All‐Inorganic Polyoxometalates Act as Superchaotropic Membrane Carriers

Polyoxometalates (POMs) are known antitumoral, antibacterial, antiviral, and anticancer agents and considered as next‐generation metallodrugs. Herein, a new biological functionality in neutral physiological media, where selected mixed‐metal POMs are sufficiently stable and able to affect membrane transport of impermeable, hydrophilic, and cationic peptides (heptaarginine, heptalysine, protamine, and polyarginine) is reported. The uptake is observed in both, model membranes as well as cells, and attributed to the superchaotropic properties of the polyoxoanions. In view of the structural diversity of POMs these findings pave the way toward their biomedical application in drug delivery or for cell‐biological uptake studies with biological effector molecules or staining agents.


Introduction
[6] Subsequently, these hydration effects have been generalized, the notion that the effects are a consequence of property changes of the bulk liquid has been abandoned, and, instead, direct interactions between chaotropic ions added later.][16][17][18] Since the utilization of boron clusters of the dodecaborate (B 12 X 12 2-) and metallacarborane (COSAN) type as membrane carriers has been introduced recently, [19][20] and their activity could be traced back to their superchaotropic nature, [8,14,21,22] this new application potential needs to be scrutinized for POMs, which offer even larger chemical diversity than boron clusters.
POMs are a large group of discrete polynuclear metal-oxoanions [10,23,24] that are typically recognized as being highly hydrophilic and water soluble.The rapidly growing number of biological POM applications [25][26][27][28][29][30][31] entails the need for a deeper understanding and analysis of the mechanism of their action.For example, their unspecific interactions with proteins [32,33] and lipids [34][35][36] have recently been interpreted in terms of the chaotropic effect and their superchaotropic character. [15]On the other hand, the omnipresent tendency for speciation and degradation of POMs in aqueous solution, frequently pronounced near neutral pH, [11,37,38] hinder their biological and in particular clinical use, [29] which until today is largely focused on antimicrobial or toxic effects, including anticancer, [28,39] antiviral, [40][41][42] antibacterial, [27,43] and enzyme inhibitory [33,[44][45][46] activity.Herein we report the development of POMs which are sufficiently stable near neutral pH to act as membrane carriers.This constitutes an entirely new biological functionality of POMs, with consequences also for the interpretation of their previously documented effects.
The capability of POMs to act as membrane carriers was explored first in large unilamellar vesicles (LUVs) loaded with carboxyfluorescein (CF). [54,55]We selected three representative cationic peptides, heptaarginine (heptaArg), heptalysine (hep-taLys), protamine, and polyarginine (polyArg), as test cargos.Inside the LUVs, the CF dye was encapsulated at self-quenching concentrations.With a suitable carrier, heptaArg can be taken up into the vesicles, bind with entrapped CF, and shuttle back to the exterior, diluting the previously quenched CF, and thereby restoring its fluorescence (Figure 1c).In the time-resolved fluorescence experiments (Figure 2a), to test whether a particular POM acts as molecular carrier, the CF emission was continuously monitored during the addition of the inorganic carrier (t = 60 s) and after injection of the impermeable selected cargo (t = 120 s).Toward the end of the experiment (t = 600 s), the surfactant Triton X-100 was added to release all the entrapped CF and allow normalization of the intensity data.
It should be noted that the addition of POMs alone (without cargo) retained the stability of the liposomes in the studied concentration range (up to 100 μm), that is, neither an increase in CF fluorescence nor changes in size distribution as obtained by dynamic light scattering experiments (DLS, see Figure S4, Supporting Information) were observed, which would have signaled membrane rupture.These negative controls allowed us to proceed to the carrier experiments, with added cargo.Among the eight inorganic clusters evaluated, we found two hits: The two mixed-metal Keggin POMs [PVW 11 ] 4-and [SiMoW 11 ] 4- triggered the desired increase in fluorescence characteristic for cargo transport.Interestingly, the planar Anderson [AlMo 6 ] 3-also showed incipient transport activity, but only for polyarginine as cargo, while other Anderson-type POM derivatives were inactive (Figure S9, Supporting Information).
Within the Keggin series, the membrane carrier potential appears to be related to the chaotropicity of the clusters, [15,16,18,56,57] which decreases with increasing net charge and charge density,  i) Activation efficiency E = Y max (pEC 50 /f), were pEC 50 is the negative logarithm of EC 50 , and f a scaling factor set to 20.6 to allocate E between 0 and 10; [ 19,55] 10% error (SD, calculated by considering error propagation with respect Y max and EC 50 ); j) From ref. [11]; k) No detectable activity at pH 5.5 and 7.4; l) Membrane disruption observed at strongly acidic pH; m) Carrier activity observed at strongly acidic pH; n) Measured by HPTS/DPX assay. [ 19,76]mely 8-, such that aged solutions showed neither an interaction with the membrane nor cargo transport (Figures S1-S3, Supporting Information). [11]In contrast, the two Al-based clusters [AlW 12 ] 5- and [AlGeW 11 ] 5-are stable at neutral pH (Figures S20-S31, Supporting Information), but their higher charge density appears to adversely affect their transport property.The sandwich cluster [(PZrW 11 ) 2 ] 8-was also evaluated (Figure S34, Supporting Information) but neither transport activity nor membrane disruption was observed.
To characterize the transport capability of the new inorganic carriers, the resulting normalized fluorescence traces were plotted versus the POM concentration, generating dose-response curves (Figure 2b; and Figures S1-S3, Supporting Information), which could be analyzed by Hill analysis (see experimental methods, Equation ( 2)).The resulting membrane transport parameters are the maximal activity (Y max ), the POM concentration that is required to achieve 50% of maximal activity (EC 50 ), and the activator efficiency [55] (E), which are compiled in Table  1).Interestingly, the transport in the vesicles does not reach the same high final Y max levels, which results in a lowering of the E values, but the onset for transport activity occurs at much lower carrier concentrations, around 1 μM (Figure 2b).
Accordingly, the required carrier concentrations are desirably low, e.g., the EC 50 value of [SiMoW 11 ] 4-is about a factor of 3 lower (16.5 vs 48 μm) than that of boron clusters, albeit higher than that of metallacarboranes. [20]The efficacy of the clusters did not markedly decrease when the sequence of addition was inverted (Figure S6, Supporting Information).This reflects the reversibility of the involved supramolecular interactions, which is not observed, for example, for PyBu, where (irreversible) clustercargo aggregation is known to interfere and significantly reduces its activity (Figure S6, Supporting Information).Another contrast [54,[58][59][60] to the amphiphilic PyBu is that the carrier activity of the POM carriers was found to be transferable to the less basic heptalysine (Figure S7, Supporting Information); in fact, the activity characteristics for oligoarginine and oligolysine transport were very similar (Table S4, Supporting Information).Note also that the carrier activity of [PVW 11 ] 4-and [SiMoW 11 ] 4- in the model membranes was comparable at pH 5.5 and pH 7.4 for all cargos listed in Table 1 (Table S4, Supporting Information), which showed that the minor hydrolysis observed at high concentration on long time scales at neutral pH (≈20% for [PVW 11 ] 4-and 45% for [SiMoW 11 ] 4-after 24 h) did not affect its carrier activity significantly because the transport experiments were carried out on a much shorter time scale (10 min).It is also important to note that POMs may become stabilized in the presence of lipids [61,62] as well as organic matter [63] (e.g., peptides), such that the active species may actually be more stable under the actual transport conditions than in the neat buffers used for the long-term stability studies by NMR and ESI-MS.To affect membrane transport, any carrier needs to display an affinity toward the cargo as well as toward the lipid bilayer.Accordingly, the interaction between [PVW 11 ] 4-and [SiMoW 11 ] 4- and the peptides was analyzed by isothermal titration calorimetry (ITC, Figure 2c,d; and Figures S10 and S11, Supporting In-formation).These experiments showed strong supramolecular interactions (on the order of 10 7 m −1 ) with binding governed by the chaotropic effect [8,18,56,64] (enthalpically driven, interaction with the peptide backbones, Table S6, Supporting Information) and Coulombic interactions (interaction with arginine and lysine side chains).The strong but reversible intermolecular interactions between the low-charged POMs and the cationic oligopeptides present an interesting finding, which was measured for both, oligo-and polypeptides.[67][68][69][70][71][72][73] Since no nanoaggregates were observed by DLS measurements (see the Experimental Section), the combined data suggest the formation of discrete POM-peptide complexes of varying stoichiometries.
U-tube transport experiments (see the Experimental Section, Figure 2e,f) confirmed that [PVW 11 ] 4-and [SiMoW 11 ] 4-act as "real" carriers that can seize and transport a hydrophilic cargo even through a hydrophobic bulk phase (such as CHCl 3 ).In the U-tube experiments, the trans phase contained only buffer, and the cis phase is loaded with combinations of CF, carrier (POM), and cargo (heptaArg).Aliquots of the trans phase are taken and evaluated by fluorescence at different times.The trans phase is initially (t = 0) nonfluorescent, but the transport of CF from the cis phase through the chloroform barrier leads to fluorescence, signaling successful transport.Figure 2f shows that there is no transport in the absence of POM, inefficient transport in the presence of POM alone, and highly efficient transport with both, POM and peptide.This is consistent with translocation of a POM•heptaArg•CF complex through the organic phase, in support of the mechanism in Figure 1c.Accordingly, the superchaotropic properties of POMs cannot only be used for biological membrane transport, but they are transferable to artificial membranes and phase-transfer processes in general, as recently independently demonstrated for catalysis and dissolution. [13,74,75]o better mimic biological membranes, which contain negatively charged components and cholesterol, equivalent transport experiments were performed in anionic liposomes [19,20] with the prototypal POM carriers, [PVW 11 ] 4-and [SiMoW 11 ] 4-.These experiments showed that, despite the apparent Coulombic repulsion, both POMs retained their carrier capabilities (Figure S8 and Table S5, Supporting Information), with only slightly reduced net efficiencies.This encouraged us to transfer the experiments to CHO-K1 (Chinese hamster ovary) cells, after we had confirmed sufficient stability of the most promising POM carriers, [PVW 11 ] 4-and [SiMoW 11 ] 4-, in the cellular nutrient mixture (F-12 Ham's medium, Figures S26-S28C, Supporting Information).
POMs are known to cause cellular damage, and their antimicrobial as well as anticancer effects rest on those. [28]Nevertheless, we tried to define a "therapeutic window" in which the cytotoxic properties would be sufficiently low to allow their carrier activity to be investigated.Indeed, with up to 10 μm of [PVW 11 ] 4- or [SiMoW 11 ] 4-, CHO-K1 cells retained about 80% viability after 24 h, according to resazurin assay (Figure S13, Supporting Information); the onset of carrier activity fell in the same range (Figure 2b; and Figure S8, Supporting Information).Accordingly, the uptake of 5( 6)-fluorescein isothiocyanate-labelled octaarginine (FITC-Arg 8 ) by these POMs was studied in live CHO-K1 cells at low carrier concentration (5 μm); the alternative amphiphilic and superchaotropic carriers, PyBu and B 12 Br 12 2-were used as references (Figure 3).Without carrier, the cells showed minor punctate fluorescence, pointing to (undesirable) endoso-mal uptake and entrapment of FITC-Arg 8 . [77,78]In the presence of the active stable POMs, however, diffuse fluorescence was observed in both, cytosol and nucleus (Figure 3).Important to note, PyBu and B 12 Br 12 2-did not achieve efficient uptake at such low carrier concentrations, which is fully consistent with their dose-response curves obtained from the liposomal model experiments (Figure 2b).Comparable results were obtained in fixed cells (Figure S14-S16, Supporting Information).Colocalization experiments with LysoTracker confirmed cytosolic transport as the major pathway, with only partial endosomal colocalization of FITC-Arg 8 (Figure S18, Supporting Information).Accordingly, the two active POMs serve as highly effective intracellular membrane carriers.Interestingly, solutions of the parent Keggin POM, [PW 12 ] 3-, also afforded some apparent cellular uptake, but the cell morphology was highly compromised (Figure S17, Supporting Information), owing to its membrane-lytic activity (Figure 1b; and Figure S5, Supporting Information).

Conclusions
Inspired by recent observations of superchaotropic behavior of large cluster ions, we have identified the first two molecular membrane carriers of the all-inorganic Keggin POM anion type that are functional in both model membranes as well as in live cells and that are stable under physiological conditions.In vesicle and cell experiments, the carriers showed transport activity at concentrations as low as 5 μM that outperform established molecular carriers.They are capable of transporting both arginine-rich and lysin-rich oligo-and polypeptides.Our experimental findings encourage the optimization of POM carriers, with the aim to further improve their cellular toxicity while maintaining hydrolytic stability.Indeed, we also found an incipient carrier activity for an Anderson-type POM, which suggests that globular shape presents no rigorous requirement for cluster ions to display this biological functionality.Further chaotropicity and membrane carrier activity modulations could be achieved by incorporation of other hetero-groups (such as SO 4 2− ) or by exploring different archetypes such as POMs of the Wells-Dawson or Lindqvisttype, which is expected to open additional research directions.coefficient of tryptophan at 280 nm ( = 5540 cm −1 m −1 ). [79]FITC-Arg 8 was from GL Biochem (Shangai, China) Ltd, and its concentration was determined by using the extinction coefficient of 5-FITC at 491 nm ( = 73 000 cm −1 m −1 ). [80]CHO-K1 cells were obtained from Sigma-Aldrich (Steinheim, Germany).
Vesicle Preparation-Zwitterionic Vesicles in Neutral Media: A thin lipid film was prepared by evaporating a lipid solution of EYPC (egg yolk phosphatidylcholine, 25 mg) in CHCl 3 (1 mL) with a stream of nitrogen and then dried in vacuo overnight.The dry film was rehydrated with 1 mL buffer (50 mm CF, 10 mm Hepes, pH 7.4 or 50 mm CF, 10 mm Tris, pH 7.4) for 30 min at room temperature, subjected to freeze-thaw cycles (7 times) and extrusions (15 times) through a polycarbonate membrane (pore size 100 nm).Extravesicular components were removed by size exclusion chromatography (NAP-25 column Sephadex G-25 DNA grade) with 10 mm Hepes, 107 mm NaCl, pH 7.4, or 10 mm Hepes, 107 mm NaCl, pH 7.4.Phospholipid concentration was calculated by the Stewart assay. [81]witterionic Vesicles in Acidic Media: Acidic vesicles were prepared analogously but the rehydration and elution buffer were 50 mm CF, 10 mm Mes, pH 5. Transport Experiments: Vesicles stock solutions (5-10 μL) were diluted with the corresponding buffer in a disposable plastic cuvette and gently stirred (total volume 2000 μL, final lipid concentration 13 μm).Carboxyfluorescein fluorescence was monitored at  em = 517 nm ( ex = 492 nm) as a function of time after addition of POMs at 60 s, analyte at 120 s and Triton X-100 (24 μL 1.2% wt/vol) at 600 s to lyse the vesicles, for calibration.Fluorescence intensities were normalized to fractional emission as where I 0 = I t before POM addition and I 0 = I ∞ after lysis.For Hill analysis, I t before lysis was defined as transport activity, Y, and plotted against POM concentration, c, and fitted to the Hill Equation ( 2), to afford the activity in absence of POM, Y 0 , the maximal activity, Y max , the concentration needed to achieve 50% of maximal activity, EC 50 , and the Hill coefficient, n ) n (2) Activator Efficiency: In the transport measurements, where different carriers (POMs) are tested with the same cargo, the activator efficiency (E) is determined by their capacity to transport the impermeable cargo to the intravesicular region and is characterized by Y max , its maximal activity, and EC 50 , the effective activator concentration.An ideal carrier reaches high Y max at low EC 50 .To count both factors, the activator efficiency is defined as were pEC 50 is the negative logarithm of the EC 50 , and f a scaling factor, which was set to 20.6 to calibrate the highest (previously known) activation efficiency, E, as 10. [55] DLS: DLS experiments of the vesicles were carried out on a Malvern Instruments DTS Nano 2000 Zeta-Sizer.A POM concentration of 100 μm was chosen as default, since this is the highest concentration at which membrane transport was evaluated.The cargo concentration was set to be same as in the transport experiments, namely 10 μm for heptaarginine, 1 μm for protamine, and 0.1 μm for polyarginine.Note that the POMs themselves, with a size of ≈1 nm, fall below the detection limit (>10 nm) of the DLS instrument.Note that DLS measurements of the POMs in the presence of different cargos (at the particular required concentrations) did not afford any detectable signals, which ruled out the formation of nanoaggregates.
ITC: All experiments were performed with a VP-ITC MicroCalorimeter from MicroCal, Int,. at atmospheric pressure and 298.15 K. Solutions were degassed and thermostatic prior to the titration experiments in a Ther-moVac accessory.A constant volume of POM ([PVW 11 ] 4-or [SiMoW 11 ] 4-) (10 μL per injection) was injected into the oligopeptide solution (heptaArg, protamine, polyArg or heptaLys) in 10 mm Tris, pH 7.4 to determine the apparent binding affinity of [PVW 11 ] 4-and [SiMoW 11 ] 4-with the oligopeptides.Dilution heats were determined by titration of the POM into buffer and subtracted from the reaction heat.The neat reaction heat was fitted with Origin 7.0 software by using the "one-set-of-sites" model to obtain the complex stability constant (K a ) and molar reaction enthalpy (∆H°).The free energy (∆G°) and entropy changes (∆S°) were obtained according to the relation: ∆G°= −RTlnK a = ∆H°−T∆S°.
U-tube Transport Experiments: The U-tubes were house-made, similar to those of Rebek [82] and Matile, [54] consisting in a small beaker with a central glass barrier, separating the two aqueous layers, cis (sampling phase) and trans (receiving phase), but allowing an interface chloroform layer below the cis and trans phases.3 mL CHCl 3 are placed into the U-tube and 1 mL of the cis (combinations of carrier (POM), cargo (heptaArg) and CF in buffer) and trans (1 mL buffer) phases were added.The organic phase was stirred at 700 rpm at room temperature.Aliquots (25 μL) from the trans phase were taken at different times, 400 μL of buffer were added and measured by fluorescence.
Cell Culture and Confocal Imaging-Fixed Cells: For confocal microscopy experiments, CHO-K1 cells were seeded into 12-well plates (Eppendorf) at a density of 1 × 10 5 cells per well (per 1 mL) and incubated for 24 h in Ham's F-12 medium containing 10% v/v fetal bovine serum and 1.0% penicillin streptomycin at 37 °C in a 5% CO 2 atmosphere.Cells were washed twice with PBS, incubated with FITC-Arg 8 (4 μm) and carrier (5, 10, and 50 μm) for 1 h at 37 °C, washed twice with Hank´s Balanced Salt solution containing 100 μg mL −1 heparin, twice again with PBS, and ultimately fixed with 1 mL cold 4% paraformaldehyde for 20 min.Cells were rewashed twice with PBS and subsequently the nuclei were stained with 2 μg mL −1 DAPI for 10 min at ambient temperature before imaging.Cells were imaged by an LSM 980 Airscan2 confocal laser scanning microscope (Zeiss, Germany) and images were processed with the instrument-specific software ZEN blue 3.7.
Cell Culture and Confocal Imaging-Living Cells: CHO-K1 cells were plated into μ-Slide 8-well plates (ibidi) at a density of ≈1 × 10 5 cells per well (per 1 mL) and incubated for 24 h in Ham's F-12 medium containing 10% v/v fetal bovine serum and 1.0% penicillin streptomycin at 37 °C in a 5% CO 2 atmosphere.The cells were washed twice with Hank's buffer, incubated for 5 min at 37 °C in Ham's F-12 medium without phenol red with carrier (5 μm), and subsequently in Ham's F-12 medium without phenol red with 5 μm FITC-Arg 8 for 30 min at 37 °C.Cells in the absence of carrier were used as control.The cells were washed twice with Hank's buffer containing 100 μg mL −1 heparin and then washed twice with Ham's F-12 medium without phenol red.The nuclear staining was performed by incubating the cells with 10 μg mL −1 Hoechst 33 342 (bis-benzimide H33342 trihydrochloride) for 10 min at 37 °C.Cells were directly imaged as stated above.
For colocalization studies live cells were incubated analogously.After medium removal, cells were washed twice with Hank's buffer containing 100 μg mL −1 heparin and twice with Ham's F-12 medium without phenol red.Cellular compartment staining was performed with 10 μg mL −1 Hoechst 33 342 (bis-benzimide H33342 trihydrochloride) and 1 μm Lyso-Tracker Red DND-99; after 5 min at 37 °C, cells were washed twice with Ham's F-12 medium without phenol red, and imaged within 5−10 min as described above.
Cellular Viability Assay: For cell viability assay in thepresence of activators, CHO-K1 cells were incubated in Ham's F-12 medium containing 10% fetal bovine serum and 1.0% penicillin streptomycin at 37 °C in a 5% CO 2 for 24 h.CHO-K1 cells were dispersed in a 96-well flat bottom tissue culture treated plates (Eppendorf) at density of 5 × 10 3 cells per well (per 100 μL) and incubated in Ham's F-12 containing 10% v/v fetal bovine serum at 37 °C in a 5% CO 2 incubator for 12 h to allow cell adhesion.Then, cells were incubated with POM, B 12 Br 12 2− , or PyBu (10, 50, and 100 μm) for another 24 h.After this incubation time, 20 μL resazurin solution (0.15 mg mL −1 ) was added to cells and incubated for another 3 h.Finally, the plates were read at  em = 590 nm ( ex = 560 nm) using a JASCO fluorometer FP-8500.The cell viability was evaluated by the following equation Cell viability = FI sample − FI blank where, FI sample is the fluorescence intensity of cells incubated with different compounds, FI blank is the fluorescence intensity of medium, and FI control is the fluorescence intensity of cells in absence of any compound.Statistical Analysis: Kinetic experiments were performed by triplicates.Data were normalized according to Equation (1) and analyzed using Origin 8.0.Cellular viability was calculated according to Equation (4).Confocal images are representative of four biological replicates.ITCs were performed by duplicates and fitted with Origin 7.0.Data are presented as mean ± standard deviation (SD).

Figure 1 .
Figure 1.a) POMs evaluated in this study, including the Keggin (top, 1-7) and Anderson (8, bottom) archetypes.(Color code: {WO 6 } blue; {MoO 6 }, orange; {XO 4 } (X = P V , Si IV , Ge IV , Al III ), yellow; substituting transition ion V V , Mo VI , purple; Zr IV , dark blue; O, red; compounds were used as acid or sodium, potassium, or tetraalkylammonium salts.b) Ordering of the studied POMs according to their chaotropicity (projected based on their charge density), their experimental pH stability, and expected membrane lytic propensity; the window of POMs with practical utility as membrane carriers is shown in yellow.c) Schematic representation of the transport of impermeable oligoarginines facilitated by POMs in model membranes.d) Cellular uptake of the labeled oligoarginine assisted by POMs.

Figure 2 .
Figure 2. a) Changes in CF emission ( ex = 492 nm,  em = 517 nm) in EYPC⊃CF vesicles as a function of time during the successive addition of i) increasing concentrations of [SiMoW 11 ] 4− (0-150 μm), ii) heptaArg (10 μm), and iii) TX100, for calibration.b) Dose-response curves for heptaArg transport for the active POMs and the corresponding Hill curve fits in comparison to the standards.c,d) Microcalorimetric titrations in 10 mm Tris, pH 7.4: thermograms (top) for the sequential injections of c) [PVW 11 ] 4− or d) [SiMoW 11 ] 4− into heptaarginine solution, and (bottom) corresponding reaction heats from the integration of the calorimetric traces.POM/oligoarginine concentrations in μm: (c) 100/8 and d) 200/10.e) Pictures of the U-tube experiment (conducted with [SiMoW 11 ] 4− ) at the opening (t = 0) and the closing (t = 24 h) of the experiment.f) CF fluorescence measured in the trans phase of the U-tube.In the U-tube experiments, CF, POM, and heptaArg were fixed as 100, 50, and 10 μm.

Table 1 .
Stability of POMs in acidic and neutral pH and membrane transport parameters of different cargo molecules at pH 7.4.