Fast and Selective Two-Stage Ratiometric Fluorescent Probes for

Article Cite This: Anal. Chem. 2017, 89, 13112−13119

pubs.acs.org/ac

Fast and Selective Two-Stage Ratiometric Fluorescent Probes for Imaging of Glutathione in Living Cells Deyan Gong,† Shi-Chong Han,‡ Anam Iqbal,†,§ Jing Qian,† Ting Cao,† Wei Liu,† Weisheng Liu,† Wenwu Qin,*,† and Huichen Guo*,‡ †

Key Laboratory of Nonferrous Metal Chemistry and Resources Utilization of Gansu Province and State Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou 730000, P. R. China ‡ State Key Laboratory of Veterinary Etiological Biology and Key Laboratory of Animal Virology of Ministry of Agriculture, Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences, Xujiaping 1, Lanzhou, Gansu Province 730046, P. R. China § Chemistry Department, University of Balochistan, Quetta, Pakistan S Supporting Information *

ABSTRACT: Two fluorescent, m-nitrophenol-substituted difluoroboron dipyrromethene dyes have been designed by nucleophilic substitution reaction of 3,5-dichloro-4,4-difluoro-4bora-3a,4a-diaza-s-indacene (BODIPY). Nonsymmetric and symmetric probes, that is. BODIPY 1 (with one nitrophenol group at the position 3) and BODIPY 2 (with two nitrophenol groups at the positions 3 and 5) were applied to ratiometric fluorescent glutathione detection. The detection is based on the two-step nucleophilic aromatic substitution of the nitrophenol groups of the probes by glutathione in buffer solution containing CTAB. In the first stage, probe 1 showed ratiometric fluorescent color change from green (λem = 530 nm) to yellow (λem = 561 nm) because of monosubstitution with glutathione (I561nm/I530nm). Addition of excess glutathione caused the second stage of ratiometric fluorescent color change from yellow to reddish orange (λem = 596 nm, I596nm/ I561nm) due to disubstitution with glutathione. Therefore, different concentration ranges of glutathione (from less to excess) could be rapidly detected by the two-stage ratiometric fluorescent probe 1 in 5 min. While, probe 2 shows single-stage ratiometric fluorescent detection to GSH (from green to reddish orange, I596nm/I535nm). Probes 1 and 2 exhibit excellent properties with sensitive, specific colorimetric response and ratiometric fluorescent response to glutathione over other sulfur nucleophiles. Application to cellular ratiometric fluorescence imaging indicated that the probes were highly responsive to intracellular glutathione. The GSH-responsive fluorescent sensors have mostly been based on the nucleophilic aromatic substitution-rearrangement reaction. Probes having labile substituent react with biothiols to form a thioether via the nucleophilic aromatic substitution. In the next step, the amino groups of Cys and Hcy, but not GSH, further substitute the sulfur through a 5- or 6-membered transition state to form the amino derivatives. The different photophysical performances of thioether- and amino-substituted fluorophores then allow the specific sensing of biothiols.10,24−26 Wang et al. reported a m-nitrophenolsubstituted cyanine dye realizing simultaneous discrimination of GSH and Cys on the basis of two distinct fluorescence turnon responses.27 Niu et al. performed a systematic study on the reactivity of BODIPY derivatives bearing a labile substituent with biothiols. They reported a probe based on p-nitrothiophenolate BODIPY derivatives for the discrimination of Cys over GSH and Hcy, the p-nitrothiophenol or p-nitrophenol group could be replaced by thiolate resulting in turn-on

I

ntracellular biothiols including glutathione (GSH), cysteine (Cys), and homocysteine (Hcy) have critical roles both in physiological and pathological processes.1,2 Among biothiols, GSH is the most abundant nonprotein thiol that has pleiotropic roles in preserving redox homeostasis in biosystems.3,4 While abnormal levels of GSH could be caused by a variety of diseases, such as liver damage, xenobiotic metabolism, AIDS, and cancer.5−7 Because of these features, it is of main concern to explore novel strategies for the monitoring of GSH in biological systems.8−10 Among the reported detection strategies, fluorescence imaging has aroused great attention.10−12 However, on account of the fact that Cys/Hcy have similar molecular structure and reactivity as GSH, the discrimination of GSH from Cys and Hcy still remains a challenge.13−18 The ratiometric fluorescent probes can provide a self-calibration of the external interference on fluorescence intensity, resulting in more effective detection both in vitro and in vivo systems.19 Until now, only a few probes provide selective ratiometric fluorescent detection for GSH.15,20,21 Moreover, compared to the single-stage response, multistage fluorescent response provides higher sensitivity and wider application.22,23 © 2017 American Chemical Society

Received: June 15, 2017 Accepted: November 21, 2017 Published: November 21, 2017 13112

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119

Article

Analytical Chemistry Scheme 1. Design of GSH Probes 1 and 2

Scheme 2. Synthetic Route for Probes 1 and 2

fluorescence response.28 However, these organic molecules usually required at least 40 min to reach the full conversion with GSH, which limited their further applications. Furthermore, these fluorescent sensors showed GSH-response without apparent fluorescence spectral shift. As a surfactant, cetyltrimethylammonium bromide (CTAB) micelles catalyze the nucleophilic substitution reaction of GSH. CTAB does not only improve the water-solubility of the probe but also speeds up the reaction and enhances sensitivity with biothiols.29,30 CTAB has also been utilized as a medium to detect H2S.31 BODIPY derivatives may be favored as fluorescent chemosensors and many of them have been synthesized to monitor a variety of analytes, such as biothiols,16,28,32 cations,33 pH,34 etc. Recently, our group have explored BODIPY fluorescent probe containing the diphenyl selenide group as the leaving group to detect H2S.35 Here, we report an investigation of the applicability of two fluorescent probes, monosubstituted BODIPY 1 and disubstituted BODIPY 2, derived from the BODIPY with mnitrophenol substituents at the 3-position and 3,5-position, for selective and rapid detection of GSH in CTAB media (Scheme 1). The m-nitrophenol at the 3-position of 1 or 2 is rapidly substituted by thiolate of GSH, whereas excessive addition of GSH promotes the second substitution at the 5-position of the probe enabling selective and sensitive GSH ratiometric fluorescence detection. Furthermore, because of the exceptional properties of the BODIPY dye, the thioether substituents induce red-shift of both the absorption and emission spectra.32,36 Interestingly, the two-step substitution of the mnitrophenol of 1 by GSH exhibited two-stage ratiometric fluorescence change. The first stage shows ratiometric fluorescent color change from green to yellow (I561nm/I530nm),

while the excessive addition of GSH caused the second stage of ratiometric fluorescent color change from yellow to reddish orange (I596nm/I561nm). To the best of our knowledge, two-stage ratiometric fluorescent responsive probe for rapid selective detection of GSH has not been reported.

■

EXPERIMENTAL SECTION Chemicals and Instruments. All chemicals and solvents were analytical and were used without further purification. NMR spectra were obtained with Bruker DRX-400 and DRX400/4 spectrometers. Mass spectra were received in E.I. Mode. pH measurements were monitored by a pH-10C digital pH meter. 1 and 2 were dissolved in MeCN and then diluted in PBS buffer (20 mM, pH 7.4) with 3 mM CTAB for all titration and selectivity experiments in a 37 °C thermostatic water bath. Preparation of Probe 1. The synthetic route for 1 is depicted in Scheme 2, and synthesis details are described below. BODIPY 3 was prepared by the previously reported method.37 To a solution of 3 (70 mg, 0.21 mmol) in 10 mL MeCN, triethylamine (500 μL, 3.61 mmol) and m-nitrophenol (28.90 mg, 0.21 mmol) were added under continuous stirring at room temperature. After 3 h, the solvent was evaporated. After silica gel column purification [petroleum ether/ethyl acetate, 5:1 (v/ v)], probe 1 (65 mg, 0.15 mmol, 70% yield) was obtained as a dark solid. mp: 116−118 °C. Rf: 0.56 (petroleum ether/ethyl acetate = 2:1). 1H NMR (CDCl3): δ 5.845 (d, 1H, J = 4.4 Hz, H-a), 6.373 (d, 1H, J = 4.0 Hz, H-b), 6.746 (d, 1H, J = 4.0 Hz, H-c), 6.888 (d, 1H, J = 4.8 Hz, H-d), 7.498 (m, 5H (Ph)), 7.627 (m, 2H (Ph)), 8.161 (m, 2H (Ph)). 13C NMR (CDCl3): δ 105.59, 115.64, 117.23, 121.11, 126.56, 128.63, 129.18, 129.68, 130.48, 130.68, 131.08, 132.64, 133.88, 141.03, 143.19, 149.27, 154.75, 164.57. Mass spectrum (ESI), m/z 440.1 (M + 13113

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119

Article

Analytical Chemistry 1); 462.1 (M + Na); 420.1 (M − F, 100%) (C21H13BClF2N3O3 requires m/z 439.0707). Preparation of Probe 2. To a solution of 3 (70 mg, 0.21 mmol) in 10 mL of MeCN, triethylamine (500 μL, 3.61 mmol) and m-nitrophenol (116 mg, 0.84 mmol) were added, and then it was heated to reflux. After about 4 h, the solvent was evaporated. After silica gel column purification [petroleum ether/ethyl acetate, 3:1 (v/v)], probe 2 (72 mg, 0.13 mmol, 63% yield) was obtained as a dark solid. mp: 128−130 °C. Rf: 0.44 (petroleum ether/ethyl acetate = 2:1). 1H NMR (CDCl3): δ 5.82 (d, 2H, J = 4 Hz, H-a), 6.83 (d, 2H, J = 4 Hz, H-b), 7.59 (m, 5H (Ph)), 7.62 (m, 4H (Ph)), 8.10 (m, 4H (Ph)). 13C NMR (CDCl3): δ 103.94, 115.10, 120.52, 126.10, 128.61, 130.49, 130.90, 131.40, 132.74, 143.16, 149.23, 155.44, 162.18. Mass spectrum (ESI), m/z 543.1 (M + 1); 523.1 (M − F, 100%) (C27H17BF2N4O6 requires m/z 542.1209). Preparation of 1-GSH and 1-2GSH. The mixture of 1 (5 mg, 0.015 mmol) and GSH (60 mg, 0.20 mmol) in 9 mL of MeCN−H2O (2:1, v/v) was stirred and refluxed at 95 °C. After about 12 h, the solvent was dried by evaporation. Then the obtained solid residue was purified by silica gel chromatography with MeOH as eluent to afford 1-GSH (monosubstituted 1 with GSH) as a reddish brown solid; and then with increased solvent polarity (MeOH/H2O = 10:1, v/v) 1−2GSH (disubstituted 1 with GSH) was obtained as a dark purple solid. Samples for ESI-MS analysis were prepared by dissolving 1-GSH in MeOH, and 1−2GSH in H2O and then these solutions were filtered through a 0.45 μm polyvinylidene fluoride (PVDF) syringe driven filter. Mass spectrum (ESI): 1GSH (C25H25BClF2N5O6S requires m/z 607.1275) m/z 607.5 (M). HRMS: calcd for 1−2GSH (C35H42BF2N8O12S2, M + 1) 879.2419, found 879.2431. Preparation of 2-2GSH. Probe 2 (5 mg, 0.015 mmol), CTAB (187 mg, 0.513 mmol). and GSH (60 mg, 0.20 mmol) were added to a 9 mL of MeCN−H2O (2:1, v/v) solution. The reaction mixture was stirred at 50 °C for 2 h. Under reduced pressure, the resulting solution was concentrated. Then the residue was purified by silica gel chromatography with MeOH as eluent to remove CTAB, and then with increased eluent polarity (MeOH/H2O = 10:1, v/v) 2-2GSH (disubstituted 2 with GSH) was obtained as a dark purple solid. The monosubstituted product of 2 with GSH could not be separated from CTAB, so, it could not be characterized. Mass spectrum (ESI): 2-2GSH (C35H41BF2N8O12S2 requires m/z 878.2346) m/z 901.2 (M + Na); 879.2 (M + 1); 859.2 (M − F, 100%); 839.2 (M − F − HF).

withdrawing effect from the two nitrophenol moieties of 2. Table S1 summarizes the spectroscopic/photophysical data of 1 and 2. Figure 1 shows the absorption of 1 and 2 changing with the addition of GSH. In the first step reaction for 1, with increasing

Figure 1. (a) Absorption changes of 1 and (b) 2 (10 μM) with GSH. Inset: (a, b) Visible changes of 1 and 2 with GSH. All spectra were obtained after 5 min of incubation with GSH.

GSH concentration in the range of 0−15 μM, the maximum of absorption at 512 nm decreased, while the intensity of a new band at 543 nm increased. The associated absorption changed at the 344/512/543 nm with four isosbestic points at 314 nm, 371, 475, and 529 nm resulting from the substitution reaction with GSH at the 3-position of the probe. It should be noted that probes 1 and 2 have different structure after the first step reaction. Probe 1 is substituted by GSH to form a thioether and chlorine at the positions 3 and 5, whereas probe 2 bears thioether and nitrophenol. This difference in structure is reflected to different fluorescent response of probes 1 and 2 after the first step reaction. Thus, simultaneously with the changes in absorption spectra, 1 displayed an intense fluorescence emission at approximately 530 nm (green, λex = 515 nm) which was quenched upon addition of GSH and a new emission band at 561 nm (yellow) grew progressively. An obvious red-shift with an isoemissive point at 548 nm was readily detected in the GSH concentration range of 0−15 μM (Figure 2a). At the GSH concentration of 10 μM, the fluorescence quantum yield is 0.39. The ratio of fluorescence intensities at 530 and 561 nm for 1 grew linearly with the GSH

■

RESULTS AND DISCUSSION Analytical Performance of Probes 1 and 2 for GSH Ratiometric Fluorescent Detection. Figure S1 and Table S1 shows the absorption and emission spectra of 1 and 2. The absorption spectrum of 1 displayed a narrow band centered at 512 nm (for 2 519 nm), ascribed to the 0−0 vibrational band of S0−S1 transition. A less obvious shoulder at shorter wavelengths belongs to 0−1 vibrational band of the S0−S1 transition. Furthermore, the broader bands of 1 and 2 at 344 nm are ascribed to the S0−S2 transition. The fluorescence emission band of 1 is centered at 530 nm (for 2 535 nm) and the fluorescence quantum yield ϕf is 0.132 (for 2 0.029). The absorption and fluorescence spectra of 1 and 2 are of similar shape as those of reported phenoxy (phenyl-O-, λabs = 520 nm and λem = 530 nm) disubstituted BODIPY derivatives.38 The low ϕf value for 2 can be attributed to a strong electron 13114

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119

Article

Analytical Chemistry

reaction with GSH were of similar shape as those of the described sulfur (ethyl 2-thioacetate −S−, λabs = 540 nm and λem = 550 nm) monosubstituted BODIPY derivatives.36 In the second step reaction, the excess addition of GSH (15− 140 μM) facilitated the substitution reaction of the chlorine and nitrophenol at the 5-position of the probes 1 and 2, respectively. This led to the decrease of absorption intensity at 344 and 512 nm and increase of the absorption intensity at 381 and 579 nm (Figure 1) with three isosbestic points at 314, 415, and 529 nm. The spectral changes can also be seen in the day light by a naked eye, as the solution color change from light yellow to pink (inset of Figure 1a). Probe 2 showed similar changes in the absorption spectra as 1 in the GSH concentration range of 0−100 μM, and almost the same UV spectra were obtained after the reactions of 1 and 2 with GSH were completed (Figures 1b and S1b, c). The results indicate that 1 and 2 can serve as “naked-eye” sensor for GSH. In the fluorescence spectra upon addition of excess GSH (15−140 μM), the emission intensity of 1 at 561 nm declined, while a new emission at 596 nm (reddish orange) grew (λex = 543 nm), resulting in a distinct red-shift with an isoemissive point at 585 nm (Figure 2b). The fluorescence quantum yield at GSH concentration of 500 μM is 0.97. Similarly, for probe 2 in the GSH concentration range of 0−50 μM, the weak emission band centered at 535 nm (green) (λex = 515 nm) was quenched and the new broad emission at 596 nm (reddish orange) increased in intensity (Figure 2c). Non-observation of the emission at 561 nm was ascribed to the quenching effect of the mnitrophenol at the 5-position of probe 2. From the linear equation in Figure S2, DL of probe 1 for GSH (15−100 μM) was calculated to be 1.3 × 10−7 M (for 2 was 7.6 × 10−8 M (5− 50 μM)). Consequently, both probes 1 and 2 are potentially useful for the quantitative determination of GSH in solution. As shown in Table S2, the DL obtained by our method is well below the listed values. The absorption and emission of 1 and 2 in the presence of excess GSH are of similar shape as those of reported sulfur (ethyl 2-thioacetate −S−, λabs = 570 nm and λem = 585 nm) disubstituted BODIPY derivatives.36 The time-dependent fluorescence response of 1 and 2 (10 μM) in the presence of GSH were investigated. It is worth noting that the fluorescence increase upon addition of 50 GSH equiv evidently illustrated fast response within 30 s and reached the maximum in the first 60 s, showing extremely fast reaction rate of 1 with GSH (Figure 3). As shown in Figure S3, 2 demonstrated similar time-dependent fluorescence response, the emission intensity at 596 nm of 2 quickly increased and reached the maximum in the first 150 s. Therefore, the exceptional time responsive detection of GSH, within 5 min, was achieved by use of probes 1 and 2. Sensitivity and Selectivity toward GSH. To investigate the selectivity in solution, various representative intracellular species were added into the solution of probes 1 or 2 (10 μM). As shown in Figure 4, only GSH caused a clear fluorescence red-shift from 530 to 596 nm, the fluorescence intensity ratio of 1 at 596 and 530 nm (I596nm/I530nm) was about 16, while 1 did not respond to any other analyte under the same condition (I596nm/I530nm were about 0.04). Similar results were obtained with 2 (Figure S4). Figure S5a, b show the emission response spectra of 1 and 2 upon addition of 50 equiv of biothiols and Na2S. It clearly demonstrates that the ratiometric fluorescence response of 1 and 2 with GSH is significantly different from that to other sulfur nucleophiles which quenched the fluorescence of the probes after 10 min. After 24 h, the

Figure 2. (a, b) Fluorescence changes of 1 (10 μM) with GSH (λex = 515 nm) and excess GSH (λex = 543 nm). (c) Fluorescence changes of 2 (10 μM) with GSH (λex = 515 nm). Inset: (a−c) Fluorescence change under a 365 nm UV lamp of 1 and 2 with GSH. All spectra were obtained after 5 min of incubation of 1 and 2 with GSH.

concentration (0−15 μM). From the linear equation in Figure S2, the DL (detection limit) of probe 1 for GSH was calculated to be 6.6 × 10−8 M (0−6 μM) and 7.6 × 10−9 M (7−15 μM) (signal-to-noise ratio (S/N) = 3). Because of the strong electron-withdrawing effect of the nitrophenol moieties at the BODIPY 5-position for 2, the increase of the new emission at 561 nm was not as strong compared to 1 (Figures 2b and S1d, e). The absorption and emission maxima of 1 and 2 after 13115

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119

Article

Analytical Chemistry

were determined from a single decay surface collected 5 min after the GSH addition (λex = 515 nm). The SPT measurements for 1 and 2 gave rise to the fluorescence decays that were best fit to two exponential functions. The decay times τ1 and τ2 for 1 and 2 are listed in Table S3. The decay time τ1 = ∼1.26 ns (for 2 1.27 ns) is attributed to the lifetime of the BODIPY fluorophore (The lifetime of phenoxy disubstituted BODIPY derivatives is 1.1−2.0 ns depending on the solvents used),38 while the τ2 = ∼0.51 ns (for 2 was 0.31) is ascribed to the lifetime of the charge transfer (CT) excited state of probes. Because of the stronger electron withdrawing effect from the two nitrophenol moieties, the shorter decay time τ2 = ∼0.31 ns with ∼94% amplitude was found for 2. After reaction of 1 and 2 with GSH in the concentration range that enables only one substitution reaction, the decay curve recorded 5 min after the GSH addition (Figure 5 and Figure 3. Time-dependent fluorescent response at 596 nm (λex = 515 nm) for 1 (10 μM) upon addition of 50 equiv of GSH.

Figure 4. Specific selectivity of 1 (10 μM) reacted with 0.5 mM of different kinds of species. All spectra were obtained after 10 min of incubation with the analytes, λex = 515 nm. (1) Probe 1, (2) GSH, (3) Na2S, (4) Cys, (5) Hcy, (6) Pro, (7) Ile, (8) Ala, (9) His, (10) Glu, (11) Tyr, (12) Lys, (13) Met, (14) Asp, (15) Phe, (16) Arg, (17) Gly, (18) Ser, (19) Na2S2, and (20) Na2S4.

fluorescence spectra did not change in appearance or intensity, showing good stability of the GSH-adducts. Therefore, the discrimination of GSH from Cys and Hcy can be achieved by probes 1 and 2. Additionally, we checked the influence of pH on the fluorescence intensity ratio of 1 (I596nm/I530nm) and 2 (I596nm/ I535nm) in the absence and presence of 50 equiv of GSH. As shown in Figure S6, when no GSH added, the fluorescence intensity ratio of 1 and 2 was slightly changed in the pH region 4−10. The fluorescence ratio of 1 and 2 upon the incubation with GSH demonstrated significantly increment and maintains relatively steady in the pH range 6−10. However, at the pH 4− 5 where GSH is not deprotonated, the substitution reaction of the m-nitrophenol group was slower and the fluorescence ratio increased slightly. Thus, 1 and 2 can be applied to the GSH detection at pH range (pH 6−10), within which most biological samples can be sensed. The pH variations demonstrate slight disturbance to the GSH detection by probes 1 and 2. Fluorescence Decay Traces. Decay traces of 1 and 2 at different contents of biothiols were obtained by single-photon timing (SPT). For solutions containing 1 and GSH or 2 and GSH mixture, the decay times τi and pre-exponential factors αi

Figure 5. Representative fluorescence decays of (a) 1 and (b) 2 (10 μM) in the absence and presence of different concentration of biothiols. All traces were obtained after 5 min of incubation with biothiols in PBS buffer (20 mM, pH 7.4) with 3 mM CTAB.

Figure S7). Upon addition of 2 equiv GSH, for the 1+GSH mixture, the decay components of ∼1.26 ns and ∼0.51 ns vanished and two new decay times were detected with τ1 ≈ 4.1 ns and τ2 ≈ 2.2 ns. The longer decay time τ1 ≈ 4.1 ns (for 2 4.51 ns) can be attributed to the lifetime of monosubstituted product with GSH, while the other decay τ2 ≈ 2.2 ns may be due to the aggregation of monosubstituted product in PBS buffer with 3 mM CTAB. 13116

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119

Article

Analytical Chemistry Scheme 3. Chemical Structures and Proposed Mechanism of 1 with Biothiols

1 and 2 with Cys/Hcy in Figure S10, show similar changes as in the reaction with GSH, indicate the disubstitution reaction also occurred after excess Cys/Hcy addition. In the first stage, the absorption at ∼512 nm of 1 and 2 declined, while new bands increased at ∼554 nm. After the addition of excess Cys/Hcy, bands at ∼554 nm declined and new bands appeared at ∼592 nm. These new bands at ∼592 nm are similar to the previously reported disubstituted BODIPY derivatives bearing nitrogen (piperidine −N−, λabs = 584 nm and aniline −N−, λabs = 594 nm).36 The proposed mechanism of probes 1 and 2 with Cys/ Hcy are summarized in Scheme 3 and Figure S8. Since the photophysical properties of BODIPY derivatives are sensitive to substituents, the thioether GSH BODIPY adducts are spectroscopically distinguishable from Cys or Hcy adducts.26,39,40 To inspect the effect of the CTAB in GSH detection, comparative experiments were carried out. Upon addition of 50 equiv of GSH to the 1 or 2 PBS buffer (20 mM, pH 7.4) solution containing 3 mM CTAB, the fluorescence intensity at 596 nm reached a maximum in 5 min (Figure S11a, b). When no CTAB was used, DMSO had to be added to dissolve probes 1 and 2, the ratio of DMSO/PBS buffer (2:8, v/v) was kept constant in all our tests. As shown in Figure S11, the emission intensity at 596 nm of 1 and 2 increased slowly and did not reach the maximum even 23 min after the addition of 50 equiv of GSH in DMSO/PBS buffer (2:8, v/v, 20 mM, pH 7.4) without CTAB. It is evident that CTAB micelles exhibit a catalytic effect on the reaction of 1 and 2 with GSH. The catalytic effect can be explained by the simultaneous presence of electrostatic and hydrophobic interactions in the reaction system. Since GSH bearing two carboxyl groups is negatively charged, it is able to strongly bind with and penetrate into positively charged CTAB micelles.29,30,41 Thus, GSH is more promoted to attack the substrate inside of the micellar aggregates, resulting in the release of m-nitrophenol group. Cellular Imaging. Inspired by the favorable properties of 1 and 2 in vitro, the applicability of the probes was investigated in confocal fluorescence imaging applications using a BHK cell model. Before the bioimaging and biosensing, MTS assays were employed to check the cytotoxicity of 1 and 2, revealing that the probes had low cytotoxicity (Figure S12). The BHK cells were treated with 1 or 2 (15 μM, 0.5 h) and 0.5 mM CTAB to evaluate the potential in bioimaging of GSH. As expected, the

Upon addition of excess GSH (10 equiv) when both chlorines were substituted with GSH, the fluorescence decay for the 1+GSH mixture became longer and was best fit to twoexponential function with two new decay times τ1 ≈ 7.5 ns and τ2 ≈ 5.4 ns. On the contrast, fluorescence decay of 2 at high GSH concentration is single exponential with the lifetime of 7.39 ns that corresponds to the emission of disubstituted product (λex = 543 nm, λem = 600 nm). The fastest decay τ2 = ∼ 0.31 ns was not detected since the second nitrophenol group was substituted by GSH for 2 at the second step reaction. In general, the products of the reaction of 1 and 2 with GSH have significantly longer decay times. Upon addition of 10 equiv of Cys or Hcy, the amplitude of the previous decay times varied and a new decay time (∼2.5 ns) appeared which demonstrate the appearance of new substances. The data show that these new substances are obviously different from 1+GSH or 2+GSH mixture. Mechanistic Study. Substitution of the nitrophenol group in probe 1 by GSH causes a significant 66 nm (for 2 61 nm) red shift in the fluorescence spectrum in agreement with the known observation that thioether substituents red-shift both the absorption and emission spectra of BODIPY dyes.28,32 Scheme 3 and Figure S8 show the proposed mechanism of probes 1 and 2 with biothiols, respectively. The MS Spectra (see Supporting Information) of 1 and 2 with GSH were detected to identify the substituted products. The expected products are 1-GSH, 1-2GSH, and 2-2GSH. In the MS and HRMS spectra, the molecular ions were detected at m/z 607.5 corresponding to [1-GSH] (calcd = 607.1); m/z 879.2431 corresponding to [1-2GSH + H] (calcd = 879.2419); and m/z 879.2 corresponding to [2-2GSH + H] (calcd = 879.2). In Figure S9, the absorption and fluorescence spectra of 2-2GSH (the isolated product) are shown to be almost the same as those of probes 1 and 2 with GSH before the purification. It is thus confirmed that the m-nitrophenol groups of 1 and 2 were substituted by sulfhydryl groups after the treatment with GSH. The disubstitution reaction occurred after excess GSH addition. By contrast, for Cys and Hcy, an intramolecular rearrangement through a 5- or 6-membered transition state would instantly take place forming the amino-substituted BODIPY. Usually, the amino substituted BODIPY dyes with strong electron-donating groups (Cl or nitrophenol group) are less reactive in the nucleophilic substitution.26 However, the absorption spectra of 13117

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119

Article

Analytical Chemistry fluorescent confocal images demonstrated that the fluorescence of probes 1 and 2 in the cytoplasm with GSH, where the bathochromic shift could be clearly seen by different color (Figure 6). The microscopy images showed that the cells

and enabled the detection of both endogenous and exogenous GSH in living cells.

■

CONCLUSIONS In summary, we have designed and synthesized two fluorescent ratiometric probes for GSH detection based on the two-step substitution reaction. Probe 1 can be used for two-stage ratiometric fluorescent detection of three concentration ranges of GSH (0−6, 7−15, and 15−100 μM) with significant stepwise red shifts in emission and remarkable emission ratio changes in 5 min. Probe 2 shows single-stage ratiometric fluorescent detection to GSH. Color changes of the probes allow for the detection of GSH by the naked eye. Moreover, fluorescent probes 1 and 2 show excellent selective sensitivity toward GSH compared to other biothiols and they can be applied to detect intracellular GSH. The strategy of the work represented here may offer an innovative approach for the development of a wide range of applied GSH multistage fluorescent ratiometric probes.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.7b02311. Spectroscopic/photophysical data of 1 and 2 in addition of GSH, comparison of the proposed method with other methods for GSH detection, photophysical properties of 1 and 2 in the absence and in addition of GSH, absorption and fluorescent spectra of 1 and 2 in the absence and presence of GSH, linear calibration curve of fluorescence intensity ratio of 1 and 2 with GSH, time course of the response at 596 nm of time-dependent fluorescence spectra of 2 with GSH, specific selectivity experiment of 2, emission response of 1 and 2 upon addition of 50 equiv of biothiols and Na2S, fluorescence intensity of 1 and 2 in the absence and presence of GSH as a function of pH, experimental fluorescence decay trace, proposed mechanism of 2 with biothiols, spectra comparison of probe 1 + GSH, 2 + GSH, and the isolated product 2-2GSH, absorption changes of 1 and 2 with Cys/Hcy, comparison of time course of the response at 596 nm of 1 and 2 with 50 equiv of GSH in buffer without and with CTAB, effects of 1 and 2 at varied concentrations on the viability of BHK cells, and NMR and MS spectrum spectra (PDF)

Figure 6. Confocal laser scanning microscopic images of (a) 1 and (b) 2: the first row of panels a and b are fluorescence images of BHK cells incubated with 1 and 2 (15 μM) and NEM (500 μM), respectively; the second row are fluorescence pictures of BHK cells treated with 1 and 2, respectively; the third row are fluorescence pictures of BHK cells of 1 and 2 pretreated with 100 μM GSH, respectively. The addition concentration of CTAB in all cell samples are 0.5 mM. Green channel is at 520−560 nm; red channel is at 580−620 nm. Scale bar = 10 μm.

■

treated with 1 and N-ethylmaleimide (NEM, 500 μM, 0.5 h, a scavenger of GSH),42 displayed fluorescence in green channel (520−560 nm) and no fluorescence in red channel (580−620 nm) when excited at 514 nm (Figure 6a); for 2, green fluorescence was weak. However, when the cells were not treated with NEM, fluorescence was seen in the green channel, as well as the red channel attributed to the cellular GSH. For 2, the fluorescence was only seen in the red channel. If the cells were incubated with exogenous GSH (100 μM, 0.5 h), and then were stained by probes 1 or 2, the fluorescence in the red channel was obviously increased and no fluorescence was detected in the green channel. These data definitely demonstrate that probes 1 and 2 showed low cytotoxicity, excellent membrane permeability and good biocompatibility

AUTHOR INFORMATION

Corresponding Authors

*Fax: +86-931-8912582. E-mail: [email protected]. *E-mail: [email protected]. ORCID

Wenwu Qin: 0000-0002-9782-6647 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors are grateful to Prof. Albert M. Brouwer and Nikola Basarić for providing us with great efforts in the manuscript revision. This work was supported by the State Key Laboratory of Veterinary Etiological Biology, Lanzhou Veterinary Research 13118

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119

Article

Analytical Chemistry

(32) Niu, L.-Y.; Guan, Y.-S.; Chen, Y.-Z.; Wu, L.-Z.; Tung, C.-H.; Yang, Q.-Z. J. Am. Chem. Soc. 2012, 134, 18928−18931. (33) Yang, C.; Gong, D.; Wang, X.; Iqbal, A.; Deng, M.; Guo, Y.; Tang, X.; Liu, W.; Qin, W. Sens. Actuators, B 2016, 224, 110−117. (34) Deng, M.; Yang, C.; Gong, D.; Iqbal, A.; Tang, X.; Liu, W.; Qin, W. Sens. Actuators, B 2016, 232, 492−498. (35) Gong, D.; Zhu, X.; Tian, Y.; Han, S.-C.; Deng, M.; Iqbal, A.; Liu, W.; Qin, W.; Guo, H. Anal. Chem. 2017, 89, 1801−1807. (36) Rohand, T.; Baruah, M.; Qin, W.; Boens, N.; Dehaen, W. Chem. Commun. 2006, 266−268. (37) Baruah, M.; Qin, W.; Basaric, N.; De Borggraeve, W. M.; Boens, N. J. Org. Chem. 2005, 70, 4152−4157. (38) Rohand, T.; Lycoops, J.; Smout, S.; Braeken, E.; Sliwa, M.; Van der Auweraer, M.; Dehaen, W.; De Borggraeve, W. M.; Boens, N. Photoch. Photobio. Sci. 2007, 6, 1061−1066. (39) Wang, F.; Guo, Z.; Li, X.; Li, X.; Zhao, C. Chem. - Eur. J. 2014, 20, 11471−11478. (40) Niu, L.-Y.; Zheng, H.-R.; Chen, Y.-Z.; Wu, L.-Z.; Tung, C.-H.; Yang, Q.-Z. Analyst 2014, 139, 1389−1395. (41) Huang, X.; Dong, Z.; Liu, J.; Mao, S.; Xu, J.; Luo, G.; Shen, J. Langmuir 2007, 23, 1518−1522. (42) Yellaturu, C. R.; Bhanoori, M.; Neeli, I.; Rao, G. N. J. Biol. Chem. 2002, 277, 40148−40155.

Institute, Chinese Academy of Agricultural Sciences, funds of the Ministry of Science and Technology of China (2014DFA31890) and the Natural Science Foundation of China (no. 21771092).

■

REFERENCES

(1) Reddie, K. G.; Carroll, K. S. Curr. Opin. Chem. Biol. 2008, 12, 746−754. (2) Weerapana, E.; Wang, C.; Simon, G. M.; Richter, F.; Khare, S.; Dillon, M. B.; Bachovchin, D. A.; Mowen, K.; Baker, D.; Cravatt, B. F. Nature 2010, 468, 790−795. (3) Economou, M.; Tsatra, I.; Athanassiou-Metaxa, M. Pediatr. Hematol. Oncol. 2003, 20, 493−495. (4) Passweg, J. R.; Tichelli, A. Haematologica 2009, 94, 310−312. (5) Tapiero, H.; Townsend, D.; Tew, K. Biomed. Pharmacother. 2003, 57, 134−144. (6) Hidalgo, J.; Garvey, J.; Armario, A. J. Pharmacol. Exp. Ther. 1990, 255, 554−564. (7) Saito, C.; Zwingmann, C.; Jaeschke, H. Hepatology 2010, 51, 246−254. (8) Fu, Z.-H.; Han, X.; Shao, Y.; Fang, J.; Zhang, Z.-H.; Wang, Y.-W.; Peng, Y. Anal. Chem. 2017, 89, 1937−1944. (9) Huang, Y.; Zhou, J.; Feng, H.; Zheng, J.; Ma, H.-M.; Liu, W.; Tang, C.; Ao, H.; Zhao, M.; Qian, Z. Biosens. Bioelectron. 2016, 86, 748−755. (10) Niu, L.-Y.; Chen, Y.-Z.; Zheng, H.-R.; Wu, L.-Z.; Tung, C.-H.; Yang, Q.-Z. Chem. Soc. Rev. 2015, 44, 6143−6160. (11) Zhou, Y.; Yoon, J. Chem. Soc. Rev. 2012, 41, 52−67. (12) Jung, H. S.; Chen, X.; Kim, J. S.; Yoon, J. Chem. Soc. Rev. 2013, 42, 6019−6031. (13) Iqbal, A.; Iqbal, K.; Xu, L.; Li, B.; Gong, D.; Liu, X.; Guo, Y.; Liu, W.; Qin, W.; Guo, H. Sens. Actuators, B 2018, 255, 1130−1138. (14) Zhang, H.; Liu, R.; Liu, J.; Li, L.; Wang, P.; Yao, S. Q.; Xu, Z.; Sun, H. Chem. Sci. 2016, 7, 256−260. (15) Liu, Z.; Zhou, X.; Miao, Y.; Hu, Y.; Kwon, N.; Wu, X.; Yoon, J. Angew. Chem., Int. Ed. 2017, 56, 5812−5816. (16) Gong, D.; Tian, Y.; Yang, C.; Iqbal, A.; Wang, Z.; Liu, W.; Qin, W.; Zhu, X.; Guo, H. Biosens. Bioelectron. 2016, 85, 178−183. (17) Wang, L.; Chen, X.; Cao, D. Sens. Actuators, B 2017, 244, 531− 540. (18) Wang, L.; Du, J.; Cao, D. Sens. Actuators, B 2014, 205, 281−288. (19) Lee, M. H.; Kim, J. S.; Sessler, J. L. Chem. Soc. Rev. 2015, 44, 4185−4191. (20) Zhang, H.; Wang, C.; Wang, K.; Xuan, X.; Lv, Q.; Jiang, K. Biosens. Bioelectron. 2016, 85, 96−102. (21) Zhang, H.; Wang, C.; Wang, G.; Wang, K.; Jiang, K. Biosens. Bioelectron. 2016, 78, 344−350. (22) Biswas, S.; Gangopadhyay, M.; Barman, S.; Sarkar, J.; Singh, N. P. Sens. Actuators, B 2016, 222, 823−828. (23) Amendola, V.; Bergamaschi, G.; Boiocchi, M.; Fabbrizzi, L.; Mosca, L. J. Am. Chem. Soc. 2013, 135, 6345−6355. (24) Kim, Y.; Mulay, S.; Choi, M.; Yu, S. B.; Jon, S.; Churchill, D. Chem. Sci. 2015, 6, 5435−5439. (25) Chen, W.; Luo, H.; Liu, X.; Foley, J. W.; Song, X. Anal. Chem. 2016, 88, 3638−3646. (26) Niu, L.-Y.; Yang, Q.-Q.; Zheng, H.-R.; Chen, Y.-Z.; Wu, L.-Z.; Tung, C.-H.; Yang, Q.-Z. RSC Adv. 2015, 5, 3959−3964. (27) Wang, X.; Lv, J.; Yao, X.; Li, Y.; Huang, F.; Li, M.; Yang, J.; Ruan, X.; Tang, B. Chem. Commun. 2014, 50, 15439−15442. (28) Niu, L.-Y.; Guan, Y.-S.; Chen, Y.-Z.; Wu, L.-Z.; Tung, C.-H.; Yang, Q.-Z. Chem. Commun. 2013, 49, 1294−1296. (29) Guo, Y.; Yang, X.; Hakuna, L.; Barve, A.; Escobedo, J. O.; Lowry, M.; Strongin, R. M. Sensors 2012, 12, 5940−5950. (30) Wang, L.; Chen, H.; Wang, H.; Wang, F.; Kambam, S.; Wang, Y.; Zhao, W.; Chen, X. Sens. Actuators, B 2014, 192, 708−713. (31) Cao, X.; Lin, W.; Zheng, K.; He, L. Chem. Commun. 2012, 48, 10529−10531. 13119

DOI: 10.1021/acs.analchem.7b02311 Anal. Chem. 2017, 89, 13112−13119