Calcium-responsive paramagnetic CEST agents

Article abstract of DOI:10.1016/j.bmc.2010.07.023

The assessment of changes in the extracellular calcium concentration by magnetic resonance imaging would be a valuable biomedical research tool to monitor brain neuronal activity. In this perspective, we report here the synthesis of novel ligands consisti

Full text of DOI:10.1016/j.bmc.2010.07.023

Bioorganic & Medicinal Chemistry 19 (2011) 1097–1105
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Calcium-responsive paramagnetic CEST agents
b
c
Goran Angelovski a, , Thomas Chauvin , Rolf Pohmann , Nikos K. Logothetis a,d, Éva Tóth b,
*
*
a Physiology of Cognitive Processes, Max Planck Institute for Biological Cybernetics, Spemannstr. 38, 72076 Tübingen, Germany
b Centre de Biophysique Moléculaire, CNRS, rue Charles Sadron, 45071 Orléans, France
c High-Field Magnetic Resonance Center, Max Planck Institute for Biological Cybernetics, Spemannstr. 38, 72076 Tübingen, Germany
d Imaging Science and Biomedical Engineering, University of Manchester, Oxford Road, Manchester M13 9PT, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 15 July 2010
The assessment of changes in the extracellular calcium concentration by magnetic resonance imaging
would be a valuable biomedical research tool to monitor brain neuronal activity. In this perspective, we
report here the synthesis of novel ligands consisting of tetraamide and bisamide derivatives of cyclen,
L
1 and L2, respectively, each bearing imino(diacetate) moieties for Ca2+ binding. Yb3+ and Eu3+ complexes
Keywords:
MRI
PARACEST
Molecular imaging
Calcium sensitive
Proton exchange
are investigated as chemical exchange saturation transfer (CEST) agents that respond to the presence of
Ca2+. A CEST effect is observed for both YbL1 and EuL1 complexes (B = 11.7 T), originating from the slow
exchange of the amide protons and those of the coordinated water, respectively, whilst no CEST is detected
for complexes of L2. Upon calcium binding, the CEST effect decreases considerably (from 60% to 20% for
YbL1 and from 35% to 10% for EuL1). A similar variation is observed in the presence of Mg2+. The affinity
constants between the lanthanide complexes and the alkaline earth metal ions have been estimated from
YbL1—Ca
aff
YbL1—Mg
aff
EuL1—Ca
aff
the variation of the CEST effect to be K
¼ 8 ꢀ 2Mꢁ1, K
¼ 23 ꢀ 3Mꢁ1 and K
¼ 10 ꢀ 3Mꢁ1
.
These low values imply the coordination of the alkaline earth ions to a single iminodiacetate arm.
Ca2+/Mg2+ binding to the lanthanide complexes slows down the exchange of the amide protons on YbL1
which is responsible for the diminished CEST effect. This has been evidenced by assessing the proton
exchange rates from the dependency of the CEST effect on the saturation time and the saturation power,
in the absence and in the presence of Ca2+ and Mg2+. The applicability of the PARACEST MRI agents for Ca2+
detection has been evaluated on a 16 T MRI scanner.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
cellular fluids is orders of magnitude higher than the intracellular
space; in mammalian body fluids, it is around 1.25 mM with minor
Calcium is one of the essential elements of the human body; it
has few rivals in its importance. It plays a central role in the phys-
iology and biochemistry of organisms and of the cell. Ca2+ is impor-
tant in signal transduction pathways (where it acts as a secondary
messenger), in neurotransmitter release from neurons, contraction
of all muscle cell types and fertilisation. Several enzymes require
calcium as a co-factor. Extracellular calcium is also important for
maintaining the potential difference across excitable cell mem-
branes, as well as proper bone formation. Around 99% of the total
calcium content in the body is found in the bone and calcified car-
tilage in the form of calcium phosphate and calcium sulphate. The
rest of calcium is present within the extracellular and intracellular
fluids. In a typical cell, the intracellular concentration of ionised
calcium is roughly 100 nM, but can increase up to 10- to 100-fold
during certain cellular functions. The Ca2+ concentration in extra-
variations.1 The development of selective organic fluorescent
probes has helped significantly in the understanding of the role
of calcium in vivo.2 However, a fundamental limitation of light-
based microscope imaging techniques employing fluorochromes,
is that they are restricted to transparent samples of a limited thick-
ness because of light absorption and scattering.
Magnetic resonance imaging (MRI) provides a good alternative
to light-based microscopy that can circumvent these limitations.
The visualisation of Ca2+ concentration and its variations in time
by imaging techniques would be particularly interesting to assess
neuronal signalling in the brain where Ca2+ acts as an important
secondary messenger.3 Significant changes in Ca2+ concentration
take place during neuronal activity.4 Tracking the dynamics of
Ca2+ concentration changes could thus contribute to the under-
standing of the basic aspects of neuronal regulation or to highlight
the abnormalities in the diseased state.5 Currently, the in vivo neu-
roimaging techniques used to study the activity of brain networks
are based on blood-oxygen-level dependent (BOLD) functional MRI
(fMRI). The BOLD technique indirectly measures neural activity by
detecting changes in blood flow to brain regions with increased
* Corresponding authors. Tel.: +49 7071 601 916; fax: +49 7071 601 919 (G.A.);
tel.: +33 2 38 25 76 25; fax: +33 2 38 63 15 17 (E.T.).
E-mail addresses: goran.angelovski@tuebingen.mpg.de (G. Angelovski), eva.ja
kabtoth@cnrs-orleans.fr (Éva Tóth).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.07.023

1098
G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
energy requirements. Given the vascular origin of the signal, BOLD
has some inherent physiological limitations.6
cient safety for potential in vivo applications.28 The iminodiacetate
moieties are known to have an affinity for Ca2+, but they do not
have a specific selectivity towards Ca2+ in the presence of other
endogenous metal ions such as Cu2+, Zn2+ or Mg2+. However, the
extracellular Ca2+ concentration is considerably higher than that
of the other cations, thus we may expect that will not interfere sig-
nificantly with the detection of calcium. We should note that the
iminodiacetate units are not expected to interfere in the formation
of the lanthanide complex; the lanthanide ion is coordinated by the
four macrocyclic nitrogens and the four amide oxygens, with one
water molecule completing the inner coordination sphere. Herein
we describe a detailed study of the CEST properties of the com-
plexes formed with Yb3+ and Eu3+ and their variation in the pres-
ence of divalent cations Ca2+ and Mg2+. MRI CEST experiments
showing the feasibility of Ca2+ detection on phantoms are also
reported.
The first contrast agent designed for sensing Ca2+ by MRI was
based on a Gd3+ complex that showed approximately an 80% in-
crease in water proton relaxivity in the presence of calcium ions,
in the micromolar concentration range corresponding to the intra-
cellular Ca2+ concentration.7 Atanasijevic et al. reported a T2 agent
based on the Ca2+ related aggregation of superparamagnetic iron
nanoparticles and calmoduline, also with an MRI response in the
intracellular concentration range.8 The main limitation of this latter
approach is the relatively long time course of the Ca-dependent
aggregation (a few seconds) inhibiting the detection of fast Ca-con-
centration changes. The concentration of free extracellular Ca2+ is in
the millimolar range and also drops up to 30–35% from the resting
state during intense stimulation (typically from 1.2 to 0.8 mM).9
Sensing Ca2+ changes at millimolar concentrations seems more
appropriate for MRI applications, as this concentration range is
more compatible with the relatively high amount of the magnetic
imaging probe needed to detect contrast changes. Targeting extra-
cellular Ca2+ also simplifies the chemical design, as additional
requirements for cell internalisation can be neglected. A series of
Gd3+ complexes have been reported as potential contrast agents
2. Results and discussion
2.1. Synthesis
The synthesis of the complexes EuL1 and YbL1 was achieved in
five steps (Scheme 1). The first involved the tetraalkylation of the
cyclen with N-Boc-N0-bromoacetylethylenediamine 129 to give
macrocycle 2. The amino groups were deprotected after treatment
of 2 in methanolic HCl, and further alkylation with tert-butyl bro-
moacetate was carried out without isolation of the intermediate
amino product.30 Finally, the octaester 3 was treated with formic
acid to give the ligand L1. Complexes EuL1 and YbL1 were prepared
by mixing the ligands with the corresponding lanthanide chloride
salts in aqueous solutions which were kept at neutral pH during
the complexation.
The bisamide complexes were prepared according to Scheme 2.
The synthesis started from the commercially available DO2AtBu
(4). Due to the presence of the acid labile tert-butyl protective
groups, the synthetic route for L1 was slightly modified. DO2AtBu
was alkylated with two molecules of the benzyl 2-(2-bromoace-
tamido)ethylcarbamate 531 to give 6. The Cbz protective groups
were removed upon the reduction with hydrogen (2.5 bar) in the
presence of 10% Pd/C. The amine 7 was alkylated with tert-butyl
bromoacetate to give the hexaester 8. Finally, all tert-butyl groups
were removed upon the treatment with formic acid to give ligand
L2, which on reaction with the appropriate lanthanide chloride salt
produced EuL2 and YbL2. All complexes were characterised by
mass spectrometry and the appropriate isotope pattern distribu-
tion was recorded.
to target the extracellular Ca2+ 10–13
The most effective compounds
.
showed a ꢂ10% relaxivity change under biologically relevant condi-
tions (in brain extracellular fluid, for a Ca2+ concentration change
between 0.8 and 1.2 mM) with a high selectivity for Ca2+
.
Recently, a new class of contrast agents have emerged that pro-
duce an image based on chemical exchange dependent saturation
transfer (CEST).14 They are an attractive alternative to gadolin-
ium-based T1 contrast agents which locally shorten the relaxation
time of bulk water protons to generate contrast in MRI.15,16 CEST
agents possess a proton with a moderate to slow exchange rate
with the bulk water. Selective saturation of the MR frequency of
this proton, followed by exchange with the water, reduces the
intensity of the bulk water MR signal. PARACEST agents include a
paramagnetic lanthanide ion which considerably shifts the MR fre-
quency of the exchangeable proton to facilitate selective detec-
tion.17–19
PARACEST agents can be particularly well adapted to the design
of responsive probes, and more recently, various examples have
been reported with selective response to enzymes,20,21 metabo-
lites,22,23 metal ions,24 pH25,26 or temperature.27 With the intention
of detecting Ca2+ via the PARACEST effect, we have prepared two
novel ligands (L1 and L2, Chart 1) based on a DOTA-tetraamide or
DOTA-bisamide chelating unit for lanthanide complexation that
bear four or two iminodiacetate arms. DOTA-tetraamide ligands
are known to form sufficiently stable complexes with trivalent lan-
thanide cations and such complexes have been those most fre-
quently investigated for application as PARACEST agents. Though
the thermodynamic stability of the DOTA-tetraamide lanthanide
complexes is considerably lower than that of the tetracarboxylate
analogues, their kinetic inertness is extremely high ensuring suffi-
2.2. CEST properties
The magnitude of the magnetisation transfer to the bulk water,
called the CEST effect, is usually calculated according to the follow-
ing equation:
Chart 1. Structure of the ligands studied.

G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
1099
Scheme 1. Synthesis of EuL1 and YbL1. Reagents and conditions: (a) K2CO3, CH3CN, 80 °C, 18 h. (b) HCl/CH3OH, 18 h. (c) tert-Butyl bromoacetate, K2CO3, CH3CN, 80 °C, 18 h.
(d) HCOOH, 60 °C, 18 h. (e) EuCl3 or YbCl3, H2O, 60 °C, 48 h.
Scheme 2. Synthesis of EuL2 and YbL2. Reagents and conditions: (a) K2CO3, DMF, 24 h. (b) H2, Pd/C, CH3OH, 18 h. (c) tert-Butyl bromoacetate, K2CO3, CH3CN, 80 °C, 18 h. (d)
HCOOH, 60 °C, 18 h. (e) EuCl3 or YbCl3, H2O, 60 °C, 48 h.
EuL1 and YbL1. The peak at 0 ppm represents direct saturation of
bulk water. The peak centred at 41 ppm in the spectrum of EuL1
%CEST ¼ ð1 ꢁ Ms=M0Þ ꢃ 100%
ð1Þ
Here Ms is the intensity of the magnetic resonance signal of the
bulk water taken after applying the radiofrequency (RF) saturation
pulse at the resonance frequency of the exchangeable proton(s) of
the contrast agent and M0 is the intensity of the reference signal,
that is, recorded after applying the RF saturation pulse symmetri-
cally on the opposite side of the bulk water signal, to correct for
non-selective saturation, mostly direct water saturation. To visual-
ise the magnetisation transfer, CEST spectra are typically used
when the ratio of the solvent water signal intensities measured
with and without presaturation (Ms/Moff) is plotted against the pre-
saturation frequency. Figure 1 shows the CEST spectra recorded for
(Fig. 1A) corresponds to the proton exchange between the Eu3+
-
bound water molecule and bulk solvent, whilst that at ꢁ11 ppm
in the spectrum of YbL1 (Fig. 1B and C) arises from the proton ex-
change between the amide proton and bulk water, based on anal-
ogy to previously reported DOTA-tetraamide complexes.19,32 As
expected, the temperature has a significant effect on the CEST
spectra (Fig. S1 in Supplementary data), indicating an acceleration
of the proton exchange at higher temperatures.
EuL2 and YbL2 show no CEST effect, either in the absence or in
the presence of Ca2+ (on addition of up to 20 equiv of Ca2+). In these

1100
G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
respectively (the m/M ratio was ꢂ0.7 for the Eu3+ complex). These
values are around two orders of magnitude higher than those re-
ported for the tetraamide EuDOTAM complex with a known PARA-
CEST effect, for instance (3.3 ꢃ 105 sꢁ1 and 8.3 ꢃ 103 sꢁ1 for m and
M isomers, respectively).34 One should also note that the water ex-
change rate (as assessed by 17O NMR spectroscopy) and the proton
exchange rate (assessed by 1H NMR) are identical for these types of
complexes.33 We have not determined the water exchange rate on
GdL2, which should be close to the water (and proton) exchange on
EuL2. However, we expect that it is in the same order of magnitude
as the one reported for the Gd3+ bis(dimethyl) amide DOTA com-
plex32 which is clearly too fast and therefore prevents the observa-
tion of a CEST effect on EuL2.
It is more surprising that the YbL2 complex does not show a
CEST peak either, as the proton exchange on the two amide nitro-
gens would be expected to be slow, within the appropriate time-
scale to produce CEST. It seems, however, that the presence of
the neighbouring carboxylates contributes to a faster proton ex-
change. The amide protons are not visible in the high resolution
1H NMR spectrum (Supplementary data). Temperature variations
within the range of 288–323 K had no effect on the recorded CEST
spectra which only showed the direct saturation of water.
To investigate the sensitivity of Eu3+ and Yb3+ complexes of L1
toward Ca2+ and Mg2+, we have measured the CEST effect as a func-
tion of the Ca2+ and Mg2+ concentrations. Some representative
CEST spectra are shown in Figure 1. In the absence of calcium ions,
the complexes exhibited a CEST effect of 60% and 35% for YbL1 and
EuL1, respectively. Upon addition of calcium the CEST effect de-
creased sensibly for both chelates. In the case of YbL1, the decrease
of the CEST effect is accompanied by a small shift of the amide pro-
ton resonance to ꢂꢁ13 ppm (see also the high resolution 1H NMR
spectra; Fig. S2 in Supplementary data). The addition of Mg2+ to the
YbL1 complex has a similar influence on the CEST effect observed
(Fig. 1C).
By plotting the relative CEST magnitude (1 ꢁ Ms/M0) for the
peak at ꢁ13 ppm as a function of the [Ca]/[YbL1] ratio, we obtain
a curve that reaches saturation from a [Ca]/[YbL1] ratio above
ꢂ40 (Fig. 2; here Ms and M0 correspond to the water signal inten-
sity at ꢁ13 ppm (on resonance) and +13 ppm (off resonance)).
These data points have been fitted to estimate an affinity constant
between YbL1 and Ca2+, by assuming that the observed overall
CEST magnitude is the sum of contributions originating from the
free YbL1 and from the Ca2+-bound YbL1 species. We have assumed
binding of a single Ca2+ cation per YbL1 complex. The conditional
affinity constant determined by this method is relatively low,
YbL1—Ca
K
¼ 8 ꢀ 2Mꢁ1, at pH 7.4. For the complexation of Ca2+ with
aff
the iminodiacetate (IDA) ligand, the formation of 1:1 and 1:2
Figure 1. Solvent water signal intensity against presaturation frequency (CEST
spectra). (A) Twenty millimolar aqueous solution of EuL1 in the absence (N) and in
the presence of 5 (j) and 100 equiv (ꢀ) of Ca2+. (B) Twenty millimolar aqueous
solution of YbL1 in the absence (N) and in the presence of 5 (j) and 100 equiv (ꢀ) of
Ca2+. (C) Twenty millimolar aqueous solution of YbL1 in the absence (N) and in the
presence of 5 (j) and 100 equiv (ꢀ) of Mg2+. Three hundred and ten Kelvin,
70
60
50
40
30
20
10
0
irradiation time 3 s, B1 = 25 lT.
complexes, the proton exchange rate on the amide protons or on
the coordinated water is probably too fast, thus we cannot detect
a CEST effect from the water coordinated to Eu3+ or from the amide
proton in YbL2, as observed for the corresponding L1 analogues. To
the best of our knowledge, no bis(amide) complex of Eu3+ or Yb3+
has ever been reported to show a CEST effect. The water exchange
rate has been previously determined on cyclen-trans-bis(amide)–
0
20
40
60
80
100
120
[Ca]/[YbL1]
bis(acetate) complexes of Gd3+ 33
For the bis(dimethyl) amide
.
Figure 2. Variation of the CEST effect observed at ꢁ13 ppm upon addition of Ca2+ to
the YbL1 complex. [YbL1] = 0.02 M, pH 7.4, T = 37 °C. The line represents the fit to
the data points as explained in the text.
analogue, the water exchange rate constants were found to be very
different for the m and M isomers, 7.0 ꢃ 107 sꢁ1 and 7.4 ꢃ 105 sꢁ1
,

G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
1101
species have been described, with thermodynamic stability con-
stants of log K1 = 3.3 and log K2 = 2.7.35 Taking into consideration
of the protonation constants of IDA, for pH 7.4 we can calculate a
we could never obtain a reasonable value for the formation of 1:2
species. Therefore we decided to fit the data with the simplest
model that gives a good quality fit.
conditional stability constant of bcond = 1.3 ꢃ 104 Mꢁ1 or Kcond
=
The interaction between YbL1 and Mg2+ has been also evaluated
YbL1—Mg
aff
19 Mꢁ1 assuming the formation of Ca(IDA)2 or Ca(IDA) species,
in a similar manner to yield K
¼ 23 ꢀ 3Mꢁ1 (Fig. S3 in Sup-
respectively. The value obtained for the YbL1–Ca2+ adduct,
plementary data. The affinity constant is slightly higher than that
obtained for the YbL1–Ca interaction. This is in good agreement
with the higher stability constant reported for the Mg(IDA) com-
plex (log K1 = 3.66)36 compared to the Ca2+ analogue. Given the
smaller size of Mg2+, the coordination of a second iminodiacetate
unit is even less probable. Indeed, no 1:2 complex has been de-
tected in the Mg2+-IDA system either.35 In the case of EuL1, the var-
iation of the CEST magnitude at 26 ppm has been plotted as a
function of the [Ca]/[EuL1] ratio, and fitted analogously to YbL1.
YbL1—Ca
K
¼ 8Mꢁ1, is close to the conditional stability constant calcu-
aff
lated for the Ca2+ complex formed with one iminodiacetate unit.
This suggests the binding of one iminodiacetate per Ca2+. The
coordination of a second IDA is likely to be hindered by steric
constraints; all amide oxygens are coordinated to the lanthanide
ion, and the ethylene linker between the amide nitrogen and the
imino nitrogen is too short to provide enough flexibility so that a
second IDA unit could turn towards the Ca2+ coordinated by a
neighbouring pending arm.
The affinity constant determined for the EuL1–Ca2+ interaction is
EuL1—Ca
We could also consider that the four IDA units are independent
K
¼ 10 ꢀ 3Mꢁ1, in good agreement with the value for YbL1
aff
with respect to Ca2+ coordination and can each coordinate a cation.
(Fig. S4 in Supplementary data). We should note that due to the
smaller variation of the CEST effect, the estimation of Kaff is less
precise than in the case of YbL1.
By analysing the data in this way, for the YbL1–Ca system we ob-
YbL1—Ca
tain K
¼ 11 ꢀ 2, and the quality of the fit is as good as that
aff
in Figure 2. However, we believe that the coordination of four cal-
cium ions is highly improbable due to steric reasons. The data can
When responsive agents are developed for MRI detection, the
selectivity towards to the cation (or other species) is an important
requirement. Our system does not show any selectivity for Ca2+
over the chemically similar and biologically available Mg2+. The
concentration of free Ca2+ is slightly higher in the extracellular
medium, 1.0 mM with respect to 0.7 mM for Mg2+ in the cerebro-
spinal fluid.9 In the plasma, the difference is more significant:
1.4 mM for free Ca2+ and 0.5 mM for free Mg2+. However, given
the predominance in number of Ca2+ channels and transporters,37
the variation in the Mg2+ concentration in comparison to variation
in the Ca2+ concentration during the neural activity should be min-
imal. Thus changes in Ca2+ concentration are more likely dominate
over the changes in Mg2+ concentration. We should also note that
be also fitted by assuming the binding of two or three Ca2+ ions to
YbL1—Ca
one LnL chelate, with constants obtained between K
¼ 8 and
aff
11. We should note that, in either of the cases, the value of the
affinity constant is close to the conditional constant for Ca(IDA)
(1:1 complex).
On the other hand, if there were more than one calcium ions
coordinated to the LnL1 complex, very likely the different sites of
coordination could not be considered as independent, but a succes-
sive binding model should be rather used. Unfortunately, all at-
tempts to try to fit more than one constant were unsuccessful. It
could be reasonable to include also the Ca(IDA)2 species. However,
Fig. 3. CEST spectra of YbL1 recorded with variable irradiation times in the absence (A) and in the presence of 100 equiv of calcium (B). Irradiation times from top to bottom:
0.25, 0.5, 1, 1.5, 2, 3, 4, 5 s. CEST spectra of YbL1 recorded with variable saturation power in the absence (C) and in the presence of 100 equiv of calcium (D). Saturation power
from top to bottom: 36, 34, 30, 28, 26, 24, 22, 20, 18 and 17 dB. T = 37 °C.

1102
G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
Zn2+ or Cu2+ (which are the most abundant divalent endogenous
cations after Ca2+ and Mg2+) are much less concentrated (micromo-
lar range) and should not interfere in the detection of Ca2+. These
factors can contribute to a reliable detection of calcium with L1 de-
spite the low selectivity.
k2ex ꢄ xCA
q ¼ R1S þ kex
ꢁ
ð5Þ
R1W þ kex ꢄ xCA
where R1,2S and R1,2W are the longitudinal and transverse relaxation
rates during saturation of the solute and of the bulk water, respec-
tively. The expression given in Eq. 2 reaches the same maximum
CEST effect as predicted by Sherry et al.18 and Aime et al.39 assum-
ing complete saturation.
The CEST effect is directly related to the rate of the proton ex-
change between the bulk solvent and the amide (for YbL1) or the
coordinated water site (for EuL1). The exchange rate can be quan-
titatively assessed from the saturation time and saturation power
dependency of the water proton intensity.38 We performed QUEST
and QUESP experiments (quantification of the exchange rate as a
function of saturation time or saturation power, respectively), to
determine the exchange rate for the YbL1 system in the absence
We have made the assumption that R1W and R2W can be approx-
imated by the relaxation rates measured for bulk water32 in a solu-
tion containing 20 mM YbL1 (with or without Ca2+/Mg2+). These
values have been therefore determined by independent measure-
ments using inversion recovery and Carr–Purcell–Meiboom–Gill
sequences and were fixed during the fit (Table 1). The longitudinal
and transverse relaxation rates of the solute R1,2S (in Eqs. 4 and 5)
have been calculated in the fit. We should note that R1S is negligi-
ble as compared to kex and has no influence on the fit. Table 1 dem-
onstrates that the presence of Ca2+ or Mg2+ salts have essentially no
effect on any of the relaxation rates. The best fitting values for kex
and the relaxation rates determined independently are listed in
Table 1.
We should note that we have also analysed the 1 ꢁ Ms/M0 val-
ues obtained as a function of the saturation time for all three sys-
tems (YbL1 without and with Ca2+/Mg2+) according to Eq. 6, as
reported by Zhang et al.17 Here we have fitted both kex and R1W
and obtained very similar values for the exchange rate (within
10%) as in the simultaneous fit of the saturation time dependent
and saturation power dependent CEST data to Eq. 2.
and in the presence of Ca2+ or Mg2+
.
Figure 3 shows the effect of saturation power and irradiation
time on the transfer of saturation from the amide proton to the
bulk water pool for YbL1 in the absence and in the presence of
100 equiv of Ca2+. Analogous data have been recorded in the pres-
ence of 100 equiv of Mg2+ (shown in the Supplementary data). The
YbL1 complex has been chosen for these experiments since the ef-
fect of the alkaline earth cations is more important than with EuL1.
The results (Fig. 3A and B) indicate that the CEST effect does not in-
crease when the saturation time is increased beyond 2 s. On the
other hand, we observe that a high pulse power produces a large
increase of the CEST effect (Fig. 3C and D).
Figure 4A and B show the dependence of the CEST effect (mea-
sured at ꢁ13 ppm) on the saturation time and on the saturation
power in the absence and in the presence of calcium, respectively.
The data have been fitted to Eqs. 2–5 to obtain the rate of proton
exchange.
Ms
M0
1 ꢁ
¼ 1
ꢀ
ꢁ
Assuming that the steady state is reached upon saturation of the
solute, the following equation applies for the measured CEST
effect,37
R1W
kex ꢄ xCA
1WþkexꢄxCAÞtsat
eꢁðR
ꢁ
þ
ð6Þ
R1W þ kex ꢄ xCA R1W þ kex ꢄ xCA
The exchange rate of the amide proton obtained for YbL1 is in
good agreement with those reported in the literature for tetraa-
mide DOTA-derivative Yb3+ complexes. For instance, kex = 1724 sꢁ1
Ms
kex
ꢄ
a
ꢄ xCA
1WþkexꢄxCAÞtsat
1 ꢁ
¼
ꢃ ½1 ꢁ eꢁðR
ꢅ
ð2Þ
M0 R1W þ kex ꢄ xCA
where xCA is the fractional concentration of the exchangeable pro-
tons of the contrast agent, tsat is the saturation time, is the satu-
ration efficiency, and kex is the rate of proton exchange on the
amide. The saturation efficiency depends on the pulse power via
the following equation:
Table 1
a
kex values calculated from the best fit of QUEST and QUESP data and relaxation rates
obtained in the fit or determined in independent inversion recovery (R1W) or CPMG
(R2W) experiments
YbL1
YbL1 + 100 equiv Ca2+
YbL1 + 100 equiv Mg2+
kex (sꢁ1
)
)
)
2100 150
0.624
588
0.269
0.271
210 20
0.624
588
0.262
0.263
0.0006
309 20
0.624
588
0.262
0.263
0.0006
x12
R1S (sꢁ1
R2S (sꢁ1
a
¼
ð3Þ
x21 þ p ꢄ q
R1W (sꢁ1
R2W (sꢁ1
XCA
)
)
where
0.00072
k2ex ꢄ xCA
p ¼ R2S þ kex
ꢁ
ð4Þ
Three hundred and ten Kelvin, pH 7.4.
R2W þ kex ꢄ xCA
Fig. 4. QUEST and QUESP results with best fits to the Eqs. 2–5. (A) Plot of the magnetisation as a function of the saturation time for YbL1 in the absence (N) and in the presence
of 100 equiv of calcium (j). (B) Plot of the magnetisation as a function of the saturation power for YbL1 in the absence (N) and in the presence of 100 equiv of calcium (j). The
curves represent the fit as explained in the text.

G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
1103
was determined for YbDOTAM (at 298 K and pH 7.4),39 kex = 2500 s
ꢁ1
for the glycinate derivative YbDOTAMGly (at 312 K and pH
8.1),40 or kex = 1111 sꢁ1 for the tetra(glycine-phenylalanine)
derivative YbDOTAMGly-Phe (at 311 K, pH 7.0).27 Upon addition
of Ca2+ or Mg2+ to the YbL1 complex, the exchange rate of the
amide proton decreases significantly which is then translated as
a diminution of the observed CEST effect. If we assume that there
is one alkaline earth cation bound per YbL1 chelate and it is bound
to a single iminodiacetate arm as suggested by the affinity constant
(vida infra), the amide protons become non-equivalent and the re-
ported exchange rate is an effective (average) value. A diminution
of the amide proton exchange rate has been previously reported for
the tetraglycinate derivative LnDOTAMGly complexes when
decreasing the pH from 8.1 to 6.0 (e.g., kex = 2500 sꢁ1 at pH 8.1
for YbDOTAMGly in comparison to kex = 125–128 sꢁ1 for Ln = Eu,
Nd, Pr at pH 6, T = 312 K).40 The slower proton exchange resulted
in a subsequent diminution or disappearance of the observed CEST
effect, similar to what we observe upon addition of the alkaline
earth metal ions to YbL1. One possible explanation of the slower
amide proton exchange on the Ca2+-bound species can be related
to a change in the protonation state of the iminodiacetate group.
In the absence of Ca2+, the nitrogen of the iminodiacetate moiety
is protonated at pH 7.4, and the presence of this proton on the
amine can contribute to a faster exchange (log KH = 9.32 for the
nitrogen of IDA, while a lower value, log KH = 8.17 was reported
for the N-acetyl-1,2-diaminoethane-N0,N0-diacetic acid that has a
structure similar to the pending arm of L1).41 Upon Ca2+ coordina-
tion, the amine is deprotonated and this catalytic effect is no longer
operating. On the other hand, the charge of the iminodiacetate unit
also changes on Ca2+ coordination from ꢁ1 to 0 which could also
influence the proton exchange on the neighbouring amide.
Although in PARACEST imaging the role of the contrast agent
concentration might be considered less critical as compared to
Gd3+-based agents (the CEST effect becomes independent of con-
centration after a certain limit), a precise knowledge of the local
concentration of the agent can be still important for an accurate
determination of the cation to be sensed in real applications. In or-
der to circumvent the problem of not knowing the local concentra-
tion of the agent, ratiometric methods have been proposed either
by using a probe endowed with multiple exchange sites (coordi-
nated water and amide protons) that have a different response to
the analyte or by applying two distinct molecules (two different
lanthanide complexes) with a different CEST response. This second
approach can be realised when a mixture of YbL1 and EuL1 is used
Fig. 6. CEST images of phantoms containing water or a solution of 20 mM YbL1 and
EuL1 before (bottom) and after (top) addition of calcium. The CEST images represent
the intensity difference between the spin-echo images for saturation at
d = ꢁ11 ppm and d = +11 ppm (YbL1) or at d = +41 ppm and d = +26 ppm (EuL1),
23 °C.
which provides differing CEST responses to Ca2+. For the two anal-
ogous complexes, one can expect identical biodistribution, there-
fore the ratio of the CEST effect detected at ꢁ13 ppm (for YbL1)
and +41 ppm (EuL1) will be independent of the absolute concentra-
tion of the complexes. This ratio has been plotted as a function of
the Ca2+ concentration (Fig. 5). Up to 0.5 M Ca2+, this ratio changes
considerably, whilst at higher concentrations it becomes constant.
To demonstrate that Ca2+ binding can be sensed in a CEST imag-
ing experiment, a phantom consisting of plastic tubes, each con-
taining 20 mM EuL1 or YbL1 in presence or absence of Ca2+ ions,
was prepared. Images were acquired after applying a selective 3 s
presaturation pulse at d = 41 ppm and d = ꢁ11 ppm for EuL1 and
YbL1, respectively. A second image was collected after presatura-
tion at d = 26 ppm and d = 11 ppm and CEST images were gener-
ated through pixel-by-pixel subtraction of the two images. The
intensities of the CEST images (Fig. 6) show a clear difference after
addition of calcium ions.
3. Conclusions
We have described the synthesis and evaluation of a new family of
calcium-activated PARACEST contrast agents. The sensors that com-
bine a Ln3+ chelating unit with an aminocarboxylate-based moiety
for calcium binding resulting in marked variations in the CEST effect
in response to Ca2+. The affinity constants between the lanthanide
complexes and the alkaline earth metal ions have been estimated
YbL1—Ca
from the variation of the CEST effect to be K
K
¼ 8 ꢀ 2Mꢁ1
,
aff
YbL1—Mg
aff
EuL1—Ca
aff
¼ 23 ꢀ 3Mꢁ1 and K
¼ 10ꢀ 3Mꢁ1. These low values
suggest the coordination of the alkaline earth ions to a single imino-
diacetate arm. Ca2+/Mg2+ binding to the lanthanide complexes slows
downtheexchangeoftheamideprotonsonYbL1 whichisresponsible
for the diminished CEST effect. EuL1 and YbL1 are able to produce a
variation of the MRI signal along with the change of calcium
concentration.
3
2
1
0
4. Experimental
4.1. General procedures
1H NMR and 13C{1H} NMR spectra were recorded on Bruker
Avance II 300 MHz ‘Microbay’ spectrometer at room temperature.
ESI-HRMS were performed on a Bruker BioApex II ESI-FT-ICR,
equipped with an Agilent ESI-Source, measured via flow injection
analysis. ESI-LRMS were performed on a ion trap SL 1100 system
(Agilent, Germany). Column chromatography was performed using
silicagel 60 (70–230 mesh ASTM) from Merck, Germany. DO2AtBu
(4) was purchased from CheMatech, Dijon, France.
0
0.2
0.4
[Ca2+] / M
0.6
0.8
Fig. 5. Calcium dependence of the relative CEST effect (independent of the LnL1
complex concentration) as resulting from the ratiometric method by simulta-
neously using EuL1 and YbL1. cEuL1 ¼ cYbL1 ¼ 20mM; 310 K, irradiation time 3 s,
B1 = 25 lT. (The line was drawn to guide the eyes.)

1104
G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
4.2. Synthesis of the ligand
due was purified by column chromatography (CH2Cl2/CH3OH
93:7) to afford 764 mg (75%) of 6 as yellow viscous oil which solid-
ifies upon standing.
4.2.1. Compound 2
Cyclen (59 mg, 0.34 mmol) was dissolved in CH3CN (10 ml),
K2CO3 (458 mg, 3.31 mmol) was added and the suspension was
stirred for 15 min. Bromide 1 (465 mg, 1.65 mmol) in CH3CN
(10 ml) was added dropwise and the mixture was heated at 80 °C
for 18 h. After cooling the solid material was removed by filtration,
the solvent was evaporated and the residue purified by column
chromatography (CH2Cl2/CH3OH 92:8) to yield 290 mg (88%) of 2.
Analytical data match to those previously reported for this
molecule.30
1H NMR (CDCl3, 300 MHz): d = 7.79 (br s, 2H), 7.29–7.16 (m,
10H), 6.06 (br s, 2H), 4.99 (s, 4H), 3.33–3.08 (br, 12H), 3.02–2.75
(br, 4H), 2.55–1.95 (br, 16H), 1.32 (s, 18H). 13C{1H} NMR (CDCl3,
75 MHz): d = 172.1, 171.7, 156.8, 136.9, 128.3, 127.8, 127.7, 81.6,
66.2, 56.2, 55.5, 50.4, 40.9, 39.6, 28.0. ESI-HRMS: for
þ
C
C
44H69N8O10
:
calcd 869.5131 [M+H]+, found 869.5165; for
44H68N8NaO10þ: calcd 891.4951 [M+Na]+, found 891.4953.
4.2.5. {4,10-Bis-[(2-amino-ethylcarbamoyl)-methyl]-7-tert-
butoxycarbonylmethyl-1,4,7,10tetraaza-cyclododec-1-yl}-
acetic acid tert-butyl ester, 7
4.2.2. tert-Butoxycarbonylmethyl-{2-[2-(4,7,10-tris-{[2-(bis-tert
-butoxycarbonylmethyl-amino)-ethylcarbamoyl]-methyl}-
1,4,7,10tetraaza-cyclododec-1-yl)-acetylamino]-ethyl}-amino-
acetic acid tert-butyl ester (3)
Compound 6 (764 mg, 0.88 mmol) was dissolved in CH3OH
(50 ml) and Pd/C (200 mg) was added. After the hydrogenation at
2.5 bar (18 h), suspension was filtered through celite and solvent
was evaporated to give 7 containing carbamate residues.
1H NMR (CDCl3, 300 MHz): d = 7.98 (br s, 2H), 3.32–3.17 (br,
8H), 3.13–2.89 (br, 8H), 2.84–2.73 (br, 4H), 2.64–2.13 (br, 16H),
1.37 (s, 18H). 13C{1H} NMR (CDCl3, 75 MHz): d = 171.9, 171.6,
81.5, 56.3, 55.5, 50.5, 50.1, 41.6, 41.3, 28.0. ESI-HRMS: for
Protected amine 2 (660 mg, 0.68 mmol), was dissolved in
1.25 M methanolic HCl (20 ml) and the solution was stirred at
room temperature for 18 h. The solvent was removed on the rotary
evaporator and the residue was dissolved in water (5 ml). The pH
of the solution was adjusted to neutral using 5 M NaOH and the
water was removed on the rotary evaporator. The solid residue
and K2CO3 (1.87 g, 13.53 mmol) were suspended in CH3CN
(20 ml) and the mixture was stirred for 30 min. tert-Butyl bromo-
acetate (1.32 g, 6.77 mmol) in CH3CN (10 ml) was finally added
dropwise and the mixture was heated at 80 °C for 18 h. After cool-
ing the suspension was filtered, the solvent was evaporated and
the residue purified by column chromatography (CH2Cl2/CH3OH
96:4) to yield 533 mg (53%) of 3 as yellow viscous oil which solid-
ifies upon standing.
C
28H56N8NaO6þ: calcd 623.4215 [M+Na]+, found 623.4223.
4.2.6. (4,10-Bis-{[2-(bis-tert-butoxycarbonylmethyl-amino)-
ethylcarbamoyl]-methyl}-7-tert-butoxycarbonylmethyl-
1,4,7,10tetraaza-cyclododec-1-yl)-acetic acid tert-butyl ester, 8
Amine
7
(106 mg, 0.18 mmol), tert-butyl bromoacetate
(172 mg, 0.88 mmol) and K2CO3 (243 mg, 1.76 mmol) were mixed
in CH3CN (10 ml) and the mixture was heated at 80 °C for 18 h.
After cooling the suspension was filtered, the solvent was evapo-
rated and the residue purified by column chromatography
(CH2Cl2/CH3OH 97:3) to yield 99 mg (53%) of 8 as yellow viscous
oil.
1H NMR (CDCl3, 300 MHz): d = 7.63 (br s, 4H), 3.34 (s, 16H),
3.24–3.14 (br, 8H), 3.12–3.03 (br, 8H), 2.80–2.70 (br, 8H), 2.67–
2.48 (br, 8H), 2.43–2.27 (br, 8H), 1.38 (s, 72H). 13C{1H} NMR (CDCl3,
75 MHz): d = 171.1, 170.8, 81.2, 57.1, 55.8, 52.7, 50.4, 37.3, 28.0.
ESI-HRMS: for C72H133N12O20þ: calcd 1485.9754 [M+H]+, found
1H NMR (CDCl3, 300 MHz): d = 7.61–7.56 (br, 2H), 3.34 (s, 8H),
3.22–3.10 (br, 8H), 3.06–2.92 (br, 4H), 2.79–2.72 (br, 4H), 2.69–
2.42 (br, 8H), 2.40–2.19 (br, 8H), 1.39 (s, 36H), 1.37 (s, 18H).
13C{1H} NMR (CDCl3, 75 MHz): d = 171.7, 171.1, 171.0, 81.3, 56.5,
55.9, 55.6, 53.0, 50.6, 37.2, 28.1. ESI-HRMS: for C52H97N8O14þ: calcd
1057.7119 [M+H]+, found 1057.7139; for C52H96N8NaO14þ: calcd
1079.6938 [M+Na]+, found 1079.6926.
1485.9847; for C72H132N12NaO20 calcd 1507.9573 [M+Na]+, found
þ
1507.9577.
4.2.3. (Carboxymethyl-{2-[2-(4,7,10-tris-{[2-(bis-carboxy-
methyl-amino)-ethylcarbamoyl]-methyl}-1,4,7,10tetraaza-
cyclododec-1-yl)-acetylamino]-ethyl}-amino)-acetic acid (L1)
Compound 3 (128 mg, 0.086 mmol) was dissolved in formic
acid (5 ml) and the solution was heated at 60 °C for 18 h. After
the solution was cooled, formic acid was removed on the rotary
evaporator. The residue was dissolved in minimal amount of water,
added dropwise to cooled acetone (ꢁ20 °C) and stored in the free-
zer overnight. The acetone was decanted from the solid material
and the crude product was obtained by drying it under reduced
pressure to afford 73 mg (82%) of L1 as colourless hygroscopic
solid.
4.2.7. (4,10-Bis-{[2-(bis-carboxymethyl-amino)-ethylcar-
bamoyl]-methyl}-7-carboxymethyl-1,4,7,10tetraaza-cyclod-
odec-1-yl)-acetic acid, L2
Compound 8 (269 mg, 0.254 mmol) was dissolved in formic
acid (5 ml) and the solution was heated at 60 °C for 18 h. After
the solution was cooled, formic acid was removed on the rotary
evaporator. The residue was dissolved in minimal amount of water,
added dropwise to cooled acetone (ꢁ20 °C) and stored in the free-
zer overnight. The acetone was decanted from the solid material
and the crude product was obtained by its drying in the vacuum
to afford 145 mg (79%) of L2 as colourless hygroscopic solid.
1H NMR (D2O, 300 MHz): d = 3.90–3.79 (br, 12H), 3.63–3.45 (br,
8H), 3.44–3.28 (br, 12H), 3.20–2.94 (br, 8H). 13C{1H} NMR (D2O,
75 MHz): d = 169.7, 56.5, 55.5, 54.8, 51.1, 48.6, 34.9. ESI-HRMS:
for C28H47N8O14þ: calcd 719.3217 [MꢁH]ꢁ, found 719.3225.
1H NMR (D2O, 300 MHz): d = 3.91–3.79 (br, 24H), 3.61–3.52 (br,
8H), 3.39–3.32 (br, 8H), 3.31–3.18 (br, 16H). 13C{1H} NMR (D2O,
75 MHz): d = 169.6, 56.6, 55.6, 54.3, 50.0, 34.7. ESI-HRMS: for
C
C
40H67N12O20ꢁ: calcd 1035.4600 [MꢁH]ꢁ, found 1035.4633; for
40H66N12O202ꢁ: calcd 517.2264 [Mꢁ2H]2ꢁ, found 517.2259.
4.2.4. {4,10-Bis-[(2-benzyloxycarbonylamino-ethylcarbamoyl)-
methyl]-7-tert-butoxycarbonylmethyl-1,4,7,10tetraaza-cycl-
ododec-1-yl}-acetic acid tert-butyl ester, 6
DO2AtBu (4, 467 mg, 1.17 mmol) was dissolved in dry DMF
(3 ml) and K2CO3 (776 mg, 5.61 mmol) was added. After stirring
for 30 min, benzyl 2-(2-bromoacetamido)ethylcarbamate (5,
882 mg, 2.80 mmol) in DMF (2 ml) was added slowly and suspen-
sion was stirred for 24 h at room temperature. After that period the
suspension was filtered, the solvent was evaporated and the resi-
4.2.8. Ln3+ complexes
The Ln3+ complexes of L1 and L2 were prepared by mixing the
ligand and the LnCl3 solutions in 1:1 ratio. The solution was stirred
at 60 °C for 48 h and at room temperature for 24 h. The pH was ad-
justed to 7.0–7.5 using a solution of NaOH (1 M). After 72 h, the
mixture was stirred for 24 h at room temperature in presence of
Chelex 100. The absence of free Ln3+ (Yb3+ or Eu3+) was verified
by colorimetric assay using xylenol orange.

G. Angelovski et al. / Bioorg. Med. Chem. 19 (2011) 1097–1105
1105
4.2.9. NMR measurements
References and notes
All CEST experiments were carried out on Bruker Avance
500 MHz spectrometer. The saturation transfer experiments were
carried out at 310 K by irradiating the sample at increments of
0.1 ppm. CEST spectra were recorded using pre-saturation pulses
of 3 s duration at 20 and 25 lT. Spectra were measured by record-
ing the bulk water signal intensity as a function of the presatura-
1. Forsen, S.; Kordel, J. Calcium in Biological Systems. In Bioinorganic Chemistry;
Bertini, I., Gray, H. B., Lippard, S. J., Valentine, J. S., Eds.; University Science
Books: Mill Valley, CA, 1994; p 107.
2. Tsien, R. Y. Chem. Eng. News 1994, 72, 34.
3. Tsien, R. W.; Tsien, R. Y. Annu. Rev. Cell Biol. 1990, 6, 715.
4. Meldolesi, J. Nature 1998, 392, 863.
5. Burgoyne, R. D. Nat. Rev. Neurosci. 2007, 8, 182.
6. Logothetis, N. K. Nature 2008, 453, 869.
tion frequency.
7. Li, W.-H.; Fraser, S. E.; Meade, T. J. J. Am. Chem. Soc. 1999, 121, 1413.
8. Atanasijevic, T.; Shusteff, M.; Fam, P.; Jasanoff, A. Proc. Natl. Acad. Sci. U.S.A.
2006, 103, 14707.
9. Somjen, G. G. Ions in the Brain: Normal Function, SEIZURES, and Stroke; Oxford
University Press: New York, 2004.
10. Dhingra, K.; Fousková, P.; Angelovski, G.; Maier, M. E.; Logothetis, N. K.; Tóth, É.
J. Biol. Inorg. Chem. 2008, 13, 35.
11. Mishra, A.; Fousková, P.; Angelovski, G.; Balogh, E.; Mishra, A. K.; Logothetis, N.
K.; Tóth, É. Inorg. Chem. 2008, 47, 1370.
12. Angelovski, G.; Fousková, P.; Mamedov, I.; Canals, S.; Tóth, É.; Logothetis, N. K.
ChemBioChem 2008, 9, 1729.
13. Dhingra, K.; Maier, M. E.; Beyerlein, M.; Angelovski, G.; Logothetis, N. K. Chem.
Commun. 2008, 3444.
14. Ward, K. M.; Aletras, A. H.; Balaban, R. S. J. Magn. Reson. 2000, 143, 79.
15. Toth, E.; Helm, L.; Merbach, A.E. Metal complexes as MRI contrast
enhancement agents. In Comprehensive Coordination Chemistry II; Elsevier:
Amsterdam, 2004.
16. Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B. Chem. Rev. 1999, 99, 2293.
17. Zhang, S.; Winter, P.; Wu, K.; Sherry, A. D. J. Am. Chem. Soc. 2001, 123, 1517.
18. Zhang, S.; Merritt, M.; Woessner, D. E.; Lenkinski, R. E.; Sherry, A. D. Acc. Chem.
Res. 2003, 36, 783.
For the QUEST experiments (quantification of the exchange rate
as a function of saturation time), data was collected by varying the
saturation time and keeping the power constant (25 lT). The satu-
ration times were 0.25, 0.5, 1, 1.5, 2, 3, 4, 5 s. QUESP data (quanti-
fication of the exchange rate as a function of saturation power) was
collected by varying the saturation power whilst the saturation
time remained constant (3 s). The saturation field strengths varied
between 150 and 1620 Hz. The QUEST and QUESP data were both
fit analytically with ScientistÒ (MicroMathÒ, Inc.).
The CaCl2 and MgCl2 solutions were made from the respective
chloride salts in distiled water. The affinity constants were deter-
mined by adding CaCl2 and MgCl2 solutions to the LnL1 solutions
at pH 7.4 and recording the CEST effect at ꢁ13 ppm and +13 ppm
for YbL1 and at +42 ppm and ꢁ42 ppm for EuL1. The titrations were
carried out at 37 °C.
19. Woods, M.; Woessner, D. E.; Sherry, A. D. Chem. Soc. Rev. 2006, 35, 500.
20. Chauvin, T.; Durand, P.; Bernier, M.; Meudal, H.; Doan, B.-T.; Noury, F.; Badet,
B.; Beloeil, J.-C.; Toth, E. Angew. Chem., Int. Ed. 2008, 47, 4370.
21. Yoo, B.; Pagel, M. D. J. Am. Chem. Soc. 2006, 128, 14032.
22. Zhang, S.; Trokowski, R.; Sherry, A. D. J. Am. Chem. Soc. 2003, 125, 15288.
23. Ren, J.; Trokowski, R.; Zhang, S.; Malloy, C. R.; Sherry, A. D. Magn. Reson. Med.
2008, 60, 1047.
24. Trokowski, R.; Ren, J.; Kalman, F. K.; Sherry, A. D. Angew. Chem., Int. Ed. 2005, 44,
6920.
25. Aime, S.; Castelli, D. D.; Terreno, E. Angew. Chem., Int. Ed. 2002, 41, 4334.
26. Zhou, J.; van Zijl, P. C. M. Prog. NMR Spectrosc. 2006, 48, 109.
27. Li, A. X.; Wojciechowski, F.; Suchy, M.; Jones, C. K.; Hudson, R. H. E.; Menon, R.
S.; Bartha, R. Magn. Reson. Med. 2008, 59, 374.
28. Pasha, A.; Lin, M.; Tircsó, G.; Rostollan, C. L.; Woods, M.; Kiefer, G. E.; Sherry, A.
D.; Sun, X. J. Biol. Inorg. Chem. 2009, 14, 421.
29. Xu, R.; Wang, Y.; Wang, X.; Jeong, E.-K.; Parker, D. L.; Lu, Z. R. Exp. Biol. Med.
2007, 232, 1081.
30. Fanning, A. M.; Plush, S. E.; Gunnlaugsson, T. Chem. Commun. 2006, 3791.
31. Barge, A.; Tei, L.; Upadhyaya, D.; Fedeli, F.; Beltrami, L.; Stefanía, R.; Aime, S.;
Cravotto, G. Org. Biomol. Chem. 2008, 6, 1176.
4.2.10. MRI measurements
Imaging experiments at a field strength of 16.4 T (correspond-
ing to a proton Larmor frequency of 700 MHz) were performed
on a Magnex magnet interfaced to a Bruker console. The samples
were immersed in water to avoid susceptibility artefacts and
heated to 37 °C throughout the experiments. Images were acquired
using a RARE (rapid acquisition with relaxation enhancement) se-
quence, preceded by a 3 s magnetisation transfer pulse with a
strength of 20
were acquired per repetition, an inplane spatial resolution of
234 m with a slice thickness of 1 mm was obtained within
lT. A repetition time of 7500 ms and four echoes
l
3 min. The data was reconstructed and difference images were
formed using home-written Matlab (The MathWorks Inc., Natick,
MA) routines.
32. Zhang, S.; Michaudet, L.; Burgess, S.; Sherry, A. D. Angew. Chem., Int. Ed. 2002,
41, 1919.
Acknowledgements
33. Zhang, S.; Kovacs, Z.; Burgess, S.; Aime, S.; Terreno, E.; Sherry, A. D. Chem. Eur. J.
2001, 7, 288.
The authors like to thank Dr. I. Mamedov for helpful discussions.
Financial support from the DFG (Grant AN 716/2-1), Max-Planck
Society, and the Centre National de la Recherche Scientifique
(France) is gratefully acknowledged. This work was performed with-
in the European COST Action D38.
34. Dunand, F. A.; Aime, S.; Merbach, A. E. J. Am. Chem. Soc. 2000, 122, 1506.
35. Ambrose, J.; Covington, A.; Thirsk, H. Trans. Faraday Soc. 1975, 65, 1897.
36. Schwarzenbach, G.; Kampitsch, E.; Steiner, R. Helv. Chim. Acta 1945, 28, 1139.
37. Kandel, E. R.; Schwartz, J. H.; Jessell, T. M., Eds., Principles of Neural Science, 4th
ed.; McGraw-Hill, 2000.
38. McMahon, M. T.; Gilad, A. A.; Zhou, J.; Sun, P. Z.; Bulte, J. W. M.; van Zijl, P. C. M.
Magn. Reson. Med. 2006, 55, 836.
39. Terreno, E.; Castelli, D. D.; Cravotto, G.; Milone, L.; Aime, S. Invest. Radiol. 2004,
39, 235.
Supplementary data
40. Aime, S.; Barge, A.; Della Castelli, D.; Fedeli, F.; Mortillaro, A.; Nielsen, F. U.;
Terreno, E. Magn. Reson. Med. 2002, 47, 639.
41. Genik-Sas-Berezowsky, R. M.; Spinner, I. H. Can. J. Chem. 1970, 48, 163.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.07.023.