Gd-complexes of DTPA-bis(amide) conjugates of tranexamic acid and its esters with high relaxivity and stability for magnetic resonance imaging

Article abstract of DOI:10.1039/b719440d

The synthesis and the characterization of a series of DTPA-bis(amide) conjugates of tranexamic acid (L1), its esters (L2-L6), and their Gd(iii) complexes of the type [Gd(L)(H₂O)]?nH₂O (L = L1-L6) are described. Except for the case of L1, all Gd-complexes exhibit greatly enhanced R₁ relaxivity. Highest R₁ reaches up to 12.9 mM^-1 s^-1 for [Gd(L2)(H₂O)]. Such high relaxivity is reflected in the intensity enhancement of the in vivo MRI study on H-ras transgenic mice bearing hepatic tumor when employing [Gd(L2)(H ₂O)] as an MRI contrast agent. Thermodynamic stability constants, conditional stability constants, and the pM values demonstrate higher stability of [Gd(L)(H₂O)]?nH₂O (L = L1-L6) than Omniscan under physiological conditions. The MTT assay performed on these complexes reveals cytotoxicity as low as that for Omniscan in the concentration range required to obtain intensity enhancement in the in vivo MRI study. The Royal Society of Chemistry 2008.

Full text of DOI:10.1039/b719440d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Gd-complexes of DTPA-bis(amide) conjugates of tranexamic acid and its
esters with high relaxivity and stability for magnetic resonance imaging
Sujit Dutta,a Ji-Ae Park,b Jae-Chang Jung,c Yongmin Chang*b,d,e and Tae-Jeong Kim*a
Received 18th December 2007, Accepted 25th February 2008
First published as an Advance Article on the web 13th March 2008
DOI: 10.1039/b719440d
The synthesis and the characterization of a series of DTPA-bis(amide) conjugates of tranexamic acid
(L1), its esters (L2–L6), and their Gd(III) complexes of the type [Gd(L)(H2O)]·nH2O (L = L1–L6) are
described. Except for the case of L1, all Gd-complexes exhibit greatly enhanced R1 relaxivity. Highest
R1 reaches up to 12.9 mM−1 s−1 for [Gd(L2)(H2O)]. Such high relaxivity is reflected in the intensity
enhancement of the in vivo MRI study on H-ras transgenic mice bearing hepatic tumor when employing
[Gd(L2)(H2O)] as an MRI contrast agent. Thermodynamic stability constants, conditional stability
constants, and the pM values demonstrate higher stability of [Gd(L)(H2O)]·nH2O (L = L1–L6) than
R
Omniscanꢀ under physiological conditions. The MTT assay performed on these complexes reveals
R
cytotoxicity as low as that for Omniscanꢀ in the concentration range required to obtain intensity
enhancement in the in vivo MRI study.
the two types, the latter is preferred because of relatively low
osmotic pressure in the body fluids after intravenous adminis-
Introduction
tration.3,4 Yet, a great number of neutral Gd-complexes incorpo-
rating DTPA-bis(amide) ligands are known and some of them
are known to exhibit poor water solubility.5–12 It is worth noting
that high relaxivity and non-cytotoxicity in addition to high water
solubility are the essential criteria for an efficient MRI CA.
We have recently demonstrated that a slight modification of
the ligand such as introduction of polar groups to the alkyl
substituents on the amide N-atoms of DTPA-bis(amide) can lead
to the formation of a series of highly water-soluble Gd-complexes.
To our disappointment, however, their use as MRI CAs has been
frustrating due to poor relaxivity.13 In an effort to overcome this
problem and at the same time meet the three-fold requirement for
an effective MRI CA mentioned above, we have developed a new
strategy to further modify DTPA-bis(amide): introduction of po-
lar alkyl substituents with high molecular weight on the amide N-
atoms in such a way that the polar groups are directed as far away
from the Gd(III) center as possible to minimize any interference
with the water exchange equilibrium between the coordinated and
the bulk water molecules. Suitable candidates for the substituents
to meet the above criteria may include trans-4-(aminomethyl)-
cyclohexanecarboxylic acid (tranexamic acid) and the correspond-
ing esters. Further rationalization for their selection may come
from the fact that tranexamic acid and its derivatives have been
used as antifibrinolytic drugs with no sign of cytotoxicity.14–16
Herein, we report the syntheses of DTPA-bis(amide) conjugates
of tranexamic acid (L1) and its esters (L2–L6) and their Gd(III)
complexes, [Gd(L)(H2O)]·nH2O (L = L1–L6). Also reported are
the studies relevant to the potential application of these complexes
as practical MRI CAs.
Magnetic resonance imaging (MRI) provides high-resolution
three-dimensional images of the internal part of the body de-
pending on the difference in the in vivo distribution of the water
molecules. Noninvasive nature and excellent spatial resolution at
the sub-millimeter level render MRI as a powerful diagnostic
imaging modality. The relatively low sensitivity of MRI can be
overcome by inducing additional contrast in the MR images by
the introduction of a contrast agent (CA) prior to the MRI test.1
These contrast agents catalytically shorten the relaxation time
of the nearby water molecules to enhance the contrast with the
background tissues in the MR images. The enhanced usage of MRI
as the diagnostic imaging modality prompts the development of
efficient MRI CAs.2 The Gd(III) ion is known to possess the highest
paramagnetism of all metal ions and Gd(III) complexes incorporat-
ing macrocyclic or acyclic poly(aminocarboxylate) ligands have so
far been used widely as MRI CAs. The Gd-based MRI CAs cur-
rently available for clinical uses may be classified into two types:
(1) an anionic type such as bis-N-methylglucamine salt of [Gd-
(DTPA)(H2O)]2− (DTPA = diethylenetriamine-N,N,Nꢀ,Nꢀꢀ,Nꢀꢀ-
R
pentaacetic acid) (Magnavistꢀ); (2) a neutral type such as [Gd-
(DTPA-BMA)(H2O)] (DTPA-BMA = N,Nꢀꢀ-bis(methylamide)
R
of DTPA) (Omniscanꢀ) and [Gd(HP-DO3A)(H2O)] (HP-
DO3A = 20-(2-hydroxypropyl) derivative of 1,4,7,10-tetraazacy-
R
clododecane-N,Nꢀ,Nꢀꢀ,Nꢀꢀꢀ-1,4,7-tetraacetic acid) (Prohanceꢀ). Of
aDepartment of Applied Chemistry, Kyungpook National University,
Daegu, 702-701, Korea. E-mail: tjkim@knu.ac.kr; Fax: +82-53-950-6594;
Tel: +82-53-950-5587
bDepartment of Medical & Biological Engineering, Kyungpook National
University, Daegu, 702-701, Korea. E-mail: ychang@knu.ac.kr; Fax: +82-
53-422-2677; Tel: +82-53-420-5471
cDepartment of Biology, Kyungpook National University, Daegu, Korea 702-
701
Experimental
General remarks
dDepartment of Diagnostic Radiology, Kyungpook National University,
Daegu, Korea 702-701
eDepartment of Molecular Medicine, Kyungpook National University,
Daegu, Korea 702-701
All reactions were carried out under an atmosphere of dinitrogen
using the standard Schlenk techniques. Solvents were purified
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2199–2206 | 2199
©

View Article Online
and dried using standard procedures. Diethylenetriamine-N,N,
Nꢀ,Nꢀꢀ,Nꢀꢀ-pentaacetic acid (DTPA) and trans-4-(aminomethyl)-
cyclohexanecarboxylic acid were obtained from TCI and
used without further purification. The N,Nꢀꢀ-bis(anhydride) of
DTPA,17 methyl trans-4-(aminomethyl)cyclohexylxarboxylate,18
ethyl trans-4-(aminomethyl)cyclohexylxarboxylate,18 L1,18 L2,18
L3,18 [Gd(L1)H2O],18 [Gd(L2)H2O],18 and [Gd(L3)H2O]18 were
prepared according to the literature methods. All other commer-
cial reagents were purchased from Aldrich and used as received un-
less otherwise stated. Deionized water was used for all experiments.
75 ◦C for 1 h, after which the mixture was cooled to RT. The solvent
was removed under a reduced pressure, and the residue triturated
thrice with hexane (30 mL). The resulting solid was taken in
methanol, treated with decolorizing charcoal, and the solvent
removed under a reduced pressure. The resulting solid was dried in
vacuo for 6 h. Yield 5.47 g (92%). 1H NMR (d6-DMSO, 400 MHz),
d 8.14 (br s, 3H, NH2HCl), 4.27 (t, J = 5.02, 2H, OCH2CH2OH),
4.00 (t, J = 5.02, 2H, OCH2CH2OH), 2.61 (d, J = 7.04, 2H,
H9), 2.25 (m, 1H, H13), 1.87 (m, 4H, H11/H12), 1.56 (m, 1H,
H10), 1.13 (m, 4H, H11/H12). 13C NMR (d6-DMSO, 100 MHz), d
175.29 (C1), 65.90 (OCH2CH2OH), 59.27 (OCH2CH2OH), 44.43
(CH2NH2), 42.39 (C2), 35.17 (C4), 29.02 (C3, C5), 28.20 (C2,
C6). Anal. Calc. for C10H19NO3HCl·0.5H2O: C, 48.68; H, 8.58; N,
5.68. Found: C, 48.50; H, 8.37; N, 5.70%. FABMS (m/z): calc. for
C10H20NO3, 202.27 ([MH]+); found, 202.05.
1
The H and 13C NMR experiments were performed on a Bruker
Advance 400 or 500 Spectrometer by Korea Basic Science Institute
(KBSI). Chemical shifts were given as d values with reference
to tetramethylsilane (TMS) as an internal standard. Coupling
constants are in Hz. FAB-Mass spectra were obtained by using a
JMS-700 model (Jeol, Japan) mass spectrophotometer. IR spectra
were run on a Mattson FT-IR Galaxy 6030E spectrophotometer
at KBSI. Elemental analyses were performed by Center for
Instrumental Analysis, KNU.
2-Methoxyethyl-trans-4-(aminomethyl)cyclohexylcarboxylate
hydrochloride. The title compound was prepared by the same
procedure for 2-hydroxyethyl-trans-4-(aminomethyl)cyclohexyl-
carboxylate hydrochloride by replacing ethylene glycol with
methoxyethanol. Yield 5.67 g (90%). 1H NMR (d6-DMSO,
400 MHz), d 8.16 (br s, 3H, NH2HCl), 4.11 (t, J = 4.52,
2H, OCH2CH2OCH3), 3.50 (t, J = 4.52, 2H, OCH2CH2OCH3),
3.25 (s, 3H, OCH3), 2.61 (d, J = 6.88, 2H, H9), 2.24 (m, 1H,
H13), 1.86 (m, 4H, H11/H12), 1.57 (m, 1H, H10), 1.11 (m, 4H,
H11/H12). 13C NMR (d6-DMSO, 100 MHz), d 173.79 (C1), 68.74
(OCH2CH2OCH3), 61.89 (OCH2CH2OCH3), 57.04 (OCH3), 43.04
(CH2NH2), 40.93 (C2), 33.74 (C4), 27.63 (C3, C5), 26.81 (C2, C6).
Anal. Calc. for C11H21NO3HCl: C, 52.48; H, 8.81; N, 5.56. Found:
C, 52.29; H, 8.67; N, 5.70%. FABMS (m/z): calc. for C11H22NO3,
216.30 ([MH]+); found, 216. 10.
Potentiometric measurements and computational method
Potentiometric titrations were performed with an automatic
titrator to determine the protonation constants of the DTPA-
bis(amide) ligands and the stability constants of corresponding
metal complexes. The autotitrating system consists of a 798
MPT Titroprocessor, a 728 stirrer and a PT-100 combination
pH electrode (Metrohm). The pH electrode was calibrated using
standard buffer solutions. All calibrations and titrations were
carried out under a CO2-free nitrogen atmosphere in a sealed glass
vessel (50 cm3) thermostatted at 25 0.1 ◦C at an ionic strength of
0.10 mol dm−3 KCl. The concentrations of the metal-ion and the
amide solutions were maintained at approximately 0.6 mmol dm−3.
A CO2-free KOH solution (0.100 mol dm−3) was used as a titrant
to minimize the changes in ionic strength during the titration.
Dioxygen and carbon dioxide were excluded from the reaction
mixtures by maintaining a positive pressure of purified nitrogen
in the titration cell. The electromotive force of the cell is given by
Allyl-trans-4-(aminomethyl)cyclohexylcarboxylate hydrochlo-
ride. The title compound was prepared by following the same
procedure as for 2-hydroxyethyl-trans-4-(aminomethyl)cyclohe-
xylcarboxylate hydrochloride by replacing ethylene glycol
with allyl alcohol. Yield 4.79 g (82%). 1H NMR (d6-DMSO,
400 MHz), d 7.81 (br s, 3H, NH2HCl), 5.71 (m, 1H), 5.16 (m, 2H,
=
=
OCH2CH CH2), 4.32 (m, 2H, OCH2CH CH2), 2.40 (d, J =
7.00, 2H, H9), 2.07 (m, 1H, H13), 1.66 (m, 4H, H11/H12), 1.33
(m, 1H, H10), 0.92 (m, 4H, H11/H12). 13C NMR (d6-DMSO,
◦
E = Eꢀ + Q log[H+] + Ej, and both Eꢀ◦ and Q were determined by
titrating a solution with a known hydrogen-ion concentration at
the same ionic strength, using the acid range of the titration. The
liquid-junction potential (Ej) was found to be negligible under the
experimental conditions employed. The protonation constants of
the ligands and the overall stability constants of various metal
complexes formed in aqueous solutions were determined from the
titration data using the computer program HYPERQUAD.19 The
accuracy of this method was verified by measuring the protonation
and the stability constants for Ca(II), Zn(II), Cu(II) and Gd(III)
complexes of [DTPA-BMA]3−. The results were compared with
literature values.20
=
100 MHz), d 174.77 (C1), 133.12 (OCH2CH CH2), 117.78
=
=
(OCH2CH CH2), 64.46 (OCH2CH CH2), 44.01 (CH2NH2),
42.32 (C2), 35.12 (C4), 29.02 (C3, C5), 28.20 (C2, C6). Anal.
Calc. for C11H19NO2·HCl: C, 56.52; H, 8.62; N, 5.99. Found: C,
56.37; H, 8.54; N, 5.76%. FABMS (m/z): calc. for C11H20NO2,
198.28 ([MH]+); found, 198.10.
Synthesis of ligands
L4. To a stirred suspension of DTPA-bis(anhydride) (1.79 g,
5 mmol) in dry DMF (15 mL) was added 2-hydroxyethyl-trans-
4-(aminomethyl)cyclohexylcarboxylate hydrochloride (2.38 g,
Synthesis of the esters of trans-4-(aminomethyl)-
cyclohexylcarboxylic acid
◦
10 mmol). The mixture was stirred at 70 C for 4 h. The solvent
was removed from the reaction mixture under reduced pressure,
and the residue dissolved in methanol (10 mL). The solution
was passed through a short column of silica gel (60 mesh) with
methanol as an eluent. The residue obtained after removal of the
solvent from the eluate was triturated with a mixture of acetone
and diethyl ether (30 : 70 v/v, 150 mL). The solid product was
2-Hydroxyethyl-trans-4-(aminomethyl)cyclohexylcarboxylate
hydrochloride. To a stirred suspension of trans-4-(aminomethyl)-
cyclohexylcarboxylic acid (3.93 g, 25 mmol) in ethylene glycol
(50 mL) cooled in a ice-bath was added thionyl chloride (3.57 g,
30 mmol) during 10 min. The reaction mixture was then heated at
2200 | Dalton Trans., 2008, 2199–2206
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
isolated by filtration, washed with acetone (3 × 30 mL), and dried
during which time a pale yellow solution resulted. Solvent from
the reaction mixture was stripped off and the residue was taken
in ethanol (40 mL). The resulting solution was refluxed for 2 h.
The reaction mixture was cooled and passed through a Celite
column to remove any solid impurities. The solvent was removed
and the residue taken up in ethanol (5 mL). The title complex
was precipitated as a white solid by adding the ethanolic solution
dropwise into acetone at 0 ◦C. The product was isolated by
filtration, washed thoroughly with acetone, dried in vacuo at 80 ◦C
for 8 h. Yield 1.89 g (88%). Anal. Calc. for C34H56GdN5O15·H2O: C,
37.94; H, 6.74; N, 6.51. Found: C, 37.92; H, 6.48; N, 6.88. FABMS
(m/z): calc. for C34H57GdN5O15, 933.09 ([MH]+), C34H55GdN5O14,
915.08 (MH − (H2O))+; found, 932.77 ([MH]+), 914.84 (MH -
(H2O))+.
◦
1
in vacuo at 70 C for 8 h. Yield 3.93 g (87%). H [d6-DMSO,
400 MHz], d 8.25 (s, 2H, CH2CONH), 4.17 (s, 2H, H2), 3.54
(m, 8H, OCH2CH2OH), 3.40 (m, 8H, H7, H5), 3.07 (m, 4H,
H3/H4), 2.94 (m, 4H, H3/H4), 2.24 (m, 2H, H13), 2.08 (m, 4H,
H9), 1.81 (m, 8H, H11/H12), 1.37 (m, 2H, H10), 1.08 (m, 8H,
H11/H12). 13C NMR (d6-DMSO, 100 MHz), d 175.46 (C1/C8),
175.09 (C1/C8), 172.96 (C14), 172.26 (C6), 65.83 (OCH2CH2OH),
65.26 (OCH2CH2OH), 63.99 (C2), 59.25 (C7), 56.49 (C5), 54.83
(C3/C4), 54.54 (C3/C4), 44.84 (C13), 43.02 (C9), 42.76 (C10),
29.59 (C11), 28.52 (C12). Anal. Calc. for C34H57N5O14·8H2O: C,
45.17; H, 8.14; N, 7.75. Found: C, 45.33; H, 7.86; N, 7.66%. FAB-
MS (m/z): calc. for C34H58N5O14, 760.85 ([MH]+), C34H57N5NaO14
782.83 ([MNa]+); found, 760.65 ([MH]+), 782.60 ([MNa]+).
[Gd(L5)H2O]. This compound was obtained essentially by
following the same procedure as that for [Gd(L4)H2O] by
replacing L4 with L5. Yield 1.98 g (90%). Anal. Calc. for
C36H60GdN5O15·8H2O: C, 39.16; H, 6.94; N, 6.34. Found: C, 39.02;
H, 6.70; N, 6.65. FABMS (m/z): calc. for C36H59GdN5O14, 943.13
(MH − (H2O))+; found, 942.78 (MH − (H2O))+.
[Gd(L6)H2O]. This compound was obtained essentially by
following the same procedure as that for [Gd(L4)H2O] by
replacing L4 with L6. Yield 1.80 g (87%). Anal. Calc. for
C36H56GdN5O13·6H2O: C, 41.89; H, 6.64; N, 6.78. Found: C, 42.09;
H, 6.47; N, 6.97. FABMS (m/z): calc. for C36H55GdN5O12, 907.1
(MH − (H2O))+; found, 906.78 (MH − (H2O))+.
L5. This compound was obtained essentially by following
the same procedure as that for L4 by replacing 2-hydroxyethyl-
trans-4-(aminomethyl)cyclohexylcarboxylate hydrochloride with
2-methoxyethyl-trans-4-(aminomethyl)cyclohexylcarboxylate hy-
drochloride. Yield 3.87 g (83%). 1H [d6-DMSO, 400 MHz],
d 8.31 (s, 2H, CH2CONH), 4.14 (s, 2H, H2), 4.09 (m, 4H,
OCH2CH2OCH3), 3.56 (m, 4H, OCH2CH2OCH3), 3.48 (m, 8H,
H7, H5), 3.36 (m, 4H, H3/H4), 3.23 (s, 6H, OCH2CH2OCH3),
3.08 (m, 4H, H3/H4), 2.93 (m, 4H, H9), 2.21 (m, 2H, H13), 1.76
(m, 8H, H11/H12), 1.37 (m, 2H, H10), 1.08 (m, 8H, H11/H12).
13C [d6-DMSO, 100 MHz], d 175.34 (C14), 172.88 (C1/C8),
172.05 (C1/C8), 170.09 (C6), 70.11 (OCH2CH2OCH3), 63.20
(OCH2CH2OCH3), 58.42 (OCH2CH2OCH3), 56.45 (C2), 54.83
(C7), 54.55 (C5), 51.85 (C3/C4), 49.72 (C3/C4), 44.84 (C13),
42.66 (C9), 37.07 (C10), 29.53 (C11), 28.52 (C12). Anal. Calc.
for C36H61N5O148H2O: C, 46.39; H, 8.33; N, 7.51. Found: C,
46.22; H, 7.97; N, 7.83. FAB-MS (m/z): calc. for C36H62N5O14,
788.9 ([MH]+); C36H61N5NaO14, 810.88 ([MNa]+); found, 788.67
([MH]+), 810.61 ([MNa]+).
Relaxivity
T1 measurements were carried out using an inversion recovery
method with a variable inversion time (TI) at 1.5 T (64 MHz). The
magnetic resonance (MR) images were acquired at 35 different
TI values ranging from 50 to 1750 ms. T1 relaxation times
were obtained from the non-linear least square fit of the signal
intensity measured at each TI value. For T2 measurements
the CPMG (Carr–Purcell–Meiboon–Gill) pulse sequence was
adapted for multiple spin–echo measurements. Thirty four images
were acquired with 34 different echo time (TE) values ranging
from 10 to 1900 ms. T2 relaxation times were obtained from the
non-linear least squares fit of the mean pixel values for the multiple
spin-echo measurements at each echo time. Relaxivities (R1 and
R2) were then calculated as an inverse of relaxation time per mM.
The determined relaxation times (T1 and T2) and relaxivities (R1
and R2) are finally image-processed to give the relaxation time map
and relaxivity map, respectively.
L6. This compound was obtained essentially by following
the same procedure as that for L4 by replacing 2-hydroxyethyl-
trans-4-(aminomethyl)cyclohexylcarboxylate hydrochloride with
allyl-trans-4-(aminomethyl)cyclohexylcarboxylate hydrochloride.
Yield 3.60 g (82%). 1H NMR (d6-DMSO, 400 MHz), d 8.28
=
(s, 2H, CH2CONH), 5.89 (m, 2H, OCH2CH CH2), 5.22 (m,
=
=
4H, OCH2CH CH2), 4.52 (d, J = 4.52, 4H, OCH2CH CH2),
4.15 (s, 2H, H2), 3.55 (m, 4H, H7), 3.48 (m, 4H, H5),
3.36 (m, 4H, H9), 3.08 (m, 4H, H3/H4), 2.94 (m, 4H,
H3/H4), 2.25 (m, 2H, H13), 1.81 (m, 8H, H11/H12), 1.38
(m, 2H, H10), 1.12 (m, 8H, H11/H12). 13C NMR (d6-DMSO,
100 MHz), d 174.94 (C14), 172.93 (C1/C8), 172.13 (C1/C8),
=
=
170.24 (C6), 133.14 (OCH2CH CH2), 117.75 (OCH2CH CH2),
In vitro determination of cell toxicity
=
64.41 (OCH2CH CH2), 56.46 (C2), 54.81 (C7), 54.55 (C5), 53.05
14D Chick cornea stroma primary cells (p1) were used. These
cell lines were obtained from Department of Biology, College
of Natural Sciences, Kyungpook National University. The cells
were grown in 100 cm2 plastic culture dishes (Corning culture
dish) in 10 mL of medium at 37 ◦C and in a humidified 5%
CO2 atmosphere. Cells were maintained in F-12-medium (Gibco)
supplemented with heat-inactivated 10% FCS, 1% chicken serum,
5 mg mL−1 insulin, 10 ng mL−1 human recombinant EGF, 100 IU
mL−1 penicillin, 100 mg mL−1 streptomycin and 200 mg mL−1
gentamicin (all purchased from Gibco). The medium was replaced
every two days, and cells were split into 96-well plate (6 × 104 cells
(C3/C4), 52.27 (C3/C4), 44.85 (C13), 42.70 (C9), 37.09 (C10),
29.58 (C11), 28.55 (C12). Anal. Calc. for C36H57N5O127H2O: C,
49.25; H, 8.15; N, 7.98. Found: C, 49.13; H, 7.80; N, 8.17. FAB-
MS (m/z): calc. for C36H58N5O12, 752.41 ([MH]+); C36H57N5NaO12,
774.39 ([MNa]+); found, 752.55 ([MH]+), 774.50 ([MNa]+).
Synthesis of complexes
[Gd(L4)H2O]. To a solution of L4 (1.81 g, 2 mmol) in dry
pyridine (10 mL) was added gadolinium(III) acetate tetrahydrate
(0.81 g, 2 mmol). The suspension was stirred for 6 h at 70 ◦C
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2199–2206 | 2201
©

View Article Online
well−1 (200 lL)). Various concentrations (0.5–10 mM) of Gd
complexes were added into the culture serum free media and incu-
bated for 24 h. Cell viability/toxicity assessment was performed
by using Cell Counting Kit (CCK-8), which was purchased from
Dojindo Laboratory, Japan. 10 lL of CC8-kit solution was directly
added to each well and the plate was incubated at 37 ◦C for 3 h.
The O.D. was read at 490 nm using an ELISA (Molecular Device,
USA Bio-rad 550 Reader) to determine the cell viability/toxicity
in each well.
Animal tumor model
The mice were maintained according to protocols approved by the
Institutional Animal Care and Use Committee, KNU. 18-Week
old H-ras12V transgenic mouse bearing hepatocellular carcinoma
(35 g) was included in this study. Animals were anaesthetized
with an intramuscular injection of 50 lL xylazine (Rompun:
20 mg mL−1) and 10 lL ketamine (Ketalar: 50 mg mL−1).
[Gd(L2)(H2O)] was injected into a tail vein at a dosage of
1.43 mmol kg−1.
MR Imaging
MR images of anaesthetized mice were obtained pre- and post-
[Gd(L2)(H2O)] injection during 75 min with a 1.5 T scanner (GE
Signa Advantage, GE Medical system, USA) and extremity coil.
The animals were placed in the magnet in a supine position with
the heads firmly fixed. After each measurement the animals were
revived from anaesthesia, and placed in their cages with free access
to food and water. During MRI measurements, the animals were
maintained at approximately 37.0 ◦C using a warm water blanket.
The imaging parameters for FSPGR (fast spoiled gradient echo)
were as follows: flip angle of 60◦, 12 × 12 mm field of view, 256 ×
128 matrix size, 22 axial slices, 2 mm slice thickness, slice gap of
0 mm, repetition time (TR) = 70 ms, echo time (TE) = 3 ms and
number of acquisitions (NEX) = 2.
Scheme 1
hexanecarboxylic acid (tranexamic acid) by treating it with a
slight excess of thionyl chloride in the corresponding alcohol
as a solvent. Simple condensation of DTPA-bis(anhydride) with
two equivalents of acid/esters in DMF resulted in corresponding
DTPA-bis(amides) (L1–L6) in almost quantitative yields. All
these ligands were characterized by analytical and spectroscopic
techniques (1H and 13C NMR and FAB-mass). IR spectroscopy is
also informative in that the presence of carbonyl groups in each
ligand can be confirmed by a pair of intense carbonyl stretching
=
bands assignable to the amide carbonyl (NHC O) and carboxylic
carbonyl (C O) groups in the range 1602–1672 cm .
−1 17
=
These ligands form Gd(III) complexes of the type [Gd(L)-
(H2O)]·nH2O (L = L1–L6) by simple complexation with an
equimolar amount of gadolinium acetate in pyridine as illustrated
in Scheme 1. All complexes were isolated as a white solid by
precipitating in cold acetone from the reaction mixture. The
complexes are highly hygroscopic, and isolated as hydrated solids.
Image analysis
The anatomical locations with enhanced contrast were identified
with respect to hepatocellular carcinoma of the liver on post-
contrast MR images. For quantitative measurement, signal in-
tensities in specific regions of interest (ROI) of 20–40 mm3 were
measured using Advantage Window software (GE Medical, USA).
Multiple areas were sampled throughout the tumor and averaged
to give a mean SI value for that tissue. The percentage of contrast
enhancement in the signal from the tumor was calculated using
eqn (1), where SI is the signal intensity.
Protonation constants and stability constants
The protonation constants (KiH) of the ligands and the stability
constants of the metal complexes are defined in eqn (2) and (3),
respectively, where HiL (i = 1, 2, ...) is the protonated ligand, L
totally deprotonated free ligand, M unhydrolyzed aqua metal ion,
and ML the non-protonated and unhydrolyzed complex.
Enhancement (%) = 100(SIpost − SIpre)/SIpre
(1)
KiH = [HiL]/[Hi−1L][H+]
(2)
(3)
Results and discussion
Synthesis
KML(therm) = [ML]/[M][L]
Scheme 1 shows the preparative method leading to the formation
of DTPA-bis(amides) functionalized by trans-4-(aminomethyl)-
cyclohexanecarboxylates (L) and their Gd(III) complexes of
the type [Gd(L)(H2O)]·nH2O (L = L1–L6). The synthesis
initially involves esterification of trans-4-(aminomethyl)cyclo-
The protonation constants of L1–L6 and the stability constants
of their Gd(III), Ca(II), Zn(II) and Cu(II) complexes were deter-
mined by potentiometric titration, and relevant data are collected
in Table 1 along with those for DTPA and DTPA-BMA for
comparative purposes. It is known that for the DTPA-bis(amide)
2202 | Dalton Trans., 2008, 2199–2206
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 1 Protonation constants (log KiH) of L1–L6, stability (KML), selectivity constants (KSel) and conditional stability constants (KꢀSel) of Gd-complexes
of L1–L6 (I = 0.10 mol dm−3), and pMa values of the complexes of Gd(III), Zn(II), Ca(II) and Cu(II) at pH 7.4
log K (25 ◦C, l = 0.10 M (KCl))
Equilibrium
L1
L2
L3
L4
9.94
5.15
3.72
L5
9.95
5.18
3.79
L6
9.81
4.81
3.67
DTPA-BMAb
DTPAc
[HL]/[L][H]
[H2L]/[HL][H]
[H3L]/[H2L][H]
11.43
10.03
5.72
9.74
5.02
3.21
9.77
5.10
3.31
9.37
4.38
3.31
—
10.49
8.60
4.28
[H5L]/[H4L][H]
pKa
4.47
—
—
—
—
—
2.64
ꢀ
29.60
17.97
18.18
18.81
18.92
18.29
17.06
26.01
[GdL]/[Gd][L]
{log KGdL (pH 7.4)}
22.38
15.71
20.42
18.08
20.55
18.18
21.10
18.56
20.80
18.25
20.99
18.58
16.85
14.84
22.46
18.14
[CaL]/[Ca][L]
{log KCaL (pH 7.4)}
14.81
8.14
7.50
5.16
7.36
4.99
7.85
5.31
7.86
5.31
7.17
4.76
7.17
5.11
10.75
6.43
[ZnL]/[Zn][L]
{log KZnL (pH 7.4)}
16.65
9.98
12.26
9.91
12.32
9.95
11.78
9.24
11.80
9.25
11.03
8.62
12.04
10.02
18.70
14.38
[CuL]/[Cu][L]
{log KCuL (pH 7.4)}
11.87
5.20
13.04
10.70
12.74
10.37
12.46
9.92
12.23
9.68
11.52
9.11
13.03
11.06
21.38
17.06
[log Ksel (Gd/Ca)]
[log Ksel (Gd/Zn)]
[log Ksel (Gd/Cu)]
14.19
5.73
12.14
12.92
8.16
7.38
13.19
8.23
7.81
13.25
9.32
8.64
12.94
9.00
8.57
13.82
9.96
9.47
9.68
4.81
3.82
11.71
3.76
1.08
log Kꢀsel
10.03
12.41
12.51
13.58
13.27
14.23
9.03
7.04
pGd
pCa
pZn
pCu
14.71
7.14
8.98
4.20
17.08
4.16
8.92
9.70
17.18
3.99
8.95
9.37
17.56
4.31
8.24
8.92
17.25
4.31
8.25
8.68
17.58
3.76
7.61
8.11
13.88
4.19
9.06
17.14
5.45
13.39
16.06
10.05
a pM = −log[Mn+
]
at pH 7.4; [Mn+
]
= 1 lmol dm−3; [L]total = 1.1 lmol dm−3
.
b Data obtained from ref. 21. c Data obtained from ref. 31.
free
total
ligands the first protonation constant (K1H) takes place at the
thermodynamic stability of their metal complexes as compared
with that of the corresponding metal complexes of DTPA-BMA,
with L1 exhibiting the highest thermodynamic stability for the
same reason described above.
H
central nitrogen atom, while the second (K2H) and the third (K3
)
at the terminal amine nitrogen atoms.12
Table 1 shows that all ligands (L1–L6) exhibit higher protona-
ꢀ
tion constants (log KiH) and pKa values than DTPA-BMA. It
is probable that the presence of carboxylic acid (L1) or esters
(L2–L6) in the amide side-arms seems to render the protonation
of the amine nitrogen(s) facile by some co-operative interaction
to exhibit higher protonation constants. When the comparison is
made among the present series, L1 shows the highest values and
even higher values than the parent DTPA. In general, substitution
of the acetate groups on the terminal amine nitrogen atoms of
However, the thermodynamic stability constant alone is insuf-
ficient to account for the stability of the complexes under the
physiological condition.21,25 The conditional stability constant or
more frequently the pM value is apt to describe the stability of
complexes under physiologically relevant conditions.26 The pM
value reflects the influence of ligand basicity and protonation
of the complex. Thus, the larger the pM value, the higher the
affinity of the ligand for the metal ion under the given condition.27
Table 1 shows that L1–L6 exhibit higher pM values with Gd(III)
than with Ca(II), Zn(II) or Cu(II); the indication is that the Gd(III)
complexes of L1–L6 are stable enough to avoid any interference
by other endogenous metal ions. In addition to the pM value,
the conditional stability constant (KꢀSel) has also to be taken into
consideration under the physiological condition. This is because
a Gd(III) complex injected as an MRI CA into the physiological
system through the blood pool competes not only with endogenous
metal ions such as Ca(II), Zn(II) and Cu(II) but also with protons
at the physiological pH.26 The conditional stability constant
(KꢀSel) can be calculated by considering all equilibria present at
physiological pH.21 The KꢀSel have been shown to correlate with the
experimental LD50 value.28 Table 1 demonstrates that L1–L6 reveal
higher log KꢀSel values than DTPA-BMA and DTPA suggesting
that their Gd(III) complexes should exhibit little cytotoxicity.
DTPA reduces its basicity, as reflected in the lower log KiH and
ꢀ
pKa values of corresponding DTPA-bis-amide ligands. High
basicity of L1–L6 will surely lead to high thermodynamic stability
of their metal complexes.
Table 1 shows the thermodynamic stability constants for the
complexes of Ca(II), Zn(II), and Cu(II) complexes. A direct
potentiometric method can not be applied for the measurement of
the stability constants of [Gd(L)(H2O)] (L = L1–L6) since they are
formed at low pH. Instead, they were determined by employing
the method of ligand–ligand competition potentiometric titration
between EDTA and L1–L6 for Gd(III) ion.22–24 Thermodynamic
stability of Gd(III) complexes measures the tendency of dissocia-
tion of the complexes in solution to generate free Gd(III) ion, which
shows acute cytotoxicity in the physiological system. The table
shows quite expectedly that high basicity of L1–L6 leads to high
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2199–2206 | 2203
©

View Article Online
Relaxivity
In vivo MRI test
Fig. 1 shows the relaxation time (T1) and relaxivity (R1) maps
The in vivo MRI test was performed with [Gd(L2)(H2O)]. The
solution of this complex was injected into the H-ras transgenic
mice bearing hepatic tumors, and the contrast enhancement of
R
on [Gd(L)(H2O)] (L = L1–L6), Omniscanꢀ, and pure water. The
phantom images were obtained with 1 mM solution of complexes
R
R
and with the same concentration of Omniscanꢀ for comparative
the MR images evaluated using Omniscanꢀ as the control. Fig. 2
purposes. Table 2 summarizes the relaxation times (T1 and T2)
shows the T1-weighted images of hepatocellular carcinoma (HCC)
before and after injection. The signal intensities (SI) after injection
are enhanced when compared with the SI obtained in the absence
of the contrast agent. The MR images of the tumor become
brighter after injection, and the demarcation from the surrounding
tissues becomes clear as a result of the contrast enhancement. The
enhanced contrast of the kidney was observed, thus confirming
the renal excretion of the contrast agent.
R
and relaxivities (R1 and R2) for the Gd-complexes, Omniscanꢀ,
and pure water.
Fig. 2 The T1-weighted images of the tumor bearing H-ras transgenic
mice: (a) pre-injection image; (b) 20 min after injection with [Gd(L2)(H2O)]
at the dosage of 1.43 mmol kg−1
.
Fig. 3 shows the MR signal intensity enhancements (IEs) as a
R
function of time after injection of [Gd(L2)(H2O)] and Omniscanꢀ
for comparative purposes. The most significant feature is that
Fig. 1 T1 and R1 maps on (a) [Gd(L)(H2O)] (L = L1–L3), OmniscanꢀR ,
and pure water; (b) [Gd(L)(H2O)] (L = L4–L6), OmniscanꢀR , and pure
water.
[Gd(L2)(H2O)] exhibits a much higher contrast enhancement than
R
Omniscanꢀ throughout the measurement (∼80 min) at the same
Gd-concentration. It is worth noting, for example, that the IE
The most significant feature of Table 2 is that all complexes
except for [Gd(L1)(H2O)] exhibit significantly high R1 values.
Increased relaxivities with L2–L6 may be explained not only in
terms of increased molecular weight as compared with DTPA-
BMA but also of some co-operative interaction between the
coordinated water and the carboxylic groups on the cyclohexyl
moiety. For instance, the rate of water exchange in L2 and L3
seems to be enhanced by the presence of hydrophobic alkyl esters.
On the other hand, other esters (L4–L6) seem to retard the water-
exchange rate for some unknown reasons. The lowest relaxivity
value found with L1 may be attributed to the interaction of
the carboxylic acid with coordinated water through hydrogen
bonding, thereby lowering the exchange rate.20
with [Gd(L2)(H2O)] becomes nearly 12 times as high as that with
R
Omniscanꢀ in 20 min after injection. Furthermore, such high IE
with [Gd(L2)(H2O)] remains almost steady during the measure-
ment, thus making it possible to acquire a number of MR images.
In vitro toxicity study
An additional observation with [Gd(L2)(H2O)] is that all
mice recovered spontaneously from anaesthesia after the MRI
measurements, thus providing motivation for toxicity studies on
this complex. Fig. 4 shows the results of the MTT assay (MTT =
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide)
performed to evaluate the cytotoxic effects of [Gd(L)(H2O)]
Table 2 T1, R1, T2 and R2 values for the Gd complexes, OmniscanꢀR , and pure watera
Sample
T1/ms
R1/mM−1 s−1
T2/ms
R2/mM−1 s−1
[Gd(L1)(H2O)]
[Gd(L2)(H2O)]
[Gd(L3)(H2O)]
[Gd(L4)(H2O)]
[Gd(L5)(H2O)]
[Gd(L6)(H2O)]
OmniscanꢀR
477.5 9.94
79.05 3.99
78.28 4.68
123.29 4.47
126.98 4.23
123.67 3.61
209.8 5.82
842.5 46.38
2.1 0.05
12.7 0.66
12.9 0.84
8.1 0.28
7.9 0.25
8.1 0.23
4.9 0.14
1.1 0.06
550.9 22.05
113.9 14.39
131.5 0.002
232.54 13.05
227.55 13.53
240.80 12.44
290.8 21.02
1220 167.20
1.8 0.07
8.7 0.88
7.2 1.9
4.3 0.24
4.4 0.26
4.2 0.21
3.4 0.25
0.82 0.12
Water
a Each value is presented as a mean value ( SD).
2204 | Dalton Trans., 2008, 2199–2206
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
described. The potentiality of Gd-complexes as practical MRI
CAs was investigated by measuring not only their R1 relaxivity but
also other relevant physicochemical properties and cytotoxicity.
All Gd-complexes except for the case of L1 exhibit much higher R1
R
relaxivity than Omniscanꢀ; the highest R1 reaches up to 2.6 times
R
that for Omniscanꢀ. Suchhigh relaxivity is reflected intheintensity
enhancement of the in vivo MRI study on H-ras transgenic mice
bearing hepatic tumors when employing these new complexes
as MRI CAs. Thermodynamic stability constants, conditional
stability constants, and the pM values also demonstrate higher
stability of these complexes under physiological conditions. The
MTT assay performed on these complexes reveals cytotoxicity as
R
low as that for Omniscanꢀ in the concentration range required to
obtain intensity enhancement in the in vivo MRI study.
Fig. 3 Time dependent MR signal intensity enhancement of the tumor in
a mouse after injection with [Gd(L2)(H2O)] and OmniscanꢀR at the dosage
Acknowledgements
of 1.43 mmol kg−1
.
This work was supported by a grant from The Advanced Medical
Technology Cluster for Diagnosis and Prediction at KNU from
MOCIE, ROK. Spectral measurements were performed by the
KBSI.
Notes and references
1 P. Caravan, Chem. Soc. Rev., 2006, 35, 512.
2 K. N. Raymond and V. C. Pierre, Bioconjugate Chem., 2005, 16, 3.
3 The Chemistry of Contrast Agents in Medical Magnetic Resonance
´
Imaging, ed. E. To´th and A. E. Merbach, Wiley, Chichester, 2001.
4 P. Caravan, J. J. Ellison, T. J. McMurry and R. B. Lauffer, Chem. Rev.,
1999, 99, 2293.
5 S. W. A. Bligh, A. H. M. S. Chowdhury, M. McPartlin, I. J. Scowen
and R. A. Bulman, Polyhedron, 1995, 14, 567.
6 H. Lammers, F. Maton, D. Pubanz, M. W. Van Laren, H. Van Bekkum,
A. E. Merbach, R. N. Muller and J. A. Peters, Inorg. Chem., 1997, 36,
2527.
7 S. Aime, F. Benetollo, G. Bombieri, S. Colla, M. Fasano and S. Paoletti,
Inorg. Chim. Acta, 1997, 254, 63.
8 X. Zhao, R. Zhuo, Z. Lu and W. Liu, Polyhedron, 1997, 16, 2755.
9 H. Lammers, A. M. van der Heijden, H. van Bekkum, C. F. G. C.
Geraldes and J. A. Peters, Inorg. Chim. Acta, 1998, 268, 249.
10 S. W. A. Bligh, A. H. M. S. Chowdhury, D. Kennedy, C. Luchinat and
G. Parigi, Reson. Med., 1999, 41, 767.
Fig. 4 Relative cell viability (%) of the 14D Chick cornea stroma primary
cells obtained by [Gd(L)(H2O)] and OmniscanꢀR . The standard deviations
(
SD) were obtained on a triplicate analysis (n = 3).
11 Y.-M. Wang, Y.-J. Wang, R.-S. Sheu, G.-C. Liu, W.-C. Liu and J.-H.
Liao, Polyhedron, 1999, 18, 1147.
12 C. F. G. C. Geraldes, A. M. Urbano, M. C. Alpoim, A. D. Sherry, K.-
T. Kuan, R. Rajagopalan, F. Maton and R. N. Muller, Magn. Reson.
Imaging, 1995, 13, 401.
13 S. Dutta, S.-K. Kim, D. B. Patel, T-J. Kim and Y. Chang, Polyhedron,
2007, 26, 3799.
14 C. M. Savahn, F. Merenyi, L. Karlson, L. Widlund and M. Gralls,
J. Med. Chem., 1986, 29, 448.
15 Y. Katsuta, Y. Yoshida, E. Kawai, M. Suetsugu, Y. Kohno, S. Inomata
and K. Kitamura, J. Dermatol. Sci., 2005, 40, 218.
16 J. P. Guo, B. P. Liu, X. C. Lv, Z. C. Tan, B. Tong, Q. Shi and D. F.
Wang, J. Chem. Eng. Data, 2007, 52, 1678.
17 W. C. Eckelman, S. M. Karesh and R. C. Reba, J. Pharm. Sci., 1975,
64, 704.
18 S. Dutta, S.-K. Kim, E. J. Lee, T.-J. Kim, D.-S. Kang, Y. Chang, S. O.
Kang and W.-S. Han, Bull. Korean Chem. Soc., 2006, 27, 1038.
19 P. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 1739.
20 S. Laurent, T. N. Parac-Vogt, K. Kimpe, C. Thirifays, K. Binnemans,
R. N. Muller and L. Vander Elst, Eur. J. Inorg. Chem., 2007, 14, 2061.
21 W. P. Cacheris, S. C. Quay and S. M. Rocklage, Magn. Reson. Imaging,
1990, 8, 467.
22 W. R. Harris and A. E. Martell, Inorg. Chem., 1976, 15, 713.
23 Y. Li, A. E. Martell, R. D. Hancock, J. H. Reibenspies, C. J. Anderson
and M. J. Welch, Inorg. Chem., 1996, 35, 404.
(L = L2–L6) using 14D Chick cornea stroma primary cells.29,30
Each set of experiments employ the complexes with the
concentration range of 0.5–10 mM in a serum-free medium,
along with the control experiment with cells in the absence of any
contrast agent. Separate experiments were also performed with
R
Omniscanꢀ for comparative purposes.
Fig. 4 demonstrates that the cell proliferation and the viability
are not affected when incubated for 24 h. No obvious change is
observed when a comparison is made between the cells of the cell
viability assessment with the contrast agents and those for the
control. These observations indicate that [Gd(L)(H2O)] (L = L2–
L6) have very low cytotoxicity in the concentration range required
for obtaining intensity enhancement in the MR images.
Conclusions
The synthesis and characterization of DTPA-bis(amide) conju-
gates of tranexamic acid (L1), its esters (L2–L6), and their Gd(III)
complexes of the type [Gd(L)(H2O)]·nH2O (L = L1–L6) are
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2199–2206 | 2205
©

View Article Online
24 C. H. Taliaferro, R. J. Motekaitis and A. E. Martell, Inorg. Chem.,
1984, 23, 1188.
25 K. Kumar, M. F. Tweedle, M. F. Malley and J. Z. Gougoutas, Inorg.
Chem., 1995, 34, 6472.
26 S. Laus, R. Ruloff, E. To´th and A. E. Merbach, Chem.–Eur. J., 2003,
9, 3555.
27 C. J. Bannochie and A. E. Martell, Inorg. Chem., 1991, 30, 1385.
28 M. F. Tweedle, J. J. Hahan, K. Kumar, S. Mantha and C. A. Chang,
Magn. Reson. Imaging, 1991, 9, 409.
29 C. Zou and Z. Shen, J. Pharmacol. Toxicol. Methods, 2007, 56, 58.
30 M.-I. Huh, Y.-M. Lee, S.-K. Seo, B.-S. Kang, Y. Chang, Y.-S. Lee, M. E.
Fini, S.-S. Kang and J.-C. Jung, J. Cell. Biochem., 2007, 101, 1222.
31 A. E. Martell and R. M. Smith, Critical Stability Constants, Plenum,
New York, 1974, vol. 1.
2206 | Dalton Trans., 2008, 2199–2206
This journal is
The Royal Society of Chemistry 2008
©