Self-disassembling diatomic nanocluster bomb unlock reciprocal synergistic multi-pathway cancer therapy

Abstract

The emerging chemodynamic therapy (CDT), which leverages tumour microenvironment (TME)-specific conversion of H₂O₂ into cytotoxic reactive oxygen species (ROS), suffers from limited efficacy due to low-level endogenous H₂O₂, inefficient ROS generation, and oxidative stress adaptation. Herein, we develop self-reporting CuFe nanodetonators (CuFe NCs) as ‘nanocluster bomb’ to integrate multi-pathway therapeutics for enhanced cancer treatment. Such NCs encapsulate oxygen-vacancy-mediated bandgap engineered CuFe peroxides within TME-responsive shells, which triggers controllable release of peroxides into the tumour cells, enabling high-level endogenous H₂O₂. ROS generation is then amplified through the robust and sustained Cu–Fe dual redox cycling and laser-assisted photothermal effect. CuFe NCs enable trimodal therapy and significantly enhance therapeutic efficacy by 198.8%. Moreover, integrated with dihydrorhodamine 123 (DHR123) as a fluorescence self-reporter, the system allows dynamic intracellular ROS monitoring and precise activation of PTT/PDT. In vitro MDA-MB-468 cells and in vivo BALB/c nude mice validations demonstrate that CuFe NCs potentiate anticancer outcomes through downregulation of Fe-S cluster proteins, upregulation of iron metabolism proteins, and elevated ACSL4 protein levels. This study presents a multivalent dual metal ion-mediated bandgap engineering approach to amplify CDT efficacy and highlights the multi-mechanistic potential of CuFe-based nanotherapeutics.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04245-0.

Introduction

Over 19.5 million people are newly diagnosed with cancer globally each year, and more than 90% of them require effective, targeted, and multimodal therapeutic strategies, such as chemotherapy, radiation therapy, and surgical management [1, 2]. Among these therapies, chemodynamic therapy (CDT) stands out thanks to its high efficiency, non-invasion, and tumour microenvironment (TME)-specific selectivity [3–5]. CDT relies on the conversion of endogenous hydrogen peroxide (H₂O₂) into highly toxic reactive oxygen species (ROS), especially hydroxyl radicals (·OH) [6]. Also, endogenous H₂O₂ induces oxidative distress (a pathological state) at lower concentrations but triggers oxidative eustress (a regulated, adaptive response) at higher levels [7, 8]. However, the low endogenous H₂O₂ concentration, insufficient conversion to ROS, and oxidative eustress-mediated mitigation of ROS damage by promoting adaptive homeostasis and stress tolerance significantly compromise the therapeutic efficacy and remain major barriers to clinical translation [9, 10]. To overcome this long-standing challenge, recent efforts have focused on two primary strategies: either increasing the local H₂O₂ levels, such as targeted release of H₂O₂ precursors at tumour sites, or accelerating the conversion of H₂O₂ into ROS using redox-active catalysts, such as Ag⁺ ions [11].

Fe-based peroxides have attracted particular attention due to their dual functionality: not only serving as a source of H₂O₂ precursors but also catalysing ·OH generation via the Fenton reaction, where Fe²⁺ catalytically convert peroxides to ·OH, thereby amplifying ROS-mediated cytotoxicity [12–14]. Despite these advantages, the overall therapeutic performance of Fe-based CDT systems remains suboptimal, primarily due to intrinsic redox kinetic limitations of Fe-based Fenton reaction [15, 16]. Specifically, the oxidation of Fe²⁺ to Fe³⁺ is relatively rapid (k = 5.7 × 10² M⁻¹·s⁻¹) in the Fenton reaction, whereas the reverse reduction of Fe³⁺ back to Fe²⁺ proceeds extremely slowly (k = 2.6 × 10^− 3 M⁻¹·s⁻¹) [17, 18]. This pronounced kinetic asymmetry leads to insufficient Fe²⁺ regeneration, disrupted ROS sustainable production, and severely hindered Fe²⁺/Fe³⁺ redox cycling under physiologically relevant (low-dose endogenous H₂O₂) conditions [19–21]. These limitations are fundamentally attributed to the large redox potential gap and the intrinsically wide bandgap of Fe-based materials [22, 23], which collectively impose a high energy barrier for electron transfer, thus suppressing the redox kinetics and catalytic continuity of the Fenton process.

Bandgap engineering via elemental doping has emerged as an effective strategy to modulate the electronic structure and narrow redox energy bandgap [24]. Elements such as Zn and Ca have been doped into Fe-based semiconductors to narrow the Fe²⁺/Fe³⁺ redox bandgap and facilitate electron transfer, thereby enhancing catalytic efficiency [25, 26]. However, these dopants exhibit minimal cytotoxicity or even promote tumour growth, failing to contribute additional therapeutic effects [27, 28]. Fortunately, Cu ions offer multiple synergistic functions [29]. Owing to favourable redox kinetics (Cu⁺ → Cu²⁺ <1.0 × 10² M⁻¹·s⁻¹, Cu²⁺ → Cu⁺ = 4.7 × 10² M⁻¹·s⁻¹) that stemmed from their narrow redox energy gap and low activation barrier for electron transfer [30, 31], Cu doping Fe-based peroxides are expect to not only generate abundant Cu⁺ species capable of rapidly reducing Fe³⁺ to Fe²⁺, thereby overcoming the intrinsic slow redox kinetic of Fe²⁺, but also enable efficient regeneration of Cu⁺, establishing a fast and sustainable Cu–Fe dual redox cycle [32, 33]. This system simultaneously addresses two critical challenges: (1) the efficient Cu⁺ regeneration through Cu²⁺-mediated glutathione (GSH) consumption [34], and (2) the generation of abundant Cu⁺ species capable of rapidly reducing Fe³⁺ to Fe²⁺, ensuring a self-propagating Cu-Fe redox continuum [35]. Beyond redox enhancement, Cu ions also initiate cuproptosis, a recently identified form of copper-triggered cell death, adding an orthogonal therapeutic pathway [36, 37]. Furthermore, the 13 kJ mol⁻¹ Cu–Fe formation enthalpy in classical alloys impedes phase mixing, whereas peroxide-mediated metal bridging unlocks new synthetic routes to intermetallic architectures [38]. Moreover, Cu–Fe bimetallic peroxide systems with reduced bandgaps possess enhanced light absorption and photothermal conversion efficiency [39]. These unique features position Cu doping bimetallic peroxide systems as an ideal platform to realize synergistic, multi-pathway tumour eradication.

Herein, we present a self-reporting CuFe nanocapsules (CuFe NCs) functioning as ‘nanocluster bomb’ that integrates multi-pathway therapeutics for synergistic cancer treatment. The nanocapsules are composed of indocyanine green (ICG), Dihydrorhodamine 123 (DHR123), and hydrothermally synthesized bimetallic CuFe peroxide nanodots (CuFe Pro NDs), all encapsulated in poly(D, L-lactic-co-glycolic acid) (PLGA) via one-step self-assembly (Scheme 1). In this design, ICG could enhance the PTT and PDT. DHR123 as self-reporter enables real-time ROS monitoring in vitro. The acidic tumour microenvironment (TME) triggers PLGA degradation, enabling the on-demand release of CuFe Pro NDs. The CuFe Pro NDs simultaneously release H₂O₂ precursors ‘bombs’ and Cu-Fe ions ‘detonators’ into tumour cells, activating an enhanced dual redox cycle and boosting ROS generation for CDT. Under laser irradiation, the CuFe Pro NDs exhibit a strong photothermal effect owing to their narrow bandgap, further accelerating the Fenton reaction and enabling simultaneous CDT, photodynamic therapy (PDT) and photothermal therapy (PTT). The laser, as converter switch, can switch the Cu-dominated to the Fe-dominated dynamic dual redox cycle. Moreover, residual Cu and Fe ions induce cuproptosis and ferroptosis, offering a multi-modal therapeutic cascade against tumour cells. CuFe NCs triggered triple-amplified CDT under laser irradiation, boosting Fenton efficiency by 198.8% and enhancing antitumour outcomes. This study not only demonstrates of bandgap regulation strategy via multivalent metal ions for enhanced CDT, but also highlights the therapeutic versatility of CuFe systems in enabling multi-pathway cancer therapy.

Scheme 1.

Schematic illustration of the synthesis of the self-monitoring organic-inorganic CuFe NCs and the ferroptosis/cuproptosis-oriented enhanced synergistic therapy against tumour. The self-reporting organic-inorganic CuFe ‘nanocluster bomb’ using CuFe NCs encapsulating CuFe Pro NPs, ICG, and DHR123 within PLGA matrix to achieve synergistic therapy through photothermal therapy (PTT), photodynamic therapy (PDT), and chemodynamic therapy (CDT) synergised with ferroptosis and cuproptosis pathways

Results and discussion

The synthesis of CuFe NCs involves two steps (Scheme 1). First, CuFe peroxide nanodots (CuFe-PrO NDs) are synthesised with an average diameter of approximately 4.8 ± 0.9 nm by a liquid stirring method at room temperature (Fig. 1A and S2A). For comparison, Cu peroxide nanodots (Cu-PrO NDs) and Fe peroxide nanodots (Fe-PrO NDs) with average diameters of 3.1 ± 0.6 nm and 2.9 ± 0.7 nm, respectively, are synthesised separately with copper chloride and ferrous chloride as single reagents (Fig. 1B, C and S2B, C). Second, the CuFe-PrO NDs, indocyanine green (ICG), and dihydrorhodamine 123 (DHR 123) are encapsulated within poly lactic-co-glycolic acid (PLGA) via the emulsion-solvent evaporation method to produce CuFe NCs with average diameters of 145.4 ± 29.6 nm (Fig. 1F and S2D). Selected area electron diffraction (SAED) maps and high-resolution transmission electron microscopy (HRTEM) images confirm Cu-PrO NDs possesses a typical polycrystalline structure, while Fe-PrO NDs owns poorest crystallinity (Figure S1). Fast Fourier Transform (FFT) and HRTEM of single CuFe-PrO NDs particle demonstrates that CuFe-PrO NDs has relatively better crystallinity, and the single Cu diffraction pattern indicates Fe atoms exist only in Cu lattice (Fig. 1D and E). The XRD results are consistent with the reported data in Ref 32 and the change in crystallinity agrees with the results of SAED (Figure S4A). For achieving the maximum utilization of each component in CuFe NCs, the encapsulation and loading efficiencies are assayed. With the increase of CuFe-PrO ND concentration, loading efficiency shows linear increase, while encapsulation efficiency gradually declines to a saturation value of 47.1% (Fig. 1G and H). Furthermore, at a fixed CuFe-PrO ND concentration of 400.0 g mL^− 1, a linear correlation between ICG concentration and loading efficiency is observed, with encapsulation efficiency declining to a saturation level of approximately 80.2% (Fig. 1I and J). Thus, CuFe NCs with an optimal ratio are prepared, wherein PLGA acts as the coating agent, encapsulating CuFe-PrO NDs at the concentration of 400.0 g mL^− 1 and ICG at 1.25 mg mL^− 1, with their microstructure shown in Fig. 1K. Elemental mapping images confirm that CuFe NCs consist predominantly of Cu, Fe, and C, indicating successful encapsulation of CuFe-PrO NDs (Fig. 1L). CuFe NCs measure 145.0 ± 29.6 nm in size, with a hydrodynamic diameter of 172.8 ± 2.7 nm (Fig. 1M and S3).

Fig. 1.

Morphology of Cu-PrO NDs, Fe-PrO NDs, CuFe-PrO NDs, and CuFe NCs. TEM images of A, Cu-PrO NDs, B, Fe-PrO NDs, and C, CuFe-PrO NDs. D, elected area electron diffraction (SAED) maps, E, high-resolution transmission electron microscopy (HRTEM) images of CuFe-PrO NDs. F, TEM images of CuFe NCs. The encapsulation efficiency and loading efficiency of G, H, CuFe-PrO NDs, and I, J, ICG. K, Representative TEM images, and L, elemental mapping of CuFe NCs. M, the particle size of Cu-PrO NDs, Fe-PrO NDs, CuFe-PrO NDs, and CuFe NCs. Data are expressed as means ± standard deviation (SD) (n-3 independent measurements)

While Cu-PrO NDs, Fe-PrO NDs and CuFe-PrO NDs all exhibit weak absorption, CuFe-PrO NDs show markedly enhanced absorbance (Fig. 2A). The distinctive absorption peaks at approximately 805 nm of CuFe NCs is attributed to the presence of ICG, which remains unaffected by PLGA modification. According to Tauc function [40], the bandgap calculations are 1.9 eV, 2.3 eV, and 1.6 eV for Cu-PrO NDs, Fe-PrO NDs, and CuFe-PrO NDs, respectively, indicating the bimetallic cooperative structure may shift the conduction band edge (CB) downwards and the valence band (VB) upwards, decreasing the net electronic band gap. (Fig. 2B). Fourier transform infrared spectroscopy (FT-IR) of the Cu-PrO NDs, Fe-PrO NDs, CuFe-PrO NDs, and CuFe NCs reveals peaks at 1282 cm^− 1, 1457 cm^− 1, and 1657 cm^− 1, corresponding to the stretching vibrations of the C-N, C-H and C = O bonds (Fig. 2C) [41]. The peaks at 1093 cm^− 1, 1184 cm^− 1, and 1366 cm^− 1 in CuFe NCs correspond to the stretching vibrations of C-O-C, C-O, and C-N bonds [42, 43]. Furthermore, the valence state of Cu in Cu-PrO NDs and CuFe-Pro NDs is confirmed to be + 2, as the Cu 2p XPS spectrum displays characteristic peaks accompanied by two satellite peaks (Fig. 2D and E) [44]. The Fe valence states in Fe-PrO NDs and CuFe-PrO NDs are determined as + 2 and + 3, corresponding to Fe 2p_1/2 and Fe 2p_3/2, respectively (Fig. 2F and G) [45]. O 1s peaks at 530.4 eV, 531.8 eV, and 532.8 eV refer to C = O (pvp), OV, and O-O (peroxo) species, respectively (Fig. 2H, I and J) [46, 47]. The intensity of C = O bonds diminishes and the concentration of OV, O-O bonds increases in Cu-PrO NDs, Fe-PrO NDs and CuFe-PrO NDs. Interestingly, a 1.4-fold higher oxygen vacancy concentration in CuFe-PrO NDs relative to Cu-PrO NDs, demonstrating that Cu-Fe co-doping serves as an effective strategy for oxygen vacancy modulation. This change of defect concentrations provides direct experimental validation of band structure tunability. Raman spectroscopy verifies that the peak at approximately 850 cm^− 1 is associated with the stretching vibrations of O − O bonds (Figure S4B) [48]. The significant increase in OV concentration of CuFe-Pro NDs indicates that the bimetal can coregulate OV and reduce the band gap, thus CuFe has more reactive active sites, the result consistent with the energy band. Electron paramagnetic resonance analysis after 6 h incubation reveals stable signal intensity of CuFe‑NCs, with spin‑electron density essentially unchanged at 5.3 10¹¹ spins/mm³, corresponding to a 98.7% retention rate (Figure S5). CuFe-PrO NDs exhibit a weak positive surface potential (2.8 mV) and negative cellular endocytosis, with others at 13.7 mV, − 3.6 mV, and − 10.1 mV for Cu-PrO NDs, Fe-PrO NDs, and CuFe NCs, respectively (Fig. 2K).

Fig. 2.

Structure and characterization of Cu-PrO NDs, Fe-PrO NDs, CuFe-PrO NDs, and CuFe NCs. A, UV-vis Absorption spectra, B, bandgap, and C, FT-IR Spectra, of Cu-PrO NDs, Fe-PrO NDs, CuFe-PrO NDs, and CuFe NCs. Cu 2p spectra of D, Cu-Pro NDs and E, CuFe NDs. Fe 2p spectra of F, Fe-Pro NDs and G, CuFe NDs. O 1s spectra of H, Cu-Pro NDs, I, Fe-Pro NDs and J, CuFe NDs. K, Zeta potentials images of Cu-PrO NDs, Fe-PrO NDs, CuFe-PrO NDs, and CuFe NCs. Values were represented as means ± SD of triplicate

Following PLGA degradation, CuFe-PrO NDs further decompose into copper and iron ions, as well as H₂O₂. H₂O₂ release further compensates for H₂O₂ content in TME, thereby enabling a sustainable Fenton reaction. DHR123 as self-reporter exhibits specificity toward OH generated through Fenton reaction via Cu²⁺ and Fe²⁺. The fluorescence intensity at 523 nm increases progressively within 60 min under the excitation of laser at 488 nm wavelengths (Fig. 3A). The maximum fluorescence intensity for CuFe NCs is 2.5 times than that at beginning, suggesting that the self-reporting CuFe NCs significantly boost the OH sustainable generation and monitor changes of OH concentration (Fig. 3B). The photothermal properties of CuFe NCs are investigated under 808 nm laser irradiation at a power density of 1.0 W cm^− 2, with temperature changes recorded using a thermal imager. As shown in Fig. 3D, the CuFe NCs solution temperature is concentration-dependent, indicating a correlation between the increasing temperature and CuFe NCs concentration. Specifically, the maximum temperature reaches 34 °C, 36 °C, 49 °C, 56 °C, 62 °C and 76 °C at the CuFe NCs concentrations of 10 µg mL^− 1, 20 µg mL^− 1, 40 µg mL^− 1, 60 µg mL^− 1, 80 µg mL^− 1 and 100 µg mL^− 1, respectively (Fig. 3C). The rapid temperature elevation in CuFe NCs originates from resonant coupling with 808 nm incident photons, generating surface hot electrons that undergo phonon-mediated thermal relaxation. CuFe NCs maintain effective photothermal output (> 43 °C) over three irradiation cycles despite progressive photobleaching-induced efficiency decay (Figure S6). In contrast, minor temperature change is observed in the PBS control solution, indicating that CuFe NCs could effectively convert light energy into heat. Acidic pH (5.2–6.2) promotes complete CuFe NCs degradation with 5.0-fold higher metal ion release than pH 7.4 at 24 h (Fig. 3E and F). Photothermal activation (808 nm) promotes 1.1-fold faster copper/iron release from CuFe NCs compared to passive diffusion conditions in PBS at pH of 5.2 (Fig. 3G and H). Time-resolved TEM imaging under laser irradiation captured the structural evolution of CuFe NCs (Figure S7). Intact nanocapsules were visualized at the initial stage, followed by partial shell fracturing at 6 h and distinct shell rupture with leakage of internal CuFe-PrO NDs at 12 h. Complete structural disintegration with extensive nanodot release was observed after 24-hour irradiation. Korsmeyer-Peppas fitting of the 24 h Cu release data at the pH value of 7.4, 6.2, and 5.2 yielded release exponents (n) of 0.7 (k = 0.09 h^− n), 0.9 (k = 0.04 h^− n), and 0.7 (k = 0.02 h^− n), respectively [49]. This indicates a non-Fickian transport mechanism for CuFe NCs, suggesting the diffusion process is governed by the decomposition of CuFe NCs. Notably, laser irradiation triggered a mechanistic transition from anomalous transport (n = 0.72, k = 0.09 h^− n) to Fickian diffusion (n = 0.3, k = 0.4 h^− n), indicating laser-enhanced release kinetics and establishing CuFe-PrO NDs as thermo-responsive agents for on-demand drug release in TME-specific combination therapies.

Fig. 3.

Fluorescence and photothermal properties of CuFe NCs. A, Time-resolved PL spectrum and B, the relative intensity of CuFe NCs at 523 nm with continuous reaction up to 60 min and recorded every 1 min. C, The maximum steady-state temperature of CuFe NCs. D, IR thermal images of aqueous suspensions of 0, 10 µg/mL, 20 µg/mL, 40 µg/mL, 60 µg/mL, 80 µg/mL, 100 µg/mL CuFe NCs recorded by an IR camera after 808 nm-laser irradiations for 0–10 min at the power intensity of 1 W cm^− 2 were recorded by a thermal infrared camera every 2 min. The E, Cu and F, Fe release responsive of CuFe NCs at pH 5.2, 6.2, and 7.4. Influence of photothermal properties on G, Cu and H, Fe release responsive at pH 5.5 with 808 nm laser irradiation for 5 min at the power intensity of 1 W cm^− 2. Values were represented as means ± SD of triplicate

To examine the Fenton catalytic activity of CuFe NCs, a colourimetric test based on o-phenylenediamine (OPDA) is conducted [50]. The Fenton catalytic activity is primarily attributed to CuFe-PrO NDs, which used in subsequent studies. CuFe-PrO NDs effectively accelerate OPDA oxidation to oxOPDA in the presence of H₂O₂, yielding a yellow color and 417 nm absorbance (Fig. 4A and B). In contrast, minimal colour change is observed under other conditions. Time-resolved optical absorption spectra for OPDA + H₂O₂ over 15-minutes indicate that oxOPDA absorbance shows a reaction time-dependent increase (Fig. 4C). The Fenton reaction occurs easily in the CuFe-PrO NDs solution at pH 7.4, 6.2, and 5.2 (Fig. 4D). Figure 4E displays that saturation would be reached more quickly with higher CuFe-PrO NDs concentrations, demonstrating their promotion of ROS generation. The CuFe-PrO NDs increase H₂O₂ concentration in the TME via decomposition. Thus, CuFe-PrO NDs exhibit efficient OPDA oxidation at high H₂O₂ concentrations, suggesting high concentration of CuFe-PrO NDs can amplify enhanced Fenton reaction for the first time (Fig. 4F). CuFe NCs also show a typical Fenton response, with increasing reaction intensity over time (Figure S8A and S8B). CuFe NCs under 808 nm laser irradiation induces a 198.8% absorbance increase within 5 min, suggesting that laser can significantly speed up OH generation and second amplified enhanced Fenton reaction (Fig. 4G). Based on our previous research, the superior ROS generation performance can be attributed to the unique CuFe dual redox functionality and synergistic coupling of multiple reaction pathways. Concurrently, surface plasmon-induced hot electron injection into the Fe³⁺-3d orbitals enhances the Fe³⁺/Fe²⁺ conversion efficiency. The GSH depletion response of CuFe-PrO NDs providing direct evidence for the Cu²⁺ to Cu⁺ reduction in CuFe-PrO NDs (Fig. 4H). GSH-dependent Fe²⁺ accumulation confirms Cu⁺-mediated Fe³⁺ reduction, enabling a sustainable Cu-Fe redox cycle (Fig. 4I). 1,3-diphenylisobenzofuran (DPBF) assays demonstrate that CuFe NCs boost O₂⁻ generation through a type II photoreaction mechanism (energy transfer to molecular oxygen generating singlet oxygen, ¹O₂), highlighting promising potential for PDT applications (Fig. 4J and K, and 4L). In CuFe NCs, Fe-3d-dominated valence band maxima with Cu-3d hybridization—engineered via bimetallic coordination and oxygen vacancies—boost Type I photodynamics through electron-transfer-mediated radical generation. Band-gap-engineered CuFe NCs exhibit tailored oxygen vacancies, simultaneously facilitating NIR-triggered synergistic catalytic therapy.

Fig. 4.

Fenton catalytic performances of CuFe-PrO NDs. A, Schematic illustration of the Fenton-like reaction procedure and the ¹O₂ generated process of CuFe NCs. B, UV-vis absorption spectra of the oxidation of oxOPDA catalysed by CuFe-PrO NDs and C, time-dependent absorbance changes under 30 min. D, Influence of various pH values (5.2, 6.2, and 7.4) on Fenton catalytic performances of CuFe-PrO NDs. E, Effect of the concentration of CuFe-PrO NDs (10 g mL^− 1, 20 g mL^− 1, 40 g mL^− 1, 60 g mL^− 1, and 80 g mL^− 1) on Fenton catalytic performances. F, Time-dependent absorbance changes of CuFe-PrO NDs as a result of the catalyzed oxidation of OPDA with various H₂O₂ concentrations (0.25 mM, 0.5 mM, 1 mM, and 2 mM). G, Photothermal enhancement of CuFe-PrO NDs Fenton catalysis assay with 808 nm laser irradiation for 5 min at the power intensity of 1 W cm^− 2. H, GSH depletion under the reduction of different concentrations of CuFe-PrO NDs. I, The detection Fe²⁺ of CuFe-PrO NDs with or without GSH. UV − vis absorption spectra of DPBF by adding J, DPBF + laser, K, CuFe-PrO NDs, and L, CuFe-PrO NDs + laser. Values were represented as means ± SD of triplicate

In vitro performances

Based on the aforementioned excellent Photothermal and Fenton properties of CuFe-PrO NDs, a series of breast cancer cells lines was used to analyse the cytotoxicity and antitumour mechanisms. The cytotoxicity of CuFe NCs against MCF-7, T-47D, MDA-MB-231, and MDA-MB-468 human breast cancer cells lines was assessed via the Cell Counting Kit-8 (CCK-8) assay, with MCF-10 A human normal mammary epithelial cells served as the control group (Fig. 5A). CuFe NCs exhibit concentration-dependent cytotoxicity against MCF-7, T-47D, MDA-MB-231, and MDA-MB-468 cells, with the survival rates at 47%, 30%, 27%, and 26%, respectively, with concentration of 80 µg/mL. In contrast, CuFe NCs show minimal cytotoxicity towards MCF-10 A cells, with a survival rate of 83%. The IC₅₀ results in MCF-7, T-47D, MDA-MB-231, and MDA-MB-468 cells confirm the antitumour efficacy of CuFe NCs against breast cancer cells (Figure S9). Notably, the lowest IC₅₀ value was observed for MDA-MB-468 cells. The invasion and migration rates of MDA-MB-468 cells treated with CuFe NCs for 24 h are 13% and 19%, respectively, implying excellent inhibition of invasion and migration compared to the control group (Fig. 5B and C and S10). During incubation, self-reporter CuFe NCs enter cells via endocytosis, enrich at lysosomes within 4 h, and eventually permeate the entire cell in 8 h (Fig. 5D). To investigate the therapeutic effect of trimodal oncotherapy, 808 nm laser irradiation is used for 5 min at 1 W cm^− 2. Compared with the control groups, a 6-hour coincubation of CuFe NCs (40 g mL^− 1) with laser irradiation exhibits a tumour cell killing rate of 75% (Fig. 5F). Calcein AM/PI co-staining assay further confirm that CuFe NCs effectively kill tumour cells through the combination of PTT and CDT (Fig. 5E). Furthermore, the therapeutic effect of CDT, PTT/PDT, and PTT/PDT/CDT treatment modalities are investigated. As depicted in Fig. 5G, the PTT/PDT/CDT combination treatment group demonstrates enhanced cytotoxicity against the MDA-MB-468 cells. The combination index (CI), calculated via the median-effect method, was 0.42, indicating moderate synergistic interaction between photothermal and chemodynamic therapies in the CuFe NCs. Then, the ROS-generating ability of CuFe NCs was evaluated under various conditions (with/without DHR123 and with/without 808-nm laser) for comparison (Fig. 5H) where DCFH-DA serves as self-reporter for intracellular ROS production. Moreover, maximum fluorescence intensity is achieved when DHR 123 as self-reporter and 808-nm laser as converter switch are combined (Figure S11). Higher ROS levels are attributed to PTT with self-reporting CuFe NCs, significantly increasing CDT efficacy.

Fig. 5.

In vitro antitumour assays of CuFe NCs. A, Cell viability of MCF-10 A, MCF-7, T-47D, MDA-MB-231, and MDA-MB-468 cells treated with CuFe NCs after incubation for 12 h. B, Invasion and C, migration assays of CuFe NCs. D, the intracellular process of CuFe NCs during an 8-hour incubation period. E, Calcein AM/PI co-staining assay with PBS, PBS plus 808 nm laser, CuFe NCs, and CuFe NCs plus 808 nm laser irradiation for 5 min at the power intensity of 1 W cm^− 2. F, Influence of photothermal properties on Cytotoxicity assays for CuFe NCs with 808 nm laser irradiation for 5 min at the power intensity of 1 W cm^− 2. G, the therapeutic effect of CDT, PTT/PDT, and PTT/PDT/CDT was evaluated by Cell viability in MDA-MB-468 cells. H, The intracellular ROS assays by fluorescence images of MDA-MB-468 cells treated with CuFe NCs in different ways. Experiment is expressed as mean ± SD (n = 3 per group) and statistical significance was established using a t-test. (*, P < 0.05; **, P < 0.01; ***, P < 0.001)

MDA-MB-468 cells cells incubated with CuFe NCs and laser-irradiated, the S phase is reduced from 18% to 10% compared to control groups, indicating S phase arrest (Fig. 6A and B). Treatment with CuFe NCs followed by laser irradiation induced early apoptosis in 36.2% of cells, indicating that CuFe NCs can promote tumor cell death through the activation of apoptotic pathways (Figure S12). CuFe-PrO NDs decompose into copper and iron ions in the TME, and iron ion accumulation in tumour tissue induce ferroptosis. GPX4 protein, an essential regulator of ferroptotic cell death, relies on GSH as a cofactor; thus, blocking GSH may indirectly trigger ferroptosis [51]. It is confirmed that intracellular redox reaction between GSH and Cu²⁺ depletes GSH and generates Cu⁺ ions (Fig. 6k). Cells incubated with CuFe NCs and exposed to laser treatment exhibit a lower CSH/GSSG ratio than the control, laser-only, and CuFe NCs-only group (Fig. 6C). CuFe NCs significantly induce lipid peroxidation in tumour cells (20.5%) and exhibit synergistic effects with radiotherapy, elevating peroxidation levels to 43.2% upon combined treatment, confirming their antitumor role through enhanced ferroptosis-associated oxidative damage (Figure S13). GPX4 protein expression decreases with increasing CuFe NCs concentrations (Fig. 6E). Reduced FTH1 protein expression may elevate Fe²⁺ concentrations, suggesting that CuFe NCs promote ferroptosis by regulating iron metabolism (Fig. 6F). In addition, ACSL4 protein expression increases, promoting the incorporation of arachidonic acids into membrane phospholipids and facilitating positive feedback in ferroptosis (Fig. 6G). Hence, CuFe NCs promote ferroptosis by down-regulating GPX4 and FTH1 protein and up-regulating ACSL4 protein (Fig. 6D). To explore additional pathways of tumour cell death, cuproptosis-associated proteins were analysed in response to copper ion enrichment at the tumour site (Fig. 6J). Reduced FDX1 protein expression may elevate Cu⁺ concentrations (more cytotoxicity than Cu²⁺), inhibit Fe-S cluster-containing proteins synthesis, and induce tumour cell death (Fig. 6H). Additionally, Reduced LIAS protein expression supports the occurrence of cuproptosis (Fig. 6I). Overall, CuFe NCs can kill tumour cells through ferroptosis and cuproptosis.

Fig. 6.

Antitumour mechanism of CuFe NCs in vitro. A, Cell cycle changes and B, statistical result of MDA-MB-468 cells after treatment with 40 g mL^− 1 of CuFe NCs and 808 nm laser (1 W cm^− 2). C, Intracellular GSH/GSSG ratios in MDA-MB-468 cells treated with PBS, PBS plus 808 nm laser, CuFe NCs, and CuFe NCs plus 808 nm laser irradiation for 5 min at the power intensity of 1 W cm^− 2. D, Western blot analysis of ferroptosis after different treatments: E, GPX4, F, FTH1, and G, ACSL4. J, Western blot analysis of cuproptosis after different treatments: H, FDX1, and I, LIAS. K, Mechanism diagram of cuproptosis and ferroptosis of CuFe NCs. Each experiment is expressed as mean ± SD (n = 3 per group) and statistical significance was established using a t-test. (*, P < 0.05; **, P < 0.01; ***, P < 0.001)

In vivo performances

To further assess the antitumour activity of CuFe NCs, MDA-MB-468-tumour BALB/c nude mice models are established. Hemolysis assay results show minor red blood cell hemolysis across various CuFe NCs concentrations (10, 20, 40, 60, 80, and 160 g mL^− 1) (Figure S14). In contrast to the significant swelling and aggregation observed in PBS, CuFe NCs exhibit stable hydrodynamic diameters over 48 h in FBS, primarily due to the formation of a stabilizing protein corona that mitigates hydrolysis-driven aggregation (Figure S15). The biodistribution and blood circulation performances of CuFe NCs demonstrate that its mainly distributed in the liver and kidney at 6 h, 12 h, and 24 h post-intravenous (i.v.) injection (Fig. 7C and S16). Tumour uptake of CuFe NCs reaches 4.7 0.6 ID/g at 24 h post-i.v. injection. CuFe NCs exhibit a moderate serum circulation half-life of 0.4 h () and 13.6 h () (Fig. 7D). These results demonstrate that CuFe NCs effectively accumulate in tumour regions and exhibit biosafety for in vitro assays. The PTT/PDT/CDT therapeutic efficiencies of CuFe NCs were evaluated in vivo by i.v. injection (Fig. 7A). Upon reaching a tumour size of approximately 100 mm³, the MDA-MB-468 tumour-bearing mice are randomly divided into four groups (n = 5): PBS, PBS with laser, CuFe NCs, and CuFe NCs with laser. Six hours post-i.v. injection of CuFe NCs (7.3 mg kg^− 1), tumour site temperature reaches 58.2 °C under 808 nm laser irradiation at 1 W cm^− 2, while the temperature of the PBS control groups shows tiny change (Fig. 7B (i), and (ii)). In addition, tumour volume measurements over 14 days indicates continuous tumour growth in PBS and PBS with laser groups. Compared to CuFe NCs-only groups, CuFe NCs with laser treatment significantly suppressed tumour growth (Fig. 7E). Body weight remained stable across the PBS, PBS with laser, CuFe NCs, and CuFe NCs with laser groups (Fig. 7F). Moreover, blood indices show minimal changes after 14 days post-i.v. injection (Figure S19). Hematoxylin and eosin (H&E) stanning images shows cell death is more pronounced in the CuFe NCs with laser group than in the CuFe NCs-only group (Fig. 7H). No significant pathological changes were observed in the heart, liver, spleen, lung, or kidney (Figure S17). The highest TUNEL expression is observed in CuFe NCs with laser groups, with reduced expression in CuFe NCs-only groups and invisible in PBS and laser-only group (Fig. 7I and S18). Ki67 expression shows an inverse pattern compared to TUNEL (Fig. 7J). TUNEL and Ki67 results verifie that laser-treated CuFe NCs could effectively induce apoptosis and inhibit tumour cell proliferation.

Fig. 7.

In vivo antitumour assays of CuFe NCs. A, Scheme illustrating the establishment and treatment of the MDA-MB-468-tumour BALB/c nude mice animal models. B, (i) Temperature imaging and (ii) temperature curves of the MDA-MB-468-tumour BALB/c nude mice after treatment with CuFe NCs under 808 nm laser (1 W cm^− 2, 5 min) irradiation. C, The biodistribution and D, blood circulation performances of CuFe NCs. E, Representative tumour images and F, the corresponding tumour volume after day 14 post-treatment from MDA-MB-468-tumour mice receiving in different treatment groups. G, Optical images of tumour in different treatment groups. H, H&E staining images, I, TUNEL staining images, and J, Ki67 staining images of MDA-MB-468-tumour mice after different treatments. Each experiment is expressed as mean ± SD (n = 5 per group) and statistical significance was established using a t-test. (*, P < 0.05; **, P < 0.01; ***, P < 0.001)

Conclusions

In summary, we have develops a ‘nanocluster bomb’ self-reporting CuFe NCs that encapsulate CuFe Pro NDs, ICG, and DHR123 within PLGA matrix improve endogenous H₂O₂ levels to enable ROS-responsive amplified catalytic therapy, synergizing PTT/PDT/CDT with ferroptosis and cuproptosis pathways. Korsmeyer‑Peppas analysis reveals pH‑dependent non‑Fickian release (n ≈ 0.7‑0.9), governed by material decomposition, while laser irradiation triggers a kinetic transition to Fickian diffusion (n = 0.3), accelerating Cu²⁺/Fe²⁺ release by 1.1‑fold and enhancing •OH generation by 198.8%. Strong therapeutic synergy was quantitatively confirmed (Combination Index < 0.7). The released ions further triggered ferroptosis and cuproptosis via protein‑expression modulation. In vivo studies demonstrated potent tumour suppression in MDA‑MB‑468 xenografts with good biosafety. This work establishes a tenable Cu/Fe bimetallic nanoplatform that integrates multimodal therapy with metal‑ion‑regulated cell death, offering a design strategy for multi‑atomic catalytic medicine.

Materials and methods

Chemicals

Iron(II) chloride (FeCl₂, 99.95%) and Copper (II) chloride (CuCl₂, 99.99%) were purchased from Aladdin Chemistry Co. Ltd. (Shanghai, China). O - Phenylenediamine (OPDA, 98%) Poly(vinylpyrrolidone) (PVP, molecular weight 8000), Poly (D, L lactic-co-glycolic acid), Poly (vinyl alcohol), dichloromethane (CH₂Cl₂), Indocyanine Green (ICG) were received from Macklin Biochemical Technology Co. (Shanghai, China). Copper and Iron standard solutions (1000 µg/mL for each) were purchased from Guobiao Testing and Certification Co., Ltd. (Beijing, China). All chemicals and solvents were of analytical grade and used as received.

Characterization

Transmission electron microscopy (TEM) images were acquired on an FEI Tecnai G2 F20 transmission electron microscope (FEI, USA) operating at 200 kV. UV − vis−NIR extinction spectra were measured using a UV-3600iPLUS UV-vis-NIR spectrophotometer (Shimadzu Scientific Instruments, Japan). Raman spectra were measured using an inVia Qontor laser confocal micro Raman spectrometer (Renishaw, UK). Fourier Infrared spectra (FIR) were measured using a Fourier Transform Infrared Spectrometer Nicolet 6700 (Thermo electron Scientific Instruments, U.S.). The crystal structures were determined by a SmartLab3 kW X-ray diffractometer (Rigaku, Japan) using Cu Kα radiation (λ = 1.5406 Å) operating at 40 kV and 30 mA. The elemental composition of CuFe NCs was determined through inductively coupled plasma-optical emission spectroscopy (ICP-OES) on an Agilent 5100 inductively coupled plasma−optical emission spectrometer (Agilent Technologies, USA). The catalytic performance was detected by using an TACAN enzyme-labeled instrument (Thermo Fisher, USA). The hydrodynamic size and zeta potential measurements were conducted on a Malvern Zetasizer Nano ZSE analyzer (Malvern Instrument Ltd., UK).

Synthesis of CuFe NDs

PVP (2 g) was added to 20 mL of deionized water, and then the 0.01 M CuCl₂2H₂O and 0.01 M FeCl₂ were added to the solution with an appropriate ratio under magnetic stirring. Next, 20 mL aqueous solution containing 0.03 M NaOH and 400 ul H₂O₂ were simultaneously added at 25 C with the stirring continued for 30 min. The resulting product was washed two more times and re-dispersed in deionized water.

Synthesis of CuFe NCs

0.4 mg CuFe NDs, 0.2 mL 0.3 mM ICG and 1 mol PLGA were dispersed in 5 mL dichloromethane, and the resulting dispersion was mixed with 40 mL 0.3% PVA aqueous solution, followed by the ultrasonic homogenization for 4 min. Finally, the solution was stirred at 40 °C for 12 h and centrifuged at 9000 rad/s for 10 min to obtain CuFe NCs.

Photothermal performance of CuFe NCs

To investigate the photothermal performance of PBS and CuFe NCs aqueous dispersions of different concentrations (10, 20, 40, 60 and 80 µg/mL) were illuminated for 10 min by a continuous 808 nm laser (1 W cm^− 2). The temperature of the solutions was monitored by using an IR thermal camera (Guide, IPT384M) and recorded every 2 min.

Fenton-like properties of CuFe NCs

Fenton-like properties of CuFe NCs were first evaluated by the catalytic oxidation of OPDA in the presence of H₂O₂. The absorbance of the color reactions was recorded after a certain reaction time using a UV-vis spectrophotometer. The steady-state kinetic assays were performed at 25 °C in 3 mL PBS solution with CuFe NCs as a catalyst in the presence of H₂O₂ and OPDA. Meanwhile, the kinetic assays of CuFe NCs using H₂O₂ as substrate (0.125, 0.25, 0.5, 0.5, 1 and 2 mM) were also performed. All reactions were monitored by measuring the absorbance after different reaction times.

Cytotoxicity assays

The cytotoxicity of CuFe NCs was investigated by the CCK-8 assay. MCF-7, MCF-10 A, Hela, T-47D, MDA-MB-231, and MDA-MB-468 cells were placed into each well of 96-well plates at a density of 1 × 10⁵ cells mL^− 1 at 37 °C. The medium was removed after 24 h, and 100 µL of new medium containing CuFe NCs at varying concentrations (2.5, 5, 10, 20, 40, and 80 µg mL^− 1) was added and treated with or without the 808 nm laser (1 W cm^− 2). The cells were washed lightly with PBS after 24 h, and 100 µL of medium containing 10 µL of CCK-8 work solution was then added and incubated for 1 h. To this end, the absorbance (450 nm) was measured.

Transwell assays

The cell migration test used a Transwell chamber (0.8 μm pore size, Corning, USA), and Matrigel (BD, USA) was used to coat the upper surface of the transwell chamber for the cell invasion assay. A 0.1% crystal violet solution (Solarbio, China) was used to stain cells that had migrated or invaded through the membrane’s subcellular cytoplasm. The cells that had migrated or invaded the membrane’s underside were ultimately imaged and quantified with a light microscope. Cells were counted using Image-Pro Plus (Media Cybernetics, USA) after they were photographed in five random fields while fixed.

Intracellular ROS assays

The intracellular ROS level was measured using the oxidant-sensitive fluorescent dye DCFH-DA. MDA-MB-468 cells were pretreated as needed and incubated at 37 °C (30 min) with serum-free culture medium containing DCFH-DA (10 mmol), in a volume of 1 mL. Following washing with PBS, HCS performed fluorescence visualization and measured fluorescence intensity in a microplate reader.

Intracellular GSH and GSSG levels assays

In 6-well plates, MDA-MB-468 cells were placed at a density of 1 × 10⁵ cells. Following treatment with CuFe NCs, cells were collected, washed with PBS, and freeze-thawed broken in a 37 °C water bath and liquid nitrogen. Complete GSH/GSSG levels were measured using an assay kit.

Animal experiments

Female Balb/c mice (4–6weeks, Hunan SJA Laboratory Animals Co., Ltd., Hunan, China) were utilized in vivo experiments. The animal use protocol listed below has been reviewed and approved by the Institutional Animal Care and Use Committee (IACUC), The Second Xiangya Hospital, Central South University, China (License Number: 20230682).

In vivo antitumour evaluation

Balb/c mice were subcutaneously injected with MDA-MB-468 cells (1 × 10⁶ cell per mouse). Until the tumour volumes reached 100 mm³ ((tumour volume = (tumour length) × (tumour width)²/2)), Balb/c mice were randomly divided in to 4 groups (n = 5): (1) Control, (2) Control + laser, (3) CuFe NCs, (4) CuFe NCs + laser. CuFe NCs were subcutaneously injected to the wound site on day 0, and the body weight and tumour volumes were recorded every two days. At the end of therapeutic period, mice were euthanised and tumours were collected for hematoxylin-eosin (H&E), monoclonal antibody Ki-67 and terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) staining to evaluate histopathological changes and apoptotic level. Major organs including heart, liver, spleen, lung and kidney of the mice were harvested and stained with H&E for histopathological analysis.

Statistical analysis

All the quantitative data in each experiment were evaluated and analysed by one-way analysis of variance and expressed as the mean values ± standard deviations (n = 3 per group). Student’s t test was used to evaluate the statistical significance of the variance in prism 9 software. Values of *P < 0.05, **P < 0.01 and ***P < 0.001 were considered statistically significant.

Author contributions

BQ conceived the study, developed the methodology, performed the investigation and formal analysis, validated the results, wrote the original draft, and provided resources; ZX curated the data, provided resources, administered the project, supervised the work, and reviewed and edited the manuscript; YZ provided resources and contributed to conceptualization; ST provided resources and contributed to conceptualization; ZZ contributed to conceptualization, methodology, investigation, formal analysis, and reviewed and edited the manuscript; GC contributed to conceptualization and formal analysis; RL contributed to methodology; ZL curated the data, provided resources, administered the project, and supervised the work.

Funding

The authors would like to acknowledge financial support by Hunan Natural Science Fund for Distinguished Young Scholars (No. 2024JJ2076), and the Project of State Key Laboratory of Powder Metallurgy, Central South University, Changsha, China.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Female Balb/c mice (4–6weeks, Hunan SJA Laboratory Animals Co., Ltd., Hunan, China) were utilized in vivo experiments. The animal use protocol listed below has been reviewed and approved by the Institutional Animal Care and Use Committee (IACUC), The Second Xiangya Hospital, Central South University, China (License Number: 20230682).

Consent for publication

No applicable.

Competing interests

The authors declare no competing interests.