Kinetic and Thermodynamic Evaluation of Kynurenic Acid Binding to

Article pubs.acs.org/JPCB

Kinetic and Thermodynamic Evaluation of Kynurenic Acid Binding to GluR1270−300 Polypeptide by Surface Plasmon Resonance Experiments Á dám Juhász,† Edit Csapó,*,†,‡ Ditta Ungor,† Gábor K. Tóth,‡ László Vécsei,§,∥ and Imre Dékány*,†,‡ †

MTA-SZTE Supramolecular and Nanostructured Materials Research Group and ‡Department of Medical Chemistry, Faculty of Medicine, University of Szeged, Dóm tér 8., Szeged H-6720, Hungary § MTA-SZTE Neuroscience Research Group and ∥Department of Neurology, University of Szeged, Semmelweis u. 6, Szeged H-6725, Hungary ABSTRACT: This work clearly demonstrates an evaluation process that is easily performed and is simply based on the fitting of temperature-dependent surface plasmon resonance (SPR) sensorgrams to provide detailed thermodynamic characterization of biologically relevant interactions. The reversible binding of kynurenic acid (KYNA) on human glutamate receptor (GluR1) polypeptide (GluR1270−300)modified gold surface has been studied at various temperatures under physiological conditions by two-dimensional SPR experiments. The registered sensorgrams were fitted by using different kinetic models without application of any commercial software. Assuming that the association of GluR1270−300− KYNA complex is first order in both reactants, the association (ka) and dissociation (kd) constants as well as the equilibrium constants (KA) and the Gibbs free-energy change (ΔG°) were given at 10, 20, 30, and 40 °C. Moreover, the enthalpy (ΔH° = −27.91 kJ mol−1), entropy (ΔS° = −60.33 J mol−1 K−1), and heat capacity changes (ΔCp = −1.28 kJ mol−1 K−1) of the model receptor−ligand system were also calculated using a spreadsheet program. Negative values of ΔG° and ΔH° indicate the exothermic formation of a stable GluR1270−300−KYNA complex, because the |ΔH| > |TΔS| relation suggests an enthalpy-driven binding process. The negative ΔH° and ΔS° values strongly support the formation of a salt bridge between KYNA and the positively charged residues of the polypeptide (Arg, Lys) at pH 7.4, confirmed by molecular docking calculations as well.

■

INTRODUCTION Interactions of proteins with small (drug) molecules play a determinant role in living organisms.1 Detailed kinetic and thermodynamic characterization of receptor−ligand interactions may decisively contribute to modern pharmaceutical developments and would be of great benefit to structure-based drug design.2 A number of different real-time and equilibrium analysis techniques are available that can be used to monitor and quantify protein−ligand association processes. However, radio-labeled assays are reproducible and fast techniques; their major disadvantages are that they are hazardous to human health, produce radioactive waste, and require special laboratory conditions.3 This has led to the development of “label-free” assays based on optical methods.4 The two-dimensional (2D) sensor techniques, especially the SPR, are capable of real-time monitoring of these interactions on a gold sensor surface without the use of labels.5 During SPR measurements, one of the interactants is immobilized from the solution onto a solid/ liquid interface, and a solution of the other interactant is passed over the functionalized gold surface at constant temperature. During this procedure, the refractive index at the interface undergoes a change, this being directly related to the number of biomolecules adsorbed on the surface of the biosensor chip. In © 2016 American Chemical Society

addition to the quantitative analysis, the SPR method simultaneously provides kinetic and thermodynamic characterizations of biomolecular interactions.6 Understanding the interactions between a biological macromolecule and drug molecule requires detailed knowledge of classic physical− chemical parameters.7 The main objective of this work was to provide important data on the interactions between a model peptide fragment of human glutamate receptor (GluR1270−300) and KYNA by using only SPR experiments. Usually, the fitting of SPR sensorgrams is carried out by using several commercial software programs, but in contrast with these evaluation procedures, we provided the kinetic and thermodynamic data of this model receptor−ligand system via a very simple method. The studied system has relevance in neuroscience. Prescott et al. have published a work about the dual action of KYNA on αamino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (also known as AMPA receptor, AMPAR) responses.8 In millimolar concentrations, KYNA is an inhibitor of AMPARs, whereas in nanomolar (or micromolar) concentrations, it Received: June 6, 2016 Revised: July 24, 2016 Published: July 26, 2016 7844

DOI: 10.1021/acs.jpcb.6b05682 J. Phys. Chem. B 2016, 120, 7844−7850

Article

The Journal of Physical Chemistry B

Figure 1. Typical sensorgram (black rough curve) and kinetic-model-fitted association and dissociation phases (gray smooth curve) of a wavelengthmodulated SPR experiment for reversible interaction. Δλ (y-axis) is derived according to the inset.

wavelength-modulated apparatus, the resonance conditions are created by a fixed incident angle while the wavelength of the light source is changed. The light source is an Ocean Optics HL-2000 type tungsten halogen source with 6.8 mW output power (output power coupled into a 600 μm UV/vis fiber and measured with an optical power meter integrated from 200 to 1100 nm) and 360−2400 nm wavelength range. The reflected light intensity is monitored in the 574−1000 nm spectrum range using an IPE AS CR S2010 spectrometer. The spectrometer communicates with a PC via USB connection and data registration is carry out by the SPR UP software. Detailed experimental conditions of the measurements were reported previously.5,10 Briefly, GluR1270−300 was immobilized on the gold surface from an aqueous solution (cpeptide = 0.03 mM; 54.6 μg in 500 μL buffer) via an Au−S covalent bond. In the second case, the reversible binding of KYNA on a polypeptide-functionalized chip was studied at 283.15, 293.15, 303.15, and 313.15 K in the concentration range of 0.5−5.0 mM. The SPR sensorgrams registered at different temperatures were used for holistic kinetic and thermodynamic characterization. Sensorgrams were recorded with the mentioned system using the SPR UP 1.1.11.3 (2014 IPE AS CR) control software, whereas the evaluation of extracted data (Windows text files) was carried out using Microsoft Excel program. Theoretical Background of the Evaluation Procedure. For calculations of the kinetic parameters (ka and kd), the A + B ↔ AB-type bimolecular reaction model was used, where A corresponds to KYNA, B is the polypeptide, and AB is the GluR1270−300−KYNA intermediate complex. Assuming that the association of the GluR1270−300−KYNA complex is first order in both reactants and the dissociation is first order in the intermediate complex gives the following overall rate law

facilitates the AMPAR responses. An understanding of the exact molecular mechanism of the action of KYNA on AMPARs might contribute future drug development for the therapeutic management of different neurological disorders such as Parkinson’s disease, cognitive impairment, or depression.8 Determination of temperature- and concentration-dependent equilibrium surface amount of KYNA on the model peptide fragment via the SPR technique as well as the calculation of isosteric heat of adsorption as a function of the surface coverage (θ) was reported previously.9 Moreover, the binding abilities of KYNA on GluR1270−300-functionalized gold surface were estimated by a molecular docking procedure.9 To supplement those results, in this work, we used SPR sensorgrams to give detailed kinetic (pseudo-first-order kinetic model) parameters of the above-mentioned polypeptide−drug molecule interactions. On the basis of the temperature dependence of KA, the ΔG°, ΔH°, ΔS°, and ΔCp values were also calculated.

■

EXPERIMENTAL SECTION Materials. KYNA (≥98.0%), NaH2PO4·H2O (≥99.0%), and Na2HPO4·12H2O (≥99.0%) were purchased from Sigma, NaCl (≥99.8%) from Molar. GluR1270−300 was synthesized in our laboratory by a solid-phase technique utilizing Boc chemistry.9 The molar mass of this 31-amino acid residue polypeptide (CKNSDARDHTRVDWKRPKYTSALTYDGVKVM) was determined by mass spectroscopy (Mwcalc = 3640.12, Mwmeas = 3639.8). SPR Spectroscopy. A temperature-controlled, two-channel in situ SPR imaging apparatus developed at the Institute of Photonics and Electronics (Prague, Czech Republic) was used. The SPR platform contains two sensing channels with two separated flow chambers; the chamber volume is 0.5 μL. The holder of the gold-coated SPR chip is a microscope glass compatible device with dimensions of 20 × 26 mm2. The SPR chip is a thin gold layer (50 nm thickness), deposited on a glass substrate. The flow system is a typical microfluidic flow system, which connects to the sensor surface with the stock solution tanks. The sample solution is moved by a peristaltic pump over the sensor surface through the flow channels. In the case of our

d[AB] = ka[A][B] − kd[AB] dt

(1)

Considering the specification of the applied SPR technique, the concentration of the KYNA solution is held constant, and the total concentration of the surface-bound polypeptide equals the sum of the free and intermediate complex form ([B]0 = [B] + 7845


Article

The Journal of Physical Chemistry B [AB]). In this way, the concentration of KYNA in eq 1 can be replaced with the concentration of the KYNA stock solution, whereas the concentration of the GluR1270−300−KYNA complex is replaceable with the difference of the total surface concentration of bound polypeptide and free form of polypeptide ([AB] = [B]0 − [B]). On the basis of the abovementioned considerations, the rearranged rate law d[AB] = (ka[A]0 + kd)[B] − kd[B]0 dt

χ2 =

(2)

(3)

As shown in Figure 1 (inset), the wavelength-modulated SPR signal is recorded as the difference between the reflectance minimum of the sample and the solvent (Δλ). A reasonable assumption is that the wavelength shift (Δλ) is proportional to the surface amount of bound analyte (concentration of GluR1270−300−KYNA complex), Δλt = α[AB].11,12 As the maximum concentration of the complex is [AB]max = [B]0, we can determine the maximum value of the wavelength shift, Δλmax = α[AB] = α[B]0, so the integrated form of the rate equation (eq 3) can be rewritten as Δλt = Δλmax (1 − e−kobst )

(4)

During the dissociation phase, the concentration of KYNA in the streaming solution is nearly zero. In this way, integrating the first-order exponential decay of the surface complex gives Δλt = Δλmax e−kdt

(5)

To provide kinetic information of the sensorgrams, some common methods, such as a nonlinear least-squares fit, can be used.13 The determination of ka and kd via kobs is a simple linear regression problem, when kobs is plotted as a function of analyte (KYNA solution) concentration; ka can be obtained from the slope of the curve, whereas the intercept provides kd. However, this calculation method transforms the errors of the primary data, and in general, the uncertainty of the intercept is much higher than that for the precision of the slope of the plotted data pairs. For this reason, it is advantageous to use a global fit for each recorded sensorgram, according to the undermentioned integrated form of the overall rate law.14 Δλt = Δλmax

ka[A] (1 − e−(ka[A] + kd)t ) ka[A] + kd

(Δλexp − Δλcalc)2 Δλcalc

(7)

where Δλcalc is the wavelength shift (proportional to the concentration of bound KYNA), obtained by calculating from the model (nm) and Δλexp is the experimental data of wavelength shift (nm). If data from the model are similar to the experimental data, χ2 will be a small number and if they are different, χ2 will be a bigger value. The interactions between macromolecules and their ligands can be characterized by the free energy, enthalpy, and entropy changes associated with the binding reaction. In most cases, different calorimetric techniques were used to determine the thermodynamic parameters of biomolecular interactions. In contrast to the classic calorimetric experiments, SPR requires an extremely small amount of the studied protein or ligand. Thermodynamic considerations of binding type interaction are necessary to conclude spontaneity of the process via the sign of ΔG°, which is the fundamental criterion of spontaneity. Reactions occur spontaneously at a given temperature if ΔG° has a negative value. A knowledge of the temperature-dependent evolution of sensorgrams enables the calculation of ka and kd using the integrated form of rate equations. Assuming that the A + B ↔ AB-type bimolecular reaction model is suitable for the formation of the GluR1270−300−KYNA complex, KA of the binding reaction could be calculated as a quotient of ka and kd (ka/kd = KA). The well-known ΔG° = −RT ln KA relationship between ΔG° and KA via the ΔG° = ΔH° − TΔS° expression of Gibbs free energy consists of a ΔH° component released or taken up during the binding process and a ΔS° component related to the change in the degree of ‘‘disorder’’ of the system due to linking. In these fundamental thermodynamic equations, R represents the universal gas constant (8.314 J mol−1 K−1) and T is the absolute temperature in K. The temperature dependence of enthalpy and entropy can be described in terms of Cp at a given temperature and constant pressure, if the heat capacity is assumed to be constant.17,18

Additionally, when the formation of the intermediate complex is favorable, the second term of eq 2 is negligible, and the invocation of an observed rate constant (kobs) eventuates a compact form of the rate law, which describes the first-order growth of the products. d[AB] = (ka[A]0 + kd)[B] = kobs[B] dt

∑

ΔH(T ) = ΔH °(T °) + ΔCp(T − T °)

(8)

ΔS(T ) = ΔS°(T °) + ΔCp ln( T T ° )

(9)

Through the temperature dependence of KA, the van’t Hoff equation provides a convenient way to obtain experimental values for enthalpy and enthalpy change for reactions in which these terms are independent of temperature. On the basis of the assumption that the entropy and enthalpy are constant with temperature changes, it describes a linear relationship between the temperature and KA. On the other hand, if the enthalpy and entropy of the biomolecular interactions change with temperature, the contribution of the heat capacity term should be calculated.19 The knowledge of the temperature dependence of KA over a wide temperature range creates an opportunity to estimate the enthalpy, entropy, and heat capacity changes of the binding process according to the integrated form of the van’t Hoff relation.20

(6)

According to the integrated rate law (eq 6), ka and kd can also be determined as a nonlinear least-squares fit problem.13 To define the best fitting of each kinetic curve, the nonlinear chisquare (χ2) error analysis method was used.15 The nonlinear chi-square test is a statistical tool necessary for the best fit of a kinetic investigation,16 obtained by judging the sum of the squared differences between the experimental and calculated data, with each squared difference divided by its corresponding value (calculated from the models). In the case of the SPRbased kinetics investigation, the equivalent mathematical statement is

ln KA =

7846

ΔCp −ΔH °(T °) ΔS°(T °) + + RT R R ⎡⎛ T − T ° ⎞ ⎤ T ⎟ − ln ⎢⎜⎝ ⎥ ⎣ T ⎠ T° ⎦

(10)


Article

The Journal of Physical Chemistry B This relationship can be used to fit the measured and estimated values of KA as a function of the reciprocal temperature, resulting in the values of enthalpy, entropy, and heat capacity change as parameters of a nonlinear parameter estimation method.21 In addition to the calculated values of enthalpy, entropy, and heat capacity by nonlinear parameter estimation, the uncertainties of these thermodynamic data are also important parameters. Small uncertainties point to the best fit of the experimental data (fitted by the integrated form of the van’t Hoff relation, eq 10 in this article).22 To calculate the standard deviations of the mentioned parameters, the weighted resampling “jackknife” procedure was used.23 Determination of enthalpy, entropy, and heat capacity values was carried out by plotting the natural logarithm of equilibrium constants against the reciprocal temperature and the experimental data pairs were fitted by nonlinear parameter estimation. After determining the enthalpy, entropy, and heat capacity values, the standard deviation of these parameters were calculated by the jackknife procedure according to the following. Nonlinear parameter estimation-based fitting of the experimental data pairs (four different ln KA vs 1/T data pairs) was repeated four times using different starting conditions. In the first case, all of the known data pairs were used for calculation, resulting in the mean value of the mentioned parameters. For the second run, the first data pair was neglected, whereas for the third case, the second data pair was neglected, and so on. After these procedures, the standard deviation of the pending parameters was calculable on the basis of the recognized four different parameter sets.

kobs of A + B ↔ AB-type bimolecular reaction was determined using a nonlinear regression method using solver add-in with a Microsoft spreadsheet (Microsoft Excel) based on eq 4. To determine the value of kobs, we minimized the sum of the square of differences between the measured and predicted (calculated by eq 4) wavelength shift (Δλ) using the solver addin function. The kobs values were plotted as a function of KYNA concentrations at 283.15, 293.15, 303.15, and 313.15 K (Figure 3), and the ka and kd values were determined according to the

RESULTS AND DISCUSSION The temperature-controlled two-channel SPR sensor platform was used to investigate the binding kinetics between GluR1270−300 and KYNA as well as to determine the equilibrium surface amount of bound KYNA at different temperatures. Experiments were performed using GluR1270−300-functionalized gold sensor chips. The typical association and dissociation SPR curves for KYNA interacting with peptide at 283.15 K are shown in Figure 2. The experimental curves registered at 283.15, 293.15, 303.15, and 313.15 K were fitted by using the integrated form of the appropriate rate equation (eqs 4−6). The fitted curves are in gray color (see Figure 2).

following equation: kobs = ka[A] + kd. The slope of the curves (presented in Figure 3) corresponds to ka, whereas the intersects provide the appropriate kd values. Global kinetic analyses of the association phases of SPR signals have also been performed using the nonlinear regression method by fitting several curves (see Figure 4, fitted curves

Figure 3. Different kobs values as a function of KYNA concentrations at 283.15 (●), 293.15 (▲), 303.15 (⧫), and 313.15 (■) K (according to the observed rate constant-based kinetic model).

■

Figure 4. Registered (black) and fitted (gray) SPR sensorgrams at 283.15 K. The concentrations of KYNA are 0.5, 1.0, 1.5, 2.0, 3.0, 4.0, and 5.0 mM. The kinetic analyses were performed by fitting the experimental data with global analyses (eq 6).

signed with gray color) from different KYNA concentrations simultaneously, using A + B ↔ AB-type bimolecular reaction as a binding model. To determine the value of ka and kd, we minimized the sum of the square of differences between the measured and predicted (calculated by eq 6) wavelength shift (Δλ) using the mentioned solver add-in function.

Figure 2. Registered (black) and fitted (gray) SPR sensorgrams at 283.15 K. The concentrations of KYNA are 0.5, 1.0, 1.5, 2.0, 3.0, 4.0, and 5.0 mM. The kinetic analyses were performed by fitting the experimental data with the observed-rate-constant-based kinetic model (eq 4). 7847


Article

The Journal of Physical Chemistry B

Table 1. ka, kd, and χ2 Values of GluR1270−300−KYNA Interaction by Fitting the Sensorgrams with Discrete Fit (A) and Global Fit (B) Kinetic Methods discrete fit kinetic method (A) (eq 4) T (K) 283.15 293.15 303.15 313.15

ka (M−1 s−1) 1.778 1.360 1.210 1.178

± ± ± ±

0.230 0.248 0.151 0.082

global fit kinetic method (B) (eq 6)

kd (s−1)

χ2

± ± ± ±

0.932 0.322 0.241 0.594

0.0162 0.0211 0.0234 0.0267

0.0007 0.0007 0.0004 0.0002

ka (M−1 s−1) 5.591 4.587 4.116 3.845

± ± ± ±

0.625 0.598 0.518 0.575

kd (s−1)

χ2

± ± ± ±

2.479 2.846 1.745 2.770

0.0082 0.0102 0.0134 0.0148

0.0009 0.0013 0.0017 0.0022

Table 2. KA and ΔG°, ΔH°, ΔS°, ΔCp Values of GluR1270−300−KYNA Interaction According to Discrete Fit (A) and Global Kinetics Evaluation (B) Methods T (K) 283.15 293.15 303.15 313.15 283.15 293.15 303.15 313.15

KA (M−1) 9 7 7 3

−11.06 −10.16 −9.95 −9.86

± ± ± ±

679 ± 107 452 ± 83 307 ± 55 259 ± 50

−15.35 −14.90 −14.43 −14.47

± ± ± ±

110 65 52 44

± ± ± ±

ΔG° (kJ mol−1)

ΔH° (kJ mol−1)

Discrete Fit Kinetic Method (A) (eq 4) 0.19 −27.91 ± 5.27 0.25 0.32 0.18 Global Fit Kinetic Method (B) (eq 6) 0.37 −27.36 ± 4.97 0.45 0.45 0.55

ΔS° (kJ mol−1 K−1)

ΔCp (kJ mol−1 K−1)

−0.06 ± 0.02

−1.28 ± 0.54

−0.04 ± 0.02

−0.69 ± 0.51

Figure 5. ln KA values as a function of 1/T for observed rate constant (A) and global kinetics evaluation (B). The gray dashed line represents the linear fitting of the experimental data, whereas the black dashed line corresponds to the nonlinear fitting of the experimental data calculated by eq 10.

KA vs 1/T) (presented in Figure 5). In many cases, this plot results in a straight line, where the slope and intercept give the inverse of both enthalpy (−ΔH°/R) and entropy (ΔS°/R) changes, respectively. For a spontaneous and thermodynamically favorable exothermic reaction, the slope is positive and an increase in the temperature results in a decrease in KA. In the course of our GluR1270−300−KYNA interaction, the abovementioned slope was positive (see Figure 5) and as Table 2 shows, it was found that the increase in the temperature resulted in the decrease in KA. Figure 5 also shows that linear (gray dashed line) and nonlinear (black dashed line) regression methods were also used to fit the experimental data. It was found that instead of the widely applied linear regression, in our case, the previously described integrated van’t Hoff relation (eq 10)-based nonlinear regression method provides the best correlation with the experimental data. Namely, in the case of the linear regression, the calculated coefficients were 0.9353 (for (A)) and 0.9793 (for (B)) while for nonlinear regression the coefficients were 0.9945 (for (A)) and 0.9959 (for (B)). In addition to R squared (R2), which indicates the goodness of the fitting technique, the presence of the heat capacity term also supports the application

On the basis of the results of the two applied kinetic methods, the determined ka and kd as well as the values of nonlinear chi-square (χ2) error analysis are listed in Table 1. Compared to the results provided by the two models (kobsbased calculation and global kinetic analysis model) we can conclude that the first evaluation method yields the best fit of the experimental results (according to lower chi-square values). Although the second model does not contain the generally large uncertainty of the intercept of the linear fitting method based on the higher chi-square values, it provides a more unpunctual fit of the experimental curves. Despite the mentioned variance of the chi-square (χ2) error analysis, the standard deviations of the determined ka and kd values do not show the mentioned correlation with the goodness of the nonlinear fit process. First, to determine the thermodynamic parameters of the examined binding process, the equilibrium constant values were calculated as a ratio of the rate constants (KA = ka/kd) at each temperature by the kobs-based kinetics evaluation (A) and by the global kinetics evaluation (B) as well. The KA values are summarized in Table 2. In the second step, according to the van’t Hoff analysis, the ln KA values were plotted against the reciprocal temperature (ln 7848


Article

The Journal of Physical Chemistry B of (eq 10)-based nonlinear regression fitting. As ΔC p determines how the enthalpy and entropy of the reaction changes with temperature, ΔH° and ΔS° will be independent of temperature if and only if ΔCp is equal to 0, which is almost never observed in biological systems. Because the R-squared values (R2 = 0.9945 and 0.9958) and the residues of the sum of the squared values are quite similar (RSS = 0.0028 and 0.0023) for both kinetic evaluation methods, the standard deviations of KA were used to better compare the results of discrete- and global fit-based kinetic evolution. In the case of the discrete fitting method, the mean of the standard deviations was nearly 9%, whereas for global fitting, this value reaches 18%. Considering the significant difference between the mean values of the standard deviations, the thermodynamic parameters provided by the observed rate constant kinetics-based evaluation were taken as a basis for interpretation of the GluR1270−300 polypeptide−KYNA interaction. Generally, the interactions between small molecules and macromolecules are based on binding forces such as van der Waals interactions, hydrophobic forces, electrostatic interactions, and hydrogen bonds. The thermodynamic parameters such as the enthalpy, entropy, and free-energy change of the binding process are the primal evidence for confirmation of the way of linking.24 Positive ΔH° and ΔS° values are frequently taken as evidence for typical hydrophobic interactions, whereas negative ΔH° and ΔS° values arise from van der Waals forces and hydrogen bond formation.25 In accordance with quantum chemical methods26 and nuclear magnetic resonance spectroscopy studies,27 strong salt-bridge interactions between the side chains of amino acids play an important role in protein−protein interactions. Salt bridges correspond to a hydrogen-bonded form, as well, but the partners are formally ionic. When the side chains of basic arginine or lysine in a protein adopt their neutral forms, a hydrogen bond becomes favorable, but an alternative interaction is also possible when this complex adopts a socalled “salt-bridge” arrangement.28 On the basis of the negative ΔH° and ΔS° values, the formation of a salt bridge between the positively charged amino or guanidino group of lysine or arginine and the negatively charged carboxylate group of KYNA is feasible. Moreover, the sign and absolute value of heat capacity might also be an excellent way to collect information on the examined system. Conventionally, large ΔCp effects are associated with hydrophobic interactions, whereas the magnitude of the effect can be correlated with changes in solvent-accessible surface areas in protein folding and protein− ligand interactions.29 A positive ΔCp corresponds to the existence of electrostatic interactions, whereas a negative value denotes a hydrophobic contribution to the binding process.30 In our case, ΔCp = −1.28 ± 0.54 and −0.69 ± 0.51 kJ mol−1 K−1 were obtained by using A and B calculation models, respectively (data in Table 2). Similar values were published previously for proteins and protein−drug interactions. Namely, the experimental (e.g., scanning microcalorimetry) and theoretical estimations of the research groups of Makhatadze31 and Bakk32 support that the magnitude of negative ΔCp upon aqueous unfolding of the polar interior part for three different (RNAse A, ubiquitin, and eglin c) proteins is between −4.9 and −1.5 kJ mol−1 K−1. Moreover, Dullweber et al. found that the isotherm titration calorimetric-based thermodynamic characterization of the drug-binding reaction of low-molecular-weight thrombin and trypsin inhibitors results in similar magnitude (−0.91 to −2.79 kJ mol−1 K−1) data.33 Taking into account the above-listed experiences and analogies,

it might be suggested that besides the salt-bridge interaction, there is a hydrophobic contribution of the examined binding process as well. This above-mentioned feasible binding of KYNA on the GluR1270−300 polypeptide-modified surface suggested by SPR experiments was previously assumed by only a molecular dynamics (MD) computer simulation method,9 in which both the formation of a salt bridge and the situation of benzene ring of KYNA in a nearby apolar pocket (a hydrophobic contribution of the interaction) were estimated.

■

CONCLUSIONS The main objective of this work was to demonstrate our improved evaluation methods to provide valuable kinetic and thermodynamic parameters of a model receptor−ligand system by using only the SPR optical technique. The values of ΔH°, ΔS°, and ΔCp of the interaction between the immobilized GluR1270−300 and KYNA were calculated to be −27.91 ± 5.27 kJ mol−1, −60.33 ± 17.95 J mol−1 K−1, and −1.28 ± 0.54 kJ mol−1 K−1, respectively. A negative value of ΔG° indicated that the binding reaction was thermodynamically favorable, and a stable GluR1270−300−KYNA complex could be formed. A negative value of ΔH° suggested that the binding reaction was exothermic; therefore, high temperatures would not favor the binding reaction. Negative ΔS° of the process indicated a decrease in the disorder of the system, whereas KYNA occupied the binding sites of the immobilized peptide fragment. Since |ΔH| > |TΔS|, the binding reaction is driven mainly by ΔH° of the reaction. Negative ΔH° and ΔS° are taken as evidence for hydrogen bonding and electrostatic interactions. Namely, the KYNA binds to a positively charged arginine or lysine residue of the peptide forming a salt bridge. Previously published molecular docking calculations also support the abovementioned interactions.9 Beyond this evidence, the existence of a negative ΔCp value denotes a hydrophobic contribution to the binding process as well.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (E.C.). *E-mail: [email protected]. Tel: +36(62)544-210. Fax: +36(62)544-042 (I.D.). Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This research was supported by the Hungarian Research Fund OTKA K 116323. REFERENCES

(1) Chautard, E.; Thierry-Mieg, N.; Ricard-Blum, S. Interaction Networks: From Protein Functions to Drug Discovery. A Review. Pathol. Biol. 2009, 57, 324−333. (2) Kuntz, I. D. Structure-Based Strategies for Drug Design and Discovery. Science 1992, 257, 1078−1082. (3) Lhoest, J. B.; Detrait, E.; van den Bosch de Aguilar, P.; Bertrand, P. Fibronectin Adsorption, Conformation, and Orientation on Polystyrene Substrates Studied by Radiolabeling, XPS, and ToF SIMS. J. Biomed. Mater. Res. 1998, 41, 95−103. (4) Cooper, M. A. Label-Free Screening of Bio-Molecular Interactions. Anal. Bioanal. Chem. 2003, 377, 834−842. (5) Homola, J. Present and Future of Surface Plasmon Resonance Biosensors. Anal. Bioanal. Chem. 2003, 377, 528−539.

7849


Article

The Journal of Physical Chemistry B (6) Casasnovas, J. M.; Springer, T. A. Kinetics and Thermodynamics of Virus Binding to Receptor. J. Biol. Chem. 1995, 13216−13224. (7) Manfredini, S.; Pavan, B.; Vertuani, S.; Scaglianti, M.; Compagnone, D.; Biondi, C.; Scatturin, A.; Tanganelli, S.; Ferraro, L.; Prasad, P.; et al. Design, Synthesis and Activity of Ascorbic Acid Prodrugs of Nipecotic, Kynurenic and Diclophenamic Acids, Liable to Increase Neurotropic Activity. J. Med. Chem. 2002, 45, 559−562. (8) Prescott, C.; Weeks, A. M.; Staley, K. J.; Partin, K. M. Kynurenic Acid Has a Dual Action on AMPA Receptor Responses. Neurosci. Lett. 2006, 402, 108−112. (9) Csapó, E.; Bogár, F.; Juhász, Á .; Sebő k, D.; Szolomájer, J.; Tóth, G. K.; Majláth, Z.; Vécsei, L.; Dékány, I. Determination of Binding Capacity and Adsorption Enthalpy between Human Glutamate Receptor (GluR1) Peptide Fragments and Kynurenic Acid by Surface Plasmon Resonance Experiments. Part 2: Interaction of GluR1270-300 with KYNA. Colloids Surf., B 2015, 133, 66−72. (10) Csapó, E.; Majláth, Z.; Juhász, Á .; Roósz, B.; Hetényi, A.; Tóth, G. K.; Tajti, J.; Vécsei, L.; Dékány, I. Determination of Binding Capacity and Adsorption Enthalpy between Human Glutamate Receptor (GluR1) Peptide Fragments and Kynurenic Acid by Surface Plasmon Resonance Experiments. Colloids Surf., B 2014, 123, 924− 929. (11) Karlsson, R.; Michaelsson, A.; Mattsson, L. Kinetic Analysis of Monoclonal Antibody-Antigen Interactions with a New Biosensor Based Analytical System. J. Immunol. Methods 1991, 145, 229−240. (12) Green, R. J.; Frazier, R. A.; Shakesheff, K. M.; Davies, M. C.; Roberts, C. J.; Tendler, S. J. Surface Plasmon Resonance Analysis of Dynamic Biological Interactions with Biomaterials. Biomaterials 2000, 21, 1823−1835. (13) Cai, R.; Liu, Z.; Ma, W.; Wu, X. Kinetic Analysis of Consecutive Reactions Using a Non-Linear Least-Squares Error-Compensation Algorithm. Analyst 1999, 124, 751−754. (14) O’Shannessy, D. J.; Brigham-Burke, M.; Soneson, K. K.; Hensley, P.; Brooks, I. Determination of Rate and Equilibrium Binding Constants for Macromolecular Interactions Using Surface Plasmon Resonance: Use of Nonlinear Least Squares Analysis Methods. Anal. Biochem. 1993, 212, 457−468. (15) Can, M. Studies of the Kinetics for Rhodium Adsorption onto Gallic Acid Derived Polymer: The Application of Nonlinear Regression Analysis. Acta Phys. Pol., A 2015, 127, 1308−1310. (16) Shah, N. B.; Duncan, T. M. Bio-Layer Interferometry for Measuring Kinetics of Protein-Protein Interactions and Allosteric Ligand Effects. J. Visualized Exp. 2014, 84, No. e51383. (17) Sebő k, D.; Csapó, E.; Preočanin, T.; Bohus, G.; Kallay, N.; Dékány, I. Adsorption of Ibuprofen and Dopamine on Functionalized Gold Using Surface Plasmon Resonance Spectroscopy at Solid-Liquid Interface. Croat. Chem. Acta 2013, 86, 287−295. (18) Cooper, A. Thermodynamic Analysis of Biomolecular Interactions. Curr. Opin. Chem. Biol. 1999, 3, 557−563. (19) Keserü, G. M.; Swinney, D. C. Thermodynamics and Kinetics of Drug Binding; John Wiley & Sons, 2015; Vol. 65. (20) Holdgate, G. A.; Ward, W. H. J. Measurements of Binding Thermodynamics in Drug Discovery. Drug Discovery Today 2005, 10, 1543−1550. (21) de Mol, N. J.; Dekker, F. J.; Broutin, I.; Fischer, M. J. E.; Liskamp, R. M. J. Surface Plasmon Resonance Thermodynamic and Kinetic Analysis as a Strategic Tool in Drug Design. Distinct Ways for Phosphopeptides to Plug into Src- and Grb2 SH2 Domains. J. Med. Chem. 2005, 48, 753−763. (22) Harris, D. C. Non-Linear Curve Fitting Using Microsoft Excel Solver. J. Chem. Educ. 1998, 75, 119−121. (23) Caceci, M. Estimating Error Limits in Parametric Curve Fitting. Anal. Chem. 1989, 61, 2324−2327. (24) Ross, P. D.; Subramanian, S. Thermodynamics of Protein Association Reactions: Forces Contributing to Stability. Biochemistry 1981, 20, 3096−3102. (25) Tian, Z. Y.; Song, L. N.; Zhao, Y.; Zang, F. L.; Zhao, Z. H.; Chen, N. H.; Xu, X. J.; Wang, C. J. Spectroscopic Study on the

Interaction between Naphthalimide-Polyamine Conjugates and Bovine Serum Albumin (BSA). Molecules 2015, 20, 16491. (26) Xie, N.-Z.; Du, Q.-S.; Li, J.-X.; Huang, R.-B. Exploring Strong Interactions in Proteins with Quantum Chemistry and Examples of Their Applications in Drug Design. PLoS One 2015, 10, No. e0137113. (27) Sahal, D.; Balaram, P. Peptide Models of Electrostatic Interactions in Proteins: NMR Studies on Two Beta-Turn Tetrapeptides Containing Asp-His and Asp-Lys Salt Bridges. Biochemistry 1986, 25, 6004−6013. (28) Nagy, P. I.; Erhardt, P. W. Theoretical Studies of Salt-Bridge Formation by Amino Acid Side Chains in Low and Medium Polarity Environments. J. Phys. Chem. B 2010, 114, 16436−16442. (29) Cooper, A.; Johnson, C. M.; Lakey, J. H.; Nollmann, M. Heat Does Not Come in Different Colours: Entropy-Enthalpy Compensation, Free Energy Windows, Quantum Confinement, Pressure Perturbation Calorimetry, Solvation and the Multiple Causes of Heat Capacity Effects in Biomolecular Interactions. Biophys. Chem. 2001, 93, 215−230. (30) Aberkane, L.; Jasniewski, J.; Gaiani, C.; Scher, J.; Sanchez, C. Thermodynamic Characterization of Acacia Gum-Beta-Lactoglobulin Complex Coacervation. Langmuir 2010, 26, 12523−12533. (31) Makhatadze, G. I.; Privalov, P. L. Energetics of Protein Structure. Adv. Protein Chem. 1995, 47, 307−425. (32) Bakk, A.; Hansen, A.; Hoye, J. S. Apolar and Polar Solvation Thermodynamics Related to the Protein Unfolding Process. Biophys. J. 2002, 82, 713−719. (33) Dullweber, F.; Stubbs, M. T.; Musil, D.; Stürzebecher, J.; Klebe, G. Factorising Ligand Affinity: A Combined Thermodynamic and Crystallographic Study of Trypsin and Thrombin Inhibition. J. Mol. Biol. 2001, 313, 593−614.

7850


Kinetic and Thermodynamic Evaluation of Kynurenic Acid Binding to

Recommend Documents