A Molecular Imaging Approach to Mercury Sensing Based on Hyperpolarized 129Xe Molecular Clamp Probe

Article abstract of DOI:10.1002/chem.201600193

Mercury pollution, in the form of mercury ions (Hg²⁺), is a major health and environmental hazard. Commonly used sensors are invasive and limited to point measurements. Fluorescence-based sensors do not provide depth resolution needed to image spatial distributions. Herein we report a novel sensor capable of yielding spatial distributions by MRI using hyperpolarized ¹²⁹Xe. A molecular clamp probe was developed consisting of dipyrrolylquinoxaline (DPQ) derivatives and twocryptophane-A cages. The DPQ derivatives act as cation receptors whereas cryptophane-A acts as a suitable host molecule for xenon. When the DPQ moiety interacts with mercury ions, the molecular clamp closes on the ion. Due to overlap of the electron clouds of the two cryptophane-A cages, the shielding effect on the encapsulated Xe becomes important. This leads to an upfield change of the chemical shift of the encapsulated Xe. This sensor exhibits good selectivity and sensitivity toward the mercury ion. This mercury-activated hyperpolarized ¹²⁹Xe-based chemosensor is a new concept method for monitoring Hg²⁺ ion distributions by MRI. In a pinch: A novel sensor is reported that is capable of yielding spatial distributions of mercury ions by MRI, using hyperpolarized ¹²⁹Xe. This clamp probe (see figure) consists of dipyrrolylquinoxaline (DPQ) derivatives and two cryptophane-A cages. The DPQ derivatives act as cation receptors, whereas cryptophane-A acts as a suitable host molecule for xenon. This sensor exhibits good selectivity and sensitivity toward the mercury ions.

Full text of DOI:10.1002/chem.201600193

DOI: 10.1002/chem.201600193
Communication
&
Sensors
A Molecular Imaging Approach to Mercury Sensing Based on
1
29
Hyperpolarized Xe Molecular Clamp Probe
[a]
[a]
[a]
[a]
[a]
[a]
Qianni Guo, Qingbin Zeng, Weiping Jiang, Xiaoxiao Zhang, Qing Luo, Xu Zhang,
[b]
[a]
[a]
Louis-S. Bouchard, Maili Liu, and Xin Zhou*
needed to map the extent of mercury exposure and changes
in its distribution over time. There are several methods to pro-
mote the sensitivity of NMR technology, such as CEST (chemi-
cal exchange saturation transformation) and hyperporlariza-
Abstract: Mercury pollution, in the form of mercury ions
2
+
(
Hg ), is a major health and environmental hazard. Com-
monly used sensors are invasive and limited to point
measurements. Fluorescence-based sensors do not pro-
vide depth resolution needed to image spatial distribu-
tions. Herein we report a novel sensor capable of yielding
[3]
129
tion. Herein we present a mercury-activated Xe molecular-
clamp probe. The mercury sensor is minimally invasive, can be
imaged with depth resolution, and its detection sensitivity as
function of depth exceeds that of fluorescence-based sen-
1
29
spatial distributions by MRI using hyperpolarized Xe. A
molecular clamp probe was developed consisting of dipyr-
rolylquinoxaline (DPQ) derivatives and twocryptophane-A
cages. The DPQ derivatives act as cation receptors where-
as cryptophane-A acts as a suitable host molecule for
xenon. When the DPQ moiety interacts with mercury ions,
the molecular clamp closes on the ion. Due to overlap of
the electron clouds of the two cryptophane-A cages, the
shielding effect on the encapsulated Xe becomes impor-
tant. This leads to an upfield change of the chemical shift
of the encapsulated Xe. This sensor exhibits good selectiv-
ity and sensitivity toward the mercury ion. This mercury-
[4]
sors. In this approach, the xenon nuclear spins are hyperpo-
larized, using the spin-exchange optical pumping (SEOP) tech-
nique, by several orders of magnitude compared to the Boltz-
[5]
mann population. This leads to amplification of the NMR sen-
sitivity by factors of more than 10000 in high field (>1 T), ena-
bling sensitive detection of weak signals at very low
[6]
129
concentrations. When developing Xe-based biosensors, the
main challenge is that Xe atoms are chemically inert, making it
difficult to create Xe-substrate fusions for the localized delivery
129
of Xe signals. Modern hyperpolarized Xe NMR sensing sys-
tems generally invoke two elements. The first is a binding site,
which is affected by the ligand topology, the guest properties
1
29
activated hyperpolarized
Xe-based chemosensor is
2
+
a new concept method for monitoring Hg ion distribu-
tions by MRI.
(
ionic radius, charge, coordination number, hardness, etc.) and
the nature of the solvent. The other element is the molecular-
recognition mechanism. The Xe center is endowed with the ca-
pability of altering its chemical shift upon binding to report on
the nature of the binding process.
Mercury pollution remains a problem of global proportions
perpetuated by the occurrence of natural geological events
and the widespread use of mercury species in human activi-
Our sensor consists of dipyrrolylquinoxaline (DPQ) deriva-
tives and two cryptophane-A cages. The DPQ derivatives act as
[
1]
2+
[7]
ties. A wide range of tools, such as Hg ion chemosensors,
have been developed to detect and monitor local mercury
cation receptors, whereas cryptophane-A acts as a suitable
[
8]
host molecule for xenon. The chemosensor 1 was designed
like a molecular clamp that can clip the target ions in its
clamp’s cave, using DPQ as the basic molecular frame, pyrrole
and the imine as the recognition site, and cryptophane-A as
the Xe NMR signal reporter moiety (Scheme 1).
[
2]
concentrations. However, these methods are invasive and re-
quire close proximity of the sensor to the sample under study,
2
+
making it impossible to obtain spatial distributions of Hg
ions within a system or region of interest. New methods are
[
a] Prof. Dr. Q. Guo, Q. Zeng, W. Jiang, X. Zhang, Prof. Dr. Q. Luo,
Prof. Dr. X. Zhang, Prof. Dr. M. Liu, Prof. Dr. X. Zhou
Key Laboratory of Magnetic Resonance in Biological Systems
State Key Laboratory of Magnetic Resonance and Atomic and
Molecular Physics, Wuhan Center for Magnetic Resonance
Wuhan Institute of Physics and Mathematics
Chinese Academy of Sciences, Wuhan (P.R. China)
E-mail: xinzhou@wipm.ac.cn
[
b] Prof. Dr. L.-S. Bouchard
Department of Chemistry and Biochemistry
Califonia NanoSystem Institute, The Molecular Biology Institute
University of California, Los Angeles, CA 90095 (USA)
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/chem.201600193.
Scheme 1. Molecular structure of chemosensor 1.
Chem. Eur. J. 2016, 22, 3967 – 3970
3967
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Due to the change in electronic density of the chromophore
induced by complexation, when the ionophore moiety is com-
plexed to a metal ion, substantial changes in the absorption
[9]
spectra can be observed. Since the recognition site is now
part of the chromophore, when conjugated with the cation in
solution, the analytical potential of the system is greatly en-
hanced. The binding affinity of the sensor to the metal ion is
controlled by the oxidation state of the nitrogen atoms of the
clamp. As shown in Figure 1, absorptions of chemosensor 1 ap-
peared at 292, 328, and 395 nm, respectively, in solution (v/v
DMSO/H O=1:4). The absorbance intensity decreased gradual-
2
2
+
À6
À5
ly with an increase in Hg concentration (about 10 –10 m,
2
2+
linearly dependent coefficient R =0.9954). With Hg /ligand
mole ratio >1:1, a clear red shift (up to 12 nm) was observed.
À5
À1
Figure 2. UV absorbance intensity of chemosensor 1 (3.0”10 molL ) at
2
+
2+
2+
2+
2+
2
+
lmax =328 nm as function of Mg , Hg , Ca , Zn , and Cd concentra-
The isosbestic point was observed at 466 nm. As Hg solution
was added to the ligand solution, the absorbance intensity
reached a minimum and did not exhibit significant changes
tions.
2
129
beyond. The R value of chemosensor 1 was 0.9954, which was
from d=66.5 to 66.1 ppm (D=38.6 Hz). The Xe NMR spectra
close to 1. The binding constant of the complex evaluated
also successfully distinguished chemosensor 1 and its chelate
5
2+
from the UV/Vis spectra was founded to be 3.57”10 . The
in the presence of Hg ions in a mouse serum solution. With-
2
+
result for chemosensor 1 were consistent with the formation
out Hg
ions, the signal of xenon caged in chemosensor
2
+
of a complex with 1:1 stoichiometry and Hg , similar to other
1 (Xe@1) appeared at d=68.4 ppm in serum. The addition of
[10]
2+
reported DPQ-based chemosensors.
Hg ions resulted an upfield increase of the chemical shift of
xenon caged in chemosensor 1 (Xe@1) from d=68.4 to
6
8.1 ppm (D=30 Hz). The chemical shift change of xenon
caged in chemosensor 1 (Xe@1) showed a relative standard
deviation at 2% (Figure S5 of the Supporting Information).
1
29
Figure 3. The upfield change of the NMR chemical shift of Xe@1 induced
by Hg . The addition of Hg ions causes the chemical shift to decrease by
nearly half a ppm.
À5
À1
2+
2+
Figure 1. UV absorbance spectra of chemosensor 1 (3.0”10 molL ) upon
2
+
addition of Hg . Inset: the absorbance intensity at lmax =328 nm as func-
tion of mercury concentration.
129
The sensitivity threshold of the hyperpolarized Xe NMR
2
+
To explore the utility of chemosensor 1 as an ion-selective
approach for Hg detection was assessed. Following the in-
troduction of 1 atm of hyperpolarized xenon gas in the NMR
2
+
chemosensor for Hg , control experiments were conducted
2
+
2+
2+
2+
2+
with Mg , Ca , Zn , and Cd ions, at various concentra-
tube, a Hg concentration of 1 mm could be readily detected
(Figure S6 of the Supporting Information). The selectivity of
À6
À5
tions (about 10 –10 m). We observed no obvious change in
2
+
the absorbance intensity (Figure 2) when other cations (e.g.,
this approach for Hg detection was also studied. The result
was in agreement with the results reported by UV/Vis. When
2
+
2+
2+
2+
2+
Mg , Ca , Zn , and Cd ) were used instead of Hg . Since
2
+
2+
2+
2+
2+
2+
2+
2+
2+
the sizes of Mg , Ca , Zn , and Cd are different, this sug-
gests that the size of mercury(II) fits well within the molecule
using Mg , Ca , Zn , and Cd instead of Hg , no obvious
change in the signal of xenon caged in chemosensor 1 (Xe@1)
could be observed. Compared with other hyperpolarized Xe-
2
+
129
clamp’s cave, indicating a prominent selectivity towards Hg
.
1
29
Xe NMR spectra successfully distinguished chemosensor
based sensors of metal ions, the downfield change of chemical
shift caused by electron-withdrawal effect of metal ions (e.g.,
2
+
1
and its chelate in the presence of Hg ions (Figure 3). In the
2
+
2+
absence of Hg ions, the signal of xenon caged in chemosen-
the chemical shift signal of xenon caged in the reported Zn
2
+
[11]
sor1 (Xe@1) appeared at d=66.5 ppm. The addition of Hg
ions resulted an upfield increase of the chemical shift of Xe@1
sensor changed from d=65.75 to 67.2 ppm),
the upfield
change of chemical shift of Xe@1 was totally different. Since
Chem. Eur. J. 2016, 22, 3967 – 3970
www.chemeurj.org
3968 ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
design of this probe is totally different from our previous de-
[
12]
signed probe, the change of chemical shift of Xe@1 is influ-
enced not only by the electron-withdrawing effect of the
metal ions, but also by the shielding effect of the two crypto-
phane-A cages. These results not only demonstrate that the
2
+
Hg ions were indeed chelated with the DPQ moiety, but also
suggest that the distance between the two cryptophane-A
cages of chemosensor 1 is reduced after binding. When bind-
2
+
ing with Hg
the molecular structure of chemosensor
1
turned from open to closed (Scheme 1). As the electron
clouds of the two cryptophane-A cages overlap the shielding
effect towards encapsulated Xe becomes important. The influ-
ence of shielding effects from electronic clouds of the two
cryptophane-A cages overlapping was stronger than the de-
shielding effect of the electron-withdrawal effect of metal ions
on the chemical shift of Xe@1. An upfield change of chemical
shift of Xe@1 was induced.
1
1
Figure 4. 2D H- H NOESY spectra of the chemosensor 1 in the: a) absence,
2+
and b) presence of Hg ions. The experiments were performed at 208C on
a Bruker Avance III 600 NMR instrument equipped with a triple resonance
clamp-probe. The mixing time for magnetization transfer by cross-relaxation
was 500 ms. Solvent suppression was accomplished using WATERGATE
before the acquisition to avoid signal loss from the exchangeable protons.
atoms that likely offset the electron-withdrawing effect of the
2
+
Hg
The change in chemical shift between the signals of encap-
.
2
+
sulated xenon in the absence and in the presence of Hg ions
relative to the NMR linewidths was sufficient to enable spectro-
scopic MRI of these chemical species. Figure 5 images were ob-
tained using a CSI method. Comparison between images a and
2
+
b in Figure 5 shows that the Hg ions can be specifically de-
tected and localized at low concentration in a short time.
These results demonstrate that highly sensitive and selective
2
+
detection of Hg can be achieved with a simple molecular
clamp chemosensor. Detection by MRI (Figure 5) and optical
(
Figure 1) methods is possible. The MRI method enables as-
sessment of spatial distributions of mercury. This mercury-acti-
Scheme 2. Conformational change of chemosensor 1 induced by its interac-
129
2
+
vated hyperpolarized Xe-based chemosensor is a new con-
tion with Hg . The blue bubbles represent electronic cloud of crypto-
phane-A. The red bubbles represent the caged Xe and the purple bubbles
represent the free Xe.
2
+
cept method for monitoring Hg ions.
To determine the binding mechanism and molecular struc-
ture changes of chemosensor 1 before and after interaction
2
+
with Hg more precisely, a NOESY experiment was performed
(
(
Figure 4). It can be seen that two strong signals from brNH
broad peak of NH at 9.945 and 10.360 ppm) of the com-
2
+
pounds vanish in the presence of the Hg . Since deprotona-
tion is usually clear evidence of the formation of a conjugate
bond, this result indicated that the two nitrogen atoms of
2
+
brNH chelated with the Hg directly. The chelation was also
confirmed by the chemical shift change of the quinoxaline CH.
The chemical shift of the quinoxaline CH proton resonance is
2
+
2+
7
.198 ppm without Hg . With the addition of Hg the reso-
nance is downfield shifted to 0.260 ppm. This indicated that
the quinoxaline CH was deshielded, which appeared to be due
to the deprotonation of brNH nitrogen atoms, and the direct
bonding to the brNH nitrogen atoms directly. The chemical
shift of the aromatic hydrogen atoms changed slightly, indicat-
ing that the cryptophane-A cages were far away from the
binding sites. Also, the molecular structure change of chemo-
sensor 1 caused a shielding effect of the aromatic hydrogen
Figure 5. Chemical shift image (CSI MRI) of chemosensor 1: a) with, and
2
+
b) without the addition of Hg
.
Chem. Eur. J. 2016, 22, 3967 – 3970
www.chemeurj.org
3969 ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
B. M. Goodson, E. Y. Chekmenev, Anal. Chem. 2014, 86, 8206; c) P. Niko-
laou, A. M. Coffey, L. L. Walkup, B. M. Guest, N. Whiting, H. Newton, S.
Barcus, I. Muradyan, M. Dabaghyan, G. D. Moroz, M. S. Rosen, S. Patz,
M. J. Barlow, E. Y. Chekmenev, B. M. Goodson, Proc. Natl. Acad. Sci. USA
2013, 110, 14150–14155.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (81227902, 21302217, and 21475147).
[7] L. Wang, X. Zhu, W. Wong, J.Guo, W. Wong, Z. Li, Dalton Trans. 2005,
3
235–3240.
1
29
Keywords: hyperporlarized Xe · mercury · molecular clamp
probe · molecular imaging · sensors
[8] a) J. M. Chambers, P. A. Hill, J. A. Aaron, Z. Han, D. W. Christianson, N. N.
Kuzma, I. J. Dmochowski, J. Am. Chem. Soc. 2009, 131, 563–569; b) P.
Berthault, G. Huber, H. Desvaux, Prog. Nucl. Magn. Reson. Spectrosc.
2
009, 55, 35–60; c) E. Dubost, J. P. Dognon, B. Rousseau, G. Milanole, C.
[
1] a) M. Nendza, T. Herbst, C. Kussatz, A. Gies, Chemosphere 1997, 35,
875; b) D. W. Boening, Chemosphere 2000, 40, 1335; c) S. Ando, K.
Koide, J. Am. Chem. Soc. 2011, 133, 2556–2566.
2] a) E. M. Nolan, S. J. Lippard, Chem. Rev. 2008, 108, 3443; b) D. G. Cho,
J. L. Sessler, Chem. Soc. Rev. 2009, 38, 1647; c) D. T. Quang, J. S. Kim,
Chem. Rev. 2010, 110, 6280; d) Q. Liu, J. Peng, L. Sun, F. Li, ACS Nano
Dugave, Y. Boulard, E. LØonce, C. Boutin, P. Berthault, Angew. Chem. Int.
Ed. 2014, 53, 9837–9840; Angew. Chem. 2014, 126, 9995–9998; d) S.
Klippel, J. Dçpfert, J. Jayapaul, M. Kunth, F. Rossella, M. Schnurr, C.
Witte, C. Freund, L. Schrçder, Angew. Chem. Int. Ed. 2014, 53, 493–496;
Angew. Chem. 2014, 126, 503–506; e) C. Witte, V. Martos, H. M. Rose, S.
Reinke, S. Klippel, L. Schrçder, C. P. R. Hackenberger, Angew. Chem. Int.
Ed. 2015, 54, 2806–2810; Angew. Chem. 2015, 127, 2848–2852.
1
[
2011, 5, 8040–8048.
[
3] a) E. Y. Chekmenev, K. W. Waddell, J. Hu, Z. Gan, R. J. Wittebort, T. A.
Cross, J. Am. Chem. Soc. 2006, 128, 9849–9855; b) Y. Feng, T. Theis, X.
Liang, Q. Wang, P. Zhou, W. S. Warren, J. Am. Chem. Soc. 2013, 135,
[
9] a) H. Sakamoto, J. Ishikawa, S. Nakao, H. Wada, Chem. Commun. 2000,
2395–2396; b) H. Ji, R. Dabestani, G. M. Brown, R. L. Hettich, Photochem.
Photobiol. 1999, 69, 513–516; c) K. Wen, S. Yu, Z. Huang, L. Chen, M.
Xiao, X. Yu, L. Pu, J. Am. Chem. Soc. 2015, 137, 4517–4524.
9
632–9635; c) R. Trokowski, J. Ren, F. K. Kµlmµn, D. Sherry, Angew.
Chem. Int. Ed. 2005, 44, 6920–6923; Angew. Chem. 2005, 117, 7080–
083; d) L. Schrçder, T. J. Lowery, C. Hillty, D. E. Wemmer, A. Pines, Sci-
[
[
10] a) Y. Hu, Q. Li, H. Li, Q. Guo, Y. Lu, Z. Li, Dalton Trans. 2010, 39, 11344–
1352; b) Q. Guo, C. Zhong, Y. Lu, C. Shi, Z. Li, J. Inclusion Phenom. Mac-
7
1
ence 2006, 314, 446–449.
4] a) N. Dave, M. Y. Chan, P. Huang, B. D. Smith, J. Liu, J. Am. Chem. Soc.
rocyclic Chem. 2012, 72, 79–88.
[
[
11] N. Kotera, N. Tassali, E. LØonce, P. Berthault, T. Brotin, J. P. Dutasta, L. De-
lacour, T. TraorØ, D. A. Buisson, F. Taran, S. Coudert, B. Rousseau, Angew.
Chem. Int. Ed. 2012, 51, 4100–4103; Angew. Chem. 2012, 124, 4176–
2010, 132, 12668–12673; b) J. Chan, S. C. Dodani, C. J. Chang, Nat.
Chem. 2012, 4, 973–984.
5] a) M. M. Spence, S. M. Rubin, A. Pines, S. Yao, F. Tian, P. G. Schultz, Proc.
Natl. Acad. Sci. USA 2001, 98, 10654–10657; b) W. Happer, E. Miron, S.
Schaefer, D. Schreiber, W. A. van Wijngaarden, X. Zeng, Phys. Rev. A
4
179.
12] J. Zhang, W. Jiang, Q. Luo, X. Zhang, Q. Guo, M. Liu, X. Zhou, Talanta
014, 122, 101–105.
[
2
1984, 29, 3092–3110; c) X. Zeng, Z. Wu, T. Call, E. Miron, D. Schreiber,
W. Happer, Phys. Rev. A 1985, 31, 260–278; d) T. G. Walker, W. Happer,
Rev. Mod. Phys. 1997, 69, 629–642; e) X. Zeng, C. Wu, M. Zhao, Chem.
Phys. Lett. 1991, 182, 538–540; f) X. Zhou, X. Sun, J. Luo, X. Zeng, M.
Liu, M. Zhan, Chin. Phys. Lett. 2004, 21, 1501–1503.
[
6] a) A. Abragam, Principles of Nuclear Magnetism, Oxford University Press,
New York, 1961; b) P. Nikolaou, A. M. Coffey, M. J. Barlow, M. Rosen,
Received: January 15, 2016
Published online on February 2, 2016
Chem. Eur. J. 2016, 22, 3967 – 3970
www.chemeurj.org
3970
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim