Hydrogen sulfide (H
2S), one of the most essential gasotransmitters [
1], is primarily generated by the metabolic pathways-regulated enzymatic processes [
2]. Endogenous H
2S participates in important activities and plays an important role in many physiological regulatory systems [
3], such as resisting inflammation [
4], reducing oxidative stress [
5], increasing cell proliferation [
6], and promoting angiogenesis [
7]. Particularly in the skin, programmed H
2S generation have been implicated in the pathogenesis of several skin diseases [
2], including diabetic wound healing [
7–
10], psoriasis [
11], and melanoma [
12]. In diabetic wound, the healing process is significantly impaired, leading to substantial side-effects in patients [
13,
14]. The impaired wound healing in diabetes is related to sustained inflammation, excessive oxidative stress, and impaired angiogenesis due to chronic hyperglycemia [
3,
15–
17]. Thus, antioxidants, reactive nitrogen species (RNS) scavengers, and their promoters are promising in improving diabetic wound healing. As an efficient scavenger for reactive oxygen species (ROS) and RNS in biological systems, H
2S has great potential in improve diabetic wound healing [
18]. Beyond that, H
2S can significantly inhibit tumor necrosis factor-
α (TNF-
α)-induced up-regulation of inducible nitric oxide synthase (iNOS), interleukin-6 (IL-6), and IL-8 in a dose dependent manner,
via suppression of p38 mitogen-activated protein kinase (p38 MAPK), extracellular signal-regulated kinase (ERK), and nuclear factor kappa-B (NF-
κB) pathways [
2,
19,
20]. Moreover, NETosis, induced by diabetes and impairs wound healing, can be inhibited by H
2S [
21,
22], meanwhile, H
2S synthase expression and H
2S content are significantly reduced in diabetic wounds [
23]. Thus, H
2S is an ideal therapeutic agent for diabetic wounds with the combined effects of anti-inflammatory and pro-healing. However, the biological effects of H
2S follow a dose-dependent pattern: low concentrations have cytoprotective effects, and high concentrations are cytotoxic [
24]. Thus, precisely controlled release and accurately monitor of H
2S is vital in promoting the related therapeutic efficacy.
H
2S donors offer an effective way to deliver prolonged H
2S with desired therapeutic concentration [
25]. Controlled releasing strategies for H
2S are generally based on polymers, hydrogels, nanoparticles and organic H
2S donors [
26–
34]. Organic H
2S donors that can trigger the release of H
2S by desire have attracted growing interest, owing to their feasibly tuned structures and robust biocompatibility. Carbonyl sulfide (COS), which can be efficiently hydrolyzed to H
2S by carbonic anhydrase (CA), has been widely used as an H
2S donor [
35]. Recently, a series of COS-based H
2S donors that can be triggered by biothiols [
36–
40], enzymes [
41,
42], ROS [
43–
45], pH [
34], and light [
46–
50] have been developed for controlled release of H
2S. Among them, photo-triggered donors have distinct advantages for robust spatiotemporal resolution, on-demand therapeutic modulation, non-invasive operation, programmability,
etc․ [
48]. However, the photo-triggers are usually UV-sensitive, including
o-nitrobenzyl (ONB) [
46], ketoprofanate [
51], or xanthone [
52]. Moreover, many triggers may produce by-products after photolysis, such as nitrosobenzaldehyde, the by-product of ONB (
Scheme 1), limiting their application
in vivo [
53,
54]. Furthermore, some photo-activated H
2S donors typically undergo a 1,6-elimination, leading to the production of 1,4-quinone methide, a highly toxic pro-oxidant that may negate the antioxidative effects of H
2S [
25]. Therefore, novel photo-triggered H
2S donors with precise spatiotemporal photo-sensitivity and robust efficiency are essential in the controlled releasing of H
2S for therapeutic interventions of many related diseases.
For active species like H
2S donors, monitoring the release of H
2S
in vivo by non-destructive methods are crucial in their dose-sensitive therapeutical process. Fluorescence imaging is one of the most feasible approaches for tracking biomolecules
in vivo. However, tracking H
2S by fluorescence always requires reactive fluorescent probes that consumes H
2S [
55]. Moreover, the potential interference from other reactive sulfur species may lead to the depletion of these fluorescent probes [
48]. Thus, H
2S donors that can simultaneously release H
2S and a fluorescent marker in a quantitative manner offer an efficient photo-triggered release of H
2S with accurate indicator for reaction progress in the complex biological environments [
25,
55].
Herein, we proposed a novel strategy for visible light-triggered releasing of H
2S based on 1,4,2-dioxazole-5-thione donors, a group of novel donors, to simultaneously release COS and bright fluorescence quantitatively. In the designed strategy, 1,4,2-dioxazole-5-thione undergoes photo-induced decomposition for COS (
Scheme 1), leading to highly reactive
N-acyl nitrene intermediates
via the elimination of COS, followed by
N-acyl nitrene rearranges to isocyanates [
56–
58]. Readily hydrolysis of isocyanate to amine triggered the release of fluorescence from the attached fluorescent scaffolds [
59,
60]. A series of photo-triggered H
2S donors, HSDs (HSD1, HSD2, and HSD3), were designed by connecting 1,4,2-dioxazole-5-thione and fluorophores with tuned spectra (
Fig. 1). The electron-withdrawing group, 1,4,2-dioxazole-5-thione, acts as both the sulfur source and the fluorescence caging group, locking the fluorescence. Through density functional theory (DFT) calculations, it is possible that the HSD1 molecule may cause fluorescence quenching through the photo-induced electron transfer (PET) effect (Fig. S18 in Supporting information). Upon photo-irradiation, the H
2S donor undergoes an efficient cleavage to COS and amine without by-products, accompanied by strong turn-on fluorescence, ensuring precise monitoring of the concentration of H
2S in real time. The probes exhibited efficient H
2S delivery and ROS scavenging
in vitro and
in vivo. Furthermore, the promoted wound healing by HSD3 has been validated in diabetic wound, demonstrating the great potential of photo-controlled release with precise tracking in therapeutic application. Overall, our H
2S donors HSDs possess the remarkable advantages of visible-light activation, no electrophilic by-products, and real-time visualization of H
2S release.
To confirm the photo-activities of 1,4,2-dioxazole-5-thione as COS/H
2S donors, we prepared HSD1, HSD2, and HSD3 by treating the corresponding
N-hydroxybenzamide with thiophosgene (
Fig. 1). Carboxylic acid derivatives were prepared following the reported procedures in literature [
61]. Briefly, carboxylic acid derivatives (1) undergo acylation reactions with oxalyl chloride to the crude acyl chloride (2). Further, the acyl chloride was added to a solution of triethylamine and hydroxylamine hydrochloride in dichloromethane to
N-hydroxybenzamide (3). Finally, the COS/H
2S donor was obtained by introducing thiophosgene and triethylamine to the reaction mixture. The H
2S donors with coumarin (HSD1), rhodamine (HSD2), and Si-substituted rhodamine (HSD3) as fluorophores are synthesized by this method (~5% yield). The structures of intermediates and final products were confirmed by
1H nuclear magnetic resonance spectroscopy (NMR),
13C NMR, and high-resolution mass spectrometry (HR-MS) analysis (Supporting information).
Photolysis of the probes was evaluated in phosphate buffer (PBS, 20 µmol/L, pH 7.4). Exposure of HSD3 to 365 nm light emitting diode (LED) leading to gradual decreases of absorption with λ
abs = 300 nm and reciprocal increase at λ
abs = 615 nm (
Fig. 2a). The absorption at 615 nm increased linearly with the irradiation duration in the first 240 s, followed with a gradually level off, likely due to consumption of HSD3. The corresponding fluorescence (λ
em = 637 nm) significantly increased ~58 times in 360 s (
Fig. 2b), and the kinetics for absorption and emission are similar (insets in
Figs. 2a and b). The absorption/emission spectra resembled that of rhodamine derivative, which was confirmed to be Si-rhodamine by high performance liquid chromatography (HPLC) and HR-MS (
Figs. 2d and h). Photo-irradiation triggers the conversion of HSD3 (HPLC retention time,
TR = 24 min) to Si-rhodamine (
TR = 14.5 min,
Fig. 2d). When solution of HSD1 was exposed to 365 nm irradiation, the maximum of the absorption spectra (λ
abs) bathochromically shifted from 290 nm to 341 nm (Fig. S1 in Supporting information). Meanwhile, the fluorescence spectra with a maximum (λ
em) at 449 nm has strikingly increased to about 75 folds (Fig. S1). HPLC analysis confirmed that 365 nm-light triggers the conversion of HSD1 (
TR = 19.0 min) to coumarin (
TR = 10.4 min, Fig. S4 in Supporting information). Upon 365 nm-irradiation of HSD2 solution, sharp absorption and fluorescence appeared (λ
abs = 529 nm; λ
em = 555 nm), with approximate 73-fold enhancement in fluorescence within 360 s (Fig. S2 in Supporting information). HPLC analysis proves the photo-triggered conversion of HSD2 (
TR = 26.0 min) to
N, N-diethylrhodamine (
TR = 11.8 min, Fig. S5 in Supporting information), as confirmed by HR-MS (Supporting information).
Specially, HSD2 can be photolysis by irradiation of broad spectra range. LED of different wavelengths, including 405, 500, and 600 nm, can efficiently trigger the conversion of HSD2 (Fig. S2). The 405 nm LED was particularly efficient with a photolysis quantum yield of 9.57% (Table S1 in Supporting information). HSD2 with fast photolysis that can be triggered by 500 nm (
t1/2 = 240 s, Fig. S2) was used in the analysis
in vitro. Photolysis of HSD1 can also be mediated by LED of 405 nm (Fig. S1), and the quantum yield of photolysis is 1.80% (Table S1). Photolysis of HSD3 can be facilely mediated by irradiation of a broad range,
e.g., 405, 500, 600, and 650 nm (
Fig. 2c and Fig. S3 in Supporting information). Although the absorption of HSD3 is low in the visible range (
ɛ: ~2275 L mol
−1 cm
−1 at 600 nm), its photolysis is fast (
t1/2 = 270 s,
Fig. 2c), and the quantum yield of photolysis is 6.37% in 405 nm (Table S1). It is noteworthy that the low-molar absorptivity does not necessarily preclude photochemical activity of HSD3 [
62]. The potential interference in the photo-triggered releasing of fluorescence in HSDs were extensively analyzed. Different bioactive species, including metal ions, endogenous amino acids, ROS,
etc. have demonstrated unnoticeable interference to the designed HSDs, in comparison with photo-triggered robust fluorescence (
Fig. 2g and Figs. S15–S17 in Supporting information). Moreover, the photolysis of HSDs is adaptative to broad ranges of pH within 5.0–10.0, demonstrating robust stability and suitability for bioimaging
in vivo (
Fig. 2e and Figs. S12–S14 in Supporting information).
To evaluate the photo-activated H
2S delivery from the donors, colorimetric methylene blue (MB) assay was employed to monitor the H
2S production of HSDs (20 µmol/L) under 365 nm-photo-irradiation in PBS containing CA (25 µg/mL). Two obvious absorption bands at 670 and 745 nm indicate the H
2S-triggered formation of MB (Fig. S7 in Supporting information). The absorption at 670 nm was used to quantify H
2S by calibration curve. For HSDs, 7.43 µmol/L (37.1% release efficiency, HSD1), 8.38 µmol/L (41.9%, HSD2), 7.96 µmol/L (39.8%, HSD3) H
2S were released, respectively (Figs. S8–S10 in Supporting information). Compared with existing H
2S donors, HSDs achieve a higher H
2S release efficiency (39.8%) than most photo-triggered donors, whose efficiencies typically range from 10% to 30% [
46,
47,
49–
51,
54]. The HSD3 reached maximum H
2S releasing in 330 s-irradiation before decreasing with extended irradiation, likely due to the volatilization and oxidation of H
2S (
Fig. 2f) [
50]. Thus, we have confirmed that 1,4,2-dioxazole-5-thione can be an effective photocage group for photo-triggered releasing of H
2S donor and fluorescence in sensitively and quantitatively tracking the photo-triggered releasing of H
2S.
HSDs have demonstrated great advantage for application
in vivo with clean and efficient photolysis. The photoactivable releasing and real-time visualization of H
2S were evaluated in immortalized human keratinocytes cell line, HaCaT Upon photo-irradiation at 405 nm, strong fluorescence was released in HSD1. Similarly, photo-irradiation at 500 and 600 nm triggered strong fluorescence in HSD2 and HSD3, respectively (
Fig. 3a). The averaged cellular fluorescence intensities from cells stained by probes were proportional to the duration of photo-exposure until the consumption of probes (
Figs. 3b–d).
H
2S delivery was subsequently explored
in vivo. We firstly investigated the H
2S delivery of HSD1 in PBS, and observed significantly enhanced fluorescence after coincubation with 3′–methoxy–3-oxo-3
H-spiro[isobenzofuran-1,9′-xanthen]−6′-yl 2-(pyridin-2-yldisulfanyl) benzoate (WSP-1), a commercial probe for H
2S (Fig. S11 in Supporting information). Next, the ability of HSD1 to deliver H
2S in HaCaT cells and Human nasal epithelial cells (HNEpCs) were evaluated. Both types of cells were incubated with HSD1 and WSP-1 without photo-activation, negligible fluorescence was detected in the red or green channels (
Fig. 4a, top panel and Fig. S25 in Supporting information). Strong fluorescence was observed in both channels after photo-activation, validating the H
2S delivery by HSD1 in live cells (
Fig. 4a, second panel and Fig. S25). After pre-incubation with acetazolamid (AAA, 2.5 µmol/L), an indiscriminate CA inhibitor [
63], green fluorescence remained unchanged, while the red channel was significantly reduced (
Fig. 4a, third panel,
Fig. 4c and Fig. S25), increased the concentration of inhibitor to 5.0 µmol/L, and the fluorescence of red channel was nearly fully quenched compared to hv group (
Fig. 4a, lower panel and
Fig. 4c), indicating the H
2S production through the hydrolysis of COS by CA.
H
2S can attenuate cellular oxidative stress by scavenging ROS. Based on the capability of H
2S-delivery and admirable results of cytotoxicity and stability of HSDs (
Fig. 4b and Figs. S22–S24 in Supporting information), the ROS scavenging was further evaluated. Phorbol 12-myristate 13-acetate (PMA) is widely used to stimulate endogenous ROS [
44]. Intracellular ROS levels were evaluated by a commercial ROS probe, 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) [
64]. The ability of HSD1 to scavenge ROS in HaCaT and HNEpC cells were evaluated. Both types of cells pre-incubated with PMA yielded strong green fluorescence compared to the control group, demonstrating that PMA significantly increased ROS levels. However, for cells pre-incubated with PMA, incubation with photoactivated HSD1 (20 µmol/L, 30 min) led to profoundly reduced green fluorescence (
Fig. 5a and Fig. S26 in Supporting information), confirming the ROS scavenging of HSD1. HSD1 without photoactivation had unnoticeable effect on the resulted ROS levels. To assess whether light exposure itself affects ROS levels, we irradiated cells after a 30 min PMA incubation. The resulting green fluorescence intensity showed minimal change, indicating that ROS generation is independent of light exposure (
Figs. 5a and b). To further confirm that the antioxidant effect of HSD1 is mediated by H
2S release, we treated cells with CA inhibitor, PMA and HSD1, followed by light irradiation. This group exhibited significantly stronger green fluorescence than the "HSD1 + hv" group, indicating the antioxidant effect of HSD1 is indeed mediated by the release of H
2S. These results revealed that photoactivated HSD1 can attenuate cellular oxidative stress, likely by scavenging ROS.
We also investigated the anti-inflammation of HSDs. Macrophage RAW 264.7 cells were incubation with lipopolysaccharide (LPS, 1.0 µg/mL) to trigger the inflammation. The inflammatory process activates the NF-
κB pathway, resulting in the increased production of nitrite (NO), TNF-
α, and IL-6 [
32,
48]. Given that the red emission of HSD3, featuring enhanced tissue penetration capability and lower biological background interference, it was chosen as the model donor for anti-inflammatory assays. Then, different concentration of HSD3 (0, 5, 10, 20 µmol/L) were added before photo-activation. Ultra-high concentrations of HSD3 over 20 µmol/L was not employed due to the potential of cytotoxicity. The HSD3 treated groups showed a dose-dependent inhibition of NO, TNF-
α, and IL-6 (
Figs. 5c–e and Figs. S27–S29 in Supporting information). Overall, we confirmed that HSD3 releases COS/H
2S in complex cellular environment and exhibits promising anti-inflammatory effect.
We further investigated the anti-inflammatory and antioxidant effects of HSDs in the wound healing process
in vivo (
Fig. 6). C57BL/6J mice were administered streptozotocin (STZ) by intraperitoneal injection to selectively damage the pancreatic
β cells irreversibly. All animal experiments were approved by the Animal Ethics Committee of East China Normal University. After continuous injection of STZ for 3 days, the blood glucose elevation in the treated mice were ≥16.7 mmol/L [
65], and the diabetic model was successfully constructed (
Fig. 6d). Then, a 5 mm diametral full-thickness wound was created on the back of diabetic mice (
Fig. 6a), following the previously described method [
48]. HSD3 with robust NIR fluorescent response was selected for analysis
in vivo. Since HSD3 can be efficiently activated by light of 365 and 600 nm, we selected 365 and 600 nm LEDs to activate HSD3 and evaluate the wound healing process. An aqueous solution of sodium alginate (SA) gel containing HSD3 (100 µmol/L) was applied to the wound in the working group. Simvastatin was used as the positive control for improving angiogenesis in wound healing [
66]. The wound healing progress was analyzed during the 9-day treatment.
We monitored the wound-healing processes in non-diabetic mice (Con 1), diabetic mice with wound treated with SA gel (Con 2), HSD3@SA gel with 365 or 600 nm photo-irradiation (HSD3+365, HSD3+600), simvastatin, photo-irradiation only (365 or 600 nm LED, 365, 600), and HSD3@SA gel without photo-irradiation (HSD3), respectively. Compared to mice in control group (Con 1), diabetic mice (Con 2) showed a delayed wound healing process with the SA gel treatment, also confirmed the success establishment of diabetic chronic wound model. Compared with diabetic mice (Con 2), the working group (HSD3+365, HSD3+600) of HSD3@SA gel with 365 or 600 nm LED irradiation showed accelerated wound healing, and the promoted healing was comparable to the simvastatin group (
Figs. 6b and c, Figs. S34 and S35 in Supporting information). The LED-treated group (365, 600) or HSD3@SA gel without photo-irradiation (HSD3) has no obviously promoting effect, proving that the promoted wound healing is the result of photo-triggered reaction rather than the photo itself. These phenomena confirmed the efficient wound healing by a broad range of light-activated HSD3 in live animals.
On the 5
th and 9
th days, wound tissues were harvested from each group before histological analysis by hematoxylin-eosin (H&E) and Masson staining for the illustration of epidermal wound margins. Re-epithelialization and granulation tissue formation are the most important indicators of wound healing [
67]. On the 5
th day, H&E staining of the tissue showed that the fibroblast proliferation, granulation tissue formation, and keratinocytes migrate from the shoulder of the wound into the wound bed to promote the re-epithelialization. The wound areas of both the photoactivated HSD3 and the simvastatin groups were smaller than that of the control group, but larger than that of the normal mice, consistent with the pictures of wound in
Fig. 6b and Fig. S34 (Supporting information). On the 9
th day, photoactivated HSD3 and simvastatin treatment exhibited a rather continuous and thick epithelial layer, but incomplete and thin epithelial layer was found in other diabetic wounds (
Fig. 6e and Fig. S36 in Supporting information). During the wound healing process, a large amount of collagen is produced and deposited in granulation tissue, providing a good indicator for wound healing [
68]. The collagen deposition was evaluated by Masson staining (
Fig. 6f). The detailed enlarged picture showed that massive collagen deposition in the photoactivated HSD3 and the simvastatin treated wounds, indicating recovery and maturation from damaged tissues (
Fig. 6g). In general, both visual observation and histological examination confirmed that photo-activated HSD3 expedited the healing of infected diabetic wounds by enhancing granulation tissue formation, facilitating the formation of epidermis, and increasing collagen deposition. Further, the
in vivo fluorescence imaging of wounds with photo-triggered fluorescence have confirmed the capability of monitoring H
2S in live animals by the developed HSDs. The quantitative results of fluorescence intensity over time showed the fluorescence basically reaching a plateau within 10 min, indicating the H
2S released completely (Figs. S30-S33 in Supporting information). These results suggest that HSDs have great potential therapeutic application prospects in accelerating the healing of chronic wounds in the future.
We proposed a robust strategy for designing HSDs, turn-on COS/H2S donors, with efficient photo-triggered release of H2S while quantitatively monitoring the amount of H2S by fluorescence across a broad palette of spectra in vivo. The photo-triggered release of H2S is fast and electrophilic by-product free, and can be accurately detected and visualized in vitro and in vivo. Furthermore, the anti-inflammatory and antioxidant effects of controlled release of H2S can be applied in different therapeutic processes. We have confirmed that HSDs can efficiently promote wound healing by controlled release of H2S in diabetic wounds. Our design strategy has provided novel insights into precisely controlling and monitoring the release of reactive species, while leveraging the advantage of photo-triggered release, holding promising potential for advanced treatments in many diseases.
PDF
Full
55
34
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
