812
Biomacromolecules 2008, 9, 812–817
Molecular Interactions between Wells-Dawson Type Polyoxometalates and Human Serum Albumin Guangjin Zhang†, Bineta Keita†, Constantin T. Craescu‡, Simona Miron‡, Pedro de Oliveira†, and Louis Nadjo*,† Laboratoire de Chimie Physique, Equipe d’Electrochimie et Photoélectrochimie, UMR 8000, CNRS, Université Paris-Sud 11, Bâtiment 350, 91405 Orsay Cedex, France, and INSERM, Institut Curie-Recherche, Centre Universitaire Paris-Sud, Bâtiment 112, 91405 Orsay Cedex, France Received October 8, 2007; Revised Manuscript Received December 17, 2007
Binding human serum albumin (HSA) of three polyoxometalates (POMs) with the Wells-Dawson structure, R2-[P2W17O61]10- (abbreviated as R2-P2W17) and two of its metal-substituted derivatives, R2-[NiP2W17O61]8and R2-[CuP2W17O61]8- (R2-P2W17Ni and R2-P2W17Cu, respectively) was studied in an aqueous medium at pH 7.5. Fluorescence quenching, circular dichroism (CD), thermal denaturation, and isothermal titration calorimetry (ITC) were used for this purpose. The results were compared with those obtained previously with the Keggin structure POM, [H2W12O40]6- (H2W12), and the wheel-shaped structure, [NaP5W30O110]14- (P5W30). All these POMs bind HSA mainly by electrostatic interactions. Comparison of the physical characteristics and HSA interaction parameters for the POMs of the present work and those studied previously showed that the overall charge of the clusters is not the single parameter governing the binding process and its consequences. In contrast, besides the influences of the structure, the dimension and/or weight of the POMs, the results have permitted highlighting of the importance of each POM atomic composition for its binding behavior.
Introduction The research on drug–protein interactions is an active field of interest due to the prospective unraveling of drug action mechanisms and the possibility of designing novel medicines. Among the thousands of proteins in a human being, human serum albumin (HSA) stands out because it acts as a drug carrier in the blood system. Also, it is the most abundant protein in the blood plasma (40 mg/mL). The abundance and versatile binding properties of HSA make it one of the best studied models to understand the physicochemical basis of drug–protein interactions. The efficiency of drugs is generally found to be determined by their binding affinity to HSA.1–4 Polyoxometalates (POMs) are early transition metal–oxygen anionic clusters. Among numerous remarkable properties, they have been reported as having promising antibacterial, antiviral (particularly anti-HIV), antitumor, and anticancer activities, which may open the way toward new, innovative, and cheap therapeutic strategies for various human diseases.5–8 In our previous papers,9,10 the interactions between HSA and two kinds of POMs having Keggin structure, [H2W12O40]6- (H2W12), and wheel-shaped structure, [NaP5W30O110]14- (P5W30), were studied. The results showed the selected POMs can effectively bind the protein mainly by electrostatic interactions. It was found that the two different POM structures led to totally different binding behaviors as far as the thermodynamics and the kinetics are concerned.9,10 Because the two POMs were very different in structure, dimensions, charges, and molecular weight, it is difficult to conclude which parameter(s) govern(s) the binding behavior. POMs are inorganic molecules that exhibit a plethora of different structures. Upon changing the structure of a POM, * Corresponding author. E-mail:
[email protected]. † Laboratoire de Chimie Physique, Equipe d’Electrochimie et Photoélectrochimie, UMR 8000, CNRS, Université Paris-Sud 11. ‡ INSERM, Institut Curie-Recherche, Centre Universitaire Paris-Sud.
Figure 1. Structure of R2-Wells-Dawson POM. The green octahedron shows either the location of the vacancy in R2-P2W17 or represents the metal substituent M in R2-MP2W17.
both its dimensions and charges are likely to change. To elucidate which parameters are crucial for the interaction, more detailed experiments with POMs having structures other than the ones used previously are required for comparison. One of the most widespread structures is the so-called Wells-Dawson type. This is a promising research target because of its attractive activities in reported biological experiments. Here, we use the lacunary Wells-Dawson POM R2-[P2W17O61]10- (abbreviated as R2-P2W17) and two of its metal-substituted derivatives, R2[NiP2W17O61]8- and R2-[CuP2W17O61]8- (R2-P2W17Ni and R2P2W17Cu, respectively) (Figure 1), which are stable in physiological conditions, to study the interaction with HSA.
10.1021/bm701120j CCC: $40.75 2008 American Chemical Society Published on Web 02/12/2008
Interactions between Wells-Dawson POMs and HSA
Biomacromolecules, Vol. 9, No. 3, 2008 813
Comparison of the present results with the ones obtained previously may shed light on the factors governing the interactions between POMs and HSA.
Experimental Section Materials. Fatty acid free human serum albumin was purchased from Sigma. The samples were dissolved in a buffer solution of pH 7.5 (0.05 M Tris, NaCl 0.15 M). The concentration of the protein was determined spectrophotometrically using an extinction coefficient of (280 ) 36600 M-1 cm-1).10 All the chemicals were of high-purity grade and were used as purchased without further purification. Ultrapure water with a resistivity of 18.2 MΩ cm was produced by passing through a RiOs 8 unit followed by a Millipore-Q Academic purification set. The three polyoxometalates, R2P2W17, R2P2W17Ni, and R2P2W17Cu, were synthesized by the methods in the literature.11 The stability of these POMs in the pH 7.5 medium used in the following was assessed by UV–visible spectroscopy for at least 6 h. Fluorescence Methods. All fluorescence spectra were recorded on a SPEX-Spectrofluorimeter (Jobin-Yvon-Horiba) equipped with a 250 W xenon lamp and a thermostatted bath. Quartz cuvettes with 1.0 cm optical path were used. The excitation wavelength was selected at 295 nm to avoid the excitation of tyrosine residues. The UV–visible spectra were recorded at room temperature on a Perkin-Elmer Lamda 19 spectrophotometer. All the measurements were performed at room temperature (20 °C) in aqueous solution unless otherwise stated. Typically, 2 mL of a solution containing an appropriate concentration of HSA was titrated by successive additions of aliquots of polyoxometalate solutions. Titrations were performed manually by using syringes. At each addition, the fluorescence spectrum was collected. Raw data were systematically corrected for inner filter effects.9 Circular Dichroism (CD) Spectroscopy. CD experiments were performed on a Jasco 715 spectropolarimeter equipped with a Peltier temperature control unit. Far-UV spectra were recorded between 200 and 260 nm at 20 °C using 1 mm quartz cells. Samples were dissolved in 10 mM Tris buffer (pH ) 7.5). Temperature denaturation curves were recorded between 5 and 95 °C with a temperature increase rate of 1 °C/min. Isothermal Titration Calorimetry (ITC). ITC experiments were performed using a MicroCal MCS-ITC instrument (MicroCal Inc., Northampton, MA). Protein and polyoxometalate solutions were properly degassed prior to the titrations to avoid the formation of bubbles in the calorimeter cell. In a standard experiment, the protein (10–30 µM) in the 1.337 mL calorimeter cell was titrated by the polyoxometalate solution (generally 10–30 times more concentrated) by up to 37 successive automatic injections of 8 µL each. The first injection of 2 µL was ignored in the final data analysis. Integration of peaks corresponding to each injection and correction for the baseline were carried out using Origin-based software provided by the manufacturer. Fitting the data to various interaction models results in the stoichimetry (n), equilibrium binding constant (Ka), and enthalpy of complex formation (∆H). The other thermodynamic parameters were calculated according to the formulas
∆G ) ∆H - T∆S
(1)
∆G ) - RT lnKa
(2)
and
where T is the absolute temperature (in the current experiment T ) 303 K), and R ) 8.3151 J mol-1 K-1.
Results and Discussion Fluorescence Quenching Experiments. First, we use fluorescence quenching to study the interaction of the three POMs with HSA. In our previous study, we have reported that the fluorescence of HSA, which is due to the single tryptophan
Figure 2. Stern-Volmer plots of the quenching of HSA fluorescence by three Wells-Dawson structure POMs. The data were corrected for inner filter absorption effects of the POMs.
residue in the protein, Trp214, was efficiently quenched in the presence of a Keggin (H2W12) and a wheel-shaped (P5W30) POM.9 Here we found that Dawson structure POMs also can quench the protein fluorescence. As shown in Figure 2, the Stern-Volmer plots for these three POMs exhibit good linear fits with different slopes. The quenching constants (Kq) can thus be calculated from the slopes of these fits. They were found to be 4.1 × 104 mol-1 L at 293 K for R2P2W17, 5.92 × 104 mol-1 L for R2P2W17Cu, and 1.15 × 105 mol-1 L for R2P2W17Ni. Thus it can be concluded that the metal-substituted POMs quench the protein fluorescence more efficiently than the lacunary form. Also, the metal used has an influence on such quenching. As we reported before, the quenching of the HSA fluorescence by binding to POMs is mainly static and concomitant with energy transfer. Thus the quenching constant is the outcome of the overlap between the absorption spectrum of the POMs and the emission spectrum of HSA. The quenching behavior depends on the absorption coefficient of the POM in the range 200–300 nm. The higher the absorption coefficient, the larger the quenching constant. Metal substitution can obviously enhance the absorption of the POM in the UV range, resulting in an enhanced quenching of metal-substituted POMs. Different metals have a distinct influence on the absorption, in our case, Ni being more efficient than Cu. When compared with the previously reported POMs, we find the quenching constants of the Wells-Dawson POMs to be larger than that of the Keggin POM, but smaller than that of the wheel-shaped POM.9 Provisionally, it must be reminded that substituted POMs are known to undergo reversible dissociation/ association processes in water solution.12 Such behavior might also impact their binding interactions with HSA, as will be discussed in the following. Circular Dichroism Spectroscopy. The interaction between the three Wells-Dawson POMs and HSA were further studied by CD spectroscopy in order to determine the effect of binding to POMs on the protein secondary structure. As appears in Figure 3, all three POMs exhibit the same trend when binding the protein. Upon increasing the concentration of the POM, the CD signal of HSA decreased, indicating a partial unfolding of the helical structure of the protein after binding to POMs. However, there are still differences between the three POMs. For R2P2W17, the unfolding of the protein reaches a plateau when the molar ratio γ (γ ) [POM]/[HSA]) is 2. Further increase of the concentration did not alter the secondary structure
814 Biomacromolecules, Vol. 9, No. 3, 2008
Zhang et al.
Figure 4. Thermal denaturation of HSA monitored by the CD signal at 222 nm in the absence and in the presence of Wells-Dawson polyoxometalates.
Figure 3. CD spectra of HSA in the absence and in the presence of different concentrations of POMs (in 10 mM Tris/HCl buffer, pH 7.5, 20 °C): (a) R2P2W17 + HSA, (b) R2P2W17Cu + HSA, (c) R2P2W17Ni + HSA.
of the protein. Only 10% of the protein unfolded at saturated binding conditions with R2P2W17. As far as R2P2W17Cu is concerned, protein unfolding stopped when γ ) 2, corresponding to a 13% loss of the helical structure. It is worth noting that protein unfolding is hardly different for R2P2W17Cu and for R2P2W17 with the same γ value. In the case of R2P2W17Ni, the process had not stopped even at γ ) 3, at which almost 25% of the protein secondary structure unfolded. The results show that the Ni-substituted POM has the strongest effect on the protein secondary structure, followed by the Cu-substituted POM and at almost equality the lacunary form. Thermal Denaturation. Thermal denaturation experiments also give some information on the differences between the three POMs regarding their interactions with the protein (Figure 4). In the absence of any ligand, 10 µM HSA shows a cooperative thermal unfolding in the range 45-85 °C with a midpoint temperature at about 64 °C. Binding of Wells-Dawson POMs has no significant destabilizing effect on the global secondary
structure. The midtransition temperatures of HSA in the presence of Wells-Dawson POMs is about 65 °C, which is similar to that of the protein alone. Actually, there is an obvious decrease of the slope when the protein was bound to POMs, indicating the stabilizing effect of POMs. As previously reported, P5W30 led to a protein unfolding of 30% when γ ) 3, whereas H2W12 had no apparent effect on the secondary structure.9,10 We find that the protein unfolding values triggered by the three present POMs are in the range defined by the two previously reported POMs. Table 1 is a compilation of all the results and corresponding parameters of the POMs studied so far. It can be seen that the interaction between the POM and HSA becomes stronger as we move down the table from the Keggin to the wheel-shaped POM. Strikingly, results in Table 1 indicate that the smaller the charge density, the larger the interaction. If we compare the Kq values and the unfolding percentage taken together with the other parameters on the table, it looks as if the charges are not a key parameter governing the interaction because it follows a trend other than that of Kq and of the unfolding percentage. In contrast, it seems that the structure, the composition, the dimension, and/or the weight of the POMs are important parameters in the interaction process. Isothermal Titration Calorimetry Experiments. To better understand the interaction between HSA and POMs, a complete characterization of the binding energetics was carried out by isothermal titration calorimetry (ITC) in an attempt to correlate the thermodynamic data with the results of other techniques. As appears in Figure 5, the thermodynamics of the binding process between Wells-Dawson structured POMs and HSA is relatively complicated. For two of the three systems, the overall interaction is a mixture of exothermic and endothermic steps and the shapes of the isothermal curves significantly differ. For R2P2W17, most of the signals are positive, which means the main interaction process is endothermic, whereas for R2P2W17Ni and R2P2W17Cu, the signals are mainly negative. As previously reported, the main driving force for the POM-HSA interaction is the exothermic electrostatic component. For R2P2W17Cu, the ITC results showed a simple 1:1 exothermic interaction with Ka ) 7.5 × 105 M-1 and ∆H ) -108.8 kJ/ mol. In an endothermic process, the driving term of the binding energy is the entropy changes associated with the interaction, which should be positive. The entropy change (∆S ) and free energy change (∆G ) were calculated to be -252.4 J/mol · K and -33.6 kJ/mol, respectively. In the case of R2P2W17Ni,
Interactions between Wells-Dawson POMs and HSA
Biomacromolecules, Vol. 9, No. 3, 2008 815
Table 1. Physical Characteristics and HSA Interaction Data Parameters for Several POMs POM H2W12 R2P2W17 R2P2W17Cu R2P2W17Ni P5W30
structure Keggin Wells-Dawson Wells-Dawson Wells-Dawson wheel
dimension (nm)
charges (charge density)
1 1 × 1.5 1 × 1.5 1 × 1.5 2×2
-6 (- 1.9) -10 (- 1.6) -8 (- 1.4) -8 (- 1.4) -14 (- 1.1)
the process is too complicated to be fitted. However, it can still be observed that the endothermic process ends when γ reaches 3. Also, it can be pointed out that the need for at least a fourbinding-site model to fit the data for R2P2W17Ni is in agreement with the results of CD experiments, which indicate a larger HSA unfolding by this POM compared to the two other POMs. Two main parameters might be retained as the most probable origins of the favorable entropic component upon protein binding to the POMs: first, the unfolding of the protein13 and, second, the increase of water entropy upon the polyanion dehydration concomitant with the POM binding to the protein. Furthermore, these two parameters might intervene to reinforce each other. However, a large positive enthalpy is observed for R2P2W17, associated with only 10% unfolding of the protein, whereas R2P2W17Ni unfolds up to 30% of the protein with a much smaller positive enthalpy variation. This observation rules out protein unfolding and privileges the dehydration process as the major parameter at the origin of the endothermic component of ITC curves in the present experiments. Finally, because of the complexity of the binding process, it is difficult to get the respective thermodynamic parameters by direct fitting the experimental data for the R2P2W17 and R2P2W17Ni systems. The results shown in Figure 5 were obtained by using a two-site model for R2P2W17. Competitive binding can shed light on the binding process. As previously reported, there is a preferential binding site for POMs in the protein, which can be specifically occupied by the Keggin-shaped POM, H2W12.10 Thus, further details on the binding of Wells-Dawson structured POMs can be obtained by performing competitive binding experiments with H2W12. As appears in Figure 6, with the protein prebound to H2W12, the subsequent interaction in the presence of a Wells-Dawson structure POM turned out to be relatively simple. For R2P2W17, there is just one endothermic binding site remaining with Ka ) 3.6 × 105 M-1, ∆H ) 22.7 kJ/mol, and a stoichiometry of 1:1. Similarly, for R2P2W17Cu, the sole exothermic binding site remaining has Ka ) 2.57 × 106 M-1, ∆H ) -82.3 kJ/mol, and a stoichiometry of 1:1. The ∆S and ∆G were calculated to be -153.4 J/mol · K and -36.6 kJ/mol, respectively. Even though not spectacular, the prior presence of H2W12 induces a significant change in the thermodynamic binding parameters of R2P2W17Cu to HSA. In the case of R2P2W17Ni, the data can be fitted to a two-binding-site model, one having a higher affinity and a negative enthalpy and the other having a lower affinity and a positive enthalpy. Actually, the results suggest the existence of at least a third binding site. Taken together, the results of the present work suggest the following conclusions: for the lacunary Wells-Dawson POMs, there are two main binding sites on HSA. For the Ni-substituted POM, there are at least three binding sites on the protein. There is only one main binding site on the protein for Cu-substituted POM. Because the three compounds virtually have the same structure, the marked differences in the binding behaviors can be attributed to their compositions, but how the composition affects the binding is unknown at present.
molar weight
Kq
3208 4882 4864 4859 8259
3.5 × 10 4.1 × 104 5.9 × 104 1.1 × 105 2 × 105 4
unfolding (%)
refs
0 10 13 25 30
9, 10 this work this work this work 9, 10
Figure 5. ITC curves corresponding to HSA titration with POMs: (a) solution of 26 µM HSA was titrated with R2P2W17 (300 µM), (b) solution of 18 µM HSA was titrated with R2P2W17Ni (300 µM), (c) solution of 20 µM HSA was titrated with R2P2W17Cu (300 µM).
816 Biomacromolecules, Vol. 9, No. 3, 2008
Zhang et al. Scheme 1. Possible Interaction Pathways for Metal-Substituted Wells-Dawson POMs
which results in rather different binding behaviors in ITC experiments. Indeed, the stability constants published for R2P2W17Ni (log K ) 5.49) and R2P2W17Cu (log K ) 6.74) were determined in molar lithium perchlorate medium, as also was the best stability constant for R2P2W17 in the presence of Li+ (log K ) 3.61).12 Tentatively, these values might support the proposed scheme. As a matter of fact, the metal-substituted Wells-Dawson type POMs, R2-P2W17M show a higher affinity for HSA than their lacunary precursor, R2-P2W17, in complete agreement with their higher stability constants. Tentatively, another possibility could explain the difference in behavior between P2W17Ni and P2W17Cu. This difference might be attributed to subtleties in the coordination geometry of the d-electron-containing metals. Specifically, Cu(II) is Jahn–Teller distorted and it most likely sits on the surface of the polytungstate cluster, not in the vacant site itself. In contrast, Ni(II) is not Jahn–Teller distorted and would most likely sit in the vacant site. Such a subtle difference may induce the big change in the molecular interaction behavior of the two complexes.
Conclusion In summary, the molecular interaction between Wells-Dawson structured POMs and HSA has been studied. It was found that within the same overall structure, the composition of the POMs plays a key role in the binding behavior. The metalsubstituted POMs showed a stronger interaction with the protein than the lacunary one. The binding process results from a mixture of both endothermic and exothermic components. This study was also useful in several respects. Even though the interaction mechanism remains mainly electrostatic in nature, the nominal negative charge of the POMs does not appear as a single parameter governing the binding process and its consequences. In contrast, besides the influences of the structure, the dimension, and/or molar weight of the POMs, the results have permitted highlighting of the importance of each POM atomic composition in its binding behavior to HSA. Because of the complexity of the observations, future studies would address specific systems of biological significance. Figure 6. ITC curves corresponding to HSA titration with POMs: (a) solution of 25 µM HSA previously treated with 25 µM of H2W12 was titrated with R2P2W17 (300 µM), (b) solution of 20 µM HSA previously treated with 20 µM of H2W12 was titrated with R2P2W17Ni (300 µM), (c) solution of 20 µM HSA previously treated with 20 µM of H2W12 was titrated with R2P2W17Cu (300 µM).
Substituted POMs are known to undergo reversible dissociation/association processes in water solution.12 Thus, as shown in Scheme 1, there are several possible interaction pathways with HSA in such solutions. The observed binding constants strongly depend on the relative values of K1, K2, K3, and K4. The composition effect is reflected in the values of K1, K2, and K4. For different substituted metals, the values significantly vary,
Acknowledgment. This work was supported by the CNRS (UMR 8000), the Université Paris-Sud 11, INSERM, Institut Curie-Recherche and Marie Curie International Incoming Fellowship from the European Community (contract no. 040487).
References and Notes (1) (2) (3) (4)
He, X. M.; Carter, D. C. Nature 1992, 358, 209–215. Carter, D. C.; Ho, J. X. AdV. Protein Chem. 1994, 45, 153–203. Peters, T. All About Albumin; Academic Press: New York, 1996. Olson, R. E.; Christ, D. D. Annu. Rep. Med. Chem. 1996, 31, 327– 336. (5) (a) Pope, M. T. Heteropoly and Isopoly Oxometalates, SpringerVerlag: New York, 1983; (b) Pope, M. T.; Müller, A. Angew. Chem., Int. Ed. Engl. 1991, 30, 34–38. (c) Keita, B; Nadjo, L.
Interactions between Wells-Dawson POMs and HSA Electrochemistry of Isopoly and Heteropoly Oxometalates. In Encyclopedia of Electrochemistry; Bard, A. J., Stratmann, M., Scholz, F., Pickett, C. J., Eds.; Wiley: New York, 2006; Vol. 7, pp 607–700. (6) (a) Baker, L. D. W.; Glick, D. C. Chem. ReV. 1998, 98, 3–49. (b) Kozhevnikov, I. V. Chem. ReV. 1998, 98, 171–198. (7) Judd, D. A.; Nettles, J. H.; Nevins, N. N.; Snyder, J. P.; Liotta, D. C.; Tang, J.; Ermolieff, J.; Schinazi, R. F.; Hill, C. L. J. Am. Chem. Soc. 2001, 123, 886–897. (8) (a) Sarafianos, S. G.; Kortz, U.; Pope, M. T.; Modak, M. J. Biochem. J. 1996, 319, 619–626. (b) Rhule, J. T.; Hill, C. L.; Judd, D. A.; Schinazi, R. F. Chem. ReV. 1998, 98, 327–357. (c) Janell, D.; Tocilj, A.; Koelin, I.; Schluenzen, F.; Gluehmann, M.; Hansen, H. A. S.; Harms, J.; Bashan, A.; Agmon, I.; Bartels, H.; Kessler, M.; Weinstein, S.; Franceshi, F. A.;
Biomacromolecules, Vol. 9, No. 3, 2008 817
(9) (10) (11) (12) (13)
Yonath, A. In Polyoxometalate Chemistry; Pope, M. T., Muller A., Eds.; Kluwer Academic Publishers: Norwell, MA, 2001; pp 391–415; (d) Yamase, T. J. Mater. Chem. 2005, 15, 4773–4782. Zhang, G.; Keita, B.; Brochon, J. C.; Oliveira, P.; Nadjo, L.; Craescu, C. T.; Miron, S. J. Phys. Chem. B 2007, 111, 1809–1814. Zhang, G.; Keita, B.; Craescu, C. T.; Miron, S.; Oliveira, P.; Nadjo, L. J. Phys. Chem. B 2007, 111, 11253–11259. Contant, R. Inorg. Synth. 1990, 27, 106–111. Contant, R. J. Chem. Res. 1984, 4, 120–121. Mohamadi-Nejad, A.; Moosavi-Movahedi, A. A.; Hakimelahi, G. H.; Sheibani, N. Int. J. Biochem. Cell Biol. 2002, 34, 1115–1124.
BM701120J
