Chemical speciation of MeHg+ and Hg2+ in aqueous solution and HEK cells nuclei by means of DNA interacting fluorogenic probes† †Electronic supplementary information (ESI) available: Experimental details, characterization data, and additional experiments. See DOI: 10.1039/c5sc00718f

Speciation of Hg2+ and MeHg+ has been achieved by in vitro approaches with fluorogenic probes supported in cultured cells.


Introduction
Environmental contamination by mercury is a serious issue because of the large amounts of mercury released into the environment by human activities such as coal-red power stations, primary metals production and the chlor-alkali industry. Another major concern is the persistence of mercury in the environment, not only as the volatile mercury metal but also as the water-soluble mercury(II) (Hg 2+ ) and methylmercury(II) (MeHg + ) cations. 1 Both species of Hg(II) are strongly interconnected in the environment because of the natural cycle of mercury that maintains Hg 2+ and MeHg + concentrations in natural water resources, that are subsequently accumulated in food products produced in these regions. 2 Undoubtedly, one of the best known neurotoxins is MeHg + , a ubiquitous environmental toxicant that leads to long-lasting neurological and developmental decits in animals and humans, 3 especially during prenatal exposure. 4 In the aquatic environment MeHg + is accumulated in sh, which represent a major source of human exposure. 5 MeHg + acts by blocking neurotransmitter release, interfering with transport of amino acids and ions, binding to sulydryl groups and inhibiting protein synthesis. MeHg + has a high mobility in the body due to formation of a complex with the amino acid cysteine. Since this complex of MeHg + and cysteine resembles the structure of the large and neutral amino acid methionine, it can enter the cell and exit as a complex with reduced glutathione, thus forming water-soluble complexes in tissues. 6 Due to their inherent toxicity, Hg(II) species have to be continuously monitored. There are many chemical probes that are useful for the fast detection of Hg 2+ contamination. Usually, they work by complexation of Hg 2+ by colorimetric or uorogenic reagents. 7 Complementing these methodologies, chemical dosimeters act by specic reactions with Hg 2+ , which subsequently undergo a colour or uorescence change. 8 Very few colorimetric or uorogenic probes are able to detect MeHg + , 9 and neither chemical probes nor dosimeters can discriminate speciation of Hg 2+ and MeHg + , 10 despite the enormous interest that MeHg + -induced neurotoxicity promotes 11 and its imaging in living systems. 12 Fluorogenic probes having sulfur atoms are able to interact with Hg(II) species, 13 therefore they could be employed to detect MeHg + by mimicking its behaviour in cells, opening the way to biomimetic selective molecular probes for Hg(II) species. Following this idea we decided to design and test some new sulfur-containing uorogenic probes for their ability to selectively interact with Hg(II) species. We have previously prepared Hg 2+ /MeHg + colorimetric dosimeters 14 and uorescent reporters for signicant analytes. 15 In the latter case, the best examples included some recently reported 14a charge-transfer uorogenic probes bearing conjugated donor and acceptor groups in their structure, as well as a secondary amine group that was not involved in the charge transfer process. We thought that a sulfur-containing functional group in such position should exert a quenching effect on the initial uorescence of the core structure by a photoinduced electron transfer (PET) from the sulfur atom. Subsequent interaction of the sulfur-containing group with thiophilic cations should therefore increase uorescence of the probe, thus making this kind of compound suitable for the detection of Hg(II) cations. Following this idea we now report the use of new uorogenic probes for the selective detection and speciation of Hg 2+ and MeHg + in aqueous-organic mixtures and in HEK293 cells.

Results and discussion
For our current purpose we have designed new uorogenic probes for Hg 2+ /MeHg + . The probes were prepared in fair yields by reaction of previously synthesized compounds 1a-d and dimethylphosphinothioic chloride 2a or O,O-diethyl phosphorochloridothioate 2b in the presence of Hünig's base (Scheme 1). In this way we obtained derivatives BD116, BD119, BD118, BD117, JG7 and JG30, in which the uorescence of the 5-(2-aminopyrimidin)-5-ylindane core is quenched by PET effect in some polar solvents (ESI †). Therefore interaction of thiophilic cations with the sulfur atom in those solvents should increase uorescence of these compounds by an OFF-ON process (Scheme 1).
The ketone derivatives BD116, BD118 and JG7 are creamy solids giving colourless solutions whereas the dicyanomethylene derivatives BD119, BD117 and JG30 are yellow solids giving pale yellow solvatochromic solutions in the most common organic solvents. The solvent effects were experimentally studied in terms of the Lippert-Mataga equations in selected examples, showing a strong inuence of general solvent effects and the absence of hydrogen-bonding interactions contributing to the Stokes shis (ESI, pp. S34-S36 †). We next tested 10 À4 M solutions of the six uorogenic probes in mixtures of different solvents (DMSO, DMF, acetone, MeCN, MeOH) and water (100%, 80 : 20, 60 : 40 and 20 : 80) by adding one to four equivalents of 5 Â 10 À3 M solutions of selected cations, as perchlorate or triate salts, or anions, as tetrabutylammonium salts, in water and recorded all changes that the uorogenic probes underwent with each analyte under a common TLC-UV light, l ¼ 366 nm, by qualitative measurements (photographs). As examples, photographs of BD116 and BD119 are shown in Fig. 1. These experiments showed that BD116 and BD119 were very efficient OFF-ON uorogenic probes for the selective detection of Hg 2+ in mixtures of methanol-water 20 : 80 v/v or acetonitrile-water 20 : 80 v/v with comparable efficiency. In contrast, the remainder of the probes were less efficient uorogenic probes in those mixtures of solvents or showed partial interaction with competing thiophilic cations such as Ag + and Au 3+ , when used in mixtures of solvents with decreasing proportion of water. None of the probes was sensitive to the presence of common anions neither in organic solvents nor in their mixtures with water.
We next tested 10 À4 M solutions of probes BD116 and BD119 by adding one to four equivalents of 5 Â 10 À3 M solution of MeHg + and compared the results to Hg 2+ . In this way, BD116 showed an enhanced uorescence in the presence of MeHg + , as well as Hg 2+ , in all tested solvent mixtures, whereas BD119 produced a uorescent response only in the presence of Hg 2+ . An immediate consequence is that the use of both probes should permit the speciation of MeHg + and Hg 2+ , because of the unique sensitivity of BD116 to MeHg + and Hg 2+ in contrast to the lack of sensitivity of BD119 to MeHg + , which was only sensitive to Hg 2+ .

Fluorescence quantitative experiments
For quantitative experiments we selected a concentration value of 2.5 Â 10 À5 M in which the absorption and the emission behaved in a linear fashion. UV/Vis and uorescent titrations were carried out by successive additions of each analyte to the cuvette containing the probe until saturation of the signal, Scheme 1 Synthesis of fluorogenic probes and their action. usually for concentrations ranging between 2.5 Â 10 À6 and 1.25 Â 10 À3 M. With these titrations we obtained stability constants, limits of detection and quantum yields in the presence of Hg 2+ and MeHg + . Titrations of BD116, 2.5 Â 10 À5 M in MeOH-H 2 O 20 : 80 v/v, with Hg 2+ showed a 20-fold increase of the initially weak uorescent emission (l exc ¼ 350 nm, l em ¼ 480 nm) and a 2-fold increase of the quantum yield. Fitting of the emission intensity titration curves to a 1 : 1 complex equation gave a stability constant K ¼ (1.9 AE 0.1) Â 10 4 M À1 for the reversible complexation between BD116 and Hg + . Replication of the titration ve times afforded a detection limit of (5.0 AE 1.0) Â 10 À6 M for the determination of Hg 2+ . Titrations of BD116 with MeHg + in the same conditions showed a 6-fold increase of emission band centred at 480 nm. Fitting of the emission intensity titration curves to a 1 : 1 complex equation gave a stability constant of K ¼ (5.7 AE 0.4) Â 10 4 M À1 , higher than the value obtained in the previous case, and replicates of the titration gave a detection limit of (9.9 AE 1.1) Â 10 À6 M for the determination of MeHg + , lower than the previous case. Fig. 2 shows the emission spectra and titration proles for complexation of BD116 and Hg 2+ or MeHg + . In turn, titrations of BD119 with Hg 2+ in the same conditions showed a 6-fold increase of the emission band (l exc ¼ 398 nm [isosbestic point], l em ¼ 554 nm). Fitting of the emission intensity titration curves to a 1 : 1 complex equation gave a stability constant of K ¼ (5.5 AE 0.5) Â 10 3 M À1 , much lower than the value obtained for the complex BD116/Hg 2+ (Fig. S27, ESI, p. S19 †) and replicates of the titration gave a detection limit of (6.79 AE 1.9) Â 10 À6 M for the determination of Hg 2+ , lower than in previous case. The stoichiometry of binding processes were conrmed by Job's plot experiments of the uorescence vs. the mole fraction of Hg 2+ or MeHg + of the three complexes (Fig. 2, Job's plot for BD116/Hg 2+ and BD116/MeHg + , Fig. S99, ESI, p. S78 † for BD119/Hg 2+ ). In all cases the curves reached a maximum for X Hg 2+ ¼ X MeHg + ¼ 0.5, conrming the formation of 1 : 1 complexes.

T-jump kinetic studies
The kinetic study of the binding processes for BD116/Hg 2+ and for BD116/MeHg + were monitored by means of the ultra-fast Tjump technique in conducting solutions having a constant ionic strength (I) of 0.1 M NaClO 4 , a concentration of the probe 2.5 Â 10 À5 M in MeOH-H 2 O 20 : 80 v/v and under metal cation excess conditions ranging from 1.25 Â 10 À4 to 3.75 Â 10 À4 M to ensure pseudo-rst order conditions. The processes were recorded by uorescent emission in 0.7 cm optical length cuvettes each containing a total volume of 1 ml. The system, initially in equilibrium, was perturbed up to a nal temperature of 25 C, and the relaxation process was observed by uorescent measurements. At least ve replicates were performed for each composition probe-cation and the outliers were discarded. One monoexponential kinetic effect was observed in both cases. Examples of the recorded kinetic traces are provided in Fig. 3A and B. Fitting of the kinetic curves by exponential functions enabled us to obtain the reciprocal relaxation times, 1/s. The kinetic analysis yielded the apparent binding reaction between the probe (P) and the metal, M (Hg 2+ or MeHg + ).
The linear dependence of the time constants (1/s) on the reactants concentrations according to eqn (2) is shown in Fig. 3C and D. From the slope/intercept values we calculated the kinetic equilibrium constant K ¼ k 1 /k 2 .
To a rst approximation, the values of [P] and [M] are deduced from the thermodynamic values of the overall binding  constant K obtained from static measurements, then they were re-evaluated from K ¼ k 1 /k 2 until convergence is reached. The linear tting to eqn (2) showed that the mechanism of PM formation was in accordance to eqn (1) in both systems. The values of k 1 , k 2 and K of the BD116/Hg 2+ and BD116/MeHg + systems are shown in Table 1.
The values for the equilibrium constant, K, denote greater affinity of the BD116 probe with MeHg + than with Hg 2+ . Table 1 shows that the values of the rate constants for formation, k 1 , and dissociation, k 2 , of the BD116/Hg 2+ complex were 8 and 20 times larger, respectively, than those of the BD116/MeHg + complex. That is, the S/MeHg + binding dissociates 20 times slower than S/Hg 2+ , revealing that in the equilibrium the hydrophobic nature of the interaction prevails. The K values were calculated also by uorescence titration, yielding very similar results (Table 1). Micromolar amounts of M were added to the probe directly into the cuvette, measuring the uorescence. Fig. 4 shows the binding isotherms according to the larger uorescence of PM compared to that of P. The monophasic form of the binding isotherm revealed that only one type of complex metal/probe was operative, according to the kinetic mechanism (eqn (1)).

Quantum mechanics calculations
To throw light to the fact that only probe BD116 was able to form stable complexes with MeHg + we performed DFT calculations by using the Gaussian 03 packing soware. 16 Full geometry optimization was performed at the B3LYP/6-31G**(Lanl2DZ for Hg) level of theory, performing the optimization in the gas phase. The preferred coordination process for all host and guest pairs took place through the sulfur atom of ligand molecules (a model of the complex BD116/MeHg + along with the frontier orbitals is shown in Fig. 5). This fact was conrmed by 1 H-NMR titration spectra of BD116 (5.0 Â 10 À3 M) in CD 3 CN with Hg 2+ , in which a downeld chemical shi for both methyl groups of the dimethylthiophosphinoic group was recorded as Hg 2+ was added, indicating that the coordination process took place by the close sulfur atom (ESI, Fig. S100, pp. S78-S79 †). The two close methylene groups of the piperazine moiety were also affected to a lesser extent by a weak downeld effect. DFT calculations also showed that complexation processes were spontaneous for BD116/Hg 2+ , BD116/MeHg + and BD119/Hg 2+ with calculated free energies of À91.8, À3.6 and À90.0 kcal mol À1 respectively, but in the case of BD119/ MeHg + the obtained free energy was positive (0.1 kcal mol À1 ), so the process was not spontaneous. This result was in agreement with the experimental behaviour of the probes.

Speciation of MeHg + and Hg 2+ in aqueous solution and in live cells
The different selectivity of probes to Hg 2+ and MeHg + was used for the chemical speciation of each analyte in aqueous test samples containing mixtures of both cations, performed for the rst time only with the help of uorogenic probes by the standard addition methodology (ESI, pp. S73-S74 †). Thus, BD119 permitted the selective analysis of Hg 2+ and BD116 permitted the determination of the sum of Hg 2+ + MeHg + . By comparison of the results obtained with each probe and a careful calibration of the uorescent titrations, to compensate for the different sensitivity of each probe to each analyte, the concentration of each analyte was obtained. In this way, a representative sample  (100 or 300 mM) in PBS for 1 h and the uorescent emission was measured by using l exc ¼ 388 or 358 nm. Cells remained viable aer incubation in the presence of the probes, which were permeable to the cellular membrane. Controls of cells without probe were measured as blanks and the relative intensity of intracellular Hg 2+ uorescence was compared. Probe BD116 showed low levels of background intracellular uorescence in the absence of Hg 2+ but the intracellular uorescence dramatically increased with addition of Hg 2+ up to 300 mM. The emission was higher in the nucleus than in the cytoplasm of cells, due to a higher accumulation of the probe in the nucleus (Fig. 6). Similar experiments with cell cultures incubated with BD116 and successive addition of MeHg + gave very weak differences in uorescence in the absence or in the presence of MeHg + . Probe BD119 showed only very little increase of uorescence with addition of Hg 2+ in the same conditions. There was no effect of the presence of MeHg + in experiments using BD119.

Interactions with DNA
The unusual accumulation of the probe BD116 in the nucleus suggested that the probe could interact with DNA. A deeper study on the interaction of BD116 and ctDNA alone, or incubated with MeHg + , was conducted at pH ¼ 7.0 and ionic strength 0.1 M with absorbance, uorescence, circular dichroism (CD), viscosity and differential scanning calorimetry techniques. For the BD116/ctDNA system (in the absence of MeHg + ), Fig. S88A-D (ESI, p. S57 †), show the formation of two types of BD116/DNA complexes, observed by both absorbance and uorescence. The analysis of the BD116/ctDNA system was performed in DMSO-water mixtures containing upto 3% DMSO. The binding process was described by the apparent reaction (4): where the probe (P) interacted with ctDNA to give the bound species (PDNA). The equilibrium constant was dened as K ¼ ). The titration curves were analysed at 344 nm according to eqn (5).
In eqn (5) C DNA and C P are the total dye and polymer concentrations, respectively; DA ¼ A À 3 P C P is the change of absorbance (A) during titration, and 3 P ¼ DA /C P , where A denotes the initial absorbance of the dye solution; D3 ¼ 3 PDNA À 3 P is the amplitude of the binding isotherm. A similar equation was used in uorescence titration experiments, by changing DA by DF and D3 by Df in eqn (5). Fitting according to eqn (5) required an iterative procedure as D3 was not known. This was put equal to zero in the le term of the equation in a rst approximation, and then calculated from the reciprocal slope of the plot; calculation was repeated until convergence was reached (usually three iterations only). A similar behavior was found when ctDNA was replaced by ctDNA incubated overnight with MeHg + (Fig. S89A-D, ESI, p. S57 †). Table 2 shows the values of the binding constants obtained for both systems by absorbance and uorescence.
The binding constants were the same order of magnitude, even though the presence of MeHg + reduces its value by a half due to the DNA/MeHg + interaction. To verify the type of binding, viscometric, CD and DSC measurements were performed at constant ctDNA and varying the probe concentration. Fig. S90A and B (ESI, p. S58 †) showed that, in the absence of MeHg + , the relative length of DNA remained the same over the  whole concentration range, whereas only a small decrease in viscosity was observed in the presence of MeHg + . Fig. S90C and D (ESI, p. S58 †) showed that the structural differences observed with CD were only modest, even though the molar dichroism changed when ctDNA was incubated with MeHg + (larger separation of the bands at 278 nm). Fig. S90E and F (ESI, p. S58 †) showed that the variation in the melting temperature obtained by DSC for the two systems remained unchanged upon increase in the probe concentration. The set of results leads to the conclusion that the BD116 probe interacts along the groove of DNA, both in the absence and in the presence of MeHg + . The small structural alterations caused by MeHg + entailed diminution of the affinity of BD116 with ctDNA.
Preparation of MeHg + uorogenic probes for live cell imaging All those experiments established that the structure of BD116 was a good starting point for the preparation of selective MeHg + uorogenic probes for live cell imaging. With this in mind, we synthesized new water soluble derivatives of 1a that retained sensitivity to Hg(II) derivatives (Scheme 2). The syntheses of probes were performed in good yields by reaction of 1a with commercial 1-azido-4-isothiocyanatobenzene 3 followed by click reaction with 2,5,8,11,14-pentaoxaheptadec-16-yne 4, usually employed to solubilize drugs for cell penetration, 17 or the methoxy and 4pentynoate-terminated poly(ethylene glycol) (n (average) ¼ 43) 5, also used for bioconjugation, 18 or the tri-PEG (2-propyn-1-yloxy) methylbenzene derivative 6, closely related to a commonly used starting material for water solubilizing dendrimers, 19 in all cases under catalysis by [Cu(NCCH 3 ) 4 ] + PF 6 À and tris((1-benzyl-1H-1,2,3-triazol-4-yl)methyl)amine (see ESI †). The simplest derivative JG15 kept the OFF-ON uorogenic selective sensing of Hg 2+ in methanol-water 90 : 10 mixture, proving the versatility of these probes, but was not soluble enough to work in pure water. The probes JG47 and JG45 showed a similar sensitivity for Hg 2+ in water as solvent, albeit JG45 was much more sensitive to MeHg + in organic solvents (ESI, Fig. S81, p. S53 †); therefore these two probes were tested for imaging and speciation of Hg(II) species in image microscopy. HEK293 cells were incubated with JG47 or JG45 solutions (100 mM in PBS with Ca 2+ and Mg 2+ ) for 1 h at 37 C. Then the plates were washed three times with PBS and incubated with Hg 2+ (100-500 mM HgClO 4 ) or MeHg + (100-400 mM MeHgCl) in PBS + Ca 2+ + Mg 2+ for 1 h and the uorescent emission was measured by exciting at l exc ¼ 388 nm. Cells remained viable aer incubation in the presence of the probes, which were permeable to the cellular membrane. Controls of cells without probe were measured as blanks and the relative intensity of intracellular Hg 2+ and MeHg + uorescence were compared. Probes JG47 or JG45 showed low levels of background intracellular uorescence in the absence of Hg 2+ or MeHg + . There was very little change in the intracellular uorescence when JG47 was tested in the presence of Hg 2+ or MeHg + . The change in uorescence was also very small for JG45 and Hg 2+ . However, the intracellular uorescence dramatically increased by addition of MeHg + up to 400 mM to the HEK293 cells incubated with JG45. The emission was observed almost exclusively in the nucleus of cells (Fig. 7) with very dim uorescence in the cytoplasm. Actually, in this case, the selective enhancement of uorescence of probe JG45 in the presence of MeHg + inside the nuclei of HEK293 cells was due to lipophilicity of both MeHg + and JG45, which tended to concentrate in the nucleus of cells, acting as an optimum lipophilic environment for interaction of probe and cation. This fact was a reection of the previously observed behaviour of JG45 in organic solvents and constitutes a new paradigm for the design of selective uorescent probes for Scheme 2 Synthesis of water-soluble fluorogenic probes. imaging MeHg + on the basis of a sensitivity linked to lipophilicity.

Conclusions
In summary, we have prepared new uorogenic probes that interact in different ways with two closely related cations of high environmental concern, Hg 2+ and MeHg + . The chemical probes were used for the chemical speciation of both cations in organic-aqueous solvents as well as in HEK293 cells. By far, the best selective speciation of Hg 2+ and MeHg + has been achieved by in vitro approaches based on the uorogenic probes supported in cultured cells, due to the particular sensitivity of the HEK293 cells to permeation by Hg 2+ , MeHg + and the uorogenic probes. These achievements provide the biochemical bases to the understanding of MeHg + selective detection and imaging, contributing to the discovery of endogenous and exogenous molecular probes that provide efficient means for speciation between Hg(II) species.