FULL PAPER
DOI:10.1002/ejic.201300419
Unveiling the Interaction of Vanadium Compounds with
Human Serum Albumin by Using
1
H STD NMR and
Computational Docking Studies
David M. Dias,
[a]
João P. G. L. M. Rodrigues,
[b]
Neuza S. Domingues,
[a]
Alexandre M. J. J. Bonvin,
[b]
and
M. Margarida C. A. Castro*
[a,c,d]
Keywords: Proteins / Docking studies / Drug delivery / Vanadium / NMR spectroscopy
The binding of the V
V
oxidation products of two vanadi-
um(IV) compounds, [VO(dmpp)
2
] and [VO(maltolato)
2
],
which have shown promising anti-diabetic properties, to hu-
man serum albumin (HSA) in aqueous aerobic solution has
been studied by
1
H saturation transfer difference (STD) NMR
spectroscopy and computational docking studies. Group epi-
tope mapping and docking simulations indicate a preference
of HSA binding to the 1:1 [VO
2
(dmpp)(OH)(H
2
O)]
–
and 1:2
[VO
2
(maltol)
2
]
–
vanadium(V) species. By using known HSA
binders, competition NMR experiments revealed that both
Introduction
Over the last few years vanadium compounds (VCs) have
attracted considerable interest from the scientific com-
munity due to their demonstrated pharmacological proper-
ties.
[1]
In particular, their potential use as oral insulin mi-
metic agents has been demonstrated by in vivo
[2,3]
and ex
vivo studies
[4]
as well as in clinical trials.
[5]
Their insulin-like
capacity to modulate several metabolic pathways may be
due to their inhibitory effect on protein tyrosine phosphat-
ases (PTPases) and/or the activation of tyrosine kinases,
which are responsible for the activation of signal transduc-
tion pathways, particularly in the insulin signalling cas-
cade.
[6]
Intensive research has been carried out to develop
VCs for use as orally administered drugs in the treatment
[a] Department of Life Sciences, Faculty of Sciences and
Technology, University of Coimbra,
P. O. Box 3046, 3001-401 Coimbra, Portugal
E-mail: gcastro@ci.uc.pt
Homepage: www.uc.pt/fctuc/dcv
[b] Bijvoet Center for Biomolecular Research, Department of
Chemistry, Faculty of Science, Utrecht University,
Padualaan 8, 3584 CH Utrecht, The Netherlands
[c] Coimbra Chemistry Centre, Rua Larga, University of Coimbra,
3004-535 Coimbra, Portugal
[d] Center of Neurosciences and Cell Biology, University of
Coimbra,
3001-401 Coimbra, Portugal
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201300419.
Eur. J. Inorg. Chem. 2013, 4619–4627 © 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim4619
complexes preferentially bind to drug site I. Docking simula-
tions carried out with HADDOCK together with restraints de-
rived from the STD results led to three-dimensional models
that are in agreement with the NMR spectroscopic data, pro-
viding useful information on molecular interaction modes.
These results indicate that the combination of STD NMR and
data-driven docking is a good tool for elucidating the interac-
tions in protein–vanadium compounds and thus for clarifying
the mechanism of drug delivery as vanadium compounds
have shown potential therapeutic properties.
of Type II diabetes mellitus (DM) at an effective non-toxic
dose. Among the many compounds synthesized to date,
only a few seem to show promise for this purpose. Bis-
(maltolato)oxovanadium(IV) (BMOV),
[7]
bis(picolinato)-
oxovanadium(IV) and bis(allixinato)oxovanadium(IV) as
well as other derivatives
[8]
have demonstrated anti-diabetic
properties
[9–11]
through indicators of insulin mimetism as-
sessed by in vitro and ex vivo studies such as glucose uptake
rates,
[12,13]
inhibition of free fatty acid (FFA) release
[14]
and
specific protein phosphorylation.
[4,10]
In vivo experiments
have confirmed their therapeutic activity and toxicity,
[15,16]
established the minimum effective dose
[17]
and provided in-
formation about bioavailability and pharmacokinetics.
BMOV and similar compounds like bis(ethylmaltolato)-
oxovanadium(IV) (BEOV) have already been presented in
clinical trials.
[18]
The vanadium complex bis(3-hydroxy-1,2-dimethyl-4-
pyridinonato)oxovanadium(IV), [V
IV
O(dmpp)
2
], has been
extensively studied. Its structure in the solid state has been
determined
[19]
and the different species formed in aqueous
solution under aerobic conditions were identified by using
different techniques and the respective formation constants
determined.
[20]
In vitro insulin mimetic and toxicity studies
with two cell lines, the mouse fibroblast SV 3T3 and the
human skin fibroblast F26,
[21]
have been conducted. Studies
with the human fibroblast cell line, 3T3-L1,
[15]
as well as
with human erythrocytes
[22]
have also been carried out to
test its cytotoxicity and glucose uptake enhancement ca-
www.eurjic.org FULL PAPER
pacity. Recently, its anti-diabetic action was recognized
through an ex vivo study with isolated primary rat adipo-
cytes, which demonstrated that this compound improves
glucose internalization, inhibition of FFA release, and has
the capacity to activate the insulin signalling cascade
through phosphorylation of the key proteins of this path-
way.
[23]
In vivo studies with obese Zucker rats by using mag-
netic resonance techniques (MRI/MRS) confirmed the pos-
itive effects of [V
IV
O(dmpp)
2
] on glucose and lipid meta-
bolism, reinforcing its promising anti-diabetic capacity.
[24]
However, the mechanism of the therapeutic action of
VCs in the human body has not yet been clarified. The in-
creased pharmacological effect of chelated V
IV
complexes
compared with VOSO
4
has been attributed, in part, to in-
creased absorption in the gastrointestinal tract.
[25]
The dif-
ferences in biodistribution and efficacy observed between
non-chelated and chelated vanadyl ions result from a small
portion of the administered complex remaining intact after
absorption.
[21,26]
A major concern is how this intact frac-
tion of the VCs is transported in the blood stream and is
taken up by the cells of peripheral tissues.
[27]
Thus, it is
important to know the interactions between the adminis-
tered VCs and endogenous macromolecular components
and/or small bioligands (such as lactate and citrate) present
in blood serum to determine whether the decomposition of
the VCs occurs or ternary complexes are formed, and to
investigate the role of serum proteins in the transport of
VCs.
[28–34]
This information is crucial in the search for a
rational drug design to improve the therapeutic efficacy of
these compounds.
The two main serum proteins involved in the transport
of vanadium species (both V
IV
and V
V
) are human serum
transferrin (Tf) and albumin (HSA). The interaction of
vanadyl (V
IV
) and vanadate (V
V
) with transferrin has been
detected by different techniques, including gel electrophore-
sis, ultrafiltration, chromatographic techniques, atomic ab-
sorption spectrometry (AAS) and inductively coupled
plasma mass spectrometry (ICP-MS).
[35,36]
Structural infor-
mation on these interactions was obtained by spectroscopic
techniques, such as UV/Vis and FTIR and FT Raman spec-
troscopy.
[37]
Vanadate binds to the two Fe
III
binding sites
of apo-transferrin as the dioxidovanadium(V) cation, VO
2
+
,
to form a (VO
2
+
)
2
–Tf complex without the need of a syner-
gistic binding anion, as shown by
51
V NMR spec-
troscopy,
[38]
whereas apo-transferrin requires such a syner-
gistic cation to bind two equivalents of the oxidovanadi-
um(IV) cation, VO
2+
, at these sites.
[39]
HSA strongly binds
one equivalent of VO
2+
in the Cu
II
site at the N terminus
and several equivalents, weakly and non-specifically, to
carboxylate side-chains of surface amino acids.
[40]
However,
when the VO
2+
/HSA ratio is 1:1 or lower, a binuclear metal
species, (VO
2+
)
2
–HSA, is formed.
[29]
EPR spectroscopy was
used as an essential tool of such extensive interaction stud-
ies of VO
2+
with Tf and HSA.
[28,29,41]
It is not yet fully clear
in which form the VO
2+
ion is transported to the target cells
in the organism as the higher concentration of HSA in the
blood could compensate its much lower affinity towards
VO
2+
relative to transferrin.
[42]
In addition, the competition
Eur. J. Inorg. Chem. 2013, 4619–4627 © 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim4620
effects of small bioligands present in the serum also need
to be taken into consideration.
[29]
The interaction of V
IV
compounds, including BMOV
and [VO(dmpp)
2
], with apo-Tf, HSA and small bioligands
present in blood serum has been studied by EPR and circu-
lar dichroism (CD).
[27–34,43]
The results indicate that, besides
the case in which the binding of the carrier ligand to vana-
dium is very weak, for example, 6-methylpicolinate, mixed
species are formed with transferrin with possible partial
carrier ligand displacement, to form cis-[VO(carrier)
x
(Tf)]
species (x = 1, 2; carrier ligand = picolinato, maltolato,
dmpp). In the case of HSA, mixed species cis-[VO(carrier)
2
-
(HSA)] involving hydrogen bonding and/or hydrophobic in-
teractions with the protein surface have also been proposed.
The formation of ternary species between the proteins
and the VO–carrier complex is an important issue as it af-
fects the distribution of species in the blood plasma. There-
fore independent evidence of their presence was an impor-
tant objective of this work. Although it is usually accepted
that the metal ion is transported in the blood in the V
IV
oxidation state,
[27,44]
almost independently of the initial oxi-
dation state of the VC, oxidation may occur, resulting in the
formation of a limited amount of diamagnetic V
V
species,
which can be studied by NMR spectroscopy. In this work
we studied the interaction in aqueous solution of vanadate/
Hdmpp complexes with HSA by using the
1
H saturation
transfer difference (STD) NMR technique.
[45]
This tech-
nique provides data on the interaction of small molecules
with a protein, validation of binding epitopes, estimation
of affinity constants and site-specific information through
competition studies. The vanadate/Hdmpp (M/L = 1:2)
solution contains both 1:1 and 1:2 V
V
species (Scheme 1),
the same species that result from the dissolution of the solid
[V
IV
O(dmpp)
2
] in water under aerobic conditions.
[20]
Scheme 1. A) Chemical structures of 3-hydroxy-1,2-dimethyl-4-pyr-
idinone (1, Hdmpp), 3-hydroxy-2-methyl-4-pyrone (2, Maltol) and
bis(maltolato)oxovanadium(IV) (3, BMOV) (with K
1
= 107.5 and
K
2
= 106.2). B) Chemical equilibrium representing the dissolution
of the solid bis[3-hydroxy-1,2-dimethyl-4-pyridinonato]oxovanadi-
um(IV), [V
IV
O(dmpp)
2
]inH
2
O under physiological conditions and
the resulting oxidation products, the vanadium(V) species
[V
V
O
2
(dmpp)(H
2
O)(OH)]
–
and [V
V
O
2
(dmpp)
2
]
–
, with M/L ratios of
1:1 and 1:2, respectively (with K
1
= 1010.48 and K
2
= 105.25).
www.eurjic.org FULL PAPER
The use of the diamagnetic vanadate/Hdmpp system in-
stead of an aerobic solution of [V
IV
O(dmpp)
2
] avoids the
presence of a small amount of paramagnetic species in solu-
tion resulting from incomplete oxidation,
[20]
which would
not allow the correct use of
1
H STD NMR experiments.
Competitive
1
H STD NMR experiments with two known
inhibitors of site I and site II of HSA, warfarin and ibup-
rofen, respectively (for a definition of site I and site II see
the Supporting Information), were performed to discrimi-
nate the preferential binding site of the V
V
coordinated spe-
cies and Hdmp.
[46]
A parallel study was also carried out
with the vanadate/maltol system to investigate the behav-
iour of the [V
IV
O(maltolato)
2
] compound (Scheme 1). The
STD results were complemented by data-driven docking
calculations performed with HADDOCK,
[47,48]
taking into
account the information from STD experiments, resulting
in 3D models that illustrate the interaction of V
V
species
with HSA, one of the main plasma proteins. Herein we
show that this procedure is a useful tool for providing im-
portant information on drug delivery, an approach not pre-
viously employed in the study of the use of vanadium com-
pounds as therapeutic drugs.
Results and Discussion
Figure 1 presents the 1D
1
H NMR spectra of solutions
containing 5 mm free Hdmpp (A), vanadate/Hdmpp (1:2,
1mm in vanadate) in the presence of 0.03 mm HSA (B) and
the
1
H STD spectrum of solution B (C).
Spectrum A shows the signals corresponding to the two
methyl groups and the aromatic 5-H and 6-H protons, as-
signed in accordance with the literature.
[20]
In the spectra
of solutions containing 1:2 vanadate/dmpp, multiple signals
are observed for each type of proton, which have been as-
signed in the figure, indicating the presence in solution of
three main species in slow exchange on the NMR timescale:
Figure 1.
1
H NMR spectra of A) an aqueous solution of 5 mm Hdmpp, B) an aqueous solution of vanadate/Hdmpp (1 mm in vanadate)
in an M/L ratio of 1:2 with 30 mm HSA at pH = 7 and C) STD NMR spectrum of solution (B). A spin-lock pulse of 30 ms was used to
remove protein resonances. The signal assignments are shown in the figure: L is the free Hdmpp and VL and VL
2
are the 1:1
[VO
2
(dmpp)(H
2
O)(OH)]
–
and 1:2 [VO
2
(dmpp)
2
]
–
complexes, respectively. The difference in the vertical scale of the reference (A,B) and
STD (C) spectra is due to the different number of scans used in the respective acquisitions (see the Exp. Sect.).
Eur. J. Inorg. Chem. 2013, 4619–4627 © 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim4621
the free dmpp and two vanadium(V) complexes, the 1:1
[VO
2
(dmpp)(H
2
O)(OH)]
–
and 1:2 [VO
2
(dmpp)
2
]
–
V
V
species
(Scheme 1). When HSA was added to the vanadate/dmpp
system (spectrum B), the resonances of dmpp in all species
broaden and lose resolution, as expected, due to their inter-
action with the protein and the small increase in solution
viscosity. Only signals of the small molecules are observed
as the protein resonances were suppressed by using a spin-
lock filter during the acquisition.
A similar study was carried out with the vanadate/maltol
system and the
1
H NMR signals of the obtained spectra
are assigned in Figure 2, in agreement with literature
data
[49]
and as confirmed by
1
H NMR spectra obtained
from solutions of vanadate/maltol at different metal/ligand
(M/L) ratios (data not shown).
In the spectrum of vanadate/maltol, M/L = 1:2, two sig-
nals for the methyl group and the 5-H and 6-H aromatic
protons are observed corresponding to free maltol and the
1:2 V
V
complex, [VO
2
(maltolato)
2
]
–
(Scheme 1), in slow ex-
change on the NMR timescale.
[49]
Again for this system, all
the resonances broaden in the presence of HSA (Figure 2,
B), reflecting the binding of maltol and [VO
2
(maltolato)
2
]
–
to the protein and the small increase in the viscosity of the
solution. The relative intensities of the resonances of the
species detected in solution for the two systems are in agree-
ment with the values of the formation constants determined
for the 1:1 and 1:2 complexes in both systems. Whereas in
the vanadate/Hdmpp system the association constants for
the 1:1 and 1:2 species are, respectively, K
1
= 1010.48 m
–1
and K
2
= 105.25 m
–1
,
[20]
in the vanadate/maltol system the
corresponding association constants are K
1
= 107.5 m
–1
and
K
2
= 106.2 m
–1
.
[49]
Figure 1 (C) shows the
1
H STD NMR spectrum of the
solution containing vanadate/dmpp (1:2) and HSA. The ap-
pearance of the resonances of the three species L, VL and
VL
2
in the spectrum indicates, qualitatively, that all these
www.eurjic.org FULL PAPER
Figure 2.
1
H NMR spectra of A) an aqueous solution of 60 mm maltol, B) an aqueous solution of vanadate/maltol (1 mm in vanadate)
in an M/L ratio of 1:2 with 30 mm HSA at pH = 7 and C) STD NMR spectrum of solution B. A spin-lock pulse of 30 ms was used to
remove protein resonances. The signal assignments are shown in the figure: L is the free maltol and VL and VL
2
are the 1:1 [VO
2
(malto-
lato)(H
2
O)(OH)]
–
and 1:2 [VO
2
(maltolato)
2
]
–
complexes, respectively. The difference in the vertical scale of the reference (A,B) and STD
(C) spectra is due to the different number of scans used in the respective acquisitions (see Exp. Sect.).
species bind to the protein. However, only through an inte-
gration of the signals in spectra B and C and the determi-
nation of the corresponding A
STD
values [by using Equa-
tions (1) and (2), see the Exp. Sect.] is it possible to draw
conclusions about the relative binding strengths of the dif-
ferent species to the protein and to identify the group epi-
tope for each one by group epitope mapping (GEM). The
A
STD
values allowed full GEM characterization of the van-
adate/Hdmpp system (Table 1).
Table 1. Values of GEM relative to the highest values of A
STD
(5-
H of the 1:1 species and 6-H of the 1:2 species for vanadate/Hdmpp
and vanadate/maltol systems, respectively) obtained from the inter-
action of all the small molecules with HSA.
GEM [%]
Small molecule 6-H 5-H CH
3
N-CH
3
Hdmpp 75 82 61 42
[VO
2
(dmpp)(H
2
O)(OH)]
–
65 100 59 41
[VO
2
(dmpp)
2
]
–
75 83 59 41
Maltol 79 93 65 –
[VO
2
(Maltol)
2
]
–
100 97 30 –
In this case, although all the species seem to interact with
HSA, the values obtained indicate that there is a slight pref-
erence for the binding of the 1:1 species and through the 5-
H proton, as this resonance is the highest saturation-receiv-
ing proton. The calculated values of GEM for all the
Hdmpp protons were normalized relative to the value for
5-H. Similar studies carried out with the vanadate/maltol
system (1 mm in vanadate) with M/L = 1:2 in the presence
of HSA (0.03 mm) led to the
1
H STD NMR spectrum
shown in Figure 2 (C) and the GEM values in Table 1. The
presence of all the resonances in the STD spectrum indi-
cates that both free maltol and the 1:2 species bind to the
protein. However, according to the GEM values, the strong-
Eur. J. Inorg. Chem. 2013, 4619–4627 © 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim4622
est interaction with HSA occurs through 6-H of the 1:2
species, although a rather similar saturation transfer is ob-
served for 5-H and 6-H.
To investigate the possible binding sites on HSA, com-
petitive
1
H STD NMR studies with the most interactive
species were carried out in the presence of warfarin and
ibuprofen, two known binders to site I and site II of HSA,
respectively.
[50]
The displacement of binding species from
these sites by the corresponding reported specific inhibitor
after being added to the initial solution of vanadate systems
and HSA reflects a direct competition and interaction of
the species at these sites.
[50]
Integration of the competitive
1
H STD NMR spectra for
the vanadate/dmpp (1:1) and HSA solution (Figure 3) led
to the A
STD
values presented in Table 2. It can be concluded
that the interaction of the 1:1 complex [VO
2
(dmpp)-
(H
2
O)(OH)]
–
with the binding site I is slightly favoured.
Although displacement of the species from both site I
and site II of HSA was observed, according to the A
STD
values, in the absence and presence of warfarin and ibup-
rofen, the largest decrease occurs in the presence of warfa-
rin for all the ligand protons of the 1:1 complex, which indi-
cates that binding to site I is favoured over site II. It is also
known that ibuprofen displaces species at site I of HSA,
albeit to a lesser extent, and so, if the interaction occurs
through site II a higher displacement would be expected
and reflected by a smaller A
STD
value.
Similarly, Figure 4 presents typical spectra of the compe-
tition studies, through the
1
H STD NMR technique, for the
vanadate/maltol (1:2) and HSA system in the presence and
absence of warfarin and ibuprofen, and the corresponding
A
STD
values are presented in Table 2.
The results obtained for this system are in agreement
with those for the vanadate/Hdmpp system. The A
STD
val-
ues in Table 2 show that, again in this system, greater dis-
placement occurs from site I, as revealed by the larger de-
www.eurjic.org FULL PAPER
Figure 3.
1
H NMR spectra of A) a solution containing vanadate/Hdmpp (1:2.5, 1 mm in vanadate) and 0.03 mm HSA to which 0.1 mm
warfarin (structure i) has been added, B) the STD spectrum of the same solution as in A, C) a solution of vanadate/Hdmpp (1:2.5, 1 mm
in vanadate) and 0.03 mm HSA to which 0.4 mm ibuprofen (structure ii) has been added and D) the STD spectrum of the same solution
as in C. A selective saturation pulse (5 s) was applied to the 0 ppm region of the protein. A 30 ms spin-lock pulse was calibrated to avoid
unwanted protein resonances. In this study, a solution of vanadate/Hdmpp, 1 mm in vanadate and an M/L ratio of 1:2.5 was used to
allow the observation of the resonances assigned to the free ligand as well as those of the 1:1 and 1:2 species, and thus to identify the
behaviour of the three species relative to protein binding. Competitor resonances are indicated by * (warfarin) and # (ibuprofen). The
difference in the vertical scale of the reference (A,C) and STD (B,D) spectra is due to the different number of scans used in the respective
acquisitions (see the Exp. Sect.).
Table 2. Values of A
STD
for all protons of the species [VO
2
(dmpp)(H
2
O)(OH)]
–
and [VO
2
(maltol)
2
]
–
, which preferentially bind to HSA,
after the addition of 0.4 mm ibuprofen or 0.1 mm warfarin to solutions containing the vanadate/Hdmpp,maltol systems and the protein.
The percentage of displacement is reflected by the decrease in the A
STD
value upon the addition of the inhibitor.
A
STD
Displacement by competition [%]
6-H 5-H CH
3
N-CH
3
6-H 5-H CH
3
N-CH
3
[VO
2
(dmpp)(H
2
O)(OH)]
–
10.67 16.42 9.67 6.73
vs. Ibuprofen 2.41 2.18 2.36 2.44 77 87 76 64
vs. Warfarin 0.02 0.04 0.03 0.06 99 98 99 99
[VO
2
(maltolato)
2
]
–
11.7 11.4 3.5 –
vs. Ibuprofen 6.1 5.8 2.6 – 48 49 26 –
vs. Warfarin 3.4 2.4 1.2 – 71 79 66 –
crease in the A
STD
values for all the protons of the [VO
2
-
(maltolato)
2
]
–
complex in the presence of warfarin. Al-
though these results do not exclude the possibility of bind-
ing to any additional sites of the protein not screened in this
study,
[50]
a common behaviour towards the HSA binding of
these V
V
species with a similar chemical structure is demon-
strated, with a slight preference of this type of compound
for site I relative to site II.
To complement the STD NMR spectroscopic data with a
more detailed structural representation, molecular docking
studies were carried out between HSA and each of the V
V
species (small molecules) by using HADDOCK.
[47,48]
Fig-
Eur. J. Inorg. Chem. 2013, 4619–4627 © 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim4623
ure 5 shows the 3D docking models for the binding of the
small molecules under study to HSA site I obtained from
HADDOCK calculations.
The results are in agreement with the NMR spectro-
scopic data, supporting the NMR observation of the pre-
dominant interactions of [VO
2
(dmpp)(H
2
O)(OH)]
–
and
[VO
2
(maltol)
2
]
–
complexes with HSA, as shown by the rela-
tively low HADDOCK score for these species. The protein
residues close to and interacting with the small molecules
remain conserved in all the calculations, which reflects the
consistency of the results. As shown in Figure 5, the aro-
matic moiety is in close proximity to the hydrophobic resi-
![Figure 5. HADDOCK models for the interaction of HSA (warfarin pocket, drug site I) with small molecules: A) the free ligand Hdmpp, B) the free ligand maltol, C) [VVO2(dmpp)2]–, D) [VVO2(maltolato)2]– and E) [VVO2(dmpp)(OH)(H2O)]–. The main interacting residues are shown by stick representation and electrostatic and van der Waals interactions are represented by blue and pink dashed lines, respectively. The figures were drawn with PyMOL.](/figures/figure-5-haddock-models-for-the-interaction-of-hsa-warfarin-mav48xtf.webp)
![Figure 1. 1H NMR spectra of A) an aqueous solution of 5 mm Hdmpp, B) an aqueous solution of vanadate/Hdmpp (1 mm in vanadate) in an M/L ratio of 1:2 with 30 mm HSA at pH = 7 and C) STD NMR spectrum of solution (B). A spin-lock pulse of 30 ms was used to remove protein resonances. The signal assignments are shown in the figure: L is the free Hdmpp and VL and VL2 are the 1:1 [VO2(dmpp)(H2O)(OH)]– and 1:2 [VO2(dmpp)2]– complexes, respectively. The difference in the vertical scale of the reference (A,B) and STD (C) spectra is due to the different number of scans used in the respective acquisitions (see the Exp. Sect.).](/figures/figure-1-1h-nmr-spectra-of-a-an-aqueous-solution-of-5-mm-xkbygkxq.webp)
![Figure 2. 1H NMR spectra of A) an aqueous solution of 60 mm maltol, B) an aqueous solution of vanadate/maltol (1 mm in vanadate) in an M/L ratio of 1:2 with 30 mm HSA at pH = 7 and C) STD NMR spectrum of solution B. A spin-lock pulse of 30 ms was used to remove protein resonances. The signal assignments are shown in the figure: L is the free maltol and VL and VL2 are the 1:1 [VO2(maltolato)(H2O)(OH)]– and 1:2 [VO2(maltolato)2]– complexes, respectively. The difference in the vertical scale of the reference (A,B) and STD (C) spectra is due to the different number of scans used in the respective acquisitions (see Exp. Sect.).](/figures/figure-2-1h-nmr-spectra-of-a-an-aqueous-solution-of-60-mm-3s1l0pli.webp)