English

• Reactive sulfur species (RSS) are a family of sulfur containing molecules that play vital roles in biological systems and are involved in diverse signal transduction pathways [1]. Typical RSS include biological thiols [2,3], hydrogen sulfide (H₂S) [4-10], hydrogen polysulfides (H₂S_n, n > 1), persulfides, polysulfides [11,12], and S-modified cysteine adducts [13], among others. As research in this field advances, certain biological effects in some physiological and pathological processes previously ascribed to H₂S have been found to be mediated by H₂S₂ [14], which exhibits stronger nucleophilicity and reducibility than biological thiols or H₂S. For instance, H₂S₂ is implicated in processes such as tumor suppressor activation, cardiovascular regulation, ion channel and transcription factor modulation, protein S-persulfidation, inflammation, and antioxidant responses [15-17]. Despite its significant biological importance, research on H₂S₂ remains in its nascent stages, with many unresolved questions persisting. Therefore, the development of effective tools for bioimaging of H₂S₂ in vivo with high specificity and high sensitivity is of great significance, especially deepening our understanding of its physiological functions in complex biological systems.

  Traditional methods such as ultraviolet (UV) spectroscopy, colorimetry, chromatography, and mass spectrometry are hard for real-time monitoring and in situ tracking of H₂S_n dynamics [18-20]. As an alternative, fluorescence-based approaches, especially when coupled with molecular probes, offer visual imaging and non-destructive analytical capabilities for the detection of chemical species within living systems [21-24]. To date, several molecular probes for monitoring H₂S_n in living systems have been reported [17,25-40], but most of these probes often suffer from limitations such as poor selectivity for polysulfides over other sulfur species, sluggish kinetics, susceptibility to esterase-mediated hydrolysis, and the generation of cytotoxic byproducts. These drawbacks compromise the accuracy and reliability of H₂S_n detection in complex biological systems. To overcome these challenges, recently, our group have discovered an H₂S₂-mediated dithiole formation reaction that showed selectivity for H₂S₂ [39,41]. However, we observed that upon prolonged reaction time (particularly beyond 1 h, Fig. 1), the novel 1,3-dichloro recognition motif also exhibited a degree of unresolved cross-reactivity with H₂S. Consequently, the development of reliable probes capable of achieving highly specific detection of H₂S₂ in vivo remains an urgent challenge.

  Figure 1

  

  Figure 1.  Time-dependent absorption changes of 10 µmol/L H₂S₂–1 (A), H₂S₂–2 (B), H₂S₂–3 (C), and H₂S₂–4 (D) in the presence of 50 µmol/L Na₂S₂.

  DownLoad: Full-Size Img PowerPoint

  In this work, we designed and synthesized a series of H₂S₂-activated near-infrared-Ⅱ (NIR-Ⅱ) cyanine molecular probes, namely H₂S₂1–4, which regulated the aromatic nucleophilic substitution reaction between the 1,3-dichloro moiety and the dual sulfhydryl groups of H₂S₂ through the electronic effect of substituents (Scheme 1). As expected, the dual ratiometric NIR-Ⅱ fluorescence (FL)/photoacoustic (PA) probe, H₂S₂–4, performed best among the four probes designed due to the strong electron-donating effect of the hydroxyl substituents, demonstrating selective recognition of H₂S₂ over H₂S with robust ratiometric responses in both NIR-Ⅱ fluorescence (I₈₉₀/I₁₀₂₀, 42-fold increase) and photoacoustic imaging (PA₈₄₀/PA₉₆₀, 4.49-fold enhancement). Moreover, H₂S₂–4 has been successfully applied to monitor endogenous H₂S₂ dynamics in Escherichia coli (E. coli)-induced acute peritonitis and hindlimb infection mouse models with high accuracy and reliability in vivo. Furthermore, this study elucidates an electronic tuning strategy that can be generalized to the design of advanced probes for precision imaging H₂S₂.

  Scheme 1

  

  Scheme 1.  Schematic illustrating the dual ratiometric NIR-Ⅱ FL/PA probe for H₂S₂ with unprecedented specificity based on electronic tuning strategy.

  DownLoad: Full-Size Img PowerPoint

  Given that H₂S₂ exhibits stronger nucleophilicity than H₂S [11], traditional fluorescent probes demonstrate high reactivity toward both species, making it challenging to design highly specific H₂S₂ probes. Generally, modulating the electron donating ability of the aromatic moiety may reduce the nucleophilic reactivity, thereby enabling the development of high-performance probes for H₂S₂ imaging. In light of the aforementioned considerations, we propose an electron tuning strategy for designing highly specific H₂S₂ probes. To validate our hypothesis, we designed and synthesized four probes with different electron-donating groups (H, CH₃, OCH₃, and OH). As illustrated in the ultraviolet-visible (UV–vis) spectroscopic tests (Fig. 1, Figs. S1 and S2 in Supporting information), H₂S₂–1 reacted rapidly with H₂S₂ and H₂S, with selective differences for a short period of time (Fig. 1A); however, with the prolongation of time (≥60 min), it showed obvious unresolved cross-reactivity with H₂S, underscoring the need for further optimization. For compounds H₂S₂–2 and H₂S₂–3, this enhancement in electron-donating ability resulted in slower spectroscopic response kinetics toward H₂S₂ and H₂S, accompanied by a gradual increase in selectivity (Figs. 1B and C). Ultimately, the introduction of a hydroxy group with electron-donating ability onto the NIR-Ⅱ cyanine (Scheme 1 and Scheme S1 in Supporting information) endowed the probe (H₂S₂–4) with the ability to remain inert toward H₂S for an extended period (≥60 min) while maintaining reactivity toward H₂S₂ (Fig. 1D). Additionally, the probe further exhibited excellent stability in phosphate-buffered saline (PBS) and saline (Fig. S3 in Supporting information). Considering both reactivity and selectivity, H₂S₂–4 was selected as the optimal probe for subsequent studies.

  The NIR-Ⅱ FL/PA dual-ratiometric responses of H₂S₂–4 to H₂S₂ were evaluated by means of UV–vis absorption spectroscopy, fluorescence spectroscopy, and PA spectroscopy (Tables S1-S4 in Supporting information). As presented in Fig. 2A, the UV–vis absorption spectra of H₂S₂–4 showed that as the concentration of H₂S₂ increased, the absorption peak at 960 nm weakened, while a new absorption signal appeared at 840 nm. A robust linear relationship (Fig. 2E) provided additional confirmation that H₂S₂–4 acts as a colorimetric probe for H₂S₂. Fluorescence titration spectra, which were obtained under different H₂S₂ concentration conditions (Fig. 2B), showed a reduction in emission intensity at 1020 nm and the emergence of a new emission peak at 890 nm. The 130 nm emission shift and well-separated dual-emission bands verified the outstanding NIR-Ⅱ fluorescence ratiometric response of H₂S₂–4 toward H₂S₂. The fluorescence intensity ratio (I₈₉₀/I₁₀₂₀) exhibited a 42-fold increase and showed a linear dependence on H₂S₂ concentration (Fig. 2F). The computed detection limit (3σ/k) for H₂S₂ via H₂S₂–4, calculated on the basis of the fluorescence ratio, was determined to be approximately 38.8 nmol/L. This result indicates that H₂S₂–4 is capable of effectively detecting H₂S₂ at relatively low concentrations. The selectivity of the probe toward biologically relevant interfering substances was evaluated further. Of note, when Na₂S and NaClO were applied together, a significant absorption change was observed which is attributed to the in situ generation of H₂S₂ from H₂S and ClO⁻ (Fig. 2C). A panel of RSS, including glutathione (GSH), cysteine (Cys), homocysteine (Hcy), S₂O₃²⁻, SO₄²⁻, Na₂S, Na₂S₄, and S₈ did not trigger any significant enhancement in the fluorescence ratio (Fig. 2D).

  Figure 2

  

  Figure 2.  Time-dependent (A) absorption changes and (B) emission changes of 10 µmol/L H₂S₂–4 in the presence of 80 µmol/L Na₂S₂, λ_ex = 808 nm. (C) Absorption changes of H₂S₂–4 in the presence of relevant reactive species. (D) The fluorescence selectivity of H₂S₂–4 toward different analytes in PBS. H₂O₂ (200 mmol/L), GSH (1 mmol/L), Hcy (100 mmol/L), Cys (60 mmol/L), Na₂S (80 µmol/L), Na₂S₄ (80 µmol/L), S₈ (80 µmol/L), Na₂S₂ (80 µmol/L) the concentrations of all other substances were 100 µmol/L, and the concentrations of all analytes mentioned later are consistent with those here. (E) Linear relationship between the absorption intensity ratio (A₈₄₀/A₉₆₀) and Na₂S₂ concentration. (F) Linear relationship between the fluorescence intensity ratio (I₈₉₀/I₁₀₂₀) and Na₂S₂ concentration. (G) Linear relationship between the PA intensity ratio (PA₈₄₀/PA₉₆₀) and Na₂S₂ concentration, illustration: PA images of H₂S₂–4 solution responding with different Na₂S₂. (H) The PA selectivity of H₂S₂–4 toward different analytes in PBS. (I) Fluorescence images of capillaries filled with H₂S₂–4 (100 µmol/L) (λ_ex = 808 nm (1100 mW/cm²)), 100 ms exposure time and a 1000 long pass (LP) filter, respectively, immersed in 1% intralipid with varying depth. (J) SBR for fluorescence imaging of H₂S₂–4 as a function of the intralipid depth in (I). (K) Fluorescence responses of the probe H₂S₂–4 (10 µmol/L) in the absence and presence of Na₂S₂ (80 µmol/L) under different pH values. Data are the mean ± SD (n = 3).

  DownLoad: Full-Size Img PowerPoint

  PA imaging of H₂S₂–4 solution samples revealed that under excitation at 840 nm, the PA signal exhibited a significant increase upon titration with Na₂S₂; by contrast, the PA signal under excitation at 960 nm underwent a decrease during this titration process. With the concentration of Na₂S₂ rising from 0 to 80 µmol/L, the PA₈₄₀/PA₉₆₀ ratio increased in a nearly linear manner, from roughly 0.43 ± 0.03 to 1.93 ± 0.07. Additionally, PA₈₄₀/PA₉₆₀ showed roughly a 4.49-fold increase (Fig. 2G and Fig. S4 in Supporting information).

  Additionally, in vitro PA selectivity experiments demonstrated that interfering species did not bring about significant changes in the PA₈₄₀/PA₉₆₀ ratio (Fig. 2H), which was consistent with the fluorescence selectivity results. Thus, H₂S₂–4 demonstrated excellent specificity for H₂S₂ in dual-ratiometric NIR-Ⅱ FL/PA detection. Furthermore, we examined the impact of pH on the reaction of H₂S₂–4 with H₂S₂. As presented in Fig. 2K, after addition of Na₂S₂, H₂S₂–4 retained efficient functionality across a pH range spanning weakly acidic to alkaline conditions. Taken together, these results unequivocally demonstrate that H₂S₂–4 is suitable for the detection of H₂S₂ via the dual-ratiometric NIR-Ⅱ FL/PA mode in vitro.

  To clarify the potential reaction mechanism, we further verified the reaction of H₂S₂–4 with Na₂S₂ via electrospray ionization mass spectrometry (ESI-MS) (Fig. S5 in Supporting information), confirming the reactivity of H₂S₂–4 toward H₂S₂. Additionally, we selected H₂S₂–1 as the model compound. As shown in Fig. S6 (Supporting information), ESI-MS analysis of the reaction mixture of H₂S₂–1 and Na₂S verified the Na₂S₂-induced response mechanism. After adding Na₂S₂ to the H₂S₂–1 solution, a molecular ion peak was observed at [M]⁺ = 591.2864 (Fig. S7 in Supporting information). In contrast, when Na₂S was added to the H₂S₂–1 solution, the intermediate ion peak at [M]⁺ = 593.2874 and the post-response compound ion peak at [M]⁺ = 591.2869 were sequentially detected over time (Fig. S6), further confirming that nucleophilic substitution occurs first, followed by dithiolane ring formation. Moreover, we used the Gaussian 09 program to calculate the natural bond orbital (NBO) charges of carbon atoms at specific sites in different probes; the NBO charge level is steadily decreasing (Fig. S8 in Supporting information). These results confirm the effectiveness of our proposed molecular regulation strategy.

  Prior to evaluating the imaging potential of H₂S₂–4 for in vitro H₂S₂ detection, we assessed its tissue penetration ability by performing imaging on a liposome model using a capillary tube loaded with H₂S₂–4. This capillary tube was immersed in a 1% liposome solution. As shown in Fig. 2I, even at a depth of 6 mm in the 1% liposome solution, the wall of the tube containing H₂S₂–4 remained clearly distinguishable in the NIR-Ⅱ region, and its signal-to-background ratio (SBR) is as shown in Fig. 2J. These results demonstrate the superior imaging resolution and contrast of H₂S₂–4, establishing its suitability for in vivo deep-tissue imaging applications.

  NIR-Ⅱ FL/PA dual-modality imaging can overcome the shortcomings of a single modality and achieve complementary advantages to further optimize the signal output of biological tissues [42-48]. In order to assess the capability of H₂S₂–4 for in vitro H₂S₂ detection, we first conducted NIR-Ⅱ FL/PA dual-ratiometric imaging experiments using centrifuge tubes (Fig. 3). When Na₂S₂ was absent, the NIR-Ⅱ fluorescence ratio of H₂S₂–4 showed little variation within the 0–30 min period, sustaining a value of roughly 1.71 ± 0.03 (Fig. 3A). After Na₂S₂ was added, the corresponding fluorescence ratio increased from 1.84 ± 0.03 to 4.22 ± 0.02 over time, representing an approximate 2.29-fold increase (Fig. 3B and Fig. S9A in Supporting information). Likewise, Fig. 3C demonstrates that the PA signal ratio of H₂S₂–4 alone stayed stable at approximately 0.55 ± 0.02; in contrast, following the addition of Na₂S₂, the PA ratio rose by a factor of 2.20 (reaching 1.21, Fig. 3D and Fig. S9B in Supporting information). These results revealed a significant change in the NIR-Ⅱ FL/PA ratio in tubes with added Na₂S₂, thus verifying the effectiveness of H₂S₂–4 in ex vivo H₂S₂ detection.

  Figure 3

  

  Figure 3.  (A) Time-dependent NIR-Ⅱ ratiometric fluorescence imaging of probe H₂S₂–4 (100 µL, 100 µmol/L) in the absence of Na₂S₂. (B) Time-dependent NIR-Ⅱ ratiometric fluorescence imaging of probe H₂S₂–4 in the presence of Na₂S₂. (C, D) Time-dependent PA ratiometric imaging of probe H₂S₂–4 in the absence and presence of Na₂S₂. Data are the mean ± SD (n = 3).

  DownLoad: Full-Size Img PowerPoint

  Biocompatibility of a probe serves as a fundamental requirement for in vivo imaging applications. Therefore, the biocompatibility of H₂S₂–4 underwent assessment. Initially, a standard cell counting kit-8 (CCK-8) assay was employed to evaluate the potential cytotoxicity of H₂S₂–4 in viable cells. As illustrated in Fig. S10A (Supporting information), even at a concentration of up to 25 µmol/L, cell viability remained above 80%, a result that verifies its favorable biocompatibility. Additionally, normal saline or H₂S₂–4 was administered via intravenous injection to the control group and the experimental mouse group, respectively. Through 7-day monitoring of the daily behavior of mice and recording of their body weight fluctuations, no deaths were detected, and no notable differences were found in body weight changes (Fig. S10B in Supporting information). Collectively, these findings demonstrate that the probe possesses outstanding biocompatibility and low toxicity, rendering it applicable for in vivo imaging in biological organisms.

  Subsequently, for assessing the in vivo detection capacity of H₂S₂–4 as a chemical tool for detecting exogenous H₂S₂, the probe and Na₂S₂ were administered to mice via subcutaneous injection. All animal experiments were approved by the Animal Ethics Committee of Guangxi Normal University. Variations in the NIR-Ⅱ fluorescence and PA dual-ratiometric signals of the probe were tracked for a 30-min period. The imaging results demonstrated that after the administration of the probe and saline, the NIR-Ⅱ ratio signal intensity (I₈₅₀/I₁₀₀₀) exhibited no significant variation (Fig. 4A). In contrast, following the administration of the probe and exogenous Na₂S₂ (Fig. 4B), as time progressed, the NIR-Ⅱ FL signal intensity exhibited an upward trend at 850 nm, whereas a decrease was observed at 1000 nm. Moreover, the FL intensity ratio increased swiftly and achieved a steady state within a 30-min window. The I₈₅₀/I₁₀₀₀ rose from 1.64 ± 0.02 to 3.25 ± 0.05, corresponding to an approximate 1.98-fold enhancement (Fig. 4E). Furthermore, the ratio in PA imaging showed a similar trend. In comparison, following the administration of the probe and saline, the mice displayed no notable variation in PA intensity in the two channels (Fig. 4C). However, as shown in Fig. 4D, following exposure to Na₂S₂, the PA signal intensity rose at 840 nm while declining at 960 nm. As a result, the PA ratio signal (PA₈₄₀/PA₉₆₀) increased from 0.35 ± 0.02 to 0.65 ± 0.03, corresponding to an approximate 1.85-fold enhancement (Fig. 4F). These experimental results further confirm the observation that H₂S₂–4 can effectively visualize exogenous H₂S₂ in vivo via NIR-Ⅱ FL/PA dual-ratiometric imaging.

  Figure 4

  

  Figure 4.  NIR-Ⅱ ratiometric fluorescence imaging of the probe H₂S₂–4 (100 µmol/L, 100 µL) in the absence (A) and presence (B) of Na₂S₂ with the filters of 850 nm LP, and 1000 nm LP, respectively. PA ratiometric imaging of the probe H₂S₂–4 in the absence (C) and presence (D) of Na₂S₂. (E) NIR-Ⅱ fluorescence ratio (I₈₅₀/I₁₀₀₀) corresponding to (A, B). (F) PA signal ratio (PA₈₄₀/PA₉₆₀) corresponding to (C, D). Note that the ratiometric signals (I₈₅₀/I₁₀₀₀ and PA₈₄₀/PA₉₆₀) reflect relative changes in H₂S₂ levels. Quantitative analysis of the signal changes in different groups. Data are the mean ± SD (n = 3). Statistical analysis was performed using one-way ANOVA followed by multiple comparison tests. *P < 0.05, **P < 0.01, ***P < 0.001.

  DownLoad: Full-Size Img PowerPoint

  Previous studies have shown that bacterial infection causes severe inflammation, leading to increased endogenous H₂S₂ levels [49,50]. Before conducting bacterial infection experiments, we examined the response of H₂S₂–4 to endogenous H₂S₂ using lipopolysaccharide (LPS, an inflammatory mediator) [51]. After LPS administration for 12 h, H₂S₂–4 was injected locally, and the NIR-Ⅱ fluorescence and photoacoustic dual-ratiometric signals of the probe were continuously monitored. As shown in Fig. S11 (Supporting information), imaging results showed no significant changes in the hind limbs of non-LPS-induced mice (Figs. S11A and C). In contrast, after LPS induction (Fig. S11B), the NIR-Ⅱ FL intensity at 850 nm increased over time, while that at 1000 nm decreased, with the NIR-Ⅱ ratio (I₈₅₀/I₁₀₀₀) rising by 1.58-fold relative to the initial value (Fig. S11E). For ratiometric PA imaging, after LPS-induced endogenous H₂S₂ production (Fig. S11D), the PA intensity at 840 nm increased and that at 960 nm decreased, resulting in a 1.51-fold increase in the PA ratio (PA₈₄₀/PA₉₆₀) (Fig. S11F). These findings further confirm that H₂S₂–4 enables efficient in vivo visualization of endogenous H₂S₂ via NIR-Ⅱ FL/PA dual-ratiometric imaging.

  BALB/c mice (25−30 g) were divided into three groups: Control, experimental, and treatment groups. The experimental and treatment groups were given an intraperitoneal injection of E. coli (5 × 10⁶ CFU/mL) in accordance with previously reported protocols. Six hours after inoculation, the treatment group was administered ciprofloxacin once every 6 h for 7 consecutive days, after which NIR-Ⅱ fluorescence/photoacoustic dual-ratiometric imaging was performed. Imaging data revealed that the control and treatment groups showed no notable changes in abdominal NIR-Ⅱ ratiometric signal intensity (Figs. 5A and C). In contrast, the experimental group exhibited time-dependent variations in NIR-Ⅱ FL signals (Fig. 5B): The signal intensity increased in the 850 nm channel, decreased in the 1000 nm channel, and rose rapidly overall. The corresponding NIR-Ⅱ ratio (I₈₅₀/I₁₀₀₀) increased from 1.52 ± 0.02 to 2.79 ± 0.05, representing an approximate 1.85-fold elevation (Fig. 5G). Consistent with this, ratiometric PA imaging exhibited similar trends. The control and treatment groups displayed no notable differences in PA intensity between the PA₈₄₀ and PA₉₆₀ channels (Figs. 5D and F). Notably, when E. coli triggered in vivo H₂S₂ generation (Fig. 5E), the PA signal intensity increased at 840 nm and decreased at 960 nm. As a result, the PA ratio signal (PA₈₄₀/PA₉₆₀) rose from 0.30 ± 0.02 to 0.53 ± 0.03, corresponding to a 1.77-fold enhancement (Fig. 5H). These findings confirmed the ability of H₂S₂–4 to efficiently visualize endogenous H₂S₂ produced during acute peritonitis in vivo.

  Figure 5

  

  Figure 5.  NIR-Ⅱ ratiometric fluorescence imaging of the probe H₂S₂–4 (100 µmol/L, 100 µL) in the absence (A) and presence (B) of E. coli induction and (C) ciprofloxacin treatment in mice. PA ratiometric imaging of the probe H₂S₂–4 in the absence (D) and presence (E) of E. coli induction and (F) ciprofloxacin treatment. (G) NIR-Ⅱ fluorescence ratiometric signal (I₈₅₀/I₁₀₀₀) obtained from (A-C). (H) PA ratiometric signal (PA₈₄₀/PA₉₆₀) obtained from (D-F). Quantitative analysis of the signal changes in different groups. Data are the mean ± SD (n = 3). Statistical analysis was performed using one-way ANOVA followed by multiple comparison tests. *P < 0.05, **P < 0.01, ***P < 0.001.

  DownLoad: Full-Size Img PowerPoint

  To further evaluate the performance of H₂S₂–4 in detecting bacterially induced endogenous H₂S₂ in vivo through the NIR-Ⅱ FL/PA dual-ratiometric imaging method, a hindlimb bacterial infection model was constructed (Fig. 6A). At 12 h post-inoculation, the treatment group was given ciprofloxacin every 6 h for 7 consecutive days, with NIR-Ⅱ FL/PA dual-ratiometric imaging subsequently performed. Imaging results demonstrated no significant changes in NIR-Ⅱ ratiometric signal intensity (I₈₅₀/I₁₀₀₀) in the hindlimbs of control and treatment group mice (Figs. 6B and D). By comparison, the experimental group showed time-dependent changes in NIR-Ⅱ FL signals: Intensity increased at 850 nm and decreased at 1000 nm, with the overall FL intensity rising rapidly (Fig. 6C). Correspondingly, the NIR-Ⅱ ratio (I₈₅₀/I₁₀₀₀) increased from 1.46 ± 0.02 to 2.52 ± 0.05, corresponding to an approximate 1.73-fold increase (Fig. 6E). Concurrently, ratiometric PA imaging exhibited a consistent trend: The control and treatment groups demonstrated no notable differences in PA intensity between the PA₈₄₀ and PA₉₆₀ channels (Figs. 6F-right and G). However, when E. coli induced in vivo H₂S₂ generation (Fig. 6F-left), the PA intensity increased at 840 nm and decreased at 960 nm, resulting in the PA ratio (PA₈₄₀/PA₉₆₀) rising from 0.40 ± 0.02 to 0.68 ± 0.03, an approximate 1.70-fold enhancement (Fig. 6H). These results further confirm that H₂S₂–4 can visualize endogenous H₂S₂ in a mouse model of inflammation induced by bacterial infection.

  Figure 6

  

  Figure 6.  (A) Schematic of E. coli-induced bacterial infections of the hindlimbs model in mice. NIR-Ⅱ ratiometric fluorescence imaging of the probe H₂S₂–4 (100 µmol/L, 100 µL) in the absence (B) and presence (C) of E. coli induction and (D) ciprofloxacin treatment in mice. (E) Corresponding FL intensity ratio (I₈₅₀/I₁₀₀₀). (F) PA ratiometric imaging of the probe H₂S₂–4 in the absence (right) and presence (left) of E. coli induction and ciprofloxacin treatment (G). (H) Corresponding PA intensity ratio (PA₈₄₀/PA₉₆₀). Quantitative analysis of the signal changes in different groups. Data are the mean ± SD (n = 3). Statistical analysis was performed using one-way ANOVA followed by multiple comparison tests. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, this study successfully addresses the challenge of selectively detecting H₂S₂ amidst competing RSS by electronic tuning of NIR-Ⅱ cyanine scaffold. Based on our strategy, a family of H₂S₂-activatable probes (H₂S₂–1 to H₂S₂–4) was rationally prepared in which substituent electronic effects regulate the aromatic nucleophilic substitution between the 1,3-dichloro moiety and H₂S₂’s dual sulfhydryl groups. Among these, the probe H₂S₂–4, equipped with strong electron-donating hydroxyl groups, demonstrated exceptional specificity for H₂S₂ over other biologically relevant sulfur species (including glutathione (GSH), cysteine (Cys), homocysteine (Hcy), Na₂S, Na₂S₄, and S₈). The probe H₂S₂–4 further exhibited robust dual ratiometric responses in both NIR-Ⅱ fluorescence and photoacoustic imaging, enabling precise and reliable quantification. The practical efficacy of H₂S₂–4 was confirmed through the successful monitoring of endogenous H₂S₂ dynamics in live mouse models of E. coli-induced acute peritonitis and bacterial hindlimb infection. More than a powerful tool for investigating H₂S₂-related pathophysiology, this work establishes a generalizable electronic-tuning strategy, paving the way for the rational design of advanced probes for precision deep-tissue imaging.

  CRediT authorship contribution statement

  Zhenni Lin: Writing – review & editing, Writing – original draft, Formal analysis, Data curation, Conceptualization. Bangping Hu: Data curation, Conceptualization. Qinian Liu: Data curation, Conceptualization. Yujie Huang: Data curation, Conceptualization. Xing-Can Shen: Supervision, Funding acquisition. Hua Chen: Writing – review & editing, Writing – original draft, Validation, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, 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.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 22167004), the Natural Science Foundation of Guangxi (No. 2023GXNSFFA026014), the National Natural Science Foundation of China (Nos. 22467004 and 22377018), the BAGUI Young Scholar Program of Guangxi, and the BAGUI Scholar Program of Guangxi, Innovation Project of Guangxi Graduate Education (No. XYCS2025106).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.112339.

  
  
• 1. [1]

     G.K. Kolluru, R.E. Shackelford, X. Shen, et al., Nat. Rev. Cardiol. 20 (2023) 109–125. doi: 10.1038/s41569-022-00741-6

  2. [2]

     X. Chen, Y. Zhou, X. Peng, et al., Chem. Soc. Rev. 39 (2010) 2120–2135. doi: 10.1039/b925092a

  3. [3]

     D. Fass, C. Thorpe, Chem. Rev. 118 (2018) 1169–1198. doi: 10.1021/acs.chemrev.7b00123

  4. [4]

     G. Yang, L. Wu, B. Jiang, et al., Science 322 (2008) 587–590. doi: 10.1126/science.1162667

  5. [5]

     M.R. Filipovic, J. Zivanovic, B. Alvarez, et al., Chem. Rev. 118 (2017) 1253–1337.

  6. [6]

     A.V. Mishanina, M. Libiad, R. Banerjee, Nat. Chem. Biol. 11 (2015) 457–464. doi: 10.1038/nchembio.1834

  7. [7]

     B.C. Yu, Y.Q. Zheng, Z.N. Yuan, et al., J. Am. Chem. Soc. 140 (2018) 30–33. doi: 10.1021/jacs.7b09795

  8. [8]

     K.R. Olson, Y. Gao, F. Arif, et al., Redox Biol. 15 (2018) 74–85. doi: 10.1016/j.redox.2017.11.009

  9. [9]

     R. Wang, Physiol. Rev. 92 (2012) 791–896. doi: 10.1152/physrev.00017.2011

  10. [10]

      T.S. Bailey, M.D. Pluth, J. Am. Chem. Soc. 135 (2013) 16697–16704. doi: 10.1021/ja408909h

  11. [11]

      T. Ida, T. Sawa, H. Ihara, et al., Proc. Natl. Acad. Sci. U. S. A. 111 (2014) 7606–7611. doi: 10.1073/pnas.1321232111

  12. [12]

      P.K. Yadav, M. Martinov, V. Vitvitsky, et al., J. Am. Chem. Soc. 138 (2016) 289–299. doi: 10.1021/jacs.5b10494

  13. [13]

      C.E. Paulsen, K.S. Carroll, Chem. Rev. 113 (2013) 4633–4679. doi: 10.1021/cr300163e

  14. [14]

      H. Liu, M.N. Radford, C.T. Yang, et al., Br. J. Pharmacol. 176 (2019) 616–627. doi: 10.1111/bph.14330

  15. [15]

      N. Sen, B.D. Paul, M.M. Gadalla, et al., Mol. Cell 45 (2012) 13–24. doi: 10.1016/j.molcel.2011.10.021

  16. [16]

      D.H. Zhang, I. Macinkovic, N.O. Devarie-Baez, et al., Angew. Chem. Int. Ed. 53 (2014) 575–581. doi: 10.1002/anie.201305876

  17. [17]

      H. Dong, Q. Zhou, L. Zhang, et al., Angew. Chem. Int. Ed. 58 (2019) 13948–13953. doi: 10.1002/anie.201907210

  18. [18]

      C. Debiemme-Chouvy, C. Wartelle, F.X. Sauvage, J. Phys. Chem. B 108 (2004) 18291–18296. doi: 10.1021/jp046977y

  19. [19]

      J. Gun, A.D. Modestov, A. Kamyshny, et al., Microchim. Acta 146 (2004) 229–237. doi: 10.1007/s00604-004-0179-5

  20. [20]

      S. Koike, K. Kawamura, Y. Kimura, et al., Free Radic. Biol. Med. 113 (2017) 355–362. doi: 10.1016/j.freeradbiomed.2017.10.346

  21. [21]

      M. Yang, J. Fan, J. Du, et al., Chem. Sci. 11 (2020) 5127–5141. doi: 10.1039/d0sc01482f

  22. [22]

      Y. Yao, J. Chen, C. Yan, et al., J. Am. Chem. Soc. 147 (2025) 46953–46965. doi: 10.1021/jacs.5c10749

  23. [23]

      C. Yan, C. Liu, B. Liu, et al., Nat. Aging 5 (2025) 2532–2545. doi: 10.1038/s43587-025-01001-1

  24. [24]

      Y. Huang, L. Wang, Q. Huang, et al., Chin. J. Chem. 43 (2025) 2603–2612. doi: 10.1002/cjoc.70160

  25. [25]

      M. Xian, Y. Wang, C. Du, et al., Acc. Chem. Res. 58 (2025) 2804–2814. doi: 10.1021/acs.accounts.5c00483

  26. [26]

      K.G. Fosnacht, M.D. Pluth, Chem. Rev. 124 (2024) 4124–4257. doi: 10.1021/acs.chemrev.3c00683

  27. [27]

      X. Zhang, Y. Yang, L. Zhang, et al., Talanta 268 (2024) 125293. doi: 10.1016/j.talanta.2023.125293

  28. [28]

      Y. Zhou, Z. Gu, C. Liu, et al., Angew. Chem. Int. Ed. 61 (2022) e202114504. doi: 10.1002/anie.202114504

  29. [29]

      X. Zhang, L. Zhang, M. Gao, et al., J. Hazard. Mater. 396 (2020) 122673. doi: 10.1016/j.jhazmat.2020.122673

  30. [30]

      X. Luo, C. Zhang, C. Yue, et al., Chem. Sci. 14 (2023) 14290–14301. doi: 10.1039/d3sc04937j

  31. [31]

      J. Franz, M. Oelschlegel, J.P. Zobel, et al., J. Am. Chem. Soc. (2024) 11272–11288.

  32. [32]

      J. Zhang, M. He, M. Huang, et al., Angew. Chem. Int. Ed. 64 (2025) e202519021. doi: 10.1002/anie.202519021

  33. [33]

      X. Tan, J. Zhou, L. Yang, et al., J. Am. Chem. Soc. 146 (2024) 30422–30433. doi: 10.1021/jacs.4c10266

  34. [34]

      Z. Zheng, J. Gao, R. Wang, et al., ACS Nano 17 (2023) 22060–22070. doi: 10.1021/acsnano.3c08767

  35. [35]

      Z. Lv, J. Wang, J. Xu, et al., Adv. Healthcare Mater. 14 (2025) 2500273. doi: 10.1002/adhm.202500273

  36. [36]

      C. Liu, W. Chen, W. Shi, et al., J. Am. Chem. Soc. 136 (2014) 7257–7260. doi: 10.1021/ja502968x

  37. [37]

      J. Guo, S. Yang, C. Guo, et al., Anal. Chem. 90 (2017) 881–887.

  38. [38]

      W. Chen, E.W. Rosser, T. Matsunaga, et al., Angew. Chem. Int. Ed. 54 (2015) 13961–13965. doi: 10.1002/anie.201506887

  39. [39]

      B. Hu, Q. Liu, Y. Jiang, et al., Angew. Chem. Int. Ed. 64 (2025) e202418378. doi: 10.1002/anie.202418378

  40. [40]

      L. Sun, Z. Zheng, Z. Zhang, et al., Anal. Chem. 96 (2024) 20049–20056. doi: 10.1021/acs.analchem.4c04848

  41. [41]

      C. Zhao, X. Yang, Chin. J. Org. Chem. 45 (2025) 1419–1420. doi: 10.2147/copd.s503625

  42. [42]

      R.R. Zhang, X.Y. Ran, K.K. Yu, et al., Adv. Mater. 37 (2025) 2415891. doi: 10.1002/adma.202415891

  43. [43]

      W. Pan, M. Rafiq, W. Haider, et al., Coord. Chem. Rev. 514 (2024) 215907. doi: 10.1016/j.ccr.2024.215907

  44. [44]

      Y. Xu, C. Li, R. Xu, et al., Chem. Sci. 11 (2020) 8157–8166. doi: 10.1039/d0sc03160g

  45. [45]

      J. Huang, K. Pu, Angew. Chem. Int. Ed. 59 (2020) 11717–11731. doi: 10.1002/anie.202001783

  46. [46]

      B. Li, H. Liu, M. Zhao, et al., Angew. Chem. Int. Ed. 63 (2024) e202404093. doi: 10.1002/anie.202404093

  47. [47]

      Y. Zhang, S. Lei, Y. Pan, et al., Chin. Chem. Lett. 37 (2026) 110977. doi: 10.1016/j.cclet.2025.110977

  48. [48]

      H. Liu, Q. Sun, M. Dong, et al., Chin. Chem. Lett. 37 (2026) 111655. doi: 10.1016/j.cclet.2025.111655

  49. [49]

      K. Buscher, H. Wang, X. Zhang, et al., Nat. Commun. 7 (2016) 10828. doi: 10.1038/ncomms10828

  50. [50]

      Y. He, X. Wang, S. He, et al., ACS Nano 19 (2025) 6977–6992. doi: 10.1021/acsnano.4c14689

  51. [51]

      F. Di Lorenzo, K.A. Duda, R. Lanzetta, et al., Chem. Rev. 122 (2022) 15767–15821. doi: 10.1021/acs.chemrev.0c01321

• 

  Figure 1  Time-dependent absorption changes of 10 µmol/L H₂S₂–1 (A), H₂S₂–2 (B), H₂S₂–3 (C), and H₂S₂–4 (D) in the presence of 50 µmol/L Na₂S₂.

  下载: 全尺寸图片 幻灯片
  

  Scheme 1  Schematic illustrating the dual ratiometric NIR-Ⅱ FL/PA probe for H₂S₂ with unprecedented specificity based on electronic tuning strategy.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Time-dependent (A) absorption changes and (B) emission changes of 10 µmol/L H₂S₂–4 in the presence of 80 µmol/L Na₂S₂, λ_ex = 808 nm. (C) Absorption changes of H₂S₂–4 in the presence of relevant reactive species. (D) The fluorescence selectivity of H₂S₂–4 toward different analytes in PBS. H₂O₂ (200 mmol/L), GSH (1 mmol/L), Hcy (100 mmol/L), Cys (60 mmol/L), Na₂S (80 µmol/L), Na₂S₄ (80 µmol/L), S₈ (80 µmol/L), Na₂S₂ (80 µmol/L) the concentrations of all other substances were 100 µmol/L, and the concentrations of all analytes mentioned later are consistent with those here. (E) Linear relationship between the absorption intensity ratio (A₈₄₀/A₉₆₀) and Na₂S₂ concentration. (F) Linear relationship between the fluorescence intensity ratio (I₈₉₀/I₁₀₂₀) and Na₂S₂ concentration. (G) Linear relationship between the PA intensity ratio (PA₈₄₀/PA₉₆₀) and Na₂S₂ concentration, illustration: PA images of H₂S₂–4 solution responding with different Na₂S₂. (H) The PA selectivity of H₂S₂–4 toward different analytes in PBS. (I) Fluorescence images of capillaries filled with H₂S₂–4 (100 µmol/L) (λ_ex = 808 nm (1100 mW/cm²)), 100 ms exposure time and a 1000 long pass (LP) filter, respectively, immersed in 1% intralipid with varying depth. (J) SBR for fluorescence imaging of H₂S₂–4 as a function of the intralipid depth in (I). (K) Fluorescence responses of the probe H₂S₂–4 (10 µmol/L) in the absence and presence of Na₂S₂ (80 µmol/L) under different pH values. Data are the mean ± SD (n = 3).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A) Time-dependent NIR-Ⅱ ratiometric fluorescence imaging of probe H₂S₂–4 (100 µL, 100 µmol/L) in the absence of Na₂S₂. (B) Time-dependent NIR-Ⅱ ratiometric fluorescence imaging of probe H₂S₂–4 in the presence of Na₂S₂. (C, D) Time-dependent PA ratiometric imaging of probe H₂S₂–4 in the absence and presence of Na₂S₂. Data are the mean ± SD (n = 3).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  NIR-Ⅱ ratiometric fluorescence imaging of the probe H₂S₂–4 (100 µmol/L, 100 µL) in the absence (A) and presence (B) of Na₂S₂ with the filters of 850 nm LP, and 1000 nm LP, respectively. PA ratiometric imaging of the probe H₂S₂–4 in the absence (C) and presence (D) of Na₂S₂. (E) NIR-Ⅱ fluorescence ratio (I₈₅₀/I₁₀₀₀) corresponding to (A, B). (F) PA signal ratio (PA₈₄₀/PA₉₆₀) corresponding to (C, D). Note that the ratiometric signals (I₈₅₀/I₁₀₀₀ and PA₈₄₀/PA₉₆₀) reflect relative changes in H₂S₂ levels. Quantitative analysis of the signal changes in different groups. Data are the mean ± SD (n = 3). Statistical analysis was performed using one-way ANOVA followed by multiple comparison tests. *P < 0.05, **P < 0.01, ***P < 0.001.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  NIR-Ⅱ ratiometric fluorescence imaging of the probe H₂S₂–4 (100 µmol/L, 100 µL) in the absence (A) and presence (B) of E. coli induction and (C) ciprofloxacin treatment in mice. PA ratiometric imaging of the probe H₂S₂–4 in the absence (D) and presence (E) of E. coli induction and (F) ciprofloxacin treatment. (G) NIR-Ⅱ fluorescence ratiometric signal (I₈₅₀/I₁₀₀₀) obtained from (A-C). (H) PA ratiometric signal (PA₈₄₀/PA₉₆₀) obtained from (D-F). Quantitative analysis of the signal changes in different groups. Data are the mean ± SD (n = 3). Statistical analysis was performed using one-way ANOVA followed by multiple comparison tests. *P < 0.05, **P < 0.01, ***P < 0.001.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (A) Schematic of E. coli-induced bacterial infections of the hindlimbs model in mice. NIR-Ⅱ ratiometric fluorescence imaging of the probe H₂S₂–4 (100 µmol/L, 100 µL) in the absence (B) and presence (C) of E. coli induction and (D) ciprofloxacin treatment in mice. (E) Corresponding FL intensity ratio (I₈₅₀/I₁₀₀₀). (F) PA ratiometric imaging of the probe H₂S₂–4 in the absence (right) and presence (left) of E. coli induction and ciprofloxacin treatment (G). (H) Corresponding PA intensity ratio (PA₈₄₀/PA₉₆₀). Quantitative analysis of the signal changes in different groups. Data are the mean ± SD (n = 3). Statistical analysis was performed using one-way ANOVA followed by multiple comparison tests. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

  下载: 全尺寸图片 幻灯片
•