed to the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. In vivo safety assessments were conducted on mice that were acclimatized for 1 week and randomly divided into five groups, each group consisting of six mice with a male-to-female ratio of 1: 1. DMSO served as the carrier control. The mice received oral doses of Cu₈Fe₂ NCs at concentrations of 25, 50, 75, and 100 mg kg^–1 every 12 h for a total of four doses. Clinical observations included monitoring body weight, behavior, food consumption, coat condition, and mortality. After 7 d, organs including the heart, liver, spleen, and kidneys were collected for routine histopathological evaluation, which involved H&E staining. Microscopic analyses were performed to identify any pathological changes in these organs. At the conclusion of the study, the mice were euthanized under deep anesthesia induced by 5% isoflurane, followed by cervical dislocation to ensure humane endpoints.

Preparation of Cu-Fe/SL films

The Cu-Fe/SL films were prepared using a modified protocol adapted from Li et al.⁴². Initially, 0.2 g of SA and 0.2 g of GL were dissolved in 10 mL of distilled water, and then 0.2 g of glycerol was added as a plasticizer. Subsequently, the film matrix was solidified with the incorporation of Cu₈Fe₂ NCs at a concentration of 1.5 mg mL^–1 to produce the Cu-Fe/SL films.

Characterization of the films

The micromorphology of the film surfaces and cross-sections was examined using SEM (Jasam-6701f, JEOL, Japan). EDS was used to analyze the distribution of elements within the films: Specifically, Cu-Fe/SL membrane samples were cut into 5 mm × 5 mm squares and fixed on carbon-coated copper sample stages using double-sided conductive tape. The target elements selected were C (Kα line, energy = 0.277 keV), S (Kα line, energy = 2.307 keV), O (Kα line, energy = 0.525 keV), Cu (Kα line, energy = 8.040 keV), and Fe (Kα line, energy = 6.400 keV). For each EDS mapping, scanning was performed on the same 100 μm × 100 μm area of the membrane to ensure clear visualization of element distribution. The dwell time per pixel was set to 50 μs, and the cumulative acquisition time for each mapping was 300 s, which minimized noise while preventing sample degradation. Oxford Instruments AZtec 6.0 software was used to process the raw EDS mapping data, with corrections applied for background noise and element peak overlaps to facilitate intuitive comparison of element distribution uniformity. Tensile properties were assessed with a universal tensile testing machine (Instron 5944 testing system, USA). FT-IR was utilized to investigate the interactions among the film components. The static water contact angle was determined using a contact angle meter (OCA20, Dataphysics, Germany).

Water solubility assessment of films

The water solubility of the films was determined using a static method⁵⁹. Film samples measuring 2 × 2 cm² were weighed and then immersed in 60 mL of water at room temperature for 24 h. After immersion, the films were dried in an oven at 105 °C until they reached a constant dry weight. The procedure was repeated three times for each sample. Water solubility was calculated using the following formula:

$$\mathrm{Water}\mathrm{solubility}\left(\%\right)=\frac{\mathbf{initial}\mathbf{dry}\mathbf{weight}-\mathbf{final}\mathbf{dry}\mathbf{weight}}{\mathbf{initial}\mathbf{dry}\mathbf{weight}}\times 100$$ 7

WVP of films

The films were sealed in beakers containing anhydrous calcium chloride, which was used to maintain a 0% relative humidity environment. These beakers were placed in a desiccator with a saturated sodium chloride solution to create a 75% relative humidity environment. The weight of the beakers was recorded periodically until a steady state of weight loss was observed, indicating that equilibrium WVP had been reached.

Long-lasting preservation test

Ripe strawberries of uniform size, free from wounds or blemishes, were selected and sterilized. Immediately following sterilization, the strawberries were randomly assigned into five groups of thirty each. The blank control group consisted of strawberries placed in a regular room environment. The positive control group involved strawberries stored in a sterile room. Strawberries in the fluconazole group were treated with a solution of 1.5 mg mL^–1 fluconazole. In the commercial antifungal group, strawberries were completely wrapped in a commercial silver-based antifungal film. Similarly, in the Cu-Fe/SL group, strawberries were completely wrapped in Cu-Fe/SL film. Sufficient air gaps were left to allow normal air exchange, thus preventing oxygen deprivation from influencing fungal growth. Throughout the preservation period, all groups, except for the positive control group, which was stored in an aseptic chamber, were kept under identical conditions (24 ± 1 °C, 40% relative humidity), and the films remained sealed. Daily photographs were taken to document visual changes in the strawberries.

Biocontrol effect of Cu₈Fe₂ NCs in strawberries

Strawberry preservation experiment

Uniformly sized, ripe strawberries free from wounds or blemishes were selected for the study. Each strawberry was inoculated three times with 10 μL of a fungal spore suspension (1 × 10⁷ spores mL^–1). Subsequently, the strawberries were randomly divided into five groups of thirty for treatment. Group A served as the blank control with no further treatment. Groups B, C, D, and E were treated by wrapping with SA film, GL film, SA/GL film, and Cu-Fe/SL film, respectively. All groups were stored at room temperature (24 ± 1 °C, 40% relative humidity). The freshness of the strawberries was monitored, and sensory evaluations were conducted throughout the storage period.

Sensory evaluations of strawberries

For the sensory evaluation of strawberries, we employed the electronic tongue test. The electronic tongue system (taste sensing system TS-5000Z, Japan) comprises reference electrodes, multichannel lipid/polymer membrane electrodes, an auto-sampler, an electronic unit for data acquisition, and a personal computer with advanced chemometric software (Intelligent Sensor Technology, Inc., Kanagawa, Japan)⁶⁰. The response intensity of each sensor was measured against an Ag/AgCl reference electrode, the most commonly used in this field⁶¹. The potentiometric differences between each coated sensor and the reference electrode determine the intensity values of the measured samples⁶².

Physicochemical analysis of strawberries

The weight loss rate of the strawberries was determined using established methods^63,64. TSS content was measured using a manual refractometer (Bellingham + Stanley, UK)⁶⁵. The pH was determined with a digital calibrated pH meter (Crison Basic 20, Spain) in accordance with AOAC standards⁶⁶. Strawberry firmness was assessed using a Texture Analyzer (CT3, Brookfield, USA). Color changes were evaluated using a Chroma Meter CR-400 (Konica Minolta Sensing, Singapore), focusing on the L* value (lightness) and a* value (redness/greenness). Vitamin C (VC) content was determined using a VC assay kit (A009, Nanjing Jiancheng Bioengineering Institute, China).

Migration of Cu and Fe in strawberries

To determination the Cu and Fe content, 1.5 mL of concentrated nitric acid and 500 μL of perchloric acid were added to strawberry samples weighing between 0.6 to 1 g. The samples were digested to completion in a water bath at 95 °C for 2 h. After digestion, the samples were resuspended in and diluted with distilled water. The concentrations of Cu and Fe were measured using inductively coupled plasma atomic emission spectrometry (ICP-AES, Model 7800, Agilent, USA). Each sample was analyzed in triplicate, and the results were averaged.

Statistical analysis

All experiments were conducted at least three times in parallel, and the experimental results were processed using the professional software SPSS 21 (IBM Statistics, USA). T-test was used for the analysis of variance, and p < 0.05 was considered indicative of a significant difference. Each data is presented as the mean ± SD.

Ethical statement

All procedures involving animals strictly adhered to the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996)⁴⁹. The study protocols were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Jilin University, under protocol number SY202306069.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

L. Liu: Investigation, Methodology, Data curation, Formal analysis, Writing—original draft; H.S. Wang and Q.L. Zhou: Investigation, Data curation, Formal analysis; B. Dong: Conceptualisation, Writing—review and editing, Supervision; X. D. Niu: Project administration, Funding acquisition, Supervision, Writing—review and editing.

Peer review

Peer review information

Nature Communications thanks Da-Wen Sun, who co-reviewed with Hongbin Pu, and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its Supplementary Information files, including Supplementary Data 1 (raw HD/PDI data), Supplementary Data 2 (zeta potential data of B. cinerea), Supplementary Data 3 (zeta potential data of Cu₈Fe₂ NCs in PDB), and Supplementary Data 4 (transcriptomic sequencing data). All data underlying this study are available from the corresponding author upon request. All omics data generated in this study have been deposited in the NCBI database under accession code PRJNA1091389 [https://www.ncbi.nlm.nih.gov/sra]. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-025-67095-8.