Development of a new fluorescent Pb²⁺ sensor

Lead is a persistent environmental contaminant,^{[1, 2, 3,Domaille, 2008 #17, 4]} as exposure to very low levels of lead can cause neurological, reproductive, cardiovascular, and developmental disorders.^{[3, 5, 6]} Children with variants in iron metabolism genes may be more susceptible to lead absorption and accumulation. ^[7],[8] The US Center for Disease Control (CDC) set standards stating that a 10–19 μg/dL blood lead level poses a potential threat and diagnostic testing is recommended.^[7] Of particular interest is Pb²⁺ as it interferes with enzymatic heme production.^[9]

Heavy metal, such as lead, poisoning has prompted demand for new techniques to selectively identify and study the actions of these metal ions.^{[7, 10],[4]} Currently, the most common detection of lead includes atomic absorption spectrometry,^[8] inductively coupled plasma mass spectrometry,^[11] and anodic stripping voltammetry,^[12] and these instrumentally intensive methods^{[6, 13]} measure only total lead content,^[1] and often times require extensive sample preparation. Thus, a simple and inexpensive method for not only detecting, but also quantitating Pb²⁺ is desirable in real time monitoring of environmental, biological, and industrial samples.

Fluorescence based sensors offer unparalleled sensitivity and thus, have garnered significant interest.^[4] Most fluorescent probes for detecting Pb²⁺ use peptides,^[14] proteins,^[15] or DNAzymes.^{[3, 6, 16–18]} These probes lack the simplicity that a small molecular probe can offer. In addition, non specific interaction and background fluorescence often act as a deterring factor, which underscores the necessity of a selective lead sensor that can function in aqueous environments.^{[1–3, 6]} To this end, a water soluble fluorescence based small molecule Pb²⁺ sensor (Leadfluor-1) has showed promise in understanding cellular Pb²⁺ trafficking.^[2] In addition to solubility and sensitivity, selectivity is an important criterion for the success of a sensor. Ideally, the sensor should have high selectivity with a high dynamic range. Herein we present the design, synthesis, and characterization of a new turn-on ratiometric fluorescent lead sensor, 4,4-dimethyl-4H-5-oxa-1,3-dithia-6,11-diaza cyclopenta[a] anthracen-2-one, Leadglow (LG, 7).

LG has a thiol-based binding site, which differs from other fluorophores with more hard donors such as oxygen or nitrogen. Lead is a soft metal and therefore favors sulfur-rich binding sites.^[19] The proposed molecule can serve as a highly sensitive and selective fluorescent lead sensor in aqueous samples. LG fluoresces at 465 nm. In the presence of Pb²⁺, LG fluoresces at 423 nm, with a 5-fold increase in emission intensity, indicating a turn-on response to lead in aqueous solution.

The synthetic procedure to LG is shown in Scheme 1. The reaction of 2-methyl-3-butyn-2-ol and 3,4-dihydro-2H-pyran in the presence of a catalytic amount of p-toluenesulfonic acid results in the protected alcohol 1 in excellent yield. Deprotonation of 1 followed by the addition of diethyl oxalate at low temperature affords 2 in moderate yield. Reaction of the α-keto ester 2 with 4-phenyl 1,3-dithiolane-2-thione allowed us to introduce the protected dithiolene moiety. Direct reaction of 2 with the 4-phenyl 1,3-dithiolane-2-thione affords the intermediate molecule 3 which was transformed to the pyrandione 4 upon addition of trifluoroacetic acid (TFA). Conversely when the same reaction was performed in xylene, the pyrandione 4 was isolated directly in moderate yields. The thione sulfur in 4 was replaced with oxygen using mercury(II) acetate giving the pyrandione, 5, in good yield. The reaction of 5 with o-phenylenediamine in methanol afforded almost quantitatively the quinoxaline compound, 6. Addition of benzylchloroformate and triethylamine to 6 leads to the formation of compound 7 (Leadglow, LG) in good yield. LG was characterized by infrared, NMR (¹H and ¹³C), and UV-visible spectroscopies, and mass spectrometry.

Scheme 1.

Synthetic scheme of Leadglow

All spectroscopic measurements were performed in 2.5% MeOH and water. NEt₄OH was added to the solution (2:1 NEt₄OH:LG) to hydrolyze the carbonyl group and expose the thiolato binding site. LG exhibits an absorption band at 415 nm (ε = 1.3 × 10⁵ M⁻¹ cm⁻¹) and an emission band of low intensity (φ = 0.12) at 465 nm. Upon incubation of a solution of LG with lead acetate solution, the absorption band shifts to 389 nm (ε = 1.1 × 10⁵ M⁻¹ cm⁻¹). The emission band also shifts to 423 nm, with a 5-fold increase in the fluorescence intensity (φ = 0.63), thus acting as a ‘turn-on’ sensor (Figure 1). In addition, LG exhibits a shift in the emission energy characteristic feature of a wavelength-ratiometric probe (blue shifted by 42 nm). Thus, like Leadfluor-1, LG acts not only as a turn on sensor, but also as a ratiometric one^[2]. Upon binding to lead, Leadflour-1 exerts a larger increase in the emission intensity (18 fold) with a quantum yield of 0.013, LG offers a higher quantum yield (0.63) for the Pb-bound species. LG is versatile and functions well at a wide pH range (Figure 2). The emission intensity of Pb²⁺ bound LG remains near constant in the pH range from 4 to 10.

Figure 1.

Spectra acquired in 2.5% MeOH and water and NEt₄OH (2:1 NEt₄OH:LG). Emission spectra of 5 μM free LG (red) and 5 μM Pb²⁺ bound LG (black). Excitation for free LG provided at 415 nm. Excitation for Pb²⁺ bound LG (1:2 Pb²⁺:LG) provided at 389 nm. Emission maximum observed at 423 nm with a 5-fold increase in emission intensity. Inset is a magnification of free LG (5 μM) emission spectra.

Figure 2.

pH profile of Pb²⁺ bound Leadglow. The complex exhibits a high, constant emission intensity from pH 4 to 10. indicating a wide functional pH range. The asterisk indicates where all selectivity studies were performed.

Binding assays were performed using Job’s method of continuous variation,^[20] which indicates a 1:2 Pb²⁺:LG complex. The apparent dissociation constant for a complex, K_d, for Pb2+ coordination to LG was found to be 217 nM (at pH 10), using the Hill-1 function. LG is very sensitive to Pb²⁺ in aqueous solution. LG binds to Pb²⁺ much stronger than Leadfluor-1 (K_d 23 μM).^[2] The ‘turn–on’ feature of the sensor allowed detection of a low level (10 ppb) of Pb²⁺, even in the presence of other metals, using a 1 μM solution of LG. The LG can be used in detecting and determining Pb²⁺ in the tested range from 1 ppb to 50 ppb. Thus LG offers a high dynamic range for Pb²⁺ detection. To further examine the sensitivity and accuracy of the sensor, we used a NIST standard of trace elements in water (SRM^® 1643e) in the concentration range of 1–50 ppb Pb²⁺ and probed with 1 μM LG. In this case, accurate fluorescence responses were observed from 50 ppb to as low as 10 ppb. LG was also used in quantitating the concentration of Pb²⁺ in solutions prepared from a lead standard (NIST SRM^® 3128). These results were compared with those obtained from ICPMS measurement. The precisions of the two methods were found to be comparable by F-test and t-test analyses.

LG is extremely selective for Pb²⁺ against other common metal ions tested. The fluorescence response of 5 μM LG in the presence of Pb²⁺ and other ions in aqueous solution are shown in Figure 3. No change in the fluorescence was observed when a solution of LG containing Pb²⁺ was incubated with 2 mM Li⁺, Na⁺, K⁺, Ca²⁺, or Mg²⁺, thus exhibiting a similar properties to Leadfluor-1.^[2] These metal ions were tested with higher concentration as they are highly abundant in mammalian cells. Similarly, the fluorescence intensity of LG containing Pb²⁺ remains unchanged in the presence of 3d transition metals. Thus, 75 μM Fe²⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, As³⁺, Sn²⁺ or Mn²⁺ ions cause no difference in fluorescence intensity. This clearly demonstrates high selectivity of LG towards Pb²⁺ which is important for a viable sensor whether investigating an environmental sample (where common heavy metal contamination includes Cd²⁺, As³⁺ and Hg²⁺) or biological sample as plausible cellular targets of toxic lead accumulation include calcium- and zinc-dependent proteins.^7,17

Figure 3.

Spectra acquired in 2.5% MeOH and water and NEt₄OH (2:1 NEt₄OH:LG). Fluorescence response of 5 μM LG to common biologically available cations. The bars represent the final fluorescence response (F_f) over the initial fluorescence response (F_i). White bars represent the addition of each ion (2 mM for Li⁺, Na⁺, K⁺, Ca²⁺, and Mg²⁺ and 75 μM for Fe²⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, Cd²⁺, Mn²⁺, Hg²⁺, As³⁺, Sn²⁺, and Pb²⁺. Black bars represent the addition of 75 μM Pb²⁺ to the solution. Excitation was provided at 389 nm. a. Li⁺, b. Na⁺, c. K⁺, d. Ca²⁺, e. Mg²⁺, f. Fe²⁺, g. Co²⁺, h. Ni²⁺, i. Cu²⁺, j. Zn²⁺, k. Cd²⁺, l. Mn²⁺, m. Hg²⁺, n. As³⁺, o. Sn²⁺, p. Pb²⁺.

In conclusion, LG is a new fluorescent sensor that can detect Pb²⁺ in aqueous solution at a wide pH range (4–10). LG is advantageous because of its sensitivity for Pb²⁺ at concentrations below the EPA limit, turn-on and ratiometric detection of Pb²⁺ over other biologically as well environmentally abundant cations, visible excitation and emission profiles, and high optical brightness. LG is useful in determining the amount of Pb²⁺ in various aqueous samples, as it can detect Pb²⁺ in a mixture of several other metals at a concentration as low as 10 ppb. This molecular system offers a wide variety of choices from tuning the excitation to specific tagging through selective substitution.

Experimental Section

Synthetic Materials and Methods

Potassium hydroxide, carbon disulfide, and mercuric acetate were purchased from Fisher Scientific and were used as they were received without further purification. The rest of the chemicals were purchased from Acros and were used as they were received without further purification. Standard Reference Material^®1643e Trace Elements in Water and 3128 Lead Standard Solution were purchased from National Institute of Standards and Technology and used as received. Column chromatography was performed on Silica gel P60 (Sorbent Technologies). CH₂Cl₂ was distilled over CaH₂, diethyl ether over Na wire/benzophenone and MeOH over Mg. All the other solvents were used without further purification. ¹H NMR and ¹³C NMR spectra were obtained in CDCl₃ or CD₃OD at 25°C on a Bruker 400 MHz spectrometer. Infrared spectra were obtained on a Nicolet 380 FT-IR (Thermo) spectrometer. APCI-MS were recorded in methanol on a Waters ZMD mass spectrometer set in positive mode (solvent: methanol; cone voltage: 20 V; corona 2.7 kV; source temperature: 130°C; flow rate: 100 μL/min) or negative mode (solvent: methanol; cone voltage: −20 V; corona 2.5 kV; source temperature: 130°C; flow rate: 100 μL/min). ESI-MS were recorded on a Waters ZMD mass spectrometer set in the negative ionization mode (solvent: methanol; cone voltage: 20 V; capillary voltage: 2.9 kV; source temperature: 130°C; flow rate: 150 μL/min). Inductively coupled plasma mass spectrometry was conducted in an Agilent 7500ce equipped with a Shield Torch SYSTEM.

3-Methyl-3-tetrahydropyranyloxy-butyne (1)

3-methyl-3tetrahydropyranyloxy-butyne was prepared according to literature procedure.^[21] Dihydropyran (14.7 mL, 160.71 mmol) was added to a cooled (−20°C) solution of 2-methyl-3-butyn-2-ol (9.2 g, 112.53 mmol) in 80 mL of dry CH₂Cl₂. A few crystals of p-toluensulphonic acid were added to the solution. The solution was stirred for 3 h and then washed with a saturated solution of NaHCO₃ (4×30 mL) and the organics were dried over MgSO₄. The solvent was removed at reduced pressure and the resulting oil was purified by vacuum distillation. Yield: 15 g (90%). ¹H NMR spectrum in CDCl₃ (ppm): δ5.06 (m, 1H), 3.95 (m, 2H), 3.50 (m, 2H), 2.43 (s, 1H), 1.85 (m, 2H), 1.70 (m, 2H), 1.51 (s, 6H). ¹³C NMR spectrum in CDCl₃ (ppm): δ 95.97, 86.25, 86.24, 71.78, 70.72, 63.12, 31.80, 30.47, 29.68, 25.29, 20.29. IR spectrum (neat, cm⁻¹): 3295, 2941, 2108, 1466, 1380.

Ethyl-5-methyl-2-oxo-5-(tetrahydro-2H-pyran-2-yloxy)hex-3-ynoate (2)

A solution of 3-methyl-3-tetrahydropyranyloxy-butyne (3.3 g, 19 mmol) in 30 mL Et₂O was cooled to 0°C. nButyllithium (11.5 mL, of a 2.5 M solution in hexane, 28 mmol) was added to the previous solution. The resulting solution was stirred for 30 min at 0°C, then cooled to −78°C and diethyl oxalate (4.3 mL, 29.4 mmol) was added. The reaction was followed by TLC (hexane/EtOAc 90:10). After ca. 2 h the reaction mixture was poured into a cold aqueous solution of NH₄Cl. The aqueous layer was extracted with Et₂O (3×25 mL). The organics were dried over MgSO₄. The solvent was removed under reduced pressure and the resulting yellow oil was purified via chromatography (silica gel, hexane/EtOAc 90:10) to afford 2 as a pale yellow liquid. Yield: 2.3 g (45%). ¹H NMR spectrum in CDCl₃ (ppm): δ 5.10 (m, 1H), 4.38 (q, 2H), 4.36 (m, 2H), 3.96 (m, 2H), 3.54 (m, 2H), 1.85 (m, 2H), 1.75 (m, 2H), 1.64 (s, 3H), 1.59 (s, 3H), 1.40 (t, 3H). ¹³C NMR spectrum in CDCl₃ (ppm): δ 169.3, 158.7, 101.4, 96.2, 81.6, 70.5, 63.1, 31.5, 29.2, 28.9, 25.2, 19.9, 13.8, 13.8. IR spectrum (neat, cm⁻¹): 2209, 1741, 1689. MS-APCI calculated for C₁₄H₂₀O₅ [M]⁻: 268.12, found 267.96.

4,4-Dimethyl-2-thioxo-4H-[1,3]dithiolo[4,5-c]pyran-6,7-dione (4)

4-phenyl-1,3-dithiolane-thione was prepared according to literature procedure.^[22] Ethyl-5-methyl-2-oxo-5-(tetrahydro-2H-pyran-2-yloxy)hex-3-ynoate (1.0 g, 3.7 mmol) and 4-phenyl-1,3-dithiolane-thione (2.0 g, 9.3 mmol) were dissolved in xylene (15 mL). The deep yellow solution was refluxed at 140°C for 8 h under Ar. The open intermediate, 3, was treated with a few catalytic drops of trifluoroacetic acid (TFA). The reaction was followed by TLC, eluent: CH₂Cl₂. The solvent was removed under reduced pressure and purified via chromatography (silica gel, CH₂Cl₂) to afford 4 as a yellow solid. Yield: 0.31 g (34%). ¹H NMR spectrum in CDCl₃ (ppm): δ 1.88 (s, 6H). ¹³C NMR spectrum in CDCl₃ (ppm): δ 205.3, 167.7, 162.7, 157.8, 153.6, 83.2, 31.9. IR spectrum (neat, cm⁻¹) 1747, 1681, 1557, 1460. MS-ESI calculated for C₇H₈O₄S₂Na [M+Na]⁺: 268.96, found 268.79.

4,4-Dimethyl-4H-[1,3]dithiolo[4,5-c]pyran-2,6,7-trione (5)

Mercuric acetate (181 mg, 0.568 mmol) was added to 4,4-dimethyl-2-thioxo-4H-[1,3]dithiolo[4,5-c]pyran-6,7-dione (100 mg, 0.406 mmol) in CH₂Cl₂/AcOH (3:1) and stirred for 30 min. The reaction mixture was filtered through a celite pad to remove the mercury salts. The resulting solution was washed with water (3×15 mL) and saturated NaHCO₃ (5×10 mL) and dried over MgSO₄. The solvent was removed under reduced pressure to afford 5 as a beige solid. Yield: 40 mg (43%). ¹H NMR spectrum in CDCl₃ (ppm): δ 1.88 (s, 6H). ¹³C NMR spectrum in CDCl₃ (ppm): δ 184.5, 163.5, 160.4, 153.5, 128.3, 84.2, 31.8. IR spectrum (neat, cm⁻¹) 1746, 1693, 1641, 1552, 1453. MS-ESI calculated for C₇H₈O₄S₂Na [M+Na]⁺: 252.96, found 252.79.

3-(5-(2-Hydroxypropan-2-yl)-2-oxo-1,3-dithiol-4-yl)quinoxalin-2(1H)-one (6)

o-phenylenediamine (50 mg, 0.46 mmol) was added to a stirred solution of 4,4-dimethyl-4H-[1,3]dithiolo[4,5-c]pyran-2,6,7-trione (100 mg, 0.43 mmol) in methanol (15 mL). The solution was stirred overnight. The solvent was removed under reduced pressure and the residue was purified via crystallization from CH₂Cl₂/hexane affording pure 6 as a light orange solid. Yield: 143 mg (98%). ¹H NMR in CDCl₃ (ppm): δ 10.40 (bs, 1H), 7.88 (d, 1H), 7.62 (t, 1H), 7.43 (d, 1H), 7.28 (t, 1H), 4.28 (s, 1H), 1.61 (s, 6H). ¹³C NMR spectrum in CD₃OD (ppm): δ 194.6, 158.7, 156.0, 148.2, 136.0, 132.6, 127.9, 125.5, 123.6, 122.3, 119.2, 77.3, 33.8. IR spectrum (neat, cm⁻¹) 1667, 1598. MS-ESI calculated for C₁₄H₁₁N₂O₃S₂ [M-H]⁻: 319.03, found 318.87.

4,4-Dimethyl-4H-5-oxa-1,3-dithia-6,11-diaza-cyclopenta[a]anthracen-2-one (Leadglow, LG) (7)

3-(5-(2-hydroxypropan-2-yl)-2-oxo-1,3-dithiol-4-yl)quinoxalin-2(1H)-one (140 mg, 0.439 mmol) was partially dissolved in CH₂Cl₂ (10 mL). Benzylchloroformate (125 μL, 0.81 mmol) and triethylamine (120 μL) were added and the resulting solution was stirred overnight. The volume of the solution was reduced to ca. 3 mL and it was purified via chromatography (silica gel, CH₂Cl₂) to give pure Leadglow, 7. Yield: 92 mg (70%). ¹H NMR spectrum in CDCl₃ (ppm): δ 7.96 (d, 1H), 7.83 (d, 1H), 7.66 (t, 1H), 7.60 (t, 1H), 1.84 (s, 6H). ¹³C NMR spectrum in CDCl₃ (ppm): δ 189.0, 152.5, 141.0, 140.7, 139.6, 133.7, 130.5, 128.6, 128.1, 127.5, 124.2, 81.2, 30.0. IR spectrum (neat, cm⁻¹): 1705, 1664, 1624, 1461, 1409. MS-APCI calculated for C₁₄H₁₁N₂O₂S₂ [M+H]⁺: 303.02, found 302.93. UV-vis in MeOH (λ_max, nm (ε, M⁻¹ cm⁻¹): 256 (10988), 367 (11194), 385 (9568). Fluorescence in MeOH: Excitation = 367, 385 nm. Emission = 415 nm. Fluorescence in 2.5% MeOH and water and NEt₄OH (1:2, LG: NEt₄OH): Excitation = 415 nm. Emission = 465 nm.

Spectroscopic Materials and Methods

All spectroscopic measurements were recorded in a solution of 2.5% methanol (for solubility reason) and water. HPLC grade MeOH and distilled water were used for each measurement. Absorption spectra were measured on a Varian Cary 300 spectrometer. Samples for absorption measurement were performed in 1-cm × 1-cm quartz cuvettes (3.5 mL volume, Starna). Fluorescence spectra were measured on a Photon Technology International Quanta Master ^™ 4 spectrofluorometer. Fluorescence measurements were performed in 1-cm × 1-cm quartz cuvettes (3.5 mL volume, NSG Precision Cells). Fluorescence quantum yields were determined in reference to fluorescein in 0.1 NaOH (φ = 0.95).^[23] The binding ratio was determined using Job’s method of continuous variation.^[20] The binding affinity of Pb²⁺ to LG was also found. Excitation was provided at 389 nm. The dissociation constant, K_d, was determined by using the Hill1 function in OriginPro 8, which is as follows: y = START + (END – START) × xⁿ / (kⁿ + xⁿ). START is the first data point where the curve begins, END is the last data point where the curve ends, x is the concentration of Pb²⁺, k is the K_d for the binding reaction, and n is the Hill coefficient or the cooperativity of the dependence on x. A 10 μM solution (1:2:4 Pb²⁺:LG:NEt₄OH) was pH adjusted down by adding AcOH and adjusted up by adding NEt₄OH. SRM^®1643e Trace Elements in Water (1–50 ppb Pb²⁺) was probed with 1 μM LG in NEt₄OH (2:1 NEt₄OH:LG) and 2.5% MeOH and water. LG was able to detect Pb²⁺ quantitatively by an increase in fluorescence response from 10–50 ppb Pb²⁺. For ICPMS measurements solutions were scanned for ²⁰⁴Pb, ²⁰⁶Pb, ²⁰⁷Pb, ²⁰⁸Pb isotopes, the sums of the counts for all isotopes were used for analyses. For quantitative analyses, (ICPMS and fluorescence) calibration curves at pH ~6.6, were created using a multi-element (31 elements) calibration standard (ICP-MSCS-M) obtained from High Purity Standards, Charleston, SC. In both cases, linearity with >99% correlation was maintained with Pb²⁺ concentration in the range 1–50 ppm. Using the linear equations, concentrations of Pb²⁺ present in samples prepared from lead standard solution (SRM^® 3128) via serial dilution were determined. The results from two methods (ICPMS and fluorescence) passed the F-test with 95% confidence limit. The t-test with 95% confidence interval showed the results of the methods are statistically equivalent.