- Email: [email protected]
Study on the binding of chiral drug duloxetine hydrochloride to human serum albumin
European Journal of Medicinal Chemistry 45 (2010) 4043e4049
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Study on the binding of chiral drug duloxetine hydrochloride to human serum albuminq Xiangping Liu a, Yingxiang Du a, b, c, * a
Department of Analytical Chemistry, China Pharmaceutical University, Nanjing 210009, China Key Laboratory of Drug Quality Control and Pharmacovigilance (Ministry of Education), China Pharmaceutical University, Nanjing 210009, China c Key Laboratory of Modern Chinese Medicines (Ministry of Education), China Pharmaceutical University, Nanjing 210009, China b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 June 2009 Received in revised form 24 May 2010 Accepted 27 May 2010 Available online 4 June 2010
Duloxetine holds a special promise as an antidepressant, and its effect depends on its binding to human serum albumin (HSA). For this reason, the binding mechanism of duloxetine with HSA was investigated. The specific binding site in HSA was identified and binding constants were determined. Duloxetine could compete with dansyl-L-proline (DLP), a site II marker for binding to site II. Binding constants between duloxetine and HSA were 1.75 103 L mol1 and 3.74 103 L mol1 at pH 7.4 and pH 8.5, respectively. The interaction process of enantiomers and HSA was susceptible to pH change. It was concluded that specific binding position of duloxetine was located in site II, and the B conformation of HSA possibly excelled the N conformation in identifying and binding to enantiomers. Ó 2010 Elsevier Masson SAS. All rights reserved.
Keywords: Duloxetine hydrochloride HSA Fluorescence technique Affinity capillary electrophoresis
1. Introduction Duloxetine hydrochloride (Fig. 1), S-(þ)-N-methyl-3-(1-napthalenyloxy)-3-(2-thienyl) propanamine hydrochloride, holds a special promise as an antidepressant drug as it is a dual inhibitor of serotonin and norepinephrine reuptake [1e3]. It is also used for stress urinary incontinence. Various features, like improved efficacy, tolerability, safety, faster recovery, fewer side effects, low affinity for neuronal receptors and dual inhibiting nature give duloxetine an edge over other existing antidepressants, such as fluoxetine and tomoxetine. Whereas its enantiomer, R-()-Nmethyl-3-(1-napthalenyloxy)-3-(2-thienyl) propanamine hydrochloride, is pharmacologically invalid and not used clinically. Most organic compounds in the blood are carried while bind to albumin. As the most abundant protein in the blood, HSA plays an important role in the transport of many endogenous substances and exogenous compounds (e.g. drugs). The distribution, metabolism and free concentration of a drug in the blood are closely related to its interaction with HSA [4]. HSA has 585 amino acid
q Supported by Program for New Century Excellent Talents in University (No.: NCET-06-0498) and the Key Project of Chinese Ministry of Education (No.: 109085). * Corresponding author. Department of Analytical Chemistry, China Pharmaceutical University, No. 24 Tongjiaxiang, Nanjing, Jiangsu 210009, China. Tel./fax: þ86 25 83221790. E-mail address: [email protected] (Y. Du). 0223-5234/$ e see front matter Ó 2010 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.ejmech.2010.05.063
residues, and consists of three homologous domains (labeled IeIII), each of which is divided into two subdomains [5]. Among these subdomains site I (in subdomain IIA) and site II (in subdomain IIIA) have physiological importance in binding to a large number of drugs [6]. Drugs of the same binding sites in HSA will compete with one another for binding to the protein, which affects their distribution and metabolism in the blood [7]. Therefore, it is important to identify the binding site of a drug in HSA. The conformation of HSA also affects the interaction process between small molecule and HSA. In the normal physiological condition, HSA undergoes conformational change known as NeB (NeutraleBasic) transition, which will affect the binding characteristics between a drug and HSA [8,9]. Some drugs are inclined to interact with the N conformation rather than the B conformation whereas others are just the opposite [10]. Recent observations have clearly recognized the fact that albumin is not a biologically inactive protein, and that it may, in fact, play an important regulatory role in the functioning of certain cell types via interaction with the cellular membrane [11e13]. During such interactions, the albumin molecule undergoes NeB-like conformational transition and adopts the B conformation at membraneewater interface [14,15]. Harmsen et al. [16] have also suggested that under increased Ca2þ concentration the B conformation predominates in blood plasma. All these studies have demonstrated that the NeB transition of HSA has physiological significance. Meanwhile, HSA is a good chiral selector to many chiral compounds [17], which influences the transport of
4044
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
Fig. 1. The structure of duloxetine hydrochloride.
chiral drugs in the blood. Duloxetine, as a chiral drug, its interaction with HSA has not been reported until now. In this work, the binding process between duloxetine and HSA was investigated, and the effect of NeB transition of HSA was also taken into account. Many techniques have been applied to examine interactions between small molecule and macromolecule, including fluorescence spectroscopy, capillary electrophoresis [18e24]. In this study, the interaction between duloxetine enantiomer and HSA was investigated by synchronous fluorescence technique, for it could distinguish the overlapped fluorescence spectra by fixing an appropriate scanning interval (Dl ¼ lemission lexcitation) [25]. The common method based on fluorescence quenching cannot work here because, as has been described in our previous work [26], the fluorescence emission spectrum of duloxetine overlapped with that of HSA mostly. The competitive binding experiment is useful in identifying the specific binding site of small drug in protein, and it has been applied to confirm many drugs’ specific binding sites in protein [6,27,28]. In present work the binding site of duloxetine in HSA was identified by monitoring the fluorescence change of markereHSA complex in the presence of duloxetine. Affinity capillary electrophoresis (ACE) refers to the separation of substances that participate in specific or non-specific affinity interactions during CE. In the past two decades, ACE has been one of the most rapidly growing analytical techniques for studying intermolecular interactions and determining binding constants [29e31]. It has advantages in low sample consumption, fast separation and diverse operation modes. In this study, binding constants between duloxetine and HSA were determined by mobility-shift ACE method. The enantioselectivity of HSA to duloxetine enantiomers mixture was also examined by ACE under the near-physiological condition, which was a complement to the preceding results.
Fig. 2. Tryptophan synchronous fluorescence spectra (Dl ¼ 60 nm) in different systems. At pH 7.4 1: HSA and 2: HSA þ 6.4 105 mol L1 duloxetine; at pH 8.5 3: HSA and 4: HSA þ 6.4 105 mol L1 duloxetine.
2.1.1. The tryptophan-214 synchronous fluorescence When Dl is 60 nm, it was observed that the fluorescence intensity was enhanced and the maximum wavelength shifted from the long wavelength to the short wavelength (shown in Fig. 2) with
2. Results 2.1. Synchronous fluorescence experiment Among 585 amino acid residues of HSA, tryptophan-214 and tyrosine-411 can emit nature fluorescence upon the excitation [32]. When Dl is 60 nm, the spectrum characteristic of tryptophan-214 is observed; when Dl is 15 nm, the spectrum characteristic of tyrosine-411 is observed [33].
Fig. 3. The effect of duloxetine on tryptophan synchronous fluorescence spectra (Dl ¼ 60 nm), a: maximum wavelength; b: F0/F; pH 7.4 (solid); pH 8.5 (dashed line).
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
4045
Fig. 4. The effect of duloxetine on tyrosine synchronous fluorescence spectra (Dl ¼ 15 nm) in pH 7.4 buffer solution with different duloxetine concentration (105 mol L1): (1) 0, (2) 1.6, (3) 3.2, (4) 4.8, (5) 6.4, (6) 8.0.
the increase of the enantiomer concentration while keeping the concentration of HSA unchanged. Effects of added duloxetine on the spectra of HSA are shown in Fig. 3. As shown in Fig. 3(a), at pH 7.4, the added duloxetine induced the maximum wavelength shift from 341 to 339 nm, while at pH 8.5, it induced the maximum wavelength shift from 341 to 338 nm. In Fig. 3(b) the symbol F0 represented fluorescence intensity in the absence of the drug and the symbol F was fluorescence intensity in the presence of the drug (in the following, all the symbols F0 and F had the same meaning). The F0/F values were smaller than 1.0, which meant duloxetine enhanced the fluorescence of HSA. It was also observed that all F0/F values at pH 8.5 (dashed line) were smaller than those at pH 7.4 (solid), which showed that the enhancement of fluorescence intensity in basic solution was greater than that in neutral solution. The effect of R-isomer on the tryptophan synchronous fluorescence spectra was also investigated in the same way. In the neutral solution (pH 7.4), the change of fluorescence spectra was similar to that induced by duloxetine. However, in basic solution (pH 8.5), both the wavelength shift and fluorescence intensity enhancement were less than those induced by duloxetine. 2.1.2. The tyrosine-411 synchronous fluorescence The effect of duloxetine on tyrosine synchronous fluorescence spectra is shown in Fig. 4. With the unchanged concentration of
Fig. 5. The effect of duloxetine on the synchronous fluorescence intensity at 300 nm, pH 7.4 (solid); pH 8.5 (dashed line).
Fig. 6. Effects of enantiomers on the fluorescence intensity at 323 nm, pH 7.4 (solid); pH 8.5 (dashed line); duloxetine (B); R-isomer (-).
HSA, as the duloxetine concentration increased, the fluorescence intensity at 300 nm decreased gradually, while the fluorescence intensities at 323 nm and 336 nm rose accordingly. Fig. 5 shows the effect of duloxetine on the synchronous fluorescence intensity at 300 nm. All the F0/F values were greater than 1.0, which indicated the fluorescence at 300 nm decreased with the increased concentration of duloxetine. At the same time, all the F0/F values at pH 8.5 (dashed line) were greater than those at pH 7.4 (solid), which indicated that the fluorescence decrease at pH 8.5 was greater than that at pH 7.4. The effect of R-isomer was similar to that of duloxetine. Effects of duloxetine and its R-isomer on the fluorescence intensity at 323 nm were investigated, respectively. Results are shown in Fig. 6. When pH value of solution was changed from 7.4 to 8.5, the fluorescence intensities at 323 nm were enhanced remarkably. It was also observed that at pH 7.4, the enhancement for duloxetine was similar to that for R-isomer, but at pH 8.5, it was greater than that for R-isomer. The similar results were acquired for fluorescence intensities at 336 nm. 2.2. Competitive binding experiment 5-dimethylaminonaphthalene-1-sulfonamide (DNSA) and DLP were chosen as specific markers for site I and site II [27], respectively. It is known that these two markers can specifically bind to HSA in certain position located in site I or site II, and the formed complex can emit characteristic fluorescence from 360 nm to 550 nm upon the excitation. A drug is added into the complex system. If the added drug quenches the characteristic fluorescence emission, it is considered that the drug binds to HSA competitively and has the same specific binding position as the marker [6,28]. Effects of enantiomers on the fluorescence intensity of DNSAeHSA complex are shown in Fig. 7. All the F0/F values at pH 7.4 were less than 1.0 and decreased with the increase of isomer amount, which implied that both duloxetine and R-isomer enhanced the fluorescence of DNSAeHSA and enantiomers could not compete with DNSA for binding to site I. At pH 8.5, F0/F values decreased first and then increased as the isomer concentration augmented, especially for duloxetine. This demonstrated that under higher concentration enantiomers presented some ability of unspecific binding to site I in the weak basic solution. The competition of enantiomers with site I marker resulted in the fluorescence decrease of HSAeDNSA complex.
4046
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
Fig. 7. Effects of enantiomers on the fluorescence intensity of DNSAeHSA at 474 nm, pH 7.4 (solid); pH 8.5 (dashed line); duloxetine (B); R-isomer (-).
Effects of enantiomers on the fluorescence intensity of DLPeHSA complex are demonstrated in Fig. 8. All F0/F values were larger than 1.0, which signified that both duloxetine and its Risomer had quenched the fluorescence of DLPeHSA complex. At pH 7.4, two enantiomers had similar abilities of quenching the fluorescence, whereas at pH 8.5, all F0/F values for duloxetine were larger than those for R-isomer. This denoted that duloxetine quenched the fluorescence stronger than R-isomer. 2.3. Determination of the binding constants In Fig. 9, electropherograms of HSA are shown in buffers containing duloxetine of different concentration. The isoelectric point of HSA is 4.7, so it is negatively charged in the buffer at pH 7.4, and is detected after DMSO (neutral marker) at the positive voltage. Binding constants of a small drug and protein can be determined by the mobility-shift ACE, in which the concentration change of a drug in the buffer can cause the mobility change of the protein due to the binding process between a drug and a protein. This method has specific advantage over other approaches, for example, the injected sample need not be highly purified. However, the
Fig. 8. Effects of enantiomers on the fluorescence intensity of DLPeHSA at 480 nm, pH 7.4 (solid); pH 8.5 (dashed line). duloxetine (B); R-isomer (-).
Fig. 9. Electropherograms of HSA in pH 7.4 buffers containing duloxetine of different concentration: (1) 0.0; (2) 100; (3) 200; (4) 300; (5) 400; (6) 500; (7) 600 (106 mol L1). The applied voltage was þ15.0 kV, and the capillary temperature was controlled at 298 K; the detection wavelength was 214 nm and the injection volume was 50 mbar/10 s.
fluctuation of the electroosmotic flow (EOF) may adversely influence the observed electrophoretic mobility. The introduction of mobility ratio (M) can avoid the influence of the EOF change, and the M value is independent of the voltage, the capillary length and solvent viscosity [20], so the electrophoretic mobility (mi) was replaced by M to determine binding constants. According to the migration times of neutral marker (teo) and protein (tp), the M values are calculated from Equ. (2) in Section 5.3.3.1. Fig. 10 shows curves plotted by 1/DM versus 1/[L]. The intercept and slope of the plotted curve were obtained, and binding constants can be calculated accordingly (listed in Table 1). 2.4. Chiral identification ability of HSA to enantiomers Electropherograms of a mixture of duloxetine and R-isomer were recorded at pH 8.5 (Fig. 11a) and pH 7.4 (Fig. 11b), respectively. The run buffer contained 1.0 105 mol L1 HSA, 0.05 mol L1
Fig. 10. Curves plotted by 1/DM versus 1/[L] based on the mobility-shift ACE. pH 7.4 (solid); pH 8.5 (dashed line); duloxetine (B); R-isomer (-).
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049 Table 1 The binding constants determined by the mobility-shift ACE at 298 K.
DuloxetineeHSA (pH ¼ 7.4) DuloxetineeHSA (pH ¼ 8.5) R-isomereHSA (pH ¼ 7.4) R-isomereHSA (pH ¼ 8.5)
Linear regression equation
Related coefficient
Binding constant (L mol1)
y ¼ 0.0047x 8.2087
r ¼ 0.9926
1.75 103
y ¼ 0.0022x 8.2221
r ¼ 0.9956
3.74 103
y ¼ 0.0046x 8.4914
r ¼ 0.9945
1.85 103
y ¼ 0.0054x 14.47
r ¼ 0.9923
2.68 103
TriseHCl with 0.10 mol L1 NaCl, which was the same as that used in the fluorescence examination. It was shown that duloxetine and R-isomer had different migration times at pH 8.5 while they were not separated at pH 7.4. The resolution was 0.71 between duloxetine and R-isomer in Fig. 11(a). 3. Discussion Results of synchronous fluorescence examination showed that the introduction of duloxetine and R-isomer had affected HSA conformation. The drug had induced synchronous fluorescence spectra change, such as the maximum wavelength and the fluorescence intensity. When the maximum wavelength shifts to the longer wavelength, it denotes that the vicinal environment of residue becomes more hydrophilic; on the other hand, when the maximum wavelength shifts to the shorter wavelength, it denotes that the vicinal environment of residue becomes more hydrophobic [34]. For tryptophan residue, the added enantiomers could induce
4047
the maximum wavelength shift to the shorter range, which implied that the microenvironment of tryptophan residue became more hydrophobic. For tyrosine residue, both two enantiomers could induce fluorescence decrease at 300 nm, owing to the fact that enantiomers had bound to tyrosine residue and thus quenched its fluorescence. It was also noticed that fluorescence at 323 and 336 nm appeared and their intensities increased as the enantiomer amount rose. Fluorescence at 323 and 336 nm should be ascribed to the added isomer. The presence of HSA could enhance the fluorescence at 323 and 336 nm, especially in weak basic solution (shown in Fig. 6.), which signified that there were new complex formation between enantiomers and HSA. The competitive binding experiment demonstrated that both two enantiomers had quenched the fluorescence of DLPeHSA complex, which denoted that they had competed with DLP for binding to site II and inhibited the formation of DLPeHSA complex. The dissociation of DLP from HSA brought about the fluorescence quenching at 480 nm. As for the DNSAeHSA system, this phenomenon of fluorescence quenching had not been observed. These results demonstrated that the specific binding position of two enantiomers was located in site II instead of site I. It is known that tyrosine-411 is located in site II and tryptophan-214 is located in site I [5,6]. Experimental results of synchronous fluorescence showed that enantiomers could bind to tyrosine-411 and quenched the fluorescence gradually, which were consistent with results of competitive binding experiments. Carter and He [35] have shown that HSA is a multi-domain molecule, and each of domains is not independent, but interacts. It accounts for the synchronous fluorescence change of tryptophan-214 and the fluorescence
Fig. 11. Electropherograms of a mixture of 1.0 104 mol L1 duloxetine and 1.0 104 mol L1 R-isomer in pH 8.5 (a) and pH 7.4 (b) run buffer. The applied voltage was þ12.0 kV, and the capillary temperature was controlled at 293 K; the detection wavelength was 220 nm and the injection volume was 50 mbar/5 s.
4048
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
enhancement of DNSAeHSA complex. Although enantiomers did not bind to site I specifically, they could also impact the vicinity environment of tryptophan by the existence of allosteric regulation between sites I and II [36]. According to fluorescence experiment results, the interaction process between the isomer and HSA was susceptible to solution pH change. In synchronous fluorescence experiment, it was observed that all spectrum changes at pH 8.5, such as the wavelength shift of tryptophan residue, the decrease of tyrosine residue fluorescence, and the enhancement of fluorescence at 323 nm (dashed line shown in Figs. 2, 3, 5 and 6), were greater than those at pH 7.4 (solid in Figs. 2, 3, 5 and 6). Moreover, the abilities of changing synchronous fluorescence spectra of HSA at pH 7.4 for two enantiomers were similar, but when pH value became 8.5 changes induced by duloxetine were greater than those induced by R-isomer. In competitive binding experiment, the regulation of pH to the binding process still existed. Duloxetine displayed some ability of unspecific binding to site I at pH 8.5, while there was no affinity at pH 7.4. As for site II, it was obvious that enantiomers had stronger ability of replacing DLP at pH 8.5 than at pH 7.4. At pH 7.4, enantiomers competed with DLP to similar extent, while at pH 8.5 they behaved differently. The fluorescence quenching induced by duloxetine was apparently greater than that induced by R-isomer. Binding constants provide a fundamental measure of the affinity of a drug to a protein. In this study binding constants between enantiomers and HSA were also different from each other at different pH. As illustrated in Table 1, binding constants obtained at pH 7.4 were smaller than those obtained at pH 8.5, which implied that interactions between isomer and HSA at pH 8.5 were stronger than those at pH 7.4. Furthermore, it was also found that binding constants (1.75 103 L mol1 for duloxetine and 1.85 103 L mol1 for R-isomer) were similar at pH 7.4, but at pH 8.5 the binding constant (3.74 103 L mol1) for duloxetine was apparently bigger than that (2.68 103 L mol1) for R-isomer. Results shown in Fig. 11 prove that the interaction was susceptible to solution pH change further. When HSA was added into the run buffer of different pH value and the mixture of two isomers was injected as sample, the partial enantioseparation of duloxetine and its R-isomer was obtained (Rs ¼ 0.71) at pH 8.5, while at pH 7.4, the separation was not observed. The longer migration time of duloxetine showed that it interacted with HSA stronger than R-isomer did. The susceptibleness to solution pH change can be ascribed to HSA conformation transition. It has been reported that we can obtain different conformation of HSA by regulating the solution pH value in vitro [37,38]. HSA exhibited N conformation in neutral solution and B conformation in weak basic solution. It was deduced that binding process changes due to pH value were possibly caused by HSA NeB conformation transition. According to results of fluorescence experiment and affinity capillary electrophoresis, two enantiomers interacted with the B conformation stronger and HSA in the B conformation had better enantioselectivity ability than the N conformation did. The 1H NMR spectra have proved that HSA is heart-shaped in N conformation, but elliptic in B conformation, and the elliptic structure is much looser than the hearted one [39]. According to our investigation, it can be deduced that the looser structure of HSA favored the interaction between duloxetine and HSA. Additionally, the pKa value of duloxetine is 10.0 [40], and pH increase could induce isomers ionization change, so the effect of isomers ionization change on the interaction process should not be ignored. 4. Conclusion Together with all the observations this study has shown that the specific binding site of duloxetine was located in site II of HSA, and
binding constants between duloxetine and HSA were 1.75 103 L mol1 at pH 7.4 and 3.74 103 L mol1 at pH 8.5, respectively. It was also demonstrated that interactions between enantiomers and HSA were susceptible to solution pH change, and the NeB transition of HSA can possibly affect its binding ability and enantioselectivity to duloxetine isomer. This study helps us understand the interaction between duloxetine and HSA in the body. Drugs of the same binding site in HSA will compete with one another for binding to the protein, which affects their distribution and metabolism in the blood. The identification of specific binding position in HSA is useful to examine the effect of other drugs on duloxetine when they are administrated together. Moreover, the fact that the binding of duloxetine to HSA was susceptible to the NeB transition showed some factors that could induce this transition, such as pH value, Ca2þ concentration and so on, should be controlled carefully when we investigated the interaction mechanism between duloxetine and HSA.
5. Experimental 5.1. Reagents Duloxetine hydrochloride (S-(þ)-N-methyl-3-(1-napthalenyloxy)-3-(2-thienyl) propanamine hydrochloride) (>99.5%) and its R-isomer (>99.5%) were supplied by Jiangsu Institute for Drug Control (Nanjing, China). Fatty acid-free (<0.005%) and globulinfree HSA, 5-dimethylaminonaphthalene-1-sulfonamide (DNSA) and dansyl-L-proline (DLP) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). The stock solutions of drugs were prepared by dissolving 0.0334 g of standard drug and diluting to 50 mL with double distilled water, the concentration was 2.0 103 mol L1. 1.0 105 mol L1 HSA solution was prepared based on its molecular weights of 68,000 in 0.05 mol L1 TriseHCl buffer solution (with 0.10 mol L1 NaCl). All of the prepared solutions were stored at 0e4 C. All other materials were of analytical reagent grade and double distilled water was used throughout. 5.2. Apparatus A Shimadzu RF-5301 spectrofluorimeter (Shimadzu, Japan) was used for the fluorescence spectra measurements equipped with a 1 cm path length quartz cell. In the synchronous fluorescence experiment both the excitation and the emission slits were all 3 nm. In the competitive binding experiment the excitation slit was 5 nm, and the emission slit was 5 nm for DNSAeHSA and 3 nm for DLPeHSA, respectively. Agilent capillary electrophoresis instrument equipped with diode array detector (Agilent Technology, USA) and a 35 cm 50 mm elasticity quartz capillary column (Yongnian Hebei province, China) was used in affinity capillary electrophoresis experiment. The value of pH is measured on a pH meter (Lei-ci pHS-25, Shanghai, China).
5.3. Procedures 5.3.1. Synchronous fluorescence examination 220 nm was set as the initial excitation wavelength. When Dl was fixed at 60 nm, synchronous fluorescence spectra of 1.0 105 mol L1 HSA were recorded with increase of drug amount; when Dl was fixed at 15 nm, the experiment was performed in the same way.
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
5.3.2. Competitive binding experiment Competitive binding experiments were performed by using fluorescent markers specific for site I and site II (site I: DNSA; site II: DLP), respectively. The marker/HSA concentration ratio was kept 1/5 in order to avoid nonspecific binding. Accordingly, concentrations of HSA and marker were 1.0 105 mol L1 and 2.0 106 mol L1, respectively. The binding site of duloxetine in HSA was investigated by monitoring fluorescence change of the markereHSA complex with the increase of the drug amount upon the excitation at 343 nm. 5.3.3. Affinity capillary electrophoresis (ACE) experiment 5.3.3.1. Determination of binding constants. The electrophoresis buffer consisted of 0.02 mol L1 TriseHCl (with 0.02 mol L1 NaCl) and drugs of different concentration (0; 100; 200; 300; 400; 500; 600 (106 mol L1)). The injection sample was 1.0 105 mol L1 HSA containing 0.1% dimethyl sulfoxide (DMSO). DMSO was used as neutral marker to correct the fluctuation of electroosmotic flow. Binding constants between small drug molecule and HSA were determined by the mobility-shift ACE. Assuming that the stoichiometry of the binding between the target and the ligand is 1:1, we can use a simple mode to determine the binding constant (shown in equ. (1)) [29].
1
mi mf
¼
1 1 1 $ þ mc mf Kb mc mf ½L
(1)
where mi is the electrophoretic mobility of the protein with different ligand concentration [L] in the buffer; mf is the electrophoretic mobility of the free protein in the absence of ligand; mc is the electrophoretic mobility of proteinedrug complex with the saturation of ligand; Kb is the binding constant between drug and protein. In order to correct the fluctuation of electroosmotic flow, the mobility ratio (M) was used to estimate binding constants.
mapp meof mi teo M ¼ ¼ ¼ 1 meof meof tp
(2)
where mapp and meof are apparent electrophoretic mobility and electroosmotic mobility, respectively; teo and tp are the migration time of neutral marker and protein with different ligand concentration, respectively. Equ. (1) can be converted to the equ. (3).
1
DM
¼
1 1 1 $ þ Kb $DMmax ½L DMmax
(3)
where DM is the change in mobility ratio as a function of [L], and DMmax is the maximum change in mobility ratio that can be achieved with the saturation of ligand for complexation. By plotting 1/DM versus 1/[L], the y intercept will give the value of 1/DMmax, and Kb ¼ intercept/slope. 5.3.3.2. Chiral identification ability of HSA to the enantiomers. The run buffer contained 1.0 105 mol L1 HSA, 0.05 mol L1 TriseHCl (with 0.10 mol L1 NaCl). A mixture of 1.0 104 mol L1 duloxetine and 1.0 104 mol L1 R-isomer was injected as sample. The
4049
pH value of run buffer was set as 7.4 and 8.5, respectively, and the electropherograms were recorded. All the above procedures were repeated at least 3 times. References [1] P. Mariappan, A. Alhasso, Z. Ballantyne, A. Grant, J. N’Dow, Eur. Urol. 51 (2007) 67e74. [2] D.E. Stewart, M.M. Wohlreich, C.H. Mallinckrodt, J.G. Watkin, S.G. Kornstein, J. Affect. Disord. 94 (2006) 183e189. [3] F.P. Bymaster, E.E. Beedle, J. Findlay, P.T. Gallagher, J.H. Krushinski, S. Mitchell, D.W. Robertson, D.C. Thompson, L. Wallace, D.T. Wong, Bioorg. Med. Chem. Lett. 13 (2003) 4477e4480. [4] C. Bertucci, E. Domenici, Curr. Med. Chem. 9 (2002) 1463e1481. [5] S. Curry, H. Mandelkow, P. Brick, N. Franks, Nat. Struct. Biol. 5 (1998) 827e835. [6] K. Yamasaki, T. Maruyama, U. Kragh-Hansen, M. Otagiri, Biochim. Biophys. Acta. 1295 (1996) 147e157. [7] F. Bree, L. Thiault, G. Gautiers, M.S. Benedetti, E. Baltes, J.P. Rihoux, J. P. Tillement, Fundam. Clin. Pharm. 16 (2002) 471e478. [8] J. Wilting, W.F. van der Giesen, L.H.M. Janssen, M.M. Weideman, M. Otagiri, J. H. Perrin, J. Biol. Chem. 255 (1980) 3032e3037. [9] S. Wanwimolruk, D.J. Birkett, Biochim. Biophys. Acta 709 (1982) 247e255. [10] B.Z. Zhao, L.M. Song, X. Liu, J. Xie, J.Q. Zhao, Bioorg. Med. Chem. 14 (2006) 2428e2432. [11] C. Heinsohn, P. Polgan, J. Fishman, L. Taylor, Biochem. Biophys. 257 (1987) 251e258. [12] R. Melsert, J.W. Hoogerbrugge, F.F.G. Rommerts, Mol. Cell. Endocrinol. 64 (1989) 35e44. [13] M.J. Fischer, A.J.M. van Oosterhout, L.M.H. Janssen, F.P. Nijkamp, Biochem. Pharmacol. 44 (1992) 351e358. [14] T. Horie, T. Mizuma, S. Kasai, S. Awazu, Am. J. Physiol. 254 (1988) G465eG470. [15] M.J.E. Fischer, O.J.M. Bos, R.F. van der Linden, J. Wilting, L.H.M. Janssen, Biochem. Pharmacol. 45 (1993) 2411e2416. [16] B.J.M. Harmsen, S.H. De Bruin, L.H.M. Janssen, J.F.R. de Miranda, G.A.J. van Os, Biochemistry 10 (1971) 3217e3221. [17] R. Mallik, D.S. Hage, J. Pharm. Biomed. Anal. 46 (2008) 820e830. [18] M. Guo, W.J. Lü, M.H. Li, W. Wang, Eur. J. Med. Chem. 43 (2008) 2140e2148. [19] Y.Q. Wang, H.M. Zhang, Q.H. Zhou, Eur. J. Med. Chem. 44 (2009) 2100e2105. [20] J. Kawaoka, F.A. Gomez, J. Chromatogr. B 715 (1998) 203e210. [21] F. Zhang, Y.X. Du, B.F. Ye, P. Li, J. Photochem. Photobio. B: Biol. 86 (2007) 246e251. [22] A. Taga, Y.X. Du, S. Suzuki, S. Honda, J. Pharm. Biomed. Anal. 30 (2003) 1587e1593. [23] X.P. Liu, Y.X. Du, W. Sun, J.P. Kou, B.Y. Yu, Spectrochim. Acta Part A 74 (2009) 1189e1196. [24] W. Sun, Y.X. Du, J.Q. Chen, J.P. Kou, B.Y. Yu, J. Lumin. 129 (2009) 778e783. [25] G.Z. Chen, X.Z. Huang, J.G. Xu, Z.Z. Zheng, Z.B. Wang, Principles of Fluorescent Spectroscopy, second ed. Science Press, Beijing, 1990. [26] X.P. Liu, Y.X. Du, X.L. Wu, Spectrochim. Acta Part A 71 (2008) 915e920. [27] M. Kurono, A. Fujii, M. Murata, B. Fujitani, T. Negoro, Biochem. Pharmacol. 71 (2006) 338e353. [28] E. Abuin, C. Calderón, E. Lissi, J. Photochem. Photobiol. A: Chem. 195 (2008) 295e300. [29] Z. Chen, S.G. Weber, Trends Anal. Chem. 27 (2008) 738e748. [30] H. Li, F. Qu, J. Xu, Y. Deng, Chin. J. Chromatogr. 26 (2008) 473e477. [31] A.Y. Shmykov, V.N. Filippov, L.S. Foteeva, B.K. Keppler, A.R. Timerbaev, Anal. Biochem. 379 (2008) 216e218. [32] G.B. Ma, F. Gao, B.Z. Ren, P. Yang, Acta Chim. Sinica 53 (1995) 1193e1197. [33] J.N. Miller, Proc. Anal. Div. Chem. Soc. 16 (1979) 203e208. [34] F. Cui, L. Qi, G. Zhang, X. Liu, X. Yao, B. Lei, Bioorg. Med. Chem. 16 (2008) 7615e7621. [35] D.C. Carter, J.X. He, Adv. Protein Chem. 45 (1994) 153e203. [36] K. Yamasaki, T. Maruyama, K. Yoshimoto, Y. Tsutsumi, Biochim. Biophys. Acta 1432 (1999) 313e323. [37] W.J. Leonard Jr., K.K. Vijai, J.F. Foster, J. Biol. Chem. 283 (1963) 1984e1988. [38] K. Adachi, H. Watarai, Anal. Chem. 78 (2006) 6840e6846. [39] O.J. Bos, J.F. Labro, M.J. Fischer, J. Wilting, L.H. Janssen, J. Biol. Chem. 264 (1989) 953e959. [40] V. Cápka, S.J. Carter, S. Viccarone, N.H. Nguyen, S.S. Merkle, ASMS Conference, Indianapolis, IN, June 2007.
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Study on the binding of chiral drug duloxetine hydrochloride to human serum albuminq Xiangping Liu a, Yingxiang Du a, b, c, * a
Department of Analytical Chemistry, China Pharmaceutical University, Nanjing 210009, China Key Laboratory of Drug Quality Control and Pharmacovigilance (Ministry of Education), China Pharmaceutical University, Nanjing 210009, China c Key Laboratory of Modern Chinese Medicines (Ministry of Education), China Pharmaceutical University, Nanjing 210009, China b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 June 2009 Received in revised form 24 May 2010 Accepted 27 May 2010 Available online 4 June 2010
Duloxetine holds a special promise as an antidepressant, and its effect depends on its binding to human serum albumin (HSA). For this reason, the binding mechanism of duloxetine with HSA was investigated. The specific binding site in HSA was identified and binding constants were determined. Duloxetine could compete with dansyl-L-proline (DLP), a site II marker for binding to site II. Binding constants between duloxetine and HSA were 1.75 103 L mol1 and 3.74 103 L mol1 at pH 7.4 and pH 8.5, respectively. The interaction process of enantiomers and HSA was susceptible to pH change. It was concluded that specific binding position of duloxetine was located in site II, and the B conformation of HSA possibly excelled the N conformation in identifying and binding to enantiomers. Ó 2010 Elsevier Masson SAS. All rights reserved.
Keywords: Duloxetine hydrochloride HSA Fluorescence technique Affinity capillary electrophoresis
1. Introduction Duloxetine hydrochloride (Fig. 1), S-(þ)-N-methyl-3-(1-napthalenyloxy)-3-(2-thienyl) propanamine hydrochloride, holds a special promise as an antidepressant drug as it is a dual inhibitor of serotonin and norepinephrine reuptake [1e3]. It is also used for stress urinary incontinence. Various features, like improved efficacy, tolerability, safety, faster recovery, fewer side effects, low affinity for neuronal receptors and dual inhibiting nature give duloxetine an edge over other existing antidepressants, such as fluoxetine and tomoxetine. Whereas its enantiomer, R-()-Nmethyl-3-(1-napthalenyloxy)-3-(2-thienyl) propanamine hydrochloride, is pharmacologically invalid and not used clinically. Most organic compounds in the blood are carried while bind to albumin. As the most abundant protein in the blood, HSA plays an important role in the transport of many endogenous substances and exogenous compounds (e.g. drugs). The distribution, metabolism and free concentration of a drug in the blood are closely related to its interaction with HSA [4]. HSA has 585 amino acid
q Supported by Program for New Century Excellent Talents in University (No.: NCET-06-0498) and the Key Project of Chinese Ministry of Education (No.: 109085). * Corresponding author. Department of Analytical Chemistry, China Pharmaceutical University, No. 24 Tongjiaxiang, Nanjing, Jiangsu 210009, China. Tel./fax: þ86 25 83221790. E-mail address: [email protected] (Y. Du). 0223-5234/$ e see front matter Ó 2010 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.ejmech.2010.05.063
residues, and consists of three homologous domains (labeled IeIII), each of which is divided into two subdomains [5]. Among these subdomains site I (in subdomain IIA) and site II (in subdomain IIIA) have physiological importance in binding to a large number of drugs [6]. Drugs of the same binding sites in HSA will compete with one another for binding to the protein, which affects their distribution and metabolism in the blood [7]. Therefore, it is important to identify the binding site of a drug in HSA. The conformation of HSA also affects the interaction process between small molecule and HSA. In the normal physiological condition, HSA undergoes conformational change known as NeB (NeutraleBasic) transition, which will affect the binding characteristics between a drug and HSA [8,9]. Some drugs are inclined to interact with the N conformation rather than the B conformation whereas others are just the opposite [10]. Recent observations have clearly recognized the fact that albumin is not a biologically inactive protein, and that it may, in fact, play an important regulatory role in the functioning of certain cell types via interaction with the cellular membrane [11e13]. During such interactions, the albumin molecule undergoes NeB-like conformational transition and adopts the B conformation at membraneewater interface [14,15]. Harmsen et al. [16] have also suggested that under increased Ca2þ concentration the B conformation predominates in blood plasma. All these studies have demonstrated that the NeB transition of HSA has physiological significance. Meanwhile, HSA is a good chiral selector to many chiral compounds [17], which influences the transport of
4044
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
Fig. 1. The structure of duloxetine hydrochloride.
chiral drugs in the blood. Duloxetine, as a chiral drug, its interaction with HSA has not been reported until now. In this work, the binding process between duloxetine and HSA was investigated, and the effect of NeB transition of HSA was also taken into account. Many techniques have been applied to examine interactions between small molecule and macromolecule, including fluorescence spectroscopy, capillary electrophoresis [18e24]. In this study, the interaction between duloxetine enantiomer and HSA was investigated by synchronous fluorescence technique, for it could distinguish the overlapped fluorescence spectra by fixing an appropriate scanning interval (Dl ¼ lemission lexcitation) [25]. The common method based on fluorescence quenching cannot work here because, as has been described in our previous work [26], the fluorescence emission spectrum of duloxetine overlapped with that of HSA mostly. The competitive binding experiment is useful in identifying the specific binding site of small drug in protein, and it has been applied to confirm many drugs’ specific binding sites in protein [6,27,28]. In present work the binding site of duloxetine in HSA was identified by monitoring the fluorescence change of markereHSA complex in the presence of duloxetine. Affinity capillary electrophoresis (ACE) refers to the separation of substances that participate in specific or non-specific affinity interactions during CE. In the past two decades, ACE has been one of the most rapidly growing analytical techniques for studying intermolecular interactions and determining binding constants [29e31]. It has advantages in low sample consumption, fast separation and diverse operation modes. In this study, binding constants between duloxetine and HSA were determined by mobility-shift ACE method. The enantioselectivity of HSA to duloxetine enantiomers mixture was also examined by ACE under the near-physiological condition, which was a complement to the preceding results.
Fig. 2. Tryptophan synchronous fluorescence spectra (Dl ¼ 60 nm) in different systems. At pH 7.4 1: HSA and 2: HSA þ 6.4 105 mol L1 duloxetine; at pH 8.5 3: HSA and 4: HSA þ 6.4 105 mol L1 duloxetine.
2.1.1. The tryptophan-214 synchronous fluorescence When Dl is 60 nm, it was observed that the fluorescence intensity was enhanced and the maximum wavelength shifted from the long wavelength to the short wavelength (shown in Fig. 2) with
2. Results 2.1. Synchronous fluorescence experiment Among 585 amino acid residues of HSA, tryptophan-214 and tyrosine-411 can emit nature fluorescence upon the excitation [32]. When Dl is 60 nm, the spectrum characteristic of tryptophan-214 is observed; when Dl is 15 nm, the spectrum characteristic of tyrosine-411 is observed [33].
Fig. 3. The effect of duloxetine on tryptophan synchronous fluorescence spectra (Dl ¼ 60 nm), a: maximum wavelength; b: F0/F; pH 7.4 (solid); pH 8.5 (dashed line).
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
4045
Fig. 4. The effect of duloxetine on tyrosine synchronous fluorescence spectra (Dl ¼ 15 nm) in pH 7.4 buffer solution with different duloxetine concentration (105 mol L1): (1) 0, (2) 1.6, (3) 3.2, (4) 4.8, (5) 6.4, (6) 8.0.
the increase of the enantiomer concentration while keeping the concentration of HSA unchanged. Effects of added duloxetine on the spectra of HSA are shown in Fig. 3. As shown in Fig. 3(a), at pH 7.4, the added duloxetine induced the maximum wavelength shift from 341 to 339 nm, while at pH 8.5, it induced the maximum wavelength shift from 341 to 338 nm. In Fig. 3(b) the symbol F0 represented fluorescence intensity in the absence of the drug and the symbol F was fluorescence intensity in the presence of the drug (in the following, all the symbols F0 and F had the same meaning). The F0/F values were smaller than 1.0, which meant duloxetine enhanced the fluorescence of HSA. It was also observed that all F0/F values at pH 8.5 (dashed line) were smaller than those at pH 7.4 (solid), which showed that the enhancement of fluorescence intensity in basic solution was greater than that in neutral solution. The effect of R-isomer on the tryptophan synchronous fluorescence spectra was also investigated in the same way. In the neutral solution (pH 7.4), the change of fluorescence spectra was similar to that induced by duloxetine. However, in basic solution (pH 8.5), both the wavelength shift and fluorescence intensity enhancement were less than those induced by duloxetine. 2.1.2. The tyrosine-411 synchronous fluorescence The effect of duloxetine on tyrosine synchronous fluorescence spectra is shown in Fig. 4. With the unchanged concentration of
Fig. 5. The effect of duloxetine on the synchronous fluorescence intensity at 300 nm, pH 7.4 (solid); pH 8.5 (dashed line).
Fig. 6. Effects of enantiomers on the fluorescence intensity at 323 nm, pH 7.4 (solid); pH 8.5 (dashed line); duloxetine (B); R-isomer (-).
HSA, as the duloxetine concentration increased, the fluorescence intensity at 300 nm decreased gradually, while the fluorescence intensities at 323 nm and 336 nm rose accordingly. Fig. 5 shows the effect of duloxetine on the synchronous fluorescence intensity at 300 nm. All the F0/F values were greater than 1.0, which indicated the fluorescence at 300 nm decreased with the increased concentration of duloxetine. At the same time, all the F0/F values at pH 8.5 (dashed line) were greater than those at pH 7.4 (solid), which indicated that the fluorescence decrease at pH 8.5 was greater than that at pH 7.4. The effect of R-isomer was similar to that of duloxetine. Effects of duloxetine and its R-isomer on the fluorescence intensity at 323 nm were investigated, respectively. Results are shown in Fig. 6. When pH value of solution was changed from 7.4 to 8.5, the fluorescence intensities at 323 nm were enhanced remarkably. It was also observed that at pH 7.4, the enhancement for duloxetine was similar to that for R-isomer, but at pH 8.5, it was greater than that for R-isomer. The similar results were acquired for fluorescence intensities at 336 nm. 2.2. Competitive binding experiment 5-dimethylaminonaphthalene-1-sulfonamide (DNSA) and DLP were chosen as specific markers for site I and site II [27], respectively. It is known that these two markers can specifically bind to HSA in certain position located in site I or site II, and the formed complex can emit characteristic fluorescence from 360 nm to 550 nm upon the excitation. A drug is added into the complex system. If the added drug quenches the characteristic fluorescence emission, it is considered that the drug binds to HSA competitively and has the same specific binding position as the marker [6,28]. Effects of enantiomers on the fluorescence intensity of DNSAeHSA complex are shown in Fig. 7. All the F0/F values at pH 7.4 were less than 1.0 and decreased with the increase of isomer amount, which implied that both duloxetine and R-isomer enhanced the fluorescence of DNSAeHSA and enantiomers could not compete with DNSA for binding to site I. At pH 8.5, F0/F values decreased first and then increased as the isomer concentration augmented, especially for duloxetine. This demonstrated that under higher concentration enantiomers presented some ability of unspecific binding to site I in the weak basic solution. The competition of enantiomers with site I marker resulted in the fluorescence decrease of HSAeDNSA complex.
4046
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
Fig. 7. Effects of enantiomers on the fluorescence intensity of DNSAeHSA at 474 nm, pH 7.4 (solid); pH 8.5 (dashed line); duloxetine (B); R-isomer (-).
Effects of enantiomers on the fluorescence intensity of DLPeHSA complex are demonstrated in Fig. 8. All F0/F values were larger than 1.0, which signified that both duloxetine and its Risomer had quenched the fluorescence of DLPeHSA complex. At pH 7.4, two enantiomers had similar abilities of quenching the fluorescence, whereas at pH 8.5, all F0/F values for duloxetine were larger than those for R-isomer. This denoted that duloxetine quenched the fluorescence stronger than R-isomer. 2.3. Determination of the binding constants In Fig. 9, electropherograms of HSA are shown in buffers containing duloxetine of different concentration. The isoelectric point of HSA is 4.7, so it is negatively charged in the buffer at pH 7.4, and is detected after DMSO (neutral marker) at the positive voltage. Binding constants of a small drug and protein can be determined by the mobility-shift ACE, in which the concentration change of a drug in the buffer can cause the mobility change of the protein due to the binding process between a drug and a protein. This method has specific advantage over other approaches, for example, the injected sample need not be highly purified. However, the
Fig. 8. Effects of enantiomers on the fluorescence intensity of DLPeHSA at 480 nm, pH 7.4 (solid); pH 8.5 (dashed line). duloxetine (B); R-isomer (-).
Fig. 9. Electropherograms of HSA in pH 7.4 buffers containing duloxetine of different concentration: (1) 0.0; (2) 100; (3) 200; (4) 300; (5) 400; (6) 500; (7) 600 (106 mol L1). The applied voltage was þ15.0 kV, and the capillary temperature was controlled at 298 K; the detection wavelength was 214 nm and the injection volume was 50 mbar/10 s.
fluctuation of the electroosmotic flow (EOF) may adversely influence the observed electrophoretic mobility. The introduction of mobility ratio (M) can avoid the influence of the EOF change, and the M value is independent of the voltage, the capillary length and solvent viscosity [20], so the electrophoretic mobility (mi) was replaced by M to determine binding constants. According to the migration times of neutral marker (teo) and protein (tp), the M values are calculated from Equ. (2) in Section 5.3.3.1. Fig. 10 shows curves plotted by 1/DM versus 1/[L]. The intercept and slope of the plotted curve were obtained, and binding constants can be calculated accordingly (listed in Table 1). 2.4. Chiral identification ability of HSA to enantiomers Electropherograms of a mixture of duloxetine and R-isomer were recorded at pH 8.5 (Fig. 11a) and pH 7.4 (Fig. 11b), respectively. The run buffer contained 1.0 105 mol L1 HSA, 0.05 mol L1
Fig. 10. Curves plotted by 1/DM versus 1/[L] based on the mobility-shift ACE. pH 7.4 (solid); pH 8.5 (dashed line); duloxetine (B); R-isomer (-).
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049 Table 1 The binding constants determined by the mobility-shift ACE at 298 K.
DuloxetineeHSA (pH ¼ 7.4) DuloxetineeHSA (pH ¼ 8.5) R-isomereHSA (pH ¼ 7.4) R-isomereHSA (pH ¼ 8.5)
Linear regression equation
Related coefficient
Binding constant (L mol1)
y ¼ 0.0047x 8.2087
r ¼ 0.9926
1.75 103
y ¼ 0.0022x 8.2221
r ¼ 0.9956
3.74 103
y ¼ 0.0046x 8.4914
r ¼ 0.9945
1.85 103
y ¼ 0.0054x 14.47
r ¼ 0.9923
2.68 103
TriseHCl with 0.10 mol L1 NaCl, which was the same as that used in the fluorescence examination. It was shown that duloxetine and R-isomer had different migration times at pH 8.5 while they were not separated at pH 7.4. The resolution was 0.71 between duloxetine and R-isomer in Fig. 11(a). 3. Discussion Results of synchronous fluorescence examination showed that the introduction of duloxetine and R-isomer had affected HSA conformation. The drug had induced synchronous fluorescence spectra change, such as the maximum wavelength and the fluorescence intensity. When the maximum wavelength shifts to the longer wavelength, it denotes that the vicinal environment of residue becomes more hydrophilic; on the other hand, when the maximum wavelength shifts to the shorter wavelength, it denotes that the vicinal environment of residue becomes more hydrophobic [34]. For tryptophan residue, the added enantiomers could induce
4047
the maximum wavelength shift to the shorter range, which implied that the microenvironment of tryptophan residue became more hydrophobic. For tyrosine residue, both two enantiomers could induce fluorescence decrease at 300 nm, owing to the fact that enantiomers had bound to tyrosine residue and thus quenched its fluorescence. It was also noticed that fluorescence at 323 and 336 nm appeared and their intensities increased as the enantiomer amount rose. Fluorescence at 323 and 336 nm should be ascribed to the added isomer. The presence of HSA could enhance the fluorescence at 323 and 336 nm, especially in weak basic solution (shown in Fig. 6.), which signified that there were new complex formation between enantiomers and HSA. The competitive binding experiment demonstrated that both two enantiomers had quenched the fluorescence of DLPeHSA complex, which denoted that they had competed with DLP for binding to site II and inhibited the formation of DLPeHSA complex. The dissociation of DLP from HSA brought about the fluorescence quenching at 480 nm. As for the DNSAeHSA system, this phenomenon of fluorescence quenching had not been observed. These results demonstrated that the specific binding position of two enantiomers was located in site II instead of site I. It is known that tyrosine-411 is located in site II and tryptophan-214 is located in site I [5,6]. Experimental results of synchronous fluorescence showed that enantiomers could bind to tyrosine-411 and quenched the fluorescence gradually, which were consistent with results of competitive binding experiments. Carter and He [35] have shown that HSA is a multi-domain molecule, and each of domains is not independent, but interacts. It accounts for the synchronous fluorescence change of tryptophan-214 and the fluorescence
Fig. 11. Electropherograms of a mixture of 1.0 104 mol L1 duloxetine and 1.0 104 mol L1 R-isomer in pH 8.5 (a) and pH 7.4 (b) run buffer. The applied voltage was þ12.0 kV, and the capillary temperature was controlled at 293 K; the detection wavelength was 220 nm and the injection volume was 50 mbar/5 s.
4048
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
enhancement of DNSAeHSA complex. Although enantiomers did not bind to site I specifically, they could also impact the vicinity environment of tryptophan by the existence of allosteric regulation between sites I and II [36]. According to fluorescence experiment results, the interaction process between the isomer and HSA was susceptible to solution pH change. In synchronous fluorescence experiment, it was observed that all spectrum changes at pH 8.5, such as the wavelength shift of tryptophan residue, the decrease of tyrosine residue fluorescence, and the enhancement of fluorescence at 323 nm (dashed line shown in Figs. 2, 3, 5 and 6), were greater than those at pH 7.4 (solid in Figs. 2, 3, 5 and 6). Moreover, the abilities of changing synchronous fluorescence spectra of HSA at pH 7.4 for two enantiomers were similar, but when pH value became 8.5 changes induced by duloxetine were greater than those induced by R-isomer. In competitive binding experiment, the regulation of pH to the binding process still existed. Duloxetine displayed some ability of unspecific binding to site I at pH 8.5, while there was no affinity at pH 7.4. As for site II, it was obvious that enantiomers had stronger ability of replacing DLP at pH 8.5 than at pH 7.4. At pH 7.4, enantiomers competed with DLP to similar extent, while at pH 8.5 they behaved differently. The fluorescence quenching induced by duloxetine was apparently greater than that induced by R-isomer. Binding constants provide a fundamental measure of the affinity of a drug to a protein. In this study binding constants between enantiomers and HSA were also different from each other at different pH. As illustrated in Table 1, binding constants obtained at pH 7.4 were smaller than those obtained at pH 8.5, which implied that interactions between isomer and HSA at pH 8.5 were stronger than those at pH 7.4. Furthermore, it was also found that binding constants (1.75 103 L mol1 for duloxetine and 1.85 103 L mol1 for R-isomer) were similar at pH 7.4, but at pH 8.5 the binding constant (3.74 103 L mol1) for duloxetine was apparently bigger than that (2.68 103 L mol1) for R-isomer. Results shown in Fig. 11 prove that the interaction was susceptible to solution pH change further. When HSA was added into the run buffer of different pH value and the mixture of two isomers was injected as sample, the partial enantioseparation of duloxetine and its R-isomer was obtained (Rs ¼ 0.71) at pH 8.5, while at pH 7.4, the separation was not observed. The longer migration time of duloxetine showed that it interacted with HSA stronger than R-isomer did. The susceptibleness to solution pH change can be ascribed to HSA conformation transition. It has been reported that we can obtain different conformation of HSA by regulating the solution pH value in vitro [37,38]. HSA exhibited N conformation in neutral solution and B conformation in weak basic solution. It was deduced that binding process changes due to pH value were possibly caused by HSA NeB conformation transition. According to results of fluorescence experiment and affinity capillary electrophoresis, two enantiomers interacted with the B conformation stronger and HSA in the B conformation had better enantioselectivity ability than the N conformation did. The 1H NMR spectra have proved that HSA is heart-shaped in N conformation, but elliptic in B conformation, and the elliptic structure is much looser than the hearted one [39]. According to our investigation, it can be deduced that the looser structure of HSA favored the interaction between duloxetine and HSA. Additionally, the pKa value of duloxetine is 10.0 [40], and pH increase could induce isomers ionization change, so the effect of isomers ionization change on the interaction process should not be ignored. 4. Conclusion Together with all the observations this study has shown that the specific binding site of duloxetine was located in site II of HSA, and
binding constants between duloxetine and HSA were 1.75 103 L mol1 at pH 7.4 and 3.74 103 L mol1 at pH 8.5, respectively. It was also demonstrated that interactions between enantiomers and HSA were susceptible to solution pH change, and the NeB transition of HSA can possibly affect its binding ability and enantioselectivity to duloxetine isomer. This study helps us understand the interaction between duloxetine and HSA in the body. Drugs of the same binding site in HSA will compete with one another for binding to the protein, which affects their distribution and metabolism in the blood. The identification of specific binding position in HSA is useful to examine the effect of other drugs on duloxetine when they are administrated together. Moreover, the fact that the binding of duloxetine to HSA was susceptible to the NeB transition showed some factors that could induce this transition, such as pH value, Ca2þ concentration and so on, should be controlled carefully when we investigated the interaction mechanism between duloxetine and HSA.
5. Experimental 5.1. Reagents Duloxetine hydrochloride (S-(þ)-N-methyl-3-(1-napthalenyloxy)-3-(2-thienyl) propanamine hydrochloride) (>99.5%) and its R-isomer (>99.5%) were supplied by Jiangsu Institute for Drug Control (Nanjing, China). Fatty acid-free (<0.005%) and globulinfree HSA, 5-dimethylaminonaphthalene-1-sulfonamide (DNSA) and dansyl-L-proline (DLP) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). The stock solutions of drugs were prepared by dissolving 0.0334 g of standard drug and diluting to 50 mL with double distilled water, the concentration was 2.0 103 mol L1. 1.0 105 mol L1 HSA solution was prepared based on its molecular weights of 68,000 in 0.05 mol L1 TriseHCl buffer solution (with 0.10 mol L1 NaCl). All of the prepared solutions were stored at 0e4 C. All other materials were of analytical reagent grade and double distilled water was used throughout. 5.2. Apparatus A Shimadzu RF-5301 spectrofluorimeter (Shimadzu, Japan) was used for the fluorescence spectra measurements equipped with a 1 cm path length quartz cell. In the synchronous fluorescence experiment both the excitation and the emission slits were all 3 nm. In the competitive binding experiment the excitation slit was 5 nm, and the emission slit was 5 nm for DNSAeHSA and 3 nm for DLPeHSA, respectively. Agilent capillary electrophoresis instrument equipped with diode array detector (Agilent Technology, USA) and a 35 cm 50 mm elasticity quartz capillary column (Yongnian Hebei province, China) was used in affinity capillary electrophoresis experiment. The value of pH is measured on a pH meter (Lei-ci pHS-25, Shanghai, China).
5.3. Procedures 5.3.1. Synchronous fluorescence examination 220 nm was set as the initial excitation wavelength. When Dl was fixed at 60 nm, synchronous fluorescence spectra of 1.0 105 mol L1 HSA were recorded with increase of drug amount; when Dl was fixed at 15 nm, the experiment was performed in the same way.
X. Liu, Y. Du / European Journal of Medicinal Chemistry 45 (2010) 4043e4049
5.3.2. Competitive binding experiment Competitive binding experiments were performed by using fluorescent markers specific for site I and site II (site I: DNSA; site II: DLP), respectively. The marker/HSA concentration ratio was kept 1/5 in order to avoid nonspecific binding. Accordingly, concentrations of HSA and marker were 1.0 105 mol L1 and 2.0 106 mol L1, respectively. The binding site of duloxetine in HSA was investigated by monitoring fluorescence change of the markereHSA complex with the increase of the drug amount upon the excitation at 343 nm. 5.3.3. Affinity capillary electrophoresis (ACE) experiment 5.3.3.1. Determination of binding constants. The electrophoresis buffer consisted of 0.02 mol L1 TriseHCl (with 0.02 mol L1 NaCl) and drugs of different concentration (0; 100; 200; 300; 400; 500; 600 (106 mol L1)). The injection sample was 1.0 105 mol L1 HSA containing 0.1% dimethyl sulfoxide (DMSO). DMSO was used as neutral marker to correct the fluctuation of electroosmotic flow. Binding constants between small drug molecule and HSA were determined by the mobility-shift ACE. Assuming that the stoichiometry of the binding between the target and the ligand is 1:1, we can use a simple mode to determine the binding constant (shown in equ. (1)) [29].
1
mi mf
¼
1 1 1 $ þ mc mf Kb mc mf ½L
(1)
where mi is the electrophoretic mobility of the protein with different ligand concentration [L] in the buffer; mf is the electrophoretic mobility of the free protein in the absence of ligand; mc is the electrophoretic mobility of proteinedrug complex with the saturation of ligand; Kb is the binding constant between drug and protein. In order to correct the fluctuation of electroosmotic flow, the mobility ratio (M) was used to estimate binding constants.
mapp meof mi teo M ¼ ¼ ¼ 1 meof meof tp
(2)
where mapp and meof are apparent electrophoretic mobility and electroosmotic mobility, respectively; teo and tp are the migration time of neutral marker and protein with different ligand concentration, respectively. Equ. (1) can be converted to the equ. (3).
1
DM
¼
1 1 1 $ þ Kb $DMmax ½L DMmax
(3)
where DM is the change in mobility ratio as a function of [L], and DMmax is the maximum change in mobility ratio that can be achieved with the saturation of ligand for complexation. By plotting 1/DM versus 1/[L], the y intercept will give the value of 1/DMmax, and Kb ¼ intercept/slope. 5.3.3.2. Chiral identification ability of HSA to the enantiomers. The run buffer contained 1.0 105 mol L1 HSA, 0.05 mol L1 TriseHCl (with 0.10 mol L1 NaCl). A mixture of 1.0 104 mol L1 duloxetine and 1.0 104 mol L1 R-isomer was injected as sample. The
4049
pH value of run buffer was set as 7.4 and 8.5, respectively, and the electropherograms were recorded. All the above procedures were repeated at least 3 times. References [1] P. Mariappan, A. Alhasso, Z. Ballantyne, A. Grant, J. N’Dow, Eur. Urol. 51 (2007) 67e74. [2] D.E. Stewart, M.M. Wohlreich, C.H. Mallinckrodt, J.G. Watkin, S.G. Kornstein, J. Affect. Disord. 94 (2006) 183e189. [3] F.P. Bymaster, E.E. Beedle, J. Findlay, P.T. Gallagher, J.H. Krushinski, S. Mitchell, D.W. Robertson, D.C. Thompson, L. Wallace, D.T. Wong, Bioorg. Med. Chem. Lett. 13 (2003) 4477e4480. [4] C. Bertucci, E. Domenici, Curr. Med. Chem. 9 (2002) 1463e1481. [5] S. Curry, H. Mandelkow, P. Brick, N. Franks, Nat. Struct. Biol. 5 (1998) 827e835. [6] K. Yamasaki, T. Maruyama, U. Kragh-Hansen, M. Otagiri, Biochim. Biophys. Acta. 1295 (1996) 147e157. [7] F. Bree, L. Thiault, G. Gautiers, M.S. Benedetti, E. Baltes, J.P. Rihoux, J. P. Tillement, Fundam. Clin. Pharm. 16 (2002) 471e478. [8] J. Wilting, W.F. van der Giesen, L.H.M. Janssen, M.M. Weideman, M. Otagiri, J. H. Perrin, J. Biol. Chem. 255 (1980) 3032e3037. [9] S. Wanwimolruk, D.J. Birkett, Biochim. Biophys. Acta 709 (1982) 247e255. [10] B.Z. Zhao, L.M. Song, X. Liu, J. Xie, J.Q. Zhao, Bioorg. Med. Chem. 14 (2006) 2428e2432. [11] C. Heinsohn, P. Polgan, J. Fishman, L. Taylor, Biochem. Biophys. 257 (1987) 251e258. [12] R. Melsert, J.W. Hoogerbrugge, F.F.G. Rommerts, Mol. Cell. Endocrinol. 64 (1989) 35e44. [13] M.J. Fischer, A.J.M. van Oosterhout, L.M.H. Janssen, F.P. Nijkamp, Biochem. Pharmacol. 44 (1992) 351e358. [14] T. Horie, T. Mizuma, S. Kasai, S. Awazu, Am. J. Physiol. 254 (1988) G465eG470. [15] M.J.E. Fischer, O.J.M. Bos, R.F. van der Linden, J. Wilting, L.H.M. Janssen, Biochem. Pharmacol. 45 (1993) 2411e2416. [16] B.J.M. Harmsen, S.H. De Bruin, L.H.M. Janssen, J.F.R. de Miranda, G.A.J. van Os, Biochemistry 10 (1971) 3217e3221. [17] R. Mallik, D.S. Hage, J. Pharm. Biomed. Anal. 46 (2008) 820e830. [18] M. Guo, W.J. Lü, M.H. Li, W. Wang, Eur. J. Med. Chem. 43 (2008) 2140e2148. [19] Y.Q. Wang, H.M. Zhang, Q.H. Zhou, Eur. J. Med. Chem. 44 (2009) 2100e2105. [20] J. Kawaoka, F.A. Gomez, J. Chromatogr. B 715 (1998) 203e210. [21] F. Zhang, Y.X. Du, B.F. Ye, P. Li, J. Photochem. Photobio. B: Biol. 86 (2007) 246e251. [22] A. Taga, Y.X. Du, S. Suzuki, S. Honda, J. Pharm. Biomed. Anal. 30 (2003) 1587e1593. [23] X.P. Liu, Y.X. Du, W. Sun, J.P. Kou, B.Y. Yu, Spectrochim. Acta Part A 74 (2009) 1189e1196. [24] W. Sun, Y.X. Du, J.Q. Chen, J.P. Kou, B.Y. Yu, J. Lumin. 129 (2009) 778e783. [25] G.Z. Chen, X.Z. Huang, J.G. Xu, Z.Z. Zheng, Z.B. Wang, Principles of Fluorescent Spectroscopy, second ed. Science Press, Beijing, 1990. [26] X.P. Liu, Y.X. Du, X.L. Wu, Spectrochim. Acta Part A 71 (2008) 915e920. [27] M. Kurono, A. Fujii, M. Murata, B. Fujitani, T. Negoro, Biochem. Pharmacol. 71 (2006) 338e353. [28] E. Abuin, C. Calderón, E. Lissi, J. Photochem. Photobiol. A: Chem. 195 (2008) 295e300. [29] Z. Chen, S.G. Weber, Trends Anal. Chem. 27 (2008) 738e748. [30] H. Li, F. Qu, J. Xu, Y. Deng, Chin. J. Chromatogr. 26 (2008) 473e477. [31] A.Y. Shmykov, V.N. Filippov, L.S. Foteeva, B.K. Keppler, A.R. Timerbaev, Anal. Biochem. 379 (2008) 216e218. [32] G.B. Ma, F. Gao, B.Z. Ren, P. Yang, Acta Chim. Sinica 53 (1995) 1193e1197. [33] J.N. Miller, Proc. Anal. Div. Chem. Soc. 16 (1979) 203e208. [34] F. Cui, L. Qi, G. Zhang, X. Liu, X. Yao, B. Lei, Bioorg. Med. Chem. 16 (2008) 7615e7621. [35] D.C. Carter, J.X. He, Adv. Protein Chem. 45 (1994) 153e203. [36] K. Yamasaki, T. Maruyama, K. Yoshimoto, Y. Tsutsumi, Biochim. Biophys. Acta 1432 (1999) 313e323. [37] W.J. Leonard Jr., K.K. Vijai, J.F. Foster, J. Biol. Chem. 283 (1963) 1984e1988. [38] K. Adachi, H. Watarai, Anal. Chem. 78 (2006) 6840e6846. [39] O.J. Bos, J.F. Labro, M.J. Fischer, J. Wilting, L.H. Janssen, J. Biol. Chem. 264 (1989) 953e959. [40] V. Cápka, S.J. Carter, S. Viccarone, N.H. Nguyen, S.S. Merkle, ASMS Conference, Indianapolis, IN, June 2007.