Interactions of different polyphenols with bovine serum albumin using fluorescence quenching and molecular docking

Food Chemistry 135 (2012) 2418–2424

Contents lists available at SciVerse ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Rapid Communication

Interactions of different polyphenols with bovine serum albumin using fluorescence quenching and molecular docking Mihaela Skrt a, Evgen Benedik a, Cˇrtomir Podlipnik b, Nataša Poklar Ulrih a,c,⇑ a

Biotechnical Faculty, University of Ljubljana, Ljubljana, Slovenia Faculty of Chemistry and Chemical Technology, University of Ljubljana, Ljubljana, Slovenia c Centre of Excellence for Integrated Approaches in Chemistry and Biology of Proteins (CipKeBiP), Jamova 39, Ljubljana, Slovenia b

a r t i c l e

i n f o

Article history: Available online 14 July 2012 Keywords: Phenolic acids Flavones Catechines Molecular docking Fluorescence emission spectrometry Bovine serum albumin

a b s t r a c t Polyphenols are responsible for the major organoleptic characteristics of plant-derived foods and beverages. Here, we investigated the binding of several polyphenols to bovine serum albumin (BSA) at pH 7.5 and 25 °C: catechins [()-epigallocatechin-3-gallate, ()-epigallocatechin, ()-epicatechin-3-gallate], flavones (kaempferol, kaempferol-3-glucoside, quercetin, naringenin) and hydroxycinnamic acids (rosmarinic acid, caffeic acid, p-coumaric acid). Fluorescence emission spectrometry and molecular docking were applied to compare experimentally determined binding parameters with molecular modelling. Among these polyphenols, ()-epicatechin-3-gallate showed the highest Stern–Volmer modified quenching constant, followed by ()-epigallocatechin-3-gallate. Similarly, ()-epicatechin-3-gallate had the highest effect on the Circular Dichroic spectrum of BSA, while the changes induced by other polyphenols were negligible. Molecular docking predicted high binding energies for ()-epicatechin-3-gallate and ()-epigallocatechin-3-gallate for the binding site on BSA near Trp213. Our data reveal that the polyphenol structures significantly affect the binding process: the binding affinity generally decreases with glycosylation and reduced numbers of hydroxyl groups on the second aromatic ring. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Polyphenols form a diverse group of aromatic secondary plant metabolites that are widely distributed throughout the plant kingdom (Vermerris & Nicholson, 2008). Due to their almost universal distribution, they are an integral part of the human diet, and they are responsible for the major organoleptic characteristics of plantderived foods and beverages (e.g. colour and taste) (Cheynier, 2005). Beneficial health effects of polyphenols are believed to be due to their antioxidative properties (Villañoa, Fernández-Pachóna, Moyáb, Troncosoa, & García-Parrilla, 2007). Indeed, they have been shown to induce antitumor, antibacterial, antimutagenic and anticancerogenic effects (Kroll, Rawel, & Rohn, 2003), although many studies have also demonstrated their harmful effects, especially when applied at high concentrations (Arts, Hollman, de Mesquita, Feskens, & Kromhout, 2001; Lambert, Sang, & Yang, 2007). These properties are partially attributed to the interactions between the

Abbreviations: BSA, bovine serum albumin; epicatechin-G, ()-epicatechin-3gallate; epigallocatechin, ()-epigallocatechin; epigallocatechin-G, ()-epigallocatechin-3-gallate; kaempferol-glu, kaempferol-3-glucoside. ⇑ Corresponding author at: Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, 1000 Ljubljana, Slovenia. Tel.: +386 1 3230780; fax: +386 1 2566296. E-mail address: [email protected] (N.P. Ulrih). 0308-8146/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2012.06.114

polyphenols and biomolecules in the body, such as proteins and lipids (Ishii et al., 2009; Nozaki, Kimura, Ito, & Hatano, 2009). However, a complete understanding of the structural and antioxidative properties of such polyphenols, in their interactions with biomolecules, has not yet been obtained. Polyphenol interactions with different proteins (Dufour & Dangles, 2005; He, Lv, & Yao, 2007; Roy et al., 2010; Soares, Mateus, & de Freitas, 2007; Wang et al., 2008), and bovine serum albumin (BSA) have been the subject of many studies of these interactions (Dufour & Dangles, 2005; Liu, Qi, & Li, 2010; Papadopoulou, Green, & Frazier, 2005; Rawel, Meidtner, & Kroll, 2005; Shi et al., 2010). Serum albumins provide a good insight into our understanding of the interactions of polyphenols under physiological conditions. BSA has been extensively used for such investigations because of its intrinsic structural similarity to human serum albumin. BSA has a wide range of physiological functions that are associated with binding, transport and distribution of biologically active compounds. The binding of active compounds to serum albumins influences the effectiveness of the biologically active compounds, as well as the activity of the serum albumins. Structurally, BSA is a single-chain polypeptide of 583 amino acids that is folded into three homologous domains, each of which contains two subdomains (A and B) (Carter & Ho, 1994). The BSA polypeptide contains two tryptophan residues (Trp134 and Trp213), and in the folded

2419

M. Skrt et al. / Food Chemistry 135 (2012) 2418–2424

BSA molecule, Trp134 is located on the surface, and Trp213 is located in a hydrophobic pocket (Kragh-Hansen, 1981). In the present study, the interactions with BSA were investigated for selected catechins [()-epigallocatechin-3-gallate (epigallocatechin-G), ()-epigallocatechin (epigallocatechin), and ()epicatechin-3-gallate (epicatechin-G)], flavones [kaempferol-3glucoside (kaempferol-glu), kaempferol, quercetin, and naringenin], and hydroxycinnamic acids (rosmarinic acid, caffeic acid, and p-coumaric acid) (Table 1), using fluorescence emission spectrophotometry and molecular docking. The molecular docking was performed to directly compare these predicted data with the experimental data, and to determine the potential modes of action of these polyphenols. The combined data will thus help us to understand the structure–activity relationships between these selected polyphenols and their binding site on the BSA molecule.

2. Materials and methods 2.1. Reagents Essentially globulin-free bovine serum albumin (BSA; P99%, lyophilized powder) was from Sigma–Aldrich (Munich, Germany). Epigallocatechin-G, epigallocatechin, epicatechin-G, kaempferolglu, and kaempferol were from Extrasynthese (Lyon, France), quercetin, naringenin, and caffeic acid from Sigma–Aldrich (Munich, Germany), rosmarinic acid from Vitiva (Markovci, Slovenia), and p-coumaric acid from Merck (Frankfurt, Germany). All of the chemicals used were of analytical grade. The BSA was dissolved in 20 mM HEPES, pH 7.5, and further purified by dialysis against 20 mM HEPES, pH 7.5, for 36 h at

4 °C. The molar concentration of BSA in aqueous HEPES solution was determined spectrophotometrically (Hewlett–Packard UV– VIS spectrophotometer, model 8453) using e280 (BSA) = 42925 M1 cm1 (calculated with ProtParam from ExPASy) and MW = 66433 g mol1 at 25 °C. The 1 mM stock solutions of each of the polyphenols were prepared in 96% ethanol, except for quercetin, which was prepared as 1 mM in pure methanol. The working solutions of the polyphenols (c = 0.1 mM) were prepared by dilution of these stock solutions in 20 mM HEPES, pH 7.5.

2.2. Fluorescence emission spectrophotometry 2.2.1. General Fluorescence spectra were recorded with a Cary Eclipse spectrofluorometer (Varian, Australia) equipped with an electro-thermal temperature controller, using a 1 cm path length quartz cuvette. Slit widths with a nominal band-pass of 5 nm were used for both excitation and emission. The intrinsic fluorescence emission spectra of BSA (c = 0.082 ± 0.001 mg ml1) were recorded from 290 nm to 450 nm, as a function of increasing concentrations of the polyphenols (titration experiment). The maximum excitation wavelengths kex and kem for BSA were observed at 280 nm and 339 nm, respectively. Based on these results the excitation wavelength of 280 nm was used to follow the BSA fluorescence. The emission spectra of BSA in the absence and presence of the polyphenols, as corrected for the solvent blank, were normalized for the dilution and corrected for the photomultiplier-tube response. The wavelengths at maximum emission intensity, kmax, and the fluorescence intensity at 339 nm were determined. All of the experiments were conducted at 25 °C in 20 mM HEPES, pH 7.5,

Table 1 Structures of the polyphenols used in this study. Catechins

Flavones

Hydroxycinnamic acids

Epigallocatechin-3-gallate R6, R8, R0 2 = H R5, R7, R0 3, R0 4, R0 5 = OH

Kaempferol-3-glucoside R1, R6, R8, R0 2, R0 3, R0 5, R0 ’6 = H R5, R7, R0 4 = OH

Rosmarinic acid

R3 =

Epigallocatechin R6, R8, R0 2 = H R3, R5, R7, R0 3, R0 4, R0 5 = OH Epicatechin-3-gallate R6, R8, R0 2, R0 5 = H R5, R7, R0 3, R0 4 = OH

R1 =

R3 =

R2 = OH

Kaempferol R1, R6, R8, R0 2, R0 3, R0 5, R0 6 = H R3, R5, R7, R0 4 = OH Quercetin R1, R6, R8, R0 2, R0 3, R0 5, R0 6 = H R3, R5, R7, R0 3, R0 4 = OH

Caffeic acid R1 = H; R2 = OH

R3 =

Naringenin R1, R6, R8, R0 2, R0 3, R0 5, R0 6 = H R5, R7, R0 4 = OH

p-Coumaric acid R1 = H; R2 = H

2420

M. Skrt et al. / Food Chemistry 135 (2012) 2418–2424

with two independent experiments, each carried out in triplicate. The data are expressed as means ± standard deviations. 2.2.2. Fluorescence quenching parameters The fluorescence emission intensities for BSA, measured at 339 nm as a function of increasing concentrations of the polyphenols, were used for the construction of the binding profiles. Fluorescence quenching is the measure of the decrease of the quantum yield of the fluorescence induced by a variety of molecular interactions with a quencher molecule. Fluorescence quenching can be dynamic, resulting from collisional encounters between the fluorophore and the quencher, or static, resulting from the formation of a ground state between the fluorophore (protein) and the quencher (Lakowicz, 2006). The fluorescence quenching is described by the linear Stern–Volmer equation:

Fo ¼ 1 þ kq so ½Q  ¼ 1 þ K SV ½Q  F

ð1Þ

In this equation, Fo and F are the fluorescence intensities in the absence and presence of the quencher, respectively, kq is the bimolecular quenching constant, so is the lifetime of the fluorescence in the absence of the quencher, and [Q] is the concentration of the quencher. The Stern–Volmer quenching constant, KSV, is given by kqso. A linear Stern–Volmer plot is generally indicative of a single class of fluorophore, all equally accessible to the quencher. If two fluorophore populations are present and one class is not accessible to the quencher, then the Stern–Volmer plots deviate from linearity. This is frequently seen for quenching of tryptophan fluorescence in proteins by polar or charged quenchers. These molecules do not rapidly penetrate the hydrophobic interior of proteins, and only those tryptophan residues on the surface of the protein are quenched. BSA has two tryptophan residues that have intrinsic fluorescence in addition to 20 tyrosine residues. It has been shown that tyrosine contribution to BSA fluorescence is quite small, even at the excitation at 280 nm (Lakowicz & Weber, 1973). Trp213 is located within a hydrophobic binding pocket, and Trp134 is located on the surface in the hydrophilic region of the molecule (Peters, 1985; Tayeh, Rungassamy, & Albani, 2009). The modified Stern–Volmer equation was applied here to analyze the fluorescence quenching:

Fo 1 1 1  ¼ þ F o  F fa ½Q  fa K A

ð2Þ

where Fo and F are the fluorescence intensities in the absence and presence of the quencher, respectively, fa is the fraction of fluorophore accessible to the quencher, [Q] is the concentration of the quencher, and KA is the modified Stern–Volmer quenching constant, which is very close to the binding constant. The plot of Fo/(Fo  F) versus 1/[Q] yields 1/fa as the intercept and 1/faKA as the slope. Thus, from the intercept and slope, the fa and KA can be obtained (Lakowicz, 2006). 2.3. Circular dichroism (CD) measurements CD spectra were obtained with an AppliedPhotophysics Chirascan™ Spectrometer (United Kingdom), using protein concentrations of 0.5 mg ml1 at pH 7.5, 20 mM HEPES buffer. Spectra were recorded at 25 °C in a 1 mm pathlength quartz cell (Hellma, Jena, Germany) from 260 to 208 with a step size of 0.5 nm, band width of 1.5 nm, and an averaging time of 1 s. For all spectra, an average of three scans was obtained. CD spectra of appropriate buffers were recorded and subtracted from the BSA or BSA:polyphenols (molar ratio 1:1) spectra. The mean residue ellipticity, Hk, was calculated by using the relation:

½Hk ¼

Mo Hk 100  c  l

ð3Þ

in which Mo is the mean residue molar mass (113.9 g mol1 for BSA), Hk is the measured ellipticity in degrees, c is the concentration in gml1, and l is the path length in decimetres. [H]k was expressed in deg cm2 dmol1. 2.4. Homology modelling and docking As there is no relevant X-ray structure of BSA in the protein databases (PDBs), we built a high-resolution structure of the BSA protein from its amino-acid sequence, using the standard set-up of the YASARA (Yet Another Scientific Artificial Reality Application) structure-homology modelling module (Vensellar et al., 2010). YASARA is an easy-to-use, reliable, universal package for molecular graphics, molecular modelling and molecular dynamics (Krieger, Koraimann, & Vriend, 2002) (http://www.yasara.com). The amino-acid sequence for the BSA protein (identified as P02769 in UniProtKB/ Swis-Prot) was used as the initial information for building a homology model, with the whole sequence used, including the 24 amino-acid residues of the signal sequence, for a total 607 amino acids. We used two different protocols for the docking. In the first protocol, the multi-step Schrödinger’s Protein Preparation Wizard (Schrödinger Suite, 2011) was used for the final preparation of the receptor site that was used as a target for GlideXP (Friesner et al., 2002, 2006; Halgren et al., 2004) molecular docking of the polyphenols listed in Table 1. As an alternative, we used Autodock version 4.0 (Moris et al., 1998) incorporated into the YASARA Structure, for docking the polyphenols into the BSA binding sites. The FoldX repairing module (Van Durme et al., 2011) was used for preparing a homology structure of BSA for docking. 3. Results and discussion 3.1. Fluorescence emission spectrometry 3.1.1. General The interactions between the polyphenols and BSA were followed by the changes in the fluorescence emission spectra of BSA in the absence and presence of the polyphenols. BSA has two trypthophan residues, Trp213 and Trp134, which show intrinsic fluorescence (Peters, 1985; Tayeh et al., 2009). Changes in the emission spectra of BSA are likely to occur in response to polyphenol binding. These will be partially due to conformational changes of the BSA protein. The fluorescence emission intensity of BSA decreases with the successive addition of epigallocatechin-G, epigallocatechin, and epicatechin-G, which indicates that these polyphenols can quench the intrinsic fluorescence of BSA (Supplementary data 1A–C). This effect was also seen with the other tested polyphenols (data not shown). The emission maximum in the spectrum of BSA was at 339 nm, which suggests that tryptophan residues are partially exposed on the surface of the BSA protein. Trp134 is located on the surface of the BSA molecule, while Trp213 is located within a hydrophobic binding pocket of BSA (Kragh-Hansen, 1981). A shift in kmax from 339 nm to 362 nm and to 356 nm towards higher wavelengths (red shift) was observed when BSA was titrated with epigallocatechin-G and epicatechin-G, respectively, which indicated that the tryptophan surroundings became more polar. This red shift in kmax of more than 23 nm is likely to indicate the presence of epigallocatechinG and epicatechin-G in the vicinity of Trp213. A shift in kmax from 339 nm to 330 nm, so towards lower wavelengths (a blue shift), was observed when BSA was titrated with the two flavones, kaempferol and quercetin, which indicates that the tryptophan surroundings became less polar. Titration of BSA with the other studied polyphenols, epigallocatechin, kaempferol-glu, naringenin,

2421

M. Skrt et al. / Food Chemistry 135 (2012) 2418–2424

rosmarinic acid, caffeic acid and p-coumaric acid, did not shift the kmax in the emission spectra of BSA significantly. The change in fluorescence intensity of Trp213 and Trp134 in the presence of some of these polyphenols might arise as a direct quenching or as a result of BSA conformational changes induced by the polyphenol–BSA complex (Bourassa, Kanakis, Tarantilis, Pollissiou, & Tajmir-Riahi, 2010). The observed spectral changes and the shift in the position of kmax of BSA are likely to reflect direct or indirect interactions of these polyphenols with the hydrophobic and hydrophilic regions located in the vicinity of Trp213, and to reflect induced conformational changes. The lack of a spectral shift is indicative that the tryptophan is not exposed to any change in polarity. 3.1.2. Fluorescence quenching parameters The order of the Trp quenching efficiencies of these polyphenols, determined from fluorescence emission intensity measurements at 339 nm is: epicatechin-G ffi epigallocatechinG > quercetin ffi kaempferol ffi naringenin ffi rosmarinic acid > pcoumaric acid ffi kaempferol-glu ffi caffeic acid ffi epigallocatechin (Supplementary data 2). Fluorescence quenching indicates the decrease in the quantum yield of the fluorescence that is induced by a variety of molecular interactions with a quencher molecule. Fluorescence quenching can be dynamic, resulting from collisional encounters between the fluorophore and the quencher, or static, resulting from the formation of a ground state complex between the fluorophore and the quencher. In our case, the bimolecular quenching constant, kq, that was obtained from the Stern–Volmer Eq. (1), using so = 5 ns for the lifetime of the fluorophore in BSA (Papadopoulou et al., 2005), is much larger than 2  1010 M1 s1 (Table 2), which indicates that the quenching is static in nature (Ghuman et al., 2005). The collision quenching constants for the interaction of epigallocatechin-G and epicatechin-G with BSA were 18.4 and 27.2  1013 M1 s1, respectively (Table 2). For kaempferol, quercetin, naringenin, and rosmarinic acid, the collision quenching constants were in the range from 8.0 to 5.4  1013 M1 s1. The lowest kq was seen for the group of polyphenols that included epigallocatechin, kaempferol-glu, caffeic acid, and p-coumaric acid (Table 2). The calculated quenching constant for naringenin was 5.40  1013 M1 s1 (Table 2), which is comparable with the kq of 3.36  1013 M1 s1 obtained for the interaction of naringenin with BSA in a 20 mM phosphate buffer, pH 7.0 (Ghuman et al., 2005). The modified Stern–Volmer constants, KA, as obtained from the plot of Fo/(Fo–F) versus 1/[Q] (Fig. 1) for the catechins epigallocatechin-G, epigallocatechin, and epicatechin-G, were 14.0 (±0.5)  105 M1, 1.4 (±0.3)  105 M1, and 15.0 (±0.3)  105 M1, respectively (Table 2). The highest KA among the polyphenols studied here was obtained for epigalloca-

Fig. 1. Fo/(Fo  F) as a function of 1/polyphenol concentration for the calculation of the modified Stern–Volmer quenching constant, KA, as the ratio of the intercept and the slope for the BSA–polyphenol complexes. Fo and F are fluorescence emission intensities of BSA at 339 nm in the absence and presence of the polyphenols, respectively. For the polyphenols, the following abbreviations were used: epigallocatechin-G (EGCG), epigallocatechin (EGC), epicatechin-G (ECG), kaempferol-glu (KAEMP-G), kaempferol (KAEMP), quercetin (QUER), naringenin (NAR), rosmarinic acid (ROA), caffeic acid (CAFF) and p-coumaric acid (pCOUM).

techin-G and epicatechin-G. The lowest KA was obtained for epigallocatechin and kaempferol-glu (Table 2). Based on these data, as listed in Table 2, we can conclude that: (i) the modified Stern–Volmer constant, KA, is highest for esterified catechins (epigallocatechin-G and epicatechin-G) and lowest for epigallocatechin, among all of the polyphenols studied here; and (ii) glycosylated kaempferol (kaempferol-glu) has a lower KA than has kaempferol. These observations are consistent with the findings of Liu et al. (2010), who reported that glycosylation of flavonoids decreases their binding constants to BSA (e.g. calycosin-glucoside and formonetin-glucoside have lower binding constants, of 1.9  104 M1 and 2.2  104 M1, respectively, than has quercetin, 10.2  104 M1). Table 2 also gives the fractions of the fluorophore accessible to the quencher, the fa. The fa values for epigallocatechin (0.40), kaempferol-glu (0.55), caffeic acid (0.41), and p-coumaric acid (0.45) are the lowest, which suggests that these polyphenols only quench the Trp134 residue on the surface of the BSA protein, and consequently they have the lowest binding affinities for BSA (Table 2). In addition to the previous conclusions given above, our results also confirm that the structurally simpler hydroxycinnamic acids, such as caffeic acid and p-coumaric acid (Table 1), have very

Table 2 Stern–Volmer (KSV), bimolecular quenching constant (kq), modified Stern–Volmer constant (KA) and the fraction of fluorophore accessible to the quencher (fa) for the interactions of the polyphenols with BSA, at 25 °C and pH 7.5. Polyphenols

Ksv ( 103) [M]1

kq ( 1012) [M  s]1

KA ( 105) [M1]

fa

Catechins ()-Epigallocatechin-3-gallate ()-Epigallocatechin ()-Epicatechin-3-gallate

922 ± 3 49.2 ± 0.2 1361 ± 13

184 ± 0.6 9.8 ± 0.1 272 ± 3

14.0 ± 0.5 1.4 ± 0.3 15.0 ± 0.3

0.98 0.40 0.90

Flavones Kaempferol-3-glucoside Kaempferol Quercetin Naringenin

68.0 ± 0.1 354 ± 0.1 400 ± 1 294 ± 0.2

13.6 ± 0.1 70.0 ± 0.1 80.0 ± 0.3 58.0 ± 0.1

1.8 ± 0.1 3.5 ± 0.4 3.4 ± 0.2 3.1 ± 0.1

0.55 1.0 1.1 0.99

Hydroxycinnamic acids Rosmarinic acid Caffeic acid p-Coumaric acid

265 ± 0.1 63.3 ± 0.5 83 ± 2

54.0 ± 0.1 12.6 ± 0.1 16.6 ± 0.4

3.2 ± 0.1 3.0 ± 0.1 4.0 ± 0.3

0.92 0.41 0.45

2422

M. Skrt et al. / Food Chemistry 135 (2012) 2418–2424

Table 3 GlideXP and Autodock docking scores for the polyphenols with BSA, according to the three models. Polyphenols

Autodock scores (kJ/mol)

Glide XP scores (kJ/mol)

Model 1

Model 2

Model 3

Model 1

Model 2

Model 3

Catechins ()-Epigallocatechin-3-gallate ()-Epigallocatechin ()-Epicatechin-3-gallate

48.5 43.5 47.3

50.6 46.9 52.3

56.5 43.1 54.4

37.7 31.8 34.7

50.6 45.2 49.0

40.2 27.6 35.1

Flavones Kaempferol-3-glucoside Kaempferol Quercetin Naringenin

46.9 39.3 41.8 40.1

43.9 41.0 42.3 38.1

44.4 41.0 41.8 37.7

30.5 22.6 37.7 34.3

48.5 – 42.3 37.6

36.4 33.1 35.1 34.3

Hydroxycinnamic acids Rosmarinic acid Caffeic acid p-Coumaric acid

48.5 31.4 30.1

53.1 33.5 31.0

43.9 28.5 27.6

32.2 38.1 36.8

35.1 33.1 25.1

33.9 19.2 21.8

Fig. 2. Docking orientations of epicatechin-G (A) and rosmarinic acid (B) for binding site 1 of BSA.

low binding affinities for BSA. All of the other polyphenols studied here have fa values greater than 0.90 (Table 2). 3.2. Circular dichroism Changes in molar ellipticity of BSA induced by binding of polphenols were followed by CD (circular dichroism) measurements in the far-UV CD range (260 nm to 208 nm). All spectra were measured at 25 °C. CD spectra of BSA measured in the far-UV CD range reflect the changes in the secondary structure. The helicity of BSA has been estimated to be 66% (Moriyama et al., 2008). In the presence of polyphenols, at molar ratio of 1–1, the molar ellipticity of BSA did not change significantly. The highest effect on the CD spectra of BSA was by epicatechin-G. The slight increase in the intensity of the CD spectrum of BSA in the presence of epicatechin-G, indicated that the binding induced small conformational changes (increase in the amount of a-helix structure). Negligible increase in the helicity of BSA was observed for epigallocatechin-G, epigallocatechin and other studied polyphenols (data not shown). 3.3. Homology modelling and docking The homology model of BSA was built from a high resolution template of the X-ray structure of human serum albumin, identified as 2BXK (Ghuman et al., 2005) in the PDBs, and it was selected for further investigations. We observed that this model includes the docking of two compounds: azapropazon and indomethacin. Indomethacin is located in the proximity of residue Trp213. This information helped us to identify where the binding site for these polyphenols is located. Therefore, on the basis that the binding site

of these polyphenols lies in the proximity of one of the tryptophan residues, three different interaction grids for docking were prepared with Schrodinger’s GlideGrid. The extensions of all three of these grids were 20 Å from: the centre of Trp134 (Model 1), the centre Trp213 (Model 2), and the gravitational centre of indomethacin and azopropazon (Model 3). These polyphenols were prepared for docking using Schrodinger’s LigPrep (Ligand Preparation) module, version 2.3, with default settings. Glide XP, version 5.0 (Friesner et al., 2004, 2006; Halgren et al., 2004), was used for docking these polyphenols to all three of the models of BSA. The Autodock grids for Model 1 and Model 2 were generated with an extension of 15 Å from Trp134 and Trp213, respectively. The grids for Model 3 were generated with an extension of 5 Å around the indomethacin and azopropazon. In all three cases, the spacing between the grid points was 0.375 Å. The Lamarckian genetic algorithm was used to search for the optimal orientation of the ligand within the binding site. Fifty independent docking runs were carried out for a convergence check of the docking results. The conformations were clustered, based on the RMSD (5.00 Å). The docking scores obtained with GlideXP and Autodock 4.0 (Moris et al., 1998) are given in Table 3, from which it can be seen that the binding sites in the proximity of Trp213 used for Model 2 and Model 3 are generally more favoured for the binding of these polyphenols than for Model 1, which was defined with Trp134. Epigallocatechin-G, epicatechin-G, and rosmarinic acid would be expected to be the best binders to BSA according to the Autodock results for Model 2, while the best experimental data were obtained for epigallocatechin-G and epicatechin-G using Model 3. The best orientations obtained with Autodock 4.0 for epicatechin-G and rosmarinic acid for the binding

M. Skrt et al. / Food Chemistry 135 (2012) 2418–2424

2423

Fig. 3. Docking orientations of epicatechin-G (A) and rosmarinic acid (B) for binding site 2 of BSA.

site 1 of BSA are shown in Fig. 2A and B, respectively. Two completely different binding modes are suggested by Autodock for epicatechin-G and rosmarinic acid. However, the orientation of epigallocatechin-G is very similar to that of epigallocatechin. The best Autodock orientations for epigallocatechin and rosmarinic acid for the binding site 2 of BSA are shown in Fig. 3A and B, respectively. From Fig. 3, it can be seen that epicatechin-G occupies the upper pocket of the binding site 2 (Fig. 3A), and it is in close contact (for p–p stacking) with Trp213; this binding is also supported by three hydrogen bonds (Ala207, 2  Asp450). In contrast, rosmarinic acid binds to the bottom part of the binding site 2 (Fig. 3B), and its interaction with Trp213 on the surface is weak, with six hydrogen bonds seen between rosmarinic acid and the binding site 2 of BSA. Autodock 4 was also used to perform ‘‘blind-docking’’ experiments for the polyphenol epicatechin-G, to determine whether there are any other favourable positions on the BSA surface for the binding of epicatechin-G, and hence for similar polyphenols. The space that was explored in this blind-docking experiment was larger than in the previous docking experiments where the space in the proximity of the BSA tryptophan residues was targeted (binding sites 1 and 2). Therefore, the resolution of this docking experiment was relatively low compared to the previous Autodock experiments. The set-up of this blind-docking experiment was: 200 runs, at a spacing of 0.853 Å, with a grid extension of 5.0 Å from all of the atoms of the BSA model – 126  102  90 points. The distribution of the Autodock scores for the blind-docking experiments is Gaussian-like (data not shown). The surface of the BSA protein has many cavities that represent potential binding sites for epicatechin-G and similar polyphenols. The distribution of the best 10 orientations of epicatechin-G at the surface of BSA was further analyzed and it shows that there are two ‘‘hot-spots’’ for the binding of epicatechin-G to BSA, and both of these are in the proximity of one of the tryptophan residues. The molecular docking data confirm that the polyphenols caffeic acid and p-coumaric acid bind weakly to BSA, while rosmarinic acid, which is an ester of caffeic acid with 3,4-dihydroxyphenyl lactic acid, has more available hydroxyl groups for interactions with Trp213. However, due to the small differences in the molecular docking scores, it was not possible to confirm the same binding affinities toward BSA for these polyphenols, as seen with the experimentally determined binding constant, KA. Just recently, it has been shown by phosphorescence and docking studies that catechin and epicatechin perturb the local environment of Trp134 in BSA and Trp214 in HSA (Durba et al., 2012). Combined with the thermodynamic analysis, we believed that hydrogen bonds and electrostatic interactions are the driving forces in the binding of the polyphenols to BSA, rather than the hydrophobic interactions. As a consequence of this, the addition of a glycosyl to the aromatic ring resulted in a decrease in the binding affinity (Dufour & Dangles, 2005; He et al., 2007). Overall, these

data thus indicate that the polyphenols–BSA binding site is mainly in the vicinity of Trp213, which is located in domain IIA of the BSA protein. 4. Conclusions In the present study, the interactions of selected catechins, flavones and hydroxycinnamic acids with BSA were investigated by combination of fluorescence emission spectrometry, circular dichroism and molecular docking. Our data reveal that the polyphenol structures significantly affect the binding process. The binding affinity is higher for esterified catechins (epigallocatechin-G and epicatechin-G) and lowest for epigallocatechin, among all of the polyphenols studied here. The glycosylated kaempferol (kaempferol-glu) has a lower binding affinity than has kaempferol and the structurally simpler hydroxycinnamic acids, such as caffeic acid and p-coumaric acid, have very low binding affinities to BSA. Circular dichroism data indicate that polyphenols, after binding to BSA, do not change the secondary structure significantly. The highest effect on secondary structure of BSA (slight increase in the amount of helicity) was by epicatechin-G. The molecular docking results confirm the experimental data, and indicate that the most possible binding site for polyphenols on BSA is in the vicinity of Trp213. Acknowledgements The authors would like to express their gratitude for financial support from the Slovenian Research Agency through the P40121 research programme. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.foodchem.2012.06.114. References Arts, I. C. W., Hollman, P. C. H., de Mesquita, H. B. B., Feskens, E. J., & Kromhout, D. (2001). Dietary catechins and epithelial cancer incidence: The Zutphen elderly study. International Journal of Cancer, 92, 298–302. Bourassa, P., Kanakis, C. D., Tarantilis, P., Pollissiou, M. G., & Tajmir-Riahi, H. A. (2010). Resveratrol, genistein, and curcumin bind bovine serum albumin. Journal of Physical Chemistry B, 114, 3348–3354. Carter, D. C., & Ho, J. X. (1994). Structure of serum albumin. Advances in Protein Chemistry, 45, 153–203. Cheynier, V. (2005). Polyphenols in foods are more complex than often thought. The American Journal of Clinical Nutrition, 81(1), 223S–229S. Dufour, C., & Dangles, O. (2005). Flavonoid-serum albumin complexation: Determination of binding constants and binding sites by fluorescence spectroscopy. Biochimica et Biophysica Acta, 1721, 164–173. Durba, R., Samrajnee, D., Shyam, S. M., Sanjib, G., Atanu, S. R., Kalyan, S. G., et al. (2012). Spectroscopic and docking studies of the binding of two stereoisomeric

2424

M. Skrt et al. / Food Chemistry 135 (2012) 2418–2424

antioxidant catechins to serum albumins. Journal of Luminiscence, 132, 1364–1375. Friesner, R. A., Banks, J. L., Murphy, R. B., Halgren, T. A., Klicic, J. J., Mainz, D. T., et al. (2004). Glide: A new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy. Journal of Medicinal Chemistry, 47, 1739–1749. Friesner, R. A., Murphy, R. B., Repasky, M. P., Frye, L. L., Greenwood, J. R., Halgren, T. A., et al. (2006). Extra precision glide: Docking and scoring incorporating a model of hydrophobic enclosure for protein–ligand complexes. Journal of Medicinal Chemistry, 49, 6177–6196. Ghuman, J., Zunszain, P. A., Petitpas, I., Bhattacharya, A. A., Otagiri, M., & Curry, S. (2005). Structural basis of the drug-binding specificity of human serum albumin. Journal of Molecular Biology, 353(1), 38–52. Halgren, T. A., Murphy, R. B., Friesner, R. A., Beard, H. S., Frye, L. L., Pollard, W. T., et al. (2004). Glide: A new approach for rapid, accurate docking and scoring. 2. Enrichment factors in database screening. Journal of Medicinal Chemistry, 47, 1750–1759. He, Q., Lv, Y., & Yao, K. (2007). Effects of tea polyphenols on the activities of aamylase, pepsin, trypsin and lipase. Food Chemistry, 101, 1178–1182. Ishii, T., Ishikawa, M., Miyoshi, N., Yasunaga, M., Akagawa, M., Uchida, K., et al. (2009). Catechol-type polyphenol is a potential modifier of protein sulfhydryls: Development and application of a new probe for understanding the dietary polyphenol actions. Chemical Research in Toxicology, 22, 1689–1698. Kragh-Hansen, U. (1981). Molecular aspects of ligand binding to serum albumin. Pharmacological Reviews, 33(1), 17–53. Krieger, E., Koraimann, G., & Vriend, G. (2002). Increasing the precision of comparative models with YASARA NOVA – a self parametrizing force field. Proteins, 47, 393–402. Kroll, J., Rawel, H. M., & Rohn, S. (2003). Reactions of plant phenolics with food proteins and enzymes under special consideration of covalent bonds. Food Science and Technology Research, 9, 205–218. Lakowicz, J. R., & Weber, G. (1973). Quenching of protein fluorescence by oxygen. Detection of structural fluctuations in proteins on the nanosecond time scale. Biochemistry, 12, 4171–4179. Lakowicz, J. R. (2006). Principles of fluorescence spectroscopy (3rd ed.). New York: Springer, pp. 954. Lambert, J. D., Sang, S., & Yang, C. S. (2007). Possible controversy over dietary polyphenols: Benefits vs risks. Chemical Research in Toxicology, 20, 583–585. Liu, E.-H., Qi, L.-W., & Li, P. (2010). Structural relationship and binding mechanisms of five flavonoids with bovine serum albumin. Molecules, 15, 9092–9103. Moris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, W. E., Belew, R. K., et al. (1998). Automated docking using a Lamarckian genetic algorithm and empirical binding free energy function. Journal of Computational Chemistry, 19, 1639–1662. Moriyama, Y., Watanabe, E., Kobayashi, K., Harano, H., Inui, E., & Takeda, T. (2008). Secondary structural change of bovine serum albumine in thermal denaturation

up to 130 °C and protective effect of sodium dodecyl sulfate on the change. Journal of Physical Chemistry B, 112, 16585–16589. Nozaki, A., Kimura, T., Ito, H., & Hatano, T. (2009). Interaction of phenolic metabolites with human serum albumin: A circular dichroism study. Chemical and Pharmaceutical Bulletin, 57, 1019–1023. Papadopoulou, A., Green, R. J., & Frazier, R. A. (2005). Interaction of flavonoids with bovine serum albumin: A fluorescence quenching study. Journal of Agriculture and Food Chemistry, 53(1), 158–163. Peters, T. Jr., (1985). Serum albumin. Advances in Protein Chemistry, 37, 161–245. Rawel, H. M., Meidtner, K., & Kroll, J. (2005). Binding of selected phenolic compounds to proteins. Journal of Agriculture and Food Chemistry, 53, 4228–4235. Roy, M. K., Koide, M., Rao, T. P., Okubo, T., Ogasawara, Y., & Juneja, L. R. (2010). ORAC and DPPH assay comparison to assess antioxidant capacity of tea infusions: Relationship between total polyphenol and individual catechin content. International Journal of Food Science & Nutrition, 61, 109–124. Schrödinger Suite 2011 Protein Preparation Wizard; Epik version 2.2, Schrödinger, LLC, New York, NY, 2011; Impact version 5.7, Schrödinger, LLC, New York, NY, 2011; Prime version 3.0, Schrödinger, LLC, New York, NY, 2011. Shi, X., Li, X., Gui, M., Zhou, H., Yang, R., Zhang, H., et al. (2010). Studies on interaction between flavonoids and bovine serum albumin by spectral methods. Journal of Luminiscence, 130, 637–644. Soares, S., Mateus, N., & de Freitas, V. (2007). Interaction of different polyphenols with bovine serum albumin (BSA) and human salivary a-amylase (HSA) by fluorescence quenching. Journal of Agriculture and Food Chemistry, 55, 6726–6735. Tayeh, N., Rungassamy, T., & Albani, J. R. (2009). Fluorescence spectral resolution of tryptophan residues in bovine and human serum albumins. Journal of Pharmaceutical and Biomedical Analysis, 50, 107–116. Van Durme, J., Delgado, J., Stricher, F., Serrano, L., Schymkowitz, J., & Rosseau, F. (2011). A graphical interface for the FoldX forcefield. Bioinformatics, 27, 1711–1712. Vensellar, H., Joosten, R. P., Vroling, B., Baakman, C. A., Hekkelman, M. L., Krieger, E., et al. (2010). Homology modeling and spectroscopy, a never ending love-story. European Biophysics Journal, 39, 551–563. Vermerris, W., & Nicholson, R. L. (2008). Phenolic compound biochemistry. Dordrecht: Springer, pp. 1–32. Villañoa, D., Fernández-Pachóna, M. S., Moyáb, M. L., Troncosoa, A. M., & GarcíaParrilla, M. C. (2007). Radical scavenging ability of phenolic compounds towards DPPH free radical. Talanta, 71, 230–235. Wang, J., Ho, L., Zhao, W., Ono, K., Rosensweig, C., Chen, L., et al. (2008). Grapederived phenolic compounds prevent a oligomerization and attenuate cognitive deterioration in a mouse model of Alzheimer’s disease. The Journal of Neuroscience, 28, 6388–6392.