J Mol Model (2014) 20:2098 DOI 10.1007/s00894-014-2098-7
ORIGINAL PAPER
Molecular interaction of PCB180 to human serum albumin: insights from spectroscopic and molecular modelling studies Senbiao Fang & Huanhuan Li & Tao Liu & Hongxia Xuan & Xin Li & Chunyan Zhao
Received: 16 September 2013 / Accepted: 27 November 2013 / Published online: 16 March 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Polychlorinated biphenyls (PCBs) are potentially hazardous to the environment because of their chemical stability and biological toxicity. In this study, we identified the binding mode of a representative PCB180 to human serum albumin (HSA) using fluorescence and molecular dynamics (MD) simulation methods. PCB180 bound exactly at subdomain IIIA of HSA based on the fluorescence study along with site marker displacement experiments. PCB180 also induced conformational changes that were governed mainly by hydrophobic forces. MD studies and free energy calculations also made important contributions to the understanding of the effects of an HSA-PCB180 system on conformational changes. The simulations on binding behavior proved that PCB180 was located only in subdomain IIIA. Hydrophobic interactions dominated the mode of binding behavior. The results obtained using the two methods correlated well with each other. Our findings provide a framework for elucidating the mechanisms of PCB180-HSA binding, and may also help in further research on the transportation, distribution, and toxicity effects of PCBs when introduced into human blood serum.
Keywords Polychlorinated biphenyls . Human serum albumin . Molecular dynamics . Fluorescence study Senbiao Fang and Huanhuan Li contributed equally to this work. Electronic supplementary material The online version of this article (doi:10.1007/s00894-014-2098-7) contains supplementary material, which is available to authorized users. S. Fang : H. Li : T. Liu : H. Xuan : C. Zhao (*) School of Pharmacy, LanzhouUniversity, Lanzhou 730000, China e-mail: [email protected] X. Li College of Food and Bioengineering, Henan University of Science and Technology, Luoyang 471003, China
Introduction Polychlorinated biphenyls (PCBs) are ubiquitous and persistent environmental pollutants that accumulate generally in the food chain [1]. PCBs with six or more chlorines can be found in multiple environmental matrices, such as air, water, sediments, adipose tissue of fish, wildlife, and in humans, particularly in breast tissue, serum, and milk [2]. The toxicological effects of exposure to PCBs include hepatotoxicity, immunotoxicity, reproductive problems, as well as respiratory, mutagenic and carcinogenic effects [3, 4]. As generally considered, PCBs theoretically encompass a group of 209 different congeners with various degrees of chlorination and substitution patterns at available sites, which strongly affect each PCB’s potency and the nature of its toxicity [6]. For now, most PCBs are currently poorly characterized, and toxicity mechanisms are not well defined [5–7]. When PCBs enter the body, their distribution and metabolism are correlated with their binding affinities towards serum albumin in blood. Serum albumin—the most abundant protein in plasma—has important functions in the maintenance of colloid osmotic blood pressure, and in the binding and transport of endogenous and exogenous ligands [8]. Thus, knowledge of the interaction mechanisms between PCBs and plasma proteins is an important factor in understanding the pharmacokinetics and structural features of these ligands, as these mechanisms strongly influence ligand distribution and so determine the biological (toxicity) effect. Human serum albumin (HSA) is a non-glycosylated singlechained polypeptide with a mass of 67 kDa, which organizes to form a heart-shaped protein with an α-helical content of approximately 67 % [9–11]. HSA consists of different numbers and sequences of amino acids, which enables it to adopt compounds in various three-dimensional structures, and possesses unique biological functions. HSA is monomeric but contains three structurally similar α-helical domains: I
2098, Page 2 of 10
(residues 1–195), II (196–383), and III (384–585), each consisting of two subdomains (A and B). The overall structure is stabilized by 17 disulfide bridges [10, 12, 13]. Aromatic and heterocyclic ligands generally bind within two hydrophobic pockets in subdomains IIA and IIIA, which are designated as sites I and II, respectively [12, 13]. The overall structure of HSA is flexible under natural conditions, while regulatory functions, such as changes in ligand-binding site conformation, are possible during the process of ligand binding [14]. The flexibility of HSA could be related directly to its myriad functions and binding properties, which further induce various modes of absorption, distribution, metabolism, excretion, and toxicity for many environmental chemicals such as PCBs [15]. Historically, screening for the binding interactions of HSA were generally thought to have been the first steps in investigating the biological behavior of chemicals. Insights into the detailed binding modes of PCBs to HSA have become fundamentally important from a pharmacological perspective. Considering that HSA is one of the main carriers for pharmaceuticals in the human organism, several experimental studies have been conducted to clarify HSA binding properties [10, 16, 17]. Various spectroscopic methods, such as fluorescence quenching, ultraviolet absorption spectroscopy, and circular dichroism were applied to determine the specific fluorescent properties of HSA; these methods are considered to be sensitive and relatively easy to use with high level of molecular detail [18–21]. Experimental studies have generally focused on physiological conditions, whereas theoretical studies such as docking and molecular dynamic (MD) simulation have long been applied to analyze biological processes rapidly and efficiently [22–24]. These computational methods can provide detailed information on the fluctuations and conformational changes of proteins and ligands. These methods are now used routinely to investigate the structural changes, dynamics and thermodynamics of biological molecules and their complexes, which can aid further understanding of the interaction mechanisms of protein–ligand systems. Particularly for PCBs, 209 PCB congeners are not wholly purified or synthesized, which makes analysis of the interaction between individual congeners and proteins difficult. Theoretical calculation and modeling of PCB–HSA binding behavior could help explain the toxicity of PCBs by facilitating annotation of the physical basis of the structure–function relationship of biomolecules. This study employed a combination of experimental and theoretical research in an attempt to determine where and how PCBs bind to HSA. Considering that non-coplanar PCBs are highly prevalent in the environment and must be included in hazard/risk assessment [25], PCB180 (as shown in Fig. S1, Supplementary materials section) was considered as the representative PCB congener. The site marker displacement experiments based on fluorescence spectroscopy were first undertaken to validate site-selective binding and to determine the
J Mol Model (2014) 20:2098
thermodynamic properties of the HSA-PCB180 system. Based on the experimental results, theoretical calculations were mapped using molecular docking and dynamics simulations to gain insight into the full structural characteristics of the PCB180-HSA complex. The results obtained by the two methods correlated well with each other. We also discussed the structural diversity of PCB180 and another non-coplanar, PCB153, which resulted in different binding modes with HSA. Our findings thus provide a framework for elucidating the mechanisms of PCB180–HSA binding. Future studies on other PCBs–HSA interactions may gain insights from our study.
Experimental methods Materials HSA (fatty acid free), phenylbutazone (PHE) and ibuprofen (IB) were purchased from Sigma (St. Louis, MO). HSA was dissolved in Tris-HCl buffer (0.1 M), and the pH was adjusted to 7.4. PCB180 was dissolved in dimethyl sulfoxide (DMSO). Other chemicals were of analytical reagent grade. Phenylbutazone and ibuprofen were initially dissolved in 0.05 M NaOH (pH 7.40) as stock solutions. The 60 μM HSA stock solutions were stored in the dark at 189 K, while phenylbutazone (500 μM), ibuprofen (2,000 μM) and PCB180 (500 μM) were stored at 277 K. Double distilled water was used for all procedures. Fluorescence titration assays Fluorescence was measured on a RF-5301 (Shimadzu, Tokyo, Japan) fluorescence spectrophotometer at λex of 280 nm with slit widths (em=3 nm and ex=5 nm). The HSA solution was fixed at 1×10−6 M concentration, then titrated by increasing the concentration of PCB180 to give a final concentration of 4×10−7 M–6.8×10−6 M. The fluorescence intensity was recorded under excitation at 280 nm and emission at 354 nm with a time interval of 10 min. In addition, the fluorescence experiments were carried out at three different temperatures of 298 K, 303 K and 308 K to evaluate the effect of temperature. For displacement experiments of PHE and IB, PCB180 (1.0 μM) was incubated with HSA (1.0 μM) in Tris-HCl buffer (0.1 M, pH 7.4) at 298 K for 30 min before displacement. Then, 2.5 mL sample was placed in a 1-cm quartz cuvette, followed by titration of aliquots of PHE (500 μM) or IB (2,000 μM) [26]. Molecular modeling methods A molecular docking study was implemented to explore possible binding positions of PCB180. In the present study, the
J Mol Model (2014) 20:2098
structure with PDB ID: 1H9Z [27] having a resolution of 2.5 Å was chosen for dockings. Water molecules, co-crystallized ligand and ions were removed (including ordered water molecules) before the protein was refined with Swiss PDB viewer. Missing hydrogens and Kollman partial charges were then added. Finally, non-polar hydrogens were merged to their corresponding carbons, and desolvation parameters were assigned to each atom. PCB180 was built with G-view and optimized at the B3LYP/6-31++G** level using Gaussian 98 software (http://www.gaussian.com/). Then rotations and torsions for PCB180 were set automatically in ADT (Autodock Tools, http://autodock.scripps.edu/). Gln390, Lys414, Glu425 and His464 were set as flexible residues. In the present work, the selection of flexible residues was based on the following criteria: (1) all binding site residues whose side chain atoms deviate more than 2.5 Å from the nearest atom in the corresponding amino acid in another crystal structure of the same receptor. (2) All binding site residues with multiple occupancies or missing density. (3) If the unit cell contains multiple independently refined copies of the receptor, all binding site residues whose side chain atoms deviate more than 1.5 Å from the nearest atom in the corresponding amino acid in another copy of the receptor in the unit were selected [28, 29]. Docking simulations were performed with AutoDock4.0 [30] using a Lamarkian genetic algorithm. The large enough grid volumes were considered, including the whole subdomain IIA and IIIA. This choice was determined by the need to maintain a small grid spacing (0.375) for the search box and at the same time to perform the docking calculation without neglecting any portion of the protein surface. Others docking parameters were set as default.
Page 3 of 10, 2098
estimate the long-range electrostatic interactions with a cutoff of 10.0 Å. All bond lengths were constrained using the SHAKE algorithm, and the integration time step was set to 2 f. using the Verlet leapfrog algorithm [32]. To eliminate possible bumps between the solute and the solvent, the entire system was minimized in two steps. Firstly, the complex was restrained with a harmonic potential with a force constant k= 100 kcal mol−1 Å−2. The water molecules and counterions were optimized using the steepest descent method of 2,500 steps, followed by the conjugate gradient method for 2,500 steps. Secondly, the entire system was optimized using the first step method without any constraint. These two minimization steps were followed by annealing simulation with a weak restraint (k=10 kcal mol−1 Å−2) for the complex and the entire system was heated gradually in the NVT ensemble from 0 to 298 K over 50 ps. After the heating phase, a 5 ns MD simulation was performed under 1 atm. The constant temperature was selected at 298 K with the NPT ensemble. Constant temperature was maintained using the Langevin thermostat with a collision frequency of 2 ps−1. Constant pressure was maintained employing isotropic position scaling algorithm with a relaxation time of 2 ps. In the present research, 20 snapshots (every 50 ps) were extracted from the last 1 ns trajectory for each complex. The binding free energy (ΔGbind) was computed for each snapshot and averaged using the MM-GBSA approach implemented as script (MMPBSA.py) in AMBER software [33].
Results and discussion Fluorescence titration assays
Molecular dynamics simulations To check the stability of the PCB180–HSA complex, the best docking result was employed for MD simulation. MD was carried out with AMBER software (version 11) [31], using AMBER ff02 force field for protein and the general AMBER force field (GAFF) for PCB180 (http://ambermd.org/). The optimized geometries of PCB180 were employed to calculate electrostatic potential (ESP)-derived charges using the restrained electrostatic potential (RESP) methodology, as implemented in the ANTECHAMBER model (http://ambermd. org/antechamber/ac.html). Hydrogen atoms were added to the initial PCB180-HSA complex model using the tleap module, setting ionizable residues as their default protonation states at a neutral pH value. The complex was solvated in a cubic periodic box of explicit TIP3P water model, which extended a minimum 10 Å distance from the box surface to any atom of the solute. Sixteen sodium ions replaced water molecules randomly in the box to neutralize a total negative charge of 16e on the entire system. The particle mesh Ewald (PME) method for simulation of periodic boundaries was used to
Fluorescence measurement is a simple method with which to explore the structural changes of HSA influenced by PCB180. First, the fluorescence emission intensity of HSA was tested by titrating various amounts of PCB180 using an excitation wavelength of 280 nm. As shown in Fig. 1a, the fluorescence intensity of HSA decreased gradually, exhibiting a concentration-dependent relationship. Quantitative analysis was then conducted using the Stern-Volmer equation [34] as listed in Table 1. The plots obtained from fluorescence measurements showed a linear relationship with R2 = 0.99 (Fig. 1b), which suggested a single class of binding site on HSA for PCB180 with a unique binding constant [35]. Identification of the binding sites The results suggested the existence of binding behavior between PCBs and HSA induced by quenchers of PCB180. However, the exact binding site of PCB180 was unknown under such conditions, and more evidence was required for identification. Thus, we conducted site marker displacement
2098, Page 4 of 10
J Mol Model (2014) 20:2098
Fig. 1 a Fluorescence spectra of human serum albumin (HSA) with various amounts of the polychlorinated biphenyl (PCB) PCB180. a–f 1 μM HSA in the presence of 0, 2, 4, 8, 15 and 25 μM PCB180, respectively; λex=280 nm, T=298 K, pH 7.40. b Plot of Stern-Volmer curve. c Quenching curves of HSA for two- and three-component systems at a molar ratio of phenylbutazone (PHE): HSA 0:1–32:1 and PHE: PCB180 (const): HSA 0:1:1–32:1:1. d Ibuprofen (IB): HSA 0:1–264:1 and PHE:PCB180 (const):HSA 0:1:1–264:1:1, λex=280 nm, T= 298 K
experiments using two probe compounds: phenylbutazone (PHE) and ibuprofen (IB) [36]. PHE generally prefers binding to site I of HSA through hydrophobic interactions, whereas IB binds to site II based on a combination of hydrophobic, hydrogen bonding, and electrostatic interactions [37]. IB and PHE were added to PCB180-HSA systems as competitors [38]. Compared with ternary (competitor-PCB180-HSA) and binary (competitor-HSA) systems, the absence of competitor IB can significantly influence quenching approaches for both binary and ternary systems (Fig. 1c). As shown in Fig. 1d, no significant changes in HSA fluorescence were observed in binary and ternary systems for another competitor, PHE. Further quantitative analysis was conducted with a binding constant (K) to support the conclusion. For competitor PHE, no notable changes were observed for the K values of binary and ternary systems, which ranged from 10.10 to 9.55. The binding constant K for IB varied significantly from 154.32 to 29.95. IB showed more significant displacement of PCB180, Table 1 Binding constant and thermodynamic parameters for the binding of the polychlorinated biphenyl (PCB) PCB180 to human serum albumin (HSA)
suggesting that IB is a stronger competitor for HSA than PHE. The data supported the view that PCB180 shared a common binding site with probe compound IB, not PHE. PCB180 was proposed to bind to one site, site II, on HSA. Thermodynamic analysis Thermodynamic processes were considered to characterize the forces acting between HSA and PCB180. Four types of non-covalent interaction, namely, hydrogen bonds, van der Waals forces, electrostatic forces, and hydrophobic interaction forces, served a function as acting forces between small molecules and macromolecules [34]. As shown in Table 1, quantitative analysis was performed at different temperatures of 298, 303 and 308 K. We then analyzed the thermodynamic parameter entropy change (ΔS0) and free energy change (ΔG0). As we know [34], a positive value of both ΔH0 and ΔS0 always indicates a typical hydrophobic interaction, thus
Temperature (K)
K (× 106M−1)
ΔG0 (KJ/mol−1)
ΔS0 (J/mol−1/K)
ΔH0 (KJ/mol−1)
298 303 308
20.9266 28.966 97.72
−7.16795 −9.24857 −11.3292
416.124
116.837
J Mol Model (2014) 20:2098
suggesting that binding of PCB180 to HSA is stabilized mainly by hydrophobic interactions. Simultaneously, the negative sign for ΔG0 exhibited spontaneity of binding and was generally related to the electrostatic interaction. For the PCB180–HSA system, the main contributor to the ΔG0 value was the large ΔS0 term. The main interaction was through hydrophobic contacts; however, the electrostatic interaction should not be neglected.
Page 5 of 10, 2098
complex were also calculated. The amino acid residues of primary binding site II had more conformational fluctuation than another binding site I (subdomain IIA), which might be induced by the binding behavior with PCB180. Residues Lys402, Pro421, and Lys466 had relatively higher RMSF values than other residues, which were located within the core of binding site II. Subdomain IIIA was also a possible binding site. Our conclusion agreed with the aforementioned observations of fluorescence experiments.
Molecular dynamic simulation Analysis of binding models Molecular modeling by docking and MD analysis was used to investigate more precise interactions between PCB180 and subdomain IIIA (site II) of HSA. The docking mode of the PCB180-HSA complex was first constructed and refined using the MD approach. Evolutions of the root-mean-square displacement (RMSD) value versus time for the whole protein, subdomain IIA (site I), and subdomain IIIA (site II) of HSA were obtained (Fig. 2). The binding location subdomain IIIA (site II), which was proven by fluorescence measurements, exhibited more fluctuation during MD simulations. The subdomain IIA (site I) was quite stable. More conformational changes of HSA were induced by the binding behavior with site II. The root mean square fluctuation (RMSF) values per residue of HSA-PCB180 Fig. 2a–d Results of molecular dynamics (MD) simulation during 5 ns evolution. a Time dependence of root-mean-square displacement (RMSD) for PCB180-HSA complex. b Root mean square fluctuation (RMSF) values of PCB180-HSA complex for each residue. c Time dependence of RMSD for PCB180-site I. d Time dependence of RMSD for PCB180-site II
The average structure of the PCB180-HSA complex within the last 1 ns of the MD trajectories was used as a refined binding model. As shown in Fig. 3a and c, the biphenyl rings of PCB180 are buried in a hydrophobic Sudlow site II. The binding pocket was formed by the packing of six helices. The two phenyls of PCB180 were not coplanar because hydrophobic interactions twisted the planar molecules against the internal steric interaction. The phenyl group was located within the binding pocket that extends to the inner hydrophobic compartment. The phenyl group was formed by several aliphatic residues, including Leu423, Val426, Ser427, Leu430, Val456, Leu457, and Leu460. The phenyl ring A with four chlorine atoms, which was in the more hydrophobic part of PCB180,
2098, Page 6 of 10
J Mol Model (2014) 20:2098
Fig. 3a–d Binding mode between PCB180 and site II of HSA. a 3D view. b 2D view. c Hydrophobic potential plot on molecular surfaces; blue, green, white and red represent decreasing hydrophobic potential. d Electrostatic potentials plot on molecular surfaces; blue, white and red represent decreasing electrostatic potential
was oriented toward the hydrophobic cavity of HSA (Fig. 3c). The other phenyl ring was situated in a larger but less hydrophobic cavity formed by residues of Tyr411, Val415, Arg417, Leu453, Val473, Arg485, Phe488, and Leu491 (Fig. 3a). Our observations agreed with the given experimental results. Hydrophobic interaction, which is a non-bonded interaction, was the main force governing the binding behavior between PCB180 and HSA. In addition to hydrophobic contacts, PCB180 also made specific interactions with nearby polar residues. PCB180
stabilized polar moieties through electrostatic interactions. Several ionic residues (Arg410 and Arg485) and polar residues (Tyr411, Thr412, Ser427, and Ser489) near PCB180, served important functions in stabilizing the PCB180-HSA complex via electrostatic interaction (Fig. 3a,b). Additional π–π conjugations between the two phenyl rings and two amino residues (Phe488 and Tyr411) also appeared (Fig. 3b). The distribution of molecular surface electrostatic potentials was also calculated and showed a good fit with the electro negativity docking pocket (Fig. 3d).
J Mol Model (2014) 20:2098
Page 7 of 10, 2098
Fig. 4 a Electrostatic potential mapped onto the molecular surface of PCB180 and PCB153; blue, white and red represent decreasing electrostatic potential. b Hydrophobic potential mapped onto the molecular surface of PCB180 and PCB153; blue, green, white and red represent decreasing hydrophobic potential
In terms of distance and orientation, we detected the formation of hydrogen bonds by examining the snapshots during the last 1 ns of the equilibrium phase of the simulations. Time evolution was performed to monitor a potential hydrogen bond between chlorine atom and residue Ser489 (Fig. S2, Supplementary materials section). The hydrogen bond occupied only 0.20 % of the last 1 ns, which indicates that the bond was not always stably formed. To probe the structural basis for the interaction with HSA, the binding mode of PCB180 was compared with another PCB congener, PCB153. The two PCBs were all buried in site II. The PCBs also showed a very different orientation, although the two compounds have similar chemical structures, and the difference existed only in the 3-position. Figure 4 presents the electrostatic charge distributions on the two PCBs, which reveals that the shapes and charge distributions of the two PCBs are similar. Compared with PCB153,
PCB180 had one more chlorine atom in the 3-position of the A ring. PCB180 had a more negative electric distribution, particularly around the ring A (Fig. 4a,b). PCB180 significantly affected the binding behavior. The A ring of PCB180 was closer to the binding cavity with positive charge, whereas PCB153 was binding in a location with more negative charge. The hydrophobic surfaces of the two PCBs were also compared. PCB180 showed greater hydrophobicity. Thus, PCB180 was located in a more hydrophobic cavity in the binding site. Key residue identification To rationalize the relative importance of polar and apolar contributions of each residue, energy contribution difference analysis was performed (Fig. 5). The most important residues were found to be Tyr411, Val426, Ser427, Leu430, Leu460,
2098, Page 8 of 10
J Mol Model (2014) 20:2098
Fig. 5 a Energy differences in each residue’s contribution to binding. b Polar and nonpolar contributions for key residues. Red bar Van der Waals and nonpolar solvation energy (ΔEmm,vdw, ΔGsol,npol); black bar electrostatic and polar solvation energy (ΔEmm,ele, ΔGsol,pol). Negative values are favorable and positive values are unfavorable for binding
Phe488, and Ser489. The contributions of these residues to the binding free energy were greater than −10 kcal/mol−1. Strong favorable contributions to binding were associated with residues located in the active site II. Residues Val473, Leu453, Ala449, Gly434, and Cys392 made only weak contributions to the binding, whereas, Leu460, Ser489, Phe488, Val426, Leu430, and Tyr411 made strong contributions (Fig. 5a). The contribution of the seven key residues was further decomposed into polar and non-polar sections (Fig. 5b). Non-polar residues (Val426, Leu430, Leu460, and Leu491) made more contributions to the binding free energy than polar residues (Tyr411, Ser427, and Ser489). Hydrophobic
Table 2 Binding free energy components for the PCB180-HSA complex
Component ΔEmm,ele ΔEmm,vdw ΔGsol,pol ΔGsol,npol ΔGele,tol ΔGnp,tol ΔH TΔS ΔGbind
Energy (kcal/mol) −1.42±0.77 −40.86±1.93 10.87±−0.94 −5.05±0.08 9.44±0.79 −45.92±2.48 −36.47 −34.67 −1.80
interactions (red bar in Fig. 5b) also contributed to binding behavior, whereas electrostatic interactions (black bar in Fig. 5b) make no contribution. Based on the results, hydrophobic interactions can be concluded to play an important role in HSA-PCB180 binding. Free energy contributions Further insight into the forces involved in ligand binding can be obtained by analyzing free energy contributions. Table 2 shows the calculated free energy contributions for each system using the MM-GBSA method [39]. Non-polar contributions (ΔGnp,tol) are equal to the sum of ΔGsol,npol and ΔEmm, vdw, which includes hydrogen bonding force and van der Waals force contributions. Electrostatics contributions (ΔGele,tol) are equal to the sum of ΔEmm,ele and ΔGsol, pol. For PCB180, the negative term of ΔEmm,ele generally indicates that electrostatic interactions between PCB180 and HSA facilitated favorable binding. A positive term indicated that the electrostatic component of the solvation free energies was consistently unfavorable for binding. A negative value of the non-polar component (ΔEmm,vdw + ΔGsol,pol) indicated that the contribution was favorable. Notably, an unfavorable electrostatics component (ΔEmm,ele + ΔGsol,pol) in PCB180-HSA complex was compensated by a highly favorable non-polar component (ΔEmm,vdw + ΔGsol,pol) of the
J Mol Model (2014) 20:2098
free energy. The highly favorable non-polar component originated mostly from the van der Waals interaction energy (ΔEmm,vdw). Numerous studies demonstrated that electrostatic contribution generally disfavors the docking of ligand and receptor molecules, because of mostly unfavorable changes in the electrostatics of solvatation. Based on the quantitative analysis of non-polar (ΔEmm,vdw + ΔGsol,npol) and electrostatic (ΔEmm, ele + ΔGsol, pol) contributions of the PCB180-HSA complex, we considered the interaction to be dominated by more favorable non-polar interactions. Favorable non-polar interactions were common for noncovalent association than for electrostatic interactions. The findings verified the fluorescence experiments from the perspective of energy. The calculated binding energy for PCB180-HSA complex was −1.80 kcal/mol−1 (−7.56 kJ/ mol −1 ), which fitted the experiment observation well (−7.17 kJ/mol−1, 298 K). In conclusion, the MD simulation and docking study were consistent with the experimental analysis. Through the combination of MD simulation, docking, and thermodynamic analysis, hydrophobic interactions were found to be the driving force behind the interaction between PCB180 and HSA, despite the existence of hydrogen bonds and electrostatic interactions.
Conclusions The interaction between PCB180 and HSA was investigated in this work using fluorescence spectroscopic techniques and a docking modeling investigation. The results obtained from the fluorescence spectroscopy suggested that PCB180 can bind to HSA through a quenching procedure, which showed that the binding behavior induced conformational changes. Site marker displacement experiments were also carried out using PHE and IB, which further proved that the binding site was located in the hydrophobic pocket subdomain IIIA (site II). Furthermore, the binding modes were characterized by docking and MD simulation to provide insight into protein complexes on an atomic level. Binding free energy calculations gave us information about the stability and thermodynamics parameter of the PCB180-HSA complex, which showed that hydrophobic interactions dominated the mode of binding behavior. The analysis may contribute to elucidate the mechanisms of PCB180-HSA binding and may throw light on future studies about PCBs-HSA interactions. Acknowledgments This work was supported financially by National Natural Science Foundation of China (NO. 31000017) (NO. 21207056), the Natural Foundation of Gansu Province (1104WCGA187) and Key Laboratory of Chemistry and Quality for Traditional Chinese Medicines of the College of Gansu Province, Gansu College of Traditional Chinese Medicine (Zzy-2011-03).
Page 9 of 10, 2098
References 1. Kavlock RJ, Daston GP, DeRosa C, Fenner-Crisp P, Gray LE, Kaattari S, Lucier G, Luster M, Mac MJ, Maczka C, Miller R, Moore J, Rolland R, Scott G, Sheehan DM, Sinks T, Tilson HA (1996) Research needs for the risk assessment of health and environmental effects of endocrine disruptors: a report of the U.S. EPAsponsored workshop. Environ Health Perspect 104:715–740 2. Wehrmeier A, Lenoir D, Schramm LW, Zimmermann R, Hahn K, Henkelmann B, Kettup A (1998) Patterns of isomers of chlorinated dibenzo-p-dioxins as tool for elucidation of thermal formation mechanisms. Chemosphere 36:2775–2801 3. Giesy JP, Kannan K (1998) Dioxin-like and Non-dioxin-like toxic effects of polychlorinated biphenyls (PCBs): implications for risk assessment. Crit Rev Toxicol 28:511–569 4. Sanchez E, Fernandez SM, Lopez-Aparicio P, Recio MN, PerezAlbarsanz MA (2000) Selective fatty acid release from intracellular phospholipids caused by PCBs in rat renal tubular cell cultures. Chem Biol Interact 125:117–131 5. Howard AS, Fitzpatrick R, Pessah I, Kostyniak P, Lein PJ (2003) Polychlorinated biphenyls induce caspase-dependent cell death in cultured embryonic rat hippocampal but not cortical neurons via activation of the ryanodine receptor. Toxicol Appl Pharmacol 190:72–86 6. Kester MH, Bulduk S, Tibboel D, Meinl W, Glatt H, Falany CN, Coughtrie MW, Bergman A, Safe SH, Kuiper GG, Schuur AG, Brouwer A, Visser TJ (2000) Potent inhibition of estrogen sulfotransferase by hydroxylated PCB metabolites: a novel pathway explaining the estrogenic activity of PCBs. Endocrinology 141:1897–1900 7. Inglefield JR, Mundy WR, Meacham CA, Shafer TJ (2002) Identification of calcium-dependent and -independent signaling pathways involved in polychlorinated biphenyl-induced cyclic AMPresponsive element-binding protein phosphorylation in developing cortical neurons. Neuroscience 115:559–573 8. Zolese G, Falcioni G, Bertoli E, Galeazzi R, Wozniak M, Wypych Z, Gratton E, Ambrosini A (2000) Steady-state and time resolved fluorescence of albumins interacting with N-oleylethanolamine, a component of the endogenous N-acylethanolamines. Proteins 40:39–48 9. Petitpas I, Grnne T, Bhattacharya AA, Curry S (2001) Crystal structures of human serum albumin complexed with monounsaturated and polyunsaturated fatty acids. J Mol Biol 314:955–960 10. Ghuman J, Zunszain PA, Petitpas I, Bhattacharya AA, Otagiri M, Curry S (2005) Structural basis of the drug-binding specificity of human serum albumin. J Mol Biol 353:38–52 11. Curry S, Mandelkow H, Brick P, Franks N (1998) Crystal structure of human serum albumin complexed with monounsaturated and polyunsaturated fatty acid. Nat Struct Biol 5:827–835 12. Fehske KJ, Mullar WE, Wollart U (1981) The location of drug binding sites in human serum albumin. Biochem Pharmacol 30:687–692 13. Khan SN, Islam B, Yennamalli R, Sultan A, Subbarao N, Khan AU (2008) Interaction of mitoxantrone with human serum albumin: Spectroscopic and molecular modeling studies. Eur J Pharm Sci 35: 371–382 14. Deeb O, Rosales-Hernández MC, Gómez-Castro C, Garduño-Juárez R, Correa-Basurto J (2010) Exploration of human serum albumin binding sites by docking and molecular dynamics flexible ligandprotein interactions. Biopolymers 93:161–170 15. Herve F, Urien S, Albengres E, Duche JC, Tillement JP (1994) Drug binding in plasma: a summary of recent trends in the study of drug and hormone binding. Clin Pharmacokinet 26:44–58 16. Houde M, Martin JW, Letcher RJ, Solomon KR, Muir DCG (2006) Biological monitoring of polyfluoroalkyl compounds: a critical review. Environ Sci Technol 40:3463–3473 17. Lejon S, Cramer JF, Nordberg PA (2008) Structural basis for the binding of naproxen to human serum albumin in the presence of fatty acids and the ga module. Acta Crystallogr F 64:64–69
2098, Page 10 of 10 18. Patel S, Datta A (2007) Steady state and time-resolved fluorescence investigation of the specific binding of Two chlorin derivatives with human serum albumin. J Phys Chem B 111:10557–10562 19. Purcell M, Neault JF, Malonga H, Arakawa H, Carpentier R, TajmirRiahi HA (2001) Interactions of atrazine and 2,4-D with human serum albumin studied by gel and capillary electrophoresis and FTIR spectroscopy. Biochim Biophys Acta 1548:129–138 20. Saquib Q, Al-Khedhairy AA, Siddiqui MA, Roy AS, Dasgupta S, Musarrat J (2011) Preferential binding of insecticide phorate with sub-domain IIA of human serum albumin induces protein damage and its toxicological significance. Food Chem Toxicol 49:1787–1795 21. Xie XY, Wang XR, Xu XM, Sun HJ, Chen XG (2010) Investigation of the interaction between endocrine disruptor bisphenol A and human serum albumin. Chemosphere 80:1075–1080 22. Fujiwara S, Amisaki T (2008) Identification of high affinity fatty acid binding sites on human serum albumin by MM-PBSA method. Biophys J 94:95–103 23. Quevedo MA, Ribone SR, Moroni GN, Brinon MC (2008) Binding to human serum albumin of zidovudine (AZT) and novel AZT derivatives. Bioorg Med Chem 16:2779–2790 24. Sudhamalla B, Gokara M, Ahalawat N, Amooru DG, Subramanyam R (2010) Molecular dynamics simulation and binding studies of βsitosterol with human serum albumin and its biological relevance. J Phys Chem B 114:9054–9062 25. Safe SH (1994) Polychlorinated-biphenyls (PCBs)-environmentalimpact, biochemical and toxic responses, and implications for risk assessment. Crit Rev Toxicol 24:87–149 26. Rahman MH, Maruyama T, Okada T, Yamasaki K, Otagiri M (1993) Study of interaction of carprofen and its enantiomers with human serum albumin-I. Stereoselective site-to-site displacement of carprofen by ibuprofen. Biochem Pharmacol 46:1721–1731 27. Petitpas I, Bhattacharya AA, Twine S, East M, Curry S (2001) Crystal structure analysis of warfarin binding to human serum albumin: anatomy of drug site I. J Biol Chem 276:22804–22809 28. Anderson AC, O’Neil RH, Surti TS, Stroud RM (2001) Approaches to solving the rigid receptor problem by identifying a minimal set of flexible residues during ligand docking. Chem Biol 8:445–457 29. Sherman W, Day T, Jacobson MP, Friesner RA, Farid R (2006) Novel procedure for modeling ligand/receptor induced fit effects. J Med Chem 49:534–553
J Mol Model (2014) 20:2098 30. Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, Olson AJ (1998) Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Comput Chem 19:1639–1662 31. Case DA, Darden TA, Cheatham TE, III Simmerling CL, Wang J, Duke RE, Luo R, Walker RC, Zhang W, Merz KM, Roberts B, Wang B, Hayik S, Roitberg A, Seabra G, Kolossvai I, Wong KF, Paesani F, Vanicek J, Liu J, Wu X, Brozell SR, Steinbrecher T, Gohlke H, Cai Q, Ye X, Wang J, Hsieh MJ, Cui G, Roe DR, Mathews DH, Seetin MG, Sagui C, Babin V, Luchko T, Gusarov S, Kovalenko A, Kollman PA (2010) AMBER 11. http://ambermd.org/ 32. Wang J, Wolf RM, Caldwell JW, Kollman PA, Case DA (2004) Development and testing of a general AMBER force field. J Comput Chem 25:1157–1174 33. Kollman PA, Massova I, Reyes C, Kuhn B, Huo SH, Chong L, Lee M, Lee T, Duan Y, Wang W, Donini O, Cieplak P, Srinivasan J, Case DA, Cheatham TE (2000) Calculating structures and free energies of complex molecules: combining molecular mechanics and continuum models. Acc Chem Res 33:889–897 34. Ross PD, Subramanian S (1981) Thermodynamics of protein association reactions: forces contributing to stability. Biochemistry 20: 3096–3102 35. Gao H, Lei L, Liu J, Qin K, Chen X, Hu Z (2004) The study on the interaction between human serum albumin and a new reagent with antitumour activity by spectrophotometric methods. J Photochem Photobiol A 167:213–221 36. Pandey RK, Constantine S, Tsuchida T, Zheng G, Medforth CJ, Aoudia M, Kozyrev AN, Rodgers MAJ, Kato H, Smith KM, Dougherty TJ (1997) Synthesis photophysical properties in vivo photosensitizing efficacy and human serum albumin binding properties of some novel bacteriochlorins. J Med Chem 40:2770–2779 37. Consuelo Jiménez M, Miranda MA, Vayá I (2005) Triplet excited states as chiral reporters for the binding of drugs to transport proteins. J Am Chem Soc 127:10134–10135 38. Sułkowska A, Maciążek-Jurczyk M, Bojko B, Równicka J, Zubik-Skupień I, Temba E, Pentak D, Sułkowski WW (2008) Competitive binding of phenylbutazone and colchicines to serum albumin in multidrug therapy: a spectroscopic study. J Mol Struct 881:97–106 39. Eliard PH, Rousseau GG (1984) Biochem J 18:395
ORIGINAL PAPER
Molecular interaction of PCB180 to human serum albumin: insights from spectroscopic and molecular modelling studies Senbiao Fang & Huanhuan Li & Tao Liu & Hongxia Xuan & Xin Li & Chunyan Zhao
Received: 16 September 2013 / Accepted: 27 November 2013 / Published online: 16 March 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Polychlorinated biphenyls (PCBs) are potentially hazardous to the environment because of their chemical stability and biological toxicity. In this study, we identified the binding mode of a representative PCB180 to human serum albumin (HSA) using fluorescence and molecular dynamics (MD) simulation methods. PCB180 bound exactly at subdomain IIIA of HSA based on the fluorescence study along with site marker displacement experiments. PCB180 also induced conformational changes that were governed mainly by hydrophobic forces. MD studies and free energy calculations also made important contributions to the understanding of the effects of an HSA-PCB180 system on conformational changes. The simulations on binding behavior proved that PCB180 was located only in subdomain IIIA. Hydrophobic interactions dominated the mode of binding behavior. The results obtained using the two methods correlated well with each other. Our findings provide a framework for elucidating the mechanisms of PCB180-HSA binding, and may also help in further research on the transportation, distribution, and toxicity effects of PCBs when introduced into human blood serum.
Keywords Polychlorinated biphenyls . Human serum albumin . Molecular dynamics . Fluorescence study Senbiao Fang and Huanhuan Li contributed equally to this work. Electronic supplementary material The online version of this article (doi:10.1007/s00894-014-2098-7) contains supplementary material, which is available to authorized users. S. Fang : H. Li : T. Liu : H. Xuan : C. Zhao (*) School of Pharmacy, LanzhouUniversity, Lanzhou 730000, China e-mail: [email protected] X. Li College of Food and Bioengineering, Henan University of Science and Technology, Luoyang 471003, China
Introduction Polychlorinated biphenyls (PCBs) are ubiquitous and persistent environmental pollutants that accumulate generally in the food chain [1]. PCBs with six or more chlorines can be found in multiple environmental matrices, such as air, water, sediments, adipose tissue of fish, wildlife, and in humans, particularly in breast tissue, serum, and milk [2]. The toxicological effects of exposure to PCBs include hepatotoxicity, immunotoxicity, reproductive problems, as well as respiratory, mutagenic and carcinogenic effects [3, 4]. As generally considered, PCBs theoretically encompass a group of 209 different congeners with various degrees of chlorination and substitution patterns at available sites, which strongly affect each PCB’s potency and the nature of its toxicity [6]. For now, most PCBs are currently poorly characterized, and toxicity mechanisms are not well defined [5–7]. When PCBs enter the body, their distribution and metabolism are correlated with their binding affinities towards serum albumin in blood. Serum albumin—the most abundant protein in plasma—has important functions in the maintenance of colloid osmotic blood pressure, and in the binding and transport of endogenous and exogenous ligands [8]. Thus, knowledge of the interaction mechanisms between PCBs and plasma proteins is an important factor in understanding the pharmacokinetics and structural features of these ligands, as these mechanisms strongly influence ligand distribution and so determine the biological (toxicity) effect. Human serum albumin (HSA) is a non-glycosylated singlechained polypeptide with a mass of 67 kDa, which organizes to form a heart-shaped protein with an α-helical content of approximately 67 % [9–11]. HSA consists of different numbers and sequences of amino acids, which enables it to adopt compounds in various three-dimensional structures, and possesses unique biological functions. HSA is monomeric but contains three structurally similar α-helical domains: I
2098, Page 2 of 10
(residues 1–195), II (196–383), and III (384–585), each consisting of two subdomains (A and B). The overall structure is stabilized by 17 disulfide bridges [10, 12, 13]. Aromatic and heterocyclic ligands generally bind within two hydrophobic pockets in subdomains IIA and IIIA, which are designated as sites I and II, respectively [12, 13]. The overall structure of HSA is flexible under natural conditions, while regulatory functions, such as changes in ligand-binding site conformation, are possible during the process of ligand binding [14]. The flexibility of HSA could be related directly to its myriad functions and binding properties, which further induce various modes of absorption, distribution, metabolism, excretion, and toxicity for many environmental chemicals such as PCBs [15]. Historically, screening for the binding interactions of HSA were generally thought to have been the first steps in investigating the biological behavior of chemicals. Insights into the detailed binding modes of PCBs to HSA have become fundamentally important from a pharmacological perspective. Considering that HSA is one of the main carriers for pharmaceuticals in the human organism, several experimental studies have been conducted to clarify HSA binding properties [10, 16, 17]. Various spectroscopic methods, such as fluorescence quenching, ultraviolet absorption spectroscopy, and circular dichroism were applied to determine the specific fluorescent properties of HSA; these methods are considered to be sensitive and relatively easy to use with high level of molecular detail [18–21]. Experimental studies have generally focused on physiological conditions, whereas theoretical studies such as docking and molecular dynamic (MD) simulation have long been applied to analyze biological processes rapidly and efficiently [22–24]. These computational methods can provide detailed information on the fluctuations and conformational changes of proteins and ligands. These methods are now used routinely to investigate the structural changes, dynamics and thermodynamics of biological molecules and their complexes, which can aid further understanding of the interaction mechanisms of protein–ligand systems. Particularly for PCBs, 209 PCB congeners are not wholly purified or synthesized, which makes analysis of the interaction between individual congeners and proteins difficult. Theoretical calculation and modeling of PCB–HSA binding behavior could help explain the toxicity of PCBs by facilitating annotation of the physical basis of the structure–function relationship of biomolecules. This study employed a combination of experimental and theoretical research in an attempt to determine where and how PCBs bind to HSA. Considering that non-coplanar PCBs are highly prevalent in the environment and must be included in hazard/risk assessment [25], PCB180 (as shown in Fig. S1, Supplementary materials section) was considered as the representative PCB congener. The site marker displacement experiments based on fluorescence spectroscopy were first undertaken to validate site-selective binding and to determine the
J Mol Model (2014) 20:2098
thermodynamic properties of the HSA-PCB180 system. Based on the experimental results, theoretical calculations were mapped using molecular docking and dynamics simulations to gain insight into the full structural characteristics of the PCB180-HSA complex. The results obtained by the two methods correlated well with each other. We also discussed the structural diversity of PCB180 and another non-coplanar, PCB153, which resulted in different binding modes with HSA. Our findings thus provide a framework for elucidating the mechanisms of PCB180–HSA binding. Future studies on other PCBs–HSA interactions may gain insights from our study.
Experimental methods Materials HSA (fatty acid free), phenylbutazone (PHE) and ibuprofen (IB) were purchased from Sigma (St. Louis, MO). HSA was dissolved in Tris-HCl buffer (0.1 M), and the pH was adjusted to 7.4. PCB180 was dissolved in dimethyl sulfoxide (DMSO). Other chemicals were of analytical reagent grade. Phenylbutazone and ibuprofen were initially dissolved in 0.05 M NaOH (pH 7.40) as stock solutions. The 60 μM HSA stock solutions were stored in the dark at 189 K, while phenylbutazone (500 μM), ibuprofen (2,000 μM) and PCB180 (500 μM) were stored at 277 K. Double distilled water was used for all procedures. Fluorescence titration assays Fluorescence was measured on a RF-5301 (Shimadzu, Tokyo, Japan) fluorescence spectrophotometer at λex of 280 nm with slit widths (em=3 nm and ex=5 nm). The HSA solution was fixed at 1×10−6 M concentration, then titrated by increasing the concentration of PCB180 to give a final concentration of 4×10−7 M–6.8×10−6 M. The fluorescence intensity was recorded under excitation at 280 nm and emission at 354 nm with a time interval of 10 min. In addition, the fluorescence experiments were carried out at three different temperatures of 298 K, 303 K and 308 K to evaluate the effect of temperature. For displacement experiments of PHE and IB, PCB180 (1.0 μM) was incubated with HSA (1.0 μM) in Tris-HCl buffer (0.1 M, pH 7.4) at 298 K for 30 min before displacement. Then, 2.5 mL sample was placed in a 1-cm quartz cuvette, followed by titration of aliquots of PHE (500 μM) or IB (2,000 μM) [26]. Molecular modeling methods A molecular docking study was implemented to explore possible binding positions of PCB180. In the present study, the
J Mol Model (2014) 20:2098
structure with PDB ID: 1H9Z [27] having a resolution of 2.5 Å was chosen for dockings. Water molecules, co-crystallized ligand and ions were removed (including ordered water molecules) before the protein was refined with Swiss PDB viewer. Missing hydrogens and Kollman partial charges were then added. Finally, non-polar hydrogens were merged to their corresponding carbons, and desolvation parameters were assigned to each atom. PCB180 was built with G-view and optimized at the B3LYP/6-31++G** level using Gaussian 98 software (http://www.gaussian.com/). Then rotations and torsions for PCB180 were set automatically in ADT (Autodock Tools, http://autodock.scripps.edu/). Gln390, Lys414, Glu425 and His464 were set as flexible residues. In the present work, the selection of flexible residues was based on the following criteria: (1) all binding site residues whose side chain atoms deviate more than 2.5 Å from the nearest atom in the corresponding amino acid in another crystal structure of the same receptor. (2) All binding site residues with multiple occupancies or missing density. (3) If the unit cell contains multiple independently refined copies of the receptor, all binding site residues whose side chain atoms deviate more than 1.5 Å from the nearest atom in the corresponding amino acid in another copy of the receptor in the unit were selected [28, 29]. Docking simulations were performed with AutoDock4.0 [30] using a Lamarkian genetic algorithm. The large enough grid volumes were considered, including the whole subdomain IIA and IIIA. This choice was determined by the need to maintain a small grid spacing (0.375) for the search box and at the same time to perform the docking calculation without neglecting any portion of the protein surface. Others docking parameters were set as default.
Page 3 of 10, 2098
estimate the long-range electrostatic interactions with a cutoff of 10.0 Å. All bond lengths were constrained using the SHAKE algorithm, and the integration time step was set to 2 f. using the Verlet leapfrog algorithm [32]. To eliminate possible bumps between the solute and the solvent, the entire system was minimized in two steps. Firstly, the complex was restrained with a harmonic potential with a force constant k= 100 kcal mol−1 Å−2. The water molecules and counterions were optimized using the steepest descent method of 2,500 steps, followed by the conjugate gradient method for 2,500 steps. Secondly, the entire system was optimized using the first step method without any constraint. These two minimization steps were followed by annealing simulation with a weak restraint (k=10 kcal mol−1 Å−2) for the complex and the entire system was heated gradually in the NVT ensemble from 0 to 298 K over 50 ps. After the heating phase, a 5 ns MD simulation was performed under 1 atm. The constant temperature was selected at 298 K with the NPT ensemble. Constant temperature was maintained using the Langevin thermostat with a collision frequency of 2 ps−1. Constant pressure was maintained employing isotropic position scaling algorithm with a relaxation time of 2 ps. In the present research, 20 snapshots (every 50 ps) were extracted from the last 1 ns trajectory for each complex. The binding free energy (ΔGbind) was computed for each snapshot and averaged using the MM-GBSA approach implemented as script (MMPBSA.py) in AMBER software [33].
Results and discussion Fluorescence titration assays
Molecular dynamics simulations To check the stability of the PCB180–HSA complex, the best docking result was employed for MD simulation. MD was carried out with AMBER software (version 11) [31], using AMBER ff02 force field for protein and the general AMBER force field (GAFF) for PCB180 (http://ambermd.org/). The optimized geometries of PCB180 were employed to calculate electrostatic potential (ESP)-derived charges using the restrained electrostatic potential (RESP) methodology, as implemented in the ANTECHAMBER model (http://ambermd. org/antechamber/ac.html). Hydrogen atoms were added to the initial PCB180-HSA complex model using the tleap module, setting ionizable residues as their default protonation states at a neutral pH value. The complex was solvated in a cubic periodic box of explicit TIP3P water model, which extended a minimum 10 Å distance from the box surface to any atom of the solute. Sixteen sodium ions replaced water molecules randomly in the box to neutralize a total negative charge of 16e on the entire system. The particle mesh Ewald (PME) method for simulation of periodic boundaries was used to
Fluorescence measurement is a simple method with which to explore the structural changes of HSA influenced by PCB180. First, the fluorescence emission intensity of HSA was tested by titrating various amounts of PCB180 using an excitation wavelength of 280 nm. As shown in Fig. 1a, the fluorescence intensity of HSA decreased gradually, exhibiting a concentration-dependent relationship. Quantitative analysis was then conducted using the Stern-Volmer equation [34] as listed in Table 1. The plots obtained from fluorescence measurements showed a linear relationship with R2 = 0.99 (Fig. 1b), which suggested a single class of binding site on HSA for PCB180 with a unique binding constant [35]. Identification of the binding sites The results suggested the existence of binding behavior between PCBs and HSA induced by quenchers of PCB180. However, the exact binding site of PCB180 was unknown under such conditions, and more evidence was required for identification. Thus, we conducted site marker displacement
2098, Page 4 of 10
J Mol Model (2014) 20:2098
Fig. 1 a Fluorescence spectra of human serum albumin (HSA) with various amounts of the polychlorinated biphenyl (PCB) PCB180. a–f 1 μM HSA in the presence of 0, 2, 4, 8, 15 and 25 μM PCB180, respectively; λex=280 nm, T=298 K, pH 7.40. b Plot of Stern-Volmer curve. c Quenching curves of HSA for two- and three-component systems at a molar ratio of phenylbutazone (PHE): HSA 0:1–32:1 and PHE: PCB180 (const): HSA 0:1:1–32:1:1. d Ibuprofen (IB): HSA 0:1–264:1 and PHE:PCB180 (const):HSA 0:1:1–264:1:1, λex=280 nm, T= 298 K
experiments using two probe compounds: phenylbutazone (PHE) and ibuprofen (IB) [36]. PHE generally prefers binding to site I of HSA through hydrophobic interactions, whereas IB binds to site II based on a combination of hydrophobic, hydrogen bonding, and electrostatic interactions [37]. IB and PHE were added to PCB180-HSA systems as competitors [38]. Compared with ternary (competitor-PCB180-HSA) and binary (competitor-HSA) systems, the absence of competitor IB can significantly influence quenching approaches for both binary and ternary systems (Fig. 1c). As shown in Fig. 1d, no significant changes in HSA fluorescence were observed in binary and ternary systems for another competitor, PHE. Further quantitative analysis was conducted with a binding constant (K) to support the conclusion. For competitor PHE, no notable changes were observed for the K values of binary and ternary systems, which ranged from 10.10 to 9.55. The binding constant K for IB varied significantly from 154.32 to 29.95. IB showed more significant displacement of PCB180, Table 1 Binding constant and thermodynamic parameters for the binding of the polychlorinated biphenyl (PCB) PCB180 to human serum albumin (HSA)
suggesting that IB is a stronger competitor for HSA than PHE. The data supported the view that PCB180 shared a common binding site with probe compound IB, not PHE. PCB180 was proposed to bind to one site, site II, on HSA. Thermodynamic analysis Thermodynamic processes were considered to characterize the forces acting between HSA and PCB180. Four types of non-covalent interaction, namely, hydrogen bonds, van der Waals forces, electrostatic forces, and hydrophobic interaction forces, served a function as acting forces between small molecules and macromolecules [34]. As shown in Table 1, quantitative analysis was performed at different temperatures of 298, 303 and 308 K. We then analyzed the thermodynamic parameter entropy change (ΔS0) and free energy change (ΔG0). As we know [34], a positive value of both ΔH0 and ΔS0 always indicates a typical hydrophobic interaction, thus
Temperature (K)
K (× 106M−1)
ΔG0 (KJ/mol−1)
ΔS0 (J/mol−1/K)
ΔH0 (KJ/mol−1)
298 303 308
20.9266 28.966 97.72
−7.16795 −9.24857 −11.3292
416.124
116.837
J Mol Model (2014) 20:2098
suggesting that binding of PCB180 to HSA is stabilized mainly by hydrophobic interactions. Simultaneously, the negative sign for ΔG0 exhibited spontaneity of binding and was generally related to the electrostatic interaction. For the PCB180–HSA system, the main contributor to the ΔG0 value was the large ΔS0 term. The main interaction was through hydrophobic contacts; however, the electrostatic interaction should not be neglected.
Page 5 of 10, 2098
complex were also calculated. The amino acid residues of primary binding site II had more conformational fluctuation than another binding site I (subdomain IIA), which might be induced by the binding behavior with PCB180. Residues Lys402, Pro421, and Lys466 had relatively higher RMSF values than other residues, which were located within the core of binding site II. Subdomain IIIA was also a possible binding site. Our conclusion agreed with the aforementioned observations of fluorescence experiments.
Molecular dynamic simulation Analysis of binding models Molecular modeling by docking and MD analysis was used to investigate more precise interactions between PCB180 and subdomain IIIA (site II) of HSA. The docking mode of the PCB180-HSA complex was first constructed and refined using the MD approach. Evolutions of the root-mean-square displacement (RMSD) value versus time for the whole protein, subdomain IIA (site I), and subdomain IIIA (site II) of HSA were obtained (Fig. 2). The binding location subdomain IIIA (site II), which was proven by fluorescence measurements, exhibited more fluctuation during MD simulations. The subdomain IIA (site I) was quite stable. More conformational changes of HSA were induced by the binding behavior with site II. The root mean square fluctuation (RMSF) values per residue of HSA-PCB180 Fig. 2a–d Results of molecular dynamics (MD) simulation during 5 ns evolution. a Time dependence of root-mean-square displacement (RMSD) for PCB180-HSA complex. b Root mean square fluctuation (RMSF) values of PCB180-HSA complex for each residue. c Time dependence of RMSD for PCB180-site I. d Time dependence of RMSD for PCB180-site II
The average structure of the PCB180-HSA complex within the last 1 ns of the MD trajectories was used as a refined binding model. As shown in Fig. 3a and c, the biphenyl rings of PCB180 are buried in a hydrophobic Sudlow site II. The binding pocket was formed by the packing of six helices. The two phenyls of PCB180 were not coplanar because hydrophobic interactions twisted the planar molecules against the internal steric interaction. The phenyl group was located within the binding pocket that extends to the inner hydrophobic compartment. The phenyl group was formed by several aliphatic residues, including Leu423, Val426, Ser427, Leu430, Val456, Leu457, and Leu460. The phenyl ring A with four chlorine atoms, which was in the more hydrophobic part of PCB180,
2098, Page 6 of 10
J Mol Model (2014) 20:2098
Fig. 3a–d Binding mode between PCB180 and site II of HSA. a 3D view. b 2D view. c Hydrophobic potential plot on molecular surfaces; blue, green, white and red represent decreasing hydrophobic potential. d Electrostatic potentials plot on molecular surfaces; blue, white and red represent decreasing electrostatic potential
was oriented toward the hydrophobic cavity of HSA (Fig. 3c). The other phenyl ring was situated in a larger but less hydrophobic cavity formed by residues of Tyr411, Val415, Arg417, Leu453, Val473, Arg485, Phe488, and Leu491 (Fig. 3a). Our observations agreed with the given experimental results. Hydrophobic interaction, which is a non-bonded interaction, was the main force governing the binding behavior between PCB180 and HSA. In addition to hydrophobic contacts, PCB180 also made specific interactions with nearby polar residues. PCB180
stabilized polar moieties through electrostatic interactions. Several ionic residues (Arg410 and Arg485) and polar residues (Tyr411, Thr412, Ser427, and Ser489) near PCB180, served important functions in stabilizing the PCB180-HSA complex via electrostatic interaction (Fig. 3a,b). Additional π–π conjugations between the two phenyl rings and two amino residues (Phe488 and Tyr411) also appeared (Fig. 3b). The distribution of molecular surface electrostatic potentials was also calculated and showed a good fit with the electro negativity docking pocket (Fig. 3d).
J Mol Model (2014) 20:2098
Page 7 of 10, 2098
Fig. 4 a Electrostatic potential mapped onto the molecular surface of PCB180 and PCB153; blue, white and red represent decreasing electrostatic potential. b Hydrophobic potential mapped onto the molecular surface of PCB180 and PCB153; blue, green, white and red represent decreasing hydrophobic potential
In terms of distance and orientation, we detected the formation of hydrogen bonds by examining the snapshots during the last 1 ns of the equilibrium phase of the simulations. Time evolution was performed to monitor a potential hydrogen bond between chlorine atom and residue Ser489 (Fig. S2, Supplementary materials section). The hydrogen bond occupied only 0.20 % of the last 1 ns, which indicates that the bond was not always stably formed. To probe the structural basis for the interaction with HSA, the binding mode of PCB180 was compared with another PCB congener, PCB153. The two PCBs were all buried in site II. The PCBs also showed a very different orientation, although the two compounds have similar chemical structures, and the difference existed only in the 3-position. Figure 4 presents the electrostatic charge distributions on the two PCBs, which reveals that the shapes and charge distributions of the two PCBs are similar. Compared with PCB153,
PCB180 had one more chlorine atom in the 3-position of the A ring. PCB180 had a more negative electric distribution, particularly around the ring A (Fig. 4a,b). PCB180 significantly affected the binding behavior. The A ring of PCB180 was closer to the binding cavity with positive charge, whereas PCB153 was binding in a location with more negative charge. The hydrophobic surfaces of the two PCBs were also compared. PCB180 showed greater hydrophobicity. Thus, PCB180 was located in a more hydrophobic cavity in the binding site. Key residue identification To rationalize the relative importance of polar and apolar contributions of each residue, energy contribution difference analysis was performed (Fig. 5). The most important residues were found to be Tyr411, Val426, Ser427, Leu430, Leu460,
2098, Page 8 of 10
J Mol Model (2014) 20:2098
Fig. 5 a Energy differences in each residue’s contribution to binding. b Polar and nonpolar contributions for key residues. Red bar Van der Waals and nonpolar solvation energy (ΔEmm,vdw, ΔGsol,npol); black bar electrostatic and polar solvation energy (ΔEmm,ele, ΔGsol,pol). Negative values are favorable and positive values are unfavorable for binding
Phe488, and Ser489. The contributions of these residues to the binding free energy were greater than −10 kcal/mol−1. Strong favorable contributions to binding were associated with residues located in the active site II. Residues Val473, Leu453, Ala449, Gly434, and Cys392 made only weak contributions to the binding, whereas, Leu460, Ser489, Phe488, Val426, Leu430, and Tyr411 made strong contributions (Fig. 5a). The contribution of the seven key residues was further decomposed into polar and non-polar sections (Fig. 5b). Non-polar residues (Val426, Leu430, Leu460, and Leu491) made more contributions to the binding free energy than polar residues (Tyr411, Ser427, and Ser489). Hydrophobic
Table 2 Binding free energy components for the PCB180-HSA complex
Component ΔEmm,ele ΔEmm,vdw ΔGsol,pol ΔGsol,npol ΔGele,tol ΔGnp,tol ΔH TΔS ΔGbind
Energy (kcal/mol) −1.42±0.77 −40.86±1.93 10.87±−0.94 −5.05±0.08 9.44±0.79 −45.92±2.48 −36.47 −34.67 −1.80
interactions (red bar in Fig. 5b) also contributed to binding behavior, whereas electrostatic interactions (black bar in Fig. 5b) make no contribution. Based on the results, hydrophobic interactions can be concluded to play an important role in HSA-PCB180 binding. Free energy contributions Further insight into the forces involved in ligand binding can be obtained by analyzing free energy contributions. Table 2 shows the calculated free energy contributions for each system using the MM-GBSA method [39]. Non-polar contributions (ΔGnp,tol) are equal to the sum of ΔGsol,npol and ΔEmm, vdw, which includes hydrogen bonding force and van der Waals force contributions. Electrostatics contributions (ΔGele,tol) are equal to the sum of ΔEmm,ele and ΔGsol, pol. For PCB180, the negative term of ΔEmm,ele generally indicates that electrostatic interactions between PCB180 and HSA facilitated favorable binding. A positive term indicated that the electrostatic component of the solvation free energies was consistently unfavorable for binding. A negative value of the non-polar component (ΔEmm,vdw + ΔGsol,pol) indicated that the contribution was favorable. Notably, an unfavorable electrostatics component (ΔEmm,ele + ΔGsol,pol) in PCB180-HSA complex was compensated by a highly favorable non-polar component (ΔEmm,vdw + ΔGsol,pol) of the
J Mol Model (2014) 20:2098
free energy. The highly favorable non-polar component originated mostly from the van der Waals interaction energy (ΔEmm,vdw). Numerous studies demonstrated that electrostatic contribution generally disfavors the docking of ligand and receptor molecules, because of mostly unfavorable changes in the electrostatics of solvatation. Based on the quantitative analysis of non-polar (ΔEmm,vdw + ΔGsol,npol) and electrostatic (ΔEmm, ele + ΔGsol, pol) contributions of the PCB180-HSA complex, we considered the interaction to be dominated by more favorable non-polar interactions. Favorable non-polar interactions were common for noncovalent association than for electrostatic interactions. The findings verified the fluorescence experiments from the perspective of energy. The calculated binding energy for PCB180-HSA complex was −1.80 kcal/mol−1 (−7.56 kJ/ mol −1 ), which fitted the experiment observation well (−7.17 kJ/mol−1, 298 K). In conclusion, the MD simulation and docking study were consistent with the experimental analysis. Through the combination of MD simulation, docking, and thermodynamic analysis, hydrophobic interactions were found to be the driving force behind the interaction between PCB180 and HSA, despite the existence of hydrogen bonds and electrostatic interactions.
Conclusions The interaction between PCB180 and HSA was investigated in this work using fluorescence spectroscopic techniques and a docking modeling investigation. The results obtained from the fluorescence spectroscopy suggested that PCB180 can bind to HSA through a quenching procedure, which showed that the binding behavior induced conformational changes. Site marker displacement experiments were also carried out using PHE and IB, which further proved that the binding site was located in the hydrophobic pocket subdomain IIIA (site II). Furthermore, the binding modes were characterized by docking and MD simulation to provide insight into protein complexes on an atomic level. Binding free energy calculations gave us information about the stability and thermodynamics parameter of the PCB180-HSA complex, which showed that hydrophobic interactions dominated the mode of binding behavior. The analysis may contribute to elucidate the mechanisms of PCB180-HSA binding and may throw light on future studies about PCBs-HSA interactions. Acknowledgments This work was supported financially by National Natural Science Foundation of China (NO. 31000017) (NO. 21207056), the Natural Foundation of Gansu Province (1104WCGA187) and Key Laboratory of Chemistry and Quality for Traditional Chinese Medicines of the College of Gansu Province, Gansu College of Traditional Chinese Medicine (Zzy-2011-03).
Page 9 of 10, 2098
References 1. Kavlock RJ, Daston GP, DeRosa C, Fenner-Crisp P, Gray LE, Kaattari S, Lucier G, Luster M, Mac MJ, Maczka C, Miller R, Moore J, Rolland R, Scott G, Sheehan DM, Sinks T, Tilson HA (1996) Research needs for the risk assessment of health and environmental effects of endocrine disruptors: a report of the U.S. EPAsponsored workshop. Environ Health Perspect 104:715–740 2. Wehrmeier A, Lenoir D, Schramm LW, Zimmermann R, Hahn K, Henkelmann B, Kettup A (1998) Patterns of isomers of chlorinated dibenzo-p-dioxins as tool for elucidation of thermal formation mechanisms. Chemosphere 36:2775–2801 3. Giesy JP, Kannan K (1998) Dioxin-like and Non-dioxin-like toxic effects of polychlorinated biphenyls (PCBs): implications for risk assessment. Crit Rev Toxicol 28:511–569 4. Sanchez E, Fernandez SM, Lopez-Aparicio P, Recio MN, PerezAlbarsanz MA (2000) Selective fatty acid release from intracellular phospholipids caused by PCBs in rat renal tubular cell cultures. Chem Biol Interact 125:117–131 5. Howard AS, Fitzpatrick R, Pessah I, Kostyniak P, Lein PJ (2003) Polychlorinated biphenyls induce caspase-dependent cell death in cultured embryonic rat hippocampal but not cortical neurons via activation of the ryanodine receptor. Toxicol Appl Pharmacol 190:72–86 6. Kester MH, Bulduk S, Tibboel D, Meinl W, Glatt H, Falany CN, Coughtrie MW, Bergman A, Safe SH, Kuiper GG, Schuur AG, Brouwer A, Visser TJ (2000) Potent inhibition of estrogen sulfotransferase by hydroxylated PCB metabolites: a novel pathway explaining the estrogenic activity of PCBs. Endocrinology 141:1897–1900 7. Inglefield JR, Mundy WR, Meacham CA, Shafer TJ (2002) Identification of calcium-dependent and -independent signaling pathways involved in polychlorinated biphenyl-induced cyclic AMPresponsive element-binding protein phosphorylation in developing cortical neurons. Neuroscience 115:559–573 8. Zolese G, Falcioni G, Bertoli E, Galeazzi R, Wozniak M, Wypych Z, Gratton E, Ambrosini A (2000) Steady-state and time resolved fluorescence of albumins interacting with N-oleylethanolamine, a component of the endogenous N-acylethanolamines. Proteins 40:39–48 9. Petitpas I, Grnne T, Bhattacharya AA, Curry S (2001) Crystal structures of human serum albumin complexed with monounsaturated and polyunsaturated fatty acids. J Mol Biol 314:955–960 10. Ghuman J, Zunszain PA, Petitpas I, Bhattacharya AA, Otagiri M, Curry S (2005) Structural basis of the drug-binding specificity of human serum albumin. J Mol Biol 353:38–52 11. Curry S, Mandelkow H, Brick P, Franks N (1998) Crystal structure of human serum albumin complexed with monounsaturated and polyunsaturated fatty acid. Nat Struct Biol 5:827–835 12. Fehske KJ, Mullar WE, Wollart U (1981) The location of drug binding sites in human serum albumin. Biochem Pharmacol 30:687–692 13. Khan SN, Islam B, Yennamalli R, Sultan A, Subbarao N, Khan AU (2008) Interaction of mitoxantrone with human serum albumin: Spectroscopic and molecular modeling studies. Eur J Pharm Sci 35: 371–382 14. Deeb O, Rosales-Hernández MC, Gómez-Castro C, Garduño-Juárez R, Correa-Basurto J (2010) Exploration of human serum albumin binding sites by docking and molecular dynamics flexible ligandprotein interactions. Biopolymers 93:161–170 15. Herve F, Urien S, Albengres E, Duche JC, Tillement JP (1994) Drug binding in plasma: a summary of recent trends in the study of drug and hormone binding. Clin Pharmacokinet 26:44–58 16. Houde M, Martin JW, Letcher RJ, Solomon KR, Muir DCG (2006) Biological monitoring of polyfluoroalkyl compounds: a critical review. Environ Sci Technol 40:3463–3473 17. Lejon S, Cramer JF, Nordberg PA (2008) Structural basis for the binding of naproxen to human serum albumin in the presence of fatty acids and the ga module. Acta Crystallogr F 64:64–69
2098, Page 10 of 10 18. Patel S, Datta A (2007) Steady state and time-resolved fluorescence investigation of the specific binding of Two chlorin derivatives with human serum albumin. J Phys Chem B 111:10557–10562 19. Purcell M, Neault JF, Malonga H, Arakawa H, Carpentier R, TajmirRiahi HA (2001) Interactions of atrazine and 2,4-D with human serum albumin studied by gel and capillary electrophoresis and FTIR spectroscopy. Biochim Biophys Acta 1548:129–138 20. Saquib Q, Al-Khedhairy AA, Siddiqui MA, Roy AS, Dasgupta S, Musarrat J (2011) Preferential binding of insecticide phorate with sub-domain IIA of human serum albumin induces protein damage and its toxicological significance. Food Chem Toxicol 49:1787–1795 21. Xie XY, Wang XR, Xu XM, Sun HJ, Chen XG (2010) Investigation of the interaction between endocrine disruptor bisphenol A and human serum albumin. Chemosphere 80:1075–1080 22. Fujiwara S, Amisaki T (2008) Identification of high affinity fatty acid binding sites on human serum albumin by MM-PBSA method. Biophys J 94:95–103 23. Quevedo MA, Ribone SR, Moroni GN, Brinon MC (2008) Binding to human serum albumin of zidovudine (AZT) and novel AZT derivatives. Bioorg Med Chem 16:2779–2790 24. Sudhamalla B, Gokara M, Ahalawat N, Amooru DG, Subramanyam R (2010) Molecular dynamics simulation and binding studies of βsitosterol with human serum albumin and its biological relevance. J Phys Chem B 114:9054–9062 25. Safe SH (1994) Polychlorinated-biphenyls (PCBs)-environmentalimpact, biochemical and toxic responses, and implications for risk assessment. Crit Rev Toxicol 24:87–149 26. Rahman MH, Maruyama T, Okada T, Yamasaki K, Otagiri M (1993) Study of interaction of carprofen and its enantiomers with human serum albumin-I. Stereoselective site-to-site displacement of carprofen by ibuprofen. Biochem Pharmacol 46:1721–1731 27. Petitpas I, Bhattacharya AA, Twine S, East M, Curry S (2001) Crystal structure analysis of warfarin binding to human serum albumin: anatomy of drug site I. J Biol Chem 276:22804–22809 28. Anderson AC, O’Neil RH, Surti TS, Stroud RM (2001) Approaches to solving the rigid receptor problem by identifying a minimal set of flexible residues during ligand docking. Chem Biol 8:445–457 29. Sherman W, Day T, Jacobson MP, Friesner RA, Farid R (2006) Novel procedure for modeling ligand/receptor induced fit effects. J Med Chem 49:534–553
J Mol Model (2014) 20:2098 30. Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, Olson AJ (1998) Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Comput Chem 19:1639–1662 31. Case DA, Darden TA, Cheatham TE, III Simmerling CL, Wang J, Duke RE, Luo R, Walker RC, Zhang W, Merz KM, Roberts B, Wang B, Hayik S, Roitberg A, Seabra G, Kolossvai I, Wong KF, Paesani F, Vanicek J, Liu J, Wu X, Brozell SR, Steinbrecher T, Gohlke H, Cai Q, Ye X, Wang J, Hsieh MJ, Cui G, Roe DR, Mathews DH, Seetin MG, Sagui C, Babin V, Luchko T, Gusarov S, Kovalenko A, Kollman PA (2010) AMBER 11. http://ambermd.org/ 32. Wang J, Wolf RM, Caldwell JW, Kollman PA, Case DA (2004) Development and testing of a general AMBER force field. J Comput Chem 25:1157–1174 33. Kollman PA, Massova I, Reyes C, Kuhn B, Huo SH, Chong L, Lee M, Lee T, Duan Y, Wang W, Donini O, Cieplak P, Srinivasan J, Case DA, Cheatham TE (2000) Calculating structures and free energies of complex molecules: combining molecular mechanics and continuum models. Acc Chem Res 33:889–897 34. Ross PD, Subramanian S (1981) Thermodynamics of protein association reactions: forces contributing to stability. Biochemistry 20: 3096–3102 35. Gao H, Lei L, Liu J, Qin K, Chen X, Hu Z (2004) The study on the interaction between human serum albumin and a new reagent with antitumour activity by spectrophotometric methods. J Photochem Photobiol A 167:213–221 36. Pandey RK, Constantine S, Tsuchida T, Zheng G, Medforth CJ, Aoudia M, Kozyrev AN, Rodgers MAJ, Kato H, Smith KM, Dougherty TJ (1997) Synthesis photophysical properties in vivo photosensitizing efficacy and human serum albumin binding properties of some novel bacteriochlorins. J Med Chem 40:2770–2779 37. Consuelo Jiménez M, Miranda MA, Vayá I (2005) Triplet excited states as chiral reporters for the binding of drugs to transport proteins. J Am Chem Soc 127:10134–10135 38. Sułkowska A, Maciążek-Jurczyk M, Bojko B, Równicka J, Zubik-Skupień I, Temba E, Pentak D, Sułkowski WW (2008) Competitive binding of phenylbutazone and colchicines to serum albumin in multidrug therapy: a spectroscopic study. J Mol Struct 881:97–106 39. Eliard PH, Rousseau GG (1984) Biochem J 18:395
