Detailed Mechanism of Phosphoanhydride Bond Hydrolysis Promoted by a Binuclear ZrIV-Substituted Keggin Polyoxometalate Elucidated by a Combination of 31P, 31P DOSY, and 31P EXSY NMR Spectroscopy

Article abstract of DOI:10.1021/acs.inorgchem.6b00385

A detailed reaction mechanism is proposed for the hydrolysis of the phosphoanhydride bonds in adenosine triphosphate (ATP) in the presence of the binuclear Zr^IV-substituted Keggin type polyoxometalate (Et₂NH₂)₈[{α-PW₁₁O₃₉Zr(μ-OH)(H₂O)}₂]·7H₂O (ZrK 2:2). The full reaction mechanism of ATP hydrolysis in the presence of ZrK 2:2 at pD 6.4 was elucidated by a combination of ³¹P, ³¹P DOSY, and ³¹P EXSY NMR spectroscopy, demonstrating the potential of these techniques for the analysis of complex reaction mixtures involving polyoxometalates (POMs). Two possible parallel reaction pathways were proposed on the basis of the observed reaction intermediates and final products. The 1D ³¹P and ³¹P DOSY spectra of a mixture of 20.0 mM ATP and 3.0 mM ZrK 2:2 at pD 6.4, measured immediately after sample preparation, evidenced the formation of two types of complexes, I1A and I1B, representing different binding modes between ATP and the Zr^IV-substituted Keggin type polyoxometalate (ZrK). Analysis of the NMR data shows that at pD 6.4 and 50 °C ATP hydrolysis in the presence of ZrK proceeds in a stepwise fashion. During the course of the hydrolytic reaction various products, including adenosine diphosphate (ADP), adenosine monophosphate (AMP), pyrophosphate (PP), and phosphate (P), were detected. In addition, intermediate species representing the complexes ADP/ZrK (I2) and PP/ZrK (I5) were identified and the potential formation of two other intermediates, AMP/ZrK (I3) and P/ZrK (I4), was demonstrated. ³¹P EXSY NMR spectra evidenced slow exchange between ATP and I1A, ADP and I2, and PP and I5, thus confirming the proposed reaction pathways.

Full text of DOI:10.1021/acs.inorgchem.6b00385

Article
pubs.acs.org/IC
Detailed Mechanism of Phosphoanhydride Bond Hydrolysis
IV
Promoted by a Binuclear Zr -Substituted Keggin Polyoxometalate
3
1
31
31
Elucidated by a Combination of P, P DOSY, and P EXSY NMR
Spectroscopy
†
,†,‡
†
,†
Thi Kim Nga Luong, Pavletta Shestakova,* Gregory Absillis, and Tatjana N. Parac-Vogt*
†
Laboratory of Bioinorganic Chemistry, Department of Chemistry, KU Leuven, Celestijnenlaan 200F, 3001 Leuven, Belgium
NMR Laboratory, Institute of Organic Chemistry with Centre of Phytochemistry Bulgarian Academy of Sciences, Acad. G. Bontchev
‡
Str., Bl.9, 1113 Sofia, Bulgaria
*
S Supporting Information
ABSTRACT: A detailed reaction mechanism is proposed for the hydrolysis of the
phosphoanhydride bonds in adenosine triphosphate (ATP) in the presence of the
IV
binuclear Zr -substituted Keggin type polyoxometalate (Et NH ) [{α-PW O Zr(μ-
2
2 8
11 39
OH)(H O)} ]·7H O (ZrK 2:2). The full reaction mechanism of ATP hydrolysis in
2
2
2
31
31
the presence of ZrK 2:2 at pD 6.4 was elucidated by a combination of P, P DOSY,
and 31P EXSY NMR spectroscopy, demonstrating the potential of these techniques
for the analysis of complex reaction mixtures involving polyoxometalates (POMs).
Two possible parallel reaction pathways were proposed on the basis of the observed
reaction intermediates and final products. The 1D 31P and P DOSY spectra of a
mixture of 20.0 mM ATP and 3.0 mM ZrK 2:2 at pD 6.4, measured immediately after
sample preparation, evidenced the formation of two types of complexes, I1A and I1B,
31
IV
representing different binding modes between ATP and the Zr -substituted Keggin
type polyoxometalate (ZrK). Analysis of the NMR data shows that at pD 6.4 and 50
°
C ATP hydrolysis in the presence of ZrK proceeds in a stepwise fashion. During the course of the hydrolytic reaction various
products, including adenosine diphosphate (ADP), adenosine monophosphate (AMP), pyrophosphate (PP), and phosphate (P),
were detected. In addition, intermediate species representing the complexes ADP/ZrK (I2) and PP/ZrK (I5) were identified and
31
the potential formation of two other intermediates, AMP/ZrK (I3) and P/ZrK (I4), was demonstrated. P EXSY NMR spectra
evidenced slow exchange between ATP and I1A, ADP and I2, and PP and I5, thus confirming the proposed reaction pathways.
7
INTRODUCTION
and a factor of 10−100 in the presence of metal ions. As a
result, natural phosphatase enzymes play an important role in
various biotechnology applications. For example, alkaline
phosphatases isolated from mammals are used to hydrolyze
or transphosphorylate numerous phosphate compounds in
■
Adenosine triphosphate (ATP, Figure 1) plays an important
role in energy transformation in all living organisms, as it is the
center for chemical energy storage through its triphosphate
1
−3
chain.
The hydrolysis of ATP to adenosine diphosphate
8
vitro. Alkaline phosphatases isolated from mouse intestine are
able to catalyze the synthesis of inorganic pyrophosphate from
inorganic phosphate during the hydrolysis of glucose 6-
phosphate, ATP, ADP, or p-nitrophenyl phosphate. Unfortu-
nately natural phosphatases with high purity are not easy to
obtain, and the enzymatic properties of these enzymes are
affected by activators and inhibitors such as amino acids and
(
ADP, Figure 1) and phosphate (P, Figure 1) is associated with
release of energy, which is an important driving force for
metabolic processes in living cells. For example, protein
4
9
degradation in animal and bacterial cells is dependent upon
ATP, and ATP hydrolysis is required for the degradation of
proteins to acid-soluble products. Moreover, ATP consumption
may promote the release of the polypeptide product from the
enzyme and, alternatively, ATP hydrolysis may permit the
translocation of the enzyme along the substrate to the next
10
metal ions in varying concentrations.
Alternatively, due to their selectivity and efficiency, artificial
phosphatases can be used for the hydrolysis of substrate
Such compounds can be synthesized
in large amounts, and their structures can be tuned and
5
cleavage site. In addition, ATP hydrolysis via highly efficient
11−13
phosphoester bonds.
metalloenzymatic ATPases also plays a central role in a variety
of processes including photosynthesis phosphorylation, oxida-
tive phosphorylation, and muscle action.
Uncatalyzed ATP hydrolysis in aqueous solution is a slow
14,15
6
designed depending on the specific purpose.
Moreover,
their use may help in understanding the precise role that active
process with a half-life between 10 and 100 days in the pH
7
range 1−12 at room temperature. However, the rate of ATP
Received: February 16, 2016
1
1
hydrolysis is increased by a factor of 10 via enzymatic catalysis
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Structures of adenosine triphosphate (ATP), adenosine diphosphate (ADP), adenosine monophosphate (AMP), triphosphate (PPP),
pyrophosphate (PP), and phosphate (P).
11
metal ions play in natural phosphatases. Several studies
demonstrated the important role of metal ions as well as their
complexes in ATP hydrolysis. For example, in the presence of
2
+
16
+
2+
7
III
17
Cu , VO2 and VO and oxidants, and the Co complex
[
III
tn Co (OH ) ] (tn = trimethylenediamine) the rate of ATP
2 2 2
4
5
hydrolysis increases up to factors of 530, 10 , and 10 ,
respectively, in comparison with pure ATP hydrolysis. The
increase in ATP hydrolysis reaction rate in the presence of
metal centers can be explained by various factors, including an
increase in the reactivity of the phosphorus centers toward the
incoming nucleophile due to the electron-withdrawing
character of the metal centers, activation and/or delivery of
the nucleophile, and/or stability of the pentacoordination
Figure 2. Structures of ZrK 2:2 (A), ZrK 1:2 (B), and ZrK 1:1 (C).
The WO groups are represented by blue octahedra, while the internal
6
IV
PO4 groups are represented by orange tetrahedra. Zr , H O
2
molecules, and OH groups are represented by violet, red, and dark
red balls, respectively.
29
(
HPNP) with a 530-fold rate enhancement in comparison
17−19
to spontaneous hydrolysis.
transition state and the leaving group.
On the one hand, the aim of the present study is to further
explore the reactivity of ZrK 2:2 toward the hydrolysis of
phosphoanhydride bonds in the biologically relevant substrate
ATP, which has a more complex structure in comparison with
previously investigated model DNA substrates (NPP, BNPP).
On the other hand, this study aims at fully elucidating the
Recent studies in our group demonstrated unprecedented
phosphoesterase activity of several metal-substituted polyox-
15,20−22
ometalates (POMs).
POMs are a large class of inorganic
oxo clusters that contain early transition metals such as V, Nb,
Ta, Mo, and W in their highest oxidation state. They have
promising applications in different research domains, including
21,22
mechanism of this hydrolytic reaction at a molecular level.
2
3
24,25
26,27
medicine, materials science,
and catalysis.
Part of
31
31
For this purpose we used a combination of P, P DOSY, and
P EXSY NMR spectroscopy to monitor the interaction
their structure can be removed, resulting in lacunary POM
structures that can act as ligands for various metal ions, such as
31
between the POM and the substrate as well as to identify all
intermediates formed during the course of the ATP hydrolytic
reaction. Diffusion ordered NMR spectroscopy (DOSY) is a
powerful technique to study both the presence of the different
compounds during the course of complex reactions and
interactions at the molecular and supramolecular level. DOSY
spectra are two-dimensional maps correlating chemical shifts of
the signals along the horizontal dimension and the translational
diffusion coefficients of the corresponding species along the
vertical dimension. Since the diffusion coefficient reflects the
size of the species, different species in complex reaction
mixtures can be clearly distinguished on the basis of their
diffusion properties, even if their signals overlap along the
chemical shift dimension due to their structural similarity.
2
+
2+
2+ 28
Zn , Co , and Cu , which are also found in metal-
IV
loenzymes. We have previously shown that a number of Zr -
substituted POM complexes can act as artificial phosphoses-
1
5,21,22
IV
terases.
Zr is an ideal active center due to its high Lewis
acidity and oxophilic properties, which facilitate both the
coordination/activation of the substrate/nucleophile and the
IV
hydrolysis of the phosphoester bond(s). Among Zr -
IV
substituted POMs, the mononuclear Zr -substituted Wells−
Dawson type POM K H[Zr(α -P W O ) ]·25H O was
1
5
2
2
17 61
2
2
demonstrated to be able to catalytically enhance the hydrolysis
rate of the DNA model compound 4-nitrophenyl phosphate
1
5
(
NPP) by nearly 2 orders of magnitude. In our previous study
IV
we also found that the binuclear Zr -substituted Keggin type
1
POM (Et NH ) [{α-PW O Zr(μ-OH)(H O)} ]·7H O (ZrK
Recently, our research group has successfully applied H and
2
2 8
11 39
2
2
2
3
1
2:2, Figure 2A) efficiently promoted hydrolysis of the extremely
P DOSY for the detailed characterization of nanosized
30
stable DNA model substrate bis-4-nitrophenyl phosphate
organic−inorganic POM-based hybrids and for the study of
different POM species present in aqueous solution.
21
22,31 31
(
BNPP) with a 320-fold rate enhancement and the RNA
P
model substrate 2-hydroxypropyl-4-nitrophenyl phosphate
exchange spectroscopy (EXSY) was applied to reveal the
B
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
reaction pathways and to gain additional information on the
slow dynamic equilibrium between the reaction products and
their complexes with the POM. In addition, several comparative
experiments were performed to evidence the interaction
between ZrK 2:2 and ADP, AMP, P, and PP (Figure 1)
formed during ATP hydrolysis. This integrated approach
enabled us to propose a detailed reaction mechanism of ATP
hydrolysis in the presence of ZrK 2:2.
RESULTS AND DISCUSSION
■
Interaction between ATP and ZrK 2:2. The aqueous
solution speciation of ZrK 2:2 under the reaction conditions
used in this study (pD 6.4 and 50 °C) was described in detail in
our previous studies and will be only briefly presented here for
21,22
the purpose of the following discussions.
Depending on the
reaction conditions such as pD, concentration, temperature,
and ionic strength, ZrK 2:2 can convert into the less active
1
0−
species [Zr(PW O ) ]
(ZrK 1:2, Figure 2B) and/or the
more active monomeric Keggin POM species [α-PW O Zr-
Figure 3. 31P NMR spectra of the hydrolysis reaction: (a) 3.0 mM ZrK
2:2; (b) 20.0 mM ATP; (c−g) 20.0 mM ATP in the presence of 3.0
mM ZrK 2:2 at pD 6.4 measured (c) after pD adjustment and after (d)
3 days, (e) 20 days, (f) 27 days, and (g) 54 days at 50 °C. Conditions:
11
39 2
1
1
39
4
−
21,22
(
μ-OH)(H O)] (ZrK 1:1, Figure 2C).
ZrK 1:1 is
2
concluded to be more catalytically active in comparison to
ZrK 2:2, as its Zr ion has more coordinated water molecules
IV
600 MHz, D O, 298 K, NS = 128.
2
that can be replaced by the substrate or that can act as a
nucleophile. Static DFT calculations showed that, depending
on pH conditions, the two most stable forms of this monomeric
Scheme 1. General Hydrolytic Pattern of ATP in the
Presence of ZrK 2:2
3−
species are the trihydrated form [α-PW O Zr(H O) ] and
1
1
39
2
3
the hydroxomonohydrated form [α-PW O Zr(μ-OH)-
1
1
39
4−
32
(
H O)] .
2
More importantly, in a recent study we have demonstrated
that a monomeric Keggin POM species (ZrK 1:1) is present in
an aqueous solution of ZrK 2:2 at nearly neutral pD and is
responsible for the hydrolysis of the phosphoester bond in
2
2
29
BNPP and HPNP. Theoretical studies showed that, when
ZrK 1:1 interacts with NPP or BNPP, it forms monodentate
complexes which are more stable than the corresponding
22
bidentate complexes. ZrK 1:2 is considered to be less active in
comparison to ZrK 2:2 since its Zr atom is coordinatively
saturated and hardly accessible for the bulky ATP ligand, as it is
sandwiched by the two Keggin moieties. Following our
2
1,22
previous findings,
we consider that the catalytically active
POM species responsible for ATP hydrolysis is the monomeric
ZrK 1:1 form, which will be abbreviated as ZrK for simplicity.
With the aim of investigating the interaction between ATP
and ZrK 2:2, a mixture of 20.0 mM ATP and 3.0 mM ZrK 2:2
at pD 6.4 was incubated at 50 °C. The interaction between
ATP and ZrK was evidenced by the characteristic changes
all signals that appeared and disappeared in 31P spectra during
the course of the reaction (Figure S2 in the Supporting
Information and Figure 3). We suggest that the reaction can
proceed via two parallel pathways, in which the first step is the
formation of the two ATP/ZrK complexes I1A (pathway A)
and I1B (pathway B). These two ATP/ZrK complexes result
from two possible binding modes between ATP and POM. In
complex I1A the POM interacts with the terminal phosphate
1
31
observed in H and P NMR spectra measured at different
time intervals (Figures S1 and S2 in the Supporting
1
Information, respectively). H NMR spectra showed broad
overlapping signals of the various species present in the
reaction mixture during the course of the ATP hydrolysis and
were therefore not useful for a detailed analysis of the reaction
mechanism. 31P NMR spectra, however, were much more
suitable for this purpose, due to the distinct difference in
chemical shifts of the signals. In order to ensure a good signal-
33
group P of ATP, while in complex I1B the POM binds to the
γ
intermediate phosphate group P . We suggest that the
β
3
1
to-noise ratio for all signals in P spectra, the concentration of
ATP was 7 times higher than the concentration of ZrK 2:2.
Some representative spectra demonstrating the changes in the
interaction of the POM and the first phosphate group P is
α
not favorable due to the steric repulsion between the bulky
POM and the adenosine residue attached to P . If the reaction
α
3
1
P spectra of the reaction mixture are shown in Figure 3.
proceeds via pathway A, ATP will be first hydrolyzed to P and
ADP, which then can subsequently interact with the POM,
forming intermediate I2. In the next step ADP will be further
hydrolyzed to AMP and a second P. In contrast, in pathway B,
ATP will be hydrolyzed first to AMP and PP. The formation of
The general reaction mechanism of ATP hydrolysis in the
presence of ZrK 2:2 and the expected reaction products and
intermediates is presented in Scheme 1. The proposed reaction
mechanism is based on careful analysis and full assignment of
C
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 4. 31P DOSY NMR spectrum (A) and 31P EXSY NMR spectrum (B) of 20.0 mM of ATP in the presence of 3.0 mM of ZrK 2:2 after
adjusting to pD 6.4 with no heating.
complexes AMP/ZrK (I3), P/ZrK (I4), and PP/ZrK (I5) was
also expected.
result of the binding to the ATP. Since a high excess of the ATP
was used, all POM should be bound to the substrate. Therefore,
we proposed that the new peak at −13.5 ppm is the peak of the
phosphorus atom from the ZrK bound to ATP in the ATP/ZrK
complex (I1A and I1B).
Analysis and Signal Assignment of 31P Spectra in
Support of the Proposed Reaction Mechanism. Figure 3b
shows that the 31P spectrum of a solution of pure ATP at pD
6
.4 is characterized by resonances at −9.32 ppm for P , −10.51
With the aim of confirming the presence of I1A and I1B, we
measured the P DOSY NMR spectrum of the reaction
γ
31
31
ppm for P , and −22.08 ppm for P . Comparison of the
P
α
β
spectra of pure ATP and an ATP/POM mixture measured
immediately after pD adjustment without heating (Figure 3c)
shows that in the presence of POM small changes in the
chemical shifts of all ATP signals (δ −9.23, 10.53, and −22.13
ppm) and a noticeable broadening of the line width of all
signals of the free ATP substrate occurs. This broadening
mixture after sample preparation (Figure 4A). The spectrum
shows that there are three main species with different diffusion
coefficients. The faster-diffusing species (δ −9.23, −10.53, and
−
10
2 −1
−22.13 ppm) with a diffusion coefficient of 4.37 × 10 m s
was assigned to the free unreacted ATP. The diffusion
31
coefficient of the species with a P chemical shift of −13.5
ppm was determined to be 2.69 × 10− m s . The diffusion
coefficient measured from the DOSY peaks at the chemical
shifts of the new signals that appeared after the addition of the
POM and assigned to I1A (δ −9.64, −12.00, and −17.99 ppm)
is identical with the diffusion coefficient of the species
characterized with a 31P peak at −13.5 ppm. This result
indicates that ATP binds to ZrK, since the diffusion of the new
species, assigned as an intermediate I1A, is significantly slower
than that of ATP.
10
2 −1
(11.97 Hz) is most pronounced for the signal of the terminal
phosphate group P . In addition, the appearance of a second set
γ
of signals was observed immediately after sample preparation at
−
9.64, −12.00, and −17.99 ppm, which were assigned to the
complex I1A between ATP and POM. We suggest that the
complex I1A between the POM and the terminal phosphate
group P of the ATP is more stable in comparison to the
γ
complex I1B between the POM and the phosphate group P in
β
the middle of the phosphate chain of the ATP, which may
33
increase the negative charge on P in the transition state and
A careful analysis of the diffusion peak of the broad
β
is sterically more hindered. This is why we assume that ATP
hydrolysis proceeds mainly through reaction pathway A. The
P spectra (Figure 3 and Figure S2 in the Supporting
Information) show that, during the course of the reaction,
the percentage of ATP as well as I1A decreased significantly
and the signal of ATP disappeared after 11 days, while the
signal of I1A disappeared completely after 20 days.
resonance at −9.23 ppm, assigned to P of the free ATP,
γ
showed that this signal has two components along the diffusion
dimension (Figure S3 in the Supporting Information). The
diffusion coefficient of the first component is identical with
those of the free ATP. The second component has a lower
3
1
diffusion coefficient of 3.24 × 10− m s , which however is
higher in comparison to the diffusion coefficient of the POM.
We suggest that this component most probably represents the
ATP involved in the less stable intermediate I1B, which exists
in a dynamic equilibrium with the free ATP. In this case its
apparent diffusion coefficient will be a population-weighted
average of the ATP diffusion coefficients in the bound and free
10
2
−1
Figure 3a shows that after pD adjustment ZrK 2:2 (δ −12.89
ppm) partially converts into a small amount of ZrK 1:2 (δ
2
1
−
13.98 and −14.04 ppm). In the presence of 20.0 mM ATP
at pD 6.4 the signal of ZrK 2:2 completely shifted to −13.5
ppm. This change in chemical shift of the POM could be a
2
1
D
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
states. This suggestion also explains the observed broadening of
the ATP signals upon addition of the POM resulting from the
exchange process and the partial overlap of the ATP and I1B
resonances. Figure S3 shows the expanded areas of the
spectrum of the reaction mixture, where the presence of
additional low-intensity signals partially overlapping with the
main ATP resonances is clearly seen. The relatively high value
of the diffusion coefficient of I1B and the very small chemical
shift differences of the signals of free ATP and I1B complex
indicate that the equilibrium is shifted toward the free ATP due
to the low stability of intermediate I1B.
The presence of a dynamic equilibrium between ATP and
3
1
the intermediates I1A and I1B was further investigated by
P
3
1
EXSY NMR. Figure 4B shows the P EXSY spectrum of a
mixture of 20.0 mM of ATP and 3.0 mM of ZrK 2:2, where the
exchange cross-peaks originating from a slow exchange between
ATP and the more stable intermediate I1A are clearly seen.
Due to the partial overlap of ATP and I1B signals the cross-
peaks between these species are not detectable.
After incubation of the sample for 3 days at 50 °C, the
appearance of new signals in addition to those of ATP and I1
was detected (Figure 3d). The singlets at 2.24 and 1.14 ppm
were assigned to the reaction products AMP and P,
respectively, while the two broad doublets at −9.26 and
Figure 5. 31P DOSY NMR spectrum of 20.0 mM of ATP in the
presence of 3.0 mM of ZrK 2:2 after 3 days at 6.4 and 50 °C.
−10.36 ppm, which are partially overlapped with the signals of
P and P of the unreacted ATP, were assigned to ADP by
γ
α
3
1
comparison with the P spectra of pure AMP, P, and ADP
Figures S4−S6 in the Supporting Information). The broad
(
and −14.04 ppm characteristic for ZrK 1:2 showed that at this
stage of the reaction ZrK 2:2 has partially converted into its less
reactive form.
resonance at around −8.94 ppm originates from free PP formed
as a reaction product following pathway B, while the low-
intensity signal at around −7.25 ppm was assigned to I5,
representing the complex between the ZrK and the PP, as
demonstrated by comparison with the reference spectra of pure
PP and of PP in the presence of ZrK 2:2 (Figure S7 in the
Supporting Information). In addition, two doublets with very
low intensity at −7.65 and −11.20 ppm were also observed. We
suggest that these signals originate from intermediate I2,
After incubation of the reaction mixture for 27 days at 50 °C,
ATP and I1 completely disappeared and their signals were no
31
longer observed in the P spectrum (Figure 3f). At this stage of
the reaction the two doublets at −9.26 and −10.36 ppm
assigned to free ADP as well as the pair of doublets of the
intermediate I2 between ADP and ZrK at −7.65 and −11.20
ppm were clearly seen. The longer incubation time resulted also
in an increase in the amount of I5 (the complex between PP
and ZrK) in the reaction mixture, as evidenced by the higher
intensity of the signal at −7.25 ppm in comparison to its
intensity after 3 days of heating at 50 °C (Figure 3d). During
the course of the reaction the intensity of ADP increased,
reaching a maximum after 7 days, and then gradually decreased
(Figure 3 and Figure S2 in the Supporting Information). The
disappearance of ATP and I1 and the higher signal intensity of
ADP and I2 at this stage of the reaction allowed better
resolution and sensitivity in the 31P DOSY spectrum of the
reaction mixture (Figure 6), which gave further evidence for the
formation of ADP and I2. The results show that the diffusion
3
1
representing a complex between ADP and ZrK. The P DOSY
NMR spectrum in Figure 5 provides further evidence for the
appearance of AMP, P, and I5. We can see from Figure 5 that P
is the smallest species, represented by the highest diffusion
−
10
2 −1
coefficient of 7.59 × 10 m s . The second species with the
diffusion coefficient of 4.67 × 10−10 m s was assigned as
AMP. The diffusion coefficient measured at the chemical shift
2
−1
−10
2 −1
of the signal assigned to I5 (2.67 × 10 m s ) is similar to
the diffusion coefficient of the ZrK and the I1A, thus
confirming that this resonance originates from the complex
between PP and ZrK. Due to the much larger molecular weight
of the POM the diffusion coefficients of all intermediates are
dominated by the POM diffusion coefficient. Unfortunately,
due to partial overlap of the ADP signals with the resonances of
P and P of the nonreacted ATP, a precise determination of
coefficient (2.75 × 10− m s ) of the species at δ −7.65 and
10
2 −1
−10
2
−11.20 ppm assigned as I2 is lower than that (4.63 × 10
m
−1
γ
α
s ) of the species at δ −9.26 and −10.36 ppm assigned as
ADP. The data are in full agreement with the chemical shifts
and diffusion coefficients determined in a comparative study of
a mixture of 20.0 mM ADP and 3.0 mM ZrK 2:2 (Figure S6 in
the Supporting Information). The chemical shifts and the
diffusion coefficients of the various species observed during the
course of the reaction are summarized in Table 1.
the ADP diffusion coefficient in this reaction mixture is
hampered. The simultaneous presence of ADP, AMP, P, and
PP in the reaction mixture indicates that the two pathways
shown in Scheme 1 for the hydrolysis of ATP in the presence of
ZrK 2:2 proceed in parallel.
With the advancement of the reaction after heating for 20
days at 50 °C (Figure 3e), the signals of ATP almost disappear,
while the intensity of I1A resonances significantly decreased. In
parallel to that, the intensity of the resonances assigned to ADP,
PP, AMP, and P as well as I5 and I2 complexes increased
The slow exchange between ADP and I2 in this reaction
31
mixture was observed by P EXSY NMR. Figure 7 shows the
expanded 31P EXSY spectrum in the region from −6 to −15
ppm, where the exchange cross-peaks between ADP and I2 are
clearly seen. The full 31P EXSY NMR spectra of the reaction
(Figure 3e). The appearance of a new set of signals at −13.98
E
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 7. 31P EXSY NMR spectrum of 20.0 mM ATP in the presence
of 3.0 mM ZrK 2:2 after 27 days at pD 6.4 and 50 °C.
Figure 6. 31P DOSY NMR spectrum of 20.0 mM of ATP in the
presence of 3.0 mM of ZrK 2:2 after 27 days at pD 6.4 and 50 °C.
monoester due to the absence of a leaving group such as p-
3
5
nitrophenyl, it is much more resistant to hydrolysis. With the
aim of providing more information about the interaction
Table 1. Chemical Shifts and Diffusion Coefficients of the
Various Species Observed during ATP Hydrolysis in the
Presence of ZrK 2:2
31
between AMP and ZrK 2:2, we recorded P NMR spectra of a
mixture of 20.0 mM AMP and 3.0 mM ZrK 2:2 in D O at pD
2
2
‑1
6.4 after pD adjustment and after different time increments at
species
chem shift (ppm)
diffusion coeff (m s )
5
0 °C (Figure S4 in the Supporting Information). The results
4.37 × 10− (±5.82 × 10
2.69 × 10− (±2.78 × 10
3.24 × 10− (±3.87 × 10
10
−13
−12
−12
)
)
)
ATP
I1A
I1B
−9.23, −10.53, −22.13
−9.64, −12.00, −17.99
−9.23
show that immediately after mixing three new resonances
appeared at −2.79, −13.20, and −13.28 ppm. We suggest that
the new resonances originate from a complex between AMP
and ZrK (I3). We suggest that the peak at −2.79 ppm and one
of the two peaks at around −13 ppm originate from the AMP/
ZrK complex I3, while the other peak at around −13 ppm is
eventually from the free ZrK. The intensity of these signals
gradually decreased when the incubation time increased, and
they completely disappear after 6 days at 50 °C, suggesting that
complex I3 was not stable and disintegrates back to AMP and
ZrK, which further converts to ZrK 1:2, as evidenced by the
two characteristic new peaks at −13.98 and −14.04 ppm.
Unfortunately, due to the very low intensity of I3 and ZrK it
10
10
4.63 × 10− (±7.49 × 10
2.75 × 10− (±4.25 × 10
10
−13
−12
)
)
ADP
I2
−9.26, −10.36
−7.65, −11.20
10
4.67 × 10− (±3.87 × 10−13
7.59 × 10− (±1.23 × 10−13
2.67 × 10− (±8.56 × 10−12
2.69 × 10− (±6.33 × 10−12
10
)
)
)
)
AMP
P
2.24
10
1.14
10
I5
−7.25
−13.5
10
ZrK 1:1
31
31
was not possible to measure P DOSY and P EXSY spectra
of this mixture. In addition to the observed new signals in the
presence of 3.0 mM of ZrK 2:2 the peak of the free AMP was
shifted upfield by 0.20 ppm and its half-width increased to
17.32 Hz. Both the change in chemical shift and the line
broadening of the signal of the free AMP indicated that it is
involved in dynamic equilibrium with the I3 complex. In a high
excess of the AMP substrate this equilibrium is shifted toward
the free AMP, as evidenced by the disappearance of the
resonances of I3 after 6 days. After 6 days of heating at 50 °C,
the signal of AMP was still very stable, suggesting that AMP
was not hydrolyzed.
mixture measured after heating at 50 °C for 20 and 27 days,
respectively, are given in Figures S8 and S9 in the Supporting
Information. Both spectra have similar cross-peak patterns.
Figure S2 in the Supporting Information shows that upon
further incubation at 50 °C the signals of ADP and I2 gradually
decreased and disappeared after 54 days (Figure 3g). In parallel
to that, the intensity of P and AMP signals increased during the
course of the reaction. After 54 days the only components
present in the mixture are the reaction products P and AMP as
well as I5. Upon further heating the intensity of I5 slowly
decreased, indicating that I5 slowly converted into P. The slow
conversion of I5 to P can be explained by the fact that the PP
Similarly, the potential formation of the complex I4 between
P and ZrK was confirmed by comparative studies of a mixture
of 20.0 mM Na HPO and 3.0 mM ZrK 2:2 (Figure S5 in the
IV
34
linkage in I5 is probably stabilized by the Zr ion of ZrK.
2
4
31
The analysis of P spectra in Figure 3 showed that no signals
indicative of the presence of I3 and I4 intermediates could be
found. We suggest that, since AMP is an unactivated phosphate
Supporting Information). In addition to the resonances of free
3
1
phosphate (1.14 ppm) and ZrK 2:2 (−12.89 ppm), the
P
spectrum of this mixture shows two weak signals at −2.32 and
F
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
13.28 ppm (Figure S5b). We suggest that these signals
Article
−
1:1 is more catalytically active in comparison to ZrK 2:2, since
its Zr ion has more coordinated water molecules that can be
IV
originate from the I4 complex and correspond to the
phosphorus atom of the phosphate (−2.32 ppm) and of the
ZrK (−13.28 ppm) involved in the I4 complex. These signals
completely disappear after 4 days at 50 °C, suggesting that
similarly to I3 this complex is not very stable and in excess
Na HPO it disintegrates to P and ZrK 1:2, as evidenced by the
replaced by the substrate or that can act as a nucleophile.
Therefore, we proposed that in this study ZrK 1:1 is also the
most active species and it is responsible for substrate
phosphoanhydride bond hydrolysis. When ATP is added to a
ZrK 2:2 solution under the same reaction conditions, ZrK 1:1
coordinates to the oxygen atom of a terminal and intermediate
phosphate group of ATP, forming intermediates I1A and I1B,
2
4
two characteristic new peaks at −13.98 and −14.04 ppm
(
Figure S5c). These results are in agreement with the
observations in the comparative study of a 20.0 mM AMP
and 3.0 mM ZrK 2:2 reaction mixture, where the excess AMP
favored the release of free substrate from I3. To further assess
the influence of sample composition on the stability of I4, we
investigated an equimolar mixture of 6.0 mM Na HPO and 6.0
mM ZrK 2:2. The P spectrum (Figure S5d) shows that in this
case the equilibrium is shifted toward the I4 complex, as
evidenced by a higher intensity of I4 signals in comparison to
the signals of free P and ZrK 2:2.
respectively. The coordination of ZrK 1:1 to P is favored, since
it is suggested to have the catalytic function of stabilizing the
γ
development of negative charge on the γ-phosphoryl oxygen of
33,36
ATP.
This coordination results in more positive charge at
the P atom, thus activating it toward nucleophilic attack. In I1A
2
4
γ
31
the OH group or the coordinated water of ZrK 1:1 can play the
role of a nucleophile, attacking the phosphorus atom and
leading to P−O bond cleavage and formation of I4 and ADP
following pathway A. In the next reaction step along this
pathway ADP binds to ZrK 1:1 in a similar way, forming
intermediate I2. The intramolecular attack of the OH group of
ZrK 1:1 to the phosphorus atom also leads to P−O bond
cleavage, resulting in the reaction products AMP and P. The
parallel hydrolytic pathway B also occurs, where ZrK 1:1
The observations based on the comparative studies of AMP/
ZrK 2:2 and Na HPO /ZrK 2:2 mixtures described above
2
4
could explain why I3 and I4 were not observed during the ATP
hydrolysis, as evidenced by the lack of signals at −2.79 ppm
3
1
(
I3) and −2.32 ppm (I4) in the P spectra of the 20.0 mM
ATP and 3.0 mM ZrK 2:2 reaction mixture (Figure 3). Under
the reaction conditions their analytical concentration is very
low, since they quickly convert to AMP and P, as demonstrated
by the comparative studies of AMP/ZrK 2:2 and Na HPO /
interacts with the P phosphorus atom of ATP in an identical
β
way, forming the intermediate I1B and then leading to the
formation of I5 and AMP. The formation of intermediates I3
and I4 is also possible; however, they are not stable under the
reaction conditions since the excess of AMP and P, whose
concentration constantly increases during the reaction, favors
their conversion back to AMP and P.
2
4
ZrK 2:2 mixtures.
The percentage of each species formed during the ATP
hydrolysis in the presence of ZrK 2:2 was determined by
3
1
integration of its respective P signal(s) and plotted as a
function of reaction time (Figure 8).
In general ZrK 1:1 can bind to the phosphate group of the
ATP substrate in a mono- or bidentate fashion. In our prevoius
study on the phosphoester bond hydrolysis in BNNP,
promoted by ZrK 2:2, we demostrated by DFT calculations
that the monodentate BNPP/ZrK 1:1 complex is more
22
favorable. The hydrolysis of the first phosphoester bond in
BNPP results in the formation of NPP, which further interacts
31
with ZrK 1:1. P EXSY NMR spectra showed that the
bidentate NPP/POM complex exists in a dynamic equilibrium
with the monobound and free NPP. We suggest that ATP
interacts with ZrK 1:1 in an identical way, resulting in both
mono- and bidentate coordination of the substrate to ZrK 1:1.
Unfortunately in this case the NMR data do not give sufficient
evidence to fully elucidate the type of the binding mode, since
apart from the broad resonance at −9.23 ppm for the I1B
which is however overlapped by the signal of P of the free
γ
ATP, we obseved only one set of NMR signals for the other
phosphorus atoms from the phosphoester chain in both bound
species. The reason could be the fast interconversion between
the two types of coordination modes with respect to the NMR
chemical shift time scale, resulting in one averaged set of
resonances.
Figure 8. Percentage of all species (ATP, I1A, ADP, I2, I5, AMP, and
P) formed during the course of the reaction as a function of reaction
time for the reaction between 20.0 mM of ATP and 3.0 mM of ZrK
During the course of ATP hydrolysis, no signal of
triphosphate (PPP, Figure 1) is detected, meaning that ATP
2
:2 at pD 6.4 and 50 °C.
cannot be immediately hydrolyzed at P to PPP, most probably
α
due to the steric hindrance of the adenosine residue attached to
Proposed Mechanism. On the basis of the above results, a
P .
α
detailed mechanism of ATP hydrolysis in the presence of ZrK
The proposed complex reaction mechanism of ATP
hydrolysis in the presence of ZrK 2:2 was further supported
by a comparative study of the interaction of ZrK 2:2 with ADP,
AMP, PP, and P used as substrates. This strategy facilitated the
analysis of the otherwise complex spectral patterns of ATP/ZrK
mixtures and allowed us to unequivocally assign all the reaction
21,22,29
2
:2 is proposed in Scheme 2.
these reaction conditions, there is an equilibrium among ZrK
:2, ZrK 1:1, and/or ZrK 1:2. ATP coordination to ZrK 1:2 is
As mentioned above under
2
not expected, since the Zr atom in ZrK 1:2 is coordinatively
saturated and hardly accessible to the bulky ATP ligand. ZrK
G
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a
Scheme 2. Proposed Reaction Mechanism for the Hydrolysis of ATP by ZrK 2:2
a
In this scheme, A in all compounds represents a ribose sugar and a nucleoside base adenosine part.
products and intermediates. The chemical shift fingerprint of
the new species observed during the hydrolysis of ATP with
those identified in the comparative reaction mixtures confirmed
that the reaction proceeds through the proposed reaction steps
and pathways. The P spectra of the comparative mixtures are
presented and described in detail in Figures S4−S7 and S10−
S12 in the Supporting Information.
Kinetic Study. The full assignment of all species that occur
during the course of ATP hydrolysis in the presence of ZrK 2:2
facilitated a kinetic study. In order to evaluate the reactivity of
ZrK 2:2 toward ATP hydrolysis, the observed rate constants of
ATP (20.0 mM) hydrolysis in the presence of 3.0 mM ZrK 2:2
at pD 6.4 and 50 °C were determined. Since the hydrolysis of
ATP in the presence of ZrK 2:2 is complicated due to the
formation of several intermediates and products, the calculation
of the rate constant of the hydrolysis was simplified by focusing
on the disappearance of both ATP and stable intermediate I1A
signals. The first-order rate constants were calculated from the
(pD 6.4 and 50 °C), the rate constant of HPNP hydrolysis
promoted by ZrCl O·8H O is 40-fold slower than that
2
2
IV
promoted by ZrK 2:2 due to the formation of insoluble Zr
29
hydroxyl polymeric gels.
3
1
To observe the influence of catalyst concentration on the
observed rate constant, as well as its influence on the different
intermediates formed in the course of hydrolytic reaction, the
rate constant of 3.0 mM ATP hydrolysis was measured in the
3
1
presence of 3.0 mM ZrK 2:2 at pD 6.4 and 50 °C. P NMR
spectra were recorded at different time intervals (Figure S14 in
the Supporting Information). As can be seen from Figure S14,
under these conditions no free ATP resonances were seen, but
the resonances of I1 (−9.64, −12.00, −17.99 ppm) were clearly
detected, suggesting that ATP completely converted into I1A.
In addition, the signal of free ZrK 2:2 at −12.89 ppm and of
ZrK at −13.5 ppm appeared simultaneously after pD
adjustment. Integration of the I1 31P resonances at different
time increments allowed the calculation of the ATP hydrolysis
simple expression −d[ATP] /dt = k [ATP] , where [ATP]
t
obs
t
t
rate constant. A linear fitting method (ln[ATP] = k t +
obs
represents the total concentration of all forms of ATP,
including nonreacted ATP and ATP, in the intermediate
ln[ATP] ) (Figure S15 in the Supporting Information), where
0
[ATP] = [I1] is the concentration of the ATP at time t, was
1
7
31
(
I1A). Integration of the ATP and I1A P NMR resonances
Figure S2 in the Supporting Information and Figure 8) at
used and the rate constant for ATP hydrolysis was calculated to
(
−5 −1
be (1.04 ± 0.11) × 10 s .
different time increments allowed the calculation of the ATP
hydrolysis rate constant. A linear fitting method (ln[ATP] =
The results from the kinetic experiments imply that the
relative ratio of the substrate and the POM influences the rate
of the ATP hydrolysis and the type of intermediates formed
during the course of the reaction. This finding is important in
view of the eventual practical applications of POMs as artificial
phosphatases, since it demonstrates that it is possible to control
the reaction rate by careful tuning of sample composition.
k t + ln[ATP] ; Figure S13 in the Supporting Information),
obs
0
where [ATP] = [ATP]nonreacted + [I1] is the concentration of
the ATP at time t, was used. The hydrolysis rate constant for
ATP hydrolysis was calculated to be (1.97 ± 0.17) × 10−
6
s
−1
at
pD 6.4 and 50 °C. We expected the rate constant of ATP
hydrolysis promoted by the Zr salt ZrCl O·8H O to be
4
+
2
2
slower than that promoted by ZrK 2:2. Unfortunately, a
comparison of these rate constants was impossible because the
activity of Zr4 in the absence of ligands could not be
determined due to the formation of precipitates. However,
our previous study showed that, under the same conditions
CONCLUSION
■
+
In this study we have proposed a detailed mechanism for ATP
37
31
31
hydrolysis in the presence of ZrK 2:2 elucidated by P,
P
DOSY, and 31P EXSY NMR spectroscopy. The results clearly
H
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
demonstrate that 31P DOSY NMR is a powerful technique to
determine in situ the presence of reaction intermediates (I1A,
I1B, I2, and I5) and products (ADP, AMP, P) formed during
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.6b00385.
■
*
S
3
1
the course of ATP hydrolysis. In addition, P EXSY NMR has
been proven to be a very useful technique to identify the
presence of slow exchange between the intermediates and the
reaction products (ATP, I1A and I1B; ADP, I2; PP, I5) and to
follow the reaction pathway. The assignment of all species that
occurred during ATP hydrolysis promoted by this POM
allowed calculation of the rate constants. The present study
31P,
31P DOSY, and 31P EXSY NMR spectra, the
interaction between ADP/AMP/P/PP and ZrK 2:2, the
interaction between the substrates structurally related to
ATP and ZrK 2:2, and ln[ATP] as a function of time
(PDF)
IV
further demonstrates the potential of Zr -substituted POMs as
15,21,22,29
artificial phosphatases
and contributes to the further
development of POMs as Lewis acid catalysts for the hydrolysis
AUTHOR INFORMATION
■
*
*
3
8−40
of biomolecules.
Understanding the detailed mechanism in
Corresponding Authors
P.S.: e-mail, psd@orgchm.bas.bg.
T.N.P.-V.: e-mail, tatjana.vogt@chem.kuleuven.be; web, www.
chem.kuleuven.be/lbc/.
this study encourages us to further exploit the hydrolytic
activity of this POM toward biologically active molecules such
as DNA/RNA fragments, pesticides (paraoxon, parathion), and
nerve agents (soman, sarin).
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
EXPERIMENTAL SECTION
■
IV
Materials. The binuclear Zr -substituted Keggin POMs
4
1,42
(
Et NH ) [{α-PW O Zr(μ-OH)(H O)} ]·7H O
(ZrK 2:2)
2
2
8
11 39
2
2
2
Notes
4
3
and (Et NH ) [Zr(PW O ) ]·7H O (ZrK 1:2) were synthesized
2
2
10
11 39
2
2
The authors declare no competing financial interest.
as described in the literature. Adenosine triphosphate (ATP),
adenosine diphosphate (ADP), adenosine monophosphate (AMP),
ACKNOWLEDGMENTS
sodium pyrophosphate (PP), Na HPO4 (P), ribose phosphate,
■
2
adenine, adenosine, DCl, and NaOD were purchased from Acros
and used without any further purification.
Sample Preparation. Solutions containing 20.0 mM of each
compound ATP, ADP, AMP, PP, and Na HPO and 3.0 mM of ZrK
T.N.P.-V. and P.S. (BOF + fellowship) thank KU Leuven for
financial support. T.K.N.L. thanks the Vietnamese Government
and KU Leuven for a doctoral fellowship. G.A. thanks FWO
Flanders for a postdoctoral fellowship. The authors acknowl-
edge the CMST COST Action CM1203 (Polyoxometalate
Chemistry for Molecular Nanoscience) for financial support in
terms of STSM applications.
2
4
2
:2 were prepared in D O for the interaction study. An additional
2
solution containing 3.0 mM ATP and 3.0 mM ZrK 2:2 was prepared in
D O for the kinetic study. The final pD of the solution was adjusted
2
with minor amounts of concentrated DCl (10%) and NaOD (15%).
The pH meter value was corrected by using the equation: pD = (pH
44
meter reading) + 0.41.
REFERENCES
■
NMR Measurements. 31P 1D NMR spectra were recorded on a
Bruker Avance 400 spectrometer. Trimethyl phosphate was used as
the 31P chemical shift external reference. The P DOSY and EXSY
spectra were measured on a Bruker Avance II+ 600 NMR
spectrometer using a 5 mm direct detection dual broad band probe,
with a gradient coil delivering a maximum gradient strength of 53 G/
cm. All experiments were performed at a temperature of 298 K. The
(
1) Bazzicalupi, C.; Bencini, A.; Bianchi, A.; Danesi, A.; Giorgi, C.;
Lodeiro, C.; Pina, F.; Santarelli, S.; Valtancoli, B. Chem. Commun.
31
2
005, 2630.
(2) Massoud, S. S.; Milburn, R. M. J. Inorg. Biochem. 1990, 39, 337.
(3) Guo, Y.; Ge, Q.; Lin, H.; Lin, H. K.; Zhu, S.; Zhou, C. Biophys.
Chem. 2003, 105, 119.
(4) Manyanga, A. S.; Sithole, A. Nucleic acids structure & function for
2
D EXSY spectra were recorded with the Bruker program noesygpph.
generalbiochemistry, biology and biotechnology; Science Math Zone:
The spectra were recorded with a spectral width of 25000 Hz, 2 K data
points in the t2 time domain and 256 t1 increments with 64 transients
each, and a relaxation delay of 3 s. The spectra were recorded for each
sample with two different mixing times of 300 and 400 ms. A sine
window function (ssb = 2) and zero-filling were applied in both
dimensions prior to Fourier transformation, to give a 4K × 4K data
matrix in the frequency domain. The P DOSY spectra were
measured using a stimulated echo based pulse sequence, with bipolar
sine shaped gradient pulses. The spectra were acquired with 64K time
domain data points in the t2 dimension, 32 gradient strength
increments, a diffusion delay of 200 ms, a total gradient pulse length of
Boston, 2014; p 48.
(
(
5) Edmunds, T.; Goldberg, A. L. J. Cell. Biochem. 1986, 32, 187.
6) Ge, R.; Lin, H.; Xu, X.; Sun, X.; Lin, H.; Zhu, S.; Ji, B.; Li, F.; Wu,
H. J. Inorg. Biochem. 2004, 98, 917.
7) Imamura, T.; Hinton, D. M.; Belford, R. L.; Gumport, R. I.;
Haight, G. P., Jr J. Inorg. Biochem. 1979, 11, 241.
(
31
(
8) Millan
́
, J. L. Purinergic Signalling 2006, 2, 335.
(9) Nayudu, R. V.; de Meis, L. FEBS Lett. 1989, 255, 163.
(10) Spencer, A. G. The Scientific Basis of Medicine: Annual Reviews
University of London; Athlone Press, 1964; p 337.
(11) Hegg, E. L.; Burstyn, J. N. Coord. Chem. Rev. 1998, 173, 133.
(12) Kitamura, Y.; Komiyama, M. Nucleic Acids Res. 2002, 30, 102e.
(13) Fang, Y.-G.; Zhang, J.; Chen, S.-Y.; Jiang, N.; Lin, H.-H.; Zhang,
Y.; Yu, X.-Q. Bioorg. Med. Chem. 2007, 15, 696.
(14) Thomas, D. T. In Metal-DNA Chemistry; American Chemical
Society: Washington, DC, 1989; Vol. 402, p 1.
(15) Vanhaecht, S.; Absillis, G.; Parac-Vogt, T. N. Dalton Transactions
2012, 41, 10028.
7
ms, 128 transients for each gradient step, and a relaxation delay of 6
ms. The gradient strength was incremented from 2 to 80% of the
maximum gradient output (from 0.68 to 27.25 G/cm). The spectra
were processed with an exponential window function (line broadening
factor 5), 64K data points in F2, and 258 data points in the diffusion
dimension. The evaluation of the diffusion coefficients was performed
by fitting the diffusion profile (the normalized signal intensity as a
function of the gradient strength G) at the chemical shift of each signal
in the DOSY spectrum with an exponential function using a variant of
the Stejskal−Tanner equation adapted to the particular pulse sequence
used.
(16) Schneider, P. W.; Brintzinger, H. Helv. Chim. Acta 1964, 47,
1717.
(17) Hediger, M.; Milburn, R. M. J. Inorg. Biochem. 1982, 16, 165.
(18) Wilcox, D. E. Chem. Rev. 1996, 96, 2435.
I
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
19) Strat
Chem., Int. Ed. Engl. 1996, 35, 2024.
20) Absillis, G.; Cartuyvels, E.; Van Deun, R.; Parac-Vogt, T. N. J.
Am. Chem. Soc. 2008, 130, 17400.
21) Luong, T. K. N.; Absillis, G.; Shestakova, P.; Parac-Vogt, T. N.
Eur. J. Inorg. Chem. 2014, 2014, 5276.
22) Luong, T. K. N.; Shestakova, P.; Mihaylov, T. T.; Absillis, G.;
Pierloot, K.; Parac-Vogt, T. N. Chem. - Eur. J. 2015, 21, 4428.
23) Stephan, H.; Kubeil, M.; Emmerling, F.; Muller, C. E. Eur. J.
Inorg. Chem. 2013, 2013, 1585.
̈
er, N.; Lipscomb, W. N.; Klabunde, T.; Krebs, B. Angew.
(
(
(
(
̈
(
24) Carraro, M.; Gross, S. Materials 2014, 7, 3956.
(
25) Proust, A.; Matt, B.; Villanneau, R.; Guillemot, G.; Gouzerh, P.;
Izzet, G. Chem. Soc. Rev. 2012, 41, 7605.
26) Izarova, N. V.; Pope, M. T.; Kortz, U. Angew. Chem., Int. Ed.
012, 51, 9492.
27) Sartorel, A.; Bonchio, M.; Campagna, S.; Scandola, F. Chem. Soc.
Rev. 2013, 42, 2262.
28) Sigel, H. Metal ions in biological systems; Marcel Dekker: New
York, 1983; Vol. 15; pp 5−138.
29) Luong, T. K. N.; Absillis, G.; Shestakova, P.; Parac-Vogt, T. N.
Dalton Transactions 2015, 44, 15690.
30) Shestakova, P.; Absillis, G.; Martin-Martinez, F. J.; De Proft, F.;
Willem, R.; Parac-Vogt, T. N. Chem. - Eur. J. 2014, 20, 5258.
31) Stroobants, K.; Absillis, G.; Shestakova, P.; Willem, R.; Parac-
Vogt, T. J. Cluster Sci. 2014, 25, 855.
32) Jimenez-Lozano, P.; Carbo, J. J.; Chaumont, A.; Poblet, J. M.;
Rodríguez-Fortea, A.; Wipff, G. Inorg. Chem. 2014, 53, 778.
(
2
(
(
(
(
(
(
́
́
(
(
33) Admiraal, S. J.; Herschlag, D. Chem. Biol. 1995, 2, 729.
34) Rainey, J. M.; Jones, M. M.; Lockhart, W. L. J. Inorg. Nucl. Chem.
1
964, 26, 1415.
(
35) Chin, J.; Banaszczyk, M. J. Am. Chem. Soc. 1989, 111, 4103.
(
(
(
36) Williams, N. H. J. Am. Chem. Soc. 2000, 122, 12023.
̈
37) Jagoda, M.; Kramer, R. Inorg. Chem. Commun. 2005, 8, 697.
38) Stroobants, K.; Absillis, G.; Moelants, E.; Proost, P.; Parac-Vogt,
T. N. Chem. - Eur. J. 2014, 20, 3894.
39) Stroobants, K.; Goovaerts, V.; Absillis, G.; Bruylants, G.;
Moelants, E.; Proost, P.; Parac-Vogt, T. N. Chem. - Eur. J. 2014, 20,
567.
40) Ly, H. G. T.; Absillis, G.; Janssens, R.; Proost, P.; Parac-Vogt, T.
N. Angew. Chem., Int. Ed. 2015, 54, 7391.
41) Nomiya, K.; Saku, Y.; Yamada, S.; Takahashi, W.; Sekiya, H.;
Shinohara, A.; Ishimaru, M.; Sakai, Y. Dalton Transactions 2009, 5504.
42) Ly, H. G. T.; Absillis, G.; Parac-Vogt, T. N. Dalton Transactions
013, 42, 10929.
43) Kato, C. N.; Shinohara, A.; Hayashi, K.; Nomiya, K. Inorg. Chem.
006, 45, 8108.
44) Glasoe, P. K.; Long, F. A. J. Phys. Chem. 1960, 64, 188.
(
9
(
(
(
2
(
2
(
J
DOI: 10.1021/acs.inorgchem.6b00385
Inorg. Chem. XXXX, XXX, XXX−XXX