Responsive MR-imaging probes for N-methyl-d-aspartate receptors and direct visualisation of the cell-surface receptors by optical microscopy

Article abstract of DOI:10.1039/c3sc50903f

A series of N-methyl-d-aspartate (NMDA) receptor-targeted MRI contrast agents has been developed, based on the known competitive NMDA antagonist, 3,4-diamino-3-cyclobutene-1,2-dione. Their use as responsive MR imaging probes has been evaluated in vitro an

Full text of DOI:10.1039/c3sc50903f

Chemical Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Responsive MR-imaging probes for N-methyl-D-
aspartate receptors and direct visualisation of the cell-
Cite this: Chem. Sci., 2013, 4, 3148
surface receptors by optical microscopy†
a
b
a
b
a
*
¨
Neil Sim, Sven Gottschalk,‡ Robert Pal, Jorn Engelmann, David Parker
a
*
and Anurag Mishra‡
A series of N-methyl-D-aspartate (NMDA) receptor-targeted MRI contrast agents has been developed,
based on the known competitive NMDA antagonist, 3,4-diamino-3-cyclobutene-1,2-dione. Their use as
responsive MR imaging probes has been evaluated in vitro and two contrast agents showed 170–176%
enhancements in relaxation rate, following incubation with a neuronal cell line model. A derivative of
the lead compound was prepared containing a biotin moiety, and both the specificity and reversibility
of binding to the NMDA cell surface receptors demonstrated using confocal microscopy.
Received 4th April 2013
Accepted 5th June 2013
DOI: 10.1039/c3sc50903f
www.rsc.org/chemicalscience
NMDAR antagonists have been considered as potential thera-
peutic and diagnostic agents.7,8
Introduction
Glutamate receptor proteins are classed into two main fami-
lies, the ligand-gated ionotropic receptors (iGluRs) and the
G-protein-coupled metabotropic receptors (mGluRs); each of
these is further branched into three subclasses. The NMDA
receptor (NMDAR) from the iGluR family has received growing
interest over the past decade owing to its role in excitatory
neurotransmission, synaptic plasticity, memory and learning.1
Following activation of the NMDAR by glutamate and co-agonist
glycine, post-synaptic depolarization occurs, leading to the
removal of a Mg2+ channel block, allowing the inux of Na+ and
Ca2+ ions into the postsynaptic cell. This overall process leads
to glutamate signal transduction from a pre- to a post-synaptic
cell.
In recent years, several central nervous system disorders
have been associated with misregulation and overstimulation of
the NMDAR, such as ischemia,2 epilepsy3 and pain amplica-
tion.4 Various neurodegenerative disorders have also been
linked to defective NMDA pathways, such as Parkinson's and
Alzheimer's diseases and even neuropsychiatric disorders, such
as schizophrenia.5,6 In seeking to combat these effects, selective
Magnetic Resonance Imaging (MRI) is regarded as one of the
most powerful, non-invasive diagnostic imaging techniques used
in clinical and biomedical research. The sensitivity and speci-
city, and hence contrast of MR images, can be further enhanced
by the use of responsive contrast agents. Notwithstanding
numerous in vivo techniques available to neuroscientists, our
current understanding of the dynamic changes and molecular
basis of neural activity is far from complete. Therefore, by
employing a targeted, responsive contrast agent that can report
changes in neural activity non-invasively, MRI could provide new
vital information on these important homeostatic processes.
One suggested way is through the use of selective glutamate-
receptor contrast agents, which can bind selectively to the
NMDAR and with sufficient affinity to act as a marker of receptor
density. For such a system to be responsive, stimulated release of
glutamate from the pre-synaptic cell, which takes place over a
duration of milliseconds, should displace the contrast agent
from the NMDAR, whilst restoration of the equilibrium state is
believed to occur over a period of about a second.9
Caravan has proposed10 that in order to observe a 10%
increase in the observed water proton relaxation rate, R1, by using
a contrast agent with relaxivity of 5 mMꢀ1 sꢀ1, a 10 mM local
concentration of the Gd-agent is needed. This corresponds to 107
Gd-chelates per cell, suggesting that this is the required NMDAR
density needed to observe a change in contrast. At present, there
aDepartment of Chemistry, Durham University, South Road, Durham, DH1 3LE,
England. E-mail: anurag.mishra@durham.ac.uk; david.parker@durham.ac.uk; Fax:
+44-191-3342051
bHigh Field MR Centre, Max Planck Institute for Biological Cybernetics,
is no typical value reported for the NMDAR density; values have
been reported to be as high as 6.4 pm mgꢀ1
. However, Sherry
11
Spemannstrasse 41, Tuebingen, 72076, Germany
et al. have reported12 that even if the bulk receptor density is low,
the receptors can still be imaged successfully at less than 10 mM
local concentrations, if the receptors cluster together and form
micro-domains of high local concentration. Furthermore, the
rate of dissociation of the contrast agent from the receptor may
† Electronic supplementary information (ESI) available (37 pages): Synthetic
procedures, microscopy and spectral imaging instrumentation and operating
procedures, cell culture, toxicity and MR experiments. See DOI:
10.1039/c3sc50903f
‡ Present address: Institute for Biological and Medical Imaging, Helmholtz
Center Munich, Neuherberg, Germany.
3148 | Chem. Sci., 2013, 4, 3148–3153
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
be slow. If the affinity of the targeted contrast agent is of the
order of log Ka ¼ 7.0, then if the rate of association is 106–107 Hz,
the dissociation rate will be of the order of 0.1–1 Hz. Therefore,
the timescale of the imaging experiment is likely to be the same
order of magnitude as the lifetime of the receptor bound
complex, potentially increasing the intrinsic sensitivity of the
experiment. Assuming that local concentrations of the contrast
agent do not change within this timescale, it should be possible
to report these changes by modulation of the water proton
relaxivity, reecting the relative amounts of receptor bound and
unbound contrast agent.
Here, we report the design, synthesis and in vitro evaluation
of a series of NMDA receptor-targeted contrast agents. The
contrast agents are based on well-known competitive NMDAR
antagonists, appended to an N-linked ‘Gd-DOTA’ core that
possesses a fast-exchanging water molecule. The proposed
contrast agents focus on modulation of the rotational correla-
tion time, sR. In the low to medium eld range (3–7 T), this term
dominates the water proton relaxivity for a low molecular
weight complex, when water exchange at the metal centre is
relatively fast.13 The contrast agents have been designed so that
in the bound state, molecular tumbling is slower than in the
unbound state. This increases the sR value, giving a signicant
relaxivity change, represented as a modulation of R1 (R1 ¼ 1/T1).
Scheme 1 Synthesis of NMDA-targeted contrast agents, [Gd.L1–4].
longitudinal proton relaxivity of [Gd.L1–4] was measured in
aqueous solution (pH 7.4) at 37 ꢁC and 1.4 T. The relaxivities for
[Gd.L1–4] were 5.17, 5.30, 4.68 and 4.80 mMꢀ1 sꢀ1, respectively.
Such values lie in the expected range for mono-aqua gadolinium
species of this molecular weight.17
Results and discussion
Ligand design and complex synthesis
Several competitive NMDAR antagonists have been reported
that contain the a-amino carboxylic acid and phosphonic acid
functionalities, separated by a carbon spacer unit.7,14 The 3,4-
diamino-3-cyclobutene-1,2-dione moiety acts as an isostere for
the a-amino carboxylic acid functionality, and has been incor-
porated into functional NMDA antagonists.14 The synthesis of
the NMDA receptor-targeted contrast agents is detailed in
Scheme 1. Monoamination followed by BOC protection of 3,4-
diethoxy-3-cyclobutene-1-2-dione gave the intermediate 1. A
Michael-type addition–elimination reaction using commercially
available glycine/b-alanine tert-butyl ester, diethyl(2-amino-
ethyl)phosphonate, or diethyl(3-aminopropyl)phosphonate15
yielded compounds 2, 5, 8 and 11, respectively. Subsequent N-
alkylation with methyl bromoacetate in the presence of base
gave the fully protected antagonist entities, 3, 6, 9 and 12.
Finally, selective ester hydrolysis, under ambient basic condi-
tions, gave compounds 4, 7, 10 and 13 in quantitative yield.
The antagonist fragments were coupled through the
carboxylic acid to the macrocyclic amine, 14, (ESI†) using
standard amide-coupling techniques, (EDC/HOBt/NMM) in
anhydrous DMF, to give the fully protected ligands, which were
hydrolyzed under acidic conditions. Complexation was ach-
ieved using GdCl3$6H2O in water at pH 6 and 60 ꢁC, to give the
complexes [Gd.L1–4].
Human Serum Albumin (HSA) is a globular protein consti-
tuting around 4.5% of plasma. As it is the major protein
constituent in the circulatory system, it is likely that the contrast
agent will bind to the protein, disrupting its rotational
dynamics leading to an enhancement of the longitudinal
relaxation rate of the water protons. Therefore, as a control
experiment, the effect of HSA on the measured relaxivity of
[Gd.L1–4] (each at 1 mM) was assessed at 1.4 T in the presence of
up to 1.6 mM HSA. This led to an increase in r1p of 54%, 193%,
39% and 66% for [Gd.L1–4], respectively (ESI†). Association
constants were estimated assuming a 1 : 1 stoichiometry of
interaction, and were found to be log Ka ¼ 3.50(ꢂ0.03) and log
Ka ¼ 3.90(ꢂ0.02) for the carboxylate analogues [Gd.L1,2], and log
Ka
¼ 2.50(ꢂ0.03) and 3.05(ꢂ0.02) for the phosphonate
analogues [Gd.L3,4] (Table 1).
Table 1 Binding affinities to serum albumin,a,b,c and relaxation properties of the
complexes reported herein (310 K, pH 7.4, 1.4 T)
[Gd.L1] [Gd.L2] [Gd.L3] [Gd.L4]
log Ka
3.50a
5.17
3.90b
5.30
11.44
18
2.50a
4.68
5.57
14
3.05b
4.80
6.70
11
ri1npitial/mMꢀ1 sꢀ1
r1p at 0.7 mM HSA/mMꢀ1 sꢀ1 7.62
rl1ipmit/mMꢀ1 sꢀ1
11
a
b
c
Error is ꢂ0.03 mMꢀ1 sꢀ1
.
Error is ꢂ0.02 mMꢀ1 sꢀ1
.
The reported
Relaxivity properties
relative affinities for the NMDA binding moiety in L1–4 are 2.3, 1.6,
0.47 and 2.6 mM respectively; these were given as IC50 values, i.e. they
are not the 1 : 1 binding constants.14
The total gadolinium concentration of [Gd.L1–4] was measured
using Evans' bulk magnetic susceptibility measurements.16 The
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci., 2013, 4, 3148–3153 | 3149

View Article Online
Chemical Science
Edge Article
By comparing each pair (carboxylic acid vs. phosphonic
acid), it is apparent that increasing the chain length of the
pendant acidic arm promotes the binding interaction between
the complex and the protein. In the case of [Gd.L2], a much
larger percentage increase in r1p was observed compared to any
other complex. Such behaviour is consistent with a stronger
complex–protein interaction, modulating sR more signicantly.
Probe-receptor binding studies
A neuronal cell line model expressing functional NMDA recep-
tors was established using the NSC-34 cell line. This hybrid cell-
line is produced by the fusion of mouse spinal cord and
neuroblastoma cells and has been used previously for the
evaluation of NMDAR antagonists.18
The expression of NMDAR on differentiated NSC-34 cells was
demonstrated using immunouorescence techniques with
different primary antibodies; one for each of the two subunits
NMDAR-1 and NMDAR-2B of the receptor, thus demonstrating
the expression of functional receptors (ESI†).
Fig. 1 (Top) Representative T1-weighted MR-images of 1 ꢃ 107 untreated
differentiated cells (control) and differentiated cells treated for 45 min with
200 mM of [Gd.DOTA] or [Gd.L1–4]. Images were recorded using a turbo spin echo
sequence with a matrix of 256 ꢃ 256 voxels over a field of view of 110 ꢃ
110 mm2, slice thickness of 1 mm (resulting in a voxel size of 0.4 ꢃ 0.4 ꢃ
1.0 mm3), TR 1000 ms, TE 13 ms, Ti 23 ms and 20 averages; (bottom) cellular 1H MR
relaxation rates R1,cell in cell suspensions (3 T, 298 K) after treatment of differ-
Cellular labelling of differentiated NSC-34 cells with [Gd.L1–4
]
was assessed by measuring the longitudinal relaxation times, T1,
of the cells on a 3 T Siemens human whole body MR scanner
equipped with a head coil, allowing the calculation of cellular
relaxation rates, R1,cell. Differentiated cells were incubated (37 ꢁC,
5% CO2) for 45 min with 200 mM [Gd.L1–4], washed with Hank's
Buffered Saline Solution (HBSS) to remove any unbound
complex, re-suspended in fresh buffer and T1-weighted MR
images were acquired (Fig. 1 and ESI†).
entiated NSC-34 cells with 200 mM [Gd.DOTA] or [Gd.L1–4
] for 45 min.
[Gd.DOTA] served as a negative control. Values are mean ꢂ SEM (n ¼ 3–6). ns: not
significant, ***P < 0.001 ANOVA with Bonferroni's multiple comparison post-test.
###P < 0.001 Student's t-test. Both tests vs. untreated control using Graphpad
Prism 5.02.
The complexes [Gd.L1–4] each contain the 3,4-diamino-3-
cyclobutene-1,2-dione moiety, that serves as an isosteric
replacement of the a-amino carboxylic acid functionality known
to be essential for many of the reported competitive NMDAR
antagonists.7 The complexes [Gd.L1,2] incorporate a carboxylic
acid residue of variable chain length (1 or 2) appended to one of
the a-amines. Although [Gd.L1] exhibits an increase in R1,cell
(118% of control), the homologue, [Gd.L2], with increased
pendant chain length (n ¼ 2) displays a statistically higher
increase in R1,cell (170% of control). This result is in keeping
with the trend in affinity constants for the antagonist moieties.14
It was hypothesised that the phosphonic acid derivatives,
Scheme 2 Possible rotameric forms of [Gd.L3] due to intramolecular hydrogen
bonding.
[Gd.L3,4] would bind more strongly to the NMDAR, leading to a much. It gave rise to the largest observed increase in R1,cell
larger increase in R1,cell. Interestingly, [Gd.L3] showed no (176% of control).
apparent receptor specic binding, despite the antagonist
It is thought that for an MRI experiment of this nature to be
portion being reported to be particularly potent.14 We have viable at least 3 ꢃ 107 bound Gd-complexes per cell are
previously shown19 that appending the sterically demanding required.21 By using Gd-ICP-MS (data from Fig. 1), the average
macrocyclic core to a known antagonist can alter the antago- number of [Gd.L4] complexes per cell was determined, that
nists' affinity for a receptor, and this could be an explanation of resulted in the measured change in R1,cell. Using this technique,
the lack of binding for [Gd.L3]. However, a more likely argument 9.3 ꢃ 108 Gd3+ ions per cell were estimated, revealing an esti-
is that rotameric forms of [Gd.L3] exist, arising from rotation mated local Gd3+ concentration of 370 mM. This value is beyond
around the C–N bond of the squaramide moiety in which the theoretical detection limit, and explains the large
intramolecular H-bonding gives rise to relatively stable 8/9-ring enhancement in cellular relaxation rate observed.
chelates. This intramolecular H-bonding interaction may
weaken the binding of the antagonist towards the NMDAR was studied by assessing the metabolic activity of differentiated
The cytotoxicity of the four gadolinium complexes [Gd.L1–4
]
(Scheme 2).20
NSC-34 cells using an MTT assay.22 None of the complexes
In contrast, [Gd.L4] with a longer 3 carbon spacer, adopts a exhibited any cytotoxic effects at concentrations of up to 200
geometry in which binding to the NMDAR is not perturbed so mM, following a 24 hour incubation (ESI†).
3150 | Chem. Sci., 2013, 4, 3148–3153
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
ꢁ
measured at pH 7.4, 37 C and 60 MHz (1.4 T). The increased
relaxivity is associated with the larger molecular volume of the
complex, giving rise to a slower rotational correlation time. Also,
an increased second sphere hydration may occur with this
complex, through multiple hydrogen bonding interactions.30
In order to study the behaviour and localisation of [Gd.L5],
live-cell laser scanning confocal microscopy (LSCM) imaging
studies were carried out. Differentiated NSC-34 cells were grown
on a specially designed microscope slide comprising of a ow-
through channel (100 mL) and subjected to various conditions.
Simultaneous treatment with a solution of [Gd.L5] (10 mM) and
AvidinAlexaFluorꢀ 488 conjugate (2.5 mM, 10 min) showed a
localization prole, resembling pit-like localization at the cell
surface (Fig. 2A). CellMaskꢁ Orange is a commercially available
dye for non-specic staining of the plasma membrane. A repeat
of the simultaneous loading experiment of [Gd.L5] and Avidi-
nAlexaFluorꢀ 488 conjugate (10 min), co-incubated with a
CellMaskꢁ Orange (5 min, 5 mg mLꢀ1), showed unequivocal
evidence for the selective localisation of [Gd.L5] at the cell
membrane (Fig. 2B/C).
Cell-surface immobilisation and visualisation of NMDA
receptors
The variation of R1,cell in the presence of [Gd.L2] and [Gd.L4] is
consistent with an increase in sR that is associated with slower
molecular tumbling of the complex when bound to the cell
surface receptors. However, internalisation of the complex via
receptor-mediated endocytosis or non-specic binding to the
cell membrane are also possible mechanisms that could lead to
an increase in R1,cell. In order to establish which of these
possibilities is responsible for the observed increase in R1,cell, a
new complex was designed that allows direct visualisation of
the contrast agent on the cell surface.
Appending a uorescent label to a non-competitive antago-
nist has recently been shown to allow for the direct visualisation
of the NMDAR-2B subunit.23 However, we24–27 and Barton and
Puckett28,29 have shown that some uorescent labels perturb
cell-uptake mechanisms and may promote internalisation of
the complex. Therefore, [Gd.L5] was designed, comprising the
antagonist binding portion of the most promising contrast
agent, [Gd.L4] (largest increase in R1,cell), in which a trans-
substituted biotin moiety is appended to the macrocyclic core.
It was envisaged that if cell-surface receptor binding is
responsible for the increase in R1,cell for [Gd.L2/4], the biotin
moiety of [Gd.L5], aer binding to an added AvidinAlexaFluorꢀ
488 conjugate, would act as a tag on the outside of the cell,
allowing direct visualisation of the complex only when it is
bound to the cell surface.
The synthesis of [Gd.L5] (Scheme 3) involves standard
protection and deprotection steps (ESI†). Following modica-
tion of the dicarboxymethyl cyclen precursor, the benzyl group
of 15 was selectively removed to allow conjugation of the pre-
formed biotin ethylenediamine. Removal of the Cbz group from
the N-terminus of the pendant lysine arm on 16 allowed for
coupling of the antagonist moiety to give 17, as described in the
synthesis of [Gd.L4]. Finally, aer selective basic and acidic
hydrolysis of the protecting groups, gadolinium was introduced
to give the complex [Gd.L5], with a relaxivity of 7.24 mMꢀ1 sꢀ1
,
Fig. 2 Live cell LSCM images of differentiated NSC-34 cells following treatment
with [Gd.L5]. (A) Simultaneous loading of [Gd.L5] (10 mM), and AvidinAlexa-
Fluorꢂ 488 conjugate (2.5 mM, 10 min) allowing visualisation of AvidinAlexa-
Fluorꢂ 488 conjugate lex/lem ¼ 488/505–555 nm; (B) As (A) but with a 5 minute
CellMaskꢃ Orange incubation allowing visualisation of CellMaskꢃ Orange lex
/
lem ¼ 543/550–660 nm; (C) RGB merge showing co-localisation of the Avidi-
nAlexaFluorꢂ 488 conjugate and CellMaskꢃ Orange on the cell surface (P ¼
0.88); (D) As image (A); (E) As image (A) but with a post-glutamate (1 mM) wash
showing that [Gd.L5] is removed from cell surface and; (F) NSC-34 cells are treated
with glutamate (1 mM) and then simultaneous loading of [Gd.L5] (10 mM), and
AvidinAlexaFluorꢂ 488 conjugate (2.5 mM, 10 min) allows recovery of 35% of the
fluorescence signal intensity as compared to image (A).
Scheme 3 Synthesis of the bifunctional MR imaging probe, [Gd.L5], in which a
biotin moiety is included, spaced from the NMDA-receptor binding sub-unit.
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci., 2013, 4, 3148–3153 | 3151

View Article Online
Chemical Science
Edge Article
Conrmation that the observed uorescence was due to the mediated targeting approach. The most promising contrast
strong (log Ka > 15) biotin–avidin interaction between [Gd.L5] and agent, [Gd.L4], showed a 176%, increase in relaxation rate in the
AvidinAlexaFluorꢀ 488 conjugate, was demonstrated in two presence of functional NMDARs on differentiated NSC-34 cells.
control experiments. Differentiated, untreated NSC-34 cells (ESI Cell-surface localisation was demonstrated for the derivative,
Fig. S3†) and differentiated cells loaded with AvidinAlexaFluorꢀ [Gd.L5], providing evidence for specicity, as no binding
488 conjugate only (ESI Fig. S3†), showed no observable uores- was observed upon incubation with an NMDA receptor negative
cence signal in the visible region of the spectrum, when using the cell-line.
previously established experimental parameters. Cell surface
The reversibility of the binding of the contrast agent to the
localisation was found to be independent of the loading receptors was also demonstrated. Addition of glutamate to the
concentration (up to 100 mM) and time (up to 45 minutes) and did receptor-bound [Gd.L5] led to displacement of the complex from
not vary for stepwise vs. simultaneous incubations of the donor the receptors, resulting in a diminution of signal intensity,
and acceptor (data not shown). No evidence for any intracellular which was partially recovered upon subsequent incubation with
staining was observed throughout these experiments; further [Gd.L5] and the AvidinAlexaFluorꢀ 488 conjugate.
evidence of cell-surface localisation has been derived from a 3D
This work suggests that the gadolinium complexes of L2 and
reconstruction (ESI: video†) of the LSCM images taken following L4 are promising MR contrast agent candidates for reporting or
simultaneous incubation of [Gd.L5] (10 mM) and the AvidinAlex- monitoring NMDAR density and have the potential to report on
aFluoro 488 conjugate (2.5 mM, 10 min).
synaptic glutamate activity. It is appreciated that for use in vivo,
direct intra-cranial injection or articially altering the perme-
ability of the blood–brain barrier (BBB) are required for such
anionic complexes to enter the brain. Therefore, future work
will attempt to seek to extend these encouraging in vitro results,
by following changes in the MR signal intensity in an appro-
priate animal model. In parallel, more potent antagonist-
conjugate systems will be examined, in which the binding
constants of the antagonist are even higher.
Establishing the specicity and reversibility of binding
The two key characteristics needed for a probe of this nature are
specicity and reversibility. Evidence for the rst of these was
obtained following simultaneous incubation of an NMDA
receptor-negative cell line, NIH-3T3, with [Gd.L5] (10 mM) and
AvidinAlexaFluorꢀ 488 conjugate (2.5 mM, 10 min). No local-
isation of any kind (cell surface or intracellular) was observed,
strengthening the argument of a receptor-mediated effect, only
in the NSC-34 cells (ESI: Fig. S4†).
Acknowledgements
The reversibility of [Gd.L5] receptor binding was also demon-
strated through a glutamate/aspartate washing experiment. As
before, differentiated NSC-34 cells underwent simultaneous incu-
bation with a solution of [Gd.L5] (10 mM) and AvidinAlexaFluorꢀ
488 conjugate (2.5 mM, 10 min) (Fig. 2D). By washing the cells with 5
successive aliquots (Vtot ¼ 500 mL) of a glutamate rich (1 mM)
culture medium, a ten-fold drop in uorescence intensity was
observed, compared to the original cell staining experiment
(Fig. 2E). Furthermore, by substituting glutamate in the wash
solution with the weaker natural agonist, aspartate (5 aliquots of 1
mM), only 38% of the original intensity was observed (ESI Fig. S6†).
The ability of the probe to displace glutamate was also
demonstrated. When differentiated NSC-34 cells were sequen-
tially treated with ve volumetric aliquots of a glutamate rich
(1 mM) culture medium, then washed with normal culture
media and nally incubated with a solution of [Gd.L5] (10 mM)
and AvidinAlexaFluorꢀ 488 conjugate (2.5 mM, 10 min), a
35% uorescence recovery was observed, compared to the
original cell staining experiment (Fig. 2F). Such behaviour
demonstrates the capability of [Gd.L5] to displace glutamate from
the receptor-binding site, for example aer a glutamate burst.
Taken together these results indicate that [Gd.L5] binds to
the cell surface glutamate-binding site of the NMDAR, via the
antagonist squaramide moiety.
We thank the EPSRC and the ERC (FCC266804) for support.
Notes and references
1 S. F. Traynelis, L. P. Wollmuth, C. J. McBain, F. S. Menniti,
K. M. Vance, K. K. Ogden, K. B. Hansen, H. J. Yuan,
S. J. Myers and R. Dingledine, Pharmacol. Rev., 2010, 62,
405–496.
2 L. Mony, J. N. C. Kew, M. J. Gunthorpe and P. Paoletti, Br. J.
Pharmacol., 2009, 157, 1301–1317.
3 G. J. Lees, Neuroscience, 1993, 54, 287–322.
4 A. B. Petrenko, T. Yamakura, A. Baba and K. Shimoji,
Anesthesia and Analgesia, 2003, 97, 1108–1116.
5 S. Cull-Candy, S. Brickley and M. Farrant, Curr. Opin.
Neurobiol., 2001, 11, 327–335.
6 E. A. Waxman and D. R. Lynch, Neuroscientist, 2005, 11, 37–
49.
¨
7 H. Brauner-Osborne, J. Egebjerg, E. Ø. Nielsen, U. Madsen
and P. Krogsgaard-Larsen, J. Med. Chem., 2000, 43, 2609–
2645.
8 W. E. Childers and R. B. Baudy, J. Med. Chem., 2007, 50,
2557–2562.
9 S. A. Hires, Y. Zhu and R. Y. Tsien, Proc. Natl. Acad. Sci.
U. S. A., 2008, 105, 4411–4416.
10 P. Caravan and Z. Zhang, Eur. J. Inorg. Chem., 2012, 1916–
1923.
Conclusions
In summary, a series of MR imaging probes targeted to the 11 D. E. Murphy, J. Schneider, C. Boehm, J. Lehmann and
NMDAR has been created, using an antagonist-receptor
M. Williams, J. Pharmacol. Exp. Ther., 1987, 240, 778–784.
3152 | Chem. Sci., 2013, 4, 3148–3153
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
12 K. Hanaoka, A. J. M. Lubag, A. Castillo-Muzquiz, T. Kodadek 21 S. Aime, C. Cabella, S. Colombatto, S. Geninatti Crich,
and A. D. Sherry, Magn. Reson. Imaging, 2008, 26, 608–617.
13 R. B. Lauffer, Chem. Rev., 1987, 87, 901–927.
E. Gianolio and F. Maggioni, J. Magn. Reson. Imaging, 2002,
16, 394–406.
14 W. A. Kinney, N. E. Lee, D. T. Garrison, E. J. Podlesny, 22 J. Carmichael, W. G. Degraff, A. F. Gazdar, J. D. Minna and
J. T. Simmonds, D. Bramlett, R. R. Notvest, D. M. Kowal
J. B. Mitchell, Cancer Res., 1987, 47, 936–942.
and R. P. Tasse, J. Med. Chem., 1992, 35, 4720–4726.
23 P. Marchand, J. Becerril-Ortega, L. Mony, C. Bouteiller,
´
´
´
´
´
´
15 P. Bako, T. Novak, K. Ludanyi, B. Pete, T. Laszlo and
P. Paoletti, O. Nicole, L. Barre, A. Buisson and C. Perrio,
G. Keglevich, Tetrahedron: Asymmetry, 1999, 10, 2373–2380.
Bioconjugate Chem., 2011, 23, 21–26.
16 D. M. Corsi, C. Platas-Iglesias, H. van Bekkum and 24 F. Kielar, A. Congreve, G.-l. Law, E. J. New, D. Parker,
J. A. Peters, Magn. Reson. Chem., 2001, 39, 723–726.
17 A. E. Merbach, The Chemistry of Contrast Agents in Medical
K.-L. Wong, P. Castreno and J. de Mendoza, Chem.
Commun., 2008, 2435–2437.
Magnetic Resonance Imaging, ed. A. E. Merbach and 25 C. P. Montgomery, B. S. Murray, E. J. New, R. Pal and
´
´
E. Toth, Wiley, New York, Chichester, 2001.
D. Parker, Acc. Chem. Res., 2009, 42, 925–937.
18 C. J. Eggett, S. Crosier, P. Manning, M. R. Cookson, 26 E. J. New, D. Parker and R. D. Peacock, Dalton Trans., 2009,
F. M. Menzies, C. J. McNeil and P. J. Shaw, J. Neurochem.,
2000, 74, 1895–1902.
672–679.
27 E. J. New, A. Congreve and D. Parker, Chem. Sci., 2010, 1,
19 A. Mishra, S. Gottschalk, J. Engelmann and D. Parker, Chem.
Sci., 2012, 3, 131–135.
111–118.
28 C. A. Puckett and J. K. Barton, Biochemistry, 2008, 47, 11711–
20 W. A. Kinney, M. Abou-Gharbia, D. T. Garrison, J. Schmid,
11716.
D. M. Kowal, D. R. Bramlett, T. L. Miller, R. P. Tasse, 29 C. A. Puckett and J. K. Barton, J. Am. Chem. Soc., 2009, 131,
M. M. Zaleska and J. A. Moyer, J. Med. Chem., 1998, 41,
236–246.
8738–8739.
30 M. Botta, Eur. J. Inorg. Chem., 2000, 399–407.
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci., 2013, 4, 3148–3153 | 3153