Interaction of human serum albumin with bendroflumethiazide studied by fluorescence spectroscopy

Journal of Photochemistry and Photobiology B: Biology 80 (2005) 139–144 www.elsevier.com/locate/jphotobiol

Interaction of human serum albumin with bendroflumethiazide studied by fluorescence spectroscopy Yue Hong Pang a, Li Li Yang a, Shao Min Shuang Michael Thompson b a

a,*

, Chuan Dong a,

Institute of Advanced Chemistry, College of Chemistry and Chemical Engineering, Shanxi University, Taiyuan 030006, PR China b Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, Ont., Canada M5S 3H6 Received 20 September 2004; received in revised form 4 February 2005; accepted 8 March 2005 Available online 23 May 2005

Abstract The interactions between bendroflumethiazide (BFTZ) and human serum albumin (HSA) have been studied by fluorescence spectroscopy. Binding constants for drug attachment to the various binding sites of HSA have been measured at different temperatures in physiological buffer solution. The effect of metal ions on BFTZ interaction with HSA was also investigated. The thermodynamic parameters, DH and DS, have been calculated to be 49.28 kJ mol
1 > 0, and 258.83 J mol
1 K
1 > 0, respectively. The distance between HSA and BFTZ, r, was determined to be 1.47 nm based on Fo¨rsterÕs non-radiative energy transfer theory. The experimental results reveal that BFTZ has a strong ability to quench the intrinsic fluorescence of HSA through a static quenching mechanism. Furthermore, the binding constants between BFTZ and HSA are remarkably independent of temperature, and decrease in the presence of various ions, usually by about 30–55%. Hydrophobic interaction occurs between BFTZ and the sub-domain II A of HSA.
2005 Elsevier B.V. All rights reserved. Keywords: Human serum albumin; Bendroflumethiazide; Steady-state fluorescence

1. Introduction Bendroflumethiazide (BFTZ, Fig. 1) is one of a number of thiazide diuretics that display significant clinical effects such as anti-hypertensive and diuretic behavior. The analysis of these compounds has gained increasing attention since Novello and Sprague [1] observed a diuretic effect with regard to chlorothiazide. Various analytical techniques such as HPLC, GC, and capillary electrophoresis coupled to UV or mass spectrometry detectors have been employed in the study of the chemistry of these molecules. The use of LC/MS/MS has especially proven to be an appropriate means to provide *

Corresponding author. Fax: +86 0351 7011688. E-mail address: [email protected] (S.M. Shuang).

1011-1344/$ - see front matter
2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jphotobiol.2005.03.006

detailed information concerning thiazide-based diuretics and other compounds belonging to this class of remedy [2]. The lowest therapeutic dose with this agent is 2.5 mg, whereas for the highly effective diuretic, the accepted dose is 2.5–10 mg. When it is used in the case of mild-to-moderate hypertension, the accepted dose is 2.5–20 mg [3]. The compound is transported in the blood while bound to albumin (often more than 94% of the drug is attached to the protein [3]). This behavior is typical of many drugs, which critically governs their distribution and pharmacokinetics [4–9]. Serum albumins are the most abundant proteins in plasma constituting 52% of the protein composition in this matrix. As the major soluble protein constituents of circulatory system, they possess many physiological functions and play a key role in the transport of many

140

Y.H. Pang et al. / Journal of Photochemistry and Photobiology B: Biology 80 (2005) 139–144

O NH2O 2S

2. Experimental

O

S NH

F 3C

N H

2.1. Apparatus CH2

Fig. 1. The structure of bendroflumethiazide (BFTZ).

endogenous and exogenous ligands. Human serum albumin (HSA) is most important and abundant constituent of blood plasma and serves as a protein storage component. Recently, the three-dimensional structure of HSA was determined through X-ray crystallographic measurement [10]. The protein contains 585 amino acids with known sequences [11] and a molecular weight of 66,500 Da [12]. Crystal structure analysis shows that the drug binding sites, I and II, are located in subdomains II A and III A, respectively. A large hydrophobic cavity is present in the II A sub-domain. Warfarin, phenylbutazone and iodipamide are typical site I (sub-domain II A) binding molecules [13]. HSA serves as a transport carrier for a variety of small species, such as fatty acids, cations and many diverse drugs [14,15], present in the systematic circulation, due to its very unique single-polypeptide globular multidomain structure [10]. Furthermore, the protein renders the feasibility to bind and carry through the bloodstream many drugs, which are poorly soluble in water [16]. It has been shown that the distribution, free concentration and the metabolism of various molecules can be significantly altered as a result of their binding to HSA [17]. Therefore, a study of the interaction of BFTZ with HSA is of major biochemical importance, and can be used as a model for an elucidation of BFTZ–protein complexation. Measurement of the quenching of the natural fluorescence of albumin is an important method for the study of its interaction with several substances [18]. This technique can reveal the accessibility of quenchers to fluorophore groups in the protein, provide an understanding of binding mechanisms to drugs, and yield clues as to the chemistry of the binding phenomenon. The fluorescence quenching technique has been widely applied in the investigation of drug–protein binding interactions because of the presence of aromatic amino acid residues such as tryptophan [19,20]. The aim of this study was to examine the binding reaction and the effect of energy transfer between BFTZ and HSA, as well as the effect of temperature and metal ions by the fluorescence quenching method. A possible mechanism for the interaction is proposed where the intrinsic fluorescence of HSA had been quenched by BFTZ through a static quenching procedure. The binding sites for BFTZ are in a hydrophobic pocket in the sub-domain II A of HSA.

All fluorescence measurements were carried out on a Hitachi Model F-4500 spectrofluorometer (Kyoto, Japan) equipped with a 150 w xenon lamp source, a thermostat bath and a quartz cells (1 · 1 cm). A Shimadzu UV-265 double-beam spectrophotometer (Tokyo, Japan), equipped with quartz cells (1 · 1 cm) was used for scanning the absorption spectra. All pH values were measured with a pHS-2 acidometre (The Second Instrument Factory of Shanghai, China). 2.2. Reagents Bendroflumethiazide (BFTZ, Beijing Biological Medicine Research Institute) and human serum albumin (HSA, Biological Identification Institute of Shanghai) were used. 4.00 · 10
4 mol/L solutions of various metal ions and 0.1 mol/L Tris–HCl buffer solutions of pH = 7.4 were prepared. All chemicals used were of reagent grade and used without further purification. In order to standardize the ionic strength, 0.1 mol/L sodium chloride (NaCl) was used for the preparation of all solutions. All stock solutions were stored at 0–4 C. 2.3. General procedure A 0.5 ml aliquot of the stock solution (2 · 10
4 mol/L) of HSA was transferred into a 5 ml volumetric flask, and an appropriate amount (0, 0.1, 0.3, 0.5, 0.8, 1.5, 2.0 ml) of 5 · 10
5 mol/L BFTZ was added. The pH was controlled by 0.1 mol/L Tris–HCl buffer solutions. The mixed solution was diluted to the mark with 0.1 mol/L sodium chloride solution and shaken thoroughly, then equilibrated for 30 min at 20 ± 1 C. Fluorescence spectra were recorded or fluorescence intensities were measured. In addition, in the presence of 5.0 · 10
6 mol/L metal ions, other fluorescence spectra could be obtained.

3. Results and discussion 3.1. Binding constant of HSA and BFTZ Fig. 2 shows the fluorescence spectrum of HSA in pH = 7.4 butter solution. The maximum excitation and emission wavelengths are at 290 and 356 nm, respectively. Addition of different concentrations of BFTZ caused a noticeable decrease in HSA fluorescence intensity. The maximum emission wavelength produces a small blue shift from 356 to 353 nm and the corresponding excitation wavelength is slightly red-shifted from 290 to 293 nm. The strong quenching of Trp 214 fluorescence

Y.H. Pang et al. / Journal of Photochemistry and Photobiology B: Biology 80 (2005) 139–144

141

namic quenching constant, the average lifetime of the biomolecule without quencher and the concentration of quencher, respectively. Obviously, K SV ¼ K q s0 .

Fig. 2. The fluorescence quenching spectrum of 2 · 10
5 mol/L HSA at pH = 7.4 in various concentrations of BFTZ CBFTZ (mol/L): (1) 0; (2) 1.0 · 10
6; (3) 3.0 · 10
6; (4) 5.0 · 10
6; (5) 8.0 · 10
6; (6) 15.0 · 10
6; (7) 20.0 · 10
6.

indicates that the HSA conformation may be changed and that intermolecular energy transfer occurs between BFTZ and HSA. Static quenching and dynamic quenching vary temperature. The quenching rate constants decrease with increasing temperature for static quenching, but the reverse effect is observed for the case of dynamic quenching [21]. A possible quenching mechanism is evident from the Stern–Volmer curves of HSA with BFTZ at different temperatures (20 C, 35 C, and 45 C) as shown in Fig. 3. The Stern–Volmer plots are linear with the slopes decreasing with increasing temperature. This indicates the occurrence of a static quenching interaction between BFTZ and HSA. In order to invoke this possibility, the mechanism is assumed to involve dynamic quenching. The quenching equation is presented by [21]: F 0 =F ¼ 1 þ K q s0 ½Q ¼ 1 þ K SV ½Q;

ð1Þ

where F and F0 are the fluorescence intensity with and without quencher, respectively. Kq, KSV, s0 and [Q] are the quenching rate constant of the biomolecule, the dy-

Fig. 3. The Stern–Volmer plots for the quenching of HSA by BFTZ at 20 C, 35 C and 45 C.

ð2Þ

Because the fluorescence lifetime of the biopolymer is 10
8 s [22], KSV is the slope of linear regressions of Fig. 3. According to Eq. (2), the quenching constant Kq can be calculated and is listed in Table 1 together with the appropriate correlation coefficients. It is noted from Table 1 that the quenching constant Kq decreases from 3.3 · 1013 L mol
1 s
1 to 2.4 · 1013 L mol
1 s
1. The order of magnitude of the quenching constant Kq is 1013. However, the maximum scatter collision quenching constant Kq of various quenchers with the biopolymer is 2 · 1010 L mol
1 s
1 [23]. Clearly, the rate constant of the protein quenching procedure initiated by BFTZ is greater than the Kq for the scatter mechanism. Accordingly, this implies that the quenching is not initiated by dynamic collision, but originates from the formation of a complex. The static quenching equation can be expressed as follows [24]: lg

ðF 0
F Þ ¼ lg K A þ n lg ½Q. F

ð3Þ

From a plot of lg ðF 0F
F Þ versus lg [Q], the binding constant of BFTZ with HSA (KA) and the binding sites (n) can be obtained from the intercept and the slope. Table 1 depicts the binding constant (KA) and the binding sites (n) between BFTZ and HSA. They exhibit good linearity. The values of the binding sites between BFTZ and HSA were all 1, the binding constants between them are remarkable and the effect of temperature is small. This suggests that there is a strong interaction and the formation of a complex between BFTZ and HSA. This clearly implies that BFTZ can be stored and removed by the proteins in the body. 3.2. Evaluation of the interaction between BFTZ and HSA The forces acting between a drug and a biomolecule are composed weak interactions of molecules such as hydrogen bond formation, van der Waals forces, electrostatic forces, and the hydrophobic interaction [25]. From the thermodynamic standpoint, DH > 0 and DS > 0 implies a hydrophobic interaction; DH < 0 and DS < 0 reflects the van der Waals force or hydrogen bond formation; with DH  0, and DS > 0 suggesting an electrostatic force [26]. Because the temperature changes are minimal, the interaction enthalpy change can be regarded as a constant. Therefore, from the following equations:
 k2 1 1 DH ln ¼ ; ð4Þ
T1 T2 R k1

142

Y.H. Pang et al. / Journal of Photochemistry and Photobiology B: Biology 80 (2005) 139–144

Table 1 The dynamic quenching constants, binding constants and binding sites of BFTZ and HSA T (C)

Kq · 10
13a (L mol
1 s
1)

r

KA · 10
5b (L mol
1)

nb

r0

20 35 45

3.3 2.9 2.4

0.9917 0.9943 0.9856

0.811 1.451 2.658

1.1554 0.9616 1.0265

0.9970 0.9965 0.9881

a b

After Stern–Volmer method [19], Eqs. (1) and (2). After static quenching equation [22], Eq. (3).

DG ¼ DH
T DS;

ð5Þ

DG ¼
RT ln K;

ð6Þ

where DH and DS are calculated to be 49.28 kJ mol
1 and 258.83 J mol
1 K
1. Since the values of both of these parameters are greater than zero it can be deduced that the force acting in the binding reaction between BFTZ and HSA is mainly of the hydrophobic type. 3.3. Energy transfer between BFTZ and HSA According to the Fo¨rster non-radiative energy transfer theory [27,28], energy transfer will occur under the following conditions: (I) the donor can produce fluorescent light; (II) the fluorescence emission spectrum of the donor and UV absorbance spectrum of the acceptor overlap; and (III) the distance of approach between the donor and the acceptor is lower than 7 nm. HSA is a donor and it has a strong intrinsic fluorescence. It can be seen from Fig. 4 that energy transfer occurs with facility between BFTZ and HSA in view of the large spectral overlap between the fluorescence emission spectrum of HSA (1) and the UV absorption spectrum of BFTZ (2). The energy transfer effect is related not only to the distance between the acceptor and the donor, but also to the critical energy transfer distance, The relation between these factors is E¼

ðR0 Þ 6

at the acceptor at wavelength k. The energy transfer efficiency is given by: E ¼1


F ; F0

ð10Þ

where J can be evaluated by integrating the spectra in Fig. 4 and Eq. (9) for k = 300–400 nm and is to be 5.86 · 10
16 cm3 L mol
1. Under these experimental conditions, we found a characteristic distance of R0 = 1.55 nm, using k2 = 2/3, n = 1.336, / = 0.13 [29]. The energy transfer effect E = 0.32 from Eq. (10) and the maximum distance between BFTZ and amino acid residue in HSA, r is 1.47 nm. This confirms that the energy transfer between BFTZ and HSA contributes to the noticeable decrease of HSA fluorescence intensity. 3.4. The effect of other ions on the binding constant between drug and protein Trace metal ions, especially the bivalent type, are essential in the human body and play an important structural role in many proteins based on the coordinate

6

ðR0 Þ þ ðr0 Þ

6

;

ð7Þ

where r is the distance between the acceptor and the donor, and R0 is the critical distance when the transfer efficiency is 50%, which, in turn, can be calculated by: ðR0 Þ6 ¼ 8.8  10
25 k 2 /n
4 J ;

ð8Þ

2

where k is the spatial orientation factor of the dipole, / the fluorescence quantum yield of the donor, n the refractive index of the medium, and J the overlap integral of the fluorescence emission spectrum of the donor and the absorption spectrum of the acceptor. Therefore, P F ðkÞeðkÞk4 Dk P J¼ ; ð9Þ F ðkÞDk where F(k) is the fluorescence intensity of the fluorescent donor at wavelength k, and e(k) the molar absorptivity

Fig. 4. Overlap of the fluorescence emission spectra of HSA (1) and absorption spectra of BFTZ (2). Experimental conditions: HSA, 1.0 · 10
5 mol/L and BFTZ, 1.0 · 10
5 mol/L, pH = 7.4.

Y.H. Pang et al. / Journal of Photochemistry and Photobiology B: Biology 80 (2005) 139–144

bond. Some references report that Zn, Mg, Mn and Co and other metal ions can form complexes with HSA [30– 32]. There are four independent regions responsible for the binding of metal ions to serum albumin [33]: (1) the Cu(II)–Ni(II) site, at the N-terminal side of the protein, (2) the Cd(II)–Zn(II) site at three histidines (His105, His146 and His247) and one aspartate (Asp249 residues, (3) a second site for Cd(II), undetermined, and (4) at the single-free thiol group of the Cys 34 residue for Hg(II), Au(I) and Ag(I). The presence of metal ions can influence directly the binding of drugs with proteins. Therefore, the effect of these metal ions on the interaction between HSA and BFTZ was investigated. As shown in Fig. 5, the absorption spectra of BFTZ recorded with increasing concentrations of Co(II) shows almost no change in the absorption intensities. Additionally, the absorption maximum does not exhibit a shift indicating that the presence of Co (II) has no influence on BFTZ. Similar phenomena are also observed for Mg (II), Zn(II), Mn(II). Therefore, it can be concluded that there is no interaction between the metal ions and BFTZ. In an additional study of the influence of metal ions on the drug–protein interaction, the effects of various metal ions on the binding constants was investigated at room temperature. The results are summarized in Table 2. In this regard, the presence of the metal ions, Mg(II), Co(II), Zn(II), Mn(II), all decrease the binding constants by 30–55%. This is likely caused by a conformational change in the vicinity of the binding site. Since such sites for the drug and metal ion for HSA are not located in the same domain, there is no direct competition between the drug and the metal ions. However,

143

Table 2 The binding constants (L mol
1) between BFTZ and HSA at room temperature in the presence of metal ions K 0A  10
6 K 0A =K A 1
ðK 0A =K A Þ

Mg(II)

Co(II)

Zn(II)

Mn(II)

1.6250 0.6090 0.3910

1.3418 0.5028 0.4972

1.8390 0.6891 0.3109

1.2583 0.4715 0.5285

the formation of metal ion–HSA complexes is likely to effect changes in the conformation of the protein, which may affect drug-binding kinetics and could even inhibit drug–HSA interaction. This will decrease the interaction between the drug and HSA resulting in a shortening the storage time of the drug in blood plasma together with an enhancement of the maximum effectiveness of the drug.

4. Conclusion The fluorescence method is highly sensitive and convenient to use in the study of intermolecular interactions. HSA solutions were excited at 290 nm in order to emit fluorescence attributable mainly to tryptophan residues. The single tryptophan residue of HSA is located at position 214 in subdomain II A. The study of the interaction of the BFTZ and HSA showed that a complex was formed between these species through the static quenching interaction. The values of the binding sites were all 1 at different temperatures. The binding sites for BFTZ are in a hydrophobic pocket at Site I of HSA within sub-domain II A. Experiments reveal that the binding constants (KA) between BFTZ and HSA are 0.811 · 105, 1.451 · 105 and 2.658 · 105 M
1 at 20 C, 35 C, 45 C. Compared with analogous values for the typical site I binding drugs, salicylate (1.9 · 105 M
1 [34]), warfarin (3.0 · 105 M
1 [35]) and flavonoids (from 104 to 106 M
1 [36,37]), it appears that the binding constants of the site I binding drugs are in the range of 104–106 M
1. These findings also indicate that site I ligands are typically bulky heterocyclic anions with charge being situated in a fairly central position with site I being considered a large flexible site. These experimental and theoretical data are of potential importance in understanding the mechanism of interaction between HSA and thiazide drugs and, accordingly, be a useful guide for efficient thiazide drug research.

Acknowledgements Fig. 5. Absorption spectra of Co(II) and BFTZ between 200 and 700 nm in Tris–Hcl buffer solution pH = 7.4 at room temperature (1) 5.0 · 10
6 mol/L Co(II); (2) 5.0 · 10
6 mol/L BFTZ; (3) 5.0 · 10
6 mol/L BFTZ + 5.0 · 10
6 mol/L Co(II); (4) 5.0 · 10
6 mol/L BFTZ + 1.0 · 10
5 mol/L Co(II); (5) 5.0 · 10
6 mol/L BFTZ + 2.0 · 10
5 mol/L Co(II).

This project received support from the National ence Foundation of China (NNSFC: 20172035 20275022). Additional support as provided by National Education Committee Foundation for

Sciand the the

144

Y.H. Pang et al. / Journal of Photochemistry and Photobiology B: Biology 80 (2005) 139–144

Outstanding Younger Teachers and the Foundation of Shanxi Province. All the authors express their deep thanks. References [1] F.C. Novello, J.M. Sprague, J. Am. Chem. Soc. 79 (1957) 2028– 2029. [2] M. Thevis, H. Schmickler, W. Scha¨nzer, Anal. Chem. 74 (2002) 3802–3808. [3] X.Q. Chen, Y.Y. Jin, G. Tang, X.B. Yao, W. Xue, fifteenth ed., PeopleÕs Medical Publishing House, Beijing, 2003. [4] C.Q. Jiang, M.X. Gao, J.X. He, Anal. Chim. Acta 452 (2002) 185–189. [5] F.L. Cui, J. Fan, J.P. Li, Z.D. Hu, Bioorg. Med. Chem. 12 (2004) 151–157. [6] B.C. Channu, H.N. Kalpana, G.B. Eregowda, C. Dass, P.J. Houghton, K.N. Thimmaiah, J. Pharm. Biomed. Anal. 21 (1999) 775–785. [7] J. Krieglstein, W. Meiler, J. Staab, Biochem. Pharmacol. 21 (1972) 985–997. [8] B. Klajnert, L. Stanislawska, M. Bryszewska, B. Palecz, Biochim. Biophys. Acta 1648 (2003) 115–126. [9] A.A. Bhattacharya, S. Curry, N.P. Franks, J. Biol. Chem. 275 (2000) 38731–38738. [10] X.M. He, D.C. Carter, Nature 358 (1992) 209–215. [11] A. Dugiaczyk, S.W. Law, O.E. Dennison, Proc. Nat. Acad. Sci. USA 79 (1982) 71–75. [12] D.C. Carter, X.M. He, S.H. Munson, P.D. Twigg, K.M. Gernert, M.B. Broom, T.Y. Miller, Science 244 (1989) 1195–1198. [13] I. Petitpas, A.A. Bhattacharya, S. Twine, M. East, S. Curry, J. Biol. Chem. 276 (2001) 22804–22809. [14] A.M.L. Zaton, J.P. Villamor, Chemico-Biol. Interact. 124 (2000) 1–11. [15] D.C. Carter, J.X. Ho, Adv. Protein Chem. 45 (1994) 153–201.

[16] T. Peters, Adv. Protein Chem. 37 (1985) 161–242. [17] U. Kragh-Hansen, Pharmacol. Rev. 33 (1981) 17–53. [18] M. Bhattacharyya, U. Chaudhuri, R.K. Poddar, Bioch. Biophys. Res. Commun. 167 (1990) 1146–1153. [19] E.S. Souza, I.Y. Hirata, L.E. Juliano, A.S. Ito, Biochim. Biophys. Acta 1474 (2000) 251–261. [20] T. Tominaga, V.E. Yushmanov, I.E. Borissevitch, H.E. Imasato, M. Tabak, J. Inorg. Biochem. 65 (1997) 235–244. [21] G.Z. Chen, X.Z. Huang, J.G. Xu, Z.Z. Zheng, Z.B. Wang, Methods of Florescence Analysis, second ed., Science Press, Beijing, 1990. [22] J.R. Lakowica, G. Weber, Biochemistry 12 (1973) 4161–4170. [23] W.R. Ware, J. Phys. Chem. 66 (1962) 455–458. [24] M. Alain, B. Michel, D. Michel, J. Chem. Edu. 63 (1986) 365–366. [25] L.H. Nie, H.C. Zhao, X.B. Wang, X. Wang, J. Beijing Normal University (Nat. Sci.) 37 (2001) 87–91. [26] D.P. Ross, S. Subramanian, Biochemistry 20 (1981) 3096–3102. [27] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum Press, New York, 1983. [28] L.A. Sklar, B.S. Hudson, R.D. Simoni, Biochemistry 16 (1977) 5100–5108. [29] M. Suzukida, H.P. Le, F. Shahid, R.A. Mcpherson, E.R. Birnbaum, D.W. Darnall, Biochemistry 22 (1983) 2415–2420. [30] O. Emiko, H. Yoshito, N. Kana, K. Susumu, J. Inorg. Biochem. 75 (1999) 213–218. [31] E.N. Chikvaidze, Biofizika 33 (1988) 723–725. [32] D. Zha, S. Li, Y. Yang, C. Tu, H. Liang, Spectrosc. Spectr. Anal. 19 (1999) 788–791. [33] S. Aime, S. Canton, S.G. Crich, E. Terreno, Magn. Reson. Chem. 40 (2002) 41–48. [34] U. Kragh-Hansen, Mol. Pharmacol. 34 (1988) 160–171. [35] K. Yamasaki, T. Maruyama, U. Kragh-Hansen, M. Otagiri, Biochim. Biophys. Acta 1295 (1996) 147–157. [36] Z. Ferenc, B. Zsolt, S. Miklos, Biochem. Pharmacol. 65 (2003) 447–456. [37] S. Bi, L. Ding, Y. Tian, D. Song, X. Zhou, X. Liu, H. Zhang, J. Mol. Struct. 703 (2004) 37–45.