A highly selective and sensitive near-infrared fluorescent probe for imaging of hydrogen sulphide in living cells and mice

Hydrogen sulphide (H2S), the third endogenous gaseous signalling molecule, has attracted attention in biochemical research. The selective detection of H2S in living systems is essential for studying its functions. Fluorescence detection methods have become useful tools to explore the physiological roles of H2S because of their real-time and non-destructive characteristics. Herein we report a near-infrared fluorescent probe, NIR-HS, capable of tracking H2S in living organisms. With high sensitivity, good selectivity and low cytotoxicity, NIR-HS was able to recognize both the exogenous and endogenous H2S in living cells. More importantly, it realized the visualization of endogenous H2S generated in cells overexpressing cystathionine β-synthase (CBS), one of the enzymes responsible for producing endogenous H2S. The probe was also successfully applied to detect both the exogenous and endogenous H2S in living mice. The superior sensing properties of the probe render it a valuable research tool in the H2S-related medical research.

Scientific RepoRts | 6:18868 | DOI: 10.1038/srep18868 tissue penetration, and minimum interference from background autofluorescence 34 41 . Despite these progresses, we are interested in developing practical NIR probes with a combination of desirable characteristics, especially low detection limits and the ability to detect endogenous H 2 S in living animals. Fluorescent probes with a lower detection limit, especially the nanomole range, or even lower, are needed owing to the low levels of endogenous H 2 S in cells/plasma/tissues. Moreover, the probes for detection of endogenous H 2 S in vivo are still sparse 44,45 . Thus, NIR probes with higher sensitivity and favourable properties to monitor endogenous H 2 S in vivo are in high demand.
Herein, we prepared NIR-HS as a NIR fluorescent probe for H 2 S detection. The key features of NIR-HS include good selectivity, high sensitivity and suitability for recognizing endogenous H 2 S in living cells and mice.

Results and Discussion
Synthesis and sensing mechanism of NIR-HS. In the design of a NIR probe for H 2 S, the hemicyanine skeleton (a NIR dye) was selected as a fluorophore in the light of its NIR emission and high stability 46 . It is known that the thiolysis of the dinitrophenyl ether reaction can be chemoselective for H 2 S over biothiols 38 . Thus, probe NIR-HS was constructed by connecting a dinitrophenyl group to hemicyanine (a NIR dye) via an ether-linkage (Fig. 1). The fluorescence of NIR-HS was quenched due to alkylation on the hydroxyl group 46 . We speculated that the reaction of sulphide with NIR-HS would cleave an ether group, and release the free fluorophore, thereby achieving fluorescence detection of sulphide. On the basis of this design, the structure of NIR-HS and the proposed sensing mechanism are illustrated in Fig. 1. The probe was readily synthesised in two steps. Treatment of IR-780 with resorcin in the presence of K 2 CO 3 afforded compound 1, which was then condensed with 1-fluoro-2,4-dinitrobenzene to generate target compound NIR-HS (please refer to the Supplementary Information for details). Finally, target probe, NIR-HS, was characterized by NMR spectroscopy and mass spectrometry (please refer to the Supplementary Information for details).
The reaction of NIR-HS with sulphide produced a red fluorescent product, which was identical to the absorption and emission of the authentic compound 1 ( Supplementary Fig. S1). Moreover, the thiolysis product was characterised by HRMS and 1 H NMR spectra (please refer to the Supplementary Information for details), demonstrating that the reaction of NIR-HS with sulphide proceeded as designed in Fig. 1.

Fluorescent properties of NIR-HS.
The fluorescent properties of NIR-HS (10 μ M) in the absence and presence of Na 2 S were determined. The free probe was almost nonfluorescent ( Fig. 2A). However, treatment of Na 2 S (100 μ M) led to a large fluorescence enhancement at 723 m (50 fold, Φ = 0.13). Figure 2B depicted elevated fluorescence intensities with increasing amounts of Na 2 S (0-300 μ M) until a plateau reached at 100 μ M Na 2 S. An excellent linear correlation between the observed fluorescence intensities and various concentrations of Na 2 S (0-100 μ M) was observed in PBS buffer (Fig. 2B inset). The in vitro detection limit for sulphide was 38 nM, which was lower than most of the reported NIR H 2 S probes. Therefore, NIR-HS is highly sensitive to low-nanomolar levels of sulphide, which facilitate the quantitative detection of endogenous/intracellular H 2 S in complex biological systems.
The fluorescence intensity in reaction of NIR-HS with Na 2 S reached the maximum value within approximately 20 min (Supplementary Fig. S2). The effects of pH on the detection of sulphide were then evaluated ( Supplementary  Fig. S3). In the pH range from 5.8 to 6.0, the emission intensities were quite low and did not change significantly. From pH 6.2 to 6.8, the fluorecence intensities were gradually increased, and the maximal fluorecence intensities were observed from pH 7.0 to 9.0. The emission profile of fluorophore (compound 1) (Supplementary Fig. S4) are consistent with the results of treating the probe with Na 2 S in different pH PBS buffer, indicating that the observed pH profile is due to the fluorophore itself. Taken together, NIR-HS is suitable for the detection of sulphide between pH 7.0 and 9.0.
Selectivity to sulphide of NIR-HS. To investigate the selectivity of NIR-HS towards sulphide, NIR-HS was treated with various species. NIR-HS displayed high selectivity for H 2 S over physiological concentrations of   Next, we tested the abilities of NIR-HS to visualize the endogenous H 2 S. MCF-7 cells express H 2 S-producing enzyme such as CSE 47 . NO could upregulate the CSE expression and stimulate the CSE activity, resulting in increased endogenous H 2 S level 48 . Therefore, SNP (Sodium Nitroprusside, a NO donor) was employed to induce the production of endogenous H 2 S in MCF-7 cells. The probe-loaded cells exhibited faint fluorescence emission without the addition of SNP (Fig. 5, panel 1A). After incubation of probe-treated cells with SNP (Fig. 5, panel 1B) for another 20 min, the fluorescence signal increased significantly, indicating the generation of endogenous H 2 S within the cells. Whereas the cells preincubated with DL-propargylglycine (PPG, an inhibitor for CSE 48 ) provided almost no fluorescence enhancement (Fig. 5, panel 1C), demonstrating that the fluorescence change is triggered by endogenously generated H 2 S. CSE and CBS are major enzymes for H 2 S production, and the overexpression of CBS or CSE could result in the elevation of endogenous H 2 S level 1,2 . We thus constructed the cells with CBS overexpression (Fig. 5, panel 2A). Cells transfected with empty vector (pCMV6) were set as control group (Fig. 5, panel 2B). As shown in Fig. 5, MCF-7 cells that were overexpressing CBS showed much stronger fluorescence (Fig. 5, panel 2A) than that from cells of the control group (Fig. 5, panel 2B), suggesting the increased endogenous level of H 2 S in CBS overexpressed cells. The western blot assay proved the overexpression of CBS in cells of panel 2A (Supplementary Fig. S12). We also quantified the fluorescence intensities of these cells, and found that the CBS overexpressed cells showed 2-fold enhanced fluorescence intensity compared to the control cells (Fig. 5, panel 2C). These results revealed the capability of NIR-HS to recognize endogenous H 2 S in living cells.
Detection of H 2 S in living mice. The prominent NIR features of NIR-HS render the probe highly favorable for fluorescence imaging of H 2 S in living animals. Inspired by these data, we further examined the suitability of the sensor to visualize exogenous and endogenous H 2 S in living mice. Kunming mice were divided into several groups. The mice were given i.p. injection of DMSO as the negative control group (Supplementary Fig. S13, panel A), and the mice were given i.p. injection of free probe as the probe-loaded group (Fig. 6, panels A). One group were pretreated with ZnCl 2 , and then injected with free probe (Supplementary Fig. S13, panel B). The other three groups were injected with different amounts of Na 2 S (1, 5 and 10 equiv.) after i.p. injection of probe (Fig. 6, panels B, C and D). The last group were given i.p. injection of SNP and followed by i.p. injection with the probe. The mice were imaged using a Night OWL IILB 983 small animal in vivo imaging system. The fluorescent images showed almost no background fluorescence in the negative control group (Supplementary Fig. S13, panel A; panel C, R = 0.12 in column A), and weak fluorescence in the probe-loaded group (Fig. 6, panel A; panel F, R = 1.0 in column A), which suggests weak fluorescence signals in probe-loaded mice may be caused by endogenous H 2 S. To confirm this assumption, we pretreated (i.p. injection) another group of mice with ZnCl 2 (an efficient eliminator of H 2 S). After 10 min of ZnCl 2 -treatment, the mice were given i.p. injection of free probe. Compared with the free probe-loaded mice, the fluorescence of ZnCl 2 -treated group is remarkably weakened (Supplementary Fig. S13, panel B; panel C, R = 0.18 in column B), indicating that the weak fluorescence in probe-loaded mice is triggered by physiological concentration of endogenous H 2 S. The mice treated with both Na 2 S (1, 5 and 10 equiv.) and the probe displayed much higher fluorescence (Fig. 6, panels B, C and D) than the mice treated with only the probe, which demonstrate that NIR-HS could respond to exogenous sulphide in mice. Moreover, the mice injected with SNP (Sodium Nitroprusside, a NO donor, could induce the production of endogenous H 2 S) and probe showed a maked elevation in the fluorescence intensities from the abdominal area of the mice (Fig. 6, panel E, R = 3.4 in column E), indicating that NIR-HS was sensitive enough to detect endogenous H 2 S in living mice. Importantly, the fluorescence intensities from the abdominal area of the mice were quantified, and the data showed that the fluorescence intensities triggered by Na 2 S were concentration-dependent (R = 1.0 in column A, R = 1.6 in column B, R = 3.5 in column C, R = 4.8 in column D) (Fig. 6, panel F). Figure 7 demonstrated that the fluorescence intensities became strong gradually within 20 min, consistent with the results of titrating the probe with Na 2 S at different time in PBS buffer ( Supplementary Fig. S2). These experiments suggested that NIR-HS is suitable for monitoring exogenous and endogenous H 2 S in living mice.
Recently, the development of fluorescent probes for H 2 S in vivo is of high interest. A few fluorescent probes have been successfully discovered for imaging of H 2 S in living animals, such as mice [49][50][51] , zebrafish [52][53][54][55] and Caenorhabditis elegans 56,57 , et al. In addition to fluorescent probes, luminescent probe and chemiluminescent probe have been applied to determining H 2 S in living mice 58,59 . Despite these progresses, the NIR fluorescence imaging of endogenous H 2 S in vivo is still highly desirable. Wallace et al. utilized fluorescent probe SF5 to investigate the regulation of leukocyte H 2 S synthesis in vivo 44 . However, probe SF5 emitted around 520 nm, the visible-light range limited its application for in vivo imaging due to the interference of background autofluorescence. Lu et al. prepared a novel bioluminescence probe for detection of endogenous H 2 S in nude mice 45 . Nevertheless, for this bioluminescence probe, Cys at 15 μ M triggered weak bioluminescence. It is well known that the concentrations of Cys in cells/tissue are much higher compared to the concentrations of endogenous H 2 S, and small response induced by Cys may interference the detection of H 2 S. Moreover, the H 2 S reaction site of this bioluminescence probe was azide group. The azide-containing H 2 S probes could undergo photoactivation under continuous excitation 60 , rendering them unsuitable for in vivo imaging. Thus, NIR probes with high sensitivity, good selectivity and favourable properties to monitor endogenous H 2 S in vivo are highly needed. Our probe NIR-HS is more suitable for biological imaging endogenous H 2 S in living mice.
Taken together, we have prepared a novel fluorescent probe NIR-HS for H 2 S detection in living cells and mice. Advantages of this H 2 S-specific probe include emission in the NIR region, a low detection limit, high sensitivity, good selectivity and low cytotoxicity. This probe not only enables fluorescence imaging of endogenous H 2 S induced by SNP in living cells, but also detects endogenous H 2 S generated in cells overexpressing cystathionine β -synthase (CBS). The probe was also successfully applied to visualizing both the exogenous and endogenous H 2 S in living mice. Probe NIR-HS shows the potential to be used as a valuable research tool in studying biological roles of H 2 S. We are currently pursuing other strategies to develop more sensitive and specific fluorescent sensors for monitoring H 2 S in living animails, as well as the H 2 S-related medical studies.

Methods
Fluorometric analysis. All fluorescence measurements were conducted at room temperature on a Hitachi F4600 Fluorescence Spectrophotometer. The probe solution (CH 3 CN) was added to a quartz cuvette. With the probe diluted to 10 μ M with 20 mM PBS buffer, Na 2 S was added (Na 2 S·9H 2 O serving as the H 2 S source in all experiments). The resulting solution was then incubated for 20 min. The samples were excited at 670 nm with the excitation and emission slit widths set at 5 nm and 10 nm, respectively. The emission spectrum was scanned from 690 nm to 850 nm at a velocity of 1200 nm/min. The photomultiplier voltage was set at 1000 V. Data are presented as the mean ± SD (n = 3).

Western blot.
For western blot analysis, the cells were washed with cold PBS and lysed with RIPAa buffer (1% Triton X-100, 1% deoxycholate, 0.1% SDS) containing protease inhibitors (1mM PMSF, 20mM NaF, 1mM NaVO 3 ). Protein concentrations were determined using BCA protein assay kits. The cells samples were separated by 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred onto nitrocellulose membranes (Amersham Pharmacia Biotech). The membranes were blocked with 5% skim milk powder in a washing buffer (Tris-buffered saline containing 0.05% (v/v) Tween 20) for 2 h at 25 °C and subsequently incubated overnight with the primary antibodies specific for CBS (1:1000) and β -actin(1:1000). Each membrane was thrice rinsed for 15 min and incubated with either alkaline phosphatase-conjugated secondary antibodies (1:1000, Goat anti-Rabbit IgG antibody) or alkaline phosphatase-conjugated secondary antibodies (1:1000, Horse anti-MouseIgG antibody), which was followed by visualization by BCIP/NBT alkaline phosphatase colour development kits. Protein bands were scanned and quantified by densitometric analysis using ImageJ version 1.34 s software.

Construction of CBS overexpressing cells.
The cDNA clones for human CBS were purchased from Origene (lot no.: RC201755). MCF-7 cells were grown to 90% confluence before being transiently transfected with pCMV6-control and pCMV6-CBS expression plasmids using Lipofectamine 2000 (Invitrogen, Shanghai, China) according to the manufacturer's instructions. Six hours after transfection, the medium containing transfection reagents was removed and incubated in fresh medium. The CBS overexpressed cells were harvested for subsequent experiments.