Interaction of Ochratoxin A with Human Serum Albumin. A Common

Yuri V. Il'ichev,† Jennifer L. Perry,† and John D. Simon*,†,‡ ... The results show that OTA and WAR share a common binding site in subdomain. ...
7 downloads 0 Views 191KB Size
460

J. Phys. Chem. B 2002, 106, 460-465

Interaction of Ochratoxin A with Human Serum Albumin. A Common Binding Site of Ochratoxin A and Warfarin in Subdomain IIA Yuri V. Il’ichev,† Jennifer L. Perry,† and John D. Simon*,†,‡ Department of Chemistry, Duke UniVersity, Durham, North Carolina 27708, and Department of Biochemistry, Duke UniVersity Medical Center, Durham, North Carolina 27710 ReceiVed: June 19, 2001; In Final Form: October 15, 2001

Optical spectroscopy was used to examine the binding of ochratoxin A (OTA) and warfarin (WAR) to human serum albumin (HSA). Both molecules in the deprotonated form showed high affinity binding to HSA. The close proximity of the highest affinity binding site of the OTA dianion to that of the WAR monoanion was suggested by depolarization of WAR emission in ternary mixtures with [OTA]/[WAR] g 3. Fluorescence polarization data also showed that both OTA and WAR simultaneously bind to HSA for 0.1 e [OTA]/ [WAR] e 1. The failure of WAR to displace OTA under these conditions is in accord with the much smaller binding constant for WAR. In all displacement experiments either the HSA-to-WAR or HSA-to-OTA concentration ratio was kept constant and close to unity. Evidence of energy transfer from electronically excited WAR to OTA when both species are bound to HSA was obtained from fluorescence emission data. The efficiency of this energy transfer provided an estimate of 27 Å as the upper limit of the distance between WAR and OTA. WAR bound to HSA strongly quenched the fluorescence of the single tryptophan residue, Trp214. However, the quenching mechanism was different from the energy transfer mechanism observed for quenching of WAR by OTA. The results show that OTA and WAR share a common binding site in subdomain IIA, with OTA having a higher binding affinity. In addition, WAR can occupy another binding site on HSA when OTA is bound. The data suggest that a secondary binding site of OTA is located in domain III, while that of WAR is in domain I.

Introduction Ochratoxin A (OTA, see Chart 1) is a mycotoxin that originates from several species of Aspergillus and Penicillium fungi and is a common contaminant of foodstuffs.1-5 OTA is primarily nephrotoxic, but it shows many other adverse effects in animals.1,4-10 Food contamination with OTA is also suspected to be responsible for several renal diseases and urinary tract tumors in humans.8-10 Toxicokinetics of OTA is partly responsible for its high toxicity in mammals.4,9,11-14 A recent study15 in a human volunteer showed that half-life of OTA in human plasma is unusually long (853 h after oral administration). The interaction of OTA with serum albumin, the most abundant plasma protein, and the importance of this interaction for the toxicokinetics are well recognized.4,11,16,17 However, the driving forces behind this interaction and the nature of the binding site(s) are not completely understood. In a previous paper18 we reported optical spectroscopic studies designed to characterize the interaction of OTA with human serum albumin (HSA). Our results showed that OTA binds to HSA with high affinity only in the completely deprotonated form (dianion). There are two unique binding sites for OTA and each site can accommodate one dianion. The binding constants for these two sites differ by a factor of ∼50. The data for energy transfer from the single tryptophan residue of HSA (Trp214) to OTA bound to the protein suggest that the highest affinity site is located in one of the two main binding regions of serum albumin, Sudlow site I and site II.19 However, it is impossible to assign the binding to a particular site based only on spectroscopic data. † ‡

Department of Chemistry. Department of Biochemistry.

CHART 1: Molecular Structures of Ochratoxin A (OTA) and Warfarin (WAR)

Identifying the binding site location may be aided by examining the competitive binding of two ligands to HSA. One of the most thoroughly characterized ligands of HSA is warfarin (WAR, see Chart 1), a widely prescribed anticoagulant.20 Numerous experimental studies19,21-33 of WAR-HSA complexes showed that the protein has one high-affinity site that can accommodate an anion of WAR. This site is located in the region of subdomain IIA that corresponds to Sudlow site I. The reported values of the binding constant are in reasonable agreement and range from 2 × 105 to 5 × 105 M-1.19,21-30 Recently Petitpas et al. reported high-resolution crystal structures of two stereoisomers of WAR bound to an HSAmyristate complex.33 The two enantiomers of WAR adopt very similar conformations within the protein matrix, both being in the open form in contrast to the lactone form, which dominates in solution.34 Warfarin occupies two chambers of the “sockshaped” binding pocket formed by all six helices of subdomain IIA. The coumarin moiety fits into the main chamber and makes hydrophobic contacts with Phe223, Leu238, Val241, Ile260, Ile264, Ser287, Ile290, and Ala291.33

10.1021/jp012315m CCC: $22.00 © 2002 American Chemical Society Published on Web 12/15/2001

Interaction of OTA with HSA

J. Phys. Chem. B, Vol. 106, No. 2, 2002 461

Figure 1. Absorbance and fluorescence emission spectra of 10 µM WAR aqueous solutions (40 mM phosphate buffer, pH 7.2) in the absence (solid) and the presence of 10 µM HSA (dashed). Excitation wavelength is 320 nm.

In addition, three of the four oxygen atoms contribute to electrostatic interactions with the protein. According to Petitpas et al.,33 the O atom of the hydroxyl group forms hydrogen bonds with the imidazole ring of His242 and a bound water molecule. The oxygen atoms of the two CdO groups are in close proximity to the side chain of Arg222. The benzyl moiety of WAR binds to a chamber formed by Phe211, Trp214, Leu219, and Leu238. Additional contacts with aliphatic parts of polar residues Arg218 and His242 also play an important role in the binding. Crystallographic studies32,33 provided direct evidence for the location of the highest affinity site of WAR in subdomain IIA. However, these studies were able to locate unambiguously only one site in contrast to the majority of other techniques, which clearly show the presence of other site(s) with binding constants e3 × 104 M-1.23,27-30 The binding parameters for these site(s) vary significantly and their location is not completely clear. The recent results of Ru¨ker and co-workers31 for three recombinant domains of HSA and earlier data of Bos et al.25 for large fragments of HSA indicate the presence of a secondary binding site in domain I. It is interesting to note that in the X-ray data of Petitpas et al.33 there is a fragment of electron density in subdomain IB, which is large enough to encompass the coumarin group of WAR. But these authors could not find the density for other parts of WAR and they did not include this site in their model. In the present paper, optical spectroscopy was used to examine the competitive binding of OTA and WAR to HSA. The results allowed us to determine the location of the highest affinity site of OTA. Furthermore, the spectral data clearly showed that OTA and WAR are able to simultaneously bind to HSA provided that 0.1 e [OTA]/[WAR] e 1. This provides a deeper insight into the nature of lower affinity sites for these two molecules. Experimental Section Warfarin (Sigma-Aldrich) was used as received. All details concerning other chemicals, solution preparation and instrumentation are described in the previous paper.18 Results and Discussion Competitive Binding of WAR and OTA. In 10 µM WAR solution (40 mM phosphate buffer, pH 7.2), the absorption and emission spectra exhibited peaks at 308 and 389 nm, respectively (Figure 1). In the presence of 10 µM HSA, the absorption

Figure 2. (A) Fluorescence polarization spectra of aqueous solutions (pH 7.2) containing 10 µM WAR, 30 µM HSA (dashed) and 10 µM WAR (solid) excited at 320 nm. (B) Fluorescence polarization spectra of aqueous solutions (pH 7.2) containing 10 µM OTA, 30 µM HSA (dashed), and 10 µM OTA (solid) excited at 375 nm.

spectrum showed only a small red shift to 310 nm, but the molar absorptivity of WAR increased substantially. The fluorescence intensity (λex ) 320 nm) increased by a factor of 8.0 and the emission spectrum narrowed and underwent a significant blue shift (λmax ) 379 nm). This fluorescence enhancement in the presence of HSA reflects an increase in both the molar absorptivity and the emission quantum yield. The increase in absorbance was a factor of 2 at 320 nm; the fluorescence quantum yield (φ) at pH ∼ 7 has been shown by Kasai-Morita et al.24 to increase from ∼0.011 to ∼0.078 upon binding to HSA. We can estimate the fraction of WAR bound to HSA (R) from binding parameters in the literature:30 for K1 ∼ 3 × 105 M-1, n1 ) 1, and K2 ∼ 1 × 104 M-1, n2 ) 2, and under our experimental conditions, R ) 0.61. Combining this figure with the above differences in absorbance and quantum yields gives a fluorescence intensity ratio of 7.4, in good agreement with our measured value of 8.0. In solution, the protonated form of WAR exists predominantly as a cyclic hemiketal.34 However, the anion of both the cyclic and open protonated forms is open,34,35 and at neutral pH we are concerned mainly with the anion (apparent pKa(aq) ∼5).35 Thus for convenience, WAR is depicted in Chart 1 in its open form. The absorption and fluorescence spectra observed in the presence of HSA suggest that the bound species is the “openform” monoanion, W-, in agreement with a crystallographic study33 showing that WAR is bound to the HSA-myristate complex in the open form. The binding of W- to HSA indicated by the fluorescence quantum yield and other spectral parameters was confirmed by emission polarization spectroscopy. Figure 2A shows polarization spectra of W- in water and in a complex with HSA. Polarization data for the dianion of OTA (A2-) discussed in detail in the previous paper18 are also presented for purpose of comparison. In aqueous solutions, W- and A2- exhibited wavelength independent polarizations of 0.18 and 0.01, respectively. The observation of a nonzero polarization for W- is consistent with the very short excited-state lifetime of this molecule in aqueous solution (∼50 ps36). The emission polarization spectra for binary complexes of W- and A2- with HSA are also shown in Figure 2. At the

462 J. Phys. Chem. B, Vol. 106, No. 2, 2002

Figure 3. Fluorescence polarization spectra measured in an aqueous solution (pH 7.2) containing 10 µM WAR and 10 µM HSA (dashed); 10 µM WAR, 10 µM HSA and 100 µM OTA (solid); 100 µM OTA and 10 µM HSA (dashed-dot) at pH 7.2 for the excitation at 320 nm (A) and 375 nm (B).

concentration ratios presented, A2- and W- were completely bound to HSA. When excited within the lowest energy absorption band, both W- and A2- bound to HSA exhibited a wavelength-independent polarization of 0.45 and 0.43, respectively, indicating that both anions are practically immobilized within their individual complexes with HSA. In addition, these high polarization values show that absorbance and emission transition moments are almost parallel for both molecules. It should be noted that our time-resolved data18 for OTA gave a limiting polarization value of 0.49. The small depolarization observed in steady-state experiments resulted exclusively from the rotational tumbling of the protein molecule. To determine whether W- and A2- share a common binding site in HSA, we studied ternary mixtures of A2- and W- with HSA. The published values of the binding constant for the highest affinity site of W- are about an order of magnitude smaller than that for A2- 18. Both species are believed to have secondary binding sites characterized by significantly lower binding constants. In our experiment, we assumed equilibrium conditions in the presence of 10 µM HSA and measured OTA and WAR binding over a range of [OTA]/[WAR] ) 0.1-10. A comparison of the absorption and emission spectra for WAR and OTA (see Figure 1 and the previous paper18) shows spectral regions where the overlap between W- and A2- spectra is small or nonexistent. As one excitation wavelength for emission polarization spectroscopy we chose 375 nm, which excites only A2-. We also used 320 nm, where both W- and A2- are excited, and the relative contribution depends on the concentration of the individual species present. Although the emission spectra of the two anions overlap there are two regions strongly dominated by a single species: 330-400 nm by Wand >450 nm by A2-. We first examined the emission polarization for binary mixtures of 100 µM OTA or 10 µM WAR with 10 µM HSA (Figure 3). Under these conditions, we expect each molecule of HSA to contain at least one A2-, located predominantly in the highest affinity site. In a 1:1 WAR-HSA mixture excited at 320 nm, P ) 0.42. Using that and the parameter θ ≈ 14, we estimated the fraction of WAR bound to HSA (see eq 2 and parameter’s definition in ref 18). From polarization data we

Il’ichev et al.

Figure 4. Fluorescence polarization spectra of aqueous solutions (pH 7.2) containing 10 µM OTA and 10 µM HSA (dashed), 10 µM OTA, 10 µM HSA, and 100 µM WAR (solid), and 100 µM WAR and 10 µM HSA (dashed-dot), for excitation at 320 nm (A) and 375 nm (B).

obtained R ) 0.68, which is in good agreement with R ) 0.61 calculated from the binding constants. When a ternary mixture of 100 µM OTA, 10 µM WAR, and 10 µM HSA was excited at 320 nm (Figure 3), the depolarization of W- in the presence of OTA was evident as P changed from 0.42 in the binary mixture to P ) 0.21 in the ternary one. In fact, the P value in the ternary mixture was close to the polarization of WAR emission in water. The polarization spectrum of A2- showed no significant deviation from the binary mixture (100 µM OTA, 10 µM HSA) to the ternary one (P ) 0.05). This indicates that A2- is bound to the same extent in the ternary mixture as in the binary complex and that WAR is almost completely displaced from HSA. The polarization spectra of OTA excited at 375 nm (see Figure 3B) confirmed this conclusion, with a polarization value of 0.12 recorded for both mixtures. Next we reversed the relative concentrations of WAR and OTA in the binary and ternary mixtures (100 µM WAR, 10 µM OTA, and 10 µM HSA). Under these concentration conditions, each HSA molecule in the binary mixtures is expected to contain at least one bound WAR or OTA located in the highest affinity site. When the excitation wavelength was 320 nm, both W- and A2- were excited; however, for the concentrations used, the ratio of the absorbances of the two molecules was A(WAR)/A(OTA) ≈ 14. Thus, the absorption was dominated by W-, but the emission contained contributions from both molecules. The spectral region of W- (350-400 nm) showed P ) 0.37 (Figure 4A), which was much higher than the value of P observed in buffer. This increase indicates that W- is partially bound to the protein. We observed a small decrease in the W- polarization in the ternary complex (P ) 0.32) compared to that in the binary solution. It is interesting to compare this P value with those calculated from the binding constants. If we take the binding parameters for the two sites and the spectral parameters of WAR mentioned above we obtain P ) 0.37, but if we exclude the highest affinity site from the calculation we get P ) 0.33 for an aqueous solution containing 100 µM WAR and 10 µM HSA (see eqs 2 and 3 in ref 18). Both values are close to those experimentally observed in the binary and ternary mixture, respectively. This result suggests that despite a 10-fold excess

Interaction of OTA with HSA WAR is mainly displaced from the highest affinity site by OTA, but it still binds in a secondary binding site to about the same extent as in the binary mixture. Polarization spectra of A2- excited at 375 nm in the binary and ternary mixtures are shown in Figure 4B. The wavelengthindependent value of P ) 0.38 for both mixtures confirmed that A2- was bound and not displaced at 0.1 e [OTA]/[WAR] < 1. This is consistent with the substantially different binding constants for the two anions. In addition, we noticed that P values for A2- excited at 320 nm in the ternary complex (P ) 0.09) were lower than those in the binary complex (P ) 0.22). In contrast, the emission polarization for 375 nm excitation was the same for these two complexes (see Figure 4A,B). Examination of the spectral data showed that the origin of this behavior was energy transfer as described in detail later. A comparison of the polarization spectra in Figure 2B, 3B, and 4B, which were obtained with the same excitation wavelength of 375 nm, proved that the depolarization seen for higher OTA-to-HSA concentration ratios was caused by an excess of free A2- present in solution. Recall that A2- in an aqueous buffer is characterized by P ) 0.01. It is noteworthy that when A2- in the binary complexes was excited at 320 nm, the polarization was lower than that observed for excitation at 375 nm (see Figures 3 and 4). At 320 nm, both the first (S1) and second (S2) excited singlet states of A2- are populated. Therefore, a smaller P value for 320 nm reflects a larger angle between the transition dipole moments of the emission and S2 absorption compared to S1 absorption. Fluorescence Quenching in HSA Complexes with WAR and OTA. Now we return to the additional depolarization of OTA emission observed for 320 nm excitation of the ternary complex compared to the binary one. This result may be explained by energy transfer from electronically excited W- to A2- in the protein matrix. Figure 5A shows the emission spectra (λexc ) 320 nm) of the equimolar binary mixtures of WAR or OTA with HSA and that of the ternary mixture in which WAR, OTA and HSA were present at the same concentrations (10 µM). The intensity of WAR emission in the ternary complex was less than that in its binary complex with HSA by a factor of ∼ 3. One could conclude that this decrease results solely from displacement of W- by A2- because the two species share a common binding site with a higher affinity to OTA, and the fluorescence efficiency for W- is much lower in water than in the protein environment. However, the similar absorption and emission polarization spectra of WAR in both mixtures (data not shown) together with the data presented above do not support such a conclusion for these concentration ratios. Furthermore, the intensity of the emission from A2- in the ternary complex was a factor of 2.4 higher than that in the binary complex. It should be noted that absorbance of OTA at 375 nm was the same for both complexes and its emission quantum yield was practically insensitive to environment. Thus we may conclude that energy is transferred from electronically excited W- to A2-. We should be able to use the observation of energy transfer from W- to A2- to estimate the distance R between the two bound molecules. The relationship between R and the energy transfer efficiency, E, is given by eq 4 in ref 18. As shown earlier, the emission intensity of W- in the ternary complex with HSA and OTA was about one-third of that observed for its binary mixture with HSA (Figure 5A). If we assume the reduced intensity in the ternary complex resulted solely from energy transfer, then E is approximately 0.7 and R/Ro ∼ 0.87.

J. Phys. Chem. B, Vol. 106, No. 2, 2002 463

Figure 5. (A) Fluorescence emission spectra of a WAR-HSA solution (solid), a WAR-OTA-HSA solution (dashed), and an OTA-HSA solution (dots) following excitation at 320 nm. (B) Fluorescence emission spectra of an HSA solution (dash-dots) and a WAR-HSA solution (solid) following excitation at 295 nm. Dotted lines correspond to the individual emission bands of Trp214 and WAR in the WARHSA mixture. Concentration for all species was 10 µM. All spectra were recorded in 40 mM phosphate buffer at pH 7.2.

Ro is the characteristic distance related to the properties of donor and acceptor.37 For freely rotating molecules Ro can easily be estimated within the Fo¨rster formalism.37 By applying this formalism to the energy transfer between W- and A2-, we obtained Ro ) 22 Å and the apparent distance R ) 20 Å. However, our polarization data showed that rotational motion of W- and A2- within the serum albumin was almost completely frozen, so the geometrical factor κ2 ) 2/3 for the free rotation is not applicable to this system. Steinberg38 demonstrated that for systems in which rotational motion is completely frozen, the time dependence of survival probability for the energy donor (and therefore the expression for energy transfer efficiency) is similar in form to that for freely rotating molecules provided that donor and acceptor molecules are randomly distributed and randomly oriented in space at the moment of excitation. For this case, only slightly smaller Ro and R values will be obtained. But the protein environment should create large constraints on the relative orientation of bound species and hence we can hardly expect completely random orientations for the transition dipole moments of W- and A2- when they are simultaneously bound to HSA. In the case of unique orientation of OTA relative to WAR and of complete restriction of rotational motion, the orientation factor κ2 can have a value between 0 and 4.37 Unfortunately, in this case even knowledge of the fluorescence depolarization resulting from energy transfer yields little insight in the value of κ2.37a By using the maximum κ2 value of 4 we were able to obtain only an upper limit of 27 Å for the distance between W- and A2-. In the previous paper18 we reported energy transfer from Trp214 to OTA bound to HSA and used it to estimate the distance between two chromophores. In this work we tried to apply the same formalism to WAR-HSA complexes based on

464 J. Phys. Chem. B, Vol. 106, No. 2, 2002 fluorescence spectra of an HSA solution and a WAR-HSA mixture excited at 295 nm (Figure 5B). Although strong quenching of Trp214 emission was observed, there was no evidence for energy transfer because the fluorescence intensity of WAR in the presence of the protein increased only to the extent expected based on its higher absorptivity and emission quantum yield in the protein environment. Under these conditions the fraction of WAR bound to HSA was ∼0.6 and therefore the apparent quantum yield in the binary mixture was ∼3.7 times higher than in water (see above for details of calculation and fluorescent parameters of bound and free WAR). Using this value and the ratio of 1.6 for the molar absorptivities at 295 nm for the bound and free ligand we obtained a fluorescence intensity ratio of ∼6, which is very close to the value of 6.4 calculated from experimental data in Figure 5B. This failure to observe increased WAR fluorescence means that the mechanism of tryptophan fluorescence quenching is different for WAR than that for OTA. Localization of Binding Sites for WAR and OTA. In the previous paper18 we presented results showing that OTA has two binding sites in HSA, each being able to accommodate one dianion. Previous studies of WAR binding isotherms suggested that the drug also binds to different sites in HSA with different affinities.27-30 The majority of these studies are supported by recent crystallographic data32,33 and demonstrate that the highest affinity site for WAR is located in the region of Sudlow site I. Figure 6A shows a detailed structure of this site and highlights Trp214 and the polar residues involved in binding. Our displacement data show a close proximity of the highest affinity sites for OTA and WAR. The domain structure of HSA suggests two reasonable positions for a secondary site for both WAR and OTA. The region in subdomain IIIA referred to as Sudlow site II19 (see Figure 1 in ref 18) is well recognized as another important region for ligand binding. There is also a homologous site in subdomain IA, although lower affinity to that site is observed for the majority of ligands,32,39 possibly because the structure of domain I is slightly different from that of the two other domains.32 However, the crystal structures of 2,3,5-triiodobenzoic acid (TIB) with HSA and the HSA-myristate complex clearly show a high affinity binding site for this ligand in subdomain IIA and two secondary binding sites in domain III and domain I.32,39,40 Apparently the affinity of TIB to Sudlow site II in domain III is higher than to the corresponding site in domain I, but when Sudlow site II is occupied by a fatty acid anion, TIB binding to subdomain IA is observed.40 A detailed view of the TIB anion and its environment in Sudlow site I, which also corresponds to the highest affinity site of WAR and OTA, is shown in Figure 6B. Some experimental data has suggested that the secondary binding site of WAR is located in domain I,25,31,33 presumably in the region where the second anion of TIB binds to the HSAmyristate complex. The results obtained in the present study support this location. The energy transfer estimate for the distance between WAR and OTA is too large to distinguish between two possible locations of a lower affinity binding site for WAR. But this estimate is quite consistent with positioning of this site both within Sudlow site II and in the region where TIB binds to domain I.41 Although at this stage we cannot make a definitive conclusion about location of a secondary binding site for OTA, it is feasible to assign it to the well-known binding region in subdomain IIIA. Our data for emission polarization support different positions of the lower-affinity sites for OTA and WAR. We find that

Il’ichev et al.

Figure 6. Binding interactions for WAR (A) and TIB (B) in subdomain IIA. The amino acids encompassing the binding site are shown in a ball-and-stick representation. For the WAR complex the two water molecules involved in hydrogen bonding within the binding site are depicted as green spheres.

polarization values can be well modeled under concentration conditions where significant occupancy of lower-affinity sites should be observed provided that independent binding of these two anions is assumed. It should be noted that both binding constants for the OTA dianion are at least an order of magnitude larger than those for the WAR anion. These data also suggest that the site with the second largest binding constant for OTA has a different location than that for WAR. We are currently trying to clarify this point by using OTA interaction with three recombinant domains of HSA and its competitive binding in the presence of different ligands. These studies will be published elsewhere. Conclusions Emission and polarization data show that WAR and OTA share a common binding site in subdomain IIA of HSA with OTA having the higher binding affinity. Under certain conditions OTA and WAR co-bind to HSA and energy transfer occurs from excited W- to A2-. On the basis of the efficiency of energy transfer, the distance between these two anions was estimated to be at most 27 Å. The data obtained in this study are consistent with the proposed location of a secondary binding site of WAR in subdomain IB and that of OTA in subdomain IIIA.

Interaction of OTA with HSA Acknowledgment. Duke University and a Faculty Recruitment Grant from the North Carolina Biotechnology Center supported this work. References and Notes (1) (a) Kuiper-Goodman, T.; Scott, P. M. Biomed. EnViron. Sci. 1989, 2, 179-248. (b) Kuiper-Goodman, T. Food Addit. Contam. 1996, 13, 5357. (2) Chu, F. S. Mutat. Res. 1991, 259, 291-306. (b) Scudamore, K. A. Food Addit. Contam. 1996, 13, 39-42. (c) Pittet, A. ReV. Med. Vet. 1998, 149, 479-492. (3) Pohland, A. E.; Nesheim, S.; Friedman, L. Pure Appl. Chem. 1992, 64, 1029-1046. (4) Marquardt, R. R.; Frohlich, A. A. J. Anim. Sci. 1992, 70, 39683988. (5) Fink-Gremmels, J. Vet. Q. 1999, 21, 115-120. (6) (a) Simon, P. J. Toxicol.-Toxin ReV. 1996, 15, 239-249. (b) Simon, P. J. Toxicol.-Toxin ReV. 1999, 18, 313-321. (7) (a) Ro¨schenthaler, R.; Creppy, E. E.; Dirheimer, G. J. Toxicol.Toxin ReV. 1984, 3, 53-86. (b) Creppy, E. E. J. Toxicol.-Toxin ReV. 1999, 18, 277-293. (8) (a) Gekle, M.; Silbernagl, S. Kidney Blood Pressure Res. 1996, 19, 225-235. (b) Gekle, M.; Sauvant, C.; Schwerdt, G.; Silbernagl, S. Kidney Blood Pressure Res. 1998, 21, 277-279. (9) Petzinger, E.; Ziegler, K. J. Vet. Pharmacol. Ther. 2000, 23, 9198. (10) Steyn, P. S.; Stander, M. A. J. Toxicol.-Toxin ReV. 1999, 18, 229243. (11) Kumagai, S. Food Chem. Toxicol. 1985, 23, 941-943. (12) Hagelberg, S.; Hult, K.; Fuchs, R. J. Appl. Toxicol. 1989, 9, 9196. (13) Fuchs, R.; Hult, K. Food Chem. Toxicol. 1992, 30, 201-204. (14) Galtier, P.; Alvinerie, M.; Charpenteau, J. L. Food Cosmet. Toxicol. 1981, 19, 735-738. (15) Studer-Rohr, I.; Schlatter, J.; Dietrich, D. R. Arch. Toxicol. 2000, 74, 499-510. (16) Chu, F. S. Arch. Biochem. Biophys. 1971, 147, 359-366. (17) (a) Uchiyama, S.; Saito, Y.; Uchiyama, M. J. Food Hyg. Soc. Jpn. 1985, 26, 651-657. (b) Uchiyama, S.; Saito, Y. J. Food Hyg. Soc. Jpn. 1987, 28, 453-460. (18) Il’ichev, Yu. V.; Perry, J. L.; Simon, J. D. J. Phys. Chem. B 2002, 106, 452. (19) (a) Sudlow, G.; Birkett, D. J.; Wade, D. N. Mol. Pharmacol. 1975, 11, 824-832. (b) Sudlow, G.; Birkett, D. J.; Wade, D. N. Mol. Pharmacol. 1976, 12, 1052-1061. (20) Palareti, G.; Legnani, C. Clin. Pharmacokinet. 1996, 30, 300-313 and references therein. (21) (a) Fehske, K. J.; Muller, W. E.; Wollert, U.; Velden, L. M. Mol. Pharmacol. 1979, 16, 778-789. (b) Fehske, K. J.; Schlafer, U.; Wollert, U.; Muller, W. E. Mol. Pharmacol. 1982, 21, 387-393. (22) (a) Maes, V.; Maras, Y.; Vercruysse, A. Toxicol. Lett. 1980, 162162. (b) Maes, V.; Engelborghs, Y.; Hoebeke, J.; Maras, Y.; Vercruysse, A. Mol. Pharmacol. 1982, 21, 100-107.

J. Phys. Chem. B, Vol. 106, No. 2, 2002 465 (23) (a) Wilting, J.; Vandergiesen, W. F.; Janssen, L. H. M.; Weideman, M. M.; Otagiri, M.; Perrin, J. H. J. Biol. Chem. 1980, 255, 3032-3037. (b) Wilting, J.; Vandergiesen, W. F.; Janssen, L. H. M. Biochem. Pharmacol. 1981, 30, 1025-1031. (c) Wilting, J.; Kremer, J. M. H.; Ijzerman, A. P.; Schulman, S. G. Biochim. Biophys. Acta 1982, 706, 96-104. (24) Kasai-Morita, S.; Horie, T.; Awazu, S. Biochim. Biophys. Acta 1987, 915, 277-283. (25) (a) Bos, O. J. M.; Remijn, J. P. M.; Fischer, M. J. E.; Wilting, J.; Janssen, L. H. M. Biochem. Pharmacol. 1988, 37, 3905-3909. (b) Bos, O. J. M.; Fischer, M. J. E.; Wilting, J.; Janssen, L. H. M. Biochem. Pharmacol. 1989, 38, 1979-1984. (26) (a) Kragh-Hansen, U. Mol. Pharmacol. 1988, 34, 160-171. (b) Kragh-Hansen, U.; Brennan, S. O.; Galliano, M.; Sugita, O. Mol. Pharmacol. 1990, 37, 238-242. (27) Pinkerton, T. C.; Koeplinger, K. A. Anal. Chem. 1990, 62, 21142122 and references therein. (28) Loun, B.; Hage, D. S. J. Chromatogr.-Biomed. Appl. 1992, 579, 225-235. (b) Loun, B.; Hage, D. S. Anal. Chem. 1994, 66, 3814-3822. (c) Loun, B.; Hage, D. S. Anal. Chem. 1996, 68, 1218-1225. (29) (a) Kosa, T.; Maruyama, T.; Otagiri, M. Pharm. Res. 1997, 14, 1607-1612. (b) Kosa, T.; Maruyama, T.; Sakai, N.; Yonemura, N.; Yahara, S.; Otagiri, M. Pharm. Res. 1998, 15, 592-598. (30) Villamor, J. P.; Zaton, A. M. L. J. Biochem. Biophys. Methods 2001, 48, 33-41. (31) (a) Dockal, M.; Carter, D. C.; Ru¨ker, F. J. Biol. Chem. 1999, 274, 29303-29310. (b) Dockal, M.; Chang, M.; Carter, D. C.; Ru¨ker, F. Protein Sci. 2000, 9, 1455-1465. (32) He, X. M.; Carter, D. C. Nature 1992, 358, 209-215. (33) Petitpas, I.; Bhattacharya, A. A.; Twine, S.; East, M.; Curry, S. J. Biol. Chem. 2001, 276, 22804-22809. (34) Valente, E. J.; Lingafelter, E. C.; Porter, W. R.; Trager, W. F. J. Med. Chem. 1977, 20, 1489-1493. (35) (a) Stella, V. J.; Mooney, K. G.; Pipkin, J. D. J. Pharm. Sci. 1984, 73, 946-948. (b) Ishihama, Y.; Oda, Y.; Asakawa, N. J. Pharm. Sci. 1994, 83, 1500-1507. (36) Fluorescence lifetime of WAR, τ, may be estimated from quantum yield data and radiative lifetimes, τR, calculated from absorbance spectra by using the Strickler-Berg equation. For WAR in water and bound to HSA we obtained radiative lifetimes of 4.9 ns and 2.9 ns. The fluorescence lifetime is then τ ) φτR ∼ 54 and 230 ps for free and bound warfarin, respectively. (37) (a) Dale, R. E.; Eisinger, J. Biopolymers 1974, 13, 1573-1605. (b) Stryer, L. Annu. ReV. Biochem. 1978, 47, 819-846. (c) Selvin, P. R. Methods Enzymol. 1995, 246, 300-341. (38) Steinberg, I. Z. J. Chem. Phys. 1968, 48, 2411-2413. (39) Carter, D. C.; Ho, J. X. AdV. Protein Chem. 1994, 45, 153-203. (40) Curry, S.; Mandelkow, H.; Brick, P.; Franks, N. Nat. Struct. Biol. 1998, 5, 827-835. (41) The distances between residues involved in binding within Sudlow site I and site II are in the range from 11 to 33 Å. The maximum distance was estimated using atoms of Ile264 of site I and Leu387 of site II from the crystal structure of HSA (PDB ID 1AO6). The minimum distance between them was obtained using atoms of Trp214 and Ala449. The distances between the sites in domain I and II are in the range 14-27 Å. This range was estimated from the distances between two anions of TIB bound to the HSA-myristate complex (1BKE).