10 ORGANIC

LETTERS

Recognition of Hg2þ Ion through 2012

Vol. 14, No. 12
Restricted Imine Isomerization: 2980–2983

Crystallographic Evidence and

Imaging in Live Cells
Amal Kumar Mandal,† Moorthy Suresh,† Priyadip Das,† E. Suresh,† Mithu Baidya,‡
Sudip K. Ghosh,*,‡ and Amitava Das*,†

Central Salt and Marine Chemicals Research Institute (CSIR), Bhavnagar 364002,
Gujarat, India, and Department of Biotechnology, Indian Institute of Technology,
Kharagpur, West Bengal 721302, India
[email protected]; [email protected]

Received April 14, 2012

ABSTRACT

A newly synthesized imine-based receptor (L) showed remarkable specificity toward the Hg2þ ion in aqueous media over other metal ions.
Coordination of L to Hg2þ induces a turn-on fluorescence response. This was explained based on the restricted imine isomerization along with
PET on coordination to Hg2þ. X-ray structural evidence tends to favor a CC bond rotation rather than CdN isomerization for adopting a favorable
conformation in L for coordination to Hg2þ. This reagent could be used for imaging the accumulation of Hg2þ ions in HeLa cells.

Mercury is one of the most prevalent toxic metals in the facile, and applicable to the environmental and biological
environment and gains access to the body orally or dermally, milieus.2 In this regard, receptor molecules that can provide
causing cell dysfunction that consequently leads to health optical feedback on binding to the Hg2þ ion in aqueous or
problems.1 The concern over its deleterious effects on human mixed aqueous environments are crucial for developing
health has actually led the sensing community to develop either colorimetric staining agents3 or fluorescent imaging
new mercury detection methods that are cost-effective, rapid, reagents4 for biological applications. Being one of the
heavier metal ions, Hg2þ generally causes an efficient
† quenching of the fluorescence through an efficient spin
Central Salt and Marine Chemicals Research Institute.
‡
Indian Institute of Technology. orbit coupling.5 However, any optical feedback that trans-
(1) (a) Hutchinson, T. C.; Meena, K. M. In Lead, Mercury, Cadmium lates the metal ion-receptor binding phenomena into
and Arsenic in the Environment; JohnWiley: New York, 1987. (b) Harada,
M. Crit. Rev. Toxicol. 1995, 25, 1–24. (c) Tchounwou, P. B.; Ayensu,
a fluorescence turn-on response is preferred over the
W. K.; Ninashvili, N.; Sutton, D. Environ. Toxicol. 2003, 18, 149–175.
(2) (a) Ros-lis, J. V.; Marcos, M. D.; Martinez-Manez, R.; Rurack, (4) (a) Lee, M. H.; Lee, S. W.; Kim, S. H.; Kang, C.; Kim, J. S. Org.
K.; Soto, J. Angew. Chem., Int. Ed. 2005, 44, 4405–4407. (b) Nolan, Lett. 2009, 11, 2101–2104. (b) Suresh, M.; Mandal, A. K.; Saha, S.;
E. M.; Lippard, S. J. Chem. Rev. 2008, 108, 3443–3480. Suresh, E.; Mandoli, A.; Liddo, R. D.; Parnigotto, P. P.; Das, A. Org.
(3) (a) Brummer, O.; La Clair, J. J.; Janda, K. D. Org. Lett. 1999, 1, Lett. 2010, 12, 5406–5409. (c) Dave, N.; Chan, M. Y.; Huang, P. J. J.;
415–418. (b) Li, T.; Li, B.; Wang, E.; Dong, S. Chem. Commun. 2009, Smith, B. D.; Liu., J. J. Am. Chem. Soc. 2010, 132, 12668–12673. (d)
3551–3553. (c) Aragay, G.; Pons, J.; Merkoc-i, A. Chem. Rev. 2011, 111, Kumar, M.; Kumar, N.; Bhalla, V.; Singh, H.; Sharma, P. R.; Kaur, T.
3433–3458. (d) Ma, X.; Wang, J.; Shan, Q.; Tan, Z.; Wei, G.; Wei, D.; Org. Lett. 2011, 13, 1422–1425.
Du, Y. Org. Lett. 2012, 14, 820–823. (e) Saha, S.; Chhatbar, M. U.; (5) (a) McClure, D. S. J. Chem. Phys. 1952, 20, 682–686. (b) Masuhara,
Mahato, P.; Praveen, L.; Siddhanta, A. K.; Das, A. Chem. Commun. H.; Shioyama, H.; Saito, T.; Hamada, K.; Yasoshima, S.; Mataga, N. J.
2012, 48, 1659–1661. Phys. Chem. 1984, 88, 5868–5873.

10.1021/ol3009733 r 2012 American Chemical Society

Published on Web 06/07/2012
fluorescence turn-off response for obvious ease in the were also confirmed from single-crystal X-ray structure
detection process. Among different signaling mechanisms, analysis.
photoinduced electron (PET)/energy transfer (PeT),6 me- Spectroscopic properties of L, L1, and L2 were investi-
talligand charge transfer (MLCT),7 and intramolecular gated in mixed aqueous organic medium (THF/aqueous
charge transfer (ICT)8 are most common. However, to phosphate buffer (6:4, v/v; pH = 7.2). The UVvis spectra
achieve a reversible switch-on fluorescence response on of L exhibited a strong band at 255 nm (ε = 58823 L
selective binding to Hg2þ, two common methodologies mol1 cm1) with two broad band at 295 nm (ε = 22768
have been adopted, namely conversion from the nonfluor- L mol1 cm1) and 385 nm (ε = 16096 L mol1 cm1).
escent lactam to a strongly fluorescent xanthene form or Electronic spectra of L were compared with those of L1 and
through achieving a structural rigidity on Hg2þ ion binding.9 L2. The bands at ∼255 and 295 nm could be attributed to
Literature reports are there for demonstrating the second N- and Q-based intracomponent charge-transfer (CT)
process using azine-based receptors.4b However, such an transitions, respectively, while the broad band at 385 nm
example using analogous imine isomerization as a signaling was ascribed to a intercomponent CT band with N as
mechanism has not been used so far for designing any the donor and Q as the acceptor fragment. Addition of a
colorimetric or fluorogenic receptor for Hg2þ ion, although perchlorate salt of alkali, alkaline earth, and common
it has been documented previously for other metal ions.9a transition-metal ions (Liþ, Naþ, Kþ, Csþ, Agþ, Ca2þ,
Herein, we report a novel imine-based molecule (L), Mg2þ, Sr2þ, Ba2þ, Cr3þ, Fe2þ, Fe3þ, Co2þ, Ni2þ, Cu2þ,
having naphthalene and quinoline as two photoactive Zn2þ, Cd2þ) did not show any change in electronic spectra
units, for the specific recognition of Hg2þ ions through of L. However, an appreciable change was observed when
restricted imine isomerization on binding of L to Hg2þ in the spectrum was recorded in the presence of Hg2þ
mixed aqueousorganic medium. To establish molecular (Supporting Information, SI).
preorganization that takes place prior to the coordination Spectral responses of L, L1, and L2 on binding to Hg2þ in
to the Hg2þ center, two reference compounds (L1 and L2) CHCl3/CH3CN (1:4 v/v) were different, which were pri-
were synthesized (Figure 1), and their optical responses on marily due to the intrinsic difference in the nature of the
binding to Hg2þ were compared with that of L. Also, donor and acceptor fragments in these three receptors.
naphthalene (N) and quinolone (Q) form a FRET pair, Systematic titration in CHCl3/CH3CN (1:4 v/v) medium
and thus, the choice of these two chromophores provides for L revealed the formation of a broader band at longer
the opportunity to achieve a larger Stokes shift on the wavelength (∼480 nm). Coordination to Hg2þ was ex-
sensing event. More importantly, the possibility of using pected to favor the intercomponent CT transition and
this reagent for imaging application and detection of Hg2þ thus the shift to longer wavelength. However, spectral
accumulated in cervical cancer cells was explored to check responses for L in THFaqueous phosphate buffer
the viability of the cells under the experimental conditions. (6:4, v/v; pH = 7.2) were quiet different (Figure 2a and
inset) and were attributed to a weaker coordination of L to
hydrated Hg2þ ion, as compared to the nonaqueous Hg2þ
ion in organic medium. Binding affinity of the reagent L
toward Hg2þ in mixed THFaqueous buffer medium was
evaluated by monitoring changes in absorbance at 385 nm
(Figure 2b) with varying [Hg2þ] (027 molar equiv).

Figure 1. Structure of the imine-based receptor molecules.

Reagent L and the two reference compounds (L1 and L2)

were synthesized following a typical one-step reaction (SI).
Various analytical and spectroscopic data agree well with
the structure proposed for these reagents and the desired
purity. Further, trans conformations for three receptors

(6) (a) de Silva, A. P.; Fox, D. B.; Huxley, A. J. M.; Moody, T. Coord.
Chem. Rev. 2000, 205, 41–57. (b) Gunnlaugsson, T.; Davis, A. P.;
O’Brien, J. E.; Glynn, M. Org. Lett. 2002, 4, 2449–2451. Figure 2. Changes in UVvis spectra recorded in 25 °C (a)
(7) Beer, P. D. Acc. Chem. Res. 1998, 31, 71. CHCl3/CH3CN (1:4 v/v) medium for L (1.8  105 M) with
(8) (a) Xu, Z.; Xiao, Y.; Qian, X.; Cui, J.; Cui, D. Org. Lett. 2005, 7, varying [Hg2þ] (0  1.1  104 M). Spectra in red indicate L.
889. (b) Wang, J. B.; Qian, X. F.; Cui, J. N. J. Org. Chem. 2006, 71, 4308. Inset: titration profile for L (2.3  105 M) with varying [Hg2þ]
(9) (a) Wu, J.; Liu, W.; Ge, J.; Zhang, H.; Wang, P. Chem. Soc. Rev. (0  6.2  104 M) in THF/aqueous phosphate buffer (6:4, v/v;
2011, 40, 3483 and the references cited therein. (b) Saha, S.; Mahato, P.; pH = 7.2) and its corresponding; (b) BH plot for evaluation of
Reddy, U. G.; Suresh, E.; Chakrabarty, A.; Baidya, M.; Ghosh, S. K.;
Das, A. Inorg. Chem. 2012, 51, 336. (c) Mahato, P.; Saha, S.; Suresh, E.;
the binding constant for L with Hg2þ. The good fit of the linear
Liddo, R. D.; Parnigotto, P. P.; Conconi, M. T.; Kesharwani, M. K.; plot (R2 = 0.99) confirmed the 2:1 binding stoichiometry.
Ganguly, B.; Das, A. Inorg. Chem. 2012, 51, 1769.

Org. Lett., Vol. 14, No. 12, 2012 2981

Weaker binding in aqueous buffer medium was also N-moiety-based luminescence in L, as observed for L2.
evident from the much lower affinity constant (KA[Hg2þ]2L However, similarities in the luminescence responses of L
= 1.48  107 M2 L2) than the one that was evaluated in and L1 on binding to Hg2þ revealed that the Q-moiety is
CHCl3/CH3CN (1:4 v/v) medium (KA[Hg2þ]2L = 2.79  109 primarily involved in the photoinduced geometrical iso-
M2 L2) under otherwise identical condition. Good linear merization either through rotation about the CdN or by
fit of the BenesiHildebrand plot (BH plot) confirmed C[Q]NCdN bond rotation.
the 2:1 binding stoichiometry and formation of [Hg2þ]2L The binding affinity of Hg2þ toward L, L1, and L2 was
complex. This binding stoichiometry in solution was also evaluated from a BenesiHildebrand (BH) plot using
confirmed from the ESImass spectral data (SI). Spectral the results of the systematic titrations either by probing the
data also confirmed formation of [Hg2þ]2L1 and [Hg2þ]2L2 change in fluorescence or absorption band intensity with
in solution. Further, the single-crystal structure for varying [Hg2þ]. Analysis of the BH plots for L and L2
[Hg2þ]2L and [Hg2þ]2L2 also corroborated this. No evi- confirmed the 2:1 binding stoichiometry (SI). No spectral
dence for the stepwise formation of 1:1 and then 2:1 evidence for stepwise complex formation was apparent.
complex could be obtained for L and L2, which was evident However, for L1, two distinct changes in fluorescence
from the electronic spectral titration profile for L1 (SI). spectra were observed when spectra were recorded with
Thus, for reagent L, the observed binding constant could varying [Hg2þ] and allowed us to evaluate stepwise binding
be better assigned as the composite binding constant for constants for formation of [Hg2þ]L1 and [Hg2þ]2L1 (SI).
the formation of {L(Hg2þ)2}(ClO4)4. The respective binding constants evaluated from different
Emission quantum yields for L (ΦL = 0.0011) and one methodologies are summarized in Table 1. The reversible
of the two model receptors (ΦL1 = 0.0118) were poor. binding of Hg2þ to L was checked by addition of
However, relative fluorescence quantum yield for L2 an aqueous solution of KI (SI). This was also demon-
(ΦL2 = 0.1090) was much higher than that of L1. Because strated by 1H NMR spectroscopic studies. Because of
of the flexible nature of the imino compound, the thermal limited solubility, the 1H NMR spectra were recorded in a
barrier between the two geometrical isomers is low, so the CDCl3DMSO-d6 (4:1, v/v) medium. On addition of Hg2þ,
relaxation through photoinduced geometrical change is the HHCdN proton showed a drastic downfield shift along
extremely rapid and efficient.10 The photoinduced electron with a decrease of signal intensity for the rest of the protons
transfer (PET) process involving the lone pairs of electrons (SI). On addition of excess of thiourea to this solution, the
of NCdN and the excited Q-based fluorophore could also initial spectrum for L was restored with a subsequent
contribute to the observed weaker emission for L and L1. increase in signal intensity (SI).
Interestingly, the fluorescence enhancement for L2 was Detailed 1H NMR spectroscopic studies were carried
2þ
only 2-fold on binding to Hg2þ (Φ[Hg ]2L2 = 0.2245), while out with an attempt to address the critical issue of bond
2þ
for similar experiment with L (Φ[Hg ]2L = 0.2390) and L1 rotation about the CdN or by C[Q]NCdN bond in L, L1,
2þ
(Φ[Hg ]2L1 = 0.2053) a switch-on fluorescence response was and L2 on binding to Hg2þ. The 1D-NOE spectra of L on
observed (Figure 3). This observed enhancement on lumi- selective irradiation of the HHCdN proton at 8.95 ppm
nescence response for L, L1 and L2 on binding to showed a correlation at 7.35 ppm, which signified a trans
Hg2þcould be ascribed due to the loss of molecular flex- orientation of the two N-atoms of Q-ring and NHCdN with
ibility of these receptors because of restricted imine iso- respect to each other in solution. No such 1D-NOE
merization. In addition, an interrupted PET process on correlation was observed in presence of Hg2þ (SI). Similar
coordination to the Hg2þ was also expected to contribute observations were also witnessed for L1 and L2 (SI).
to this observed switched-on emission response.11 It is A single-crystal X-ray structure for L and L2 revealed
worth mentioning here that the presence of lower lying that quinolyl and imine nitrogen atoms oriented trans to
CT states (vide supra) could account for the absence of any each other (Figure 4a,c), minimizing the protonproton
interactions. Such an orientation is typical for pyridy-
limine ligands.12 Crystal structures for L and [(Hg)2L]I4
(Figures 4a,b) also revealed change in the twist angle
(θQ‑N)13 of 14.6° (42.2° for L to 56.8° for [(Hg)2L]I4)
between the terminal Q and central N ring. This change
in θQN for L2 and [(Hg)2L2]I4 was 14.1° (42.2 for L2 to
56.3 for [(Hg)2L2]I4) (Figures 4c,d). Thus, structural evi-
dence suggest a difference of ∼14° in θQ‑N to make effective

(10) King, N. R.; Whale, E. A.; Davis, F. J.; Gilbert, A.; Mitchell,
G. R. J. Mater. Chem. 1997, 7, 625–630.
(11) (a) Naik, L. R.; Suresh, H. M.; Inamdar, S. R.; Math, N. N.
Spectrosc. Lett. 2005, 38, 645. (b) Koner, A. L.; Mishra, P. P.; Jha, S.;
Figure 3. Fluorescence spectral changes in THFaqueous phos- Dutta, A. J. Photochem. Photobiol. A. 2005, 170, 21–26.
phate buffer medium (6:4, v/v; pH = 7.2) at 25 °C: (a) L (2.0  (12) (a) Kesslen, E. C.; Euler, W. B. Chem. Mater. 1999, 11, 336–340.
105 M) with varying [Hg2þ] (05.2  104 M), λexc = 295 nm; (b) Tuna, F.; Hamblin, J.; Clarkson, G.; Errington, W.; Alcock, N. W.;
Hannon., M. J. Chem.;Eur. J. 2002, 8, 4957–4964.
(b) fluorescence scanning of L (2.0  105 M) in the presence of (13) θQN is the twist angle between the Q and N rings, while
various metal ions (6.0  104 M) (λexc = 295 nm). θ[CdNCNaph] represents the dihedral angle between planes involving
the imine functionality and the naphthalene (N) ring.

2982 Org. Lett., Vol. 14, No. 12, 2012

Table 1. Association Constants for Complex Formation with
Hg2þ with Different Receptors in Different Solvent Mediaa
K2L (105)
K (M2 L2) K1L1 and K2L1d K2L2 (105) (M2 L2)

absb 148 8700 and 1100 M1 L 5.84

absc 27900 1.29  109 M2 L2e 364 Figure 5. Bright field transmission of living HeLa cells incubated
fluob 186 10000 and 1500 M1 L 7.83 with (a) L (20 μM) and (b) L (20 μM) with Hg2þ (10 μM); (c)
2þ 2þ fluorescence image of HeLa cells incubated with L (20 μM) with
a
K2L = K[Hg2þ]2L; K1L1 = K[Hg ]L1; K2L1 = K[Hg ]2L1; K2L2 =
2þ Hg2þ (10 μM); (d) merged image of (b) and (c).
K[Hg ]2L2. b Absorption titration in THFaqueous phosphate buffer
medium (6:4, v/v; pH = 7.2). c Signifies absorption titration in CHCl3/
CH3CN (1:4 v/v) medium. d Stepwise (1:1 and then 2:1 with respect to
Hg2þ and L1) complex formation was observed for reagent L1 only in fluorescence was observed for Hela cells pretreated with
THFaqueous buffer medium. e Composite binding constant was ob- Hg2þ and then with L/L1, while no such fluorescence was
tained in CHCl3:CH3CN (1:4 v/v) medium.
observed eitehr for untreated HeLa Cells or cells treated
with only Hg(ClO4)2 (Figure 5).
So, these two probe reagents (L and L1) were cell
membrane permeable and could detect Hg2þ uptake in
living cells exposed to [Hg2þ] as low as 10 μM. The
reversibility of the binding to L/L1 and the detection of
Hg2þ in HeLa cells was established by treating those
stained cells with KI (20 μM) solution (SI).
In summary, we synthesized an imine-based chemo-
sensor that selectively binds Hg2þ ion in aqueous media.
We exploited two signal transaction mechanisms, namely,
restriction of molecular flexibility and/or PET to achieve
fluorescence on response on binding to Hg2þ ions. Crytallo-
Figure 4. Single crystal structure: (a) L, (b) [(Hg)2L]I4, (c) L2; (d) graphic evidence tends to favor the CC bond rotation,
[(Hg)2L2]I4 (solvent molecule was omitted for clarity). rather than the geometrical isomerization at CdN as one of
the primary reason for the restriction of the molecular
flexibility on binding to Hg2þ. Perhaps for a binucleating
binding to Hg2þ either by L or L2, while θ involving the receptors like these, cistrans isomerization is energetically
N-ring and imine plane (θ[CdNCNaph])13 was 41.8° and more demanding than the CC bond rotation. Results of the
42.4° for L and L2, respectively (SI). Interestingly, a much fluorescence microscopic studies revealed that L and L1 are
smaller change in θ[CdNCNaph] on binding to the Hg(II) cell permeable and could be used as nontoxic imaging
center was observed (θ[CdNCNaph] of 46.01° for [(Hg)2L]I4 reagents for detection of the uptake of Hg2þ in the HeLa cells.
and 45.1° for [(Hg)2L2]I4), which indicated a nominal
perturbation in θ[CdNCNaph] upon coordination to the Acknowledgment. A.D. thanks DST and CSIR (India)
Hg2þ center. Thus, structural evidence tends to indicate for financial support. A.K.M., M.S., and P.D. acknowledge
that C[Q]NCdN bond rotation was more plausible than CSIR for Senior Research Fellowships. M.B. acknowledges
geometrical isomerization around the CdN. UGC (New Delhi) for Jr. Research Fellowship. S.K.G.
We explored the possibility of using L and L1 for the acknowledges DST for partial grant for the Confocal
recognition of the Hg2þ ion in living cervical cancer cells facility.
(HeLa cells) using fluorescence imaging experiments.
MTT assay studies revealed that >80% Hela cells re- Supporting Information Available. Synthetic details,
mained viable even after 12 h of incubation with these characterization data for the compounds, and selected
probe molecules having concentration comparable to that spectroscopic data. Three crystal structures have been
used for imaging studies (SI). For fluorescence imaging deposited with the Cambridge Crystallographic Data
experiments, HeLa cells preexposed to 10 μM Hg2þ were Centre and allocated deposition nos. CCDC 874976
washed with an appropriate buffer solution and subs- 874978. This material is available free of charge via the
quently incubeted either with L or with L1 for 30 min. Internet at http://pubs.acs.org.
On viewing through a fluorescence microscope using a
330380 nm UV filter as an excitation source, a bright blue The authors declare no competing financial interest.

Org. Lett., Vol. 14, No. 12, 2012 2983