Copper-coordinated engineered nanoreactors with self-amplifying O₂ and cascade chem-drug synthesis for enhanced sonodynamic-mediated oncotherapy

Abstract

Efficient synthesis of chemotherapeutic drugs within the tumor microenvironment (TME) holds great potential for enhancing sonodynamic-mediated synergistic oncotherapy. Herein, an engineered nanoreactor composed of Cu²⁺, indocyanine green (ICG), and 1,5-dihydroxynaphthalene (DHN) (termed DCI) with self-amplifying reactive oxygen species (ROS) and cascade chemo-drug synthesis in tumor cells was constructed by coordination-driven co-assembly of sonosensitizer ICG, transition metal copper (Cu²⁺), and nontoxic DHN. Upon intravenous administration, DCI selectively accumulated in tumor tissues and performed excellent photoacoustic (PA) and near-infrared fluorescence (NIRF) imaging capabilities compared to free ICG. Once DCI was internalized into tumor cells, it rapidly disassembled and generated a chemo-drug under the stimuli of acidic lysosome and sonodynamic radiation. The released Cu²⁺ efficiently decomposed endogenous hydrogen peroxide (H₂O₂) into hydroxyl radical (·OH) and O₂ and glutathione (GSH). The generated O₂ efficiency alleviated tumor hypoxia and enhanced sonodynamic therapy (SDT) efficiency. Notably, a large amount of the singlet oxygen (¹O₂) generated by SDT could efficiently oxidize the released nontoxic DHN into hydroxynaphthoquinone (HPN) with higher toxicity. Overall, this work provides a simple, flexible, and effective strategy for oncotherapy.

Graphical Abstract

Tumor microenvironment-activated copper-coordinated engineered nanoreactors with self-amplifying O2 and cascade chem-drug synthesis for enhanced sonodynamic-mediated oncotherapy.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-03890-1.

Introduction

The TME comprises stromal cells, blood vasculature, infiltrating immune cells, various cytokines, etc [1–4]. It exhibits a series of abnormal physiological features, such as acidification, hypoxia, high concentration of GSH and H₂O₂, which are closely related to tumor growth, invasion, and metastasis [5, 6]. TME-activated chemotherapy drugs have recently been considered a promising candidate for enhancing tumor treatment. For instance, tirapazamine, as the bio-reductive agent, can be activated within hypoxic TME, demonstrating approximately 300-fold greater toxicity toward tumor cells under hypoxic conditions compared to non-hypoxic conditions [7]. However, there are still many challenges, including the unsustainable drug release, making it challenging for application of TME-activated chemotherapy drugs including the unsustainable drug release, making it challenging to achieve therapeutic concentrations within TME before reaching dose-limiting toxicity, or insufficient induction of hypoxia, causing low treatment efficiency [8].

As an emerging noninvasive tumor treatment strategy, photodynamic therapy and SDT have recently garnered extensive attention owing to their high tissue-penetration capacity at the targeting sites and minimal damage to the surrounding healthy tissues [9–11]. Under low intensity ultrasound (US), sonosensitizer converts O₂ into ¹O₂ and other ROS, thus inhibiting the tumor growth [12–14]. Then, the generated ROS effectively mediated irreversible damage to tumor cells through multiple mechanisms: (1) depolarization of mitochondrial membrane potential, (2) induction of cytoskeleton contraction, (3) promotion of chromatin condensation, and (4) initiation of DNA fragmentation [15]. Meanwhile, the generation of ¹O₂ can efficiently activate TME-responsive chemotherapy drugs [16]. However, the generation efficiency of ¹O₂ by SDT is limited by the O₂ abundance. Therefore, it is necessary to rapidly replenish the O₂ source within TME to increase the ¹O₂ production rate and thereby activate chemotherapy drugs more effectively.

Transition metal-based nanoreactors have emerged as promising TME-responsive therapeutic platforms leveraging their size-dependent quantum effects and many enzyme-like activities for precision oncology applications [17, 18]. Transition metals, including rhenium, ruthenium, Cu, etc., performed excellent effects on regulating redox balance and catalytic reactions due to their distinctive electron distribution characteristics and spatial structures [19–21]. For instance, Cu²⁺ can consume GSH and catalyze the abundant H₂O₂ within TME into ·OH, which breaks the redox balance within tumor cells [22, 23]. Additionally, it can catalyze H₂O₂ into O₂, thus alleviating hypoxia within TME [24]. Moreover, Cu²⁺ can catalyze the generation of reactive oxygen species (ROS) via Cuproptosis-related pathways, primarily by targeting lipoylated proteins in the tricarboxylic acid (TCA) cycle, leading to proteotoxic stress and ultimately amplifying lipid peroxidation [25–27]. These characteristics of Cu-based nanoreactors enable them to work in synergy with other therapeutic approaches, such as SDT and photodynamic therapy, to enhance tumor treatment.

Herein, we engineered copper-coordinated nanoreactors through a co-assembly strategy employing ICG, Cu²⁺, and DHN. The resulting nanoreactors demonstrate TME-responsive activation, enabling self-amplified oxygen generation and cascaded therapeutic outcomes. Following cellular internalization, DCI undergoes pH/US-dual-responsive rapid disassembly within acidic lysosomes. The released Cu²⁺ catalyzes endogenous H₂O₂ decomposition via Fenton-like reactions to yield ·OH while simultaneously generating O₂, thereby alleviating tumor hypoxia and enhancing SDT efficacy. Meanwhile, US-activated ICG produces cytotoxic ¹O₂, which oxidizes the nontoxic DHN into HPN, thereby amplifying therapeutic specificity. Notably, Cu²⁺-mediated GSH depletion further disrupts redox homeostasis, creating a self-reinforcing oxidative stress. In 4T1 tumor-bearing mice, the DCI nanoreactors exhibit exceptional TME-targeting capabilities and programmable therapeutic efficacy, which achieve synergistic SDT and chemo-dynamic therapy (CDT). This work presents a novel nanoreactor that integrates imaging-guided precision, hypoxia alleviation, and redox dysregulation, offering a robust and versatile strategy for multimodal oncotherapy (as illustrated in Scheme 1).

Fig. 1.

Schematic illustration of DCI synthesis and TME-activated DCI nanoreactors for self-amplifying ROS generation and cascade CDT and SDT

Result and discussion

Construction and characterization of DCI

Based on the efficient therapeutic effect and excellent safety of the coordination-driven co-assembly nanoreactors, we developed a programmable multifunctional metal-organic coordination nanoreactor utilizing nontoxic DHN, the Fenton-like reagent Cu²⁺, and the FDA-approved sonosensitizer ICG as functional components. In this system, Cu²⁺ was stabilized via dynamic coordinate covalent bonds with hydroxyl groups of DHN and the sulfonate groups of ICG, forming a tetra-coordination structure (Fig. 2A). The DCI was prepared by mixing an aqueous solution of ICG and CuCl₂ with a methanol solution of DHN at room temperature. During this process, its color changed from dark green to purple (Fig. 2G). Then, the DCI was obtained by ultracentrifugation and redispersion. When the mass ratio of ICG/Cu²⁺/DHN was 1:1:1, the DCI showed regular spherical shape, structural integrity, and excellent dispersion, which was in line with the Dh obtained by DLS measurement (PDI = 0.1, Dh = 100 nm) (Fig. 2B-C). The zeta potential of DCI was negative (below − 18.8 mV) (Fig. 2E). Additionally, results showed that the particle size and zeta potential of DCI after freeze-thaw was still maintained at about 100 nm and below − 18 mV, respectively, and its PDI was still < 0.3 after freeze-thaw, indicating that DCI had a good freeze-thaw stability (Fig. 2D and S1). These all facilitate the effective enhanced permeability and retention (EPR) effect, ensuring good passive targeting to tumors [28]. The elemental mapping images reflected that the DCI is composed of C, N, O, Cu, and S elements (Fig. 2F). These elements were uniformly distributed throughout DCI, indicating that ICG, DHN, and Cu²⁺ might be successfully co-assembled through dynamic coordination interactions. In addition, ICP-MS and HPLC measurements showed that the encapsulation efficiencies of ICG, Cu²⁺, and DHN were 92.6%, 94.3%, and 92.2%, respectively. That is, in DCI, the mass ratios of ICG, Cu²⁺, and DHN were 1: 1.02: 1, This result was highly consistent with the 1:1:1 feeding ratio set during our synthesis, fully demonstrating the reliability and accuracy of the preparation method and ensuring the nanoreactors have a consistent and well-defined chemical composition. The critical aggregation concentration (CAC) results of DCI also demonstrated a strong coordination-driven self-assembly ability between DHN, Cu²⁺, and ICG (CAC approximately 1.25 µg/mL).

Fig. 2.

(A) (Upper right) Coordination interaction between Cu²⁺ and DHN/ICG; (Lower right) Spatial coordination of Cu²⁺ and DHN/ICG in DCI and the length of the coordination bonds. (B) TEM image of DCI. (C) D_h of DCI (0.2, 0.4, 0.8 mg/mL) dispersed in water. (D) D_h of DCI presents the particle size and PDI of DCI after lyophilization and reconstitution. (Inset: Tyndall effect of DCI). (E) Zeta potential of DCI (0.2, 0.4, 0.8 mg/mL) in water. (F) TEM-EDS mapping images of DCI. (G) The fluorescence intensity ratio of pyrene (I₃₇₄/I₃₈₄) as a logarithmic function of DCI concentration (Color changes of DCI synthesized over different times). (H) Stability of DCI after 6 days of incubation in water, PBS (pH 7.4), DMEM, and 10% DMEM containing Fetal bovine serum, including (I) D_h and Zeta potential of DCI (n = 3). (J) Dispersion of DCI within 24 h. Data are expressed as mean ± SD (n = 3). * p < 0.05; ** p < 0.01; and *** p < 0.001

To further elucidate the coordination mechanism of DCI, X-ray diffraction (XRD), Fourier-transform infrared (FT-IR) spectroscopy, UV-vis-NIR spectroscopy, and X-ray photoelectron spectroscopy (XPS) were measured. As shown in Fig. 3A, DHN and CuCl₂ exhibited distinct crystalline peaks, whereas DCI displayed a broad amorphous peak with the disappearance of certain crystalline peaks from the precursors, suggesting the formation of a metal–organic amorphous structure among DHN, ICG, and Cu²⁺.In the UV-vis-NIR spectrum of DCI (Fig. 3B), the absorption peak of DHN at 298 nm red-shifted to 302 nm, and the absorption peak of ICG at 782 nm blue-shifted to 776 nm, indicating electronic interactions within the coordinated complex. FT-IR spectra (Fig. 3C) further revealed a shift in the hydroxyl stretching vibration of DHN from 3327 cm⁻¹ to 3202 cm⁻¹, and the sulfonate group stretching vibration of ICG shifted from 1032 cm⁻¹ to 1039 cm⁻¹, confirming coordination through hydroxyl and sulfonate groups with Cu²⁺ [29–31]. Additionally, XPS results in Fig. 3D also revealed peaks at 284.71, 285.21, 286.29, 532.0, 534.5, and 933.27 eV, corresponding to the binding energies of C-C/C = C, C-N, C-O, C = O, O-S, and Cu-O bonds. Furthermore, in the Cu 2p3/2 spectrum shown in Fig. 3E, the peak at 933.27 eV assigned to the binding energies of Cu 2p_3/2 could be attributed to the formation of Cu-O bonds [32].

Fig. 3.

(A) XRD spectra of DHN, ICG, CuCl₂, DHN/ICG, and DCI. (B) UV-vis-NIR spectra of DHN, ICG, DHN/ICG, and DCI. The solvent of DHN was methanol. The solvent of ICG, DHN/ICG, and DCI was water. (C) FTIR spectra of DHN, ICG, CuCl₂, DHN/ICG, and DCI. (D) XPS spectrum of DCI. (E) High-resolution Cu 2p spectrum. (F) Fluorescence intensity changes of ICG in water at room temperature under light protection for 6 days. (G) Fluorescence intensity changes of DCI in water at room temperature under light protection for 6 days. (H) UV-vis-NIR absorption of ICG in PBS at room temperature under light protection for 6 days. (I) UV-vis-NIR absorption of DCI in PBS at room temperature under light protection for 6 days. (J) Drug release curve of DHN under different treatments. (K) Drug release curve of Cu²⁺ under different treatments. (L) PA signal and PA image of ICG and DCI (The final concentrations of DCI were 3.125, 6.25, 12.5, 25, and 50 µg/mL). The data are presented as the mean ± SD (n = 3)

In vitro stability and on-demand drug release

The good colloidal and optical stability of nanoreactors is an indispensable prerequisite for their further application in vitro and in vivo. We further evaluated the stability of DCI under simulated physiological conditions. Results showed that there were no significant differences in Dh and zeta potential after being incubated in water, PBS, DMEM, and DMEM (containing 10% fetal bovine serum) for 6 days (Fig. 2H-J), indicating preserved structural integrity under systemic circulation conditions, a critical prerequisite for achieving tumor-targeted accumulation via the EPR effect. Furthermore, the optical stability of DCI was evaluated by UV-vis-NIR and fluorescence spectroscopy. The absorption and fluorescence intensity of ICG in DCI showed no obvious attenuation during storage for 6 days, compared with free ICG (Fig. 3F-I), indicating that the optical stability of ICG was improved after coordinating with Cu²⁺. This phenomenon can be attributed to the partial encapsulation of ICG in the nanoreactors, which prevents excessive contact with the surrounding aqueous environment [33]. On-demand drug release is essential to prevent premature leakage before reaching TME. We further evaluated the in vitro drug release behavior of DCI. As shown in Fig. 3J-K, in PBS (pH 7.4), the cumulative release rates of DHN and Cu²⁺ were limited to 25% and 25.3% after 48 h, respectively. In contrast, under acidic conditions (PBS, pH 5.0), the values increased to 79.7% and 81.4% over the same duration. Most strikingly, US radiation triggered a near-complete release, with DHN and Cu²⁺ reaching 95.6% and 98%, respectively. These results showed that DCI could exhibit remarkable structural stability and superior stimuli-responsive release capability to both pH and US stimulation. These results suggested that DCI underwent rapid structural disintegration within tumor cells following cellular internalization, triggered synergistically by US radiation and the acidic lysosomal microenvironment (pH 5.0), consistent with the in vitro release (Fig. 3J-K). In addition, the drug release could be promoted in the presence of GSH (Figure S2). This on-demand drug release behavior not only enhances the therapeutic effect of the nanoreactors but also reduces the side effects on normal cells or tissues [34]. However, in future studies, the relationship between the coordination bond strength and the US-induced cavitation effect should be analyzed quantitatively. For example, comparing the bond stability or dissociation rate (e.g., via stability constant or △G°) with the US intensity or cavitation dose required for DCI disassembly and drug release should be conducted.

Self-amplification of ROS

It is reported that the TME is closely related to tumor growth, invasion, and metastasis [35]. Therefore, developing the cascade-reaction nanoreactors that can regulate TME is considered a promising strategy for tumor treatment. Consequently, we systematically investigated the ability of DCI to convert H₂O₂ into ·OH using ESR and methylene blue (MB) color change assays. As shown in Fig. 4A, after incubation with MB for 30 min, the MB in the system containing DCI, H₂O₂, and GSH had faded, and its absorbance at 664 nm decreased. Then, we also used 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as the free radical probe to investigate the generation of ·OH. As shown in Fig. 4B, the system containing DCI and H₂O₂ showed a significant DMPO-OH signal peak that was markedly intensified, indicating that DCI exhibits excellent chemiluminescent activity. It is worth noting that the ESR intensity of DMPO-OH was further increased in the presence of GSH. In addition, as shown in Figs. 4C, E, ·OH production exhibited a concentration- and time-dependent enhancement, correlating with elevated GSH levels and prolonged incubation duration.

Fig. 4.

UV-vis-NIR spectra (A) and electron spin resonance (ESR) spectra (B) of MB, MB + H₂O₂, MB + DCI + H₂O₂, and MB + DCI + H₂O₂ + GSH after 30 min of reaction (inset: optical images of MB, MB + H₂O₂, MB + DCI + H₂O₂, and MB + DCI + H₂O₂ + GSH after 30 min of reaction). (C) Chemiluminescent activity of DCI at varying concentrations of GSH. (D) Schematic representation of the Fenton-like reaction involving DCI. (E) Relative chemiluminescent activity of DCI at different time points. (F) SOSG fluorescence spectra of DCI and (G) ICG at different concentrations after 5 min of US irradiation (1 MHz, 0.97 W/cm², 50% duty cycle). (H) SOSG fluorescence spectra of ICG, DHN/ICG, and DCI (ICG concentration = 20 µg/mL) after 30 min of US irradiation (1 MHz, 0.97 W/cm², 50% duty cycle). (I) UV-vis-NIR spectra of PBS (pH 4.5) containing DCI and DHN after 30 min of US irradiation (1 MHz, 0.97 W/cm², 50% duty cycle). (J) ¹H NMR spectra of DHN and HPN. (K) Comparison of ¹H NMR spectra of synthesized HPN and commercial HPN. (L) UV-vis-NIR spectra of PBS (pH 4.5) containing DHN after 30 min of US irradiation (1 MHz, 0.97 W/cm², 50% duty cycle). (M) Fluorescence spectra of DHN and HPN. (N) Yield of HPN after five cycles. Data are expressed as mean ± SD (n = 3). Data are expressed as mean ± SD (n = 3). * p < 0.05; ** p < 0.01; and *** p < 0.001

SDT-mediated oxidation of DHN

To evaluate the SDT efficacy of DCI, ¹O₂ production was quantified via singlet oxygen sensor green (SOSG) assays. As shown in Figs. 4F, G, the ¹O₂ yield demonstrated a concentration-dependent escalation with both DCI and ICG under optimized US radiation parameters (1 MHz, 0.97 W/cm², 50% duty cycle, 5 min), confirming their SDT capability. Furthermore, Fig. 4H shows that DCI generated significantly higher ¹O₂ levels than free ICG under the same sonication conditions. This enhancement likely arises from the improved structural integrity of the DCI nanoreactors post-US exposure, which suppresses energy loss pathways associated with fluorescence emission during sonoluminescence. Collectively, these findings demonstrate that DCI exhibits enhanced SDT performance compared to ICG, as evidenced by its elevated ¹O₂ yields under US radiation. This superior SDT efficacy establishes a mechanistic basis for investigating SDT-triggered oxidative damage in DHN-mediated therapeutic processes.

Evidence has shown that ¹O₂ plays a crucial role in the photo-oxidation of DHN, and its oxidized product HPN has been widely used in anticancer therapies [36]. As mentioned above, DCI could generate a high abundance of ¹O₂ when exposed to US. Based on this, we evaluated the efficiency of the oxidation of DHN. As shown in Fig. 4I, the absorbance of DHN at 298 nm was decreased under US irradiation (1 MHz, 0.97 W/cm², 50% duty cycle), while its absorbance at 419 nm was increased. Additionally, the ¹H NMR spectrum revealed that the product is HPN (Fig. 4J, K). As shown in Fig. 4L, the UV spectrum of PBS (pH 4.5) containing DHN did not change significantly after 30 min of ultrasound irradiation. Interestingly, we observed that the fluorescence of DHN was quenched during its oxidation to HPN (Fig. 4M). Furthermore, the yield of HPN could reach up to 90%, with its cyclic yield remaining nearly unchanged (Fig. 4N). In summary, DCI exhibited high oxidation efficiency, and the oxidation process is cyclically stable.

In vitro cellular uptake and subcellular localization

The confocal laser scanning microscopy (CLSM) and flow cytometry were employed to evaluate the cellular uptake of DCI by 4T1 cells. As shown in Fig. 5A, the intracellular fluorescence intensity of DCI-treated cells exhibited a significant, time-dependent increase. In contrast, no notable time-dependent change in intracellular fluorescence intensity was observed for ICG treatment. Moreover, at the same time, the intracellular fluorescence intensity of DCI-treated cells was substantially higher than that of ICG-treated cells. These results indicate that DCI exhibits significantly enhanced cellular uptake efficiency, which may be attributed to its well-defined nanostructure and larger particle size compared to small molecular agents [37].

Fig. 5.

In vitro cellular uptake and antitumor effects. (A) CLSM images and flow cytometry images of 4T1 cells incubated with DCI for 1, 4, 8, and 12 h. (B) Cell viability of 4T1 cells (left) and L02 cells (right) incubated with different formulations without and with SDT (1 MHz, 0.97 W/cm², and 50% cycle). Cell apoptosis analysis through Annexin V-FITC/PI double-staining (C) and Bio-TEM images (D) of 4T1 cells incubated with different formulations without and with SDT. Mitochondria are shown by the red arrow. (E) CLSM images of ROS and O₂ generated in 4T1 cells incubated with different formulations without and with SDT. Data are expressed as average ± SD value (n = 3), *p < 0.05, **p < 0.01 and ***p < 0.001. Scale bar in TEM images refers to 500 nm. Scale bar in CLSM images refers to 40 µm

Moreover, the detailed internalization mechanism of DCI was investigated using Lysotracker Green (green fluorescence) by CLSM. As shown in Figure S3B, after 1 h of incubation with DCI, only a small amount of red fluorescent signal was detected in acidic lysosomes, indicating minimal internalization of DCI into acidic lysosomes. However, after an additional 8 h incubation, a substantial amount of red fluorescent signal was found in the acidic lysosomes, which was significantly higher than that of free ICG. These results suggest that once DCI is taken up by tumor cells through the endocytosis pathway, a large amount of DCI can be internalized into acidic lysosomes. This finding agrees with previous studies indicating that nano-systems are typically endocytosed by lysosomes after entering cells [38].

In vitro cytotoxicity

We investigated the antitumor effect of DCI by 3-(4,5-dimethylthiazol-2-yl)−2,5-diphenyltetrazolium bromide (MTT) assay. As shown in Fig. 5B, the cell viability of all groups exhibited a concentration-dependent decrease, which could be attributed to the fact that after DCI is efficiently taken up by the cells, the released Cu²⁺ ions generate a large amount of ROS via a Fenton-like reaction, leading to a significant reduction in cell viability (IC₅₀ of 24 µg/mL). Under US irradiation, the cell viability of all groups, except for the DHN-treated group, was further reduced compared with that without US irradiation. Notably, the DCI-treated group with US radiation exhibited optimal antitumor effect, which may be attributed to the fact that DCI combined CDT and SDT. This coordinated system exerts an efficiently programmed synergistic antitumor effect, with an IC₅₀ of 4 µg/mL. Compared to existing copper-based photodynamic therapy platforms [39, 40], DCI demonstrates distinct advantages in multiple aspects. Conventional copper-based photodynamic systems typically rely on a single type of ROS or initiate drug synthesis under static conditions. In contrast, DCI exhibits stronger cytotoxicity at similar drug concentrations due to its self-amplifying ROS generation and cascaded drug synthesis strategy, thereby significantly enhancing antitumor efficacy. Importantly, all dosage forms, including DCI, showed no significant toxic side effects on normal L02 liver cells, demonstrating that our metal-organic coordination nano-systems, designed based on the characteristics of the tumor microenvironment, can effectively switch anti-tumor effects.

We further systematically evaluated the in vitro anti-tumor effects of DCI using Calcein-AM/PI, Annexin V-FITC, and transmission electron microscopy (TEM). Live/dead cell staining by Calcein-AM/PI (Figure S4) and apoptosis analysis by Annexin V-FITC (Fig. 5C) also further demonstrated the significant killing effect of DCI on 4T1 cells, especially after US radiation. Notably, Bio-TEM of 4T1 cells (Fig. 5D) further corroborated that after US radiation, ICG, ICG/DHN, and DCI-treated cells all showed severe cellular crumpling, subcellular organelle damage, or cellular fragmentation. We also observed that when under laser radiation, ICG/DHN and DCI-treated cells had their mitochondria significantly impaired, which was not the case in other groups, suggesting that HPN induces mitochondrial dysfunction and ultimately leads to cell apoptosis.

Self-amplification of intracellular ROS

Intracellular ROS generation was quantified using the DCFH-DA probe. As shown in Fig. 5E and S3C, in the absence of US irradiation, no DCFH fluorescence signals were observed in DHN, ICG, and DHN/ICG-treated cells. In contrast, strong DCFH fluorescence signals were detected in DCI-treated cells, indicating that DCI could generate ·OH through the Cu²⁺-based Fenton-like reaction within the cells. Furthermore, no DCFH fluorescence signal was observed in DHN-treated cells under US irradiation, while stronger DCFH fluorescence signals were generated in ICG and DHN/ICG-treated cells under US irradiation. This may be because ICG could produce ¹O₂ under US irradiation. Notably, the fluorescence intensity of DCFH in DCI-treated cells was 2.53 times greater than in DHN/ICG-treated cells under US irradiation. This can be attributed to the efficient cellular uptake, which resulted in substantial accumulation of DCI in the tumor cells. Under the combined influence of lysosomal acidity and US irradiation, the degradation of DCI released ICG and Cu²⁺, which then generated large amounts of ¹O₂ and ·OH through highly efficient SDT and CDT effects, respectively. This led to a dramatic increase in ROS abundance. These results were consistent with the results obtained by flow cytometry (Figure S6). To further validate the SDT effect of DCI, the intracellular O₂ content was investigated using the [Ru (DPP)₃] Cl₂ (RDPP) probe by CLSM. As shown in Fig. 5E, it showed lower fluorescence signals without US irradiation. In contrast, RDPP in ICG, DHN/ICG, and DCI-treated groups with US irradiation had higher RDPP fluorescence, and DCI-treated cells had the strongest fluorescence signals. These results not only confirm the superior SDT effect of DCI but also demonstrate its ability to rapidly deplete O₂ within tumor cells during ultrasound treatment [35, 41]. Then we investigated the ¹O₂ level in 4T1 cells with different treatments. As shown in Figure S5, after treated by US irradiation, the intracellular ¹O₂ level was increased and the DCI + US group showed the highest ¹O₂, which indicated that US irradiation could efficiency convert O₂ to ¹O₂.

Intracellular ROS accumulation induced mitochondrial dysfunction. Therefore, the mitochondrial membrane potential (ΔΨm) and the expression level of Bax, Bcl-2, and cleaved caspase-3 were investigated. As shown in Figure S7, the results of the control group showed that the vast majority of cells maintained high ΔΨm (only 6.70% of cells exhibited depolarization), confirming the reliability of the experimental system. Treatment with DHN or the photosensitizer (ICG) did not induce significant changes in ΔΨm, indicating negligible intrinsic toxicity of these drugs. The key finding was that DCI nanoreactors under ultrasound (US) activation exhibited the most potent and widespread collapse of ΔΨm, far surpassing all other groups (including ICG combined with US). As shown in Figure S8, the drug could upregulate the expression of pro-apoptotic proteins (Bax, cleaved caspase-3) and downregulate the expression of anti-apoptotic protein (Bcl-2), thus activating the mitochondrial apoptotic pathway. These results strongly suggest that DCI serves as a highly efficient sonosensitizer, and its combination with SDT) specifically targets and disrupts mitochondrial function, thereby effectively inducing apoptosis. This mechanism is likely critical for its antitumor efficacy.

In vivo NIRF and PA imaging

The non-invasive imaging based on the combination of NIRF and PA imaging has been widely used in tracing the drug in vivo and guiding tumor diagnosis and treatment due to its high photosensitivity, resolution, contrast, deep tissue penetration, etc [37]. As a classical sonosensitizer approved by the FDA, ICG can be used as an NIRF and PA imaging probe. It is worth mentioning that our study showed that DCI had a superior photostability effect and PA imaging performance in vitro, compared to ICG (Fig. 3L). The enhanced PA signal of DCI at equivalent concentrations can be attributed to factors as follows: First, free ICG is highly unstable and prone to degradation in aqueous solution, whereas its formation into the DCI nano-complex with Cu²⁺ and DHN significantly improves its stability, thereby showing a stronger PA signal. Second, the polymethine chain of free ICG undergoes significant rotational and vibrational relaxation, which dissipates energy. In contrast, this molecular motion is restricted within the rigid DCI nanostructure, leading to more efficient energy absorption and a stronger PA signal. Third, free ICG tends to form H-aggregates in water, which typically quenches its PA signal. This undesirable aggregation is suppressed in the stable DCI complex.

To further explore the in vivo biological distribution and tumor enrichment of DCI, DCI and ICG were injected into 4T1 tumor-bearing mice via the tail vein. Figure 6A and B show the PA signal intensity of free ICG and DCI in tumors, heart, liver, spleen, lungs, and kidneys of mice after injection. A stronger PA signal indicates a higher level of accumulation. As shown in Fig. 6A, the PA signals of both ICG and DCI in tumors reached their peak at 8 h, then gradually decreased, indicating maximum accumulation at 8 h post-injection. At each time point, the PA signal intensity of DCI was significantly higher than that of ICG, especially at 24 h, where the DCI group still showed markedly stronger PA signals than the ICG group. Additionally, as shown in Fig. 6B, at 24 h post-injection, the PA signal intensity of DCI was only slightly higher than that of ICG in the heart, liver, spleen, lungs, and kidneys; however, it was significantly stronger in tumors-approximately twice that of ICG. These results demonstrate the excellent tumor-targeting accumulation ability and long-circulating stability of DCI in vivo. Consistent results were obtained from in vivo fluorescence imaging experiments. As shown in Figs. 6C and D, at each time point, the fluorescence intensity of DCI at the tumor site was higher than that of free ICG. At 24 h post-injection, almost no fluorescence was detected in the tumors of the ICG group, while the DCI group still exhibited significantly stronger fluorescence intensity than the ICG group. These experimental results further confirm the outstanding tumor-targeting accumulation capability and long-circulating stability of DCI in vivo. Therefore, DCI enables in vivo NIRF/PA dual-modal imaging and selectively accumulates at tumor sites via EPR-mediated passive targeting. This provides an early opportunity for imaging-guided, multi-modal non-invasive tumor diagnosis and therapy.

Fig. 6.

Antitumor effects of programmable synergistic therapy with DCI in vivo. In vivo PA (A and B) and NIRF imaging (C and D) of nude mice following injection of ICG and DCI. (E) Schematic illustration of in vivo programmable synergistic therapy in nude mice with DCI and representative photographs (the dashed red line indicates the tumor region) (F). Tumor volume change curves (G) and (H) tumor growth inhibition rate of 4T1 tumor-bearing mice during treatment of different formulations. Tumor weight of tumors (J), Survival rate (K), and body weight (L) change of nude mice during treatment of different formulations. Data are expressed as average ± SD value (n = 5), *p < 0.05, **p < 0.01 and ***p < 0.001

The biosafety in vivo

In vivo, the body weights of mice showed no remarkable differences among groups (Fig. 6L). Furthermore, in vivo imaging revealed no evident accumulation or retention of the drugs in major organs, and histopathological examination (H&E staining) further confirmed the absence of obvious damage to these organs (Figure S9 and 10). The hemolytic properties of various drug were evaluated, and their effects on liver and kidney functions in mice were assessed by measuring serum levels of ALT, AST, BUN, and creatinine after administration. All drugs (DHN, ICG, DHN/ICG, DCI) exhibited hemolysis rates below 5%, showing no significant difference compared to the PBS group. Furthermore, this result confirms their excellent hemocompatibility and absence of acute hemolysis risk (Figure S11). Additionally, results showed that after drug administration in mice, the level of ALT, AST, BUN, Cr in serum inhibited no significant difference compared to the saline group, indicating good biocompatibility of the tested drugs (Figure S12).

In vivo antitumor effects

To further confirm whether DCI possessed in vivo excellent antitumor effects and safety, we divided the 4T1 tumor-bearing mice into nine groups: Group (a) PBS, (b) DHN, (c) ICG, (d) DHN/ICG, (e) DCI, (f) DHN + US, (g) ICG + US, (h) DHN/ICG + US, (i) DCI + US. Next, mice were intravenously administered with different formulas, and the mice of groups (g), (h), and (i) were locally exposed to US radiation (1 MHz, 0.97 W/cm², 50% cycle, 8 min) after intravenous injection for 8 h. The tumor volume of mice receiving different treatments was measured every 2 days. Our results showed that, in the absence of US radiation, mice injected with DCI have a certain degree of tumor inhibition effect, while mice injected with other dosage forms have almost no tumor inhibition effect, which can be interpreted as the high-efficiency accumulation of DCI in tumors under the mediation of passive targeting, and high-efficiency uptake by tumor cells. The released Cu²⁺ can produce a large amount of highly toxic ·OH through a Fenton-like reaction to achieve tumor inhibition [42]. Under US irradiation, mice injected with other dosage forms still had almost no tumor inhibition effect. That may be because they were all small molecules with no obvious passive targeting effect, and they could not efficiently accumulate in the tumor lesion. However, the tumors of mice injected with DCI were almost ablated. To further verify the CDT and SDT effects of DCI in vivo, we also used a DCFH-DA probe to detect the ROS levels of DCI at tumor lesion sites. As shown in Fig. 7A, under US radiation, tumor tissues of nude mice injected with DCI showed the strongest DCFH fluorescence signal, while tumor tissues of nude mice with other dosage forms had a weaker DCFH signal (although ICG and DHN/ICG groups could produce a certain amount of ROS at the cellular level, they were all small-molecule dosage forms. There is no obvious passive targeting effect in vivo, so ROS cannot be efficiently enriched in tumor lesion sites, resulting in low ROS abundance.

Fig. 7.

(A) DCFH fluorescence distribution, (B) Bio-TEM, (C) TUNEL, Ki67, and H&E images of tumor tissue after treatment with different treatments (n = 5)

In addition, we further verified the in vivo programmed synergistic therapy effect of DCI under US radiation using pathological-level analyses, including Bio-TEM, TUNEL, Ki67, and H&E. Under US irradiation, tumor tissues of mice injected with DCI showed not only obvious cell shrinkage or chromatin concentration but also obvious loss of mitochondrial morphology compared with those injected with other dosage forms, which was similar to the results of cell Bio-TEM mentioned above. Therefore, we have reason to speculate that HPN converted by DHN through acoustic oxidation can specifically induce mitochondrial dysfunction, which provides a theoretical basis for further research on the anti-tumor mechanism of HPN in the future. At the same time, we also found that under US irradiation, the results of TUNEL, Ki67, and H&E (see Fig. 7C) in tumor tissues of mice injected with DCI showed that DCI could not only significantly inhibit the proliferation of 4T1 cells, but also induce their apoptosis. For the systemic immune response, we measured the inflammatory cytokine (IL-6, TNF-α, IL-1β) levels in serum following different treatments. As shown in Figure S13. Compared to the saline group, the DCI group showed significantly increased levels of IL-6, TNF-α, and IL-1β. After US treatment, these inflammatory factors were further elevated, indicating that DCI, particularly when assisted with US treatment, can significantly activate the anti-tumor immune response.

Conclusion

We have successfully developed an engineered dual-mode imaging copper-coordinated nanoreactor by coordination-driven co-assembly of Cu²⁺, ICG, and DHN. After acidic lysosome and sonodynamic radiation triggering, this nanoplatform rapidly disassembles and releases the drug. The released Cu²⁺ decomposed H₂O₂ into ·OH and O₂. The generated O₂ efficiency alleviated tumor hypoxia and enhanced SDT efficiency. A large amount of the ¹O₂ generated by SDT oxidized DHN into HPN with higher toxicity. Taken together, this study provides a new idea for tumor diagnosis and treatment.

Author contributions

Zhongxiong Fan: Conceptualization, Writing-original draft, Supervision, Funding acquisition. Lijun Ding: Writing-original draft, Data curation. Fukai Zhu: Data curation. Nan Yang: Data curation. Yanrong Shi: Data curation. Qian Sun: Data curation. Jun Hai: Supervision and Funding acquisition. Jinyao Li: Supervision and Funding acquisition. Guanghao Su: Supervision and Funding acquisition. Zhenqing Hou: Supervision.

Funding

This work was supported by grants from the National Natural Science Foundation of China (82560364), Key research and development program in Xinjiang Uygur Autonomous Region (2023B02030-1), Project of Jiangsu Province Key Research and Development Plan (Social Development) (BE2022732), and Key Laboratory in Science and Technology Development Project of Suzhou (CN) (SZS2024012).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

All animal experiments were conducted in accordance with the guidelines of the Ethics Committee of Xiamen University, and the experiments were approved by the Ethics Committee of Xiamen University.

Consent for publication

All authors agree for publication.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.100x/s1xxxx0xx-xxxx-y.