A Domino‐Synthesized Dicoordinate Copper(I) Bis‐imidazopyridine Complex Triggering Cuproptosis/Ferroptosis for Enhanced Cancer Immunotherapy

ABSTRACT

Novel synthetic methods offer significant potential to accelerate drug development, yet there remains a largely unexplored area for efficiently synthesizing metal‐based anticancer agents. Herein, we report a novel solvothermal domino reaction of pyridine‐2‐methylamine (and its 4‐OCH₃‐substituted derivative), benzaldehyde, and CuCl₂∙2H₂O, which simultaneously achieves ligand synthesis and coordination assembly in one pot and facilely affords innovative dinuclear dicoordinate copper(I) complexes (Cu1 and Cu2) with the in situ‐formed bulky steric‐hindering tetraarylethane ligands featuring the bis‐imidazo[1,5‐a]pyridine scaffold. The unique geometry of Cu1 and Cu2 confers physiological stability and vacant coordination sites for efficiently catalyzing Fenton‐like reactions. Further studies reveal that Cu2 effectively elevates intracellular copper ion levels to induce cuproptosis, concurrently disrupting cellular redox homeostasis to trigger ferroptosis. The concurrent cuproptosis‐ferroptosis activation finally elicits significantly enhanced immunogenic cell death (ICD), which facilitates the antitumor activity of Cu2. Moreover, in combination with immune checkpoint inhibitor αPD‐1, Cu2 exhibits improved immunotherapy effects. This work introduces the first small‐molecule copper complex that achieves immunotherapy potentiation through cuproptosis‐ferroptosis‐ICD induction and provides a new pathway for accessing innovative metal‐based antitumor agents through such a rationally designed domino reaction.

A paradigm‐shifting domino synthe tic strategy is developed to concurrently achieve organic ligand formation and coordination assembly in one pot, efficiently transforming simple substrates into the dinuclear dicoordinate copper(I) complex Cu2. Cu2 functions as the first small‐molecule cooper complex for cuproptosis‐ferroptosis synergy and immunogenic cell death (ICD) activation, exhibiting concurrent primary and distant tumor inhibition.

1. Introduction

Innovations in synthetic chemistry have driven drug discovery and development over the past century, improving human health, particularly in anticancer therapeutics [1, 2, 3]. Among various potential anticancer agents, metal complexes demonstrate unique therapeutic potentials due to their structural programmability, subcellular targeting capability, and distinct physicochemical properties [4, 5]. While conventional strategies for synthesizing metal complexes typically involve inefficient stepwise procedures, which include multi‐step organic ligand synthesis followed by coordination with metal ions, and each step often requires tedious separation and purification. Therefore, high‐efficiency synthetic methods integrating organic ligand synthesis and coordination assembly are critically needed to accelerate the access of novel metal‐based antitumor agents.

Over the past two decades, we have established the solid‐state/solution structural information correlation principle, which leverages combined crystallography (precise solid‐sate structural information) and mass spectrometry (supplementary information in solution) to elucidate complicated coordination assembly processes. With this principle, we synthesized a series of 3d‐metal coordination clusters and revealed their assembly mechanisms [6, 7, 8, 9, 10, 11]. Subsequently, we observed that certain heteroaromatic alcohol ligands underwent in situ covalent bond reorganization (e.g., C‐C coupling, dehydration) during assembly [7, 9]. Further investigation revealed that heteroaromatic aldehydes and amines could participate in covalent domino reactions facilitated by 3d metal ions under solvothermal conditions, directly yielding metal complexes with novel in situ‐formed organic ligands [12, 13, 14, 15, 16, 17, 18, 19]. Unlike conventional solvothermal methods that rely on pre‐synthesized ligands, our approach enables the in situ generation of novel organic ligands from simple substrates via covalent domino reactions, followed immediately by their coordination assembly into metal complexes. Therefore, the solvothermal domino synthetic strategy provides an efficient pathway to innovative metal complexes.

Recently, increasing innovative metal complexes exhibit promising anticancer effects by activating non‐apoptotic pathways [20], which can overcome apoptosis resistance to directly kill tumor cells and further induce immunogenic cell death (ICD) for enhanced immunotherapy [21, 22, 23, 24, 25, 26, 27, 28, 29, 30]. Cuproptosis, a newly identified cell death pathway triggered by intracellular copper overload, has garnered significant research interest [31]. Meanwhile, copper ions (Cu⁺/Cu²⁺) also act as potent enhancers of Fenton‐like reactions via their excellent redox activity, catalyzing cytotoxic hydroxyl radical (∙OH) generation from endogenous H₂O₂, potentially leading to mitochondrial dysfunction and ferroptosis [28, 32]. Notably, the concurrent activation of cuproptosis and ferroptosis markedly amplifies the ICD effect by generating robust immunostimulatory signals, thereby further potentiating the efficacy of cancer immunotherapy [33, 34, 35, 36]. However, most reported copper complexes for cuproptosis induction adopt four‐ or five‐coordinate geometries (Table S1), and their crowded coordination spheres shield the copper centers, leaving limited sites for directly binding with lipoylated proteins in cuproptosis induction. We propose that low‐coordinate copper complexes may be suitable for effective cuproptosis activation. Meanwhile, such copper complexes with vacant coordination sites can also expose their copper centers to intracellular excess H₂O₂, thereby inducing ferroptosis through efficient Fenton‐like catalysis.

Herein, we developthe solvothermal domino reaction of pyridyl‐2‐methylamine (and its 4‐OCH₃‐substituted derivative), benzaldehyde (lacking effective coordination sites), and copper chloride, which in situ furnishes bis‐imidazopyridine‐substituted tetraarylethane ligands featuring two spatially isolated N‐coordination sites, subsequently coordinating with two Cu⁺ ions to yield dinuclear dicoordinate copper(I) complexes (Cu1 and Cu2). Further antitumor mechanistic studies reveal that Cu2 can trigger concurrent cuproptosis and ferroptosis, subsequently eliciting ICD via damage‐associated molecular patterns (DAMPs) release, driving dendritic cell (DC) maturation and remodeling of the tumor immune microenvironment. Finally, in combination with the immune checkpoint inhibitor αPD‐1, Cu2 effectively inhibits tumor growth through potentiated immune responses. For the first time (Tables S2 and S3), we report the small‐molecule dicoordinate copper(I) complex Cu2, synthesized via a novel one‐pot solvothermal domino reaction, that simultaneously induces cuproptosis, ferroptosis, and ICD to potentiate cancer immunotherapy.

2. Results and Discussion

2.1. Design, Synthesis, and Characterization of Cu1 and Cu2

Imidazo[1,5‐a]pyridine scaffolds, which are found in natural marine sponges, as well as their synthetic derivatives and metal complexes have demonstrated potential antitumor activities [37, 38]. Previously, we employed solvothermal domino reactions to synthesize two novel iron complexes (Fe1 and Fe‐NBM) bearing in situ‐formed ligands with imidazo[1,5‐a]pyridine scaffolds, which demonstrated potential for tumor photodynamic therapy [14] and ICD activation [18]. These findings motivated us to design low‐coordinate copper complex ligands featuring imidazole[1,5‐a]pyridine scaffolds by reducing coordination sites of corresponding substrates for effective cuproptosis and ferroptosis activation. Based on the synthetic procedures for Fe1 and Fe‐NBM (Figure S1), we judiciously selected benzaldehyde (lacking effective coordination sites) as the aldehyde component, reacting it with pyridin‐2‐ylmethanamine (and its 4‐OCH₃‐substituted derivative) and CuCl₂∙2H₂O in MeOH at 160°C for 48 h. This reaction efficiently afforded two novel dinuclear dicoordinate copper(I) complexes (Cu1 and Cu2) featuring in situ‐generated bis‐imidazo[1,5‐a]pyridine‐substituted tetraarylethane ligands with reduced N‐coordination sites (Figures 1A, S2, and S3). Notably, both copper complexes were isolated by simple filtration of the reaction mixture, and no additional purification was required.

FIGURE 1.

Synthesis and characterization of Cu1 and Cu2. (A) Synthetic route for the preparation of Cu1 and Cu2. (B) Crystal structures of Cu1 and Cu2. Color code: C atoms are light gray, N atoms are light blue, O atoms are red, Cl atoms are bright green and Cu atoms are turquoise. H atoms are deleted for clarity. (C) Proposed reaction mechanism and observed intermediates with ESI‐MS analysis for Cu2. (D) The pH‐responsive generation of •OH by Cu2 in the presence of TMB and H₂O₂. (E) Cu2 exhibits superior Fenton‐like catalytic activity to Cu1 in the presence of TMB and H₂O₂ at pH 5.5. (F) ESR detection of •OH generation promoted by Cu1 and Cu2 using DMPO as a spin trap at pH 6.5.

Cu1, Cu2, and their corresponding in situ‐formed organic ligands (Bim‐1 and Bim‐2) were fully characterized using ¹H NMR and ¹³C NMR spectroscopy (Figure S4 and S5). Single crystal X‐ray diffraction (SXRD) analysis (Figure 1B and Table S4) reveals that both Cu1 and Cu2 are dinuclear, containing a centrosymmetric tetraarylethane ligand (copper‐to‐ligand ratio is 2:1) [39]. Each methylene group of the ligand bears one phenyl and one imidazo[1,5‐a]pyridine unit, where the latter is further functionalized with an additional phenyl group and coordinates to a Cu⁺ ion via the imidazole ring nitrogen atom. The coordination sphere is completed by a terminal Cl⁻, resulting in a slightly distorted linear geometry at each copper center (Table S5). The N‐Cu‐Cl bond angles of Cu1 and Cu2 are 172.51°and 175.27°, respectively (Table S6). Complementary to the SXRD results, X‐ray photo electron spectroscopy analysis verified the presence of Cu⁺ ions in the obtained two copper complexes (Figure S8) [40]. Electrospray ionization mass spectrometry (ESI‐MS) analysis (Figures S9 and S10, Tables S7 and S8) of Cu1 and Cu2 revealed that the dominant fragment ions could be assigned to the partial dissociation of coordination bonds (e.g., Cu‐N or Cu‐Cl) such as [(Cu1) − Cu^ICl + H]⁺ (m/z = 665.1427), [(Cu1) − Cu^ICl₂]⁺ (m/z = 629.1781), and [(Cu2)− Cu^ICl₂]⁺ (m/z = 689.1949), or to the complete dissociation of coordination bonds, such as [(Cu1)−Cu^I ₂Cl₂ + CH₃]⁺ (m/z = 581.2721), [(Cu2) − Cu^I ₂Cl₂ + H/Na/CH₃]⁺ (m/z = 627.2735, 649.2553 or 641.2891). Besides, thermogravimetric analysis of Cu1 and Cu2 (Figure S11) showed no significant mass loss below 200°C, indicating the absence of solvent molecules (e.g., MeOH or H₂O) in their crystal structures, which was consistent with the SXRD data. Upon continuous heating, Cu1 and Cu2 could be stable up to 200°C, followed by gradual decomposition to release volatile organic fragments from ligands and ultimately form a CuCl/amorphous carbon composite. The unique multi‑phenyl substituted tetraarylethane ligands may endow Cu1 and Cu2 with a certain degree of structural stability in solution (Figure S12).

Next, we performed time‐dependent ESI‐MS analysis (Figures S13–S15; Table S9) to track the domino reaction process, which revealed a 9‐step reaction sequence (Figure 1C, take Cu2 as an example). Key steps in the proposed mechanism involve aldimine condensation of (4‐methoxypyridin‐2‐yl)methanamine (L1) and benzaldehyde (L2) to generate the imine intermediate L3 (m/z = 227.1181), which coordinated with Int‐A ([Cu^II(L1)Cl₂], m/z = 235.9778, confirmed by SXRD) to generate Int‐B (m/z = 462.0893). Intramolecular nucleophilic addition of L3 was followed by deprotonation and β‐H elimination to yield the imidazo[1,5‐a]pyridine unit L4 (m/z = 225.1029), and the [Cu^II]‐H specie Int‐E (m/z = 202.0167) that subsequently reacted with Cu²⁺ to give Cu⁺. Next, the intermolecular nucleophilic addition of L4 and L2 formed diarylmethanol intermediate Int‐G (m/z = 331.1440), which underwent homolytic C‐O bond cleavage and C‐C bond coupling, and finally afforded the product Cu2 (m/z = 627.2769). Additionally, when the control experiment used 2,2,6,6‐tetramethyl‐1‐piperidinyloxy (TEMPO) as the radical scavenger, no Cu2‐related peaks were detected (Figure S15), indicating the involvement of a radical process in the domino reaction.

The solubilities of Cu1 and Cu2 in dimethyl sulfoxide (DMSO) were determined to be 2.02 and 1.96 mg/mL by UV‐vis spectroscopy (Figure S16). Time‐dependent ¹H NMR validated the structural stability of Cu1 and Cu2 over 48 h, both in DMSO‑d₆ containing 10% D₂O (Figure S17) and in the presence of an equimolar amount of glutathione (GSH, Figure S18). Besides, UV‐vis spectroscopy demonstrated their biological stability in phosphate buffer saline (PBS, pH 7.4), PBS with GSH, and PBS with 10% fetal bovine serum over 48 h, respectively (Figure S19, S20, and S21). Liquid chromatography‐mass spectrometry analysis showed that Cu1 and Cu2 could also maintain stability in mouse plasma over 48 h without significant change in their peak areas (Figure S22). Furthermore, UV‐vis spectroscopy confirmed that the valence of copper centers within Cu1 and Cu2 maintained +1 in PBS over 48 h (Figure S23). We next investigated the stability of Cu1 and Cu2 in PBS at varying pH (7.4, 6.5, and 5.5) over 24 h using UV‐vis spectroscopy. At pH 7.4 (Figures S24A and S25A), both Cu1 and Cu2 maintained stability with negligible absorption loss. In contrast, decreases in the absorption of Cu1 and Cu2 were observed at pH 6.5 and 5.5 (Figures S24B–D and S25B–D), which was attributed to partial ligand protonation under acidic conditions. These results demonstrated that Cu1 and Cu2 exhibit higher stability under neutral conditions but lower stability under acidic conditions over 24 h. Additionally, Cu1 and Cu2 displayed similar emission peaks at 463 and 486 nm (excited at 343 nm), respectively (Figure S26).

Cyclic voltammetry measurement using [(nBu)₄N]PF₆ as the supporting electrolyte showed that Cu1 and Cu2 possessed a similar reversible redox cycle in PBS (Figure S27). This electrochemical behavior corroborates the ability of Cu1 and Cu2 to undergo redox cycling under physiological conditions, a critical prerequisite for Fenton‐like reactivity. To evaluate the capability of the synthesized copper(I) complexes to trigger Fenton‐like reactions under various pH conditions, we used 3,3’,5,5’‐tetramethylbenzidine (TMB) as the ∙OH probe. As shown in Figures 1D and S28, the results showed that the ∙OH generation capability of Cu1 and Cu2 increased as pH decreased, consistent with the consensus that more acidic conditions would enhance Fenton‐like reaction efficiency [41]. At physiological pH (7.4), negligible ∙OH generation promoted by Cu1 or Cu2 was observed, thereby reducing potential ROS damage to healthy tissues. However, Cu1 and Cu2 initiated a significant Fenton‐like reaction under acidic conditions (pH 6.5 and 5.5), which indicated the potential of disrupting intracellular redox homeostasis within acidic conditions (e.g., lysosomes). Notably, Cu2 exhibited higher catalytic rate for ∙OH generation than Cu1 (Figures 1E and S29). Electron spin resonance (ESR) measurement using 5,5‐dimethylpyrroline‐N‐oxide (DMPO) as the spin trap agent further confirmed the generation of ∙OH through Fenton‐like reactions promoted by Cu1 and Cu2 (Figure 1F). Subsequently, theoretical calculations were further performed to elucidate the mechanism underlying the enhanced Fenton‐like catalytic activity of Cu2. Ground‐state geometries of Cu1 and Cu2 were optimized at the B3LYP/def2‐SVP/SMD level, while excitation energies were computed using PBE0/def2‐TZVP/SMD (Figure S30). The HOMO‐LUMO energy gap of Cu2 (3.87 eV) was narrower than that of Cu1 (4.00 eV), implying enhanced electron transfer efficiency in Cu2 that facilitates Fenton‐like reactions.

2.2. Cu2 Can Trigger both Cuproptosis and Ferroptosis In Vitro

We next evaluated the cytotoxicity of Cu1 and Cu2 against different tumor cell lines, including breast adenocarcinoma MCF‐7, MDA‐MB‐231 and 4T1, lung adenocarcinoma A549, and colon adenocarcinoma HCT116 using the 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) assay. In cytotoxicity assays, Cu2 demonstrated greater cytotoxicity than Cu1 across all tested cancer cell lines. The enhanced antitumor activity of Cu2 may be attributed to its superior Fenton‐like catalytic ability. Notably, Cu2 exhibited the strongest activity against MCF‐7 cells, with an IC₅₀ of 0.40 ± 0.02 µg/mL, which was much stronger than Cu1 (IC₅₀ = 2.88 ± 0.52 µg/mL) and oxaliplatin (OXA, a widely used clinical metal‐based antitumor drug, IC₅₀> 20 µg/mL) (Table 1). Calcein AM/propidium iodide (PI) staining further validated a dose‐dependent of Cu2‐induced cell death, evidenced by a marked reduction in green fluorescence of live MCF‐7 cells (Figure 2A). To further investigate the contribution of the ligands (Bim‐1 and Bim‐2) in Cu1 and Cu2, a series of MTT assays were performed to evaluate the cytotoxicity against MCF‐7 cells of CuCl (IC₅₀> 100 µM), Bim‐1 (IC₅₀ = 7.10 ± 2.29 µM), CuCl + Bim‐1 (2:1) (IC₅₀ = 7.56 ± 1.59 µM), Bim‐2 (IC₅₀ = 6.26 ± 1.28 µM), and CuCl + Bim‐2 (2:1) (IC₅₀ = 4.52 ± 1.20 µM) (Table S10). These results showed that compared with Cu1 (IC₅₀ = 3.77 ± 0.68 µM), CuCl + Bim‐1 (2:1) was approximately 2.0‐fold less potent. Notably, compared with Cu2 (IC₅₀ = 0.49 ± 0.02 µM), CuCl + Bim‐2 (2:1) was approximately 9.2‐fold less potent, which indicated that the coordination significantly enhances the cytotoxicity against tumor cells. Furthermore, Cu2 demonstrated a good selectivity between tumor cells (MCF‐7) and normal cells (MCF‐10A) with a selectivity index (SI) of 12.8, which is better than Cu1 (SI = 2.51), CuCl (SI < 0.85), Bim‐1 (SI = 1.00), CuCl + Bim‐1 (SI = 0.78), Bim‐2 (SI = 1.3) and CuCl + Bim‐2 (SI = 1.12). These results indicated that Cu2 exhibited stronger antitumor effects and better selectivity. Accordingly, Cu2 was chosen for further studies.

TABLE 1.

The cytotoxicity of Cu1, Cu2, and OXA towards different tumor cell lines.

Complex	IC50 values (µM)
	MCF‐7	MDA‐MB‐231	4T1	A549	HCT116
Cu1	3.77 ± 0.68	3.48 ± 0.44	3.71 ± 0.43	6.47 ± 1.26	6.28 ± 0.21
Cu2	0.49 ± 0.02	1.07 ± 0.15	1.60 ± 0.29	0.68 ± 0.16	0.70 ± 0.16
OXA	> 50	14.85 ± 5.89	45.7 ± 1.44	42.6 ± 9.6	46.0 ± 7.8

FIGURE 2.

Cu2 can trigger both cuproptosis and ferroptosis in vitro. (A) Fluorescence images of MCF‐7 cells for calcein‐AM/PI assay after treated with different concentrations of Cu2. Scale bar: 400 µm. (B) CLSM images of MCF‐7 cells upon incubation with Cu2 for 18 h. The Pearson correlation coefficient was calculated by Fiji software. Scale bar: 50 µm. (C) The impact of various inhibitors (TTM, 3 µM; 3‐MA, 100 µM; Nec‐1, 6 µM; Fer‐1, 6 µM; z‐VAD‐fmk 20 µM) on the cell death induced by Cu2 (n = 3). (D) The dose‐dependent intracellular copper content of MCF‐7 cells following treated by Cu2 (n = 3). (E) The representative CLSM images to determine the oligomerization of DLAT in MCF‐7 cells after treated by different concentrations of Cu2. Scale bar: 50 µm. (F) Western blot assay of free lipoylated DLAT, LIAS, and FDX1 in MCF‐7 cells after various concentrations of Cu2 treatments. The uncropped images can be found in Figure S42. (G, H) The generation of ROS in MCF‐7 cells after treatment with different concentrations of Cu2 measured by flow cytometry and fluorescence images. Scale bar: 100 µm. (I, J) Dose dependent changes of cellular MDA and GSH levels in MCF‐7 cells after different concentrations of Cu2 treatments (n = 3). (K) Fluorescence images of MCF‐7 cells for JC‐1 assay after incubated with various concentrations of Cu2. Scale bar: 200 µm. The data are presented as the mean values ± SD or images from one of three representative experiments. Statistical analysis was performed using one‐way ANOVA with Dunnett's multiple comparisons. (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05).

To further investigate the mechanism underlying the remarkable cytotoxicity of Cu2, the intracellular distribution of Cu2 was initially observed under a confocal laser scanning microscope (CLSM). After incubation of Cu2 and MCF‐7 cells, the blue fluorescence of Cu2 showed obvious colocalization with both the endoplasmic reticulum (ER) and mitochondria (Mito), with Pearson correlation coefficients (PCCs) of 0.90 and 0.87, respectively (Figure 2B). This subcellular distribution of Cu2 suggested mitochondria‐targeted copper delivery for potential cuproptosis induction. Furthermore, Cu2 also showed colocalization in lysosomes (PCC = 0.56), which is beneficial for its Fenton‐like catalysis under acidic conditions, thereby elevating intracellular ROS levels to activate ferroptosis. Similarly, comparable colocalization results were observed for Bim‑2 (Figure S31), suggesting that the ligand predominantly determines the intracellular distribution of Cu2. Furthermore, 3‐methyladenine (3‐MA, the autophagy inhibitor), necrostatin‐1 (Nec‐1, the necrosis inhibitor), ferrostatin‐1 (Fer‐1, the ferroptosis inhibitor) ammonium tetrathiomolybdate (TTM, the cuproptosis inhibitor), and z‐VAD‐fmk (the apoptosis inhibitor) were used to determine the predominant modes of cell death induced by Cu2, and found that only Fer‐1 and TTM could efficiently mitigating the cell death caused by Cu2 (Figure 2C). The above results indicated that Cu2 might trigger both cuproptosis and ferroptosis pathways.

Cuproptosis is triggered by the direct binding of Cu⁺ to lipoylated components of the tricarboxylic acid cycle. This interaction leads to the oligomerization of lipoylated proteins, such as dihydrolipoamide S‐acetyltransferase (DLAT), and the degradation of iron‐sulfur cluster proteins, including lipoic acid synthetase (LIAS) and ferredoxin 1 (FDX1) [31, 42]. Consequently, mitochondrial proteotoxic stress is induced, ultimately resulting in cell death. Cu2 localizes in Mito and features a dicoordinate geometry that might promote intracellular copper ion accumulation for effective cuproptosis induction. To evaluate the copper overload induced by Cu2, the intracellular copper concentrations of MCF‐7 cells with varied treatments were determined by Cell Copper Colorimetric Assay Kit. As shown in Figure 2D, after 1.6 µg/mL Cu2 treatment, copper ion levels in MCF‐7 cells increased significantly (up to 3.39‐fold more than the Control group), which revealed that Cu2 could effectively induce copper overload and provided a necessary precondition for inducing cuproptosis. Moreover, after treatment with increasing concentrations of Cu2, DLAT oligomerization became clearly observable under fluorescence confocal microscopy through the enhanced green fluorescence (Figure 2E). Additionally, a western blot (WB) assay was conducted to evaluate three key markers of cuproptosis, including the levels of free lipoylated DLAT, oligomerized DLAT, LIAS, and FDX1. Figures 2F and S32 suggested that Cu2 could induce oligomerization of lipoylated DLAT and downregulation of LIAS and FDX1, which consequently leaded to cuproptosis.

By sharing pathways of oxidative stress and mitochondrial dysfunction, cuproptosis synergistically potentiates ferroptosis [43]. Subsequently, we further investigated whether ferroptosis could also be induced by Cu2. Ferroptosis is characterized by several key features, including ROS surge that promotes GSH depletion, lipid peroxidation (LPO) accumulation, glutathione peroxidase 4 (GPX4) downregulation, and mitochondrial dysfunction [44, 45]. First, the impact of Cu2 on cellular ROS levels was detected using fluorescent probe 2’,7’‐dichlorodihydrofluorescein diacetate (DCFH‐DA). As shown in Figure 2H, MCF‐7 cells treated with Cu2 displayed remarkably bright green fluorescence compared with the control group. Flow cytometry analysis presented similar results in Cu2‐treated cells in a dose‐dependent manner (Figure 2G), illustrating the robust ROS generation promoted by Cu2. Moreover, we used BDPY 581/591 C11, a lipid peroxidation probe, to evaluate the changes in intracellular LPO levels, and the enhanced green fluorescence indicating the decreased LPO levels in MCF‐7 cells (Figure S33). In addition, the increase of intracellular malondialdehyde (MDA) content, which is the final metabolite of LPO and a crucial factor in the induction of ferroptosis [24], was also observed following Cu2 treatment (Figure 2I). Next, the GSH levels in MCF‐7 cells were measured. As shown in Figure 2J, the GSH levels decreased significantly following Cu2 treatment in a dose‐dependent manner, demonstrating that Cu2 could amplify the intracellular oxidative damage by GSH depletion. As the primary regulator of ferroptosis, GPX4 functions as an LPO scavenger utilizing GSH as the reducing substrate [46], which was further investigated by WB assay. As shown in Figure S34, GPX4 expression was prominently downregulated upon Cu2 treatment in a dose‐dependent manner, leading to the block of LPO elimination. Besides, mitochondrial membrane potential (MMP) detected by JC‐1 staining confirmed the oxidative mitochondrial stress caused by elevating ROS. As shown in Figure 2K, Cu2‐treated MCF‐7 cells exhibited significantly reduced red fluorescence (JC‐1 aggregates) and increased green fluorescence (JC‐1 monomers) compared to the control group, which demonstrated the mitochondrial dysfunction caused by Cu2. Moreover, the alternations in cell morphology upon Cu2 treatment were detected by transmission electron microscopy (TEM). Compared with the control group, MCF‐7 cells treated with Cu2 showed obvious mitochondrial damage (Figure S35), including reduced mitochondrial volume, cristae disappearance, mitochondrial membrane rupture, and decreased mitochondrial membrane density. Collectively, these results well corroborated the concurrent activation of cuproptosis and ferroptosis by Cu2 in vitro.

2.3. Cu2 Can Induce ICD In Vitro

Recent studies have indicated that both ferroptosis and cuproptosis can also trigger ICD [33, 34, 35, 36], which is closely related to ER stress to trigger the unfolded protein response (UPR) and release of DAMPs [47, 48, 49]. WB results demonstrated that Cu2 significantly increased the phosphorylation level of eIF‐2α and upregulated the expression of C/EBP homologous protein (CHOP) and glucose‐regulated protein 78 (GRP78) in MCF‐7 cells (Figure 3A), which a closely associated with ER stress. Besides, TEM images showed the swelling and vacuolization of ER in MCF‐7 cells (Figure S36), supporting the ER stress caused by Cu2.

FIGURE 3.

Cu2 can induce ICD in vitro. (A) Western blot assay of ER stress related hallmarks in MCF‐7 cells, including GRP78, p‐eIF2α, and CHOP, after different concentrations of Cu2 treatments. The uncropped images can be found in Figure S44. (B, C) The exposure of CRT was imaged through immunostaining of MCF‐7 cells. Scale bar: 50 µm. Flow cytometry further showing dose‐dependent CRT release in MCF‐7 cells. (D) Dose dependent HMGB1 released from MCF‐7 cells detected by an enzyme‐linked immunosorbent assay (ELISA) kit after Cu2 treatments (n = 3). (E) Fluorescence imaging for HMGB1 release from MCF‐7 cells after treatment with different concentrations of Cu2. Scale bar: 50 µm. (F) The release of ATP levels in the extracellular supernatant of MCF‐7 cells after various concentrations Cu2 treatments (n = 3). (G) Flow cytometry analysis for CD80 and CD86 expression on DC2.4 cells after being incubated with PBS or various concentrations of Cu2. The data are presented as the mean values ± SD or images from one of three representative experiments. Statistical analysis was performed using one‐way ANOVA with Dunnett's multiple comparisons. (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05).

Accordingly, we then evaluated three ICD related DAMPs, including calreticulin (CRT), high‐mobility group box 1 (HMGB1), and adenosine triphosphate (ATP). Following treatment with Cu2, obvious CRT exposure on the surface of tumor cells was observed by confocal microscopy and flow cytometry (Figure 3B,C), which facilitated the engulfment of tumor cells by antigen‐presenting cells. Moreover, immunofluorescence images, enzyme‐linked immunosorbent assay (ELISA), and WB assay revealed more HMGB1 release in the supernatant and less remaining in the cells (Figures 3D,E and S37). Additionally, extracellular ATP levels also exhibited a concentration‐dependent increase following treatment with Cu2 (Figure 3F). Furthermore, DC maturation, a hallmark of adaptive immunity, was then evaluated to confirm the adaptive immune response induced by Cu2. To investigate whether the ICD caused by Cu2 could lead to DC maturation, we incubated DC2.4 cells with 4T1 cells (treated with different concentrations of Cu2) by using a transwell system. As shown in Figure 3G, Cu2 triggered a significantly elevated expression of CD80/CD86 in a dose‐dependent manner, which was up to 35.47% higher than the control group, validating the augmented DC maturation owing to Cu2. Accordingly, in vitro ICD studies of Bim‐2 and positive controls with OXA (a well‐known ICD inducer) were also conducted. As illustrated in Figure S38, Bim‐2 alone did not trigger typical ICD hallmarks, whereas 0.8 µg/mL (0.97 µM) Cu2 induced stronger ICD responses than 75 µM OXA, suggesting that the observed effects were indeed due to metal coordination and that Cu2 functioned as an effective ICD inducer. These results collectively manifested that ferroptosis and cuproptosis induced by Cu2 could effectively induce ICD in vitro.

2.4. Cu2 Exhibits Ideal Antitumor Efficacy In Vivo

Given the above results demonstrating the potential of Cu2 to induce both cuproptosis and ferroptosis in vitro, its therapeutic potential was assessed in murine model. A 4T1 unilateral tumor mouse model was established to evaluate the antitumor efficacy of Cu2 in vivo (Figure 4A). OXA was used for the positive control of antitumor effects. After the tumor volume reached approximately 100 mm³, all mice were randomly divided into seven groups: (1) Control; (2) 2 mg/kg Cu2; (3) 4 mg/kg Cu2; (4) 8 mg/kg Cu2; (5) 4 mg/kg OXA; (6) 8 mg/kg OXA; (7) 12 mg/kg OXA. OXA and Cu2 were administered intratumorally (i.t.) on day 1, 4, 7 and 10 in OXA and Cu2 treated groups, respectively. The tumor volume was routinely monitored for 14 days. Notably, both Cu2 and OXA could effectively inhibit the tumor growth in a dose‐dependent manner, while 4 mg/kg Cu2 group exhibited similar antitumor efficacy to 8 mg/kg OXA group (Figure 4B–D). Accordingly, Hematoxylin and Eosin (H&E) stained images of tumor sections also revealed the obvious shrinking and discrete cancer cells in Cu2 group, suggesting the ideal antitumor efficacy of Cu2 (Figure S39). Meanwhile, unlike the 12 mg/kg OXA treatment group, which demonstrated significant body weight loss and hepatotoxicity, the 8 mg/kg OXA treatment group demonstrated slight body weight loss, while 4 mg/kg Cu2 demonstrated no significant change in mice body weight, nor structural or pathological changes in major organs (Figures 4E and S39), indicating the better biosafety of Cu2 than OXA. To further investigate the mechanism underlying the antitumor efficacy of Cu2 in vivo, histological analysis was performed on tumor tissue sections. As shown in Figure 4F, compared with the control group, the significantly enhanced fluorescent signals of DLAT and obviously reduced green fluorescence signals of GPX4 indicated cuproptosis and ferroptosis induced by Cu2 in vivo, which corresponded to our previous finding in vitro. These results collectively indicated the ideal antitumor efficacy of Cu2 and the biosafety of Cu2 for therapeutic applications as an inducer of both cuproptosis and ferroptosis in vivo.

FIGURE 4.

Cu2 exhibits ideal antitumor efficacy in vivo. (A) Schematic illustration of the experimental design to evaluate the antitumor effects of Cu2 by intratumoral treatment. OXA (4, 8, and 12 mg/kg) and Cu2 (2, 4, and 8 mg/kg) were administered intratumorally (i.t.) on Days 1, 4, 7, and 10 in OXA and Cu2 group, respectively (n = 5). (B) Tumor weight changes of 4T1 unilateral tumor mouse model in different groups (mean ± SD, n = 5). (C) Photographs of tumors from mice post various treatments collected on day 14. (D). Tumor volume changes of 4T1 unilateral tumor mouse model in different groups (mean ± SD, n = 5). (E) Body weight change of 4T1 unilateral tumor mouse model in different groups (mean ± SD, n = 5). (F) Representative immunofluorescence images of GPX4 and DLAT tumor dissection obtained from excised tumors with different treatments. Scale bar: 100 µm. The data are presented as the mean values ± SD or images from one of five representative experiments. Statistical analysis was performed using one‐way ANOVA with Tukey's post‐hoc test. (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05).

2.5. Cu2 Can Activate a Robust Immune Response and Enhance Tumor Immunotherapy Efficacy

To further investigate the immunogenicity triggered by Cu2 as an ICD inducer in vivo, a bilateral 4T1 tumor model was subsequently constructed. Initially, 4T1 cells were inoculated on the left side of the mice to establish primary tumors. Three days later, 4T1 cells were further inoculated on the right side of the mice as distant tumors (Figure 5A). Five days later, all mice were randomized into five groups: (1) Control; (2) αPD‐1 (3.0 mg/kg); (3) OXA (8.0 mg/kg); (4) Cu2 (4.0 mg/kg), and (5) Cu2 (4.0 mg/kg) + αPD‐1 (3.0 mg/kg) and received intratumoral injections of OXA and Cu2 into the primary tumors only while αPD‐1 was administrated by intravenous injection. The volume and weight of the tumor were routinely monitored for 14 days. As shown in Figures 5B–F and S40, both the Cu2 group and the Cu2 + αPD‐1 group effectively inhibited the growth of primary and distant tumors, which demonstrated better antitumor effects compared to the OXA and αPD‐1 groups. At the end of the treatment, the inhibition rates for the primary and distant tumors in the Cu2 group were approximately 60.41% and 78.12%, respectively, while Cu2 + αPD‐1 group exhibited the highest inhibition rates, which were approximately 85.88% (primary) and 93.42% (distant). To confirm the immunotherapeutic progression of Cu2 as an ICD inducer, the exposure of CRT and the release of HMGB1 were evaluated by IHC. As demonstrated in Figure 5G, increased green fluorescence of CRT, as well as disappeared red fluorescence of HMGB1 in Cu2 and Cu2 + αPD‐1 groups indicated the enhanced ICD effect in distant tumor induced by the combination of Cu2 and αPD‐1, which would stimulate strong activation in T cell immunity subsequently.

FIGURE 5.

Cu2 can activate a robust immune response and enhance tumor immunotherapy efficacy. (A) Schematic illustration of the subcutaneous 4T1 bilateral tumor model. All mice were randomized into five groups: (1) Control, (2) αPD‐1 (3.0 mg/kg), (3) OXA (8.0 mg/kg), (4) Cu2 (4.0 mg/kg), and (5) Cu2 (4.0 mg/kg) + αPD‐1 (3.0 mg/kg). OXA (8.0 mg/kg) and Cu2 (4.0 mg/kg) were administered intratumorally (i.t.) and αPD‐1 (3.0 mg/kg) were administered intravenously (i.v.) in primary tumors on Days 1, 4, 7, and 10 respectively (n = 5). In Cu2 + αPD‐1 group, αPD‐1 was administered intravenously (i.v.) on Days 1, 4, 7, and 10 respectively. (B) Photographs of primary and distant tumors from mice post various treatments collected on day 14. Primary tumor (C) volume and (D) weight changes of mice with various treatments (n = 5). Distant tumor (E) volume and (F) weight changes of mice with various treatments (n = 5). (G) Representative immunofluorescence image of CRT and HMGB1 tumor dissection obtained from excised tumors with different treatments. Scale bar: 100 µm. The flow cytometry of (H) CD3⁺CD4⁺ T cells and (I) CD3⁺CD8⁺ T cells in spleen after various treatments for 14 days. The data are presented as the mean values ± SD or images from one of five representative experiments. Statistical analysis was performed using one‐way ANOVA with Tukey's post‐hoc test. (****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05).

Accordingly, flow cytometric analysis was conducted to investigate T cell‐mediated immune responses induced by Cu2. In the Cu2 group, the proportion of CD3⁺CD4⁺ T lymphocytes in the Cu2 group was approximately 34.8%, and in the Cu2 + αPD‐1 group, it was about 40.8%, which was 21.0% and 27.0% higher than the Control group (Figure 5H). The proportion of CD3⁺CD8⁺ T lymphocytes of Cu2 group was approximately 15.7%, while in the Cu2 + αPD‐1 group, it was about 25.2%, which was 10.2% and 19.7% higher than the Control group (Figure 5I). In addition, as the key factor of T cell‐mediated immune responses, the secretion levels of immunostimulatory cytokines including interferon γ (IFN‐γ) IL‐6 and TNF‐α in Cu2 and Cu2 + αPD‐1 groups were further evaluated by ELISA. Both Cu2 and Cu2 + αPD‐1 groups showed an increased secretion levels of IFN‐γ, IL‐6 and TNF‐α (Figure S41), which further improved the effective induction of antitumor immune responses. Moreover, Cu2 + αPD‐1 groups demonstrated higher secretion levels of IFN‐γ, IL‐6 and TNF‐α than Cu2 group, indicating the enhanced immunotherapy efficacy of αPD‐1 combined with Cu2. These results confirmed that Cu2 could activate a robust immune response and enhanced the immunotherapy efficacy of αPD‐1.

3. Conclusion

Conventional stepwise synthetic strategies for metal complexes have limited efficient access to innovative metal‐based antitumor agents. To address this, we presented a paradigm‐shifting domino synthetic strategy that seamlessly integrated organic ligand formation and coordination assembly in one pot, and efficiently converted simple precursors to novel dinuclear dicoordinate copper(I) complexes (Cu1 and Cu2) with in situ‐generated tetraarylethane ligands featuring the bis‐imidazo[1,5‐a]pyridine scaffold. Crucially, crystallography and time‐dependent ESI‐MS elucidated the intricate mechanism of this domino process as a precisely orchestrated 9‐step sequence, involving critical transformations such as aldimine condensation, coordination, nucleophilic addition, β‐H elimination, homolytic C‐O bond cleavage, and C‐C coupling.

Besides, although concurrent induction of cuproptosis and ferroptosis with potential ICD offers potent antitumor efficacy, small‐molecule copper complexes capable of elevating intracellular copper ions while exhibiting sufficient Fenton‐like activity to trigger these pathways remain scarce. To address this issue, Cu2 with unique dicoordinate geometry has been rationally designed, which enables: (1) mitochondria‐targeted copper delivery to effectively induce cuproptosis; (2) sufficient Fenton‐like catalysis for ROS generation to disrupt intracellular redox homeostasis and trigger ferroptosis; (3) concurrent activation of cuproptosis‐ferroptosis inducing ICD. Moreover, Cu2 potentiates immune checkpoint inhibitor therapy by enhancing tumor immunogenicity in vivo. To the best of our knowledge, Cu2 represents the first small‐molecule copper complex capable of integrated cuproptosis‐ferroptosis‐ICD induction. Our findings underscore the immense potential of rationally designed synthetic chemistry in unlocking new mechanisms for cancer therapy and pave the way for the development of next‐generation metal‐based immunotherapeutic agents targeting synergistic non‐apoptotic pathways.

Author Contributions

Ning Tian: methodology, data curation, investigation, writing – original draft, formal analysis. Haoyu Ju: methodology, investigation, writing – original draft, formal analysis, data curation. Yu Liu: investigation, methodology, software, formal analysis, validation. Jinmei Huang: investigation, methodology, software, formal analysis. Zhenggang Luan: investigation, methodology, software, formal analysis. Qifeng Hou: software, formal analysis, methodology, investigation. Qing Chen: supervision, validation, data curation, resources. Bin Zhang: conceptualization, methodology, writing – original draft, validation, data curation, supervision. Jin Huang: conceptualization, methodology, writing – review and editing, supervision, project administration, funding acquisition. Ming‐Hua Zeng: project administration, funding acquisition, writing – review and editing, methodology, conceptualization, supervision, validation.

Conflicts of Interest

The authors declare no conflict of interest.

Supporting information

Supporting File 1: anie73452‐sup‐0001‐SuppMat.docx.

Supporting File 2: anie73452‐sup‐0002‐cif.zip.

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.