Microscopic visualization of metabotropic glutamate receptors on the surface of living cells using bifunctional magnetic resonance imaging probes

Article abstract of DOI:10.1021/cn400175m

A series of bimodal metabotropic glutamate-receptor targeted MRI contrast agents has been developed and evaluated, based on established competitive metabotropic Glu receptor subtype 5 (mGluR₅) antagonists. In order to directly visualize mGluR₅ binding of these agents on the surface of live astrocytes, variations in the core structure were made. A set of gadolinium conjugates containing either a cyanine dye or a fluorescein moiety was accordingly prepared, to allow visualization by optical microscopy in cellulo. In each case, surface receptor binding was compromised and cell internalization observed. Another approach, examining the location of a terbium analogue via sensitized emission, also exhibited nonspecific cell uptake in neuronal cell line models. Finally, biotin derivatives of two lead compounds were prepared, and the specificity of binding to the mGluR₅ cell surface receptors was demonstrated with the aid of their fluorescently labeled avidin conjugates, using both total internal reflection fluorescence (TIRF) and confocal microscopy.

Full text of DOI:10.1021/cn400175m

Research Article
pubs.acs.org/chemneuro
Microscopic Visualization of Metabotropic Glutamate Receptors on
the Surface of Living Cells Using Bifunctional Magnetic Resonance
Imaging Probes
Anurag Mishra,*,†,∥ Ritu Mishra,‡ Sven Gottschalk,§,∥ Robert Pal,† Neil Sim,† Joern Engelmann,§
Martin Goldberg,‡ and David Parker*,†
†Department of Chemistry and ‡School of Biological and Biomedical Sciences, Durham University, South Road, Durham DH1 3LE,
England
§High Field MR Centre, Max Planck Institute for Biological Cybernetics, Spemannstrasse 41, Tuebingen 72076, Germany
S
* Supporting Information
ABSTRACT: A series of bimodal metabotropic glutamate-receptor targeted MRI contrast agents has been developed and
evaluated, based on established competitive metabotropic Glu receptor subtype 5 (mGluR5) antagonists. In order to directly
visualize mGluR5 binding of these agents on the surface of live astrocytes, variations in the core structure were made. A set of
gadolinium conjugates containing either a cyanine dye or a fluorescein moiety was accordingly prepared, to allow visualization by
optical microscopy in cellulo. In each case, surface receptor binding was compromised and cell internalization observed. Another
approach, examining the location of a terbium analogue via sensitized emission, also exhibited nonspecific cell uptake in neuronal
cell line models. Finally, biotin derivatives of two lead compounds were prepared, and the specificity of binding to the mGluR5
cell surface receptors was demonstrated with the aid of their fluorescently labeled avidin conjugates, using both total internal
reflection fluorescence (TIRF) and confocal microscopy.
KEYWORDS: mGluR5, imaging agents, lanthanides, MRI, microscopy
he human brain is a complex system in both intuitive and
complexity.8 MR imaging procedures can be substantially
Tcomputational terms. It is involved in the processing of improved when applied in combination with paramagnetic
contrast agents (CAs) to enhance sensitivity and image quality.
Gadolinium (Gd3+)-based CAs are used because of their high
spin paramagnetism and the slow electronic relaxation of Gd3+,
which influence the longitudinal and transverse relaxation times
(T1 and T2) of surrounding water protons, thereby altering the
image contrast during MR measurements.9 The use of
nonspecific CAs show enhanced sensitivity and image quality,
thereby improving the accuracy of prognoses in clinical
applications. The sensitivity and specificity of imaging can be
augmented several fold by introduction of “responsive CAs
(RCAs)”. These agents respond to changes in their
surroundings by sensing changes in the biochemical environ-
ment, following spatially localized neural activation. In
principle, these RCAs may exhibit relaxivity change by
modulation of three parameters: (i) the number of water
molecules bound directly to the gadolinium ion (q); (ii) the
cognitive, sensory, and motor information via neurons which
are one of the basic building blocks of the nervous system.1,2
Chemical neurotransmitters are involved in signal transmission
through neurons. One of the active chemical messengers,
glutamate (Glu), is abundantly distributed in the mammalian
central nervous system (CNS), and plays a critical role in
mediating excitatory signals through both G-protein-coupled
metabotropic receptors and ligand-gated ionotropic receptors
present on the postsynaptic neuronal cells.1−3 The metabo-
tropic Glu receptor subtype 5 (mGluR5) is known to be actively
involved in modulating excitatory signals via a heterogeneous
family of G-protein-coupled receptors that are activated by
Glu.4,5 An imbalance of Glu concentration and loss of their
corresponding receiving mGluR5 in the synaptic cleft have been
implicated in a number of CNS disorders, such as pain, anxiety,
depression, Parkinson’s disease, and addiction.1,4,6,7
The noninvasive imaging of the brain using a powerful
technique such as magnetic resonance imaging (MRI) [high
spatial resolution imaging (≤200 μm)] has revolutionized our
understanding of the brain’s organisational and operational
Received: September 23, 2013
Revised: November 19, 2013
Published: November 19, 2013
© 2013 American Chemical Society
128
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
Figure 1. Structures of the imaging probes examined in this study.
a
Scheme 1.
a
Reagents, conditions, and yields: (a) DIPEA, DMF, RT, 18-22%; (b) H2O, pH = 8, RT, 32-41%.
water exchange rate (kex); and (iii) the molecular rotational
correlation time (τR). These changes can be detected and
quantified by MRI.9 We and others have published reports of
pH, enzyme, and Ca2+ sensitive RCAs, for example.10,11
A key challenge in neuroimaging is to develop appropriate
‘chemical tools’ for MRI to detect synaptic glutamate
fluctuations noninvasively. One approach could be to target
Glu directly, by making small chemical entities. The design of
such chemical entities is hampered by low selectivity toward
Glu in comparison to similar amino acids such as aspartate or
glutamine. In our approach, we follow an indirect path and
target mGluR5 located on the postsynaptic membranes that are
modulated by glutamate upon neuronal firing. In order to
devise a probe that functionally reports the Glu fluctuation, it is
important that even during signal transduction the probe
should mainly remain on the membrane surface bound to the
receptors and not be internalized by receptor-mediated
endocytosis.
Recently, we introduced a new set of chemical tools based on
an antagonistic approach to target mGluR5 and detect changes
in glutamate concentration.12 The results from a functional
cellular calcium assay (potent antagonistic effects; 3.9 0.9 and
3.1
0.3, respectively) and the enhancement of receptor
mediated cellular relaxation rates (R1,cell = 32% and 28%,
respectively) at 3T revealed that two MRI probes
([Gd.L3 and 8]) (Figure 1) specifically interacted with cellular
mGluR5.13 However, the observed increase in R1,cell could also
result from internalization of the complex, via receptor-
mediated endocytosis. Probe−receptor interaction assays at
low temperature using MRI suggested that the probe interacted
with cell surface receptors but may also have been taken up into
cells.13 Nonetheless, there is no direct proof to confirm that the
compound is indeed specifically interacting with mGluR5 on
the live cell surface and is not getting internalized in cells
through receptor-mediated endocytosis. Furthermore, due to
the intrinsic insensitivity of MR imaging compared to optical
129
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
a
Scheme 2.
a
Reagents, conditions, and yields : (a) Na2CO3, MeCN, 60°C, 40%; (b) K2CO3, MeCN, 80°C, 30−40%; (c) NaOH, H2O:MeOH (1:9), RT; (d) A/
B, HOBt, NMM, EDC, DMF, RT, 20−28%; (e) 11/12/17/18, CH2Cl2:TFA (1:9), RT; (f) L4/L9/17/18, LnCl3.6H2O (Ln3+=Gd3+/Tb3+), H2O,
pH 5.5, 80°C; (g) (i) 15, 0.15 M Ba(OH)2, glyme, H2O, 80°C, (ii) CH2Cl2:TFA (1:9), RT, 78%; (h) (i) 16, 10% PdC, MeOH, 40 psi, RT, (ii)
CH2Cl2:TFA (1:9), RT, 65%; (i) D-biotin, HATU, DIPEA, DMF, RT, 15−22%.
imaging, it is impossible to visualize the complexes
[Gd.L3 and 8] directly in/on cells. Hence, a series of bifunctional
imaging probes was designed to allow real-time visualization of
the live cell surface receptors, using high-resolution fluores-
cence microscopic methods (Figure 1).
tert-butyl 2-bromopentanedioate (C)14 yielded the protected
DO2A-GA (8). Subsequent treatment of 8 with 2-bromo-N-
((5-oxo-5H-chromeno[2,3-b]pyridin-2-yl)methyl)acetamide
(E) and Hunig’s base in anhydrous MeCN yielded tetrasub-
̈
stituted intermediate 9, from which the carboxylic acid 10 was
obtained following basic hydrolysis of the benzyl group. The
azaxanthone bromoamide (E) was synthesized via alkylation of
2-(aminomethyl)-5-H-chromeno[2,3-b]pyridin-5-one14 (as pre-
viously reported) with 2-bromoacetyl bromide in MeCN. The
acid, 10, was coupled with the amines A and B, using N′-(3-
dimethylaminopropyl)-N-ethylcarbodiimide (EDC), 1-hydrox-
ybenzotriazole (HOBt), and N-methylmorpholine (NMM) in
DMF, to afford the protected amides, 11 and 12, respectively.
Following removal of the ester groups with TFA, the desired
lanthanide complexes, [Ln.L4 and 9], were formed by treatment
with LnCl3·6H2O (Gd3+ and Tb3+) at pH 6.5.
RESULTS AND DISCUSSION
■
Synthesis of Bifunctional Imaging Probes. In order to
allow the direct visualization of the cell surface mGluR5, we
designed characterized and evaluated three sets of imaging
probes (fluorescently labeled [L1,2,6 and 7] and MRI probes
[Gd.L4−5 and 9−10] (Figure 1). These probes were synthesized
using alkyne and dipyridyl amide derived antagonistic amines
[A and B]. These amine precursors were synthesized using a
range of standard synthetic transformations including Suzuki/
Sonogashira coupling, alkylation, amide formation, protection,
and deprotection (Supporting Information Scheme S1).
The biotin coupled antagonistic gadolinium complexes,
[Gd.L5 and 10], were synthesized in 7 steps as shown in Scheme
2. Alkylation of 8 with (S)-tert-butyl 6-(benzyloxycarbonylami-
no)-2-bromohexanoate (D)15 gave the compound 13, which
was further treated with base, cleaving the benzyl group to
afford the acid 14. The tetraesters 15 and 16 were synthesized
by coupling the amines A/B and acid 14 by using standard
carbodiimide methodology [EDC/HOBt/NMM]. The inter-
mediate amines, 17 and 18, were obtained by successive
deprotections (benzylcarbamate removal by 0.15 M Ba(OH)2,
glyme for alkyne amine; hydrogenation using PdC as the
catalyst for dipyridyl amine and tert-butyl groups with
The four fluorescent probes [L1,2,6 and 7] were synthesized
using either cyanine 5.5 NHS ester with Hunig’s base in DMF
̈
or fluorescein isothiocyanate at pH 8−8.5 in water, in single
step conjugation reactions with the antagonistic amines, A and
B (Scheme 1). In FITC conjugation, the pH was not allowed to
exceed 8.5 because at pH > 9, lactone cleavage of the FITC part
occurs, leading to the formation of a side product.
The azaxanthone based antagonistic lanthanide complexes,
[Ln.L4 and 9], were synthesized as depicted in Scheme 2.
Alkylation of the tert-butyl ester of 1,7-bis(carboxymethyl)-
1,4,7,10-tetraazacyclododecane (DO2A) by the (S)-5-benzyl 1-
130
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
CH2Cl2:TFA). Ligands 17 and 18 were loaded with Gd3+ using
GdCl3·6H2O in water at pH 6.5. The amines of Gd.17 and
Gd.18 were coupled to the carboxylic acid group of D-biotin,
using HATU and DIPEA in DMF, to afford the final complexes
[Gd.L5 and 10], respectively.
Fluorescently Labeled Probes. In the first attempt to
visualize mGluR5 on live astrocytes, the first set of four optical
probes [L1 and 6] (1 μM) and [L2 and 7] (10 μM) were incubated
on cultured astrocytes for 15 or 45 min (37 °C, 5% CO2). After
incubation, the cells were washed twice with Hank’s balanced
salt solution (HBSS) to remove unbound material. The Cy5.5
conjugated probes, [L1 and 6] (Figure 1), showed a complete
intracellular distribution in mGluR5 expressing astrocytes. No
signal was obtained in TIRF microscopy, but a punctate
intracellular distribution was observed in epifluorescence mode
The final complexes [Ln.L4 and 9] and [Gd.L5 and 10] were
purified by reverse phase HPLC, and the metal ion
concentration of [Tb.L4 and 9] and [Gd.L4,5 and 9,10] was
measured using Evans’ bulk magnetic susceptibility measure-
ments.16 The longitudinal proton relaxivities (r1p) of
[Gd.L4,5 and 9,10] were measured in aqueous solution (pH 7.4)
at 37 °C and 1.4 T, and found to be 7.49, 7.23, 7.98, and 7.88
mM−1 s−1, respectively. Such values are slightly higher than
(data not shown). The mitochondrial distribution of [L1 and 6
]
was confirmed using the commercial stain, MitoTracker
(Invitrogen) (500 nM) and coincubation with [L1 and 6] (1
μM) (Supporting Information Figure S2). Such observations
indicate that in this case the cellular distribution of the probes is
not driven by the receptor binding moiety but instead is
determined by the larger Cy5.5 fluorophore subunit. The
structural and charge similarity of Cy5.5 to the mitochondrial
dye may partly explain the intracellular distribution of the
probes.
9
those reported for related monoaqua gadolinium complexes,
as a consequence of their bigger molecular volume and slower
rotational correlation time, coupled to a fast water exchange
rate and a significant second sphere contribution to the overall
relaxivity. The relative importance of this effect has been
emphasized earlier.17,18
Biochemical Evaluation of Imaging Probes. Cellular
Receptor Expression and Cytotoxicity. The in vitro receptor
binding behavior of the newly modified imaging probes was
investigated on cortical rat astrocytes that are known to express
functional mGluR5 abundantly.19 Optical microscopy studies
were undertaken with cultured mGluR5 expressing secondary
astrocytes (Supporting Information). The secondary astrocytes
were obtained from frozen-harvested, confluent primary
astrocyte cultures which were gently thawed and then
recultured on surface-modified glass chamber slides (Support-
ing Information). The medium was changed after 24 h to allow
the cells to attach to the chamber slide surface. Fresh G-5
supplement (Invitrogen) containing medium was used to
increase the expression of mGluR5,20 and the cells were used
for the microscopy studies after 4 days. The expression of
mGluR5 was confirmed by immunofluorescence staining
methods (Supporting Information Figure S1).12,13 In vitro
MRI studies were done on primary astrocytes as explained
earlier.12,13 None of the imaging probes (up to 200 μM)
induced any significant cytotoxicity on astrocytes after 24 h of
incubation, as observed in a metabolic activity test for
mitochondrial dehydrogenase activity (XTT assay) (data not
shown).
Microscopic Techniques Used to Visualize mGluR5 on Live
Astrocytes. To investigate the cellular distribution of these
imaging probes, two fluorescence imaging techniques have been
used. Total internal reflection fluorescence (TIRF) microscopy
was used to analyze the localization of the probes near the
plasma membrane. Due to its experimental design, TIRF
microscopy facilitates very shallow depth penetration and
primarily illuminates the fluorophores near to the coverslip
adhered cell surface; hence, the signal from intracellular regions
is reduced to a minimum.21 A signal from the imaging probes at
a maximum distance of 150 nm from the plasma membrane
could be obtained employing this microscopy technique. Using
the same cell preparations, live-cell laser scanning confocal
microscopy (LSCM) studies were also performed. Extracellular
receptor tagging and/or intracellular localization of probes in
the entire cell was observed. Depth profiling “z-stacks” were
recorded, selectively scanning through well-defined subcellular
sections of the cell. These z-stacks were subjected to three-
dimensional reconstruction, in order to clearly depict the
distribution of the imaging probes in/on the astrocytes.
On the other hand, the fluorescein conjugated probes
(Figure 1), [L2 and 7] (10 μM, 45 min), could be visualized on
the cell surface of astrocytes using TIRF microscopy. However,
by examining the epifluorescence images recorded by TIRF and
the intracellular planes obtained from LSCM, it was confirmed
that these probes were also present in intracellular regions
(Supporting Information Figure S3). The conjugate L7 was
brighter both on the cell surface and inside, consistent with
better receptor binding/uptake of this probe compared to L2.
In the past, we22−25 and others26,27 have shown that a
consequence of appending a membrane permeable (relatively
lipophilic) fluorescent moiety can be to perturb cell-uptake
mechanisms, thereby promoting nonspecific cell internalization
of the probe. Therefore, these fluorescently labeled probes
[L1,2,6 and 7] were not considered to be appropriate to visualize
cell-surface mGluR5 and therefore should not be compared
with the MR performance of the parent probes [Gd.L3 and 8
]
that gave rise to the observed receptor mediated R1,cell
enhancements.
Luminescent Probes Based on Terbium Emission. In an
alternative approach, azaxanthone-based luminescent probes,
[Tb.L4 and 9] (Figure 1), were designed to tag and visualize
mGluR5. In this set of imaging probes, we preserved the parent
molecular structures [Gd.L3 and 8] and appended the lumino-
phore on the trans-position of [Gd.L3 and 8] (i.e., about 10 Å
away) to allow visualization by optical microscopy. Terbium-
(III) complexes of [L4 and 9] were incubated with cultured
secondary astrocytes for either 10 or 45 min, using 10 or 100
μM loading concentrations, at 37 °C, 5% CO2, and cells were
washed twice with HBSS to remove any unbound probe that
may be present. In order to present high quality microscopy
images with adequate brightness and S/N ratio, for example, to
compensate for lower probe emission/uptake, a 100 μM
loading concentration at 45 min was finally used. The cellular
localization profile of the complexes was observed by examining
the terbium emission (450−570 nm) by fluorescence
microscopy, following excitation of the azaxanthone chromo-
phore at 355 nm. Optical sections through the cell were
examined by LCSM, and confirmed that these two complexes
were also distributed within the cell, consistent either with cell
surface receptor mediated or nonspecific uptake (Supporting
Information Figure S4).
131
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
In the past, we have identified that such amide-linked
azaxanthone sensitizing moieties promote probe uptake into
the cell by macropinocytosis.28 We have also observed that
small structural modifications to the chromophore of a family of
emissive Tb3+ complexes have an influence on the cellular
compartmentalization profile.25,29 Therefore, by virtue of the
lipophilic nature of the azaxanthone moiety, the distribution of
these molecules may also not be directly compared to the
behavior of the parent complexes, [Gd.L3 and 8].
astrocytes were sequentially incubated with [Gd.L10] (10 μM,
10 min), AvidinAlexaFluor488 dye (2.5 μM, 5 min), CellMask
Orange (5 μg mL−1 for 5 min), and Hoechst 33342 (1 μg mL−1
for 5 min). During successive incubation, the cells were washed
twice with HBSS to remove unreactive compounds. The TIRF
images displayed a uniform labeling of the plasma membrane
by CellMaskOrange (Figure 2B), while the [Gd.L10]-
Microscopic Visualization by Complex Conjugate Inter-
actions Based on Biotin/Avidin. In order to target and
visualize mGluR5 on the cell surface selectively, we needed to
perturb the parent MRI probe [Gd.L3 and 8] structure
minimally, minimizing changes to complex charge and
hydrophilicity. Accordingly, a final set of probe molecules
[Gd.L5 and 10] was prepared, each being derived from 12 (N4,
C8) membered macrocyclic ring, tert-butyl ester of trans-N1,N7-
DO2A [(7-carbonylmethyl-1,4,7,10-tetraazacyclododec-1-yl)-
acetic acid].
The probe structure retained the same linkage to the
antagonists (A and B), and D-biotin was appended to the side
chain lysine amine, linked to the trans-related ring nitrogen N10
of the macrocycle (Figure 1). This design was based on the
hypothesis that if the cell-surface mGluR5 binding is mainly
responsible for the increase in R1,cell for [Gd.L3 and 8], the
antagonistic moiety of the newly designed [Gd.L5 and 10] would
interact with the cell specific receptors and the remote, trans-
substituted biotin moiety would be available to bind to the
emissive AvidinAlexaFluor488 (Invitrogen) to allow visual-
ization of probe localization by optical microscopy techniques.
Secondary astrocytes expressing mGluR5 were incubated
with [Gd.L5 and 10] (100 μM, 10−45 min) and then washed
with HBSS to remove the unbound probes. Following
subsequent incubation with AvidinAlexaFluor488 (25 μM, 5
min, 37 °C, 5% CO2), cells were washed with HBSS again to
eliminate any unbound dye. These AvidinAlexaFluor488-
[Gd.L5 or 10]-mGluR5 tagged cells were imaged by both TIRF
and LSCM. The TIRF images revealed a strong signal near the
plasma membrane, consistent with receptor tagging; the
punctate distribution resembled the pattern of the mGluR5
distribution on the cell surface (Supporting Information Figures
S1 and S5). The intracellular sections of these labeled
astrocytes were imaged by LCSM, confirming the membrane
binding of the probes. The intracellular regions of the cells were
not significantly stained, even following a 45 min incubation
(Supporting Information Figure S5). Further demonstration/
visualization of the selective cell-surface labeling of
[Gd.L5 and 10] was provided by examining a HCC projection
(color coded topological depth map reconstruction from the
recorded z-stacks (voxel size 120 × 120 × 780 nm3) of the
optical sections reconstructed topological surface map projec-
tion of the chosen 7 μm deep section of the scanned raster). A
HCC projection allows the color coded depth profile
information of all the sections to be projected onto one
plane (Supporting Information Figure S6 and Videos S1 and
S2). These studies also revealed that [Gd.L10] labeled better at
low concentrations (10 μM [Gd.L10], 2.5 μM of AvidinAlexa-
Fluor488) compared to [Gd.L 5] (Supporting Information
Figure S7).
Figure 2. Detailed visualization of the cell surface using TIRF
microscopy: (left column) and LSCM (right) of live mGluR5
expressing secondary astrocytes after labeling with [Gd.L10] (10 μM
for 10 min), (A) Avidin Alexa-Fluor488 (green) (2.5 μM for 5 min),
(B) CellMask Orange (red) (5 μg mL−1 for 5 min), and (C) Hoechst
33342 (blue) (1 μg mL−1 for 5 min). The figure shows the
AvidinAlexa-Fluor488 conjugated to receptors labeled with [Gd.L10],
CellMask Orange staining of the cell membrane, and Hoechst 33342
labeling of cell nuclei. Hoechst images in TIRF (left column, C) were
acquired in epifluorescence mode. (D) Overlay of (A), (B), and (C)
images. Bar size is 11 μm in the TIRF images, and 30 μm in the LSCM
images.
AvidinAlexaFluor488 conjugate exhibited a punctate cell surface
localization (Figure 2A left). The images obtained by LSCM
confirmed the results from the TIRF analysis, proving that the
conjugate is predominantly located on the outside of the cell
surface; significant intracellular staining was not observed by
LCSM, only a few punctate perinuclear fluorescent conjugates
(Figure 2A right). The Hoechst 33342 images were acquired in
epifluorescence mode on the TIRF microscope to locate the
nuclei in the astrocytes (Figure 2C left). An overlay of the
In order to confirm the cell-surface tagging of mGluR5, a
detailed study with [Gd.L 10] was undertaken. The nonspecific
plasma membrane stain dye CellMask Orange was used, parallel
with the nuclear blue stain dye, Hoechst 33342. The secondary
132
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
Figure 3. (A) T1-weighted MR images of 1 × 107 untreated cells (control) and cells treated for 45 min with 200 μM [Gd.DOTA] or [Gd.L3−5,8−10].
Images were recorded using a turbo spin echo sequence with a matrix of 256 × 256 voxels over a field of view of 110 × 110 mm2, slice thickness of 1
mm resulting in a voxel size of 0.4 × 0.4 × 1.0 mm3, TR 1000 ms, TE 13 ms, Ti 23 ms, and 20 averages. (B) Cellular relaxation rates R1,cell of primary
astrocytes after treating with increasing concentrations of [Gd.DOTA] or [Gd.L3−5, 8−10] for 45 min. Control represents cells incubated with HBSS
without CA. [Gd.DOTA] served as a negative control. Data are means of n = 1−6 SEM. *P < 0.05, **P < 0.01, ***P < 0.001 vs control. ANOVA
with Bonferroni’s multiple comparison post-test. Note: The concentrations for [Gd.L10] were 142 and 200 μM.
LSCM image depicts the orange/green envelope of the plasma
membrane, with the blue nuclei shown in the center (Figure 2D
right).
scanner (Figure 3A). The T1-weighted images were used to
calculate cellular longitudinal relaxation rates R1,cell, which
represents cellular labeling (Figure 3B).12,13 The complexes
[Gd.L3] and [Gd.L8] showed statistically significant increases
in R1,cell, as indicated recently (Figure 3B).13 Furthermore, the
trans-substituted biotin probes [Gd.L5] and [Gd.L10] showed
similar behavior (200 μM; increase in R1,cell = 38% and 36%,
respectively) to that of the parent analogues ([Gd.L3 and 8]).
The modulations in R1,cell (R1 = 1/T1) on interacting with
mGluR5 were attributed to slower molecular tumbling of the
complex [increase in rotational correlation time (τR)] and fast
water exchange rate with bulk water (kex) when bound to the
Control Experiments. The specificity of the compounds
toward mGluR5 was also investigated by using mGluR5-negative
NIH-3T3 fibroblast cells. As described for the secondary
astrocytes, sequential loading (equivalent amount and
incubation time) of [Gd.L10] and AvidinAlexaFluor488 dye
were incubated with NIH-3T3 cells. No staining (cell surface or
intracellular localization) was observed, strengthening the
argument that receptor-mediated binding occurs only on the
mGluR5 (data not shown). In a further control experiment, the
labeling of AvidinAlexaFluor488 on astrocytes did not show any
localization using either microscopy technique (data not
shown), emphasizing that this dye itself cannot be taken up
by astrocytes. A preconjugated construct ([Gd.L5 and 10]-
AvidinAlexaFluor488) also bound with mGluR5 and no
significant uptake in cells were observed (data not shown).
Cellular Receptor-MR Imaging Probes Studies in Cell
13
cell surface receptors.
The azaxanthone-based gadolinium probes, [Gd.L4 and 9
]
behaved quite differently. No significant changes in R1,cell
were observed with [Gd.L4], whereas the higher concentration
(200 μM) of [Gd.L9] led to a small increase in R1,cell after 45
min of incubation, albeit to a lesser extent than its
corresponding parent probe [Gd.L8] (200 μM [Gd.L9]:
113.5% of control vs 200 μM [Gd.L8] with 128.2% of control).
A significant loss in the apparent R1,cell of [Gd.L4 and 9] can be
explained by reference to the luminescence studies of the
corresponding Tb complexes [Tb.L4 and 9]. The lanthanide
luminescence microscopy studies had shown the well-defined
punctate vesicular distribution profile, possibly associated with
lysosomal localization, as seen previously with Ln(III)
complexes of related structure and overall complex charge.23
Therefore, it is reasonable to speculate that confinement within
vesicles restricts the exchange of water molecules that can
access the paramagnetic metal center. Such a “relaxation
quenching” effect has already been reported by Aime and co-
1
Suspensions: Cellular H-MR Relaxation Enhancement. The
structural modification of the mGluR5 parent MRI probes
12,13
[Gd.L3 and 8
]
can have a profound effect on their receptor
binding efficiency. Therefore, it was necessary to investigate the
receptor binding behavior of these newly modified imaging
probes [Gd.L4,5,9,10] by MRI, comparing it with the perform-
ance of the parent complexes [Gd.L3 and 8]. Primary astrocytes
were incubated with [Gd.L3−5 and 8−10] (100−200 μM probe
concentrations, 45 min incubation time at 37 °C, 5% CO2),
washed with buffer to remove unbound molecules, and
resuspended in fresh buffer, and subsequently, T1-weighted
images were recorded on a 3T Siemens human whole body MR
133
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
workers30,31 and has been interpreted in terms of a three
compartment relaxation model.32
(S)-Di-tert-butyl 2,2′-(4-(1-tert-butoxy-5-(2-(3-(6-(6-methylpyri-
din-2-ylcarbamoyl)pyridin-3-yl)phenoxy)ethylamino)-1,5-dioxo-
pentan-2-yl)-10-(2-oxo-2-((5-oxo-5H-chromeno[2,3-b]pyridin-2-yl)-
methylamino)ethyl)-1,4,7,10-tetraazacyclododecane-1,7-diyl)-
diacetate (12). A solution of 10 (0.17 g, 1.86 mmol), amine (B) (65
mg, 1.86 mmol), NMM (0.05 mL, 3.71 mmol), and HOBt (28 mg,
2.04 mmol) in anhydrous DMF (2 mL) was stirred at 0−5 °C for 15
min, and then EDC (40 mg, 2.04 mmol) was added. The reaction
mixtures were stirred for 12 h at room temperature. The completion of
reaction was verified by TLC. The solution was poured into water (20
mL) and extracted with EtOAc (3 × 20 mL). The combined organic
layers were dried over anhydrous Na2SO4 and filtered, and the filtrate
evaporated under reduced pressure. The residue was purified by
column chromatography (silica gel, 10% MeOH in CH2Cl2, Rf = 0.15)
SUMMARY AND CONCLUSIONS
■
In conclusion, we have developed probes that specifically target
mGluR5 on astrocytes and have described the direct
observation of cell surface receptor binding using optical
imaging techniques. By employing a two-step approach, based
on a remotely linked biotin−avidin interaction, the binding
ability of the parent molecule was minimally perturbed and
characterization by both MRI and optical methods was
rendered possible.
1
to give 12 as a light yellow gum (61 mg, 22%). H NMR (400 MHz,
The next step will focus on the use of longitudinal studies for
testing these imaging probes using intracranial injection directly
in the mouse brain to map the mGlu5 receptor density. The
blood-brain barrier (BBB) is the most crucial bottleneck in
attaining molecular delivery through the blood capillaries. For
further clinical application there is an obvious need for
noninvasive delivery of diagnostic agents to the brain. An
inability to enter the brain through the blood flow, either
developing artificial methods for overcoming this barrier are
required for such anionic complexes delivery in the brain or the
successful RCAs could be conjugated with already established
BBB permeating agents.33 We will also extend these
encouraging results by noninvasively mapping of mGluR5 by
MRI in depressed, anxiety, and drug-abuse mouse models. The
potential development of medications for the treatment of
addiction and other neuropsychiatric disorders, these imaging
probes could also provide new pathological information of the
brain.
CDCl3): δ 1.37 (s, 9H, 3 × CH3), 1.44 (s, 9H, 3 × CH3), 1.46 (s, 9H,
3 × CH3), 1.92−2.03 (m, 2H, CH−CH2), 2.18−2.36 (m, 2H, CH2−
CO−NH), 3.56 (s, 3H, CH3), 2.64−3.88 (br. m, 22H, CO−CH, NH−
CH2, CO−CH2, ring CH2, CO−CH, NH−CH2), 4.06 (t, J = 8.00 Hz,
2H, O−CH2), 4.69 (d, J = 7.5 Hz, 2H, CH2), 6.89 (d, J = 8.5 Hz, 1H,
HPy), 6.97 (d, J = 7.5 Hz, 2H, HAr), 7.30−7.40 (m, 5H, 5 × HAr, HPy),
7.44 (s, 1H, NH), 7.46 (s, 1H, NH), 7.68 (t, J = 8.0 Hz, 1H, HPy), 7.70
(t, J = 8.0 Hz, 1H, HPy), 7.77 (d, J = 7.0 Hz, 1H, HPy), 7.95 (d, J = 8.0
Hz, 1H, HPy), 8.26 (d, J = 8.5 Hz, 1H, HPy), 8.26 (d, J=8.5 Hz, 1H,
HAr), 8.55 (d, J = 8.5 Hz, 1H, HPy), 8.72 (s, 1H, HPy), 10.61 (br. s., 1H,
NH). ESI HRMS ( ): calcd C58H76N8O11S, m/z 1183.6186 [M+H]+;
found, 1183.6221 [M+H]+.
Common Synthesis of L4 and L9. Compounds 11/12 (1 equiv)
were dissolved in TFA/CH2Cl2 (9:1 mL) and stirred overnight. The
solvent was removed by evaporation and dried under reduced
pressure.
(S)-2,2′-(4-(1-Carboxy-4-(2-(3-((2-methylthiazol-4-yl)ethynyl)-
phenoxy)ethylamino)-4-oxobutyl)-10-(2-oxo-2-((5-oxo-5H-
chromeno[2,3-b]pyridin-2-yl)methylamino)ethyl)-1,4,7,10-tetraaza-
cyclododecane-1,7-diyl)diacetic Acid (L4). L4 was obtained as an off-
Overall, these synthetic imaging probes could have
importance not only in studying cell receptor distribution,
but also in allowing the study of signaling and activation
processes in the brain, by allocating specific regions of the brain
to be monitored by MRI, following functional stimulation.
1
white sticky solid (13.5 mg, 32%). H NMR (400 MHz, D2O): δ
1.76−2.04 (m, 2H, CH−CH2), 2.95−3.13 (m, 2H, CH2−CO−NH),
3.21 (s, 3H, CH3), 2.75−3.82 (br. m, 25H, CO−CH, NH−CH2, CO−
CH2, CH2 ring), 3.85−4.24 (m, 4H, CH2-CO, NH−CH2), 6.88 (d, J =
8.0 Hz, 1H, HAr), 7.02 (s, 1H, HAr), 7.20 (t, J = 8.0 Hz, 1H, HAr),
7.35−7.69 (m, 4H, HAr, HPy), 7.81 (s, 1H, S−CH-C), 8.16 (t, J = 7.5
Hz 1H, HAr), 8.57 (d, J = 8.0 Hz, 1H, HAr), 8.78 (d, J = 7.5 Hz, 1H,
HPy). ESI LRMS ( ): calcd C46H52N8O11S, m/z 925.3 [M+H]+;
found, 925.7 [M+H]+.
METHODS
■
Synthesis of Imaging Probes. (S)-Di-tert-butyl 2,2′-(4-(1-tert-
butoxy-5-(2-(3-((2-methylthiazol-4-yl)ethynyl)phenoxy)-
ethylamino)-1,5-dioxopentan-2-yl)-10-(2-oxo-2-((5-oxo-5H-
chromeno[2,3-b]pyridin-2-yl)methylamino)ethyl)-1,4,7,10-tetraaza-
cyclododecane-1,7-diyl)diacetate (11). A solution of 10 (0.2 g, 2.18
mmol), 2-[3-{(2-methylthiazol-4-yl)ethynyl}phenoxy]ethanamine (A)
(57 mg, 2.18 mmol), NMM (0.06 mL, 4.32 mmol), and HOBt (32
mg, 2.4 mmol) in anhydrous DMF (2 mL) was stirred at 0−5 °C for
15 min, and then N′-(3-dimethylaminopropyl)-N-ethyl-carbodiimide
[EDC] (47 mg, 2.40 mmol) was added. The reaction mixtures were
stirred for 12 h at room temperature. The completion of reaction was
verified by TLC. The solution was poured into water (20 mL) and
extracted with EtOAc (3 × 20 mL). The combined organic layers were
dried over anhydrous Na2SO4 and filtered, and the filtrate evaporated
under reduced pressure. The residue was purified by column
chromatography (silica gel, 10% MeOH in CH2Cl2, Rf = 0.15) to
(S)-2,2′-(4-(1-Carboxy-4-(2-(3-(6-(6-methylpyridin-2-
ylcarbamoyl)pyridin-3-yl)phenoxy)ethylamino)-4-oxobutyl)-10-(2-
oxo-2-((5-oxo-5H-chromeno[2,3-b]pyridin-2-yl)methylamino)-
ethyl)-1,4,7,10-tetraazacyclododecane-1,7-diyl)diacetic Acid (L9). L9
was obtained as an off-white sticky solid (15 mg, 29%). 1H NMR (400
MHz, MeOD): δ 2.02−2.58 (m, 4H, CHCH2CH2CONH), 2.68 (s,
3H, CH3), 2.72−4.02 (br. m, 27H, NH−CH2, CO−CH, O−CH2,
CO−CH2, ring CH2), 4.37 (d, J = 7.5 Hz, 2H, CH2), 6.89 (d, J = 8.5
Hz, 1H, HPy), 7.13 (d, J = 7.5 Hz, 1H, HPy), 7.31 (d, J = 7.5 Hz, 1H,
HPy), 7.35 (d, J = 8.5 Hz, 1H, HAr), 7.38 (t, J = 8.0 Hz, 1H, HAr), 7.48
(t, J = 8.0 Hz, 1H, HPy), 7.57 (d, J = 8.5 Hz, 2H, HAr), 7.71 (d, J = 7.0
Hz, 2H, HAr), 7.92 (t, J = 7.5 Hz, 1H, HAr), 8.03 (d, J = 8.0 Hz, 1H,
HPy), 8.06 (d, J = 8.0 Hz, 1H, HAr), 8.10 (d, J = 8.0 Hz, 1H, HPy), 8.48
(d, J = 8.5 Hz, 1H, HPy), 8.75 (s, 1H, HPy). ESI LRMS (+): calcd
C52H58N10O12S, m/z 1015.2 [M+H]+; found, 1015.5 [M+H]+.
Common Synthesis of [Ln.L4] and [Ln.L9]. Gadolinium complexes
[Ln.L4] and [Ln.L9] were prepared from corresponding solutions of
1
give 11 as a light yellow gum (53 mg, 22%). H NMR (700 MHz,
CDCl3): δ 1.42 (s, 9H, 3 × CH3), 1.44 (s, 9H, 3 × CH3), 1.47 (s, 9H,
3 × CH3), 2.01−2.20 (m, 2H, CH−CH2), 2.23−2.40 (m, 2H, CH2−
CO−NH), 2.76 (s, 3H, CH3), 2.79−3.60 (br. m, 22H, CO−CH, NH−
CH2, CO−CH2, ring CH2), 3.62−3.75 (m, 3H, CO−CH, NH−CH2),
4.08 (t, J = 8.00 Hz, 2H, O−CH2), 4.70 (d, J = 7.5 Hz, 2H, CH2), 6.13
(br. s, 1H, NH), 6.87 (d, J = 8.5 Hz, 1H, HAr), 7.05 (s, 1H, HAr), 7.17
(d, J = 8.0 Hz, 1H, HAr), 7.19 (d, J = 8.0 Hz, 1H, HPy), 7.25 (t, J = 8.0
Hz, 1H, HAr), 7.43 (d, J = 7.5 Hz, 1H, HAr), 7.59 (t, J = 8.0 Hz, 1H,
HAr), 7.75 (s, 1H, SCHC), 7.78 (d, J = 7.5 Hz, 1H, HAr), 8.31 (t, J =
8.0 Hz, 1H, HAr), 8.64 (d, J = 7.5 Hz, 1H, HPy), 9.73 (br. s, 1H, NH).
ESI HRMS ( ): calcd C58H76N8O11S, m/z 1093.5433 [M+H]+;
found, 1093.5453 [M+H]+.
the ligands L4/L9 (1 equiv) and solutions of LnCl3·6H2O (Ln3+
=
Gd3+, Tb3+; 1.1 equiv). The reaction mixture was stirred at 60 °C for
20 h. The pH was periodically checked and adjusted to 6.0 using
solutions of NaOH (1 M) and HCl (1 N) as needed. After
completion, the reaction mixture was cooled down and passed through
Chelex-100 to trap free Ln3+ ions, and the Ln3+-loaded complexes were
eluted. The fractions were dialyzed (500 MW cutoff; Spectra/Pro
biotech cellulose ester dialysis membrane, Spectrum Laboratories) and
lyophilized to obtain off-white solids. The absence of free Ln3+ was
checked with xylenol orange indicator. These complexes were
134
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
characterized by ESI-LRMS in positive/negative modes, and the
appropriate isotope pattern distributions for Ln3+ were recorded.
[Tb.L4]. ESI LRMS (−): calcd C46H49N8O11STb, m/z 1079.24 [M−
H]−; found, 1079.27 [M-H]−.
139.2, 146.1, 147.2, 150.5, 156.3, 157.0, 159.5, 162.6, 172.8, 172.8,
173.1, 174.7, 175.1. ESI HRMS ( ): calcd C67H97N9O13, m/z
1236.7278 [M+H]+; found, 1236.7267 [M+H]+.
2,2′-(4-((R)-5-Amino-1-carboxypentyl)-10-((S)-1-carboxy-4-(2-(3-
((2-methylthiazol-4-yl)ethynyl)phenoxy)ethylamino)-4-oxobutyl)-
1,4,7,10-tetraazacyclododecane-1,7-diyl)diacetic Acid (17). A sol-
ution of 15 (20 mg, 0.18 mmol) in H2O (1 mL), 0.15 M Ba(OH)2
(0.06 mL, 0.54 mmol), and glyme (5 mL) was stirred at 60 °C for
overnight. The reaction mixture was evaporated under reduced
pressure. The crude intermediate product was dissolved in TFA/
CH2Cl2 (9:1 mL) and stirred overnight. The solvent was removed by
evaporation and dried under reduced pressure to give 17 as a yellow
gum (10.9 mg, 78%). 1H NMR (700 MHz, MeOD): δ 1.48−1.64 (m,
2H, CH2), 1.68−2.01 (m, 4H, CH2), 2.04−2.19 (m, 2H, CH2CO−
NH), 2.71 (s, 3H, CH3), 2.90−3.40 (br. m, 22H, ring CH2, CO−CH,
CH2−NH2, CH−CH2), 3.47−3.53 (m, 2H, CO−NH−CH2), 3.54−
3.61 (m, 4H, CH2−CO), 4.05 (t, J = 8.00 Hz, 2H, O−CH2), 6.96−
7.04 (m, 1H, HAr), 7.06−7.17 (m, 2H, HAr), 7.29 (t, J = 8.0 Hz, 1H,
HAr), 7.64 (s, 1H, C−CH-S). ESI LRMS (+): calcd C37H53N7O10S, m/
z 788.3 [M+H]+; found, 788.5 [M+H]+.
2,2′-(4-((R)-5-Amino-1-carboxypentyl)-10-((S)-1-carboxy-4-(2-(3-
(6-(6-methylpyridin-2-ylcarbamoyl)pyridin-3-yl)phenoxy)-
ethylamino)-4-oxobutyl)-1,4,7,10-tetraazacyclododecane-1,7-diyl)-
diacetic Acid (18). A solution of 16 (30 mg, 0.24 mmol), (10%) Pd−
C (1/5, w/w) in MeOH (5 mL) under H2 (40 psi) was stirred at room
temperature in a Parr apparatus for 4 h. The reaction mixture was
filtered through Celite, and the filtrate evaporated under reduced
pressure. The crude intermediate product was dissolved in TFA/
CH2Cl2 (9:1 mL) and stirred overnight. The solvent was removed by
evaporation and dried under reduced pressure to give 18 as a yellow
gum (14.3 mg, 65%). 1H NMR (700 MHz, MeOD): δ 1.40−1.55 (m,
2H, CH2), 1.60−1.73 (m, 2H, CH2), 1.77−1.90 (m, 2H, CH2), 1.93−
2.08 (m, 2H, CH2), 2.19−2.38 (m, 2H, CH2), 2.44 (s, 3H, CH3),
2.70−3.47 (br. m, 24H, CO−CH, NH−CH2, CO−CH2, ring CH2),
3.48−3.67 (m, 2H, NH−CH2), 4.02 (t, J = 8.00 Hz, 2H, O−CH2),
6.84 (d, J = 8.0 Hz, 1H, HPy), 7.03 (d, J = 8.0 Hz, 2H, HAr), 7.63 (d, J =
8.0 Hz, 2H, HAr), 7.76 (t, J = 7.5 Hz, 1H, HPy), 8.10 (d, J = 8.00 Hz,
1H, HPy), 8.13 (d, J = 8.00 Hz, 1H, HPy), 8.14 (d, J = 8.00 Hz, 1H,
HPy), 8.87 (s, 1H, HPy). ESI LRMS (+): calcd C43H59N9O11, m/z 878.4
[M+H]+; found, 878.5 [M+H]+.
[Gd.L4]. ESI LRMS (−): calcd C46H49GdN8O11S, m/z 1078.24
[M−H]−; found, 1078.60 [M-H]−. r1p = 7.49 mM−1 s−1 (60 MHz, 310
K).
[Tb.L9]. ESI LRMS (−): calcd C52H55N10O12Tb, m/z 1169.32 [M−
H]−; found, 1169.69 [M-H]−.
[Gd.L9]. ESI LRMS (−): calcd C52H55GdN10O12, m/z 1168.32 [M−
H]−; found, 1168.74 [M-H]−. r1p = 7.98 mM−1 s−1 (60 MHz, 310 K).
Di-tert-butyl 2,2′-(4-((R)-6-(benzyloxycarbonylamino)-1-tert-bu-
toxy-1-oxohexan-2-yl)-10-((S)-1-tert-butoxy-5-(2-(3-((2-methylthia-
zol-4-yl)ethynyl)phenoxy)ethylamino)-1,5-dioxopentan-2-yl)-
1,4,7,10-tetraazacyclododecane-1,7-diyl)diacetate (15). A solution
of 14 (90 mg, 0.96 mmol), 2-[3-{(2-methylthiazol-4-yl)ethynyl}-
phenoxy]ethanamine (A) (25 mg, 0.96 mmol), NMM (0.028 mL, 1.98
mmol), and HOBt (15 mg, 1.11 mmol) in anhydrous DMF (1 mL)
was stirred at 0−5 °C for 15 min, and then EDC (22 mg, 1.11 mmol)
was added. The reaction mixtures were stirred for 18 h at room
temperature. The completion of reaction was verified by TLC. The
solution was poured into water (20 mL) and extracted with EtOAc (3
× 20 mL). The combined organic layers were dried over anhydrous
Na2SO4 and filtered, and the filtrate evaporated under reduced
pressure. The residue was purified by column chromatography (silica
gel, 10% MeOH in CH2Cl2, Rf = 0.35) to give 11 as a light yellow gum
1
(22 mg, 20%). H NMR (400 MHz, CDCl3): δ 1.19−1.36 (m, 4H,
CH2), 1.38 (s, 9H, 3 × CH3), 1.41 (s, 9H, 3 × CH3), 1.43 (s, 9H, 3 ×
CH3), 1.44 (s, 9H, 3 × CH3), 1.84−2.62 (m, 6H, CH2), 2.72 (s, 3H,
CH3), 2.73−3.55 (br. m, 23H, CO−CH, NH−CH2, CO−CH2, ring
CH2), 3.58−3.77 (m, 3H, CO−CH, NH−CH2), 4.02 (t, J = 8.00 Hz,
2H, O−CH2), 5.06 (s, 2H, CH2), 6.19 (br. s, 1H, NH), 6.34 (br. s.,
1H, NH), 6.88 (d, J = 9.0 Hz, 1H, HAr), 7.04 (s, 1H, HAr), 7.14 (t, J =
8.5 Hz, 1H, HAr), 7.24 (d, J = 9.0 Hz, 1H, HAr), 7.27−7.35 (m, 5H,
HAr), 7.37 (s, 1H, C−CH-S). 13C NMR (101 MHz, CDCl3): δ 19.1,
23.1, 26.4, 27.6, 27.7, 28.2, 29.2, 32.5, 32.6, 38.8, 43.6, 53.3, 54.9, 55.6,
55.9, 58.1, 66.3, 66.6, 68.8, 69.9, 81.8, 82.0, 82.3, 83.3, 88.4, 115.5,
117.0, 122.4, 123.5, 124.6, 128.4, 129.5, 134.9, 136.6, 144.1, 156.5,
158.1, 165.7, 170.5, 172.8, 173.4, 174.8. ESI HRMS ( ): calcd
C61H91N7O13S, m/z 1146.6519 [M+H]+; found, 1146.6498 [M+H]+.
Di-tert-butyl 2,2′-(4-((R)-6-(benzyloxycarbonylamino)-1-tert-bu-
toxy-1-oxohexan-2-yl)-10-((S)-1-tert-butoxy-5-(2-(3-(6-(6-methyl-
pyridin-2-ylcarbamoyl)pyridin-3-yl)phenoxy)ethylamino)-1,5-dioxo-
pentan-2-yl)-1,4,7,10-tetraazacyclododecane-1,7-diyl)diacetate
(16). A solution of 14 (90 mg, 0.96 mmol), 5-[4-(2-aminoethoxy)-
phenyl]-N-(6-methylpyridin-2-yl)picolinamide (B) (35 mg, 0.96
mmol), NMM (0.028 mL, 1.98 mmol), and HOBt (15 mg, 1.11
mmol) in anhydrous DMF (1 mL) was stirred at 0−5 °C for 15 min,
and then EDC (22 mg, 1.11 mmol) was added. The reaction mixtures
were stirred for 18 h at room temperature. The completion of reaction
was verified by TLC. The solution was poured into water (20 mL) and
extracted with EtOAc (3 × 20 mL). The combined organic layers were
dried over anhydrous Na2SO4 and filtered, and the filtrate evaporated
under reduced pressure. The residue was purified by column
chromatography (silica gel, 10% MeOH in CH2Cl2, Rf = 0.4) to
Common Synthesis of Gadolinium Complexes [Gd.17] and
[Gd.18]. Gadolinium complexes [Gd.17] and [Gd.18] were prepared
from corresponding solutions of the ligands 17/18 (1 equiv) and
solutions of GdCl3·6H2O (1.1 equiv). The reaction mixture was stirred
at 60 °C for 20 h. The pH was periodically checked and adjusted to 6.0
using solutions of NaOH (1 M) and HCl (1 N) as needed. After
completion, the reaction mixture was cooled down and passed through
Chelex-100 to trap free Gd3+ ions, and the Gd3+-loaded complexes
were eluted. The fractions were dialyzed (500 MW cutoff; Spectra/Pro
biotech cellulose ester dialysis membrane, Spectrum Laboratories) and
lyophilized to obtain off-white solids. The absence of free Gd3+ was
checked with xylenol orange indicator. These complexes were
characterized by ESI-LRMS in positive/negative modes, and the
appropriate isotope pattern distributions for Gd3+ were observed.
[Gd.17]. Yield =12 mg, 100% (an off white solid). ESI LRMS (+):
calcd C37H49GdN7O10S, m/z 942.25 [M+H]+ and 471.32 [M+2H]2+;
1
give 11 as a light yellow gum (31 mg, 25%). H NMR (700 MHz,
CDCl3): δ 1.19−1.39 (m, 4H, CH2), 1.44 (s, 9H, 3 × CH3), 1.45 (s,
9H, 3 × CH3), 1.47 (s, 9H, 3 × CH3), 1.48 (s, 9H, 3 × CH3), 1.95−
2.11 (m, 2H, CH2), 2.13−2.27 (m, 2H, CH2), 2.33 (t, J = 8.00 Hz, 2H,
CH2), 2.53 (s, 3H, CH3), 2.71−3.52 (br. m, 23H, CO−CH, NH−
CH2, CO−CH2, ring CH2), 3.58−3.77 (m, 3H, CO−CH, NH−CH2),
4.12 (t, J = 8.00 Hz, 2H, CH2), 4.84 (br. s., 1H, NH), 5.09 (s, 2H,
CH2), 6.55 (br. s., 1H, NH), 6.94 (d, J = 7.5 Hz, 1H, HPy), 7.08 (d, J =
8.50 Hz, 2H, HAr), 7.28−7.37 (m, 5H, HAr), 7.57 (d, J = 8.50 Hz, 2H,
HAr), 7.66 (t, J = 8.00 Hz, 1H, HPy), 8.03 (d, J = 8.00 Hz, 1H, HPy),
8.25 (d, J = 8.00 Hz, 1H, HPy), 8.30 (d, J = 8.00 Hz, 1H, HPy), 8.80 (s,
1H, HPy), 10.49 (br. s., 1H, NH). 13C NMR (176 MHz, CDCl3): δ
24.1, 24.3, 26.4, 27.8, 28.0, 28.1, 29.2, 31.7, 34.9, 36.4, 38.5, 52.6, 53.4,
55.9, 60.4, 61.1, 66.4, 66.5, 68.3, 69.4, 81.8, 81.9, 82.0, 82.3, 110.7,
115.5, 119.2, 122.4, 128.1, 128.3, 128.5, 134.9, 135.5, 136.5, 138.5,
found, 942.79 [M+H]+ and 471.98 [M+2H]2+
.
[Gd.18]. Yield = 17 mg, 100% (an off white solid). ESI LRMS (+):
calcd C43H55GdN9O11, m/z 516.33 [M+2H]2+; found, 516.55 [M
+2H]2+
.
Common Synthesis of [Gd.L5] and [Gd.L10]. A solution of
[Gd.17]/[Gd.18] (1 equiv), D-biotin (1.1 equiv), DIPEA (2 equiv),
and HATU (1.1 equiv) in anhydrous DMF (1 mL) was stirred
overnight at room temperature. The completion of reaction was
verified by LRMS. The solution was evaporated under reduced
pressure, purified by RP-HPLC, and lyophilized to obtain off-white
solids.
[Gd.L5]. Yield = 3.26 mg, 22%. ESI-LRMS ( ): calcd
C47H63GdN9O12S2, m/z 1166.32 [M−H]− and 584.32 [M+2H]2+;
135
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
found, 1166.80 [M−H]− and 584.07 [M+2H]2+, tR = 12.6 min, r1p
=
ACKNOWLEDGMENTS
■
7.23 mM−1 s−1 (60 MHz, 310 K).
[Gd.L10]. Yield = 3.1 mg, 15%. ESI-LRMS ( ): C53H69GdN11O13S,
m/z 1256.40 [M−H]− and 629.40 [M+2H]2+; found, 1256.90 [M−
H]− and 629.17 [M+2H]2+, tR = 12.9, r1p = 7.88 mM−1 s−1 (60 MHz,
310 K).
The authors would like to thank Dr. Tim Hawkins for his
technical assistance.
REFERENCES
■
Total Internal Reflection Fluorescence (TIRF) Microscopy.
TIRF microscopy was performed on an inverted Leica TIRF MC
Imaging system with excitation lasers set to 405, 488, and 561 nm to
visualize Hoechst, GFP, and Cell mask, respectively. Acquisition was
performed using an 100×/HCX Plan Apo Oil TIRF 1.47 objective.
Imaging and image capture were performed using LASAFsoftware.
Confocal Microscopy. Cell images and colocalization experiments
were obtained using a Leica SP5 II microscope. A HeNe laser was used
to visualize SA-AF488 fluorescence. The microscope was equipped
with a triple channel imaging detector, comprising two conventional
PMT systems and a HyD hybrid avalanche photodiode detector. The
latter part of the detection system, when operated in the BrightRed
mode, is capable of improving imaging sensitivity above 550 nm by
25%, reducing signal-to-noise by a factor of 5. The pinhole was always
determined by the Airy disc size, calculated from the objective in use
(HCX PL APO 63×/1.40 NA LbdBlue), using the lowest excitation
wavelength (488 nm). Scanning speed was adjusted to 400 Hz in a
bidirectional mode, to ensure both sufficient light exposure and
enough time to collect the emitted light from the optical probes (1024
× 1024 frame size, a pixel size of 120 × 120 nm and depth of 0.89
μm). The three-dimensional reconstruction was achieved using a novel
saturation elimination algorithm update of the existing ImageJ 1.46r
3D plug-in using LSCM images recorded on the above detailed Leica
SP5 II microscope. In these z-stack images, a deliberate 20% overlap in
the applied axial resolution was introduced, determined by the applied
optics and experimental parameters detailed above.
(1) von Bohlen und Halbach, O., and Dermietzel, R. (2006)
Neurotransmitters and Neuromodulators, Wiley-VCH, Weinheim,
Germany.
(2) Featherstone, D. E., and Shippy, S. A. (2008) Regulation of
synaptic transmission by ambient extracellular glutamate. Neuroscientist
14, 171−181.
(3) Rothstein, J. D. (2000) Neurobiology. Bundling up excitement.
Nature 407, 141−143.
(4) Rousseaux, C. G. (2008) A Review of Glutamate Receptors II:
Pathophysiology and Pathology. J. Toxicol. Pathol. 21, 133−173.
(5) Conn, P. J., and Pin, J. P. (1997) Pharmacology and functions of
metabotropic glutamate receptors. Annu. Rev. Pharmacol. 37, 205−237.
(6) Cleva, R. M., and Olive, M. F. (2011) Positive Allosteric
Modulators of Type 5 Metabotropic Glutamate Receptors (mGluR5)
and Their Therapeutic Potential for the Treatment of CNS Disorders.
Molecules 16, 2097−2106.
(7) Ferraguti, F., and Shigemoto, R. (2006) Metabotropic glutamate
receptors. Cell Tissue Res. 326, 483−504.
(8) Logothetis, N. K., Guggenberger, H., Peled, S., and Pauls, J.
(1999) Functional imaging of the monkey brain. Nat. Neurosci. 2,
555−562.
(9) Merbach, A. E., and Toth, E. (2001) The Chemistry of Contrast
Agents in Medical Magnetic Resonance Imaging, Wiley, New York,
Chichester.
(10) Bonnet, C. S., and Toth, E. (2010) MRI probes for sensing
biologically relevant metal ions. Future Med. Chem. 2, 367−384.
(11) Mishra, A., Logothetis, N. K., and Parker, D. (2011) Critical in
vitro evaluation of responsive MRI contrast agents for calcium and
zinc. Chem.Eur. J. 17, 1529−1537.
ASSOCIATED CONTENT
* Supporting Information
■
S
Synthesis of intermediate amines A and B, experimental
methods, Figures S1−S7, and Videos S1 and S2. This material
is available free of charge via the Internet at http://pubs.acs.org.
(12) Mishra, A., Gottschalk, S., Engelmann, J., and Parker, D. (2012)
Responsive imaging probes for metabotropic glutamate receptors.
Chem Sci 3, 131−135.
(13) Gottschalk, S., Engelmann, J., Rolla, G. A., Botta, M., Parker, D.,
and Mishra, A. (2013) Comparative in vitro studies of MR imaging
probes for metabotropic glutamate subtype-5 receptor targeting. Org.
Biomol. Chem. 11, 6131−6141.
(14) Atkinson, P., Findlay, K. S., Kielar, F., Pal, R., Parker, D., Poole,
R. A., Puschmann, H., Richardson, S. L., Stenson, P. A., Thompson, A.
L., and Yu, J. H. (2006) Azaxanthones and azathioxanthones are
effective sensitisers for europium and terbium luminescence. Org.
Biomol. Chem. 4, 3708−3708.
AUTHOR INFORMATION
Corresponding Authors
*Fax: +49-8931872017. E-mail: anurag.mishra@helmholtz-
muenchen.de (A.M.).
■
*E-mail: david.parker@durham.ac.uk (D.P.).
Present Address
∥A.M. and S.G.: Institute for Biological and Medical Imaging,
(15) Vibhute, S. M., Engelmann, J., Verbic, T., Maier, M. E.,
Logothetis, N. K., and Angelovski, G. (2013) Synthesis and
characterization of pH-sensitive, biotinylated MRI contrast agents
and their conjugates with avidin. Org. Biomol. Chem. 11, 1294−1305.
(16) Corsi, D. M., Platas-Iglesias, C., van Bekkum, H., and Peters, J.
A. (2001) Determination of paramagnetic lanthanide(III) concen-
trations from bulk magnetic susceptibility shifts in NMR spectra.
Magn. Reson. Chem. 39, 723−726.
(17) Botta, M. (2000) Second coordination sphere water molecules
and relaxivity of gadolinium(III) complexes: Implications for MRI
contrast agents. Eur. J. Inorg. Chem., 399−407.
(18) Aime, S., Botta, M., Parker, D., and Williams, J. A. G. (1996)
Extent of hydration of octadentate lanthanide complexes incorporating
phosphinate donors: Solution relaxometry and luminescence studies. J.
Chem. Soc., Dalton Trans., 17−23.
Helmholtz Zentrum Munchen (GmbH), Neuherberg 85764,
Germany.
̈
Author Contributions
A.M. conceived the project with D.P.’s consultancy. A.M. and
N.S. performed the designing, chemical synthesis and character-
isation of imaging probes. R.M., A.M., S.G., and R.P. performed
the cell culture experiments, microscopic studies, and MRI
experiments for the characterisation of the imaging probes. J.E.,
M.G., and D.P. supported the research. All authors analyzed the
data and wrote the paper.
Funding
This work was supported by a Marie Curie Intra European
Fellowship (PIEF-GA-2009-237253) [A.M.], the ERC
(FCC266804) [A.M., N.S., R.P.], by the Biotechnology and
Biological Sciences Research Council (BBSRC), UK (BB/
G011818/1) [R.M.], the Ministry for Education and Research,
BMBF (FKZ:01EZ0813) [S.G.], and the Max Planck Society.
(19) Danial, N. N., Gramm, C. F., Scorrano, L., Zhang, C. Y., Krauss,
S., Ranger, A. M., Datta, S. R., Greenberg, M. E., Licklider, L. J.,
Lowell, B. B., Gygi, S. P., and Korsmeyer, S. J. (2003) BAD and
glucokinase reside in a mitochondrial complex that integrates
glycolysis and apoptosis. Nature 424, 952−956.
Notes
(20) Peavy, R. D., Chang, M. S. S., Sanders-Bush, E., and Conn, P. J.
(2001) Metabotropic glutamate receptor 5-induced phosphorylation
The authors declare no competing financial interest.
136
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137

ACS Chemical Neuroscience
Research Article
of extracellular signal-regulated kinase in astrocytes depends on
transactivation of the epidermal growth factor receptor. J. Neurosci. 21,
9619−9628.
(21) Mattheyses, A. L., Simon, S. M., and Rappoport, J. Z. (2010)
Imaging with total internal reflection fluorescence microscopy for the
cell biologist. J. Cell Sci. 123, 3621−3628.
(22) Mishra, A., Pfeuffer, J., Mishra, R., Engelmann, J., Mishra, A. K.,
Ugurbil, K., and Logothetis, N. K. (2006) A new class of Gd-based
DO3A-ethylamine-derived targeted contrast agents for MR and optical
imaging. Bioconjugate Chem. 17, 773−780.
(23) Montgomery, C. P., Murray, B. S., New, E. J., Pal, R., and Parker,
D. (2009) Cell-Penetrating Metal Complex Optical Probes: Targeted
and Responsive Systems Based on Lanthanide Luminescence. Acc.
Chem. Res. 42, 925−937.
(24) New, E. J., Duan, R., Zhang, J. Z., and Hambley, T. W. (2009)
Investigations using fluorescent ligands to monitor platinum(IV)
reduction and platinum(II) reactions in cancer cells. Dalton Trans.,
3092−3101.
(25) New, E. J., Parker, D., Smith, D. G., and Walton, J. W. (2010)
Development of responsive lanthanide probes for cellular applications.
Curr. Opin. Chem. Biol. 14, 238−246.
(26) Puckett, C. A., and Barton, J. K. (2008) Mechanism of Cellular
Uptake of a Ruthenium Polypyridyl Complex. Biochemistry 47, 11711−
11716.
(27) Puckett, C. A., and Barton, J. K. (2009) Fluorescein Redirects a
Ruthenium-Octaarginine Conjugate to the Nucleus. J. Am. Chem. Soc.
131, 8738−8739.
(28) Smith, D. G., Law, G. L., Murray, B. S., Pal, R., Parker, D., and
Wong, K. L. (2011) Evidence for the optical signalling of changes in
bicarbonate concentration within the mitochondrial region of living
cells. Chem. Commun. 47, 7347−7349.
(29) New, E. J., Congreve, A., and Parker, D. (2010) Definition of
the uptake mechanism and sub-cellular localisation profile of emissive
lanthanide complexes as cellular optical probes. Chem. Sci. 1, 111−118.
(30) Terreno, E., Geninatti Crich, S., Belfiore, S., Biancone, L.,
Cabella, C., Esposito, G., Manazza, A. D., and Aime, S. (2006) Effect of
the intracellular localization of a Gd-based imaging probe on the
relaxation enhancement of water protons. Magn. Reson. Med. 55, 491−
497.
(31) Geninatti Crich, S., Cabella, C., Barge, A., Belfiore, S., Ghirelli,
C., Lattuada, L., Lanzardo, S., Mortillaro, A., Tei, L., Visigalli, M.,
Forni, G., and Aime, S. (2006) In vitro and in vivo magnetic resonance
detection of tumor cells by targeting glutamine transporters with Gd-
based probes. J. Med. Chem. 49, 4926−4936.
(32) Strijkers, G. J., Hak, S., Kok, M. B., Springer, C. S., Jr., and
Nicolay, K. (2009) Three-compartment T1 relaxation model for
intracellular paramagnetic contrast agents. Magn. Reson. Med. 61,
1049−1058.
(33) Malakoutikhah, M., Teixido, M., and Giralt, E. (2011) Shuttle-
mediated drug delivery to the brain. Angew. Chem., Int. Ed. Engl. 50,
7998−8014.
137
dx.doi.org/10.1021/cn400175m | ACS Chem. Neurosci. 2014, 5, 128−137