Phosphinic derivative of DTPA conjugated to a G5 PAMAM dendrimer: An 17O and 1H relaxation study of its Gd(III) complex

Article abstract of DOI:10.1039/b517847a

A DTPA-based chelate containing one phosphinate group was conjugated to a generation 5 polyamidoamine (PAMAM) dendrimer via a benzylthiourea linkage. The Gd(iii) complex of this novel conjugate has potential as a contrast agent for magnetic resonance imaging (MRI). The chelates bind Gd³⁺ via three nitrogen atoms, four carboxylates and one phosphinate oxygen, and one water molecule completes the inner coordination sphere. The monomer Gd(iii) chelates bearing nitrobenzyl and aminobenzyl groups ([Gd(DTTAP-bz-NO₂)(H ₂O)]^2- and [Gd(DTTAP-bz-NH₂)(H ₂O)]^2-) as well as the dendrimeric Gd(iii) complex G5-(Gd(DTTAP))₆₃) were studied by multiple-field, variable temperature ¹⁷O and ¹H NMR. The rate of water exchange is faster than that of [Gd(DTPA)(H₂O)]^2- and very similar on the two monomeric complexes (8.9 and 8.3 × 10⁶ s^-1 for [Gd(DTTAP-bz-NO₂)(H₂O)]^2- and [Gd(DTTAP-bz-NH₂)(H₂O)]^2-, respectively), while it is decreased on the dendrimeric conjugate (5.0 × 10⁶ s ^-1). The Gd(iii) complex of the dendrimer conjugate has a relaxivity of 26.8 mM^-1 s^-1 at 37 °C and 0.47 T (corresponding to ¹H Larmor frequency of 20 MHz). Given the contribution of the second sphere water molecules to the overall relaxivity, this value is slightly higher than those reported for similar size dendrimers. The experimental ¹⁷O and ¹H NMR data were fitted to the Solomon-Bloembergen-Morgan equations extended with a contribution from second coordination sphere water molecules. The rotational dynamics of the dendrimeric conjugate was described in terms of global and local motions with the Lipari-Szabo approach. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b517847a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Phosphinic derivative of DTPA conjugated to a G5 PAMAM dendrimer:
an 17O and 1H relaxation study of its Gd(III) complex†
b
b
b,c
Petra Lebdusˇkova´,a,b Ange´lique Sour, Lothar Helm,* Eva To´th, Jan Kotek,a Ivan Lukesˇ*a and
´
Andre´ E. Merbachb
Received 20th December 2005, Accepted 4th April 2006
First published as an Advance Article on the web 19th April 2006
DOI: 10.1039/b517847a
A DTPA-based chelate containing one phosphinate group was conjugated to a generation 5
polyamidoamine (PAMAM) dendrimer via a benzylthiourea linkage. The Gd(III) complex of this novel
conjugate has potential as a contrast agent for magnetic resonance imaging (MRI). The chelates bind
Gd3+ via three nitrogen atoms, four carboxylates and one phosphinate oxygen, and one water molecule
completes the inner coordination sphere. The monomer Gd(III) chelates bearing nitrobenzyl and
aminobenzyl groups ([Gd(DTTAP-bz-NO2)(H2O)]2− and [Gd(DTTAP-bz-NH2)(H2O)]2−) as well as the
dendrimeric Gd(III) complex G5-(Gd(DTTAP))63) were studied by multiple-field, variable temperature
17O and 1H NMR. The rate of water exchange is faster than that of [Gd(DTPA)(H2O)]2− and very
similar on the two monomeric complexes (8.9 and 8.3 × 106 s−1 for [Gd(DTTAP-bz-NO2)(H2O)]2− and
[Gd(DTTAP-bz-NH2)(H2O)]2−, respectively), while it is decreased on the dendrimeric conjugate (5.0 ×
106 s−1). The Gd(III) complex of the dendrimer conjugate has a relaxivity of 26.8 mM−1 s−1 at 37 ◦C and
0.47 T (corresponding to 1H Larmor frequency of 20 MHz). Given the contribution of the second
sphere water molecules to the overall relaxivity, this value is slightly higher than those reported for
similar size dendrimers. The experimental 17O and 1H NMR data were fitted to the
Solomon–Bloembergen–Morgan equations extended with a contribution from second coordination
sphere water molecules. The rotational dynamics of the dendrimeric conjugate was described in terms
of global and local motions with the Lipari–Szabo approach.
fine-tuning these parameters, one can obtain a contrast agent of
Introduction
very high efficiency. Since little is known of the relation between
molecular structure and electronic spin relaxation, the design of
new contrast agents is mainly focused on the optimisation of sM
and sR.
Paramagnetic metal complexes are widely used as contrast agents
for magnetic resonance imaging (MRI). They increase the re-
laxation rate of the water proton spins in their surroundings,
which enhances the image contrast. Gd(III) is often applied as
the paramagnetic metal ion for its electronic properties (the seven
unpaired electrons make it highly paramagnetic, and its electronic
relaxation is relatively slow due to the ground state S symmetry).1
The term relaxivity (r1) defines the enhancement of the relax-
ation rate of the water protons per mmol of Gd3+ and it is directly
proportional to the efficiency of the contrast agent. The relaxivity
depends on several parameters, such as the residence time of the
inner sphere water molecule (sM), the reorientation time of the
complex (sR) and the electronic relaxation times (T1e and T2e). By
The exchange of the first coordination sphere water is generally
too slow on currently used contrast agents.1 It could be accelerated
by introducing a bulky phosphinic or phosphonic group instead of
a carboxylate:2–4 the increased steric crowding and negative charge
brought by the phosphorus-containing moiety forces the bound
water molecule to leave faster. The phosphonic and phosphinic
functions can also induce hydrogen bond formation, which will
prolong the residence time of the outer sphere molecules in the
proximity of the paramagnetic centre and therefore influence
their relaxation times.5,6 Those water molecules constitute the so-
called second hydration sphere.7 The rotation of the paramagnetic
complex can be slowed down by increasing the molecular volume
and maintaining a rigid internal structure. This can be achieved
by self-assembly of small chelates (like micelle formation), or by
covalent or non-covalent conjugation to macromolecules, such as
peptides, polysaccharides or dendrimers.1,8–12
PAMAM (polyamidoamine) dendrimers are water soluble,
spheroidal particles, composed of repeated polyamidoamine units
and primary amino groups on the surface, which can be deriva-
tised. The generation 5 dendrimer, used in this work, contains 128
primary amino groups.13,14 An example of a previously studied
phosphinic chelate conjugated to PAMAM dendrimer is the
G2-(Gd(DO3A-PABn))16 complex3 which shows high relaxivity
aDepartment of Inorganic Chemistry, Charles University, Hlavova 2030,
12840 Prague, Czech Republic. E-mail: lukes@natur.cuni.cz; Fax: +420 2
2195 1253; Tel: +420 2 2195 1259
bInstitut des Sciences et Inge´nierie Chimiques, Ecole Polytechnique Fe´de´rale
de Lausanne, CH-1015 Lausanne, Switzerland. E-mail: lothar.helm@
epfl.ch; Fax: +41 21 693 98 75; Tel: +41 21 693 98 76
cCentre de Biophysique Mole´culaire, CNRS, rue Charles-Sadron, 45071
Orle´ans, Cedex 2, France
† Electronic supplementary information (ESI) available: Equations used
for the analyses of 17O NMR and 1H NMRD data; tables of the variable
temperature 17O NMR and 1H NMRD relaxation data and figures of 17
O
NMR chemical shift data; calculation of the number of chelating groups
per molecule of dendrimer and a full list of parameters for the fitting of
[Gd(DTTAP-bz-NO2)(H2O)]2−, [Gd(DTTAP-bz-NH2)(H2O)]2− and G5-
(Gd(DTTAP))63. See DOI: 10.1039/b517847a
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3399–3406 | 3399
©

View Article Online
(14.1 mM−1 s−1, 37 ◦C, 20 MHz) due to fast water exchange
and a second sphere contribution. A phosphinate containing
DTPA derivative, DTTAP-bz-NH2, was conjugated to inuline
and a relaxivity of 18.3 mM−1 s−1 was reported for its Gd(III)
complex (37 ◦C, 0.47 T).15 This relaxivity was lower than expected
according to the molecular mass of the conjugate, probably due
to the internal flexibility of the macromolecule.
In this paper, we report the conjugation of the same monomeric
chelate DTTAP-bz-NH2 to a generation 5 PAMAM dendrimer
(Schemes 1 and 2), which has a greater molecular weight and
should also be more rigid than inulin, and therefore a more
significant increase in relaxivity is expected for the Gd(III) com-
plex. The monomeric chelates [Gd(DTTAP-bz-NO2)(H2O)]2− and
[Gd(DTTAP-bz-NH2)(H2O)]2− as well as the dendrimeric conju-
2−
gate [PAMAM-G5(N{CS}N-bz-Gd{DTTAP}{H2O} )63] (later
in the text referred to as G5-(Gd(DTTAP))63) were evaluated using
variable temperature multiple-field NMR spectroscopy and the
experimental data were fitted to the commonly used equations
based on the Solomon–Bloembergen–Morgan approximation,1
extended by a second sphere contribution to the overall relax-
1
ivity. The experimental H and 17O longitudinal relaxation rates
obtained for the dendrimer were evaluated using the Lipari–
Szabo approach yielding local and global rotational correla-
tion times.9,16,17 All the ligands discussed are depicted in Scheme 3.
Experimental
Materials and methods
All reagents and solvents were commercially available. The
ethylenediamine core PAMAM G5 dendrimer with primary
amines on the surface was purchased as a 5.26 w/w% aqueous
Scheme 1 Structure of the dendritic ligand.
Scheme 2 Synthesis of the dendrimeric chelate; (i) H2/Pd/C, (ii) CSCl2, CCl4, HCl, (iii) PAMAM G5, 1 M LiOH.
Scheme 3 Structures of monomeric chelates discussed in the text.
3400 | Dalton Trans., 2006, 3399–3406 This journal is
The Royal Society of Chemistry 2006
©

View Article Online
solution from Dendritech Inc. (Midland, MI). MILIPORE ul-
trafiltration membranes YM-10 were obtained from Amicon
mixture was vigorously stirred at RT for 6 h. The completion of the
reaction was checked using TLC (iPrOH : NH3 : H2O 7 : 3 : 3, Rf =
0.54, the spot remained uncoloured under ninhydrin detection).
All the solvents were evaporated and the solid product was dried
overnight under vacuum at 45 ◦C to yield DTTAP-bz-NCS which
was used without further purification for the next reaction.
NMR characterization was in agreement with the previously
published data of this compound.15
R
(Centriprepꢀ, regenerated cellulose 10 000 MWCO, Bedford,
MA).
The luminescence measurements were performed on an Aminco
ABS2 spectrofluorimeter. Luminescence decay were obtained at
615 nm after excitation at 395 nm.
1H NMRD profiles were performed on a Stelar Spinmaster FFC
fast field cycling relaxometer covering magnetic fields from 2.35 ×
10−4 T to 0.47 T (proton Larmor frequency range 0.01–20 MHz).
The temperature was fixed by a gas flow and controlled by a
VTC90 temperature control unit. At higher fields, relaxivities were
measured on Bruker Minispecs mq30 (30 MHz), mq40 (40 MHz)
and mq60 (60 MHz) and on Bruker 100 MHz (2.35 T) and
200 MHz (4.7 T) cryomagnets connected to a Bruker Avance-200
console and on a Bruker DRX 400 (400 MHz, 9.4 T). In each case,
the temperature was measured by the substitution technique.18
Longitudinal and transverse 17O relaxation rates were measured
at temperatures from 279 K to 359 K. The data were recorded on
Bruker ARX 400 (9.4 T, 54.2 MHz) and Bruker Avance-200 (4.7 T,
27.1 MHz) spectrometers. VT 3000 temperature control units were
used to maintain a constant temperature, which was measured
by a substitution technique.18 The samples were sealed in glass
spheres, fitting into 10 mm NMR tubes. Longitudinal relaxation
rates, 1/T1, were obtained by the inversion recovery method
while transverse relaxation rates, 1/T2, were measured by the
Carl–Purcell–Meiboom–Gill spin-echo technique. As external
references, acidified water (pH 3.3) was used for the monomeric
chelates and the Y(III) complex (of a concentration and pH
identical to those of the Gd(III) complex) for the dendrimer.
Elemental analysis was performed by Solvias AG, Basel,
Switzerland. Analysis of the 17O NMR and 1H NMRD experimen-
tal data was performed with the Visualiseur/Optimiseur programs
on a Matlab platform version 6.5.19
MS (ESI), m/z (%): 561.3 [M + H+], calcd. for C21H29N4O10SP:
M = 560.515
PAMAM G5-(DTTAP)63
DTTAP-bz-NCS obtained above was dissolved in H2O (8.5 ml)
with the addition of LiOH (3.8 ml, 1 M) until everything dissolved.
The resulting pH was 7. The aqueous solution of PAMAM G5
dendrimer (2.347 g, 5.26 w/w% in water, 0.5482 mmol of NH2
groups) was added dropwise to the reaction mixture. The pH was
increased to a value of 8–9 and the reaction mixture was stirred at
30 ◦C for 5 days. The pH was maintained constant by addition of
LiOH (1 M).
The reaction mixture was then filtered through a filter paper and
then through a 22 lm disposable filter. Free chelate was removed by
R
ultrafiltration (Centriprepꢀ YM-10) using centrifuge (3800 rpm).
The resulting yellow solution was evaporated and vacuum dried
overnight at 45 ◦C to yield 342 mg of a yellow solid.
Assessment of Gd(III) loading on the dendrimer
The number of chelates linked to the dendrimer was assessed
by 1H NMR, by complexometric titration using GdCl3 and
20
by elemental analysis. The integration of the 1H NMR spectrum
(aromatic vs. aliphatic regions) gave an estimation of 48% loading
of the total number of primary amino groups.
The complexometric titration allows the determination of the
number of chelating units per mass of the product. A solution
of a known amount of the dendrimeric conjugate was titrated by
a GdCl3 solution (buffered with urotropin/HCl, pH 5.8, xylenol
orange indicator). The complexometric titration in combination
with the elemental analysis allowed calculating the average loading
of the dendrimer which was found to be 63 molecules of chelate per
128 amino groups on one molecule of dendrimer (49%). Details
are given in the ESI†.
Diethylenetriamine-Nꢀ-methylene(p-aminobenzyl)-phosphinic-
N,N,Nꢀꢀ,Nꢀꢀ-tetraacetic acid (DTTAP-bz-NH2)
The synthesis of DTTAP-bz-NO2 and DTTAP-bz-NH2 was per-
formed according to the literature.15 The purification of DTTAP-
bz-NH2 was improved from the original procedure, performing a
chromatography on a SiO2 column (iPrOH : NH3 : H2O 10 : 3 :
3). Rf (DTTAP-bz-NH2) = 0.33. The cations were removed on a
column of strong anion exchanger (AG 1-X4, 100–200 mesh, Bio-
Rad, AcO− form, 130 ml). After washing the column with water,
the product was eluted using HCl solution (5%). The solvents were
evaporated under reduced pressure into the resulting yellow oil.
All the amount obtained was used for the next reaction.
MS (ESI) for DTTAP-bz-NO2, m/z (%): 549.4 [M + H+], calcd.
for C20H29N4O12P: M = 548.44
Elemental analysis calcd. (%) for C1262H2528N506O252·63
C21H29N4O10PS (64137.32): C 48.4, H 6.84, O 22.0, N 16.6, P
3.04, S 3.15; found: C 48.8, H 7.05, O 21.2, N 17.3, P 2.71, S 2.81.
Sample preparation
Gadolinium(III) and yttrium(III) complexes for 17O NMR and 1H
NMRD measurements were prepared by mixing a slight excess of
the ligands (5%) with metal chloride solutions. GdCl3 and YCl3
solutions were prepared by dissolving metal oxides in slight excess
of HCl (Merck, p.a., 25%) followed by evaporation of solvents
and dissolution in double distilled water. The concentration of the
stock solutions was determined by titration with Na2H2EDTA
solution (indicator xylenol orange). The pH of the complex
solutions was adjusted using NaOH (1 M) and HCl (1 M). The
absence of the free metal ion was checked with xylenol orange
MS (ESI) for DTTAP-bz-NH2, m/z (%): 519.3 [M + H+], calcd.
for C20H31N4O10P: M = 518.45
Diethylenetriamine-Nꢀ-methylene(p-isothiocyanatobenzyl)-
phosphinic-N,N,Nꢀꢀ,Nꢀꢀ-tetraacetic acid (DTTAP-bz-NCS)
425 mg (0.64 mmol) of DTTAP-bz-NH2·4HCl was dissolved in
HCl (7.3 ml, 3M). Carbon tetrachloride (4.2 ml, 44 mmol) and
thiophosgene (2.0 ml, 26 mmol) were added and the reaction
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3399–3406 | 3401
©

View Article Online
indicator at pH 5.8. The samples used for 17O NMR were enriched
to 1% with 17O-enriched water (Isotrade GmbH).
Results and discussion
Synthesis of the dendrimeric conjugate
The conjugation of the chelate to the PAMAM G5 dendrimer
was performed via thiourea linkage according to Scheme 2. The
NCS-chelate was used in 20% molar excess and the loading ran
to 49% of the total possible substitution places. Even on addition
of a novel portion of 200% excess of the DTTAP-bz-NCS, no
higher loading of the dendrimer was achieved. For carboxylate
derivatives, higher loadings have usually been reported.20 One
possible explanation could be that the phosphinate groups form
stronger hydrogen bonds with the remaining unsubstituted amines
than the carboxylates and this disables further conjugation.
Nevertheless, the non-complete substitution of the dendrimer can
be an advantage as it leaves space for further modification and
functionalization of the primary amines. For instance, one can
attach any targeting group to deliver the agent to a desired place
in the body, or an antenna that can act as a reporter in another
imaging modality (dyes for near infrared imaging, radioactive
tracers for SPECT, etc.) giving rise to ‘multimodal’ contrast agents.
17O NMR and 1H relaxivity measurements on the monomers [Gd-
(DTTAP-bz-NO2)(H2O)]2− and [Gd(DTTAP-bz-NH2)(H2O)]2−
Variable temperature 17O NMR measurements were performed
at 9.4 T (54.2 MHz) on aqueous solutions of the chelates and
were referenced to an acidified water solution (aqueous HClO4,
pH 3.3). The data obtained for [Gd(DTTAP-bz-NO2)(H2O)]2−
and [Gd(DTTAP-bz-NH2)(H2O)]2− are very similar (Fig. 1A,
1B, 2A, 2B). The temperature dependence of the transverse
17O relaxation rates indicates relatively fast water exchange (the
maximum of the ln(1/T2r) versus 1/T curve (Fig. 1A, 1B) is shifted
towards low temperatures with respect to analogous data21 for
[Gd(DTPA)(H2O)]2−).
By analogy to similar Gd(III) complexes,2 coordination of one
inner-sphere water molecule was expected. It was confirmed by
the luminescence decay measurement method. The value of lumi-
nescence lifetime measured for the [Eu(DTTAP-bz-NH2)(H2O)]2−
complex (s = 582 ls) corresponds with the hydration number q =
1 according to well-proved relationships.22,23
Reduced 17O chemical shifts were also measured for these
two gadolinium complexes (data available in the ESI†), however,
they were not included in the calculations since the values are
systematically smaller than what is expected for a chelate with one
coordinated water molecule. A possible explanation could be the
influence of the second sphere water molecules. As it has been
shown by quantum chemical calculations, the sign of spin density
on Gd(III) complexes can depend on spatial arrangement and
therefore second sphere 17O chemical shifts could have a different
sign compared to the first sphere shift.24
Fig. 1 Reduced longitudinal (ꢀ) and transverse (᭹) relaxation rates of:
(A) [Gd(DTTAP-bz-NO2)(H2O)]2−; (B) [Gd(DTTAP-bz-NH2)(H2O)]2−
;
(C) G5-(Gd(DTTAP))63 (4.7 T –ꢁ, ᭺; 9.4 T –ꢀ, ᭹); the lines are
calculated using Solomon–Bloembergen–Morgan theory (A, B) and the
Lipari–Szabo approach (C).
(r1 decreases with increasing temperature) follows the general
trend for small molecular weight complexes (Fig. 2A, 2B). The
r1 values of [Gd(DTTAP-bz-NO2)(H2O)]2− and [Gd(DTTAP-bz-
NH2)(H2O)]2− are slightly higher than those of currently used
contrast agents (5.24 and 5.02 mM−1 s−1 for [Gd(DTTAP-bz-
NO2)(H2O)]2− and [Gd(DTTAP-bz-NH2)(H2O)]2−, respectively,
vs. 4.02 mM−1 s−1 for [Gd(DTPA)(H2O)]2−) but are very similar
to the phosphorus containing analogue [Gd(DTTAP-ph)(H2O)]2−
Proton relaxivities were measured as a function of the
1H Larmor frequency at four temperatures (5, 25, 37 and
50 ◦C) and at a pH close to physiological (pH ∼ 6.5). The
shape of the NMRD curves and their temperature dependence
3402 | Dalton Trans., 2006, 3399–3406
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
To obtain the best least-square fits, a second sphere water coordi-
nation shell had to be taken into account in the set of equations.2
This term was necessary not only to describe proton relaxation,
but also longitudinal oxygen-17 relaxation. Longitudinal 17O
relaxation is influenced by dipolar and quadrupolar contributions,
both related to sR. The second sphere water feels the rotation of
the chelate, consequently the T1 values measured on bulk water
will be affected by the presence of second sphere water molecules.
On the other hand, 17O T2 relaxation is mainly driven by the scalar
contribution thus it is not sensitive to sR and consequently to the
presence of a second hydration shell.
The total number of parameters used for the simultaneous
fitting was 23; several of them are common to both sets of 17O
1
and H experimental data. As discussed above and similarly to
the previously reported inulin-(Gd(DTTAP))23,15 the number of
inner sphere water molecules was fixed to one. The number of
second sphere water molecules was fixed to 2, based on analogous
phosphinates studied by Kotek et al.2 The distances between
oxygen/proton of the coordinated water and gadolinium were
1
˚
˚
fixed to usual values: 2.5 A and 3.1 A, respectively. The distance
between the outer sphere water proton and the gadolinium was
˚
fixed to 3.5 A. To describe the contribution of the second sphere
water, four additional parameters were included: the average
298
2nd
residence time of the second sphere water (sM ) was fixed to 50 ps,
based on the simulations of Borel et al.25 Its activation enthalpy
(DH‡2nd) was set to 35 kJ mol−1, as it was found for similar molecules
by Kotek et al.,2 but the quality of the fit was not significantly
influenced upon changing this parameter between 15–50 kJ mol−1.
The distances of the second sphere oxygen and proton from the
˚
˚
gadolinium ion were fixed to 4.1 A and 3.5 A, respectively. For the
quadrupolar coupling constant v(1 + g2/3)1/2 we used the value
for pure water (7.58 MHz).21 The value of the hyperfine coupling
constant (A/ꢀ), was fixed to −3.8 × 106 rad s−1, as previously
obtained for the [Gd(DTPA)(H2O)]2− complex.21 The activation
energy of sV, Ev, had to be fixed to a value of 1 kJ mol−1; when left
as a variable, it iterated to negative values.
The rotational dynamics of the complexes has been evaluated
by using different rotational correlation times for the motion of
the Gd–water oxygen and Gd–water proton vectors (sRH and sRO ),
as observed by 17O NMR and 1H NMRD.26 Based on geometrical
considerations, the ratio of those two rotational correlation times
298
RH
298
RO
(s /s ) has to lie between 0.65 and 1. The best least-square fits
were obtained using parameters listed in Table 1. It is evident that
the two monomeric compounds [Gd(DTTAP-bz-NO2)(H2O)]2−
and [Gd(DTTAP-bz-NH2)(H2O)]2− are very similar and that the
reduction of the nitro to an amino group does not significantly
affect the relaxation properties.
In comparison to [Gd(DTPA)(H2O)]2−, the water exchange
rates of [Gd(DTTAP-bz-NO2)(H2O)]2− and [Gd(DTTAP-bz-
NH2)(H2O)]2− are significantly faster, which can be explained by
the presence of the more bulky phosphinate. The space around the
Gd(III) is more sterically crowded which forces the coordinated
water molecule to leave. However, the faster water exchange has
not much consequence on the r1 value of the monomers as ke2x98
does not limit relaxivity for these small complexes. The relaxivities
are higher than for current contrast agents due to the second
sphere effect. Nevertheless, faster water exchange will be retained
after conjugation of the monomer to a macromolecule, where the
rotation is slowed down and the exchange rate becomes more
Fig. 2 1H NMRD profiles: (A) [Gd(DTTAP-bz-NO2)(H2O)]2−; (B)
[Gd(DTTAP-bz-NH2)(H2O)]2−; (C) G5-(Gd(DTTAP))63 (ꢀ5 ◦C, ᭺15 ◦C,
ꢀ 25 ◦C, ᭜ 37 ◦C and ᭺ 50 ◦C; the lines are calculated using
Solomon–Bloembergen–Morgan theory).
complex2 (Scheme 3, Table 1). This originates from the contribu-
tion of second sphere water molecules, present in the phosphinate
containing complexes.
The relaxation data obtained by the variable temperature 17O
1
NMR and H NMRD measurements were analysed simultane-
ously (Fig. 1A with 2A, Fig. 1B with 2B for [Gd(DTTAP-bz-
NO2)(H2O)]2− and [Gd(DTTAP-bz-NH2) (H2O)]2−, respectively).
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3399–3406 | 3403
©

View Article Online
Table 1 Parameters resulting from simultaneous fits of 1H NMRD and 17O NMR data; r1 values correspond to 37 ◦C and 0.47 T
a
b
Parameter
[Gd(DTTAP-bz-NH2)(H2O)]2− [Gd(DTTAP-bz-NO2)(H2O)]2−
[Gd(DTTAP-ph)(H2O)]2−
[Gd(DTPA)(H2O)]2−
r1/mM−1 s−1
5.02
8.9 1.3
48.1
+49.8 10
116
20.5 0.8
28.2 0.5
0.403 0.003
25
30
5.24
8.3 0.6
48.5
5.36
11
37.0
—
296
19
26
0.32
22.75
—
8
0.37
4.02
3.3
51.6
298
kex [106]/s−1
DH‡/kJ mol−1
DS‡/J mol−1 K−1
sRO 298/ps
3
2
+50.3
8
+53.0
2
117 10
58
ER/kJ mol−1
sV298/ps
18
1
17.3
25
0.46
20
19.4
1.2
26.3 0.4
0.40 0.01
25
30
2.8 0.5
D2 [1020]/s−2
D298
[10−10]/m2s−1
2
1
2
1
GdH
E
DGdH/kJ mol−1
2
dgL [10−2
]
3.3 0.6
0.81 0.04
sRH298/sRO
0.85 0.08
—
298
a Ref. 2. b Ref. 21.
critical. The other parameters obtained in the fit have values
comparable to analogous small paramagnetic Gd(III) complexes.
Fig. 3 displays the experimental and fitted 1H NMRD profiles
reduced longitudinal and transverse relaxation rates as a function
of temperature.
Proton relaxivities of the dendrimer G5-(Gd(DTTAP))63 were
◦
1
at 37 C, as well as the calculated contribution from the second
measured as a function of the H Larmor frequency at tempera-
and outer sphere water. It is shown that the major contribution
to the overall relaxivity comes from the inner sphere water and
a very significant contribution is represented by the outer sphere.
The second sphere contribution makes up to about 20% of the
total relaxivity.
tures 5, 15, 25, 37 and 50 ◦C and pH 6.25. The shape of the NMRD
curves of the dendrimer complex shows a hump at higher frequen-
cies, typical for macromolecules, with a maximum at ∼30 MHz.
In comparison to the monomers, a great relaxivity increase is
observed (at 0.47 T and 37 ◦C the relaxivity is 26.8 mM−1s−1 for G5-
(Gd(DTTAP))63 vs. 5.02 and 5.24 mM−1s−1 for [Gd(DTTAP-bz-
NH2)(H2O)]2− and [Gd(DTTAP-bz-NO2)(H2O)]2−, respectively).
This relaxivity hump results from the slow rotation of the
dendrimer. The relaxivity measured at the maximum of the hump
(30 MHz) increases with temperature up to ∼25 ◦C and decreases
then if the temperature is increased further (Fig. 4). This implies
that at lower temperatures the water exchange rate is a limiting
factor for r1, while at higher temperatures the relaxivity is mainly
limited by rotation which is too fast. This trend is in contrast
to that reported previously for dendrimeric conjugates. For the
fast exchanging EPTPA derivatives, the relaxivities increase with
decreasing temperature in the whole temperature range, showing
that rotation limits r1,20 while for other dendrimeric complexes
based on DOTA or DTPA an opposite trend was reported, as a
result of slow water exchange.11
Fig. 3 1H NMRD profile of [Gd(DTTAP-bz-NH2)(H2O)]2− recorded at
37 ◦C (᭜) with the fitted curve (full line) and calculated contributions form
outer (dashed line) and second (dotted line) spheres.
17O NMR and 1H relaxivity measurements on the dendrimer
complex G5-(Gd(DTTAP))63
The 17O NMR measurements were referenced to an Y(III) complex
solution of the same pH and concentration as the Gd(III) analogue.
The reason for using the diamagnetic Y(III) complex was a
difference (at low temperatures) between T1 values of the acidified
water reference and the Y(III) complex. Similarly to previous
observations on dendrimeric systems20 T1 values of the Y(III)
complex were up to 30% lower than those of the acidified water.
The measurements were performed at two magnetic fields, 9.4 T
(54.2 MHz) and 4.7 T (27.1 MHz), and it was shown that 17O
T1 strongly depends on the magnetic field. Fig. 1C shows the
Fig. 4 Temperature dependence of the relaxivity of G5-(Gd(DTTAP))63.
The experimental data correspond to 30 MHz (ꢀ) and 1 MHz (᭺) of 1H
Larmor frequency.
3404 | Dalton Trans., 2006, 3399–3406
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
The longitudinal 17O relaxation rates were analysed by using the
Lipari–Szabo approach16,17 that distinguishes between rotational
correlation times for the fast local and slow global motion of the
complex. The local rotational correlation time is related to the
motion of the Gd–water oxygen vector while the global rotational
correlation time describes the slow overall motion of the entire
macromolecule that the model considers spherical.1,9 The degree
of spatial restriction of the local motion with regard to the global
rotation is given by the model free generalized order parameter
S2. For a completely free internal motion S2 = 0, while for an
internal motion completely correlated to the global one S2 = 1.
The second sphere contribution has been also included in the
model, as described above for the monomers. The equations used
for the calculations are listed in the ESI†.
chelates. The motion of the Gd(III) segments on the dendrimeric
surface is less limited by spatial restrictions than in the similar
PAMAM G5 chelate G5-(Gd(EPTPA))111 or in Gadomer 1712
20
and does not fully take advantage of the slow global motion of the
molecule.
The water exchange rate slightly decreased upon conjugation
to the dendrimer (5.0 × 106 s−1, Fig. 1C). A similar trend was
previously observed on conjugation to inulin15 and on a smaller
dendrimeric complex G2-(Gd(DO3A-PABn))16.3 This can be due
to the possible electrostatic interactions between primary amines
of dendrimer and negatively charged chelate, which changes the
electron density of the chelate and therefore also the interaction
with the inner sphere water. All parameters obtained can be found
in Table 2 or in the ESI†.
All Gd–O and Gd–H distances were fixed at the same values
as for the monomers as well as the hyperfine and quadrupolar
coupling constants. Finally, we were not able to fit simultaneously
1H NMRD and 17O NMR data. The observed temperature
dependence of the relaxivities in the high-field ‘hump’ could only
be reproduced by a dramatically lower DH‡ as compared to the
value calculated from 17O NMR. Therefore only the 17O NMR
data were fitted. The transverse 17O relaxation rates allow an
accurate determination of the water exchange rate, while the
longitudinal 17O relaxation rates give access to the rotational
parameters.
Comparison with other macromolecular Gd(III) chelates
The overall relaxivity of the dendrimer G5-(Gd(DTTAP))63 at
◦
37 C and 0.47 T is 26.8 mM−1 s−1, higher than that of the
inulin-conjugate.15 This is directly proportional to the increase
in the molecular weight (see Table 3). As a consequence of
the second sphere water, the relaxivity of G5-(Gd(DTTAP))63 is
even higher than that of the similar dendrimeric conjugate20 G5-
(Gd(EPTPA))111. Table 3 compares the principal parameters of
several monomeric and macromolecular complexes, all of them
derived from DTPA21 or DOTA (Gadomer 17,12 G2-(Gd(DO3A-
PABn))163). The rigidity is described by the parameter S2 and one can
see that the internal flexibility increases from Gadomer 17 towards
PAMAM G5 dendrimers. The very low local correlation time
compared to Gadomer 17 is due to less restricted local motion.20
1H NMRD profiles at 37 ◦C of various Gd(III) complexes are
compared in Fig. 5.
The parameters obtained are shown and compared to an
20
analogous dendrimer, G5-(Gd(EPTPA))111 in Table 2. The local
correlation time (sl) is much shorter than the global one (sg),
71 ps vs. 4417 ps, and the general order parameter S2 is relatively
small (0.28) which indicates an important flexibility of the Gd(III)
Table 2 Selected parameters obtained from the 17O NMR fit of G5-
(Gd(DTTAP))63] compared to an analogous dendrimer;20 r1 values cor-
respond to 37 ◦C and 0.47 T
Conclusion
Parameter
G5-(Gd(DTTAP))63
G5-(Gd(EPTPA))111
A DTPA-based ligand bearing one phosphinic acid was con-
jugated to a generation 5 polyamidoamine (PAMAM) den-
drimer via a benzylthiourea linkage. Gd(III) complexes of the
monomers [Gd(DTTAP-bz-NO2)(H2O)]2− and [Gd(DTTAP-bz-
NH2)(H2O)]2− and the dendrimeric G5-(Gd(DTTAP))63 were
synthesised and characterised by 1H NMRD and 17O NMR
to evaluate parameters relevant to potential MRI applications.
The experimental data were analysed by using the Solomon–
Bloembergen–Morgan set of equations, extended by the Lipari–
Szabo approach for the dendrimer. A contribution from the second
r1310/mM−1s−1
26.8
5.0 0.2
23.9
150 30
298
kex [106]/s−1
DH‡/kJ mol−1
DS‡/J mol−1K−1
sgO298/ps
56.2
1
20.0
4
+71.9
3
−21 12
4417 340
4040 300
E
gO/kJ mol−1
21
71
21
2
3
2
25
150 15
31
2
slO298/ps
ElO/kJ mol−1
3
S2
0.28 0.01
1.7 0.1
0.43 0.03
—
dgL [10−2
]
2
Table 3 Comparison of principal parameters for various Gd(III) complexes; r1 values correspond to 37 ◦C and 0.47 T
Complex
sM298/ns
kex [106]/s−1
298
sgO298/ps
slO298/ps
S2
r1/mM−1s−1
Mw/kDa
Ref.
[Gd(DTPA)(H2O)]2−
[Gd(DTTAP-bz-NH2)(H2O)]2−
inulin-(Gd(DTTAP))23
G2-(Gd(DO3A-PABn))16
G5-(Gd(DTTAP))63
G5-(Gd(EPTPA))111
Gadomer 17
303
113
170
30
199
6.7
100
3.3
8.9
5.9
33
5.0
150
1
—
—
—
—
4417
4040
3050
58
116
866
—
71
150
760
—
—
—
—
0.28
0.43
0.50
3.7
5.02
18.3
14.1a
26.8
23.9
17.18
0.57
0.69
23
11.8
64
110
17.5
21
This work
15
3
This work
20
12
a J. Rudovsky´ et al., unpublished results.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3399–3406 | 3405
©

View Article Online
COST Action D18 and the European-founded EMIL programme
(LSHC-2004-503569).
References
´
1 E. To´th, L. Helm and A. E. Merbach, Relaxivity of Gadolinium(III)
Complexes: Theory and Mechanism in The Chemistry of Contrast
Agents in Medical Magnetic Resonance Imaging, ed. A. E. Merbach
´
and E. To´th, Wiley, Chichester, UK, 2001, pp. 45–120.
2 J. Kotek, P. Lebdusˇkova´, P. Hermann, L. Vander Elst, R. N. Muller,
C. F. G. C. Geraldes, T. Maschmeyer, I. Lukesˇ and J. A. Peters, Chem.–
Eur. J., 2003, 9, 5899.
3 J. Rudovsky´, P. Hermann, M. Botta and I. Lukesˇ, Chem. Commun.,
2005, 2390.
4 J. Rudovsky´, J. Kotek, P. Hermann, I. Lukesˇ, V. Mainero and S. Aime,
Org. Biomol. Chem., 2005, 3, 112.
5 J. Rudovsky´, P. C´ıgler, J. Kotek, P. Hermann, P. Vojt´ısˇek, I. Lukesˇ, J. A.
Peters, L. Vander Elst and R. N. Muller, Chem.–Eur. J., 2005, 11, 2373.
6 P. Vojt´ısˇek, P. C´ıgler, J. Kotek, J. Rudovsky´, P. Hermann and I. Lukesˇ,
Inorg. Chem., 2005, 44, 5591.
Fig. 5 Comparison of 1H NMRD profiles at 37 ◦C of Gd(III) com-
plexes: [Gd(DTPA)(H2O)]2− (line), [Gd(DTTAP-bz-NH2)(H2O)]2− (ꢂ),
Gadomer 17 (ꢀ), inulin-(Gd(DTTAP))23 (᭹), G5-(Gd(EPTPA))111 (ꢁ) and
G5-(Gd(DTTAP))63 (ꢃ).
7 M. Botta, Eur. J. Inorg. Chem., 2000, 399.
8 E. C. Wiener, M. W. Brechbiel, H. Brothers, R. L. Magin, O. A. Gansow,
D. A. Tomalia and P. C. Lauterbur, Magn. Reson. Med., 1994, 31, 1.
´
9 E. To´th, D. Pubanz, S. Vauthey, L. Helm and A. E. Merbach, Chem.–
Eur. J., 1996, 2, 1607.
hydration sphere was included in the analysis of the longitudinal
relaxation rates.
10 L. D. Margerum, B. K. Campion, M. Koo, N. Shargill, J.-J. Lai, A.
Marumoto and P. C. Sontum, J. Alloys Compd., 1997, 249, 185.
11 L. H. Bryant, Jr., M. W. Brechbiel, C. Wu, J. W. M. Bulte, V. Herynek
and J. A. Frank, J. Magn. Reson. Imaging, 1999, 9, 348.
The rigidity of the dendrimer complex, expressed by the term S2,
20
is lower than for the analogous G5-(Gd(EPTPA))111
,
probably
´
12 G. M. Nicolle, E. To´th, H. Schmitt-Willich, B. Radu¨chel and A. E.
due to the limited loading of G5-(Gd(DTTAP))63. Despite this
lower rigidity, G5-(Gd(DTTAP))63 has a high relaxivity at higher
magnetic fields, related to the presence of second sphere water
molecules.
Merbach, Chem.–Eur. J., 2002, 8, 1040.
13 J. C. Roberts, M. K. Bhalgat and R. T. Zera, J. Biomed. Mater. Res.,
1996, 30, 53.
14 D. A. Tomalia, B. Huang, D. R. Swanson, H. M. Brothers and J. W.
Klimash, Tetrahedron, 2003, 59, 3799.
The unsubstituted amino groups on the dendrimer surface
may offer a convenient way to further derivatization of the
macromolecule. The attachment of targeting groups could allow
for delivering the contrast agent to a place of interest in the body.
Functionalization of the dendrimer with other probes can lead
to multimodal imaging agents (dyes for near infrared imaging,
radioactive tracers for SPECT, etc.).
15 P. Lebdusˇkova´, J. Kotek, P. Hermann, L. Vander Elst, R. N. Muller, I.
Lukesˇ and J. A. Peters, Bioconjugate Chem., 2004, 15, 881.
16 G. Lipari and A. Szabo, J. Am. Chem. Soc., 1982, 104, 4546.
17 G. Lipari and A. Szabo, J. Am. Chem. Soc., 1982, 104, 4559.
18 C. Amman, P. Meyer and A. E. Merbach, J. Magn. Reson., 1982, 46,
319.
19 F. Yerly, VISUALISEUR 2.3.4., Lausanne, Switzerland, 1999; F. Yerly,
OPTIMISEUR 2.3.4., Lausanne, Switzerland, 1999.
´
20 S. Laus, A. Sour, R. Ruloff, E. To´th and A. E. Merbach, Chem.–Eur. J.,
2005, 11, 3064.
21 D. H. Powell, O. M. N. Dhubhghaill, D. Pubanz, L. Helm, Y. S. Lebedev,
W. Schlaepfer and A. E. Merbach, J. Am. Chem. Soc., 1996, 118, 9333.
22 S. Lis, J. Alloys Compd., 2002, 341, 45.
23 N. Arnaud and J. Georges, Spectrochim. Acta, Part A, 2003, 59, 1829.
24 O. V. Yazyev, L. Helm, V. G. Malkin and O. L. Malkina, J. Phys. Chem.
A, 2005, 109, 10997.
25 A. Borel, L. Helm and A. E. Merbach, Chem.–Eur. J., 2001, 7, 600.
26 F. A. Dunand, A. Borel and A. E. Merbach, J. Am. Chem. Soc., 2002,
124, 710.
Acknowledgements
This research was financially supported by the Swiss National Sci-
ence Foundation and the Swiss Secretariat of State for Education
and Research (SER). We also thank to GACR (No. 203/03/0168)
ˇ
for support. The work was carried out in the frame of the EC
3406 | Dalton Trans., 2006, 3399–3406
This journal is
The Royal Society of Chemistry 2006
©