Polyoxomolybdate promoted hydrolysis of a DNA-model phosphoester studied by NMR and EXAFS spectroscopy

Article abstract of DOI:10.1021/ic201498u

Hydrolysis of (p-nitrophenyl)phosphate (NPP), a commonly used phosphatase model substrate, was examined in molybdate solutions by means of ¹H, ³¹P, and ⁹⁵Mo NMR spectroscopy and Mo K-edge Extended X-ray Absorption Fine Str

Full text of DOI:10.1021/ic201498u

ARTICLE
pubs.acs.org/IC
Polyoxomolybdate Promoted Hydrolysis of a DNA-Model
Phosphoester Studied by NMR and EXAFS Spectroscopy
†
‡
,†
Gregory Absillis, Rik Van Deun, and Tatjana N. Parac-Vogt*
†
Department of Chemistry, Katholieke Universiteit Leuven, Celestijnenlaan 200F, B-3001 Leuven, Belgium
‡
Inorganic and Physical Chemistry Department, Universiteit Gent, Krijgslaan 281, Building S3, B-9000 Gent, Belgium
ABSTRACT:
Hydrolysis of (p-nitrophenyl)phosphate (NPP), a commonly used phosphatase model substrate, was examined in molybdate solutions
1
31
95
by means of H, P, and Mo NMR spectroscopy and Mo K-edge Extended X-ray Absorption Fine Structure (EXAFS) spectroscopy.
ꢁ
5
ꢁ1
At 50 °C and pD 5.1 the cleavage of the phosphoester bond in NPP proceeds with a rate constant of 2.73 ꢀ 10
s
representing an
acceleration of nearly 3 orders of magnitude as compared to the hydrolysis measured in the absence of molybdate. The pD dependence
of kobs exhibits a bell-shaped profile, with the fastest cleavage observed in solutions where [Mo O ] is the major species in solution.
Mixing of NPP and [Mo O ] resulted in formation of these two intermediate complexes that were detected by P NMR
6ꢁ
7
24
6ꢁ
31
7
24
31
31
spectroscopy. Complex A was characterized by a P NMR resonance at ꢁ4.27 ppm and complex B was characterized by a P NMR
resonance at ꢁ7.42 ppm. On the basis of the previous results from diffusion ordered NMR spectroscopy, performed with the
4
ꢁ
hydrolytically inactive substrate phenylphosphonate (PhP), the structure of these two complexes was deduced to be (NPP) Mo O
5 21
2
4ꢁ
(
complex A) and (NPP) Mo O (H O) (complex B). The pH studies point out that both complexes are hydrolytically active and
2 12 36 2 6
lead to the hydrolysis of phosphoester bond in NPP. The NMR spectra did not show evidence of any paramagnetic species, excluding
the possibility of Mo(VI) reduction to Mo(V), and indicating that the cleavage of the phosphomonoester bond is purely hydrolytic. The
95
Mo K-edge XANES region also did not show any sign of Mo(VI) to Mo(V) reduction during the hydrolytic reaction. Mo NMR and
6
ꢁ
Mo K-edge EXAFS spectra measured during different stages of the hydrolytic reaction showed a gradual disappearance of [Mo O ]
7
24
6ꢁ
duringthehydrolytic reactionand appearanceof[P Mo O ] , whichwasthe finalcomplexobserved at the endofhydrolytic reaction.
2
5 23
’
INTRODUCTION
Early transition metals such as V, Nb, Ta, Mo, and W in their
recalcitrant pancreatic cancer cells, and that its activity is
1
6ꢁ21
comparable to that of commercially available drugs.
Despite
the fact that the antitumor activity of polyoxomolybdate has been
well documented, its mode of action on a molecular level has not
been fully understood. Several studies have shown that polyox-
omolybdates interact with biological molecules containing
phosphate moieties, leading to the formation of a pentamolyb-
highest oxidation state are able to form negatively charged metal-
oxide clusters, commonly known as polyoxometalates (POMs).
This class of compounds is characterized by a broad variety of
chemical and physical properties and exhibits a wide structural
versatility. Accordingly, many applications in various research
domains, including medicine and catalysis have been reported.
It has been demonstrated that because of their strong acidity and
tunable redox properties, several heteropolyacids were used both
22ꢁ26
1ꢁ4
dodiphosphate type of structure.
In the presence of mo-
lybdate, the high energy phospoanhydride bonds in ATP are fully
27ꢁ29
hydrolyzed.
Several studies implicate the interaction of
5ꢁ9
molybdate with phosphate groups in ATP as one of the key
in homogeneous and heterogeneous catalysis.
decade there has been a growing interest in the biological activity
of POMs, especially after it was shown that many of them have
In the past
6
ꢁ 26ꢁ29
factors in the antitumor activity of [Mo O ]
.
7
24
Interestingly, although there is much literature data on the
interactions between polyoxomolybdates and phosphate moi-
eties, the interactions with monophosphoesters have been scar-
cely explored. The hydrolysis of monophosphates is commonly
described as one of the most important reactions in biological
systems. Monophosphates are involved in several essential
processes such as signal transduction, energy storage, and replica-
10ꢁ15
potent antiviral, antibacterial, and antitumor properties.
It
has been generally accepted that noncovalent binding influenced
by the size, shape, and charge density of polyoxometalates is the
main factor governing their biological activity, but on the other
hand, very little is known about the reactivity of POMs towards
biological molecules and their building blocks.
30
tion of genetic material. While most nucleases utilize metal ions
Among the POMs exhibiting biological activity, the polyoxo-
6
ꢁ
molybdate [Mo O ] has been extensively studied. A number
7
24
of in vivo studies have shown that it exhibits antitumor properties
Received: July 14, 2011
against a range of different cancer cell lines, including the highly
Published: October 26, 2011
r 2011 American Chemical Society
11552
dx.doi.org/10.1021/ic201498u
|Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
1
31
as cofactors, the exact role of the metal in the hydrolytic
mechanism isnot fully understood. Numerous model compounds
have been used to identify the possible mechanism by which metal
NMR Spectroscopy. H and P spectra were recorded on a Bruker
Avance 300 spectrometer and on a Bruker Avance 400 spectrometer.
O with 0.05 wt % 3-(trimethylsilyl) propionic acid as an internal
standard was used as a solvent. Trimethyl phosphate was used as a
ppm P reference. Mo NMR spectra were recorded on a Bruker
D
2
31ꢁ35
ions promote phosphoester bond hydrolysis.
With the
3
1
95
0
exception of the work by Yatsimirsky and his co-workers, in
which charge neutral lanthanide complexes have been implicated
Avance 600 (39 MHz) spectrometer. UVꢁVis absorption spectra have
36
been measured on a Varian Cary 5000 spectrophotometer.
as hydrolytically active species, the metal complexes that are
EXAFS Spectroscopy. Measurements were performed in trans-
mission mode using a Si(111) double crystal monochromator on the
Dutch-Belgian Beamline (DUBBLE, BM26A) at the European Synchro-
tron Radiation Facility (ESRF, Grenoble, France). Higher harmonics
were rejected by the Si mirrors themselves (suppression factor about
000). The molybdenum K-edge spectra were collected using Oxford
Instruments ionization chambers, filled with 30% Ar/70% He at 1 bar for
and pure Ar at 1.5 bar for I
were collected in equidistant k-steps of 0.05 Å across the EXAFS region.
A Mo metal foil (first inflection point at 19999 eV) was used for energy
calibration. As the Mo phase function is strongly nonlinear at k-values
below k = 5 Å , in a first approach, the filtering and back-transformation
procedure was made by using only k-values larger than k = 5 Å (k-range
active as catalysts for the hydrolysis of phosphoester bonds are
37ꢁ42
generally positively charged.
This seems to be an essential
property for electrostatic interaction with the negatively charged
phosphate group. The mechanism by which metal complexes
accelerate the rate of phosphate ester hydrolysis is generally
described as an interplay of several factors among which the
Lewis acid activation via coordination of phosphoryl oxygen(s) to
the metal ion, coordination of a nucleophile such as water or
hydroxide ligand to the metal, and leaving group activation via
coordination of the leaving group oxygen to the metal appear to
4
6a
1
I
0
t
at ambient temperature and pressure. Data
ꢁ1
30ꢁ33
be of crucial importance.
As a result, the metal complexes
ꢁ1
that act as artificial phosphoesterases have to meet a number of
criteria such as an overall positive charge, free coordination site for
the binding of the phosphoryl oxygen, the presence of coordi-
nated water or hydroxide, or the presence of functional groups for
acidꢁbase catalysis, among others.
ꢁ1
ꢁ1
was then 5ꢁ16.9 Å ), but this had only a minor effect on the structural
parameters obtained. Therefore, it was decided to use the whole k-range
ꢁ
1
from 2.6 to 16.9 Å in the final procedure. Solutions of the polyox-
omolybdates were pipetted out of the reaction vessel at the indicated
times. This vessel was kept at 50 °C, and the solutions were then left to
cool to room temperature (thereby slowing down the reaction to a near
standstill) and were sealed in polyethylene cuvets with a path length of
8 mm for the EXAFS measurements. The initial [MoO4] concentration
had been set to 700 mM and the concentration of NPP was 100 mM.
EXAFS data extraction and data fitting were performed using the program
EXAFSPAK. Theoretical phase and amplitude functions were calcu-
lated using FEFF 8.2 using the crystal structure from reference 58. The
amplitude reduction factor, S
Kinetics. In a typical kinetic experiment, the hydrolysis of NPP in the
Despite these requirements, our recent studies have revealed
6ꢁ
that [Mo O ] , which is a negatively charged and coordina-
7
24
tively saturated metal complex, lacking coordinated water or
hydroxide, effectively hydrolyses phosphodiester bonds in bis-
2ꢁ
(
p-nitrophenyl) phosphate (BNPP) and 2-hydroxypropyl-4-
nitrophenyl phosphate (HPNP), commonly used as DNA and
4
3,44
46b
RNA model compounds.
These kinetic studies strongly
6
ꢁ
47
suggest that [Mo O ] is the hydrolytically active complex
7
24
2
0
, was kept constant at 1 throughout the fit.
and that the cleavage occurs by a mechanism which is different to
that of other currently known hydrolytically active metal com-
plexes. The evidence suggested that the origin of hydrolytic
1
presence of varying concentrationsof molybdate wasfollowedby H NMR
6ꢁ
spectroscopy. ThepHofthesolution wasmeasuredinthe beginning andat
the end of the hydrolytic reaction, and the difference was typically less than
0.1 pH unit. The pD value of the solution was obtained by adding 0.41 to
the pH reading, according to formula pD = pH + 0.41. The reaction
samples were kept at constant temperature and the rate constants for the
hydrolysis were determined by following the appearance of the p-
nitrophenol resonances in the H NMR spectra at different time intervals.
The observed first order rate constants (kobs) were calculated by the
integral method from at least 90% conversion. For reactions performed in
activity of [Mo O ] toward phosphodiesters may lay in its
7
24
high internal lability and an intramolecular exchange process
43,44
which resultsin partial detachmentof oneMoO tetrahedron.
4
48
Because of the rapid exchange between the free and the bound
phosphodiester, the structure of the active complex could not be
elucidated by spectroscopic techniques. Further to this, the
interaction between polyoxomolybdates and (p-nitrophenyl)-
phosphate (NPP), a commonly used phosphatase model sub-
strate, has been investigated by using the analogous, but
hydrolytically inactive model substrate phenylphosphonate
1
2
H O solutions, the same kinetic method was used, and the reaction
products were detected by the H NMR water suppression technique.
1
4
ꢁ
(
PhP). Two distinct complexes [(PhP) Mo O ]
and
2
5 21
4ꢁ
[(PhP)Mo O ]
could be detected by diffusion ordered
7 25
’
RESULTS AND DISCUSSION
43
NMR spectroscopy (DOSY NMR). On the basis of the DOSY
4
ꢁ
NMR results with PhP, the complexes (NPP) Mo O
5 21
2
Hydrolysis of NPP by Polyoxomolybdate. Acidification of a
molybdate solution leads to the formation of isopolyoxomolyb-
4
ꢁ
complex A) and (NPP) Mo O (H O)
6
(
(complex B) have
2
12 36
2
been suggested as the active species leading to hydrolysis of the
phosphoester bond in the NPP substrate. In this study we further
employ kinetics experiments, multinuclear NMR, and Extended
X-ray Absorption Fine Structure (EXAFS) spectroscopy to fully
examine the hydrolysis of NPP in the presence of polyoxo-
molybdates and give a complete account on this novel reaction.
dates so that at any given pH below 6.5 the presence of two or
49
more species has been detected. The first cluster that is formed
6
ꢁ
in mildly acidic solutions is heptamolybdate [Mo O ] and its
7
24
protonated forms, while further acidification leads to the forma-
6ꢁ
tion of octamolybdate [Mo O ]
:
8
26
2
7
½MoO4ꢂ ꢁ þ 8Hþ h ½Mo7O24ꢂ6ꢁ þ 4H2O
’
EXPERIMENTAL SECTION
2
8
½MoO4ꢂ ꢁ þ 12Hþ h ½Mo8O26ꢂ4ꢁ þ 6H2O
Materials. Sodium molybdate, (p-nitrophenyl)phosphate, and phe-
nylphosphate were purchased from Acros and used without further
purification. The pH of the solutions for the NMR studies was adjusted
with D SO and NaOD, both from Acros.
The initial reaction between a solution containing 700 mM
2
ꢁ
[MoO
According to the thermodynamic model developed previously
]
and 100 mM NPP was performed at pD 5.10.
4
2
4
1
1553
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
Scheme 1. Hydrolytic Cleavage of NPP Promoted by
6
ꢁ
[
Mo O ]
7 24
4
ꢁ
Figure 3. Structure of complex A, (NPP) Mo O
5 21
and complex B,
2
4
ꢁ
NPP) Mo O (H O)
6
(
.
2
12 36
2
1
2ꢁ
Figure 1. H NMR spectra of the reaction between 700 mM [MoO ]
4
and 100 mM NPP at pD 5.10 and 50 °C after (a) 1 h, (b) 3 h, and (c) 24
h after mixing.
Figure 4. Fraction of NPP, complex A, complex B, phosphate, and
6
ꢁ
[
[
P
MoO
2
Mo
5
O
]
23
]
of a solution containing 100 mM NPP and 700 mM
2
ꢁ
31
4
at pD 5.2 followed by P NMR.
31
The P NMR spectra (Figure 2) recorded shortly upon
6ꢁ
mixing of NPP and [Mo O ] have shown the absence of free
7
24
NPP resonance at ꢁ5.95 ppm and appearance of two new
resonances at ꢁ4.27 ppm and ꢁ7.42 ppm. This indicates that
NPP was fully bound into two types of complexes designated A
and B, which were present in about a 9:1 ratio.
3
1
Figure 2. P NMR spectra recorded during the reaction of 100 mM
2
ꢁ
DOSY NMR spectroscopy suggested that complex A has a
NPP with 700 mM [MoO
after 2 h, (c) after 5 h, and (d) after 13 h.
4
]
at pD 5.1 recorded (a) upon mixing, (b)
4
ꢁ
formula (NPP) Mo O
, consisting of five MoO octahedra
2
5
21
6
coupled by one corner- and four edge-shared units to form a ring
structure, which is capped on either side by two phosphate
6
ꢁ
in our group, at this pD value [Mo O ] is the major species in
7
24
4
5
1
31
solution. During the course of the reaction, H NMR spectra of
the aromatic region indicated the disappearance of the NPP
resonances at 8.25 and 7.33 ppm and the appearance of the p-
nitrophenol (NP) resonances at 8.20 and 6.95 ppm (Figure 1).
This is clear evidence that the cleavage of the labile phosphoester
bond in NPP occurred. The NMR spectra did not show evidence
of any paramagnetic species, excluding the possibility of Mo(VI)
reduction to Mo(V), and indicating that the cleavage of the
phosphodiester bond is purely hydrolytic. (Scheme 1).
moieties, and is characterized by a P NMR resonance at ꢁ4.27
4
5
4ꢁ
ppm. Complex B with the formula (NPP)
Mo12 O)
18(H O) entities
2 3
O
36(H
2
2
6
is built up of two weakly binding (NPP)Mo
O
6
31
and gives a P NMR signal at ꢁ7.42 ppm (Figure 3).
During the course of the reaction a small peak at ꢁ3.12 ppm
appeared and disappeared again. Spiking experiments identified
this peak as free phosphate. At the end of the hydrolysis reaction a
single peak at ꢁ2.05 ppm was detected. This peak could be
6
ꢁ
unambiguously assigned to [P
Mo
O
]
. The separately pre-
2
5
23
1
6ꢁ
A close inspection of the H NMR spectra revealed the existence
pared pentamolybdodiphosphate ion [P
2
Mo
5
O
]
23
, which
of another minor species in solution, characterized by a doublet
at 7.83 ppm. Although present during the course of reaction, this
species fully disappear upon completion of NPP hydrolysis.
typically forms in mildly acidic solutions containing molybdate
and phosphate ions exhibited the same NMR features as the
5
0
final complex observed at the end of hydrolytic reaction.
1
1554
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
Figure 6. Distribution of complexes A and B as a function of pD. The
pD dependence of kobs (solid triangles) was added for comparison.
Figure 5. pD dependence of kobs for the hydrolysis of 100 mM NPP in
2ꢁ
the presence of 700 mM [MoO
4
]
at 50 °C (solid line). Distribution of
molybdate species as a function of pD was added for comparison (dotted
line).
6ꢁ
between pD 4.0 and 6.0 where [Mo O ] predominates in
7
24
solution. Although at the examined pD range other molybdenum-
2ꢁ
4ꢁ
26
(
VI) species such as [MoO4] and [Mo O ] are present in
8
Figure 4 summarizes the species distribution which was
obtained by integration of the P NMR spectra, recorded at
different time increments.
solution, they seem to be catalytically much less active. Slow
3
1
4ꢁ
hydrolysis at lower pD values, where [Mo
O ]
8 26
predominates,
implies low catalytic activity of this species. Furthermore, hydro-
2ꢁ
The first-order rate constants (kobs) for the hydrolysis of NPP
lysis hardly occurred at pD 8.0 where [MoO
]
4
is the only
were calculated from the increase in intensity of the p-nitrophenol
Mo(VI) species present in solution, suggesting that the mono-
meric form is virtually hydrolytically inactive.
1
(
NP) H NMR resonances at different reaction times. A first-
ꢁ
5
ꢁ1
order exponential fit resulted in a rate constant of 2.73 ꢀ 10 s .
The same reaction was studied by following the increase of the NP
absorption band maximum at 315 nm by UVꢁVis spectroscopy and
Binding of NPP to Polyoxomolybdates. While the pD
6
ꢁ
dependence of the rate constant implies that [Mo
O
]
24
is
7
31
the hydrolytically active species in solution, the P NMR and
DOSY measurements with the PhP analogue suggest that NPP is
completely bound into complexes A and B upon mixing with
ꢁ
5
resulted in a comparable value for the rate constant (3.20 ꢀ 10
ꢁ
1
s ). It has to be noted that despite the presence of an activating
nitrophenyl group, the phosphoester bond in NPP is still extremely
stable. At 50 °C the half-life for PꢁO bond cleavage was estimated
to be 135 days. In that respect, the molybdate accelerates the rate of
phosphoester bond hydrolysis by nearly 3 orders of magnitude.
Contrary to NPP hydrolysis, the hydrolysis of phenylphosphate
6
ꢁ 45,51,52
[Mo
O
]
.
The pD dependent formation of complexes
7
24
A and B (Figure 6) shows that the ratio between complexes A and
B is strongly dependent on the pD. Upon decrease in pD of the
solution, the amount of complex B gradually increased. While at
pD 6.04 complex A is the only NPP containing structure, at pH
4.4 only about 40% of this complex is detected in solution. The
(
PP) was very slow. After 1 day at 50 °C only a small amount (2%)
1
31
of a 100 mM solution of PP was hydrolyzed to the corresponding
phenol and inorganic phosphate in the presence of 700 mM
H and P NMR experiments at pD 7.0 and 8.0 show no
evidence of the formation of the complex A and B. This may
explain the slow hydrolysis of NPP at these pD values and suggest
that the formation of complexes A and B is essential for the
hydrolysis to be observed. To correlate the distribution of
complexes A and B to the reaction rate, the pD dependency of
the reaction rate constant was added for comparison. No clear
correlation can be observed, implying that both complexes A and
B lead to hydrolysis of NPP. It is however possible that the NPP
hydrolysis rate constant is different for both complexes and that
kobs therefore represents a cumulative value.
2ꢁ
[MoO ] , while in the case of NPP almost 90% conversion
4
was achieved under the same conditions. Further accurate analysis
of the PP reaction was not possible because of the formation of a
white precipitate.
pD Dependence of the Hydrolysis Rate. To identify the
hydrolytically active molybdate complex, the effect of pD on the
hydrolysis of NPP was studied. The fact that the reactivity follows
the pD profile of a certain species suggests that kobs is propor-
tional to the concentration of that species. Since the kinetic
6
ꢁ
measurements were performed in D O solutions, the rates
The hydrolysis of NPP in the presence of [Mo O ] was
7 24
95
2
were plotted as a function of pD, which was obtained by adding
also analyzed by Mo NMR spectroscopy (Figure 7). A 700 mM
48
2ꢁ
0
.41 pH units to the recorded pH value. This is illustrated in
solution of [MoO4] at pH 5.1 is characterized by two peaks: a
2ꢁ
Figure 5, where the values of kobs and the fraction of polyoxo-
metalate species are plotted as a function of pD.
sharp peak at 0 ppm which corresponds to [MoO4] and a
broad resonance at 35 ppm that can be assigned to
6
ꢁ 53
The pD dependence of kobs shows a bell-shaped profile with the
fastest cleavage observed at pD 5.0. Comparison of the rate profile
with the concentration profile of polyoxometalates shows a good
[Mo O ]
.
Upon adding NPP a significant broadening of
7
24
the 0 ppm resonance was observed. At the end of the hydrolytic
reaction the intensity of the peak at 35 ppm was strongly
diminished and the species at 0 ppm remained as a major
6ꢁ
overlap of kobs and the concentration of [Mo O ] , implying
7
24
95
that presence of heptamolybdate is required for the hydrolytic
reaction to occur. Indeed, the highest rates were observed
component in solution. Mo NMR control experiments of a
6
ꢁ
100 mM [P Mo O ] solution confirmed that this species is
2
5
23
1
1555
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
6
ꢁ
Scheme 2. Binding of NPP to [Mo O ]
7
24
9
5
2ꢁ
4
Figure 7. Mo NMR spectra of (a) 700 mM solution of [MoO ] at
2ꢁ
pD 5.1, (b) 700 mM solution of [MoO
4
]
upon adding 100 mM
solution of NPP at pD 5.1, (c) mixture b at the end of hydrolytic
6
ꢁ
reaction, and (d) 100 mM solution of [P Mo O ] at pD 5.1.
2
5 23
Figure 9. ln(kobs/T) as a function of temperature for the hydrolysis
of NPP.
31
Since P NMR experiments indicated that the binding
6
ꢁ
between NPP and [Mo O ] is fast, it is reasonable to assume
7
24
that the formation of the product (k ) is much slower than
2
reaching the equilibrium between NPP and molybdate, k + k
k . Because the formation of nitrophenol (NP) is a first-order
2
.
1
ꢁ1
reaction, kobs can be written as in eq 1.
6
ꢁ
0
k2½Mo7O ꢂ
2
4
kobs ¼ kꢁ1
k1
ð1Þ
6
ꢁ
0
þ ½Mo O ꢂ
7
24
The ratio between k and k is equal to the formation constant,
6
ꢁ
1
ꢁ1
Figure 8. kobs as a function of the concentration of [Mo
concentration of NPP was kept constant at 20 mM.
7
O
24
]
. The
ꢁ4 ꢁ1
K . By fitting eq 1 to the data in Figure 8, k (5.60 ꢀ 10 s ) and
f
2
ꢁ1
K (361.2 M ) were calculated. It is important to note that since
f
at this pH both complexes A and B were present in solution, the
calculated binding constant represents an average value.
characterized by a broad resonance at 0 ppm. Considering the very
6ꢁ
similar environment of molybdenum atoms in [P Mo O ] and
Temperature Dependence of the Hydrolytic Reaction.
The effect of temperature on kobs was determined in the
temperature range between 293 and 353 K. As expected, an
increase of temperature resulted in higher reaction rate constants.
2
5 23
95
the complex A, it is likely that the Mo chemical shift in these
two complexes is very similar. Bearing in mind the results from
3
1
P NMR measurements, which unambiguously prove complete
q
ꢁ1
binding of NPP to the polyoxomolybdate structure, the broad-
ening of the peak at 0 ppm can be interpreted as the formation of
The activation Gibbs function (ΔG = 38 kJ mol at 310 K),
q ꢁ1 q
ꢁ1
enthalpy (ΔH = 96 kJ mol ) and entropy (ΔS = ꢁ122 J mol
6
ꢁ
ꢁ1
complex A which is gradually converted into [P Mo O ] in
K ) were obtained from the Eyring plot (Figure 9). There is a
2
5 23
q
q
the course of NPP hydrolysis. The disappearance of the
close correspondence in the ΔH and ΔS values to the previously
6ꢁ
[
Mo O ] peak at 35 ppm occurs as a consequence of the
reported values for the hydrolysis of phopshoesters catalyzed by
7
24
54ꢁ56
equilibrium shift due to the binding of these species into complex
A, which is being consumed during the hydrolytic reaction.
To estimate the equilibrium constant for the binding be-
various metal complexes.
The negative activation entropy
typically implies a largely ordered transition state and the increase
of coordination number of phosphorus from four to five through
formation of a trigonal-bipyramidal transition state. However, the
activation parameters need to be interpreted with caution as they
are usually derived from composite rate constants that also include
the contribution from the binding of the ligand to the complex.
Activation entropy is considered to be a mixture of factors
6
ꢁ
tween NPP and [Mo O ] , kinetic experiments using a fixed
7
24
2
ꢁ
amount of NPP and increasing amounts of [MoO ] at pH 5.2
4
1
and 50 °C were examined by H NMR spectroscopy (Figure 8).
The overall reaction scheme can be written as presented in
Scheme 2.
1
1556
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
2
ꢁ
4 4
Figure 10. (a) Influence of the concentration of NaClO on the rate constant for the reaction between 100 mM NPP with 700 mM [MoO ] . (b)
log(kobs/k
0
) a function of the square root of the ionic strength.
(
changes in solvation, conformation, molecularity, etc.) that are
not easily predicted or rationalized. Therefore, assessing whether
the change in activation entropy reflects a change in the reaction
mechanism or in the binding of the phosphoester to the complex is
not straightforward.
Effect of Ionic Strength of the Hydrolytic Rate. Ionic
strength plays an important role in reactions between charged
species. It highly affects the interaction between both reactants
and has a pronounced influence on the reaction rate. The effect
of salt ions on kobs was investigated by adding increasing amounts
of NaClO or NaCl to a reaction mixture containing NPP and
4
2
ꢁ
[MoO4] . As can be seen from Figure 10a, a reaction rate
decrease was observed upon addition of increasing amounts of
NaClO . The same effect was observed upon addition of NaCl.
4
This can be related to the OlsonꢁSimonson effect occurring for
reactions between ions of the same sign, which predicts a rate
decrease upon increasing the concentration of oppositely
3
Figure 11. Raw k weighted EXAFS (left) and corresponding Fourier
transforms (right) of the 9 solutions.
5
7
charged ions. The effect of added salt on the rate constant
for the reaction between two charged species A and B can be also
described by the BrønstedꢁBjerrum relationship (eq 2) in which
log(k /k ) is plotted as a function of the square root of the ionic
time, a gradual decrease of the intensity of the Mo Mo shell at
3.2 Å occurs, indicating a steady disappearance of the
3
3 3
6
ꢁ
[Mo
only the [Mo
total of eight samples have been measured. The first one
O
]
cluster. Apart from a reference solution containing
obs
0
7
24
6
ꢁ
strength (Figure 10b).
7
O
24
]
species, in which no NPP was present, a
kobs
1
=2
(
[
solution 1) was measured just after mixing NPP with
Mo O ] at room temperature. Then, the mixture was heated
log
¼ 1:018Z Z I
ð2Þ
A
B
6
ꢁ
24
k0
7
to 50 °C and samples were taken from this solution at different
In this equation k represents the reaction rate constant in the
0
time increments (solutions 2 until 8).
absence of salt, Z and Z the charge of both reactants, and I the
A
B
3
The raw k weighted Mo K-edge EXAFS spectra and the
ionic strength. Non-linearity of this plot implies that the salt
concentration does not only influence the hydrolysis rate, but
that it also exhibits a secondary effect, which is not surprising as it
is well-known that ionic strength can also play a role in speciation
corresponding Fourier transforms of the 9 solutions are shown in
Figure 11. The Fourier transforms show two distinctive sets of
peaks: a first set, centered at about R + Δ = 1.3 Å, corresponding
2
ꢁ4,49
to the overlapping Mo O shells in the range of real distances
and stability of the polyoxometalate species.
3 3 3
EXAFS Spectroscopy. The molybdenum K-edge (located at
1.7ꢁ2.4 Å, and a second, centered at about R + Δ = 3 Å,
1
9999 eV) was used to investigate the coordination sphere of the
corresponding to the Mo Mo shells at real distances of 3.2 and
3
3 3
31
molybdenum atoms during the hydrolysis of NPP. While
P
3.4 Å, respectively.
In the left part of Figure 11, a gradual change in the raw k
3
NMR spectroscopy allows for the detection of the complexes A
6
ꢁ
and B, as well as of the final complex [P Mo O ] , which all
weighted EXAFS spectra can be seen. The oscillations between
k = 10 and k = 16 show the clearest changes. The Fourier
transforms in the right part of Figure 11 show the clearest change
at around R + Δ = 3 Å, indicating that structural changes are
occurring in the polyoxomolybdates present in the solution.
2
5 23
contain phosphorus atoms, it does not give an insight into the
fate of other polyoxomolybdate species that do not contain
phosphorus atoms. The EXAFS measurements were performed
in a time period of approximately 24 h. During this period of
1
1557
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
Therefore, a back-transformation from FT to χ(k) in the range
of the Mo Mo shells was made to give the corresponding
3
3 3
EXAFS oscillations. These were then fitted with two Mo Mo
3
3 3
shells. This Fourier-filtered data range, taken from the spectra in
Figure 11, is shown in Figure 12. The experimental curves are
given as the black solid lines, and the fitted curves are given as the
red dashed lines. Careful calculation from the crystal data
6
ꢁ
indicates that every Mo atom in the [Mo O ] cluster “sees”
7
24
on average 1.71 Mo neighbors at 3.2 Å, and also 1.71 Mo
neighbors at 3.4 Å. These coordination numbers N have been
fixed in the shell fit procedure for the reference solution, whereas
2
the distances R and DebyeꢁWaller factors σ were left free. The
fitted distances of 3.23 and 3.40 Å, respectively, correspond well
to the distances obtained from the crystal structure. This
indicates that the only polyoxomolybdate present in the refer-
Figure 12. Back Fourier-transformed EXAFS signals (left) and Fourier
transform windows (right) taken from Figure 11. Black solid lines are
experimental curves; red dashed lines are fitted curves.
6ꢁ
ence solution is [Mo O ] , as could be expected.
7
24
In a second step, the DebyeꢁWaller factors that were obtained
from the fit of the reference solution were taken and fixed during
the shell fit procedure of solutions 1 (time 0) until 8 (time 24 h)
Table 1. EXAFS Structural Parameters for the Solution
to obtain the coordination numbers in the Mo Mo shells. The
6
ꢁ
3 3 3
Containing 700 mM [Mo O ] and 100 mM NPP at
7
24
EXAFS structural parameters for the 9 solutions are summarized
in Table 1.
pD = 5.1
a
a
2
2
N
R (Å)
σ (Å )
ΔEk=0 (eV)
F
From the right part of Figure 12, it is clear that the amplitude
of the Mo Mo shell at a real distance of 3.2 Å decreases
b
b
3 3 3
Reference
1.71
1.71
1.57
1.64
1.03
1.23
0.95
1.34
0.80
1.27
0.72
1.28
0.64
1.26
0.61
1.26
0.42
1.23
3.23
3.40
3.22
3.40
3.23
3.40
3.23
3.40
3.23
3.40
3.23
3.40
3.23
3.40
3.23
3.40
3.22
3.40
0.00526
0.00603
4.0
0.25
steadily from the reference solution to the solution at time 24 h,
whereas the amplitude of the shell at a real distance of 3.4 Å stays
more or less the same (from Table 1, it can be seen that in fact
also this shell decreases somewhat in intensity). This indicates
(no NPP)
b
Solution 1
t = 0
0.00526
0.00603
0.00526
0.00603
0.00526
0.00603
0.00526
0.00603
0.00526
0.00603
0.00526
0.00603
0.00526
0.00603
0.00526
0.00603
3.8
0.23
0.11
0.09
0.07
0.06
0.05
0.05
0.04
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
6ꢁ
Solution 2
t = 30 min
Solution 3
t = 1 h 45 min
Solution 4
t = 3 h
10.1
10.6
10.7
10.9
11.1
11.1
11.7
that the [Mo O ] cluster is gradually converted into the
7
24
6
ꢁ
[P Mo O ] cluster.
2
5 23
From looking at Figure 12 and Table 1, comparing the signals
and data from the reference solution with those of the solution at
time 0, it is noteworthy to mention that right after mixing the
6
ꢁ
NPP solution with the [Mo O ] solution, a small fraction of
7
24
6ꢁ
24
the [Mo O ] cluster seems to have immediately reacted, even
7
Solution 5
t = 4 h 30 min
Solution 6
t = 5 h 30 min
Solution 7
t = 6 h 35 min
Solution 8
t = 24 h
though the kinetics of the hydrolytic reaction are quite slow at
room temperature. This is in agreement with the suggested
6
ꢁ
equilibrium that is established between the [Mo O ] cluster
7
24
31
and the intermediate species A and B detected in P NMR.
Assuming that the reaction has reached an end point at t = 24 h,
6ꢁ
one can estimate the remaining fraction of [Mo O ] , by taking
7
24
the ratio between the coordination number for the Mo Mo
3 3
3
shellat 3.2 Å, from solutions 8 and 1. This indicates that about 25%
6ꢁ
a
of the [Mo O ] cluster remains in solution. Theoretically, if all
7 24
6
Errors in coordination numbers N are (10%; errors in distances R are
0.01 Å. Value fixed during shell-fit procedure. The overall goodness
of the fits, F, is given by χ , weighted by the magnitude of data. The
threshold energy, Ek=0, was arbitrarily defined at 19999 eV and varied as
ꢁ
6ꢁ
23
b
[
Mo O ] would be converted to [P Mo O ] , the coordi-
7 24 2 5
(
2
nation number for the Mo Mo shell at 3.4 Å should increase
3
3 3
6ꢁ
to a value of 2, since every Mo atom in the [P Mo O ] cluster
2
5 23
a global fit parameter resulting in the energy shift ΔEk=0
.
“sees” on average two other molybdenum atoms at about 3.4 Å.
6
ꢁ
24
However, both the fact that only about 75% of the [Mo O ]
7
6
ꢁ
6ꢁ
23
Because the [Mo O ] species is in constant equilibrium
cluster is converted into the [P Mo O ] cluster and the
7
24
2 5
2
ꢁ
with the [MoO4] monomer, it is not possible to obtain a
reasonable fit including the Mo O shells. The peaks at R + Δ =
Å, however, allow the identification of the polyoxomolybdate in
solution. Indeed, close examination of the crystal structure of the
Mo O ] cluster reveals that there are basically two sets of
Mo Mo distances in this molecule: one centered at 3.2 Å, and
another one centered at 3.4 Å. In the [P Mo O ] cluster,
however, which would result from the reaction between
fact that both of these clusters are in equilibrium with the
2
ꢁ
[MoO4] monomer (in which the molybdenum atom does
not “see” any other molybdenum atoms) account for the
observed slight decrease in the coordination number of the
3
3 3
3
6
ꢁ
[
Mo Mo shell at 3.4 Å.
7
24
3 3 3
3
3 3
5
6
6ꢁ
23
2
5
’
CONCLUSION
6ꢁ
[
Mo O ] and phosphate, all Mo Mo distances are more
In conclusion, we report on a DNA model phosphoester bond
7
24
3 3 3
or less the same and centered at 3.4 Å. So, monitoring the
amplitude of the Mo Mo shell at 3.2 Å allows to follow the
cleavage promoted by a highly negatively charged polyoxometa-
6
ꢁ
7 24
late cluster. Binding of NPP to [Mo O ] results in the
3
3 3
6
ꢁ
change in the concentration of [Mo O ]
.
formation of two complexes A and B in which the phosphoester
7
24
1
1558
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
group is incorporated into the polyoxomolybdate skeleton. The
sharing of oxygen atoms with Mo(VI) centra may lead to bond
strain and cause polarization of the PꢁO ester bond and its
activation toward external attack by water. The cleavage seems to
occur by a mechanism which is different to that of other currently
known hydrolytically active metal complexes. The pD depen-
dence of kobs shows a bell-shaped profile with the fastest cleavage
observed at pD 5.0. At physiological pH, the dominant molyb-
denum species is the mononuclear [MoO4] which is proven to
be hydrolytically inactive. However, in the lower pH environment
associated with hypoxic tumors the catalytically active species
at pH 5 may play a role in the biological activity of POMs.
We are currently examining the phosphoesterase activity of poly-
oxomolybdate toward oligonucleotides and other DNA substrates.
These studies are important since they may shed more light on the
molecular origin of the biological activity of polyoxomolybdates.
(17) Yamase, T.; Fujita, H.; Fukushima, K. Inorg. Chim. Acta 1988,
151, 15.
18) Ogata, A.; Mitsui, S.; Yanagie, H.; Kasano, H.; Hisa, T.; Yamase,
T.; Eriguchi, M. Biomedicine & Pharmacotherapy 2005, 59, 240.
19) Yanagie, H.; Ogata, A.; Mitsui, S.; Hisa, T.; Yamase, T.;
Eriguchi, M. Biomed. Pharmacother. 2006, 60, 349.
20) Ogata, A.; Yanagie, H.; Ishikawa, E.; Morishita, Y.; Mitsui, S.;
(
(
(
Yamashita, A.; Hasumi, K.; Takamoto, S.; Yamase, T.; Eriguchi, M. Br. J.
Cancer 2008, 98, 399.
(21) Kolesarova, A.; Capcarova, M.; Sirotkin, A. V.; Medvedova, M.;
Kalafova, A.; Filipejova, T.; Kovacik, J. J. Environ. Sci., Health, Part A
2011, 46, 170.
2ꢁ
(22) Katsoulis, D. E.; Lambriandou, A. N.; Pope, M. T. Inorg. Chim.
Acta 1980, 46, L55.
(
23) Geraldes, C. F. G. C.; Castro, M. C. C. A. J. Inorg. Biochem.
986, 28, 319.
24) Piperaki, P.; Katsaros, N.; Katakis, D. Inorg. Chim. Acta 1982,
7, 37.
25) Hill, L. M. R.; George, G. N.; Duhm-Klair, A. K.; Young, C. G.
J. Inorg. Biochem. 2002, 88, 274.
26) Kwak, W.; Pope, M. T.; Scully, T. F. J. Am. Chem. Soc. 1975,
7, 5735.
27) Cartuyvels, E.; Van Hecke, K.; Van Meervelt, L.; G €o rller-
Walrand, C.; Parac-Vogt, T. N. J. Inorg. Biochem. 2008, 102, 1589.
1
(
6
(
’
AUTHOR INFORMATION
(
Corresponding Author
9
*E-mail: Tatjana.Vogt@chem.kuleuven.be. Fax: +32 16/327992.
(
(
(
(
(
28) Weil-Malherbe, H.; Green, R. H. Biochem. J. 1951, 49, 286.
29) Ishikawa, E.; Yamase, T. J. Inorg. Biochem. 2006, 100, 344.
30) Cleland, W. W.; Hengge, A. C. Chem. Rev. 2006, 106, 3255.
31) Williams, N. H.; Takasaki, B.; Wall, M.; Chin, J. Acc. Chem. Res.
’
ACKNOWLEDGMENT
G.A. thanks FWO-Flanders (Belgium) for the doctoral fellow-
ship. T.N.P.-V. thanks K. U. Leuven for the financial support
START 1/09/028). We thank the FWO-Vlaanderen and The
1999, 32, 485.
(
(32) Morrow, J. R.; Iranzo, O. Curr. Opin. Chem. Biol. 2004, 8, 192.
(33) Mancin, F.; Scrimin, P.; Tecilla, P.; Tonellato, U. Chem.
Commun. 2005, 2540.
Netherlands Organization for Scientific Research (NWO) for
providing beam time at the DutchꢁBelgian Beamline
(34) Ott, R.; Kr €a mer, R. Angew. Chem., Int. Ed. 2000, 39, 3255.
(DUBBLE, BM26A, ESRF). Sergei Nikitenko and Miguel Sil-
(35) Fanning, A. M.; Plush, S. E.; Gunnlaugsson, T. Chem. Commun.
veira (DUBBLE) are acknowledged for technical support during
the EXAFS measurements.
2
006, 3791.
36) Aguilar-Perez, F.; Gomez-Tagle, P.; Collado-Fregoso, E.;
Yatsimirsky, A. K. Inorg. Chem. 2006, 45, 9502.
37) Feng, G.; Natale, D.; Prabaharan, R.; Mareque-Rivas, J. C.;
(
(
’
REFERENCES
Williams, N. H. Angew. Chem., Int. Ed. 2006, 45, 7056.
(38) Schneider, H.-J.; Rammo, J.; Hettich, R. Angew. Chem., Int. Ed.
Engl. 1993, 32, 1716.
(39) O’Donoghue, A.; Pyun, S. Y.; Yang, M. Y.; Morrow, J. R.;
Richard, J. P. J. Am. Chem. Soc. 2006, 128, 1615.
(40) Komiyama, M.; Kina, S.; Matsumura, K.; Sumaoka, J.; Tobey,
S.; Lynch, V. M.; Anslyn, E. J. Am. Chem. Soc. 2002, 24, 13731.
(41) Miti ꢀc , N.; Smith, S. J.; Neves, A.; Guddat, L. W.; Gahan, L. R.;
Schenk, G. Chem. Rev. 2006, 106, 3338.
(
(
1) Pope, M. T.; M €u ller, A. Angew. Chem., Int. Ed. Engl. 1991, 30, 34.
2) Pope, M. T. Polyoxo anions: Synthesis and Structure. In
Comprehensive Coordination Chemistry II;Wedd, A. G., Ed.; Elsevier
Science: New York, 2004; Vol.4, pp 635ꢁ678.
(
3) Hill, C. L. Polyoxometalates: Reactivity. In Comprehensive Co-
ordination Chemistry II; Wedd, A. G., Ed.; Elsevier Science: New York,
004; Vol.4, pp 679ꢁ759.
4) Pope, M. T., M €u ller, A., Eds.; Polyoxometalate Chemistry: From
2
(
Topology via Self-Assembly to Applications; Kluwer: Dordrecht, The
Netherlands, 2001.
(42) Neves, A.; Lanznaster, M.; Bortoluzzi, A. J.; Peralta, R. A.;
Casellato, A.; Castellano, E. E.; Herrald, P.; Riley, M. J.; Schenk, G. J. Am.
Chem. Soc. 2007, 129, 7486.
(43) Cartuyvels, E.; Absillis, G.; Parac-Vogt, T. N. Chem. Commun.
2008, 85.
(44) Absillis, G.; Cartuyvels, E.; Van Deun, R.; Parac-Vogt, T. N.
J. Am. Chem. Soc. 2008, 130, 17400.
(45) Van Lokeren, L.; Cartuyvels, E.; Absillis, G.; Willem, R.; Parac-
Vogt, T. N. Chem. Commun. 2008, 2774.
(
(
5) Kozhevnikov, I. V. Chem. Rev. 1998, 98, 198.
6) Hill, C. L.; Prossermccartha, C. M. Coord. Chem. Rev. 1995,
1
2
2
43, 407.
(7) Proust, A.; Thouvenot, R.; Gouzerh, P. Chem. Commun.
008, 1837.
(
8) Mizuno, N.; Yamaguchi, K.; Kamata, K. Coord. Chem. Rev. 2005,
49, 1944.
(
(
(
(
(
9) Neumann, R. Prog. Inorg. Chem. 1998, 47, 317.
10) Rhule, J. T.; Hill, C. L.; Judd, D. A. Chem. Rev. 1998, 98, 327.
11) Yamase, T. J. Mater. Chem. 2005, 15, 4773.
12) Hasenknopf, B. Front. Biosci. 2005, 10, 275.
(46) (a) Borsboom, M.; Bras, W.; Cerjak, I.; Detollenaere, D.;
van Loon, D. G.; Goedtkindt, P.; Konijnenburg, M.; Lassing, P.; Levine,
Y. K.; Munneke, B.; Oversluizen, M.; van Tol, R.; Vlieg, E. J. Synchrotron
Radiat. 1998, 5, 518. (b) George, N. G.; Pickering, I. J. EXAFSPAK, a
suite of computer programs for analysis of X-ray absorption spectra; Stanford
Synchrotron Radiation Laboratory: Stanford, CA, 2000.
(47) Ankudinov, A. L.; Ravel, B.; Rehr, J. J.; Conradson, S. D. Phys.
Rev. B 1998, 58, 7565.
13) Wang, X. H.; Li, F.; Liu, S. X.; Pope, M. T. J. Inorg. Biochem.
2
005, 99, 452.
14) Gerth, H. U. V.; Rompel, A.; Krebs, B.; Boos, J.; Lanvers-
Kaminsky, C. Anti-Cancer Drugs 2005, 16, 101.
15) Cindric, M.; Novak, T. K.; Kraljevic, S.; Kralj, M.; Kamenar, B.
Inorg. Chim. Acta 2006, 359, 1673.
16) Fujita, H.; Fujita, T.; Sakurai, T.; Seto, Y. Chemotherapy 1992,
0, 173.
(
(
(48) Glasoe, P. K.; Long, F. A. J. Phys. Chem. 1960, 64, 188.
(49) Cruywagen, J. J. Adv. Inorg. Chem. 2000, 49, 127.
€
(
(50) Pettersson, L.; Andersson, I.; Ohman, L.-O. Inorg. Chem. 1986,
4
25, 4726.
1
1559
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560

Inorganic Chemistry
ARTICLE
(51) Yagasaki, A.; Andersson, I.; Pettersson, L. Inorg. Chem. 1987,
2
4
4
1
4
6, 3926.
(
52) Lyxell, D.-G.; Pettersson, L.; Persson, I. Inorg. Chem. 2001,
0, 584.
(
53) Maksimovskaya, R. I.; Maksimov, G. M. Inorg. Chem. 2007,
6, 3688.
(
54) Deal, K. A.; Hengge, A. C.; Burstyn, J. N. J. Am. Chem. Soc. 1996,
18, 1713.
(55) Deck, K. M.; Tseng, T. A.; Burstyn, J. N. Inorg. Chem. 2002,
1, 669.
(
56) Fry, F H.; Fishermann, J.; Belousoff, M. J.; Spiccia, L.; Br €u gger,
J. Inorg. Chem. 2005, 44, 941.
(
(
57) Olson, A. R.; Simonson, T. R. J. Chem. Phys. 1946, 17, 1167.
58) Sjobom, K.; Hedman, B. Acta Chem. Scand. 1973, 27, 3673.
1
1560
dx.doi.org/10.1021/ic201498u |Inorg. Chem. 2011, 50, 11552–11560