Relaxivity and transmetallation stability of new benzyl-substituted derivatives of gadolinium-DTPA complexes

Article abstract of DOI:10.1002/hlca.200490098

In our efforts of finding new specific contrast agents of higher relaxivity and selectivity, we have prepared the two new benzyl-functionalized DTPA ('diethylenetriamine pentaacetate') gadolinium complexes (S)-3 and (R,S)-4, and compared their properties with those of the known regioisomers (S)-2 and (S)-1. The theoretical fitting of the reduced transverse relaxation rates of the ¹⁷O-nucleus of H₂O gave values for the water-residence time (τ_M) of 86-143 ns at 310 K, values that are not limiting the proton relaxivity at body temperature. ¹H-NMRD (nuclear magnetic-relaxation dispersion) Profiles showed that the relaxivity of 1-4 (r₁=4.3-5.1 s^-1 mM^-1 at 20 MHz and 310 K) is higher than for the Gd-DTPA parent compound 5. Transmetallation assessment demonstrated that all substituted compounds, except for (S)-2, are more stable than 5. The highest stability towards Zn²⁺-induced transmetallation was achieved with complexes 3, 1, and 4 (in decreasing order). Apparently, the steric hindrance of the benzyl substituents in positions 5, 4, and 2, respectively, favorably reduces the accessibility of Zn ions. From a synthetic point of view, 4-substituted DTPA complexes of type 1 are more readily accessible than 5-substituted compounds of type 3. Therefore, the former seem to be superior for linking substituted DTPA complexes to macromolecules or specific vectors.

Full text of DOI:10.1002/hlca.200490098

Helvetica Chimica Acta Vol. 87(2004)
1077
Relaxivity and Transmetallation Stability of New Benzyl-Substituted
Derivatives of GadoliniumÀDTPA Complexes
by Sophie Laurent, FranÁois Botteman, Luce Vander Elst, and Robert N. Muller*
NMR Laboratory, Department of Organic Chemistry, University of Mons-Hainaut, B-7000 Mons
(Phone/fax: ‡32-65-373 520; e-mail: robert.muller@umh.ac.be)
In our efforts of finding new specific contrast agents of higher relaxivity and selectivity, we have prepared
the two new benzyl-functionalized DTPA (−diethylenetriamine pentaacetate×) gadolinium complexes (S)-3 and
(R,S)-4, and compared their properties with those of the known regioisomers (S)-2 and (S)-1. The theoretical
fitting of the reduced transverse relaxation rates of the 17O-nucleus of H2O gave values for the water-residence
time (tM) of 86 143 ns at 310 K, values that are not limiting the proton relaxivity at body temperature.
1H-NMRD (nuclear magnetic-relaxation dispersion) Profiles showed that the relaxivity of 1 4 (r1 ˆ 4.3
5.1 sÀ1 mmÀ1 at 20 MHz and 310 K) is higher than for the GdÀDTPA parent compound 5. Transmetallation
assessment demonstrated that all substituted compounds, except for (S)-2, are more stable than 5. The highest
stability towards Zn2‡-induced transmetallation was achieved with complexes 3, 1, and 4 (in decreasing order).
Apparently, the steric hindrance of the benzyl substituents in positions 5, 4, and 2, respectively, favorably
reduces the accessibility of Zn ions. From a synthetic point of view, 4-substituted DTPA complexes of type 1 are
more readily accessible than 5-substituted compounds of type 3. Therefore, the former seem to be superior for
linking substituted DTPA complexes to macromolecules or specific vectors.
Introduction. Synthesis of paramagnetic complexes capable to target specific
pathologies is often performed through amide linkages with vectorizing moieties. Such
a functionalization, however, has an adverse effect on the exchange rate of coordinated
H2O molecules, which is slowed down compared to that of the parent compounds, such
as gadolinium (Gd) complexes of either DTPA (−diethylenetriamine pentaacetate×) or
DOTA (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetate). This effect inhibits the
possible gain in relaxivity brought by the lengthening of the molecular tumbling time
[1][2]. In this context, studies of the water-exchange rate of different C-functionalized
complexes are pertinent.
Recently, several paramagnetic complexes carrying benzyl (Bn) groups have been
synthesized and characterized by multinuclear NMR [3]. It has been observed that a
Bn group in 4- position (trivial atom-numbering) has a beneficial influence on the H2O-
exchange rate of the Gd complex. This group, however, can theoretically be attached at
four different positions of the DTPA skeleton (see formulae 1 5). In this work, we
report the relaxivity of the four possible regioisomerically substituted GdÀDTPA
complexes 1 4, two of which have not been synthesized before.
Results and Discussion. 1. Synthesis. The syntheses of the Gd complexes (S)-1 and
(S)-2 have been reported previously [3][4]. The synthesis of a DTPA derivative
substituted in 5-position (complex of type 3) was described by Brechbiel [5]. For the
synthesis of the corresponding Gd-free, benzyl-substituted ligand, the dipeptide 6,

1078
Helvetica Chimica Acta Vol. 87(2004)
obtained in ca. 30% yield by condensation of the Phe and Gly derivatives 7 and 8,
respectively, was transformed into the substituted diethylenetriamine derivative 9
(Scheme 1). Thereby, 8 was prepared from N-Boc-protected Phe by N-hydroxysuccin-
imide activation1) [6][7]. The poor coupling efficiency for the condensation of 7 and 8
was attributed to steric hindrance, since modification of the experimental conditions
(reaction time, solvent, temperature) did not improve the yield. The reduction of 6 to
(S)-9 was achieved with borane, and the product (hydrochloride from) was purified by
precipitation. Finally, multiple alkylation of (S)-9 with tert-butyl 2-bromoacetate
afforded the pentaester 10, which was hydrolyzed with HCl to the final ligand (S)-11.
Scheme 1. Synthesis of the New Chiral Ligand 11 Required for the Preparation of the Gd Complex (S)-3
Two methods have been proposed for the insertion of a Bn group in 2-position of
DTPA (ligand 12 for complexes of type 4): either reductive amination of a protected
diethylenetriamine derivative with an a-keto acid (Scheme 2,a) or selective mono-
deprotonation of protected DTPA, followed by alkylation (Scheme 2,b).
1
)
Similar yields were obtained for activation with 1-hydroxy-1H-benzotriazole (HOBt).

Helvetica Chimica Acta Vol. 87(2004)
1079
The reductive-amination procedure has been described by Westerberg for the
synthesis of a (4-nitrophenyl)methyl derivative [8]. We, thus, adapted this method for
the preparation of 12 (Scheme 2,a). The central N-atom of terminally Boc-protected
diethylenetriamine, obtainable by −Boc-on× methodology from the unprotected
precursor, was alkylated at the central N-atom with 2-chloro-N,N-diethylacetamide.
The following reductive amination with NaBH3CN was slow (4 d) and gave yields of
only ca. 20%. Moreover, the complex mixture of products obtained required
purification of 13 by HPLC. The latter was then alkylated with 2-bromoacetic acid,
followed by ester hydrolysis under strongly alkaline conditions (7m NaOH solution,
reflux). The overall yield of 12 was only ca. 4%. Therefore, we decided to investigate
the second route shown in Scheme 2,b.
Scheme 2. Two Different Synthetic Approaches towards the New Ligand 12 (racemate) Required for the
Preparation of the Gd Complex (R,S)-3. Route b (selective deprotonation and alkylation) was much more
convenient than route a (reductive amination).

1080
Helvetica Chimica Acta Vol. 87(2004)
The three-step synthesis of 12 via carbanion 14 turned out to be very convenient
(Scheme 2,b) since it is not only fast, but avoids the difficulties inherent to polyamine
monoalkylation. The carbanion 14 was obtained in situ at low temperature by treating
the DTPA pentamethyl ester 15 with a strong base. The success of this step strongly
depended on the nature of the base, the solvent, the electrophile, and the ratios of
reactants used. Hence, we optimized the reaction conditions, using different benzyl
halides, bases (lithium diisopropylamide; LDA), potassium hexamethyldisilazide
(KHMDS), and solvents (THF, THF/HMPA (−hexamethylphosphortriamide×)) in
various ratios. The optimal conditions were met with the use of 2 equiv. of KHMDS and
3 equiv. of benzyl iodide (BnI) in the presence of HMPA. Interestingly, in all protocols
[9][10], monosubstitution in 2-position has been observed predominantly, but the
percentage of unreacted pentaester and of polyalkylation were not negligible. In our
case, monoalkylation of 15 to 16 proceeded in 32% isolated yield. Finally, HCl-
promoted ester hydrolysis (86%) of 16 afforded the desired ligand 12.
2. Relaxometric Studies of Gd Complexes. 2.1. Qualitative and Quantitative
Evaluation of the Water-Residence Time. The influence of the water-residence time,
tM, on the proton relaxivity r1, defined as the increase of relaxation rate in the presence
of 1 mm complex, can be qualitatively assessed by the temperature dependence of the
relaxivity at 20 MHz. For Gd complexes, a decrease in temperature induces an increase
in relaxivity, when the residence time of a coordinated H2O molecule is smaller than the
relaxation time of its H-nuclei. The observed increase of r1 for complexes 1 4 on
lowering the temperature is, thus, characteristic of a nonlimiting H2O exchange
(Fig. 1). As for the unsubstituted parent complex, GdÀDTPA (5), no significant
limiting effect of tM was observed even at low temperatures.
The quantitative evaluation of tM was carried out through the analysis of the
temperature dependence of the reduced transverse relaxation rate of the H2O 17O-
nucleus (Fig. 2). Obviously, the exchange rate of H2O was different for the 4-Bn-
substituted complex (S)-1: the maximum of the curve was observed at lower
temperatures for (S)-1 and for (S)-2, whereas for (R,S)-4 and (S)-3 it occured at a
temperature close to that observed for 5.
The theoretical adjustment of the experimental data, performed as described
previously [3][11][12], allowed for the determination of various parameters: A/ꢀh, the
hyperfine-coupling constant between the O-nucleus of bound H2O and Gd3‡; tv, the
correlation time modulating the electronic relaxation of Gd3‡; Ev, the activation energy
related to tv; the B-parameter, which is related to the mean-square of the zero-field-
splitting energy (B ˆ 2.4D2); and DH= and DS=, the enthalpy and entropy of activation,
respectively, of the water-exchange process. Water-residence times and other param-
eters of the theoretical adjustment are given in Table 1.
The water-residence time of the Gd complex (R,S)-4 (308 ns at 298 K) was
comparable to that of the Gd complexes of BOPTA2) (321 ns at 298 K) [13] and of
COPTA3) (290 ns at 298 K) [14]. Complexes 1 and 2 were characterized by smaller
values of tM than GdÀDTPA (5), whereas, surprisingly, 3 and 4 had tM values close to
that of the parent compound. Substitution at the 4- and 7-position of DTPA, such as in
2
)
)
BOPTA ˆ −Benzyloxypropionictetraacetate× (trivial name).
3
COPTA ˆ 4-Carboxy-5,8,11-tris(carboxymethyl)-1-cyclohexyl-2-oxa-5,8,11-triazatridecan-13-oic acid.

Helvetica Chimica Acta Vol. 87(2004)
1081
Fig. 1. Temperature dependence of the proton longitudinal relaxivity of the Gd complexes 1
5
Table 1. Parameters Derived by Theoretical Adjustment of the Experimental 17O-NMR Transverse-Relaxation
Rates of the Gd Complexes 1 5. For a description of the physical symbols, see text.
(S)-1
(S)-2
(S)-3
(R,S)-4
5
t3M10 [ns]
86 Æ 8
87 Æ 5
122 Æ 45
50.5 Æ 0.9
49.9 Æ 2.8
À 2.8 Æ 1.1
2.73 Æ 0.13
Æ 0.9
143 Æ 15
143 Æ 25
DH= [kJ molÀ1
]
50.1 Æ 0.1
51.8 Æ 0.4
À 3.2 Æ 0.1
2.80 Æ 0.06
10.3 Æ 0.3
8.7 Æ 3.9
46.2 Æ 0.1
39.0 Æ 0.2
À 2.8 Æ 0.1
2.79 Æ 0.12
15.6 Æ 0.715.5
3.7 Æ 2.6
46.4 Æ 0.2
35.7 Æ 0.3
À 3.1 Æ 0.6
2.38 Æ 0.13
15.6 Æ 1.0
9.4 Æ 1.9
51.5 Æ 0.3
52.1 Æ 0.6
À 3.4 Æ 0.1
2.60 Æ 0.06
11.5 Æ 0.3
4.5 Æ 4.2
DS= [J molÀ1 KÀ1
]
A/ꢀh [106 rad sÀ1
]
B [1020 sÀ2
t3v10 [ps]
]
Ev [kJ molÀ1
]
7.3 Æ 5.6
complexes of types 1 and 2 seems, thus, to be best with regard to tM (smaller values).
This observation may be rationalized by means of more-pronounced steric hindrance in
the coordination spheres of 1 and 2 as compared to the novel compounds 3 and 4
(Table 1).
2.2. 1 H-NMRD Profiles. The NMRD (nuclear magnetic-relaxation dispersion)
profiles of the Gd complexes 1 4, shown in Fig. 3, were rather similar, and were in
agreement with an increased relaxivity of the Bn-substituted derivatives at both low
and high fields as compared to the parent complex 5. The relaxivities of (R,S)-4 were
slightly lower than those of (S)-1 (S)-3, especially at low field, but still higher than for
GdÀDTPA (5).
The theoretical adjustment of the NMRD profiles [3] takes into account both outer-
and inner-sphere contributions. By analogy to other GdÀDTPA derivatives, the

1082
Helvetica Chimica Acta Vol. 87(2004)
Fig. 2. Logarithmic 17O-NMR reduced tranverse-relaxation rates (ln (1/Tr2 ), in sÀ1 (vertical axes); with 1/Tr2 ˆ
55.55/(Tp2 ¥ [Gd complex])) as a function of reciprocal temperature (1/T; in K; horizontal axes) for aqueous
solutions of the Gd complexes a) (S)-1, b) (S)-2, c) (S)-3, and d) (R,S)-4. The calculated lines lacking any data
points represent the parent compound GdÀDTPA (5).
number of H2O molecules in the first coordination sphere q was fixed to 1, the distance
d of closest approach for the outer-sphere contribution was set to 0.36 nm, the relative
translational correlation time was fixed to 3.3 Â 10À9 m2 sÀ1, and tM was fixed to the
value determined by the 17O-NMR experiments. The electronic relaxation time at very
low magnetic field is given by tSO ˆ 1/(5B ¥ tV). The resulting parameters are
summarized in Table 2.
The increased relaxivity of complexes (S)-1 and (S)-2 has been reported [3] to arise
mainly from both a slightly lower mobility and a reduced distance of interaction (r)
between the H-atom of the coordinated H2O molecule and the Gd-ion. Indeed, for the
above two complexes, a distance r of 0.31 nm resulted in fitted values of tR (84 and

Helvetica Chimica Acta Vol. 87(2004)
1083
Fig. 3. 1H-NMRD (Nuclear magnetic-relaxation dispersion) relaxivity profiles of the Gd complexes 1 5 in H2O.
The lines have been obtained by theoretical fitting of the data points.
Table 2. Proton Longitudinal Relaxivity (r1; in sÀ1 mmÀ1; measured at 20 MHz/0.47T) for the Gd Complexes
1
5. Also given are tM, tR, tSO, tV, and r values (see text) obtained by theoretical fitting of experimental
1H-NMRD profiles.
(S)-1
(S)-2
(S)-3
(R,S)-4
5
r1310 [sÀ1 l mmolÀ1
t3M10 [ns]a)
]
4.8
86
5.1
87122
4.6
143
4.3
143
3.9
r [nm]
0.291 0.30
61 72b)
77 91b)
14 17b)
0.296 0.30
0.30
68
90
0.31
7
87
16
0.31
c
t3R10 [ps]
68
7
)
2
2
59
t
[ps]
86 93c)
82
310
SO
c
t3V10 [ps]
27
)
35 19
23
a) Derived by 17O-NMR spectroscopy. b) The lower t value refers to a distance r of 0.291 nm, the higher one to
0.300 nm. c) The lower t value refers to a distance r of 0.296 nm, the higher one to 0.300 nm.
87ps, resp.) significantly larger than those obtained by 2H-relaxometry of the
corresponding labeled diamagnetic lanthanum complexes (t3R10 ˆ 65 Æ 7ps for La
analogs of (S)-1 and (S)-2). Fittings were, thus, performed with shorter distances r to
account for these tR values. Values of r of 0.291 0.300 nm (for (S)-1) and 0.296
2
0.300 nm (for (S)-2) agreed with the tR values obtained by H-relaxometry. Similarly,
for complex (S)-3, a value of r equal to 0.300 nm and a tR value of 68 ps were in
agreement with the observed relaxivity enhancement. According to the recent report of
Caravan et al. [15], r should be equal to 0.31 Æ 0.01 nm. However, our values of

1084
Helvetica Chimica Acta Vol. 87(2004)
0.300 nm seem to be more adequate. Interestingly, for complex (R,S)-4, a reduction of r
was not observed.
2.3. Transmetallation Stability. Transmetallation of Gd complexes with Zn2‡ in
phosphate-buffer solution (pH 7) has been shown to result in a decrease of the
paramagnetic relaxation rate as a consequence of the precipitation of gadolinium
phosphate [16]. This phenomenon was not observed for macrocyclic complexes, but
was clearly encountered with GdÀDTPA (5) and its bisamide derivatives. Trans-
metallation assessment of the Gd complexes 1 4 showed that all Bn-substituted
compounds, except for (S)-2, were more stable than 5 (Fig. 4). The observed initial
increase in the relaxation rate was most likely due either to a slight decrease of the
temperature during the mixing of the Gd complexes with Zn solutions, or to solvation
phenomena such as increase in the solvent accessibility of the Gd ions during
transmetallation. Because a −free× Gd-ion is surrounded by eight H2O molecules, its
relaxivity is much higher relative to that of a chelated Gd-ion, which has only one H2O
molecule in its first coordination sphere. Consequently, Rp1 is increased. A few minutes
later, the production of insoluble GdPO4 induces a decrease of Rp1. With regard to
transmetallation, the highest stabilities were achieved, in decreasing order, for
compounds with Bn substituents in the 5-, 4-, and 2-positions (i.e., complexes 3, 1,
and 4, resp.). Surprisingly, substitution in 7-position (complex 2) had a contradictory
effect on the transmetallation stability with respect to Zn2‡ (Fig. 4).
Fig. 4. Transmetallation stability towards Zn2‡ of the Gd complexes 1 5 in aqueous solution. The trans-
metallation properties are expressed by evolution of the ratio X ˆ Rp1(t)/R1p(t0) as a function of time t (t0 ˆ 0:
addition of Zn2‡). High values for X mean high transmetallation stability.

Helvetica Chimica Acta Vol. 87(2004)
1085
Conclusions. We have undertaken a comparative study of four possible Bn-
substituted GdÀDTPA complexes with respect to both relaxivity and transmetallation
stability. GdÀDTPA Complexes of types 1 and 2, substituted on the central arm, i.e., in
4- and 7-position (trivial atom-numbering; see formulae) exhibit a faster exchange of
H2O molecules than those substituted in other positions. This finding is relevant for the
design of novel NMR contrast agents upon coupling substituted GdÀDTPA complexes
to macromolecules or specific vectors. Also, synthetically, coupling in 4-position is
easier.
Our study indicates that the steric hindrance of the first coordination sphere
depends on substituent position. Depending on its proximity to the coordinated H2O,
substituents could accelerate solvent expulsion. Substituents close to the first
coordination shell are known to be capable of modifying the number of H2O molecules
coordinated to lanthanide complexes [17][18]. Therefore, it is not unlikely that
substituents may also affect the residence time of H2O in the first coordination sphere.
It has been reported that tM is reduced for various derivatives of GdÀDTPA complexes
[11][12][19], and that amide derivatives of GdÀDTPA or GdÀDOTA are charac-
terized by lower water-exchange rates than their unsubstituted congeners [20 23].
It should be mentioned that the differential effects on the water-exchange rates of
substituted GdÀDTPA complexes could also reflect differential exchange rates of the
single stereoisomers. Some recent work has, however, ruled out a major influence of the
co-existence of stereoisomers in the case of bisamide derivatives of GdÀDTPA
complexes [24].
In conclusion, given that the synthesis of 5-Bn-substituted compounds of type 3 is
more tedious and that the relaxivities and transmetallation properties of 3 and 1 are not
too different, the 4-position (i.e., complexes of type 1) seems to be optimal to link these
paramagnetic ligands to a macromolecule or to a tailored vector to produce potent
contrast agents of both higher relaxivity and selectivity.
Experimental Part
1. General. All chemicals were purchased from Aldrich (Bornem, Belgium) and were used without further
purification. Products were purified by anion- (Dowex 1X8-400) or cation- (Dowex 50X8-400) exchange
1
chromatography. H- and 13C-NMR spectra: Bruker AMX-300 instrument (300 and 75 MHz, resp.); chemical
shifts dH and dC in ppm rel. to SiMe4 (ˆ0 ppm) and t-BuOH (Me resonance at 31.2 ppm), resp.; multiplicities
stated as s (singlet), d (doublet), t (triplet), q (quartet), and m (multiplet). Electrospray-ionization (ESI) mass
spectra were obtained on a Q-tof-2 mass spectrometer (Micromass, Manchester, UK), using samples dissolved
in MeOH/H2O.
2. 17O-NMR Experiments. Spectra were recorded on a Bruker AMX-300 spectrometer, with 2-ml samples
placed in 10-mm tubes (o.d.). The temperature was regulated by air or N2 flow controlled by a BVT-2000 unit.
17O-NMR Transverse-relaxation times T2 of distilled H2O (pH 6.5 7.0) were measured by means of
CarrÀPurcellÀMeiboomÀGill sequence, and the data were subsequently processed by a two-parameter fit.
The 17O-NMR T2 values of coordination H2O of the complexes were obtained from line-width measurements.
All 17O-NMR spectra were proton-decoupled. The sample conc. was below 40 mm, i.e., (S)-1, 38.7; (S)-2, 20.75;
(S)-3, 20.84; and (R,S)-4, 19.07m m.
3. 1H-NMRD Profiles. Samples of 0.6 ml in 10-mm tubes (o.d.) were inserted into a Field Cycling
Relaxometer (Field Cycling Systems, Honesdale, PA, USA). Magnetic fields of 0.24 mT and 1.2 Twere applied,
which correspond to proton Larmor frequencies of 0.01 and 50 MHz, resp. Additional relaxation rates at 1.4 and
7.05 Twere determined on aMinispec Mq-60 and a Bruker AMX-300 spectrometer, resp. Relaxation rates were
also measured at 0.47T on a Minispec PC-20 apparatus (Bruker, Karlsruhe, Germany), thermostabilized by

1086
Helvetica Chimica Acta Vol. 87(2004)
means of tetrachloroethylene flow. Fitting of the 1H-NMRD profiles was performed with data-processing
software that uses different theoretical models for nuclear-relaxation phenomena (Minuit, CERN Library)
[25][26].
4. Transmetallation. Transmetallation with Zn2‡ was evaluated by the decrease of the longitudinal
relaxation rate of H2O at 310 K and 20 MHz of buffered phosphate solutions (pH 7; [KH2PO4] ˆ 26 mm,
[Na2HPO4] ˆ 41 mm), containing equimolar amounts (2.5 mm) of one of the GdÀDTPA complexes (1 5) and
of Zn2‡ [16].
5. Syntheses. 5.1. Synthesis of Ligand 11 for the Gd Complex (S)-3 (see Scheme 1). (S)-2-Amino-3-
phenylpropanamide (l-Phe-NH2; 7). l-Phenylalanine methyl ester hydrochloride (25.34 g, 177.5 mmol) was
dissolved in anh. MeOH (150 ml), and the soln. was mixed with an equimolar amount of Et3N (16 ml). Then,
Et2O (150 ml) was added under stirring to precipitate the ammonium salt. The soln. was maintained at À108 for
2 h, and the salt was filtered off. The filtrate was evaporated under reduced pressure, and the remaining yellow
oil was dissolved in anh. MeOH (300 ml). NH3 Gas was bubbled under stirring into this soln. for 2 h, and the
mixture was stirred for 2 d. The solvent was evaporated, the pale yellow residue was dissolved in H2O (100 ml),
the resulting soln. was adjusted to pH 11, stirred overnight at 508, and concentrated to a minimum. The sample
was purified by anion-exchange chromatography (Dowex AG 1X8-400, HCOONa; column: 20 Â 2.4 cm). The
column was washed with H2O. At pH ca. 6, compound 7 was eluted. The soln. was treated several times with H2O
(20 ml) and evaporated to eliminate formic acid: 95% of 7. 1H-NMR (D2O, 298 K): 2.75 3.05 (m, CH2); 3.65
(dd, CH); 7 .25 (m, Ph).
2-{[(tert-Butoxy)carbonyl]amino}acetic Acid (N-Boc-Gly). Et3N (30 ml) was added to glycine (10 g,
133.21 mmol) in H2O (120 ml). Boc-on4) (36.15 g, 146.75 mmol) in 1,4-dioxane (80 ml) was added dropwise
under stirring and an inert atmosphere. The green soln. turned yellow upon stirring for 4 h at r.t. The mixture
was poured in H2O (250 ml) and extracted with Et2O (5 Â 50 ml). The aq. phase was adjusted to pH 2 with 3m
HCl and extracted with AcOEt (5 Â 50 ml). The combined org. phase was evaporated, and the remaining yellow
oil was dissolved in H2O (75 ml) and lyophilized to a white powder: 87% yield. 1H-NMR (CDCl3, 298 K): 10.8
(br. s, OH); 4.0 (s, CH2); 5.2 (br. s, NH); 1.5 (s, t-Bu).
tert-Butyl N-{2-[(2,5-Dioxopyrrolidin-1-yl)oxy]-2-oxoethyl}carbamate (8). N-Boc-Gly (10 g, 57.1 mmol)
and N-hydroxysuccinimide (7.66 g, 66.51 mmol) were mixed in THF (90 ml) under an inert atmosphere.
Dicyclohexylcarbodiimide (DCC; 13.21 g, 64 mmol) in THF (40 ml) was added dropwise. The mixture was
stirred at r.t. overnight. Then, AcOH (0.1 ml) was added to stop the reaction. The soln. was stirred for 1 h and
was then filtered, and the filtrate was evaporated under reduced pressure. The residue was suspended in i-PrOH
(100 ml) and stirred vigorously for 1 h. The resulting suspension was filtered, and the remaining solid was dried
with a drying pistol at 508: 68% of 8. 1H-NMR (CDCl3, 298 K): 2.7( s, 2 CH2); 1.4 (s, t-Bu); 4.9 (br. s, NH); 4.2 (s,
CH2). 13C-NMR (CDCl3, 298 K): 168.2; 167.2; 150.3; 35.2; 79.4; 28.2; 24.6.
tert-Butyl (S)-N-{2-[(2-Amino-1-benzyl-2-oxoethyl)amino]-2-oxoethyl}carbamate (6). Compound
7
(9.32 g, 62.4 mmol) and Et3N (60 ml, 330.6 mmol) were dissolved in anh. MeCN (75 ml) and anh. MeOH
(75 ml). Compound 8 (17g, 62.4 mmol), dissolved in anh. THF (200 ml), was added dropwise to the mixture
under an inert atmosphere, and the resulting soln. was stirred for 4 d at r.t. The mixture was evaporated, and the
residue was suspended in MeCN (100 ml) and stirred for 2 h. The precipitate was filtered off, and the filtrate was
evaporated. The resulting residue was dissolved in H2O (100 ml) and CHCl3 (100 ml). The aq. phase was
adjusted to pH 11 with a sat. Na2CO3 soln. The mixture was stirred with brine (50 ml), and the two phases were
separated. The aq. phase was extracted with CHCl3 (5 Â 25 ml). The combined org. phase was dried (MgSO4)
and evaporated. The remaining solid was purified by column chromatography (CC) (SiO2; CHCl3/MeOH 1 :1):
27% of 6. 1H-NMR (CDCl3, 298 K): 6.0 (br. s, NH2); 5.4 (br. s, NH); 6.5 (br. s, NH); 1.4 (s, t-Bu); 3.4 (s, CH2);
4.25 (t, CH); 2.5 2.7( dd, CH2); 7.0 (m, Ph). 13C-NMR (CDCl3, 298 K): 136.2; 129.1; 128.7; 127.2; 175.4; 166.7;
156.3; 79.4; 28.8; 54.8; 37.1; 40.2.
(S)-N2-(2-Aminoethyl)-3-phenylpropane-1,2-diamine Tris(hydrochloride) (9). Compound
6 (4 g,
14.42 mmol) was dissolved in trifluoroacetic acid (TFA; 50 ml) at 08, and then stirred for 5 h at r.t. The
mixture was repeatedly evaporated, and the residue was redissolved in H2O (50 ml) to eliminate TFA. The
residue was dissolved in H2O (50 ml), and the soln. was adjusted to pH 10 with aq. Na2CO3 soln. This soln. was
evaporated, and the residue was dissolved in MeOH. The soln. was filtered and evaporated to afford the
corresponding Boc-deprotected intermediate (ˆ2-[(2-aminoacetyl)amino]-3-phenylpropanamide; not shown
4
)
−2-(tert-Butoxycarbonyloxyimino)-2-phenylacetonitrile×; (systematic name: N-[(tert-butoxycarbonyl)oxy]-
benzenecarboximidoyl cyanide.

Helvetica Chimica Acta Vol. 87(2004)
1087
in Scheme 1). This intermediate was suspended in anh. THF (100 ml), and a BH3 soln. (1m in THF, 150 ml) was
added dropwise at À108 under an inert atmosphere. The mixture was stirred for 1 h at this temp. and was then
allowed to warm to r.t. The soln. was refluxed overnight. After cooling to 58, anh. MeOH (50 ml) was slowly
added (to quench excess BH3). The soln. was evaporated under reduced pressure, and the residue was again
quenched with MeOH (50 ml, as above). The solvent was evaporated, and the remaining oil, dissolved in anh.
EtOH (100 ml), was filtered. The filtrate was cooled in a bath of ice, saturated with HCl, and stirred for 2 h. The
precipitating hydrochloride 9 was collected and dried with a drying pistol at 508: 40% of 9. 1H-NMR (D2O,
298 K): 2.9 4.0 (m, CH, 4 CH2); 7.1 7.3 (m, Ph). 13C-NMR (D2O, 298 K): 137.7; 129.8; 129.1; 127.1; 56.6; 43.3;
42.6; 40.0; 36.0.
tert-Butyl (S)-8-Benzyl-6,9,12-tris[2-(tert-butoxy)-2-oxoethyl]-2,2-dimethyl-4-oxo-3-oxa-6,9,12-triazatetra-
decan-14-oate (10). Compound 9 (1.39 g, 4.58 mmol) was dissolved in H¸nig×s base ((i-Pr)2NEt; 7.98 ml,
45.81 mmol) and anh. DMF (50 ml). tert-Butyl 2-bromoacetate (5.17ml, 32.06 mmol), dissolved in anh. DMF
(30 ml), was added dropwise at 58 under an inert atmosphere. The mixture was stirred at this temp. for 1 h, then
at 508 for 48 h. The filtrate was evaporated under reduced pressure, and the residual oil was dissolved in AcOEt
(50 ml) and H2O (50 ml). The aq. phase was extracted with AcOEt (3 Â 50 ml). The combined org. phase was
extracted with H2O (50 ml) and sat. aq. NaHCO3 soln. (50 ml), dried (MgSO4), and evaporated. The residual oil
was purified by CC (SiO2; Et2O): 46% of 10. 1H-NMR (CDCl3, 298 K): 7.2 (m, Ph); 1.5 (s, 5 t-Bu); 2.5 3.0 (m,
4 CH2); 3.2 3.6 (m, 5 CH2); 3.7( m, CH). 13C-NMR (CDCl3, 315 K): 28.4; 36.4; 66.6; 50.3; 53.7; 55.2; 56.0; 56.5;
‡
‡
62.1; 80.5; 81.1; 81.5; 139.5; 129.3; 128.3; 125.9; 172.9; 171.0; 170.8. ESI-MS: 765 ([M ‡ H] ), 787 ([M ‡ Na] ).
(S)-2-({1-Benzyl-2-[bis(carboxymethyl)amino]ethyl}{2-[bis(carboxymethyl)amino]ethyl}amino)acetic
Acid (11). Compound 10 (1.6 g, 2.09 mmol) was treated with conc. HCl (75 ml), and the soln. was stirred for 24 h
at r.t. The mixture was filtered and evaporated. The residue was dissolved in a minimum of H2O, adjusted to
pH 10 with a sat. aq. NaHCO3 soln., and extracted with AcOEt (5 Â 30 ml). The aq. phase was adjusted to pH 2
‡
with 3m HCl, concentrated, and purified by cation-exchange chromatography (Dowex AG 50W-X8, H -form;
column: 15 Â 2.4 cm). The column was washed with H2O until reaching a pH of ca. 6. The product was then
eluted with 2m aq. NH3 soln. (500 ml). The soln. was evaporated, and the residue was dissolved in a minimum of
1
MeOH. The soln. was filtered, and the product was precipitated with cold Et2O: 70% of 11. H-NMR (D2O,
298 K, pH > 7): 7.3 (m, Ph); 2.3 3.5 (m, 9 CH2, CH). 13C-NMR (D2O, 320 K, pH > 7): 180.6; 180.3; 179.4; 65.3;
‡
61.8; 60.5; 59.3; 54.3; 53.3; 52.6; 140.3; 129.7; 129.2; 126.7; 32.3. ESI-MS: 550 ([M ‡ 3Na] ).
5.2. Synthesis of the Ligand 12 for the Gd Complex (R,S)-4 (see Scheme 2,b). Methyl 5,8,11-Tris(2-methoxy-
2-oxoethyl)-3-oxo-2-oxa-5,8,11-triazatridecan-13-oate (15). To DTPA5) (5 g, 12.7mmol), suspended in anh.
MeOH (200 ml), was added at 08 under inert gas SOCl2 (10 ml, 140 mmol) under stirring. The clear soln. was
stirred overnight, and then refluxed for 4 h. The residual solvent was evaporated under reduced pressure. The
residue was dissolved in AcOEt (100 ml), and a sat. aq. NaHCO3 soln. was slowly added under stirring until a
pH of ca. 7was reached. The soln. was filtered, the phases were separated, and the aq. phase was extracted with
AcOEt (5 Â 30 ml). The combined org. phase was dried (MgSO4), filtered, evaporated, and dried to afford 15
(95%). Yellow oil. 1H-NMR (CDCl3, 298 K): 2.75 (t, 2 CH2); 2.85 (t, 2 CH2); 3.7( s, 4 Me); 3.65 (s, Me); 3.6 (s,
4 CH2); 3.5 (s, CH2). 13C-NMR (CDCl3, 298 K): 171.5; 171.8; 54.8; 54.6; 52.4; 51.8; 51.2; 51.0. ESI-MS: 464
‡
‡
([M ‡ H] ), 486 ([M ‡ Na] ).
Methyl (R,S)-4-Benzyl-5,8,11-tris(2-methoxy-2-oxoethyl)-3-oxo-2-oxa-5,8,11-triazatridecan-13-oate (16).
To a soln. of KHMDS6) (8.64 ml, 4.32 mmol) in toluene, diluted with anh. THF (10 ml), was added 15 (1 g,
2.16 mmol) in anh. THFat À788 under an inert atmosphere over a period of 20 min. The yellow soln. was stirred
at this temp. for an additional 20 min. Then, a soln. of BnI (0.77 ml, 6.48 mmol) [27] in a mixture of anh. THF
(12 ml) and HMPA (0.108 ml) was added dropwise at À788 over a period of 30 min, and this soln. was stirred for
an additional 30 min. Then, the cooling bath was removed, and the mixture was stirred for 1 h at r.t., then
overnight at 508. The solvents were evaporated, and the residue was dissolved in AcOEt (50 ml) and sat. aq.
NaHCO3 soln. (50 ml). The aq. phase was extracted with AcOEt (3 Â 15 ml), and the combined org. phase was
dried (MgSO4) and evaporated. The remaining yellow oil was prepurified by CC (SiO2; AcOEt). The crude
product fractions were combined, evaporated, and purified by CC (SiO2; AcOEt/hexane gradient): 32% of 16.
1H-NMR (CDCl3, 298 K): 7.2 (m, Ph); 2.65 2.85 (m, 4 CH2); 2.95 (m, CH2); 3.4 (s, CH2); 3.5 3.7( m, 5 Me,
3 CH2, CH). 13C-NMR (CDCl3, 310 K): 172.7; 172.3; 172.1; 171.8; 138.1; 129.4; 128.4; 126.6; 66.3; 36.7; 55.2;
‡
‡
55.0; 53.2; 52.8; 52.7; 52.4; 52.3; 52.1; 51.6; 51.4; 50.6. ESI-MS: 554 ([M ‡ H] ); 576 ([M ‡ Na] ).
5
)
)
Systematic name: 2-(bis{2-[bis(carboxymethyl)amino]ethyl}amino)acetic acid.
Potassium hexamethyldisilazide.
6

1088
Helvetica Chimica Acta Vol. 87(2004)
(R,S)-2-[{2-[{2-[Bis(carboxymethyl)amino]ethyl}(carboxymethyl)amino]ethyl}(carboxymethyl)amino]-3-
phenylpropanoic Acid (12). Compound 16 (1.09 g, 1.97mmol) was treated with conc. HCl (45 ml). The soln. was
stirred for 24 h at r.t., filtered, and evaporated. The residue was dissolved in a minimum of H2O, and the pH was
adjusted to 10 with a sat. aq. NaHCO3 soln. The aq. phase was extracted with AcOEt (5 Â 30 ml), concentrated,
and adjusted to pH 2 with 3m HCl. The product was purified by ion-exchange chromatography (Dowex AG
‡
50W-X8, H -form; column: 15 Â 2.4 cm). The column was washed with H2O until reaching a pH of ca. 6. Then,
the product was eluted with 2m aq. NH3 soln. (500 ml). The product fractions were evaporated, and the
remainder was dissolved in a minimum of MeOH. The MeOH soln. was filtered and poured in cold ether,
whereupon 12 (86%) precipitated. 1H-NMR (CDCl3, 298 K): 6.8 (m, Ph); 2.4 3.9 (m, 9 CH2, CH). 13C-NMR
(CDCl3, 310 K): 174.7; 174.2; 171.6; 170.6; 138.1; 131.7; 130.5; 126.7; 62.9; 57.7; 56.8; 56.6; 55.7; 54.6; 52.9; 52.7;
‡
34.4. ESI-MS: 560 ([M ‡ 2K] ).
5.3. Gadolinium-Complex Formation. The ligand (11 or 12; 0.5 mmol), dissolved in H2O (2 ml), was treated
dropwise with an equimolar aq. soln. of GdCl3 ¥ 6 H2O, keeping the pH between 5.5 and 6.5 by means of NaOH.
The stirred soln. was heated at 608 for 48 h. The medium was then adjusted to pH 9 with aq. NaOH soln.
(precipitation of free Gd3‡ as Gd(OH)3), and the resulting precipitate was filtered off. The filtrate was adjusted
to pH 7, and then treated withChelex resin for 2 h (elimination of residual free Gd3‡, as assessed by means of the
xylenol-orange test). The final concentration of the Gd-complexes (compounds 1 5) were determined on
samples treated with HNO3 by comparing their proton paramagnetic-relaxation rates at 20 MHz with those of
reference samples containing known amounts of Gd3‡
.
We thank Mrs. Patricia de Francisco for her help in preparing this manuscript. This work was supported by
the ARC Program 00/05-258 of the French Community of Belgium. F. B. thanks the FRIA for financial support.
The support and sponsorship provided by the COST Action D18 (−Lanthanide Chemistry for Diagnosis and
Therapy×) are kindly acknowledged.
REFERENCES
[1] F. A. Dunand, E. Toth, R. Hollister, A. E. Merbach, J. Biol. Inorg. Chem. 2001, 6, 247.
[2] S. Aime, E. Gianolio, E. Terreno, G. B. Giovenzana, R. Pagliarin, M. Sisti, G. Palmisano, M. Botta, M. P.
Lowe, D. Parker, J. Biol. Inorg. Chem. 2000, 5, 488.
¬
¬
[3] S. Laurent, L. Vander Elst, S. Houze, N. Guerit, R. N. Muller, Helv. Chim. Acta 2000, 83, 394.
[4] M. A. Williams, H. Rapoport, J. Org. Chem. 1993, 58, 1151.
[5] M. W. Brechbiel, O. A. Gansow, Bioconjugate Chem. 1991, 2, 187.
[6] O. Renn, C. F. Meares, Bioconjugate Chem. 1992, 3, 563.
[7] M. S. Ali, S. M. Quadri, Bioconjugate Chem. 1996, 7, 576.
[8] D. A. Westerberg, P. L. Carney, P. E. Rogers, S. J. Kline, D. K. Johnson, J. Med. Chem. 1989, 32, 236.
[9] J. F. W. Keana, J. S. Mann, J. Org. Chem. 1990, 55, 2868.
[10] H. Nemoto, J. Cai, H. Nakamura, M. Fujiwara, Y. Yamamoto, J. Org. Chem. 1999, 81, 170.
¬
[11] R. N. Muller, B. Raduchel, S. Laurent, J. Platzek, C. Pierart, P. Mareski, L. Vander Elst, Eur. J. Inorg. Chem.
1999, 1949.
[12] L. Vander Elst, F. Maton, S. Laurent, F. Seghi, F. Chapelle, R. N. Muller, Magn. Reson. Med. 1997, 38, 604.
[13] F. Uggeri, S. Aime, P. L. Anelli, M. Botta, M. Brocchetta, C. de Haen, G. Ermondi, M. Grandi, P. Paoli,
Inorg. Chem. 1995, 34, 633.
[14] S. Aime, S. G. Crich, E. Gianolio, E. Terreno, A. Beltrami, F. Uggeri, Eur. J. Inorg. Chem. 1998, 1283.
[15] P. Caravan, A. V. Astashkin, A. M. Raitsimring, Inorg. Chem. 2003, 42, 3972.
[16] S. Laurent, L. Vander Elst, F. Copoix, R. N. Muller, Invest. Radiol. 2001, 36, 115.
[17] R. A. Moats, S. E. Fraser, T. J. Meade, Angew. Chem., Int. Ed. 1997, 36, 726.
[18] M. P. Lowe, D. Parker, O. Reany, S. Aime, M. Botta, G. Castellano, E. Gianolio, R. Pagliarin, J. Am. Chem.
Soc. 2001, 123, 7601.
[19] E. Toth, F. Connac, L. Helm, K. Adzamli, A. E. Merbach, J. Biol. Inorg. Chem. 1998, 3, 606.
[20] S. Aime, M. Botta, M. Fasano, S. Paoletti, L. Anelli, F. Uggeri, M. Virtuani, Inorg. Chem. 1994, 33, 4707.
[21] E. Toth, D. Pubanz, S. Vauthey, L. Helm, A. E. Merbach, Chem. Eur. J. 1996, 2, 1607.
[22] S. Aime, A. Barge, J. I. Bruce, M. Botta, J. A. K. Howard, J. M. Moloney, D. Parker, A. S. de Sousa, M.
Woods, J. Am. Chem. Soc. 1999, 121, 5762.
[23] F. Botteman, G. Nicolle, L. Vander Elst, S. Laurent, A. E. Merbach, R. N. Muller, Eur. J. Inorg. Chem. 2002,
2686.

Helvetica Chimica Acta Vol. 87(2004)
1089
[24] U. Cosentino, D. Pitea, G. Moro, V. Barone, A. Villa, R. N. Muller, F. Botteman, Theor. Chem. Acc. 2004,
111, 204.
[25] R. N. Muller, D. Declercq, P. Vallet, F. Giberto, B. Daminet, H. W. Fischer, F. Maton, Y. Van Haverbeke,
−Proc. ESMRMB, 7th Annual Congress×, Strasbourg, 1990, p. 394.
[26] P. Vallet, Ph. D. Thesis, University of Mons-Hainaut, Belgium, 1992.
[27] A. I. Vogel, −Practical Organic Chemistry×, 3rd edn., Longmans, 1966, p. 538
Received October 27, 2003