Biochimica et Biophysica Acta 1674 (2004) 205 214 www.bba-direct.

com

Interactions of human serum albumin with chlorogenic acid and ferulic acid
Juan Kanga,b, Yuan Liua, Meng-Xia Xiea,*, Song Lia, Min Jianga, Ying-Dian Wanga
b

Analytical and Testing Center of Beijing Normal University, Beijing 100875, Peoples Republic of China College of Chemistry and Molecular Engineering of Peking University, Beijing 100083, Peoples Republic of China Received 26 March 2004; received in revised form 9 June 2004; accepted 24 June 2004 Available online 20 July 2004

Abstract The interactions of chlorogenic acid and ferulic acid with human serum albumin (HSA) have been investigated by fluorescence and Fourier transformed infrared (FT-IR) spectrometry. Fluorescence results showed that one molecule of protein combined with one molecule of drugs at the molar ratio of drug to HSA ranging from 1 to 10, and their binding affinities (K A ) are 4.37104 M1 and 2.23104 M1 for chlorogenic acid and ferulic acid, respectively. The primary binding site for chlorogenic acid is most likely located on IIA and that for ferulic acid in IIIA. The main mechanism of protein fluorescence quenching was static quenching process. Combining the curve-fitting results of infrared amide I and amide III bands, the alterations of protein secondary structure after drug complexation were estimated. With increasing the drug concentration, the protein a-helix structure decreased gradually and the reduction of protein a-helix structure reached about 7% and 5% for protein binding with chlorogenic acid and ferulic acid individually at the drug to protein molar ratio of 30. This indicated a partial unfolding of HSA in the presence of the two acids. From the fluorescence and FT-IR results, the binding mode was discussed. D 2004 Elsevier B.V. All rights reserved.
Keywords: Chlorogenic acid; Ferulic acid; Human serum albumin; FT-IR spectroscopy; Fluorescence spectroscopy

1. Introduction Phenolic acids, as a kind of important active components in Chinese traditional medicine, exist widely in leaves, roots and especially fruits of the plants. With different chemical structures and characters, they have been confirmed to have multiple biological and pharmacological properties [1,2]. Chlorogenic acid and ferulic acid (structure shown in Fig. 1) are the major phenolic compounds found in numerous plant species [3]. They have the phenolic groups enabling them to act as the natural antioxidant [4 6], and they can bind to enzymes and other multisubunit proteins and modify their structural properties and alter their biological activities [79].
* Corresponding author. Tel.: +8610 5880 7981; fax: +8610 5880 0076. E-mail address: mengxia-xie@263.net (M.-X. Xie). 0304-4165/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.bbagen.2004.06.021

Human serum albumin (HSA) is the most abundant protein constituent of blood plasma and serving as a protein storage component. It is a globular protein consisting of a single peptide chain of 585 amino acids. HSA can bind with a variety of substrates, including metal cations, fatty acids, amino acids and diverse drugs. It has been shown that the distribution, free concentration and the metabolism of various drugs can be significantly altered as a result of their binding to HSA [10]. Therefore, investigating the interaction of active components in Chinese herbs with HSA can provide useful information of drug actions and can be used as a model for elucidating the drugprotein complex [11,12]. Chlorogenic acid may interact covalently or non-covalently with bovine serum albumin [13,14], and the interaction caused the structural changes of the protein and has effects on its functional properties. Equilibrium dialysis, ultrafiltration, calorimetry and other methods have

206

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214

Fig. 1. Chemical structure of chlorogenic acid (A) and ferulic acid (B).

been used to study the interaction of chlorogenic acid, ferulic acid and related organic anions with bovine serum albumin [1518]. Fluorescence spectroscopy has become a popular method for studying the interaction of drugs and proteins [1921]. It can provide the binding information and reflect the conformation changes of proteins in various environments. Fourier transform infrared spectroscopy (FT-IR) is a new method that can be used in the investigation of drugprotein interaction. It can estimate the alterations of protein secondary structure [22,23]. In this work, the interactions between chlorogenic acid and ferulic acid and HSA have been investigated by FT-IR and fluorescence spectroscopy. Through the binding strength, position and alteration of protein secondary structure, the binding modes were discussed.

2.3. Fluorescence and UV absorption spectroscopic measurement Fluorescence emission spectra were recorded from 300 to 500 nm (excited at 280 nm) and 315 to 500 nm (excited at 295 nm), respectively, on a Fluorolog-TAU-3 fluorescence spectrometer (JY, USA) at room temperature. Buffer solutions of drugs in corresponding concentrations were used as reference when measuring the fluorescence spectra of proteindrug mixtures. The fluorescence emission spectrum of HSA solution in concentration of 5.0106 mol l1 was also measured at excitation wavelength of 295 nm from 260 to 520 nm for calculating overlapping areas. The UV absorbance spectra of the two drugs with concentration of 5.0105 mol l1 were recorded on a Cintra-10e UVVis spectrometer (GBC, Australia) from 260 to 520 nm and shown in Fig. 6. 2.4. FT-IR spectroscopic measurements All infrared spectra were recorded at room temperature on a Nicolet Nexus 670 FT-IR spectrometer equipped with a zinc selenide (ZnSe) attenuated total reflectance (ATR) accessory, a deuterated triglycine sulfate (DTGS) detector and a KBr beam splitter. Prior to obtaining sample spectrum, an open beam background spectrum through the blank ATR crystal was recorded. For each spectrum, a 512-scan interferogram was collected at a resolution of 4 cm1. A pre-recorded water vapor absorption spectrum under identical condition was automatically subtracted during data collection. 2.5. FT-IR spectra processing procedures and estimation of protein secondary structure The free protein infrared spectrum was obtained by subtracting the spectrum of the protein solution to that of buffer solution using the band located at 22001800 cm1 as internal standards [24]. The difference process between the spectra of protein and drug solutions and that of the same concentration of drugs were generated to obtain the protein spectra after interacting with drugs as that described in Ref. [25]. A straight baseline passing through the ordinates at 16001700 cm1 and 13301220 cm1 was subtracted to get the amide I and amide III bands.

2. Materials and methods 2.1. Materials HSA (fatty acid-free, 99%) fraction V and ferulic acid were purchased from Sigma Chemical Co. (St. Louis, MO, USA), chlorogenic acid was purchased from Acros Organic Co. (New Jersey, USA) and other chemicals were of reagent grade. All reagents were used as supplied without further purification. 2.2. Preparation of stock solutions The drugs and protein were dissolved in 20 mmol/l phosphate buffer solution (pH 7.4) individually. A series of drugprotein solutions were prepared by mixing protein solution with different concentration of drug solutions. For FT-IR determination, the final protein concentration was 4.0104 mol/l, and the molar ratios of drug to protein were 0, 0.1, 1, 5, 10, 20 and 30. For fluorescence determination, the final protein concentration was 5.0106 mol/l, and the molar ratios of drug to protein were 0, 0.01, 0.1, 0.3, 0.5, 0.7, 1, 3, 5, 7, 10, 20, and 30. The mixtures were stirred to ensure the formation of homogeneous solution and put into the refrigerator overnight prior to determination. The reference solutions were also prepared according to above procedures without protein.

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214

207

Fourier self-deconvolution and second derivative resolution enhancement were applied to narrow the widths of infrared bands and increase the separation of the overlapping components [12,23]. The resolution enhancement resulting from self-deconvolution and the second derivative is such that the number and the position of the bands to be fitted are determined. The curve-fitting was accomplished with a curve-fitting process employing Galactic peaksolve software (version 1.0). The number of bands was entered into the program along with their respective positions and half-heights. The program iterates the curve-fitting process to achieve the best Gaussian-shaped curves that fit the original protein spectrum. A best fit is determined by the root mean square (rms) of differences between the original protein spectrum and the sum of all individual resolved bands. The assignment of component bands in amide I and amide III of HSA has been studied [12,26,27] in our previous work, 16101640 cm1 to h-sheet, 16401650 cm1 to random coil, 16501658 cm1 to a-helix and 16601700 cm1 to h-turn structure in amide I; 13301290 cm1 to a-helix, 12901270 cm1 to h-turn, 12701250 cm1 to random coil and 12501220 cm1 to h-sheet in amide III. The percentages of each secondary structure were

calculated from the integrated areas of the component bands in amide III and amide I, respectively.

3. Results and discussion 3.1. Fluorescence quenching of HSA in the presence of drugs Fluorescence quenching of protein could be used to retrieve many drugprotein binding information [28,29]. When the excitation wavelength is 280 nm, both tryptophan and tyrosine amino acid residues in protein have fluorescence emission. While at the excitation wavelength of 295 nm, only tryptophan has fluorescence emission. The fluorescence emission wavelength of HSA is about 350 nm at both excitation wavelengths. The fluorescence emission intensities of HSA were quenched when the protein interacted with chlorogenic acid and ferulic acid. Fig. 2 showed the fluorescence quenching of HSA with the increase of drug concentrations. A series of molar ratios of drug to protein were selected, and it could be found that the fluorescence emission nearly completely quenched when the molar ratio reached to 30. Chlorogenic

Fig. 2. HSA fluorescence emission in the presence of chlorogenic acid (A), (B) and ferulic acid (C), (D). (A), (C) k ex=280 nm; (B), (D) k ex=295 nm; molar ratio of drug to protein is 0, 0.01, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 2, 3, 4, 5, 10, 20, 30 (from top to bottom).

208

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214

acid and ferulic acid also have fluorescence emission at about 440 and 420 nm, respectively, under the same experimental conditions (see Fig. 3), but using the same concentration of drugs as reference solution could eliminate their interferences when measuring the protein fluorescence emission spectra at different drug concentrations. From Fig. 2, it can be seen that the fluorescence emission wavelengths of HSA have obvious red shifts after the protein interacted with drugs, most significantly in higher drug concentrations. When the drug to protein molar ratio is 30, the fluorescence emission wavelengths of HSA shift about 18 and 8 nm after binding with chlorogenic acid and ferulic acid, respectively. The red shifts indicated that the tryptophan residue in protein has been brought to a more hydrophilic environment [30,31]. Regression curves were plotted according to the Stern Volmer equation (Eq. (1)) [32,33] for fluorescence quenching of HSA induced by chlorogenic acid and ferulic acid (see Fig. 4). F0 1 KQ s0 Q 1 KD Q F 1

In Eq. (1), F 0 and F are the fluorescence intensities in the absence and presence of quencher, [ Q] is the quencher concentration and K D is the SternVolmer quenching constant, which can be written as K D =K Q s 0, where K Q is the bimolecular quenching constant and s 0 is the lifetime of the fluorophore in the absence of quencher. The curves were not linear when drug to protein molar ratios were higher than 10. The SternVolmer quenching constant K D was obtained by the slope of regression curves in the linear range, and the bimolecular quenching constant K Q was calculated (s 0 is about 108 s, as to Ref. [34]). The results in Table 1 showed that the bimolecular quenching constants are larger than the limiting diffusion rate constant of the biomolecule (2.01010 M1 s1) [34,35], so the dynamic quenching is not the main reason that causes the fluorescence quenching of HSA.

Fig. 4. SternVolmer plots of fluorescence quenching of HSA in the presence of chlorogenic acid (A) and ferulic acid (B) (k ex=295 nm).

3.2. Binding constants of HSAdrug complex It was postulated that the HSA fluorescence quenching was a static quenching process. The number of binding sites (n) can be obtained from the regression curve based on the following equation (Eq. (2)) deduced from our previous report [12] in the linear range and listed in Table 1. lg F 0 F lg KA n lgQ F 2

The results showed that the n was about 1 for both drugs, which suggested that one molecule of HSA combined one molecule of drugs in the drug to protein molar ratio ranging from 1 to 10. Double reciprocal curves of fluorescence quenching of HSA were plotted (see Fig. 5) in the same drug

Table 1 Quenching constants K Q and binding constants K A between chlorogenic acid or ferulic acid with HSA (k ex=295 nm) Drugs CLA FA Fig. 3. Fluorescence emission spectra of chlorogenic acid and ferulic acid (k ex=295 nm). K Q (1012 M1 s1) 5.39F0.64 3.02F0.59 K A (104 M1) 4.37F0.59 2.23F0.32 n 1.11F0.11 1.17F0.12

CLA: chlorogenic acid; FA: ferulic acid; mean values for five different experiments.

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214

209

3.3. Binding distance There is a good overlapping between the fluorescence emission spectrum of free HSA and absorption UV spectra of the two drugs (Fig. 6). As the fluorescence emission of protein was affected by the excitation light around 295 nm, the spectrum ranging from 300 to 520 nm was chosen to calculate the overlapping integral. According to Ffrsters theory [37,38], the energy transfer efficiency E is defined as the following equation (Eq. (4)). Where r is the distance from the ligand to the tryptophan E R6 0 r6

R6 0

amino acid residue of the protein and R 0 is the Ffrster critical distance, at which 50% of the excitation energy is transferred to the acceptor. It can be calculated from donor

Fig. 5. Double reciprocal curve of protein fluorescence quenching in the presence of chlorogenic acid (A) and ferulic acid (B) (k ex=295 nm).

concentration range based on Eq. (3) [12]. From the regression equation of the curves, the binding constants (K A ) of protein to drugs can be obtained (see Table 1). F0 1 1 1 KA Q F0 F 3

From Table 1, it can be seen that the binding affinity of protein to chlorogenic acid was stronger than that of protein to ferulic acid. It can be explained by the difference of the drug structure. Both molecules of chlorogenic acid and ferulic acid have phenolic hydroxyl, COOH and phenyl ring (see Fig. 1). Chlorogenic acid has two phenolic hydroxyl groups and three hydroxyl groups connected to saturated hexacyclic ring, but ferulic acid has only one phenolic hydroxyl group. The hydroxyl group in drug molecule can bind with polypeptide chain of protein by forming hydrogen bonds [11,36]. So the binding affinity of protein to chlorogenic acid, which has more phenolic hydroxyl groups and more chances to form hydrogen bonds, was stronger than that of ferulic acid.

Fig. 6. Overlap between the fluorescence emission and UV absorption (1), the fluorescence emission spectrum of HSA (k ex=295 nm) (2), UV absorption spectra of chlorogenic acid (A) and ferulic acid (B); [HSA]=5.0106 mol/l, [ Q]=5.0105 mol/l.

210

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214

emission and acceptor absorption spectra using the Ffrster formula (Eq. (5)), R6 8:8 1025 K 2 n4 UJ : 0 5

Where K 2 is a factor that describes the relative orientations of the donor and acceptor, U is the fluorescence quantum yield of donor in the absence of acceptor; n is the refractive index of the medium between the donor and acceptor. Given that the fluorescence quantum yield U of tryptophan is 0.118, the refractive index n of the medium is the average value of water and organic solute (1.366) and K 2 is 2/3 for random orientation [39], J is the integral that defines the amount of overlapping areas between the normalized fluorescence emission spectrum of the donor and the UV absorption spectrum of the acceptor and was calculated by dividing the area of the overlapped region to a very small rectangle (Eq. (6)). J X Fk ek k4 Dk .X Fk Dk 6

conformation has been changed after protein binding with chlorogenic and ferulic acid [30,31,41,42]. If the change of protein structure included the transforming of protein secondary structure in the drugHSA complex, it can be reflected in the infrared absorption spectra. Table 2 lists the curve-fitted results of protein secondary structure before and after binding with different concentration of drugs. The free HSA in buffer solution contained major a-helix 54% and h-sheet 23%, which are consistent with the recent spectroscopic studies of the HSA [25] and the determination results from amide I matched well with those from amide III (maximum difference is lower than 1%). In order to relate the alterations of protein secondary structure to their fluorescence quenching, a fluorescence quenching fraction Q was defined as follows; F0 F 100%: F0

F k is the fluorescence intensity of donor in the absence of the acceptor at wavelength k and e k is the UV molar absorption coefficient of the acceptor at k. E can be estimated from the following formula: E 1 F=F0 : 7

Where F 0 and F is the fluorescence intensity of donor in the absence and presence of equal amount of acceptor, respectively. If the value of E, K 2, U and n are known, R 0 and r can be calculated. The calculated results showed that R 0 for chlorogenic acid and ferulic acid were 2.53 and 1.95 nm and their binding distances r were 3.57 and 2.45 nm, respectively. Theres only one tryptophan residue in HSA, Trp214, and the distance from the bound ligands to Trp214 are both less than 7 nm, which suggested that a non-radioactive energy transfer mechanism may be among the quenching mechanisms [32,34]. While the binding distance r of the drugs was more than their respective critical distance R 0, the fluorescence quenching was more likely induced by static quenching other than non-radioactive energy transferring [32,34]. 3.4. Alterations of protein secondary structure induced by drugs When drugs bind to a globular protein, the intramolecular forces responsible for maintaining the secondary and tertiary structures can be altered, resulting in a conformational change of the protein [40]. The distinct fluorescence quenching of the Trp214 suggested that the drugHSA combination has changed the microenvironment of tryptophan. The red shifts of fluorescence emission wavelength, shown in Fig. 2, also indicated that the

Where F 0 and F is the fluorescence intensity of HSA before and after binding with drugs. The protein fluorescence quenching fractions excited at 280 and 295 nm were calculated, respectively, at different drug to protein molar ratios (see Table 3). From Table 2, it can be seen that there were no obvious changes of protein secondary structure in low drug concentration (C drug/C HSAb0.1), which is consistent with the weak fluorescence quenching fraction listed in Table 3. When drug to protein molar ratio reached 1, the percentage of protein a-helix structure decreased about 1%, which signified that the drugs bound with the amino acid residue of the main polypeptide chain of protein and destroyed their hydrogen binding networks. In this drug concentration, the protein fluorescence quenching became obvious; the quenching fraction Q is about 20% for chlorogenic acid and 15% for ferulic acid. When the molar ratio of drug to protein increased from 1 to 10, the protein quenching fraction Q also increased significantly, from about 20% to 75% for chlorogenic acid and 15% to 70% for ferulic acid. In this range of drug concentration, the percentage of protein a-helix structure decreased about 3% and 2%, and h-sheet structure decreased about 2% and 1% for chlorogenic acid and ferulic acid, respectively, with the increase of drug concentration due to forming drugprotein complex; in the meantime, the contents of h-turn and random coil structure increased more than 2% for protein binding with chlorogenic acid, and the content of random coil structure increased about 3%, but h-turn structure remained unchanged for ferulic acid. The percentage of protein a-helix structure decreased about 3% with increasing the molar ratio of chlorogenic acid to protein from 10 to 30, but the content of h-sheet structure has very little alteration. The interaction of ferulic acid to protein caused about 2% reduction of a-helix structure and

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214 Table 2 Percentages of protein secondary structure before and after interaction with the two acids at different concentrations Medicine Molar ratio (drugprotein) 0:1 0.1:1 1:1 CLA 5:1 10:1 20:1 30:1 0:1 0.1:1 1:1 FA 5:1 10:1 20:1 30:1 Amide band (%) I III I III I III I III I III I III I III I III I III I III I III I III I III I III a-Helix (%) 54.4 54.2 54.5 53.8 52.8 52.9 52.1 51.3 49.9 49.3 47.7 48.4 46.7 47.5 54.4 54.2 54.6 54.4 53.2 53.2 52.4 52.2 50.9 50.7 49.9 49.3 49.1 48.7 h-Sheet (%) 23.0 23.1 23.3 23.3 23.2 23.1 22.3 22.2 21.4 21.6 21.2 21.1 20.6 21.3 23.0 23.1 23.0 23.2 22.8 22.9 22.1 22.0 22.1 22.2 21.3 21.2 20.6 21.0 h-Turn (%) 16.5 16.4 16.1 16.3 17.3 16.7 17.6 18.1 19.1 19.7 21.2 20.9 22.3 21.1 16.5 16.4 15.8 16.1 16.8 16.7 16.9 16.5 17.3 17.2 19.1 19.3 20.2 20.0

211

Random (%) 6.1 6.3 6.2 6.6 6.7 7.3 8.0 8.4 9.6 9.3 9.62 9.6 10.4 10.1 6.1 6.3 6.6 6.4 7.2 7.2 8.7 9.3 9.7 9.9 9.8 10.2 10.1 10.4

Mean values for five different experiments and the mean deviation was 0.52%.

about 1% decrease of h-sheet structure in this range of drug concentration. The content of h-turn structure increased about 2% and 3%, and random structure increased less than 1% for protein interacting with chlorogenic acid and ferulic acid, respectively, in the high drug concentration. Fig. 7 showed the original and curve-fitting spectra of amide I and amide III bands before and after the protein binging with the two drugs at the drug to protein molar ratio of 30. In the high drug concentration, the fluorescence quenching fraction Q (excited at 295 nm) does not have much difference
Table 3 Fluorescence quenching fraction (%) at different drug concentrations Medicine Molar ratio (drugprotein) 0.1:1 1:1 5:1 10:1 20:1 30:1 0.1:1 1:1 5:1 10:1 20:1 30:1 k ex (nm) 280 nm 1.7F0.7 18.3F3.7 56.2F2.3 69.0F1.8 85.9F3.3 93.6F2.4 1.3F1.0 17.7F2.4 50.9F2.1 75.6F2.5 92.9F2.6 96.0F1.3 295 nm 1.9F0.5 19.6F4.2 61.0F3.1 75.6F3.0 90.4F2.9 96.5F1.3 2.0F1.9 14.8F3.3 47.0F2.3 71.3F1.7 91.2F2.9 94.7F1.2

between the two drugs, and the Q induced by ferulic acid exceeds that induced by chlorogenic acid at the excitation wavelength of 280 nm. 3.5. Binding mode of HSA with drugs From the fluorescence results of protein binding with chlorogenic acid and ferulic acid, the binding mode varied with drug concentrations. In lower drug concentration (drug to protein molar ration ranged from 1 to 10), the calculation based on Eq. (2) above showed that there was only one binding site, which indicated that one molecule of drug bound to one molecule of protein. In higher drug concentrations (drug to protein molar ration ranged from 10 to 30), the number of binding sites was more than one and cant be calculated because the relationship of lg[( F 0F)/F] and lg[ Q] was beyond the linear range. HSA comprises three homologous domains and each domain consisted of two subdomains, subdomain A and subdomain B [4345]. Two major ligand binding sites, site I and site II, are located within specialized cavities in subdomains IIA and IIIA [46,47]. Many ligands, such as digitoxin and ibuprofen, were found to bind preferentially to the cavity in IIIA, but aspirin and iodinated aspirin analogues, such as triiodobenzoic acid (TIB), show nearly equal distributions between binding sites located in IIA and IIIA [44].

CLA

FA

Mean values for five different experiments.

212

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214

Fig. 7. Curve-fitted amide I and amide III (13301220 cm1) regions of free HSA, CLAHSA and FAHSA in the drug to protein molecular ratio of 30.

Chlorogenic acid and ferulic acid belong to small aromatic acids like TIB, which have both aromatic ring and carboxylate group, so it is possible for them to bind to the cavity in IIA or IIIA. Because the structures of the two acids, which contain hydroxyl groups, were different from that series of molecules such as TIB, chlorogenic acid and ferulic acid have only one binding site on HSA in the drug to protein molar ratio ranging from 1 to 10. From Table 3, it can be seen that the Q from Trp and Tyr (excited at 280 nm) is smaller than that from only Trp

(excited at 295 nm) when protein bound with chlorogenic acid, but the result was reverse for the proteinferulic acid complex. This illustrated that ferulic acid has more influence on the microenvironment of Tyr. In subdomain IIA, there is only one tyrosine residue, Tyr263, while in subdomain IIIA there are four tyrosine residues, Tyr401, Tyr411, Tyr452 and Tyr497. Hence, we presumed that chlorogenic acid most likely binds to the cavity in subdomain IIA and ferulic acid binds to the cavity in subdomain IIIA.

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214

213

The interaction of chlorogenic acid and ferulic acid with protein includes the hydrophobic forces between the aromatic ring of drugs and the hydrophobic amino acid residues, the static forces between carboxylate group and the basic amino acid residues and the hydrogen bonding between the hydroxyl groups and polypeptide chain. The drugs entered the hydrophobic binding cavities located in subdomain IIA or IIIA by the hydrophobic interactions, and the hydroxyl groups in the aromatic rings can interact with CMO and NUH of the main polypeptide chain by forming strong hydrogen bonding, resulting in the rearrangement of polypeptide carbonyl hydrogen bonding network and the reduction of protein a-helix structure. The alterations of protein secondary structure were slightly significant for protein interacting with chlorogenic acid than ferulic acid, and this was consistent with their binding affinities (see Table 1). It also indicated that the phenolic hydroxyl groups in the drug molecules played an important role for their binding with protein. The combinations of HSA with chlorogenic and ferulic acids signified that the two drugs could be stored and carried by the protein and have their biological and pharmacological actions. The drugprotein binding mechanism can be used as a mode to elucidate the biological and pharmacological properties of drugs. It can explain the phenomenon that chlorogenic acid has higher inhibitory effects on the activities of enzymes than ferulic acid has [7]. The higher binding affinity of chlorogenic acid to protein caused more significant alteration of protein structure, resulting in higher inhibitory effects on the activities of enzymes. Chlorogenic acid can significantly suppress the invasion of AH109A, a rat ascites hepatoma cell line, and Yagasaki et al. [48] suggested that the 3,4-dihydroxyl groups of the drugs might be involved in the suppression. In the higher drug concentration (drug to protein molar ratio ranged from 10 to 30), the drugs can bind with protein not only in high-affinity binding sites located in IIA or IIIA, but also in other weak-affinity binding sites [49,50], and brought continued reduction of protein a-helix structure. The interaction of drugs with protein caused the protein structure mainly transforming from a-helix to h-turn structure, and it illustrated that the hydrophobic or neutral residues formerly constructing a-helix structure were exposed to the surface and composed of h-turn structure at high drug concentrations. In fact, the binding mode may be more complex in high drug concentration. On the protein surface, there are many polar and charged amino acid residues, and drugs may bind with them by electrostatic interaction. Basic amino acids, + which have positive side chain (NH3 ) with strong polarity are commonly distributed on the surface of the protein and contribute little to the protein conformation stability [51]. In conclusion, both chlorogenic acid and ferulic acid can bind with HSA to form drugprotein complexes. The combination caused significant reductions of protein ahelix structure, which signified the rearrangement of carbonyl hydrogen bonding network of the main polypep-

tide chain. The protein fluorescence quenching mechanism was mainly a static quenching process and the hydroxyl groups on the aromatic ring of drug molecules played an important role during the binding processes. References
[1] Y. Koshihara, T. Neichi, S.I. Murota, Caffeic acid is a selective inhibitor for leukotriene biosynthesis, Biochim. Biophys. Acta 792 (1984) 92 97. [2] T. Tanaka, T. Kojima, T. Kawamori, A. Wang, M. Suzui, K. Okamoto, H. Mori, Inhibition of 4-nitroquinoline-1-oxide-induced rat tongue carcinogenesis by the naturally occurring phenolic caffeic, ellagic, chlorogenic and ferulic acids, Carcinogenesis 14 (1993) 1321 1325. [3] A.N. Cokugras, E. Bodur, Comparative effects of two plant growth regulator; indole-3-acetic acid and chlorogenic acid on human and horse serum butyrylcholinesterase, Pestic. Biochem. Physiol. 77 (2003) 24 33. [4] M. Nardini, F. Natell, V. Gentili, M. Di Felice, C. Scaccini, Effect of caffeic acid dietary supplementation on the antioxidant defense system in rat: an in vivo study, Arch. Biochem. Biophys. 342 (1997) 157 160. [5] K. Yagi, N. Ohishi, Action of ferulic acid and its derivatives as antioxidants, J. Nutr. Sci. Vitaminol. 25 (1979) 127 130. [6] K. Hiroe, H. Masashi, H. Kanae, A. Kayo, T. Hisaji, Antioxidant properties of ferulic acids and its related compounds, J. Agric. Food Chem. 50 (2002) 2161 2168. [7] R. Sascha, M.R. Harshadrai, K. Jurgen, Inhibitory effects of plant phenols on the activity of selected enzymes, J. Agric. Food Chem. 50 (2002) 3566 3571. [8] C.-T. Yeh, G.-C. Yen, Effects of phenolic acids on human phenolsulfotransferases in relation to their antioxidant activity, J. Agric. Food Chem. 57 (2003) 1474 1479. [9] P. Suryaprakash, R.P. Kumar, V. Prakash, Thermodynamics of interaction of cafferic acid and quinic acid with multisubunit proteins, Int. J. Biol. Macromol. 27 (2000) 219 228. [10] T. Peters, All About Albumin: Biochemistry, Genetics and Medical Applications, Academic Press, San Diego, 1995. [11] B. Bartolome , I. Estrella, M.T. Herna ndez, Interaction of low molecular weight phenolics with proteins (BSA), J. Food Sci. 65 (2000) 617 621. [12] M. Jiang, M.X. Xie, D. Zheng, Y. Liu, X.Y. Li, X. Cheng, Spectroscopic studies on the interaction of cinnamic acid and its hydroxyl derivatives with human serum albumin, J. Mol. Struct. 692 (2004) 71 80. [13] V.E.P. Stephanie, G. Harry, J.W.G.V. Antonie, G.A.V. Koningsveld, G.A.H. De Jong, A.G.J. Voragen, Effects of non-covalent interactions with 5-O-caffeoylquinic acid (chlorogenic acid) on the heat denaturation and solubility of globular proteins, J. Agric. Food Chem. 51 (2003) 5088 5095. [14] M.R. Harshadrai, R. Sascha, K. Hans-Peter, K. Jurgen, Structural changes induced in bovine serum albumin by covalent attachment of chlorogenic acid, Food Chem. 78 (2002) 443 445. [15] T. Adzet, J. Camarasa, E. Escubedo, M. Merlos, In vitro study of caffeic acidbovine serum albumin interaction, Eur. J. Drug Metab. Pharmacokinet. 13 (1988) 11 14. [16] B.K. Muralidhara, V. Prakash, Interaction of 3V-O-caffeoyl d-quinic acid with human serum albumin, Eur. J. Pept. Protein Res. 46 (1995) 1 8. [17] W.L. Peticolas, I.M. Klotz, Structural specificities in the interactions of some organic ions with serum albumin, II, J. Am. Chem. Soc. 78 (1956) 5262 5272. [18] J.D. Teresi, J.M. Luck, The combination of organic anions with serum albumin: VI. Quantitative studies by equilibrium dialysis, J. Biol. Chem. (1948) 653 661.

214

J. Kang et al. / Biochimica et Biophysica Acta 1674 (2004) 205214 [36] B. Bartolome, M.T. Hernandez, I. Eatrella, Evaluation of the capability of low molecular weight phenolic compounds to interact with proteins, Polyphenols Commun. 2 (1996) 439 440. [37] L.A. Sklar, B.S. Hudson, R.D. Simoni, Conjugated polyene fatty acids as fluorescent probes: binding to bovine serum albumin, Biochemistry 16 (1977) 5100 5108. [38] F.-L. Cui, J. Fan, J.-P. Li, Z.-D. Hu, Interactions between 1-benzoyl4-p-chloro-phenyl thiosemicarbazide and serum albumin: investigation by fluorescence spectroscopy, Bioorg. Med. Chem. 12 (2004) 151 157. [39] C.B. Berde, B.S. Hudson, R.D. Simoni, L.A. Sklar, Human serum albumin. Spectroscopic studies of binding and proximity relationships for fatty acids and bilirubin, J. Biol. Chem. 254 (1979) 391 400. [40] M.A. Khan, S. Muzammil, J. Musarrat, Differential binding of tetracyclines with serum albumin and induced structure alterations in drug-bound protein, Int. J. Biol. Macromol. 30 (2002) 243 249. [41] B. Klajnert, M. Bryszewska, Fluorescence studies on PAMAM dendrimers interactions with bovine serum albumin, Bioelectrochemistry 55 (2002) 33 35. [42] T. Yuan, A.M. Weljie, H.J. Vogel, Tryptophan fluorescence quenching by methionine and selenomethionine residues of calmodulin: orientation of peptide and protein binding, Biochemistry 37 (1998) 3187 3195. [43] B. Meloun, L. Moravek, V. Kostka, Complete amino acid sequence of human serum albumin, FEBS Lett. 58 (1975) 134 137. [44] X.M. He, D.C. Carter, Atomic structure and chemistry of human serum albumin, Nature 358 (1992) 209 2115. [45] D. Michael, C. Martina, C.C. Daniel, R. Florian, Five recombinant fragments of human serum albumintool for the characterization of the warfarin binding site, Prot. Sci. 9 (2000) 1455 1465. [46] B. Sjoholm, B. Ekman, A. Kober, I. Ljungstedt-Pahlman, B. Seiving, T. Sjodin, Binding of drugs to human serum albumin: XI. The specificity of three binding sites as studied with albumin immobilized in microparticles, Mol. Pharmacol. 16 (1979) 767 777. [47] K.J. Fehske, U. Schlafer, W.E. Muller, Characterization of an important drug binding area on human serum albumin including the high-affinity binding sites of warfarin and azapropazone, Mol. Pharmacol. 21 (1982) 387 393. [48] K. Yagasaki, Y. Miura, R. Okauchi, T. Furuse, Inhibitory effects of chlorogenic acids and its related compounds on the invasion of hepatoma cells in culture, Cytotechnology 33 (2000) 229 235. [49] T. Oida, Proton nuclear magnetic resonance study on the multimode interactions of human serum albumin with drug molecules, J. Biochem. (Tokyo) 100 (1986) 99 113. [50] A.A. Bhattacharya, T. Grune, S. Curry, Crystallographics analysis reveals common modes of binding of medium and long-chain fatty acids to human serum albumin, J. Mol. Biol. 303 (2000) 721 732. [51] K. Yagasaki, Y. Miura, R. Okauchi, T. Furuse, Inhibitory effects of chlorogenic acid and its related compounds on the invasion of hepatoma cells in culture, Cytotechnology 33 (2000) 229 235.

[19] L. Trynda-Lemiesz, A. Karacyzn, B.K. Keppler, H. Kozlowski, Studies on the interactions between human serum albumin and trans-indazolium (bisindazole) tetrachlororuthenate(III), J. Inorg. Biochem. 78 (2000) 341 346. [20] L. Trynda-Lemiesz, H. Kozlowski, Some aspects of the interactions of adriamycin with human serum albumin, Bioorg. Med. Chem. 4 (1996) 1709 1713. [21] S. Anna, Interaction of drugs with bovine and human serum albumin, J. Mol. Struct. 614 (2002) 227 232. [22] W. Dzwolak, M. Kato, A. Shimizu, Y. Taniguchi, Fourier-transform infrared spectroscopy study of the pressure-induced changes in the structure of the bovine a-lactalbumin: the stabilizing role of the calcium ion, Biochim. Biophys. Acta 1433 (1999) 45 55. [23] J.T. Pelton, L.R. Mclean, Spectroscopic methods for analysis of protein secondary structure, Anal. Biochem. 277 (2000) 167 176. [24] A.C. Dong, P. Huang, W.S. Caughey, Protein secondary structure in water from second-derivative amide I infrared spectra, Biochemistry 29 (1990) 3303 3312. [25] M. Purcell, J.F. Neault, H. Malonga, H. Arakawa, R. Carpentier, H.A. Tajmir-Riahi interactions of atrazine, Interactions of atrazine and 2,4D with human serum albumin studied by gel and capillary electrophoresis and FTIR spectroscopy, Biochim. Biophys. Acta 1548 (2001) 129 138. [26] M.X. Xie, Y. Liu, Studies on amide III infrared bands for the secondary structure determination of proteins, Chem. J. Chin. Univ. 24 (2) (2003) 226 231. [27] Y. Liu, M.X. Xie, J. Kang, D. Zheng, Studies on the interaction of total saponins of Panax notoginseng and human serum albumin by Fourier transform infrared spectroscopy, Spectrochim. Acta, Part A 59 (2003) 2747 2758. [28] M.R. Eftink, C.A. Ghiron, Exposure of tryptophanyl residues in proteins quantitative determination by fluorescence quenching studies, Biochemistry 15 (1976) 672 680. [29] S. Lehres, Solute perturbation of protein fluorescence: the quenching of the tyrpophyl fluorescence of model compounds and of lysozyme by zodide ion, Biochemistry 10 (1971) 3254 3263. [30] P.R. Callis, 1La and 1Lb transitions of tryptophan: applications of theory and experimental observation to fluorescence of proteins, Methods Enzymol. 237 (1997) 113 150. [31] B. Kljnert, L. Stanislawska, M. Bryszewska, B. Palecz, Interactions between PAMAM dendrimers and bovine serum albumin, Biochim. Biophys. Acta 1648 (2003) 115 126. [32] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, 2nd ed., Kluwer Academic/Plenum Publishers, New York, 1999. [33] E.L. Gelamo, C.H.T.P. Silva, H. Imasato, M. Tabak, Interaction of bovine and human serum albumins with ionic surfactants: spectroscopy and modeling, Biochim. Biophys. Acta 1594 (2002) 84 99. [34] M.R. Eftink, Fluorescence quenching reactions: probing biological macro-molecular structures, in: T.G. Dewey (Ed.), Biophysical and Biochemical Aspects of Fluorescence Spectroscopy, Plenum, New York, 1991, pp. 105 133. [35] W.R. Ware, Oxygen quenching of fluorescence in solution: an experimental study of the diffusion process, J. Phys. Chem. 66 (1962) 455 458.