Magnetic resonance and optical imaging probes for NMDA receptors on the cell surface of neurons: Synthesis and evaluation in cellulo

Article abstract of DOI:10.1039/c4ob01848f

A second generation of N-methyl-d-aspartate (NMDA) receptor-targeted MRI contrast agents has been synthesised and evaluated in cellulo, based on established bicyclic NMDA receptor antagonists. Their use as responsive MR imaging probes has been evaluated i

Full text of DOI:10.1039/c4ob01848f

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Magnetic resonance and optical imaging probes
for NMDA receptors on the cell surface of
neurons: synthesis and evaluation in cellulo†
Cite this: Org. Biomol. Chem., 2014,
12, 9389
Neil Sim,a Robert Pal,a David Parker,*a Joern Engelmann,b Anurag Mishra*a,c and
Sven Gottschalkb,c
A second generation of N-methyl-D-aspartate (NMDA) receptor-targeted MRI contrast agents has been
synthesised and evaluated in cellulo, based on established bicyclic NMDA receptor antagonists. Their use
as responsive MR imaging probes has been evaluated in suspensions of NSC-34 cells, and one agent
exhibited significant enhancements in measured longitudinal and transverse water proton relaxation rates
(19 and 38% respectively; 3 T, 298 K). A biotin derivative of the lead compound was prepared and the
specificity and reversibility of binding to the NMDA cell surface receptors demonstrated using confocal
laser scanning microscopy. Competitive and reversible binding of glutamate to the receptors was also
visualised, suggesting that the receptor-targeted approach may allow MRI to be used to monitor neuronal
events associated with modulation of local glutamate concentrations.
Received 29th August 2014,
Accepted 8th October 2014
DOI: 10.1039/c4ob01848f
www.rsc.org/obc
chemical composition as a function of neural activity. This
rationale has driven the recent development of contrast agents
Introduction
Receptors for N-methyl-D-aspartate (NMDA) are found on the that selectively target the various glutamate receptors that are
surface of neurons (e.g. astrocytes) and constitute part of the found on the cell surface of astrocytes in high abundance.7–9
ligand-gated ionotropic family of glutamate receptor proteins.1 Such responsive contrast agents must be able to bind selec-
These receptors play a key role in excitatory neurotransmission tively and reversibly to the receptor, so that changes in local
and in memory and learning. Many central nervous system contrast agent levels can be modulated following the gluta-
(CNS) disorders have been related to the mis-regulation or mate bursts that periodically occur following a stimulus. Given
over-stimulation of NMDA receptors including epilepsy, ische- that release of glutamate from a pre-synaptic cell takes place
mia, Parkinson’s and Alzheimer’s diseases.2–4 Selective antago- over a period of milliseconds and should displace the contrast
nists of the NMDA receptors (NMDAR) have also been studied, agent from the cell surface receptor, MRI may be able to
exploring their potential as imaging or therapeutic agents for monitor the restoration to equilibrium that occurs over a
these and related CNS disorders.5,6
period of a second or two.10 As the glutamate levels drop and
One of the goals of molecular imaging is to provide more the contrast agent competes in binding to the receptor, gluta-
detailed spatiotemporal information on the chemical pro- mate is displaced, leading to local signal intensity enhance-
cesses occurring in the brain that are associated with neuro- ment once more. Such a perturbational approach requires that
transmission and its relationship to learning, cognitive the contrast agents bind reversibly and with relatively high
behaviour and disease status and progression. The utility of affinity to the target site, with minimal non-specific binding.
MRI is improved by the use of contrast agents, especially those
For these reasons, we have set out to examine conjugates of
that enhance specificity and sensitivity by targeting certain gadolinium contrast agents with established competitive
regions of interest, or that allow the monitoring of changes in antagonists for glutamate receptors. When bound to the recep-
tor site, the local rotational dynamics and second sphere of
hydration around the gadolinium moiety are perturbed
aDepartment of Chemistry, Durham University, South Road, Durham DH1 3LE, UK.
E-mail: david.parker@dur.ac.uk, anurag.mishra@helmholtz-muenchen.de
bHigh Field MR Centre, Max Planck Institute for Biological Cybernetics,
leading to relaxivity enhancement and increased contrast. In
our initial work with NMDA receptor targeted gadolinium con-
jugates, we reported the synthesis and behaviour of competi-
Spemannstrasse 41, Tuebingen, D-72076, Germany
cInstitute for Biological and Medical Imaging, Helmholtz Center, Munich,
tive NMDAR antagonists based on
a
3,4-diamino-3-
cyclobutene-1,2-dione developed by Kinney.11a The NMDA
receptor is a tetrameric complex, most commonly built up
Neuherberg, D-85764, Germany
†Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ob01848f
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9389

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 1
from alternating GluN1 and GluN2 subunits. Based on the from the paramagnetic Gd3+ centre, the less likely that receptor
work of Kinney,11b these receptor binding moieties are a class binding would be compromised. However, the longer the
of competitive antagonists. Therefore, they must bind at the chain the less motional coupling will occur between the gado-
GluN2 subunit where the glutamate binding site is located. linium moiety and the slowly tumbling macromolecule, com-
However, the GluN2 subunit can be one of four genetic pro- paring free and receptor bound complexes and hence the
ducts (GluN2A–D) and the relative expression levels of each smaller may be any relaxivity enhancement.
can vary between each receptor. In initial work,13 immuno-
fluorescence staining experiments were undertaken to estab- conjugate, [Gd·L7], in which the biotin moiety allows selective
lish the presence of the GluN2B subunit. tagging to fluorescently labelled avidin conjugates to allow
In addition, we have examined the behaviour of the biotin
The targeting moieties based on the competitive antagonist visualization of the avidin–biotin probe conjugate using confo-
sub-unit were conjugated to a modified Gd-DOTA complex that cal microscopy.
possesses a single fast-exchanging water molecule, e.g. [Gd·La–d],
Scheme 1. When water exchange at the Gd centre is fast
(typically of the order of 107 s−1 at 298 K), then the relaxivity,
(the increment of the water proton relaxation rate per unit
complex concentration) in the low to mid-field range (1 to 3 T),
is dominated by rotational dynamics and hence is more sensi-
tive to receptor binding.12 Initial studies in cell suspensions of
the NSC-34 cell line showed relaxation enhancements of up to
75% on cell surface binding versus controls lacking the NMDA
receptors.13 The conjugates were non-toxic, as revealed by stan-
dard IC50 studies using an MTT assay that assesses pertur-
bation of mitochondrial redox status, consistent with evidence
from optical microscopy studies that revealed no evidence for
internalization of the conjugate. Furthermore, the confocal
microscopy studies established both the specificity and the
Results and discussion
Synthesis of the ligands and complexes
reversibility of probe binding, in the absence and presence of
added glutamate.13
The synthesis of the required bicyclic antagonist moieties fol-
Here, we report the synthesis and evaluation of an lowed methods adapted from the work of Kinney.11b Ortho-
additional series of complexes constituting a second-gene- gonal protection of 1,3-diamino-propan-2-ol was achieved in
ration, [Gd·L1–L6] (Scheme 1) in which the monocyclic antago- high yield following stepwise Boc and Z protection steps.14a
nist group is replaced by a bicyclic moiety, for which receptor Subsequent mesylation of the free alcohol, followed by substi-
binding affinities were established to be about an order of tution with NaN3 in DMF afforded the azide, 3 (Scheme 2). The
magnitude higher, in the original small molecule work.11b The azide group served as a ‘masked’ amine, which was trans-
nature of the spacing chain differs in each case and four formed to a primary amine group in the final step of the syn-
examples with shorter linkages (Gd·L1–Gd·L4) were examined thesis. A reagent for selective removal of the CBz protecting
together with two analogues (Gd·L5, Gd·L6) in which the group from 3 was required. This ruled out the use of standard
spacing chain was longer. The rationale behind this selection deprotection methods, such as hydrogenation over a palla-
of targets was that the further away the targeting vector was dium catalyst, as this would also reduce the azide group.
9390 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 2 Synthesis of the competitive antagonist moieties (adapted from ref. 11b).
Instead, Ba(OH)2 in refluxing aqueous glyme was used, and at pH 6, and the complexes were purified by RP-HPLC follow-
gave the free amine, 4, in good yield.14b N-Alkylation with, for ing treatment with Chelex resin to remove any excess Gd salts.
example, diethyl-(2-bromo)ethyl phosphonate in the presence The overall isolated yields were rather low (20–30%), and gave
of Na2CO3 gave the mono-substituted amine, 7, in modest the paramagnetic complexes as a mixture of diastereoisomers
yield, which was used in a Michael addition–elimination reac- in each case.
tion with diethoxy-3-cyclobutene-1,2-dione. The phosphonate
The synthesis of the biotin conjugate, [Gd·L7] proceed via a
ester 11 was obtained as a pair of rotamers in a 1 : 1.2 ratio as similar reaction sequence (Scheme 4). Z-Deprotection of the
deduced by 31P NMR spectroscopy. Cyclisation was achieved in differentially protected penta-ester, 23,15 followed be reaction
high yield after selective hydrolysis of the Boc protecting with succinic anhydride in DMF afforded the monocarboxylic
group, followed by heating the free amine in ethanol in the acid, 24. Following amide coupling (EDC, HOBt, NMM) to the
presence of a tertiary amine. By studying the 1H NMR spec- antagonist derivative, 19, hydrolysis of the methyl ester in
trum of the cyclised azide, 15, it was possible to deduce that dilute base enabled coupling of the resultant carboxylic acid
the conformation adopted by this and related bicyclic inter- with an ethylenediamine derivative of biotin. Stepwise depro-
mediates involved selective population of a twist-chair isomer tection of the Boc and phosphinate ethyl groups gave the
(ESI†). In a similar manner, compounds 13, 14 and 16 were ligand L7 and complexation with GdCl3 proceeded well at pH 6
prepared.
to afford [Gd·L7]. This complex could be used in confocal
With the azide in hand, mild, selective reduction to unmask microscopy experiments in the presence of a fluorescently
the primary amine functionality was accomplished using a labeled avidin host molecule, in order to verify cell surface
standard Staudinger reaction, using triphenylphosphine in a binding to the NMDA receptors of NSC-34 cells.
THF–H2O solvent mixture. The reduction proceeded cleanly to
Relaxivity properties and cell suspension MRI studies
yield the primary amines 17–20, as racemic mixtures.
The final step in the synthesis of [Gd·L1–6] involved amide The longitudinal proton relaxivities of the seven Gd complexes
bond formation with the carboxylic acid of a DO3A-glutarate were measured in water at 1.4 T and 310 K, and the values fell
derivative (Scheme 3).15 This step was undertaken using EDC/ in the expected range17 for such mono-aqua complexes of MW
HOBt for the formation of the active ester in the presence of 850 to 1450, for which the relaxivity value increases with mole-
NMM to deprotonate the primary amine. For the phosphonate cular volume, as it scales with the overall rotational correlation
series, removal of the ethyl ester groups was undertaken using time. Values were also measured in the presence of human
TMS-Br in DMF. Complexation with GdCl3 proceeded cleanly serum albumin (Table 1).
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9391

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 3
Scheme 4
Human serum albumin (HSA) constitutes around 4.5% of in each case. At a protein concentration of 0.7 mM, the highest
plasma and is the major protein constituent in the circulatory relaxivity values were measured for [Gd·L3]2− and [Gd·L6]2−
,
system. Each contrast agent will bind to the protein to some consistent with their higher binding affinities.
extent, disrupting its rotational dynamics and leading to an
A neuronal cell line model that expresses functional NMDA
enhancement of the longitudinal relaxation rate of the water receptors had been established earlier using the NSC-34 cell
protons. As a control experiment, the effect of added HSA on line, produced by fusing mouse spinal cord and neuroblas-
the measured relaxivity of [Gd·L1–6] (each at 1 mM) was toma cells. The cell line has been used in several studies of the
assessed at 1.4 T following incremental addition of up to evaluation of NMDAR antagonists.18 The expression of func-
1.6 mM HSA. Increases in r1p were observed and association tional NMDA receptors on differentiated NSC-34 cells was
constants were estimated by assuming a 1 : 1 stoichiometry of demonstrated as reported earlier using immuno-fluorescence
interaction. No particular trend was evident, correlating log K techniques with primary antibodies.13 The labelling of differ-
values with structure; indeed the weakest interaction occurred entiated NSC-34 cells with [Gd·L1–6] was assessed by measur-
with the phosphonates, [Gd·L4]2− and [GdL5]2−, whilst the ana- ing the longitudinal relaxation times, T1, of the water signal in
logues with one less, [Gd·L3]2−, and one more methylene cell suspensions using a Siemens human whole body MR
groups, [Gd·L6]2−, bound to albumin 100 times more strongly scanner, equipped with a head coil operating at 3 T. These
9392 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 1 Estimated binding affinities for serum albumin,a,b,c,d and relax-
ation properties of the Gd complexes (310 K, pH 7.4, 1.4 T)
addition of only a single methylene group to the targeting
moiety, accords with the lower receptor affinity of such antago-
nists in earlier work.11b The set of complexes [Gd·L1–4] each
gave less than 6 to 12% enhancements in R1 and R2 respectively,
suggesting that the receptor binding moiety may be too close
to the bulky Gd complex, leading to inhibition of probe
binding to the cell surface receptor, or that local hydration or
water exchange dynamics are compromised for the receptor
bound complex in these cases (vide infra).
[Gd·L1] [Gd·L2] [Gd·L3] [Gd·L4] [Gd·L5] [Gd·L6]
log Ka
2.2
1.9
4.54
3.7
5.76
1.6
5.21
1.6
4.97
3.7
5.18
ri1npitial/mM−1 s−1 c 4.91
rl1ipmit/mM−1 s−1 c 18.0
22.0
5.42
7.50
6.63
30.0
5.89
25.0
5.51
8.50
6.82
r1p/mM−1 s−1 b
5.99
(at 0.7 mM HSA)
a Errors in log K values associated with experimental variation and
statistical fitting were estimated to be 0.1. The binding constants
were estimated assuming a 1 : 1 stoichiometry; the relative affinity
constant (IC50 value) for a structurally similar NMDA binding moiety
(the anionic alcohol analogue of the phosphinate) was reported to be
The IC50 values of the gadolinium complexes [Gd·L1–6] were
assessed using an MTT assay that probes the perturbation of
mitochondrial redox activity of the differentiated NSC-34
cells.19 Following a 24-hour incubation, each complex exhibi-
ted no cytotoxic effect at concentrations of up to 200 μM (data
not shown), consistent with the hypothesis that the anionic
probes were not being internalized into the cell. This hypo-
thesis was corroborated by measuring the Gd cell uptake using
ICP-MS. After a 24 h incubation, followed by washing of the
cells with fresh buffer three times, less than 2% Gd3+ cell
uptake was measured in each case, and the gadolinium was
found in the washings (ca. 98%).
19 nM, determined by
a
radiolabelled competition assay.11b
b Statistical errors associated with the measurement of r1p values are
0.03. c The r1lipmit and ri1npitial values were derived from the iterative least-
squares fitting analysis; values for the complexes of L4 and L5 are
subject to the largest error. d For [Gd·L7]2−, the relaxivity value in water
was 7.23 mM−1 s−1 (310 K, 60 MHz).
measurements allowed the calculation of cellular relaxation
rates, R1,cell. Differentiated cells were incubated for 45 min
with 200 μM [Gd·L1–6] (37 °C, 5% CO2), washed with Hank’s
Cellular optical imaging
Buffered Saline Solution (HBSS) to remove unbound complex, It was hypothesised that the observed increase in relaxation
re-suspended in fresh buffer and R1 was determined in cell rates per cell were due to binding of the Gd probe to NMDA
suspensions using an inversion recovery sequence (Fig. 1). receptors on the cell surface. However, cell labelling of the Gd
Values for R2 were also measured.
complex via non-specific binding to the cell membrane or by
The largest enhancements in relaxation were evident with internalisation involving receptor-mediated endocytosis are
[Gd·L5]2− for which the enhancement was 19% in R1 and plausible mechanisms that may lead to the observed increase
37.5% in R2 over the untreated control and [Gd·DOTA]− treated in R1,cell. To distinguish these possibilities, the biotin complex
cells. Moreover, this was one of the complexes exhibiting the was designed to allow tracking of the contrast agent on the cell
weakest binding to serum albumin (Table 1). Given that surface.
albumin binding may be considered as a model for ‘non-
By attaching a fluorescent label to a non-competitive
specific’ binding to a protein, such behaviour is encouraging. antagonist, the direct visualisation of the NMDAR-2B subunit
The much lower R1,cell value for [Gd·L6]2−, notwithstanding the has been demonstrated recently.20 Accordingly, [Gd·L7] was
Fig. 1 Cellular 1H MR relaxation rates, (a) R1,cell and (b) R2,cell in cell suspensions (3 T, 298 K) after treatment of differentiated NSC-34 cells with
200 μM [Gd·DOTA]− or [Gd·L1–6] for 45 minutes. Under the same conditions, the Gd3+ complexes of La–Ld gave corresponding changes of +18, +70,
−9 and +76% respectively.13 [Gd·DOTA]− served as a negative control. Values are mean SEM (n = 4–8). *P < 0.05, **P < 0.01, ***P < 0.001 statistical
significant difference vs. untreated control; aP < 0.001 statistical significant difference vs. [Gd·DOTA]− (ANOVA with Dunnett’s multiple comparison
post test, assessed using Graphpad Prism 5.04).
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9393

View Article Online
Paper
Organic & Biomolecular Chemistry
designed, bearing the antagonist binding moiety of [Gd·L5], as culture medium that was enriched in glutamate (1 mM), a
this probe had shown the largest increase in R1,cell values. The nine-fold drop in fluorescence intensity was observed, com-
complex includes a remote trans-substituted biotin sub-unit, pared to the original cell staining experiment (Fig. 2D). The
attached to the macrocyclic core. It was hypothesised that if ability of the probe to displace glutamate was also demon-
cell-surface receptor binding is responsible for the increase in strated. When differentiated NSC-34 cells were sequentially
R1,cell for [Gd·L5], the biotin moiety of [Gd·L7], after forming a treated the avidin/biotin components and with five volumetric
tightly bound complex to added AvidinAlexaFluor® 488 conju- aliquots of a glutamate rich (1 mM) culture medium, then
gate, would be tagged to the outside of the cell. The presence washed with normal culture media and finally incubated with
of the fluorescent dye on the avidin sub-unit allows direct visua- a solution of [Gd·L7] (10 μM) and AvidinAlexaFluor® 488 conju-
lisation of the biotin complex using fluorescence microscopy, gate (2.5 μM, 10 min), 38% of signal was observed after
only when bound to the cell surface.
10 minutes and 98% of the original fluorescence signal inten-
In order to study the behaviour and localisation of [Gd·L7], sity was recovered after 45 minutes (Fig. 2E). Such behaviour
laser scanning confocal microscopy (LSCM) imaging studies established the ability of the probe [Gd·L7] to displace gluta-
were carried out using live cells. The differentiated NSC-34 mate from the NMDA receptor-binding site, as may occurred
cells were grown on a microscope slide with a flow-through after a glutamate burst during neural signalling.
channel and were subjected to various different experimental
In summary, these results show that [Gd·L7] is able to bind
conditions. Following incubation of the cells with a solution of to the cell surface glutamate-binding site of the NMDAR, via
[Gd·L7] (10 μM) and AvidinAlexaFluor® 488 conjugate (2.5 μM, the antagonist entity.
10 min) together, the cells were examined by microscopy and a
The competitive antagonist, 27, was synthesized in accord-
localization profile was observed resembling ‘pit-like’ regions ance with the methods of Kinney, and has been shown to bind
at the cell surface (Fig. 2A). The dye CellMask™ Orange is com- to the NMDA receptor with an affinity of the order of 20 nM.11b
mercially available and can be used to stain the cell plasma Co-incubation of 27 (10 μM) with [Gd·L7]2− and the Alexa-Fluor
membrane non-specifically. By repeating the simultaneous avidin conjugate gave rise to <7% of signal intensity (Fig. 2F
loading experiment of [Gd·L7] and AvidinAlexaFluor® 488 vs. Fig. 2A). By carrying out the following incubation sequence
conjugate (10 min) and co-incubating with CellMask™ sequentially, first 27, then a Glu wash and finally incubation
Orange (5 min, 5 μg mL−1), unequivocal evidence for the with [Gd·L7]2−, the observed signal grew to 18%. Evidently, the
selective localisation of [Gd·L7] at the cell membrane was small molecule has a higher affinity for the NMDA receptor
obtained (Fig. 2B and 2C). Furthermore, by using the same than the Gd probe, [Gd·L7]2−, but the restoration of some
microscope settings as used previously,13 it was possible to observed signal (18%) after the glutamate wash (Fig. 2G)
demonstrate that cells treated with [Gd·L7] were 38% brighter suggests that it may be important not to have a probe that is
than when they were treated with the acyclic analogue, consist- bound too tightly, or glutamate may not compete for receptor
ent with the bicyclic targeting moiety having a higher receptor binding.
affinity.
Finally, competition experiments with equal concentrations
and [Gd·L7]2− were under-
Confirmation that the observed fluorescence was due to of the parent complexes, [Gd·L3/5 2−
]
the strong biotin–avidin interaction between [Gd·L7] and Avidin- taken (Table 2 and Fig. 2H and I), and the observation of
AlexaFluor® 488 conjugate on the cell surface was given about 50% of the original signal in each case is consistent
through various control experiments. Differentiated cells with each of these complexes having a similar binding affinity
loaded with AvidinAlexaFluor® 488 conjugate only and differ- for the NMDA receptor site.
entiated but untreated NSC-34 cells each gave rise to no fluo-
rescence signal in the visible region, using the same
experimental parameters as above. Cell surface localisation
did not vary for stepwise vs. simultaneous incubations of the
Conclusions
biotin and avidin moieties and was independent of time A second generation of MR imaging probes has been created
(5 to 45 minutes) and the loading concentration (up to that target the NMDAR, based on a competitive antagonist
100 μM). No staining was observed with NIH-3T3 cells approach. Of the series of six probes examined, [Gd·L5],
(mouse skin fibroblasts) that do not possess the NMDA showed a 19%, increase in the measured water proton relax-
receptors at the cell surface. In each of these control experi- ation rate in the presence of functional NMDARs on differen-
ments, no evidence for any significant intracellular staining tiated NSC-34 cells. Cell-surface localisation was demonstrated
was observed.
using confocal microscopy for the biotin-functionalised deriva-
tive, [Gd·L7]. No evidence was found for probe internalisation
during the time-periods used, neither was any evidence found
Competition experiments and reversibility of receptor binding
The reversibility of [Gd·L7] receptor binding was demonstrated for cell-surface binding, following probe incubation with an
using a glutamate competition experiment. Differentiated NMDA receptor negative cell-line, consistent with the inherent
NSC-34 cells were incubated with a solution of [Gd·L7] (10 μM) specificity of this approach.
and AvidinAlexaFluor® 488 conjugate (2.5 μM, 10 min). After
Binding of the Gd contrast agents to the NMDA receptors
washing the cells with five successive aliquots (Vtot = 500 μL) of was shown to be reversible. Following addition of glutamate to
9394 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 2 Live cell laser scanning confocal microscopy (LSCM) images of differentiated NSC-34 cells following treatment with [Gd·L7]. (A) Simultaneous
loading of [Gd·L7] (10 μM), and AvidinAlexaFluor® 488 conjugate (2.5 μM, 10 min) allowing visualisation of AvidinAlexaFluor® 488 conjugate λex/λem
=
488/505–555 nm. (B) As in (A), but with a 5 minute incubation of the non-specific cell-surface membrane dye CellMask™ Orange, allowing visua-
lisation of CellMask™ Orange λex/λem = 543/550–660 nm. (C) RGB merge showing co-localisation of the AvidinAlexaFluor® 488 conjugate and Cell-
Mask™ Orange on the cell surface. (D) As in image (A) but with a post glutamate (1 mM) wash showing that [Gd·L7] is removed from the cell surface.
(E) As image (D) and then simultaneous loading of [Gd·L7] (10 μM), and AvidinAlexaFluor® 488 conjugate (2.5 μM, 45 min) allows 98% recovery of the
fluorescence signal intensity as compared to image (A). (F) Loading of the antagonist, 27, (10 μM, 10 min) followed by simultaneous loading of
[Gd·L7] (10 μM), and AvidinAlexaFluor® 488 conjugate (2.5 μM, 10 min). (G) Loading of 27 (10 μM, 10 min), followed by a glutamate wash (1 mM) and
then simultaneous loading of [Gd·L7] (10 μM), and AvidinAlexaFluor® 488 conjugate (2.5 μM, 45 min). (H) Simultaneous loading of [Gd·L7] (10 μM),
AvidinAlexaFluor® 488 conjugate (2.5 μM) and [Gd·L3] (10 μM, 10 min). (I) Simultaneous loading of [Gd·L7] (10 μM), AvidinAlexaFluor® 488 conjugate
(2.5 μM) and [Gd·L5] (10 μM, 10 min). See also the data given in Table 2.
the receptor-bound conjugate [Gd·L7], displacement of the
These in vitro and in cellulo studies suggest that such
complex from the receptors occurred, leading to a reduction of probes could be used in MRI studies in vivo, when the probes
signal intensity in microscopy, that could be partially recov- are injected intra-cranially to maintain a steady state local
ered by subsequent incubation with [Gd·L7] and the Avidin- concentration, that may allow time-dependent modulation of
AlexaFluor® 488 conjugate.
glutamate concentration to be observed.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9395

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 2 Changes in the observed fluorescence intensity of [Gd·L7]-Avidin-Alexa-488 conjugate (2.5 μM), under various conditions
[Gd·L7]a, Glu
Antagonist 27c,
[Gd·L7]a, post
washb, [Gd·L7]a Antagonist, 27c, Glu washb,
[Gd·L7]a and
[Gd·L7]a and
[Gd·L7]a Glu washb
(10/45 min)
[Gd·L7]a
[Gd·L7]a
[Gd·L3]d (1 : 1) [Gd·L5]d (1 : 1)
Relative fluorescence intensity/% 100
11
38/98
<7
18
43
56
a Concentration of [Gd·L7] was 10 μM. b Concentration of added glutamate in the medium was 1 mM. c Concentration of the antagonist 27, was
10 μM in each case. d Concentration of [Gd·L3/5] = 10 μM.
Melting points were recorded using a Gallenkamp (Sanyo)
apparatus and are uncorrected.
Experimental
Materials and methods
Relaxivity measurements were carried out at 310 K, 60 MHz
All solvents used were laboratory grade and anhydrous sol- (1.4 T) on a Bruker Minispec mq60 instrument. The mean
vents, when required, were freshly distilled over the appropri- value of three independent measurements was recorded. The
ate drying agent. Water was purified by the ‘PuriteSTILLplus’ relaxivities of the compounds were calculated as the slope of
system, with conductivity of ≤4 μS cm−1. All reagents used the function shown in the equation below,
were purchased from commercial suppliers (Acros, Aldrich,
Fluka, Merck and Strem) and were used without further purifi-
cation unless otherwise stated. Reactions requiring anhydrous
conditions were carried out using Schlenk line techniques
under an atmosphere of argon.
1
1
¼
þ r1½GdꢀLnꢁ
ð1Þ
T1;obs T1;d
where T1,obs is the measured T1, T1,d is the diamagnetic contri-
bution of the solvent (calculated to be 4000 ms) and [Gd·Ln]
is the concentration in mM of the appropriate Gd3+ complex
(n = 1–6). Errors for all relaxivity values were less than
Thin layer chromatography was performed on neutral alu-
minium sheet silica gel plates (Merck Art 5554) and visualised
under UV irradiation (254 nm), or using Dragendorff reagent
staining. Preparative column chromatography was performed
using silica gel (Merck Silica Gel 60, 230–400 mesh).
Reverse phase HPLC was conducted at 298 K using a Shi-
madzu system. XBridge C18 4.6 × 100 mm, i.d. 5 μm analytical
column and XBridge C18 OBD 19 × 100 mm, i.d. 5 μm prepara-
tive columns were used to analyse and purify the complexes. A
gradient elution with a solvent system composed of H2O +
0.1% HCOOH–MeOH + 0.1% HCOOH was performed for a
total run time of 20.0 min.
0.3 mM−1 s−1
.
The apparent binding constant for the interaction of the
Gd3+ complex with Human Serum Albumin (HSA) was calcu-
lated using eqn (2) below:
ꢀ
ꢁ
2
ðR ꢂ R0Þ=ðR1 ꢂ R0Þ
ðR ꢂ R0Þ
ðR1 ꢂ R0Þ
ðR ꢂ R0Þ
ðR1 ꢂ R0Þ
ðR ꢂ R0Þ
þ ½GdꢀLnꢁ ꢃ
ꢂ ½GdꢀLnꢁ ꢃ
K
ðR1 ꢂ R0Þ
½Xꢁ ¼
1 ꢂ
1H, 13C and 31P NMR spectra were recorded in commercially
available deuterated solvents on a Varian Mercury-400 (1H
399.960, 13C 100.572), Bruker Avance-400 (1H 400.052, 13C
100.603 and 31P 161.91), Varian Inova-500 (1H 499.722, 13C
125.671,), Appleby VNMRS-600 (1H 599.832, 13C 150.828), or
Varian VNMRA-700 (1H 699.731, 13C 175.948 and 31P 283.26)
spectrometer. All chemical shifts are given in ppm with coup-
ling constants in Hz.
Low resolution electrospray mass spectra (LR-MS) were
recorded on a Waters Micromass LCT or Thermo-Finnigan
LTQ FT instrument operating in positive or negative ion mode
as stated, with MeOH as the carrier solvent. Accurate mass
spectra [High resolution electrospray mass spectra (HR-MS)]
were recorded using the Thermo-Finnigan LTQ FT mass
spectrometer.
½GdꢀXꢁ
K ¼
½Xf ꢁ½Gdfꢁ
ð2Þ
where [X] is the total concentration of HSA in the solution;
[Gd·Ln]: the total concentration of the complex; K: the binding
constant; R: relaxation rate of a given concentration of X; R0:
the initial relaxation rate; R1: final relaxation rate; [Gd·X]: the
concentration of the HSA-coordinated complex; [Xf]: the con-
centration of free HSA in the mixture; [Gdf]: the concentration
of the free complex.
Details of cell experiments, microscopy, and the MTT
assays used have been given in earlier papers in this
series.7–9,13 The other compounds, ligands and Gd complexes
not reported below are described in the ESI.†
9396 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Ligand and complex synthesis
Benzyl tert-butyl(2-hydroxypropane-1,3-diyl)dicarbamate
(907 mg, 2.80 mmol) was dissolved in anhydrous dichloro-
methane (16 mL), to which triethylamine (430 μL, 3.08 mmol)
and methanesulfonyl chloride (238 μL, 3.08 mmol) were added
at 0 °C. The resulting solution was stirred at this temperature
for 5 minutes before being warmed to room temperature for a
further 1 hour until complete conversion of the starting
material was observed by ESI-MS. The solvent was removed
under reduced pressure and the residue partitioned between
brine and DCM (30 mL, 1 : 1). The organic portion was separ-
ated and the aqueous layer extracted with dichloromethane
(3 × 30 mL). The combined organic portions were dried
over MgSO4, filtered and reduced to give a colourless oil
which was used directly in the next step without further
(3-Amino-2-hydroxypropyl)carbamic acid 1,1-dimethylethyl
ester,11b 1.
A
solution of 1,3-diamino-2-hydroxypropane (5.0 g,
55.5 mmol) in anhydrous acetonitrile (50 mL) was main-
tained at room temperature and treated with a solution of di-
tert-butyl dicarbonate (4.04 g, 18.5 mmol) in anhydrous aceto-
nitrile (20 mL) over a 2-hour period with vigorous stirring.
The subsequent suspension was left to stir overnight. After a
period of 18 hours, the reaction mixture was concentrated
and the residue re-dissolved in brine (50 mL). The pH was
adjusted to 5 by treatment with HCl (1 N), washed with
dichloromethane (3 × 50 mL) and made basic (pH 12) by the
addition of NaOH solution (2.5 M). The product was extracted
using copious amounts of CHCl3 (5 × 100 mL), dried over
MgSO4, filtered and concentrated to yield 1 as a white solid
(1.23 g, 34%). 1H NMR (400 MHz, CDCl3) δ 1.43 (9H, s,
C(CH3)3), 2.42–2.72 (3H, br, m), 2.78–2.92 (2H, br, m),
3.01–3.15 (1H, br, m), 3.20–3.35 (1H, br, m), 3.56–3.72 (1H,
br, m), 5.05 (1H, br, s). 13C NMR (125 MHz, CDCl3) δ 28.4
(C(CH3)3), 44.1, 44.5 (CH2), 71.0 (COH), 79.5 (C(CH3)3), 156.7
(CO). MS (ES+) m/z 191.1 [M + H]+; C8H19N2O3 requires
191.1396; found 191.1392. M. Pt. 78–80 °C [lit.11b 77–79 °C].
Benzyl tert-butyl(2-hydroxypropane-1,3-diyl)dicarbamate, 2.14a
1
purification (1.20 g, 100%). H NMR (400 MHz, CDCl3) δ 1.43
(9H, s, C(CH3)3), 3.04 (3H, s, CH3), 3.22–3.59 (4H, m, CH2),
4.68 (1H, br, NH), 5.10 (2H, s, CH2Ph), 5.13 (1H, m, CH),
5.55 (1H, br, NH), 7.31–7.38 (5H, m, Ar-H). MS (ES+) m/z 403.1
[M + H]+.
Benzyl tert-butyl(2-azidopropane-1,3-diyl)dicarbamate, 3.
To a solution of the mesylate (1.20 g, 2.93 mmol) in anhydrous
dimethylformamide (13 mL) was added NaN3 (580 mg,
8.93 mmol) at room temperature. After addition, the cloudy
mixture was heated to 60 °C under a stream of argon for
20 hours. Once no further conversion of starting materials has
been revealed by ESI-MS, the mixture was cooled and the
solvent removed under reduced pressure. The residue was par-
titioned between EtOAc–H2O (40 mL, 1 : 1) and the organic
portion separated. The aqueous was extracted with EtOAc
(3 × 40 mL) and the combined organic portions dried
over MgSO4, filtered, concentrated to give the crude residue
which was purified by column chromatography (hexane–
EtOAc, 90 : 10 to 70 : 30 using 5% increments; Rf = 0.51) to
give a colourless oil (605 mg, 59%). 1H NMR (400 MHz,
CDCl3) δ 1.44 (9H, s, C(CH3)3), 3.07–3.28 (2H, m, CH2),
3.31–3.49 (2H, m, CH2), 3.66 (1H, quin, J = 6, CH), 5.04
(1H, br, NH), 5.11 (2H, s, CH2Ph), 5.46 (1H, br, NH), 7.28–7.39
(5H, m, Ar-H). 13C NMR (101 MHz, CDCl3) δ 28.4 (C(CH3)3),
40.9, 41.3 (CH2), 60.9 (CN3), 67.2 (CH2Ph), 80.2 (C(CH3)3),
128.2, 128.4, 128.7, 136.4 (C-Ar), 156.4, 156.9 (CO). MS (ES+)
m/z 372.1 [M + Na]+; C16H23N5O4Na requires 372.1648; found
372.1626.
To a solution of (3-amino-2-hydroxypropyl)carbamic acid 1,1-
dimethylethyl ester (1.28 g, 6.72 mmol) in anhydrous dichloro-
methane (43 mL) was added benzyl chloroformate (1.05 mL,
7.39 mmol) and diisopropylethylamine (1.29 mL, 7.39 mmol)
at 0 °C. The resulting solution was stirred under argon at this
temperature for 30 min before being allowed to warm to room
temperature and stirred overnight. After complete consump-
tion of starting materials has been revealed by TLC, the reac-
tion mixture was concentrated and the crude residue purified
by column chromatography (DCM–MeOH, 100% to 95 : 5 using
1% increments Rf = 0.26) to give a pale green oil (1.64 g, 75%).
1H NMR (400 MHz, CDCl3) δ 1.43 (9H, s, C(CH3)3), 3.06–3.35
(4H, m, CH2), 3.76 (1H, quint, J = 6, CH), 5.09 (2H, s, CH2Ph),
5.14 (1H, br, OH), 5.49 (2H, br, NH), 7.29–7.38 (5H, m, Ar-H).
13C NMR (101 MHz, CDCl3) δ 28.5 (C(CH3)3), 43.7, 44.1 (CH2),
67.1 (CH2Ph), 70.9 (COH), 80.1 (C(CH3)3), 128.2, 128.3, 128.7,
136.5 (Ar-C), 157.4, 157.6 (CO). MS (ES+) m/z 347.2 [M + Na]+;
C16H24N2O5Na requires 347.1583; found 347.1593.
tert-Butyl(3-amino-2-azidopropyl)carbamate, 4.
11,11-Dimethyl-3,9-dioxo-1-phenyl-2,10-dioxa-4,8-diazadode-
can-6-yl methylsulfonate.
Benzyl tert-butyl (2-azidopropane-1,3-diyl)dicarbamate (605 mg,
1.73 mmol) was dissolved in anhydrous glyme (40 mL) to
which a solution of Ba(OH)2 in H2O (0.15 M, 20 mL) was
added. The resulting solution was heated to 80 °C under
argon, and the reaction progress monitored by ESI-MS and
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9397

View Article Online
Paper
Organic & Biomolecular Chemistry
TLC. After 48 hours, the reaction was cooled to room temp- of methyl 3-((2-azido-3-((tert-butoxycarbonyl)amino)propyl)-
erature and the solvent removed under reduced pressure. amino)propanoate (206 mg, 0.68 mmol) in anhydrous ethanol
The residue was partially dissolved in brine and the pH (5 mL) over 30 min. The resulting solution was stirred under
adjusted to 5 using HCl (1 N). The aqueous portion was argon at room temperature and the progress of the reaction
washed with dichloromethane (2 × 100 mL) before adjusting followed by TLC. When no further reaction was observed by
the pH to 12 using NaOH solution (2.5 M). The product was TLC, the solution was concentrated under reduced pressure.
extracted with copious amounts of CHCl3 (5 × 100 mL), dried The crude residue was then purified by column chromato-
over MgSO4, filtered and concentrated to give a colourless oil graphy (DCM–MeOH, 100% to 95 : 5 using 0.5% increments;
(257 mg, 69%). 1H NMR (400 MHz, CDCl3) δ 1.45 (9H, s, Rf = 0.59) to give a colourless oil, (234 mg, 81%). This compound
C(CH3)3), 1.56 (2H, bs. s), 2.76 (1H, dd, J = 13, 7, CHHNH), exists as a pair of diastereoisomers. 1H NMR (400 MHz, CDCl3)
2.88 (1H, dd, J = 13, 5, CHHNH), 3.20 (1H, dt, J = 13, 7, δ 1.41 (9H, s, C(CH3)3), 1.43 (3H, t, J = 7, OCH2CH3), 2.66 (2H,
CHHNH2), 3.35 (1H, dt, J = 13, 5 CHHNH2), 3.50–3.60 (1H, m, t, J = 7, CH2CH2CO2Me), 3.08–3.57 (3H, m, CH2), 3.66 (3H, s,
CH), 4.89 (1H, br, NH). 13C NMR (101 MHz, CDCl3) δ 28.5 CH3) 3.71–4.07 (4H, m, CH2CH), 4.75 (2H, q, J = 7, OCH2CH3),
(C(CH3)3), 42.1, 43.4 (CH2), 64.7 (CH), 78.4 (C(CH3)3), 156.6 5.09 (1H, br, NH). 13C NMR (101 MHz, CDCl3) δ 15.9
(CO). MS (ES+) m/z 216.1 [M + H]+; C8H18N5O2 requires (OCH2CH3), 28.4 (C(CH3)3), 33.9, 42.2, 46.2, 51.1 (CH2), 52.1
216.1461; found 216.1458.
(CH3), 61.2 (CN3), 70.1 (OCH2CH3), 80.1 (C(CH3)3), 155.9 (CO),
Methyl 3-((2-azido-3-((tert-butoxycarbonyl)amino)propyl)- 171.2 (CvC), 172.7 (CO), 177.4 (CvC), 183.0, 188.4 (CO). MS
amino)propanoate, 6.
(ES+) m/z 851.1 [2M + H]+; C36H55N10O14 requires 851.3899;
found 851.3918.
Methyl 3-(4-azido-8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-
en-2-yl)propanoate, 14.
A
solution of tert-butyl (3-amino-2-azidopropyl)carbamate
(200 mg, 0.93 mmol), diisopropylethylamine (243 μL,
1.4 mmol) and methyl-3-bromopropionate (102 μL,
0.93 mmol) in anhydrous DMF (8 mL) was stirred at room
temperature for 48 hours until no further reaction was
observed by TLC. The mixture was concentrated under reduced
pressure before being partitioned between EtOAc–H2O (1 : 1,
30 mL). The organic portion was separated and the aqueous
extracted with EtOAc (3 × 25 mL). The combined organic
layer were dried over MgSO4, filtered and concentrated, allow-
ing the crude residue to be purified by column chromato-
graphy (DCM–MeOH, 100% to 94 : 6 using 1% increments; Rf =
0.47) to give a pale yellow oil (220 mg, 79%). 1H NMR
(400 MHz, CDCl3) δ 1.43 (9H, s, C(CH3)3), 1.50 (1H, br, NH),
2.77 (2H, dd, J = 12, 7, CH2), 2.89 (2H, dd, J = 12, 5, CH2), 3.09
(2H, t, J = 6, CH2), 3.36 (2H, t, J = 6, CH2), 3.71 (3H, s, CH3),
4.02 (1H, m, CH), 5.19 (1H, br, NH). 13C NMR (101 MHz,
CDCl3) δ 28.4 (C(CH3)3), 32.5, 42.0, 44.8, 49.5 (CH2), 52.3 (CH3),
59.6 (CN3), 80.4 (C(CH3)3), 156.5, 172.4 (CO). MS (ES+) m/z 302.2
[M + H]+; C12H24N5O4 requires 302.1828; found 302.1831.
To a solution of methyl 3-((2-azido-3-((tert-butoxycarbonyl)-
amino)propyl)(2-ethoxy-3,4-dioxocyclobut-1-en-1-yl)amino)pro-
panoate (78 mg, 0.18 mmol) in anhydrous dichloromethane
(1 mL) was added TFA (1 mL), and the resulting solution
stirred at room temperature for 1 hour. Reaction was complete
after this period, as indicated by TLC, and the solvents were
removed under reduced pressure. The residue was re-dissolved
in dichloromethane and again concentrated under reduced
pressure. This process was repeated 5 times to ensure com-
plete removal of excess TFA. The TFA salt was dissolved in
anhydrous ethanol (2 mL) to which a solution of triethylamine
(102 μL, 0.736 mmol) in anhydrous ethanol (1.5 mL) was
added over a 20 minute period. The resulting solution was
heated to reflux for 18 hours, until no further reaction was
observed by TLC. At this point, the solvent was removed and
the crude residue purified by column chromatography (DCM–
MeOH, 100% to 90 : 10 using 1% increments; Rf = 0.52) to give
a colourless oil (42 mg, 82%). 1H NMR (700 MHz, DMSO-d6)
δ 2.69 (2H, overlapping dt, J = 14, 7, CH2CO2Me), 3.40 (1H, dd,
J = 14, 7, Ha(eq)), 3.55 (1H, ddd, J = 14, 7, 4, Ha(axial)), 3.57–3.59
(2H, br, m, Hc (ax/eq)), 3.60 (3H, s, CH3), 3.86 (1H, dt, J = 14, 7,
CH2CH2CO2Me), 3.94 (1H, dt, J = 14, 7, CH2CH2CO2Me), 4.31
(1H, m, CH), 8.53 (1H, br, NH). 13C NMR (176 MHz, DMSO-d6)
δ 33.0 (Ce), 45.9 (CH2CH2CO2Me), 47.5 (Ca), 51.5 (CH3), 54.4
(CH2CH2CO2Me), 58.9 (CN3), 167.7, 168.3 (CvC), 171.1, 180.9,
Methyl 3-((2-azido-3-((tert-butoxycarbonyl)amino)propyl)-
(2-ethoxy-3,4-dioxocyclobut-1-en-1-yl)amino)propanoate, 10.
To a solution of 3,4-diethoxy-3-cyclobutene-1,2-dione (102 μL, 181.0 (CO). MS (ES+) m/z 280.1 [M + H]+; C11H14N5O4 requires
0.68 mmol) in anhydrous ethanol (1 mL), was added a solution 280.1046; found 280.1067.
9398 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Methyl 3-(4-amino-8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)- (CH2), 61.7 (d, 2J = 7, P(OCH2CH3)2), 61.9 (CN3), 79.9 (C(CH3)3),
en-2-yl)propanoate, 18.
156.1 (CO). 31P NMR (CDCl3, 162 MHz) δ 32.01. MS (ES+) m/z
394.2 [M + H]+; C15H32N5O5P requires 394.2219; found 394.2215.
tert-Butyl(2-azido-3-((3-(diethoxyphosphoryl)propyl)(2-ethoxy-
3,4-dioxocyclobut-1-en-1-yl)amino)propyl)carbamate, 12.
To a solution of methyl 3-(4-azido-8,9-dioxo-2,6-diazabicyclo-
[5.2.0]non-1(7)-en-2-yl)propanoate (38 mg, 0.14 mmol) in an-
hydrous THF (3 mL) was added H2O (100 µL) and PPh3 (54 mg,
0.20 mmol). The suspension was stirred under argon at 60 °C
and over time, became a clear solution. After stirring for
16 hours, the solvent was removed under reduced pressure
and the crude residue partitioned between DCM and H2O
(10 mL, 1 : 1). The aqueous layer was separated and washed
twice with DCM (10 mL), before lyophilisation yielded a white
To a solution of 3,4-diethoxy-3-cyclobutene-1,2-dione (152 μL,
1.03 mmol) in anhydrous ethanol (1 mL), was added a solution
of tert-butyl (2-azido-3-((3-(diethoxyphosphoryl)propyl)amino)-
propyl)carbamate (396 mg, 1.0 mmol) in anhydrous ethanol
(7 mL) over 30 min. The resulting solution was stirred under
argon at room temperature and the progress of the reaction
followed by TLC. After no further reaction was observed by
TLC, the solution was concentrated under reduced pressure.
The crude residue was then purified by column chromato-
graphy (DCM–MeOH, 100% to 95 : 5 using 0.5% increments; Rf =
0.44) to give a colourless oil, existing as a pair of rotamers
1
solid (34 mg, 99%), m.p. >250 °C. H NMR (700 MHz, DMSO-
d6) δ 1.78–2.18 (2H, br, NH2), 2.69 (2H, overlapping dt, J = 14,
7, CH2CO2Me), 3.11 (2H, m, Ha), 3.21 (1H, dd, J = 14, 7,
Hc(axial)), 3.27–3.31 (1H, m, Hb), 3.36 (1H, d, J = 14, Hc(eq)),
3.59 (3H, s, CH3), 3.87 (1H, dt, J = 14, 7, CH2CH2CO2Me), 3.92
(1H, dt, J = 14, 7, CH2d′), 8.53 (1H, br, NH). 13C NMR
(176 MHz, DMSO-d6)
δ
32.9 (CH2CH2CO2Me), 46.2
1
(417 mg, 79%). H NMR (700 MHz, CDCl3) δ 1.32 (6H, t, J = 7,
(CH2CH2CO2Me), 51.5 (CH3), 51.6 (CH), 51.8 (CH2), 58.8 (CH2),
167.3, 168.0 (CvC), 171.3, 180.1, 181.2 (CO). MS (ES+) m/z
254.1 [M + H]+; C11H16N3O4 requires 254.1141; found 254.1166.
tert-Butyl(2-azido-3-((3-(diethoxyphosphoryl)propyl)amino)-
propyl)carbamate, 8.
P(OCH2CH3)2), 1.44 (9H, s, C(CH3)3), 1.46 (3H, t, J = 7,
OCH2CH3), 1.65–1.96 (4H, m, PCH2CH2CH2), 3.11–3.42 (2H,
m, CH2), 3.48–3.89 (5H, m, CH2CH + PCH2CH2CH2), 4.03–4.15
(4H, m, P(OCH2CH3)2), 4.75–4.80 (2H, m, OCH2CH3), 5.04 (1H,
br, NH). 13C NMR (175 MHz, CDCl3) δ 16.0 (OCH2CH3), 16.6
3
2
(d, J = 6, P(OCH2CH3)2), 22.3 (d, J = 5, NCH2CH2CH2P), 22.6
1
(d, J = 143, NCH2CH2CH2P), 28.4 (C(CH3)3), 42.2 (CH2), 50.6
(d, 3J = 18, NCH2CH2CH2P), 50.8 (CH2), 61.1 (CN3), 62.0 (d, 2J =
7, P(OCH2CH3)2), 70.2 (OCH2CH3), 80.3 (C(CH3)3), 156.0 (CO),
172.7, 177.2 (CvC), 182.9, 188.5 (CO). 31P NMR (CDCl3,
283 MHz) δ 30.14. MS (ES+) m/z 518.1 [M + H]+; C21H37N5O8P
requires 518.2380; found 518.2375.
A
solution of tert-butyl (3-amino-2-azidopropyl)carbamate
(395 mg, 1.84 mmol), sodium carbonate (293 mg, 2.76 mmol)
and diethyl-3-bromopropyl phosphonate (560 μL, 2.91 mmol)
in anhydrous ethanol (8 mL) was boiled under reflux for
16 hours until no further reaction was observed by TLC. The
mixture was concentrated under reduced pressure before being
partitioned between DCM–H2O (1 : 1, 30 mL). The organic
portion was separated and the aqueous extracted with dichloro-
methane (3 × 25 mL). The combined organic fractions were
dried over MgSO4, filtered and concentrated, allowing the
crude residue to be purified by column chromatography
Diethyl-3-(4-azido-8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-
en-2-yl)propyl)phosphonate, 16.
(DCM–MeOH, 100% to 92 : 8 using 1% increments; Rf = 0.35) To a solution of tert-butyl (2-azido-3-((3-(diethoxyphosphoryl)-
to give the ester as a colourless oil (396 mg, 55%). 1H NMR propyl)(2-ethoxy-3,4-dioxocyclobut-1-en-1-yl)amino)propyl)car-
(400 MHz, CDCl3) δ 1.31 (6H, t, J = 7, P(OCH2CH3)2), 1.43 (9H, bamate (201 mg, 0.39 mmol) in anhydrous dichloromethane
s, C(CH3)3), 1.71–1.86 (5H, m, PCH2CH2CH2 + NH), 2.66 (1H, (1 mL) was added TFA (1 mL), and the resulting solution
dd, J = 12, 7, CHH), 2.69 (2H, m, PCH2CH2CH2), 2.71 (1H, dd, stirred at room temperature for 1 hour. Reaction was complete
J = 12, 5, CHH), 3.17 (1H, dt, J = 14, 7, CHH), 3.33 (1H, dt, J = after this period, as indicated by TLC, and the solvents were
3
14, 5, CHH), 3.63–3.72 (1H, m, CH), 4.02–4.15 (4H, qd, JH–H
=
removed under reduced pressure. The residue was re-dissolved
7, JH–P = 3, P(OCH2CH3)2), 5.01 (1H, br, NH). 13C NMR in dichloromethane and again reduced under reduced
3
3
(101 MHz, CDCl3) δ 16.6 (d, J = 6, P(OCH2CH3)2), 23.0 (d, J = pressure. This process was repeated 5 times to ensure com-
5, NCH2CH2CH2P), 23.4 (d, 1J = 143, NHCH2CH2P), 28.5 plete removal of excess TFA. The TFA salt was dissolved in
(C(CH3)3), 42.6 (CH2), 50.0 (d, 2J = 18, NCH2CH2CH2P), 50.8 anhydrous ethanol (5 mL) to which a solution of triethylamine
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9399

View Article Online
Paper
Organic & Biomolecular Chemistry
(218 μL, 1.56 mmol) in anhydrous ethanol (2 mL) was added A solution of (R)-5-benzyl 1-tert-butyl 2-[4,7,10-tris(2-tert-butoxy-
over a 20 minute period. The resulting solution was heated to 2-oxoethyl)-1,4,7,10-tetraazacyclododecan-1-yl]pentanedioate3
reflux for 18 hours, until no further reaction was observed by (677 mg, 0.86 mmol) and Pd-C (10%) in MeOH (10 mL), was
TLC. At this point, the solvent was removed under reduced agitated at room temperature, under H2 (40 psi), in a Parr appar-
pressure and the crude residue purified by column chromato- atus for 6 h. The reaction mixture was filtered through Celite
graphy (DCM–MeOH, 100% to 90 : 10 using 1% increments; Rf = and the filtrate evaporated under reduced pressure, to give the
1
0.26) to give a colourless oil (126 mg, 87%). 1H NMR acid as a hygroscopic solid (530 mg, 88%). H NMR (400 MHz,
(400 MHz, CDCl3) δ 1.30 (6H, t, J = 7, P(OCH2CH3)2), 1.69–2.01 CDCl3) δ 1.27 (9H, s, C(CH3)3), 1.28 (18H, s, C(CH3)3), 1.29 (9H,
(4H, m, PCH2CH2CH2), 3.49 (1H, dd, J = 13, 7, Hc(axial)), s, C(CH3)3), 1.86–2.16 (4H, m, CH2CH2CO), 2.18–3.18 (16H, br.
3.52–3.62 (3H, m, Hc/a(eq), Ha(axial)), 3.71–4.11 (7H, m, CH + m, CH2), 3.23 (4H, s, COCH2), 3.24 (2H, s, COCH2), 3.27–3.35
NCH2CH2CH2P + P(OCH2CH3)2), 8.16 (1H, br, s, NH).). 13C (1H, m, COCH), 3.67 (1H, s, OH). 13C NMR (101 MHz, CDCl3)
3
NMR (101 MHz, CDCl3) δ 16.5 (d, J = 6, P(OCH2CH3)2), 21.8 δ 27.6, 27.7, 27.9, 28.0 (C(CH3)3), 33.7, 49.7, 52.5, 55.3, 55.6, 56.0,
2
1
(d, J = 5, NCH2CH2CH2P), 22.2 (d, J = 143, NCH2CH2CH2P), 60.1 (CH2), 69.6 (CH), 81.7, 82.0, 82.5 (C(CH3)3), 171.2, 172.6,
48.5 (CH2), 51.3 (d, J = 18, NCH2CH2CH2P), 55.3 (CH2), 59.5 175.2, 176.0 (CO). MS (ES+): m/z 701.1 [M + H]+; C35H65N4O10
3
(CN3), 62.0 (d, 2J = 7, P(OCH2CH3)2), 167.6, 168.6 (CvC), 180.9, requires 701.4701; found 701.4704.
182.2 (CO). 31P NMR (CDCl3, 162 MHz) δ 30.80. MS (ES+) m/z
(R)-4-((6-tert-Butoxy)-6-oxo-5-(4,7,10-tris(2-(tert-butoxy)-
372.1 [M + H]+; C14H23N5O5P requires 372.1437; found 372.1436. 2-oxoethyl)-1,4,7,10-tetraazacyclododecan-1-yl)hexyl)amino)-
Diethyl-3-(4-amino-8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-
en-2-yl)propyl)phosphonate, 20.
4-oxobutanoic acid, 22.
To a solution of diethyl (3-(4-azido-8,9-dioxo-2,6-diazabicyclo-
[5.2.0]non-1(7)-en-2-yl)propyl)phosphonate (179 mg, 0.48
mmol) in anhydrous THF (10 mL) was added H2O (250 µL)
and PPh3 (189 mg, 0.72 mmol). The suspension was
stirred under argon at 60 °C and over time, became a clear
solution. After stirring for 16 hours, the solvent was removed
under reduced pressure and the crude residue partitioned
between DCM and H2O (20 mL, 1 : 1). The aqueous layer was
separated and washed twice with DCM (10 mL), before lyophi-
lisation yielded a white solid (155 mg, 94%), m.p. >250 °C.
1H NMR (700 MHz, DMSO-d6) δ 1.23 (6H, t, J = 7 P(OCH2CH3)2),
1.72–1.81 (4H, m, PCH2CH2CH2), 3.23–3.38 (2H, br, NH2),
To a pre-stirred solution of (R)-tri-tert-butyl 2,2′,2″-(10-(6-
amino-1-(tert-butoxy)-1-oxohexan-2-yl)-1,4,7,10-tetraazacyclodo-
decane-1,4,7-triyl)triacetate7 (325 mg, 0.46 mmol) and diiso-
propylethylamine (160 μL, 0.92 mmol) in anhydrous DMF
(3 mL), was added succinic anhydride (46 mg, 0.46 mmol) and
the resulting solution stirred at room temperature and the
reaction monitored by ESI-MS. After 16 hours, the solvent was
removed under reduced pressure and the crude residue puri-
fied by column chromatography (DCM–MeOH, 100% DCM to
87 : 13; Rf = 0.32) to give a yellow gum (228 mg, 62%). 1H NMR
(600 MHz, CDCl3) δ 1.34 (9H, s, C(CH3)3), 1.35 (18H, s,
C(CH3)3), 1.36 (9H, s, C(CH3)3), 1.41–1.64 (6H, m), 1.96–2.06
(3H, m), 2.11–2.27 (4H, m), 2.37–2.54 (7H, m), 2.65–2.80 (6H, m),
2.83–2.96 (3H, m), 3.10–3.20 (2H, m), 3.21–3.31 (3H, m),
3.32–3.33 (1H, m), 8.27–8.33 (1H, br. s, OH), 8.85 (1H, br.s, NH).
13C NMR (151 MHz, CDCl3) δ 24.5 (CH2), 26.5, 27.7, 27.8, 27.8
(C(CH3)3), 28.8, 30.0, 31.8, 39.1, 44.5, 47.1, 48.0, 48.3, 48.4, 52.4,
52.5, 53.5, 55.4, 55.6, 55.7 (CH2), 61.1 (CH), 81.9, 82.0, 82.0
(C(CH3)3), 172.5, 172.7, 174.4, 174.8, 175.1 (CO). MS (ES+) m/z
800.4 [M + H]+; C40H74N5O11 requires 800.5385; found 800.5366.
[Conjugate 2].
3
3.39–3.77 (7H, m, CH2 + CH), 3.95–4.01 (4H, qd, JH–H = 7,
3JH–P = 3, P(OCH2CH3)2), 8.57 (1H, s, NH). 13C NMR (176 MHz,
DMSO-d6) δ 16.3 (d, 3J = 6, P(OCH2CH3)2), 21.3 (d, 2J = 5,
1
NCH2CH2CH2P), 21.5 (d, J = 143, NCH2CH2CH2P), 47.9, 50.3
3
2
(CH), 50.7 (d, J = 18, NCH2CH2CH2P), 54.8 (CH2), 61.0 (d, J =
7, P(OCH2CH3)2), 167.9, 168.4 (CvC), 181.2, 181.4 (CO).
31P NMR (162 MHz, DMSO-d6) δ 31.27. MS (ES+) m/z 346.1
[M + H]+; C14H25N3O5P requires 346.1532; found 346.1550.
(R)-5-tert-Butoxy-5-oxo-4-[4,7,10-tris(2-tert-butoxy-2-oxoethyl)-
1,4,7,10-tetraazacyclododecan-1-yl]pentanoic acid.15 21.
9400 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
(R)-5-tert-Butoxy-5-oxo-4-[4,7,10-tris(2-tert-butoxy-2-oxoethyl)- complete consumption of the starting materials was revealed
1,4,7,10-tetraazacyclododecan-1-yl]pentanoic acid (94 mg, by ESI-MS. After this period, the solvent was removed under
0.13 mmol), EDC (31 mg, 0.16 mmol) and HOBt (22 mg, reduced pressure and the crude oil taken up into EtOAc
0.16 mmol) were dissolved in anhydrous DMF (2 mL) and (30 mL). NaHCO3 (30 mL) was added, the layers separated and
stirred at room temperature under an atmosphere of argon for the aqueous washed with EtOAc (3 × 50 mL). The combined
20 minutes. After this period, a pre-stirred solution of methyl organic portions were dried over MgSO4, filtered and the
3-(4-amino-8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl)- solvent removed under reduced pressure. The crude residue
propanoate (34 mg, 0.13 mmol) and NMM (30 μL, 0.27 mmol) was purified by column chromatography (DCM–MeOH, 100%
in anhydrous DMF (1.5 mL) was added dropwise and the to 90 : 10 in 1% increments; Rf = 0.41) to yield a dark yellow
resulting solution stirred at room temperature until complete viscous oil. This product was characterized as a pair of diastereo-
consumption of the starting materials was revealed by ESI-MS. isomers (57 mg, 28%). 1H NMR (700 MHz, CDCl3) δ 1.27
After this period, the solvent was removed under reduced (6H, t, J = 7, P(OCH2CH3)2), 1.40, 1.40, 1.41, 1.42 (36H, s,
pressure and the crude oil taken up into EtOAc (20 mL). C(CH3)3), 1.66–1.74 (2H, m, PCH2CH2CH2), 1.81–1.89 (3H, m,
NaHCO3 (20 mL) was added, the layers separated and the PCH2CH2CH2 + CHH), 1.96–2.09 (4H, m), 2.15–2.23 (3H, m),
aqueous fraction washed with EtOAc (3 × 40 mL). The com- 2.32–2.40 (1H, m), 2.45–2.56 (4H, m), 2.59–2.64 (3H, m),
bined organic portions were dried over MgSO4, filtered and the 2.70–2.76 (2H, m), 2.76–2.82 (2H, m), 2.87–2.97 (3H, m),
solvent removed under reduced pressure. The crude residue 3.28–3.35 (2H, m), 3.36–3.38 (1H, m), 3.40–3.44 (1H, m),
was purified by column chromatography (DCM–MeOH, 100% 3.46–3.54 (2H, m), 3.61–3.68 (1H, m), 3.72–3.77 (1H, m, Hc)
to 90 : 10 in 1% increments; Rf = 0.37) to yield a pale brown 3.79–3.87 (2H, m, PCH2CH2CH2), 3.99–4.07 (4H, m,
viscous oil. This product was characterized as a pair of diaster- P(OCH2CH3)2), 4.13–4.19 (1H, m, Hb), 7.56 (1H, br, NH), 8.85
eoisomers (33 mg, 26%). 1H NMR (600 MHz, CDCl3) (1H, br, NH). 13C NMR (176 MHz, CDCl3) δ 16.6 (d, 3J = 6,
δ 1.42–1.44 (36H, overlapping s, C(CH3)3), 1.82–1.93 (1H, m), P(OCH2CH3)2), 22.0 (d, 2J = 5, NCH2CH2CH2P), 22.9 (d, 1J = 143,
1.95–2.11 (4H, m), 2.17–2.27 (3H, m), 2.34–2.41 (1H, m), NCH2CH2CH2P), 27.9, 28.0 (overlapping C(CH3)3), 35.3, 38.7,
2.49–2.58 (4H, m), 2.60–2.70 (5H, m, CH2 + CH2), 2.71–2.78 44.5, 47.1, 48.4, 50.2 (CHNH), 51.5 (d, 3J = 16, NCH2CH2CH2P),
e
(2H, m), 2.78–2.85 (2H, m), 2.89–2.99 (3H, m), 3.29–3.38 (2H, 52.8, 55.6, 55.9, 56.2, 60.7 (CH), 61.7 (overlapping d, 2J = 7,
m), 3.40–3.52 (4H, m, Ha + CH), 3.64 (3H, s, CH3), 3.77–3.87 P(OCH2CH3)2), 82.0, 82.1, 82.1, 82.3 (C(CH3)3), 167.4,
(2H, m, Hc), 3.94 (1H, dt, J = 14, 7, CH2d), 4.03 (1H, dt, J = 14, 167.8 (CvC), 172.6, 172.8, 172.9, 173.4, 175.3, 181.4,
7, CH2d′), 4.13–4.19 (1H, m, CHNH), 7.72 (1H, br, NH), 8.70 181.8 (CO). 31P NMR (283 MHz, CDCl3) δ 30.85. MS (ES+) m/z
(1H, br, NH). 13C NMR (151 MHz, CDCl3) δ 27.9, 28.0 (overlap- 1028.5 [M + H]+; C49H87N7O14P requires 1028.605; found
ping C(CH3)3), 33.8 (Ce), 35.3, 45.6, 47.2 (Cd), 48.2, 48.7, 49.8, 1028.606.
50.2 (CHNH), 51.2 (CH3), 52.8, 55.6, 55.9, 56.4, 60.8 (CH), 82.0,
82.1, 82.1, 82.2 (C(CH3)3), 167.7, 167.9 (CvC), 171.4, 172.6,
172.8, 172.9, 173.5, 175.3, 181.2, 182.0 (CO). MS (ES+) m/z
936.6 [M + H]+; C46H78N7O13 requires 936.5658; found 936.5672.
[Conjugate 4].
[Gd·L2].
The [Conjugate 2] (10 mg, 0.01 mmol) was dissolved in MeOH
(200 µL) with stirring. To this was added NaOH (0.5 mg) as a
solution in H2O (2.5 M) and the resulting solution stirred at
room temperature overnight. Complete removal of the methyl
ester was verified by ESI-MS, at which point the solvent was
removed under reduced pressure. The residue was completely
(R)-5-tert-Butoxy-5-oxo-4-[4,7,10-tris(2-tert-butoxy-2-oxoethyl)- dried under high vacuum before being suspended in DCM
1,4,7,10-tetraazacyclododecan-1-yl]pentanoic acid (138 mg, (1 mL), to which trifluoroacetic acid (1 mL) was added. The
0.20 mmol), EDC (45 mg, 0.24 mmol) and HOBt (32 mg, resulting yellow solution was stirred at room temperature over-
0.24 mmol) were dissolved in anhydrous DMF (3 mL) and night. Complete removal of the tert-butyl ester groups was veri-
stirred at room temperature under an atmosphere of argon for fied by ESI-MS, at which point excess solvent was removed
20 minutes. After this period, a pre-stirred solution of diethyl under reduced pressure. The residue was repeatedly re-dis-
(3-(4-amino-8,9-dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-2-yl)- solved in DCM (2 mL) and the solvent removed under reduced
propyl)phosphonate (68 mg, 0.20 mmol) and NMM (43 μL, pressure to remove excess TFA. This process yielded the proto-
0.39 mmol) in anhydrous DMF (2 mL) was added dropwise nated salt of L2 as a pale-brown solid. MS (ES+) m/z 698.3 [M +
and the resulting solution stirred at room temperature until H]+. The salt of L2 (7.5 mg, 0.01 mmol) was dissolved in H2O
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9401

View Article Online
Paper
Organic & Biomolecular Chemistry
(0.5 mL) and the pH adjusted to about 5.5 by the addition
of NaOH (0.1 M). GdCl3·6H2O (4.8 mg, 0.013 mmol) was
added as a solution in H2O (0.5 mL) and the reaction mixture
stirred at 60 °C overnight. The pH of the solution was periodi-
cally checked and maintained between 5–6 with the addition
of NaOH–HCl (0.1 M). Upon completion of complexa-
tion, excess gadolinium was removed by the addition of
Chelex-100™ with stirring. The Chelex trap was filtered
and the complex eluted with excess H2O. Removal of the water
by lyophilisation gave a white solid that was purified by
RP-HPLC. HR-MS (ES+) C29H41154GdN7O13 requires 849.1971
[M + 2H]+; found 849.1956. r1p = 4.54 mM−1 s−1 (60 MHz,
310 K). RP-HPLC: tR = 7.1 min [2–30% MeOH in H2O over
10 min].
[Gd·L6].
[Conjugate 6] (10 mg, 0.01 mmol) was dissolved in DCM
(1 mL) with stirring. To this solution was added trifluoroacetic
acid (1 mL) and the resulting solution stirred at room tempera-
ture for overnight. Hydrolysis of the tert-butyl ester groups was
verified by ESI-MS, at which point the excess solvent was
removed under reduced pressure. The residue was repeatedly
re-dissolved in DCM (2 mL) and the solvent removed under
reduced pressure to remove excess TFA. This process yielded
the phosphonate ethyl ester as a light-brown solid. This
residue was dissolved in DMF (1 mL) to which bromotrimethyl-
silane (10 μL, 0.07 mmol) was added dropwise. The resulting
mixture was heated to 60 °C overnight until complete hydro-
lysis was indicated by ESI-MS. The solvent was removed under
reduced pressure, before the residue re-dissolved in H2O. The
pH was adjusted to 6 and the aqueous phase washed with
DCM (3 × 3 mL) and diethyl ether (3 × 3 mL). The aqueous
solvent was then removed by lyophilisation to give the proto-
nated salt of L12 as a light brown solid. MS (ES+) m/z 847.8
[M + H]+. The salt of L12 (7.6 mg, 0.01 mmol) was dissolved in
H2O (1 mL) and the pH adjusted to 6. GdCl3·6H2O (4.0 mg,
0.011 mmol) was added as a solution in H2O (0.5 mL) and the
reaction mixture stirred at 60 °C overnight. The pH of the solu-
tion was periodically checked and maintained between 5 and 6
with the addition of NaOH–HCl (0.1 M). Upon completion of
complexation, excess gadolinium was removed by the addition
of Chelex-100™ with stirring. The Chelex trap was filtered and
the complex eluted with excess H2O. Removal of the water by
lyophilisation gave the complex as a white solid that was puri-
fied by RP-HPLC. HR-MS (ES+) C34H53158GdN8O15P requires
1002.261 [M + 2H]+; found 1002.267. r1p = 5.18 mM−1 s−1
(60 MHz, 310 K). RP-HPLC: tR = 7.8 min [2–30% MeOH in H2O
over 10 min].
[Gd·L4].
[Conjugate 4] (17 mg, 0.02 mmol) was dissolved in DCM
(1 mL) with stirring. To this solution was added trifluoroacetic
acid (1 mL) and the resulting solution stirred at room tempera-
ture for overnight. Hydrolysis of the tert-butyl ester groups was
verified by ESI-MS, at which point the excess solvent was
removed under reduced pressure. The residue was repeatedly
re-dissolved in DCM (2 mL) and the solvent removed under
reduced pressure to remove excess TFA. This process yielded
the phosphonate ethyl ester as a light-brown solid. This
residue was dissolved in DMF (1 mL) to which bromotrimethyl-
silane (18 μL, 0.13 mmol) was added dropwise. The resulting
mixture was heated to 60 °C overnight until complete hydro-
lysis occurred as indicated by ESI-MS. The solvent was
removed under reduced pressure, before the residue re-dis-
solved in H2O. The pH was adjusted to 6 and the aqueous
phase washed with DCM (3 × 3 mL) and diethyl ether (3 ×
3 mL). The aqueous solvent was then removed by lyophilisa-
tion to give the protonated salt of L4 as a light brown solid. MS
(ES+) m/z 748.7 [M + H]+. The salt of L4 (12.5 mg, 0.017 mmol)
was dissolved in H2O (0.5 mL) and the pH adjusted to 5.5.
GdCl3·6H2O (6.8 mg, 0.018 mmol) was added as a solution in
H2O (0.5 mL) and the reaction mixture stirred at 60 °C over-
night. The pH of the solution was periodically checked and
maintained between 5 and 6 with the addition of NaOH–HCl
(0.1 M). Upon completion of complexation, excess gadolinium
was removed by the addition of Chelex-100™ with stirring. The
Chelex trap was filtered and the complex eluted with excess
H2O. Removal of the water by lyophilisation gave the complex,
Compound 26.
[Gd·L4] as a white solid that was purified by RP-HPLC. HR-MS (2R)-1-tert-Butyl 5-methyl 2-(4,10-bis(2-(tert-butoxy)-2-oxoethyl)-
(ES+) C29H44154GdN7O14P requires 899.1893 [M + 2H]+; found 7-((2R)-1-(tert-butoxy)-6-(4-((2-(2-(diethoxyphosphoryl)ethyl)-8,9-
899.1891. r1p = 5.21 mM−1 s−1 (60 MHz, 310 K). RP-HPLC: tR
7.5 min [2–30% MeOH in H2O over 10 min].
=
dioxo-2,6-diazabicyclo[5.2.0]non-1(7)-en-4-yl)amino)-4-oxobuta-
namido)-1-oxohexan-2-yl)-1,4,7,10-tetraazacyclododecan-1-yl)-
9402 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
pentanedioate (50 mg, 0.042 mmol) was dissolved in methanol H2O. The pH was adjusted to 6 and the aqueous phase washed
(1 mL) and NaOH (2 mg, 0.051 mmol) was added as a solution with DCM (3 × 2 mL) and diethyl ether (3 × 2 mL). The
in H2O (250 μL). The light yellow solution was stirred at room aqueous solvent was then removed by lyophilisation to give the
temperature and progress of the reaction monitored by protonated salt of L7 as a light brown solid. MS (ES+) m/z 587.0
ESI-MS. Upon complete hydrolysis of the methyl ester after [M + 2H]2+. The salt of L7 (8.1 mg, 0.007 mmol) was dissolved
1 hour, the solvent was removed under reduced pressure to in H2O (0.5 mL) and the pH adjusted to 6. GdCl3·6H2O
give a pale yellow solid, which was dried fully on a high (3.1 mg, 0.008 mmol) was added as a solution in H2O (0.5 mL)
vacuum line. This yielded the carboxylic acid (49 mg, 0.042), to and the reaction mixture stirred at 60 °C overnight. The pH of
which EDC (9.6 mg, 0.050 mmol) and HOBt (6.8 mg, the solution was periodically checked and maintained between
0.050 mmol) were added and dissolved in anhydrous DMF 5 and 6 with the addition of NaOH–HCl (0.1 M). Upon
(1 mL). The mixture was stirred at room temperature under an completion of complexation, excess gadolinium was removed
atmosphere of argon for 20 minutes. After this period, a pre- by the addition of Chelex-100™ with stirring. The Chelex
stirred solution of N-(2-aminoethyl)-5-((3aS, 4S, 6aR)-2-oxohexa- trap was filtered and the complex eluted with excess H2O.
hydro-1H-thieno[3,4-d]imidazol-4-yl)pentanamide16 (12 mg, Removal of the water by lyophilisation gave the complex as a
0.042 mmol) and NMM (9 μL, 0.083 mmol) in anhydrous DMF white solid that was purified by RP-HPLC. HR-MS (ES+)
(0.5 mL) was added dropwise and the resulting solution stirred C48H75154GdN12O18PS requires 1324.399 [M + 2H]+; found
at room temperature until complete consumption of the start- 1324.399. r1p = 7.23 mM−1 s−1 (60 MHz, 310 K). RP-HPLC: tR
ing materials was revealed by ESI-MS. After this period, the 16.7 min [2–30% MeOH in H2O over 10 min].
solvent was removed under reduced pressure and the crude oil
=
taken up into EtOAc (15 mL). NaHCO3 (15 mL) was added, the
layers separated and the aqueous washed with EtOAc (3 ×
20 mL). The combined organic portions were dried over
MgSO4, filtered and the solvent removed under reduced
Acknowledgements
This work was supported by the ERC (DP, NS, AM; FCC
266804) and the Ministry for Education and Research, BMBF
[FKZ:01EZ0813 (SG)], and the Max Planck Society.
pressure. The crude residue was purified by column chromato-
graphy (DCM–MeOH–NH4OH, 100% to 80 : 15 : 5; Rf = 0.15) to
yield a yellow oil (10 mg, 17%). 1H NMR (700 MHz, CDCl3)
δ 1.29–1.37 (6H, m, P(OCH2CH3)2), 1.40–1.51 (40H, m),
1.56–1.75 (5H, m), 2.22–2.80 (20H, m), 2.80–3.51 (31H, m),
3.60–3.93 (5H, m), 5.05–4.15 (4H, m, P(OCH2CH3)2), 4.21–4.52
(3H, m), 7.87–8.51 (3H, m). 31P NMR (283 MHz, CDCl3)
δ 27.17. MS (ES+) m/z 727.4 [M + 2H]2+; C68H118N12O18PS
requires 1453.815; found 1453.819.
Notes and references
1 (a) P. Paoletti, C. Bellone and Q. Zhou, Nat. Rev. Neurosci.,
2013, 14, 383; (b) M. Sheng, J. Cummings, L. A. Roldan,
Y. N. Jan and L. Y. Jan, Nature, 1994, 368, 144.
2 S. Cull-Candy, S. Brickley and M. Farranti, Curr. Opin. Neuro-
biol., 2001, 11, 327.
[Gd·L7].
3 E. A. Waxman and D. R. Lynch, Neuroscientist, 2005, 11, 37.
4 I. Ahmed, S. K. Bose, N. Pavese, A. Ramlackhansingh,
F. Turkheimer, G. Hotton, A. Hammers and D. J. Brooks,
Brain, 2011, 134, 979.
5 S. X. Cai, Curr. Top. Med. Chem., 2006, 6, 651.
6 W. E. Childers and R. B. Baudy, J. Med. Chem., 2007, 50,
2557.
Compound 26 (10 mg, 0.007 mmol) was dissolved in DCM
(1 mL) with stirring. To this solution was added trifluoroacetic
acid (1 mL) and the resulting solution stirred at room tempera-
ture for overnight. Hydrolysis of the tert-butyl ester groups was
verified by ESI-MS, at which point the excess solvent was
removed under reduced pressure. The residue was repeatedly
re-dissolved in DCM (2 mL) and the solvent removed under
7 A. Mishra, S. Gottschalk, J. Engelmann and D. Parker,
Chem. Sci., 2012, 3, 131.
8 S. Gottschalk, J. Engelmann, G. Rolla, M. Botta, D. Parker
and A. Mishra, Org. Biomol. Chem., 2013, 11, 6131.
9 A. Mishra, R. Mishra, S. Gottschalk, R. Pal, N. Sim,
J. Engelmann, M. Goldberg and D. Parker, ACS Chem.
Neurosci., 2014, 5, 128.
reduced pressure to remove excess TFA. This process yielded 10 S. A. Hires, Y. Zhu and R. Y. Tsien, Proc. Natl. Acad.
the phosphonate ethyl ester as a light-brown solid. This Sci. U. S. A., 2008, 105, 4411.
residue was dissolved in DMF (0.8 mL) to which bromotri- 11 (a) W. A. Kinney, N. E. Lee, D. T. Garrison, E. J. Podlesny,
methylsilane (7 μL, 0.055 mmol) was added dropwise. The
resulting mixture was heated to 60 °C overnight until complete
hydrolysis as verified by ESI-MS. The solvent was removed
under reduced pressure, before the residue re-dissolved in
J. T. Simmonds, D. Bramlett, R. R. Notvest, D. M. Kowal
and R. P. Tasse, J. Med. Chem., 1992, 35, 4720;
(b) W. A. Kinney, M. Abou-Gharbia, D. T. Garrison,
J. Schmid, D. M. Kowal, D. R. Bramlett, T. L. Miller,
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9389–9404 | 9403

View Article Online
Paper
Organic & Biomolecular Chemistry
R. P. Tasse, M. M. Zaleska and J. A. Moyer, J. Med. Chem., 17 The Chemistry of Contrast Agents in Medical Magnetic Reson-
1998, 41, 236.
ance Imaging, ed. A. E. Merbach and É. Tóth, Wiley,
12 P. Caravan and Z. Zhang, Eur. J. Inorg. Chem., 2012, 1916.
New York, Chichester, 2001.
13 N. Sim, S. Gottschalk, R. Pal, J. Engelmann, D. Parker and 18 C. J. Eggett, S. Crosier, P. Manning, M. R. Cookson,
A. Mishra, Chem. Sci., 2013, 4, 3148.
F. M. Menzies, C. J. McNeil and P. J. Shaw, J. Neurochem.,
2000, 74, 1895.
19 J. Carmichael, W. G. Degraff, A. F. Gazdar, J. D. Minna and
J. B. Mitchell, Cancer Res., 1987, 47, 936.
20 P. Marchand, J. Becerril-Ortega, L. Mony, C. Bouteiller,
P. Paoletti, O. Nicole, L. Barré, A. Buisson and C. Perrio,
Bioconjugate Chem., 2011, 23, 21.
14 (a) S. Corporation, WO2006/19768 A1; 2006; (b) L. E. Overman
and M. J. Sharp, Tetrahedron Lett., 1988, 29, 901.
15 K.-P. Eisenwiener, P. Powell and H. R. Mäcke, Bioorg. Med.
Chem. Lett., 2000, 10, 2133.
16 A. Eisenführ, P. S. Arora, G. Sengle, L. R. Takaoka, J. S. Nowick
and M. Famulok, Bioorg. Med. Chem., 2003, 11, 235.
9404 | Org. Biomol. Chem., 2014, 12, 9389–9404
This journal is © The Royal Society of Chemistry 2014