A BODIPY-based fluorescent probe for the differential recognition of Hg(ii) and Au(iii) ions

In recent years, research on the development of molecular sensors for analysing diverse biologically and environmentally important analyte species has increased. While the vast majority of sensors addressed in such literature are designed to recognize a specic target, less common are sensors capable of differentiating multiple targets. Differential detection of multiple analyte species can be achieved best by recognizing each species through a different signal output (i.e., emission wavelength). Incorporating multiple binding motifs onto a single signal-transducing molecule (chromophore/uorophore) and, as an alternative, combining different transducing molecules have both appeared as efficient routes for sensors with multiple output modes. Despite recent advances in the eld, it remains a challenge to differentiate metal species with similar chemical natures. For example, the ionic species of gold (Au) and mercury (Hg) share several similarities in terms of binding properties, since both have strong binding affinities toward sulphur species. When accumulated in the biological system, they thus have great potential to interact with sulphur-bearing biomolecules such as enzymes, proteins, and DNA. As a result, these metal species can disturb a series of cellular processes that cause toxicity in humans. Tracking these metal species in a living environment with the aid of a uorescent molecular sensor is thus crucial to evaluating their roles in certain biological processes.

In recent years, research on the development of molecular sensors for analysing diverse biologically and environmentally important analyte species has increased. 1 While the vast majority of sensors addressed in such literature are designed to recognize a specic target, less common are sensors capable of differentiating multiple targets. Differential detection of multiple analyte species can be achieved best by recognizing each species through a different signal output (i.e., emission wavelength). 2 Incorporating multiple binding motifs onto a single signal-transducing molecule (chromophore/uorophore) and, as an alternative, combining different transducing molecules have both appeared as efficient routes for sensors with multiple output modes. 3 Despite recent advances in the eld, it remains a challenge to differentiate metal species with similar chemical natures. For example, the ionic species of gold (Au 3+ ) and mercury (Hg 2+ ) share several similarities in terms of binding properties, since both have strong binding affinities toward sulphur species. When accumulated in the biological system, they thus have great potential to interact with sulphur-bearing biomolecules such as enzymes, proteins, and DNA. As a result, these metal species can disturb a series of cellular processes that cause toxicity in humans. 3,4 Tracking these metal species in a living environment with the aid of a uorescent molecular sensor is thus crucial to evaluating their roles in certain biological processes.
In their structure, most molecular sensors devised for mercury ions use sulphur moieties as a recognition motif. 5 As such, it is always possible that gold and mercury species interfere with each other during their analysis, which could primarily explain why molecular sensors able to differentiate these two metal ions are extremely rare. 6 Molecular tools that can differentiate multiple analytes of a similar chemical nature (e.g., gold 7 and mercury 5,8 ions) are therefore clearly in high demand.
Herein, we designed a molecular sensor with a single uorophore core appended with semithiocarbazone functionality as the metal ion recognition motif. The uorophore core, based on a BODIPY dye, 9 is designed to be inactive (i.e., non-emissive) in its initial state yet expected to become active in response to the metal species (Scheme 1). The differential detection of Hg 2+ and Au 3+ relies on different modulation mechanisms and can be realized with two distinct uorescence changes, based on either an Hg 2+ -ligand coordination event or a gold-mediated chemical transformation.
The title compound, BOD-ZN, was prepared (40% overall) by the synthetic route outlined in Scheme 1 and, its structure was unambiguously conrmed by 1 H-NMR, 13 C-NMR, and HRMS spectroscopy. 10 The sensing behaviour of BOD-ZN toward the addition of a range of metal ion species was studied by UV/Vis and uorescence spectroscopy. As shown in Fig. 1a, the UV/Vis spectrum of free BOD-ZN (phosphate buffer/ethanol 1 : 4, pH 7.0) displays a maximum absorption band at 533 nm, which belongs to the BODIPY chromophore. The uorescence spectrum of BOD-ZN collected upon excitation at 460 nm exhibits a very weak emission band at 601 nm. Reasonably, BOD-ZN was nearly nonemissive, since the molecular structure of the probe bears a C]N functionality that diminishes the emission of the BODIPY core caused by a non-radiative deactivation process involving the rapid isomerization of the C]N group.
Our investigation began with the evaluation of the optical behaviour of BOD-ZN in response to the addition of Au 3+ ions (e.g., AuCl 3 ). The addition of Au 3+ (1 equiv.) to BOD-ZN prompted the appearance of a new emission band at 512 nm that was assigned to the formation of a new BODIPY derivative (Fig. 1b). The appearance of this new band was accompanied with a distinct change in the solution's emission colour; the redemitting probe solution became distinctly green, as was clearly visible to the naked eye (Fig. S19, ESI †).
The compound displaying such green emission was isolated and further characterized as BOD-AL, the hydrolysis product of BOD-ZN (Scheme 2). Evidently, the recognition of Au 3+ was based on an Au 3+ -mediated hydrolysis reaction that resulted in the formation of a highly emissive BODIPY derivative (BOD-AL).
A systematic titration of BOD-ZN with Au 3+ reveals that emission band intensity increases linearly with the increase in concentration of Au 3+ in the range of 0.1-100 mM (Fig. S6, ESI †). At the same time, the kinetic study showed that the spectral response toward the addition of Au 3+ was rapid (<1 min) and that the emission intensity plateaued within 20 min. Due to the addition of 10 equiv. of Au 3+ , which thereby enhanced intensity by over 200-fold. Moreover, the minimum amount of Au 3+ detectable was evaluated to be 128.0 nM based on the signal-tonoise ratio (S/N ¼ 3) (Fig. S2, ESI †).
Having established that the detection of Au 3+ ions relies on an irreversible chemical reaction, we investigated the selectivity prole of BOD-ZN in response to other metal species. The probe proved to be highly specic for Au 3+ ions, since no change was detected in the spectrum when Au + ions were present. Furthermore, BOD-ZN showed no spectral response to other metal ions such as Cu 2+ , Ag + , Zn 2+ , Pb 2+ , Ni 2+ , Na + , Mg 2+ , Li + , K + , Pd 2+ , Co 2+ , Cd 2+ , Ca 2+ , Ba 2+ , Fe 3+ , and Cr 3+ , except for Hg 2+ .
We recognized the detection of Hg 2+ with a different emission output. Given the addition of Hg 2+ , the weakly emissive probe solution immediately (<5 s) turned profoundly yellow, possibly due to the blockage of the C]N isomerization via strong coordination to the nitrogen electron pair which prevents the electron transfer process, and thus enhances the uorescence emission (Scheme 3).
Meanwhile, in the uorescence spectrum of BOD-ZN, a new emission band appeared at 542 nm and increased linearly with the increased concentration of Hg 2+ above the range of 0.1-100 mM (Fig. 2b). The detection limit of BOD-ZN for detecting Hg 2+ was 160.0 nM.
In sharp contrast to the detection of Au 3+ , detecting Hg 2+ ions proceeded reversibly. The reversibility of the binding event between BOD-ZN and Hg 2+ was conrmed by the addition of Na 2 S to the solution pre-treated with Hg 2+ , which sharply decreased the emission intensity. The regeneration of uorescence was again made possible by introducing Hg 2+ ions into the solution, and the off-on switching ability of the system with Hg 2+ proved the reversibility of the process (Fig. S15, ESI †).
At the same time, the binding process of Hg 2+ to BOD-ZN could be clearly followed by the aid of 1 H-NMR spectroscopy. Scheme 2 Hydrolysis mechanism of BOD-ZN in the presence of Au 3+ ions.
During Hg 2+ incubation lasting 5 min, we observed pronounced differences in the 1 H-NMR spectrum of BOD-ZN. For one, the resonance of the H a proton signal at 8.01 ppm belonging to the hydrogen atom of the aldimine shied to a higher frequency, while the resonance of the methyl protons (H b and H c ) in close proximity to the recognition motif shied to a lower frequency. Furthermore, a reorganization of the phenyl ring proton signals strongly suggests the structural modication in the phenyl thiourea motif (Fig. 3).
The stoichiometry of the sensing event was established by following the Benesi-Hildebrand method and, accordingly, the related binding constant was determined as 4.2 Â 10 4 M À2 . 11 With HRMS analysis, we also conrmed the binding of Hg 2+ ions to BOD-ZN. HRMS data of the solution (Hg 2+ /BOD-ZN) indicated the formation of an Hg 2+ ion complex, (m/z ¼ 1204.36841 found; 1204.36354 calc.), with a binding stoichiometry of 1 : 2. 10 Given all of the above, the structure of the binding complex is most likely that shown in Scheme 3.
Having claried the detection of both metal species, we assessed the interference of other metal ions in the detection of Au 3+ and Hg 2+ . Although the spectral response of BOD-ZN induced by Au 3+ ions showed no interference with other metal ions, the detection of Hg 2+ was disturbed in the presence of Au 3+ ions. More specically, in the presence of Au 3+ and Hg 2+ , the uorescence spectrum initially displayed an emission band at 542 nm for the Hg 2+ /BOD-ZN binding complex. However, the band disappeared within 10 min, as a new band appeared at 512 nm, which indicates that Au 3+ ions also mediate the hydrolysis of the Hg 2+ binding complex (Fig. S18, ESI †). Notably, the same sensing behaviour was also observed by adding Au 3+ ions to a solution pre-treated with Hg 2+ .
Relying on the impressive sensing properties of BOD-ZN, we next investigated its capacity for imaging Hg 2+ and Au 3+ ions in living cells. As Fig. 4a and a 0 show, the images of human lung adenocarcinoma (A549) cells incubated with BOD-ZN did not display any uorescence until the addition of the metal species. However, upon incubation with Au 3+ or Hg 2+ , the cells started to emit a distinct uorescence emission consistent with results obtained in the solution. Based on the nucleus staining experiment using DAPI as the staining dye, we concluded that the probe passes through the cell membrane and detects both   metal species from within the cell. This preliminary cell imaging study suggested that BOD-ZN can be used efficiently for the in vitro imaging of Au 3+ and Hg 2+ species in living cells.

Conclusions
To close, we have studied the design, spectral behaviour, and cell-imaging capacity of a unique uorescent molecular structure that can efficiently differentiate Hg 2+ and Au 3+ ions. The differential detection of Hg 2+ and Au 3+ was recognized in two distinct uorescence changes: one resulting from a reversible Hg 2+ /sensor complex formation, the other an irreversible Au 3+mediated hydrolysis reaction. This novel molecular structure displayed the ability to recognize both metal species at nanomolar levels.