An ESIPT-AIE nanosensor for ONOO⁻ imaging: decoding heavy metal stress and ferroptosis in living plants

Abstract

Peroxynitrite (ONOO⁻) serves as a critical redox signaling molecule in plant stress responses and ferroptosis, yet real time monitoring within complex plant matrices remains challenging. To address this, we synthesized a ratiometric nanosensor (DA) through molecular self-assembly. The DA particles (258 nm in diameter) exhibited enhanced sensitivity driven by an excited-state intramolecular proton transfer (ESIPT)-triggered restricted intramolecular motion mechanism, resulting in distinctive aggregation-induced emission (AIE) behavior. This nanoscale configuration improved tissue penetration and eliminated the aggregation-caused quenching (ACQ) effect commonly observed in traditional rhodamine derivatives. Meanwhile, the DA probe featured a large Stokes shift of 167 nm and an ultra-low limit of detection (LOD) of 6.4 nM. Leveraging these optical advantages, the nanosensor enabled real-time visualization of ONOO⁻ dynamics in plant tissues and quantitative assessment of ONOO⁻ accumulation under cadmium (Cd²⁺), sodium chloride (NaCl), and erastin induced stress. This work represents the first application of an ESIPT-AIE hybrid probe for ONOO⁻ detection in plants, providing a powerful analytical platform for elucidating oxidative stress mechanisms and advancing strategies to enhance crop resilience.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04516-w.

Introduction

Reactive oxygen species (ROS) function as vital redox regulators and signaling mediators in biological systems, governing diverse physiological processes such as cellular proliferation, stress adaptation, and immune homeostasis [1]. However, excessive ROS accumulation under stress conditions disturbs redox balance, inducing oxidative injury. Among ROS derivatives, peroxynitrite (ONOO⁻)—a potent oxidant generated through the rapid reaction between superoxide and nitric oxide (NO)—has emerged as a central mediator in plant stress physiology [2, 3]. Environmental stressors including heavy metals, salinity-alkalinity, and temperature extremes provoke pathological ONOO⁻ buildup, initiating cascades of protein nitration, lipid peroxidation, and nucleic acid damage that culminate in cell death and compromised crop productivity [4, 5]. Recent findings further implicate ONOO⁻ in regulating ferroptosis, an iron-dependent form of programmed cell death characterized by glutathione depletion, redox disequilibrium, and membrane disintegration [6, 7]. Despite these insights, the mechanistic role of ONOO⁻ in plant stress signaling and ferroptosis remains poorly defined [8, 9], largely due to the limited sensitivity of current detection techniques (with reported ONOO⁻ probe detection limits mostly concentrated in the 50 nM–5 μM range in complex biological matrices [10]) and the inherent complexity of plant tissues. This gap underscores an urgent demand for advanced ratiometric ONOO⁻ probes with superior specificity and quantitative precision to elucidate its roles in plant stress physiology and programmed cell death pathways.

Fluorescent probe technology has emerged as a powerful approach for ONOO⁻ detection, owing to its exceptional sensitivity, real-time imaging capability, and non-invasive nature [11–15]. While widely applied in biomedical contexts including inflammation, cancer, and neurodegenerative disease research to visualize oxidative stress mechanisms [16], ONOO⁻ is increasingly recognized as a pivotal regulator of plant stress adaptation and ferroptosis. Nevertheless, its detection in plants remains challenging due to its extreme reactivity and nanomolar physiological concentrations [17], which constrain the performance of plant-compatible fluorescent probes [18–20]. The pioneering study by Setsukinai et al. (2003) utilized aminophenylfluorescein (APF) for non-specific ONOO⁻ imaging in tobacco (Nicotiana tabacum) BY-2 cells [21]. More recently, Lu et al. (2024) developed the RAW-A probe, enabling endogenous ONOO⁻ visualization in Arabidopsis thaliana [22], while Cao et al. (2025) introduced the NFP probe to monitor ONOO⁻ dynamics during seed germination under abiotic stress [23]. Despite these advances, current probes exhibit critical shortcomings: most rely on single-emission intensity readouts, making them susceptible to artifacts caused by probe concentration variations, uneven distribution, and environmental fluctuations. Furthermore, common issues such as aggregation-caused quenching (ACQ), small Stokes shifts, and limited sensitivity hinder accurate quantification and real time tracking of rapid ONOO⁻ fluctuations in living plant systems [21–25].

To overcome these limitations, we focused on aggregation-induced emission (AIE) fluorophores, first conceptualized by Tang et al., which exhibit distinct advantages over conventional dyes [26]. AIE systems demonstrate enhanced fluorescence intensity, superior photostability, intrinsic resistance to ACQ, and through nanoscale engineering improved tissue penetration and biocompatibility [27]. Particularly promising are AIE luminogens incorporating excited-state intramolecular proton transfer (ESIPT), which feature large Stokes shifts (> 150 nm). ESIPT arises from intramolecular hydrogen bonding between proton donor and acceptor groups, proceeding through a four-step phototautomerization mechanism that yields emissions with minimal self-absorption and reduced spectral overlap, thereby mitigating signal interference [28, 29]. These features render ESIPT-based probes especially advantageous for high-fidelity plant imaging [30, 31]. Importantly, combining ESIPT and AIE mechanisms represents an emerging paradigm in functional fluorophore design. These hybrid systems offer synergistic benefits: ESIPT-mediated O–H⋯N hydrogen bonding rigidifies molecular conformations and amplifies AIE via restricted intramolecular motion (RIM) [32–34], while AIE activation suppresses non-radiative decay and stabilizes nanoaggregate formation. Despite their potential, ESIPT-AIE-based probes have not yet been exploited for ONOO⁻ detection in plant systems. Harnessing the cooperative effects of these mechanisms offers a compelling route toward next-generation fluorescent probes with unparalleled sensitivity, stability, and imaging performance.

In this study, we harnessed the ESIPT-triggered RIM process to activate the AIE mechanism, enabling the design and synthesis of three novel fluorescent probes: DA, DA-1, and DA-2. These probes integrate a hydroxy benzothiazole (HBT)-modified rhodamine fluorophore that concurrently exhibits ESIPT and AIE properties, producing a synergistic enhancement of green fluorescence. Upon reaction with ONOO⁻, the DA probe undergoes selective amide bond cleavage, resulting in spiro-ring opening and the release of a strong red fluorescence signal. This probe offers several distinctive advantages for ONOO⁻ detection in plants: an exceptional ratiometric fluorescence response with reduced spectral overlap, high selectivity against interfering species, and excellent biocompatibility for in vivo imaging. These attributes enabled the visualization of ONOO⁻ dynamics in various plant models, including onion epidermal cells and tobacco (Nicotiana tabacum) root tips. Moreover, DA successfully detected ONOO⁻ in situ under multiple stress conditions, including heavy metal exposure, saline-alkaline stress, and erastin-induced ferroptosis (Scheme 1), establishing DA as a potent tool for monitoring plant stress responses, advancing crop health diagnostics, and supporting food security initiatives.

Scheme 1.

Response mechanism, morphology of the fluorescence probe DA, the process of in vitro detection of ONOO⁻, and the schematic scheme for visualizing the distribution of ONOO⁻ in plant-related physiological processes. (SLC7A11: solute carrier family 7 member 11; GPX4: glutathione peroxidase 4; GSH: glutathione; NOS: nitric oxide synthase; PUFA: polyunsaturated fatty acid; Fe.²⁺: ferrous ion; ∙OH: hydroxyl radical; and H₂O₂: hydrogen peroxide)

Experimental section

Synthesis and characterization of DA

Synthesis of DA: as shown in Scheme S2, DA-1 (300 mg, 0.561 mmol) was dissolved in 10 mL dichloromethane, and 52 µL dimethylcarbamic chloride (0.566 mmol) was added. The reaction was stirred at room temperature for 6 h while monitoring progress by TLC. The mixture was vacuum-dried and purified by thin-layer chromatography using dichloromethane-methanol (100: 1) to yield a white powdered solid named DA (48.6 mg, yield: 14.30%, m.p. 248.3–249.2 ℃). ¹H NMR (400 MHz, DMSO-d₆), δ 11.83 (s), 8.47 (s), 8.06 (d, J = 8.0 Hz), 7.93–7.90 (m), 7.66 (s), 7.63–7.59 (m), 7.47–7.43 (m), 7.36 (t, J = 8.0 Hz), 7.14–7.11 (m), 6.83 (d, J = 1.0 Hz), 6.51 (d, J = 8.5 Hz), 6.37–6.34 (m), 3.33 (q, J = 7.0 Hz), 2.61 (s), 1.08 (t, J = 7.0 Hz). ¹³C NMR (101 MHz, DMSO-d₆) δ 165.13, 164.48, 157.27, 156.84, 155.15, 153.04, 151.76, 151.73, 148.85, 134.69, 133.93, 130.18, 129.87, 129.37, 126.80, 125.25, 124.61, 123.31, 122.40, 115.64, 111.93, 108.46, 103.20, 97.33, 65.23, 44.13, 36.53, 26.87, 13.00. HRMS: m/z [M + H]⁺ calcd for C₃₄H₃₂N₅O₄S⁺: 606.2170, found: 606.2141.

DFT calculations

All quantum-chemical calculations were performed using the Gaussian 16 software package. Full optimization of the ground-state molecular geometry was performed at the B3LYP/6-31G(d) level of theory within an aqueous solvation model based on density functional theory (DFT) [16].

Fluorescence and UV–Vis spectra measurements

All spectral analyses—including fluorescence emission, titration, competition, and UV–Vis spectra—were acquired using PTI QuantaMaster 8000 and TU-1900 spectrophotometers. Fluorescence measurements were performed at room temperature with an excitation wavelength of 365 nm, and emission spectra were collected from 450 to 700 nm, slits were set to 2 nm. All fluorescence measurements were conducted in MeCN-H₂O (f_w = 90%) mixed solvent. For selectivity and competition assays, DA (1 mM) was incubated with varying concentrations of ONOO⁻ and potential interfering species under identical conditions [18]. All experiments were repeated three times, and data are presented as mean ± standard deviation (SD). Error bars represent the SD of three independent measurements.

Toxicity test of the probe DA

Uniformly sized mung bean (Vigna radiata) and tobacco (Nicotiana tabacum) seeds were soaked in ultrapure water for 6 h and then transferred to Petri dishes containing DA probe solutions at gradient concentrations (0, 10, 20, 40, 60, 80, and 100 μM). The dishes were maintained at 25 ℃ under a 16 h light/8 h dark photoperiod and 90% relative humidity. Germination progress was observed and recorded throughout the experiment [35].

Imaging exogenous ONOO⁻ in onion epidermal cells

For in situ imaging of exogenous ONOO⁻, onion (Allium cepa) epidermal cells were incubated with DA (10 μM), followed by exposure to ONOO⁻ solutions at different concentrations (0, 10, 25, 50, and 70 μM). To evaluate time-dependent fluorescence response, cells pretreated with DA (10 μM) were incubated with 70 μM ONOO⁻ for 0, 15, 30, 45, and 60 min. Imaging was then performed using confocal microscopy [36].

Imaging exogenous ONOO⁻ in plant tissues

Two-week-old tobacco (Nicotiana tabacum) seedlings were rinsed thoroughly with deionized water to remove surface residues, immersed in DA (10 μM), and subsequently treated with ONOO⁻ solutions of varying concentrations (0, 10, 25, 50, and 70 μM). After triple rinsing with deionized water, the seedlings were imaged using confocal microscopy [30].

Imaging endogenous ONOO⁻ in plant tissues

For cadmium ion (Cd²⁺) stress analysis, 1-week-old tomato (Lycopersicon esculentum) seedlings were irrigated with Cd²⁺ solutions at different concentrations (0, 25, 50, 70, and 100 μM) for 7 days. After cultivation, the seedlings were thoroughly rinsed three times with distilled water, incubated in DA (10 μM), and washed again before imaging the root tips with confocal microscopy [37].

Salt stress experiment

Tomato (Lycopersicon esculentum) seedlings grown for 1 week were irrigated with sodium chloride (NaCl) solutions of varying concentrations (0, 25, 50, 70, and 100 μM) for 7 days. Following cultivation of the tomato seedlings (Lycopersicon esculentum), they were rinsed thoroughly three times with distilled water to remove any surface residues. They were then immersed in a DA probe solution (10 μM) and incubated. After incubation, the seedlings were washed three times with distilled water, and the root tips were mounted on slides for imaging using confocal microscopy [23].

Erastin-induced ferroptosis experiment

Tobacco (Nicotiana tabacum) seedlings grown for 2 weeks were cultured in Hoagland’s solution containing erastin at concentrations of 0, 10, 20, 40, 60, and 80 μM for 1 day. After treatment, the tobacco (Nicotiana tabacum) seedlings were rinsed three times with distilled water to remove surface residues, then immersed in a DA probe solution (10 μM) and incubated. Following incubation, the seedlings were washed three times with distilled water, and root tips were prepared on slides for confocal microscopy imaging [28].

Results and discussion

Exploration of the design and properties of probes

Three novel rhodamine-based fluorescent probes (DA, DA-1, and DA-2) were thus constructed, each incorporating a tailored amide-based recognition site for ONOO⁻ detection [38]. ONOO⁻ mediated cleavage induces spiro-ring opening, resulting in a pronounced fluorescence change. The probes retain the intrinsic photophysical robustness of rhodamine derivatives while displaying enhanced ESIPT-AIE synergy due to HBT integration. The rhodamine fluorophore DD was synthesized following previously reported methods [39]. Synthetic routes for DA, DA-1, and DA-2 are depicted in Scheme S1–S3, and their structures were verified by NMR and HRMS analyses (Figure S1–S10, Table S1).

The aggregation behavior of DA, DA-1, and DA-2 was examined in MeCN-H₂O mixtures (Fig. 1). In pure acetonitrile, all probes exhibited weak fluorescence; however, as the water fraction increased (f_w = 0–99%), a marked fluorescence enhancement at 516 nm was observed (Fig. 1B–1G), confirming AIE activation in aqueous environments. Solid-state DA emitted intense green fluorescence under UV excitation (Fig. 1H), further verifying its dual ESIPT-AIE character. Scanning electron microscope (SEM) analysis revealed nano-spherical aggregates of DA with an average diameter of 258 nm (Fig. 1I, 1J). This aggregated morphology strengthens intermolecular hydrogen bonding, corroborating the AIE mechanism. To further confirm the ESIPT process, time-resolved fluorescence lifetime measurements were conducted on DA in the AIE-active state, which revealed a tri-exponential decay with lifetimes of τ₁ = 0.53 ns, τ₂ = 4.04 ns, and τ₃ = 15.78 ns (average lifetime 6.61 ns, Figure S11). The sub-nanosecond component (τ₁ = 0.53 ns) is the definitive spectroscopic signature of short-lived enol-form emission, while the longer-lived components correspond to keto-form emissions from different microenvironments within the nano-aggregates—a kinetic fingerprint of a functioning ESIPT process [40, 41]. Solvent polarity-dependent emission studies (Figure S12) demonstrated dual fluorescence bands corresponding to the enol (high-energy) and keto (low-energy) forms, hallmarks of the ESIPT process [42, 43]. In MeCN-H₂O (f_w = 90%), the HBT unit formed an intramolecular O–H⋯N hydrogen bond, restricting benzothiazole ring rotation, suppressing non-radiative decay, and promoting nanoparticle formation with amplified green emission. Among the three probes, DA exhibited the strongest ONOO⁻ responsive fluorescence enhancement in the AIE-active solvent system (Figure S13) and was therefore selected for further investigation.

Fig. 1.

AIE

Characterization of DA

The mechanism and morphology of dopamine-based self-assembled nanostructures (DA) were systematically investigated using molecular dynamics (MD) simulations, Fourier-transform infrared (FT-IR) spectroscopy, transmission electron microscopy (TEM), and SEM. MD simulations were performed using GROMACS, employing gmx hbond for hydrogen bonding analysis, gmx gyrate for radius of gyration, and gmx rms for root mean square deviation (RMSD). Structural visualization and trajectory analysis were conducted in VMD 1.9.3 and PyMOL 3.0 [44, 45], with representative conformations extracted at 20 ns intervals (Fig. 2B). The simulations revealed the initiation of nanocluster formation at 20 ns, achieving structural stabilization by 40 ns (Figure S14A). The equilibrium structure obtained at 100 ns (Fig. 2C) displayed well-organized water molecules surrounding DA aggregates. Quantitative analysis yielded a mean RMSD of 3.092 ± 0.311 nm, confirming system stability beyond 40 ns. The solvent-accessible surface area (SASA) was analyzed to assess molecular exposure to the solvent. The average SASA value of 80.759 ± 18.020 nm² exhibited a marked decline during the first 40 ns, reflecting reduced solvent exposure as nanoclusters condensed. Beyond 40 ns, SASA values plateaued, indicating the formation of compact, stable aggregates with minimal solvent accessibility (Figure S14B). To identify the driving forces underpinning DA self-assembly, intermolecular interactions were examined in detail. Hydrogen bonding and π–π stacking were found to dominate the process, with Coul-SR and LJ-SR parameters quantifying hydrogen bonding and π–π stacking strengths, respectively. Energy analysis revealed mean interaction energies of –2481.128 ± 61.131 kJ/mol (Coul-SR) and –2879.801 ± 342.076 kJ/mol (LJ-SR; Figure S14C), establishing π–π stacking as the predominant stabilizing force. Structural analysis identified an average of 12 ± 2.030 hydrogen bonds (Figure S14D), confirming that hydrogen bonding cooperatively contributes to nanocluster stabilization. As illustrated in Fig. 2D, extensive hydrogen bonding networks form between the hydroxyl and carbonyl groups of DA molecules (Fig. 1-2D), while conjugated aromatic rings promote π–π stacking (Fig. 2D-2). Collectively, these findings demonstrate that DA nanoparticle formation arises through synergistic electrostatic (hydrogen bonding) and hydrophobic (π–π stacking) interactions (Fig. 2A). FT-IR spectroscopic analysis furnished critical experimental validation for the self-assembly mechanism deduced from MD simulations (Figure S15). Specifically, the marked attenuation of the carbonyl (C = O) stretching vibration peak at 1670 cm⁻¹ directly verifies that carbonyl groups act as hydrogen-bond acceptors during intermolecular interactions. Moreover, the broadening of the absorption band at 3250 cm⁻¹ (attributed to O–H/N–H stretching), the wavenumber shift of the peak at 3050 cm⁻¹ (corresponding to aromatic C–H stretching), and the characteristic spectral variations in the fingerprint region (700–850 cm⁻¹, arising from aromatic C–H out-of-plane bending vibrations) collectively substantiate, from an experimental standpoint, the formation of spherical nanostructures with highly ordered aromatic ring alignment. This structural organization is primarily driven by synergistic hydrogen bonding and π–π stacking interactions. These findings were further corroborated by TEM and SEM characterizations, which visually evidenced the distinct spherical morphology of the as-synthesized DA nanoparticles (Figure S16 and 1I).

Fig. 2.

MD simulation of DA. (A) Schematic diagram of the self-aggregation process of DA. (B) Structural changes in the DA system per 20 ns during the simulation. (C) Structural changes in the DA system in the initial and end states. (D) Molecular interaction of the DA system

The responsiveness of the probe DA to ONOO⁻

The photophysical behavior of DA was systematically investigated to evaluate its ONOO⁻ sensing performance (Fig. 3A). UV–Vis and fluorescence spectroscopy, key analytical techniques for characterizing analyte responses, were employed to examine DA’s reactivity toward ONOO⁻. In the UV–Vis spectrum, DA displayed a dominant absorption peak at 349 nm. Upon ONOO⁻ addition, a new absorption band emerged at 557 nm, accompanied by a visible color transition from colorless to red, enabling visual detection of ONOO⁻ (Fig. 3B). Fluorescence spectra (Fig. 3C) revealed intense green emission centered at 516 nm, arising from the synergistic effects of ESIPT and AIE. Interaction with ONOO⁻ suppressed both ESIPT and AIE processes, causing a pronounced redshift to 580 nm and reducing the fluorescence quantum yield from Φ = 0.35 at 516 nm (probe DA in the AIE state) to Φ = 0.17 at 580 nm (after reaction with ONOO⁻). Additionally, the large Stokes shift (167 nm) effectively minimized excitation and scattering interference in imaging experiments (Fig. 3D). Collectively, these results identify DA as a ratiometric fluorescent probe capable of reliable ONOO⁻ detection.

Fig. 3.

Fluorescence spectrum of DA. (A) Detection diagram of DA for ONOO⁻. (B) The absorbance of DA (10 μM) in the absence and presence of ONOO⁻ (70 μM). Inset: Photos of DA and DA + ONOO⁻ under natural light. (C) Fluorescence spectra of DA (10 μM) in the absence and presence of ONOO⁻ (70 μM). Inset: Photos of DA and DA + ONOO⁻ under ultraviolet light. (D) The normalized absorption spectra of DA and the normalized fluorescence spectra of DA after adding ONOO⁻. (E) 3D fluorescence spectra obtained during the titration of the probe DA (10 μM) with ONOO⁻ concentrations (0–80 μM) in the mixed solution. (F) Fluorescence intensity ratio (I_{580 nm}/I_{516 nm}) changes with ONOO⁻ concentrations (0–200 μM). (G) Fluorescence intensity ratio of the probe DA (10 μM) to different analytes (ONOO⁻ 70 μM, other analytes 100 μM) (I_{580 nm}/I_{516 nm}). Analyte: (1) blank, (2) Ag⁺, (3) Al³⁺, (4) Arg, (5) ATP, (6) Ba²⁺, (7) Ca²⁺, (8) ClO⁻, (9) Co²⁺, (10) Cr³⁺, (11) Cu²⁺, (12) Cys, (13) EDTA, (14) Fe²⁺, (15) Fe³⁺, (16) Gla, (17) H₂O₂, (18) H₂S, (19) Hcy, (20) ¹O₂, (21) K⁺, (22) Li⁺, (23) (Cys)₂, (24) Mg²⁺, (25) NO, (26) NO₂⁻, (27) Na⁺, (28) Ni²⁺, (29) NO₃⁻, (30) ∙OH, (31) SO₃²⁻, (32) SO₄²⁻, (33) tBuOOH, (34) Zn²⁺, (35) Pd²⁺, (36) Sn²⁺, (37) Sn⁴⁺, (38) Hg²⁺, (39) N₂H₄. (H) Competition experiment: Fluorescence intensity of DA (10 μM) on ONOO⁻ (70 μM) (with/without the above analyte [100 μM]). Green columns: fluorescence intensity ratio (I_{580 nm}/I_516 nm) for DA with the interferent, orange columns: after adding ONOO⁻ into the premixed DA with the interferent. (I) Changes in the ratio of fluorescence intensity (I_{580 nm}/I_{516 nm}) with or without ONOO⁻ (70 μM) in DA (10 μM) at different pH values. (J) Changes in the ratio of fluorescence intensities (I_{580 nm}/I_{516 nm}) of DA (10 μM) and ONOO⁻ (70 μM) over time. All fluorescence measurements were performed in MeCN-H₂O (f_w = 90%) solvent at room temperature. λ_ex = 365 nm, slits = 2 nm, λ_em = 516 nm for probe DA and λ_em = 580 nm for DA + ONOO⁻. Error bars represent standard deviation from three independent measurements (n = 3)

To quantify DA’s ratiometric response toward ONOO⁻, fluorescence titration experiments were performed. As shown in Fig. 3E and 3F, gradual addition of ONOO⁻ (0–200 μM) led to a progressive decrease in emission intensity at 516 nm and a concomitant increase at 580 nm. The fluorescence intensity ratio (I_{580 nm}/I_{516 nm}) increased steadily, reaching a > 100-fold enhancement at saturation. Colorimetric evolution was further confirmed by the CIE 1931 chromaticity diagram (Figure S17 and S18), where emission coordinates shifted from green (0.29, 0.59) to orange (0.54, 0.44) with increasing ONOO⁻ concentration, offering an intuitive visualization for ONOO⁻ identification. A strong linear relationship (R² = 0.98844) was observed within 0–70 μM, with the limit of detection (LOD) of 6.4 nM criterion, signal-to-noise ratio method (), Figure S19), surpassing the sensitivity of most reported ONOO⁻ probes (Table S2). These findings confirm DA’s exceptional sensitivity and quantitative capability for ONOO⁻ detection.

Given the complexity of endogenous species in plant systems, rigorous evaluation of DA’s selectivity toward ONOO⁻ was conducted. Systematic assessments were performed against 38 potential interferents, including metal ions [silver ion (Ag⁺), aluminium ion (Al³⁺), barium ion (Ba²⁺), calcium ion (Ca²⁺), cobalt ion (Co²⁺), chromic ions (Cr³⁺), cupric ion (Cu²⁺), ferrous ion (Fe²⁺), ferric ion (Fe³⁺), potassium ion (K⁺), lithium ion (Li⁺), magnesium ion (Mg²⁺), sodium ion (Na⁺), nickel ion (Ni²⁺), palladium(II) ion (Pd²⁺), stannous ion (Sn²⁺), stannic ion (Sn⁴⁺), mercury (II) ion (Hg²⁺), zinc ion (Zn²⁺)]; small molecules [adenosine triphosphate (ATP), ethylenediaminetetraacetic acid (EDTA), γ-carboxyglutamic acid (Gla), L-arginine (Arg), L-cysteine (Cys), L-homocysteine (Hcy), L-cystine (Cys)₂, hydrazine (N₂H₄)]; and reactive species [hypochlorite ion (ClO⁻), hydrogen peroxide (H₂O₂), hydrogen sulfide (H₂S), singlet oxygen (¹O₂), nitric oxide (NO), nitrite ion (NO₂⁻), nitrate ion (NO₃⁻), hydroxyl radical (∙OH), sulfite ion (SO₃²⁻), sulfate ion (SO₄²⁻), tert-butyl hydroperoxide (tBuOOH)]. Among these, only ONOO⁻ elicited a pronounced fluorescence response at 580 nm (Fig. 3G). Competitive assays using DA (10 μM), ONOO⁻ (70 μM), and each interferent (100 μM) demonstrated a consistent fluorescence enhancement (Fig. 3H), confirming DA’s exceptional selectivity and strong anti-interference capability for ONOO⁻ detection.

The fluorescence response of DA was further evaluated across pH 6–9, reflecting physiological conditions in plant systems (Fig. 3I). The probe maintained stability within this range, exhibiting substantial fluorescence enhancement upon ONOO⁻ addition, thus confirming its reliability across biologically relevant pH conditions.

Stability studies (Figure S20) also demonstrated that the fluorescence intensity ratio (I_{580 nm}/I_{516 nm}) remained constant over 60 min at room temperature and across a temperature range of 0–60 ℃, confirming the excellent temporal and thermal stability of DA.

Time-dependent fluorescence kinetics (Fig. 3J) revealed that, in the absence of ONOO⁻, the emission intensity ratio (I_{580 nm}/I_{516 nm}) remained constant. Upon ONOO⁻ exposure, the ratio sharply increased, reaching equilibrium within 1 min, highlighting DA’s rapid response kinetics and suitability for real-time biological detection.

Sensing mechanism

Probe DA initially exhibits AIE driven by RIM, supported by an ESIPT process between the benzothiazole nitrogen and the phenolic hydroxyl group, which suppresses benzothiazole rotation. This mechanism is corroborated by DFT and time-dependent density functional theory (TD-DFT) calculations (Fig. 4B and 4D). Molecular electrostatic potential (ESP) analysis confirms the intramolecular O–H⋯N hydrogen bond prerequisite for ESIPT (Figure S21). Frontier molecular orbital analysis shows that the E* to K* transition reduces the HOMO–LUMO energy gap by 0.55 eV, corresponding to the red-shifted emission of the K* tautomer. Potential energy curve scanning reveals a negligible energy barrier for proton transfer in the S₁ state, strongly favoring ESIPT upon photoexcitation, while the S₀ state exhibits an unfavorable barrier (Figure S22). Upon nucleophilic attack by ONOO⁻ at the hydrazine site, amide bond cleavage and spirolactam ring opening occur, producing a π-extended DD fluorophore with a reduced band gap of 2.70–2.37 eV (DFT/TD-DFT, Fig. 4C and 4F). This structural conversion is further validated by ¹H NMR titration (Fig. 4E), where the emergence of a characteristic xanthene proton signal at 5.6 ppm and downfield shift of proton H unambiguously confirm spiro-ring opening. The transformation induces a fluorescence redshift from 516 to 580 nm, terminating both ESIPT and AIE and disassembling the aggregates (SEM, Figure S23). HRMS confirms the reaction pathway by detecting both the DD product (m/z = 521.1511, Figure S24) and residual DA (m/z = 606.2151, Figure S25). This ONOO⁻-triggered dual transformation—structural rearrangement and aggregate disruption—enables a distinct ratiometric fluorescence response, making DA a robust and selective probe for ONOO⁻ detection in complex biological environments.

Fig. 4.

Probe DA’s ONOO⁻ sensing and ESIPT mechanism. (A) Sensing mechanism of the probe DA for ONOO⁻. (B) Under aqueous system conditions, the HOMO and LUMO of DA-enol and DA-keto were calculated at the B3LYP functional and 6-31G (d) basis set levels using Gaussian 16 software. (C) DFT-calculated HOMO and LUMO energy levels and orbital distributions of the ring-opened product DD formed. (D) Frontier molecular orbitals (HOMOs and LUMOs) of DA and corresponding energy levels of the enol and keto forms in the S₁ state. (E) Partial ¹H NMR spectra (500 MHz, DMSO-d₆) of probe DA before (bottom) and after (top) addition of ONOO⁻. (F) TD-DFT-calculated HOMO and LUMO energy levels of the ring-opened product DD formed

Imaging of ONOO⁻ in living cells

To assess biocompatibility, germination assays were conducted using mung bean (Vigna radiata) and tobacco (Nicotiana tabacum) models, confirming that DA exhibits no significant phytotoxicity, validating its safety for biological applications. Seeds were incubated in Petri dishes containing DA solutions (0–100 μM) for 2 and 7 days. Quantitative analysis showed sustained germination viability (> 90% for V. radiata, > 85% for N. tabacum) even at 100 μM DA (Fig. 5). No morphological abnormalities were observed in radicles or cotyledons. These results demonstrate the negligible phytotoxicity of DA, confirming its excellent biocompatibility and suitability for in planta fluorescence imaging applications.

Fig. 5.

DA’s dose effects on mung bean (Vigna radiata) and tobacco (Nicotiana tabacum) germination. (A) Germination rate of mung bean (Vigna radiata) sprouts under the influence of different concentrations of DA. (B) Germination rate of tobacco (Nicotiana tabacum) under the influence of different concentrations of DA. (C) Germination diagrams of a graph showing the effects of different concentrations of DA on the germination of mung bean (Vigna radiata) and tobacco (Nicotiana tabacum). Significant differences between the experimental groups are indicated by *p < 0.05, **p < 0.01, and ***p < 0.001

Capitalizing on the DA probe’s superior optical performance and biosafety, we employed it for in situ monitoring of ONOO⁻ in plant tissues (Fig. 6). Using onion (Allium cepa) epidermal peels—chosen for their paradermal architecture and optical transparency confocal microscopy revealed distinct DA accumulation within cytoplasmic compartments. As shown in Fig. 6A and 6C, incubation with 10 μM DA generated a characteristic green fluorescence signal. Subsequent exposure to ONOO⁻ (0–70 μM) elicited a concentration-dependent ratiometric shift, manifested as progressive quenching of green emission (516 nm) accompanied by enhancement of red fluorescence (580 nm). Time-course imaging further demonstrated that prolonged ONOO⁻ exposure caused a gradual decline in green intensity concurrent with time-dependent amplification of red signals (Fig. 6B and 6D). Complementary imaging of tobacco (Nicotiana tabacum) root tissues (Figure S26 and S27) confirmed that DA exhibits high tissue permeability and reliable performance for dynamic visualization of ONOO⁻ fluctuations in plants. Importantly, control groups treated with NO (100 μM) or H₂O₂ (100 μM) in the presence of DA exhibited only green fluorescence without any red signal, demonstrating that the red fluorescence response is specifically triggered by ONOO⁻ and not by other reactive nitrogen or oxygen species. These results further validate the selective ONOO⁻ sensing capability of DA in complex plant tissues.

Fig. 6.

Detection of ONOO⁻ in onion (Allium cepa) epidermal cells. (A) Confocal fluorescence images for onion (Allium cepa) epidermal cells pretreated with the probe DA (10 μM) by adding various ONOO⁻ (0, 10, 25, 50, and 70 μM). (B) Fluorescence images of onion (Allium cepa) epidermal cells incubated with DA after the addition of the same concentration of ONOO⁻ for different times of incubation. (C) Quantification of the fluorescence images in (A). (D) Quantification of the fluorescence images in (B). (Green fluorescence: excited at 488 nm, collected at 451–579 nm. Red fluorescence: excited at 561 nm, collected at 580–635 nm. Scale bar: 100 μm)

Fluorescence imaging of ONOO⁻ in plants under Cd²⁺ stress

Cd²⁺ exerts potent phytotoxic effects by inducing ROS-mediated oxidative damage and suppressing antioxidant defenses, ultimately impairing growth and cellular homeostasis [46, 47]. Elevated ONOO⁻ levels serve as endogenous biomarkers of Cd²⁺ toxicity [48]. To visualize ONOO⁻ generation under Cd²⁺ stress, tomato (Lycopersicon esculentum) seedlings were treated with 10 μM DA during exposure to Cd²⁺. As depicted in Fig. 7B, seedling growth declined progressively with increasing Cd²⁺ concentrations (0–100 μM). Fluorescence imaging revealed dose-dependent ratiometric changes in root tips and stems, characterized by decreasing green and increasing red emissions (Fig. 7D, 7E, S28A, S28B). Notably, co-treatment with the ONOO⁻ scavenger uric acid (UC) under 100 μM Cd²⁺ stress restored green fluorescence and suppressed red emission relative to Cd²⁺ exposure alone, confirming that the spectral shifts originated from ONOO⁻ formation. These results establish ONOO⁻ as a robust biomarker of heavy-metal-induced oxidative stress and provide mechanistic insight into plant adaptive responses, thereby offering potential strategies to mitigate Cd²⁺ toxicity through targeted stress regulation.

Fig. 7.

Fluorescence imaging of ONOO⁻ showing tomato (Lycopersicon esculentum) stress responses to Cd²⁺/NaCl via. (A) Schematic diagram of the stress experiment. (B) Growth map of tomato (Lycopersicon esculentum) under Cd²⁺ stress. (C) Tomato (Lycopersicon esculentum) growth map under salt stress. (D) Confocal fluorescence images of endogenous ONOO⁻ under stress. Tomato (Lycopersicon esculentum) seedlings were incubated with Cd²⁺ (0, 25, 50, 70, and 100 μM) and then treated with the probe DA (10 μM). After incubation with Cd²⁺ 100 μM, incubation with UC and determination of the values of I_Green and I_Red. (E) Quantification of the fluorescence images in (D). (F) Confocal fluorescence images of endogenous ONOO⁻ under NaCl stress. Tomato (Lycopersicon esculentum) seedlings were incubated with NaCl (0, 25, 50, 70, and 100 μM) and then treated with the probe DA (10 μM). After incubation with NaCl 100 μM, incubation with UC and determination of the values of I_Green and I_Red. (G) Quantification of the fluorescence images in (F). (Green fluorescence: excited at 488 nm, collected at 451–579 nm. Red fluorescence: excited at 561 nm, collected at 580–635 nm. Scale bar: 100 μm)

Fluorescence imaging of ONOO⁻ in plants under NaCl stress

Soil salinity similarly compromises crop productivity by disrupting cellular redox balance, driving excessive production of reactive oxygen/nitrogen species and consequent cytotoxic ONOO⁻ accumulation in plant tissues [49–51]. Real time tracking of ONOO⁻ dynamics using DA in salt-stressed tomato (Lycopersicon esculentum) seedlings therefore elucidates the molecular basis of stress adaptation. As shown in Fig. 7C, seedling height decreased with rising NaCl concentrations (0–100 μM). Correspondingly, fluorescence imaging demonstrated a gradual reduction in green intensity with a concomitant rise in red fluorescence within root tips and rhizome regions (Fig. 7F, 7G, S28C, S28D). Control experiments with UC following 100 μM NaCl exposure verified that these spectral transitions stemmed from ONOO⁻ generation. Collectively, these observations provide direct experimental validation that salt stress provokes ONOO⁻ overproduction, advancing our understanding of the molecular pathways underlying salinity-induced oxidative injury.

Fluorescence imaging of ONOO⁻ in plants under erastin stress

Abiotic stresses can also induce ferroptosis—an iron-dependent form of programmed cell death characterized by lipid peroxidation and ROS accumulation [52, 53]. ONOO⁻ amplifies this process by oxidizing membrane lipids and promoting structural destabilization [54–58]. Despite its pathological relevance, direct evidence linking ONOO⁻ to plant ferroptosis remains limited, underscoring the need for advanced imaging strategies. To investigate the spatiotemporal dynamics of ONOO⁻ and its association with lipid peroxidation under ferroptotic stress in plants, erastin was employed as a classical ferroptosis inducer. Treatment with erastin resulted in a dose-dependent reduction in the expression levels of two key regulatory proteins: solute carrier family 7 member 11 (SLC7A11)—a core component of System Xc⁻ (a cystine/glutamate antiporter responsible for importing cystine for glutathione synthesis), and glutathione peroxidase 4 (GPX4), indicating disruption of the cellular antioxidant defense system and subsequent uncontrolled lipid peroxidation. Under these conditions, the rapid accumulation of ROS reacts with NO to generate the highly reactive oxidant ONOO⁻. Using a specific fluorescent probe, we visualized a concomitant, dose-dependent increase in ONOO⁻ levels in erastin-treated tobacco (Nicotiana tabacum) seedlings, thereby confirming its role as a key downstream mediator in the ferroptosis pathway. Notably, co-treatment with UC—a known scavenger of ONOO⁻—restored basal green fluorescence and suppressed red emission (Fig. 8), suggesting mitigation of oxidative damage. These results provide the first direct molecular and visual evidence linking ONOO⁻ accumulation to the canonical ferroptosis signaling cascade in plant systems.

Fig. 8.

Ferroptosis-mediated ONOO⁻ generation visualized in erastin-treated tobacco (Nicotiana tabacum). (A) Schematic diagram of fluorescence imaging of erastin-induced ferroptosis (drawn using BioRender). (B) Effect of erastin on SLC7A11 expression levels. (C) Effect of erastin on GPX4 expression levels. (D) Confocal fluorescence images of endogenous ONOO⁻ produced by the root tips of tobacco (Nicotiana tabacum) induced by erastin. The tobacco (Nicotiana tabacum) seedlings were treated with erastin (0, 20, 40, 60, and 80 μM) for 1 day and then treated with the probe DA (10 μM). Then incubate with 80 μM erastin for 1 day and incubate with UC to determine the I_Green and I_Red values, and (E) quantification of the fluorescence images in (D). (F) Confocal fluorescence images of endogenous ONOO⁻ produced by the root stems of tobacco (Nicotiana tabacum) induced by erastin. Tobacco (Nicotiana tabacum) seedlings were treated with erastin (0, 20, 40, 60, and 80 μM) and then treated with the probe DA (10 μM), followed by culture with 80 μM erastin and then with UC to determine I_Green and I_Red. (G) Quantification of the fluorescence images shown in (F). Significant differences between the experimental groups are indicated by *p < 0.05, **p < 0.01, and ***p < 0.001. (Green fluorescence: excited at 488 nm, collected at 451–579 nm. Red fluorescence: excited at 561 nm, collected at 580–635 nm. Scale bar: 100 μm)

Conclusions

Based on the synergistic ESIPT–AIE mechanism, we successfully developed a novel nano-engineered ratiometric fluorescent probe [59–61], DA, with a particle size of approximately 258 nm. The probe incorporates a rigid fluorophore scaffold stabilized by intramolecular hydrogen bonds, which effectively suppresses non-radiative transitions, while its AIE-active structure mitigates ACQ. Nano-engineering further enhances the probe’s biocompatibility and tissue penetration. DA exhibits a high signal-to-noise ratio for ONOO⁻ detection (I_{580 nm}/I_{516 nm} > 100-fold), an ultra-low LOD of 6.4 nM; outperforming most reported ONOO⁻ probes), a large Stokes shift (167 nm), and outstanding selectivity. In vivo imaging confirmed DA’s capacity for spatiotemporal visualization of ONOO⁻ in plant tissues—including onion (Allium cepa) epidermis and tobacco (Nicotiana tabacum) roots—revealing that ONOO⁻ accumulation under Cd²⁺, NaCl, and erastin stress correlates with oxidative damage. Overall, this study presents an innovative ESIPT-AIE-based probe that elucidates the mechanistic role of ONOO⁻ in plant oxidative stress and establishes a new analytical platform for precision agriculture.

Author contributions

Guang-Ye Wang: Conceptualization, Investigation, Formal analysis, Writing – original draft. Shuai Tan: Investigation. Feng Gao: Investigation. Wei Niu: Investigation. Shi-Tao Liu: Investigation. Ya-Ping Wu: Investigation. Tian-Li Lu: Investigation. A-Ling Tang: Investigation, Conceptualization, Project administration. Xiang Zhou: Investigation, Supervision, Writing – review & editing. Song Yang: Conceptualization, Supervision, Project administration, Resources, Writing – review & editing. Funding acquisition.

Funding

This work was supported by National Key Research and Development Program of China (2024YFE0214300, 2022YFD1700300), National Natural Science Foundation of China (32372610, U23A20201, 32160661, 32202359), the Central Government Guides Local Science and Technology Development Fund Projects [Qiankehezhongyindi (2023) 001], [Qiankehezhongyindi (2024) 007], Scientific and Technological Innovation Platform Research Project of Guizhou Province (CXPTXM [2025] 012), Major Scientific and Technological Achievement Transformation Project of Guizhou Province (Qian Ke He Zhong Yin Di [2024]027), Outstanding Young Scientific and Technological Talent Project of Guizhou Province (Qian Ke He Ping Tai Ren Cai YQK [2023]004)." Thanks.This work was supported by National Key Research and Development Program of China (2024YFE0214300, 2022YFD1700300), National Natural Science Foundation of China (32372610, U23A20201, 32160661, 32202359), the Central Government Guides Local Science and Technology Development Fund Projects [Qiankehezhongyindi (2023) 001], [Qiankehezhongyindi (2024) 007], Scientific and Technological Innovation Platform Research Project of Guizhou Province (CXPTXM [2025] 012), Major Scientific and Technological Achievement Transformation Project of Guizhou Province (Qian Ke He Zhong Yin Di [2024]027), Qiankehe Platform DLSYS9 (2025) Major 001, Outstanding Young Scientific and Technological Talent Project of Guizhou Province (Qian Ke He Ping Tai Ren Cai YQK [2023]004).

Data availability

Data will be made available on request.

Declarations

Competing interests

The authors declare no competing interests.