Three-channel fluorescent sensing via organic white light-emitting dyes for detection of hydrogen sulfide in living cells

Biomaterials 34 (2013) 7429e7436

Contents lists available at SciVerse ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Three-channel fluorescent sensing via organic white light-emitting dyes for detection of hydrogen sulfide in living cells Jiaoliang Wang, Weiying Lin*, Weiling Li State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha, Hunan 410082, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 April 2013 Accepted 10 June 2013 Available online 29 June 2013

To demonstrate the feasibility of development of three-channel based fluorescent sensors based on organic white light-emitting dyes, in this work, for proof-of-principle, we initially judiciously designed an organic white light-emitting dye, which was further used as a robust platform to engineer a new fluorescent sensor for monitoring H2S with turn-on fluorescence signals in blue, green, and red emission channels in both solution and living cells. This work should open a new avenue for design of three- and multiple-channel based fluorescent sensors for various analytes. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Fluorescent sensors ESIPT Three-channel Hydrogen sulphide Cellular imaging

1. Introduction Fluorescent sensing is widely applied in diverse fields such as chemistry, biology, and medicine due to its high sensitivity and simple operation [1e3]. In addition, by combining with fluorescence microscopy, fluorescent sensing and bio-imaging can be exploited as a powerful approach to investigate biomolecules of interest with high temporal and spatial resolution. Fluorescent sensors are requisite molecular tools for sensing and bio-imaging. Up to date, a large volume of fluorescent sensors have been developed. However, most of them exhibit fluorescence signal variations only in one channel. By contrast, dual-channel based sensors have fluorescence signal changes in two distinct channels. This may reduce the potentials errors due to false positive or artifacts arising from environmental factors [4e6]. Inspired by the spectral feature of dual-channel based sensors, we reasoned that three-channel based sensors could have fluorescence signal variations in three distinct channels. Thus, in principle, three-channel based fluorescent sensors should be much more reliable to eliminate potential false positive or artifacts, as the fluorescent signals in three different channels can be used for mutual corroboration. However, to our best knowledge, no organic three-channel based fluorescent sensors have been developed yet. Although the

* Corresponding author. Fax: þ86 731 88821464. E-mail address: [email protected] (W. Lin). 0142-9612/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biomaterials.2013.06.013

development of organic three-channel based fluorescent sensors is still an unmet challenge, we envisioned that this type of sensors could be constructed based on appropriate organic white lightemitting dyes, which emit light with three different channels, red, green, and blue. In this communication, to test the hypothesis that three-channel based fluorescent sensors could be constructed based on organic white light-emitting dyes, for proof-of-principle, we initially formulated a strategy (Fig. 1) to design compound 1 (Scheme 1) as a new organic white light-emitting dye, which was further applied as a platform to develop the fluorescent sensor 4 for hydrogen sulfide (H2S) (Scheme 1). Significantly, the fluorescent sensor 4 could respond to H2S with turn-on fluorescence signals in blue, green, and red emission channels in both solution and living cells. As organic white light-emitting dyes often bear excited-state intramolecular proton transfer (ESIPT) components, herein, we proposed a strategy to design organic white light-emitting dyes by conjugately connecting a blue fluorescent dye with an ESIPT dye (Fig. 1). In a typical ESIPT process, photoexcitation elicits a shift in electron density that facilitates proton migration from a dye in the enol form to the keto form. [1] Thus, an ESIPT dye normally possesses two different emission channels. We hypothesized that appropriately combing a blue light emitting dye with a dualchannel ESIPT dye may afford an organic white light-emitting dye with three emission channels. Notably, although the ESIPT mechanism has been widely used to design dual-channel based fluorescent sensors [7e9], the

7430

J. Wang et al. / Biomaterials 34 (2013) 7429e7436

in materials science, their three-channel characteristic in solution state has not been employed in the fluorescent sensing field. Based on the above strategy, compound 1 was designed as an organic white light-emitting dye, which is composed of two subunits, 1H-phenanthro[9,10-d]imidazol and 3-hydroxychromone moieties connected by a rigid and conjugated phenyl linker. The choice of these two subunits is based on the considerations that 1H-phenanthro[9,10-d]imidazol dyes [10,11] emit blue light with emission from 380 to 420 nm and 3-hydroxychromone is a typical ESIPT dye with dual channel emission [12,13]. Fig. 1. The proposed design strategy of organic white light-emitting dyes.

2. Experimental sections 2.1. Materials and instruments

three-channel feature of organic white light-emitting dyes has not been previously exploited in the development of three-channel based fluorescent sensors. In other words, although organic white light-emitting dyes in solid state have found important applications

Unless otherwise stated, all reagents were purchased from commercial suppliers and used without further purification. Solvents used were purified and dried by standard methods prior to use. Twice-distilled water was used throughout all experiments. Melting points were determined with a Beijing taike XT-4 microscopy

O H N CHO

OH

N

H2O2

a

N

b

O

5

1

O2N

NO2

NO2

d

F C4H9Br

O

HO

H N

O2N

c

O

H N N

HO

OH CHO

N

O 4

O

N a

H2O2

N

b

N 2

HO

N

O

O 1

O

O

6

H N

O

O C4H9Br

N

e

N

O

O 3

Scheme 1. Synthetic route to compounds 1, 2, 3, and 4. a: NaOH (10%), EtOH, 60  C; b: NaOH (10%), EtOH, rt; c: K2CO3, DMF, 50  C; d: K2CO3, DMF, 75  C; e: K2CO3, acetone, TBAI, reflux.

J. Wang et al. / Biomaterials 34 (2013) 7429e7436 and were uncorrected. ESI-MS analyses were performed using a Waters Micromass ZQ-4000 spectrometer. High resolution mass spectrometric (HRMS) analyses were measured on a Finnigan MAT 95 XP spectrometer. Electronic absorption spectra were obtained on a LabTech UV Power spectrometer; Photoluminescent spectra were recorded with a HITACHI F4600 fluorescence spectrophotometer; fluorescence spectrophotometer with the excitation and emission slit widths at 5.0 nm. Cells imaging were performed with a Nikon eclipase TE300 inverted fluorescence microscopy. 1H and 13C spectra were measured on an INOVA-400 spectrometer using TMS as an internal standard. TLC analyses were performed on silica gel plates and column chromatography was conducted over silica gel (mesh 200e300), both of which were obtained from the Qingdao Ocean Chemicals.

7431

for 1.5 h. Subsequently, the solvent was removed under reduced pressure, and the resulting residue was purified by chromatography on silica gel (acetone:dichloromethane, 1 : 20, V/V) to give compound 4 as a orange yellow solid (0.105 g, 0.17 mmol, yield 70.1%). mp 154e156  C. 1H NMR (400 MHz, d6-DMSO), d ¼ 13.66 (s, 1H), 8.93 (d, J ¼ 2.8 Hz, 1H), 8.86 (s, 2H), 8.57 (d, J ¼ 8.0 Hz, 2H), 8.51 (d, J ¼ 8.8 Hz, 2H), 8.37e8.34 (dd, J ¼ 9.2, 2.4 Hz, 1H), 8.27 (d, J ¼ 8.4 Hz, 2H), 8.11 (d, J ¼ 8.0 Hz, 1H), 7.98 (d, J ¼ 3.6 Hz, 2H), 7.76 (s, 2H), 7.71 (s, 1H), 7.69e7.66 (t, J ¼ 4.8 Hz, 2H), 7.61e 7.59 (m, 1H); 13C NMR (100 MHz, d6-DMSO): 171.1, 156.2, 155.2, 153.3, 147.5, 141.5, 138.1, 134.8, 134.4, 133.1, 129.2, 128.9, 127.6, 127.2, 126.1, 125.7, 125.0, 123.2, 121.7, 118.8, 117.8. HRMS (ESI): m/z calcd for [C36H20N4O7]:620.1043, Found: 620.1016. 2.3. Preparation of the test solution

2.2. Syntheses 2.2.1. Synthesis of 2-(4-(1H-phenanthro[9,10-d]imidazol-2-yl)phenyl)-3-hydroxy4H-chromen-4-one (1) The starting compound 5 was synthesized according to a reported method [10]. o-hydroxy acetophenone (0.136 g, 1 mmol) was added to a solution of 10% NaOH in EtOH/H2O (5:1, v/v) with stirring. After 20 min, compound 5 (0.322 g, 1 mmol) in 5 mL ethanol was introduced slowly. The mixture was stirred at room temperature for 2e3 h, and then heated at 60 C for 16 h. The red precipitate was filtrated and washed with cold EtOH. The red solid was dissolved in 10 mL ethanol with 1.5 mL of 10% NaOH, and then 1.5 mL of 30% H2O2 solution was added dropwise. The reaction mixture was further stirred for 1.5 h. The resulting orange reaction mixture was poured on crushed ice water. The yellow solid thus obtained was filtrated, washed with water and dried. The crude product was purified by silica gel column chromatography (dichloromethane/methanol, 40:1, v/v) to give compound 1 as a yellow solid powder (244.8 mg, 0.539 mmol, 90%). mp 175e178  C. 1H NMR (400 MHz, d6DMSO), d ¼ 10.23 (s, 1H), 9.86 (s, 1H), 8.86 (d, J ¼ 8.4 Hz, 2H), 8.71 (s, 1H), 8.61 (d, J ¼ 8.4 Hz, 2H), 8.46 (d, J ¼ 8.4 Hz, 2H), 8.15 (d, J ¼ 7.6 Hz, 1H), 7.85 (d, J ¼ 3.6 Hz, 2H), 7.77e7.73 (t, J ¼ 7.6 Hz, 2H), 7.67e7.64 (t, J ¼ 7.2 Hz, 2H), 7.52e7.48 (m, 1H); 13C NMR (100 MHz, d6-DMSO): 172.9, 154.6, 148.3, 144.6, 139.5, 133.8, 131.7, 131.2, 128.0, 127.8, 127.1, 126.2, 125.4, 124.8, 124.6, 123.9, 122.3, 122.2, 121.3, 118.5. HRMS (ESI): m/z calcd for [C30H18N2O3]:454.1317, Found: 455.1390. 2.2.2. Synthesis of 2-(4-(1-butyl-1H-phenanthro [9,10-d]imidazol-2-yl) phenyl) 3hydroxy-4H-chromen-4-one (2) Compound 6 was synthesized according to a reported method [14]. The synthetic procedure for compound 2 from compound 6 is the same as that for the preparation of compound 5 from compound 1 as described above. The crude product was purified by silica gel column chromatography (dichloromethane : petroleum ether, 10:1, v/v) to provide compound 2 as a yellow green solid powder (0.415 g, 0.815 mmol, 81.5%). mp 155e158  C. 1H NMR (400 MHz, d6-DMSO), d ¼ 9.94 (s, 1H), 8.99 (d, J ¼ 8.0 Hz, 1H), 8.87 (d, J ¼ 8.4 Hz, 1H), 8.64 (d, J ¼ 6.8 Hz, 1H), 8.50 (d, J ¼ 8.8 Hz, 2H), 8.44 (d, J ¼ 8.0 Hz, 1H), 8.17 (d, J ¼ 7.2 Hz, 1H), 8.03 (d, J ¼ 8.8 Hz, 2H), 7.85e7.79 (m, 3H), 7.76e7.67 (m, 2H), 7.53e7.49 (m, 1H), 6.97e6.93 (t, J ¼ 7.8 Hz, 1H), 4.78e4.75 (t, J ¼ 7.0 Hz, 2H), 1.85e1.80 (q, J ¼ 7.2 Hz, 2H), 1.20e1.11 (m, 2H), 0.73e0.70 (t, J ¼ 7.4 Hz, 3H); 13C NMR (100 MHz, d6-DMSO): 173.4, 172.2, 161.4, 154.9, 144.6, 140.0, 137.7, 135.9, 134.2, 132.4, 131.9, 130.5, 130.2, 128.6, 128.1, 128.0, 127.7, 127.1, 126.4, 126.0, 125.5, 125.1, 124.9, 124.8, 123.9, 123.1, 122.2, 121.6, 121.5, 119.4, 118.7, 117.3, 113.2, 46.6, 32.0, 19.1, 13.5. HRMS (ESI): m/z calcd for [C34H26N2O3]:510.1943, Found: 511.2016. 2.2.3. Synthesis of 3-butoxy-2-(4-(1-butyl-1H-phenanthro[9,10-d]imidazol-2-yl) phenyl)-4H- chromen-4-one (3) Compound 1 (1.135 g, 2.5 mmol), n-butylbromide (1.7 g, 12.5 mmol), tetra-nbutylammonium iodide (TBAI) (0.033 g, 0.14 mmol), anhydrous potassium carbonate (2.760 g, 20.0 mmol), and acetone (20 mL) were refluxed for 24 h and the reaction was monitored by TLC. After acetone was removed under reduced pressure, the solid was collected and purified by silica-gel column chromatography (dichloromethane: petroleum ether, 5:1, v/v) to give compound 3 as a yellow green solid powder (1.064 g, 1.88 mmol, 75%). mp 162e164  C. 1H NMR (400 MHz, d6DMSO), d ¼ 8.98 (d, J ¼ 8.4 Hz, 1H), 8.87 (d, J ¼ 8.4 Hz, 1H), 8.63 (d, J ¼ 7.6 Hz, 1H), 8.43 (d, J ¼ 8.0 Hz, 1H), 8.30 (d, J ¼ 8.4 Hz, 2H), 8.15e8.13 (dd, J ¼ 8.0 Hz, 1H), 8.03e8.01 (d, J ¼ 8.4 Hz, 2H), 7.88e7.84 (m, 1H), 7.81e7.78 (t, J ¼ 6.4 Hz, 2H), 7.75e 7.65 (m, 3H), 7.55e7.51 (t, J ¼ 7.4 Hz, 1H), 4.77e4.73 (t, J ¼ 7.2 Hz, 2H), 4.10e4.07 (t, J ¼ 6.4 Hz, 2H), 1.86e1.79 (m, 2H), 1.66e1.63(m, 2H), 1.41e1.32 (m, 2H), 1.15e1.10 (m, 2H), 0.88e0.84 (t, J ¼ 7.4 Hz, 3H), 0.73e0.69 (t, J ¼ 7.4 Hz, 3H); 13C NMR (100 MHz, d6-DMSO): 193.5, 163.6, 151.5, 144.2, 136.6, 135.6, 130.6, 129.7, 129.3, 128.9, 128.2, 127.4, 126.9, 126.4, 125.0, 124.5, 123.2, 123.0, 122.7, 121.2, 120.8, 120.0, 118.9, 118.7, 47.0, 32.3, 19.5, 13.5. HRMS (ESI): m/z calcd for [C38H34N2O3]: 566.2569, Found: 567.2653. 2.2.4. Synthesis of 2-(4-(1H-phenanthro[9,10-d]imidazol-2-yl)phenyl)-3-(2,4dinitro-phenoxy)-4H-chromen-4-one (4) Compound 1 (0.109 g, 0.24 mmol), 1-fluoro-2,4-dinitrobenzene (0.116 g, 0.6 mmol), and anhydrous potassium carbonate (0.331 g, 2.4 mmol) were added into a three-neck bottom flask under N2 atmosphere, and then anhydrous DMF (2.0 mL) was injected into the reaction flask with a syringe. The mixture was stirred at 75  C

A stock solution of sensor 4 (5.0  104 M) was prepared in DMSO. The test solution of sensor 4 (5 mM) in 3 mL pH 7.4, 25 mM phosphate buffer: DMSO, v/v, 8: 2, with 0.5% Tween-20 was prepared by placing 0.05 mL of the sensor stock solution, 2.05 mL DMSO, and 0.9 mL of 25 mM sodium phosphate buffer (pH ¼ 7.4). The solutions of various testing species were prepared from NaHCO3; NaN3; NaOCl; KO2; Na2SO3; NaS2O3; KSCN; NaNO2; H2O2; NO; cysteine; glutathione; and NaHS, respectively. 2.4. DFT calculations Density functional theory (DFT) calculations were carried out with the GAUSSIAN 09 program package [15]. All the calculations were performed on systems in the gas phase using the Becke’s three-parameter hybrid functional with the LYP correlation functional (B3LYP) and 6-31G(d, p) basis set. 2.5. Cytotoxicity assay of sensor 4 MG63 cells were grown in the modified Eagle’s medium (MEM) supplemented with 10% FBS (fetal bovine serum) in an atmosphere of 5% CO2 and 95% air at 37  C. Immediately before the experiments, the cells were placed in a 96-well plate, followed by addition of increasing concentrations of sensor 4 (99% MEM and 1% DMSO). The final concentrations of the probe were kept from 5 to 100 mM (n ¼ 3). The cells were then incubated at 37  C in an atmosphere of 5% CO2 and 95% air at 37  C for 24 h, followed by the MTT assays. Untreated assay with MEM (n ¼ 3) was also conducted under the same conditions. 2.6. Cell culture and fluorescence imaging Living MG63 cells were seeded in a 12-well plate in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum for 24 h. Living MG63 cells were then incubated with NaHS (50 mM), cysteine (100 mM) or glutathione (100 mM) in the culture medium for 30 min at 37  C. After washing with PBS three times to remove the remaining NaHS, cysteine, or glutathione, the cells were further incubated with sensor 4 (5 mM) for 30 min at 37  C. After washing the cells with PBS three times, the fluorescence images were acquired with a Nikon Eclipse TE300 equipped with a CCD camera. For the control experiment, MG63 cells were incubated with sensor 4 (5 mM) in the culture medium for 30 min at 37  C. After washing with PBS three times, the fluorescence images were acquired with a Nikon Eclipse TE300 equipped with a CCD camera.

3. Results and discussions 3.1. Synthesis and spectroscopic evaluation of compound 1 The synthetic route to compound 1 and control compounds 2 and 3 is shown in Scheme 1. The starting compound 5 was prepared by a reported procedure [14]. Condensation of compound 5 with 1(2-hydroxyphenyl) ethanone under basic conditions gave a chalcone intermediate, which was further treated with H2O2 under basic conditions to afford target compound 1 by an oxidative cyclization [12,16]. Alkylation of compound 5 gave compound 6, which was then transformed into the control 2 by the similar synthetic sequences as formation of compound 1 from compound 5. Alkylation of compound 1 provided the control 3. With compound 1 in hand, we examined its optical properties in DMF (Fig. 2). Compound 1 exhibits three strong absorption bands at 337, 358, and 374 nm, showing the characteristic absorption of phenanthro [9,10-d]imidazol and 3-hydroxychromone subunits [10,12]. As designed, upon excitation at a single wavelength (350 nm), compound 1 displays three different emission bands at around 446, 527, and 569 nm (Fig. 2). The emission band at 446 nm

7432

J. Wang et al. / Biomaterials 34 (2013) 7429e7436

is inhibited by intermolecular H-bonding to solvents. As shown in Fig. S2, the CIE coordinate of compound 1 in DMF is (0.31, 0.35) in a CIE 1931 chromaticity diagram, in good agreement with the above finding that compound 1 emits white light under ambient temperature and pressure. By contrast, in other polar solvents including acetone and DMSO, compound 1 only emits near white light. 3.2. Investigation of the ESIPT mechanism

Fig. 2. Normalized absorption (o) and emission (C) spectra of compound 1 (10 mM) in DMF.

can be attributed to the emission of the 1H-phenanthro[9,10-d] imidazol moiety in a polar solvent [10], and the emission bands at 527 and 569 nm can be assigned to the enol and keto forms (Fig. 2) of the 3-hydroxychromone component [12], respectively. Furthermore, the assignment of the enol and keto emission bands is consistent with the HOMOeLUMO energy gaps for the enol and keto tautomers (Fig. S1). Notably, compound 1 exhibits the typical feature of ESIPT dyes with a large pseudo Stokes shift. Gratifyingly, in DMF, the emission intensities of the three emission bands of compound 1 are approximately equal. Consequently, the visual emission color of compound 1 in DMF is bright white (Inset of Fig. 5), indicating that compound 1 is promising as an organic dye which can emit white light under ambient temperature and pressure. As an ESIPT process can be influenced by solvents [17], we then proceeded to preliminarily examine the solvent effect on the emission properties of compound 1 (Fig. 3). The solvent effect due to the re-orientation of solvent molecules around the excited dyes is shown by a red-shift in the emission spectra of compound 1 with increasing the polarity of the solvents. For instance, in dichloromethane, a solvent of moderate polarity (ε ¼ 3.4), the enol/keto emission bands are at 514/547 nm. By contrast, in higher polar solvents such as acetone (ε ¼ 5.4), DMF (ε ¼ 6.4), and DMSO (ε ¼ 7.2), the enol/keto emission bands are at 508/559, 527/569, and 532/571 nm, respectively. On the other hand, in methanol, an alcoholic solvent, only the enol band is evident, as the ESIPT process

To examine the ESIPT mechanism as shown in Fig. 4, the enol and keto tautomers of dye 1 were optimized by density functional theory (DFT) calculations. The studies indicate that the enol tautomer bears an intramolecular H-bond with a distance of 1.94 Å, whereas the keto tautomer has an intramolecular H-bond with a distance of 1.79 Å (Fig. 4). These values are consistent with the typical distances of intramolecular H-bonds of the enol and keto tautomers in 3-hydroxychromone derivatives [12,18]. To further investigate the ESIPT mechanism, we decided to synthesize control compounds 2 and 3 (Scheme 1) and to study their emission profiles. When compared to dye 1, the control 2 has a butyl group on the 1H-phenanthro[9,10-d]imidazol subunit, whereas the control 3 bears two butyl groups on the 1H-phenanthro[9,10-d]imidazol and 3-hydroxychromone subunits, respectively. As shown in Fig. 5, the control 2 displays only two emission bands at 513 and 564 nm, ascribed to the enol and keto tautomers, while the characteristic emission band of 1H-phenanthro[9,10-d] imidazol unit is not present due to alkylation on the imidazol core by a butyl group. This is further confirmed by comparison with the emission profile of the control 3. Only one emission band attributed to the enol form is shown, while the emssion bands of the keto form and 1H-phenanthro[9,10-d]imidazol unit are not present, as the ESIPT is inhibited by alkylation on the key hydroxyl group of the chromone subunit [7e9] and the typical emission band of 1Hphenanthro[9,10-d]imidazol unit is blocked owing to alkylation on the imidazol core of 1H-phenanthro[9,10-d]imidazol subunit. The emission colors and absorption profiles of compounds 1e3 are shown in Figs. S3eS4. Taken together, these studies are in good agreement with the proposed ESIPT mechanism of compound 1 shown in Fig. 4. The above studies indicate that the emission properties of compound 1 can be readily tuned by blocking the critical hydroxyl group. Thus, we envisioned that incorporation of an appropriate trigger on the key hydroxyl group may afford a robust platform for development of fluorescent sensors provided that the trigger can be removed by a target of interest to generate compound 1 (Scheme 2). 3.3. Design of the fluorescent H2S sensor 4

Fig. 3. Emission spectra of compound 1 (10 mM) in various organic solvents. Excited at 350 nm.

For proof-of-principle, we created compound 4 (Scheme 1.) as a fluorescent sensor for hydrogen sulfide (H2S). H2S has been recently recognized as the third gasotransmitter together with carbon monoxide (CO) and nitric oxide (NO) [19], and it plays an important physiological role in many biological processes. However, abnormal levels of H2S are associated with many types of diseases [20e23]. Thus, it is of importance to develop fluorescent sensors for H2S. The design of compound 4 is based on the following considerations. Firstly, the removal of dinitrophenyl trigger can be mediated via a thiolysis reaction [24,25]. In addition, we envisioned that 2,4dinitrobenyl moiety may quench the emission of dye 1 due to the notorious fluorescence quenching effect of the nitro group [26]. Thereby, we anticipated that compound 4 is essentially nonfluorescent in all the three emission channels. However, upon interaction with hydrogen sulfide, the 2,4-dinitrobenyl moiety can be uncaged to release compound 1, and consequently, the fluorescent turn-on signals in all the three emission channels can be observed.

J. Wang et al. / Biomaterials 34 (2013) 7429e7436

7433

Fig. 4. Top: Excited-state intramolecular proton transfer (ESIPT) process in compound 1. Bottom: View of the intramolecular H-bonding interactions of compound 1. The optimization was conducted by B3LYP (6-31G(d, p)).

Fig. 5. Emission spectra of white light-emitting dye 1 (C) and control compounds 2 (B) and 3 (+) (10 mM) in DMF. Excited at 350 nm. Inset: photograph of white lightemitting dye 1 in DMF.

Although a few well-designed small-molecule fluorescent H2S sensors with a single or dual-emission channel have been developed very recently [27e34], no fluorescent H2S sensors with threeemission channels have been reported yet. 3.4. Spectroscopic properties of the fluorescent H2S sensor 4 Indeed, as designed, sensor 4 is essentially non-fluorescent in buffer/tween (pH 7.4, 25 mM phosphate buffer: DMSO, v/v, 8: 2, with 0.5% Tween-20). However, addition of increasing

concentrations of NaHS (NaHS was used as a hydrogen sulfide source in all experiments [27e29,31]) elicited a dramatic change in the emission profiles. 3-, 6-, and 16-fold fluorescence intensity enhancements at the three emission channels around 440, 510, and 570 nm, respectively, were noted (Fig. 6, Fig. S5), demonstrating that hydrogen sulfide induced a large fluorescence enhancement in all three emission channels. Thus, to our best knowledge, sensor 4 represents the new paradigm of organic three-channel based fluorescent sensors. In good agreement with the changes in the emission profiles, treatment of NaHS with sensor 4 caused a significant emission color change from darkness to bright white (Scheme 3). The detection limit (S/N ¼ 3) of sensor 4 was determined to be about 10 mM in buffer/tween, which is comparable to that of the known small-molecule fluorescent hydrogen sulfide probes (Fig. S6). In addition, a mass spectroscopic analysis confirms that the three-channel fluorescent turn-on response is indeed due to the transformation of sensor 4 to dye 1 in the presence of NaHS (Fig. S7). Furthermore, the selectivity studies suggest that sensor 4 is selective for H2S over other species tested including smallmolecule thiols, for example, glutathione at a very high concentration (10 mM) (Fig. S8). To study the practical applicability, the effect of pH on the fluorescence response of sensor 4 to hydrogen sulfide was investigated. As shown in Fig. S9, in the absence of hydrogen sulfide, almost no change in fluorescence intensity was observed in the free sensor 4 over a wide pH range of 5.5e8.5, indicating that the free sensor was stable in the wide pH range. Upon treatment with hydrogen sulfide, the maximal fluorescence signal was observed in the pH range of 5.5e8.5. Thus, the observation that sensor 4 had the maximal sensing response at physiological pH, suggests that sensor 4 was promising for biological applications.

Scheme 2. Platform for design of fluorescent sensors based on dye 1.

7434

J. Wang et al. / Biomaterials 34 (2013) 7429e7436

3.5. Fluorescence images of H2S in living cells

Fig. 6. Fluorescence spectral changes of sensor 4 (5 mM) upon addition of increasing concentrations (0e10 equiv.) of NaHS (lex ¼ 350 nm) in buffer/tween. Inset: the fluorescence enhanced factor (FEF) of sensor 4 with and without NaHS in blue, green, and red emission channels. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

In order to be useful as hydrogen sulfide imaging agents, fluorescent probes should have low cytotoxicity. Thus, we investigated the potential toxicities of sensor 4 against a representative cell line: MG63 cells. The living cells were incubated with various concentrations (5e100 mM) of sensor 4 for 24 h, and then the cell viability was determined by the standard 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assays [35]; the results indicated that sensor 4 do not exhibit marked cytotoxicity at concentrations below 50 mM (Fig. S10). The promising results of three-channel fluorescent turn-on responses of sensor 4 in the solution prompted us to further examine the possibility of three-channel sensing in living cells. Toward this end, sensor 4 was incubated with the living MG63 cells pre-treated with or without NaHS. As shown in Fig. 7aec, the cells treated with only the sensor exhibit relatively weak fluorescence in the blue, green, and red channels. By contrast, when the cells pre-treated with NaHS, then incubated with sensor 4, stronger fluorescence in the blue, green, and red channels were observed (Fig. 7def). Quantification of average fluorescence emission intensities indicate that approximately 3-, 10-, and 20fold fluorescence enhancement at the blue, green, and red channels, respectively, was obtained (Fig. 7g). Thus, these data establish that sensor 4 is cell membrane permeable and able to respond to

Scheme 3. The emission color changes upon reaction of sensor 4 with NaHS.

Fig. 7. (aec) Fluorescent images of living MG63 cells treated with sensor 4 (5 mM) for 30 min. (a) Fluorescent image from the blue channel. (b) Fluorescent image from the green channel. (c) Fluorescent image from the red channel. (def) Fluorescent images of living MG63 cells pre-treated with NaHS (50 mM) for 30 min, then incubated with sensor 4 (5 mM) for 30 min. (d) Fluorescent image from the blue channel. (e) Fluorescent image from the green channel. (f) Fluorescent image from the red channel. (g) The fluorescence enhanced factor (FEF) of sensor 4 with and without NaHS in three emission channels (blue, green, and red). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

J. Wang et al. / Biomaterials 34 (2013) 7429e7436

H2S with turn-on signals in three different emission channels in living cells. This represents the new report of three-channel turnon sensing in living cells by using an organic white light-emitting dye-based fluorescent sensor. Recently, He’s group reported that incubation of living cells with the potential sulphide sources such as cysteine or glutathione (at the 100 mM level) could stimulate the production of endogenous H2S, which was then detected by a fluorescent H2S probe [33]. They attributed the response to the result of the free sulphide generated intracellularly by the perturbed cellular levels of cysteine or glutathione, as the millimolar levels of thiols already exist inside cells.

7435

Thus, encouraged by the above promising results of imaging exogenous H2S in the living cells, we decided to further investigate the feasibility of our probe to detect endogenously produced H2S in living cells by using the protocol developed by He’s group [33]. As shown in Fig. 8def, the cells pre-incubated with 100 mM cysteine, then treated with sensor 4, exhibited much brighter emission in the blue, green, and red channels when compared to the cells treated with only sensor 4 (Fig. 8aec). Similarly, the cells pre-incubated with 100 mM glutathione, then treated with sensor 4, displayed much more intense emission in the blue, green, and red channels (Fig. 8gei) when compared to the cells treated with only sensor 4. The FEF data (Fig. 8j, k) of the cells incubated with 100 mM cysteine

Fig. 8. (aec) Fluorescent images of living MG63 cells treated with sensor 4 (5 mM) for 30 min. (a) Fluorescent image from the blue channel. (b) Fluorescent image from the green channel. (c) Fluorescent image from the red channel. (def) Fluorescent images of living MG63 cells pre-treated with 100 mM cysteine for 30 min, then incubated with sensor 4 (5 mM) for 30 min. (d) Fluorescent image from the blue channel. (e) Fluorescent image from the green channel. (f) Fluorescent image from the red channel. (gei) Fluorescent images of living MG63 cells pre-treated with 100 mM glutathione for 30 min, then incubated with sensor 4 (5 mM) for 30 min. (g) Fluorescent image from the blue channel. (h) Fluorescent image from the green channel. (i) Fluorescent image from the red channel. (j) The fluorescence enhanced factor (FEF) of cells treated with sensor 4 in the presence or absence of 100 mM cysteine in three emission channels (blue, green, and red). (k) The fluorescence enhanced factor (FEF) of cells treated with sensor 4 in the presence or absence of 100 mM glutathione in three emission channels (blue, green, and red). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

7436

J. Wang et al. / Biomaterials 34 (2013) 7429e7436

or glutathione are consistent with those of the cells treated with NaHS (Fig. 7g). Thus, these results indicate that sensor 4 is capable of monitoring endogenously produced H2S in living cells with three distinct emission channels. 4. Conclusions We have demonstrated that three-channel based fluorescent sensors could be constructed based on organic white lightemitting dyes using fluorescent sensor 4 as a representative example for proof-of-principle. We have judiciously designed dye 1 as an organic white light-emitting dye which can emit white light under ambient temperature and pressure. In addition, by employing dye 1 as a robust platform, we have engineered the fluorescent sensor 4, which can respond to H2S with turn-on fluorescence signals in blue, green, and red emission channels. Significantly, we have further shown that sensor 4 is capable of monitoring H2S in living cells in three different emission channels. Thus, our work extends the utility of the three-channel characteristic of organic white light-emitting dyes into the fluorescent sensing field. This work should open a new avenue for design of three- and multiple-channel based fluorescent sensors for various analytes. Acknowledgments This work was financially supported by NSFC (20872032, 20972044, 21172063), NCET (08-0175), HNNSF(13JJ3117) and the Doctoral Fund of Chinese Ministry of Education (20100161110008). Appendix A. Supplementary data Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.biomaterials.2013.06.013. References [1] Wu J, Liu W, Ge J, Zhang H, Wang P. New sensing mechanisms for design of fluorescent chemosensors emerging in recent years. Chem Soc Rev 2011;40(7):3483e95. [2] Yang Y, Zhao Q, Feng W, Li F. Luminescent chemodosimeters for bioimaging. Chem Rev 2013;113(1):192e270. [3] Chen X, Pradhan T, Wang F, Kim J, Yoon J. Fluorescent chemosensors based on spiroring-opening of xanthenes and related derivatives. Chem Rev 2012;112(3):1910e56. [4] Terai T, Nagano T. Fluorescent probes for bioimaging applications. Curr Opin Chem Biol 2008;12(5):515e21. [5] Jung HS, Han JH, Pradhan T, Kim S, Lee SW, Sessler JL, et al. A cysteine-selective fluorescent probe for the cellular detection of cysteine. Biomaterials 2012;33(3):945e53. [6] Guo Z, Kim GH, Shin I, Yoon J. A cyanine-based fluorescent sensor for detecting endogenous zinc ions in live cells and organisms. Biomaterials 2012;33(31):7818e27. [7] Chen WH, Xing Y, Pang Y. A highly selective pyrophosphate sensor based on ESIPT turn-on in water. Org Lett 2011;13(6):1362e5. [8] Santra M, Roy B, Ahn KH. A “reactive” ratiometric fluorescent probe for mercury species. Org Lett 2011;13(3):3422e5. [9] Liu B, Wang H, Wang T, Bao Y, Du F, Tian J, et al. A new ratiometric ESIPT sensor for detection of palladium species in aqueous solution. Chem Commun 2012;48(23):2867e9.

[10] Lin W, Yuan L, Tan W, Feng J, Long L. Construction of fluorescent probes via protection/deprotection of functional groups: a ratiometric fluorescent probe for Cu2þ. Chem Eur J 2009;15(6):1030e5. [11] Wang Z, Lu P, Chen S, Gao Z, Shen F, Zhang W, et al. Phenanthro[9,10-d] imidazole as a new building block for blue light emitting materials. J Mater Chem 2011;21(14):5451e6. [12] Svechkarev DA, Baumer VN, Syzova ZA, Doroshenko AO. New benzimidazolic 3-hydroxychromone derivative with two alternative mechanisms of the excited state intramolecular proton transfer reaction. J Mol Struct 2008;882(1):63e9. [13] Carturan S, Quaranta A, Maggioni G, Bonafini M, DellaMea G. 3Hydroxyflavone-based wavelength shifting systems for near UV optical sensors. Sens Actuators A 2004;113(3):288e92. [14] Yan YN, Pan W, Song HC. The synthesis and optical properties of novel 1, 3, 4oxadiazole derivatives containing an imidazole unit. Dyes Pigm 2010;86(3): 249e58. [15] Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, et al. Wallingford CT: Gaussian, Inc.; 2010. [16] Kamboj RC, Arora R, Kumar D, Sharma G. 3-alkoxy-6-chloro-2-(thiophen-3yl)-4H -chromen-4-ones: regioselective cyclization via g-hydrogen abstraction. J Photochem Photobiol A Chem 2011;220(2):124e33. [17] Klymchenko AS, Pivovarenko VG, Demchenko AP. Elimination of the hydrogen bonding effect on the solvatochromism of 3-Hydroxyflavones. J Phys Chem A 2003;107(21):4211e6. [18] Chepeleva LV, Matsakova AY, Kondratyuk ZA, Yaremenko FG, Doroshenko AO. Pyrazolyc 3-hydroxychromones: regulation of ESIPT reaction by the “flavonollike” intramolecular hydrogen bonding to carbonyl group oxygen, which dominates over the “alternative” H-bond to heterocyclic nitrogen. J Photochem Photobiol A Chem 2010;209(2):163e73. [19] Wang R. Two’s company, three’s a crowd: can H2S be the third endogenous gaseous transmitter? FASEB J 2002;16(13):1792e8. [20] Tanizawa K. Production of H2S by 3-mercaptopyruvate sulphurtransferase. J Biochem 2011;149(4):357e9. [21] Boehning D, Snyder SH. Novel neural modulators. Annu Rev Neurosci 2003;26:105e31. [22] Abe K, Kimura H. The possible role of hydrogen sulfide as an endogenous neuromodulator. J Neurosci 1996;16(3):1066e71. [23] Kamoun P. Endogenous production of hydrogen sulfide in mammals. Amino Acids 2004;26(3):243e54. [24] Shaltiel S. Thiolysis of some dinitrophenyl derivatives of amino acids. Biochem Biophys Res Commun 1967;29(2):178e83. [25] Philosof-Oppenheimer R, Pechtz I, Fridkin M. The 2,4-dinitrophenyl group for protection of hydroxyl function of tyrosine during solid-phase peptide synthesis. Int J Pept Protein Res 1995;45(2):116e21. [26] Ueno T, Urano Y, Kojima H, Nagano T. Mechanism-based molecular design of highly selective fluorescence probes for nitrative stress. J Am Chem Soc 2006;128(33):10640e1. [27] Lippert AR, New EJ, Chang CJ. Reaction-based fluorescent probes for selective imaging of hydrogen sulfidein living cells. J Am Chem Soc 2011;133(26): 10078e80. [28] Peng H, Cheng Y, Dai C, King AL, Predmore BL, Lefer DJ, et al. A fluorescent probe for fast and quantitative detection of hydrogen sulfide in blood. Angew Chem Int Ed 2011;50(41):1e5. [29] Xu W, Sheng C, CaoY He W, Wang W. Fluorescent probes for the detection of hydrogen sulfide in biological systems. Angew Chem Int Ed 2012;51(10): 2282e4. [30] Montoya LA, Pluth MD. Selective turn-on fluorescent probes for imaging hydrogen sulfide in living cells. Chem Commun 2012;48(39):4767e9. [31] Liu C, Pan J, Li S, Zhao Y, Wu LY, Berkman CE, et al. Capture and visualization of hydrogen sulfide by a fluorescent probe. Angew Chem Int Ed 2011;50(44): 10327e9. [32] Sasakura K, Hanaoka K, Shibuya N, Mikami Y, Kimura Y, Komatsu T, et al. Development of a highly selective fluorescence probe for hydrogen sulfide. J Am Chem Soc 2011;133(45):18003e5. [33] Qian Y, Karpusl J, Kabil O, Zhang SY, Zhu HL, Banerjee R, et al. Selective fuorescent probes for live-cell monitoring of sulphide. Nat Commun 2011;2: 495e502. [34] Xu Z, Xu L, Zhou J, Xu Y, Zhu W, Qian X. A highly selective fluorescent probe for fast detection of hydrogen sulfide in aqueous solution and living cells. Chem Commun 2012;48(88):10871e3. [35] Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983;65(1):55e63.