Carrier-free GSH-responsive in situ nanoreactor for copper-overload augment cuproptosis/chemodynamic therapy

Abstract

Recently, copper (Cu)-based cancer therapy, an intracellular Cu-dependent process, has received tremendous attention because of the emergence of cuproptosis. However, the inherent regulatory mechanism limits intracellular Cu accumulation. Herein, a carrier-free GSH-responsive in situ nanoreactor (TECJ) is fabricated, of which Cu ionophore elesclomol (ES), GSH-responsive nitric oxide (NO) donor O₂-(2,4-dinitrophenyl) 1-[(4-ethoxycarbonyl) piperazin-1-yl] diazen-1-ium-1,2-diolate (JSK) and Cu²⁺ are integrated via self-assembly and coated with D-α-tocopherol polyethylene glycol 1000 succinate (TPGS). TECJ enhances the systemic stability and tumor accumulation of ES and JSK through both the coating function and the additional inner force provided by TPGS. After internalization, TECJ specifically releases the cargoes via GSH-responsive dissociation. Then, ES efficiently transports extracellular Cu²⁺ into the cytoplasm for Cu influx, while NO released from JSK aggravates mitochondrial dysfunction and blocks adenosine triphosphate supply to inhibit ATP7A expression and reduce Cu efflux, therefore dually resulting in Cu-overload and amplifying cuproptosis. Furthermore, Cu-induced Fenton-like reaction together with TPGS-mediated ROS generation triggers chemodynamic therapy (CDT), and NO transfers ROS into more toxic ONOO⁻ for accelerating tumor death while reducing tumor self-alleviation by inhibiting DNA repair. Afterwards, the process above synergistically activates immunogenic cell death and cascades immunotherapy. Finally, TECJ successfully suppresses tumor growth and prevents tumor metastasis.

Graphical abstract

1. Introduction

Copper (Cu) is an extraordinarily crucial trace element. On the one hand, it acts as a cofactor of enzymes and plays a vital role in keeping mitochondrial respiration and biosynthesis and preserving the oxidation/reduction balance [[1], [2], [3]]. On the other hand, Cu dysregulation may cause toxicity and has been fully discussed against tumors through two major pathways, cuproptosis and chemodynamic therapy (CDT) [4]. Cuproptosis is an emerging programmed cell death process [5], which is distinct from other cell death pathways such as apoptosis, pyroptosis or necroptosis. Once accumulated into the cytoplasm, Cu²⁺ binds to Ferredoxin 1 (FDX1), resulting in lipoylated protein aggregation and the disappearance of Fe-S cluster proteins, thereby leading to proteotoxic stress for cell death [6,7]. Besides cuproptosis, excess intracellular Cu²⁺ can also be oxidized to Cu⁺ and subsequently react with the high level of H₂O₂ via a Fenton-like reaction to generate a large amount of ·OH, which is considered a kind of CDT [8]. Afterwards, tumor cells undergoing both cuproptosis and CDT trigger immunogenic cell death (ICD) [9] and activate damage-associated molecular patterns (DAMPs) as “eat me” signals, including calreticulin (CRT) exposure and the release of adenosine triphosphate (ATP) and high mobility group box 1 (HMGB1), etc., ultimately promoting the recruitment and maturation of dendritic cells (DCs) and intertumoral infiltration of cytotoxicity T lymphocytes for cancer immunotherapy [10,11].

Although the antitumor activity of Cu is recognized, the inherent intracellular Cu content is far from the baseline of tumor inactivation. For this contradiction, Cu ionophores are applied to transport Cu [12]. Up to now, a series of Cu ionophores have been developed, such as elesclomol (ES) [13], disulfiram [14] and Cu-based nanomaterials [15], etc. However, as the application goes on, intracellular Cu improvement is always less than expected and is even kept within a safe threshold [16]. It is ascribed to the homeostasis regulation mechanisms of Cu, such as the ATP7A-mediated Cu efflux pathway [17]. Once intracellular Cu overload occurs, the efflux mechanism will rapidly respond for self-protection and reduce the antitumor efficacy. It is worth mentioning that this is also considered one of the main reasons for the contradiction between in vitro and in vivo outcomes of the Cu-based treatments. Compared with the excessive given Cu dosage during the in vitro characterization, the tumor accumulation of Cu in vivo is limited, making the negative effects of efflux more authentic and sensitive during the in vivo characterization than the in vitro one. To this end, the binary regulation of both the influx and efflux of Cu may be an attractive strategy to enhance intracellular Cu content and Cu-based antitumor efficacy.

In this study, ES was selected to transport Cu because of its extremely high transfer rate, mitochondria-selective delivery capacity [18,19] and peerless biosafety. To compensate for the inadequate Cu tumor perfusion, Cu was added to form the ES-Cu (EC) chelate as the initial Cu loading. As for reducing Cu efflux, blocking ATP supply may be a promising strategy, because ATP7A is an ATP-dependent transmembrane efflux pump [20,21]. In the past few decades, gas-based therapy has exhibited great advantages against cancer with superior safety [[22], [23], [24]]. Among them, nitric oxide (NO) as an endogenous gasotransmitter is seriously recommended. It can disturb mitochondrial respiration homeostasis and decrease mitochondrial membrane potential to inhibit ATP generation [25]. Furthermore, NO has also been reported to convert intra-tumoral reactive oxygen species (ROS) into the highly toxic peroxynitrite (ONOO⁻), which serves as a poly ADP-ribose polymerase (PARP) inhibitor to impede DNA repair and accelerate tumor death [26]. However, the short half-life and non-specific release reduce its feasibility [27]. To this end, O₂-(2,4-dinitrophenyl) 1-[(4-ethoxycarbonyl) piperazin-1-yl] diazen-1-ium-1,2-diolate (JSK), a NO donor, was applied, which can remain stable in the systemic circulation and specifically release NO in tumor cells via a GSH-responsive manner, the same manner as the EC release [[28], [29], [30]]. Therefore, the combination of ES and JSK is an attractive strategy for Cu-overload and Cu-based tumor therapy, while the low solubility of ES and JSK limits their application.

Herein, an in situ nanoreactor was fabricated. As shown in Scheme 1, ES, JSK and Cu²⁺ were self-assembled into ECJ via π-π stacking and coordination force. Then, D-α-tocopherol polyethylene glycol 1000 succinate (TPGS), as an amphiphilic FDA-approved biomaterial with ROS generation capacity [31], was applied to encapsulate the ECJ, forming TECJ. During intravenous administration, the coating function and additional inner force provided by TPGS enhance the systemic stability and tumor accumulation of ES and JSK. After internalization, all the cargoes were released via the GSH-responsive manner. Then, ES transports extracellular Cu²⁺ into the cytoplasm, while NO aggravates mitochondrial dysfunction and the cascaded reduction of ATP7A expression and Cu efflux, therefore resulting in two-pronged Cu-overload and amplifying cuproptosis. Meanwhile, Cu-induced Fenton-like reaction together with TPGS-mediated mitochondrial disruption accelerates ROS generation and triggers CDT. At the same time, NO reacts with ROS into more toxic ONOO⁻ and inhibits DNA repair to accelerate the tumor death and reduce tumor self-alleviation. Finally, the process above synergistically activates ICD and cascades immunotherapy against tumor growth and tumor metastasis. In summary, TECJ is a promising Cu-based antitumor option.

Scheme 1.

(A) Schematic illustration of the preparation of TECJ. (B) Potential mechanism of TECJ for cuproptosis/CDT and immunotherapy.

2. Results and discussions

2.1. Characterization of the nanoparticles

TECJ was prepared as shown in Fig. 1A. ECJ was formed by self-assembly of ES, Cu²⁺ and JSK via coordination force (ES ∼ Cu²⁺) and π-π stacking (ES ∼ JSK), and TPGS was decorated on the surface of ECJ via hydrophobic interaction, forming TECJ. EC was fabricated without JSK. The particle size and ζ-potential were both measured. As shown in Fig. 1B and C, the nanoparticles displayed comparable particle sizes (∼130 nm) and ζ-potentials (∼20 mV). The uniform particle size distribution was displayed and verified the applicability of the preparation method (Fig. 1D–F). Then, the morphology was observed by TEM. As shown in Fig. 1G–I, all the nanoparticles formed a spheroid-like structure. An obvious hydrophilic shell could be observed in TECJ, indicating that TPGS was successfully decorated. The stability of nanoparticles is a vital parameter closely related to the entire process of the application. Herein, the in vitro stability investigation was performed with PBS, DMEM and 10% FBS. As shown in Fig. S1, TECJ displayed superior storage stability with the decoration of TPGS, while EC and ECJ gradually aggregated during the incubation, especially in the medium with 10% FBS. This was because TPGS could enhance the stability of TECJ via both the coating function and additional inner force.

Fig. 1.

Characterization of the nanoparticles. (A) Self-assembly process of TECJ. (B-C) Particle size and ζ-potential of the nanoparticles. (D-F) Particle size distribution of EC, ECJ and TECJ, respectively. (G-I) Morphology of EC, ECJ and TECJ observed by TEM. Scale bar: 200 nm. Data are presented as mean ± SD, n = 3.

As the formation of a multi-component self-assembly process is often controversial, the potential self-assembly mechanism was investigated by a molecular dynamic simulation assay. As shown in Fig. S2A, the root mean square deviation (RMSD) curve was nearly flat after 20 ns, indicating that the ECJ system reached a stable state. Then, the self-assembly process was visualized (Fig. S2B). Most of the monomers completed the polymerization process and generated cluster-like structures within 10 ns. However, the aggregation between cluster-like structures was delayed, continuously transforming between micelle-cluster-micelle structures. Subsequently, the molecular interaction was investigated and analyzed (40 ns). As shown in Fig. S2C, JSK provided π-π stacking and supplemented by hydrogen bond interaction owing to its unsaturated aromatic ring and a large number of oxygen atoms, which were considered the main driving force during the self-assembly process. The density variation of the system was also investigated by the Radius of gyration (Rg). As shown in Fig. S2D, it was consistent with the self-assembly process. The degree of asphericity (A_d) was detected to evaluate the morphology. During the process, the A_d of the ECJ system remained below 0.06 (Fig. S2E), indicating a spherical-like structure. Considering the molecular simulation outcomes of EC [32], the addition of JSK slightly improved the self-assembly of the nanoparticles.

Then, the self-assembly process of TECJ was further investigated. As shown in Fig. 2A, TECJ was also stable after 20 ns. Compared with ECJ, TECJ monomers were more likely to aggregate with each other to form uniform nanoparticles (Fig. 2B), benefiting from the additional inner force induced by the addition of TPGS. Furthermore, as the results of spatial conformational distributions analysis (Fig. 2C), the long-tail structure of TPGS tended to connect the components within the system and wrap around ES, JSK and Cu²⁺, significantly reducing the unconstrained alteration of the polymer clusters and narrowing the distance between the clusters, thereby enhancing the aggregation of TECJ. Then, Rg was measured. As shown in Fig. 2D, TECJ gained a higher Rg value and lower curve fluctuation (∼0.3 nm), compared with the ECJ group (curve fluctuation: ∼0.7 nm). It was suggested that the addition of TPGS constructed both the loose coating and the tight core. As predicted, TECJ reduced A_d value, confirming that TECJ had a higher sphericity degree (Fig. 2E). All the results above proved the optimized self-assembly and storage stability (Fig. S1).

Fig. 2.

Molecular dynamics simulation assay and drug release assay. (A) Dynamics simulation analysis of RMSD. (B) Visualization of the self-assembly process of TECJ during 50 ns. (C) Interaction force analysis of TECJ during the self-assembly process at 20 ns. (D-E) Rg and A_d analysis of TECJ. (F) GSH-responsive drug behavior of TECJ. (G) Morphology variation of TECJ under GSH conditions. Scale bar: 200 nm. (H) Mechanism of GSH-responsive NO release by JSK. (I) NO release of ECJ and TECJ. (J) Accumulative ES release of the nanoparticles. Data are presented as mean ± SD, n = 3.

The release rate of drugs also affects the therapeutic effect of nanoparticles. TECJ was considered to dissociate via the GSH-responsive manner (Fig. 2F). The alteration of the morphology was visualized by TEM. As shown in Fig. 2G, after incubation with GSH for 2 h, TECJ aggregated with each other, and the particle size significantly increased. At the 12 h time point, intact TECJ was rarely observable and replaced by a large number of black pixel dots. The results above confirmed the GSH-responsive release manner of TECJ, which was due to both the high GSH affinity to Cu²⁺ and the two-step oxidation-reduction reaction of JSK (Fig. 2H). Then, the release of NO and ES was detected. As shown in Fig. 2I, compared with ECJ, TECJ reduced the leakage of NO in the absence of GSH, benefiting from both the coating function and interaction force of TPGS. As the addition of GSH, ECJ and TECJ rapidly released NO. Although the release process of TECJ was delayed, there was no significant difference between the two groups at the 12 h time point. As for ES, except for the TPGS, JSK could also slightly reduce the leakage of ES by providing extra inner force (Fig. 2J). Theoretically, as the release of NO, the proportion of hydrophilic components in the JSK-ES system decreased, which further improved the interaction between JSK residues and ES to keep the nanoparticles stable. Then, the release of ES in the GSH condition was discussed. It was reduced in ECJ and TECJ groups, while there was no significant difference compared with that of EC. To this end, TECJ, as a carrier-free nanoreactor, is considered to enhance the circulation stability of ES and JSK and maintain sufficient tumor in situ release capacity, potentially improving the availability of ES and JSK.

2.2. In vitro antitumor efficacy

The in vitro antitumor efficacy was investigated. Mediums containing gradient concentrations of CuCl₂ were applied to evaluate the antitumor efficacy of the nanoparticles, and the IC₅₀ value was calculated. As shown in Fig. 3A–D, all the nanoparticles performed a Cu²⁺ concentration-dependent cytotoxicity, proving that ES effectively transported Cu into the cytoplasm and completed the first phase of copper overload. Compared with the ES group, EC displayed improved antitumor activity with the initial Cu loading, and ECJ was more effective because of the addition of the NO producer JSK. As expected, TECJ gained the superior performance among the groups, benefiting from the TPGS decoration, which synergistically generated ROS.

Fig. 3.

In vitro cell cytotoxicity and the potential mechanism. (A-C) Cell cytotoxicity of the nanoparticles with gradient CuCl₂ concentrations against the 4T1 cell line, n = 6. (D) The IC₅₀ value of the nanoparticles. (E-F) The cellular uptake efficiency of the nanoparticles measured by FCM. (G-I) Intracellular ROS, NO and ONOO⁻ generation visualized by DCFH-DA, DAF-FM DA and DAX-J2 PON, respectively. Scale bar: 20 μm. Data are presented as mean ± SD.

Then, the cellular uptake efficiency of the nanoparticles was detected by confocal laser scanning microscope (CLSM) and flow cytometry (FCM), because the concentration of intercellular toxic components commonly directs the cytotoxicity. Coumarin 6 (C6) was selected as the fluorescence probe. As shown in Fig. 3E and F and S3, all the nanoparticles showed a favorable tumor internalization performance compared with the control group. As the outcomes were comparable, the differences in cytotoxicity among the nanoparticles were determined by the unit antitumor activity, which further confirmed the superiority of the addition of JSK and TPGS. To clarify the functions of TECJ, factors, processes and cell death pathways were further investigated.

2.3. In vitro ROS, NO, ONOO⁻ generation and mitochondrial dysfunction

The generation of ROS, NO and ONOO⁻ was visualized by CLSM with 2,7-dichlorofluorescein diacetate (DCFH-DA), 3-amino-4-aminomethyl-2′,7′-difluorescein diacetate (DAF-FM DA) and DAX-J2 PON, respectively. As shown in Fig. 3G, the EC group displayed slight green fluorescence, indicating limited ROS generation due to the ES-induced Cu influx, while bright green fluorescence was observed after incubation with ECJ, because NO could aggravate mitochondrial dysfunction and enhance the production of ROS. Among the groups, TECJ gained the highest fluorescence intensity with the decoration of TPGS, which could promote ROS generation via interacting with mitochondrial respiratory complex Ⅱ [33]. Subsequently, NO and ONOO⁻ levels were analyzed. As shown in Fig. 3H and I, ECJ and TECJ effectively produced a large amount of NO, which further reacted with ROS to generate the highly toxic ONOO⁻. It is noted that despite no additional NO generation for TECJ, it increased the generation of ONOO⁻, because the extra ROS accelerated the ROS/RNS reaction. All these results confirmed the factor production capacity of TECJ.

The activity of mitochondria is crucial for cells, and all of the ROS, NO and ONOO⁻ are mitochondrial inactivators. Herein, the mitochondrial membrane potential and mitochondrial activity were detected to reflect mitochondrial homeostasis. Mitochondrial membrane potential was measured by a JC-1 fluorescence probe. As shown in Fig. 4A, compared with EC and ECJ, both the down-regulation of the red fluorescence (J-aggregate) and the up-regulation of the green one (J-monomer) were amplified after being treated with ECJ and TECJ, implying that the mitochondrial membrane potential was significantly decreased due to the high levels of NO and ONOO⁻ produced along with JSK. Then, Rh123 was applied to visualize the mitochondrial activity. As shown in Fig. 4B, TECJ displayed negligible green fluorescence compared with the other groups, indicating the highest mitochondrial toxicity with the addition of the ROS generator TPGS. The results confirmed that TECJ could sufficiently disrupt the mitochondrial function. Then, intracellular ATP content was also measured. As shown in Fig. 4C, a significant decrease in ATP production was found in the ECJ and TECJ groups, which provided additional evidence of mitochondrial dysfunction.

Fig. 4.

In vitro mitochondrial dysfunction and DNA impairment. (A) Mitochondrial membrane potential detected by CLSM with the JC-1 fluorescence probe. (B) Mitochondrial activity measured by CLSM with the Rh123 fluorescence probe. (C) Intracellular ATP level after incubation with the nanoparticles. (D) ATP7A expression observed by CLSM. Scale bar: 20 μm. (E) Intracellular Cu uptake with a gradient Cu feeding. The significant analysis is compared with the EC group. (F) In vitro γ-H2AX expression after incubation with the nanoparticles. (G-I) Nitro-tyrosine and PARP expression measured by Western blot. Data are presented as mean ± SD. “*”, “**” and “***” represented P < 0.05, P < 0.01 and P < 0.001, respectively.

2.4. In vitro ATP7A expression and Cu overload

The outcomes of Cu-based therapeutics were correlated positively with the intracellular Cu content. However, once Cu overload occurs, the ATP-dependent Cu transporter ATP7A will pump out the excess Cu to maintain homeostasis. Herein, the expression of ATP7A was investigated. As shown in Fig. 4D, compared with other groups, TECJ displayed the lowest ATP7A expression due to the reduction of ATP supply. The expression of ATP7A was also confirmed by WB assay (Fig. S4). Subsequently, intracellular Cu content was measured with a gradient concentration of Cu feeding. As shown in Fig. 4E, all the nanoparticles performed a Cu-dependent intracellular Cu transport manner, while TECJ gained the highest Cu content mainly due to the downregulation of ATP7A expression. This satisfied Cu overload was considered the binary regulation of both the influx and efflux. The effective Cu influx by the transport of ES, together with inhibition of Cu efflux by inhibiting ATP7A expression significantly increased intracellular Cu content, which was decisive for the Cu-based therapy.

In addition to Cu overload, TECJ could also amplify the antitumor efficacy by inhibiting tumor self-repair. In general, ROS-mediated cell death is based on DNA double-strand breaks, which can be reflected by γ-H2AX [34]. As shown in Fig. 4F, compared with other groups, TECJ showed up-regulated fluorescence intensity, indicating that TECJ effectively enhanced DNA damage. It is well acknowledged that DNA damage can be quickly repaired with the help of PARP, while ONOO⁻ may serve as a potential PARP inhibitor [35]. As shown in Fig. 4G–I, both Nitro-Tyrosine (ONOO⁻ marker) increase and PARP decrease were significantly observed in the TECJ group, confirming that TECJ alleviated the PARP-associated DNA-repair process by the generation of ONOO⁻ and facilitated tumor cell death.

2.5. In vitro cell death pathway

Then, the cell death pathway was characterized. In vitro cell apoptosis assay was investigated by Annexin V-FITC/PI detection kit. As shown in Fig. 5A and B, compared with the Control and EC groups, ECJ gained a higher apoptosis ratio with the addition of JSK, while TECJ performed best among the groups due to the collaborative efforts of all the components.

Fig. 5.

In vitro cell death pathway. (A-B) In vitro cell apoptosis assay of the nanoparticles against the 4T1 cell line. (C) In vitro FDX1 expression after incubation with the nanoparticles measured by CLSM. Scale bar: 50 μm. (D) In vitro DLAT oligomerization observation by CLSM. Scale bar: 20 μm. (E) HMGB1 release of 4T1 tumor cells after treatment. (F) Matured DCs in the 4T1/BMDCs co-culture model measured by FCM. (G) The potential mechanism of TECJ-induced tumor death. Data are presented as mean ± SD. “*”, “**” and “***” represented P < 0.05, P < 0.01 and P < 0.001, respectively.

With a significant improvement in intracellular Cu content, the nanoparticles were endowed with the potential of cuproptosis. To confirm the cuproptosis process, FDX1 and DLAT, as hallmarks, were both measured. As shown in Fig. 5C and S5, bright red fluorescence was observed in the control group, indicating a high level of FDX1 expression, while the cells treated with EC, ECJ and TECJ displayed gradually decaying fluorescence intensity, demonstrating the amplified cuproptosis cell death pathway. Then, DLAT oligomerization was evaluated. As shown in Fig. 5D, there was a uniform green fluorescence distribution in the Control group, proving that oligomerization rarely occurred in the cytoplasm. However, a large number of highlighted green pixel dots were found in ECJ and TECJ-treated cells, which was the representation of DLAT aggregation. Then, the WB assay was carried out. As shown in Fig. S6, the cuproptosis was indeed enlarged as the addition of JSK and TPGS. Herein, TECJ was pronounced to promote cuproptosis.

Both cuproptosis and CDT have been reported to trigger ICD. Tumor cells undergoing ICD can increase the expression of CRT and the release of HMGB1, which are known as DAMPs. CRT exposure activates DCs to engulf dying debris, and HMGB1 release promotes the stable association between the dying cells and DCs [36,37]. Herein, CRT exposure and HMGB1 release were detected. As shown in Fig. 5E and S7, cells treated with TECJ exhibited the highest CRT exposure and HMGB1 release among the groups. Then, the pre-treated tumor cells were co-incubated with bone marrow-derived dendritic cells (BMDCs) to verify the immune activation potential. As shown in Fig. 5F and Fig. S8, the maturation percentage of DCs was significantly increased in the TECJ groups, indicating that TECJ could effectively increase the immunogenicity of tumor cells and induce a cascaded immune response.

Therefore, TECJ was speculated to be functional by the following process (Fig. 5G). After internalization, all the cargoes were released via the GSH-responsive manner. Then, ES transports extracellular Cu²⁺ into the cytoplasm. Cu, TPGS and NO generated a large amount of ROS and RNS and resulted in mitochondrial dysfunction. Afterwards, the ATP production was reduced, and the expression of ATP7A was down-regulated, thereby inhibiting Cu efflux for Cu-overload and amplifying CDT and cuproptosis. Meanwhile, ONOO⁻ inhibited DNA repair to reduce tumor self-alleviation and accelerate the tumor death. Finally, the inactivated cells triggered ICD for tumor immunotherapy.

2.6. In vivo antitumor efficacy

Then, the antitumor efficacy of the nanoparticles was evaluated in vivo. As both the application of ES and JSK were criticized due to the unsatisfactory system circulation stability and low tumor accumulation, the in vivo biodistribution assay was carried out against the 4T1 tumor-bearing mice with DiR-labeled nanoparticles. At 12 h, the mice were sacrificed, and the major organs were obtained for ex vivo imaging. As shown in Fig. 6A–C, all the nanoparticles displayed a gradually attenuated biodistribution, and the fluorescence was mainly located in the tumor and liver at 12 h. The fluorescence intensity in tumors of TECJ was significantly higher than that of EC and ECJ, indicating that TPGS decoration effectively prolonged system circulation and enhanced tumor accumulation. Furthermore, DiO-labeled nanoparticles were used to visualize the intertumoral biodistribution. As shown in Fig. 6D, TECJ displayed the highest green fluorescence compared with other reference groups, which was consistent with the in vivo biodistribution assay.

Fig. 6.

In vivo biodistribution and antitumor efficacy. (A) In vivo biodistribution of the nanoparticles against the 4T1 tumor-bearing mice. (B-C) Ex vivo biodistribution analysis. (D) Intertumoral biodistribution of the nanoparticles in the 4T1 tumor-bearing mice. (E) Individual tumor volume alteration. (F) The average tumor volume of the 4T1 tumor-bearing mice after treatment. (G) Tumor weight after treatment. (H) The alteration of body weight during the treatment. Data are presented as mean ± SD. “*”, “**” and “***” represented P < 0.05, P < 0.01 and P < 0.001, respectively.

To validate the antitumor efficacy in vivo, a 4T1 tumor-bearing mouse model was established, and the mice were randomized and treated with PBS, ES, EC, ECJ and TECJ, respectively. Individual tumor volume of the mice was recorded and analyzed (Fig. 6E–G). Compared with the ES group, EC displayed higher antitumor efficacy with the initial Cu loading, which partly compensated for the intracellular Cu and amplified the Cu-based cuproptosis/CDT process. Then, JSK was introduced into the following formulation. ECJ gained enhanced performance in the presence of JSK, which produced NO via a GSH-responsive manner in the tumor tissues. As expected, TECJ achieved the most pronounced antitumor activity because of the synergies of ES, JSK and TPGS, which not only enhanced the intracellular Cu accumulation by the binary regulation for cuproptosis but also achieved ROS/RNS-induced CDT with the collaborative efforts of Cu, NO and TPGS. The body weight was also recorded. As shown in Fig. 6H, there was no significant alteration in body weight, indicating favorable biosafety. The biosafety was evaluated with healthy Balb/c mice, including body weight, blood biochemistry assay and H&E staining of major organs. All the results confirmed that TECJ displayed acceptable biosafety (Figs. S9–11).

Subsequently, the histological assay was applied. As shown in Fig. 7A and B, H&E and TUNEL staining showed that tumor tissues treated with TECJ displayed extensive nuclear shrinkage, nucleic loss and fragmentation with apparent apoptosis. FDX1 immunofluorescence staining was also applied in vivo, the results showing that TECJ did induce cuproptosis (Fig. 7C). Then, the cuproptosis behavior was also confirmed by in vivo WB assay (Fig. S12).

Fig. 7.

Histological analysis. (A-D) H&E staining, TUNEL staining, FDX1 immunofluorescence and CRT immunofluorescence of tumor tissues after treatment. Scale bar: 100 μm. (E) Matured DC in TDLNs after treatment. (F-G) CD8⁺ T lymphocyte infiltration and its IFN-γ secretion in tumor tissues. (H-J) TNF-α, IFN-γ and IL-10 levels after treatment. Data are presented as mean ± SD. “*”, “**” and “***” represented P < 0.05, P < 0.01 and P < 0.001, respectively.

2.7. In vivo immune response analysis

To further demonstrate the impact of TECJ on immunotherapy, CRT immunofluorescence was investigated. As shown in Fig. 7D, compared with other groups, TECJ significantly induced CRT exposure, which served as an “eat me” signal to stimulate DC maturation and activate the antitumor immune response.

Then, the tumor tissues and tumor-draining lymph nodes (TDLNs) were collected and analyzed. The DC maturation assay was investigated. As shown in Fig. 7E and S13, TECJ effectively increased the subpopulation of matured DCs (gated as CD80⁺CD86⁺), which was 1.44- and 2.38-fold higher than that of ECJ and EC, respectively. Subsequently, matured DCs could present antigens to T lymphocytes and activate T cells in tumor sites to evoke the immune response. Herein, CD8⁺ T lymphocyte infiltration was evaluated. As shown in Fig. 7F and S14, the percentage of CD8⁺ T lymphocyte cells in tumor tissues of the TECJ group was 1.55- and 2.17-fold higher than that of ECJ and EC, proving effective CD8⁺ T lymphocyte infiltration. IFN-γ secreted by CD8⁺ T cells is an important weapon for tumor killing [38]. Mice treated with TECJ displayed sufficient IFN-γ secretion from CD8⁺ T lymphocytes compared with other groups (Fig. 7G and S15), implying the superior antitumor immune response. Immunotherapy heavily relies on systemic cytokine levels. As shown in Fig. 7H–J, after being treated with TECJ, pro-inflammatory cytokines (TNF-α, IFN-γ) were significantly elevated, while the anti-inflammatory one (IL-10) was reduced, confirming that TECJ induced a robust antitumor immune response.

2.8. In vivo anti-metastatic activity

Encouraged by the satisfactory in vivo results, we are also interested in investigating whether TECJ could inhibit tumor metastasis. As the lung is a high-risk site for metastasis of breast cancer, the pulmonary nodules and the histological alteration were both evaluated. As shown in Fig. 8A, compared with other reference groups, the fewest pulmonary nodules (red circles) were found after being treated with TECJ, implying superior anti-metastatic activity. It is reported that NO could reduce the exosmosis of platelets to restrain epithelial-mesenchymal transition (EMT) and inhibit tumor metastasis [39]. The hallmarks of EMT (N-cadherin and E-cadherin) were measured by Western blot. As shown in Fig. 8B–D, with the addition of JSK, ECJ and TECJ significantly up-regulated the expression of E-cadherin and down-regulated N-cadherin compared with other groups. Furthermore, TECJ performed better than ECJ by the decoration of TPGS, which enhanced the systemic circulation stability. Moreover, blocking ATP supply could inhibit the ATP-dependent tumor-derived microvesicles (TMVs), which were also regarded as a hallmark of tumor metastasis [40]. Therefore, TGM2 was also investigated by immunofluorescence. As shown in Fig. 8E, ECJ and TECJ reduce the expression of TGM2, indicating that ECJ and TECJ could inhibit ATP supply and TMV generation for anti-metastasis. Then, the in situ NO generation was evaluated. As NO could react with ROS to generate RNS, 3-nitrotyrosine (3-NT) generation was investigated via immunohistochemistry. As shown in Fig. 8F, TECJ could effectively generate NO (brown pixel dots) compared with other groups. Herein, it was concluded that ECJ and TECJ played a vital role in tumor metastasis inhibition with the addition of JSK, especially TECJ. The potential mechanism for anti-metastasis of TECJ is illustrated in Fig. 8G. On one hand, TECJ generated NO together with TPGS/Cu to induce mitochondrial dysfunction and inhibit the secretion of TMV. On the other hand, it produced ROS and NO, which further reacted to RNS to inhibit the EMT process.

Fig. 8.

Anti-metastasis assay. (A) H&E staining of lung tissues in tumor-bearing mice. Red circle: metastasis nodule. (B-D) E/N cadherin expression measured by Western blot. (E-F) TGM2 and 3-NT expression in tumor tissues after treatment. (G) The potential mechanism for anti-metastasis of TECJ. Data are presented as mean ± SD. “*”, “**” and “***” represented P < 0.05, P < 0.01 and P < 0.001, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3. Conclusions

In this study, a carrier-free in situ nanoreactor was fabricated for amplifying cuproptosis/CDT and immunotherapy against tumors. It showed that TECJ could effectively accumulate in tumor regions with favorable in vivo stability. After internalization, TECJ dissociated via the GSH-responsive manner and released all the cargoes, which rapidly generated ROS and RNS for mitochondrial dysfunction to inhibit the expression of ATP7A and induce Cu-overload for amplifying cuproptosis/CDT. Then, the cells underwent ICD promoted the immunotherapy. Overall, TECJ effectively inhibits tumor growth and metastasis, which provides a promising strategy for tumor treatment.

4. Methods and materials

4.1. Materials and reagents

ES, JSK, Hoechst 33242, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), C6, DAF-FM DA, DCFH-DA and TPGS were all purchased from Aladdin Biotechnology (Shanghai, China). Annexin V-FITC/PI kit, live/dead kit, mitochondrial membrane potential probe (JC-1), culture medium and fetal bovine serum were all purchased from Meilun Biotechnology Co., Ltd. (Dalian, China). DAX-J2 PON was purchased by AAT Bioquest (Sunnyvale, CA). All other chemical reagents were analytical grade.

4.2. Cells and animals

Murine 4T1 tumor cells were purchased from Meilun Biotech. Co., Ltd. (Dalian, China). Cells were cultured in DMEM culture medium with 1% penicillin/streptomycin and supplemented with 10% FBS (37 °C, 5% CO₂). BMDCs were cultured as reported [41].

Balb/c mice (18-22 g) were purchased from Qinglongshan Biotech Co., Ltd. (Nanjing, China). All animal experiments were performed in compliance with the Institutional Animal Care and Use Committee (IACUC) at China Pharmaceutical University (24-03-027).

4.3. Preparation and characterization of TECJ

TECJ was fabricated by a one-spot assembly method. Briefly, ES (5 mg) and JSK (5 mg) were co-dissolved in 1 mL DMSO, and copper chloride (CuCl₂) was added to the mixture, then TPGS (50 mg) was added. The mixture was stirred at room temperature and the unreacted CuCl₂ (10 mg) was removed via centrifugation. Then, the supernatant was dispersed in distilled water with stirring and dialyzed to remove the organic solvent. After concentration, TECJ was obtained. ECJ was prepared in the same way without TPGS, and EC was fabricated without both JSK and TPGS.

The particle size distribution and ζ-potential were measured by Malvern Zetasizer (Nano ZS, UK). The morphology was observed by transmission electrostatic microscopy (TEM, HT7700, Hitachi, Japan). To investigate the storage stability, the nanoparticles were placed into different culture mediums (including PBS, 10% FBS and DMEM, respectively) at 37 °C. At the predetermined time points, the particle size was measured. In vitro NO release behavior was investigated by the Griess reagent. Briefly, EC, ECJ and TECJ were added into PBS containing GSH with gradient concentrations at 37 °C. At the predetermined time points, NO concentration was detected by the Griess reagent. In vitro ES release behavior was investigated by the dialysis method. Briefly, the nanoparticles were transferred into a dialysis bag and placed into a flask containing 100 mL of the release medium (PBS with/without gradient-concentration GSH). Then, the flasks were placed in a shaking incubator (100 rpm, 37 °C). At the predetermined time points, 1 mL of the release medium was taken and an equal volume of fresh medium was supplied. The concentration of the drug was measured by HPLC.

4.4. Molecular simulation assay

To investigate the self-assembly process of TECJ, a molecular simulation assay was conducted. The structures of EC optimized complexes were obtained, and the 3D structure of JSK and TPGS were constructed by ChemDraw and Chem3D software and optimized by the ACPYPE online tool.

The self-assembly behavior of 100-molecule ES-Cu²⁺-JSK complex and 100-molecule ES-Cu²⁺-JSK with 50 molecules of TPGS after 50 ns polymerization was predicted by the GROMACS2021 software based on the AMBER-GAFF force field. The monomer components were randomly distributed in a square cubic periodic box with a side length of 20 nm, and the water molecule model was SPCE with a simulation step size of 2 fs. The cut-off radius for van der Waals interactions was set to 1.4 nm. Long-range electrostatic interactions were calculated as PME with a cut-off distance of 1.4 nm. The systems used were subjected to energy minimization of 50,000 steps. The NVT tethered zones were temperature controlled (25 °C) by the V-rescale thermostatic method, resulting in a pre-equilibrium of 0.1 ns. The NPT tether was controlled by the isotropic Parrinello-Rahman with a pressure of 1 bar and simulated at 25 °C for 500 ps. Periodic boundary conditions were used throughout the simulation, and the bond lengths and bond angles of all hydrogen atoms were constrained by the LINCE algorithm. 50 ns sampling was carried out with parameters analogous to those of NPT. At the end of the simulation, the trajectories were visualized and analyzed by VMD and PyMOL.

4.5. In vitro cell cytotoxicity assay

In vitro cell cytotoxicity was investigated by an MTT assay. Briefly, 4T1 tumor cells were seeded into a 96-well plate. After attachment, the culture medium was replaced by the nanoparticles with gradient concentrations. After 24 h incubation, MTT was added and incubated for another 4 h. Then, all the medium was removed, and DMSO was added to dissolve the purple crystal. The cell viability was measured by a microplate reader. The IC₅₀ value of the nanoparticles was calculated by GraphPad Prism 8.0 software.

4.6. In vitro cellular uptake assay

C6 was used to trace the cellular uptake behavior of the nanoparticles. Briefly, cells were seeded into a 6-well plate with a round coverslip. After attachment, the C6-loaded nanoparticles were added and incubated for 4 h. Then, the cells were washed with PBS, fixed with 4% paraformaldehyde (PFA) and stained with Hoechst 33242. The cellular uptake was observed by CLSM. For the FCM assay, cells were harvested after incubation with the nanoparticles. After being washed and collected, the fluorescence intensity was measured by FCM.

4.7. Intracellular NO, ROS, ONOO⁻, ATP, ATP7A and Cu detection

DAF-FM DA, DCFH-DA and DAX-J2 PON were used to detect intracellular NO, ROS and ONOO⁻, respectively. Briefly, cells were seeded into a 6-well plate with a round coverslip. The nanoparticles were added with additional CuCl₂. After being incubated for 4 h, the cells were washed with PBS and stained with the probes. The intracellular fluorescence signal was eventually observed by CLSM.

For ATP detection, the cells were pre-treated as described above, and the intracellular ATP content was measured by an ATP assay kit. Furthermore, ATP7A expression was observed by CLSM. Briefly, after incubation with the nanoparticles, the cells were fixed, permeabilized and blocked. Then, the cells were incubated with anti-ATP7A antibody overnight and stained. The ATP7A expression was observed by CLSM. Meanwhile, ATP7A was also measured by Western blot according to the standard operating procedure.

Intracellular Cu content was measured by a Cu colorimetric assay kit. Briefly, cells were seeded into a 24-well plate to allow attachment. Subsequently, the nanoparticles were added and incubated for 12 h. To investigate the Cu transport capability, a gradient of CuCl₂ was added to the culture medium before incubation. At the predetermined time point, the cells were washed with PBS, and the intracellular Cu content was measured.

4.8. In vitro mitochondrial evaluation

JC-1 and Rh123 were used to investigate the mitochondrial membrane potential and mitochondrial activity of 4T1 cells. Briefly, cells were incubated with the nanoparticles for 12 h and stained with a JC-1 or Rh123 probe. After being washed with PBS, the fluorescence was observed by CLSM.

4.9. In vitro DNA repair

γ-H2AX immunofluorescence was carried out to detect DNA damage. Briefly, cells were seeded into a 6-well plate with a round coverslip. After treatment, the cells were washed, fixed and stained with the antibody and fluorescence probe. The DNA repair was observed by CLSM. Furthermore, PARP and Nitro-Tyrosine were also detected by Western blot according to the standard operating procedure.

4.10. In vitro cell death pathway

4.10.1. In vitro apoptosis

Annexin V-FITC/PI apoptosis kit was utilized to investigate the apoptosis. Briefly, cells were seeded into a 6-well plate for attachment. Then, the nanoparticles with additional CuCl₂ were added and incubated for 24 h. The cells were washed and collected. Subsequently, Annexin V-FITC and PI were added according to the protocols of the kit and eventually measured by FCM.

4.10.2. In vitro cuproptosis

FDX1 and DLAT expression were detected by immunofluorescence and Western blot according to the standard operating procedure, which was mentioned above.

4.10.3. In vitro ICD

In vitro CRT expression was investigated by CLSM. Briefly, 4T1 cells were seeded into a cover glass in a 6-well plate to allow attachment. The nanoparticles with additional CuCl₂ were added and incubated for 24 h. The cells were washed with PBS and fixed with 4% PFA. Subsequently, the cells were stained with an anti-CRT primary antibody and Alexa488-conjugated secondary antibody, and the nuclei were stained with Hoechst 33258 and observed using CLSM.

HMGB1 and ATP release were measured by the corresponding kits. Briefly, the cells were seeded into a 6-well plate. After attachment, the cells were incubated with the nanoparticles with additional CuCl₂ for 24 h. The culture medium was collected and centrifuged. Then, the supernatant was collected, and the concentrations of HMGB1 and ATP were measured by the corresponding kits.

In vitro DC maturation was detected with the 4T1/DC co-culture model. Briefly, 4T1 tumor cells were seeded into a 6-well plate and incubated with the nanoparticles with additional CuCl₂ for 6 h. Then, the cells were harvested and co-incubated with BMDCs for 24 h. The co-culture cells were collected and washed with PBS. Matured DCs were stained with a cocktail of antibodies, including anti-CD11c-FITC, anti-CD80-PE and anti-CD86-APC, respectively. Matured DCs were gated as CD11c⁺CD80⁺CD86⁺.

4.11. In vivo biodistribution assay

DiR was used as a fluorescence probe to detect the biodistribution behavior of the nanoparticles in tumor-bearing mice. Briefly, the 4T1 tumor-bearing mice were randomly divided. 0.2 ml of the DiR-loaded nanoparticles (containing an equal amount of DiR) was intravenously administrated. At an interval time point, the mice were anesthetized and detected by the IVIS system. At the end of the study, the major organs were obtained, and the fluorescence signal was measured by the IVIS system.

4.12. In vivo antitumor activity of TECJ

The 4T1 tumor-bearing mice with a tumor volume was approximately 100 mm³ were randomly divided into several groups, including Control, ES, EC, ECJ and TEC (n = 6). The nanoparticles were intravenously administrated every other day, 5 times. The body weight and tumor volume were calculated every other day. On day 14, the mice were sacrificed and tumors were harvested, weighed and stored for histological assay.

Healthy Balb/c mice were selected for the biosafety assay. Briefly, mice were randomly divided into several groups, and the nanoparticles were intravenously administrated every other day, 5 times. Body weight was measured. At day 14, the mice were sacrificed, and the major organs were collected for H&E staining. Blood samples were also collected for biochemistry analysis.

4.13. In vivo histological evaluation

For tumor tissues, the samples were fixed and sliced for H&E staining, TUNEL staining, FDX1 immunofluorescence and CRT immunofluorescence assay according to the standard operating procedure, respectively. The expression of FDX1 and DLAT was also measured by WB according to the standard operating procedure.

4.14. Immune cells analysis

Tumor tissues were collected, cut and digested into a single-cell suspension. The cells were washed with PBS and resuspended in FCM staining buffer and stained. Cell suspension of TDLN was collected and stained with CD11c-FITC, CD80-PE and CD86-APC, respectively. Matured DCs were gated as CD11c⁺CD80⁺CD86⁺. Then, the cell suspension of tumor tissues was stained with CD3-PerCP-Cy5.5, CD4-FITC, CD8-PE and IFN-γ-APC, respectively. CD4⁺ T cells and CD8⁺ T cells were gated as CD3⁺CD4⁺CD8⁻ and CD3⁺CD4⁻CD8⁺, respectively. IFN-γ secretion by CD8⁺ T cells was gated as CD3⁺CD4⁻CD8⁺ IFN-γ⁺. Blood cytokines were measured by the corresponding ELISA kits.

4.15. Anti-metastasis study

Tumor-bearing mice were treated with the nanoparticles as previously described, and the mice were sacrificed at day 28th. The lung tissues were obtained, and the metastasis nodules were calculated according to H&E staining. Meanwhile, E/N-cadherin expression was detected by Western blot. TGM2 and 3-NT expression was detected by immunofluorescence and immunohistochemistry and observed by CLSM.

4.16. Statistical analysis

All data were presented as mean ± SD. The significant difference was analyzed by one-way ANOVA. P < 0.05 was considered significant. *P < 0.05, **P < 0.01, ***P < 0.001.

CRediT authorship contribution statement

Yuhan Hu: Investigation, Methodology, Writing – original draft. Qifan Zhou: Data curation, Funding acquisition, Investigation, Methodology, Supervision, Validation, Writing – original draft. Haitong Wen: Investigation, Methodology, Writing – original draft. Tianyi Huang: Investigation, Methodology. Yongqing Liu: Investigation, Methodology. Ye Liu: Investigation, Methodology. Xin Jin: Data curation, Funding acquisition, Supervision, Validation, Writing – review & editing. Xiaoyu Lu: Data curation, Funding acquisition, Supervision, Validation, Writing – review & editing. Hangyi Wu: Conceptualization, Data curation, Funding acquisition, Supervision, Validation, Writing – review & editing.

Declaration of competing interest

The authors have no conflicts of interest to declare.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.