A pyrophosphate-responsive gadolinium(III) MRI contrast agent

Article abstract of DOI:10.1002/chem.201001397

This study shows that the relaxivity and optical properties of functionalised lanthanide-DTPA-bis-amide complexes (lanthanide=Gd³⁺ and Eu³⁺, DTPA=diethylene triamine pentaacetic acid) can be successfully modulated by addition of specific anions, without direct Ln ³⁺/anion coordination. Zinc(II)-dipicolylamine moieties, which are known to bind strongly to phosphates, were introduced in the amide "arms" of these ligands, and the interaction of the resulting Gd-Zn₂ complexes with a range of anions was screened by using indicator displacement assays (IDAs). Considerable selectivity for polyphosphorylated species (such as pyrophosphate and adenosine-5′- triphosphate (ATP)) over a range of other anions (including monophosphorylated anions) was apparent. In addition, we show that pyrophosphate modulates the relaxivity of the gadolinium(III) complex, this modulation being sufficiently large to be observed in imaging experiments. To establish the binding mode of the pyrophosphate and gain insight into the origin of the relaxometric modulation, a series of studies including UV/Vis and emission spectroscopy, luminescence lifetime measurements in H₂O and D₂O, ¹⁷O and ³¹P NMR spectroscopy and nuclear magnetic resonance dispersion (NMRD) studies were carried out. Pick an anion: The anion-binding properties of functionalised Ln-DTPA-bis-amide complexes (Ln=Gd³⁺, Eu³⁺, DTPA=diethylene triamine pentaacetic acid) were investigated by using indicator displacement assays (IDAs) (see picture). A remarkable selectivity for polyphosphorylated species was observed. This has been successfully used to modulate the relaxivity of the Gd³⁺ complex with pyrophosphate.

Full text of DOI:10.1002/chem.201001397

FULL PAPER
DOI: 10.1002/chem.201001397
A Pyrophosphate-Responsive Gadolinium ACHUTNGRNENUG( III) MRI Contrast Agent
[
a]
[b]
[c]
[d]
Andrew J. Surman, Cꢀlia S. Bonnet, Mark P. Lowe, Gavin D. Kenny,
[
d]
[b]
[a]
Jimmy D. Bell, ꢁva Tꢂth, and Ramon Vilar*
Abstract: This study shows that the re-
laxivity and optical properties of func-
tionalised lanthanide-DTPA-bis-amide
screened by using indicator displace-
ment assays (IDAs). Considerable se-
lectivity for polyphosphorylated species
(such as pyrophosphate and adenosine-
5’-triphosphate (ATP)) over a range of
other anions (including monophos-
phorylated anions) was apparent. In
addition, we show that pyrophosphate
modulates the relaxivity of the
tion being sufficiently large to be ob-
served in imaging experiments. To es-
tablish the binding mode of the pyro-
phosphate and gain insight into the
origin of the relaxometric modulation,
a series of studies including UV/Vis
and emission spectroscopy, lumines-
3
+
complexes (lanthanide=Gd
and
3
+
Eu
,
DTPA=diethylene triamine
pentaacetic acid) can be successfully
modulated by addition of specific
3
+
anions, without direct Ln /anion coor-
dination. Zinc(II)-dipicolylamine moi-
eties, which are known to bind strongly
to phosphates, were introduced in the
amide “arms” of these ligands, and the
interaction of the resulting Gd–Zn2
complexes with a range of anions was
cence lifetime measurements in H O
2
1
7
31
and D O, O and P NMR spectrosco-
2
gadolinium
A
H
U
G
R
N
U
G
py and nuclear magnetic resonance dis-
persion (NMRD) studies were carried
out.
Keywords: contrast agents · gadoli-
nium · lanthanides · MRI · sensors
[
1]
Introduction
vide invaluable anatomical information. Recently, there
has been increasing interest in developing responsive con-
trast agents, which would be able to report biological events
Interest in lanthanide ACHUTNGRNENUG( III) complexes has increased dramati-
[2]
cally in recent years as their unique magnetic and spectro-
scopic properties offer enormous potential for a variety of
imaging applications, especially as contrast agents for nucle-
ar magnetic resonance imaging (MRI). The current genera-
tion of contrast agents (CAs) in clinical use is able to pro-
by modulation of the relaxivity of the CAs. Responsive
MRI contrast agents have been reported, which respond to
several different conditions, analytes or biomolecules, such
[3]
[4]
[5]
[6]
as pH, temperature, p(O ), enzyme activity and metal
2
[7]
ions. As in the broader field of molecular recognition, CAs
that are responsive to anionic analytes have lagged behind
due to the intrinsic difficulty in binding anions selectively in
aqueous media, especially in vivo. This may be attributed to
several factors, especially high solvation energies, pH sensi-
tivity and, in the case of in vivo imaging, highly competitive
endogenous anions.
In spite of the above, in recent years examples of CAs
that interact with anions have appeared: either responding
to anion concentration, or by using a change in binding af-
finity to endogenous anions to signal interaction with a dif-
ferent analyte, most notably enzymes. Additionally, bind-
ing of endogenous anions has been frequently encountered
as an unintended consequence of attempts to increase con-
trast agent relaxivity (for example by reduction of the coor-
dination number of Gd or by reduction of steric hindrance
around free coordination sites). Generally, however, these
contrast agents interact with anions by direct lanthanide–
[
a] Dr. A. J. Surman, Dr. R. Vilar
Department of Chemistry
Imperial College London
London SW7 2AZ (UK)
Fax : (+44)2075941139
E-mail: r.vilar@imperial.ac.uk
[
[
b] Dr. C. S. Bonnet, Dr. ꢀ. Tꢁth
Centre de Biophysique Molꢂculaire-CNRS
Rue Charles-Sadron, 45071 Orleans Cedex 2 (France)
[8]
[
c] Dr. M. P. Lowe
[9]
Department of Chemistry
University of Leicester, Leicester, LE1 7RH (UK)
d] G. D. Kenny, Prof. J. D. Bell
Metabolic and Molecular Imaging Group, MRC
Clinical Sciences Centre, Hammersmith Hospital
Imperial College, London, W12 0HS (UK)
3
+
[10]
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem201001397.
Chem. Eur. J. 2011, 17, 223 – 230
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
223

anion coordination, limiting the use of other established
anion-binding moieties. As in other areas of anion recogni-
tion and sensing, a move towards modular design is desira-
ble, in which the anion-binding moieties are separate from
the “reporting” group (the lanthanide in this case), and
might allow the development of anion-responsive MRI
agents.
ther equilibrium, leading to the design of another generation
of complexes in which this divergent binding was preven-
[7d]
ted. Herein we report a study of the interaction between
the bis-zinc(II) complexes [Ln(L)Zn ](NO ) (with Ln=
AHCTUNGTRENNUNG
2
3 4
3
+
3+
Gd and Eu ) and different phosphate-containing anions.
We show that pyrophosphate (PPi) binds strongly to these
complexes resulting in a change in the relaxivity (Ln=Gd
3
+
)
3
+
The work herein presented, involves the incorporation of
and optical properties (Ln=Eu ) of the lanthanide com-
known selective phosphate-binding moieties into
gadolinium(III) complex, endeavouring to achieve signifi-
cant modulation of relaxivity on interaction with specific
a
plexes.
ACHTUNGTRENNUNG
phosphorylated species, without direct gadolinium
anion coordination.
A
H
U
G
R
N
U
G
Results and Discussion
Over the past few years, it has been shown that zinc(II)–
dipicolylamine (Zn–DPA) complexes bind strongly and se-
lectively to a range of phosphate species (including inorgan-
ic phosphates and phosphorylated organics such as phospho-
peptides and amino acids). Therefore, this type of com-
plex was chosen as the phosphate-binding moiety to be in-
Synthesis of complexes and characterisation of their anion
binding: The [Ln(L)] and [Ln(L)Zn ] ACHTUNGTRNEGNU( NO ) complexes
2 3 4
3
+
3+
(with Ln=Gd and Eu ) were synthesised by adaptations
of literature procedures (see the Supporting Information for
details of their syntheses). Briefly, alkylation of di-(2-pico-
lyl)amine with N-(2-bromoethyl)phthalimide followed by
[11]
[14]
corporated in our lanthanide
A
H
U
G
R
N
U
G
deprotection yielded amino ethyl di-(2-picolyl)amine. This
[7c]
plexes have been used to develop a variety of phosphate
hosts. Often, two or more Zn-DPA units are placed in either
a rigid or flexible arrangement to further improve selectivity
of the host for the specific phosphate guest. Several of the
resulting receptors have been used as chemical sensors by
amine was treated with DTPA-bis(anhydride)
to yield
ligand H L which was treated in situ with the corresponding
3
3
+
3+
LnCl to yield [Ln(L)] (Ln=Gd and Eu ). These neutral
3
complexes were purified by reverse-phase flash chromatog-
raphy. The [Ln(L)Zn ] ACHTUNGTRNENUG( NO ) complexes were prepared in
2 3 4
[11b,c,12]
using indicator displacement assays (IDAs).
The DTPA–bis(amide) complex (DTPA=diethylenetri-
aminepentaacetic acid) [Ln(L)] has previously been report-
situ by reaction of aqueous [Ln(L)] with two equivalents of
Zn(NO ) ·6H O. Quantitative formation of such complexes
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
ACHTUNGTRENNUNG
3
2
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
has been established in the literature, and although in this
case they may be isolated, they are extremely hygroscopic
making their handling and analytical characterisation diffi-
cult.
Once the lanthanide–zinc complexes had been prepared,
it was next necessary to assess their ability to bind a range
of different anions. For this, indicator displacement assays
[
13]
3+
ed as a prototype zinc-sensitive optical probe (Ln=Eu
or Tb ) and MRI contrast agent
plex was designed to bind Zn between the two converging
DPA “arms”, blocking water exchange and altering the opti-
cal/MRI properties of the complex. Unforeseen binding of
3
+
[7c,d]
3+
(Ln=Gd ). The com-
2
+
2
+
Zn ions by both DPA substituents independently to yield
[
Gd(L)Zn ]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(NO ) was thought to be responsible for a fur-
were carried out by using [Gd(L)Zn ] ACHTUNGTNERGUN( NO ) , which was
2 3 4
2
3
4
shown to bind to different dyes. With pyrocatechol violet
PV) a desirably strong binding, and considerable colour
(
change upon binding, was observed: bound PV is bright
blue, “free” PV is yellow-brown. Other dyes such as gallo-
cyanin (GC) underwent less obvious colour changes on
binding, but their weaker binding was interesting to contrast
with that of PV. Screening of a variety of common anions
for binding to the metalloreceptor in an IDA that uses PV
(
Figure 1) and GC (see the Supporting Information) showed
a clear selectivity for polyphosphate species (e.g., pyrophos-
phate and adenonsine-5’-triphosphate (ATP)) over mono-
phosphates and other anions; a degree of GC displacement
by phosphorylated amino acids was also observable (see the
Supporting Information).
The indicator binding stoichiometry was investigated by
constructing Jobs plots of PV binding to [Gd(L)Zn2]
ACHTUNGTRENNUNG( NO )
3 4
(
see the Supporting Information), in which the maximum
corresponded to a 1:1 stoichiometry (although considerable
asymmetry hinted at a more complex system). Titrations of
PV/[Gd(L)Zn2] CAHTUNGTRNEUNG( NO ) conjugates with displacing anions
3 4
gave responses consistent with simple 1:1 models, however,
subsequent work showed a 1:1 model to be inadequate (see
224
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 223 – 230

A Pyrophosphate-Responsive Contrast Agent
FULL PAPER
ing
formation
constants
logbHG =4.5ꢁ0.2 and logb
=
HG2
7
.4ꢁ0.2, corresponding to step-
wise
binding
constants
and
logKHG =4.5ꢁ0.2
logKHG2 =2.9ꢁ0.2. By using
these values (with the presump-
tion that there is little differ-
ence in the PPi binding of the
3
+
3+
Gd and the Eu complexes)
it was possible to fit the T data
1
by least-squares fitting to a the-
oretical
binding
model
(
Figure 4). This allowed us to
estimate the relaxivities (r ) of
1
the three different complexes
proposed to be in equilibrium.
Figure 1. Indicator displacement assays that use [Gd(L)Zn
ent anions. a) Shows the changes observed when 1:1 mixture of [Gd(L)Zn
2
]
A
H
U
G
R
N
U
(NO
3
)
4
, pyrocatechol violet and a range of differ-
(NO and pyrocatechol violet
were mixed with one equivalent (Lane 1) and ten equivalents (Lane 2) of the corresponding anion. b) Addi- The
tion of up to 100 equivalents of inorganic phosphate (Pi) to a 1:1 mixture of [Gd(L)Zn (NO and pyrocate-
2
]
A
H
U
T
E
N
N
3 4
)
r1
value
estimated
is
2
]
A
H
U
G
R
N
U
G
3 4
)
for
3
[Gd(L)Zn ]
A
H
U
G
E
N
N
(NO )
3 4
2
chol violet does not lead to any significant colour change, whereas addition of one equivalent of pyrophos-
phate (PPi) induces a clear colour change from blue (bound PV) to orange (unbound/displaced PV). Concen-
ꢂ1 ꢂ1
.13 mm s , whereas those for
the mono- and bispyrophos-
tration of [Gd(L)Zn
2 3 4
] ACHTUNGTRENNUNG( NO ) =0.05 mm, 10 mm HEPES, pH 7.4, RT. Abbreviations used in this figure: PhPi=
phenyl phosphate, Tere=terephthalate, T(P)=Fmoc-phosphotyrosine, S(P)=Fmoc-phosphoserine, AMP=ad- phate
complexes
and 4.22 mm s ,
are
ꢂ1 ꢂ1
ꢂ1 ꢂ1
enosine-5’-monophosphate, ATP=adenosine-5’-triphosphate.
1.97 mm
s
respectively (all at 500 MHz).
below) and unambiguous fitting of displacement data to a
more complex model to determine binding constants was
not possible.
Once strong binding to pyrophosphate had been estab-
lished, the effect on relaxivity of the [Gd(L)Zn ] ACTHUNGTRNEUNG( NO )
2 3 4
complex on interaction was investigated. T measurements
1
showed a modulation of the complexꢄs relaxivity on titration
with pyrophosphate solutions (see Figure 2a), which was
also observable in MRI images of sample tube “phantoms”
(
4.7 T, see Figure 2b). Negative control experiments with
[Gd(L)], in the absence of zinc(II) showed little response
over the same range of pyrophosphate concentration (see
the Supporting Information), confirming that the zinc(II)-
DPA moieties in [Gd(L)Zn2] ACHTUNGRTNEUNG( NO ) are responsible for the
3 4
interaction between the complex and pyrophosphate.
The inflection in the relaxometric response of
[
Gd(L)Zn ] ACHTUNGTRENNUNG( NO ) to PPi (see Figure 2a) on addition of
2 3 4
around one equivalent PPi suggests a sequential 1:1 and
then 1:2 binding, as described in Equation (1) (where KHG
@
KHG2, H denoting the host, a [Gd(L)Zn ] species, and G de-
2
noting the guest, PPi).
KHG
KHG2
½HꢀG
H½HGꢀG
H½HGꢀ2
ð1Þ
To gain further insight into these equilibria, the analogous
europium complex [Eu(L)Zn ](NO ) was prepared and its
ACHTUNGTRENNUNG
2
3 4
luminescence was studied. Small changes in the emission
spectrum of the complex were observed on addition of pyro-
phosphate (see Figure 3a). The subtle change in the form of
the hypersensitive DJ=2 transition (l=617 nm) is clear. On
titration, it was possible to fit the variation to a stepwise 1:1/
Figure 2. a) Plot of apparent relaxivity at 500 MHz of [Gd(L)Zn
2 3 4
] ACHTUNGTRENNUNG( NO )
as a function of the PPi concentration (with fitting of speciation model
III
from Eu
analogue titration data); [Gd(L)Zn
2
]
A
H
U
G
R
N
U
G
3
)
4
starting at
1
.21 mm. b) Image (4.7 T, 200 MHz) of sample tubes containing 1 mm
Gd(L)Zn (NO and increasing [PPi]: i) 0, ii) 0.75 and iii) 8 equiv PPi
(10 mm HEPES, pH 7.4, RT).
[
2
]
A
H
U
G
E
N
N
3 4
)
1
:2 model (by using SpecFit/32, 2006, see Figure 3b), yield-
Chem. Eur. J. 2011, 17, 223 – 230
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
225

R. Vilar et al.
3
1
Figure 5. P NMR spectra of a solution of PPi alone (bottom), and in so-
lution mixed with [Eu(L)Zn (NO : 0.5 (middle) and 8 equivalents
top) PPi compared to Eu (400 MHz, 228C, 10 mm HEPES, pH 7.4).
2
]
A
H
U
G
E
N
N
3 4
)
3
+
(
[
Eu(L)Zn ] ACHNUTGTNERNUG( NO ) were recorded (see Figure 5). The
2 3 4
P NMR spectrum of pyrophosphate in the absence of any
3
1
added europium complex showed a single resonance at d =
ꢂ6.3 ppm (Figure 5, bottom spectrum). Upon addition of a
small excess of [Eu(L)Zn ] ACHTUNGRTNEUNG( NO ) to PPi ([PPi]=0.5 equiv
2 3 4
3
+
compared to Eu ), it was expected that practically all PPi
would be coordinated in a 1:1 fashion (Figure 4). The
3
1
P NMR spectrum of this mixture gave a very broad set of
signals between d ꢃ ꢂ3.0 and 7.0 ppm and a single reso-
nance shifted far upfield (d = ꢂ121 ppm). In addition, the
sharp and intense resonance at d = ꢂ6.3 ppm (associated to
uncoordinated PPi) disappeared. These observations are
consistent with the pyrophosphate being held in close prox-
imity to the paramagnetic europium centre on convergent
binding between the two Zn-DPA “arms” in a 1:1 complex
Figure 3. a) Emission spectra of [Eu(L)Zn
ence of zero (c) and one (a) equivalent PPi. b) Variation of the
emission intensity (l=617 nm) of [Eu(L)Zn (NO on titration with PPi
data from two titrations are presented). The error bars represent arbitra-
2 3 4
] ACHTUNTGNERNU(G NO ) (1.0 mm) in the pres-
2
]
A
H
U
G
E
N
N
3 4
)
(
see Figure 2); the multiple resonances are indicative of a
(
number of inequivalent conformations in equilibrium. As
ry +/ꢂ 5% error. [Eu(L)Zn
2 3 4
] ACHTUNGTRNENUG( NO ) =0.8 mm (lex =395 nm, 10 mm
[15]
has been discussed extensively elsewhere,
in Ln-DTPA-
HEPES, pH 7.4, RT).
bisamide complexes, the DTPA ligands are present as multi-
AHCTUNGTRENNUNG
ple isomers (trans, syn, anti, cis) in solution, and hence, the
flexible amide “arms” may potentially adopt multiple con-
formations in each of these. It is likely that pyrophosphate
binds to more than one of these isomers and therefore, sev-
3
1
eral slightly different P environments are present, giving
rise to the several broad signals. The small resonance at d =
ꢂ121 ppm, suggests that in one of the conformations the py-
rophosphate is brought very close to, if not into contact
with, the europium centre. However, as will be discussed in
the next section, if direct coordination of pyrophosphate to
europium occurs it can only represent a very small propor-
tion of the species in equilibrium, because the hydration
number (q) measured for this system in the presence of PPi
is practically one (see Table 1) and no hydration equilibrium
is observed.
Figure 4. Calculated fractional speciation of [Eu(L)Zn
tion, either as free [Eu(L)Zn (c), [Eu(L)Zn
Eu(L)Zn (PPi) (a) species.
2
] in a model titra-
2
] ACHTUTGNENRN(GU PPi) (g), or
2
]
[
2
]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
2
Where pyrophosphate is present in a large excess ([PPi]=
3
+
With the aim of establishing the possible mode of binding
between the complexes and pyrophosphate, P NMR spec-
tra of PPi in the presence of 0, 0.5 and 8 equivalents
8 equiv compared to Eu ) a mixture of [Eu(L)Zn2]
ACHTUNGTRENNUNG( PPi)
2
3
1
and free PPi is expected (Figure 4). Indeed, this is consistent
with the recorded P NMR spectrum with one intense peak
3
1
226
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 223 – 230

A Pyrophosphate-Responsive Contrast Agent
FULL PAPER
Table 1. Measured lifetimes of [Eu(L)Zn
presence of 0, 0.75, and 8.00 equivalents PPi (10 mm HEPES, pH/pD 7.4,
Eu]=1 mm, RT).
2
]
A
H
U
G
R
N
N
(NO
3
)
4
luminescence in the
phosphate (see Table 1 and the Supporting Information) in-
dicate no significant change in the mean hydration state of
the europium ACHTUNGRNENUG( III) centre (q=1 throughout). The lack of a
[
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
[PPi]
t
H O [ms]
t
D O [ms]
q
calcd
2
2
hydration equilibrium was corroborated by variable temper-
ature high-resolution UV/Vis spectroscopic study in which
0
0
8
.00
.75
.00
0.59
0.62
0.59
2.46
2.45
2.52
1.08
0.98
1.07
[17]
the absorbances characteristic of multiple hydration states
were not observed (see the Supporting Information for fur-
ther details).
17
Variable-temperature transverse and longitudinal O re-
laxation rate and chemical shift measurements were made at
at d = ꢂ6.3 ppm (free pyrophosphate) and two distinct
peaks at d = ꢂ4.4 and ꢂ4.8 ppm. The latter two resonances
are consistent with a divergent 1:2 binding mode in which
each Zn-DPA unit is bound to one pyrophosphate (see
Figure 2), with one of the signals corresponding to the phos-
phorous bound to the zinc centre and the other correspond-
ing to the more distant (non-coordinated) phosphorous (see
Figure 2 for a schematic representation of this binding
mode).
11.7 T in aqueous solutions of [Gd(L)Zn ] ACHTUNGTENRUNG( NO ) with 0,
2 3 4
0.75 and 10 equivalents of pyrophosphate (with respect to
1
7
the complex). The O relaxation rates were analysed ac-
cording to the Solomon–Bloembergen–Morgan theory,
yielding the parameters shown in Table 2. The resulting ex-
1
7
Table 2. Parameters obtained from the analysis of O NMR data for the
Gd(L)Zn (NO complex with no PPi (A), 0.75 (B) and 10 equivalents
of PPi (C) in water. Reported data for monoaqua complexes of [Gd-
[
2
]
A
H
U
G
E
N
N
3 4
)
Because the majority of PPi in aqueous solution at pH 7.4
3
ꢂ
[19]
is monoprotonated HP O
(ꢃ83% at RT, calculated by
A
H
U
G
R
N
U
G
A
H
U
G
E
N
N
(DOTA=1,4,7,10-
2
7
[16]
using the CurTiPot software from pK values 9.32, 6.70,
a
A
B
C
A
H
U
G
E
N
N
ACHTUNGTRENNUNG
2.1 and 0.91), it is potentially useful to understand whether
2
ex
98[a]
PPi is deprotonated on binding. To endeavour to clarify the
situation, an un-buffered aqueous solution of the
k
0.28ꢁ0.02 0.17ꢁ0.01 0.31ꢁ0.02
3.3ꢁ0.2
4.1ꢁ0.2
6
ꢂ1
AHCTUNGTRENNUG[ 10 s ]
DH
ꢀ
25.5ꢁ0.2 20ꢁ0.3
570ꢁ15 430ꢁ12
25.3ꢁ0.8 19.4ꢁ0.3
19.6ꢁ0.2
683ꢁ11
24.2ꢁ0.3
51.6ꢁ1.4
49.8ꢁ1.5
[
[
Gd(L)Zn ] ACHTUNGTRENNUNG( NO ) complex (pH 7.4, initial concentration of
2 3 4
ꢂ1
A
H
U
G
R
N
U
G
]
]
Gd(L)Zn ]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(NO ) =0.4 mm) was titrated with
a
pH-
298[b]
R
[c]
[c]
2
3
4
t
103ꢁ10
90ꢁ15
ꢂ12 ꢂ1
matched un-buffered solution of PPi (pH 7.4, [PPi]=5 mm).
The pH of the titration solution was monitored by using a
digital pH-meter; the results are shown in Figure 6. The
[10
s
]
E
r
–
–
ꢂ1
AHCTUNGTRENNUNG[ kJmol
[
a] Exchange rate. [b] Rotational correlation time. [c] The t
R
values are
1
7
calculated from a simultaneous fit of O NMR and nuclear magnetic res-
onance dispersion (NMRD) data.
2
98
change rate, k , for the complex in the presence of
ex
0
.75 equivalents PPi is notably lower than that with no PPi
or where it is in large excess, which are similar, and of a sim-
3
+
ilar magnitude than those reported for other Gd –DTPA–
ꢂ6 ꢂ1
298
bis
A
H
U
G
R
N
N
(amide) complexes (generally, 0.22ꢅ10
s ).
s
<kex <0.45ꢅ
This supports the hypothesis of 1:1
ACHTUNGERTN(NUNG PPi) species predominating in the presence of
2
ꢂ6 ꢂ1 [18]
1
0
a
[
Gd(L)Zn ]
Figure 6. Changes of the pH on interaction of a complex ([Gd(L)Zn
(NO
(*) or [Gd(L)] (~)) and PPi between (pH 7.4 (un-buffered) aque-
ous solutions).
2
]-
0.75 equivalents PPi, in which the convergent binding of PPi
hinders water exchange without displacing coordinated
water, and of removal of this hindrance on divergent 1:2
binding (on addition of a large excess of PPi).
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
3 2
)
sharp drop in pH on addition of one equivalent of PPi clear-
ly demonstrates the displacement of protons from PPi on
The values calculated for the rotational correlation time,
, vary rather unusually: reduction on addition of
2
R
98
t
1
:1 binding. That the proton displacement effect is mediated
0.75 equivalents PPi, followed by an increase on addition of
a further excess. The tR in the presence of excess PPi being
considerably larger than that of the complex alone might be
4
+
298
by the binding interaction with [Gd(L)Zn2] , was con-
firmed by a similar titration with [Gd(L)], as a negative con-
trol.
rationalised by the bulk of the putative [Gd(L)Zn2]
ACHTUNGTRENNUNG( PPi)
2
species, and the rigidity conferred by electrostatic repulsion
2
98
Mechanism of relaxivity modulation: To study the mecha-
nism of this response, luminescence and absorbance spec-
troscopy was used to investigate whether this was mediated
of the two “arms”. The decrease in tR on binding one
equivalent of PPi (unusual on increase of M ) might be the
w
consequence of the formation of a more compact and more
spherical species. There might also be an extended hydro-
gen-bonding network around the less contracted complexes
(those without PPi or with an excess of PPi), resulting in
3
+
by changes in the Ln hydration state (q). Luminescence
lifetime measurements in H O and D O of [Eu(L)Zn ]-
2
2
2
ACHTUNGTRENNUNG( NO ) in the presence of 0, 0.75 and 8 equivalents of pyro-
3 4
Chem. Eur. J. 2011, 17, 223 – 230
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
227

R. Vilar et al.
more water molecules being held around the complexes.
used as contrast agent in a competitive environment such as
that encountered in the cell.
2
98
This would give rise to a longer tR for these species, and
the less spherical character of the species would also lead to
a greater outer-sphere contribution to relaxivity (especially
for the larger 1:2 complexes).
Experimental Section
2
98
To further investigate the possible role of changes in tR
,
1
All synthetic procedures including the description of general instrumen-
tation used for characterisation are described in detail in the Supporting
Information.
H NMRD profiles were also obtained for the same three
systems ([Gd(L)Zn2](NO ) with 0, 0.75 and 10 equiv) at
AHCTUNGTRENNUNG
3
4
2
5
98 K in the proton Larmor frequency range of 0.01–
00 MHz (see the Supporting Information). Due to the
Indicator displacement assays (IDA): The anion-binding ability of
[
2 3 4
Gd(L)Zn ] ACHTUNRGTNEG(NU NO ) was screened by examining the colour (naked eye)
large number of variables to be adjusted simultaneously,
lack of independent access to data on electronic properties
and relatively small variations in water proton relaxivity, it
is not possible to unambiguously fit the NMRD and
and spectroscopic (UV/Vis) changes on addition of salts to a 1:1 mixture
of the gadolinium–lanthanide complex and the selected indicator mole-
cule (i.e., one previously shown to change colour on interaction with the
complex). In each well of a 96-well plate a solution was prepared (in
1
2 3 4
0 mm HEPES buffer, pH 7.4) containing: [Gd(L)Zn ACHUTNRTGNEUNG( NO ) ] (0.05 mm),
17
O NMR data to give reliable values for the electronic and
an indicator (either pyrocatechol violet (PV) or gallocyanine, (GC),
rotational parameters. Qualitatively, the shape of the pro-
files observed does not vary dramatically on addition of PPi,
however this is not inconsistent with relatively small changes
0.05 mm) and a salt (at either 0.05 mm and 1 equiv, or at 0.5 mm and
+
+
+
1
4 4
0 equiv, all were Bu N , H N or Na salts, see Figure 1 for the studied
anions). Absorption spectra of the solutions were measured in 96-well
plates by using a UV/Vis spectrometer at room temperature (218C).
298
in t contributing to the modest degree of r modulation.
R
1
Relaxometric titration of complex [Gd(L)Zn
phate at 500 MHz: The modulation of [Gd(L)Zn
2
]
A
H
U
G
R
N
U
G
3
)
4
with pyrophos-
3 4
) relaxivity on
Taken together, these studies rule out hydration equilibria
and suggest that both changes in rotational correlation time
2
]
A
H
U
G
R
N
U
binding pyrophosphate was studied by measuring relaxivity of these com-
plex on titration with pyrophosphate. Pyrophosphate in deuterated aque-
3
+
and the exchange rate of Gd -coordinated water play a
role in mediating the observed relaxivity response. Because
second and outer-sphere relaxivity is not easily defined or
studied, it is likely that a greater solvent-accessible surface
area and greater possibility for hydrogen-bonding interac-
ous buffer solutions was added stepwise to a solution of [Gd(L)Zn
(NO in the same deuterated buffer (starting concentration=1.21 mm),
and the T relaxation time was measured. These measurements were per-
2
]-
A
H
U
G
R
N
U
G
3 4
)
1
formed by using a Bruker DRX-500 NMR spectrometer in a tempera-
ture-controlled environment (208C); an inversion-recovery pulse se-
quence (with the correct delay for a 908 pulse determined specifically for
tion also contribute to them, hence, the greater r with ten
1
each measurement) was applied to determine the T
1
relaxation time of
equivalents of PPi compared to the value with no PPi
added. Similarly, contraction of the complexꢄs solvent-acces-
sible surface area may also play a role in the reduction of r1
on binding one equivalent PPi.
the residual H O present in the solution, by using Bruker Topspin soft-
2
ware version 3.0a, 2009, for non-linear regression of the data produced;
relaxivity was calculated by using the following Equation (2):
1
=T1 obs ¼ 1=T1 diaꢂr
1
½Gdꢀ
ð2Þ
subtracting the diamagnetic contribution to T
A correction to take into account the difference of viscosities of H
1
, and accounting for [Gd].
Conclusion
2
O and
2
D O was then used to allow direct comparison with other relaxivity
[
20]
The [Ln(L)Zn ]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(NO ) complexes herein presented, have
data, by using Equation (3):
2
3 4
been shown by IDAs to bind polyanionic species, especially
polyphosphates, such as pyrophosphate, with a great degree
of selectivity over a range of other anions (including mono-
phosphates). Significant modulation in the relaxivity of
r
1
ðH
2
OÞ ¼ r
1
ðD
2
OÞ ꢄ ðhH2
O
=hD2
O
Þ
ð3Þ
ꢂ1
ꢂ1
(
the values hH2O =1.01 MPas and hD2O =1.25 MPas at 208C were
used).
[
Gd(L)Zn ] ACHTUNGTRENNUNG( NO ) has been shown on addition of pyrophos-
2 3 4
MRI imaging experiments: Imaging was performed on a 4.7 T (200 MHz)
Varian Direct Drive system by using VnmrJ version 2.3A software
(Varian, Palo Alto, CA, USA). A quadrature volume-coil was used, with
an outer diameter of 66 mm and an internal diameter of 38 mm. Aqueous
solutions (typically 1 mm) were scanned in 250ml PCR tubes in a Perspex
phate. The behaviour of the system may be explained by the
stepwise formation of 1:1 and 1:2 complexes between
[Gd(L)Zn ] ACHTUNGTRENNUNG( NO ) and PPi bringing about changes in rota-
2 3 4
tion dynamics and accessibility of both the inner and outer
coordination spheres of the complex to the bulk water, and
hence the efficiency of relaxation enhancement. With this,
the main motivation for carrying out these studies, namely,
to show that relaxivity can be modulated by binding of
anions outside of the inner coordination sphere of gadolini-
um, has been accomplished. It should, however, be noted,
holder placed inside the RF coil. For T
Gd]=1.0 mm was acquired by using saturation recovery, a series of spin-
echo scans were run with an echo of the following parameters: TE=
0.97 ms, FOV=40ꢅ40 mm, matrix=256ꢅ256, 2 mm thick coronal slice
and varying TR=0.1, 0.3, 0.5, 0.7, 1.0, 3.0, 5.0, 7.0, 10.0 and 15.0 ms. De-
tails of image analysis can be found in the Supporting Information.
1
measurements of the samples
[
1
[
Eu(L)Zn
2
] luminescence lifetime measurements for determination of q:
(NO in D O and
Luminescence lifetime measurements of [Eu(L)Zn
2
]
A
H
U
G
R
N
U
G
3
)
4
2
that the interaction of [Gd(L)Zn ]
A
H
U
G
R
N
U
G
2
H O solution (10 mm HEPES, pD 7.4 and pH 7.4, respectively) and in
2
3
4
phosphates such as ATP (as revealed by IDA), is likely to
prevent the use of this compound as a contrast agent re-
sponsive to a particular polyphosphate in a cellular environ-
ment (due to lack of specificity). Further work would be re-
quired to improve the selectivity of the complex if it is to be
the presence of 0, 0.75 and 8 equivalents PPi were recorded by using a
Jobin Yvon Horiba FluoroMax-P spectrometer (with DataMax for Win-
dows v2.2 software). Samples were held in a 10ꢅ10 mm quartz cuvette
and a cutoff filter (550 nm) was used to avoid second-order diffraction ef-
fects. Lifetimes were measured by direct excitation (l=395 nm) with a
short 40 ms pulse of light (500 pulses per point) followed by monitoring
228
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 223 – 230

A Pyrophosphate-Responsive Contrast Agent
FULL PAPER
the intensity of emission (l=595 nm, DJ=1) emitted during a fixed gate
time of 0.1 ms, at a delay time later. Delay times were set at 0.1 ms inter-
vals; excitation and emission slits were both set to 5 nm.
by using a pipette, with a pH reading taken between each addition by
using an electronic pH meter (typically allowing 2 min for equilibration).
Lifetimes are obtained by iterative least-squares fitting (Microsoft Excel
ꢂkt
Solver of the data obtained to a standard first order decay: Iobs =I
0
e
+
offset minimising in terms of the emission decay constant (k), where Iobs
is the observed intensity at time t, and I is the initial intensity. The lumi-
Acknowledgements
0
nescence lifetimes quoted (tH2O and tD2O) are the reciprocal of the respec-
tive decay constants (i.e., tH2O =1/kH2O). The hydration state (q) is then
This work has been supported by the EU project STRP 516982 (HET-
EROMOLMAT) and the Medical Research Council (MRC). A.J.S. and
R.V. thank EPSRC for a studentship and fellowship, respectively.
calculated from the emission decay constants obtained in H
2 2
O and D O
solutions by using Equation (4):
[
1] a) P. Caravan, J. J. Ellison, T. J. McMurry, R. B. Lauffer, Chem. Rev.
q
Eu ¼ 1:2½ðkH2
O
ꢂkD2
O
Þꢂ0:25ꢂ0:15ꢀ
ð4Þ
1
2
999, 99, 2293; b) S. Aime, M. Botta, E. Terreno, Adv. Inorg. Chem.
005, 57, 173.
where the correction of 2ꢅ0.075 for two amide NꢂH oscillators adjacent
[2] M. P. Lowe, Curr. Pharm. Biotechnol. 2004, 5, 519.
III
[21]
to the Eu is included, as established by Parker et al.
[3] a) S. Aime, A. Barge, D. D. Castelli, F. Fedeli, A. Mortillaro, F. U.
Nielsen, E. Terreno, Magn. Reson. Med. 2002, 47, 639; b) R. Hov-
land, C. Glogard, A. J. Aasen, J. Klaveness, J. Chem. Soc. Perkin
Trans. 2 2001, 929; c) N. Raghunand, C. Howison, A. D. Sherry, S.
Zhang, R. J. Gillies, Magn. Reson. Med. 2003, 49, 249; d) M. P.
Lowe, D. Parker, Inorg. Chim. Acta 2001, 317, 163; e) M. P. Lowe,
D. Parker, O. Reany, S. Aime, M. Botta, G. Castellano, E. Gianolio,
R. Pagliarin, J. Am. Chem. Soc. 2001, 123, 7601.
1
NMRD profile: The H NMRD profiles were recorded on a Stelar smar-
tracer FFC fast-field-cycling relaxometer covering magnetic fields from
ꢂ4
2
.35ꢅ10 to 0.25 T, which corresponds to a proton Larmor frequency
range of 0.01–10 MHz. The relaxivity at highest fields was recorded by
using a Bruker WP80 with the spinmaster smartracer PC NMR console
at variable field from 20 to 80 MHz. The temperature was controlled by
a VTC90 temperature control unit and fixed by a gas flow. Samples were
[
4] S. R. Zhang, C. R. Malloy, A. D. Sherry, J. Am. Chem. Soc. 2005,
27, 17572.
2 3 4
solutions of [Gd(L)Zn ] ACHTUNGTRENNUNG( NO ) (1 mm) in the presence of the required
1
amount of PPi (pH 7.4, 10 mm HEPES). The relaxivity at 200 MHz was
recorded by using a 4.7 T Varian Direct Drive system as described for
imaging experiments (see above). Relaxivity at 500 MHz was measured
[
5] a) L. Burai, E. Toth, S. Seibig, R. Scopelliti, A. E. Merbach, Chem.
Eur. J. 2000, 6, 3761; b) L. Burai, R. Scopelliti, E. Toth, Chem.
Commun. 2002, 2366; c) S. Aime, M. Botta, E. Gianolio, E. Terreno,
Angew. Chem. 2000, 112, 763; Angew. Chem. Int. Ed. 2000, 39, 747.
6] a) A. Y. Louie, M. M. Huber, E. T. Ahrens, U. Rothbacher, R.
Moats, R. E. Jacobs, S. E. Fraser, T. J. Meade, Nat. Biotechnol. 2000,
in D
ating and calculating relaxivity at described for relaxometric titrations
see above—T titrations with PPi).
3 4
O NMR analyses of [Gd(L)Zn ) on interaction with PPi: The
2
O buffer solutions by using a Bruker DRX-500 spectrometer, oper-
[
(
1
1
7
2
]
A
H
U
G
E
N
N
1
8, 321; b) A. L. Nivorozhkin, A. F. Kolodziej, P. Caravan, M. T.
Greenfield, R. B. Lauffer, T. J. McMurry, Angew. Chem. 2001, 113,
987; Angew. Chem. Int. Ed. 2001, 40, 2903; c) K. Hanaoka, K. Ki-
1
7
2
]
A
H
U
G
R
N
U
G
3 4
)
2
(
kuchi, T. Terai, T. Komatsu, T. Nagano, Chem. Eur. J. 2008, 14, 987.
7] a) W.-h. Li, S. E. Fraser, T. J. Meade, J. Am. Chem. Soc. 1999, 121,
ture range of 276–348 K, on a Bruker Avance 500 (11.75 T, 67.8 MHz)
[
spectrometer. The temperature was calculated according to previous cali-
1
413; b) W.-h. Li, G. Parigi, M. Fragai, C. Luchinat, T. J. Meade,
[
22]
bration with ethylene glycol and methanol. An acidified water solution
Inorg. Chem. 2002, 41, 4018; c) K. Hanaoka, K. Kikuchi, Y. Urano,
T. Nagano, J. Chem. Soc. Perkin Trans. 2 2001, 1840; d) K. Hanaoka,
K. Kikuchi, Y. Urano, M. Narazaki, T. Yokawa, S. Sakamoto, K. Ya-
maguchi, T. Nagano, Chem. Biol. 2002, 9, 1027; e) E. L. Que, C. J.
Chang, J. Am. Chem. Soc. 2006, 128, 15942; f) C. S. Bonnet, E. Toth,
Future Med. Chem. 2010, 2, 367.
1
7
was used as reference (HClO
) were measured by the inversion-recovery pulse sequence, and the
transverse relaxation times (T ) were obtained by the Carr–Purcell–Mei-
boom–Gill spin-echo technique.
spheres fitted into 10 mm NMR tubes, to eliminate susceptibility correc-
4
, pH 3.3). Longitudinal O relaxation times
[
23]
(T
1
2
[24]
The samples were sealed in glass
[
25]
17
tions to the chemical shifts. To improve sensitivity in the O NMR ex-
[
8] a) P. Atkinson, Y. Bretonniere, D. Parker, Chem. Commun. 2004,
1
7
17
2
periments, O-enriched water (10% H O, CortecNet) was added to the
solutions to yield around 2% O enrichment. The O NMR data have
been evaluated according to the Solomon–Bloembergen–Morgan theory
of paramagnetic relaxation.
formed by using Micromath Scientist ; the hydration number q=1 was
4
2
38; b) P. Atkinson, B. S. Murray, D. Parker, Org. Biomol. Chem.
006, 4, 3166; c) S. Aime, D. Delli Castelli, F. Fedeli, E. Terreno, J.
1
7
17
Am. Chem. Soc. 2002, 124, 9364; d) C. H. Huang, J. R. Morrow, J.
Am. Chem. Soc. 2009, 131, 4206; e) K. Nwe, C. M. Andolina, C. H.
Hang, J. R. Morrow, Bioconjugate Chem. 2009, 20, 1375; f) J. I.
Bruce, R. S. Dickins, L. J. Govenlock, T. Gunnlaugsson, S. Lopinski,
M. P. Lowe, D. Parker, R. D. Peacock, J. J. B. Perry, S. Aime, M.
Botta, J. Am. Chem. Soc. 2000, 122, 9674; g) R. S. Dickins, S. Aime,
A. S. Batsanov, A. Beeby, M. Botta, J. I. Bruce, J. A. K. Howard,
C. S. Love, D. Parker, R. D. Peacock, H. Puschmann, J. Am. Chem.
Soc. 2002, 124, 12697; h) R. S. Dickins, T. Gunnlaugsson, D. Parker,
R. D. Peacock, Chem. Commun. 1998, 1643; i) J. A. Mosely, B. S.
Murray, D. Parker, Eur. J. Mass Spectrom. 2009, 15, 145.
[
17]
17
The analysis of O NMR data was per-
[
26]
fixed having been independently confirmed.
5
7
High-resolution UV/Vis spectroscopy: UV/Vis spectra of the
D
0
! F
0
III
transitions of Eu complexes were obtained with a PerkinElmer Lambda
1
9 spectrometer in the region l
= 577–581 nm with data steps of
0
.01 nm. A constant temperature was maintained by using a liquid
heated/cooled cell with a 10 cm path length, and spectra were obtained at
98 and 338 K. Aqueous solutions of [Eu(L)Zn (NO were made up to
2
2
]
A
H
U
G
E
N
N
3 4
)
approximately 0.02m concentrations, and the pH adjusted to 7.4 with
small additions of aqueous NaOH/HCl; spectra were recorded in the
presence of 0, 0.75 and 8 equivalents PPi; data from multiple acquisitions
over some time (typically around 20 h) was combined by using the visual-
isation program running on a MATLAB platform version 6.5. The re-
corded spectra can be found in the Supporting Information.
[
9] M. Giardiello, M. P. Lowe, M. Botta, Chem. Commun. 2007, 4044.
[
[
10] D. Messeri, M. P. Lowe, D. Parker, M. Botta, Chem. Commun. 2001,
742.
2
11] a) H. N. Lee, Z. C. Xu, S. K. Kim, K. M. K. Swamy, Y. Kim, S. J.
Kim, J. Yoon, J. Am. Chem. Soc. 2007, 129, 3828; b) M. J. McDo-
nough, A. J. Reynolds, W. Y. G. Lee, K. A. Jolliffe, Chem. Commun.
2006, 2971; c) E. J. OꢄNeil, B. D. Smith, Coord. Chem. Rev. 2006,
250, 3068; d) S. Tamaru, I. Hamachi, Struct. Bonding 2008, 129, 95.
[12] B. P. Morgan, S. He, R. C. Smith, Inorg. Chem. 2007, 46, 9262.
[13] K. Hanaoka, K. Kikuchi, H. Kojima, Y. Urano, T. Nagano, Angew.
Chem. 2003, 115, 3104; Angew. Chem. Int. Ed. 2003, 42, 2996.
pH-metric binding titration: An un-buffered aqueous solution of a com-
plex (either [Gd(L)Zn
2
]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(NO
3
)
4
or [Gd(L)], 5 mm solution, ꢃ12 mL) was
made up, and its pH adjusted to 7.4 by using aqueous HCl/NaOH. A sim-
ilar solution was made up of tetrasodium pyrophosphate (also un-buf-
fered, pH adjusted to 7.4). Aliquots (typically 50 mL) of the pyrophos-
III
phate solution were added to the stirring solution of the Gd complex
Chem. Eur. J. 2011, 17, 223 – 230
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
229

R. Vilar et al.
[
[
14] a) C. Incarvito, M. Lam, B. Rhatigan, A. L. Rheingold, C. J. Qin,
A. L. Gavrilova, B. Bosnich, J. Chem. Soc. Dalton Trans. 2001, 3478;
b) O. Horner, E. Anxolabehere-Mallart, M. F. Charlot, L. Tcherta-
nov, J. Guilhem, T. A. Mattioli, A. Boussac, J. J. Girerd, Inorg.
Chem. 1999, 38, 1222.
15] a) C. F. G. C. Geraldes, A. M. Urbano, M. A. Hoefnagel, J. A.
Peters, Inorg. Chem. 1993, 32, 2426; b) C. F. G. C. Geraldes, A. M.
Urbano, M. C. Alpoim, M. A. Hoefnagel, J. A. Peters, J. Chem. Soc.
Chem. Commun. 1991, 656; c) H. Lammers, F. Maton, D. Pubanz,
M. W. Van Laren, H. Van Bekkum, A. E. Merbach, R. N. Muller,
J. A. Peters, Inorg. Chem. 1997, 36, 2527; d) Chemistry of Contrast
Agents in Medical Magnetic Resonance Imaging (Eds.: A. E. Mer-
bach, E. Toth), Wiley-VCH, Weinheim, 2001, p. 471.
[19] D. H. Powell, O. M. Ni Dhubhghaill, D. Pubanz, L. Helm, Y. S. Leb-
edev, W. Schlaepfer, A. E. Merbach, J. Am. Chem. Soc. 1996, 118,
9333.
[20] A. Nonat, C. Gateau, P. H. Fries, M. Mazzanti, Chem. Eur. J. 2006,
12, 7133.
[21] A. Beeby, I. M. Clarkson, R. S. Dickins, S. Faulkner, D. Parker, L.
Royle, A. S. de Sousa, J. A. G. Williams, M. Woods, J. Chem. Soc.
Perkin Trans. 2 1999, 493.
[
[
22] D. S. Raiford, C. L. Fisk, E. D. Becker, Anal. Chem. 1979, 51, 2050.
23] R. L. Vold, J. S. Waugh, M. P. Klein, D. E. Phelps, J. Chem. Phys.
1
968, 48, 3831.
[
[
24] S. Meiboom, D. Gill, Rev. Sci. Instrum. 1958, 29, 688.
25] A. D. Hugi, L. Helm, A. E. Merbach, Helv. Chim. Acta 1985, 68,
5
08.
[
16] CurTiPot software, 3.4.1 ed., I. G. R. Gutz, 2008.
[
26] Micromath Scientist, version 2.0, Salt Lake City, 1995.
[
17] E. Toth, L. Helm, A. E. Merbach, Chemistry of Contrast Agents in
Medical Magnetic Resonance Imaging, Wiley, New York, 2001,
pp. 45–119.
Received: May 20, 2010
Revised: September 16, 2010
Published online: December 7, 2010
[
18] L. Helm, A. E. Merbach, Chem. Rev. 2005, 105, 1923.
230
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 223 – 230