Article pubs.acs.org/IC
A Switch-On NIR Probe for Specific Detection of Hg2+ Ion in Aqueous Medium and in Mitochondria Hridesh Agarwalla,∇,† Pankaj S. Mahajan,† Debashis Sahu,‡ Nandaraj Taye,§ Bishwajit Ganguly,*,‡ Santosh B. Mhaske,*,† Samit Chattopadhyay,*,⊥,§ and Amitava Das*,∥,† †
Organic Chemistry Division, CSIR-National Chemical Laboratory, Pune 411008, India Chromatin and Disease Biology Lab, National Centre for Cell Science; Pune 411007, India § Computation and Simulation Unit (Analytical Discipline and Centralized Instrument Facility), CSIR−Central Salt and Marine Chemicals Research Institute, Bhavnagar 364002, India ∥ CSIR-Central Salt & Marine Chemicals Research Institute, Bhavnagar 364002, Gujarat, India ‡
S Supporting Information *
ABSTRACT: A new 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (BODIPY)-based probe molecule (L) is synthesized for specific binding to Hg2+ ion in physiological condition with an associated luminescence ON response in the near-IR region of the spectrum. Appropriate functionalization in the 5-position of each of two pyrrole moieties with styryl functionality in a BODIPY core helped us in achieving the extended conjugation and a facile intramolecular charge transfer transition with a narrow energy gap for frontier orbitals. This accounted for a poor emission quantum yield for the probe molecule L. Binding to Hg2+ helped in interrupting the facile intramolecular charge transfer (ICT) process that was initially operational for L. This resulted in a 2+ hypsochromic shift of absorption band and a turn-on luminescence response with λEms Max of 650 nm on specific binding to Hg . Observed spectral changes are rationalized based on quantum chemical calculations. Interestingly, this reagent is found to be localized preferentially in the mitochondria of the live human colon cancer (Hct116) cells. Mitochondria is one of the major targets for localization of Hg2+, which actually decreases the mitochondrial membrane potential and modifies various proteins having sulfudryl functionality(ies) to cause cell apoptosis. Considering these, ability of the present reagent to specifically recognize Hg2+ in the mitochondrial region of the live Hct116 cells has significance.
■
be 2 ppb.3 Therefore, considerable efforts have been put together for developing appropriate reagent and the subsequent methodology for selective and sensitive detection of Hg2+ ion that could be present in an environmental or biological sample at a very low concentration. Over the years, several methods have been developed, for example, atomic emission spectroscopy,4 inductive coupled plasma mass spectroscopy,5 inductively coupled plasma−atomic emission spectroscopy6 for detection of Hg2+ ion in environmental samples. On the one hand, most of these methods are expensive, requiring tedious sample preparation and involvement of highly skilled manpower. On the other hand, use of molecular sensors that allow detection and quantitative estimation of the Hg2+ ion through measurable changes in optical responses have a distinct edge primarily due to the ease in detection process/methodology as well as these methodologies allow detection of the Hg2+ ion present in the ultra trace quantity. Among various optical sensors, fluorescence-based sensors are obviously preferred, as these allow developing
INTRODUCTION With the advent of the modern era and industrialization the problem of diffusion of heavy metal ions into the environment has been significantly enhanced primarily due to the burning of fossil fuels and coals, mining, chemical industry, and battery industry. Subsequently, detection of heavy metal ions became one of the key issues for environmental sample analysis and impact assessment as well as for clinical biology/diagnosis and analytical chemistry.1 Among these heavy metal ions, mercury is considered to be one of the most toxic and is known to be a potent neurotoxin in its various forms for human physiology and other living organisms. Among various forms, Hg2+ is the most common form of ionic mercury that exists in the environment. Redox transformation between elemental Hg(0) and divalent Hg(II) is a key process that leads to the transport of mercury in groundwater systems, and this is actually induced by diverse anaerobic bacteria.2a Bioaccumulation of Hg2+ in human body through food chain is a major source of serious neural disorder, Hunter-Rusell syndrome, and diseases like Alzheimer’s and Minamata.2 Considering the deleterious effects of Hg2+ in human health, the Environmental Protection Agency has set the maximum allowed level of Hg2+ in drinking water to © XXXX American Chemical Society
Received: September 14, 2016
A
DOI: 10.1021/acs.inorgchem.6b02233 Inorg. Chem. XXXX, XXX, XXX−XXX
Inorganic Chemistry
■
Article
RESULTS AND DISCUSSIONS In this present study, we prepared a new BODIPY derivative L having an appended dipicolylamine moiety (Scheme 1) for
noninvasive methodologies with the possibility of using such reagent for in vitro or in vivo imaging application, clinical diagnosis, and studying biochemical processes associated with Hg2+ in living organisms.7,8 There have been many previous attempts for developing fluorescence-based sensors for Hg2+ ion using small molecular probes, oligonucleotides, gold nanoparticles, etc.9−11 Among such attempts, reagents that are capable of exhibiting modified luminescence response on binding to Hg2+ in the near-infrared (NIR) region (>650 nm) of the spectrum have a special significance for any imaging application, as this helps in delineating the response(s) from the background auto fluorescence of intrinsic biomolecules that generally occur in the visible region.12 Additionally, deep-red light as the excitation source allows minimum photobleaching of the probe/photodamage of the living tissues/cells, deeper tissue penetration, and low light scattering.13 Further, organellespecific dye localization also offers us the option to monitor Hg2+ localization in that specific cellular compartment, and this is important for clinical impact assessment.14,15 It is welldocumented that mitochondria are one of the major targets for localization of heavy metals. Several reports suggest mercury decreases mitochondrial membrane potential in hepatic and renal mitochondria.16 It has been argued that mercury reacts with proteins having sulfudryl group in the mitochondrial inner membrane and generates oxidative stress to induce nephrotoxicity and eventually cell apoptosis.17 However, the exact mechanism for this process is yet to be fully established.18 Considering these, there is a definite need and scope for developing an efficient NIR active molecular receptor that could effectively recognize and image the localization of Hg2+ in the mitochondrial region of live cells. To the best of our knowledge there are only two previous reports in contemporary literature that have addressed this crucial issue.19 Boron dipyrrolemethanes (BODIPY) are an important class of fluorescent molecules, which are being used for imaging application owing to high emission quantum yield, narrow absorption/emission bands, higher photostability, ease in derivatization for tuning their optical/physicochemical properties, and in general for appropriate lipophilicity for improved cell membrane permeability.20 Simple BODIPY derivatives generally exhibit shorter emission (∼500 nm) band. However, such emission responses could be fine-tuned for achieving luminescence responses in the NIR region by adopting proper synthetic strategy by appropriate functionalization. One such strategy involves extending conjugation at pyrrolic position. Styryl substitution at 5-position helps in achieving BODIPY derivatives that have absorption and emission bands in the NIR region, which make them ideally suited for imaging application.21 In this article, we have described design and synthesis of a distyryl derivative of BODIPY that shows a selective turn-ON emission response in NIR region upon interaction with Hg2+ in physiological condition. The reagent shows high specificity toward Hg2+ with very low detection limit. Our description on luminescence responses are rationalized based on computational studies. This nontoxic reagent could permeate through the cell membrane of the human colon cancer (Hct116) cells and is found to be localized specifically in mitochondrial region enabling us to map the distribution of Hg2+ in that specific organelle.
Scheme 1. Methodology Adopted for Synthesis of the Probe L
coordination to a metal ion. Synthetic procedure as well as purification of the probe molecule (L) and all intermediates (2, 3, and 4) are given in the Experimental Section. The purity of these intermediates was ensured by standard analytical and spectroscopic analysis. Spectroscopic and analytical data confirmed the desired purity for the receptor molecule L (Supporting Information Figures S1−S4). The newly synthesized probe L was found to have limited solubility in pure aqueous HEPES buffer medium. Accordingly, we used a predominantly aqueous buffer (10 mM aqueous HEPES buffer−acetonitrile, 3:7 (v/v); pH 7.2) medium for our entire studies. Electronic spectrum recorded for L in various solvents showed a hump at ∼625 nm and another prominent band with maximum at ∼680 nm. Both these bands were found to move toward longer wavelength with increase in solvent polarity (Supporting Information Figure S5), which signified the charge-transfer (CT) nature of the transitions associated with these absorption bands where amine functionalities (−NMe2 and −N(CH2C6H5N)2) are expected to be the donor fragments and the dipyrrole-core as the predominant acceptor fragment (vide infra). Other low-energy bands appeared at ∼330 and 420 nm, which could be attributed to the predominatly π−π*-based transitions. In aqueous HEPES buffer−acetonitrile (3:7, v/v; pH 7.2) medium the green colored solution of the reagent L displayed an intense absorption band at 685 nm (ε = 5150 M−1 cm−1). We presumed that on interaction between the dipicolyl amine moiety and a metal ion, this intramolecular charge transfer (ICT) transition would be interrupted or disfavored, and accordingly a blue shift in the absorption band was anticipated. The metal ion recognition ability of the reagent L was systematically examined by recording UV−vis spectra in absence and presence (100 mol equiv) of various metal ions in aqueous HEPES buffer−acetonitrile (3:7, v/v; pH 7.2) medium keeping [L] fixed at 2 μM. Metal ions that are common for human physiology and are generally available in surface water, such as Na+, K+, Cs+, Ba2+, Ca2+, Mg2+, Cd2+, Ni2+, Cu2+, Zn2+, Pb2+, Co3+, Cr3+, Fe2+, Fe3+, and Hg2+, were utilized for our studies. Results revealed that barring Hg2+, all other metal ions failed to induce any detectable change in the observed electronic spectrum of the reagent L. For Hg2+ ion, a 15 nm blue-shifted spectrum with little decrease in molar absorptivity (Figure 1) was observed. The blue shift of the absorption band may be correlated to a disfavored ICT process. However, changes were more prominent when steady-state emission spectrum for L was recorded in absence and presence of various metal ions mentioned above. The emission spectrum B
DOI: 10.1021/acs.inorgchem.6b02233 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
reagent L (2 μM) in the essentially aqueous buffer medium remained unchaged for all metal ions, except for Hg2+. A new enhanced emission band appeared at 650 nm (λExt = 620 nm). The excitation spectrum recorded for L (λEms Max = 750 nm) showed band maxima at 680 nm, while this in the presence of excess Hg2+ showed a band at 620 nm (λEms Max = 650 nm), and this implied that the final emission states for L and L·Hg2+ were distinctly different (Supporting Information Figure S7). A little and insignificant enhancement of emission band intensity at 650 nm was observed in the presence of Zn2+ (100 mol equiv). The observed preference of L toward Hg2+ over Zn2+ could be due to the higher hydration enthalpy of Zn2+ (−2046 kJ mol−1) than Hg2+ (−1824 kJ mol−1).29 A significant increase in the emission quantum yield (LΦ = 0.026 and L·Hg2+Φ = 0.12) was observed on binding of the reagent L to Hg2+. (Cresyl violet perchlorate in ethanol with Φ = 0.54 was used as reference.) This remarkable change in the emission spectrum confirmed the specificity of the reagent L toward Hg2+ ion. However, a little and no change in emission spectral pattern for L in the presence of excess Zn2+ ion and all other metal ions (except Hg(II)), respectively, did not completely exclude the possibility of a weak/very weak coordination complex formation between L and Zn2+ or other metal ions in aqueous HEPES buffer− acetonitrile (3:7, v/v; pH 7.2) medium. We also examined the preference of the reagent L (2 μM) toward Hg2+ (100 μM) in the presence of even larger (400 μM) excess of aforementioned cationic analytes (Supporting Information Figure S8). Results of this interference study revealed that the spectral change observed for L on binding to Hg2+ remained unaffected even in the presence of large excess of all other competing cations, and this confirmed the specificity of the reagent toward Hg2+. Binding affinity of the reagent L toward Hg2+ was studied by systematic emission titration studies using 2 μM of L and varying [Hg2+] (0−150 mol equiv) in aqueous HEPES buffer− acetonitrile (3:7, v/v; pH 7.2) medium (Figure 2). With increase in [Hg2+], a gradual increase in emission intensity at 650 nm was observed. Using parameters that were obtained from this titration plot, formation constant (Ka) for L·Hg2+ was evaluated using Benesi−Hildebrand equation, and it was found to be (4.93 ± 0.2) × 103 M−1. A good linear fit of Benesi− Hildebrand plot also confirmed the 1:1 binding stoichiometry for the complex formation between Hg2+ and L (Supporting Information Figure S9). Linear change in emission intensity at
Figure 1. Absorption spectra of probe L (2 μM) in absence and presence of different metal ions (Na+, K+, Cs+, Ba2+, Ca2+, Mg2+, Cd2+, Ni2+, Cu2+, Zn2+, Pb2+, Co3+, Cr3+, Fe2+, Fe3+) in CH3CN−aqueous HEPES buffer (10 mM, 3:7, v/v; pH 7.2) medium.
of the reagent L (2 μM) in an essentially aqueous HEPES buffer medium (pH 7.2) showed a weak band at ∼750 nm (LΦ = 0.026) on excitation at either 680 or 620 nm. This suggested that both excited states were in equilibrium at room temperature. Weak emission response for L could only be ascribed to a relatively narrow energy gap for frontier orbitals. Solvatochromic effects were more pronounced for emission spectra recorded in solvents of varying polarities (Figure 2 and Supporting Information Figure S5). This further corroborated our presumption about the ICT-based transitions associated with these bands. Possibility of the twisted intramolecular charge transfer (TICT) process was also considered. Accordingly, emission spectra for L were recorded in a highly viscous and polar solvent like glycerol. In a highly viscous solvent, deactivation processes associated with nonrigid conformations for TICT process would be restricted, and this would have accounted for emission intensity enhancements. On the contrary, for L a decrease in emission intensity with bathochromic shift was observed. These helped us to presume that the emission band at 750 nm with low-emission quantum yield is primarily associated with facile CT process (Supporting Information Figure 6). On addition of 100 mol equiv of different metal ions described above, emission spectrum of
Figure 2. Emission spectra of probe L (2 μM) in (a) absence and presence of different metal ions, (b) presence of varying concentration of Hg2+, (c) Benesi−Hildebrand plot of 1/[F − F0] vs 1/[Hg2+] in CH3CN−aqueous HEPES buffer (10 mM; 3:7, v/v; pH 7.2). λExt = 620 nm was used for studies. C
DOI: 10.1021/acs.inorgchem.6b02233 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Figure 3. (a) Partial 1H NMR spectra for L recorded in absence and presence of Hg2+ in CD3CN; (b) molecular structure for Hg2+·L showing the possible binding mode for Hg2+ ion.
Figure 4. M06-2X/6-31G(d) optimized geometries with their relative energies (Erel) of (a) probe (L), (b) [L + Hg2+ + 2ClO4−], (c) [L + Hg2+ + 2H2O], (d) [L + Hg2+ + 1ClO4−], and (e) [L + Hg2+ + 1H2O + 1ClO4−]. The energies are in kilocalories per mole, and the distances are in angstroms. Key: gray, C; green, Cl; orange, B; blue, N; white, H; red, O; cyano blue, F; deep maroon, Hg2+.
650 nm was observed for the concentration range of [Hg2+] = 0−1.5 × 10−5 M, and lowest detection limit (LOD) was evaluated as 1.7 × 10−7 M by using 3σ/k method (Supporting Information Figure S10). Thus, this reagent could be used for quantitative estimation of Hg2+ that could be present in aqueous environment in sub-micromolar region. Emission spectral responses were also examined in the presence of
varying [HgCl2] for ascertaining the role of the effective acidity of the Hg center in binding to L.9f Systematic titration profile with varying [HgCl2] helped us in evaluating the association constant (KaHgCl2 = (2.96 ± 0.3) × 103 M−1) using B−H plot, while LOD was evaluated as 4.7 × 10−7 M (Supporting Information Figures S11 and S12). Slightly lower value for KaHgCl2 presumably reflects the relatively lower ionic character D
DOI: 10.1021/acs.inorgchem.6b02233 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Figure 5. (a) DFT (M06-2X) calculated HOMO and LUMO of probe (L) and [L + Hg2+ + 2ClO4−] and (b) simulated UV−vis spectra.
[L + Hg2+ + ClO4−] and two ClO4− (abbreviated as [L + Hg2+ + 2ClO4−]) to the Hg2+ ion for our studies (Table S1). Relative complexation energies (Erel) with respect to the most stable complex [L + Hg2+ + 2ClO4−] (Erel= Ecomplex − Estable complex) for all the above-referred complexes were evaluated (Figure 4, Supporting Information Table S1). The calculated results revealed that Hg2+ complex coordinated to two ClO4− (L + Hg2+ + 2ClO4−) has the lowest energy (Figure 4). This lowest-energy complex supports the results of the mass spectral analysis described previously. To get a better insight in the transitions associated with L and L + Hg2+ + 2ClO4− (lowest-energy form of the complex) and the spectroscopic behavior, we simulated the UV−vis spectrum for both species using TD-DFT calculations.27 The highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) for two species are shown in Figure 5. Figure 5a reveals that the HOMO orbital coefficients of the reagent L are delocalized over the π framework of the BODIPY and extended up to the styryl group. In the case of LUMO, the electron densities are mainly localized in the BODIPY core, which indicates a CT transition from styryl to BODIPY moieties. On complexation with Hg2+, electron densities for HOMO are located mainly on the dimethylamino arm and the BODIPY core with less electron density at dipicolylamine arm. The redistribution of electron density on binding to Hg2+ indicates a disfavored ICT process, which also supports results of the UV−vis spectra. The first excited states of L (oscillator strength, f = 0.3756) and the complex [L + Hg2+ + 2ClO4−] (oscillator strength, f = 0.4245) are characterized as the HOMO−LUMO one-electron excited states. The theoretical vertical excitation wavelengths for the reagent L and the complex, [L + Hg2+ + 2ClO4−] in the gas phase are at 588 and 514 nm, respectively (Figure 5). The calculated blue shift in the UV−vis spectra also corroborates well with experimental UV−vis spectra discussed previously. Cell Study. For evaluating the feasibility of using the reagent L as an imaging agent for mapping the cellular uptake of Hg2+, cytotoxicity of L toward human colon cancer cells (Hct116
for Hg center in HgCl2, as compared to Hg(ClO4)2. The pH influence of the reagent in sensing was also investigated (Supporting Information Figure S13). This reagent showed higher sensitivity for the pH range of 6−8, which was just ideal for studies under physiological studies. The probe reacts with Hg2+ very quickly and reaches plateu in
