A biotin-guided hydrogen sulfide fluorescent probe and its application in living cell imaging

Hydrogen sulfide (H2S), a well-known signaling molecule, exerts significant regulatory effects on the cardiovascular and nervous systems. Therefore, monitoring the metabolism of H2S offers a potential mechanism to detect various diseases. In addition, biotin is significantly used as a targeting group to detect cancer cells exclusively. In this work, a biotin-guided benzoxadizole-based fluorescent probe, NP-biotin, was developed for H2S detection and evaluated in normal liver cell (LO2) and liver cancer cell (HepG2) lines. Results reveal that NP-biotin can detect cellular H2S with high sensitivity and selectivity. Moreover, NP-biotin has been confirmed to possess the ability to target cancer cells under the guidance of the biotin group.


Introduction
Like carbon monoxide (CO) and nitric oxide (NO), hydrogen sulde (H 2 S) is well known as a gaseous mediator. H 2 S regulates the cardiovascular system 1-5 and nervous systems 6,7 and also exerts anti-inammatory effects. 8,9 So far, many studies have suggested that endogenous H 2 S is mainly produced from cysteine by CBS (cystathionine b-synthase) or CSE (cystathionine g-lyase) enzymes, which are responsible for the synthesis of H 2 S in vivo. [10][11][12] In addition, several pathways for H 2 S synthesis have been reported, 13,14 in which the enzymatic actions of CBS and CSE on cysteine have been regarded as the predominant driving force.
As a signaling molecule, various concentrations (from nM to mM) of H 2 S are found in different tissues and biological uids, 15 thus the excess generation or paucity of H 2 S indicate disease status. In recent years, it has been reported that dysregulation of H 2 S metabolism is related to neurodegenerative diseases, such as Parkinson's, Alzheimer's, and Huntington's diseases. 16 However, research on the physiological and pathological functions of H 2 S is still in its preliminary stage compared the extensive studies on CO and NO. Monitoring the production and distribution of cellular H 2 S could help to understand how it stimulates biological response and interacts with signaling pathways, further illustrating its relationship with diseases.
Compared with the traditional methods, using the uorescence imaging technique to detect cellular H 2 S has numerous advantages. Several benets include monitoring the production of H 2 S in real-time and displaying the spatial distribution of H 2 S without destructive sampling. Besides, the uorescence imaging has high sensitivity and selectivity towards its target based on uorescent probes. [17][18][19][20][21][22][23][24] To assist the uorescence imaging of H 2 S, it is necessary to develop a uorescent probe with excellent performance. Based on reports of the good selectivity of the piperazinyl-NBD-based probe towards H 2 S, we selected piperazinyl-NBD as a response group for H 2 S in designing our probe. 25 Many cancer cells oen overexpress vitamin receptors (such as folate and biotin) on the surface of the cell membrane to aid the transduction of signals and uptake of nutrients, which could promote the fast growth and proliferation of cancer cells. 26,27 Based on the specic recognition between vitamins and their corresponding receptors, vitamins are commonly used in a drug delivery system to target cancer cells exclusively. [28][29][30] In previous reports, researchers conrmed that the biotin-modied probes would be more selectively taken up by biotin-positive cancer cells than by biotin-negative cells. [31][32][33][34] Taking this into consideration, the biotin group was introduced into our probe to increase its cancer-targeting ability.
In this work, a biotin-guided piperazinyl-NBD-based uorescent probe NP-biotin for H 2 S is reported. Biotin was selected as a cancer-targeting group, and piperazinyl-NBD was used as a response group for H 2 S. NP-biotin exhibited great sensitivity and selectivity for H 2 S. In the living cell imaging, NP-biotin successfully detected cellular H 2 S and targeted the cancer cells via the binding with biotin receptors under guidance of the biotin group.

Materials and instruments
All reagents are obtained commercially. UV-Vis spectra were obtained from spectrometer (Beckman DU 800, USA) and uorescence spectra were measured on a uorescence spectrophotometer (SPEX Flurolog 3-TCSPC instrument, USA). 1 H and 13 C NMR spectra were recorded on nuclear magnetic resonance spectrometer (Bruker AVIII-400, Germany) and mass spectra were recorded on mass spectrometry (AB Sciex QSTAR Elite, USA). Water was prepared by the Milli-Q purication system.

Synthesis of NP-biotin
The synthesis procedure of our probe, NP-biotin, is shown in Scheme 1. Under argon atmosphere, NBD-PZ (50 mg, 0.20 mmol) was added into a mixture of biotin (68 mg, 0.20 mmol), DIPEA (0.10 mL), EDCI (58 mg, 0.30 mmol), and HOBt (68 mg, 0.50 mmol) in DMF (5.0 mL). The mixture was stirred overnight at room temperature. Aer the addition of 20 mL water, the mixture was extracted with ethyl acetate and washed three times by water, then dried with Na 2 SO 4 . Aer the solvent was removed, the crude product was puried by silica column chromatography with dichloromethane/methanol (20 : 1) to obtain the nal NP-biotin product, which displayed an orange color, with a yield of 45%. The structure of NP-biotin was conrmed by 1 H and 13 C NMR spectrum ( Fig. S1 and S2 †). 1

Spectroscopic measurements
The stock solution of NP-biotin (1 mM, DMSO) was prepared then stored in the dark until use. The stock solution of Na 2 S (the source of H 2 S) and other analytes (10 mM) were dissolved in water then diluted by PBS (10 mM, pH ¼ 7.4) when used. For the spectroscopic measurements, the stock solution of NP-biotin was diluted to work solution (10 mM) by PBS (10 mM, pH ¼ 7.4, 1% DMSO). In the titration experiment, Na 2 S was added step-wise into the NP-biotin work solution to observe the behavior of NP-biotin towards different concentrations of Na 2 S. Subsequently, the selectivity of NP-biotin was investigated by addition of the various analytes (100 mM) mentioned above.
Also, the availability of NP-biotin was tested in solutions of different pH values, ranging 3.0-12.0. The kinetics between NPbiotin and Na 2 S (100 mM) were explored by recording the uorescence change of NP-biotin at 550 nm upon the addition of Na 2 S. The uorescence spectrum of NP-biotin was recorded under the excitation at 480 nm, and the slit width of the excitation and emission was set to 5 nm.
Cell culture and living cell imaging HepG2 and LO2 cells were cultured in Dulbecco's modied Eagle's medium supplemented with 10% fetal bovine serum in an atmosphere with 5% CO 2 at 37 C. Cells were seeded in glassbottom culture dishes until attached and then incubated with the 10 mM probe followed by washing with PBS before imaging on a laser scanning confocal microscope (Olympus FV-1000-IX81). Emission was collected at the green channel (500-600 nm, excitation at 488 nm).

Cytotoxicity assay
Cells were seeded in a 96-well cell culture dish (1000-10 000 cells per well). Aer cell attachment, the cell culture medium containing the probe from 0-100 mM was added for 24 h incubation. Then, the medium containing probe was replaced with the medium containing 10% CCK-8 reagent. Aer another 4 h of incubation, the absorbance of each well at 450 nm was measured with the microplate reader (TECAN Innite Series M1000 Pro).

Results and discussion
Optical response of NP-biotin to H 2 S To explore the response of NP-biotin towards different equivalents of Na 2 S, we performed a titration experiment. The uorescence spectra of NP-biotin were recorded aer the Na 2 S stepwise Scheme 1 The synthesis of NP-biotin. addition. As shown in Fig. 1a, NP-biotin shows strong uorescence emission at 550 nm, and the emission peak gradually decreased as the concentration of Na 2 S increased. A good linear relationship is seen between the uorescence intensity at 550 nm and Na 2 S concentration in the range of 0-40 mM (Fig. 1b). Based on the 3s/s principle, 35 the limit of detection was calculated to be 3.69 nM, which is much lower than the physiological concentration of H 2 S (10-100 mM). 36-38 NP-biotin exhibited great sensitivity for Na 2 S, which indicates its good potential in living cell imaging.
The selectivity was evaluated by Na 2 S and other analytes, including biothiols GSH, Cys, and Hcy. As shown in Fig. 2a, only Na 2 S could efficiently quench the uorescence of the probe, and the uorescence intensity at 550 nm was reduced about 100-fold aer the addition of Na 2 S (Fig. 2b). It is worth mentioning that GSH, Cys, and Hcy had almost no effect on the uorescence performance of NP-biotin. Furthermore, the response of NPbiotin to Na 2 S was also explored in the presence of other analytes (Fig. S3 †). The results show that the existence of other species do not affect the quenching of NP-biotin by Na 2 S. It is also interesting to note that the probe showed excellent selectivity towards Na 2 S, which implies it could be applied in complex biological systems.
To investigate the behavior of NP-biotin with or without Na 2 S at different pH values, a series of phosphate buffers with different pH values were prepared. The difference between the uorescence intensity at 550 nm in the absence (black line in Fig. 3a) and presence (red line in Fig. 3a) of Na 2 S could reect the availability of NP-biotin towards Na 2 S under different pH conditions. In Fig. 3b, a dramatic reduction of uorescence intensity at 550 nm occurred in the pH range from 6.0-9.0 and reached a maximum at pH 7.0, which is closed to the physiological pH of humans. Impressively, these results indicate that NP-biotin works well in physiological pH conditions.
To explore the kinetics between NP-biotin and Na 2 S, the time-dependent uorescence intensity of NP-biotin at 550 nm was monitored aer the addition of Na 2 S. As shown in Fig. 4, the uorescence intensity of NP-biotin decreased signicantly as soon as the Na 2 S was added and reached a minimum aer about 1 h. The pseudo-rst-order rate was calculated to be 7.4 Â 10 À3 S À1 by tting the uorescence intensity with a single exponential function. These data imply the possibility of NPbiotin to detect H 2 S in a real-time manner. Fig. 2 Fluorescence spectra (a) and F 0 /F at 550 nm (b) of 10 mM NPbiotin with various species (Na 2 S, GSH, Cys, Hcy, Na 2 S 2 O 4 , KNO 3 , Na 2 S 2 O 3 , NaOCN, Na 2 S 2 O 5 , KBr, NaNO 2 , NaN 3 , K 2 P 2 O 7 , NaI, Na 2 SO 4 , CH 3 COONa, KF, Na 2 SO 3 , KSCN and NaHSO 3 ) under excitation at 480 nm (F 0 represents the fluorescence intensity of NP-biotin and F represents the fluorescence intensity of NP-biotin in the presence of other guests respectively).  In order to verify the mechanism of the reaction between the probe and Na 2 S, the methanol solution before and aer the reaction was analyzed by mass spectrometry. The sensing mechanism was explored by ESI-HRMS. As shown in Fig. S4  : 195.9820)) was detected in the mass spectrum of the probe solution aer Na 2 S was added. These data conrm that in the proposed sensing mechanism (Fig. 5) based on the nucleophilicity, H 2 S cleaves the C-N bond between piperazinyl and benzoxadizole of NPbiotin to ultimately produce NBD-SH.

Fluorescence imaging of NP-biotin in living cells
In the following cell imaging experiments, HepG2, 39-41 a liver cancer cell line, was used as biotin receptor-positive cell, and a normal liver cell line LO2 (ref. 42) was selected as biotin receptor-negative cell.
To explore the application of NP-biotin in the detection of intracellular hydrogen sulde, we examined the effects of probes at different concentrations on cell viability using the CCK-8 Kit. HepG2 cells were treated by the probe with different concentrations for 24 h. The cytotoxicity results show that the cell viability of HepG2 had almost no change with 0-50 mM probe treatment and remained about 80% with the 100 mM probe, which indicates that the probe exhibits low cytotoxicity ( Fig. S6 †). Hence, we selected a concentration of 10 mM for subsequent cell imaging experiments.
To verify the ability of probe molecules to detect exogenous hydrogen sulde, LO2 cells were selected as the target cells. Aer incubation with 10 mM NP-biotin for 1 h, strong uorescence was observed inside LO2 cells (Fig. 6a). Aer 50 mM Na 2 S was added, the uorescence gradually decreased and almost turned off 10 min (Fig. 6c). These results demonstrate that the probe could detect exogenous hydrogen sulde.
Comparatively, in uorescent imaging of HepG2, uorescence was observed on the cell membrane but almost no uorescence could be observed in the cytoplasm (Fig. 7a). It has been reported that cancer cells usually express excessive hydrogen sulde compared to normal cells; 43,44 thus, the weak uorescence inside the HepG2 cells may be caused by the high concentration of hydrogen sulde. NMM (N-methylmaleimide), a hydrogen sulde scavenger, was tested to verify our speculation. HepG2 was pretreated with 1 mM NMM, which was used to consume endogenous H 2 S before incubation with the probe. The uorescence appeared when the cells were pretreated with NMM, as shown in Fig. 7b. At the moment, the pH of endosome and lysosome in HepG2 cells was acidic, but the signicant uorescence of the NP-biotin was shown in the cells. So this experiment can conrm that the uorescence of the NP-biotin would not be signicantly reduced by being exposed to acidic pH in endosome and lysosome. From what has been discussed above these results indicate that the weak uorescence was caused by the high concentration of H 2 S in HepG2.
Cell imaging was further used to investigate the tumor targeting ability of the probe. In this assay, both biotin    receptor-positive cells (HepG2) and biotin receptor-negative cells (LO2) were selected as subjects. It is suggested that the bright uorescence on the cell membrane of HepG2 may be attributed to the biotin receptors, which could recognize and bind to the biotin group of our probe. [45][46][47] To conrm this, biotin (nal concentration of 2 mM) was added with probe, in which biotin was used as competition for biotin receptors on the cell membrane with the probe. As shown in Fig. 8, the presence of biotin signicantly reduced the uorescence intensity on the cell membrane of HepG2, which indicates that biotin occupied a certain amount of biotin receptors and limited the sites available for NP-biotin. This phenomenon demonstrates the targeting ability of NP-biotin towards cancer cells via biotin receptors. As expected, the co-incubation of biotin with the probe had no inuence on LO2 cells, a biotin receptor negative cell line (Fig. 9). These results prove that NPbiotin possesses a cancer-targeting function by binding to the biotin receptor.
Aer incubation with 10 mM probe for 1 h and washing by PBS, HepG2 and LO2 cells were observed every 5 min by the microscope. The uorescence on the cell membrane of HepG2 gradually decreased over time, which suggests that the probe was transported from the membrane to cytoplasm and then recognized by the H 2 S inside HepG2 cells (Fig. 10a). On the contrary, the uorescence of LO2 cells did not decrease (Fig. 10b), indicating that the probe in the cytoplasm was not completely quenched. This result demonstrates that HepG2 cells exhibit a higher H 2 S concentration than LO2 cells, which agrees with the previous studies that reported cancer cells usually express more H 2 S than normal cells. 43,44

Conclusions
In this work, a novel biotin-guided piperazinyl-NBD-based uorescent probe (NP-biotin) for sensing cellular H 2 S was successfully developed. Results conrm that NP-biotin possesses excellent sensitivity and selectivity toward H 2 S, which implies its good application in live cell imaging. The cell imaging results show that the NP-biotin could detect cellular H 2 S in complex biological systems. This work further reports that, by selecting HepG2 and LO2 as biotin receptor positive and negative cells, respectively, NP-biotin can target cancer cells via the recognition between the biotin group of NP-biotin and biotin receptors.

Conflicts of interest
There are no conicts to declare.  This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 36135-36140 | 36139