Integrated ESI-MS/EPR/computational characterization of the binding of metal species to proteins: vanadium drug–myoglobin application

Giuseppe Sciortinoab, Daniele Sanna*c, Valeria Ugonea, Jean-Didier Maréchal*b and Eugenio Garribba*a
aDipartimento di Chimica e Farmacia, Università di Sassari, Via Vienna 2, I-07100 Sassari, Italy. E-mail: garribba@uniss.it; Tel: +39 079 229487
bDepartament de Química, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallés, Barcelona, Spain. E-mail: Jeandidier.Marechal@uab.cat; Tel: +34 935814936
cIstituto CNR di Chimica Biomolecolare, Trav. La Crucca 3, I-07040 Sassari, Italy. E-mail: daniele.sanna@cnr.it; Tel: +39 079 2841207

Received 20th February 2019 , Accepted 16th April 2019

First published on 15th May 2019


An integrated experimental/computational strategy to study the binding modes of metal species to proteins is presented. With this multistep method based on the combined application of spectrometric (ESI-MS), spectroscopic (EPR) and computational (docking and QM) techniques, the interaction of VIVO2+ and four potential drugs VIVOL2 (L = 1,2-dimethyl-3-hydroxy-4(1H)-pyridinonate or dhp, L-mimosinate or mim, maltolate or ma, acetylacetonate or acac) with myoglobin (Mb) was characterized. ESI-MS allows the determination of the number of moieties (VOL+ or VOL2) bound to the protein, EPR helps distinguish the type of coordinating donors, and docking and full QM models allow the prediction of the specific residues involved in the V coordination as well as the 3D structure of the adducts. The results indicate that VIVO2+ ions bind to three different sites with the involvement of three residues of the polypeptide chain (His, Asp and Glu). In the systems with dhp and mim, mixed species {VOL2}n(Mb) with n = 2 (dhp) and 3–4 (mim) were formed with the equatorial coordination of one exposed His residue. With ma and acac, adducts with the general formula {VOL}n(Mb) with n = 2 were identified with the simultaneous binding of two residues (Glu, Asp or His) to two adjacent positions of the VOL+ moiety. This approach is generalizable and could be applied to other metal ions and proteins using – depending on the metal features – different spectroscopic techniques.


Introduction

The interactions of metals and coordination compounds with proteins play a major role in biology and medicine.1 If on one hand many metal ions, such as Mn, Fe, Cu, Zn, play various roles in the humans (such as storage, transport or enzymatic functions), on the other hand the discovery of cisplatin, cis-[Pt(NH3)2Cl2],2 has stimulated the research field now known as medicinal inorganic chemistry. Compounds of Gd, Bi, Au and Pt are currently used as contrast agents in magnetic resonance imaging or in the therapy of several diseases, such as ulcer, arthritis and cancer.3 For all these metallic species, the interaction with proteins can determine the transport in the blood, the active species in the cells and their mechanism of action, while activation and deactivation processes are possible after the metal binding.

The assumption that metallodrugs reach the target cells in the same form as they are administered can be considered an oversimplification.4 When the effectiveness and potential medical application of a metal compound is taken into account, information on its species distribution is necessary. In fact, in the blood and cellular environment several chemical processes can occur – depending on the thermodynamic stability of the specific metal compound – such as ligand exchange, complexation and/or, in several cases, redox reactions. In all these biotransformations, proteins play a central role, both for the high affinity toward specific metals and their large concentration in the blood. Among them, human serum transferrin (hTf), human serum albumin (HSA), immunoglobulins (Ig) and hemoglobin (Hb) have special importance.5

In light of these assumptions, the study of how a metal species interacts with proteins is fundamental. This includes determining how many metal moieties bind to proteins and predicting, as accurately as possible, the three-dimensional structure of the complex formed between the metal and protein. NMR and EPR are the most used spectroscopic techniques with diamagnetic and paramagnetic species, respectively. They allow the identification of the residues involved in the metal binding and provide insight into the protein region occupied by the metal,6 but often a complete 3D description and the determination of the specific amino acids bound to the metal are not possible. To support the spectroscopic results, often computational methods were applied and we recently developed a new approach for updating metallocompound–protein docking simulations taking into account the formation of a coordination bond.7

Vanadium compounds have attracted growing interest over the last few years for their possible use in medicine as antiparasitic, antiviral, anti-HIV and antituberculosis agents, but mainly as potential antidiabetic and antitumor drugs.8 Among the antidiabetic compounds, BMOV (bis(maltolato)oxidovanadium(IV)) is usually considered the reference compound for the new molecules with insulin-enhancing action,9 and its derivative BEOV (bis(ethylmaltolato)oxidovanadium(IV)) has reached phase IIa of clinical trials.10 VIVOL2 complexes, where L is a bidentate monoanionic ligand, represent a good model for a pharmacologically active metal complex because under physiological conditions they can exhibit three different types of interactions with proteins: formation of binary VO2+–protein species, formation of ternary VOL–protein or VOL2–protein adducts and – obviously – no binding.5 Among these interactions, the formation of binary and ternary species of VIVO2+ with hTf, HSA, IgG and Hb,11,12 the inhibition of phosphatase by H2VVO4 which is the basis of the V antidiabetic activity,13,14 and the reduction of VV in the cytosol to VIVO species15 must be highlighted. Therefore, several types of V–protein binding are possible, and the limited knowledge of the species formed after this interaction is one of the factors due to which V compounds are not in the pipeline of the pharmaceutical companies.16

Myoglobin is a small and well characterized protein, found in heart and skeletal muscles, where it stores and transports molecular oxygen, and consists of a polypeptide of 153 amino acid residues and a heme group.17 In blood, myoglobin is mainly bound to plasma globulins, in particular haptoglobin which prevents the negative consequences of its circulation in the extracellular environment.18 Mb is also expressed in different tumors with hypoxia, such as renal cell, prostate, breast and non-small cell lung cancer,19 and, for this reason, increasing attention has been focused on its possible use as a marker in the diagnosis of tumors.20 Myoglobin has also been used as a model to obtain information on the binding of the metal species to larger proteins.21

In this work, the interaction with myoglobin of the VIVO2+ ion and four important VIVO compounds with antidiabetic and antitumor activity, [VO(dhp)2], cis-[VO(mim)2(H2O)]2−, cis-[VO(ma)2(H2O)], and [VO(acac)2], where dhp, ma, mim and acac indicate 1,2-dimethyl-3-hydroxy-4(1H)-pyridinonate, L-mimosinate (a pyridinone derivative), maltolate and acetylacetonate (see Scheme S1 of the ESI) was studied by an integrated method based on the application of spectrometric (ESI-MS), spectroscopic (EPR) and computational (DFT and docking) techniques. Going beyond our previous studies on V systems, in which we determined the most stable binding site and enantiomeric preference of proteins for VOL2 species,7a,22 here we present a multistep approach to evaluate the binding of all the possible species generated in solution by a V drug candidate, VO2+, VOL+ and VOL2. In particular, ESI-MS allows the determination of the number of moieties (VOL+ or VOL2) bound to Mb, EPR can be used to distinguish the type of residues involved in the coordination (His, Asp/Glu, Ser/Thr, Cys, etc.) of VO2+, VOL+ and VOL2 fragments and computational techniques help predict which specific residues interact with the metal ion as well as provide a 3D representation of the adducts. Three goals are expected from this paper: (i) the results could provide a major step forward in understanding the interaction of pharmacologically active V compounds with globular proteins, (ii) the approach could be considered the basis for other investigations of the V drug binding to proteins and (iii) the presented methodology could be applied to the study of systems formed by other metal complexes and other proteins.

Results and discussion

Interaction of the VIVO2+ ion with myoglobin

To the best of our knowledge only two papers on the system formed by oxidovanadium(IV) and myoglobin have been reported. Tamura et al. replaced the Fe2+ ion with VIVO2+ in the porphyrin ring,23 and Sakurai et al. described the enhanced oxygen affinity of myoglobin in the presence of VIVO2+.24 This last effect was explained assuming that VIVO2+ can reduce, being oxidized to VV, Fe3+ of met-Mb to Fe2+ which, having a high affinity for O2, forms oxy-Mb.24

Anisotropic EPR spectra with VIVO2+/Mb molar ratios of 1/1, 2/1 and 3/1 were recorded (Fig. 1, traces a–c) and interpreted on the basis of the shift of MI = 7/2 resonances, those more sensitive to the change in the equatorial coordination. Mb used in our study contains iron in the reduced form and we have not observed any redox reaction between Fe and V. The signals due to Fe2+ do not interfere with the resonances of VIV, allowing an easy analysis of the spectra. The results can be compared with those obtained with hemoglobin (Fig. 1, traces d–e), which were interpreted some years ago assuming the presence of two sites with different coordination modes and values of the 51V hyperfine coupling constant Az. They were named β and γ with Az values of 166.8 × 10−4 cm−1 and 163.3 × 10−4 cm−1, respectively.25 On the basis of the DFT calculations on VIVO2+ model complexes (using acetate as a model for Asp or Glu coordination and 1-methylimidazole for His binding), the coordination modes (His-N; Asp/Glu-COO; H2O; H2O) for the site β of Hb and (His-N; His-N; Asp/Glu-COO; H2O) for that named γ were proposed.25 However, no information on the localization of the binding site and on the specific residues involved was obtained with EPR and DFT data. With myoglobin, the behaviour is comparable and EPR spectra change with the ratio VIVO2+/Mb. When the ratio is 1/1 only one site with gz = 1.950 and Az = 163.8 × 10−4 cm−1 (MI = 7/2 resonances indicated with I in Fig. 1) was revealed, while at ratios 2/1 and 3/1 the resonances of another site with gz = 1.946 and Az = 167.0 × 10−4 cm−1 (II) were detected. Obviously, these data can be interpreted assuming that the first VIVO2+ ion binds to the stronger site and the second ion interacts with a weaker site.


image file: c9qi00179d-f1.tif
Fig. 1 High field region of the X-band anisotropic EPR spectra recorded on frozen solutions (120 K) containing: (a) VIVO2+/Mb 1/1 (VIVO2+ 1.0 × 10−3 M); (b) VIVO2+/Mb 2/1 (VIVO2+ 1.0 × 10−3 M); (c) VIVO2+/Mb 3/1 (VIVO2+ 1.0 × 10−3 M); (d) VIVO2+/Hb 1/1 (VIVO2+ 6.2 × 10−4 M) and (e) VIVO2+/Hb 2/1 (VIVO2+ 6.2 × 10−4 M). With I and II the MI = 7/2 resonances of the sites γ and β are indicated.

While EPR provides information on the type of residue bound to V (His and Asp/Glu for Mb), docking is able to distinguish the specific amino acids involved in the metal binding through direct coordination bonds.7,22 To identify the residues that interact with the VIVO2+ ion, a preliminary analysis of the potential binding regions, where at least three potential coordinating residues featured a β-carbon within a distance of 2.5 Å from each queried grid point, was accomplished.26 The zones that satisfied these criteria were further analyzed with docking calculations applying the Gold rotamer libraries27 for the selected residues and their neighbours as described in the Experimental and computational section. The results suggest that the sites with at least three residues coordinated (and, therefore, with a rather high thermodynamic stability) are only three: one site with two His bound to V and two sites with only one His (Table 1). In analogy with what was done for the sites found with Hb,25 they were named γ (2 His coordinated to V) and β (1 His coordinated to V). The γ site involves the coordination of His24, His119 and Asp20 and a water molecule which completes the equatorial coordination in a square pyramidal structure (Fig. 2a). For the β sites there are two possibilities (named β1 and β2): in the first one His82, Glu83, Asp141 and one water molecule interact with the VIVO2+ ion (β1, Fig. 2b), while in the second one His81 with a N donor, Glu83 with a COO group, Gly80 with a carbonyl CO of the backbone and a H2O molecule are involved (β2, Fig. 2c). Even though the first option has lower scoring and population than the second one, it is proposed here because it is in the terminal part of the protein, whose great mobility would allow the structure to relax significantly.


image file: c9qi00179d-f2.tif
Fig. 2 QM optimized binding sites of Mb for the VIVO2+ ion: (a) site γ; (b) site β1 and (c) site β2.
Table 1 Binding sites of Mb for the VIVO2+ ion determined by docking methods
Site Residues Coordination V–Na V–Oa Fmaxb Fmeanc Pop.d
a Distance in Å.b Fitness value for the most stable pose of each cluster (Fmax).c Mean Fitness value of the GoldScore scoring function for each cluster (Fmean).d Percent population of the cluster.
γ His24; His119; Asp20 N, N, COO; H2O 2.174, 2.730 2.499 43.5 42.7 62%
β1 His82; Glu83; Asp141 N, COO, COO; H2O 2.474 2.700, 2.046 24.2 24.2 2%
β2 His81; Glu83; Gly80 N, COO, CO; H2O 2.356 2.258, 2.894 28.8 28.1 24%


To further validate the reliability of the docking approach, the most stable solutions of the docking runs were refined by the QM cluster method28 (structures presented in Fig. 2). ΔG for the three sites was calculated for the reaction:

 
[VO(H2O)4]2+ + Site β/γ ⇄ VO(Site β/γ)(H2O) + 3H2O (1)

The values of ΔGaq, always negative, are in good agreement with those based on pure scoring and population of GoldScore terms: −31.7 kcal mol−1 for γ, −31.9 kcal mol−1 for β1, and −29.1 kcal mol−1 for β2. The order of binding affinity for the three sites is γ ∼ β1 < β2, which means that the γ and β1 sites will populate before the β2 site (see Fig. 1). QM refinement allowed us to improve the prediction of the bond angles and distances with respect to the docking (Table 2), which are comparable to those expected for V–N(imidazole), V–O(carboxylate) and V–O(water) lengths.29 Overall, the site γ resembles that of V-substituted carboxypeptidase, which has two His (with the two rings parallel and perpendicular to the V[double bond, length as m-dash]O bond), one Glu and one H2O and is characterized by an Az of 165.9 × 10−4 cm−1 (to be compared with 163.8 × 10−4 cm−1 for Mb).30 Furthermore, the experimental Az value for the three sites γ, β1 and β2 were compared with those expected on the basis of the ‘additivity relationship’31 and calculated by DFT methods: the results indicate that the order of Az is γ < β1 < β2, as suggested by the experimental data.

Table 2 Structural, spectroscopic and energetic data for the binding sites of the VIVO2+ ion to Mb
Site V–Na V–Oa V–OH2a gzexptl Azexptl[thin space (1/6-em)]b Azexpctd[thin space (1/6-em)]b,c Azcalcd[thin space (1/6-em)]b,d ΔGaqe
a Distance in Å.b A in 10−4 cm−1.c Az expected on the basis of the ‘additivity relationship’.d Az calculated by DFT methods.e ΔG in kcal mol−1.
γ 2.087, 2.112 1.966 2.074 1.950 163.8 167.9 171.0 −31.7
β1 2.106 1.948, 2.002 2.047 1.946 167.0 169.9 174.0 −31.9
β2 2.103 1.998, 2.003 2.055 1.946 167.0 171.3 178.0 −29.1


Interaction of VO(dhp)2 with Mb

The antidiabetic action of the bis-chelated VIVO complex of dhp has been demonstrated from several years in terms of free fatty acid release from isolated rat adipocytes and an increase of glucose uptake both in vitro and in vivo.32,33 Recently, the antitumoral effect in malignant melanoma cell lines has been discussed.34 Potentiometric and EPR results suggest that, in the binary system and in the pH range 5–8, the bis-chelated species is present in solution as a mixture of the square pyramidal [VO(dhp)2] and distorted octahedral cis-[VO(dhp)2(H2O)], in equilibrium between each other.35

The full ESI-MS spectrum of myoglobin (Fig. S1a of the ESI) is characterized by several groups of peaks that correspond to the different charge states (from +6 to +15) of Mb. In the deconvoluted spectrum, two series of peaks can be recognized with experimentally determined masses of 16[thin space (1/6-em)]951 (weak signal) and 17[thin space (1/6-em)]566 Da (strong signal) (Fig. S2). The mass difference is that between the holo and apo forms of myoglobin and this suggests that heme remains bound to the protein under the used experimental conditions.36

The ESI-MS spectra were measured on the system [VO(dhp)2]/Mb with protein concentrations of 5 and 50 μM and different molar ratios. The raw spectrum at the ratio of 3/1 with a concentration of 5 μM is reported in Fig. S1b. The comparison with the raw spectrum of myoglobin in Fig. S1a shows that the charge state distribution does not change significantly after the VOL+/VOL2 binding, suggesting that – as pointed out by Fenn and by Dobo and Kaltashov37,38 – the conformation of the protein remains almost unaltered. The region with m/z between 1950 and 2010 (peak of charge +9) provides other interesting information (Fig. S3): each peak is separated by the other ones by m/z in the range 2.4–2.6, which corresponds to one Na+ ion and is further split by ca. 0.1 m/z due to the different isotopic distribution of the adducts formed. It can also be observed that the most intense peak in the spectrum of myoglobin falls at m/z 1967.3, while it is revealed at m/z 1959.9 in the ternary system, the difference of approximately 66 Da being due to the replacement of three Na+ (23 Da) with three H+ (1 Da). Moreover, the additional peaks at m/z 1982.6 and 1998.1 can be attributed to the adducts formed by [VO(dhp)+] (mass: 205.1 Da) and [VO(dhp)2] (mass: 343.2 Da) with Mb. Overall the results and analysis are comparable with those for the interaction of cisplatin with insulin.39 An ultrazoom of the two most intense peaks with charge +9 recorded in the systems with Mb only and with Mb and [VO(dhp)2] is presented in Fig. S4a and S4b.

The deconvoluted ESI-MS spectra measured on the system [VO(dhp)2]/Mb with a protein concentration of 5 μM and molar ratios of 3/1 and 5/1 are reported in Fig. 3. The multipeak signal in the range 17[thin space (1/6-em)]560–17[thin space (1/6-em)]740 Da is attributed to the holo form of the protein and the series of peaks around the more intense signal at 17[thin space (1/6-em)]630 Da to the Na+ ions interacting with the protein. The deconvoluted spectra display the same ion distribution observed in the raw spectra, with a series of peaks around that more intense. The formation of species with compositions {n[VO(dhp)2]}–Mb (n = 1–2) and {[VO(dhp)+] + [VO(dhp)2]}–Mb was observed, where VO(dhp)+ and VO(dhp)2 moieties are bound to myoglobin. It can be noticed that an increase in the [VO(dhp)2]/Mb ratio increases the number of fragments (VOL+ or VOL2) bound to the protein and, when the ratio is 5/1, the protein can bind up to 2 VO(dhp)+/VO(dhp)2 moieties. If the spectra are recorded with a protein concentration of 50 μM, a similar behaviour is observed, with a slight increase of the relative amount of all the detected adducts. Thus, the ESI-MS technique indicates that 2 moieties of VO(dhp)2 can interact with myoglobin.


image file: c9qi00179d-f3.tif
Fig. 3 Deconvoluted ESI-MS spectra recorded on the system containing [VO(dhp)2] and myoglobin (5 μM): molar ratio 3/1 (top) and 5/1 (bottom). With a and b the moieties VO(dhp)+ and VO(dhp)2 are indicated. Mass is expressed in Da.

Anisotropic EPR spectra were recorded in the ternary system with VIVO2+, Mb and dhp at pH 7.4 as a function of the ratio VO(dhp)2/Mb (traces b–g in Fig. 4). In the system containing the protein the resonances of a species not present in the binary system were revealed (II in Fig. 4). The value of Az (162.7 × 10−4 cm−1) is intermediate between those of cis-[VO(dhp)2(H2O)] and [VO(dhp)2] and is practically coincident with that of cis-[VO(dhp)2(MeIm)] (III, 163.0 × 10−4 cm−1[thin space (1/6-em)]11b,c), in which the equatorial water molecule in the cis-octahedral species is replaced by 1-methylimidazole, a well-known model for His coordination.11b,c,12b This indicates the formation of a mixed species with composition VO(dhp)2(Mb), where an accessible His-N binds equatorially the VO(dhp)2 moiety, similar to what was observed for other proteins, such as apo- and holo-hTf, HSA, IgG and Hb.11b,d,h,25a The percent amount of this species depends on the ratio VO(dhp)2/Mb: at low molar ratios (for example 1/1) VO(dhp)2(Mb) is the major species in solution, but at higher ratios (from 2/1) the percentage of [VO(dhp)2] (Ia) and cis-[VO(dhp)2(H2O)] (Ib) significantly increases. In other words, the first two equivalents of cis-VO(dhp)2 find His residues suitable for metal coordination, forming the adduct II, while the next ones are not able to interact with the protein and the signals of the two species Ia and Ib appear. This is revealed when the ratio is VO(dhp)2/Mb ∼ 2, indicating that the general stoichiometry of the adducts may be denoted as {VO(dhp)2}n(Mb), with n ∼ 2. The whole process, presented in Scheme 1, can be summarized stating that at low VO(dhp)2/Mb ratios the equilibrium is shifted to the right due to the formation of the protein adduct (II), but on increasing this ratio – when no binding sites are available – it shifts to the left with the formation of [VO(dhp)2] (Ia) and cis-[VO(dhp)2(H2O)] (Ib).


image file: c9qi00179d-f4.tif
Fig. 4 High field region of the X-band anisotropic EPR spectra recorded on frozen solutions (120 K) containing: (a) VIVO2+/dhp/MeIm 1/2/4; (b) VIVO2+/dhp/Mb 1/2/1; (c) VIVO2+/dhp/Mb 2/4/1; (d) VIVO2+/dhp/Mb 4/8/1; (e) VIVO2+/dhp/Mb 6/12/1; (f) VIVO2+/dhp/Mb 8/16/1; (g) VIVO2+/dhp/Mb 10/20/1 and (h) VIVO2+/dhp 1/2. VIVO2+ concentration was 1.0 × 10−3 M. II and the dotted line indicate the MI = 7/2 resonances of the adduct {VO(dhp)2}n(Mb), and III that of the model species cis-[VO(dhp)2(MeIm)], while with Ia and Ib and the dash-dotted lines the MI = 7/2 resonances of [VO(dhp)2] and cis-[VO(dhp)2(H2O)] are denoted. With the bold line the spectrum with the maximum value of n for the adduct {VO(dhp)2}n(Mb) is also shown.

image file: c9qi00179d-s1.tif
Scheme 1 Complexation process of VIVO complexes formed by dhp (L) with myoglobin.

To complement the results obtained by ESI-MS spectrometry and EPR spectroscopy, docking calculations were carried out. The structures of the eight enantiomers OC-6 of [VO(dhp)2(H2O)] (i.e., the Δ and Λ series of OC-6-34, OC-6-34, OC-6-23 and OC-6-24[thin space (1/6-em)]40) were optimized by DFT, the equatorial water was removed and the moiety cis-VO(dhp)2 – where one of the four equatorial sites is free to interact with the protein – was docked with Mb (PDB: 4DC8[thin space (1/6-em)]41) using the parameters recently optimized for vanadium.7 To identify the candidate residues, a relative SES (Solvent Excluded Surface) calculation was preliminarily performed on the X-ray structure of Mb, which shows that the potential donors are His81, His113, His116, and His119. GOLD rotamer libraries27 were applied during the calculations for the identified histidines and the neighboring residues. The results were interpreted on the basis of the values of Fmax and Fmean and of the population of each pose (Table 3) and indicate that only His81 and His113 are able to interact with cis-VO(dhp)2. The most stable structures are shown in Fig. 5; the isomers that show stronger interactions are OC-6-32-Δ with His81 and OC-6-34-Λ with His113. As can be observed, the plane of the aromatic ring of histidine is almost parallel to V[double bond, length as m-dash]O.


image file: c9qi00179d-f5.tif
Fig. 5 Most stable adducts predicted by docking methods for the interaction of the cis-VO(dhp)2 moiety with myoglobin: (a) OC-6-32-Δ-VO(dhp)2 with His81 and (b) OC-6-34-Λ-VO(dhp)2 with His113.
Table 3 Binding sites for the interaction of cis-VOL2 moieties with myoglobina
Species His81 His113 His116 His119
Fmaxb Fmeanc Pop.d Fmaxb Fmeanc Pop.d Fmaxb Fmeanc Pop.d Fmaxb Fmeanc Pop.d
a With boldface text, the histidine residues which could interact with the VOL2 moiety are indicated.b Fitness value for the most stable pose of each cluster (Fmax), expressed as the mean obtained on all the possible enantiomers for VOL2 species.c Mean Fitness value of the GoldScore scoring function for each cluster (Fmean).d Percent population of the cluster.
VO(dhp)2 45.3 42.0 98.3% 49.4 47.4 99.8% 34.3 33.1 60.8% 33.1 31.5 1.0%
VO(mim)22− 56.3 47.9 49.3% 66.3 52.8 31.5% 44.9 39.6 36.3% 64.1 57.0 4.0%


Finally, to model the binding of the VO(dhp)+ moiety to Mb, docking simulations were performed on the enantiomers SPY-5-13-A and SPY-5-13-C optimized by DFT methods, after removing the two H2O molecules and activating the two equatorial sites (see the Experimental and computational section). The results show three potential sites with the simultaneous coordination of His24 and His119, His82 and Asp141, and Glu83 and Asp141, the order of relative affinity being (His24, His119) > (His82; Asp141) > (Glu83; Asp141). The data are listed in Table 4 and the three sites are presented in Fig. S5 of the ESI. To the best of our knowledge this is the first case of a completely characterized binding of a VOL+ species to a protein with the contemporaneous coordination of two accessible residues.

Table 4 Docking solutions for the interaction of VO(dhp)+ with myoglobin
Site Residues Isomer V–Da Fmaxb Fmeanc Pop.d
a Distance in Å; D = N, O.b Fitness value for the most stable pose of each cluster (Fmax).c Mean Fitness value of the GoldScore scoring function for each cluster (Fmean).d Percent population of the cluster.
1st His24; His119 SPY-5-13-A 2.212, 2.322 40.8 38.5 10%
SPY-5-13-C 2.103, 2.390 46.5 42.3 37%
2nd His82; Asp141 SPY-5-13-A 2.437, 2.434 36.6 32.6 28%
SPY-5-13-C 2.300, 2.334 34.3 32.5 23%
3rd Glu83; Asp141 SPY-5-13-A 2.474, 2.586 28.8 27.8 8%


Interaction of VO(mim)2 with Mb

L-Mimosine (mim), α-amino-β-(3-hydroxy-4-oxo-1,4-dihydropyridin-1-yl)propanoic acid or leucenol is a relatively rare plant α-amino acid. It is a specific inhibitor of initiation of DNA replication,42 tyrosinase,43 and other processes, such as folate and deoxyribonucleotide metabolism.44 Over the last few years, L-mimosine has been found to have several biological properties such as anticancer, antiinflammatory, antifibrosis, antiinfluenza, antiviral, herbicidal and insecticidal action.45 L-Mimosine is a strong binder toward iron46 and it is an interesting ligand because two alternative coordination modes are possible: one “pyridinone-like” and another “α-amino acid-like”. With respect to dhp, discussed in the previous section, the presence of the amino acid side chain on the pyridinone ring – able to form hydrogen bonds with protein residues – could change the type of interaction with myoglobin.

With the VIVO2+ ion at pH 7.4, L-mimosine forms a stable bis-chelated cis-octahedral species with the ligand in the coordination mode (CO, O), similar to dhp and maltol.47 Its stoichiometry is [VO(mim)2(H2O)]2−, with the carboxylic group of the amino acid function in the deprotonated form, and the spin Hamiltonian parameters are gz = 1.943 and Az = 169.0 × 10−4 cm−1.47 The presence of VO(mim)22− in aqueous solution has been confirmed by ESI-MS spectrometry. The list of the species identified in the system VIVO2+/mim 1/2 is reported in Table S1 of the ESI and the simulation of the isotopic pattern for the peak attributed to [VO(mim)2]2−, revealed at m/z = 229.5 in the negative-ion mode ESI-MS spectrum, is displayed in Fig. S6. As pointed out in the literature, the water molecule in the equatorial position of the VIVO2+ ion could not be revealed because a monodentate solvent ligand can be removed from the first metal coordination sphere during the ionization process.48 To confirm the cis-octahedral arrangement of [VO(mim)2]2−, the values of Az were calculated for all the possible isomers (the two square pyramidal isomers SPY-5-12 and SPY-5-13, and the four cis-octahedral isomers OC-6-23, OC-6-24, OC-6-32 and OC-6-34, Scheme S3) by DFT methods and compared with the experimental ones. From the data it emerges that the formation of square pyramidal isomers can be excluded and that the best agreement between calculated and experimental Az is found for OC-6-34 and OC-6-24 (Table S2).

The raw ESI-MS spectrum with a molar ratio of V/Mb of 3/1 and a Mb concentration of 5 μM are reported in Fig. S1c of the ESI and a zoomed region with the two most intense peaks with charge +9 in Fig. S4c. It can be observed that the charge distribution is significantly different from that revealed in the systems with myoglobin alone and with [VO(dhp)2] and Mb, indicating that the protein conformation changes after the binding of the V complex. It is plausible to suppose that this conformational change favors the exposure on the protein surface of a larger number of sites which can be protonated during the ionization, explaining the presence of peaks with higher charge.37,38 Moreover, the m/z values for the peak with charge +9 shifts from 1959.9 (dhp) to 1952.8 (mim), suggesting the replacement of three Na+ ions with three H+ ions.

The deconvoluted ESI-MS spectra recorded on the system VO(mim)2/Mb with different molar ratios (3/1 and 5/1) and a protein concentration of 50 μM are reported in Fig. 6. It can be noticed that, besides the peak of the free protein at ca. 17[thin space (1/6-em)]610 Da, different adducts are formed. In particular, up to 3 fragments of VO(mim)22− (with a mass of 459 Da) can be bound simultaneously to the protein. The following species were observed: {n[VO(mim)2]}–Mb (n = 1–3) and {n[VO(mim)2] + [mim]}–Mb (n = 0–2). The detection of the adduct {n[VO(mim)2] + [mim]}–Mb implies that a part of the ligand is in the free form (due to partial hydrolysis of VO(mim)2 at the concentrations used) and that it binds to the protein. This interaction was not observed in the other systems and this can be related to the specific structure of L-mimosinate which has polar side-chain groups (NH2/NH3+ and COO) able to form hydrogen bonds with myoglobin.


image file: c9qi00179d-f6.tif
Fig. 6 Deconvoluted ESI-MS spectra recorded on the system containing cis-[VO(mim)2(H2O)]2− and myoglobin (50 μM): molar ratio 3/1 (top) and 5/1 (bottom). With L and b the moieties mim2− and VO(mim)22− are indicated. Mass is expressed in Da.

As in the system with dhp, the EPR spectra change with the ratio VO(mim)2/Mb (Fig. 7). When this ratio is 4, only the resonances of species II are detected. These were attributed to {VO(mim)2}n(Mb), where the moiety cis-VO(mim)22− interacts with His residues of myoglobin. The comparison of the Az of VO(mim)2(Mb) (trace b of Fig. 7, gz = 1.947 and Az = 163.9 × 10−4 cm−1) with that of cis-[VO(mim)2(MeIm)]2− (III in the trace a, gz = 1.947 and Az = 164.3 × 10−4 cm−1) confirms His binding. The resonances of the binary complex cis-[VO(mim)2(H2O)]2− (I) can be observed at a VO(mim)2/Mb ratio ∼ 4 (which must be compared with the system with dhp, for which this happens with a lower ratio, ∼2), indicating that there is a larger number of His residues that bind VO(mim)2; this is in agreement with the ESI-MS data.


image file: c9qi00179d-f7.tif
Fig. 7 High field region of the X-band anisotropic EPR spectra recorded on frozen solutions (120 K) containing: (a) VIVO2+/mim/MeIm 1/2/4; (b) VIVO2+/mim/Mb 1/2/1; (c) VIVO2+/mim/Mb 2/4/1; (d) VIVO2+/mim/Mb 4/8/1; (e) VIVO2+/mim/Mb 6/12/1; (f) VIVO2+/mim/Mb 8/16/1; (g) VIVO2+/mim/Mb 10/20/1 and (h) VIVO2+/mim 1/2. VIVO2+ concentration was 1.0 × 10−3 M. II and the dotted line indicate the MI = 7/2 resonances of the adduct {VO(mim)2}n(Mb), and III that of the model species cis-[VO(mim)2(MeIm)]2−, while with I and the dash-dotted line the MI = 7/2 resonances of cis-[VO(mim)2(H2O)]2− are denoted. With the bold line the spectrum with the maximum value of n for the adduct {VO(mim)2}n(Mb) is also shown.

The comparison of the results obtained by ESI-MS and EPR techniques in the systems with 1,2-dimethyl-3-hydroxy-4(1H)-pyridinonate and L-mimosinate at first sight could seem strange, considering that mim is a more hindered ligand than dhp and that, for this reason, a lower number of fragments bound to myoglobin should be expected. To throw light on this aspect, docking calculations were carried out (Table 3). These suggest that, in contrast with VO(dhp)2, at least three of the exposed histidines (His81, His113, and His119) and probably also His116 can interact with VO(mim)2 with Fmax values being significantly higher than those of dhp (the mean values of Fmax calculated on the four histidines are 40.5 for VO(dhp)2 and 57.9 for VO(mim)2); the adduct formed can be indicated as {VO(mim)2}n(Mb), with n = 1–3 or 1–4. The reason for the higher affinity of VO(mim)2 for His lies in the formation of a network of hydrogen bonds between the non-coordinating amino group of L-mimosinate and polar residues of Mb (particularly Lys79 and Glu83), which is not predicted with VO(dhp)2. This is illustrated in Fig. 8 for the binding of His81. Interestingly, the residues His113, His116, and His119 are those proposed by Zhao et al. and Moreno-Gordaliza et al. for the interaction with myoglobin of the hydrolysis products of cisplatin and transplatin, cis-[Pt(NH3)2(H2O)2] and trans-[Pt(NH3)2(H2O)2].49,50


image file: c9qi00179d-f8.tif
Fig. 8 Most stable adducts predicted by docking methods for the interaction of the moieties cis-VO(dhp)2 and cis-VO(mim)22− with the residue His81 of myoglobin: (a) OC-6-32-Δ-VO(dhp)2 and (b) OC-6-34-Δ-VO(mim)2. The hydrogen bonds between the amino group of the L-mimosinate ligand and the residues Lys79 and Glu83 are highlighted with orange lines.

Interaction of VO(ma)2 with Mb

As mentioned in the Introduction, the antidiabetic properties of [VO(ma)2] are well-known and it is considered the benchmark compound for all the insulin-enhancing potential agents.9,10,51 In the binary system VIVO2+/maltol with a molar ratio of 1/2, only the complex cis-[VO(ma)2(H2O)], with an equatorial–equatorial and equatorial–axial arrangement of the ligands, exists at pH 7.4.52

ESI-MS spectra of the system [VO(ma)2]/Mb were recorded at different protein concentrations (5 and 50 μM) and different molar ratios between the complex and Mb (3/1, 5/1 and 10/1). The spectrum and a selected region with the two most intense peaks of the adducts with charge +9 are shown in Fig. S1d and S4d of the ESI. The distribution of the charge states is different compared with those of the systems with Mb and [VO(dhp)2]/Mb and similar to that revealed with L-mimosinate. This confirms two different conformations of myoglobin depending on the V complex bound (the first stabilized by the interaction with the V species of dhp, and the second one by the binding of V complexes of mim and ma).

The deconvoluted spectra recorded with a protein concentration of 50 μM (Fig. S7 of the ESI) show that both the fragments VO(ma)+ (192 Da) and VO(ma)2 (317 Da) can bind to the protein under these experimental conditions. In particular, the species {n[VO(ma)+]}–Mb (with n = 1–4), [VO(ma)2]–Mb and a small amount of {n[VO(ma)+] + [VO(ma)2]}–Mb (with n = 1–2) are detected.

The most important difference between the spectra recorded in the systems containing [VO(dhp)2] and [VO(ma)2] is that, in the first case the interaction of the fragment VO(dhp)2 with Mb is favored, while in the second one the preferential binding of VO(ma)+ compared to VO(ma)2 is observed. This could depend on the lower thermodynamic stability of VO(ma)2 than that of VO(dhp)2.35a,52

The formation of the species {VO(ma)}n(Mb) detected by ESI-MS spectrometry was confirmed by EPR and docking techniques. EPR spectra recorded as a function of pH using a molar ratio of VIVO2+/ma/Mb of 1/1/1 unveils the formation of one species at pH 4.80 (indicated with II in Fig. 9) and another at pH 6.20 (III). This can be explained postulating that the two equatorial water ligands of the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 species [VO(ma)(H2O)2]+ are replaced by two carboxylate groups of Asp or Glu residues at acidic pH values (when histidines are still protonated and cannot interact with the metal ion) and one His and one Asp/Glu or two His at pH above 6.


image file: c9qi00179d-f9.tif
Fig. 9 High field region of the X-band anisotropic EPR spectra recorded on frozen solutions (120 K) containing: (a) VIVO2+/ma 1/1 (pH 5.25); (b) VIVO2+/ma/Mb 1/1/1 (pH 4.80); (c) VIVO2+/ma/Mb 1/1/1 (pH 5.30); (d) VIVO2+/ma/Mb 1/1/1 (pH 6.20); and (e) VIVO2+/ma/MeIm 1/2/4 (pH 7.40). VIVO2+ concentration was 1.0 × 10−3 M. II and III indicate the MI = 7/2 resonances of the adducts {VO(ma)}n(Mb) with (Glu; Asp) coordination and {VO(ma)}n(Mb) with (His; Asp) or (His; His) coordination, while with I and IV the MI = 7/2 resonances of [VO(ma)(H2O)2]+ and cis-[VO(ma)2(MeIm)] are denoted.

To prove these assumptions, docking calculations were carried out allowing the moiety VO(ma)+, with two vacant coordination sites, to interact with Mb. From the DFT optimized structures of SPY-5-13-A and SPY-5-13-C enantiomers (see Scheme S2 of the ESI), the two equatorial H2O molecules were removed and the coordination vacancies were activated. An analysis of the potential coordination sites was first performed to identify the possible binding regions with at least two residues able to coordinate the VO(ma)+ moiety. Moreover, to explain the experimental EPR spectra in Fig. 9 and distinguish the different coordination modes, two series of simulations were performed taking into account the histidine protonation state assessed by PROPKA simulations:53 at pH 4.80 (in the protonated form and unable to bind VIVO2+) and pH 6.20 (in the neutral form and free to coordinate VIVO2+). The docking results indicate that at pH 4.80 two binding sites are possible, one with Glu83 and Asp141 and another with Glu105 and Asp109, which replace the two water ligands in two adjacent equatorial positions (Fig. 10a and b). At pH 6.2, after the deprotonation of histidine residues, two sites were identified as well: the first one involving His82 and Asp141 is derived from that at pH 4.8 upon the replacement of Glu83 with His82, while the second one is formed by two histidines, His24 and His119 (Fig. 11a and b). Therefore, in agreement with the ESI-MS measurements, the results indicate that there are at least three independent sites able to bind the VO(ma)+ moiety: (Glu83; Asp141)/(His82; Asp141), (Glu105; Asp109) and (His24; His119). The Az values expected for these sites with the ‘additivity relationship’31 (Azexpct) agree well with the experimental values (Table 5).


image file: c9qi00179d-f10.tif
Fig. 10 Most stable adducts predicted by docking methods for the interaction of the VO(ma)+ moiety with myoglobin at a low pH value (with the His residues protonated): (a) SPY-5-13-C-VO(ma)(H2O)2 with Glu83 and Asp141 and (b) SPY-5-13-C-VO(ma)(H2O)2 with Glu105 and Asp109. The hydrogen bonds are highlighted with orange lines.

image file: c9qi00179d-f11.tif
Fig. 11 Most stable adducts predicted by docking methods for the interaction of the VO(ma)+ moiety with myoglobin at a high pH value (with the His residues deprotonated): (a) SPY-5-13-C-VO(ma)(H2O)2 with His82 and Asp141 and (b) SPY-5-13-A-VO(ma)(H2O)2 with His24 and His119.
Table 5 Structural, docking and spectroscopic data for the binding sites of the VO(ma)+ moiety to Mb
pH Residues V–Na V–Oa Fmaxb Pop.c gzexptl Azexptl[thin space (1/6-em)]d Azexpct[thin space (1/6-em)]d,e
a Distance in Å.b Fitness value for the most stable pose of each cluster (Fmax).c Percent population of the cluster.d A measured in 10−4 cm−1.e Az expected on the basis of the ‘additivity relationship’.f Best solution of the first cluster.g Best solution of the second cluster.
4.8 Glu83; Asp141f 2.199, 2.489 39.4 49% 1.945 168.7 166.6
4.8 Glu83; Asp141g 2.145, 2.457 39.2 40% 1.945 168.7 166.6
4.8 Glu105; Asp109 2.474 2.505, 2.221 41.9 39% 1.945 168.7 166.6
6.2 His82; Asp141f 2.571 2.358 34.2 46% 1.947 165.4 165.2
6.2 His82; Asp141g 2.467 2.369 35.9 27% 1.947 165.4 165.2
6.2 His24; His119 2.155, 2.428 41.7 36% 1.947 165.4 163.8


The results obtained from EPR spectroscopy using a higher V concentration (1 mM instead of 5 or 50 μM), which favours the 1[thin space (1/6-em)]:[thin space (1/6-em)]2 species cis-[VO(ma)2(H2O)], are presented in Fig. 12. When 1-methylimidazole is added to an aqueous solution containing cis-[VO(ma)2(H2O)], the mixed species cis-[VO(ma)2(MeIm)] is formed after the replacement of a water ligand by the more basic MeIm (III in trace a of Fig. 12).11e As expected, this results in a decrease of the Az value (for the ‘additivity relationship’, the contribution of an imidazole-N to Az is 40.7 × 10−4 cm−1, while that of water-O is 45.6 × 10−4 cm−1[thin space (1/6-em)]31). In the ternary system with Mb, when the ratio VO(ma)2/Mb is <4 the EPR resonances coincide with that of III, suggesting that a His-N replaces H2O in the equatorial plane and that a species {VO(ma)2}n(Mb) with n = 3–4 is formed (II). This means that only 3–4 residues of His are able to interact with the cis-VO(ma)2 moiety. The spin Hamiltonian parameters for this species are gz = 1.947 and Az = 164.8 × 10−4 cm−1, to be compared with gz = 1.948 and Az = 164.8 × 10−4 cm−1 of cis-[VO(ma)2(MeIm)].11e At molar ratio VO(ma)2/Mb = 4, the resonances of cis-[VO(ma)2(H2O)] appear and this indicates that the moiety cis-VO(ma)2 does not find other accessible histidines and remains free in solution (resonances indicated with I in traces d–f). On the basis of the ESI-MS measurement and docking calculations discussed previously for dhp and mim, the histidine residues involved in the V binding could be His81, His113, His116, and His119.


image file: c9qi00179d-f12.tif
Fig. 12 High field region of the X-band anisotropic EPR spectra recorded on frozen solutions (120 K) containing: (a) VIVO2+/ma/MeIm 1/2/4; (b) VIVO2+/ma/Mb 1/2/1; (c) VIVO2+/ma/Mb 2/4/1; (d) VIVO2+/ma/Mb 4/8/1; (e) VIVO2+/ma/Mb 8/16/1; (f) VIVO2+/ma/Mb 16/32/1 and (h) VIVO2+/ma 1/2. VIVO2+ concentration was 1.0 × 10−3 M. II and the dotted line indicate the MI = 7/2 resonances of the adduct {VO(ma)2}n(Mb), and III that of the model species cis-[VO(mim)2(MeIm)]2−, while with I and the dash-dotted line the MI = 7/2 resonances of cis-[VO(ma)2(H2O)] are denoted. With the bold line the spectrum with the maximum value of n for the adduct {VO(ma)2}n(Mb) is also shown.

Interaction of VO(acac)2 with Mb

Among the V compounds with antidiabetic and anticancer actions, [VO(acac)2] is worth being mentioned. It corrects the hyperglycemia and impairs hepatic glycolysis in streptozotocin-induced diabetic rats (STZ-rats) more effectively than inorganic salt VOSO4,54 and exhibits a greater capacity to enhance insulin receptor (IR) kinase activity in cells than other VIVO chelate complexes.55 Potentiometric and spectroscopic studies indicate that in the system VIVO2+/acac 1/2, around the physiological pH, the VIVO2+ ion is present as the neutral complex [VO(acac)2].56,57

The raw ESI-MS spectrum recorded in the ternary system with Mb is presented in Fig. S1e of the ESI. The charge distribution suggests that the configuration of myoglobin is comparable to that revealed with [VO(mim)2]2− and [VO(ma)2] rather than with [VO(dhp)2]. The zoomed region containing the two most intense peaks with charge +9 is shown in Fig. S4e. The deconvoluted spectra at different molar ratios (3/1 and 5/1) with a protein concentration of 5 μM are reported in Fig. 13. Only the adducts with the composition {n[VO(acac)+]}–Mb were identified, with n = 1–3. It can be noticed that the number of peaks and their relative abundance increase with the ratio between the metal and protein. As was observed in the system with maltol, even in this case the interaction of the mono-chelated 1[thin space (1/6-em)]:[thin space (1/6-em)]1 species [VO(acac)(H2O)2]+ is favoured with respect to that of the bis-chelated [VO(acac)2], and this finding can be related to the intermediate strength of the ligand.56 When the spectra were recorded at a higher concentration (50 μM Mb) an increase in the intensities of all the adduct signals was observed (Fig. S8). It can be noticed that the peaks of the species {n[VO(acac)+] + [VO(acac)2]}–Mb (with n = 1–2) were also identified; this is in agreement with the thermodynamic data which suggest that, with the increase in the metal concentration, the relative amount of VO(acac)2 increases with respect to VO(acac)+.56


image file: c9qi00179d-f13.tif
Fig. 13 Deconvoluted ESI-MS spectra recorded on the system containing [VO(acac)2] and myoglobin (5 μM): molar ratio 3/1 (top) and 5/1 (bottom). With a the moiety VO(acac)+ is indicated. Mass is expressed in Da.

Docking simulations were performed in order to characterize the interaction of the VO(acac)+ moiety with myoglobin and predict the binding of residues to V. In analogy with the previous calculations, the regions with at least two potential coordinating residues were examined applying the GOLD rotamer libraries.27 The results are comparable with those obtained with VO(dhp)+ and VO(ma)+ and are presented in Table S3 of the ESI: at acidic pH the binding site involving Glu83 and Asp141 is predicted (Fmax = 34.5), while at higher pH – when the histidine residues are deprotonated – three different sites were identified, with the most stable involving His24 and His119 (Fmax = 44.8), the second His82 and Asp141 (Fmax = 38.7) and the third one Gln124 (carbonyl donor) and His116 (Fmax = 32.8). The most stable sites around neutrality are shown in Fig. 14. Therefore, the stoichiometry of the species formed is, according to ESI-MS data, {VO(acac)}n(Mb) with n = 3.


image file: c9qi00179d-f14.tif
Fig. 14 Most stable adducts predicted by docking methods for the interaction of the VO(acac)+ moiety with myoglobin at physiological pH: (a) with His24 and His119 and (b) with His82 and Asp141.

These findings confirm the recent results of Costa Pessoa and co-workers – obtained by spectroscopy (CD), mass spectrometry (ESI and MALDI-TOF) and size exclusion chromatography (SEC) – on the interaction of VO(acac)+ with apo-hTf. They revealed a mixed species with the composition VO(acac)(apo-hTf), formed after the binding of the VO(acac)+ moiety in the hTf iron sites of transferrin.12l In light of our data, the formation of the adduct {VO(acac)}n(Mb) is favored by the low concentrations used in the ESI-MS experiments.

Fig. S9 of the ESI shows the EPR spectra recorded in the system VO(acac)2/Mb as a function of the ratio (from 1/1 to 10/1), plus the benchmark systems VIVO2+/acac and VIVO2+/acac/MeIm. In this case, the V concentration (1 mM) favours the 1[thin space (1/6-em)]:[thin space (1/6-em)]2 complex [VO(acac)2] and allows us to rule out the presence of [VO(acac)(H2O)2]+ at physiological pH. No variation is observed and this can be explained taking into account that the geometry of the bis-chelated compound formed by acetylacetonate is square pyramidal rather than cis-octahedral (see Scheme S1): this hinders the coordination of an accessible Mb residue on the equatorial plane of the VIVO2+ ion. In other words, since [VO(acac)2] has no equatorial water molecules to be replaced by other ligands, it does not have any tendency (at least at these V concentrations) to form mixed complexes with the stoichiometry cis-VO(acac)2(Mb), in contrast to what was previously observed for dhp, mim, and ma. This is confirmed by the fact that no ternary species are formed in the system containing VIVO2+, acetylacetonate and 1-methylimidazole. Similar observations were made with other proteins, such as transferrin, albumin, immunoglobulin or hemoglobin.11b,d,h,25a To demonstrate this insight, EPR spectra were also recorded at RT because any strong metal–protein interaction would result in an anisotropic spectrum: in fact, the rotational motion of a species VO(acac)2–Mb would be slower than the EPR timescale.58 The only signal detected at RT is the isotropic one of [VO(acac)2] in solution not bound to the protein (Fig. S10 of the ESI). The absence of a rigid limit or slowly tumbling resonances indicate that [VO(acac)2] does not interact with Mb or that the interaction is so weak that only a very small fraction of it (not observable by EPR spectroscopy) is bound to the protein through a long axial coordination bond.

Docking methods were applied to study the interaction between [VO(acac)2] and myoglobin, using a recent modification of GoldScore to predict surface interactions of VIVO compounds; in particular, it was found that the value of 16–17 GoldScore units marks the transition from a rigid limit EPR spectrum with strong interactions and VIVO complexes bound on the protein surface to a slowly tumbling or an isotropic spectrum with weak or no interactions and VIVO species unbound and free to rotate in solution.58 The docking calculations show several clusters (Fig. S11 of the ESI) with comparable Fitness indicating the absence of any binding specificity; the value of Fmax (11.5) is lower than the limit of 16–17. As pointed out above, the contradiction between these experiments and those of the team of Costa Pessoa, who found an adduct with the composition VO(acac)(apo-hTf) after the interaction of [VO(acac)2] with apo-transferrin,12l can be explained with the different V concentrations used, 1 × 10−3 M in the EPR measurement shown in Fig. S9 and S10 and ca. 10−5 M in the MALDI and SEC measurements carried out by Costa Pessoa et al.,12l which favours the hydrolysis and the formation of VO(acac)+ that – in its turn – can interact with the protein having two accessible equatorial coordination positions.

Binding sites of Mb for VIVO species

Taking all the collected data together, this study provides a unique opportunity to discuss (i) what are the most probable binding sites of VIVO based drugs with Mb and (ii) how the resulting interactions are a function of the type of ligand bound to the metal.

cis-Octahedral vanadium pro-drugs, [VOL2(H2O)], can exchange their weak aqua ligand to form adducts preferentially with His-N. The coordinating residues must be solvent exposed and not sterically hindered and, for this reason, no preferential binding region can be identified. Moreover, neighbouring hydrogen bond donors/acceptors are frequently involved in the stabilization of these adducts, mainly in the case of organic ligands with Lewis complementary groups. Such second coordination sphere interactions are highly responsible for the relative affinity between different VOL2 moieties (for example, VO(mim)2 is, on average, 17 Goldscore unit favoured with respect to VO(dhp)2) as well as the chiral discrimination observed. Similar conclusions have also been reported in the case of bis-chelated quinolone complexes interacting with serum proteins.22 It clearly appears that the interaction of the metallodrugs with carboxylate containing side chains (Asp or Glu) only occurs when histidines are absent or highly buried.7a

Square pyramidal VOL+ moieties can be coordinated in regions in which at least two solvent exposed donor containing amino acid side chains are located featuring β-carbons within a distance of ∼2.5 Å from the potential position of the metal. The relative affinity of the binding sites, obtained from the studies as a function of pH, is: (His-N; His-N) > (His-N; COO-Asp/Glu) > (COO-Asp/Glu; COO-Asp/Glu). As for the adducts with a mono-coordination, the hydrogen bond stabilization plays a pivotal role in the binding recognition process.

Finally, the binding of the bare VO2+ ion is favoured for regions containing two or three residues featuring the β-carbons close to ∼2.5 Å from the potential position of the metal. The set of residues, taking into account the small size of VO2+, can be exposed on the surface as well as buried into the protein scaffold. The metal first coordination sphere is filled by H2O or OH ligands depending on the biological pH. Based on this and other published studies,5,59,60 the preferential coordination mode of VO2+ in a biological environment (at physiological pH) is mainly represented by His-N and Asp/Glu-COO donors.

A general overview of the Mb binding sites is given in Fig. 15, in which the coordinating residues are highlighted depending on their coordination capability toward the formation of the adducts cis-VOL2(Mb), VOL(Mb) or VO(Mb)(H2O). Comparing the behaviour of VOL2, VOL+ and VO2+, it can be concluded that the number of VOL2 moieties interacting with Mb is larger than those of VO2+ and VOL+ since the histidines bind independently if they are exposed on the protein surface; in contrast, only three binding regions are possible for the interaction of VOL+ and VO2+ since two or three neighbouring residues, respectively, must be able to bind V contemporaneously. For this reason, in the systems with dhp and mim, the general formula of the adducts formed at physiological pH is {VOL2}n(Mb) where n = 2–4, while with ma and acac it is {VOL}n(Mb) where n = 2 with the two VOL+ moieties bound to the regions containing His81, His82, Glu83, and Asp141 on one hand, and Asp20, His24 and His119 on the other. The same situation could be statistically found also for other proteins.


image file: c9qi00179d-f15.tif
Fig. 15 Mb global representation and binding sites. The coordinating residues are highlighted depending on their coordination capability toward VOL+ and VO2+ moieties. The coordinating histidine residues of cis-VOL2 are depicted with the red disks.

This clearly demonstrates that the binding of VIVO2+ and its complexes on Mb strongly depends on the organic ligand contained in the metallodrug candidate and the interaction could take place in a variety of protein regions. Such a flexibility in the location of the drug should be taken into account for future drug design projects.

Conclusions

In this study an integrated methodology based on experimental (ESI-MS spectrometry and EPR spectroscopy) and theoretical (docking and quantum mechanics) techniques was used to study the interaction of VIVO2+ ion and four pharmacologically active VIVOL2 drugs with myoglobin. The systems were characterized and, both for VIVO2+ and VIVOL2, we demonstrated how many metallic species could bind to myoglobin and in which region and which residues of the protein they interact with. In particular, ESI-MS allows the determination of the number of moieties (VOL+ or VOL2) bound to the protein, EPR helps distinguish the type of residues involved in the coordination (His or Asp/Glu for Mb), and docking/QM calculations allow the prediction of the specific residues interacting with V and provide the 3D structure of the binding site. This opens new avenues in the characterization of the complex systems, such as those containing metal species and proteins, and presents an alternative approach to the X-ray analysis (often not possible) or NMR determination (not easily applicable for paramagnetic ions).

The combined application of EPR and ESI-MS techniques allows the study of the behaviour of the systems at two different concentrations (in the order of hundreds of μM for EPR and tens of μM for ESI-MS), covering the possible range of metal concentrations found under the physiological conditions. The results indicate that, if the thermodynamic stability of VIVOL2 is high and the cis-octahedral geometry is favored, as in the case of 1,2-dimethyl-3-hydroxy-4(1H)-pyridinonate and L-mimosinate, they survive even at low concentrations as cis-VOL2(H2O) which can interact with the protein after the replacement of the water molecule mainly with a His-N or, secondarily, with an Asp/Glu-O donor. If the stability of the V species is lower, as in the case of maltolate and acetylacetonate, they undergo hydrolysis at low concentration forming the moiety VOL(H2O)2+ that binds to protein with the contemporaneous coordination of two accessible residues (His-N and Asp/Glu-O for Mb). In both the cases, docking calculations, recently validated using a new parametrization for metals,7 are able to find the specific residues that interact with the metal and the possible secondary bonds which stabilize the adduct. The adducts identified with docking can be further refined with quantum mechanics calculations, which allow the prediction of geometries, bond distances and angles very close to those expected for simple coordination compounds. The structure of the ligand is also important to favour or disfavour the binding of the metal moieties since the formation of hydrogen bonds or electrostatic interactions with the residues on the protein surface can occur, as shown for mim and dhp. When the coordination geometry of the bis-chelated complexes is square pyramidal (as for VO(acac)2) there is no interaction or there are only weak interactions on the protein surface.

As a last comment, we would like to stress that this approach is generalizable and can be applied to systems containing other metal ions and proteins using different spectroscopic techniques depending on the metal features. ESI-MS can be employed to study the interaction of metal complexes with small proteins and the adducts formed in solution are detectable using soft ionization conditions, EPR is applicable when the metal ions are paramagnetic but it can be replaced by other spectroscopic techniques, such as NMR or UV-Vis, for diamagnetic systems, and finally docking and quantum mechanics calculations can be carried out on any metal species using the new series of metal parameters in the GOLD program generated recently7 or improving the parameters for the specific metal.

Experimental and computational section

Chemicals

Water was deionized prior to use through the purification system Millipore MilliQ Academic or purchased from Sigma-Aldrich (LC-MS grade). VIVO2+ solutions were prepared from VOSO4·3H2O following literature methods.61 1,2-Dimethyl-3-hydroxy-4(1H)-pyridinone (dhp), 2-amino-3-(3-hydroxy-4-oxo-1,4-dihydropyridin-1-yl)-propanoic acid or L-mimosine (mim), 3-hydroxy-2-methyl-4H-pyran-4-one or maltol (ma), acetylacetone (acac), 1-methylimidazole (MeIm) and 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES) were Aldrich or Fluka products of the highest grade available and used as received.

Concerning the solid compounds, [VO(acac)2] is a Sigma-Aldrich product, while [VO(dhp)2] and [VO(ma)2] were synthesized according to the procedure established in the literature.35a,62

Myoglobin from the equine heart (Mb) was purchased from Sigma with code M1882.

ESI-MS measurements

The solutions for ESI-MS measurements were prepared in two different ways: (i) dissolving in LC-MS grade water the solid complex to have a V concentration of (1.0–2.0) × 10–3 M (systems with dhp, ma and acac) or (ii) dissolving in LC-MS grade water (1.0–2.0) × 10–3 M VOSO4·3H2O and the ligand to obtain a metal-to-ligand ratio of 1/2 and adding a stoichiometric amount of BaCl2·2H2O to remove the interfering sulphate ion as BaSO4, which was immediately filtered (system with mim). Subsequently the solutions were diluted in LC-MS grade water with an aliquot of a stock myoglobin solution (500 μM) to obtain (V complex)/Mb molar ratios of 3/1, 5/1 and 10/1 with a final protein concentration of 5 or 50 μM. After the preparation of the solution, ESI-MS spectra were recorded immediately. Argon was bubbled in the solution during all the operations.

Mass spectra were obtained in positive-ion mode (ternary systems with Mb) and negative-ion mode (binary systems with mim) on a Q Exactive™ Plus Hybrid Quadrupole-Orbitrap™ (Thermo Fisher Scientific) mass spectrometer. The solutions were infused at a flow rate of 5.00 μL min−1 into the ESI chamber and the spectra were recorded in the m/z ranges 300–4500 (ternary systems with Mb) and 50–750 (binary systems with mim) at a resolution of 140[thin space (1/6-em)]000 and accumulated for at least 5 min in order to increase the signal-to-noise ratio. The instrumental conditions used for the measurements were as follows. Positive mode: spray voltage: 2300 V, capillary temperature: 250 °C, sheath gas: 10 (arbitrary units), auxiliary gas: 3 (arbitrary units), sweep gas: 0 (arbitrary units), probe heater temperature: 50 °C. Negative mode: spray voltage: −1900 V, capillary temperature: 250 °C, sheath gas: 20 (arbitrary units), auxiliary gas: 5 (arbitrary units), sweep gas: 0 (arbitrary units), probe heater temperature: 14 °C.

ESI-MS spectra were analyzed by using the Thermo Xcalibur 3.0.63 software (Thermo Fisher Scientific) and the average deconvoluted monoisotopic masses were obtained through the Xtract tool integrated in the software.

EPR and measurements

The solutions for EPR measurements were prepared dissolving in ultra-pure water VOSO4·3H2O and the ligand to obtain a VIVO2+ concentration of 1.0 × 10−3 M and a metal to ligand molar ratio of 1/2. Argon was bubbled through the solutions to ensure the absence of oxygen and avoid the oxidation of the VIVO2+ ion. To the solutions, HEPES was added as a buffer with a concentration of 1.0 × 10−1 M. Previous EPR studies proved that HEPES does not interact with the VIVO2+ ion and VIVO complexes under the conditions used for the experiments.

The pH value of the V solutions was increased to 7.4 and to a volume of 1 mL of the mixtures, Mb was added to achieve a concentration of 1.0 × 10−3 M and a ratio (V complex)/Mb of 1/1. Subsequently, various aliquots of the solutions containing the V complex were added to obtain a Mb concentration between 6.25 × 10−5 M and 5.0 × 10−4 M and a ratio of (V complex)/Mb between 2/1 and 16/1. In all the solutions Ar was always bubbled.

EPR spectra were recorded with an X-band Bruker EMX spectrometer equipped with a HP 53150A microwave frequency counter. Anisotropic spectra were recorded on frozen solutions (120 K) or at room temperature (298 K). When the samples were transferred into the EPR tubes, the spectra were immediately measured. To increase the signal to noise ratio, signal averaging was used.63 The instrumental parameters were as follows: microwave frequency, 9.40–9.41 GHz at 120 K and 9.83–9.84 GHz at 298 K; microwave power, 20 mW; time constant, 81.92 ms; modulation frequency, 100 kHz; modulation amplitude, 0.4 mT; and resolution, 4096 points. The spectra were interpreted analyzing the high field region, the part more sensitive to the changes in the equatorial coordination and amount of the species in solution;64 for this reason, in all the figures only this part is reported. In all the text, Az values were reported using the absolute value, following the formalism used in the literature.

The model systems VIVO2+/L and VIVO2+/L/MeIm, where L = dhp, mim, ma, and acac, were considered to study the binding of VIVO species to Mb.

DFT and docking calculations

DFT calculations on the VIVO complexes were carried out with the software Gaussian 09 (revision D.01).65 The geometry and harmonic frequencies were computed at the B3P86/6-311g(d,p) level of theory using the continuum SMD model66 for water. This choice ensures a good degree of accuracy in the prediction of the structures of first-row transition metal complexes67 and, in particular, of vanadium compounds.68 For L-mimosine, all the geometrical isomers of [VO(mim)2]2− and cis-[VO(mim)2(H2O)]2− were simulated (Scheme S3 of the ESI).

The simulation of the 51V hyperfine coupling constants (A) was performed on the optimized structures with the functional BHandHLYP and the basis set 6-311+g(d) level of theory, according to the procedures previously published.69 The theory background was described in detail in ref. 70.

Docking calculations were carried out through the GOLD 5.2 software71 applying the recently developed methodology able to account for the coordination bond formation.7 The calculations were performed on the X-ray diffraction (XRD) structure available in the Protein Data Bank (PDB) for myoglobin (code: 4DC8[thin space (1/6-em)]41). The XRD structure was first cleaned removing all the small molecules and crystallographic water molecules and subsequently the hydrogen atoms were added with the UCSF Chimera program.72 The protonation state of the amino acid side chains was computed by means of the PROPKA algorithm.53

The simulations were carried out building an evaluation sphere of 12 Å radius, centred for each calculation in the region of interest. The protein side chain flexibility was taken into account considering the Gold rotamer libraries.27 Genetic algorithm (GA) parameters have been set to 50 GA runs and a minimum of 100[thin space (1/6-em)]000 operations. The rest of the GA parameters were set to default.

The scoring (Fitness of GoldScore) was evaluated applying the modified version of GoldScore scoring function, which has been validated in previously published papers.7

 
Fitness (F) = α·Sexthbond + β·SextvdW + γ·Sinthbond + δ·(SintvdWStors) (2)
where Sexthbond and SextvdW are the scoring terms related to the hydrogen (hbond) and van der Waals (vdW) intermolecular interactions. Sinthbond represents the intramolecular hbond interactions and Stors evaluates the change in stability due to molecular torsions. α, β, γ, and δ are empirical coefficients optimized to weigh the different interactions. To identify the potential Mb binding sites for the VIVO species, relative Solvent Excluded Surface (SES)73 calculations were preliminarily performed, focusing on the most exposed potential coordinating residues. Additionally, in the case of docking involving the moieties with at least two coordination vacancies, the protein space was first probed for the zones with at least three (for the VIVO2+ ion) or two (for VO(dhp)+, VO(ma)+ and VO(acac)+ moieties) potential coordinating residues featuring a β-carbon within a distance of 2.5 Å from each queried grid point.26

The docking calculations were performed using, in all the cases, the selected DFT energy minima of the analyzed compound. The metal structures were preliminary treated removing the leaving groups – one or two H2O ligands in the case of the interaction of cis-VOL2(H2O) or VOL(H2O)2+ with Mb and four H2O molecules for VO2+ – and adding in the vacancies one, two or four dummy hydrogen atoms according to the procedure recently established.7 All dockings were computed considering the possible dihedral changes along the aliphatic bonds applying the GOLD algorithm. The solutions were analyzed by means of GaudiView.74

The best solutions of the calculations were evaluated through two main criteria: the mean (Fmean) or the highest value (Fmax) of the scoring (Fitness of GoldScore, eqn (2)) associated with each pose and (ii) the percent population of the cluster containing the best pose.

The refinement optimization of the protein–complex adducts found by docking calculations was performed with full QM calculations, cutting out from the protein the VIVO2+ ion plus the coordinated and neighbour interacting amino acid side chains and freezing the coordinates of the backbone atoms where the truncation was made as reported in the literature by Siegbahn and Himo.28 The geometry relaxation and ΔG calculations relative to the reaction of adduct formation were performed at B3P86/6-311++g(d,p) in the SMD continuum model66 for water.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was supported by the Spanish grant CTQ2017-87889-P and Generalitat de Catalunya 2017SGR1323, COST Action CM1306, Fondazione di Sardegna (project FDS15Garribba) and FFABR 2017 “Fondo per il finanziamento delle attività base di ricerca”. G. S. is also grateful to the Universitat Autònoma de Barcelona for the support to his Ph.D.

References

  1. (a) R. Crichton, Biological Inorganic Chemistry. A New Introduction to Molecular Structure and Function, Second Edition, Elsevier, Oxford, 2012 Search PubMed; (b) D. Rehder, Bioinorganic Chemistry, Oxford University Press, Oxford, 2014 Search PubMed.
  2. (a) B. Rosenberg, L. Vancamp, J. E. Trosko and V. H. Mansour, Nature, 1969, 222, 385–386 CrossRef CAS PubMed; (b) B. Rosenberg and L. VanCamp, Cancer Res., 1970, 30, 1799–1802 CAS; (c) Cisplatin. Chemistry and Biochemistry of a Leading Anticancer Drug, ed. B. Lippert, Verlag Helvetica Chimica Acta, Zürich, 1999 Search PubMed; (d) S. Medici, M. Peana, V. M. Nurchi, J. I. Lachowicz, G. Crisponi and M. A. Zoroddu, Coord. Chem. Rev., 2015, 284, 329–350 CrossRef CAS.
  3. (a) Z. Guo and P. J. Sadler, Angew. Chem., Int. Ed., 1999, 38, 1512–1537 CrossRef CAS; (b) N. Farrell, in Comprehensive Coordination Chemistry II, ed. J. A. McCleverty and T. J. Meyer, Pergamon, Oxford, 2 003, vol. 9, pp. 809–840 Search PubMed; (c) Metallotherapeutic Drugs and Metal-Based Diagnostic Agents. The Use of Metals in Medicine, ed. M. Gielen and E. R. T. Tiekink, John Wiley & Sons Ltd, Chichester, 2005 Search PubMed; (d) C. Jones and J. Thornback, Medicinal Applications of Coordination Chemistry, The Royal Society of Chemistry, Cambridge, 2007 Search PubMed; (e) J. C. Dabrowiak, Metals in Medicine, John Wiley & Sons, Ltd, Chichester, 2009 CrossRef; (f) K. H. Thompson, in Encyclopedia of Inorganic and Bioinorganic Chemistry, ed. R. B. King, John Wiley & Sons, Ltd, Chichester, 2011 Search PubMed; (g) Bioinorganic Medicinal Chemistry, ed. E. Alessio, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, 2011 Search PubMed; (h) Interrelations between Essential Metal Ions and Human Diseases, ed. A. Sigel, H. Sigel and R. K. O. Sigel, Springer Science+Business Media, Dordrecht, 2013 Search PubMed; (i) N. P. E. Barry and P. J. Sadler, Chem. Commun., 2013, 49, 5106–5131 RSC; (j) Metallo-Drugs: Development and Action of Anticancer Agents, ed. A. Sigel, H. Sigel, E. Freisinger and R. K. O. Sigel, De Gruyter GmbH, Berlin, 2018 Search PubMed; (k) Essential Metals in Medicine: Therapeutic Use and Toxicity of Metal Ions in the Clinic, ed. P. L. Carver, De Gruyter GmbH, Berlin, 2019 Search PubMed.
  4. (a) K. D. Mjos and C. Orvig, Chem. Rev., 2014, 114, 4540–4563 CrossRef CAS PubMed; (b) K. A. Doucette, K. N. Hassell and D. C. Crans, J. Inorg. Biochem., 2016, 165, 56–70 CrossRef CAS PubMed; (c) A. Levina, D. C. Crans and P. A. Lay, Coord. Chem. Rev., 2017, 352, 473–498 CrossRef CAS; (d) T. Kiss, É. A. Enyedy and T. Jakusch, Coord. Chem. Rev., 2017, 352, 401–423 CrossRef CAS.
  5. J. Costa Pessoa, E. Garribba, M. F. A. Santos and T. Santos-Silva, Coord. Chem. Rev., 2015, 301–302, 49–86 CrossRef CAS.
  6. R. O. Louro and R. Crichton, Practical Approaches to Biological Inorganic Chemistry, Elsevier, Oxford, 2013 Search PubMed.
  7. (a) G. Sciortino, D. Sanna, V. Ugone, G. Micera, A. Lledós, J.-D. Maréchal and E. Garribba, Inorg. Chem., 2017, 56, 12938–12951 CrossRef CAS PubMed; (b) G. Sciortino, J. Rodríguez-Guerra Pedregal, A. Lledós, E. Garribba and J.-D. Maréchal, J. Comput. Chem., 2018, 39, 42–51 CrossRef CAS PubMed; (c) G. Sciortino, E. Garribba and J.-D. Maréchal, Inorg. Chem., 2019, 58, 294–306 CrossRef CAS PubMed.
  8. (a) J. Costa Pessoa and I. Tomaz, Curr. Med. Chem., 2010, 17, 3701–3738 CrossRef; (b) D. Gambino, Coord. Chem. Rev., 2011, 255, 2193–2203 CrossRef CAS; (c) D. Rehder, Future Med. Chem., 2012, 4, 1823–1837 CrossRef CAS PubMed; (d) D. Rehder, in Interrelations between Essential Metal Ions and Human Diseases, ed. A. Sigel, H. Sigel and R. K. O. Sigel, Springer Science+Business Media, Dordrecht, 2013, ch. 5, pp. 139–169 Search PubMed; (e) M. Aureliano and C. A. Ohlin, J. Inorg. Biochem., 2014, 137, 123–130 CrossRef CAS PubMed; (f) J. Costa Pessoa, S. Etcheverry and D. Gambino, Coord. Chem. Rev., 2015, 301–302, 24–48 CrossRef; (g) D. Rehder, Future Med. Chem., 2016, 8, 325–338 CrossRef CAS PubMed; (h) I. E. Leon, J. F. Cadavid-Vargas, A. L. Di Virgilio and S. B. Etcheverry, Curr. Med. Chem., 2017, 24, 112–148 CrossRef CAS PubMed; (i) D. Crans, C. L. Yang, A. Haase and X. Yang, in Metallo-Drugs: Development and Action of Anticancer Agents, ed. A. Sigel, H. Sigel, E. Freisinger and R. K. O. Sigel, De Gruyter GmbH, Berlin, 2018, vol. 18, pp. 251–280 Search PubMed; (j) D. Rehder, ChemTexts, 2018, 4, 20 CrossRef; (k) D. C. Crans, L. Henry, G. Cardiff and B. I. Posner, in Metal Ions in Life Sciences, ed. P. L. Carver, De Gruyter GmbH, Berlin, 2019, vol. 19, pp. 203–230 Search PubMed.
  9. (a) K. H. Thompson and C. Orvig, in Met. Ions Biol. Syst., ed. H. Sigel and A. Sigel, Marcel Dekker, New York, 2004, vol. 41, pp. 221–252 Search PubMed; (b) K. H. Thompson, B. D. Liboiron, G. R. Hanson and C. Orvig, in Medicinal Inorganic Chemistry, ed. J. L. Sessler, S. R. Doctrow, T. J. McMurry and S. J. Lippard, American Chemical Society, Washington DC, 2005, vol. 903, ch. 21, pp. 384–399 Search PubMed; (c) K. H. Thompson and C. Orvig, J. Inorg. Biochem., 2006, 100, 1925–1935 CrossRef CAS PubMed.
  10. K. H. Thompson, J. Lichter, C. LeBel, M. C. Scaife, J. H. McNeill and C. Orvig, J. Inorg. Biochem., 2009, 103, 554–558 CrossRef CAS PubMed.
  11. (a) D. Sanna, G. Micera and E. Garribba, Inorg. Chem., 2009, 48, 5747–5757 CrossRef CAS PubMed; (b) D. Sanna, G. Micera and E. Garribba, Inorg. Chem., 2010, 49, 174–187 CrossRef CAS PubMed; (c) D. Sanna, P. Buglyó, G. Micera and E. Garribba, J. Biol. Inorg. Chem., 2010, 15, 825–839 CrossRef CAS PubMed; (d) D. Sanna, G. Micera and E. Garribba, Inorg. Chem., 2011, 50, 3717–3728 CrossRef CAS PubMed; (e) D. Sanna, L. Biro, P. Buglyo, G. Micera and E. Garribba, Metallomics, 2012, 4, 33–36 RSC; (f) D. Sanna, L. Bíró, P. Buglyó, G. Micera and E. Garribba, J. Inorg. Biochem., 2012, 115, 87–99 CrossRef CAS PubMed; (g) D. Sanna, V. Ugone, G. Micera and E. Garribba, Dalton Trans., 2012, 41, 7304–7318 RSC; (h) D. Sanna, G. Micera and E. Garribba, Inorg. Chem., 2013, 52, 11975–11985 CrossRef CAS PubMed; (i) T. Koleša-Dobravc, E. Lodyga-Chruscinska, M. Symonowicz, D. Sanna, A. Meden, F. Perdih and E. Garribba, Inorg. Chem., 2014, 53, 7960–7976 CrossRef PubMed; (j) D. Sanna, V. Ugone, L. Pisano, M. Serra, G. Micera and E. Garribba, J. Inorg. Biochem., 2015, 153, 167–177 CrossRef CAS PubMed; (k) D. Sanna, V. Ugone, G. Micera, P. Buglyo, L. Biro and E. Garribba, Dalton Trans., 2017, 46, 8950–8967 RSC; (l) D. Sanna, V. Ugone, M. Serra and E. Garribba, J. Inorg. Biochem., 2017, 173, 52–65 CrossRef CAS PubMed; (m) D. Sanna, J. Palomba, G. Lubinu, P. Buglyó, S. Nagy, F. Perdih and E. Garribba, J. Med. Chem., 2019, 62, 654–664 CrossRef CAS PubMed.
  12. (a) G. R. Willsky, A. B. Goldfine, P. J. Kostyniak, J. H. McNeill, L. Q. Yang, H. R. Khan and D. C. Crans, J. Inorg. Biochem., 2001, 85, 33–42 CrossRef CAS PubMed; (b) B. D. Liboiron, K. H. Thompson, G. R. Hanson, E. Lam, N. Aebischer and C. Orvig, J. Am. Chem. Soc., 2005, 127, 5104–5115 CrossRef CAS PubMed; (c) B. D. Liboiron, in High Resolution EPR, ed. L. Berliner and G. Hanson, Springer, New York, 2009, vol. 28, ch. 12, pp. 507–549 Search PubMed; (d) T. Jakusch, D. Hollender, E. A. Enyedy, C. S. Gonzalez, M. Montes-Bayon, A. Sanz-Medel, J. Costa Pessoa, I. Tomaz and T. Kiss, Dalton Trans., 2009, 2428–2437 RSC; (e) I. Correia, T. Jakusch, E. Cobbinna, S. Mehtab, I. Tomaz, N. V. Nagy, A. Rockenbauer, J. Costa Pessoa and T. Kiss, Dalton Trans., 2012, 41, 6477–6487 RSC; (f) G. Gonçalves, A. I. Tomaz, I. Correia, L. F. Veiros, M. M. C. A. Castro, F. Avecilla, L. Palacio, M. Maestro, T. Kiss, T. Jakusch, M. H. V. Garcia and J. Costa Pessoa, Dalton Trans., 2013, 42, 11841–11861 RSC; (g) G. C. Justino, E. Garribba and J. Costa Pessoa, J. Biol. Inorg. Chem., 2013, 18, 803–813 CrossRef CAS PubMed; (h) S. Mehtab, G. Gonçalves, S. Roy, A. I. Tomaz, T. Santos-Silva, M. F. A. Santos, M. J. Romão, T. Jakusch, T. Kiss and J. Costa Pessoa, J. Inorg. Biochem., 2013, 121, 187–195 CrossRef CAS PubMed; (i) J. Costa Pessoa, G. Gonçalves, S. Roy, I. Correia, S. Mehtab, M. F. A. Santos and T. Santos-Silva, Inorg. Chim. Acta, 2014, 420, 60–68 CrossRef CAS; (j) M. F. A. Santos, I. Correia, A. R. Oliveira, E. Garribba, J. Costa Pessoa and T. Santos-Silva, Eur. J. Inorg. Chem., 2014, 3293–3297 CrossRef CAS; (k) J. Costa Pessoa, J. Inorg. Biochem., 2015, 4–24 CrossRef CAS PubMed; (l) I. Correia, I. Chorna, I. Cavaco, S. Roy, M. L. Kuznetsov, N. Ribeiro, G. Justino, F. Marques, T. Santos-Silva, M. F. A. Santos, H. M. Santos, J. L. Capelo, J. Doutch and J. Costa Pessoa, Chem. – Asian J., 2017, 12, 2062–2084 CrossRef CAS PubMed; (m) C. G. Azevedo, I. Correia, M. M. C. dos Santos, M. F. A. Santos, T. Santos-Silva, J. Doutch, L. Fernandes, H. M. Santos, J. L. Capelo and J. Costa Pessoa, J. Inorg. Biochem., 2018, 180, 211–221 CrossRef CAS PubMed.
  13. D. C. Crans, J. J. Smee, E. Gaidamauskas and L. Yang, Chem. Rev., 2004, 104, 849–902 CrossRef CAS PubMed.
  14. D. Rehder, Bioinorganic Vanadium Chemistry, John Wiley & Sons Ltd, Chichester, 2008 Search PubMed.
  15. (a) L. C. Cantley and P. Aisen, J. Biol. Chem., 1979, 254, 1781–1784 CAS; (b) I. G. Macara, K. Kustin and L. C. Cantley Jr., Biochim. Biophys. Acta, Gen. Subj., 1980, 629, 95–106 CrossRef CAS; (c) A. Heinz, K. A. Rubinson and J. J. Grantham, J. Lab. Clin. Med., 1982, 100, 593–612 CAS; (d) T. V. Hansen, J. Aaseth and J. Alexander, Arch. Toxicol., 1982, 50, 195–202 CrossRef CAS PubMed; (e) W. R. Harris, S. B. Friedman and D. Silberman, J. Inorg. Biochem., 1984, 20, 157–169 CrossRef CAS PubMed; (f) M. Garner, J. Reglinski, W. E. Smith, J. McMurray, I. Abdullah and R. Wilson, J. Biol. Inorg. Chem., 1997, 2, 235–241 CrossRef CAS; (g) K. De Cremer, M. Van Hulle, C. Chéry, R. Cornelis, K. Strijckmans, R. Dams, N. Lameire and R. Vanholder, J. Biol. Inorg. Chem., 2002, 7, 884–890 CrossRef CAS PubMed; (h) X. Yang, K. Wang, J. Lu and D. C. Crans, Coord. Chem. Rev., 2003, 237, 103–111 CrossRef CAS; (i) T. C. Delgado, A. I. Tomaz, I. Correia, J. Costa Pessoa, J. G. Jones, C. F. G. C. Geraldes and M. M. C. A. Castro, J. Inorg. Biochem., 2005, 99, 2328–2339 CrossRef CAS PubMed.
  16. T. Scior, J. A. Guevara-Garcia, Q.-T. Do, P. Bernard and S. Laufer, Curr. Med. Chem., 2016, 23, 2874–2891 CrossRef CAS PubMed.
  17. J. Schaller, S. Gerber, U. Kämpfer, S. Lejon and C. Trachsel, Human Blood Plasma Proteins: Structure and Function, John Wiley & Sons Ltd, Chichester, 2008 Search PubMed.
  18. O. Fatunmbi, R. R. Abzalimov, S. N. Savinov, A. Gershenson and I. A. Kaltashov, Biochemistry, 2016, 55, 1918–1928 CrossRef CAS PubMed.
  19. (a) U. Oleksiewicz, N. Daskoulidou, T. Liloglou, K. Tasopoulou, J. Bryan, J. R. Gosney, J. K. Field and G. Xinarianos, Lung Cancer, 2011, 74, 411–418 CrossRef PubMed; (b) G. Kristiansen, M. Rose, C. Geisler, F. R. Fritzsche, J. Gerhardt, C. Lüke, A. M. Ladhoff, R. Knüchel, M. Dietel, H. Moch, Z. Varga, J. P. Theurillat, T. A. Gorr and E. Dahl, Br. J. Cancer, 2010, 102, 1736 CrossRef CAS PubMed; (c) G. Kristiansen, J. Hu, D. Wichmann, D. P. Stiehl, M. Rose, J. Gerhardt, A. Bohnert, A. ten Haaf, H. Moch, J. Raleigh, M. A. Varia, P. Subarsky, F. M. Scandurra, E. Gnaiger, E. Gleixner, A. Bicker, M. Gassmann, T. Hankeln, E. Dahl and T. A. Gorr, J. Biol. Chem., 2011, 286, 43417–43428 CrossRef CAS PubMed; (d) A. Bicker, D. Dietrich, E. Gleixner, G. Kristiansen, T. A. Gorr and T. Hankeln, Hum. Mol. Genet., 2014, 23, 479–490 CrossRef CAS PubMed; (e) A. Bicker, A. M. Brahmer, S. Meller, G. Kristiansen, T. A. Gorr and T. Hankeln, PLoS One, 2015, 10, e0142662 CrossRef PubMed; (f) S. E. Flonta, S. Arena, A. Pisacane, P. Michieli and A. Bardelli, Am. J. Pathol., 2009, 175, 201–206 CrossRef CAS PubMed; (g) U. Flögel and C. V. Dang, J. Clin. Invest., 2009, 119, 766–768 CrossRef.
  20. (a) C. L. Behnes, J. Bedke, S. Schneider, S. Küffer, A. Strauss, F. Bremmer, P. Ströbel and H. J. Radzun, Exp. Mol. Pathol., 2013, 95, 307–312 CrossRef CAS PubMed; (b) S. Meller, A. Bicker, M. Montani, K. Ikenberg, B. Rostamzadeh, V. Sailer, P. Wild, D. Dietrich, B. Uhl, T. Sulser, H. Moch, T. A. Gorr, C. Stephan, K. Jung, T. Hankeln and G. Kristiansen, Virchows Arch., 2014, 465, 419–427 CrossRef CAS PubMed.
  21. M. Wenzel and A. Casini, Coord. Chem. Rev., 2017, 352, 432–460 CrossRef CAS.
  22. D. Sanna, V. Ugone, G. Sciortino, P. Buglyo, Z. Bihari, P. L. Parajdi-Losonczi and E. Garribba, Dalton Trans., 2018, 47, 2164–2182 RSC.
  23. M. Tamura, N. Shimidzu and K. Hayashi, Biochem. Biophys. Res. Commun., 1977, 75, 1029–1036 CrossRef CAS PubMed.
  24. H. Sakurai, T. Goda and S. Shimomura, Biochem. Biophys. Res. Commun., 1982, 107, 1349–1354 CrossRef CAS PubMed.
  25. (a) D. Sanna, M. Serra, G. Micera and E. Garribba, Inorg. Chem., 2014, 53, 1449–1464 CrossRef CAS PubMed; (b) D. Sanna, M. Serra, G. Micera and E. Garribba, Inorg. Chim. Acta, 2014, 420, 75–84 CrossRef CAS.
  26. G. Sciortino, E. Garribba, J. Rodríguez-Guerra Pedregal and J.-D. Maréchal, ACS Omega, 2019, 4, 3726–3731 CrossRef CAS.
  27. S. C. Lovell, J. M. Word, J. S. Richardson and D. C. Richardson, Proteins: Struct., Funct., Bioinf., 2000, 40, 389–408 CrossRef CAS.
  28. P. E. M. Siegbahn and F. Himo, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2011, 1, 323–336 CAS.
  29. L. F. Vilas Boas and J. Costa Pessoa, in Comprehensive Coordination Chemistry, ed. G. Wilkinson, R. D. Gillard and J. A. McCleverty, Pergamon Press, Oxford, 1985, vol. 3, pp. 453–583 Search PubMed.
  30. R. J. DeKoch, D. J. West, J. C. Cannon and N. D. Chasteen, Biochemistry, 1974, 13, 4347–4354 CrossRef CAS.
  31. (a) D. N. Chasteen, in Biological Magnetic Resonance, ed. L. J. J. Berliner and J. Reuben, Plenum Press, New York, 1981, vol. 3, ch. 2, pp. 53–119 Search PubMed; (b) T. S. Smith II, R. LoBrutto and V. L. Pecoraro, Coord. Chem. Rev., 2002, 228, 1–18 CrossRef; (c) E. Garribba, E. Lodyga-Chruscinska, G. Micera, A. Panzanelli and D. Sanna, Eur. J. Inorg. Chem., 2005, 1369–1382 CrossRef CAS.
  32. M. Rangel, A. Tamura, C. Fukushima and H. Sakurai, J. Biol. Inorg. Chem., 2001, 6, 128–132 CrossRef CAS.
  33. (a) M. Passadouro, A. M. Metelo, A. S. Melão, J. R. Pedro, H. Faneca, E. Carvalho and M. M. C. A. Castro, J. Inorg. Biochem., 2010, 104, 987–992 CrossRef CAS PubMed; (b) A. M. Metelo, R. Pérez-Carro, M. M. C. A. Castro and P. López-Larrubia, J. Inorg. Biochem., 2012, 115, 44–49 CrossRef CAS PubMed; (c) N. Domingues, J. Pelletier, C. G. Ostenson and M. M. C. A. Castro, J. Inorg. Biochem., 2014, 131, 115–122 CrossRef CAS PubMed.
  34. C. Rozzo, D. Sanna, E. Garribba, M. Serra, A. Cantara, G. Palmieri and M. Pisano, J. Inorg. Biochem., 2017, 174, 14–24 CrossRef CAS PubMed.
  35. (a) P. Buglyo, T. Kiss, E. Kiss, D. Sanna, E. Garribba and G. Micera, J. Chem. Soc., Dalton Trans., 2002, 2275–2282 RSC; (b) M. Rangel, A. Leite, M. João Amorim, E. Garribba, G. Micera and E. Lodyga-Chruscinska, Inorg. Chem., 2006, 45, 8086–8097 CrossRef CAS PubMed.
  36. (a) V. Katta and B. T. Chait, J. Am. Chem. Soc., 1991, 113, 8534–8535 CrossRef CAS; (b) Y. Konishi and R. Feng, Biochemistry, 1994, 33, 9706–9711 CrossRef CAS; (c) L. Konermann, F. I. Rosell, A. G. Mauk and D. J. Douglas, Biochemistry, 1997, 36, 6448–6454 CrossRef CAS PubMed; (d) J. A. Loo, Int. J. Mass Spectrom., 2000, 200, 175–186 CrossRef CAS.
  37. J. B. Fenn, J. Am. Soc. Mass Spectrom., 1993, 4, 524–535 CrossRef CAS PubMed.
  38. A. Dobo and I. A. Kaltashov, Anal. Chem., 2001, 73, 4763–4773 CrossRef CAS PubMed.
  39. (a) E. Moreno-Gordaliza, B. Cañas, M. A. Palacios and M. M. Gómez-Gómez, Anal. Chem., 2009, 81, 3507–3516 CrossRef CAS PubMed; (b) E. Moreno-Gordaliza, B. Cañas, M. A. Palacios and M. M. Gómez-Gómez, Analyst, 2010, 135, 1288–1298 RSC.
  40. Nomenclature of Inorganic Chemistry – IUPAC Recommendations, ed. N. G. Connelly, T. Damhus, R. M. Hartshorn and A. T. Hutton, The Royal Society of Chemistry, Cambridge, 2005. According to IUPAC, the configuration index of square pyramidal (SPY-5) and octahedral (OC-6) complexes consists of two digits. For SPY-5 the first digit is the priority number of the donor on the C4 symmetry axis of the idealized pyramid and the second digit is the priority number of the donor trans to that with the lowest priority number in the plane perpendicular to the C4 axis (if there is more than one of the highest priority ligand in the plane, the priority number of the trans ligand having the largest numerical value is selected). For OC-6 the first digit is the priority number of the ligating atom trans to the ligating atom of priority number 1, and the second digit of the configuration index is the priority number of the ligating atom trans to the most preferred ligand in the plane that is perpendicular to the reference axis. The procedure for assigning priority numbers to the donor atoms is based upon the standard sequence rules developed for chiral carbon compounds by Cahn, Ingold and Prelog (CIP rules) Search PubMed.
  41. H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T. N. Bhat, H. Weissig, I. N. Shindyalov and P. E. Bourne, Nucleic Acids Res., 2000, 28, 235–242 CrossRef CAS PubMed.
  42. D. M. Gilbert, A. Neilson, H. Miyazawa, M. L. DePamphilis and W. C. Burhans, J. Biol. Chem., 1995, 270, 9597–9606 CrossRef CAS PubMed.
  43. M. D. Aytemir and G. Karakaya, in Medicinal Chemistry and Drug Design, ed. D. Ekinci, InTech, Rijeka, 2012, ch. 1, pp. 1–26 Search PubMed.
  44. P. P. Corbi, A. C. Massabni and C. M. Costa-Neto, J. Coord. Chem., 2005, 58, 1477–1483 CrossRef CAS.
  45. B. C. Q. Nguyen and S. Tawata, Phytother. Res., 2016, 30, 1230–1242 CrossRef CAS.
  46. K. H. Thompson, C. A. Barta and C. Orvig, Chem. Soc. Rev., 2006, 35, 545–556 RSC.
  47. E. Lodyga-Chruscinska, E. Garribba, G. Micera and A. Panzanelli, J. Inorg. Biochem., 1999, 75, 225–232 CrossRef.
  48. V. B. Di Marco and G. G. Bombi, Mass Spectrom. Rev., 2006, 25, 347–379 CrossRef CAS.
  49. T. Zhao and F. L. King, J. Inorg. Biochem., 2010, 104, 186–192 CrossRef CAS PubMed.
  50. E. Moreno-Gordaliza, B. Cañas, M. A. Palacios and M. M. Gómez-Gómez, Talanta, 2012, 88, 599–608 CrossRef CAS PubMed.
  51. (a) C. Orvig, P. Caravan, L. Gelmini, N. Glover, F. G. Herring, H. Li, J. H. McNeill, S. J. Rettig and I. A. Setyawati, J. Am. Chem. Soc., 1995, 117, 12759–12770 CrossRef CAS; (b) K. H. Thompson and C. Orvig, Coord. Chem. Rev., 2001, 219–221, 1033–1053 CrossRef CAS.
  52. P. Buglyó, E. Kiss, I. Fábián, T. Kiss, D. Sanna, E. Garribba and G. Micera, Inorg. Chim. Acta, 2000, 306, 174–183 CrossRef.
  53. M. H. M. Olsson, C. R. Søndergaard, M. Rostkowski and J. H. Jensen, J. Chem. Theory Comput., 2011, 7, 525–537 CrossRef CAS PubMed.
  54. (a) B. A. Reul, S. S. Amin, J.-P. Buchet, L. N. Ongemba, D. C. Crans and S. M. Brichard, Br. J. Pharmacol., 1999, 126, 467–477 CrossRef CAS PubMed; (b) S. S. Amin, K. Cryer, B. Zhang, S. K. Dutta, S. S. Eaton, O. P. Anderson, S. M. Miller, B. A. Reul, S. M. Brichard and D. C. Crans, Inorg. Chem., 2000, 39, 406–416 CrossRef CAS.
  55. (a) M. W. Makinen and M. J. Brady, J. Biol. Chem., 2002, 277, 12215–12220 CrossRef CAS PubMed; (b) H. Ou, L. Yan, D. Mustafi, M. W. Makinen and M. J. Brady, JBIC, J. Biol. Inorg. Chem., 2005, 10, 874–886 CrossRef CAS PubMed; (c) M. W. Makinen and M. Salehitazangi, Coord. Chem. Rev., 2014, 279, 1–22 CrossRef CAS PubMed.
  56. D. C. Crans, A. R. Khan, M. Mahroof-Tahir, S. Mondal, S. M. Miller, A. la Cour, O. P. Anderson, T. Jakusch and T. Kiss, J. Chem. Soc., Dalton Trans., 2001, 3337–3345 RSC.
  57. E. Garribba, G. Micera and D. Sanna, Inorg. Chim. Acta, 2006, 359, 4470–4476 CrossRef CAS.
  58. G. Sciortino, D. Sanna, V. Ugone, A. Lledós, J.-D. Maréchal and E. Garribba, Inorg. Chem., 2018, 57, 4456–4469 CrossRef CAS PubMed.
  59. D. N. Chasteen, in Met. Ions Biol. Syst, ed. H. Sigel and A. Sigel, Marcel Dekker, New York, 1995, vol. 31, ch. 16, pp. 231–247 Search PubMed.
  60. D. Sanna, V. Pecoraro, G. Micera and E. Garribba, J. Biol. Inorg. Chem., 2012, 17, 773–790 CrossRef CAS PubMed.
  61. I. Nagypál and I. Fábián, Inorg. Chim. Acta, 1982, 61, 109–113 CrossRef.
  62. A. Levina, A. I. McLeod and P. A. Lay, Chem. – Eur. J., 2014, 20, 12056–12060 CrossRef CAS PubMed.
  63. D. Sanna, E. Garribba and G. Micera, J. Inorg. Biochem., 2009, 103, 648–655 CrossRef CAS PubMed.
  64. T. Kiss, T. Jakusch, J. Costa Pessoa and I. Tomaz, Coord. Chem. Rev., 2003, 237, 123–133 CrossRef CAS.
  65. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresma n, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, revision C.01, Gaussian, Inc., Wallingford, CT, 2010 Search PubMed.
  66. A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem. B, 2009, 113, 6378–6396 CrossRef CAS PubMed.
  67. (a) M. Bühl and H. Kabrede, J. Chem. Theory Comput., 2006, 2, 1282–1290 CrossRef PubMed; (b) M. Bühl, C. Reimann, D. A. Pantazis, T. Bredow and F. Neese, J. Chem. Theory Comput., 2008, 4, 1449–1459 CrossRef PubMed.
  68. (a) G. Micera and E. Garribba, Int. J. Quantum Chem., 2012, 112, 2486–2498 CrossRef CAS; (b) E. Lodyga-Chruscinska, D. Sanna, E. Garribba and G. Micera, Dalton Trans., 2008, 4903–4916 RSC.
  69. (a) G. Micera, V. L. Pecoraro and E. Garribba, Inorg. Chem., 2009, 48, 5790–5796 CrossRef CAS PubMed; (b) G. Micera and E. Garribba, Eur. J. Inorg. Chem., 2010, 4697–4710 CrossRef CAS; (c) G. Micera and E. Garribba, Eur. J. Inorg. Chem., 2011, 3768–3780 CAS; (d) D. Sanna, K. Varnágy, S. Timári, G. Micera and E. Garribba, Inorg. Chem., 2011, 50, 10328–10341 CrossRef CAS PubMed; (e) D. Sanna, P. Buglyó, L. Bíró, G. Micera and E. Garribba, Eur. J. Inorg. Chem., 2012, 1079–1092 CrossRef CAS; (f) D. Sanna, P. Buglyo, A. I. Tomaz, J. Costa Pessoa, S. Borovic, G. Micera and E. Garribba, Dalton Trans., 2012, 41, 12824–12838 RSC; (g) L. Pisano, K. Varnagy, S. Timári, K. Hegetschweiler, G. Micera and E. Garribba, Inorg. Chem., 2013, 52, 5260–5272 CrossRef CAS PubMed; (h) D. Sanna, K. Várnagy, N. Lihi, G. Micera and E. Garribba, Inorg. Chem., 2013, 52, 8202–8213 CrossRef CAS PubMed; (i) E. Lodyga-Chruscinska, A. Szebesczyk, D. Sanna, K. Hegetschweiler, G. Micera and E. Garribba, Dalton Trans., 2013, 42, 13404–13416 RSC; (j) S. Kundu, D. Mondal, K. Bhattacharya, A. Endo, D. Sanna, E. Garribba and M. Chaudhury, Inorg. Chem., 2015, 54, 6203–6215 CrossRef CAS PubMed; (k) S. P. Dash, S. Majumder, A. Banerjee, M. F. N. N. Carvalho, P. Adão, J. Costa Pessoa, K. Brzezinski, E. Garribba, H. Reuter and R. Dinda, Inorg. Chem., 2016, 55, 1165–1182 CrossRef CAS PubMed.
  70. (a) G. Micera and E. Garribba, Dalton Trans., 2009, 1914–1918 RSC; (b) S. Gorelsky, G. Micera and E. Garribba, Chem. – Eur. J., 2010, 16, 8167–8180 CrossRef CAS PubMed; (c) G. Micera and E. Garribba, J. Comput. Chem., 2011, 32, 2822–2835 CrossRef CAS PubMed; (d) D. Sanna, G. Sciortino, V. Ugone, G. Micera and E. Garribba, Inorg. Chem., 2016, 55, 7373–7387 CrossRef CAS PubMed.
  71. (a) G. Jones, P. Willett and R. C. Glen, J. Mol. Biol., 1995, 245, 43–53 CrossRef CAS; (b) G. Jones, P. Willett, R. C. Glen, A. R. Leach and R. Taylor, J. Mol. Biol., 1997, 267, 727–748 CrossRef CAS PubMed; (c) M. L. Verdonk, J. C. Cole, M. J. Hartshorn, C. W. Murray and R. D. Taylor, Proteins: Struct., Funct., Bioinf., 2003, 52, 609–623 CrossRef CAS PubMed.
  72. E. F. Pettersen, T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng and T. E. Ferrin, J. Comput. Chem., 2004, 25, 1605–1612 CrossRef CAS PubMed.
  73. M. Connolly, J. Appl. Crystallogr., 1983, 16, 548–558 CrossRef CAS.
  74. (a) J. Rodríguez-Guerra, Insilichem/gaudiview: Pre-alpha public releas, Zenodo, 2017 Search PubMed; (b) J. Rodríguez-Guerra Pedregal, G. Sciortino, J. Guasp, M. Municoy and J.-D. Maréchal, J. Comput. Chem., 2017, 38, 2118–2126 CrossRef PubMed.

Footnote

Electronic supplementary information (ESI) available: Species identified from the ESI-MS spectra of the system VIVO2+/mim (Table S1), 51V hyperfine coupling constants calculated for the bis-chelated VIVO complexes formed by mim (Table S2), docking solutions for the interaction of VO(acac)+ with Mb (Table S3), raw ESI-MS spectra recorded for all the systems (Fig. S1), deconvoluted ESI-MS spectrum of Mb (Fig. S2), zoomed region for the multipeak with charge +9 of Mb and [VO(dhp)2]/Mb (Fig. S3), ultrazoomed region for the two most intense peaks with charge +9 of all the systems (Fig. S4), docking solutions for the interaction of VO(dhp)+ with Mb (Fig. S5), experimental and calculated isotopic patterns for the peak of [VO(mim)2]2− (Fig. S6), deconvoluted ESI-MS spectra recorded on the system containing VO(ma)2/Mb and VO(acac)2/Mb (Figs. S7 and S8), the high field region of the X-band EPR spectra recorded on the system VIVO2+/acac/Mb at 120 and 298 K (Figs. S9 and S10), cluster distribution for the interaction of VO(acac)2 with Mb (Fig. S11), structures in aqueous solution and physiological pH of the bis-chelated VIVO complexes formed by dhp, ma, mim and acac (Scheme S1), enantiomers of [VO(dhp)(H2O)2]+ and [VO(ma)(H2O)2]+ (Scheme S2) and possible isomers of the bis-chelated VIVO complex formed by mim (Scheme S3). See DOI: 10.1039/c9qi00179d

This journal is © the Partner Organisations 2019