Mixed-Valence Pentadecavanadate with Ca²⁺-ATPase Inhibition Potential and Anti-Breast Cancer Activity

Abstract

Polyoxovanadates are a subclass of polyoxometalates (POMs) known to interact with proteins and to present anticancer, antimicrobial, and antiviral activities. Herein, we aimed to pursue the study of the breast anticancer activity of a mixed-valence polyoxovanadate, [Cl@V^V₇V^IV₈O₃₆]⁶⁻ (V₁₅) against MCF-7 and MDA-MB-231 cancer cell lines and to analyze its Ca²⁺-ATPase inhibition potential. ⁵¹V NMR and UV-Vis/NIR studies of V₁₅ indicated its stability in HEPES and RPMI media. For the Ca²⁺-ATPase activity, V₁₅ showed an IC₅₀ value of 14.2 μM and a mixed type of inhibition. The electrostatic potential map of V₁₅ and other POMs were correlated with the enzyme activity inhibition. V₁₅ also exhibited cytotoxicity against MDA-MB-231 (IC₅₀ = 17.2 μM) and MCF-7 (IC₅₀ = 15.1 μM) breast cancer cell lines. Using V₁₅ concentrations equivalent to half and 1/4 of the IC_50, it was observed that MDA-MB-231 cell migration was reduced by 90 and 70%, after 24 h, respectively. Moreover, V₁₅ caused morphological changes from fusiform to an epithelial-like (amoeboid) shape. Finally, V₁₅ induced the increase in RIPK1, MLKL, and RIPK3 gene expression, up to 3, 10, and 15-fold, respectively, pointing out that the mechanisms of cell death in the triple-negative breast cancer cell line may occur by necroptosis.

1. Introduction

Polyoxovanadates (POVs), an important subclass of polyoxometalates (POMs), are polyatomic oxide anions that contain vanadium in its highest oxidation states. Decavanadate, [H_nV₁₀O₁₈]^(6–n)– (V₁₀)_, is the most extensively studied POV regarding its biological activity, such as antitumor and antibacterial activity, and is used in protein crystallography, serving as a model compound for understanding the interactions between POVs and proteins [1]. The potential of V₁₀ as an antitumor agent has been explored with promising results, for example, for the hybrid compound Metformin-V₁₀ that showed stronger antiproliferative effects on melanoma cells [2]. Many of the in vivo and in vitro biological effects of V₁₀ have been linked to inhibition of P-type ATPases [1].

Calcium ions are universal secondary messengers that play a vital role in the functioning of mitochondrial processes such as adenosine triphosphate (ATP) production, reactive oxygen species (ROS) generation, and apoptosis [3]. An imbalance in Ca²⁺ levels is associated with several diseases, including cancer, diabetes, Parkinson’s disease, and Alzheimer’s disease [4,5,6]. The Ca²⁺-ATPase of the sarcoplasmic reticulum, found in muscle, is a transmembrane ion transporter. It is one of the most representative enzymes belonging to the family of P-type ATPases involved in cytosolic Ca²⁺ homeostasis. This enzyme is vital for the proper functioning of cells, and in the specific case of Ca²⁺-ATPase, it regulates muscle contraction and relaxation [7]. Ca²⁺-ATPase inhibitors have been used as drugs for cardioprotection, immunosuppression, and antitumor agents. The latter effect occurs due to the increase in cytoplasmic Ca²⁺ levels, which consequently activates apoptotic factors, leading to cell death [8,9]. As previously described, these P-type ATPases were considered one of the potential targets for POMs, which are expected to develop into the next generation of anticancer drugs that selectively target cancer cells [10]. Moreover, the calcium homeostasis dysfunction, a well-known hallmark of cancer, has recently been reviewed for its role in chemoresistance [11].

Recently, a POV with all vanadium centers in +V oxidation state, PV_{14 ox}, was tested regarding its effect on U87 glioblastoma cells, and it was also reported to inhibit Ca²⁺-ATPase and Na⁺/K⁺-ATPase in vitro and ex vivo, respectively [12]. Besides targeting these enzymes, also classified as P-type ATPases, POMs were also described to act as agonists of P2X receptors in hippocampal neuronal HT-22 cells, with the activation of metabotropic purinergic P2 receptors responsible for about 80% contribution to the cytosolic Ca²⁺ increase [13]. Therefore, by affecting cellular Ca²⁺ homeostasis implicitly, POMs might interfere with several physiological and pathological processes, among others [13,14].

Mixed-valence polyoxovanadates (MV-POVs) can adopt various structural patterns, including wheel and bowl-type structures, Lindqvist-type [15], and a range of pseudo-spherical, host–guest vanadium-oxido {V₁₄}, {V₁₅}, {V₁₆}, {V₁₇}, and {V₁₈} clusters, with different combinations of +III, +IV, and +V oxidation states. Studies using Density Functional Theory, DFT, have demonstrated that in such compounds, the electrons in the aggregate tend to be uniformly distributed over all the vanadium centres [16]. The pentadecavanadates, [X@V₁₅O₃₆]^q–, where X = Cl⁻, Br⁻, and CO₃²⁻, and net charge q = 4, 6, or 8, consist of fifteen edge-shared {O=VO₄} square pyramids arranged in a pseudo-spherical structure with approximate C_2v symmetry, encapsulating a single anion. These cage-like MV-POVs have been studied for their magnetic and electrochemical properties and applied to build a wood-based solar evaporation generator that achieves 90% efficiency for water purification [17].

Although MV-POVs encompass the oxidation states most frequently observed under physiological conditions, their therapeutic potential remains comparatively underexplored in relation to other polyoxometalates [18]. Specifically, the [(CH₃)₄N]₆[Cl@V^IV₈V^V₇O₃₆] (V₁₅) [19] has been evaluated in its interaction with different biological targets in the last 15 years (Figure 1). In studies developed by our research group, V₁₅ showed a chemoprotective effect of pUC19 plasmid DNA of up to 70%, minimizing the action of potentially carcinogenic agents such as the alkylating agent diethyl sulfate. Moreover, Escherichia coli cultures have been used as a more complex cell model system, resulting in approximately 40% chemoprotection for V₁₅ and [I@V^IV₁₂V^V₆O₄₂]⁷⁻ (V₁₈I) [20]. Correlation with spectroscopic studies has suggested a relationship between stability in solution and chemoprotective performance, where MV-POVs containing encapsulated halides V₁₅ and V₁₈I have shown more promising results than V₁₀ and those containing phosphate encapsulated [20]. The studies developed in vitro have enabled us to evaluate the reactivity patterns of MV-POVs in a controlled biological environment, thereby establishing structure-activity-reactivity relationships.

Studies involving ABC transporter P-glycoprotein showed that V₁₅ can inhibit the multidrug resistance associated with this membrane transporter. Currently, the resistance to multiple drugs with different chemical structures and mechanisms of action is a major challenge in conventional cancer therapies [21]. Recently, it was demonstrated that V₁₅ interacts with Chinese Hamster Ovary cells, decreasing the packing of membrane lipids and increasing signal transduction pathways mediated by the aggregation of the luteinizing hormone receptor (LHR) [22]. The LHR is a G-protein-coupled receptor that plays a key role in reproductive physiology in both males and females [22]. In addition, it was recently reported that a crystallographic structure of a model lysozyme, in which different MV-POV, including [(H₂O)@V₁₅O₃₆]^5–, non-covalently interact with enzyme surface residues [23]. Recently, the crystalline structure of a Cl@V₁₅-ferritin adduct was reported as obtained from a suspension containing [V^IVO(acac)₂], where acac = acetylacetonato, and H-chain ferritin in basic medium, under mild conditions [24]. This physiologically relevant protein was involved in the spontaneous formation and stabilization of the mixed-valence [Cl@V₁₅O₃₆]⁶⁻, containing 8V(IV): 7V(V), which corresponds to the classical polyoxidoanion described by Muller and used in this work, differing only in the cations. These achievements should encourage further research into mixed-valence cage-like chemistry and delve into their biological properties.

Among the few in vivo studies with POVs accessing anticancer activity, V₁₀ prevented and suppressed tumor growth (B16-F10 melanoma allografted mice) in 70% [25]. A heterometallic Mo/V-POV reduced breast tumor growth in BALB/c female mice [26]. The V₁₀ was evaluated for its toxicity to Sparus aurata fish [27], while V₁₅ was assessed in mice through an acute 28-day repeated toxicity study. To our knowledge, no other POV has been studied so far, finding a moderated toxicity following a single oral dose of V₁₅ from 25 to 2000 mg kg⁻¹, even after 14 days of exposure. However, repeated daily exposure over 28 days caused high toxicity at doses above 25 mg/kg, and V₁₅ was classified as level 5 in the oral assay [28]. Taking all together, not only V₁₅, but also other compounds with physiologically accessible oxidation states of vanadium must be considered to better understand MV-POV chemistry, with a view to the safe future use of drugs based on vanadium.

Figure 1. Timeline of the cage-like MV-POV [(CH₃)₄N]₆[Cl@V₁₅O₃₆] (V₁₅), emphasizing its role in biological applications, with ball-and-stick and polyhedral representations (the cations were omitted for clarity) [19,20,21,22,24,28,29].

Figure 1. Timeline of the cage-like MV-POV [(CH₃)₄N]₆[Cl@V₁₅O₃₆] (V₁₅), emphasizing its role in biological applications, with ball-and-stick and polyhedral representations (the cations were omitted for clarity) [19,20,21,22,24,28,29].

The previous studies motivated us to further explore the biological activities of V₁₅ to analyze its anticancer potential. Breast cancer is the most common cancer and causes the highest mortality among women worldwide, and is subdivided according to molecular mechanisms and aggressiveness [30]. The metastatic disease, responsible for more than 90% of cancer-related deaths, is defined by the ability of cancer cells to detach from the primary tumor and, through the bloodstream, invade nearby tissues, forming secondary tumors in different organs. The molecular classification of breast cancer subtypes mainly involves the estrogen and progesterone receptor genes (RE and RP, respectively) and the HER2 protein [31]. There are 4 molecular subtypes, with 70% of tumors classified as luminal A and B, characterized by the expression of hormone receptors RE and RP and negative for HER2 (Luminal A) or weakly positive for HER2 (Luminal B). Luminal subtypes are the least aggressive, with luminal B being more aggressive than luminal A. In order of aggressiveness, 15–20% are HER2 positive (overexpressing the HER2 protein) and 15% are triple negative (TNG), i.e., they do not have hormone receptors or overexpress HER2 [31]. Despite optimal systemic chemotherapy, fewer than 30% of women with metastatic breast cancer survive 5 years after diagnosis, and virtually all women with metastatic triple-negative breast cancer will ultimately die of their disease [32].

For luminal tumors, therapy uses tamoxifen (hormone therapy) or aromatase inhibitors, and for HER2 tumors, the monoclonal antibody trastuzumab (anti-HER2 antibody) is used. Advanced breast cancer, especially TNG, presents a highly metastatic profile and drug resistance. The lack of specific therapeutic targets means that TNG tumors are treated only with chemotherapy agents (e.g., doxorubicin and/or paclitaxel). However, chemotherapy agents present severe acute side effects ranging from vomiting and intense malaise to adverse effects such as cardiotoxicity, neurotoxicity, and neuropathy [33,34]. In this sense, vanadium compounds, including POVs, are promising chemotherapy agents, besides the well-known inorganic-based drugs containing cisplatin [35,36], as several studies have shown that they are active at different stages of carcinogenesis, such as prevention, therapy, and early diagnosis [37].

Nowadays, in breast cancer, we have found many research fields and/or perspectives such as cancer genomics, cancer therapy resistance, artificial intelligence tools, DNA repair and immunosurveillance, tumor microenvironment, clinical cancer biomarkers, non-invasive diagnostics, signaling in triple-negative breast cancer, chemotherapeutic agents for breast cancer, association with other cancer types, among many others [38]. Additionally, cancer research is conducted to discover new inorganic compounds, such as POMs, that combine inhibitory effects and/or reversal of drug resistance, as well as for breast cancer photothermal therapy using POMs-based nanoparticles [26,35,36,39].

The aim of this work is to pursue the studies of the putative biological activities of the mixed-valence pentadecavanadate, [(CH₃)₄N]₆[Cl@V₁₅O₃₆] (V₁₅), analyzing the Ca²⁺-ATPase inhibition potential and its anti-breast cancer activity in the luminal breast cancer cell line (MCF-7) and the TNG breast cancer cell line (MDA-MB-231). Specifically, in this work, we undertook the following tasks: (i) the evaluation of the V₁₅ stability and speciation under the experimental conditions employed using ⁵¹V NMR and UV/Vis spectroscopy; (ii) the determination of the IC₅₀ value for inhibition of Ca²⁺-ATPase activity, using a enzymatic couple assay; (iii) the determination of V_max and K_m kinetic parameters, regarding the native ligand ATP, using the couple enzymatic assay and the Lineweaver-Burk plot of Ca²⁺-ATPase in the absence or presence of inhibitor for determination of the type of enzyme inhibition; (iv) comparison of the electrostatic potential map of [V₁₀O₂₈]⁶⁻ (V₁₀), [MnV₁₃O₃₈]⁷⁻(MnV₁₃), [V^V₁₄O₃₈(PO₄)]⁹⁻ (PV_14ox) and also [Nb₁₀O₂₈]⁶⁻ (Nb₁₀) with [Cl@V^V₇V^IV₈O₃₆]⁶⁻ (V₁₅); (v) determination of the V₁₅ IC₅₀ values against MDA-MB-231 and MCF-7 cell lines, at 24 h; (vi) analysis of the effect of V₁₅ in MDA-MB-231 cell migration; (vii) to verify V₁₅ modulation of epithelial–mesenchymal transition (EMT) in MDA-MB-231 by cell morphology alteration; (viii) the putative mechanism of cell death induced by V₁₅ searching for necroptosis and apoptosis.

2. Results and Discussion

2.1. Stability Studies

Although the characterization of POV structures in the solid state is important, understanding their behavior in solution is essential to identify the vanadium species present in equilibrium, which may include bioactive forms that contribute to the overall biological activity. It is known that the stability of POVs depends on the counter cation [40], the encapsulated anion [16], and the buffer or culture medium compositions used in the biological assay [41]. In the physiological medium, interaction of POVs with peptides and proteins may also change the nuclearity and oxidation state of the vanadium compound [42]. The stability study of V₁₅ was already performed in solution containing DMEM medium (pH 7.4) [22], water (pH 6.3) [29], PIPES (pH 7.5) [29], and LB (Luria–Bertani) medium [20]. These correlations are essential for the rational selection of new biological applications.

Herein, the speciation of 1.0 mM V₁₅ in enzymatic medium at pH 7.0 (25 mM HEPES, 100 mM KCl, 5.0 mM MgCl₂, 50 µM CaCl₂), and in RPMI medium (pH 7.4) after incubation at room temperature, was monitored by ⁵¹V NMR (Figure 2). Spectra were also recorded in water for comparison. In water/D₂O and RPMI, NMR spectra were recorded at 0, 3, 10, 24, 48, and 72 h, reflecting the conditions used in the cellular viability assay. In the enzymatic medium, spectra were acquired at 0 and 1 h, aligning with the time course of the Ca²⁺-ATPase inhibition assay.

After dissolving V₁₅ in H₂O, only a signal at δ –560 ppm was detected, attributed to the mononuclear [H₂VO₄]⁻ complex (V₁), indicating that V₁₅ hydrolyses partially [43]. Since V₁₅ contains seven V(V) and eight V(IV) centers, the intact mixed-valence cluster was not detected by ⁵¹V NMR. Over time, the concentration of vanadium(V) species increases, and after approximately 10 h, the signals characteristic of the three vanadium(V) environments of V₁₀ appeared (Figure 2c) [43]. The intensity of these signals increased over time, with isomeric shifts at—423 (V_A), –504 (V_B), and –522 (V_C) ppm, consistent with the protonated [HV₁₀O₂₈]⁵⁻ species [43].

In enzymatic medium, the ⁵¹V NMR of V₁₅ reveals signals for V₁, [H₂V₂O₇]²⁻ (V₂) and [V₄O₁₂]⁴⁻ (V₄) at –560, –574 and –578 ppm, respectively (Figure 2a). The speciation of POVs in solution is strongly pH-dependent, with V₁₀ forming predominantly under acidic environments. This explains why V₁₀ forms upon V₁₅ hydrolysis in water but does not form in neutral buffered solutions. Given that RPMI medium contains nutrients such as amino acids, phosphate, and glucose, which are capable of binding metal ions, a more complex speciation profile was expected. Indeed, NMR signals corresponding to oligovanadates V₂, V_4, and [V₅O₁₅]^5- (V₅) at δ –585 ppm were observed (Figure 2d) during the time of the experiment. Additionally, the signal at δ –558 ppm of V₁ represents a combination of V₁ and a phosphate-vanadate (PV) complex (V₁ + PV), as these species rapidly exchange on the NMR timescale [22].

The electronic paramagnetic resonance (EPR) spectroscopy at X-band complements the ⁵¹V NMR studies, as EPR detects the paramagnetic V(IV), while ⁵¹V NMR detects V(V) species. The 1.00 mM freshly prepared solutions of V₁₅ in water, enzymatic, and RPMI media at 77 K (Figure S1) gave weak broad lines compatible with those previously described for the mixed-valence polyoxidoanions [20,22,29]. Although some variations in the spectra can be observed over 72 h, the poor signal-to-noise ratio prevented detailed analysis. However, no resolvable hyperfine structure, characteristic of vanadium(IV) mononuclear species, was detected. These results reinforce V₁₅ stability in different conditions and a possible oxidation of vanadium(IV) to vanadium(V) upon hydrolysis.

Additionally, the stability of V₁₅ in enzymatic and RPMI medium was monitored using UV/Vis/NIR spectroscopy over a period of 168 h, and the results were compared with those registered for the aqueous solution (Figure 3). In the initial time, the spectra displayed a broad band centered at 950 nm assigned to a partially delocalized intervalence charge-transfer transition (IVCT, type II, V(IV)→V(V)) [44] characteristic of MV-POVs [29]. A second, higher-intensity band, starting at approximately 500 nm and extending to the ultraviolet region, was also observed, and attributed to ligand-metal charge transfers (LMCT, pπ (O, oxo) → d(V)). In the aqueous solution, some oxidation of V(IV) to V(V) occurred, causing a gradual reduction in IVCT absorbance. After 168 h, the V₁₅ concentration had dropped to 46% of its initial value, as found by the calibration curve (Figure S2). The aggregate also undergoes partial hydrolysis, resulting in a mixture of V₁₅, V₁, and V₁₀, as suggested by NMR data along with the evolution of UV/Vis/NIR spectrum profiles and the change in the color of the solution from intense green to yellowish green. It is well known that the octahedral V(V) in V₁₀ causes a bathochromic shift in the LMCT bands compared to V(V) in other geometries, and that this transition is responsible for the yellow color of the anion [45].

Despite the speciation observed in NMR spectra in the initial time, V₁₅ showed greater stability in the enzymatic medium containing HEPES pH 7.0, a buffer with a weak affinity for binding metal ions, in RPMI medium pH 7.4 (Figure 2b,d), reducing the MLCT band intensity by ca. 10% after 168 h. RPMI, on the other hand, has a composition rich in amino acids, including aspartate, that might interact with the POV and stabilize it or lead to redox reactions, depending on the conditions (temperature, pH, concentration, and ionic force of the medium). In the face of the above, V₁₅ is more kinetically stable to hydrolysis for days under physiological conditions and is not completely oxidised or broken down into oligovanadates before exerting the biological effect reported here.

2.2. Inhibition of Ca²⁺-ATPase

The effect of V₁₅ on the activity of sarcoplasmic reticulum Ca²⁺-ATPase from skeletal muscle was evaluated in vitro. V₁₅ inhibited enzyme activity in a concentration-dependent manner, with a half-maximal inhibitory concentration (IC₅₀) value of 14.2 μM (Figure 4a).

Among other POVs, decavanadate (V₁₀) exhibits an IC₅₀ of 15 μM [46]. In previously reported studies, the bicapped Keggin-type [V^V₁₄O₃₈(PO₄)]⁹⁻ (PV_{14 ox}) showed an IC₅₀ of 5 μM [12], while K₇MnV₁₃O₃₈ (MnV₁₃) and K₅MnV₁₁O₃₃ (MnV₁₁) presented IC₅₀ values of 31 μM and 58 μM, respectively [47]. Notably, both V₁₅ and V₁₀ carry the same net charge of −6 at physiological pH. Although their sizes and structures differ, the similarity in IC₅₀ values may be attributed to their charge and the predominance of electrostatic interactions and hydrogen bonds with the target protein, as previously described for other POMs [48]. V₁₅ solution at the IC₅₀ concentration (15 μM) was incubated in the enzymatic medium for 60 min prior to measuring its inhibitory effect on Sarcoplasmic/Endoplasmic Reticulum Calcium ATPase (SERCA) hydrolysis activity. After incubation, V₁₅ ATPase inhibitory activity was reduced to 42.4%, instead of 50% without incubation, indicating that the POV remains relatively stable in solution, even after 60 min. We emphasize that the IC₅₀ values were determined during the first two minutes after the POV addition. In fact, no incubation was made with the POMs in the medium to prevent any decomposition. Therefore, it is suggested that V₁₅ is the most probable species responsible for the observed inhibitory effects on the Ca²⁺-ATPase, as described elsewhere for similar experimental conditions with other POMs [12,46,47,48,49,50]. On the other hand, if putative V₁₅ decomposition into V₁ species occurs, it would lead to a lower inhibition potential activity for this ATPase, considering that the IC₅₀ values for V₁ species were reported to be lower (IC₅₀ = 80 μM) [49].

It was determined that V₁₅ exerts a mixed-type inhibition on the ATPase activity, as evidenced by a decrease in V_max and an increase in K_m relative to the control measured by the slope and intercept of the double-reciprocal plot in Figure 4b. This suggests that V₁₅ interacts with at least two distinct binding sites on the enzyme. In addition to the ATP-binding site, V₁₅ may also bind to other domains of the Ca²⁺-ATPase, regardless of whether or not the enzyme is bound to its substrate. Mixed-type inhibition has also been reported for other POVs, such as PV_14ox and MnV₁₃. In contrast, V₁₀ exhibits non-competitive inhibition with respect to ATP, suggesting mechanistic differences despite similar IC₅₀ values between V₁₀ and V₁₅. It is known that V₁₀ binds to all conformational states of the enzyme—E1, E2, phosphorylated or not—whereas monomeric vanadate binds exclusively to the E2 conformation [49].

Furthermore, V₁₀ can induce cysteine oxidation accompanied by the reduction of V(V) to V(IV), implying a potential role for V(IV) species in biological activity [51]. MV-POVs have already been observed, for instance, in lysozyme crystals [42]. X-ray structure studies have shown that [(OH₂)@V₁₅O₃₆]⁵⁻ and [(OH₂)@V₁₅O₃₃]⁺, both described as mixed-valence (V^V₇/V^IV₈), crystallize with lysozyme in a solution treated with monomeric vanadyl complex, [V^IVO(acac)₂] (acac = acetylacetonato). The interaction with the protein occurs through hydrogen bonds with lysine, tyrosine, and arginine residues [42]. Moreover, [V₁₅O₃₃(OH₂)]⁺ species was observed to form covalent bonds with the protein, where two POV oxygen atoms are replaced by those of aspartyl carboxylates, and a third is replaced by a glycine carbonyl oxygen. These interactions suggest that lysozyme structure provides a stabilizing environment for the V15-like-POVs [42,52].

These observations indicate that not only the surface charge, but also the POV specific structure and the presence of vanadium in the +IV oxidation state contribute to its interaction with Ca²⁺-ATPase and the resulting inhibition of its enzymatic activity.

Electrostatic Potential (ESP) Map

The inhibition of Ca²⁺-ATPase by POMs is frequently associated with their high negative surface charge and has been correlated with their charge density and size [48,53]. In fact, it was described that Preyssler-type anion [NaP₅W₃₀O₁₁₀]¹⁴⁻ (P₅W₃₀) [50] and [H₁₀Se₂W₂₉O₁₀₃]¹⁴⁻ (Se₂W₂₉¹⁴⁻) with a similar charge density (q/m = 0.47 and 0.48, respectively) shows the highest POMs Ca²⁺-ATPase inhibitory potential, respectively, IC₅₀ = 0.37 and 0.3 μM whereas [H₂P₂W₁₂O₄₈]¹²⁻ (P₂W₁₂) with a charge density of 1.00 presents a higher IC₅₀ value (11 μM) [48]. Besides the charge of the POM, charge density, and size, other features may contribute to POMs’ interactions with proteins and the inhibition of enzymatic activities. For example, the isostructural decavanadate (V₁₀) and decaniobate (Nb₁₀) with analogous structure, net charge, and approximately the same size but with different inhibition potential for the Ca²⁺-ATPase activity (See

Table 1. Surface maxima and minima of ESP (E_max and E_min, respectively), ΔE, and the volume enclosed by the calculated isosurface (V_isosurface 0.001 e/bohr³) and charge density (q/m).

Parameter	Nb10	MnV13	V10	V15	PV14 ox
IC50 μM	35	31	15	14.2	5
Emax/kcal mol−1	−323.31	−350.29	−329.76	−306.01	−439.53
Emin/kcal mol−1	−382.80	−468.82	−412.55	−354.20	−562.68
ΔE/kcal mol−1	59.5	118.5	82.79	48.19	123.1
Visosurface/Å3	629.94	558.32	744.23	790.49	835.19
q/m	0.6	0.54	0.6	0.4	0.64
(q/m)/Visosurface/10–4	9.524	9.644	8.062	5.060	7.697

ΔE = E_max − E_min; q/m = is expressed as charge (q) of the POM divided by its number of metal atoms (m); V_isosurface = volume of electron density in the POM.

presented below), have been shown to present different binding sites for G-actin [51]. Thus, it was shown that V₁₀ binds preferentially to the alpha binding site surrounded by Arg and Lys residues, being this the binding site of the natural substrate of the protein (ATP), whereas Nb₁₀ binds to a binding site beta [54]. Besides this specific POV-actin interaction, several distinct modes of POVs' interactions with proteins have been recently described [54]. Regarding the Ca²⁺-ATPase, V₁₀ was found to crystallize in a highly positively charged groove formed by three protein domains [52], highlighting its ability to bind electrostatically to this region.

The Electrostatic Potential (ESP) map is a useful tool for examining the most probable sites of non-covalent interactions, such as electrostatic attractions and hydrogen bonds. The ESP is broadly used to visualize how the electrostatic potential varies across the surface of a molecule, and it also aids in analyzing intermolecular interactions based on electrostatic complementarity. ESP calculations were performed for V₁₅ and for other selected discrete POVs that have been previously evaluated for their Ca²⁺-ATPase inhibitory activity [V₁₀O₂₈]⁶⁻ (V₁₀), [Nb₁₀O₂₈]⁶⁻ (Nb₁₀), [MnV₁₃O₃₈]⁷⁻ (MnV₁₃), which are expected to exist as deprotonated species under the enzymatic assay conditions [12]. The fully oxidized bicapped Keggin anion [V^V₁₄O₃₈(PO₄)]⁹⁻ (PV_{14 ox}) was also assessed despite its low kinetic stability in aqueous solution at pH 7, because it coexists in equilibrium with oligovanadates in freshly prepared solutions [55].

For better comparison, all volume electron densities were represented as isosurfaces (V_isosurface, Figure 5), and are drawn using a continuous color gradient at the same electrostatic potential scale. All values on the isosurfaces are negative, with the most electron-rich areas in red, indicating regions of highest attraction for a positively charged species or ions. The blue areas represent the less negative potential values, while the green regions indicate intermediate electron-rich areas. Except for the heterometallic MnV₁₃, the most negative electrostatic surface potential values of the POMs are localized in the vicinity of the doubly and triply bridging oxygen atoms, while the terminal oxygen atoms exhibited the least negative electrostatic potential. For MnV₁₃, the most negative potentials are positioned in the centre of the aggregate.

To shed light on the factors involved in the Ca²⁺-ATPase inhibitory activity promoted by POVs, some molecular and electrostatic features were attempted to correlate with IC₅₀ values depicted in Table 1. The relative order for both E_max and E_min is V₁₅ < Nb₁₀ < V₁₀ < MnV₁₃ < PV_{14 ox}, while the IC₅₀ order is Nb₁₀ < MnV₁₃ < V₁₀~V₁₅ < PV_{14 ox}. The analysis of the results reveals two important parameters: the surface maxima and minima values in ESP (E_max and E_min), and the volume enclosed by the isosurface (V_isosurface).

Regarding the border inhibitors, PV_{14 ox} shows the most negative E_min, the highest ΔE value (123 kcal mol⁻¹), and the biggest V_isosurface of 835.19 ų, being the most nucleophilic POV and exhibiting the strongest inhibition of Ca²⁺-ATPase (Table 1). On the other hand, the poorest inhibitor, Nb₁₀, presents E_min of—382.80 kcal mol⁻¹ and a ΔE value of only 59 kcal mol⁻¹; moreover, its V_isosurface is 25% smaller than the PV_{14 ox} one. The E_max and E_min for V₁₀ are slightly more negative than those for Nb₁₀, but its V_isosurface being 18% larger might be associated with its inhibitory activity, which is 1.75 times more potent. The MnV₁₃, in turn, exhibits the second-lowest Emin value, but its IC₅₀ is comparable to that of Nb₁₀. Here, the impact of the volume is significant because it showed the lowest V_isosurface, i.e., 29% and 33% smaller than V₁₅ and PV_{14 ox}.

The V₁₅ is a key aggregate to demonstrate the correlation between the Ca²⁺-ATPase activity with the ESP and the V_isosurface, because even showing the lowest E_max, E_min, and a ΔE of only 48 kcal mol⁻¹, its IC₅₀ value is close to the V₁₀ (15 and 14.2 μM). The V_isosurface calculated for V₁₅ is only 5.9% bigger than V₁₀ and 5.4% smaller than PV_{14 ox}, but considerably higher than Nb₁₀, achieving 25%. These results showed that, despite their structural differences and variations in vanadium oxidation states, V₁₀ and V₁₅ display similar availability of negative charge to participate in electrostatic interactions with the expected binding site of this enzyme [48]. However, the relationship between delocalization and electron density distribution in ESP is not straightforward, and there are still few systematic structural studies with mixed valence POVs to trace a linear correlation. As can be observed in Table 1, V₁₅ presents a charge density (q/m) of 0.40. It was reported that POMs with moderate charge density (q/m = 0.33) interact strongly with protein surfaces due to a process named chaotropic effect [56]. It was suggested that this chaotropic effect contributes to a structure–activity relationship for the affinity of POMs towards proteins [56]. Still, further studies are needed to unravel at the molecular level the POMs features that correlate with specific biological activities.

2.3. Inhibition of Cell Viability in Breast Cancer Cells

The MTT assay was performed to evaluate cell viability using V₁₅ in the non-tumorigenic, human mammary epithelial cell line (HB4a), luminal breast cancer cell line (MCF-7), and the triple-negative breast cancer cell line (MDA-MB-231). The cells were initially introduced into the wells to allow cell adhesion for 2 h, followed by the addition of increasing amounts of V₁₅ (Figure S3). The results of cell viability inhibition of the cell lines treated with V₁₅ showed a dose-dependent effect for the three cell lines, with an IC₅₀ value of 1.02 μM for HB4a, 15.1 μM for MCF-7, and 17.2 μM for MDA-MB-231 (Figure 6 and

Table 2. IC₅₀ values determined for V₁₅ regarding Ca²⁺-ATPase inhibition, breast cancer cell lines (MCF-7 and MDA-MB-231), and normal breast cell line (HB4a) viabilities, after 24 h of incubation.

Assay	IC50 (µM)
Ca2+-ATPase	14.2
HB4a	1.02
MCF-7	15.1
MDA-MB-231	17.2

). The IC₅₀ value for luminal breast cancer cells was 16 times more potent than the commercial drug 5-fluorouracil under similar conditions (24 h, RPMI) [57].

The non-tumorigenic cell line showed higher sensitivity to V₁₅ compared to the two cancer cell lines, indicating that the POV does not exhibit selectivity regarding this experimental model. Despite the lack of selectivity, in our previous studies with V₁₅ using peripheral blood mononuclear cells (PBMC), an IC₅₀ value of 12.9 μM was determined in RPMI after 48 h [28]. These data, along with early toxicity studies in mice, may guide future in vivo research when focusing on the urgent need to develop new therapies for the triple-negative breast cancer subtype. In such studies, a V₁₅ concentration in the order of 1.0 μM (see Figure S2) would be reasonable, because it produced a cytotoxic effect of approximately 30–40% on MDA-MB-231 cells.

The IC₅₀ value of V₁₅ in MCF-7 and MDA-MB-231 cell lines was selected to monitor cell viability over 24, 48, and 72 h. The graphs presented in Figure S4 demonstrate that the POV is a time-dependent potent inhibitor of cell growth in both cell lines, corroborating the value determined for the IC₅₀ after 24 h of incubation. The time curve showed a significant decrease in the viability of MDA-MB-231 cells after incubation, with only 45.4% (24 h), 15.3% (48 h), and 8.3% (72 h) of cells remaining alive. Similar results were observed with MCF-7, showing 64.3% (24 h), 12.1% (48 h), and 3.03% (72 h). Therefore, V₁₅ showed a cumulative effect in the interference of cell viability over time.

The IC₅₀ values obtained for V₁₅ in the MCF-7 and MDA-MB-231 cell lines are promising, as they are lower than those found in the literature for other polyoxovanadates. For MCF-7 cells, a literature POV hybrid covalently bonded to amino acid esters, [V₆O₁₃{(OCH₂)₃C₁₀H₁₄O₅}₂]²⁻, exhibited a cytotoxic effect (IC₅₀ = 53.01 μM) twice as strong as that of the traditional commercial drug 5-fluorouracil, in DMEM at 48 h [58]. Two decavanadates with betaine showed anticarcinogenic activity against MCF-7 cells at 24 h, with IC₅₀ values close to 200 μM [59] while the super-Keggin MV-POV [(H₂O)@V₁₈O₄₂]¹²⁻ (H₂O@V₁₈) presented IC₅₀ values of 45.95 µM and >500 µM for 24 h, respectively [60]. According to spectroscopic studies, V₁₅ is stable in RPMI medium, suggesting that the activity could be attributed to this MV-POV. To the best of our knowledge, no speciation studies are available for H₂O@V₁₈. The H₂O@V₁₈ presents the lowest cytotoxic effect, being approximately 3 and 30 times less potent than V₁₅ for MCF-7 and MDA-MB-231 cells, respectively.

2.4. Inhibition of Cell Migration Using the Wound Healing Assay in MDA-MB-231 Cell Line

The Wound Healing assay is a simple method used to study cell migration. Reducing migration is desirable for a therapy that aims not only to diminish a tumor but also to delay or even prevent metastasis. The MDA-MB-231 cell line exhibits mesenchymal characteristics (fusiform shape), which is indicative of its migratory and invasive potential [61,62]. The cancer cells were treated with V₁₅ at concentrations of 8.60 and 4.30 μM, equivalent to half and a quarter of the IC₅₀ value, and the wound area was photographed after removing non-viable cells. Figure 7a displays selected images where, in the control, the typical migration of the cell line over the scratch is observed. In both samples treated with V₁₅, the wound remained open, indicating that cell migration was impaired.

The wounds in the wells containing cells treated with both concentrations of V₁₅ for 24 h were compared with the wounds of the control (non-treated cells) incubated for 24 h (Figure 7b). In the control samples, the wound area remained open by only 13%. For V₁₅, a dose-dependent cell migration was observed after 24 h, and consequently, the gap area remained 66% and 86% open for the lower and higher concentrations, respectively. Similar results were reported for decavanadate with tetra-(benzylammonium), which also inhibited cell migration of MDA-MB-231 at concentrations of 0.70 and 2.8 μM [63].

Figure 7. Inhibition of cell migration as analyzed by the Wound Healing Assay in the MDA-MB-231 cell line without treatment (negative control) and treated with V₁₅ for 0 and 24 h. (a) Selected photos of cell migration over the wound, from a total of 6 repetitions. The zoom is related to the wound area. (b) The graph was generated by taking six images in each condition, with the aid of ImageJ.JS 1.53m software [64]. The asterisks indicate the statistical treatment for six repetitions compared to the control at 24 h. **** means p < 0.001. The bars represent the standard deviation.

Figure 7. Inhibition of cell migration as analyzed by the Wound Healing Assay in the MDA-MB-231 cell line without treatment (negative control) and treated with V₁₅ for 0 and 24 h. (a) Selected photos of cell migration over the wound, from a total of 6 repetitions. The zoom is related to the wound area. (b) The graph was generated by taking six images in each condition, with the aid of ImageJ.JS 1.53m software [64]. The asterisks indicate the statistical treatment for six repetitions compared to the control at 24 h. **** means p < 0.001. The bars represent the standard deviation.

2.5. Scanning Electron Microscopy (SEM Images) in MDA-MB-231 Cell Line

SEM analysis was performed on the MDA-MB-231 cell line at the same concentrations as those used in the Wound Healing assay (4.30 and 8.60 µM). The SEM images showed cells with a fusiform shape, characteristic of this mesenchymal subtype of triple-negative cell line (Figure 8a), which changed to a more epithelial-like (amoeboid) morphology (Figure 8b,c). This morphology is generally related to a lower migratory capacity, which corroborates the cell migration assays. Finally, the microvesicles that appeared after treatment with the highest dose of V₁₅ might be associated with the beginning of the cell death process. These morphological changes in the cells may be related to interactions with the cytoskeleton and possibly with the actin [65].

A process that is related to the progression of the disease to the metastatic form is called epithelial–mesenchymal transition (EMT). EMT is a reversible process, transitioning from epithelial to mesenchymal cells through a change in morphology. EMT occurs naturally in several processes, such as wound healing and embryonic development. However, when associated with the formation of fibroblasts (tissue regeneration, trauma, or inflammation), it can cause fibrosis and organ damage. When in neoplastic cells, there is an increase in cell proliferation and invasive and metastatic phenotypes, increasing the malignancy and aggressiveness of the disease [65]. The EMT phenotypical alteration involves genetic and epigenetic modifications in cancer cells when they initiate the metastasis process. Here, we observed a potential reverse process (mesenchymal–epithelial transition), indicating that V₁₅ can switch off one of the most important hallmarks of cancer, thereby reducing the mechanism of progression to metastatic moving cells [65]. However, assays carried out at concentrations up to 1.0 μM of V₁₅ caused only minor changes in cell morphology (Figure S5).

2.6. Gene Expression in MDA-MB-231 Cell Line

There are three classic mechanisms of cell death: apoptosis, necrosis, and necroptosis. Apoptosis is a programmed and “organized” cell death, essential for the development and homeostasis of the organism, and it does not trigger an inflammatory response. It is characterized by cell contraction and formation of apoptotic bodies, and since the cell contents are not released into the bloodstream, there is no inflammatory response. This mechanism can be triggered by internal signals, such as DNA damage or cell death signalling [66]. Necrosis is an unprogrammed cell death, usually triggered by injury, anoxia, or extreme tissue damage. Unlike apoptosis, it is an inflammatory process that results in the release of cellular contents into the bloodstream. It is characterized by cell swelling and rupture of the plasma membrane, leading to the release of cellular contents and the consequent inflammatory response. This mechanism can be caused mainly by a lack of blood supply [66].

Necroptosis is a cell death that resembles classical necrosis in morphology; however, it is a programmed inflammatory cell death. Unlike necrosis, it depends on a signaling pathway mediated by kinases called RIPK1 and RIPK3, with the pseudokinase MLKL acting as its effector molecule [66]. It is characterized by the controlled rupture of the plasma membrane and the release of cellular contents, but with some regulation compared to necrosis [66].

Based on the results presented above for the POVs, it was decided to investigate the cell death mechanism only for V₁₅ in the MDA-MB-231 cell line, since it alters cell morphology, which may indicate interaction with the membrane and the triggering of processes inside the cell. The selected genes were those related to the mechanisms of apoptosis (BCL2, antiapoptosis; TP53, apoptosis mediated by DNA damage; CASP8, apoptosis via extrinsic pathway; and BAX, pro-apoptosis), and necroptosis (RIPK1, MLKL, and RIPK3). The graphs of relative gene expression compared with control without treatment and normalized by GAPDH constitutive gene (Figure 9) showed that those referred to necroptosis mechanism, RIPK1, MLKL, and RIPK3, increased 3, 10, and 15-fold, respectively (Figure 9a). The expression of these genes indicates that one of the mechanisms of cell death caused by V₁₅ in the TNG breast cancer cell line may occur by necroptosis. On the other hand, no statistically significant effect on the general apoptosis mechanism could be related to V₁₅ in this model (Figure 9b).

Interaction with V₁₅ can trigger the necroptosis pathway, signaled by the activation of the RIPK3, RIPK1, and MLKL genes, which are factors involved in the formation of the necrosome, activated when caspase inactivation or inhibition occurs. The necrosome, along with other factors, migrates to the cell membrane, causing rupture and leakage, leading to cell death [67]. Some studies with POMs confirm that this class of compounds can cross the cell membrane of cancer cells, localizing itself within the cytoplasm; however, the mechanism of action of POMs has not yet been thoroughly explored [68,69,70,71,72,73]. About forty years ago, Yamase performed the first study on the mechanism of POMs in cancer cells [74]. The proposed mechanism involves the reduction and reoxidation of POM, as well as of the cellular components of the electron transport chain, which can inhibit the formation of ATP, leading to cell death by apoptosis [74]. This proposed mechanism is based on a series of studies that suggested that the cytotoxicity of bioactive POMs directly relates to their redox potential; however, other factors, such as size, structure, and composition, also significantly influence the activity of POMs against cancer cells [10].

POMs might affect cancer cells through several mechanisms, including the ones referred above, such as inducing apoptosis and affecting bioenergetics processes, but other mechanisms were described, such as interfering with DNA replication and gene expression, cell cycle arrest, generating reactive oxygen species (ROS), disrupting membrane transports, and/or inactivating essential enzymes within cancer cells [10,75]. Of course, those specific mechanisms for anticancer activity might depend on the POM’s structure and composition in order to selectively target cancer cells and reduce toxicity to healthy cells. When comparing and sorting the cell viability IC₅₀ values in ascending order for 20 human cancer cell lines tested for different types of POMs, the first polyoxovanadates (POVs), polyoxomolybdates (POMos), polyoxopaladates (POPds), and polyoxotungstates (POTs) [75]. When comparing clinically approved drugs and POMs, in many cases, better results were observed for POMs in relation to the drugs [75]. Therefore, POVs such as V₁₅ could become in the future an alternative to existing drugs in cancer therapy [75].

Alterations in calcium levels related to cell death by necroptosis have recently been revised, and it has been proposed that classical necroptosis is mediated by death receptors that work in synergy with caspase inhibitory signals [76]. However, a non-classical model of necroptosis has emerged. Based on the argument that cytosolic calcium signaling can drive the necroptotic mode of cell death in both dependent/independent receptor mechanisms when the cytoplasmic calcium levels are strictly related. Here we have found an inhibition of Ca²⁺-ATPase activity and a putative correlation with a decrease in cell proliferation, a phenotypic switch from mesenchymal to epithelial, and cell death by necroptosis. More detailed studies are in progress to clarify this probably new mechanism of drug interaction with promising therapeutic future intervention [76].

3. Materials and Methods

3.1. Generalities

The synthesis reactions were carried out in air, in deionized water (ultrapure water, resistivity less than 3.5 μS·cm⁻¹ at 25 °C). All the chemicals purchased were of reagent grade and were used without further purification. Roswell Park Memorial Institute (RPMI) medium and fetal bovine serum (FBS), both from Thermo Fisher Scientific (Waltham, MA, USA); 4-(2-hydroxyethyl)1-piperazine ethane sulfonic acid (HEPES) from Sigma-Aldrich (St. Louis, MO, USA); 3-(4,5-dimethyltiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT), sodium cacodylate, glutaraldehyde, and osmium tetroxide were provided by Merck (Darmstadt, Germany). The AllPrep DNA/RNA Mini Kit (Qiagen, Hilden, Germany), SuperScript IV from Invitrogen (Walthan, MA, USA), and SYBR Green Power Master Mix from Applied Biosystems (Foster City, CA, USA) were used as received.

Powder X-ray diffractogram (PXRD) was registered on a Shimadzu XRD-6000 equipment (Kyoto, Japan), with Cu-Kα radiation (λ = 1.5418 Å) using silicon powder as an internal standard, using a voltage of 40 kV, current of 30 mA, and scan speeds of 0.02°s⁻¹ (at 2ϴ). The ⁵¹V NMR spectra were recorded using a Bruker Avance 400 MHz (9.4 T) equipment (Ettlingen, Germany) , equipped with a direct detection multinuclear probe (5 mm). The spectra were acquired at 303 K with 500 μL of each sample solution, using 90° pulses. The spectra were registered with 2048 scans, with a recycle delay of 0.100 s and acquisition times of 0.218 s, along a spectral window of 714 ppm (+ 50 ppm to −650 ppm) with a ⁵¹V core acquisition frequency of 105.25 MHz. Electron Paramagnetic Resonance spectra were recorded in an X-band Bruker ELEXSYS E-500 spectrometer (Ettlingen, Germany) from solution frozen at 77 K. The spectra were referenced using pure VOCl₃ in a capillary (external reference), fixing the signal at 0.00 ppm. UV-vis/NIR spectra were recorded on a PerkinElmer LAMBDA 1050 UV-Vis/NIR spectrophotometer (Waltham, MA, USA) equipped with three PMT/InGaAs/PbS detectors at room temperature in a wavelength range of 265–1300 nm. Scanning Electron Microscopy (SEM) analyses were performed using a TESCAN VEGA3 LMU equipment (Brno, Czech Republic) , with a resolution of 3.0 nm, at magnifications ranging from 500× to 10,000×.

Synthesis and Stock Solution of V₁₅

The (Me₄N)₆[Cl@V₁₅O₃₆], V_15, was synthesized following the procedure described in the literature [29] and in the Supplementary Material. A detailed scaled-up synthesis was described for some of us in reference [28]. IR spectra were registered for a freshly synthesized sample, after two months and two years of aging. Main bands in IR spectrum (Figure S6) of dark green crystals: ν_as and ν_s (CH₃) weak bands at 3024 and 2952 cm⁻¹ respectively, strong δ_as and δ_s(CH₃) absorptions at 1485 cm⁻¹ and 1446 cm⁻¹ respectively, and ν(V=O) at 983 cm⁻¹; ν(V−O−V) at 580, 660 and 725 cm⁻¹. The diffraction peaks at angles 2θ of 7.4, 8.6, 15.4, 17.6, 19.9, 22.3, 23.1, 25.7, 26.7, 27.2, 30.7, 32.8, 33.2, 41.2 and 41.7° (Figure S7) are in good agreement with those previously reported, confirming that the solid was isolated free of the fresnoite-type oxide (NH₄)₂V₃O₈, a common insoluble byproduct in the synthesis of mixed-valence POVs [28].

New data on V₁₅ stability in solid state over a period of 2 years were obtained by the infrared spectroscopic analysis, indicating that this MV-POV remained stable for a long time of storage (Figure S6). Stock solutions of V₁₅ were prepared at a concentration of 1.0 mM (pH 4.67) in heated water, considering a molecular weight of 1820.42 g mol⁻¹ to yield a green solution. Freshly prepared solutions were used in all experiments to prevent oxidation of the vanadium. The aqueous solution was diluted in RPMI culture medium or enzymatic medium to the desired concentration specified in each biological assay. The enzymatic medium contained: 25 mM HEPES (pH 7.0), 100 mM KCl, 5 mM MgCl₂, and 50 μM CaCl₂.

3.2. Spectroscopic Studies in Solution

The vanadium species formed in a 1.0 mM V₁₅ solution were determined using vanadium nuclear magnetic resonance (⁵¹V NMR) spectroscopy. For sample preparation, 450 μL of 1.11 mM stock solution of V₁₅ in water, RPMI pH 7.4 (9:1), without the addition of fetal bovine serum or enzymatic medium, received 50 μL of D₂O, to render a final concentration of 1.0 mM. The spectra in aqueous solution and in RPMI medium were recorded at several points (0, 3, 10, 24, 48, and 72 h) to determine whether there were changes in the speciation during the time of the cellular cytotoxicity experiment. The spectra in the enzymatic medium were recorded at 0 h and 1 h.

Ultraviolet, visible, and near-infrared (UV/Vis/NIR) spectroscopy was used to follow the consumption of V₁₅ at times of 0, 3, 10, 24, 48, 72 and 168 h, measuring the absorption band from 265 to 1300 nm in a PerkinElmer LAMBDA 1050 UV/Vis/NIR spectrophotometer (Waltham, MA, USA) equipped with three PMT/InGaAs/PbS. All spectra were recorded at ambient temperature in water, RPMI, and enzymatic media. A 1.0 mM solution of V₁₅ in each condition was incubated for up to 168 h at 25 °C, and the spectrum was recorded at 24 h intervals. Due to the high molar absorptivity of V₁₅, all samples had to be diluted to 0.020 mM prior to analysis to guarantee a detection within the linear range.

3.3. Computational Methods

The crystallographic data for the selected polyoxovanadates were available in the Cambridge Crystallographic Data Centre with CCDC numbers 722216 [Nb₁₀O₂₈]⁶⁻ [77], 1610770 [MnV₁₃O₃₈]⁷⁻ [78], 1839623 [V₁₀O₂₈]⁶⁻ [79], 794586 [Cl@V^V₇V^IV₈O₃₆]⁶⁻ [29], 1886317 [V^V₁₄O₃₈(PO₄)]⁹⁻ [12]. The structure of V₁₀ was subject to geometric optimisation after removal of hydrogens, using Density Functional Theory (DFT) with the WB97X-D3 functional [80,81] and DEF2-TZVP basis set [82], utilising ORCA 5.0.4 software [83]. Electrostatic potential (ESP) map calculations were conducted using Multiwfn 3.8 software with an isosurface defined as 0.001 e/bohr³ [84,85,86]. The isosurfaces were rendered using the VMD—Visual Molecular Dynamics 1.9.3 software [87]. For [Nb₁₀O₂₈]⁶⁻, the wave function was calculated using the relativistic Hamiltonian Douglas–Kroll–Hess (DKH), with the basis SARC-DKH-TZVP for niobium and DKH-def2-TZVP with the auxiliary set SARC/J for the oxygens [88].

3.4. Inhibition of Ca²⁺-ATPase Activity by V₁₅

Isolated sarcoplasmic reticulum (SR) vesicles prepared from rabbit skeletal muscles as described elsewhere [46] were suspended in 0.1 M KCl, 10 mM HEPES (pH 7.0), diluted 1:1 with 2.0 M sucrose, and frozen in liquid nitrogen prior to storage at −80 °C.

The activity of Ca²⁺-ATPase and its inhibition of the POV solution were measured spectrophotometrically by monitoring the change in absorbance at 340 nm at room temperature, both in the absence (100% activity) and in the presence of several concentrations of V₁₅ (2.0, 5.0, 10, 15, 20, and 25 μM) for determination of IC₅₀ (Figure 2a), for determination of the type of inhibition (Figure 2b) the activity was measured without inhibitor and with inhibitor (V₁₅ at 15 μM) using increasing ATP concentrations (0.10 mM, 0.25 mM, 0.50 mM, 1.0 mM, 2.5 mM, and 3.75 mM).

The measurements were performed using the coupled enzyme pyruvate kinase/lactate dehydrogenase assay, as described elsewhere [46]. Aqueous stock solutions of V₁₅ were prepared prior to use at 1.0 mM in water. The reaction medium contained the following: 25 mM HEPES (pH 7.0), 100 mM KCl, 5 mM MgCl₂, and 50 μM CaCl₂. For the coupled enzyme assay, the following components were added: 0.42 mM phosphoenolpyruvate, 18 IU lactate dehydrogenase, 7.5 IU pyruvate kinase, 0.25 mM NADH, and 2.5 mM ATP. The POV solution was added to the assay medium, and the experiment was initiated immediately after the addition of SR vesicles containing the Ca²⁺-ATPase (10 μg mL⁻¹), and the absorbance was followed for about 1 min to measure the basal activity. Subsequently, 4% (w/w) calcium ionophore A23187 was added to the cuvette, and the decrease in absorbance was recorded for 3 min (uncoupled ATPase activity). ATPase activity and inhibition were calculated based on the rate of absorbance decrease per minute in the absence and presence of V₁₅. Data are presented as means standard deviation (± SD). The results represent the average of triplicate experiments. The inhibitory effect of V₁₅ was expressed as the IC₅₀ value, defined as the concentration of V₁₅ that induces 50% inhibition of Ca²⁺-ATPase activity.

3.5. Cell Culture

The human breast cancer cell lines, MDA-MB-231 and MCF-7, were provided by the ATCC (American Type Culture Collection) from Manassas, VA, USA). The cells were cultivated in RPMI 1640 medium supplemented with 10% FBS at 5% CO₂ and 37 °C in a humidified atmosphere.

3.6. MTT Assay

The MDA-MB-231 and MCF-7 cells were cultured in RPMI medium containing approximately 5 × 10³ cells/well in a 96-well plate. The V₁₅ was solubilized in water and included after 2 h of cellular adhesion. After incubating for 24 h, the in vitro cytotoxicity was measured by the MTT assay [89], and the results were analyzed in a MULTISKAN SkyHigh multiplate spectrophotometer (Vantaa, Finland) at a wavelength of 540 nm. IC₅₀ values were calculated using GraphPad Prism 7 Software for Windows. Cell viability values were expressed as arithmetic means, with standard deviation calculated considering three independent experiments. Differences between multiple groups and controls were detected by one-way ANOVA followed by Tukey’s multiple comparisons test. A value of p < 0.05 was considered statistically significant [90].

3.7. Cell Migration by Wound Healing Assay

The MDA-MB-231 cell line was cultured in 6-well plates with the addition of 6 × 10⁵ cells per well. After cells confluence, FBS was removed for 4 h, and using a tip, a straight opening through the cell monolayer was made [91]. V₁₅ (4.30 and 8.60 μM) was added, and the cultures were incubated with RPMI medium containing 1% FBS. Each well was washed 3 times with 1.0 mL of PBS 1x to remove the dead cells. The area of the opening in the cultures was photographed using an inverted microscope at 0, 24 h, and the wound area was measured using ImageJ^® 1.53m software to analyze cell migration.

3.8. Scanning Electron Microscopy (SEM) Images

The preparation of cell line samples for SEM was carried out following Gonçalves’ protocol [92]. Briefly, circular coverslips measuring 0.130 mm in diameter were placed in a 24-well plate, and 30 × 10³ cells per well were cultivated. After 2 h of cell adhesion, the control (same water volume) and V₁₅ (4.30 and 8.60 μM) were added. After 24 h, the cells were washed with 0.100 M sodium cacodylate and fixed with 2.5% glutaraldehyde in 0.100 M sodium cacodylate buffer at pH 7.20 and kept for 1 h at room temperature. Then, the coverslips were washed twice with 0.100 M sodium cacodylate with a 10 min interval between each wash and fixed with 1% osmium tetroxide in 0.100 M sodium cacodylate buffer at pH 7.2 for 30 min in the absence of light. They were washed again twice with 0.100 M sodium cacodylate with a 10 min interval between washes. In the dehydration step, the cells were washed with ethanol at different concentrations (from 50% up to 100%) with a 10-min interval between washes. After the critical point was performed, the samples were metallized with gold, and SEM images were obtained with the Electronic Scanning Microscope TESCAN ORSAY HOLDING MIRA3 (Brno, Czech Republic).

3.9. Gene Expression Quantification

The MDA-MB-231 cells (8 × 10⁵) were plated in P100 culture plates in quintuplicate and treated for 48 h with V₁₅ at the obtained IC₅₀ concentration. The total RNA was extracted using the AllPrep DNA/RNA Mini Kit (Qiagen, Hilden, Germany), and Reverse transcription was performed with SuperScript IV Reverse Transcription Kits (Invitrogen, Waltham, MA, USA) according to the manufacturer’s instructions. Quantitative PCR experiments were performed using SYBR Green Power Master Mix, following the manufacturer’s protocols in the Step One Plus equipment (Applied Biosystems, Foster City, CA, USA). The reference transcript used for gene normalization was GAPDH. In terms of efficiency, we used the pattern curve for each primer, and the acceptable value was more than 98% efficiency. The obtained results were analyzed using the_Ct method and further processed using the 2^−ΔΔCt [93]. The primer sequences are listed in Table S1.

4. Conclusions

In the present study, a mixed-valence polyoxovanadate, [ClV^V₇V^IV₈O₃₆] (V₁₅), was studied against the MDA-MB-231 cancer cell line and analyzed its Ca²⁺-ATPase inhibition potential. Moreover, the electrostatic potential map of V₁₅, V₁₀, PV_{14 ox}, MnV_13, and Nb₁₀ was correlated with the Ca²⁺-ATPase activity inhibition. We believed that this kind of interdisciplinary collaborations, although not always easy to be achieved, are needed and represents a key cog in the “clock of the knowledge”, dynamically joining several scientific wheels such as chemistry, biology, and medicine, among others, for achieving perfect shared innovation and development [94], particularly for the understanding of the effects of POMs in biological systems and their putative applications in medicine.

Thus, we first analyzed the V₁₅ inhibitory potential for Ca²⁺-ATPase activity using a coupled enzymatic method. It was determined that V₁₅ presented an IC₅₀ value of 14.2 μM and a mixed type of inhibition (Figure 10a). By affecting a key protein involved in cellular Ca²⁺ homeostasis, V₁₅ might interfere with several physiological and pathological processes. ⁵¹V-NMR and UV/Vis studies of V₁₅ indicated its stability in the enzymatic and RPMI media used. In fact, V₁₅ also exhibited cytotoxicity against MDA-MB-231 (IC₅₀ = 17.2 μM) and MCF-7 (IC₅₀ = 15.1 μM) breast cancer cells, determined after 24 h (Figure 10b). Furthermore, using V₁₅ concentrations equivalent to half and 1/4 of the IC₅₀, MDA-MB321 cell migration was prevented by 70% and 90%, after 24 h, respectively (Figure 10c). Also, the fusiform morphology characteristic of the MDA-MB-231 cell line was lost to an epithelial-like (amoeboid) morphology upon exposure to V₁₅, which is a potential advantage to control the tumor growth and progression (Figure 10d). Finally, it was observed that V₁₅ induces an increase in the expression of the genes RIPK1, MLKL, and RIPK3, up to 3, 10, and 15-fold, respectively, pointing out that the mechanisms of cell death caused by V₁₅ in the triple-negative breast cancer cell line may occur by necroptosis (Figure 10e). Putting it all together, V_15, which targets Ca²⁺-ATPase with high affinity, has shown promising anticancer activities against breast cancer cells, and is worthy of being explored for therapeutic applications as well as for other areas of research.