ꢀ. Tꢁth, P. Durand et al.
more slowly than the proton at d=ꢀ23 ppm. A similar dif-
ference in the exchange rate has been previously reported
for the two magnetically non-equivalent NH2 protons of the
Yb3+-DOTAM complex.[35] To the best of our knowledge,
no proton-exchange rates have been reported for amine
groups in metal complexes. The exchange rate of the amine
function of acetyl-lysine-NH2 was 4000 sꢀ1, whilst exchange
rates for arginine of 700 sꢀ1 and 1200 sꢀ1 correspond to the
eNH and hNH2 groups, respectively.[36,37] For the amine pro-
tons of both complexes YbL2 and YbL3, we observe a sys-
tematic decrease of the exchange rate with increasing pH
value, which is also translated by a diminution of the ob-
served CEST effect. This trend is similar to that observed
for aniline protons[37] and for the amine of the TmL2 com-
plex.[38]
However this trend is opposite to the typical behavior of
the amide protons in Ln3+-DOTAM complexes, for which
the proton exchange is a base-catalyzed process, with a con-
comitant increase of the CEST effect with increasing basici-
ty.[39] For instance, the CEST effect measured in a 30 mm so-
lution of the tetraglycinate-derivative Yb-DOTAM-Gly
changes from 0 to about 65% between pH 5.5 and 8.0,[39]
whilst for complex YbL3, it decreases from 65% at pH 6.3
to 15% at pH 9.0. The aniline-like behavior of the amine
groups of complexes YbL2 and YbL3 could be explained by
their near-sp2 hybridization caused by lanthanide coordina-
tion of the gem-diamine, moiety as previously discussed. For
the proton exchange of amines, both acid- and base-catalysis
has been reported.[36] It is interesting to note that a plot of
log kex values versus the pH value for both complexes (and
for both protons of YbL2) gives a straight line, in accordance
with the general formula of acid catalysis (kex =k0+kH[H+];
also see the Supporting Information).
various pH values by using the dependency of the CEST
effect on concentration, saturation time, and saturation
power. We also applied a concentration-independent analy-
sis of the saturation power dependency data, as recently
proposed by Sherry and co-workers.[33] These different meth-
ods gave similar results. In contrast to the typically base-cat-
alyzed amide-proton exchange on [LnACTHNUTRGNEUNG(dota)]/tetraamide
complexes, the exchange of the amine protons becomes
faster upon decreasing the pH value, which then leads to an
increasing CEST effect at lower pH values.
Experimental Section
NMR measurements: 1H and 13C NMR spectra were recorded at
300 MHz and 75 MHz, respectively, and calibrated using tetramethylsi-
lane as an internal reference. Variable-temperature 17O NMR measure-
ments of aqueous solutions of the GdIII complexes were obtained on a
Bruker Avance 500 spectrometer (11.75 T, 67.8 MHz) and referenced to
an acidified water solution (aqueous HClO4, pH 4). Longitudinal
17O NMR relaxation times (T1) were measured by the inversion-recovery
pulse sequence,[40] and the transverse relaxation times (T2) were obtained
by the Carr–Purcell–Meiboom–Gill spin-echo technique.[41] To eliminate
susceptibility corrections to the chemical shifts, the samples were sealed
in glass spheres that fitted into 10 mm NMR tubes.[42] To improve sensi-
tivity in the 17O NMR spectra, 17O-enriched water (10% H217O, Cortec-
net) was added to the solutions to yield 1% 17O enrichment. The temper-
ature was calculated according to a previous calibration with ethylene
glycol and methanol.[43] The concentrations and pH values of the samples
were:
[GdL1]=14.2 mmolkgꢀ1
,
pH 7.10;
[GdL2]=13.7 mmolkgꢀ1
,
,
pH 4.50; [GdL2]=10.3 mmolkgꢀ1
,
pH 6.60; [GdL3]=24.7 mmolkgꢀ1
pH 6.90; [GdL4]=12.6 mmolkgꢀ1, pH 7.20. The pH values of the solu-
tions were adjusted by using diluted solutions of NaOH and HCl. The
saturation-transfer experiments were performed on a Bruker Avance 500
spectrometer by irradiating the sample at 0.1 ppm increments. CEST
spectra were recorded using pre-saturation pulses of 3 s duration at
25 mT, unless otherwise stated. Spectra were measured by recording the
signal intensity of the bulk water as a function of the presaturation fre-
quency. For QUEST experiments (quantification of the exchange rate as
a function of saturation time), data were collected by varying the satura-
tion time (0.25, 0.5, 1, 1.5, 2, 2.5, 3, 4, 5 s) at constant power (25 mT). The
QUESP data (quantification of the exchange rate as a function of satura-
tion power) were collected by varying the saturation power whilst the
saturation time remained constant (3 s). The QUEST and QUESP data
were fitted with Scientist (MicroMath, Inc.).
Conclusions
We have designed and synthesized DOTA derivatives of a-
aminoglycine and their corresponding lanthanide complexes
as building blocks for self-immolative imaging probes. A
synthetic pathway giving access to the platform with or with-
out the self-immolative linker and without the difficulties as-
sociated with the intrinsic instability of these ligands was
elaborated. The physicochemical properties of several Gd3+
complexes were investigated: a chelate bearing a self-immo-
lative arm and a sugar unit as a selective substrate for b-gal-
actosidase, its enzymatic product, and two model com-
pounds. They all have one inner-sphere water molecule and
consequently only slightly different proton relaxivities. This
similarity precludes the application of the GdL1 complex as
enzyme-specific T1 relaxation agents. The water exchange
for all four systems is slightly slower than for GdDOTA.
NMRD measurements: The 1/T1 nuclear magnetic relaxation dispersion
(NMRD) profiles of the GdIII complexes were recorded on a Stelar
SMARtracer FFC fast-field-cycling relaxometer covering magnetic fields
from 2.35ꢃ10ꢀ4 T to 0.25 T, which corresponded to a proton Larmor fre-
quency range of 0.01–10 MHz. The relaxivity at higher fields was record-
ed using a Bruker WP80 adapted to variable field measurements and
controlled by the SMARtracer PC NMR console. The temperature was
controlled by a VTC90 temperature-control unit and fixed by a gas flow.
The temperature was determined according to a previous calibration
with a platinum resistance temperature probe. The relaxivity at 500 MHz
was measured on a Bruker Avance 500 (11.75 T) spectrometer. The si-
multaneous-least-squares fit of the 17O NMR and 1H NMRD data were
performed by using Micromath Scientist version 2.0 (Salt Lake City, UT,
USA). The reported errors correspond to one standard deviation ob-
tained by statistical analysis.
The complexes containing exocyclic NH2 or NHACTHNURTGNENG(U CH3)
UV/Vis spectrophotometry: Absorbance spectra were recorded on a
Perkin–Elmer Lambda 19 spectrometer in thermostated cells between 25
and 508C for EuL1 (cEu ꢃ2.5 mm, pHꢃ6.80) and EuL2 (cEu ꢃ2.4 mm, pH
ꢃ6.90). The measurements were carried out in a cylindrical cuvette with
10 cm optical path-length between l=577–581.0 nm.
amine groups are not protonated at physiological pH, as evi-
denced by pH-potentiometric measurements. Their Yb3+ an-
alogues show a pH-dependent PARACEST effect. The
proton-exchange rate was determined for both complexes at
1416
ꢄ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 1408 – 1418