Cuproptosis tracker: Visualizing organelle dynamics with a dual-targeted fluorescent probe

Abstract

Cuproptosis often interacts with mitochondrial (Mito) dysfunction and lipid droplets (LDs) metabolism disturbances, thus resulting in programmed cell death, whereas their dynamic interaction lacks a rational analyzing tool. Herein, we show a Mito-LDs dual-targeted fluorescent reporter (MLR) for tracking the Mito-LDs interaction during cuproptosis by dynamic monitoring of intracellular sulfur dioxide (SO₂) dynamics. MLR integrates a coumarin-derived SO₂-responsive core linked via piperazine to a benzopyronium Mitol anchor, enabling one-step synthesis with exceptional sensitivity (0.34 μM) and rapid response (<10 s). Live-cell imaging demonstrated MLR's SO₂-triggered translocation from Mito to LDs during cuproptosis, directly visualizing inter-organelle communication. Dual fluorescence channel imaging associated SO₂ fluctuations with Mito-LDs targeting, revealing the interaction between LDs-Mito during Cu²⁺ and elesclomol (ES) induced apoptosis. In addition to imaging, MLR-based test strips and hydrogels can achieve rapid (<1 min) on-site SO₂ detection. As a dual-organelle tracer for cuproptosis, MLR overcomes single-target probe limitations, offering a transformative platform to analyze spatiotemporal organelle dynamics for advancing diagnostic tools development.

Graphical abstract

Highlights

• •Dual-organelle tracking enabling SO₂-mediated fluorescence tracing during cuproptosis with low LOD and rapid response.
• •Achieving SO₂ specific detection with dual channel ratio imaging to reveal SO₂ metabolic law in cuproptosis.
• •Rapid on-site SO₂ detection, extending to real-time tracking of copper toxicity in zebrafish.

1. Introduction

Cuproptosis manifests as copper-dependent cell death through interconnected mechanisms: excessive Cu²⁺ binds lipoylated tricarboxylic acid cycle (TCA) cycle proteins, inducing aggregation and iron-sulfur cluster destabilization, while ferredoxin-1 (FDX1)-mediated Cu⁺ generation amplifies proteotoxicity via pyruvate dehydrogenase lipoylation [[1], [2], [3]]. Copper overload can trigger a Fenton-like reaction, produce reactive oxygen species (ROS), and cooperate with protein toxic stress to damage the hub of ROS metabolism and sulfur antioxidant synthesis in mitochondria (Mito) [[4], [5], [6]]. Mito dysfunction consumes sulfur dioxide (SO₂) (a key ROS scavenger), thereby exacerbating oxidative damage and destroying redox balance [7,8]. Meanwhile, Mito damage makes copper death metabolism dependent on lipid droplets (LDs), resulting in LDs accumulation and lipotoxicity, which reveals the close connection between Mito-LDs during copper death [9,10]. Although sulfur antioxidants such as SO₂ can reduce oxidative stress, their depletion during cell death can cause a self-sustaining cycle of redox collapse. At this time, tools are needed to dynamically map the process of organelle interaction and sulfur metabolism in copper death [[11], [12], [13], [14], [15]].

Given the complexity of this mechanism, fluorescent probes have become an indispensable tool for the dynamic tracking of copper ions and their metabolites, owing to their advantages such as noninvasiveness, real-time sampling capability, high sensitivity, and superior spatiotemporal resolution [[16], [17], [18]]. Recently, a large number of SO₂ fluorescent probes targeting Mito or LDs have been developed. For example, bifunctional probes, such as TPA-SO₂ for dual-parameter imaging of LDs polarity and SO₂ during ferroptosis, and AND-logic-gate probes for synergistic detection of Mito bisulfite and viscosity, have shown high specificity and ultralow detection limit [8,19]. Although some progress has been made in the relationship between the cuproptosis process and organelles, there are still gaps in monitoring the sulfide dynamics between organelles. Most of the reported probes only focus on a single organelle, as emphasized earlier. There is a need to study the relationship between cuproptosis and double organelles. As far as we know, the use of fluorescent probes that monitor SO₂ dynamics to explore the interaction between Mito-LDs during cuproptosis has not been reported. To effectively trace these interactions during cuproptosis, an ideal probe should possess the following characteristics: (i) dual-targeting capability for both Mito and LDs, (ii) rapid SO₂ responsiveness, and (iii) high specificity and sensitivity toward SO₂.

To this end, we designed a coumarin-benzopyran hybrid probe Mito-LDs dual-targeted fluorescent reporter (MLR), which can simultaneously target Mito and LDs. In MLR, the α, β-unsaturated ketone moiety selectively reacts with SO₂ through Michael addition [20,21], destroying the electron distribution of benzopyran cations and inducing a fluorescence signal transition. MLR achieved nanomolar sensitivity, and short-term response to SO₂ detection and allowed successfully visualizing the dynamic change of intracellular SO₂ level between Mito and LDs during cuproptosis. By analyzing the distribution of SO₂ in living cells and zebrafish, MLR achieves rapid and accurate tracking of the synergy between organelles and provides different insights into the treatment strategies for copper toxicity and redox imbalance recovery.

2. Materials and methods

2.1. Ethics approval

All experimental procedures and animal protocols in this study were approved by the Animal Care & Welfare Ethical Committee of Guangxi Medical University, Nanning, China (Approval No.: 202405032). The study was conducted in accordance with the principles of the Declaration of Helsinki.

2.2. Experimental instruments and chemicals

All the chemicals were reagent grade and used as received without further purification. Super-dry solvent including tetrahydrofuran (THF), 1,4-dioxane, and dichloromethane (DCM) with molecular sieves was purchased from Innochem (Beijing, China) and used as received. Dry solvents were purchased from Beijing Chemical Works or Energy Co., Ltd (Beijing, China). All products were purified by silica gel (200–300 mesh) column chromatography. TLC plates and silica gel were purchased from Yantai Jiangyou Silica Gel Development Co, Ltd (Yantai, Shandong, China). At room temperature, proton nuclear magnetic resonance (¹H NMR) and carbon-13 nuclear magnetic resonance spectra (¹³C NMR spectra) were recorded on Bruker AVANCE600 MHZ (Östliche Rheinbrückenstr., Karlsruhe, Germany) with CDCl₃, DMSO-d₆ as a solvent, and tetramethylsilane (TMS) as the internal reference. High-resolution mass spectrometry (HRMS) analysis was performed in the electrospray ionization (ESI) mode of the Bruker apex Ultra mass spectrometer (Bruker, Östliche Rheinbrückenstr., Karlsruhe, Germany). The pH was determined using equipment from Sedoris Scientific Instruments Company (Göttingen, Lower Saxony, Germany). Fluorescence spectra were measured on Shimadzu F6000 fluorescence spectrophotometer (Shimadzu, Kyoto, Japan). The ultraviolet–visible (UV–Vis) absorption spectrum was obtained on the Cary 60 UV–Vis spectrophotometer (Santa Clara, CA, USA). The cholecystokinin octapeptide (CCK-8) kit was purchased from Biyuntian Co., Ltd (Shanghai, China). Fluorescence imaging experiments were performed using an Olympus FW300 confocal microscope and Olympus upright fluorescence microscope (Olympus, Tokyo, Japan).

2.3. Synthesis of MLR

Compound 3 (18 mg, 0.05 mmol), compound 6 (198 mg, 0.05 mmol), 4-dimethylamino-pyridine (DMAP) (0.61 mg, 0.005 mmol), and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDCI) (19 mg, 0.1 mmol) were dissolved in 2 mL of anhydrous DCM and stirred at N₂ protection at room temperature for 24 h. Reducing the pressure by removing the solvent yielded a crude product. Probe MLR was isolated by column chromatography (200–300 silica gel, leach: V_DCM:V_MeOH = 20:1), purple solid, 213 mg, 85.14% yield. ¹H NMR (600 MHz, DMSO-d₆) δ8.61 (d, J = 4.2 Hz, 1H), 8.24 (s, 2H), 7.98–7.87 (m, 2H), 7.77 (d, J = 8.7 Hz, 1H), 7.62 (d, J = 7.3 Hz, 2H), 7.51 (t, J = 7.3 Hz, 1H), 7.46 (t, J = 7.5 Hz, 2H), 7.33 (dd, J = 42.7, 21.3 Hz, 2H), 7.15 (s, 2H), 6.84 (d, J = 7.8 Hz, 1H), 6.62 (s, 1H), 3.95–3.80 (m, 2H), 3.76–3.56 (m, 10H), 3.49 (d, J = 6.3 Hz, 4H), 1.24 (t, J = 6.9 Hz, 6H), 1.16 (t, J = 6.9 Hz, 6H). ¹³C NMR (600 MHz, DMSO-d₆) δ 167.45, 163.51, 158.68, 158.43, 156.02, 155.51, 154.79, 151.80, 148.41, 133.33, 132.60, 132.26, 131.68, 131.17, 129.06, 129.12, 128.46, 126.88, 120.73, 118.44, 117.56, 117.04, 116.94, 114.39, 110.19, 108.66, 106.76, 105.28, 82.36, 47.32, 46.27, 45.67, 44.74, 32.01, 29.89, 12.79. HRMS (ESI) m/z: calculated for C₄₅H₄₅N₄O₄⁺ [M +H]⁺: 705.3435, found: 705.3439. The synthetic route of the probe was mentioned in Figs. S1–S11.

2.4. Optical experiments

The probe MLR was dissolved in dimethyl sulfoxide as a reserve liquid with a concentration of 1 mg/mL. Various analytes (Mg²⁺, K⁺, Ca²⁺, Zn²⁺, I⁻, H₂O₂, H₂PO₄⁻, S₂O₃²⁻, SO₃²⁻, SO₄²⁻, Cl⁻, S²⁻, PO₄³⁻, HSO₃⁻, NO²⁻, NO₃⁻, AC⁻, Asp, Arg, Pro, Glu, Lys, Ala, Ile, glutathione (GSH), Hcy, Cys, Gly, and Glu) at 10 μM were selected to examine the anti-interference capability of the MLR in the testing system.

2.5. Fluorescence imaging in HeLa cells

MLR (10 μM) was incubated with HeLa cells for 30 min, and the cells were washed twice with phosphate-buffered saline (PBS) to remove residual probes. Then the cells were cultured with 50 and 200 μM NaHSO₃ (SO₂ was added to the probe system in the form of NaHSO₃) or 200 and 400 μM Cys for 30 min. In addition, HeLa cells seeded into 6-well culture plates were divided into six groups: (1) Blank, (2) Control, (3) Cu²⁺, (4) elesclomol (ES), (5) Cu²⁺ + ES, and (6) Cu²⁺ + ES + tetrathiomolybdate (TTM). In detail, HeLa cells were cultured with the above group for 2 h. Subsequently, the cells were stained with MLR (10 μM) for 30 min to detect SO₂. Confocal laser scanning microscope (CLSM) (Olympus FV3000, Hachioji, Tokyo, Japan) was used to obtain the fluorescent cell images. The orange channel (500−600 nm) was obtained under 488 nm excitation, and the red channel (550−650 nm) was collected under 561 nm excitation.

2.6. Mito-LDs co-localization imaging

Blue FX490 (a Mito localization commercial dye) and BODIPY 630/650X (an LDs titration commercial dye) were purchased from AAT Bioquest, Inc. (Pleasanton, California, USA). HeLa cells were cultured with 1 μM MitoLite^TM Blue FX490 and BODIPY 630/650X for 15 min, and then stained with MLR (5 μM). Subsequently, the cells were induced with Cu²⁺ + ES (1 μM) to dynamic observation of organelle colocalization changes in a time-dependent manner. Then, the images were captured by CLSM. The orange channel (500−600 nm) was obtained under 488 nm excitation, and the red channel (550−650 nm) was collected under 561 nm excitation.

2.7. Fluorescence imaging in zebrafish

Before imaging, zebrafish (3–7 dpf) were divided into three groups and transferred to confocal imaging dishes. One group was incubated with 10 μM MLR for 60 min before imaging. For the other two groups, zebrafish were pre-incubated with 100 and 200 μM NaHSO₃ or 200 and 400 μM Cys for 30 min, followed by 10 μM MLR for 30 min before imaging. Also, zebrafish seeded into 6-well culture plates were divided into six groups: (1) Blank, (2) control, (3) Cu²⁺, (4) ES, (5) Cu²⁺ + ES, and (6) Cu²⁺ + ES + TTM. In detail, zebrafish were cultured with the above group for 2 h. Subsequently, the cells were stained with MLR (10 μM) for 30 min to detect SO₂. Then, the images were captured by CLSM.

2.8. Statistical analysis

All experiments were performed in triplicate or more, with data expressed as mean ± standard deviation (SD). Origin (version 2021) were used to generate figures and to perform statistical analysis. All density functional theory (DFT) calculations were performed with Gaussian16, A03 software. Fluorescence imaging and intensity analysis were conducted using Olympus FV3000 (Hachioji, Tokyo, Japan) and ImageJ-win 64 software, Western blot (WB) band analysis was performed with Image Studio Lite, and flow cytometry data were analyzed using FlowJo.

3. Results and discussion

3.1. Synthesis and characterization of MLR

Probe MLR was synthesized according to the synthetic workflow depicted in Scheme S1. Coumarin derivatives were used as the design nucleus of fluorescent probes due to their modifiable structure and high quantum yield [22]. MLR, as a dual-organelle targeted fluorescent probe, was designed by combining the Mito-targeted benzopyran cation and the LDs-affinity phenylethynylcoumarin moiety in a donor-π-acceptor (D-π-A) framework. To meet the requirements of dual targeting (cationic Mito and lipophilicity of LDs), we strategically replaced a phenylethynylcoumarin unit (validated LDs targeting group) with a π-conjugated bridge and used its hydrophobic interaction to enrich LDs. This amphiphilic design ensured sequential localization: cationic benzopyran guiding initial Mito accumulation, while phenylethynylcoumarin-mediated lipophilicity promoting subsequent LDs distribution. The red emission (λ_em = 637 nm, Stokes shift = 112 nm) was realized by D-π-A conjugation, which effectively avoided the autofluorescence of cells. Meanwhile, the fluorescence tracking of SO₂ change was realized by Michael's addition of α, β-unsaturated ketones (Fig. 1A). By constructing the dual targeting ability of organelles and the fluorescence conversion of SO₂ response, MLR realized the simultaneous tracking of SO₂ changes between Mito and LDs, which played a key role in the development of organelle interaction analysis tools (Fig. 1B). The structure of MLR was confirmed using HRMS, ¹H NMR, and ¹³C NMR (Figs. S1–S9).

Fig. 1.

Schematic diagram to show the design of mitochondrial (Mito)-lipid droplets (LDs) dual-targeted fluorescent reporter (MLR) for realtime monitoring of sulfur dioxide (SO₂) in living cells and zebrafish. (A) The design and application of multifunctional MLR in response to SO₂ in cuproptosis. (B) The proposed reactive oxygen species (ROS)→SO₂ generation route by autoxidation of Cu²⁺ and elesclomol (ES) and a schematic representation of closer Mito-LDs contact in cells during cuproptosis.

Subsequently, the UV–Vis absorption response of MLR to SO₂ was investigated (Fig. 2A). As shown in Figs. 2B and S12, the probe exhibited a strong absorption band centered at 570 nm, which gradually decreased upon the addition of SO₂, while a new absorption band centered at 455 nm significantly increased. The fluorescence signal changes from red to orange fluorescence. This absorption shift, from 570 nm to 455 nm, can be attributed to the changes in the Förster Resonance Energy Transfer (FRET) effect induced by the Michael addition reaction with SO₂ [[23], [24], [25]]. The reaction kinetics between MLR and SO₂ were measured by monitoring fluorescence intensity at different time intervals. The reaction reached the steady-state concentration period within 30 s (Figs. 2C, 2D and S12). To determine the optimal detection range and limit of detection (LOD), MLR was titrated with SO₂ varying in concentrations. Upon the addition of SO₂, MLR exhibited a distinct fluorescence signal, which intensified with increasing SO₂ concentration (Figs. 2E and S13). The fluorescence intensity of MLR showed a linear response within the range of 1–100 μM SO₂ (R² = 0.996) (Fig. 2F), and the LOD was determined to be 0.34 μM. The fluorescence response of the probe MLR was calculated using rhodamine 6G (Φ_ref = 0.94) as the reference and characterized as follows: the molar extinction coefficient of ε₅₇₀ = 1.22 × 10⁴ M⁻¹cm⁻¹ and quantum yields of Φ_native = 0.38 → Φ_adduct = 0.29.

Fig. 2.

Optical spectra and emission mechanism. (A) Schematic illustration of the reaction of mitochondrial (Mito)-lipid droplets (LDs) dual-targeted fluorescent reporter (MLR) to sulfur dioxide (SO₂). (B) Ultraviolet–visible (UV–Vis) absorption of MLR before and after reduction by SO₂ in Hank's balanced salt solution (HBSS) buffer (1 M, pH 7.4). (C) Time-dependent changes in fluorescence emission intensity at 575 and 637 nm of MLR in response to SO₂ in HBSS buffer (1 M, pH 7.4) at 37 °C. λ_ex = 455 nm. (D) Time-dependent changes in fluorescence emission intensity over the 500−750 nm range of MLR in response to SO₂ in HBSS buffer (1 M, pH 7.4) at 37 °C. λ_ex = 455 nm. (E) Fluorescence intensity changes of MLR (10 μM) with SO₂ (0–100 μM). λ_ex = 455 nm. (F) Linear relationship for fluorescence intensity ratios (I₅₇₅ /I₆₃₇) of SO₂ concentrations (1.0–100μM). (G) Fluorescent intensity ratios (I₅₇₅/I₆₃₇) of the probe MLR in the absence or presence of SO₂ at various pH values. (H) Fluorescence intensity of MLR for long-term-stability study. (I) Fluorescence intensity ratios (I₅₇₅/I₆₃₇) of MLR (10 μM) upon addition of various analytes (1 mM). λ_ex = 455 nm. In (B-D, G), 10 μM MLR and 100 μM SO₂ were used. Error bars represent mean ± standard deviation (SD) (n = 3).

In the complex microenvironment of biological systems, such as in cells, a probe for detecting SO₂ is required for high specificity. In this study, the probe functioned well across the physiological pH range of 6.0–9.0 (Fig. 2G), and showed excellent stability (relative standard deviation ≤2%) (Fig. 2H), suggesting its great potential for detecting SO₂ in living organisms. The selectivity of MLR was evaluated via fluorescence spectral analysis of 100 μM SO₂ in the presence of various potential competing analytes, including different amino acids (Lys, Arg, Gly, Ile, Trp, Thr, Phe, Glu, and Val), biological thiols (GSH and Hcy), inorganic salts, and ROS. As shown in Figs. 2I and S14, the fluorescence intensity of MLR at 575 nm largely increased only in response to SO₂, demonstrating its high selectivity in SO₂ detection.

The molecular architecture of MLR was constructed through a nucleophilic substitution reaction between the amino group of compound 3 and the sulfonate ester of compound 6 (Fig. 3A), yielding a bifunctional probe with dual-organelle targeting capability. This design strategically integrated a Mito-directing moiety and an LDs-anchoring unit, interconnected by a conjugated π-bridge that spatially links the donor and acceptor fluorophores to establish a FRET-based sensing framework (Fig. 3B). In its native state (apo-MLR), the extended D-A separation (>10 nm) suppresses energy transfer, manifesting solely as blue donor emission (Fig. 3C). The SO₂-responsive mechanism involves nucleophilic attack at the electron-deficient C=N bond, triggering molecular orbital reconfiguration (Fig. 3D). This reaction induces charge redistribution that contracts the D-A distance to <10 nm, thereby activating FRET-mediated orange acceptor fluorescence (Fig. 3C). Molecular electrostatic potential mapping (Fig. 3D) corroborates this mechanism, revealing enhanced electrophilic character at the reaction center post-SO₂ binding. Electronic structure analysis (Fig. 3E) quantifies the thermodynamic driving force, demonstrating a 1.54 eV reduction in the Highest Occupied Molecular Orbital-Lowest Unoccupied Molecular Orbital (HOMO-LUMO) gap (HOMO: –5.25→–5.17 eV; LUMO: –2.84→–1.22 eV) that facilitates structural reorganization. In addition, by comparing the response performance of the current fluorescent probes targeting Mito and LDs to SO₂ [8,[26], [27], [28], [29]], it can be seen that MLR designed by us realizes the sensitivity monitoring of SO₂ nanomolar level (<10 s) and fewer synthesis steps (Fig. 3F and Table S1). At the same time, to evaluate the radiation imaging potential of MLR, Red/Green/Blue (RGB) analysis based on the smartphone was carried out (Figs. S14–S18) [[30], [31], [32], [33]]. As shown in Fig. 3G, there was a linear relationship between the R/G signal ratio and SO₂ concentration. The fluorescence color of MLR treated with different SO₂ concentrations can be identified in the International Commission on Illumination (CIE) chromaticity diagram. These coordinated structural, electronic, and optical responses establish MLR as a high-sensitivity probe for real-time SO₂ dynamics monitoring in subcellular.

Fig. 3.

Molecular design, response mechanism, and performance verification of mitochondrial (Mito)-lipid droplets (LDs) dual-targeted fluorescent reporter (MLR). (A) The synthetic route of probe MLR. (B) The förster resonance energy transfer (FRET) switch mechanism enabling Mito to LDs targeting transition upon sulfur dioxide (SO₂) binding. (C) The highest occupied molecular orbital-lowest unoccupied molecular orbital (HOMO-LUMO) energy gap variation (ΔE = 2.41 → 1.87 eV) between donor and acceptor units before/after SO₂ interaction. (D) Comparative electrostatic potential (ESP) analysis (−0.15–0.15 a.u.) showing charge redistribution upon SO₂ conjugation. (E) Frontier orbital shifts (ΔE = 2.41 → 3.95 eV) validating reduced electron transfer driving force post-SO₂ binding. (F) The performance of MLR was compared with some representative SO₂ detection targeted fluorescent probes. (G) Photograph of the International Commission on Illumination (CIE) chromaticity diagram of MLR after SO₂ treatment.

3.2. Cellular evaluation of MLR in SO₂ imaging

Following the characterization of MLR properties, we systematically evaluated its capability for SO₂ detection in live HeLa cells (Fig. 4A). Cytotoxicity assessment revealed that both HeLa and MCF-10A cells maintained >85% viability at MLR concentrations up to 25 μM (Fig. S19), indicating minimal cytotoxicity and suitability for long-term imaging. Subsequent cell imaging under physiological conditions (Figs. S20 and S21) confirmed the stability of the probe. Also, we used three different cells to assess exogenous SO₂ detection, the cells in the control group incubated with 10 μM MLR showed obvious red fluorescence, while the orange fluorescence signal was negligible (Fig. 4B(i)). The cells were pretreated with different concentrations of SO₂ (50, 200 μM). As shown in Figs. 4B(ii) and (iii), after SO₂ treatment, red fluorescence gradually decreased, while orange fluorescence significantly increased. Studies have demonstrated that sulfur-containing amino acids (e.g., L-cysteine) [34] and sulfonamides in mammalian systems can generate endogenous SO₂ through metabolic processes. To verify the probe's capability in detecting endogenous SO₂, we conducted cellular experiments using Cys as an endogenous SO₂ inducer in HeLa cells. In the control experiment, cells incubated with MLR alone showed red fluorescence (Fig. 4B(i)), while cells in the experimental group were induced by 200, 400 μM Cys to produce endogenous SO₂ (Figs. 4B(iv) and (v)). After reacting with MLR, the intracellular orange fluorescence gradually increased and the red fluorescence decreased. Quantitative analysis of fluorescence intensity confirmed MLR's capacity for detecting intracellular SO₂. Significantly, we observed that the fluorescence intensity of the orange channel gradually increased with the decrease of the fluorescence intensity of the red channel. This indicates that MLR can visualize SO₂ produced in cells through ratiometric fluorescence signals.

Fig. 4.

Fluorescence imaging analysis of mitochondrial (Mito)-lipid droplets (LDs) dual-targeted fluorescent reporter (MLR)-based sulfur dioxide (SO₂) detection during cuproptosis. (A) Experimental workflow for cellular induction. (B) Exo/endogenous SO₂-dependent fluorescence imaging and intensity quantification in HeLa cells: the MLR-treated control group showed intense red fluorescence and faint orange signal (i), exogenous SO₂ (50 and 200 μM) weakened red fluorescence and elevated orange signal progressively (ii, iii), and L-Cys-triggered endogenous SO₂ induced analogous spectral changes in HeLa cells (iv, v). (C) Confocal microscopy images of SO₂ produced by HeLa cells after incubation with 1 μM Cu²⁺ + elesclomol (ES) at 0, 30, 60 and 90 min. (D) The fluorescence intensity of SO₂ produced by HeLa cells after incubation with 1 μM Cu²⁺ + ES at 0, 30, 60 and 90 min. (E) The SO₂ produced during the incubation of HeLa cells with Cu²⁺ + ES was compared by confocal microscopy images, and MLR was co-localized with commercial organelle localization probes methylthymol blue complexone (MTB) (i, ii) and 4,4-difluoro-boradiazaindacene (BODIPY) 630/650X (iii, iv). (F) Corresponding to the change of co-localization coefficient of Mito and LDs in (E) and the quantification of fluorescence channel data. (Red channel: λ_ex = 570 nm, λ_em = 600−700 nm; Orange channel: λ_ex = 455 nm, λ_em = 500–600 nm). Error bars represent mean ± standard deviation (SD) (n = 3).

3.3. Ratiometric imaging SO₂-driven Mito-LDs interplay during cuproptosis

We further employed probe MLR to dynamically monitor the change of SO₂ content during cuproptosis, simultaneously tracking Mito-LDs contact alterations. As shown in Fig. 4C, HeLa cells preloaded with MLR (10 μM) initially exhibited predominant red channel fluorescence with negligible orange signal. Upon treatment with the cuproptosis inducer ES, a time-dependent fluorescence shift occurred. Orange channel intensity progressively increased within 30 min, reflecting the increase of intracellular SO₂ levels. After incubation with ES for 60 min, Cu²⁺ + ES further triggered a surge in intracellular ROS. At the same time, the intracellular SO₂ level decreased, resulting in a gradual decrease in the fluorescence of red and orange channels. As the cells were incubated for 90 min, their state gradually became blurred, eventually leading to apoptosis. At the same time, apoptosis led to a decrease in ROS levels, and red and orange fluorescence almost disappeared. Quantitative fluorescence analysis confirmed MLR's capability for monitoring SO₂ content fluctuations during cuproptosis (Fig. 4D).

To further validate MLR's dual-targeting capacity for Mito and LDs, we first used different organelle-targeted commercial probes to co-locate with MLR to ensure that the probe was initially targeted to Mito. These dyes targeted lysosomes (Lyso-Tracker Green, LTG) and the endoplasmic reticulum (ER-Tracker Blue, ETB). The results showed low Pearson Correlation Coefficient (PCC) values of 0.15 (LTG), 0.19 (ETB) (Fig. S22). Also, to verify the differential imaging effect of the probe on cell types, we selected three different types of cell lines (such as normal cell RAW, cancer cell HepG2, etc.) for co-localization analysis. The results showed that the co-localization coefficient of the probe for different cells was >0.90 (Fig. S23). Then we performed co-localization studies in Cu²⁺ + ES treated cells using commercial fluorescent probes MitoLite™ Blue FX490 (MTB, 1 μM, targeting Mito) and BODIPY 630/650X (1 μM, targeting LDs) markers. In Cu²⁺ + ES treated cells, the initial PCC for the overlap between the red channel (MLR) and blue channel (MTB) is 0.92, while between the orange channel (MLR-SO₂) and the deep red channel (BODIPY) was 0.78 (Figs. 4E(i), 4E(iii), and 4F). This indicates that the MLR is first localized within the Mito. With the increased induction time, intracellular SO₂ levels increased. At this time, the overlap coefficient of the red channel and the blue channel decreases to 0.76, while the overlap coefficient of the orange channel and the dark red channel rises to 0.96 (Fig. S24). Corresponding to MLR's reaction with accumulating SO₂ to form lipophilic neutral adducts that subsequently localized to LDs. With the deepening of the induced cuproptosis, ROS overproduction triggered SO₂ depletion, and the correlation coefficient between the red channel and the blue channel decreased to 0.15, and the correlation coefficient between the orange channel and the dark red channel also decreased to 0.29, and concomitant fluorescence decay in both channels (Figs. 4E(ii), 4E(iv), and 4F) corroborated this dynamic process. This PCC trajectory and fluorescence change demonstrates enhanced Mito-LDs contact during early cuproptosis, followed by organelle disassociation in late apoptosis, as evidenced by MLR's dual-channel ratiometric response.Cuproptosis is a copper-dependent cell death pathway initiated by ferredoxin-1 (FDX1)-mediated reduction of ES-transported Cu²⁺ to redox-active Cu⁺, a process generating ROS [[35], [36], [37]]. During this cascade, FDX1 simultaneously catalyzes dihydrolipoamide S-acetyltransferase (DLAT) lipoylation via lipoid acid synthase (LIAS) interaction, enabling Cu⁺ binding to sulfonated DLAT Cys residues that drive oligomerization and iron-sulfur (Fe-S) cluster depletion [38]. The resultant proteotoxic stress from Fe-S protein loss and DLAT aggregation ultimately executes cell death (Fig. 5A). Then we passed intracellular ROS levels were quantified via 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) fluorescence, revealing significantly enhanced signal in Cu²⁺ + ES-treated cells versus the other experimental groups (Figs. 5B and S25), indicative of ES-mediated cuproptosis effect potentiation. Meanwhile, the cytotoxicity of Cu²⁺-ES-induced cuproptosis on cell viability was assessed using calcein-acetyloxymethyl ester (calcein-AM)/propidium iodide (PI). Fluorescence images showed that the cell viability of the Cu²⁺ + ES group was significantly lower than other cell treatment groups (Figs. 5C and S25). In addition, immunoblotting demonstrated a marked reduction of DLAT monomers and LIAS protein levels specifically in Cu²⁺ + ES groups, and slightly elevated FDX1 protein levels, while DLAT showed fluorescence oligomerization phenomenon (Figs. 5D and S26). Consistent with reported levels of cuproptosis biomarkers. We also used flow cytometry to further verify that Cu²⁺ + ES induced higher ROS production and 42.8% apoptosis rate (11.2 times higher than ES alone (3.82%)) during cuproptosis (Figs. 5E, F, and G). Abnormal changes of Mito membrane potential and irregular distribution and aggregation of intracellular LDs occurred in cuproptosis (Fig. S27), mechanistically substantiates Cu²⁺ + ES induced cuproptosis through Fe-S cluster depletion and proteotoxic cascade.

Fig. 5.

In vitro evaluation of elesclomol (ES) efficacy in cuproptosis. (A) Mechanistic schema of Cu²⁺ + ES-induced cuproptosis: dihydrolipoamide S-acetyltransferase (DLAT), and lipoid acid synthase (LIAS) proteins. (B) The distribution of reactive oxygen species (ROS) and its corresponding fluorescence intensity in different treatment groups of 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) stained cells were measured by a confocal microscopy. (C) The viability of different cell treatment groups was evaluated by calcein-acetyloxymethyl ester (calcein-AM) (green, alive)/propidium iodide (PI) (red, dead) co-staining. (D) Western blot (WB) of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), DLAT, ferredoxin-1 (FDX1), and LIAS proteins. (E) Flow cytometry quantification of ROS levels in different cell treatment groups. (F) Confocal laser scanning microscope (CLSM) time images of mitochondrial (Mito) membrane potential. (G) Apoptosis/necrosis ratios under different conditions by Annexin V-FITC/PI assay. In different treatment groups, Cu²⁺, ES, Cu²⁺ + ES, Cu²⁺ + ES + tetrathiomolybdate (TTM) (a copper chelator) were used at 1 μM. Error bars represent mean ± standard deviation (SD) (n = 3).

3.4. Ratiometric SO₂ tracking reveals cuproptosis stress in zebrafish

We next assessed MLR's biosensing capability in zebrafish, leveraging their optical transparency and tractability for in vivo imaging (Fig. 6A) [39,40]. Following 1 h with 10 μM MLR is co-incubated with zebrafish, confocal imaging revealed pronounced red fluorescence confirming effective probe ingestion (Fig. 6B(i). Notably, zebrafish pre-exposed to 50 μM exogenous SO₂ (60 min) followed by MLR treatment 1h exhibited emerging orange fluorescence inversely correlated with red signal attenuation (Figs. 6B(ii) and (iii)). This ratiometric fluorescence response demonstrated SO₂-dependent MLR activation. Furthermore, L-Cys-induced endogenous SO₂ generation elicited concentration-dependent orange fluorescence enhancement concurrent with red channel diminution (Figs. 6B(iv) and (v)), validated by quantitative intensity analysis (Fig. 6C). These results conclusively demonstrate MLR's capacity for dynamic, dual-channel SO₂ detection in zebrafish.

Fig. 6.

In vivo zebrafish fluorescence imaging. (A) Mechanistic schema of mitochondrial (Mito)-lipid droplets (LDs) dual-targeted fluorescent reporter (MLR)-based in vivo sulfur dioxide (SO₂) detection. (B) MLR-labeled zebrafish showed red fluorescence, and exogenous/Cys-induced endogenous SO₂ shifted signals to orange, proving its ratiometric SO₂ response in vivo: Zebrafish incubated with 10 μM MLR for 1 h showed strong red fluorescence, confirming successful probe uptake (i); pretreatment with exogenous SO₂ triggered orange fluorescence and attenuated red signals, verifying SO₂-responsive ratiometric activation of MLR (ii, iii); Cys stimulated endogenous SO₂ induced dose-dependent orange fluorescence elevation alongside reduced red signals (iv, v). (C) Exo/endogenous SO₂-dependent fluorescence imaging intensity quantification in zebrafish. (D) Confocal imaging of 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) stained zebrafish after incubation with Cu²⁺ + elesclomol (ES) and the fluorescence quantitative analysis of the produced reactive oxygen species (ROS). (E) The fluorescence quantitative analysis of SO₂ in zebrafish incubated with different treatment groups (ES, Cu²⁺, Cu²⁺ + ES, Cu²⁺ + ES + tetrathiomolybdate (TTM)) by MLR staining. Red channel: λ_ex = 570 nm, λ_em = 600–700 nm; Orange channel: λ_ex = 455 nm, λ_em = 500–600 nm. Error bars represent mean ± standard deviation (SD) (n = 3).

Prior studies have established that copper ions induce structural damage to zebrafish gill tissues and compromise antioxidant systems, triggering immune-inflammatory cascades [41,42]. To investigate oxidative stress dynamics mediated by cuproptosis associated ROS and reactive sulfur species (RSS), zebrafish were pretreated with cuproptosis modulators (inducers/inhibitors) for 30 min before incubation MLR (1 h). DCFH-DA-stained ROS tracking revealed time-dependent fluorescence intensification in Cu²⁺ + ES-treated zebrafish, peaking at 30 min post-treatment before signal decay. Concurrent morphological analysis documented zebrafish developmental abnormalities: yolk sac enlargement, cardiac chamber dilation, and multi-organ edema. These pathologies culminated in organismal death, correlating with fluorescence-quantified cuproptosis progression (Fig. 6D). As shown in Fig. 6E, orange fluorescence was observed in the Cu²⁺ + ES group, while Cu²⁺ + ES + TTM co-treatment restored red fluorescence. This ratiometric response indicates that cuproptosis-induced oxidative stress activates endogenous ROS production, with subsequent SO₂ generation serving as a transient protective mechanism. Notably, the Cu²⁺ + ES inducer demonstrated dual toxicity: (1) direct tissue damage through antioxidant system disruption, and (2) inflammatory mediator induction via SO₂ overproduction. The MLR probe successfully tracked SO₂ change during these oxidative stress cascades, demonstrating its utility for dynamic monitoring of sulfur species in live vertebrate models.

4. Conclusions

In summary, we constructed a Mito-LDs dual-targeted fluorescent probe-MLR for real-time monitoring of the dynamic changes of SO₂ and the interaction between organelles during cuproptosis in living cells and zebrafish. The probe exhibits FRET-mediated red fluorescence (λ_em = 637 nm) under 575 nm excitation through the piperazine bridge structure of benzopyranium salt and coumarin derivative. After the double bond cleavage of the benzopyranium salt triggered by SO₂, the probe was converted into a lipophilic neutral structure, resulting in the release of the FRET effect. The characteristic performance was red fluorescence quenching accompanied by orange fluorescence enhancement (λ_em = 575 nm), and the fluorescence intensity ratio (I₅₇₅/I₆₃₇) changed by 23.8 times. Compared with the existing organelle-targeting probes [43,44], MLR has the following advantages: (i) dual-organelle targeting for simultaneously tracking the conversion process from Mito to LDs, (ii) dynamic visualization of organelle interactions in living animals; (iii) one-step synthesis and rapid response (10 s). We envision future studies integrating the MLR architecture with multiplexed imaging systems through multi-organelle co-localization strategies to enable simultaneous detection of cuproptosis biomarkers [[45], [46], [47], [48], [49]].

CRediT authorship contribution statement

Furao Li: Writing – original draft, Validation, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Chunyan Liang: Software, Methodology, Investigation, Formal analysis, Data curation. Xifeng Mo: Resources, Methodology, Investigation, Formal analysis. Xiaohuan Xu: Methodology, Formal analysis, Data curation. Yongbiao Wei: Visualization, Software, Resources. Chunyan Zhou: Visualization, Software, Resources. Ting Meng: Validation, Resources, Formal analysis. Hui Zhang: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Fan Yang: Writing – review & editing, Writing – original draft, Supervision, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: