1. INTRODUCTION
Traumatic skin defects are associated with serious health problems and may deteriorate into lethal trauma without proper regeneration. Hence, the repair of skin tissue trauma is essential to decrease suffering and prolong the life expectancy of patients [1, 2]. Wound healing is a continuous, highly coordinated multi-tissue dynamic process involving three overlapping stages: inflammation, fibrous tissue hyperplasia, and matrix remodeling [3, 4]. Initially, the stress response triggers a severe inflammatory reaction in the body. Meanwhile, the overexpression of cytokines, such as C-reactive protein (CRP) and interleukin-6 (IL-6), and the formation of reactive oxygen species (ROS) free radicals is activated by the dysregulation of multiple intracellular signaling pathways, thus causing systemic inflammatory response syndrome [5]. A moderate inflammatory response protects wounds and promotes wound healing. However, an excessive inflammatory response can lead to immune dysfunction, thereby inhibiting wound healing or even causing life-threatening problems in patients [6, 7]. Hence, safe, effective, and feasible strategies must be designed to accelerate wound healing, with a focus on balancing the immune response and regulating the inflammatory pathway.
H2S is found in almost every living organ system, and plays a major role in diverse pathologies and disease states [8–12]. H2S has multiple functions, such as protecting the central nervous system, mitigating oxidative stress in the cardiovascular system, and modulating blood pressure homeostasis [13]. Moreover, H2S participates in anti-oxidant, anti-inflammatory, and cytoprotective effects through its direct interaction with major signaling pathways associated with various disease states [14–17]. Therefore, developing molecules with the ability to regulate H2S levels at wound sites may facilitate control of the inflammatory response and promote early wound healing. Furthermore, research and development on small-molecule H2S donors is crucial to understanding the mechanisms underlying wound healing and may facilitate optimization of interventional therapy.
Although substantial efforts have been dedicated to the design and application of activatable H2S donors [18–20], the visualization and accurate detection of released H2S in living systems remains a crucial challenge, owing to the inconvenience and insufficient sensitivity of many traditional analytical technologies, such as electrochemistry [21, 22], gas chromatography [23, 24], and colorimetry [25, 26]. In recent years, fluorescence imaging has attracted attention among researchers because of its high selectivity, high sensitivity, convenience, low invasiveness, and real-time imaging capability [27–40]. Excellent H2S release efficiency with long-wavelength fluorescence emission and the ability to visualize H2S release are prerequisites for competent H2S donors. Several H2S donors with intrinsic fluorescence signal changes have been reported, which are activated by various exogenous or endogenous stimuli, such as light [41–43], pH modulation [44, 45], enzymes [46, 47], cellular thiols [48–51], and ROS [52, 53], and subsequently release H2S. However, some have drawbacks, such as uncontrollable H2S release, low H2S release efficiency, and short-wavelength emission, which are unsuitable for applications in to treating diseases in living systems. Recently, our group has developed a Cys-triggered NIR fluorescent H2S donor [54] that is suitable for anti-inflammation applications in living cells. However, its main fluorescence products are not generated in the same pathway as H2S, and the observed fluorescence is not correlated with H2S concentration in some cases, thus preventing precise detection of H2S.
Herein, we developed a biothiol-activated NIR fluorescence H2S donor, PRO-ST ( Figure 1a ), for real-time visualization of H2S release. PRO-ST consists of a dicyanoisophorone-based NIR fluorescence moiety (TCOO), a biothiol-trigger moiety (4-isothiocyanate benzyl alcohol), and a sulfur-source group (thiophosgene) ( Figure 1b ). The optical response of PRO-ST under biothiol-induced cleavage was evaluated, and the reaction mechanism underlying H2S release was verified with the kinetics and release efficiency of H2S. The anti-inflammatory and wound healing abilities of PRO-ST were also studied in lipopolysaccharide (LPS)-induced inflammatory macrophages and in mice, respectively.

a) Schematic illustration of the design of PRO-ST for visualizing H2S release and promoting wound healing via inhibition of inflammation. b) Schematic diagram of the mechanism of biothiol-activated COS/H2S release with fluorescence response from PRO-ST. c) Synthesis route of PRO-ST. (i) thiophosgene, anhydrous CH2Cl2, room temperature, 0.5 h; (ii) thiophosgene, anhydrous THF, KOH, room temperature, 1 h; (iii) 4-dimethylaminopyridine (DMAP), anhydrous CH2Cl2, room temperature, 10 min.
2. RESULTS AND DISCUSSION
The synthesis route of PRO-ST is presented in Figure 1c . The compound (4-isothiocyanatophenyl)methanol was synthesized by stirring (4-aminophenyl)methanol 1 and thiophosgene in a dichloromethane solution. The middle component 3 was obtained from TCOO by caging the oxygen atom with a thiochloroformate group. Compound 3 was then reacted with 2 in the presence of DMAP to yield the desired H2S donor PRO-ST ( Figure S1–8 and Methods, Supporting Information).
To evaluate the fluorescence response induced by biothiol-triggered cleavage, we investigated the fluorescence spectral properties of PRO-ST before and after treatment with GSH, Cys, and Hcy. As shown in Figure 2a , PRO-ST had intrinsic hypofluorescence with a low fluorescence quantum yield (ΦF=0.008). This phenomenon was attributed to the strong electron-withdrawing property of the carbon-sulfur double bond in PRO-ST, which diminished the electron-donating ability of the aromatic hydroxy group and finally quenched the fluorescence of TCOO. After addition of GSH (20 equiv), the fluorescence intensity of PRO-ST at 666 nm increased by 45-fold over its initial state. In addition, a distinct Stokes shift of 136 nm was observed. Similarly, the addition of Cys and Hcy significantly enhanced the fluorescence of PRO-ST, by 37-fold and 17-fold, respectively. The activation of fluorescence responses might have resulted from the rapid nucleophilic attack on the isothiocyanate in PRO-ST by the -SH of biothiols, followed by intramolecular 1,6-elimination with the release of the fluorophore (TCOO) and COS ( Figure 1b ).

a) Fluorescence spectra of PRO-ST in the absence or presence of GSH/Cys/Hcy under aqueous DMF/PBS solutions (pH 7.4, 20 mM, v/v, 3/7). Inset: corresponding fluorescence imaging of PRO-ST after addition of GSH/Cys/Hcy, acquired at 666 nm with an IVIS spectrum imaging system. b) Fluorescence intensity of PRO-ST at 666 nm toward different substances (200 μM) including GSH, Cys, Hcy, H2S, BSA, amino acids, cations, and reactive oxygen and nitrogen species. c) Linear correlation between the fluorescence intensity of PRO-ST and various concentrations of GSH/Cys/Hcy. d) Time-dependent fluorescence intensity of PRO-ST after addition of GSH/Cys/Hcy. e) Confocal fluorescence microscopy images of RAW-246.7 cells incubated with PRO-ST: (first column) untreated cells; (second column) cells preincubated with NEM; (third-fifth column) NEM-preincubated cells treated with GSH/Cys/Hcy. f) Relative fluorescence intensity of the corresponding fluorescence images in e). Note: the concentration of PRO-ST, C-7AZ, and biothiols was 10 μM, 3 μM, and 200 μM, respectively, unless otherwise indicated.
To confirm this assumption, we performed mass spectrum analysis, which indicated that PRO-ST decomposed into the expected TCOO (m/z, 289.2), nucleophilic addition-elimination quinone product (m/z, 309.1), and thioester intermediate (m/z, 659.1) after being treated with NAC (a compound with an acetyl group attached to the amino group of cysteine) ( Figure S9 , Supporting Information). Moreover, the mass spectrum of PRO-ST revealed the same results: the common fluorophore TCOO and the corresponding nucleophilic addition-elimination quinone product (m/z, 453.1, 267.0, and 281.0) were generated after reaction with GSH, Cys, and Hcy, respectively ( Figure S10–12 , Supporting Information). To further demonstrate the role of -SH in biological thiols, as control compounds, we selected β-mercaptoethanol, NAC, cystine, and cysteamine, whose molecular structures are similar to those of GSH, Cys, and Hcy. The compounds containing an -SH group activated PRO-ST and produced a fluorescence signal, whereas the compounds containing an amino group without -SH did not ( Figure S13 , Supporting Information), thus indicating that -SH, rather than the amino group, participates in nucleophilic attack on isothiocyanates. Therefore, fluorescence signals of PRO-ST are activated by the -SH group in biological thiols through preferential attack on the isothiocyanate via nucleophilic addition, in agreement with our hypothesized mechanism.
The selectivity of PRO-ST toward biothiols (GSH, Cys, and Hcy) was investigated through testing various biologically relevant species, such as H2S, BSA, amino acids (Arg, Gly, Val, Ser, Leu, Asn, Asp, His, Pro, Tyr, Thr, Glu, Lys, Phe, Met, and Ile), metal ions (Ca2+, Na+, Mg2+, K+, Zn2+, Cu2+, Al3+, Fe3+ Hg2+, and Fe2+), and reactive oxygen and nitrogen species (1O2, O2 .−, H2O2, ClO−, ·OH, ONOO−, and NO2 −) ( Figure 2b and S14 Supporting Information). None of them induced significant fluorescence changes in PRO-ST, thus indicating the high selectivity of PRO-ST toward biothiols. The changes in the fluorescence spectra of PRO-ST toward biothiols enabled quantification of the signals on the basis of the fluorescence intensity at 666 nm ( Figure S15–17 , Supporting Information). Good linear correlations between the fluorescence signal and the concentration were observed for GSH, Cys, and Hcy, with a limit of detection of 28.1, 41.4, and 33.2 nM, respectively ( Figure 2c ). The kinetics of GSH-, Cys-, and Hcy-induced optical changes in PRO-ST were also investigated ( Figure 2d ), and the fluorescence intensity of PRO-ST at 666 nm reached a plateau in approximately 60 min, thus suggesting that PRO-ST was activated by biothiols slowly. These data demonstrated that PRO-ST has a specific and controllable fluorescence response toward biothiols.
To verify whether biothiols in cells activate PRO-ST and release the fluorophore TCOO, we treated murine macrophage (RAW 264.7) cells with PRO-ST and performed imaging. Before the bioimaging experiment, we performed cell viability studies with PRO-ST and its reaction product TCOO. When the concentration of PRO-ST and TCOO increased to 50 μM, the survival rate of RAW 264.7 cells still exceeded 75% ( Figure S18 , Supporting Information). After incubation with PRO-ST, the cytoplasm of the cells clearly exhibited red fluorescence. However, when pretreated with N-ethylmaleimide (NEM), an effective sulfhydryl scavenger [55], the cytoplasm of the cells displayed faint red fluorescence. Further quantification demonstrated that the fluorescence intensity of NEM-preincubated cells was 9% lower than that of cells without preincubation ( Figure 2f ). In addition, when NEM-pretreated RAW 246.7 cells were cultured with exogenous GSH, Cys, and Hcy, recovery of strong fluorescence signals was observed. These results suggested that the intense fluorescence in RAW 246.7 cell cytoplasm was induced by the intracellular biothiols, which triggered the activation of PRO-ST. Moreover, the biocompatibility of PRO-ST in vivo was confirmed in zebrafish and mouse imaging experiments ( Figure S19–20 , Supporting Information).
As shown in Scheme 1A, COS, released from PRO-ST in the presence of biothiols was converted to H2S by carbonic anhydrase (CA) [56]. To verify the underlying mechanism, we used C-7AZ, an H2S-responsive fluorescence probe [57], to detect H2S ( Figure 3a and S21 , Supporting Information). As shown in Figure 3a , when C-7AZ was added to the reaction solution, no significant difference in fluorescence intensity at 450 nm was observed in the absence of CA with respect to the control group (λ ex=350 nm). However, the fluorescence intensity of the GSH group in the presence of CA was clearly enhanced, to a level approximately 10-fold higher than that of the control group. Similarly, the fluorescence intensity of the Cys and Hcy groups was 8- and 5-fold high than that of the control group, respectively. These results demonstrated that the release of H2S from PRO-ST depended on the generation of COS and the catalysis of CA.

a) Fluorescence intensity of C-7AZ in the presence of PRO-ST and biothiols, with or without CA, λ ex=350 nm, λ em=450 nm. b) Time-dependent absorption intensity of MB reacted with PRO-ST (25 μM), GSH/Cys/Hcy (250 μM), and CA (25 μg/mL), λ abs=667 nm. c) Confocal fluorescence microscopy images of RAW 246.7 cells. (I) Cells treated with PRO-ST. (II) Cells treated with C-7AZ; (III) Cells treated with PRO-ST and C-7AZ; (IV) NEM-pretreated cells incubated with PRO-S and C-7AZ; (V–VII) NEM-pretreated cells incubated with GSH, Cys, and Hcy, then incubated with PRO-ST and C-7AZ sequentially. (VIII) Cells were incubated with AAA (2.5 μM), PRO-ST, and C-7AZ sequentially. Blue channel: λ ex=405 nm, λ em=450 ± 20 nm; red channel: λ ex=514 nm, λ em=660 ± 30 nm. d) Relative fluorescence intensity of the corresponding fluorescence images in c). Effects of PRO-ST on nitrite (e), TNF-α (f), and IL-6 (g) in LPS-pretreated RAW264.7 cells. Results are expressed as mean, SD (n=3), ###P < 0.001 vs the 0 μM PRO-ST group, ***P < 0.001 vs the control group. h) PRO-ST suppresses NF-κb transactivation. Results are expressed as mean, SD (n=3), ***P < 0.001 vs the PRO-ST group, nsNo statistically significant difference. Note: The concentration of PRO-ST, C-7AZ, and biothiols was 10 μM, 3 μM, and 200 μM, respectively, unless otherwise indicated.
To investigate the release efficiency of H2S from PRO-ST triggered by biothiols, we performed a kinetic study using methylene blue (MB) assays [58]. At 667 nm, the absorption intensity of MB solution increased gradually in the presence of CA and reached a plateau approximately 2 h after incubation of PRO-ST with GSH ( Figure 3b ). Moreover, 18.2 μM H2S showed 73% release efficiency, according to the calibration curve between MB and H2S ( Figure S22 and S23 , Supporting Information). Cys and Hcy also triggered PRO-ST release of H2S (14.1 and 9.0 μM) under the same conditions, with a moderate release efficiency of 57% and 36%, respectively ( Figure S24 and S25 , Supporting Information). Therefore, these results indicated that PRO-ST may be a promising candidate for the highly efficient and controllable release of H2S.
To assess the ability of PRO-ST to release H2S in the cellular environment, we introduced C-7AZ to verify the release of H2S [57], as validated by fluorescence enhancement in the blue channel. As shown in Figures 3c and 3d , RAW 246.7 cells treated with only C-7AZ showed slight fluorescence in the blue channel, possibly because of endogenous H2S (column II). However, after being treated with PRO-ST and C-7AZ, both the red channel and blue channel showed a clear enhancement in fluorescence signals (column III). Moreover, after preincubation with the potent sulfhydryl scavenger NEM, the fluorescence of the cells in both channels decreased significantly when PRO-ST and C-7AZ were added successively (column IV). When the NEM-pretreated cells were incubated with exogenous GSH, Cys, and Hcy, the fluorescence of the cells in both channels was restored (columns V, VI, and VII, respectively). These results suggested that biothiols were a major factor responsible for the changes in cellular fluorescence signals. To validate the requirement for CA in the conversion from COS to H2S [56], we pretreated RAW 246.7 cells with acetazolamide (AAA, a CA inhibitor [59]), PRO-ST, and C-7AZ successively. As shown in Figure 3c (VIII column), negligible fluorescence signals were observed only in the blue channels, in agreement with the previous results obtained in an aqueous solution ( Figure 3a ). These data thus confirmed the suitability of PRO-ST for controllable and visualized H2S release in cells.
Previous studies have verified that H2S has anti-inflammatory effects at physiological concentrations. In addition to the mechanism underlying the release of H2S, the anti-inflammatory effect of PRO-ST in RAW264.7 cells was further evaluated. LPS, a known endotoxin [60], was used to establish a cellular inflammation model. The introduction of LPS results in upregulation of nuclear factor kappa-B (NF-κb), a pleiotropic transcription factor associated with inflammation, through intracellular signaling pathways [61]. Activated NF-κb subsequently induces the regulation of various cytokine inflammatory mediators, including TNF-α, NO2 −, and the cytokine IL-6 [62]. Appropriate concentrations of H2S effectively inhibit the activation of the NF-κb signaling pathway and regulate the excessive release of inflammatory mediators [15]. Therefore, the anti-inflammatory effects of PRO-ST were evaluated on the basis of the accumulation of NO2 −, TNF-α, and IL-6 in LPS-exposed RAW 264.7 cells, as measured with a Griess Reagent Kit and ELISA kit. As shown in Figure 3e–g , the LPS-exposed RAW264.7 cells showed significantly higher levels of NO2 −, TNF-α, and IL-6 than the control group. When the LPS-incubated cells were treated with PRO-ST, accumulation of NO2 −, TNF-α, and IL-6 was significantly impaired, thus indicating good anti-inflammatory performance of PRO-ST. Similar inhibitory effects were also observed in the positive control group (treated with dexamethasone, an effective anti-inflammatory drug). Notably, the LPS-pretreated cells incubated with the byproduct TCOO did not show inhibition of LPS-induced production of NO2 −, TNF-α, and IL-6, thus indicating that H2S released from PRO-ST played a crucial role in the anti-inflammatory effect.
The relationship between the accumulation of inflammatory mediators and NF-κb was also evaluated. A 96-well ELISA kit was used for measuring the NF-κb DNA binding activity. As shown in Figure 3h , LPS-exposed cells displayed strongly upregulated nuclear translocation of NF-κb, whereas exposure to PRO-ST alone scarcely affected the translocation of NF-κb with respect to that in the PBS group. Moreover, PRO-ST significantly suppressed LPS-induced NF-κb transactivation, thus indicating that the anti-inflammatory effect of PRO-ST may be achieved by blocking NF-κb transactivation.
A wealth of evidence has suggested that wound healing involves complex processes such as inflammation, proliferation, and remodeling [63]. In chronic wounds, decreasing the inflammatory response is considered to effectively promote wound healing [64]. H2S has been found to inhibit the excessive release of inflammatory mediators [13]. Hence, we evaluated the wound healing effects of PRO-ST in a murine full-thickness cutaneous wound model by periodically recording the conditions of wound contraction ( Figure 4a ). To dress the wounds, we gelled PRO-ST with sodium alginate, a natural polysaccharide used as an excipient for pharmaceutical preparations with high stability, solubility, viscosity, and safety. As shown in Figure 4b , the wound shrank from the edge over time in all experimental groups, owing to the body’s natural self-healing capability. Moreover, in comparison with the control group treated with PBS buffer and the TCOO-treated group, mice with PRO-ST gel applied on the 3rd day of post-surgery showed significantly accelerated wound closure; the therapeutic effect was comparable to that in the chitin-treated group (positive control). After 7 days of treatment, the wound area was smaller ( Figure 4c ) and showed almost complete healing. In addition, fluorescence imaging experiments indicated a significant decrease in fluorescence intensity an increase in the number of treatment days ( Figures 4d and 4e ). Therefore, PRO-ST successfully enabled visualization of wound healing.

a) Schematic illustration of the overall experiment using PRO-ST sodium alginate gel therapy for cutaneous wounds. b) Representative images of full-thickness cutaneous wounds at different time intervals treated with PBS buffer, chitin gel, TCOO gel, or PRO-ST gel. c) Relative wound areas (% of day 0) of groups treated with PBS buffer, chitin gel, TCOO gel, or PRO-ST gel on days 1, 3, 5, 7, 9, and 11 after injury. Statistical analysis was performed with one-way ANOVA. d) Fluorescence imaging of cutaneous wounds in living mice, λ ex=514 nm. e) Relative fluorescence intensity of the corresponding fluorescence images in d). f) Hematoxylin and eosin (H&E)-stained sections of healthy skin and wound tissues on day 3 and day 11, treated with PBS buffer, chitin gel, TCOO gel, or PRO-ST gel.
Controlling inflammation is a crucial step in wound healing. We next performed hematoxylin and eosin (H&E) staining histological analyses ( Figure 4f ). On the 3rd day post-surgery, abundant inflammatory cell infiltration was found in the negative control group treated with PBS buffer and TCOO gel. The PRO-ST gel group, as well as the positive control group treated with chitin gel, exhibited less inflammatory cell infiltration than the negative control group treated with PBS, presumably because of the anti-inflammatory effect of H2S. Furthermore, on the 11th day post-surgery, an abundance of multinucleate giant cells and neo-epidermis was observed in the PRO-ST gel group, with histomorphology similar to that of normal skin; similar findings were also observed in the chitin gel group. In contrast, no significant improvement was observed in the negative control group. The histological results demonstrated that the H2S released from PRO-ST promotes re-epithelialization and epidermal repair occurring in the wound area.
3. CONCLUSION
In summary, we developed a biothiol-activated NIR fluorescent H2S donor, PRO-ST, for real-time visualization of H2S release. Compared with the previously reported donor (Table S1, Supporting Information), PRO-ST enables controllable release of H2S with higher efficiency (up to 73%). Furthermore, PRO-ST overcame the deficiencies of the previous H2S donor, which does not generate H2S through the same pathway as the main fluorescent product. More importantly, PRO-ST can be activated to release H2S by multiple biothiol stimulants, thus avoiding excessive consumption of a single stimulant and consequently maintaining the local redox balance. Additionally, PRO-ST promoted real-time anti-inflammation and in situ wound-healing in biological systems, and therefore has potential applications in the clinical treatment of skin trauma.