Kinetics of Ga(NOTA) formation from weak Ga-citrate complexes

Article abstract of DOI:10.1021/ic201445e

Gallium complexes are gaining increasing importance in biomedical imaging thanks to the practical advantages of the ⁶⁸Ga isotope in Positron Emission Tomography (PET) applications. ⁶⁸Ga has a short half-time (t_1/2 = 68 min); thus the ⁶⁸Ga complexes have to be prepared in a limited time frame. The acceleration of the formation reaction of gallium complexes with macrocyclic ligands for application in PET imaging represents a significant coordination chemistry challenge. Here we report a detailed kinetic study of the formation reaction of the highly stable Ga(NOTA) from the weak citrate complex (H₃NOTA = 1,4,7-triazacyclononane-1,4, 7- triacetic acid). The transmetalation has been studied using ⁷¹Ga NMR over a large pH range (pH = 2.01-6.00). The formation of Ga(NOTA) is a two-step process. First, a monoprotonated intermediate containing coordinated citrate, GaHNOTA(citrate), forms in a rapid equilibrium step. The rate-determining step of the reaction is the deprotonation and slow rearrangement of the intermediate accompanied by the citrate release. The observed reaction rate shows an unusual pH dependency with a minimum at pH 5.17. In contrast to the typical formation reactions of poly(amino carboxylate) complexes, the Ga(NOTA) formation from the weak citrate complex becomes considerably faster with increasing proton concentration below pH 5.17. We explain this unexpected tendency by the role of protons in the decomposition of the GaHNOTA(citrate)* intermediate which proceeds via the protonation of the coordinated citrate ion and its subsequent decoordination to yield the final product Ga(NOTA). The stability constant of this intermediate, log K _{GaHNOTA(citrate)} = 15.6, is remarkably high compared to the corresponding values reported for the formation of macrocyclic lanthanide(III)-poly(amino carboxylates). These kinetic data do not only give mechanistic insight into the formation reaction of Ga(NOTA), but might also contribute to establish optimal experimental conditions for the rapid preparation of Ga(NOTA)-based radiopharmaceuticals for PET applications.

Full text of DOI:10.1021/ic201445e

ARTICLE
pubs.acs.org/IC
Kinetics of Ga(NOTA) Formation from Weak Ga-Citrate Complexes
ꢀ
Jean-Fran c- ois Morfin and Eva T ꢀo th*
Centre de Biophysique Mol ꢀe culaire, Centre National pour la Recherche Scientifique (CNRS), rue Charles, Sadron, 45071 Orl ꢀe ans,
France
S Supporting Information
b
ABSTRACT:
68
Gallium complexes are gaining increasing importance in biomedical imaging thanks to the practical advantages of the Ga isotope in
68
68
Positron Emission Tomography (PET) applications. Ga has a short half-time (t = 68 min); thus the Ga complexes have to be
1/2
prepared in a limited time frame. The acceleration of the formation reaction of gallium complexes with macrocyclic ligands for
application in PET imaging represents a significant coordination chemistry challenge. Here we report a detailed kinetic study of the
formation reaction of the highly stable Ga(NOTA) from the weak citrate complex (H NOTA = 1,4,7-triazacyclononane-1,4,7-
3
71
triacetic acid). The transmetalation has been studied using Ga NMR over a large pH range (pH = 2.01ꢀ6.00). The formation of
Ga(NOTA) is a two-step process. First, a monoprotonated intermediate containing coordinated citrate, GaHNOTA(citrate)*,
forms in a rapid equilibrium step. The rate-determining step of the reaction is the deprotonation and slow rearrangement of the
intermediate accompanied by the citrate release. The observed reaction rate shows an unusual pH dependency with a minimum at
pH 5.17. In contrast to the typical formation reactions of poly(amino carboxylate) complexes, the Ga(NOTA) formation from the
weak citrate complex becomes considerably faster with increasing proton concentration below pH 5.17. We explain this unexpected
tendency by the role of protons in the decomposition of the GaHNOTA(citrate)* intermediate which proceeds via the protonation
of the coordinated citrate ion and its subsequent decoordination to yield the final product Ga(NOTA). The stability constant of this
intermediate, log KGaHNOTA(citrate)* = 15.6, is remarkably high compared to the corresponding values reported for the formation of
macrocyclic lanthanide(III)-poly(amino carboxylates). These kinetic data do not only give mechanistic insight into the formation
reaction of Ga(NOTA), but might also contribute to establish optimal experimental conditions for the rapid preparation of
Ga(NOTA)-based radiopharmaceuticals for PET applications.
’
INTRODUCTION
Three radioisotopes of gallium ( Ga, Ga, Ga) have
form. The precipitation of Ga(OH) starts already under acidic
3
5
66
67
68
conditions, from pH ∼3.3, though at nanomolar concentration,
1
67
typically used in radiopharmaceutical applications, no precipita-
attractive nuclear properties for medical use. Ga has been applied
as gallium citrate for over 30 years as an alternative to
Single Photon Emission Computed Tomography (SPECT) imag-
ing, though lately it has been proved that the citrate does not
prevent the transchelation of the Ga ion to biomolecules such as
transferrin. The two other isotopes are positron emitters and can be
used in Positron Emission Tomography (PET) imaging for differ-
ent diagnostic purposes considering their very different half-times
2
3+
99 m
tion occurs. A safe biomedical use of Ga requires chelating
agents which prevent interaction of the metal with hydroxide but
also with endogenous ligands which can transport the isotope to
the area of biological interest. The ligands have to form gallium
complexes of high stability and kinetic inertness to avoid any
ligand exchange or transmetalation in the biological medium. In
the context of biomedical imaging, macrocyclic poly(amino
carboxylates) have been widely used for metal complexa-
Tc for
3+
2
68
66
(
68 min for Ga and 9.49 h for Ga), decay and the energy of the
3 66 67
68
3+
6
tion, including Ga for PET applications. These ligands are
particles emitted. In contrast to Ga and Ga which are cyclotron
68
68
68
known to form thermodynamically highly stable complexes with
products, Ga is available from a commercial Ge/ Ga generator,
which largely reduces the production costs and makes this PET
isotope easily accessible.
The Ga ion has specific characteristics which render its
aqueous coordination chemistry difficult. Because of its strong
tendency for hydrolysis, in neutral aqueous media gallium(III)
can exist only as insoluble hydroxides or in a strongly chelated
3+
various metal ions, like Ga and lanthanides, which also show
4
good kinetic inertness. Derivatives of H DOTA (1,4,7,10-tetra-
4
3
+
azacyclododecane-1,4,7,10-tetraacetic acid) and H NOTA
3
Received: July 8, 2011
Published: September 09, 2011
r 2011 American Chemical Society
10371
dx.doi.org/10.1021/ic201445e
|Inorg. Chem. 2011, 50, 10371–10378

Inorganic Chemistry
ARTICLE
3
+
Scheme 1. Formation Kinetics of Ga(NOTA) Studied via Monitoring the Transfer of Ga from the Citrate- to the NOTA-
a
Complex.
a
The protonation states indicated apply to pH 5.0.
7
1
(
1,4,7-triazacyclononane-1,4,7-triacetic acid), grafted to mol-
complexes with NOTA at different pHs and 25 °C. Ga NMR
ecules of biological interest have been commonly applied as
was used to monitor the time course of the ligand exchange. The
high quadrupole moment of the Ga nucleus results in a
significant line-broadening, which limits the observation of the
Ga NMR signal to few, very symmetric species, such as
6
8
3
68
71
vectors of Ga with promising results. For instance, Ga-
NOTA-RGD was produced rapidly with a high yield and showed
favorable properties like high stability, affinity and specificity for
angiogenesis PET imaging. The nine-member macrocyclic
NOTA is known to form a Ga complex of particularly high
stability (log KGa(NOTA) = 31.0). This is related to the smaller
size of its macrocyclic cavity as compared to the twelve-member
DOTA, which is not well adapted to the small Ga ion (63 pm).
Given the relatively short half-life of Ga, fast formation
kinetics is required for the preparation of the complexes. In the
literature, various conditions including the use of different
buffers, pHs, and temperatures have been reported for radiola-
beling of DOTA- or NOTA-derivatives. The formation reac-
tions between these macrocyclic chelates and various metals,
such as lanthanide ions but also Ga , are typically slow. The
acceleration of the complexation between the radiometal and the
chelating agents would be highly desirable since the optimization
of the labeling conditions could provide a considerable gain in the
time window available for imaging. Generally speaking, kinetic
data on Ga complexes are rather scarce, mainly because of the
difficulties associated with the strong hydrolysis of Ga . Indeed,
71
7
3+
ꢀ
71
[Ga(H O) ] , [Ga(OH) ] , or Ga(NOTA); thus, no Ga
2 6 4
12
3+
NMR signalcan be detected forthe citrate or acetatecomplexes.
6
We monitored the increase of the Ga(NOTA) signal in compar-
ison to the signal of [Ga(OH)4] which is used as reference
ꢀ
3+
8
(Figure 1).
68
The presence of acetate and citrate prevents the formation of
3+
insoluble Ga hydroxides. The absence of the hydroxide pre-
cipitate allows studying the kinetics of the metal transfer from the
citrate or acetate to the NOTA complex over a large pH range,
between pH 2 and 6. The excess of the exchanging NOTA ligand
ensures pseudo-first order conditions; thus, the rate of the
Ga(NOTA) complex formation may be described by eq 1:
9
3+
d½GaNOTAꢁ
¼
kobs½Gaꢁ0
ð1Þ
dt
3+
3
+
where [Ga] is the total concentration of free Ga and kobs is the
0
3+
pseudo-first order rate constant. The formation reaction was
investigated at different pH values and with varying concentra-
tions of NOTA for three different citrate concentrations. For all
systems and at all pHs, kobs was found to increase with the NOTA
concentration, showing saturation curves (Figure 2 and in
Supporting Information, Figure S7).
3+
to study the complexation reactions of Ga at pH > 3, the metal
must be prechelated in the form of a weak complex. While dis-
sociation kinetic data have been reported on Ga chelates,
including Ga(NOTA),
3+
10,11
to the best of our knowledge, no
3+
detailed formation kinetic study has been published on Ga
Under the experimental conditions and at pH between 4.5
and 6.3, in the Ga -citrate system the major species is
macrocyclic complexes. The objective of the present work was to
investigate the kinetics of Ga(NOTA) formation via ligand
exchange from the weak gallium citrate complex over a large
pH range (Scheme 1). Citrate has been chosen as a precomplex-
ing agent since it is probably the most often applied buffer in the
radiolabeling procedures with Ga . Its role, however, is not
limited to buffer the system, but it also behaves as a weak chelator
to avoid hydrolysis and formation of insoluble Ga(OH) in the
preparation of radiopharmaceuticals. On the other hand, the
3+
ꢀ
3-
[
Ga(citrate)OH] (log βMLOH = 7.1), while [Ga(citrate) ]
2
5
(
log βML2 = 15.0) appears only at a citrate excess. In these
experiments, in addition to the citrate, acetate is used as a buffer
in much higher concentration (0.1 M) than citrate. Acetate ions
can also form weak complexes with Ga . On the basis of litera-
ture data, we find that two species are predominant in the Ga -
68
3+
3+
13
3+
3
2
+
acetate system in the pH range of our study: [Ga (OH) Ac] is
2
2
2
+
the major species (log βM2(OH)2 L = ꢀ1.16), while [GaAc]
choice of NOTA is explained by the utmost importance of high
stability of its Ga complex in biomedical applications. In
addition, this ligand allows monitoring the formation reaction
3
+
is present in lower concentration (log β = 2.41). How-
ML
ever, citrate complexes are considerably more stable and in the
presence of citrate, no acetate complexes form (see species
distribution curves in Supporting Information).
Although the main purpose of this work was to assess the
formation kinetics of Ga(NOTA) in the presence of citrate, for
comparison, we have also performed a formation kinetic study in
a more limited pH range in solutions containing only acetate.
Like in the presence of citrate (Figure 2), saturation curves were
obtained with increasing NOTA concentration in the samples
with only acetate as well. Figure 3 compares the reaction rate
7
1
by Ga NMR, which is only applicable for highly symmetrical
3
+
Ga species such as Ga(NOTA). For comparison, the formation
kinetics of Ga(NOTA) from the weak acetate complex has been
also assessed.
’
RESULTS AND DISCUSSION
The formation kinetics of Ga(NOTA) was studied by
following the transchelation of gallium citrate or gallium acetate
1
0372
dx.doi.org/10.1021/ic201445e |Inorg. Chem. 2011, 50, 10371–10378

Inorganic Chemistry
ARTICLE
71
Figure 1. Representative Ga NMR spectra in the formation reaction of Ga(NOTA) from Ga-citrate. cGa = 4 mM, ccitr = 4.7 mM, cNOTA = 40 mM,
pH = 4.51 (0.1 M acetate buffer), I = 1 M NaNO , 25 °C. The spectra were recorded in a 10 mm NMR tube which contains a 5 mm insert with a 0.1 M
3
Na[Ga(OH) ] solution as reference.
4
Figure 2. Representative series of pseudo-first-order rate constants, kobs, as a function of NOTA concentration. cGa = 4 mM. The curves represent the fit
to the data points as explained in the text.
constants in acetate only with those measured in solutions
containing citrate and acetate. The absolute values of the observed
reaction rate constants obtained in the presence of acetate only
are considerably lower than those obtained in the presence of
citrate, showing the very important role of citrate in accelerating
the kinetics of Ga(NOTA) formation. The effect of citrate to
accelerate the formation of Ga(NOTA) is strongly dependent on
the pH, as clearly visible even in the relatively limited pH range
available for comparison between the studies with or without
citrate. While at pH 5.17 the kobs values differ only by a factor of
2ꢀ2.5, at pH 4.5, the reaction rate constants are about 10 times
higher in the presence of citrate than in its absence (Figure 3).
The comparison reveals another surprising phenomenon.
Without citrate, the formation rate of Ga(NOTA) increases with
increasing pH, as it is typically observed in the formation
14
reactions of NOTA- or DOTA-complexes. This behavior is
ꢀ
related to the OH -mediated deprotonation of the protonated
intermediate (see below). In the presence of citrate, the rate
1
0373
dx.doi.org/10.1021/ic201445e |Inorg. Chem. 2011, 50, 10371–10378

Inorganic Chemistry
ARTICLE
Figure 3. Comparison of the observed pseudo-first-order rate constants, kobs, as a function of NOTA concentration, in the absence and in the presence of
.7 mM citrate, at pH = 5.17 (4); 5.00 (ꢀ);4.77(O), 4.51 (ꢃ), cGa =4mM,cacetate = 0.1 M. The curves represent the fit to the data points as explained in the text.
4
Scheme 2
constants show this usual tendency only between pH 6.01 and
pH 5.17 under our experimental conditions, while at pH below
In the intermediate, the proton is supposed to be attached to a
macrocycle nitrogen, analogously to LnHNOTA*, LnH DOTA*
or LnH TRITA* intermediates (H TRITA = 1,4,7,11-tetraaza-
cyclotridecane-1,4,7,11-tetraacetic acid).
pulsion between the metal ion and this proton could preclude
the Ga from entering the macrocyclic cavity of the ligand
during the initial stages of complex formation.
2
5
.17, the observed rate constants strongly increase with decreas-
2
4
15ꢀ17
ing pH. This difference in the pH dependency of the reaction
rates between the citrate-free and the citrate containing samples
suggests a different mechanism of complex formation which will
be assessed in the following sections.
Electrostatic re-
3+
Formation of Ga(NOTA) in the Absence of Citrate. The
saturation curves observed for varying concentrations of NOTA
are characteristic of a complex formation reaction proceeding in
twosteps: a rapid, equilibrium formation of an intermediate which
is followed by the slow transformation of the intermediate to yield
The rate-determining step is the deprotonation and the
rearrangement of the intermediate followed by the entrance of
the metal ion into the macrocycle cage:
d½GaNOTAꢁ
ꢂ
¼
kobs½Gaꢁ ¼ k½GaHNOTA ꢁ
ð3Þ
14
0
the final product in the rate-determining step (Scheme 2).
In the literature, formation kinetic studies have been reported
dt
15
for LnNOTA complexes. The intermediate has been identified
as the monoprotonated LnHNOTA* complex which then de-
protonates to yield the final LnNOTA chelate. The monopro-
tonated nature of the LnHNOTA* intermediate was evidenced
by monitoring the pH variation occurring during the complex
formation in a slightly buffered solution. Similarly to the pH
dependency of the formation rates observed for Ga(NOTA) in
the absence of citrate (Figure 3), the rates of formation of the
LnNOTA complexes also increased with increasing pH.
In analogy to the LnNOTA complex formation, we can
assume that, in the absence of citrate, the formation reaction of
Ga(NOTA) proceeds via a two-step process. The intermediate is
supposed to be the monoprotonated complex, GaHNOTA*,
with a stability constant as defined in eq 2:
The concentration of the noncomplexed ligand can be described
by eq 4 using the protonation constants of NOTA, where the fully
deprotonated form can be neglected in the pH range of the study.
þ
þ 2
½
NOTAꢁfree ¼ ½HNOTAꢁð1 þ K ½H ꢁ þ K K ½H ꢁ Þ
2
2
3
¼
α½HNOTAꢁ
ð4Þ
ð5Þ
The total metal concentration can be expressed by
ꢂ
½
Gaꢁ ¼ ½GaHNOTA ꢁ þ ½Gaꢁ
0
free
Then kobs can be deduced as in eq 6:
ꢀ
ꢁ
KGaHNOTA
ꢂ
k ꢃ
½NOTAꢁ
α
kobs ¼
ꢀ
ꢁ
ð6Þ
ꢂ
½
GaHNOTA ꢁ
KGaHNOTA
ꢂ
KGaHNOTAꢂ ¼
ð2Þ
1 þ
½NOTAꢁ
½
Gaꢁ½HNOTAꢁ
α
1
0374
dx.doi.org/10.1021/ic201445e |Inorg. Chem. 2011, 50, 10371–10378

Inorganic Chemistry
ARTICLE
Table 1. Stability Constants, KGaHNOTA*, of the Reaction
Intermediates in the Formation of LnNOTA and Ga(NOTA)
Complexes
Table 2. Formation Rate Constants of the Complexes
Ga(NOTA) in the
Ga(NOTA) in the
a
a
b
presence of citrate
absence of citrate GdNOTA
M
log KMHNOTA*
ꢀ
1
ꢀ1
4
5
7
k
k
k
k
OH (M
s
)
(4.2 ( 0.9) ꢃ 10
(1.14 ( 0.04) ꢃ 10 7.1 ꢃ 10
3
+
ꢀ
1
ꢀ4
Ga
4.2 ( 0.3
0
1
2
(s
)
(4.3 ( 0.5) ꢃ 10
9 ( 2
0
3
+
a
ꢀ
1
ꢀ1
ꢀ1
Ce
3.2
(M
(M
s
s
)
)
3
+
a
ꢀ
2
4
Gd
3.6
(9.6 ( 0.6) ꢃ 10
3
+
a
a
b
Er
3.8
This work. Ref 15.
a
Ref 15.
shown by the 2 orders of magnitude difference in the kOH values
between Ga(NOTA) and GdNOTA (Table 2).
Formation of Ga(NOTA) in the Presence of Citrate. In the
presence of citrate, the pH dependency of the observed pseudo-
first-order rate constants is different from that observed in the
absence of citrate and follows an unexpected trend. At any of the
three different citrate concentrations, the plateau of the satura-
tion curves shows a minimum at pH 5.17, and then increases for
higher and also for lower pH values. We should note that there is
no significant difference between the three citrate concentra-
tions, thus the citrate has no effect on the kinetics in the
concentration range studied. As in the absence of citrate, the
saturation curves in Figure 2 imply a fast equilibrium formation
of an intermediate, followed by its transformation in a slower,
rate-determining step to yield the final product. To assess the
nature of the intermediate, we have monitored the pH variation
in the time course of the reaction in slightly buffered samples
Figure 4. Formation rate constants for Ga(NOTA), without citrate, as a
function of 1/[H ], at 25 °C. The curve represents the fit of the data
points to eq 7 as explained in the text.
+
3
+
(0.01 M acetate), upon mixing a solution containing Ga and
ꢀ
ꢀ
4
citrate (5 ꢃ 10 M, each, 5 mL) with a solution of H2.04NOTA
4
The pseudo-first-order rate constants measured at various pHs
were fitted to eq 6, and the stability constant of the intermediate,
KGa(NOTA)H*, and the rate constants, k, were calculated. Table 1
shows the stability constant of the monoprotonated intermediate
(5 mL, 5 ꢃ 10 M) at a starting pH 4.27 (at this pH and
concentration, the NOTA ligand has an average protonation of
2.04). Immediately after mixing the two solutions, the pH shows
a slight increase to pH 4.33, which is followed by a successive,
slow increase to pH 4.39 (within 1 day). The pH variation along
the formation reaction has been found reproducible in several
experiments. Because of the protonation equilibria involving
also the citrate ion, the situation is more complicated than it was
3+
in comparison to those calculated for the corresponding Ln
analogues. The stability constant obtained for the GaHNOTA*
3+
intermediate is higher than those calculated for the Ln analo-
gues. This difference reflects the higher stability constant of
Ga(NOTA) with respect to those of LnNOTA complexes
1
5
16
for the analogous experiment with LnNOTA, LnDOTA, or
LnTRITA complexes. Nevertheless, in the second step, the
18
17
(
log KGa(NOTA) = 31.0 vs log KGdNOTA = 13.7). On the other
hand, we have to note that the stability constant calculated for the
GaHNOTA* intermediate is still a conditional constant since
pH would be clearly expected to decrease if this second step
indeed corresponds to the deprotonation of the monoproto-
nated intermediate. The observation of an opposite pH change
is not compatible with the classical complex formation mechan-
ism involving a monoprotonated intermediate. On the basis of
this result, we hypothesize that the intermediate also contains a
3+
the [Ga] concentration in eq 2 includes all Ga species not
complexed to NOTA (in the form of acetate or eventually
hydroxo complexes).
+
The rate constants, k, are inversely proportional to the [H ]
ꢀ
3+
concentration (Figure 4). This implies an OH catalyzed
citrate ligand that remains coordinated to the Ga ion
deprotonation process of the intermediate, as described by eq 7.
(GaHNOTA(citrate)*) and the rate determining step is the
deprotonation of this intermediate accompanied by the release
of the citrate. The citrate released in this second step will
become partially protonated (log KH1 = 5.70, log KH2 = 4.35,
log KH3 = 2.91) which explains the pH increase observed
(Scheme 3). Although we do not have any direct evidence of
the nature of the intermediate, the suggested composition is the
most plausible explanation to account for the pH variation
observed.
We should note that the assumption of an intermediate that
contains coordinated hydroxide might also lead to the observed
pH increase in the second step. However, we refuted this
hypothesis since, if it was the case, a similar intermediate with
similar behavior should be observed also in the absence of
citrate.
k ¼ k þ k ꢃ 1=½H ꢁ ¼ k þ kOH½OHꢀꢁ
0
0
þ
ꢀ
0
0
¼
k þ k =K ½OH ꢁ
ð7Þ
0
w
ꢀ
The k values versus the [OH ] concentration have been fitted
to eq 7. When fitting k , we obtained a very small value with a
0
large error; therefore k was fixed to 0 and the kOH value obtained
0
is given in Table 2. The zero intercept of the straight line in
Figure 4 shows that the spontaneous transformation of the
monoprotonated intermediate is very slow and negligible under
our experimental conditions. The dominant pathway over this
pH range is the OH-catalyzed deprotonation of [GaHNOTA]*.
The formation of Ga(NOTA) via this pathway, however, is
3+
considerably slower than the formation of the Ln analogues, as
1
0375
dx.doi.org/10.1021/ic201445e |Inorg. Chem. 2011, 50, 10371–10378

Inorganic Chemistry
ARTICLE
Scheme 3
The stability constant of GaHNOTA(citrate)* is defined as
Table 3. Stability Constants of the Reaction Intermediate,
KGaHNOTA(citrate)*, and the Rate Constants, k, Calculated for
the Different pH Values and Citrate Concentrations
ꢂ
GaHNOTAðcitrateÞ ꢁ
½
K
GaHNOTAðcitrateÞꢂ ¼
ð8Þ
½
Gaꢁ½HNOTAꢁ½citrateꢁ
4
ꢀ1
citrate (mM)
pH
log KGaHNOTA(citrate)*
k ꢃ 10 (s
)
Considering that the transformation of the intermediate is the
rate-determining step of the complex formation, the rate of the
pseudo-first-order reaction can be described as
4
.7
6.00
16.0
15.9
15.8
15.7
15.6
15.6
14.9
16.7
15.5
15.7
14.9
15.8
15.9
15.5
15.0
8.8 ( 0.3
6.9 ( 0.3
5.6 ( 0.3
4.9 ( 0.3
6.7 ( 0.4
9.0 ( 0.4
19.9 ( 0.8
8.0 ( 0.5
5.5 ( 0.5
6.3 ( 0.4
19.5 ( 0.9
9.4 ( 0.6
5.7 ( 0.6
8.4 ( 0.5
22.4 ( 0.9
5
5
.83
.51
d½GaNOTAꢁ
ꢂ
5.17
.77
4.51
.03
¼
kobs½Gaꢁ ¼ k½GaHNOTAðcitrateÞ ꢁ ð9Þ
0
dt
4
The total metal concentration can be expressed by
4
ꢂ
½
Gaꢁ ¼ ½GaHNOTAðcitrateÞ ꢁ þ ½Gaꢁ
ð10Þ
0
free
9
1
.8
6.00
5.17
4.77
where [Ga]free is the total metal concentration not complexed to
NOTA. Under our conditions, the concentration of gallium
hydroxo complexes is negligible.
4.02
ꢂ
9.0
6.00
5.17
4.77
4.03
½
Gaꢁfree ¼ Ga 1 þ βGaðcitrateÞ½citrateꢁfree þ βGaðcitrateÞ23ꢀ ½citrateꢁfree
β
GaðcitrateðOHÞꢀ
½citrateꢁfree
þ
Þ ¼ β½Gaꢁ
ð11Þ
þ
½
H ꢁ
As above, the concentration of the noncomplexed ligand can
be described by eq 4, and the concentration of the noncomplexed
citrate can be expressed by eq 12, using the corresponding
protonation constants.
average: 15.6 ( 0.7
GaHNOTA(citrate)* intermediate is expected to be highly stable
in comparison to LnHNOTA* intermediates in which the
citrateꢁfree ¼ ½citrateꢁð1 þ KHcitrate½Hþꢁ
3+
½
coordination number of the Ln ion is very low (CN = 4) as
compared to the preferred CN of lanthanide ions (CN = 8 or 9).
Differences in the steric strain can be also important for the
differences in the stability. In the intermediate, the citrate is
þ KHcitrateKH citrate½Hþꢁ2
2
þ 3
þ KHcitrateKH citrateKH citrate½H ꢁ Þ
3+
2
3
expected to coordinate to the Ga as a tripodal ligand with two
¼
δ½citrateꢁ
ð12Þ
ð13Þ
deprotonated carboxylate donors and the alcoholic oxygen, as it
was proved by an X-ray crystallographic study for the Ga-
Then, kobs can be deduced as in eq 13
19
(
citrate) complex.
2
kK
ꢂ½citrateꢁ½NOTAꢁ=αβδ
The variation of the plateau of the kobs versus cNOTA curves
Figure 2) as a function of pH has already shown the unusual
pH dependency of the complex formation reaction, which is
confirmed by the pH dependency of the k values (Table 3).
When they are plotted against the inverse proton concentration,
GAHNOTAðcitrateÞ
kobs ¼ 1
(
þ KGaHNOTAðcitrateÞꢂ½citrateꢁ½NOTAꢁ=αβδ
The pseudo-first-order rate constants were fitted to eq 13 at
each pH and at the three different citrate concentrations
+
(4.7 mM, 9.8 mM, and 19 mM). The rate constants, k, and the
1/[H ], they show a curve with a minimum for all three citrate
stability constant of the intermediate, KGaHNOTA(citrate)*, were
calculated (Table 3).
concentrations (Figure 5). This is in contrast with the results
obtained in the absence of citrate (Figure 4) and with all previous
observations of complex formation reactions of LnDOTA or
LnNOTA analogues where a linear dependency of k on 1/[H ]
was reported.
At the various pHs and citrate concentrations, very similar
values have been obtained for the stability constant of the
GaHNOTA(citrate)* intermediate. The stability constant is
remarkably high, however, since the stoichiometry of the inter-
mediate is supposed to be different from that in the absence of
citrate or that detected in LnNOTA formation, we cannot
directly compare this value to the stability constants of those
intermediates. On the other hand, high stability is expected for
+
To explain this unusual pH dependency of the k values, we
have to consider the transformation of the intermediate which
was assumed to be the ternary complex involving the coordina-
3+
tion of the Ga ion to a citrate ion and to the NOTA protonated
on the macrocyclic nitrogen. The formation of the final Ga-
(NOTA) complex implies the protonation of the citrate ion
followed by its immediate dissociation and the deprotonation of
the macrocycle nitrogen. These two processes might occur
simultaneously or subsequently. The protonation of the coordi-
nated citrate is expected to be catalyzed by protons, while the
3+
GaHNOTA(citrate)* since the Ga ion has a complete coordi-
nation sphere with an overall coordination number of CN = 6,
involving the three carboxylates of the NOTA, as well as two
carboxylates and one hydroxyl from the citrate. Since the
3
+
typical coordination number is 6 in Ga complexes, the
1
0376
dx.doi.org/10.1021/ic201445e |Inorg. Chem. 2011, 50, 10371–10378

Inorganic Chemistry
ARTICLE
Figure 6. Percentage of the Ga(NOTA) complex formed after 2 min of
reaction time as a function of pH. cGa = 4 mM, ccitrate = 4.7 mM and
cNOTA = 40 mM, in 0.1 M acetate buffer, 25 °C.
Figure 5. Formation rate constants for Ga(NOTA), as a function of
+
These kinetic data clearly confirm two predominant processes
depending on the pH: a base catalysis at above pH ∼5, described
1
4
/[H ], 25 °C, measured at different citrate concentrations, ccitrate
=
.7 mM (]), 9.8 mM (0), and 19.0 mM (4). The curve represents the
by k , and a strong acid catalysis at lower pH, described by k
1
fit to all data points as explained in the text.
OH
and k . The fit also resulted in a nonzero value for the k , which
2
0
points to a relatively important contribution of the spontaneous
transformation of the intermediate (Table 2).
At pHs below 3, the complex formation is too fast; thus, the
Scheme 4. Intermediate, Assumed to Be a Ternary Complex
Involving Citrate Coordination, May Be Transformed via
Spontaneous and OH - or H -Catalyzed Pathways
ꢀ
+
reaction could not be followed, and the formation rates could not
71
be calculated from the Ga NMR measurements. Indeed, this
NMR method is limited to relatively slow reactions since at least
71
1
min is needed to record the first Ga NMR spectra after mixing
the reactants. At low pH (<3), the reaction was almost com-
plete in a few minutes; nevertheless, these data confirm that the
formation becomes considerably faster with decreasing pH. To
give a qualitative picture of the acceleration of the formation
reaction of Ga(NOTA) with increasing proton concentration,
Figure 6 shows the percentage of the Ga(NOTA) complex
3+
formed 2 min after mixing the Ga -citrate and NOTA solutions.
This empirical curve clearly shows that the formation rate of
Ga(NOTA) is remarkably higher at pH 2ꢀ3 as compared to
pH 5. Below pH 2.5, the formation rate seems to decrease with
increasing acidity.
deprotonation of the macrocycle nitrogen is a hydroxide-catalyzed
+
ꢀ
process. In addition to the H or OH catalyzed pathways, a
spontaneous transformation of the intermediate might also occur.
On the basis of this hypothesis, the rate determining step can involve
the following pathways depending on the pH, as shown in Scheme 4.
’ CONCLUSION
According to the observed pH dependency of the k values
Figure 5), the first pathway, catalyzed by OH ions, is pre-
We have studied the kinetics of Ga(NOTA) formation from
Ga -citrate and Ga -acetate complexes in a large pH-range.
Ga NMR was used to monitor the increase of the Ga(NOTA)
ꢀ
(
3+
3+
dominant for pH values above 5.17 where the formation rate of
the product, k, is inversely proportional to the proton concentra-
tion. The last two pathways, involving the k and k rate
71
concentration. We have evidenced that the use of citrate for
3+
1
2
precomplexation of the Ga not only avoids hydroxide pre-
+
constants, imply a H catalyzed dissociation of the intermediate.
At increasing proton concentration, protonation of the citrate
anion occurs which will destabilize the intermediate, and lead to
the dissociation of the citrate, followed by the deprotonation of
the macrocyclic nitrogen and the subsequent entering of the
metal into the macrocyclic cavity.
cipitation but also results in much higher formation rates than
those observed in an acetate buffer without citrate. In the
presence of citrate, the formation kinetics has an unexpected
pH dependency. The reaction rates show a minimum at pH ∼5,
then become faster with increasing or decreasing proton con-
centration. The first step of the reaction is the equilibrium
formation of an intermediate which is monoprotonated on the
macrocyclic nitrogen. In the presence of citrate, the intermediate
is assumed to be a highly stable ternary complex involving also
the coordination of a citrate ion GaHNOTA(citrate)*. In the
second step, this intermediate can be transformed to yield
The rate of formation of Ga(NOTA) from the intermediate
can be then described by eq 14. The rate constants k versus
+
1
/[H ] as shown in Figure 5 have been fitted to eq 14,
simultaneously for all three citrate concentration. The values of
kOH, k , k , and k obtained in the fit are summarized in Table 2.
0
1
2
ꢀ
+
Ga(NOTA) via both OH - or H -catalyzed pathways which
accounts for the unusual pH-dependency of the reaction rates.
We hypothesize that the H catalysis is related to the protonation
ꢀ
þ
þ 2
k ¼ k0 þ kOH½OH ꢁ þ k1½H ꢁ þ k2½H ꢁ
ð14Þ
+
1
0377
dx.doi.org/10.1021/ic201445e |Inorg. Chem. 2011, 50, 10371–10378

Inorganic Chemistry
ARTICLE
of the coordinated citrate which leads to its decoordination and
’ ACKNOWLEDGMENT
to the subsequent formation of the final complex. The effect of
OH ions, classically observed in analogous metal complex
This work was financially supported by the Centre National
pour la Recherche Scientifique (CNRS) and the Agence National
de la Recherche (ANR) and was performed in the frame of the
COST Action D38 “Metal-Based Systems for Molecular Imaging
Applications”.
ꢀ
formation reactions, can be explained by the OH- catalysis of
the deprotonation of the macrocycle nitrogen. Even if not all
mechanistic aspects can be directly proved, our experimental data
clearly indicate that the Ga(NOTA) formation in the presence of
citrate is faster at pH 2ꢀ3 than at higher pHs. These unexpected
results, which have to be validated in radiochemical concentra-
tions, might contribute to the development of more efficient
’ REFERENCES
(1) Wadas, T. J.; Wong, E. H.; Weisman, G. R.; Anderson, C. J.
68
67
radiolabeling strategies for Ga or Ga complexes used as radio-
pharmaceuticals in diagnostic applications.
Chem. Rev. 2010, 110, 2858.
2) Bartholoma, M. D.; Louie, A. S.; Vallian, J. F.; Zubieta, J. Chem.
Rev. 2010, 110, 2903.
3) Audi, G.; Bersillon, O.; Blachot, J.; Wapstra, A. H. Nucl. Phys. A
003, 729 (1), 3.
4) Fani, M.; Andr ꢀe , J. P.; M €a cke, H. R. Contrast Media Mol. Imag.
008, 3, 53.
(
(
’
EXPERIMENTAL SECTION
2
2
(
Sample Preparation. H NOTA and sodium citrate were pur-
3
chased from Chematech and Aldrich, respectively. The Ga(NO3)3
solution was prepared by dissolving 99.99% Ga metal in HNO (the
final pH was 1.3). The Ga concentration was determined by adding
excess of Na H EDTA solution to the Ga(NO ) solution, and titrating
(5) Jackson, G. E.; Byrne, M. J. J. Nucl. Med. 1996, 37, 379.
3
(6) Martell, A. E.; Motekaitis, R. J.; Clarke, E. T.; Delgado, R.; Sun,
3+
Y.; Ma, R. Supramol. Chem. 1996, 6, 353.
2
2
3 3
(7) Jeong, J. M.; Hong, M. K.; Chang, Y. S.; Lee, Y.-S.; Kim, Y. J.;
Cheon, G. J.; Lee, D. S.; Chung, J.-K.; Lee, M. C. J. Nucl. Med. 2008,
49, 830.
(8) Clarke, E. T.; Martell, A. E. Inorg. Chim. Acta 1991, 181, 273.
(9) (a) Eder, M.; Krivoshein, A. V.; Backer, M.; Backer, J. M.;
Haberkorn, U.; Eisenhut, M. Nucl. Med. Biol. 2010, 37, 405. (b) Liu,
Z.; Niu, G.; Wang, F.; Chen, X. Eur. J. Nucl. Med. Mol. Imaging 2009,
2
+
back the Na
orange indicator. Solutions of NOTA were prepared by dissolving
NOTA in 1 M NaNO and 0.2 M acetate buffer. Ga-citrate solutions
were made by mixing a Ga(NO ) solution with solutions of sodium
2 2
H EDTA with Zn at pH 5.8 in the presence of xylenol
H
3
3
3
3
citrate (10 mM, 20 mM, 40 mM) to give a finale Ga/citrate ratio of
1
0
.18:1, 2.45:1, and 4.75:1. The pH of all solutions was adjusted by adding
.1 or 1 M HCl and NaOH. The final acetate concentration in the kinetic
36, 1483. (c) Hanin, F.-X.; Pauwels, S.; Bol, A.; Breeman, W.; de Jong,
M.; Jamar, F. Nucl. Med. Biol. 2010, 37, 157. (d) Velikyan, I.; M €a cke, H.;
Langstrom, B. Bioconjugate Chem. 2008, 19, 569.
measurements was 0.1 M.
Kinetic Studies. The rates of Ga(NOTA) formation from Ga-
citrate species were measured at 25 °C and at an ionic strength of 1 M
(10) (a) Craig, A. S.; Parker, D.; Adams, H.; Bailey, N. A. J. Chem.
Soc., Chem. Commun. 1989, 1793. (b) Broan, C.; Cox, J. P.; Craig, A. S.;
Kataky, R.; Parker, D.; Harrison, A.; Randall, A. M.; Ferguson, G.
J. Chem. Soc., Perkin Trans. 2 1991, 87.
7
1
NaNO
3
.
Ga NMR spectra were recorded on a Bruker Avance 500
71
spectrometer ( Ga, 152.5 MHz, 11.75 T). Chemical shifts were
externally referenced to a 0.01 M Ga(NO3)3 solution in 0.1 M
(11) Kubí ꢁc ek, V.; Havlí ꢁc kov ꢀa , J.; Kotek, J.; Tircs ꢀo , G.; Hermann, P.;
HNO
were mixed in a 10 mm NMR tube wherein there was a reference (5 mm
NMR tube) with a Ga(NO solution in 0.1 M NaOH. The kinetic
3
; 0.0 ppm. Identical volumes of NOTA and Ga-citrate solutions
ꢀ
T ꢀo th, E.; Luke ꢁs , I. Inorg. Chem. 2010, 49, 10960.
12) Prata, M. I. M.; Santos, A. C.; Geraldes, C. F. G. C.; De Lima,
(
3
)
3
J. J. P. J. Inorg. Biochem. 2000, 79, 359.
(13) Claus ꢀe n, M.; O€ hman, L.-O.; Kubicki, J. D.; Persson, P. J. Chem.
Soc., Dalton Trans. 2002, 2559.
(14) Br €u cher, E., Sherry, D. Stability and Toxicity of Contrast
Agents. In The Chemistry of Contrast Agents in Medical Magnetic
Resonance Imaging; Merbach, A. E., T ꢀo th, E., Eds.; Wiley & Sons:
Chichester, U.K., 2001; p 243.
experiments were performed at a constant temperature of 25.0 °C
maintained by the NMR spectrometer and/or by a thermostatted bath.
The pH was checked in each sample after the kinetic experiment, and
even at the lowest pHs, where the acetate has no more buffer effect, its
maximum change did not exceed 0.05.
3+
The concentration of Ga was 4 mM in all experiments; the
concentration of citrate was 4.76, 9.8, or 19.0 mM, and the concentration
of NOTA varied between 5 and 40 mM. To monitor the formation of the
(
(
15) Br €u cher, E.; Sherry, A. D. Inorg. Chem. 1990, 29, 1555.
16) T ꢀo th, E.; Br €u cher, E.; Lazar, I.; T ꢀo th, I. Inorg. Chem. 1994,
ꢀ
71
33, 4070.
17) Balogh, E.; Tripier, R.; Ruloff, R.; T ꢀo th, E. Dalton Trans.
005, 1058.
18) Br €u cher, E.; Cortes, S.; Chavez, F.; Sherry, A. D. Inorg. Chem.
991, 30, 2092.
19) O’Brien, P.; Salacinksi, H.; Motevalli, M. J. Am. Chem. Soc.
997, 119, 12695.
20) Scientist for Windows, version 2.01; Micromath Inc.: Salt Lake
City, UT, 1995.
Ga(NOTA) complex, the intensity of the Ga NMR signal was
measured in 2 min intervals with respect to the signal of the reference.
All fits of the kinetic data were performed with the program Micromath
ꢀ
(
2
1
1
(
20
Scientist.
(
’
ASSOCIATED CONTENT
(
S
Supporting Information. Table of literature data for
b
3+
protonation and stability constants of species involving Ga ,
acetate, citrate, and NOTA. Species distribution diagrams. This
material is available free of charge via the Internet at http://pubs.
acs.org.
’
AUTHOR INFORMATION
Corresponding Author
*
2
E-mail: eva.jakabtoth@cnrs-orleans.fr. Phone: + 33 2 38 25 76
5. Fax: + 33 2 38 63 15 17.
1
0378
dx.doi.org/10.1021/ic201445e |Inorg. Chem. 2011, 50, 10371–10378