Pyridine-based lanthanide complexes combining MRI and NIR luminescence activities

Article abstract of DOI:10.1002/chem.201102310

A series of novel triazole derivative pyridine-based polyamino- polycarboxylate ligands has been synthesized for lanthanide complexation. This versatile platform of chelating agents combines advantageous properties for both magnetic resonance (MR) and opt

Full text of DOI:10.1002/chem.201102310

FULL PAPER
DOI: 10.1002/chem.201102310
Pyridine-Based Lanthanide Complexes Combining MRI and NIR
Luminescence Activities**
Cꢀlia S. Bonnet,[a] Frꢀdꢀric Buron,[b] Fabien Caillꢀ,[a, b] Chad M. Shade,[c]
Bohuslav Drahosˇ,[a] Laurent Pellegatti,[a, b] Jian Zhang,[c] Sandrine Villette,[a]
Lothar Helm,[d] Chantal Pichon,[a] Franck Suzenet,*[b] Stꢀphane Petoud,*[a, c] and
ꢁva Tꢂth*[a]
Abstract: A series of novel triazole de-
rivative pyridine-based polyamino–
polycarboxylate ligands has been syn-
thesized for lanthanide complexation.
This versatile platform of chelating
agents combines advantageous proper-
ties for both magnetic resonance (MR)
and optical imaging applications of the
corresponding Gd3+ and near-infrared
luminescent lanthanide complexes. The
thermodynamic stability constants of
the Ln3+ complexes, as assessed by pH
potentiometric measurements, are in
the range logKLnL =17–19, with a high
106 sꢀ1) is faster than that of clinically
used magnetic resonance imaging
(MRI) contrast agents and proceeds
number of emitted photons and better
detection sensitivity. The most conju-
gated system PheTPy, bearing
a
through
a
dissociatively activated
phenyl–triazole pendant on the pyri-
dine ring, is particularly promising as it
displays the lowest excitation and trip-
let-state energies associated with good
quantum yields for both Nd3+ and
Yb3+ complexes. Cellular and in vivo
toxicity studies in mice evidenced the
non-toxicity and the safe use of such
bishydrated complexes in animal ex-
periments. Overall, these pyridinic li-
gands constitute a highly versatile plat-
form for the simultaneous optimization
of both MRI and optical properties of
the Gd3+ and the luminescent lantha-
nide complexes, respectively.
mechanism, as evidenced by the posi-
¼
tive activation volumes (DV =7.2–
8.8 cm3 molꢀ1). The new triazole ligands
allow
a considerable shift towards
lower excitation energies of the lumi-
nescent lanthanide complexes as com-
pared to the parent pyridinic complex,
which is a significant advantage in the
perspective of biological applications.
In addition, they provide increased ep-
selectivity for lanthanides over Ca2+
,
silon values resulting in a larger
Cu2+, and Zn2+. The complexes are
bishydrated, an important advantage to
obtain high relaxivities for the Gd3+
chelates. The water exchange of the
Keywords: bimodal · lanthanides ·
luminescence · magnetic resonance
imaging · near infrared
Gd3+
complexes
(kex298 =7.7–9.3ꢃ
Introduction
ˇ
[a] Dr. C. S. Bonnet, Dr. F. Caillꢀ, Dr. B. Drahos, Dr. L. Pellegatti,
Dr. S. Villette, Prof. Dr. C. Pichon, Prof. Dr. S. Petoud, Dr. ꢁ. Tꢂth
Centre de Biophysique Molꢀculaire, CNRS
In the last decades, imaging applications have revolutionized
not only clinical diagnostics, but also cellular and integrative
biology. Today, in the field of molecular imaging, it is be-
coming more and more important to ascertain observations
obtained in one imaging modality by a complementary
imaging method. Each of the state-of-the-art clinical and
biological imaging modalities has its own advantages and
range of application, which makes them complementary for
a specific problem. Typically, magnetic resonance imaging
(MRI) is characterized by a superb spatial resolution and
low sensitivity, requiring high concentration of the contrast
agents to be detected. On the other hand, fluorescence and
luminescence techniques are endowed with high sensitivity
and low resolution. Some modalities, such as MRI or com-
puter tomography (CT) provide high-resolution morphologi-
cal images. In a bimodal application, they might allow the
precise anatomical localization of images obtained by mo-
lecular imaging techniques exclusively based on the visuali-
zation of an imaging probe, like in optical methods. Further-
rue Charles Sadron, 45071 Orlꢀans (France)
E-mail: stephane.petoud@cnrs-orleans.fr
eva.jakabtoth@cnrs-orleans.fr
[b] Dr. F. Buron, Dr. F. Caillꢀ, Dr. L. Pellegatti, Dr. F. Suzenet
Institut de Chimie Organique et Analytique
UMR 6005 CNRS, Universitꢀ d’Orlꢀans
rue de Chartres, 45067 Orlꢀans (France)
[c] Dr. C. M. Shade, Dr. J. Zhang, Prof. Dr. S. Petoud
Department of Chemistry
The University of Pittsburgh
219 Parkman Avenue
15260 Pittsburgh (USA)
[d] Prof. Dr. L. Helm
Laboratory of Inorganic and Bioinorganic Chemistry
Ecole Polytechnique Fꢀdꢀrale de Lausanne
BCH, 1015 Lausanne (Switzerland)
[**] MRI=magnetic resonance imaging; NIR=near infrared.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201102310.
Chem. Eur. J. 2012, 18, 1419 – 1431
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1419

more, in MRI or optical imaging, the modulation of the con-
trast/signal properties by physiological parameters allows
the development of responsive probes, whereas such appli-
cations are intrinsically impossible in nuclear imaging. In
this context, the successive and, in an optimal case, the si-
multaneous use of two or more imaging modalities becomes
prevalent and starts to enter also the clinical practice. In
order to respond to the increasing demand, the development
of bimodal instruments is in important progression.
Ideally, bimodal imaging probes integrate a reporter opti-
mized for each of the imaging techniques within the same
molecular entity, with the following main benefits: i) simpli-
fied chemical design, characterization, and formulation of
the imaging probe and ii) identical biodistribution, resulting
in crossed examinations by the different imaging modalities,
easier interpretation, and limited risk of false negatives or
false positives. Lanthanide complexes are particularly well
adapted for the common design of MRI and optical bimodal
probes because the lanthanide cations have different mag-
netic and optical properties but similar chemical reactivities.
Stable Gd3+ complexes are commonly used as contrast-en-
hancing agents in MRI. On the other hand, luminescent lan-
thanide complexes have unique advantages over organic flu-
orescent probes and luminescent semiconductor nanocrys-
tals: excellent spectral discrimination due to their narrow
emission bands, absence of re-absorption effects due to the
large energy gap between the absorption and emission
bands, long luminescence lifetimes, and high resistance to
photobleaching. The majority of MRI optical bimodal
agents reported so far are Gd3+ complexes or iron oxide
nanoparticles conjugated to fluorescent organic dyes.[1,2]
These probes have been mainly used to enable the colocali-
zation of the acquired MR and optical images providing a
powerful way to validate in vivo experiments. Bimodal
probes have to account for the differences in sensitivity of
the imaging modality. An MRI reporter (Gd3+ complex) has
to be administered in at least approximately 1000-fold
higher concentration than the luminescent optical probe.
The development of lanthanide chelators combining MRI
and luminescence activities has been relatively modest so
far.[3–9] The conditions needed to satisfy requirements for
both imaging techniques were often considered to be non-
compatible: the inner-sphere-coordinated H2O molecule(s)
required for MRI is expected to induce strong quenching of
the Ln3+ luminescence through non-radiative deactivation,
in particular when near-infrared (NIR) emitting lanthanides
are concerned.[10–13] In a preliminary communication, we
have reported two prototypes of a simple pyridine-based
scaffold for Ln3+ complexation that simultaneously satisfy
MRI and luminescence requirements.[14] The ligands Py and
MeOPy (Scheme 1) ensure i) high MRI efficacy when com-
plexed to Gd3+, due to bishydration of the chelate, ii) effi-
cient sensitization of a NIR-emitting Ln3+ when complexed
to Nd3+ due to the efficient energy transfer. They also fulfill
the requirements of thermodynamic stability and kinetic in-
ertness, prerequisite for biological applications of lanthanide
complexes. We were able to prove for the first time that the
Scheme 1. Pyridine-based ligands used for Ln3+ complexation.
presence of two H2O molecules bound to the Ln3+, benefi-
cial for MRI applications of the Gd3+ analogue, is not an
absolute limitation for the development of NIR luminescent
probes. The relaxivity, which describes the efficiency of a
Gd3+ complex as contrast enhancing agent in MRI, is direct-
ly proportional to the number of water molecules in the
inner hydration sphere of the cation. The increase of the hy-
dration number is the most straightforward way of increas-
ing the relaxivity at all magnetic fields. The formation of ter-
nary complexes in the presence of various endogenous
anions, such as carbonate, phosphate, or citrate, has been
previously excluded for LnPy complexes.[14]
The in vivo application of bimodal MR and luminescence
imaging agents based on lanthanide complexes by using a
common chelator remains to be demonstrated. The design
of ligands satisfying the requirements for both techniques is
a coordination chemistry challenge and we believe that the
results obtained, even without direct in vivo application, are
a major contribution to the development of bimodal imag-
ing probes.
A major limitation of the first two compounds for applica-
tion is the excitation wavelength, (the apparent maxima of
the electronic envelopes are located at l=270 and 260 nm,
respectively, for the complexes formed with Py and with
MeOPy). These excitation wavelengths correspond to high-
energy radiation, which can be harmful to the biological sys-
tems to be observed. Such wavelengths are also not compat-
ible with most commercially available instruments dedicated
to biological observations, such as microscopes. Here we
present novel triazole-containing derivatives of the parent
pyridinic ligand (Scheme 1) that allow a shift towards lower
excitation energies of the luminescent lanthanide complexes
while preserving good complex stability and water exchange
properties. Such a shift towards lower excitation energies is
highly important in the perspective of biological applica-
tions, and in addition, it also allows monitoring the biocon-
jugation reaction.
1420
www.chemeurj.org
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1419 – 1431

Pyridine-Based Lanthanide Complexes
FULL PAPER
These triazole derivatives can be considered as model
compounds of conjugation through click chemistry of the
chelating unit to various macromolecules for targeting pur-
poses. Protonation constants and complex-stability constants
have been determined by pH potentiometry for the novel li-
gands, as well as for the parent Py compound and its me-
thoxy derivative. The parameters describing water exchange
and rotational dynamics have been assessed by variable tem-
perature 17O NMR experiments for the Gd3+ complexes. In
addition, we have measured variable pressure 17O transverse
relaxation rates in order to gain insight into the water ex-
change mechanism. Photophysical studies have been carried
out to describe and quantify the luminescence properties of
the NIR-emitting lanthanide complexes, with a particular in-
terest for the effects of the substituents on the excitation
wavelengths of the complexes. Finally, a study involving cel-
lular, as well as in vivo toxicity tests on mice has been per-
formed in order to address the general concerns that are as-
sociated with the potential use of a bishydrated chelate as
imaging probe. Although several bishydrated Gd3+ com-
plexes have been proposed as MRI contrast agents, in vivo
experiments are restricted to few of them, including deriva-
tives of the DTTA (diethylenetriamine tetraacetate) chela-
tor.[15,16] Although the MRI evaluation of DTTA-based com-
plexes in mice did not reveal toxicity problems, their limited
thermodynamic stability and kinetic inertness raised con-
cerns about their possible toxicity. Therefore, it seemed par-
ticularly important to carry out a complete in vitro and in
vivo toxicity study with our bishydrated LnPy complexes.
ligands (i.e., Py and MeOPy).[14] The optimization of both
magnetic resonance and optical properties as well as target-
ing biological entities remain highly challenging. With these
challenges in mind, and considering the numerous bioor-
thogonal ligation reactions described in the literature,[17–20]
we focused on the click chemistry concept and more specifi-
cally on the copper-catalyzed azide–alkyne cycloaddition re-
action.[21–26] This strategy was motivated by the expected
positive influence of the extra triazole ring conjugated to
the pyridine chelator on the absorption properties of the re-
sulting enlarged antenna. Recently, Hovinen et al. have pre-
sented the synthesis of lanthanideACTHNUTRGNEUNG(III) chelates based on
1H-1,2,3-triazol-4-yl pyridine subunits B’ (X=N, Y =C,
Scheme 2) and concluded on its negative effect on the physi-
cal properties of the chelate.[27,28] In our case, the presence
of the azide group on the pyridine ring could allow a
copper-free click chemistry approach[29–33] and lead to the
direct conjugation of one nitrogen from the triazole ring B
(Y=N, X=C, Scheme 2), which is expected to have a more
important impact on the electronic structure and the physi-
cal properties of the corresponding complexes.[29–33] Conse-
quently, we decided to synthesize these novel chelators of
type B with the 1H-1,2,3-triazol-1-yl group substituted by a
phenyl (R’=phenyl, Y=N, X =C, Scheme 2) or an alkyl
substituent (R’=pentyl, Y=N, X =C, Scheme 2).
The synthesis of the key intermediate 1 was envisaged by
exploiting the ability of 4-halogenopyridines to easily react
with
(Scheme 3). We first exploited the convenient access to 4-
bromo-2,6-bis[N,N-bis-(ethoxycarbonylmethyl)aminome-
a nucleophile according to an SNAr mechanism
AHCTUNGTRENNUNG
thyl]pyridine (4) described by Takalo et al. (this compound
was synthesized as described for the corresponding tert-butyl
ester).[34] Starting from chelidamic acid, reaction with phos-
phorus pentabromide and subsequent trapping of the bis-
acid bromide intermediate with methanol generated the cor-
responding bis-methyl ester 3. The ester groups were re-
duced with sodium borohydride and the resulting diol was
brominated with phosphorus tribromide. Primary bromines
were substituted by ethyl iminodiacetate to afford com-
pound 4. Bromine at the C4 position was reacted with
sodium azide in hot DMF to give the desired intermediate
1.[35]
Results and Discussion
Syntheses of the ligands: We have recently described bimo-
dal luminescent and MRI contrast agents on pyridine-based
We then focused on the Huisgen 1,3-dipolar cycloaddi-
tion. A few examples of this click chemistry reaction have
been reported on lanthanide complexes,[36–39] whereas publi-
cations describing the reaction with metal-free ligands are
very scarce.[27] Starting with phenylacetylene as a model sub-
Scheme 2. Triazole derivative pyridine-based ligands.
Scheme 3. a) PBr5, 908C, 4 h; b) MeOH, 08C to RT, 12 h, 98%; c) NaBH4, EtOH, heating to reflux, 3 h, 94%; d) PBr3, heating to reflux, 12 h, 78%;
e) HN(CH2CO2Et)2, K2CO3, CH3CN, 508C, 12 h, 95%; f) NaN3, DMF, 1008C, 48 h, 48%.
ACHTUNGTRENNUNG
Chem. Eur. J. 2012, 18, 1419 – 1431
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1421

F. Suzenet, S. Petoud, ꢁ. Toth et al.
Table 1. Protonation constants measured in KCl (0.1m) at 298 K.
strate, numerous copper complexes (such as CuI,
CuBr·Me2S, or Cu(CH3CN)4PF6) were tested in THF in cat-
Py[a]
PyOMe[a]
C5TPy[b]
DTPA[c]
ACHTUNGTRENNUNG
LogKH
alytic or stoichiometric quantities, in the presence or ab-
sence of a base and under classical conditions, and did not
afford any triazole derivative. To solve this problem, we de-
cided to use a copper-coordinating solvent and found that a
catalytic amount of Cu+ in acetonitrile, which was heated to
reflux, provided very efficient reaction conditions. The two
new ligands C5TPy and PheTPy were isolated in 90 and
89% yield, respectively, after almost quantitative saponifica-
tion of the ethyl esters (Scheme 4).
logKH1
logKH2
logKH3
logKH4
logKH5
SlogKHi
8.95
7.85
3.38
2.48
–
8.91
7.85
3.01
2.42
–
8.11 (8)
7.55 (7)
3.28 (8)
2.3 (1)
–
10.59
8.65
4.28
2.73
2.06
28.31
22.66
22.19
21.24
[a] From reference [14]. [b] The numbers in bracket indicate the error in
the last digit. [c] From reference [44].
ethylene triamine pentaacetic acid (DTPA), the first proto-
nation constant of the three ligands is lower (1.5–2 log
units). This result can be likely explained by the first proto-
nation step of DTPA occurring on the central nitrogen
atom, then the second one on a terminal nitrogen atom ac-
companied by the proton transfer from the central to the
second terminal amine group, leaving the central nitrogen
atom deprotonated. The value of logKH3 is also higher for
DTPA than for our ligands as it corresponds to the protona-
tion of the central nitrogen atom that we do not observe,
and, in our case, it does correspond to the protonation of a
carboxylate function.
Scheme 4. Synthesis of the triazole derivatives.
As expected, there are no significant differences if we
compare logKH3 and logKH4 of the pyridine-based ligands,
as the structural changes that occur on the pyridine ring, are
too far away to affect the carboxylate moieties. In a similar
manner, the four protonation constants of Py and MeOPy
are very similar and the addition of the methoxy group does
not have a significant effect on the protonation constants.
The structural changes induced in compound C5TPy are
more important, in particular the conjugation of the aromat-
ic ring is significantly enhanced. In terms of protonation
constants, the main effect is observed on logKH1, which is
nearly one order of magnitude lower. This can be explained
by the +M mesomeric effect of the triazole ring, thus en-
hancing the electron density available on the pyridine nitro-
gen atom. Consequently, the hydrogen bonding between the
proton of the monoprotonated ligand and the pyridine nitro-
gen atom will be stronger than for the other ligands.
Protonation constants of the ligands and stability of the cor-
responding complexes: The potentiometric studies of three
representative ligands, namely Py, MeOPy, and C5TPy, show
the presence of four protonation constants for each ligand.
The titration curves are presented in Figure 1 and in the
Complex stability constants, logKML, and complex proto-
nation constants, logKMLH [Eq. (1) and (2)], have been de-
termined for complexes formed with various Ln3+ ions by
direct potentiometric titrations, as the formation of the com-
plexes was fast.
Figure 1. Potentiometric titration curves of solutions containing C5TPy
(L) (1.32 mm) with 0 or 1 equivalent of CaCl2, ZnSO4, CuCl2, or GdCl3 in
water. I=0.1m KCl, 298 K.
½MLꢁ
ð1Þ
ð2Þ
KML
¼
½Mꢁ½Lꢁ
Supporting Information (Figures S1 and S2), and the corre-
sponding calculated protonation constants are shown in
Table 1. The first two protonation constants correspond to
the protonation of the amine nitrogen atoms, whereas the
two last protonation steps occur on carboxylic functions.
The protonation of the pyridine nitrogen atom is not ob-
served in the experimental pH range (pH 2.5–11), as was
½MLHꢁ
KMLH
¼
½Hꢁ½MLꢁ
The titration curves obtained at a 1:1 molar ratio are pre-
sented in Figure 1, and Figures S1–S3 in the Supporting In-
formation. The stability constants obtained from the fitting
of the experimental curves for the different complexes are
summarized in Table 2. The stability constants of the various
the case with similar ligands containing
a pyridine
moiety.[40–42] For C5TPy, the protonation of the triazole ring
occurs at too low pH to be observed.[43] In comparison to di-
1422
www.chemeurj.org
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1419 – 1431

Pyridine-Based Lanthanide Complexes
FULL PAPER
Table 2. Stability constants, pm values, and selectivity constants of the
different complexes measured by potentiometric titration in KCl (0.1m)
at 298 K.
The stability constant of GdC5TPy (logKML ꢂ17), is more
than one order of magnitude lower than the other ones,
which can be related to the lower basicity of the ligand. The
pGd values calculated under the commonly used conditions
(pH 7.4, cGd =1, cL =10 mm, T=258C) are all in the same
order of magnitude and only slightly lower than those of
some commercially available MRI contrast agents (Table 2).
It has been demonstrated that the in vivo toxicity of Ln3+
complexes is related to the release of free Ln3+. This release
can be a consequence of Zn2+, Cu2+, and Ca2+ transmetalla-
tion in vivo. We have also investigated the complex stabili-
ties of those bioavailable cations (Table 2). In the case of
Cu2+ and Zn2+, the experimental data could be well fitted
with the introduction of a monoprotonated complex. To
take into account the competition with these endogenous
cations, Cacheris et al. have defined a Gd3+ selectivity con-
stant, Ksel.[45] The comparison of our ligands with classical
polyamino–polycarboxylate chelators, such as DTPA, shows
similar selectivity. The kinetic inertness of a lanthanide com-
plex is another key feature to determine its in vivo toxici-
ty.[46] As demonstrated in the preliminary communication,
the kinetic inertness of GdPy is remarkable for a bishydrat-
ed chelate and can be attributed to the absence of dinuclear
Zn2+ complexes that are an important driving force in the
transmetallation of GdDTPA, and to the presence of the
rigid pyridinic skeleton.[14] No dinuclear complexes were ob-
served with MeOPy and C5TPy at a 2:1 metal-to-ligand
ratio, and in general, similarly high kinetic inertness is ex-
pected for their chelates as well.
LogK
Py[a]
PyOMe[a]
C5TPy
DTPA[b]
logKNdL
logKGdL
18.76
18.60
18.1[c]
18.39(1)
–
15.84
15.3[c]
3.81
18.28
18.76
17.1(1)
17.2(1)
21.62
22.39
logKYbL
logKLuL
logKZnL
18.39(1)
–
16.30
–
22.64
22.46
18.2
17.0(1)
14.63(5)
logKZnLH
logKCaL
3.98
9.88
3.77(4)
8.8(1)
5.60
10.75
9.43
9.2[c]
15.69
3.45
logKCuL
logKCuLH
pGd[d]
15.64
3.37
17.62
6.76
14.55(7)
3.8(1)
17.03
21.2
4.80
18.87
7.06
17.44
7.07
[e]
logKsel
6.86
[a] From reference [14] except for log KYbL. [b] From reference [42].
[c] From reference [47]. [d] Logarithmic concentration of the free lantha-
nide ion (pLn=ꢀlog[Ln]free), calculated for [Ln]=10ꢀ6 m, [L]=10ꢀ5 m at
ꢀ1 ꢀ1
ꢀ1
=
pH=7.4) [e] Ksel =Ktherm(aHꢀ1+aCaꢀ1+aZnꢀ1+aCu
)
with aH
CaLA
1
+
=
KH1[H+]+KH1KH2[H+]2+…; aCaꢀ1 =K
CuLA
[Ca2+]; aCaꢀ1 =K [Zn2+]; aCu
ꢀ1
K
[Cu2+]; calculated for [Ca2+]=2.5 mm, [Zn2+]=50 mm, [Cu2+]=1 mm
at pH 7.4.
Gd3+ complexes lie between the stability constant of
GdDTPA–BMA (BMA=bis(methyl amide); logK=16.85)
and that of GdDTPA, DTPA having one more complexing
carboxylate unit. For a wide range of linear and macrocyclic
polyamino–polycarboxylate ligands, it has been shown that
the affinity for lanthanide ions is correlated to the basicity
of the molecule. Figure 2 shows the correlation between the
affinity constants of a series of cyclic and linear poly(amino-
carboxylate) ligands for Gd3+ and their basicity (SlogKH). It
can be seen that the affinities found for our ligands correlate
very well with those for the previously reported systems.
17O NMR measurements and water exchange: In addition to
the hydration number, the rate of water exchange and the
dynamics of rotation are the most important parameters
that influence the relaxivity, thus the MRI efficiency of a
Gd3+ complex. In order to assess the parameters character-
izing water exchange and rotational dynamics of this pyri-
dinic family of Gd3+ complexes, we have measured variable-
temperature transverse and longitudinal 17O NMR relaxa-
tion rates and chemical shifts at 11.7 T (Figure 3 and Figur-
es S4 and S5 in the Supporting Information). Given its low
solubility, GdPheTPy did not allow 17O NMR measure-
ments, which typically require concentrations above around
10 mm. The relatively low solubility of GdC5TPy also pre-
vented to calculate reliable values for the reduced chemical
shifts and longitudinal relaxation rates (the data were exper-
imentally measured but the difference between the
GdC5TPy solution and the diamagnetic reference was too
small). Luminescence lifetime measurements performed
previously on the analogous EuPy complex have proven the
Similar constants were obtained for Nd3+, Eu3+, Yb3+
,
and Lu3+ complexes, and no specific trend is detected along
the lanthanide series.
presence of two water molecules in the first coordination
[14]
sphere of the Ln3+
.
Given the similar structure of the
Figure 2. Correlation between the affinity constants of the Gd3+ com-
plexes and the basicity (SlogKa) of a series of linear and cyclic poly(ami-
three complexes (same ligand denticity), we assumed bishy-
dration for all complexes. This was confirmed by the experi-
mental values of the reduced 17O chemical shifts, which are
directly proportional to the hydration number q.
~
nocarboxylate) ligands ( :nitrilotriacetic acid (NTA), ethylenediaminete-
&
traacetic acid (EDTA), DTPA; : DO3A 1,4,7,10-tetraazacyclododecane-
1,4,7-triacetic acid, HP-DO3A (HP=hydroxypropyl), 1,4,7,10-tetraazacy-
As it can be seen in Figure 3 and Figures S4 and S5 in the
^
clododecane-1,4,7,10-tetraacetic acid (DOTA);
MeOPy, C5TPy).
correspond to Py,
Supporting Information, the 17O-reduced transverse relaxa-
Chem. Eur. J. 2012, 18, 1419 – 1431
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1423

F. Suzenet, S. Petoud, ꢁ. Toth et al.
stant, A/ꢀh, the rotational correlation time, tR298, and its acti-
vation energy ER, and the parameters describing electron
spin relaxation, namely the mean square of the zero field
splitting energy, D2, the correlation time for the modulation
of the zero-field splitting, tV298, whereas its activation
energy, EV has been fixed to 1 kJmolꢀ1. The parameters re-
sulting from the best fit are summarized in Table 3.
Table 3. Parameters obtained from the fitting of the 17O NMR relaxation
rates and chemical shifts.
GdPy
(9.3ꢃ0.5)
4.2[b]
(50.4ꢃ0.8)
52.2[b]
(+58ꢃ3)
(20.2ꢃ0.2)
(92ꢃ2)
(2.8ꢃ0.1)
(0.96ꢃ0.03) (1.00ꢃ0.03)
(ꢀ3.7ꢃ0.1) (ꢀ3.4ꢃ0.1)
(+8.8ꢃ0.9)
GdPyOMe
GdC5TPy
(7.7ꢃ0.4)
GdDTPA[a]
298
kex [106 sꢀ1
]
N
A
ACHTUNGTRENNUNG
3.3
¼
DH [kJmolꢀ1
]
N
(46.0ꢃ0.8)
51.6
¼
DS [Jmolꢀ1
)
N
A
+53.0
17.3
58[d]
25
0.46
ꢀ3.8
ER [kJmolꢀ1
]
N
n.d.[c]
n.d.
298
tRO [ps]
ACHTUNGTRENNUNG
298
tv [ps]
A
ACHTUNGTRENNUNG
D2 [1020sꢀ1
]
E
A/ꢀh [106 radsꢀ1
]
ACHTUNGTRENNUNG
¼
DV [cm3 molꢀ1
]
–
A
[a] From ref. [50]. [b] From ref. [6]. [c] Not determined due to the lack
298
of T1 data (due to low solubility of the complex) [d] tRH [ps].
17
~
Figure 3. Temperature dependence of reduced O transverse ( ) and
In general terms, bishydrated complexes are expected to
have faster water exchange than monohydrated chelates, be-
cause their inner sphere is inherently more flexible, which
facilitates the transfer to the transition state in the water ex-
change process. This has proved true, for instance, for the
Gd3+ complexes of the DTTA derivative ligands (kex298 =8–
9ꢃ106 sꢀ1)[50,51,52] with respect to GdDTPA. Likewise, for the
pyridine-based complexes the water exchange rates are two
to four times higher than that of GdDTPA and they are
very similar to the DTTA analogues. This means that al-
though the pyridine rigidifies the ligand skeleton itself, this
does not have any negative influence on the flexibility of
the inner sphere to limit water exchange. On the other
hand, the value of the exchange rate is little influenced by
appending a methoxy group or a triazole derivative on the
pyridine, which do not affect directly the coordination
sphere of the metal. A slightly lower water exchange rate
has been previously published for GdPy.[6]
&
longitudinal ( ) relaxation rates (top) as well as reduced chemical shifts
(bottom) of GdPy at 11.7 T. The curves represent the simultaneous fit to
the experimental data points.
tion rates (1/T2r) first slightly increase (up to ꢂ290 K), then
decrease with increasing temperature, indicating that the
complexes are in the intermediate and then, at higher tem-
peratures, in the fast exchange region. In this fast exchange
regime, 1/T2r is defined by the transverse relaxation rate of
the bound water oxygen, 1/T2m, which is in turn influenced
by the water exchange rate, kex, the longitudinal electronic
relaxation rate, 1/T1e, and the scalar coupling constant, A/ꢀh.
The reduced 17O chemical shifts are determined by A/ꢀh. The
transverse 17O relaxation is governed by the scalar relaxa-
tion mechanism and thus contains no information on the ro-
tational motion of the system. On the other hand, longitudi-
nal 17O relaxation is dominated by dipolar and quadrupolar
terms, both related to the rotational dynamics of the com-
plex. The transverse and longitudinal 17O relaxation rates
and chemical shifts have been analyzed according to the So-
lomon–Bloembergen–Morgan theory of paramagnetic relax-
ation (see the Supporting Information). In the analysis of
the data, several parameters have been fixed to common
The mechanism of the water exchange process can be de-
termined from variable pressure 17O transverse relaxation
measurements, which give access to the volume of activa-
¼
¼
tion, DV , a potent tool in assigning the mechanism.[53] DV
is defined as the difference between the partial molar
volume of the transition state and the reactants and is relat-
ed to the pressure dependence of the exchange rate constant
through Equation (3).
values. Among these, rGd has been fixed to 2.5 ꢅ, based
ꢀ
O
on available crystal structures and electron-nuclear double
Resonance (ENDOR) results,[48] and the quadrupolar cou-
pling constant, cACHTUNGTRENNUNG
(1+h2/3)1/2, has been set to the value for
(
)
z
DV
RT
1
tm
pure water (7.58 MHz).[49] The empirical constant describing
the outer sphere contribution to the 17O chemical shift, Cos,
was fixed to 0. In the fit, the following parameters have
been adjusted: the water exchange rate, kex298, the activation
T
ð3Þ
¼ kex ¼ ðkexÞ0 exp
ꢀ
P
where (kex)0T is the water exchange rate at zero pressure and
the temperature T. Thus, the exchange reaction is either
¼
enthalpy for water exchange, DH , the scalar coupling con-
1424
www.chemeurj.org
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1419 – 1431

Pyridine-Based Lanthanide Complexes
FULL PAPER
slowed down or accelerated by increasing pressure, when
was fixed to the values obtained in the fit of the variable
temperature data. When fitted, values identical to those
from the variable temperature fit were obtained. In the
analysis, only data measured at 708C have been included.
The experimental variable pressure 17O relaxation data to-
gether with the best fits are presented in Figure 4 and Fig-
ure S6 in the Supporting Information, and the resulting acti-
vation volumes are reported in Table 3. The positive activa-
tion volumes evidencing a dissociative exchange are also in
agreement with the positive values of the activation entropy.
The character of the water exchange is slightly less dissocia-
tive for the pyridine-based complexes than for GdDTPA.
GdDTPA has a highly dissociatively activated water ex-
change process, with an activation volume close to the limit-
¼
¼
DV is positive or negative, respectively. DV is assumed to
be a direct measure of the degree of bond forming and
bond breaking occurring in the transition state and the con-
current lengthening and shortening of non-exchanging
ligand-to-metal distances. The mechanism of the water ex-
change has been established for a large number of monohy-
drated Gd3+ complexes;[54] however, very few data are avail-
able on bishydrated chelates.[55] Transverse 17O relaxation
rates, 1/T2r, were measured at 9.4 T and two different tem-
peratures (25 and 708C) for GdPy and GdC5TPy. Although
no significant changes in the relaxation rates were observed
with pressure changes at 258C, at 708C 1/T2r increases with
pressure for both complexes (Figure 4 and Figures S6 and S7
ing value for
a
dissociative water exchange (+
12.9 cm3 molꢀ1).[57]
Concerning the other parameters determined in the fit of
the variable-temperature 17O NMR data, we note that the
rotational correlation times obtained from the 17O longitudi-
nal relaxation rates are in accordance with the size of the
complexes in comparison to similar small molecular weight
chelates (the value reported for GdDTPA in Table 3 corre-
ꢀ
sponds to the rotation of the Gd H vector, which is always
ꢀ
shorter than the rotational correlation time of the Gd O
vector).[58]
Photophysical properties: In order to evaluate the effect of
the different substituents on the pyridine ring from a photo-
physical point of view, we have recorded and analyzed the
UV/Vis absorption, excitation, and emission spectra of the
Nd3+ and Yb3+ complexes formed with the four ligands in
Figure 4. Pressure dependence of reduced 17O transverse relaxation rates
of GdPy at 9.4 T and 708C. The curve represents the simultaneous fit to
the experimental data points.
2-{4-(2-hydroxyethyl)-1-piperazinyl}ethanesulfonic
acid
(HEPES) buffer at pH 7. The absorption spectra (Figures S8
and S9 in the Supporting Information) reveal several inter-
esting features. All spectra contain several bands that are
well defined or that appear as shoulders located at signifi-
cantly different wavelengths one to another implying that
the addition of substituents on the central pyridine groups
induces significant modifications of the electronic structure
of the lanthanide sensitizers. As one of the goals of this
work is to develop lanthanide complexes that can be excited
at lower energy in order to minimize interactions with bio-
logical material, the PheTPy group appears as the most
promising sensitizer based on its absorption. On the other
hand, the sensitizer MeOPy induces a shift of the absorption
towards higher energy, a less favorable trend.
For all the complexes, the epsilon values indicate that we
are monitoring p–p* transitions centered on the chromo-
phore. The substitution of the pyridine group results in a
dramatic change of the epsilon of the complexes following
the order C5TPy>PheTPy>Py>MeOPy, if we consider
the absorption at wavelength above l=270 nm. This in-
crease in absorption (more than a factor 3 at l=270 nm) is
advantageous for the overall luminescence properties of the
complexes as the number of photons emitted by the lantha-
nide cations will be directly proportional to these epsilon
values, and allow for a more sensitive detection.
in the Supporting Information). At 258C, the contribution
of the water exchange rate to the measured 17O 1/T2r values
is relatively small (kex represents ꢂ20% in the correlation
time, 1/ts1 =kex+1/T1e, see the Supporting Information).
Therefore, no pressure effect is observable on the measured
T2 values. At 708C, the systems are in the fast exchange
regime where the transverse relaxation rates are proportion-
al to 1/kex. The increase of 1/T2r with pressure is thus due to
the slowing down of the water exchange process. This sug-
gests a dissociative interchange (Id) mechanism, confirmed
by the positive values calculated for the activation volume.
Ideally, the pressure-dependent measurements should be
performed in the slow exchange region, where the trans-
verse relaxation rates are directly determined by the water
exchange rate. If it is not possible, as in our case, it is advan-
tageous to do the measurements at higher temperatures,
where the contribution of the electron spin relaxation is less
important. A/ꢀh has been previously found independent of
pressure for different lanthanideACTHNUTRGNEUNG
(III) aqua ions.[56] As noth-
ing is known about the pressure dependence of the electron-
ic relaxation, in the analysis of the variable pressure
17O NMR data we have checked that a reasonable pressure
°
variation of tv (jDVv j ꢄ4 cm3 molꢀ1) had no significant
effect on the calculated activation volume. In the fit, (kex)T
0
Chem. Eur. J. 2012, 18, 1419 – 1431
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1425

F. Suzenet, S. Petoud, ꢁ. Toth et al.
The excitation spectra of the four Yb3+ complexes upon
monitoring the Yb3+-centered emission at l=978 nm are
displayed in Figure 5. The addition of the methoxy group on
cence of both Nd3+ and Yb3+. Upon excitation through the
ligands Py, MeOPy, C5TPy, and PheTPy at l=270, 250, 275,
and 330 nm, respectively, the luminescence spectra obtained
contain bands corresponding to the f–f NIR transitions of
Nd3+ or Yb3+. The spectra of the Yb3+ complexes formed
with the four ligands are characterized by a splitted band
ranging from l=920 to 1050 nm, which is assigned to the
2
2F5/2! F7/2 transition. Those of Nd3+ are characterized by
three main sharp bands at l=873, 1059, and 1328 nm, which
4
4
are attributed to the transitions from the F3/2 level to the I9/
4
4
2, I11/2, and I13/2 sublevels, respectively. Sensitization of the
NIR luminescence by the pyridine, and the pyridine–triazole
scaffolds is clearly demonstrated by the excitation spectra,
which closely overlay with the bands located at the lowest
energy observed on the absorption spectra for all eight lan-
thanide complexes (Figures S8–S13 in the Supporting Infor-
mation). This results indicate that all four chromophores are
able to act as antennae for Yb3+ and Nd3+ through the p–
p* transitions.
Figure 5. Normalized excitation spectra in HEPES buffer (0.1m, pH=
6.8) of YbPy (1.53 mm), YbMeOPy (1.63 mm), YbC5TPy (0.44 mm), and
YbPheTPy (1.32 mm) complexes; emission fixed at l=978 nm.
The sensitization process of Ln3+ ions usually involves
mainly the triplet excited states of the sensitizer of the Ln3+
ion. In order to systematically compare our different sensi-
tizer systems, we have quantified the energies of these excit-
ed states. Therefore, we have recorded the steady-state fluo-
rescence, and the time-resolved phosphorescence spectra of
the four Gd3+ complexes formed with the different sensitiz-
ers. Indeed, with this metal cation, no energy transfer to the
lanthanide can occur because the Gd3+ electronic levels are
too high in energy to accept a transfer from the singlet and/
or triplet states of the sensitizers used in this study. The
measurements of the fluorescence spectra obtained at room
temperature resulted in the identification of a unique broad
band that is assigned to the S1–S0 transition with apparent
maxima ranging from n˜ =27800 cmꢀ1 (l=360 nm) for Py to
n˜ =24400 cmꢀ1 (l=410 nm) for PheTPy (Table 4). In the
phosphorescence spectra recorded with delay after flash, the
band corresponding to the singlet state disappears, and a
band corresponding to the triplet state becomes apparent.
These bands have apparent maxima located at n˜ =
25400 cmꢀ1 (l=393 nm), n˜ =25900 cmꢀ1 (l=386 nm), n˜ =
the pyridine ring results in a significant shift of the excita-
tion band towards higher energy, detrimental for the use of
these complexes in biological conditions. On the other hand,
the addition of the triazole unit on the pyridine moiety is fa-
vorable by shifting the excitation bands towards lower ener-
gies. A red shift of approximately 20 nm is observed be-
tween the apparent maxima of excitation energies of YbPy
and YbPheTPy. The latter compound can even be excited at
wavelengths up to l=400 nm (using the side of the band lo-
cated at lower energy). Such excitation at lower energy is fa-
vorable for the sensitization of the NIR-emitting lanthanide
cations in biological media. This result is ascribed to the in-
crease of the conjugation by the addition of the phenyl ring
to the triazole group. A similar trend can be observed on
the excitation spectra of the Nd3+ complexes (Figure S10 in
the Supporting Information).
Spectra depicted in Figure 6 and Figures S11–S13 in the
Supporting Information clearly show that all chromophores
presented in this work are able to sensitize NIR lumines-
Figure 6. Absorption, normalized emission (lex =295 and 330 nm respectively), and excitation (lem =1060 and 978 nm respectively) spectra of NdPheTPy
(0.127 mm) and YbPheTPy (1.32 mm) in HEPES buffer (0.01m, pH 7.0).
1426
www.chemeurj.org
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1419 – 1431

Pyridine-Based Lanthanide Complexes
FULL PAPER
Table 4. Singlet and triplet state energies [cmꢀ1] recorded in aqueous sol-
utions of the complexes (HEPES buffer, pH 7).
quenching effect due to small energy gaps in these NIR-
emitting cations.
Py
MeOPy
C5TPy
PheTPy
It is interesting to compare the performances of these
Nd3+ and Yb3+ complexes to the corresponding complexes
formed with the ligand T2soxMe developed by Bꢆnzli
et al.[59] This tetrapodal chromophoric ligand has been de-
signed to sensitize Nd3+ and Yb3+ and to maximize their
protection from the non-radiative deactivation generated by
water molecules of the surrounding environment (no water
molecules are directly bound to both Nd3+ and Yb3+). Such
requirement was considered as a major parameter in order
to obtain complexes able to emit a large number of photons.
Comparing the Nd3+ complexes formed with both types of
ligand systems, it is exciting to notice that despite the pres-
ence of two water molecules directly bound to Nd3+, the
complexes formed with the various ligands display quantum
yields in the same order of magnitude than the one of
NdT2soxMe. Concerning the Yb3+ complexes, the quantum
yield values presented in this work are significantly smaller
than those formed with the T2soxMe sensitizer.
singlet[a]
triplet[b]
27800
25400
26500
25900
25700
23500
24400
22300
[a] Steady-state fluorescence spectra recorded at 298 K (1 mm). [b] Time-
resolved phosphorescence spectra recorded at 77 K with 0.1 ms delay
after flash. Apparent maxima of emission bands used for the calculation
of the energies of the electronic levels.
23500 cmꢀ1 (l=425 nm), and n˜ =22300 cmꢀ1 (l=448 nm)
for Py, MeOPy, C5TPy and PheTPy, respectively (see
Table 4 and Figure S14 in the Supporting Information). We
can observe an important shift (n˜ =3600 cmꢀ1) towards
lower energies by comparing the MeOPy systems to the
PheTPy systems. This difference in energy indicates that the
electronic structure of this family of sensitizers can be tuned
with a wide dynamic range upon substitution of the pyridin-
ic moiety.
In order to quantify the intramolecular energy transfer,
absolute quantum yields and luminescence lifetimes were
measured. Experimental luminescence lifetimes of the emis-
sionin the Yb3+ complexes formed with the four different li-
gands (upon ligand excitation) have been recorded, fitted,
and the results are summarized in Table 5. The fairly similar
values for all four complexes indicate that the quenching
can be assumed to be similar for the various complexes.
These similar quenching strengths indicate that the mea-
sured quantum yield values reflect mainly the effect of the
ligand-to-lanthanide energy transfer within the four com-
plexes.
It is worth noticing that despite the important changes of
the electronic structure of the sensitizers indicated by the
important differences in energy of their triplet states, the
corresponding quantum yields remain within the same
range.
We can draw several conclusions from the photophysical
measurements on this series of complexes. i) The presence
of the substituents induces significant changes in the epsilon
values of the corresponding NIR-emitting lanthanide com-
plexes. An increase in epsilon value upon substitution of the
central pyridine ring is useful as the detection sensitivity will
be directly proportional to the number of emitted photons,
and consequently to the epsilon value. ii) The introduction
of the methoxy group on the pyridine was detrimental as it
increases the excitation energies of the complexes, and re-
sulted in a decrease of the quantum yields. iii) In contrast to
what Hovinen has previously observed,[27] the introduction
of the triazole ring on the pyridine scaffold enables a signifi-
cant shift of the excitation wavelengths towards higher
values. The main difference between the two studies is that,
in our work, the triazole ring was appended through the ni-
trogen atom whereas they did it through a carbon atom.
The quantum yields were measured upon ligand excita-
tion and upon monitoring the lanthanide emission in water
at pH 7 and room temperature. The results are summarized
in Table 6. Those quantum yields are all within the same
order of magnitude with each other and are remarkable for
bishydrated complexes in aqueous solution, in which the
ꢀ
presence of two proximate O H oscillators induces a large
Table 5. Luminescence lifetimes of Yb3+ complexes formed with the cor-
responding ligands as obtained in aqueous solutions (HEPES buffer,
10 mm, pH 7); lex =266 nm.
Py
MeOPy
0.407(2)
C5TPy
0.43(3)
PheTPy
0.45(2)
Toxicity of the complexes: To assess the toxicity of this
family of bishydrated complexes, we have determined the
cell viability and the proliferation index of cells treated with
EuPy. In addition, the in vivo toxicity was also studied fol-
lowing chronic treatment of mice.
t [ms]
0.384(1)
Table 6. Absolute quantum yields of lanthanide complexes at 298 K in
HEPES buffer (0.01m, pH 7).
lexc [nm]
Ln3+
f [%]
Cytotoxicity evaluation: The effect on the cell viability was
first assessed by MTT assay. The MTT substrate is only re-
duced when mitochondrial reductase enzymes are active,
and therefore, the conversion can be directly related to the
number of viable cells. Hela and C6 cells were incubated
with different concentrations of EuPy during 24 h at 378C.
There was no effect on the viability of Hela cells whereas
C6 glioma cells appeared to be more sensitive, with (77ꢃ
Py
267
266
249
253
290
285
320
320
Nd3+
Yb3+
Nd3+
Yb3+
Nd3+
Yb3+
Nd3+
Yb3+
0.0097(9)
0.022(2)
0.0075(7)
0.017(2)
0.0095(9)
0.017(2)
0.007(1)
0.0068(9)
MeOPy
C5TPy
PheTPy
Chem. Eur. J. 2012, 18, 1419 – 1431
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1427

F. Suzenet, S. Petoud, ꢁ. Toth et al.
26)% of cells viable following their incubation with EuPy
(500 mm, Figure S15, in the Supporting Information). It
should be noted that this effect is not strong if we take into
account the long term incubation time. When it was carried
out only for 3 h at 378C, no significant cytotoxicity effect
was observed for either cell types (data not shown).
The effect of EuPy incubation on cell proliferation was
also evaluated by labeling cells with PKH26. This lipophilic
dye stably integrates into the cell membrane without dis-
turbing its surface marker expression and can be used to
track cell division because of the progressive halving of the
dye fluorescence intensity associated with the cell after each
division. Cells were labeled with PKH26 before EuPy incu-
bation. The cell proliferation index of these cells calculated
at one week post-treatment was compared to that of control
cells. In Figure 7 the proliferation index of cells treated with
vivo chronic toxicity that could be induced by the treatment.
Mice were checked up every day for close observations and
no signs of toxicity were observed. Again, no significant var-
iation of the body weight of mice treated chronically with
200 mmol EuPy per kg weight was observed (Figure S16B in
the Supporting Information). Vital organs (liver, spleen,
kidney, and lung) were harvested after 7 weeks for histopa-
thological analyses. No histopathology changes in terms of
inflammation, necrosis or structural tissue organization were
observed in vital organs of treated groups in comparison to
vital organs of untreated mice (Figure 8). It is worth noting
that no signs of inflammation were observed after 3 d post-
injection. In parallel, we collected blood samples of control
and treated mice to determine the level of transaminase en-
Figure 7. Effect of EuPy on the cell proliferation index. One week after
incubation, the proliferation index was measured by flow cytometry by
using the fluorescence intensity of PKH26.
different concentrations of EuPy was normalized to that of
the control condition (untreated cells). For all concentra-
tions tested, no significant effect on the cell proliferation
index was observed. These results indicate that EuPy incu-
bation did not alter any cellular mechanism that could stop
the proliferation of these cells.
In vivo evaluation of the toxicity: The absence of significant
cytotoxicity has prompted us to investigate the in vivo toxic-
ity of EuPy. We carried out a first set of experiments con-
sisting of one intra-peritoneal injection of EuPy at 8, 40, 80,
200, or 400 mmolkgꢀ1 in mice. Mice were daily examined for
any changes in morphology and behavior. Except for the
highest amount (400 mmolkgꢀ1), which was lethal to mice,
all mice have survived during 14 d without exhibiting any
abnormalities. They did not show any symptoms of toxicity,
such as fatigue, loss of appetite, change in fur color, or
weight loss. There was no significant variation of their body
weight during 14 d as shown in Figure S16A in the Support-
ing Information. A second set of experiments that consisted
of repeated intra-peritoneal injections (one per week during
6 weeks) of 200 mm of EuPy was performed to assess the in
Figure 8. Histological specimens of mice tissues (lung, kidney, liver, and
spleen) collected from mice euthanized at 7th week, stained with hema-
toxylin and eosin (H and E) showed normal histology. Mice were weekly
injected with 200 mmol EuPy (a, c, e, g) or 150 mm NaCl (b, d, f, h). Con-
trol animal (b) and EuPy-treated liver sections (a) showing normal hep-
atic structure. Control kidney (d) and EuPy-treated sections (c) showing
normal renal cortex and glomerular. Control animal (f) and EuPy-treated
lung sections (e) with no pathologic changes. Normal splenic architecture
of spleen sections from control animal (h) and from EuPy-treated mice
(g). Scale bar: 200 mm. Image magnification: 5ꢃ
1428
www.chemeurj.org
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1419 – 1431

Pyridine-Based Lanthanide Complexes
FULL PAPER
Potentiometric studies: Aqueous solutions of carbonate-free KOH
(0.1 molLꢀ1) and HCl (0.1 molLꢀ1) were prepared from Fisher Chemicals
concentrates. Potentiometric titrations were performed in an aqueous so-
lution of KCl (0.1 molLꢀ1) under nitrogen atmosphere and the tempera-
ture was controlled to ꢃ0.18C with a circulating water bath. The p[H]
value (p[H]=ꢀlog[H+], concentration in molarity) was measured in each
titration with a combined pH glass electrode (Metrohm) filled with KCl
(3m) and the titrant addition was automated by use of a 702 SM titrino
system (Metrohm). The electrode was calibrated in hydrogen ion concen-
tration by titration of HCl with KOH in 0.1 molLꢀ1 electrolyte solu-
tion.[60] A plot of meter reading versus p[H] allows the determination of
the electrode standard potential (E8) and the slope factor (f). Continuous
potentiometric titrations with HCl and KOH (0.1 molLꢀ1) were conduct-
ed on aqueous solutions containing Py (3 mL, 2.00 mm), MeOPy (3 mL,
3.00 mm), or C5TPy (5 mL, 1.32 mm) in KCl (0.1 molLꢀ1) with 2 min
waiting between successive points. The titrations of the metal complexes
were performed with the same ligand solutions containing one equivalent
of the metal cation, with 2 min waiting time between two titration points.
The same experiments were performed with two equivalents of Zn2+ as
well, to check the absence of dinuclear complex formation. Experimental
data were refined by using the computer program Hyperquad 2008[61] or
PSEQUAD.[62] All equilibrium constants are concentration quotients
rather than activities and are defined as:
zymes. There was no significant difference in the transami-
nase activity of blood samples from treated and control
mice ((1.44ꢃ0.163) mmolmLꢀ1 vs. (1.22ꢃ0.062) mmolmLꢀ1
of oxaloacetate). These measurements are in line with the
absence of liver alteration observed by histological analyses
as shown in Figure 8.
Biodistribution and pharmacokinetic studies will be re-
quired to determine how this agent is eliminated by the kid-
neys and liver.
Conclusion
We have synthesized a series of novel triazole derivatives of
a versatile pyridine-based scaffold for lanthanide complexa-
tion, which combine favorable properties for both optical
and MR imaging applications. These bishydrated lanthanide
complexes preserve good thermodynamic stability with a
good selectivity towards lanthanide ions with respect to en-
dogenous cations. The water exchange of the Gd3+ com-
plexes is faster than that of clinically used MRI contrast
agents and proceeds through a dissociatively activated
mechanism, as evidenced by the positive activation volumes.
Thanks to the triazole ring, these compounds allow a consid-
erable shift towards lower excitation energies of the lumi-
nescent lanthanide complexes as compared to the parent
pyridinic complex, an advantage in the perspective of bio-
logical applications. The most conjugated system of the
study, PheTPy, displays the lowest excitation and triplet-
state energies, with quantum yields in the same range than
the other complexes for both near-infrared-emitting Nd3+
and Yb3+ complexes. The PheTPy system combines three
advantages with respect to the luminescence properties:
i) increase of epsilon value resulting in a larger number of
emitted photons and detection sensitivity, ii) lower energy of
excitation, which becomes more suitable for biological appli-
cations, and iii) good quantum yields for NIR-emitting com-
plexes in aqueous conditions. Cellular and in vivo toxicity
studies evidenced the non-toxicity and the safe use of such
bishydrated complexes in animal studies. Overall, these pyri-
dinic ligands constitute a highly versatile platform for the si-
multaneous optimization of both MRI and optical properties
of the Gd3+ and the luminescent lanthanide complexes, re-
spectively.
½MmLlHhꢁ
h
Kmhl
¼
ð4Þ
m
l
½Mꢁ ½Lꢁ ½Hꢁ
The ionic product of water at 258C and 0.1 molLꢀ1 ionic strength is
pKw =13.77.[42] Fixed values were used for pKw, ligand acidity constants,
as well as total concentrations of metal, ligand, and acid. All values and
errors (one standard deviation) reported are at least the average of three
independent experiments.
Temperature-dependent 17O NMR measurements: The longitudinal and
transverse 17O relaxation rates (1/T1,2) and the chemical shifts were mea-
sured in aqueous solutions of GdL in the temperature range 280–350 K
on a Bruker Avance 500 (11.7 T, 67.8 MHz) spectrometer. The tempera-
ture was calculated according to previous calibration with ethylene glycol
and methanol.[63] An acidified water solution (HClO4, pH 3.3) was used
as external reference. Transverse relaxation times (T2) were obtained by
the Carr–Purcell–Meiboom–Gill spin-echo technique[64] and longitudinal
relaxation times were measured by the inversion recovery sequence.[65]
The technique of the 17O NMR measurements on Gd3+ complexes has
been described elsewhere.[66] The samples were sealed in glass spheres
fitted into 10 mm NMR tubes to avoid susceptibility corrections of the
chemical shifts.[67] To improve the sensitivity, 17O-enriched water (10%
H217O, CortectNet) was added to the solutions to reach around 1% en-
richment. The concentration and pH of the solutions were as follows :
[GdPy]=42.35 mm,
pH 6.67;
[GdMeOPy]=45.0 mm,
pH 6.60;
[GdC5TPy]=2.25 mm, pH 7.19. The 17O NMR data have been treated ac-
cording to the Solomon–Bloembergen–Morgan theory of paramagnetic
relaxation[68] (see the Supporting Information). The least-square fits of
the 17O NMR data were performed by using Micromath Scientist.[69] The
reported errors correspond to two times the standard deviation.
Pressure-dependent 17O NMR measurements: The transverse 17O relaxa-
tion rates (1/T2) were measured in aqueous solutions in the pressure
range 1–200 MPa at 708C on
a Bruker Avance 400 spectrometer
equipped with a home-made high-pressure probe head. The temperature
was controlled by circulating fluid from a temperature bath and was mea-
sured by means of a built-in Pt resistor. The pressure dependence of the
Experimental Section
Synthesis: See the Supporting Information.
transverse relaxation rate of acidified water, used as a reference, was de-
[70]
scribed by assuming an activation volume of +0.97 cm3 molꢀ1
.
The con-
Liquid sample preparation: The ligand concentrations were determined
by adding an excess of lanthanide solution to the ligand solution and ti-
trating the metal excess with standardized Na2H2EDTA in urotropine
buffer (pH 5.6–5.8) in the presence of xylenol orange as an indicator. The
concentration of the metal solutions was determined similarly by com-
plexometric titrations. For each lanthanide complex, after mixing the ap-
propriate amount of ligand and metal ion, the absence of free metal was
checked with the xylenol orange test.
centration and pH of the solutions were as follows: [GdPy]=17 mm,
pH 6.3; [GdC5TPy]=3 mm, pH 6.7.
Luminescence measurements: Solutions of complexes were prepared in
HEPES buffer (0.01m or 0.1m, pH 7) with the following concentrations:
[YbPy]=1.53 mm, [NdPy]=1.62 mm, [YbMeOPy]=1.63 mm, [NdMeO-
Py]=1.25 mm, [YbC5TPy]=0.44 mm, [NdC5TPy]=0.432 mm, [YbPheT-
Py]=1.32 mm, and [NdPheTPy]=0.123 mm. UV/Vis absorption spectra
Chem. Eur. J. 2012, 18, 1419 – 1431
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1429

F. Suzenet, S. Petoud, ꢁ. Toth et al.
were recorded at 258C by using a Perkin–Elmer Lambda 9 spectropho-
tometer. Emission and excitation spectra were measured by using a
modified Jobin–Yvon Horiba Fluorolog-322 spectrofluorimeter equipped
with an electro-optical systems (Inc. DSS-IGA02TL) cooled to 77 K
near-infrared detector. Singlet and triplet states were recorded on the
Gd3+ complexes upon appropriate excitation wavelength at room tem-
perature (fluorescence mode) and 77 K (phosphorescence mode with
0.1 ms delay), respectively. Luminescence quantum yields were collected
either with an integration sphere developed by Frꢀdꢀric Gumy and Prof.
Jean-Claude G. Bꢆnzli (Laboratory of Lanthanide Supramolecular
Chemistry, ꢁcole Polytechnique Fꢀderale de Lausanne (EPFL), BCH
1402, CH-1015 Lausanne, Switzerland) and manufactured by GMP
(Renens Switzerland), by using a quartz tube sample holder or in a 1 cm
quartz cuvette by using the relative method against a reference. With this
method, quantum yields were calculated by using Equation (5):
by measuring cell proliferation one week after sonoporation.[72] Two days
before EuPy incubation, Hela cells were labeled with PKH 26 according
to the manufacturerꢇs protocol (Sigma–Aldrich). The PKH 26 fluores-
cence intensity was measured by using its lexc and lem of 488 and approxi-
mately 585 nm, respectively. The rate of cell generation and the cell pro-
liferation index were estimated from the flow cytometry data by using
the ModFit LT software (CellQuestPro, Becton-Dickinson; Verity Soft-
ware House, http://www.vsh.com/products/mflt/index.asp).
In vivo toxicity studies: Mice were randomly divided into two groups
(n=4 mice per group): group 1: control, group 2: EuPy treated mice.
Each mouse of group 2 was administered with EuPy diluted in aqueous
NaCl (150 mm) at 200 mmolkgꢀ1 per week through intra-peritoneal injec-
tion during seven weeks. The control group mice were treated with
buffer alone. At the end of seven weeks, mice of the two groups were
starved over night and were euthanized on the next day to determine the
toxicity through examination of hematological and histological analysis.
Liver, spleen, kidneys, tendons, and mesenteric lymph nodes were har-
vested and fixed in p-formaldehyde (10%) at room temperature. Then
organs were embedded in paraffin wax (Tissue Tek, Sakura Bayer Diag-
nostics), sliced in serial sections, and stained with hematoxyline eosine sa-
franin (HES; Tissue Tek), and toxicity analyses were performed by No-
vaxia (St Laurent Nouan, France). Each slide was inspected microscopi-
cally by a pathologist to search for any evidence of damage such as adhe-
sions, necrosis, cellular infiltration, or hypervascularity.
2
Fx=Fr ¼ ½ArðlrÞ=AxðlxÞꢁ½IðlrÞ=IðlxÞꢁ½hx2=hr ꢁ½Dx=Drꢁ
ð5Þ
where subscript r stands for the reference and x for the sample. A is the
absorbance at the excitation wavelength, I is the intensity of the excita-
tion light at the same wavelength, h is the refractive index, and D is the
measured integrated luminescence intensity. Spectra were corrected for
the excitation (variations in lamp output and monochromator response)
and the emission (spectral response of the detector and of the emission
monochromator), and for the absorption of neutral density filters, where
applicable. The calculated values were determined by integrating the
emission profiles, averaged from three independent trials, and substitu-
tion into the ratio of emitted photons over absorbed photons. Lantha-
nide-centered luminescence lifetimes were measured at 298 K by using
either a Continuum Powerlite 8100 or a Quantel YG 980 (266 nm, fourth
harmonic) as the excitation source. Emission was collected at a right
angle to the excitation beam, and wavelengths were selected by using an
interferential filter (990 nm, BP20). The signal was monitored either by a
Hamamatsu H10330-45 or cooled R316-2 near-infrared detector, and was
collected on a 500 MHz band pass digital oscilloscope (Tektronix TDS
724C). Experimental luminescence decay curves were treated with
Origin 8.0 software (MicroCal LLC, Northampton, USA, 2004) by using
exponential fitting models. Three decay curves were collected on each
sample, and reported lifetimes are an average of at least three successful
independent measurements.
Serum analysis: The blood samples collected through cardiac puncture,
in sterile vials were immediately used for serum biochemical analysis,
which was carried out to determine the level of metabolic liver transami-
nase enzymes indicating liver damage. Transaminase activity was deter-
mined as described by Morin and Prox 1973.[73] The amount of oxaloace-
tate produced is indicative of the level of transaminase enzymes present
in the blood samples. Measurements were done in four mice treated with
EuPy (200 mm) and three control mice. Results are given as mean S.D.
Acknowledgements
This work was financially supported by the Institut National du Cancer,
La Ligue contre le Cancer, France, and was carried out in the frame of
the European COST Action D38. S.P. acknowledges support from Institut
National de la Santꢀ et de la Recherche Mꢀdicale (INSERM). Loꢈc
Lebꢉgue is acknowledged for his contribution in the toxicity studies.
Cell culture: Hela cells (ATCC CCL-2) were immortal cells that were de-
rived from the cervix adenocarcinoma of a young American female. C6
glioma cells (ATCC CCL-107) were from a rat glial tumor. Reagents for
cell culture were purchased from Invitrogen. Cells were cultured in Dul-
beccoꢇs modified Eagleꢇs Medium (DMEM) supplemented with 10%
heat-inactivated fetal calf serum (FCS), l-glutamine (2 mm), streptomycin
(100 unitsmLꢀ1), and penicillin (100 mgmLꢀ1) at 378C in a humidified at-
mosphere in the presence of 5% CO2. Their medium was changed two
times a week and subcultivated at a ratio of 1:2. The animal studies were
carried out according to the guidelines of the French Ministry of Agricul-
ture for experiments with laboratory animals (law 87848, accreditation of
C. Pichon). Assessment of cellular toxicity: One day before the experi-
ments, cells were seeded on multiwell plate at 2ꢃ105 per well. The cell
incubation was performed as following: the medium was discarded and
cells were rinsed with fresh medium. They were incubated at 378C either
for 3 or 24 h in the presence of indicated concentration of EuPy (0, 125,
250, and 500 mm) diluted in the cell medium. The cellular toxicity was
evaluated with the MTT and PKH26 assays as described below.
[1] L. Frullano, T. J. Meade, J. Biol. Inorg. Chem. 2007, 12, 939.
[2] W. J. M. Mulder, G. J. Strijkers, C. D. M. Donega, G. Storm, A. W.
Griffioen, K. Nicolay in Nanoparticles in Biomedical Imaging,
Springer, New York, 2007.
[3] R. J. Aarons, J. K. Notta, M. M. Meloni, J. Feng, R. Vidyasagar, J.
Narvainen, S. Allan, N. Spencer, R. A. Kauppinen, J. S. Snaith, S.
Faulkner, Chem. Commun. 2006, 909.
[4] S. G. Crich, L. Biancone, V. Cantaluppi, D. Duo, G. Esposito, S.
Russo, G. Camussi, S. Aime, Magn. Reson. Med. 2004, 51, 938.
[5] T. Koullourou, L. S. Natrajan, H. Bhavsar, S. J. A. Pope, J. Feng, J.
Narvainen, R. Shaw, E. Scales, R. Kauppinen, A. M. Kenwright, S.
Faulkner, J. Am. Chem. Soc. 2008, 130, 2178.
[6] S. Laurent, L. Vander Elst, M. Wautier, C. Galaup, R. N. Muller, C.
Picard, Bioorg. Med. Chem. Lett. 2007, 17, 6230.
[7] I. Nasso, C. Galaup, F. Havas, P. Tisnes, C. Picard, S. Laurent, L.
Vander Elst, R. N. Muller, Inorg. Chem. 2005, 44, 8293.
[8] A. Nonat, C. Gateau, P. H. Fries, M. Mazzanti, Chem. Eur. J. 2006,
12, 7133.
[9] G. Tallec, D. Imbert, P. H. Fries, M. Mazzanti, Dalton Trans. 2010,
39, 9490.
[10] H. Uh, S. Petoud, C. R. Chim. 2010, 13, 668.
MTT assay: Forty-eight hours after EuPy incubation, the cytotoxicity was
evaluated by using the colorimetric assay with MTT.[71] MTT (5 mgmLꢀ1
in phosphate-buffered saline (PBS)) was added to each well and incubat-
ed for 4 h. MTT was converted to an insoluble dye in living cells and was
then solubilized with acidic isopropanol. Absorbance was measured at a
wavelength of l=570 nm and expressed as a percentage of the absorb-
ance observed for non-treated cells cultured in the wells under the same
conditions as EuPy-treated cells.
PKH 26 assay: A simple, fast, and reliable functional PKH 26 assay
(Sigma–Aldrich) was used to determine the cytotoxicity effect of EuPy
[11] E. G. Moore, A. P. S. Samuel, K. N. Raymond, Acc. Chem. Res.
2009, 42, 542.
1430
www.chemeurj.org
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1419 – 1431

Pyridine-Based Lanthanide Complexes
FULL PAPER
[12] C. P. Montgomery, B. S. Murray, E. J. New, R. Pal, D. Parker, Acc.
Chem. Res. 2009, 42, 925.
[43] M. Witanowski, W. Sicinska, Z. Biedrzycka, Z. Grabowski, G. A.
Webb, J. Magn. Reson. Ser. A 1995, 112, 66.
[13] K. Binnemans, Chem. Rev. 2009, 109, 4283.
[44] P. Letkeman, A. E. Martell, Inorg. Chem. 1979, 18, 1284.
[45] W. P. Cacheris, S. C. Quay, S. M. Rocklage, Magn. Reson. Imaging
1990, 8, 467.
[46] L. Sarka, L. Burai, E. Brucher, Chem. Eur. J. 2000, 6, 719.
[47] S. Jakab, Z. Kovacs, L. Burai, E. Brucher, Magy. Kem. Foly. 1993,
99, 391.
[48] P. Caravan, J. J. Ellison, T. J. McMurry, R. B. Lauffer, Chem. Rev.
1999, 99, 2293.
[49] B. Halle, H. Wennerstrom, J. Chem. Phys. 1981, 75, 1928.
[50] D. H. Powell, O. M. NiDhubhghaill, D. Pubanz, L. Helm, Y. S. Lebe-
dev, W. Schlaepfer, A. E. Merbach, J. Am. Chem. Soc. 1996, 118,
9333
[14] L. Pellegatti, J. Zhang, B. Drahos, S. Villette, F. Suzenet, G. Guillau-
met, S. Petoud, E. Toth, Chem. Commun. 2008, 6591.
[15] J. B. Livramento, C. Weidensteiner, M. I. M. Prata, P. R. Allegrini,
C. Geraldes, L. Helm, R. Kneuer, A. E. Merbach, A. C. Santos, P.
Schmidt, E. Toth, Contrast Media Mol. Imaging 2006, 1, 30.
[16] P. L. de Sousa, J. B. Livramento, L. Helm, A. E. Merbach, W. Meme,
B. T. Doan, J. C. Beloeil, M. I. M. Prata, A. C. Santos, C. Geraldes,
E. Toth, Contrast Media Mol. Imaging 2008, 3, 78.
[17] J. Hovinen, P. M. Guy, Bioconjugate Chem. 2009, 20, 404.
[18] T. Kurpiers, H. D. Mootz, Angew. Chem. 2009, 121, 1757; Angew.
Chem. Int. Ed. 2009, 48, 1729.
[19] R. K. V. Lim, Q. Lin, Chem. Commun. 2010, 46, 1589.
[20] A. S. Raghavan, H. C. Hang, Drug Discovery Today 2009, 14, 178.
[21] M. D. Best, Biochemistry 2009, 48, 6571.
[22] D. C. Dieterich, Curr. Opin. Neurobiol. 2010, 20, 623.
[23] V. Hong, S. I. Presolski, C. Ma, M. G. Finn, Angew. Chem. 2009, 121,
10063; Angew. Chem. Int. Ed. 2009, 48, 9879.
[24] H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. 2001, 113,
2056; Angew. Chem. Int. Ed. 2001, 40, 2004.
[25] R. K. V. Lim, Q. Lin, Sci. China: Chem. 2010, 53, 61.
[26] G. C. Tron, T. Pirali, R. A. Billington, P. L. Canonico, G. Sorba,
A. A. Genazzani, Med. Res. Rev. 2008, 28, 278.
[51] J. B. Livramento, L. Helm, A. Sour, C. OꢇNeil, A. E. Merbach, E.
Toth, Dalton Trans. 2008, 1195.
[52] J. B. Livramento, A. Sour, A. Borel, A. E. Merbach, V. Toth, Chem.
Eur. J. 2006, 12, 989.
[53] S. F. Lincoln, A. E. Merbach, A. G. Sykes in Advances in Inorganic
Chemistry, Academic Press, New York, 1995, p. 1.
[54] L. Helm, A. E. Merbach, Chem. Rev. 2005, 105, 1923.
[55] M. K. Thompson, M. Botta, G. M. Nicolle, L. Helm, S. Aime, A. E.
Merbach, K. N. Raymond, J. Am. Chem. Soc. 2003, 125, 14274.
[56] C. Cossy, L. Helm, A. E. Merbach, Inorg. Chem. 1989, 28, 2699.
[57] T. W. Swaddle, M. K. S. Mak, Can. J.Chem. 1983, 61, 473.
[58] F. A. Dunand, A. Borel, A. E. Merbach, J. Am. Chem. Soc. 2002,
124, 710.
[27] J. Ketola, J. Kataffisto, H. Hakala, J. Hovinen, Helv. Chim. Acta
2007, 90, 607.
[28] J. Hovinen, WO/2008/025886, 2008.
[59] S. Comby, D. Imbert, C. Vandevyver, J.-C. G. Bꢆnzli, Chem. Eur. J.
2007, 13, 936.
[60] A. E. Martell, R. J. Motekaitis, Determination and Use of stability
Constants, VCH, Weinheim, 1992.
[29] J. M. Baskin, J. A. Prescher, S. T. Laughlin, N. J. Agard, P. V. Chang,
I. A. Miller, A. Lo, J. A. Codelli, C. R. Bertozzi, Proc. Natl. Acad.
Sci. USA 2007, 104, 16793.
[30] J. A. Codelli, J. M. Baskin, N. J. Agard, C. R. Berozzi, J. Am. Chem.
Soc. 2008, 130, 11486.
[31] M. F. Debets, C. W. J. van der Doelen, F. Rutjes, F. L. van Delft,
ChemBioChem 2010, 11, 1168.
[61] P. Gans, A. Sabatini, A. Vacca, Talanta 1996, 43, 1739.
[62] L. Zꢀkany, I. Nagpal in Computational Methods for the Determina-
tion of Formation Constants (Ed.: D. J. Leggett), Plenum Press, New
York, 1985, p. 291.
[32] J. C. Jewett, C. R. Bertozzi, Chem. Soc. Rev. 2010, 39, 1272.
[33] X. H. Ning, J. Guo, M. A. Wolfert, G. J. Boons, Angew. Chem. 2008,
120, 2285; Angew. Chem. Int. Ed. 2008, 47, 2253.
[34] H. Takalo, V. M. Mukkala, H. Mikola, P. Liitti, I. Hemmila, Biocon-
jugate Chem. 1994, 5, 278.
[35] Z. E. Chamas, X. M. Guo, J. L. Canet, A. Gautier, D. Boyer, R.
Mahiou, Dalton Trans. 2010, 39, 7091.
[36] J. M. Bryson, W. J. Chu, J. H. Lee, T. M. Reineke, Bioconjugate
Chem. 2008, 19, 1505.
[37] M. Jauregui, W. S. Perry, C. Allain, L. R. Vidler, M. C. Willis, A. M.
Kenwright, J. S. Snaith, G. J. Stasiuk, M. P. Lowe, S. Faulkner,
Dalton Trans. 2009, 6283.
[38] D. E. Prasuhn, Jr., R. M. Yeh, A. Obenaus, M. Manchester, M. G.
Finn, Chem. Commun. 2007, 1269.
[39] G. J. Stasiuk, M. P. Lowe, Dalton Trans. 2009, 9725.
[40] N. Chatterton, C. Gateau, M. Mazzanti, J. Pecaut, A. Borel, L.
Helm, A. Merbach, Dalton Trans. 2005, 1129.
[41] Z. Pꢊlinkꢊs, A. Roca-Sabio, M. Mato-Iglesias, D. Esteban-Gꢂmez,
C. Platas-Iglesias, A. de Blas, T. Rodrꢋguez-Blas, E. Tꢂth, Inorg.
Chem. 2009, 48, 8878.
[63] D. S. Raiford, C. L. Fisk, E. D. Becker, Anal. Chem. 1979, 51, 2050.
[64] S. Meiboom, D. Gill, Rev. Sci. Instrum. 1958, 29, 688.
[65] R. L. Vold, J. S. Waugh, M. P. Klein, D. E. Phelps, J. Chem. Phys.
1968, 48, 3831.
[66] K. Micskei, L. Helm, E. Brucher, A. E. Merbach, Inorg. Chem.
1993, 32, 3844.
[67] A. D. Hugi, L. Helm, A. E. Merbach, Helv. Chim. Acta 1985, 68,
508.
[68] E. Toth, L. Helm, A. E. Merbach in The Chemistry of Contrast
Agents in Medical Magnetic Resonance Imaging (Eds.: A. E. Mer-
bach, E. Toth), Wiley, New York, 2001.
[69] Micromath Scientist,] version 2.0, Micromath Research, Salt Lake
City (USA), 1997.
[70] Y. Ducommun, W. L. Earl, A. E. Merbach, Inorg. Chem. 1979, 18,
2754.
[71] T. Mosmann, J. Immunol. Methods 1983, 65, 55.
[72] P. K. Horan, S. E. Slezak, Nature 1989, 340, 167.
[73] L. G. Morin, J. Prox, Clin. Chem. 1973, 19, 776.
[42] R. M. Smith, R. J. Motekaitis, A. E. Martell, National Institute of
standards and technology ed., NIST Standard Reference Database,
1997.
Received: July 26, 2011
Published online: December 30, 2011
Chem. Eur. J. 2012, 18, 1419 – 1431
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1431