Noncovalent binding of some new lipophilic gadolinium DTPA complexes to human serum albumin. A structure - Affinity relationship

Article abstract of DOI:10.1002/cbdv.201000194

Four Gadolinium·DTPA complexes bearing long lipophilic alkyl chains were synthesized: two bis[amide] and two 4-substituted derivatives. In two of them (one bis[amide] and one 4-substituted), the alkyl chain ends with a carboxylate function. Their relaxometric properties in H₂O show the self aggregation of Gd·DTPA-BdodecylAmide, the better stability of the 4-substituted derivatives vs. Zn transmetallation, and the very good stability of Gd·(4-(carboxyundecylisothiourea-Bz)DTPA). Amongst the four compounds, only Gd·(4-(carboxyundecylisothiourea-Bz)DTPA) shows a strong interaction with human serum albumin (HSA) as demonstrated by proton relaxometry and ESI mass spectrometry. These data highlight the importance of the negative charge on the alkyl chain in the context of the interaction of Gd·(4-substituted DTPA) derivatives with HSA. Copyright

Full text of DOI:10.1002/cbdv.201000194

2846
CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
Noncovalent Binding of Some New Lipophilic Gadolinium DTPAComplexes
to Human Serum Albumin. A Structure–Affinity Relationship
by Sophie Laurent, Luce Vander Elst, Virginie Henrotte, and Robert N. Muller*
Department of General, Organic and Biomedical Chemistry, Laboratory of NMR and Molecular
Imaging, University of Mons, 19 Avenue Maistriau, B-7000 Mons
(phone: þ32-65-373520; fax: þ32-65-372533; e-mail: robert.muller@umons.ac.be)
Four Gadolinium·DTPA complexes bearing long lipophilic alkyl chains were synthesized: two
bis[amide] and two 4-substituted derivatives. In two of them (one bis[amide] and one 4-substituted), the
alkyl chain ends with a carboxylate function. Their relaxometric properties in H2O show the self
aggregation of Gd·DTPA-BdodecylAmide, the better stability of the 4-substituted derivatives vs. Zn
transmetallation, and the very good stability of Gd·(4-(carboxyundecylisothiourea-Bz)DTPA).
Amongst the four compounds, only Gd·(4-(carboxyundecylisothiourea-Bz)DTPA) shows a strong
interaction with human serum albumin (HSA) as demonstrated by proton relaxometry and ESI mass
spectrometry. These data highlight the importance of the negative charge on the alkyl chain in the context
of the interaction of Gd·(4-substituted DTPA) derivatives with HSA.
Introduction. – Magnetic resonance imaging (MRI) is a powerful and non-invasive
technique routinely used in medical diagnosis. However, some examinations require
the use of contrast media. Most of the clinically used MR contrast agents are
compounds containing a paramagnetic ion capable of enhancing the relaxation rates of
the H2O H-atoms of tissues where they accumulate. In this way, they enable the
characterization of lesions and abnormalities in tissues and blood vessels. Gadolinium
chelates are the most widely used T1-agents. For clinical use, the recommended dose is
0.1 mmol/kg. Because of their low molecular masses, they diffuse in the interstitial
space – and, therefore, their biodistribution is nonspecific – and are rapidly excreted by
the kidneys. To overcome the problem of rapid clearance, complexes of Gd with
reversible binding to blood proteins have been developed. The theory behind these
agents is that the reversible binding allows an extended blood half-time and reduces the
fraction of free chelate available for glomerular filtration in the kidneys. In addition,
the binding to a macromolecule results in an increase of the rotational correlation time
of the Gd chelate leading to an enhancement of its efficiency. For example, MS-325
(Vasovistꢀ), developed by Lauffer et al. [1], consists of a diphenylcyclohexyl group
attached to Gd·DTPA (diethylenetriamine-N,N,N’,N’,N’’-pentaacetic acid) by a
phosphodiester linkage. The presence of the lipophilic group allows the noncovalent
binding to human serum albumin (HSA). The reversible binding leads to a high
relaxivity (the relaxivity is defined as the relaxation rate enhancement produced by one
mmole per l of the paramagnetic compound) [2][3] and, as a result, a low dose of
0.03 mmol/kg produces excellent blood–tissue contrast enhancement. In humans, MR
angiography with MS-325 has shown to provide a safe and effective evaluation in phase
ꢁ 2010 Verlag Helvetica Chimica Acta AG, Zꢂrich

CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
2847
III studies [4]. Competition studies with fluorescent probes suggested that MS-325
binds primarily to site II of HSA in accordance with the Sudlow nomenclature [5]. In
fact, the three-domain structure of the protein provides a variety of binding sites. The
main binding sites were denoted site I and site II: site I corresponds to the hydrophobic
pocket in subdomain IIA, and site II to the pocket within subdomain IIIA. Bulky
heterocyclic anions with the charge located at a fairly central position in the molecule
were usually identified as binding to site I. Neutral aromatic compounds or anionic
aromatic structures with the charge located more peripherally (like MS-325) were
usually identified to bind to site II.
To study the importance of the hydrophobic character and of the charge of contrast
agents with regard to their affinity for HSA, four substituted paramagnetic DTPA
complexes with an hydrophobic alkyl chain longer than ten C-atoms introduced
through bis[amide] functions or at C(4) were synthesized: Gd·DTPA-BdodecylAmide
(Gd·1), Gd·DTPA-BcarboxylundecylAmide (Gd·2), Gd·(4-(dodecylisothiourea-
Bz)DTPA) (Gd·3), and Gd·[4-(carboxylundecylisothiourea-Bz)DTPA] (Gd·4). The
complexes were characterized in vitro in aqueous solution and in the presence of HSA.
Their stability was studied by transmetallation with Zn ions in phosphate buffer, and
their affinities for HSA were evaluated by two techniques, proton relaxation rate
measurements and ESI mass spectrometry.
Results and Discussion. – Synthesis. DTPA-BdodecylAmide (1) and DTPA-
BcarboxyundecylAmide (2)1) were obtained by reaction of the DTPA-bisanhydride
with dodecanamine and 12-aminododecanoic acid, respectively, as described by Kimpe
et al. [6] (Scheme 1). The ligands 4-(dodecylisothiourea-Bz)DTPA (3) and 4-(carboxy-
undecylisothiourea-Bz)DTPA (4)1) were synthesized as described in [7] by reaction of
1-(4-isothiocyanatobenzyl)diethylenetriaminepentaacetic acid and the corresponding
amine (Scheme 2). All these ligands were then complexed with GdCl3.
Proton Relaxation in H2O Bisamide Derivatives Gd · 1 and Gd · 2. Because of the
low solubility of Gd·1 in H2O, the proton longitudinal relaxation rates were recorded in
a solution of EtOH/H2O 5 :95. Fig. 1 shows the paramagnetic relaxation rates Rp1 (R1p is
defined as the observed relaxation rate minus the proton relaxation rate of the solvent)
1
)
For systematic names of the ligands 1–4, see Exper. Part.

2848
CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
Scheme 1. Synthesis of DTPA-BdodecylAmide (1) and DTPA-BcarboxyundecylAmide (2)
Scheme 2. Synthesis of 4-(Dodecylisothiourea-Bz)DTPA (3) and 4-(Carboxyundecylisothiourea-
Bz)DTPA (4)
for all complexes recorded over a field range extending from 0.24 mT to 1.4 T. The
nuclear magnetic relaxation dispersion (NMRD) profile of Gd·1 (Fig. 1,a) exhibits a
marked increase of Rp1, as compared to the parent complex Gd·DTPA, and a maximum
around 20 to 60 MHz. This hump is characteristic of a decreased rotational mobility of
the paramagnetic Gd complex and can be explained by the formation of spontaneous
micelles in the solution. The theoretical fitting of these data cannot be accurately
performed, since the size of the micelles is not controlled. Nevertheless, one can
estimate from the high-field data that the mean tR is in the ns range.
Fig. 1. NMRD Profiles of the four complexes at a concentration of 1 mm in H2O at 310 K. The plain and
the dashed lines were obtained by fitting the data with r values equal to 0.31 and 0.30 nm, respectively.

CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
2849
As expected, the less hydrophobic Gd·2 is more soluble in H2O, and its NMRD
profile (Fig. 1,a) obtained at a concentration of 1 mm has a the classical shape of a low-
molecular-weight Gd complex with a relaxivity at 20 MHz of 4.4 sꢀ1 mmꢀ1. The
theoretical fitting of the data according to the model of inner- and outersphere
paramagnetic contributions [8–10] gives a tR value of 94 ps (Table) in good agreement
with the expected value for a non-aggregated structure. The presence of carboxylate
groups on the alkyl chain thus efficiently prevents the formation of micelles in solution
and increases the solubility.
Table. Parameters Obtained by the Theoretical Fitting of the Relaxometric Data. The following
parameters were fixed: D¼3·10ꢀ9 m2 sꢀ1, d¼0.36 nm, q¼1.
r [nm]
Gd·2a)
Gd·3b)
Gd·4b)
0.31
0.31
0.3
0.31
0.3
tR [ps]
tSO [ps]
tV [ps]
94ꢁ4
84ꢁ2
19ꢁ2
149ꢁ3
85ꢁ1
39ꢁ2
121ꢁ2
73ꢁ1
32ꢁ2
179ꢁ4
105ꢁ2
40ꢁ2
150ꢁ3
87ꢁ1
31ꢁ2
a) tM ¼1500 ns. b) tM ¼100 ns.
4-Substituted Benzyl Derivatives Gd · 3 and Gd · 4. The Rp1 values of Gd·3 and Gd·4
in H2O at a concentration of 1 mm are greater than that of the parent complex over the
whole frequency range (Fig. 1,b) with relaxivities at 20 MHz of 6.7 and 7.2 sꢀ1 mmꢀ1
,
respectively. The absence of humps around 20 MHz excludes the formation of large
aggregates. The fitted parameters (Table) of both complexes were obtained by setting
the tM value at 100 ns, a value close to those obtained for other Gd·(4-substituted
DTPA) derivatives [2][11–14]. Two values of r (0.3 and 0.31 nm) were used. Although
the quality of the fittings is quite similar, the tR values seem more reasonable for an r¼
0.3 nm. The larger tR value of Gd·4 can be related to the presence of the hydrated
carboxylate group.
Stability. In phosphate buffer, no significant change of the relaxivity was observed,
indicating that the complexes are stable in this medium. The possible transmetallation
process by Zn ions was then assessed by measuring the evolution of the Rp1 value of a
phosphate solution buffered at pH 7.0 containing equimolar amounts of Gd complex
and ZnII ions (2.5 mm). The evolution of Rp1 allows estimation of the extent of the
transmetallation process. After ca. 500 min, ca. 98% of Gd·1 is dissociated, whereas the
dissociation can be evaluated to be 26% for Gd·2, and much lower for complexes Gd·3
and Gd·4 (18 and 2%, resp.; Fig. 2). As already reported [15][16], bisamide derivatives
of Gd·DTPA are less stable than 4-substituted derivatives. However, the presence of
the carboxylate groups on the alkyl chains greatly enhanced the stability of Gd·2
probably because of their interaction with the Zn ions. Similarly, the better stability of
Gd·4 as compared to Gd·3 can be related to the presence of the carboxylate group of
the alkyl substituent which can interact with the Zn cation. As compared to the parent
compound Gd·DTPA, Gd·2, Gd·3, and Gd·4 are more stable regarding the Zn
transmetallation process. Moreover, Gd·4 shows a very good stability, similar to that of
the macrocyclic Gd·DOTA complex.

2850
CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
&
*
Fig. 2. Evolution of the transmetallation process by Zn ions vs. time. · · ·: Gd·DTPA; : Gd·1; : Gd·2;
~
^
: Gd·3; : Gd·4.
Interaction with HSA. Proton Relaxation Method. Complex Gd·1. In the presence
of HSA, it was not necessary to add EtOH in order to obtain a clear solution. As
compared to the solution without HSA, the proton relaxation rates are greater over the
whole magnetic field range and maximum at ca. 20–30 MHz (Fig. 3,a). At 20 MHz, the
value of Rp1 is increased by 20% (14.5 sꢀ1 in H2O and 17.4 sꢀ1 in HSA solution). This
slight increase indicates either an absent or a weak affinity for the protein, or the
presence of a moderate interaction with the protein leading to macromolecular species
of similar rotational correlation times of bound and unbound Gd complex. However,
the increase of solubility in the presence of HSA supports the presence of an
interaction with HSA.
Fig. 3. NMRD Profiles of the different complexes at a concentration of 1 mm in the presence and absence
&
*
of 4% HSA at 310 K. a) Gd·1 in 4% HSA ( ) and in H2O (–––); Gd·2 in 4% HSA ( ) and in H2O (· · ·).
&
*
b) Gd·3 in 4% HSA ( ) and in H2O (–––); Gd·4 in 4% HSA ( ) and in H2O (· · ·).
Complex Gd·2. The Rp1 value in HSA solution shows at all fields a small increase
with a slight hump at high fields reaching a Rp1 of 7.40 sꢀ1 mmꢀ1 at 20 MHz (Fig. 3,a).

CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
2851
This can be due to the increase of viscosity resulting from the presence of HSA and/or
to a weak interaction between Gd·2 and HSA.
Complex Gd·3. In the presence of HSA, Rp1 increases to 9.4 sꢀ1 mmꢀ1 at 20 MHz
(Fig. 3,b). As explained above for Gd·2, this weak change does not indicate a strong
binding to HSA.
Complex Gd·4. Fig. 3,b, shows a clear hump at high fields with an Rp1 value of
34.6 sꢀ1 at 20 MHz, close to the value obtained in the literature for MS-325. This
complex thus shows a marked affinity for HSA. The binding of Gd·4 to HSA was,
therefore, investigated more deeply in vitro by measuring the Rp1 evolution of 4% HSA
solutions containing increasing amounts of Gd·4. This quantification of the interaction
was performed at 20 MHz, the frequency range of the NMRD profile where the effect
of the interaction is maximal. In pure H2O, the relaxation rates of Gd·4 increased
linearly with the concentration, and a relaxivity (r1) of 7.2 sꢀ1 mmꢀ1 was determined.
When HSA is present, the increase of Rp1 is much higher and not linear, in agreement
with a strong binding. The fitting of the titration data using Eqn. 1 gave a K value of ca.
9410ꢁ2560 mꢀ1 with 2.0ꢁ0.2 equiv. and independent binding sites and an rc1 value of
43.0ꢁ2.4 sꢀ1 mmꢀ1 (Fig. 4).
Fig. 4. Evolution of the paramagnetic relaxation rate of H2O H-atoms in the presence of various
concentrations of Gd·[4-(carboxyundecylisothiourea-Bz)DTPA] (Gd·4) without (· · ·) and with 4% of
HSA (310 K, 20 MHz). The fitting was performed using Eqn. 1 and with the lower limit of rc1 as largest
values of the apparent relaxivity (42 sꢀ1 mmꢀ1).
Electrospray Mass Spectrometry (ESI-MS). The two more stable complexes Gd·3
and Gd·4 were investigated by ESI-MS. Fig. 5 shows that there is no clear additional
peak after addition of a fourfold excess of Gd·3 to 5 mm HSA as compared to HSA
alone, although some broadening of the peak is observed. These data reflect an absence
of or a weak interaction between the Gd complex and HSA, and agree with the
relaxometric results.

2852
CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
Fig. 5. ESI Mass spectrum of a) HSA (5 mm) and b) HSA (5 mm) with Gd·[4-(dodecylisothiourea-
Bz)DTPA] (Gd·3) at a concentration of 20 mm
On the contrary, under the same conditions, when Gd·4 is added to HSA solution
several additional peaks are clearly detected (Fig. 6,c), indicating the presence of a
strong binding of the Gd complex to HSA and the presence of several stoichiometries.
When the concentrations of Gd·4 and HSA are equal (5 mm; Fig. 6,a), three signals are
seen in each massive: the first signal on the left hand side corresponds to the protein
alone, the larger signal corresponds to a complex between the protein and one molecule
of contrast agent, and the weaker signal on the right hand side results from a complex
with a stoichiometry 1:2. When the concentration of contrast agent is increased to
10 mm (Fig. 6,b), the intensity of the signal corresponding to the protein alone
decreased, and a new stoichiometry appeared. With a greater excess of contrast agent,
additional stoichiometries appeared (Fig. 6,c). These data confirm the significant
interaction of Gd·4 with HSA and the presence of several binding sites.
Fig. 6. ESI Mass spectra of HSA (5 mm) with Gd·[4-
(carboxylundecylisothiourea-Bz)DTPA] (Gd·4) at
a) 5 mm, b) 10 mm, and c) 20 mm

CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
2853
Conclusions. – These data confirm the better stability of Gd·(4-substituted DTPA)
derivatives as compared to bis[amide] ones. They also demonstrate i) that the presence
of a negative charge on the hydrophobic alkyl chain of the Gd·(4-substituted DTPA)
derivative is essential for a marked interaction with HSA, and ii) that bis[amide]
derivatives do not bind efficiently to HSA even when the alkyl chains terminates by a
negatively charged group. In conclusion and as already reported, the grafting of
substituents on the 4-position of Gd·DTPA is beneficial regarding both the stability
versus zinc transmetallation and the potential interaction with albumin.
The authors thank Mrs. Patricia de Francisco for her help in preparing the manuscript. This work was
supported by the FNRS and the ARC Program 05/10-335 of the French Community of Belgium. The
support and sponsorship concerted by COST Action D38 ꢃMetal-Based Systems for Molecular Imaging
Applicationsꢄ and the EMIL program are kindly acknowledged.
Experimental Part
Chemicals. 1-(4-Isothiocyanatobenzyl)diethylenetriaminepentaacetic acid was purchased from
Macrocyclics (Dallas, USA). 12-Aminododecanoic acid and dodecanamine, and non-defatted HSA
(A-1653) were purchased from Sigma-Aldrich (B-Bornem). All reagents were commercially available
(Aldrich or Fluka, B-Bornem) and were used without further purification.
1
Instruments. H- and 13C-NMR Spectra: Bruker AMX 300 spectrometer; recorded in CDCl3 rel. to
Me4Si; d in ppm, J in Hz. ESI-MS: Q-TOF 2 or Q-TOF Ultima (Micromass, Manchester, UK); samples
were solubilized in MeOH. Proton Relaxometry: NMRD Profiles were obtained at 310 K on a fast field
cycling relaxometer (Stelar, I-Mede). The range of magnetic fields covered extends from 0.24 mT to
0.24 T, which corresponds to frequencies varying from 0.01 to 10 MHz. Additional longitudinal
relaxation rates were measured on Bruker Minispecs mq20 and mq60, and Bruker AMX-300
spectrometer at 20 MHz (0.47 T), 60 MHz (1.41 T), and 300 MHz (7.05 T), resp.
The data were analyzed using the classical description which takes into account two contributions:
the inner-sphere model described by Solomon [8] and Bloembergen [9], which refers to short-distance
interactions, and the outer-sphere contribution described by Freed [10], which accounts for the larger-
distance interactions. Various parameters related to the structure and the dynamics of the complex are
used in these models. Some of them are quite similar for all low-molecular-weight complexes, like the
distance of closest approach (d¼0.36 nm), the distance between the H-atoms of the coordinated H2O
molecules and the Gd-atom (r¼0.30–0.31 nm), and the relative diffusion constant which is close to the
value of pure H2O (D¼3.0ꢂ10ꢀ9 m2 sꢀ1 at 310 K) [11]. The structure of the complex determines other
parameters, i.e., q, the number of coordinated H2O molecules; tM, the residence time of these H2O
molecules; tR, the rotational correlation time of the complex; tSO, the electronic relaxation time at very
low fields; and tV, the correlation time describing the modulation of the zero-field splitting (ZFS).
The evolution of the transmetallation process was followed by proton relaxometry on a Minispec
20 MHz as described in [15][16].
Studies of Noncovalent Interactions. Proton Relaxivity Method. The proton relaxation rates (R1obs
)
were measured at a fixed field strength (0.47 T) and temp. (310 K) using a Minispec (Bruker, D-
Karlsruhe). The HSA concentration was set to 0.6 mm (4%), and the concentration of contrast agents
ranged from 0 to 2 mm. The proton data obtained in HSA solution were fitted using Eqn. 1.
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
8
<
9
2
3
2
ꢀ1
ðNP0 þ L0 þ Kꢀ1Þ ꢀ4L0P0
=
NP0 þ L0 þ K
ꢀ
ꢀ
ꢁ
f
f
P
c
4
5
R1obs ¼ 1000 r1L0 þ r1 ꢀ r1
(1)
:
;
2
where RP1obs is defined as the observed relaxation rate minus the proton relaxation rate of the solution
without contrast agent, K is the association constant, P0 is the protein concentration, L0 is the
concentration of the paramagnetic complex studied, N is the number of independent and identical

2854
CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
interaction sites, and rc1 and r1f [sꢀ1 mmꢀ1] are the relaxivities of the complex HSA–contrast agent and the
free contrast agent, resp.
ESI-MS Method. The nanospray source was operated in the pos.-ion mode at a cap. voltage of 1.4 kV.
The samples, dissolved in AcONH4 (50 mm), were injected through needles at a flow rate of a few nl/min.
For this procedure, HSA was desalted by five dilution-concentration steps using Microcon YM-10 from
Millipore (B-Brussels). The concentration of albumin samples, measured by spectrophotometry (UV, l
280 nm; 8452A diode array spectrophotometer Hewlett-Packard, B-Brussels), injected into the mass
spectrometer was 5 mm. Each spectrum is the sum of ca. 400 scans.
Synthesis of Gd Complexes. Ligands. The DTPA bisamides were synthesized as described in the
literature [6][17], and their structures were confirmed by proton and carbon-13 NMR spectroscopy and
mass spectrometry.
DTPA-BdodecylAmide (¼6-(Carboxymethyl)-3,9-bis[2-(dodecylamino)-2-oxoethyl]-3,6,9-triazaun-
1
decanedioic Acid; 1). H-NMR: 8–7.6 (m, 3 OH, 2 NH); 3.7–2.9 (m, 11 CH2); 1.5 (m, 2 CH2); 1.4–1.1
(m, 18 CH2); 0.9 (m, 3 CH2). 13C-NMR: 178.2; 177.1; 170.1; 59.9; 58.7; 57.8; 56.4; 52.6; 38.9; 31.9; 29.7;
29.6; 29.5; 29.3; 28.9; 26.3; 25.7; 25.6; 22.7; 14.1. ESI-MS: 728 ([MþH]þ ).
DTPA-BcarboxylundecylAmide (¼16,19,22-Tris(carboxymethyl)-14,24-dioxo-13,16,19,22,25-penta-
azaheptatriacontanedioic Acid; 2). 1H-NMR: 8.1–7.7 (m, 5 OH, 2 NH); 3.5–3 (m, 11 CH2); 2.4–2.2 (m,
2 CH2); 1.5 (m, 2 CH2); 1.4–1.1 (m, 18 CH2). 13C-NMR: 181.7; 179.8; 176.2; 170.5; 59.9; 58.8; 57.9; 52.3;
49.9; 45.8; 38.9; 34.1; 30.7; 30.1; 29.6; 29.4; 28.8; 26.3; 25.7; 23.6. ESI-MS: 832 ([Mþ2Na]þ ).
Ligands 3 and 4 were synthesized by adding an aq. soln. of dodeanamide and aminolauric acid, resp.,
to a soln. of 1-(4-isothiocyanatobenzyl)diethylenetriaminepentaacetic acid (1.5 equiv.). The pH of the
soln. was set between 9 and 10, and the soln. was stirred for 48 h at r.t. After reaction, the product was
dialyzed on a spectra/Por ꢀ Biotech Cellulose Ester (CE) membrane with a cut-off of 500 (VWR, B-
Leuven) and purified by column chromatography (CC) on silica gel 60 RP-18 (40–63 mm; Merck, D-
Darmstadt) with MeOH/H2O 40 :60 (v/v).
4-(Dodecylisothiourea-Bz)DTPA (¼ N-(2-[Bis(carboxymethyl)amino]-3-{4-[(dodecylcarbamo-
thioyl)amino]phenyl}propyl)-N-{2-[bis(carboxymethyl)amino]ethyl}glycine; 3). 1H-NMR: 7.2–6.8
(AA’BB’, 4 arom. H); 5 (quint., CH); 4 (s, 2 CH2); 3.7 (s, CH2); 3.5–3.4 (m, 2 CH2); 3.3 (s, 2 CH2);
2.9–2.5 (m, 3 CH2); 1.7–1.5 (m, 2 CH2); 1.4–1.3 (m, 8 CH2); 0.9 (t, Me). 13C-NMR: 181.9; 177.8; 175.5;
170.6; 138.5; 133.1; 130.6; 118.7; 61.1; 60.6; 58.9; 57.1; 56.7; 52.9; 49.6; 44.5; 40.8; 33.1; 31.9; 30.1; 29.7;
29.5; 29.4; 29.1; 28.8; 22.7; 14.1. ESI-MS: 727 ([MþH]þ ).
4-(Carboxylundecylisothiourea-Bz)DTPA (¼12-{[(4-{2-[Bis(carboxymethyl)amino]-3-[{2-[bis(carb-
oxymethyl)amino]ethyl}(carboxymethyl)amino]propyl}phenyl)carbamothioyl]amino}dodecanoic Acid;
4). 1H-NMR: 7.1–6.7 (AA’BB’, 4 arom. H); 5 (quint., CH); 4 (s, 2 CH2); 3.7 (s, CH2); 3.5–3.4 (m,
2 CH2); 3.3 (s, 2 CH2); 2.9–2.5 (m, 3 CH2); 2.3 (t, CH2); 1.7–1.5 (m, 4 CH2); 1.4–1.3 (m, 7 CH2).
13C-NMR: 182.7; 182.0; 177.5; 175.5; 170.6; 138.5; 133.4; 130.6; 118.7; 61.1; 60.5; 58.4; 57.1; 56.7; 53.4;
49.6; 44.5; 40.8; 34.0; 31.9; 30.1; 30.0; 29.6; 29.4; 29.1; 28.9; 23.6. ESI-MS: 757 ([MþH]þ ).
Complexes. Equimolar amounts of ligand (1 mmol) [6] and GdCl3 ·6 H2O were mixed. The pH was
adjusted to 5 with pyridine, and the soln. was stirred overnight. The soln. was filtered, and the solvent was
evaporated under vacuum. Xylenol or arsenazo tests confirmed the absence of free Gd [18]. The product
was treated with 50 ml of dist. H2O and then lyophilized, and its structure was confirmed by MS. The
complex was desalted by dialysis on membranes with a molecular weight cutoff (MWCO) value of 500.
Gd·DTPA-BdodecylAmide (Gd·1). HR-ESI-MS: 883.4548 ([MþH]þ , C38H71GdN5O8þ ; calc.
883.4544).
Gd·DTPA-BcarboxylundecylAmide (Gd·2). HR-ESI-MS: 941.3869 ([MþH]þ , C38H65GdN5O1þ2
;
calc. 941.3871).
Gd·(4-(Dodecylisothiourea-Bz)DTPA) (Gd·3). HR-ESI-MS: 925.2379 ([MþH]þ , C34H51GdN5-
Na2O10Sþ ; calc. 925.2393).
Gd·(4-(Carboxylundecylisothiourea-Bz)DTPA) (Gd·4). HR-ESI-MS: 977.1959 ([MþH]þ , C34H47-
GdN5Na3O12Sþ ; calc. 977.1954).

CHEMISTRY & BIODIVERSITY – Vol. 7 (2010)
REFERENCES
2855
[1] R. B. Lauffer, D. J. Parmelee, H. S. Ouellet, R. P. Dolan, H. Sajiki, D. M. Scott, P. J. Bernard, E. M.
´
Buchanan, K. Y. Ong, Z. Tyeklar, K. S. Midelfort, T. J. McMurry, R. C. Walovitch, Acad. Radiol.
1996, 3, S356.
´
[2] R. N. Muller, B. Radꢂchel, S. Laurent, J. Platzek, C. Pierart, P. Mareski, L. Vander Elst, Eur. J. Inorg.
Chem. 1999, 1949.
[3] P. Caravan, N. J. Cloutier, M. T. Greenfield, S. A. McDermid, S. U. Dunham, J. W. M. Bulte, J. C.
Amedio Jr., R. J. Looby, R. M. Supkowski, W. De W. Horrocks, T. J. McMurry, R. B. Lauffer, J. Am.
Chem. Soc. 2002, 124, 3152.
[4] M. Goyen, M. Edelman, P. Perreault, E. OꢄRiordan, H. Bertoni, J. Taylor, D. Siragusa, M.
Sharafuddin, E. R. Mohler III, R. Breger, E. K. Yucel, K. Shamsi, R. M. Weisskoff, Radiology 2005,
236, 825.
[5] G. Sudlow, D. J. Birkett, D. N. Wade, Mol. Pharmacol. 1976, 12, 1052.
´
[6] K. Kimpe, T. N. Parac-Vogt, S. Laurent, C. Pierart, L. Vander Elst, R. N. Muller, K. Binnemans, Eur.
J. Inorg. Chem. 2003, 3021.
[7] C. Henoumont, V. Henrotte, S. Laurent, L. Vander Elst, R. N. Muller, J. Inorg. Biochem. 2008, 102,
721.
[8] I. Solomon, Phys. Rev. 1955, 99, 559.
[9] N. Bloembergen, J. Chem. Phys. 1957, 27, 572.
[10] J. H. Freed, J. Chem. Phys. 1978, 68, 4034.
[11] L. Vander Elst, A. Sessoye, S. Laurent, R. N. Muller, Helv. Chim. Acta 2005, 88, 574.
[12] L. Vander Elst, F. Maton, S. Laurent, F. Seghi, F. Chapelle, R. N. Muller, Magn. Reson. Med. 1997,
38, 604.
´
´
[13] S. Laurent, L. Vander Elst, S. Houze, N. Guerit, R. N. Muller, Helv. Chim. Acta 2000, 83, 394.
[14] S. Laurent, F. Botteman, L. Vander Elst, R. N. Muller, Eur. J. Inorg. Chem. 2004, 463.
[15] S. Laurent, L. Vander Elst, F. Copoix, R. N. Muller, Invest. Radiol. 2001, 36, 115.
[16] S. Laurent, L. Vander Elst, R. N. Muller, Contrast Media Mol. Imaging 2006, 1, 128.
[17] W. P. Cacheris, R. C. Grabiak, A. C. Lee, T. J. Richard, T. H. Goodin, R. J. Kaufman, ꢃEmulsions of
Lipophilic Paramagnetic Complexes as MRI Contrast Mediaꢄ, in ꢃProceedings of the 4th Special
Topic Seminar of the European Magnetic Resonance Forumꢄ, Eds. P. A. Rinck, R. N. Muller, 1994,
pp. 37–51.
[18] A. Barge, G. Cravotto, E. Gianolio, F. Fedeli, Contrast Media Mol. Imaging 2006, 1, 184.
Received July 10, 2010