[Gd(AAZTA)]? Derivatives with n-Alkyl Acid Side Chains Show Improved Properties for Their Application as MRI Contrast Agents**

Article abstract of DOI:10.1002/chem.202004479

Herein, the synthesis and an extensive characterization of two novel Gd(AAZTA) (AAZTA=6-amino-6-methylperhydro-1,4-diazepine tetra acetic acid) derivatives functionalized with short (C₂ and C₄) n-alkyl acid functions are reported. Th

Full text of DOI:10.1002/chem.202004479

Accepted Article
Accepted Article
Title: [Gd(AAZTA)]- derivatives with n-alkyl acid side chains show
improved properties for their application as MRI contrast agents
Authors: Zsolt Baranyai, Flávio Vinicius Crizóstomo Kock, Attila
Forgács, Nicol Guidolin, Rachele Stefania, Adrienn Vágner,
Eliana Gianolio, and Silvio Aime
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202004479
Link to VoR: https://doi.org/10.1002/chem.202004479
01/2020

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
[Gd(AAZTA)]- derivatives with n-alkyl acid side chains show
improved properties for their application as MRI contrast agents
Flávio Vinicius Crizóstomo Kock,[a,b] Attila Forgács,[c,d] Nicol Guidolin,[e] Rachele Stefania,[b] Adrienn
Vágner,[c] Eliana Gianolio,*[b] Silvio Aime[b] and Zsolt Baranyai*[e]
Abstract: Herein the synthesis and an extensive characterization of
two novel Gd(AAZTA) derivatives functionalized with short (C2 and
C4) n-alkyl acid function are reported. The carboxylate functionality
is the site for further conjugations for the design of more specific
contrast agents (CAs). Interestingly it has been found that the
synthetized complexes display enhanced properties for their use as
MRI contrast agents by their own. The stability constants determined
using potentiometric titration and UV-Vis spectrophotometry were
slightly higher than the one reported for the parent Gd(AAZTA)
complex. This observation might be accounted in terms of the larger
sigma-electron donation of the acyl substituents in respect to the one
provided by the methyl group in the parent complex. As far as
concerns the kinetic stability, transmetallation experiments with
endogenous ions (e.g. Cu2+, Zn2+ and Ca2+) led to infer that the Gd3+
ions present in these Gd(AAZTA) derivatives show somewhat
smaller susceptibility to chemical exchange towards these ions at
25C, close to the physiological condition. The 1H-NMR spectra of
the complexes with EuIII and YbIII displayed a set of signals
consistent with half the number of methylene protons present on
each ligand. The number of resonances resulted invariant over a
large range of temperature to suggest the occurrence of a fast
interconversion between structural isomers. The relaxivity values
(298K, 20MHz) were consistent with q=2 being equal to 8.8 mM-1.s-1
for the C2 derivative and 9.4 mM-1s-1 for the C4 one, i.e sensibly
larger than the one reported for Gd(AAZTA) (7.1 mM-1.s-1). VT-T2
17O-NMR measurements showed, for both complexes, the presence
of two populations of coordinated water molecules, one in fast and
one in slow exchange with the bulk water. Since the high resolution
1H-NMR spectra of the analogues with Eu(III) and Yb(III) did not
show the occurrence of distinct isomers (as frequently observed in
other macrocyclic Lanthanide(III)-containing complexes), we
surmised the presence of two fast-interconverting isomers in solution.
The analysis of the 17O NMR VT-T2 profiles vs. temperature allowed
to establish their relative molar fraction being 35% for the isomer
with the fast exchanging water and 65% for the isomer with the
water molecules in slower exchange. Finally, 1H NMRD profiles over
an extended range of applied magnetic field strengths have been
satisfactory fitted on the basis of the occurrence of the two
interconverting species.
Introduction
The design of novel contrast agents (CAs) for magnetic
resonance imaging (MRI) is still a task of strong interest in the
search of safe, rapid and non-invasive clinical diagnostics.[1] Up
to now these CAs are represented by paramagnetic metal
complexes characterized by a high longitudinal paramagnetic
relaxivity (r1p). Several paramagnetic ions, among them Mn2+,
Mn3+, Fe3+, and Gd3+ can exerts a strong paramagnetic effect on
the water proton relaxation being Gd3+ ions the candidate of
choice due to the high paramagnetism associated to its seven
unpaired electrons and the relatively long electron relaxation
time.[2–4]
It was immediately evident, in the early days of their suggestion
as MRI-CAs, that free Gd3+ ions could not be administered to
living systems because of their antagonistic role in respect to the
endogenous Ca2+ and Zn2+ ions .[3] Therefore, the complexation
of Gd3+ ions with ligands to yield highly stable chelates has been
a fundamental step in the search of new contrast agents.
Nowadays, the clinically used Gd3+-based contrast agents
(GBCAs)[5] own only one water molecule coordinated (q=1) to
the paramagnetic ion,[6] in spite of the fact that the relaxivity
scales up with the number of coordinated water molecules. This
choice relies on the early observation that the coordination with
octa-dentate ligands yields highly stable complexes which
ensure for a limited in vivo release of Gd3+ ions. However, the
search for systems containing two or three coordinated water
molecules still endowed with a stability compatible with their in
vivo use, is an area of intense activity.[7,8] In this context, the
AAZTA ligand (6-amino-6-methylperhydro-1,4-diazepine tetra
acetic acid) (and its derivatives) has been considered as an
excellent system for the design of q=2 GBCAs (Scheme 1). In
[a]
[b]
F. V. C. Kock, São Carlos Institute of Chemistry, University of São
Paulo, Avenida Trabalhador São Carlense 400, 13566-590, São
Paulo, Brazil
F. V. C. Kock, Dr. E. Gianolio, R. Stefania, Prof. S. Aime, Dep. of
Molecular Biotechnologies and Health Science, University of Turin,
Via Nizza 52, 10125, Turin, Italy.
fact high relaxivities were reported for AAZTA chelates forming
[9]
E-mail: eliana.gianolio@unito.it
lipid-based aggregates
or able to bind to human serum
[c]
[d]
[e]
Dr. A. Forgács, Dr. A. Vágner, Dep. of Inorganic and Analytical
Chemistry, University of Debrecen, H-4010, Debrecen, Egyetem tér
1., Hungary
Dr. A. Forgács, MTA-DE Redox and Homogeneous Catalytic
Reaction Mechanisms Research Group, Egyetem tér 1, Debrecen,
H-4032 Hungary
N. Guidolin, Dr. Zs. Baranyai, Bracco Imaging SpA, Bracco
Research Center, Via Ribes 5, 10010, Colleretto Giacosa (TO),
Italy.
albumin (HSA)[10] . For these applications bifunctional AAZTA
derivatives were designed to embody remote free functional
groups ready for the conjugation step with the selected
moieties.[11–15] The aim of the herein reported work was to
address a detailed physico-chemical characterization of two
novel Gd(AAZTA) derivatives functionalized with short (C2 and
C4) spaced acid function (Scheme 1), designed for being used in
further conjugations with a variety of targeting vectors, lipophilic
chains or to form multimeric species.
E-mail: zsolt.baranyai@bracco.com
Supporting information for this article is given via a link at the end of
the document. Experimental details of the thermodynamic, kinetic
and relaxometric studies; VT-1H NMR spectra of Eu(III) and Yb(III)-
complexes
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
protecting groups were carried out by using 1 M lithium
hydroxide and tetrahydrofuran (THF) solution to receive the
bifunctional chelator AAZTA-C4-(tBu)4.
Scheme 1. Structure of H4AAZTA, H4AAZTA-C2-COOH and H4AAZTA-C4-
COOH ligands
Results and Discussion
The H4AAZTA-C2-COOH and H4AAZTA-C4-COOH ligands are
the derivative of H4AAZTA in which the methyl group is
substituted by n-propionic and n-valeric acid pendants (Scheme
1) for conjugation purposes.[15,16] Generally three N- and four
carboxylate O-atoms can simultaneously bind the metal ion in
the AAZTA-complexes.[17–22] The presence of n-propionate and
n-valerate pendants instead of the methyl substituent might
affect the equilibrium, kinetic, relaxation and structural properties
of the metal complexes formed with AAZTA-C2-COO- and
AAZTA-C4-COO- ligands. Indeed, the presence of the n-valeric
Scheme 3. The synthesis of H4AAZTA-C2-COOH
The synthesis route of the ligand H4AAZTA-C2-COOH (5a) is
shown in Scheme 3, it was obtained from its bifunctional
prochelating agent AAZTA-C2-(Me)4 (4a) by ester hydrolysis of
all protective groups with lithium hydroxide. Compound 4a was
successfully obtained over 5-steps, starting from the synthesis of
hydroxymethyl diazepine (1a) by double nitro-Mannich reaction
between N,N′-dibenzylethylenediamine, paraformaldehyde and
2-nitroethanol. Then, compound 1a was reacted with tert-butyl
acrylate and t-BuOK in THF, via tandem retro-Henry and
Michael reactions,[16] to afford compound 2a in good yield.
Hydrogenolysis using palladium on carbon and H2 and
subsequent alkylation with methyl bromoacetate yielded the
compound 3a. Deprotection of the tert-butyl ester with TFA
yielded the bifunctional agent derivative 4a.
acid pendant used for the conjugation of DATA ligand (DATA5m
)
to biologically active molecules evidently influence the physico-
chemical properties of the GaIII-complexes.[13–15] In order to get
more insights into the role the conjugating moieties may have on
the physico-chemical properties, the behaviour of the metal
complexes formed with the AAZTA-C2-COO- and AAZTA-C4-
COO- has been investigated close to physiological condition
(0.15 M NaCl, 25C).
Synthesis of ligands and their LnIII-complexes (Ln = Gd3+
and Eu 3+
)
The LnIII-complexes of AAZTA-C4-COO- and AAZTA-C2-COO-
were obtained by the addition of stoichiometric amounts of EuCl3
or GdCl3 to the ligand solution at pH 6.5. The formation of LnIII-
complexes has confirmed by mass spectrometry.
Solution equilibria of the AAZTA-C2-COO- and AAZTA-C4-
COO- ligands and its complexes
Protonation equilibria of the AAZTA-C2-COO- and AAZTA-C4-
COO- ligands: The protonation constants of the ligand, defined
by Eq. (1) were determined by pH-potentiometry and the values
are listed in Table 1. (standard deviations are shown in
parentheses). The charges of the ligands and complexes will be
indicated only when they are deemed really necessary.
Scheme 2. The synthesis of H4AAZTA-C4-COOH
[HiL]
KiH 
The synthesis route of the ligand H4AAZTA-C4-COOH (5) is
shown in Scheme 2. It was obtained from its bifunctional
prochelating agent AAZTA-C4-(tBu)4 (4) after cleavage of all
protective groups with trifluoroacetic acid (TFA). Compound 4
(1)
[Hi1L][H ]
where i=1, 2, …7. The protonation sequence of the AAZTA
ligand was already fully characterized by both spectroscopic and
potentiometric methods.[18] The first protonation process was
shown to occur at the nitrogen atoms of the endo- and exocyclic
nitrogen atoms of the ligand backbone (the protonation occurs
partially at all nitrogen atoms). The second protonation takes
place at the endocyclic nitrogen, whereas the first proton is
moved to the exocyclic nitrogen due to the electrostatic
repulsion between the protonated nitrogen atoms. Further
protonations occur at the ring-carboxylate groups attached to the
non-protonated endocyclic nitrogen, the non-protonated
endocyclic nitrogen and/or the carboxylate pendant arms of the
was successfully synthesized over 4-steps
following the
[12]
protocol described by Manzoni et al.
Briefly, compound 1, a
diazepane ring, was formed via a double nitro-Mannich reaction
between N,N′-dibenzylethylenediamine, paraformaldehyde and
6-nitrohexanoic acid methyl ester in nearly quantitative yield.
Hydrogenolysis using palladium on carbon and H2 removed the
benzyl protecting groups at the endocyclic amines and
simultaneously reduced the nitro group to an amine. The product
was directly reacted with tert-butyl bromoacetate to form tetra
alkylated compound 3. Deprotection of the methyl ester
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
exocyclic nitrogen atom, respectively (Scheme S1). According to
the similarities to the parent AAZTA, the observed protonation
sequence of AAZTA-C2-COO- and AAZTA-C4-COO- ligands
was very similar. A comparison of the protonation constants
obtained in the presence of the same background electrolyte
The stability and protonation constants of the metal complexes
formed with the AAZTA-C2-COO- and AAZTA-C4-COO- ligand
are defined by Eqs. (2) and (3).
[ML]
(2)
KML

H
H
[M][L]
(0.15 M NaCl, Table 1) indicated that the logK1 and logK3
values of AAZTA-C2-COO- and AAZTA-C4-COO- ligands are
slightly larger than the corresponding protonation constants of
the parent AAZTA. By taking into account the protonation
constant of n-propionic acid and n-valeric acid (propionic acid:
[MHiL]
KMH L

(3)
[MHi1L][H ]
i
H
H
logK1 =4.53, n-valeric acid: logK1 =4.69, 1.0
M
NaClO4,
where i=1, 2, …4. The protonation and stability constants of the
CaII-, ZnII- and LnIII-complexes formed with AAZTA-C2-COO-
and AAZTA-C4-COO- ligands have been calculated from the pH-
potentiometric titration curves obtained at 1:1 metal/ligand
concentration ratios. The best fitting was obtained by assuming
the formation of ML, MHL, MH2L, MH3L and MH4L species in
equilibrium. The formation of the deprotonated [M(L)H-1]n-
complexes takes place at pH>9.0, as indicated by the base
consumption in the titration curves (M= Zn2+ and Cu2+, n=2;
M=Ln3+, n=3). These processes can be interpreted by assuming
a hydrolysis of the metal ion which takes place by the
coordination of an OH- ion with the formation of [M(L)H-1]n-
species. The protonation of the [M(L)H-1]n- species is
characterized by the equilibrium constant KM(L)H-1 (Eq. (4))
25C),[23] one can assume that the third protonation of AAZTA-
C2-COO- and AAZTA-C4-COO- occurs at the carboxylate group
of the n-propionic and the n-valeric side chain. The last three
protonation constants of AAZTA-C2-COO- and AAZTA-C4-COO-
H
H
H
(logK4 , logK5 and logK6 ) characterizing the protonation of the
ring carboxylate and non-protonated exocyclic nitrogen or
carboxylate groups, are very similar and comparable with the
corresponding logKiH values (logK3 , logK4 and logK5 ) of
AAZTA ligand. The somewhat higher logK1H value of AAZTA-C2-
COO- and AAZTA-C4-COO- might be explained by the electron
donating properties of the n-propionic acid and the n-valeric side
chain, which might influence the basicity of the exo- and
H
H
H
H
endocyclic nitrogen atoms. It is important to note that the logK1
value of AAZTA and its derivative ligands determined in 0.15 M
NaCl solution is significantly smaller than the value obtained in
0.1 M KCl or 0.1 M Me4NCl solutions,[18] which might be
explained by the formation of the NaI complexes that competes
with the first protonation process.
[ML]
KM(L)H

(4)
[M(L)H1][H ]
1
The stability and protonation constants of the CaII-, ZnII, CuII-
and LnIII-complexes obtained by using pH potentiometric titration
and UV/Vis spectrophotometric techniques are presented and
compared with those of the corresponding AAZTA complexes in
Table 1. Protonation constants of AAZTA, AAZTA-C2-COOH and AAZTA-
C4-COOH ligands (25C)
Table 2. The UV/Vis spectrophotometric studies of Cu2+
-
AAZTA-C4-
COO-
AAZTA-C2-
COO-
AAZTA
AAZTA-C2-COO- and Cu2+ - AAZTA-C4-COO- systems are
summarized in ESI file. The comparison of the stability constants
reported in Table 2 indicates that CaII-, ZnII and LnIII-complexes
formed by AAZTA-C2-COO- and AAZTA-C4-COO- ligands are
are comparable and slightly higher than those of the
corresponding AAZTA complexes due to the similar basicities of
the AAZTA-C2-COO-, AAZTA-C4-COO- ligands and AAZTA
ligands. Interestingly, the stability constants for the Cu(AAZTA-
C2-COO-) and Cu(AAZTA-C4-COO-) complexes are smaller by
0.6 and 1.6 logK unit than that of the parent Cu(AAZTA).
Moreover, the stability constant of the Cu(AAZTA-C4-COO-) is
about 1 logK lower than that of the Cu(AAZTA-C2-COO-). Based
on the distorted octahedral structure of the Cu(AAZTA)2- (CuII-
ion is coordinated by the exo- and endocyclic nitrogens and
carboxylate oxygen atoms in equatorial positions, whereas the
other endocyclic nitrogen and one of the exocyclic carboxylate
oxygen atoms are in axial positions),[19] it can be reasonably
assumed that the coordination of the more basic exo- and
endocyclic nitrogen atoms results in the further distortion of the
CuII-complexes[19] which might cause the slight decrease of the
stability constant of the Cu(AAZTA-C2-COO-) and Cu(AAZTA-
C4-COO-) upon the increase of the basicity of the exo- and
endocyclic nitrogen atoms.
I
0.15 M NaCl
0.15 M NaCl[a]
0.1 M KCl[b]
11.23
6.52
3.78
2.24
1.56

H
H
H
H
H
H
H
logK1
logK2
logK3
logK4
logK5
logK6
logK7
10.48 (1)
6.90 (1)
10.22 (1)
6.53 (1)
10.06
6.50
3.77
2.33
1.51

4.68 (1)
4.33 (2)
3.73 (1)
3.62 (1)
2.60 (1)
2.91 (1)
1.80 (2)
2.03 (2)
1.09 (3)
1.23 (3)


logKiH
31.27/26.59[c]
30.86/26.53[c]
24.16
25.33
a Ref. [19] b Ref. [18] c The protonation constant of the propionic acid and the
, ,
n-valeric acid was not considered due to a negligible role in metal binding.
Complexation properties with M2+/3+ cations: The logKiH values,
presented in Table 1., clearly indicate that the total basicity of
the AAZTA-C2-COO- and AAZTA-C4-COO- is slightly higher
than of AAZTA. By taking into account the higher basicity of the
AAZTA-C2-COO- and AAZTA-C4-COO-, the stability constants
of their LnIII complexes are expected to be somewhat higher
than that of the parent AAZTA complex. However, the stability
constants of the Ln(AAZTA-C2-COO-)2- and Ln(AAZTA-C4-
COO-)2- complexes might be influenced by the flexibility of the
coordination cage by the presence of the side chains.
The stability constants of Ln(AAZTA-C2-COO-) and Ln(AAZTA-
C4-COO-) complexes like the parent Ln(AAZTA) complexes
increase from La3+ to Lu3+ (Table 2). These data clearly reflect
the effect of the flexible coordination cage formed by the three
nitrogen and four carboxylate oxygen atoms of the ligands that
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
results in an improved size match for the Ln3+ ions with the
smaller ionic radius.
carboxylate oxygen) that can be protonated in the pH range of 2
to 4.
Table 2. Stability and protonation constants of AAZTA-C4-COO-, AAZTA-C2-
COO- and AAZTA complexes formed with Ca2+, Mn2+, Zn2+ and Cu2+ ions
(0.15 M NaCl, 25C)
Relaxation properties of the Gd(AAZTA-C2-COO-) and
Gd(AAZTA-C4-COO-)
Relaxivity values (r1p) of Gd(AAZTA), Gd(AAZTA-C2-COO-) and
Gd(AAZTA-C4-COO-) as a function of pH are shown in Figure 1.
AAZTA-C4-
COO-
AAZTA-C2-
COO-
AAZTA
0.15 M NaCl[a]
0.1 M KCl[b]
I
0.15 M NaCl
CaL
CaHL
CaH2L
12.05 (2)
4.77 (2)
3.27 (2)
17.05 (3)
4.74 (3)
3.76 (2)
2.76 (2)
11.39 (4)
18.94 (5)
4.79 (2)
3.91 (2)
2.88 (2)
1.31 (2)
11.02 (3)
11.91 (1)
4.36 (1)
3.58 (3)
11.75
3.41

12.76
3.34

ZnL
16.93 (3)
4.48 (3)
3.61 (3)
2.74 (3)
11.44 (5)
19.96 (5)
4.54 (3)
3.86 (2)
2.94 (2)
1.40 (4)
11.23 (2)
16.02
3.95
2.53

18.01
3.87
2.36

ZnHL
ZnH2L
ZnH3L
ZnLH-1
CuLc
Figure 1. Relaxivity values (r1p) of Gd(AAZTA) (), Gd(AAZTA-C2-COOH)
() and Gd(AAZTA-C4-COOH) () as a function of pH ([GdL]=1.0 mM, 21
MHz, 25°C).
11.36
20.60
3.86
2.43
1.37

11.25
22.27
4.00
2.72

The relaxivity values of Gd(AAZTA-C2-COO-) and Gd(AAZTA-
C4-COO-) are 8.8 mM-1s-1 and 9.4 mM-1s-1 , respectively, at
pH=7.4, 21 MHz and 298 K. The pH dependence of relaxivity
shown in Figure 1 has similar behaviours for the two complexes.
The r1p values remain almost constant in the range of pH 2-11
while significantly increase below pH 2 and decrease above
pH=11 for both complexes. The high relaxivity values obtained in
strong acidic conditions have to be associated to acidic
hydrolysis suffered by the complexes. At low pH, these species
CuHL
CuH2L
CuH3L
CuH4L
CuLH-1

10.62
10.81
LaL
LaHL
LaH2L
17.24 (2)
4.68 (1)
2.37 (2)
16.98 (3)
4.36 (1)
2.59 (2)
16.48
1.90

17.53
1.97

are highly protonated with
a consequent release of the
paramagnetic Gd3+ ions, which are characterized by high
relaxivity (12.98 mM-1s-1 at 21MHz and 25°C). From pH 11-12,
hydrolysis of the coordinated water molecules may take place
with a consequent decrease of the relaxivity values.
GdL
GdHL
GdH2L
20.33 (2)
4.74 (1)
2.07 (1)
20.06 (5)
4.45 (2)
2.11 (3)
18.93
2.18

20.24
1.89

LuL
22.06 (3)
4.65 (2)
21.65 (4)
4.47 (4)
21.22
21.85
LuHL


a
[13]
b
[19]
c
Ref.
,
Ref.
,
Spectrophotometry, [H+]=0.01–1.0 M, I =[H+]+[Na+] =
0.15 M in samples at [H+]<0.15 M.
The CaII-, ZnII and LnIII-complexes of AAZTA-C2-COO- and
AAZTA-C4-COO-, similarly to the behaviour shown by the
analogous AAZTA complexes can be protonated at low pH
values. The logKMHL values of CaII-, ZnII-, CuII and LnIII-
complexes with AAZTA-C2-COO- and AAZTA-C4-COO- ligands
H
(Table 2) are very similar to the logK3 value of the free ligands
Figure 2. Molecular weight (MW) dependence of the relaxivity
(0.47 T, 298 K) for typical GdIII-complexes with two inner-sphere
water molecules. See ref [10] and references herein.
(Table 1). The experimental evidences clearly indicate that the
propionic acid and the n-valeric acid side chains of AAZTA-C2-
COO- and AAZTA-C4-COO- ligands do not take place in the
coordination to metal ions, as this functionality can
protonate/deprotonate independently. At lower pH values, one
and two smaller protonation constants could be determined for
the ZnII- and CuII-complexes, which might be explained by the
presence of one or two weakly coordinated donor atom (a
When comparing the relaxivity values of the herein investigated
GdIII-complexes with those of previously reported q=2 GdIII-
complexes it comes evident that they are higher than what
expected in the plot of relaxivity as a function of molecular
weight (Figure 2). This enhancement in relaxivity can, in
principle, be ascribed to either
a lengthened molecular
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
reorientation due to some kind of aggregation and/or
a
water molecules on the same molecule or to the presence of two
distinct solution species endowed with different structures. The
observation of a 35/65 distribution of the two kind of molecules
between the slow and fast regime in Gd(AAZTA-C4-COO-)
derivative lends support to the hypothesis of the presence of the
isomeric species as frequently observed in the case of solutions
of macrocyclic LnIII complexes.[27]
contribution from second sphere water molecules.[24] The first
hypothesis was checked by measuring the relaxation rate of
Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-) as a function of
their concentration in the range 0.05-3.5 mM (ESI-Figure S2), at
neutral pH, 25°C and 21MHz. A slight deviation from the linearity
is observed, for both Gd-complexes, above concentration ca 0.1
mM, indicating the likely occurrence of
a concentration
dependent aggregation process. Linear fitting of data at
concentration < 0.1 mM afforded relaxivity values of 7.7 and 8.6
mM-1s-1 for Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-),
respectively. These values still fall slightly above the line of
expected values on the basis of molecular weight, leading to
foresee an additional contribution from second sphere water
molecules.
In order to determine the kinetic parameters governing the water
exchange process, the transversal relaxation rate (R2p) of 17O
nucleus of 17OH2 were determined by 17O NMR spectroscopy at
variable temperature (273 – 346 K)[25] in the presence of
Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-) complexes
(Figure 3). Overall all three investigated systems show similar
transverse relaxivity values (R2p = R2(complex)-R2(control)) and are
consistent with the presence of two inner sphere water
molecules as for the parent Gd(AAZTA). However, while
Gd(AAZTA) shows a single, pretty-defined bell-shaped profile, in
the case of Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-)
derivatives, there is a clear evidence of an additional species
that shows up in the low temperature measurements (273-283
K).
Figure 4. 1/T1 NMRD profile from 0.01 to 80 MHz of a 1mM aqueous solution
of Gd(AAZTA-C2-COOH) (a) and Gd(AAZTA-C4-COOH) (b) at pH = 7 and
25C () and 37C ().
To get more insight into the solution structures of the two n-alkyl
acid derivatives of Gd(AAZTA) we went to acquire the high
1
resolution H-NMR spectra of the corresponding EuIII and YbIII
complexes. Recently published VT-1H NMR spectra of the
Ln(AAZTA) complexes were also considered.[20] As shown in
1
Figures S2-S6 the H-NMR spectra showed a pattern made of
Figure 3. Comparison of the temperature dependence of the reduced
half the number of methylenic protons present in each of the
considered AAZTA ligands. The observed pattern did not
change over an extensive range of temperature (274 – 313 K)
indicating the occurrence of a fast dynamic process that
hampers the detection of “frozen” structures. Actually, by
passing from water to a D2O/DMSO-d6 (60/30 vol %) mixture it
was possible to acquire spectra down to 248 K (Figures S2 and
S5). Inspection in the lowest temperature spectra allowed to
access differences in the broadening of the resonances that may
be suggestive of an incipient coalescence process. The
observed behaviour is taken as an additional support for the
hypothesis for the presence, in solution, of two interconverting
isomers, differing for the exchange regime of the coordinated
water molecules.
transverse 17O water relaxation (R2p
)
for Gd(AAZTA-C2-COO-) (),
Gd(AAZTA-C4-COO-) () and Gd(AAZTA) () at 14.09 T pH=7.4 ([GdL]=20.0
mM).
The profiles obtained for the two Gd(AAZTA) derivatives suggest
that the inner sphere water molecules in these species display
both fast- and slow-regime exchange rate. The analysis of the
observed profiles using the modified Swift-Connick equations[26]
yielded the following results: for Gd(AAZTA-C2-COO-), the
water molecules in fast exchange with the bulk solvent
molecules are characterized by a M value of 7.0 ns and weight
for about 35% of the total whereas 65% of water molecules are
characterized by a M value of 330 ns. In the case of Gd(AAZTA-
C4-COO-) an analogous population distribution is observed with
the corresponding values being 7.6 ns and 286 ns, respectively.
In principle, the presence of fast- and slow- exchanging
populations might be ascribed either to the structurally different
More information on the determinants of the observed relaxivity
was obtained recording the 1/T1 NMRD profiles over an
extended range of magnetic field strengths (0.01-80 MHz as
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
proton Larmor Frequency). The NMRD profiles measured at
neutral pH, 25oC and 37oC for Gd(AAZTA-C2-COO-) and
Gd(AAZTA-C4-COO-), are shown in Figure 4. Experimental data
were fitted using the SBM theory to extract the relevant
relaxometric parameters (∆2, V and R) reported in Table 3. In
the fitting procedure, the values of τM for the two derivatives
were kept fixed to the weighted average of the values
determined by 17O-NMR-R2p.
The kinetic inertness of the Gd(AAZTA) was assessed by
investigating the rates of the transmetallation reaction with the
use of Cu2+ or Eu3+ as exchanging ions.[18] For a direct
comparison of the kinetic properties of Gd(AAZTA-C2-COO-),
Gd(AAZTA-C4-COO-) and Gd(AAZTA), the same method and
identical conditions were used as previously used for
Gd(AAZTA).[18] The rates of the transmetallation reactions were
studied by spectrophotometry in the presence of Cu2+ as
exchanging metal ion. Mechanism proposed for the
transmetallation reactions of the Gd(AAZTA-C2-COO-) and
Gd(AAZTA-C4-COO-) with Cu2+ are summarized in Scheme 4.
Experimental details and equations used to evaluate the kinetic
parameters are summarized in ESI. The rate and equilibrium
constants that characterize the transmetalation reaction of
Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-) are shown and
compared with those of Gd(AAZTA) in Table 4. In our
experimental condition (pH=2.8 – 5.0) the dominating species is
Gd(HL) with the protonated n-propionate and n-valerate
pendants.
Table 3. Relaxation parameters of the Gd(AAZTA-C2-COO-), Gd(AAZTA-
C4-COO-) and Gd(AAZTA) complexes as derived from fitting of 17OR2p vs
temp. and NMRD profiles (25°C).
Gd(AAZTA-C4-
COOH)
Gd(AAZTA-C2-
COOH)
Gd(AAZTA)[g]
r1p
9.4
8.8
7.1
(0.47T, 298K)
2 ×1019 / s-2 [a]
2.3±0.21
32.6±2.9
115±2.2
3.2±0.24
26.1±1.7
92±1.5
2.15
31.0
74

V / ps [b]
R / ps [c]

species
1
7.0 (35%)
7.6 (35%)
286 (65%)
M
/
90
ns [d]
species
2
331 (65%)
q[e]
2
1
2
1
2
0
[f]
qSS
Scheme 4. Proposed mechanism of the transmetallation reactions for the
Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-) complexes. The lower and
the last upper paths are valid for the Gd(AAZTA-C2-COO-) and Gd(AAZTA-
C4-COO-) complex, respectively.
[a] Squared mean transient zero-field splitting (ZFS) energy as obtained from
[b]
fitting of NMRD profiles.
Correlation time for the collision-related
modulation of the ZFS Hamiltonian as obtained from fitting of NMRD
profiles. [c] Re-orientational correlation time as obtained from fitting of NMRD
Table 4. Rate (ki) and equilibrium constants (Ki) and half-lives (t1/2=ln2/kd) for
the transmetallation reactions of Gd(AAZTA-C4-COO-), Gd(AAZTA-C2-
COO-) and Gd(AAZTA) complexes (25C)
[d]
profiles.
Exchange life-time of the coordinated water molecules as
[e]
obtained from fitting of 17OR2pvsT profiles of Fig.3 .
Number of inner
sphere water molecules as obtained from fitting of NMRD profiles. [f] Number
of second sphere water molecules as obtained from fitting of NMRD
profiles., [g] Ref.[17]
Gd(AAZTA-C4-
COO-)
Gd(AAZTA-C2-
COO-)
Gd(AAZTA)[a]
The best fitting of experimental data afforded the presence of
one second sphere water molecule for both Gd(AAZTA-C2-
COO-) and Gd(AAZTA-C4-COO-) which was not present in the
parent Gd(AAZTA). This result supported the suggestion related
to the observation that the proton relaxivity values at 21.5 MHz
(Figure 2) fall slightly above the line of expected values
calculated on the basis of the molecular weights. This additional
contribution from second sphere water molecules is likely related
to the presence on the herein reported complexes of the acidic
alkyl function, which is expected to enhance the overall
hydrophilic character of the complexes.
I
0.15 M NaCl
0.1 M KCl[b]
kGdHL / s-1
(7  2)10-6
(2.3  0.2)10-3
1.2  0.3

(2  2)10-5
(2.2  0.1)10-3

4.510-3
kGd(H2L) / s-1
kGd(H3L) / s-1
kGd(L)Cu / s-1
KGd(HL) / M-1
KGd(H2L) / M-1
KGd(L)Cu / M-1
kd / s-1 pH=7.4
t1/2 / h pH=7.4
a Ref. [18]


2.110-5
233
(1.0  0.1)10-5
28183 (pH-pot.)
128 (pH-pot.)
166  40
54954 (pH-pot.)
118 (pH-pot.)


Transmetallation of Gd(AAZTA-C2-COO-) and Gd(AAZTA-
C4-COO-) complexes with Cu2+
9
1.410-8
2.310-8
4.010-8
4.3103
Besides the high thermodynamic stability, GBCAs must be
characterized by high kinetic inertness in order to avoid the in
vivo dissociation of the GdIII-complexes. Actually it has been
realized that the kinetic inertness is more important than the
absolute value of the stability constant for the in vivo
applications of GdIII complexes.[3,28–30] Body fluids are very
complex systems and the in vivo study of the rate of dissociation
reactions of GdIII complexes is considered a difficult task.[31–33]
However, in vitro studies are usually quite suitable to predict the
kinetic behaviour of the GdIII complexes under in vivo conditions.
1.3104
1.0104
According to the proposed mechanism, the dissociation of the
Gd(AAZTA) derivatives might take place by the proton- and
metal-assisted pathways.[18] The proton assisted dissociation of
the GdIII-complexes generally takes place by the protonation of
the complexes via the formation of protonated intermediates.
The protonation of the Gd(AAZTA) derivatives presumably
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
occurs on the carboxylate groups. The dissociation that might
take place via proton transfer from the carboxylate group to the
nitrogen atom, results in the displacement of the Gd3+ ion from
the coordination cage and the release of the Gd3+. The kGd(HL)
rate constants characterizing the dissociation rate of the
monoprotonated Gd(AAZTA-C2-COOH) and Gd(AAZTA-C4-
COOH) species are about 44 and 1500 times smaller that that of
monoprotonated Gd(HAAZTA). Since the protonation of the
Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-) takes place on
the distant n-propionate and the n-valerate side chain, the
proton transfer to the nitrogen atom is less probable than that of
Gd(HAAZTA) species with the protonated carboxylate group of
the AAZTA skeleton. On the other hand, the dissociation rate of
derivatives C2 and C4 spaced, respectively, in comparison to
Gd(AAZTA) (r1p = 7.1 mM-1.s-1). 17O- and 1H-NMR spectra
results suggest the presence in solution of species containing
water molecules in two distinct motion regimes. The observation
of fast and slow water exchanging isomers is not uncommon in
lanthanide(III) macrocyclic complexes and it is usually
accompanied by the observation of the corresponding sets of
signals in the corresponding 1H-NMR spectra of the analogs with
other LnIII-ions. Herein the interconversion appears too fast to be
slowed down in the high resolution NMR spectra.
A corollary to these observations is that, for the first time to our
knowledge, it has been possible, in the field of LnIII containing
complexes, to extract fundamental information on the presence
of two interconverting isomers from the 17O NMR data when the
high resolution NMR spectra were completely non informative.
the
double
protonated
Gd(HAAZTA-C2-COOH)
and
Gd(HAAZTA-C4-COOH) species (kGd(H2L)) are comparable with
the kGd(HL) value of the Gd(HAAZTA) species. Since the second
protonation of Gd(AAZTA-C2-COO-) and Gd(AAZTA-C4-COO-)
and the first protonation of Gd(AAZTA) occur on the carboxylate
group of the AAZTA unit, the dissociation of these protonated
GdIII-complexes might takes place by the similar probability.
However, the somewhat lower kGd(H2L) values of Gd(HAAZTA-
C2-COOH) and Gd(HAAZTA-C4-COOH) than that of kGd(HL) of
Gd(HAAZTA) may be explained by the interaction between the
Gd3+ ion and the basic nitrogen atoms of AAZTA-C2-COO- and
AAZTA-C4-COO-. The stability constants of the heter-dinuclear
Gd(AAZTA-C2-COO-)Cu intermediate (KGd(L)Cu) is about 16 times
higher than that of Gd(AAZTA)Cu, which can be explained by
the coordination of the CuII-ion by the more basic n-propionate
side in Gd(AAZTA-C2-COO-)Cu intermediate. kGd(L)Cu rate
constant, characterizing the dissociation of dinuclear
Gd(AAZTA-C2-COO-)Cu intermediate is about two times smaller
than that of Gd(AAZTA)Cu. Because of the stronger interaction,
the functional groups of the AAZTA-C2-COO- ligand are
transferred with the much smaller rate from the GdIII to the
attacking CuII ion in respect to what was found in the case of
Gd(AAZTA). This hypothesis is further supported by the lack of
the Cu2+ assisted dissociation of the Gd(AAZTA-C4-COO-)
complex due to the sronger coordination of the Gd3+ ion with the
more basic nitrogen atoms of the ligand. Somewhat larger
kinetic inertness of Gd(AAZTA-C4-COO-) and Gd(AAZTA-C2-
COO-) complexes is also confirmed by the calculated
dissociation rate (kd) and half-life (t1/2) values at pH=7.4. kd and
t1/2 values in Table 4 reveal that the dissociation of Gd(AAZTA-
C2-COO-) and Gd(AAZTA-C4-COO-) complexes slightly slower
than that of Gd(AAZTA).
Experimental
1. General
All standard chemicals were acquired from Merck and VWR. The new
compounds were characterized by using 1H and 13C NMR. 1H and 13C-
NMR spectra were recorded at 298K on a Bruker spectrometer (600 MHz
for 1H). UPLC (Ultra-performance liquid chromatography)-analytical
characterizations were carried out using a Waters UPLC-H-Class system
equipped with Acquity QDa MS detector and dual-wavelength UV/Vis
TUV detector. Purification by using HPLC was performed with a system
equipped with Waters 2767 autosampler and autoinjector, Waters 2525
pumps, Waters 3100 MS Detector and Waters 2998 photodiode array
(PDA) detector, on Atlantis dC18 OBD Prep Column, 100Å, 5 µm, 19 mm
X 100 mm. Water 0.1% TFA (A) and acetonitrile 0.1% TFA (B) were used
as eluents.
2. Synthesis of H4AAZTA-C4-COOH ligand (Scheme 2)
Compound 1: N,N’-dibenzylethylenediamine (28.6 mmol, 6.7 mL) was
suspended in EtOH (50 mL) and the mixture was stirred until a clear
solution was obtained. Paraformaldehyde (86 mmol, 2.6 g) was added
and the suspension was stirred and heated at 80°C for 1.5 h. The
solution of 6-nitrohexanoic acid methyl ester (28.6 mmol, 5 g) in EtOH
(10 mL) was added dropwise. The new solution was left to cool to room
temperature and stirred for 18 hours. The mixture was evaporated and
the residue dissolved in ethyl acetate (EtOAc) was purified by flash
chromatography (silica gel column, 90:10 petroleum ether/EtOAc) to give
1 as a pale yellow oil (7.4 g, 60%). 1H NMR (CDCl3) : δ 7.32 (m, 10H),
3.76 (d, 2H, J= 13Hz), 3.67 (s, 3H), 3.60 (d, 2H, J=13Hz), 3.53 (d, 2H,
J=14.2), 2.98 (d, 2H, J=14.2), 2.64 (m, 4H), 2.13 (t, 2H, J=7.5), 1.59 (m,
2H), 1.32 (m, 2H), 0.79 (m, 2H). 13C NMR (CDCl3): δ 173.1, 138.5, 128.5,
127.6, 126.5, 94.1, 63.4, 60.8, 58.1, 50.9, 36.0, 32.9, 23.9, 21.9. ESI-MS
(m/z): [M+H]+ 440.5 (obsd.), 440.5 (calcd. for C25H34N3O4).
Conclusions
Compound 3: Palladium on carbon (1.0 g) was slowly added to a solution
of compound 1 (15.5 mmol; 6.8 g) in MeOH (300 mL). The suspension
was stirred at 40°C for 5 h under hydrogen atmosphere. The suspension
was filtered on a bed of Celite and concentrated in vacuo. The residue
was dissolved in acetonitrile (MeCN) (75 mL) and then freshly ground
K2CO3 (80.2 mmol; 11.0 g) and Na2SO4 (15.0 mmol; 2.1 g) were added.
t-Butyl bromoacetate (71.0 mmol; 10.4 g) was added and the orange
mixture was stirred and heated at 80°C for 20 hours. The mixture was
filtered, evaporated and the residue was purified by chromatography
(silica gel column, 30: 5 n-hexane/EtOAc to give a pale yellow oil (6.0 g;
57%).1H NMR (CDCl3): δ 3.66 (s, 3H; OCH3), 3.61 (s, 4H), 2.69-3.01 (br,
6H), 2.32 (t, 2H, J= 7.5 Hz), 1.53-1.64 (br, 6H), 1.45 (s, 36 H). 13C NMR
(CDCl3): δ 173.3. 172.0, 81.3, 80.4, 65.5, 62.8, 61.5, 60.6, 51.2, 50.8,
35.8, 33.2, 27.5, 24.8, 21.5. ESI-MS (m/z): [M+H]+ 686.6 (obsd.), 686.5
(calcd. for C35H64N3O10);
Two functionalized Gd(AAZTA) derivatives have been
successfully synthetized and characterized. They appear
promising systems for further conjugation to substrates of
interest. Moreover it was found that they show improved
relaxation properties and slightly better thermodynamic and
kinetic behaviour when compared with those ones of the parent
Gd(AAZTA). Therefore, the body of these results leads to
conclude that these complexes are good candidates as contrast
agents for further MRI applications either by their own or as
systems for the design of targeting agents. Relaxometric studies
demonstrated that these complexes presented improved
relaxivity values (r1p = 8.8 mM-1.s-1 and 9.4 mM-1.s-1 for the
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
Compound 4: A 1 M solution of LiOH (23.0 mmol; 23.0 mL) was added
then freshly ground K2CO3 (2.0 g, 15 mmol) and Na2SO4 (0.36 g, 2.6
mmol) were added. Methyl bromoacetate (1.24 mL, 13.2 mmol) was
added and the mixture was stirred and heated at 80°C for 20 hours. The
mixture was filtered, more K2CO3 (2.0 g, 15 mmol), Na2SO4 (0.36 g; 2.6
mmol) and methyl bromoacetate (0.62 mL, 6.6 mmol,) were added and
the new mixture heated at 80°C for 12 h. The mixture was filtered,
evaporated and the residue was purified by chromatography (silica gel
column, 30: 5 n-hexane/EtOAc to give a pale yellow oil (0.7 g; 47%).1H
NMR (CDCl3): δ 3.75 (s, 4H), 3.69 (s, 6H), 3.67 (s, 6H), 3.36 (s, 4H), 3.01
(d, 2H, J= 14.3Hz) 2.8 (m, 2H), 2.69 (m, 4H), 2.35 (m, 2H), 1.84 (m, 2H),
1.43 (s, 9H). 13C NMR (CDCl3): δ 172.9, 172.6, 171.0, 79.3, 63.6, 62.1,
60.4, 58.3, 50.8, 50.6, 31.3, 28.0, 27.4. ESI-MS (m/z): [M+H]+ 532.2
(obsd.), 532.3 (calcd. for C24H42N3O10).
dropwise to
a solution of compound 3 (2.9 mmol; 2.0 g) in
tetrahydrofuran (THF) (50 mL) cooled at 0°C. The solution was then
stirred at room temperature for 28 h. The pH was brought to 7.0 by
addition of HCl 6 M. Water (45 mL) was added and THF was evaporated.
The aqueous residue was extracted with EtOAc (3 x 75 mL). The organic
phases were collected, dried with Na2SO4, filtered and concentrated in
vacuo. The crude was purified by chromatography (silica gel column, 3:2
n-hexane/ EtOAc) to give 4 as a pale yellow oil. (1.0 g; 54%). 1H NMR
(CDCl3): δ 3.63 (s, 4H), 3.16 (br, 6H), 2.41 (t, 2H), 1.61-1.67 (br, 6H),
1.48 (s, 36 H). 13C NMR (CDCl3): δ 178.7, 173.2, 171.0, 81.6, 80.7, 65.3,
63.3, 61.5, 60.3, 37.4, 34.3, 29.0, 27.5, 21.4. ESI-MS (m/z): [M+H]+ 672.6
(obsd.), 672.4 (calcd. for C34H62N3O10); [M+Na]+ 694.6 (obsd.), 694.8
(calcd).
Compound 4a: Trifluoroacetic acid (2 mL) was added dropwise to a
solution of 3a (0,5 g, 0.94 mmol) and triisopropylsilane (0.050 mL) in
CH2Cl2 (2mL) cooled to 0-5 °C. The solution was stirred at room
temperature for 1 h. The mixute was co-evaporated with CH2Cl2 (3
×10mL) and taken into next step without further purification. 1H NMR
(CDCl3): δ 3.89 (s, 2H), 3.83 (s, 2H), 3.79 (s, 6H), 3.75 (s, 6H), 3.71 (s,
4H), 3.47-3.31 (m, 6H), 3.16 (d, 2H, J=14.5Hz), 2.42 (t, 2H, J= 7.4), 1.79
(t, 2H, J= 7.4). 13C NMR (CDCl3): δ 175.4, 173.8, 167.3, 61.4, 60.0, 57.4,
52.9, 51.9, 51.8, 49.9, 29.2, 27.2. ESI-MS (m/z): [M+H]+ 476.4 (obsd.),
476.2 (calcd. for C20H34N3O10).
Compound 5 (H4AAZTA-C4-COOH): Trifluoroacetic acid (20 mL) was
added dropwise to
a solution of 4 (1,0 g, 1.50 mmol) and
triisopropylsilane (0.6 mL) in CH2Cl2 (5 mL) cooled to 0-5 °C. The
solution was stirred at room temperature for 20 h, then evaporated. The
residue dissolved in H2O (5 mL) was purified by HPLC
under isocratic conditions (98:2, A:B) at a flow rate of 20 mL/min. The
pure product was obtained as a white powder (0.40 g, yield 60%) and
characterized by UPLC-UV-MS-ESI(+) using a Acquity UPLC HSS T3
Column, 100Å, 1.8 µm, 2.1 mm X 100 mm and 0.05% TFA in water (A)
and 0.05% TFA in acetonitrile (B) as solvents. Elution: initial condition
0% B, isocratic 0 % B over 0.56 min, gradient 0-10% B over 2.5 min, 10-
30% B over 6 min, 30-100% B over 10 min, flow rate 0.4 mL/min and UV
Compound 5a (H4AAZTA-C2-COOH): A 1 M solution of LiOH (20.0
mmol; 20.0 mL) was added dropwise to a solution of compound 3a (0.6
mmol; 0.4 g) in tetrahydrofuran (THF) (40 mL) cooled at 0°C. The
solution was then stirred at room temperature for 28 h. The pH was
brought to 7.0 by addition of HCl 6 M. Water (45 mL) was added and
THF was evaporated. The aqueous residue was purified by HPLC
under isocratic conditions (98:2, A:B) at a flow rate of 20 mL/min. The
pure product was obtained as a white powder (0.10 g, yield 39%) and
characterized by UPLC-UV-MS-ESI(+) using method described for
compound 5, (tR = 2.5 min, purity 92.0%). 1H NMR (D2O): δ 3.80 (s, 4H),
3.71 (s, 4H), 3.40 (m, 2H), 3.32 (s, 6H), 2.38 (t, 2H, J=8.0), 1.68 (t, 2H,
J= 8.0). 13C NMR (D2O): δ 176.3, 176.1, 170.4, 61.3, 59.2, 57.7, 52.2,
50.2, 28.5, 27.1. ESI-MS (m/z): [M+H]+ 420.2 (obsd.), 420.1 (calcd. for
C16H26N3O10).
1
detection at 220 nm (tR, retention time = 3.9 min, purity 95.0%). H NMR
(D2O): δ 3.80 (d, 4H, J= 5.1 Hz), 3.71 (s, 4H), 3.60 (m, 2H), 3.44 (s, 6H),
2.31 (t, 2H, J=7.2), 1.50 (m, 2H), 1.26 (m, 2H). 13C NMR (D2O): δ 177.8,
175.6, 169.8, 61.9, 59.3, 57.8, 51.8, 51.5, 33.2, 32.6, 23.8, 21.4. ESI-MS
(m/z): [M+H]+ 448.3 (obsd.), 448.4 (calcd. for C18H30N3O10).
3. Synthesis of H4AAZTA-C2-COOH (Scheme 3)
Compound 1a: N,N’-dibenzylethylenediamine (22.0 mmol, 5.3 mL) was
suspended in EtOH (50 mL) and the mixture was stirred until a clear
solution was obtained. Paraformaldehyde ( 66.0 mmol, 2.0 g) was added
and the suspension was stirred and heated at 80°C for 1.5 h. The
solution of 2-nitroethanol (22.0 mmol, 2 g) in EtOH (10 mL) was added
dropwise. The new solution was left to cool to room temperature and
stirred for 18 hours. The mixture was evaporated and the residue
dissolved in ethyl acetate (EtOAc) was purified by flash chromatography
(silica gel column, 80:20 petroleum ether/EtOAc) to give 1 as a pale
4. Synthesis of LnIII-complexes
Aqueous solutions of GdCl3 or EuCl3 and ligands were mixed at 1:1
concentration ratio ([GdCl3]=[EuCl3]=[H4AAZTA-C4-COOH]=[H4AAZTA-
C2-COOH]=50 mM). The solutions were stirred for 4 hours. The pH of
the solutions was maintained 7 by the addition of solid NaOH. The
occurrence of residual free Ln3+ ion was assessed by UV-Vis
spectroscopy using the xylenol orange test [34]. The amount of the
residual free Ln3+ ion was less than 0.3% (mol/mol). The LnIII-complexes
were characterized by direct-infusion method using ESI-MS (Waters
3100) in negative ion mode Gd(AAZTA-C2-COOH): ESI- MS (-): m/z:
calculated for C16H21GdN3O10 [M-H]-: 573.0, found: 573.1; Eu(AAZTA-
C2-COOH): ESI- MS (-): m/z: calculated for C16H21EuN3O10 [M-H]-: 568.0,
found: 568.1; Gd(AAZTA-C4-COOH): ESI- MS (-): m/z: calculated for
C18H25GdN3O10 [M-H]-: 601.0, found: 601.2; Eu(AAZTA-C4-COOH): ESI-
MS (-): m/z: calculated for C18H25EuN3O10 [M-H]-: 596.0, found: 596.2.
The concentration of the GdIII-complexes was determined by 1H-NMR
relaxometry after remineralization for 24 hours at 120°C in concentrated
HCl solution (37%, v/v).
1
yellow oil (7.0 g, 90%). H NMR (CDCl3) : δ 7.33 (m, 10H), 3.74 (d, 2H,
J= 13Hz), 3.70 (s, 2H), 3.65 (d, 2H, J=13Hz), 3.53 (d, 2H, J=14.4), 3.05
(d, 2H, J=14.3), 2.70 (m, 2H), 2.63 (m, 2H). 13C NMR (CDCl3): δ 138.0,
128.3, 127.7, 126.7, 94.0, 65.0, 63.1, 58.37, 58.0. ESI-MS (m/z): [M+H]+
356.2 (obsd.), 356.2 (calcd. for C20H26N3O3).
Compound 2a: Compound 1a (18.0mmol, 6.3g) was dissolved in dry
THF (40 mL), tert-butyl acrylate (5.3 mL, 36.0 mmol) was added and the
mixture was stirred at room temperature for 5 minutes. t-BuOK (2.4 g,
21.6 mmol) was added to the solution, and stirring was continued at room
temperature for 3 hours. After removal of the solvent under reduced
pressure, the residue was taken up in methanol and purified by
chromatography (silica gel column, 90:10 petroleum ether/EtOAc) to give
2a as a pale yellow oil (3.9 g, 48%). 1H NMR (CDCl3): δ 7.31 (m, 10H),
3.76 (d, 2H, J= 12.5Hz), 3.61 (d, 2H, J= 12.7Hz), 3.50 (d, 2H, J=14Hz),
2.98 (d, 2H, J=14), 2.61 (m, 4H), 1.96 (m, 2H), 1.74 (m, 2H), 1.43 (s, 9H).
13C NMR (CDCl3): δ 170.3, 138.3, 128.3, 127.6, 126.6, 93.4, 80.0, 63.2,
61.0, 57.9, 31.0, 28.4, 27.3. ESI-MS (m/z): [M+H]+454.3 (obsd.), 454.3
(calcd. for C26H36N3O4).
5. Thermodynamic studies
Materials: The chemicals used for the experiments were of the highest
analytical grade. The concentration of the CaCl2, ZnCl2, CuCl2 and GdCl3
solutions were determined by complexometric titration with standardized
Na2H2EDTA and xylenol orange (ZnCl2, and LnCl3), murexid (CuCl2) and
Patton & Reeder (CaCl2) as indicators. The concentration of the
H4AAZTA, H4AAZTA-C2-COOH and H4AAZTA-C4-COOH was
determined by pH-potentiometric titration in the presence and absence of
a large (40-fold) excess of CaCl2. The pH-potentiometric titrations were
made with standardized 0.2 M NaOH.
Compound 3a: Palladium on carbon (0.3 g) was slowly added to a
solution of compound 2a (1.3 g, 2.8 mmol) in MeOH (50 mL). The
suspension was stirred at 40°C for 5 h under hydrogen atmosphere. The
suspension was filtered on
a bed of Celite and concentrated in
vacuo.The residue was dissolved in acetonitrile (MeCN) (50 mL) and
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
Equilibrium measurements: The stability and protonation constants of
CaII, ZnII and LnIII -complexes formed with AAZTA-C2-COO- and AAZTA-
C4-COO- ligands were determined by pH-potentiometric titration. The
metal-to-ligand concentration ratio was 1:1 (the concentration of the
ligand was generally 0.002 M). In calculating the equilibrium constants of
the metal complexes, the best fitting of the NaOH - pH data pairs, were
obtained by assuming the formation of ML, MHL, MH2L, MH3L, MH4L and
MLH-1 complexes in the 1.7-12.0 pH range. For the pH measurements
and titrations, Metrohm 888 Titrando titration workstation Metrohm-
6.0234.110 combined electrode was used. Equilibrium measurements
were carried out at a constant ionic strength (0.15 M NaCl) in 6 ml
samples at 25 C. The solutions were stirred, and N2 was bubbled
through them. The titrations were made in the pH range of 1.7-12.0. KH-
phthalate (pH=4.005) and borax (pH=9.177) buffers were used to
calibrate the pH meter, For the calculation of [H+] from the measured pH
constant, monochloroacetic acid (pH range 2.8
–
3.1), N,N’-
dimethylpiperazine (pH range 3.1 – 4.1) and N-methylpiperazine (pH
range 4.1 – 5.2) buffers (0.01 M) were used. The pseudo-first-order rate
constants (kd) were calculated by fitting the absorbance data to Eq. (5)
At  (A0  Ap )ek t  Ap
d
(5)
where At, A0 and Ap are the absorbance values at time t, the start of the
reaction and at equilibrium, respectively.
7. Relaxometric studies
The observed longitudinal water protons relaxation rates (R1obs) were
measured using a Stelar Spinmaster (Mede, Pavia, Italy) spectrometer
operating at 0.47 T and employing a standard inversion-recovery (IR)
pulse sequence (16 experiments, 2 scans). A typical 90o degree pulse
width was 3.5 µs and the reproducibility for the longitudinal relaxation
times (T1) were ±0.5%. NMRD profiles were obtained using a Stelar
SpinMaster FFC-NMR relaxometer from 0.01-20 MHz and a Bruker
WP80 NMR electromagnet for variable higher-field measurements (21.5 -
80 MHz), both equipped with a Stelar VTC-91 air-flow heater equipped
with copper/constant thermocouple (uncertainty ±0.10C) for temperature
control. .
values, the method proposed by Irving et al. was used as follows.[35]
A
0.01M HCl solution was titrated with standardized NaOH solution at 0.15
M NaCl ionic strength. The differences (A) between the measured
(pHread) and calculated pH (-log[H+]) values were used to obtain the
equilibrium H+ concentration from the pH values measured in the titration
experiments (A=0.024). For the equilibrium calculations, the
stoichiometric water ionic product (pKw) was also needed to calculate
[OH] values under basic conditions. The VNaOH – pHread data pairs of the
HCl – NaOH titration obtained in the pH range 10.5 – 12.0 were used to
calculate the pKw value (pKw=13.85).
8. NMR studies
The VT-1H NMR spectra of EuIII- and YbIII-complexes were recorded on
the Bruker Avance III (9.4 T) and Avance 600 (14.09 T) spectrometer,
equipped with 5 mm probe and using a D2O solution as internal lock. The
temperature was controlled with Bruker thermostating units, and high
resolution spectra have been acquired by varying the temperature from
278 K to 350 K. Variable temperature 17O NMR measurements were
performed with Bruker Avance spectrometer, equipped with a 5mm probe
and using a capillary containing D2O as external lock. The experimental
settings were: spectral width 1000 Hz, 90o degree pulse (7 µs),
acquisition time 10 ms, 1000 scans, and no sampling spinning. Aqueous
solutions containing 2.6% of the 17O isotope (Yeda, Israel) were used.
The stability constant of the Cu(AAZTA)2-, Cu(AAZTA-C2-COO-)3-
Cu(AAZTA-C4-COO-)3- complexes was determined by spectrophotometry
studying the competition reaction between the Cu2+ and H+ for the
AAZTA, AAZTA-C2-COO- and AAZTA-C4-COO- at the absorption band
of Cu(AAZTA)2-, Cu(AAZTA-C2-COO-)3- Cu(AAZTA-C4-COO-)3- in the
wavelength range of 400-800 nm. The concentration of Cu2+, AAZTA,
AAZTA-C2-COO- and AAZTA-C4-COO- was 0.002 M, while that of the H+
was varied between 0.01 and 1.0 M (6 samples). The H+ concentration in
the samples was adjusted with the addition of calculated amounts of 3 M
HCl. The ionic strength of samples was adjusted to 0.15 M (H+]0.15
M→[Na+]+[H+]=0.15 M). The samples were kept at 25 C for 7 days in
order to attain the equilibrium (the time needed to reach the equilibria
was determined by spectrophotometry). The absorbance values of the
samples were determined at 11 wavelengths (575, 595, 615, 635, 655,
675, 695, 715, 735, 755 and 775 nm). For the equilibrium calculations,
the protonation constants of the Cu(AAZTA)2-, Cu(AAZTA-C2-COO-)3-
Cu(AAZTA-C4-COO-)3- and the molar absorptivities of the Cu2+, CuL,
Cu(HL), Cu(H2L), Cu(H3L) and Cu(H4L) species were used. The
protonation constants of the complexes Cu(AAZTA)2-, Cu(AAZTA-C2-
COO)3- Cu(AAZTA-C4-COO)3- were determined by pH-potentiometric
titrations, made at 1:1 metal to ligand concentration ratios. The molar
absorptivities of the Cu2+, CuL, Cu(HL), Cu(H2L), Cu(H3L) and Cu(H4L)
species were determined at 11 wavelengths (575, 595, 615, 635, 655,
675, 695, 715, 735, 755 and 775 nm) by recording the spectra of 1, 2 and
The observed transverse relaxation rate (R2) were calculated from the
obs
signal width at half-height (1/2): R2
= 1/2. Paramagnetic
contributions to the observed transversal relaxation rate (R2p) were
obs
calculated by subtracting from R2
the diamagnetic contribution
measured at each temperature value on pure water enriched with 2.6%
17O isotope.
Acknowledgements
FVCK: The author would like to thank the Brazillian agency
FAPESP (Proc. 2012/23169-8, 2015/16624-9 and 2018/16040-
5) for the financial support. FA: The research was supported by
the EU and co-financed by the European Regional Development
Fund under the projects GINOP-2.3.2-15-2016-00008.
3
mM Cu(AAZTA)2-, Cu(AAZTA-C2-COO-)3- Cu(AAZTA-C4-COO-)3-
solutions in the pH range 1.7 – 6.0 (0.15 M NaCl, 25C). The pH was
adjusted by stepwise addition of concentrated NaOH or HCl. The
spectrophotometric measurements were made with the use of
PerkinElmer Lambda 365 UV-Vis spectrophotometer at 25 C, using 1.0
cm cells. The protonation and stability constants were calculated with the
PSEQUAD program.[36]
Conflicts of interest
There are no conflicts of interest to declare.
6. Kinetic studies of thee trans-metallation reactions
The kinetic inertness of the Gd(AAZTA-C2-COO-)2- and Gd(AAZTA-C4-
COO-)2- complexes was characterized by the rates of the exchange
reactions taking place between the GdL and Cu2+. The transmetallation
reactions with Cu2+ were studied by spectrophotometry, following the
formation of the CuII complexes at 300 nm with a PerkinElmer Lambda
365 UV-Vis spectrophotometer. The concentration of the GdL complexes
was 110-3 M, while the concentration of the Cu2+ was 20, 30, 40 and 50
times larger, in order to guarantee pseudo-first-order conditions. The
temperature was maintained at 25C and the ionic strength of the
solutions was kept constant, 0.15 M for NaCl. The exchange rates were
studied in the pH range about 2.8 – 5.0. For keeping the pH values
Keywords: Ligands • Lanthanides • Relaxation •
Thermodynamic • Kinetics
[1]
[2]
L. M. De León-Rodríguez, A. F. Martins, M. Pinho, N. Rofsky, A. D.
Sherry, J Magn Reson Imaging 2015, 42, 545–565.
C. S. Bonnet, P. H. Fries, S. Crouzy, P. Delangle, J. Phys. Chem. B
2010, 114, 8770–8781.
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
[3]
P. Caravan, J. J. Ellison, T. J. McMurry, R. B. Lauffer, Chem Rev 1999,
[28] T. J. Clough, L. Jiang, K.-L. Wong, N. J. Long, Nature Comm. 2019, 10,
1420–1433.
99, 2293–2352.
[4]
A. S. Merbach, L. Helm, É. Tóth, The Chemistry of Contrast Agents in
Medical Magnetic Resonance Imaging, Wiley, 2013.
[29] J. Wahsner, E. M. Gale, A. Rodríguez-Rodríguez, P. Caravan, Chem.
Rev. 2019, 119, 957–1057.
[5] C. F. G. C. Geraldes, S. Laurent, Cont. Med. & Mol. Imag. 2009, 4, 1–23.
[30] E. Brücher, G. Tircsó, Z. Baranyai, Z. Kovács, A. D. Sherry, in The
Chemistry of Contrast Agents in Medical Magnetic Resonance Imaging
(Eds.: A.S. Merbach, L. Helm, E. Toth), John Wiley & Sons, Ltd, 2013,
pp. 157–208.
[6]
[7]
[8]
[9]
S. Aime, M. Botta, M. Fasano, E. Terreno, Chem. Soc. Rev. 1998, 27,
19–29.
S. Hajela, M. Botta, S. Giraudo, J. Xu, K. N. Raymond, S. Aime, J. Am.
Chem. Soc. 2000, 122, 11228–11229.
[31] P. M. May, P. W. Linder, D. R. Williams, J Chem Soc Dalton 1977, 588–
595.
E. J. Werner, J. Kozhukh, M. Botta, E. G. Moore, S. Avedano, S. Aime,
K. N. Raymond, Inorg Chem 2009, 48, 277–286.
[32] Z. Baranyai, Z. Palinkas, F. Uggeri, A. Maiocchi, S. Aime, E. Brucher,
Chem-Eur J 2012, 18, 16426–16435.
E. Gianolio, G. B. Giovenzana, D. Longo, I. Longo, I. Menegotto, S.
Aime, Chemistry – A European Journal 2007, 13, 5785–5797.
[33] Z. Baranyai, E. Brucher, F. Uggeri, A. Maiocchi, I. Toth, M. Andrasi, A.
Gaspar, L. Zekany, S. Aime, Chem-Eur J 2015, 21, 4789–4799.
[10] E. Gianolio, C. Cabella, S. Colombo Serra, G. Valbusa, F. Arena, A.
Maiocchi, L. Miragoli, F. Tedoldi, F. Uggeri, M. Visigalli, P. Bardini, S.
Aime, J. Biol. Inorg. Chem. 2014, 19, 715–26.
[34] A. Barge, G. Cravotto, E. Gianolio, F. Fedeli, Contrast Media
Molecular Imaging 2006, 1, 184–188.
&
[11] I. Mamedov, J. Engelmann, O. Eschenko, M. Beyerlein, N. K. Logothetis,
Chem. Commun. 2012, 48, 2755–2757.
[35] H. M. Irving, M. G. Miles, L. D. Pettit, Analytica Chimica Acta 1967, 38,
475–488.
[12] L. Manzoni, L. Belvisi, D. Arosio, M. P. Bartolomeo, A. Bianchi, C.
Brioschi, F. Buonsanti, C. Cabella, C. Casagrande, M. Civera, M. De
Matteo, L. Fugazza, L. Lattuada, F. Maisano, L. Miragoli, C. Neira, M.
Pilkington-Miksa, C. Scolastico, ChemMedChem 2012, 7, 1084–93.
[13] E. Farkas, J. Nagel, B. P. Waldron, D. Parker, I. Tóth, E. Brücher, F.
Rösch, Z. Baranyai, Chemistry – A European Journal 2017, 23, 10358–
10371.
[36] L. Zékány, I. Nagypál, in Computational Methods for the Determination
of Formation Constants (Ed.: D. J. Leget), Plenum Press, New York,
1985, pp. 291–353.
[14] B. P. Waldron, D. Parker, C. Burchardt, D. S. Yufit, M. Zimny, F. Roesch,
Chem Commun (Camb) 2013, 49, 579–81.
[15] J.-P. Sinnes, J. Nagel, B. P. Waldron, T. Maina, B. A. Nock, R. K.
Bergmann, M. Ullrich, J. Pietzsch, M. Bachmann, R. P. Baum, F. Rösch,
EJNMMI Research 2019, 9, 48.
[16] J. Martinelli, G. Gugliotta, L. Tei, Org. Lett. 2012, 14, 716–719.
[17] S. Aime, L. Calabi, C. Cavallotti, E. Gianolio, G. B. Giovenzana, P. Losi,
A. Maiocchi, G. Palmisano, M. Sisti, Inorg. Chem. 2004, 43, 7588–90.
[18] Z. Baranyai, F. Uggeri, G. B. Giovenzana, A. Benyei, E. Brucher, S.
Aime, Chem-Eur J 2009, 15, 1696–1705.
[19] Z. Baranyai, F. Uggeri, A. Maiocchi, G. B. Giovenzana, C. Cavallotti, A.
Takács, I. Tóth, I. Bányai, A. Bényei, E. Brucher, S. Aime, Eur. J. Inorg.
Chem. 2013, 2013, 147–162.
[20] Z. Baranyai, D. D. Castelli, C. Platas-Iglesias, D. Esteban-Gomez, A.
Bényei, L. Tei, M. Botta, Inorg. Chem. Front. 2020, 7, 795–803.
[21] A. Vagner, E. Gianolio, S. Aime, A. Maiocchi, I. Toth, Z. Baranyai, L. Tei,
Chem Commun (Camb) 2016, 52, 11235–8.
[22] D. Delli Castelli, L. Tei, F. Carniato, S. Aime, M. Botta, Chem Commun
(Camb) 2018, 54, 2004–2007.
[23] L. Barcza, K. Mihályi, Zeitschrift für Physikalische Chemie 1977, 104,
199–212.
[24] M. Botta, European Journal of Inorganic Chemistry 2000, 2000, 399–
407.
[25] K. Micskei, L. Helm, E. Brucher, A. E. Merbach, Inorganic Chemistry
1993, 32, 3844–3850.
[26] T. J. Swift, R. E. Connick, J. Chem. Phys. 1962, 37, 307–320.
[27] a) L. R. Tear, C. Carrera, E. Gianolio and S. Aime, Chem. – Eur. J.
2020, 26, 6056-6063; b) D. Delli Castelli, M. C. Caligara, M. Botta, E.
Terreno and S. Aime, Inorg. Chem. 2013, 52, 7130-7138; c) C. Platas-
Iglesias, Eur. J. Inorg. Chem. 2012, 2012, 2023-2033; d) S. Aime, M.
Botta, Z. Garda, B. E. Kucera, G. Tircso, V. G. Young and M. Woods,
Inorg. Chem. 2011, 50, 7955-7965.
This article is protected by copyright. All rights reserved.

10.1002/chem.202004479
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
Layout 1:
FULL PAPER
The good properties of Gd(AAZTA) as
MRI contrast agent are further
improved by simply attaching a short
n-alkyl function on the ligand's
surface. The VT-T2 measurements
of the 17OH2 resonance suggest the
occurrence of two interconverting
isomers in solution characterized by
very different exchange rates of the
coordinated water molecules.
Flávio Vinicius Crizóstomo Kock, Attila
Forgács, Nicol Guidolin, Rachele
Stefania, Adrienn Vágner, Eliana
Gianolio,* Silvio Aime and Zsolt
Baranyai*
Page No. – Page No.
[Gd(AAZTA)]- derivatives with n-alkyl
acid side chains show improved
properties for their application as MRI
contrast agents
This article is protected by copyright. All rights reserved.