Spectrometric and voltammetric studies of the interaction between quercetin and bovine serum albumin using warfarin as site marker with the aid of chemometrics

Spectrochimica Acta Part A 71 (2009) 1865–1872

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Spectrometric and voltammetric studies of the interaction between quercetin and bovine serum albumin using warfarin as site marker with the aid of chemometrics Yongnian Ni a,b,∗ , Xia Zhang b , Serge Kokot c a b c

State Key Laboratory of Food Science and Technology, Nanchang University, Nanchang, Jiangxi 330047, China Department of Chemistry, Nanchang University, Nanchang, Jiangxi 330047, China School of Physical and Chemical Sciences, Queensland University of Technology, Brisbane, Queensland 4001, Australia

a r t i c l e

i n f o

Article history: Received 4 April 2008 Received in revised form 2 July 2008 Accepted 3 July 2008 Keywords: Quercetin Bovine serum albumin Warfarin Chemometrics Multivariate curve resolution-alternating least squares

a b s t r a c t The interaction of quercetin, which is a bioflavonoid, with bovine serum albumin (BSA) was investigated under pseudo-physiological conditions by the application of UV–vis spectrometry, spectrofluorimetry and cyclic voltammetry (CV). These studies indicated a cooperative interaction between the quercetin–BSA complex and warfarin, which produced a ternary complex, quercetin–BSA–warfarin. It was found that both quercetin and warfarin were located in site I. However, the spectra of these three components overlapped and the chemometrics method – multivariate curve resolution-alternating least squares (MCR-ALS) was applied to resolve the spectra. The resolved spectra of quercetin–BSA and warfarin agreed well with their measured spectra, and importantly, the spectrum of the quercetin–BSA–warfarin complex was extracted. These results allowed the rationalization of the behaviour of the overlapping spectra. At lower concentrations ([warfarin] < 1 × 10−5 mol L−1 ), most of the site marker reacted with the quercetin–BSA, but free warfarin was present at higher concentrations. Interestingly, the ratio between quercetin–BSA and warfarin was found to be 1:2, suggesting a quercetin–BSA–(warfarin)2 complex, and the estimated equilibrium constant was 1.4 × 1011 M−2 . The results suggest that at low concentrations, warfarin binds at the high-affinity sites (HAS), while low-affinity binding sites (LAS) are occupied at higher concentrations. © 2008 Elsevier B.V. All rights reserved.

1. Introduction Serum albumin (SA) is one of the most abundant proteins, and plays an important role in the transport and deposition of a variety of ligands in blood [1–4]. Distribution and metabolism of many biologically active compounds, e.g. drugs and natural products, in the body have been correlated with their affinities towards SA. Thus, investigation of such molecules with respect to their binding with SA is important. A number of biochemical and molecular biological investigations have revealed that proteins are frequently the “targets” for therapeutically active flavonoids of both natural and synthetic origins [5]. Flavonoids are common constituents of plants and therefore have been identified in a broad range of fruits and vegetables as well as in beverages such as tea, red wine, coffee, and beer.

∗ Corresponding author at: State Key Laboratory of Food Science and Technology, Nanchang University, Nanchang, Jiangxi 330047, China. Tel.: +86 791 3969500; fax: +86 791 3969500. E-mail address: [email protected] (Y. Ni). 1386-1425/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2008.07.004

Flavonoids consist of two aromatic rings (A and B, Table 1) linked by an oxygen-containing heterocyclic ring (C) [6]. The most abundant naturally occurring flavonoid, quercetin (3,5,7,3 ,4 pentahydroxyflavone, Table 1), inhibits the activity of enzymes such as calcium phospholipid-dependent protein kinase [7] among others. Recent work has shown that SA plays an important role in the transport and disposition of flavonoids, especially of quercetin [8,9], and the research with these compounds has been focused on: (i) their potential biological and therapeutic applications, and (ii) their unusual fluorescence emission properties. The first point refers to the growing evidence for flavonoids therapeutic activities of high potency and low systemic toxicity against, for example, cancers, AIDS, and allergies [5,10]. There are also the well-known antioxidatiant effects of flavonoids [11–13]. The second aspect relates to such properties as the very large Stokes shifts, dual fluorescence behaviour, and high sensitivity of the emission parameters to the environment (e.g., hydrogen bonding effects, pH, temperature) [14–16]. The multiplicity of binding sites on SA available for small molecules creates a challenge to assess interactions between differ-

1866

Y. Ni et al. / Spectrochimica Acta Part A 71 (2009) 1865–1872

Table 1 Structural formulae of quercetin, warfarin and ibuprofen Name

Molecular formula

Quercetin

C15 H10 O7

Warfarin

C19 H16 O4

Structural formula

In this paper, the interaction between quercetin and BSA was investigated by UV–vis spectrometry, spectrofluorimetry, and cyclic voltammetry using warfarin (Table 1) as site I marker and ibuprofen (Table 1) as site II marker. The responses from each technique were investigated independently, and the qualitative and quantitative conclusions were compared. The chemometrics method – multivariate curve resolution-alternating least squares (MCR-ALS) was applied to resolve the two-way UV–vis spectra so as to extract the equilibrium concentration profiles of the reacting species and the associated quantitative information. 2. Experimental 2.1. Apparatus

Ibuprofen

C13 H18 O2

ent ligands and SA, whether these are competitive or cooperative. However, it is important to address this issue in order to obtain a fuller description of the ligand-binding properties of SA [17]. SA has two major drug-binding sites (sites I and II), which are located within particular cavities in subdomains IIA and IIIA, respectively [18]. Three main regions or subsites have been described in the site I complex: bilirubin–phenol red, warfarin–phenylbutazone, and digitoxin–salicylate. The site II ligands include l-tryptophan, l-thyroxin, octanoate, diazepam and other benzodiazepines, iopanate and clofibrate [19]. In a competitive binding interaction, two ligands (the drug and the site marker) usually have an affinity for the same binding site on the protein molecule, and the extent of drug–protein binding may be revealed by monitoring the change in the fluorescence response intensity of the system [20]. The interaction of quercetin with SA has been studied by spectrometry. Kanakis et al. [7] examined the interaction of human serum albumin (HSA) with quercetin, kaempferol and delphinidin in an aqueous solution under physiological conditions. Fourier transform infrared spectroscopy (FTIR) and UV–vis spectroscopic methods were applied to determine the polyphenolic binding mode, the binding constant, and the effects of flavonoid complexation on the secondary structure of the protein. Dufour and Dangles [21] investigated the binding of different flavonoids to human and bovine serum albumin (BSA) by fluorescence spectroscopy with the use of three methods – the quenching of the albumin fluorescence, the enhancement of the flavonoid fluorescence, and the quenching of the fluorescence of the quercetin–albumin complex by a second flavonoid. Bi et al. [22] studied the interaction of HSA with the flavonoids: quercetin, rutin, hyperin and baicalin, and estimated the binding constants and the number of binding sites of the flavonoids. Zsila et al. [6] probed the binding of quercetin to HSA by circular dichroism, electronic absorption spectroscopy, and molecular modeling methods. Sengupta and Sengupta [23] also investigated the interaction of quercetin with HSA by the fluorescence quenching method. Their work indicated that the quercetin molecules bind at a dynamically restricted site near tryptophan213 in the inter-domain cleft region of HSA. Furthermore, the binding constant (K = 1.9 × 105 M−1 ) and Gibbs free energy change (G0 = −30.12 kJ mol−1 ) for the interaction have been calculated. Recently, the interaction of salicylic acid with BSA was studied with spectrofluorimetry and chemometrics – particularly with the application of parallel factor analysis (PARAFAC) [24].

The fluorescence spectra were measured on a PerkinElmer LS 55 luminescence spectrometer equipped with a thermostatic bath (Model ZC-10, Tianheng Instruments Factory, Ningbo, China) and a 10 mm quartz cuvette. The excitation and emission slits were set at 10 nm while the scanning rate was 1500 nm min−1 . The UV–vis absorption spectra were measured on an Agilent UV-8453 spectrophotometer with the use of a 10 mm cell. Voltammetric studies were carried out on a CHI-660A electrochemical workstation (Chenhua Instrumental Company, Shanghai) with a three-electrode system: the working electrode – a glassy carbon, the reference electrode – an Ag/AgCl, and the counter electrode – a platinum wire. 2.2. Materials A BSA solution (1.5 × 10−3 mol L−1 ) was prepared by dissolving 5.0 g BSA (Purified BSA; Mr = 66,000; “essentially globulin and fatty acid free” quality; Huamei Biological Co. Ltd, Shanghai), in 50 mL sodium chloride solution (50 mmol L−1 ) and stored at 4 ◦ C. Its purity was estimated to be 99.5% based on the measured absorbance value at 279 nm compared to the reference value of 0.667 for 1.0 g L−1 pure BSA [25]. A quercetin stock solution (5.0 × 10−3 mol L−1 ) was prepared from its crystals by dissolving them in a small amount of ethanol, followed by further dilution to the desired volume with water. Warfarin (3.0 × 10−3 mol L−1 ) and ibuprofen (2 × 10−4 mol L−1 ) stock solutions were prepared by dissolving their crystals in distilled water, and diluting to the desired concentration. The pHs of all solutions were adjusted with the Tris–HCl buffer (0.05 mol L−1 , pH 7.4), which was prepared by mixing and diluting 25 mL buffer (0.2 mol L−1 ) and 45 mL HCl (0.1 mol L−1 ) to 100 mL. All chemicals were analytical grade reagents. Doubly distilled water was used throughout. 2.3. Procedures 2.3.1. Voltammetric characterization of quercetin interacting with BSA The concentration of quercetin was kept at 8.28 × 10−6 mol L−1 , and the BSA solution was added at different concentrations in the range of 0–1.13 × 10−5 mol L−1 . A given quercetin–BSA system was stirred for 10 s and then, the cyclic voltammetry experiment was performed at pH 7.4 (scanning potential: from 0 to 0.8 V; sweep rate = 100 mV s−1 ). 2.3.2. Fluorescence quenching study A 2.7 × 10−7 mol L−1 BSA solution was added to quercetin solutions of different concentrations in the range of 0.0–2.38 × 10−7 mol L−1 . The solutions were mixed thoroughly, and equilibrated for 10 min at 300 K. Fluorescence spectra were then

Y. Ni et al. / Spectrochimica Acta Part A 71 (2009) 1865–1872

measured in the range of 260–600 nm at the excitation wavelength of 280 nm. The measurements were repeated but at 310 K. Each spectrum was background corrected by subtracting a spectrum of the buffer blank. 2.3.3. Fluorescence enhancement study Two experiments were performed: (i) sample series 1 − constant [BSA] = 1.55 × 10−5 mol L−1 , and [quercetin] was varied from 0 to 2.10 × 10−5 mol L−1 ; (ii) sample series 2 − constant [quercetin] = 1.65 × 10−5 mol L−1 , and [BSA] was varied from 0 to 1.80 × 10−5 mol L−1 . Fluorescence spectra were recorded for both sets of samples (i and ii) with the excitation wavelength of 465 nm. 2.3.4. Identification of the binding site of quercetin on the BSA 2.3.4.1. Fluorescence spectroscopy. As described later in Section 3, Section 3.3.2, the experiments, carried out according to the procedures in Section 2.3.3, showed that the ratio of quercetin to BSA in their complex was 1:1. This information was used in the method as described below. Quercetin and BSA solutions were mixed thoroughly in a 1:1 ratio, and then the warfarin solution was added to the mixture. The fluorescence spectra (ex = 465 nm) of the quercetin–BSA and warfarin mixtures were obtained with the increasing concentration of warfarin (0–2.0 × 10−5 mol L−1 ). 2.3.4.2. UV–vis spectrometry. The concentrations of quercetin and BSA were kept at 1.65 × 10−5 and 1.50 × 10−5 mol L−1 , respectively, and the concentration of warfarin was varied from 0 to 5.89 × 10−5 mol L−1 . 2.4. Chemometrics 2.4.1. Multivariate curve resolution-alternating least squares (MCR-ALS) MCR-ALS is a soft modeling method, which has been shown to be a powerful tool for spectroscopic investigations of molecular complex formation processes [26,27]. The multivariate curve resolution method facilitates a bilinear decomposition of the experimental data matrix with the use of the following algebraic model: D = CS T + E

(1)

The goal of MCR-ALS is the decomposition of the data matrix, D, into the ‘true’ pure response profiles associated with the data variance in the rows and columns, and represented by matrices C (concentration) and ST (spectra), respectively. The procedure of MCR-ALS calculates the C and S in turn by least squares, and the iterative process is repeated if the model has not converged. In general, it is regarded that convergence has been achieved when the relative difference between the values from two consecutive iterations (lack-of-fit) fails the set threshold value (often 0.1%) or when a preselected number of iterations is exceeded. Lack-of-fit (LOF) is defined as the difference between the input data, D, and the data from the CST product obtained by MCR-ALS. This value is calculated according to the expression:



LOF =

ij

∗

(dij − dij )



ij

(dij )

2

1867

[29]. In the forward-step EFA, PCA is performed sequentially on a series of sub-matrices, which are constructed by successively adding spectra to the previous sub-matrix. When a singular value or factor becomes significant with respect to the noise level, it is considered that the appearance of an event has been detected. The backward-step EFA is initiated by starting repetitive PCAs with the last few spectra in the matrix, and systematically adding spectra in the reverse order. The resulting backward-EFA plot, maps the disappearance of significant factors. 3. Results and discussion 3.1. Voltammetric interaction of quercetin with BSA Cyclic voltammograms of quercetin in the presence of BSA (Fig. 1) showed that quercetin has an oxidation peak at 465 mV and a reduction peak at 430 mV (trace1, Fig. 1). The peak potential difference (ϕ) was 35 mV, which indicates a quasi-reversible, 2e− redox process (for a reversible electrode process, ϕ = 59.1/n) [30], and the ϕ1/2 was 447.5 mV. When BSA was added to the quercetin solution, the oxidation and the reduction peaks shifted towards high and low potentials, respectively, and a decrease of oxidation current was observed. Initially, the reduction peak current increased somewhat, but then it tended to decrease. In comparison, the oxidation peak changes were larger than those for the reduction peak. When the [BSA] increased, the decrease in peak current became increasingly large. At the [BSA] of 1.13 × 10−5 mol L−1 (trace 13, Fig. 1), the oxidation peak shifted to 520 mV, and the reduction one to about 400 mV. Thus, the peak potential difference was 120 mV, and the ϕ1/2 potential was 460 mV, which suggested that the process became more irreversible when BSA was added to the quercetin solution. These observations suggested that a quercetin–BSA complex was formed, but this complex did not appear to be electroactive [31]. This observation is consistent with the view that the quercetin–BSA interaction occurred between the most hydrophobic segment of the quercetin molecule (benzopyran4-one) [32] and the hydrophobic part of the BSA cavity. A study of the interaction between quercetin and HSA using molecular models found that the benzopyran-4-one moiety was located within the binding pocket of HSA, but the B-ring protruded from it and pointed toward the interface of IIB and IIIA domains [6]. The structure of BSA

0.5

(2)

where dij are the experimental values and dij * are the corresponding calculated values using the MCR-ALS decomposition. 2.4.2. Initial estimation of concentration profiles by EFA of the matrix D EFA has been applied in spectroscopic studies of multi-equilibria systems [28]. It is based on repetitive PCAs of sub-matrices of D

Fig. 1. Cyclic voltammograms of quercetin mixed with different concentrations of BSA. [quercetin] = 8.28 × 10−6 mol L−1 , [BSA] = 0, 9.38 × 10−7 , 1.88 × 10−6 , 2.81 × 10−6 , 3.75 × 10−6 , 4.69 × 10−6 , 5.62 × 10−6 , 6.56 × 10−6 , 7.5 × 10−6 , 8.44 × 10−6 , 9.38 × 10−6 , 1.03 × 10−5 and 1.13 × 10−5 mol L−1 (1–13).

1868

Y. Ni et al. / Spectrochimica Acta Part A 71 (2009) 1865–1872

is similar to HSA and a similar conclusion can be drawn. The electroactive parts of the quercetin i.e., the 3 - and 4 -OH groups, were embedded within the BSA structure, and this precluded its interaction at the electrode surface and therefore its participation in the redox reaction. Furthermore, the concentration of free quercetin at the electrode surface decreased during the course of the experiment, and hence, the peak current was also reduced. Consequently, the observed potential shifts of the oxidation and reduction peaks can be attributed to the changes of the molecular environment of quercetin as a result of its interaction with BSA. The oxidation of quercetin (QH2 ) in Tris–HCl buffer involves a two-electron process, which can be represented as follows:

where Q is the oxidized form of quercetin. The relationship between the formal potentials and equilibrium constants, for twoelectron redox process is given by Eq. (3) [33]: Eb␪ − Ef␪ = 0.0295 log(K2 /K1 ) Ef␪

(3)

Eb␪

where and are the formal potentials of the free and bound quercetin. Thus, for a shift of 12.5 mV (460–447.5 = 12.5 mV), K2 /K1 for QH2 /Q is 2.65, i.e., the reduced form is bound about 2.65 times more strongly than the oxidized form. Assuming that the interaction of quercetin and BSA produces only a single BSA/m-quercetin molecular complex, the binding number, m, and the binding constant, ˇ, can be determined as follows [34]: Given that: BSA + mquercetin  BSA − mquercetin [BSA − mquercetin] ˇ= [quercetin]m [BSA]

(4)

the relationships (5)–(8), and Eq. (9) may be derived. These relate the changes in the measured current, I, to the [quercetin]. From these equations, m and ˇ may be extracted. CBSA = [BSA] + [BSA − mquercetin]

(5)

I = k [BSA − mquercetin]

(6)

Imax = kCBSA

(7)

Imax − I = k[BSA]

(8)

and the final equation is:



log

I (Imax − I)



= log ˇ + m log[quercetin]

(9)

where I is the peak current change of the responses measured at the same concentration of quercetin in the absence and presence of BSA; Imax is the maximum peak current change. The [BSA] was kept constant at 6.0 × 10−6 mol L−1 and the concentration of the added quercetin was in the range of 4.96 × 10−6 to 3.47 × 10−5 mol L−1 . Thus, the resulting linear plot (r = 0.9946) gave an m value of 1.14 ± 0.06 and a ˇ of (4.2 ± 0.3) × 107 M−1 .

Fig. 2. Absorption spectra of quercetin in the presence [quercetin] = 1.65 × 10−5 mol L−1 , [BSA] = 1.50 × 10−5 mol L−1 .

of

BSA.

(300–400 nm) and Band II (240–280 nm). Band I has been associated with the cinnamoyl system (B + C rings, Table 1), and Band II with the benzoyl moiety, which is formed by the A + C ring [6,35]. In the UV–vis spectra of quercetin (Fig. 2), the band at 378 nm was assigned as the Band I of quercetin dissolved in the Tris–HCl buffer, and that at 267 nm as the Band II. The intensity of the spectral band of quercertin at 378 nm was significantly reduced in the presence of BSA (1:1 ratio), and a 10 nm red shift was noted. These observations indicate the presence of a specific interaction between quercetin and BSA. 3.3. Fluorescence measurements of the interaction between quercetin and BSA 3.3.1. Fluorescence quenching study When excited at 280 nm, BSA shows characteristic emission fluorescence with a maximum at 350 nm. This has been attributed to the tryptophan (Trp) residues when the BSA is present under pseudo-physiological conditions [36]. In the presence of quercetin at different concentrations, the fluorescence of BSA was quenched, and this spectral information was analysed by the well-known Stern–Volmer model [37]: F0 = 1 + KSV [Q] F

(10)

where F0 and F are the steady state fluorescence intensities in the absence and presence of quencher, respectively; KSV is the Stern–Volmer quenching constant and [Q] is the concentration of quencher (in this work – quercetin). Linear Stern–Volmer plots which can be expressed as F0 /F = 0.8806 + 0.072 [quercetin] (r = 0.9957) at 300 K and F0 /F = 0.8825 + 0.062 [quercetin] (r = 0.9946) at 310 K were obtained. The quenching constant, KSV, decreases with increasing temperature. This indicates that the probable quenching mechanism of quercetin–BSA binding is initiated by a static quenching procedure, which involves a fluorophore–quencher complex formation [38]. When small molecules bind independently to a set of equivalent sites on a macromolecule, the equilibrium between free and bound molecules is given by the equation:

 (F − F)  0

3.2. UV–vis spectra of quercetin in the absence and presence of BSA

log

The absorption spectra of flavonoids in the 240–450 nm range usually have two main bands commonly referred to as Band I

where Ka and n are the binding constant and the number of binding sites, respectively.

F

= log Ka + n log [quercetin]

(11)

Y. Ni et al. / Spectrochimica Acta Part A 71 (2009) 1865–1872

1869

Fig. 3. Fluorescence contour plots for quercetin–BSA.

This equation, when applied to the spectral data discussed above, yielded a linear plot of log{(F0 − F)/F}=7.53+ 1.11 log[quercetin] (r = 0.997), and Ka = (3.4 ± 0.3) × 107 M−1 and n = 1.11 ± 0.04, which are similar to the values obtained from the cyclic voltammetry studies (Section 3.1).

3.3.2. Fluorescence enhancement study The interaction between quercetin and BSA was further studied at high concentration under the same experimental conditions (see Section 2.3.3). The fluorescence contour plot (Fig. 3) indicates that suitable excitation and emission wavelengths occur at 465 and 525 nm, respectively. No emission fluorescence was observed for either BSA or quercetin solution alone, but when the two solutions were mixed, an emission peak appeared at 525 nm. These observations can be attributed to the formation of a fluorescent complex between quercetin and BSA. When this occurs, then on the one hand, the molecular rigidity of quercetin is strengthened, the structural plane is enlarged, and the fluorescence quantum yield is increased; but on the other hand, the hydrophobic interior of BSA becomes less polar than the buffer solution. Thus, the appearance of the fluorescence emission peak can be attributed to the suppression of knr (rate constant of non-radiative processes) and the reduced polarity of the environment [39]. The fluorescence intensity increased gradually with the increasing concentration of quercetin, indicating an interaction between quercetin and BSA. When the ratio of quercetin/BSA reached 1, the fluorescence intensity arrived at a stable level, suggesting that the complexation reaction was complete and a stable complex between quercetin and BSA was formed. When the BSA solution was added under constant concentration of quercetin, the observations were the same. This confirmed the formation of the 1:1 quercetin–BSA complex. These results support the conclusions from the UV–vis spectroscopic study, i.e., a quercertin–BSA complex was formed.

3.4. Binding of quercetin with BSA in the presence of site markers BSA has a large hydrophobic cavity that can accommodate two or more ligands. When two ligands bind to BSA simultaneously, two types of interaction can occur [21]:

Fig. 4. Absorption spectra of the interaction of quercetin, BSA and warfarin. (a) Absorption spectra – 250–450 nm range; (b) absorption spectra – 350–450 nm range. [quercetin] = 1.65 × 10−5 mol L−1 , [BSA] = 1.50 × 10−5 mol L−1 , [warfarin] = 0, 0.31, 0.62, 0.93, 1.24, 1.51, 1.86, 2.17, 2.48, 2.79, 3.10, 3.41, 3.72, 4.03, 4.34, 4.65, 4.96, 5.27, 5.58 and 5.89 (×10−5 mol L−1 ) (1–20).

(1) competitive binding +BSA

+L2

L1 −→ L1–BSA−→ L2–BSA + L1 (2) non-competitive binding

+BSA

+L2

L1 −→ L1–BSA−→L1–BSA–L2 From X-ray crystallographic studies, warfarin, a coumarin, which is structurally related to flavonoids, has been found to bind to the subdomain IIA (site I), while ibuprofen is considered to bind to subdomain IIIA (site II) [4]. In this work, when the site I marker, warfarin was added to the quercetin/BSA mixture, the observed 388 nm band associated with the quercetin–BSA complex (Fig. 4) gradually decreased in intensity and shifted towards the blue end of the spectrum. These effects were observed with increasing concentration of the added warfarin. The band intensity at 278 nm increased with the appearance of a peak at 309 nm. Warfarin does have a band at approximately 309 nm and this could contribute to the observed growth of the peak with increasing concentration of the compound. It would appear from the decrease in the 388 nm band that a ternary com-

1870

Y. Ni et al. / Spectrochimica Acta Part A 71 (2009) 1865–1872

Fig. 5. Fluorescence spectra of the interaction of quercetin, BSA and warfarin. [BSA] = 1.55 × 10−5 mol L−1 , [quercetin] = 2.10 × 10−5 mol L−1 , [warfarin] = 0, 1.0 × 10−6 , 2.0 × 10−6 , 3.0 × 10−6 , 4.0 × 10−6 , 5.0 × 10−6 , 6.0 × 10−6 , 7.0 × 10−6 , 8.0 × 10−6 , 9.0 × 10−6 ,1.0 × 10−5 , 1.1 × 10−5 , 1.2 × 10−5 , 1.3 × 10−5 , 1.4 × 10−5 , 1.5 × 10−5 , 1.6 × 10−5 , 1.7 × 10−5 , 1.8 × 10−5 and 1.9 × 10−5 mol L−1 (1–20).

plex of quercetin–BSA–warfarin does form but the bands are too broad and overlap significantly, which makes any detailed interpretation difficult. Thus, further clarification was sought with the use of the MCR-ALS method for spectral resolution (Section 3.5). Fluorescence measurements were carried out to obtain more information about the above interaction. The fluorescence intensity of the 525 nm band (Fig. 5) decreased with the addition of warfarin to the quercetin–BSA solution. This suggested either a displacement of quercetin from its complex or the formation of a ternary complex. It should be noted that under the experimental conditions in this work neither warfarin nor the BSA–warfarin complex fluoresced at or near 525 nm. Hence, if warfarin were to displace quercetin, the fluorescence at 525 nm should be quenched completely. However, this was not observed, and the fluorescence intensity appeared to have leveled off (inset, Fig. 5). Thus, this indicated the formation of a complex species with reduced fluorescence. These observations are consistent with the formation of a quercetin–BSA–warfarin ternary complex where two small molecules share, rather than compete, for the same binding location in site I. Further investigations of the binding site were carried out using of ibuprofen – a site II marker. When this substance was added to the quercetin–BSA complex, no obvious changes were observed in the UV–vis spectra. This demonstrated that ibuprofen and quercetin do not share the same binding site. They bind at their own binding sites independently.

Fig. 6. Comparison of the measured absorption spectra for the quercetin–BSA, warfarin and quercetin–BSA–warfarin with those obtained from ALS modeling. Solid line: resolved from ALS, and dashed line: measured spectra.

of the free warfarin and quercetin–BSA agreed well with their measured counterparts (dashed line, Fig. 6). And importantly, the spectrum of the quercetin–BSA–warfarin complex was also estimated. When the resolved spectra are compared with the measured ones (Fig. 4a), it can be seen that the band at 278 nm is composed of both the quercetin–BSA and the ternary complex absorptions with the latter band growing in intensity with the addition of warfarin. This accounts for the increase in the band at approximately 278 nm. The band at 388 nm decreases as the quercetin–BSA complex interacts with the warfarin, and the band at approximately 309 nm is a composite between the residual free warfarin and the new additional band due to the growing concentration of the ternary complex. Thus, the resolution of the overlapping spectra by the MCR-ALS method permits the rationalization of the complex overlapping spectrum. This argument also suggests that the binding of warfarin follows the non-competitive mechanism. From the equilibrium concentration profiles (Fig. 7), it was found that the added warfarin did not react completely with the quercetin–BSA complex to form the ternary complex. There was

3.5. Application of multivariate curve resolution-alternating least squares (MCR-ALS) to the two-way absorption spectra The absorption data matrix, D (20 objects × 201 variables) was processed with the aid of the SVD model, and the extracted eigenvalues were 37.27, 5.59, 0.190 and 0.02, indicating that there were arguably three significant factors for the prediction of the three separate chemical components in the system, i.e., the quercetin–BSA complex, warfarin and quercetin–BSA–warfarin ternary complex, but as noted previously the spectra of these three substances in mixtures overlapped considerably. Therefore, the estimation of concentrations of each component at equilibrium during the titration process was not possible by conventional methods. The MCR-ALS method was applied to resolve the two-way UV–vis spectra (Fig. 4a). The resolved spectra (solid line, Fig. 6)

Fig. 7. Equilibrium concentration profiles of warfarin, quercetin–BSA, and quercetin–BSA–warfarin extracted by the MCR-ALS chemometrics method.

Y. Ni et al. / Spectrochimica Acta Part A 71 (2009) 1865–1872

still some free warfarin present in solution. It can be seen that the concentration of quercetin–BSA–warfarin complex increased as the concentration of quercetin–BSA decreased. However, the amount of the warfarin initially added to the reaction mixture could not be accounted for by the sum of the free warfarin and that bound in the proposed 1:1 quercetin–BSA–warfarin complex. On the other hand, if a 1:2 complex, i.e., quercetin–BSA–(warfarin)2 , was considered, the mass balance was then reasonably achieved. This can be readily checked especially at the lower concentration of added warfarin (4–6 mol L−1 ). Thus, the interaction process can be described by the following reactions: quercetin + BSA → quercetin–BSA

(12)

quercetin–BSA + 2warfarin → quercetin–BSA–(warfarin)2

(13)

The expression for the equilibrium constant was: K = [quercetin– BSA–(warfarin)2 ]/[quercetin–BSA][warfarin]2 , and the value of K was estimated to be 1.4 × 1011 M−2 . X-ray crystallographic analysis and other methods have demonstrated that the binding of warfarin on BSA is in site I [4,40–42]. However, most studies by equilibrium dialysis [43,44] have supported the view that warfarin possess one high-affinity binding site (HAS) and several low-affinity binding sites (LAS) in the serum albumin protein. Both the HAS and LAS are located in site I for warfarin [44,45]. When a molar ratio of warfarin/BSA is less than 1, warfarin binds mainly in the HAS located in site I. When warfarin/BSA molar ratio is larger than 1, warfarin begins to bind to its LAS. According to the equilibrium concentration profiles (Fig. 7), the [free warfarin] was low when the (warfarin/BSA) < 1, but the fraction of free warfarin began to increase when the (warfarin/BSA) > 1. These observations are consistent with the initial strong binding of the warfarin to HAS followed by association with the LAS as the warfarin/BSA increases. From these results, the indications are that the binding site of quercetin is in site I in the BSA molecule. It is likely to be near to or overlapping the binding sites of warfarin, i.e., HAS and LAS but there is insufficient information at present to pinpoint the site(s). 4. Conclusion In this paper, UV–vis spectrometry, spectrofluorimetry, and cyclic voltammetry were applied to investigate the interaction of quercetin and BSA with the aid of the chemometrics method – MCR-ALS. • The results of cyclic voltammetry suggested that the quercetin–BSA complex is not electroactive. The binding constant of the reduced form of quercetin is 2.65 times stronger than the oxidized form. The values of binding number m, and constant ˇ were 1.14 ± 0.06 and (4.2 ± 0.3) × 107 M−1 , respectively. • The UV–vis spectra and fluorescence spectra demonstrated that a 1:1 quercetin–BSA fluorescent complex was formed, and the Stern–Volmer analysis of the fluorescence data gave a Ka (binding constant) = (3.4 ± 0.3) × 107 M−1 and n (binding number) = 1.11 ± 0.04. • Warfarin (site I marker) and ibuprofen (site II marker) were used to identify the binding site of quercetin on BSA. The results indicated that site I, rather than site II, was involved, and a ternary complex, quercetin–BSA–warfarin, was formed. The complex nature of the UV–vis spectra from the mixture of warfarin and quercetin–BSA and their product precluded the extraction of further information. • The application of the MCR-ALS for the resolution of the two-way UV–vis spectra produced the estimates of pure spectra of warfarin and the quercetin–BSA complex, which were in good agreement

1871

with the measured responses. Additionally and importantly, the spectrum of the interaction product, quercetin–BSA–warfarin, was extracted. Comparison of these extracted spectra provided a rationalization for the spectral changes observed during the addition of warfarin. Equilibrium concentration profiles were also obtained. From these profiles, the ratio between quercetin–BSA and warfarin was found to be 1:2 rather than 1:1, and interestingly, the interaction product was better represented by the formula quercetin–BSA–(warfarin)2 . Both the HAS and the LAS binding sites were invoked to rationalize this result. Acknowledgements The authors gratefully acknowledge the financial support of this study by the National Science Foundation of China (No. 20562009), the State Key Laboratory of Chemo/Biosensing and Chemometrics of Hunan University (No. 2005-22), the Jiangxi Province Natural Science Foundation (No. 0620041), and the program for Changjiang Scholars and Innovative Research Team in Universities (No. IRT0540). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36]

K.J. Fehske, W.E. Muller, U. Wollert, Biochem. Pharmacol. 30 (1981) 687–692. J.J. Vallner, J. Pharm. Sci. 66 (1977) 447–465. D.C. Carter, J.X. Ho, Adv. Protein Chem. 45 (1994) 153–203. X.M. He, D.C. Carter, Nature 358 (1992) 209–215. D.W. Lamson, M.S. Brignall, Altern. Med. Rev. 5 (2000) 196–208. F. Zsila, Z. Bikadi, M. Simonyi, Biochem. Pharmacol. 65 (2005) 447–456. C.D. Kanakis, P.A. Tarantilis, M.G. Polissiou, S. Diamantoglou, H.A. Tajmir-Riahi, J. Mol. Struct. 798 (2006) 69–74. O. Dangles, C. Dufour, C. Manach, C. Morand, C. Remesy, Methods Enzymol. 335 (2001) 319–333. C. Manach, C. Morand, O. Texier, M.L. Favier, G. Agullo, C. Demigne, F. Regerat, C. Remesy, J. Nutr. 125 (1995) 1911–1922. J.B. Harborne, Flavonoids in the environment: structure-activity relationships, in: V. Cody, E. Middleton, J.B. Harborne, A. Beretz (Eds.), Plant Flavonoids in Biology and Medicine. II. Biochemical, Cellular, and Medicinal properties, Alan R Liss, New York, 1988, pp. 17–27. G. Rusak, H.O. Gutzeit, J.L. Mullwer, Nutr. Res. 25 (2005) 141–153. J.C. Lee, J. Kim, J.K. Park, G.H. Chung, Y.S. Jang, Exp. Cell Res. 291 (2003) 386–397. M. Hultberg, A. Isaksson, A. Andersson, B. Hultberg, Clin. Biochem. 39 (2006) 160–163. O.S. Wolfbeis, in: S.G. Schulman (Ed.), Molecular Luminescence Spectroscopy: Methods and Applications, Part-I, Wiley, New York, 1985, p. 167. B. Sengupta, J. Guharay, P.K. Sengupta, J. Surf. Sci. Technol. 14 (1998) 150–156. J. Guharay, B. Sengupta, P.K. Sengupta, Proteins Struct. Funct. Genet. 43 (2001) 75–81. R. Simard, P.A. Zunszain, J.A. Hamilton, S. Curry, J. Mol. Biol. 361 (2006) 336–351. G. Sudlow, D. Birkett, D. Wade, Mol. Pharmacol. 11 (1975) 824–832. V.T.G. Chuang, A. Kuniyasu, H. Nakayama, Y. Matsushita, S. Hirono, M.K. Otagiri, Biochim. Biophys. Acta 1434 (1999) 18–30. A. Angelakou, G. Valsami, P. Macheras, M. Koupparis, Eur. J. Pharm. Sci. 9 (1999) 123–130. C. Dufour, O. Dangles, Biochim. Biophys. Acta 1721 (2005) 164–173. S.Y. Bi, L. Ding, Y. Tian, D.Q. Song, X. Zhou, X. Liu, H.Q. Zhang, J. Mol. Struct. 703 (2004) 37–45. B. Sengupta, P.K. Sengupta, Biochem. Biophys. Res. Commun. 299 (2002) 400–403. Y.N. Ni, S.J. Su, S. Kokot, Anal. Chim. Acta 580 (2006) 206–215. T. Peter, All About Albumin, Biochemistry Genetics and Medical Applications, Academic Press, San Diego, CA, 1996, pp. 9–75. R. Tauler, A. Izquierdo-Ridorsa, E. Casassas, Chemom. Intell. Lab. Syst. 18 (1993) 293–300. R. Tauler, A.K. Smilde, B.R. Kowalski, J. Chemom. 9 (1995) 31–58. R. Tauler, B.R. Kowalski, S. Flemming, Anal. Chem. 65 (1993) 2040–2047. H. Gampp, M. Maeder, C.J. Meyer, A.D. Zuberbuhler, Talanta 32 (1985) 1133–1139. J. Wang, Analytical Electrochemistry, second ed., Wiley, New York, 2000, p31. Y.H. Wu, X.B. Ji, S.S. Hu, Bioelectrochemistry 64 (2004) 91–97. Z.W. Zhu, C. Li, N.Q. Li, Microchem. J. 71 (2002) 57–63. M.T. Carter, M. Rodriguez, A.J. Bard, J. Am. Chem. Soc. 111 (1989) 8901–8909. W. Sun, K. Jiao, Talanta 56 (2002) 1073–1080. T.J. Mabry, K.R. Markham, M.B. Thomas, The Systematic Identification of Flavonoids, Springer-Verlag, New York, 1970, pp. 41–61. F. Moreno, J. Gonzalez-Jimenez, Chem. Biol. Interact. 121 (1999) 237–252.

1872

Y. Ni et al. / Spectrochimica Acta Part A 71 (2009) 1865–1872

[37] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum Press, New York, 1983, p. 257. [38] H.W. Zhao, M. Ge, Z.X. Zhang, W.F. Wang, G.Z. Wu, Spectrochim. Acta, A 65 (2006) 811–817. [39] S. Deepa, A.K. Mishra, J. Pharm. Biomed. Anal. 38 (2005) 556–563. [40] T.C. Pinkerton, K.A. Koeplinger, Anal. Biochem. 62 (1990) 2114–2122. [41] F.J. Diana, K. Veronich, A.L. Kapoor, J. Pharm. Sci. 78 (1989) 195–199.

[42] B. Loun, D.S. Hage, Anal. Chem. 66 (1994) 3814–3822. [43] K. Yamasaki, T. Maruyama, U. Kragh-Hansen, M. Otagiri, Biochim. Biophys. Acta 1295 (1996) 147–157. [44] T. Kosa, T. Maruyama, M. Otagiri, Pharmaceut. Res. 14 (1997) 1607– 1612. [45] S. Baroni, M. Mattu, A. Vannini, R. Cipollone, S. Aime, P. Ascenzi, M. Fasano, Eur. J. Biochem. 268 (2001) 6214–6220.