Synthesis and physicochemical characterization of new C-functionalized derivatives of the gadolinium(III) complex with 3,6,10-tris(carboxymethyl)-3,6, 10-triazadodecanedioic acid (H5ttda) exhibiting fast water exchange - Potential paramagnetic reporters for molecular imaging

Article abstract of DOI:10.1002/hlca.200790057

To confirm the observation that [Gd(ttda)] derivatives have a significantly shorter residence time τ_M of the coordinated H₂O molecule than [Gd(dtpa)], four new C-functionalized [Gd(ttda)] complexes, [Gd(4-Me-ttda)] (1), [Gd(4-Ph-ttda)] (2), [Gd(9-Me-ttda)] (3), and [Gd(9-Ph-ttda)] (4), were prepared and characterized (H₅ttda = 3,6,10-tris(carboxymethyl)-3,6,10-triazadodecanedioic acid; H₅dtpa = 3,6,9-tris(carboxymethyl)-3,6,9-triazaundecanedioic acid). The temperature dependence of the proton relaxivity for these complexes at 0.47 T and of the ¹⁷O transverse relaxation rate of H₂¹⁷O at 7.05 T confirm that the proton relaxivity is not limited by the H ₂O-exchange rate. The residence time of the H₂O molecules in the first coordination sphere of the gadolinium complexes at 310 K, as calculated from ¹⁷O-NMR data, is 13, 43, 2.9, and 56 ns for 1, 2, 3, and 4, respectively. At 310 K, the longitudinal relaxivity of 2 is higher than for the parent compound [Gd(ttda)] and the other complexes of the series. The stability of the new compounds was studied by transmetallation with Zn ²⁺ ions. All the new complexes are more stable than the parent compound [Gd(ttda)].

Full text of DOI:10.1002/hlca.200790057

562
Helvetica Chimica Acta – Vol. 90 (2007)
Synthesis and Physicochemical Characterization of New C-Functionalized
Derivatives of the Gadolinium(III) Complexwith 3,6,10-Tris(carboxymethyl)-
3,6,10-triazadodecanedioic Acid (H5ttda) Exhibiting Fast Water Exchange –
Potential Paramagnetic Reporters for Molecular Imaging
by Sophie Laurent, Luce Vander Elst, Antoine Vroman, and Robert N. Muller*
Department of General, Organic and Biomedical Chemistry, NMR and Molecular Imaging Laboratory,
University of Mons-Hainaut, B-7000 Mons
(phone/fax: þ 32-65-373520; e-mail: robert.muller@umh.ac.be)
To confirm the observation that [Gd(ttda)] derivatives have a significantly shorter residence time tM
of the coordinated H2O molecule than [Gd(dtpa)], four new C-functionalized [Gd(ttda)] complexes,
[Gd(4-Me-ttda)] (1), [Gd(4-Ph-ttda)] (2), [Gd(9-Me-ttda)] (3), and [Gd(9-Ph-ttda)] (4), were
prepared and characterized (H5ttda ¼ 3,6,10-tris(carboxymethyl)-3,6,10-triazadodecanedioic acid;
H5dtpa ¼ 3,6,9-tris(carboxymethyl)-3,6,9-triazaundecanedioic acid). The temperature dependence of
the proton relaxivity for these complexes at 0.47 T and of the 17O transverse relaxation rate of H217O at
7.05 T confirm that the proton relaxivity is not limited by the H2O-exchange rate. The residence time of
the H2O molecules in the first coordination sphere of the gadolinium complexes at 310 K, as calculated
from 17O-NMR data, is 13, 43, 2.9, and 56 ns for 1, 2, 3, and 4, respectively. At 310 K, the longitudinal
relaxivity of 2 is higher than for the parent compound [Gd(ttda)] and the other complexes of the series.
The stability of the new compounds was studied by transmetallation with Zn2þ ions. All the new
complexes are more stable than the parent compound [Gd(ttda)].
Introduction. – Current research on tracers for magnetic resonance molecular
imaging (MRMI) is devoted to the development of contrast agents with a high
relaxivity and high specificity towards molecules overexpressed under some patho-
logical conditions. One of the factors limiting the proton relaxivity of the gadolinium
complexes is the exchange rate of the coordinated H2O with bulk. An optimal H2O
residence time tM is comprised between 10 and 50 ns depending on the field strength
[1]. Previous studies have shown that the C(4)-functionalized derivatives of [Gd(dtpa)]
(H5dtpa ¼ 3,6,9-tris(carboxymethyl)-3,6,9-triazaundecanedioic acid) are characterized
by a H2O residence time tM that is shorter than that of the parent compound [2 – 4] and
have a higher stability towards transmetallation by zinc, contrarily to the bis-amides
that have longer tM and lower stability [5]. [Gd(ttda)] (H5ttda ¼ 3,6,10-tris(carboxy-
methyl)-3,6,10-triazadodecanedioic acid) has been shown to have a much faster H2O-
exchange rate than [Gd(dtpa)] [6][7], but its stability towards transmetallation by Zn2þ
ions is very low. Similarly, [Gd(ttda)]-derived bis-amides are characterized by short tM
values (20 – 30 ns at 310 K) but also show a very poor stability towards transmetallation
[8].
In this work, four new C-functionalized [Gd(ttda)] complexes, [Gd(4-Me-ttda)]
(1), [Gd(4-Ph-ttda)] (2), [Gd(9-Me-ttda)] (3), and [Gd(9-Ph-ttda)] (4) were
ꢀ 2007 Verlag Helvetica Chimica Acta AG, Zürich

Helvetica Chimica Acta – Vol. 90 (2007)
563
synthesized with the objective to combine the beneficial effects of the C-substitution
previously observed, i.e., to maintain a high-H2O-exchange rate, and to increase the
stability towards transmetallation by Zn2þ ions.
1
The new complexes were characterized by H- and 17O-NMR relaxometry at
various temperatures with the objective to determine the H2O residence time and its
1
1
influence on the H relaxivity. The H-NMRD profiles were recorded at 310 K and
analyzed by means of the classical model of the inner- and outer-sphere theories. The
stability towards Zn2þ transmetallation was tested by a previously described procedure
[5].
1. Results and Discussion. –1.1. Syntheses. Four C-functionalized H5ttda derivatives
were synthesized, two of them carrying a Me or Ph group at the ethane-1,2-diyl moiety
(Scheme 1), and two of them carrying a Me or Ph group at the propane-1,3-diyl moiety
(Scheme 2) bridge. The synthetic scheme is inspired from the literature [9].
C(4)-Substituted Derivatives 11. The methyl ester hydrochloride 7 of the commer-
cial a-amino acid 6 is treated with propane-1,3-diamine to give the corresponding
amide 8 (Scheme 1). This amide is reduced, the obtained amine 9 alkylated with tert-
butyl bromoacetate, and the pentaester 10 hydrolyzed to give the polyaminocarboxylic
acid ligand 11.
C(9)-Substituted Derivatives 17. The methyl ester hydrochloride 13 of the
commercial b-amino acid 12 is treated with ethane-1,2-diamine and the product 14
reduced. The final ligand 17 is obtained after alkylation of the 1,4,8-triazaoctane 15,
with tert-butyl bromoacetate (!16), followed by hydrolysis (Scheme 2).
1.2. Physicochemical Characterization. 1.2.1. Proton Relaxivity. The variation of the
proton relaxivity r1 [sꢀ1 mmꢀ1] as a function of temperature reflects the temperature
dependence of the inner- and the outer-sphere relaxation mechanisms. While the outer-
sphere relaxivity always increases when the temperature is decreased, the inner-sphere
relaxivity may either increase or decrease: if the H2O exchange between the first
coordination sphere and the bulk is very fast (i.e., tM is smaller than the relaxation time
of the bound nuclei T1M over the whole temperature range), the inner-sphere relaxivity
increases when temperature decreases. By contrast, the inner-sphere relaxivity
decreases when the temperature is lowered if a slow-exchange regime is reached (tM
is larger than T1M).

564
Helvetica Chimica Acta – Vol. 90 (2007)
Scheme 1. Synthesis of H5(4-Me-ttda) (11a; R ¼ Me) and H5(4-Ph-ttda) (11b; R ¼ Ph)
a) MeOH, HCl. b) Et3N, Et2O, propane-1,3-diamine. c) BH3 · THF. d) BrCH2COOtBu, iPr2EtN. e) HCl.
As a result, when the exchange rate is not limiting, the global relaxivity increases on
decreasing the temperature. But if the exchange rate becomes limiting, the relaxivity
tends to reach a plateau or even to decrease at low temperatures. Clearly, the relaxivity
evolution observed for the four [Gd(ttda)] derivatives 1 – 4 shows that tM does not limit
the proton relaxivity (Fig. 1).
1.2.2. 17O-NMR Relaxometry. The H2O residence time tM can be calculated from the
temperature dependence of the 17O water transverse relaxation rate of the gadolinium
complex solutions [10 – 14]. The theoretical adjustment of the experimental data
depends on six parameters: DS= and DH=, the entropy and enthalpy of activation,
respectively; A/ꢀh, the hyperfine-coupling constant between the O-nucleus and the
gadolinium; tV, the correlation time describing the electronic relaxation times; B,
related to the amplitude of the zero field splitting energy; and EV, the activation energy
relative to tV.
The curves representing ln(1/TR2 ) obtained for the four C-functionalized [Gd(ttda)]
derivatives 1 – 4, as well as for the parent compound [Gd(ttda)] (5), do not reach a
maximum in the temperature range investigated and are thus characteristic of a water
exchange that is faster than that for [Gd(dtpa)] (Fig. 2).
The theoretical adjustment of the experimental data shows that the methyl
derivatives [Gd(4-Me-ttda)] (1) and [Gd(9-Me-ttda)] (3) have tM values ranging
between 3 and 13 ns at 310 K, whereas larger values (43 and 56 ns, resp.) are obtained
for the phenyl derivatives [Gd(4-Ph-ttda)] (2) and [Gd(9-Ph-ttda)] (4) (Table 1). In
these fittings, the number of coordinated H2O-molecules was fixed to one to achieve a 9

Helvetica Chimica Acta – Vol. 90 (2007)
565
Scheme 2. Synthesis of H5(9-Me-ttda) (17a; R ¼ Me) and H5(9-Ph-ttda) (17b; R ¼ Ph)
a) MeOH, HCl. b) Et3N, Et2O, ethane-1,2-diamine. c) BH3 · THF. d) BrCH2COOtBu, iPr2EtN. e) HCl.
Fig. 1. Temperature dependence of the proton longitudinal relaxivity r1 of the [Gd(ttda)] derivatives 1 – 4
(B0 ¼ 0.47 T)

566
Helvetica Chimica Acta – Vol. 90 (2007)
Fig. 2. Reduced transverse relaxation rate of 17O (1/T2R ¼ 55.55/(Tp2 *[Gd-complex])) as a function of
temperature for complexes 1 – 4 and the parent complex
Table 1. Parameters of the Theoretical Adjustment of the 17O-NMR Transverse Relaxation Rate Evolution
with Temperature. Errors in parentheses.
310
298
tM
[ns]
DH=
[kJmolꢀ1
DS=
[J molꢀ1 Kꢀ1] [106 radsꢀ1] [1020 sꢀ2
A/ꢀh
B
tV
[ps]
EV
[kJ molꢀ1
]
]
]
1
2
3
4
13.3 (1.9)
43.3 (4.2)
2.9 (0.5)
56.1 (22.8) 42.6 (0.22) 31.1 (3.3)
3.6 (3.6); 40.3 (1.3); 46.4 (4.0);
6.3a), 2.1b) 27.9b)
15.3 (0.44) ꢀ 45.1 (0.9) ꢀ 4.1 (0.3) 1.26 (0.81) 26.2 (6.4) 20.0 (15.4)
35.6 (0.13) 10.8 (0.4)
ꢀ 3.4 (0.1) 6.27 (0.1) 12.2 (0.2) 19.8 (0.4)
ꢀ 3.3 (0.1) 4.99 (0.6) 21.6 (2.7) 20.0 (19.0)
ꢀ 2.8 (0.1) 5.57 (0.21) 2.4 (0.2) 12.6 (1.0)
22.4 (0.03) ꢀ 9.6 (0.1)
[Gd/ttda)]
ꢀ 2.8 (0.3); 4.01 (0.38); 19.5 (1.7);
0.9 (14.9);
11.0b)
51.5 (0.3) 52.1 (0.6)
ꢀ 3.9b)
1.82b)
25a), 22.4b) 1.6a), 1.0b)
4.5 (4.2)
[Gd(ttda)]c) 143 (26)
ꢀ 3.4 (0.1) 2.60 (0.06) 12.3 (0.3)
a) From [6]. b) From [7]. c) From [4][15].
coordination shell of the Gd3þ ion (3 N-atoms, 5 COOꢀ groups, and 1 H2O molecule). It
seems that the Ph derivatives 2 and 4 have a longer tM than the parent compound 5.
However, the values are still close to the optimal value; thus covalent or noncovalent
binding of such complexes to a macromolecular structure should result in very efficient
complexes.
1.2.3. NMRD Profiles. The nuclear magnetic relaxation dispersion (NMRD)
profiles of aqueous solutions of the complexes are shown in Fig. 3. At low fields, [Gd(4-
Me-ttda)] (1), [Gd(9-Me-ttda)] (3), and the parent compound [Gd(ttda)] (5) have

Helvetica Chimica Acta – Vol. 90 (2007)
567
Fig. 3. 1H-NMRD Profiles of aqueous solutions of [Gd(4-Me-ttda)] (1), [Gd(4-Ph-ttda)] (2), [Gd(9-
Me-ttda)] (3), and [Gd(9-Ph-ttda)] (4), each compared to those of [Gd(ttda)] and [Gd(dtpa)]
rather similar relaxivities, but at high fields, complex 3 has a slightly lower relaxivity.
The relaxivity at low field of both Ph derivatives 2 and 4 is similar and larger than for
the parent compound, but at high fields, the 4-phenyl derivative 2 has a larger relaxivity
than [Gd(ttda)]. The parameters obtained from the theoretical adjustment of the
NMRD curves with the classical equations describing the inner-sphere [16 – 17] and
outer-sphere relaxations [18] are summarized in Table 2. In these fittings, some
parameters were fixed to usual values: the number of H2O molecules coordinated to
the Gd3þ ion (q ¼ 1), the distance between the proton nuclei of the inner-sphere H2O
molecule and the Gd3þ (r¼ 0.31 nm), the relative diffusion constant (D ¼ 3.3 10ꢀ9 m2
sꢀ1), and the distance of closest approach for the outer sphere contribution (d ¼
0.36 nm). tM was fixed to the values obtained by 17O-NMR. tV and tSO (the electronic
relaxation time at zero field tSO ¼ 5BtV), describing the electronic relaxation times, and
tR (the rotational correlation time) were optimized for the outer-sphere and the inner-
sphere contributions simultaneously.
For complex 1, the tR value obtained is in good agreement with those characterizing
[Gd(ttda)] and [Gd(dtpa)]; but for the isomer 3, the fitting with an r value of 0.31 nm
310
resulted in an unrealistic tR value (tR310 < 50 ps). A more reasonable value (tR
¼
57 ps) was obtained by increasing r to 0.32 nm. Similarly, for the 9-phenyl derivative
4, a larger r value (r¼ 0.32 nm) had to be used to get an acceptable tR (tR ¼ 68 ps).
310

568
Helvetica Chimica Acta – Vol. 90 (2007)
Table 2. Proton Longitudinal Relaxivity at 0.47 T (20 MHz) and 1.4 T (60 MHz) and Parameters of the
Fittings of the Proton NMRD Profiles (T 310 K)
r1 at 0.47 T
r1 at 1.4 T
[sꢀ1 mmꢀ1
]
tM [ns]a)
tR [ps]
tSO [ps]
tV [ps]
r [nm]a)
[sꢀ1 mmꢀ1
]
[Gd(4-Me-ttda)] 1
[Gd(4-Ph-ttda)] 2
4.1
4.8
3.5
4.5
13
43
61
97
75
48
57
56
68
64
57
54
96
91
83
148
171
171
196
90
23
22
20
14
20
23
38
11
0.31
0.31
0.30
0.31
0.32
0.31
0.32
0.31
[Gd(9-Me-ttda)] 3
[Gd(9-Ph-ttda)] 4
[Gd(ttda)] 5
3.6
3.9
3.9
3.8
3.0
3.4
3.5
3.4
2.9
56
3.6
[Gd(dtpa)] 6b)
143
87
25
0.31
a) Parameter fixed. b) From [15].
On the contrary, for [Gd(4-Ph-ttda)] 2, it was necessary to decrease the r value to
310
0.30 nm to get a reasonable tR (tR ¼ 75 ps). It seems thus that contrary to the 4-substi-
tution, which is rather beneficial, the 9-substitution negatively affects the distance r.
1.3. Transmetallation. The stability of [Gd(ttda)] derivatives towards transmetalla-
tion with Zn2þ is estimated by the decrease of the proton relaxation rate measured in
phosphate-buffer solutions containing Zn2þ ions. When Gd3þ ions are substituted by
Zn2þ ions, Gd3þ ions are indeed released and precipitate as phosphates. Consequently,
the relaxation rate decreases according to the amount of released Gd3þ ions. In a first
step, the stability of the complexes has to be tested in the phosphate buffer in the
absence of Zn2þ ions. For [Gd(dtpa)] or [Gd(dtpa-bma)] (dtpa-bma ¼ 6-(carboxy-
methyl)-3,9-bis[2-(methylamino)-2-oxoethyl]-3,6,9-triazaundecanedioic acid), no
marked change of the relaxation rate is observed after several days, but for [Gd(ttda)]
derivatives, the relaxation rates decrease markedly indicating a rather poor stability in
phosphate buffer (Fig. 4). This decrease is larger for complex 1 than for the parent
complex and smaller for complexes 2 – 4.
After addition of Zn2þ ions, the relaxation-rate decrease is very fast for the methyl
derivatives 1 and 3 and for the parent compound [Gd(ttda)] (Fig. 5) indicating a fast
and large transmetallation process. On the other hand, the Ph substitution has a
beneficial effect on stability. Among the two isomers, the 9-substituted derivative 4 is
the best regarding stability. It is more stable than [Gd(dtpa-bma)], and its stability is
only slightly lower than the one of [Gd(dtpa)]. Although markedly more stable than
the parent compound, [Gd(4-Ph-ttda)] (2) is still less stable than the commercial
[Gd(dtpa-bma)]. Thus, the Ph group seems to stabilize the [Gd(ttda)] structure,
particularly when the substitution is at C(9).
2. Conclusions. – Four new [Gd(ttda)] derivatives have been synthesized, [Gd(4-
Me-ttda)] (1), [Gd(4-Ph-ttda)] (2), [Gd(9-Me-ttda)] (3) and [Gd(9-Ph-(ttda)] (4),
with the objective to obtain a paramagnetic reporter characterized by a very short tM in
order to preserve a high relaxivity after coupling to the vector. From the analysis of the

Helvetica Chimica Acta – Vol. 90 (2007)
569
Fig. 4. Evolution of the proton relaxation rate of [Gd(4-Me-ttda)] (1), [Gd(4-Ph-ttda)] (2), [Gd(9-Me-
ttda)] (3), and [Gd(9-Ph-ttda)] (4) in phosphate-buffer solution. The data for [Gd(ttda)] are given for
comparison.
1H- and 17O-NMR relaxometric data, it appears that: i) All C-functionalized derivatives
1 – 4 have a tM value close to the optimal value in the imaging field range. ii) The Me
derivatives [Gd(4-Me-ttda)] (1) and [Gd(9-Me-ttda)] (3) have a lower low-field
relaxivity than the Ph derivatives [Gd(4-Ph-ttda)] (2) and [Gd(9-Ph-ttda)] (4). iii)
Complexes 1, 3, and 4 have a high-field relaxivity similar to the parent complex
[Gd(ttda)], whereas [Gd(4-Ph-ttda)] (2) has a relaxivity increased by ca. 30% at 1.4 T
as compared to [Gd(ttda)] or [Gd(dtda)]. iv) The stability in the phosphate buffer
solution is significantly increased for the two Ph derivatives 2 and 4 but not for the Me
derivatives. v) The presence of Ph substituents has a beneficial effect on the
transmetallation process by Zn2þ ions, with [Gd(9-Ph-ttda)] 4 showing the best
stability. Indeed, 36% of the relaxivity of [Gd(9-Ph-ttda)] 4 is preserved after 4 days.
It turns out that both Ph derivatives [Gd(4-Ph-ttda)] (2) and [Gd(9-Ph-ttda)] (4)
are interesting complexes since their relaxivity could reach very high values after
inclusion in slowly tumbling systems. Actually, it can be assumed that the longitudinal
relaxivity of [Gd(4-Ph-ttda)] (2) and [Gd(9-Ph-ttda)] (4) included in supramolecular
structures with tR values ranging between 20 and 30 ns could reach values larger than 50
and 70 sꢀ1 mmꢀ1, respectively, in the imaging field region and that their r2 values could
be larger than 150 sꢀ1 mmꢀ1.
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 00/05-258 of the French Community of Belgium. The
support and sponsorship concerted by COST Action D18 ꢁLanthanide Chemistry for Diagnosis and
Therapyꢂ and the EMIL Network of Excellence of the 6th Framework Program of the European
Community are kindly acknowledged.

570
Helvetica Chimica Acta – Vol. 90 (2007)
Fig. 5. Stability with respect to transmetallation. Data for [Gd(ttda)], [Gd(dtpa)], and [Gd(dtpa-bma)]
are given for comparison.
Experimental Part
1. General. H- and 13C-NMR Spectra: Bruker-AMX-300 (Bruker, Karlsruhe, Germany), in D2O or
CDCl3, for d(C), tBuOH as internal standard (Me at d(C) 31.2). MS: Q-Tof 2 mass spectrometer
(Micromass, Manchester, UK); samples in MeOH/H2O 1:1; injection rate 5 ml/min.
2. 17O-NMR Spectroscopy. 17O-NMR Measurements of solns. were performed with 2-ml samples in
10-mm external diameter tubes and a Bruker-AMX-300 spectrometer. The temp. was regulated by air or
N2 flow controlled by a BVT-2000 unit. 17O Transverse relaxation times of distilled water (pH 6.5 – 7)
were measured by using a CPMG sequence and a subsequent two-parameter fit of the data points. The
908 and 1808 pulse lengths were 25 and 50 ms, resp. The 17O T2 of water in a complex soln. was obtained
from linewidth measurement. Broadband proton decoupling was applied during the acquisition of all
17O-NMR spectra. Concentration of the samples was lower than 25 mm.
3. 1H-NMRD Profiles. Proton nuclear magnetic relaxation dispersion (NMRD) profiles were
measured on a Stelar Spinmaster FFC, fast field cycling NMR relaxometer (Stelar, Mede (PV), Italy)
over a range of magnetic fields extending from 0.24 mT to 0.35 T and corresponding to 1H Larmor

Helvetica Chimica Acta – Vol. 90 (2007)
571
frequencies from 0.01 to 15 MHz. Measurements were performed with 0.6-ml samples in 10-mm o.d.
tubes. Additional relaxation rates at 20 and 60 MHz were obtained with a Bruker-Minispec-PC-120 and
mq-60 spectrometer (Bruker, Karlsruhe, Germany), resp. Fitting of the 1H-NMRD was adjusted with a
data-processing software that uses different theoretical models describing observed nuclear-relaxation
phenomena (Minuit, CERN Library) [19][20].
4. Transmetallation. Transmetallation by Zn2þ ions was evaluated by the decrease of the water
longitudinal relaxation rate at 310 K and 20 MHz (Bruker Minispec PC 20) of buffered phosphate solns.
(pH 7) containing 2.5 mm of the Gd-complex and 2.5 mm of Zn2þ [6].
5. Synthesis of the Ligands 11 and 17. 5.1. (S)-3,6,10-Tris(carboxymethyl)-4-methyl-3,6,10-triazado-
decane dioic Acid ¼ N-{(2S)-2-[Bis(carboxymethyl)amino]propyl}-N-{3-[bis(carboxymethyl)amino]pro-
pyl}glycine; 11a; H5(4-Me-ttda)). Methyl l-Alaninate Hydrochloride 7a. l-Alanine (6a; 10 g,
122.2 mmol) was suspended in HCl-saturated dry MeOH (200 ml). The soln. was stirred overnight at
r.t. Then MeOH was evaporated: 7a (98%). 1H-NMR (CDCl3): 4.2 – 4.1 (m, HꢀC(2)); 3.8 (s, MeO); 2.2
br. s, NH2); 1.8 (d, J ¼ 7, Me(3)).
N1-(3-Aminopropyl)-l-alaninamide (8a). A soln. of 7a (17.72 g) in MeOH (36 ml) was treated with
Et3N (26 ml) and Et2O (390 ml) to liberate the amino ester. The precipitate was filtered off and the
filtrate evaporated. To the residual oil, propane-1,3-diamine (147 ml) was added dropwise. The soln. was
stirred for 19 h at r.t. The excess of propane-1,3-diamine was evaporated, MeOH (20 ml) was added, and
the soln. again evaporated. These last 2 operations were repeated 3 ꢁ : 8a (48%). Yellow oil which was
used without further purification. 1H-NMR (CDCl3): 7.7 (br. s, NH, 2 NH2); 3.7 – 3.6 (m, HꢀC(2)); 3.5 (t,
J ¼ 7, 1 CH2); 2.8 (t, J ¼ 7, 1 CH2); 1.6 – 1.5 (m, 1 CH2); 1.4 (d, J ¼ 7, Me(3)).
(S)-2-Methyl-1,4,8-triazaoctane Trihydrochloride (2S)-N1-(3-Aminopropyl)propane-1,2-diamine Tri-
hydrochloride; 9a). A soln. of 8a (7.6 g) in THF (100 ml) was stirred under Ar at ꢀ 108 for 30 min. Then
1m borane in THF (330 ml) was added dropwise, and the soln. was maintained at ꢀ 108 for 1 h. Stirring
was continued for 20 h under reflux. The mixture was cooled at ꢀ 108 and anh. MeOH (33 ml) was
injected to destroy the excess of borane. The mixture was allowed to warm up to r.t., and solvents were
evaporated. Anh. MeOH (33 ml) was added and the mixture evaporated. The residue was recovered in
anh. MeOH (132 ml) sat. with gaseous HCl and heated to reflux for 6 h. The mixture was cooled and left
for 68 h at 08. The precipitate was filtered, and the filtrate was distilled under reduced pressure. The
residue was dissolved in H2O and extracted with Et2O. The org. phase was discarded and the aq. phase
evaporated: 9a (52%). 1H-NMR (D2O, pH ca. 6): 3.6 (t, J ¼ 7, 1 CH2); 3.4 (t, J ¼ 7, 1 CH2); 3.1 – 2.9 (m,
1 CH2, CH); 1.9 – 1.85 (m, 1 CH2); 1.2 (d, J ¼ 7, Me ) .
Di(tert-butyl) (S)-3,6,10-Tris[(2-tert-butoxy)-2-oxoethyl]-4-methyl-3,6,10-triazadodecanedioate
(¼ N-{(2S)-2-{Bis[2-(tert-butoxy)-2-oxoethyl]amino}propyl}-N-{3-{bis[2-(tert-butoxy)-2-oxoethyl]ami-
no}propyl}glycine tert-Butyl Ester; 10a). To a soln. of 9a (6.55 g) and iPr2EtN (50 ml) in DMF (200 ml) at
r.t., tert-butyl bromoacetate (26 ml) was added under N2. Stirring was continued for 14 h at r.t. After
filtration, the solvents were evaporated. The oil was dissolved in AcOEt (400 ml) and H2O (150 ml). The
aq. phase was extracted with AcOEt (3 ꢁ 50 ml). The org. phases were extracted with H2O (50 ml) and
sat. NaHCO3 (50 ml). The org. phase was dried (MgSO4) and concentrated and the residue purified by
1
column chromatography (silica gel (Merck 60), AcOEt): 10a (10%). H-NMR (CDCl3): 3.5 – 3.1 (m,
t
8 CH2, CH); 2.8 – 2.7 (m, CH2); 1.4 (s, 5 Bu); 1.3 (d, J ¼ 7, 1 Me ) .
Ligand 11a (H5(4-Me-ttda)). The pentaester 10a was hydrolyzed with conc. HCl soln. (25 ml) for
24 h. The precipitate was discarded and the soln. washed with Et2O (2 ꢁ 50 ml). The aq. phase was then
1
evaporated and the product isolated by lyophilization: 11a (52%). H-NMR (D2O): 4.0 – 3.9 (m, CH);
3.8 (s, 4 CH2); 3.75 (s, 1 CH2); 3.1 – 3.05 (m, 3 CH2); 1.8 – 1.7 (m, 1 CH2); 1.2 (d, J ¼ 7, Me ) . 13C-NMR
(D2O): 173.0; 171.3; 169.0; 168.9; 168.4; 63.8; 61.5; 59.9; 55.7; 54.5; 52.8; 52.6; 50.1; 49.3; 30.1; 13.4. EI-
MS: 444 (40, [M þ Na]þ), 422 (100, [M þ H]þ).
5.2. (S)-3,6,10-Tris(carboxymethyl)-4-phenyl-3,6,10-triazadodecane dioic Acid ¼ N-{(2S)-2-[Bis(car-
boxymethyl)amino]-2-phenylethyl}-N-{3-[bis(carboxymethyl)amino]propyl}glycine; 11b; H5(4-Ph-ttda).
As described for 11a (Sect. 5.1), from (2S)-2-phenylglycine methyl ester.
(2S)-N1-(3-Aminopropyl)-2-phenylglycinamide (8b). Yield 98%. 1H-NMR (D2O): 7.4 – 7.2 (m, Ph);
4.4 (s, HꢀC(2)); 3.1 (t, J ¼ 7, 1 CH2); 2.5 (t, J ¼ 7, 1 CH2); 1.5 – 1.4 (m, 1 CH2).

572
Helvetica Chimica Acta – Vol. 90 (2007)
(S)-2-Phenyl-1,4,8-triazaoctane Trihydrochloride (¼ N1-[(2S)-2-Amino-2-phenylethyl]propane-1,3-
1
diamine Trihydrochloride; 9b): Yield 54%. H-NMR (CDCl3): 7.4 – 7.2 (m, Ph); 4.0 (t, J ¼ 7, CH); 3.9
(br. s, 2 NH2, NH); 3.7 (t, J ¼ 7, CH2); 3.2 (t, J ¼ 7, CH2); 3.1 (dd, J ¼ 13, 4, CH2); 1.8 – 1.7 (m, CH2).
Di(tert-Butyl) (S)-3,6,10-Tris[2-(tert-butoxy)-2-oxoethyl]-4-phenyl-3,6,10-triazadodecanedioate
(¼ N-{(2S)-2-{Bis[2-(tert-butoxy)-2-oxoethyl]amino}-2-phenylethyl}-N-{3-{bis[2-(tert-butoxy)-2-oxoe-
thyl]amino}propyl}glycine tert-Butyl Ester; 10b): Yield 25%. 1H-NMR (CDCl3): 7.25 – 7.15 (m, 3 H, Ph);
7.1 – 7.0 ( m, 2 H, Ph); 5.1 (t, J ¼ 6, CH); 3.9 (s, 2 CH2); 3.8 (s, 1 CH2); 3.7 (s, 2 CH2); 3.0 (dd, J ¼ 13, 4,
1 CH2); 2.5 (t, J ¼ 7, 1 CH2); 2.4 (t, J ¼ 7, 1 CH2); 2.2 – 2.1 (m, 1 CH2).
Ligand 11b (H5(4-Ph-ttda)): Yield 67%. 1H-NMR (D2O): 7.5 – 7.3 (m, Ph); 5.1 (t, J ¼ 6, CH); 4.1 (s,
1 CH2); 4 (d, J ¼ 7, 2 CH2); 3.9 (s, 2 CH2); 3.5 – 3.3 (m, 2 CH2); 3.1 (t, J ¼ 7, 1 CH2); 2.1 – 2.0 (m, CH2).
13C-NMR (D2O): 175.4; 172.9; 172.4; 169.1; 168.3; 139.4; 134.4; 133.3; 131.6; 60.1; 58.0; 55.8; 53.1; 52.1;
51.7; 50.2; 49.2; 48.3; 30.6. EI-MS: 508 (30, [M þ Na]þ), 486 (100, [M þ H]þ).
5.3. (RS)-3,6,10-Tris(carboxymethyl)-9-methyl-3,6,10-triazadodecanedioic Acid (¼ N-{(3RS)-3-
[Bis(carboxymethyl)amino]butyl}-N-{2-[bis(carboxymethyl)amino]ethyl}glycine; 17a; H5(9-Me-ttda)).
As described for 11a (Sect. 5.1), from (3RS)-3-aminobutanoic acid, but with ethane-1,2-diamine instead
of propane-1,3-diamine in the second step.
(3RS)-3-Aminobutanoic Acid Methyl Ester Hydrochloride (13a): Yield 98%. 1H-NMR (D2O): 3.8 (s,
MeO); 3.1 – 2.7 (m, 1 CH2); 2.5 – 2.4 (m, CH); 1.5 (d, J ¼ 7, Me(4)).
(3RS)-3-Amino-N-(2-aminoethyl)butanamide (14a): Yield 92%. 1H-NMR (D2O): 3.2 – 3.1 (m, CH);
3.1 – 2.9 (m, 1 CH2); 2.5 (t, J ¼ 7, 1 CH2); 2.1 (t, J ¼ 7, 1 CH2); 0.8 (dd, J ¼ 10, 4, Me(4)).
(RS)-7-Methyl-1,4,8-triazaoctane Trihydrochloride ¼ (3RS)-N1-(2-Aminoethyl)butane-1,3-diamine
Trihydrochloride; 15a): Yield 58%. 1H-NMR (D2O): 3.7 – 3.0 (m, 3 CH2, CH); 1.9 – 1.7 (m, 1 CH2); 1.2
(d, J ¼ 7, Me ) .
Di(tert-butyl (RS)-3,6,10-Tris[2-(tert-butoxy)-2-oxoethyl]-9-methyl-3,6,10-triazadodecanedioate
(¼ N-{(3RS)-3-{Bis[2-(tert-butoxy)-2-oxoethyl]amino}butyl}-N-{2-{bis[2-(tert-butoxy)-2-oxoethyl]ami-
1
no}ethyl}glycine tert-Butyl Ester; 16a): Yield 30%. H-NMR (CDCl3): 3.6 (s, 1 CH2); 3.5 – 3.1 (m, CH,
t
5 CH2); 2.9 – 2.7 (m, 2 CH2); 1.7 – 1.4 (m, CH2, 5 Bu); 1.0 (dd, J ¼ 10, 4, Me).
Ligand 17a (H5(9-Me-ttda)): Yield 55%. 1H-NMR (D2O): 3.7 (s, 1 CH2); 3.2 – 2.5 (m, CH, 5 CH2);
2.5 (t, J ¼ 7, 1 CH2); 2.4 (t, J ¼ 7, 1 CH2); 1.8 – 1.7 (m, 1 CH2); 1 (dd, J ¼ 10, 7, Me). 13C-NMR (D2O): 176.7;
173.0; 171.2; 169.1; 168.4; 63.9; 63.7; 60.0; 56.3; 55.7; 54.7; 53.4; 53.2; 49.3; 32.2; 13.5. EI-MS: 466 (26,
[M þ 2 Na]þ), 444 (42, [M þ Na]þ), 422 (100, [M þ H]þ).
5.4. (RS)-3,6,10-Tris(carboxymethyl)-9-phenyl-3,6,10-triazadodecanedioic Acid (¼ N-{2-[Bis(car-
boxymethyl)amino]ethyl}-N-{(3RS)-3-[bis(carboxymethyl)amino]-3-phenylpropyl}glycine (17b; H5(9-
Ph-ttda)). As described for 11a (Sect. 5.1), from (3RS)-3-amino-3-phenylpropano acid but with
ethane-1,2-diamine instead of propane-1,3-diamine in the second step.
1
(3RS)-3-Amino-3-phenylpropanoic Acid Methyl Ester Hydrochloride (13b): Yield 98%. H-NMR
(CDCl3): 8.7 (br. s, NH2); 7.5 – 7.4 (m, 2 H, Ph); 7.3 – 7.1 (m, 3 H, Ph); 4.6 (t, J ¼ 7, CH); 3.8 (s, MeO, 3 H);
3.3 (dd, J ¼ 10, 4, 1 CH2).
(3RS)-3-Amino-N-(2-aminoethyl)-3-phenylpropanamide (14b): Yield 98%. 1H-NMR (CDCl3): 7.4 –
7.2 ( m, Ph); 4.3 (t, J ¼ 7, CH); 3.2 – 3.1 (m, 1 CH2); 2.9 – 2.8 (m, 2 CH2); 1.7 (br. s, 2 NH2, NH).
7-Phenyl-1,4,8-triazaoctane Trihydrochloride (¼(1RS)-N3-(2-Aminoethyl)-1-phenylpropane-1,3-di
amine; 15b): Yield 58%. 1H-NMR (CDCl3): 7.5 – 7.4 (m, 2 H, Ph); 7.3 – 7.2 (m, 3 H, Ph); 5 (br. s,
2 NH2, NH); 4.3 (t, J ¼ 7, CH); 3.1 – 2.8 (m, 2 CH2); 1.65 – 1.55 (m, 1 CH2); 1.5 – 1.4 (m, 1 CH2).
Di(tert-Butyl) (RS)-3,6,10-Tris[2-(tert-butoxy)-2-oxoethyl]-9-phenyl-3,6,10-triazadodecanedioate
(¼ N-{2-{Bis[2-(tert-butoxy)-2-oxoethyl]amino}ethyl}-N-{(3RS)-3-{bis[2-(tert-butoxy)-2-oxoethyl]ami-
no}-3-phenylpropyl}glycine tert-Butyl Ester: Yield 27%. 1H-NMR (CDCl3): 7.3 (s, Ph); 3.7 (t, J ¼ 6, CH);
t
3.5 (s, 2 CH2); 3.45 (s, 2 CH2); 3 (s, CH2); 2.8 – 2.5 (m, 4 CH2); 1.4 (s, 5 Bu).
Ligand 17b (H5(9-Ph-ttda)): Yield 92%. 1H-NMR (D2O): 7.5 – 7.4 (m, Ph); 4.4 (t, J ¼ 6, CH); 3.8 – 3.5
(m, 5 CH2); 3.3 – 3.0 (m, 3 CH2); 2.5 – 2.4 (m, 1 CH2). 13C-NMR (D2O): 176.5; 170.7; 169.7; 168.2; 167.9;
137.2; 130.8; 130.0; 128.4; 61.4; 60.7; 59.6; 55.2; 55.1; 53.9; 51.6; 50.8; 49.8; 31.4. EI-MS: 530 (30, [M þ 2
Na]þ), 486 (100, [M þ H]þ).
6. Synthesis of the Corresponding Gd-Complexes. The Gd3þ complexes were prepared by mixing aq.
solns. of equimolar amounts of hexahydrated GdCl3 and one of the ligands 11a,b or 17a,b. The pH was

Helvetica Chimica Acta – Vol. 90 (2007)
573
adjusted to 6.5 – 7 with NaOH. The absence of free Gd3þ ions was checked with arsenazo(III) indicator.
The mass of the complexes was confirmed by ES-MS.
{N-{(2S)-2-[Bis(carboxymethyl)amino]propyl}-N-{3-[bis(-
carboxymethyl)amino]propyl}glycinato(5ꢀ)}gadolinate(2ꢀ) (1; [Gd(4-Me-ttda)]): ES-MS: 620 (100,
[M þ 2 Na]þ).
{N-{(2S)-2-[Bis(carboxymethyl)amino]-2-phenylethyl}-N-{3-[bis(carboxymethyl)-
amino]propyl}glycinato(5ꢀ)}gadolinate(2ꢀ) (2; [Gd-4-Ph-ttda)]): ES-MS: 682 (100, [M þ 2 Na]þ).
{N-{(3RS)-3-[(Bis(carboxymethyl)amino]butyl}-N-{2-[bis(carboxymethyl)amino]ethyl}glycinato(5ꢀ)}-
gadolinate(2ꢀ) (3; [Gd(9-Me-ttda)]): ES-MS: 620 100, [M þ 2 Na]þ). {N-{2-[Bis(carboxymethyl)ami-
no]ethyl}-N-{(3RS)-3-[bis(carboxymethyl)amino]-3-phenylpropyl}glycinato(5 ꢀ)}gadolinate(2ꢀ) (4;
[Gd(9-Ph-ttda)]): ES-MS: 682 (100, [M þ 2 Na]þ).
REFERENCES
[1] E. Toth, L. Helm, A. E. Merbach, in ꢁThe Chemistry of Contrast Agents in Medical Magnetic
Resonance Imagingꢂ, Ed. E. Toth and A. E. Merbach, Wiley & Sons, New-York, 2001, p. 45.
[2] L. Vander Elst, F. Maton, S. Laurent, F. Seghi, F. Chapelle, R. N. Muller, Magn. Reson. Med. 1997, 38,
604.
´
[3] R. N. Muller, B. Raduchel, S. Laurent, J. Platzek, C. Pierart, P. Mareski, L. Vander Elst, Eur. J. Inorg.
Chem. 1999, 1949.
´
´
[4] S. Laurent, L. Vander Elst, S. Houze, N. Guerit, R. N. Muller, Helv. Chim. Acta 2000, 83, 394.
[5] S. Laurent, L. Vander Elst, F. Copoix, R. N. Muller, Invest. Radiol. 2001, 36, 115.
[6] T.-H. Cheng, T.-M. Ou, C.-R. Li, G.-C. Liu, Y.-M. Wang, Helv. Chim. Acta 2002, 85, 1033.
[7] S. Laus, R. Ruloff, E. Toth, A. Merbach, Chem. – Eur. J. 2003, 9, 3555.
[8] S. Laurent, L. Vander Elst, R. N. Muller, Magn. Reson. Mater. Phys. Biol. Med. 2003, 16(Suppl. 7),
S218.
[9] M. W. Brechbiel, O. A. Gansow, R. W. Atcher, J. Schlom, J. Esteban, D. E. Simpson, D. Colcher,
Inorg. Chem. 1986, 25, 2772.
[10] S. Laurent, F. Botteman, L. Vander Elst, R. N. Muller, Eur. J. Inorg. Chem. 2004, 463.
[11] S. Laurent, F. Botteman, L. Vander Elst, R. N. Muller, Helv. Chim. Acta 2004, 87, 1077.
[12] K. Micskei, L. Helm, E. Brucher, A. E. Merbach, Inorg. Chem. 1993, 32, 3844.
[13] G. Gonzalez, D. H. Powell, V. Tissieres, A. E. Merbach, J. Phys. Chem. 1994, 98, 53.
[14] F. Botteman, G. Nicolle, L. Vander Elst, S. Laurent, A. E. Merbach, R. N. Muller, Eur. J. Inorg.
Chem. 2002, 2686.
[15] S. Laurent, L. Vander Elst, R. N. Muller, Contrast Med. Mol. Imaging 2006, 1, 128.
[16] I. Solomon, Phys. Rev. 1955, 99, 559.
[17] N. Bloembergen, J. Chem. Phys. 1957, 27, 572.
[18] J. H. Freed, J. Chem. Phys. 1978, 68, 4034.
[19] P. Vallet, ꢁRelaxivity of Nitroxide Stable Free Radicals. Evaluations by Field Cycling Method and
Optimisationꢂ, Ph.D. Thesis, University of Mons-Hainaut, Belgium, 1992.
[20] R. N. Muller, D. Declercq, P. Vallet, F. Giberto, B. Daminet, H. W. Fisher, F. Maton, Y. Van
Haverbeke, ꢁProc. ESMRMB, 7th Annual Congressꢂ, 1990, p. 394.
Received October 30, 2006