Single-Atom Switching as a General Approach to Designing Colorimetric and Fluorogenic Probes for Mercury Ions

Abstract

By performing a single-atom replacement within common fluorophores, we have developed a facile and general strategy to prepare a broad-spectrum class of colorimetric and fluorogenic probes for the selective detection of mercury ions in aqueous environments. Thionation of carbonyl groups from existing fluorophore cores results in a great reduction of fluorescence quantum yield and loss of fluorescence emission. The resulting thiocaged probes are efficiently desulfurized to their oxo derivatives in the presence of mercury ions, leading to pronounced changes in chromogenic and fluorogenic signals. Because these probes exhibit high selectivity, excellent sensitivity, good membrane-permeability, and rapid responses towards mercury ions, they are suitable for visualization of mercury in both aqueous and intracellular environments.

Graphical Abstract

1. INTRODUCTION

Mercury is a serious environmental pollutant and toxin,[1] linked to severe pathogenic effects on the nervous system, immune system, and kidneys.[2,3] Thus, the development of new and generally-applicable methods for the selective detection and visualization of Hg²⁺ ions and other heavy metals is critically important.[4–10] Among a range of sophisticated Hg²⁺ sensors, including nanoparticles,[11–15] polymers,[16–19] enzymes,[20–24] and DNA-based probes,[25–29] chemosensors are particularly attractive because of their high sensitivity, low cost, and ease of use.[30–36] Based on the strong thiophilic affinity of Hg²⁺, fluorescent changes associated with Hg²⁺-induced cyclization,[37–40] coordination,[41–44] ring opening complexation,[45–48] and hydrolysis[49–53] have been used in the design of Hg²⁺ sensors. In particular, the hydrolysis reaction of thiocarbonyl group mediated by S-Hg bond formation is of great interest due to ease of design, excellent selectivity, and sensitivity to Hg²⁺ ions. Fluorophore scaffolds with short wavelength emission, including coumarin,[53] and 8-hydroxyquinoline,[54] have been exclusively used for the design of Hg²⁺ sensors based on this hydrolysis mechanism.

In 2019, we employed a single sulfur-for-oxygen atom substitution as a general approach to preparing photoactivatable probes or heavy-atom free photosensitizers.[55–57] We found that a single sulfur-for-oxygen replacement in the carbonyl groups of existing fluorophore cores (e.g., Nile red, naphthalimide, acridone, and coumarin) leads to a substantial reduced fluorescence due to a photoinduced electron transfer-quenching mechanism. Compared to the corresponding carbonyl compounds, thio-based fluorophores exhibit significant red-shifts and enhanced absorption due to the reduced energy gap between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbitals (LUMO) in these fluorophores. The significant decrease in quantum yields of all tested thio-based fluorophores suggests that replacement of a single oxygen atom with sulfur can be a general approach for preparing fluorogenic Hg²⁺ sensors with diverse structures.

In this study, we have used a single sulfur-for-oxygen atom substitution as a general strategy to prepare a class of colorimetric and fluorogenic chemosensors for the selective detection of Hg²⁺ in aqueous environments (Figure 1). Upon addition of Hg²⁺, thio-caged sensors are efficiently desulfurized to their oxo derivatives, thus restoring strong emission of the fluorophores across a broad spectral range (410 nm - 625 nm). Furthermore, our newly developed thio-imide probes exhibit a much faster response to Hg²⁺ than the previously reported thiocoumarins, probably due to stronger electron-donating ability of the nitrogen atom compared to oxygen atom.

Fig. 1. General reaction mechanism of the desulfurization reaction.

Changes in the fluorescence intensity of thiocaged fluorophores (5.0 × 10⁻⁶ M) were monitored in H₂O/CH₃CN (50:50) in the presence of different metal ions (5.0 × 10⁻⁴ M). The following excitation and emission wavelengths are used: SACD (390 nm, 408 nm), SCou (375 nm, 467 nm), SDMAP-NH (370 nm, 504 nm), SDMN-NH (430 nm, 540 nm), SDMNP-NH (370 nm, 513 nm), SNile Red (550 nm, 700 nm).

2. EXPERIMENTAL DETAILS

2.1. Materials and apparatus

¹H and ¹³C NMR spectra were measured on a Bruker AVANCE III HD 600 MHz spectrometer in CDCl₃ and D₂O/CD₃CN 1:1. UV-vis absorption spectra were measured on a Thermo Scientific Evolution 220 UV-Vis Spectrophotometer. Fluorescence measurements were conducted on a Horiba FluoroMax-4 spectrofluorometer. ESI mass spectroscopy was performed on a Bruker MicroToF ESI LC-MS System. The cation solutions were prepared from NaCl, MgCl₂, KCl, CaCl₂, MnCl₂, FeCl₃, CoCl₂, NiCl₂, CuCl₂, ZnCl₂, Pd(OAc)₂, AgNO₃, CdCl₂, SnCl₂, CsCO₃, PtCl₂, PbCl₂; in 10 mM HEPES buffer (pH = 7.4), respectively at a concentration of 1 mM.

2.2. Synthesis

SCou, SDMAP, and SNile Red were synthesized according to the previously reported protocol.[55] DMN-O and DMNP-O were synthesized using the reported synthetic route.[56] DMN-NH and DMNP-NH were synthesized using the procedure reported previously.[58]

DMN-NH

DMN-O (80 mg, 0.37 mmol) and urea (45 mg, 0.74 mmol, 2 eq) were heated at 170 °C for 2 h in a Schlenk tube. The mixture was cooled to room temperature, triturated with water (10 mL) and extracted three times with DCM. The organic phase was dried over Na₂SO₄, and the solvent was removed under reduced pressure. The residue was purified by silica gel column chromatography (30% ethyl acetate in hexanes) to afford DMN-NH (73 mg, 82% yield) as a yellow solid. ¹H NMR (600 MHz, Chloroform-d) δ 8.56 (dd, J = 7.3, 1.1 Hz, 1H), 8.47 (dd, J = 7.3, 1.1 Hz, 1H), 8.47 (d, J = 8.2 Hz, 1H), 8.34 (s, 1H), 7.67 (dd, J = 8.4, 7.3 Hz, 1H), 7.13 (d, J = 8.2 Hz, 1H), 3.13 (s, 6H). ¹³C NMR (151 MHz, CDCl₃) δ 164.47, 163.84, 157.73, 132.63, 132.07, 131.82, 131.05, 125.78, 124.96, 123.07, 114.66, 113.36, 44.95. ESIMS [M+H]⁺ calcd. for C₁₄H₁₂N₂O₂ 241.1, found 241.0.

DMNP-NH

DMNP-O (40 mg, 0.18 mmol) and urea (23 mg, 0.37 mmol, 2 eq) were stirred at 170 °C for 2 h under nitrogen. The mixture was cooled to room temperature, triturated in water, and extracted five times with DCM. The organic phase was dried over Na₂SO₄ and the solvent was removed under reduced pressure and purified by silica gel column chromatography (30 to 50% ethyl acetate in hexanes) to give DMNP-NH (35 mg, 78% yield) as a yellow solid. ¹H NMR (600 MHz, DMSO-d₆) δ 11.23 (s, 1H), 8.20 (s, 1H), 8.09 (s, 1H), 8.02 (d, J = 9.2 Hz, 1H), 7.39 (dd, J = 9.2, 2.5 Hz, 1H), 7.24 (d, J = 2.5 Hz, 1H), 3.09 (s, 6H). ¹³C NMR (151 MHz, DMSO) δ 168.63, 168.53, 149.65, 136.65, 130.45, 128.68, 126.26, 123.48, 122.90, 121.14, 117.32, 106.88, 39.26. ESIMS [M+H]⁺ calcd. for C₁₄H₁₂N₂O₂ 241.1, found 241.0.

SDMN-NH

DMN-NH (12 mg, 0.05 mmol), and Lawesson’s reagent (40 mg, 0.1 mmol) were stirred in dry toluene (5 mL) under nitrogen at 110 °C, in a Schlenk tube for 36 h. The reaction was cooled to room temperature, the solvent was removed under reduced pressure and purified by silica gel chromatography (10% ethyl acetate in hexanes) to give SDMP-NH as a dark purple solid (8.6 mg, 63% yield).¹H NMR (600 MHz, DMSO-d₆) δ 13.58 (s, 1H), 8.81 (d, J = 7.6 Hz, 1H), 8.63 (d, J = 8.9 Hz, 1H), 8.57 (d, J = 8.3 Hz, 1H), 7.66 (t, J = 8.0 Hz, 1H), 7.16 (d, J = 9.0 Hz, 1H), 3.27 (s, 6H). ¹³C NMR (151 MHz, DMSO) δ 189.16, 186.10, 158.43, 138.10, 136.17, 134.28, 129.00, 127.72, 125.44, 122.63, 120.03, 113.83, 44.95. ESIMS [M+H]⁺ calcd. for C₁₄H₁₂N₂S₂ 273.1, found 273.0.

SDMNP-NH

DMN-NH (12 mg, 0.05 mmol), and Lawesson’s reagent (40 mg, 0.1 mmol) were stirred in dry toluene under nitrogen at 110 °C, in a Schlenk tube for 12 h. The reaction was cooled to room temperature, the solvent was removed under reduced pressure and purified by silica gel chromatography (15% ethyl acetate in hexanes) to give SDMP-NH as a dark purple solid (9.8 mg, 72% yield). ¹H NMR (600 MHz, DMSO-d₆) δ 13.73 (s, 1H), 8.22 (s, 1H), 8.10 (s, 1H) 8.09 (d, J = 9.2 Hz, 2H), 7.38 (dd, J = 9.2, 2.5 Hz, 1H), 7.35 (d, J = 2.5 Hz, 1H), 3.11 (s, 6H). ¹³C NMR (151 MHz, DMSO) δ 198.09, 197.27, 150.05, 136.63, 133.14, 131.38, 128.09, 125.87, 124.22, 121.46, 117.41, 107.74, 39.24. ESIMS [M+H]⁺ calcd. for C₁₄H₁₂N₂S₂ 273.1, found 273.0.

2.3. Fluorescence assays

Thionated fluorophores (10 μM) in acetonitrile, were mixed with 1mM solutions of each corresponding metal dissolved in 10mM HEPES buffer. For detection limit experiments, 10 μM solutions of thionated fluorophores in acetonitrile were mixed with the corresponding concentration of mercury (2.5, 5, 7.5, and 10 μM). Blank measurements were performed with 5 μM solutions of the corresponding fluorophore in the solution containing 50% acetonitrile and 50% 10 mM HEPES.

2.4. UV-Vis experiments

Blank reference was measured using an acetonitrile/water (1:1) mixture. 10 μM solutions of thionated fluorophores in acetonitrile, were mixed with 1mM solutions of each corresponding metal dissolved in 10mM HEPES buffer.

2.5. NMR experiments

Fluorophore (1 mM) were prepared in 1:1 CD₃CN/D₂O (DMSO-d₆ for SDMNP-NH), followed by the addition of HgCl₂ (10 eq). ¹H NMR spectra were monitored at different time points until no change was observed.

2.6. Quantum yield determination

Quantum yields measurements were determined using fluorescein (quantum yield of 0.92 in 0.1 M NaOH), Quinine sulfate (quantum yield of 0.546 in H₂SO₄ 0.5 M), and Rhodamine B (quantum yield of 0.31 in water) as standards.

The fluorescence quantum yield, Φ_f (sample), was calculated according to the equation below:

$$\frac{{\text{Φ}}_{f,\mathrm{ s}ample}}{{\text{Φ}}_{f,ref}}=\frac{O{D}_{ref}\cdot I_{sample}\cdot d_{sample}^2}{O{D}_{sample}\cdot I_{ref}\cdot d_{\text{r}ef}^2}$$

Φ_f: quantum yield of fluorescence;

I: integrated emission intensity;

OD: optical density at the excitation wavelength;

d: refractive index of solvents, d_DMSO=1.48; d_water=1.33.

2.7. Cell culture

HeLa cells were cultured in DMEM (Invitrogen, Carlsbad, CA), supplemented with 10% fetal bovine serum in a humidified atmosphere of 5% CO₂ at 37 °C. For imaging experiments, exponentially growing cells were seeded in 8-chamber (Lab-Tek 155411). The cells were cultured at 37 °C in a 5% CO₂ atmosphere for 24 h before adding reagents.

2.8. Cell imaging

For labeling, the medium was removed and the cells were rinsed three times with PBS (pH 7.4). Then, the HeLa cells were incubated with probes SCou, SDMAP, and SDMNP-NH with a final concentration of 4 μM in PBS at 37 °C for 30 min as the control. For Hg²⁺ imaging, another set of HeLa cells was preloaded with the corresponding probes with a final concentration of 4 μM in PBS at 37 °C for 30 min, rinsed three times with PBS, and further treated with increasing amounts of Hg²⁺ (2, 5, and 10 μM) in PBS at 37 °C for additional 30 minutes. The cells were rinsed three times with PBS, then the imaging was carried out on a Carl Zeiss Confocal LSM800 microscope with 40× lenses. All probes were scanned by 405 laser light at 70% laser intensity.

3. RESULTS AND DISCUSSION

To determine if a single sulfur-for-oxygen atom replacement within common fluorophore scaffolds represents a general approach to preparing Hg²⁺ sensors, we prepared thio-acridone (SACD), thio-coumarin (SCou), thio-4-dimethylaminophthalimide (SDMAP-NH), thio-4-dimethylamino-1,8-naphthalimide (SDMN-NH), thio-4-dimethylaminonaphthalimide (SDMNP-NH), and thio-based Nile Red (SNile Red) from their oxo forms using Lawesson’s reagent in toluene according to our previously described methodology.[55,56] All compounds were fully characterized by ¹H NMR, ¹³C NMR, and mass spectrometry as described in the supporting information. Compared to carbonyl analogs, thiocarbonyl fluorophores exhibit a red-shifted absorption spectrum and larger extinction coefficients (Table S1).

Importantly, the fluorescence quantum yield of thio-caged fluorophores in DMSO is significantly lower than that of the corresponding oxo forms, indicating the likelihood that the Hg²⁺-induced hydrolysis reaction might be used to re-activate the fluorescence of thio-caged probes. To test the selectivity of thio-caged fluorophores as Hg²⁺ sensors, the fluorescence intensity changes in different thio-caged probes were evaluated 30 minutes after the addition of different metal ions (Figure 1). In the presence of Hg²⁺ ions, SCou, SDMAP-NH, SDMN-NH, SDMNP-NH, but not SACD and SNile Red, exhibited remarkable increases in fluorescence intensity, probably due to differences in the electron-donating ability of the atom adjacent to the carbonyl group. In contrast, the addition of other metal ions, including Pt²⁺, Pb²⁺, Na⁺, Mg²⁺, K⁺, Ca²⁺, Mn²⁺, Fe²⁺, and others, did not result in a significant change in fluorescence intensity (Figure 1). Additionally, competing experiments showed the probe selectivity reacts mercury. Using the mixture of Hg²⁺ and other ions listed in Figure 1, the fluorescent response was similar to the single addition of HgCl₂ (Figure S1). The only exceptions were the addition of Ag⁺ to SCou and SDMN-NH. Silver ions are known to form complexes with thionated dyes. Therefore, with the addition of Ag⁺, the fluorescence intensity was relatively lower. SDMAP and SDMNP-NH react quickly with Hg²⁺, and no significant difference was observed when adding silver ions (Figure S1).

To further characterize the Hg²⁺-triggered fluorescent activation process in SCou, SDMAP, and SDMNP, we recorded ¹H NMR, absorption, and emission spectra. Upon addition of Hg²⁺ ions, aromatic protons of SCou shift from 8.02, 7.92, 7.37, 7.27, and 7.12 ppm to 8.72, 8.29, 7.80, 7.73, 7.41 ppm, corresponding to the formation of the SCou-Hg complex. Within 72 h, signals corresponding to the SCou-Hg complex gradually disappeared and were replaced by new peaks appearing at 8.32, 7.97, 7.33, 7.20, and 6.61 ppm, in good agreement with the spectrum of commercially available Cou (Figure 2A). The absorption peak of SCou at 472 nm was initially red-shifted to 511 nm, followed by a blue-shift to 381 nm with the addition of Hg²⁺ ions (Figure 2B). After 12h of exposure to Hg²⁺, a 144-fold increase in fluorescence intensity was observed at 467 nm (Figure 2C and 2D).

Fig. 2.

Spectral changes of SCou and SDMNP-NH after Hg²⁺ addition. A) ¹H NMR monitoring of SCou 1 mM with 10 eq. of HgCl₂ in CD₃CN/D₂0 1:1. Image of samples before (left) and after (right) addition of HgCl₂ under UV light (365 nm) irradiation. B) UV-vis monitoring of 5 μM SCou after the addition of HgCl₂ (500 μM) in CH₃CN/HEPES buffer 1:1. C) Fluorescence spectra monitoring of 5 μM SCou after the addition of HgCl₂ (500 μM) in CH₃CN/HEPES buffer 1:1. D) Turn-on fold increase of 5 μM SCou after the addition of Hg ²⁺ (500 μM) at 467 nm in CH₃CN/HEPES buffer 1:1. E) ¹H NMR monitoring of SDMNP-NH 1 mM with 10 eq. of HgCl₂ in DMSO-d₆. Images of samples before (left) and after (right) addition of HgCl₂ under UV light (365 nm) irradiation. F) UV-vis monitoring of 5 pM SDMNP-NH after the addition of HgCl₂ (500 μM) in CH₃CN/HEPES buffer 1:1. G) Fluorescence spectra monitoring of 5 pM SDMNP-NH after the addition of HgCl₂ (500 μM) in CH₃CN/HEPES buffer 1:1. H) Turn-on fold increase of 5 μM SDMNP-NH after the addition of HgCl₂ (500 μM) at 513 nm in CH₃CN/HEPES buffer 1:1.

SDMAP-NH and SDMNP-NH exhibited more rapid responses to Hg²⁺ ions. Upon treatment of SDMAP-NH and SDMNP-NH with 10 equiv. of HgCl₂ in aqueous solution, changes in ¹H NMR, absorption, and emission spectra were recorded. As shown in Figure 2E, signals corresponding to SDMNP-NH phenyl moiety protons shifted from 8.22, 8.10, 8.09, 7.38, 7.35 ppm to 8.20, 8.09, 8.02, 7.39, 7.24 ppm, respectively. After the addition of Hg²⁺, SDMAP-NH and SDMNP-NH exhibit a purple color, with absorption bands at 557 nm (Figure S2) and 558 nm (Figure 2F), that rapidly disappears within 10 min, yielding colorless solutions. Furthermore, SDMAP-NH and SDMNP-NH exhibited fluorescence turn-on ratios of 38 at 504 nm (Figure S2) and 52 at 513 nm (Figure 2G and H), respectively. Therefore, our data suggest that SDMAP-NH and SDMNP-NH can be fully desulfurized within 10 mins, compared to 12h for SCou. SDMN-NH exhibits a similar 34-fold fluorescence turn-on ratio at 540 nm after 10 h (Figure S3). The solution colors change immediately from yellow to red for SCou, purple to light-yellow for SDMAP and SDMNP-NH, and purple to pink for SDMN-NH. These changes allow for the naked-eye detection of mercury ions at varying concentrations (Figure S4).

Mass spectrometry analysis of samples incubated with Hg²⁺ indicates losses of 16 amu for SCou and 32 amu for SDMAP-NH, SDMN-NH, and SDMNP-NH, corresponding to the regeneration of fluorophores in the oxo form (Figures S5-S8). The results obtained with UV-vis absorbance and 1H NMR agree with the two-step desulfurization mechanism described previously.[53] After the addition of mercury, a complex is formed immediately, assisted by a neighboring electron-donating atom (oxygen or nitrogen). The second step involves the hydrolysis of mercury sulfide, restoring the corresponding fluorescent dye. (Figure 1).

Detection limits for SCou, SDMAP, SDMN, and SDNMP were measured according to the methodology used by Talukdar.[59] Low mercury concentrations were used to obtain linear relationships between fluorescence intensity and Hg²⁺ concentration (Figure S9). SCou proved to be the most sensitive of the probes tested, with a detection limit of 2.70 ± 0.06 nM after an incubation time of 8 h. With a similar incubation time, SDMN has an Hg²⁺ detection limit of 160 nm. By comparison, SDMAP and SDMNP exhibit detection limits of 288 ± 7 nM and 770 ± 8 nM, respectively, measured 15 minutes after Hg²⁺ ion addition. A comparison with other sensing platforms is also summarized in table S2.

Next, we studied the utility of thio-caged probes for the detection of Hg²⁺ ions in living cells. Hela cells were incubated for 30 min with 4 μM SCou, SDMAP-NH, and SDMNP-NH, followed by the addition of different concentrations of HgCl₂ for an additional 20 min. As shown in Figure 3, cells without HgCl₂ treatment exhibited no fluorescence signal. By comparison, SCou-, SDMAP-NH-, and SDMNP-NH-treated cells exhibited distinct intracellular fluorescence in the prescence of more than 5 μM Hg²⁺ ions. This illustrates both the membrane permeability of thio-caged probes and their ability to detect intracellular Hg²⁺. Importantly, no obvious cytotoxicity was observed in cells treated with these probes.

Fig. 3.

Confocal microscope images of HeLa cells incubated with SCou, SDMAP-NH, and SDMNP-NH (4 μM) with varying concentrations of HgCl₂ (0, 2, 5, and 10 μM, respectively). Images were obtained 30 min after mercury addition. Scalebar = 20 μm.

4. CONCLUSION

In conclusion, we have used a single sulfur-for-oxygen atom substitution within available fluorophores to develop a general method for the preparation of colorimetric and fluorogenic Hg²⁺ chemosensors. Using this approach, we developed a series of thio-caged probes across a broad spectral range. Fluorescence spectroscopy can be used alongside these probes to provide rapid recognition of Hg²⁺ ions, with high selectivity and sensitivity in aqueous samples. The colorimetric assay developed in this work also allows for the immediate, qualitative detection of Hg²⁺. SDMAP and SDMNP-NH offer rapid detection, while SCou requires a longer incubation time but affords higher sensitivity. Furthermore, the detection of intracellular Hg²⁺ using these thio-caged chemosensors suggests their potential applications in biological and environmental research.

HIGHLIGHTS.

Single-atom replacement can quench fluorescent dyes via one-step modification.

Mercury ions can activate thiocaged dyes in aqueous and cellular environments.

Thiocaged dyes are applied to detect mercury based on colorimetric and fluorescent signals.